Us Japan Hybrid14

Seismic Analysis, Behavior,
and Design of Unbonded

Post-Tensioned Hybrid
Coupled Wall Structures
December 2006
Qiang Shen, Yahya C. Kurama, and Brad D. Weldon
Report #NDSE-06-02
rigid elements for PT anchorages

3-D brick elements
gap/contact
surfaces
rigid elements for

PT anchorages
truss element
for PT tendon
wall contact region
coupling shear force, Vb (kN)
300
initial stiffness
of subassemblage
with "fixed" beamto-wall connections
angle
fracture
Test 3
-300
-10
0
beam chord rotation, b (%)
10
Structural Engineering Research Report

Department of Civil Engineering and Geological Sciences
University of Notre Dame
Notre Dame, Indiana
This page intentionally left blank.
Seismic Analysis, Behavior,

and Design of Unbonded
Post-Tensioned Hybrid
Coupled Wall Structures
December 2006
Report #NDSE-06-02
by
Qiang Shen
Former Graduate
Research Assistant
Yahya C. Kurama
Associate Professor
and
Brad D. Weldon
Graduate Research
Assistant
Structural Engineering Research Report

Department of Civil Engineering and Geological Sciences
University of Notre Dame
Notre Dame, Indiana
ABSTRACT
This report focuses on the development of a new type of hybrid coupled wall
structure for seismic regions. Coupling of concrete wall piers is achieved by posttensioning steel beams to the walls using unbonded post-tensioning tendons. Different
from conventional hybrid coupled walls, the coupling beams of the new system are not
embedded into the walls. Steel top and seat angles are used at the beam-to-wall
connections for energy dissipation.
Unbonded post-tensioned hybrid coupling systems offer important advantages over
conventional coupling systems, such as simplified construction, ability to undergo large
lateral displacements with little damage, and self-centering capability. Analytical
investigations are conducted on the nonlinear behavior of floor-level coupled wall
subassemblages and multi-story structures under combined lateral and gravity loads. Both
finite element and fiber element analytical models are developed. The effects of design
parameters (e.g., amount of post-tensioning, beam/wall properties) on the behavior of the
structures (e.g., strength, energy dissipation, deformation capacity) are investigated.
Systems with precast concrete walls as well as monolithic cast-in-place concrete walls are
considered. The behavior of the proposed coupled wall system is compared with the
behaviors of uncoupled walls and conventional systems with embedded steel coupling
beams. The analytical results are used to develop approximate procedures to estimate the
nonlinear load-deformation behavior of the structures without sophisticated analytical
models.
The results from eleven half-scale experiments of floor-level unbonded posttensioned hybrid coupled wall subassemblages are also summarized in this report. The
tests results are used to validate and improve the analytical models, evaluate critical
structural components that can limit lateral strength and ductility, and make
recommendations for practical applications.
Finally, a performance-based seismic design approach is developed for unbonded
post-tensioned hybrid coupled wall structures. Two prototype eight-story hybrid coupled
wall systems are designed using the proposed procedures. Evaluations of the global and
local behavior of the structures are conducted based on nonlinear static lateral load
analyses as well as dynamic time-history analyses under selected ground motion records.
The results from these analyses are ultimately used to critically evaluate the validity of
the design approach and procedures in achieving the target design performance objectives
for the structures.
This report may be downloaded from http://www.nd.edu/~concrete/.
CONTENTS
FIGURES............................................................................................................................x
TABLES....................................................................................................................... xxxii
SYMBOLS ...................................................................................................................... xlii
ACKNOWLEDGMENTS .............................................................................................. lvi
CHAPTER 1 INTRODUCTION ......................................................................................1
1.1 Problem Statement..................................................................................................1
1.2 Proposed Hybrid Coupled Wall System .................................................................3
1.3 Research Objectives................................................................................................5
1.4 Research Scope.......................................................................................................5
1.5 Summary of Approach............................................................................................6
1.6 Research Significance.............................................................................................7
1.7 Organization of Report ...........................................................................................8
CHAPTER 2 LITERATURE REVIEW ......................................................................10
2.1 Coupled Wall Structural Systems ........................................................................10
2.2 Concrete Coupled Wall Structures ......................................................................12
2.2.1 Behavior and Design of Concrete Coupling Beams ...................................13
2.2.2 Analysis and Modeling of Concrete Coupled Wall Structures ...................14
2.3 Hybrid Structures with Embedded Steel Coupling Beams...................................15
2.3.1 Behavior and Design of Embedded Steel Coupling Beam
Subassemblages ...........................................................................................15
2.3.2 Embedded Steel Coupling Beams with Concrete Encasement ...................18
2.3.3 Multi-Story Hybrid Coupled Walls with Embedded Beams ......................18
2.3.4 Degree of Coupling .....................................................................................20
2.4 Wall Shear Force Demands in Coupled Wall Structures .....................................21
2.5 Unbonded Post-Tensioning in Building Construction .........................................21
2.5.1 Unbonded Post-Tensioning ........................................................................22
2.5.2 Unbonded Post-Tensioned Precast Concrete Moment Frames ..................23
2.5.3 Post-Tensioned Steel Moment Frames .......................................................29
2.5.4 Unbonded Post-Tensioned Precast Concrete Structural Walls ...................31
2.6 Behavior of Top and Seat Angles ........................................................................37
2.6.1 Kishi and Chen (1990) and Lorenz et al. (1993) ........................................39
2.6.2 Sims (2000) .................................................................................................42
2.6.3 Ricles et al. (2001) ......................................................................................43
ii
2.6.4 Garlock et al. (2003) ...................................................................................45

2.7 Chapter Summary ................................................................................................46
CHAPTER 3 ANALYTICAL MODELING OF COUPLED WALL
SUBASSEMBLAGES ..............................................................................48
3.1 Prototype Subassemblage ....................................................................................48
3.1.1 Subassemblage Material Properties and Idealizations ................................51
3.2 Analytical Modeling Assumptions ......................................................................53
3.3 Fiber Element Subassemblage Model .................................................................54
3.3.1 Modeling of Wall Regions ..........................................................................54
3.3.2 Modeling of Coupling Beams and Flange Cover Plates .............................59
3.3.3 Modeling of Gap Opening ..........................................................................60
3.3.4 Modeling of Beam Post-Tensioning Tendons and Anchorages .................62
3.3.5 Modeling of Top and Seat Angles ..............................................................63
3.4 Verification of Fiber Element Subassemblage Model .........................................66
3.4.1 Finite Element Subassemblage Model ........................................................66
3.4.2 Finite Element Analysis Results .................................................................68
3.4.3 Comparisons Between Finite Element and Fiber Element Model
Results..........................................................................................................69
3.5 Advantages and Limitations of Fiber Element Model .........................................72
3.6 Modeling of Embedded Steel Coupling Beams ...................................................73
3.7 Chapter Summary ................................................................................................77
CHAPTER 4 NONLINEAR BEHAVIOR OF COUPLED WALL
SUBASSEMBLAGES................................................................................79
4.1 Overview of Subassemblage Analyses ................................................................79
4.2 Behavior Under Monotonic Loading ...................................................................80
4.3 Behavior Under Cyclic Loading ..........................................................................83
4.4 Parametric Analyses ............................................................................................85
4.5 Chapter Summary ................................................................................................87
CHAPTER 5 IDEALIZED SUBASSEMBLAGE MOMENT-ROTATION
RELATIONSHIP.......................................................................................89
5.1 Linear-Elastic Behavior .......................................................................................89
5.2 Idealized Nonlinear Relationship ........................................................................91
5.2.1 Beam Softening State .................................................................................92
5.2.2 Beam PT-Yielding State .............................................................................95
5.3 Verification of Idealized Relationship ...............................................................100
5.4 Chapter Summary ..............................................................................................103
CHAPTER 6 ANALYTICAL MODELING OF MULTI-STORY COUPLED
WALLS ...................................................................................................104
6.1 Prototype Coupled Wall Structures ...................................................................104
6.1.1 Walls .........................................................................................................104
iii
6.1.2 Coupling Beams ........................................................................................107

6.1.3 Material Properties and Idealizations .......................................................109
6.2 Analytical Modeling Assumptions ....................................................................112
6.3 Multi-Story Coupled Wall Models ....................................................................113
6.3.1 Cast-in-Place Concrete Walls in CIP-UPT system ...................................114
6.3.2 Precast Concrete Walls in PRE-UPT system ............................................118
6.4 Verification of Wall Models ..............................................................................122
6.5 Advantages and Limitations of Analytical Models ...........................................122
6.6 Modeling of Walls with Embedded Coupling Beams .......................................124
6.7 Chapter Summary ..............................................................................................124
CHAPTER 7 NONLINEAR BEHAVIOR OF MULTI-STORY COUPLED
WALLS. ................................................................................................125
7.1 Overview of Analyses ........................................................................................125
7.2 Behavior Under Monotonic Loading .................................................................127
7.2.1 Coupled Wall Base Shear Force versus Roof Drift Behaviors .................127
7.2.2 Wall Pier Deflected Shapes ......................................................................129
7.2.3 Wall Pier Base Flexural Steel Strains and Stresses ..................................129
7.2.4 Wall Pier Base Axial Forces .....................................................................131
7.2.5 Wall Pier Base Moments ..........................................................................134
7.2.6 Degree of Coupling ...................................................................................134
7.2.7 Wall Pier Base Concrete Strains ...............................................................136
7.2.8 Coupling Beam Shear Force versus Chord Rotation Behaviors ...............138
7.2.9 Coupling Beam Axial Forces ....................................................................141
7.2.10 Coupling Beam End Strains ....................................................................146
7.2.11 Tension Angle Force versus Deformation Behaviors .............................146
7.3 Behavior Under Cyclic Loading ........................................................................153
7.3.1 Coupled Wall Base Shear Force versus Roof Drift Behaviors .................153
7.4 Parametric Investigation ....................................................................................159
7.4.1 Base Shear versus Roof Drift Behavior ....................................................160
7.4.2 Degree of Coupling ...................................................................................160
7.5 Chapter Summary ..............................................................................................164
CHAPTER 8 IDEALIZED COUPLED WALL LATERAL LOADDISPLACEMENT RELATIONSHIPS ................................................167
8.1 Linear-Elastic Behavior .....................................................................................167
8.2 Systems with Cast-in-Place Concrete Walls ......................................................171
8.2.1 Coupled Wall Softening State ..................................................................171
8.2.2 Coupled Wall Ultimate State ....................................................................175
8.3 Systems with Precast Concrete Walls ................................................................180
8.3.1 Coupled Wall Softening State ..................................................................180
8.3.2 Coupled Wall PT-Yielding State ..............................................................183
iv
8.3.3 Coupled Wall Ultimate State ....................................................................189

8.4 Verification of Idealized Relationships .............................................................191
8.5 Chapter Summary ..............................................................................................197
CHAPTER 9 SUBASSEMBLAGE EXPERIMENTS ...............................................198
9.1 Experimental Program .......................................................................................198
9.1.1 Test Set-Up ...............................................................................................198
9.1.2 Loading Block and Reaction Block ..........................................................202
9.1.3 Coupling Beams ........................................................................................203
9.1.4 Beam Post-Tensioning Strands .................................................................206
9.1.5 Beam-to-Wall Connection Regions ..........................................................206
9.1.6 Loading .....................................................................................................210
9.1.7 Material Properties ....................................................................................212
9.1.8 Instrumentation .........................................................................................224
9.1.9 Test Procedure ..........................................................................................227
9.2 Results of Experiments ......................................................................................229
9.2.1 Behavior of Properly Designed Subassemblages (Tests 1-4) ...................229
9.2.2 Effect of Coupling Beam Flange Cover Plates (Tests 5-6) ......................238
9.2.3 Effect of Angle Thickness (Tests 7-8) ......................................................241
9.2.4 Effect of Beam Post-Tensioning Steel Area (Test 9) ...............................242
9.2.5 Effect of Beam Depth (Tests 10-11) .........................................................243
9.3 Recommendations for Application ....................................................................247
9.4 Chapter Summary ..............................................................................................247
CHAPTER 10 REVISED ANALYTICAL MODEL .................................................249
10.1 Revised Subassemblage Model .......................................................................249
10.1.1 Modeling of Coupling Beams and Wall-Contact Regions .....................249
10.1.2 Modeling of Beam Post-Tensioning Tendons ........................................251
10.1.3 Modeling of Top and Seat Angles ..........................................................253
10.2 Evaluation of Revised Subassemblage Model..................................................256
10.2.1 Coupling Shear Force versus Chord Rotation Behaviors .......................257
10.2.2 Coupling Beam Post-Tensioning Forces ................................................261
10.2.3 Behavior at Coupling Beam Ends ...........................................................261
10.2.4 Behavior Along Beam Span ...................................................................262
10.2.5 Behavior of Top and Seat Angles ...........................................................268
10.3 Revised Idealized Moment-Rotation Relationship ..........................................270
10.3.1 Beam Softening State .............................................................................271
10.3.2 Beam PT-Yielding State .........................................................................271
10.4 Chapter Summary ............................................................................................274
CHAPTER 11 PERFORMANCE-BASED SEISMIC DESIGN APPROACH .......275
11.1 Overview ..........................................................................................................275
11.2 Performance (Damage) Levels ........................................................................277
11.2.1 Immediate Occupancy Performance Level .............................................277
v
11.2.2 Life Safety Performance Level ...............................................................277

11.2.3 Collapse Prevention Performance Level ................................................277
11.3 Structure Limit States and Capacities ..............................................................277
11.3.1 Steel Coupling Beam Limit States ..........................................................278
11.3.2 Concrete Wall Pier Limit States .............................................................279
11.4 Seismic Demand Levels ..................................................................................281
11.5 Structure Demands ...........................................................................................281
11.6 Relationships Between Performance Levels and Limit States ........................282
11.6.1 Steel Coupling Beams .............................................................................282
11.6.2 Cast-in-Place Concrete Wall Piers ..........................................................282
11.6.3 Precast Concrete Wall Piers ....................................................................283
11.7 Design Objectives ............................................................................................284
11.7.1 Coupling Beams ......................................................................................284
11.7.2 Systems with Cast-in-Place Concrete Wall Piers ...................................284
11.7.3 Systems with Precast Concrete Wall Piers .............................................285
11.8 Seismic Design Criteria ...................................................................................285
11.9 Design Acceleration Response Spectra ...........................................................286
11.9.1 Design Spectra for Survival-Level Demand ...........................................286
11.9.2 Design Spectra for Design-Level Demand .............................................287
11.10 Equivalent Nonlinear SDOF Representation .................................................288
11.10.1 Transformation from MDOF System to SDOF System .......................288
11.10.2 Equivalent Nonlinear SDOF Model .....................................................293
11.10.3 Application of Equivalent SDOF BP Model ........................................294
11.10.4 Evaluation of Equivalent SDOF BP Model ..........................................298
11.11 Seismic Displacement Demand Relationships ..............................................299
11.11.1 Previous Research .................................................................................303
11.11.2 Displacement Demand Relationships for SDOF BP Model .................306
11.12 Nonlinear Static Procedure ............................................................................310
11.12.1 Coupled Wall Design Base Shear Demand, Qwd ..................................310
11.12.2 Vertical Distribution of Seismic Lateral Forces ...................................312
11.12.3 Horizontal Distribution of Story Forces ...............................................313
11.12.4 Load Combinations ...............................................................................314
11.13 Maximum Base Shear Demand .....................................................................315
11.14 Chapter Summary ..........................................................................................317
CHAPTER 12 DESIGN OF PROTOTYPE STRUCTURES ...................................319
12.1 Design Overview .............................................................................................319
12.1.1 Design Material Properties .....................................................................321
12.1.2 Selection of Initial Dimensions for Structural Members ........................321
12.1.3 Design Gravity Loads and Seismic Weight ............................................322
12.1.4 Design Load Combination and Capacity Reduction Factors ..................323
12.1.5 Design Scope and Limitations ................................................................324
12.2 Performance-Based Seismic Design Parameters .............................................324
12.2.1 Allowable Target Roof Drift Demands ..................................................324
vi
12.2.2 Allowable Inter-Story Drift Demands ....................................................325

12.2.3 Acceleration Response Spectra Used in Design .....................................325
12.3 Linear-Elastic Structure Properties ..................................................................326
12.3.1 Linear Elastic Stiffness ...........................................................................327
12.3.2 Period, Mode Shape, and Effective Mass ...............................................327
12.4 Design Base Shear Demand .............................................................................328
12.4.1 Displacement Demand Relationships .....................................................328
12.4.2 Strength Ratios ........................................................................................329
12.4.3 Building Design Base Shear Demands ...................................................330
12.4.4 Coupled Wall Design Base Shear Demands ...........................................330
12.5 Preliminary Design Checks .............................................................................331
12.5.1 Peak Roof Drift Demands .......................................................................331
12.5.2 Peak Inter-Story Drift Demands .............................................................332
12.5.3 Diagonal Tension at Wall Pier Bases .....................................................332
12.6 Design of Steel Coupling Beams .....................................................................334
12.6.1 Design Coupling Beam Shear Force Demand ........................................334
12.6.2 Angle and Post-Tensioning Contributions to Coupling Resistance .......336
12.6.3 Coupling Beam Cross Section ................................................................339
12.6.4 Beam Post-Tensioning ............................................................................340
12.6.5 Top and Seat Angles and Connections ...................................................344
12.6.6 Coupling Beam Shear Force versus Rotation Behavior .........................346
12.6.7 Coupling Beam Rotation Demands ........................................................348
12.6.8 Yielding of Beam Post-Tensioning Tendons ..........................................349
12.6.9 Fracture of Top and Seat Angles ............................................................350
12.6.10 Shear Slip at Beam-to-Wall Interfaces .................................................351
12.6.11 Wall-Contact Regions ...........................................................................352
12.7 Design of Wall Piers in Structure P1-CWUPT ................................................354
12.7.1 Wall Pier Flexural Reinforcement ..........................................................355
12.7.2 Concrete Confinement at Wall Pier Bases ..............................................358
12.7.4 Shear Slip at Coupled Wall Base ............................................................362
12.8 Design of Wall Piers in Structure P2-PWUPT ................................................363
12.8.1 Wall Pier Post-Tensioning Steel Reinforcement ....................................363
12.8.2 Yielding of Wall Pier Post-Tensioning Steel ..........................................365
12.8.3 Concrete Confinement at Wall Pier Bases ..............................................367
12.8.5 Shear Slip at Coupled Wall Base ............................................................371
12.9 Summary of Design and Final Checks ............................................................373
12.9.1 Lateral Load versus Deformation Relationships ....................................373
12.9.2 Design Criteria ........................................................................................375
CHAPTER 13 BEHAVIOR OF PROTOTYPE STRUCTURES UNDER STATIC
LATERAL LOADING .........................................................................379
13.1 Introduction ......................................................................................................379
vii
13.2 Behavior Under Monotonic Loading ...............................................................380

13.2.1 Coupled Wall Base Shear Force versus Roof Drift Behaviors ...............380
13.2.2 Wall Pier Base Flexural Steel Strains and Stresses ................................384
13.2.3 Wall Pier Base Axial Forces ...................................................................386
13.2.4 Wall Pier Base Shear Forces ...................................................................392
13.2.5 Wall Pier Base Moments ........................................................................393
13.2.6 Degree of Coupling .................................................................................397
13.2.7 Wall Pier Base Concrete Strains .............................................................397
13.2.8 Coupling Beam Shear Force versus Chord Rotation Behaviors .............401
13.2.9 Coupling Beam Axial Forces and Post-Tensioning Forces ....................404
13.2.10 Coupling Beam End Strains ..................................................................407
13.2.11 Tension Angle Force versus Deformation Behaviors ...........................407
13.3 Behavior Under Reversed Cyclic Loading ......................................................414
13.3.1 Coupled Wall Base Shear Force versus Roof Drift Behaviors ...............414
13.3.2 Wall Pier Post-Tensioning Forces in Structure P2-PWUPT ..................416
13.3.3 Wall Pier Base Axial Forces ...................................................................416
13.3.4 Wall Pier Base Shear Forces ...................................................................419
13.3.5 Wall Pier Base Moments ........................................................................419
13.3.6 Degree of Coupling .................................................................................423
13.3.7 Wall Pier Base Concrete Strains .............................................................424
13.3.8 Coupling Beam Shear Force versus Chord Rotation Behaviors .............427
13.3.9 Coupling Beam Axial Forces and Post-Tensioning Forces ....................429
13.3.10 Coupling Beam End Strains ..................................................................430
13.3.11 Angle Force versus Deformation Behaviors .........................................437
13.4 Structure with Embedded Steel Coupling Beams ............................................437
13.4.1 Behaviors Under Monotonic and Cyclic Loading ..................................437
13.5 Chapter Summary ............................................................................................441
CHAPTER 14 GROUND MOTION RECORDS FOR DYNAMIC ANALYSES ..442
14.1 Overview ..........................................................................................................442
14.2 Design-Level Ground Motion Records ...........................................................443
14.3 Survival-Level Ground Motion Records .........................................................448
CHAPTER 15 BEHAVIOR OF PROTOTYPE STRUCTURES UNDER
EARTHQUAKE LOADING ................................................................454
15.1 Analytical Model for Dynamic Analyses ........................................................454
15.1.1 Building Mass .........................................................................................454
15.1.2 Viscous Damping ....................................................................................455
15.1.3 Time Step ................................................................................................456
15.2 Structure P1-CWUPT ......................................................................................456
15.2.1 Floor/Roof Drifts ....................................................................................456
15.2.2 Inter-Story Drifts .....................................................................................458
15.2.3 Floor/Roof Accelerations ........................................................................462
15.2.4 Wall Base Axial Forces ..........................................................................464
viii
15.2.5 Wall Base Shear Forces ..........................................................................468

15.2.6 Wall Base Diagonal Tension and Shear Slip Failure ..............................477
15.2.7 Wall Base Moments ................................................................................477
15.2.8 Wall Base Strains ....................................................................................484
15.2.9 Coupling Beam Chord Rotations ............................................................487
15.2.10 Coupling Beam Axial Forces ................................................................489
15.2.11 Coupling Beam Post-Tensioning Forces ..............................................494
15.2.12 Coupling Beam Shear Forces ...............................................................495
15.2.13 Coupling Beam Shear Slip Failure .......................................................501
15.2.14 Angle Deformations ..............................................................................504
15.3 Structure P2-PWUPT .......................................................................................509
15.3.1 Floor/Roof Drifts ....................................................................................509
15.3.2 Inter-Story Drifts ....................................................................................511
15.3.3 Floor/Roof Accelerations ........................................................................516
15.3.4 Wall Post-Tensioning Forces ..................................................................518
15.3.5 Wall Base Axial Forces ..........................................................................522
15.3.6 Wall Base Shear Forces ..........................................................................526
15.3.7 Wall Base Diagonal Tension and Shear Slip Failure ..............................534
15.3.8 Wall Base Moments ................................................................................538
15.3.9 Wall Base Strains ....................................................................................544
15.3.10 Coupling Beam Chord Rotations ..........................................................547
15.3.11 Coupling Beam Axial Forces ................................................................549
15.3.12 Coupling Beam Post-Tensioning Forces ..............................................554
15.3.13 Coupling Beam Shear Forces ...............................................................555
15.3.14 Coupling Beam Shear Slip Failure .......................................................561
15.3.15 Angle Deformations ..............................................................................562
15.4 Uncoupled Concrete Wall Systems .................................................................568
15.5 Chapter Summary ............................................................................................571
CHAPTER 16 SUMMARY, CONCLUSIONS, AND FUTURE WORK ................572
16.1 Summary ..........................................................................................................572
16.2 Conclusions ......................................................................................................575
16.3 Future Work .....................................................................................................581
REFERENCES ..............................................................................................................582
ix
FIGURES
CHAPTER 1
Fig. 1.1 Eight story coupled wall structure ..........................................................................2
Fig. 1.2 Hybrid coupled wall subassemblage with embedded beam: (a) undeformed
configuration; (b) deformed configuration ..........................................................................2
Fig. 1.3 Proposed coupled wall system: (a) subassemblage at a floor level; (b) deformed
configuration; (c) coupling forces........................................................................................4
CHAPTER 2
Fig. 2.1 Coupled structural walls .......................................................................................10
Fig. 2.2 Lateral load resisting mechanisms in structural walls (adapted from Paulay and
Priestley 1992): (a) uncoupled wall; (b) walls coupled with strong beams; (c) walls
coupled with weak beams ..................................................................................................11
Fig. 2.3 Shear in concrete coupling beams (from Paulay and Priestley 1992): (a) diagonal
tension failure; (b) sliding shear failure; (c) beam with diagonal reinforcement ..............13
Fig. 2.4 Modeling of coupled wall systems (adapted from Paulay and Priestley 1992): (a)
walls with flexible beams; (b) walls with stiff beams........................................................14
Fig. 2.5 Hybrid coupled wall subassemblages with embedded steel beams (from Harries
et al. 1997); (a) shear-critical beam; (b) beam shear force versus relative end
displacement behavior; (c) damage at end of test; (d) flexure-critical beam; (e) beam
shear versus relative end displacement behavior; (f) damage at end of test......................16
Fig. 2.6 Hybrid coupled wall system with embedded steel beams (adapted from Harries
1996): (a) plan view; (b) analytical model.........................................................................19
Fig. 2.7 Unbonded post-tensioned precast concrete frame: (a) elevation; (b) beam-column
subassemblage; (c) idealized exaggerated displaced shape ...............................................24
Fig. 2.8 Unbonded post-tensioned precast beam-column subassemblages (from Priestley
and MacRae1996): (a) interior joint specimen; (b) exterior joint specimen; (c) test setup;
x
(d) photograph of test setup; (e) interior joint behavior; (f) exterior joint behavior; (g)
photograph of interior joint................................................................................................25
Fig. 2.9 Unbonded post-tensioned precast beam-column subassemblages (from El-Sheikh
et al. 1999, 2000): (a) moment-rotation behavior; (b) analytical model............................27
Fig. 2.10 Partially post-tensioned precast beam-column subassemblages (adapted from
Kurama 2002) ...................................................................................................................28
Fig. 2.11 Post-tensioned steel exterior beam-column connection (adapted from Ricles et
al. 2001) ............................................................................................................................30
Fig. 2.12 Post-tensioned steel beam-column connection details (from Ricles et al. 2002):
(a) specimen with wide flange column; (b) specimen with composite concrete filled steel
tube column........................................................................................................................32
Fig. 2.13 Lateral load versus displacement behavior of post-tensioned steel beam-column
subassemblages (from Ricles et al. 2001) .........................................................................33
Fig. 2.14 Analytical modeling of post-tensioned steel beam-column subassemblages
(adapted from Ricles et al. 2001): (a) interior joint region; (b) beam-column
subassemblage; (c) modeling of post-tensioning anchorages............................................33
Fig. 2.15 Unbonded post-tensioned precast concrete walls (adapted from Kurama et al.
1999a): (a) wall elevation and cross-section; (b) gap opening and shear slip behavior; (c)
analytical model; (d) behavior under cyclic lateral loads; (e) dynamic response..............34
Fig. 2.16 Experiments of precast concrete walls (from Holden et al. 2003 and Restrepo
2003): (a) test setup; (b) hysteretic response of emulative wall; (c) hysteretic response of
partially post-tensioned wall; (d) photograph of partially post-tensioned wall at 3% drift
............................................................................................................................................37
Fig. 2.17 Top-and-seat angle connection: (a) deformed configuration; (b) angle
parameters ..........................................................................................................................38
Fig. 2.18 Angle model (adapted from Kishi and Chen 1990 and Lorenz et al. 1993): (a)
cantilever model of tension angle; (b) assumed yield mechanism; (c) free body diagram
of angle horizontal leg .......................................................................................................40
Fig. 2.19 Isolated angle experiments (from Sims 2000): (a) plan view of test set up; (b)
angle force-deformation relationship of Specimen 9; (c) angle force-deformation
relationship of Specimen 11...............................................................................................44
xi
Fig. 2.20 Modeling of top and seat angles (from Ricles et al. 2001): (a) fiber Models A
and B; (b) fiber model and spring model behaviors; (c) comparisons with test results.....45
CHAPTER 3
Fig. 3.1 Prototype coupled wall subassemblage: (a) elevation; (b) top and seat angle
connection details...............................................................................................................50
Fig. 3.2 Subassemblage material properties and idealizations: (a) steel; (b) beam posttensioning strand; (c) wall unconfined concrete; (d) wall-contact region confined concrete
............................................................................................................................................52
Fig. 3.3 Analytical model: (a) subassemblage; (b) wall-contact elements and beam
elements .............................................................................................................................55
Fig. 3.4 Fiber element, segments, and fibers .....................................................................56
Fig. 3.5 Schematic showing fiber discretization of wall-contact elements........................57
Fig. 3.6 Length of first beam fiber element segment.........................................................60
Fig. 3.7 Fiber discretization of beam elements ..................................................................61
Fig. 3.8 Cyclic material models: (a) compression-only steel (SC) fiber; (b) compressiononly confined concrete (CCC) fiber; (c) compression-tension steel (S) fiber; (d) beam PT
element...............................................................................................................................63
Fig. 3.9 Angle model: (a) idealized displaced shape; (b) cyclic angle load-deformation
model; (c) angle load-deformation behavior for prototype subassemblage.......................65
Fig. 3.10 Finite element model: (a) two dimensional (2D) model; (b) three-dimensional
(3D) model; (c) close-up view of beam-to-wall interface .................................................67
Fig. 3.11 Comparison between 2D and 3D finite element models ....................................68
Fig. 3.12 Rigid wall versus deformable wall finite element models .................................69
Fig. 3.13 Maximum principal stress distributions: (a) beam stresses; (b) wall stresses ........
............................................................................................................................................70
Fig. 3.14 Model verification: (a) rigid wall comparison; (b) deformable wall comparison;
(c) contact depth comparison .............................................................................................71
xii
Fig. 3.15 Modeling of embedded steel coupling beams: (a) Specimen S4 (Harries et al.
1997); (b) analytical model; (c) steel stress-strain relationship .........................................74
Fig. 3.16 Modeling of embedded steel coupling beams: (a) Specimen 3 (Remmetter et al.
1992); (b) analytical model; (c) steel stress-strain relationship .........................................76
Fig. 3.17 Verification of embedded steel coupling beam model: (a) Specimen S4 (Harries
et al. 1997); (b) Specimen 3 (Remmetter et al. 1992) .......................................................77
CHAPTER 4
Fig. 4.1 Simplified subassemblage model .........................................................................79
Fig. 4.2 Displaced shape: (a) subassemblage; (b) analytical model ..................................80
Fig. 4.3 Subassemblage behavior under monotonic loading: (a) beam end moment versus
chord rotation behavior; (b) total force in beam post-tensioning tendons; (c) contact
depth; (d) maximum beam compression strain; (e) tension angle behavior ......................81
Fig. 4.4 Subassemblage behavior under cyclic loading: (a) with L881-1/8 angles; (b)
total force in post-tensioning tendons; (c) with L883/4 angles; (d) without angles......84
Fig. 4.5 Subassemblage moment-rotation behaviors from parametric investigation: (a) abt
and fbpi; (b) abt; (c) fbpi; (d) lw; (e) tc; (f) tbf; (g) dbw; (h) lb ..................................................86
CHAPTER 5
Fig. 5.1 Linear-elastic subassemblage model: (a) model with wall-contact elements; (b)
model without wall-contact elements ................................................................................90
Fig. 5.2 Idealized bilinear subassemblage moment-rotation relationship..........................92
Fig. 5.3 Coupling beam free body diagram at softening state ...........................................93
Fig. 5.4 Coupling beam free body diagram at PT-yielding state .......................................96
Fig. 5.5 Gap opening at coupling beam PT-yielding state.................................................97
Fig. 5.6 Second order effects .............................................................................................99
Fig. 5.7 Verification of moment estimations: (a) abt and fbpi; (b) abt; (c) fbpi; (d) lw; (e) tc;
(f) tbf; (g) dbw; (h) lb ..........................................................................................................101
xiii
Fig. 5.8 Verification of rotation estimations: (a) abt and fbpi; (b) abt; (c) fbpi; (d) lw; (e) tc;
(f) tbf; (g) dbw; (h) lb ..........................................................................................................102
CHAPTER 6
Fig. 6.1 Plan view of prototype structures .......................................................................105
Fig. 6.2 Wall elevation and cross-section at base: (a) CIP-UPT system; (b) PRE-UPT
system ..............................................................................................................................106
Fig. 6.3 Floor level details: (a) elevation; (b) top and seat angle connections ................108
Fig. 6.4 Wall material properties and idealizations: (a) mild reinforcing steel for CIP-UPT
system; (b) post-tensioning bar for PRE-UPT system; (c) concrete................................110
Fig. 6.5 Multi-story coupled wall analytical models: (a) CIP-UPT system; (b) PRE-UPT
system ..............................................................................................................................113
Fig. 6.6 CIP-UPT wall-height elements: (a) elevation; (b) fiber slice schematic near wall
base; (c)fiber slice schematic away from wall base.........................................................115
Fig. 6.7 CIP-UPT system cyclic material models: (a) confined concrete (CC) fiber; (b)
unconfined concrete (CU) fiber; (c) mild steel reinforcement (S) fiber; (d) steel fiber
stress-strain relationship under monotonic loading and envelope relationship under cyclic
loading; (e) concrete in tension........................................................................................117
Fig. 6.8 PRE-UPT wall-height elements: (a) elevation; (b) fiber slice schematic near wall
base; (c) fiber slice schematic away from wall base........................................................119
Fig. 6.9 PRE-UPT system cyclic material models: (a) compression-only confined
concrete (CCC) fiber; (b) wall PT element......................................................................121
Fig. 6.10 Monolithic cast-in-place reinforced concrete wall-height element model
verification: (a) Wall RW2 (from Thomsen 1995); (b) measured behavior; (c) predicted
behavior............................................................................................................................123
CHAPTER 7
Fig. 7.1 Gravity and lateral forces for multi-story coupled wall analyses.......................126
xiv
Fig. 7.2 Base shear versus roof drift behaviors under monotonic lateral loads: (a) CIPUPT system; (b) PRE-UPT system..................................................................................128
Fig. 7.3 Wall deflected shapes: (a) CIP-UPT system; (b) PRE-UPT system ..................130
Fig. 7.4 Wall pier flexural steel stresses: (a) CIP-UPT system; (b) PRE-UPT system ...131
Fig. 7.5 Wall pier base axial forces: (a) CIP-UPT system; (b) PRE-UPT system...........132
Fig. 7.6 Components of wall pier base axial forces: (a) CIP-UPT system; (b) PRE-UPT
system ..............................................................................................................................133
Fig. 7.7 Wall base moments: (a) CIP-UPT system; (b) PRE-UPT system......................135
Fig. 7.8 Degree of coupling .............................................................................................136
Fig. 7.9 Wall pier base neutral axis depths: (a) CIP-UPT system; (b) PRE-UPT system .....
..........................................................................................................................................137
Fig. 7.10 Wall pier base extreme confined concrete compression strains: (a) CIP-UPT
system; (b) PRE-UPT system ..........................................................................................138
Fig. 7.11 Coupling beam shear force versus chord rotation behaviors for CIP-UPT system
..........................................................................................................................................139
Fig. 7.12 Coupling beam shear force versus chord rotation behaviors for PRE-UPT
system ..............................................................................................................................140
Fig. 7.13 Calculation of coupling beam chord rotation, b in multi-story coupled wall........
..........................................................................................................................................141
Fig. 7.14 Coupling beam axial forces and post-tensioning forces in CIP-UPT system: (a)
2nd floor; (b) 3rd floor; (c) 4th floor; (d) 5th floor; (e) 6th floor; (f) 7th floor; (g) 8th floor; (h)
roof...................................................................................................................................143
Fig. 7.15 Effect of beam post-tensioning sequence on beam axial forces: (a) 2nd floor to
roof; (b) roof to 2nd floor.................................................................................................144
Fig. 7.16 Simulation of beam post-tensioning using external forces...............................145
Fig. 7.17 Coupling beam end neutral axis (i.e., contact) depths for CIP-UPT system....147
Fig. 7.18 Coupling beam end neutral axis (i.e., contact) depths for PRE-UPT system...148
xv
Fig. 7.19 Coupling beam end extreme steel compression strains for CIP-UPT system ........
..........................................................................................................................................149
Fig. 7.20 Coupling beam end extreme steel compression strains for PRE-UPT system .......
..........................................................................................................................................150
Fig. 7.21 Coupling beam tension angle force-deformation behaviors for CIP-UPT system
..........................................................................................................................................151
Fig. 7.22 Coupling beam tension angle force-deformation behaviors for PRE-UPT system
..........................................................................................................................................152
Fig. 7.23 Coupled wall base shear force versus roof drift behaviors under cyclic lateral
loads: (a) CIP-UPT system; (b) PRE-UPT system; (c) CIP-EMB system; (d) CIP-UPT
system without angles ......................................................................................................154
Fig. 7.24 Coupling beam axial forces in CIP-UPT system cyclic loading: (a) 2nd floor;
(b) 3rd floor; (c) 4th floor; (d) 5th floor; (e) 6th floor; (f) 7th floor; (g) 8th floor; (h) roof ........
..........................................................................................................................................156
Fig. 7.25 Coupling beam post-tensioning forces in CIP-UPT system cyclic loading: (a)
2nd floor; (b) 3rd floor; (c) 4th floor; (d) 5th floor; (e) 6th floor; (f) 7th floor; (g) 8th floor; (h)
roof...................................................................................................................................157
Fig. 7.26 Coupling beam axial forces: (a) equilibrium of horizontal forces; (b) wall
floor/roof shear forces......................................................................................................158
Fig. 7.27 Base shear versus roof drift behaviors of parametric CIP-UPT systems: (a) lw;
(b) ws; (c) tw; (d) dbw; (e) abt; (f) fbpi; (g) lb; (h) ta............................................................161
Fig. 7.28 Base shear versus roof drift behaviors of parametric PRE-UPT systems: (a) lw;
(b) wp; (c) fwpi; (d) dbw; (e) abt; (f) fbpi; (g) lb; (h) ta .........................................................162
CHAPTER 8
Fig. 8.1 Eight story linear-elastic coupled wall model ....................................................169
Fig. 8.2 Idealized base shear force versus roof drift relationship for prototype CIP-UPT
system ..............................................................................................................................171
Fig. 8.3 Coupled wall softening state for systems with cast-in-place concrete walls: (a)
free body diagram; (b) coupling beam shear forces; (c) strain and stress distributions at
wall bases .........................................................................................................................173
xvi
Fig. 8.4 Coupled wall ultimate state for systems with cast-in-place concrete walls: (a)
strain and stress distributions at base of compression-side wall; (b) beam chord rotation;
(c) components of beam chord rotation ...........................................................................176
Fig. 8.5 Coupled wall ultimate state for systems with cast-in-place concrete walls: (a)
coupling beam shear forces; (b) free body diagram.........................................................178
Fig. 8.6 Idealized base shear force versus roof drift relationship for prototype PRE-UPT
system ..............................................................................................................................180
Fig. 8.7 Coupled wall softening state for systems with precast concrete walls: (a) free
body diagram; (b) concrete strain and stress distributions at wall bases .........................182
Fig. 8.8 Coupled wall PT-yielding state for systems with precast concrete walls: (a)
idealized coupled wall displaced shape; (b) beam shear forces.......................................185
Fig. 8.9 Coupled wall PT-yielding state for systems with precast concrete walls: (a) free
body diagram; (b) concrete stress distributions at base of tension-side wall; (c) concrete
stress distributions at base of compression-side wall ......................................................187
Fig. 8.10 Coupled wall ultimate state for systems with precast concrete walls strain and
stress distributions at base of compression-side wall ......................................................190
Fig. 8.11 Verification of idealized relationship for systems with cast-in-place walls Fws
and Fwu .............................................................................................................................192
Fig. 8.12 Verification of idealized relationship for systems with cast-in-place walls ws
and wu .............................................................................................................................193
Fig. 8.13 Verification of idealized relationship for systems with precast walls Fws, Fwy,
and Fwu .............................................................................................................................194
Fig. 8.14 Verification of idealized relationship for systems with precast walls ws, wy,
and wu .............................................................................................................................195
Fig. 8.15 Determination of coupled wall softening state in fiber element model: (a)
systems with cast-in-place concrete walls; (b) systems with precast concrete walls ......196
CHAPTER 9
Fig. 9.1 Photograph of test set-up ....................................................................................199
Fig. 9.2 Elevation of experimental set-up: (a) side view; (b) end view...........................201
xvii
Fig. 9.3 Bracing frame side view .....................................................................................202

Fig. 9.4 Test structure ......................................................................................................203
Fig. 9.5 Loading block: (a) 3D view; (b) side view; (c) south end view .........................204
Fig. 9.6 Reaction block: (a) 3D view; (b) side view; (c) north end view ........................204
Fig. 9.7 Photograph of steel form for casting of reaction block ......................................205
Fig. 9.8 Beam end view ...................................................................................................205
Fig. 9.9 Beam side view...................................................................................................206
Fig. 9.10 Photograph of post-tensioning anchorage system ............................................207
Fig. 9.11 Photograph of wall test region prior to casting of concrete..............................207
Fig. 9.12 Photograph of spiral reinforcement used inside wall test region......................208
Fig. 9.13 Photograph showing shim plates at beam-to-reaction block interface .............209
Fig. 9.14 Photograph showing angle-to-beam and angle-to-wall connections................210
Fig. 9.15 Displacement history ........................................................................................211
Fig. 9.16 Photograph of mono-strand post-tensioning jack and pump ............................212
Fig. 9.17 Beam steel stress-strain behavior .....................................................................213
Fig. 9.18 Cover plate steel specimen ...............................................................................214
Fig. 9.19 Cover plate steel stress-strain behavior ............................................................215
Fig. 9.20 Angle steel specimen ........................................................................................216
Fig. 9.21 Angle steel stress-strain behavior: (a) L8x4x1/2 first heat; (b) L8x4x1/2
second heat; (c) L8x4x5/8................................................................................................217
Fig. 9.22 Spiral wire specimens.......................................................................................219
Fig. 9.23 Spiral wire stress-strain behavior .....................................................................219
Fig. 9.24 Unloading stiffness of spiral wire steel ............................................................221
xviii
Fig. 9.25 Post-tensioning strand stress-strain behavior ...................................................222

Fig. 9.26 Post-tensioning strand test setup.......................................................................223
Fig. 9.27 Concrete cylinder specimens ............................................................................224
Fig. 9.28 Instrumentation: (a) overall view; (b) beam-to-wall joint region.....................225
Fig. 9.29 Measurement of post-tensioning forces: (a) load cells; (b) placement of load
cells ..................................................................................................................................226
Fig. 9.30 DT11-DT15 at displaced position of coupling beam during Test 1 .................227
Fig. 9.31 Measured coupling shear force versus beam chord rotation (Vb-b)
relationships: (a) Test 1; (b) Test 2; (c) Test 3.................................................................231
Fig. 9.32 Test 3: (a) displaced shape overall; (b) displaced shape local....................233
Fig. 9.33 Test 3 total beam post-tensioning force.........................................................234
Fig. 9.34 Test 3 contact depth.......................................................................................235
Fig. 9.35 Test 3 beam flange steel and wall concrete strains........................................236
Fig. 9.36 Test 3 angle fracture ......................................................................................237
Fig. 9.37 Angle behavior (Tests 4, 6, 8, and 11): (a) test set-up; (b) deformed shape; (c)
Vb-b relationships ...........................................................................................................239
Fig. 9.38 Test 5: (a) Vb-b relationship; (b) beam flange steel strains.............................240
Fig. 9.39 Vb-b relationship from Test 7..........................................................................242
Fig. 9.40 Test 9: (a) Vb-b relationship; (b) total beam post-tensioning force; (c) beam
flange steel strains............................................................................................................244
Fig. 9.41 Test 10: (a) Vb-b relationship; (b) total beam post-tensioning force; (c) beam
flange steel strains............................................................................................................245
xix
CHAPTER 10
Fig. 10.1 Revised analytical model: (a) subassemblage; (b) wall-contact elements and
beam elements..................................................................................................................250
Fig. 10.2 Kinking behavior: (a) idealized deformed shape; (b) gap/contact PT kink
element.............................................................................................................................252
Fig. 10.3 Angle model: (a) angle displacements and forces; (b) horizontal angle element;
(c) vertical angle element.................................................................................................254
Fig. 10.4 Coupling shear force versus rotation (Vb-b) behaviors: (a) Test 1; (b) Test 3;
(c) Test 5; (d) Test 7; (e) Test 9; (f) Test 10 ....................................................................258
Fig. 10.5 Test 3 predictions using previous model: (a) Vb-b behavior; (b) angle
behavior............................................................................................................................259
Fig. 10.6 Test 3 predictions using revised model: (a) horizontal angle element; (b)
vertical angle element; (c) model with vertical angle elements removed........................260
Fig. 10.7 Post-tensioning forces: (a) Test 3; (b) Test 9; (c) Test 10 ................................263
Fig. 10.8 Horizontal displacements of loading block: (a) Test 3; (b) Test 9; (c) Test 10......
..........................................................................................................................................264
Fig. 10.9 Beam cover plate/flange strains: (a) Test 1; (b) Test 3; (c) Test 5; (d) Test 7; (e)
Test 9; (f) Test 10.............................................................................................................265
Fig. 10.10 Average compressive strains at beam end: (a) SG22 from Test 1; (b)
calculation of average strains...........................................................................................266
Fig. 10.11 Beam web strains: (a) Test 1; (b) Test 3; (c) Test 5; (d) Test 7; (e) Test 9; (f)
Test 10..............................................................................................................................267
Fig. 10.12 Calculation of average compressive strains at beam web ..............................268
Fig. 10.13 Angle behavior under monotonic loading-Tests 4, 6, and 8...........................269
Fig. 10.14 Angle behavior under cyclic loading: (a) Tests 1 and 3; (b) difference between
Tests 1 and 3 during b=6% cycle....................................................................................270
xx
CHAPTER 11
Fig. 11.1 Outline of proposed performance-based seismic design approach (adapted from
Kurama 1997) .................................................................................................................276
Fig. 11.2 Limit states, performance levels, and design objectives: (a) steel coupling beam;
(b) coupled wall system with cast-in-place concrete wall piers; (c) coupled wall system
with precast concrete wall piers.......................................................................................283
Fig. 11.3 Design acceleration response spectra ...............................................................287
Fig. 11.4 First mode component of roof-displacement time-history ...............................290
Fig. 11.5 BP hysteresis model with s=r<1 (adapted from Farrow and Kurama 2003) ......
..........................................................................................................................................293
Fig. 11.6 Idealized base shear force versus roof displacement relationships: (a) CIP-UPT
system; (b) PRE-UPT system ..........................................................................................294
Fig. 11.7 Equivalent SDOF representation of CIP-UPT system: (a) smooth MDOF
hysteresis; (b) MDOF BP model; (c) hysteretic energy dissipation; (d) equivalent SDOF
BP model..........................................................................................................................297
Fig. 11.8 Equivalent SDOF representation of PRE-UPT system: (a) smooth MDOF
hysteresis; (b) MDOF BP model; (c) hysteretic energy dissipation; (d) equivalent SDOF
BP model..........................................................................................................................298
Fig. 11.9 Peak SDOF versus MDOF displacements: (a) CIP-UPT system; (b) PRE-UPT
system ..............................................................................................................................300
Fig. 11.10a SDOF and MDOF roof-displacement time-histories for CIP-UPT system
(SAC LA21-LA30) ..........................................................................................................301
Fig. 11.10b SDOF and MDOF roof-displacement time-histories for CIP-UPT system
(SAC LA31-LA40) ..........................................................................................................302
Fig. 11.11a SDOF and MDOF roof-displacement time-histories for PRE-UPT system
(SAC LA21-LA30) ..........................................................................................................303
Fig. 11.11b SDOF and MDOF roof-displacement time-histories for PRE-UPT system
(SAC LA31-LA40) ..........................................................................................................304
Fig. 11.12 Definitions of R and (from Farrow and Kurama 2003) ............................305
xxi
Fig. 11.13 Distribution of inertial forces: (a) H=0.67hw; (b) H=0.3hw; (c) bending
moment diagrams (adapted from Paulay and Priestley 1992) ........................................316
CHAPTER 12
Fig. 12.1 Plan view of prototype buildings......................................................................320
Fig. 12.2 Design-level and survival-level acceleration response spectra ........................326
Fig. 12.3 Design coupling beam shear force demand......................................................335
Fig. 12.4 Angle strength ratio, ar: (a) ar=0.66; (b) ar=0.91; (c) ar=1.27 ....................338
Fig. 12.5 Post-tensioning contribution to coupling resistance: (a) preliminary; (b) revised .
..........................................................................................................................................339
Fig. 12.6 Coupling beam and post-tensioning details: (a) Structure P1-CWUPT; (b)
Structure P2-PWUPT.......................................................................................................342
Fig. 12.7 Angle contribution to coupling resistance ........................................................344
Fig. 12.8 Top and seat angle and connection details: (a) Structure P1-CWUPT; (b)
Fig. 12.9 Coupling beam shear force versus chord rotation (Vb-b) relationships: (a)
Structure P1-CWUPT; (b) Structure P2-PWUPT............................................................347
Fig. 12.10 Coupled wall idealized displaced shape .........................................................348
Fig. 12.11 Tension angle deformation .............................................................................350
Fig. 12.12 Wall contact region stresses: (a) side view; (b) beam end view.....................353
Fig. 12.13 Wall-contact region concrete confinement.....................................................355
Fig. 12.14 Wall pier flexural reinforcement: (a) trial section 1; (b) trial section 2; (c)
flexural steel area versus base shear resistance; (d) final wall cross section ...................357
Fig. 12.15 Concrete confinement at wall pier bases ........................................................360
Fig. 12.16 Wall pier post-tensioning steel reinforcement................................................365
Fig. 12.17 Concrete confinement at wall pier bases ........................................................369
xxii
Fig. 12.18 Prestress losses in wall pier post-tensioning bars...........................................372

Fig. 12.19 Coupled wall base shear force versus roof drift (F-) relationships: (a)
Structure P1-CWUPT; (b) Structure P2-PWUPT............................................................374
CHAPTER 13
Fig. 13.1 Coupled wall base shear force versus roof drift behaviors: (a) Structure P1CWUPT; (b) Structure P2-PWUPT .................................................................................382
Fig. 13.2 Wall pier flexural mild steel stresses at base of Structure P1-CWUPT: (a)
tension-side wall; (b) compression-side wall...................................................................386
Fig. 13.3a Wall pier post-tensioning bar stresses in Structure P2-PWUPT: tension-side
wall; (b) compression-side wall .......................................................................................387
Fig. 13.3b Wall pier post-tensioning bar stresses in Structure P2-PWUPT: compressionside wall ...........................................................................................................................388
Fig. 13.4a Wall pier post-tensioning bar stress-strain behaviors for Structure P2-PWUPT:
tension-side wall ..............................................................................................................389
Fig. 13.4b Wall pier post-tensioning bar stress-strain behaviors for Structure P2-PWUPT:
compression-side wall......................................................................................................390
Fig. 13.5 Wall pier base axial forces: (a) Structure P1-CWUPT; (b) Structure P2-PWUPT
..........................................................................................................................................391
Fig. 13.6 Components of wall pier base axial forces: (a) Structure P1-CWUPT; (b)
Fig. 13.7 Wall base shear forces: (a) Structure P1-CWUPT; (b) Structure P2-PWUPT.......
..........................................................................................................................................394
Fig. 13.8 Ratio of compression-side wall base shear force to total coupled wall base shear
forces: (a) Structure P1-CWUPT; (b) Structure P2-PWUPT ..........................................395
Fig. 13.9 Wall base moments: (a) Structure P1-CWUPT; (b) Structure P2-PWUPT .....396
Fig. 13.10 Degree of coupling: (a) Structure P1-CWUPT; (b) Structure P2-PWUPT ....398
Fig. 13.11 Wall pier base neutral axis (i.e., contact) depths: (a) Structure P1-CWUPT; (b)
xxiii
Fig. 13.12 Wall pier base extreme confined concrete compression strains: (a) Structure
P1-CWUPT; (b) Structure P2-PWUPT ...........................................................................400
Fig. 13.13 Coupling beam shear force versus chord rotation behaviors for Structure P1CWUPT............................................................................................................................402
Fig. 13.14 Coupling beam shear force versus chord rotation behaviors for Structure P2PWUPT ............................................................................................................................403
Fig. 13.15 Coupling beam axial forces and post-tensioning forces in Structure P1CWUPT............................................................................................................................405
Fig. 13.16 Coupling beam axial forces and post-tensioning forces in Structure P2PWUPT ............................................................................................................................406
Fig. 13.17 Coupling beam end neutral axis (i.e., contact) depths for Structure P1-CWUPT
..........................................................................................................................................408
Fig. 13.18 Coupling beam end neutral axis (i.e., contact) depths for Structure P2-PWUPT
..........................................................................................................................................409
Fig. 13.19 Coupling beam end extreme steel compression strains for Structure P1CWUPT............................................................................................................................410
Fig. 13.20 Coupling beam end extreme steel compression strains for Structure P2PWUPT ............................................................................................................................411
Fig. 13.21 Coupling beam tension angle force-deformation behaviors for Structure P1CWUPT............................................................................................................................412
Fig. 13.22 Coupling beam tension angle force-deformation behaviors for Structure P2PWUPT ............................................................................................................................413
Fig. 13.23 Cyclic coupled wall base shear force versus roof drift behaviors: (a) Structure
P1-CWUPT; (b) Structure P2-PWUPT ...........................................................................415
Fig. 13.24 Cyclic wall post-tensioning forces for Structure P2-PWUPT: (a) left-side wall
pier; (b) right-side wall pier; (c) total coupled wall.........................................................417
Fig. 13.25 Cyclic wall base axial forces for Structure P1-CWUPT: (a) left-side wall pier;
(b) right-side wall pier; (c) total coupled wall .................................................................418
Fig. 13.26 Cyclic wall base axial forces for Structure P2-PWUPT: (a) left-side wall pier;
(b) right-side wall pier; (c) total coupled wall .................................................................420
xxiv
Fig. 13.27 Cyclic wall pier base shear forces for Structure P1-CWUPT: (a) left-side wall
pier; (b) right-side wall pier .............................................................................................421
Fig. 13.28 Cyclic wall pier base shear forces for Structure P2-PWUPT: (a) left-side wall
pier; (b) right-side wall pier .............................................................................................421
Fig. 13.29 Cyclic wall base moments for Structure P1-CWUPT: (a) left-side wall pier
base moment; (b) right-side wall pier base moment; (c) coupling base moment; (d) total
coupled wall base moment...............................................................................................422
Fig. 13.30 Cyclic wall base moments for Structure P2-PWUPT: (a) left-side wall pier
base moment; (b) right-side wall pier base moment; (c) coupling base moment; (d) total
coupled wall base moment...............................................................................................423
Fig. 13.31 Degree of coupling: (a) Structure P1-CWUPT; (b) Structure P2-PWUPT ..........
..........................................................................................................................................425
Fig. 13.32 Cyclic wall pier base neutral axis (i.e., contact) depths and extreme confined
concrete compression strains for Structure P1-CWUPT: (a) left-side wall pier contact
depth; (b) right-side wall pier contact depth; (c) left-side wall pier concrete strains; (d)
right-side wall pier concrete strains .................................................................................426
depth; (b) right-side wall pier contact depth; (c) left-side wall pier concrete strains; (d)
right-side wall pier concrete strains .................................................................................427
Fig. 13.34 Cyclic coupling beam shear force versus chord rotation behaviors for Structure
P1-CWUPT ......................................................................................................................428
Fig. 13.35 Cyclic coupling beam shear force versus chord rotation behaviors for Structure
P2-PWUPT ......................................................................................................................429
Fig. 13.36 Cyclic coupling beam axial forces and post-tensioning forces in Structure P1CWUPT............................................................................................................................431
Fig. 13.37 Cyclic coupling beam axial forces and post-tensioning forces in Structure P2PWUPT ............................................................................................................................432
Fig. 13.38 Cyclic coupling beam end neutral axis (i.e., contact) depths for Structure P1CWUPT............................................................................................................................433
Fig. 13.39 Cyclic coupling beam end neutral axis (i.e., contact) depths for Structure P2PWUPT ............................................................................................................................434
xxv
Fig. 13.40 Cyclic coupling beam end extreme steel compression strains for Structure P1CWUPT............................................................................................................................435
Fig. 13.41 Cyclic coupling beam end extreme steel compression strains for Structure P2PWUPT ............................................................................................................................436
Fig. 13.42 Cyclic coupling beam angle force versus deformation behaviors for Structure
P1-CWUPT ......................................................................................................................438
Fig. 13.43 Cyclic coupling beam angle force versus deformation behaviors for Structure
P2-PWUPT ......................................................................................................................439
Fig. 13.44 Coupled wall base shear force versus roof drift behavior of Structure P3CWEMB: (a) monotonic loading; (b) cyclic loading ......................................................440
CHAPTER 14
Fig. 14.1a Acceleration time-histories of SAC design-level ground motions (LA01LA10)...............................................................................................................................444
Fig. 14.1b Acceleration time-histories of SAC design-level ground motions (LA11LA20)...............................................................................................................................445
Fig. 14.2a 5%-damped linear-elastic pseudo-acceleration response spectra of SAC
design-level ground motions (LA01-LA10) ....................................................................446
Fig. 14.2b 5%-damped linear-elastic pseudo-acceleration response spectra of SAC
design-level ground motions (LA11-LA20) ....................................................................447
Fig. 14.3a Acceleration time-histories of SAC survival-level ground motions (LA21LA30)...............................................................................................................................449
Fig. 14.3b Acceleration time-histories of SAC survival-level ground motions (LA31LA40)...............................................................................................................................450
survival-level ground motions (LA21-LA30)..................................................................451
survival-level ground motions (LA31-LA40)..................................................................452
xxvi
CHAPTER 15
Fig. 15.1 Roof drift time history for Structure P1-CWUPT ............................................458
Fig. 15.2 Peak roof drift demands for Structure P1-CWUPT: (a) survival level; (b) design
level..................................................................................................................................459
Fig. 15.3 Inter-story drift time histories for Structure P1-CWUPT.................................461
Fig. 15.4 Peak inter-story drift demands for Structure P1-CWUPT: (a) survival level; (b)
design level ......................................................................................................................463
Fig. 15.5 Floor/roof acceleration time histories for Structure P1-CWUPT.....................465
Fig. 15.6 Peak floor/roof acceleration demands for Structure P1-CWUPT: (a) survival
level; (b) design level.......................................................................................................466
Fig. 15.7 Wall pier base axial force time histories for Structure P1-CWUPT: (a) left-side
wall; (b) right-side wall....................................................................................................469
Fig. 15.8 Wall pier base axial force time histories for Structure P1-CWUPT: (a) left-side
Fig. 15.9 Peak wall pier base axial force demands for Structure P1-CWUPT: (a) survival
Fig. 15.10 Coupled wall base shear force time history for Structure P1-CWUPT..........472
Fig. 15.11 Peak coupled wall base shear force demands for Structure P1-CWUPT: (a)
survival level; (b) design level.........................................................................................473
Fig. 15.12 Wall pier base shear force time histories for Structure P1-CWUPT: (a) leftside wall; (b) right-side wall ............................................................................................475
Fig. 15.13 Peak wall pier base shear force demands for Structure P1-CWUPT: (a)
Fig. 15.14 Wall pier base moment time histories for Structure P1-CWUPT: (a) left-side
Fig. 15.15 Peak wall pier base moment demands for Structure P1-CWUPT: (a) survival
Fig. 15.16 Coupled wall base moment time history for Structure P1-CWUPT ..............482
xxvii
Fig. 15.17 Peak coupled wall base moment demands for Structure P1-CWUPT: (a)
Fig. 15.18 Wall pier base extreme confined concrete compression strain time histories for
Structure P1-CWUPT: (a) left-side wall; (b) right-side wall...........................................485
Fig. 15.19 Peak wall base confined concrete compression strain demands for Structure
P1-CWUPT: (a) survival level; (b) design level..............................................................486
Fig. 15.20 Coupling beam chord rotation time histories for Structure P1-CWUPT........488
Fig. 15.21 Peak coupling beam chord rotation demands for Structure P1-CWUPT: (a)
Fig. 15.22 Coupling beam midspan axial force time histories for Structure P1-CWUPT.....
..........................................................................................................................................492
Fig. 15.23 Peak coupling beam midspan axial force demands for Structure P1-CWUPT:
(a) survival level; (b) design level ...................................................................................493
Fig. 15.24 Coupling beam post-tensioning force time histories for Structure P1-CWUPT ..
..........................................................................................................................................496
Fig. 15.25 Peak coupling beam post-tensioning force demands for Structure P1-CWUPT:
Fig. 15.26 Coupling beam shear force time histories for Structure P1-CWUPT ............499
Fig. 15.27 Peak coupling beam shear force demands for Structure P1-CWUPT: (a)
Fig. 15.28 Coupling beam shear slip capacity and shear force demand time histories for
Structure P1-CWUPT ......................................................................................................503
Fig. 15.29 Minimum coupling beam shear slip capacity to shear force demand ratios for
Structure P1-CWUPT: (a) survival level; (b) design level ..............................................505
Fig. 15.30 Coupling beam connection angle deformation time histories for Structure P1CWUPT............................................................................................................................507
Fig. 15.31 Peak coupling beam connection angle tensile deformation demands for
Structure P1-CWUPT: (a) survival level; (b) design level ..............................................508
Fig. 15.32 Roof drift time history for Structure P2-PWUPT...........................................510
xxviii
Fig. 15.33 Peak roof drift demands for Structure P2-PWUPT: (a) survival level; (b)
design level ......................................................................................................................512
Fig. 15.34 Inter-story drift time histories for Structure P2-PWUPT ...............................514
Fig. 15.35 Peak inter-story drift demands for Structure P2-PWUPT: (a) survival level; (b)
design level ......................................................................................................................515
Fig. 15.36 Floor/roof acceleration time histories for Structure P2-PWUPT ...................517
Fig. 15.37 Peak floor/roof acceleration demands for Structure P2-PWUPT: (a) survival
Fig. 15.38 Total coupled wall post-tensioning force time histories for Structure P2PWUPT: (a) LA38; (b) LA33 ..........................................................................................522
Fig. 15.39 Minimum and maximum total coupled wall post-tensioning forces for
Structure P2-PWUPT: (a) survival level; (b) design level...............................................523
Fig. 15.40 Wall pier base axial force time histories for Structure P2-PWUPT: (a) left-side
Fig. 15.41 Wall pier base axial force time histories for Structure P2-PWUPT: (a) left-side
Fig. 15.42 Peak wall pier base axial force demands for Structure P2-PWUPT: (a) survival
Fig. 15.43 Coupled wall base shear force time history for Structure P2-PWUPT ..........529
Fig. 15.44 Peak coupled wall base shear force demands for Structure P2-PWUPT: (a)
Fig. 15.45 Wall pier base shear force time histories for Structure P2-PWUPT: (a) left-side
Fig. 15.46 Peak wall pier base shear force demands for Structure P2-PWUPT: (a)
Fig. 15.47 Coupled wall base shear slip capacity and base shear force demand time
histories for Structure P2-PWUPT...................................................................................536
Fig. 15.48 Minimum coupled wall base shear slip capacity to base shear force demand
ratio for Structure P2-PWUPT: (a) survival level; (b) design level.................................537
xxix
Fig. 15.49 Wall pier base moment time histories for Structure P2-PWUPT: (a) left-side
Fig. 15.50 Peak wall pier base moment demands for Structure P2-PWUPT: (a) survival
Fig. 15.51 Coupled wall base moment time history for Structure P2-PWUPT...............542
Fig. 15.52 Peak coupled wall base moment demands for Structure P2-PWUPT: (a)
Fig. 15.53 Wall pier base extreme confined concrete compression strain time histories for
Structure P2-PWUPT: (a) left-side wall; (b) right-side wall ...........................................545
Fig. 15.54 Peak wall base confined concrete compression strain demands for Structure
P2-PWUPT: (a) survival level; (b) design level ..............................................................546
Fig. 15.55 Coupling beam chord rotation time histories for Structure P2-PWUPT ........548
Fig. 15.56 Peak coupling beam chord rotation demands for Structure P2-PWUPT: (a)
Fig. 15.57 Coupling beam midspan axial force time histories for Structure P2-PWUPT .....
..........................................................................................................................................552
Fig. 15.58 Peak coupling beam midspan axial force demands for Structure P2-PWUPT:
Fig. 15.59 Coupling beam post-tensioning force time histories for Structure P2-PWUPT...
..........................................................................................................................................556
Fig. 15.60 Peak coupling beam post-tensioning force demands for Structure P2-PWUPT:
Fig. 15.61 Coupling beam shear force time histories for Structure P2-PWUPT.............559
Fig. 15.62 Peak coupling beam shear force demands for Structure P2-PWUPT: (a)
Fig. 15.63 Coupling beam shear slip capacity and shear force demand time histories for
Fig. 15.64 Minimum coupling beam shear slip capacity to shear force demand ratios for
xxx
Fig. 15.65 Coupling beam connection angle deformation time histories for Structure P2PWUPT ............................................................................................................................566
Fig. 15.66 Peak coupling beam connection angle tensile deformation demands for
Fig. 15.67 Peak roof drift demand for uncoupled wall systems: (a) Structure P4-CWUNC;
(b) Structure P5-PWUNC ................................................................................................570
xxxi
TABLES
CHAPTER 3
Table 3.1 Prototype subassemblage coupling beam properties .........................................48
Table 3.2 Prototype subassemblage beam post-tensioning properties...............................49
Table 3.3 Prototype subassemblage beam-to-wall connection top and seat angle
properties............................................................................................................................49
Table 3.4 Prototype subassemblage wall properties ..........................................................49
Table 3.5 Prototype subassemblage material properties....................................................51
Table 3.6 Wall-contact element fiber discretization ..........................................................58
Table 3.7 Beam fiber element discretization .....................................................................62
Table 3.8 Prototype subassemblage beam-to-wall connection top and seat angle
design capacities ................................................................................................................65
CHAPTER 4
Table 4.1 Parametric study of coupled wall subassemblages ............................................87
CHAPTER 6
Table 6.1 CIP-UPT wall properties .................................................................................106
Table 6.2 PRE-UPT wall properties ................................................................................107
Table 6.3 Walls CIP-UPT and PRE-UPT coupling beam properties ..............................108
Table 6.4 Walls CIP-UPT and PRE-UPT beam post-tensioning properties....................108
Table 6.5 Walls CIP-UPT and PRE-UPT beam-to-wall connection top and seat
angle properties................................................................................................................109
xxxii
Table 6.6 Wall material properties...................................................................................109

Table 6.7 CIP-UPT system wall-height element discretization.......................................116
Table 6.8 PRE-UPT system wall-height element discretization......................................120
CHAPTER 7
Table 7.1 Wall Pier Gravity loads for CIP-UPT and PRE-UPT systems ........................126
Table 7.2 Parametric CIP-UPT systems ..........................................................................159
Table 7.3 Parametric PRE-UPT systems .........................................................................159
Table 7.4 Degree of coupling of parametric CIP-UPT systems ......................................163
Table 7.5 Degree of coupling of parametric PRE-UPT systems .....................................163
Table 7.6 Second floor beam initial axial force divided by total initial PT force in
parametric CIP-UPT systems...........................................................................................164
Table 7.7 Second floor beam initial axial force divided by total initial PT force in
parametric PRE-UPT systems..........................................................................................164
CHAPTER 9
Table 9.1 Subassemblage test specimens.........................................................................200
Table 9.2 Beam steel (W1068) material properties .......................................................213
Table 9.3 Cover plate steel material properties ...............................................................215
Table 9.4 Angle steel material properties ........................................................................218
Table 9.5 Spiral wire steel material properties ................................................................220
Table 9.6 Post-tensioning strand properties.....................................................................223
Table 9.7 Concrete properties ..........................................................................................224
xxxiii
CHAPTER 11
Table 11.1 MDOF system properties...............................................................................295
Table 11.2 Idealized MDOF BP model properties ..........................................................296
Table 11.3 Equivalent SDOF BP model properties .........................................................298
Table 11.4 Regression coefficients a and b for BP models (based on IND spectra) ......308
Table 11.5 Regression coefficients a and b for BP models (based on DES spectra) .....309
CHAPTER 12
Table 12.1 General properties of prototype structures.....................................................320
Table 12.2 Seismic weights and gravity loads.................................................................323
Table 12.3 Acceleration response spectra........................................................................325
Table 12.4 Linear-elastic stiffnesses................................................................................327
Table 12.5 Linear-elastic modal analysis results .............................................................328
Table 12.6 Displacement demand relationships ..............................................................329
Table 12.7 Displacement ductility demands and strength ratios .....................................329
Table 12.8 Design base shear demands ...........................................................................330
Table 12.9 Peak roof drift demands.................................................................................331
Table 12.10 Peak inter-story drift demands.....................................................................332
Table 12.11 Diagonal tension at wall pier bases..............................................................333
Table 12.12 Design coupling beam shear force demands................................................335
Table 12.13 Effect of ar on coupling beam behavior .....................................................338
Table 12.14 Components of coupling beam design forces ..............................................339
Table 12.15 Beam post-tensioning ..................................................................................342
xxxiv
Table 12.16 Top and seat angles......................................................................................346

Table 12.17 Coupling beam shear force versus rotation relationships ............................348
Table 12.18 Coupling beam rotation demands ................................................................349
Table 12.19 Tension angle deformation demands ...........................................................351
Table 12.20 Shear slip at beam-to-wall interfaces...........................................................352
Table 12.21 Wall-contact regions ....................................................................................354
Table 12.22 Wall pier flexural reinforcement..................................................................358
Table 12.23 Concrete confinement at wall pier bases .....................................................361
Table 12.25 Shear slip at coupled wall base ....................................................................363
Table 12.26 Wall pier post-tensioning steel reinforcement .............................................365
Table 12.27 Yielding of wall pier post-tensioning steel ..................................................367
Table 12.28 Concrete confinement at wall pier bases .....................................................369
Table 12.30 Shear slip at coupled wall base ....................................................................373
Table 12.31 Idealized F- behaviors for Structures P1-CWUPT and P2-PWUPT.........375
Table 12.32 Seismic design criteria for Structures P1-CWUPT and P2-PWUPT...........376
CHAPTER 13
Table 13.1 Wall pier gravity loads in Structures P1-CWUPT and P2-PWUPT..............380
Table 13.2 Comparisons between estimation and analytical results................................385
xxxv
CHAPTER 14
Table 14.1 SAC design-level ground motions records LA01-LA20 ...............................448
Table 14.2 SAC Survival-level ground motions records LA21-LA40 ............................453
CHAPTER 15
Table 15.1 Period and damping ratios for prototype structures.......................................456
Table 15.2 Peak roof drift demands for Structure P1-CWUPT under survival level
ground motions ................................................................................................................457
Table 15.3 Peak roof drift demands for Structure P1-CWUPT under design level
ground motions ................................................................................................................457
Table 15.4 Peak inter-story drift demands for Structure P1-CWUPT under survival
level ground motions........................................................................................................460
Table 15.5 Peak inter-story drift demands for Structure P1-CWUPT under design
Table 15.6 Peak floor/roof acceleration demands for Structure P1-CWUPT under
survival level ground motions..........................................................................................464
Table 15.7 Peak floor/roof acceleration demands for Structure P1-CWUPT under
design level ground motions ............................................................................................464
Table 15.8 Peak wall pier base tensile axial force demands for Structure P1-CWUPT
under survival level ground motions................................................................................467
Table 15.9 Peak wall pier base tensile axial force demands for Structure P1-CWUPT
under design level ground motions ..................................................................................467
Table 15.10 Peak wall pier base compressive axial force demands for Structure
P1-CWUPT under survival level ground motions ...........................................................467
Table 15.11 Peak wall pier base compressive axial force demands for Structure
P1-CWUPT under design level ground motions .............................................................468
Table 15.12 Peak coupled wall base shear force demands for Structure P1-CWUPT
xxxvi
Table 15.13 Peak coupled wall base shear force demands for Structure P1-CWUPT
Table 15.14 Peak wall pier base shear force demands for Structure P1-CWUPT
Table 15.15 Peak wall pier base shear force demands for Structure P1-CWUPT
Table 15.16 Peak wall pier base shear force distribution ratios for Structure
Table 15.17 Peak wall pier shear force distribution ratios for Structure P1-CWUPT
Table 15.18 Peak wall pier base moment demands for Structure P1-CWUPT under
Table 15.19 Peak wall pier base moment demands for Structure P1-CWUPT under
Table 15.20 Peak coupled wall base moment demands for Structure P1-CWUPT
Table 15.21 Peak coupled wall base moment demands for Structure P1-CWUPT
Table 15.22 Peak wall pier base confined concrete compression strain demands for
Structure P1-CWUPT under survival level ground motions ...........................................484
Structure P1-CWUPT under design level ground motions.............................................484
Table 15.24 Peak coupling beam chord rotation demands for Structure P1-CWUPT
Table 15.25 Peak coupling beam chord rotation demands for Structure P1-CWUPT
Table 15.26 Peak coupling beam midspan axial force demands for Structure
xxxvii
Table 15.28 Peak coupling beam post-tensioning force demands for Structure
Table 15.30 Peak coupling beam shear force demands for Structure P1-CWUPT
Table 15.31 Peak coupling beam shear force demands for Structure P1-CWUPT
Table 15.32 Minimum coupling beam shear slip capacity-demand ratios for
Structure P1-CWUPT under survival level ground motions ...........................................501
Table 15.33 Minimum coupling beam slip capacity-demand ratios for Structure
Table 15.34 Peak angle tensile deformation demands for Structure P1-CWUPT
Table 15.35 Peak angle tensile deformation demands for Structure P1-CWUPT
Table 15.36 Peak roof drift for Structure P2-PWUPT under survival level ground
motions.............................................................................................................................509
Table 15.37 Peak roof drift for Structure P2-PWUPT under design level ground
motions.............................................................................................................................509
Table 15.38 Peak inter-story drift for Structure P2-PWUPT under survival level
ground motions ................................................................................................................513
Table 15.39 Peak inter-story drift for Structure P2-PWUPT under design level
ground motions ................................................................................................................513
Table 15.40 Peak floor/roof acceleration for Structure P2-PWUPT under survival
Table 15.41 Peak floor/roof acceleration for Structure P2-PWUPT under design
Table 15.42 Minimum total coupled wall post-tensioning forces for Structure
P2-PWUPT under survival level ground motions ...........................................................520
xxxviii
Table 15.43 Minimum total coupled wall post-tensioning forces for Structure
P2-PWUPT under design level ground motions..............................................................520
Table 15.44 Maximum total coupled wall post-tensioning forces for Structure
Table 15.45 Maximum total coupled wall post-tensioning forces for Structure
Table 15.46 Minimum wall pier base compression axial force demands for Structure
Table 15.47 Minimum wall pier base compression axial force demands for Structure
Table 15.48 Maximum wall pier base compression axial force demands for Structure
Table 15.49 Maximum wall pier base compression axial force demands for Structure
Table 15.50 Peak coupled wall base shear force demands for Structure P2-PWUPT
Table 15.51 Peak coupled wall base shear force demands for Structure P2-PWUPT
Table 15.52 Peak wall pier base shear force demands for Structure P2-PWUPT
Table 15.53 Peak wall pier base shear force demands for Structure P2-PWUPT
Table 15.56 Minimum coupled wall base shear slip capacity-demand ratios for
Structure P2-PWUPT under survival level ground motions............................................535
Table 15.57 Minimum coupled wall base shear slip capacity-demand ratios for
Structure P2-PWUPT under design level ground motions ..............................................535
xxxix
Table 15.58 Peak wall pier base moment demands for Structure P2-PWUPT under
Table 15.59 Peak wall pier base moment demands for Structure P2-PWUPT under
Table 15.60 Peak coupled wall base moment demands for Structure P2-PWUPT
Table 15.61 Peak coupled wall base moment demands for Structure P2-PWUPT
Structure P2-PWUPT under design level ground motions ..............................................544
Table 15.64 Peak coupling beam chord rotation demands for Structure P2-PWUPT
Table 15.65 Peak coupling beam chord rotation demands for Structure P2-PWUPT
Table 15.66 Peak coupling beam midspan axial forces demands for Structure
P2-PWUPT under survival level ground motion .............................................................551
Table 15.70 Peak coupling beam shear force demands for Structure P2-PWUPT
Table 15.71 Peak coupling beam shear force demands for Structure P2-PWUPT
Table 15.72 Minimum coupling beam shear slip capacity-demand ratios for
xl
Table 15.73 Minimum coupling beam slip capacity-demand ratios for Structure
Table 15.74 Peak angle tensile deformation demands for Structure P2-PWUPT
Table 15.75 Peak angle tensile deformation in Structure P2-PWUPT under design
Table 15.76 Peak roof drift demands for Structure P4-CWUNC under survival
Table 15.77 Peak roof drift demands for Structure P5-PWUNC under survival
xli
SYMBOLS
a
abp
absp
abt
af
awp
awsp
Aa
Aaw
Ab
Abc
Abf
Abfc
Abg
Abp
Abw
Ac
Aw
Awc
Awp
Aws
b
bbf
bc
c
cb
cb,pty
cb,sof
ccw
ccws
ccwu
ccwy
ctw
ctws
ctwy
[C]
Ca
Casx
Cax
= regression coefficient for R--T relationship

= area of one coupling beam post-tensioning strand
= area of spiral wire in wall pier contact region adjacent to coupling beam
= area of coupling beam post-tensioning multi-strand tendon
= area of steel fiber
= area of wall pier post-tensioning bar
= area of spiral wire at wall pier base
= gross area of angle leg gross cross section
= total area of post-tensioning steel used in angle-to-wall pier connection
= area of coupling beam cross section
= area of coupling beam cross-section including flange cover plates (if any)
= area of coupling beam flange
= area of coupling beam flange plus flange cover plate
= shear area of coupling beam cross section
= total area of coupling beam post-tensioning steel per floor/roof level
= effective area of weld for coupling beam cover plate
= area of coupling beam flange cover plate
= area of wall pier gross cross section
= net concrete area of wall pier cross section
= total area of wall pier post-tensioning steel reinforcement
= total area of wall pier boundary flexural mild steel reinforcement
= width of coupling beam flange
= width of coupling beam flange cover plate
= neutral axis (contact) depth of coupling beam at beam-to-wall pier interface
= value of cb at coupling beam PT-yielding state
= value of cb at coupling beam softening state
= neutral axis depth of compression-side wall pier at base
= value of ccw at coupled wall softening state
= value of ccw at coupled wall ultimate state
= value of ccw at coupled wall PT-yielding state
= neutral axis depth of tension-side wall pier at base
= value of ctw at coupled wall softening state
= value of ctw at coupled wall PT-yielding state
= damping matrix for structure
= force in compression angle
= slip force of angle-to-beam connection bolt group
= compression force in angle for loading parallel to coupling beam
xlii
Cb
Cb,pty
Cb,sof
Ceq
Cvx
Cwc
CC
CCC
CM
CS
CU
CUC
dab
dabb
dabw
db
dbc
dbp
dbsp
dbw
dwsp
dwp
dwp1
D
Dbsp
Dm
Dwsp
DOC
DOCws
DOCwu
e
eacc
eapp
E
Ea
Eb
Ebp
Ec
Ehc
= total compression force at end of coupling beam

= value of Cb at coupling beam PT-yielding state
= value of Cb coupling beam softening state
= generalized damping for first (fundamental) vibration mode ({1}T[C]{1})
= factor for distribution of Qsd over structure height
= compression force in wall pier contact region adjacent to coupling beam
= confined concrete fiber
= compression-only confined concrete fiber
= building center of mass at floor/roof level
= building center of stiffness at floor/roof level
= unconfined concrete fiber
= compression-only unconfined concrete fiber
= diameter of angle-to-beam or angle-to-column/wall pier connector
= diameter of angle-to-beam connector
= diameter of angle-to-column/wall pier connector
= depth of coupling beam
= depth of coupling beam including flange cover plates (if any)
= diameter of coupling beam post-tensioning strand
= diameter of spiral wire in wall pier contact region adjacent to coupling beam
= depth of coupling beam web
= diameter of spiral wire at wall pier base
= distance of wall pier post-tensioning bar from compression end of wall pier
= value of dwp for first post-tensioning bar that yields
= unfactored design dead loads
= center-to-center diameter of spiral in wall pier contact region adjacent to
coupling beam
= coefficient to estimate Qh,max
= center-to-center diameter of spiral at wall pier base
= degree of coupling
= value of DOC at coupled wall softening state
= value of DOC at coupled wall ultimate state
= eccentricity between floor/roof center of mass (CM) and center of stiffness
(CS)
= accidental eccentricity between floor/roof center of mass (CM) and center of
stiffness (CS)
= applied eccentricity between floor/roof center of mass (CM) and center of
stiffness (CS)
= effect of seismic loads
= modulus of elasticity of angle steel
= modulus of elasticity of coupling beam steel
= modulus of elasticity of coupling beam post-tensioning strand
= modulus of elasticity of wall pier concrete
= hysteretic energy dissipation during a base shear force versus roof
displacement cycle
xliii
Es
Esp
Ewp
Ews
f
fam
fau
fawpi
fawpu
fay
fbi
fbm
fbpi
fbpm
fbpu
fbpy
fbu
fby
f c'
fcc
fccw
fcl
fct
fcwp
fEXX
frm
fru
fry
fsm
fspm
fspu
fspy
fsy
ftwp
fwpi
fwpm
fwpr
fwpu
fwpy
= modulus of elasticity of steel

= modulus of elasticity of spiral steel
= modulus of elasticity of wall pier post-tensioning bar
= modulus of elasticity of wall pier mild steel reinforcement
= lateral forces applied at floor/roof levels
= maximum (peak) strength of angle steel
= 0.85fam
= initial stress in angle-to-wall pier connection strands
= maximum (peak) strength of angle-to-wall pier connection strands
= yield strength of angle steel
= initial compressive stress in coupling beam (including flange cover plates, if
any) due to post-tensioning before application of lateral loads
= maximum (peak) strength of coupling beam steel
= initial stress in coupling beam post-tensioning strands before application of
lateral loads
= maximum (peak) strength of coupling beam post-tensioning strands
= fbpm
= yield (limit of proportionality) strength of coupling beam post-tensioning
strands
= 0.85fbm
= yield strength of coupling beam steel
= maximum (peak) compressive strength of unconfined concrete
= maximum (peak) compressive strength of confined concrete
= value of fcc for wall pier contact region adjacent to coupling beam
= linear limit (limit of proportionality) stress of confined concrete in
compression
= tensile strength of unconfined and confined concrete
= stress in post-tensioning bar of compression-side wall pier
= weld electrode strength
= maximum (peak) strength of coupling beam flange cover plate steel
= 0.85frm
= yield strength of coupling beam flange cover plate steel
= maximum (peak) strength of steel
= maximum (peak) strength of spiral steel
= 0.85fspm
= yield strength of spiral steel
= yield strength of steel
= stress in post-tensioning bar of tension-side wall pier
= initial stress in wall pier post-tensioning bar before application of lateral loads
= maximum (peak) strength of wall pier post-tensioning bars
= residual stress of wall pier post-tensioning bar after losses under cyclic
loading
= fwpm
= yield (limit of proportionality) strength of wall pier post-tensioning bar
xliv
fwp1
fwp2
fwp3
fwp4
fwp5
fwp6
fwp7
fwp8
fwse
fwsm
fwsy
fx
f
F
Fa
Fbe
Fbp
Fcw
Fdes
Felas
Fep
Feq
Feq,be
Feq,bp
Feq,ep
Fg,ss
Fnlin
Fp,ss
Fs,ss
Ftw
Fv
Fwbs
Fw,dt
Fws
Fwsi
= stress in wall pier post-tensioning bar layer 1

= extreme tension stress of flexural mild steel reinforcement at compressionside wall pier base
= maximum (peak) strength of wall pier mild steel reinforcement
= yield strength of wall pier mild steel reinforcement
= lateral force applied at floor/roof level x
= damping adjustment coefficient
= base shear force
= site coefficient for spectral response acceleration at short periods
= yield strength of BE component of equivalent nonlinear single-degree-offreedom BP hysteresis model
= yield strength of equivalent nonlinear single-degree-of-freedom BP hysteresis
model
= base shear force in compression-side wall pier
= base shear force at first significant yield of structure
= linear-elastic lateral force demand for structure
= yield strength of EP component of equivalent nonlinear single-degree-offreedom BP hysteresis model
= force in equivalent single-degree-of-freedom system
= yield strength of BE component of equivalent nonlinear single-degree-offreedom BP hysteresis model
= yield strength of equivalent nonlinear single-degree-of-freedom BP hysteresis
model
= yield strength of EP component of equivalent nonlinear single-degree-offreedom BP hysteresis model
= shear friction capacity at base of coupled wall due to gravity loads
= lateral force corresponding to unlin
= shear friction capacity at base of coupled wall due to wall pier post-tensioning
steel reinforcement
= shear friction capacity at base of coupled wall due to wall pier mild steel
reinforcement
= base shear force in tension-side wall pier
= site coefficient for spectral response acceleration at 1 sec. period
= wall pier shear force below floor level
= diagonal tension shear force capacity for wall pier
= coupled wall base shear force at coupled wall softening state
= Fwbs minus Fwts
xlv
Fw,ss
Fwts
Fwu
Fw,unc
Fwy
Fy
g
gb,pty
G
Gb
Gc
hb
hs
hw
hwp
H
H1
I
Ia
Ib
Iw
ka
kbe
kep
K
[K]
Kaixc
Kaixt
Kaiy
Kaxs
Kbi
[Kbi]
[Kci]
Keq
Kli
Ksi
Kwi
la
lal
lb
lbe
lb,cr
lb,eff
lbpu
= coupled wall base shear slip capacity

= wall pier shear force above floor level
= coupled wall base shear force at coupled wall ultimate state
= uncoupled wall base shear force (one wall pier)
= coupled wall base shear force at coupled wall PT-yielding state
= base shear force at global yield of structure
= acceleration of gravity
= width of gap opening at coupling beam centerline at beam PT-yielding state
= wall floor and roof gravity (dead plus live) forces
= shear modulus of coupling beam steel
= shear modulus of wall pier concrete
= height of coupling beam from wall base
= height of story
= height of wall
= plastic hinge height at base of compression-side wall pier
= resultant height of coupled wall lateral forces
= resultant height of coupled wall first mode inertial forces
= occupancy importance factor (ICC 2000)
= moment of inertia of angle leg cross section
= moment of inertia of coupling beam cross section
= moment of inertia of wall pier gross cross section
= angle fillet length measured from angle heel to toe of angle fillet
= linear-elastic stiffness of single-degree-of-freedom BE hysteresis model
= linear-elastic stiffness of single-degree-of-freedom EP hysteresis model
= linear-elastic stiffness
= stiffness matrix of structure
= initial compression stiffness of angle for loading parallel to coupling beam
= initial tension stiffness of angle for loading parallel to coupling beam
= initial stiffness of angle for loading perpendicular to coupling beam
= shooting stiffness of angle for loading parallel to coupling beam
= initial linear-elastic lateral stiffness of a coupling beam subassemblage
= initial linear-elastic stiffness matrix for coupling beam
= initial linear-elastic stiffness matrix of wall pier
= generalized stiffness for first (fundamental) vibration mode ({1}T[K]{1})
= initial linear-elastic lateral stiffness of lateral load resisting system
= initial linear-elastic lateral stiffness of entire building
= initial linear-elastic lateral stiffness of coupled wall
= length of angle
= length of angle leg
= length of coupling beam
= embedment length of coupling beam
= length of first coupling beam fiber element or segment adjacent to wall pier
= effective length for embedded coupling beam
= unbonded length of coupling beam post-tensioning tendons
xlvi
lc
lgh
lgv
lg1
lg2
lw
lwc
lwpu
L
Lc
Leq
meff
meff,2
mw
M
[M]
Ma
Map
Ma0
Mb
Mb,bfy
Mb,dec
Mbp
Mb,pty
Mb,sof
Mby
Mcw
Mcws
Mcwu
Mcwy
Meff
Meff,2
Meq
Mtw
Mtws
Mtwu
Mtwy
= length of coupling beam flange cover plate

= angle-to-beam connection gage length measured from heel of angle to
centroid of angle-to-beam connector group
= angle-to-column/wall pier connection gage length measured from heel of
angle to center of innermost angle-to-wall pier connectors
= length of angle vertical leg (leg perpendicular to coupling beam) assumed to
act like a cantilever
= effective gage length of angle vertical leg (leg perpendicular to coupling
beam) between two plastic hinges
= length of wall pier
= length at each end of a wall pier where concrete confinement is provided
= unbonded length of wall pier post-tensioning bar
= unfactored design live loads
= arm for Ntwb-Ncwb axial force couple between tension-side and compressionside wall pier centroids
= earthquake excitation factor ({1}T[M]{1})
= effective first mode mass for coupled wall (meff=Meff/nw)
= effective second mode mass for coupled wall
= seismic mass for coupled wall
= seismic mass for entire building
= lumped diagonal mass matrix for structure
= bending moment in angle leg
= plastic moment in angle vertical leg (leg perpendicular to coupling beam)
considering shear-flexure interaction
= plastic moment in angle vertical leg (leg perpendicular to coupling beam)
without considering shear-flexure interaction
= coupling beam end moment
= value of Mb at coupling beam flange yielding state
= value of Mb at coupling beam decompression state
= plastic moment of coupling beam cross section
= value of Mb at coupling beam PT-yielding state
= value of Mb at coupling beam softening state
= yield moment of coupling beam cross section
= compression-side wall pier base moment
= value of Mcw at coupled wall softening state
= value of Mcw at coupled wall ultimate state
= value of Mcw at coupled wall PT-yielding state
= effective first mode mass for entire building (Meff=Leq2/Meq)
= effective second mode mass for entire building
= generalized mass for first (fundamental) mode of vibration ({1}T[M]{1})
= tension-side wall pier base moment
= value of Mtw at coupled wall softening state
= value of Mtw at coupled wall ultimate state
= value of Mtw at coupled wall PT-yielding state
xlvii
Mw
Mwb
Mwbs
Mws
Mwu
Mw,unc
Mwy
Mx
MIV
n
nab
nabb
nabw
nb
nbp
nbt
nw
nwp
Nap
Nb
Nbi
Ncw
Ncwb
Ncwg
Ncwp
Ncws
Ncwu
Ncwy
Ntw
Ntwb
Ntwg
Ntwp
Ntws
Ntwu
Ntwy
Pb
Pbi
Pbj
Pbm
Pb,pty
Pbu
= total coupled wall base moment

= contribution of coupling beam shear forces to Mw
= contribution of coupling beam shear forces to Mw at coupled wall softening
state
= value of Mw at coupled wall softening state
= value of Mw at coupled wall ultimate state
= base moment of uncoupled wall (one wall pier)
= value of Mw at coupled wall PT-yielding state
= portion of M assigned to floor/roof level x
= maximum incremental velocity of ground motion record
= number of vibration modes
= number of angle-to-beam or angle-to-column/wall pier connectors
= number of angle-to-beam connectors
= number of angle-to-column/wall pier connectors
= number of coupling beams over wall height
= number of strands per coupling beam post-tensioning tendon
= number of coupling beam post-tensioning tendons
= number of coupled walls in a building
= number of post-tensioning bars in wall pier
= axial force in angle vertical leg (leg perpendicular to coupling beam) acting
with Map and Vap
= axial force in coupling beam at midspan
= initial axial force in coupling beam before application of lateral loads
= axial force in compression-side wall pier at base
= component of Ncw due to coupling beam shear forces
= component of Ncw due to gravity loads
= component of Ncw due to wall pier post-tensioning force
= value of Ncw at coupled wall softening state
= value of Ncw at coupled wall ultimate state
= value of Ncw at coupled wall PT-yielding state
= axial force in tension-side wall pier at base
= component of Ntw due to coupling beam shear forces
= component of Ntw due to gravity loads
= component of Ntw due to wall pier post-tensioning force
= value of Ntw at coupled wall softening state
= value of Ntw at coupled wall ultimate state
= value of Ntw at coupled wall PT-yielding state
= total force in coupling beam post-tensioning tendons
= total initial force in coupling beam post-tensioning tendons before application
of lateral loads
= total force in coupling beam PT elements before losses
= maximum total force in coupling beam post-tensioning tendons
= value of Pb at coupling beam PT-yielding state
= total maximum strength of coupling beam post-tensioning tendons
xlviii
Pby
Pcw
Pcwi
Pcws
Pcwy
Ptw
Ptwi
Ptws
Ptwy
Pwi
Pwu
PGA
PGAd
PGAs
Qbd
Qbd,a
Qbd,p
Qb,max
Qdd
Qds
QE
Qh,max
Qsd
Qwd
Qw,dt
Qw,max
Q1,max
{R}
Rd
Rn
R
Rd
Rs
sbsp
sl
sw
swsp
= total yield strength of coupling beam post-tensioning tendons

= total force in compression-side wall pier post-tensioning bars
= total initial force in compression-side wall pier post-tensioning bars before
application of lateral loads
= value of Pcw at coupled wall softening state
= value of Pcw at coupled wall softening state at coupled wall PT-yielding state
= total force in tension-side wall pier post-tensioning bars
= total initial force in tension-side wall pier post-tensioning bars before
application of lateral loads
= value of Ptw at coupled wall softening state
= value of Ptw at coupled wall PT-yielding state
= total initial force in wall pier post-tensioning bars
= maximum post-tensioning tendon force based on design maximum strength of
post-tensioning steel (Pwu=awpfwpu)
= peak ground acceleration
= design-level peak ground acceleration
= survival-level peak ground acceleration
= design coupling beam shear force demand
= portion of Qbd resisted by top and seat angles
= portion of Qbd resisted by coupling beam post-tensioning force
= maximum coupling beam shear force demand
= design-level base shear force demand for entire building
= survival-level base shear force demand for entire building
= effect of horizontal seismic loads
= higher mode component of Qw,max
= design base shear force demand for entire building
= coupled wall design base shear force demand
= wall pier diagonal tension shear force demand
= coupled wall maximum base shear force demand including higher mode
effects
= first mode component of Qw,max
= vector of rigid body displacements of structure in horizontal direction
= strength ratio based on first significant yield of structure
= nominal weld metal design strength
= strength ratio based on global yield of structure
= design-level strength ratio
= survival-level strength ratio
= pitch for spiral reinforcement in wall pier contact region adjacent to coupling
beam
= distance (perpendicular to strong direction) of lateral load resisting system
shear center from center of story stiffness (CS)
= distance (perpendicular to strong direction) of coupled wall shear center from
center of story stiffness (CS)
= pitch for spiral reinforcement at wall pier base
xlix
S
Sa
Sad
Sas
SDS
SD1
SMS
SM1
SS
S1
SC
t
ta
tbc
tbf
tbw
tc
te
tf
tsh
tw
twc
T
Ta
Ta,pty
Tax
Tayx
Tayy
Teq
Tm
Tn
Ts
T0
T1
u
uad
uas
ub
ube
ubp
= compression-tension steel fiber

= site-adjusted spectral response acceleration
= site-adjusted design-level spectral response acceleration
= site-adjusted survival-level spectral response acceleration
= site-adjusted design-level spectral response acceleration at short periods
= site-adjusted design-level spectral response acceleration at 1 sec. period
= site-adjusted survival-level spectral response acceleration at short periods
= site-adjusted survival-level spectral response acceleration at 1 sec. period
= mapped spectral response acceleration at short periods
= mapped spectral response acceleration at 1 sec. period
= compression-only steel fiber
= time
= thickness of angle leg
= thickness of coupling beam flange plus cover plate
= thickness of coupling beam flange
= thickness of coupling beam web
= thickness of coupling beam flange cover plate
= thickness of wall pier embedded steel plate
= thickness of concrete fiber
= thickness of shim plate at coupling beam-to-wall pier interface
= thickness of wall pier
= confined thickness of wall pier enclosed by centerline of confinement
reinforcement
= period
= tension force in angle
= value of Ta at coupling beam PT-yielding state
= tension force in angle for loading parallel to coupling beam
= tension yield force of angle for loading parallel to coupling beam
= yield force of angle for loading perpendicular to coupling beam
= period of equivalent single-degree-of-freedom system
= period for mth mode
= period for nth mode
= SM1/SMS
= 0.2SM1/SMS
= first (fundamental) vibration mode period
= lateral displacement of structure at roof level
= design-level deformation demand for tension angle
= survival-level deformation demand for tension angle
= vertical displacement of loading block relative to reaction block
= yield displacement of BE component of equivalent nonlinear single-degreeof-freedom BP hysteresis model
= yield displacement of equivalent nonlinear single-degree-of-freedom BP
hysteresis model
l
ub,pty
uc
udes
uelas
uep
ueq
u& eq
&u& eq
ueq,be
ueq,bp
ueq,ep
&u& g
unlin
uwpy
uy
u1
{U}
&}
{U
&& }
{U
V
Vap
Va,ss
Va0
Vb
Vba,sof
Vbp,sof
Vb,pty
Vb,ss
Vb,sof
Vb,ws
Vb,wu
Vb,wy
Vp,ss
wab
= elongation of beam post-tensioning tendons at coupling beam PT-yielding

state
= amplitude of roof displacement during loading cycle
= lateral displacement of structure at first significant yield
= linear elastic lateral displacement corresponding to Felas
= yield displacement of EP component of equivalent nonlinear single-degree-offreedom BP hysteresis model
= displacement of equivalent single-degree-of-freedom system
= velocity of equivalent single-degree-of-freedom system
= acceleration of equivalent single-degree-of-freedom system

= yield displacement of BE component of equivalent nonlinear single-degreeof-freedom BP hysteresis model
= yield displacement of equivalent nonlinear single-degree-of-freedom BP
hysteresis model
= yield displacement of EP component of equivalent nonlinear single-degree-offreedom BP hysteresis model
= horizontal ground acceleration
= peak (maximum/minimum) lateral displacement
= elongation of extreme wall pier post-tensioning bar at coupled wall PTyielding state
= lateral displacement of structure at global yield
= first vibration mode component of roof lateral displacement
= relative lateral displacement vector of structure
= relative lateral velocity vector of structure
= relative lateral acceleration vector of structure
= force
= plastic shear force in angle vertical leg (leg perpendicular to coupling beam)
considering shear-flexure interaction
= contribution of top and seat angles to Vb,ss
= plastic shear force in angle vertical leg (leg perpendicular to coupling beam)
without considering shear-flexure interaction
= shear force in coupling beam
= contribution of top and seat angles to Vb,sof
= contribution of coupling beam post-tensioning force to Vb,sof
= value of Vb at coupling beam PT-yielding state
= shear slip capacity of coupling beam at beam-to-wall pier interface
= value of Vb at coupling beam softening state
= value of Vb at coupled wall softening state
= value of Vb at coupled wall ultimate state
= value of Vb at coupled wall PT-yielding state
= contribution of coupling beam axial force to Vb,ss
= width of angle connector head/nut measured across flat sides
li
wabw
ww
W
{Y}
Yi
Y1
za
zcb
a
bg
cw
d
tw
wg
wp
ar
cw
d
r
s
tw
a,pty
ax
ay
ayx
ayy
ins
isd
iss
d
s
t
td
ts
= width of angle-to-column/wall pier connector head/nut measured across flat

sides
= seismic weight for coupled wall
= total seismic weight for entire building
= vector of modal amplitudes
= amplitude of ith vibration mode shape
= amplitude of first vibration mode shape
= arm for angle force Ta+Ca couple
= arm for coupling beam compression force Cb couple
= post-yield stiffness ratio of equivalent nonlinear single-degree-of-freedom BP
hysteresis model
= post-yield stiffness ratio for angle in tension
= shear coefficient for coupling beam initial stiffness
= confined concrete stress-block parameter for compression-side wall pier at
base
= Rayleigh damping factor for mass proportional damping
= confined concrete stress-block parameter for tension-side wall pier at base
= shear coefficient for wall pier initial linear-elastic stiffness matrix
= post-yield stiffness ratio for wall pier post-tensioning bars
= angle strength ratio
= confined concrete stress-block parameter for compression-side wall pier at
base
= Rayleigh damping factor for stiffness proportional damping
= equivalent nonlinear single-degree-of-freedom BP hysteresis model strength
ratio
= equivalent nonlinear single-degree-of-freedom BP hysteresis model stiffness
ratio
= confined concrete stress-block parameter for tension-side wall pier at base
= participation factor (Leq/Meq)
= value of ax at coupling beam PT-yielding state
= deformation in tension angle parallel to coupling beam (horizontal
displacement of angle heel from column/wall pier)
= deformation in tension angle perpendicular to coupling beam
= value of ax at yield of angle
= value of ay at yield of angle
= roof drift (roof lateral displacement divided by structure height)
= inter-story drift of coupled wall
= design-level inter-story drift demand of coupled wall
= survival-level inter-story drift of coupled wall
= design-level roof drift demand of coupled wall
= survival-level roof drift demand of coupled wall
= target allowable roof drift demand of coupled wall
= design-level target allowable roof drift demand of coupled wall
= survival-level target allowable roof drift demand of coupled wall
lii
ws
wu
wue
wug
w,unc
wup
wy
am
au
ay
be
bm
bpu
bpy
bu
by
'c
cc
cce
ccu
ce
cu
cws
rm
ru
ry
sm
spm
spu
spy
sy
tce
wps
wpi
wp1
wp2
wp3
wp4
wp5
wp6
wp7
= roof drift of coupled wall at coupled wall softening state

= roof drift of coupled wall at coupled wall ultimate state
= linear-elastic component of wu
= component of wu due to gap opening at compression-side wall pier base
= roof drift of uncoupled wall (one wall pier)
= component of wu due to plastic hinge rotation at compression-side wall pier
base
= roof drift of coupled wall at coupled wall PT-yielding state
= strain of angle steel at fam
= strain of angle steel at fau
= yield strain of angle steel
= extreme strain of steel at coupling beam end
= strain of coupling beam steel at fbm
= strain of coupling beam post-tensioning strand at fbpu
= yield (limit of proportionality) strain of coupling beam post-tensioning strand
= strain of coupling beam steel at fbu
= yield strain of coupling beam steel
= strain of unconfined concrete at f c'
= strain of confined concrete at fcc
= extreme compression strain of confined concrete at compression-side wall
pier base
= ultimate (crushing) strain of confined concrete
= extreme compression strain of confined concrete at wall pier base
= ultimate (crushing) strain of unconfined concrete
= strain in extreme flexural reinforcement of compression-side wall pier at
coupled wall softening state
= strain of coupling beam flange cover plate steel at frm
= strain of coupling beam flange cover plate steel at fru
= yield strain of coupling beam flange cover plate steel
= strain of steel at fsm
= strain of spiral steel at fspm
= strain of spiral steel at fspu
= yield strain of spiral steel
= yield strain of steel
= extreme compression strain of confined concrete at tension-side wall pier base
= strain in wall pier post-tensioning bar under survival-level earthquake
= initial strain in wall pier post-tensioning bar before application of lateral loads
= strain in wall pier post-tensioning bar layer 1
liii
wp8
wpu
wpy
wse
wsm
wsy
b
b,bfy
b,dec
be,pty
bg,pty
b,pty
b,sof
b,wu
b,wy
d
s
aw
bw
t
td
ts
wf
'
"
b
c
eq
m
n
0
1
3
bsp
wp

= strain of wall pier post-tensioning bar at fwpu
= yield (limit of proportionality) strain of wall pier post-tensioning bar
= extreme tension strain of compression-side wall pier flexural reinforcement at
base
= strain of wall pier mild steel reinforcement at fwsm
= yield strain of wall pier mild steel reinforcement
= coupling beam chord rotation (i.e., transverse displacement between tangents
at beam ends divided by beam length)
= value of b at coupling beam flange yielding state
= value of b at coupling beam decompression state
= linear-elastic component of b,pty
= component of b,pty due to gap opening at coupling beam-to-wall pier interface
= value of b at coupling beam PT-yielding state
= value of b at coupling beam softening state
= value of b at coupled wall ultimate state
= value of b at coupled wall PT-yielding state
= design-level coupling beam chord rotation demand
= survival-level coupling beam chord rotation demand
= plastic hinge rotation in angle vertical leg (leg perpendicular to coupling
beam)
= peak lateral displacement ductility
= coefficient of friction between angle vertical leg and wall pier
= coefficient of friction between coupling beam and wall pier
= allowable target peak lateral displacement ductility
= design-level allowable target peak lateral displacement ductility
= survival-level allowable target peak lateral displacement ductility
= coefficient of friction between wall pier base and foundation
= factor for distribution of story shear force in building plan
= value of due to story translations
= value of due to applied and accidental torsion effects
= Poisson ratio for coupling beam steel
= Poisson ratio for wall pier concrete
= viscous damping ratio
= viscous damping ratio of equivalent single-degree-of-freedom system
= assumed viscous damping ratio for mth mode
= assumed viscous damping ratio for nth mode
= viscous damping ratio for acceleration response spectra used in design
= viscous damping ratio for first (fundamental) vibration mode
= viscous damping ratio for third vibration mode
= reliability factor based on system redundancy (ICC 2000)
= spiral reinforcement ratio in wall pier contact region adjacent to coupling
beam
= post-tensioning steel reinforcement ratio at wall pier base
liv
ws
wsp
wc
cwu
{i}
{1}
1x
w
[]
eq
1
= flexural mild steel reinforcement ratio at wall pier base

= spiral reinforcement ratio at wall pier base
= nominal concrete compression stress in wall pier contact region adjacent to
coupling beam
= curvature of compression-side wall pier base at coupled wall ultimate state
= ith linear-elastic vibration mode-shape
= first (fundamental) linear-elastic vibration mode shape
= coordinate of {1} at floor/roof level x
= a resistance factor for fillet weld design
= matrix containing linear-elastic vibration mode shapes
= frequency of equivalent single-degree-of-freedom system
= first (fundamental) mode frequency
lv
ACKNOWLEDGMENTS
This investigation was funded by the National Science Foundation (NSF) under
Grant No. CMS 98-10067 as a part of the United States-Japan Cooperative Earthquake
Research Program on Composite and Hybrid Structures. The support of the NSF
Program Director Dr. Shih-Chi Liu and Coordinators Dr. Subash Goel and Dr. Stephen
Mahin is gratefully acknowledged.
The authors thank the American Concrete Institute (ACI) Committee 335 on
Composite and Hybrid Structures and Dr. Kent Harries for providing guidance.
In addition, the following companies have contributed to the project through
material donations and assistance, and this support is gratefully acknowledged:
Ambassador Steel; CSR American Precast, Inc.; Dayton/Richmond Concrete
Accessories; Dywidag Systems International, U.S.A, Inc.; Flexalloy Inc.; Insteel Wire
Products; Ivy Steel & Wire.
The authors also thank several individuals for their assistance in the conduct of this
work: Brent Bach, Keli Engvall, Julio Escada, and Michael May.
The opinions, findings, and conclusions expressed in this report are those of the
authors and do not necessarily reflect the views of the NSF or of the individuals and
organizations acknowledged above.
lvi
CHAPTER 1
INTRODUCTION
This chapter provides an introduction for the dissertation as follows: (1) problem
statement; (2) proposed hybrid coupled wall system; (3) research objectives; (4) research
scope; (5) summary of approach; (6) research significance; and (7) organization of
dissertation.
1.1
Problem Statement
Previous earthquakes have shown that reinforced concrete structural walls are one
of the most effective primary lateral load resisting systems for seismic regions. The
lateral stiffness and strength of these systems can be significantly increased by coupling
two or more walls together. Coupling of the walls can be achieved by using reinforced
concrete or steel beams at the floor and roof levels. The research described in this
dissertation is on hybrid systems that use steel beams to couple reinforced concrete
walls.
In conventional hybrid coupled wall systems, coupling of the walls is achieved
by embedding the coupling beams into the walls. As an example, Fig. 1.1 shows a typical
eight-story coupled wall system and Fig. 1.2a shows a floor-level hybrid coupled wall
subassemblage that consists of two concrete wall regions and a steel coupling beam
embedded into the walls (indicated by the shaded area in Fig. 1.1). Fig. 1.2b shows an
idealized exaggerated deformed configuration of the subassemblage under lateral forces
acting on the walls from left to right. The desired behavior of this system can be achieved
if the coupling beams yield before the walls, behave in a ductile manner, and dissipate
energy. This requires that the coupling beams are designed and detailed to develop and
sustain their full moment (or shear) capacity under large cyclic deformations.
Hybrid coupled wall systems with embedded steel beams have been investigated
as part of the National Science Foundation U.S.-Japan Cooperative Earthquake Research
Program on Composite and Hybrid Structures and other research programs (Harries 1992,
1995, 2001; Harries et al. 1992, 1993, 1994, 1997, 2000; Shahrooz et al. 1992, 1993a,
1993b, 2003; Gong and Shahrooz 1997, 1998, 2001a, 2001b; Gong et al. 1997; Shahrooz
and Gong 1998). Properly-designed and detailed embedded steel coupling beam systems
possess large ductility and energy dissipation capacities; however, they are prone to
significant damage during lateral displacements in the form of flexural and/or shear
yielding in the beams, and spalling and cracking in the wall concrete in the embedment
regions (Harries et al. 1997). Furthermore, the selection of the beam shape and
1
dimensions is affected by the wall reinforcement in the embedment regions, often

requiring fabricated sections for the beams instead of readily available rolled sections.
wall
coupling
beam
wall
concrete
wall region
embedded
steel coupling
beam
embedment
region
(a)
hw
(b)
lw
lb
lw
Fig. 1.1 Eight story coupled

wall structure
Fig. 1.2 Hybrid coupled wall subassemblage with

embedded beam: (a) undeformed configuration;
(b) deformed configuration
In comparison with the system in Fig. 1.2, the coupling beams investigated in this
dissertation are not embedded into the walls. Instead, coupling is achieved by posttensioning the beams and the walls together at the floor and roof levels. Compared with
current industry practice and research, the use of post-tensioning to couple structural wall
systems is a significant innovation and has the following primary advantages:
(1) The design and detailing requirements for the beams and the wall coupling
regions are expected to be simpler than the requirements for conventional systems.
(2) The coupling beams and the wall coupling regions can be designed to remain
essentially damage-free during a severe ground motion, with most of the damage
occurring in certain selected components of the beam-to-wall connections that can be
replaced after the earthquake.
(3) The post-tensioning force provides a self-centering capability to the structure
(i.e., ability of the structure to return towards the original undisplaced position upon
unloading from a large nonlinear lateral displacement).
2
This dissertation provides a comprehensive study on the seismic behavior and

design of this new system.
1.2
Proposed Hybrid Coupled Wall System
As an example, Fig. 1.3a shows a subassemblage of the proposed hybrid coupled

wall system at a floor level. The subassemblage consists of a steel coupling beam and two
concrete wall regions (as indicated by the shaded area in Fig. 1.1) coupled using posttensioning. The post-tensioning force is provided by multi-strand high strength tendons
that are not in contact with the beam along the span. Inside the walls, the bond between
the post-tensioning steel and the concrete is prevented by placing the tendons inside ducts
that are not grouted. This type of construction, where the post-tensioning steel is
anchored to the coupled wall system only at two locations at the outer ends of the walls,
is called unbonded post-tensioning. The advantages of using unbonded post-tensioning
as compared with bonded post-tensioning are described later.
The beam-to-wall connection regions include top and seat angles connected to the
beam flanges and to steel plates embedded inside the walls (with welded headed studs) as
shown in Fig. 1.3a. Fig. 1.3b shows the expected exaggerated deformed shape of the
subassemblage under lateral loads acting on the walls from left to right. The nonlinear
deformations of the beam occur primarily as a result of gap opening at the beam-to-wall
interfaces. Friction develops at the beam-to-wall interfaces due to the post-tensioning
force to resist the coupling shear force in the vertical direction.
In a properly-designed subassemblage, the desired behavior is yielding of the top
and seat angles due to the opening of the gaps, with little yielding and damage in the
beam and walls. The purpose of the angles is to provide redundancy and energy
dissipation during an earthquake. The angles also provide a part of the coupling
resistance of the subassemblage, prevent sliding of the beam at the beam-to-wall
interfaces (together with friction resistance against sliding as induced by post-tensioning),
and can serve as temporary beam supports during construction. The yielded angles can be
replaced after the earthquake.
As the subassemblage deforms, gap opening results in large concentrated
compressive stresses in the contact regions near the beam-to-wall interfaces (Fig. 1.3b).
Steel cover plates welded to the flanges may be used at the beam ends to stabilize and
strengthen the flanges, and prevent or delay the yielding of the flanges in compression.
The embedded plates inside the walls distribute the compressive stresses to the concrete.
In addition, confinement reinforcement is used behind the embedded plates to transfer
and distribute the compressive stresses without excessive deformations in the concrete.
Fig. 1.3c shows the coupling forces that develop from the formation of a diagonal
compression strut in the beam upon lateral displacement of the walls, where Vb is the
coupling shear force in the beam. The amount of coupling between the walls can be
controlled by controlling the total beam post-tensioning force Pb (which controls the total
3
compression force in the beam, Cb), the tension and compression angle forces Ta and Ca,
the beam depth db, and the beam length lb.
concrete wall
region
PT anchorage
angl e
connection concrete wall

region
region
steel
coupling
beam
A
hs
db
PT tendon
steel embedded
plate
lw
cover
plat e
concrete
confinement
lb
abp
section at A -A
lw
(a)
beam-to-wall
contact
region
gap opening
reference line
(b)
diagonal
compression strut
Ca
Cb
Vb
Ta
coupling beam
zcb C d
b b
za
lb
Ta
Vb =
Vb
Ca
Cb zcb+ (Ta + Ca ) z a
lb
(c)
Fig. 1.3 Proposed coupled wall system: (a) subassemblage at a floor level;
(b) deformed configuration; (c) coupling forces
As a result of post-tensioning, the initial lateral stiffness of the coupled wall
subassemblage in Fig. 1.3a before the initiation of gap opening is similar to the initial
stiffness of a comparable system with embedded steel coupling beams. Gap opening
results in a reduction in the lateral stiffness of the subassemblage, allowing the system to
soften and undergo large nonlinear rotations without significant damage (except for the
damage in the angles). The post-tensioning force controls the depth and width of the gaps.
As the walls displace laterally, the tensile forces in the post-tensioning tendons increase,
4
resulting in an increase in the compression axial force in the coupling beam. Upon
unloading, the post-tensioning tendons provide a restoring force that tends to close the
gaps, thus pulling the beam and the walls back towards their undisplaced position with
little or no residual displacement (i.e., self-centering capability).
Unbonding of the post-tensioning tendons has two important advantages: (1) it
results in a uniform strain distribution in the tendons between the anchors, and thus,
significantly delays the nonlinear straining (i.e., yielding) of the steel (yielding of the
post-tensioning steel is not desirable since it results in a loss of prestress under cyclic
loading); and (2) it significantly reduces the tensile stresses transferred to the wall
concrete, thus reducing damage due to concrete cracking.
1.3
Research Objectives
The broad objective of the research described in this dissertation is to investigate

the seismic behavior and design of hybrid coupled wall structures that use unbonded
post-tensioning. The research has three specific objectives:
(1) To develop and verify analytical models for the nonlinear behavior of unbonded
post-tensioned hybrid coupled wall structures under seismic loading.
(2) To conduct parametric evaluations on the seismic behavior of prototype coupled
wall systems.
(3) To develop seismic design procedures, tools, and application guidelines for
these structures.
1.4
Research Scope
The research described in this dissertation investigates the seismic behavior and
design of reinforced concrete walls coupled using unbonded post-tensioned steel beams.
The investigation is limited to the coupling of two identical walls that are sufficiently
slender such that their nonlinear behavior under lateral loads is governed by axial and
flexural effects rather than by shear effects. Coupling of more than two walls is not
investigated. Note that the research focuses on rectangular cross-sections for the walls
and I-shaped cross sections for the coupling beams; however, it may be possible to
extend the findings to other types of cross-sections, such as flanged walls and hollow
beams.
In order to achieve the research objectives described above, a two-dimensional (in
the plane of the walls) analytical model for unbonded post-tensioned hybrid coupled wall
structures is developed using the DRAIN-2DX program (Prakash et al. 1993). It is
assumed that the earthquake-induced forces at the floor and roof levels of the building are
transferred to the primary lateral load resisting structure through rigid diaphragms.
Interactions between the coupled wall systems and the rest of the building (i.e., the floor
5
system/diaphragm, columns, foundation-soil system) are not included in the analytical

model. The DRAIN-2DX model is verified by comparing the analytical results with
results obtained using the ABAQUS program (Hibbitt et al. 1998) as well as with results
obtained from experiments of floor-level unbonded post-tensioned hybrid coupled wall
subassemblages and experiments of uncoupled walls.
The analytical models are used to conduct parametric nonlinear static push-over
analyses and nonlinear cyclic lateral load analyses of isolated unbonded post-tensioned
hybrid coupled wall subassemblages and multi-story coupled wall structures. The
following parameters are considered in the analyses: (1) amount of post-tensioning; (2)
location of the post-tensioning steel; (3) length of the coupling beams; (4) cross-sectional
properties of the coupling beams; (5) properties of the beam-to-wall connection angles; (6)
cross-sectional properties of the walls; (7) precast concrete walls versus monolithic castin-place reinforced concrete walls; and (8) coupling degree of coupled wall system.
The results from the parametric investigations are used to develop simplified
closed-form procedures and tools that can be used in the seismic design and nonlinear
response evaluation of unbonded post-tensioned hybrid coupled wall structures. A
performance-based seismic design approach is proposed with the primary objective of
limiting the peak lateral displacements of the structure during a design level and/or
survival level earthquake to below selected values of allowable target displacement.
Two eight-story prototype unbonded post-tensioned hybrid coupled wall structures
are designed using the proposed seismic design approach as follows: (1) a structure with
monolithic cast-in-place reinforced concrete walls; and (2) a structure with precast
concrete walls, Both structures are designed for a region with high seismicity (Los
Angeles) and a site with a medium soil profile [Site Class D in IBC 2000 (ICC 2000)].
The seismic behavior of the structures is assessed and the proposed design
approach is critically evaluated based on nonlinear static and nonlinear dynamic timehistory analyses. Ductility demands on the coupling beams and on the walls are
established. The seismic behavior of the prototype structures is compared with the
behavior of conventional hybrid coupled wall structures that use embedded steel beams.
1.5
Summary of Approach
The research described in this dissertation consists of the following eight tasks:
(1) Conduct a literature survey of previous analytical and experimental research on

the seismic behavior and design of coupled wall structures, as well as structures with
unbonded post-tensioning.
(2) Develop and verify analytical models that describe the nonlinear hysteretic
behavior of hybrid coupled wall structures with unbonded post-tensioned steel beams and
structures with embedded steel beams.
6
(3) Using the analytical models, conduct parametric investigations on the behavior
of unbonded post-tensioned hybrid coupled wall subassemblages as well as multi-story
coupled wall systems under lateral loads.
(4) Using the results from the parametric investigations, develop simplified closedform procedures to estimate the nonlinear behavior of unbonded post-tensioned hybrid
coupled wall structures under lateral loads.
(5) Develop a performance-based approach for the seismic design of unbonded
post-tensioned hybrid coupled wall structures.
(6) Use the design approach developed in Task (5) to design prototype building
structures that use unbonded post-tensioned hybrid coupled walls as the primary lateral
load resisting system.
(7) Select a set of ground motion records representative of the seismic regions and
site soil characteristics used in the design of the prototype structures.
(8) Evaluate the seismic response of the prototype coupled walls, including
comparisons with walls that use embedded steel beams, based on nonlinear static and
nonlinear dynamic time-history analyses using the analytical models developed in Task
(2) and the ground motions selected in Task (7).
These research tasks are described fully in the following chapters of the dissertation.
Each task addresses one or more of the project objectives outlined above. Tasks (1), (2),
and (4) are needed to achieve Objective (1). Tasks (3), (7), and (8) achieve Objective (2).
Finally, Tasks (1), (4) (5), (6), (7), and (8) are used to achieve Objective (3).
1.6
Research Significance
Compared with current industry practice and current research, the use of unbonded
post-tensioning to couple structural walls is a significant innovation. The research results
described in this dissertation are needed to demonstrate that unbonded post-tensioned
steel beams can provide significant and stable levels of coupling between concrete walls
over large nonlinear cyclic deformations and to develop seismic building design
recommendations for this new type of hybrid structural system.
It may also be possible to use the proposed system to couple existing walls as part
of a seismic retrofit and strengthening scheme. The post-tensioning tendons can be placed
outside the walls for this purpose. The use of post-tensioning allows the designer to
control the amount of coupling between the walls as shown in Fig. 1.3c. This may be
particularly important for the strengthening of existing walls since additional axial forces
(both tension and compression) develop in the walls as a result of coupling. Thus, the
amount of coupling may be limited by the existing design details of the foundations and
the walls near the base.
7
This research, which was conducted as a part of the U.S.-Japan Cooperative

Earthquake Research Program on Composite and Hybrid Structures funded by the
National Science Foundation, is essential and fundamental to the study of the seismic
behavior and design of unbonded post-tensioned hybrid coupled wall systems. Results
from the research can be used as the foundation on which future experimental research on
multi-story structures can be built and codified building design specifications can be
formulated.
1.7
Organization of Dissertation
The remainder of this dissertation is organized into the following fifteen chapters
(Chapters 2-16) in accordance with the research approach summarized above (Tasks 1-8).
Chapter 2 presents a literature survey of previous analytical and experimental
research on the seismic behavior and design of coupled wall structures as well as
structures with unbonded post-tensioning.
Chapter 3 describes analytical models to investigate the nonlinear behavior of
floor-level hybrid coupled wall subassemblages.
Chapter 4 describes a parametric investigation on the nonlinear lateral load
behavior of unbonded post-tensioned hybrid coupled wall subassemblages using the
analytical models in Chapter 3.
Chapter 5 presents simplified closed-form procedures to estimate the nonlinear
behavior of unbonded post-tensioned hybrid coupled wall subassemblages under lateral
loads.
Chapter 6 describes analytical models for multi-story hybrid coupled wall
structures based on the subassemblage models in Chapter 3.
Chapter 7 describes a parametric investigation on the nonlinear lateral load
behavior of multi-story unbonded post-tensioned hybrid coupled wall structures using the
analytical models in Chapter 6.
Chapter 8 presents simplified closed-form procedures to estimate the nonlinear
behavior of multi-story unbonded post-tensioned hybrid coupled wall structures under
lateral loads.
Chapter 9 describes an experimental investigation on the nonlinear behavior of
floor-level unbonded post-tensioned hybrid coupled wall subassemblages.
Chapter 10 compares the results from the experimental investigation in Chapter 9
with estimations using the subassemblage analytical models from Chapter 3, and
recommends modifications to the analytical models based on these comparisons.
8
Chapter 11 proposes a performance-based approach for the seismic design of

unbonded post-tensioned hybrid coupled wall structures.
Chapter 12 describes the design of prototype structures using the design approach
proposed in Chapter 11.
Chapter 13 evaluates the behavior of the prototype structures under static lateral
loading and provides comparisons with coupled wall structures that use embedded steel
coupling beams.
Chapter 14 describes the selection and properties of a set of ground motion records
representative of the seismic regions and site soil characteristics used in the design of the
prototype structures.
Chapter 15 presents an evaluation of the seismic response of the prototype
structures based on nonlinear dynamic time-history analyses, including comparisons with
uncoupled wall structures.
Finally, Chapter 16 presents a summary and conclusions from this dissertation, and
discusses future research.
CHAPTER 2
LITERATURE REVIEW
This chapter provides an overview of previous research related to this report as
follows: (1) coupled wall structural systems; (2) concrete coupled wall structures; (3)
hybrid structures with embedded steel coupling beams; (4) wall shear force demands in
coupled wall structures; (5) unbonded post-tensioning in building construction; and (6)
behavior of top and seat angles.
2.1
Coupled Wall Structural Systems
In concrete structural walls, a regular pattern of openings is often required to

accommodate windows, doors, and/or mechanical penetrations. Efficient seismic
structural systems particularly suited for ductile response with very good energy
dissipation characteristics can be achieved when these openings are arranged in a rational
pattern (Park and Paulay 1975; Paulay and Priestley 1992). Examples are shown in Fig.
2.1 where a number of walls are interconnected or coupled to each other by beams at the
floor and roof levels. These systems are generally referred to as coupled structural walls
with the implication that the connecting (i.e., coupling) beams, which may be relatively
short and deep, are substantially weaker than the walls. The walls, which behave
predominantly as cantilevers, can then impose sufficient rotations on the coupling beams
to make them yield. If suitably detailed, the beams can dissipate a significant amount of
energy distributed over the entire height of the structure.
wall
wall
coupling
beam
Fig. 2.1 Coupled structural walls

The behavior and mechanisms of lateral resistance of a single (i.e., uncoupled) wall
and two coupled wall structures are compared in Fig. 2.2. The gravity loads acting on the
walls are ignored for this example and it is assumed that a lateral force in the plane of the
10
walls is applied at the top. The base moment resistance, Mw,unc of the uncoupled wall (Fig.
2.2a) is developed in the traditional form by flexural stresses, while axial forces as well
as moments are resisted in the coupled wall systems (Figs. 2.2b and 2.2c). When a
coupled wall structure is pushed from left to right under lateral loads, tensile axial forces
(Ntwb) develop in the left wall and compressive axial forces (Ncwb=Ntwb) develop in the
right wall due to the coupling effect. The magnitude of these wall axial forces is equal to
the sum of the shear forces of all the coupling beams at the upper floor and roof levels,
and thus, depends on the stiffness and strength of those beams.
Vb
Vb
Vb
Vb
Vb
weak
beam
strong beam
tension
Vb
Vb
Vb
Vb
Vb
Vb
Vb
compression
strain gradient
Mw,unc
(a)
Mcw
Mtw
Ntwb
Lc
Ncwb=Ntwb
(b)
Mtw
Ntwb
Mcw
Lc
Ncwb=Ntwb
(c)
Fig. 2.2 Lateral load resisting mechanisms in structural walls (adapted from Paulay and
Priestley 1992): (a) uncoupled wall; (b) walls coupled with strong beams;
(c) walls coupled with weak beams
As a result of the axial forces that develop in the walls, the lateral stiffness and
strength of a coupled wall system are significantly larger than the combined stiffness and
strength of the individual constituent walls (i.e., wall piers) with no coupling. The total
base moment, Mw of the coupled wall structures in Figs. 2.2b and 2.2c can be written as:
M w = M tw + M cw + N cwb Lc
(2.1)
where, Mtw and Mcw are the base moments in the tension-side and compression-side walls,
respectively, Ncwb=Ntwb, and Lc is the distance between the centroids of the tension-side
and compression-side walls. Then, the contribution of the wall axial forces due to
11
coupling to the total lateral resistance of the structure can be expressed by the degree (or
level) of coupling as
DOC =
N cwb Lc
N cwb Lc
=
Mw
M tw + M cw + N cwb Lc
(2.2)
The degree of coupling, DOC, can be controlled by changing the strength and
stiffness of the beams relative to the walls as shown in Figs. 2.2b and 2.2c. Note that the
degree of coupling varies as the wall deforms into the nonlinear range of behavior under
lateral loading. This is discussed in more detail later in the report.
The coupling degree is a key parameter in the seismic behavior and design of
coupled wall structures. Too little coupling (i.e., too small a coupling degree) yields a
system that behaves in a manner similar to uncoupled walls and the benefits due to
coupling are minimal. Too much coupling (i.e., too large a coupling degree) will add
excessive stiffness to the system, causing the coupled walls to perform as a single pierced
wall with little or no energy dissipation provided by the beams. The desirable range for
the amount of coupling lies in between these two extremes and should be selected
properly in seismic design.
Coupling beams are analogous to, and serve the same structural role as link beams
in eccentrically braced frames. To achieve desirable performance, the failure mechanism
should involve the formation of plastic hinges in most or all of the coupling beams and
also at the base of each wall. In this manner, the dissipation of seismic input energy is
distributed over the height of the structure (rather than being concentrated in a few
stories), similar to the strong column-weak beam design philosophy for ductile moment
resisting frames.
Despite their structural advantages such as increased lateral strength and stiffness
and good energy dissipation, coupled wall systems are not used to their full potential in
current design practice. This may be because of the complex characteristics of these
structures where achieving a code compliant design and optimizing the global system
behavior are often difficult.
2.2
Concrete Coupled Wall Structures
In the last three to four decades, an extensive amount of experimental and

analytical research has been conducted on the seismic behavior and design of monolithic
cast-in-place reinforced concrete coupled wall structural systems (e.g., Paulay 1971, 1977,
1986; Paulay and Binney 1974; Santhakumar 1974; Aristizabal-Ochoa and Sozen 1976;
Srichatrapimuk 1976; Paulay and Santhakumar 1976; Lybas and Sozen 1977; Shiu et al.
1978, 1981; Aristizabal-Ochoa et al. 1979; Fintel and Ghosh 1980, 1982; Aktan and
Bertero 1981, 1984, 1987; Aktan et al. 1982; Aristizabal-Ochoa 1982, 1987; Saatcioglu et
al. 1987; Pekau and Cistra 1989; Subedi 1991a, 1991b; Chaallal 1992; Chaallal and
Gauthier 1995, 2000, 2001; Chaallal and Ghlamallah 1996; Chaallal et al. 1996a, 1996b,
12
1997; Tassios et al. 1996; Chaallal and Nollet 1997; Harries et al. 1998; Teshigawara
1998; Sugaya et al. 1999; Harries et al. 2000; Munshi and Ghosh 2000; Harries 2001;
Cosenza and Pecce 2001; Nollet and Chaallal 2002; Paulay 2002). The behavior and
design of the monolithic cast-in-place reinforced concrete coupling beams in these
structures are significantly different than the behavior and design of the unbonded posttensioned steel coupling beams investigated in this research, and thus, a complete review
of the previous research listed above is beyond the scope of this report. However, some
of the important findings and conclusions regarding the seismic modeling, behavior, and
design of monolithic cast-in-place reinforced concrete coupled wall structures are
summarized below.
2.2.1
Behavior and Design of Concrete Coupling Beams
The primary role of coupling beams during an earthquake is the transfer of forces
from one wall pier to the other. In considering the seismic behavior and design of coupled
wall structures, it should be noted that significantly larger nonlinear deformations can
occur in the coupling beams than in the walls that are coupled. Previous research has
shown that short monolithic cast-in-place reinforced concrete coupling beams with
conventional transverse steel reinforcement (i.e., vertical hoops) inevitably fail in
diagonal tension or sliding shear (Figs. 2.3a and 2.3b), with limited or no ductility
capacity (Paulay and Priestley 1992). The displacement capacity of these structures is
often exceeded by the demand (Harries 2001). The consideration of large seismic
ductility demands under many load reversals has led to the development of concrete
coupling beams that use diagonal steel reinforcement as shown in Fig. 2.3c. This allows
for the transfer of diagonal tension and compression forces to the reinforcement during
the lateral displacements of the structure, resulting in a considerably ductile behavior
with good energy dissipation characteristics.
(a)
(b)
(c)
Fig. 2.3 Shear in concrete coupling beams (from Paulay and Priestley 1992):
(a) diagonal tension failure; (b) sliding shear failure;
(c) beam with diagonal reinforcement
13
2.2.2
Analysis and Modeling of Concrete Coupled Wall Structures
Previous research has often adopted an equivalent frame analogy for the
nonlinear inelastic hysteretic modeling and analysis of coupled wall structural systems as
shown in Fig. 2.4. The cross-sectional properties of the walls and the beams are
concentrated at the centerline of each member. Where the coupling beams frame into the
walls, it is usually necessary to use rigid end zones (or kinematic constraints) in the
analytical model. Rigid end zones may also be needed in the walls depending on the
stiffness of the coupling beams. Nonlinear shear deformations in walls with aspect ratios,
hw/lw (as shown in Fig. 2.4) larger than 4 are often neglected (Paulay and Priestley 1992).
outline of
structure
the model
structure
hw
hw
infinitely
rigid
regions
lw
lb
lw
lw
(a)
lb
lw
(b)
Fig. 2.4 Modeling of coupled wall structures (adapted from Paulay and Priestley
1992): (a) walls with flexible beams; (b) walls with stiff beams
It is emphasized that the validity of the frame analogy used to model the behavior
of coupled wall structures can vary considerably depending on the stiffnesses assumed
for the members (including the effect of concrete cracking) and the lengths of the rigid
end zones. Furthermore, axial elongation and shortening of the tension and
compression side walls due to axial-flexural interaction cannot be represented using
frame analogy. As a result, the modeling of the coupled wall structures in this report is
done using fiber beam-column elements as described in Chapter 3. A significant
advantage of using fiber-based elements instead of frame analogy is that a reasonably
accurate model can be constructed accounting for axial-flexural interaction including
14
axial as well as flexural deformations in the members, nonlinear behavior of concrete and
steel, and gap opening at the beam-to-wall interfaces.
2.3
Hybrid Structures with Embedded Steel Coupling Beams
As described above, ductility is an important issue to be considered in the seismic

design of coupled wall structures. The coupling beams in a coupled wall structure must
behave in a ductile manner and dissipate energy during the cyclic displacements of the
walls during a severe earthquake. While reinforced concrete coupling beams can be
designed and detailed to behave in a relatively ductile manner by using diagonal
reinforcement as shown in Fig. 2.3c, the construction of these beams is cumbersome and
it is difficult to design a constructible diagonally reinforced coupling beam especially for
larger span-to-depth ratios (e.g., larger than 1.5, see Harries et al. 2005). In an attempt to
address the limitations of reinforced concrete coupling beams, previous research carried
out by Raths (1974), Marcakis and Mitchell (1980), and Mattock and Gaafar (1982) on
precast concrete connections with embedded steel members led to the use of embedded
steel beams (see Fig. 1.2) to couple reinforced concrete walls (e.g., Harries 1992, 1995;
Harries et al. 1992, 1993, 1994, 1997, 2000; Remmetter et al. 1992; Shahrooz et al. 1992,
1993a, 1993b, 2003; Gong and Shahrooz 1997, 1998, 2001a, 2001b; Gong et al. 1997;
Shahrooz and Gong 1998; El-Tawil et al. 2002; El-Tawil and Kuenzli 2002; Hassan and
El-Tawil 2004).
Using embedded steel coupling beams, it is possible to achieve better ductility and
energy dissipation with relatively simpler design and detailing requirements than
reinforced concrete coupling beams. Furthermore, reinforced concrete coupling beams
with diagonal reinforcement often result in impractically deep cross sections. The depth
of the beams can be considerably reduced by using steel cross sections as compared with
concrete beams, which is particularly desirable in high-rise apartment buildings. Thus,
the advantages of embedded steel coupling beams become especially apparent in cases
where height restrictions do not permit the use of deep reinforced concrete beams, or
where the required strength, stiffness, or deformation capacity cannot be developed
economically with a concrete beam, as demonstrated in recent construction projects
(Harries and Shahrooz 2005).
2.3.1
Behavior and Design of Embedded Steel Coupling Beam Subassemblages
Harries et al. (1993) and Shahrooz et al. (1993a) showed that properly-designed
and detailed embedded steel coupling beam systems possess large ductility and energy
dissipation capacities; however, they are prone to significant damage during lateral
displacements in the form of flexural and/or shear yielding in the beams, and spalling and
cracking in the wall concrete in the embedment regions.
Harries et al. (1997) conducted floor-level subassemblage experiments of
embedded steel coupling beams as depicted in Fig. 2.5a. The test results showed that the
seismic design and detailing of these systems should include: (1) moment and shear
15
strength requirements for the beams; (2) laterally unsupported length requirements to
prevent lateral instability of the beams; (3) stiffener requirements for the beams within
the clear span between the walls; (4) minimum embedment length requirements for the
beams; and (5) additional detailing requirements for the embedment regions. The
embedment regions are designed and detailed for large cyclic nonlinear-inelastic
deformations under shear forces and bending moments that correspond to the full
capacity of the beam cross-section. The design requirements for the embedment regions
aim to ensure that these regions remain essentially linear-elastic while the coupling
beams undergo inelastic deformations in their clear span. These requirements include: (1)
full-depth stiffeners on both sides of the beam web; (2) cover plates for the beam flanges;
(3) increased thickness for the beam web; and (4) additional wall reinforcement. A full
description of the design and detailing requirements for embedded steel coupling beams
is given in Harries et al. (1997).
Specimen S3
Specimen S3
(a)
(b)
(c)
(d)
(e)
(f)
Specimen S4
Fig. 2.5 Hybrid coupled wall subassemblages with embedded steel beams (from Harries
et al. 1997); (a) shear-critical beam; (b) beam shear force versus relative end
displacement behavior; (c) damage at end of test; (d) flexure-critical beam; (e) beam
shear force versus relative end displacement behavior; (f) damage at end of test.
The nonlinear behavior of embedded steel coupling beams can be characterized as
shear critical or flexure critical. Shear-critical beams behave in a predominantly
shear mode, remaining essentially elastic in flexure. Similarly, flexure-critical beams
behave in a predominantly flexural mode. It is typically more advantageous to design
steel coupling beams as shear yielding members since such members exhibit a more
desirable mode of energy dissipation (Harries et al. 1997).
16
As an example, Figs. 2.5a and 2.5b show the design details and the beam shear
force versus relative end displacement behavior, respectively, of a shear-critical
embedded steel coupling beam (Harries et al. 1997). As a result of proper seismic
detailing, the behavior of the test specimen shows stable hysteresis loops with large
ductility capacity and large energy dissipation. However, as shown in Fig. 2.5c, the
specimen has significant damage at the end of the test in the form of excessive yielding of
the beam over the entire clear span and pronounced concrete cracking in the walls.
Extensive spalling of the cover concrete was also observed in another test specimen.
Earlier tests conducted by Harries et al. (1993) on specimens without proper detailing
exhibited premature local flange and web buckling and a more pinched hysteretic
behavior.
Similarly, Fig. 2.5d shows a flexure critical coupling beam. The reversed cyclic
loading behavior of the specimen in Fig. 2.5e illustrates relatively large and stable
hysteresis loops throughout the test. Fig. 2.5f shows the flexural hinges that form in the
beam near the face of each wall, with significant concrete cover spalling and cracking in
the embedment regions. In order to reduce the amount of damage in the embedment
regions, a recent study by Shahrooz et al. (2003) has investigated steel coupling beams
with shear fuses to concentrate the inelastic deformations in a repairable or replaceable
component placed at the beam midspan. In this manner, the wall embedment regions are
protected from damage due to the coupling forces.
The experiments conducted by Harries et al. (1997) and Shahrooz et al. (1993a) are
used later in this report to develop analytical models for hybrid coupled wall systems
with embedded steel beams. These models are then used to compare the behavior of
embedded systems with the unbonded post-tensioned system investigated in this report.
As described above, despite excellent performance in terms of energy dissipation and
ductile hysteretic behavior, the connection regions of embedded steel coupling beams are
prone to significant damage under lateral loading. In such cases, repair is expensive and
may be impractical. The unbonded post-tensioned hybrid coupled wall system described
in this report has three important advantages over systems with embedded beams as
follows:
(1) The selection of the section shape and size for an unbonded post-tensioned steel
coupling beam is not affected by the reinforcement details inside the walls, and the use of
widely available rolled cross sections is recommended. In comparison, the wall
reinforcement inside the embedment regions and the seismic detailing requirements in the
form of stiffener and cover plates often result in the use of fabricated beam cross sections
in embedded systems.
(2) The nonlinear deformations in unbonded post-tensioned steel coupling beams
occur primarily as a result of the opening of gaps at the beam ends. Thus, corresponding
to a given nonlinear wall lateral displacement level, the amount of damage in the walls
17
and beams of an unbonded post-tensioned system is expected to be much smaller than the
corresponding damage in systems with embedded beams.
(3) The post-tensioning tendons used for coupling at the floor and roof levels
provide a restoring force that pulls the entire coupled wall structure back towards its
original undisplaced position (with little residual displacements) during and at the end of
an earthquake, thus providing self-centering capability (i.e., ability of the structure to
return towards zero displacement upon unloading from a nonlinear displacement). In
comparison, embedded hybrid coupled wall structures have relatively small self-centering
capability, indicating that significant permanent displacements (i.e., residual lateral
displacements from plumb position) can remain at the end of a severe earthquake.
2.3.2
Embedded Steel Coupling Beams with Concrete Encasement
As part of the NSF-funded U.S.-Japan Cooperative Earthquake Research Program

on Composite and Hybrid Structures, Gong and Shahrooz (1997, 1998) and Gong et al.
(1997) investigated the seismic behavior and design of embedded steel coupling beams
with concrete encasement. Steel coupling beams are often encased inside concrete
members such as door lintels. However, current design guidelines ignore the effect of this
encasement on the behavior of the beams. For example, stiffener details for the beams are
determined ignoring the effect of the encasement, and the embedment length is
determined based on the capacity of the steel beam cross section alone.
Experiments conducted by Gong and Shahrooz (1997, 1998, 2001a, 2001b) showed
that nominal concrete encasement around embedded steel coupling beams has the
following effects: (1) concrete encasement provides an effective means of preventing
web buckling, thus, web stiffeners in the clear span can be eliminated; (2) concrete
encasement increases the beam moment and shear capacities, thus, the embedment
regions have to be designed based on the capacity of the composite section including the
encasement; and (3) concrete encasement increases the stiffness of the beams, which
affects the degree of coupling and, thus, the distribution of forces between the walls.
Further information on steel coupling beams with concrete encasement is not
within the scope of this report.
2.3.3
Multi-Story Hybrid Coupled Walls with Embedded Beams
Compared with research on floor-level coupled wall subassemblages, previous

research on multi-story hybrid coupled wall structures is limited and includes analytical
investigations only (e.g., Harries 1996, 2001; El-Tawil and Kuenzli 2002; El-Tawil et al.
2002; Hassan and El-Tawil 2004; Harries and McNeice 2006). Harries (1996) conducted
an analytical investigation of the seismic behavior of two eighteen story hybrid coupled
walls with embedded steel beams (Fig. 2.6a). The structures were modeled as twodimensional equivalent frames (Fig. 2.6b) using beam-column frame elements located
18
at the wall centerlines and beam elements located at the beam centerlines. The dynamic
response of the structures under four ground motion records was obtained from nonlinear
dynamic time-history analyses.
EQ
550x550mm column (typ.)

steel coupling beam
N
30 m
6 m (typ.)
concrete wall
200 mm thick flat plate

rigid link
6 m (typ.)
42 m
5.5 m
end bays
(a)
(b)
Fig. 2.6 Hybrid coupled wall system with embedded steel beams (adapted from Harries
1996): (a) plan view; (b) analytical model
As part of a conventional strength-based design approach, Harries (2001) conducted
an assessment of the seismic behavior of coupled wall structures with different types of
coupling beams (i.e., conventionally reinforced concrete beams, diagonally reinforced
concrete beams, and embedded steel beams). Through a review of available experimental
investigations on coupling beam subassemblages and analytical investigations on multistory coupled wall systems, it was concluded that the estimated coupling beam ductility
demands from nonlinear dynamic time-history analyses often exceed the expected
available ductility capacities from the experiments. Therefore, coupled wall structures are
prone to premature failure in the course of a significant seismic event. Limits to the
degree of coupling, DOC, were proposed as follows:
- DOC 50% for conventionally reinforced concrete coupling beams;
- DOC 55% for diagonally reinforced concrete coupling beams;
- DOC 60% for shear critical embedded steel coupling beams; and
- DOC 65% for flexure critical steel coupling beams.
19
The higher coupling levels allowed for steel beams were based on previous
experimental results that these beams are capable of sustaining larger levels of ductility
than reinforced concrete coupling beams.
Harries (2001) also observed that an increase in the degree of coupling causes an
increase in the strength of the beams relative to the strength of the individual walls. If the
degree of coupling is too high, this may lead to an undesirable structural response where
the walls act as the primary energy dissipating components, rather than the coupling
beams. It was argued that in order to ensure that the coupling beams act as the primary
energy dissipating members, the walls should be sufficiently strong to allow all of the
coupling beams to achieve their probable strengths. Thus, a lower limit of 1.5 was
recommended for the wall overstrength factor used in design (applied to the required wall
strength to achieve the probable strengths of the coupling beams).
2.3.4
Degree of Coupling
From the above discussion, it is clear that the degree of coupling, DOC, is an
important parameter and its effect on the behavior of coupled wall systems must be
understood before the successful design of a coupled wall structure can be achieved. As a
part of the NSF-funded U.S.-Japan Cooperative Earthquake Research Program on
Composite and Hybrid Structures, El-Tawil et al. (2002) investigated the effect of the
coupling degree on the behavior of hybrid coupled wall systems comprised of three
reinforced concrete walls with embedded steel beams.
Pushover lateral load analyses of a series of walls with different amounts of
coupling were conducted. It was concluded that moderately coupled hybrid wall systems
(DOC=30% to 45%) are well suited for use in regions of high seismic risk. Overcoupling
(e.g., DOC=60%) can lead to problems due to the development of large shear forces and
compressive axial loads in the compression-side wall, and tension forces in the tensionside wall. Large tension forces, for example, can lead to more widespread cracking and
yielding in the walls, reduce the axial-flexural and shear strength of the walls, and cause
foundation uplift.
El-Tawil et al. (2002) recommended that a target level of coupling should be
selected as part of the design approach in order to ensure structures with desirable seismic
characteristics. Assuming that the coupling beam properties are uniform over the height
of the structure, the coupling axial force acting on the compression-side wall can be
expressed as:
N cwb = nbVb
(2.3)
where, nb is the total number of coupling beams at the floor and roof levels and Vb is the
coupling beam shear strength.
20
Then, with a target design value for the degree of coupling, DOC, design coupled
wall base moment, Mw, and distance between the wall centroids, Lc, known, the required
design shear strength of the coupling beams, Vb can be determined by considering
Equations (2.2) and (2.3) together as follows:
Vb =
2.4
DOC M w
nb Lc
(2.4)
Wall Shear Force Demands in Coupled Wall Structures
Due to the effect of higher modes of response during an earthquake, base shear
force demands of coupled wall structures can be much greater than the forces predicted
from a static analysis (Harries, 2001; Chaallal and Gauthier 2000). This is similar to the
effect of higher modes on the base shear demands of uncoupled wall structures as
investigated by many researchers including Aktan and Bertero (1984, 1985), Keintzel
(1990), Ghosh and Markevicius (1990), Aoyama (1993), Eberhard and Sozen (1993),
Kabeyasawa (1993), and Otani et al. (1994).
In addition, there can be a significant redistribution of shear forces from the
tension-side wall to the compression-side wall, further increasing the shear demands on
individual wall piers. As much as 80% redistribution of shear forces between the tension
and compression side walls was observed by a recent experimental research program on
multi-story reinforced concrete coupled wall structures in Japan (Teshigawara et al. 1998;
Sugaya et al. 2000). Chaallal et al. (2000) investigated the effect of the coupling degree
on the distribution of the total wall base shear between individual wall piers.
The seismic design of a coupled wall structure should consider both of these effects.
More information on the wall shear demands in coupled wall structures is provided in
Chapter 11.
2.5
Unbonded Post-Tensioning in Building Construction
In the last decade, a significant amount of research has been conducted on the use
of unbonded post-tensioning in precast concrete and steel moment frames and precast
concrete walls for building construction in seismic regions. Important findings from this
previous research are summarized in this section. Applications of unbonded posttensioning in bridge construction have also been investigated previously; however, an
overview of this research is outside the scope of this report.
Note that the lateral load behavior of unbonded post-tensioned steel frames and
precast concrete frames is governed by the opening of gaps at the beam-to-column joints,
similar to the opening of gaps at the beam-to-wall joints of the proposed coupled wall
system (see Fig. 1.3b). Thus, there are similarities in the analytical modeling and seismic
design of these different types of structures.
21
2.5.1
Unbonded Post-Tensioning
In unbonded post-tensioned concrete structures, the post-tensioning tendons can be

placed either internal or external to the structural members (e.g., beams). Internal tendons
are placed inside ducts that, unlike the ducts in bonded post-tensioned construction, are
not filled with grout. This way, the bond between the post-tensioning tendons and
concrete is intentionally prevented. Greased tendons may be used to reduce the friction
forces that may develop between the ducts and the tendons. Furthermore, the ducts can be
oversized to prevent the tendons from coming into contact with the ducts during the
lateral displacements of the structure. The elimination of the grouting operation offers
considerable advantages in the application of post-tensioning. Therefore, unbonded posttensioned construction often has a lower unit cost than bonded post-tensioned
construction.
Internal and external unbonded post-tensioned construction types differ in the
displaced shape of the tendon. The displaced shape of internal tendons follows the
displaced shape of the structural member, unless the ducts are sufficiently oversized. The
displaced shape of external tendons is generally different from that of the member except
at deviator or saddle point locations anchored to the structure. Note that the use of
deviator points in seismic applications is not recommended in order to achieve symmetric
behavior under reversed cyclic loading.
There is a fundamental difference in the flexural behavior of an unbonded posttensioned structural member and a bonded post-tensioned member. For a bonded posttensioned member, it is reasonable and customary to assume that the change in the tendon
strain is the same as the change in the concrete strain adjacent to the tendon. In other
words, after grouting, an assumption of strain compatibility between the tendon and the
adjacent concrete can be made. If it is assumed that plane sections remain plane during
flexural deformations, the strain at any point in the tendon can be calculated from the
deformations of the corresponding section (i.e., curvature and average axial strain),
because the change in strain at any point in a bonded tendon depends only on the
deformations at that section.
For an unbonded post-tensioned member, the usual assumption of strain
compatibility between the tendon and the adjacent concrete cannot be made. Rather, the
change in the tendon strain is uniform (ignoring friction forces) over the unbonded length
of the tendon. The change in strain at any point in an unbonded tendon is the average
change in strain in the concrete over the entire unbonded length. In other words, the
tendon strain depends on the total change in the length of the concrete over the unbonded
length. This makes the tendon strain (and stress) dependent on member deformations
rather than section deformations. As a result of the uniform distribution of strains,
unbonded tendons reach the inelastic strain range at larger overall member deformations
than bonded tendons, which is one of the reasons why they are preferred for seismic
applications.
22
The application of unbonded post-tensioning in steel structural members is similar

to the application of external post-tensioning in concrete members. The post-tensioning
tendons can be placed outside or inside (as in the case of hollow cross-sections) the steel
member.
2.5.2
Unbonded Post-Tensioned Precast Concrete Moment Frames
In the last decade, a significant amount of research has been conducted on the
seismic behavior and design of unbonded post-tensioned precast concrete frame
structures (e.g., Cheok and Lew 1991, 1993; Cheok et al. 1993; Priestley and Tao 1993;
MacRae and Priestley 1994; Priestley and MacRae 1996; El-Sheikh et al. 1997, 1999,
2000).
As a part of the PREcast Seismic Structural Systems (PRESSS) Research Program
funded by the National Science Foundation (NSF), the Precast/Prestressed Concrete
Institute (PCI), and the Precast/Prestressed Concrete Manufacturers Association of
California (PCMAC), Priestley and Tao (1993) analytically investigated the concept of
joining precast concrete beam and column members using post-tensioning tendons
unbonded over a length of each beam through and on either side of the column as shown
in Figs. 2.7a and 2.7b. The nonlinear behavior of these structures occurs primarily due to
the opening of gaps at the beam-to-column interfaces (Fig. 2.7c).
Cheok and Lew (1991, 1993) and Cheok et al. (1993) conducted reversed cyclic
lateral load experiments of 1/3-scale unbonded post-tensioned precast interior beamcolumn joint subassemblages at the National Institute of Standards and Technology
(NIST). MacRae and Priestley (1994) and Priestley and MacRae (1996) conducted
additional cyclic lateral load experiments on two 2/3-scale unbonded post-tensioned
precast beam-column subassemblages representing an interior and an exterior joint. The
subassemblages from this research program are illustrated in Figs. 2.8a and 2.8b with the
test setup shown in Figs. 2.8c and 2.8d. El-Sheikh et al. (1997, 1999, 2000) developed
analytical models and a performance based seismic design approach for multi-story
unbonded post-tensioned precast concrete frame structures.
Some of the important conclusions from the previous research on unbonded posttensioned precast concrete frame structures are summarized below.
(1) As a result of post-tensioning, the initial linear-elastic lateral stiffness of an
unbonded post-tensioned precast concrete beam-column joint is similar to the initial
stiffness of a monolithic cast-in-place reinforced concrete joint.
(2) Unbonded post-tensioned precast concrete frames can soften (indicating a
significant reduction in the lateral stiffness) and go through large nonlinear cyclic lateral
displacements without significant damage, and thus, without significant strength
degradation (as compared to precast concrete frames with bonded tendons or monolithic
23
cast-in-place reinforced concrete frames). The lateral load versus displacement behaviors
of the interior and exterior beam-column joint specimens tested by Priestley and MacRae
(1996) are shown in Figs. 2.8e and 2.8f, respectively, with the interior joint at 2.8% drift
shown in Fig. 2.8g. The nonlinear displacements occur primarily through the opening of
gaps at the beam-to-column interfaces as shown in Fig. 2.7c.
(3) The use of unbonded post-tensioning tendons results in a uniform strain
distribution in the steel, and thus, delays the nonlinear straining (i.e., yielding) of the
tendons. As long as the post-tensioning tendons remain elastic and the anchors for the
tendons do not deteriorate, the initial level of prestress can be maintained during the
cyclic lateral displacements of the structure.
beam
column
(a)
precast
column
unbonded
post-tensioning
gap opening at
beam-to-column
interfaces
precast beam
bonded length
(b)
(c)
Fig. 2.7 Unbonded post-tensioned precast concrete frame: (a) elevation;

(b) beam-column subassemblage; (c) idealized exaggerated displaced shape.
24
(b)
(a)
(c)
(d)
(e)
(f)
(g)
Fig. 2.8 Unbonded post-tensioned precast beam-column subassemblages (from

Priestley and MacRae 1996): (a) interior joint specimen; (b) exterior joint specimen;
(c) test setup; (d) photograph of test setup; (e) interior joint behavior;
(f) exterior joint behavior; (g) photograph of interior joint
25
(4) Unbonding of the post-tensioning tendons significantly reduces the amount of

tensile stresses transferred to the concrete, thus reducing damage due to concrete cracking.
(5) Large compression strains develop at the corners of the beams near the beamto-column interfaces due to gap opening and post-tensioning. Therefore, transverse
reinforcement (e.g., spiral or closed hoop reinforcement) is necessary to confine the
concrete near the ends of the beams (see test specimen details with spiral reinforcement
in Figs. 2.8a and 2.8b).
(6) Upon unloading from a large nonlinear lateral displacement, the post-tensioning
force provides a restoring effect (i.e., self-centering capability) that tends to close the
gaps and pull the frame back towards its original undeformed (i.e., plumb) position
without significant residual lateral displacements. The resulting hysteretic behavior of the
structure is close to nonlinear-elastic as indicated by the loading and unloading cycles
that are close to each other for the subassemblages in Figs. 2.8e and 2.8f.
(7) The friction resistance that develops due to post-tensioning is sufficient to
prevent shear slip at the beam-to-column interfaces and to transfer the shear forces from
gravity and lateral loads without the need for corbels or shear keys at the beam ends. As
long as the post-tensioning force is maintained, the shear slip resistance at the beam-tocolumn interfaces is maintained during large cyclic lateral displacements of the structure.
(8) Unbonded post-tensioning alters the shear resistance mechanism in the beams
and in the beam-to-column joint panel regions (as compared with bonded post-tensioned
and monolithic cast-in-place reinforced concrete structures), in which the shear force is
transferred by a diagonal compression strut, greatly simplifying the design of the beam
and joint shear reinforcement and reducing damage.
(9) Because of the small amount of damage, the inelastic energy dissipation of
unbonded post-tensioned precast concrete frame structures is small (see Figs. 2.8e and
2.8f) as compared to monolithic cast-in-place reinforced concrete frame structures.
(10) The nonlinear moment versus rotation behavior of a properly-designed
unbonded post-tensioned precast beam-column joint subassemblage can be idealized
using a trilinear relationship as shown in Fig. 2.9a from El-Sheikh et al. (1999). The five
limit states marked on the smooth moment-rotation relationship (developed using a fiber
element based analytical model) are as follows: Point (1) represents the initiation of gap
opening (i.e., decompression) at the beam-to-column interfaces; Point (2) represents a
significant reduction in the stiffness of the subassemblage (i.e., softening), which occurs
primarily due to increased gap opening at the beam-to-column interfaces; Point (3)
represents the initiation of cover concrete spalling at the beam ends; Point (4) represents
the initiation of yielding (i.e., nonlinear straining) of the unbonded post-tensioning steel;
and Point (5) represents the axial-flexural failure of the subassemblage due to the
crushing of the confined concrete at the beam ends.
26
(a)
(b)
Fig. 2.9 Unbonded post-tensioned precast beam-column subassemblages (from El-Sheikh

et al. 1999, 2000): (a) moment-rotation behavior; (b) analytical model
Fig. 2.9b shows the analytical model of an interior unbonded post-tensioned precast
concrete beam-column joint subassemblage developed by El-Sheikh et al. (1999, 2000).
This model, which was used to generate the smooth subassemblage moment-rotation
relationship in Fig. 2.9a, includes the following elements in the DRAIN-2DX structural
analysis program (Prakash et al. 1993): (1) fiber beam-column elements to model the
lengths of the beams with unbonded post-tensioning steel; (2) linear-elastic frame beamcolumn elements to model the column as well as the lengths of the beams with bonded
post-tensioning steel, where only linear-elastic deformations are expected to occur; (3)
truss elements to model the unbonded length of the post-tensioning tendons; (4) a zerolength rotational spring element to model the joint panel zone shear deformations of the
column member; (5) kinematic constraints (rigid links) for the beam elements and rigid
end zones for the column elements to model the axial and flexural stiffnesses (assumed to
be infinitely large) within the panel zone; and (6) kinematic constraints to model the ends
of the unbonded length of the post-tensioning tendons. The gap opening behavior at the
beam-to-column interfaces is modeled using compression-only stress-strain relationships
for the fibers at the beam ends. The results obtained from the analytical model were
found to compare well with experimental results reported by Cheok and Lew (1993).
The greatest setback to the use of unbonded post-tensioned precast concrete frame
structures in seismic regions is that their lateral displacement demands during a severe
earthquake may be larger than acceptable as a result of small energy dissipation. Previous
research (e.g., Priestley and Tao 1993) has shown that the peak lateral displacements of
these structures can be, on average, 1.35-1.40 times the peak displacements of
comparable monolithic cast-in-place reinforced concrete frame structures. In order to
reduce the lateral displacement demands during a seismic event, the use of bonded mild
(e.g., Grade 60) steel reinforcement through the precast beam-column joints, in addition
to the unbonded post-tensioning steel, has been investigated (Cheok and Stone 1994;
Cheok et al. 1996, 1998; Stanton et al. 1997; Stone et al. 1998; Nakaki et al. 1999;
Priestley et al. 1999; Stanton and Nakaki 2002).
27
These partially prestressed systems are often referred to as hybrid precast

concrete frame structures due to the mixed use of mild steel and post-tensioning steel
reinforcement across the beam-column joints as shown in Fig. 2.10. Properly designed
and detailed mild steel reinforcement would yield in tension and compression during the
cyclic gap opening displacements that occur at the beam-to-column interfaces, thus
dissipating energy. The bond between the mild steel bars and the concrete may be
prevented over a predetermined length at the ends of the beams (by wrapping the bars) to
prevent fracturing of the mild steel and to reduce the cracking of the concrete during the
deformations of the bars in tension.
column
mild steel bars

(transverse reinf.
not shown)
section A-A
fiber reinforced grout
wrapped length
mild steel bar

beam
trough
PT tendon
beam
beam-to-column interface
unbonded
length
central duct
for PT tendon
section B-B
Fig. 2.10 Partially post-tensioned precast beam-column subassemblages

(adapted from Kurama 2002)
The concept of combining post-tensioning steel for flexural strength and selfcentering with mild steel for inelastic energy dissipation was recently demonstrated
through 0.60-scale tests of a five-story precast concrete building structure as part of the
final phase of the PRESSS (PREcast Seismic Structural Systems) Research Program
(Nakaki et al. 1999; Priestley et al. 1999), leading to the development of design
recommendations (ACI Innovation Task Group 1 and Collaborators 2001a, 2001b, 2003)
and the successful application of this new structural system in a 39-story precastprestressed concrete apartment building in San Francisco, California (Englekirk 2002).
Note that in most recent applications, the post-tensioning tendons are typically left
unbonded over the entire length of the frame (i.e., without any bonded regions within the
beam spans) and are anchored only at the outer ends of the exterior joints.
Additional efforts to reduce the seismic displacement demands of post-tensioned
precast concrete frame structures include a new type of friction damper developed for use
locally at the beam ends (Morgen and Kurama 2003, 2004a, 2004b, 2005).
28
2.5.3
Post-Tensioned Steel Moment Frames
The use of post-tensioning in steel moment resisting frames for seismic regions was
recently investigated by Peng et al. (2000), Ricles et al. (2000, 2001, 2002), Peng (2001),
Christopoulos et al (2002), Garlock (2002), and Rojas et al. (2002). Similar to the
unbonded post-tensioned precast concrete frame system described above, these frame
structures are constructed by post-tensioning steel beams to steel columns using high
strength post-tensioning strands. The strands extend over the entire length of the frame
and are anchored only at the outer ends of the exterior joints as shown in Fig. 2.11. Top
and seat angles are used at beam-to-column connections to provide energy dissipation
and redundancy under seismic loading (analogous to the use of mild steel reinforcement
in the partially prestressed precast concrete system in Fig. 2.10).
The nonlinear behavior of post-tensioned steel frame structures is governed by the
opening of gaps at the beam-to-column interfaces, similar to the gaps shown in Figs. 1.3b
and 2.7c. Reinforcing cover plates are used at the beam ends (analogous to the spiral
concrete confinement in Fig. 2.8) to strengthen and control the yielding of the flanges
under the high compressive stresses that develop due to post-tensioning and gap opening.
Shim plates are used between the columns and the beam flanges to force the bearing
between the beams and the columns to occur at the flanges, so that one or both flanges
remain in contact with the column after the flanges yield in compression under cyclic
loading. The shim plates are terminated at the beam web, thus preventing contact
between the web and the column. Moreover, Ricles et al. (2001) used washer plates in the
angle-to-column connections (see Fig. 2.11) to better control the location of the plastic
hinge in the angle vertical leg near the column bolts and to reduce the prying forces on
the bolts. The role of the washer plates on the behavior of the angles is described later.
The beams in a post-tensioned steel moment frame system are similar to the
coupling beams investigated in this report. The primary differences between these two
types of beams are: (1) the beams in a frame structure are post-tensioned to adjacent steel
columns with the post-tensioning steel anchored at the outer ends of the frame; whereas,
the beams in a coupled wall structure are post-tensioned to adjacent concrete walls; and
(2) the beam length to depth aspect ratios in a frame structure are typically much larger
than the beam aspect ratios in a coupled wall structure.
As a result of the similarities between these two types of post-tensioned steel beams,
the previous experimental and analytical research on post-tensioned steel frame structures
is important for the research described in this report. The experimental program
conducted by Ricles et al. (2001, 2002) involved the testing of full-scale interior beamcolumn subassemblages. Each cruciform shaped specimen (Fig. 2.12) consisted of two
wide flange steel beams post-tensioned to a steel column. Several parameters were
investigated including angle leg thickness, angle gage length, beam flange cover plates,
connection shim plates, and post-tensioning force.
29
shim plate
washer plate
column
angle connection bolts
beam
post-tensioning
anchors
post-tensioning strand
flange cover plate

angle
washer plate
Fig. 2.11 Post-tensioned steel exterior beam-column connection

(adapted from Ricles et al. 2001)
The experimental results from Ricles et al. (2001) shown in Fig. 2.13 demonstrate
that properly-designed post-tensioned steel frame beam-column subassemblages possess
exceptional cyclic lateral strength and ductility, where yielding mainly occurs in the top
and seat angles while the beam and column members remain essentially linear-elastic.
The initial linear-elastic lateral stiffness of a post-tensioned steel beam-column
connection is comparable to that of a conventional fully-restrained (welded) connection,
and no significant permanent deformations remain in the structure following the
application of several nonlinear cycles of drift (indicating a large self-centering capability
due to the post-tensioning force). The compression yielding that occurs in the beam
flanges should be controlled in order to maintain the lateral stiffness, strength, and selfcentering capability of the structure. The angle geometry has a significant effect on the
connection moment capacity, energy dissipation, and ductility.
Ricles et al. (2002) used the subassemblage experimental results to develop simple
procedures for estimating the connection behavior. Similarly, Ricles et al. (2001) used
DRAIN-2DX (Prakash et al. 1993) to develop an analytical model for post-tensioned
steel frame beam-column subassemblages. As an example, Fig. 2.14 shows the model for
an interior subassemblage. Fiber beam-column elements are used to represent the beam
and column members, and two parallel bilinear truss elements are used to represent each
angle. The modeling of the top and seat angles is discussed in more detail later. The posttensioning tendons in the subassemblage are also modeled using truss elements. The posttensioning anchorages are represented by kinematically constraining (i.e., slaving) the
displacements of the nodes at the ends of the truss elements to corresponding nodes at the
ends of the beam fiber elements (Fig. 2.14). The flexibility of the column panel zone in
30
shear is modeled by placing a zero-length rotational spring element at the center of the
panel. The moment-rotation behavior of the rotational spring represents the shear
stiffness of the panel zone. Gap opening behavior at the beam-to-column interfaces is
modeled by using compression-only stress-strain relationships for the steel fibers near the
beam ends. This is similar to the modeling of gap opening in unbonded post-tensioned
precast concrete frame and wall structures (El-Sheikh et al. 1999, 2000; Kurama et al.
1996, 1999a) and the modeling of gap opening at the beam-to-wall interfaces of the
hybrid coupled wall structure investigated in this report.
2.5.4
Unbonded Post-Tensioned Precast Concrete Structural Walls
As part of the NSF-funded PRESSS Research Program, Kurama et al. (1996, 1997,
1998a, 1998b, 1999a, 1999b, 2002) conducted an analytical investigation on the seismic
behavior and design of structural walls that are constructed by post-tensioning precast
concrete wall panels across horizontal joints using unbonded high strength posttensioning bars (Fig. 2.15a). The expected seismic characteristics of properly-designed
walls were found to be similar to the behavior of unbonded post-tensioned precast
concrete frames described previously, with an ability to soften and undergo large
nonlinear lateral displacements with little damage. The nonlinear displacements of the
walls occur primarily due to the opening of gaps at the horizontal joints between the wall
panels and between the base panel and the foundation, as shown in Fig. 2.15b. Shear slip
along the horizontal joints (Fig. 2.15b) is not a desired mode of lateral displacement and
is prevented by design.
An analytical model (Fig. 2.15c) was developed using the DRAIN-2DX program
(Prakash et al. 1993) to investigate the lateral load behavior of the walls (Kurama et al.
1996, 1999a). Fiber beam-column elements were used to model the precast concrete wall
panels and truss elements were used to model the unbonded post-tensioning bars.
Anchorages to the post-tensioning bars were modeled by kinematically constraining (i.e.,
slaving) the displacements of the truss element nodes to the wall fiber element nodes at
the roof and at the foundation. Gap opening behavior along the horizontal joints was
represented by using compression-only stress-strain relationships for the fibers used to
model the wall panels.
The fiber wall model in Fig. 2.15c accounts for axial-flexural interaction, nonlinear
hysteretic behavior of the post-tensioning steel and concrete including crushing of
concrete, and gap opening behavior along the horizontal joints. Comparisons of analysis
results obtained from the fiber wall model with results obtained from ABAQUS (Hibbitt,
Karlsson, and Sorensen) finite element models that use gap/contact elements to model
gap opening (Allen and Kurama 2002a, 200b) indicate that the fiber wall model is
capable of predicting both the global (e.g., base shear versus roof lateral displacement)
behavior and the local (e.g., gap opening and contact) behavior of the walls reasonably
well (Kurama 2000).
31
Fig. 2.12 Post-tensioned steel beam-column connection details (from Ricles et al.
2002): (a) specimen with wide flange column; (b) specimen with composite concrete
filled steel tube column
32
Fig. 2.13 Lateral load versus displacement behavior of post-tensioned steel beam-column
subassemblages (from Ricles et al. 2001)
(a)
(b)
(c)
dc
node ES is slaved to node BM
in translational degrees
of freedom
fiber
element
post-tensioning strands
(truss elements)
panel zone
spring
ES
db
BM
db
beam fiber
element
fiber element
end segment
beam outline
parallel
springs
master node
slaved node
Fig. 2.14 Analytical modeling of post-tensioned steel beam-column subassemblages

(adapted from Ricles et al. 2001): (a) interior joint region; (b) beam-column
subassemblage; (c) modeling of post-tensioning anchorages
33
PT anchorage
wall panel
unbonded PT
bar
horizontal joint
spiral
reinforcement
foundation
spiral
reinforcement
PT duct and
unbonded PT
bar
bonded
wire mesh
tw
precast wall
gap opening
shear slip
lw
(a)
(b)
node
truss element
kinematic
constraint
fiber element
base shear force (kN)
4000
-4000
-2
-1
(c)
0
1
roof drift (percent)
(d)
roof drift (percent)
Hollister
PGA=1.0g
unbonded PT precast wall
-4
15
30
time (seconds)
(e)
Fig. 2.15 Unbonded post-tensioned precast concrete walls (adapted from Kurama et al.
1999a): (a) wall elevation and cross-section; (b) gap opening and shear slip behavior;
(c) analytical model; (d) behavior under cyclic lateral loads; (e) dynamic response
34
A performance based seismic design approach for the walls was developed by
Kurama et al. (1997, 2002) with two objectives: (1) to resist design level ground motions
without significant damage; and (2) to resist severe survival level ground motions with
damage, but without collapse. The seismic response of a series of prototype walls was
evaluated based on nonlinear static push-over and reversed cyclic lateral load analyses as
well as nonlinear dynamic time-history analyses. As an example, Fig. 2.15d shows the
expected base shear force versus roof lateral drift (i.e., roof lateral displacement divided
by the wall height) behavior of a six-story wall designed for a region with high seismicity
and a site with a medium soil profile (corresponding to Site Class D in IBC 2000)
under cyclic lateral loads (Kurama et al. 1997). The behavior of the wall is nearly
nonlinear-elastic with a large self-centering capability (i.e., small residual displacements)
but with little inelastic energy dissipation.
The dynamic analyses were conducted under an ensemble of recorded and
generated ground motions selected for the seismic regions and site soil characteristics
used in the design of the walls (Kurama et al. 1997). As an example, Fig. 2.15e shows the
roof drift time history of the wall in Fig. 2.15d under the Hollister ground motion
(recorded on a site with a medium soil profile during the 1989 Loma Prieta earthquake),
scaled to a peak acceleration of 1.0g to represent a survival level ground motion. Based
on the results of the dynamic analyses, recommendations were made for the peak
nonlinear lateral displacement demands and peak base shear demands for the walls.
Based on these analytical investigations, Kurama et al. (1996, 1997, 1998a, 1998b,
1999a, 1999b, 2002) concluded that unbonded post-tensioned precast concrete walls
provide a feasible primary lateral load resisting structural system for use in regions with
high seismicity. Similar to unbonded post-tensioned precast frames, the greatest
disadvantage of these wall structures is that their lateral displacement demands during a
severe earthquake may be larger than acceptable as a result of small inelastic energy
dissipation. In order to reduce the seismic displacement demands, Perez (1998), Perez et
al. (2004a, 2004b), and Priestley et al. (1999) investigated the use of ductile yielding
connectors along vertical joints between two or more walls. The use of supplemental
viscous fluid-dampers and friction dampers to reduce the displacements of the walls was
investigated by Kurama (2000, 2001). More recently, partially post-tensioned walls that
use bonded mild (e.g., Grade 60) steel reinforcement crossing the horizontal joints in
addition to the unbonded post-tensioning steel were investigated experimentally by
Rahman and Restrepo (2000), Holden et al. (2001, 2003), and Restrepo (2003), and
analytically by Kurama (2002, 2004). Additional research on unbonded post-tensioned
walls with supplemental energy dissipation includes an analytical study on the use of
dampers in series with the post-tensioning tendons at the wall base, including different
tendon profiles (Ajrab et al. 2004).
Kurama (2002) conducted an analytical parameter study to investigate the effect of
mild steel reinforcement on the lateral displacement response of unbonded post-tensioned
precast concrete walls during earthquakes. A total of 12 wall structures with different
35
amounts of mild steel and post-tensioning steel reinforcement were analyzed considering
regions with high seismicity (e.g., coastal California) as well as regions with moderate
seismicity (e.g., eastern U.S.). The effect of the number of stories on the behavior of the
walls was examined by including ten-story, six-story, and four-story walls in the
investigation.
It was shown that the energy dissipation of the walls can be increased considerably
by using mild steel reinforcement, while maintaining a high level of self-centering
capability. The use of mild steel has two important advantages on the dynamic behavior
of the walls: (1) the maximum roof drift decreases; and (2) the number of large roof drift
cycles decreases because the response of the wall decays faster. Based on the parametric
investigation, a seismic design approach was developed for partially post-tensioned
precast walls (Kurama 2004), with the main objective of limiting the lateral
displacements of a wall to a predetermined allowable displacement.
Rahman and Restrepo (2000), Holden et al. (2001, 2003), and Restrepo (2003)
tested three half-scale partially post-tensioned precast concrete wall specimens,
representative of a four story building under quasi-static reversed cyclic lateral loading.
Fig. 2.16a shows the test set-up. Two of the specimens were identical, with the exception
that gravity load was applied on one specimen but not on the other. The specimens were
reinforced with two 0.5 inch diameter low relaxation post-tensioning strands and two 62
ksi mild steel bars crossing the horizontal joint between the wall and the foundation. The
mild steel bars had a diameter of 0.79 in. but they were milled down to a diameter of 0.63
in. near the base of the wall in order to control yielding. The third specimen used steel
fiber reinforced concrete for the wall and unbonded carbon fiber tendons instead of highstrength steel tendons to apply the post-tensioning force. A fourth specimen was also
tested as a code compliant conventionally reinforced specimen designed to emulate the
behavior of a ductile monolithic cast-in-place reinforced concrete wall.
The experimental results showed that the specimen emulating monolithic cast-inplace reinforced concrete construction dissipated more energy than the partially posttensioned systems; however, it sustained significant structural damage and residual
displacements (see Fig. 2.16b). In contrast, the amount of structural damage was
markedly reduced in the partially post-tensioned specimens and the residual
displacements upon unloading were negligible. Figs. 2.16c and 2.16d show the hysteretic
lateral load versus drift behavior of one of the partially post-tensioned specimens and the
extent of damage in the structure at 3% drift (Restrepo 2003).
The analytical and experimental research efforts summarized above have recently
led to the development of draft acceptance criteria for the use of post-tensioned precast
concrete structural walls in seismic regions (ACI Innovation Task Group and
Collaborators 2004).
36
load cell
5000
4500
4200
units: mm
1350
PT anchor actuator
support to
apply axial
load
load cell
PT ducts
ducts for mild

steel bars
shear dowels
foundation
beam
PT anchor
1350
tie-down
650
500
0
3000
4960
(a)
(b)
lateral force, kN
120
-120
-4.0
0
lateral drift, %
5.0
(c)
(d)
Fig. 2.16 Experiments of precast concrete walls (from Holden et al. 2003 and Restrepo
2003): (a) test setup; (b) hysteretic response of emulative wall; (c) hysteretic response of
partially post-tensioned wall; (d) photograph of partially post-tensioned wall at 3% drift
2.6
Behavior of Top and Seat Angles
As discussed in the preceding sections, it is possible to reduce the lateral

displacement demands of unbonded post-tensioned frame and wall structures by using
supplemental passive energy dissipation, without losing the desirable characteristics of
these structures. Using steel top and seat angles at the frame or wall joints where gap
opening occurs is one way of providing energy dissipation and increasing structural
redundancy as shown in Figs. 1.3 and 2.11.
37
A significant amount of previous research has been conducted on steel frame beamto-column connections that use top and seat angles (e.g., Azizinamini 1985; Aktan et al.
1989; Youssef-Agha et al. 1989; Kishi and Chen 1990; Lorenz et al. 1993; Bernuzzi et al.
1994, 1997; Leon and Shin 1994; Mander et al. 1994; Sarraf and Bruneau 1996; Bernuzzi
1998; Kasai 1998; Kukreti and Abolmaali 1999; Swanson and Leon 1999; Shen and
Astaneh-Asl 1999, 2000; Sims 2000; Ricles et al. 2000, 2001; Garlock 2002; Garlock et
al. 2003). Fig. 2.17a shows the idealized exaggerated deformed configuration of a topand-seat-angle steel frame connection, which is a semi-rigid connection comprised of a
pair of steel angles bolted to the flanges of the beam and the column. Fig. 2.17b shows
some of the design parameters for the connection angles.
vertical leg
tension angle
horizontal leg
ax
vertical leg
beam
column
wabw
bolt head
innermost bolt
bolts
db
fillet
loading lgv
horizontal leg
ka
ta
compression angle
heel
(a)
l gh
(b)
Fig. 2.17 Top-and-seat angle connection:

(a) deformed configuration; (b) angle parameters
The flexural stiffness, strength, and ductility of a top-and-seat-angle beam-tocolumn connection are influenced by many design parameters, including: (1) beam depth,
db; (2) angle length, la (which is usually equal to the beam flange width); (3) angle leg
thickness, ta; (4) angle steel yield strength, fay; (5) angle-to-column connection gage
length, lgv, measured from the heel of the angle to the center of the innermost angle-tocolumn connectors; (6) angle-to-beam connection gage length, lgh, measured from the
heel of the angle to the centroid of the angle-to-beam connector bolt group; (7) angle
fillet length, ka, measured from the angle heel to the toe of the angle fillet; (8) angle
connector bolt type and diameter, dab; (9) bolt head/nut width measured across flat sides,
wab; and (10) number of bolts, nab. The previous research conducted by Kishi and Chen
38
(1990) and Lorenz et al. (1993), Sims (2000), Ricles et al. (2001), and Garlock et al.
(2003) on the behavior of the connection angles is described below.
2.6.1
Kishi and Chen (1990) and Lorenz et al. (1993)
Kishi and Chen (1990) and Lorenz et al. (1993) developed a model for the analysis
and design of steel top-and-seat-angle beam-to-column connections based on equilibrium
considerations with assumed internal force distributions and boundary conditions for the
angles. An overview of their model is given below since it is used to describe the
behavior of the angles at the beam-to-wall connections in this report. Other researchers
(e.g., Goto et al. 1991; Matsuoka et al. 1993) have also used this model to investigate the
behavior of semi-rigid steel frame connections. Several other similar top-and-seat angle
connection models exist (e.g., Sarraf and Bruneau 1996; Aktan et al. 1989).
Note that the model developed by Kishi and Chen (1990) and Lorenz et al. (1993)
can only be used to estimate the initial stiffness and yield force capacity of the tension
angle (see Fig. 2.17a) in a top-and-seat angle connection under monotonic loading. The
post-yield behavior of the tension angle, the behavior of the compression angle, and the
behavior of the connection under cyclic loading are not addressed in their investigation.
Tension angle initial stiffness, Kaixt
The initial linear-elastic stiffness of the tension angle is determined by assuming
that the vertical leg of the angle is fixed along the innermost edge of the line of angle-tocolumn connection bolt heads and is pulled as a cantilever by the beam flange as shown
in Fig. 2.18. Considering the shear deformations of the vertical angle leg, it can be shown
that the initial linear-elastic stiffness of the angle, Kaixt is given as:
K aixt =
Tax
3Ea I a
=
2
ax l g1(l g1
+ 0.78ta2 )
(2.5)
where, Tax is the force in the tension angle parallel to the beam, ax is the horizontal
displacement of the heel of the angle (see Fig. 2.17a), EaIa is the flexural stiffness of the
angle vertical leg cross section, ta is the thickness of the angle, and lg1 is the length of the
vertical leg that is assumed to act as a cantilever beam. Using centerline dimensions for
the angle legs and referring to Fig. 2.18a,
l g1 = l gv
wabw ta
2
2
(2.6)
where, wabw is the width (across flats) of the angle-to-column connector bolt head/nut and
lgv is the angle vertical leg connection gage length measured from the heel of the angle to
the center of the innermost angle-to-column connectors as shown in Fig. 2.18a.
39
vertical leg
wabw
bolt head
innermost bolt
lgv
fixed support
l g1
ka
ta
ka
(a)
lgv
plastic
hinges
g2
ka
ayx
ayx
(b)
N ap
Map
Vap
Tayy
Tayx
Ma
(c)
Fig. 2.18 Angle model (adapted from Kishi and Chen 1990 and Lorenz et al. 1993):
(a) cantilever model of tension angle; (b) assumed yield mechanism;
(c) free body diagram of angle horizontal leg
40
Tension angle yield force capacity, Tayx

The yield force capacity, Tayx of the tension angle is determined based on an
assumed plastic hinge mechanism in the vertical leg as shown in Fig. 2.18b. One of the
plastic hinges is located along the innermost edge of the angle-to-column connection bolt
heads/nuts and the other plastic hinge is located along the toe of the fillet. The work
equation for the tension angle at this plastic mechanism state is given by
2M ap = Vapl g 2
(2.7)
where, Map is the plastic moment and Vap is the plastic shear force in the vertical leg, and
is the plastic hinge rotation as shown in Fig. 2.18b. Based on the free body diagram of
the angle horizontal leg in Fig. 2.18c, the tension angle yield force, Tayx is equal to the
plastic shear force Vap in the vertical leg.
The distance lg2 in Equation (2.7) is the effective gage length for the assumed
plastic hinge mechanism equal to the distance between the two plastic hinges as:
l g 2 = l gv
wabw
ka
2
(2.8)
where, wabw is the width of angle-to-column connection bolt head/nut measured across
flat sides and ka is the distance from heel to toe of fillet of the angle. Since lg2 is usually
rather short, a shear-flexure interaction equation is considered in the calculation of the
capacity of the angle as follows:
M ap
M a0
Vap
=1
+
Va 0
(2.9)
where, Ma0 and Va0 are the plastic moment and plastic shear force in the vertical leg
without considering shear-flexure interaction, respectively as
f ay la (ta )2
M a0 =
Va 0 =
f ay lata
2
(2.10)
(2.11)
in which fay is the yield strength of the angle steel and la is the length of the angle.
Substituting Equations (2.7), (2.10), and (2.11) into Equation (2.9), the angle yield force
Tayx=Vap can be determined from
41
l g 2 Vap Vap
+
=1
ta Va 0 Va 0
(2.12)
Note that in Fig. 2.18c, the horizontal leg of the angle is cut along the innermost
edge of the angle-to-beam connection bolt heads/nuts. At this location, it is assumed that
the moment Ma is small, and is ignored. Moreover, the shear force of the horizontal leg,
Tayy and the corresponding axial force, Nap at the plastic hinge are also ignored.
2.6.2
Sims (2000)
Sims (2000) conducted a parametric finite element investigation of top-and-seatangle beam-to-column connections. The column was assumed to be rigid and fixed, and
reversed cyclic transverse displacements were applied at the beam end. A companion
experimental program was also conducted. The finite element model, which was
developed using the ANSYS program (Swanson Analysis Systems 1997), included
material and geometric nonlinearities and captured aspects of bolt behavior such as preload, slip, axial and flexural flexibility, and inelasticity.
Five failure modes for the angles were identified from the finite element analyses:
(1) bolt thread failure; (2) bolt/angle flexural mechanism; (3) shear mechanism; (4)
vertical-leg/horizontal-leg combination mechanism; and (5) vertical-leg mechanism. The
vertical-leg mechanism is similar to the failure mechanism in Fig. 2.18b. According to
Sims (2000), this failure mechanism represents an ideal mode with a relatively more
uniform distribution of strain, resulting in the lowest maximum strain in the angle. It was
found that for angle configurations with vertical-leg mechanism failure modes, the model
developed by Kishi and Chen (1991) and Lorenz et al. (1993) can be used to estimate the
angle yield force, Tayx. However, since the plastic deformations in the horizontal leg (i.e.,
leg parallel to the beam) and the effect of bolt response are neglected in this analytical
model, the angle yield force for the other failure modes can not be estimated especially
for angle configurations with non-negligible bolt response (e.g., bolt elongation, slip,
yielding, fracture).
The following additional conclusions were made based on the finite element
analyses:
(1) Bolt response reduces the strength, stiffness, and ductility of the angle system.
Due to the less ductile properties of high strength bolts compared to angle sections,
configurations with bolt fracture are less ductile than those with angle fracture. Thus,
failure of the bolts should be avoided.
(2) For short effective gage length lg2 values, the maximum strain in the angle
occurs along the innermost edge of the vertical leg connection bolt heads/nuts. For long
lg2 values, the maximum strain occurs along the toe of the fillet in the vertical leg. For
42
intermediate lg2 values, the maximum strain in the angle occurs along the toe of the fillet
in the horizontal leg (leg parallel to the beam). Strong and ductile top and seat angle
configurations exist between lg2 values of 1 to 2 in. This range represents an actual gage
length lgv of between 3 to 4 in., depending on the angle and bolt sizes. In this range,
optimal efficiency of the angle in terms of the amount of angle strength per plastic strain
(i.e., maximum force capacity of the angle for a given plastic strain demand) could be
reached.
(3) The boundary condition of the horizontal leg connection bolts is an important
factor affecting the behavior of the angles. Neglecting the rotations of the horizontal leg
results in an overestimation of the tension angle yield force capacity.
(4) Catenary effects at large angle deformations play a very significant role on the
angle behavior. Inclusion of large deflections (i.e., second order geometric effects) in the
analytical model is necessary to accurately capture these effects.
The isolated angle experiments conducted by Sims (2000) showed good agreement
with the finite element analytical results. The plan view of the experimental setup is
shown in Fig. 2.19a. Two angles were placed back-to-back and loaded against a support
fixture modeling a rigid column, where the applied force of 2Tax represents the combined
resistance of the angle pair. Three monotonic and eleven cyclic loading experiments were
conducted. Figs. 2.19b and 2.19c show representative angle force versus deformation
(Tax-ax) relationships from two cyclic experiments, each with a vertical-leg angle
failure mechanism. One of the important conclusions from this experimental program
was that bolt slip in the angle horizontal leg is a cause of significant stiffness reduction of
the angle. One option for eliminating this bolt slip is to weld the angle horizontal leg to
the beam flange. In the absence of this measure, a large number of slip critical bolts may
be required for the angle-to-beam connections.
2.6.3
Ricles et al. (2001)
As described previously in Fig. 2.14, Ricles et al. (2001) modeled each angle in
their post-tensioned steel beam-column connection using two parallel bilinear truss
element springs in the DRAIN-2DX program (Prakash et al. 1993). The truss element
springs represent the behavior of the angle when gap opening occurs at the beam-tocolumn interface. Elongation and shortening of the springs result in forces that are
equivalent to the shear force in the vertical leg of the angle. The combined forcedeformation behavior of the two parallel bilinear springs produces a trilinear forcedeformation relationship for each angle as described below.
Ricles et al. (2001) determined the properties of the parallel spring model by fitting
a trilinear curve to the expected force-deformation behavior of the angles, where the
angle deformation is equivalent to ax in Fig. 2.17. The expected angle force-deformation
behavior was obtained from analyses of a model of the angle using fiber beam-column
43
elements in the DRAIN-2DX program. Each angle leg was modeled using one fiber
element having several segments, accounting for shear and flexural deformations. Fig.
2.20a shows two angle fiber element models (Models A and B) and the loading and
boundary conditions used to determine the angle force-deformation behavior. The effects
of prying forces were anticipated to be small due to the relatively long angle gage lengths,
and thus, were ignored.
strong floor
loading tee
support
fixture
displacement, ax
actuator
test angles
applied force, 2Tax

load cell
bolts
plan view
(a)
(b)
(c)
Fig. 2.19 Isolated angle experiments (from Sims 2000): (a) plan view of test setup;
(b) angle force-deformation relationship of Specimen 9;
(c) angle force-deformation relationship of Specimen 11
The angle force-deformation behaviors from Models A and B and the
corresponding trilinear parallel spring models are shown in Fig. 2.20b. The accuracy of
the angle models was verified by comparing the analytical results with experimental
results for post-tensioned steel beam-column connection subassemblages (Fig. 2.13a). As
shown in Fig. 2.20c, the analytical results for the subassemblages were found to be not
too sensitive to the differences between Models A and B. The behavior and modeling of
the angles under cyclic loading are not discussed in detail by Ricles et al. (2001).
44
Fig. 2.20 Modeling of top and seat angles (from Ricles et al. 2001): (a) fiber Models A
and B; (b) fiber model and spring model behaviors; (c) comparisons with test results.
2.6.4
Garlock et al. (2003)
Garlock et al. (2003) experimentally investigated the behavior of the angles in a

bolted top-and-seat angle beam-to-column connection under cyclic loading. Seven
specimens were tested to determine how the angle size, gage length, and angle-to-column
connection washer plate (see Fig. 2.11) affect the connection stiffness, strength, energy
dissipation, and resistance to low cycle fatigue. All seven specimens developed a yield
mechanism with three plastic hinges two hinges in the angle leg attached to the column
(similar to the hinges in Fig. 2.18b) and one hinge along the toe of the fillet in the angle
leg attached to the beam. An analytical model was developed based on the experimental
results to predict the angle initial stiffness, force at yield mechanism, post-yield stiffness,
and unloading behavior. The main conclusions from this research include:
(1) Following the formation of the plastic hinge mechanism, the angle stiffness
greatly decreases but does not become zero. This post-yield stiffness is nearly linear and
is comprised of both geometric hardening and material hardening. The geometric
hardening accounts for slightly less than half of the total post-yield stiffness.
45
(2) Angles with smaller ratios of gage length to angle thickness (lg2/ta) dissipate
more energy for a given value of angle displacement and are stronger and stiffer than
angles with larger lg2/ta ratios. However, angles with smaller lg2/ta ratios have less
resistance to low-cycle fatigue. Thus, the angle gage length and thickness must be
selected considering the stiffness, strength, energy dissipation, and ductility requirements
of the connection.
(3) The washer plates that were used in the angle-to-column connections have no
significant influence on the bolt prying forces and are not effective in influencing the
location of the plastic hinge for angles with small lg2/ta ratios. Furthermore, angles
without washer plates are capable of undergoing several more cycles until failure,
implying that the small decrease in gage length caused by the washer plates reduces the
angles resistance to low-cycle fatigue. Thus, washer plates are not recommended for use
in bolted top-and-seat angle connections.
2.7
Chapter Summary
A summary of the important findings from the literature review presented in this
chapter is given below.
(1) Coupled wall structural systems are effective primary lateral load resisting
systems for seismic regions. The coupling degree is an important parameter for the
seismic behavior and design of these structures.
(2) Concrete coupling beams require the use of diagonal reinforcement to prevent
premature failure modes due to diagonal tension and sliding shear.
(3) Embedded steel coupling beams possess large ductility capacity and energy
dissipation; however, they are prone to significant damage during a large earthquake,
possibly resulting in considerable residual deformations after unloading.
(4) Unbonded post-tensioning has been successfully applied to different types of
building systems in seismic regions. In an unbonded post-tensioned structure, the
nonlinear behavior occurs primarily as a result of the opening of gaps along the joints
between the structural members, and not as a result of material nonlinearity. Thus, these
structures can undergo large lateral displacements with little damage and possess large
self-centering capability due to the restoring effect of the post-tensioning force. The
greatest setback to the use of unbonded post-tensioned structures in seismic regions is
that their lateral displacement demands during a severe earthquake may be larger than
acceptable as a result of small energy dissipation.
(5) Fiber beam-column elements has been successfully used to model the nonlinear
hysteretic behavior of different types of unbonded post-tensioned structures, including
the opening of gaps at the joints between the structural members.
46
(6) Steel top and seat angles are effective structural components that can be used at
the gap opening connections to provide energy dissipation and structural redundancy
during an earthquake. In a properly-designed post-tensioned top and seat angle
connection, the only components to receive significant damage are the angles, which can
be replaced after the earthquake.
47
CHAPTER 3
ANALYTICAL MODELING OF
COUPLED WALL SUBASSEMBLAGES
This chapter describes the analytical modeling of hybrid coupled wall
subassemblages as follows: (1) prototype subassemblage; (2) analytical modeling
assumptions; (3) fiber element subassemblage model; (4) verification of fiber element
subassemblage model; (5) advantages and limitations of fiber element model; and (6)
modeling of embedded steel coupling beams. The subassemblage models described in
this chapter are later used to construct analytical models of multi-story coupled wall
structures in this report.
3.1
Prototype Subassemblage
A prototype floor-level unbonded post-tensioned hybrid coupled wall

subassemblage (similar to Fig. 1.3) was designed as part of an eight-story structure for a
site with a stiff soil profile (Site Class D in IBC-2000) in a region with high seismicity
(coastal California). The structural design properties of the subassemblage are
summarized in Tables 3.1-3.4 and described below.
The elevation view of the prototype subassemblage is shown in Fig. 3.1a, with the
top and seat angle connection details provided in Fig. 3.1b. The design of similar
subassemblages is described in detail in Chapter 12. The story height hs = 3.96 m (13 ft),
wall length lw = 3.05 m (10 ft), wall thickness tw = 356 mm (14 in.) uniform, and
coupling beam length lb = 3.05 m (10 ft).
TABLE 3.1
PROTOTYPE SUBASSEMBLAGE COUPLING BEAM PROPERTIES
Coupling Beam Properties
Section
W21182
Beam length, lb
3.05 m (10 ft)
Beam depth, db
577 mm (22.72 in.)
Flange thickness, tbf
37.6 mm (1.48 in.)
Flange width, bbf
318 mm (12.5 in.)
Web thickness, tbw
21 mm (0.83 in.)
Beam cross section area, Ab
34581 mm2 (53.6 in2)
Beam moment of inertia, Ib
19.7 m4 (4730 in4)
Flange cover plate thickness, tc
38.1 mm (1.5 in.)
Flange cover plate width, bc
318 mm (12.5 in.)
Flange cover plate length, lc
406 mm (16 in.)
48
TABLE 3.2
PROTOTYPE SUBASSEMBLAGE BEAM POST-TENSIONING PROPERTIES
Beam Post-Tensioning Properties
Beam post-tensioning strand diameter, dbp
15.2 mm (0.6 in.)
Number of strands per beam post-tensioning tendon, nbp
3
Beam post-tensioning tendon area, abt
420 mm2 (0.651 in2)
Number of beam post-tensioning tendons, nbt
6
Total beam post-tensioning steel area, Abp
2520 mm2 (3.906 in2)
Unbonded length of beam post-tensioning tendons, lbpu
9.14 m (30 ft)
Initial stress in beam post-tensioning steel, fbpi
0.6fbpu
TABLE 3.3
PROTOTYPE SUBASSEMBLAGE BEAM-TO-WALL CONNECTION
TOP AND SEAT ANGLE PROPERTIES
Beam-to-Wall Connection Top and Seat Angle Properties
Section
L881-1/8
Angle length, la
318 mm (12.5 in.)
Angle leg length, lal
203 mm (8 in.)
Angle leg thickness, ta
28.6 mm (1.125 in.)
Angle fillet length, ka
44.5 mm (1.75 in.)
Number of angle-to-wall connection bolts, nabw
8
Number of angle-to-beam connection bolts, nabb
8
Angle-to-wall connection bolt diameter, dabw
22.2 mm (7/8 in.)
Angle-to-beam connection bolt diameter, dabb
25.4 mm (1 in.)
Angle-to-wall connection bolt head width across flats, wabw
36.5 mm (1.4375 in.)
Angle-to-wall connection gage length, lgv
106 mm (4.1875 in.)
Angle-to-beam connection gage length, lgh
127 mm (5 in.)
TABLE 3.4
PROTOTYPE SUBASSEMBLAGE WALL PROPERTIES
Wall Properties
Wall length, lw
Wall thickness, tw
Wall storey height, hs
Wall-contact region embedded plate thickness, te
Wall-contact region spiral diameter, Dbsp
Wall-contact region spiral wire diameter, dbsp
Wall-contact region spiral pitch, sbsp
Wall-contact region spiral reinforcement ratio, bsp
3.05 m (10 ft)

356 mm (14 in.)
3.96 m (13 ft)
31.8 mm (1.25 in.)
292 mm (11.5 in.)
12.8 mm (0.504 in.)
31.8 mm (1.25 in.)
5.57%
The coupling beam has a W21182 cross section and the top and seat angles have a
L881-1/8 cross section. Note that the steel section designations in this report follow
standard U.S. nomenclature. Also note that this research focuses on beams with I sections;
however, the findings can be extended to other types of steel beams, such as with hollow
(e.g., box-shaped) sections. One of the advantages of unbonded post-tensioned coupling
beams over embedded steel beams is that the selection of the beam cross section is not
affected by the reinforcing details inside the walls. The beam flange width is limited to
the wall thickness.
The length of the top and seat angles la is equal to the beam flange width, bbf = 318
mm (12.5 in.). Flange cover plates with thickness tc = 38.1 mm (1.5 in.) and length lc =
49
406 mm (16 in.) are used at the beam ends. The width of the cover plates is assumed to
be equal to the beam flange width; however, note that the leg thicknesses of the fillet
welds between the cover plate and the flange must be considered in practice. The
thickness of the embedded plates inside the walls, te = 31.8 mm (1.25 in.), and the width
of the plates is equal to the wall thickness, tw = 356 mm (14 in.).
coupling beam
W21 182
connection
region
l c =406 mm
PT
anchorage
(16 in.)
PT tendon spiral
V
W21 182
wall
region
db
hs=3.96m
(13 ft)
cover A
B
embedded
plate
plate te=31.75 mm angle
L8x8x1-1/8
l b = 3.05 m (10 ft)
l w = 3.05 m (10 ft)
cover plate
tc =38.1mm
(1.5in.)
l w= 3.05 m (10 ft)
section at A A
28.6
(1.125)
A49022.2(7/8)bolt
38.1
(1.5)
lg1=73.7(2.9)
76.2 38.1
(3) (1.5)
A490 25.4(1)bolt
wabw=36.5
(1.4375)
lgv =106
(4.1875)
ka
k a=44.5(1.75)
203
(8)
ta=28.6
(1.125)
lgh=127(5)
cross section
elevation
section at B B
60.3
76.2 38.1
(2.375) (3) (1.5)
l a=318(12.5)
88.9
(3.5)
dbc
3-strand PT tendon
a bt=420mm2, fbpi =0.6fbpu
(0.651 in 2 )
(a)
38.1 76.2
(1.5) (3)
W21 182
76.2
(3)
36.5
(1.4375)
318
60.3
(12.5)
(2.375)
88.9
(3.5)
77.8
(3.0625)
76.2
(3)
28.6
(1.125)
38.1
(1.5)
plan
unit: mm (inch)
(b)
Fig. 3.1 Prototype coupled wall subassemblage: (a) elevation;
(b) top and seat angle connection details
The prototype subassemblage is post-tensioned using six multi-strand tendons,
three tendons on each side of the beam web as shown in Fig. 3.1a. Each tendon consists
of three 15.2 mm (0.6 in.) nominal diameter high-strength seven-wire strands with a total
area of abt = 420 mm2 (0.651 in2). The tendons are prestressed to fbpi = 0.60fbpu, where fbpu
= 1862 MPa (270 ksi) is the design maximum strength of the post-tensioning strands. The
stress fbpi=0.60fbpu represents the design initial stress assumed to act in the beam posttensioning tendons just before the application of lateral loads. The initial stress at the
ends of the coupling beam (including the flange cover plates) due to post-tensioning is fbi
= 48 MPa (6.96 ksi), which is equal to 0.13fby where fby = 362 MPa (52.5 ksi) is the
assumed yield strength of the beam steel.
50
Two steel spirals are used to confine the concrete in each wall region, one spiral
near each beam flange as shown in Fig. 3.1a. The spiral center-to-center diameter Dbsp =
292 mm (11.5 in.), spiral W20 wire diameter dbsp = 12.8 mm (0.504 in.), and spiral pitch
sbsp = 31.8 mm (1.25 in.), resulting in a spiral reinforcement ratio of bsp = 5.57% (defined
as the ratio of the volume of spiral reinforcement to the volume of spiral confined
concrete core).
The angle-to-wall connections consist of eight 22.2 mm diameter ASTM A490
bolts and the angle-to-beam connections consist of eight 25.4 mm diameter slip-critical
ASTM A490 bolts as shown in Fig. 3.1b.
3.1.1 Subassemblage Material Properties and Idealizations
This section describes assumed material properties for: (1) the steel used in the
coupling beam, angles, cover plates, and wall embedded plates; (2) the beam posttensioning strands; and (3) the unconfined and spiral confined concrete in the wallcontact regions. Multi-linear idealizations to the assumed smooth uniaxial material stressstrain relationships are also provided for modeling purposes.
The subassemblage material design properties are listed in Table 3.5.
TABLE 3.5
PROTOTYPE SUBASSEMBLAGE MATERIAL PROPERTIES
Beam, Angle, Cover
Plate, Wall Embedded
Plate Steel
Yield
362 MPa
strength, fsy (52.5 ksi)
Yield
0.00181
strain, sy
Maximum 470 MPa
strength, fsm (68.2 ksi)
Strain at max.
0.14
strength, sm
Beam Post-Tensioning Wall-Contact Region

Strands
Spiral Wire
Yield
strength, fbpy
Yield
strain, bpy
Maximum
strength, fbpu
Strain at max.
strength, bpu
1689 MPa
Yield
(245 ksi) strength, fspy
Yield
0.00860
strain, spy
1862 MPa Maximum
(270 ksi) strength, fspm
Strain at max.
0.05
strength, spm
Wall-Contact Region Concrete

Unconfined Concrete
414 MPa Linear-elastic 30441 MPa

(60 ksi) stiffness, Ec (4415 ksi)
41.4 MPa
Maximum
0.00207
(6 ksi)
strength, fc
621 MPa Strain at max.
0.002
(90 ksi) strength, c
Ultimate
0.003
0.08
strain, cu
Spiral Confined Concr.

Linear-elastic
stiffness, Ec
Maximum
strength, fcc
Strain at max.
strength, cc
Ultimate
strain, ccu
30441 MPa
(4415 ksi)
91.0 MPa
(13.2 ksi)
0.0140
0.0323
Fig. 3.2a shows the assumed smooth (dashed line) and idealized tri-linear (solid
line) stress-strain relationships of the steel used in the coupling beam, angles, cover plates,
and wall embedded plates. The yield strength and maximum strength of the steel are
equal to fsy = 362 MPa (52.5 ksi) and fsm= 470 MPa (68.2 ksi), respectively. The postyield stiffness of the idealized stress-strain relationship is determined from the portion of
the smooth stress-strain relationship between the yield strain, sy, and a strain of sm= 0.14
at which the maximum strength, fsm, is assumed to be reached. The ratio of the idealized
post-yield stiffness to the initial linear-elastic stiffness Es=199955 MPa (29000 ksi) is
equal to 0.00393.
The assumed smooth (dashed line) and idealized bi-linear (solid line) stress-strain
relationships of the post-tensioning strands are shown in Fig. 3.2b. The yield strength and
maximum strength of the post-tensioning strands are equal to fbpy=1689 MPa (245 ksi)
51
and fbpu=1862 MPa (270 ksi), respectively. The yield strength fbpy is assumed to be equal
to the linear limit stress (i.e., the limit of proportionality) of the smooth stress-strain
relationship. Since the post-tensioning tendons are not bonded to the concrete, the peak
strains in the strands during an earthquake are expected to remain small. Thus, the postyield stiffness of the idealized bi-linear stress-strain relationship in Fig. 3.2b was
determined by assuming that the maximum strength, fbpu=1862 MPa (270 ksi) is reached
at a strain of 0.0116, to represent the portion of the smooth stress-strain relationship
between the yield strain, bpy, and the largest strain expected in the strands as the
subassemblage is displaced. The ratio of the idealized post-yield stiffness to the initial
linear-elastic stiffness Ebp=196508 MPa (28500 ksi) of the post-tensioning strands is
equal to 0.292.
500
2000 (0.0116, 1862)
steel stress (MPa)
beam PT strand stress (MPa)
( sm , fsm )=
(0.14, 470)
400
300
( sy , f sy )= (0.00181, 362)
200
smooth relationship
idealized tri-linear relationship
100
TENSION
0
0
0.12
0.08
0.04
0.16
1200
800
smooth relationship
idealized bi-linear relationship
400
TENSION
0
0.20
0.02
0.01
steel strain
(0.002, 41.4)
(0.00676, 86.2)
concrete stress (MPa)
30
unconfined
concrete
(0.000750, 22.8)
smooth relationship
idealized multi-linear
relationship
COMPRESSION
0
0.001
0.003
0.004
(0.0140, 91.0)
80
60
(0.00170, 51.7)
40
smooth relationship
idealized multi-linear
relationship
20
0
0.005
(0.0323, 84.8)
concrete
crushing
spiral confined concrete
COMPRESSION
(0.004, 0.07)
0.002
0.05
(b)
concrete crushing
(0.003, 35.9)
0.04
100
(0.00150, 38.5)
15
0.03
beam post-tensioning strand strain
(a)
45
( bpu , fbpu)=
(0.05, 1862)
(bpy, fbpy )=
(0.00860, 1689)
1600
0.01
0.02
concrete strain
concrete strain
(c)
(d)
(0.04, 0.07)
0.04
0.03
Fig. 3.2 Subassemblage material properties and idealizations: (a) steel; (b) beam posttensioning strand; (c) wall unconfined concrete; (d) wall-contact region confined concrete
The wall unconfined concrete maximum compressive strength is assumed to be
f =41.4 MPa (6 ksi) reached at a strain of c' =0.002, and the concrete initial linear'
c
52
elastic stiffness is assumed to be Ec=4730 f c' =30442 MPa (or Ec=57000 f c' =4415201
psi). It is assumed that the crushing of the unconfined concrete occurs at a strain of
cu=0.003. Fig. 3.2c shows the smooth (dashed line) and idealized multi-linear (solid line)
compressive stress-strain relationships of the unconfined concrete.
Similarly, Fig. 3.2d shows the smooth (dashed line) and idealized multi-linear
(solid line) compressive stress-strain relationships of the spiral confined concrete in the
wall-contact regions of the prototype subassemblage. The maximum compressive
strength of the spiral confined concrete is fcc=91.0 MPa (13.2 ksi) reached at a strain of
cc=0.014, and the ultimate (i.e., crushing) strain is ccu=0.0323.
The smooth concrete stress-strain relationships in Figs. 3.2c and 3.2d were
determined using a concrete model developed by Mander et al. (1988). The spiral
confined concrete stress-strain parameters, which include the maximum strength, fcc,
strain at maximum strength, cc, and ultimate (i.e., crushing) strain, ccu, were calculated
based on the unconfined concrete properties and the amount, distribution, yield strength
[assumed as fspy=414 MPa (60 ksi)], and strain at maximum strength (assumed as
spm=0.08) of the spiral reinforcement. According to the concrete confinement model by
Mander et al. (1988), the ultimate (crushing) strain ccu of the confined concrete is
reached when the spiral reinforcement fractures, resulting in a complete loss of
confinement and a sudden loss in the compression resistance of the concrete. The
concrete confinement provided by the wall transverse reinforcement other than the spirals
(e.g., wire mesh) was ignored.
In general, the slopes of the first segment of the idealized multi-linear stress-strain
relationships in Fig. 3.2 match the linear-elastic stiffness of the smooth relationships.
Furthermore, the maximum compressive strength, strain at maximum strength, and
ultimate strain of the concrete are matched by the idealized stress-strain relationships in
Figs. 3.2c and 3.2d.
3.2
Analytical Modeling Assumptions

The following assumptions are made for the subassemblage analytical modeling:
(1) The objective of this research is to investigate the behavior of isolated coupled
wall structures under earthquake induced lateral loads. The interaction between the
coupled walls and other structural members (e.g., slabs supported by the coupling beams
and the walls) is not within the scope of the analytical model, and thus, is ignored. Note
that the floor and roof slabs may affect the expected and desired behavior of a coupled
wall structure; however, this is not investigated in the report.
(2) The coupled wall system undergoes in-plane deformations only. Torsional and
out-of-plane deformations are not modeled.
(3) Local and/or global instability of the coupling beams is prevented by proper
design and detailing.
53
(4) Shear slip of the coupling beams at the beam-to-wall interfaces is prevented by
proper design and detailing.
(5) The welds between the cover plates (if used) and the coupling beam flanges are
properly designed and detailed for the maximum forces and deformations. The cover
plates are sufficiently long such that yielding of the coupling beam where the plates are
terminated does not occur.
(6) The anchorages for the coupling beam post-tensioning tendons are properly
designed and detailed for the maximum post-tensioning forces.
(7) The top and seat angles are loaded in the horizontal direction parallel to the
coupling beam and the angles form a ductile vertical-leg failure mechanism as shown in
Fig. 2.18.
(8) The angle-to-beam and angle-to-wall connections are properly designed and
detailed for the maximum angle forces and deformations.
3.3
Fiber Element Subassemblage Model
This section describes a fiber element based analytical model for unbonded posttensioned hybrid coupled wall subassemblages as follows: (1) modeling of wall regions;
(2) modeling of coupling beams and flange cover plates; (3) modeling of gap opening; (4)
modeling of beam post-tensioning tendons and anchorages; and (5) modeling of top and
seat angles. The DRAIN-2DX program (Prakash et al. 1993) is used as the analytical
platform. As an example, the model for the prototype subassemblage in Fig. 3.1 is shown
in Fig. 3.3. The verification of the subassemblage model is described later.
3.3.1
Modeling of Wall Regions
The wall regions are modeled using the fiber beam-column element in DRAIN2DX. As shown in Fig. 3.4, each fiber element is divided into a number of segments. The
cross-section properties remain constant within a fiber segment, but can vary from one
segment to another. Within a segment, parallel fibers in the direction of the element
model the cross-section (or slice) at the mid-length of the segment. Each fiber is
characterized by a uniaxial multi-linear material stress-strain relationship (similar to the
relationships in Fig. 3.2), a cross section area, and a distance from the longitudinal
reference axis of the element. The force versus deformation behavior of the slice is
determined by the integration of the stress versus strain behavior of the fibers over the
cross section. The theoretical formulations for the development of the fiber element in
DRAIN-2DX are described by Prakash et al. (1993), and thus, are not discussed here in
further detail.
Each wall region in an unbonded post-tensioned hybrid coupled wall
subassemblage is modeled using two sets of fiber beam-column elements as shown in Fig.
3.3a. The first set consists of elements that are in the vertical direction to model the axial54
flexural and shear behavior of the wall region along its height. These elements, referred
to as the wall-height elements, are used to model the cross section of each wall in the
horizontal X-Z plane. More detailed information on the wall-height elements is provided
in Chapter 6.
LEFT WALL REGION
20
wall21
height
elements
RIGHT WALL REGION

angle element
35 beam elements
34
28
29
30
36
kinematic
constraint 22
23
beam PT
element
37
lc
38
40
39
wallcontact
elements
Y
31
32
33
X
Z
41
kinematic
constraint
lw
lb
lw
midstory node
(a)
20 slope=1:3
24
21
29
Node 1
3
2
22
25
beam
7
6
11
9
8
10
cover plate
13
12
15 17
18
14 16
19
32
26
te
27
23
0.5lw1.5dbc d bc 0.5d bc
(b)
Fig. 3.3 Analytical model: (a) subassemblage;

(b) wall-contact elements and beam elements
The second set of fiber elements is used to model the local behavior of the wallcontact regions to the left and right of the coupling beam. These elements, referred to as
the wall-contact elements, are in the horizontal direction. As indicated in Fig. 1.3, large
concentrated compression stresses develop in the contact regions between the beam and
the walls upon gap opening. The wall-contact elements model the deformation of the
concrete in the wall-contact regions under these large stresses. The fiber cross section
properties of the wall-contact elements are determined from the properties of an
effective wall cross section in the vertical Y-Z plane as described below.
The thickness of the effective wall cross section is assumed to be equal to the wall
thickness, tw including the unconfined cover concrete. The depth of the effective wall
cross section is equal to the beam depth, db plus two times the thickness of the cover plate,
tc (denoted as dbc = db + 2tc) at the beam-to-wall interface and is assumed to increase away
55
from the interface (with a slope of 1:3) as shown in Fig. 3.3b. The compression stresses
in the wall-contact region decrease away from the interface as a result of an increase in
the depth of the compression region inside the wall. The increase in the depth of the
effective wall cross section represents this increase in the depth of the compression
region away from the interface.
fiber element
Node J
Node I
segment
slice
fiber
Fig. 3.4 Fiber element, segments, and fibers

Four wall-contact elements are used between the center of the left wall region
(Node 1) and the beam-to-wall interface (Node 5). The Y-translational degree-of-freedom
(DOF) of Node 5 is kinematically constrained to that of Node 1. The rotational and Xtranslational DOFs of Node 5 are not constrained. It is assumed that no slip occurs at the
beam-to-wall interface (Assumption 4). The total length of the first two wall-contact
elements adjacent to Node 5 is equal to 0.5dbc. The first element represents the embedded
steel plate and has a length equal to the embedded plate thickness, te. The length of the
third element is equal to dbc. The total length of the wall-contact elements between Nodes
1 and 5 is equal to one half of the wall length, lw. Note that if 1.5dbc0.5lw, only three
wall-contact elements should be used and the length of the third element should be
adjusted accordingly. The modeling of the right wall region between Nodes 19 and 15 is
similar to the modeling of the left wall region between Nodes 1 and 5.
Fig. 3.5 shows a schematic of the fiber discretizations used in the four wall-contact
elements of the prototype subassemblage model in Fig. 3.3b. The first fiber element
representing the embedded plate (between Nodes 5 and 4) includes steel fibers only (with
no concrete fibers) with the trilinear steel stress-strain relationship in Fig. 3.2a. The other
wall-contact elements include concrete fibers only (with no steel fibers) with the multilinear concrete stress-strain relationships in Figs. 3.2c and 3.2d. The steel and concrete
56
fibers used in the wall-contact elements model the transfer of compression stresses
between the coupling beam and the walls, and, are assumed to have no resistance in
tension to represent the opening of gaps at the beam-to-wall interfaces. In Fig. 3.5,
CUC refers to compression-only unconfined concrete fibers, CCC refers to
compression-only confined concrete fibers, and SC refers to compression-only
steel fibers. More information on the modeling of gap opening at the beam-to-wall
interfaces, as well as on the assumed cyclic behaviors of the compression-only steel and
compression-only concrete fibers is provided later.
slice 1
slice 2
slice 4
slice 3
wall mid-length
node
Node 1
te
0.5l w1.5dbc
d bc
0.5d bc
CUC: compression only unconfined concrete fibers

CCC: compression only confined concrete fibers
SC: compression only steel fibers
slice 1
CUC
slice 2
CUC
CUC
CUC
CCC
CCC
slice 3
CUC
slice 4
CCC
lw/12+5dbc/4+te/6
CUC
CUC
4dbc/3+te/6
SC
CUC
13dbc/12+te/3
dbc+te/3
CUC
CCC
CCC
CCC
tw
tw
CUC
tw
CUC
tw
Note: actual number of fibers in each slice not shown
Fig. 3.5 Schematic showing fiber discretization of wall-contact elements

The actual number of fibers and the thickness of each fiber, tf used in the wallcontact elements modeling the prototype subassemblage are given in Table 3.6. Note that
refinement studies using models with larger numbers of elements/segments/fibers than
the discretization in Table 3.6 showed no appreciable differences in the analysis results.
57
TABLE 3.6
WALL-CONTACT ELEMENT FIBER DISCRETIZATION
Element Element
Nodes
Length
Number of Segment Length/

Segments Element Length
and Slices and Slice Number
1-2
544 mm
(21.42 in.)
1 (Slice 1)
(see Fig. 3.5)
2-3
653.3 mm
(25.72 in.)
1 (Slice 2)
3-4
294.9 mm
(11.61 in.)
0.6 (Slice 3)
0.4 (Slice 3)
4-5
31.75 mm
(1.25 in.)
1 (Slice 4)
Fiber Thickness, tf
mm (in.)
1 CUC fiber; tf=7.37(0.29)
22 CUC fibers;tf=12.7(0.5)
1 CUC fiber; tf=22.4(0.88)
1 CUC fiber; tf=3.05(0.12)
1 CCC fiber; tf=3.05(0.12)
2 CUC fibers; tf=12.7(0.5)
2 CCC fibers; tf=12.7(0.5)
6 CUC fibers; tf=25.4(1)
6 CCC fibers; tf=25.4(1)
1 CUC fiber; tf=9.65(0.38)
1 CCC fiber; tf=9.65(0.38)
1 CUC fiber; tf=15.7(0.62)
1 CUC fiber; tf=25.4(1)
1 CUC fiber; tf=10.9(0.43)
1 CUC fiber; tf=22.35(0.88)
1 CUC fiber; tf=3.05(0.12)
1 CCC fiber; tf=3.05(0.12)
1 CUC fiber; tf=9.65(0.38)
1 CCC fiber; tf=9.65(0.38)
1 CUC fiber; tf=15.7(0.62)
1 CUC fiber; tf=25.4(1)
1 CUC fiber; tf=5.33(0.21)
1 CCC fiber; tf=5.33(0.21)
1 CUC fiber; tf=9.65(0.38)
1 CCC fiber; tf=9.65(0.38)
1 CUC fiber; tf=15.7(0.62)
1 SC fiber; tf=1.78(0.07)
8 SC fibers; tf=6.35(0.25)
6 SC fibers; tf=12.7(0.5)
8 SC fibers; tf=25.4(1)
Depth of
Total Number of
Cross Section Fibers in Slice
1488 mm
(58.58 in.)
104
1089 mm
(42.86 in.)
72
773 mm
(30.42 in.)
76
664 mm
(26.14 in.)
46
Notes:
(1) Only fiber discretization for top half of cross section is given for each slice, listed in order from extreme
fiber to innermost fiber. Fiber discretization for bottom half of each cross section is symmetric about middepth of section.
(2) Modeling of wall-contact elements between Nodes 19-15 is similar to modeling between Nodes 1-5.
58
3.3.2
Modeling of Coupling Beams and Flange Cover Plates
Fiber beam-column elements are used to model the axial-flexural and shear
behavior of the coupling beams. The fiber cross section properties of these elements,
which are referred to as the beam elements, are determined from the properties of the
coupling beam cross section. The beam elements near the beam ends also include fibers
to model the flange cover plates, assuming that the welds between the cover plates and
the beam flanges are properly designed and detailed for the maximum forces and
deformations (Assumption 5). Second order effects (often referred to as P- effects) in
the coupling beam (due to the rotation of the beam with respect to the left and right wall
regions) are included in the fiber beam elements.
Typically, a larger number of fiber elements/segments/fibers are used near the ends
of a coupling beam where the nonlinear behavior is expected to concentrate as compared
with the midspan regions. The exact discretization of the fiber elements along the length
of the beam is flexible with the following exceptions: (1) there are two beam element
nodes (Nodes 7 and 13 in Fig. 3.3b) located at the same X-coordinates as the angle
element Nodes 35, 37 and 38, 40, respectively (the angle element nodes are kinematically
constrained to these beam element nodes as described later); and (2) new fiber elements
or segments are used where the beam cover plates are terminated (in order to define the
change in beam geometry).
Furthermore, the length of the first beam fiber element segment adjacent to the
beam-to-wall interface is important in modeling the compression deformations that occur
at the beam ends. The length where significant yielding of the steel in compression
occurs in an unbonded post-tensioned coupling beam is quite different than an embedded
steel coupling beam due to the gap opening behavior and the compression stresses from
post-tensioning. El-Sheikh et al. (1997) provide recommendations for the confined
concrete crushing length at the ends of unbonded post-tensioned precast concrete beams.
Similar to their recommendation, a length of lb,cr=2cb,sof (Fig. 3.6) is used for the length of
the first fiber element segment in this research, where cb,sof is the estimated depth of the
compression (i.e., contact) region at the beam softening state (see Chapters 4 and 5).
The middle portion of the coupling beam between the cover plates can be modeled
using a relatively smaller number of fiber elements (e.g., four elements as shown in Fig.
3.3). It is recommended that a node is placed at the beam mid-span (Node 10 in Fig. 3.3b).
Fig. 3.7 shows a schematic of typical fiber discretizations used at the beam ends
and near the beam midspan of the prototype subassemblage model in Fig. 3.3. The beam
elements include steel fibers only with the tri-linear steel stress-stain relationship shown
in Fig. 3.2a. In order to model the effect of gap opening, the steel fibers used in the first
beam element segment adjacent to each beam-to-wall interface are assumed to have no
resistance in tension. The other beam fiber element segments use full tensioncompression steel fibers. In Fig. 3.7, SC refers to compression-only steel fibers and
59
S refers to compression-tension steel fibers. More information on the modeling of

gap opening at the beam-to-wall interfaces, as well as on the assumed cyclic behaviors of
the compression-only and tension-compression steel fibers used in the beam elements is
provided later.
beam-to-wall
interface
c b,sof
l b,cr=2c b,sof
45
45
tc
cover plate
t bf
beam flange
elevation
beam web
Fig. 3.6 Length of first beam fiber element segment

The actual number of fibers and the thickness of each fiber, tf used in the beam
elements modeling the prototype subassemblage are given in Table 3.7. Note that
refinement studies using models with larger numbers of elements/segments/fibers than
the discretization in Table 3.7 showed no appreciable differences in the analysis results.
3.3.3
Modeling of Gap Opening
Gap opening at the beam-to-wall interfaces is one of the most important

characteristics governing the behavior of unbonded post-tensioned coupling beams. As a
result of the opening of gaps and due to post-tensioning, large compression stresses
develop in the beam-to-wall contact regions, while the tensile stresses in the coupling
beam and the wall region on either side of a gap are equal or close to zero. The
compression behavior in the contact regions is modeled using the nonlinear uniaxial
compressive stress-strain relationships of the fibers. To model the gap opening behavior,
the tensile strength and stiffness of the fibers adjacent to the beam-to-wall interfaces are
set to zero.
The steel fibers of the first beam element and first wall-contact element (with
length te to model the embedded plate) adjacent to the beam-to-wall interfaces are
assigned the compression-only steel stress-strain relationship in Fig. 3.8a (i.e., the SC
fiber type in Figs. 3.5 and 3.7). Similarly, the concrete wall-contact elements between the
embedded plate and the center of each wall region use the compression-only concrete
60
stress-strain relationship in Fig. 3.8b (i.e., the CCC and CUC fiber types for confined
concrete and unconfined concrete, respectively, in Fig. 3.5). Only the confined concrete
(CCC) fiber behavior is shown in Fig. 3.8b the cyclic behavior of the unconfined
concrete (CUC) fiber has similar characteristics. More details on the cyclic stress-strain
relationship of the concrete fiber in DRAIN-2DX can be found in Kurama et al. (1996)
and Kurama (1997). The beam elements away from the beam-to-wall interfaces have the
full compression-tension steel stress-strain relationship in Fig. 3.8c (i.e., the S fiber type
in Fig. 3.7). Note that Bauschinger effect is not modeled in the compression-tension steel
fibers since the beam elements away from the interfaces remain essentially linear-elastic.
l gh l b,cr
l b,cr
l c l gh
Node 5
slice 1
slice 2
slice 1
l b/2 lc
9
10
slice 3
beam midspan node

11
SC
13
14
15
SC: compression-only steel fibers

S: compression-tension steel fibers
slice 2
tc
dbc
12
slice 3
(no cover plate)
tf
db
bf
Note: actual number of fibers in each slice not shown
Fig. 3.7 Fiber discretization of beam elements

Through this model, the gap opening displacements that occur at the beam-to-wall
interfaces are represented as distributed tensile deformations that occur in the adjacent
beam and wall-contact elements. The reduction in the lateral stiffness of the
subassemblage as a result of gap opening is modeled by the zero stiffness of the fibers
that go into tension near the interfaces when the precompression stresses due to the posttensioning force are overcome by the flexural stresses that develop due to the lateral loads
(i.e., when decompression occurs).
The process of gap opening/closing under the action of lateral loading/unloading
causes softening/re-stiffening at the beam-to-wall interface regions. This process is
captured in the fiber element model by having an increasing number of fibers subjected to
tension during loading, and then by having the fibers subjected to tension go back into
compression during unloading.
61
TABLE 3.7
BEAM FIBER ELEMENT DISCRETIZATION
Element
Nodes
Element
Length

5-6
76.2 mm (3 in.)
1 (Slice 1)
(see Fig. 3.7)
6-7
50.8 mm (2 in.)
1 (Slice 2)
7-8
279.4 mm
(11 in.)
0.5 (Slice 2)
0.5 (Slice 2)
8-9
508 mm (20 in.)
0.5 (Slice 3)
0.5 (Slice 3)
9-10
609.6 m (24 in.)
0.417 (Slice 3)
0.583 (Slice 3)
Fiber Thickness, tf
mm (in.)
15 SC fibers; tf=2.54(0.1)
1 SC fiber; tf=2.03 (0.08)
14 SC fibers; tf=2.54(0.1)
1 SC fiber; tf=7.11(0.28)
12 SC fibers; tf=20.32(0.8)
15 S fibers; tf=2.54(0.1)
1 S fiber; tf=2.03 (0.08)
14 S fibers; tf=2.54(0.1)
1 S fiber; tf=7.11(0.28)
12 S fibers; tf=20.32(0.8)
15 S fibers; tf=2.54(0.1)
1 S fiber; tf=2.03 (0.08)
14 S fibers; tf=2.54(0.1)
1 S fiber; tf=7.11(0.28)
12 S fibers; tf=20.32(0.8)
1 S fiber; tf=2.03(0.08)
14 S fibers; tf=2.54(0.1)
1 S fiber; tf=7.11(0.28)
12 S fibers; tf=20.32(0.8)
1 S fiber; tf=2.03(0.08)
14 S fibers; tf=2.54(0.1)
1 S fiber; tf=7.11(0.28)
12 S fibers; tf=20.32(0.8)
Depth of Total Number of

653.3 mm
(25.72 in.)
86
653.3 mm
(25.72 in.)
86
653.3 mm
(25.72 in.)
86
577 mm
(22.72 in.)
56
577 mm
(22.72 in.)
56
Notes:
(1) Only fiber discretization for top half of cross section is given for each slice, listed in order from extreme
fiber to innermost fiber. Fiber discretization for bottom half of each cross section is symmetric about middepth of section.
(2) Modeling of beam elements between Nodes 15-10 is similar to modeling between Nodes 5-10.
3.3.4
Modeling of Beam Post-Tensioning Tendons and Anchorages
Each beam unbonded post-tensioning (PT) tendon is modeled using a truss element,
referred to as the beam PT element. Post-tensioning tendons located at the same
elevation within the beam cross section are modeled using the same truss element. The
cyclic stress-strain behavior of the truss elements, based on the bi-linear relationship in
Fig. 3.2b, is shown in Fig. 3.8d. The post-tensioning loads are simulated by tensile forces
in the truss elements, which are equilibrated by compressive forces in the beam and wallcontact elements. Note that the compression forces that develop in the beam and wallcontact elements result in an elastic shortening and subsequent loss in the forces in the
truss elements modeling the PT tendons. Thus, larger tensile forces are applied to the
truss elements in order to achieve the desired amount of initial force (i.e., force just
before the application of lateral loads) in the post-tensioning tendons after elastic
shortening takes place.
Each beam PT element is connected to two nodes representing the anchorages
between the post-tensioning tendons and the walls at the outer ends (e.g., Nodes 28-31,
62
29-32, and 30-33 for the three pairs of post-tensioning tendons in Fig. 3.3). Based on
Assumption 6, the displacements of each anchorage node are kinematically constrained to
the node at the center of the corresponding wall region (Nodes 1 and 19) as shown in Fig.
3.3. Second order effects (often referred to as P- effects) in the post-tensioning
tendons (due to the relative displacements of the tendon nodes in a direction
perpendicular to the tendon) are modeled in the beam PT elements.
500
100
TENSION
stress (MPa)
stress (MPa)
50
TENSION
(-0.04, -0.07)
-50
( sy , fsy ) (-0.00181, -362)
(-0.14, -470)
-500
-0.2
-0.1
0
strain
0.1
(-0.00170, -51.7)
(-0.0323, -84.8)
-100
-0.05
0.2
(-0.00676, -86.2)
(-0.0140, -91.0)
0
strain
0.05
compression-only steel (SC) fiber type
compression-only confined concrete (CCC) fiber type
(a)
(b)
500
2000
(0.14, 470)
( sy , fsy )
( bpy , fbpy )
1500
(0.00860, 1689)
(0.00181, 362)
(0.0116, 1862)
stress (MPa)
stress (MPa)
1000
TENSION
0
500
TENSION
0
-500
-1000
-1500
(-0.00181, -362)
-500
-0.2
-0.1
0
strain
0.1
0.2
-2000
-0.02
-0.01
0
strain
compression-tension steel (S) fiber type
beam PT element
(c)
(d)
0.01
0.02
Fig. 3.8 Cyclic material models: (a) compression-only steel (SC) fiber; (b) compressiononly confined concrete (CCC) fiber; (c) compression-tension steel (S) fiber;
(d) beam PT element
3.3.5
Modeling of Top and Seat Angles
The behavior of the top and seat angles at the beam-to-wall interfaces is also
modeled using fiber elements, referred to as angle elements. Each angle element is
63
connected to two nodes as shown in Fig. 3.3 (Nodes 34-35, 36-37, 38-39, and 40-41 for
the four angles). Both nodes are located at the same elevation as the middle of the
horizontal leg of the angle parallel to the beam. Based on Assumption 8, the first angle
node (e.g., Node 34), which is located at the wall face, is kinematically constrained to a
wall-height element node at the same elevation (Node 21). The second angle node (Node
35) is located at the centroid of the bolt group connecting the angle to the beam flange
and is kinematically constrained to a corresponding beam node (Node 7).
The angle elements represent the expected behavior of the top and seat angles
under loading parallel to the beam flanges as shown in Fig. 3.9a. As described in Chapter
2, the behavior of steel angles under similar loading conditions has been investigated
experimentally and analytically by other researchers (e.g., Kishi and Chen 1990; Lorenz
et al. 1993; Swanson and Leon 1999; Shen and Astaneh-Asl 1999, 2000; Sims 2000;
Garlock et al. 2003). The angle load versus deformation relationship is governed by many
factors including the size and gage length of the connection bolts, number and layout of
bolts, and angle leg thickness.
The cyclic load-deformation relationship used to model the behavior of each angle
in a coupled wall subassemblage is shown in Fig. 3.9b. This cyclic behavior is obtained
by using two fibers in parallel fiber 1 is a tension-only concrete fiber and fiber 2 is a
tension-compression steel fiber. The combination of the load-deformation relationships of
these two fibers in parallel models the expected behavior of an angle under cyclic loading.
Under tensile loading, the yield strength Tayx and initial stiffness, Kaixt for an angle are
determined as described by Kishi and Chen (1990) and Lorenz et al. (1993) and as
summarized in Chapter 2. The vertical leg of the angle is assumed to be fixed along the
innermost edge of the line of bolt heads and is pulled by the coupling beam flange (Fig.
3.9a). The yield strength of the angle is reached when the vertical-leg plastic hinge
mechanism shown in Fig. 3.9a develops, considering the interaction between the bending
moment and shear in the vertical leg. The yield deformation of the angle is assumed to
be ayx=Tayx/Kaixt. The post-yield stiffness in tension is assumed to be equal to aKaixt with
a=0.02 based on previous experimental results reported by Sims (2000).
Under compression loading, the yield strength of an angle is assumed to be equal
to the slip force Casx for the bolts connecting the horizontal leg of the angle to the beam
flange. The subassemblage analyses described in Chapter 4 show that extremely small
deformations occur in the angle once the beam flange comes in contact with the wall.
Thus, the development of the shear/bearing strength of the angle-to-beam connection
bolts or the development of the compression bearing capacity of the angle horizontal leg
cross section is not expected and is not modeled.
Based on the angle model described above, Fig. 3.9c shows the load-deformation
relationship developed for the top and seat angles in the prototype subassemblage. Some
of the angle design capacities are given in Table 3.8. The design slip-critical strength,
shear strength, and tension strength of the angle-to-beam and angle-to-wall connection
64
bolts were determined based on the recommendations of the American Institute of Steel
Construction (AISC 1998).
distance from heel
gage length, l gv
ayx to toe of fillet, k a

ta
T ayx
T ayy
Ma
plastic hinges
seat angle at
tension yielding
bolt head
(a)
fiber 1 force
TENSION
TENSION
a K aixt
(TayxC asx , ayx ) a K aixt
K aixt
=
deformation
C asx
C asx
angle behavior
angle force (kN)
fiber 1
16.5 kN/mm
1200
0
-543
14
angle deformation (mm)
fiber 2
(b)
a=0.02
fiber 1 force (kN)
(952,1.15)
TENSION
deformation
deformation
0
1200
(Casx , ayx )
-1200
-2
fiber 2 force
1200
(409,1.15)
16.5 kN/mm
-1200
-2
fiber 2 force (kN)
angle force
(Tayx , ayx )
(543,1.15)
0
-543
14
-1200
-2
14
(c)
Fig. 3.9 Angle model: (a) idealized displaced shape; (b) cyclic angle load-deformation
model; (c) angle load-deformation behavior for prototype subassemblage
TABLE 3.8
PROTOTYPE SUBASSEMBLAGE BEAM-TO-WALL CONNECTION
TOP AND SEAT ANGLE DESIGN CAPACITIES
Beam-to-Wall Connection Top and Seat Angle Design Capacities
Angle initial stiffness in tension, Kaixt
824 kN/mm (4707 kips/in.)
Angle yield strength in tension, Tayx
952 kN (214 kips)
Total design slip-critical strength of angle-to-beam connection
68 kN x 8 bolts = 543 kN (122 kips)
bolts at factored loads, Casx
Total design shear strength of angle-to-beam connection bolts
197 kN x 8 bolts = 1573 kN (354 kips)
Total design tension strength of angle-to-wall connection bolts
227 kN x 8 bolts = 1814 kN (408 kips)
65
The shape of the idealized angle cyclic behavior in Fig. 3.9c resembles the
experimentally observed behavior in Fig. 2.19 by Sims (2000). Note that this angle model
represents a simplified idealization of the actual behavior of the angles in a coupled wall
subassemblage. The ultimate failure of the angles is not modeled. The effect of bolt
response (e.g., yielding of the bolts) on the behavior of the angles is ignored (Assumption
8) and the angles are assumed to form a ductile plastic mechanism with two hinges in the
vertical leg as shown in Fig. 3.9a. The rotation of the horizontal angle leg with respect to
the vertical leg, which occurs as a result of the rotation of the beam with respect to the
walls, is not considered. Furthermore, only the axial force in the angle horizontal leg
parallel to the beam is represented by the model, ignoring the other angle forces
transferred to the beam (e.g., the shear force in the horizontal leg). Similar assumptions
have been made by other researchers (e.g., Ricles et al. 2001).
3.4
Verification of Fiber Element Subassemblage Model
The coupled wall subassemblage model described above (referred to as the fiber
element model) is verified using a finite element model developed with the ABAQUS
program (Hibbitt et al. 1998).
3.4.1
Finite Element Subassemblage Model
Two finite element models of the prototype coupled wall subassemblage described
previously were developed: (1) a two-dimensional (2D) model using nonlinear
rectangular plane stress elements to represent the wall regions and the coupling beam;
and (2) a three-dimensional (3D) model using nonlinear brick elements to represent the
wall regions and the coupling beam. In both models, truss elements were used to
represent the post-tensioning tendons and gap/contact surfaces were used to represent the
opening of gaps at the beam-to-wall interfaces. Adequate friction was provided at the
gap/contact surfaces to prevent sliding of the beam with respect to the walls.
Fig. 3.10 shows the 2D and 3D finite element models of the prototype
subassemblage in exaggerated deformed configurations. Discrete gaps form between the
beam and the walls as a result of the gap/contact surfaces. The left wall region is referred
to as the reaction block and is fixed along the top and bottom horizontal surfaces. The
right wall region, referred to as the loading block, is allowed to translate in the horizontal
and vertical directions, but not allowed to rotate assuming that the left and right wall
regions do not rotate relative to each other. A force V is applied on the loading block in
the vertical direction. These boundary conditions result in deformations in the
subassemblage that are similar to the deformations with respect to the reference line in
Fig. 1.3b.
Note that the flange cover plates and the top and seat angles are not included in the
finite element models. Thus, the verification of the fiber element model is done based on
the prototype subassemblage with the flange cover plates and the angles removed. Other
66
researchers have shown the difficulties in the finite element modeling of the nonlinear
behavior of top and seat angle connections, in particular the boundary conditions for the
angle legs, prying, friction, slip, and interaction between the angles, bolts, and nuts (Sims
2000). Thus, the modeling of the angles using results from previous research as described
in Section 3.3.5 is considered to be satisfactory. An improved angle model based on
large-scale experiments of unbonded post-tensioned hybrid coupled wall subassemblages
is described later in Chapter 10.
gap/contact surfaces
V
rigid elements
for PT
anchorages
truss element
undeformed shape
plane stress elements
reaction block
loading block
(a)
3-D brick elements
gap/contact
surfaces
rigid elements for

PT anchorages
truss element for PT-tendon
(b)
wall contact region
(c)
Fig. 3.10 Finite element models: (a) two dimensional (2D) model; (b) three-dimensional
(3D) model; (c) close-up view of beam-to-wall interface
Fig. 3.11 compares the load-deformation relationships of the prototype
subassemblage using the 2D and 3D finite element models. Note that this comparison is
67
based on an approximately half-scale model of the prototype subassemblage in Fig. 3.1.

The load is equal to the coupling shear force in the beam, Vb=V, and the deformation is
equal to the beam chord rotation, b, which is calculated as the vertical displacement of
the right end of the beam divided by the beam length. The results obtained using the 2D
model (with plane stress elements) and the 3D model (with brick elements) are very
similar. Thus, the 2D model is adopted for further finite element analyses of the
subassemblage due to its relative simplicity.
250
beam shear force, Vb(kN)
200
150
approximately half-scale model of
prototype subassemblage
100
2D finite element model (ABAQUS)
50
3D finite element model (ABAQUS)
2.0
1.0
3.0
beam chord rotation, b (percent)
Fig. 3.11 Comparison between 2D and 3D finite element models

3.4.2
Finite Element Analysis Results
Two 2D finite element models of the full-scale prototype subassemblage were

developed: (1) a rigid wall model in which the deformations in the reaction and loading
blocks are prevented; and (2) a deformable wall model in which the deformations in the
wall regions are included. Fig. 3.12 compares the load-deformation relationships of the
subassemblage using the rigid wall and deformable wall finite element models. The
deformations in the wall regions have a considerable effect on the initial stiffness of the
subassemblage. Most of these deformations occur in the wall-contact regions near the
beam-to-wall interfaces, demonstrating the need for modeling the deformations in the
wall-contact regions and the need for using concrete confinement in these regions as
described previously. The deformations in the walls away from the contact regions
remain small.
Fig. 3.13 shows the maximum principal stress distributions corresponding to a
beam chord rotation of b=5% from the deformable wall finite element model. The
68
maximum principal stresses remain mostly compressive, demonstrating the effect of gap
opening and post-tensioning on the behavior of the structure. The coupling beam stresses
in Fig. 3.13a clearly show the formation of a diagonal compression strut along the beam
span, with the maximum principal stresses at the beam end reaching close to 414 MPa
(60 ksi) in compression. The stresses outside the diagonal compression strut remain small
as a result of gap opening.
beam shear force, Vb (kN)
800
600
400
prototype
subassemblage
200
deformable wall FE model (ABAQUS)
rigid wall FE model (ABAQUS)
2
3
4
1
Fig. 3.12 Rigid wall versus deformable wall finite element models
Similarly, Fig. 3.13b shows the maximum principal stresses in the reaction block,
where the stress concentrations in the post-tensioning anchorage region and the wallcontact region can be seen. The wall-contact region stresses are similar to the beam
flange stresses near the beam-to-wall interface (demonstrating the need for the steel
embedded plates and the concrete confinement) and quickly diminish away from the
interface as assumed in the wall-contact elements used in the fiber element model. As a
result of gap opening, the concrete stresses outside the wall-contact region remain small.
3.4.3
Comparisons Between Finite Element and Fiber Element Model Results
Fig. 3.14a compares the load-deformation relationships of the prototype

subassemblage using rigid wall fiber element and finite element models. In this
comparison, the wall regions are assumed to be rigid so that the modeling of the coupling
beam and gap opening at the beam-to-wall interfaces can be evaluated. The results
indicate that the fiber element model provides a remarkable representation of the
nonlinear behavior of the beam including the effect of gap opening.
69
diagonal compression strut

b = 5%
coupling beam
maximum principal stresses at beam end
(a)
beam end
PT
anchor
end
b = 5%
reaction block
maximum principal stresses

in wall contact region
(b)
Fig. 3.13 Maximum principal stress distributions: (a) beam stresses; (b) wall stresses
70
1000
800
600
prototype
subassemblage
400
200
finite element model (ABAQUS)

fiber element model (DRAIN-2DX)
2
3
4
1
(a)
800
600
1.0
contact depth/beam depth, cb/db
1000
db= 718 mm
d b = 577 mm
(prototype subassemblage)
400
200

2
3
4
1
(b)
0.8

0.6
0.4
0.2
prototype
subassemblage
2
3
4
1
beam chord rotation, b(percent)
(c)
Fig. 3.14 Model verification: (a) rigid wall comparison;

(b) deformable wall comparison; (c) contact depth comparison
Similarly, Fig. 3.14b compares the load-deformation relationships of the fiber
element and finite element models including the deformations in the wall regions. The
objective of this comparison is to evaluate the modeling of the wall regions, in particular,
the local deformations in the wall-contact regions using the wall-contact elements.
Comparisons are provided for two subassemblages with different beam depths: db = 577
mm (i.e., the prototype subassemblage) and db = 718 mm. Again, a reasonable agreement
between the two models is observed.
Finally, Fig. 3.14c compares the contact depth, cb between the beam and the
reaction block in the prototype subassemblage using the deformable wall fiber element
and finite element models. The contact depth is normalized with respect to the beam
depth, db.
71
The results in Fig. 3.14 show that the fiber element model is capable of predicting
both the global (e.g., load-deformation) behavior and the local (e.g., gap opening and
contact) behavior of unbonded post-tensioned hybrid coupled wall subassemblages
remarkably well. The fiber element model is used to conduct the analyses described in
the remainder of this report because of its simplicity and computational efficiency as
compared with the finite element model.
3.5
Advantages and Limitations of Fiber Element Model
A significant advantage of the fiber element model described in this chapter is that
a reasonably accurate representation of the behavior of the coupling beams, the posttensioning tendons, the wall-contact regions, and the beam-to-wall connection regions
can be developed using uniaxial stress-strain models for concrete, wall steel, beam steel,
and post-tensioning steel, the properties of the top and seat angles, and the geometry and
dimensions of the coupling beam and wall regions. The fiber element model provides an
effective tool to conduct nonlinear monotonic and cyclic load analyses of unbonded posttensioned hybrid coupled wall subassemblages accounting for: (1) nonlinear axialflexural interaction in the coupling beams; (2) linear-elastic shear behavior in the
coupling beams; (3) local deformations in the wall-contact regions; (4) nonlinear cyclic
behavior of the angles; (5) nonlinear cyclic behavior of the steel and concrete, including
yielding of the steel and crushing of the concrete; (6) gap opening and closing at the
beam-to-wall interfaces; and (7) elongation of the post-tensioning tendons due to gap
opening. Analytical models of multi-story coupled wall structures can be constructed by
combining the subassemblage fiber element models for the floor and roof levels as
described in Chapter 6.
The main limitations of the fiber element subassemblage model are as follows:
(1) The desired behavior of unbonded post-tensioned coupling beams is governed,
primarily, by the opening of gaps at the beam-to-wall interfaces. Shear slip at the beamto-wall interfaces should be prevented by design, and thus, is not modeled.
(2) The fiber element model accounts for the linear and nonlinear axial-flexural
deformations (including gap opening) and linear shear deformations of the coupling
beams; however, nonlinear shear deformations are not modeled. Nonlinear shear
deformations are not expected to occur in the coupling beams investigated in this report
as a result of the gap opening behavior and the transfer of shear forces through the
formation of a diagonal compression strut in the beam. Note that nonlinear shear
deformations may be significant in beams with small span-to-depth aspect ratios and with
thin webs. Modeling of these types of beams is not within the scope of this report.
(3) The degradation (if any) in the coupling stiffness and resistance of the
subassemblages due to increasing lateral displacements is modeled; however, any
72
additional degradation under repeated displacement cycles to a constant amplitude is not

captured.
(4) Local and/or global instability of the coupling beams should be prevented by
design, and thus, is not modeled.
(5) As discussed in Chapter 2, the nonlinear behavior of top and seat angles is
complicated and the modeling of the angles in this report has a number of limitations.
Angle plastic hinge mechanisms other than the vertical-leg mechanism in Fig. 3.9a can be
modeled by changing the assumed angle force-deformation behavior; however, this is not
investigated. Nonlinear behavior and/or failure of the angle-to-beam and angle-to-wall
connections should be prevented by design, and thus, is not modeled. Similarly, low
cycle fatigue fracture of the angle legs at the plastic hinges is not modeled.
(6) The model assumes that the post-tensioning tendons remain straight during the
lateral displacements of a subassemblage. Kinking of the tendons at the beam-to-wall
interfaces can occur if the tendons come into contact with the ducts under large
displacements of the subassemblage or if the ducts are not oversized (in which case
kinking would occur as soon as gap opening occurs). A revised analytical model is
presented in Chapter 10 based on the subassemblage experiments described in Chapter 9.
The revised model includes the kinking of the post-tensioning tendons at the beam-towall interfaces, as well as other improvements to the analytical model described in this
chapter.
3.6
Modeling of Embedded Steel Coupling Beams
This section describes an analytical model for hybrid coupled wall subassemblages
with embedded steel beams, similar to the subassemblage in Fig. 2.5d. This analytical
model is used later in the report to compare the behaviors of hybrid coupled wall
structures with unbonded post-tensioned beams and with embedded beams.
As described in Chapter 2, the nonlinear behavior of embedded steel coupling
beams can be characterized as shear critical or flexure critical. The analytical model
described below is limited to flexure critical beams with a predominantly flexural, rather
than shear, mode of behavior. As an example, Fig. 3.15a shows a flexure critical
embedded steel coupling beam subassemblage (Specimen S4 as shown in Figs, 2.5d, 2.5e,
and 2.5f) tested by Harries et al (1997). The beam has a W1422 (U.S. designation)
section with a clear span length of lb=1200 mm and an embedment length of lbe=600 mm.
Flange cover plates with 5 mm thickness are used over the embedded lengths of the beam.
73
LEFT WALL REGION
RIGHT WALL REGION

5 mm cover plate
on embedded flanges
2- 1062 mm stiffeners
symmetric on web
flexure critical W1422
349 mm
8127 mm flange
6 mm web
1200 mm
600 mm
600 mm
(a)
LEFT WALL REGION
RIGHT WALL REGION

kinematic
kinematic
constraint
constraint
beam elements
16
wall-height
elements
17
7
18
19
l be
21
12 13 14
2 3 45 6
Node 1
20
15
10 11
22
(1/3)l be
23
effective beam length l b,eff
lw
lb
lw
(b)
steel fiber stress (MPa)
600
400
200
0
200
400
600
0.08
0.04
0.04
0.08
steel fiber strain

(c)
Fig. 3.15 Modeling of embedded steel coupling beams: (a) Specimen S4 (Harries et al.
1997); (b) analytical model; (c) steel stress-strain relationship
74
Fig. 3.15b shows a fiber element analytical model for Specimen S4. Similar to the
modeling of unbonded post-tensioned subassemblages, the DRAIN-2DX program
(Prakash et al. 1993) is used as the analytical platform. Fiber beam-column elements are
used to model the coupling beam and the wall regions (referred to as beam elements and
wall elements, respectively, in Fig. 3.15b). Previous researchers (Shahrooz et al. 1993a,
1993b; Gong et al. 1997, 1998; Harries et al. 1997) have shown that the flexibility inside
the embedment regions needs to be taken into account to ensure that the subassemblage
forces and displacements are computed with reasonable accuracy. Based on previous
experimental results (Shahrooz et al. 1993a, 1993b, Gong et al. 1997, 1998), the effective
length, lb,eff of the coupling beam is assumed to be equal to the clear span length, lb plus
one-third of the embedment length, lbe at each end as
lb,eff = lb + (2 / 3)lbe
(3.1)
Nodes 2 and 14 represent the effective length of the coupling beam in Fig. 3.15b.
The displacements and rotations of these nodes are kinematically constrained to
corresponding nodes at the center of each wall region (Nodes 1 and 15, respectively).
Fig. 3.15c shows the cyclic stress-strain relationship used for the steel fibers
modeling the coupling beam. Note that the Bauschinger effect of the steel is considered
in this model under cyclic loading.
Similarly, Fig. 3.16 shows the analytical model for an embedded steel coupling
beam specimen tested by Remmetter et al. (1992). Only one half the length of the beam is
tested embedded inside a wall specimen.
Fig. 3.17a compares the measured hysteretic load versus displacement behavior of
Specimen S4 (Harries et al. 1997) with the predicted behavior from the analytical model.
Similarly, Fig. 17b compares the measured and predicted behaviors of Specimen 3 tested
by Remmetter et al. (1992). Good agreement is observed between the analytical model
and the test results for both specimens.
75
(a)
13 mm
TEST WALL REGION
TEST WALL REGION
13 mm
kinematic
constraint
beam elements
203 mm
229 mm
102 mm
2 3 45 6
Node 1
13 mm web
7
l be
305 mm
305 mm
(1/3)l be
effective beam length l b,eff
lw
lb
(b)
500
steel fiber stress (MPa)
400
300
200
100
0
100
200
300
400
500
-0.08 -0.06 -0.04
-0.02
0.02
0.04
0.06
0.08
steel fiber strain
(c)
Fig. 3.16 Modeling of embedded steel coupling beams: (a) Specimen 3 (Remmetter et al.
1992); (b) analytical model; (c) steel stress-strain relationship
76
Fig. 3.17 Verification of embedded steel coupling beam model: (a) Specimen S4 (Harries
et al. 1997); (b) Specimen 3 (Remmetter et al. 1992)
3.7
Chapter Summary
This chapter describes a fiber element analytical model to investigate the nonlinear
behavior of floor-level unbonded post-tensioned hybrid coupled wall subassemblages.
The details of a prototype coupled wall subassemblage are also presented. The purpose of
the analytical model is to provide a basis for comparisons with subassemblage
experimental results as described later in the report and to provide a basis for the
modeling of multi-story coupled wall structures. A preliminary verification of the fiber
element subassemblage model is done based on comparisons with finite element models.
An analytical model for embedded steel coupling beam subassemblages is also developed
77
in the chapter to investigate the differences in the seismic behaviors of coupled wall
structures with unbonded post-tensioned and embedded steel beams later in the report.
78
CHAPTER 4
NONLINEAR BEHAVIOR OF COUPLED WALL SUBASSEMBLAGES
This chapter describes an analytical investigation on the nonlinear moment versus
rotation behavior of unbonded post-tensioned hybrid coupled wall subassemblages as
follows: (1) overview of subassemblage analyses; (2) behavior under monotonic loading;
(3) behavior under cyclic loading; and (4) parametric analyses. These subassemblage
analyses are necessary to form a foundation for the investigation of multi-story coupled
wall behavior later in the report. The prototype subassemblage and the fiber element
model presented in Chapter 3 are used to conduct the analyses.
4.1
Overview of Subassemblage Analyses
Based on the fiber element and finite element analysis results described in Chapter
3, it was found that the deformations in the wall-height elements of the subassemblage
model are not significant. Thus, as shown in Fig. 4.1, the wall-height elements are
removed from the fiber element model for the purpose of the subassemblage analyses.
The reaction block is assumed to be fixed at Node 1 (see also Fig. 3.3), and the loading
block (Node 19) is allowed to translate in the horizontal and vertical directions, but not
allowed to rotate. The degrees of freedom (DOFs) of angle element Nodes 34 and 36 are
constrained to Node 1 and the DOFs of Nodes 39 and 41 are constrained to Node 19.
fixed node
28
29
30
angle element
35 beam elements
slope =1:3
34
wallcontact
elements
39
38
15
5
1
19
kinematic
constraint
36
37
PT
40
element
31
32
33
41
kinematic
constraint
lw
lb
lw
Fig. 4.1 Simplified subassemblage model

Following the application of the beam post-tensioning forces, a vertical force V is
applied at Node 19. The subassemblage analyses are conducted in displacement control
based on the relative displacement, ub (see Fig. 4.2a) of the loading block (Node 19) in
79
the vertical direction with respect to the reaction block (Node 1). Note that since the
loading and reaction blocks are not allowed to rotate, the relative vertical displacement
between Nodes 5 and 15 at the beam ends is equal to the control displacement ub as
shown in Fig. 4.2b.
RIGHT WALL REGION
V
LEFT WALL REGION
(a)
ub
15
V
19
ub
5
1
(b)
Fig. 4.2 Displaced shape: (a) subassemblage; (b) analytical model
4.2
Behavior Under Monotonic Loading
Fig. 4.3a shows the beam end moment versus chord rotation (Mb-b) behavior of
the prototype subassemblage in Chapter 3 as determined using the simplified fiber
element model described above. The corresponding total force in the beam posttensioning tendons, Pb is shown in Fig. 4.3b, contact depth at the beam-to-wall interfaces,
cb is shown in Fig. 4.3c, extreme beam flange and cover plate compression strains are
shown in Fig. 4.3d, and tension angle force versus deformation behavior is shown in Fig.
4.3e. The beam end moment is equal to the coupling moment calculated as Mb=Vlb/2 and
the beam chord rotation is calculated as b=ub/lb. The contact depth, cb is normalized with
respect to the beam depth, dbc including the flange cover plates. The dashed horizontal
lines in Figs. 4.3b, 4.3c, and 4.3d represent, respectively, the total initial force in the
beam post-tensioning tendons, Pbi, the thickness of the beam flange, tbf plus the cover
plate thickness, tc, and the assumed yield strain of the beam steel, sy.
80
beam end moment, Mb (kN.m)
3000
M bp
M by
beam
PT yielding
2000
beam flange
yielding
cover plate yielding

tension angle yielding
1000
softening
decompression
2
4
(a)
beam contact dpeth, cb /beam depth, dbc
beam PT tendon force, Pb (kN)
5000
4000
Pbi =2814 kN
3000
2000
1000
2
4
1
beginning of gap opening
0.8
0.6
0.4
0.2
(tc +tbf )/dbc =0.116

2
4
(b)
(c)
1600
0
sy =0.00181
beam
flange strain
tension angle force (kN)
extreme beam compression strain
-0.01
beam cover
plate strain
-0.02
1200
800
400
-0.035
0
2
4
(d)
-200
-5
10
20
tension angle deformation (mm)
35
(e)
Fig. 4.3 Subassemblage behavior under monotonic loading: (a) beam end moment versus
chord rotation behavior; (b) total force in beam post-tensioning tendons; (c) contact
depth; (d) maximum beam compression strain; (e) tension angle behavior
81
As the subassemblage deforms, it goes through five states of response: (1)

decompression (i.e., initiation of gap opening), as shown by the markers in Fig. 4.3; (2)
yielding of the angles in tension, as shown by the markers; (3) yielding of the flange
cover plates in compression at the beam-to-wall interfaces, as shown by the markers; (4)
yielding of the beam flanges in compression at the beam-to-wall interfaces, as shown by
the markers; and (5) yielding of the beam post-tensioning tendons, as shown by the +
markers.
The first state in the moment-rotation behavior of the subassemblage is
decompression at the beam-to-wall interfaces. The beam moment and rotation at this state
are referred to as Mb,dec and b,dec, respectively. Decompression represents the beginning
of nonlinear behavior due to gap opening (as indicated by a reduction in the contact depth
in Fig. 4.3c) when the precompression due to post-tensioning is overcome by the applied
lateral load. The lateral stiffness of the subassemblage in Fig. 4.3a before the
decompression state is similar to the stiffness of a comparable subassemblage with an
embedded steel beam. The results indicate that the effect of gap opening (i.e., reduction
in contact depth) on the subassemblage stiffness is small until the gaps extend over a
significant portion of the beam depth and reach close to the compression flange of the
beam.
Fig. 4.3a shows that a significant reduction in the stiffness of the subassemblage
occurs between 0.5% and 0.7% rotation due to a combined effect of increased gap
opening, local deformation of the walls in the contact regions, yielding of the angles in
tension, and yielding of the flange cover plates in compression. This combined state is
referred to as the beam softening state, where the term softening is used to describe the
reduction in the stiffness of the structure. The beam moment and rotation at the softening
state are referred to as Mb,sof and b,sof, respectively. It is observed in Fig. 4.3b that the
increase in the beam post-tensioning force due to gap opening at the softening state is
relatively small. Furthermore, the results in Fig. 4.3c indicate that the beam contact depth
remains relatively constant after the softening state is reached.
The moment resistance of the subassemblage continues to increase beyond the
softening state. This occurs as a result of the post-yield stiffness of the tension angles and
the stiffness of the post-tensioning tendons (which are still linear-elastic). The softening
state is followed by the yielding of the beam flanges in compression (at the beam-to-wall
interfaces) at a beam moment Mb,bfy and rotation b,bfy. It is observed in Fig. 4.3d that the
maximum compression strains in the flange cover plates are significantly larger than the
maximum beam flange strains.
The subassemblage is designed such that the yielding of the post-tensioning
tendons, referred to as the beam PT-yielding state, occurs at a large rotation (as shown in
Fig. 4.3b). The linear limit point on the stress-strain relationship of the post-tensioning
strands (corresponding to fbpy=1689 MPa in Fig. 3.2b) is used to identify the yielding of
the tendons. Note that the elongations in the tendons due to the deformation of the
subassemblage are the same regardless of their location in the beam cross section since
82
the left and right wall regions are not allowed to rotate relative to each other as shown in
Fig. 4.2. Thus, the tendons yield at the same beam rotation as long as they are prestressed
by the same amount. The beam moment and rotation at the PT-yielding state are referred
to as Mb,pty and b,pty, respectively.
As described in Chapter 3, the prestressing of the beam post-tensioning tendons is
simulated by tensile forces in the beam PT elements, which are equilibrated by
compression forces in the beam and wall-contact elements. Since a small amount of posttensioning force is lost due to the elastic shortening of the beam and wall-contact
elements, a larger amount of tensile force is applied to the beam PT elements in order to
achieve the desired amount of initial force (i.e., force just before the application of lateral
loads) in the post-tensioning tendons after elastic shortening. For the prototype
subassemblage, a total force of Pbj=2906 kN (653.4 kips) is applied in the beam PT
elements to achieve the desired initial post-tensioning force of Pbi=nbtabtfbpi=2814 kN
(633 kips) where, abt=420 mm2 is the beam post-tensioning tendon area, nbt=6 is the
number of beam post-tensioning tendons, and fbpi=0.6fbpu (with fbpu=1862 MPa) is the
desired initial stress in the tendons.
The behavior of the subassemblage beyond the beam PT-yielding state is not
investigated in Fig. 4.3. The dashed horizontal lines in Fig. 4.3a show the yield moment,
Mby and the plastic moment, Mbp of the W21182 section used for the prototype beam.
The unbonded post-tensioned subassemblage reaches the beam softening state at
approximately Mb,sof = 0.60Mby and the PT-yielding state at approximately Mb,pty =
0.80Mbp (or Mb,pty = 0.91Mby). The beam moment Mb,sof is taken as the moment when the
tension angles yield. The results indicate that, as a result of gap opening, the strength of
an unbonded post-tensioned subassemblage is smaller than the strength of a comparable
subassemblage with a properly-designed and detailed embedded steel beam that achieves
the full yield and plastic capacity of the beam cross-section.
4.3
Behavior Under Cyclic Loading
Fig. 4.4a shows the hysteretic moment-rotation behavior of the prototype

subassemblage under reversed cyclic loading, as determined using the fiber element
model in Fig. 4.1. The subassemblage has stable behavior under large nonlinear cyclic
rotations. The thick curve represents the behavior under monotonic loading as described
above. The hysteresis loops show that the subassemblage has smaller inelastic energy
dissipation but larger self-centering capability as compared to systems with embedded
steel coupling beams (e.g., see Fig. 3.17).
The large self-centering capability of the subassemblage indicates that the posttensioning tendons provide a sufficient amount of restoring force to yield the tension
angles back in compression and close the gaps at the beam-to-wall interfaces. The total
force, Pb in the post-tensioning tendons corresponding to the hysteretic behavior in Fig.
4.4a is shown in Fig. 4.4b. The total post-tensioning force is normalized with respect to
the total strength Abpfbpu of the tendons, where Abp=nbtabt. Since the yielding of the posttensioning tendons is significantly delayed as a result of unbonding, only a small amount
83
of loss in the initial post-tensioning force occurs during the reversed cyclic displacements
of the structure.
Figs. 4.4c and 4.4d investigate the effect of the top and seat angles on the hysteretic
behavior of the subassemblage. The moment-rotation behavior in Fig. 4.4c is for a
subassemblage with smaller angles having a L883/4 cross-section. Similarly, Fig. 4.4d
shows the behavior of the prototype subassemblage with the angles removed.
1
initial PT
force
monotonic
cyclic
Pb/(A bp f bpu )
beam end moment, M b (kN.m)
2500
final PT
force
0.5
L8x8x1-1/8
-2500
-6
0
-6
(a)
(b)
2500
2500
monotonic
cyclic
-2500
-6
L8x8x3/4
monotonic
cyclic
-2500
-6
no angles
(d)
(c)
Fig. 4.4 Subassemblage behavior under cyclic loading: (a) with L881-1/8 angles;
(b) total force in post-tensioning tendons; (c) with L883/4 angles; (d) without angles
The cyclic behavior of the subassemblage without angles is very close to nonlinearelastic indicating that the angles provide most of the inelastic energy dissipation of the
structure and that significant damage is confined to the angles only, which are replaceable
after an earthquake. The angle size and amount of post-tensioning can be determined to
satisfy different design requirements as described in detail later in the report. A large
enough post-tensioning force should be provided to yield the tension angles back in
compression during unloading and to develop a sufficient amount of self-centering
84
capability in the subassemblage. If necessary, supplemental energy dissipation devices

can be used to increase the energy dissipation of the system.
4.4
Parametric Analyses
This section presents a parametric investigation of the nonlinear moment-rotation

behavior of unbonded post-tensioned hybrid coupled wall subassemblages under
monotonic loading. Selected structural properties of the prototype subassemblage are
varied, and then, a monotonic analysis of each subassemblage is conducted using the
fiber element model. The results are used to determine how the behavior of the system
can be controlled by design. The varied properties are: (1) area of post-tensioning tendon,
abt; (2) initial stress in post-tensioning steel, fbpi; (3) length of wall, lw; (4) thickness of
flange cover plate, tc; (5) thickness of beam flange, tbf; (6) depth of beam web, dbw; and (7)
length of beam, lb.
The subassemblage moment-rotation relationships from the parametric
investigation are given in Figs. 4.5a-h. Each figure shows the moment-rotation behavior
of three parametric subassemblages. Subassemblage (1) is the same as the prototype
subassemblage from Chapter 3 and the properties of Subassemblages (2) and (3) are
varied as shown in Table 4.1. The markers in Fig. 4.5 represent the states of behavior
identified in Fig. 4.3a.
In Fig. 4.5a, the area abt and initial stress fbpi of the post-tensioning tendons are
varied simultaneously such that the initial post-tensioning force, Pbi = abtfbpi remains
constant. The results indicate that the moments Mb,dec and Mb,sof remain constant when Pbi
is constant. In Figs. 4.5b and 4.5c, abt and fbpi are varied one at a time, respectively. There
is an increase in Mb,dec and Mb,sof, and a decrease in b,bfy as abt is increased. An increase
in fbpi results in decreases in b,bfy and b,pty, while Mb,bfy and Mb,pty are not significantly
affected.
As expected, an increase in the wall length, lw in Fig. 4.5d results in an increase in
b,pty since the unbonded length of the post-tensioning tendons is increased. In Fig. 4.5e,
the most significant effect of an increase in the flange cover plate thickness tc is an
increase in b,bfy. The variables tbf and dbw in Figs. 4.5f and 4.5g represent the beam
flange thickness and web depth, respectively. Both parameters affect the stiffness and
strength of the subassemblage. Finally, Fig. 4.5h shows that the beam length lb does not
have a significant effect on the subassemblage moment-rotation behavior, other than a
change in b,pty due to a change in the unbonded length of the post-tensioning tendons.
Note that the location of the beam post-tensioning tendons is not investigated as a
part of the parametric analyses since the elongations in the tendons due to the
displacements of the subassemblage are the same regardless of the distance of the
tendons from the beam centerline (assuming that the tendons are prestressed by the same
amount and that the left and right wall regions do not rotate relative to each other as
shown in Fig. 4.2). Thus, the subassemblage moment-rotation behavior is not affected by
the location of the post-tensioning tendons.
85
3000
2
1
3
abt=420 mm 2
f bpi=0.60fbpu
2
2 a bt=504 mm
decompression
f bpi=0.50fbpu
2
beam flange yielding 3 abt=360 mm
f bpi=0.70fbpu
PT tendon yielding
3000
1
2
3
decompression
beam flange yielding
PT tendon yielding
3000
1
decompression
PT tendon yielding
fbpi=0.60fbpu
fbpi=0.50fbpu
fbpi=0.70fbpu
3
1 l w =3.05 m
2
decompression
PT tendon yielding
l w=3.81 m
l w=4.57 m
(d)
2
1
3
1
2
decompression
PT tendon yielding
tc =38.1 mm
t c =63.5 mm
tc =12.7 mm
3000
3
2
1
2
3
tbf =37.6 mm
t bf =63.0 mm
tbf =88.4 mm
decompression
PT tendon yielding
(e)
(f )
3
3000
3000
(c)
2
1
1
2
3
decompression
PT tendon yielding
(b)
3000
2
1
3
(a)
abt=420.0 mm 2
abt=560.0 mm 2
abt=280.0 mm2
dbw=502 mm
dbw=604 mm
dbw=705 mm
(g)
3000
2
1
2
decompression
PT tendon yielding
l b=3.05 m
l b=3.81 m
l b=4.57 m
(h)
Fig. 4.5 Subassemblage moment-rotation behaviors from parametric investigation:

(a) abt and fbpi; (b) abt; (c) fbpi; (d) lw; (e) tc; (f) tbf; (g) dbw; (h) lb
86
TABLE 4.1
PARAMETRIC STUDY OF COUPLED WALL SUBASSEMBLAGES
Parameters Varied
Wall length, lw
Beam flange thickness, tbf
Beam web depth, dbw
Beam length, lb
Subassemblage (1)
420 mm2
0.6fbpu
420 mm2
0.6fbpu
3.05 m
38.1 mm
37.6 mm
502 mm
3.05 m
Subassemblage (2)
504 mm2 (1.20)
0.5fbpu (0.83)
560 mm2 (1.33)
0.5fbpu (0.83)
3.81 m (1.25)
63.5 mm (1.67)
63.0 mm (1.67)
604 mm (1.20)
3.81 m (1.25)
Subassemblage (3)
360 mm2 (0.86)
0.7fbpu (1.17)
280 mm2 (0.67)
0.7fbpu (1.17)
4.57 m (1.50)
12.7 mm (0.33)
88.4 mm (2.35)
705 mm (1.40)
4.57 m (1.50)
Note: Numbers in parentheses show parameter values for Subassemblage (2) or (3) divided by value for
Subassemblage (1).
The parametric investigation described above provides useful information to

control the behavior of unbonded post-tensioned hybrid coupled wall subassemblages in
design. For example, the results in Fig. 4.5 show that the coupling moment, shear force,
and initial stiffness of the subassemblages can be controlled using the post-tensioning
steel area and the beam depth. the coupling shear force, which is equal to Vb=2Mb/lb, can
also be controlled using the beam length. the rotation capacity at the pt-yielding state can
be controlled by controlling the initial stress and unbonded length of the post-tensioning
tendons.
4.5
Chapter Summary
This chapter presents an analytical investigation on the nonlinear behavior of

unbonded post-tensioned hybrid coupled wall subassemblages under lateral loads. The
effects of structural design parameters such as the amount of post-tensioning, beam
properties, and angle properties on the behavior of the subassemblages, including the
lateral resistance and energy dissipation is investigated. The following conclusions are
made based on the results:
(1) As a result of post-tensioning, the initial stiffness of an unbonded posttensioned hybrid coupled wall subassemblage is similar to the initial stiffness of a
comparable embedded steel coupling beam system. Top and seat angles are used at the
beam-to-wall interfaces for inelastic energy dissipation and redundancy during an
earthquake. The structure can be designed to provide significant and stable levels of
coupling over large nonlinear cyclic deformations with most of the damage occurring in
the angles, which can be replaced after the earthquake.
(2) The effect of gap opening on the subassemblage stiffness is small until the gaps
extend over a significant portion of the beam depth.
(3) The behavior of a properly designed and detailed unbonded post-tensioned
coupled wall subassemblage under lateral loads is characterized by the following five
states: (1) decompression; (2) yielding of the angles in tension; (3) yielding of the flange
87
cover plates (if used) in compression; (4) yielding of the flanges in compression; and (5)
yielding of the beam post-tensioning tendons.
(4) The total beam post-tensioning force remains relatively constant until the
softening state is reached.
(5) The contact depth at the beam-to-wall interfaces remains relatively constant
after the softening state is reached.
(6) The post-tensioning force provides a restoring effect that yields the tension
angles back in compression, closes the gaps, and pulls the walls and the coupling beam
back towards their undisplaced position upon unloading from a large nonlinear
deformation. This results in a large self-centering capability of the structure. The initial
post-tensioning force is maintained as long as the yielding of the tendons is prevented.
(7) The coupling moment can be controlled by the area of the post-tensioning steel
as well as the beam depth, and the yielding of the beam post-tensioning steel can be
controlled by the initial stress and unbonded length of the post-tensioning tendons.
(8) As a result of gap opening at the beam-to-wall interfaces, the coupling moment
strength of an unbonded post-tensioned steel coupling beam is smaller than the moment
strength of an embedded steel coupling beam with the same cross section. This loss in
strength is somewhat compensated by the top and seat angles used at the beam ends.
88
CHAPTER 5
IDEALIZED SUBASSEMBLAGE MOMENT-ROTATION RELATIONSHIP
This chapter presents an idealized beam end moment versus chord rotation
relationship for unbonded post-tensioned hybrid coupled wall subassemblages under
monotonic lateral loading. Procedures are developed to estimate the structure behavior
using basic principles of equilibrium, compatibility, and constitutive relationships. These
procedures are used later in the report as design tools and to develop idealized lateral load
versus displacement relationships for multi-story coupled wall structures. The chapter is
divided into the following sections: (1) linear-elastic behavior; (2) idealized nonlinear
relationship; and (3) verification of idealized relationship.
5.1
Linear-Elastic Behavior
The linear-elastic behavior of an unbonded post-tensioned hybrid coupled wall

subassemblage under lateral loads can be approximated using the simplified model in Fig.
5.1a. The cross-sectional properties and boundary conditions of the model are determined
from the fiber element model in Figs. 3.3a and 3.3b, based on the following general
assumptions/approximations:
(G1) The deformations in the wall-height elements are negligible.
(G2) The left and right wall regions do not rotate relative to each other.
In the linear-elastic range of behavior, the effects of the post-tensioning tendons
and the top and seat angles are ignored. Gap opening at the beam ends causes geometric
nonlinear effects on the structure, and thus, is not included in the linear-elastic model.
Due to symmetry, only one half the length of the subassemblage is used, from the
center of the reaction block (Node 1) to the midspan of the coupling beam (Node 10).
The center of the reaction block is fixed and the beam midspan is free. A vertical force Vb
is applied at the free end as shown in Fig. 5.1a. The Y-translational DOF of Node 5,
which represents the beam-to-wall interface, is restrained. Thus, the model is statically
indeterminate to the first degree.
The cross-sectional properties of the model in Fig. 5.1a vary along the length. The
properties between Nodes 1 and 5 are the same as the properties used for the wall-contact
elements in the fiber element model in Fig. 3.3b. The cross-sectional properties between
Nodes 5 and 8 and between Nodes 8 and 10 are determined based on the beam cross
section with and without the cover plates (if used), respectively.
89
lw
20
slope=1:3
21
29
3
4
24
X
beam
7
6
11
9
10
25
13
12
15 17
18
14 16
cover plate
22
27
lb
0.5l w
1.5dbc
0.5d
dbc bc
32
26
te
23
19
fiber element model
Vb
1
0.5l w
lc
10
linear-elastic model
0.5l b- l c
(a)
Vb
5
lc
10
0.5l b - l c
(b)
Fig. 5.1 Linear-elastic subassemblage model: (a) model with wall-contact elements;
(b) model without wall-contact elements
The linear-elastic lateral stiffness of the subassemblage is defined using the beam
end moment versus chord rotation (Mb-b) relationship as:
Kbi =
Mb
b
90
(5.1)
and can be determined using an appropriate linear-elastic structural analysis procedure

based on the model in Fig. 5.1a. Closed-form expressions for Kbi can also be developed.
Note that a simpler linear-elastic subassemblage model can be obtained by ignoring
the deformations in the wall-contact elements, resulting in a small overestimation of the
initial stiffness as shown in Fig. 3.12. The linear-elastic subassemblage model without the
wall-contact elements is shown in Fig. 5.1b. This model is statically determinate.
Ignoring the effect of the beam flange cover plates (if any) and including the shear
deformations of the beam, the subassemblage initial stiffness, Kbi can be determined as:
K bi =
where,
bg =
6 Eb I b
1
lb 1 + 2 bg
6 Eb I b
Gb Abg lb2
(5.2)
(5.3)
with Gb (shear modulus of the coupling beam steel) and Abg (shear area of the coupling
beam cross section) given as,
Gb =
Eb
and Abg = dbtbw
2(1 + vb )
(5.4)
and lb is the coupling beam span length (as shown in Fig. 5.1), Ib is the moment of inertia,
db is the depth, and tbw is the web thickness of the coupling beam cross section (ignoring
the beam flange cover plates), and b and Eb are the Poissons ratio and modulus of
elasticity of the beam steel, respectively (assumed as b=0.3 and Eb=199955 MPa,
resulting in Gb=76907 MPa).
For the prototype subassemblage in Chapter 3, initial linear-elastic stiffness values
of Kbi=4.25108 kN-mm (3.77106 kip-in.) and Kbi=5.01108 kN-mm (4.43108 kip-in.)
result from the models in Figs. 5.1a and 5.1b, respectively, which compare well with the
value of Kbi=4.17108 kN-mm (3.69106 kip-in.) from the DRAIN-2DX fiber element
model including the beam cover plates and the wall-contact elements.
5.2
Idealized Nonlinear Relationship
The nonlinear beam end moment versus chord rotation behavior of an unbonded
post-tensioned hybrid coupled wall subassemblage is idealized using a bilinear
relationship. As an example, Fig. 5.2 shows the bilinear moment-rotation relationship
estimated for the prototype subassemblage in Chapter 3. The idealized moment-rotation
relationship is identified by the coupling beam softening state (at Mb,sof, b,sof) and the PTyielding state (at Mb,pty, b,pty). A procedure to estimate Mb,sof, b,sof, Mb,pty, and b,pty is
developed below using basic principles of equilibrium, compatibility, and constitutive
relationships. This estimation procedure can be used as a tool in the seismic analysis and
91
design of coupled wall structural systems, without the need to develop nonlinear fiber
element models.
3000
bilinear estimation
fiber element model
2000
softening
(M b,sof , b,sof )
prototype
subassemblage
1000
PT-yielding
(M b,pty , b,pty )
decompression
PT tendon yielding
estimation points
2
4
Fig. 5.2 Idealized bilinear subassemblage moment-rotation relationship

The following general assumptions/approximations are made for the development
of the bilinear subassemblage relationship, in addition to assumptions (G1)-(G2)
described previously:
(G3) Local and/or global instability of the coupling beam does not occur.
(G4) The post-tensioning anchorages, angle-to-beam and angle-to-wall connections,
and cover-plate-to-beam-flange welds (if cover plates are used) are properly designed and
detailed for the maximum forces and deformations.
(G5) The coupling beam does not slip with respect to the walls at the beam-to-wall
interfaces.
(G6) The cover plates (if used) are sufficiently long such that yielding of the
coupling beam does not occur where the plates are terminated.
5.2.1
Beam Softening State
The beam end moment Mb,sof and chord rotation b,sof at the softening state of an
unbonded post-tensioned hybrid coupled wall subassemblage are estimated based on the
following assumptions, in addition to the general assumptions (G1)-(G6) above:
92
(S1) The force in each tension angle is equal to the yield force, Tayx (see Fig. 3.9),
and acts at the middle of the horizontal leg of the angle parallel to the beam flange.
(S2) The force in each compression angle is equal to the slip force, Casx (see Fig.
3.9), of the angle-to-beam connection bolts and acts at the middle of the horizontal leg of
the angle parallel to the beam flange.
(S3) The stress in the beam post-tensioning tendons is equal to the initial stress, fbpi
(as demonstrated in Fig. 4.3b).
(S4) The compression stresses in the beam contact regions at the beam-to-wall
interfaces have a linear (i.e., triangular) distribution.
(S5) The largest compression stress in the beam flange cover plates (or the beam
flanges if cover plates are not used) at the beam-to-wall interfaces is equal to the yield
strength of the steel, fsy.
Based on these assumptions, Fig. 5.3 shows the softening-state free body diagram
of a coupling beam between the midspan and the beam-to-wall interface. Since the
subassemblage displacements in Fig. 3.10 result in a point of inflection at the beam
midspan, only a shear force and an axial force act in the beam at this section. The shear
force is equal to the coupling shear force at the softening state, Vb,sof and the axial force is
equal to the total force in the post-tensioning tendons, Pbi (since no external axial loads
are applied to the subassemblage). The triangular compressive stress distribution shown
at the beam end is based on Assumptions (S4) and (S5).
C asx
cb,sof
C b,sof
dbc/2-(1/3)cb,sof
za=d bc+t a
linear stress
distribution with
maximum stress f sy
beam centerline
V b,sof
t bf +tc
V b,sof
Pbi
Tayx
l b/2
Fig. 5.3 Coupling beam free body diagram at softening state

The steps used in the estimation of Mb,sof and b,sof are presented below:
93
Step 1
Based on Assumptions (G3), (G4), (S1), (S2), and (S3), and using the free body
diagram in Fig. 5.3, estimate the total beam compression force at the beam-to-wall
interface as:
Cb, sof = Pbi + Tayx Casx
(5.5)
where, Pbi is the total initial post-tensioning force as:
Pbi = abt fbpi
(5.6)
Step 2
Estimate the depth of the compression (i.e., contact) region, cb,sof at the beam-towall interface using a linear stress distribution based on Assumptions (S4) and (S5).
Assuming that the compression region is above the coupling beam web (i.e., cb,softbf+tc)
and the width of the cover plate is the same as the beam flange width bbf, then, cb,sof can
be estimated as:
cb, sof =
2Cb, sof
(5.7)
f sy bbf
Note that a non-rectangular compression region should be used if the compression

region extends into the beam web (i.e., cb,sof>tbf+tc). The cover plate thickness tc=0 if no
cover plates are used on the coupling beam.
Step 3
Estimate the moment Mb,sof at the beam-to-wall interface by taking moments about
the centerline of the beam as:
M b, sof = Vb, sof
cb, sof
lb
d
= Cb, sof ( bc
) + 0.5(Casx + Tayx )(dbc + ta )
2
2
3
(5.8)
where, dbc is equal to the total beam depth including the cover plate thicknesses (i.e.,
dbc=db+2tc).
Step 4
Estimate the beam chord rotation at the softening state, b,sof using the initial lateral
stiffness, Kbi from a linear-elastic analysis of the subassemblage (see Section 5.1). Ignore
the rotation that occurs due to gap opening since the effect of gap opening on the stiffness
of the subassemblage before the softening state is small as shown in Fig. 4.3. Thus,
b, sof =
94
M b, sof
K bi
(5.9)
Step 5
The nonlinear behavior of the coupled wall subassemblage can also be represented
using the coupling shear force versus beam chord rotation relationship. Estimate the
coupling shear force at the softening state, Vb,sof, as:
Vb, sof =
5.2.2
2M b, sof
lb
(5.10)
Beam PT-Yielding State
The beam end moment, Mb,pty and chord rotation, b,pty at the PT-yielding state of
an unbonded post-tensioned hybrid coupled wall subassemblage are estimated using an
iterative procedure based on the following assumptions, in addition to the general
assumptions (G1)-(G6) described previously:
(P1) The beam post-tensioning tendons are stressed to the same initial stress, fbpi.
(P2) Strain hardening of the beam and cover plate (if any) steel in compression is
ignored. Thus, the largest compression stress in the beam and cover plates is equal to fsy.
(P3) The compressive stresses in the beam and in the cover plates at the beam-towall interfaces have a uniform (i.e., rectangular) distribution.
(P4) An idealized load-deformation relationship is available for the angles in
tension.
(P5) The force in each compression angle is equal to the slip force, Casx (see Fig.
3.9), of the angle-to-beam connection bolts and acts at the middle of the horizontal leg of
the angle parallel to the beam flange.
Based on these assumptions, Fig. 5.4 shows the PT-yielding-state free body
diagram of a coupling beam between the midspan and the beam-to-wall interface. The
beam shear force is equal to the coupling shear force at the beam PT-yielding state, Vb,pty
and the beam axial force at midspan is equal to the total force in the post-tensioning
tendons at yield, Pb,pty. The rectangular compressive stress distribution shown at the beam
end is based on Assumptions (P2) and (P3).
The steps used in the estimation of Mb,pty and b,pty are presented below:
Step 1
Select an initial value for the tension angle force at PT-yielding, Ta,pty. Since the
tension angles are assumed to yield at the softening state, Ta,pty>Tayx. An initial value of
Ta,pty may be selected as:
Ta, pty = Tayx
95
(5.11)
C asx
cb,pty
C b,pty
dbc/2-(1/2)cb,pty
rectangular stress
distribution with
maximum stress f sy
za=dbc+t a
beam centerline
V b,pty
t bf +t c
V b,pty
Pb,pty
Ta,pty
l b/2
Fig. 5.4 Coupling beam free body diagram at PT-yielding state
Step 2
Based on the free body diagram in Fig. 5.4, estimate the total beam compression
force at the beam-to-wall interface as:
Cb, pty = Pb, pty + Ta, pty Casx
(5.12)
where, Ta,pty is from Step 1. As described in Chapter 4, the post-tensioning tendons go

through the same elongation regardless of their locations within the beam depth. Thus,
based on Assumption (P1), the tendons yield simultaneously and
Pb, pty = abt fbpy
(5.13)
Step 3
Estimate the depth of the compression (i.e., contact) region, cb,pty at the beam-towall interface using a uniform contact stress distribution based on Assumptions (P2) and
(P3). Assuming that the compression region is above the coupling beam web and the
width of the cover plate (if any) is the same as the beam flange width bbf, then, cb,pty can
be estimated as:
cb, pty =
Cb, pty
f sy bbf
(5.14)
A non-rectangular compression region should be used if cb,pty>tbf+tc. The cover plate

thickness tc=0 if no cover plates are used on the coupling beam.
96
Step 4
Estimate the elongation of the beam post-tensioning tendons between the initial
prestress state and the PT-yielding state as:
ub, pty =
( fbpy fbpi )lbpu
(5.15)
Ebp
where, lbpu=lb+2lw is the unbonded length and Ebp is the Youngs modulus of the tendons.
Step 5
Estimate the width of the beam-to-wall gap at the centerline of the beam, gb,pty. Fig.
5.5 shows the gap opening that occurs at the beam-to-wall interface at the PT-yielding
state. Since the left and right wall regions do not rotate relative to each other, the
centerline gap at the two ends of the beam have the same width, and thus,
gb, pty =
ub, pty
(5.16)
top
angle
ta
tc
cb,pty
PT-tendon
0.5dbc
g b,pty
bg,pty
coupling
beam
a,pty
seat
angle
lw
reaction block
Fig. 5.5 Gap opening at coupling beam PT-yielding state
97
Step 6
Estimate the elongation of the tension angle, a,pty due to gap opening. Based on Fig.
5.5:
a , pty =
dbc + 0.5ta cb, pty

0.5dbc cb, pty
gb, pty
(5.17)
Step 7
Determine the tension angle force Ta,pty using the angle deformation a,pty from Step
6 and the idealized angle load-deformation relationship from Assumption (P4). Repeat
Steps 1-7 using the value of Ta,pty from Step 7 in Step 1 until a satisfactory agreement
between the Ta,pty values from Steps 1 and 7 is achieved. Analyses of the parametric
coupled wall subassemblages from Chapter 4 indicate that convergence can be achieved
in four to five iterations.
Step 8
Based on the free body diagram in Fig. 5.4, estimate the moment Mb.pty at the
beam-to-wall interface by taking moments about the centerline of the beam as:
M b, pty = Vb, pty
lb
= 0.5Cb, pty (dbc cb, pty ) + 0.5(Casx + Ta, pty )(dbc + ta )
2
(5.18)
Step 9
Referring to Fig. 5.5, estimate the rotation of the beam due to gap opening at the
PT-yielding state as:
bg , pty =
gb, pty
0.5dbc cb, pty
(5.19)
where, gb,pty and cb,pty are determined in Steps 5 and 3, respectively.

Step 10
Estimate the elastic rotation of the beam at the PT-yielding state as:
be, pty =
M b, pty
K bi
(5.20)
Step 11
Estimate the total beam chord rotation at the PT-yielding state as:
b, pty = 1.075(bg , pty + be, pty )

where, bg,pty and be,pty are from Steps 9 and 10, respectively.
98
(5.21)
Based on the analyses of the parametric subassemblages from Chapter 4, a factor of

1.075 is necessary in the estimation of b,pty to account for the additional beam rotation
that occurs due to the deformations in the wall-contact regions at the PT-yielding state.
Step 12
Consider second order effects (often referred to as P- effects) as follows:
If the beam post-tensioning ducts are not oversized, then, the tendons
kink at the beam-to-wall interfaces as soon as gap opening occurs and
displace parallel to the coupling beam. Thus, the second order effects in
the tendons counteract the second order effects in the coupling beam and,
M b, pty ( with second order effects) = M b, pty ( from Step 8)
(5.22)
If the beam post-tensioning ducts are large enough such that the posttensioning tendons remain straight (i.e., the tendons do not kink at the
beam-to-wall interfaces). Then, referring to Fig. 5.6, the displaced shapes
of the beam and the post-tensioning tendons differ, resulting in second
order effects as:
M b, pty ( with second order effects) = M b, pty(from Step 8)
reaction block
Pb, ptyb, pty lb
Pb,pty (beam axial

force)
(1
lb
)
lb + 2lw
(5.23)
b,pty
b,pty
b,pty (1-
PT tendon
ne
erli
ent
c
m
bea
lb
)
l b+2 l w
lb
l b+2 l w
Pb,pty
(total force
in PT tendons)
beam
midspan
Fig. 5.6 Second order effects

The parametric analyses from Chapter 4, which assume that the posttensioning tendons remain straight, indicate that the change in Mb,pty due
to second order effects results in a negligible change in b,pty and, thus,
there is no need to update the rotation b,pty from Step 10. Note also that a
consideration of second order effects is not necessary in the estimation of
99
the softening state moment Mb,sof since the beam rotation at the softening
state (i.e., b,sof) is small.
-
If the beam post-tensioning ducts are oversized, but not enough to

prevent kinking of the post-tensioning tendons during the entire
displacement history of the subassemblage, then, an Mb,pty value in
between the values given by Equations (5.22) and (5.23) should be used.
Step 13
Estimate the coupling shear force at the PT-yielding state, Vb,pty, as:
Vb, pty =
5.3
2 M b, pty
lb
(5.24)
Verification of Idealized Relationship
The bilinear estimation of the subassemblage moment versus rotation behavior

described above is verified by comparing the estimated moment and rotation values
corresponding to the softening and PT-yielding states with values determined using the
fiber element model. The Mb,sof and b,sof values from the fiber element model are
determined at the tension angle yielding state from the subassemblage moment-rotation
relationship. Fig. 5.7 shows the comparisons for the Mb,sof and Mb,pty values and Fig. 5.8
shows the comparisons for the b,sof and b,pty values for the parametric subassemblages in
Fig. 4.5. The effect of the wall-contact elements is included in the estimation of Kbi
(using the model in Fig. 5.1a) and second order effects are included in the estimation of
Mb,pty. The results indicate that the estimated moment and rotation values are remarkably
close to the values determined using the fiber element model for a wide range of
parameters. The maximum error in the estimation of Mb,sof, Kbi, Mb,pty, and b,pty is less
than 5%. Thus, it is concluded that the estimation procedure described above can be used
to conduct approximate, simplified analyses of unbonded post-tensioned hybrid coupled
wall subassemblages with different properties.
100
3000
3000
M b,pty
M b,sof
analytical model
bilinear estimation
0
0.8
M b,pty
M b,sof
analytical model
bilinear estimation
0
initial PT stress, fbpi /fbpu
area of PT tendon, a bt (mm2 )
(a)
600
(b)
3000
3000
M b,pty
M b,sof
analytical model
bilinear estimation
0
initial PT stress, f bpi /f bpu
0.8
M b,pty
M b,sof
analytical model
bilinear estimation
0
wall length, l w (m)
(c)
(d)
3000
3000
M b,pty
M b,sof
analytical model
bilinear estimation
0
80
M b,pty
M b,sof
analytical model
bilinear estimation
0
100
thickness of cover plate, t c (mm)
thickness of beam flange, t bf (mm)
(e)
(f )
3000
3000
M b,pty
M b,sof
analytical model
bilinear estimation
0
800
M b,pty
M b,sof
analytical model
bilinear estimation
0
beam length, l b (m)
depth of beam web, dbw (mm)
(g)
(h)
Fig. 5.7 Verification of moment estimations: (a) abt and fbpi;

(b) abt; (c) fbpi; (d) lw; (e) tc; (f) tbf; (g) dbw; (h) lb
101
b,pty
analytical model
bilinear estimation
b,sof
0
0.8
b,pty
analytical model
bilinear estimation
b,sof
0
area of PT tendon, abt (mm2 )
(b)

(a)
8
b,pty
analytical model
bilinear estimation
b,sof
0
0.8
b,pty
analytical model
bilinear estimation
b,sof
0
b,pty
analytical model
bilinear estimation
b,sof
thickness of cover plate, t c (mm)
80
b,pty
analytical model
bilinear estimation
b,sof
0
(e)
8
b,pty
analytical model
bilinear estimation
b,sof
0
thickness of beam flange, t bf (mm)
100
(f )
(d)
(c)
600
800
b,pty
analytical model
bilinear estimation
b,sof
0
depth of beam web, d bw(mm)
(g)
(h)
Fig. 5.8 Verification of rotation estimations: (a) abt and fbpi;

(b) abt; (c) fbpi; (d) lw; (e) tc; (f) tbf; (g) dbw; (h) lb
102
5.4
Chapter Summary
This chapter describes an analytical procedure to estimate the nonlinear lateral load
versus displacement behavior of unbonded post-tensioned hybrid coupled wall
subassemblages by quantifying two selected limit states. The estimation procedure, which
is verified with results from the DRAIN-2DX program (Prakash et al. 1993), can be used
to conduct approximate, simplified analyses of coupled wall subassemblages with
different structural properties. This procedure is used later in Chapter 8 to estimate the
behavior of multistory coupled wall systems under lateral loads and in the seismic design
approach proposed in Chapters 11 and 12.
103
CHAPTER 6
ANALYTICAL MODELING OF MULTI-STORY COUPLED WALLS
This chapter describes the analytical modeling of multi-story hybrid coupled wall
structures as follows: (1) prototype coupled wall structures; (2) analytical modeling
assumptions; (3) multi-story coupled wall models; (4) verification of wall models; (5)
advantages and limitations of analytical models; and (6) modeling of walls with
embedded coupling beams.
6.1
Prototype Coupled Wall Structures
This section describes the design properties of two prototype eight-story unbonded
post-tensioned hybrid coupled wall structures. The plan view of the building for which
the structures were designed is shown in Fig. 6.1. This plan layout was determined based
on the recommendations from the U.S.-Japan Cooperative Earthquake Research
Program on Composite and Hybrid Structures (Goel and Yamanouchi 1992) funded by
the National Science Foundation. The building was designed for a site with a stiff soil
profile [Site Class D in IBC-2000 (ICC 2000)] in a region with high seismicity (e.g.,
coastal California).
A set of four identical unbonded post-tensioned hybrid coupled wall structures
provide the entire lateral load resistance in the east-west direction of the building. The
structural walls providing lateral load resistance in the north-south direction of the
building are not addressed in this report. The properties of the prototype coupled wall
structures are presented below.
6.1.1
Walls
Each prototype coupled wall structure includes two identical concrete walls. The
wall height hw=32.6m (107 ft) [with 4.88m (16 ft) for the first story and 3.96m (13 ft) for
the upper stories], wall length lw=3.05m (10 ft), and wall thickness tw=356mm (14 in.
uniform).
Two types of structures are considered: (1) a structure with monolithic cast-in-place
reinforced concrete walls (referred to as Wall CIP-UPT); and (2) a structure with precast
concrete walls (referred to as Wall PRE-UPT). The elevation views and cross-sections
near the base (single wall only) of these two structures are shown in Figs. 6.2a and 6.2b,
and the general design properties of the walls are listed in Tables 6.1 and 6.2, respectively.
Note that this report focuses on walls with rectangular cross sections; however, it may be
possible to extend the results to other types of cross sections (e.g., T- or U-shaped walls)
as well.
104
(28ft)
8.53m
concrete
wall
prototype system
(28ft)
8.53m
steel beam
(28ft)
8.53m
6.10m
(20ft)
6.10m
(20ft)
6.10m
(20ft)
6.10m
(20ft)
6.10m
(20ft)
PLAN VIEW
Fig. 6.1 Plan view of prototype structures

The monolithic cast-in-place reinforced concrete walls in Fig. 6.2a are conventional
walls with bonded mild steel reinforcement to provide flexural resistance. The precast
concrete walls in Fig. 6.2b are similar to the wall shown in Fig. 2.15, and are constructed
by joining two-story tall precast concrete wall panels along horizontal joints using high
strength post-tensioning bars running in the vertical direction. The post-tensioning bars
are prestressed to fwpi=0.625fwpu, where fwpu=1035MPa (150 ksi) is the design maximum
strength of the post-tensioning steel. The stress fwpi=0.625fwpu represents the design initial
stress assumed to act in the wall post-tensioning bars just before the application of lateral
loads on the walls.
The post-tensioning bars in the precast concrete walls are unbonded over the entire
wall height, and, are anchored to the walls only at the roof and foundation levels. The
behavior of these walls under lateral loads is governed by the opening of gaps at the
horizontal joints between the wall panels and between the wall and the foundation (see
Fig. 2.15), similar to the opening of gaps at the ends of the coupling beams. More
detailed information on this type of precast concrete wall can be found in Chapter 2.
105
roof
8th floor
8th floor
7th floor
7th floor
6th floor
6th floor
unbonded PT
bar
5th floor
mild steel reinforcement

not shown
h w=32.6m
4th floor
3rd floor
3rd floor
spiral
reinforcement
(18#7@3in.)
1822.2mm@76mm
5th floor
4th floor
2nd floor
lw
roof
PT anchorage
lb
lw
(10#5@18in.)
1015.9mm@457mm
precast concrete
wall panel
2nd floor
spiral
reinforcement
foundation
foundation
(18#7@3in.)
1822.2mm@76mm
h w=32.6m
lw
lb
lw
(121-3/8in. PT bars@8in.)
1235mm PT bars@203mm, fwpi=0.625 f wpu
tw=356mm
(14in.)
l w=3.05m (10ft)
l w=3.05m (10ft)
wsp= 1.94%
tw=356mm
(14in.)
(a)
(b)
Fig. 6.2 Wall elevation and cross-section at base:

(a) CIP-UPT system; (b) PRE-UPT system
TABLE 6.1
CIP-UPT WALL PROPERTIES
CIP-UPT Wall Properties
Wall length, lw
3.05 m (10 ft)
Wall thickness, tw
356 mm (14 in.)
Wall height, hw
32.6 m (107 ft)
Wall main flexural reinforcement
mild steel bars (see Fig. 6.2a)
Diameter of boundary flexural reinforcing bars
22.2 mm (7/8 in.)
Area of each boundary flexural reinforcing bar
387 mm2 (0.6 in2)
Number of boundary flexural reinforcing bars
18
Total area of boundary flexural reinforcement
6968 mm2 (10.8 in2)
Wall base region spiral diameter, Dwsp
296 mm (11.67 in.)
Wall base region spiral wire diameter, dwsp
8.4 mm (0.329 in.)
Wall base region spiral pitch, swsp
38.1 mm (1.5 in.)
Wall base region spiral reinforcement ratio, wsp
1.94%
Wall base region spiral confined length
508 mm
106
wsp = 7.46%
TABLE 6.2
PRE-UPT WALL PROPERTIES
PRE-UPT Wall Properties
Wall length, lw
3.05 m (10 ft)
Wall thickness, tw
356 mm (14 in.)
Wall height, hw
32.6 m (107 ft)
Wall main flexural reinforcement
high strength post-tensioning bars (see Fig. 6.2b)
Diameter of wall post-tensioning bars
35 mm (1.375 in.)
Area of each wall post-tensioning bar
1019 mm2 (1.58 in2)
Number of wall post-tensioning bars in each wall
12
Total area of wall post-tensioning steel in each wall
12232 mm2 (18.96 in2)
Initial stress in wall post-tensioning steel, fwpi
0.625fwpu
Unbonded length of wall post-tensioning bars, lwpu
32.6 m (107 ft)
Wall base region spiral diameter, Dwsp
152 mm (6.0 in.)
Wall base region spiral wire diameter, dwsp
10.7 mm (0.422 in.)
Wall base region spiral pitch, swsp
31.8 mm (1.25 in.)
Wall base region spiral reinforcement ratio, wsp
7.46%
Wall base region spiral confined length
508 mm
It is assumed that adequate concrete confinement is provided in the walls using

spiral reinforcement near the base (see Fig. 6.2). For Wall CIP-UPT, the center-to-center
spiral diameter Dwsp=296 mm (11.67 in.), spiral W8.5 wire diameter dwsp=8.4 mm (0.329
in.), and pitch swsp=38.1 mm (1.5 in.), resulting in a spiral reinforcement ratio,
wsp=1.94% (defined as the ratio of the volume of spiral reinforcement to the volume of
spiral confined concrete core). Similarly, for Wall PRE-UPT, the center-to-center
overlapping spiral diameter Dwsp=152 mm (6.0 in.), spiral W14 wire diameter dwsp=10.7
mm (0.422 in.), and pitch swsp=31.8 mm (1.25 in.), resulting in wsp=7.46%. The larger
amount of concrete confinement used at the base of Wall PRE-UPT is due to the
additional concrete compression stresses from the wall post-tensioning.
6.1.2
Coupling Beams
The coupling beams in Walls CIP-UPT and PRE-UPT are assumed to be identical,
with properties that remain constant over the height of the structures as shown in Fig. 6.3.
The properties of the coupling beams in the multi-story prototype structures are the same
as the prototype subassemblage in Chapter 3, with the following exceptions: (1) the
thickness of the flange cover plates is reduced to tc=28.6mm (1.125 in.); (2) the gage
length of the angle-to-wall connection bolts measured from the heel of the angle to the
center of the innermost angle-to-wall bolts is increased to lgv=114 mm (4.5 in.); and (3)
the area and initial stress of the beam post-tensioning steel are different as described
below. More information on the design properties of the coupling beams in Walls CIPUPT and PRE-UPT is provided below and listed in Tables 6.3-6.5.
As shown in Fig. 6.3, each coupling beam in the multi-story CIP-UPT and PREUPT systems is post-tensioned using four multi-strand tendons, two tendons on each side
of the beam web. Each tendon consists of four 12.7mm (0.5 in.) nominal diameter highstrength seven-wire strands with a total area of abt=395mm2. The tendons are prestressed
to fbpi=0.625fbpu, where fbpu=1862 MPa (270 ksi) is the design maximum strength of the
post-tensioning strands. The stress fbpi=0.625fbpu represents the design initial stress
assumed to act in the beam post-tensioning tendons just before the application of lateral
loads on the walls.
107
coupling beam
W21 182 connection
region
l c =406 mm
PT anchorage
(16 in.)
PT tendon spiral
W21 182
W21 182 cover plate
tc =28.6mm
(1.125in.)
wall
region
dbc
db
hs=3.96m
(13 ft)
embedded
plate te=31.75 mm
l w= 3.05 m (10 ft)
4-strand PT tendon
a bt=395mm2, fbpi = 0.625 fbpu
(0.612 in 2 )
cover A
B
plate
angle
L8x8x1-1/8
l b= 3.05 m (10 ft)
(a)
28.6
(1.125)
38.1 76.2
(1.5) (3)
88.9
(3.5)
lgv=114
(4.5)
ka
k a=44.5(1.75)
ta=28.6
(1.125)
lgh=127(5)
cross-section
38.1
(1.5)
76.2 38.1
(3) (1.5)
A490 25.4(1)bolt
wabw=36.5
(1.4375)
section at B B
60.3
76.2 38.1
(2.375) (3) (1.5)
l a=318(12.5)
A49022.2(7/8)bolt
l g1 =81.8(3.2)
section at A A
l w= 3.05 m (10 ft)
76.2
(3)
36.5
(1.4375)
318
52.4
(12.5)
(2.0625)
88.9
(3.5)
85.7
(3.375)
76.2
(3)
28.6
(1.125)
38.1
(1.5)
elevation
plan
unit: mm (inch)
(b)
Fig. 6.3 Floor level details: (a) elevation; (b) top and seat angle connections
TABLE 6.3
WALLS CIP-UPT AND PRE-UPT COUPLING BEAM PROPERTIES
CIP-UPT and PRE-UPT Coupling Beam Properties
Section
W21182
Beam length, lb
3.05 m (10 ft)
Beam depth, db
577 mm (22.72 in.)
Flange thickness, tbf
37.6 mm (1.48 in.)
Flange width, bbf
318 mm (12.5 in.)
Web thickness, tbw
21 mm (0.83 in.)
Beam cross section area, Ab
34581 mm2 (53.6 in2)
Beam moment of inertia, Ib
19.7 m4 (4730 in4)
28.6mm (1.125 in.)
Flange cover plate width, bc
318 mm (12.5 in.)
Flange cover plate length, lc
406 mm (16 in.)
TABLE 6.4
WALLS CIP-UPT AND PRE-UPT BEAM POST-TENSIONING PROPERTIES
CIP-UPT and PRE-UPT Beam Post-Tensioning Properties
Beam post-tensioning strand diameter, dbp
12.7 mm (0.5 in.)
Number of strands per beam post-tensioning tendon, nbp
4
395 mm2 (0.612 in2)
Number of beam post-tensioning tendons, nbt
4
Total beam post-tensioning steel area, Abp
1579 mm2 (2.448 in2)
Unbonded length of beam post-tensioning steel, lbpu
9.14 m (30 ft)
0.625fbpu
108
TABLE 6.5
WALLS CIP-UPT AND PRE-UPT BEAM-TO-WALL CONNECTION
TOP AND SEAT ANGLE PROPERTIES
CIP-UPT and PRE-UPT Beam-to-Wall Connection Top and Seat Angle Properties
Section
L881-1/8
Angle length, la
318 mm (12.5 in.)
Angle leg length, lal
203 mm (8 in.)
Angle leg thickness, ta
28.6 mm (1.125 in.)
Angle fillet length, ka
44.5 mm (1.75 in.)
Number of angle-to-wall connection bolts, nabw
8
Number of angle-to-beam connection bolts, nabb
8
Angle-to-wall connection bolt diameter, dabw
22.2 mm (7/8 in.)
Angle-to-beam connection bolt diameter, dabb
25.4 mm (1 in.)
Angle-to-wall connection bolt head width across flats, wabw
36.5 mm (1.4375 in.)
Angle-to-wall connection gage length, lgv
114 mm (4.5 in.)
Angle-to-beam connection gage length, lgh
127 mm (5 in.)
6.1.3
Material Properties and Idealizations
This section describes assumed material properties for: (1) the flexural mild steel
reinforcement used in Wall CIP-UPT; (2) the post-tensioning bars used in Wall PREUPT; and (3) the spiral confined concrete at the wall bases. Multi-linear idealizations to
the assumed smooth uniaxial stress-strain relationships are also provided for modeling
purposes.
Some of the design material properties for the prototype structures are listed in
Table 6.6. Note that the material properties for the coupling beams, top and seat angles,
flange cover plates, wall embedded plates, and beam post-tensioning strands are assumed
to be the same as the properties described for the prototype subassemblage in Chapter 3
(see Figs. 3.2a and 3.2b). Similarly, the compressive stress-strain relationships of the
spiral confined concrete in the wall-contact regions and the unconfined concrete are
assumed to be the same as the prototype subassemblage in Figs. 3.2c and 3.2d.
TABLE 6.6
WALL MATERIAL PROPERTIES
Wall CIP-UPT
Wall PRE-UPT
Mild Steel Bars
Post-Tensioning Bars
828MPa
Yield
414 MPa
Yield
strength, fwsy (60 ksi) strength, fwpy (120 ksi)
Yield
Yield
0.00207
0.00414
strain, wsy
strain, wpy
Maximum 672 MPa Maximum 1035 MPa
strength, fwsm (97.5 ksi) strength, fwpu (150 ksi)
Strain at max.
Strain at max.
0.06
0.08
strength, wsm
strength, wpu
Wall Base Region

Spiral Wire
Yield
414MPa
strength, fspy (60 ksi)
Yield
0.00207
strain, spy
Maximum 621 MPa
strength, fspm (90 ksi)
Strain at max.
0.08
strength, spm
Wall Base Region Spiral Confined Concrete

Wall CIP-UPT
Wall PRE-UPT
Linear-elastic 30441 MPa Linear-elastic 30441 MPa
stiffness, Ec (4415 ksi) stiffness, Ec (4415 ksi)
Maximum 63.4 MPa Maximum 100 MPa
strength, fcc (9.2 ksi) strength, fcc (14.5 ksi)
Strain at max.
Strain at max.
0.00728
0.0162
strength, cc
strength, cc
Ultimate
Ultimate
0.0182
0.0385
strain, ccu
strain, ccu
Fig. 6.4a shows the assumed idealized multi-linear stress-strain relationships for the
mild steel reinforcement in Wall CIP-UPT. The design yield strength and maximum
strength of the steel are equal to fwsy=414 MPa (60 ksi) and fwsm=672 MPA (97.5 ksi),
respectively.
109
wall post-tensioning bar stress

(MPa)
reinforcing steel stress

(MPa)
1200
800
(0.06,672)
600
400
(0.02,563)
(0.03,614)
(0.002,414)
200
0
0
0.01
0.03
0.02
0.04
0.05
0.06
(0.0110,873)
(wpu ,fwpu )=
(0.08, 1035)
(wpy,f wpy)=(0.00414, 828)
0.1
0.05
reinforcing steel strain
wall post-tensioning bar strain
(a)
(b)
110
(0.0162,100)
concrete
(0.0385,93.1) crushing
(0.00767,95.2)
Wall PRE-UPT base confined concrete
(0.00400,60.0)
(0.00728,63.4)
concrete
(0.0182,55.9) crushing
Wall CIP-UPT base confined concrete
(0.00165,50.3)
(0.002,41.4)
(0.004,40.7)
(0.00150,38.5)
unconfined concrete
(0.000750,22.8)
(0.00109,33.1)
compression
0
(0.000132,4.00)
(0.00132,0.069)
tension
20
0.005
0.01
0.03
0.02
0.04
0.05
concrete strain
(c)
Fig. 6.4 Wall material properties and idealizations: (a) mild reinforcing steel for CIPUPT system; (b) post-tensioning bar for PRE-UPT system; (c) concrete
Similarly, the assumed smooth (dashed line) and idealized bi-linear (solid line)
stress-strain relationships for the post-tensioning bars used in Wall PRE-UPT are shown
in Fig. 6.4b. The smooth stress-strain relationship is from test results of 25.4 mm (1 in.)
diameter Grade 150 hot-rolled threaded post-tensioning bars reported by Dywidag
Systems International (2002). The yield strength and maximum strength of the posttensioning bars are taken as fwpy=828 MPa (120 ksi) and fwpu=1035 MPa (150 ksi),
respectively. The yield strength fwpy is assumed to be equal to the linear limit stress (i.e.,
the limit of proportionality) of the smooth stress-strain relationship.
110
Since the wall post-tensioning bars are not bonded to the concrete, the peak strains
in the bars during an earthquake are expected to remain small. Thus, the post-yield
stiffness of the idealized stress-strain relationship in Fig. 6.4b was determined from the
portion of the smooth stress-strain relationship between the yield strain, wpy, and a
maximum strain of 0.0110 (with a stress of 873 MPa) expected in the wall posttensioning bars as the structure is displaced laterally. The ratio of the idealized post-yield
stiffness to the initial linear-elastic stiffness Ewp=199955 MPa (29000 ksi) of the posttensioning bars is equal to 0.0328.
Fig. 6.4c shows the smooth (dashed lines) and idealized multi-linear (solid lines)
compressive stress-strain relationships of the spiral confined concrete at the bases of
Walls CIP-UPT and PRE-UPT. The unconfined concrete stress-strain relationship is also
shown for comparison. The maximum compressive strength of the spiral confined
concrete near the base of Wall CIP-UPT is fcc=63.4 MPa (9.2 ksi) reached at a strain of
cc=0.00728, and the ultimate (i.e., crushing) strain is ccu=0.0182. The maximum
compressive strength of the spiral confined concrete near the base of Wall PRE-UPT is
fcc=100 MPa (14.5 ksi) reached at a strain of cc=0.0162, and the ultimate (i.e., crushing)
strain is ccu=0.0385.
The smooth concrete stress-strain relationships in Fig. 6.4c were determined using
a concrete model developed by Mander et al. (1988). The spiral confined concrete stressstrain parameters, which include the maximum strength, fcc, strain at maximum strength,
cc, and ultimate (i.e., crushing) strain, ccu, were calculated based on the unconfined
concrete properties and the amount, distribution, yield strength [assumed as fspy=414 MPa
(60 ksi)], and strain at maximum strength (assumed as spm=0.08) of the spiral
reinforcement. According to the concrete confinement model by Mander et al. (1988), the
ultimate (crushing) strain ccu of the confined concrete is reached when the spiral
reinforcement fractures, resulting in a complete loss of confinement and a sudden loss in
the compression resistance of the concrete. The concrete confinement provided by the
wall transverse reinforcement other than the spirals (e.g., wire mesh) was ignored.
As described in more detail later, the tensile strength of the concrete in Wall PREUPT is ignored in order to model the gap opening behavior along the horizontal joints of
the walls. For Wall CIP-UPT, the tensile strength of the confined and unconfined
concrete is assumed to be equal to fct=0.6
f c' =4 MPa (or fct=7.5
f c' =581 psi) as shown
in Fig. 6.4c.
In general, the slopes of the first segment of the idealized multi-linear stress-strain
relationships in Fig. 6.4 match the linear-elastic stiffness of the smooth relationships.
Furthermore, the maximum compressive strength, strain at maximum strength, and
ultimate strain of the concrete are matched by the idealized stress-strain relationships in
Fig. 6.4c.
111
6.2
Analytical Modeling Assumptions
In addition to the assumptions made for the subassemblage model in Chapter 3, the
following assumptions are made for the multi-story coupled wall analytical modeling.
(1) The objective of this research is to investigate the behavior of isolated coupled
wall structures under earthquake induced lateral loads. The interaction between the
coupled walls and other structural members (e.g., slabs supported by the coupling beams
and the walls) is not within the scope of the analytical model, and thus, is ignored. Note
that the floor and roof slabs may affect the expected and desired behavior of a coupled
wall structure; however, this is not investigated in the report.
(2) The earthquake induced lateral forces are in the plane of the walls.
(3) The entire seismic force in the direction of the coupled walls is resisted by the
coupled walls.
(4) The seismic forces at each floor and roof level are transferred to the coupled
walls from the floor and roof diaphragms by adequate connections between the coupled
walls and the diaphragms. For in-plane forces, the floor and roof diaphragms are assumed
to be rigid.
(5) The total seismic lateral force acting on a coupled wall structure at a floor or
roof level is distributed equally between the left and right walls.
(6) The coupled wall system undergoes in-plane deformations only. Torsional and
out-of-plane deformations are not modeled.
(7) The walls are sufficiently slender such that their nonlinear behavior is governed
by axial-flexural deformations. Shear deformations are assumed to be linear-elastic.
(8) Shear slip at the bases and along the horizontal joints of the walls is prevented
by design.
(9) The foundations are fixed to the ground (i.e., soil-foundation-structure
interaction is not considered).
(10) The mild steel reinforcement in the CIP-UPT system is assumed to be
adequately anchored and fully bonded to the concrete, ignoring any slip due to anchorage
and/or bond failure.
(11) The anchorages for the wall post-tensioning bars in the PRE-UPT system are
properly designed for the maximum post-tensioning forces.
112
6.3
Multi-Story Coupled Wall Models
Analytical models for multi-story unbonded post-tensioned hybrid coupled wall

structures are constructed by joining fiber element models for the coupled wall
subassemblages at the floor and roof levels. The DRAIN-2DX Program (Prakash et al.
1993) is used as the analytical platform. As an example, analytical models for Walls CIPUPT and PRE-UPT, and the corresponding floor level subassemblage models are shown
in Figs. 6.5a and 6.5b, respectively. Note that only selected beam element nodes and wall
PT element nodes (in the PRE-UPT system) are shown in the multi-story wall models to
maintain clarity. Detailed information on the subassemblage model is provided in
Chapter 3.
wall
beam
wall
wall
54
55 27 57
27
26
25
52
24
23
22
49
20
19
46
18
17
16
43
24
50
23
22
wall PT element
18
44
17
16
Y
41
40
11
10
9
87
36
34 35
31 32
30
29
28
3
2
1
left wall region
wallcontact
elements
PT
element
kinematic
constraint
lb
(a)
40
anchor.
node
lw
left wall region
41
39
37
38
36
34 35
54
31 32
30
29
60
28 62
3
2
56 1
right wall region

angle element
beam elements
44
42
33
54
47
45
43
12
39
50
48
46
14
13
53
51
49
15
38
37
52
20
19
45
42
wall
59 54 61
21
47
14
13
12
lw
51
15
11
10
kinem.
constr.
26
25
48
21
wallheight
elements
53
beam
33
58
wall PT element
wallheight
elements
angle element
beam elements
kinem.
constr.
beam PT
kinematic element
constraint
lb
lw
right wall region

wallcontact
elements
anchor.
node
lw
(b)
Fig. 6.5 Multi-story coupled wall analytical models:

As shown in Fig. 6.5 and described in Chapter 3, each concrete wall in the CIPUPT and PRE-UPT systems is modeled using two sets of fiber beam-column elements.
The wall-contact elements, which model the local behavior of the wall-contact regions
to the left and right of the coupling beams, are discussed in detail in Chapter 3. The
113
wall-height elements, which model the axial-flexural and shear behavior of each wall
along its height, are described in this section. The modeling of the wall unbonded posttensioning bars in the PRE-UPT system is also described herein.
Each wall-height element consists of a number of parallel steel and/or concrete
fibers in the direction of the height of the wall. Each fiber has a location in the wall cross
section, a cross-sectional area, and a uniaxial stress-strain relationship. Typically, a larger
number of fiber elements/segments/fibers are used near the base of a wall where the
nonlinear behavior is expected to concentrate as compared with regions away from the
base. The stress-strain relationship of each fiber is a multi-linear idealization of the
smooth uniaxial stress-strain relationship for the mild steel, confined concrete, or
unconfined concrete (e.g., cover concrete) in the walls. Second order effects (often
referred to as P- effects) in the walls are modeled by the wall-height elements.
6.3.1
Cast-in-Place Concrete Walls in CIP-UPT system
The fiber wall-height elements for the prototype CIP-UPT system are described
using Fig. 6.6. A total of 26 elements are used over the height of each wall as shown in
Fig. 6.6a. The fiber cross sections (Figs. 6.6b and 6.6c) of the walls include steel fibers to
model the flexural mild steel reinforcing bars (with the multi-linear steel stress-strain
relationship in Fig. 6.4a) and concrete fibers to model the unconfined and spiral confined
concrete (with the multi-linear concrete stress-stain relationships in Fig. 6.4c).
In Figs. 6.6b and 6.6c, CU refers to unconfined concrete fibers, CC refers to
spiral confined concrete fibers, and S refers to steel fibers. In the boundary regions near
the base of each wall (Fig. 6.6b), spiral confined concrete as well as unconfined concrete
and steel fibers are used. Away from the wall base (Fig. 6.6c), only unconfined concrete
and steel fibers are used. The actual number of fibers, thickness of each concrete fiber, tf,
and area of each steel fiber, af, used in the wall-height elements for the CIP-UPT system
are given in Table 6.7. Note that refinement studies using models with larger numbers of
elements/segments/fibers than the discretization in Table 6.7 showed no appreciable
differences in the analysis results.
The cyclic behavior of the concrete fibers modeling the spiral confined concrete
(CC fiber type) at the base of the CIP-UPT system is shown in Fig. 6.7a. The cyclic
behavior of the unconfined concrete (CU) fibers has similar characteristics to the
hysteretic model for the confined concrete fibers, as shown in Fig. 6.7b. More details on
the cyclic stress-strain characteristics of the concrete fiber in DRAIN-2DX can be found
in Kurama et al. (1996) and Kurama (1997).
Fig. 6.7c shows the cyclic (solid lines) and monotonic (dashed lines) behavior of
the steel fibers used to model the mild steel reinforcement (S fiber type) in the CIP-UPT
system. Note that in Fig. 6.7c, a lower yield strength is used for the steel fibers under
cyclic loading than under monotonic loading in order to model the Bauschinger effect in
the reinforcing steel under cyclic loading (see also Fig. 6.7d for a comparison between
114
the monotonic steel stress-strain relationship and the envelope relationship under cyclic
loading). The use of a lower yield strength for the reinforcing steel results in an earlier
reduction in the stiffness (i.e., softening) of the walls under lateral loading. In order to
compensate for this reduction, modifications to the unconfined and confined concrete
stress-strain behavior in tension are also made, as shown in Fig. 6.7e, such that the
monotonic lateral load behavior of the coupled wall structure using the cyclic and
monotonic reinforcing steel fiber models are similar.
wall
27
26
25
24
23
22
21
CU
20
19
CU
CC
18
17
slice 2
tw
16
15
14
13
boundary region
slice 1
(b)
boundary region
CU
12
11
10
9
8
7
lw
6
5
4
3
2
Node1
slice 2
slice 1
(c)
(a)
Fig. 6.6 CIP-UPT wall-height elements: (a) elevation; (b) fiber slice schematic near wall
base; (c) fiber slice schematic away from wall base
115
TABLE 6.7
CIP-UPT SYSTEM WALL-HEIGHT ELEMENT DISCRETIZATION
Element
Nodes
1-2
2-3
3-4
4-5

Fiber Thickness, tf, and Area, af Length of Total Number of
mm (in.) and mm2 (in2)
CONCRETE FIBERS
1 CU fiber; tf =31.75(1.25)
1 CU fiber; tf =19(0.75)
1 CC fiber; tf =19(0.75)
5 CU fibers; tf =25.4(1)
5 CC fibers; tf =25.4(1)
292 mm (11.5 in.)
1 (Slice 1)
3048 mm
762 mm (30 in.)
1
4 CC fibers; tf =50.8(2)
83
(see Fig. 6.6b)
(120 in.)
2756 mm (125.5 in.)
5 CU fibers; tf =127(5)
721 mm (12.5 in.)
5 CC fibers; tf =127(5)
Element
Length
STEEL FIBERS
6 S fibers; af =1161 (1.8)
3 S fibers; af =400 (0.62)
5-6
6-7
7-8
292 mm (12.5 in.)

3378 mm (131 in.)
292 mm (12.5 in.)
8-9
9-10
10-11
292 mm (12.5 in.)

3378 mm (131 in.)
292 mm (12.5 in.)
11-12
12-13
13-14
292 mm (12.5 in.)

3378 mm (131 in.)
292 mm (12.5 in.)
14-15
15-16
16-17
292 mm (12.5 in.)

3378 mm (131 in.)
292 mm (12.5 in.)
17-18
18-19
19-20
292 mm (12.5 in.)

3378 mm (131 in.)
292 mm (12.5 in.)
20-21
21-22
22-23
292 mm (12.5 in.)

3378 mm (131 in.)
292 mm (12.5 in.)
23-24
24-25
25-26
26-27
292 mm (12.5 in.)

3378 mm (131 in.)
292 mm (12.5 in.)
345 mm (12.5 in.)
1 (Slice 2)
(see Fig. 6.6c)
CONCRETE FIBERS
1 CU fiber; tf =31.75(1.25)
1 CU fiber; tf =19(0.75)
3048 mm
(120 in.)
53
STEEL FIBERS
6 S fibers; af =1161 (1.8)
3 S fibers; af =400 (0.62)
Notes:
(1) Only fiber discretization for left half of cross section is given for each slice, listed in order from extreme
fiber to innermost fiber. Fiber discretization for right half of each cross section is symmetric about midlength of section.
(2) Modeling of left and right walls is same.
(3) Each steel fiber represents all of bars located at same distance from longitudinal reference axis of
element (at mid-length of cross section). Innermost bar layer is located at reference axis (see Fig. 6.6).
116
tension
(-0.00109, -33.1)
-40
(-0.0182, -55.8)
(-0.00728, -63.4)
(-0.00400, -60.0)
-0.015
-0.005
-80
-0.02
tension
-0.01
unconfined concrete
stress (MPa)
confined concrete
stress (MPa)
compression
compression
-45
-0.006
-0.004
(-0.002, -41.4)
-0.002
strain
confined concrete (CC) fiber type
strain
unconfined concrete
(a)
(b)
0.002
700
for monotonic
loading
stress (MPa)
stress (MPa)
800
400
0
for cyclic loading with

Bauschinger effect
-400
-800
-0.06
cyclic
monotonic
-0.04
-0.02
0.02
0.04
0.06
strain
0.03
0.06
(c)
(d)
concrete tension stress (MPa)
mild steel reinforcement (S) fiber type
strain
mild steel reinforcement (S) fiber type
4.5
(0.000132, 4.00)
for cyclic loading with

Bauschinger effect
for monotonic
loading
(0.00132, 0.069)
0
0
(0.00461, 0.069)
0.0025
0.005
concrete tension strain
(e)
Fig. 6.7 CIP-UPT system cyclic material models: (a) confined concrete (CC) fiber;
(b) unconfined concrete (CU) fiber; (c) mild steel reinforcement (S) fiber;
(d) steel fiber stress-strain relationship under monotonic loading and envelope
relationship under cyclic loading; (e) concrete in tension
117
6.3.2
Precast Concrete Walls in PRE-UPT system
The modeling of the precast concrete walls in the PRE-UPT system is different
than the modeling of the cast-in-place concrete walls in the CIP-UPT system because of
the use of unbonded post-tensioning bars in the precast walls and the opening of gaps
along the horizontal joints of the walls. This is described in detail in Kurama et al. (1999)
and Kurama (2002).
The fiber wall-height elements for the PRE-UPT system are described using Fig.
6.8. Similar to the CIP-UPT system, a total of 26 elements are used over the height of
each wall as shown in Fig. 6.8a. The fiber cross sections (Figs. 6.8b and 6.8c) of the
precast concrete wall panels include concrete fibers to model the unconfined and spiral
confined concrete (with the multi-linear concrete stress-strain relationships in Fig. 6.4c).
The actual number of fibers and thickness of each fiber, tf used in the wall-height
elements for the PRE-UPT system are given in Table 6.8. Refinement studies using
models with larger numbers of elements/segments/fibers than the discretization in Table
6.8 showed no appreciable differences in the analysis results.
Note that the fiber elements modeling the wall panels in the PRE-UPT system do
not contain any steel fibers for two reasons: (1) the post-tensioning bars are unbonded
from the wall panels, and thus, are modeled separately using truss elements; and (2) the
bonded wall panel steel reinforcement (e.g., wire mesh) is not continuous across the
horizontal joints, and thus, does not contribute to the flexural resistance of the wall. The
modeling of the post-tensioning bars and the modeling of gap opening along the
horizontal joints of the walls are described below.
Modeling of Gap Opening
Gap opening at the horizontal joints, especially between the base panel and the
foundation, is one of the most important characteristics governing the behavior of
unbonded post-tensioned precast concrete walls. The modeling of gap opening at the
horizontal joints is similar to the modeling of gap opening at the joints between the
coupling beams and the walls as described in Chapter 3. As a result of the opening of
gaps and due to post-tensioning, large compressive stresses develop near the regions of a
precast wall panel in contact with another panel or with the foundation, while the tensile
stresses in a significant portion of the panel are equal or close to zero (Allen and Kurama
2002). The compression behavior in the contact regions is modeled using the nonlinear
uniaxial compressive stress-strain relationships of the concrete fibers. To model the gap
opening behavior, the tensile strength and stiffness of the fibers representing the wall
panels are set to zero as described in Kurama et al. (1996, 1999) and verified in Kurama
(2000).
In Figs. 6.8b and 6.8c, CUC refers to compression-only unconfined concrete
fibers and CCC refers to compression-only confined concrete fibers. Fig. 6.9a shows
the cyclic behavior of the compression-only concrete fibers modeling the spiral confined
concrete (CCC fiber type) at the base of the PRE-UPT system. The cyclic behavior of the
118
compression-only unconfined concrete (CUC) fibers has similar characteristics to the

hysteretic model for the confined concrete fibers.
Through this model, the gap opening displacements that occur at the horizontal
joints of the PRE-UPT system are represented as distributed tensile deformations that
occur in the fiber elements over the height of the wall panels. The reduction in the
flexural stiffness of the structure as a result of gap opening (Kurama et al. 1999a, 1999b)
is modeled by the zero stiffness of the concrete fibers that go into tension when the
precompression stresses due to the gravity and post-tensioning forces are overcome by
the flexural stresses that develop at the tension sides of the left and right walls due to the
lateral loads (i.e., when decompression occurs).
wall
27
26
25
CUC
CUC
CCC
24
23
22
21
20
19
slice 1
boundary region
18
17
slice 2
boundary region
(b)
16
15
14
13
UCC
12
11
10
9
8
7
lw
slice 2
6
5
4
3
2
1
slice 1
(c)
Fig. 6.8 PRE-UPT wall-height elements: (a) elevation; (b) fiber slice schematic near wall
base; (c) fiber slice schematic away from wall base
119
TABLE 6.8
PRE-UPT SYSTEM WALL-HEIGHT ELEMENT DISCRETIZATION
Element
Nodes
Element
Length
1-2
2-3
3-4
4-5
292 mm (11.5 in.)

762 mm (30 in.)
2756 mm (108.5 in.)
721 mm (28.4 in.)
5-6
6-7
7-8
292 mm (11.5 in.)

3378 mm (133 in.)
292 mm (11.5 in.)
8-9
9-10
10-11
292 mm (11.5 in.)

3378 mm (133 in.)
292 mm (11.5 in.)
11-12
12-13
13-14
292 mm (11.5 in.)

3378 mm (133 in.)
292 mm (11.5 in.)
14-15
15-16
16-17
292 mm (11.5 in.)

3378 mm (133 in.)
292 mm (11.5 in.)
17-18
18-19
19-20
292 mm (11.5 in.)

3378 mm (133 in.)
292 mm (11.5 in.)
20-21
21-22
22-23
292 mm (11.5 in.)

3378 mm (133 in.)
292 mm (11.5 in.)
23-24
24-25
25-26
26-27
292 mm (11.5 in.)

3378 mm (133 in.)
292 mm (11.5 in.)
345 mm (13.6 in.)

1 (Slice 1)
(see Fig. 6.8b)
1 (slice 2)
(see Fig. 6.8c)
Fiber Thickness, tf
mm (in.)
1 CUC fiber; tf=31.75(1.25)
1 CUC fiber; tf=19(0.75)
1 CCC fiber; tf=19(0.75)
5 CUC fibers; tf=127(5)
5 CCC fibers; tf=127(5)
1 CUC fiber; tf=31.75(1.25)

1 CUC fiber; tf=19(0.75)
Length of Total Number of

3048 mm
(120 in.)
66
3048 mm
(120 in.)
36
Notes:
(1) Only fiber discretization for left half of cross section is given for each slice, listed in order from extreme
fiber to innermost fiber. Fiber discretization for right half of each cross section is symmetric about midlength of section.
(2) Modeling of left and right walls is the same.
The process of gap opening/closing under the action of lateral loading/unloading

causes softening/re-stiffening at the horizontal joints. This process is captured in the fiber
elements by having an increasing number of fibers subjected to tension during loading,
and then by having the fibers subjected to tension go back into compression during
unloading.
120
1100
20
TENSION
-0.04
(0.0110,873)
stress (MPa)
(-0.00165, -50.3)
(-0.0162, -100)
(-0.0385, -93.1)
-110
-0.06
(0.00414,828)
(-0.06, -0.069)
stress (MPa)
(-0.00767, -95.2)
-0.02
0.02
strain
0.01
0.02
strain
compression only confined concrete (CCC) fiber type
(a)
wall PT element
(b)
Fig. 6.9 PRE-UPT system cyclic material models:

(a) compression-only confined concrete (CCC) fiber; (b) wall PT element
Modeling of Wall Post-Tensioning Bars and Anchorages
The wall unbonded post tensioning bars in the PRE-UPT system are modeled using
truss elements, referred to as wall PT elements, running over the entire wall height as
shown in Fig. 6.5b. Note that only selected wall PT elements are shown in Fig. 6.5b to
maintain clarity. The cyclic stress-strain behavior of the truss elements, based on the
bilinear relationship in Fig. 6.4b, is shown in Fig. 6.9b. The post tensioning loads are
simulated by tensile forces in the truss elements, which are equilibrated by compression
forces in the wall-height elements. Note that the compression forces that develop in the
wall-height elements result in an elastic shortening and subsequent loss in the forces in
the truss elements modeling the post-tensioning bars. Thus, larger tensile forces are
applied to the truss elements in order to achieve the desired amount of initial force in the
post-tensioning bars after elastic shortening.
Each wall PT element is connected to two nodes representing the anchorages
between the post-tensioning bar and the walls (e.g., Nodes 55-56 and 57-58 in Fig. 6.5b).
It is assumed that the bars are placed inside oversized ducts in the wall panels, and thus,
remain straight during the lateral displacements of the structure. At the bases of the walls,
the wall PT element nodes (Nodes 56, 58, 60 and 62 in Fig. 6.5b) are assumed fixed to
the foundation. At the top of the walls (i.e., roof), the displacements of the PT element
nodes (Nodes 55, 57, 59 and 61 in Fig. 6.5b) are kinematically constrained to the
corresponding wall-height element node (Node 27) to model the anchorages between the
post-tensioning bars and the walls (see Fig. 6.5b). Second order effects (often referred to
as P- effects) in the post-tensioning bars (due to the relative displacements of the bar
nodes in a direction perpendicular to the bar) are modeled in the wall PT elements.
121
6.4
Verification of Wall Models
Test results of multi-story hybrid coupled wall systems are not available. Thus, the
model for the coupled wall subassemblages at the floor levels and the models for the
concrete walls are verified separately. The verification of the subassemblage model based
on comparisons with ABAQUS results is discussed in Chapter 3, and the verification of
the precast concrete wall model can be found in Kurama et al. (1999) and Kurama (2000).
These results show that the fiber-element analytical models are capable of adequately
predicting both the global (e.g., load-deformation) behavior and the local (e.g., gap
opening and contact) behavior of the coupled wall subassemblages and the precast
concrete walls. Further verification and improvement of the subassemblage model based
on experimental results can be found in Chapters 9 and 10.
The verification of the cast-in-place wall-height element model is done based on
previous experiments conducted by Thomsen (1995). Fig. 6.10a shows the elevation and
cross section of the wall test specimen used in the verification, which is referred to as
Wall RW2 with full details provided in Thomsen (1995). A reversed cyclic lateral force
and a constant vertical force were applied at the top of the wall. The behavior of the wall
was dominated by nonlinear axial-flexural deformations rather than shear deformations as
described by Thomsen (1995). Fig. 6.10b shows the lateral force versus top displacement
relationship of the wall measured during the test and Fig. 6.10c shows the predicted
behavior using a fiber element analytical model similar to the model in Fig 6.6. A
reasonable comparison between the measured and predicted results is observed. Thus, it
is concluded that the analytical model is capable of providing an adequate representation
of the nonlinear axial-flexural behavior of monolithic cast-in-place reinforced concrete
walls under cyclic lateral loads.
6.5
Advantages and Limitations of Analytical Models
A significant advantage of the fiber element models used in this report is that a
reasonably accurate representation of multi-story hybrid coupled wall systems can be
developed using only the geometry and dimensions of the structure, and uniaxial stressstrain models for concrete and steel. The model provides an effective tool to conduct
static and dynamic analyses of multi-story structures under earthquake-induced loads,
accounting for axial-flexural interaction in the wall and beam members, behavior of the
top and seat angles, gap opening at the beam and wall joints, and hysteretic behavior of
the steel and concrete.
The main limitations of the multi-story coupled wall models presented in this
chapter are as follows:
(1) The fiber element models account for the linear and nonlinear axial-flexural
deformations (including gap opening) and linear shear deformations of the walls;
however, nonlinear shear deformations are not modeled. The nonlinear shear
deformations of the walls may be significant, depending on the wall height-to-length
122
aspect ratio and the amount of coupling. According to Paulay and Priestley (1992), shear
deformations in walls with aspect ratios smaller than 4.0 may need to be considered in
seismic analysis and design. The aspect ratios of the walls investigated in this report vary
between 5.35 and 14.3. Thus, nonlinear shear deformations are not expected to play an
important role in the behavior of the walls.
(b)
lateral load, kips (kN)
-2.8
40
(178)
2.8
-2
0
2
top displacement, inches (mm)
4
(102)
20
0
-20
(-178)
-40
-4
(-102)
(a)
lateral drift, percent

-1.4
0
1.4
(c)
Fig. 6.10 Monolithic cast-in-place reinforced concrete wall-height element model

verification: (a) Wall RW2 (from Thomsen 1995);
(b) measured behavior; (c) predicted behavior
(2) The degradation (if any) in the flexural stiffness and resistance of the walls due
to increasing lateral displacements is modeled; however, any additional degradation
under repeated displacement cycles to a constant amplitude is not captured.
(3) Buckling and low cycle fatigue fracture of the mild steel reinforcement are not
modeled.
(4) For systems with precast concrete walls, the desired behavior of the walls is
governed, primarily, by the opening of gaps at the horizontal joints between the wall
panels and between the base panel and the foundation. Shear slip at the horizontal joints
should be prevented by design, and thus, is not modeled.
123
(5) The precast concrete wall model assumes that the post-tensioning bars remain
straight during the lateral displacements of the structure. Kinking of the bars at the
horizontal joints can occur if the bars come into contact with the ducts inside the wall
panels; however, this behavior is not desired and should be prevented by using
sufficiently oversized ducts.
6.6
Modeling of Walls with Embedded Coupling Beams
The modeling of multi-story hybrid coupled wall systems with embedded steel
coupling beams is similar to the modeling of the CIP-UPT system described above,
except for the use of embedded steel coupling beam subassemblage models at the floor
and roof levels as described in Chapter 3.
6.7
Chapter Summary
This chapter describes fiber element analytical models to investigate the nonlinear
behavior of multi-story hybrid coupled wall systems. The details of two prototype
unbonded post-tensioned hybrid coupled wall structures are also presented. The
analytical models are constructed by joining subassemblage models at the floor and roof
levels. Systems with monolithic cast-in-place reinforced concrete walls as well as with
precast concrete walls are considered. Verification of the cast-in-place reinforced
concrete wall model is presented based on comparisons with previous experiments of
isolated (i.e., uncoupled) wall specimens.
124
CHAPTER 7
NONLINEAR BEHAVIOR OF MULTI-STORY
COUPLED WALLS
This chapter describes an analytical investigation on the nonlinear behavior of
multi-story unbonded post-tensioned hybrid coupled wall systems under lateral loading as
follows: (1) overview of analyses; (2) behavior under monotonic loading; (3) behavior
under cyclic loading; and (4) parametric investigation. The behavior of unbonded posttensioned hybrid coupled wall structures is compared with the behavior of structures that
use embedded steel coupling beams and structures without coupling.
7.1
Overview of Analyses
The two coupled wall systems (Walls CIP-UPT and PRE-UPT) and the fiber
element models described in Chapter 6 are used as the basis for the multi-story coupled
wall analyses in this chapter. The analyses are conducted by applying loads on the
structures in the following order: (1) beam and wall pier (for the PRE-UPT system) posttensioning forces; (2) gravity loads applied at the floor and roof levels; and (3) lateral
forces applied at the floor and roof levels in displacement control.
The beam and wall PT element forces are applied all at once. Then, the gravity
loads, G are applied as 100% of the unfactored design dead load (D) plus 25% of the
unreduced unfactored design live load (L) to represent the amount of gravity loads that
may be acting on the structure during an earthquake. As shown in Fig. 7.1, the gravity
loads are applied at the wall-height element nodes at the floor and roof levels, with no
loads applied along the length of the coupling beams. Table 7.1 provides a summary of
the gravity loads applied at the floor and roof levels of the left and right wall piers in the
CIP-UPT and PRE-UPT systems. Note that the gravity loads for these two systems are
assumed to be the same in order to facilitate comparisons between the behaviors of the
two structures.
The displacement controlled nonlinear lateral load analyses are conducted based on
the lateral displacement of Node 54 (as shown in Fig. 7.1) with respect to the wall base.
An inverse triangular distribution of lateral loads, F (applied from left to right) over the
height of the walls is used. For each coupled wall system, the lateral loads are assumed to
be equally divided between the left and right walls and are applied at the same nodes as
the gravity loads as shown in Fig. 7.1. Note that inertia force distributions significantly
different than the assumed inverse triangular lateral load distribution may occur during an
earthquake (e.g., Ghosh and Markevicius 1990; Paulay and Priestley 1992; Eberhard and
125
Sozen 1993; Kabeyasawa 1993; Aoyama 1993; Otani et al. 1994; Kurama et al. 1997,
1999b); however, this is not investigated in this chapter.
Gr
fr
roof
27
G8
f8
f7
54
8th floor
50
G 7 7th floor
G7
20
47
G6
6th floor
17
G5
G4
f4
G6
44
5th floor
14
f3
G8
23
f6
f5
Gr
G5
41
4th floor
G4
11
38
G 3 3rd floor
G3
35
G 2 2nd floor
G2
f2
lw
hw
32
lb
lw
Fig. 7.1 Gravity and lateral forces for multi-story coupled wall analyses
TABLE 7.1
WALL PIER GRAVITY LOADS FOR CIP-UPT AND PRE-UPT SYSTEMS
Level
Roof
8th floor
7th floor
6th floor
5th floor
4th floor
3rd floor
2nd floor
Gravity Loads, G
1.0D (kN)
0.25L (kN)
207
3.8
258
15.6
258
15.6
258
15.6
258
15.6
258
15.6
258
15.6
332
15.6
126
7.2
Behavior Under Monotonic Loading
This section describes the behavior of the prototype CIP-UPT and PRE-UPT
coupled wall systems under combined gravity loads and monotonic lateral loads as
follows: (1) base shear force versus roof drift behaviors; (2) wall pier deflected shapes; (3)
wall pier base flexural steel strains and stresses; (4) wall pier base axial forces; (5) wall
pier base moments; (6) degree of coupling; (7) wall pier base concrete strains; (8)
coupling beam shear force versus chord rotation behaviors; (9) coupling beam axial
forces; (10) coupling beam end strains; and (11) tension angle force versus deformation
behaviors.
7.2.1
Coupled Wall Base Shear Force versus Roof Drift Behaviors
As examples of representative behavior, the thick solid lines in Figs. 7.2a and 7.2b
show the coupled wall base shear force versus roof drift (F-) behaviors of the CIP-UPT
and PRE-UPT systems, respectively. The roof drift is equal to the average lateral
displacement of the left and right side walls at the roof (i.e., at Nodes 27 and 54 in Fig.
7.1) divided by the wall height to the base, hw.
The prototype CIP-UPT and PRE-UPT systems were designed such that they have
similar F- relationships under monotonic lateral loads as shown in Figs. 7.2a and 7.2b.
This was done to facilitate comparisons between the behaviors of the two structures. As
the prototype systems displace laterally, they go through a number of states including:
(1) Tension-side wall (left-side wall for lateral loads applied from left to right)
softening ( markers in Fig. 7.2) This state is defined as the state when the neutral axis
at the base of the tension-side wall reaches the centerline of the wall.
(2) Compression-side wall (right-side wall for lateral loads applied from left to
right) softening (+ markers) This state is defined as the state when the neutral axis at
the base of the compression-side wall reaches the centerline of the wall.
(3) First wall pier steel yielding ( markers) This state is defined as the first
yielding of the wall pier mild flexural reinforcing bars for the CIP-UPT system and the
wall pier post-tensioning bars for the PRE-UPT system. Note that the linear limit point on
the stress-strain curve is used to identify the yielding of the post-tensioning bars (see Fig.
6.4c).
(4) Coupling beam softening ( markers) This state is defined as the state when
the average coupling beam shear force in the multi-story structure (i.e., the sum of the
shear forces in the coupling beams divided by the number of beams) is equal to
1.05Vb,sof, where Vb,sof is the subassemblage softening state coupling shear force (see
Chapters 4 and 5) and the factor 1.05 is described in Chapter 8.
127
wall base shear forces F, Ftw , Fcw and Fw,unc (kN)
4000
F, CIP-UPT system
Fcw , compression-side wall in coupled system

tension side wall softening
comp. side wall softening
first wall steel yielding
coupling beam softening
first beam PT yielding
comp. side wall conc. crush
2Fw,unc , two uncoupled

wall piers
Ftw , tension-side wall

in coupled system
wall roof drifts, and w,unc (percent)
(a)
wall base shear forces F, Ftw , Fcw and Fw,unc (kN)
4000
F, PRE-UPT system
Fcw , compression-side
wall in coupled system

ten. side wall conc. linear limit
comp. side wall conc. crush
2Fw,unc , two uncoupled wall piers

Ftw , tension-side wall in coupled system
2
1
(b)
Fig. 7.2 Base shear force versus roof drift behaviors under monotonic lateral loads:
128
(5) Tension-side wall concrete linear limit ( markers) This state is defined as the
state when the maximum confined concrete compression stress at the base of the tensionside wall reaches the linear limit (i.e., limit of proportionality) stress of the confined
concrete, fcl.
(6) First beam PT yielding (* markers) This state is defined as the state when first
yielding of the beam post-tensioning tendons at the floor and roof levels occurs (see also
Chapter 4).
(7) Compression-side wall concrete crushing ( markers) This state is defined as
the state when the extreme strain in the spiral confined concrete at the base of the
compression-side wall reaches the crushing strain, ccu (see Fig. 6.4c).
For each prototype system, the thin solid lines in Figs. 7.2a and 7.2b show the shear
forces Ftw and Fcw at the bases of the tension-side and compression-side walls. Similarly,
the dashed lines show the total base shear force versus roof drift relationship (2Fw,unc,
w,unc) of two wall piers without coupling (i.e., two wall piers with no coupling beams).
Note that the base shear force versus roof drift relationships of the tension-side and
compression-side walls do not pass through the origin since shear forces develop in the
walls upon the application of the coupling beam post-tensioning forces, with no external
lateral loads. The sum of the tension-side and compression-side wall base shear forces,
Ftw and Fcw is approximately equal to the total coupled wall base shear force, F. In both
structures, most of the coupled wall base shear force is resisted by the compression-side
wall as a result of the larger compressive axial force that develops as compared with the
axial force in the tension-side wall.
7.2.2
Wall Pier Deflected Shapes
The solid and dashed lines in Figs. 7.3a and 7.3b show the deflected shapes of the
tension-side and compression-side walls in the CIP-UPT and PRE-UPT systems,
respectively, at coupled wall roof drift values of =0.5, 1, 1.5, 2, and 2.5%. The results
demonstrate that the lateral displacements and rotations of the tension-side and
compression-side walls are similar.
7.2.3
Wall Pier Base Flexural Steel Strains and Stresses
Fig. 7.4a shows the stresses of the mild reinforcement at four different locations in
the tension-side and compression-side walls of the CIP-UPT system, respectively, versus
the coupled wall roof drift, . The results indicate that more layers of mild steel
reinforcement in the tension-side wall is in tension as compared with the reinforcement in
the compression-side wall.
129
35
30
1.5%
roof drift,
=0.5%
2.0%
1.0%
wall height, hw (m)
25
2.5%
20
15
10
5
0
0
CIP-UPT system
tension-side wall
compression-side wall
100
200
300
400
500
600
700
800
900
wall floor/roof deflection (mm)
(a)
35
30
1.5%
roof drift,
=0.5%
2.0%
1.0%
wall height, hw (m)
25
2.5%
20
15
10
5
0
0
PRE-UPT system
tension-side wall
100
200
300
400
500
600
wall floor/roof deflection (mm)
700
800
900
(b)
Fig. 7.3 Wall deflected shapes: (a) CIP-UPT system; (b) PRE-UPT system
130
tension
first wall steel

600 yielding
wall pier mild steel bar stresses at base (MPa)
800
1
2
3
400
3
200
200
CIP-UPT system
tension-side wall
400
600
0
0.5
1
1.5
2
roof drift, (percent)
2.5
200
CIP-UPT system
200
400
4
600
0
0.5
1
1.5
2
2.5
tension
wall pier PT bar stresses
fwp1 through fwp6 (MPa)
wall pier PT bar stresses

fwp1 through fwp6 (MPa)
400
800
fwpi
fwp1 through fwp6
400
PRE-UPT system
tension-side wall
200
1000
tension
tension
600
(a)
1000
600
800
0.5
1
1.5
2
2.5
800
600
fwp1 through fwp6
400
PRE-UPT system
200
0
fwpi
0.5
1
1.5
2
2.5
(b)
Fig. 7.4 Wall pier flexural steel stresses: (a) CIP-UPT system; (b) PRE-UPT system
Similarly, Fig. 7.4b shows the stresses, fwp1 through fwp6 in the six pairs of posttensioning bars (see Fig. 6.2b) in the tension-side and compression-side walls of the PREUPT system, respectively, versus the coupled wall roof drift, . The results indicate that
the stresses in the post-tensioning bars of the tension-side wall are larger than the stresses
in the post-tensioning bars of the compression-side wall. The stresses in the posttensioning bars located in the compression (i.e., contact) region of the compression-side
wall remain close to the initial stress, fwpi.
7.2.4
Wall Pier Base Axial Forces
The solid and dashed lines in Figs. 7.5a and 7.5b show the axial forces Ntw and Ncw
at the bases of the tension-side and compression-side walls in the CIP-UPT and PREUPT systems, respectively, versus the coupled wall roof drift, . A positive wall pier
axial force represents a compressive force.
131
wall pier base axial forces Ntw and Ncw (kN)
12000
10000
compression-side wall axial force, Ncw
8000
6000
4000
CIP-UPT system
2000
0
-2000
tension-side wall axial force, Ntw
-4000
-6000
0.5
1.5
2.5
(a)
wall pier base axial forces Ntw and Ncw (kN)
20000
18000
compression-side wall axial force, Ncw
16000
14000
12000
PRE-UPT system
10000
8000
6000
4000
tension-side wall axial force, Ntw
2000
0
0
0.5
1.5
2.5
(b)
Fig. 7.5 Wall pier base axial forces: (a) CIP-UPT system; (b) PRE-UPT system
132
Figs. 7.6a and 7.6b show the contributions of the wall pier gravity loads, Ntwg and
Ncwg, wall pier post-tensioning forces (for the PRE-UPT system only), Ntwp and Ncwp, and
coupling beam shear forces, Ntwb and Ncwb, on the wall pier base axial forces, Ntw and Ncw
in the CIP-UPT and PRE-UPT systems, respectively. Note that the coupling beam shear
forces result in tensile axial forces in the tension-side walls and compressive axial forces
in the compression-side walls.
The results in Figs. 7.5 and 7.6 show that the post-tensioning of the precast
concrete walls in the PRE-UPT system results in significantly larger compressive axial
forces than the forces in the cast-in-place concrete walls of the CIP-UPT system. As the
structures are displaced laterally, the axial forces in the compression-side walls increase
and the axial forces in the tension-side walls decrease due to the coupling beam shear
forces. While tensile forces develop at the base of the tension-side wall in the CIP-UPT
system, the wall pier axial forces in the PRE-UPT system remain compressive throughout
the analysis as a result of the wall post-tensioning forces.
12000
12000
8000
6000
4000
N twg
2000
0
-2000
N tw
-4000
-6000
-8000
0.5
1
1.5
2
8000
2.5
4000
N cwg
2000
0
-2000
-4000
-8000
0
N cwb
6000
-6000
N twb
0
N cw
10000
CIP-UPT system
tension-side wall
wall pier base axial force

components Ncwg and Ncwb (kN)

components Ntwg and Ntwb (kN)
10000
CIP-UPT system
0.5
1
1.5
2
2.5
2.5
(a)
20000
PRE-UPT system
tension-side wall
15000
10000
N twp
5000
N tw
N twg
0
-5000
-10000

components Ncwg , Ncwp , and Ncwb (kN)

components Ntwg , Ntwp , and Ntwb (kN)
20000
N twb
0
0.5
1
1.5
2
2.5
15000
N cw
10000
N cwp
5000
N cwb
N cwg
-10000
PRE-UPT system
-5000
0.5
1
1.5
2
(b)
Fig. 7.6 Components of wall pier base axial forces: (a) CIP-UPT system;
(b) PRE-UPT system
133
7.2.5
Wall Pier Base Moments
Figs. 7.7a and 7.7b show the contributions of the tension-side wall base moment,
Mtw, the compression-side wall base moment, Mcw and the base moment due to the
coupling force couple, Mwb to the total coupled wall base moment resistance Mw of the
CIP-UPT and PRE-UPT systems, respectively, versus the coupled wall roof drift, . For
comparison the sum of the bending moments, Mw,unc at the bases of the wall piers without
coupling (i.e., two wall piers with no coupling beams) are also shown in Fig. 7.7a and
7.7b.
The results in Fig. 7.7 show that most of the coupled wall base moment comes from
the coupling force couple, indicating a large degree of coupling in both structures.
Similar to the base shear forces in Fig. 7.2, the compression-side wall has a larger
contribution to the coupled wall base moment than the contribution of the tension-side
wall, as a result of the larger compressive axial forces that develop in the compressionside wall. The base moment versus roof drift relationships for the tension-side and
compression-side walls do not pass through the origin since bending moments develop in
the walls upon the application of the coupling beam post-tensioning forces, with no
external lateral loads.
7.2.6
Degree of Coupling
As shown in Fig. 7.2, coupling of the left and right wall piers results in a significant
increase in the lateral strength and initial stiffness of the system as a whole. This increase
can be quantified using the degree of coupling, DOC, from Equation (2.2).
Fig. 7.8 shows the degree of coupling [from Equation (2.2)] for the CIP-UPT and
PRE-UPT systems (solid and dashed lines, respectively), plotted against the roof drift, .
The coupling degree for both systems is around DOC=60% and remains relatively
constant as the structures are displaced laterally. Based on recent research by El-Tawil
and Kuenzli (2002), the 60% coupling for the prototype walls may be large and can lead
to tension forces in the tension-side wall (as shown in Fig. 7.5a for the CIP-UPT system),
large compressive and shear forces in the compression-side wall (as shown in Figs. 7.2
and 7.5), and large shear forces at the beam-to-wall interfaces. The design of unbonded
post-tensioned hybrid coupled wall structures to achieve a desired target level of coupling
is discussed in Chapter 12.
134
wall base moments Mw, Mtw , Mcw , Mwb , and 2Mw,unc(kN.m)

wall base moments Mw, Mtw , Mcw , Mwb , and 2Mw,unc(kN.m)
90000
CIP-UPT system
80000
70000
Mw
60000
50000
Mwb
40000
30000
2M w,unc
20000
M cw
10000
M tw
0.5
1.5
2.5
(a)
90000
PRE-UPT system
80000
Mw
70000
60000
50000
Mwb
40000
30000
2M w,unc
20000
M cw
10000
M tw
0.5
1.5
2.5
(b)
Fig. 7.7 Wall base moments: (a) CIP-UPT system; (b) PRE-UPT system
135
100
degree of coupling, DOC (percent)
90
80
70
60
50
40
CIP-UPT system
30
PRE-UPT system
20
10
0
0.5
1.5
2.5
Fig. 7.8 Degree of coupling

7.2.7
Wall Pier Base Concrete Strains
Figs. 7.9a and 7.9b show the neutral axis (i.e., contact) depths ctw and ccw
(measured from the compression corners of the wall piers) at the bases of the tension-side
and compression-side walls in the CIP-UPT and PRE-UPT systems, respectively, versus
the coupled wall roof drift, . The ctw and ccw values are normalized with respect to the
wall pier length, lw=3.05 m (10 ft). The corresponding extreme confined concrete
compression strains, tce and cce of the tension-side and compression-side wall piers at
the base are shown in Figs. 7.10a and 7.10b, respectively, with the confined concrete
crushing strains, ccu shown by the dashed horizontal lines.
For both structures, the neutral axis depths at the bases of the tension-side and
compression-side walls change little at large roof drift values. As expected, the
compression-side walls have larger neutral axis depths and larger concrete compression
strains than the tension-side walls as a result of the larger compressive axial forces due to
the coupling effect. Furthermore, larger wall pier compression strains develop in the
PRE-UPT system due to the additional compressive axial forces from the wall posttensioning bars.
136
wall pier base neutral axis depth, ctw and ccw / wall pier length, lw
1
0.9
CIP-UPT system
0.8
0.7
0.6
0.5
0.4
0.3
ccw / lw (compression-side wall)
0.2
0.1
0
0
ctw / lw (tension-side wall)

0.5
1.5
2
2.5
(a)
1
0.9
PRE-UPT system
0.8
0.7
0.6
0.5
0.4
ccw / lw (compression-side wall)
0.3
0.2
0.1
0
0
0.5
1.5
2
2.5
(b)
Fig. 7.9 Wall pier base neutral axis depths: (a) CIP-UPT system; (b) PRE-UPT system
137
wall pier base extreme confined concrete

compression strain, tce and cce
tce(tension-side wall)
cce
-0.01
(compression-side wall)
-0.02
ccu = 0.0182
-0.03
CIP-UPT system
-0.04
-0.05

compression strain, tce and cce
0.5
1
1.5
2
2.5
(a)
tce (tension-side wall)
-0.01
cce
-0.02
PRE-UPT system
-0.03
-0.04
ccu = 0.0385
-0.05
0
0.5
1
1.5
2
2.5
(b)
Fig. 7.10 Wall pier base extreme confined concrete compression strains:
7.2.8
Coupling Beam Shear Force versus Chord Rotation Behaviors
The solid lines in Figs. 7.11a-7.11h show the shear force versus chord rotation (Vbb) behaviors of the eight coupling beams in the CIP-UPT system, respectively. The solid
lines in Figs. 7.12a through 7.12h show similar relationships for the coupling beams in
the PRE-UPT system. The chord rotation values for the coupling beams are measured
from a tangent drawn at the left end of each beam as shown in Fig. 7.13.
For comparison, the dashed lines in Figs. 7.11 and 7.12 show the coupling beam
shear force versus chord rotation behaviors of isolated subassemblages from the CIPUPT and PRE-UPT systems, respectively. For both structures, the coupling beams from
the multistory and subassemblage models have similar Vb-b behaviors.
138
800
600
CIP-UPT system
2nd floor
400
multi-story
isolated
200
0
0
coupling beam shear force ,Vb (kN)
1000
1200
2
3
4
1000
800
600
CIP-UPT system
4th floor
400
multi-story
isolated
200
800
600
CIP-UPT system
3rd floor
400
multi-story
isolated
200
0
2
3
4
1000
800
600
CIP-UPT system
5th floor
400
multi-story
isolated
200
0
1200
2
3
4
1000
800
600
CIP-UPT system
6th floor
400
multi-story
isolated
200
0
0
0
1200
1000
800
600
CIP-UPT system
8th floor
400
multi-story
isolated
200
0
2
3
4
800
600
CIP-UPT system
7th floor
400
multi-story
isolated
200
0
0
1000
5
1000
1200
5
1200
1200
1000
800
600
CIP-UPT system
roof
400
multi-story
isolated
200
0
2
3
4
2
3
4
Fig. 7.11 Coupling beam shear force versus chord rotation behaviors
for CIP-UPT system
139
800
600
400
multi-story
isolated
200
1000
800
600
400
multi-story
isolated
200
PRE-UPT system
4th floor
0
1
2
3
4
1000
800
600
multi-story
isolated
200
PRE-UPT system
6th floor
0
0
1200
2
3
4
800
600
multi-story
isolated
200
PRE-UPT system
8th floor
0
0
2
3
4
800
600
400
multi-story
isolated
200
0
1200
2
3
4
800
600
400
multi-story
isolated
200
PRE-UPT system
5th floor
0
1
2
3
4
1200
1000
800
600
400
multi-story
isolated
200
PRE-UPT system
7th floor
0
0
1200
2
3
4
1000
800
600
400
multi-story
isolated
200
PRE-UPT system
roof
0
0
2
3
4
for PRE-UPT system
140
1000
PRE-UPT system
3rd floor
1000
400
1000
1200
400
1200
2
3
4
1200
PRE-UPT system
2nd floor
0
0
1000
1200
beam chord
beam chord rotation
transverse
displacement between
tangents at beam ends
tangent at left end

beam length
beam chord rotation =
transverse displacement between

tangents at beam ends
beam length
Fig. 7.13 Calculation of coupling beam chord rotation, b in multi-story coupled wall
7.2.9
Coupling Beam Axial Forces
Axial forces develop in the coupling beams of an unbonded post-tensioned hybrid

coupled wall structure, primarily, as a result of the post-tensioning forces applied at the
floor and roof levels. As described in Chapter 3, the prestressing of the beam posttensioning tendons is simulated in the coupled wall analytical models by tensile forces in
the beam PT elements, which are equilibrated by compression forces in the beam and
wall-contact elements. In order to reach equilibrium, the post-tensioning forces result in
an elastic shortening of the beam and wall-contact elements, which in turn result in a
reduction in the beam PT element forces. Thus, larger tensile forces are applied to the
beam PT elements in order to achieve the desired amount of initial post-tensioning forces
in the tendons after elastic shortening. For the prototype CIP-UPT and PRE-UPT systems,
a total force of Pbj=1868 kN (420 kips) is applied in the beam PT elements at each floor
and roof level to result, approximately, in the desired total beam initial post-tensioning
force of Pbi=nbtabtfbpi=1837 kN (413 kips) before the application of lateral loads, where
abt=395 mm2 is the beam post-tensioning tendon area, nbt=4 is the number of beam posttensioning tendons, and fbpi=0.625fbpu (with fbpu=1862 MPa) is the initial stress in the
tendons.
The analysis results for the CIP-UPT and PRE-UPT systems indicate that the
fixed foundation conditions assumed at the bases of the left and right wall piers and the
in-plane lateral stiffness of the wall piers affect the transfer of the post-tensioning forces
as compressive axial forces into the coupling beams. As an example, Fig. 7.14a shows the
axial force, Nb at the midspan of the 2nd floor coupling beam (as marked in Fig. 7.1) of
141
the CIP-UPT system, as well as the total force, Pb in the 2nd floor post-tensioning tendons
as the structure is displaced (thick and thin solid lines, respectively). A positive beam
axial force indicates a compressive force. Figs. 7.14b-7.14h show similar comparisons
between the beam axial forces and the beam PT tendon forces at the upper floor and roof
levels. The dashed lines in Fig. 7.14 show the total initial force, Pbi (i.e., at =0%) in the
post-tensioning tendons at each floor and roof level, which is approximately equal to the
desired initial post-tensioning force of 1837 kN (413 kips).
Fig. 7.14a shows that the initial axial force, Nbi in the 2nd floor beam before the
application of lateral loads on the structure (i.e., at =0%), is about 70% of the total
initial force, Pbi in the 2nd floor post-tensioning tendons. A significant portion of the
initial post-tensioning force is not transferred into the beam due to the fixed conditions
assumed at the base of the walls. The initial axial forces in the upper level coupling
beams are close to the total post-tensioning forces (as shown in Figs. 7.14b-7.14h),
indicating that the effect of the foundation/wall stiffness quickly diminishes above the
base.
Upon lateral displacements of the structure, the fixed foundation conditions restrain
the opening of gaps at the ends of the lower level coupling beams, resulting in the
development of additional axial forces in the beams as the walls displace. This effect is
most pronounced in the 2nd floor beam and results in the development of larger axial
forces, Nb in the beam than the total post-tensioning force Pb as shown in Fig. 7.14a.
Note that the results in Fig. 7.14 are based on analyses where the beam PT element
forces at the floor and roof levels are applied all at once. These analyses may not reflect
actual conditions from practice, where the coupling beam post-tensioning forces would
most likely be applied sequentially, beginning from the 2nd floor and ending at the roof
level. Based on the principle of superposition, the sequence of post-tensioning over the
height of the structure should not affect the axial forces that develop in the coupling
beams as long as the structure remains linear-elastic during the post-tensioning process.
Figs. 7.15a and 7.15b demonstrate two different cases where the post-tensioning
forces are applied sequentially from the 2nd floor level to the roof and from the roof to the
2nd floor level, respectively, of the CIP-UPT system. Since the DRAIN-2DX analysis
program (Prakash et al. 1993) does not allow sequential application of initial element
forces, the beam post-tensioning forces in Fig. 7.15 were simulated by external forces
applied at the floor and roof levels as shown in Fig. 7.16. The total element force of
Pbj=1868 kN (420 kips) corresponding to the desired initial post-tensioning force of
Pbi=1837 kN (413 kips) was used as the magnitude of these external forces. The structure
was assumed to remain linear-elastic (including the beam-to-wall connections) during the
entire post-tensioning operation.
142
beam initial PT force, Pbi

0
2.5
coupling beam midspan axial force, Nb
(a)
CIP-UPT system
4th floor
2500
2000
1500
1000
beam midspan axial force, Nb

beam PT force, Pb
500
0
0.5
1
1.5
2
(c)
2.5
and beam PT force, Pb (kN)
CIP-UPT system
6th floor
2500
2000
1500
1000

beam PT force, Pb
500

0
0
0.5
1
1.5
2
2.5
(e)
3000
CIP-UPT system
8th floor
2500
2000
1500
1000

beam PT force, Pb
500
0
0
0.5
1
1.5
2
2.5
2000
1500
1000

beam PT force, Pb
500

0
1
1.5
2
2.5
CIP-UPT system
5th floor
2500
2000
1500
1000

beam PT force, Pb
500
0
0.5
1
1.5
2
(d)
2.5
3000
CIP-UPT system
7th floor
2500
2000
1500
1000

beam PT force, Pb
500

0
0
0.5
1
1.5
2
2.5
(f )
3000
CIP-UPT system
roof
2500
2000
1500
1000

beam PT force, Pb
500
0
0
0.5
1
1.5
2
2.5
(h)
Fig. 7.14 Coupling beam axial forces and post-tensioning forces in CIP-UPT system:
(a) 2nd floor; (b) 3rd floor; (c) 4th floor; (d) 5th floor; (e) 6th floor;
(f) 7th floor; (g) 8th floor; (h) roof
143
(b)
(g)
0.5
3000
3000
1
1.5
2
3000
0.5
CIP-UPT system
3rd floor
2500
500

beam PT force, Pb
1000
1500
2000
3000
2500
CIP-UPT system
2nd floor
3000
2200
post-tensioning post-tensioning post-tensioning post-tensioning post-tensioning post-tensioning post-tensioning post-tensioning
beam midspan axial force, Nb (kN)
of 2nd floor
of 3rd floor
of 4th floor
of 5th floor
of 6th floor
of 7th floor
of 8th floor
of roof
Pbi =
1837 kN
2nd floor
beam
3rd floor
beam
4th floor
beam
5th floor
beam
6th floor
beam
7th floor
beam
8th floor
beam
roof
beam
CIP-UPT system
0
sequence of beam post-tensioning
(a)
2200
post-tensioning post-tensioning post-tensioning post-tensioning post-tensioning post-tensioning post-tensioning post-tensioning

of roof
of 6th floor
of 8th floor
of 3rd floor
of 5th floor
of 4th floor
of 2nd floor
of 7th floor
Pbi =
1837 kN
roof
beam
8th floor
beam
7th floor
beam
6th floor
beam
5th floor
beam
4th floor
beam
3rd floor
beam
2nd floor
beam
CIP-UPT system
0
sequence of beam post-tensioning
(b)
Fig. 7.15 Effect of beam post-tensioning sequence on beam axial forces:

(a) 2nd floor to roof; (b) roof to 2nd floor
In Fig. 7.15a, the thin to thick solid lines represent the application of the external
forces in Fig. 7.16 sequentially from the 2nd floor level to the roof. The application of the
force at the 2nd floor level results in the development of a compressive axial force in the
2nd floor coupling beam (around Nb=1200 kN) as well as a compressive force in the 3rd
floor beam (around Nb=300 kN). The axial forces in the coupling beams at the 4th floor
level and above are little influenced. Upon the application of the external force at the 3rd
floor level, the axial force in the 2nd floor beam increases to about Nb=1500 kN, the axial
force in the 3rd floor beam increases to about Nb=1600 kN, and the axial force in the 4th
144
floor beam increases to about Nb=300 kN. The axial forces that develop in the coupling
beams following the entire post-tensioning sequence over the height of the structure are
similar to the forces in Fig. 7.14. The dashed line in Fig. 7.15a represents the desired
initial post-tensioning force of Pbi=1837 kN (413 kips). As discussed previously, the
initial (i.e., after all floor and roof level beams are post-tensioned but before the
application of lateral forces) compressive force in the 2nd floor beam is significantly
smaller than the total force in the 2nd floor post-tensioning tendons due to the fixed
boundary conditions at the wall bases.
roof
Pbj
26
Pbj
23
Pbj
20
Pbj
17
Pbj
14
Pbj
11
Pbj
Pbj
53
Pbj
50
Pbj
47
Pbj
44
Pbj
41
Pbj
38
Pbj
35
Pbj
32
Pbj
8th floor
7th floor
6th floor
5th floor
hw
4th floor
3rd floor
2nd floor
lw
lb
lw
Fig. 7.16 Simulation of beam post-tensioning using external forces

Similarly in Fig. 7.15b, the external forces are applied sequentially from the roof to
the 2 floor level. By comparing the results in Figs. 7.14, 7.15a, and 7.15b, it is
concluded that the axial forces that develop in the coupling beams are not affected by the
sequence of post-tensioning over the height of the structure, as long as the structure
remains linear-elastic during this process. The results also indicate that the application of
nd
145
post-tensioning at a floor or roof level primarily affects the beam axial forces at that level
and the two adjacent levels above and below.
Note that the results in Fig. 7.15 are limited to structures that remain linear-elastic
during post-tensioning. While significant material nonlinearity is not expected to occur as
a result of post-tensioning, geometric nonlinearity can occur during sequential posttensioning of an unbonded post-tensioned system if gaps are allowed to open at the ends
of the beams that are not yet prestressed. According to Fig. 7.15a, when the 2nd floor
coupling beam is post-tensioned, small tensile forces (not shown in the figure) develop in
the 4th floor beam and above. The opening of gaps at the beam ends as a result of these
tensile forces, and the effect of these gaps on the subsequent axial forces that develop in
the coupling beams upon completion of the entire post-tensioning operation cannot be
investigated using a linear-elastic model.
The coupled wall analyses described in the remainder of this report are based on the
model where the beam post-tensioning forces are simulated using element forces (not
external forces) and applied simultaneously at all floor and roof levels.
7.2.10 Coupling Beam End Strains
Figs. 7.17 and 7.18 show the neutral axis (i.e., contact) depths, cb at the left ends of
the coupling beams in the CIP-UPT and PRE-UPT systems, respectively. The cb values
are normalized with respect to the beam depth, dbc=634 mm (24.97 in). The behaviors at
the right ends of the coupling beams are similar. The corresponding extreme cover plate
steel compression strains, be at the left ends of the coupling beams are shown in Figs.
7.19 and 7.20, respectively, with the steel yield strain, sy shown by the dashed horizontal
lines. In general, the coupling beams over the height of each structure have similar
behaviors.
7.2.11 Tension Angle Force versus Deformation Behaviors
Figs. 7.21 and 7.22 show the force versus deformation behaviors of the tension
angles at the left ends of the coupling beams in the CIP-UPT and PRE-UPT systems,
respectively. The behaviors of the tension angles at the right ends of the coupling beams
are similar. For both structures, the tension angles at the second floor level have the
smallest deformations because the fixed boundary conditions at the wall bases restrain the
gap opening at the beam-to-wall interfaces.
146
0.8
coupling beam end neutral axis

depth, cb /beam depth, dbc

1
CIP-UPT system
2nd floor
0.6
0.4
0.2
tbc=66 mm
0.6
0.4
0.2
tbc=66 mm
0
0
2
3
4
CIP-UPT system
4th floor
0.8
0.6
0.4
0.2
tbc=66 mm
CIP-UPT system
5th floor
0.8
0.6
0.4
0.2
tbc=66 mm

b
1

0
0
CIP-UPT system
6th floor
0.8
0.6
0.4
0.2
tbc=66 mm
CIP-UPT system
7th floor
0.8
0.6
0.4
0.2
tbc=66 mm
0
0
2
3
4
2
3
4
1
1
2
3
4
1

CIP-UPT system
3rd floor
0.8
CIP-UPT system
8th floor
0.8
0.6
0.4
0.2
tbc=66 mm
CIP-UPT system
roof
0.8
0.6
0.4
0.2
tbc=66 mm
0
0
2
3
4
2
3
4
Fig. 7.17 Coupling beam end neutral axis (i.e., contact) depths for CIP-UPT system
147
0.8


1
PRE-UPT system
2nd floor
0.6
0.4
0.2
tbc=66 mm
1
0.8
PRE-UPT system
3rd floor
0.6
0.4
0.2
tbc=66 mm
0
0
0.8
PRE-UPT system
4th floor
0.6
0.4
0.2
tbc=66 mm
0.8
PRE-UPT system
5th floor
0.6
0.4
0.2
tbc=66 mm
0
0

b
1
1
1

PRE-UPT system
6th floor
0.8
0.6
0.4
0.2
tbc=66 mm
PRE-UPT system
7th floor
0.8
0.6
0.4
0.2
tbc=66 mm
0
0
2
3
4
2
3
4
1
1
PRE-UPT system
8th floor
0.8
0.6
0.4
0.2
tbc=66 mm
PRE-UPT system
roof
0.8
0.6
0.4
0.2
tbc=66 mm
0
0
2
3
4
2
3
4
Fig. 7.18 Coupling beam end neutral axis (i.e., contact) depths for PRE-UPT system
148
0
steel yield strain, sy=0.00181
coupling beam end extreme

compression strain, be

-0.01
-0.02
-0.03
-0.04
0
CIP-UPT system
2nd floor
1

-0.01
-0.02
-0.03
-0.04
0
CIP-UPT system
3rd floor
1


-0.01
-0.02
CIP-UPT system
4th floor
-0.03
-0.02
CIP-UPT system
5th floor
-0.03

0


-0.04
0
-0.01
-0.02
CIP-UPT system
6th floor
-0.03
2
3
4

-0.01
-0.02
CIP-UPT system
7th floor
-0.03
-0.04
0
2
3
4
5
0


-0.01
-0.01
-0.02
CIP-UPT system
8th floor
-0.03
-0.04
0
0
-0.04
0
-0.04
0
2
3
4

-0.01
-0.02
CIP-UPT system
roof
-0.03
-0.04
0
2
3
4
5
Fig. 7.19 Coupling beam end extreme steel compression strains for CIP-UPT system
149
0


-0.01
-0.02
-0.03
-0.04
0
PRE-UPT system
2nd floor
2
3
4
5


-0.01
-0.02
-0.03
PRE-UPT system
4th floor
-0.04
0
2
3
4
5

-0.01
-0.02
-0.03
-0.04
0
PRE-UPT system
5th floor
2
3
4
5
0


PRE-UPT system
3rd floor
-0.01
-0.02
PRE-UPT system
6th floor
2
3
4
5

-0.01
-0.02
-0.03
-0.04
0
PRE-UPT system
7th floor
2
3
4
5
0


-0.03
-0.04
0
-0.02
-0.04
0
-0.03

-0.01
-0.01
-0.02
-0.03
PRE-UPT system
8th floor
-0.04
0
2
3
4
5

-0.01
-0.02
-0.03
-0.04
0
PRE-UPT system
roof
2
3
4
5
Fig. 7.20 Coupling beam end extreme steel compression strains for PRE-UPT system
150
1200
1000
1000
1200
800
600
CIP-UPT system
2nd floor
400
200
0
-200
-5
800
600
CIP-UPT system
3rd floor
400
200
0
-200
10
15
20
25
30
-5
1200
1200
1000
1000
800
600
400
CIP-UPT system
4th floor
200
0
-200
-5
10
15
20
25
30
10
15
20
25
30
25
30
25
30

1000
30
0
-200
-5
800
600
400
CIP-UPT system
6th floor
800
600
400
CIP-UPT system
7th floor
200
0
-200
10
15
20
25
30
-5

1200
1200
1000
1000
800
600
400
CIP-UPT system
8th floor
200
0
10
15
20
25
CIP-UPT system
5th floor
200
1000
-200
-5
20
400
1200
15
600
1200
-200
-5
10
800
200
800
600
400
CIP-UPT system
roof
200
0
-200
10
15
20
25
30
-5
10
15
20
Fig. 7.21 Coupling beam tension angle force-deformation behaviors for CIP-UPT system
151
1200
1000
1000
1200
800
600
400
PRE-UPT system
2nd floor
200
0
-200
5
10
15
20
25
400
PRE-UPT system
3rd floor
200
0
30
-5
1200
1200
1000
1000
800
600
400
PRE-UPT system
4th floor
200
0
-200
5
10
15
20
25
30
800
600
400
PRE-UPT system
5th floor
200
0
-200
-5
5
10
15
20
25
30
-5
1200
1200
1000
1000
600
-200
-5
800
600
PRE-UPT system
6th floor
400
200
0
-200
5
10
15
20
25
30
800
600
PRE-UPT system
7th floor
400
200
0
-200
-5
5
10
15
20
25
30
-5
1200
1200
1000
1000
800
800
600
400
PRE-UPT system
8th floor
200
0
-200
5
10
15
20
25
30
800
600
400
PRE-UPT system
roof
200
0
-200
-5
10
15
20
25
30
-5
10
15
20
25
30
Fig. 7.22 Coupling beam tension angle force-deformation behaviors for PRE-UPT system
152
7.3
Behavior Under Cyclic Loading
This section describes the behavior of the prototype CIP-UPT and PRE-UPT
systems under combined gravity loads and reversed cyclic lateral loads. The structures
are displaced to roof drift values of =0.5, 1, 1.5, 2, and 2.5%, with one cycle of loading
at each displacement amplitude. The base shear versus roof drift behavior of the
prototype systems is compared with structures that use embedded steel coupling beams.
The effect of the top and seat angles on the behavior of the prototype structures is
investigated.
7.3.1
Coupled Wall Base Shear Force versus Roof Drift Behaviors
Figs. 7.23a and 7.23b show the base shear versus roof drift (F-) behaviors of the
prototype CIP-UPT and PRE-UPT systems, respectively, under combined gravity loads
and cyclic lateral loads. The results indicate that the PRE-UPT system has considerable
inelastic energy dissipation and a large self-centering capability (i.e., ability to return
towards the zero displacement position upon unloading from a nonlinear displacement).
The CIP-UPT system has larger inelastic energy dissipation and a somewhat reduced but
still large self-centering capability. Both structures have stable behavior under large
nonlinear reversed cyclic lateral displacements.
The large self-centering capability of the structures indicate that the beam posttensioning tendons provide a sufficient amount of restoring force to pull the walls back
towards their original undisplaced position upon unloading. The increased self-centering
capability of the PRE-UPT system occurs as a result of the additional restoring effect of
the post-tensioning bars used as flexural reinforcement in the precast concrete walls. The
larger inelastic energy dissipation in the CIP-UPT system occurs, primarily, as a result of
the yielding of the wall mild steel reinforcement near the base of the walls. Note that the
unbonded post-tensioning bars used as flexural reinforcement in the precast concrete
walls have little contribution to the inelastic energy dissipation of the PRE-UPT system
since the yielding of the bars is significantly delayed as a result of unbonding (see Figs.
7.2b and 7.4b).
For comparison, Fig. 7.23c shows the expected behavior of the cast-in-place
reinforced concrete walls with embedded steel coupling beams (referred to as the CIPEMB system) representing conventional construction. Note that the lateral strength of a
post-tensioned coupling beam is smaller than the lateral strength of an embedded beam of
the same size since the post-tensioned beam cannot develop the full yield and plastic
capacity of the cross section as described in Chapter 4. To facilitate a comparative
investigation, smaller coupling beams are used in the CIP-EMB system (with W2193
cross sections) than the beams in the CIP-UPT system (with W21182 cross sections)
such that the monotonic F- relationships of the two systems are similar. The modeling
of the CIP-EMB system is similar to the modeling of the CIP-UPT system (see Chapter
6), except for the use of embedded steel coupling beam subassemblage models at the
153
4000
4000
coupled wall base shear force, F (kN)
floor and roof levels as described in Chapter 3. The embedded lengths of the coupling
beams in the CIP-EMB system are assumed to be equal to lbe=1.83 m (6 ft).
CIP-UPT system
-4000
-2.5
(a)
(b)
2.5
4000
-4000
-2.5
2.5
4000
CIP-EMB system
-4000
-2.5
PRE-UPT system
2.5
CIP-UPT system
w/o angles
-4000
-2.5
(c)
(d)
2.5
Fig. 7.23 Coupled wall base shear force versus roof drift behaviors under cyclic lateral
loads: (a) CIP-UPT system; (b) PRE-UPT system; (c) CIP-EMB system;
(d) CIP-UPT system without angles
Comparing the cyclic behaviors of the CIP-UPT and CIP-EMB systems in Figs.
7.23a and 7.23c, the CIP-EMB system has significantly larger energy dissipation, which
occurs as a result of the yielding of the coupling beams. However, the self-centering
capability of the CIP-EMB system is small indicating the possibility of significant
residual (i.e., permanent) lateral displacements after a large earthquake.
The results in Figs. 7.23a and 7.23b are for systems that include top and seat angles
at the beam-to-wall interfaces. To investigate the amount of energy dissipation provided
by the angles, Fig. 7.23d shows the behavior of the CIP-UPT system with the angles
removed. In this case, the area of the post-tensioning steel in the coupling beams is
154
doubled [to a value of abt=790 mm2 (1.224 in2)] such that the monotonic F- relationship
of the system without angles is similar to that of the system with angles.
As expected, the energy dissipation of the system without angles in Fig. 7.23d is
significantly smaller than the system with angles in Fig. 7.23a. Since unbonded posttensioned coupling beams without top and seat angles do not dissipate much energy (see
Chapter 4), the energy dissipated in Fig. 7.23d is provided primarily by the inelastic
behavior of the walls near the base.
7.3.2
Coupling Beam Axial Forces
Figs. 7.24 and 7.25 show the axial forces, Nb at the midspan of the coupling
beams and the total forces in the post-tensioning tendons, Pb respectively, at the floor and
roof levels of the CIP-UPT system corresponding to the analysis results in Fig. 7.23a. A
positive beam axial force indicates a compressive force. The and markers indicate
peak roof drift (with =0.5, 1, 1.5, 2, 2.5%) and zero roof drift positions, respectively,
during the lateral load analysis of the structure. The beam axial forces and posttensioning forces follow similar trends. The peak forces occur when the structure is
displaced to the peak roof drift during each cycle. The peak beam axial forces are similar
to or larger than (especially for the second floor level during the small displacement
cycles in Fig. 7.24a) the peak post-tensioning forces. The larger peak beam axial forces
as compared to the post-tensioning forces are due to the restraining effect of the left and
right wall piers on the opening of gaps at the beam ends. Upon returning to the original
undisplaced position (with =0%), the axial forces and post-tensioning forces generally
decrease back towards their initial levels of Nbi and Pbi (horizontal dashed lines) before
the application of lateral forces, except for the second floor beam axial force as shown in
Fig. 7.24a.
Upon unloading of the structure from =2.0%, Fig. 7.24a shows that the axial force
in the second floor coupling beam drops well below the initial axial force level of
Nbi=1467 kN even though the total force in the post-tensioning tendons in Fig. 7.25a is
close to the initial force, Pbi. The smallest axial force reached in the 2nd floor coupling
beam upon unloading from =2.5% is around Nb=199 kN, indicating that the beam has
lost most of its initial precompression at this stage of the analysis. This type of behavior
is not observed in the upper level coupling beams, which maintain most of their initial
axial force, Nbi levels throughout the lateral loading history. Note that despite the loss in
precompression upon unloading, the peak 2nd floor beam axial forces are similar to or
larger than the peak post-tensioning forces upon loading, similar to the behavior shown in
Fig. 7.14a.
155
2500
3000
2.5% -2.5%
2.0%
-2.0%
=0.5% -0.5% 1.0%-1.0% 1.5% -1.5%
2000
1500
1000
500
Nbi
zero roof drift CIP-UPT system
peak roof drift 2nd floor
3000
2500
2000
1500
Nbi
1000
500
CIP-UPT system
3rd floor
cyclic loading duration
(b)
3000
(a)
2500
2000
1500
Nbi
1000
500
CIP-UPT system
4th floor
zero roof drift

peak roof drift
3000
2500
2000
1500
Nbi
1000
500
CIP-UPT system
5th floor
(d)
(c)
3000
2500
2000
Nbi
1000
500
CIP-UPT system
6th floor
zero roof drift

peak roof drift
3000
2500
2000
1500
Nbi
1000
500
CIP-UPT system
7th floor
(f )
(e)
3000
2500
2000
1000
500
Nbi
CIP-UPT system
8th floor
zero roof drift

peak roof drift
1500
zero roof drift

peak roof drift
1500
zero roof drift

peak roof drift
zero roof drift

peak roof drift
3000
2500
2000
1500
1000
500
Nbi
CIP-UPT system
roof
zero roof drift

peak roof drift
(g)
(h)
Fig. 7.24 Coupling beam axial forces in CIP-UPT system cyclic loading:
156
3000
2.5% -2.5%
2.0% -2.0%
1.5% -1.5%
2500
1.0% -1.0%
=0.5% -0.5%
2000
1500
Pbi
1000
500
CIP-UPT system
2nd floor
2500
beam PT force, P (kN)
beam PT force, Pb (kN)
3000
zero roof drift

peak roof drift
2000
1500
1000
500
Pbi
CIP-UPT system
3rd floor
(b)
3000
3000
2500
2500
(a)
2000
1500
Pbi
1000
500
CIP-UPT system
4th floor
2000
1500
Pbi
1000
500
zero roof drift

peak roof drift
CIP-UPT system
5th floor
(d)
3000
2500
2500
(c)
3000
2000
Pbi
1000
500
CIP-UPT system
6th floor
zero roof drift

peak roof drift
2000
1500
Pbi
1000
500
CIP-UPT system
7th floor
(f )
3000
3000
2500
2500
(e)
2000
Pbi
1000
500
zero roof drift

peak roof drift
1500
zero roof drift

peak roof drift
1500
zero roof drift

peak roof drift
CIP-UPT system
8th floor
zero roof drift

peak roof drift
2000
1500
1000
500
Pbi
CIP-UPT system
roof
zero roof drift

peak roof drift
(g)
(h)
Fig. 7.25 Coupling beam post-tensioning forces in CIP-UPT system cyclic loading:
157
The horizontal (axial/shear) force equilibrium diagram in Fig. 7.26a is used to

explain the reduction of the precompression force in the 2nd floor coupling beam upon
unloading under cyclic loading. Only the left wall region above and below a floor/roof
level and half of the coupling beam is shown as a free body. The axial force in the
coupling beam, Nb is in equilibrium with the total force in the post-tensioning tendons, Pb,
the external lateral force, f, applied at the floor level, and the wall top and bottom shear
forces, Fwts and Fwbs. Let the difference between the wall bottom shear force Fwbs and the
wall top shear force Fwts for each floor and roof level be defined as Fwsi. The Fwsi values
for the second floor and roof levels of the CIP-UPT system are compared in Fig. 7.26b
(solid and dashed lines, respectively). The and markers indicate peak roof drift (with
=0.5, 1, 1.5, 2, 2.5%) and zero roof drift positions, respectively, during the lateral load
analysis of the structure.
The results in Fig. 7.26b show that, after the application of the initial beam posttensioning forces and at the beginning of the lateral load analysis, Fwsi is equal to zero at
the roof level (dashed line) while a non-zero Fwsi value is developed at the 2nd floor level
(solid line) because of the wall foundation boundary conditions. This non-zero Fwsi value
is the reason for the difference in the initial beam axial force, Nbi and total posttensioning force, Pbi at the 2nd floor level in Figs. 7.14a and 7.24a. The results in Fig.
7.26b show that there is a decrease and then reversal in the 2nd floor Fwsi value as the
structure is loaded laterally. Upon unloading of the structure back to =0%, the Fwsi value
at the 2nd floor level begins to increase, especially following the displacements to =2%
and 2.5%. This buildup in the Fwsi value at =0% results in the decrease in the coupling
beam axial force, Nb observed in Fig. 7.24a. In comparison, the Fwsi value at the roof
level (dashed line in Fig. 7.26b) is close to zero upon unloading to =0%. The behaviors
of the upper floor level coupling beams are similar to the roof beam as indicated in Figs.
7.24b-h and 7.25b-h.
wall top shear force, Fwts
left wall region
PT force, Pb
beam axial
force, Nb
lateral force, f
Fsi = Fwbs - Fwts (kN)
2000
CIP-UPT system
roof
2nd floor
zero roof drift
peak roof drift
=0.5%
l b /2
2.5%
2.0%
1.5%
1.0% -1.0%
-1.5%
-2.0%
-0.5%
-1000
wall bottom shear force, Fwbs
(a)
(b)
Fig. 7.26 Coupling beam axial forces: (a) equilibrium of horizontal forces;
(b) wall floor/roof shear forces
158
-2.5%
7.4
Parametric Investigation
The pushover lateral load behavior of a series of parametric unbonded posttensioned hybrid coupled wall systems, as determined by varying the structural properties
of the CIP-UPT and PRE-UPT systems described above, is investigated in this section.
The results are used to determine how the behavior of each system can be controlled by
design.
The following parameters are investigated: (1) wall length, lw; (2) wall
reinforcement ratio, ws or wp (defined as the total area of wall mild steel or posttensioning steel divided by the wall gross cross section area); (3) wall thickness, tw
(investigated only for systems with cast-in-place concrete wall piers); (4) initial stress in
wall post-tensioning bars, fwpi (investigated only for systems with precast concrete wall
piers); (5) beam web depth, dbw; (6) beam post-tensioning tendon area, abt; (7) initial
stress in beam post-tensioning tendons, fbpi; (8) beam length, lb; and (9) top and seat angle
thickness, ta.
Three coupled wall systems are investigated for each parameter: System (1) is the
same as the prototype CIP-UPT or PRE-UPT system and the properties of Systems (2)
and (3) are varied as shown in Tables 7.2 and 7.3. All other structural properties of
Systems (2) and (3) are the same as System (1). A discussion of the results from the
parametric analyses is provided below.
TABLE 7.2
PARAMETRIC CIP-UPT SYSTEMS
Parameter Varied
Wall length, lw
Wall flexural reinforcement ratio, ws
Wall thickness, tw
Beam web depth, dbw
Beam PT tendon area, abt
Initial stress of beam PT tendons, fbpi
Beam length, lb
Angle thickness, ta
System (1)
3.05 m
1.47%
356 mm
502 mm
395 mm2
0.625fbpu
3.05 m
28.6 mm
System (2)
3.81 m (1.25)
1.84% (1.25)
254 mm (0.71)
629 mm (1.25)
592 mm2 (1.50)
0.725fbpu (1.16)
3.81 m (1.25)
19.1 mm (0.67)
System (3)
2.29 m (0.75)
2.21% (1.50)
457 mm (1.28)
375 mm (0.75)
197 mm2 (0.50)
0.525fbpu (0.84)
4.57 m (1.50)
0 mm (0)
Note: Numbers in parentheses show parameter values for Systems (2) and (3) divided by value for System
(1).
TABLE 7.3
PARAMETRIC PRE-UPT SYSTEMS
Parameter Varied
Wall length, lw
Wall flexural reinforcement ratio, wp
Initial stress of wall PT bars, fwpi
Beam web depth, dbw
Beam length, lb
Angle thickness, ta
System (1)
3.05 m
1.13%
0.625fwpu
502 mm
395 mm2
0.625fbpu
3.05 m
28.6 mm
System (2)
3.81 m (1.25)
1.41% (1.25)
0.725fwpu (1.16)
629 mm (1.25)
592 mm2 (1.50)
0.725fbpu (1.16)
3.81 m (1.25)
19.1 mm (0.67)
System (3)
2.29 m (0.75)
1.69% (1.50)
0.525fwpu (0.84)
375 mm (0.75)
197 mm2 (0.50)
0.525fbpu (0.84)
4.57 m (1.50)
0 mm (0)
Note: Numbers in parentheses show parameter values for Systems (2) and (3) divided by value for System
(1).
159
7.4.1 Base Shear versus Roof Drift Behavior

The base shear versus roof drift behaviors of the parametric coupled wall systems
in Tables 7.2 and 7.3 are given in Figs. 7.27 and 7.28, respectively. The markers shown
on the figures represent the states of behavior defined in Fig. 7.2.
The base shear versus roof drift behaviors of the parametric coupled wall systems
show that the lateral strength of the structures can be controlled by the beam posttensioning tendon area abt, beam depth dbw, and angle thickness ta, as well as the wall
length lw. The yielding of the beam post-tensioning steel, which is significantly delayed
due to the use of unbonded tendons, can be controlled by the initial stress in the posttensioning steel, fbpi. More information on the parameters that control the behavior of the
coupling beams can be found in Chapter 4 and more information on the parameters that
control the behavior of unbonded post-tensioned precast concrete walls (e.g., the
eccentricity of the wall post-tensioning bars) can be found in Kurama et al. (1996, 1999a).
7.4.2 Degree of Coupling
As described previously, the degree of coupling, DOC, is an important factor for
the lateral load behavior of coupled wall structures. For this purpose, Tables 7.4 and 7.5
show the degrees of coupling for the parametric coupled wall systems in Tables 7.2 and
7.3. As shown in Fig. 7.8, the degree of coupling varies by a small amount during lateral
loading. The degree of coupling values in Tables 7.4 and 7.5 were calculated using
Equation (2.2) at a roof drift of =0.5%.
In Table 7.4, the degree of coupling for the cast-in-place concrete wall systems (see
Fig. 7.27) ranges between 47%-68%. Similarly, the degree of coupling for the precast
concrete wall systems in Fig. 7.28 ranges between 44%-63% (see Table 7.5). In Chapter
13, the behavior of coupled wall systems with lower coupling degrees (DOC=30-35%)
will be demonstrated. Since the different parameters in Tables 7.4 and 7.5 are varied by
different amounts, it is difficult to directly compare the relative effects of these
parametric variations on the degree of coupling. However, the results indicate that the
wall length, beam depth, and angle thickness have larger influences on the coupling
degree than the other parameters investigated. The beam length, beam post-tensioning
steel area, wall flexural reinforcement area, and initial stresses in the beam and wall posttensioning steel have smaller effects, and the wall thickness has the smallest effect on the
degree of coupling. The results for the systems with cast-in-place concrete and precast
concrete walls are similar.
160
4500
CIP-UPT system
2
1
3
1 l w=3.05m
tension side wall soften. 2 lw=3.81m
comp. side wall soften. 3 l =2.29m
w
coupling beam soften.
n
comp. wall conc. crush.
3.5
4500
CIP-UPT system
1
tension side wall soften.

n 1 =1.47%
ws
comp. side wall soften.
first wall steel yielding 2 ws=1.84%
3 ws=2.21%
(b)
4500
CIP-UPT system
3
2
tension side wall soften. 1 tw=356mm

comp. side wall soften. 2 t =254mm
w
coupling beam soften. 3 tw=457mm
3.5
4500
CIP-UPT system
=502mm
tension side wall soften. 1 dbw
first wall steel yielding 2 dbw=629mm
coupling beam soften. 3 d =375mm
bw
2
1
3
1 abt =
tension side wall soften. 395mm2
comp. side wall soften. 2 a bt=
592mm2
first beam PT yielding 3 a bt=
197mm2
3.5
4500
2
3
tension side wall soften. 1 l =3.05m

b
first wall steel yielding 2 l b=3.81m
3 l b=4.57m
f bpi=0.625f bpu
f bpi=0.725f bpu
f bpi=0.525f bpu
CIP-UPT system
2
1
3
3.5
(f )
4500
CIP-UPT system
1
2
3

comp. side wall conc. crushing
(e)
3.5
(d)
4500
2
1
(c)
CIP-UPT system
3.5
(a)
3
2
3.5
(g)
4500
1 t a=28.6mm (L881-1/8)
2 ta=19.1mm (L883/4)
3 ta=0 (no angle)
CIP-UPT
1
2

comp. side wall soften. reached at
coupling beam soften. =4.11%
3.5
(h)
Fig. 7.27 Base shear versus roof drift behaviors of parametric CIP-UPT systems:
(a) lw; (b)ws; (c) tw; (d) dbw; (e) abt; (f) fbpi; (g) lb; (h) ta
161
4500
PRE-UPT system
1
3
1 l w=
tension side wall soften. 3.05m

coupling beam soften. 2 lw=
3.81m
ten. wall conc. lin. lim .
first wall steel yielding 3 l =
w
2.29m
3.5
4500
PRE-UPT system

comp. side wall soften. 1 wp=1.13%
coupling beam soften. 2 =1.41%
wp
first wall steel yielding 3 wp=1.69%
1 fwpi=0.625f wpu
2 fwpi=0.725fwpu
3 fwpi=0.525fwpu
PRE-UPT system
2
1
3

tens. wall conc. linear limit
comp. wall conc. crushing
3.5
(c)
4500
1 abt=395mm
2
2 abt=592mm
2
3 abt=197mm
PRE-UPT system
2
1
3

3.5
4500
PRE-UPT system
3
1 dbw=
502mm
d =
coupling beam soften. 2 bw
629mm
first wall steel yielding 3 d =
bw
375mm
l b=3.05m
l b=3.81m
l b=4.57m
PRE-UPT system
1
2
3

3.5
(d)
4500
1
2
3
fbpi=0.625f bpu
fbpi=0.725f bpu
fbpi=0.525f bpu
PRE-UPT system
2
1
3

3.5
(f )
1
2
3
2
1
(e)
4500
3.5
(b)
(a)
4500
3
2
1
3.5
(g)
4500
1 t a=28.6mm (L881-1/8)
2 ta=19.1mm (L883/4)
3 ta=0 (no angle)
1
2
3
PRE-UPT
system
0

3.5
(h)
Fig. 7.28 Base shear versus roof drift behaviors of parametric PRE-UPT systems:
(a) lw; (b) wp; (c) fwpi; (d) dbw; (e) abt; (f) fbpi; (g) lb; (h) ta
162
TABLE 7.4
DEGREE OF COUPLING OF PARAMETRIC CIP-UPT SYSTEMS
Degree of Coupling, DOC (%)
System (1)
System (2)
System (3)
62
58 (0.94)
68 (1.10)
62
59 (0.95)
57 (0.92)
62
64 (1.03)
61 (0.98)
62
66 (1.06)
56 (0.90)
62
65 (1.05)
56 (0.90)
62
63 (1.02)
60 (0.97)
62
59 (0.95)
56 (0.90)
62
55 (0.89)
47 (0.76)
Parameter Varied
Wall length, lw
Wall thickness, tw
Beam web depth, dbw
Beam length, lb
Angle thickness, ta
Note: Numbers in parentheses show DOC values for Systems (2) and (3) divided by DOC value for System
(1).
TABLE 7.5
DEGREE OF COUPLING OF PARAMETRIC PRE-UPT SYSTEMS
Parameter Varied
Wall length, lw
Beam web depth, dbw
Beam length, lb
Angle thickness, ta
Degree of Coupling, DOC (%)

System (1)
System (2)
System (3)
58
54 (0.93)
63 (1.09)
58
55 (0.95)
52 (0.90)
58
56 (0.97)
60 (1.03)
58
62 (1.07)
52 (0.90)
58
62 (1.07)
52 (0.90)
58
60 (1.03)
57 (0.98)
58
55 (0.95)
53 (0.91)
58
52 (0.90)
44 (0.76)
Note: Numbers in parentheses show DOC values for Systems (2) and (3) divided by DOC value for System
(1).
7.4.3 Coupling Beam Axial Forces

As described previously using Figs. 7.14a and 7.26, a significant portion of the
total initial post-tensioning force, Pbi at the 2nd floor level is not transferred into the 2nd
floor coupling beam as a result of the fixed base conditions assumed for the walls. This
effect quickly diminishes above the 2nd floor level. In order to investigate this behavior
further, Tables 7.6 and 7.7 show the 2nd floor beam initial axial force, Nbi divided by total
initial post-tensioning force Pbi for the parametric coupled wall structures in Tables 7.2
and 7.3. It is clear that among the parameters investigated, the wall length has the largest
influence on the amount of initial axial force transferred into the 2nd floor coupling beam,
since it results in the largest effect on the lateral stiffness of the left and right side wall
piers.
163
TABLE 7.6
SECOND FLOOR BEAM INITIAL AXIAL FORCE DIVIDED BY TOTAL INITIAL
PT FORCE IN PARAMETRIC CIP-UPT SYSTEMS
Parameter Varied
Wall length, lw
Wall thickness, tw
Beam web depth, dbw
Beam length, lb
Angle thickness, ta
System (1)
73%
73%
73%
73%
73%
73%
73%
73%
Nbi / Pbi
System (2)
68%
75%
77%
76%
75%
75%
73%
73%
System (3)
83%
74%
74%
74%
75%
75%
71%
72%
TABLE 7.7
SECOND FLOOR BEAM INITIAL AXIAL FORCE DIVIDED BY TOTAL INITIAL
PT FORCE IN PARAMETRIC PRE-UPT SYSTEMS
Design Parameter
Wall length, lw
Beam web depth, dbw
Beam length, lb
Angle thickness, ta
7.5
System (1)
76%
76%
76%
76%
76%
76%
76%
76%
Nbi / Pbi
System (2)
69%
76%
76%
77%
76%
76%
75%
74%
System (3)
84%
76%
76%
75%
76%
76%
74%
72%
Chapter Summary
This chapter presents an analytical investigation on the nonlinear lateral load

behavior of multi-story unbonded post-tensioned hybrid coupled wall structures. The
properties, and wall properties on the behavior of the structures, including the degree of
coupling, energy dissipation, and lateral displacement capacity is investigated. Systems
with precast concrete walls as well as monolithic cast-in-place reinforced concrete walls
are considered. The behavior of post-tensioned coupled wall systems is compared with
the behavior of systems with embedded steel coupling beams and systems without
coupling. Some of the conclusions based on the analysis results are as follows:
(1) Post-tensioned steel beams can provide significant and stable levels of coupling
between concrete walls over large nonlinear cyclic deformations, similar to the levels of
coupling that can be developed using embedded steel beams. As a result of the use of
unbonded post-tensioning steel and the opening of gaps at the beam-to-wall interfaces,
little damage is expected to occur in the coupling beams and in the regions of the walls
near the beams.
164
(2) As a coupled wall structure is displaced laterally, the axial forces in the
compression-side wall increase and the axial forces in the tension-side wall decrease due
to the coupling beam forces. This can lead to tension forces in the tension-side wall and
large compressive and shear forces in the compression-side wall. Thus, large levels of
coupling should be avoided in design.
(3) The lateral displacements and rotations of the tension-side and compressionside walls in a coupled wall structure under lateral loading are similar.
(4) As compared with embedded steel coupling beams, post-tensioned beams can
provide a large restoring force to the walls, reducing the residual lateral displacements
upon unloading from a nonlinear displacement. This results in a large self-centering
capability of the structure.
(5) Unbonded post-tensioned hybrid coupled walls dissipate less energy than walls
with embedded steel coupling beams. Most of the energy dissipation is provided by the
yielding of the top and seat angles used at the beam-to-wall interfaces and the nonlinear
behavior of the walls near the base.
(6) Unbonded post-tensioned coupled wall structures with cast-in-place reinforced
concrete walls have larger inelastic energy dissipation and a somewhat reduced but still
large self-centering capability as compared with structures that use post-tensioned precast
concrete walls.
(7) In order to achieve the same lateral strength, coupled wall structures with
unbonded post-tensioned steel coupling beams require beams of larger cross section than
walls with embedded steel coupling beams.
(8) The lateral strength of unbonded post-tensioned hybrid coupled walls can be
controlled by the area of the beam post-tensioning steel, as well as other parameters such
as the wall length, beam depth, and thickness of the beam-to-wall connection angles. The
yielding of the post-tensioning steel can be controlled by the initial stress in the steel, and
can be significantly delayed due to the use of unbonded tendons.
(9) The degree of coupling remains relatively constant as a coupled wall structure is
displaced laterally into the nonlinear range.
(10) A significant portion of the total initial force in the 2nd floor coupling beam
post-tensioning tendons is not transferred into the 2nd floor coupling beam as a result of
the fixed base conditions assumed for the walls. The initial axial forces in the upper level
coupling beams are close to the total initial post-tensioning forces, indicating that the
effect of the foundation/wall stiffness quickly diminishes above the base. The wall pier
length has the largest influence on the amount of initial post-tensioning force transferred
into the 2nd floor beam, since it results in the largest effect on the lateral stiffness of the
wall piers.
165
(11) Upon lateral displacements of the structure, the fixed foundation conditions
restrain the opening of gaps at the ends of the 2nd floor coupling beam, resulting in the
development of additional axial forces in the beam as the walls displace.
(12) During large nonlinear cyclic lateral displacements of an unbonded posttensioned hybrid coupled wall system, the 2nd floor coupling beam can loose most of its
initial precompression even if the total force in the post-tensioning tendons remains close
to the initial force. This behavior is not observed in the upper level coupling beams,
which maintain most of their initial axial force levels.
(13) The initial axial forces that develop in the coupling beams are not affected by
the sequence of post-tensioning over the height of the structure.
166
CHAPTER 8
IDEALIZED COUPLED WALL LATERAL
LOAD-DISPLACEMENT RELATIONSHIPS
This chapter presents idealized base shear force versus roof drift relationships for
unbonded post-tensioned hybrid coupled wall structures under monotonic lateral loading.
Procedures are developed to estimate the structure behavior using basic principles of
equilibrium, compatibility, and constitutive relationships. These procedures are used later
in the report as design tools for unbonded post-tensioned hybrid coupled wall structures.
The chapter is divided into the following sections: (1) linear-elastic behavior; (2) systems
with cast-in-place concrete walls; (3) systems with precast concrete walls; and (4)
verification of idealized relationships.
8.1
Linear-Elastic Behavior
For design purposes, the linear-elastic behavior of a multi-story unbonded posttensioned hybrid coupled wall structure under lateral loads can be determined using a
linear-elastic structural analysis program [e.g., SAP2000 (CSI 1999)]. The cross-sectional
properties and boundary conditions of the model can be determined from the multi-story
coupled wall fiber element model in Fig. 6.5 and from the subassemblage linear-elastic
model in Fig. 5.1. Note that in the linear-elastic range of behavior, the effects of the beam
and wall post-tensioning tendons and the beam-to-wall connection angles can be ignored.
Gap opening behavior at the beam-to-wall joints and at the precast concrete wall panel
joints causes geometric nonlinear effects in the structure, and thus, is not included in the
linear-elastic model. As a result, unbonded post-tensioned coupled wall systems with
precast concrete walls can be modeled the same as systems with monolithic cast-in-place
reinforced concrete walls.
In lieu of considering the local deformations in the wall-contact regions as
described in Chapter 5 using the subassemblage model in Fig. 5.1a, a simpler linearelastic analytical model can be developed as shown in Fig. 8.1 for an eight story structure
(see also Fig. 5.1b). The deformations in the wall-contact elements are ignored and rigid
zones are used between the ends of the coupling beams and the wall centerlines (similar
to the model in Fig. 2.4a). Ignoring the flange cover plates (if any) and including shear
deformations, the initial linear-elastic stiffness matrix for the steel coupling beams is
given as:
167
Eb Ab
l
b
0
[Kbi ] = E A
b b
lb
12Eb I b
6Eb I b
lb3 (1 + 2bg )
6Eb I b
lb2 (1 + 2bg )
4Eb I b (1 + bg /2)
lb2 (1 + 2bg )
lb (1 + 2bg )
12Eb I b
3
lb (1 + 2bg
6Eb I b
2
lb (1 + 2bg
Eb Ab
lb
0
Eb Ab
lb
6Eb I b
2
lb (1 + 2bg
)
2Eb I b (1 bg )
lb (1 + 2bg )
0
12Eb I b
lb3 (1 + 2bg )
6Eb I b
lb2 (1 + 2bg )
0
12Eb I b
3
lb (1 + 2bg
6Eb I b
lb2 (1 + 2bg )
2Eb I b (1 bg )
lb (1 + 2bg )
6Eb I b
2
lb (1 + 2bg )
4Eb I b (1 + bg /2)
lb (1 + 2bg )
6Eb I b
2
lb (1 + 2bg
(8.1)
where,
bg =
6 Eb I b
Gb Abg lb2
(8.2)
with Gb (shear modulus of the coupling beam steel) and Abg (shear area of the coupling
beam cross section) given as:
Gb =
Eb
and Abg = dbtbw
2(1 + vb )
(8.3)
and lb is the coupling beam span length (as shown in Fig. 8.1), Ib is the moment of inertia,
Ab is the gross area, db is the depth, and tbw is the web thickness of the coupling beam
cross section (ignoring the beam flange cover plates), and b and Eb are the Poissons
ratio and modulus of elasticity of the beam steel, respectively (assumed as b=0.3 and
Eb=199955 MPa, resulting in Gb=76907 MPa).
168
steel beam
26
53
23
50
20
47
17
44
14
41
outline of
coupled wall
model
coupled wall
hw
infinitely
rigid link
38
11
steel
beam
8
35
32
28
lw
lb
concrete wall
lw
Fig. 8.1 Eight story linear-elastic coupled wall model
169
Similarly, the initial linear-elastic stiffness matrix for the concrete walls is given as:
Ec Aw
l
w
0
0
[Kci ] = E A
c w
l
w
12Ec I w
3
lw (1 + 2 wg )
6Ec I w
2
lw (1 + 2 wg )
6Ec I w
4Ec I w (1 + wg / 2)
lw2 (1 + 2 wg )
lw (1 + 2 wg )
12Ec I w
lw3 (1 + 2 wg )
6Ec I wg
lw2 (1 + 2 wg )
Ec Aw
lw
0
12Ec I w
3
lw (1 + 2 wg )
6Ec I w
2
lw (1 + 2 wg )
Ec Aw
lw
6Ec I w
lw2 (1 + 2 wg )
2Ec I w (1 wg )
lw (1 + 2 wg )
0
12Ec I w
lw3 (1 + 2 wg )
6Ec I w
2
lw (1 + 2 wg )
6Ec I w
2
lw (1 + 2 wg )
2Ec I w (1 wg )
lw (1 + 2 wg )
6Ec I w
2
lw (1 + 2 wg )
4Ec I w (1 + wg / 2)
lw (1 + 2 wg )
(8.4)
where,
wg =
6 Ec I w
Gc Awlw2
(8.5)
with the shear modulus of the wall concrete, Gc given as:
Gc =
Ec
2(1 + vc )
(8.6)
and lw is the wall length (as shown in Fig. 8.1), Iw is the moment of inertia and Aw is the
area of the wall cross section, and c and Ec are the Poissons ratio and modulus of
elasticity for the wall concrete, respectively (assumed as c=0.18 and Ec=30441 MPa for
f c' =41.4 MPa, resulting in Gc=12901 MPa).
A Matlab algorithm (2000) was developed using the beam and wall element
stiffness matrices above to conduct linear-elastic structural analyses of multi-story
coupled wall structures under lateral loads. The initial linear-elastic lateral stiffness of the
structure is defined using the base shear versus roof displacement (F-u) relationship as:
K wi =
170
F
u
(8.7)
For the prototype CIP-UPT and PRE-UPT systems described in Chapter 6, an

initial linear-elastic stiffness of Kwi=41 kN/mm (233 kip/in.) results from the Matlab
algorithm, which compares well with the value of Kwi=40 kN/mm (227 kip/in.) from the
DRAIN-2DX fiber element model including the beam cover plates and the wall-contact
elements.
8.2
Systems with Cast-in-Place Concrete Walls
The nonlinear base shear force versus roof drift (F-) behavior of an unbonded
post-tensioned hybrid coupled wall system with monolithic cast-in-place reinforced
concrete walls is idealized using a bilinear relationship. As an example, Fig. 8.2 shows
the bilinear F- relationship estimated for the prototype CIP-UPT system in Chapter 6.
The idealized base shear versus roof drift relationship is identified by the coupled wall
softening state (at Fws, ws) and the coupled wall ultimate state (at Fwu, wu). A procedure
to estimate Fws, ws, Fwu, and wu is developed below using basic principles of
equilibrium, compatibility, and constitutive relationships. This estimation procedure can
be used as a tool in the seismic analysis and design of coupled wall structural systems,
without the need to develop nonlinear fiber element models.
4000
base shear force, F (kN)
fiber element model

bilinear estimation
(Fws, ws)
(Fwu, wu)
CIP-UPT system
compression-side wall softening

compression-side wall concrete crushing
estimation points
1
2
Fig. 8.2 Idealized base shear force versus roof drift relationship
for prototype CIP-UPT system
8.2.1
Coupled Wall Softening State
The coupled wall softening state is defined as the state corresponding to a

significant reduction in the lateral stiffness of the coupled wall system. For a system with
171
cast-in-place concrete walls, the base shear force Fws and roof drift ws at this state are
estimated using an iterative procedure with the following assumptions:
(1) The coupling beams are at the beam softening state and the coupling shear force
corresponding to the beam softening state, Vb,sof, can be estimated as described in Chapter
5. This assumption is justified, on average, based on the marker in Fig. 7.2a.
(2) The strain distributions at the bases of the walls are linear.
(3) The concrete compressive stresses at the bases of the walls have linear
distributions.
(4) The neutral axis at the base of the compression-side wall (i.e., right side wall for
lateral loads applied from left to right) is located at the centerline of the wall (i.e.,
ccws=0.5lw). The + marker in Fig. 7.2a shows that this assumption is reasonable.
(5) The strain and stress in the extreme flexural reinforcement of the tension-side
wall is equal to the yield strain wsy and stress fwsy, respectively. The marker in Fig. 7.2a
shows that this assumption is reasonable.
As an example, Fig. 8.3a shows a free body diagram to determine the wall
softening state for the prototype CIP-UPT system described in Chapter 6. Note that the
axial forces and moments developed at the beam ends are not shown in the free body
diagram. The steps used in the estimation of Fws and ws are presented below:
Step 1
Estimate the axial forces at the bases of the tension-side and compression-side
walls, Ntws and Ncws, respectively, as:
N tws = G 1.05 Vb,ws
(8.8a)
N cws = G + 1.05 Vb,ws
(8.8b)
where, G represents the gravity loads applied at the floor and roof levels and Vb,ws
represents the shear forces in the coupling beams (see Fig. 8.3a). Based on Assumption
(1), the coupling beam shear forces can be calculated as Vb,ws=Vb,sof (see Fig. 8.3b)
corresponding to the beam softening state for an isolated coupled wall subassemblage
from Chapter 5 [see Equation (5.10)].
Note that under monotonic lateral loading, the coupling beam shear forces that
develop in the lower stories of a multi-story structure are typically larger than the beam
shear force estimated using an isolated subassemblage. This is because of the additional
axial forces that develop in the lower story beams as shown in Fig. 7.14. The 1.05 factor
in Equations (8.8a) and (8.8b) represents this effect, on an average and approximate sense,
for the beams used in a multi-story coupled wall system. This factor is discussed further
in Chapter 14.
172
53
26
1.05Vb,ws
50
23
1.05Vb,ws
resultant
lateral force, Fws
20
47
1.05Vb,ws
17
44
1.05Vb,ws
14
41
hw
1.05Vb,ws
11
38
1.05Vb,ws
3000
35
beam shear force, Vb (kN.m)
1.05Vb,ws
32
1.05Vb,ws
1
28
CIP-UPT system
N tws
N cws
M tws
M cws
lw
lb
(Vb,pty,b,pty)
(Vb,ws=Vb,sof)
0 b,sof
lw
(a)
tension wsy
(b)
strain distribution
tension-side wall
strain distribution
ctws
tension cws
compression
compression
ccws=l w /2
lw
lw
tension-side wall
concrete stresses
f wsy
concrete stresses
tension-side wall
steel stresses (17 layers of
bars as shown in Fig. 6.2a)
f cws
steel stresses (17 layers
of bars as shown in Fig. 6.2a)
(c)
Fig. 8.3 Coupled wall softening state for systems with cast-in-place concrete walls:
(a) free body diagram; (b) coupling beam shear forces;
(c) strain and stress distributions at wall bases
173
Step 2
Select initial values for the neutral axis depth at the base of the tension-side wall,
ctws and the strain in the extreme flexural reinforcement of the compression-side wall,
cws< wsy when the coupled wall softening state is reached.
Step 3
Determine the axial strains and stresses in the concrete and flexural steel at the
bases of the tension-side and compression-side walls using Assumptions (2)-(5), and the
ctws and cws values selected in Step 2 (see Fig. 8.3c). Ignore the tensile stresses in the
concrete.
Step 4
Use the concrete and steel stresses from Step 3 to determine the axial forces at the
bases of the tension-side and compression-side walls, Ntws and cws.
Step 5
Repeat Steps 2-4 until a satisfactory agreement between the Ntws and cws values
from Steps 1 and 4 is achieved.
Step 6
Use the concrete and steel stresses from Step 3 to determine the moments at the
bases of the tension-side and compression-side walls, Mtws and Mcws, respectively, by
taking moments about the wall centerlines.
Step 7
Using the free body diagram in Fig. 8.3a, determine the total moment at the base of
the coupled wall system by taking moments about the system centerline as:
M ws = M tws + M cws + 1.05 (lw + lb ) Vb, ws
(8.9)
The total coupled wall base shear force can be determined as:
Fws =
M ws
H
(8.10)
where, H is the resultant height of the lateral loads applied on the walls.
Step 8
Determine the degree of coupling at the coupled wall softening state (to be used
later) as:
DOCws =
1.05 (lw + lb ) Vb, ws
M ws
174
(8.11)
Note that this definition for the degree of coupling [based on Equation (2.2)] is different
from the definition used in Chapter 7 based on the lateral load resistances of the coupled
and uncoupled wall systems.
Step 9
Determine the coupled wall roof drift at the wall softening state, ws using the
initial lateral stiffness, Kwi from a linear-elastic analysis of the structure (see Section 8.1)
as:
ws =
Fws
hw K wi
(8.12)
where, hw is the wall height.

8.2.2
Coupled Wall Ultimate State
The coupled wall ultimate state is defined as the state corresponding to the crushing
of the confined concrete at the base of the compression-side wall. For a system with castin-place concrete walls, the base shear Fwu and roof drift wu at this state are estimated
using an iterative procedure with the following assumptions:
(1) The maximum compression strain in the confined concrete at the base of the
compression-side wall is equal to the ultimate (i.e., crushing) strain of the confined
concrete, ccu. This assumption is based on the definition of the coupled wall ultimate
state above.
(2) The strain distributions at the bases of the walls are linear.
(3) The height of the plastic hinge, hwp at the base of the compression-side wall is
equal to the confined thickness of the wall, twc (i.e., the wall thickness enclosed by the
centerline of the reinforcing bars or wires used as concrete confinement).
(4) The coupling degree at the coupled wall ultimate state, DOCwu is the same as
the coupling degree at the wall softening state, DOCws. This assumption is justified based
on Fig. 7.8.
(5) The rotations and lateral displacements of the tension-side and compressionside walls are the same. This assumption is justified based on Fig. 7.3a.
(6) The moment at the base of the tension-side wall at the coupled wall ultimate
state, Mtwu is the same as the moment at the wall softening state, Mtws. Fig. 7.7a shows
that this assumption is reasonable.
As an example, Fig. 8.4a shows the assumed strain and stress distributions [based
on Assumptions (1)-(2)] at the base of the compression-side wall when the ultimate state
175
of the prototype CIP-UPT system is reached. The steps used in the estimation of Fwu and
wu are presented in the following. Note that the differences between the unconfined
cover concrete and the confined concrete are ignored (as shown in Fig. 8.4a) in the
procedure below. A revised procedure can be developed by considering the crushing of
the unconfined cover concrete in the proposed formulation.
tension coupling
beam
side wall chord
strain distribution
cwu
tension
compression
side wall
compression
ccu
b,wu wup(l w -ccwu )/l b
ccwu
lw
tension
wup
concrete stresses
compression
steel stresses (17 layers,
as shown in Fig. 6.2a)
wup
wup
tension
lw
ccwu
lb
compression
(a)
tension-side wall deforms
no compression
side wall
(b)
compression-side
tension-side wall deforms wall does not deform
coupling
beam
chord
b,wu
wup
coupling
beam chord
wup
tension-side wall
deforms
deforms
wup(lw -c cwu )/lb

wup
wup
wup
wup (lw -c cwu )
wup
ccwu
(c)
Fig. 8.4 Coupled wall ultimate state for systems with cast-in-place concrete walls:
(a) strain and stress distributions at base of compression-side wall; (b) beam chord
rotation; (c) components of beam chord rotation
176
Step 1
Select an initial value for the neutral axis depth at the base of the compression-side
wall ccwu, when the coupled wall ultimate state is reached. An initial value of
ccwu=ccws=lw/2 may be selected.
Step 2
Determine the axial strains and stresses in the concrete and flexural steel at the base
of the compression-side wall using Assumptions (1) and (2), and the ccwu value selected
in Step 1 (see Fig. 8.4a). Ignore the tensile stresses in the concrete.
Step 3
Use the concrete and steel stresses from Step 2 to determine the axial force at the
base of the compression-side wall, Ncwu.
Step 4
Estimate the curvature at the base of the compression-side wall as:
cwu =
ccu
(8.13)
ccwu
Step 5
Estimate the coupled wall roof drift due to the plastic rotation at the base of the
compression-side wall as:
wup = 1.85
cwu hwp
DOCwu
(8.14)
where, hwp=twc and DOCwu=DOCws based on Assumptions (3) and (4), respectively. This
equation was determined based on the results obtained from the parametric coupled wall
analyses in Chapter 7.
Step 6
Estimate the chord rotation in the coupling beams at the coupled wall ultimate state
as:
b, wu = wup 1 +
lw ccwu
lb
(8.15)
based on the idealized coupled wall displaced shape in Fig. 8.4b and Assumption (5).
Note that the beam chord rotation includes two components; one is wup from Step 5 due
to the rotation of the tension-side wall and the other is wup(lw-ccwu)/lb due to the rotation
of the compression-side wall, as shown in more detail in Fig. 8.4c. The transverse (i.e.,
vertical) displacements at the ends of the coupling beams supported by the tension-side
wall are ignored in the estimation of b,wu in Equation (8.15).
177
Step 7
Estimate the shear force in the coupling beams Vb,wu corresponding to the beam
rotation b,wu using the idealized subassemblage shear force versus rotation relationship
described in Chapter 5 (see Fig. 8.5a).
G
G
f
53
26
1.05Vb,wu
50
23
1.05Vb,wu
resultant
lateral force, Fwu
47
20
1.05Vb,wu
17
44
1.05Vb,wu
hw
41
14
1.05Vb,wu
3000
38
11
1.05Vb,wu
35
1.05Vb,wu
(Vb,pty,b,pty)
V b,wu
(Vb,sof, b,sof)
prototype
subassemblage
32
1.05Vb,wu
1
28
CIP-UPT system
b,wu
N twu
N cwu
M twu
M cwu
lw
(a)
lb
lw
(b)
Fig. 8.5 Coupled wall ultimate state for systems with cast-in-place concrete walls:
(a) coupling beam shear forces; (b) free body diagram
Step 8
Determine the total axial force at the base of the compression-side wall as:
N cwu = G + 1.05 Vb, wu
(8.16)
based on the free body diagram at the coupled wall ultimate state as shown in Fig. 8.5b.
Note that the axial forces and moments developed at the beam ends are not shown in Fig.
178
8.5b. The 1.05 factor in Equation (8.16) is described in Section 8.2.1 for the coupled wall
softening state.
Step 9
Repeat Steps 1-8 until a satisfactory agreement between the Ncwu values in Steps 3
and 8 is achieved.
Step 10
Use the concrete and steel stresses from Step 2 to determine the moment at the base
of the compression-side wall, Mcwu, by taking moments about the centerline of the wall.
Step 11
Using the free body diagram in Fig. 8.5b, determine the total moment at the base of
M wu = M twu + M cwu + 1.05 (lw + lb ) Vb, wu
(8.17)
where, Mtwu=Mtws based on Assumption (6).

Fwu =
M wu
H
(8.18)
Based on the multi-story coupled wall analyses in Chapter 7, second order effects
(often referred to as P- effects) in the tension-side and compression-side walls have a
small influence on the base shear force versus roof drift relationship of the structure, and
thus, can be neglected. Note that the second order effects in the coupling beams and the
beam post-tensioning tendons are included in the calculation of Vb,wu as described in
Chapter 5. Note also that a consideration of second order effects is not necessary in the
estimation of the coupled wall softening state since the displacements of the structure at
the softening state are relatively small.
Step 12
Estimate the coupled wall roof drift as:
wu = wup + wue
(8.19)
where, wup is from Step 5 and wue is the linear-elastic roof drift determined as:
wue =
Fwu
hw K wi
179
(8.20)
with the initial lateral stiffness of the coupled wall system Kwi determined from a linearelastic analysis of the structure as described in Section 8.1.
8.3
Systems with Precast Concrete Walls
The nonlinear base shear force versus roof drift (F-) behavior of an unbonded
post-tensioned hybrid coupled wall system with precast concrete walls is idealized using
a trilinear relationship. As an example, Fig. 8.6 shows the trilinear F- relationship
estimated for the prototype PRE-UPT system in Chapter 6. The idealized base shear
versus roof drift relationship is identified by the coupled wall softening state (at Fws, ws),
the wall PT-yielding state (at Fwy, wy), and the coupled wall ultimate state (at Fwu, wu).
A procedure to estimate Fws, ws, Fwy, wy, Fwu, and wu is developed below using basic
principles of equilibrium, compatibility, and constitutive relationships. This estimation
procedure can be used as a tool in the seismic analysis and design of coupled wall
structural systems, without the need to develop nonlinear fiber element models.
4000
fiber element model

trilinear estimation
PRE-UPT system
(Fwu,wu)
(Fws,ws)
(Fwy,wy)

tension-side wall concrete linear limit
comp.-side wall concrete crushing
estimation points
1
2
Fig. 8.6 Idealized base shear force versus roof drift relationship
for prototype PRE-UPT system
8.3.1
Coupled Wall Softening State
The coupled wall softening state is defined as the state corresponding to a

significant reduction in the lateral stiffness of the coupled wall system. For a system with
precast concrete walls, the base shear Fws and roof drift ws at this state are estimated
using the following assumptions:
(1) Assumptions (1)-(4) from the estimation of the coupled wall softening state for
a system with cast-in-place concrete walls, described in Section 8.2.1, are also valid for a
180
system with precast concrete walls. Assumptions (1) and (4) are justified based on the
and + markers in Fig. 7.2b, respectively.
(2) The stresses in the wall post-tensioning bars are equal to the initial stress fwpi.
This assumption is justified based on Fig. 7.4b.
(3) The maximum confined concrete compression stress at the base of the tensionside wall is equal to the linear limit stress of the confined concrete, fcl. This assumption is
justified based on the marker in Fig. 7.2b.
As an example, Fig. 8.7a shows a free body diagram to determine the coupled wall
softening state for the prototype PRE-UPT system described in Chapter 6. Note that the
axial forces and moments developed at the beam ends are not shown in the free body
diagram. The assumed concrete stress and strain distributions for the tension-side and
compression-side walls are shown in Fig. 8.7b. The steps used in the estimation of Fws
and ws are presented below:
Step 1
Determine the axial forces at the bases of the tension-side and compression-side
N tws = G 1.05 Vb, ws + Ptws
(8.21)
N cws = G + 1.05 Vb, ws + Pcws
(8.22)
where, G represents the gravity loads applied at the floor and roof levels,
Ptws=Pcws=awpfwpi is the total initial post-tensioning force in each wall (awp is the area of
a post-tensioning bar), and Vb,ws represents the shear forces in the coupling beams (see
Fig. 8.7a). The coupling beam shear forces are calculated as Vb,ws=Vb,sof (see Fig. 8.3b)
corresponding to the beam softening state for an isolated coupled wall subassemblage
from Chapter 5 [see Equation (5.10)]. The 1.05 factor in Equations (8.21) and (8.22) is
described in Section 8.2.1.
Step 2
Determine the neutral axis (i.e., contact) depth at the base of the tension-side wall
(i.e., the depth over which the wall is in contact with the foundation) as:
ctws =
2 N tws
f cl t w
(8.23)
For the parametric walls in Chapter 7, fcl=0.5fcc is assumed, where fcc is the maximum
compressive strength of the confined concrete.
181
Ptws
Pcws
PRE-UPT system
G
53
26
1.05Vb,ws
50
23
tension-side wall
strain distribution
1.05Vb,ws
tension
resultant
lateral force, Fws
20
47
1.05Vb,ws
compression
c tws
17
lw
44
1.05Vb,ws
14
hw
41
tension-side wall
concrete stresses
1.05Vb,ws
11
f cl
N tws
l w /2
38
ctws /3
1.05Vb,ws
strain distribution
35
1.05Vb,ws
5
tension
compression
32
lw/2
1.05Vb,ws
1
28
compression-side
wall concrete stresses
N tws
N cws
M tws
M cws
lw
lb
N cws
l w/3
lw
(a)
(b)
Fig. 8.7 Coupled wall softening state for systems with precast concrete walls:
(a) free body diagram; (b) concrete strain and stress distributions at wall bases
Step 3
From Fig. 8.7b, determine the moments at the bases of the tension-side and
compression-side walls by taking moments about the wall centerlines using Equations
(8.24) and (8.25), respectively, as:
M tws = N tws (
182
lw ctws
)
2
3
(8.24)
M cws = N cws
lw
3
(8.25)
Step 4
Using the free body diagram in Fig. 8.7a, determine the total moment at the base of
M ws = M tws + M cws + 1.05 (lw + lb ) Vb, ws
(8.26)
Fws =
M ws
H
(8.27)
Step 5
Estimate the coupled wall roof drift at the wall softening state, ws using the initial
lateral stiffness, Kwi from a linear-elastic analysis of the structure (see Section 8.1) as:
ws =
Fws
hw K wi
(8.28)
where, hw is the wall height.

8.3.2
Coupled Wall PT-Yielding State
The coupled wall PT-yielding state for a system with precast concrete walls is
defined as the state when the first wall pier post-tensioning bar yields. The base shear Fwy
and roof drift wy at this state are estimated using an iterative procedure with the
following assumptions:
(1) The maximum stress in the wall pier post-tensioning bars is equal to the yield
stress, fwpy (corresponding to the linear limit strain of the post-tensioning steel, see
Chapter 6). This assumption is based on the definition of the coupled wall PT yielding
state above.
(2) The stresses in the wall pier post-tensioning bars located within the contact
depth at the bases of the tension and compression-side walls are equal to the initial
stresses, fwpi (i.e., the stresses just before the application of the lateral forces). The results
in Fig. 7.4b indicate that this assumption is in general valid.
(3) The rotations and lateral displacements of the tension-side and compressionside walls are the same. This assumption is justified based on Fig. 7.3b.
183
The steps used in the estimation of Fwy and wy are presented in the following. Note
that the differences between the unconfined cover concrete and the confined cover
concrete are ignored in the procedure below. A revised procedure can be developed by
considering the crushing of the unconfined cover concrete in the proposed formulation.
Step 1
Estimate the elongation (due to the lateral displacements of the walls) of the wall
pier post-tensioning bar that yields (usually the extreme bar on the tension side of the
structure) as:
uwpy =
( f wpy f wpi ) lwpu

Ewp
(8.29)
where, lwpu=hw is the unbonded length and Ewp is the Youngs modulus of the wall pier
post-tensioning bars.
Step 2
Select initial values for the neutral axis (i.e., contact) depths at the bases of the
tension-side and compression-side walls, ctwy and ccwy, respectively, when the coupled
wall PT-yield state is reached. Initial values of ctwy=ctws and ccwy=ccws=lw/2 may be
selected.
Step 3
Estimate the coupled wall roof drift at the coupled wall PT-yielding state as:
wy =
uwpy
d wp1 ctwy
(8.30)
where, dwp1 is the distance of the first post-tensioning bar that yields from the
compression corner of the tension-side wall as shown in Fig. 8.8a and uwpy and ctwy are
from Steps 1 and 2, respectively.
Step 4
Based on the idealized coupled wall displaced shape in Fig. 8.8a (see also Figs.
8.4b and 8.4c), estimate the chord rotation in the coupling beams as:
b, wy = wy 1 +
lw ccwy
lb
(8.31)
Note that the transverse (i.e., vertical) displacements at the ends of the coupling
beams supported by the tension-side wall are ignored in the estimation of b,wy in
Equation (8.31).
184
tension coupling
side wall
beam
chord
compression
side wall
wy
b,wy
l wpu=hw
first wall PT
bar that yields
wy
wy(l w -ccwy )/lb

wy
uwpy
wy
wy
dwp1
lw
ctwy
ccwy
lb
lw
(a)
3000
(Vb,pty,b,pty)
V b,wy
(Vb,sof, b,sof)
prototype
subassemblage
b,wy
(b)
Fig. 8.8 Coupled wall PT-yielding state for systems with precast concrete walls:
(a) idealized coupled wall displaced shape; (b) beam shear forces
Step 5
Estimate the shear force in the coupling beams Vb,wy corresponding to the beam
rotation b,wy using the idealized subassemblage shear force versus rotation relationship
described in Chapter 5 (see Fig. 8.8b).
185
The third term on the right hand side of Equation (8.33) for the compression-side
wall post-tensioning bars accounts for the shortening of the post-tensioning bars due to
the coupling forces transferred to the wall pier at each floor and roof level (see the free
body diagram of the coupled wall system at the wall PT-yielding state in Fig. 8.9a). Note
that a similar term to account for the elongation of the tension-side wall post-tensioning
bars due to the coupling forces is not included in Equation (8.32) in order to determine a
conservative (i.e., lower) value for the stresses in the post-tensioning bars.
Step 6
Estimate the stress of each post-tensioning bar in the tension-side and compressionside walls using Equations (8.32) and (8.33), respectively, as:
ftwp = f wpi +
f cwp = f wpi +
Ewp wy (d wp ctwy )
lwpu
Ewp wy (d wp ccwy )
lwpu
1.05Vb, wy Ewp hb
lwpu ( Awc Ec + Awp Ewp )
(8.32)
(8.33)
where, dwp is the distance of each bar from the compression corner of the wall pier, Awc is
the net concrete area of the wall pier cross section, Ec is the Youngs modulus of concrete,
and Awp=awp is the total area of the post-tensioning steel in the wall pier cross section.
Note that fwpyftwpfwpi and fwpy>fcwpfwpi. In Equation (8.33), hb is equal to the height of
each coupling beam from the base of the wall, with the term hb calculated as the sum of
the hb values for the coupling beams over the height of the structure.
Step 7
Determine the axial forces at the bases of the tension-side and compression-side
walls, Ntwy and Ncwy, respectively, as:
N twy = G 1.05 Vb, wy + Ptwy
(8.34)
N cwy = G + 1.05 Vb, wy + Pcwy
(8.35)
based on the free body diagram of the coupled wall system at the PT-yielding state as
shown in Fig. 8.9a. Note that the axial forces and moments developed at the beam ends
are not shown in Fig. 8.9a. In Equations (8.34) and (8.35), Ptwy=awpftwp and
Pcwy=awpfcwp are the total forces in the post-tensioning bars of each wall pier, G
represents the gravity loads applied at the floor and roof levels, and Vb,wy represents the
shear forces in the coupling beams. The 1.05 factor is described in Section 8.2.1.
186
Ptwy
PRE-UPT system
G
f
Pcwy
G
53
26
1.05Vb,wy
50
23
1.05Vb,wy
resultant
lateral force, Fwy
47
20
1.05Vb,wy
17
44
1.05Vb,wy
14
41
hw
1.05Vb,wy
11
38
1.05Vb,wy
8
35
1.05Vb,wy
5
32
1.05Vb,wy
28
N twy
N cwy
M cwy
M twy
lw
lb
lw
(a)
tension-side wall
c twy
c cwy
N twy
Note:
wall PT bars not shown
Note:
wall PT bars not shown
twc twy/2
N cwy c /2
cw cwy
0.5l w
0.5l w
tension-side wall
twc twy
cwc cwy
cwfcc
tw fcc
N twy
lw
N cwy
cwc cwy /2
lw
twc twy /2
(b)
(c)
Fig. 8.9 Coupled wall PT-yielding state for systems with precast concrete walls:
(a) free body diagram; (b) concrete stress distributions at base of tension-side wall;
(c) concrete stress distributions at base of compression-side wall
187
Step 8
Estimate the neutral axis depths at the bases of the tension-side and compressionside walls using assumed equivalent rectangular compression stress distributions as:
ctwy =
ccwy =
N twy
tw tw f cctwc
N cwy
cw cw f cctwc
(8.36)
(8.37)
where, tw and tw are confined concrete compression stress block parameters (Mander et
al. 1988, Paulay and Priestley 1992) for the tension-side wall, and cw and cw are
confined concrete compression stress block parameters for the compression-side wall. It
is assumed that significant nonlinear behavior (but no crushing) occurs in the confined
concrete at the bases of the tension-side and compression-side walls when the coupled
wall PT-yielding state is reached. Figs. 8.9b and 8.9c show the smooth compression
stress distributions and the assumed equivalent rectangular stress distributions for the
tension-side and compression-side walls, respectively. For the parametric coupled wall
structures investigated in this chapter, confined concrete compression stress block
parameters of tw=0.68, tw=0.73, and cw=0.87, cw=0.83 were used.
Step 9
Repeat Steps 2-8 using the values of ctwy and ccwy from Step 8 in Step 2 until a
satisfactory agreement between the ctwy and ccwy values from Steps 2 and 8 is achieved.
Step 10
Determine the moments at the bases of the tension-side and compression-side walls,
Mtwy and Mcwy, respectively, by taking moments about the wall pier centerlines as:
M twy = 0.5N twy (lw twctwy ) + [awp ftwp (d wp 0.5lw )]
(8.38)
M cwy = 0.5N cwy (lw cwccwy ) + [awp f cwp (d wp 0.5lw )]
(8.39)
where, dwp is the distance of each post-tensioning bar from the compression corner of the
wall pier. Note that the wall pier post-tensioning bars are not shown in Figs. 8.9b and
8.9c for clarity.
Step 11
Using the free body diagram in Fig. 8.9a, estimate the total moment at the base of
M wy = M twy + M cwy + 1.05 (lw + lb ) Vb, wy
188
(8.40)
Fwy =
M wy
H
(8.41)
Based on the multi-story coupled wall analyses in Chapter 7, second order effects
(often referred to as P- effects) in the tension-side and compression-side walls have a
small influence on the base shear force versus roof drift relationship of the structure, and
thus, can be neglected. Note that the second order effects in the coupling beams and the
beam post-tensioning tendons are included in the calculation of Vb,wy as described in
Chapter 5. Note also that a consideration of second order effects is not necessary in the
estimation of the coupled wall softening state since the displacements of the structure at
the wall softening state are relatively small.
8.3.3
Coupled Wall Ultimate State
The coupled wall ultimate state is defined as the state corresponding to the crushing
of the confined concrete at the base of the compression-side wall. For a system with
precast concrete walls, the base shear Fwu and roof drift wu at this state are estimated
using the following assumptions:
(1) Assumptions (1), (2), (3), and (5) from the estimation of the coupled wall
ultimate state for systems with cast-in-place concrete walls are also valid for a system
with precast concrete walls.
(2) The neutral axis depth at the base of the compression-side wall at the coupled
wall ultimate state is the same as the neutral axis depth at the coupled wall PT-yield state
(i.e., ccwu=ccwy). This assumption is justified based on Fig. 7.9b.
As an example, Fig. 8.10 shows the assumed strain and stress distributions at the
base of the compression-side wall when the ultimate state of the prototype PRE-UPT
system is reached. The steps used in the estimation of Fwu and wu are presented in the
following. Note that the differences between the unconfined cover concrete and the
confined concrete are ignored (as shown in Fig. 8.10) in the procedure below. A revised
procedure can be developed by considering the crushing of the unconfined cover concrete
in the proposed formulation.
Step 1
Estimate the curvature at the base of the compression-side wall as:
cwu =
ccu
ccwu
189
(8.42)
strain distribution
cwu
tension
compression
ccu
ccwu
lw
tension
concrete stresses
compression
Fig. 8.10 Coupled wall ultimate state for systems with precast concrete walls strain
and stress distributions at base of compression-side wall
Step 2
Estimate the coupled wall roof drift due to gap opening at the base of the
wug = 1.5cwu hwp
(8.43)
where, hwp=twc. This equation was determined based on the results obtained from the
parametric coupled wall analyses in Chapter 7.
Step 3
Estimate the coupled wall base shear force at the coupled wall ultimate state as:
Fwu = Fwy + 0.5 wug wy
) Fwy Fws
wy
(8.44)
ws
Equation (8.44) assumes that the slope of the idealized base shear force versus roof drift
relationship from the coupled wall PT yielding state to the coupled wall ultimate state is
one half of the slope from the coupled wall softening state to coupled wall PT yielding
state.
Step 4
Estimate the coupled wall roof drift as:
wu = wug + wue
190
(8.45)
where, wug is from Step 2 and wue is the linear-elastic roof drift determined as:
wue =
Fwu
hw K wi
(8.46)
with the initial lateral stiffness of the coupled wall system Kwi determined from a linearelastic analysis of the structure as described in Section 8.1.
8.4
Verification of Idealized Relationships
The idealized F- relationships described above are verified by comparing the

estimated base shear force and roof drift values with values determined using the fiber
element model from Chapter 6. A Matlab (2000) algorithm was developed to carry out
the estimation procedures. Fig. 8.11 shows the comparisons for Fws and Fwu, and Fig. 8.12
shows the comparisons for ws and wu for the parametric coupled wall systems with
cast-in-place concrete walls in Fig. 7.27. Similarly, Fig. 8.13 shows the comparisons for
Fws, Fwy, and Fwu, and Fig. 8.14 shows the comparisons for ws, wy, and wu for the
parametric coupled wall systems with precast concrete walls in Fig. 7.28.
The Fwu and wu values from the fiber element model correspond to the crushing of
the confined concrete at the base of the compression-side wall (based on the definition of
the wall ultimate state) as indicated by the markers in Figs. 7.27 and 7.28. Similarly,
the Fwy and wy values correspond to the initiation of yielding in the wall post-tensioning
bars (based on the definition of the wall PT-yielding state) as indicated by the markers
in Fig. 7.28.
The Fws and ws values from the fiber element model correspond to a significant
reduction in the lateral stiffness of the coupled wall system. This state was determined
based on the assumptions used in the estimation of the coupled wall softening state as
follows. For each coupled wall system, the base shear corresponding to three states were
determined using the fiber element model. The Fws value was taken as the average of the
maximum and minimum base shear forces from these three states. The ws value was
taken as the roof drift corresponding to Fws. Figs. 8.15a and 8.15b show the procedure to
determine Fws and ws for the prototype CIP-UPT and PRE-UPT systems, respectively.
For coupled wall systems with cast-in-place walls, Fws values for the following
states were used in accordance with Assumptions (1), (4), and (5) from the estimation of
the coupled wall softening state: (1) softening of the compression-side wall, which is
reached when the neutral axis at the base of the wall reaches the centerline of the wall (+
marker); (2) first yielding of the wall pier steel, which is reached when the extreme
reinforcement in the tension-side wall yields ( marker); and (3) softening of the coupling
beams, which is reached when the average coupling beam shear force in the structure
(i.e., the sum of the shear forces in the coupling beams divided by the number of beams)
is equal to 1.05Vb,sof, where Vb,sof is the softening state coupling shear force for an
isolated floor-level coupled wall subassemblage (see Chapters 4 and 5) and the factor
1.05 is described previously.
191
4500
4500
CIP-UPT system
CIP-UPT system
Fwu
Fws
fiber element model
estimation
4.0
4500
4500
fiber element model

estimation
wall reinf. ratio, (percent ) 2.5
ws
CIP-UPT system
Fwu
Fws
fiber element model
estimation
wall thickness, tw(mm)
500
4500
CIP-UPT system
Fwu
Fws
fiber element model
estimation
beam PT tendon area, abt (mm2 ) 700
4500
CIP-UPT system
Fwu
Fws
fiber element model
estimation
Fwu
Fws
fiber element model
estimation
beam web depth, dbw(mm)
700
CIP-UPT system
Fwu
Fws
fiber element model
estimation
0 initial beam PT stress, f /f

0.8
bpi bpu

4500
Fws
4500
CIP-UPT system
Fwu
5.0
CIP-UPT system
Fwu
Fws
fiber element model
estimation
angle thickness, t a (mm)
30
Fig. 8.11 Verification of idealized relationship for systems with cast-in-place walls
Fws and Fwu
192
4.5
fiber element model

estimation
4.5
wu
CIP-UPT system
ws
4.0
4.5
wu
CIP-UPT system
ws
wall thickness, tw(mm)
500
ws
wall reinf. ratio, (percent) 2.5
ws
fiber element model

estimation
wu
CIP-UPT system
ws
beam web depth, dbw(mm) 700
wu
CIP-UPT system
ws
beam PT tendon area, a bp(mm2 ) 700
fiber element model

estimation
wu
CIP-UPT system
ws

0.8
bpi bpu
4.5
fiber element model

estimation
CIP-UPT system
4.5
fiber element model

estimation
4.5
wu
4.5
fiber element model

estimation
4.5
fiber element model

estimation
wu
CIP-UPT system
ws
5.0
wu
CIP-UPT system
fiber element model

estimation
ws
30
Fig. 8.12 Verification of idealized relationship for systems with cast-in-place walls
ws and wu
193
4500
Fwy
Fws
fiber element model

estimation
0
Fwy
Fws
fiber element model
estimation
initial wall PT stress, fwpi/fwpu 0.8
Fwy
Fws
fiber element model
estimation
Fwu
700
Fwy
Fws
fiber element model

estimation
Fwu
PRE-UPT system
Fwu
4500
PRE-UPT system
Fwy
Fws
fiber element model
estimation
0.8
bpi bpu
beam PT tendon area, abp(mm2) 700
4500
4500
PRE-UPT system
wall reinf. ratio, wp(percent) 2.5
PRE-UPT system
4500
Fwu
Fwy
Fws
fiber element model
estimation
0
fiber element model

estimation
Fwu
PRE-UPT system
Fws
4500
4500
Fwy
4.0
Fwu
PRE-UPT system
PRE-UPT system
4500
Fwu
beam length, l b(m)
5.0
PRE-UPT system
Fwu
Fwy
Fws
fiber element model
estimation
angle thickness, ta (mm)
30
Fig. 8.13 Verification of idealized relationship for systems with precast walls
Fws, Fwy, and Fwu
194
4.5
fiber element model

estimation
4.5
wu
wy
PRE-UPT system
ws
4.0
wall length, l w(m)
4.5
4.5
fiber element model

estimation
wu
wy
PRE-UPT system
ws
initial wall PT stress, fwpi/fwpu 0.8
4.5
fiber element model

estimation
4.5
wu
wy
PRE-UPT system
ws
beam PT tendon area, abp(mm 2 ) 700
4.5
fiber element model

estimation
4.5
wu
wy
PRE-UPT system
ws
beam length, lb (m)
5.0
fiber element model

estimation
wu
wy
ws
PRE-UPT system
wall reinf. ratio, wp(percent) 2.5
fiber element model

estimation
wu
PRE-UPT system
wy
ws
700
fiber element model

estimation
wu
wy
PRE-UPT system
ws
initial beam PT stress, fbpi /f bpu 0.8
fiber element model

estimation
wu
wy
PRE-UPT system
ws
30
Fig. 8.14 Verification of idealized relationship for systems with precast walls
ws, wy, and wu
195
4000
CIP-UPT system
Fws
Fws is equal to the average of the

maximum and minimum base
shear forces from these three states

ws
1
2
(a)
4000
PRE-UPT system
Fws
Fws is equal to the average of the
maximum and minimum base
shear forces from these three states

ws
1
2
(b)
Fig. 8.15 Determination of coupled wall softening state in fiber element model:
(a) systems with cast-in-place concrete walls; (b) systems with precast concrete walls
For coupled wall systems with precast concrete walls, the first two states are the
same as the first and third states listed for systems with cast-in-place walls above. In
accordance with Assumption (3) from the estimation of the coupled wall softening state
196
for systems with precast concrete walls, the third state was taken as the state when the
maximum confined concrete compression stress at the base of the tension-side wall
reaches the linear limit stress of the confined concrete, fcl=0.5fcc ( marker).
The comparisons in Figs. 8.11 through 8.14 show that the estimated coupled wall
base shear force and roof drift values are close to the values determined using the fiber
element model for a wide range of parameters. The wall-contact elements and the beam
flange cover plates were ignored in the estimation of Kwi (using the model in Fig. 8.1).
The maximum error in the estimation of Kwi is less than 5%, the maximum error in the
estimation of Mws, Mwy, and Mwu is less than 7.5% and the maximum error in the
estimation of wy and wu is less than 15%. Thus, it is concluded that the estimation
procedures described above can be used to conduct approximate analyses of unbonded
post-tensioned hybrid coupled wall systems with different properties.
These approximate analyses can be used for two purposes: (1) to evaluate the
behavior of a given system under lateral loads; and (2) as part of a design approach that
requires the estimation of the nonlinear lateral load versus displacement relationship of
the structure. Note that while the proposed analysis procedures involve a significant
number of steps (some of which are iterative), the calculations can be coded into a
computer algorithm, for example using Matlab (2000). This allows for push-over
analyses of multi-story coupled wall systems in seismic evaluation/design, without the
need to develop a sophisticated analytical model (e.g., the fiber element model). A
performance-based seismic design approach using the above estimation procedures is
described in Chapters 11 and 12.
8.5
Chapter Summary
This chapter describes analytical procedures to estimate the linear and nonlinear
lateral load versus displacement behavior of multi-story unbonded post-tensioned hybrid
coupled wall structures by quantifying selected limit states for the walls. The estimation
procedures, which are verified with results from the DRAIN-2DX program (Prakash et al.
1993), can be used to conduct approximate analyses of coupled wall systems with cast-inplace reinforced concrete and precast concrete walls as part of a performance-based
seismic evaluation and design approach as demonstrated in Chapters 11 and 12.
197
CHAPTER 9
SUBASSEMBLAGE EXPERIMENTS
This chapter describes the results from eleven half-scale experiments to investigate
the nonlinear behavior of unbonded post-tensioned hybrid coupled wall subassemblages.
Each test specimen includes a steel coupling beam and the adjacent concrete wall regions
at a floor level. The experiments were conducted in the Structural Systems Laboratory at
the University of Notre Dame by two graduate students, Michael A. May and Brad D.
Weldon, as part of their research towards a Master of Science Degree in Civil
Engineering. A large portion of the material presented in this chapter is based on their
work (May et al. 2006).
The chapter is divided into the three main sections: (1) experimental program; (2)
results of experiments; and (3) recommendations for application. The results from the
experimental investigation described in this chapter are used in the remainder of this
report as follows. Chapter 10 provides comparisons between the test results and
analytical predictions and makes revisions to the analytical models developed previously
in the report. Then, Chapters 11 and 12 propose a design approach to ensure the desirable
behavior of the proposed coupling system and describe the design of two prototype
coupled wall systems. Finally, Chapters 13 through 15 evaluate the seismic response of
the prototype structures based on nonlinear static and nonlinear dynamic analyses under
earthquake-induced loads, including a critical evaluation of the proposed design approach.
9.1
Experimental Program
Fig. 9.1 shows a photograph of the half-scale test set-up. Eleven specimens were
tested (Tests 1-11) with the following parameters: beam cover plate, beam posttensioning steel area, angle leg thickness, angle vertical leg gage length, and beam depth.
Test 3 was designed to represent a half-scale model of the prototype subassemblage
described in Chapter 3. The other test specimens were determined based on parameter
variations from Test 3 and were not necessarily designed for all of the anticipated forces
and deformations. This allowed the experimental program to demonstrate a number of
critical failure modes for the proposed system. More details on the test specimens are
provided below, with variations between the specimens summarized in Table 9.1.
9.1.1
Test Set-Up
Figs. 9.2a and 9.2b show the side view and end view, respectively, of the test set-up,
which includes two reinforced concrete wall regions placed at each end of a steel
coupling beam. The left (south) wall region, referred to as the reaction block, is fixed to
198
a 1524 mm (60 in.) thick strong floor in the laboratory. The right (north) wall region,
referred to as the loading block, is connected to two hydraulic actuators hanging from a
stiff steel loading frame. The actuators are used in displacement control to move the
loading block up or down without rotating it; thus, modeling the idealized displaced
shape of the subassemblage with respect to the reference line in Fig. 1.3b.
The loading block is free to move in the horizontal (north-south) direction. Out-ofplane displacements are not modeled in the test program. An inner steel bracing frame
(see Fig. 9.3, not shown in Fig. 9.2) is used to prevent out-of-plane movement of the
loading block at three points, and of the coupling beam at the midspan on each side. At
each bracing point, steel bracing plates are adjusted up against corresponding bracing
plates on the coupling beam and the loading block. Grease is applied between the
matching bracing plates to allow in-plane movements of the beam and the loading block,
while out-of-plane displacements are prevented.
SOUTH
NORTH
Fig. 9.1 Photograph of test set-up
199
TABLE 9.1
SUBASSEMBLAGE TEST SPECIMENS
Loading aCoupling
Cover
Type
Beam plates (mm)
Test
No.
Test
Date
3/13/02
Cyclic
W10x68
3/18/02
Cyclic
W10x68
4/8/02
Cyclic
W10x68
7/18/02 Monot.
W10x68
2/12/03
Cyclic
W10x68
3/12/03 Monot.
W10x68
5/15/03
Cyclic
W10x68
5/29/03 Monot.
W10x68
6/5/03
Cyclic
W10x68
10 11/14/03 Cyclic
W14x99
11 11/25/03 Cyclic
W14x99
203x124
x19.1
203x124
x19.1
203x124
x19.1
203x124
x19.1
Angles
lgv dabp Pbi

(mm) (mm2) Pbu
Pby
Pbi
A bfcfby
Pbm
A bfcfby
A bfc fbm
Test
Parameter
560
0.55
0.25
0.37
0.30
No angles
560
0.54
0.24
0.38
0.30
Retest
beam
560
0.52
0.23
0.35
0.30
Base test
Angle test
560
0.53
0.47
0.70
0.60
No cover
plates
Angle test
560
0.56
0.50
0.70
0.60
Thicker
angles
Angle test
840
0.55
0.74
0.96
0.90
560
0.52
0.46
0.68
0.60
Four
95.3
L8x4x1/2
Two
95.3
L8x4x1/2
Four
114
L8x4x1/2
Two
114
L8x4x1/2
Four
114
L8x4x5/8
Two
114
L8x4x5/8
Four
114
L8x4x5/8
Four
138
L8x4x5/8
Four
138
L8x4x5/8
Larger PT
area
Larger
beam depth
Angle test
Notes:
a
U.S. shape. Coupling beam flanges were saw cut to a width of 159 mm to satisfy similitude requirements.
b
U.S. shape.
c
Angle vertical leg gage length measured from center of innermost angle-to-wall connectors to heel of
angle.
d
abp=total area of coupling beam post-tensioning steel.
e
Pbi=total initial force measured in coupling beam post-tensioning strands; Pbu= abpfbpu, where fbpu=design
maximum strength of post-tensioning strands (1862 MPa).
f
Abfc=cross section area of one beam flange, Abf plus cover plate, Ac (if any); fby= measured yield strength
of coupling beam steel (386 MPa), where, difference between measured yield strengths of beam steel and
cover plate steel in Tests 1-3 is ignored.
g
Pbm=maximum total force measured in coupling beam post-tensioning strands.
h
Pby= abpfbpy, where fbpy=design yield strength corresponding to linear limit stress (i.e., limit of
proportionality) of beam post-tensioning strand stress-strain relationship (1689 MPa); fbm=measured
maximum strength of coupling beam steel (524 MPa), where, difference between measured maximum
strengths of beam steel and cover plate steel in Tests 1-3 is ignored.
Note that the subassemblage test set-up in Fig. 9.2 does not model the wall pier
shear forces that develop in a multi-story structure, and thus, does not capture the effect
of the wall piers on the coupling beam axial forces during the application of posttensioning as well as the application of lateral loads on the structure. As discussed in
Chapter 7, the effect of the wall piers on the coupling beam axial forces may be
significant in the lower floor beams in a multi-story structure; however, they are
negligible for the coupling beams in the upper floor and roof levels.
200
ACTUATOR BEAM
W30X326
END BEAM W33X118
END BEAM W33X118
SIDE
BEAM
ACTUATOR
ACTUATOR
ANGLE
L5X5X3/4
WALL TEST
REGION
REACTION BLOCK
NORTH
ANGLE
L5X5X3/4
ANGLE L7X4X3/4
CONNECTION
BEAM
W14X283
COUPLING
BEAM
CROSS BRACING
C12X25
PT STRAND
DEAD END
ANCHORS
COLUMN W12X87
LIVE END
ANCHORS
PT STRAND
ANGLE L7X4X3/4 LOADING BLOCK
TEST FLOOR
3988 mm
(a)
WEST
51 mm
STIFFENER
823 mm
ACTUATOR BEAM W30X326
ACTUATOR
2223 mm
COLUMN W12X87
END BEAM
W33X118
BRACE C12X25
COUPLING
BEAM
LOADING BLOCK
610 mm
914 mm
424 mm
51 mm
CONNECT ION
BEAM
W14X283
COLUMN
W12X87
TEST
FLOOR
2997 mm
(b)
Fig. 9.2 Elevation of experimental set-up: (a) side view; (b) end view
201
loading block
out-of-plane support
NORTH
coupling beam
out of plane support
W12x87
1981 mm
loading
block
reaction block
W12x87
coupling
beam
W12x87
616 mm
TEST
FLOOR
loading block
out-of-plane supports
2159 mm
connection beam
W14x283
W12x87
3000 mm
Fig. 9.3 Bracing frame side view

9.1.2
Loading Block and Reaction Block
Fig. 9.4 shows a more detailed view of the test structure. The horizontal lengths of
the loading and reaction blocks are equal to lw=1524 mm (60 in.; one half the length of
the full-scale prototype wall described in Chapter 3). The loading block is used to
transmit the forces from the actuators to the coupling beam and is connected to the
actuators via two W14283 (see Fig. 9.2) steel connection beams bolted to the actuators.
The loading block has a uniform width (thickness) of 889 mm (35 in.) along its length
(see Fig. 9.5) in order to accommodate the connections to the two actuators and to
prevent damage to the block during testing.
A total of twenty-eight 15.2 mm (0.6 in.) diameter post-tensioning strands, each
prestressed to approximately 89 kN, are used to secure the loading block to the two
connection beams. The strands run through 25.4 mm (1 in.) diameter vertical ducts in the
loading block and are anchored to the bottom flanges of the connection beams (dead end)
and to four 1524 mm by 102 mm by 19 mm (60 in. by 4 in. by 3/4 in.) plates embedded
to the bottom of the loading block (live end). In addition to the strands, eight high
strength steel threaded rods prestressed to approximately 156 kN each, are used near the
coupling beam end of the loading block to increase the clamping force between the block
and the connection beams.
Unlike the loading block, the reaction block does not have a uniform thickness (see
Fig. 9.6). Adjacent to the coupling beam, the width of the reaction block is 191 mm (7.5
in.), modeling the thickness of the prototype wall in Chapter 3. This region of the reaction
block, referred to as the wall test region, is 762 mm (30 in.) long and 1219 mm (48 in.)
high. Note that half-scale similitude requires a wall thickness of tw=178 mm (7 in.);
however, a slightly larger thickness was necessary to accommodate the placement of
oversized post-tensioning ducts in the block with adequate concrete cover and
reinforcement. The other regions of the reaction block are 1473 mm (58 in.) wide to
provide lateral stability to the block and to allow tie down to the laboratory strong floor
202
[at a spacing of 1000 mm (39.4 in.) between four groups of four anchors]. Fig. 9.7 shows
a photograph of the steel form used to cast the reaction block.
N
steel loading frame
actuator
spreader beam
anchor plate
838x508x25mm wall test
region
REACTION BLOCK
actuator
steel connection beam
shim plate 254x191x13mm

L8x4x1/2
22mm A490 bolt
W10x68
LOADING BLOCK
(thickness=0.89m)
15mm
PT strand
anchor plate
838x508x25mm
15mm beam PT strand

cover plate 203x124x19mm
15mm
PT strand
25mm
tie-down bar
1.0m
l w= 1.52m
embedded plate
838x178x16mm
strong floor
embedded plate 15mm

15mm
PT strand
838x229x38mm PT strand
2
a bp=140mm
four embedded plates
fbpi= 0.52fbpu
tsh=12.7mm
1524x102x19mm
tsh=12.7mm
l b = 1.47m
l w = 1.52m
Fig. 9.4 Test structure

9.1.3
Coupling Beams
Steel U.S. wide-flange shapes (Table 9.1) are used for the coupling beams. The
beam length is equal to lb1473 mm (58 in.; one half the full-scale prototype beam
length). In order to satisfy half-scale similitude requirements with the prototype beam
(with a W21182 cross-section), the flanges of the test beams were saw cut to a width of
bbf = 159 mm (6.25 in.; one half the full-scale flange width). Figs. 9.8 and 9.9 show the
beam-end-view and beam-side-view, respectively, of the test set-up near the wall test
region. The ends of the beams were trimmed (after welding of the flange cover plates, if
any) using a saw cutter to achieve even contact surfaces at the beam-to-wall interfaces.
203
838x508x25 mm
anchor plate
838x229x38 mm
embedded plate
spiral
reinforcement
1524x102x19 mm
embedded plates (4)
25 mm diameter
connection duct
914 mm NORTH
END
NORTH END
1524 mm
60 mm OD duct for
beam PT strand
25 mm diameter duct
for angle-to-wall
connection PT strand
25 mm diameter duct
for angle-to-wall
(b)
838x229x38 mm
embedded plate
838x229x38 mm
embedded plate
914 mm
60 mm OD duct for
beam PT strand
SOUTH END
EAST
SIDE
spiral
reinforcement
(a)
889 mm
(c)
Fig. 9.5 Loading block: (a) 3D view; (b) side view; (c) south end view
762 mm
spiral
reinforcement
NORTH
END
1219 mm
1676 mm
SOUTH
END
1.47m
838x508x25 mm
embedded plate
60mm diameter
tie-down duct
1.0m
0.19m
838x178x16 mm
embedded plate
457 mm
1524 mm
wall test region
1.68m
(b)
1.22m
60 mm dia.
duct for beam
PT strand
refer to (c) for details

0.64m
0.76m
NORTH
END
1.52m
EAST
SIDE
330 mm
838 mm
WEST
SIDE
0.46m
(a)
spiral
reinforcement
25 mm dia. duct
for angle-to-wall
191 mm
1473 mm
(c)
Fig. 9.6. Reaction block: (a) 3D view; (b) side view; (c) north end view
204
Fig. 9.7 Photograph of steel form for casting of reaction block
WALL TEST REGION
25mm duct for

15mm angle-to-wall
angle vertical leg
spiral
15.9 mm E70
SMAW weld
angle horizontal leg

cover plate
shim plate
EAST
76mm
102mm
76mm
coupling beam W10x68

60mm OD duct for
15mm beam
PT strand
22mm A490 bolt
embedded plate
838x178x16mm
160 mm
191mm
Fig. 9.8 Beam end view

205
WEST
NORTH
1.22m
wall test region

mesh
W4.0xW4.0-4x4
shim plate 254x191x13mm

L8x4x1/2
22mm A490 bolt
W10x68
165mm 165mm
spiral
(diameter=
133mm,
wire
diameter=
10mm,
pitch=
32mm,
bsp=7.66%,
length=
889mm)
15mm beam PT strand
cover plate
203x124x19mm
embedded plate
838x178x16mm
0.76m
Fig. 9.9 Beam side view

9.1.4
Beam Post-Tensioning Strands
Seven-wire strands, each strand with a nominal diameter of dbp=15.2 mm (0.6 in.)
and area of abp=140 mm2 (0.217 in2), are used to apply the beam post-tensioning force.
Six test specimens had two pairs of strands (placed on either side of the coupling beam
web as shown in Fig. 9.9), one specimen had three pairs of strands, and four specimens
had no post-tensioning strands. The resulting total post-tensioning steel area, Abp=abp, in
each specimen is given in Table 9.1. Note that the post-tensioning steel area required for
similitude with the full-scale prototype beam (see Chapter 3) is equal to 630 mm2 (0.977
in2), which is slightly larger than the area provided in Test 3 (560 mm2). Each posttensioning strand is run through an oversized Dywidag Spiro duct [with
60.1 mm (2.37 in.) outside diameter] inside the reaction and loading blocks, and is
anchored to the far ends of the blocks using steel wedge/barrel anchorage systems (Fig.
9.10). The strands are not bonded to the blocks or connected to the beam between the
anchors. Embedded steel anchor plates are used at the far ends of the reaction and loading
blocks (see Fig. 9.4) to distribute the forces from the beam post-tensioning tendons. The
post-tensioning operation is carried out from the north end of the loading block (live end).
206
Fig. 9.10 Photograph of post-tensioning anchorage system

9.1.5
Beam-to-Wall Connection Regions
The beam-to-wall connection regions (see Fig. 9.9) include embedded steel plates,
wire mesh reinforcement, and spiral reinforcement inside the loading and reaction blocks;
and flange cover plates (in Tests 1-4; see Table 9.1), shim plates, and top and seat angles
at the beam ends. Each of these beam-to-wall connection components is described in
more detail below. A photograph of the wall test region (adjacent to the beam-to-wall
interface) before the casting of the concrete is provided in Fig. 9.11.
Fig. 9.11 Photograph of wall test region prior to casting of concrete

Embedded Plates Embedded steel plates, placed flush with the outside of the
reaction and loading blocks during casting (with nominal welded studs), are used to
207
distribute the contact stresses in the concrete. The embedded plate inside the reaction
block has a thickness of te=16.0 mm (0.625 in.; one half the embedded plate thickness in
the full-scale prototype wall) and a width of 178 mm (7 in.). A thicker 38 mm (1.5 in.)
plate is used in the loading block to prevent damage to the block during testing.
Spiral Reinforcement Spiral reinforcement (Fig. 9.12) is used behind the
embedded plates to resist and distribute the compression stresses without excessive
deformations in the concrete. Two spirals are placed horizontally in each of the reaction
and loading blocks, one spiral near each coupling beam flange as shown in Figs. 9.9 and
9.11. The spiral center-to-center diameter Dbsp = 133 mm (5.25 in.), spiral smooth wire
diameter dbsp = 10 mm (0.4 in.), and spiral pitch sbsp = 32 mm (1.25 in.), resulting in a
spiral reinforcement ratio, bsp = 7.66%. Note that this spiral reinforcement ratio is
significantly larger than the value of bsp = 5.57% used in the prototype subassemblage in
Chapter 3. The length of the spiral is 889 mm (35 in.) inside the reaction block and 305
mm (12 in.) inside the loading block. In addition, W4.0W4.0-44 welded wire mesh
[with 5.72 mm (0.22 in.) wire diameter] is used on both faces of the 191 mm (7.5 in.)
thick wall test region of the reaction block, with a concrete cover of 12.7 mm (0.5 in.) to
the center of the wire mesh.
Fig. 9.12 Photograph of spiral reinforcement used inside wall test region
Flange Cover Plates Beam flange cover plates with thickness of tc=19.1 mm (0.75
in.), width of 124 mm (4.875 in.), and length of lc=203 mm (8 in., one half the full-scale
cover plate length) are used in Tests 1-4 to strengthen and stabilize the coupling beam
flanges in compression. As shown in Figs. 9.8 and 9.9, the cover plates are placed flush
with the beam ends and are fillet welded to the flanges using 15.9 mm (0.625 in.) E70
SMAW (Shielded Metal Arc Welding) welds along all edges except at the beam ends.
Shim Plates Previous experiments of post-tensioned steel moment frame systems
(Ricles et al. 2001, 2002) have shown that yielding and permanent deformation of the
beam flanges in compression may result in a loss of contact at the flanges upon unloading
to the zero displacement position, leading to a reduction in the initial lateral stiffness of
the structure under reversed cyclic loading. Shim plates are used between the coupling
beam flanges and the embedded plates in the walls (see Fig. 9.13) in order to force the
bearing between the coupling beam and the walls to occur at the flanges (so that one or
both flanges remain in contact with each wall during the reversed cyclic displacements of
208
the structure). The shim plates are welded to the embedded plates and are terminated at
the beam web (Figs. 9.8 and 9.9), thus preventing contact between the web and the
concrete blocks. The shim plates in the test specimens have a thickness of tsh=12.7 mm
(0.5 in.), width of 191 mm (7.5 in.), and length of 254 mm (10 in.). In practical
applications, shim plates may also be used for construction/tolerance purposes.
Fig. 9.13 Photograph showing shim plates at beam-to-reaction-block interface

Top and Seat Angles Top and seat angles are used at the beam ends to yield and
dissipate energy under cyclic loading. The angles in Tests 3-6 have L841/2 crosssections and the angles in Tests 7-11 have L845/8 cross-sections. The angle length is
equal to the beam flange width of bbf =159 mm (6.25 in). As shown in Figs. 9.8 and 9.9,
the angles are oriented such that the shorter leg, which is parallel to the beam flange, is
bolted to the beam flange using two A490 22.2 mm slip critical bolts (inside oversized
27.0 mm diameter holes) in Tests 3-9 (see Fig. 9.14) and two 25.4 mm slip critical bolts
(inside standard 27.0 mm diameter holes) in Tests 10 and 11. The bolts are installed using
the turn-of-nut pretensioning method described in the Manual of Steel Construction
Load and Resistance Factor Design (AISC 2001).
The connection between the vertical leg of each angle and the reaction block or the
loading block is achieved using two (Tests 6, 10, and 11, see Fig. 9.14) or four (Tests 3-5
and 7-9) 15.2 mm (0.6 in.) diameter seven-wire unbonded post-tensioning strands, each
prestressed to 0.4-0.5 times the design maximum (i.e., peak) strength of the posttensioning steel [fbpu=1862 MPa (270 ksi)]. The connection gage lengths, lgv for the
209
vertical legs of the angles (i.e., the length between the center of the innermost angle-towall connectors and the angle heel) are given in Table 9.1. The purpose of the longer
angle vertical legs is to allow for parameter variations in lgv. Each post-tensioning strand
in an angle-to-wall connection is passed through a 25.4 mm (1 in.) diameter duct and is
unbonded along its length (i.e., the length of the concrete block) to prevent the yielding of
the connection strand as the angles are pulled by the beam during testing. Each strand is
anchored at the ends of the concrete block using steel wedge/barrel anchorage systems,
thus putting the concrete into compression. The post-tensioning operation for the angleto-wall connection strands is carried out from the outer end of each block.
Note that it would also be possible to use bolted or welded angle-to-wall
connections. These connection types were not used in this research to prevent the
development of large tensile stresses in the concrete. Furthermore, yielding and fracture
of the angle-to-wall connection bolts/welds may be possible, thus reducing the ductility
of the connection and preventing the angles from yielding and dissipating energy
effectively.
Fig. 9.14 Photograph showing angle-to-beam and angle-to-wall connections

9.1.6
Loading
Two 979 kN (220 kip) servo-controlled hydraulic actuators are connected to the top
of the loading block (Fig. 9.4) and are operated in displacement control to move the block
up or down at a rate of approximately 12.7 mm per minute. The two actuators are moved
by the same displacement (resulting in actuator forces in opposite directions), thus
restraining the block from rotating in the vertical plane (i.e., the plane of the test set-up)
to achieve the idealized displaced shape in Fig. 1.3c. A two-channel Schenck Pegasus
5910 Digital Controller and a computer are used to send the displacement command
signal to the two actuators near-simultaneously and prevent significant lag between the
actuators.
210
Fig. 9.15 shows the cyclic actuator displacement history in Tests 5, 7, 9, 10 and 11,
which consists of a series of 6-cycle sets with target amplitudes corresponding to beam
chord rotation values (i.e., relative vertical displacement between the coupling beam ends
divided by the beam length) of b=0.125, 0.25, 0.5, and 0.75%, followed by 3-cycle sets
each with amplitudes of b=1, 1.5, 2, 3, 4, 5, 6, 7, and 8%. After 1% rotation, three
smaller cycles with a target amplitude of b=0.25% are imposed between two consecutive
large rotation cycle sets. A slightly different displacement history was used in Tests 1 and
3, consisting of 3-cycle sets with target amplitudes of b=0.10, 0.25, 0.50, 0.75, 1, 2, 3, 4,
5, 6, 7, and 8%. After 1% rotation, one small cycle at b=0.25% was imposed between
two consecutive large rotation cycle sets. Selected displacement cycles from Test 1 were
used in Test 2, consisting of 3-cycle sets with target amplitudes of b=0.10, 0.25, 0.50,
0.75, 1, 2, 4, 6, and 8%.
10
8%
Tests 5, 7, 9, 10, 11 only

6
5
4
0.75 1
0.25 0.5
1.5 2
0.125
0.25
-10
test duration
Fig. 9.15 Displacement history

The beam and angle post-tensioning strands are prestressed using a mono-strand
hydraulic jack (Fig. 9.16). Table 9.1 shows the total initial forces (after short-term losses)
measured in the beam post-tensioning strands, Pbi normalized by the total design
maximum strength Pbu=abpfbpu. Due to the relatively short length of the strands in the
half-scale test specimens, considerable prestress was lost upon seating of the posttensioning anchors during post-tensioning. As a result, the average initial stresses in the
beam post-tensioning strands are smaller than the design initial stress of 0.60fbpu in the
prototype subassemblage.
A nominal compression (i.e., downward) axial force is applied to the 191 mm (7.5
in.) thick wall test region of the reaction block (adjacent to the coupling beam) using
eight 25.4 mm (1 in.) diameter vertical tie-down steel bars (Fig. 9.4) that bolt the block to
the strong floor [through 60 mm (2.36 in.) diameter ducts] on either side of the wall test
region. Each tie-down rod is pre-tensioned to about 89 kN (20 kips) resulting in a total
211
force of about 712 kN (160 kips) in the wall test region. This axial force stayed relatively
constant during each test. Eight additional tie-down bars are used through the 1473 mm
(58 in.) wide region of the reaction block to secure it to the strong floor.
Fig. 9.16 Photograph of mono-strand post-tensioning jack and pump

9.1.7
Material Properties
American Society for Testing and Materials (ASTM) standards were followed to
determine the properties of the coupling beam steel, cover plate steel, angle steel, spiral
steel, and post-tensioning strand (ASTM 370), and the wall concrete (ASTM C39/C
39M).
Coupling beam steel
The material samples for the coupling beam steel (ASTM A572 Grade 50 steel)
were saw cut from the flange in the direction along the length of the beam. A total of
three samples (only for the W1068 section) were machined into standard (ASTM .505)
12.8 mm (0.5 in.) round specimens with 50.8 mm (2 in.) gage length.
Fig. 9.17 shows the measured monotonic tensile stress-strain relationships of the
three specimens, as obtained from an Instron Model 5590-67HVL hydraulic universal
testing machine with 267 kN (60 kip) capacity. The strains were measured using an
Instron Model 2630-114 extensometer with 50.8 mm (2 in.) gage length. The failure of
all three specimens occurred within this gage length.
212
700
600
stress (MPa)
500
400
300
200
yield strength
maximum (peak) strength
ultimate strain
100
0
-0.05
0.05
0.10
0.15
0.20
0.25
0.30
0.35
strain
Fig. 9.17 Beam steel stress-strain behavior
Table 9.2 summarizes the material properties of the coupling beam steel specimens.
The samples show a distinct yield point. The yield strength was determined as the lower
yield point on the measured stress-strain relationship. The yield strain was determined by
dividing the yield strength with the Youngs modulus. The Youngs modulus was
calculated based on the slope from two points on the linear-elastic portion of the
measured stress-strain relationship. The , , and markers in Fig. 9.17 correspond to
the yield strength, maximum (peak) strength, and ultimate strain values, respectively, in
Table 9.2.
TABLE 9.2
BEAM STEEL (W1068) MATERIAL PROPERTIES
Specimen No.
1
2
3
Average
fby (MPa)
390
386
383
386
by (%)
0.19
0.18
0.20
0.19
fbm (MPa)
519
531
521
524
bm (%)
16.5
16.3
16.8
16.5
fbu (MPa)
441
451
443
445
Notes:
fby=lower yield strength; by=fby divided by measured Youngs modulus ( markers in Fig. 9.17).
fbm=maximum (i.e., peak) strength; bm=strain at fbm ( markers in Fig. 9.17).
fbu=0.85fbm; bu=ultimate strain at 0.85fbm ( markers in Fig. 9.17).
213
bu (%)
26.5
30.0
28.1
28.2
Cover plate steel

The material samples for the coupling beam flange cover plate steel (ASTM A572
Grade 50 steel) were saw-cut in the longitudinal direction (i.e., in the direction along the
length of the beam). A total of three samples were machined into standard (ASTM .505)
12.8 mm (0.5 in.) round specimens with 50.8 mm (2 in.) gage length, similar to the
coupling beam material.
Fig. 9.18 shows one of the specimens after failure and Fig. 9.19 shows the
measured monotonic tensile stress-strain relationships of the three specimens, as obtained
from an Instron Model 5590-67HVL hydraulic universal testing machine with 267 kN
(60 kip) capacity. The strains were measured using an Instron Model 2630-114
extensometer with 50.8 mm (2 in.) gage length. The failure of all three specimens
occurred within this gage length.
Table 9.3 summarizes the material properties of the cover plate steel specimens.
The samples show a distinct yield point. The yield strength was determined as the lower
yield point on the measured stress-strain relationship. The yield strain was determined by
dividing the yield strength with the Youngs modulus. The Youngs modulus was
calculated based on the slope from two points on the linear-elastic portion of the
measured stress-strain relationship. The , , and markers in Fig. 9.19 correspond to
the yield strength, maximum (peak) strength, and ultimate strain values, respectively, in
Table 9.3.
Fig. 9.18 Cover plate steel specimen
214
700
600
stress (MPa)
500
400
2
1
300
200
yield strength
ultimate strain
100
0
-0.05
0.05
0.10
0.15
0.20
0.25
0.30
0.35
strain
Fig. 9.19 Cover plate steel stress-strain behavior
TABLE 9.3
COVER PLATE STEEL MATERIAL PROPERTIES
Specimen No.
1
2
3
Average
fry (MPa)
399
380
383
387
ry (%)
0.19
0.20
0.17
0.19
frm (MPa)
572
570
573
572
rm (%)
14.3
15.1
15.2
14.9
fru (MPa)
486
485
487
486
ru (%)
26.9
24.8
21.2
24.3
Notes:
fry=lower yield strength; ry=fry divided by measured Youngs modulus ( markers in Fig. 9.19).
frm=maximum (i.e., peak) strength; rm=strain at frm ( markers in Fig. 9.19).
fru=0.85frm; ru=ultimate strain at 0.85frm ( markers in Fig. 9.19).
Angle steel
The material samples for the angle steel (ASTM A36-96) were saw-cut from the
longer (i.e., vertical) leg in the direction perpendicular to the angle length. A total of nine
samples (three samples each for two heats of L841/2 angles and three samples for the
L845/8 angles) were machined into 6.40 mm (0.25 in.) round specimens (proportional
to standard ASTM .505 specimens because of the limited angle thickness) with 50.8 mm
215
(2 in.) gage length. The L841/2 angles in Test 3 were from the first heat and the angles
in Tests 4-6 were from the second heat.
Fig. 9.20 shows one of the specimens during testing and Figs. 9.21a-c show the
measured monotonic tensile stress-strain relationships of the nine specimens, as obtained
from an Instron Model 5590-67HVL hydraulic universal testing machine with 267 kN
(60 kip) capacity. The strains were measured using an Instron Model 2630-114
extensometer with 50.8 mm (2 in.) gage length. The failure of all three specimens
occurred within this gage length.
Table 9.4 lists the material properties of the angle steel specimens. The samples
show a distinct yield point. The yield strength was determined as the lower yield point on
the measured stress-strain relationship. The yield strain was determined by dividing the
yield strength with the Youngs modulus. The Youngs modulus was calculated based on
the slope from two points on the linear-elastic portion of the measured stress-strain
relationship. The , , and markers in Fig. 9.21 correspond to the yield strength,
maximum (peak) strength, and ultimate strain values, respectively, in Table 9.4.
Fig. 9.20 Angle steel specimen
216
(a)
700
600
L8x4x1/2 Angle 1st heat
stress (MPa)
500
400
300
200
yield stength
100
0
-0.05

ultimate strain
0.05
0.10
0.15
0.20
0.25
0.30
0.35
strain
(b)
700
600
L8x4x1/2 Angle 2nd heat
stress (MPa)
500
400
1
300
200
yield stength
100

ultimate strain
0
-0.05
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.30
0.35
strain
(c)
700
600
L8x4x5/8 Angle
stress (MPa)
500
400
300
yield stength
200
100

ultimate strain
0
-0.05
0.05
0.10
0.15
0.20
0.25
strain
Fig. 9.21 Angle steel stress-strain behavior: (a) L8x4x1/2 first heat;
(b) L8x4x1/2 second heat; (c) L8x4x5/8
217
TABLE 9.4
ANGLE STEEL MATERIAL PROPERTIES
Specimen No.
1
2
3
Average
Specimen No.
1
2
3
Average
Specimen No.
1
2
3
Average
fay (MPa)
324
317
308
316
ay (%)
0.18
0.17
0.24
0.20
fay (MPa)
334
321
327
327
ay (%)
0.19
0.16
0.17
0.17
fay (MPa)
376
374
380
377
ay (%)
0.21
0.28
0.37
0.29
L8x4x1/2 Angle 1st heat

fam (MPa)
am (%)
472
15.3
476
22.8
469
20.3
472
19.5
L8x4x1/2 Angle 2nd heat
fam (MPa)
am (%)
486
16.2
487
19.3
483
20.9
485
18.8
L8x4x5/8 Angle
fam (MPa)
am (%)
548
16.9
544
15.2
548
15.2
547
15.8
fau (MPa)
401
405
399
402
au (%)
21.3
31.4
29.3
27.3
fau (MPa)
413
414
411
413
au (%)
19.8
26.7
28.8
25.1
fau (MPa)
467
462
466
465
au (%)
24.8
22.1
19.8
22.2
Notes:
fay=lower yield strength; ay=fay divided by measured Youngs modulus ( markers in Fig. 9.21).
fam=maximum (i.e., peak) strength; am=strain at fam ( markers in Fig. 9.21).
fau=0.85fcm; au=ultimate strain at 0.85fam ( markers in Fig. 9.21).
Spiral wire steel

The monotonic tensile stress-strain relationships of three spiral wire steel (ASTM
A82) samples were obtained by using a 2669 kN (600 kip) SATEC Model 600XWHVL
universal testing machine. The strains were obtained using an MTS Model 634.25E-24
extensometer with a 50.8 mm (2 in.) gage length.
In order to ensure failure to occur within the extensometer gage length, the crosssection of each wire specimen was reduced by hand-grinding a short length of the wire.
The diameter of the reduced cross-section, where necking and fracture occurred, was
used to calculate the stresses from the measured forces. Fig. 9.22 shows the three
specimens that were tested and Fig. 9.23 shows the measured stress-strain relationships
for the specimens. The material properties from the tests are summarized in Table 9.5.
The , , and markers in Fig. 9.23 correspond to yield strength, maximum (peak)
strength, and ultimate strain values, respectively, in Table 9.5.
218
Fig. 9.22 Spiral wire specimens
700
600
stress (MPa)
500
400
1 3 2
300
yield strength
200
100
0
-0.05

ultimate strain
0.05
0.10
0.15
0.20
0.25
strain
Fig. 9.23 Spiral wire stress-strain behavior
219
0.30
0.35
TABLE 9.5
SPIRAL WIRE STEEL MATERIAL PROPERTIES
Specimen
No.
1
2
3
Average
Reduced
Diameter (mm)
10
10
10
10
Esp
(MPa)
183669
179663
176360
179897
fspy
(MPa)
595
594
594
594
spy
(%)
0.52
0.53
0.54
0.53
fspm
(MPa)
626
630
637
631
spm
(%)
3.45
3.73
3.96
3.71
fspu
(MPa)
532
536
541
536
spu
(%)
8.24
8.51
8.36
8.37
Notes:
Esp=Measured Youngs modulus.
fspy, spy=yield strength and strain based on a 0.2% strain off-set at zero stress ( markers in Fig. 9.23).
fspm=maximum (i.e., peak) strength; spm=strain at fspm ( markers in Fig. 9.23).
fspu=0.85fspm; spu=ultimate strain at 0.85fspm ( markers in Fig. 9.23).
Note that the spiral wire specimens were bent prior to testing (see Fig. 9.22), as a
result of being stored in a spool. The stress-strain relationships in Fig. 9.23 and the values
in Table 9.5 show the behaviors of the specimens after they were straightened under
tensile loading; with the initial linear-elastic portion of the curves determined from the
measured linear-elastic unloading stiffnesses of the straightened specimens. In order to
determine the point on the original stress-strain curves (not shown in Fig. 9.23) where the
samples were straightened, eight unloading-reloading cycles were applied the cycles
were set to load to 414 MPa, unload to 276 MPa, reload to 441 MPa, unload to 303 MPa,
and continue similar cycles at 28 MPa increments until a stress of 607 MPa was reached.
The unloading stiffness from each of these cycles was determined and then compared
with the linear-elastic Youngs modulus, Esp determined from an additional (ninth)
unloading-reloading cycle from a strain of 0.035.
Since the Youngs modulus, Esp determined during unloading from a strain of 0.035
corresponded to a straight specimen, it was assumed that the complete straightening of
the specimen occurred at a stress when the unloading stiffness interpolated from the first
eight unloading-reloading cycles equaled the unloading stiffness, Esp from the cycle at
0.035 strain (see Fig. 9.24). The portion of the measured stress-strain curve before this
stress was ignored and the linear-elastic origin for each curve was determined by
unloading from this point with a slope equal to Esp.
The strain values listed in Table 9.5 were calculated from this revised origin for the
data. Note that, since the stress-strain relationships in Fig. 9.23 do not show a distinct
yield point, the yield strength and strain were determined based on a 0.2% strain off-set
at zero stress upon unloading from the yield point using the measured Youngs modulus,
Esp.
220
unloading stiffness (MPa)
200000
Esp
stress at complete
straightening of wire
160000
300
400
500
600
unloading stress (MPa)
700
Fig. 9.24 Unloading stiffness of spiral wire steel

Post-tensioning strand
Fig. 9.25 shows the monotonic tensile stress-strain relationships of three single
strand post-tensioning steel specimens (ASTM A416 low-relaxation seven-wire strands),
as obtained from a 2669 kN (600 kip) SATEC Model 600XWHVL universal testing
machine. The strains were obtained using an MTS Model 634.25E-24 extensometer with
a 50.8 mm (2 in.) gage length.
Single strand steel wedge/barrel anchors (see Fig. 9.10) were used to pull the posttensioning strands until failure, resulting in anchor conditions similar to those in the
coupled wall subassemblage experiments. The strand specimens were approximately 1.5
m (60 in.) long between anchors and were carefully aligned between the loading heads of
the testing machine (see Fig. 9.26). No prestress was applied on the strands prior to
loading. The failure of all three specimens occurred due to the fracture of a posttensioning wire (one wire in a seven-wire strand) inside an anchor, at an average stress of
fbpu=1698 MPa (246 ksi), well below the design maximum strength of fbpu=1862 MPa
(270 ksi). Table 9.6 summarizes the material properties for the post-tensioning strand
specimens. The and markers in Fig. 9.25 correspond to the limit of proportionality
point and the maximum (peak) strength/ultimate strain values, respectively, in Table 9.6.
Note that the ultimate strain, bpu at wire fracture of the post-tensioning strands is reached
when the peak stress, fbpu is reached.
221
2000
1800
2
3
1600
stress (MPa)
1400
1200
1000
800
600
400
proportionality limit
maximum (peak) strength/ultimate strain
200
0
0.005
0.01
0.015
strain
Fig. 9.25 Post-tensioning strand stress-strain behavior
Note that the stress-strain relationships in Fig. 9.25 do not show the initial portions
of the measured curves where the post-tensioning strand specimens had reduced stiffness.
The specimens do not show a distinct yield point. The linear-elastic origin for each curve
was determined by unloading from the limit of proportionality point with a slope equal to
the measured Youngs modulus. The Youngs modulus was calculated based on the slope
from two points on the linear-elastic portion of the measured stress-strain relationship.
The stress-strain relationships in Fig. 9.25 and the values in Table 9.6 were calculated
from this revised origin for the data. The strain at the limit of proportionality was
determined by dividing the stress at the limit of proportionality with the measured
Youngs modulus.
222
PT strand
PT anchor
extensometer
TESTING MACHINE
Fig. 9.26 Post-tensioning strand test setup
TABLE 9.6
POST-TENSIONING STRAND PROPERTIES
Specimen No.
1
2
3
Average
bpy (%)
0.56
0.55
0.53
0.55
fbpy (MPa)
1138
1103
1069
1103
fbpu (MPa)
1696
1724
1674
1698
bpu (%)
0.98
1.16
1.01
1.05
Notes:
fbpy=limit of proportionality; bpy=fbpy divided by measured Youngs modulus ( markers in Fig. 9.25).
fbpu=maximum (i.e., peak) strength; bpu=ultimate strain at fbpu ( markers in Fig. 9.25).
Wall concrete
The compressive strengths of the unconfined concrete used in the reaction and
loading blocks were obtained by conducting three standard 152305 mm (6x12 in.)
cylinder tests using a 2669 kN (600 kip) SATEC Model 600XWHVL universal testing
machine. Table 9.7 shows the results obtained from the concrete cylinder tests and Fig.
9.27 shows two sample cylinders after failure. The concrete cylinders were cast at the
same time as the reaction and loading blocks and were kept under the same conditions as
the blocks until testing.
223
TABLE 9.7
CONCRETE PROPERTIES
Specimen No.
1
2
3
Block
Loading Block
Reaction Block
Reaction Block
Cast Date
02/07/2001
02/07/2001
02/07/2001
Test Date
11/29/2001
11/29/2001
10/17/2003
fc (MPa)
65.7
66.8
67.6
Note: fc=maximum (i.e., peak) strength of unconfined loading block or reaction block concrete.
Fig. 9.27 Concrete cylinder specimens

9.1.8
Instrumentation
A National Instruments SCXI data acquisition system and a desktop computer were
used to achieve close-to-simultaneous scanning, conditioning, and recording of the data
measured during each test. The test instrumentation includes load cells that are numbered
with a prefix of LC, displacement/rotation transducers that use a prefix of DT, and strain
gages that have a prefix of SG as follows (Fig. 9.28).
Load cells
Load cells (1) two load cells (LC1 and LC2) to measure the actuator forces; (2)
up to six load cells (LC3-LC8) to measure the forces in the beam post-tensioning strands;
(3) up to two load cells (LC9-LC10) to measure the forces in the angle-to-wall
connection post-tensioning strands; and (4) eight load cells (LC11-LC18) to measure the
vertical forces applied to the wall test region of the reaction block.
224
see detail
at right
reaction block
DT6
LC1
DT1
305mm
LC9
64m
LC3-LC8
305mm
SG1
1118mm
102mm
DT9
25mm
SG23-SG25
DT5
loading block
102mm
DT10
140mm
57, 32, 25 mm
beam PT strands
64mm
tiltmeter
(DT16) at
beam
midspan
LC11-LC18
38mm
191mm
SG4
LC10 SG3
DT8
LC2
DT2
191mm
SG2
DT17
DT7
actuators
DT3
DT4
(a)
6.35mm
19mm
25mm
SG15(E)
SG17(E)
SG16(W)
SG18(W)
cover plate
140mm
DT14
152mm
SG5(E)
SG6(W)
SG7(E)
SG8(W)
SG9(E or W)
DT11
flange centerline
76mm
test wall
region
DT17
beam
DT12
25mm
140mm
web centerline
shim plate
152mm
DT15
SG26-SG28
64, 32, 25 mm
embedded plate
SG10(E)
SG11(W)
SG19(E)
SG12(E)
SG13(W)
SG14(E)
DT13
SG21(E)
SG22(W)
SG20(W)
19mm
6.35mm
25mm
(b)
Fig. 9.28 Instrumentation: (a) overall view; (b) beam-to-wall joint region
The post-tensioning strand load cells LC3-LC10 were manufactured and calibrated
at the University of Notre Dame by instrumenting a second post-tensioning barrel using
225
four strain gages (Fig. 9.29a). During testing, the load cell barrel is placed between the
post-tensioning anchor barrel and the bearing plate (Fig. 9.29b), placing the load cell in
compression. The post-tensioning load cells were calibrated using a 2669 kN (600 kip)
SATEC Model 600XWHVL universal testing machine. Following several cycles of
shake-down loading under various orientations of each load cell, eight increments of load
up to 267 kN (60 kips) in compression were used to relate the force from the testing
machine to the readings from the four strain gages. The load cells showed linear behavior
within the calibration load range. More information on the post-tensioning load cells can
be found in May et al. (2006).
Load cells LC11-LC18 each consisted of a single electrical resistance strain gage
attached to a steel tie-down bar (a total of 8 bars) used to apply the vertical force in the
wall test region.
(a)
(b)
Fig. 9.29 Measurement of post-tensioning forces:

(a) load cells; (b) placement of load cells
Displacement/rotation transducers in the vertical plane
Displacement/rotation transducers in the vertical plane (1) two displacement
transducers (DT1 and DT2) to measure the actuator axial displacements; (2) three
displacement transducers (DT3-DT5) to measure the loading block displacements; (3)
three displacement transducers (DT6-DT8) to measure the reaction block displacements;
(4) two displacement transducers (DT9 and DT10) to measure the coupling beam vertical
displacements; (5) three displacement transducers (DT11-DT13) to measure the gap
opening displacements between the coupling beam and the reaction block; (6) two
displacement transducers (DT14 and DT15) to measure the local horizontal deformations
in the high compression regions of the wall test region in the reaction block; and (7) up to
two tiltmeters (DT16-DT17) to measure the coupling beam rotations at the midspan and
at the reaction block end. Displacement transducers DT3-DT10 were string pots and
displacement transducers DT11-DT15 were Linear Variable Displacement Transducers
(LVDTs). Fig. 9.30 shows DT11-DT15 at a displaced position of the beam during Test 1.
226
South
North
Fig. 9.30 DT11-DT15 at displaced position of coupling beam during Test 1

Electrical resistance strain gages
Electrical resistance strain gages (1) four strain gages (SG1-SG4) to measure the
horizontal strains of the coupling beam web along its length; (2) up to eighteen strain
gages (SG5-SG22) to measure the horizontal strains in the coupling beam flanges and
cover plates at the reaction block end (the letters E and W in Fig. 9.28 indicate gages
placed on the east and west sides of the beam, respectively); and (3) six strain gages
(SG23-SG28) to measure the strains in the confined concrete regions of the wall test
region. Gages SG23-SG28 were attached to horizontal 9.5 mm (0.375 in.) diameter
deformed steel reinforcing bars tied to the spiral reinforcement inside the concrete. 120
ohm strain gages with a gage factor of 2.0 were used for all strain gage applications.
9.1.9
Test Procedure
The procedure for testing a subassembly requires a two day process. The first day
involves the alignment of the coupling beam and the loading and reaction blocks,
application of the gravity load to the wall test region of the reaction block, posttensioning of the beam, and connection of the top and seat angles at the beam-to-wall
interfaces. On the second day, the actual testing of the structure occurs by moving the
loading block vertically through a predetermined displacement history.
On the first day of preparations, the initial step is to determine the self weight of the
loading block and the portion of the self weight of the beam that rests on the loading
block. This step is completed by lifting the loading block using the two servo-controlled
hydraulic actuators. The actuators are displaced upward until the loading block is off of
its temporary support jacks on the laboratory floor. The reading from each of the two
227
actuator load cells is then recorded. The recorded load from each actuator is used in the
following day to subtract out the self weight of the system.
The next step is to make sure that the coupling beam, the loading block, and the
reaction block are properly aligned. The loading block is rested back on three temporary
support jacks and the hydraulic actuators are turned off. The positions of the loading
block and the coupling beam are adjusted on the support jacks until the entire
subassemblage is aligned using a laser alignment tool. The seat angles are used as
temporary supports for the beam at the two ends.
Following proper alignment of the structure, all of the instrumentation is zeroed or
initialized. The vertical forces on the wall test region of the reaction block are applied (by
tightening the nuts on the tie-down bars) or checked from a previous test based on the
measurements from LC11-LC18. Then, the post-tensioning of the coupling beam to the
reaction and loading blocks is carried out by first applying a small amount of stress to
each of the main strands until the gaps (necessary for construction tolerances) between
the coupling beam and the reaction and loading blocks are closed and the subassemblage
is snugly connected. The LVDTs that measure the gap opening displacements between
the coupling beam and the reaction block (DT11-DT13) are re-zeroed, so that the LVDT
measurements do not include the gap that exists between the beam and the reaction block
before post-tensioning. After the LVDTs have been re-zeroed, the four main beam posttensioning strands are stressed up to their full desired value of about 0.50-0.55fbpu. The
post-tensioning strand load cells (LC3-LC6) are used to determine when the desired
prestress level has been reached. Finally, the positions of the two servo-controlled
hydraulic actuators are recorded.
The top and seat angles at the ends of the coupling beam are bolted to the beam
flanges and then connected to the reaction block and the loading block using unbonded
post-tensioning strands. The angle connections are tightened after the post-tensioning of
the main coupling beam strands to ensure that most of the initial post-tensioning force is
transferred into the coupling beam. The angle-to-beam connection bolts are tightened to
slip-critical level by using the turn-of-nut pretensioning method described in the LRFD
Manual of Steel Construction (AISC 2001).
On the day of the test (i.e., second day), the first step is to initialize all instruments
based on the measurements from the previous day. The two actuators are then moved to
their positions recorded during the setup on the previous day. The servo-controller is then
switched to load control and the self-weight forces recorded during the previous day are
applied on the actuators to remove the effect of self-weight from the specimen. The
actuator positions are recorded again and the coupling beam chord rotation and the beam
shear force and moment are zeroed. The string pots and tiltmeters are also re-zeroed at
this stage. The servo-controller is put back into displacement control mode and, if
necessary, the actuators are moved back to their positions recorded just after the
application of the self-weight forces. The structure is now ready to run the predetermined
displacement history.
228
A summary of the complete procedure used to set up and conduct the tests is given
below.
Day 1 (preparation day)
1. Determine actuator forces due to self weight of subassemblage
2. Level and align coupling beam and loading block using temporary floor jacks
3. Zero/initialize all instrument readings
-Post-tensioning strand load cells
-String pots
-LVDTs measuring gap opening
-LVDTs measuring deformations in wall test region
-Tiltmeters
-Strain gages
4. Apply/check vertical forces on wall test region of reaction block
5. Apply a small amount of prestress to main post-tensioning strands until coupling
beam is snugly connected to reaction and loading blocks
6. Re-zero LVDTs used to measure gap opening
7. Apply full coupling beam post-tensioning force
8. Record actuator positions
9. Tighten angle-to-beam connection bolts to slip-critical level
10. Prestress angle-to-wall connection strands
Day 2 (test day)
1. Initialize all instruments
2. Move actuators to position recorded on step 8 of Day 1
3. Switch to load control
4. Apply self weight forces recorded on step 1 of Day 1
5. Record actuator positions
6. Zero beam rotation and beam shear/moment
7. Re-zero string pots and tiltmeters
8. Switch to displacement control
9. Move actuators to position recorded in step 5 of Day 2
10. Run predetermined displacement history
9.2
Results of Experiments
The experimental results are evaluated below as follows: (1) behavior of properlydesigned subassemblages; (2) effect of coupling beam flange cover plates; (3) effect of
angle thickness; (4) effect of beam post-tensioning steel area; and (5) effect of beam
depth.
9.2.1
Behavior of Properly Designed Subassemblages (Tests 1-4)
The behavior of properly-designed unbonded post-tensioned hybrid coupled wall

subassemblages is evaluated below using the results from Tests 1-4 (Table 9.1). In Test 1,
229
top and seat angles were not used at the beam-to-wall connections to study the behavior
of the subassemblage and to verify the analytical models without the angles. In Test 2,
the coupling beam used in the first test was retested after replacing the post-tensioning
strands. In Test 3, a new beam with top and seat angles at the beam-to-wall connections
was used. Test 4 was a monotonic test with only two tension angles and no beam posttensioning steel to obtain more information on the angle behavior.
Preliminary tests of a subassemblage without top and seat angles showed slip of the
coupling beam at the beam-to-wall interfaces. To prevent the beams in Tests 1 and 2 from
slipping, restraining plates with thickness of 9.5 mm (0.375 in.) were connected to the
shim plates above and below the beam. The restraining plates did not have any effect on
the behavior of the subassemblages other than prevent slip of the beam with respect to the
wall blocks. The coupling beam in the other tests (except Test 10 as described later) did
not slip (since top and seat angles were used in these tests); and thus, restraining plates
were not necessary.
Figs. 9.31a-c show the coupling shear force versus rotation (Vb-b) behaviors from
Tests 1-3, respectively. The force Vb is determined from the actuator forces and the
rotation b is the coupling beam chord rotation determined from transducers DT9 and
DT10 measuring the vertical displacements of the beam at the ends. Only the first cycle
during each set of displacement cycles of equal amplitude is shown, except where
significant differences occur in the subsequent cycles.
Fig. 9.31a shows that the behavior of the subassemblage without angles is
essentially bilinear-elastic, caused mainly by gap opening at the beam ends. A small
amount of reduction (i.e., loss) in prestress occurred during the test as discussed in more
detail later. The reduction in prestress resulted in a small reduction in the Vb resistance;
however, it did not cause a reduction in the strength of the specimen (see Fig. 9.31a).
Failure of the subassemblage occurred when one of the seven wires in the top west beam
post-tensioning strand fractured at the end of the second cycle to b=-8% in the negative
(i.e., counter clockwise) direction. The fracture occurred inside a post-tensioning anchor
(similar to the failures in the material tests), where the anchor teeth grip the strand at the
live (north) end, at a stress of 1466 MPa [212.6 ksi, well below the average maximum
strength of fbpu=1698 MPa (246.3 ksi) from the material tests], and resulted in a small
reduction in Vb. The largest stress reached in the four beam post-tensioning strands
during the test was 1576 MPa (228.5 ksi). The subassemblage was unloaded and the
experiment was terminated after the strand fracture. No damage was observed in the
reaction block or the loading block during the test, allowing both blocks to be reused in
all of the subsequent Tests 2-11.
230
Fig. 9.31 Measured coupling shear force versus beam chord rotation (Vb-b)
relationships: (a) Test 1; (b) Test 2; (c) Test 3
231
The coupling beam used in Test 1 did not receive any damage other than a small
amount of compression yielding in the cover plates at the beam ends. To demonstrate this,
the beam was retested after replacing the post-tensioning strands. Note that only selected
displacement cycles from Fig. 9.15 were used in this second test. The behavior of the
retested beam in Fig. 9.31b is very similar to the behavior of the original beam in Fig.
9.31a. The small differences observed between the two tests are possibly because all or
most of the compression yielding in the beam cover plates occurred in Test 1 and no
further appreciable beam yielding took place in Test 2. The results demonstrate that
unbonded post-tensioned hybrid coupled wall subassemblages can go through large
nonlinear reversed cyclic displacements without receiving significant damage in the walls
or the beams. The beams do not need to be replaced after a large earthquake as long as
the yielded or fractured strands are replaced.
Figs. 9.31a and 9.31b show that there is a small increase in the post-softening
stiffness (i.e., stiffness after the softening state, see Chapter 4) of the subassemblage after
about b=4% to 5% rotation. This occurred as a result of the kinking of the beam posttensioning strands when they came into contact with the oversized post-tensioning ducts
at the beam-to-wall interfaces, leading to increased second-order effects in the strands.
Kinking of the strands did not have any undesirable effects on the behavior of the
structure (note that the strand fracture in Test 1 occurred inside an anchor and not along
the length of the strand). The post-tensioning strands in Test 2 did not fracture, and the
test was terminated after unloading from the third cycle to b=8%, without failure.
Fig. 9.31c shows the Vb-b behavior of the subassemblage with top and seat angles
from Test 3. As described earlier, this specimen is a half-scale model of the prototype
subassemblage described in Chapter 3. The hysteresis loops indicate desirable seismic
characteristics with stable behavior up to b=8% rotation and significant energy
dissipation. Comparing Figs. 9.31a and 9.31c, the increase in strength and energy
dissipation occurs as a result of the angles.
Fig. 9.32a shows the overall displaced shape of the beam in Test 3 at b=+8%
rotation in the positive (i.e., clockwise) direction and Fig. 9.32b shows a close-up view of
the test region near the reaction block at +8% rotation. The displaced shape of the
subassemblage is similar to the displaced shape with respect to the reference line in Fig.
1.3c, with most of the beam rotation occurring as a result of the gaps at the beam ends.
The rotation of the beam with respect to the walls results in the yielding of the top and
seat angles in tension and compression as shown in Figs. 9.32a and 9.32b.
The straight dashed line in Fig. 9.31c shows the theoretical initial (i.e., linearelastic) stiffness of the same subassemblage assuming fixed beam-to-wall connections
(representing an embedded steel beam). As a result of post-tensioning, the measured
initial stiffness of the test beam before the initiation of gap opening is similar to the initial
stiffness of an embedded beam. The hysteresis loops in Fig. 9.31c indicate that the beam
post-tensioning strands provide a restoring force such that the gaps are closed upon
unloading, thus pulling the beam towards its undeformed position with little residual
displacement (i.e., large self-centering capability). The initial stiffness of the
subassemblage is preserved even after unloading from very large nonlinear rotations.
232
(a)
(b)
Fig. 9.32 Test 3: (a) displaced shape overall; (b) displaced shape local
233
The sum of the coupling beam post-tensioning forces, Pb in Test 3 (normalized with
the total design maximum strength of the post-tensioning strands, Pbu=abpfbpu) is plotted
in Fig. 9.33. Before the initiation of gap opening, the total force in the post-tensioning
strands is similar to the initial post-tensioning force, Pbi. As the specimen is displaced, the
strand forces increase, thus resisting gap opening. Prestress losses are observed upon
unloading from increased displacements; however, these losses are small because of the
use of unbonded post-tensioning strands. Unlike Test 1, fracture of the post-tensioning
strands did not occur in Test 3.
Pb / abp f bpu
initial force
prestress loss
final force
Test 3
0
-10
0
10
Fig. 9.33 Test 3 - total beam post-tensioning force

Fig. 9.34 shows the depth of contact, cb between the coupling beam and the
reaction block in Test 3, as determined from displacement transducers DT11-DT13 at the
beam end (see Fig. 9.28). The contact depth is normalized with respect to the depth of the
beam including the cover plates, dbc=302 mm (11.89 in.). Each marker represents the
measured contact depth upon first loading to a peak b value in the positive or negative
direction during the displacement history followed in the test. Under small rotations of
the beam, the entire beam depth is in contact with the reaction block (note that the beam
depth between the shim plates is not in contact). The initiation of gap opening (i.e.,
decompression) occurs before b=0.25% and results in a rapid reduction of
the contact depth. Values of cb below the horizontal dashed line indicate that the contact
234
depth is within the shim plate. A contact depth of cb=12.5 mm (0.49 in., approximately
0.66 times the cover plate thickness, tc) is reached at b=2%, beyond which the contact
depth changes relatively little.
normalized contact depth, c b /d bc
Test 3
negative (ccw) rotation
positive (cw) rotation
shim plate depth
beam chord rotation, | b| (percent)
10
Fig. 9.34 Test 3 - contact depth

To investigate the amount of yielding in the coupling beam flanges in Test 3, the
deformations were measured using strain gages SG5-SG14 located at the flange
centerline. As an example, the thick solid line in Fig. 9.35 shows the top flange strains in
SG5 at the beam end (see Fig. 9.28). Application of the initial post-tensioning force
results in a compressive strain at the beginning of the test. The strain readings cycle but
remain compressive as the beam is rotated in the positive and negative directions. As a
result of gap opening, the compression strain in SG5 drops to a minimum value during
the first cycle to b=+0.25%. Upon reversed loading to b=-0.25%, the strain in SG5
reaches the measured average (from three specimens) yield strain of by=-0.0019
(indicated by the thick dashed horizontal line). As shown in Fig. 9.35, there is an
accumulation of plastic (i.e., residual) compressive strains in the beam flange upon
unloading from b values beyond -0.25%. The largest strain measured in SG5 is -0.0035
(approximately equal to 1.8by). It is concluded that the amount of yielding in the flanges
of the beam in Test 3 was negligible.
235
Fig. 9.35 Test 3 - beam flange steel and wall concrete strains
Similarly, in order to investigate the performance of the reaction block, the
deformations in the 191 mm (7.5 in.) thick wall test region were measured using DT14DT15 and SG23-SG28 (see Fig. 9.28). As an example, the thin solid line in Fig. 9.35
shows the strains from SG24, which remain compressive during the test. The thin dashed
horizontal line corresponds to the assumed unconfined concrete spalling/crushing strain
of -0.003. It is concluded that the wall test region did not receive any damage during the
test, including cracking and/or spalling of the cover concrete (see Fig. 9.32b).
Initiation of low cycle fatigue cracks was observed in the vertical legs of the
tension angles at about b=7%. The cracks occurred at the critical section adjacent to the
fillet. The specimen was able to sustain three displacement cycles at b=8% with a steady,
but not excessively large, reduction in strength and post-softening stiffness (see Fig.
9.31c). This reduction in stiffness and strength occurred due to increased cracking and
necking of the vertical legs of the tension angles. Failure of the specimen eventually
occurred as a result of the complete fracture of the vertical leg of the seat angle at the
north (right) end of the beam when b=+9% was reached for the first time. The resistance
of the specimen at this stage was, approximately, 90% of the peak resistance. Fig. 9.36
shows the fractured angle at b=+9%. All four angles had sustained significant damage at
this stage resulting in a considerable amount of energy dissipation as shown in Fig. 9.31c.
The subassemblage was unloaded and the test was terminated upon fracture of the first
angle.
236
Test 3
b=+9%
fracture
Fig. 9.36 Test 3 angle fracture

As shown in Fig. 9.32b, the angle-to-wall connections performed extremely well,
allowing the angles to go through large nonlinear deformations without damaging the
concrete. The integrity of the angle-to-wall connections was preserved during the entire
test since the connection strands did not yield. The angle-to-beam connections also
behaved satisfactorily, with no slip between the angles and the beam up to b=5% and
negligible slip afterwards, indicating that the slip-critical bolts were adequate. Slip
between the coupling beam and the reaction and loading blocks did not occur
237
demonstrating that the angles provided adequate vertical support to the beam together
with friction resistance due to the beam post-tensioning force.
Fig. 9.37a shows the set-up for Test 4, where a subassemblage with only two
tension angles and no beam post-tensioning steel was tested monotonically. The coupling
beam from Test 3 was re-used in Test 4 with two new angles. Since no post-tensioning
strands were used in the coupling beam, the entire lateral resistance of the subassemblage
was from the two angles in tension. Fig. 9.37b shows the displaced shape from a similar
test (Test 6) upon loading to b=+8% and upon subsequent unloading to zero rotation.
The permanent gaps at the beam ends demonstrate the role of the beam post-tensioning
force in yielding the tension angles back in compression and closing the gaps upon
unloading. The monotonic Vb-b relationships from Test 4, as well as from two other
similar monotonic tests (Tests 6 and 8 as will be described later), and
one cyclic test with four angles and no beam post-tensioning steel (Test 11), are shown in
Fig. 9.37c. The tensile yield strengths of the angles were determined from these curves
and were used to develop/verify analytical models for the angles. Unlike the cyclic
loading experiments, the angles in the monotonic experiments did not show any evidence
of deterioration/failure.
9.2.2
Effect of Coupling Beam Flange Cover Plates (Tests 5-6)
A coupling beam with no flange cover plates was used in Test 5 to explore the need
for the cover plates in strengthening and stabilizing the flanges. As a second difference,
the gage length lgv for the angle vertical legs in Test 5 was longer than the gage length in
Test 3 by an amount equal to the cover plate thickness of tc=19.1 mm (0.75 in., Table 9.1).
The measured Vb-b behavior from Test 5, shown in Fig. 9.38a, illustrates the stable
behavior of the structure. Failure of the subassemblage occurred as a result of the
complete fracture of the vertical leg of the top angle at the north end of the beam when
b=-7% was reached for the third time, after which the subassemblage was unloaded and
the experiment was terminated. The angle-to-beam and angle-to-wall connections
performed as desired, with no slip or failure during the entire test.
As compared with Test 3 (Fig. 9.31c), the peak coupling shear strength and energy
dissipation from Test 5 (Fig. 9.38a) are smaller. This is possibly due to the increase in the
angle gage length, which led to a reduction in the angle contribution to the coupling
resistance. To investigate the effect of the gage length on the angle behavior, a second
monotonic angle test (Test 6) was conducted, with the smaller angle strength
demonstrated as shown in Fig. 9.37c. The coupling beam from Test 5 was re-used in Test
6, with no beam post-tensioning strands and after replacing the angles.
238
Fig. 9.37 Angle behavior (Tests 4, 6, 8, and 11): (a) test set-up;
(b) deformed shape; (c) Vb-b relationships
239
300
Test 5
-300
-10
0
10
(a)
0
beam steel yield strain
compressive strain
SG9(W)
SG8(W)
SG5(E)
Test 5
-0.014
test duration
(b)
Fig. 9.38 Test 5: (a) Vb-b relationship; (b) beam flange steel strains
240
To investigate the nonlinear behavior in the coupling beam from Test 5, Fig. 9.38b
shows the flange strains at the top south end of the beam (from strain gages SG5, SG8
and SG9 at the flange centerline). The strain readings remain compressive (similar to Fig.
9.35) and increase under negative b (i.e., when the loading block is moved upward), and
decrease under positive b. The markers show the maximum and minimum readings in
strain gage SG5 (at the beam end) corresponding to peak b values in the negative and
positive directions, respectively. The flange strain reaches the beam yield strain by=0.0019 (indicated by the dashed horizontal line) at approximately b=-1.5%, after which
there is an accumulation of plastic (i.e., residual) strains upon unloading. The delayed
flange yielding in Test 5 as compared with Test 3 could be due to uneven initial contact
at the beam-to-wall interfaces. The largest strain measured in SG5 is -0.012
(approximately equal to 6by). The measurements from the three gages indicate that the
beam strains quickly diminish away from the beam end. It is concluded that the lack of
cover plates resulted in an increase in the beam flange strains (as compared with Test 3,
Fig. 9.35); however, this did not have any adverse effects on the performance of the beam.
9.2.3
Effect of Angle Thickness (Tests 7-8)
The primary variable between Tests 7 and 5 is the thickness of the top and seat
angles; thicker L845/8 angles were used in Test 7 as compared with L841/2 angles
in Test 5. Fig. 9.39 shows the measured Vb-b hysteresis loops from Test 7, which
demonstrate desirable characteristics with stable behavior up to b=6%, large selfcentering capability, and increased energy dissipation. Initiation of low cycle fatigue
cracking was apparent in the vertical legs of the tension angles at b=6%, resulting in a
reduction in the coupling resistance and post-softening stiffness. Failure of the specimen
occurred as a result of the complete fracture of the vertical leg of the seat angle at the
north end of the beam when b=+7% was reached for the second time.
Compared with Test 5, the increase in the coupling strength and energy dissipation
in Test 7 (see Figs. 9.38a and 9.39) occurred as a result of the thicker top and seat angles.
This is shown using the results from a third monotonic angle test (Test 8) in Fig. 9.37c.
The coupling beam from Test 7 was re-used in Test 8, with no beam post-tensioning
strands and after replacing the angles.
As indicated by the self-centered hysteretic behavior in Fig. 9.39, the beam posttensioning force in Test 7 was sufficient to yield the tension angles back in compression
and close the gaps. The beam flanges remained stable during the entire test. The angle-towall connections also behaved satisfactorily, with no evidence of yielding in the
connection strands or damage in the concrete blocks.
Different from the previous tests, slipping of the angle horizontal legs with respect
to the coupling beam was more evident at rotations larger b=6%, indicating that the slip
critical capacity of the angle-to-beam bolts was not sufficient beyond this rotation. The
slip behavior resulted in the spikes in the hysteresis loops of Fig. 9.39. Early stiffening of
the hysteresis curves at small rotations during unloading is also evident in Fig. 9.39,
241
which occurred as the slipped angles came into contact with the loading and reaction
blocks prior to the closing of the gaps at the beam ends.
coupling shear force, Vb(kN)
300
-300
Test 7
-10
0
10
Fig. 9.39 Vb-b relationship from Test 7

9.2.4
Effect of Beam Post-Tensioning Steel Area (Test 9)
Test 9 differs from Test 7 in the total area of the beam post-tensioning steel; six
15.2 mm diameter strands were used in Test 9 as compared with four strands in Test 7
(Table 9.1). No modifications were made in the angle-to-beam connections. Figs. 9.40a
and 9.40(b) show the measured Vb-b hysteresis loops and the total normalized beam
post-tensioning force from Test 9, respectively. The results demonstrate stable behavior
up to b=7% rotation and increased self-centering capacity. The strength of the
subassemblage is also increased as compared with Test 7.Slip in the angle-to-beam
connections of Test 9 was more evident as compared with Test 7, (with first slip
occurring at b=4%) and resulted in the jagged hysteresis loops in Fig. 9.40a. The
initiation of low cycle fatigue cracking in the angle vertical legs was apparent at b=6%.
Ultimate failure of the specimen occurred as a result of the complete fracture of the
vertical leg of the seat angle at the south end of the beam when b=-8% was reached for
the second time.
242
Table 9.1 shows the nominal beam flange initial stress after gap opening, calculated
by normalizing the total measured initial force in the post-tensioning strands Pbi=abpfbpi
by Abfcfby, where Abfc is the cross section area of one flange plus cover plate (if any) and
fby is the measured beam steel yield strength. Note that, for simplicity, the small
difference between the measured yield strengths of the beam steel and cover plate steel in
Tests 1-3 was ignored in this calculation. As a result of the larger beam post-tensioning
force, the nominal flange initial stress in Test 9 is significantly larger than the initial
stress in the other test beams. Fig. 9.40c shows the strain measurements from SG6, SG8,
and SG9 in the top flange at the south end of the beam in Test 9. Flange yielding is
increased significantly (with a maximum compression strain of, approximately, -0.084,
which is equal to 44by) as a result of the larger post-tensioning force; however, the beam
remained stable during the test. As compared with Tests 3 (Fig. 9.33), 5, and 7, a larger
amount of prestress was lost during Test 9 (Fig. 9.40b), possibly because of the increased
compression yielding, and thus, shortening of the beam. Minor spalling of the cover
concrete in the wall test region occurred at the level of the beam flanges approximately
76 mm (3 in.) away from the beam-to-wall interface. The reaction block was patched and
reused in the subsequent tests. It is concluded that the increased post-tensioning force
resulted in increased, but still small levels of damage in the coupling beam and the wall
concrete.
9.2.5
Effect of Beam Depth (Tests 10-11)
The main difference between Tests 10 and 7 is the coupling beam depth; a W1499
beam was used in Test 10 as compared with a W1068 beam in Test 7 (Table 9.1). The
flanges of both beams were saw cut to a width of 159 mm (6.25 in.) as described earlier.
Additional differences are: (1) the gage length, lgv for the angle vertical legs in Test 10
(138 mm, 5.44 in.) was longer than the gage length in Test 7 (114 mm, 4.5 in., see Table
9.1); (2) two A490 25.4 mm slip critical bolts were used in the angle-to-beam
connections of Test 10 as compared with two A490 22.2 mm bolts in Test 7; and (3)
two post-tensioning strands were used in the angle-to-wall connections of Test 10 as
compared with four strands in Test 7.
Figs. 9.41a and 9.41b show the measured Vb-b hysteresis loops and the total
normalized beam post-tensioning force from Test 10, respectively. Although the
subassemblage performed well up to b=5%, the behavior of the structure beyond this
rotation was dominated by the failure of the beam post-tensioning strands. A wire in the
bottom east post-tensioning strand fractured at a stress of 1364 MPa (well below the
average maximum strength of fbpu=1698 MPa from the material tests) at the end of the
first cycle to b=+5%, followed by simultaneous fractures of one wire each in the top east
strand (at a stress of 1465 MPa) and top west strand (at a stress of 1497 MPa) at the end
of the first cycle to b=+6%, and subsequent post-tensioning wire fractures in the bottom
east strand (second fracture) and bottom west strand (at a stress of 1525 MPa) during the
second cycle to b=+7% and third cycle to -7%, respectively.
243
300
(a)
Test 9
-300
-10
10
Pb/ a bp f bpu
(b)
0
Test 9
0
-10
0
SG9(W)
compressive strain
(c)
0
10
beam steel
yield strain
SG8(W)
SG6(W)
Test 9
-0.09
test duration
Fig. 9.40 Test 9: (a) Vb-b relationship; (b) total beam post-tensioning force;
(c) beam flange steel strains
244
Fig. 9.41 Test 10: (a) Vb-b relationship; (b) total beam post-tensioning force;
(c) beam flange steel strains
245
All strand wire fractures occurred inside the post-tensioning anchors, where the
anchor teeth grip the strands at the live (north) end. The effects of these fractures on the
Vb-b hysteresis loops and on the total beam post-tensioning force are evident from Figs.
9.41a and 9.41b. As shown in Table 9.1, the initial stresses in the beam post-tensioning
strands used in Test 10 were similar to the initial stresses in the previous tests. However,
the largest stress reached in the strands during Test 10 was 1627 MPa, significantly
higher than the stresses reached during the other tests. The increase in the gap widths as a
result of the increase in the beam depth in Test 10 resulted in increased elongations, and
thus, increased stresses in the post-tensioning strands.
Other than the premature strand failures, the results from Test 10 demonstrated the
increased coupling shear strength due to the increased beam depth. Slip in the angle-tobeam connections was not observed due to the use of the larger bolts. The angle-to-wall
connections also behaved well; however, a small amount of slip between the beam and
the wall blocks occurred during second loading to b=-5%, after a significant loss in the
beam post-tensioning force due to strand fracture. It is possible that the slip occurred as a
result of a reduction in the friction resistance at the beam-to-wall interfaces due to the
loss in the total post-tensioning force.
Despite the fracture of multiple post-tensioning strand wires, the subassemblage
maintained all or most of its coupling strength. Significant low cycle fatigue deterioration
in the top and seat angles was observed upon first loading to b=7%. The ultimate failure
of the specimen occurred as a result of the complete fracture of the vertical leg of the seat
angle at the south end of the beam when b=-7% was reached for the third time.
Fig. 9.41c shows the strains from beam flange gages SG5, SG12, and SG14. The strain in
SG5 reached the beam steel yield strain at b=-0.25% (with a maximum compression
strain of, approximately, -0.0097, which is equal to 5.1by).
Test 11 was conducted to investigate the behavior of the angles in Test 10. The
coupling beam from Test 10 was re-used with no beam post-tensioning strands and after
replacing the angles. Different from the previous angle Tests 4, 6, and 8, the
subassemblage in Test 11 had all four angles and was tested cyclically under the same
displacement history as Test 10. Fig. 9.37c shows the measured Vb-b hysteresis loops.
The stiffening behavior at large displacements occurred as the compression angles came
into contact with the reaction and loading blocks. Failure occurred as a result of the
complete fracture of the seat angle at the north end of the beam during loading to b=+8%
for the third time. While the fracture of the angles in all of the previous tests occurred in
the vertical leg as shown in Fig. 9.36, the failure of the angle in Test 11 occurred in the
horizontal leg. As a result of the different angle failure mechanisms, it is questionable
how well the results from Test 11 represent the behavior of the angles in Test 10,
especially at large beam rotations.
246
9.3
Recommendations for Application
Based on the test results described above, it is concluded that properly-designed

unbonded post-tensioned hybrid coupled wall subassemblages (e.g., Test 3) offer
significant advantages for seismic regions, such as a self-centering capability and an
ability to undergo large nonlinear rotations without significant damage (except in the
sacrificial top and seat angles). The beam post-tensioning anchors and the angle-to-wall
and angle-to-beam connections are critical components that can adversely affect the
performance of these structures. Anchor types other than the steel wedge/barrel system
used in this experimental program may improve the performance of the post-tensioning
strands. It may also be possible to grout a short length of the post-tensioning tendons near
the anchors to prevent premature fracture of the strands inside the anchors. Kinking of the
post-tensioning strands during the experiments did not have any adverse effects on the
performance of the strands.
The slip critical angle-to-beam connections used in the test specimens worked well.
Slip at the angle-to-beam connections in Tests 7 and 9 was observed at relatively large
rotations; however, Test 10 showed that it is possible to prevent this slip by ensuring that
the total slip capacity of the connection is larger than the maximum tension force
developed in the angles. If the required slip capacity cannot be achieved using a bolted
angle-to-beam connection, it may also be possible to weld the angles to the beam flanges.
The post-tensioned angle-to-wall connections performed well, with no yielding in the
connection strands and no damage in the concrete.
The beam-to-wall connections should be designed to transfer the maximum
coupling shear forces without slip. Significant loss of the beam post-tensioning force in
Test 10 resulted in a loss in the friction resistance against slip at the beam ends. It may be
possible to increase the contribution of the top and seat angles against slip by fillet
welding the vertical legs of the angles to the walls. It is important that this weld does not
prevent the development of a ductile yield mechanism in the angle.
The experimental results showed that the use of cover plates on the coupling beam
flanges is not necessary for nominal flange initial stress values (see Table 9.1) of up to
75% of the steel yield strength, as long as compact sections are used for the beams. The
use of confinement reinforcement inside the wall contact regions and steel plates at the
beam-to-wall interfaces is necessary to prevent damage to the wall concrete. The shim
plates used in the experiments ensured contact between the beam flanges and the walls
throughout the cyclic displacement history; and thus, their use in practice is
recommended.
9.4
Chapter Summary
This chapter describes an experimental program on the nonlinear behavior of

unbonded post-tensioned hybrid coupled wall subassemblages. The test results
demonstrate that the proposed system can be designed to have excellent stiffness,
247
strength, ductility, and energy dissipation characteristics under large reversed cyclic
loading. The post-tensioning anchors and the angle-to-wall and angle-to-beam
connections are critical components that can adversely affect the performance of the
structure.
As a result of post-tensioning, the initial linear elastic stiffness of an unbonded
post-tensioned hybrid coupled wall subassemblage is similar to the initial stiffness of a
subassemblage with an embedded steel beam. The nonlinear deformations of properly
designed systems occur primarily as a result of the opening of gaps at the beam ends. The
post-tensioning force provides a restoring effect that closes these gaps and pulls the walls
and the beam back towards their original undisplaced position upon unloading from a
large nonlinear deformation, resulting in a self-centering capability. The structure can be
designed to provide significant and stable levels of coupling with most of the damage
occurring in the beam-to-wall connection angles, which can be replaced after an
earthquake. It is concluded that unbonded post-tensioned steel beams provide an effective
and feasible means to couple reinforced concrete walls.
248
CHAPTER 10
REVISED ANALYTICAL MODEL
This chapter makes modifications to the analytical models developed previously for
unbonded post-tensioned hybrid coupled wall subassemblages to better capture the
experimental measurements and observations from Chapter 9. These revised models are
used in the remainder of this report for the modeling, design, and analysis of coupled wall
structures. The chapter is divided into the following sections: (1) revised subassemblage
model; (2) evaluation of revised subassemblage model; and (3) revised idealized
moment-rotation relationship.
10.1 Revised Subassemblage Model
This section makes modifications to the fiber element coupled wall subassemblage
analytical model in Chapter 3, based on the experimental results from Chapter 9. The
revised model is shown in Figs. 10.1a and 10.1b. The modeling of the wall-height
elements and gap opening/closing behavior at the beam-to-wall interfaces remain
unchanged. The modeling of the coupling beams and wall-contact regions, the beam posttensioning tendons, and the top and seat angles are revised as follows. The assumptions
for the subassemblage model can be found in Chapter 3.
10.1.1 Modeling of Coupling Beams and Wall-Contact Regions
As described in Chapter 9, the use of shim plates at the beam-to-wall interfaces is
recommended to prevent contact between the coupling beam web and the walls, so that
one or both flanges of the beam remain in contact with each wall during the reversed
cyclic displacements of the structure, even after the yielding of the flanges in
compression. In practical applications, shim plates may also be used for
construction/tolerance purposes.
The experiments described in Chapter 9 show that the shim plates may have a
significant effect on the behavior of the subassemblages by limiting the contact depth at
the beam-to-wall interfaces. The previous analytical model shown in Fig. 3.3a and 3.3b
does not include the shim plates, allowing full contact between the beam and the walls at
zero rotation. Three changes are made in the revised model to account for the shim plates
(Fig. 10.1b): (1) the thickness of the shim plates is included in the model; (2) an
effective beam cross-section is used in the fiber elements near the beam ends instead of
the full cross-section; and (3) the effective cross-section of the wall-contact elements in
Fig. 3.3b is revised. The dotted regions in Fig. 10.1b show the effective cross-sections
used in the beam and wall-contact elements in the revised model, including the shim
plates.
249
wallheight
38 elements
beam element
52-54
wallcontact
elements 41
48,49
44,45
58-60
PT kink elements
39
42
50,51
40
55-57
kinematic
constraint
LEFT WALL REGION
lw
PT
46,47 element
vertical angle elem.
horizontal angle elem.
t sh
lb
43
61-63
shim
plate
RIGHT WALL REGION
tsh
lw
(a)
22
cover plate
23
15-18
beam
24 slope=3:4
25
11
13
39
9
5 7
3
1
8
12
2
10
4 6
14
26
27
midstory 28
slope=3:4
node
29
te
0.5l w1.5dbc d
bc
19
30
31
32
21 33
20
shim plate
42
34
35
36
37
0.5dbc
(b)
Fig. 10.1 Revised analytical model: (a) subassemblage;

(b) wall-contact elements and beam elements
The depth of the effective beam cross-section adjacent to the shim plates is
assumed to be equal to the depth (including the cover plates, if any) in contact with the
shim plates (i.e., the depth between the shim plates is not included in the effective beam
depth). The effective beam depth is assumed to increase away from the shim plates with a
slope of 3:4 as shown in Fig. 10.1b. The shim plates are modeled as part of the wallcontact elements. The effective depth for the wall-contact elements is determined with a
3:4 slope spread away from the shim plates, similar to the effective beam depth. The
compressive contact stresses in the beam and walls decrease away from the shim plates as
a result of an increase in the depth of the compression region. The increase in the depth of
the effective beam and wall cross-sections represents this increase in the depth of the
compression regions away from the shim plates.
250
The widths of the effective wall and beam cross sections are assumed to be equal to
the wall thickness and beam flange/web width, respectively. The beam elements include
second order effects due to the rotation of the beam with respect to the walls. The
discretization of the fiber elements along the length of the beam is flexible with the
following exceptions: (1) there are two beam nodes (Nodes 7 and 15 in Fig. 10.1b)
located at the same X-coordinates as the angle element nodes (the angle element nodes
are constrained to these beam nodes as described later); and (2) new fiber elements or
segments are used where the effective beam depth becomes equal to the full depth of the
beam and where the beam cover plates (if any) are terminated (in order to define the
change in beam geometry). Furthermore, the length of the first beam fiber element
segment adjacent to the shim plates is important in estimating the maximum compression
deformations that occur at the beam ends. Similar to the previous model in Chapter 3, a
length of lb,cr=2cb,sof (Fig. 3.6) is used in this research, where cb,sof is the estimated depth
of the compression (i.e., contact) region at the beam softening state.
Fiber wall-contact elements are used between the center of the left wall region
(Node 1) and the beam-to-shim-plate interface (Node 6). The Y-translational DOF of
Node 6 is kinematically constrained to Node 1. The rotational and X-translational DOFs
of Node 6 are not constrained. It is assumed that no slip occurs at the beam-to-shim-plate
interface. The discretization of the wall-contact elements between Nodes 1 and 6 is
flexible with the following exceptions: (1) the length of the first wall-contact element
segment adjacent to Node 6 is equal to the shim plate thickness; and (2) new fiber
elements or segments are used to define changes in the properties of the effective wall
cross-section (e.g., to model the wall embedded plates). The modeling of the right wall
region is similar to the left wall region.
10.1.2 Modeling of Beam Post-Tensioning Tendons
As shown in Fig. 10.1a, three truss elements (e.g., see elements between Nodes 3853-59-41) are used to model a beam post-tensioning tendon in the revised model, whereas
only one truss element is used to model each tendon in the previous model (Fig. 3.3a).
The objective of this modification is to model the kinking of the post-tensioning tendons
at the beam ends, which occurs when the displacement of the structure is large enough to
cause the tendons to come into contact with the post-tensioning ducts (Fig. 10.2a). As
observed in the subassemblage experiments described in Chapter 9, kinking of the beam
post-tensioning tendons results in an increase in the coupling resistance and stiffness due
to increased second order effects in the tendons.
In the revised analytical model, the truss elements representing a beam posttensioning tendon are connected to each other at nodes (e.g., Node 53) located at the
beam-to-wall interfaces. These nodes are free to move in the horizontal direction (since
the post-tensioning tendons are unbonded), but are kinematically constrained from
moving in the vertical direction by gap/contact PT kink elements above and below.
Each PT kink element is connected to a second node (e.g., Node 52), which is
kinematically constrained to a wall-height element node at the same elevation (Node 24).
251
loading block
reaction block
coupling
beam
kink
duct
PT
anchor
only one PT
tendon shown
(a)
force (kN)
45
initial gap
0
-40
displacement (mm)
(b)
Fig. 10.2 Kinking behavior: (a) idealized deformed shape;

(b) gap/contact PT kink element
The force-displacement relationship of the gap/contact PT kink elements is shown
in Fig. 10.2b. An initial gap is defined to model the distance between the outside of the
post-tensioning tendon and the inside of the post-tensioning duct. The initial gap is zero if
the ducts are not oversized with respect to the tendon diameter. The forces in the PT kink
elements are initially zero. Contact between the post-tensioning tendons and the ducts
occurs when the displacement of the structure is large enough to close the initial gap.
Once in contact, the PT kink elements constrain further vertical displacements of the
post-tensioning truss element nodes at the beam-to-wall interfaces, modeling the effect of
kinking. Second order effects are included in the truss elements to capture the increase in
the coupling resistance and stiffness of the subassemblage due to kinking.
252
10.1.3 Modeling of Top and Seat Angles

In the previous analytical model, the contribution of the top and seat angles to the
coupling resistance was represented using fiber elements placed parallel to the coupling
beam (see Chapter 3). Based on the experimental results from Chapter 9, three
modifications are made to this model: (1) zero-length translational spring elements are
used to model the angles instead of fiber elements; (2) the hysteretic model for the angle
elements placed parallel to the coupling beam is revised; and (3) the shear force in the
horizontal legs of the angles is modeled.
In the revised model, each angle is represented using two translational zero-length
spring elements in the X- and Y-directions, respectively (Fig. 10.1a). The first spring
element, which is placed in the horizontal (i.e., X) direction parallel to the beam,
represents the axial force in the horizontal leg of the angle and is referred to as the
horizontal angle element. The second element, referred to as the vertical angle
element, is in the vertical (i.e., Y) direction and represents the shear force in the angle
horizontal leg. Both elements are connected to the same pair of nodes (e.g., Nodes 44 and
45) with identical coordinates at the centroid of the angle-to-beam connection bolts and at
the same elevation as the middle of the angle horizontal leg thickness. It is assumed that
the angle-to-wall and angle-to-beam connections are properly designed for the maximum
angle forces and deformations. Based on this assumption, one of the angle nodes (e.g.,
Node 44) is kinematically constrained to a wall-height element node at the same
elevation (Node 23) and the other angle node (Node 45) is kinematically constrained to a
corresponding beam node (Node 7).
The behavior of an angle as it is loaded by the beam is governed by many factors
including the angle leg thickness, and the number, size, layout, and gage length of the
angle connectors. Fig. 10.3a shows the idealized deformed shape of a seat angle as it is
pulled and rotated by the coupling beam. It is assumed that the failure of the angle occurs
through the formation of two plastic hinges in the vertical leg. As described in Sims
(2000), other angle failure modes (e.g., an additional plastic hinge in the horizontal leg)
may be possible; however, this was not observed in the subassemblage experiments
presented in Chapter 9, and thus, is not modeled. The relatively short connection gage
length for the horizontal legs of the angles in the half-scale test specimens may have
played a role in the observed angle failure mode. Full scale subassemblage tests need to
be conducted to investigate the possibility of other angle failure mechanisms.
Fig. 10.3a shows the angle free body diagram between the plastic hinge adjacent to
the angle fillet on the vertical leg and the centroid of the angle-to-beam connector bolts
on the horizontal leg. Using equilibrium of the angle forces, it can be shown that:
Tayx = Vap
253
(10.1)
M ap + Vap (ka
Tayy =
Tayy
t
l gh a
2
Ma
ta
)
2
(10.2)
lgh
Tayx
ka
plastic hinge
Tayy
ta
Ma=0
Tayx
Vap
angle-to-wall
connector
Map
centroid of
Tayy angle-to-beam
connection bolts
(a)
angle force 2T ,5
ayx
ayx
Tayx,ayx
Kaixt
Kaixc
angle force
0.06Kaiy
Tayy,ayy Kaiy
TENSION
0
deformation
0 deformation
Kaxs
Casx
2Tayx
(b)
(c)
Fig. 10.3 Angle model: (a) angle displacements and forces;

(b) horizontal angle element; (c) vertical angle element
where, Tayx is the axial force in the angle horizontal leg, Tayy is the shear force in the
horizontal leg, Map is the plastic hinge moment and Vap is the plastic shear force in the
vertical leg including shear-flexure interaction, ka is the distance from heel to toe of fillet
of the angle, lgh is the gage length of the angle-to-beam connectors (measured from heel
of the angle to the centroid of the angle-to-beam connector bolts), and ta is the angle leg
254
thickness. The angle moment Ma at the centroid of the angle-to-beam connector bolts is
ignored.
Horizontal angle element force-deformation model
The force-deformation relationship of the horizontal angle element is shown in
Fig. 10.3b, where the hysteretic characteristics were determined based on the
subassemblage experiments as described later. A new type of zero-length spring element
was developed in DRAIN-2DX to model the hysteretic behavior in Fig. 10.3b. Under
tensile loading, the yield strength, Tayx=Vap [see Equation (10.1)] and initial stiffness,
Kaixt are determined the same as the previous model in Chapter 3 using a method
developed by Kishi and Chen (1990) and Lorenz et al. (1993). In this model, the vertical
leg is assumed to be fixed along the innermost edge of the line of angle-to-wall
connectors and is pulled horizontally by the coupling beam flange like a cantilever. The
rotation of the horizontal leg with respect to the vertical leg, which occurs as a result of
the rotation of the beam with respect to the walls, as shown in Fig. 10.3a, is ignored. The
yield strength, Tayx is reached when the two plastic hinges in Fig. 10.3a develop,
considering the interaction between the bending moment and shear force in the vertical
leg.
Based on the subassemblage experiments, it is assumed that the maximum strength
of the horizontal angle element in tension is equal to 2 times the yield strength, Tayx, and
is reached at an angle deformation of 5 times the yield deformation, ayx=Tayx/Kaixt.
Under compression, the initial stiffness of an angle as it is pushed back horizontally
toward the wall by the coupling beam is assumed to be equal to:
K aixc =
1 Ea Aa
K aixt
40 l gh
(10.3)
where, Ea is the Youngs modulus for the angle steel, Aa is the gross cross-section area of
the angle horizontal leg. The angle unloading stiffness from a tensile force is assumed to
be the same as the initial stiffness in tension, Kaixt, and the stiffness upon crossing the
zero-force axis is assumed to be equal to the shooting stiffness, Kaxs.
The angle force-deformation behavior is assumed to shoot towards the smallest of
the following three forces on the initial linear-elastic loading branch in compression: (1) a
compressive force with magnitude equal to the unloading force in tension; (2) the
compression force reached assuming Kaxs=Kaixt; and (3) the yield strength of the angle
in compression. The yield strength of the angle in compression is assumed to be equal to
the slip force, Casx of the angle-to-beam connection bolts. The development of the full
bearing capacity of the angle horizontal leg cross-section is not expected, and, is not
modeled since analyses and experiments of coupled wall subassemblages show that
extremely small deformations occur in an angle once the beam flange comes into contact
with the wall.
255
Note that slip of the angle-to-beam connector bolts can also occur when the angle is
pulled away from the wall (i.e., tension loading direction in Fig. 10.3b); however, this is
not a desirable type of behavior since it would be difficult to control the amount of slip.
In Fig. 10.3b, the slip critical capacity of the angle-to-beam bolts, Casx is larger than the
assumed angle capacity in tension, 2Tayx, and thus, slip does not occur in tension. A
design approach to achieve the desired behavior of the angles is described in Chapter 12.
Vertical angle element force-deformation model
The vertical angle element models the shear force in the angle horizontal leg using
an elasto-plastic force-deformation behavior as shown in Fig. 10.3c. The yield force
Tayy is determined from Equation (10.2), with Map and Vap calculated as recommended by
Kishi and Chen (1990) ignoring the rotation of the horizontal angle leg with respect to the
vertical leg. Assuming that Tayx and Tayy are reached at the same coupling beam chord
rotation and that the rotation of the beam occurs about the compression corner, the initial
stiffness, Kaiy of the vertical angle element can be determined as:
K aiy = K aixt
Tayy dbc
Tayx l gh
(10.4)
where, dbc is the depth of the coupling beam including the cover plates (if any). The postyield stiffness of the vertical angle element is assumed to be equal to 6% of the initial
stiffness, Kaiy.
Note that the modeling of the angles as described above assumes that the
contributions of the vertical and horizontal angle elements can be superposed, even
though this assumption is in general not valid in the nonlinear range. Note also that the
contribution of the vertical angle element to the subassemblage behavior is small as
compared with the horizontal angle element, and thus, can be ignored. The evaluation of
the revised analytical model below includes the vertical angle elements to achieve better
comparisons with the experimental results; however, these elements are ignored in the
multi-story analyses in the reminder of this report.
10.2 Evaluation of Revised Subassemblage Model
This section provides a critical evaluation of the revised analytical model based on
the subassemblage experiments in Chapter 9. Comparisons between experimentally
measured and analytically predicted behaviors are presented with respect to the following:
(1) coupling shear force versus chord rotation behaviors; (2) coupling beam posttensioning forces; (3) behavior at coupling beam ends; (4) behavior along beam span; and
(5) behavior of top and seat angles. The analytical models use the material test results
from Chapter 9.
256
10.2.1 Coupling Shear Force versus Chord Rotation Behaviors

Fig. 10.4 shows the predicted coupling shear force versus chord rotation (Vb-b)
behaviors for Tests 1, 3, 5, 7, 9, and 10 using the revised analytical model. The measured
behaviors in Chapter 9 are not plotted with the predicted results to maintain clarity in the
hysteresis curves. Note that the analytical cycle rotation amplitudes are very close to but
not the same as the actual rotations reached during each experiment. For comparison, Fig.
10.5a shows the predicted behavior for Test 3 using the previous model from Chapter 3.
The previous model is able to capture the envelope of the measured Vb-b behavior;
however, the behavior of the specimen during unloading is not represented well. The
force-deformation behavior of an angle element during Test 3 using the previous model
and the behavior of a horizontal angle element from the revised model are shown in Figs.
10.5b and 10.6a, respectively. The improved hysteretic characteristics of the revised
model in Fig. 10.4b as compared with Fig. 10.5a are primarily due to the differences in
the modeling of the angles as well as the other modifications presented previously.
For comparison, the force-deformation behavior of a vertical angle element from
the revised analytical model for Test 3 is shown in Fig. 10.6b; and Fig. 10.6c shows the
predicted Vb-b behavior for Test 3 using the revised analytical model with the vertical
angle elements removed. It is observed that the vertical angle elements have relatively a
small effect on the behavior of the structure.
The results in Fig. 10.4 indicate that the revised analytical model provides a good
representation of the global behavior of the test specimens across the entire parameter
range. The increase in the coupling resistance and stiffness due to the kinking of the posttensioning strands in Test 1 is captured. Note that this effect is not evident in the Vb-b
behaviors of the other test specimens since the kinking of the post-tensioning strands
occurs at large rotations (due to the use of oversized post-tensioning ducts) and the
increase in stiffness due to kinking is counteracted by the reduction in stiffness due to
deterioration in the top and seat angles. The failure of the test specimens (i.e., reduction
in the coupling resistance) due to low cycle fatigue fracture of the angles or due to
fracture of the post-tensioning strands is not captured by the analytical model. Slip at the
beam-to-wall interfaces or at the angle-to-wall and angle-to-beam connections is also not
included in the model.
257
300
300
0
10
0
10
(d)
-10
300
Test 5
Test 7
-300
-10
0
10
-10
0
10
300
300
(e)
-300
-300
(c)
-300
Test 3
Test 1
-300
-10
300
(b)
(a)
(f )
Test 9
-10
0
10
-300
Test 10
-10
0
Fig. 10.4 Coupling shear force versus rotation (Vb-b) behaviors: (a) Test 1;
(b) Test 3; (c) Test 5; (d) Test 7; (e) Test 9; (f) Test 10
258
10
300
Test 3
-300
-10
10
(a)
angle force (kN)
250
TENSION
Test 3
-250
-5
10
15
20
(b)
Fig. 10.5 Test 3 predictions using previous model:
(a) Vb-b behavior; (b) angle behavior
259
25
(a)
250
angle force (kN)
Tayx,ayx
2Tayx,5ayx
TENSION
Kaixt
0
Kaxs
Kaixc Casx
-250
-5
Test 3
10
15
20
25
(b)
250
angle force (kN)
TENSION
Tayy,ayy
Kaiy
0
Tayy,ayy
Test 3
250
4

300
(c)
without
vertical angle
elements
-300
Test 3
-10
10
Fig. 10.6 Test 3 predictions using revised model: (a) horizontal angle element; (b)
vertical angle element; (c) model with vertical angle elements removed
260
10.2.2 Coupling Beam Post-Tensioning Forces

Fig. 10.7 shows the predicted results for the total post-tensioning forces (i.e., sum
of the forces in the post-tensioning strands) from Tests 3, 9, and 10 using the revised
analytical model. The total post-tensioning force is normalized by the design maximum
strength of the strands, Pbu=abpfbpu, where abp=140 mm2 (0.217 in2) is the area of a single
strand and fbpu=1862 MPa (270 ksi) is the design maximum strength of the posttensioning steel. Comparisons of the predicted forces with the measured forces in Chapter
9 indicate that the analytical model is capable of predicting not only the increase in the
post-tensioning force due to the displacements of the subassemblage, but also the loss in
the post-tensioning force (upon unloading to zero displacement) due to nonlinear
behavior in the structure. The fracture of the post-tensioning tendons in Test 10 is not
modeled.
The results in Fig. 10.7 show that the analytical model overestimates slightly the
increase in the total beam post-tensioning force as the subassemblage is displaced. To
investigate this difference, Fig. 10.8 compares the measured and predicted displacements
of the loading block in the horizontal direction as the beam is rotated. The positive
displacements indicate that the loading block moves away from the reaction block. These
horizontal displacements, which occur primarily due to the opening of gaps at the beamto-wall interfaces, are directly related to the total elongations of the post-tensioning
strands, and thus, the post-tensioning forces. Note that the measured displacements in Fig.
10.8 are upon first loading to peak b values during the loading history followed in each
test. The displacements corresponding to the same b amplitude in the positive and
negative loading directions are averaged. The comparisons indicate that the analytical
model overestimates the horizontal displacements of the loading block at large rotations,
and thus overestimates the elongations of the post-tensioning tendons, resulting in the
overestimated post-tensioning forces in Fig. 10.7.
10.2.3 Behavior at Coupling Beam Ends
To evaluate the capability of the revised analytical model in predicting the behavior
at the coupling beam ends, Fig. 10.9 compares the measured and predicted results for the
average compressive strains in the beam contact regions from Tests 1, 3, 5, 7, 9, and 10.
Beam cover plate (for Test 1) and flange (for Tests 3, 5, 7, 9, and 10) strain gages SG5SG22 were used for this purpose, with locations as given in Chapter 9.
The procedure to calculate the average strains from the measured data is
demonstrated using strain gages SG15-SG22 in Fig. 10.10. The following steps are used:
(1) for each strain gage, take the average of the peak compression strain values (occurring
at peak b values; e.g., see markers in Fig. 10.10a for SG22 during Test 1) from the
three cycles during each set of displacement cycles; (2) take the average of the average
peak compression strains from the gages located at the same distance from the beam end
(i.e., average from strain gages SG15, SG16, SG19 and SG20 located 6.35 mm from the
beam end, and strain gages SG17, SG18, SG21 and SG22 located 25.35 mm from the
261
beam end, as shown in Fig. 10.10b); (3) assume a uniform strain distribution over the
influence length for each set of strain gages (i.e., uniform strain distribution over a
length of 15.85 mm for SG15, SG16, SG19 and SG20 and over a length of 19 mm for
SG17, SG18, SG21 and SG22); (4) use the uniform strain distributions to calculate the
total compressive deformation at the end of the beam (over a total length of 34.85 mm for
strain gages SG15-SG22 in Fig. 10.10b); and (5) calculate the average strains by dividing
the total deformation with the total length considered (i.e., 34.85 mm in Fig. 10.10b).
The predicted average strains for the same length of cover plate or flange are
determined using the analytical model. The comparisons in Fig. 10.9 indicate that, while
there are discrepancies between the measured and predicted values, the analytical model
is capable of providing a satisfactory representation of the contact behavior at the beam
ends. No attempt is made to compare the measured and predicted strains at individual
gage locations, since the measured strains varied considerably due to uneven contact at
the beam ends.
10.2.4 Behavior Along Beam Span
To evaluate the capability of the revised analytical model in predicting the behavior
of the coupling beams along the span, Fig. 10.11 compares the measured and predicted
average maximum and minimum compressive strain values in strain gages SG1-SG4
from Tests 1, 3, 5, 7, 9, and 10. The locations of strain gages SG1-SG4 are given in
Chapter 9.
Note that when b is positive (i.e., clockwise rotation), maximum compression
strains of similar value are measured in gages SG2 and SG3, and minimum compressive
strains of similar value are measured in gages SG1 and SG4. This is because of the
symmetric double curvature loading condition that develops along the length of the beam
(with zero curvature at beam midspan) as the subassemblage is displaced laterally. As the
beam is cycled in the positive and negative loading directions, the maximum and
minimum readings in these gage pairs alternate. The measured strains in Fig. 10.11 are
calculated as the average maximum and minimum readings from all four gages upon first
loading to the peak b values in the positive and negative directions during the
displacement history followed in each test (see Fig. 10.12). The comparisons indicate that,
while there are discrepancies between the measured and predicted results, the analytical
model provides reasonable estimates of the beam web strains along the span.
262
Pb /a bpf bpu
(a)
Test 3
-10
10
Pb /a bpf bpu
(b)
0
Test 9
0
-10
0
1
Pb /a bpf bpu
(c)
10
Test 10
0
-10
0
10
Fig. 10.7 Post-tensioning forces: (a) Test 3; (b) Test 9; (c) Test 10
263
(a)
25
experiment
prediction
horizontal displacement
of loading block (mm)
Test 3
(b)
25
10
experiment
prediction
Test 9
0
0
(c)
25
10
experiment
prediction
Test 10
0
0
10
Fig. 10.8 Horizontal displacements of loading block: (a) Test 3; (b) Test 9; (c) Test 10
264
experiment
prediction
-0.06
0
Test 1
experiment
prediction
-0.06
0
10
beam chord rotation,

b (percent)
(b)
compressive strain
compressive strain
(a)
Test 3
10

b (percent)
experiment
prediction
-0.06
0
Test 5
Test 7
10

b (percent)
experiment
prediction
compressive strain
(e)
compressive strain
-0.06
0
experiment
prediction
-0.06
10

b (percent)
(d)
compressive strain
compressive strain
(c)
Test 9

b (percent)
10
-0.06
0
(f )
experiment
prediction
Test 10

b (percent)
Fig. 10.9 Beam cover plate/flange strains: (a) Test 1; (b) Test 3;
(c) Test 5; (d) Test 7; (e) Test 9; (f) Test 10
265
10
Fig. 10.10 Average compressive strains at beam end: (a) SG22 from Test 1;
(b) calculation of average strain
266
(a)
minimum
compressive strain
compressive strain
minimum
maximum
experiment
prediction
-0.0015
0
Test 1
maximum
experiment
prediction
-0.0015
0
10

b (percent)
compressive strain
compressive strain
maximum
-0.0015
0
Test 5
compressive strain
compressive strain

b (percent)
Test 9
experiment
prediction
0
Test 10

b (percent)
Fig. 10.11 Beam web strains: (a) Test 1; (b) Test 3; (c) Test 5;
(d) Test 7; (e) Test 9; (f) Test 10
267
(f )
maximum
-0.0015
10
10

b (percent)
minimum
maximum
-0.0015
Test 7
minimum
experiment
prediction
experiment
prediction
-0.0015
(e)
(d)
maximum
10

b (percent)
10

b (percent)
minimum
minimum
experiment
prediction
Test 3
(c)
(b)
10
positive b
beam
SG1
wall
SG3
SG2
wall
SG4
beam
negative b
wall
SG2
SG4
SG1
SG3
average maximum
=
compression strain
(SG2+SG3)at positive
average minimum
=
compression strain
(SG1+SG4)at positive
wall
+(SG1+SG4)at negative
4
b
+(SG2+SG3)at negative
Fig. 10.12 Calculation of average compressive strains in beam web

10.2.5 Behavior of Top and Seat Angles
Fig. 10.13 shows the measured versus predicted Vb-b behaviors from Tests 4, 6,
and 8. As described in Chapter 9, these experiments were conducted under monotonic
loading using subassemblages with two angles in tension and no beam post-tensioning
strands. The predicted curves up to the yield point of the angles ( markers in Fig.
10.13) indicate that the method developed by Kishi and Chen (1990) provides reasonable
estimates for the initial stiffness, Kaixt and yield strength, Tayx of the test specimens. The
post-yield stiffnesses of the angles under monotonic loading were determined from the
experimental results in Fig. 10.13.
268
80
0
0
angle "yield"
Test 4
Test 6
Test 8
experiment
prediction
10
Fig. 10.13 Angle behavior under monotonic loading - Tests 4, 6, and 8

Under reversed cyclic loading, the post-yield behavior of the top and seat angles is
complicated due to kinematic hardening of the steel and low cycle fatigue of the angle
vertical legs. Low cycle fatigue of the angles did not occur in Tests 4, 6, and 8 since the
loads were applied monotonically. Furthermore, the angles in Test 11, which was
conducted under reversed cyclic loading, formed a different failure mechanism than the
other cyclic tests as described in Chapter 9. In the absence of a reliable method to predict
the post-yield behavior of the angles under reversed cyclic loading, the following
approach was taken.
As described in Chapter 9, the only significant difference between Tests 1 and 3 is
the use of four top and seat angles at the beam ends. The contribution of these angles to
the coupling resistance was determined by subtracting the Vb-b relationship of Test 1
from that of Test 3 as shown in Fig. 10.14a. The resulting Vb-b behavior for a cycle of
loading to b=6% is shown by the thick line in Fig. 10.14b.
For comparison, the predicted difference between Tests 1 and 3 using the angle
models in Figs. 10.3b and 10.3c is shown by the thin line in Fig. 10.14b. The results
indicate that the revised angle model provides a reasonable representation of the behavior
of the angles under cyclic loading. Failure of the angles due to low cycle fatigue is not
modeled.
269
300
-300
-10
Test 3
Test 1
0
10
(a)
100
-100
-8
experiment
prediction
0
(b)
Fig. 10.14 Angle behavior under cyclic loading: (a)Tests 1 and 3;
(b) difference between Tests 1 and 3 during b=6% cycle
10.3 Revised Idealized Moment-Rotation Relationship
This section makes modifications to the idealized bilinear subassemblage beam end
moment versus chord rotation (Mb-b) relationship in Chapter 5 based on the revised
270
subassemblage model described above. The most significant update to the idealized
moment-rotation relationship is the use of the horizontal angle element force-deformation
model in Fig. 10.3b. This modification is necessary to better capture the measured
behavior from the subassemblage experiments as demonstrated in Figs. 10.5 and 10.6.
The contribution of the vertical angle elements (see Fig. 10.3c) to the subassemblage
behavior is ignored.
The revised moment-rotation relationship is identified by the following two limit
states, similar to the previous relationship in Chapter 5:
(1) Beam softening state (at Mb,sof, b,sof).
(2) Beam PT-yielding state (at Mb,pty, b,pty).
10.3.1 Beam Softening State
The coupling beam end moment, Mb,sof and chord rotation, b,sof at the beam
softening state of an unbonded post-tensioned hybrid coupled wall subassemblage are
estimated as described in Chapter 5 with the following modifications:
(1) The force in each tension angle is equal to 1.5Tayx, where Tayx is the yield
strength of the horizontal angle elements as shown in Fig. 10.3b.
(2) The force in each compression angle is equal to (1/40)fayAaCasx , where fay is
the angle steel yield strength and Casx is the slip force for the angle-to-beam connector
bolts as shown in Fig. 10.3b.
The subassemblage initial stiffness, Kbi can be determined ignoring the
deformations in the wall-contact regions as described in Chapter 5.
10.3.2 Beam PT-Yielding State
The coupling beam end moment, Mb,pty and chord rotation, b,pty at the beam PTyielding state of an unbonded post-tensioned hybrid coupled wall subassemblage are
estimated as described in Chapter 5 with the following modifications:
(1) The maximum strength of the angles in tension is reached and maintained (i.e.,
angle failure does not occur) before the yielding of the beam post-tensioning tendons.
Thus, the force in each tension angle at the PT-yielding state is equal to 2Tayx, where
2Tayx is the maximum strength of the horizontal angle elements as shown in Fig. 10.3b.
(2) The force in each compression angle is equal to Casx, where Casx is the slip force
for the angle-to-beam connector bolts as shown in Fig. 10.3.
The steps used in the estimation of Mb,pty and b,pty are presented below. Note that
different from Chapter 5, the revised procedure does not require an idealized load271
deformation relationship for the angles in tension since the force in each tension angle is
assumed to be equal to 2Tayx. Thus, there is no need for iteration to estimate Mb,pty and
b,pty.
Step 1
Estimate the total beam compression force at the beam-to-wall interface as:
Cb, pty = Pb, pty + 2Tayx Casx
(10.5)
where, Pb,pty is the total force in the post-tensioning tendons. As described in Chapter 5,
the post-tensioning tendons go through the same elongation regardless of their locations
within the beam depth. Thus, the tendons yield simultaneously and
Pb, pty = Pby = abt fbpy
(10.6)
Step 2
Estimate the depth of the compression (i.e., contact) region, cb,pty at the beam-towall interface using a uniform contact stress distribution. Assuming that the compression
region is above the coupling beam web and the width of the cover plate (if used) is the
same as the beam flange width bbf, then, cb,pty can be estimated as:
cb, pty =
Cb, pty
fby bbf
(10.7)
A non-rectangular compression region should be used if cb,pty>tbf+tc. The cover plate

thickness tc=0 if no cover plates are used on the coupling beam.
Step 3
Estimate the coupling beam end moment at the beam PT-yielding state, Mb,pty by
taking moments about the centerline of the beam as:
M b, pty = Vb, pty
lb
= 0.5Cb, pty (dbc cb, pty ) + 0.5(Casx + 2Tayx )(dbc + ta )
2
(10.8)
Step 4
Estimate the elongation of the beam post-tensioning tendons between the initial
prestress state and the PT-yielding state as:
ub, pty =
( fbpy fbpi )lbpu

Ebp
(10.9)
where, lbpu=lb+2lw is the unbonded length and Ebp is the Youngs modulus of the posttensioning tendons.
272
Step 5
Estimate the width of the beam-to-wall gap at the centerline of the beam, gb,pty.
Fig. 5.5 shows the gap opening that occurs at the beam-to-wall interface at the PTyielding state. Using symmetry,
gb, pty =
ub, pty
2
(10.10)
Step 6
Estimate the rotation of the beam due to gap opening at the PT-yielding state as:
bg , pty =
gb, pty
0.5dbc cb, pty
(10.11)
where, gb,pty and cb,pty are determined in Steps (5) and (2), respectively.
Step 7
Estimate the elastic rotation of the beam at the PT-yielding state using the
subassemblage linear-elastic stiffness, Kbi as:
M b, pty
be, pty =
K bi
(10.12)
Step 8
Estimate the total beam chord rotation at the PT-yielding state as:
b, pty = 1.075(bg , pty + be, pty )
(10.13)
where, the factor 1.075 is described in Chapter 5, and bg,pty and be,pty are from Steps (6)
and (7), respectively.
Step 9
Consider second order effects on Mb,pty as described in Chapter 5.
Step 10
Estimate the coupling shear force at the beam PT-yielding state, Vb,pty, as:
Vb, pty =
2 M b, pty
273
lb
(10.14)
10.4 Chapter Summary
This chapter uses the experimental results from Chapter 9 to make necessary
modifications to the fiber element subassembly analytical model described in Chapter 3.
The revised analytical model is critically evaluated by comparing the analytical
estimations with the experimental measurements. The idealized coupling beam end
moment versus chord rotation relationship developed in Chapter 5 is also modified based
on the experimental results.
274
CHAPTER 11
PERFORMANCE-BASED SEISMIC DESIGN APPROACH
This chapter proposes a performance-based seismic design approach for multi-story
unbonded post-tensioned hybrid coupled wall structures. The chapter is divided into the
following sections: (1) overview; (2) performance (damage) levels; (3) structure limit
states and capacities; (4) seismic demand levels; (5) structure demands; (6) relationships
between performance levels and limit states; (7) design objectives; (8) seismic design
criteria; (9) ground motion response spectra; (10) equivalent nonlinear SDOF
representation; (11) seismic displacement demand relationships; (12) nonlinear static
procedure; and (13) maximum base shear demand. The design of two prototype structures
using the proposed design approach is presented in Chapter 12.
11.1 Overview
Performance-based seismic design refers to a noble concept that aims to allow the
designer to specify and predict, with reasonable accuracy, the performance (degree of
damage) of a structure for a specified level of ground motion intensity. Over the last three
decades following the 1971 San Fernando Earthquake in California, the need for
performance-based seismic design has been recognized. For example, greater emphasis
has been given to damage control by controlling drift, and the concept of an importance
factor has been introduced for critical structures that should remain functional after a
ground motion. It has also been well recognized that increasing the lateral strength of a
structure alone may not necessarily enhance safety nor reduce damage. For example, in
the seismic design of a frame building, ensuring a weak beam/strong column yield
mechanism is much more important than increasing the total lateral strength of the
structure.
Over the past 10-15 years, significant research has been conducted on performancebased seismic engineering. Two of the more recent developments in this area are as
follows:
(1) Recognizing the importance of limiting the nonlinear lateral displacements of a
structure during an earthquake, there is a gradual shift from force-based design
approaches to displacement-based approaches.
(2) The use of nonlinear pushover analyses in seismic design is becoming more
standard practice, especially, for unusual or complex structures, or structures of higher
importance.
275
Even though the idea of performance-based engineering is not new and significant
progress has been made in recent years, the application of performance-based seismic
design concepts in model U.S. building codes is limited, primarily, because of lack of
reliable methods to assess seismic demands and structure capacities with reasonable
certainty. The FEMA-356 (ASCE 2000) prestandard and commentary applies some of the
latest methodologies in performance-based seismic engineering to achieve various levels
of seismic performance in the seismic rehabilitation of existing buildings.
The performance-based seismic design and evaluation guidelines in FEMA-356 are
comprised of three basic components: (1) performance objectives categorized by four
discrete structural performance levels and two intermediate structural performance ranges;
(2) seismic demand prediction using one of four alternative analysis procedures; and (3)
acceptance criteria using force and/or deformation limits that are meant to satisfy the
desired performance objective(s).
The seismic design approach proposed in this report for unbonded post-tensioned
hybrid coupled wall structures utilizes similar concepts as the provisions in FEMA-356,
focusing on a displacement-based procedure. The components used in this seismic design
approach are shown in Fig. 11.1 (Kurama et al. 1997) as: (1) performance (damage)
levels; (2) structure limit states and capacities; (3) seismic demand levels; and (4)
structure demands. The performance levels are related to the seismic demand levels
through design objectives and the structure capacities are related to the structure
demands through design criteria as described below.
PERFORMANCE (DAMAGE)
DESIGN
OBJECTIVES
SEISMIC DEMAND
LEVELS
LEVELS
QUANTIFICATION
OF PERFORMANCE
STRUCTURE LIMIT STATES
QUANTIFICATION
OF DEMAND
DESIGN
CRITERIA
AND
STRUCTURE
DEMANDS
CAPACITIES
Fig. 11.1 Outline of proposed performance-based seismic design approach

(adapted from Kurama 1997)
276
11.2 Performance (Damage) Levels

Performance-based seismic design requires the identification of specific seismic
performance levels. A performance level is defined in terms of the damage expected in
various structural and non-structural members during a ground motion. In general, less
damage corresponds to better performance. The proposed design approach uses three
performance levels in order of increasing damage as: (1) Immediate Occupancy; (2) Life
Safety; and (3) Collapse Prevention. These performance levels are described below.
11.2.1 Immediate Occupancy Performance Level
The Immediate Occupancy performance level refers to a post-earthquake damage
state in which only very limited structural damage has occurred. The basic vertical- and
lateral-force-resisting systems of the building retain nearly all of their pre-earthquake
strengths and stiffnesses. Although some minor structural repairs may be appropriate,
these would generally not be required prior to re-occupancy. The risk of life-threatening
injury as a result of structural damage is very low.
11.2.2 Life Safety Performance Level
The Life Safety performance level refers to a post-earthquake damage state in
which significant damage to the structure has occurred but an adequate margin against
either total or partial collapse remains. Some structural members and components may be
severely damaged, but this does not result in large falling debris hazards, either inside or
outside the building. Injuries may occur during the earthquake; however, the overall risk
of life-threatening injury as a result of structural damage is low. It would be possible to
repair the structure. While the damaged structure is not an imminent collapse risk, it
would be prudent to implement structural repairs or install temporary bracing prior to reoccupancy.
11.2.3 Collapse Prevention Performance Level
The collapse prevention performance level refers to a post-earthquake damage state
in which the structure is on the verge of partial or total collapse. Substantial damage to
the structure has occurred, potentially including significant degradation in the stiffness
and resistance of the lateral-force system. Degradation in the vertical load capacity of the
structure is possible; however, all major components of the gravity-load system continue
to carry their gravity load demands. Significant risk of injury due to falling hazards from
structural debris may exist. The structure may not be practical to repair and may not be
safe for reoccupancy, as aftershock activity could induce collapse.
11.3 Structure Limit States and Capacities
Performance-based seismic design requires the identification of limit states to
describe and quantify the damage in various members or components of the structure. For
277
each limit state, a force or deformation limit is defined as the corresponding capacity. The
structure limit states and capacities depend on the type of structure. For example, the
limit states for a steel frame structure may include yielding, buckling, and fracture of the
beam and column members, whereas the limit states for a reinforced concrete frame may
include yielding of the reinforcement and crushing of the concrete. The limit states and
capacities that are considered in the seismic design of unbonded post-tensioned hybrid
coupled wall structures are given below.
It is assumed that the lateral load behavior of the coupled wall structure is governed
by the axial-flexural behavior of the walls and gap opening behavior at the beam-to-wall
joints. The desired failure mode of the structure occurs through the crushing of the
confined concrete at the wall base after the yielding of the wall flexural reinforcement.
The following undesirable limit states should be prevented by design: (1) shear failure of
the walls, including diagonal tension failure and shear slip at the base; (2) local and/or
global instability of the walls or the coupling beams; (3) failure of the concrete wallcontact regions adjacent to the coupling beams; (4) fracture or anchorage failure of the
wall reinforcement; (5) fracture of the top and seat angles at the beam-to-wall
connections; (6) slip of the beams with respect to the walls; (7) failure of the angle-tobeam or angle-to-wall connections; and (8) failure of the post-tensioning anchorages or
fracture of the post-tensioning steel.
11.3.1 Steel Coupling Beam Limit States
As described in Chapters 4 and 5, the following limit states are identified for
properly-designed unbonded post-tensioned steel coupling beams:
(1) Decompression (i.e., initiation of gap opening) at the beam-to-wall interfaces.
(2) Yielding of the top and seat angles in tension.
(3) Yielding of the beam flange cover plates (if any) in compression.
(4) Yielding of the beam flanges in compression.
(5) Yielding of the beam post-tensioning tendons.
The following two states are used to identify an idealized bilinear lateral force
versus deformation relationship for the beams (see Chapters 5 and 10):
Beam Softening State This state identifies a significant reduction in the lateral
stiffness of the beam due to, primarily, a combined effect of increased gap opening and
yielding of the angles in tension. The beam end moment and chord rotation at the
softening state are referred to as Mb,sof and b,sof, respectively.
278
Beam PT-Yielding State This state corresponds to the yielding of the beam posttensioning tendons. It is assumed that at the beam PT-yielding state, the maximum strain
in the beam post-tensioning tendons is equal to the linear limit strain of the posttensioning strand stress-strain relationship. The beam end moment and chord rotation at
the PT-yielding state are referred to as Mb,pty and b,pty, respectively.
11.3.2 Concrete Wall Pier Limit States
As described in Chapters 7 and 8, both monolithic cast-in-place reinforced concrete
wall piers and precast concrete wall piers are considered in this report.
Cast-in-place concrete wall piers
The limit states identified for a properly-designed cast-in-place reinforced concrete
wall are:
(1) Cracking of concrete at the base of the tension-side wall (i.e., left side wall for
lateral loads applied from left to right).
(2) Spalling of unconfined (i.e., cover) concrete at the base of the compression-side
wall (i.e., right side wall for lateral loads applied from left to right).
(3) Softening of the tension-side wall. This state is defined as the state when the
neutral axis at the base of the tension-side wall reaches the centerline of the wall.
(4) Softening of the compression-side wall. This state is defined as the state when
the neutral axis at the base of the compression-side wall reaches the centerline of the
wall.
(5) Yielding of the flexural mild steel reinforcement at the base of the tension-side
wall.
(6) Crushing of the confined concrete at the base of the compression-side wall.
The following two states are used to identify an idealized bilinear lateral load
versus displacement relationship for coupled wall structures with cast-in-place concrete
walls (see Chapter 8):
Coupled Wall Softening State This state corresponds to a significant reduction in
the lateral stiffness of the coupled wall system. It is assumed that at the coupled wall
softening state, all of the coupling beams have reached their softening states (see above
for the definition of the beam softening state), the extreme flexural mild steel
reinforcement in the tension-side wall has reached the yield strain, and the neutral axis at
the base of compression-side wall is at the centerline of the wall (i.e., the compression
279
side wall has softened). The coupled wall base shear force and roof drift at the wall
softening state are referred to as Fws and ws, respectively.
Coupled Wall Ultimate State This state corresponds to the crushing of the
confined concrete at the base of the compression-side wall. It is assumed that at the
coupled wall ultimate state, the maximum compression strain in the confined concrete at
the base of the compression-side wall is equal to the ultimate strain of the confined
concrete. The coupled wall base shear force and roof drift at the wall ultimate state are
referred to as Fwu and wu, respectively.
Precast concrete wall piers
The limit states identified for a properly-designed precast concrete wall are:
(1) Decompression at the base of the tension-side wall (i.e., left side wall for lateral
loads applied from left to right).
(2) Spalling of unconfined (i.e., cover) concrete at the base of the compression-side
wall (i.e., right side wall for lateral loads applied from left to right).
(3) Softening of the tension-side wall. This state is defined as the state when the
neutral axis at the base of the tension-side wall reaches the centerline of the wall.
(4) Softening of the compression-side wall. This state is defined as the state when
the neutral axis at the base of the compression-side wall reaches the centerline of the
wall.
(5) Yielding of the post-tensioning steel in the tension-side wall.
(6) Crushing of the confined concrete at the base of the compression-side wall.
The following three states are used to identify an idealized tri-linear lateral force
versus displacement relationship for coupled wall structures with precast concrete walls
(see Chapter 8):
Coupled Wall Softening State This state corresponds to a significant reduction in
the lateral stiffness of the coupled wall system. It is assumed that at the coupled wall
softening state, all of the coupling beams have reached their softening states (see above
for the definition of the beam softening state), the maximum concrete compression stress
at the base of the tension-side wall is equal to the linear limit stress of the confined
concrete, and the neutral axis at the base of compression-side wall is at the centerline of
the wall (i.e., the compression-side wall has softened). The coupled wall base shear force
and roof drift at the wall softening state are referred to as Fws and ws, respectively.
280
Coupled Wall PT-Yielding State This state corresponds to the yielding of the wall
post-tensioning steel. It is assumed that at the coupled wall PT-yielding state, the
maximum strain in the wall post-tensioning bars (which occurs in the tension-side wall)
is equal to the linear limit strain of the post-tensioning bar stress-strain relationship. The
coupled wall base shear force and roof drift at the wall PT-yielding state are referred to as
Fwy and wy, respectively.
Coupled Wall Ultimate State This state corresponds to the crushing of the
confined concrete at the base of the compression-side wall. It is assumed that at the
coupled wall ultimate state, the maximum compression strain in the confined concrete at
the base of the compression-side wall is equal to the ultimate strain of the confined
concrete. The coupled wall base shear force and roof drift at the wall ultimate state are
referred to as Fwu and wu, respectively.
11.4 Seismic Demand Levels
In addition to performance levels and structure limit states and capacities,
performance-based seismic design requires a set of seismic demand levels. These seismic
demand levels are generally defined in terms of selected levels of ground motion (with
selected return periods) for a given site. Two seismic demand levels are considered in this
report using a design level ground motion and a survival level ground motion.
The survival level ground motion is the same as the maximum considered
earthquake level in IBC 2000 (ICC 2000), which has a 2% probability of being
exceeded in 50 years, approximately corresponding to a 2500-year return period. The
acceleration response spectrum for the design-level ground motion is determined by
multiplying the spectrum for the survival-level ground motion by a factor of 2/3 [referred
to as the seismic margin in the NEHRP provisions (BSSC 1998)]. This demand level
roughly corresponds to an earthquake with a 10% probability of being exceeded in 50
years, or a return period of approximately 500 years, for coastal California and a lower
probability of occurrence (a return period of approximately 1400 years) for the eastern
United States (BSSC 1998).
11.5 Structure Demands
Structure demands quantify lateral displacement (e.g., roof drift) and force (e.g.,
base shear) requirements of a structure for the selected seismic demand levels. The
structure demands considered in the design of an unbonded post-tensioned hybrid
coupled wall system include:
(1) The survival-level coupled wall roof drift demand, s.
(2) The design-level coupled wall roof drift demand, d.
(3) The survival-level coupling beam chord rotation demand, s.
281
(4) The design-level coupling beam chord rotation demand, d.

(5) The coupled wall design base shear force demand, Qwd, as determined from a
design ground motion acceleration spectrum based on the first mode response of the
structure.
(6) The maximum coupled wall base shear force demand, Qw,max, including the
effect of higher modes.
The estimation of these structure demands is discussed in detail later in this chapter.
11.6 Relationships Between Performance Levels and Limit States
Performance-based design requires the development of relationships between the
target performance levels and the structure limit states. These relationships are necessary
to determine when the structure reaches a specified performance level. The following
relationships are used in the design of unbonded post-tensioned hybrid coupled wall
structures.
11.6.1 Steel Coupling Beams
Immediate Occupancy performance levelThis performance level is assumed to
be reached when the beam softening state is reached (i.e., at Mb,sof, b,sof).
Life Safety performance levelThis performance level is assumed to be reached
when the beam chord rotation reaches 2/3 times the rotation at the beam PT-yielding state
(i.e., at 2b,pty/3).
Collapse Prevention performance levelThis performance level is assumed to
be reached when the beam PT-yielding state is reached (i.e., at Mb,pty, b,pty).
These three performance levels are shown on the typical idealized end moment
versus chord rotation (Mb-b) relationship of an unbonded post-tensioned steel coupling
beam in Fig. 11.2a.
11.6.2 Cast-in-Place Concrete Wall Piers
be reached when the coupled wall softening state is reached (i.e., at Fws, ws).
when the wall roof drift reaches 2/3 times the roof drift at the coupled wall ultimate state
(i.e., at 2wu/3).
282

be reached when the coupled wall ultimate state is reached (i.e., at Fwu, wu).
These three performance levels are shown on the typical idealized base shear force
versus roof drift (F-) relationship of an unbonded post-tensioned hybrid coupled wall
system with cast-in-place concrete wall piers in Fig. 11.2b.
beam end moment, M b
Survival-Level Demand
( s )
Design-Level Demand
( d)
Life Safety
(2/3)b,pty
Immediate Occupancy
( M b,sof , b,sof )
Collapse Prevention
(M b,pty , b,pty)
beam chord rotation, b
(a)
(s)
(s)
Design-Level Demand
( d)
Collapse Prevention
(Fwu, wu)
Life Safety
(2/3) wu
Immediate Occupancy
(Fws , ws )
base shear force, F
base shear force, F
Design-Level Demand
( d)
Life Safety
(Fwy , wy)
Collapse Prevention
(Fwu, wu)
Immediate Occupancy
(Fws , ws )
CIP-UPT system
roof drift,
(b)
roof drift,
(c)
Fig. 11.2 Limit states, performance levels, and design objectives: (a) steel coupling beam;
(b) coupled wall system with cast-in-place concrete wall piers;
(c) coupled wall system with precast concrete wall piers
11.6.3 Precast Concrete Wall Piers
be reached when the coupled wall softening state is reached (i.e., at Fws, ws).
283

when the coupled wall PT-yielding state is reached (i.e., at Fwy, wy).
be reached when the coupled wall ultimate state is reached (i.e., at Fwu, wu).
These three performance levels are shown on the typical idealized base shear force
versus roof drift (F-) relationship of an unbonded post-tensioned hybrid coupled wall
system with precast concrete wall piers in Fig. 11.2c.
11.7 Design Objectives
Design objectives relate the target performance levels to the seismic demand levels
as shown in Fig. 11.1. The design approach developed in this report has two objectives as
follows:
(Design Objective 1) To achieve the Collapse Prevention performance level under
the survival demand level; and
(Design Objective 2) To achieve the Life Safety performance level under the design
demand level.
11.7.1 Coupling Beams
As shown in Fig. 11.2a, the design objectives for the coupling beams are assumed
to be achieved if the following conditions are met:
(1) The coupling beam chord rotation capacity at the beam PT-yielding state is
larger than the survival-level chord rotation demand Design Objective 1. Thus,
b, pty > s
(11.1)
(2) 2/3 times the coupling beam chord rotation capacity at the beam PT-yielding
state is larger than the design-level chord rotation demand Design Objective 2. Thus,
2
b, pty > d
3
(11.2)
(3) All other failure modes (e.g., shear failure, instability, fracture of top and seat
angles) are prevented.
11.7.2 Systems with Cast-in-Place Concrete Wall Piers
As shown in Fig. 11.2b, the design objectives for the cast-in-place concrete wall
piers are assumed to be achieved if the following conditions are met:
284
(1) The coupled wall roof drift capacity at the wall ultimate state is larger than the
survival-level wall roof drift demand Design Objective 1. Thus,
wu > s
(11.3)
(2) 2/3 times the coupled wall roof drift capacity at the wall ultimate state is larger
than the design-level wall roof drift demand Design Objective 2. Thus,
2
wu > d
3
(11.4)
(3) All other failure modes (e.g., diagonal tension failure, shear slip) are prevented.
11.7.3 Systems with Precast Concrete Wall Piers
As shown in Fig. 11.2c, the design objectives for the precast concrete wall piers can
be achieved if the following conditions are met:
(1) The coupled wall roof drift capacity at the wall ultimate state is larger than the
survival-level wall roof drift demand Design Objective 1. Thus,
wu > s
(11.5)
(2) The coupled wall roof drift capacity at the wall PT-yielding state is larger than
the design-level wall roof drift demand Design Objective 2. Thus,
wy > d
(11.6)
(3) All other failure modes (e.g., diagonal tension failure, shear slip) are prevented.
11.8 Seismic Design Criteria
As indicated in Fig. 11.1, seismic design criteria (or acceptance criteria) specify
required comparisons between structure capacities and structure demands. If the structure
capacities are larger than the structure demands, then, the design objectives are expected
to be achieved. If the capacities are smaller than the demands, the structure has to be
redesigned.
The seismic design of a coupled wall structure involves establishing structure
demands and providing structure capacities until all of the design criteria are satisfied.
Equations (11.1)-(11.6) represent six of the design criteria used the proposed design
approach. The other required design criteria used in the report are provided in Chapter 12.
285
11.9 Design Acceleration Response Spectra
This section describes the design acceleration response spectra for the survivallevel and design-level seismic demands used in the proposed design approach.
11.9.1 Design Spectra for Survival-Level Demand
As described earlier, the survival-level ground motion is the same as the maximum
considered ground motion in IBC 2000 (ICC 2000), which has a 2% probability of being
exceeded in 50 years (corresponding to a return period of approximately 2500 years). The
design acceleration response spectrum for the survival-level ground motion is determined
as follows.
According to IBC 2000, the maximum considered earthquake spectral response
acceleration at short periods, SMS, and at 1 second period, SM1, are found as:
S MS = Fa S S
(11.7)
S M 1 = Fv S1
(11.8)
where, Fa and Fv are site coefficients given in IBC 2000, and SS and S1 are mapped
spectral response acceleration values at short periods and at 1 second period, respectively,
obtained from Maps 1 through 24 in IBC 2000, from contour maps published by the
Federal Emergency Management Agency (BSSC 1998), or from the United States
Geological Survey (USGS 1996).
The general response spectrum for the survival-level ground motion, Sas is plotted
using SMS and SM1 (see Fig. 11.3). For periods less than or equal to T0, the spectral
response acceleration, Sas, is given as:
S as = 0.6
S MS
T + 0.4S MS
T0
(11.9)
where, T is the period and T0=0.2SM1/SMS.

For periods between T0 and Ts=SM1/SMS, the spectral response acceleration is taken
as equal to SMS.
For periods greater than Ts, the spectral response acceleration is determined as:
S as =
SM 1
T
286
(11.10)
The peak acceleration of the survival-level ground motion, PGAs can be obtained
by substituting T=0 into Equation (11.9) as:
PGAs = 0.4S MS
(11.11)
Design Spectral Response Acceleration, Sas and Sad (g)
Note that IBC 2000 requires a site-specific geotechnical investigation and dynamic
site response analyses for certain sites with soft soil characteristics.
S MS
survival level spectrum
S DS=(2/3)SMS
S M1 /T
0.4S MS
S D1 /T=(2/3)(SM1 /T)
design level
spectrum
(2/3)0.4S MS
T0 =0.2SM1 /S MS
TS =S M1/S MS
Period, T (sec.)
Fig. 11.3 Design acceleration response spectra

11.9.2 Design Spectra for Design-Level Demand
According to IBC 2000, the design-level ground motion spectrum, Sad is

determined by multiplying the spectrum for the maximum considered earthquake by a
factor of 2/3 (referred to as the seismic margin). Thus,
S ad =
2
S as
3
(11.12)
The peak acceleration of the design-level ground motion PGAd can be calculated as:
287
PGAd = (2 / 3)0.4 S MS
(11.13)
11.10 Equivalent Nonlinear SDOF Representation
The proposed seismic design approach requires the estimation of the nonlinear
displacement demands of a coupled wall structure under design-level and survival-level
ground motions. The most accurate analytical method to estimate the seismic demands of
a structure is to conduct nonlinear multi-degree-of-freedom (MDOF) dynamic timehistory analyses. However, since this procedure is time-consuming and is usually not
employed in practice for most regular structures, the proposed design approach uses a
nonlinear static procedure (NSP).
The nonlinear static procedure is based, in part, on the assumption that the lateral
displacement response of a building during an earthquake can be related to the response
of an equivalent single-degree-of-freedom (SDOF) system following simplifying
assumptions for mode shapes and mass participation (SEAOC 1996). This implies that
the displacement response of the structure is controlled by a single mode (often the first
mode), and that the shape of this mode remains essentially constant throughout the
nonlinear response history. Previous studies (Saiidi and Sozen 1981; Fajfar and
Fischinger 1988; Qi and Moehle 1991; Miranda 1991; Lawson et al. 1994) have indicated
that these assumptions lead to reasonable predictions of the peak seismic displacements
of MDOF buildings provided that the response is dominated by the first mode.
An equivalent nonlinear SDOF model that can be used to represent the nonlinear
lateral displacement response of MDOF unbonded post-tensioned hybrid coupled wall
structures is presented below. It is assumed that there are no significant plan or vertical
discontinuities/irregularities (e.g., set-backs) in the structure and that the building is not
excessively tall (e.g., not taller than 240 ft or, approximately, 18 stories as described in
IBC 2000).
11.10.1Transformation from MDOF System to SDOF System
Multistory buildings can be represented as MDOF systems with masses lumped at

the floor and roof levels. For structures responding linear-elastically to a base excitation
u&&g (t ) , the MDOF equation of motion can be expressed as:
[ M ]{U&& (t )} + [C ]{U&& (t )} + [ K ]{U (t )} = [ M ]{R}u&&g (t )
where: [M]
[C]
[K]
u&&g (t )
{R}
(11.14)
is the lumped diagonal mass matrix of the system;

is the damping matrix of the system;
is the stiffness matrix of the system;
is the horizontal ground acceleration time-history;
is a vector of rigid body displacements in the horizontal direction
({R}={1} considering horizontal degrees of freedom only); and
288
{U(t)}
is the relative lateral displacement vector of the MDOF system at
the floor and roof levels, with {U& (t )} and {U&& (t )} representing the first and second time
derivatives (i.e., velocity and acceleration), respectively.
First mode displacements
The vector {U(t)} can be decomposed using mass-orthogonal mode shapes as:
n
{U } = {1}Y1 + {2 }Y2 + + {n }Yn = {i }Yi = [ ]{Y }
(11.15)
i =1
where, {i} is the ith linear-elastic vibration mode-shape of the MDOF system, n is the
number of modes, [] is a matrix containing the n mode shapes, Yi is the ith modal
amplitude, and {Y} is the modal amplitude vector. The orthogonality property of the
mode shapes leads to
{i }T [ M ]{U } = {i }T [ M ]{i }Yi
(11.16)
Let u(t) be equal to the roof displacement of the MDOF system. The first mode
component of the roof-displacement time-history, u1 can be calculated as:
u1 = Y1 =
{1}T [ M ]{U }
{1}T [ M ]{1}
{1}T [ M ]{U }
M eq
(11.17)
where, Y1 is the amplitude of the first mode, {1} is the first linear-elastic mode-shape of
the structure normalized with respect to the roof, and Meq is the generalized mass for
the first mode.
The first mode representation of the nonlinear lateral displacement response of
unbonded post-tensioned hybrid coupled wall structures is demonstrated using nonlinear
dynamic time-history analyses of the CIP-UPT system described in Chapters 6, 7, and 8.
The fiber element model in Chapter 6 is used to conduct the analyses, with the wallcontact elements removed to reduce the computation time. The structure is assumed to
have a viscous damping ratio of =3% in the first and third modes using Rayleigh
damping.
The solid line in Fig. 11.4 shows the roof-displacement time-history of the CIPUPT system under the LA22 ground motion from the SAC Steel Project (Somerville et al.
1997). The dashed line represents the first mode component of the roof-displacement
time-history, u1 as calculated using Equation (11.17). Comparison between the solid and
dashed lines in Fig. 11.4 shows that u1 is very close to the total roof-displacement, u
indicating that the lateral displacement response of the structure can be estimated based
on the first mode as:
289
{U (t )} = {1}Y1 (t ) = {1}u1 (t )
(11.18)
500
total roof displacement, u(t)
first mode component, u1(t)
400
CIP-UPT system
roof displacement, u (mm)
300
200
100
0
-100
-200
-300
SAC LA22
-400
-500
0
10
time, t (sec.)
15
20
Fig. 11.4 First mode component of roof-displacement time-history

Equivalent SDOF system
Pre-multiplying both sides of Equation (11.14) by the transpose of the first mode
shape vector {1}T, and substituting Equation (11.18)
M equ&&1(t ) + Cequ&1 (t ) + K equ1 (t ) = Lequ&&g (t )
2
u&&1 (t ) + 2 eqequ&1 (t ) + eq
u1 (t ) =
Leq
M eq
u&&g (t )
(11.19)
(11.20)
where,
M eq = {1}T [ M ]{1}
(11.21)
Ceq = {1}T [C ]{1}
(11.22)
K eq = {1}T [ K ]{1}
(11.23)
290
2
eq
= K eq / M eq
2 eqeq = Ceq / M eq
(11.24)
Leq = {1}T [ M ]{1}
(11.26)
(11.25)
The scalars Meq, Ceq, and Keq represent the generalized mass, damping, and
stiffness for the first mode, respectively. The scalar Leq is often referred to as the
earthquake excitation factor and represents the extent to which the ground motion tends
to excite response in the first mode of vibration.
Equation (11.20) is the equation of motion of a unit mass SDOF system subjected
to a factored earthquake, where the factor, given by =Leq/Meq, is generally known as the
participation factor. Multiplying Equation (11.20) by Meq/Leq, a new displacement
coordinate is defined as:
ueq (t ) =
M eq
Leq
u1 (t ) =
u1 (t )
(11.27)
and Equation (11.20) can be rewritten as:

2
u&&eq (t ) + 2eqequ&eq (t ) + eq
ueq (t ) = u&&g (t )
(11.28)
Equation (11.28) describes the displacement response of an equivalent unit mass

SDOF system under the original earthquake acceleration record. Here, ueq is the
2
displacement of the equivalent SDOF system and the product eq
ueq is the force per unit
of mass acting on the equivalent system.
The period of the equivalent SDOF system, Teq, is equal to
Teq =
eq
M eq
2
= 2
K eq
K eq
(11.29)
M eq
Using the undamped first mode free vibration equation of motion as:
[ K ]{1} = 12 [ M ]{1}
(11.30)
and multiplying both sides of the equation by {1}T, it can be shown that eq and Teq of
the equivalent SDOF system are equal to the first mode frequency 1 and period T1 of the
MDOF system, respectively.
291
Relationships between MDOF and SDOF displacements and forces
Based on the above equivalent SDOF formulation, the roof displacement u(t) of the
MDOF system can be estimated as:
u (t ) = u1 (t ) = ueq (t )
(11.31)
and the displacements, {U(t)} of the structure over the height are given as:
{U (t )} = {1}u1 (t ) = {1} ueq (t )
(11.32)
The base shear, F(t) of the MDOF system can be obtained as:
F (t ) = {1}T { f (t )}
(11.33)
where, {f(t)} is a vector containing the resisting forces at the floor and roof levels. This
vector of forces can be expressed as the product of the stiffness matrix, [K], and the
displacement vector, {U(t)}, resulting in
F (t ) = {1}T [ K ]{1} ueq (t )
(11.34)
Substituting Equations (11.26), (11.30), and =Leq/Meq

F (t ) =
where the term
L2eq
M eq
L2eq
M eq
2
eq
ueq
(11.35)
is often referred to as the effective first mode mass, Meff.
Note that Equation (11.28) can also be written for an equivalent SDOF system with
mass equal to the effective first mode mass, Meff as:
2
M eff u&&eq (t ) + M eff 2 eqequ&eq (t ) + M eff eq
ueq (t ) = M eff u&&g (t )
(11.36)
The corresponding SDOF force can be written as:

Feq (t ) =
2
M eff eq
ueq
L2eq
M eq
292
2
eq
ueq = F (t )
(11.37)
and thus, the resisting force of the SDOF system with mass equal to Meff is the same as
the base shear force of the MDOF system.
11.10.2 Equivalent Nonlinear SDOF Model
The objective of developing an equivalent SDOF system is to provide a tool to

estimate the nonlinear lateral displacement response of a MDOF system without the need
for MDOF dynamic time-history analyses. This section presents a hysteretic SDOF
model for unbonded post-tensioned hybrid coupled wall structures. For design purposes,
the smooth nonlinear lateral load versus displacement relationship of the structure is
idealized using a multi-linear relationship.
The lateral load versus displacement behavior of unbonded post-tensioned hybrid
coupled wall systems possesses hysteretic characteristics similar to the bilinearelastic/elasto-plastic (BP) model developed by Farrow and Kurama (2001, 2003, 2004).
As shown in Fig. 11.5, the BP hysteresis model is constructed by placing a bilinearelastic (BE) hysteresis model in parallel with an elastic-perfectly-plastic (EP) hysteresis
model. The BE model represents the self-centering capability of the structure provided by
the post-tensioning force and the EP model represents the energy dissipation capability
provided by the top and seat angles and yielding of the mild steel reinforcement at the
bases of the cast-in-place concrete walls.
Feq
(1+s )k be
(rFbe,
(Fbe , ube)
ueq
0
k be
Feq
Feq
r
u )
s be
[(1+r)Fbe,
ueq
0
s k be
(1+ s)k be
1+r
u ]
1+s be
ueq
0
(1+s )kbe
Bilinear-Elastic (BE)
Elasto-Plastic (EP)
Bilinear-Elastic/
Elasto-Plastic (BP)
(s = r < 1)
Fig. 11.5 BP hysteresis model with s=r<1 (adapted from Farrow and Kurama 2003)
As shown in Fig. 11.5 the relationships between the lateral stiffnesses and strengths
of the BE and EP hysteresis models that make up a BP model are:
k ep = s k be
(11.38)
Fep = r Fbe
(11.39)
293
where, kbe and kep are the linear-elastic stiffnesses, and Fbe and Fep are the yield strengths
of the BE and EP models, respectively, and s and r are the BP model stiffness and
strength ratios, respectively.
Note that BP models with only one yield point are used in this report as illustrated
in Fig. 11.4. In this case, s must be equal to r. More information on the BP hysteresis
model can be found in Farrow and Kurama (2001, 2003, 2004).
11.10.3 Application of Equivalent SDOF BP Model
This section describes the use of the equivalent SDOF BP hysteresis model to
represent the nonlinear lateral displacement response of multi-story unbonded posttensioned hybrid coupled wall structures. The procedure is demonstrated using the CIPUPT and PRE-UPT coupled wall structures from Chapters 6, 7, and 8. Note that the
structure properties given below are for one coupled wall system only, and not for the
entire building, which consists of four identical coupled wall systems in the E-W
direction.
First, idealized base shear force versus roof displacement (F-u) relationships of the
CIP-UPT and PRE-UPT structures are developed using the procedures described in
Chapter 8. The idealized relationships give the base shear force and roof displacement
values corresponding to the coupled wall softening, PT-yielding, and ultimate states. Figs.
11.6a and 11.6b show comparisons between the idealized relationships (dashed lines) and
the smooth relationships from static push-over analyses of the two structures using the
fiber element model described in Chapter 6 (solid lines). A reasonable representation of
the behavior of the coupled wall systems is observed.
4500
4500
( Fwu , wu )
( Fwu , wu )
( Fws , ws )
( Fwy , wy)
( Fws , ws )
CIP-UPT system
400
PRE-UPT system
800
400
(a)
(b)
Fig. 11.6 Idealized base shear force versus roof displacement relationships:
294
800
Note that a trilinear idealized relationship is described in Chapter 8 and shown in

Fig. 11.6b for the PRE-UPT structure; however, the SDOF BP model uses a bilinear
relationship. Thus, the procedure demonstrated below utilizes only the coupled wall
softening and ultimate states for the structures, and the PT-yielding state for the PREUPT system is ignored (dotted line in Fig. 11.6b ). The results from the idealized MDOF
base shear force versus roof displacement relationships, as well as the results from a
linear-elastic modal analysis of the MDOF structures are given in Table 11.1 below.
TABLE 11.1
MDOF SYSTEM PROPERTIES
CIP-UPT
Wall Softening State

uws
Fws
kN (kips)
mm (in.)
2282(513)
50 (1.98)
Wall Ultimate State

Fwu
uwu
kN (kips)
mm (in.)
3856 (867)
630 (24.8)
PRE-UPT
2700(607)
4052 (911)
Structure
62 (2.43)
752 (29.6)
T1 (sec.)
meff (kg)
0.83
3%
1.42
1137000
0.85
3%
1.42
1142000
where, meff is the effective first mode mass assigned to the coupled wall system (assumed
to be equal to the total building effective first mode mass, Meff divided by the number of
coupled wall systems, nw).
Then, the idealized MDOF base shear force versus roof displacement (F-u)
relationships are used to determine BP models to represent the hysteretic characteristics
of the structures. There are four independent unknowns for each BP model, Fbe, ube, ,
and r, which can be combined with Equations (11.38) and (11.39) to determine the BE
and EP components assuming s=r. Referring to Fig. 11.5, the four independent
unknowns for a BP model are calculated using the following four relationships.
(1) The yield strength of the BP model is the same as the base shear capacity of the
MDOF system at the coupled wall softening state. Thus,
Fbe (1 + r ) = Fws
(11.40)
(2) The yield displacement of the BP model is the same as the roof displacement
capacity of the MDOF system at the coupled wall softening state. Thus,
ube = uep = uws
(11.41)
(3) The post-yield stiffness of the BP model is the same as the lateral stiffness of
the MDOF system between the coupled wall softening and ultimate states. Thus,
(1 + r )
Fbe Fwu Fws

=
ube uwu uws
(11.42)
Note that represents the post-yield stiffness ratio of the combined BP model.
295
(4) The hysteretic energy dissipation of the BP model during a selected

displacement cycle is the same as the energy dissipation from the smooth base shear force
versus roof displacement relationship of the MDOF system during the same displacement
cycle.
Let the hysteretic energy dissipation from the smooth MDOF base shear force
versus roof displacement relationship during a roof displacement cycle to uc, measured
by the area inside the hysteresis loop, be equal to Ehc. Then, it can be shown that,
4 r Fbe (uc ube ) = Ehc
(11.43)
Substituting Equations (11.40) and (11.41) into Equation (11.43), it can be shown
that,
r =
Ehc
4 Fws (uc uws ) Ehc
(11.44)
Note that an exact application of Equation (11.44) requires a smooth cyclic base
shear force versus roof displacement relationship for the MDOF structure (to determine
Ehc), which may not be available for design. In practical design applications, the use of an
assumed value of r (based on the expected hysteretic characteristics of the structure) is
typically adequate because of the relatively small effect of r on the seismic demands
(Farrow and Kurama 2003, 2004). This is demonstrated in Chapter 12.
The properties of the idealized MDOF BP models for the CIP-UPT and PRE-UPT
structures are determined using Equations (11.40), (11.41), (11.42) and (11.44) and
presented in Table 11.2. As an example, Fig. 11.7a shows the cyclic base shear force
versus roof displacement (F-u) relationship of the CIP-UPT structure using the fiber
element model described in Chapter 6. The structure is subjected to roof displacements of
u=163 mm, 245 mm, 326 mm, 408 mm, 489 mm, 571 mm, 652 mm, 734 mm,
and 815 mm (corresponding to roof drift values of =0.5%, 0.75%, 1.0%, 1.25%,
1.5%, 1.75%, 2.0%, 2.25%, and 2.5%, respectively). The displacement cycle to
uc=489 mm (thick lines in Fig. 11.7a) is used in Equation (11.44) above.
TABLE 11.2
IDEALIZED MDOF BP MODEL PROPERTIES
BE Component
ube
Fbe
(1+r)
kN (kips) mm (in.)
CIP-UPT 1557 (350) 50 (1.98)
0.088
PRE-UPT 2139 (481) 62 (2.43)
0.056
Structure
EP Component
Combined BP Model
Fep
uep
Fbp
ubp
r
kN (kips) mm (in.) kN (kips) mm (in.)

729 (164) 50 (1.98) 2282 (513) 50 (1.98) 0.47 0.060
556 (125) 62 (2.43) 2700 (607) 62 (2.43) 0.26 0.045
Fig. 11.7b shows the cyclic force versus displacement (F-u) behavior of the
idealized MDOF BP model. The thick lines indicate the selected roof displacement cycle
to uc=489 mm. The dashed and solid lines in Fig. 11.7c show the hysteretic energy, Eh
296
dissipated by the smooth and idealized relationships (calculated as the area enclosed by
each hysteresis loop) in Figs. 11.7a and 11.7b, respectively. The energy dissipations
during uc=489 mm are the same because of Equation (11.44). The energy dissipation of
the idealized MDOF BP model is smaller than the energy dissipation of the smooth
model during smaller displacement cycles and larger than the energy dissipation of the
smooth model during larger displacement cycles.
6000
MDOF BP model force, F (kN)
6000
u c = +489 mm
CIP-UPT system
-6000
-1000
CIP-UPT system
1000
roof displacement (mm)
MDOF BP model displacement, u (mm)
(a)
(b)
6000
SDOF BP model force, Feq (kN)
smooth MDOF model
hysteretic energy
dissipation, Eh (kN.m)
u c = +489 mm
-6000
-1000
1000
3000
MDOF BP model
Ehc=1261 kN.m
CIP-UPT system
0
0
u c = +489 mm
-6000
-1000
1000
CIP-UPT system
0
1000
SDOF BP model displacement, ueq (mm)

(c)
(d)
Fig. 11.7 Equivalent SDOF representation of CIP-UPT system: (a) smooth MDOF
behavior; (b) MDOF BP model; (c) hysteretic energy dissipation;
(d) equivalent SDOF BP model
Finally, the properties of the equivalent SDOF BP model are determined by
dividing the displacements of the MDOF BP model with the roof displacement
participation factor, as shown in Fig. 11.7d.
Similarly, Fig. 11.8 shows the development of an equivalent SDOF BP model for
the PRE-UPT structure. Table 11.3 lists the equivalent SDOF BP model properties for the
CIP-UPT and PRE-UPT structures.
297
6000
MDOF BP model force, F (kN)
6000
u c = +489 mm
PRE-UPT system
-6000
-1000
u c = +489 mm
PRE-UPT system
-6000
-1000
1000
3000
SDOF BP model force, Feq (kN)
BP MDOF model
hysteretic energy
dissipation, Eh (kN.m)
1000
6000
smooth MDOF model
Ehc=951 kN.m
PRE-UPT system
0
MDOF BP model displacement, u (mm)

(b)

(a)
u c = +489 mm
-6000
-1000
1000
PRE-UPT system
0
1000
SDOF BP model displacement, ueq (mm)

(c)
(d)
Fig. 11.8 Equivalent SDOF representation of PRE-UPT system: (a) smooth MDOF
behavior; (b) MDOF BP model; (c) hysteretic energy dissipation;
(d) equivalent SDOF BP model
TABLE 11.3
EQUIVALENT SDOF BP MODEL PROPERTIES
BE Component
Fbe
ube
(1+r)
kN (kips) mm (in.)
CIP-UPT 1557 (350) 35 (1.39)
0.088
EP Component
Combined BP Model
uep
Fbp
ubp
Fep
r
kN (kips) mm (in.) kN (kips) mm (in.)

729 (164) 35 (1.39) 2282 (513) 35 (1.39) 0.47 0.060
PRE-UPT 2139 (481) 43 (1.71)
556 (125)
Structure
0.056
43 (1.71) 2700 (607) 43 (1.71)
0.26
0.045
11.10.4 Evaluation of Equivalent SDOF BP Model
In order to evaluate the capability of the SDOF BP hysteresis model to capture the
nonlinear displacement response characteristics of unbonded post-tensioned hybrid
coupled wall structures, nonlinear dynamic time-history analyses of the smooth MDOF
and equivalent SDOF BP models for the CIP-UPT and PRE-UPT structures were
conducted under 20 ground motion records (LA21-LA40) from the SAC Steel Project
298
(Somerville et al. 1997). More information on these ground motion records can be found
in Chapter 14.
The equivalent SDOF analyses were conducted using the BP models given in Table
11.3 and shown in Figs. 11.7d and 11.8d for the CIP-UPT and PRE-UPT systems,
respectively. The other dynamic properties of the equivalent SDOF systems are as
follows:
Equivalent BP model period = T1 (see Table 11.1)
Equivalent BP model mass = meff (see Table 11.1)
Equivalent BP model damping ratio = 1 = 3% of critical damping (see Table
11.1).
Similarly, the MDOF analyses were conducted with a viscous damping ratio of 3%
in the first and third modes (i.e., 1=3=3%) using Rayleigh damping. The damping
coefficients were calculated based on the linear-elastic properties of each structure and
were kept constant during the nonlinear dynamic time-history analyses.
Figs. 11.9a and 11.9b show the ratios between the peak displacements from the
SDOF and MDOF analyses of the CIP-UPT and PRE-UPT structures, respectively. Note
that the displacements from the SDOF analyses are multiplied by the roof displacement
participation factor, . The results are plotted against the maximum incremental
velocities (MIV) of the SAC ground motion records (LA21-LA40), where MIV for a
ground motion is calculated as the maximum area under the acceleration time history of
the record between two successive zero acceleration crossings. On average (dashed lines),
the ratio between the peak displacements of the SDOF and MDOF systems is close to 1.0,
with little scatter in the results from the 20 ground motions.
Similarly, Figs. 11.10 and 11.11 show comparisons between the SDOF and MDOF
roof-displacement time-histories for the CIP-UPT and PRE-UPT systems, respectively,
under the 20 SAC ground motions (LA21-LA40). It is concluded that the nonlinear
equivalent SDOF BP model is capable of predicting not only the peak roof displacements
but also the roof-displacement time-histories of the structures reasonably well.
11.11 Seismic Displacement Demand Relationships
This section describes design relationships between the lateral strength and the
mean peak lateral displacements of BP type SDOF models. These relationships are used
to establish design demands for MDOF unbonded post-tensioned hybrid coupled wall
structures, in lieu of conducting nonlinear dynamic time-history analyses.
299
SDOF/MDOF peak displacements
1.5
average=1.03
1
CIP-UPT system
0.5
50
100
150
200
250
300
350
maximum incremental velocity, MIV (cm/sec.)
400
(a)
SDOF/MDOF peak displacements
1.5
average=0.97
1
0.5
PRE-UPT system
50
100
150
200
250
300
350
400
(b)
Fig. 11.9 Peak SDOF versus MDOF displacements:
300
2000
SAC LA21
roof disp., u
(mm)
roof disp., u
(mm)
2000
SDOF BP model
smooth MDOF model
CIP-UPT system
-2000
10
15
SAC LA22
0
CIP-UPT system
-2000
20
SDOF BP model
smooth MDOF model
10
time , t (sec.)
2000
SDOF BP model
smooth MDOF model
SAC LA23
roof disp., u
(mm)
roof disp., u
(mm)
2000
CIP-UPT system
-2000
10
12
14
CIP-UPT system
-2000
16
10
roof disp., u
(mm)
roof disp., u
(mm)
CIP-UPT system
0
10
CIP-UPT system
-2000
15
SDOF BP model
smooth MDOF model
SAC LA26
15
2000
SAC LA27
roof disp., u
(mm)
2000
roof disp., u
(mm)
10
time , t (sec.)
time , t (sec.)
SDOF BP model
smooth MDOF model
CIP-UPT system
0
10
15
SAC LA28
0
CIP-UPT system
-2000
20
SDOF BP model
smooth MDOF model
10
15
20
time , t (sec.)
time , t (sec.)
2000
roof disp., u
(mm)
2000
roof disp., u
(mm)
20
2000
SDOF BP model
smooth MDOF model
SAC LA25
SDOF BP model
smooth MDOF model
SAC LA29
0
-2000
15
time , t (sec.)
2000
-2000
20
SDOF BP model
smooth MDOF model
SAC LA24
time , t (sec.)
-2000
15
time , t (sec.)
CIP-UPT system
0
10
15
20
25
-2000
30
time , t (sec.)
SDOF BP model
smooth MDOF model
SAC LA30
CIP-UPT system
0
10
15
20
25
time , t (sec.)
Fig. 11.10a SDOF and MDOF roof-displacement time-histories for CIP-UPT system
(SAC LA21-LA30)
301
30
2000
SAC LA31
roof disp., u
(mm)
roof disp., u
(mm)
2000
0
SDOF BP model
smooth MDOF model
-2000
CIP-UPT system
10
15
SAC LA32
0
SDOF BP model
smooth MDOF model
-2000
20
10
time , t (sec.)
roof disp., u
(mm)
roof disp., u
(mm)
0
SDOF BP model
smooth MDOF model
0
CIP-UPT system
10
15
SAC LA34
0
SDOF BP model
smooth MDOF model
-2000
20
roof disp., u
(mm)
roof disp., u
(mm)
0
SDOF BP model
smooth MDOF model
0
CIP-UPT system
10
15
SAC LA36
0
SDOF BP model
smooth MDOF model
-2000
20
CIP-UPT system
10
15
20
time , t (sec.)
2000
2000
SAC LA37
SDOF BP model
smooth MDOF model
roof disp., u
(mm)
roof disp., u
(mm)
20
2000
SAC LA35
time , t (sec.)
CIP-UPT system
0
10
15
20
25
SAC LA38
0
CIP-UPT system
-2000
30
SDOF BP model
smooth MDOF model
10
15
20
25
30
time , t (sec.)
time , t (sec.)
2000
SAC LA39
roof disp., u
(mm)
2000
roof disp., u
(mm)
15
time , t (sec.)
2000
SDOF BP model
smooth MDOF model
-2000
CIP-UPT system
10
time , t (sec.)
-2000
20
2000
SAC LA33
-2000
15
time , t (sec.)
2000
-2000
CIP-UPT system
CIP-UPT system
0
10
15
20
25
SAC LA40
0
-2000
30
time , t (sec.)
SDOF BP model
smooth MDOF model
CIP-UPT system
0
10
15
20
25
time , t (sec.)
Fig. 11.10b SDOF and MDOF roof-displacement time-histories for CIP-UPT system
(SAC LA31-LA40)
302
30
2000
SAC LA21
roof disp., u
(mm)
roof disp., u
(mm)
2000
SDOF BP model
smooth MDOF model
PRE-UPT system
-2000
10
15
PRE-UPT system
-2000
0
20
SDOF BP model
smooth MDOF model
SAC LA22
10
time , t (sec.)
SDOF BP model
smooth MDOF model
PRE-UPT system
0
10
12
14
SAC LA24
0
PRE-UPT system
-2000
16
SDOF BP model
smooth MDOF model
10
time , t (sec.)
roof disp., u
(mm)
roof disp., u
(mm)
SDOF BP model
smooth MDOF model
PRE-UPT system
0
10
SAC LA26
0
PRE-UPT system
-2000
15
SDOF BP model
smooth MDOF model
10
2000
roof disp., u
(mm)
roof disp., u
(mm)
2000
SAC LA27
SDOF BP model
smooth MDOF model
PRE-UPT system
0
10
15
SDOF BP model
smooth MDOF model
-2000
20
SAC LA28
PRE-UPT system
0
10
15
20
time , t (sec.)
time , t (sec.)
2000
SAC LA29
roof disp., u
(mm)
2000
roof disp., u
(mm)
15
time , t (sec.)
time , t (sec.)
SDOF BP model
smooth MDOF model
-2000
20
2000
SAC LA25
-2000
15
time , t (sec.)
2000
-2000
20
2000
SAC LA23
roof disp., u
(mm)
roof disp., u
(mm)
2000
-2000
15
time , t (sec.)
PRE-UPT system
0
10
15
20
25
SAC LA30
0
-2000
30
SDOF BP model
smooth MDOF model
PRE-UPT system
0
10
15
20
25
30
time , t (sec.)
time , t (sec.)
Fig. 11.11a SDOF and MDOF roof-displacement time-histories for PRE-UPT system
(SAC LA21-LA30)
11.11.1 Previous Research
Peak nonlinear displacement demands for different types of hysteretic SDOF

systems have been studied by many researchers (e.g., Newmark and Hall 1973; Newmark
and Riddell 1980; Al-Sulaimani and Roessett 1985; Lin and Mahin 1985; Elghadamsi
and Mohraz 1987; Nassar and Krawinkler 1991; Miranda 1993; Priestley and Tao 1993;
Rahnama and Krawinkler 1993; Hall et al. 1995; Naeim 1995; Shome et al. 1998; Kwan
and Billington 1999; Borzi and Elnashai 2000; Christopoulos et al. 2002; Farrow and
Kurama 2001, 2003, 2004). These studies show that the SDOF peak displacement
demand depends on structure (e.g., lateral strength, period, hysteretic behavior), site (e.g.,
soil type), and ground motion characteristics.
303
2000
roof disp., u
(mm)
roof disp., u
(mm)
2000
SAC LA31
0
SDOF BP model
smooth MDOF model
-2000
PRE-UPT system
10
15
SAC LA32
0
SDOF BP model
smooth MDOF model
-2000
20
10
time , t (sec.)
PRE-UPT system
-2000
10
15
SAC LA34
0
SDOF BP model
smooth MDOF model
-2000
20
roof disp., u
(mm)
roof disp., u
(mm)
0
SDOF BP model
smooth MDOF model
0
PRE-UPT system
10
15
SAC LA36
0
SDOF BP model
smooth MDOF model
-2000
20
10
SDOF BP model
smooth MDOF model
SAC LA37
0
PRE-UPT system
0
10
15
20
15
20
2000
roof disp., u
(mm)
roof disp., u
(mm)
2000
PRE-UPT system
time , t (sec.)
time , t (sec.)
25
SAC LA38
0
PRE-UPT system
-2000
30
SDOF BP model
smooth MDOF model
10
15
20
25
30
time , t (sec.)
time , t (sec.)
2000
roof disp., u
(mm)
2000
roof disp., u
(mm)
20
2000
SAC LA35
SDOF BP model
smooth MDOF model
0
-2000
15
time , t (sec.)
2000
-2000
PRE-UPT system
10
time , t (sec.)
-2000
20
2000
SDOF BP model
smooth MDOF model
SAC LA33
15
time , t (sec.)
roof disp., u
(mm)
roof disp., u
(mm)
2000
PRE-UPT system
PRE-UPT system
SAC LA39
0
10
15
20
25
SDOF BP model
smooth MDOF model
0
-2000
30
time , t (sec.)
PRE-UPT system
SAC LA40
0
10
15
20
25
30
35
time , t (sec.)
Fig. 11.11b SDOF and MDOF roof-displacement time-histories for PRE-UPT system
(SAC LA31-LA40)
For medium and long period systems (i.e., T0.5 sec.) with no strength degradation,
it has been shown that the peak displacement demand does not depend strongly on the
lateral strength. In other words, for a given ground motion, a nonlinear system has
approximately the same peak displacement demand as a corresponding linear-elastic
system with the same period. For shorter periods (i.e., T<0.5 sec.), the peak
displacements tend to increase as the lateral strength of a structure decreases.
Previous studies on the development of relationships between the lateral strength,
period, and peak displacement demands of SDOF systems have primarily considered
models with significant hysteretic energy dissipation. Theses system include bilinear
elasto-plastic models and stiffness degrading models. An idealized bilinear lateral forcedisplacement relationship (Fig. 11.12) is often used with the yield strength, Fy, quantified
304
with respect to the strength required to keep the structure linear-elastic during an
earthquake (i.e., the linear-elastic force demand), Felas, using the strength ratio, R. The
ductility demand, , provides an estimate of the peak nonlinear displacement demand of
the structure, unlin and is usually presented in relationship to R.
Felas
change in
displacement
force reduction
2
K
Fnlin
F
Fy= elas
R
F
Fdes
= elas
Rd
K
3
significant yield
udes u y
linear-elastic behavior
nonlinear behavior
idealized bilinear beh.

unlin =u y
u elas
Fig. 11.12 Definitions of R and (from Farrow and Kurama 2003)

The hysteretic energy dissipation of unbonded post-tensioned hybrid coupled wall
structures is typically smaller than bilinear elasto-plastic systems and stiffness degrading
systems. As part of a comprehensive statistical investigation of SDOF demand index
relationships for performance-based seismic design, Farrow and Kurama (2001, 2003,
2004) developed design relationships for the peak displacement demands of SDOF BP
type models. These relationships relate the strength reduction factor R to the peak
displacement ductility demand and the SDOF period of vibration T using a two-step
nonlinear regression analysis scheme and the form of the regression equations from
Nassar and Krawinkler (1991) as:
R ( , T , ) = [c( 1) + 1]1 / c
(11.45)
Felas
u
, = nlin
Fy
uy
(11.46)
with,
R =
c(T , ) =
Ta
b
T +1 T
a
305
(11.47)
where, is the post-yield stiffness ratio, is the displacement ductility demand, and a
and b are regression coefficients determined from a statistical evaluation of SDOF
nonlinear dynamic time-history analyses using the BP system.
The form of the regression relationship given by Equations (11.45-11.47) is based
on a previous relationship proposed by Osteraas and Krawinkler (1990). The first step
regression is carried out in the R- domain to relate the period, T and other parameter
dependencies (i.e., hysteresis type and site conditions) to the c coefficient. The second
step regression is carried out in the c-T domain to relate the parameter dependencies to
the a and b coefficients. More information on the regression procedure to determine the
design R--T relationships can be found in Farrow and Kurama (2001, 2003, 2004).
Note that the definition of R in the proposed design approach is based on a point
of global yield for the structure (i.e., Fy, uy), which is different from the definition of
the strength ratio in current U.S. seismic design provisions (referred to as the response
modification factor, Rd) based on a point of significant yield in the lateral load system
(i.e., Fdes, udes in Fig. 11.12). The definition based on the global yield point takes into
account the over-strength in the structure, and thus, is expected to result in better
measures of the seismic demands through uy, instead of udes.
Note also that the strength ratio R in this research is used to quantify the structure
yield strength. Current U.S. seismic design provisions (ICBO 1997, BSSC 1998, ICC
2000) are force-based and, in this context, provide prescribed Rd values to specify the
required design strength of a structure. These force-based approaches may be replaced
with displacement-based and performance-based approaches in future design provisions.
In accordance with this transition, the use of R in this report is from a displacementbased perspective, where the problem of interest is the estimation of the required
structure strength for a given target peak displacement ductility demand, t. In this
context, R is used to quantify the yield strength Fy of the structure with respect to the
linear-elastic force demand Felas, and is considered as a ground-motion-dependent
structure property, not as a code-specified constant.
11.11.2 Displacement Demand Relationships for SDOF BP Model
This section describes the development of peak seismic displacement demand (i.e.,
R--T) relationships for SDOF BP systems. A MATLAB (2000) algorithm, CDSPEC
(Capacity-Demand SPECtra), was developed by Farrow and Kurama (2001a,b, 2003,
2004) to determine R--T relationships for various types of SDOF models, including BP
type models. Ground motion ensembles from the SAC Steel Project (Somerville et al.
1997) were used and augmented by records generated at the University of Notre Dame to
conduct statistical analyses of SDOF demands for selected seismic design conditions,
including site soil type, site seismicity, and seismic demand level.
The CDSPEC program was used in this report to develop R--T relationships [in
the form of Equations (11.45), (11.46), (11.47)] for BP models considering a wider range
306
of seismic design conditions than originally studied by Farrow and Kurama (2001, 2003,
2004). Farrow and Kurama (2004) used three different reference linear-elastic
acceleration response spectra to calculate the linear-elastic force demand, Felas for use in
Equation (11.46) as follows: (1) individual (IND) response spectra of the ground motions
used in each ensemble; (2) average (AVG) spectra of each ground motion ensemble; and
(3) smooth design (DES) response spectra from model building seismic design provisions.
In the case of the IND spectra, Felas is calculated as the structure mass times the
spectral acceleration (at the structure period) from each ground motion in an ensemble.
Thus, the structure lateral strength is recalculated for each ground motion in the ensemble.
As described in Farrow and Kurama (2004), most of the previous research efforts on the
development of peak SDOF displacement demand relationships are based on this method.
In the case of the AVG spectra, Felas is calculated as the structure mass times the
average spectral acceleration (at the structure period) from the ground motion ensemble.
Thus, the same structure lateral strength is used for all of the ground motions in an
ensemble. The AVG spectra provide an overall representation, in an average sense, of the
ground motion characteristics in each ensemble.
In the case of the DES spectra, Felas is calculated as the structure mass times the
spectral acceleration (at the structure period) from a smooth design response spectrum.
Thus, the structure lateral strength does not vary with the individual ground motions used
in an ensemble, similar to the use of the AVG spectra described above.
The previous research on seismic displacement demand relationships has mostly
used IND spectra to determine the structure lateral strength. In comparison, Farrow and
Kurama (2001, 2004) argue that R--T relationships developed based on smooth DES
spectra from current design codes (e.g., IBC 2000) are more suitable for use in seismic
design. This is because, in practical seismic design applications, the lateral strength of a
structure does not vary with the individual ground motions considered in an ensemble.
R--T relationships from DES spectra utilize the design spectra used in the
determination of the design base shear demand for the structure. Thus, the resulting
seismic displacement demand estimates are compatible with the lateral strength of the
structure.
Tables 11.4 and 11.5 show the a and b regression coefficients, corresponding to
IND spectra and DES spectra respectively, for the R--T relationships developed as part
of the proposed seismic design approach in this report. Displacement demand
relationships for the following seismic design conditions are developed:
1) Two types of BP models with r=1/3 and r=1/2.
2) One post-yield stiffness ratio, =0.10.
307
3) Three site seismicities low seismicity (Boston), medium seismicity (Seattle),

and high seismicity (Los Angeles).
4) Two seismic demand levels design-level and survival-level. The design and
survival seismic demand levels correspond to the design-level ground motion (with a
10% probability of exceedance in 50 years) and the maximum considered earthquake
ground motion (with a 2% probability of exceedance in 50 years) in IBC 2000 (ICC
2000), respectively.
5) Two types of site soil characteristics stiff soil and soft soil. The stiff and
soft soil characteristics correspond to Site Classes D and E in IBC 2000 (ICC 2000),
respectively.
TABLE 11.4
REGRESSION COEFFICIENTS a AND b FOR BP MODELS
(BASED ON IND SPECTRA)
Hysteresis
Model
Site
Seismicity
Demand
Level
Design
Los Angeles
Survival
BP
(r=1/3,
=0.10)
Design
Seattle
Survival
Design
Boston
Survival
Design
Los Angeles
Survival
BP
(r=1/2,
=0.10)
Site Soil
Class
Design
Seattle
Survival
Design
Boston
Survival
Regression Coefficient
a
b
3.82
0.87
0.65
1.02
1.1*
0.89*
2.39
0.64
0.61
0.68
1.33
0.63
0.92
0.61
0.43
0.59
0.93
0.62
3.52
0.84
0.55
0.99
0.96
0.87*
2.15
0.61
0.50
0.65
1.21
0.60
0.83
0.59
0.34
0.56
0.83
0.59
Note: These a and b regression coefficients are also reported by Farrow and Kurama (2001, 2003).
308
TABLE 11.5
REGRESSION COEFFICIENTS a AND b FOR BP MODELS
(BASED ON DES SPECTRA)
Hysteresis
Model
Site
Seismicity
Demand
Level
Design
Los Angeles
Survival
BP
(r=1/3,
=0.10)
Design
Seattle
Survival
Design
Boston
Survival
Design
Los Angeles
Survival
BP
(r=1/2,
=0.10)
Site Soil
Class
Design
Seattle
Survival
Design
Boston
Survival
a
b
1.17
0.59
0.93
0.96
2.84
0.88*
-1.03
0.29
-3.12
0.32
0.14
0.68
-4.92
0.027
-5.80
0.039
-0.25
0.67
1.07
0.55
0.84
0.94
2.66
0.85
-1.19
0.24
-3.44
0.27
0.038
0.65
-5.23
-0.012
-6.09
-0.0029
-0.35
0.64
Note: These a and b regression coefficients are also reported by Farrow and Kurama (2001, 2004).
The design-level and survival-level DES spectra used in the development of the
regression coefficients in Table 11.5, and the spectral acceleration coefficients (SMS, SM1,
SDS, and SD1) that define the DES spectra can be found in Farrow and Kurama (2001,
2004). The IND spectra used in the development of the regression coefficients in Table
11.4 are also given in Farrow and Kurama (2001, 2004).
The dynamic analyses to determine the regression coefficients in Tables 11.4 and
11.5 were conducted following the procedure described in Farrow and Kurama (2001,
2003, 2004) using a viscous damping ratio of =5%, five strength ratios of R=1(linearelastic), 2, 4, 6, and 8, and thirty structure periods, exponentially spaced, between T=0.1
to 3.0 sec. A total of nine far-field ground motion ensembles were used (see Farrow
and Kurama 2001, 2003, 2004), corresponding to the site seismicities, seismic demand
levels, and soil classes listed above. Each ensemble contains 20 ground motion records,
resulting in a total of 180 records used in the SDOF dynamic time-history analyses. Note
309
that regression coefficients were not developed for sites with a soft soil type under the
survival demand level (blank cells in Tables 11.4 and 11.5), since IBC 2000 requires a
site-specific geotechnical investigation and dynamic site response analyses for this
condition.
The SDOF dynamic time-history analyses were conducted for selected values of R,
with the objective of determining the demands from each analysis. This approach is
referred to as the constant-R approach in Farrow and Kurama (2001, 2004). In Nassar
and Krawinkler (1991), the constant- approach was used, where the objective is to
determine the R values required to limit the demand to pre-selected values. Unlike the
constant-R approach, for each pre-selected demand in the constant- approach, a series
of nonlinear dynamic time-history analyses are conducted based on an initial value of R,
until the desired demand is obtained. Farrow and Kurama (2001, 2004) show that the
constant-R approach results in more conservative peak displacement demand
relationships than the constant- approach.
The use of the displacement demand relationships developed above is demonstrated
in the design of two prototype unbonded post-tensioned hybrid coupled wall structures in
Chapter 12. Note that, as demonstrated by Farrow and Kurama (2001, 2003, 2004), the
effect of structure hysteretic characteristics (e.g., and r) on the peak displacement
demands is small as compared with the effects of site seismicity, demand level, and soil
type. Thus, the use of assumed values for and r (based on the expected hysteretic
characteristics of the structure) is typically adequate for seismic design as shown in
Chapter 12.
11.12 Nonlinear Static Procedure
This section describes the main steps of the nonlinear static procedure (NSP) used
in the proposed seismic design approach.
11.12.1 Coupled Wall Design Base Shear Demand, Qwd
The determination of the coupled wall design base shear demand, Qwd requires that
trial values are selected for the number, nw and gross dimensions of the coupled wall
systems in a building. Then, a linear-elastic modal analysis of the structure is conducted
to determine the first (i.e., fundamental) mode shape, {1} and period, T1. Alternatively,
{1} and T1 may be estimated using approximate procedures [e.g., as described in the
Equivalent Lateral Load procedure of IBC 2000 (ICC 2000)].
The coupled wall design base shear demand, Qwd is determined by distributing the
total design base shear demand, Qsd for the entire building vertically over the height and
horizontally to the lateral load resisting members in the plan of the structure as described
in the Equivalent Lateral Force Procedure of IBC 2000. The total building design base
shear demand, Qsd is calculated as:
310
M eff f S as M eff f Sad

Qsd = max(Qds , Qdd ) = max
,
Rs
Rd
(11.48)
where, Qds and Qdd are the survival-level and design-level total building base shear
demands, Meff =Leq2/Meq is the design effective linear-elastic first mode mass of the
building (with Leq={1}T[M]{1} and Meq={1}T[M]{1} as described previously), f is a
damping adjustment coefficient, Sas and Sad are the site-adjusted linear-elastic design
spectral response accelerations (based on Teq=T1) for the survival-level and design-level
ground motions used in design (as determined from IBC 2000), and Rs and Rd are the
survival-level and design-level strength ratios calculated based on an equivalent SDOF
BP system used to represent the structure.
The damping adjustment coefficient f accounts for the difference (if any) in the
assumed viscous damping ratio, 1 for the structure being designed and the viscous
damping ratio, o corresponding to the Sas/Rs and Sad/Rd values used in design. For
example, a value of 1=3% is often assumed for dry-jointed precast concrete structures
whereas the Sas/Rs and Sad/Rd values from IBC 2000 are based on o=5%. The damping
adjustment coefficient accounts for the difference between 1 and o, and can be
calculated as (AIJ 1993):
f =
(1 + 25o )
(1 + 251 )
(11.49)
For o=5% and 1=3%, f is equal to 1.13. Note that this value is similar to the
damping adjustment coefficients given in the FEMA 356 Prestandard and Commentary
for the Seismic Rehabilitation of Buildings (FEMA 2000).
In accordance with the general principles of a displacement-based design approach,
the strength ratios Rs and Rd are used to determine the required base shear strengths, Qds
and Qdd to limit the peak roof drift demands of the structure to below selected allowable
target roof drift values, ts and td for the survival and design seismic demand levels,
respectively. This requires an iterative procedure for each selected allowable target roof
drift as follows:
1) Assume survival-level and design-level displacement ductility demands, ts
and td, corresponding to ts and td, respectively.
2) Determine Rs and Rd (corresponding to ts and td) from Equations (11.45)(11.47) for the equivalent SDOF BP system assumed to represent the structure.
3) Determine the total building design base shear demand, Qsd from Equation
(11.48).
311
4) If Qds governs Qsd in Step 3, then, calculate a revised Rd=(Rd from Step
2)(Qdd/Qsd) and determine the corresponding revised td. If Qdd governs Qsd,
then revise Rs instead.
5) Determine the coupled wall design base shear demand, Qwd by distributing Qsd
vertically over the height and horizontally to the lateral load resisting
members in the plan of the structure.
6) Estimate the nonlinear survival-level and design-level peak roof drift demands
s and d of the structure as:
s = ts
Qwd
hw K wi
(11.50)
d = td
Qwd
hw K wi
(11.51)
where, hw is the structure height and Kwi is the linear-elastic lateral stiffness of a
coupled wall system.
7) Compare the estimated nonlinear roof drift demands s and d with the
allowable drifts ts and td, respectively. Repeat Steps (1-6) if any of the
allowable drifts, ts or td, is exceeded or if both s and d are significantly
smaller than ts and td indicating an over-design. If the design objectives
cannot be achieved, the trial wall and beam dimensions (e.g., wall length, lw,
and thickness, tw) and/or the number of coupled wall systems, nw in the
structure may need to be revised.
The application of this iterative procedure to the design of two prototype coupled
wall structures is presented in Chapter 12.
11.12.2 Vertical Distribution of Seismic Lateral Forces
This section describes the distribution of the total building design base shear force,
Qsd as lateral forces applied at the floor and roof levels over the height of the structure.
The first linear-elastic mode shape of the structure, {1} is used to determine the lateral
force fx applied at level x as follows:
with
f x = CvxQsd
(11.52)
M x1x
Leq
(11.53)
Cvx =
where,
312
Cvx=vertical distribution factor

Mx= portion of the building seismic mass assigned to level x
1x=coordinate of { 1 } at level x
Leq={1}T[M]{1}
[M]=diagonal lumped mass matrix of the structure.
Alternatively, the Equivalent Lateral Force procedure described in IBC 2000 (ICC
2000) can be used to distribute the base shear force over the height of the structure.
11.12.3 Horizontal Distribution of Story Forces
Once the total building design base shear force is distributed over the structure
height, then, the total design story shear force and story moment demands can be
determined from the floor/roof lateral forces using equilibrium. The horizontal (i.e., in the
plan of the structure) distribution of the design story shear and moment demands between
the lateral load resisting systems in the structure depends on the in-plane stiffness of the
floor/roof diaphragms. This report focuses on the design of buildings that can be assumed
to have rigid diaphragms (in the plane of the diaphragms). Thus, the horizontal
distribution of the story forces is based only on the distribution of story mass and
stiffness in the plan of the structure.
The horizontal distribution of the story forces should consider applied torsional
moments resulting from the difference (if any) in the locations of the center of mass and
center of stiffness at each floor or roof level under consideration. The center of mass (CM)
is the location where the inertial forces generated at a floor/roof level are assumed to be
concentrated and the center of stiffness (CS) represents the point where the application of
a lateral force would cause only relative floor/roof translations without twist. Since the
inertial forces act through CM rather CS, relative floor rotations (twist) as well as floor
translations occur if CM and CS differ. For regular structures, CM and CS are at the same
location over the height of the structure.
In addition, an accidental torsion may be introduced during the seismic response
of a structure because of differences between the estimated and actual stiffnesses of the
lateral load resisting members, different amounts of stiffness degradation in the lateral
load resisting members during nonlinear response, differences between assumed and
actual distributions and values for the floor and roof masses, etc. For this reason,
allowance is made in IBC 2000 (ICC 2000) for an accidental torsional moment caused by
an assumed re-location of the center of mass (in each orthogonal direction of the building)
from its actual location by a distance equal to 5% of the dimension of the building
perpendicular to the direction of the applied forces.
The portion of the building design story shear force resisted by one coupled wall
system is given as:
Qwd = vQsd
313
(11.54)
The distribution factor v can be expressed as:

v = v'+v"
(11.55)
where, v' represents the effect of story translations and v" represents the effect of story
twist from applied and accidental eccentricities, e=eapp+eacc. Since regular structures with
no applied torsion are considered in this report, eapp=0 and e=eacc (equal to 5% of the
dimension of the building perpendicular to the direction of the applied forces).
Based on the assumption of rigid floor and roof diaphragms, it can be shown that,
v' =
v" =
and
K wi
K si
esw K wi
(sl2 Kli )
(11.56)
(11.57)
where, Kwi and Ksi are the initial linear elastic lateral stiffnesses of the coupled wall
system being designed and of the entire building, respectively, in the direction considered,
sw is the distance (in the perpendicular direction) of the shear center of the coupled wall
system from the center of story stiffness (CS), and sl and Kli represent similar quantities
to sw and Kwi for all of the lateral load resisting systems in the plan of the structure
(including the coupled wall systems and any other lateral load systems in both directions
of the building). In determining v, the stiffnesses of the lateral load resisting systems in
their weak (i.e., transverse) direction are ignored.
11.12.4 Load Combinations
The IBC 2000 (ICC 2000) provisions specify two load combinations (lc1 and lc2
below) for seismic design. In lc1, the effects of gravity loads and seismic loads are
additive, and in lc2, the effects of gravity loads counteract the seismic load effects as
follows:
lc1:
1.2 D + 1.0 E + 0.5 L + 0.2S , where E = Q E + 0.2 S DS D resulting in

(1.2 + 0.2S DS ) D + Q E + 0.5L + 0.2S
(11.58)
lc2:
0.9 D + 1.0 E , where E = QE 0.2S DS D resulting in

(0.9 0.2S DS ) D + Q E
where,
D
E
QE
= effect of dead load;

= combined effect of horizontal and vertical seismic forces;
= effect of horizontal seismic forces;
314
(11.59)
SDS
= site-adjusted design-level spectral acceleration at short periods,
SDS=(2/3)SMS;
= reliability factor based on system redundancy as described in IBC 2000;

L
= effect of live loads; and
S
= effect of snow loads.
11.13 Maximum Base Shear Demand
The displacement-based seismic design procedure described in this chapter

assumes that the nonlinear response of the structure is governed by the first (i.e.,
fundamental) linear elastic mode of vibration. While this assumption has been shown to
provide reasonable estimates for the maximum displacement demands of most regular
structures, the maximum base shear demands under earthquake loading may be grossly
underestimated due to the effects of higher modes of vibration, especially in wall
structures (e.g., Ghosh and Markevicius 1990; Aoyama 1993; Eberhard and Sozen 1993;
Kabeyasawa 1993; Otani et al. 1994).
During a predominantly first-mode response of a wall structure under earthquake
loading, the distribution of inertial forces is similar to the one shown in Fig. 11.13a. The
force pattern is similar to that of standard code-specified equivalent lateral forces, with a
resultant force located near H=0.67hw above the base (for regular structures). At some
instants during the nonlinear vibrations of the structure, the response accelerations may
be strongly influenced by the higher modes (especially the second and third modes),
resulting in inertial force distributions similar to the one in Fig. 11.13b, with a resultant
force located much closer to the wall base (as low as H=0.3hw). For a flexural wall (i.e., a
wall with nonlinear behavior dominated by flexure rather than shear effects) with given
base moment capacity, Mwu the resulting maximum base shear demand can be several
times larger than the base shear demand estimated using the first mode distribution of
inertial forces. The bending moment diagrams corresponding to the lateral force patterns
in Figs. 11.13a and 11.13b are shown in Fig. 11.13c.
In the linear-elastic range of response, the maximum base shear demand during a
ground motion can be determined from modal analysis. The modal base shear force,
which is equal to the sum of the inertial forces that develop in that mode, can be
determined as the product of the effective modal mass and the modal acceleration. The
effective modal mass is the part of the total mass responding to the ground motion in that
mode, and depends on the mode shape and mass distribution over height. The modal
acceleration depends on the modal period and linear-elastic response acceleration
spectrum (for a given value of viscous damping). The effective modal mass decreases for
higher modes, and thus, in the linear-elastic range, the largest modal contribution to the
total base shear demand is from the first mode, and the higher mode contributions are
usually negligible.
315
single or coupled
F=Mwu/(0.67hw) wall system
single or coupled
wall system
(a)
(b)
F=Mwu/(0.3hw)
hw
H=0.67hw
H=0.3hw
F
base moment, Mwu
base moment, Mwu
(a)
(b)
Mwu
(c)
Fig. 11.13 Distribution of inertial forces: (a) H=0.67hw; (b) H=0.3hw;

(c) bending moment diagrams (adapted from Paulay and Priestley 1992)
In the nonlinear range of response, the effect of higher modes on the maximum
base shear demand increases. Nonlinear behavior results in a decrease in the lateral
stiffness (i.e., softening) of the structure, which results in elongation of the modal periods.
Period elongation typically results in a decrease in modal acceleration for the first mode
and an increase in modal acceleration for the higher modes. Nonlinear behavior also
changes the mode shapes, resulting in, for a typical wall, an increase in the effective
modal mass for the first mode and a decrease in the effective modal mass for the higher
modes. The effect of nonlinear behavior on the effective modal mass is small compared
to the effect on the modal acceleration, and thus, the contribution of higher modes to the
base shear demand increases and can be comparable to or even higher than the
contribution of the first mode. In the nonlinear range of response under earthquake
loading, the effect of higher modes on the maximum base shear demand of wall
structures cannot be neglected.
Maximum base shear demands for unbonded post-tensioned hybrid coupled wall
structures can be estimated using a method developed for uncoupled wall systems by
Kabeyasawa (1993) and Aoyama (1993). In this method, the maximum base shear
demand, Qw,max is estimated as the sum of a first mode component Q1,max and a higher
mode component Qh,max as:
Qw, max = Q1, max + Qh, max
(11.60)
The first mode component is estimated using the maximum base moment capacity
(corresponding to the coupled wall ultimate state), Mwu and the resultant height of the
first mode distribution of inertial forces, H1 as:
Q1, max =
M wu
H1
316
(11.61)
The IBC 2000 (ICC 2000) distribution of equivalent lateral forces is similar to the
first mode distribution of inertial forces and can also be used to determine H1.
The higher mode component of maximum base shear demand is estimated as:
Qh, max = Dm mw ( PGAs )
(11.62)
where, mw is the total mass assigned to the coupled wall system (assumed to be equal to
the total building mass, M divided by the number of coupled wall systems, nw) and PGAs
is the peak acceleration for the survival-level ground motion.
The coefficient Dm depends on the mode shapes and mass distribution over the
height of the structure. As described in Kurama (1997), the expression for Dm developed
by Kabeyasawa (1993) and Aoyama (1993) can be written as:
Dm =
mw meff + 0.7meff , 2
mw
= 1
meff
mw
+ 0.7
meff , 2
mw
(11.63)
where, meff is the effective first mode mass and meff,2 is the effective second mode mass
assigned to the coupled wall system (assumed to be equal to the total building effective
first and second mode masses, Meff and Meff,2, respectively, divided by the number of
coupled wall systems, nw).
Performance-based seismic design requires estimates of structural capacities and

seismic demands under different levels of seismic intensity. The main steps of the
performance-based design procedure described in this chapter are summarized below.
Application of this procedure to the design of two prototype structures is described in
Chapter 12.
Step 1
Define the seismic demand levels, desired performance objectives, and allowable
target roof drifts, ts and td, of the structure.
Step 2
Define the lateral system layout of the building, and then, select initial trial
dimensions for the walls and the coupling beams.
Step 3
Perform a linear-elastic analysis of the structure to compute the linear elastic
stiffnesses Kwi and Ksi, first mode period, T1, first mode vibration shape, {1},
participation factor, , and effective first mode mass, Meff.
317
Step 4
Define the design acceleration response spectra and determine the survival-level
and design-level response accelerations Sas and Sad corresponding to the first mode period,
T1.
Step 5
Select and r parameters for an equivalent SDOF BP model to represent the
structure.
Step 6
Assume survival-level and design-level displacement ductility demands, ts and td,
corresponding to ts and td, respectively, and calculate strength ratios Rs and Rd from
available R--T relationships for the equivalent SDOF BP model.
Step 7
Determine the total building design base shear force, Qsd using the nonlinear static
procedure described previously and revise Rs or Rd as necessary.
Step 8
Determine the coupled wall design base shear demand, Qwd by distributing Qsd
vertically over the height and horizontally to the lateral load resisting members in the
plan of the structure.
Step 9
Estimate the survival-level and design-level peak roof drift demands, s and d, of
the structure and compare with the allowable target roof drifts, ts and td. If the design
requirements cannot be achieved, the assumed ts and td values, trial wall and beam
dimensions, and/or the number of coupled wall systems in the structure may need to be
revised.
Step 10
Complete the detailed design of the concrete walls and the steel coupling beams to
achieve a desired amount of coupling in each coupled wall system. This is described in
Chapter 12. Consider all of the load combinations required by the governing model
building code.
Step 11
Check all applicable seismic design criteria (i.e., acceptance criteria) to ensure that
the target performance objectives are achieved (see Chapter 12).
318
CHAPTER 12
DESIGN OF PROTOTYPE STRUCTURES
This chapter describes the seismic design of two prototype unbonded posttensioned hybrid coupled wall structures, referred to as Structures P1-CWUPT and P2PWUPT, based on the performance-based design approach introduced in Chapter 11. The
chapter is divided into the following sections: (1) design overview; (2) performancebased seismic design parameters; (3) linear-elastic structure properties; (4) design base
shear demand; (5) preliminary design checks; (6) design of steel coupling beams; (7)
design of wall piers in Structure P1-CWUPT; (8) design of wall piers in Structure P2PWUPT; and (9) summary of design and final checks. The design of the prototype
structures is critically evaluated in Chapters 13 and 15 based on nonlinear static analyses
and nonlinear dynamic time-history analyses under a series of ground motion records
representative of the seismic conditions used in the design.
12.1 Design Overview
Two eight-story office building structures are designed for a stiff soil site
[corresponding to Site Class D in IBC 2000 (ICC 2000)] in Los Angeles, California. The
plan view of the structures, which is identical for the two buildings, is shown in Fig. 12.1.
The lateral load resistance of each structure in the E-W direction is provided by four
identical unbonded post-tensioned hybrid coupled wall systems. This chapter focuses on
the seismic design of these coupled wall systems. The seismic design of the lateral load
resisting systems in the transverse N-S direction of the prototype structures is not
addressed.
Two types of structures are considered: (1) with monolithic cast-in-place reinforced
concrete wall piers (referred to as Structure P1-CWUPT); and (2) with precast concrete
wall piers (referred to as Structure P2-PWUPT). These coupled wall systems are similar
to the CIP-UPT and PRE-UPT systems described in Chapter 6, but the design properties
are different.
The selection of the structure layout is based on the recommendations of the
National Science Foundation U.S.-Japan Cooperative Research Program on Composite
and Hybrid Structures (Goel and Yamanouchi 1992). The center-to-center plan
dimensions of the structure are 7.92 m 5 bays = 39.6 m (26 ft 5 bays = 130 ft) in the
E-W direction and 9.75 m 1 bay + 7.32 m 2 bays = 24.4 m (32 ft 1 bay + 24 ft 2
bays = 80 ft) in the N-S direction.
319
The structure height is 32.6 m (107 ft) with 4.88 m (16 ft) for the first story and
3.96 m (13 ft) for the remaining seven stories. A summary of the general properties of the
prototype structures is given in Table 12.1 and described below.
(24 ft)
9.75 m
(32 ft)
(24 ft)
slab
7.32 m
7.32 m
gravity
beam
wall
coupling
beam
prototype coupled
wall system
wall
gravity
beam
N-S
column
7.92 m
7.92 m
7.92 m
7.92 m
7.92 m
(26 ft)
(26 ft)
(26 ft)
(26 ft)
(26 ft)
Fig. 12.1 Plan view of prototype buildings

TABLE 12.1
GENERAL PROPERTIES OF PROTOTYPE STRUCTURES
tw
lb Beam db bbf tbf
W
M
ww mw
hw lw
tbw
Ab
(m) (m) (mm) (m) section (mm) (mm) (mm) (mm) (mm2) (kN) (ton) (kN) (ton)
W21
P1-CWUPT 4 32.6 4.88 356 2.13
560 318 29.2 18.3 27871 73712 7524 18428 1881
147
W21
P2-PWUPT 4 32.6 4.88 356 2.13
560 318 29.2 18.3 27871 74568 7608 18642 1902
147
Structure nw
where, nw is the number of coupled wall systems in the E-W direction, hw, lw, and tw are
the wall height, length, and thickness, respectively, lb, db, bbf, tbf, tbw, and Ab are the
coupling beam length, depth, flange width, flange thickness, web thickness, and cross
section area, respectively, W and M are the total building seismic weight and mass, and
ww and mw are the seismic weight and mass assigned to one coupled wall system only,
respectively.
320
12.1.1 Design Material Properties

The material properties used in the design of the prototype structures are as follows:
Cast-in-Place Concrete Wall Piers
Max. (i.e., peak) strength of unconfined conc. (normal wt.),
Youngs modulus of wall concrete,
Yield strength of mild steel reinforcement,
Youngs modulus of mild steel reinforcement,
fc=41.4 MPa (6 ksi)

Ec=30441 MPa (4415 ksi)
fwsy=414 MPa (60 ksi)
Ews=199955 MPa (29000 ksi)
Precast Concrete Wall Piers

Max. (i.e., peak) strength of unconfined conc. (normal wt.),
Youngs modulus of wall concrete,
Yield (i.e., linear-limit) strength of post-tensioning bars,
Maximum (i.e., peak) strength of post-tensioning bars,
Youngs modulus of post-tensioning bars,
fc=41.4 MPa (6 ksi)

Ec=30441 MPa (4415 ksi)
fwpy=828 MPa (120 ksi)
fwpu=1035MPa (150 ksi)
Ewp=199955 MPa (29000 ksi)
Steel Coupling Beams

Yield strength of beam steel (A572 Grade 50 steel),
Maximum strength of beam steel (A572 Grade 50 steel),
Youngs modulus of beam steel,
Yield strength of angle steel (A572 Grade 50 steel),
Youngs modulus of angle steel,
Yield (i.e., linear-limit) strength of post-tensioning strands,
Maximum (i.e., peak) strength of post-tensioning strands,
Youngs modulus of post-tensioning strands,
fby=345 MPa (50 ksi)

fbm=448 MPa (65 ksi)
Eb=199955 MPa (29000 ksi)
fay=345 MPa (50 ksi)
Ea=199955 MPa (29000 ksi)
fbpy=1689 MPa (245 ksi)
fbpu=1862 MPa (270 ksi)
Ebp=196508 MPa (28500 ksi)
12.1.2 Selection of Initial Dimensions for Structural Members

As described in Chapter 11, the proposed design approach requires the selection of
initial dimensions for the structural members, provided below. Note that these initial
dimensions may need to be modified during the detailed design of each structure.
Wall piers
Each coupled wall system includes two identical concrete wall piers with
rectangular cross sections. The wall pier length, lw=4.88 m (16 ft) and wall thickness, tw=
356 mm (14 in.). The wall height, hw=32.6 m (107 ft) with 4.88 m (16 ft) for the first
story and 3.96 m (13 ft) for the upper stories. This results in a height, hw to length, lw
aspect ratio of 6.69 for each wall pier.
Coupling beams
To simplify the design process, the coupling beams at the floor and roof levels are
assumed to be identical. A W21147 section is selected, with length lb=2.13 m (7 ft),
depth db=560 mm (22 in.), and other geometric properties as provided in Table 12.1. This
results in a coupling beam length, lb to depth, db aspect ratio of 3.8.
321
Floor and roof slabs

For the structure with cast-in-place reinforced concrete wall piers (i.e., Structure
P1-CWUPT), a 178 mm (7 in.) thick cast-in-place reinforced concrete slab is assumed for
the floor and roof levels.
For the structure with precast concrete wall piers (i.e., Structure P2-PWUPT), 1.22
m 203 mm (4 ft 8 in.) hollow-core panels with 64 mm (2.5 in.) cast-in-place concrete
topping is used for the floor and roof slabs. The hollow-core floor and roof panels span in
the E-W direction.
Other structural members
Geometric properties of the other structural members in the prototype structures
(i.e., walls in the N-S direction, gravity beams, and columns as shown in Fig. 12.1) are
needed to estimate the total structure self weight. The gravity beams are assumed to have
W2476 steel cross sections and the columns are assumed to have 610 mm 610 mm
(24 in. 24 in.) concrete cross sections (for simplicity, the column size is assumed to be
uniform over the structure height). The concrete walls in the transverse direction (N-S
direction) are assumed to be 4.88 m (16 ft) long and 381 mm (15 in.) thick. Since the
interaction between these structural members and the unbonded post-tensioned coupled
wall systems in the prototype buildings is ignored, detailed design of the structural
members other than the coupled wall systems is not addressed in this report.
12.1.3 Design Gravity Loads and Seismic Weight
A summary of the gravity loads used in the design of Structures P1-CWUPT and
P2-PWUPT is given below. The structures are assumed to have similar gravity loads.
Floor loads
Dead load: structure self weight + 1.44 kPa (30 psf) superimposed dead load
Live load: 2.39 kPa (50 psf)
Cladding: 8.03 kN/m (550 plf) on perimeter of building
Roof loads
Dead load: structure self weight + 1.44 kPa (30 psf) superimposed dead load
Live load: 0.57 kPa (12 psf)
Parapet:
8.03 kN/m (550 plf) on perimeter of building
Loads that are not considered include elevators, stairways, and any heavy
equipment.
Based on the Precast/Prestressed Concrete Institute (PCI) Design Handbook (PCI
2004), the weight of the 1.22 m 203 mm (4 ft 8 in.) hollow core floor/roof slab panels
in Structure P2-PWUPT is 2.68 kPa (56 psf) and the weight of the topping is 1.50 kPa
(31 psf).
322
The 1.44 kPa (30 psf) superimposed dead load includes 0.96 kPa (20 psf) for
partitions (ICBO 1997) and 0.48 kPa (10 psi) for ceiling, finish, electrical equipment,
mechanical ducts, piping, etc.
The floor and roof live loads are taken from the American Society of Civil
Engineers (ASCE) Standard for Minimum Design Loads for Buildings and Other
Structures (ASCE 2003).
Based on Fig. 12.1 and the preliminary member dimensions, the seismic weights at
the floor and roof levels of the prototype structures are calculated from the structure self
weight plus the superimposed dead loads as shown in Table 12.2. The unfactored
tributary floor/roof gravity dead and live loads for each coupled wall system are also
provided in Table 12.2. Note that the dead and live loads are for the entire coupled wall
system (two wall piers).
TABLE 12.2
SEISMIC WEIGHTS AND GRAVITY LOADS
Structure
P1-CWUPT
P2-PWUPT
Floor/Roof
Level
roof
3 -8th floor
2nd floor
roof
rd th
3 -8 floor
2nd floor
rd
Seismic Weights
for Entire Building
(kN)
7953
9348
9672
8056
9456
9776
Dead Loads per

Coupled Wall System
(kN)
659
659
731
667
667
739
Live Loads per

Coupled Wall System
(kN)
33
139
139
33
139
139
The floor and roof seismic weights in Table 12.2 are for the entire building. The
resulting total building seismic weights (i.e., sum of the floor and roof seismic weights),
W are shown in Table 12.1. The seismic weights, ww for individual coupled wall systems
are determined by dividing W with the number of coupled wall systems, nw. The
corresponding seismic masses, M and mw are calculated by dividing the seismic weights,
W and ww by the gravitational acceleration, g.
12.1.4 Design Load Combination and Capacity Reduction Factors
No capacity reduction factors are applied in the design of the prototype structures.
A single hypothetical load combination is used as follows:
Design Load Combination: 1.0 D + 0.25 L + 1.0QE
where,
D=effect of dead loads given in Table 12.2
323
(12.1)
L=effect of live loads given in Table 12.2 (including roof live load, with no
live load reduction)
QE=effect of horizontal seismic forces
The 25% design live load is assumed to represent the amount of live load that may
be acting on the structures during an earthquake. The analytical evaluation of the
structures in Chapters 13 and 15 is also conducted under this hypothetical load condition.
Note that code specified capacity reduction factors and load factors/combinations (e.g.,
ICC 2000, see Chapter 11) should be used in practice; however, this is not done in this
report to allow for more direct comparisons between the design estimations and the
analysis results.
12.1.5 Design Scope and Limitations
Only the design of the coupled wall structures in the E-W direction of the prototype
buildings is considered in this report. The prototype buildings are assumed to have no
plan or vertical irregularities and torsional loading effects (applied and accidental) are
ignored. The floor and roof diaphragms are assumed to be rigid in their planes. The effect
of the other structural members (e.g., floor/roof diaphragms, gravity system) on the
behavior and design of the coupled wall structures is not considered.
12.2 Performance-Based Seismic Design Parameters
This section describes the parameters used in the performance-based seismic design
of the prototype structures as follows: (1) allowable target roof drift demands; (2)
allowable target inter-story drift demands; and (3) acceleration response spectra used in
design.
12.2.1 Allowable Target Roof Drift Demands
The roof drift is defined as the lateral displacement of the structure at the roof level
divided by the structure height. As described in Chapter 11, the proposed design
approach requires the selection of allowable target roof drift demands for the survival
and design seismic demand levels. The following allowable target roof drift demands
are selected for the design of the prototype structures:
Survival-level demand,
ts=1.75%
Design-level demand,
td=1.17%
The design-level allowable roof drift demand, td is selected as 2/3 times the
survival-level demand, ts in order to provide an adequate margin of safety between the
life safety and collapse prevention performance levels used in design (see Chapter 11).
324
12.2.2 Allowable Inter-Story Drift Demands

The inter-story drift is defined as the relative lateral displacement between two
adjacent floor/roof levels in a structure divided by the story height. For wall structures,
the inter-story drift demand is largest in the top story and is slightly larger than the roof
drift demand. According to IBC 2000 (ICC 2000), the peak inter-story drift under the
design-level earthquake should not exceed 2%. The 2% inter-story drift limit is often
specified for controlling damage in basic access and life safety systems.
Because of the relatively small allowable target roof drift demands specified for
design (ts=1.75% and td=1.17%), it is expected that the peak inter-story drift demands
of the prototype structures will satisfy the 2% limit in IBC 2000 (ICC 2000) not only for
the design-level earthquake, but also for the survival-level earthquake. This is
investigated further in Chapter 15. For design purposes, allowable target inter-story drift
demands of 2% and 1.33% are used for the survival and design seismic demand levels.
12.2.3 Acceleration Response Spectra Used in Design
The survival-level and design-level acceleration response spectra used in the design
of the prototype structures are constructed based on the recommendations of IBC 2000
(ICC 2000) as described in Chapter 11. The mapped spectral response accelerations at
short periods, SS, and at 1 second period, S1, for the maximum considered earthquake
(which is referred to as the survival-level earthquake in this report) can be obtained from
the contour maps in IBC 2000, from contour maps published by the Federal Emergency
Management Agency (BSSC 1998), or from the United States Geological Survey (USGS
1996).
The SS and S1 values used in the design of the prototype structures (Table 12.3)
were obtained from the online USGS National Seismic Hazard Mapping Project (USGS
1996) by specifying the latitude and longitude coordinates for the geographic center for
Los Angeles. These acceleration coefficients were verified manually by inspecting
hardcopy contour maps from IBC 2000.
TABLE 12.3
ACCELERATION RESPONSE SPECTRA
Site
SS (g) S1 (g) Site Soil Fa
Fv SMS (g) SM1 (g) SDS (g) SD1 (g) PGAs (g) PGAd (g)
Los Angeles 2.05 0.81 Class D 1.0 1.5 2.05

1.22
Note: g, acceleration of gravity=9.81 m/sec2 (386 in./sec2).
1.37
0.81
0.82
0.55
The site-adjusted survival-level spectral response accelerations for short periods,

SMS, and at 1 second period, SM1 (see Table 12.3) are determined by multiplying the SS
and S1 values with the Fa and Fv site coefficients, respectively, for Site Class D.
325
Finally, the design-level spectral response accelerations at short periods, SDS, and at
1 second period, SD1 (listed in Table 12.3) are determined by multiplying the SMS and SM1
values by 2/3 (i.e., the seismic margin).
The resulting survival-level and design-level acceleration response spectra used in
the design of the prototype structures are shown in Fig. 12.2. Note that these response
spectra are the same as the DES spectra used to generate the a and b regression
coefficients in Table 11.5 for Site Class D in Los Angeles.
Design Spectral Response Acceleration, Sas and Sad (g)
2.5
= 5%
S MS = 2.05
2.0
survival level spectrum
1.5
S DS = 1.37
S M1 /T=1.22/T
1.0
PGA S = 0.4SMS =0.82
0.5
design level spectrum
S D1 /T=0.81/T
PGA d= 0.4S M1 =0.55

0.5S 1= 0.41
T0 =0.2S M1/S MS = 0.12
0.5
TS =SM1 /SMS = 0.59
1.0
1.5
Period, T (sec.)
Fig. 12.2 Design-level and survival-level acceleration response spectra

Based on the provisions of IBC 2000 (ICC 2000), it is assumed that the prototype
buildings are in Seismic Use Group I with an Occupancy Importance Factor of I=1.0. The
buildings are assigned to Seismic Design Category E since S1>0.75g. Note that for
Seismic Design Category E, the spectral acceleration cannot be taken less than
0.5S1=0.41g. This limit is shown by the dashed horizontal line in Fig. 12.2.
12.3 Linear-Elastic Structure Properties
Linear-elastic analyses are conducted to determine the linear-elastic lateral stiffness,
first (fundamental) vibration mode period and shape, and effective first and second mode
masses of the prototype buildings. Out-of-plane and torsion effects are ignored, and it is
assumed that the four coupled wall systems in the E-W direction of the buildings are
connected through rigid floor and roof diaphragms.
326
12.3.1 Linear Elastic Stiffness

The linear-elastic lateral stiffness of the coupled wall systems, Kwi (i.e., slope of the
base shear force versus roof displacement relationship in the initial linear-elastic range) in
the prototype structures is estimated using the Matlab (2000) algorithm described in
Chapter 8. Note that other appropriate structural analysis programs can also be used to
conduct this linear-elastic analysis.
The input parameters to estimate Kwi include the gross wall and beam dimensions
and the linear-elastic properties for wall concrete and beam steel. The deformations in the
wall-contact regions adjacent to the coupling beams are ignored and the coupling beams
are assumed to be fixed to the walls with no top and seat angles. Shear deformations in
the walls and the coupling beams are included.
The resulting linear-elastic stiffness of each coupled wall system, Kwi and of the
building as a whole, Ksi (determined by multiplying Kwi with the number of coupled wall
systems, nw) are given in Table 12.4. Note that the estimated lateral stiffnesses of
Structures P1-CWUPT and P2-PWUPT are the same since the same dimensions and
linear-elastic material properties are assumed for the lateral load resisting members in
both structures.
TABLE 12.4
LINEAR-ELASTIC STIFFNESSES
Structure
Kwi (kN/mm)
Ksi (kN/mm)
P1-CWUPT
P2-PWUPT
105
105
420
420
12.3.2 Period, Mode Shape, and Effective Mass

The Matlab algorithm described in Chapter 8 is also used to conduct linear-elastic
modal analyses of the prototype structures. The structure seismic masses are lumped at
the floor and roof levels, and are assumed to be distributed equally between the four
coupled wall systems.
The resulting first mode period, T1, first mode shape, {1}, coupled wall effective
masses for the first and second modes, meff and meff,2, and building effective masses for
the first and second modes, Meff and Meff,2, are given in Table 12.5. The building effective
masses are determined by multiplying the coupled wall effective masses with the number
of coupled wall systems, nw in each building.
327
TABLE 12.5
LINEAR-ELASTIC MODAL ANALYSIS RESULTS
Structure
T1
(sec.)
P1-CWUPT
0.61
P2-PWUPT
0.62
{1}
(base to roof)
{0.0532 0.153 0.282 0.428
0.579 0.728 0.868 1.00}
{0.0532 0.153 0.282 0.428
0.579 0.728 0.868 1.00}
meff
(ton)
meff,2
(ton)
Meff
(ton)
Meff,2
(ton)
1343
327
5372
1308
1360
331
5440
1324
12.4 Design Base Shear Demand

This section describes the determination of the coupled wall design base shear
demand, Qwd using the nonlinear static procedure outlined in Chapter 11.
12.4.1 Displacement Demand Relationships
As described in Chapter 11, an equivalent single-degree-of-freedom (SDOF) BP
model is used to establish displacement ductility demand, relationships for the
prototype structures. The form of the ductility demand relationship is:
R ( , T , ) = [c( 1) + 1]1 / c
c(T , ) =
Ta
b
T a +1 T
+
(12.2)
(12.3)
BP model regression coefficients a and b for selected seismic design conditions are
provided in Chapter 11. The coefficients used in the design of the prototype buildings,
taken from Chapter 11, are listed in Table 12.6. Note that two sets of regression
coefficients, based on IND spectra and DES spectra, are provided in Chapter 11. The
coefficients used in the design of the prototype structures are based on the DES spectra in
Fig. 12.2, and thus, the resulting displacement demand estimates are expected to be
compatible with the lateral strengths of the prototype structures. This is described in
detail in Farrow and Kurama (2001, 2004).
It is assumed that the hysteretic characteristics of Structures P1-CWUPT and P2PWUPT can be represented using SDOF BP models with r=1/2 and r=1/3, respectively.
The larger r value used for Structure P1-CWUPT reflects the larger energy dissipation
expected to occur near the bases of the cast-in-place wall piers in the structure, as
compared with the smaller energy dissipation provided by the precast concrete wall piers
in Structure P2-PWUPT. The post-yield stiffness for both structures is assumed to be
equal to =0.10.
328
TABLE 12.6
DISPLACEMENT DEMAND RELATIONSHIPS
Structure
Period, T
(sec.)
BP
Model
Site
Seismicity
Site Soil
Class
P1-CWUPT
0.61
r=1/2,
=0.10
Los Angeles
P2-PWUPT
0.62
r=1/3,
=0.10
Los Angeles
Demand
Level
Survival
(DES spectra)
a
b
2.66
0.85
Design
1.07
0.55
Survival
2.84
0.88
Design
1.17
0.59
Note that, as described in Chapter 11, the use of assumed values for and r (based
on the expected hysteretic characteristics of the prototype buildings) is often adequate for
seismic design, because of the relatively small effect of these parameters on the peak
displacement demands. The most important factors affecting the displacement demands
are the site seismicity, soil class, and demand level as discussed in Farrow and Kurama
(2004).
12.4.2 Strength Ratios
The strength ratios, Rs and Rd, for the survival and design seismic demand levels,
respectively, are determined from Equations (12.2) and (12.3) using the a and b
regression coefficients in Table 12.6, the estimated structure linear-elastic first mode
period, T1 (see Table 12.5), and the survival-level and design-level target displacement
ductility demands, ts and td (corresponding to the allowable target roof drift demands,
ts and td used in design).
For the prototype structures, the following target displacement ductility ratios are
assumed for the survival and design seismic demand levels:
Survival-level demand,
ts=12 corresponding to ts=1.75%
Design-level demand,
td=8 corresponding to td=1.17%
The resulting strength ratios, Rs and Rd, used in the design of the prototype
structures are provided in Table 12.7.
TABLE 12.7
DISPLACEMENT DUCTILITY DEMANDS AND STRENGTH RATIOS
Structure
P1-CWUPT
P2-PWUPT
Demand Level
ts or td (%)
ts or td
Rs or Rd
Survival
1.75
12
6.2
Design
1.17
6.1
Survival
1.75
12
6.1
Design
1.17
5.9
329
12.4.3 Building Design Base Shear Demands
As described in Chapter 11, the building design base shear demand, Qsd is
determined from:
M eff f S as M eff f S ad
Qsd = max(Qds , Qdd ) = max
,
Rs
Rd
(12.4)
with the Meff, Rs, and Rd values for the prototype structures given in the previous
sections. The survival-level and design-level site-adjusted linear-elastic spectral response
accelerations, Sas and Sad (corresponding to the period T1 from Table 12.5) as determined
from Figure 12.2, are listed in Table 12.8.
It is assumed that the prototype structures have a viscous damping ratio of 1=3%
in the first mode. Since the design spectra in Fig. 12.2 correspond to a viscous damping
ratio of o=5%, a damping adjustment coefficient, f equal to 1.13 (see Chapter 11), is
used in Equation (12.4). The resulting building design base shear demands are provided
in Table 12.8. It is observed that the lateral strength of both structures is governed by the
survival-level demand.
TABLE 12.8
DESIGN BASE SHEAR DEMANDS
Structure
P1-CWUPT
1.13
P2-PWUPT
1.13
Demand Level Sas or Sad (g) Qds or Qdd (kN) Qsd (kN)
Survival
1.99
19128
Design
1.32
13113
Survival
1.98
19572
Design
1.31
14441
Qwd (kN)
19128
4782
19572
4893
12.4.4 Coupled Wall Design Base Shear Demands
To allow for more direct comparisons between the design target seismic
performance objectives and the estimated demands from the dynamic analyses in Chapter
15, the coupled wall design base shear demands, Qwd are determined by distributing the
total building design base shear demands, Qsd equally between the coupled wall systems
in each prototype structure as:
Qwd =
Qsd
nw
(12.5)
Thus, any torsional effects are ignored in the calculation of Qwd. The resulting
coupled wall design base shear demands for the prototype structures are provided in
Table 12.8.
330
Note that in practice, the distribution of the building design base shear demand to
the lateral load resisting members should be done based on the governing seismic design
provisions, including any applied torsion (due to the differences between the center of
mass and center of stiffness at the floor and roof levels) and accidental torsion (see
Chapter 11).
12.5 Preliminary Design Checks
This section describes the preliminary design checks for the prototype structures
prior to conducting a more detailed design of the wall and beam members as described
later in the chapter. Three preliminary design checks are conducted: (1) peak roof drift
demands; (2) peak inter-story drift demands; and (3) diagonal tension at wall pier bases.
12.5.1 Peak Roof Drift Demands
Once the coupled wall design base shear demands are determined, the trial
structures are checked to verify that the displacement-based design objectives can be
achieved. The peak lateral roof drift demands of the prototype structures are estimated
based on the linear elastic stiffness, Kwi and the roof drift ductility demands, ts and td as:
s = ts
Qwd
hw K wi
(12.6)
d = td
Qwd
hw K wi
(12.7)
Note that since Qsd and Qwd for both structures are governed by Qds, the designlevel strength ratios need to be revised as: Rd=(Rd assumed in Table 12.7)(Qdd/Qsd). The
resulting revised Rd ratios and the corresponding revised design-level roof drift ductility
demands from Equations (12.2) and (12.3) are given in Table 12.9.
The estimated survival-level and design-level roof drift demands, s and d for
both prototype structures, also listed in Table 12.9, satisfy the allowable target roof drift
demands, ts and td, respectively. Thus, the preliminary design checks for the peak roof
drift demands of the structures are satisfied.
TABLE 12.9
PEAK ROOF DRIFT DEMANDS
Structure
P1-CWUPT
P2-PWUPT
Demand Level
ts or td (%)
Rs or Rd
ts or td
s or d (%)
Survival
1.75
6.2
12
1.67
Design
1.17
4.2
5.1
0.71
Survival
1.75
6.1
12
1.71
Design
1.17
4.4
5.6
0.79
331
If the design requirements are not met (i.e., any of the allowable target roof drifts,
ts or td, is exceeded or both s and d are significantly smaller than ts and td
indicating an over-design), then, the assumed displacement ductility demands, ts and td,
the trial wall and beam dimensions (e.g., wall length, lw, beam length, lb), and/or the
number of coupled wall systems, nw in the structures may need to be revised.
12.5.2 Peak Inter-Story Drift Demands
The peak inter-story drift demands for the prototype structures are estimated based
on the linear elastic deflected shape of each structure under the first mode lateral inertia
force distribution. The inter-story drift demands are determined as the inter-story drifts in
the top story (where the inter-story drift is largest over the structure height) when the roof
drift reaches the estimated peak roof drift demand from Table 12.9, and are calculated as
the average inter-story drifts of the left and right wall piers. The resulting peak inter-story
drift demand estimates are shown in Table 12.10, together with the allowable target
demands used in design. It is observed that both structures satisfy the inter-story drift
demand requirements.
TABLE 12.10
PEAK INTER-STORY DRIFT DEMANDS
Structure
P1-CWUPT
P2-PWUPT
Demand Level
Allowable Demand (%)
Demand Estimate (%)
Survival
2.00
1.92
Design
1.33
0.81
Survival
2.00
1.97
Design
1.33
0.90
12.5.3 Diagonal Tension at Wall Pier Bases
To prevent shear (diagonal tension) failure of the concrete wall piers in the
prototype structures, the diagonal tension capacity, Fw,dt of the wall piers is required to be
greater than the maximum wall pier base shear demand, Qw,dt. Thus,
Fw,dt > Qw,dt
(12.8)
For preliminary design, it assumed that 75% of the total maximum coupled wall
base shear force demand, Qw,max is resisted by the compression-side wall, constituting the
most critical condition for diagonal tension failure. Thus:
Qw, dt = 0.75 Qw, max
(12.9)
As described in Chapter 11, the maximum base shear demand, Qw,max, of a coupled
wall structure is calculated as the sum of a first mode component, Q1,max, and a higher
332
(12.10)
For the preliminary design of the prototype structures, the base shear demand for
the first mode component is estimated as the design base shear demand, Qwd, multiplied
by an over-strength factor of 1.25 as:
Q1, max = 1.25 Qwd
(12.11)
The higher mode component of the maximum base shear demand, Qh,max is
estimated as:
(12.12)
with,
Dm =
mw
= 1
meff
mw
+ 0.7
meff ,2
(12.13)
mw
where, mw, PGAs, meff, and meff,2 are given in Tables 12.1, 12.3, and 12.5. The resulting
maximum coupled wall base shear demands for the prototype structures are listed in
Table 12.11.
TABLE 12.11
DIAGONAL TENSION AT WALL PIER BASES
Structure
Demand Level
Q1,max (kN)
Dm
Qh,max (kN)
Qw,max (kN)
Qw,dt (kN)
Fw,dt (kN)
P1-CWUPT
Survival
5976
0.41
6168
12144
9108
9261
P2-PWUPT
Survival
6116
0.41
6236
12352
9264
9261
Based on ACI 318-02 (ACI Committee 318 2002), the diagonal tension (shear)
capacity of a concrete wall cannot be taken larger than
Fw,dt = 10 Aw f c'
(12.14)
where, Aw=lwtw is the gross cross-sectional area of a wall pier and f c' is the compressive
strength of unconfined concrete.
The largest attainable wall pier shear force capacities, Fw,dt for both prototype
structures, listed in Table 12.11, are greater than or sufficiently close to the corresponding
maximum wall pier shear force demands, Qw,dt. Thus, the preliminary design checks for
the diagonal tension failure of the concrete wall piers are satisfied.
333
12.6 Design of Steel Coupling Beams
This section describes the detailed design of the coupling beams in the prototype
structures. The design includes the following components: (1) design coupling beam
shear force demand; (2) angle and post-tensioning contributions to coupling resistance; (3)
coupling beam cross section; (4) beam post-tensioning; (5) top and seat angles and
connections; (6) coupling beam shear force versus rotation behavior; (7) coupling beam
rotation demands; (8) yielding of beam post-tensioning tendons; (9) fracture of top and
seat angles; (10) shear slip at beam-to-wall interfaces; and (11) wall-contact regions.
The design approach aims to prevent the following failure modes in the structure:
(1) yielding of the beam post-tensioning tendons or failure at the post-tensioning anchors;
(2) local and/or global instability of the coupling beams; (3) fracture of the top and seat
angles; (4) failure of the beam-to-wall, angle-to-wall, and angle-to-beam connections;
and (5) failure of the concrete wall-contact regions.
Note that the discussion below is limited to coupled wall systems with two wall
piers of the same length; however, it may be possible to extend the design
recommendations to other wall configurations. As described previously, the coupling
beams over the height of each structure are assumed to be identical.
12.6.1 Design Coupling Beam Shear Force Demand
The design coupling beam shear force demand, Qbd is determined using the coupled
wall design base shear demand, Qwd (see Table 12.8). A target coupling degree, DOC, is
selected for the structure, from which the design coupling beam shear force demand can
be calculated as:
Qbd =
( DOC ) Qwd H
1.05nb (lw + lb )
(12.15)
where H is the resultant height of the equivalent lateral forces applied on the coupled wall
system (based on the first mode distribution of inertial forces or the equivalent lateral
force distribution from IBC 2000), nb is the number of coupling beams over the height of
the structure (usually equal to the number of stories), lw is the length of a single wall pier,
lb is the coupling beam length, and the factor 1.05 is described in Chapter 8. The design
parameters in Equation (12.15) are shown in Fig. 12.3.
It is assumed that when the coupled wall design base shear demand, Qwd is reached,
the coupled wall system and the coupling beams are at their softening states as described
in Chapters 8 and 10, respectively. Thus, the design approach requires that the coupling
beam shear force capacity at the softening state, Vb,sof be larger than the design coupling
beam shear force demand, Qbd as:
334
Qbd
Qbd
Qbd
Qbd
Qbd
Qbd
Qwd
Qbd
Qbd
Qbd
hw
Qbd
Qbd
Qbd
Qbd
Qbd
Qbd
Qbd
nb = 8
lw
lb
lw
Fig. 12.3 Design coupling beam shear force demand
Vb, sof > Qbd
(12.16)
A moderate level of coupling (e.g., 0.30DOC0.45) is recommended for design,

since over-coupling (e.g., DOC0.60) can cause excessive axial and shear forces in the
wall piers as shown in Chapter 7.
A target coupling degree of DOC=0.35 is selected for the design of the prototype
structures. The resulting design coupling beam shear force demands, Qbd are shown in
Table 12.12.
TABLE 12.12
DESIGN COUPLING BEAM SHEAR FORCE DEMANDS
Structure
DOC
H (m)
Qbd (kN)
P1-CWUPT
0.35
22.6
618
P2-PWUPT
0.35
22.6
654
335
12.6.2 Angle and Post-Tensioning Contributions to Coupling Resistance
Once the design coupling beam shear force demands for the prototype structures
are determined, the next step is the design of the coupling beams to satisfy Equation
(12.16). For design purposes, the total coupling beam shear force demand, Qbd is divided
into two components as:
Qbd = Qbd , a + Qbd , p
(12.17)
where, Qbd,a is resisted by the top and seat angles and Qbd,p is resisted by the beam posttensioning steel.
Then, the coupling beam strength requirement in Equation (12.16) is re-written as:
Vba,sof > Qbd,a
(12.18a)
Vbp,sof > Qbd, p
(12.18b)
and
where, Vba,sof represents the contribution of the top and seat angles to Vb,sof, and Vbp,sof is
the contribution of the beam post-tensioning force to Vb,sof.
Introducing the angle strength ratio, ar as:
ar =
Qbd,a
Qbd, p
(12.19)
Then,
Qbd,a =
Qbd ar
ar + 1
(12.20)
Qbd, p =
Qbd
ar + 1
(12.21)
and
336
The proposed seismic design approach requires that a value for ar be selected for
the structure. If the selected ar value is too large (i.e., the angle contribution is too large),
then the coupling beams may not have enough restoring post-tensioning force to yield the
tension angles back in compression and close the gaps at the beam-to-wall interfaces. If
the ar value is too small (i.e., the angle contribution is too small), then the energy
dissipation of the system may be small.
For example, Figs. 12.4a-c show the beam shear force, Vb versus chord rotation, b
behaviors of three coupling beam subassemblages (similar to the subassemblages
investigated in Chapters 3-5) under monotonic and cyclic loading. The angle behavior is
assumed to be similar to the angle behavior described in Chapter 10. The solid lines on
the right hand side of Fig. 12.4 show the monotonic beam shear force versus chord
rotation behavior of each subassemblage and the dashed lines show the behavior of the
same subassemblages with the angles removed.
The three subassemblages in Fig. 12.4 are designed to have similar coupling shear
resistances; however, the beam post-tensioning steel areas, Abp and the angle tension yield
forces, Tayx are varied (see Table 12.13), resulting in different angle and post-tensioning
contributions (i.e., Vba,sof and Vbp,sof) to the total coupling shear resistance at the softening
state (Vb,sof). The calculation of Vba,sof and Vbp,sof is described later. The initial stress of
the beam post-tensioning tendons in the three subassemblages is not varied in this
investigation.
The angle strength ratios for the subassemblages, calculated as ar=Vba,sof/Vbp,sof,
vary between 0.66 and 1.27 (see Table 12.13). As the ar value is increased, the energy
dissipation of the subassemblage increases (as indicated by the larger hysteresis loops in
Fig. 12.4c as compared with Fig. 12.4a); however, the self-centering capability decreases
(as indicated by the larger residual rotations of the subassemblage upon unloading). Thus,
ar is an important design parameter that influences the energy dissipation and selfcentering capability of the coupling beams. Note, however, that the energy dissipation
and self-centering capability of a multi-story coupled wall structure are affected by the
properties of the concrete wall piers as well.
The ar values used in the design of the prototype structures described in this
chapter, as well as the resulting Qbd,a and Qbd,p values are provided in Table 12.14. A
larger ar value is used for Structure P2-PWUPT in order to provide as much energy
dissipation as possible through the top and seat angles, since the post-tensioned precast
concrete wall piers are expected to provide little energy dissipation but a large amount of
self-centering capability as shown in Chapter 7. Structure P1-CWUPT on the other hand
uses a much smaller ar value to increase the self-centering capability of the entire
coupled wall system through the beam post-tensioning tendons.
337
1500
1200
ar = 0.66
1000
beam shear force,Vb (kN)
1000
500
0
-500
-1000
-5
subassemblage with angles

subassemblage without angles
0
0
(a)
1200
ar = 0.91
1000
1000
500
0
-500
Subassemblage 2
tension angle strength (2Tayx)
beam PT-yielding
tension angle yielding (Tayx)
600
400

200
-1500
-10
-5
10
(b)
1500
beam PT-yielding
800
-1000
1200
ar = 1.27
1000
beam shear force,V (kN)
1000
500
beam PT-yielding
400
10
1500
600
200
beam PT-yielding
800
Subassemblage 1
-1500
-10
0
-500
Subassemblage 3
-1000
800
600
-5
beam PT-yielding
400
200
-1500
-10
beam PT-yielding
10
(c)

0
Fig. 12.4 Angle strength ratio, ar: (a) ar=0.66; (b) ar=0.91; (c) ar=1.27
TABLE 12.13
EFFECT OF ar ON COUPLING BEAM BEHAVIOR
Subassemblage Abp (mm2) Tayx (kN)
1
2
3
1680
1400
1120
273
367
440
Vb,sof (kN)
Vba,sof (kN)
Vbp,sof (kN)
ar
774
747
716
307
356
400
467
391
316
0.66
0.91
1.27
338
TABLE 12.14
COMPONENTS OF COUPLING BEAM DESIGN FORCES
Structure
ar
Qbd,a (kN) Qbd,p (kN)
P1-CWUPT 0.051
30
588
P2-PWUPT 1.22
360
294
12.6.3 Coupling Beam Cross Section
The coupling beam cross section is determined based on Fig. 12.5a, which shows a
preliminary free body diagram of the beam at the softening state, excluding the top and
seat angle forces. It is assumed that: (1) flange cover plates are not used; (2) the neutral
axis depth at the beam-to-wall interfaces is equal to the beam flange thickness, tbf; (3) the
contact stresses at the beam ends have linear (i.e., triangular) distributions; and (4) the
maximum beam contact stress is equal to the beam steel yield strength, fby.
fby
Vbp,sof
tbf
linear stress
distribution
db
coupling beam
Vbp,sof=Qbd,p
db-(2/3)tbf
lb
(a)
Pbi
fby
Vbp,sof
cb,sof
linear stress
distribution
db-(2/3)cb,sof
coupling beam
Vbp,sof=Qbd,p
Pbi
(b)
Fig. 12.5 Post-tensioning contribution to coupling resistance: (a) preliminary; (b) revised
Then, the coupling shear resistance due to the post-tensioning force, Vbp,sof can be
written as:
Vbp , sof =
0.5 bbf tbf fby

2t
db bf
lb
3
339
(12.22)
For design, Equation (12.22) is used to select a beam cross-section with depth, db,
flange width, bbf, and flange thickness, tbf, that satisfy the post-tensioning component of
the design coupling beam shear force demand, Qbd,p as required in Equation (12.18b). The
W21147 cross section selected for the beams in the prototype structures satisfies this
requirement.
In general, any cross section with compact web and flanges can be used for the
coupling beams. Even though this report focuses on beams with I-sections, other types of
beams such as with hollow sections may also be used. The beam flange width, bbf should
not be larger than the wall thickness, tw so that the entire flange is in contact with the wall.
12.6.4 Beam Post-Tensioning
The next step is the design of the beam post-tensioning tendons using the revised
softening-state free body diagram in Fig. 12.5b. The total required initial beam posttensioning force, Pbi is determined as:
Pbi =
Qbd , p lb
2
db cb, sof
3
(12.23)
where, cb,sof is the depth of the compression (i.e., contact) region at the beam end. The
following assumptions are made in the expression for Pbi: (1) the beam post-tensioning
force at the beam softening state is equal to the initial post-tensioning force, Pbi; (2) the
axial force in the beam is equal to the post-tensioning force; (3) the contact stresses at the
beam ends have linear (i.e., triangular) distributions; and (4) the contact region is outside
the web. A non-rectangular compression region should be used if the contact region
extends into the web.
The depth of the contact region at the beam softening state, cb,sof, depends on the
total compression force in the beam, which is assumed to be equal to Pbi. The effect of the
top and seat angles is ignored and it is assumed that the largest compressive stress in the
beam is equal to the yield strength of the steel, fby. Then, the contact depth can be
estimated as:
cb, sof =
2 Pbi
fby bbf
(12.24)
Equations (12.23) and (12.24) are solved iteratively to determine Pbi and cb,sof. An
initial value of cb,sof=tbf [similar to Equation (12.22)] can be used for this purpose, with
convergence achieved in 5 to 10 iterations. Once Pbi is known, the total required beam
post-tensioning steel area, Abp, can be determined as:
340
Abp = abp =
Pbi
fbpi
(12.25)
where, fbpi is the design initial stress in the post-tensioning steel. It is recommended that
multi-strand tendons are used with fbpi not more than 0.6fbpu, where fbpu is the design
maximum strength of the strands.
The total number of beam post-tensioning strands and tendons can be determined
from:
nbp nbt =
Pbi
abp fbpi
(12.26)
where, nbt is the number of post-tensioning tendons, nbp is the number of strands per
tendon, and abp is the area of a single strand. In the design of the prototype structures,
15 mm (0.6 in.) diameter post-tensioning strands with abp=140 mm2 (0.217 in2) and
fbpi=0.6fbpu are used. The placement and distribution of the post-tensioning strands and
tendons is depicted in Fig. 12.6 and the design is summarized in Table 12.15. Two eightstrand tendons are used in Structure P1-CWUPT, while only one eight-strand tendon is
necessary in Structure P2-PWUPT as a result of the larger ar value used as described
previously.
Note that the elongations in the beam post-tensioning strands due to the lateral
displacements of the structure are the same regardless of the distance of the tendons from
the beam centerline (see Chapter 5). Thus, the strands can be placed symmetrically
anywhere between the beam flanges on both sides of the web. The strands should be
placed inside ungrouted post-tensioning ducts to achieve unbonded conditions, with a
short length near the anchors grouted to prevent premature fracture of the strand wires at
the anchors as observed in the subassemblage experiments described in Chapter 9. The
subassemblage experiments did not show any advantage of using oversized posttensioning ducts significantly larger than necessary for construction fit and tolerances,
thus, this is not considered to be necessary.
Based on the experiments described in Chapter 9, two upper limits to the total beam
post-tensioning steel area and force are recommended to prevent excessive local
distortions and possible instability of the beam as follows. First, the total initial beam
post-tensioning force, Pbi should be limited as:
Pbi < 0.25Abc fby
(12.27)
where, Abc is the gross cross section area of the beam including the flange cover plates (if
any). This requirement also ensures that gap opening (i.e., decompression) at the beam
ends occurs before the yielding of the beam in compression. It can be shown that gap
341
opening and beam compression yielding would occur simultaneously if Pbi=0.5Abcfby

(ignoring the effect of the angles). It is assumed that the cover plate steel (if used) has the
same design strength as the beam steel.
wall
wall
angle
angle
t a = 22
t a = 22
W21x147
t bf = 29
W21x147
t bf = 29
tbw= 18
tbw= 18
76
15 strand
total 16
15 strands
76
d b=560
d b=560
total 8
15 strands
76
A bp =1120 mm2
A bp = 2240 mm 2
l a =203
b bf =318
la=bbf=318
units: mm
t w=356
t w =356
beam end view
beam end view
(a)
(b)
Fig. 12.6 Coupling beam and post-tensioning details:

(a) Structure P1-CWUPT; (b) Structure P2-PWUPT
TABLE 12.15
BEAM POST-TENSIONING
Abp
Pby
Pbi
(mm2) (kN) (kN)
2240 2500 3784
Pbu
(kN)
4170
Abcfby
(kN)
9608
P2-PWUPT
1
8
1120 1251 1892 2085
*
Note: cb,sof for Wall P1-CWUPT extends into beam web.
9608
Structure
nbt
nbp
P1-CWUPT
342
Abfcfbm cb,sof
(kN) (mm)
4159 *288
4159
29
Vbp,sof
(kN)
623
Qbd,p
(kN)
588
316
294
Second, it is recommended that

Pby < Abfc fbm
(12.28)
Pby = abp fbpy
(12.29)
where,
is the total yield strength of the beam post-tensioning steel, Abfc is the cross section area
of one beam flange plus cover plate (if any), and fbm is the beam steel maximum (i.e.,
peak) strength. In the calculation of Pby, it is assumed that all of the beam post-tensioning
strands yield simultaneously since they are prestressed by the same amount.
The coupling beams in both prototype structures satisfy Equations (12.27) and
(12.28) as shown in Table 12.15. Note that for Structure P1-CWUPT, Pbi is a little larger
than 0.25Abcfby; however, the difference is smaller than 5 percent and is acceptable. If
any of these two limits is not satisfied, a coupling beam with a larger cross section may
be selected or the total post-tensioning steel area may be reduced. Alternatively, cover
(i.e., reinforcing) plates may be used to strengthen the beam flanges (note that flange
cover plates are not needed in the prototype structures). The cover plates (if needed)
should be sufficiently long to prevent beam flange yielding where the plates are
terminated and should extend over a length equal to, at least, the total beam depth, dbc=
db+2tc including the cover plates. The welds between the cover plates and the beam
flanges should be designed for a force equal to
1.25Cb, pty
for the 2 nd and 3rd floor levels
Cb, pty
for the upper floor/roof levels
(12.30)
where, the calculation of Cb,pty is described later and the 1.25 factor accounts for the
additional axial forces that develop in the lower floor coupling beams in a structure due
to the lateral stiffness of the wall piers near the foundations as described in Chapter 7.
The post-tensioning anchors should be designed for the maximum tendon forces,
Pbu based on the design maximum strength of the post-tensioning steel, fbpu as:
Pbu = abp fbpu
(12.31)
Note that Equation (12.22) provides an approximation for the coupling shear
resistance due to the post-tensioning force, Vbp,sof. Once the design of the beam posttensioning tendons is finalized, a revised value for Vbp,sof can be determined as:
343
Vbp,sof =
2c
Pbi
dbc b,sof
lb
3
(12.32)
where cb,sof is from Equation (12.24). No cover plates are used in the coupling beams of
the prototype structures, and thus, dbc=db (see Fig. 12.5b). As shown in Table 12.15, the
design requirement given by Equation (12.18b) is satisfied for both prototype structures.
12.6.5 Top and Seat Angles and Connections
The top and seat angles are designed based on Fig. 12.7, which shows a free body
diagram of the coupling beam at the softening state, excluding the post-tensioning forces.
The angle forces are assumed to be acting parallel to the beam. The sum of the tension
and compression angle forces (Tax and Cax, respectively) to satisfy the angle component
of the design coupling beam shear force demand, Qbd,a can be determined as:
Cax
cover plate (if any)
tc
ta
Tax
coupling beam
dbc+t a Vba,sof
Vba,sof=Qbd,a
Tax
Cax
Fig. 12.7 Angle contribution to coupling resistance
Tax + Cax =
Qbd,alb
dbc + ta
(12.33)
where, ta is the angle leg thickness. Assuming Tax=1.5Tayx (see assumed angle behavior
for the coupling beam softening state in Chapter 10) and Cax=Tayx, and ignoring ta, the
required angle tension yield strength is
Tayx =
2 Qbd,alb
5 dbc
(12.34)
The angle thickness, ta and gage length of the angle-to-wall connections, lgv are
determined to achieve the required angle tension yield strength Tayx as described in
Chapter 3 following Kishi and Chen (1990) and Lorenz et al. (1993). The final design
details for the top and seat angles and connections used in the prototype structures are
shown in Fig. 12.8. Angles with L881/2 and L887/8 cross sections are used in
Structures P1-CWUPT and P2-PWUPT, respectively. The angle length is taken as 203
mm (8 in.) in Structure P1-CWUPT. Longer angles with length equal to the beam flange
344
width, bbf=318 mm (12.5 in.) are used in Structure P2-PWUPT. Note that it may also be
possible to use thinner angles with 318 mm length in Structure P1-CWUPT; however,
this was not investigated.
12.7 76.2
(0.5) (3)
76.2 38.1
(3) (1.5)
l a=203(8)
25.4
(1)
l g1=123.8(4.875)
50.8
(2)
50.8
(2)
50.8
(2)
PT strand anchor
50.8
(2)
44
(1.75)
15(0.6)strand
lgv =152
(6)
51
(2.0)
25.4
(1)
102
(4)
203
(8)
139.7
(5.5)
51
(2.0)
A49025.4(1)bolt
ka
12.7
(0.5)
t a =12.7 (0.5)
k a=28.6(1.125)
unit: mm (inch)
lgh =127(5)
(a)
PT strand anchor
l gv =108
(4.25)
44
(1.75)
A49025.4(1)bolt
ka
k a=38.1(1.5)
lgh =127(5)
66.7
76.2 38.1
(2.625) (3) (1.5)
31.8
(1.25)
63.5
(2.5)
la=318(12.5)
31.8 63.5 63.5 63.5 63.5 31.8
(1.25) (2.5) (2.5) (2.5) (2.5) (1.25)
15(0.6)strand
lg1=74.6(2.94)
22.2
(0.875)
63.5
(2.5)
34.9
(1.375)
318
60.3 (12.5)
(2.375)
63.5
(2.5)
85.7
(3.375)
63.5
(2.5)
22.2
(0.875)
31.8
(1.25)
t a =12.7 (0.5)
unit: mm (inch)
(b)
side view
(cross-section)
beam end view

(elevation)
top view
(plan)
Fig. 12.8 Top and seat angle and connection details:

The resulting Tayx and Casx values, as well as the coupling beam shear force
resistance due to the top and seat angles, Vba,sof are given in Table 12.16. The Vba,sof
values are determined from
Vba,sof =
(Tax + Cax )(dbc + ta )

lb
345
(12.35)
with Tax=1.5Tayx and Cax= (1/40)fayAaCasx, where Casx is the slip capacity of the angleto-beam connection bolts, fay is the angle steel yield strength, and Aa is the gross area of
the angle leg cross section. It is observed that the prototype structures satisfy the design
requirement given by Equation (12.18a).
TABLE 12.16
TOP AND SEAT ANGLES
Structure
Tayx (kN) Casx (kN) Vba,sof (kN) Qbd,a (kN)
P1-CWUPT
53
425
44
30
P2-PWUPT
516
1063
432
360
In order to prevent slip of the angle-to-beam connection bolts as the angles are
pulled away from the walls, it is recommended that the slip capacity of the angle-to-beam
connection bolts, Casx be greater than the maximum anticipated angle tension force, 2Tayx
as:
Casx > 2 Tayx
(12.36)
where, the factor 2 accounts for the over-strength observed in the angles from the
subassemblage experiments, as discussed in Chapter 9. Table 12.16 shows that the angles
in the prototype structures satisfy this requirement. The angle-to-wall connections should
also be designed for the maximum anticipated angle tension force, 2Tayx, including
prying effects. As shown in Fig. 12.8, unbonded post-tensioning strands are used in the
angle-to-wall connections of the prototype structures, similar to the subassemblage test
specimens in Chapter 9. More information on the design of the angle-to-beam and angleto-wall connections is beyond the scope of this report.
12.6.6 Coupling Beam Shear Force versus Rotation Behavior
The coupling beam shear force versus chord rotation (Vb-b) behavior can be
estimated from the structure design properties determined above. As shown by the solid
lines in Fig. 12.9, the revised procedure in Chapter 10 is used to develop an idealized bilinear Vb-b relationship for each prototype structure based on the beam softening state
(at Vb,sof, b,sof) and the PT-yielding state (at Vb,pty, b,pty). Key values calculated for the
bi-linear coupling beam shear force versus chord rotation relationships are provided in
Table 12.17. Note that the coefficient of 1.075 for b,pty in Equation (10.13) is ignored
conservatively. It is observed that the coupling beam shear force capacities at the
softening state satisfy the design coupling beam shear force demands from Table 12.12,
and thus the design requirement given in Equation (12.16) is met.
346
1000
900
beam
800 softening
700
beam PT-yielding
(Vb,pty, b,pty)
(Vb,sof , b,sof)
600
500
decompression
400
300
P1-CWUPT
200
bi-linear estimation
DRAIN-2DX fiber element model
100
0
(a)
1000
900 beam
beam PT-yielding
softening
800
(Vb,pty, b,pty)
(Vb,sof , b,sof)
700

600
500
400
300
P2-PWUPT
decompression
200
100
0
0
(b)
Fig. 12.9 Coupling beam shear force versus chord rotation (Vb-b) relationships:
347
TABLE 12.17
COUPLING BEAM SHEAR FORCE VERSUS ROTATION RELATIONSHIPS
Beam Softening State
Structure
P1-CWUPT
Cb,sof
(kN)
2502
cb,sof
(mm)
*261
Vb,sof
(kN)
689
Beam PT-Yielding State
b,sof
(%)
0.17
Cb,pty
(kN)
3465
cb,pty
(mm)
*30
P2-PWUPT
1251
26
738
0.19
1861
19
Note: *cb,sof for Walls P1-CWUPT and P2-PWUPT extend into beam web.
Vb,pty
(kN)
770
bg,pty
(%)
6.99
be,pty
(%)
0.25
b,pty
(%)
7.24
947
6.63
0.27
6.90
For comparison, the dashed lines in Fig. 12.9 show the smooth Vb-b behaviors for
Structures P1-CWUPT and P2-PWUPT using the fiber element subassemblage analytical
model from Chapter 3. In general, the idealized bi-linear relationships capture the
structure capacities from the fiber element models reasonably well.
12.6.7 Coupling Beam Rotation Demands
The survival-level and design-level coupling beam chord rotation demands, s and
d, respectively, are estimated based on the idealized coupled wall displaced shape in Fig.
12.10 as:
tension
side wall
coupling
beam
chord
(l w /lb)
compression
side wall
lw
lb
lw
Fig. 12.10 Coupled wall idealized displaced shape
348
s =
s (lw + lb )
lb
(12.37)
d =
d (lw + lb )
lb
(12.38)
where, s and d are the survival-level and design-level coupled wall roof drift demands,
respectively. The above expressions for s and d assume, conservatively, that the
rotations of the tension-side and compression-side wall piers occur about the compression
corners of the wall piers at the base.
The estimated s and d demands corresponding to s and d demands in Table 12.9
are given in Table 12.18.
TABLE 12.18
COUPLING BEAM ROTATION DEMANDS
Structure
P1-CWUPT
P2-PWUPT
Demand Level
s and d (%)
Survival
5.49
Design
2.33
Survival
5.62
Design
2.59
12.6.8 Yielding of Beam Post-Tensioning Tendons
As discussed in Chapter 11, the coupling beam chord rotation capacity at the beam
PT-yielding state is required to be larger than the survival-level beam chord
rotation demand (i.e., yielding of the beam post-tensioning tendons is not allowed to
occur under the survival seismic demand level Design Objective 1). Thus,
b, pty > s
(12.39)
where, b,pty is the coupling beam chord rotation capacity at the PT-yielding state.
In addition, it is required that 2/3 times the coupling beam chord rotation capacity
at the beam PT-yielding state be larger than the design-level beam chord rotation demand
(Design Objective 2). Thus,
2
b, pty > d
3
(12.40)
The b,pty capacities for the coupling beams in the prototype structures, provided in
Table 12.17, satisfy the rotation demands in Table 12.18. The initial stress in the beam
349
post-tensioning strands, fbpi, may be reduced if the design requirements in Equations

(12.39) and (12.40) are not satisfied.
12.6.9 Fracture of Top and Seat Angles
Fracture of the top and seat angles prior to the survival demand level should be
prevented. As shown in Fig. 12.11, the maximum deformation of the tension angles, uas
(i.e., the separation of the angle heel from the wall) under the survival-level beam
rotation demand, s can be estimated as:
uas = s (dbc + 0.5ta cb, pty )
top
angle
ta
(12.41)
tc
cb,pty
PT-tendon
dbc
s
coupling
beam
u as
seat
angle
lw
concrete wall
Fig. 12.11 Tension angle deformation
It is assumed that the depth of the compression (i.e., contact) region at the beam-towall interface is equal to cb,pty (see Table 12.17). The resulting survival-level tension
angle deformation demands, uas for the coupling beams in the prototype structures are
shown in Table 12.19.
For comparison with the analysis results in Chapter 13, Table 12.19 also includes
the estimated design-level tension angle deformation demands, uad for the prototype
structures. These uad values were obtained by replacing s with the design-level beam
rotation demand d in Equation (12.41) and using cb,pty as the contact depth at the beam
ends under the design demand level.
350
TABLE 12.19
TENSION ANGLE DEFORMATION DEMANDS
Structure
uas (mm)
uad (mm)
P1-CWUPT
30
13
P2-PWUPT
31
14
More research needs to be conducted on the deformation capacity of the top and
seat angles to ensure that the survival level deformation demand, uas can be reached
without low cycle fracture of the angle legs. In the absence of a reliable method to
estimate the deformation capacity of the angles, it is required that
uas < 38 mm (1.5 in.)
(12.42)
This requirement is satisfied for both structures as shown in Table 12.19. If the
deformation demands of the tension angles are larger than allowable, then a coupling
beam with smaller depth may need to be selected.
12.6.10 Shear Slip at Beam-to-Wall Interfaces
The beam-to-wall connections should be designed to transfer the maximum

coupling beam shear forces without slip at the beam ends. The design requirement is
stated as:
Vb, ss > Qb, max
(12.43)
where, Qb,max is the maximum shear slip demand as:
Qb,max = 1.25 Vb, pty for the 2nd and 3rd floor level
Qb,max = Vb, pty
for the upper floor / roof levels
(12.44)
with Vb,pty from Table 12.17 and the 1.25 factor accounting for the increase in the lower
floor coupling beam shear forces due to the lateral stiffness of the wall piers near the
foundations as described in Chapter 7.
The shear slip capacity at the beam-to-wall interfaces, Vb,ss is equal to
Vb, ss = V p, ss + Va, ss
(12.45)
where, Vp,ss is the slip capacity due to the coupling beam end axial force and Va,ss is the
slip capacity from the top and seat angle-to-wall connections.
The slip capacity due to the beam end axial force is assumed to be equal to
351
V p, ss = bw Pbi
(12.46)
where, bw is the coefficient of friction between the beam and the wall shim plates and Pbi
is the total initial beam post-tensioning force (assuming that the beam end axial force
when Qb,max is reached is equal to the initial post-tensioning force). A value of bw=0.33
for steel against steel friction surfaces can be used for design (AISC 2001).
The calculation of Va,ss depends on the type of the angle-to-wall connections. For
unbonded post-tensioned angle-to-wall connections (as shown in Fig. 12.8), it is
recommended that
Va,ss = 2aw f awpi Aaw
(12.47)
where, aw=bw=0.33 is the coefficient of friction between the angle vertical leg and the
wall shim plates; fawpi is the initial stress of the post-tensioning strands used in the angleto-wall connections; and Aaw is the total area of the post-tensioning steel used in an
angle-to-wall connection (for one angle). A value of fawpi=0.5fawpu is used in the design of
the prototype structures, where fawpu=1862 MPa (270 ksi) is the maximum strength of the
angle-to-wall connection strands.
The estimated shear slip capacities and demands at the beam-to-wall interfaces of
the prototype structures are given in Table 12.20. Both structures satisfy the design
requirement in Equation (12.43).
TABLE 12.20
SHEAR SLIP AT BEAM-TO-WALL INTERFACES
Qb,max (kN)
Structure
aw and bw
Aaw
(mm2)
fawpi
(MPa)
Va,ss
(kN)
Pbi
(kN)
Vp,ss
(kN)
Vb,ss
(kN)
P1-CWUPT
0.33
560
931
344
2500
825
1169
963
770
P2-PWUPT
0.33
1400
931
860
1251
413
1273
1184
947
2nd and 3rd upper
12.6.11 Wall-Contact Regions
The wall-contact regions adjacent to the coupling beams should be designed for a
compressive force equal to
Cwc = 1.25Cb, pty
for the 2 nd and 3rd floor levels
Cwc = Cb, pty
for the upper floor/roof levels
(12.48)
where, Cb,pty from Table 12.17 is the compression force in the beam at the beam PTyielding state and the 1.25 factor accounts for the additional axial forces that develop in
352
the lower floor coupling beams in a structure due to the lateral stiffness of the wall piers
near the foundations as described in Chapter 7.
Shim plates should be used at the beam-to-wall interfaces, with the total thickness
of the shim plate, tsh and wall embedded plate, te equal to at least the thickness of the
beam flange, tbf. The shim plates should be terminated at the coupling beam web, thus
preventing contact between the web and the walls as described in Chapter 9.
Concrete confinement should be provided in the wall-contact regions near the
coupling beam flanges. Assuming a 3:4 spread for the contact stresses as shown in Fig.
12.12 and assuming that the contact depth at the beam-to-shim-plate interface is equal to
cb,pty, the nominal concrete compressive stress behind the embedded plate can be
estimated as:
wc =
(cb, pty
Cwc
+ 1.5te + 1.5t sh )(bbf + 1.5te + 1.5tsh )
(12.49)
beam flange
te tsh
0.75(te+tsh)
shim plate
cb,pty
tbf
Cwc
wc
3:4 spread for
contact stresses
beam web
embedded plate
wall concrete
(a)
0.75(te+tsh)
area over which wc is

assumed to act
0.75(te+tsh)
bbf
cb,pty
beam flange
contact region
(b)
Fig. 12.12 Wall-contact region stresses: (a) side view; (b) beam end view
353
Since the cb,pty values in Table 12.17 are calculated based on an isolated coupled
wall subassemblage, they should be multiplied by 1.25 for the design of the 2nd and 3rd
floor level wall-contact regions using Equation (12.49). Further, Equation (12.49) should
be modified if cb,pty>tbf+tc and the beam compression region is not rectangular. In order to
simplify the calculation in that case, it would be a reasonable approximation to replace
cb,pty with tbf+tc, provided that the beam web thickness tbw is much smaller than the flange
width bbf. Note also that the term (bbf+1.5te+1.5tsh) in Equation (12.49) cannot be taken
greater than the wall thickness, tw.
The maximum compressive strength of the confined concrete in the wall-contact
regions, fccw should be larger than wc to limit the deformations in the wall-contact
regions. Thus,
f ccw > wc
(12.50)
The final design details for the wall-contact regions in the prototype structures,
including the shim plate, embedded plate, and concrete confinement reinforcement are
given in Table 12.21 and Fig. 12.13. Two steel spirals are used to confine the concrete in
the wall-contact regions at each beam end, one spiral near each beam flange. The stressstrain behaviors of the confined concrete in the prototype structures, as estimated using
Mander et al. (1988), are shown in Fig. 12.13. The spiral reinforcement properties are
kept constant at all of the floor and roof levels in each structure.
As shown in Table 12.21, the resulting confined concrete strengths are adequate to
satisfy the design requirement in Equation (12.50) for both prototype structures.
TABLE 12.21
WALL-CONTACT REGIONS
Cwc (kN)
te
tsh
2nd and
(mm)
(mm)
upper
3rd
P1-CWUPT 4730
3784
38
38
P2-PWUPT 2365
1892
16
16
Structure
cb,pty (mm)
2nd and
3rd
92.8
23.7
upper
30.1
19.0
wc (MPa)
2nd and
3rd
92.4
82.1
fccw (MPa)
2nd and
upper
3rd
74.4
98.7
63.7
90.4
upper
98.7
90.4
12.7 Design of Wall Piers in Structure P1-CWUPT
This section describes the design of the cast-in-place reinforced concrete wall piers
in Structure P1-CWUPT. The design includes the following components: (1) wall pier
flexural reinforcement; (2) concrete confinement at wall pier bases; (3) diagonal tension
at wall pier bases; and (4) shear slip at coupled wall base. Only the cross-sections of the
wall piers at the base are designed. The design and curtailment of the reinforcement over
the height of the structures are not addressed in this report.
354
confined concrete
dbsp
=13 mm
sbsp=25 mm
4absp
=6.83 %
bsp=
Dbspsbsp
Dbsp=292 mm
wire
W20
spiral wire
absp=127mm2
Structure P1-CWUPT
confined concrete
sbsp=32 mm
4absp
bsp=
=5.46 %
Dbspsbsp
Dbsp=292 mm
wire
W20
dbsp
=13 mm
absp=127mm2
spiral wire
Structure P2-PWUPT
100
100
(0.016, 98.7)
(0.014, 90.4)
80
stress (MPa)
stress (MPa)
80
60
40
60
40
smooth relationship
idealized multi-linear relationship
20
smooth relationship
20
Structure P1-CWUPT
Structure P2-PWUPT
0
0
0.01
0.02
strain
0.03
0.04
0.01
0.02
strain
0.03
0.04
Fig. 12.13 Wall-contact region concrete confinement

12.7.1 Wall Pier Flexural Reinforcement
The design of the wall pier flexural reinforcement to satisfy the design coupled wall
base shear force demand, Qwd is based on the coupled wall softening state described in
Chapters 7 and 8. The design criterion is given as:
Fws > Qwd

where, Fws is the coupled wall base shear force capacity at the wall softening state.
355
(12.51)
The following assumptions are made for the design of the wall flexural
reinforcement, similar to the assumptions made in Chapter 8 for the estimation of the
coupled wall softening state: (1) the coupling beam shear forces are equal to Vb,sof; (2) the
strain distributions at the bases of the wall piers are linear; (3) the concrete compressive
stresses at the bases of the wall piers have linear distributions; (4) the neutral axis depth
at the base of the compression-side wall (i.e., right-side wall for lateral loads applied
from left to right) is located at the centerline of the wall (i.e., ccws=0.5lw); and (5) the
strain and stress in the extreme flexural reinforcement of the tension-side wall are equal
to the yield strain wsy and yield stress fwsy, respectively.
An iterative design procedure is followed as described below.
Step 1
Select trial details for the amount and layout of the flexural reinforcement in the
wall pier cross sections. A sample trial section (Trial 1) for Structure P1-CWUPT is
given in Fig. 12.14a. Note that the cross section of only one of the wall piers is shown
and the details of the other wall pier are identical. It is assumed that 1415.9mm
reinforcing bars at 457 mm spacing are used within the central region of each wall pier
and eleven layers of reinforcing bars (each layer with a total steel area of 1290 mm2) at
76 mm spacing are used within the end regions.
Step 2
Ntws = G 1.05 Vb, sof
(12.52)
N cws = G + 1.05 Vb, sof
(12.53)
where, G represents the factored gravity loads applied at the floor and roof levels of a
wall pier, Vb,sof is from Table 12.17, and the 1.05 factor is described in Chapter 8. It is
assumed that the coupled wall gravity loads (see Table 12.2 for the unfactored gravity
loads) are distributed equally between the tension-side and compression-side wall piers.
Step 3
Determine the bending moment resistances of the tension-side and compressionside walls, Mtws and Mcws respectively, and the total base moment and base shear force
resistances of the coupled wall system, Mws and Fws respectively, using the procedure
described in Chapter 8.
Step 4
Compare Fws with Qwd to check the adequacy of the trial cross section.
356
(14#5@18in.)
1415.9mm@457mm
1290mm2@76mm
1290mm2@76 mm
tw=356mm
(14in.)
l w=4.88m (16ft)
(a)
(14#5@18in.)
3226mm2@127mm 1415.9mm@457mm
3226mm2@76 mm
tw=356mm
(14in.)
l w=4.88m (16ft)
base shear at softening state, Fws (kN)
(b)
6000
2
3
5000
Fws = Qwd
4000
3226 mm2
2581 mm2
1935 mm2
1290 mm2
3000
2013 mm2
2000
trial sections 1-4
final design point
1000
0
500
1000
1500
required steel area

2000
2500
3000
3500
total trial steel area (mm2)

(c)
(22#11@3in.)
2234.9mm@76mm
(14#5@18in.)
1415.9mm@457mm
(22#11@3in.)
2234.9mm@76mm
tw=356mm
(14in)
l w=4.88m (16ft)
(d)
Fig. 12.14 Wall pier flexural reinforcement: (a) trial section 1; (b) trial section 2;
(c) flexural steel area versus base shear resistance; (d) final wall cross section
357
Step 5
If necessary (as determined from the comparison in Step 4), repeat Steps 2-4 for
other trial sections (e.g., see Fig. 12.14b for Trial 2).
Step 6
Plot the coupled wall base shear force capacity at the wall softening state, Fws
determined for the trial cross sections against the corresponding reinforcing steel areas
used in the end regions of each section as shown in Fig. 12.14c. Use interpolation to
determine the required area of wall flexural reinforcement to satisfy the coupled wall
design base shear force demand Qwd and select the reinforcement.
The resulting wall pier cross section for Structure P1-CWUPT is shown in Fig.
12.14d with the design details summarized in Table 12.22. A positive wall axial force
represents a compressive force. A total of 4434.9 mm reinforcing bars and 1415.9
mm bars are used in each wall pier, with the total pier flexural reinforcement area, Aws
given in Table 12.22. It is observed that the coupled wall base shear capacity at the wall
softening state Fws is smaller than but sufficiently close to the design base shear demand,
Qwd, and thus, the design is considered acceptable for Equation (12.51).
TABLE 12.22
WALL PIER FLEXURAL REINFORCEMENT
Aws
Mcws Mwbs
G Vb,sof Ntws Ncws ctws
Mtws
Mws Fws Qwd
(mm2) (kN) (kN) (kN) (kN) (mm) (kN.m) (kN.m) (kN.m) (kN.m) (kN) (kN)
P1-CWUPT 47084 2798 5515 -2993 8589 1270 29241 32099 40607 101946 4603 4782
Structure
12.7.2 Concrete Confinement at Wall Pier Bases
As discussed in Chapter 11, the coupled wall roof drift capacity at the wall ultimate
state is required to be larger than the survival-level coupled wall roof drift demand (i.e.,
crushing of the confined concrete at the bases of the wall piers is not allowed to occur
under the survival seismic demand level Design Objective 1). Thus,
wu > s
(12.54)
In addition, it is required that 2/3 times the coupled wall roof drift capacity at the
wall ultimate state be larger than the design-level coupled wall roof drift demand (Design
Objective 2). Thus,
2
wu > d
3
(12.55)
where, wu is the coupled wall roof drift capacity at the wall ultimate state and s and d
are the survival-level and design-level roof drift demands, respectively.
358
Note that since d is smaller than (2/3)s for the prototype structures, the design
requirement in Equation (12.55) is automatically satisfied when Equation (12.54) is
satisfied.
The following assumptions are made in the design of the confinement
reinforcement at the bases of the wall piers, similar to the assumptions made in Chapter 8
for the estimation of the coupled wall ultimate state: (1) the maximum compression strain
in the confined concrete at the base of the compression-side wall is equal to the ultimate
(i.e., crushing) strain of the confined concrete, ccu; (2) the strain distributions at the bases
of the wall piers are linear; (3) the height of the plastic hinge, hwp at the base of the
compression-side wall is equal to the confined thickness of the wall, twc (estimated as the
wall thickness minus the unconfined cover concrete thicknesses); (4) the coupling degree
at the wall ultimate state, DOCwu is equal to the design coupling degree, DOC; and (5)
the neutral axis depth at the base of the compression-side wall at the wall ultimate state is
equal to ccwu=0.2lw.
The design procedure is described below.
Step 1
Estimate the curvature at the base of the compression-side wall when the survivallevel roof drift demand, s is reached as:
cwu =
(DOC)wu s (DOC)s
=
1.85hwp
1.85twc
(12.56)
where hwp=twc and DOCwu=DOC based on Assumptions (3) and (4), respectively. This
equation follows from the estimation of the coupled wall ultimate state in Chapter 8.
Step 2
Estimate the required ultimate strain for the confined concrete at the base of the
ccu = ccwucwu
(12.57)
where ccwu=0.2lw based on Assumption (5).

Step 3
Design the concrete confinement to achieve ccu. A concrete confinement model,
such as the model developed by Mander et al. (1988), can be used for this purpose.
As shown in Fig. 12.15, spiral reinforcement is used to confine the concrete at the
bases of the wall piers in Structure P1-CWUPT. The stress-strain behavior of the
confined concrete, as estimated using Mander et al. (1988), is also shown in Fig. 12.15. It
359
is recommended that the confinement reinforcement be provided over a height equal to at

least the first story height. The concrete confinement should be used over a wall length,
lwc where the concrete compression strains exceed the assumed unconfined concrete
crushing strain, cu as:
lwc = ccwu (1
cu
)
ccu
(12.58)
confined concrete
Dwsp=299 mm
dwsp
= 6 mm
swsp=38 mm
wire W4.5
wsp=
4awsp
Dwspswsp
spiral wire
awsp=29mm2
=1.02%
60
(0.00502, 53.9)
stress (MPa)
50
(0.0128, 44.8)
40
30
20
smooth relationship
10
0
0
0.005
0.01
0.015
strain
Fig. 12.15 Concrete confinement at wall pier bases
Table 12.23 shows the wall pier confined concrete ultimate strain required by
Equation (12.57) as well as the strain capacity for the confined concrete. Since the strain
capacity is larger than the required strain, crushing of the confined concrete under the
survival demand level is prevented.
360
TABLE 12.23
CONCRETE CONFINEMENT AT WALL PIER BASES
Structure
ccu required ccu provided lwc (mm)
P1-CWUPT
0.0090
0.012
732
Fw,dt > Qw,dt
(12.59)
Because of the compression forces that develop in the compression-side wall pier
due to coupling, a larger portion of the total coupled wall base shear force is carried by
the compression-side wall. It is assumed that the portion of the total maximum coupled
wall base shear demand, Qw,max carried by the compression-side wall can be estimated as:
Qw, dt =
M cws + M wbs
Qw, max
M tws + M cws + M wbs
(12.60)
where, Qw,dt is the maximum wall pier base shear demand, Mtws and Mcws are the base
moment resistances of the tension-side and compression-side wall piers at the softening
state (see Table 12.22), and Mwbs is the contribution of the coupling beam shear forces to
the coupled wall base shear resistance at the softening state.
(12.61)
The base shear demand for the first mode component is estimated as the coupled
wall base shear capacity at the wall ultimate state, Fwu as:
Q1,max = Fwu
(12.62)
The estimation of Fwu is described in Chapter 8. The higher mode component of the
maximum base shear demand, Qh,max is estimated as:
with,
361
(12.63)
Dm =
mw
= 1
meff
mw
+ 0.7
meff ,2
(12.64)
mw
where, mw, PGAs, meff, and meff,2 are given in Tables 12.1, 12.3, and 12.5.
The resulting maximum base shear demands for Structure P1-CWUPT are listed in
Table 12.24. Note that in the preliminary design check described in Section 12.5.3, Qw,dt
was assumed to be equal to 75% of the maximum coupled wall base shear force demand,
Qw,max. Based on the design values in Table 12.24, Equation (12.60) gives a distribution
ratio of 71%.
TABLE 12.24
Structure
Demand Level
P1-CWUPT
Survival
Q1,max (kN) Dm
6988
Qh,max (kN)
0.41
6168
Qw,max (kN) Qw,dt (kN) Fw,dt (kN)

13156
9341
9261
Based on ACI 318-02 (ACI Committee 318 2002), the maximum attainable shear
capacity of a concrete wall can be estimated as:
Fw,dt = 10Aw f c'
(12.65)
The maximum attainable wall pier shear force capacity, Fw,dt for Structure P1CWUPT, listed in Table 12.24, is smaller than but sufficiently close to the maximum wall
pier base shear force demand, Qw,dt. Thus, the design is considered acceptable for
Equation (12.59). The selection and placement of the wall pier transverse reinforcement
to satisfy Qw,dt is not discussed in this report.
12.7.4 Shear Slip at Coupled Wall Base
To prevent shear slip failure between the coupled wall structure and the foundation,
the shear slip capacity of the structure at the base, Fw,ss is required to be larger than the
survival-level maximum base shear force demand, Qw,max. Thus,
Fw, ss > Qw, max
(12.66)
The total shear slip capacity Fw,ss is equal to the slip capacity provided by the two
wall piers, which can be calculated as the sum of two components. The first component
Fg,ss is the slip capacity due to the wall gravity load and the second component Fs,ss is the
slip capacity due to the wall longitudinal mild steel reinforcement at the base as:
362
Fw, ss = Fg , ss + Fs, ss
(12.67)
Fg , ss = 2 wf G
(12.68)
Fs , ss = 2 wf Aws f wsy
(12.69)
with
where, wf is the coefficient of friction between the wall and the foundation, G represents
the factored gravity loads acting at the floor and roof levels of a wall pier, and Aws is the
total area of developed vertical reinforcement crossing the wall-to-foundation joint at the
base of a single wall pier. A value of wf=1.0 is recommended for design. It is assumed
that the coupled wall gravity loads (see Table 12.2 for the unfactored gravity loads) are
distributed equally between the tension-side and compression-side wall piers.
The estimated shear slip capacity for Structure P1-CWUPT, shown in Table 12.25,
satisfies the design requirement given by Equation (12.65).
TABLE 12.25
SHEAR SLIP AT COUPLED WALL BASE
Structure
Demand Level
wf
P1-CWUPT
Survival
1.0
G (kN) Fg,ss (kN) Aws (mm2) Fs,ss (kN) Fw,ss (kN) Qw,max (kN)
2798
5596
47084
38956
44552
13156
12.8 Design of Wall Piers in Structure P2-PWUPT
This section describes the design of the post-tensioned precast concrete wall piers
in Structure P2-PWUPT. The design includes the following components: (1) wall pier
post-tensioning steel reinforcement; (2) yielding of wall pier post-tensioning steel; (3)
concrete confinement at wall pier bases; (4) diagonal tension at wall pier bases; and (5)
shear slip at coupled wall base.
12.8.1 Wall Pier Post-Tensioning Steel Reinforcement
The design of the wall pier post-tensioning steel reinforcement to satisfy the design
coupled wall base shear force demand, Qwd is based on the coupled wall softening state
described in Chapters 7 and 8. The design criterion is given as:
Fws > Qwd
(12.70)
where, Fws is the coupled wall base shear force capacity at the wall softening state.
The following assumptions are made in the design of the wall pier post-tensioning
steel reinforcement, similar to the assumptions made in Chapter 8 for the estimation of
363
the coupled wall softening state: (1) the coupling beam shear forces are equal to Vb,sof; (2)
the strain distributions at the bases of the wall piers are linear; (3) the concrete
compressive stresses at the bases of the wall piers have linear distributions; (4) the neutral
axis depth at the base of the compression-side wall (i.e., right-side wall for lateral loads
applied from left to right) is located at the centerline of the wall (i.e., ccws=0.5lw); (5) the
stresses in the wall pier post-tensioning bars are equal to the initial stress fwpi; and (6) the
maximum concrete compression stress at the base of the tension-side wall is equal to the
design unconfined concrete strength, f c' .
An iterative design procedure is followed as described below.
Step 1
Assume a trial value for the total area of the post-tensioning steel reinforcement in
each wall pier, Awp. Then, the total initial wall pier post-tensioning force can be
determined as:
Pwi = Awp f wpi
(12.71)
Step 2
N tws = G 1.05 Vb, sof + Pwi
(12.72)
N cws = G + 1.05 Vb, sof + Pwi
(12.73)
where, G represents the factored gravity loads applied at the floor and roof levels of a
wall pier, Vb,sof, is from Table 12.17, and the 1.05 factor is described in Chapter 8. It is
assumed that the coupled wall gravity loads (see Table 12.2 for the unfactored gravity
loads) are distributed equally between the tension-side and compression-side wall piers.
Step 3
Determine the neutral axis (i.e., contact) depth at the base of the tension-side wall
(i.e., the depth over which the wall is in contact with the foundation) as:
ctws =
2 N tws
f c't w
(12.74)
Step 4
Determine the moment resistances of the tension-side and compression-side walls,
Mtws and Mcws respectively, and the total base moment and base shear force resistances of
the coupled wall system, Mws and Fws respectively, using the procedure described in
Chapter 8.
364
Step 5
Repeat Steps 1-4 until the coupled wall base shear resistance Fws satisfies the
design demand Qwd.
Step 6
Once the required area of the wall pier post-tensioning steel Awp is known, the
number and placement of the wall post-tensioning tendons can be determined. High
strength post-tensioning bars are used for the precast concrete wall piers in this report.
The total number of post-tensioning bars in each wall pier can be determined from:
nwp =
Awi
awp
(12.75)
where, awp is the area of one bar.

The resulting wall pier cross-section for Structure P2-PWUPT is shown in Fig.
12.16 with the design details summarized in Table 12.26. A total of sixteen 44 mm
(1.75 in.) diameter deformed post-tensioning bars spaced at 203 mm (8 in.) are used in
each wall pier, with awp=1690 mm2 (2.62 in2) and fwpi=0.675fwpu, where fwpu=1035 MPa
(150 ksi) is the design ultimate strength of the bars. Since the coupled wall base shear
force capacity at the wall softening state Fws in Table 12.26 is larger than the design base
shear demand, Qwd, Equation (12.70) is satisfied.
(161-3/4in. PT bars@8in.)
1644mm PT bars@203mm, fwpi=0.675 fwpu
wsp = 5.57%
tw=356mm
(14in.)
layer 1
4
5
6
l w=4.88m (16ft)
Fig. 12.16 Wall pier post-tensioning steel reinforcement

TABLE 12.26
WALL PIER POST-TENSIONING STEEL REINFORCEMENT
Structure
Awp Pwi G Vb,sof Ntws Ncws ctws Mtws Mcws Mwbs

Mws
Fws Qwd
(mm2) (kN) (kN) (kN) (kN) (kN) (mm) (kN.m) (kN.m) (kN.m) (kN.m) (kN) (kN)
P2-PWUPT 27045 18879 2829 5904 15509 27907 1844 27217 45327 45422 117970 5226 4893
12.8.2 Yielding of Wall Pier Post-Tensioning Steel
As discussed in Chapter 11, the coupled wall roof drift capacity at the wall PTyielding state is required to be larger than the design-level coupled wall roof drift demand
(i.e., yielding of the wall pier post-tensioning tendons is not allowed to occur under the
design seismic demand level Design Objective 2). Thus,
365
wy > d
(12.76)
where, wy is the coupled wall roof drift capacity at the wall PT-yielding state (described
in Chapters 7 and 8) and d is the design-level roof drift demand.
For the wall and beam properties determined previously, wy can be estimated
based on the following assumptions: (1) the maximum stress in the wall pier posttensioning tendons is equal to the yield stress, fwpy (corresponding to the linear limit
strain of the post-tensioning steel); and (2) the neutral axis depth at the base of the
tension-side wall is one half the neutral axis depth from the wall softening state, thus,
ctwy=0.5ctws.
Step 1
Estimate the elongation (due to the lateral displacements of the walls) of the wall
pier post-tensioning tendon that yields (usually the extreme tendon on the tension side of
the structure) as:
uwpy =
( f wpy f wpi ) lwpu

Ewp
(12.77)
where, lwpu=hw is the unbonded length and Ewp is the Youngs modulus of the wall posttensioning steel.
Step 2
Estimate the roof drift capacity at the wall PT-yielding state as:
wy =
u wpy
d wp1 ctwy
(12.78)
where, dwp1 is the distance of the post-tensioning tendon that yields from the compression
corner of the tension-side wall and ctwy=0.5ctws from Assumption (2).
The estimated maximum elongation in the post-tensioning bars of Structure P2PWUPT at the wall PT-yielding state, uwpy is given in Table 12.27. The resulting wy
capacity for the structure is larger than the design-level roof drift demand, d. Thus
Equation (12.76) is satisfied. If this design requirement is not satisfied, then the initial
stress in the wall pier post-tensioning tendons, fwpi, may need to be reduced.
366
TABLE 12.27
YIELDING OF WALL PIER POST-TENSIONING STEEL
Structure
uwpy (mm)
wy (%)
d (%)
P2-PWUPT
23
1.06
0.79
Note that the wall pier post-tensioning anchors should be designed for the
maximum tendon forces, Pwu based on the design maximum strength of the posttensioning steel, fwpu as Pwu=awpfwpu.
12.8.3 Concrete Confinement at Wall Pier Bases
As discussed in Chapter 11, the coupled wall roof drift capacity at the wall ultimate
state is required to be larger than the survival-level coupled wall roof drift demand (i.e.,
crushing of the confined concrete at the bases of the wall piers is not allowed to occur
under the survival seismic demand level Design Objective 1). Thus,
wu > s
(12.79)
where, wu is the coupled wall roof drift capacity at the wall ultimate state and s is the
survival-level roof drift demand.
The following assumptions are made in the design of the confinement
reinforcement at the bases of the wall piers, similar to the assumptions made in Chapter 8
for the estimation of the coupled wall ultimate state: (1) the maximum compression strain
in the confined concrete at the base of the compression-side wall is equal to the ultimate
(i.e., crushing) strain of the confined concrete, ccu; (2) the strain distributions at the bases
of the wall piers are linear; (3) the height of the plastic hinge, hwp at the base of the
compression-side wall is equal to the confined thickness of the wall, twc (estimated as the
wall thickness minus the unconfined cover concrete thicknesses); and (4) the neutral axis
depth at the base of the compression-side wall at the wall ultimate state is equal to
ccwu=0.2lw.
Step 1
Estimate the curvature at the base of the compression-side wall when the survivallevel roof drift demand, s is reached as:
cwu =
s
1.5hwp
(12.80)
where, hwp=twc from Assumption (3). This equation follows from the estimation of the
coupled wall ultimate state in Chapter 8.
367
Step 2
Estimate the required ultimate strain for the confined concrete at the base of the
ccu = ccwucwu
(12.81)
where, ccwu=0.2lw from Assumption (4).

Step 3
Design the concrete confinement to achieve ccu. A concrete confinement model,
such as the model developed by Mander et al. (1988), can be used for this purpose.
As shown in Fig. 12.17, spiral reinforcement is used to confine the concrete at the
bases of the wall piers in Structure P2-PWUPT. The stress-strain behavior of the confined
concrete, as estimated using Mander et al. (1988), is also shown in Fig. 12.17. It is
recommended that the confinement reinforcement be provided over a height equal to at
least the first story height. The concrete confinement should be used over a wall length,
lwc where the concrete compression strains exceed the assumed unconfined concrete
crushing strain, cu as:
lwc = ccwu (1
cu
)
ccu
(12.82)
Table 12.28 shows the wall pier confined concrete ultimate strain required by
Equation (12.81) as well as the strain capacity for the confined concrete. Since the strain
capacity is equal to the required strain, crushing of the confined concrete under the
survival demand level is prevented.
Fw,dt > Qw,dt
(12.83)
Because of the compression forces that develop in the compression-side wall pier
due to coupling, a larger portion of the total coupled wall base shear force is carried by
the compression-side wall. It is assumed that the portion of the total maximum coupled
wall base shear force demand, Qw,max carried by the compression-side wall can be
estimated as:
Qw, dt =
M cws + M wbs
Qw, max
M tws + M cws + M wbs
368
(12.84)
confined concrete
swsp=32 mm
Dwsp=292 mm
wire W20
dwsp
=12.7 mm
wsp=
4awsp
Dwspswsp
spiral wire
awsp=129mm2
100
(0.0140, 91.0)
=5.57%
(0.0323, 84.8)
stress (MPa)
80
60
40
smooth relationship
20
0
0
0.01
0.02
strain
0.03
0.04
Fig. 12.17 Concrete confinement at wall pier bases

TABLE 12.28
CONCRETE CONFINEMENT AT WALL PIER BASES
Structure
ccu required ccu provided lwc (mm)
P2-PWUPT
0.032
0.032
878
where, Qw,dt is the maximum wall pier base shear demand, Mtws and Mcws are the base
moment resistances of the tension-side and compression-side wall piers at the softening
state (see Table 12.22), and Mwbs is the contribution of the coupling beam shear forces to
the coupled wall base shear resistance at the softening state.
(12.85)
The base shear demand for the first mode component is estimated as the coupled
wall base shear capacity at the wall ultimate state, Fwu as:
369
Q1,max = Fwu
(12.86)
The estimation of Fwu is described in Chapter 8. The higher mode component of the
maximum base shear demand, Qh,max is estimated as:
(12.87)
with,
Dm =
mw
= 1
meff
mw
+ 0.7
meff ,2
(12.88)
mw
where, mw, PGAs, meff, and meff,2 are given in Tables 12.1, 12.3, and 12.5.
The resulting maximum base shear demands for Structure P2-PWUPT are listed in
Table 12.29. Note that in the preliminary design check described in Section 12.5.3, Qw,dt
was assumed to be equal to 75% of the maximum coupled wall base shear force demand,
Qw,max. Based on the design values in Table 12.29, Equation (12.84) gives a distribution
ratio of 76%.
TABLE 12.29
Structure
P2-PWUPT
Demand Level Q1,max (kN)

Survival
6863
Dm
Qh,max (kN)
Qw,max (kN)
Qw,dt (kN)
Fw,dt (kN)
0.41
6236
13099
9955
9261
Based on ACI 318-02 (ACI Committee 318 2002), the maximum attainable shear
capacity of a concrete wall can be estimated as:
Fw,dt = 10 Aw f c'
(12.89)
The maximum attainable wall pier shear force capacity, Fw,dt for Structure P2PWUPT, listed in Table 12.29, is smaller than but sufficiently close to the maximum wall
pier base shear force demand, Qw,dt. Thus, the design is considered acceptable for
Equation (12.83). The selection and placement of the wall pier transverse reinforcement
to satisfy Qw,dt is not discussed in this report.
370
12.8.5 Shear Slip at Coupled Wall Base
To prevent shear slip failure between the coupled wall structure and the foundation,
the shear slip capacity of the structure at the base, Fw,ss is required to be larger than the
survival-level maximum base shear force demand, Qw,max. Thus,
Fw, ss > Qw, max
(12.90)
The total shear slip capacity Fw,ss is equal to the slip capacity provided by the two
wall piers, which can be calculated as the sum of two components. The first component
Fg,ss is the slip capacity due to the wall gravity load and the second component Fp,ss is the
slip capacity due to the wall post-tensioning force as:
Fw, ss = Fg , ss + Fp , ss
(12.91)
Fg , ss = 2 wf G
(12.92)
Fp, ss = 2 wf awp f wpr
(12.93)
with
where, wf is the coefficient of friction between the wall and the foundation, G represents
the factored gravity loads acting at the floor and roof levels of a wall pier, awp is the area
of one wall post-tensioning bar, and fwpr represents the residual stresses in the wall posttensioning bars after losses under cyclic loading. A value of wf=0.7 is recommended for
design. It is assumed that the coupled wall gravity loads (see Table 12.2 for the
unfactored gravity loads) are distributed equally between the tension-side and
compression-side wall piers.
Under reversed cyclic loading up to the survival-level roof drift demand s,
prestress losses occur in the wall pier post-tensioning bars if the maximum bar strain
exceeds the yield (i.e., linear limit) strain. This loss of prestress must be considered in
calculating Fp,ss based on Equation (12.93). Referring to Fig. 12.18, the residual stress
after losses in a post-tensioning bar can be estimated as:
f wpr = f wpy + f wpi wp f wpy Ewp wps + wp Ewp wps
(12.94)
where, fwpy is the yield strength, fwpi is the initial stress (i.e., stress before the application
of lateral loads), wp is the post-yield stiffness ratio, Ewp is the linear-elastic stiffness, and
wps is the maximum strain reached in the wall post-tensioning bar under the survivallevel ground motion. Equation (12.94) can be simplified by assuming, conservatively,
that pt=0, which gives
371
f wpr = f wpy + f wpi Ewp wps
(12.95)
wpEwp
fwpy+(wps-wpy)wpEwp
wall PT bar stress
fwpy
Ewp
(wps-wpi)Ewp
fwpi
fwpr
wpi
wpy=fwpy/Ewp wps
wall PT bar strain

Fig. 12.18 Prestress losses in wall pier post-tensioning bars
It should be conservatively assumed that the survival-level roof drift demand, s is
reached in the positive and negative directions of loading, and then, the structure is
brought back to zero displacement. Thus, under symmetric reversed cyclic loading, the
post-tensioning bars of the right-side wall pier will have the same amount of prestress
losses as the bars in the left-side wall pier. When the lateral loads are applied from left to
right, the maximum strains reached in the post-tensioning bars of the left-side wall pier
(i.e., tension-side wall pier) can be estimated as:
wps =
s (d wp ctwu )
lwpu
+ wpi
(12.96)
where, dwp is the distance of each bar from the compression corner of the tension-side
wall pier, ctwu is the neutral axis (i.e., contact) depth at the base of the tension-side wall
pier at the coupled wall ultimate state, lwpu is the unbonded length of the wall posttensioning bar (usually equal to the wall height, hw), and wpi is the initial strain of the bar.
The neutral axis depth at the base of the tension-side wall is taken as ctwu=ctwy=0.5ctws.
Using the above procedure, Table 12.30 shows the residual post-tensioning bar
stresses, fwpr in the tension-side wall of Structure P2-PWUPT after displacing the
structure in one direction to the estimated survival-level roof drift demand, s, and then,
unloading it to zero lateral displacement. Note that fwpr=fwpi if wps is less than or equal to
the yield strain, wpy.
372
When the structure is displaced in the opposite direction, the post-tensioning bar
layers 5 through 8 may experience larger tension strains. In other words, considering
loading in both the positive and negative directions of displacement, the residual bar
stresses in post-tensioning bar layers 5 through 8 may be smaller than the values shown
in Table 12.30. In the design of Structure P2-PWUPT, the total residual post-tensioning
bar force was conservatively calculated as four times the sum of the residual forces in
post-tensioning bars layers 1 through 4, as shown in Table 12.30. It can be seen that the
estimated shear slip capacity of the structure satisfies the design requirement given by
Equation (12.90).
TABLE 12.30
SHEAR SLIP AT COUPLED WALL BASE
wf
G
(kN)
Fg,ss
ctwu lwpu
(kN) (mm) (m)
0.7
2829
3961
922
Bar
Layer No.
1
2
3
4
32.6
5
6
7
8
dwp
(m)
3.15
2.95
2.74
2.54
2.34
2.13
1.93
1.73
4
Fp,ss Fw,ss Qw,max
fwpr
4 f wpr ( 2a wp )
(MPa) i = 1
(kN) (kN) (kN)
620
642
664
686
33594
23516 27477 13099
698
698
698
698
12.9 Summary of Design and Final Checks
This section provides a summary of the designs for Structures P1-CWUPT and P2PWUPT and conducts final design checks as follows: (1) lateral load versus deformation
relationships; and (2) design criteria.
12.9.1 Lateral Load versus Deformation Relationships
Based on the structure design properties from the previous sections, the solid lines
in Figs. 12.19a and 12.19b show the idealized coupled wall base shear force versus roof
drift (F-) relationships for Structures P1-CWUPT and P2-PWUPT, respectively. The
procedures described in Chapter 8 and the idealized Vb-b relationships in Fig. 12.9 are
used to develop a bi-linear F- relationship for Structure P1-CWUPT and a tri-linear F-
relationship for Structure P2-PWUPT. The bi-linear relationship for Structure P1CWUPT is based on the coupled wall softening state (at Fws, ws) and the coupled wall
ultimate state (at Fwu, wu). In addition to these two states, the tri-linear relationship for
Structure P2-PWUPT also uses the wall PT-yielding state (at Fwy, wy). A first-modeshape distribution of lateral loads over the height of the walls is assumed with the lateral
loads equally divided between the left and right wall piers.
The results from the idealized F- relationships (star markers in Fig. 12.19) are
summarized in Table 12.31. Comparisons of the estimated capacities with the design
demands from the previous sections indicate that the design of both prototype structures
is satisfactory.
373
8000
7000
wall ultimate
(Fwu,wu)
6000
wall softening
(Fws,ws)
5000
4000
3000
P1-CWUPT
2000
1000
0
0
0.4
0.8
1.2
1.6
(a)
8000
7000
wall ultimate
wall softening
6000
(Fwu,wu)
(Fws,ws)
wall PT-yielding
(Fwy,wy)
5000
4000
3000
P2-PWUPT
2000
tri-linear estimation
1000
0
0.4
0.8
1.2
1.6
(b)
Fig. 12.19 Coupled wall base shear force versus roof drift (F-) relationships:
374
TABLE 12.31
IDEALIZED F- BEHAVIORS FOR STRUCTURES P1-CWUPT AND P2-PWUPT
Structure
Wall Softening State Wall PT-Yielding State
Wall Ultimate State
Fws (kN)
ws (%)
Fwy (kN)
wy (%)
Fwu (kN)
wu (%)
P1-CWUPT
4603
0.13
6988
1.97
P2-PWUPT
5226
0.15
6383
1.06
6863
1.79
For comparison, the dashed lines in Fig. 12.19 show the smooth F- behaviors of
Structures P1-CWUPT and P2-PWUPT using the fiber element analytical model from
Chapter 6. In general, the idealized relationships capture the structure capacities from the
fiber element models reasonably well. More detailed information on the expected
behavior of the prototype structures using the fiber element models is given in Chapter 13.
12.9.2 Design Criteria
As demonstrated throughout the previous sections of this chapter, the seismic

design of a coupled wall structure involves establishing structure demands and providing
structure capacities until required design criteria are satisfied. A summary of the final
design comparisons between the estimated demands and capacities of Structures P1CWUPT and P2-PWUPT are given in Table 12.32, with the required design criteria listed
below. Note that the structure design capacities and demands in Table 12.32 are
estimated using the idealized design relationships and procedures from Chapters 5, 8, 10,
and 11, without the simplifying assumptions made during the design of the structures. As
expected, the comparisons between the capacities and the demands indicate that the
prototype structures satisfy all of the required design criteria reasonably well.
(1) To satisfy the displacement-based design requirement, the roof drift capacity of
the coupled wall structure at the wall ultimate state, wu should be larger than the
allowable survival-level roof drift demand, s. Thus,
wu > s
(12.97)
(2-1) For systems with cast-in-place concrete walls, 2/3 times the roof drift capacity
of the coupled wall structure at the wall ultimate state, wu should be larger than the
allowable design-level roof drift demand, d. This requirement ensures that an adequate
margin of reserve displacement capacity exists between the survival-level and designlevel demands, and between the collapse prevention and life safety performance levels.
Thus,
(2 / 3)wu > d
(12.98)
(2-2) For systems with precast concrete walls, the roof drift capacity of the coupled
wall structure at the wall PT-yielding state, wy should be larger than the allowable
design-level roof drift demand, d. This requirement ensures that yielding of the wall
375
post-tensioning steel does not occur under the design-level earthquake and that an
adequate margin of reserve displacement capacity exists between the survival-level and
design-level demands. Thus,
wy > d
(12.99)
(3) To prevent premature softening of the structure, an adequate lateral strength

needs to be ensured. The base shear capacity of the coupled wall system at the wall
softening state, Fws should be larger than the design base shear force demand, Qwd. Thus,
Fws > Qwd
(12.100)
TABLE 12.32
SEISMIC DESIGN CRITERIA FOR STRUCTURES P1-CWUPT AND P2-PWUPT
P1-CWUPT
Design Criterion
(1)
(2)
P2-PWUPT
Capacity Demand Capacity Demand
wu > s
(2 / 3)wu > d (P1-CWUPT)
wy > d (P2-PWUPT)
1.97%
1.67%
1.79%
1.71%
1.31%
0.71%
1.06%
0.79%
(3)
Fws > Qwd
(4)
Fw, ss > Qw, max
(5)
Fw, dt > Qw, dt
9261 kN
Vba,sof > Qbd,a
44 kN
30 kN
432 kN
360 kN
Vbp,sof > Qbd, p
623 kN
588 kN
316 kN
294 kN
(7)
b, pty > s
7.24%
5.49%
6.90%
5.62%
(8)
(2 / 3)b, pty > d
4.83%
2.33%
4.60%
2.59%
(9)
Vb, ss > Qb, max
1169 kN
963 kN
1273 kN
1184 kN
(10)
f ccw > wc
(11)
Pbi < 0.25Abc fby
2402 kN
2500 kN 2402 kN
1251 kN
(12)
Pby < Abfc fbm
4159 kN
3784 kN 4159 kN
1892 kN
(13)
uas < 38 mm (1.5 in.)
38 mm
30 mm
38 mm
31 mm
(14)
Casx > 2Tayx
425 kN
106 kN
1063 kN
1032 kN
(6)
4603 kN
4782 kN 5226 kN
4893 kN
44614 kN 13156 kN 27477 kN 13099 kN

9341 kN 9261 kN
9955 kN
98.7 MPa 92.4 MPa 90.4 MPa 82.1 MPa
(4) To prevent shear slip failure between the coupled wall structure and the
foundation, the shear slip capacity of the structure at the base, Fw,ss should be larger than
the survival-level maximum base shear force demand, Qw,max. Thus,
376
Fw, ss > Qw, max
(12.101)
(5) To prevent diagonal tension failure of the wall piers, the diagonal tension
capacity of each wall pier, Fw,dt should be larger than the survival-level maximum wall
pier shear force demand, Qw,dt. Thus,
Fw, dt > Qw, dt
(12.102)
(6) To ensure that the target coupling degree, DOC and angle strength ratio, ar
selected during design are satisfied, the contribution of the top and seat angles, Vba,sof and
the contribution of the beam post-tensioning force, Vbp,sof to the coupling beam shear
force capacity at the beam softening state, Vb,sof should be larger than the corresponding
demands, Qbd,a and Qbd,p, respectively. Thus,
Vba, sof > Qbd , a
(12.103)
Vbp, sof > Qbd , p
(12.104)
(7) To prevent yielding of the beam post-tensioning steel under the survival seismic
demand level, the chord rotation capacity of the coupling beams at the beam PT-yielding
state, b,pty should be larger than the survival-level beam chord rotation demand, s. Thus,
b, pty > s
(12.105)
(8) 2/3 times the chord rotation capacity of the coupling beams at the beam PTyielding state, b,pty should be larger than the design-level beam chord rotation demand, d.
This requirement ensures that an adequate margin of reserve displacement capacity exists
between the survival-level and design-level demands, and between the collapse
prevention and life safety performance levels. Thus,
(2 / 3)b, pty > d
(12.106)
(9) To prevent shear slip at the ends of the coupling beams, the shear slip capacity
at the beam-to-wall interfaces, Vb,ss should be larger than the survival-level maximum
beam shear force demand, Qb,max. Thus,
Vb, ss > Qb, max
(12.107)
(10) To transfer the coupling beam contact stresses with little deformation in the
wall concrete, the compressive strength of the confined concrete in the beam-to-wallcontact regions, fccw should be larger than the maximum anticipated concrete contact
stresses under the survival seismic demand level, wc. Thus,
377
f ccw > wc
(12.108)
(11) To prevent excessive local distortions and possible instability of the coupling
beams, the total initial beam post-tensioning force, Pbi should be smaller than 0.25 times
the axial yield strength of the coupling beam cross section (plus cover plates, if used).
Thus,
Pbi < 0.25Abc fby
(12.109)
(12) To prevent excessive local distortions and possible instability of the coupling
beams, the total yield strength of the beam post-tensioning tendons, Pby should be smaller
than the maximum axial strength of one coupling beam flange (plus cover plate, if used).
Thus,
Pby < Abfc fbm
(12.110)
(13) To prevent low cycle fatigue fracture of the top and seat angles, the maximum
anticipated deformation of the angles in tension under the survival demand level should
be smaller than the deformation capacity of the angles in tension. More research needs to
be conducted on the behavior and capacity of the top and seat angles to ensure that the
design deformation can be reached without fracture. Based on the subassemblage
experiment results in Chapter 9, it is suggested that the deformation of the tension angles
under the survival demand level uas be smaller than 38 mm (1.5 in.). Thus,
uas < 38 mm (1.5 in.)
(12.111)
(14) To prevent slip of the angle-to-beam connection bolts as the angles are pulled
away from the walls, the slip capacity of the angle-to-beam bolts, Casx should be larger
than the maximum anticipated strength of the angles in tension, 2Tayx. Thus,
Casx > 2Tayx
(12.112)
(15) The angle-to-wall connections should also be designed for the maximum
anticipated strength of the angles in tension. No detailed design information is provided
in the report on this topic.
(16) The coupling beam and wall pier (in the case of precast concrete walls) posttensioning anchors should be designed for the maximum tendon forces based on the
design maximum strength of the post-tensioning steel. No detailed design information is
provided in the report on this topic.
(17) If cover plates are used to strengthen the coupling beam flanges, the welds
between the cover plates and the flanges should be designed for the maximum anticipated
forces. No detailed design information is provided in the report on this topic.
378
CHAPTER 13
BEHAVIOR OF PROTOTYPE STRUCTURES
UNDER STATIC LATERAL LOADING
This chapter discusses the expected behavior of prototype Structures P1-CWUPT
and P2-PWUPT from Chapter 12 under static lateral loading. Evaluations of the global
behavior of the structures as well as the local behavior of the wall piers and the coupling
beams are provided. The behavior of Structure P1-CWUPT is compared with the
behavior of a similar structure with embedded steel coupling beams. The chapter is
organized into the following sections: (1) introduction; (2) behavior under monotonic
loading; (3) behavior under reversed cyclic loading; and (4) structure with embedded
steel coupling beams.
13.1 Introduction
The nonlinear behavior of each prototype coupled wall structure is investigated
under static lateral loads combined with gravity loads using the DRAIN-2DX (Prakash et
al. 1993) fiber element model described in Chapter 6 including the modifications
recommended in Chapter 10. The wall-contact elements from Chapter 6 and the vertical
angle elements from Chapter 10 are not used in the modeling of the structures in order to
simplify the models and reduce computation time. Previous analytical investigations have
shown that these elements do not have a significant effect on the behavior of multi-story
coupled wall structures, and thus, they can be ignored.
The analyses are conducted by applying loads on the structures in the following
order: (1) beam and wall pier (for Structure P2-PWUPT) post-tensioning forces; (2)
gravity loads applied at the floor and roof levels; and (3) equivalent lateral forces applied
at the floor and roof levels in displacement control.
The beam and wall pier post-tensioning element forces are applied all at once. Then,
the gravity loads, G are applied under the same load combination used in design with
100% of the unfactored design dead load (D) plus 25% of the unreduced unfactored
design live load (L) to represent the amount of gravity loading that may be acting on the
structure during an earthquake. The use of the same load combination in the design and
analysis of the structures allows for more direct comparisons between the target seismic
performance objectives and the estimated structure demands and capacities.
As described in Chapter 7, the gravity loads are applied at the wall-height element
nodes at the floor and roof levels, with no loads applied along the length of the coupling
379
beams. Table 13.1 provides a summary of the gravity loads applied on the left and right
wall piers in Structures P1-CWUPT and P2-PWUPT.
TABLE 13.1
WALL PIER GRAVITY LOADS IN STRUCTURES P1-CWUPT AND P2-PWUPT
Level
Roof
8th floor
7th floor
6th floor
5th floor
4th floor
3rd floor
2nd floor
Structure P1-CWUPT
1.0D (kN)
0.25L (kN)
330
4
330
17
330
17
330
17
330
17
330
17
330
17
366
17
Structure P2-PWUPT
1.0D (kN)
0.25L (kN)
334
4
334
17
334
17
334
17
334
17
334
17
334
17
370
17
The displacement controlled nonlinear lateral load analyses are conducted based
on the lateral displacement of Node 54 (as shown in Fig. 7.1) with respect to the wall
base. A first mode distribution of lateral loads applied from left to right over the height of
the walls is used. For each coupled wall system, the lateral loads are equally divided
between the left and right wall piers and are applied at the wall-height element nodes at
the floor and roof levels (at the same nodes where the gravity loads are applied). Note
that inertial force distributions significantly different than the assumed first mode lateral
load distribution may occur during an earthquake (e.g., Ghosh and Markevicius 1990;
Paulay and Priestley 1992; Eberhard and Sozen 1993; Kabeyasawa 1993; Aoyama 1993;
Otani et al. 1994; Kurama et al. 1997, 1999b); however, this is not investigated in this
chapter.
13.2 Behavior Under Monotonic Loading
This section evaluates the behavior of Structures P1-CWUPT and P2-PWUPT
under monotonic lateral loads combined with gravity loads as follows: (1) coupled wall
base shear force versus roof drift behaviors; (2) wall pier base flexural steel strains and
stresses; (3) wall pier base axial forces; (4) wall pier base shear forces; (5) wall pier base
moments; (6) degree of coupling; (7) wall pier base concrete strains; (8) coupling beam
shear force versus chord rotation behaviors; (9) coupling beam axial forces and posttensioning forces; (10) coupling beam end strains; and (11) tension angle force versus
deformation behaviors.
Note that the lateral load behaviors of Structures P1-CWUPT and P2-PWUPT are
similar to the behaviors of the CIP-UPT and PRE-UPT systems discussed in Chapter 7.
Thus, the sections below make frequent references to Chapter 7 in evaluating the results.
13.2.1 Coupled Wall Base Shear Force versus Roof Drift Behaviors
Figs. 13.1a and 13.1b show the coupled wall base shear force versus roof drift (F-)
behaviors of Structures P1-CWUPT and P2-PWUPT, respectively (same as the dashed
380
lines in Fig. 12.19a and 12.19b, respectively). The base shear force is calculated as the
sum of the lateral forces applied over the height of the structures, and the roof drift is
calculated as the average lateral displacement of the tension-side wall (i.e., left-side wall
for lateral loads applied from left to right) and the compression-side wall (i.e., right side
wall) at the roof (i.e., at Nodes 27 and 54 in Fig. 7.1) divided by the wall height to the
base, hw.
The following states are identified on the F- relationships in Fig. 13.1, similar to
the states described in Chapters 7 and 8:
(1) Tension-side wall softening ( markers) This state is defined as the state
when the neutral axis at the base of the tension-side wall pier reaches the centerline of the
wall pier.
(2) First beam gap opening ( markers) This state is defined as the state when gap
opening initiates at the beam-to-wall interfaces of the structure.
(3) Tension-side wall gap opening ( marker) This state is defined as the state
when gap opening initiates at the base of the tension-side wall pier in Structure P2PWUPT.
(4) First beam flange yielding in compression ( markers) This state is defined as
the state when compression yielding of the coupling beam flanges occurs for the first
time in the structure.
(5) Compression-side wall softening (+ markers) This state is defined as the state
when the neutral axis at the base of the compression-side wall pier reaches the centerline
of the wall pier.
(6) First angle yielding in tension ( markers) This state is defined as the state
when the force in the tension angles reaches the angle yield strength, Tayx for the first
time in the structure.
(7) Compression-side wall gap opening ( marker) This state is defined as the
state when gap opening initiates at the base of the compression-side wall pier in Structure
P2-PWUPT.
(8) Coupled wall softening ( markers) As discussed in Chapter 8, the base shear
force at this state, Fws is taken as the average of the maximum and minimum base shear
forces from the compression-side wall softening, first wall pier steel yielding, and
coupling beam softening states for Structure P1-CWUPT and from the compression-side
wall softening, tension-side wall concrete linear limit, and coupling beam softening states
for Structure P2-PWUPT. The roof drift at the coupled wall softening state, ws is taken
as the roof drift corresponding to Fws.
381
8000
7000
6000
Qwd=4782kN
5000
tension-side wall softening

first beam gap opening
first beam flange yielding in comp.
coupled wall softening
first angle yielding in tension
first wall pier steel yielding
first angle strength in tension
4000
3000
2000
1000
0
0
s=1.67%
d=0.71%
0.2
0.4
0.6
0.8
1.2
1.4
1.6
1.8
(a)
8000
7000
6000
5000
Qwd=4893kN
4000
3000
2000
1000
0
0
s=1.71%
d=0.79%
0.2
0.4
0.6
first beam gap opening

tension-side wall gap opening
first beam flange yielding in comp.
first angle yielding in tension
compression-side wall gap opening
tension-side wall softening
coupled wall softening
first angle strength in tension
first wall pier steel yielding
0.8
1
1.2
1.4
1.6
1.8
(b)
Fig. 13.1 Coupled wall base shear force versus roof drift behaviors:
382
(9) First wall pier steel yielding ( markers) This state is defined as the first
yielding of the wall pier mild flexural reinforcing bars in Structure P1-CWUPT and as the
first yielding of the wall pier post-tensioning bars in Structure P2-PWUPT (also referred
to as the coupled wall PT-yielding state, reached at Fwy and wy, as described in Chapter
8). Note that the linear limit point on the steel stress-strain relationship is used to identify
the yielding of the post-tensioning bars.
(10) Tension-side wall concrete linear limit ( markers) This state is defined as
the state when the maximum confined concrete compression stress at the base of the
tension-side wall pier reaches the linear limit (i.e., limit of proportionality) stress of the
confined concrete.
(11) Coupling beam softening ( markers) This state is defined as the state when
the average coupling beam shear force in the structure (i.e., the sum of the shear forces in
the coupling beams divided by the number of beams) is equal to 1.05Vb,sof, where Vb,sof is
the softening state coupling shear force for an isolated subassemblage (see Chapters 10
and 12) and the factor 1.05 is described in Chapter 8.
(12) First angle strength in tension ( markers) This state is defined as the state
when the force in the tension angles reaches the full angle tension strength, 2Tayx for the
first time (based on the revised tension angle model in Chapter 10) in the structure.
(13) Compression-side wall concrete crushing (
markers) This state, also
referred to as the coupled wall ultimate state (reached at Fwu and wu) in Chapter 8, is
defined as the state when the extreme strain in the confined concrete at the base of the
compression-side wall pier reaches the crushing strain, ccu.
Fig. 13.1 shows that, due to the larger compressive axial forces in the wall piers as
a result of post-tensioning, Structure P2-PWUPT reaches the tension-side wall softening
state ( markers) and the compression-side wall softening state (+ markers) at larger base
shear forces than Structure P1-CWUPT. Furthermore, as a result of the smaller coupling
beam post-tensioning forces while the beam cross section is kept the same, Structure P2PWUPT reaches the first beam gap opening state ( markers) at a smaller base shear
force than Structure P1-CWUPT. Due to the earlier gap opening at the coupling beam
ends, Structure P2-PWUPT also reaches the first beam flange yielding in compression
state ( markers) at a smaller base shear force than Structure P1-CWUPT.
The horizontal dashed lines in Figs. 13.1a and 13.1b show the coupled wall base
shear demands, Qwd from Chapter 12 (Qwd=4782 kN for Structure P1-CWUPT and
Qwd=4893 kN for Structure P2-PWUPT). It is observed that the base shear capacities of
the prototype structures at the coupled wall softening state, Fws ( markers) satisfy the
Qwd demands.
Similarly, the vertical dashed lines in Figs. 13.1a and 13.1b show the design-level
and survival-level roof drift demands, d and s, respectively, from Chapter 12
383
(d=0.71%, s=1.67% for Structure P1-CWUPT, and d=0.79%, s=1.71% for Structure
P2-PWUPT). As required by design, for both structures, the roof drift capacities at the
coupled wall ultimate state ( markers), wu satisfy the s demands. In addition, the roof
drift capacity at the coupled wall PT-yielding state ( marker), wy for Structure P2PWUPT satisfies the d demand.
Note that, for both structures, yielding of the coupling beam post-tensioning
tendons (i.e., beam PT-yielding state) occurs after the coupled wall ultimate state, and,
thus is not shown in Fig. 13.1. As required by design, yielding of the beam posttensioning tendons does not occur under the survival level demand.
A more detailed evaluation of the behavior of the prototype structures is provided
in the following sections. A summary of the results from the analyses is given in Table
13.2, together with estimated values from the idealized relationships and procedures in
Chapters 5, 8, 10, and 11. It is observed that the coupled wall base shear force, F and roof
drift, can be obtained with good accuracy using the idealized relationships. Note that
the estimated values in Table 13.2 do not use the simplifying assumptions made during
the design of the structures in Chapter 12.
13.2.2 Wall Pier Base Flexural Steel Strains and Stresses
Figs. 13.2a and 13.2b show the stresses in selected layers of flexural mild steel
reinforcement at the bases of the tension-side and compression-side wall piers from
Structure P1-CWUPT, respectively, versus the coupled wall roof drift, .
Similarly, Figs. 13.3a and 13.3b show the stresses, fwp1 through fwp8, in the eight
pairs of post-tensioning bars of the tension-side and compression-side wall piers from
Structure P2-PWUPT, respectively. The corresponding stress-strain (fwp-wp) behaviors of
the post-tensioning bars are shown in Figs. 13.4a and 13.4b, respectively. Similar to the
PRE-UPT system in Chapter 7, the post-tensioning bar strains in the tension-side wall
pier are larger than the strains in the compression-side wall pier since the tension-side
wall is subjected to tension forces from the coupling beams. As expected, the posttensioning bars that are further away from the compression (i.e., right) corner at the base
of each wall pier reach larger tensile strains.
The vertical dashed lines in Fig. 13.3 show the design-level and survival-level roof
drift demands (d and s, respectively) from Chapter 12. As required by design, none of
the wall pier post-tensioning bars yields under the design-level roof drift demand, d. As
the structure is displaced further to the survival demand level, the four pairs of posttensioning bars furthest away from the compression corner of the tension-side wall (i.e.,
fwp1 through fwp4) yield beginning at a roof drift of, approximately, wy=1.12%,
corresponding to the coupled wall PT-yielding state (see Fig. 13.1b and Table 13.2). In
the compression-side wall, only the first pair of bars (fwp1) furthest away from the
compression corner yields at a roof drift of, approximately, =1.5%. As a result of the
384
use of unbonded tendons, the strains in the wall pier post-tensioning bars remain small,
even for the bars that yield (see Fig. 13.4).
As shown in Table 13.2, the total post-tensioning forces in the tension-side and
compression-side wall piers, Ptw and Pcw, and the roof drift at the wall PT-yielding state,
wy can be estimated with good accuracy using the idealized relationships and procedures
from Chapters 5, 8, 10, and 11.
TABLE 13.2
COMPARISONS BETWEEN ESTIMATION AND ANALYTICAL RESULTS
P1-CWUPT
Wall Softening
State
P2-PWUPT
Wall Ultimate
State
Wall Softening Wall PT-Yielding

State
State
Wall Ultimate
State
DRAIN estimate DRAIN estimate DRAIN estimate DRAIN estimate DRAIN estimate
coupled wall base
4670
4603
7335
6988
5182
5226
6628
6383
6912
6863
shear force, F (kN)
coupled wall base
105644 101946 165800 161323 117150 117970 149895 144019 156302 154844
moment, Mw (kN-m)
coupled wall roof
0.31
0.13
2.01
1.97
0.33
0.15
1.12
1.06
1.75
1.79
drift, (%)
tension wall base
1735
1797
1352
992
952
shear force, Ftw (kN)
compression wall base
2940
5591
3741
5431
5711
shear force, Fcw (kN)
tension wall base
24250 29241 41787 41604 26702 27217 32680 32866 34894
moment, Mtw (kN-m)
41810 32099 74083 76242 43855 45327 59427 58235 61800
moment, Mcw (kN-m)
degree of coupling,
36
40
28
40
37
39
34
37
34
34
DOC (percent)
tension wall base
-2727
-2993 -3759 -3372 15435 15509 15741 16039 17699 16489
axial force, Ntw (kN)
8384
8589
9416
9047
27769 27907 30460 31132 31821 32742
axial force, Ncw (kN)
tension wall
19158 18879 20928 21128 21862 23018
PT force, Ptw (kN)
compression wall
18384 18879 19620 20403 20786 21764
PT force, Pcw (kN)
tension wall neutral
1285
1270
615
635
2159
1844
963
922
838
922
axis depth, ctw (mm)
comp. wall neutral
2019
2438
955
932
2591
2438
1219
1219
1118
1219
axis depth, ccw (mm)
average coupling beam
694
689
823
743
820
738
1001
860
1026
926
shear force, Vb (kN)
average coupling beam
0.87
0.17
5.53
6.24
0.95
0.19
3.46
2.82
5.48
4.86
chord rotation, b (%)
385
600
P1-CWUPT
tension-side wall
tension
400
200
0
200
400
600
800
0
0.5
1
1.5
(a)
800
800
P1-CWUPT
600 compression-side wall
tension
2
400
200
0
200
400
600
800
0
0.5
1
1.5
(b)
Fig. 13.2 Wall pier flexural mild steel stresses at base of Structure P1-CWUPT:
(a) tension-side wall; (b) compression-side wall
13.2.3 Wall Pier Base Axial Forces
The solid and dashed lines in Figs. 13.5a and 13.5b show the axial forces, Ntw and
Ncw, at the bases of the tension-side and compression-side wall piers in Structures P1CWUPT and P2-PWUPT, respectively. A positive wall pier axial force indicates a
compressive force. The markers shown in Figs. 13.5a and 13.5b are the same as the
markers in Figs. 13.1a and 13.1b, respectively.
Figs. 13.6a and 13.6b show the contributions of the wall pier gravity loads, Ntwg
and Ncwg, wall pier post-tensioning forces (for Structure P2-PWUPT only), Ntwp and Ncwp,
and coupling beam shear forces, Ntwb and Ncwb, on the wall pier axial forces, Ntw and Ncw
in Structures P1-CWUPT and P2-PWUPT, respectively. The vertical lines in Figs. 13.5
and 13.6 show the design-level and survival-level roof drift demands, d and s,
respectively, from Chapter 12.
Similar to the results in Chapter 7, Figs. 13.5 and 13.6 show that the posttensioning of the precast concrete walls in Structure P2-PWUPT results in significantly
larger compressive axial forces than the forces in the cast-in-place concrete walls of
Structure P1-CWUPT. The coupling shear forces result in tensile axial forces in the
tension-side walls and compressive axial forces in the compression-side walls (Ntwb and
Ncwb, respectively). Thus, as the structures are displaced laterally, the compressive axial
forces in the compression-side walls increase and the axial forces in the tension-side
walls decrease.
386
1000
P2-PWUPT
tension-side wall
......
(MPa)
600
400
d=0.79%
P2-PWUPT
tension-side wall
800
s=1.71%
200
layer 1
0
600
400
d=0.79%
200
layer 2
0.5
1.5
0.5
1.5
1000
P2-PWUPT
tension-side wall
(MPa)
P2-PWUPT
tension-side wall
800
wp4
800
wall pier PT bar stress, f
wall pier PT bar stress, f wp3 (MPa)
1
1000
600
s=1.71%
d=0.79%
400
200
layer 3
600
s=1.71%
d=0.79%
400
200
layer 4
0
0
0.5
1.5
0.5
1.5

1000
P2-PWUPT
tension-side wall
P2-PWUPT
tension-side wall
800
wp6
800
(MPa)
1000
s=1.71%
0
0
600
400
200
layer 5
s=1.71%
d=0.79%
600
400
200
layer 6
d=0.79%
s=1.71%
0
0
0.5
1.5
0.5
1.5

1000
P2-PWUPT
tension-side wall
P2-PWUPT
tension-side wall
800
wp8
800
(MPa)
1000
wp2
800
1000
600
400
s=1.71%
d=0.79%
200
layer 7
600
400
s=1.71%
d=0.79%
200
layer 8
0
0
0.5
1.5
0.5
1
Fig. 13.3a Wall pier post-tensioning bar stresses

in Structure P2-PWUPT: tension-side wall
387
1.5
600
800
1000
P2-PWUPT
wp1
(MPa)
1000
......
400
d=0.79%
s=1.71%
200
P2-PWUPT
800
600
400
d=0.79%
layer 2
0
0.5
1.5
0.5
(MPa)
wp3
600
400
s=1.71%
d=0.79%
200
layer 3
P2-PWUPT
800
600
400
s=1.71%
d=0.79%
200
layer 4
0.5
1.5
0.5
1.5

1000
P2-PWUPT
1000
wp5
(MPa)
0
0
1000
P2-PWUPT
800
600
400
s=1.71%
d=0.79%
200
layer 5
0
P2-PWUPT
800
600
400
d=0.79%
200
s=1.71%
layer 6
0
0.5
1.5
0.5
1.5

1000
P2-PWUPT
800
1000
wp7
(MPa)
1.5
1000
s=1.71%
200
layer 1
0
800
600
400
s=1.71%
d=0.79%
200
layer 7
0
P2-PWUPT
800
600
400
s=1.71%
d=0.79%
200
layer 8
0
0.5
1.5
0.5
1
Fig. 13.3b Wall pier post-tensioning bar stresses

in Structure P2-PWUPT: compression-side wall
388
1.5
......
1000
1000
800
600
400
P2-PWUPT
tension-side wall
layer 1
200
2
800
600
400
P2-PWUPT
tension-side wall
layer 2
200
0
0
wall pier PT bar strain, wp1 (x10-3)
800
600
400
P2-PWUPT
tension-side wall
layer 3
200
0
1
wall pier PT bar strain,
wp3
600
400
P2-PWUPT
tension-side wall
layer 4
200
(x10 )
wp4
(x10-3)
1000
800
wp5
(MPa)
800
-3
1000
0
0
600
400
P2-PWUPT
tension-side wall
layer 5
200
800
600
400
P2-PWUPT
tension-side wall
layer 6
200
0
0

1000
1000
800
wp7
(MPa)
1000
wp3
(MPa)
1000
600
400
P2-PWUPT
tension-side wall
layer 7
200
800
600
400
P2-PWUPT
tension-side wall
layer 8
200
0
0
wall pier PT bar strain,wp7 (x10-3)
Fig. 13.4a Wall pier post-tensioning bar stress-strain behaviors

for Structure P2-PWUPT: tension-side wall
389
600
400
P2-PWUPT
layer 1
200
0
1
wp1
400
P2-PWUPT
layer 2
200
3
wp2
(x10-3)
1000
(MPa)
4
800
wp4
800
600
400
P2-PWUPT
layer 3
200
600
400
P2-PWUPT
layer 4
200
0
0
wp3
(x10-3)
wp4
(x10-3)
1000
(MPa)
1000
800
wp6
800
600
400
P2-PWUPT
layer 5
200
600
400
P2-PWUPT
layer 6
200
0
0

1000
800
(MPa)
1000
600
400
P2-PWUPT
layer 7
200
800
wp8
(x10-3)
1000
600
0
0
800
wp2
800
(MPa)
1000
1. . . . . .
wp1
(MPa)
1000
600
400
P2-PWUPT
layer 8
200
0
0
wp7
(x10-3)
wp8
Fig. 13.4b Wall pier post-tensioning bar stress-strain behaviors

for Structure P2-PWUPT: compression-side wall
390
(x10-3)
8000
6000
wall pier base axial forces N
tw
and N
cw
(kN)
10000
4000
P1-CWUPT
2000
d=0.71%
s=1.67%
-2000
-4000
0.5
1
1.5
(a)
30000
25000
wall pier base axial forces N
tw
and N
cw
(kN)
35000
20000
d=0.79%
s=1.71%
15000
10000
5000
0
0
P2-PWUPT
0.5
1.5
(b)
Fig. 13.5 Wall pier base axial forces: (a) Structure P1-CWUPT; (b) Structure P2-PWUPT
391
10000
8000
P1-CWUPT
wall pier base axial force components

Ncwg and N cwb (kN)

Ntwg and Ntwb (kN)
10000
6000
4000
Ntwg
2000
d=0.71%
0
-2000
s=1.67%
Ntw
-4000
Ntwb
-6000
-8000
0.4
0.8
1.2
1.6
Ncwb
4000
Ncwg
2000
0
d=0.71%
-2000
-4000
s=1.67%
P1-CWUPT
-6000
0
0.4
(a)
0.8
1.2
1.6
2.0
35000
P2-PWUPT
30000
25000
d=0.79%
Ntwp

N cwg , N cwp , and N cwb (kN)

N twg , N twp , and N twb (kN)
Ncw
6000
-8000
2.0
35000
s=1.71%
20000
15000
Ntw
10000
Ntwg
5000
0
-5000
-10000
0
8000
Ntwb
0.5
1
1.5
30000
Ncwp
20000
15000
Ncwb
10000
5000
Ncwg
0
-5000
-10000
Ncw
25000
P2-PWUPT
0
0.5
d=0.79%
s=1.71%
1
1.5
(b)
Fig. 13.6 Components of wall pier base axial forces:
For Structure P1-CWUPT, the axial force at the base of the tension-side wall
reverses direction and goes into tension at a roof drift of approximately =0.04%, while
the wall pier axial forces in Structure P2-PWUPT remain compressive throughout the
analysis as a result of the wall post-tensioning forces. A small increase in the axial force
of the tension-side wall of Structure P2-PWUPT is observed after a roof drift of,
approximately, =0.5% because of the increase in the wall post-tensioning bar forces due
to increased gap opening at the wall-to-foundation horizontal joint.
As shown in Table 13.2, the wall pier base axial forces, Ntw and Ncw, can be
estimated with good accuracy using the idealized relationships and procedures from
Chapters 5, 8, 10, and 11.
13.2.4 Wall Pier Base Shear Forces
The thin solid lines in Figs. 13.7a and 13.7b show the shear forces, Ftw and Fcw, at
the bases of the tension-side and compression-side wall piers in Structures P1-CWUPT
392
and P2-PWUPT, respectively. The vertical lines show the design-level and survival-level
roof drift demands, d and s, from Chapter 12. The markers shown in Figs. 13.7a and
13.7b are the same as the markers in Figs. 13.1a and 13.1b, respectively.
Similar to the results in Chapter 7, the base shear force versus roof drift
relationships for the tension-side and compression-side walls do not pass through the
origin since shear forces develop in the wall piers upon the application of the coupling
beam post-tensioning forces, with no external lateral loads.
The sum of the tension-side and compression-side wall base shear forces, Ftw+Fcw,
shown using the thick solid lines in Fig. 13.7, is approximately equal to the total coupled
wall base shear force, F from the applied lateral forces. The corresponding Fcw/F ratios
for the prototype structures are depicted using the solid lines in Fig. 13.8 and show that,
in both structures, most of the coupled wall base shear force is resisted by the
compression-side wall (roughly 75% at large roof drift values), as a result of the larger
compressive axial forces that develop as compared with the smaller axial forces in the
tension-side wall. This will be discussed further in Chapter 15 based on the dynamic
analysis results of the prototype structures.
Note that since Fcw and Ftw can not be directly obtained in design, the proposed
design approach in Chapter 12 uses the ratio (Mcw+Mwb)/(Mtw+Mcw+Mwb) to estimate the
distribution of the total coupled wall base shear force demand to the compression-side
wall pier, where, Mcw and Mtw are the compression-side and tension-side wall base
moments and Mwb is the contribution of the coupling beam shear forces to the total
coupled wall base moment. Fig. 13.8 shows that the ratio used in design (dashed lines) is
reasonably close to the Fcw/F ratio from the fiber element model at large roof drift levels.
13.2.5 Wall Pier Base Moments
Figs. 13.9a and 13.9b show the contributions of the tension-side wall base moment,
Mtw, the compression-side wall base moment, Mcw, and the base moment due to the
coupling force couple, Mwb, to the total coupled wall base moment resistance, Mw of
Structures P1-CWUPT and P2-PWUPT, respectively. The vertical lines represent the
design-level and survival-level roof drift demands, d and s, respectively, from Chapter
12. The markers shown in Figs. 13.9a and 13.9b are the same as the markers in Figs.
13.1a and 13.1b, respectively.
Similar to the base shear forces in Fig. 13.7, the base moment versus roof drift
relationships for the tension-side and compression-side walls do not pass through the
origin since bending moments develop in the wall piers upon the application of the
coupling beam post-tensioning forces, with no external lateral loads. It is observed that
the compression-side wall has a larger contribution to the coupled wall base moment than
the contribution of the tension-side wall, as a result of the larger compressive axial forces
that develop in the compression-side wall. The base moment resistances of the tensionside and compression-side walls, Mtw and Mcw, in both structures increase with increasing
roof drift regardless of the wall pier axial forces.
393
8000
wall base shear forces F, Ftw , and Fcw (kN)
7000
coupled wall base shear force, F
6000
5000
base shear force Fcw
4000
d=0.71%
3000
s=1.67%
2000
1000
tension-side wall
base shear force Ftw
P1-CWUPT
0
0
0.2
0.4
0.6
0.8
1.2
1.4
1.6
1.8
1.8
(a)
8000
7000
wall base shear forces F, Ftw , and Fcw (kN)
coupled wall base shear force, F

6000
base shear force Fcw
5000
4000
3000
d=0.79%
s=1.71%
2000
tension-side wall
base shear force Ftw
1000
P2-PWUPT
0
0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
(b)
Fig. 13.7 Wall base shear forces: (a) Structure P1-CWUPT; (b) Structure P2-PWUPT
394
distribution of wall base shear forces
(Mcw+Mwb)/(Mcw+Mtw+Mwb)
0.8
0.6
Fcw/F
0.4
d=0.71%
s=1.67%
0.2
P1-CWUPT
0
0.4
0.8
1.2
1.6
(a)
distribution of wall base shear forces
(Mcw+Mwb)/(Mcw+Mtw+Mwb)
0.8
0.6
Fcw/F
0.4
d=0.79%
s=1.71%
0.2
P2-PWUPT
0
0.4
0.8
1.2
1.6
(b)
Fig. 13.8 Ratio of compression-side wall base shear force to total coupled wall base shear
force: (a) Structure P1-CWUPT; (b) Structure P2-PWUPT
395
180000
Mw , M tw , Mcw , Mwb , and 2Mw,unc (kN.m)
wall base moments
P1-CWUPT
Mw
140000
2Mw, unc
100000
d=0.71%
s=1.67%
Mcw
60000
Mwb
Mtw
20000
0
0
0.4
0.8
1.2
1.6
roof drift, and w,unc (percent)
(a)
180000
Mw , M tw , Mcw , Mwb , and 2Mw,unc (kN.m)
wall base moments
P2-PWUPT
Mw
140000
s=1.71%
d=0.79%
100000
2Mw, unc
Mcw
60000
Mwb
Mtw
20000
0
0
0.4
0.8
1.2
1.6
roof drift, and w,unc (percent)
(b)
Fig. 13.9 Wall base moments: (a) Structure P1-CWUPT; (b) Structure P2-PWUPT
396
As shown in Table 13.2, the wall base moments, Mw, Mtw, and Mcw can be
estimated with good accuracy using the idealized relationships and procedures from
Chapters 5, 8, 10, and 11.
The dashed lines in Fig. 13.9a and 13.9b show the sum of the base moment
resistances, 2Mw,unc, of the left-side and right-side wall piers without coupling (i.e., two
wall piers with no coupling beams) for Structures P1-CWUPT and P2-PWUPT,
respectively. Comparing 2Mw,unc with the total coupled wall base moment, Mw, there is a
significant increase in the lateral stiffness and strength of the walls as a result of coupling.
The degree of coupling, DOC for the structures, calculated using Equation (2.2), is shown
in Fig. 13.10. The vertical lines show the design-level and survival-level roof drift
demands, d and s, respectively, from Chapter 12, and the horizontal lines show the
target degree of coupling DOC=35% used in the design of the structures. The markers
shown in Figs. 13.10a and 13.10b are the same as the markers in Figs. 13.1a and 13.1b,
respectively.
It is observed that, in the initial range of response (<~0.3%), the coupling degree
values for the prototype structures are larger than the target value used in design;
however, as the structures are displaced laterally into the nonlinear range, the coupling
degree decreases. For Structure P1-CWUPT, the coupling degree, DOC reaches the target
value at approximately =0.37% and drops to a value of DOC=28% when s is reached.
For Structure P2-PWUPT, the coupling degree is very close to the target value for
>0.3%. The larger decrease in the DOC value of Structure P1-CWUPT as compared
with the DOC value of Structure P2-PWUPT as the structures are displaced into the
nonlinear range may be because of the larger strain hardening that occurs in the flexural
reinforcement of the monolithic cast-in-place reinforced concrete wall piers in Structure
P1-CWUPT.
As shown in Table 13.2, the coupling degree of the walls, DOC can be estimated
with good accuracy using the idealized relationships and procedures from Chapters 5, 8,
10, and 11.
13.2.7 Wall Pier Base Concrete Strains
Figs. 13.11a and 13.11b show the neutral axis (i.e., contact) depths, ctw and ccw
(measured from the compression corners of the wall piers) at the bases of the tension-side
and compression-side wall piers in Structures P1-CWUPT and P2-PWUPT, respectively.
The ctw and ccw values are normalized with respect to the wall pier length, lw=4.88 m. The
corresponding extreme confined concrete compression strains, tce and cce of the tensionside and compression-side wall piers at the base are shown in Figs. 13.12a and 13.12b,
respectively, with the confined concrete crushing strains, ccu shown by the dashed
horizontal lines. For each structure, the markers in Figs. 13.11 and 13.12 are the same as
the markers in Fig. 13.1.
397
100
P1-CWUPT
80
60
d=0.71%
s=1.67%
40
DOC=35%
20
0.4
0.8
1.2
1.6
(a)
100
P2-PWUPT
80
60
s=1.71%
d=0.79%
40
DOC=35%
20
0.4
0.8
1.2
1.6
(b)
Fig. 13.10 Degree of coupling: (a) Structure P1-CWUPT; (b) Structure P2-PWUPT
398
0.8
P1-CWUPT
0.6
d=0.71%
d=1.67%
ccw / lw
0.4
0.2
0.4
0.8
1.2
1.6
(a)
1
0.8
P2-PWUPT
ccw / lw
0.4
0.2
s=1.71%
d=0.79%
0.6
0.4
0.8
1.2
1.6
(b)
Fig. 13.11 Wall pier base neutral axis (i.e., contact) depths:
399
tce
wall pier base extreme confined
concrete compression strain, tce and cce
-0.005
cce
P1-CWUPT
-0.01
ccu = 0.012
-0.015
-0.02
d=0.71%
s=1.67%
-0.025
-0.03
-0.035
0
0.4
0.8
1.2
1.6
(a)

concrete compression strain, tce and cce
-0.005
tce
-0.01
-0.015
d=0.79%
s=1.71%
-0.02
cce
P2-PWUPT
-0.025
-0.03
-0.035
0
ccu = 0.032
0.4
0.8
1.2
1.6
(b)
Fig. 13.12 Wall pier base extreme confined concrete compression strains:
400
The vertical lines in Figs. 13.11 and 13.12 show the design-level and survival-level
roof drift demands, d and s, respectively, from Chapter 12. For both prototype
structures, the ctw and ccw values change little between the design and survival demand
levels and the ccw values for the compression-side wall piers are very close to 0.2lw as
assumed in design for Structure P2-PWUPT (see Chapter 12).
As required by design, crushing of the confined concrete at the bases of the wall
piers in the prototype structures does not occur under the survival-level demand. Similar
to the results in Chapter 7, the compression-side wall piers have larger concrete
compression strains than the tension-side wall piers as a result of the larger compression
forces due to the coupling effect. Furthermore, the wall pier compression strains in
Structure P2-PWUPT are significantly larger than the strains in Structure P1-CWUPT
due to the additional compressive axial forces from the wall post-tensioning bars. This
results in a significantly larger amount of wall pier concrete confinement reinforcement at
the base of Structure P2-PWUPT as shown in Chapter 12.
As required by design, crushing of the confined concrete under the survival demand
level is prevented in both prototype structures, since the maximum concrete compression
strains at s are smaller than the corresponding confined concrete crushing strains, ccu.
Table 13.2 shows that the wall pier contact depths, ctw and ccw, and the coupled
wall roof drift at the wall ultimate state, wu can be estimated with good accuracy using
the idealized relationships and procedures from Chapters 5, 8, 10, and 11.
13.2.8 Coupling Beam Shear Force versus Chord Rotation Behaviors
The solid lines in Figs. 13.13a-13.13(h) show the shear force versus chord rotation
(Vb-b) behaviors of the eight coupling beams in Structure P1-CWUPT, respectively.
Similar results for the coupling beams in Structure P2-PWUPT are shown in Fig. 13.14.
The chord rotation values for the coupling beams are measured from a tangent drawn at
the left end of each beam as described in Chapter 7.
For comparison, the dashed lines in Figs. 13.13 and 13.14 show the coupling beam
shear force versus chord rotation behaviors of isolated subassemblages from Structures
P1-CWUPT and P2-PWUPT, respectively, and the horizontal lines show the design
coupling shear force demand, Qbd from Chapter 12. For both structures, the coupling
beams from the multi-story and subassemblage models have similar Vb-b behaviors
except for the second floor beam due to the effect of the fixed boundary conditions
assumed at the wall pier bases. The design coupling shear force demands are satisfied for
all of the coupling beams in the structures.
401
P1-CWUPT
2nd floor
1200
s=4.47%
(analysis)
d=1.61%
(analysis)
1000
800
600
1400
s=5.49%
(design)
Qbd=618 kN
400
multi-story
isolated
200
d=2.33%
(design)
1000
d=1.97%
(analysis)
800
600
Qbd=618 kN
400
multi-story
isolated
200
s=5.49%
(design)
P1-CWUPT
4th floor
1200
1400
d=2.21%
(analysis)
1000
800
600
Qbd=618 kN
400
d=2.33%
(design)
multi-story
isolated
200
s=5.32%
(analysis)
s=4.96%
(analysis)
1400
d=2.31%
(analysis)
1000
800
600
Qbd=618 kN
d=2.33%
(design)
400
multi-story
isolated
200
s=5.47%
(analysis)
1200
d=2.34%
(design)
1000
800
600
Qbd=618 kN
400
multi-story
isolated
200
d=2.33%
(analysis)
s=5.49%
(design)
P1-CWUPT
6th floor
s=5.49%
(design)
P1-CWUPT
5th floor
1200
1400
d=2.33%
(design)
s=5.52%
(analysis)
1400
s=5.49%
(design)
P1-CWUPT
7th floor
1200
d=2.34%
(design)
1000
800
600
Qbd=618 kN
400
multi-story
isolated
200
d=2.33%
(analysis)
s=5.53%
(analysis)
0
0
1400
s=5.49%
(design)
P1-CWUPT
8th floor
1200
d=2.33%
(analysis)
1000
800
600
Qbd=618 kN
d=2.33%
(design)
400
s=5.52%
(analysis)
multi-story
isolated
200
s=5.49%
(design)
P1-CWUPT
3rd floor
1200
1400
1400
s=5.49%
(design)
P1-CWUPT
roof
1200
d=2.32%
(analysis)
1000
800
600
Qbd=618 kN
d=2.33%
(design)
400
s=5.50%
(analysis)
multi-story
isolated
200
0
0
0
for Structure P1-CWUPT
402
multi-story
isolated
1200
s=5.62%
(design)
1000
800
1400
Qbd=654 kN
600
d=2.04%
400
(analysis)
d=2.59%
(design)
s=4.85%
(analysis)
P2-PWUPT
2nd floor
200
0
0
1400
1000
800
Qbd=654 kN
600
d=2.31%
(analysis)
400
P2-PWUPT
3rd floor
200
0
1400
s=5.62%
(design)
multi-story
isolated
1000
800
Qbd=654 kN
600
d=2.45%
(analysis)
400
d=2.59%
(design)
s=5.39%
(analysis)
P2-PWUPT
4th floor
200
0
0
Qbd=654 kN
600
d=2.56%
(analysis)
400
s=5.51%
(analysis)
d=2.59%
(design)
P2-PWUPT
6th floor
200
0
0
Qbd=654 kN
600
d=2.53%
(analysis)
400
Qbd=654 kN
600
d=2.53%
(analysis)
400
0
0
s=5.48%
(analysis)
d=2.59%
(design)
P2-PWUPT
8th floor
200
1400
s=5.62%
(design)
multi-story
isolated
1200
1000
800
Qbd=654 kN
600
d=2.55%
(analysis)
400
s=5.51%
(analysis)
d=2.59%
(design)
P2-PWUPT
7th floor
200
0
0
800
s=5.48%
(analysis)
s=5.62%
(design)
1000
d=2.59%
(design)
P2-PWUPT
5th floor
200
1400
1200
s=5.62%
(design)
800
multi-story
isolated
1000
800
s=5.62%
(design)
1000
multi-story
isolated
1200
1400
1200
s=5.21%
(analysis)
1400
multi-story
isolated
d=2.59%
(design)


1200
s=5.62%
(design)
multi-story
isolated
1200
1400
multi-story
isolated
1200
s=5.62%
(design)
1000
800
Qbd=654 kN
600
d=2.51%
(analysis)
400
0
0
s=5.46%
(analysis)
d=2.59%
(design)
P2-PWUPT
roof
200
for Structure P2-PWUPT
403
The dashed vertical lines in Figs. 13.13 and 13.14 show the design-level and
survival-level coupling beam chord rotation demands, d and s, respectively, from
Chapter 12 and the solid vertical lines show the beam chord rotation demands from the
nonlinear lateral load analysis of each structure (using the DRAIN-2DX fiber element
model) corresponding to the design-level and survival-level roof drift demands, d and s,
from Chapter 12. The results indicate that the beam chord rotation demands from design
(Chapter 12) are close to or larger than (resulting in conservative estimates) the rotation
demands from the nonlinear analyses of the structures, verifying the idealized structure
deflected shape in Fig. 12.10. As described previously, yielding of the beam posttensioning tendons is prevented under the survival level demand.
Table 13.2 shows that the average coupling beam shear forces, Vb and the average
beam chord rotation values, b over the height of each structure can be estimated with
good accuracy using the idealized relationships and procedures from Chapters 5, 8, 10,
and 11. Note that the b values at the wall softening state are not estimated well since the
initial linear elastic stiffness of the structure is used in the estimation.
13.2.9 Coupling Beam Axial Forces and Post-Tensioning Forces
The solid lines in Figs. 13.15 and 13.16 show the midspan axial force versus
chord rotation (Nb-b) behaviors of the coupling beams in Structures P1-CWUPT and P2PWUPT, respectively. A positive beam axial force indicates a compressive force. For
comparison, the dashed lines show the total forces in the post-tensioning tendons, Pb of
each coupling beam, the horizontal lines show the total yield strength of the posttensioning tendons, Pby, and the vertical lines show the design-level and survival-level
beam chord rotation demands, d and s, respectively, from the nonlinear analysis of each
structure (using the DRAIN-2DX model) corresponding to the design-level and survivallevel roof drift demands, d and s, from Chapter 12.
As discussed in detail in Chapter 7, the axial force in the second floor coupling
beam is significantly different than the total post-tensioning tendon force because of the
fixed conditions assumed at the bases of the wall piers. The axial forces in the upper level
beams are close to the post-tensioning tendon forces, indicating that the effect of the
foundation conditions quickly diminishes above the base.
404
5000
coupling beam midspan axial

force, N band PT force, Pb (kN)

5000
Pby=3784 kN
4000
Nb
3000
Pb
2000
P1-CWUPT d=1.61%
2nd floor (analysis)
1000
0
s=4.47%
(analysis)
2
3
4


Pby=3784 kN
4000
Nb
3000
Pb
P1-CWUPT
4th floor
1000
s=5.32%
(analysis)
d=2.21%
(analysis)
2000
2
3
4
s=4.96%
(analysis)
2
3
4
4000
3000
Pby=3784 kN
Nb
Pb
2000
1000
Pby=3784 kN
4000
Nb
3000
Pb
2000
s=5.52%
(analysis)
d=2.34%
(analysis)
P1-CWUPT
6th floor
2
P1-CWUPT
5th floor
1000
2
3
4
Nb
3000
Pb
2000
Pby=3784 kN
3000
Pb
Nb
2000
P1-CWUPT
8th floor
1
d=2.33%
(analysis)
2
3
4
d=2.34%
(analysis)
P1-CWUPT
7th floor
1000
s=5.53%
(analysis)

4000
0
0
s=5.47%
(analysis)
Pby=3784 kN
4000
5000
1000
d=2.31%
(analysis)
5000



d=1.97%
(analysis)
P1-CWUPT
3rd floor
1000
5000
Pb
2000
5000
5000
Nb
3000
Pby=3784 kN
4000
s=5.52%
(analysis)
5000
Pby=3784 kN
4000
Pb
3000
Nb
d=2.32%
(analysis)
2000
P1-CWUPT
roof
1000
0
s=5.50%
(analysis)
Fig. 13.15 Coupling beam axial forces and post-tensioning forces

in Structure P1-CWUPT
405
5000
P2-PWUPT
2nd floor
4000

force,N b and PT force, Pb (kN)

5000
s=4.85%
(analysis)
d=2.04%
(analysis)
Nb
3000
Pby=1892 kN
2000
Pb
1000
0
1
2
3
4
Pby=1892 kN
Nb
2000
Pb
1000
2
3
4
5000
P2-PWUPT
4th floor
4000
d=2.45%
(analysis)
3000
s=5.39%
(analysis)
Nb
Pby=1892 kN
2000


d=2.31%
(analysis)
3000
5000
Pb
1000
P2-PWUPT
5th floor
4000
d=2.53%
(analysis)
3000
s=5.48%
(analysis)
Pby=1892 kN
2000
Nb
Pb
1000
0
0
2
3
4
2
3
4
5000

5000

s=5.21%
(analysis)
0
0
P2-PWUPT
6th floor
4000
d=2.56%
(analysis)
3000
s=5.51%
(analysis)
Pby=1892 kN
2000
Nb
Pb
1000
P2-PWUPT
7th floor
4000
d=2.55%
(analysis)
3000
s=5.51%
(analysis)
Pby=1892 kN
2000
Pb
Nb
1000
0
0
2
3
4
2
3
4
5000

5000

P2-PWUPT
3rd floor
4000
P2-PWUPT
8th floor
4000
d=2.53%
(analysis)
3000
s=5.48%
(analysis)
Pby=1892 kN
2000
Pb
Nb
1000
P2-PWUPT
roof
4000
d=2.51%
(analysis)
3000
s=5.46%
(analysis)
Pby=1892 kN
2000
Pb
Nb
1000
0
0
Fig. 13.16 Coupling beam axial forces and post-tensioning forces

in Structure P2-PWUPT
406

Figs. 13.17 and 13.18 show the neutral axis (i.e., contact) depths, cb (normalized
with respect to the beam depth, db=559 mm) at the left ends of the coupling beams in
Structures P1-CWUPT and P2-PWUPT, respectively. The behaviors at the right ends of
the coupling beams are similar. The vertical lines show the design-level and survivallevel beam chord rotation demands, d and s, respectively, from the nonlinear analysis of
each structure (using the DRAIN-2DX model) corresponding to the design-level and
survival-level roof drift demands, d and s, from Chapter 12 and the horizontal lines
represent the thickness of the beam flange, tbf. It is observed that the beam-to-wall contact
depths of the prototype structures remain within the flange (except for the second floor
beam of Structure P1-CWUPT) during most of the nonlinear response.
The corresponding extreme steel compression strains, be at the left ends of the
coupling beams are shown in Figs. 13.19 and 13.20, respectively. In general, the coupling
beams over the height of each structure have similar behaviors, with the largest strains
occurring in the second floor beam. The extreme compression strains in the beams of
Structure P2-PWUPT are significantly smaller than the strains in the beams of Structure
P1-PWUPT due to the smaller beam post-tensioning forces.
13.2.11 Tension Angle Force versus Deformation Behaviors
Figs. 13.21 and 13.22 show the force versus deformation behaviors of the tension
angles at the left ends of the coupling beams in Structures P1-CWUPT and P2-PWUPT,
respectively. The behaviors of the tension angles at the right ends of the coupling beams
are similar. The dashed vertical lines show the design-level and survival-level tension
angle deformation demands, uad and uas, from Chapter 12 and the solid vertical lines show
the tension angle deformation demands from the nonlinear lateral load analysis of each
structure (using the DRAIN-2DX fiber element model) corresponding to the design-level
and survival-level roof drift demands, d and s, from Chapter 12. The results indicate
that the tension angle deformation demands from design (Chapter 12) are reasonably
close to or larger than (resulting in conservative estimates) the angle deformation
demands from the nonlinear analyses of the structures, verifying the idealized beam end
displaced shape in Fig. 12.11.
For both structures, the tension angles at the second floor level have the smallest
deformations, because, the fixed boundary conditions assumed at the bases of the wall
piers restrain the opening of gaps at the beam-to-wall interfaces. The largest survivallevel tension angle deformations from the nonlinear analyses of Structures P1-CWUPT
and P2-PWUPT are 32 mm and 31 mm, respectively, which are both smaller than the
design limit of 38 mm from Chapter 12. The smaller angle forces in Structure P1CWUPT, as designed in Chapter 12 with the use of a small top and seat angle ratio
(ar=0.051), are noticeable in the plots.
407
P1-CWUPT
2nd floor
0.8
0.6
s=4.47%
(analysis)
d=1.61%
(analysis)
0.4
0.2
tbf=29 mm

depth, cb /beam depth, db

P1-CWUPT
3rd floor
0.8
d=1.97%
(analysis)
0.6
0.4
0.2
tbf=29 mm
0
P1-CWUPT
4th floor
0.8


d=2.21%
(analysis)
0.6
s=5.32%
(analysis)
0.4
0.2
tbf=29 mm
d=2.31%
(analysis)
P1-CWUPT
5th floor
0.8
0.6
s=5.47%
(analysis)
0.4
0.2
tbf=29 mm
0
0

1


P1-CWUPT
6th floor
0.8
0.6
d=2.33%
(analysis)
s=5.52%
(analysis)
0.4
0.2
tbf=29 mm
P1-CWUPT
7th floor
0.8
0.6
d=2.33%
(analysis)
s=5.53%
(analysis)
0.4
0.2
tbf=29 mm
0
0

1
P1-CWUPT
8th floor
0.8


s=4.96%
(analysis)
d=2.33%
(analysis)
0.6
s=5.52%
(analysis)
0.4
0.2
tbf=29 mm
0
P1-CWUPT
roof
0.8
d=2.32%
(analysis)
0.6
s=5.50%
(analysis)
0.4
0.2
tbf=29 mm
0
Fig. 13.17 Coupling beam end neutral axis (i.e., contact) depths for Structure P1-CWUPT
408


P2-PWUPT
2nd floor
0.8
0.6
s=4.85%
(analysis)
d=2.04%
(analysis)
0.4
0.2
tbf=29 mm
0
P2-PWUPT
3rd floor
0.8
0.6
d=2.31%
(analysis)
0.2
tbf=29 mm
0


P2-PWUPT
4th floor
0.8
0.6
s=5.39%
(analysis)
d=2.45%
(analysis)
0.4
0.2
tbf=29 mm
P2-PWUPT
5th floor
0.8
0.6
d=2.53%
(analysis)
0.4
0.2
s=5.48%
(analysis)
tbf=29 mm
0
0

1


P2-PWUPT
6th floor
0.8
0.6
s=5.51%
(analysis)
d=2.56%
(analysis)
0.4
0.2
tbf=29 mm
P2-PWUPT
7th floor
0.8
0.6
d=2.55%
(analysis)
s=5.51%
(analysis)
0.4
0.2
tbf=29 mm
0
0

1


s=5.21%
(analysis)
0.4
P2-PWUPT
8th floor
0.8
d=2.53%
(analysis)
0.6
s=5.48%
(analysis)
0.4
0.2
tbf=29 mm
P2-PWUPT
roof
0.8
d=2.51%
(analysis)
0.6
s=5.46%
(analysis)
0.4
0.2
tbf=29 mm
0
0
Fig. 13.18 Coupling beam end neutral axis (i.e., contact) depths for Structure P2-PWUPT
409


0
-0.02
-0.04
s=4.47%
(analysis)
-0.06
d=1.61%
(analysis)
-0.08
-0.1
P1-CWUPT
2nd floor
-0.12
0
-0.06
d=1.97%
(analysis)
-0.08
-0.02
-0.04
d=2.21%
(analysis)
-0.06
-0.08
-0.1
s=5.32%
(analysis)
P1-CWUPT
4th floor
-0.12
0
-0.02
-0.04
d=2.31%
(analysis)
-0.06
-0.08
s=5.47%
(analysis)
P1-CWUPT
5th floor
-0.1
-0.12

0
-0.02
-0.04
d=2.34%
(analysis)
-0.06
-0.08
s=5.52%
(analysis)
P1-CWUPT
6th floor
-0.1
-0.02
-0.04
d=2.34%
(analysis)
-0.06
-0.08
s=5.53%
(analysis)
P1-CWUPT
7th floor
-0.1
-0.12
-0.12
0

0


s=4.96%
(analysis)
P1-CWUPT
3rd floor
-0.1


-0.04
-0.12
2

-0.02
-0.02
-0.04
d=2.33%
-0.06
(analysis)
-0.08
s=5.52%
(analysis)
P1-CWUPT
8th floor
-0.1
-0.02
-0.04
d=2.32%
(analysis)
-0.06
-0.08
s=5.50%
(analysis)
P1-CWUPT
roof
-0.1
-0.12
-0.12
0
Fig. 13.19 Coupling beam end extreme steel compression strains

410

compressive strain, be

0
-0.02
-0.04
-0.06
s=4.85%
(analysis)
d=2.04%
(analysis)
-0.08
-0.1
P2-PWUPT
2nd floor
-0.12
0
d=2.31%
(analysis)
P2-PWUPT
3rd floor
-0.1
-0.02
-0.04
-0.06
-0.08
s=5.39%
(analysis)
d=2.45%
(analysis)
P2-PWUPT
4th floor
-0.02
-0.04
-0.06
-0.08
d=2.53%
(analysis)
P2-PWUPT
5th floor
-0.1
s=5.48%
(analysis)
-0.12
0

-0.02
-0.04
-0.06
-0.08
d=2.56%
(analysis)
P2-PWUPT
6th floor
-0.1
s=5.51%
(analysis)
-0.02
-0.04
-0.06
-0.08
d=2.55%
(analysis)
P2-PWUPT
7th floor
-0.1
s=5.51%
(analysis)
-0.12
-0.12
0

-0.02
-0.04
-0.06
-0.08
s=5.48%
(analysis)
d=2.53%
(analysis)
P2-PWUPT
8th floor
-0.1

s=5.21%
(analysis)
-0.08


-0.06
-0.12

-0.04
-0.12
2
-0.1
-0.02
-0.12
-0.02
-0.04
-0.06
-0.08
d=2.51%
(analysis)
P2-PWUPT
roof
-0.1
s=5.46%
(analysis)
-0.12
0
Fig. 13.20 Coupling beam end extreme steel compression strains

411
150
uas =
30 mm
(design)
P1-CWUPT
2nd floor
150
100
uad =
13 mm
(design) u =
uad =
as
6 mm
19 mm
(analysis)
(analysis)
50
100
50
uad =
9 mm
(analysis)
0
0
10
15
20
25
30
35

150
P1-CWUPT
4th floor
uas =
30 mm
(design)
100
uad =
13 mm
(design)
50
uas =
29 mm
(analysis)
uad =
12 mm
(analysis)
20
P1-CWUPT
5th floor
25
30
35
uas =
30 mm
(design)
100
uad =
13 mm
(design)
50
uas =
30 mm
(analysis)
uad =
12 mm
(analysis)
10
15
20
25
30
35

150
50
uad =
13 mm
(analysis)
15
20
150
100
uad =
13 mm
(design)
10
25
30
35
uas =
30 mm
(design)
P1-CWUPT
6th floor
15
0
0
uas =
31 mm
(analysis)
uas =
30 mm
(design)
P1-CWUPT
7th floor
100
50
uad =
13 mm
(analysis)
uad =
13 mm
(design)
uas =
31 mm
(analysis)
0
0
10
15
20
25
30
35

150
50
uad =
13 mm
(analysis)
15
20
150
100
uad =
13 mm
(design)
10
25
30
35
uas =
30 mm
(design)
P1-CWUPT
8th floor
10
uad =
13 mm
(design)
uas =
25 mm
(analysis)
150
uas =
30 mm
(design)
P1-CWUPT
3rd floor
uas =
31 mm
(analysis)
uas =
30 mm
(design)
P1-CWUPT
roof
100
50
uad =
13 mm
(analysis)
uad =
13 mm
(design)
uas =
31 mm
(analysis)
0
0
10
15
20
25
30
35
10
15
20
25
Fig. 13.21 Coupling beam tension angle force-deformation behaviors

412
30
35
1200
P2-PWUPT
2nd floor
1000
uad =
14 mm
(design)
800
600
uad =
8 mm
(analysis)
400
200
uas =
31 mm
(design)
1200
uas =
23 mm
(analysis)
P2-PWUPT
3rd floor
1000
800
uad =
14 mm
(design)
600
400
0
0
10
15
20
25
30
35
1000
800
P2-PWUPT
4th floor
600
uad =
14 mm
(design)
uas =
29 mm
(analysis)
uad =
13 mm
(analysis)
400
200
uas =
31 mm
(design)
15
20
25
30
35
30
35
1200
1000
800
P2-PWUPT
5th floor
600
uad =
14 mm
(design)
uad =
13 mm
(analysis)
400
200
uas =
31 mm
(design)
uas =
30 mm
(analysis)
0
0
10
15
20
25
30
35
10
15
20
25

1200
1000
800
P2-PWUPT
6th floor
600
uad =
14 mm
(design)
200
uas =
31 mm
(design)
uas =
31 mm
(analysis)
uad =
14 mm
(analysis)
400
1200
10
1200
1000
800
P2-PWUPT
7th floor
600
uad =
14 mm
(design)
uad =
14 mm
(analysis)
400
200
uas =
31 mm
(design)
uas =
31 mm
(analysis)
0
0
10
15
20
25
30
35
10
15
20
25
30
35

1200
1000
800
P2-PWUPT
8th floor
600
400
uad =
14 mm
(design)
uas =
31 mm
(design)
uas =
31 mm
(analysis)
uad =
14 mm
(analysis)
200
1200
uas =
28 mm
(analysis)
uad =
11 mm
(analysis)
200
uas =
31 mm
(design)
1000
800
P2-PWUPT
roof
600
400
uad =
14 mm
(design)
uas =
30 mm
(analysis)
uad =
13 mm
(analysis)
200
uas =
31 mm
(design)
0
0
10
15
20
25
30
35
10
15
20
25
Fig. 13.22 Coupling beam tension angle force-deformation behaviors

413
30
35
13.3 Behavior Under Reversed Cyclic Loading

This section describes the behavior of Structures P1-CWUPT and P2-PWUPT
under reversed cyclic lateral loads combined with gravity loads. The structures are
displaced to roof drift values of =0.5, 0.75, 1, 1.25, 1.5, and 1.75%, with one cycle of
loading at each displacement amplitude. Note that the largest roof drift of 1.75% is
slightly larger than the survival-level roof drift demand, s for both structures. The
section is organized as follows: (1) coupled wall base shear force versus roof drift
behaviors; (2) wall pier post-tensioning forces in Structure P2-PWUPT; (3) wall pier base
axial forces; (4) wall pier base shear forces; (5) wall pier base moments; (6) degree of
coupling; (7) wall pier base concrete strains; (8) coupling beam shear force versus chord
rotation behaviors; (9) coupling beam axial forces and post-tensioning forces; (10)
coupling beam end strains; and (11) angle force versus deformation behaviors.
13.3.1 Coupled Wall Base Shear Force versus Roof Drift Behaviors
Figs. 13.23a and 13.23b show the coupled wall base shear force versus roof drift
(F-) behaviors from the reversed cyclic lateral load analyses of Structures P1-CWUPT
and P2-PWUPT, respectively. For comparison, the behaviors from the monotonic lateral
load analyses in Fig. 13.1 are shown using the thick solid lines.
The results in Fig. 13.23 indicate that Structure P1-CWUPT has a larger amount of
inelastic energy dissipation than Structure P2-PWUPT, which occurs as a result of the
yielding of the mild steel reinforcement at the base of the wall piers. In contrast, the
unbonded post-tensioning bars used as flexural reinforcement in the precast concrete wall
piers of Structure P2-PWUPT provide little inelastic energy dissipation; since the
yielding of the bars is significantly reduced as a result of the use of unbonded bars (see
Figs. 13.3 and 13.4).
It is also observed that Structure P1-CWUPT has a smaller amount of selfcentering capability than Structure P2-PWUPT. As an example, the residual roof drifts of
Structures P1-CWUPT and P2-PWUPT upon unloading from a roof drift of 1.75% are
equal to 0.49% and 0.10%, respectively. The increased self-centering capability of
Structure P2-PWUPT occurs as a result of the additional restoring forces provided by the
post-tensioning bars used as flexural reinforcement in the precast concrete wall piers.
Note that the residual roof drifts of Structure P1-CWUPT in Fig. 13.23a are larger
than the residual roof drifts of the CIP-UPT system in Fig. 7.23a. This is because,
Structure P1-CWUPT has a much smaller coupling degree (DOC35%) than the CIPUPT system (DOC60%), and thus, the beam post-tensioning tendons provide a smaller
amount of restoring force to pull the wall piers back towards their original undisplaced
position. Note that the hysteresis loop of Structure P1-CWUPT in Fig. 13.23a appears
more like the one in Fig. 7.23d (without top and seat angles) than Fig. 7.23a (with top and
seat angles). This is possibly because the top-and-seat angles in Structure P1-CWUPT are
designed to have relatively small angle tension yield strength (53 kN).
414
8000
6000
Qwd=4782kN
4000
2000
P1-CWUPT
0
-2000
Qwd=4782kN
-4000
-6000
-8000
-2
s=1.67%
d=0.71%
-1.5
-1
-0.5
0.5
1.5
(a)
8000
6000
Qwd=4893kN
4000
P2-PWUPT
2000
0
-2000
-4000
Qwd=4893kN
-6000
d=0.79%
-8000
-2
-1.5
-1
-0.5
0.5
s=1.71%
1
1.5
(b)
Fig. 13.23 Cyclic coupled wall base shear force versus roof drift behaviors:
415
The vertical dashed lines in Fig. 13.23 show the design-level and survival-level
roof drift demands, d and s, respectively, and the horizontal dashed lines show the
coupled wall base shear demands, Qwd from Chapter 12. As expected, the base shear
capacities of the prototype structures satisfy the Qwd demands.
13.3.2 Wall Pier Post-Tensioning Forces in Structure P2-PWUPT
Figs. 13.24a and 13.24b show the total force in the post-tensioning bars of the leftside and right-side wall piers, respectively, from the reversed cyclic lateral load analysis
of Structure P2-PWUPT and Fig. 13.24c shows the total coupled wall post-tensioning
force as the sum of the left-side and right-side wall pier post-tensioning bar forces. For
comparison, the thick solid lines represent the behavior from the monotonic lateral load
analysis of the structure and the vertical dashed lines show the design-level and survivallevel rood drift demands, d and s, respectively, from Chapter 12.
The results in Fig. 13.24 show that the variations in the wall pier post-tensioning
bar forces during the cyclic lateral displacements of the structure are small as a result of
the use of unbonded bars. As described previously, a small amount of yielding in the wall
pier post-tensioning bars occurs when the structure is displaced beyond the PT-yielding
state. The yielding of the wall pier post-tensioning bars results in a small amount of loss
in the post-tensioning bar forces under cyclic loading; however, this loss is mostly
negligible. Thus, it is concluded that the initial post-tensioning forces in the wall piers of
Structure P2-PWUPT are mostly maintained throughout the lateral displacement history.
13.3.3 Wall Pier Base Axial Forces
Figs. 13.25a and 13.25b show the axial forces at the bases of the left-side and rightside wall piers, respectively, from the reversed cyclic lateral load analysis of Structure
P1-CWUPT. A positive wall pier axial force indicates a compressive force. For
comparison, the behaviors from the monotonic analysis results in Fig. 13.5 are shown
using the thick solid lines, and the vertical dashed lines show the design-level and
survival-level roof drift demands, d and s, respectively, from Chapter 12.
The thick solid line in Fig. 13.25c shows the total coupled wall base axial force (i.e.,
sum of the left-side and right-side wall pier base axial forces) plotted against the cyclic
loading duration. The left-side and right-side wall pier base axial forces (thin solid and
dashed lines, respectively) are also plotted in Fig. 13.25c, with markers showing the
peak roof drifts during each displacement cycle. Since the coupling-induced axial forces
in the left-side and right-side wall piers cancel each other, the total coupled wall axial
force in Structure P1-CWUPT is equal to the gravity loading, and thus, remains constant
during the cyclic analysis.
416
25000
wall pier PT force (kN)
20000
15000
P2-PWUPT
left side wall
d=0.79%
10000
s=1.71%
5000
0
-2
-1
0
1
(a)
25000
wall pier PT force (kN)
20000
15000
P2-PWUPT
right side wall
d=0.79%
10000
s=1.71%
5000
0
-2
-1
0
1
(b)
45000
total coupled wall PT force (kN)
40000
35000
30000
25000
d=0.79%
20000
s=1.71%
15000
10000
5000
0
-2
-1
0
1
(c)
Fig. 13.24 Cyclic wall post-tensioning forces for Structure P2-PWUPT:
(a) left-side wall pier; (b) right-side wall pier; (c) total coupled wall
417
10000
8000
wall pier base axial force (kN)
6000
d=0.71%
4000
s=1.67%
2000
0
-2000
P1-CWUPT
left-side wall
-4000
-6000
-8000
-10000
-2
-1.5
-1
-0.5
0.5
1.5
(a)
10000
8000
6000
4000
2000
0
-2000
d=0.71%
-4000
s=1.67%
P1-CWUPT
right-side wall
-6000
-8000
-10000
-2
-1.5
-1
-0.5
0.5
1.5
left wall, right wall, and total coupled wall base axial forces (kN)
(b)
10000
8000
6000
4000
2000
0
-2000
-4000
-6000
-8000
=0.5%
-0.5%
=0.75% -0.75% 1.0% -1.0% 1.25% -1.25%
=1.75% -1.75%
1.5% -1.5%
left-side wall base axial force
P1-CWUPT
right-side wall base axial force

sum of left-side and right-side wall base axial forces
-10000
(c)
Fig. 13.25 Cyclic wall base axial forces for Structure P1-CWUPT:
418
Similar plots for the wall pier base axial forces in Structure P2-PWUPT are given
in Figs. 13.26a, 13.26b, and 13.26c. Different from Structure P1-CWUPT, the left-side
and right-side wall pier axial forces in Structure P2-PWUPT remain compressive
throughout the analysis, as a result of the wall pier post-tensioning forces. Note also that
the total coupled wall base axial force in Structure P2-PWUPT (thick solid line in Fig.
13.26c) varies during the cyclic loading history, since the wall pier post-tensioning forces
vary as shown in Fig. 13.24.
13.3.4 Wall Pier Base Shear Forces
Figs. 13.27a and 13.27b show the shear forces at the bases of the left-side and
right-side wall piers, respectively, from the reversed cyclic lateral load analysis of
Structure P1-CWUPT. For comparison, the behaviors from the monotonic analysis results
in Fig. 13.7 are shown using the thick solid lines, and the vertical dashed lines show the
design-level and survival-level roof drift demands, d and s, respectively, from Chapter
12. Similar plots for the wall pier base shear forces in Structure P2-PWUPT are given in
Figs. 13.28a and 13.28b.
It is observed that in each structure, the base shear versus roof drift behavior of the
left-side wall pier under positive roof drift is similar to the behavior of the right-side wall
pier under negative roof drift and vice versa. The sum of the left-side and right-side wall
pier base shear forces is approximately equal to the total coupled wall base shear force F
in Fig. 13.23, which is symmetric in the positive and negative roof drift directions.
13.3.5 Wall Pier Base Moments
Figs. 13.29a and 13.29b show the bending moments at the bases of the left-side
and right-side wall piers, respectively, from the reversed cyclic lateral load analysis of
Structure P1-CWUPT. Similarly, Fig. 13.29c shows the contribution of the coupling
beam forces to the total coupled wall base moment, and Fig. 13.29d shows the total
coupled wall base moment, Mw (i.e., sum of the left-side wall pier base moment, rightside wall pier base moment, and base moment due to the coupling forces). For
comparison, the behaviors from the monotonic lateral load analysis results in Fig. 13.9
are shown using the thick solid lines, and the vertical dashed lines show the design-level
and survival-level roof drift demands, d and s, respectively, from Chapter 12. Similar
plots for the wall base moments in Structure P2-PWUPT are given in Fig. 13.30.
By inspecting the different contributions to the total coupled wall base moment in
Fig. 13.29, it is concluded that almost all of the inelastic energy dissipation in Structure
P1-CWUPT is provided by the monolithic cast-in-place reinforced concrete wall piers.
This is expected since a small value for the coupling beam top and seat angle ratio
(ar=0.051) is used in the design of Structure P1-CWUPT in Chapter 12, and thus, the
angle contribution to the coupling beam forces is small. In contrast, the results in Fig.
13.30 show that the coupling beams provide almost all of the inelastic energy dissipation
in Structure P2-PWUPT, since the energy dissipation from the left-side and right-side
wall piers is small and a large value for the coupling beam angle ratio (ar=1.22) is used
in the design of the structure in Chapter 12.
419
60000
50000
d=0.79%
P2-PWUPT
left-side wall
40000
s=1.71%
30000
20000
10000
0
-2
-1.5
-1
-0.5
0
0.5
1.5
(a)
60000
50000
P2-PWUPT
right-side wall
d=0.79%
s=1.71%
40000
30000
20000
10000
-2
-1.5
-1
-0.5
0.5
1.5
left wall, right wall, and total coupled wall base axial forces (kN)
(b)
60000
P2-PWUPT
50000
40000
1.5% -1.5% 1.75% -1.75%
1.0% -1.0% 1.25% -1.25%
=0.5% -0.5% 0.75% -0.75%
30000
20000
10000
left-side wall base axial force

right-side wall base axial force
sum of left-side and right-side wall base axial forces
0
(c)
Fig. 13.26 Cyclic wall base axial forces for Structure P2-PWUPT:
420
6000
8000
P1-CWUPT
left-side wall
wall pier base shear force (kN)
8000
4000
2000
0
d=0.71%
-2000
-4000
s=1.67%
-6000
-8000
-2
-1
0
1
4000
2000
0
-2000
d=0.71%
-4000
s=1.67%
-6000
-8000
P1-CWUPT
right-side wall
6000
-2
-1
(a)
0
1
(b)
Fig. 13.27 Cyclic wall pier base shear forces for Structure P1-CWUPT:
(a) left-side wall pier; (b) right-side wall pier
6000
8000
P2-PWUPT
left-side wall
8000
4000
2000
0
-2000
d=0.79%
-4000
s=1.71%
-6000
-8000
-2
-1
0
1
4000
2000
0
-2000
(a)
d=0.79%
-4000
s=1.71%
-6000
-8000
P2-PWUPT
right-side wall
6000
-2
-1
0
1
(b)
Fig. 13.28 Cyclic wall pier base shear forces for Structure P2-PWUPT:
(a) left-side wall pier; (b) right-side wall pier
421
200000
P1-CWUPT
left-side wall
100000
wall pier base moment (kN.m)
200000
0
d=0.71%
-100000
P1-CWUPT
right-side wall
100000
d=0.71%
-100000
s=1.67%
s=1.67%
-200000
-2
-1
0
-200000
-2
-1
(a)
total coupled wall base moment, Mw (kN.m)
coupled wall base moment due

to coupling beam forces (kN.m)
P1-CWUPT
100000
0
d=0.71%
s=1.67%
-200000
-2
-1
(b)
200000
-100000
0
0
(c)
200000
P1-CWUPT
100000
d=0.71%
-100000
s=1.67%
-200000
-2
-1
0
(d)
Fig. 13.29 Cyclic wall base moments for Structure P1-CWUPT: (a) left-side wall pier
base moment; (b) right-side wall pier base moment; (c) coupling base moment;
(d) total coupled wall base moment
422
200000
P2-PWUPT
left-side wall
100000
200000
d=0.79%
-100000
s=1.71%
-200000
-2
-1
0
d=0.79%
-100000
s=1.71%
-200000
P2-PWUPT
right-side wall
100000
-2
-1
0
(a)
total coupled wall base moment, Mw (kN.m)
coupled wall base moment due

to coupling beam forces (kN.m)
P2-PWUPT
100000
d=0.79%
s=1.71%
-200000
-2
-1
0
(c)
(b)
200000
-100000
200000
P2-PWUPT
100000
-100000
d=0.79%
s=1.71%
-200000
-2
-1
0
(d)
Fig. 13.30 Cyclic wall base moments for Structure P2-PWUPT: (a) left-side wall pier
base moment; (b) right-side wall pier base moment; (c) coupling base moment;
(d) total coupled wall base moment
Figs. 13.31a and 13.31b show the degree of coupling, DOC [using Equation (2.2)],
from the reversed cyclic lateral load analyses of Structures P1-CWUPT and P2-PWUPT,
respectively. For comparison, the behaviors from the monotonic analysis results in Fig.
13.10 are shown using the thick solid lines, and the vertical dashed lines show the designlevel and survival-level roof drift demands, d and s, respectively, from Chapter 12.
The results in Fig. 13.31 show that the degree of coupling shows great variation
during the cyclic lateral displacements of the structures. The monotonic DOC curve
provides a good representation of the envelope behavior from the cyclic analysis results.
It is observed that in between the peak displacements in the positive and negative
directions, the DOC values from the cyclic analyses can drop down to a value close to
zero and can also take a value greater than 100%. Values of DOC close to zero indicate
that the contribution of the coupling beam forces to the total coupled wall base moment
resistance is negligible at those instances in the cyclic analysis. Conversely, values of
423
DOC greater than 100% (not plotted in Fig. 13.31) indicate that the coupled wall base
moment due to the coupling beam forces is greater than the total coupled wall base
moment resistance, which means that the sum of the left-side and right-side wall pier
base moments has a negative contribution to the total coupled wall base moment at those
instances.
The horizontal lines in Fig. 13.31 show the target degree of coupling DOC=35%
used in the design of the structures in Chapter 12. In general, the envelope DOC values
from the cyclic analyses are reasonably close to the target value, especially for Structure
P2-PWUPT, as described in more detail based on the monotonic analysis results in
Section 13.2.6.
13.3.7 Wall Pier Base Concrete Strains
Figs. 13.32a and 13.32b show the neutral axis (i.e., contact) depths (measured from
the compression corners of the wall piers) at the bases of the left-side and right-side wall
piers, respectively, from the reversed cyclic lateral load analysis of Structure P1-CWUPT.
Note that the compression corners of the wall piers alternate between the right and left
corners as the structure is displaced cyclically. The contact (i.e., compression) depths are
normalized with respect to the wall length, lw=4.88 m and are plotted against the cyclic
loading duration. The corresponding extreme confined concrete compression strains at
the left and right ends of the wall piers are shown in Figs. 13.32c and 13.32d, respectively,
with the confined concrete crushing strains, ccu shown by the horizontal lines. Similar
plots for the wall pier base contact depths and extreme confined concrete strains in
Structure P2-PWUPT are given in Fig. 13.33.
Examining the results in Fig. 13.33, it is observed that the contact depths at the
bases of the wall piers in Structure P2-PWUPT remain larger than zero throughout the
entire loading duration, indicating that both wall piers maintain contact with the
foundation. This is because of the additional compressive wall pier axial forces in
Structure P2-PWUPT provided by the wall post-tensioning bars. Full contact at the bases
of the wall piers in Structure P2-PWUPT is indicated by the normalized contact depths in
Figs. 13.33a and 13.33b reaching a value of 1. In contrast, full contact of the wall piers in
Structure P1-CWUPT rarely occurs and full uplift of the tension-side wall pier (as
indicated by the normalized contact depths in Figs. 13.32a and 13.32b reaching a value of
0) is quite common, during which the entire lateral resistance of the wall pier is provided
by the flexural mild steel reinforcement at the base.
424
100
80
60
40
DOC=35%
d=0.71%
20
s= 1.67%
P1-CWUPT
0
(a)
100
80
P2-PWUPT
d=0.79%
60
s= 1.71%
40
DOC=35%
20
0
2
0
(b)
Fig. 13.31 Degree of coupling: (a) Structure P1-CWUPT; (b) Structure P2-PWUPT
425
wall pier base neutral axis depth/wall pier length
0.8
P1-CWUPT
left-side wall
0.6
0.4
0.2
0
0.8
0.6
0.4
0.2
0
(a)
(b)
0
-0.005
P1-CWUPT
left end of left-side wall
right end of left-side wall
ccu=-0.012

compression strain

compression strain
-0.01
P1-CWUPT
right-side wall
-0.005
-0.01
P1-CWUPT
left end of right-side wall
right end of right-side wall
ccu=-0.012
-0.015
-0.015
(c)
(d)
depth; (b) right-side wall pier contact depth; (c) left-side wall pier concrete strains;
(d) right-side wall pier concrete strains
Similar to the monotonic analysis results in Section 13.2.7, significantly larger
extreme confined concrete compression strains occur in the wall piers of Structure P2PWUPT as compared with Structure P1-CWUPT, due to the additional compressive wall
pier axial forces from the wall post-tensioning bars. Furthermore, in each wall pier, larger
compression strains occur at the outer end than the strains at the inner end (coupling
beam end) of the pier. As required by design, crushing of the confined concrete under the
survival demand level is prevented in both prototype structures, since the maximum
confined concrete compression strains are smaller than the corresponding confined
concrete crushing strains, ccu.
426
0.8
0.6
0.4
0.2
P2-PWUPT
left-side wall
0.8
0.6
0.4
0.2
P2-PWUPT
right-side wall
(a)
(b)
-0.02
P2-PWUPT
ccu=-0.032

compression strain

compression strain
-0.04
-0.02
P2-PWUPT
ccu=-0.032
-0.04
(c)
(d)
concrete strains for Structure P2-PWUPT: (a) left-side wall pier contact depth; (b) rightside wall pier contact depth; (c) left-side wall pier concrete strains;
(d) right-side wall pier concrete strains
13.3.8 Coupling Beam Shear Force versus Chord Rotation Behaviors
Figs. 13.34 and 13.35 show the cyclic shear force versus chord rotation (Vb-b)
behaviors of the coupling beams in Structures P1-CWUPT and P2-PWUPT, respectively.
For comparison, the behaviors from the monotonic analysis results in Figs. 13.13 and
13.14 are shown using the thick solid lines, and the vertical solid lines show the designlevel and survival-level beam chord rotation demands, d and s, respectively, from the
monotonic nonlinear analysis of each structure (using the DRAIN-2DX model)
corresponding to the design-level and survival-level roof drift demands, d and s, from
Chapter 12.
As expected, a larger amount of inelastic energy dissipation is provided by the
coupling beams in Structure P2-PWUPT, since a larger value of the top and seat angle
ratio (ar=1.22) is used in design. The differences between the behaviors of the second
427
1500
P1-CWUPT
2nd floor
1000
500
0
d=1.61%
-500
-1000
s=4.47%
-1500
-6
-4
-2
0
2
4
6
1500
P1-CWUPT
4th floor
1000
500
0
d=2.21%
-500
s=5.32%
-1000
-1500
-6
-4
-2
0
2
4
6
1500
P1-CWUPT
6th floor
1000
500
0
d=2.33%
-500
s=5.52%
-1000
-1500
-6
-4
-2
0
2
4
1500
1000
floor beam and the upper floor beams (as a result of the fixed foundation conditions at the
base of the structure) are more pronounced in Structure P1-CWUPT.
P1-CWUPT
8th floor
500
0
d=2.33%
-500
-1000
-1500
-6
s=5.52%
-4
-2
0
2
4
1500
1000
P1-CWUPT
3rd floor
500
0
d=1.97%
-500
-1000
-1500
-6
s=4.96%
-4
-2
0
2
4
1500
1000
P1-CWUPT
5th floor
500
0
-500
d=2.31%
-1000
-1500
-6
-4
s=5.47%
-2
0
2
4
1500
1000
P1-CWUPT
7th floor
500
0
s=5.53%
-500
d=2.33%
-1000
-1500
-6
-4
-2
0
2
4
1500
1000
P1-CWUPT
roof
500
0
-500
s=5.50%
d=2.32%
-1000
-1500
-6
-4
-2
0
2
4
Fig. 13.34 Cyclic coupling beam shear force versus chord rotation behaviors
428
P2-PWUPT
2nd floor
1000
500
0
s=4.85%
-500
-1000
d=2.04%
-4
1500
-2
0
2
4
6
-1500
-6
P2-PWUPT
4th floor
1000
500
0
s=5.39%
-500
-1000
d=2.45%
-1500
-6
-4
-2
0
2
4
6
1500
P2-PWUPT
6th floor
1000
500
0
-500
d=2.56%
-1000
-1500
-6
-4
-2
s=5.51%
1500

1500
P2-PWUPT
8th floor
1000
500
0
-500
d=2.53%
-1000
-1500
-6
-4
s=5.48%
-2
0
2
4
1500
1000
P2-PWUPT
3rd floor
500
0
-500
d=2.31%
-1000
-1500
-6
-4
1500
1000
s=5.21%
-2
0
2
P2-PWUPT
5th floor
500
0
-500
d=2.53%
-1000
-1500
-6
-4
-2
0
2
s=5.48%
-1500
1000
P2-PWUPT
7th floor
500
0
-500
d=2.55%
-1000
-1500
-6
-4
1500
1000
-2
s=5.51%
P2-PWUPT
roof
500
0
-500
d=2.51%
-1000
-1500
-6
s=5.46%
-4
-2
Fig. 13.35 Cyclic coupling beam shear force versus chord rotation behaviors
13.3.9 Coupling Beam Axial Forces and Post-Tensioning Forces
The solid lines in Figs. 13.36 and 13.37 show the midspan axial force, Nb versus
the cyclic loading duration behaviors of the coupling beams in Structures P1-CWUPT
429
and P2-PWUPT, respectively. A positive beam axial force indicates a compressive force.
For comparison, the dashed lines show the total forces in the post-tensioning tendons, Pb
of each coupling beam and the horizontal lines show the total yield strength of the
coupling beam post-tensioning tendons, Pby.
The larger coupling beam post-tensioning forces in Structure P1-CWUPT, as
designed in Chapter 12 with the use of a small top and seat angle ratio (ar=0.051), are
noticeable in the plots. Since the yielding of the beam post-tensioning tendons is
prevented by design, there is no prestress loss in the beam post-tensioning tendons other
than a small, negligible amount that occurs due to the compression yielding and
shortening of the coupling beams under the effect of the beam axial forces.
Similar to the monotonic analysis results in Section 13.2.9, significant differences
are observed between the behaviors of the second floor coupling beam and the upper
floor beams in both structures. The axial forces in the lower floor beams of Structure P1CWUPT go through larger variations and differences from the total beam post-tensioning
force under cyclic loading as compared with the lower floor beams of Structure P2PWUPT. Furthermore, despite the larger coupling beam post-tensioning forces in
Structure P1-CWUPT, complete loss of axial force is observed in the second floor
beam during cyclic response, while the beam axial forces in Structure P2-PWUPT remain
compressive throughout the analysis. This is because of the nonlinear inelastic behavior
that occurs at the bases of the monolithic cast-in-place reinforced concrete wall piers in
Structure P1-CWUPT as compared with the nonlinear but mostly elastic behavior at the
bases of the precast concrete wall piers in Structure P2-PWUPT.
Figs. 13.38 and 13.39 show the neutral axis (i.e., contact) depths, cb at the left ends
of the coupling beams from the reversed cyclic lateral load analyses of Structures P1CWUPT and P2-PWUPT, respectively. The behaviors at the right ends of the coupling
beams are similar. The cb values are measured from the compression corners of the
beams and are normalized with respect to the beam depth, db=559 mm, where the
horizontal lines represent the thickness of the beam flange, tbf. Note that the compression
corners of the beams alternate between the top and bottom corners as the structure is
displaced cyclically. The corresponding extreme steel compression strains, be at the left
end top corners of the coupling beams are shown in Figs. 13.40 and 13.41.
It is observed that the coupling beams in both structures maintain contact with the
wall piers at both ends during the entire nonlinear displacement history, indicating that
the beam post-tensioning tendons provide a sufficient restoring force to yield the tension
angles back in compression and close the gaps at the beam ends.
Similar to the monotonic analysis results in Figs. 13.19 and 13.20, the extreme
compression strains in the coupling beams of Structure P2-PWUPT are smaller than the
430
strains in the beams of Structure P1-PWUPT due to the smaller beam post-tensioning
force.
beam PT force Pb
beam midspan axial force Nb
4000 Pby=3784 kN
5000
P1-CWUPT
2nd floor
3000
2000
1000


5000
4000
beam PT force Pb
Pby=3784 kN
3000
2000
1000
0
5000
P1-CWUPT
4th floor
Pby=3784 kN
3000
2000
1000
beam PT force Pb


5000
4000
4000
Pby=3784 kN
2000
1000
beam PT force Pb
0
5000
P1-CWUPT
6th floor
Pby=3784 kN
3000
2000
beam PT force Pb


5000
1000
0
4000
2000
1000
5000
P1-CWUPT
8th floor
Pby=3784 kN
3000
2000
beam PT force Pb


beam PT force Pb
5000
1000
P1-CWUPT
7th floor
Pby=3784 kN
3000
4000
P1-CWUPT
5th floor
3000
4000
P1-CWUPT
3rd floor
4000
Pby=3784 kN
3000
2000
1000
beam PT force Pb
0
Fig. 13.36 Cyclic coupling beam axial forces and post-tensioning forces
in Structure P1-CWUPT
431
P1-CWUPT
roof
3000
P2-PWUPT
2nd floor
Pby=1892 kN
2000
4000
beam PT force Pb
coupling beam midspan

axial force, N and PT force, P (kN)

5000
1000
0
5000
4000
2000

P2-PWUPT
4th floor
Pby=1892 kN

beam PT force Pb
2000
1000
0
1000
0
5000
beam PT force Pb
4000
P2-PWUPT
6th floor
2000


beam PT force Pb
Pby=1892 kN
1000
0
2000
5000
beam PT force Pb
4000
3000
2000
P2-PWUPT
8th floor
1000
Pby=1892 kN
P2-PWUPT
7th floor
1000
0
5000
beam PT force Pb
2000


beam PT force Pb
Pby=1892 kN
5000
3000
Pby=1892 kN
1000
4000
P2-PWUPT
5th floor
5000
3000
3000
4000
Pby=1892 kN
5000
3000
P2-PWUPT
3rd floor
3000
4000
beam PT force Pb
4000
3000
2000
Pby=1892 kN
1000
0
Fig. 13.37 Cyclic coupling beam axial forces and post-tensioning forces
in Structure P2-PWUPT
432
P2-PWUPT
roof


P1-CWUPT
2nd floor
0.8
0.6
0.4
0.2
tbf=29 mm
0
0.8
P1-CWUPT
3rd floor
0.6
0.4
0.2
tbf=29 mm
0

1
0.8
0.6
P1-CWUPT
4th floor
0.4
0.2
tbf=29 mm


0.8
0.6
P1-CWUPT
5th floor
0.4
0.2
tbf=29 mm
0

1
0.8
0.6
P1-CWUPT
6th floor
0.4
0.2
tbf=29 mm


0.8
0.6
P1-CWUPT
7th floor
0.4
0.2
tbf=29 mm
0

1
0.8
0.6
P1-CWUPT
8th floor
0.4
0.2
tbf=29 mm


0.8
0.6
P1-CWUPT
roof
0.4
0.2
tbf=29 mm
0
Fig. 13.38 Cyclic coupling beam end neutral axis (i.e., contact) depths
433
P2-PWUPT
2nd floor
0.8
0.6
0.4
0.2
tbf=29 mm


0.8
P2-PWUPT
3rd floor
0.6
0.4
0.2
tbf=29 mm
0

1
0.8
0.6
P2-PWUPT
4th floor
0.4
0.2
tbf=29 mm


0.8
0.6
P2-PWUPT
5th floor
0.4
0.2
tbf=29 mm
0

1
0.8
0.6
P2-PWUPT
6th floor
0.4
0.2
tbf=29 mm


0.8
0.6
P2-PWUPT
7th floor
0.4
0.2
tbf=29 mm
0

1
0.8
0.6
P2-PWUPT
8th floor
0.4
0.2
tbf=29 mm


0.8
0.6
P2-PWUPT
roof
0.4
0.2
tbf=29 mm
0
0
Fig. 13.39 Cyclic coupling beam end neutral axis (i.e., contact) depths
434
Fig. 13.40 Cyclic coupling beam end extreme steel compression strains
435
Fig. 13.41 Cyclic coupling beam end extreme steel compression strains
436
13.3.11 Angle Force versus Deformation Behaviors

Figs. 13.42 and 13.43 show the cyclic force versus deformation behaviors of the
top angles at the left ends of the coupling beams in Structures P1-CWUPT and P2PWUPT, respectively. The behaviors of the other angles at each floor/roof level are
similar. The vertical lines show the design-level and survival-level tension angle
deformation demands, uad and uas, from the monotonic nonlinear lateral load analysis of
each structure (using the DRAIN-2DX fiber element model) corresponding to the designlevel and survival-level roof drift demands, d and s, from Chapter 12.
Similar to the monotonic analysis results, for both structures, the angles at the
second floor level have the smallest tension deformations, because, the fixed boundary
conditions assumed at the bases of the wall piers restrain the opening of gaps at the beamto-wall interfaces. The smaller angle forces in Structure P1-CWUPT, as designed in
Chapter 12 with the use of a small top and seat angle ratio (ar=0.051), are noticeable in
the plots.
13.4 Structure with Embedded Steel Coupling Beams
Similar to the CIP-EMB system in Chapter 7, the coupling beams in Structure P1CWUPT are modified to represent a conventional hybrid system with embedded steel
coupling beams, referred to as Structure P3-CWEMB. As described in Chapter 4, the
lateral strength of a post-tensioned steel coupling beam is smaller than the lateral strength
of an embedded beam of the same size, since the post-tensioned beam cannot develop the
full yield and plastic moment capacity of the cross section. Thus, to facilitate a
comparative investigation, smaller coupling beams are used in Structure P3-CWEMB
(with a W12136 section) than the beams in Structure P1-CWUPT (with a W21147
section) such that the monotonic F- relationships of the two coupled wall structures are
similar. All of other design parameters for Structures P1-CWUPT and P3-CWEMB are
the same.
It is assumed that the embedment lengths of the coupling beams in Structure P3CWEMB are equal to lbe=1.22 m (48 in.) at each end, with a clear span length of lb=2.13
m (7 ft). In the analytical modeling of the structure, the lengths of the embedded steel
coupling beams are taken as lb,eff=(2/3)lbe+lb as described in Chapter 3.
13.4.1 Behaviors Under Monotonic and Cyclic Loading
The solid curves in Figs. 13.44a and 13.44b show the coupled wall base shear force
versus roof drift (F-) behaviors of Structure P3-CWEMB under monotonic and reversed
cyclic lateral loads, respectively. For comparison, the monotonic F- behaviors of
Structures P1-CWUPT and P2-PWUPT are shown with the dashed curves in Fig. 13.44a.
437
100
100
P1-CWUPT
2nd floor
50
0
uad =
6 mm
(analysis)
-50
-100
0
10
15
20
25
-100
30
-150
-5
35
100
angle force (kN)
100
50
0
uad =
12 mm
(analysis)
-150
-5
P1-CWUPT
4th floor
uas =
29 mm
(analysis)
10
15
20
25
-150
-5
35
100
50
0
-150
-5
P1-CWUPT
6th floor
uas =
31 mm
(analysis)
10
15
20
25
30
-150
-5
100
P1-CWUPT
8th floor
uad =
13 mm
(analysis)
-100
-150
-5
uas =
31 mm
(analysis)
10
15
20
25
30
uad =
13 mm
(analysis)
35
P1-CWUPT
7th floor
uas =
31 mm
(analysis)
10
15
20
25
0
-50
30
35
uad =
13 mm
(analysis)
-150
-5
P1-CWUPT
roof
uas =
31 mm
(analysis)
10
15
20
25
Fig. 13.42 Cyclic coupling beam angle force versus deformation behaviors
438
30
50
-100
35
P1-CWUPT
5th floor
uas =
30 mm
(analysis)
10
15
20
25
-50
100
50
35
150
-50
-100
35
30
50
150
uad =
12 mm
(analysis)
-100
30
10
15
20
25
-50
100
uad =
13 mm
(analysis)
150
-100
uas =
25 mm
(analysis)
50
150
-50
uad =
9 mm
(analysis)
-50
150
-50
P1-CWUPT
3rd floor
50
150
-100
angle force (kN)
uas =
19 mm
(analysis)
angle force (kN)
angle force (kN)
-150
-5
angle force (kN)
angle force (kN)
150
angle force (kN)
angle force (kN)
150
30
35
1500
1500
P2-PWUPT
2nd floor
0
-500
uad =
8 mm
(analysis)
uas =
23 mm
(analysis)
500
10
15
20
25
-500
30
35
-1500
-5
1500
1000
1000
angle force (kN)
1500
500
0
P2-PWUPT
4th floor
-500
uad =
13 mm
(analysis)
-1000
-1500
-5
uas =
29 mm
(analysis)
10
15
20
25
30
35
500
0
uas =
31 mm
(analysis)
-1500
-5
10
15
20
25
30
35
-1500
-5
1000
uad =
14 mm
(analysis)
-1000
-1500
-5
uas =
31 mm
(analysis)
30
35
uas =
30 mm
(analysis)
P2-PWUPT
5th floor
10
15
20
25
uad =
14 mm
(analysis)
35
uas =
31 mm
(analysis)
10
15
20
25
30
35
500
0
P2-PWUPT
roof
-500
uad =
13 mm
(analysis)
-1500
-5
uas =
30 mm
(analysis)
10
15
20
25
Fig. 13.43 Cyclic coupling beam angle force versus deformation behaviors
439
30
P2-PWUPT
7th floor
-1000
P2-PWUPT
8th floor
10
15
20
25
-500
1000
35
1500
500
30
500
1500
-500
-1000
P2-PWUPT
6th floor
10
15
20
25
uad =
13 mm
(analysis)
-1000
1000
-1500
-5
-500
1000
uad =
14 mm
(analysis)
uas =
28 mm
(analysis)
1500
-1000
uad =
11 mm
(analysis)
500
1500
-500
P2-PWUPT
3rd floor
-1000
angle force (kN)
angle force (kN)
-1500
-5
angle force (kN)
angle force (kN)
500
-1000
angle force (kN)
1000
angle force (kN)
angle force (kN)
1000
30
35
8000
6000
4000
P3-CWEMB
2000
P1-CWUPT
0
0.4
0.8
1.2
2.0
1.6
8000
6000
4000
P3-CWEMB
2000
P2-PWUPT
0
0.4
0.8
1.2
1.6
2.0
(a)
8000
6000
4000
2000
0
-2000
-4000
-6000
-8000
-2
-1.5
-1
-0.5
0.5
1.5
(b)
Fig. 13.44 Coupled wall base shear force versus roof drift behavior of Structure P3CWEMB: (a) monotonic loading; (b) cyclic loading
As expected, Structure P3-CWEMB has a larger inelastic energy dissipation and
smaller self-centering capability than Structures P1-CWUPT and P2-PWUPT. As an
example, the residual roof drift of Structure P3-CWEMB upon unloading from a roof
drift of 1.75% is equal to 1.07%, whereas the residual roof drifts of Structures P1CWUPT and P2-PWUPT upon unloading from the same roof drift are equal to 0.49%
and 0.10%, respectively (see Section 13.3.1). The increased energy dissipation capability
440
of Structure P3-CWEMB occurs as a result of the yielding of the embedded steel

coupling beams.
This chapter investigates the behavior of prototype Structures P1-CWUPT and P2PWUPT from Chapter 12 under both monotonic and cyclic lateral loading, using the
DRAIN-2DX (Prakash et al. 1993) fiber element model described in Chapter 6 including
the modifications recommended in Chapter 10. Evaluations of the global behavior of the
structures as well as the local behavior of the wall piers and the coupling beams are
provided. The analysis results indicate that the design procedures developed in Chapter
12 are valid. The estimations of the various design parameters and structure capacities in
Chapter 12 match reasonably well with the analytical results in this chapter.
441
CHAPTER 14
GROUND MOTION RECORDS FOR DYNAMIC ANALYSES
This chapter describes the ground motion records used in the nonlinear dynamic
time-history analyses of Structures P1-CWUPT and P2-PWUPT from Chapter 12 as
follows: (1) overview; (2) design-level ground motion records; and (3) survival level
ground motion records. The dynamic analyses of the prototype structures are described in
Chapter 15.
14.1 Overview
A considerable number of dynamic analyses using different ground motion records
are often needed to derive clear conclusions about the seismic behavior of nonlinear
structures. The results from these dynamic analyses depend on the characteristics of the
ground motions as well as the vibration characteristics of the structures. Ground motion
characteristics depend on many factors, including the seismicity of the region, magnitude,
epicentral distance, fault rupture mechanism, fault rupture direction, and site soil profile.
A total of forty far-fault ground motion records compiled by the SAC steel
project (Somerville et al. 1997) are used in the dynamic analyses of the prototype
structures in this report. SAC is a joint venture with the following partners: the Structural
Engineers Association of California (SEAOC), the Applied Technology Council (ATC),
and California Universities for Research in Earthquake Engineering (CUREE). The
ground motions are for stiff soil sites [site class D in IBC 2000 (ICC 2000)] in Los
Angeles, similar to the site conditions used in the design of the structures in Chapter 12,
and are grouped into two ensembles based on the seismic demand level: (1) twenty
design-level records (with a 10% probability of being exceeded in 50 years); and (2)
twenty survival-level records (with a 2% probability of being exceeded in 50 years).
For each ground motion, two horizontal components, rotated 45 degrees away from
the fault-normal and fault-parallel orientations, are provided by the SAC steel project.
These ground motions were scaled by the SAC steel project based on target linear-elastic
smooth acceleration response spectra as described in Somerville et al. (1997). The
probabilistic ground motion spectra published by the United States Geological Survey
(Frankel et al. 1996, BSSC 1998, ICC 2000), modified to represent site class D, were
used as the target spectra. In order to preserve the variability in the characteristics of the
individual ground motions, the shapes of the acceleration response spectra of the records
were not modified in the scaling procedure. Instead, for each ground motion, a single
scaling factor was found that minimized the weighted sum of the squared error between
the average 5%-damped linear-elastic acceleration response spectra of the two horizontal
442
components and the corresponding target response spectrum in the period range of 0.3 to
4 sec. This scale factor was then applied to both components of the ground motion, thus
retaining the ratios between the components at all periods. The weights used in the
determination of the ground motion scaling factors were 0.1, 0.3, 0.3, and 0.3 for periods
of 0.3, 1, 2, and 4 sec., respectively.
SAC ground motion records LA01-LA20 (20 recorded motions) are used in this
report to represent the design-level seismic demand and SAC records LA21-LA40 (10
simulated and 10 recorded motions) are used to represent the survival-level seismic
demand, as described in more detail below. The names of these records are kept the same
as the names used by SAC (Somerville et al. 1997).
Note that reliable nonlinear dynamic time-history analysis results depend on the use
of realistic ground motion records with phasing and response spectral characteristics that
are appropriate for the magnitude, distance, site conditions, and wave propagation
properties of the region. According to Somerville et al. (1997), the SAC ground motion
records provide a sample of this variability in phasing and spectra through a set of time
histories that are realistic not only in their average properties but also in their individual
characteristics. Thus, it is expected that the dynamic analysis results presented in Chapter
15 are characteristic of the conditions considered in the design of the prototype structures.
Nevertheless, it should be stated that the findings and conclusions are conditioned on the
ability of the SAC ground motions to suitably represent site soil conditions, site
seismicity, and seismic demand level. More information on the ground motion records
used in this report can be found in Farrow and Kurama (2001).
14.2 Design-Level Ground Motion Records
The acceleration time-histories and the 5%-damped linear-elastic pseudoacceleration response spectra of the SAC design-level LA01-LA20 ground motions are
shown in Figs. 14.1 and 14.2, respectively. The dashed lines in Fig. 14.2 show the mean
acceleration spectra of the 20 ground motions in the ensemble and the thick solid lines
show the design-level spectrum used in the design of the prototype structures (see Fig.
12.2), respectively.
Table 14.1 provides the following information on the SAC LA01-LA20 ground
motion records: (1) record site location; (2) earthquake magnitude; (3) epicentral distance;
(4) SAC scale factor; (5) peak ground acceleration, PGA; and (6) maximum incremental
velocity, MIV. The MIV of a ground motion is equal to the maximum area under the
acceleration time-history of the record between two successive zero-acceleration
crossings. Previous research has shown that a strong correlation exists between the MIV
and the severity of a ground motion (Farrow and Kurama 2001; Kurama and Farrow
2003).
443
acceleration (g)
la01
0
5
20
time (sec.)
40
60
acceleration (g)
5
la03
0
5
20
time (sec.)
40
60
acceleration (g)
5
la05
0
5
20
time (sec.)
40
60
acceleration (g)
5
la07
0
5
20
time (sec.)
40
60
acceleration (g)
acceleration (g)
acceleration (g)
acceleration (g)
acceleration (g)
acceleration (g)
la09
0
5
20
time (sec.)
40
60
5
la02
0
5
20
time (sec.)
40
60
40
60
40
60
40
60
40
60
5
la04
0
5
20
time (sec.)
5
la06
0
5
20
time (sec.)
5
la08
0
5
20
time (sec.)
5
la10
0
5
20
time (sec.)
Fig. 14.1a Acceleration time-histories of SAC design-level ground motions

(LA01-LA10)
444
acceleration (g)
acceleration (g)
5
la11
0
5
20
40
60
5
la12
0
5
20
5
la13
0
5
20
40
60
acceleration (g)
acceleration (g)
la15
0
20
20
40
60
acceleration (g)
acceleration (g)
la17
0
20
20
40
60
acceleration (g)
acceleration (g)
la19
0
20
40
60
40
60
la18
0
5
20
time (sec.)
60
time (sec.)
40
la16
0
time (sec.)
60
time (sec.)
40
la14
0
time (sec.)
60
time (sec.)
40
time (sec.)
acceleration (g)
acceleration (g)
time (sec.)
40
60
time (sec.)
5
la20
0
5
20
time (sec.)
Fig. 14.1b Acceleration time-histories of SAC design-level ground motions

(LA11-LA20)
445
period (sec.)
5
la03
0
period (sec.)
5
la05
0
period (sec.)
5
la07
0
period (sec.)
5
la09
0
period (sec.)
acceleration (g)
acceleration (g)
acceleration (g)
acceleration (g)
la01
acceleration (g)
acceleration (g)
acceleration (g)
acceleration (g)
acceleration (g)
acceleration (g)
5
la02
0
period (sec.)
5
la04
0
period (sec.)
5
la06
0
period (sec.)
5
la08
0
period (sec.)
5
la10
0
period (sec.)

design-level ground motions (LA01-LA10)
446
acceleration (g)
la11
0
period (sec.)
3
acceleration (g)
5
la13
0
period (sec.)
3
acceleration (g)
5
la15
0
period (sec.)
3
acceleration (g)
5
la17
0
period (sec.)
acceleration (g)
acceleration (g)
acceleration (g)
acceleration (g)
acceleration (g)
acceleration (g)
la19
0
period (sec.)
5
la12
0
period (sec.)
5
la14
0
period (sec.)
5
la16
0
period (sec.)
5
la18
0
period (sec.)
5
la20
0
period (sec.)

design-level ground motions (LA11-LA20)
447
TABLE 14.1
SAC DESIGN-LEVEL GROUND MOTION RECORDS LA01-LA20
Record
Name
LA01
LA02
LA03
LA04
LA05
LA06
LA07
LA08
LA09
LA10
LA11
LA12
LA13
LA14
LA15
LA16
LA17
LA18
LA19
LA20
Site
Location
Imperial Valley, 1940, EI Centro
Imperial Valley, 1940, EI Centro
Imperial Valley, 1979, Array #05
Landers, 1992, Barstow
Landers, 1992, Barstow
Landers, 1992, Yermo
Landers, 1992, Yermo
Loma Prieta, 1989, Gilroy
Loma Prieta, 1989, Gilroy
Northridge, 1994, Newhall
Northridge, 1994, Newhall
Northridge, 1994, Rinaldi RS
Northridge, 1994, Rinaldi RS
Northridge, 1994, Sylmar
Northridge, 1994, Sylmar
North Palm Springs, 1986
North Palm Springs, 1986
EQ
Magnitude
6.9
6.9
6.5
6.5
6.5
6.5
7.3
7.3
7.3
7.3
7.0
7.0
6.7
6.7
6.7
6.7
6.7
6.7
6.0
6.0
Distance
(km)
10
10
4.1
4.1
1.2
1.2
36
36
25
25
12
12
6.7
6.7
7.5
7.5
6.4
6.4
6.7
6.7
Scale
Factor
2.01
2.01
1.01
1.01
0.84
0.84
3.20
3.20
2.17
2.17
1.79
1.79
1.03
1.03
0.79
0.79
0.99
0.99
2.97
2.97
PGA
(cm/sec2)
452
663
386
479
296
230
413
417
510
353
652
951
665
645
523
569
558
801
999
968
MIV
(cm/sec.)
88.9
81.0
103
74.6
106
81.6
59.1
71.5
135
76.4
79.4
86.8
84.6
132
124
165
102
139
48.6
122
14.3 Survival-Level Ground Motion Records

The acceleration time-histories and the 5%-damped linear-elastic pseudoacceleration response spectra of the SAC survival-level LA21-LA40 ground motions are
shown in Figs. 14.3 and 14.4, respectively. The dashed lines in Fig. 14.4 show the mean
acceleration spectra of the 20 ground motions in the ensemble and the thick solid lines
show the survival-level spectrum used in the design of the prototype structures (see Fig.
12.2), respectively.
Table 14.2 provides further information on the SAC LA21-LA40 ground motion
records similar to Table 14.1 above for the design-level records.
448
20
time (sec.)
40
60
la23
0
5
20
time (sec.)
40
60
la25
0
5
20
time (sec.)
40
60
la27
0
5
20
time (sec.)
40
60
la29
0
5
20
time (sec.)
40
60
acceleration (g)
acceleration (g)
acceleration (g)
acceleration (g)
la21
acceleration (g)
acceleration (g)
acceleration (g)
acceleration (g)
acceleration (g)
acceleration (g)
la22
0
5
20
time (sec.)
40
60
40
60
40
60
40
60
40
60
la24
0
5
20
time (sec.)
la26
0
5
20
time (sec.)
la28
0
5
20
time (sec.)
la30
0
5
20
time (sec.)
Fig. 14.3a Acceleration time-histories of SAC survival-level ground motions

(LA21-LA30)
449
20
time (sec.)
40
60
la33
0
5
20
time (sec.)
40
60
la35
0
5
20
time (sec.)
40
60
la37
0
5
20
time (sec.)
40
60
la39
0
5
20
time (sec.)
40
60
acceleration (g)
acceleration (g)
acceleration (g)
acceleration (g)
la31
acceleration (g)
acceleration (g)
acceleration (g)
acceleration (g)
acceleration (g)
acceleration (g)
la32
0
5
20
time (sec.)
40
60
40
60
40
60
40
60
40
60
la34
0
5
20
time (sec.)
la36
0
5
20
time (sec.)
la38
0
5
20
time (sec.)
la40
0
5
20
time (sec.)
Fig. 14.3b Acceleration time-histories of SAC survival-level ground motions

(LA31-LA40)
450
period (sec.)
5
la23
0
period (sec.)
5
la25
0
period (sec.)
5
la27
0
period (sec.)
5
la29
0
period (sec.)
acceleration (g)
acceleration (g)
acceleration (g)
acceleration (g)
la21
acceleration (g)
acceleration (g)
acceleration (g)
acceleration (g)
acceleration (g)
acceleration (g)
5
la22
0
period (sec.)
5
la24
0
period (sec.)
5
la26
0
period (sec.)
5
la28
0
period (sec.)
5
la30
0
period (sec.)

survival-level ground motions (LA21-LA30)
451
la31
0
acceleration (g)
acceleration (g)
5
la32
0
5
la33
0
acceleration (g)
acceleration (g)
la35
acceleration (g)
acceleration (g)
la37
acceleration (g)
acceleration (g)
la39
la38
0
1
period (sec.)
period (sec.)
la36
period (sec.)
period (sec.)
la34
period (sec.)
period (sec.)
2
period (sec.)
acceleration (g)
acceleration (g)
period (sec.)
period (sec.)
5
la40
0
1
period (sec.)

survival-level ground motions (LA31-LA40)
452
TABLE 14.2
SAC SURVIVAL-LEVEL GROUND MOTION RECORDS LA21-LA40
Record
Name
LA21
LA22
LA23
LA24
LA25
LA26
LA27
LA28
LA29
LA30
LA31
LA32
LA33
LA34
LA35
LA36
LA37
LA38
LA39
LA40
Site
Location
Kobe, 1995
Kobe, 1995
Loma Prieta, 1989
Loma Prieta, 1989
Northridge, 1994
Northridge, 1994
Northridge, 1994
Northridge, 1994
Tabas, 1974
Tabas, 1974
Elysian Park (simulated)
Palos Verdes (simulated)
EQ
Magnitude
6.9
6.9
7.0
7.0
6.7
6.7
6.7
6.7
7.4
7.4
7.1
7.1
7.1
7.1
7.1
7.1
7.1
7.1
7.1
7.1
453
Distance
(km)
3.4
3.4
3.5
3.5
7.5
7.5
6.4
6.4
1.2
1.2
17
17
11
11
11
11
1.5
1.5
1.5
1.5
Scale
Factor
1.15
1.15
0.82
0.82
1.29
1.29
1.61
1.61
1.08
1.08
1.43
1.43
0.97
0.97
1.10
1.10
0.90
0.90
0.88
0.88
PGA
(cm/sec2)
1258
903
410
464
852
925
909
1304
793
973
1271
1164
767
668
973
1079
698
761
491
613
MIV
(cm/sec.)
274
242
86.9
211
202
269
166
226
92.2
128
208
260
188
161
343
329
263
302
117
279
CHAPTER 15
BEHAVIOR OF PROTOTYPE STRUCTURES
UNDER EARTHQUAKE LOADING
This chapter investigates the response of prototype Structures P1-CWUPT and P2PWUPT from Chapter 12 under earthquake loading. A series of nonlinear dynamic timehistory analyses of the structures are conducted using the twenty survival-level and
twenty design-level ground motion records from Chapter 14. Evaluations of the global
beams are provided. The effect of coupling is investigated by comparing the dynamic
analysis results for the prototype coupled wall systems with results obtained from
dynamic analyses of uncoupled concrete wall piers.
The chapter is divided into the following sections: (1) analytical model for dynamic
analyses; (2) Structure P1-CWUPT; (3) Structure P2-PWUPT; and (4) uncoupled
concrete wall systems.
15.1 Analytical Model for Dynamic Analyses
The dynamic analyses of the prototype structures are conducted using the fiber
element model from Chapter 13 with the DRAIN-2DX Program (Prakash et al. 1993) as
the analytical platform. The modeling of the building mass and viscous damping, as well
as the time step used in the dynamic analyses are described below.
15.1.1 Building Mass
The total building mass is calculated as the seismic weight, W from Chapter 12
divided by the acceleration due to gravity, g. Similar to the assumptions used in the
design of the prototype structures, the total building mass is assumed to be distributed
equally between the four coupled wall systems in the E-W direction and the mass
assigned to each coupled wall system is assumed to be distributed equally between the
two concrete wall piers.
In the analytical model of a coupled wall system, the mass assigned to each wall
pier is distributed to the floor and roof levels based on the distribution of the seismic
weight over the height of the structure in Chapter 12. The wall pier floor and roof masses
are lumped at the horizontal degrees-of-freedom of the fiber element nodes at these levels.
454
15.1.2 Viscous Damping

The inherent structural damping is modeled using a viscous damping matrix, [C]
that is proportional to the lumped mass matrix, [M] and the structure stiffness matrix, [K]
as (Prakash et al. 1993):
[C ] = d [ M ] + d [ K ]
(15.1)
This form of damping is called Rayleigh damping as described in detail in

standard textbooks on structural dynamics (e.g., Clough and Penzien 1993, Chopra 1995).
The component of [C] proportional to [M] (i.e., mass proportional damping) introduces
dashpots at each degree-of-freedom with mass. These dashpots have damping
coefficients given by d[M]. The component of [C] proportional to [K] (i.e., stiffness
proportional damping) introduces dashpots parallel to the structural elements, with
damping coefficients given by d[K]. In the DRAIN-2DX Program, the stiffness
proportional damping is based on the initial linear-elastic stiffness of the structural
elements and remains constant during a nonlinear analysis. Since the mass proportional
damping also remains constant, the damping matrix [C] remains constant during an
analysis.
The Rayleigh damping factors, d and d, are determined using assumed damping
ratios m and n (as a percentage of critical damping) for two selected modes of vibration
(with periods Tm and Tn). Because detailed information about the variation of the
damping ratio with period is not available, it is assumed that the damping ratios for the
two modes are the same (i.e., m=n=). In this case, the Rayleigh damping factors are
given by:
4 2
d
d (Tm + Tn ) TmTn
(15.2)
Usually, Tm is taken as the first mode period and Tn is the period of one of the
higher modes of vibration that is assumed to contribute significantly to the dynamic
response of the structure. The dynamic analyses of the prototype structures were
conducted using a damping ratio of =3% for the first and third linear-elastic modes of
vibration. The assumed damping ratio of 3% is considered to be a conservative lowerbound value (to result in higher seismic demands from the dynamic analyses) for the type
of structures investigated in this report.
Periods and the resulting damping ratios for the first eight linear-elastic vibration
modes of Structures P1-CWUPT and P2-PWUPT are given in Table 15.1. Equation (15.2)
ensures that the assumed damping ratio of =3% is achieved for the first and third modes.
The second mode has a damping ratio smaller than 3%, while modes higher than the third
mode have damping ratios larger than 3% increasing monotonically with decreasing
period. Table 15.1 shows that even though a damping ratio of 3% is specified only for the
455
first and third modes, the resulting damping ratios for the second and fourth modes are
reasonable.
TABLE 15.1
PERIOD AND DAMPING RATIOS FOR PROTOTYPE STRUCTURES
Structure
P1-CWUPT
Structure
P2-PWUPT
period, T (sec.)
1st
mode
0.59
2nd
mode
0.14
3rd
mode
0.060
4th
mode
0.036
5th
mode
0.030
6th
mode
0.029
7th
mode
0.027
8th
mode
0.026
damping ratio, (%)
3.0
1.8
3.0
4.5
5.4
5.6
6.0
6.2
period, T (sec.)
0.61
0.15
0.066
0.040
0.030
0.029
0.028
0.027
damping ratio, (%)
3.0
1.9
3.0
4.9
6.3
6.6
6.8
6.9
15.1.3 Time Step

The accuracy of nonlinear dynamic time-history analysis depends on the time step
used in the numerical time-marching algorithm. For a single-degree-of-freedom system,
Clough and Penzien (1993) recommend a time step shorter than or equal to one-tenth the
linear elastic period of vibration. For the dynamic analyses of the prototype structures, a
constant time step of 0.01 sec. was used. This time step is shorter than one-tenth the
period of the first two modes of the structures (see Table 15.1). Previous analyses
(Kurama et al. 1997) of uncoupled concrete structural walls with similar period ranges
show that the differences between results from dynamic analyses conducted using a time
step of 0.001 sec. and 0.01 sec. are not significant. Thus, the time step of 0.01 sec. is
considered to be adequate to capture the dynamic characteristics of the structures in this
report.
15.2 Structure P1-CWUPT
This section discusses the results from the nonlinear dynamic time-history analyses
of Structure P1-CWUPT as follows: (1) floor/roof drifts; (2) inter-story drifts; (3)
floor/roof accelerations; (4) wall base axial forces; (5) wall base shear forces; (6) wall
base diagonal tension and shear slip failure; (7) wall base moments; (8) wall base strains;
(9) coupling beam chord rotations; (10) coupling beam axial forces; (11) coupling beam
post-tensioning forces; (12) coupling beam shear forces; (13) coupling beam shear slip
failure; and (14) angle deformations.
15.2.1 Floor/Roof Drifts
The peak coupled wall roof drift demands from the dynamic analyses of Structure
P1-CWUPT under the twenty survival-level SAC ground motion records (LA21-LA40)
and twenty design-level records (LA01-LA20) are listed in Tables 15.2 and 15.3,
respectively. As described in Chapter 13, the roof drift, is calculated as the average
lateral displacement of the left and right wall piers at the roof divided by the wall height
to the base, hw.
456
TABLE 15.2
PEAK ROOF DRIFT DEMANDS FOR STRUCTURE P1-CWUPT
UNDER SURVIVAL LEVEL GROUND MOTIONS
ground
LA21 LA22 LA23c LA24
motion
peak roof drift
1.91 1.39 0.69 1.45
(%)
ground
LA31 LA32 LA33 LA34
motion
peak roof drift
1.74 1.92 1.15 1.26
(%)
LA25a LA26 LA27 LA28 LA29 LA30

1.79
2.49
1.43 1.79
estimate
(design)
standard
deviation
1.67
0.82
mean
standard
deviation/mean
1.77
0.46
0.76 0.98
LA35b LA36 LA37 LA38 LA39 LA40

3.62
3.51
2.30 2.63
0.77 1.82
The peak demand under this ground motion is closest to the mean demand.
The peak demand under this ground motion is the largest demand.
c
The peak demand under this ground motion is the smallest demand.
b
TABLE 15.3
PEAK ROOF DRIFT DEMANDS FOR STRUCTURE P1-CWUPT
UNDER DESIGN LEVEL GROUND MOTIONS
ground
LA01a LA02 LA03 LA04 LA05 LA06c LA07 LA08
motion
peak roof drift
0.75
0.81 0.90 0.29 0.39 0.26 0.55 0.68
(%)
ground
LA11 LA12 LA13 LA14 LA15 LA16 b LA17 LA18
motion
peak roof drift
0.58
0.53 1.09 1.18 0.89 1.50 0.70 0.88
(%)
LA09 LA10
0.89
0.67
LA19 LA20
0.63
1.21
estimate
(design)
standard
deviation
0.71
0.32
mean
standard
deviation/mean
0.77
0.41
c
b
The mean peak roof drift demands from the survival-level dynamic analyses in
Table 15.2 is 1.77%, which is close to the estimated survival-level roof drift demand of
s=1.67% and the allowable target roof drift demand of ts=1.75% from Chapter 12. As
an example, Fig. 15.1 shows the roof drift time history of Structure P1-CWUPT under the
LA25 ground motion, for which the peak roof drift demand is closest to the mean demand
from the twenty records (see Table 15.2). The dashed horizontal lines indicate the design
estimate for the peak roof drift demand, s from Chapter 12. It is observed that the
oscillations of the structure are generally centered around the original zero-drift position
demonstrating a large self-centering capability.
The mean peak roof drift demand from the design-level analyses in Table 15.3 is
0.77%, which is close to the estimated design-level roof drift demand of d=0.71% and
smaller than the allowable target roof drift demand of td=1.17% from Chapter 12. It is
concluded that the displacement-based design objectives under the survival-level and
design-level ground motions are, on average, satisfied.
457
2
estimate (design), s=1.67%
1
0
P1-CWUPT
LA25 (survival level)
-1

-2
0
8
10
time, t (sec.)
12
14
16
Fig. 15.1 Roof drift time history for Structure P1-CWUPT

Fig. 15.2 plots the peak coupled wall roof drift demands from Tables 15.2 and 15.3
( markers) against the maximum incremental velocity (MIV) of the ground motion
records (see Tables 14.2 and 14.1 for the MIV values of the LA21-LA40 and the LA01LA20 records, respectively), where the dashed and solid horizontal lines represent the
design estimate (from Chapter 12) and the mean peak roof drift demand, respectively, for
each data set. It is observed that the peak roof drift demands tend to be larger under
ground motions with larger MIV (i.e., there seems to be a correlation between MIV and
the peak roof drift demand, especially for the survival-level records). Furthermore, there
is a significant amount of scatter in the data indicating that, while the design objectives
are satisfied on average, the peak roof drift demand may be significantly larger or
significantly smaller than the design estimate, depending on the earthquake. This is
investigated in more detail in Kurama and Farrow (2003).
15.2.2 Inter-Story Drifts
The peak coupled wall inter-story drift demands from the dynamic analyses of
Structure P1-CWUPT under the twenty survival-level SAC ground motion records
(LA21-LA40) and twenty design-level records (LA01-LA20) are listed in Tables 15.4
and 15.5, respectively. The inter-story drift is defined as the relative lateral displacement
between two adjacent floor/roof levels in a structure divided by the story height and is
calculated as the average inter-story drift of the left and right wall piers.
458
4
3.5
P1-CWUPT
LA21-LA40 (survival level)
peak roof drift (percent)
3
2.5
mean=1.77%
1.5 estimate (design), s=1.67%

1
0.5
0
0
50
100
150
200
250
300
350
(a)
4
3.5
400
P1-CWUPT
LA01-LA20 (design level)
3
2.5
2
1.5
1
mean=0.77%
0.5
0
0
estimate (design), d=0.71%

50
100
150
(b)
Fig. 15.2 Peak roof drift demands for Structure P1-CWUPT:
(a) survival level; (b) design level
459
200
TABLE 15.4
PEAK INTER-STORY DRIFT DEMANDS FOR STRUCTURE P1-CWUPT
ground
motion
peak inter-story
drift (%)
ground
motion
peak inter-story
drift (%)
LA21 LA22 LA23c LA24 LA25 LA26 LA27 LA28 LA29 LA30
2.25 1.64
0.85
1.67
2.08
2.85
1.64
2.06
0.93
1.24
estimate
(design)
standard
deviation
1.92
0.90
LA31 LA32 LA33 LA34 LA35b LA36 LA37 LA38 LA39 LA40a mean
2.19 2.28
1.38
1.50
4.00
3.92
2.60
2.93
0.93
2.05
2.05
standard
deviation/mean
0.44
c
b
TABLE 15.5
PEAK INTER-STORY DRIFT DEMANDS FOR STRUCTURE P1-CWUPT
ground
motion
peak inter-story
drift (%)
ground
motion
peak inter-story
drift (%)
LA01 LA02a LA03 LA04 LA05 LA06c LA07 LA08 LA09 LA10
0.87
0.96
1.07
0.35
0.46
0.33
0.73
0.86 1.06 0.80
estimate
(design)
standard
deviation
0.81
0.36
LA11 LA12 LA13 LA14 LA15 LA16b LA17 LA18 LA19 LA20 mean
0.73
0.82
1.35
1.40
1.02
1.73
0.84
1.04 0.81 1.50
0.94
standard
deviation/mean
0.39
c
b
The mean peak inter-story drift demand from the twenty survival-level analyses in
Table 15.4 is 2.05%, which is close to the estimated survival-level inter-story drift
demand of 1.92% (occurring in the eighth story) from Chapter 12. As an example, Fig.
15.3 shows the inter-story drift time histories of Structure P1-CWUPT under the LA40
ground motion, for which the peak inter-story drift demand is closest to the mean demand
from the twenty records (see Table 15.4). It is observed that the drifts of all eight stories
are in phase, and thus, the deflected shape of the structure is governed by the first
vibration mode. As expected, the peak inter-story drift demand (2.05%) occurs in the
eighth (top) story (indicating a flexure dominated deflected shape) and is larger than the
corresponding peak roof drift demand of 1.82% from Table 15.2. The dashed horizontal
lines show the design estimate for the peak inter-story drift demand in the eighth story
from Chapter 12.
The mean peak inter-story drift demand from the twenty design-level analyses in
Table 15.5 is 0.94%, which is close to the estimated design-level inter-story drift demand
of 0.81% from Chapter 12. As discussed in Chapter 12, according to IBC 2000 (ICC
2000), the peak inter-story drift demand under the design-level ground motion should not
exceed 2%. Because of the relatively small allowable target roof drift demands used in
design (ts=1.75% and td=1.17%), the peak inter-story drift demands in Tables 15.4 and
15.5 satisfy the 2% limit not only for the design-level ground motions, but also for many
of the survival-level motions.
460
3
inter-story drift (percent)
3
P1-CWUPT
first story
2
1
0
-1
-2
-3
5
10
time, t (sec.)
15
0
-1
-2
20
10
time, t (sec.)
15
20
-3
0
P1-CWUPT
third story
1
0
-1
-2
-3
P1-CWUPT
fourth story
1
0
-1
-2
-3
10
time, t (sec.)
15
20
10
time, t (sec.)
15
20
3
P1-CWUPT
second story
P1-CWUPT
fifth story
2
1
0
-1
-2
-3
P1-CWUPT
sixth story
2
1
0
-1
-2
-3
10
time, t (sec.)
15
20
10
time, t (sec.)
15
20
3
estimate (design) =1.92%

2
P1-CWUPT
seventh story
1
0
-1
-2
-3
0
10
time, t (sec.)
15
20
2
P1-CWUPT
eighth story
1
0
-1
-2
-3
0

5
10
time, t (sec.)
Fig. 15.3 Inter-story drift time histories for Structure P1-CWUPT
461
15
20
Note that, for both ground motion sets, the ratio between the mean and estimated
peak inter-story drift demands is similar to the ratio between the mean and estimated peak
roof drift demands. In general, the trends observed for the peak roof drift demands in the
previous section are also valid for the peak inter-story drift demands.
Fig. 15.4 plots the peak coupled wall inter-story drift demands from Tables 15.4
and 15.5 ( markers) against the maximum incremental velocity (MIV) of the ground
motion records (see Tables 14.2 and 14.1), where the dashed and solid horizontal lines
represent the design estimate (for the eighth story, from Chapter 12) and the mean peak
inter-story drift demand, respectively, for each data set. Similar to the peak roof drift
demands discussed in the previous section, there is a significant amount of scatter in the
peak inter-story drift demands, and the demands tend to be larger under the ground
motions with larger MIV (i.e., there seems to be a correlation between MIV and the peak
inter-story drift demand, especially for the survival level records).
15.2.3 Floor/Roof Accelerations
The peak absolute floor/roof acceleration demands from the dynamic analyses of
and 15.7, respectively. The accelerations are calculated as the average total absolute
lumped mass accelerations (i.e., relative accelerations with respect to the ground plus the
ground acceleration) at the floor/roof levels of the left and right wall piers.
As an example, Fig. 15.5 shows the floor/roof acceleration time histories of
Structure P1-CWUPT under the survival-level LA36 ground motion, for which the peak
acceleration demand (occurring at the third floor in this case) is closest to the mean
demand from the twenty records (see Table 15.6). It is observed that the peak
acceleration demand does not necessarily occur at the roof level.
Fig. 15.6 plots the peak floor/roof acceleration demands ( markers) from Tables
15.6 and 15.7 against the maximum incremental velocity (MIV) of the ground motion
records, as well as the peak ground accelerations (PGA, markers) of the records (see
Tables 14.2 and 14.1). The solid horizontal lines represent the mean peak floor/roof
acceleration demand and the mean peak ground acceleration for each data set. As
expected, the peak floor/roof acceleration demands are larger than the peak ground
accelerations. There is a significant amount of scatter in the peak floor/roof acceleration
demands; however, unlike the peak roof drift and peak inter-story drift demands
discussed in the previous sections, no significant correlation exists between the peak
floor/roof acceleration demand and the MIV. It is also interesting to note that the designlevel LA20 ground motion record results in a larger peak floor/roof acceleration demand
than all of the other records, including the survival-level records.
462
4
P1-CWUPT
peak inter-story drift (percent)
3.5
3
2.5
mean=2.05%
2
estimate (design)=1.92%
1.5
1
0.5
0
50
100
150
200
250
300
350
400
(a)
4
3.5
P1-CWUPT
3
2.5
2
1.5
mean=0.94%
estimate (design)=0.81%
0.5
0
50
100
150
200
(b)
Fig. 15.4 Peak inter-story drift demands for Structure P1-CWUPT:
463
TABLE 15.6
PEAK FLOOR/ROOF ACCELERATION DEMANDS FOR STRUCTURE P1-CWUPT
ground
LA21 LA22 LA23 LA24
motion
peak floor/roof
1.76 1.53 1.03 1.26
acceleration (g)
ground
LA31 LA32 LA33 LA34
motion
peak floor/roof
2.48 2.01 1.43 1.25
acceleration (g)
standard
deviation
LA25 LA26 LA27 LA28 LA29b LA30

1.59
1.44
1.17
1.70
2.65
2.52
LA35 LA36a LA37 LA38c LA39 LA40

1.98
1.59
1.16
0.92
1.34
1.00
0.51
mean
standard
deviation/mean
1.59
0.32
c
b
TABLE 15.7
PEAK FLOOR/ROOF ACCELERATION DEMANDS FOR STRUCTURE P1-CWUPT
ground
motion
peak floor/roof
acceleration (g)
ground
motion
peak floor/roof
acceleration (g)
LA01 LA02 LA03 LA04 LA05 LA06c LA07 LA08 LA09

1.13 1.76 1.27
0.87
0.74
0.71
1.07 1.04 1.17
LA11 LA12 LA13 LA14a LA15 LA16 LA17 LA18 LA19

1.45 2.26 1.52
1.33
1.22
1.08
1.14 1.16 2.41
LA10
standard
deviation
0.98
0.57
LA20b
mean
standard
deviation/mean
2.94
1.36
0.42
c
b
15.2.4 Wall Base Axial Forces

The peak tensile and compression wall pier base axial force demands from the
dynamic analyses of Structure P1-CWUPT under the twenty survival-level SAC ground
motion records (LA21-LA40) and twenty design-level records (LA01-LA20) are listed in
Tables 15.8 through 15.11. The wall pier base axial forces are obtained directly from the
fiber beam-column elements used for the wall piers in the analysis model, where a
positive force indicates compression.
Since the proposed design approach does not require the estimation of the survivallevel and design-level wall pier axial forces, the results from the dynamic analyses are
compared with the results from the static idealized bi-linear pushover analysis of the
structure in Chapter 13. The tension-side and compression-side wall pier base axial forces,
Ntwu and Ncwu, corresponding to the estimated coupled wall ultimate state are given in
Tables 15.8 and 15.10, respectively (see also Table 13.2). The results show that the peak
tensile and compressive axial force demands from the dynamic analyses are
underestimated by the axial forces, Ntwu and Ncwu, from the idealized pushover analysis
by 10-15%. This is possibly because of the underestimated coupling beam shear forces as
described later in more detail.
464
3
absolute floor acceleration (g)
3
P1-CWUPT
second floor
2
1
0
-1
-2
-3
0
10
time, t (sec.)
15
P1-CWUPT
fourth floor
2
1
0
-1
-2
5
10
time, t (sec.)
15
1
0
-1
-2
10
time, t (sec.)
15
10
time, t (sec.)
15
20
P1-CWUPT
fifth floor
2
1
0
-1
-2
5
10
time, t (sec.)
15
20
P1-CWUPT
seventh floor
2
1
0
-1
-2
-3
0
20
10
time, t (sec.)
15
20
3
P1-CWUPT
eighth floor
2
1
absolute roof acceleration (g)
-2
3
P1-CWUPT
sixth floor
0
-1
-2
-3
0
-1
-3
0
20
-3
0
-3
0
-3
0
20
P1-CWUPT
third floor
10
time, t (sec.)
15
20
P1-CWUPT
roof
2
1
0
-1
-2
-3
0
10
time, t (sec.)
15
Fig. 15.5 Floor/roof acceleration time histories for Structure P1-CWUPT
465
20
absolute peak floor/roof acceleration

and peak ground acceleration (g)
3.5
peak floor/roof acceleration

peak ground acceleration, PGA
P1-CWUPT
3
2.5
2
mean floor/roof acceleration = 1.59 g
1.5
1
0.5
0
0
mean PGA=0.87 g
50
100
150
200
250
300
350
400
(a)
4

3.5
P1-CWUPT

LA20
3
2.5
2
mean floor/roof acceleration = 1.36 g
1.5
1
0.5
mean PGA=0.59 g
0
0
50
100
150
200
(b)
Fig. 15.6 Peak floor/roof acceleration demands for Structure P1-CWUPT:
466
TABLE 15.8
PEAK WALL PIER BASE TENSILE AXIAL FORCE DEMANDS FOR
STRUCTURE P1-CWUPT UNDER SURVIVAL LEVEL GROUND MOTIONS
ground
motion
peak wall pier base
tension force (kN)
ground
motion
peak wall pier base
tension force (kN)
LA21
LA22
LA23c LA24 LA25 LA26 LA27 LA28 LA29 LA30b
estimate
(idealized)
standard
deviation
-3811
-3840
-3427 -3681 -3893 -3723 -3829 -3844 -3550 -3924
-3372
-142
LA31
LA32
LA33a LA34 LA35 LA36 LA37 LA38 LA39 LA40
mean
standard
deviation/mean
-3862
-3858
-3728 -3847 -3811 -3863 -3877 -3650 -3488 -3880
-3769
0.038
c
b
TABLE 15.9
PEAK WALL PIER BASE TENSILE AXIAL FORCE DEMANDS FOR
STRUCTURE P1-CWUPT UNDER DESIGN LEVEL GROUND MOTIONS
ground
motion
peak wall pier base
tension force (kN)
ground
motion
peak wall pier base
tension force (kN)
LA01 LA02 LA03 LA04 LA05 LA06c LA07 LA08 LA09 LA10
standard
deviation
-3462 -3670 -3702 -2997 -2987 -2626 -3452 -3284 -3607 -3429
-349
LA11 LA12a LA13 LA14 LA15 LA16 LA17 LA18 LA19 LA20b
mean
standard
deviation/mean
-3272 -3464 -3724 -3616 -3713 -3828 -3420 -3588 -4012 -4093
-3497
0.10
c
b
TABLE 15.10
PEAK WALL PIER BASE COMPRESSION AXIAL FORCE DEMANDS FOR
ground
motion
peak wall pier base
compression force (kN)
ground
motion
peak wall pier base
estimate
(idealized)
standard
deviation
9977 10853 11160 10097 10586 9501 10319
9047
596
LA31 LA32 LA33 LA34a LA35 LA36b LA37 LA38 LA39 LA40
mean
standard
deviation/mean
11120 10831 9977 10368 11129 11209 10853 10697 9381 10066
10414
0.057
10635 10235 9279
c
b
467
TABLE 15.11
PEAK WALL PIER BASE COMPRESSION AXIAL FORCE DEMANDS FOR
ground
motion
peak wall pier base
ground
motion
peak wall pier base
standard
deviation
LA01 LA02 LA03 LA04 LA05 LA06c LA07 LA08 LA09 LA10a
9465 9728 9901 8749 8669 8389 9252 9118 9585
9496
556
LA11 LA12 LA13 LA14 LA15 LA16 LA17 LA18 LA19 LA20b
mean
standard
deviation/mean
9118 9150 10275 9781 9710 10422 9421 9496 9817 10502
9502
0.059
c
b
As an example, Fig. 15.7 shows the axial force time histories at the bases of the
left-side and right-side wall piers in Structure P1-CWUPT under the survival-level LA33
ground motion, for which the peak wall pier base tensile axial force demands (occurring
in the right-side wall pier) is closest to the mean demand from the twenty records (see
Table 15.8). Similarly, Fig. 15.8 shows the axial force time histories at the bases of the
ground motion, for which the peak compressive axial force (occurring in the right-side
wall pier) is closest to the mean demand from the twenty records (see Table 15.10). The
dashed horizontal lines indicate the estimated idealized wall pier base axial forces, Ntwu
and Ncwu, from the bi-linear static pushover analysis in Chapter 13 at the coupled wall
ultimate state.
Fig. 15.9 plots the peak tensile ( markers) and peak compressive ( markers) wall
pier base axial force demands from Tables 15.8-15.11 against the maximum incremental
velocity (MIV) of the ground motion records (see Tables 14.2 and 14.1), where the
dashed and solid horizontal lines represent the estimated idealized wall pier base axial
force (from Chapter 13) and the mean peak wall pier base axial force demand,
respectively, for each data set. It is observed that there is very little scatter or variation in
the peak wall pier base axial force demands in each data set, and that the peak wall pier
base axial force demands under the survival-level and design-level ground motions are
similar.
15.2.5 Wall Base Shear Forces
The peak total coupled wall base shear force demands from the dynamic analyses
of Structure P1-CWUPT under the twenty survival-level SAC ground motion records
and 15.13, respectively. The coupled wall base shear force, F is equal to the sum of the
floor and roof level wall pier inertia forces, which are calculated as the lumped mass
times the total absolute acceleration (i.e., relative acceleration with respect to the ground
plus the ground acceleration) at each floor/roof level.
468
Comparisons of the peak coupled wall base shear force demands in Tables 15.12
and 15.13 with the coupled wall base shear forces reached during the static push-over
analysis of the structure in Chapter 13 show that the forces from the dynamic analyses are
significantly larger than the forces from the static analysis. As discussed in Chapter 11,
the differences between the static and dynamic analysis forces are because of the
contribution of higher modes of vibration to the dynamic response of the structure.
10000
10000
5000
-5000
-10000
0
estimate (idealized), Ntwu=3372 kN

P1-CWUPT
left-side wall
5
10
time, t (sec.)
15
5000
-5000
-10000
0
20

P1-CWUPT
right-side wall
5
10
time, t (sec.)
(a)
15
20
(b)
Fig. 15.7 Wall pier base axial force time histories for Structure P1-CWUPT:
(a) left-side wall; (b) right-side wall
10000
estimate
(idealized)
Ncwu=9047 kN
5000
10000
-5000
P1-CWUPT
left-side wall
-10000
estimate
(idealized),
Ncwu=9047 kN
5000
-5000
P1-CWUPT
right-side wall
-10000
0
8
10
time (sec.)
12
14
16
(a)
8
10
time (sec.)
12
(b)
Fig. 15.8 Wall pier base axial force time histories for Structure P1-CWUPT:
469
14
16
15000
mean=10414 kN
peak wall pier base axial force (kN)
10000
estimate (idealized), Ncwu=9047 kN
peak compressive axial force
peak tensile axial force
5000
P1-CWUPT
0
estimate (idealized), Ntwu=-3372 kN
mean=-3769 kN
-5000
-10000
-15000
0
50
100
150
200
250
300
350
400
(a)
15000
mean=9502 kN
peak wall pier base axial force (kN)
10000
5000
peak compressive axial force

peak tensile axial force
P1-CWUPT
-5000
mean=-3497 kN
-10000
-15000
50
100
150
200
(b)
Fig. 15.9 Peak wall pier base axial force demands for Structure P1-CWUPT:
470
TABLE 15.12
PEAK COUPLED WALL BASE SHEAR FORCE DEMANDS FOR STRUCTURE
P1-CWUPT UNDER SURVIVAL LEVEL GROUND MOTIONS
ground
motion
peak coupled wall
ground
motion
peak coupled wall
LA21 LA22 LA23c LA24 LA25 LA26 LA27 LA28 LA29 LA30b
estimate
(design)
standard
deviation
13001 11007 7392 9873 12253 12822 9383 12768 14891 15683
13156
2351
LA31 LA32 LA33 LA34 LA35 LA36 LA37 LA38a LA39 LA40
mean
standard
deviation/mean
14217 14771 11699 10397 15585 13601 10855 11887 8392c 10514
12050
0.20
c
b
TABLE 15.13
P1-CWUPT UNDER DESIGN LEVEL GROUND MOTIONS
ground
motion
peak coupled wall
ground
motion
peak coupled wall
standard
deviation
8876 10763 9630 6125 5860 5602 6971 8032 9099
8522
3002
LA11a LA12 LA13 LA14 LA15 LA16 LA17 LA18 LA19 LA20b mean
9843 15359 10641 10191 9109 10599 8199 11263 14593 16797
9804
standard
deviation/mean
0.31
c
b
The mean peak coupled wall base shear force demand from the twenty survivallevel ground motions is 12050 kN, which is close to and smaller than the estimated
maximum coupled wall base shear force demand of Qw,max=13156 kN from Chapter 12.
Thus, the method used in Chapter 12 to estimate the coupled wall base shear force
demand of the prototype structures including the effect of higher modes appears to be
valid. Note that the design approach does not require the estimation of the maximum
coupled wall base shear force demand under the design-level earthquake; and thus, no
design-level estimate is provided in Table 15.13.
As an example, Fig. 15.10 shows the coupled wall base shear force time history of
coupled wall base shear force demand is closest to the mean demand from the twenty
records (see Table 15.12). The dashed horizontal lines indicate the design estimate for the
maximum coupled wall base shear force demand, Qw,max from Chapter 12.
Fig. 15.11 plots the peak coupled wall base shear force demands from Tables
15.12 and 15.13 ( markers) against the maximum incremental velocity (MIV) of the
ground motion records (see Tables 14.2 and 14.1), where the dashed and solid horizontal
lines represent the design estimate (from Chapter 12) and the mean peak coupled wall
base shear force demand, respectively, for each data set. Significant scatter is observed in
471
the data indicating that, while the survival-level design estimate matches the mean
demand from the dynamic analyses quite well, the coupled wall base shear force demand
may be considerably larger or smaller than the estimated demand, depending on the
earthquake. It is also interesting to note that the design-level LA20 ground motion record
results in a larger peak coupled wall base shear force demand than all of the other records,
including the survival-level records. Some correlation can be observed between the peak
coupled wall base shear force demands in Fig. 15.11 and the peak acceleration demands
in Fig. 15.6.
The peak wall pier base shear force demands from the dynamic analyses of
and 15.15, respectively. The wall pier base shear forces are calculated from the
equilibrium of the floor/roof level wall pier inertia forces, with the coupling beam
midspan axial forces and beam post-tensioning tendon forces from the analysis model
(obtained from the fiber beam-column elements and the truss elements representing the
coupling beams and the beam post-tensioning tendons, respectively).
15000
estimate (design), Qw,max=13156 kN
10000
5000
0
-5000
P1-CWUPT
-1000 LA38 (survival level)
-15000
0

10
15
time, t (sec.)
20
25
Fig. 15.10 Coupled wall base shear force time history for Structure P1-CWUPT
472
20000
peak coupled wall base shear force (kN)
P1-CWUPT
16000
12000
mean=12050 kN
8000
4000
50
100
200
150
250
300
350
400
(a)
20000
P1-CWUPT
LA20
16000
12000
mean=9804 kN
8000
4000
50
100
150
200
(b)
Fig. 15.11 Peak coupled wall base shear force demands for Structure P1-CWUPT:
473
TABLE 15.14
PEAK WALL PIER BASE SHEAR FORCE DEMANDS FOR
ground
motion
peak wall pier base
shear force (kN)
ground
motion
peak wall pier base
shear force (kN)
estimate
(design)
standard
deviation
5098 7055 8330 8794 6519 8152 9324 9935
9341
1326
LA31 LA32b LA33 LA34 LA35 LA36 LA37 LA38 LA39 LA40
mean
standard
deviation/mean
9551 10015 7644 7666 8337 9057 7451 8497 5798 7679
8020
0.17
LA21 LA22 LA23c LA24 LA25 LA26 LA27 LA28a LA29 LA30
8770 6720
c
b
TABLE 15.15
PEAK WALL PIER BASE SHEAR FORCE DEMANDS FOR STRUCTURE
ground
motion
peak wall pier base
shear force (kN)
ground
motion
peak wall pier base
shear force (kN)
standard
deviation
LA01 LA02 LA03 LA04 LA05 LA06c LA07 LA08 LA09a LA10
5993 6025 6950 3936 3981 3193 4599 4886 6313
6052
1671
LA11 LA12 LA13 LA14 LA15 LA16 LA17 LA18 LA19 LA20b mean
5833 9415 7304 6489 6035 6815 5777 7584 8736
9085
6250
standard
deviation/mean
0.27
c
b
The mean peak wall pier base shear force demand from the twenty survival-level
analyses is 8020 kN, which is reasonably close to and smaller than the estimated
maximum wall pier base shear force demand of Qw,dt=9341 kN from Chapter 12. Thus,
the method used in Chapter 12 to estimate the maximum wall pier base shear force
demand appears to be valid. Note that the design approach does not require the estimation
of the maximum wall pier base shear force demand under the design-level earthquake;
and thus, no design-level estimate is provided in Table 15.15.
As an example, Fig. 15.12 shows the shear force time histories at the bases of the
ground motion, for which the peak wall pier base shear force demand (occurring in the
left-side wall pier) is closest to the mean demand from the twenty records (see Table
15.14). The dashed horizontal lines indicate the design estimate for the maximum wall
pier base shear force demand, Qw,dt from Chapter 12. It is observed that there are residual
base shear forces in the left-side and right-side wall piers at the end of the ground motion,
even though the coupled wall base shear force is close to zero.
474
10000
10000
estimate (design), Qw,dt=9341 kN
5000
P1-CWUPT
left-side wall
-5000

-10000
0

5000
P1-CWUPT
right-side wall
-5000
-10000
10
time, t (sec.)
15
20
(a)
10
time, t (sec.)
15
20
(b)
Fig. 15.12 Wall pier base shear force time histories for Structure P1-CWUPT:
Fig. 15.13 plots the peak wall pier base shear force demands from Tables 15.14 and
15.15 ( markers) against the maximum incremental velocity (MIV) of the ground
represent the design estimate (from Chapter 12) and the mean peak wall pier base shear
force demand, respectively, for each data set. Similar to the peak coupled wall base shear
force demands in Fig. 15.11, significant scatter is observed in the peak wall pier base
shear force demands and the demands for some of the design-level ground motion
records are similar to the largest demand from the survival-level records.
If the peak wall pier base shear force demands in Tables 15.14 and 15.15 are
divided by the corresponding peak coupled wall base shear force demands in Tables
15.12 and 15.13, the wall pier base shear force distribution ratios are obtained. The mean
wall pier base shear force distribution ratio from the twenty survival-level analyses is
0.67, which is close to the design ratio of Qw,dt/Qw,max=0.71 from Chapter 12.
Note that the peak wall pier base shear force demands in Tables 15.14 and 15.15
and the corresponding peak coupled wall base shear force demands in Tables 15.12 and
15.13 do not necessarily occur at the same time during each dynamic analysis. Thus,
Tables 15.16 and 15.17 show the peak ratio of the wall pier base shear force to the
corresponding coupled wall base shear force occurring at the same time during each
analysis. The mean value of the peak wall pier base shear force distribution ratio from the
twenty survival-level analyses in Table 15.16 is 0.69, which is close to the value used in
design and the value calculated from the peak wall pier and peak coupled wall base shear
force demands above.
475
20000
peak wall pier base shear force (kN)
P1-CWUPT
16000
12000
8000
mean=8020 kN
4000
50
100
150
200
250
300
350
400
(a)
20000
P1-CWUPT
16000
12000
8000
4000
mean = 6250 kN
50
100
150
200
(b)
Fig. 15.13 Peak wall pier base shear force demands for Structure P1-CWUPT:
476
TABLE 15.16
PEAK WALL PIER BASE SHEAR FORCE DISTRIBUTION RATIOS FOR
ground
motion
peak base shear force
distribution ratio
ground
motion
distribution ratio
LA21 LA22 LA23 LA24 LA25 LA26 LA27 LA28 LA29 LA30
0.67
0.66
0.75
0.71
0.68
0.69
0.74
0.71
0.65
0.63
estimate
(design)
standard
deviation
0.71
0.047
LA31 LA32 LA33 LA34 LA35 LA36 LA37 LA38 LA39 LA40 mean
0.67
0.68
0.72
0.74
0.53
0.67
0.69
0.71
0.69
0.73
0.69
standard
deviation/mean
0.069
TABLE 15.17
PEAK WALL PIER SHEAR FORCE DISTRIBUTION RATIOS FOR
ground
motion
distribution ratio
ground
motion
distribution ratio
standard
deviation
0.69
0.64
0.74
0.64
0.68
0.57
0.72
0.65
0.75
0.73
0.064
0.69
0.63
0.77
0.76
0.76
0.76
0.74
0.69
0.63
0.56
0.69
standard
deviation/mean
0.093
15.2.6 Wall Base Diagonal Tension and Shear Slip Failure

It is shown in the previous section that the mean peak wall pier base shear force
demand under the twenty survival-level ground motion records in Table 15.14 is smaller
than the estimated design maximum wall pier base shear force demand from Chapter 12.
Thus, it is concluded that diagonal tension failure of the wall piers in Structure P1CWUPT can be, on average, prevented as long as the design diagonal tension capacity
Fw,dt=9261 kN from Chapter 12 is achieved and maintained.
Similarly, the mean peak coupled wall base shear force demand under the twenty
survival-level ground motion records in Table 15.12 is smaller than the estimated design
maximum coupled wall base shear force demand from Chapter 12. Furthermore, the
design shear slip capacity estimated for Structure P1-CWUPT in Chapter 12 is equal to
Fw,ss=44614 kN, which is larger than the peak coupled wall base shear force demands
under all of the survival-level and design-level ground motion records in Tables 15.12
and 15.13. Thus, it is concluded that shear slip failure of the structure does not occur.
15.2.7 Wall Base Moments
The peak wall pier base moment demands from the dynamic analyses of Structure
P1-CWUPT under the twenty survival-level SAC ground motion records (LA21-LA40)
respectively. The wall pier base moments are calculated from the equilibrium of the
floor/roof level wall pier inertia forces with the coupling beam midspan axial forces and
beam post-tensioning tendon forces from the analysis model (obtained from the fiber
477
beam-column elements and the truss elements representing the coupling beams and the
beam post-tensioning tendons, respectively).
TABLE 15.18
PEAK WALL PIER BASE MOMENT DEMANDS FOR STRUCTURE P1-CWUPT
ground
motion
peak wall pier base
moment (kN.m)
ground
motion
peak wall pier base
moment (kN.m)
LA21 LA22 LA23c LA24 LA25 LA26 LA27a LA28 LA29 LA30
estimate
(idealized)
standard
deviation
77583 71584 61415 70781 76046 79854 73640 77549 65381 71177
76242
6169
LA31 LA32 LA33 LA34 LA35 LA36b LA37 LA38 LA39 LA40
mean
standard
deviation/mean
79255 78351 72386 73109 82509 83051 80532 79594 62828 76815
74672
0.083
c
b
TABLE 15.19
PEAK WALL PIER BASE MOMENT DEMANDS FOR STRUCTURE P1-CWUPT
ground
motion
peak wall pier base
moment (kN.m)
ground
motion
peak wall pier base
moment (kN.m)
standard
deviation
63573 65121 68285 45508 50976 38153 57891 58602 65641 62997
8716
LA11 LA12 LA13 LA14 LA15 LA16 LA17 LA18 LA19a LA20b mean
standard
deviation/mean
59789 60625 67821 69595 67211 72521 63347 65257 62082 73753 61937
0.14
c
b
Since the proposed design approach does not require the estimation of the survivallevel and design-level wall pier base moments, the results from the dynamic analyses are
compared with the results from the static idealized bi-linear pushover analysis of the
structure in Chapter 13. The bending moment at the base of the compression-side wall
pier (which has a larger base moment than the tension side wall pier), Mcwu,
corresponding to the estimated coupled wall ultimate state is given in Table 15.18 (see
also Table 13.2). The results show that the wall pier base moment, Mcwu from the
idealized pushover analysis provides a good estimate of the peak wall pier base moment
demands from the dynamic analyses.
As an example, Fig. 15.14 shows the bending moment time histories at the bases of
the left-side and right-side wall piers in Structure P1-CWUPT under the survival-level
LA27 ground motion, for which the peak wall pier base moment demand (occurring in
the right-side wall pier) is closest to the mean demand from the twenty records (see Table
15.18). The dashed horizontal lines indicate the estimated idealized compression-side
wall pier base moment, Mcwu from the bi-linear static pushover analysis in Chapter 13 at
the coupled wall ultimate state.
478
80000
estimate (idealized), Mcwu=76242 kN.m
80000
40000
0
P1-CWUPT
left-side wall
-40000
40000
0
P1-CWUPT
right-side wall
-40000
-80000
-80000
0
6
8
time, t (sec.)
10
12
(a)
6
8
time, t (sec.)
10
12
(b)
Fig. 15.14 Wall pier base moment time histories for Structure P1-CWUPT:
Fig. 15.15 plots the peak wall pier base moment demands from Tables 15.18 and
represent the estimated idealized wall pier base moment (from Chapter 13) and the mean
peak wall pier base moment demand, respectively, for each data set. A considerable
amount of scatter is observed in the peak wall pier base moment demands in each data set.
The peak total coupled wall base moment demands from the dynamic analyses of
and 15.21, respectively. The coupled wall base moment, Mw is calculated from the
equilibrium of the floor and roof level wall pier inertia forces.
Similar to the wall pier base moments, the peak coupled wall base moment
demands from the dynamic analyses are compared with the results from the static
idealized bi-linear pushover analysis of the structure in Chapter 13. The coupled wall
base moment, Mwu corresponding to the estimated coupled wall ultimate state is given in
Table 15.20 (see also Table 13.2). The results show that the coupled wall base moment,
Mwu from the idealized pushover analysis underestimates the peak coupled wall base
moment demands from the dynamic analyses, even though the wall pier base moment
demands are estimated reasonably well. This is possibly because of the underestimated
coupling beam shear forces as described later in more detail.
479
90000
peak wall pier base moment (kN.m)
80000
mean = 74672 kN.m
70000
60000
50000
40000
30000
P1-CWUPT
20000
10000
0
50
100
150
200
250
300
350
400
(a)
90000
80000
70000
60000
mean = 61937 kN.m
50000
40000
30000
P1-CWUPT
20000
10000
0
50
100
150
200
(b)
Fig. 15.15 Peak wall pier base moment demands for Structure P1-CWUPT:
480
TABLE 15.20
PEAK COUPLED WALL BASE MOMENT DEMANDS FOR
ground
motion
peak coupled wall
base moment (kN.m)
ground
motion
peak coupled wall
base moment (kN.m)
LA21 LA22a LA23c LA24 LA25 LA26 LA27 LA28 LA29 LA30
estimate
(idealized)
standard
deviation
192510 183300 146210 174890 189995 212398 171370 192660 153340 169320
161323
20742
mean
standard
deviation/mean
191080 197140 169270 175180 217570 220260 193770 202460 152960 182400
184404
0.11
c
b
TABLE 15.21
PEAK COUPLED WALL BASE MOMENT DEMANDS FOR
ground
LA01a LA02 LA03 LA04
motion
peak coupled wall
149230 154190 159220 113430
base moment (kN.m)
ground
LA11 LA12 LA13 LA14
motion
peak coupled wall
144120 144360 172870 168900
base moment (kN.m)
LA05 LA06c LA07 LA08 LA09 LA10
standard
deviation
121560 101580 136710 145150 159340 147730
19993
LA15 LA16b LA17 LA18 LA19 LA20
mean
155660 180920 146650 156230 145510 176880 149010
standard
deviation/mean
0.13
c
b
As an example, Fig. 15.16 shows the coupled wall base moment time history of
coupled wall base moment demand is closest to the mean demand from the twenty
records (see Table 15.20). The dashed horizontal lines indicate the estimated idealized
coupled wall base moment, Mwu from the bi-linear static pushover analysis in Chapter 13
at the coupled wall ultimate state.
Fig. 15.17 plots the peak coupled wall base moment demands from Tables 15.20
represent the estimated idealized coupled wall base moment (from Chapter 13) and the
mean peak coupled wall base moment demand, respectively, for each data set. Similar to
the peak wall pier base moment demands, a considerable amount of scatter is observed in
the peak coupled wall base moment demands in each data set.
481
250000
coupled wall base moment, Mw (kN.m)
200000
P1-CWUPT
150000
estimate (idealized), Mwu=161323 kN.m
100000
50000
0
-50000
-100000
-150000
-200000
-250000
0
8
10
time, t (sec.)
12
14
16
Fig. 15.16 Coupled wall base moment time history for Structure P1-CWUPT
482
peak coupled wall base moment (kN.m)
300000
P1-CWUPT
200000
mean=184404 kN.m
100000
50
100
150
200
250
300
350
400
(a)
300000
P1-CWUPT
200000
mean=149010 kN.m
100000
50
100
150
200
(b)
Fig. 15.17 Peak coupled wall base moment demands for Structure P1-CWUPT:
483
15.2.8 Wall Base Strains

The peak confined concrete compression strain demands at the bases of the wall
piers from the dynamic analyses of Structure P1-CWUPT under the twenty survival-level
SAC ground motion records (LA21-LA40) and twenty design-level records (LA01-LA20)
are listed in Tables 15.22 and 15.23, respectively. The concrete compression strains are
determined directly from the fiber beam-column elements used for the wall piers in the
analysis model.
TABLE 15.22
PEAK WALL PIER BASE CONFINED CONCRETE COMPRESSION
STRAIN DEMANDS FOR STRUCTURE P1-CWUPT
ground
motion
peak confined
concrete strain (10-3)
ground
motion
peak confined
-8.2
-5.4
-3.0
-5.1
-7.2
-11.0
-6.0
-8.3
-3.5
-4.9
estimate
(design)
standard
deviation
-9.0
-4.4
LA31 LA32a LA33 LA34 LA35 LA36b LA37 LA38 LA39c LA40 mean
-9.1
-8.4
-5.4
-5.6 -17.6 -19.0 -11.6 -11.7
-3.3
-8.0
standard
deviation/mean
-8.1
0.54
c
b
TABLE 15.23
PEAK WALL PIER BASE CONFINED CONCRETE COMPRESSION
STRAIN DEMANDS FOR STRUCTURE P1-CWUPT
ground
motion
peak confined
ground
motion
peak confined
standard
deviation
-3.1
-3.5
-4.2
-1.2
-1.4
-1.0
-2.2
-2.3
-3.7
-3.0
-1.4
LA11 LA12 LA13 LA14 LA15 LA16 LA17a LA18 LA19 LA20b mean
-2.6
-2.6
-4.2
-4.6
-4.2
-5.8
-3.1
-3.5
-3.0
-6.1
-3.3
standard
deviation/mean
0.42
c
b
The mean peak wall pier base confined concrete compression strain demand from
the survival-level dynamic analyses in Table 15.22 is -0.0081, which is smaller than and
close to the estimated wall pier base confined concrete strain demand of -0.0090 from
Chapter 12. Thus, the method used in Chapter 12 to estimate the wall pier base confined
concrete compression strain demand seems to be valid. Note that the design approach
does not require the estimation of the confined concrete compression strain demand under
the design-level earthquake; and thus, no design-level estimate is provided in Table 15.23.
As an example, Fig. 15.18 shows the extreme confined concrete compression strain
time histories at the base of Structure P1-CWUPT under the survival-level LA32 ground
484
motion, for which the peak wall pier base confined concrete compression strain demand
(occurring at the left end of the left-side wall pier) is closest to the mean demand from the
twenty records (see Table 15.22). A total of four strain time histories are shown for the
left and right ends of the left-side and right-side wall piers, where the dashed horizontal
lines indicate the design estimate for the wall pier base confined concrete compression
strain demand from Chapter 12.
Fig. 15.19 plots the peak wall pier base confined concrete compression strain
demands from Tables 15.22 and 15.23 ( markers) against the maximum incremental
dashed and solid horizontal lines represent the design estimate and the mean peak wall
pier base confined concrete compression strain demand, respectively, for each data set.
P1-CWUPT
-0.09
0

concrete compression strain

estimate (design)=0.009
5
10
time, t (sec.)
15
20
P1-CWUPT
0
10
time, t (sec.)
P1-CWUPT
0
(a)
-0.09
-0.09


The trends for the peak wall pier base confined concrete compression strain
demands in Fig. 15.19 are similar to the trends for the peak coupled wall roof drift
demands in Fig. 15.2.
(b)
20
15
20
P1-CWUPT
0
20
15
-0.09
15
10
time, t (sec.)
10
time, t (sec.)
Fig. 15.18 Wall pier base extreme confined concrete compression strain time histories
for Structure P1-CWUPT: (a) left-side wall; (b) right-side wall
485
peak wall pier base confined concrete compression strain
0
P1-CWUPT
-0.002
-0.004
-0.006
mean = -0.0081
-0.008
estimate (design) = -0.009
-0.01
-0.012
-0.014
-0.016
-0.018
-0.02
50
100
150
200
250
300
350
400
(a)
0
0.002
0.004
mean =-0.0033
0.006
0.008
0.01
P1-CWUPT
0.012
0.014
0.016
0.018
0.02
0
50
100
150
200
(b)
Fig. 15.19 Peak wall pier base confined concrete compression strain demands
for Structure P1-CWUPT: (a) survival level; (b) design level
486
15.2.9 Coupling Beam Chord Rotations

The peak coupling beam chord rotation demands from the dynamic analyses of
and 15.25, respectively. As described in Chapter 7, the coupling beam chord rotation, b
is measured from a tangent drawn at the left end of each beam.
The mean peak coupling beam chord rotation demand from the survival-level
dynamic analyses in Table 15.24 is 3.08%, which is significantly smaller than the
estimated survival-level coupling beam chord rotation demand of s=5.49% from Chapter
12. As an example, Fig. 15.20 shows the chord rotation time histories for the eight
coupling beams in Structure P1-CWUPT under the survival-level LA31 ground motion,
for which the peak coupling beam chord rotation demand (occurring in the roof beam) is
closest to the mean demand from the twenty records (see Table 15.24).
TABLE 15.24
PEAK COUPLING BEAM CHORD ROTATION DEMANDS FOR STRUCTURE P1CWUPT UNDER SURVIVAL LEVEL GROUND MOTIONS
ground
motion
peak coupling beam
3.27 2.21 1.17 2.44
chord rotation (%)
ground
motion
peak coupling beam
3.06 3.34 1.97 2.21
chord rotation (%)
LA25 LA26 LA27 LA28 LA29 LA30

3.17
4.35
2.46
3.10
1.28 1.71
estimate
(design)
standard
deviation
5.49
1.45
LA35b LA36 LA37 LA38 LA39 LA40 mean

6.37
6.13
4.04
4.62
1.33 3.18
3.08
standard
deviation/mean
0.47
c
b
TABLE 15.25
PEAK COUPLING BEAM CHORD ROTATION DEMANDS FOR STRUCTURE
ground
LA01 LA02 LA03 LA04
motion
peak coupling beam
1.79 1.35 0.93 0.35
chord rotation (%)
ground
LA11 LA12 LA13 LA14
motion
peak coupling beam
0.71 0.62 1.98 2.05
chord rotation (%)
LA05 LA06c LA07 LA08a LA09 LA10

0.58
0.16
0.97
1.20
1.30
1.12

1.48
2.63
1.17
1.50
0.90
c
b
487
2.25
estimate
(design)
standard
deviation
2.33
0.65
mean
standard
deviation/mean
1.25
0.52
4
P1-CWUPT
2nd floor
coupling beam chord

rotation, b (percent)
coupling beam chord

2
4
2
4
10
time, t (sec.)
15
20
15
20
15
20
15
20
15
20
2
4
10
time, t (sec.)
15
20
10
time, t (sec.)
4
P1-CWUPT
6th floor
coupling beam chord

coupling beam chord

10
time, t (sec.)
P1-CWUPT
5th floor

4
P1-CWUPT
7th floor
2

4
4
0
10
time, t (sec.)
15
20
10
time, t (sec.)
4
P1-CWUPT
8th floor
coupling beam chord

coupling beam chord

4
P1-CWUPT
4th floor
coupling beam chord

coupling beam chord

P1-CWUPT
3rd floor
2
4
0
10
time, t (sec.)
P1-CWUPT
roof
2
4
15
20
10
time, t (sec.)
Fig. 15.20 Coupling beam chord rotation time histories for Structure P1-CWUPT
Similarly, the mean peak coupling beam chord rotation demand from the designlevel analyses in Table 15.25 is 1.25%, which is significantly smaller than the estimated
design-level coupling beam chord rotation demand of d=2.33% from Chapter 12. It is
concluded that the displacement-based design objectives under the survival-level and
design-level ground motions are satisfied; however, the procedure to estimate the
488
coupling beam chord rotation demands may need to be improved to prevent an overlyconservative design.
Fig. 15.21 plots the peak coupling beam chord rotation demands from Tables 15.24
represent the design estimate (from Chapter 12) and the mean peak coupling beam chord
rotation demand, respectively, for each data set. It is observed that the coupling beam
chord rotation demands tend to follow similar trends to the coupled wall roof drift
The peak axial force demands at the midspan of the coupling beams from the
dynamic analyses of Structure P1-CWUPT under the twenty survival-level SAC ground
Tables 15.26 and 15.27, respectively. The coupling beam midspan axial forces, Nb are
determined directly from the fiber beam-column elements at the beam midspan locations
in the analysis model.
As discussed in Chapter 13, the largest coupling beam axial forces develop in the
lowest beam at the second floor level as a result of the fixed boundary conditions
assumed at the wall pier bases. The mean peak coupling beam midspan axial force
demand from the survival-level dynamic analyses in Table 15.26 is 4131 kN, which is
reasonably close to and smaller than the design force of, approximately, 1.25Pby=4730
kN from Chapter 12. Thus, the axial force used in Chapter 12 for the design of the
coupling beams appears to be valid. Note that the design approach does not require the
estimation of the coupling beam axial force under the design-level earthquake; and thus,
no design-level estimate is provided in Table 15.27.
As an example, Fig. 15.22 shows the midspan axial force time histories for the
eight coupling beams in Structure P1-CWUPT under the survival-level LA25 ground
motion, for which the peak coupling beam midspan axial force demand (always occurring
in the second floor beam) is closest to the mean demand from the twenty records (see
Table 15.26). The dashed horizontal lines indicate the design estimate for the peak
coupling beam midspan axial force, which is equal to 1.25Pby for the second and third
floor beams and Pby for the upper level beams. It is observed that the peak midspan axial
force demands in the upper level beams are very close to the design estimate. Note also
that there is a considerable loss in the second floor beam axial force as discussed in
Chapter 13, resulting in a smaller axial force in this beam at the end of the earthquake as
compared with the upper level beams.
489
peak coupling beam chord rotation (percent)
7
6
estimate (design), s = 5.49%
5
P1-CWUPT
mean = 3.08%
3
2
1
0
0
50
100
150
200
250
300
350
400
(a)
7
6
P1-CWUPT
5
4
3
estimate (design), d = 2.33%
2
mean = 1.25%
1
0
0
50
100
150
200
(b)
Fig. 15.21 Peak coupling beam chord rotation demands for Structure P1-CWUPT:
490
TABLE 15.26
PEAK COUPLING BEAM MIDSPAN AXIAL FORCE DEMANDS FOR
ground
motion
peak coupling beam
midspan axial force (kN)
ground
motion
peak coupling beam
estimate
(design)
standard
deviation
4730
257
mean
standard
deviation/mean
4204 4309 4033
4131
0.062
LA21 LA22 LA23c LA24 LA25a LA26 LA27 LA28 LA29 LA30
4231 4037 3588
3933 4110
4178 4506
4303
4586
4053 4211 3785 4086
4318 4326 3616 4200
c
b
TABLE 15.27
ground
LA01 LA02 LA03 LA04
motion
peak coupling beam
4670 3833 3947 3051
ground
LA11 LA12 LA13 LA14
motion
peak coupling beam
3431 4897 3851 5106
LA05 LA06c LA07 LA08 LA09 LA10b
standard
deviation
3048 2955 3363 3419 3845 5800
744
LA15 LA16a LA17 LA18 LA19 LA20
mean
standard
deviation/mean
3879 4079 4590 3830 3846 4577
4001
0.19
c
b
Fig. 15.23 plots the peak coupling beam midspan axial force demands from Tables
lines represent the design estimate (from Chapter 12) and the mean peak coupling beam
midspan axial force demand, respectively, for each data set. The mean peak coupling
beam midspan axial force demands under the survival-level and design-level ground
motions are similar; however, the design-level demands show a larger amount of scatter.
It is interesting to note that the peak coupling beam midspan axial force demands under
some of the design-level ground motion records are larger than the largest axial force
demand under the survival-level records.
491
4000
P1-CWUPT
2nd floor
LA25 (survival level)v
3000
2000
1000
0
6
8
10
time, t (sec.)
12
14
4000
estimate (design), Pby=3784 kN
3000
2000
P1-CWUPT
4th floor
1000
0
2
8
10
time, t (sec.)
12
14
3000
2000
P1-CWUPT
6th floor
0
0
8
10
time, t (sec.)
12
14
5000
4000
3000
2000
P1-CWUPT
8th floor
1000
0
0
8
10
time, t (sec.)
12
14
3000
2000
1000
0
2
8
10
time, t (sec.)
12
14
16
14
16
14
16
14
16
5000
4000
3000
2000
P1-CWUPT
5th floor
1000
0
0
16
P1-CWUPT
3rd floor
16
4000
estimate (design), 1.25Pby=4730 kN
4000
16
5000
1000
5000
16
coupling beam midspan axial force, Nb (kN)
5000
0
5000
8
10
time, t (sec.)
12
5000
4000
3000
2000
P1-CWUPT
7th floor
1000
0
0
8
10
time, t (sec.)
12
5000
4000
3000
2000
P1-CWUPT
roof
1000
0
0
8
10
time, t (sec.)
12
Fig. 15.22 Coupling beam midspan axial force time histories for Structure P1-CWUPT
492
peak coupling beam midspan axial force (kN)
6000
5000

mean=4131 kN
4000
3000
2000
P1-CWUPT
1000
0
0
50
100
150
200
250
300
350
(a)
6000
400
5000
4000
mean=4001 kN
3000
2000
P1-CWUPT
1000
0
0
50
100
150
200
(b)
Fig. 15.23 Peak coupling beam midspan axial force demands for Structure P1-CWUPT:
493

The peak coupling beam post-tensioning force demands from the dynamic analyses
of Structure P1-CWUPT under the twenty survival-level SAC ground motion records
and 15.29, respectively. The beam post-tensioning forces, Pb are calculated as the sum of
the forces in the truss elements representing the beam post-tensioning tendons at each
floor/roof level in the analytical model.
TABLE 15.28
PEAK COUPLING BEAM POST-TENSIONING FORCE DEMANDS FOR
ground
motion
peak coupling beam
post-tensioning force (kN)
ground
motion
peak coupling beam
estimate
(design)
standard
deviation
3784
351
LA31a LA32 LA33 LA34 LA35b LA36 LA37 LA38 LA39 LA40
mean
standard
deviation/mean
3574
3490
0.10
3658
3357 2913
3671 3181
3317 3627
3285 4052
3851 3354 3601 2940 3112
4022 3820 3877 2976 3611
c
b
TABLE 15.29
ground
LA01 LA02a LA03 LA04
motion
peak coupling beam
2914 2968 3038 2642
ground
LA11 LA12 LA13 LA14
motion
peak coupling beam
2865 2901 3201 3212

2704
2634
2851 2926 3009 2897
LA15 LA16b LA17 LA18 LA19 LA20 mean

3027
3431
2908 3030 2867 3302 2966
standard
deviation
204
standard
deviation/mean
0.069
c
b
The mean peak coupling beam post-tensioning force demand from the survivallevel dynamic analyses in Table 15.28 is close to and smaller than the total yield force of
the coupling beam post-tensioning tendons, Pby=3784 kN from Chapter 12; and thus, the
yielding of the beam post-tensioning tendons is prevented as required by design. As an
example, Fig. 15.24 shows the post-tensioning force time histories for the eight coupling
beams in Structure P1-CWUPT under the survival-level LA31 ground motion, for which
the peak coupling beam post-tensioning force demand (occurring in the roof beam) is
closest to the mean demand from the twenty records (see Table 15.28). The dashed
horizontal lines indicate the total yield force of the coupling beam post-tensioning
tendons, Pby. It is observed that there is almost no loss in the coupling beam posttensioning forces during the ground motion since the tendons do not to yield. The
494
smallest post-tensioning force develops in the second floor coupling beam, since the
fixed foundation conditions at the bases of the wall piers restrain the opening of gaps at
the beam ends, and thus, limit the elongation of the post-tensioning tendons during the
ground motion.
Fig. 15.25 plots the peak coupling beam post-tensioning force demands from
Tables 15.28 and 15.29 ( markers) against the maximum incremental velocity (MIV) of
the ground motion records (see Tables 14.2 and 14.1), where the dashed and solid
horizontal lines represent the design estimate (from Chapter 12) and the mean peak
coupling beam post-tensioning force demand, respectively, for each data set. It is
observed that yielding of the coupling beam post-tensioning tendons occurs under some
of the survival-level ground motion records, since the design roof drift from Chapter 12 is
exceeded (see Fig. 15.2).
15.2.12 Coupling Beam Shear Forces
The peak coupling beam shear force demands from the dynamic analyses of
and 15.31, respectively. The coupling beam shear forces, Vb are calculated based on the
equilibrium of the wall pier floor/roof level applied gravity loads and the wall pier axial
forces above and below each floor/roof level as determined from the fiber beam-column
elements used for the wall piers in the analytical model.
The peak coupling beam shear force demands in Tables 15.30 and 15.31 always
occur in the second floor beam. The mean peak beam shear force demand from the
survival-level dynamic analyses in Table 15.30 is 1223 kN, which is considerably larger
than the estimated demand of Qb,max=1.25Vb,pty=963 kN from Chapter 12. Thus, the force
used in Chapter 12 for the shear slip design of the coupling beam-to-wall interfaces does
not appear to be valid for the second floor level. Note that the design approach does not
require the estimation of the coupling shear force under the design-level earthquake; and
thus, no design-level estimate is provided in Table 15.31.
495
4000
coupling beam PT force, Pb (kN)
4000
3000
2000
P1-CWUPT
2nd floor
1000
0
5
10
time, t (sec.)
15
P1-CWUPT
3rd floor
1000
10
time, t (sec.)
15
20
15
20
15
20
15
20
4000
2000
20
4000
3000
2000
P1-CWUPT
4th floor
1000

3000
2000
P1-CWUPT
5th floor
1000
0
0
10
time, t (sec.)
15
20
10
time, t (sec.)
4000
4000
3000
0
0

3000
2000
P1-CWUPT
6th floor
1000

3000
2000
P1-CWUPT
7th floor
1000
0
0
10
time, t (sec.)
15
20
10
time, t (sec.)
4000
4000

3000
2000
P1-CWUPT
8th floor
1000
0
0
10
time, t (sec.)
15
20

3000
2000
P1-CWUPT
roof
1000
0
0
10
time, t (sec.)
Fig. 15.24 Coupling beam post-tensioning force time histories for Structure P1-CWUPT
496
4500
peak coupling beam PT force (kN)
4000
3500

mean=3490 kN
3000
2500
2000
1500
P1-CWUPT
1000
500
0
0
50
100
150
200
250
300
350
400
(a)
4500
4000
3500
3000
mean=2966 kN
2500
2000
1500
P1-CWUPT
1000
500
0
0
50
100
150
200
(b)
Fig. 15.25 Peak coupling beam post-tensioning force demands for Structure P1-CWUPT:
497
TABLE 15.30
PEAK COUPLING BEAM SHEAR FORCE DEMANDS FOR STRUCTURE
ground
motion
peak coupling beam
shear force (kN)
ground
motion
peak coupling beam
shear force (kN)
estimate
(design)
standard
deviation
963
151
LA31 LA32 LA33 LA34 LA35 LA36 LA37 LA38 LA39c LA40
mean
standard
deviation/mean
1286 1152
1223
0.12
LA21 LA22b LA23 LA24 LA25 LA26 LA27 LA28a LA29 LA30
1432 1659
1101 1115 1210 1335 1103 1232
1115 1141 1375 1423 1156 1136
1140
1087
1145
1114
c
b
TABLE 15.31
ground
LA01 LA02 LA03 LA04
motion
peak coupling beam
1089 1133 1127 1050
shear force (kN)
ground
LA11 LA12 LA13 LA14
motion
peak coupling beam
1107 1129 1091 1055
shear force (kN)
LA05 LA06c LA07 LA08 LA09a LA10
standard
deviation
1057 1019 1093 1086 1098 1059
63
mean
standard
deviation/mean
1065 1161 1067 1051 1314 1158
1100
0.057
c
b
As an example, Fig. 15.26 shows the shear force time histories for the eight
coupling beams in Structure P1-CWUPT under the survival-level LA28 ground motion,
for which the peak coupling shear force demand (occurring in the second floor beam) is
horizontal lines indicate the design estimate for the coupling beam shear force, Qb,max
from Chapter 12, which is equal to 1.25Vb,pty for the second and third floor beams and
Vb,pty for the upper level beams. It is observed that the peak shear force demands in the
upper level beams also exceed the design estimate. Since the coupling beam shear forces
from the static analyses in Chapter 13 are estimated reasonably well by the proposed
approximate procedures, the underestimation in the dynamic analysis results may be due
to the effect of higher modes on the peak coupling beam shear force demands; however
Fig. 15.27 plots the peak coupling beam shear force demands from Tables 15.30
represent the design estimate (from Chapter 12) and the mean peak coupling beam shear
force demand, respectively for each data set. It is observed that there is considerable
scatter in the peak coupling beam shear force demands under the survival-level ground
motions but not so much scatter under the design-level ground motions.
498
-1000
P1-CWUPT
2nd floor
-2000
coupling beam shear force, Vb (kN)
10
time, t (sec.)
15
-1000
P1-CWUPT
4th floor
-2000
5
10
time, t (sec.)
15
estimate (design), Qb,max = Vb,pty = 770 kN
1000
-1000
P1-CWUPT
6th floor
-2000
0
10
time, t (sec.)
15
-1000
P1-CWUPT
8th floor
-2000
0
10
time, t (sec.)
15
-1000
P1-CWUPT
3rd floor
-2000
10
time, t (sec.)
15
-1000
P1-CWUPT
5th floor
-2000
5
10
time, t (sec.)
15
20
2000
1000
-1000
P1-CWUPT
7th floor
-2000
0
10
time, t (sec.)
15
20
2000
1000
-1000
P1-CWUPT
roof
-2000
0
10
time, t (sec.)
15
Fig. 15.26 Coupling beam shear force time histories for Structure P1-CWUPT
499
20
1000
20
2000
20
2000
1000
20
2000
estimate (design), Qb,max = 1.25Vb,pty = 963 kN
1000
0
1000
2000
20
2000
1000
2000
20
2500
peak coupling beam shear force (kN)
P1-CWUPT
2000
1500
mean = 1223 kN
1000
500
0
0
50
100
150
200
250
300
350
400
(a)
2500
2000
P1-CWUPT
1500
1000
mean = 1100 kN
500
0
0
50
100
150
200
(b)
Fig. 15.27 Peak coupling beam shear force demands for Structure P1-CWUPT:
500
15.2.13 Coupling Beam Shear Slip Failure

The minimum coupling beam shear slip capacity to shear force demand ratio from
the dynamic analyses of Structure P1-CWUPT under the twenty survival-level SAC
ground motion records (LA21-LA40) and twenty design-level records (LA01-LA20) are
listed in Tables 15.32 and 15.33, respectively. The coupling beam shear slip capacity is
computed as Vb,ss=Vp,ss+Va,ss where, Va,ss is the slip capacity from the top and seat angleto-wall connections determined as described in Chapter 12 and Vp,ss is the slip capacity
due to the beam axial force calculated as the axial force at the beam end, Cb times the
assumed coefficient of friction, bw=0.33. Note that the axial force at the beam ends is
larger than the axial force at the beam midspan as a result of the top and seat angle forces.
The beam end axial forces are determined directly from the fiber beam-column elements
at the beam ends in the analysis model. The coupling beam shear slip capacity, Vb,ss
varies with time during an earthquake since Vp,ss due to the beam axial force varies. The
slip capacity from the top and seat angle-to-wall connections, Va,ss is assumed to remain
constant during an earthquake since the integrity of the angle-to-wall connections is
assumed to be maintained.
TABLE 15.32
MINIMUM COUPLING BEAM SHEAR SLIP CAPACITY-DEMAND RATIOS FOR
ground
LA21 LA22c LA23 LA24
motion
min. beam shear slip
0.75 0.42 1.48 0.99
capacity-demand ratio
ground
LA31 LA32 LA33 LA34
motion
1.00 0.79 1.12 1.13
LA25 LA26 LA27 LA28a LA29 LA30

1.31 1.15
1.40
1.10
1.21
1.27

0.96 0.85
1.54
1.17
1.55
0.94
estimate
(design)
standard
deviation
1.21
0.285
mean
standard
deviation/mean
1.11
0.26
The minimum ratio under this ground motion is closest to the mean ratio.
The minimum ratio under this ground motion is the largest ratio.
c
The minimum ratio under this ground motion is the smallest ratio.
b
TABLE 15.33
MINIMUM COUPLING BEAM SLIP CAPACITY-DEMAND RATIOS FOR
ground
motion
1.17 1.17 1.18 1.18
ground
LA11 LA12 LA13 LA14
motion
1.18 1.16 1.18 1.19
standard
deviation

1.18
1.18
1.18
1.18
1.18
1.18
0.018
LA15 LA16 LA17 LA18b LA19c LA20 mean

1.18
1.18
1.18
1.19
1.10
1.16
1.17
standard
deviation/mean
0.0156
The minimum ratio under this ground motion is closest to the mean ratio.
The minimum ratio under this ground motion is the largest ratio.
c
The minimum ratio under this ground motion is the smallest ratio.
b
The design estimate for the survival-level minimum coupling beam shear slip
capacity to shear force demand ratio in Table 15.32 is determined as the ratio between the
501
design estimates for the coupling beam shear slip capacity, Vb,ss and the beam shear force
demand, Qb,max, from Chapter 12. It is observed that the mean ratio from the survivallevel dynamic analysis results is smaller than the design estimate, and thus, the design
procedure to prevent shear slip failure at the beam ends is not conservative. This is
possibly because of the underestimated coupling beam shear force demands as described
previously.
The mean values of the minimum coupling beam shear slip capacity to shear force
demand ratios in Tables 15.32 and 15.33 are both larger than 1.0. Thus, shear slip failure
at the ends of the coupling beams in Structure P1-CWUPT, on average, does not occur,
even though the design approach underestimates the peak coupling beam shear force
demands from the dynamic analyses (see Fig. 15.27).
As an example, Fig. 15.28 shows comparisons between the shear slip capacity
(solid lines) and shear force demand (dashed lines) time histories for the eight coupling
beams in Structure P1-CWUPT under the survival-level LA28 ground motion, for which
the minimum coupling beam shear slip capacity-demand ratio is closest to the mean ratio
from the twenty records (see Table 15.32). For comparison, the solid and dashed
horizontal lines indicate the design estimates for the coupling beam shear slip capacity,
Vb,ss and shear force demand, Qb,max=1.25Vb,pty (second and third floor level beams) or
Qb,max=Vb,pty (upper level beams), respectively. As described in Chapter 12, the design
estimate for the coupling beam shear slip capacity due to the beam end axial force, Vp,ss is
computed as the coupling beam initial post-tensioning force, Pbi times the assumed
coefficient of friction at the beam ends, bw=0.33. In addition to the underestimated
maximum shear force demands, the design approach seems to overestimate the minimum
shear slip capacities from the dynamic analysis, also contributing to the unconservative
design to prevent shear slip failure at the beam ends.
502
1500
estimate (design), Vb,ss =1169 kN
1000
estimate (design), Qb,max =963 kN
500
0
0
10
time, t (sec.)
15
20
2500
P1-CWUPT
4th floor
LA28
2000
capacity
demand
1500
1000
500
0
0
2500
10
time, t (sec.)
P1-CWUPT
6th floor
LA28
2000
15
20
capacity
demand
1500
1000
500
0
0
10
time, t (sec.)
2500
20
capacity
demand
P1-CWUPT
8th floor
LA28
2000
15
1500
1000
500
0
0
10
time, t (sec.)
15
20
beam shear slip capacity and demand (kN)
2000
capacity
demand
beam shear slip capacity and demand (kN) beam shear slip capacity and demand (kN)
P1-CWUPT
2nd floor
LA28

beam shear slip capacity and demand (kN) beam shear slip capacity and demand (kN)
2500
2500
P1-CWUPT
3rd floor
LA28
2000
capacity
demand
1500
1000
500
0
0
10
time, t (sec.)
15
20
2500
P1-CWUPT
5th floor
LA28
2000
1500
capacity
demand
1000
500
0
0
2500
10
time, t (sec.)
15
P1-CWUPT
7th floor
LA28
2000
1500
20
capacity
demand
1000
500
0
0
10
time, t (sec.)
15
2500
capacity
demand
P1-CWUPT
roof
LA28
2000
1500
20
1000
500
0
0
10
time, t (sec.)
15
Fig. 15.28 Coupling beam shear slip capacity and shear force demand time histories
503
20
It is observed in Fig. 15.28 that the second floor coupling beam shear force demand
exceeds the design estimate Qb,max; however, shear slip does not occur during the
earthquake. This is because, the times during the ground motion when the shear force
demand exceeds the design estimate also correspond to times when the shear slip
capacity exceeds the design estimate Vb,ss due to significant increases in the beam axial
force (which are conservatively ignored in the design approach in Chapter 12). The
results in Fig. 15.28 show that shear slip is not critical at the ends of the upper floor
coupling beams.
Fig. 15.29 plots the minimum coupling beam shear slip capacity-demand ratios
from Tables 15.32 and 15.33 ( markers) against the maximum incremental velocity
(MIV) of the ground motion records (see Tables 14.2 and 14.1), where the dashed and
solid horizontal lines represent the design estimate (from Chapter 12) and the mean ratio,
respectively, for each data set. Note that while shear slip at the ends of the coupling
beams is not expected to occur under any of the design-level ground motion records,
shear slip at the ends of the coupling beams (especially the second floor beams) may
occur under some of the survival-level ground motion records. It is also interesting to
note that there is little scatter or variation in the minimum coupling beam shear slip
capacity-demand ratio under the design level ground motions and that the mean ratios
under the design-level and survival-level ground motion sets are similar.
15.2.14 Angle Deformations
The peak tensile deformation demands in the top and seat connection angles of the
coupling beams from the dynamic analyses of Structure P1-CWUPT under the twenty
survival-level SAC ground motion records (LA21-LA40) and twenty design-level records
(LA01-LA20) are listed in Tables 15.34 and 15.35. The angle deformations, ax are
determined directly from the horizontal angle elements in the analytical model as the
deformations parallel to the coupling beams (i.e., horizontal displacement of angle heel
from wall).
The mean peak angle tensile deformation demand from the survival-level dynamic
analyses in Table 15.34 is 33 mm, which is larger than the estimated deformation of
uas=30 mm from Chapter 12, despite the over-estimated coupling beam chord rotation
demands in Fig. 15.21. Thus, the method used in Chapter 12 to estimate the angle
deformations needs to be improved.
504
minimum coupling beam shear slip capacitydemand ratio
2
1.8
1.6
1.4
estimate (design) =1.25
1.2
1
mean=1.11
0.8
0.6
0.4
0.2
0
0
P1-CWUPT
50
100
150
200
250
300
350
400
minimum coupling beam shear slip capacitydemand ratio
(a)
2
1.8
1.6
1.4
1.2
mean=1.17
1
0.8
P1-CWUPT
0.6
0.4
0.2
0
0
50
100
150
200
(b)
Fig. 15.29 Minimum coupling beam shear slip capacity to shear force demand ratios
505
TABLE 15.34
PEAK ANGLE TENSILE DEFORMATION DEMANDS FOR STRUCTURE
ground
estimate
standard
LA21 LA22 LA23c LA24 LA25 LA26 LA27 LA28a LA29 LA30
motion
(design)
deviation
peak angle tensile
35
26
12
25
34
46
26
33
14
19
30
15
deformation (mm)
ground
standard
LA31 LA32 LA33 LA34 LA35b LA36 LA37 LA38 LA39 LA40 mean
motion
deviation/mean
peak angle tensile
33
36
21
24
68
65
43
49
14
33
33
0.47
deformation (mm)
c
b
TABLE 15.35
ground
estimate
standard
LA01 LA02 LA03 LA04 LA05 LA06c LA07 LA08a LA09 LA10
motion
(design)
deviation
peak angle tensile
19
14
10
4
6
2
10
13
14
12
13
7
deformation (mm)
ground
standard
motion
deviation/mean
peak angle tensile
8
7
21
22
16
28
12
16
10
24
13
0.52
deformation (mm)
a
c
b
As an example, Fig. 15.30 shows the deformation time histories, ax for the bottom
(i.e., seat) angles at the left ends of the eight coupling beams in Structure P1-CWUPT
under the survival-level LA28 ground motion, for which the peak angle tensile
deformation demand (occurring in the roof beam) is closest to the mean demand from the
twenty records (see Table 15.34). The dashed horizontal lines indicate the design estimate
for the peak angle tensile deformation, uas, from Chapter 12.
Fig. 15.31 plots the peak angle tensile deformation demands from Tables 15.34 and
represent the design estimate (from Chapter 12) and the mean peak angle tensile
deformation demand, respectively, for each data set. It is observed that the peak angle
tensile deformation demands tend to follow similar trends to the peak coupled wall roof
drift demands in Fig. 15.2 and the peak coupling beam chord rotation demands in Fig.
15.21.
506
40
angle deformation, ax (mm)
40
estimate (design), uas =30 mm
30
P1-CWUPT
2nd floor
left end seat angle
LA28
20
10
0
10
5
10
time, t (sec.)
15
10
time, t (sec.)
15
20
40
10
20
40
30
P1-CWUPT
4th floor beam
left end seat angle
LA28
20
10
0
10
30
P1-CWUPT
5th floor beam
left end seat angle
LA28
20
10
0
10
10
time, t (sec.)
15
20
10
time, t (sec.)
15
20
40
40
P1-CWUPT
3rd floor beam
left end seat angle
LA28
20
10
0
30
20
P1-CWUPT
6th floor beam
left end seat angle
LA28
10
0
10
30
20
P1-CWUPT
7th floor beam
left end seat angle
LA28
10
0
10
10
time, t (sec.)
15
20
10
time, t (sec.)
15
20
40
40
30
30
P1-CWUPT
8th floor beam
left end seat angle
LA28
20
10
0
10
30
P1-CWUPT
roof beam
left end seat angle
LA28
20
10
0
10
10
time, t (sec.)
15
20
10
time, t (sec.)
15
Fig. 15.30 Coupling beam connection angle deformation time histories

507
20
70
P1-CWUPT
peak angle tensile deformation (mm)
60
50
40
mean=33 mm
30
estimate (design), uas=30 mm
20
10
0
0
50
100
150
200
250
300
350
400
(a)
70
60
P1-CWUPT
50
40
30
20
estimate (design), uad=13 mm

mean=13 mm
10
0
0
50
100
150
200
(b)
Fig. 15.31 Peak coupling beam connection angle tensile deformation demands
508
15.3 Structure P2-PWUPT

This section discusses the results from the nonlinear dynamic time-history analyses
of Structure P2-PWUPT as follows: (1) floor/roof drifts; (2) inter-story drifts; (3)
floor/roof accelerations; (4) wall post-tensioning forces; (5) wall base axial forces; (6)
wall base shear forces; (7) wall base diagonal tension and shear slip failure; (8) wall base
moments; (9) wall base strains; (10) coupling beam chord rotations; (11) coupling beam
axial forces; (12) coupling beam post-tensioning forces; (13) coupling beam shear forces
failure; (14) coupling beam shear slip failure; and (15) angle deformations.
15.3.1 Floor/Roof Drifts
The peak coupled wall roof drift demands from the dynamic analyses of Structure
P2-PWUPT under the twenty survival-level SAC ground motion records (LA21-LA40)
respectively. As described before, the roof drift, is calculated as the average lateral
displacement of the left and right wall piers at the roof divided by the wall height to the
base, hw.
TABLE 15.36
PEAK ROOF DRIFT DEMANDS FOR STRUCTURE P2-PWUPT
ground
motion
peak roof drift
(%)
ground
motion
peak roof drift
(%)
LA21 LA22 LA23 LA24a LA25 LA26 LA27 LA28 LA29 LA30
2.80
1.47
0.73
1.81
1.72
2.53
1.31
2.16
0.84
0.97
LA31 LA32 LA33 LA34 LA35 LA36b LA37 LA38 LA39c LA40
2.29
2.31
1.48
1.04
3.38
3.56
2.08
2.34
0.67
1.53
estimate
(design)
standard
deviation
1.71
0.84
mean
standard
deviation/mean
1.84
0.45
c
b
TABLE 15.37
PEAK ROOF DRIFT DEMANDS FOR STRUCTURE P2-PWUPT
ground
motion
peak roof drift
0.69 0.76 0.74 0.31
(%)
ground
LA11 LA12 LA13 LA14
motion
peak roof drift
0.62 0.55 1.03 1.16
(%)

0.35
0.28
0.53 0.58
1.03 0.61

0.85
1.46
0.88 0.84
0.61 1.07
c
b
509
estimate
(design)
standard
deviation
0.79
0.30
mean
standard
deviation/mean
0.75
0.40
The mean peak roof drift demand from the survival-level dynamic analyses in
Table 15.36 is 1.84%, which is close to the estimated survival-level roof drift demand of
s=1.71% and the allowable target roof drift demand of ts=1.75% from Chapter 12. As
an example, Fig. 15.32 shows the roof drift time-history of Structure P2-PWUPT under
the LA24 ground motion, for which the peak roof drift demand is closest to the mean
demand from the twenty records (see Table 15.37). The dashed horizontal lines indicate
the design estimate for the peak roof drift demand, s from Chapter 12.
1
P2-PWUPT
-1
-2
0
10
time, t (sec.)
15
20
Fig. 15.32 Roof drift time history for Structure P2-PWUPT

Comparing Fig. 15.32 for Structure P2-PWUPT with Fig. 15.1 for Structure P1CWUPT, it is observed that the oscillations of both structures are generally centered
around the original zero-drift position; however, Structure P2-PWUPT appears to have a
slightly larger self-centering capability than Structure P1-CWUPT. This is also true under
the other ground motion records included in the investigation. The increased selfcentering capability in Structure P2-PWUPT occurs as a result of the post-tensioning
steel used as flexural reinforcement in the precast concrete wall piers.
The mean peak roof drift demand from the design-level analyses of Structure P2PWUPT in Table 15.37 is 0.75%, which is close to the estimated design-level roof drift
demand of d=0.79% and smaller than the allowable target roof drift demand of
td=1.17% from Chapter 12. As for Structure P1-CWUPT, it is concluded that the
displacement-based design objectives for Structure P2-PWUPT under the survival-level
and design-level ground motions are, on average, satisfied.
Fig. 15.33 plots the peak coupled wall roof drift demands from Tables 15.36 and
motion records (see Tables 14.2 and 14.1 for the MIV values of the LA21-LA40 and the
LA01-LA20 records, respectively), where the dashed and solid horizontal lines represent
510
the design estimate (from Chapter 12) and the mean peak roof drift demand, respectively,
for each data set. Similar to Structure P1-CWUPT, it is observed that the peak roof drift
demands for Structure P2-PWUPT tend to be larger under ground motions with larger
MIV (i.e., there seems to be a correlation between MIV and the peak roof drift demand,
especially for the survival-level records). Furthermore, there is a significant amount of
scatter in the data indicating that, while the design objectives are satisfied on average, the
peak roof drift demand may be significantly larger or significantly smaller than the
design estimate, depending on the earthquake. This is investigated in more detail in
Kurama and Farrow (2003).
15.3.2 Inter-Story Drifts
The peak coupled wall inter-story drift demands from the dynamic analyses of
Structure P2-PWUPT under the twenty survival-level SAC ground motion records
and Table 15.39, respectively. The inter-story drift is defined as the relative lateral
displacement between two adjacent floor/roof levels in a structure divided by the story
height and is calculated as the average inter-story drift of the left and right wall piers.
The mean peak inter-story drift demand from the twenty survival-level analyses in
Table 15.38 is 2.04%, which is close to the estimated survival-level inter-story drift
demand of 1.97% (occurring in the eighth story) from Chapter 12. As an example, Fig.
15.34 shows the inter-story drift time histories of Structure P2-PWUPT under the LA24
ground motion, for which the peak inter-story drift demand is closest to the mean demand
from the twenty records (see Table 15.38). It is observed that the drifts of all eight stories
are in phase, and thus, the deflected shape of the structure is governed by the first
vibration mode. As expected, the peak inter-story drift demand (2.04%) occurs in the
eighth (top) story (indicating a flexure dominated deflected shape) and is larger than the
corresponding peak roof drift demand of 1.81% from Table 15.36. The dashed horizontal
lines show the design estimate for the peak inter-story drift demand in the eighth story
from Chapter 12.
The mean peak inter-story drift demand from the twenty design-level analyses in
Table 15.39 is 0.88%, which is close to the estimated design-level inter-story drift
demand of 0.90% from Chapter 12. As discussed in Chapter 12, according to IBC 2000
(ICC 2000), the peak inter-story drift demand under the design-level ground motion
should not exceed 2%. Because of the relatively small allowable target roof drift demands
used in design (ts=1.75% and td=1.17%), the peak inter-story drift demands in Tables
15.38 and 15.39 satisfy the 2% limit not only for the design-level ground motions, but
also for many of the survival-level motions.
511
4
3.5
P2-PWUPT
3
2.5
2
1.5
mean=1.84%
1
0.5
0
0
50
100
150
200
250
300
350
400
(a)
4
3.5
P2-PWUPT
2.5
2
1.5
1
estimate (design), d=0.79%

mean=0.75%
0.5
0
0
50
100
150
(b)
Fig. 15.33 Peak roof drift demands for Structure P2-PWUPT:
512
200
TABLE 15.38
PEAK INTER-STORY DRIFT DEMANDS FOR STRUCTURE P2-PWUPT
ground
LA21 LA22 LA23 LA24a
motion
peak inter-story
2.95 1.76 0.83 2.04
drift (%)
ground
LA31 LA32 LA33 LA34
motion
peak inter-story
2.63 2.50 1.64 1.17
drift (%)

1.91
2.70
1.48
2.36
0.95
1.15
LA35 LA36b LA37 LA38 LA39c LA40

3.62
3.80
2.25
2.52
0.81
1.71
estimate
(design)
standard
deviation
1.97
0.87
mean
standard
deviation/mean
2.04
0.43
c
b
TABLE 15.39
PEAK INTER-STORY DRIFT DEMANDS FOR STRUCTURE P2-PWUPT
ground
motion
peak inter-story
0.83 0.89 0.83 0.38
drift (%)
ground
LA11 LA12 LA13 LA14
motion
peak inter-story
0.74 0.78 1.28 1.35
drift (%)

0.42
0.35
0.61
0.70
1.18
0.71

0.99
1.66
1.03
0.95
0.70
1.26
estimate
(design)
standard
deviation
0.90
0.34
mean
standard
deviation/mean
0.88
0.39
c
b
Note that, for both ground motion sets, the ratio between the mean and estimated
peak inter-story drift demands is similar to the ratio between the mean and estimated peak
roof drift demands. In general, the trends observed for the peak roof drift demands in the
previous section are also valid for the peak inter-story drift demands.
Fig. 15.35 plots the peak coupled wall inter-story drift demands from Tables 15.38
represent the design estimate (for the eighth story, from Chapter 12) and the mean peak
inter-story drift demand, respectively, for each data set. Similar to the peak roof drift
demands discussed in the previous section, there is a significant amount of scatter in the
peak inter-story drift demands, and the demands tend to be larger under the ground
motions with larger MIV (i.e., there seems to be a correlation between MIV and the peak
inter-story drift demand, especially for the survival level records).
513
1
0
-1
-2
P2-PWUPT
first story
-3
10
time, t (sec.)
15
-1
-2
20
1
0
P2-PWUPT
third story
-1
-2
-3
P2-PWUPT
second story
10
time, t (sec.)
15
20
15
20
15
20
1
0
-1
-2
P2-PWUPT
fourth story
-3
0
10
time, t (sec.)
15
20
-3
0
1
0
-1
-2
P2-PWUPT
fifth story
-3
10
time, t (sec.)
1
0
-1
-2
P2-PWUPT
sixth story
-3
0
10
time, t (sec.)
15
20
1
0
-1
P2-PWUPT
seventh story
-2
-3
0
10
time, t (sec.)
15
20
10
time, t (sec.)
1
0
-1
P2-PWUPT
eighth story
-2
-3
0

5
10
time, t (sec.)
Fig. 15.34 Inter-story drift time histories for Structure P2-PWUPT
514
15
20
3.5
P1-CWUPT
3
2.5
mean=2.04%
2
1.5
1
0.5
0
50
100
150
200
250
300
350
400
(a)
4
P2-PWUPT
3.5
3
2.5
2
1.5
mean=0.88%
0.5
0
50
100
150
200
(b)
Fig. 15.35 Peak inter-story drift demands for Structure P2-PWUPT:
515
15.3.3 Floor/Roof Accelerations

The peak absolute floor/roof acceleration demands from the dynamic analyses of
and 15.41, respectively. The accelerations are calculated as the average total absolute
lumped mass accelerations (i.e., relative accelerations with respect to the ground plus the
ground acceleration) at the floor/roof levels of the left and right wall piers.
TABLE 15.40
PEAK FLOOR/ROOF ACCELERATION DEMANDS FOR STRUCTURE P2-PWUPT
ground
LA21b LA22 LA23 LA24 LA25
motion
peak floor/roof
2.59 2.28 1.36 1.88 1.78
acceleration (g)
ground
LA31 LA32 LA33 LA34 LA35
motion
peak floor/roof
2.27 2.36 1.90 1.53 2.34
acceleration (g)
standard
deviation
LA26a LA27c LA28 LA29 LA30

1.98
1.17
2.25
2.42
2.57
0.48
LA36 LA37 LA38 LA39 LA40 mean

2.58
1.46
1.81
1.08
1.51
standard
deviation/mean
1.96
0.25
c
b
TABLE 15.41
PEAK FLOOR/ROOF ACCELERATION DEMANDS FOR STRUCTURE P2-PWUPT
ground
motion
peak floor/roof
acceleration (g)
ground
motion
peak floor/roof
acceleration (g)
standard
Deviation
LA01 LA02 LA03a LA04 LA05 LA06c LA07 LA08 LA09 LA10
1.57
1.81
1.48
0.89
0.70
0.62
1.51
0.88
1.32
1.12
0.55
LA11 LA12 LA13 LA14 LA15 LA16 LA17 LA18 LA19 LA20b mean
1.63
2.30
1.67
1.21
1.30
1.31
1.01
1.68
2.09
2.88
1.45
standard
deviation/mean
0.38
c
b
As an example, Fig. 15.36 shows the floor/roof acceleration time histories of

Structure P2-PWUPT under the survival-level LA22 ground motion, for which the peak
acceleration demand (occurring at the roof level in this case) is closest to the mean
demand from the twenty records (see Table 15.40). It is noted that the peak acceleration
demand does not always occur at the roof level.
516
3
3
P2-PWUPT
second floor
2
1
0
-1
-2
-3
0
P2-PWUPT
third floor
2
1
0
-1
-2
-3
10
15
time, t (sec.)
P2-PWUPT
fourth floor
2
1
0
-1
-2
P2-PWUPT
fifth floor
2
1
0
-1
-2
-3
10
15
time, t (sec.)
1
0
-1
-2
P2-PWUPT
seventh floor
2
1
0
-1
-2
-3
10
15
time, t (sec.)
10
15
time, t (sec.)
3
3
P2-PWUPT
eighth floor
absolute roof acceleration (g)
15
3
P2-PWUPT
sixth floor
1
0
-1
-2
-3
0
10
time, t (sec.)
-3
0
15
-3
0
10
time, t (sec.)
P2-PWUPT
roof
2
1
0
-1
-2
-3
10
15
time, t (sec.)
10
time, t (sec.)
Fig. 15.36 Floor/roof acceleration time histories for Structure P2-PWUPT
517
15
Fig. 15.37 plots the peak floor/roof acceleration demands ( markers) from Tables
15.40 and 15.41 against the maximum incremental velocity (MIV) of the ground motion
records, as well as the peak ground accelerations (PGA, markers) of the records (see
Tables 14.2 and 14.1). The solid horizontal lines represent the mean peak floor/roof
acceleration demand and the mean peak ground acceleration for each data set. As
expected, the peak floor/roof acceleration demands are larger than the peak ground
accelerations. There is a significant amount of scatter in the peak floor/roof acceleration
demands; however, unlike the peak roof drift and peak inter-story drift demands
discussed in the previous sections, no significant correlation exists between the peak
floor/roof acceleration demand and the MIV. It is also interesting to note that the designlevel LA20 ground motion record results in a larger peak floor/roof acceleration demand
than all of the other records, including the survival-level records. Comparing Figs. 15.37
and 15.6, it is observed that the peak floor/roof acceleration demands of Structure P2PWUPT are, on average, larger than the peak acceleration demands of Structure P1CWUPT.
15.3.4 Wall Post-Tensioning Forces
The minimum total coupled wall post-tensioning forces from the dynamic analyses
of Structure P2-PWUPT under the twenty survival-level SAC ground motion records
and 15.43, respectively. Similarly, the maximum total coupled wall post-tensioning
forces from the dynamic analyses of the structure under the survival-level and designlevel ground motion records are listed in Tables 15.44 and 15.45, respectively. The total
coupled wall post-tensioning forces are calculated as the sum of the forces in the truss
elements representing the left-side and right-side wall pier post-tensioning bars in the
analytical model.
The mean minimum total coupled wall post-tensioning force from the survivallevel dynamic analyses in Table 15.42 is 34077 kN, which is close to the estimated
survival-level minimum total coupled wall post-tensioning force of 33594 kN from
Chapter 12 (see Section 12.8.5 and Table 12.30). Thus, the design procedure to estimate
the prestress losses in the wall pier post-tensioning bars appears to be valid. Under the
design-level earthquake, the minimum total coupled wall post-tensioning force is
assumed to be equal to the total initial post-tensioning force, 2Pwi=37758 kN from
Chapter 12, which is similar to the mean minimum coupled wall post-tensioning force
from the design-level dynamic analyses in Table 15.43.
518


2.5
2 mean floor/roof acceleration=1.96 g
1.5
1 mean PGA=0.87 g
0.5
0
0
P2-PWUPT
50
100
150
200
250
300
350
400
(a)

2.5

LA20
1.5
mean floor/roof acceleration=1.45 g
1
mean PGA=0.59 g
0.5
P2-PWUPT
0
0
50
100
150
200
(b)
Fig. 15.37 Peak floor/roof acceleration demands for Structure P2-PWUPT:
519
TABLE 15.42
MINIMUM TOTAL COUPLED WALL POST-TENSIONING FORCES FOR
STRUCTURE P2-PWUPT UNDER SURVIVAL LEVEL GROUND MOTIONS
ground
motion
min. coupled wall posttensioning force (kN)
ground
motion
estimate
(design)
standard
deviation
29247 36167 37164 35136 36066 32258 36867 33147 37086 36988 33594
LA31 LA32 LA33 LA34 LA35 LA36c LA37 LA38a LA39b LA40
mean
3791
standard
deviation/mean
32460 32972 36444 37096 26383 23662 34951 34049 37178 36223 34077
0.11
The force under this ground motion is closest to the mean force.
The force under this ground motion is the largest force.
c
The force under this ground motion is the smallest force.
b
TABLE 15.43
MINIMUM TOTAL COUPLED WALL POST-TENSIONING FORCES FOR
STRUCTURE P2-PWUPT UNDER DESIGN LEVEL GROUND MOTIONS
ground
motion
ground
motion
estimate
(design)
standard
deviation
37178 37121 37202 37271 37267 37278 37221 37220 37132 37221
37758
143
LA11 LA12 LA13 LA14 LA15a LA16c LA17 LA18 LA19 LA20
mean
standard
deviation/mean
37203 37133 37084 37034 37138 36629 37175 37132 37148 36965
37138
0.0039
c
b
TABLE 15.44
MAXIMUM TOTAL COUPLED WALL POST-TENSIONING FORCES FOR
ground
motion
max. coupled wall posttensioning force (kN)
ground
motion
estimate
(idealized)
standard
deviation
42498 41369 39116 41969 42280 43575 41196 43130 39465 40229
44782
1619
LA31 LA32 LA33a LA34 LA35 LA36b LA37 LA38 LA39c LA40
mean
standard
deviation/mean
42390 42428 41897 40335 44102 44183 43304 43473 38702 41680
41866
0.039
The peak force under this ground motion is closest to the mean force.
The peak force under this ground motion is the largest force.
c
The peak force under this ground motion is the smallest force.
b
520
TABLE 15.45
MAXIMUM TOTAL COUPLED WALL POST-TENSIONING FORCES FOR
ground
motion
ground
motion
LA01 LA02 LA03a LA04 LA05 LA06c LA07 LA08 LA09 LA10
estimate
(idealized)
standard
deviation
38941 39370 39056 37600 37634 37427 38274 38300 40111 38487
41531
1087
mean
standard
deviation/mean
38448 38046 40029 40403 39413 41453 39464 39399 38511 40576
39047
0.028
c
b
Since the proposed design approach in Chapter 12 does not require the estimation
of the survival-level and design-level maximum total coupled wall post-tensioning forces,
the results from the dynamic analyses in Tables 15.44 and 15.45 are compared with the
results from the static idealized tri-linear pushover analysis of the structure in Chapter 13.
The sum of the tension-side and compression-side wall pier post-tensioning forces, Ptw
and Pcw, corresponding to the estimated coupled wall ultimate state and PT-yielding state
are given in Tables 15.44 and 15.45, respectively (see also Table 13.2). The results show
that the total coupled wall post-tensioning forces from the idealized pushover analysis
provide reasonable estimates for the maximum post-tensioning forces from the dynamic
analyses.
As an example, Fig. 15.38 shows the total coupled wall post-tensioning force timehistories of Structure P2-PWUPT under the survival level LA38 and LA33 ground
motions, for which the minimum and maximum total coupled wall post-tensioning forces
are closest to the corresponding mean forces from Tables 15.42 and 15.44, respectively.
The dashed horizontal lines indicate the design estimate for the minimum post-tensioning
force from Chapter 12 and the estimated idealized maximum post-tensioning force from
the tri-linear static pushover analysis in Chapter 13.
Fig. 15.39 plots the minimum ( markers) and maximum ( markers) total
coupled wall post-tensioning forces from Tables 15.42-15.45 against the maximum
incremental velocity (MIV) of the ground motion records (see Tables 14.2 and 14.1),
where the dashed and solid horizontal lines represent the design/idealized estimate and
the mean total coupled wall post-tensioning force, respectively, for each data set. It is
observed that there is very little scatter or variation in the design-level minimum total
coupled wall post-tensioning forces since yielding of the wall pier post-tensioning bars is
prevented under the design level earthquake, and thus, the minimum post-tensioning
force is equal to the initial force. In comparison, significant prestress losses occur in the
wall pier post-tensioning bars under the survival-level ground motions, which should be
considered in the design of the structure against shear slip failure at the base as discussed
in more detail later.
521
50000
max. estimate (idealized)=44782 kN
40000
min. estimate (design)=33594 kN
30000
20000
P2-PWUPT
10000
00
10
15
20
25
total coupled wall post-tensioning force (kN)
total coupled wall post-tensioning force (kN)
50000
30
max. estimate (idealized)=44782 kN

40000
min. estimate (design)=33594 kN
30000
20000
P2-PWUPT
10000
00
10
15
20
25
30
time, t (sec.)
time, t (sec.)
(a)
(b)
Fig. 15.38 Total coupled wall post-tensioning force time histories

for Structure P2-PWUPT: (a) LA38; (b) LA33
15.3.5 Wall Base Axial Forces
As a result of the wall pier post-tensioning forces, the axial forces in both wall piers
remain compressive during the dynamic analyses of Structure P2-PWUPT. The minimum
wall pier base compressive axial force demands under the twenty survival-level SAC
listed in Tables 15.46 and 15.47, respectively. Similarly, Tables 15.48 and 15.49 give the
maximum wall pier base compressive axial force demands under the twenty survivallevel and design-level records, respectively. The minimum wall pier base axial force
demands are important in the design of the structure to prevent shear slip failure as
described later. The wall pier base axial forces in Tables 15.46-15.49 are obtained
directly from the fiber beam-column elements used for the wall piers in the analysis
model, where a positive force indicates compression.
Since the proposed design approach does not require the estimation of the survivallevel and design-level wall pier axial forces, the results from the dynamic analyses are
compared with the results from the static idealized tri-linear pushover analysis of the
structure in Chapter 13. The tension-side wall pier base axial forces, Ntwu and Ntwy,
corresponding to the estimated coupled wall ultimate state and PT-yielding state are
given in Tables 15.46 and 15.47, respectively (see also Table 13.2). Similarly, the
compression-side wall pier base axial forces, Ncwu and Ncwy, corresponding to the
estimated coupled wall ultimate state and PT-yielding state are given in Tables 15.48 and
15.49, respectively. The results show that the Ncwu and Ncwy forces from the idealized
pushover analysis are reasonably close to the maximum wall pier compressive axial force
demands from the dynamic analyses; however, the Ntwu and Ntwy forces significantly
overestimate the minimum wall pier compressive axial force demands. This is possibly
because of the underestimated beam shear forces as described later in more detail.
522
minimum and maximum total wall post-tensioning forces (kN)
50000
estimate (idealized)=44782 kN
40000 mean=41866 kN
mean=34077 kN
estimate (design)=33594 kN
30000
20000
minimum total coupled wall PT force
maximum total coupled wall PT force
10000
P2-PWUPT
0
50
250
300
350
100
150
200
400
minimum and maximum total wall post-tensioning forces (kN)
(a)
50000
mean=39047 kN estimate (idealized)=41531 kN
40000
mean=37138 kN
estimate (design)=37758 kN
30000
20000
minimum total coupled wall PT force
maximum total coupled wall PT force
10000
P2-PWUPT
0
50
100
150
200
(b)
Fig. 15.39 Minimum and maximum total coupled wall post-tensioning forces
for Structure P2-PWUPT: (a) survival level; (b) design level
523
TABLE 15.46
MINIMUM WALL PIER BASE COMPRESSION AXIAL FORCE DEMANDS FOR
ground
motion
min. wall pier base
ground
motion
min. wall pier base
LA21 LA22 LA23b LA24 LA25 LA26 LA27 LA28a LA29 LA30
estimate
(idealized)
standard
deviation
11142 12690 13237 12819 12890 12241 13139 12299 13006 12899
16489
912
LA31 LA32 LA33 LA34 LA35 LA36c LA37a LA38 LA39 LA40
mean
standard
deviation/mean
11934 12018 12864 13219 10217 10141 12299 12245 13193 12695
12359
0.074
c
b
TABLE 15.47
MINIMUM WALL PIER BASE COMPRESSION AXIAL FORCE DEMANDS FOR
ground
motion
min. wall pier base
ground
motion
min. wall pier base
LA01 LA02 LA03 LA04 LA05 LA06b LA07 LA08 LA09 LA10a
estimate
(idealized)
standard
deviation
13028 13113 13366 14701 14634 15123 13531 13184 13148 13393
16039
631
mean
standard
deviation/mean
13486 13486 12828 13099 12975 13162 13068 13122 13322 12957
13436
0.047
c
b
TABLE 15.48
MAXIMUM WALL PIER BASE COMPRESSION AXIAL FORCE DEMANDS FOR
ground
motion
max. wall pier base
ground
motion
max. wall pier base
LA21 LA22 LA23 LA24 LA25a LA26 LA27 LA28 LA29 LA30
estimate
(idealized)
standard
deviation
36078 33885 31625 34868 35157 37581 33622 36411 32066 32132
32742
2447
mean
standard
deviation/mean
35686 36500 34579 32724 39440 39618 36652 37101 31025 34254
35050
0.070
c
b
524
TABLE 15.49
MAXIMUM WALL PIER BASE COMPRESSION AXIAL FORCE DEMANDS FOR
ground
motion
max. wall pier base
ground
motion
max. wall pier base
LA01a LA02 LA03 LA04 LA05 LA06c LA07 LA08 LA09 LA10
estimate
(idealized)
standard
deviation
31305 31603 31478 28325 28779 27982 30429 30816 32773 30785
31132
1599
mean
standard
deviation/mean
30665 29597 32751 32777 31999 33867 32146 31839 30945 33098
31198
0.051
c
b
As an example, Fig. 15.40 shows the axial force time histories at the bases of the
left-side and right-side wall piers in Structure P2-PWUPT under the survival-level LA28
ground motion, for which the minimum wall pier base compressive axial force demand
(occurring in the right-side wall pier) is closest to the mean demand from the twenty
records (see Table 15.46). Similarly, Fig. 15.41 shows the axial force time histories at the
bases of the left-side and right-side wall piers in Structure P2-PWUPT under the survivallevel LA25 ground motion, for which the maximum wall pier base compressive axial
force demand (occurring in the right-side wall pier) is closest to the mean demand from
the twenty records (see Table 15.48). The dashed horizontal lines indicate the estimated
idealized wall pier base axial forces, Ntwu and Ncwu, from the tri-linear static pushover
analysis in Chapter 13 at the coupled wall ultimate state.
40000
30000
20000
10000 estimate (idealized), N =16489 kN P2-PWUPT
twu
left-side wall
0
0
5
10
15
20
time (sec.)
(a)
40000
30000
20000
10000 estimate (idealized), N =16489 kN P2-PWUPT
twu
right-side wall
0
0
5
10
15
20
time (sec.)
(b)
Fig. 15.40 Wall pier base axial force time histories for Structure P2-PWUPT:
525
40000
40000
30000
20000
P2-PWUPT
left-side wall
10000
30000
20000
P2-PWUPT
right-side wall
10000
0
0
0
8
10
time, t (sec.)
12
14
16
(a)
8
10
time, t (sec.)
12
14
16
(b)
Fig. 15.41 Wall pier base axial force time histories for Structure P2-PWUPT:
Fig. 15.42 plots the minimum ( markers) and maximum ( markers) compressive
wall pier base axial force demands from Tables 15.46-15.49 against the maximum
incremental velocity (MIV) of the ground motion records (see Tables 14.2 and 14.1),
where the dashed and solid horizontal lines represent the estimated idealized wall pier
base axial force and the mean wall pier base axial force demand, respectively, for each
data set. It is observed that there is very little scatter or variation in the wall pier base
axial force demands in each data set, and that the wall pier base axial force demands
under the survival-level and design-level ground motions are similar.
15.3.6 Wall Base Shear Forces
The peak total coupled wall base shear force demands from the dynamic analyses
and 15.51, respectively. The coupled wall base shear force, F is equal to the sum of the
floor and roof level wall pier inertia forces, which are calculated as the lumped mass
times the total absolute acceleration (i.e., relative acceleration with respect to the ground
plus the ground acceleration) at each floor/roof level.
Comparisons of the peak coupled wall base shear force demands in Tables 15.50
and 15.51 with the coupled wall base shear forces reached during the static push-over
analysis of the structure in Chapter 13 show that the forces from the dynamic analyses are
significantly larger than the forces from the static analysis. As discussed in Chapter 11,
the differences between the static and dynamic analysis forces are because of the
contribution of higher modes of vibration to the dynamic response of the structure.
526
maximum and minimum wall pier base compressive axial forces (kN)
maximum and minimum wall pier base compressive axial forces (kN)
50000
P2-PWUPT
40000
mean=35050 kN
30000
maximum compressive axial force

minimum compressive axial force
20000

mean=12359 kN
10000
50
100
150
200
250
300
350
400
(a)
50000
P2-PWUPT
40000
mean=31198 kN
30000
maximum compressive axial force
minimum compressive axial force
20000
10000
00
estimate (idealized),
Ncwy=31132 kN
estimate (idealized), Ntwy=16039 kN

mean=13436 kN
50
100
150
200
(b)
Fig. 15.42 Peak wall pier base axial force demands for Structure P2-PWUPT:
527
TABLE 15.50
P2-PWUPT UNDER SURVIVAL LEVEL GROUND MOTIONS
ground
motion
peak coupled wall
ground
motion
peak coupled wall
LA21 LA22 LA23c LA24 LA25 LA26a LA27 LA28b LA29 LA30
estimate
(design)
standard
deviation
13579 9632 8558 11283 11955 12477 10190 16350 12892 16139 13099
2315
standard
deviation/mean
14246 15196 10869 10446 15416 13262 11346 11909 8990 11037 12289
0.19
c
b
TABLE 15.51
P2-PWUPT UNDER DESIGN LEVEL GROUND MOTIONS
ground
LA01 LA02 LA03 LA04
motion
peak coupled wall
10043 10355 9280 7348
ground
motion
peak coupled wall
9679 13892 10859 10507
LA05c LA06 LA07 LA08 LA09 LA10

6264
6103 7351 7689 8953
7867
standard
deviation
2625
LA15 LA16 LA17 LA18 LA19 LA20b mean
standard
deviation/mean
9031 10114 8599 11683 14137 16152 9795
0.27
c
b
The mean peak coupled wall base shear force demand from the twenty survivallevel ground motions is 12289 kN, which is close to and smaller than the estimated
maximum coupled wall base shear force demand of Qw,max=13099 kN from Chapter 12.
Thus, the method used in Chapter 12 to estimate the coupled wall base shear force
demand of the prototype structures including the effect of higher modes appears to be
valid. Note that the design approach does not require the estimation of the maximum
coupled wall base shear force demand under the design-level earthquake; and thus, no
design-level estimate is provided in Table 15.51.
As an example, Fig. 15.43 shows the coupled wall base shear force time history of
coupled wall base shear force demand is closest to the mean demand from the twenty
records (see Table 15.50). The dashed horizontal lines indicate the design estimate for the
maximum coupled wall base shear force demand, Qw,max from Chapter 12.
528
15000
10000
P2-PWUPT
5000
0
-5000
-10000
-15000
0
10
15
time, t (sec.)
Fig. 15.43 Coupled wall base shear force time history for Structure P2-PWUPT
Fig. 15.44 plots the peak coupled wall base shear force demands from Tables 15.50
motion records (see Tables 14.2 and 14.1), where the dashed and solid horizontal
lines represent the design estimate (from Chapter 12) and the mean peak coupled wall
base shear force demand, respectively, for each data set. Significant scatter is observed in
the data indicating that, while the survival-level design estimate matches the mean
demand from the dynamic analyses quite well, the coupled wall base shear force demand
may be considerably larger or smaller than the estimated demand, depending on the
earthquake. It is also interesting to note that the design-level LA20 ground motion record
results in a peak coupled wall base shear force demand that is similar to the largest
demand from the survival-level records. Some correlation can be observed between the
peak coupled wall base shear force demands in Fig. 15.44 and the peak acceleration
The peak wall pier base shear force demands from the dynamic analyses of
and 15.53, respectively. The wall pier base shear forces are calculated from the
equilibrium of the floor/roof level wall pier inertia forces with the coupling beam axial
forces and beam post-tensioning tendon forces from the analysis model (obtained from
the fiber beam-column elements and the truss elements representing the coupling beams
and the beam post-tensioning tendons, respectively).
529
20000
16000
12000
mean=12289 kN
8000
4000
P2-PWUPT
50
100
150
200
250
300
350
(a)
20000
400
P2-PWUPT
16000
LA20
12000
mean=9795 kN
8000
4000
50
100
150
200
(b)
Fig. 15.44 Peak coupled wall base shear force demands for Structure P2-PWUPT:
530
TABLE 15.52
ground
motion
peak wall pier base
shear force (kN)
ground
motion
peak wall pier base
shear force (kN)
LA21 LA22 LA23c LA24 LA25 LA26 LA27 LA28b LA29 LA30
estimate
(design)
standard
deviation
10379 7560 6824 8636 9361 9424 7812 11849 8781 10841
9955
1497
mean
standard
deviation/mean
10182 11646 8353 8016 11487 10153 9079 9287 6904
9264
0.16
8697
c
b
TABLE 15.53
ground
motion
peak wall pier base
shear force (kN)
ground
motion
peak wall pier base
shear force (kN)
standard
deviation
LA01 LA02 LA03 LA04 LA05 LA06c LA07 LA08 LA09a LA10
7714 7944 6663 5178 4619 3802 5599 5723 7095
5878
1646
LA11 LA12 LA13 LA14 LA15 LA16 LA17 LA18 LA19 LA20b
mean
standard
deviation/mean
6104 9323 8548 7189 7163 7967 6791 8729 8263 10415
7036
0.23
c
b
The mean peak wall pier base shear force demand from the twenty survival-level
analyses is 9264 kN, which is close to and smaller than the estimated maximum wall pier
base shear force demand of Qw,dt=9955 kN from Chapter 12. Thus, the method used in
Chapter 12 to estimate the maximum wall pier base shear force demand appears to be
valid. Note that the design approach does not require the estimation of the maximum wall
pier base shear force demand under the design-level earthquake; and thus, no design-level
estimate is provided in Table 15.53.
As an example, Fig. 15.45 shows the shear force time histories at the bases of the
left-side and right-side wall piers in Structure P2-PWUPT under the survival-level LA38
ground motion, for which the peak wall pier base shear force demand (occurring in the
left-side wall pier) is closest to the mean demand from the twenty records (see Table
15.52). The dashed horizontal lines indicate the design estimate for the maximum wall
pier base shear force demand, Qw,dt from Chapter 12. It is observed that there are residual
base shear forces in the left-side and right-side wall piers at the end of the ground motion,
even though the coupled wall base shear force is close to zero.
Fig. 15.46 plots the peak wall pier base shear force demands from Tables 15.52 and
531
represent the design estimate (from Chapter 12) and the mean peak wall pier base shear
force demand, respectively, for each data set. Similar to the peak coupled wall base shear
force demands in Fig. 15.44, significant scatter is observed in the peak wall pier base
shear force demands and the demand for the design-level LA20 ground motion record is
similar to the largest demand from the survival-level records.
If the peak wall pier base shear force demands in Tables 15.52 and 15.53 are
divided by the corresponding peak coupled wall base shear force demands in Tables
15.50 and 15.51, the wall pier base shear force distribution ratios are obtained. The mean
wall pier base shear force distribution ratio from the twenty survival-level analyses is
0.75, which is close to the design ratio of Qw,dt/Qw,max=0.76 from Chapter 12.
Note that the peak wall pier base shear force demands in Tables 15.52 and 15.53
and the corresponding peak coupled wall base shear force demands in Tables 15.50 and
15.51 do not necessarily occur at the same time during each dynamic analysis. Thus,
Tables 15.54 and 15.55 show the peak ratio of the wall pier base shear force to the
corresponding coupled wall base shear force occurring at the same time during each
analysis. The mean value of the peak wall pier base shear force distribution ratio from the
twenty survival-level analyses in Table 15.54 is 0.76, which is the same as the value used
in design and close to the value calculated from the peak wall pier and peak coupled wall
base shear force demands above.
10000
10000
P2-PWUPT
left-side wall
5000
-5000
5
10
15
20
time, t (sec.)
25
5000
-5000
-10000
-10000
0
P2-PWUPT
right-side wall
LA38 (survival)
30
(a)
10
15
20
time, t (sec.)
25
(b)
Fig. 15.45 Wall pier base shear force time histories for Structure P2-PWUPT:
532
30
20000
16000
12000
8000
mean=9264 kN
P2-PWUPT
4000
0
0
50
100
150
200
250
300
350
400
(a)
20000
16000
12000
LA20
8000
4000
P2-PWUPT
mean=7036 kN
50
100
150
200
(b)
Fig. 15.46 Peak wall pier base shear force demands for Structure P2-PWUPT:
533
TABLE 15.54
ground
motion
distribution ratio
ground
motion
distribution ratio
0.76
0.78
0.82
0.79
0.78
0.75
0.79
0.72
0.69
0.67
estimate
(design)
standard
deviation
0.76
0.037
0.74
0.77
0.77
0.78
0.76
0.77
0.80
0.78
0.78
0.81
0.76
standard
deviation/mean
0.049
TABLE 15.55
ground
LA01 LA02 LA03 LA04
motion
0.77 0.77 0.72 0.70
distribution ratio
ground
LA11 LA12 LA13 LA14
motion
0.76 0.69 0.79 0.68
distribution ratio
standard
deviation

0.74
0.62
0.77
0.77
0.79
0.76
0.061
LA15 LA16 LA17 LA18 LA19 LA20 mean

0.79
0.79
0.79
0.75
0.58
0.64
0.73
standard
deviation/mean
0.083
15.3.7 Wall Base Diagonal Tension and Shear Slip Failure

It is shown in the previous section that the mean peak wall pier base shear force
demand under the twenty survival-level ground motion records in Table 15.52 is smaller
than the estimated design maximum wall pier base shear force demand from Chapter 12.
Thus, it is concluded that diagonal tension failure of the wall piers in Structure P2PWUPT can be, on average, prevented as long as the design diagonal tension capacity
Fw,dt=9261 kN from Chapter 12 is achieved and maintained.
Similarly, the mean peak coupled wall base shear force demand under the twenty
survival-level ground motion records in Table 15.50 is smaller than the estimated design
maximum coupled wall base shear force demand from Chapter 12. However, it cannot be
simply concluded that shear slip failure of the structure is prevented since the shear slip
capacity at the base of the structure, Fw,ss varies during a ground motion. As described in
Chapter 12, the coupled wall base shear slip capacity is computed as Fw,ss= Fg,ss+Fp,ss
where, Fg,ss is the slip capacity due to the applied gravity loads and Fp,ss is the shear
friction capacity due to due to the wall pier post-tensioning forces. The coupled wall base
shear slip capacity varies with time during an earthquake since Fp,ss varies. The shear slip
capacity from the gravity loads, Fg,ss is assumed to remain constant during an earthquake.
The minimum coupled wall base shear slip capacity to base shear force demand
ratios from the dynamic analyses of Structure P2-PWUPT under the twenty survival-level
are listed in Tables 15.56 and 15.57, respectively. The coupled wall base shear slip
capacity is calculated using an assumed coefficient of friction between the wall piers and
534
the foundation of wf=0.7 (see Chapter 12). The design estimate for the survival-level
minimum coupled wall base shear slip capacity to base shear force demand ratio in Table
15.56 is determined as the ratio between the design estimates for the coupled wall base
shear slip capacity, Fw,ss and base shear force demand, Qw,max, from Chapter 12. It is
observed that the mean ratio from the survival-level dynamic analysis results is larger
than and reasonably close to the design estimate, and thus, the design procedure to
prevent shear slip failure at the base of the structure seems to be valid.
TABLE 15.56
MINIMUM COUPLED WALL BASE SHEAR SLIP CAPACITY-DEMAND RATIOS
FOR STRUCTURE P2-PWUPT UNDER SURVIVAL LEVEL GROUND MOTIONS
ground
motion
min. wall shear slip
ground
motion
LA21 LA22 LA23b LA24 LA25 LA26 LA27 LA28c LA29 LA30
2.38
3.33
3.70
3.00
2.82
2.59
3.02
1.96
2.43
estimate
(design)
standard
deviation
2.10
0.51
mean
standard
deviation/mean
2.70
0.19
1.97
2.23
2.28
2.87
2.94
2.21
1.99
3.04
2.75
3.47
3.07
The peak ratio under this ground motion is closest to the mean ratio.
The peak ratio under this ground motion is the largest ratio.
c
The peak ratio under this ground motion is the smallest ratio.
b
TABLE 15.57
MINIMUM COUPLED WALL BASE SHEAR SLIP CAPACITY-DEMAND RATIOS
FOR STRUCTURE P2-PWUPT UNDER DESIGN LEVEL GROUND MOTIONS
ground
LA01 LA02 LA03a LA04
motion
3.15 3.01 3.32 4.13
ground
LA11 LA12 LA13 LA14
motion
3.15 2.25 2.96 2.93
standard
deviation

4.83
4.94
4.17
4.02
3.56
3.86
0.81
LA15 LA16 LA17 LA18 LA19c LA20 mean

3.49
3.27
3.70
2.69
2.17
1.92
3.38
standard
deviation/mean
0.24
c
b
The minimum coupled wall base shear slip capacity to base shear force demand
ratios under all of the survival-level and design-level ground motion records in Tables
15.56 and 15.57 are larger than 1.0. Thus, shear slip failure at the base of Structure P2PWUPT is prevented. As an example, Fig. 15.47 shows comparisons between the
coupled wall base shear slip capacity (solid line) and base shear force demand (dashed
line) time histories for the structure under the survival-level LA38 ground motion, for
which the minimum coupled wall base shear slip capacity-demand ratio is closest to the
mean ratio from the twenty records (see Table 15.56). For comparison, the solid and
dashed horizontal lines indicate the design estimates for the coupled wall base shear slip
capacity, Fw,ss and base shear force demand, Qw,max, respectively. As described in Chapter
12, the design estimate for the coupled wall base shear slip capacity due to the wall pier
post-tensioning forces, Fp,ss is computed from estimated residual stresses in the wall pier
535
coupled wall base shear slip capacity and demand (kN)
post-tensioning bars upon unloading the structure after displacing it laterally up to the
survival-level roof drift demand s in the positive and negative directions.
40000
capacity
demand
wall shear slip capacity
30000
estimate (design), Fw, ss=27477 kN
20000
P2-PWUPT
LA38
estimate (design), Qw, max=13099 kN
10000
wall shear slip demand
0
10
15
20
25
30
time, t (sec.)
Fig. 15.47 Coupled wall base shear slip capacity and base shear force demand
time histories for Structure P2-PWUPT
Fig. 15.48 plots the minimum coupled wall base shear slip capacity-demand ratios
respectively, for each data set. It is observed that there is significant scatter or variation in
the minimum coupled wall base shear slip capacity-demand ratio; however, as stated
previously, shear slip at the base of the structure is not expected to under any of the
ground motion records.
536
minimum coupled wall base shear slip capacitydemand ratio

minimum coupled wall base shear slip capacitydemand ratio
P2-PWUPT
mean=2.70
0
0
50
100
150
200
250
300
350
400
(a)
6
P2-PWUPT
4
mean=3.38
3
0
0
50
100
150
200
(b)
Fig. 15.48 Minimum coupled wall base shear slip capacity to base shear force demand
ratio for Structure P2-PWUPT: (a) survival level; (b) design level
537
15.3.8 Wall Base Moments

The peak wall pier base moment demands from the dynamic analyses of Structure
P2-PWUPT under the twenty survival-level SAC ground motion records (LA21-LA40)
respectively. The wall pier base moments are calculated from the equilibrium of the
floor/roof level wall pier inertia forces with the coupling beam midspan axial forces and
beam post-tensioning tendon forces from the analysis model (obtained from the fiber
beam-column elements and the truss elements representing the coupling beams and the
beam post-tensioning tendons, respectively).
TABLE 15.58
PEAK WALL PIER BASE MOMENT DEMANDS FOR STRUCTURE P2-PWUPT
ground
motion
peak wall pier base
moment (kN.m)
ground
motion
peak wall pier base
moment (kN.m)
estimate
(idealized)
standard
deviation
68533 64929 59484 66816 66692 70284 64455 68737 59924 62613
61800
4242
LA31 LA32 LA33a LA34 LA35 LA36b LA37 LA38 LA39c LA40
mean
standard
deviation/mean
67991 68759 66511 62568 72826 72973 69494 69708 57540 66002
66342
0.064
c
b
TABLE 15.59
PEAK WALL PIER BASE MOMENT DEMANDS FOR STRUCTURE P2-PWUPT
ground
motion
peak wall pier base
moment (kN.m)
ground
motion
peak wall pier base
moment (kN.m)
estimate
(idealized)
standard
deviation
59212 60071 59009 46988 47191 39803 54750 54592 62579 55371
58235
6351
LA11 LA12 LA13 LA14 LA15 LA16b LA17 LA18 LA19a LA20
mean
standard
deviation/mean
54885 53270 63054 61461 60139 64771 60365 59845 56501 63833
56884
0.11
c
b
Since the proposed design approach does not require the estimation of the survivallevel and design-level wall pier base moments, the results from the dynamic analyses are
compared with the results from the static idealized tri-linear pushover analysis of the
structure in Chapter 13. The bending moments at the base of the compression-side wall
pier (which has a larger base moment than the tension-side wall pier), Mcwu and Mcwy,
corresponding to the estimated coupled wall ultimate state and coupled wall PT-yielding
state are given in Tables 15.58 and 15.59, respectively (see also Table 13.2). The results
show that the wall pier base moments, Mcwu and Mcwy, from the idealized pushover
538
analysis provide good estimates of the peak wall pier base moment demands from the
dynamic analyses.
As an example, Fig. 15.49 shows the bending moment time histories at the bases of
the left-side and right-side wall piers in Structure P2-PWUPT under the survival-level
LA33 ground motion, for which the peak wall pier base moment demand (occurring in
the left-side wall pier) is closest to the mean demand from the twenty records (see Table
15.54). The dashed horizontal lines indicate the estimated idealized compression-side
wall pier base moment, Mcwu from the tri-linear static pushover analysis in Chapter 13 at
the coupled wall ultimate state.
Fig. 15.50 plots the peak wall pier base moment demands from Tables 15.54 and
represent the estimated idealized wall pier base moment (from Chapter 13) and the mean
peak wall pier base moment demand, respectively, for each data set. A considerable
amount of scatter is observed in the peak wall pier base moment demands in each data set.
The peak total coupled wall base moment demands from the dynamic analyses of
and 15.61, respectively. The coupled wall base moment, Mw is calculated from the
equilibrium of the floor and roof level wall pier inertia forces.
80000
80000
40000 P2-PWUPT
left-side wall
0
-40000
-80000
0
10
time (sec)
15
20
(a)
40000

P2-PWUPT
right-side wall
-40000
-80000
0
10
time (sec)
15
(b)
Fig. 15.49 Wall pier base moment time histories for Structure P2-PWUPT:
539
20
80000
mean = 66342 kN.m
60000
40000
P2-PWUPT
20000
50
100
150
200
250
300
350
400
(a)
80000
estimate (design), Mcwy=58235 kN.m
60000
mean = 56884 kN.m
40000
20000
P2-PWUPT
50
100
150
200
(b)
Fig. 15.50 Peak wall pier base moment demands for Structure P2-PWUPT:
540
TABLE 15.60
PEAK COUPLED WALL BASE MOMENT DEMANDS FOR STRUCTURE
ground
motion
peak coupled wall
base moment (kN.m)
ground
motion
peak coupled wall
base moment (kN.m)
LA21b LA22 LA23 LA24 LA25a LA26 LA27 LA28 LA29 LA30
estimate
(idealized)
standard
deviation
249711 193240 172936 201761 200660 221160 188434 211976 169190 168147
154844
23852
LA31 LA32 LA33 LA34 LA35 LA36 LA37 LA38 LA39c LA40
mean
standard
deviation/mean
213558 217696 187010 172921 224774 233441 203306 206712 159868 191164
199380
0.12
c
b
TABLE 15.61
PEAK COUPLED WALL BASE MOMENT DEMANDS FOR STRUCTURE
ground
motion
peak coupled wall
base moment (kN.m)
ground
motion
peak coupled wall
base moment (kN.m)
LA01a LA02 LA03 LA04 LA05 LA06c LA07 LA08 LA09 LA10
estimate
(idealized)
standard
deviation
162830 162900 172330 126050 126930 112440 149330 155540 177680 153920
144019
20633
mean
standard
deviation/mean
155490 141440 181900 181390 169780 192600 171880 163410 155460 177350
159530
0.13
c
b
Similar to the wall pier base moments, the peak coupled wall base moment
demands from the dynamic analyses are compared with the results from the static
idealized tri-linear pushover analysis of the structure in Chapter 13. The coupled wall
base moments, Mwu and Mwy, corresponding to the estimated coupled wall ultimate state
and PT-yielding state are given in Tables 15.60 and 15.61, respectively (see also Table
13.2). The results show that the coupled wall base moments, Mwu and Mwy, from the
idealized pushover analysis underestimate the peak coupled wall base moment demands
from the dynamic analyses even though the wall pier base moment demands are
estimated reasonably well. This is possibly because of the underestimated coupling beam
shear forces as described later in more detail.
As an example, Fig. 15.51 shows the coupled wall base moment time history of
coupled wall base moment demand is closest to the mean demand from the twenty
records (see Table 15.60). The dashed horizontal lines indicate the estimated idealized
coupled wall base moment, Mwu from the tri-linear static pushover analysis in Chapter 13
at the coupled wall ultimate state.
Fig. 15.52 plots the peak coupled wall base moment demands from Tables 15.60
541
represent the estimated idealized coupled wall base moment (from Chapter 13) and the
mean peak coupled wall base moment demand, respectively, for each data set. Similar to
the peak wall pier base moment demands, a considerable amount of scatter is observed in
the peak coupled wall base moment demands in each data set.
250000
200000
P2-PWUPT
coupled wall base moment, Mw (kN.m)
150000
100000
50000
0
-50000
-100000
-150000
-200000
-250000
0
10
15
time, t (sec.)
Fig. 15.51 Coupled wall base moment time history for Structure P2-PWUPT
542
300000
P2-PWUPT
mean=199380 kN.m
200000

100000
50
100
150
200
250
300
350
400
(a)
300000
P2-PWUPT
200000
mean=159530 kN.m
100000
50
100
150
200
(b)
Fig. 15.52 Peak coupled wall base moment demands for Structure P2-PWUPT:
543
15.3.9 Wall Base Strains

The peak confined concrete compression strain demands at the bases of the wall
piers from the dynamic analyses of Structure P2-PWUPT under the twenty survival-level
are listed in Tables 15.62 and 15.63, respectively. The concrete compression strains are
determined directly from the fiber beam-column elements used for the wall piers in the
analysis model.
TABLE 15.62
PEAK WALL PIER BASE CONFINED CONCRETE COMPRESSION STRAIN
DEMANDS FOR STRUCTURE P2-PWUPT UNDER SURVIVAL LEVEL GROUND
MOTIONS
ground
motion
peak confined
ground
motion
peak confined
-7.5
-2.1
-0.82 -3.1
-2.5
-6.1
-1.9
-5.0
-0.86
-1.7
estimate
(design)
standard
deviation
-3.2
0.029
LA31 LA32a LA33 LA34 LA35 LA36b LA37 LA38 LA39c LA40 mean
-5.5
-4.9
-2.9
-1.3 -10.0 -10.7
-5.2
-5.6
-0.65
-2.8
-4.0
standard
deviation/mean
0.73
c
b
TABLE 15.63
PEAK WALL PIER BASE CONFINED CONCRETE COMPRESSION STRAIN
DEMANDS FOR STRUCTURE P2-PWUPT UNDER DESIGN LEVEL GROUND
MOTIONS
ground
LA01 LA02 LA03 LA04
motion
peak confined
-8.1 -10.0 -7.5 -2.5
ground
LA11 LA12 LA13 LA14
motion
peak confined
-5.7 -5.9 -12.6 -12.3
standard
deviation

-2.6
-1.7
-5.1
-4.9 -12.2 -5.5
0.0046
LA15 LA16b LA17a LA18 LA19 LA20 mean

-8.3
-20.0
-8.2
-8.9
-6.0 -15.6
-8.2
standard
deviation/mean
0.56
c
b
The mean peak wall pier base confined concrete compression strain demand from
the survival-level dynamic analyses in Table 15.62 is -0.040, which is considerably larger
than the estimated wall pier base confined concrete strain demand of -0.032 from Chapter
12. Thus, the method used in Chapter 12 to estimate the wall pier base confined concrete
compression strain demand needs to be improved. Note that the design approach does not
require the estimation of the confined concrete compression strain demand under the
design-level earthquake; and thus, no design-level estimate is provided in Table 15.63.
544


As an example, Fig. 15.53 shows the extreme confined concrete compression strain
time histories at the base of Structure P2-PWUPT under the survival-level LA32 ground
motion, for which the peak wall pier base confined concrete compression strain demand
(occurring at the left end of the left-side wall pier) is closest to the mean demand from the
twenty records (see Table 15.62). A total of four strain time histories are shown for
the left and right ends of the left-side and right-side wall piers, where the dashed
horizontal lines indicate the design estimate for the wall pier base confined concrete
compression strain demand from Chapter 12.

P2-PWUPT
estimate (design)=-0.032
10
time (sec)
15
20

P2-PWUPT
-0.05
0
10
time (sec)
(a)


P2-PWUPT
LA32
-0.05
0
-0.05
0
15
(b)
15
20
15
20
-0.05
0
20
10
time (sec)

P2-PWUPT
10
time (sec)
Fig. 15.53 Wall pier base extreme confined concrete compression strain time histories
for Structure P2-PWUPT: (a) left-side wall; (b) right-side wall
Fig. 15.54 plots the peak wall pier base confined concrete compression strain
demands from Tables 15.62 and 15.63 ( markers) against the maximum incremental
dashed and solid horizontal lines represent the design estimate and mean peak wall pier
base confined concrete compression strain demand, respectively, for each data set. The
trends for the peak wall pier base confined concrete compression strain demands in Fig.
15.54 are similar to the trends for the peak coupled wall roof drift demands in Fig. 15.33.
545
-0.02
-0.04
mean=-0.0404
-0.06
-0.08
-0.1
-0.12
0
P2-PWUPT
50
100
150
200
250
300
350
maximum incremental velocity, MIV (cm/sec)
400
(a)
0
-0.02
mean=-0.0082
-0.04
-0.06
-0.08
-0.1
-0.12
0
P2-PWUPT
50
100
150
maximum incremental velocity, MIV (cm/sec)
200
(b)
Fig. 15.54 Peak wall pier base confined concrete compression strain demands
546
15.3.10 Coupling Beam Chord Rotations

The peak coupling beam chord rotation demands from the dynamic analyses of
and 15.65, respectively. As described in Chapter 7, the coupling beam chord rotation, b
is measured from a tangent drawn at the left end of each beam.
TABLE 15.64
PEAK COUPLING BEAM CHORD ROTATION DEMANDS FOR STRUCTURE P2PWUPT UNDER SURVIVAL LEVEL GROUND MOTIONS
ground
motion
peak coupling beam
chord rotation (%)
ground
motion
peak coupling beam
chord rotation (%)
4.70
2.63
1.24
3.18
2.98
4.31
2.28
3.71
1.42
1.69
3.97
3.94
2.56
1.79
5.77
6.09
3.58
4.03
1.16
2.68
estimate
(design)
standard
deviation
5.62
1.41
mean
standard
deviation/mean
3.19
0.44
c
b
TABLE 15.65
PEAK COUPLING BEAM CHORD ROTATION DEMANDS FOR STRUCTURE
ground
motion
peak coupling beam
1.13 1.24 0.70 0.43
chord rotation (%)
ground
LA11 LA12 LA13 LA14
motion
peak coupling beam
0.61 0.62 1.84 1.97
chord rotation (%)

0.49
0.27
0.79
0.93 1.36
0.94
estimate
(design)
standard
deviation
2.59
0.58
LA15 LA16b LA17 LA18 LA19 LA20 mean

1.38
2.44
1.46
1.33 0.85
1.83
1.13
standard
deviation/mean
0.51
c
b
The mean peak coupling beam chord rotation demand from the survival-level
dynamic analyses in Table 15.64 is 3.19%, which is significantly smaller than the
estimated survival-level coupling beam chord rotation demand of s=5.62% from Chapter
12. As an example, Fig. 15.55 shows the chord rotation time histories for the eight
coupling beams in Structure P2-PWUPT under the survival-level LA24 ground motion,
for which the peak coupling beam chord rotation demand (occurring in the roof beam) is
closest to the mean demand from the twenty records (see Table 15.64). The design
estimate for the peak coupling beam chord rotation demand, s from Chapter 12 is
indicated in the figure.
547
4
P2-PWUPT
2nd floor
coupling beam chord

coupling beam chord

P2-PWUPT
3rd floor
4
0
10
time, t (sec.)
15
20
20
15
20
15
20
15
20
2
4
10
time, t (sec.)
15
20
10
time, t (sec.)
4
P2-PWUPT
6th floor
coupling beam chord

coupling beam chord

15
P2-PWUPT
5th floor

4
P2-PWUPT
7th floor
4
0
10
time, t (sec.)
15
20
10
time, t (sec.)
4
P2-PWUPT
8th floor
coupling beam chord

coupling beam chord

10
time, t (sec.)
4
P2-PWUPT
4th floor
coupling beam chord

coupling beam chord

2
4
0
10
time, t (sec.)
P2-PWUPT
roof
2
4
15
20
10
time, t (sec.)
Fig. 15.55 Coupling beam chord rotation time histories for Structure P2-PWUPT
Similarly, the mean peak coupling beam chord rotation demand from the designlevel analyses in Table 15.65 is 1.13%, which is smaller than the estimated design-level
coupling beam chord rotation demand of d=2.59% from Chapter 12. It is concluded that
the displacement-based design objectives under the survival-level and design-level
ground motions are satisfied; however, the procedure to estimate the coupling beam
548
chord rotation demands may need to be improved to prevent an overly-conservative

design.
Fig. 15.56 plots the peak coupling beam chord rotation demands from Tables 15.64
represent the design estimate (from Chapter 12) and the mean peak coupling beam chord
rotation demand, respectively, for each data set. It is observed that the coupling beam
chord rotation demands tend to follow similar trends to the coupled wall roof drift
The peak axial force demands at the midspan of the coupling beams from the
dynamic analyses of Structure P2-PWUPT under the twenty survival-level SAC ground
Tables 15.66 and 15.67, respectively. The coupling beam midspan axial forces, Nb are
determined directly from the fiber beam-column elements at the beam midspan locations
in the analysis model.
As discussed in Chapter 13, the largest coupling beam axial forces develop in the
lowest beam at the second floor level as a result of the fixed boundary conditions
assumed at the wall pier bases. The mean peak coupling beam midspan axial force
demand from the survival-level dynamic analyses in Table 15.66 is 3784 kN, which is
much larger than the design force of, approximately, 1.25Pby=2365 kN from Chapter 12.
Thus, the axial force used in Chapter 12 for the design of the coupling beams appears
unconservative and need to be improved. Note that the design approach does not require
the estimation of the coupling beam axial force under the design-level earthquake; and
thus, no design-level estimate is provided in Table 15.67.
As an example, Fig. 15.57 shows the midspan axial force time histories for the
eight coupling beams in Structure P2-PWUPT under the survival-level LA24 ground
motion, for which the peak coupling beam midspan axial force demand (always occurring
in the second floor beam) is closest to the mean demand from the twenty records (see
Table 15.66). The dashed horizontal lines indicate the design estimate for the peak
coupling beam midspan axial force, which is equal to 1.25Pby for the second and third
floor beams and Pby for the upper level beams. It is observed that the peak midspan axial
force demands in the upper level beams also exceed the design estimate. Looking at the
coupling beam midspan axial force time histories in Fig. 15.57 and since the coupling
beam axial forces from the static analyses in Chapter 13 are estimated reasonably well by
the proposed approximate procedures, the large peak axial force demands from the
dynamic analyses may be due to the effect of higher modes during the response of the
structure; however, this was not investigated. Note that there is a considerable loss in the
second floor beam axial force as discussed in Chapter 13, resulting in a smaller axial
force in this beam at the end of the earthquake as compared with the upper level beams.
549
7
6
estimate (design), s = 5.62%
P2-PWUPT
4
mean=3.19%
3
2
1
0
50
100
150
200
250
300
maximum incremental velocity (cm/sec)
350
400
(a)
7
6
P2-PWUPT
5
4
3
estimate (design), d = 2.59%
2
mean=1.13%
1
0
50
100
150
200
maximum incremental velocity (cm/sec)
(b)
Fig. 15.56 Peak coupling beam chord rotation demands for Structure P2-PWUPT:
550
TABLE 15.66
ground
motion
peak coupling beam
ground
motion
peak coupling beam
estimate
(design)
standard
deviation
4519 3708 3082 3719 4047 3718 3408 4016 3513 3891
2365
497
LA31b LA32 LA33 LA34 LA35 LA36 LA37 LA38 LA39c LA40 mean
4750 4351 3486 3346 4330 4355 3619 3503 2809 3507
standard
deviation/mean
3784
0.13
c
b
TABLE 15.67
ground
motion
peak coupling beam
ground
motion
peak coupling beam
3199 3404 2825 2160 1972 1835
2585 2444 3212
2611
LA11 LA12 LA13 LA14 LA15 LA16 LA17a LA18 LA19 LA20b mean
2628 3403 3439 3375 3439 3389
2993 3450 3071
4069 2975
standard
deviation
578
standard
deviation/mean
0.19
c
b
Fig. 15.58 plots the peak coupling beam midspan axial force demands from Tables
lines represent the design estimate (from Chapter 12) and the mean peak coupling beam
midspan axial force demand, respectively, for each data set. A significant amount of
scatter is observed in the dynamic analysis results.
551
estimate (design),
1.25Pby =2365 kN
2000
1000
0

10
time, t (sec.)
15
5000
P2-PWUPT
4th floor
4000
3000
estimate (design)
Pby =1892 kN
2000
1000
0
0
5000
10
time, t (sec.)
15
20
P2-PWUPT
6th floor
4000
3000
estimate (design)
Pby =1892 kN
2000
1000
0
5000
10
time, t (sec.)
15
20
P2-PWUPT
8th floor
4000
3000
estimate (design)
Pby =1892 kN
2000
1000
0
10
time, t (sec.)
15
5000
20
P2-PWUPT
3rd floor
4000
estimate (design),
1.25Pby =2365 kN
3000
2000
1000
0
0
20
3000
4000
P2-PWUPT
2nd floor
5000
10
time, t (sec.)
15
20
10
time, t (sec.)
15
20
10
time, t (sec.)
15
20
10
time, t (sec.)
15
20
5000
P2-PWUPT
5th floor
4000
3000
estimate (design)
Pby =1892 kN
2000
1000
0
0
5000
P2-PWUPT
7th floor
4000
3000
estimate (design)
Pby =1892 kN
2000
1000
0
5000
P2-PWUPT
roof
4000
3000
estimate (design)
Pby =1892 kN
2000
1000
0
Fig. 15.57 Coupling beam midspan axial force time histories for Structure P2-PWUPT
552
6000
5000
mean=3784 kN
4000
3000
estimate (design), 1.25Pby =2365 kN
2000
P2-PWUPT
1000
0
0
50
100
150
200
250
300
350
400
(a)
6000
5000
4000
mean=2975 kN
3000
2000
1000
0
0
P2-PWUPT
50
100
150
200
(b)
Fig. 15.58 Peak coupling beam midspan axial force demands for Structure P2-PWUPT:
553

The peak coupling beam post-tensioning force demands from the dynamic analyses
and 15.69, respectively. The beam post-tensioning forces, Pb are calculated as the sum of
the forces in the truss elements representing the beam post-tensioning tendons at each
floor/roof level in the analytical model.
TABLE 15.68
ground
motion
peak coupling beam
ground
motion
peak coupling beam
LA21 LA22 LA23 LA24 LA25a LA26 LA27 LA28 LA29 LA30
1937 1748 1496 1831 1793
estimate
(design)
standard
deviation
1892
169
1922 1678 1898 1528 1586
LA31 LA32 LA33 LA34 LA35 LA36b LA37 LA38 LA39c LA40 mean
1912 1908 1726 1592 1993
2009 1887 1912 1486 1748
standard
deviation/mean
1780
0.095
c
b
TABLE 15.69
ground
motion
peak coupling beam
ground
motion
peak coupling beam
LA01 LA02a LA03 LA04 LA05 LA06c LA07 LA08 LA09 LA10
1493 1516 1494 1367 1376
1356
1432 1456 1591 1458
1465 1486 1621 1643 1538
1728
1552 1528 1456 1620 1509
standard
deviation
97
standard
deviation/mean
0.064
c
b
The mean peak coupling beam post-tensioning force demand from the survivallevel dynamic analyses in Table 15.68 is smaller than the total yield force of the coupling
beam post-tensioning tendons, Pby=1892 kN from Chapter 12; and thus, the yielding of
the beam post-tensioning tendons is prevented as required by design. As an example, Fig.
15.59 shows the post-tensioning force time-histories for the eight coupling beams in
Structure P2-PWUPT under the survival-level ground motion LA25, for which the peak
coupling beam post-tensioning force demand (occurring in the 6th floor beam) is closest
to the mean demand from the twenty records (see Table 15.68). The dashed horizontal
lines indicate the total yield force of the coupling beam post-tensioning tendons, Pby. It is
observed that there is almost no loss in the coupling beam post-tensioning forces during
the ground motion since the tendons do not yield. The smallest post-tensioning force
554
develops in the second floor coupling beam, since the fixed foundation conditions at the
bases of the wall piers restrain the opening of gaps at the beam ends, and thus, limit the
elongation of the post-tensioning tendons during the ground motion.
Fig. 15.60 plots the peak coupling beam post-tensioning force demands from
Tables 15.68 and 15.69 ( markers) against the maximum incremental velocity (MIV) of
the ground motion records (see Tables 14.2 and 14.1), where the dashed and solid
horizontal lines represent the design estimate (from Chapter 12) and the mean peak
coupling beam post-tensioning force demand, respectively, for each data set. It is
observed that yielding of the coupling beam post-tensioning tendons occurs under some
of the survival-level ground motion records, since the design roof drift from Chapter 12 is
exceeded (see Fig. 15.2).
15.3.13 Coupling Beam Shear Forces
The peak coupling beam shear force demands from the dynamic analyses of
and 15.71, respectively. The coupling beam shear forces, Vb are calculated based on the
equilibrium of the wall pier floor/roof level applied gravity loads and the wall pier axial
forces above and below each floor/roof level as determined from the fiber beam-column
elements used for the wall piers in the analytical model.
The peak coupling beam shear force demands in Tables 15.70 and 15.71 always
occur in the second floor beam. The mean peak beam shear force demand from the
survival-level dynamic analyses in Table 15.70 is 1955 kN, which is considerably larger
than the estimated demand of Qb,max=1.25Vb,pty=1184 kN from Chapter 12. Thus, the
force used in Chapter 12 for the shear slip design of the coupling beam-to-wall interfaces
does not appear to be valid for the second floor level. Note that the design approach does
not require the estimation of the coupling beam shear force under the design-level
earthquake; and thus, no design-level estimate is provided in Table 15.71.
As an example, Fig. 15.61 shows the shear force time histories for the eight
coupling beams in Structure P2-PWUPT under the survival-level LA40 ground motion,
for which the peak coupling shear force demand (occurring in the second floor beam) is
horizontal lines indicate the design estimate for the coupling beam shear force, Qb,max
from Chapter 12, which is equal to 1.25Vb,pty for the second and third floor beams and
Vb,pty for the upper level beams. It is observed that the peak shear force demands in the
upper level beams also exceed the design estimate. Since the coupling beam shear forces
from the static analyses in Chapter 13 are estimated reasonably well by the proposed
approximate procedures, the underestimation in the dynamic analysis results may be due
to the effect of higher modes on the peak coupling beam shear force demands; however,
555
1000
P2-PWUPT
2nd floor
500
0
4
8
10
time, t (sec.)
12
14
1000
P2-PWUPT
4th floor
0
2
8
10
time, t (sec.)
12
14
estimate (design), Pby =1892 kN

1500
P2-PWUPT
6th floor
500
0
0
8
10
time, t (sec.)
12
14

1500
1000
P2-PWUPT
8th floor
0
0
8
10
time, t (sec.)
12
14
P2-PWUPT
3rd floor
500
0
4
8
10
time, t (sec.)
12
14
1000
P2-PWUPT
5th floor
500
0
2
8
10
time, t (sec.)
12
14
16
2000

1500
1000
P2-PWUPT
7th floor
500
0
0
8
10
time, t (sec.)
12
14
16
2000

1500
1000
P2-PWUPT
roof
500
0
0
8
10
time, t (sec.)
12
14
Fig. 15.59 Coupling beam post-tensioning force time histories for Structure P2-PWUPT
556
16
1500
16
2000
16
2000
500
1000
16
2000
1000

1500

1500
500
2000
16
2000

1500
2000
16
2500
2000
estimate (design), Pby =1892 (kN)

mean=1780 kN
1500
1000
P2-PWUPT
500
0
0
50
100
150
200
250
300
350
400
(a)
2500
2000
1500
mean=1509 kN
1000
500
0
0
P2-PWUPT
50
100
150
200
(b)
Fig. 15.60 Peak coupling beam post-tensioning force demands for Structure P2-PWUPT:
557
TABLE 15.70
ground
LA21b LA22 LA23 LA24
motion
peak coupling beam
2473 1730 1623 1832
shear force (kN)
ground
LA31 LA32 LA33 LA34
motion
peak coupling beam
2166 2108 1877 1712
shear force (kN)

1983 2082 1655 2073
1690
1708
estimate
(design)
standard
deviation
1184
266
LA35 LA36 LA37 LA38 LA39c LA40a mean

2388 2317 2184 1988
1583
1930
standard
deviation/mean
1955
0.17
c
b
TABLE 15.71
ground
motion
peak coupling beam
1517 1645 1561 1245
shear force (kN)
ground
LA11 LA12 LA13 LA14
motion
peak coupling beam
1436 1525 1668 1628
shear force (kN)
standard
deviation
1263 1183 1396 1392 1695 1418
157
LA15 LA16 LA17 LA18 LA19 LA20b mean

1570 1681 1495 1584 1481 1748
1507
standard
deviation/mean
0.10
c
b
Fig. 15.62 plots the peak coupling beam shear force demands from Tables 15.70
represent the design estimate (from Chapter 12) and the mean peak coupling beam shear
force demand, respectively for each data set. It is observed that there is considerable
scatter in the peak coupling beam shear force demands under the survival-level and
design-level ground motions.
558
1000
0
P2-PWUPT
2nd story
-1000
-2000
10
15
20
time, t (sec.)
25
0
P2-PWUPT
4th story
-1000
-2000
5
10
15
20
time, t (sec.)
25
estimate (design), Qb,max=Vb, pty=947 kN
1000
0
P2-PWUPT
6th story
-1000
-2000
0
10
15
20
time, t (sec.)
25
0
P2-PWUPT
8th story
-1000
-2000
0
10
15
20
time, t (sec.)
25
0
P2-PWUPT
3rd story
-1000
-2000
10
15
20
time, t (sec.)
25
0
P2-PWUPT
5th story
-1000
-2000
5
10
15
20
time, t (sec.)
25
30
2000
1000
0
P2-PWUPT
7th story
-1000
-2000
0
10
15
20
time, t (sec.)
25
30
2000
1000
0
P2-PWUPT
roof story
-1000
-2000
0
10
15
20
time, t (sec.)
25
Fig. 15.61 Coupling beam shear force time histories for Structure P2-PWUPT
559
30
1000
30
2000
30
2000
1000
1000
30
2000
estimate (design), Qb,max =1.25Vb, pty= 1184 kN
0
1000
2000
30
2000
0
estimate (design), Qb,max=1.25Vb, pty=1184 kN
2000
30
2500
2000 mean=1955 kN
1500
estimate (design), Qb,max =1.25Vb, pty=1184 (kN)
1000
P2-PWUPT
500
0
0
50
100
150
200
250
300
350
400
(a)
2500
2000
1500
mean=1507 kN
1000
500
0
0
P2-PWUPT
50
100
150
200
(b)
Fig. 15.62 Peak coupling beam shear force demands for Structure P2-PWUPT:
560
15.3.14 Coupling Beam Shear Slip Failure

The minimum coupling beam shear slip capacity to shear force demand ratio from
the dynamic analyses of Structure P2-CWUPT under the twenty survival-level SAC
listed in Tables 15.72 and 15.73, respectively. The coupling beam shear slip capacity is
computed as Vb,ss=Vp,ss+Va,ss where, Va,ss is the slip capacity from the top and seat angleto-wall connections determined as described in Chapter 12 and Vp,ss is the slip capacity
due to the beam axial force calculated as the axial force at the beam end, Cb times the
assumed coefficient of friction, bw=0.33. Note that the axial force at the beam ends is
larger than the axial force at the beam midspan as a result of the top and seat angle forces.
The beam end axial forces are determined directly from the fiber beam-column elements
at the beam ends in the analysis model. The coupling beam shear slip capacity, Vb,ss
varies with time during an earthquake since Vp,ss due to the beam axial force varies. The
slip capacity from the top and seat angle-to-wall connections, Va,ss is assumed to remain
constant during an earthquake since the integrity of the angle-to-wall connections is
assumed to be maintained.
TABLE 15.72
MINIMUM COUPLING BEAM SHEAR SLIP CAPACITY-DEMAND RATIOS FOR
ground
motion
ground
motion
estimate
(design)
standard
deviation
0.82
1.08
0.103
LA31 LA32 LA33 LA34 LA35 LA36c LA37a LA38 LA39 LA40
mean
standard
deviation/mean
0.74
0.139
LA21 LA22 LA23 LA24 LA25 LA26 LA27b LA28 LA29 LA30
0.61
0.60
0.82
0.69
0.79
0.81
0.79
0.83
0.64
0.59
0.78
0.48
0.85
0.74
0.76
0.77
0.80
0.84
0.78
c
b
TABLE 15.73
MINIMUM COUPLING BEAM SLIP CAPACITY-DEMAND RATIOS FOR
ground
LA01 LA02 LA03a LA04
motion
0.90 0.78 0.87 1.05
ground
LA11 LA12 LA13 LA14
motion
0.91 0.84 0.87 0.86
standard
deviation
LA05 LA06b LA07 LA08 LA09c LA10

1.03
1.09
0.93
0.94
0.74
0.93
0.0921
LA15 LA16 LA17 LA18 LA19 LA20 mean

0.87
0.81
0.91
0.91
0.83
0.75
0.89
standard
deviation/mean
0.104
c
b
The design estimate for the survival-level minimum coupling beam shear slip
capacity to shear force demand ratio in Table 15.72 is determined as the ratio between the
561
design estimates for the coupling beam shear slip capacity, Vb,ss and the beam shear force
demand, Qb,max, from Chapter 12. It is observed that the mean ratio from the survivallevel dynamic analysis results is smaller than the design estimate, and thus, the design
procedure to prevent shear slip failure at the beam ends is not conservative. This is
possibly because of the underestimated coupling beam shear force demands as described
previously.
The mean values of the minimum coupling beam shear slip capacity to shear force
demand ratios in Tables 15.72 and 15.73 are both less than 1.0. Thus, shear slip failure at
the ends of the coupling beams in Structure P2-PWUPT, on average, might occur. As an
example, Fig. 15.63 shows comparisons between the shear slip capacity (solid lines) and
shear force demand (dashed lines) time histories for the eight coupling beams in Structure
P2-PWUPT under the survival-level LA37 ground motion, for which the minimum
coupling beam shear slip capacity-demand ratio is closest to the mean ratio
from the twenty records (see Table 15.72). For comparison, the solid and dashed
horizontal lines indicate the design estimates for the coupling beam shear slip capacity,
Vb,ss and shear force demand, Qb,max=1.25Vb,pty (second and third floor level beams) or
Qb,max=Vb,pty (upper level beams), respectively. As described in Chapter 12, the design
estimate for the coupling beam shear slip capacity due to the beam end axial force, Vp,ss is
computed as the coupling beam initial post-tensioning force, Pbi times the assumed
coefficient of friction at the beam ends, bw=0.33. In addition to the underestimated
maximum shear force demands, the design approach seems to overestimate the minimum
shear slip capacities from the dynamic analysis, also contributing to the unconservative
design to prevent shear slip failure at the beam ends. As a result, it is observed from Fig.
15.63 that shear slip can occur not only at the ends of the lower floor level coupling
beams but also the upper level beams.
Fig. 15.64 plots the minimum coupling beam shear slip capacity-demand ratios
respectively, for each data set. The results clearly point to the need for improvement in
the design approach to prevent shear slip failure at the ends of the coupling beams.
15.3.15 Angle Deformations
The peak tensile deformation demands in the top and seat connection angles of the
coupling beams from the dynamic analyses of Structure P2-PWUPT under the twenty
survival-level SAC ground motion records (LA21-LA40) and twenty design-level records
(LA01-LA20) are listed in Tables 15.74 and 15.75. The angle deformations, ax are
determined directly from the horizontal angle elements in the analytical model as the
deformations parallel to the coupling beams (i.e., horizontal displacement of angle heel
from wall).
562
2000
1500

1000
500
0
0
10
15
20
time, t (sec.)
25
30
2500
capacity
demand
P2-PWUPT
4th floor
LA37
2000
1500
1000
500
0
0
10
15
20
time, t (sec.)
25
30
2500
capacity
demand
P2-PWUPT
6th floor
LA37
2000
1500
1000
500
0
0
10
15
20
time, t (sec.)
25
30
2500
capacity
demand
P2-PWUPT
8th floor
LA37
2000
1500
1000
500
0
0
10
15
20
time, t (sec.)
25
30
beam shear slip capacity and demand (kN) beam shear slip capacity and demand (kN) beam shear slip capacity and demand (kN)
capacity
demand
P2-PWUPT
2nd floor
LA37
beam shear slip capacity and demand (kN) beam shear slip capacity and demand (kN) beam shear slip capacity and demand (kN)
2500
2500
capacity
demand
P2-PWUPT
3rd floor
LA37
2000
1500
1000
500
0
0
10
15
20
time, t (sec.)
25
2500
capacity
demand
P2-PWUPT
5th floor
LA37
2000
1500
1000
500
0
0
10
15
20
time, t (sec.)
25
30
2500
capacity
demand
P2-PWUPT
7th floor
LA37
2000
1500
1000
500
0
0
10
15
20
time, t (sec.)
25
30
2500
capacity
demand
P2-PWUPT
roof
LA37
2000
1500
1000
500
0
0
10
15
20
time, t (sec.)
25
Fig. 15.63 Coupling beam shear slip capacity and shear force demand time histories
563
30
30
minimum couping beam shear slip capacitydemand ratio
2
1.8
1.6
1.4
1.2
1
0.8
mean =0.74
0.6
0.4
P2-PWUPT
0.2
0
0
50
100
150
200
250
300
350
400
minimum couping beam shear slip capacitydemand ratio
(a)
2
1.8
1.6
1.4
1.2
1
0.8
mean =0.89
0.6
0.4
P2-PWUPT
0.2
0
0
50
100
150
200
(b)
Fig. 15.64 Minimum coupling beam shear slip capacity to shear force demand ratios
564
TABLE 15.74
estimate
standard
ground
(design)
deviation
motion
peak angle tensile
46
26
12
31
29
42
22
36
14
17
31
14
deformation (mm)
standard
ground
LA31 LA32 LA33 LA34 LA35 LA36b LA37 LA38 LA39c LA40 mean
deviation/mean
motion
peak angle tensile
39
39
25
18
57
60
35
40
11
26
31
0.45
deformation (mm)
c
b
TABLE 15.75
estimate
standard
ground
LA01 LA02 LA03 LA04 LA05c LA06 LA07 LA08a LA09 LA10
(design)
deviation
motion
peak angle tensile
12
13
7
5
5
3
8
10
15
10
14
6
deformation (mm)
standard
ground
deviation/mean
motion
peak angle tensile
7
7
20
21
15
26
16
14
9
20
12
0.51
deformation (mm)
c
b
The mean peak angle tensile deformation demand from the survival-level dynamic
analyses in Table 15.74 is 31 mm, which is equal to the estimated deformation of uas=31
mm from Chapter 12. However, the estimated angle deformations are based on overestimated coupling beam rotation demands (see Fig. 15.56). Thus, it is concluded that the
method used in Chapter 12 to estimate the angle deformations needs to be improved.
As an example, Fig. 15.65 shows the deformation time histories, ax for the bottom
(i.e., seat) angles at the left ends of the eight coupling beams in Structure P2-PWUPT
under the survival-level LA24 ground motion, for which the peak angle tensile
deformation demand (occurring in the roof beam) is closest to the mean demand from the
twenty records (see Table 15.74). The dashed horizontal lines indicate the design estimate
for the peak angle tensile deformation, uas, from Chapter 12.
Fig. 15.66 plots the peak angle tensile deformation demands from Tables 15.74 and
motion records(see Tables 14.2 and 14.1), where the dashed and solid horizontal lines
represent the design estimate (from Chapter 12) and the mean peak angle tensile
deformation demand, respectively, for each data set. It is observed that the peak angle
tensile deformation demands tend to follow similar trends to the peak coupled wall roof
drift demands in Fig. 15.33 and the peak coupling beam chord rotation demands in Fig.
15.56.
565
40
30
40
P2-PWUPT
2nd floor
left end seat angle
LA24
20
10
0
10
10
time, t (sec.)
15
10
time, t (sec.)
15
20
40
30
10
20
40
P2-PWUPT
4th floor
left end seat angle
LA24
20
10
0
10
30
P2-PWUPT
5th floor
left end seat angle
LA24
20
10
0
10
10
time, t (sec.)
15
20
40
10
time, t (sec.)
15
20
40
30
P2-PWUPT
3rd floor
left end seat angle
LA24
20
10
0
P2-PWUPT
6th floor
left end seat angle
LA24
20
10
0
10
30
P2-PWUPT
7th floor
left end seat angle
LA24
20
10
0
10
10
time, t (sec.)
15
20
40
10
time, t (sec.)
15
20
40
30
30
P2-PWUPT
8th floor
left end seat angle
LA24
20
10
0
10
30
P2-PWUPT
roof
left end seat angle
LA24
20
10
0
10
10
time, t (sec.)
15
20
10
time, t (sec.)
15
Fig. 15.65 Coupling beam connection angle deformation time histories

566
20
70
60
P2-PWUPT
50
40
30
estimate (design), uas=31 mm

mean=31 mm
20
10
0
0
50
100
150
200
250
300
maximum incremental velocity (cm/sec.)
350
400
(a)
70
60
P2-PWUPT
50
40
30
20
estimate (design), uad=14 mm
10
0
0
mean=12 mm
50
100
150
maximum incremental velocity (cm/sec.)
200
(b)
Fig. 15.66 Peak coupling beam connection angle tensile deformation demands
567
15.4 Uncoupled Concrete Wall Systems

This section discusses the dynamic peak roof drift response of uncoupled wall pier
systems (two wall piers with no coupling beams in between), referred to as Structures P4CWUNC and P5-PWUNC corresponding to the prototype coupled wall Structures P1CWUPT and P2-PWUPT, respectively. The analytical models for the uncoupled systems
are created by removing the coupling beams in Structures P1-CWUPT and P2-PWUPT.
The behavior of the uncoupled wall systems under static monotonic lateral loading
can be found in Chapter 13. For dynamic analysis, it is assumed that the building masses
for the uncoupled systems are the same as those for the corresponding coupled wall
structures. As a result of the reduced lateral stiffness, the fundamental periods for
Structures P4-CWUNC and P5-PWUNC are 0.95 sec. and 1.02 sec., respectively, which
are significantly longer than the fundamental periods of the coupled wall systems in
Chapter 12.
The peak roof drift demands for Structures P4-CWUNC and P5-PWUNC under the
twenty survival-level SAC ground motion records (LA21-LA40) are listed in Tables
15.76 and 15.77, respectively.
Figs. 15.67a and 15.67b plot the peak roof drift demands of Structures P4-CWUNC
and P5-PWUNC from Tables 15.76 and 15.77, respectively, against the maximum
incremental velocity (MIV) of the 20 survival level ground motion records (LA21-LA40).
It is observed that the peak roof drift demands of the uncoupled wall pier systems are
significantly larger than the peak roof drift demands of the corresponding coupled wall
systems (see Figs. 15.2a and 15.33a). The ratio between the mean peak roof drift
demands of the coupled and uncoupled wall systems is equal to 0.72 and 0.55 for the
cast-in-place and precast concrete wall systems, respectively.
568
TABLE 15.76
PEAK ROOF DRIFT DEMANDS FOR STRUCTURE P4-CWUNC
ground
LA21 LA22 LA23 LA24 LA25 LA26a
motion
peak roof drift
2.43 1.55 1.28 2.28 1.82 2.48
(%)
ground
motion
peak roof drift
2.04 1.80 2.27 1.76 4.24 4.40
(%)
standard
deviation

2.39
2.26
1.17
1.70
1.06
LA37 LA38b LA39 LA40 mean

3.09
4.68
1.63
3.98
standard
deviation/mean
2.46
0.43
c
b
TABLE 15.77
PEAK ROOF DRIFT DEMANDS FOR STRUCTURE P5-PWUNC
ground
LA21a LA22 LA23
motion
peak roof drift
3.33 2.02 1.67
(%)
ground
LA31 LA32 LA33
motion
peak roof drift
2.65 2.47 2.70
(%)
standard
deviation
LA24 LA25 LA26 LA27 LA28 LA29c LA30

3.25
2.09
3.18
4.16
1.97
1.35
2.50
1.61
LA34 LA35 LA36b LA37 LA38 LA39 LA40 mean

3.31
6.34
6.99
3.77
5.52
1.93
c
b
569
5.61
3.34
standard
deviation/mean
0.48
8
7
6
5
P4-CWUNC
4
3
mean = 2.46%
2
1
0
0
50
100
150
200
250
300
350
400
350
400
(a)
8
7
P5-PWUNC
5
4
3
2
1
0
0
50
100
150
200
250
300
(b)
Fig. 15.67 Peak roof drift demands for uncoupled wall systems:
(a) Structure P4-CWUNC; (b) Structure P5-PWUNC
570

This chapter discusses the response of prototype Structures P1-CWUPT and P2PWUPT from Chapter 12 under earthquake loading. A series of nonlinear dynamic timehistory analyses of the structures are conducted using the twenty survival-level and
twenty design-level ground motion records from Chapter 14. Evaluations of the global
beams are provided. The results of the dynamic analyses indicate that the design
approach and procedures presented in Chapters 11 and 12 are mostly valid.
The following components of the proposed design approach provide, on average,
good correlation with the dynamic analysis results: (1) estimation of peak floor/roof drift
demands; (2) estimation of peak inter-story drift demands; (3) estimation of peak coupled
wall base shear forces; (4) estimation of coupling beam post-tensioning forces; and (5)
estimation of wall pier post-tensioning forces.
The following components of the design approach need further improvement: (1)
estimation of peak coupling beam chord rotations; (2) estimation of peak coupling beam
axial forces; (3) estimation of peak coupling beam shear forces; and (4) estimation of
peak top and seat angle deformations. The deficiencies of the proposed procedures in
providing satisfactory designs for the performance of the coupling beams may be due to
the development of significant higher mode effects under dynamic loading; however, this
is not investigated in the report.
Significant scatter is observed in the response of the prototype structures under the
ground motion records used in the dynamic analyses. This is possibly due to a large
variation in the intensity of the ground motion records. The large scatter in the dynamic
analysis results indicate that the seismic demands under some of the ground motions
(especially survival-level motions) can be significantly larger than the estimated design
demands.
Effective ground motion scaling methods are needed to reduce the scatter in the
dynamic analysis results. It may be possible to obtain more uniform levels of seismic
demands for the structures by scaling the ground motion records to a constant maximum
incremental velocity (MIV).
571
CHAPTER 16
SUMMARY, CONCLUSIONS, AND FUTURE WORK
16.1 Summary
This report investigates a new method to couple concrete walls using unbonded
post-tensioned steel beams, without embedding the beams into the walls. The report
includes analytical investigations on the nonlinear behavior of floor-level coupled wall
subassemblages as well as multi-story coupled wall structures under lateral loads,
experimental investigations of half-scale coupled wall subassemblage test specimens,
performance-based seismic design, and nonlinear dynamic time-history response
evaluations of prototype coupled wall structures under earthquake loading.
An introduction and overview of the report can be found in Chapter 1. Then,
Chapter 2 provides a summary of previous research related to the report as follows: (1)
coupled wall structural systems; (2) concrete coupled wall structures; (3) hybrid
structures with embedded steel coupling beams; (4) wall shear force demands in coupled
wall structures; (5) unbonded post-tensioning in building construction; and (6) behavior
of top and seat angles.
Chapter 3 describes a fiber element analytical model to investigate the nonlinear
behavior of floor-level hybrid coupled wall subassemblages. The details of a prototype
coupled wall subassemblage are also presented. The purpose of the analytical model is to
provide a basis for comparisons with subassemblage experimental results as described
later in the report and to provide a basis for the modeling of multi-story coupled wall
structures. A preliminary verification of the fiber element subassemblage model is done
based on comparisons with finite element models. An analytical model for embedded
steel coupling beam subassemblages is also developed to investigate the differences in
the seismic behaviors of coupled wall structures with unbonded post-tensioned and
embedded steel beams later in the report.
Chapter 4 presents an analytical investigation on the nonlinear behavior of
unbonded post-tensioned hybrid coupled wall subassemblages under lateral loads. The
properties, and angle properties on the behavior of the subassemblages, including the
lateral resistance and energy dissipation are investigated. These subassemblage analyses
form a foundation for the investigation of multi-story coupled wall behavior later in the
report. The prototype subassemblage and the fiber element model presented in Chapter 3
are used to conduct the analyses.
572
Chapter 5 presents an idealized beam end moment versus chord rotation

relationship for unbonded post-tensioned hybrid coupled wall subassemblages under
monotonic lateral loading. Procedures are developed to estimate the structure behavior
using basic principles of equilibrium, compatibility, and constitutive relationships. These
procedures, which are verified based on comparisons with the analytical model from
Chapter 3, can be used to conduct approximate, simplified analyses of coupled wall
subassemblages with different structural properties. The proposed procedures are used
later in the report as design tools and to develop idealized lateral load versus
displacement relationships for multi-story coupled wall structures.
Chapter 6 describes fiber element analytical models to investigate the nonlinear
behavior of multi-story unbonded post-tensioned hybrid coupled wall systems. The
details of two prototype coupled wall structures are also presented. The analytical models
are constructed by joining subassemblage models at the floor and roof levels. Systems
with monolithic cast-in-place reinforced concrete wall piers as well as with precast
concrete wall piers are considered. Verification of the cast-in-place reinforced concrete
wall pier model is presented based on comparisons with previous experiments of isolated
(i.e., uncoupled) wall specimens.
Chapter 7 presents an analytical investigation on the nonlinear behavior of multistory unbonded post-tensioned hybrid coupled wall structures under lateral loading. The
properties, and wall properties on the behavior of the structures, including the degree of
coupling, energy dissipation, and lateral displacement capacity are investigated. Systems
with precast concrete wall piers as well as monolithic cast-in-place reinforced concrete
wall piers are considered. The lateral load behavior of unbonded post-tensioned hybrid
coupled wall structures is compared with the behavior of structures that use embedded
steel coupling beams and structures without coupling.
Chapter 8 presents idealized base shear force versus roof drift relationships for
unbonded post-tensioned hybrid coupled wall structures under monotonic lateral loading.
Procedures are developed to estimate the structure behavior using basic principles of
equilibrium, compatibility, and constitutive relationships. These procedures, which are
verified based on comparisons with the analytical models from Chapter 6, can be used to
conduct approximate, simplified analyses of coupled wall systems with cast-in-place
reinforced concrete and precast concrete wall piers. The proposed procedures are used
later in the report as design tools for unbonded post-tensioned hybrid coupled wall
structures.
Chapter 9 describes the results from eleven half-scale experiments to investigate
the nonlinear behavior of unbonded post-tensioned hybrid coupled wall subassemblages.
Each test specimen includes a steel coupling beam and the adjacent concrete wall regions
at a floor level. The experiments were conducted in the Structural Systems Laboratory at
the University of Notre Dame by two other graduate students, Michael A. May and Brad
D. Weldon, as part of their research towards a Master of Science Degree in Civil
573
Engineering. A large portion of the material presented in this chapter is based on their
work (May et al. 2006).
Chapter 10 uses the experimental results from Chapter 9 to make necessary
modifications to the fiber element subassemblage analytical model described in Chapter
3. The revised analytical model is critically evaluated by comparing the analytical
estimations with the experimental measurements. The idealized coupling beam end
moment versus chord rotation relationship developed in Chapter 5 is also modified based
on the experimental results. These revised models are used in the remainder of the report
for the modeling, design, and analysis of coupled wall structures.
Chapter 11 proposes a two-level performance-based seismic design approach for
unbonded post-tensioned hybrid coupled wall structures. The objectives of the design
approach are to achieve the Collapse Prevention performance level under the
survival demand level and to achieve the Life Safety performance level under the
design demand level. The chapter describes a set of design criteria to achieve these
objectives by limiting the peak displacement demands of a structure to allowable target
design-level and survival-level displacements. In addition, design tools and procedures
are developed to estimate the seismic demands and capacities for unbonded posttensioned hybrid coupled wall structures.
Chapter 12 describes the seismic design of two prototype unbonded post-tensioned
hybrid coupled wall structures based on the performance-based design approach
introduced in Chapter 11.
Chapter 13 discusses the expected behavior of the prototype structures from
Chapter 12 under static lateral loading, including comparisons with the behavior of a
similar structure with embedded steel coupling beams. The analyses are conducted using
the model described in Chapter 6 including the modifications recommended in Chapter
10. Evaluations of the global behavior of the structures as well as the local behavior of
the wall piers and the coupling beams are provided.
Chapter 14 describes twenty design-level ground motion records and twenty
survival-level ground motion records used in the nonlinear dynamic time-history analyses
of the prototype structures from Chapter 12. These ground motion records are considered
to be representative of the seismic conditions used in the design of the structures.
Finally, Chapter 15 investigates the response of the prototype structures from
Chapter 12 under earthquake loading. A series of nonlinear dynamic time-history
analyses of the structures are conducted using the survival-level and design-level ground
motion records from Chapter 14. Evaluations of the global behavior of the structures as
well as the local behavior of the wall piers and the coupling beams are provided. The
effect of coupling is investigated by comparing the dynamic analysis results for the
prototype coupled wall systems with results obtained from dynamic analyses of the
structures with the coupling beams removed.
574
16.2 Conclusions
The research presented in this report has produced a number of conclusions. The
most important conclusion is that unbonded post-tensioned steel beams provide an
effective and feasible system to couple reinforced concrete wall piers in seismic regions.
Other important conclusions from this work are listed below.
Behavior of unbonded post-tensioned hybrid coupling systems
(1) As a result of post-tensioning, the initial lateral stiffness of an unbonded posttensioned coupling beam system is similar to the initial stiffness of a comparable system
with embedded steel coupling beams.
(2) In the nonlinear range, the lateral displacements of unbonded post-tensioned
coupling beams occur primarily as a result of the opening of gaps at the beam ends.
(3) The effect of gap opening on the lateral stiffness of the structure is small until
the gaps extend over a significant portion of the beam depth.
(4) In the proposed coupling system, top and seat angles are used at the beam-towall interfaces to provide energy dissipation during an earthquake. The angles also
provide a part of the coupling force of the structure, prevent sliding of the beam at the
beam-to-wall interfaces (together with friction resistance against sliding as induced by
post-tensioning), and can serve as temporary beam supports during construction.
(5) In a properly-designed unbonded post-tensioned hybrid coupling system, the
desired behavior under lateral loads is yielding of the top and seat angles due to the
opening of gaps at the beam ends, with little yielding and damage in the beam and the
wall piers. The yielded angles can be replaced after an earthquake.
(6) The behavior of properly-designed and detailed unbonded post-tensioned
hybrid coupling systems under lateral loads can be characterized by the following five
states: (i) decompression; (ii) yielding of the connection angles in tension; (iii) yielding
of the beam flange cover plates (if used) in compression; (iv) yielding of the beam
flanges in compression; and (v) yielding of the post-tensioning tendons. The yielding of
the angles in tension, accompanied by increased gap opening at the beam ends,
corresponds to a significant reduction in the lateral stiffness of the structure, which is
referred to as the softening state.
(7) The total beam post-tensioning force remains relatively constant until the
softening state is reached.
(8) The contact depth at the beam-to-wall interfaces remains relatively constant
after the softening state is reached.
575
(9) The hysteretic behavior of unbonded post-tensioned hybrid coupling systems

without yielding components (i.e., without the top and seat angles at the beam-to-wall
interfaces of the proposed system) is similar to a nonlinear-elastic type of behavior with
little energy dissipation.
(10) Unbonded post-tensioned hybrid coupling systems dissipate less energy per
displacement cycle than systems with embedded coupling beams. This is because, a
significant portion of the strength of the structure is provided by the post-tensioning steel,
which is designed to remain mostly linear elastic.
(11) Upon unloading from a large nonlinear lateral displacement, the posttensioning force provides a restoring effect that yields the tension angles back in
compression, closes the gaps, and pulls the wall piers and the beams back towards their
undisplaced position, reducing the residual lateral displacements. This results in a large
self-centering capability in unbonded post-tensioned coupling systems as compared with
systems using conventional embedded steel or monolithic concrete coupling beams.
(12) The use of unbonded tendons delays the yielding (i.e., nonlinear straining) of
the post-tensioning steel and reduces the tensile stresses transferred to the wall concrete.
Upon unloading from a large nonlinear lateral displacement of the structure, the initial
post-tensioning force is maintained as long as the yielding of the tendons is prevented.
The yielding of the post-tensioning steel can be controlled by the initial stress and
unbonded length of the post-tensioning tendons.
(13) The strength of an unbonded post-tensioned hybrid coupling system can be
controlled by the area of post-tensioning steel as well as the beam depth, the angle
thickness, and the angle vertical leg gage length.
(14) As a result of the opening of gaps at the beam-to-wall interfaces, the lateral
strength of an unbonded post-tensioned steel coupling beam is smaller than the strength
of a properly-designed and detailed embedded steel coupling beam with the same cross
section. This loss in strength is somewhat compensated by the top and seat angles used at
the beam ends.
(15) Experiments of floor-level coupled wall subassemblages demonstrate that
properly-designed unbonded post-tensioned hybrid coupling systems have excellent
stiffness, ductility, energy dissipation, and self-centering characteristics under large
nonlinear reversed cyclic lateral displacements. Little damage occurs in the coupling
beams and in the coupling regions of the wall piers, with most of the damage being
concentrated in the sacrificial top and seat angles. Corresponding to a given nonlinear
lateral displacement level, the amount of damage in the wall piers and beams of an
unbonded post-tensioned system is much smaller than the corresponding damage in
conventional systems with embedded steel or monolithic concrete beams.
576
(16) The post-tensioning anchors are critical components that can adversely affect
the performance of these structures. Premature fracture of single wires inside the anchors
of the unbonded post-tensioning strands was observed during the subassemblage
experiments. Anchor types other than the steel wedge/barrel system used in this
experimental program may improve the performance of the post-tensioning strands. It
may also be possible to grout a short length of the post-tensioning tendons near the
anchors to prevent premature fracture of the strands inside the anchors.
(17) Kinking of the post-tensioning strands at the beam-to-wall interfaces during
the experiments did not have any adverse effects on the performance of the strands.
(18) The angle-to-wall and angle-to-beam connections are also critical
components of unbonded post-tensioned hybrid coupling systems. Slip at these
connections can adversely affect the performance of these structures.
(19) The slip critical angle-to-beam connections used in the test specimens
worked well. If the required slip capacity cannot be achieved using a bolted angle-tobeam connection, it may also be possible to weld the angles to the beam flanges.
(20) The post-tensioned angle-to-wall connections used in the test specimens
performed well, with no yielding in the connection strands and no damage in the wall
concrete.
(21) Significant loss of the beam post-tensioning force can result in a loss of the
friction resistance against slip of the beam with respect to the wall piers. It may be
possible to increase the contribution of the top and seat angles to prevent slip at the beam
ends by fillet welding the vertical legs of the angles to the walls. It is important that this
weld does not prevent the development of a ductile yield mechanism in the angle.
(22) The experimental results showed that the use of cover plates on the coupling
beam flanges is not necessary for nominal flange initial stress values of up to 75% of the
steel yield strength, as long as compact sections are used for the beams.
(23) The use of confinement reinforcement inside the wall-contact regions and
steel plates at the beam-to-wall interfaces is necessary to prevent damage to the wall
concrete. Shim plates used in the experiments ensured contact between the beam flanges
and the wall piers, and maintained the initial lateral stiffness of the structure under cyclic
loading. Thus, the use of shim plates in practice is recommended.
Behavior of multi-story coupled wall structures
(1) Unbonded post-tensioned steel beams can provide stable levels of coupling,
similar to the levels of coupling that can be developed using conventional embedded steel
or monolithic concrete beams, between concrete walls over large nonlinear cyclic lateral
displacements.
577
(2) In order to achieve the same lateral strength, coupled wall structures with
unbonded post-tensioned steel coupling beams require beams of larger cross section than
walls with embedded steel coupling beams.
(3) The lateral strength of unbonded post-tensioned hybrid coupled walls can be
controlled by the area of the beam post-tensioning steel, as well as other parameters such
as the wall length, beam depth, and thickness and vertical leg gage length of the beam-towall connection angles.
(4) As compared with embedded steel and monolithic concrete coupling beams,
unbonded post-tensioned beams can provide a large restoring force to the wall piers,
reducing the residual lateral displacements upon unloading from a nonlinear
displacement. This results in a large self-centering capability of the coupled wall
structure. Thus, corresponding to a given peak nonlinear lateral displacement level during
an earthquake, the amount of residual displacement (i.e., permanent drift after the
earthquake) in an unbonded post-tensioned system is expected to be smaller than the
residual displacement in a system with embedded steel or monolithic concrete beams.
(5) Unbonded post-tensioned hybrid coupled walls dissipate less energy than
walls with embedded steel coupling beams. Most of the energy dissipation is provided by
the yielding of the top and seat angles at the beam-to-wall interfaces and the inelastic
behavior of the wall piers near the base.
(6) Unbonded post-tensioned coupled wall structures with cast-in-place reinforced
concrete wall piers have larger energy dissipation and a somewhat reduced but still large
self-centering capability as compared with structures that use post-tensioned precast
concrete wall piers.
(7) The lateral displacements and rotations of the tension-side and compressionside wall piers in a coupled wall structure under lateral loading are similar.
(8) The degree of coupling remains relatively constant as a coupled wall structure
is displaced laterally into the nonlinear range.
(9) A significant portion of the total initial force in the 2nd floor coupling beam
post-tensioning tendons is not transferred into the 2nd floor coupling beam as a result of
the fixed base conditions assumed for the wall piers. The initial axial forces in the upper
level coupling beams are close to the total initial post-tensioning forces, indicating that
the effect of the foundation/wall pier stiffness quickly diminishes above the base. The
wall pier length has the largest influence on the amount of initial post-tensioning force
transferred into the 2nd floor beam, since it results in the largest effect on the lateral
stiffness of the wall piers.
578
(10) Upon lateral displacements of the structure, the fixed foundation conditions
restrain the opening of gaps at the ends of the 2nd floor coupling beam, resulting in the
development of additional axial compression forces in the beam as the walls displace.
(11) During large nonlinear cyclic lateral displacements of an unbonded posttensioned hybrid coupled wall system, the 2nd floor coupling beam can loose most of its
initial compression even if the total force in the post-tensioning tendons remains close to
the initial force. This behavior is not observed in the upper level coupling beams, which
maintain most of their initial axial compression levels.
(12) The initial axial forces that develop in the coupling beams are not
significantly affected by the sequence of post-tensioning over the height of the structure.
Analytical modeling and procedures
(1) Comparisons between experimental results and analytical predictions of floorlevel unbonded post-tensioned hybrid coupled wall subassemblages show that the
analytical models provide good representations of not only the global load-deformation
behavior, but also the local behavior (e.g., gap opening behavior) of the proposed
coupling system.
(2) The test results indicate that the method developed by Kishi and Chen (1990)
provides reasonable estimates for the initial stiffness and yield strength of the top and seat
angles in tension.
(3) The subassemblage analytical models can be combined at the floor and roof
levels to develop models of multi-story coupled wall structures.
(4) In addition to the analytical models, relatively simple procedures are
developed to estimate the nonlinear lateral load-deformation behavior of unbonded posttensioned hybrid coupled wall structures. These procedures, which are verified with
results from the analytical models, can be used as design tools to conduct approximate
analyses of structures with different properties.
Seismic design and response evaluation
(1) The proposed seismic design procedures and tools can be used to determine
the properties of the wall piers, coupling beams, post-tensioning tendons, top and seat
angles, angle-to-beam and angle-to-wall connections, and beam-to-wall connections and
contact regions in an unbonded post-tensioned hybrid coupled wall structure.
(2) Since the coupling beams in an unbonded post-tensioned hybrid coupled wall
structure are not embedded into the wall piers, the selection of the beam shape and
dimensions is not affected by the reinforcing steel details in the walls, and the use of
widely available rolled cross sections is recommended. In comparison, the wall
579
reinforcement inside the embedment regions and the seismic detailing requirements in the
form of stiffener and cover plates often result in the use of fabricated beam cross sections
in embedded systems.
(3) Large levels of coupling can lead to large tension forces in the tension-side
wall pier and large compressive and shear forces in the compression-side wall pier, and
thus, should be avoided in design.
(4) Monotonic and cyclic static lateral load analyses of prototype unbonded posttensioned hybrid coupled wall structures indicate that the proposed design procedures are
valid. The design estimations of the various design parameters and structure capacities
match reasonably well with the static analysis results.
(5) Nonlinear dynamic time-history analyses of the prototype structures under
design-level and survival-level ground motions indicate that the following components of
the design approach provide, on average, good correlation with the analysis results: (1)
estimation of peak floor/roof drift demands; (2) estimation of peak inter-story drift
demands; (3) estimation of peak coupled wall base shear forces; (4) estimation of
coupling beam post-tensioning forces; and (5) estimation of wall pier post-tensioning
forces (in systems with precast concrete wall piers).
(6) Similarly, the following components of the design approach need further
improvement: (1) estimation of peak coupling beam chord rotations; (2) estimation of
peak coupling beam axial forces; (3) estimation of peak coupling beam shear forces; and
(4) estimation of peak top and seat angle deformations. The deficiencies of the proposed
procedures in providing satisfactory designs for the performance of the coupling beams
may be due to the development of significant higher mode effects under dynamic loading.
(7) Significant scatter is observed in the response of the prototype structures under
the ground motion records used in the dynamic analyses of the structures. This is possibly
due to a large variation in the intensity of the ground motion records. The large scatter in
the dynamic analysis results indicate that the seismic demands under some of the ground
motions (especially survival-level motions) can be significantly larger than the estimated
design demands. Effective ground motion scaling methods are needed to reduce the
scatter in the seismic demands.
(8) The dynamic analysis results indicate that it may be possible to obtain more
uniform levels of seismic demands for the walls by scaling the ground motion records to
a constant maximum incremental velocity (MIV).
(9) Nonlinear dynamic time-history analyses indicate that the proposed BP-type
equivalent nonlinear single-degree-of-freedom system provides an effective tool to
estimate the lateral displacement response of unbonded post-tensioned hybrid coupled
wall structures under earthquake loading.
580
(10) The roof displacement time-history results demonstrate the self-centering

capability of the prototype structures as indicated by the oscillations about close to the
zero-displacement position, with little residual (i.e., permanent) displacements at the end
of a ground motion.
(11) The dynamic analysis results also show that the lateral displacement
responses of the prototype structures decay (i.e., decrease) at a reasonable rate. This
indicates the presence of an adequate level of energy dissipation in the structures.
(12) The peak lateral displacement demands of the prototype coupled wall
structures are smaller than the displacement demands of the wall piers with the coupling
beams removed.
(13) The peak absolute roof accelerations of the prototype structures do not seem
to have a strong correlation with the ground motion intensity.
16.3 Future Work
The following areas of future work are recommended based on the research
described in this report:
(1) Experiments of multi-story unbonded post-tensioned hybrid coupled wall
structures, especially substructures near the base.
(2) Investigation of low cycle fatigue fracture of the top and seat angles.
(3) Investigation of wire fracture inside the post-tensioning strand anchors.
(4) More detailed investigations of the angle-to-beam and angle-to-wall
connections, including welded connection configurations.
(5) Investigation of other energy dissipation systems that utilize the gap opening
displacements at the beam-to-wall interfaces.
(6) Investigation of higher mode effects on the response of the coupling beams.
(7) Investigation of high-rise (i.e., twenty stories and more) unbonded posttensioned hybrid coupled wall structures.
(8) Investigation of unbonded post-tensioning as a possible method for the
seismic retrofit and strengthening of existing concrete walls. The post-tensioning tendons
can be placed outside the walls and above and below the slab for this purpose.
581
REFERENCES
ACI Committee 318, Building Code Requirements for Structural Concrete (ACI 318-99)
and Commentary (ACI 318R-99), American Concrete Institute, Farmington Hills, MI,
1999.
ACI Committee 318, Building Code Requirements for Structural Concrete (ACI 318M02) and Commentary (ACI 318RM-02), American Concrete Institute, Farmington Hills,
MI, 2002.
ACI Innovation Task Group 1 and Collaborators, Acceptance Criteria for Moment
Frames Based on Structural Testing (T1.1-01) and Commentary (T1.1R-01), American
Concrete Institute, Farmington Hills, MI, 2001, 11 pp.
ACI Innovation Task Group and Collaborators, Acceptance Criteria for Special Precast
Structural Walls Based on Validation Testing and Commentary, DRAFT, ITG T5.1,
2004.
ACI Innovation Task Group 1 and Collaborators, Special Hybrid Moment Frames
Composed of Discretely Jointed Precast and Post-Tensioned Concrete Members, ACI
Proposed Standard T1.2-03 and Commentary ACI T1.2R-03, American Concrete
Institute, Farmington Hills, MI, 2003.
AISC, Load and Resistance Factor Design Manual of Steel Construction, Vol. I and
Vol. II, Second Edition, American Institute of Steel Construction, Chicago, IL, 1998.
Ajrab, J., Pekcan, G., and Mander, J., Rocking Wall-Frame Structures with
Supplemental Tendon Systems, Journal of Structural Engineering, American Society of
Civil Engineers, Vol. 130, No. 6, June 2004, pp. 895-903.
Aktan, A. and Bertero, V., Seismic Response of R/C Frame-Wall Structures, Journal of
Structural Engineering, American Society of Civil Engineers, Vol. 110, No. 8, August
1984, pp. 1803-1821.
Aktan, A. and Bertero, V., RC Structural Walls: Seismic Design for Shear, Journal of
Structural Engineering, American Society of Civil Engineers, Vol. 111, No. 8, August
1985, pp. 1775-1791.
Aktan, A. and Bertero, V., Evaluation of Seismic Response of RC Buildings Loaded to
Failure, Journal of Structural Engineering, American Society of Civil Engineers, Vol.
113, No. 5, May 1987, pp. 1092-1108.
582
Aktan, A., Bertero, V., and Piazza, M. Prediction of the Seismic Responses of R/C
Frame-Coupled Wall Structures, Earthquake Engineering Research Center, Report No.
UCB/EERC-82/12, Berkeley, CA, 1982, 182 pp.
Aktan, H., Yousef-Agha, W., and Olowokere, O., Seismic Response of Low-Rise Steel
Frames, Journal of Structural Engineering, American Society of Civil Engineers, Vol.
115, March 1989, pp. 594-607.
Al-Sulaimani, G. and Roessett, J., Design Spectra for Degrading Systems, Journal of
Structural Engineering, Vol. 111, No. 12, December 1985, pp. 2611-2623.
ANSYS, Release 5.4, Swanson Analysis Systems IP, Inc., Canonsburg, PA, 1997.
Aoyama, H., Earthquake Resistant Design of Reinforced Concrete Frame Buildings with
Flexural Walls, in Earthquake Resistance of Reinforced Concrete Structures, A
Volume Honoring Hiroyuki Aoyama, Editor: T. Okada, Department of Architecture,
Faculty of Engineering, University of Tokyo, Japan, November 1993, pp. 78-100.
Aristizabal-Ochoa, J., Seismic Behavior of Slender Coupled Wall Systems, Journal of
Structural Engineering, American Society of Civil Engineers, Vol. 113, No. 10, October
1987, pp. 2221-2234.
ASCE, Minimum Design Loads for Buildings and Other Structures, ASCE 7-02,
American Society of Civil Engineers, 2003, 376 pp.
ASCE, Prestandard and Commentary for the Seismic Rehabilitation of Buildings,
FEMA-356, American Society of Civil Engineers, Reston, Virginia, and Federal
Emergency Management Agency (FEMA), Washington, D.C., November 2000.
Azizinamini, A., Cyclic Characteristics of Bolted Semi-Rigid Steel Beam-to-Column
Connections, Ph.D. Dissertation, University of South Carolina, 1985.
Borzi, B. and Elnashai, A., Refined Force Reduction Factors for Seismic Design,
Engineering Structures, Vol. 22, No. 10, October 2000, pp. 1244-1260.
BSSC, NEHRP Guidelines for the Seismic Rehabilitation of Buildings, FEMA-273,
Building Seismic Safety Council, Federal Emergency Management Agency (FEMA),
Washington, D.C., October 1997.
BSSC, NEHRP Recommended Provisions for Seismic Regulations for New Buildings
and Other Structures, Part 1 and Part 2, FEMA-302 and FEMA-303, Building Seismic
Safety Council, Federal Emergency Management Agency (FEMA), Washington, D.C.,
February 1998.
583
Chaallal, O., Finite Element Model for Seismic RC Coupled Walls Having Slender
Coupling Beams, Journal of Structural Engineering, American Society of Civil
Engineers, Vol. 118, No. 10, October 1992, pp. 2936-2943.
Chaallal, O. and Gauthier, D., Seismic Shear Demand on Wall Segments of Ductile
Coupled Shear Walls, Canadian Journal of Civil Engineering, Vol. 27, No. 3, 2000, pp.
506-522.
Chaallal, O., Gauthier, D., and Malenfant, P., Classification Methodology for Coupled
Shear Walls, Journal of Structural Engineering, American Society of Civil Engineers,
Vol. 122, No. 12, December 1996, pp. 1453-1458.
Chaallal, O., Guizani, L., and Malenfant, P., Drift-Based Methodology for Seismic
Proportioning of Coupled Shear Walls, Canadian Journal of Civil Engineering, Vol. 23,
No. 5, October 1996, pp. 1030-1040.
Cheok, G. and Lew, H., Model Precast Concrete Beam-to-Column Connections Subject
to Cyclic Loading, PCI Journal, Precast/Prestressed Concrete Institute, Vol. 38, No. 4,
July-August 1993, pp. 80-92.
Cheok, G. and Stone, W., Performance of 1/3 Scale Model Precast Concrete BeamColumn Connections Subjected to Cyclic Inelastic Loads Report No. 4, Report
NISTIR 5436, National Institute of Standards and Technology, Gaithersburg, MD, 1994,
59 pp.
Cheok, G., Stone, W., and Lew, H., Performance of 1/3-Scale Model Precast Concrete
Beam-Column Connections Subjected to Cyclic Inelastic Loads - Report No. 3, NISTIR
5246, National Institute of Standards and Technology, Gaithersburg, MD, August 1993.
Cheok, G., Stone, W., Kunnath, S., Seismic Response of Precast Concrete Frames with
Hybrid Connections, ACI Structural Journal, American Concrete Institute, Vol. 95, No.
5, September-October 1998, pp. 527-539.
Cheok, G., Stone, W., and Nakaki, S., Simplified Design Procedure for Hybrid Precast
Concrete Connections, NISTIR 5765, National Institute of Standards and Technology,
Gaithersburg, MD, 1996, 81 pp.
Chopra, A., Dynamics of Structures: Theory and Applications to Earthquake Engineering,
Prentice Hall, Inc., Englewood Cliffs, New Jersey, 1995, 729 pp.
Christopoulos, C., Filiatrault, A., and Folz, B., Seismic Response of Self-Centering
Hysteretic SDOF Systems, Earthquake Engineering and Structural Dynamics, Vol. 31,
No. 5, May 2002, pp. 1131-1150.
584
Christopoulos, C., Filiatrault, A., Uang, C., and Folz, B., Posttensioned Energy
Dissipating Connections for Moment-Resisting Steel Frames, Journal of Structural
Engineering, American Society of Civil Engineers, Vol. 128, No. 9, September 2002, pp.
1111-1120.
Clough, R. and Penzien, J., Dynamics of Structures, McGraw-Hill, New York, 1993.
Cosenza, E. and Pecce, M., Shear and Normal Stresses Interaction in Coupled Structural
Systems, Journal of Structural Engineering, American Society of Civil Engineers, Vol.
127, No. 1, January 2001, pp. 84-88.
CSI, SAP2000-Integrated Finite Element Analysis and Design of Structures, Computers
and Structures, Inc., Berkeley, CA, 1999.
Deierlein, G., Proposed Ground Motions for Dynamic Analyses, 4th Joint Technical
Coordinating Committee Meeting, U.S.-Japan Cooperative Earthquake Research Program
on Composite and Hybrid Structures, October 1997, Monterey, CA, 12 pp.
Dywidag Systems International http://www.dywidag-systems.com/download/us/DSI_
Bar_Post_Tensioning.pdf, 2002
Eberhard, M. and Sozen, M., Behavior-Based Method to Determine Design Shear in
Earthquake-Resistant Walls, Journal of Structural Engineering, American Society of
Civil Engineers, Vol. 119, No. 2, February 1993, pp. 619-640.
Elghadamsi, F. and Mohraz, B., Inelastic Earthquake Spectra, Earthquake Engineering
and Structural Dynamics, Vol. 15, 1987, pp. 91-104.
El-Sheikh, M., Sause, R., Pessiki, S., and Lu, L.W., Seismic Analysis, Behavior, and
Design of Unbonded Post-Tensioned Precast Frames, Research Report No. EQ-97-02,
Department of Civil and Environmental Engineering, Lehigh University, Bethlehem, PA
18015, November 1997, 316 pp.
El-Sheikh, M., Sause, R., Pessiki, S., and Lu, L.W., Seismic Behavior, and Design of
Unbonded Post-Tensioned Precast Concrete Frames, PCI Journal, Precast/Prestressed
Concrete Institute, Vol. 44, No. 3, May-June 1999, pp. 54-71.
El-Tawil, S., Kuenzli, C., and Hassan, M., Pushover of Hybrid Coupled Walls I: Design
and Modeling, Journal of Structural Engineering, Vol. 128, No. 10, 2002, pp. 12711281.
El-Tawil, S. and Kuenzli, C., Pushover of Hybrid Coupled Walls II: Analysis and
Behavior, Journal of Structural Engineering, Vol. 128, No. 10, 2002, pp. 1282-1289.
585
Fajfar, P. and Fischinger, M., N2-A Method for Non-Linear Seismic Analysis of
Regular Structures, 9th World Conference on Earthquake Engineering, 1988, TokyoKyoto, Japan.
Farrow, K. and Kurama, Y., Capacity-Demand Index Relationships for PerformanceBased Seismic Design, Structural Engineering Research Report #NDSE-01-02,
Department of Civil Engineering and Geological Sciences, University of Notre Dame,
Notre
Dame,
IN,
November
2001,
(available
for
download
at
http://www.nd.edu/~concrete).
Farrow, K. and Kurama, Y., CDSPEC (Capacity-Demand SPECtra) Program,
Department of Civil Engineering and Geological Sciences, University of Notre Dame,
Notre
Dame,
IN,
November
2001,
(available
for
download
at
http://www.nd.edu/~concrete)
Farrow, K. and Kurama, Y., SDOF Demand Index Relationships for Performance-Based
Seismic Design, Earthquake Spectra, Vol. 19, No. 4, November 2003, pp. 799-838.
Farrow, K. and Kurama, Y., SDOF Displacement Ductility Demands Based on Smooth
Ground Motion Response Spectra, Engineering Structures, Elsevier, Vol. 26, No. 12,
October 2004, pp. 1713-1733.
FEMA, NEHRP Recommended Provisions for Seismic Regulations for New Buildings
and Other Structures, Federal Emergency Management Agency, 1997.
Fintel, M. and Ghosh, S., Seismic Resistance of a 31 Storey Frame-Wall Building Using
Dynamic Inelastic Response History Analysis, 7th World Conference on Earthquake
Engineering, Istanbul, 1980, pp. 379-386.
Fintel, M. and Ghosh, S., Case Study of Aseismic Design of a 16 Storey Coupled Wall
Structure Using Inelastic Dynamic Analysis, ACI Structural Journal, American
Garlock, M., Ricles, J., and Sause, R., Cyclic Load Tests and Analysis of Bolted Topand-Seat Angle Connections, Journal of Structural Engineering, Vol. 129, No. 12,
December 2003, pp. 1615-1625.
Ghosh, S. and Markevicius, V., Design of Earthquake Resistant Shearwalls to Prevent
Shear Failure, 4th U.S. National Conference on Earthquake Engineering, Earthquake
Engineering Research Institute, Palm Springs. CA, Vol. 2, May 1990, pp. 905-913.
Gong, B. and Shahrooz, B., Steel/Composite Coupling Beams-Behavior and Design,
4th Joint Technical Coordinating Committee Meeting, U.S.-Japan Cooperative
Earthquake Research Program on Composite and Hybrid Structures, October 1997,
Monterey, CA, 20 pp.
586
Gong, B., Shahrooz, B., and Gillum, A., Seismic Behavior and Design of Composite
Coupling Beams, in Composite Construction in Steel and Concrete III, New York:
American Society of Civil Engineers, 1997, pp. 258-271.
Gong, B. and Shahrooz, B., Seismic Behavior of Composite Coupled Wall Systems,
University of Cincinnati, College of Engineering, Report No. UC-CII 98/01, June 1998,
274 pp.
Gong, B. and Shahrooz, B., Concrete-Steel Composite Coupling Beams. I: Component
Testing, Journal of Structural Engineering, American Society of Civil Engineers, Vol.
127, No. 6, June, 2001a, pp. 625-631.
Gong, B. and Shahrooz, B., Concrete-Steel Composite Coupling Beams. II:
Subassembly Testing and Design Verification, Journal of Structural Engineering,
American Society of Civil Engineers, Vol. 127, No. 6, June, 2001b, pp. 632-638.
Goto, Y., Suzuki, S., and Chen, W., Analysis of Critical Behavior of Semi-Rigid Frames
With or Without Load History in Connection, Journal of Solid Structures, Vol. 27, 1991,
pp. 467-483.
Hall, J., Heaton, T., Halling, M., and Wald, D., Near-Source Ground Motion and Its
Effects on Flexible Buildings, Earthquake Spectra, Vol. 11, No. 4, 1995, pp. 569-605.
Harries, K. Seismic Response of Steel Beams Coupling Concrete Walls, M.S. Thesis,
McGill University, May 1992, 95 pp.
Harries, K., Seismic Design and Retrofit of Coupled Walls Using Structural Steel,
Ph.D. Dissertation, McGill University, Canada, 1995, 229 pp.
Harries, K., Design and Analysis of Four Prototype Eighteen Story Coupled Wall
Structures, McGill University, Structural Engineering Series Report No. 96-01, 1996,
122 pp.
Harries, K., Ductility and Deformability of Coupling Beams in Reinforced Concrete
Coupled Walls, Earthquake Spectra, Vol. 17, No. 3, August 2001, pp. 457-478.
Harries, K. and McNeice, D., Performance-Based Design of High-Rise Coupled Wall
Systems, The Structural Design of Tall and Special Structures, Vol. 15 No. 3, 2006, pp.
289-306.
Harries, K. and Shahrooz, B., Hybrid Coupled Wall Systems, Concrete International,
American Concrete Institute, Vol. 27, No. 5, May 2005, pp. 45-51.
587
Harries, K., Cook, W., Redwood, R., and Mitchell, D., Concrete Walls Coupled by
Ductile Steel Link Beams, 10th World Conference on Earthquake Engineering, Vol. 6,
Madrid, July 1992, pp. 3205-3211.
Harries, K., Mitchell, D., Cook, W., and Redwood, R., Seismic Response of Steel
Beams Coupling Concrete Walls, Journal of Structural Engineering, American Society
of Civil Engineers, Vol. 119, No. 12, December 1993, pp. 3611-3629.
Harries, K., Cook, W., Mitchell, D., and Redwood, R., The Use of Steel Beams to
Couple Reinforced Concrete Walls, 7th Canadian Conference on Earthquake
Engineering, Montreal, June 1994, pp. 509-516.
Harries, K., Mitchell, D., Redwood, R., and Cook, W., Seismic Design of Coupled
Walls-A Case for Mixed Construction, Canadian Journal of Civil Engineering, Vol. 24,
1997, pp. 448-459.
Harries, K., Mitchell, D., Redwood, R., and Cook, W., Nonlinear Seismic Response
Predictions of Walls Coupled with Steel and Concrete Beams, Canadian Journal of
Civil Engineering, Vol. 25, No. 5, October 1998, pp. 803-818.
Harries, K., Gong, B., and Shahrooz, B., Behavior and Design of Reinforced Concrete,
Steel, and Steel-Concrete Coupling Beams, Earthquake Spectra, Vol. 16, No. 4,
November 2000, pp. 775-799.
Harries, K., Fortney, P., Shahrooz, B., and Brienen, P., Design of Practical Diagonally
Reinforced Concrete Coupling Beams A Critical Review of ACI 318 Requirements,
ACI Structures Journal, American Concrete Institute, Vol. 102, No. 6, 2005, pp. 876-882.
Hassan, M. and El-Tawil, S., Inelastic Dynamic Behavior of Hybrid Coupled Walls,
Journal of Structural Engineering, American Society of Civil Engineers, Vol. 130, No. 2,
February 2004, pp. 285-296.
Hibbitt, Karlsson, and Sorensen, ABAQUS Users Manual, Hibbitt, Karlsson, &
Sorensen, Inc., Version 5.8, 1998.
Holden T., Restrepo J., and Mander J., Seismic Performance of Precast Reinforced and
Prestressed Concrete Walls, Journal of Structural Engineering, American Society of
Civil Engineers, Vol. 129, No. 3, March 2003, pp. 286-296.
Holden, T., Restrepo, J., and Mander, J., A Comparison of the Seismic Performance of
Precast Wall Construction: Emulation and Hybrid Approaches, Rep. No. 2001-4, Dept.
of Civil Engineering, Univ. of Canterbury, Christchurch, New Zealand, 2001.
ICBO, Uniform Building Code, International Conference of Building Officials,
Whittier, CA, 1997.
588
ICC, 2000 International Building Code (2000 IBC), International Code Council, Falls
Church, VA, December 2000.
Kabeyasawa, T., Ultimate-State Design of Wall-Frame Structures, in Earthquake
Resistance of Reinforced Concrete Structures, A Volume Honoring Hiroyuki Aoyama,
Editor: T. Okada, Department of Architecture, Faculty of Engineering, University of
Tokyo, Japan, November 1993, pp. 431-440.
Keintzel, E., Seismic Design Shear Forces in RC Cantilever Shear Wall Structures,
European Earthquake Engineering, Vol. IV, No. 3, 1990, pp. 7-16.
Kishi, N. and Chen, W., Moment-Rotation Relations of Semirigid Connections with
Angles, Journal of Structural Engineering, American Society of Civil Engineers, Vol.
116, No. 7, July 1990, pp. 1813-1834.
Kurama, Y., Seismic Analysis, Behavior, and Design of Unbonded Post-Tensioned
Precast Concrete Walls, Ph.D. Dissertation, Department of Civil and Environmental
Engineering, Lehigh University, Bethlehem, PA 18015, May 1997, 491 pp.
Kurama, Y., Seismic Design of Unbonded Post-Tensioned Precast Walls with
Supplemental Viscous Damping, ACI Structural Journal, American Concrete Institute,
Vol. 97, No. 4, July-August 2000, pp. 648-658.
Kurama, Y., Simplified Seismic Design Approach for Friction-Damped Unbonded PostTensioned Precast Walls, ACI Structural Journal, American Concrete Institute, Vol. 98,
No. 5, September-October 2001, pp. 705-716.
Kurama, Y., Hybrid Post-Tensioned Precast Concrete Walls for Use in Seismic
Regions, PCI Journal, Precast/Prestressed Concrete Institute, Vol. 47, No. 5,
September-October 2002, pp. 36-59.
Kurama, Y., Seismic Design of Partially Post-Tensioned Precast Concrete Walls, PCI
Journal, Precast/Prestressed Concrete Institute, Vol. 50, No. 4, July-August 2005, pp.
100-125.
Kurama, Y. and Farrow, K., Ground Motion Scaling Methods for Different Site
Conditions and Structure Characteristics, Earthquake Engineering and Structural
Dynamics, Vol. 32, No. 15, 2003, pp. 2425-2450.
Kurama, Y., Pessiki, S., Sause, R., Lu, L., and El-Sheikh, M., Analytical Modeling and
Lateral Load Behavior of Unbonded Post-Tensioned Precast Concrete Walls, Research
Report No. EQ-96-02, Department of Civil and Environmental Engineering, Lehigh
University, Bethlehem, PA 18015, November 1996, 190 pp.
589
Kurama, Y., Sause, R., Pessiki, S., Lu, L., and El-Sheikh, M., Seismic Design and
Response Evaluation of Unbonded Post-Tensioned Precast Concrete Walls, Research
Report No. EQ-97-01, Department of Civil and Environmental Engineering, Lehigh
University, Bethlehem, PA 18015, November 1997, 184 pp.
Kurama, Y., Sause, R., Pessiki, S., and Lu, L., Seismic Behavior and Design of
Unbonded Post-Tensioned Precast Concrete Walls, 6th U.S. National Conference on
Earthquake Engineering, Earthquake Engineering Research Institute, Seattle, WA, May
31-June 4, 1998a, 12 pp.
Kurama, Y., Sause, R., Pessiki, S., and Lu, L., Seismic Design of Unbonded PostTensioned Precast Concrete Walls, Structural Engineers World Congress, San Francisco,
CA, July 19-23, 1998b, 8 pp.
Kurama, Y., Sause, R., Pessiki, S., and Lu, L., Lateral Load Behavior and Seismic
Design of Unbonded Post-Tensioned Precast Concrete Walls, ACI Structural Journal,
American Concrete Institute, Vol. 96, No. 4, July-August 1999a, pp. 622-632.
Kurama, Y., Pessiki, S., Sause, R., and Lu, L., Seismic Behavior and Design of
Unbonded Post-Tensioned Precast Concrete Walls, PCI Journal, Precast/Prestressed
Concrete Institute, Vol. 44, No. 3, May-June 1999b, pp. 72-89.
Kurama, Y., Sause, R., Pessiki, S., and Lu, L., Seismic Response Evaluation of
Unbonded Post-Tensioned Precast Walls, ACI Structural Journal, American Concrete
Institute, Vol. 99, No. 5, September-October 2002, pp. 641-651.
Kwan, W. and Billington, S., Seismic Behavior of Precast Concrete Pier Cap-Beam-toColumn Joints with Horizontal and Vertical Prestressing, Earthquake Engineering: the
8th Canadian Conference, 1999, pp. 439-444.
Lawson, R., Vance, V., and Krawinkler, H., Nonlinear Static Push-Over Analysis - Why,
When, and How? 5th US Conference in Earthquake Engineering, Earthquake
Engineering Research Institute, Oakland, California, Vol. 1, 1994.
Lin, J. and Mahin, S., Effect of Inelastic Behavior on the Analysis and Design of
Earthquake Resistant Structures, Report No. UCB/EERC-85/08, Earthquake
Engineering Research Center, University of California, Berkeley, CA, June 1985.
Lorenz, R., Kato, B., and Chen, W., Semi-Rigid Connections in Steel Frames, Council
on Tall Buildings and Urban Habitat, Committee 43, McGraw-Hill, Inc., 1993.
MacRae, G. and Priestley, M., Precast Post-Tensioned Ungrouted Concrete BeamColumn Subassemblage Tests, Report No. PRESSS-94/01, Department of Applied
Mechanics and Engineering Sciences, Structural Systems, University of California, San
Diego, CA, March 1994, 124 pp.
590
Mander, J., Priestley, M., and Park, R., Theoretical Stress-Strain Model for Confined
Concrete, Journal of Structural Engineering, American Society of Civil Engineers, Vol.
114, No. 8, August 1988, pp. 1804-1826.
Mander, J., Chen, S., and Pekcan, G., Low-Cycle Fatigue Behavior of Semi-Rigid Topand-Seat Angle Connections, Engineering Journal, Vol. 31, No. 3, 1994, pp. 111-121.
Marcakis, K. and Mitchell, D., Precast Concrete Connections with Embedded Steel
Members, PCI Journal, Precast/Prestressed Concrete Institute, July-August 1980, pp.
88-116.
MATLAB, 2000, The Math Works, Inc., Natick, MA.
Matsuoka, K., Kishi, N., Chen, W., and Goto, Y., Analysis Program for the Design of
Flexibly Jointed Frames, Computers and Structures, Vol. 49, Nov. 1993, pp. 705-713.
Mattock, A. and Gaafar, G., Strength of Embedded Steel Sections as Brackets, ACI
Structural Journal, American Concrete Institute, Vol. 79, No. 2, March-April 1982, pp.
83-93.
May, M., Kurama, Y. and Shen, Q., Experimental Evaluation of Unbonded PostTensioned Hybrid Coupled Wall Subassemblages, Structural Engineering Research
Report, No. NDSE-06-01, Civil Engineering and Geological Sciences, University of
Notre Dame, Notre Dame, Indiana, April 2006.
Miranda, E., Site-Dependent Strength-Reduction Factors, Journal of Structural
Engineering, Amer. Soc. of Civil Eng., Vol. 119, No. 12, December 1993, pp. 3503-3519.
Miranda, E., Seismic Evaluation and Upgrading of Existing Buildings, Ph.D.
Dissertation, Department of Civil Engineering, University of California, Berkeley,
California, 1991.
Morgen, B. and Kurama, Y., A Cast-Steel Friction Damper for Post-Tensioned Precast
Concrete Beam-to-Column Joints, National Technical and Operating Conference,
Session on Castings for Building, Steel Founders Society of America, Chicago, IL,
November 6-8, 2003.
Morgen, B. and Kurama, Y., A Friction Damper for Post-Tensioned Precast Concrete
Beam-to-Column Joints, PCI Journal, Precast/Prestressed Concrete Institute, Vol. 49,
No. 4, July-August 2004a, pp. 112-133.
Morgen, B. and Kurama, Y., A Friction Damper for Post-Tensioned Precast Concrete
Beam-to-Column Joints, 13th World Conference on Earthquake Engineering, Vancouver,
BC, Canada, August 1-6, 2004b, 15 pp. (CD-ROM)
591
Morgen, B. and Kurama, Y., An Innovative Friction-Damped Precast Concrete Frame

Structure for Seismic Regions, ASCE Structures Congress, American Society of Civil
Engineers, New York, NY, April 20-24, 2005, 12 pp.
Munshi, J. and Ghosh, S., Displacement-Based Seismic Design for Coupled Wall
Systems, Earthquake Spectra, Vol. 16, No.3, August 2000, pp. 621-642.
Naeim, F., On Seismic Design Implications of the 1994 Northridge Earthquake
Records, Earthquake Spectra, Vol. 11, No. 1, 1995, pp. 91-109.
Nassar, A. and Krawinkler, H., Seismic Demands for SDOF and MDOF Systems,
Report No. 95, Department of Civil Engineering, Stanford University, June 1991, 204 pp.
Newmark, N. and Hall, W., Procedures and Criteria for Earthquake Resistant Design,
Building Science Series, No. 46, 1973, National Bureau of Standards, U.S. Department of
Commerce, Washington, D.C.
Newmark, N. and Riddell, R., Inelastic Spectra for Seismic Design, 7th World
Conference on Earthquake Engineering, Vol. 4, 1980, Ankara, pp. 129-136.
Otani, S., Teshigawara, M., Hayashi, M., Ishii, T., Kawabata, I., and Kani, N.,
Earthquake Response Characteristics of Reinforced Concrete Structural Walls, 4th
Meeting of the U.S.-Japan Joint Technical Coordinating Committee on Precast Seismic
Structural Systems, Tsukuba, Japan, May 1994.
Paulay, T., Ductility of Reinforced Concrete Shear Walls for Seismic Areas, in
Reinforced Concrete Structures in Seismic Zones, American Concrete Institute, SP-53,
1977, pp. 127-147.
Paulay, T., A Displacement-Focused Seismic Design of Mixed Building Systems,
Earthquake Spectra, Vol. 18, No.4, November 2002, pp. 689-718.
Paulay, T. and Priestley, M., Seismic Design of Reinforced Concrete and Masonry
Buildings, John Wiley & Sons, Inc., 1992, 744 pp.
Paulay, T. and Santhakumar, A., Ductile Behavior of Coupled Shear Walls, Journal of
Structural Engineering, American Society of Civil Engineers, Vol. 102, No. ST1,
January 1976, pp. 93-108.
PCI, PCI Design Handbook, Precast/Prestressed Concrete Institute, Sixth Edition, 2004.
Pekau, O. and Cistra, V., Behavior of Nonlinear Coupled Shear Walls with Flexible
Bases, Canadian Journal of Civil Engineering, Vol. 18, February 1989, pp. 45-54.
592
Peng, S., Ricles, J., Sause, R., and Lu, L., Experimental Evaluation of a Post-Tensioned
Moment Connection for Steel and Composite Frames in Seismic Zones, 6th ASCCS
International Conference on Steel-Concrete Composite Structure, Los Angeles, March
22-24, 2000, pp. 721-728.
Perez, F., Lateral Load Behavior of Precast Concrete Walls with Ductile Vertical Joint
Connectors, M.S. Thesis, Department of Civil and Environmental Engineering, Lehigh
University, Bethlehem, PA, December 1998.
Perez, F., Pessiki, S., and Sause, R., Seismic Design of Unbonded Post-Tensioned
Precast Concrete Walls with Vertical Joint Connectors, PCI Journal, Precast/Prestressed
Concrete Institute, Vol. 49, No.1, January-February, 2004a, pp. 58-79.
Perez, F., Pessiki, S., and Sause, R., Lateral Load Behavior of Unbonded PostTensioned Precast Concrete Walls with Vertical Joints, PCI Journal, Precast/Prestressed
Concrete Institute, Vol. 49, No.2, March-April, 2004b, pp. 48-64.
Prakash, V., Powell, G., and Campbell, S., DRAIN-2DX Base Program Description and
User Guide; Version 1.10, Rep. No. UCB/SEMM-93/17, Department of Civil
Engineering, University of California, Berkeley, CA, 1993.
Priestley, M., The PRESSS Program: Current Status and Proposed Plans for Phase III,
PCI Journal, Precast/Prestressed Concrete Institute, Vol. 41, No. 2, March-April 1996,
pp. 22-40.
Priestley, M. and MacRae, G., Seismic Tests of Precast Beam-to-Column Joint
Subassemblages With Unbonded Tendons, PCI Journal, Precast/Prestressed Concrete
Institute, Vol. 41, No. 1, January-February 1996, pp. 64-81.
Priestley, M. and Tao, J., Seismic Response of Precast Prestressed Concrete Frames
with Partially Debonded Tendons, PCI Journal, Precast/Prestressed Concrete Institute,
Vol. 38, No.1, January-February 1993, pp. 58-69.
Priestley, M., Sritharan S., Conley J., and Pampanin S., Preliminary Results and
Conclusions from the PRESSS Five-Story Precast Concrete Test Building, PCI Journal,
Precast/Prestressed Concrete Institute, Vol. 44, No.6, November-December 1999, pp. 4267.
Qi, X. and Moehle, J., Displacement Design Approach for Reinforced Concrete
Structures Subjected to Earthquakes, Report No. EERC 91/02, Earthquake Engineering
Research Center, University of California, Berkeley, California, 1991.
Rahman, A. and Restrepo, J., Earthquake Resistant Precast Concrete Buildings: Seismic
Performance of Cantilever Walls Prestressed Using Unbonded Tendons, Rep. No. 20005, Dept. of Civil Engineering, Univ. of Canterbury, Christchurch, New Zealand, 2000.
593
Rahnama, M. and Krawinkler, H., Effects of Soft Soil and Hysteresis Model on Seismic
Demands, John A. Blume Earthquake Engineering Center, Report No. 108, Department
of Civil Engineering, Stanford University, July 1993.
Raths, C., Embedded Structural Steel Connections, PCI Journal, Precast/Prestressed
Remmetter, M., Qin, F., and Shahrooz, B., Seismic Resistance of Coupled Structural
Walls, Cincinnati Infrastructure Institute, University of Cincinnati, College of
Engineering, Report No. UC-CII 92/01, Cincinnati, Ohio, October 1992, 174 pp.
Restrepo, J., Self-Centering Precast Post-Tensioned Cantilever Walls Theory and
Experimental Work, Structures Congress, American Society of Civil Engineers, Seattle,
May 29-31, 2003.
Ricles, J., Sause, R., Garlock, M., and Zhao, C., Posttensioned Seismic-Resistant
Connections for Steel Frames, Journal of Structural Engineering, American Society of
Civil Engineers, Vol. 127, No. 2, February, 2001, pp. 113-121.
Ricles, J., Sause, R., Peng, S., and Lu, L., Experimental Evaluation of Earthquake
Resistant Posttensioned Steel Connections, Journal of Structural Engineering, American
Society of Civil Engineers, Vol. 128, No.7, 2002, pp. 850-859.
Saatcioglu, M., Derecho, A., and Corley, W., Parametric Study of Earthquake-Resistant
Coupled Walls, Journal of Structural Engineering, American Society of Civil Engineers,
Vol. 113, No. 1, January 1987, pp. 141-157.
Saiidi, M. and Sozen, M., Simple Nonlinear Seismic Analysis of R/C Structures,
Journal of the Structural Engineering Division, American Society of Civil Engineers,
Vol. 107, No. ST5, 1981, pp. 937-951.
Sarraf, M. and Bruneau, M., Cyclic Testing of Existing and Retrofitted Riveted
Stiffened Seat Angle Connections, Journal of Structural Engineering, American Society
of Civil Engineers, Vol. 122, No. 7, July, 1996, pp. 762-775.
SEAOC, Recommended Lateral Force Requirements and Commentary, Sixth Edition,
Structural Engineers Association of California, Seismology Committee, Sacramento, CA,
1996.
Shahrooz, B., Remmetter, M., and Qin, F., Seismic Design and Performance of
Composite Coupled Walls, Journal of Structural Engineering, American Society of
Civil Engineers, Vol. 119, No. 11, November 1993a, pp. 3291-3309.
594
Shahrooz, B., Remmetter, M., and Qin, F., Seismic Response of Composite Coupled
Walls, in Composite Construction in Steel and Concrete II, edited by W. Easterling
and W. Roddis, American Society of Civil Engineers, 1993b, pp. 429-441.
Shahrooz, B., Fortney, P., and Rassati, G., Seismic Performance of Hybrid Corewall
Buildings, International Workshop on Steel and Concrete Composite Construction,
National Center for Research on Earthquake Engineering (NCREE) of Taiwan, Taipei,
Oct. 2003, pp. 79-88.
Shen, J. and Astaneh-Asl, A., Hysteretic Behavior of Bolted Angle Connections,
Journal of Construction Steel Research, Vol. 51, No. 3, July, 1999, pp. 201-218.
Shen, J. and Astaneh-Asl, A., Hysteretic Model of Bolted Angle Connections, Journal
of Construction Steel Research, Vol. 54, No. 3, April, 2000, pp. 317-343.
Shome, N., Cornell, C., Bazzurro, P., and Carballo, J., Earthquakes, Records, and
Nonlinear Responses, Earthquake Spectra, Vol. 14, No. 3, 1998, pp. 469-500.
Sims, J., Flange Angle Behavior in Semi-Rigid Connections for Steel PR Frames, M.S.
Thesis, Department of Civil Engineering and Geological Sciences, University of Notre
Dame, IN, April 2000, 187 pp.
Srichatrapimuk, T., Earthquake Responses of Coupled Shear Wall Buildings,
Earthquake Engineering Research Center, Report No. UCB/EERC-76/27, 1976, 117 pp.
Stanton, J. and Nakaki, S., Design Guidelines for Precast Concrete Seismic Structural
Systems, PRESSS Report No. 01/03-09, UW Report No. SM 02-02, Department of
Civil Engineering, University of Washington, Seattle, WA, February 2002.
Stanton, J., Stone, W., and Cheok, G., A Hybrid Reinforced Precast Frame for Seismic
Regions, PCI Journal, Precast/Prestressed Concrete Institute, Vol. 42, No. 2, MarchApril 1997, pp. 20-32.
Stone, W., Cheok, G., and Stanton, J., Performance of Hybrid Moment-Resisting Precast
Beam-Column Concrete Connections Subjected to Cyclic Loading, ACI Structural
Journal, American Concrete Institute, Vol. 91, No. 2, March-April 1995, pp. 229-249.
Subedi, N., RC-Coupled Shear Wall Structures. I: Analysis of Coupling Beams,
March 1991a, pp. 667-680.
Subedi, N., RC-Coupled Shear Wall Structures. II: Ultimate Strength Calculations,
March 1991b, pp. 681-698.
595
Sugaya, K., Teshigawara, M., Kato, M. and Matsushima, Y., Experimental Study on
Carrying Shear Force Ratio of 12-Story Coupled Shear Wall, 12th World Conference on
Earthquake Engineering, paper 2152, 2000.
Swanson, J. and Leon, R., SAC Steel Project, http://www.ce.gatech.edu/~sac/, School
of Civil and Environmental Engineering, Georgia Institute of Technology, GA, 1999.
Tassios, T., Moretti, M., and Bezas A., On the Behavior and Ductility of Reinforced
Concrete Coupling Beams of Shear Walls, ACI Structural Journal, American Concrete
Institute, Vol. 93, No. 6, November-December 1996, pp. 711-720.
Teshigawara, M., Sugaya, K., Kato, M., and Matsushima, Y., Seismic Test on 12-Story
Coupled Shear Wall with Flange Walls, reprinted in List of Technical Papers Written in
English Authored by Japan-Side Researchers, U.S.-Japan Cooperative Earthquake
Research Program on Composite and Hybrid Structures, December 2000.
University of Michigan, Recommendations for U.S. Japan Cooperative Earthquake
Research Program Phase 5 - Composite and Hybrid Structures, Report No. UMCEE 9229, Department of Civil and Environmental Engineering, University of Michigan, Ann
Arbor, MI, 48109, September 10-12, 1992.
USGS, USGS National Seismic Hazard Mapping Project, U.S. Geological Survey,
WWW Document, http://eqhazmaps.usgs.gov/, 1996.
596
STRUCTURAL ENGINEERING RESEARCH REPORT SERIES

LIST OF TECHNICAL REPORTS
NDSE-01-01 Design of Rectangular Openings in Unbonded Post-Tensioned Precast
Concrete Walls, by M. Allen and Y. C. Kurama, April 2001, 142 pp. (this
report may be downloaded from http://www.nd.edu/~concrete/).
NDSE-01-02 Capacity-Demand Index Relationships for Performance-Based Seismic
Design, by K.T. Farrow and Y. C. Kurama, November 2001, 260 pp. (this
report may be downloaded from http://www.nd.edu/~concrete/).
NDSE-06-01 Experimental Evaluation of Unbonded Post-Tensioned Hybrid Coupled
Wall Subassemblages, by M. A. May, Y. C. Kurama, and Q. Shen, April
2006, 212 pp. (this report may be downloaded from
http://www.nd.edu/~concrete/).
NDSE-06-02 Seismic Analysis, Behavior, and Design of Unbonded Post-Tensioned
Hybrid Coupled Wall Structures, by Q. Shen, Y. C. Kurama, and B. D.
Weldon, December 2006, 596 pp. (this report may be downloaded from
http://www.nd.edu/~concrete/).

Us Japan Hybrid14

Transféré par

Informations du document

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

Us Japan Hybrid14

Transféré par

Droits d'auteur :

Formats disponibles

Seismic Analysis, Behavior,

and Design of Unbonded

rigid elements for PT anchorages

rigid elements for

coupling shear force, Vb (kN)

Structural Engineering Research Report

This page intentionally left blank.

Seismic Analysis, Behavior,

Structural Engineering Research Report

This page intentionally left blank.

This report may be downloaded from http://www.nd.edu/~concrete/.

This page intentionally left blank.

2.6.4 Garlock et al. (2003) ...................................................................................45

6.1.2 Coupling Beams ........................................................................................107

8.3.3 Coupled Wall Ultimate State ....................................................................189

11.2.2 Life Safety Performance Level ...............................................................277

12.2.2 Allowable Inter-Story Drift Demands ....................................................325

13.2 Behavior Under Monotonic Loading ...............................................................380

15.2.5 Wall Base Shear Forces ..........................................................................468

Fig. 9.3 Bracing frame side view .....................................................................................202

Fig. 9.25 Post-tensioning strand stress-strain behavior ...................................................222

Fig. 12.18 Prestress losses in wall pier post-tensioning bars...........................................372

Table 6.6 Wall material properties...................................................................................109

Table 12.16 Top and seat angles......................................................................................346

= regression coefficient for R--T relationship

= total compression force at end of coupling beam

= modulus of elasticity of steel

= stress in wall pier post-tensioning bar layer 1

= coupled wall base shear slip capacity

= length of coupling beam flange cover plate

= total coupled wall base moment

= total yield strength of coupling beam post-tensioning tendons

= compression-tension steel fiber

= elongation of beam post-tensioning tendons at coupling beam PT-yielding

= acceleration of equivalent single-degree-of-freedom system

= width of angle-to-column/wall pier connector head/nut measured across flat

= roof drift of coupled wall at coupled wall softening state

= strain in wall pier post-tensioning bar layer 8

= flexural mild steel reinforcement ratio at wall pier base

This page intentionally left blank.

dimensions is affected by the wall reinforcement in the embedment regions, often

Fig. 1.1 Eight story coupled

Fig. 1.2 Hybrid coupled wall subassemblage with

This dissertation provides a comprehensive study on the seismic behavior and

Proposed Hybrid Coupled Wall System

As an example, Fig. 1.3a shows a subassemblage of the proposed hybrid coupled

connection concrete wall

The broad objective of the research described in this dissertation is to investigate

system/diaphragm, columns, foundation-soil system) are not included in the analytical

(1) Conduct a literature survey of previous analytical and experimental research on

This research, which was conducted as a part of the U.S.-Japan Cooperative

Chapter 11 proposes a performance-based approach for the seismic design of

This page intentionally left blank.

Coupled Wall Structural Systems

In concrete structural walls, a regular pattern of openings is often required to

Fig. 2.1 Coupled structural walls

Concrete Coupled Wall Structures

In the last three to four decades, an extensive amount of experimental and

Behavior and Design of Concrete Coupling Beams

Analysis and Modeling of Concrete Coupled Wall Structures

Hybrid Structures with Embedded Steel Coupling Beams

As described above, ductility is an important issue to be considered in the seismic

Behavior and Design of Embedded Steel Coupling Beam Subassemblages

Embedded Steel Coupling Beams with Concrete Encasement