Académique Documents
Professionnel Documents
Culture Documents
Paolo Morandi
University of Pavia and EUCENTRE, Italy
Sanja Hak
University of Zagreb and EUCENTRE, Italy
Guido Magenes
University of Pavia, EUCENTRE and IUSS Pavia, Italy
Paolo Morandi
University of Pavia and EUCENTRE, Italy
Sanja Hak
University of Zagreb and EUCENTRE, Italy
Guido Magenes
University of Pavia, EUCENTRE and IUSS Pavia, Italy
October, 2017
Pavia, Italy
Nessuna parte di questa pubblicazione può essere riprodotta o trasmessa in qualsiasi forma o con qualsiasi
mezzo elettronico, meccanico o altro senza l’autorizzazione scritta dei proprietari dei diritti e dell’editore.
No parts of this publication may be copied or transmitted in any shape or form, and by any type of electronic,
mechanical or different means, without the prior written permission of the copyright holder and the publisher.
EUCENTRE Foundation
Via Adolfo Ferrata, 1 - 27100 Pavia, Italy
Phone (+39) 0382.5169811 - Fax (+39) 0382.529131
E-mail: info@eucentrepress.it - Web: www.eucentrepress.it
ISBN: 978-88-85701-03-8
ABSTRACT
In Italy and other European countries the use of strong clay block masonry infills in RC structures
as external enclosures is progressively accepted, providing besides well-known advantageous
architectural properties and sustainability assets of masonry as a building material, additional
efficiency of sound and thermal insulation, but pointing also to possible advantages for the
achievement of a satisfactory seismic response, due to increased strength and deformation
capacity.
Despite extensive experimental and numerical investigations in the past related to masonry infilled
RC structures, the seismic response of commonly adopted strong infill typologies, regarding the in-
plane deformation capacity, the attainment of performance criteria, typical damage and/or failure
mechanisms, and particularly the behaviour under combined in-plane and out-of-plane actions, has
not yet received sufficient attention. Furthermore, observations from field surveys after major
seismic events repeatedly point towards a high seismic vulnerability of masonry infills, not only in
existing buildings designed for gravity loads, but also in recently constructed RC structures verified
according to contemporary seismic code regulations. Hence, motivated by the need to improve
further existing design procedures for the verification of RC structures with masonry infills and to
provide the necessary data for infill typologies which are widely used in seismic regions, the
present study is focused on three major research objectives: (i) to increase the understanding of
the seismic behaviour of strong clay block masonry infills based on experimental test results, (ii) to
assess the response of RC structures with strong clay block masonry infills in function of different
design conditions based on nonlinear analyses of prototype buildings, and consequently (iii) to
propose a more effective approach to the design of new RC buildings, ensuring adequate damage
control for masonry infills.
The first objective of this work is achieved through the analysis of results obtained during a
systematic experimental campaign, carried out at the Department of Civil Engineering and
Architecture of the University of Pavia and at EUCENTRE of Pavia. The experimental
investigations are based on results of cyclic in-plane and out-of-plane static tests on full-scale,
single-storey masonry infilled RC frame specimens, including the case of full infill and infill with a
central opening. Subsequently, regarding the second and the third objective of the study, extensive
numerical parametric analyses of prototype RC building configurations with strong infills are carried
out for different design conditions and infill layouts. Finally, the effectiveness of related seismic
design procedures commonly adopted in Europe is studied in the light of obtained experimental
and numerical results and possible improvements are proposed. Attached to this report, a proposal
of guidelines for the seismic design of masonry infills in RC structures with calculation examples
(also translated in Italian) is reported.
i
ACKNOWLEDGEMENTS
This work has been carried out at EUCENTRE and at the University of Pavia. The research has
been funded by ANDIL Assolaterizi and by the Project DPC-RELUIS 2013-2016. The received
financial support is gratefully acknowledged.
Special thanks go to Mr Milad Oliaee for his precious contribution in the execution of non-linear
analyses on the infilled building configurations with strong masonry panels.
Eng. Andrea Rossi, Dr Riccardo Milanesi, eng. Luca Albanesi and eng. Carlo Manzini, who have
contributed at different levels, for the final editing and formatting of the present document, are also
sincerely acknowledged. Moreover, the contribution in the translation to English of the design
guidelines by Dr Milanesi has been appreciated. The initial effort by Luca Albanesi in performing
the seismic design and the non-linear analyses on the bare and fully infilled RC frames is resulted
of paramount importance for this research.
ii
TABLE OF CONTENTS
ABSTRACT ....................................................................................................................................... I
ACKNOWLEDGEMENTS ................................................................................................................ II
1 INTRODUCTION ......................................................................................................................... 1
1.1 MOTIVATION AND OBJECTIVES ............................................................................................ 1
1.2 SCOPE OF RESEARCH ......................................................................................................... 3
1.3 REPORT OUTLINE ............................................................................................................... 3
iv
8.4.1 SIMPLIFIED SEISMIC ANALYSIS OF NON-STRUCTURAL ELEMENTS.......................... 143
8.4.2 OUT-OF-PLANE RESISTANCE VERIFICATION OF INFILLS DAMAGED IN-PLANE .......... 144
v
LIST OF TABLES
vii
LIST OF FIGURES
Figure 1.1 Example of severe masonry infill damage, Emilia 2012 earthquake in Italy .................... 1
Figure 2.1 (a) In-plane, and (b) Out-of-plane test setup (Angel et al. [1994]) ................................... 6
Figure 2.2 (a) In-plane test setup; (b) Force displacement response (Combescure et al., [1996]) ... 6
Figure 2.3 Cracking patterns of the non-reinforced infill: (a) at 0.20% drift, (b) at 0.30% drift, (b) at
0.40% drift, (b) after out-of-plane test (Calvi & Bolognini [1999]) .............................................. 6
Figure 2.4 RC frame detailing: (a) Common practise; (b) New proposal (Crisafulli [1997]) .............. 7
Figure 2.5 (a) In-plane test setup; (b) Fully and partially infilled specimens (da Porto et al. [2012]) 7
Figure 2.6 (a) Layout of the 4-storey frame test specimen; (b) Uniformly infilled frame; (c)
Maximum inter-storey drift profiles (Negro & Verzeletti [1996]) ................................................. 8
Figure 2.7 (a) Layout of two-storey frame specimen; (b) Free column top displacements of infilled
structure: analysis vs. estimated test results (Fardis et al. [1999]) ............................................ 9
Figure 2.8 Current design procedure for masonry infilled RC structures according to Eurocode 8 11
Figure 3.1 Selected strong masonry unit ........................................................................................ 17
Figure 3.2 (a) Preparation of prism-shaped mortar test specimens, (b) Mortar testing in flexure, (c)
Mortar testing in compression .................................................................................................. 17
Figure 3.3. Masonry mortar joints: (a) Bed joint - plan view & during construction, (b) Mortar joints
adjacent to RC members - front view & during construction .................................................... 18
Figure 3.4. (a) Front view of bare frame specimen; (b) Cross section of infilled frame specimen .. 19
Figure 3.5. Infilled frame specimen configurations: (a) Fully infilled; (b) Partially infilled; (c) Vertical
infill strip ................................................................................................................................... 19
Figure 3.6. (a) 4-storey prototype frame configuration for design [m]; (b) Single-storey single-bay
substructure; (c) Adopted design acceleration spectra according to NTC08 [2008] ................ 20
Figure 3.7. Reinforcement details: (a) Frame and foundation front view; (b) Foundation plan view21
Figure 3.8. Construction of test specimens – RC frame: (a) Positioning of foundation and column
reinforcement; (b) Casting of foundation; (c) Casting of columns; (d) Positioning of beam
reinforcement (beam-column joint); (e) Casting of beams; (f) Finalised bare frames .............. 22
Figure 3.9. Construction of test specimens – Masonry infill: (a) Infill construction; (b) Finalised
infilled frames ........................................................................................................................... 22
Figure 3.10. Horizontal load introduction during quasi-static cyclic tests ........................................ 23
Figure 3.11. Layout of the in-plane setup for quasi-static cyclic tests ............................................. 24
Figure 3.12. Layout of the out-of-plane setup for quasi-static cyclic tests: (a) Plan view; (b) Side
view.......................................................................................................................................... 25
Figure 3.13. Front view of system for out-of-plane load transfer from actuator to masonry infill
(TA1, TA2 & TA3) .................................................................................................................... 26
Figure 3.14. Front view of system for out-of-plane load transfer from actuator to masonry infill
(TA4) ........................................................................................................................................ 26
Figure 3.15. Front view of system for out-of-plane load transfer from actuator to masonry infill
(TA5) ........................................................................................................................................ 27
viii
Figure 3.16. Plan view of out-of-plane load transfer system ........................................................... 27
Figure 3.17. Masonry characterisation – details and instrumented specimen: Vertical compression
................................................................................................................................................. 29
Figure 3.18. Masonry characterisation test results: Vertical compression ...................................... 30
Figure 3.19. Damaged masonry wallettes after vertical compression tests .................................... 30
Figure 3.20. Masonry characterisation – details and instrumented specimen: Lateral compression
................................................................................................................................................. 30
Figure 3.21. Masonry characterisation test results: Lateral compression ....................................... 31
Figure 3.22. Damaged masonry wallettes after lateral compression tests...................................... 31
Figure 3.23. Masonry characterisation – details and instrumented specimen: shear strength ....... 32
Figure 3.24. Characterisation test results: Initial shear strength ..................................................... 32
Figure 3.25. Damaged masonry triplets after shear strength tests ................................................. 33
Figure 4.1. Bare frame specimen instrumentation – in-plane: (a) Front view; (b) Back view .......... 38
Figure 4.2. Infilled frame specimen instrumentation (TA1_IP, TA2_IP & TA3_IP): (a) Front view; (b)
Back view................................................................................................................................. 39
Figure 4.3. Infilled frame specimen instrumentation (TA4_IP): (a) Front view; (b) Back view ........ 40
Figure 4.4. Layout of strain gauges installed on the RC frame reinforcement (TNT, TA1, TA2) .... 41
Figure 4.5. Drift/displacement history - Bare frame TNT ............................................................... 41
Figure 4.6. In-plane test results - Bare frame TNT ........................................................................ 42
Figure 4.7. Observed crack pattern - Bare frame TNT, 3.50% drift (19D) ...................................... 42
Figure 4.8. Bare frame TNT during the in-plane test at 19D loading amplitude (3.50% drift) ......... 43
Figure 4.9. Details of plastic hinge regions at 19D (3.50% drift): (a) RC beam; (b) RC column ..... 43
Figure 4.10. Drift/displacement history - Infilled frame TA1_IP ...................................................... 44
Figure 4.11. In-plane test results - Infilled frame TA1_IP ............................................................... 44
Figure 4.12. Observed crack pattern - Infilled frame TA1: (a) 0.50% drift (09D); (b) 1.50% drift
(14D) ........................................................................................................................................ 45
Figure 4.13. Infilled frame TA1 during the in-plane test after 14D loading (1.50% drift) ................. 46
Figure 4.14. Details of in-plane damaged TA1 infill corners (back side) after 14D loading (1.50%
drift): (a) Upper left; (b) Upper right ......................................................................................... 46
Figure 4.15. Drift/displacement history - Infilled frame TA2_IP ...................................................... 47
Figure 4.16. In-plane test results - Infilled frame TA2_IP ................................................................ 47
Figure 4.17. Observed crack pattern - Infilled frame TA2: (a) 1.00% drift (12D); (b) 2.50% drift
(17D) ........................................................................................................................................ 49
Figure 4.18. Infilled frame TA2 during the in-plane test at 17D loading amplitude (2.50% drift) ..... 49
Figure 4.19. Drift/displacement history - Bare frame TA3_IP ........................................................ 50
Figure 4.20. In-plane test results - Infilled frame TA3_IP ................................................................ 50
Figure 4.21. Observed crack pattern - Infilled frame TA3: (a) 0.50% drift (09D); (b) 1.00% drift
(12D) ........................................................................................................................................ 51
Figure 4.22. Details of damaged TA3 infill corners (back side): (a) Upper left; (b) Upper right ...... 51
Figure 4.23. Infilled frame TA3 during the in-plane test after 12D loading (1.00% drift) ................. 52
ix
Figure 4.24. Drift/displacement history - Infilled frame TA4_IP ...................................................... 52
Figure 4.25. In-plane test results - Infilled frame TA4_IP ................................................................ 53
Figure 4.26. Observed crack pattern - Infilled frame TA4 (a) 0.50% drift (09D); (b) 1.00% drift (12D)
................................................................................................................................................. 54
Figure 4.27. Infilled frame TA4 during the in-plane test after 12D loading (1.00% drift) ................. 55
Figure 4.28. Details of in-plane damaged TA4 infill corners after 12D loading (1.00% drift) at the
bottom (next to opening): (a) Left panel; (b) Right panel ......................................................... 55
Figure 4.29. Infilled frame specimen instrumentation (TA1_OP, TA2_OP & TA3_OP): (a) Front
view; (b) Back view .................................................................................................................. 59
Figure 4.30. Infilled frame specimen instrumentation (TA4_OP): (a) Front view; (b) Back view..... 60
Figure 4.31. Infilled frame specimen instrumentation (TA5_OP): (a) Front view; (b) Back view..... 61
Figure 4.32. Drift/displacement history - Infilled frame TA1_OP .................................................... 62
Figure 4.33. Out-of-plane test results - Infilled frame TA1_OP ...................................................... 63
Figure 4.34. Infill TA1 during the out-of-plane test at 09D loading amplitude (75 mm displacement)
................................................................................................................................................. 64
Figure 4.35. Details of in-plane & out-of-plane damaged infill TA1 after 09D loading (75 mm
displacement): (a) Diagonal infill crack; (b) Column-infill interface .......................................... 64
Figure 4.36. Crack pattern (in black – out-of-plane damage; in grey – previous in-plane damage) of
infill TA1 after the out-of-plane test .......................................................................................... 65
Figure 4.37. Drift/displacement history - Infilled frame TA2_OP .................................................... 65
Figure 4.38. Out-of-plane test results - Infilled frame TA2_OP ...................................................... 66
Figure 4.39. Infill TA2 during the out-of-plane test at 09D loading amplitude (75 mm displacement)
................................................................................................................................................. 67
Figure 4.40. Deformed shape of infill TA2 during the out-of-plane test at 09D loading amplitude (75
mm displacement): (a) Front view; (b) Back view .................................................................... 67
Figure 4.41. Crack pattern (in black – out-of-plane damage; in grey – previous in-plane damage) of
infill TA2 after the out-of-plane test .......................................................................................... 68
Figure 4.42. Drift/displacement history - Infilled frame TA3_OP .................................................... 69
Figure 4.43. Out-of-plane test results - Infilled frame TA3_OP ...................................................... 69
Figure 4.44. Infill TA3 during the out-of-plane test at 09D loading amplitude (75 mm displacement)
................................................................................................................................................. 70
Figure 4.45. Details of in-plane & out-of-plane damaged infill TA3 after 09D loading (75 mm
displacement): (a) Diagonal infill crack; (b) Local damage at force introduction...................... 70
Figure 4.46. Crack pattern (in black – out-of-plane damage; in grey – previous in-plane damage) of
infill TA3 after the out-of-plane test .......................................................................................... 71
Figure 4.47. Drift/displacement history - Infilled frame TA4_OP .................................................... 71
Figure 4.48. Out-of-plane test results - Infilled frame TA4_OP, LEFT panel ................................. 72
Figure 4.49. Out-of-plane test results - Infilled frame TA4_OP, RIGHT panel ............................... 72
Figure 4.50. Infill TA4 during the out-of-plane test after 09D loading (40 mm actuator displacement)
................................................................................................................................................. 73
Figure 4.51. Deformed shape of infill TA4 during the out-of-plane test at 09D loading amplitude (40
mm actuator displacement): (a) Front view; (b) Back view ...................................................... 73
Figure 4.52. Drift/displacement history - Infilled frame TA5_OP .................................................... 74
x
Figure 4.53. Out-of-plane test results - Infilled frame TA5_OP ...................................................... 74
Figure 4.54. Infill TA5 during the out-of-plane test at 09D loading amplitude (75 mm displacement):
(a) Deformed shape; (b) Region of hinging at the centre; (c) Region of support at the bottom 75
Figure 5.1. Average experimental response: (a) Strong infill without and with opening; (b) Strong
and slender infill without opening ............................................................................................. 77
Figure 5.2. Simplified masonry infill response: (a) Strong infill without and with opening; (b) Strong
and slender infill without opening ............................................................................................. 78
Figure 5.3. Crisafulli [1997] stress-strain model: (a) Monotonic response; (b) Cyclic response ..... 80
Figure 5.4 (a) Degradation of strut cross section area; (b) Axial force – strain relationship (Oilaee
et al. [2015a]) ........................................................................................................................... 81
Figure 5.5 Comparison of numerical and experimental response: Bare frame ............................... 81
Figure 5.6 Comparison of numerical and experimental cyclic response: Strong infill (TA2, Oliaee et
al. [2015a]) ............................................................................................................................... 82
Figure 5.7 Comparison of numerical and experimental cyclic response: Strong infill with opening
(TA4, Oliaee et al. [2015a]) ...................................................................................................... 82
Figure 5.8 Comparison of numerical and experimental strength envelope: Strong infill (TA2) and
strong infill with opening (TA4; Oliaee et al. [2015a]) .............................................................. 83
Figure 5.9 Comparison of numerical and experimental response: Infilled frame - Strong infill (TA2,
Oliaee et al. [2015a]) ............................................................................................................... 84
Figure 5.10 Comparison of numerical and experimental response: Infilled frame - Strong infill with
opening (TA4, Oliaee et al. [2015a]) ........................................................................................ 84
Figure 5.11 Comparison of numerical and experimental strength envelope: Infilled frame - Strong
infill (TA2) and strong infill with opening (TA4, Oliaee et al. [2015a]) ...................................... 85
Figure 5.12 Shear crack development at increasing drift levels: (a) 0.30%; (b) 1.0%; (c) 1.5%; (d)
2.5%......................................................................................................................................... 87
Figure 5.13. Deflection of fully infilled specimens at 50 mm target displacement: (a) TA2, (b) TA1,
(c) TA3 and at 75 mm target displacement: (d) TA2, (e) TA1, (f) TA3 ..................................... 87
Figure 5.14. Force-displ. response: (a) TA1, (b) TA2, (c) TA3, (d) Comparison of TA1, TA2 and
TA3 .......................................................................................................................................... 88
Figure 5.15. Specimen TA5: (a) Force-displacement response; (b) Deflection along the height.... 89
Figure 5.16. Specimen TA5: (a) Progression of damage; (b) Damage at 75 mm target
displacement ............................................................................................................................ 89
Figure 5.17. (a) Assumption of uniformly distributed out-of-plane action; (b) Formation of out-of-
plane arching mechanism ........................................................................................................ 90
Figure 5.18. (a), (b) Application of the loads in the single bending out-of-plane test, and c) Forces
acting in the upper part of the panel ........................................................................................ 91
Figure 5.19. Moment-mid-height displacement response of the panel subjected to single bending
................................................................................................................................................. 91
Figure 5.20. Out-of-plane strength in function of in-plane drift: (a) Strong infill; (b) Strong infill with
opening .................................................................................................................................... 92
Figure 5.21. Experimental strength reduction coefficient: (a) Strong infill; (b) Strong infill with
opening .................................................................................................................................... 93
Figure 6.1. Plan and elevation view of 3-story, 6-story and 9-story frame systems ........................ 95
Figure 6.2. (a) Partially infilled (P), and (b) Fully infilled (F) 2-D 3-story frame configurations........ 96
xi
Figure 6.3. (a) Bare central bays & infills with opening (BO), and (b) Infilled central bays & infills
with opening (PO) 2-D 3-story frame configurations ................................................................ 96
Figure 6.4. Normalised elastic acceleration and displacement response spectra for specific sites in
Italy: (a) at the DLS; (b) at the ULS ......................................................................................... 98
Figure 6.5. Ratio of spectral ordinates at the DLS and the ULS for specific sites in Italy: (a) on
ground type A; (b) on ground type B ........................................................................................ 99
Figure 6.6. Rigid beam and column ends ..................................................................................... 101
Figure 6.7. (a) Concentrated-plasticity frame member; (b) Modified Takeda hysteretic rule; (c)
Moment-curvature bilinear idealisation .................................................................................. 102
Figure 6.8. (a) Equivalent diagonal single-strut model; (b) Strut element in frame model ............ 103
Figure 6.9. (a) Compression at the centre of the infill; (b) Compression at the corners of the infill;
(c) Sliding shear failure; (d) Diagonal tension failure ............................................................. 104
Figure 6.10. Code compatible earthquake acceleration and displacement spectra at the DLS.... 108
Figure 6.11. Code compatible earthquake acceleration and displacement spectra at the ULS.... 109
Figure 6.12. Design spectrum compatible earthquake records at the DLS .................................. 110
Figure 6.13. Design spectrum compatible earthquake records at the ULS .................................. 111
Figure 7.1. Displacement and drift profiles at ULS: 0.35gS DCM frames - Bare B ....................... 115
Figure 7.2. Displacement and drift profiles at ULS: 0.35gS DCM frames - Infill P (Oliaee et al.
[2015b]).................................................................................................................................. 115
Figure 7.3. Displacement and drift profiles at ULS: 0.35gS DCM frames - Infill BO (Oliaee et al.
[2015b]).................................................................................................................................. 116
Figure 7.4. Displacement and drift profiles at ULS: 0.35gS DCM frames - Infill PO (Oliaee et al.
[2015b]).................................................................................................................................. 116
Figure 7.5. Displacement and drift profiles at ULS: 0.35gS DCM frames - Infill F (Oliaee et al.
[2015b]).................................................................................................................................. 117
Figure 7.6. Average maximum inter-storey drifts of 3-storey frames – Bare (B) and infilled frames
(P, BO, PO, F & T1): (a) DLS; (b) ULS (Oliaee et al. [2015b]) ............................................. 118
Figure 7.7. Average maximum inter-storey drifts of 6-storey frames – Bare (B) and infilled frames
(P, BO, PO, F, T1): (a) DLS; (b) ULS (Oliaee et al. [2015b]) ................................................ 118
Figure 7.8. Average maximum inter-storey drifts of 9-storey frames – Bare (B) and infilled frames
(P, BO, PO, F & T1): (a) DLS; (b) ULS (Oliaee et al. [2015b]) ............................................. 119
Figure 7.9. Average damage distribution for RC frames in DCM at the DLS: (a) 0.25gS; (b) 0.35gS
(Oliaee et al. [2015b]) ............................................................................................................ 120
Figure 7.10. Average damage distribution for RC frames in DCM at the ULS: (a) 0.25gS; (b)
0.35gS (Oliaee et al. [2015b]) ................................................................................................ 121
Figure 7.11. Average damage distribution for RC frames in DCH at the DLS: (a) 0.25gS; (b)
0.35gS (Oliaee et al. [2015b]) ................................................................................................ 121
Figure 7.12. Average damage distribution for RC frames in DCH at the ULS: (a) 0.25gS; (b)
0.35gS (Oliaee et al. [2015b]) ................................................................................................ 121
Figure 7.13. (a) Masonry infill secant stiffness; (b) Estimated infill strength reduction due to
opening .................................................................................................................................. 125
Figure 7.14. Average drift demands, 3-storey, 6-storey & 9-storey frames (P): (a) DLS; (b) ULS 128
Figure 7.15. Average drift demands, 3-storey, 6-storey & 9-storey frames (BO): (a) DLS; (b) ULS
............................................................................................................................................... 128
xii
Figure 7.16. Average drift demands, 3-storey, 6-storey & 9-storey frames (PO): (a) DLS; (b) ULS
............................................................................................................................................... 128
Figure 7.17. Average drift demands, 3-storey, 6-storey & 9-storey frames (F): (a) DLS; (b) ULS 128
Figure 7.18. Force reduction and displacement ductility: (a) Equal displacement rule; (b) Equal
energy rule ............................................................................................................................. 132
Figure 7.19. (a) Definition of bilinear curve for Cj = 1.0; (b) Variation of bilinear curve for different
values of Cj ............................................................................................................................ 133
Figure 7.20. Average and predicted drift demands, 3-storey DCH frame (P, BO, PO & F), j = 2: (a)
DLS; (b) ULS ......................................................................................................................... 134
Figure 7.21. Average and predicted drift demands, 6-storey DCH frame (P, BO, PO & F), j = 2: (a)
DLS; (b) ULS ......................................................................................................................... 134
Figure 7.22. Average and predicted drift demands, 6-storey DCH frame (P, BO, PO & F), j = 5: (a)
DLS; (b) ULS ......................................................................................................................... 135
Figure 7.23. Average and predicted drift demands, 9-storey DCH frame (P, BO, PO & F), j = 3: (a)
DLS; (b) ULS ......................................................................................................................... 135
Figure 7.24. Average and predicted drift demands, 9-storey DCH frame (P, BO, PO & F), j = 5: (a)
DLS; (b) ULS ......................................................................................................................... 135
Figure 7.25. Average and predicted drift demands, 9-storey DCH frame (P, BO, PO & F), j = 7, 8,
9: (a) DLS; (b) ULS ................................................................................................................ 136
Figure 8.1. Possible improvements in the current European design approach ............................ 137
Figure 8.2. Critical length of columns: (a) Ground floor; (b) Full contact with the infill on one side
............................................................................................................................................... 139
Figure 8.3. Partial contact of column and infill: (a) Critical length lcr and (b) Decreased column
shear span ratio for clear length lcl; (c) Flexural column capacity at ends of contact length lc 139
Figure 8.4. Strut force activation in function of average inter-storey drift: (a) Strong infill; (b)
Slender/weak infill .................................................................................................................. 142
Figure 8.5. (a) Position of a masonry infill in the RC structure; (b) Normalised seismic coefficient
applicable to non-structural elements (Eurocode 8) .............................................................. 143
Figure 8.6. Simplified out-of-plane resistance reduction coefficient βa,j: (a) Strong infill; (b) Strong
infill with opening ................................................................................................................... 145
xiii
LIST OF SYMBOLS
xiv
Fh,max = Maximum horizontal force for in-plane setup design
fm = Mortar strength
fm' = Equivalent diagonal strut strength
Fo,max = Maximum horizontal force for out-of-plane setup design
fv = Masonry shear strength
Fv = Vertical force acting on one column of the test specimen
fv0 = Masonry initial shear strength under zero compressive stress
= Characteristic masonry initial shear strength under zero compressive
fv0k
stress
Fw,s = Equivalent diagonal strut force
fw = Masonry infill strength
fw,0 = Control masonry infill strength
Fw = Horizontal component of the equivalent diagonal strut force
Fw,act = Horizontal component of the activated equivalent diagonal strut force
fwh = Horizontal masonry compression strength
fws = Shear resistance under diagonal compression
fwu = Sliding shear resistance of the mortar joints
fwv = Vertical masonry compression strength
fy = Reinforcement yield strength
g = Acceleration of gravity
G = Masonry shear modulus
h = Centreline storey height
H = Total height of the building
h0,j = Height of equivalent single-storey single-bay frame for storey j
hb,0,j = Beam height of equivalent single-storey single-bay frame for storey j
hc = Column cross sectional height
hw = Masonry infill height
γI = Importance factor of the building
= Dimensionless parameter for the definition of the equivalent diagonal
K1, K2
strut width
KI,0,j = Control stiffness of masonry infills in storey j
KI,i,j = Tangent stiffness of masonry infill in bay i of storey j
KI,j = Stiffness coefficient for masonry infills in storey j
KS,j = Structural stiffness coefficient for storey j
kw = Equivalent diagonal strut axial stiffness per unit length
lc = Equivalent strut contact length
xv
lcl = Clear column height
lcr = Critical length of the column
Li,j = Clear length of bay i in storey j
lp = Length of masonry infill opening
lt = Total column height
Lw = Masonry infill length
Lw,i,j = Masonry infill length in bay i of storey j
= Simplified moment of resistance due to the vertical arching
MR
mechanism
Mw = Earthquake magnitude
mw = Masonry infill mass per unit height
nc = Number of reverse cycles
Nmax = Maximum beam axial force
nb,j = Number of bays in storey j
np = Number of structural planes
ns = Number of storeys
PR = Probability of exceedance
= Vertical component of the maximum compressive force activated in
Pu
the vertical arching mechanism
q = Behaviour factor of the structure
qa = Behaviour factor of the non-structural element
qd = Displacement global ductility
r = Moment-curvature bi-linear factor
= Fraction of out-of-plane resistance for damage limitation state
ra,1
conditions
ra,2 = Fraction of out-of-plane resistance for ultimate limit state conditions
= Average ratio of inelastic vs. elastic inter-storey displacement
rd
demands
rp = Strength/stiffness reduction factor accounting for openings
S = Soil factor
Sa = Seismic force coefficient of the non-structural element
Sd(T) = Design acceleration response spectrum
Se(T) = Elastic acceleration response spectrum
T = Period of vibration
Ta = Fundamental vibration period of the non-structural element
TC = Acceleration response spectrum corner period
TDLS = Reference return period of the Damage limitation seismic action
xvi
T1 = Initial period of vibration of the building
= Reference return period of the No-collapse (Ultimate limit state)
TULS
seismic action
tw = Masonry infill thickness
tw,0 = Control masonry infill thickness
VC = Column shear force
VC,Ed,cl = Shear demand due to local effects on clear column height
VC,Ed,l = Column shear demand induced by local effects
= Column shear demand induced by local effects considering activated
VC,Ed,l,a
strut force
VC,Ed,M = Shear demand due to local capacity design requirements
= Shear demand due to horizontal component of equivalent diagonal
VC,Ed,w
strut force
= Shear demand due to horizontal component of activated equivalent
VC,Ed,w,a
diagonal strut force
VC,Rd = Column shear resistance
W = Masonry infill unit weight
wa = Seismic force expressed as equivalent pressure acting on the infill
Wa = Weight of the non-structural element
wj = Masonry infill density coefficient
= Out-of-plane resistance of the infill due to vertical arching
wR
mechanism
= Out-of-plane strength due to arching action of damaged masonry
wR,β
infill
= Distance of the centre of mass of the masonry infill from the level of
zw
application of the seismic action
= Ratio of the design ground acceleration on type A ground to the
α
acceleration of gravity
αre = Reloading strain factor
αun = Unloading stiffness coefficient
βa,j = Out-of-plane resistance reduction coefficient for storey j
βre = Reloading stiffness coefficient
γa = Importance factor of the non-structural element
γRd = Overstrength coefficient
γun = Unloading stiffness factor
δ = Inter-storey drift
δDLS = Design drift limit at the damage limitation limit state
δinf,j = Lower bound drift demand for storey j
δj = Drift demand of bare frame structure in storey j
xvii
δm ’ = Drift at masonry infill damage limitation deformation
δm,0’ = Reference drift at masonry infill damage limitation deformation
δm,j’ = Equivalent infill drift capacity for storey j
δmax,j,i = Maximum drift of the j-th storey attained for record i
δsup,j = Upper bound drift demand for storey j
Δt = Integration time step
δu = Drift at masonry infill ultimate deformation capacity
δULS = Design drift limit at the ultimate limit state
ΣVEd = Sum of the seismic shear forces acting on vertical primary seismic
members of the structure
xviii
P. Morandi, S. Hak, G. Magenes EUCENTRE 1
Research Report
1 INTRODUCTION
Figure 1.1 Example of severe masonry infill damage, Emilia 2012 earthquake in Italy
2 Experimental and numerical seismic performance of strong clay masonry infills
Consequently, shifting the principal attention towards the concern about acceptable infill damage,
considering also the influence of simultaneous in-plane and out-of-plane seismic actions, further
important aspects related to the behaviour of masonry infills in RC structures have lately been
recognised. In particular, the infill response with reference to the deformation capacity and the
attainment of performance criteria assigned to different limit states has been studied more
extensively for commonly adopted masonry typologies, such as slender/weak clay masonry infills
with and without light reinforcement (Calvi and Bolognini [2001], Hak et al. [2013a]), as well as
reinforced and unreinforced strong clay block masonry infills (da Porto et al. [2012]). At the same
time, research efforts have increasingly been directed to possible improvements of the
performance in seismic conditions through the introduction of new products and innovative
construction techniques (e.g. Mohammadi et al. [2011], Markulak et al. [2013], Preti et al. [2012,
2014]).
In line with European design practise, multimodal dynamic response spectrum analyses are
commonly adopted in the design of buildings and structural models are usually established
neglecting the contribution of non-structural elements to strength and stiffness characteristics of
the load-bearing structure. Hence, masonry infill panels are considered only in terms of masses
and vertical loads, while analyses and safety verifications are performed on bare frame structural
models. This hypothesis can generally be considered reasonable, in particular due to the strongly
non-linear response of masonry infills and the complexity of related modelling procedures,
especially in the case of infills with openings, but also given the variety of available masonry infill
typologies and construction techniques, as well as the possible variability of infill layouts during the
life of the structure. However, the satisfactory seismic response has to be ensured also for the
infilled configuration and performance requirements defined for damage limitation and life safety
limit states, accounting not only for the structural response, but also the masonry infill behaviour.
Considering the response of different RC frame configurations, previous numerical investigations
(Hak et al. [2012, 2013a]) have shown that, in terms of in-plane damage, the behaviour of
slender/weak unreinforced masonry infill walls in RC frames designed in compliance with present
European seismic code regulation (CEN [2004a]) can result to be unsatisfactory. Regarding the
effective limitation of infill damage, major deficiencies in the current design approach seem to arise
due to the application of a single inter-storey drift limit at the damage limitation limit state, being
equal to 0.50% for the case of rigid masonry infills. Such unique damage control parameter,
commonly applied to the bare structure independently of the structural layout, the infill distribution
and the adopted masonry typology, allows only a rough control of the actual damage and may not
always be able to guarantee the achievement of efficient design choices, ensuring a sufficient level
of safety without overly penalising the RC structure. Specifically, referring to both limit states, the
current design approach may be insufficient in some cases, such as for sparse and/or weak infills
having low displacement capacities. On the other hand, for densely distributed and/or enhanced
infill types or innovative infill typologies with significantly increased levels of displacement capacity,
overly conservative requirements may be imposed. Consequently, with the aim to achieve more
flexible design criteria applicable to different building configurations, a revision of the drift limit at
the damage limitation limit state as well as the introduction of an additional drift limitation
requirement at the ultimate limit state, has been found necessary (Hak et al. [2012]), and possible
improvements in the design approach have been proposed with reference to slender/weak infill
typoogies (Hak et al. [2013a]).
Motivated by the need to improve further existing design procedures for the verification of RC
structures with masonry infills and to provide the necessary data for infill typologies which are
widely used in some seismic regions, but have not been thoroughly investigated in the past, the
principle goal of this work is to explore the experimental seismic performance of strong clay block
masonry infills. Furthermore, this study aims to investigate the influence of the tested strong infill
typology on the response of masonry infilled RC structures in function of different design
parameters as well as to verify the applicability of improved design procedures, which are thougt to
enhance the infill damage control, without overly increasing the dimensions of the RC structural
elements.
P. Morandi, S. Hak, G. Magenes EUCENTRE 3
Research Report
In Chapter 3 the experimental campaign carried out in the laboratories of the Department of Civil
Engineering and Architecture of the University of Pavia and EUCENTRE of Pavia is introduced, the
design of the new experimental setup for in-plane and out-of-plane tests on bare and infilled RC
specimens is presented and the design of the specimens, carried out following European and
Italian code provisions, is summarised. Furthermore, the characterisation tests, carried out in order
to evaluate the mechanical properties of material components, are described and the obtained
results are presented.
In Chapter 4 the adopted test protocol is presented for both in-plane and out-of-plane tests on
single-storey, single-bay bare and infilled RC frame specimens, the experimental results are
summarised in terms of force-displacement response and observations gathered during the tests,
in particular regarding the damage propagation are discussed.
Chapter 5 presents most important issues regarding the interpretation of the obtained experimental
results, addressing the definition of infill performance levels, the calibration of a simple numerical
model for the representation of infills, the possible occurrence of local effects due to the presence
of infills, as well as out-of-plane response mechanisms and the related out-of-plane strength
reduction due to previous in-plane damage.
Chapter 6 introduces the framework of the numerical study of this work. Firstly, a description of the
considered prototype building configurations is provided, referring to 2-dimensional RC frame
structures analysed in a series of extensive parametric studies. In addition, modelling assumptions
adopted in the elastic structural models used to evaluate design action effects in terms of forces
and displacements, as well as in the inelastic models used for nonlinear time-history analyses are
discussed and the earthquake records selected to match the design response spectra are
presented.
In Chapter 7 the in-plane response of the 2-dimensional bare and infilled case study RC frame
systems is evaluated. The performance of the masonry infills in terms of damage is estimated at
both the damage limitation and the ultimate limit state in order to assess to which extent the
current design procedure is able to control the infill damage. Furthermore, drift demands for bare
and infilled structures are compared and a simplified method for the prediction of drift demands
corresponding to infilled frames based on a relative infill vs. structural stiffness coefficient is
established for the case of strong infills.
Chapter 8 deals with possible improvements in the current design approach for new RC structures,
in particular regarding a more effective infill damage control, discussing the limitation of inter-storey
drift demands in function of infill properties and design conditions, based on the correlation of drift-
demands for bare and infilled structures. Moreover, possible verifications of the out-of-plane infill
resistance in function of previous in-plane damage are considered with reference to the obtained
experimental results, and the evaluation of increased column shear demands is discussed.
A summary and the conclusions of the study are given in Chapter 9, followed by recommendations
for future developments.
Finally, attached to this report, a proposal of guidelines for the seismic design of masonry infills in
RC structures with calculation examples (also translated in Italian) is reported.
P. Morandi, S. Hak, G. Magenes EUCENTRE 5
Research Report
2 LITERATURE REVIEW
2.1 Introduction
The current study is related to the behaviour of strong clay block masonry infills and their role in
the seismic response of infilled RC structures, evaluated based on an extensive experimental
campaign as well as the implementation of obtained results for the accomplishment of systematic
numerical analyses for different building configurations. Therefore, the principal investigations
strongly rely on principles commonly adopted in the evaluation of static cyclic tests and the
nonlinear modelling of RC frame structures and masonry infills. Furthermore, the given
considerations are related to seismic design procedures according to European and Italian national
code provisions.
Hence, a number of significant experimental studies based on static cyclic tests, commonly related
to accompanying analytical investigations, will be reviewed. Since for scaled specimens of RC
frames with infills, the difficulty to use adequately scaled masonry elements may induce artificially
weak frames in relation with the infills, possibly leading to unrealistic results (CEB [1996]), in the
case of quasi-static tests primary attention is given to experimental investigations on full scale or
close to full-scale specimens. In particular, regarding the assessment of masonry infill damage and
the definition of corresponding performance levels, which is, for the case of strong infill, one of the
important research interests related to the scope of this work, the response obtained for scaled
specimens may be unconservative. Clearly, the addressed achievements present only a brief
overview of selected previous investigations, which are of fundamental importance for the current
study and/or deal with issues of major interest, such as the evaluation and control of infill damage,
the correlation of in-plane and out-of-plane infill response, the behaviour of similar contemporary
infill typologies etc. In addition, particular attention will be devoted to important investigations
related to the verification and application of the design philosophy for masonry infilled RC frames
adopted in Eurocode 8 – Part 1 (CEN [2004a]) and currently adopted seismic design methods with
reference to the prevention of infill damage will be addressed. Related to modelling issues in
numerical analyses of masonry infilled RC structures, the well-established equivalent diagonal strut
modelling approach will be adopted in this work, based on the findings by Decanini et al. [1993]
and Crisafulli [1997], relying on the newly available experimental results to represent the response
of strong masonry infills as realistically as possible. A comprehensive overview of the state-of-the-
art related to the mathematical macro-modelling of masonry infills can be found e.g. in the work by
Asteris et al. [2011a].
panels can resist, based on arching action for a strip of infill that spans between two fully
restrained supports. Simplified expressions evaluating the out-of-plane strength of panels were
developed, including the influence of previous in-plane damage.
(a) (b)
Figure 2.1 (a) In-plane, and (b) Out-of-plane test setup (Angel et al. [1994])
An extensive experimental program has been accomplished (Combescure et al. [1996]) at the
National Laboratory for Civil Engineering of Portugal (LNEC) on 2:3 scale single-storey single-bay
models, subjected to in-plane cyclic horizontal loading (Figure 2.2a). A comparison of the
experimental results with a numerical prediction performed by means of a plasticity based model
showed that both, the failure pattern and the global characteristics of the structural behaviour had
been captured. The force displacement curve obtained for a RC frame specimen, infilled with a
15.0 cm thick horizontally hollow clay brick masonry infill is shown in Figure 2.2b.
(a) (b)
Figure 2.2 (a) In-plane test setup; (b) Force displacement response (Combescure et al., [1996])
At the University of Pavia in Italy an experimental study related to the in-plane and out-of-plane
behaviour of RC frames with different types of weak clay brick masonry infill, including
unreinforced and weakly reinforced typologies, was carried out (Calvi and Bolognini [1999, 2001]).
Full-scale single-story single-bay frame specimens were designed according to modern seismic
code provisions. Examples of cracking patterns at four different loading stages for a specimen with
unreinforced masonry infill are shown in Figure 2.3. Major conclusion indicated that the presence
of little reinforcement significantly improved the response of a single infilled frame, particularly
concerning the attainment of damage limit states. Furthermore, the fundamental role of the state of
damage for masonry infills in the definition of limit states for RC frames was stressed, since in
general for well designed frames a high damage or a potential for out-of-plane expulsion may
precede any significant damage to the frame.
(a) (b) (c) (d)
Figure 2.3 Cracking patterns of the non-reinforced infill: (a) at 0.20% drift, (b) at 0.30% drift, (b) at 0.40%
drift, (b) after out-of-plane test (Calvi & Bolognini [1999])
A research program, involving the static cyclic in-plane testing of 3:4 scale single-storey single-bay
masonry infilled RC frames, representing the bottom part of two-storey structure, was conducted at
the University of Canterbury, New Zealand (Crisafulli [1997], Figure 2.4). An advanced nonlinear
cyclic model was proposed for the representation of masonry infills. In particular, the analytical
P. Morandi, S. Hak, G. Magenes EUCENTRE 7
Research Report
model was characterised by the possibility to represent the material with different levels of
accuracy according to available data, and to account for the interaction between the masonry infill
and the surrounding frame. The cyclic compressive behaviour of masonry was represented by
several hysteresis rules, considering different response mechanisms for loading, unloading and
reloading, assuming that the envelope curve was independent of the load pattern and loading
history and coincides approximately with the stress-strain curve under monotonic loading. For the
needs of the numerical analyses carried out in this study, the hysteretic model proposed by
Crisafulli [1997] has been calibrated based on the obtained experimental results and used to
represent the cyclic response of the adopted equivalent diagonal strut model; hence, further details
are addresses in Section 5.1.2.
(a) (b)
Figure 2.4 RC frame detailing: (a) Common practise; (b) New proposal (Crisafulli [1997])
At the University of Padova, an experimental research program related to the combined in-plane
and out-of-plane response of contemporary clay masonry infill typologies in RC frames designed
according to current seismic design provisions has been accomplished (da Porto et al. [2012]).
Static cyclic in-plane tests and subsequent monotonic out-of-plane tests have been carried out on
full scale single-storey single-bay frame specimens (Figure 2.5a) considering an unreinforced and
a corresponding reinforced infill typology. In addition to the fully infilled configuration, the presence
of a significant central opening along the full height of the infill has been accounted for, as
illustrated in Figure 2.5b.
(a) (b)
Figure 2.5 (a) In-plane test setup; (b) Fully and partially infilled specimens (da Porto et al. [2012])
the activation of different energy dissipation mechanisms. Based on the test results, further issues
related to irregularities induced by non-structural masonry panels in framed buildings were
discussed in the work by Negro and Colombo [1997]. The irregular distribution of infills was found
to possibly result in unacceptably high ductility demands in the frame. Furthermore, from a
comparison of experimental and numerical results on the uniformly infilled structure, it was
demonstrated that a regular distribution of infills may result in potentially negative effects that may
or may not be counterbalanced by the positive effect of due to increased stiffness, strength and
energy dissipation capacity induced by the presence of infills. Therefore, a safe design procedure
should not neglect the influence of the non-structural masonry infills.
(a) (b) (c)
Figure 2.6 (a) Layout of the 4-storey frame test specimen; (b) Uniformly infilled frame; (c) Maximum inter-
storey drift profiles (Negro & Verzeletti [1996])
Through shaking table tests carried out on the MASTER shaking table of the ISMES establishment
in Bergamo (Franchioni [1999]) and nonlinear dynamic analyses, Fardis et al. [1999] have studied
the bi-directional response of a full-scale two-storey RC frame specimen with infills in both storeys
on two adjacent sides and the corresponding bare frame configurations (Figure 2.7a). The infilled
specimen was meant to represent typical RC structures which are symmetric and torsionally
balanced in both horizontal directions, but may develop lateral-torsional coupling due to a strongly
eccentric layout of infills in plan. Given the wide agreement of that negative effects imposed on RC
structural members due to the presence of infills are often associated with irregularities in the
distribution of infills in plan and/or in elevation, the main interest was to investigate whether and to
what extent the seismic demands on RC elements are larger than in the bare structure under the
same bi-directional ground motion. Specifically, related to the design of new structures for
earthquake resistance according to Eurocode 8 – Part 1 (CEN [2004a]), the code requirement to
include, in the case of strong regularities in plan, a sensitivity study of the effect of the position and
stiffness of the infills is addressed. The fact that masonry infills are traditionally non-engineered is
mentioned as a major problem with a design procedure in which infills are explicitly included in the
model for the analysis. Hence, the relevant characteristics to represent the masonry infills
adequately, especially in the nonlinear range and under cyclic actions, are almost impossible to
estimate reliably at the design stage, and hard to control at the construction stage. Finally, during
the service life of a structure the geometry of masonry infills may be subject to alterations beyond
the control of the designer and of the building code. Consequently, relying on the opinion that the
only prudent approach for new structures is to design the structural system against conservative
bounds for the seismic demands in the presence of infills, relying as little as possible on
quantitative information on the infill properties, the specimen was designed as a bare frame
structure in accordance with the requirements for ductility class medium. The results of the shaking
table tests for the bare and the infilled frame configuration along with nonlinear dynamic analyses
of the response and a series of numerical sensitivity studies, showed that the response of such
torsionally unbalanced structure to a bidirectional ground motion with two independent components
conforming to the same spectrum, was essentially rotational about the common corner of the two
infilled sides, with predominant period about the same as that of the bare structure in the two
horizontal directions. Moreover, the peak displacements of the free column in each horizontal
direction were found to be about the same as, or smaller than, those of the bare structure, but take
place simultaneously (see Figure 2.7b). Implication for the design of RC frame structures with such
infill eccentric in plan were specified, indicating that the columns near the common corner of the
open sides should be proportioned for simultaneous occurrence of the full peak force or
P. Morandi, S. Hak, G. Magenes EUCENTRE 9
Research Report
deformation demands due to both horizontal components of the seismic action. In all other aspects
conclusions pointed to the fact that the RC frame structure could be designed as bare.
(a) (b)
Figure 2.7 (a) Layout of two-storey frame specimen; (b) Free column top displacements of infilled structure:
analysis vs. estimated test results (Fardis et al. [1999])
Summarising the conclusions of several numerical studies, such as Fardis and Panagiotakis
[1997a, 1997b], supported also by experimental evidence (Negro et al. [1996], Fardis et al. [1999])
design implications were given in the work by Fardis [2000], pointing out, amongst other
conclusions, that for regularly infilled structures the overall effect of infills was found to be
beneficial and designing the structure as bare was sufficient. Further discussion and supplemental
recommendations related to the design of RC frames with infills that do not have any positive
structural connection, according to Eurocode 8 – Part 1 (CEN [2004a]), considering non-structural
masonry infills as a source of overstrength and as a second line of defence, were provided by
Fardis [2006]. The mandatory code requirements to be imposed against local or global adverse
effects of infills, without explicitly accounting for the individual infill panels in the model used for the
seismic analysis of the building, were reviewed and clarified, the seismic assessment of existing
buildings with infilled RC frames was addressed and a simple masonry infill modelling approach
was presented.
infills, most frequently referenced in Europe, i.e. regulations adopted in the USA, in Canada and in
New Zealand, can be found in Hak et al. [2013a].
For regular buildings whose response is not significantly affected by contributions from modes of
vibration higher than the fundamental mode in each principal direction, the application of the
equivalent static lateral force method is permitted according to Eurocode 8 – Part 1 (CEN [2004a]).
For the determination of the fundamental period of vibration period T1 of the building, expressions
based on methods of structural dynamics (e.g., the Rayleigh method) or simplified empirical
formulae may be used. For structures up to 40 m height with concrete or masonry shear walls, the
expression given in Equation (2.1) is suggested, where H denotes the total height of the building
and Ct is defined equal to 0.085 for moment resistant space steel frames, 0.075 for moment
resistant space concrete frames and for eccentrically braced steel frames and 0.050 for all other
structures. Alternatively, for structures with concrete or masonry shear walls, the value Ct may be
evaluated based on an estimation of the total effective area of the shear walls in the first storey of
the building (see also Tomažević [1999]).
T1 =C 1H 3/ 4 (2.1)
Indicating a certain lack of clarity related to essential issues in the design of infilled RC structures,
it has to be pointed that different authors, studying simplified methods for the estimation of
fundamental periods of vibrations for infilled RC structures in code provisions available around the
world, approach the interpretation of Eurocode 8 - Part 1 (CEN [2004a]) recommendations in
diverse ways (see e.g., Kaushik et al. [2006], Kose [2009]), identifying a series of difficulties.
Studying relationships for periods of vibration and their employment in linear seismic analysis,
Pinho and Crowley [2009] have concluded that the application of Equation (2.1) may be
reasonable for the period of vibration of RC frames with rigid infills, using the value of Ct
recommended for other structures (i.e., Ct = 0.050), but believe that many designers following the
code would use Ct = 0.075 for all reinforced concrete buildings, regardless of the details of the
masonry infills, even though many reinforced concrete buildings in Europe are constructed with
stiff masonry infills which are often not isolated from the RC frame.
However, it has to be pointed out that the initial stiffness (and period of vibration) of the infilled
frame configuration and the corresponding force demands are related to the case of undamaged
infills which are actually capable of transmitting a portion of the horizontal actions. At the design
seismic action, adopted for the dimensioning of the RC members, masonry infills are already
damaged to a certain degree, causing a reduced contribution to the strength and stiffness of the
structure, and the structural behaviour gradually approaches that of the bare configuration.
Nevertheless, when the design of the RC structure is carried out on the bare configuration, a
discrepancy between the structural ductility imposed using the behaviour factors, assigned
according to the code for ductile RC structures, and the ductility of commonly adopted non-
structural masonry infills may arise. Note that in the specific rules for steel building in Eurocode 8 -
Part 1 (CEN [2004a]) actually an upper limit reference value of the behaviour factor equal to 2.0 is
specified for moment resisting frames with concrete or masonry infills in contact with the frame.
Given the variety of available masonry infill typologies and construction techniques, the possible
variability of infill layouts during the life of the structure and the complexity of masonry infill
modelling, in particular for infills with openings, the design of regular infilled RC structures on
corresponding bare structural configurations can, in principle, be considered an acceptable
simplification for practical applications. This fact is supported also by experimental evidence
(Fardis et al. [1999]), but applies only if the infill damage is limited appropriately and RC structural
elements are verified against increased shear demands (Fardis [2006]). Explicitly, to satisfy the
code requirements, in the design of infilled RC frames a rational conceptual design has to be
followed, additional measures need to be taken, avoiding irregularities in plan and elevation,
damage limitation requirements have to be fulfilled and potential local effects due to the presence
of infills have to be accounted for.
adequate degrees of reliability related to life safety and damage limitation are satisfied.
Consequently, ultimate limit states (ULS), associated with no-collapse or life safety requirements,
have to be verified at a reference design seismic action for which a reference return period TULS is
defined in the National Annex of each country. The return period TULS = 475 years is adopted in
Italy for ordinary buildings. The code states that structures shall be designed and constructed to
withstand the ULS seismic action without local or global collapse, thus retaining its structural
integrity and a residual load bearing capacity after the seismic event. The verification of damage
limitation limit states (DLS) is required for the corresponding seismic action with a return period
TDLR, having a larger probability of occurrence than the design seismic action. For the needs of this
study, the damage limitation limit state is associated with a seismic action having a reference
return period of TDLR = 50 years, in accordance to the value set in the Italian National Code.
Structures shall be designed and constructed to withstand the DLS seismic action without the
occurrence of damage and the associated limitations of use, the cost of which would be
disproportionately high in comparison with the cost of the structure itself. A reliability differentiation
is implemented in the code through the classification of structures into different importance classes
and the assignment of corresponding importance factors γI.
The seismic hazard for a certain seismic region in general is described by a single parameter, the
reference peak ground acceleration agR on type A ground (i.e., rock or other rock-like geological
formation), corresponding to the reference return period TULS of the design seismic action for the
ultimate limit state requirement and an importance factor γI = 1.0. The earthquake ground motion
at a given site is represented by an elastic ground acceleration response spectrum Se(T). In order
to account for the lower return period of the DLS seismic action with respect to the ULS design
seismic action, the simple application of a reduction factor νDLS is allowed for the evaluation of
displacements and drifts, although the Italian National Code provides site-specific design spectra
for both the ULS and the DLS ground motion intensity. The possible influence of local ground
conditions on the seismic action for ground types other than rock is accounted for applying the soil
factor S.
The design procedure to be followed according to Eurocode 8 - Part 1 (CEN [2004a]) provisions for
the seismic design of RC structures is illustrated in Figure 2.8. For the definition of ULS design
seismic forces acting on the structure, in the case of a linear elastic analysis, elastic acceleration
response spectrum ordinates Se(T), defined corresponding to the peak ground acceleration
assigned to the considered site, are reduced introducing the behaviour factor q of the structure.
Hence, the design for resistance to seismic forces is accomplished performing the elastic analysis
based on a reduced elastic response spectrum, the design spectrum Sd(T), accounting for the
capacity of the structural system to resist seismic actions in the nonlinear range. Conversely, the
evaluation of seismic action effects for the DLS is performed with reference to the corresponding
elastic response spectrum.
Figure 2.8 Current design procedure for masonry infilled RC structures according to Eurocode 8
12 Experimental and numerical seismic performance of strong clay masonry infills
The damage limitation requirements, according to Eurocode 8 - Part 1 (CEN [2004a]), are
considered satisfied, as given in Equation (2.2), if the inter-storey drift δj for each storey j induced
by the damage limitation seismic action does not exceed the inter-storey drift limit δDLS, equal to
0.50% for buildings with brittle non-structural elements attached to the structure in each storey of
the building. In the case of ductile non-structural elements the drift limit δDLS is set equal to 0.75%
and in the case of non-structural elements fixed such that they do not interfere with structural
deformations, or without non-structural elements, the drift δDLS equal to 1.0% shall not be
exceeded.
δ j < δ DLS (2.2)
No further specifications are provided referring to the definition of brittle and ductile non-structural
elements. Probably in most cases, unreinforced masonry infills will intuitively be classified as
brittle. However, the fact that the code specifies an upper limit value of the behaviour factor qa
equal to 2.0 for non-structural elements in the case of exterior and interior walls, partitions and
façades, could also lead to the interpretation that non-structural infills can possess a certain level
of ductility.
Even though not clearly stated in the code, the drift verifications are commonly carried out on the
bare frame structural configuration and such approach is followed in the design carried out for the
prototype structures analysed within the scope of this study. In fact, considering masonry infills as
brittle, the requirement to apply a drift limit δDLS equal to 0.50% for infilled structures at the damage
limitation limit state would imply that any type of rigid infills could remain substantially undamaged
up to this level of drift (satisfying the damage limitation requirements), even in the case of infills
with openings. Evidently, the evaluation of displacements based on a linear elastic analysis
strongly depends on the values of stiffness assumed for load-bearing elements. Eurocode 8 - Part
1 (CEN [2004a]) states that the effect of cracking on the element stiffness has to be taken into
account and, unless a more accurate analysis of the cracked elements is performed, the elastic
flexural and shear stiffness properties may be taken to be equal to one-half of the corresponding
stiffness of the uncracked RC elements.
At the ultimate limit state, the safety verification has to be accomplished in terms of resistance to
seismic action effects for both structural and non-structural elements. In particular, for non-
structural elements that might, in case of failure, cause risk for human life or affect the main
structure of the building or services of critical facilities, the verification of resistance for the design
seismic action is foreseen and a simplified procedure is proposed for the evaluation of the seismic
force acting on the non-structural element. Accordingly, the effects of the seismic action may be
determined applying a horizontal force Fa, acting at the centre of mass of the non-structural
element in the most unfavourable direction. To satisfy the safety verification, the seismic force Fa
that can be expressed as equivalent pressure wa acting on the masonry infill, needs to be smaller
than the corresponding out-of-plane resistance wR, as given in Equation (2.3).
wa < wR (2.3)
Nevertheless, in Eurocode 8 - Part 1 (CEN [2004a]) no recommendation for the calculation of the
corresponding resistance of the infills is provided. In principle, resistance models based on full
vertical arching action, as for instance described in Eurocode 6 – Part 1-1 (CEN [2004b]), could be
used, although they are usually applied to elements subjected to non-seismic actions (wind or
gravity loads), and they may be considered appropriate only for undamaged infills. An expected
reduction of the out-of-plane strength of the infills according to the eventual damage due to
previous in-plane actions is not taken into consideration in codified procedures.
and are in contact with the frame without special separation joints or structural connections (e.g.,
through ties, belts, posts or shear connectors). However, in order to achieve advantageous
behaviour, in particular for infills that might be vulnerable to out-of-plane failure, the same
recommendations may also be adopted for infilled structures designed according to provisions for
low or medium ductility. For wall or wall-equivalent dual concrete systems, the interaction with the
masonry infills may be neglected, while for engineered masonry infills constituting part of the
seismic resistant structural system, analysis and design shall be carried out in accordance with the
criteria and rules provided for confined masonry.
Any possible consequences of irregularity in plan or elevation produced by the infills should be
taken into account. In particular, strongly irregular, unsymmetrical or non-uniform arrangements of
infills in plan should be avoided, taking into account the extent of openings and perforations. If due
to the unsymmetrical arrangement of the infills (e.g., existence of infills mainly along two
consecutive faces of the building), severe irregularities are induced in plan, spatial models should
be used for the analysis of the structure. In this case, infills should be included in the model and a
sensitivity analysis regarding the position and the properties of the infills should be performed (e.g.,
by disregarding one out of three or four infill panels in a planar frame, especially on the more
flexible sides). Special attention should be paid to the verification of structural elements on the
flexible sides of the plan against the effects of any torsional response caused by the infills. Infill
panels with more than one significant opening or perforation should be disregarded in models for
these analyses. Furthermore, it is stated that for structures with masonry infills not regularly
distributed, but not in such a way as to constitute a severe irregularity in plan, these irregularities
may be taken into account increasing by a factor of 2.0 the effects of the accidental eccentricity
defined in the code, indicating that also in this case the analysis of a bare frame structure may be
considered sufficient. In the case of considerable irregularities in elevation (e.g., drastic reduction
of infills in one or more storeys compared to the others), the seismic action effects in the vertical
elements of the respective storeys have to be amplified, if no more precise model is used, by the
magnification factor η, defined in Equation (2.4), where ΔVRw denotes the total reduction of the
resistance of masonry infills in the considered storey, compared to the more infilled storey above,
ΣVEd is the sum of the seismic shear forces acting on its vertical primary seismic members and q is
the behaviour factor of the structure.
∆V Rw (2.4)
η =1+ ≤q
∑ V Ed
For values of η smaller than 1.1, the modification of action effects may be neglected. In addition,
the code requires that high uncertainties related to the behaviour of masonry infills are considered.
The variability of mechanical properties of the infill walls and of their attachment to the surrounding
frame, possible modifications during the exploitation period of the building and the non-uniform
degree of damage suffered during the earthquake itself are also mentioned. However, no specific
solutions are provided to ensure the achievement of the given requirements.
In addition, possible adverse local effects due to the frame-infill interaction, such as potential shear
failure of columns under shear forces induced by the diagonal strut action of infills, are addressed
in Eurocode 8 - Part 1 (CEN [2004a]) and the application of appropriate measures for the
protection of structural elements is required. In particular, for infills in full contact with the adjacent
structural elements, shear demands induced by local effects due to the presence of infills VC,Ed,l
should not exceed the column resistance VC,Rd, according to Equation (2.5).
VC , Ed , l < VC , Rd (2.5)
For infilled frame structures of all ductility classes, except in the case of low seismicity, according
to the code, appropriate measures should be taken to avoid brittle failure and premature
disintegration of the infill walls, as well as the partial or total out-of-plane collapse of slender
masonry infills. Particular attention should be devoted to infill panels with a slenderness ratio (i.e.,
the smaller of the length or height to thickness ratio) greater than 15. As possible measures to
improve both in-plane and out-of-plane integrity and behaviour, light wire meshes, wall ties fixed to
14 Experimental and numerical seismic performance of strong clay masonry infills
the columns and cast into the bedding planes of the masonry, as well as concrete posts and belts
across the infill panels and through the full thickness of the wall are mentioned.
P. Morandi, S. Hak, G. Magenes EUCENTRE 15
Research Report
and ultimate limit state conditions, while in the out-of-plane direction the achievement of ultimate
levels of drift was envisaged. Constant vertical loads have been imposed on the RC columns
during the tests.
Table 3.1 Summary of performed quasi-static cyclic experimental tests
Reference Selected
typology typology
Masonry unit properties Strong masonry
Lightweight block
Tongue & groove system
with dry head joint
Nominal vertical resistance [MPa] 8.0 8.0
Vertical resistance (preliminary characterisation) [MPa] - 12.9
Nominal horizontal resistance [MPa] - 2.2
Horizontal resistance (preliminary characterisation) [MPa] - 2.2
Thickness [mm] 350 350
Height [mm] 230 - 250 235
Percentage of holes [%] 50 - 55 50
Minimum shell thickness [mm] 4.5 - 5.0 4.8
P. Morandi, S. Hak, G. Magenes EUCENTRE 17
Research Report
Figure 3.2 (a) Preparation of prism-shaped mortar test specimens, (b) Mortar testing in flexure, (c) Mortar
testing in compression
Thus, the selected typology represents a traditional strong single leaf unreinforced masonry infill of
35.0 cm thickness and consists of vertically hollowed tongue and groove block units, having a
volumetric percentage of holes ≈50% and a minimum thickness of webs and shells equal to 4.8
mm and 6.8 mm, respectively. The infills should be constructed after full hardening of the RC
frame, adopting traditional bed joints, having a thickness of about 1.0 cm. The bed joint mortar,
belonging to class M5, is applied in two strips in the longitudinal direction of the wall with an
intermediate cavity of about 2.0 cm (see Figure 3.3a). Full contact between the infill and the
surrounding RC members is assumed to be achieved filling the remaining vertical gaps on the two
sides of the infill and the horizontal gap at the top of the infill with mortar, as illustrated in Figure
3.3b.
18 Experimental and numerical seismic performance of strong clay masonry infills
(a)
(b)
Figure 3.3. Masonry mortar joints: (a) Bed joint - plan view & during construction, (b) Mortar joints adjacent
to RC members - front view & during construction
(b) Definition of RC Frame Layout. For the experiments the choice of a single RC frame
configuration has been envisaged and the constraints of available laboratory facilities for various
stages of the tests have been respected, with reference to specimen dimensions, available
equipment and instrumentation, as well as attachment points on the strong floor in the laboratory.
Since the quasi-static experiments have been carried out at the laboratory of the Department of
Civil Engineering and Architecture of the University of Pavia, for the in-plane tests the available
reaction wall could be utilised to introduce horizontal actions by means of a hydraulic actuator,
while for the out-of-plane force introduction a new steel reaction frame has been designed.
Preliminary dimensions of the single-storey single-bay RC frame specimen to be tested with and
without infill have been chosen with the aim to realistically represent the part of a full-scale RC
frame structure. Hence, a clear span of 4.22 m and a clear height of 2.95 m have been adopted
such that the number of masonry blocks that has to be cut in order to fit the dimensions of the
frame is reduced to a minimum. Beam and column lengths have been extended beyond the beam-
column panel zone with the aim to provide sufficient space for the appropriate anchorage of
reinforcement rebars and to facilitate the introduction of both horizontal and vertical forces. For the
foundation of the frame an inverted T-shaped cross section has been adopted consisting of a 0.40
m thick and 1.70 m wide slab and a stud of 0.35 m width and 0.20 m height. The width of the
foundation was primarily dictated by the position of the available holes in the strong floor. The front
view of the bare frame specimen and the cross section of the infilled frame specimens are shown
in Figure 3.4a and Figure 3.4b, respectively, while the front view of infilled configurations with full
(TA1, TA2 and TA3) or partial infill (TA4) and the vertical infill strip (TA5) is illustrated in Figure 3.5.
P. Morandi, S. Hak, G. Magenes EUCENTRE 19
Research Report
(a) (b)
Figure 3.4. (a) Front view of bare frame specimen; (b) Cross section of infilled frame specimen
(c)
Figure 3.5. Infilled frame specimen configurations: (a) Fully infilled; (b) Partially infilled; (c) Vertical infill strip
(a) (b)
(c)
Figure 3.6. (a) 4-storey prototype frame configuration for design [m]; (b) Single-storey single-bay
substructure; (c) Adopted design acceleration spectra according to NTC08 [2008]
The design has been carried out for the bare structural configuration and the given additional
measures for masonry infilled RC frames have been considered. Variable floor loads equal to 2.0
kN/m2 and permanent floor loads equal to 6.5 kN/m2 have been assigned to all storeys. The
building is designed for ductility class high (DCH); hence, the behaviour factor according to the
code results to be q = 5.85. The seismic action is represented by the site specific acceleration
spectra for the Italian site of Isernia (Scapoli) with a ULS design peak ground acceleration on
ground type A being ag = 0.35g. Soil class B is assumed resulting in a ULS design peak ground
acceleration equal to 0.35gS, where S = 1.076. Recall that according to the Italian National Code
provisions site specific spectra are defined also at the damage limitation state, for the considered
site with a peak ground acceleration equal to 0.142gS, where S = 1.2 (see Figure 3.6c). Material
properties corresponding to concrete class C28/35 and reinforcing steel class B450C are assumed
for the design of the frame. The safety verifications have been carried out for design action effects
evaluated from a multimodal spectral analysis on a simple linear-elastic model of the planar 4-
storey bare frame structure, using a commercial program for structural analyses. Rigid beam and
column ends are assumed in the joint panel zone and values of cracked flexural and shear
stiffness equal to 50% of the gross section stiffness are adopted. Referring to the verification of
local shear demands due to the presence of infills, the full length of the column has been
considered as critical length and the transversal reinforcement has been detailed accordingly.
Local ductility requirements have been satisfied and provisions for the detailing of beam-column
joints have been followed, as required by the code for RC frame members of buildings designed
for high ductility. Adequate values of reinforcement anchorage length have been adopted and the
anchorage of the column rebars has been enhanced through welded transversal rebars. Besides
the load cases related to the analysis of the frame specimen as a sub-structural unit of the 4-storey
prototype building, the single-storey single-bay frame has been verified for additional design
situations, accounting for specific load conditions arising in the laboratory, during transportation
and during the tests. In particular for the in-plane quasi-static tests, in order to introduce horizontal
cyclic actions using a single hydraulic actuator in push and pull, a longitudinal precompression at
least equal to the maximum expected resistance of the specimen has to be imposed on the beam
element. The increased values of beam moment resistance due to axial compression have a
P. Morandi, S. Hak, G. Magenes EUCENTRE 21
Research Report
significant influence on the frame response and have been accounted for in the verification of
capacity design principles in order to ensure that the formation of plastic hinges will start at the
bottom of the columns and at the beam ends. The foundation of the specimen has been designed
for crack control to adequately present a stiff support for the frame specimen. The relevant load
conditions occurring during transportation and during the tests have been accounted for. Material
properties corresponding to concrete class C35/45 and reinforcing steel class B450C are assumed
for the design of the foundation. A detailed plan showing the geometry of the specimen and
reinforcement details is given in Figure 3.7.
(a)
(b)
Figure 3.7. Reinforcement details: (a) Frame and foundation front view; (b) Foundation plan view
(b) RC Frame and Masonry Infill Construction. The bare and infilled frame specimens have
been constructed such that, first of all, the reinforcement of the foundation and of the columns has
been placed in position (Figure 3.8a), then the T-shaped foundation has been cast in two phases,
firstly the bottom part, i.e. the slab, and then the stud (Figure 3.8b). The foundation has been cast
leaving the holes for the anchorage of the specimen to the strong floor. Moreover, eight steel studs
having a load-bearing capacity of 200 kN have been cast in the foundation, of which four were
used for the transportation of the specimen and four for the introduction of vertical loads on the
columns. Subsequently, the columns have been cast up to the level below the beam-column joints
(Figure 3.8c). Then, the reinforcement of the beams has been placed in position (Figure 3.8d) and
22 Experimental and numerical seismic performance of strong clay masonry infills
the beam has been cast together with the studs of the columns above the beam (Figure 3.8e). The
finalised bare frames are shown in Figure 3.8f.
(a) (b)
(c) (d)
(e) (f)
Figure 3.8. Construction of test specimens – RC frame: (a) Positioning of foundation and column
reinforcement; (b) Casting of foundation; (c) Casting of columns; (d) Positioning of beam reinforcement
(beam-column joint); (e) Casting of beams; (f) Finalised bare frames
After all, for all except one frame, the masonry infill has been constructed (Figure 3.9a, following
the indications given in the description of the selected masonry infill typology. The finalised infilled
frames are shown in Figure 3.9b.
(a) (b)
Figure 3.9. Construction of test specimens – Masonry infill: (a) Infill construction; (b) Finalised infilled frames
attached to the strong floor through the holes in the foundation by means of ten prestressed steel
bars. A force transfer through pure friction between the foundation and the strong floor was
assumed, adopting a safe-sided friction coefficient equal to 0.20. Introducing a precompression of
400 kN for each steel bar, the maximum predicted in-plane reaction, i.e. Fh,max = 700 kN, could be
safely transmitted.
Given the estimated maximum horizontal load-bearing capacity of the most resistant specimen, an
actuator of 1000 kN capacity with an internal load cell was adopted. The length of the actuator at
mid stroke equals 3566.2 mm, allowing ± 266.7 mm of displacement. To transmit the horizontal
force to the frame, a steel plate of 80 mm thickness was placed at each end of the RC beam and
the two plates were connected by means of four prestressed steel bars. Such system allowed the
achievement of reverse loading cycles in both directions using a single actuator on one end of the
specimen, ensuring that the steel plates are always in contact with the RC frame. Hence, a total
prestressing force equal to the maximum horizontal force, i.e. Fh,max = 700 kN, needed to be
achieved and each of the four steel bars has to attain a prestressing force of at least 175 kN. Note
that different axial load conditions occur in the RC beam when positive (push) and negative (pull)
actions are imposed by the actuator on the frame. For the hypothetical case when the
precompression equals exactly the maximum applied horizontal load, as illustrated in Figure 3.10,
when positive loads are applied, the full precompression and half of the applied load are imposed
on the beam, i.e., Nmax= 1.5Fh,max. In the reverse direction, at the onset of negative force
application, the precompression theoretically drops to zero and only half of the applied load is
imposed on the beam, i.e., Nmax= 0.5Fh,max. The variation of axial load in the beam may significantly
influence the response of the frame and has been accounted for in the design of the specimen.
The effective precompression should be slightly higher than the maximum expected force to be
applied by the actuator in order to account for the instantaneous losses of precompression due to
the shortening of the beam and the anchorage of the bars in the steel plates. Time-dependent
losses and other short-term losses, which are commonly considered in verifications for prestressed
concrete, such as due to friction, can in this case be neglected. An equivalent prestressed system
consisting of two 80 mm thick steel plates and four steel bars prestressed to a total force of at least
Fh,max = 700 kN was adopted to attach the actuator to the reaction wall.
Figure 3.11. Layout of the in-plane setup for quasi-static cyclic tests
(b) Out-of-plane Setup. For the needs of the experimental campaign the construction of a newly
designed steel reaction frame was foreseen that was aimed to serve (i) as a reaction frame and an
out-of-plane restraint during the out-of-plane quasi-static tests, (ii) as an out-of-plane restraint
during the in-plane quasi-static tests, (iii) as an out-of-plane restraint during future shaking table
tests. The static out-of-plane tests needed to be carried out immediately following the in-plane
experimentation, since they were performed on masonry infills that have previously sustained a
certain level of in-plane damage. Hence, the out-of-pane reaction frame was constructed next to
the in-plane setup, such that the specimens for the application of cyclic loading in both directions
can remain connected to the strong floor in the same position. Therefore, the out-of-plane setup
could easily be utilised also during the in-plane quasi-static tests to prevent possible out-of-plane
displacements of the RC frame. The steel setup was assembled on a 35.0 cm thick RC foundation
in order to provide a sufficiently rigid support and to ensure a satisfactory transfer of forces to the
strong floor, given the limited number of possible attachment points. The position of the out-of-
plane setup, centred with respect to the frame specimen, is illustrated in Figure 3.12a. The steel
structure consists of a central reaction plane that serves as a support for the actuator introducing
the horizontal load on the masonry infill and two external restraining planes (see Figure 3.12b).
The three structural planes are connected in the transversal direction with a steel beam allowing
the attachment of the actuator. Additionally, above the actuator, the two restraining planes are
connected with a steel profile that can provide support for the self-weight of the actuator. In the
restraining planes, the steel reaction frame can be connected to the beam of the RC frame
specimen.
P. Morandi, S. Hak, G. Magenes EUCENTRE 25
Research Report
(a)
(b)
Figure 3.12. Layout of the out-of-plane setup for quasi-static cyclic tests: (a) Plan view; (b) Side view
L-shaped steel profiles attached to the cantilever beams reaching above the specimen are used as
restraints to prevent out-of-plane displacements of the RC frame and allow the application of out-
of-plane forces on the infill. The load was applied at the mid-height of the masonry infill by means
of a servo-controlled hydraulic actuator with an internal load cell, having a capacity of 650 kN and
450 kN in push and pull, respectively. The length of the actuator at mid stroke is 2534.0 mm,
allowing ±203.0 mm of displacement.
The design of the steel reaction frame has been carried out assuming a maximum horizontal force
of Fo,max = 600 kN and providing sufficient stiffness to limit the design displacements to a maximum
value of 0.5 mm. The horizontal force resisted by the masonry infill can be in part transmitted to the
foundation of the specimen, i.e., directly to the strong floor, while in part it is sustained by the out-
of-plane restraints of the steel frame, counteracting the reaction of the actuator. Consequently, the
obtained system is partially self-equilibrated; if adequate load transmission is ensured at the points
of restraint on the beam of the RC frame, the transfer of about 40% of the horizontal force through
the restraints may be assumed.
The application of a system for the load transfer from the actuator to the masonry infill consisting of
a series of hinged steel beams and plates was envisaged, allowing the introduction of the
horizontal force in a relatively large number of discrete points. In particular, for the fully infilled
frame specimens (TA1, TA2 and TA3) the application of forces in two lines and eight points per
line was foreseen (Figure 3.13), reproducing an approximately linear load distribution at the mid-
height of the panel. For the partially infilled configuration (TA4) the force was introduced through
26 Experimental and numerical seismic performance of strong clay masonry infills
four points on each part of the infill (Figure 3.14), while for the 1.38 m wide infill strip (TA5) the load
introduction through eight central points (Figure 3.15) was aimed to reproduce vertical bending. A
plan view of the load transfer system is given in Figure 3.16.
Figure 3.13. Front view of system for out-of-plane load transfer from actuator to masonry infill (TA1, TA2 &
TA3)
Figure 3.14. Front view of system for out-of-plane load transfer from actuator to masonry infill (TA4)
P. Morandi, S. Hak, G. Magenes EUCENTRE 27
Research Report
Figure 3.15. Front view of system for out-of-plane load transfer from actuator to masonry infill (TA5)
The characterisation of masonry units was carried out following European norms, i.e. Specification
for masonry units - Part 1: Clay masonry units (EN 771-1:2003/A1:2005, CEN [2005]) and Methods
of test for masonry units – Determination of compressive strength (EN 772-1 CEN [2000])
prescribes the determination (i) of the dimensions for ten units, (ii) of the configuration for ten units,
(iii) of the gross and net density for ten units, (iv) of the percentage of holes for ten units, (v) of the
thickness of webs and shells for ten units, (vi) of the combined thickness of webs and shells for
three units, (vii) of the vertical compression strength for ten units, and (viii) of the horizontal
compression strength for ten units. With the aim to evaluate, in addition to the average vertical
compression strength of the masonry units, also the characteristic value, as indicated in the
explanatory of the Italian National Code (NTC08 Circolare esplictiva [2009]), 30 units have been
tested in vertical compression. A summary of the results obtained from tests of vertical and
horizontal compression strength is given in Table 3.5.
Table 3.5 Summary of compression test results on masonry units
Vertical compression Horizontal compression
f fnorm. f fnorm.
[MPa] [MPa] [MPa] [MPa]
Mean strength, fb 8.64 9.81 2.78 3.15
St. dev. 0.79 - 0.30 -
C.o.v. 9.2% - 10.8% -
Characteristic strength, fbk 7.34 - 1.95 -
According to the norm related to mortar testing Methods of test for mortar for masonry -
Determination of flexural and compressive strength of hardened mortar (EN 1015-11/A1 CEN
[2006]), flexural and compressive tests on three prism-shaped specimens of mortar (of 40 x 40 x
160 mm dimensions) were carried out for each batch of mortar. Average values of flexural and
compressive strength equal to 2.15 MPa and 7.68 MPa, respectively, have been found.
Related to the resistance of masonry as a composite material in Methods of test for masonry -
Determination of compressive strength (EN1052-1, CEN [1998]), the requirements related to the
determination of the vertical and the horizontal strength on masonry wallettes are defined. For both
directions six specimens have been tested. On all masonry units and wallettes to be tested, in the
regions of support or force application, a cement mortar topping of about 1.0 cm was applied with
the aim to ensure even contact surfaces.
(b) Vertical Compression Strength. The dimensions, instrumentation and construction of the
wallettes for vertical compression tests are illustrated in Figure 3.17, while the corresponding
results are summarized in Table 3.7 and Figure 3.18. Examples of typical failure patterns obtained
due to vertical compression are shown in Figure 3.19.
Figure 3.17. Masonry characterisation – details and instrumented specimen: Vertical compression
Figure 3.20. Masonry characterisation – details and instrumented specimen: Lateral compression
P. Morandi, S. Hak, G. Magenes EUCENTRE 31
Research Report
Table 3.8 Summary of lateral compression test results on masonry wallettes
Specimen f E f E
n° [MPa] [MPa] [MPa] [MPa]
1 1.29 310 Mean strength
1.08 494
2 1.11 635 fmhm
3 0.85 488 St. dev. 0.16 162
4 1.23 265 C.o.v. 14.5% 32.8%
5 1.11 550 Characteristic strength
0.85 -
6 0.91 712 fmhk
Figure 3.23. Masonry characterisation – details and instrumented specimen: shear strength
The corresponding results are summarized in Table 3.9 and Figure 3.24, illustrating the evaluation
of the initial shear strength under zero compression fv0 and the corresponding characteristic value
fv0k, as well as the corresponding friction coefficient µ = tanα and its characteristic value µk = tanαk.
Examples of typical failure patterns obtained during the shear sliding tests are shown in Figure
3.25.
Table 3.9 Summary of shear strength test results on masonry triplets
Subsequently, using the system for vertical force introduction, on both columns simultaneously, the
load has been imposed by the hydraulic jacks at a velocity of ca. 2.0 kN/s. The achieved vertical
load of 400 kN per column has been kept constant during the entire test. Finally, reverse cycles of
horizontal in-plane loading (first pull, then push) were imposed on the frame in the height of the
beam by means of the servo-controlled hydraulic actuator. Firstly, two different levels of force-
controlled loading were accomplished, while subsequently displacement-controlled loading cycles
at increasing levels of in-plane drift were imposed, as summarised in Table 4.1. Clearly, not all
levels of drift have been achieved in each test. For each level of loading (target force or
displacement) three complete reverse loading cycles have been carried out and the duration of
load application has been kept approximately constant.
36 Experimental and numerical seismic performance of strong clay masonry infills
(b) Instrumentation. In order to measure the displacements and deformations of the specimen
during the in-plane test, displacement transducers (linear potentiometers) have been adopted, i.e.
in total, 22 potentiometers have been used for the bare frame (TNT_IP), 24 for the fully infilled
frame configurations (TA1_IP, TA2_IP and TA3_IP), and 32 for the partially infilled frame (TA4_IP).
In addition, an optical acquisition system has been installed to measure the in-plane displacements
of optical markers uniformly distributed (ca. 40 to 50 cm spaced) on the RC frame as well as on the
masonry infill. Typical layouts of the displacement transducers adopted for the bare frame and the
infilled frame specimens (Table 4.2) are schematically represented in Figure 4.1, Figure 4.2 and
Figure 4.3.
The centreline deformation of the RC frame has been measured using a typical instrumentation
scheme consisting of potentiometers along the column heights (3, 5), the beam length (4) and the
corresponding frame diagonals (6, 7). In addition, to record the deformed shape of the beam,
vertical displacements of the beam have been measured in two points within the span of the
specimen (8, 9). Horizontal displacements of the frame have been measured on top at the
centreline of the beam (13 at the beam end, 14 at the centre of the beam) as well as at the half-
height of the column (15, 16). Deformations in the plastic hinge regions at the bottom of the
columns have been recorded using pairs of potentiometers (17 and 18, 19 and 20) for each
column.
Any potential horizontal displacement of the foundation has been measured at the foundation
centre on one side of the specimen (10), while the occurrence of possible uplift has been recorded
using one potentiometer on each end of the foundation (11, 12). In order to control the out-of-plane
displacements of the specimen during in-plane loading, two potentiometers (21, 22) have been
placed in the beam centreline within the frame span orthogonally to the plane of the specimen. For
the fully infilled frame specimens deformations of the masonry infill were measured using two
potentiometers (23, 24) placed on the diagonals in the central part of the panel. In the case of
partial infill five potentiometers were placed on each part of the wall, two in the vertical direction
(27 and 29, 30 and 32), one at the top of the infill (28, 31) and two diagonally (23 and 24, 25 and
26). In addition, displacements possibly occurring in the actuator hinges have also been recorded
(1, 2).
P. Morandi, S. Hak, G. Magenes EUCENTRE 37
Research Report
Table 4.2 Adopted displacement transducers for in-plane tests
TNT TA1_IP TA2_IP TA3_IP TA4_IP
No. l0 [mm]
1 25 25 25 25 25
2 25 25 25 25 25
3 100 100 100 100 100
4 50 50 50 50 50
5 100 100 100 100 100
6 250 100 100 100 100
7 250 100 100 100 100
8 100 100 100 100 100
9 100 100 100 100 100
10 25 25 25 25 25
11 25 25 25 25 25
12 25 25 25 25 25
13 250 250 250 250 250
14 250 250 250 250 250
15 250 100 100 100 100
16 250 100 100 100 100
17 100 50 50 50 50
18 100 50 50 50 50
19 100 50 50 50 50
20 100 50 50 50 50
21 25 25 25 25 25
22 25 25 25 25 25
23 - 100 100 100 100
24 - 100 100 100 100
25 - - - - 100
26 - - - - 100
27 - - - - 50
28 - - - - 50
29 - - - - 50
30 - - - - 50
31 - - - - 50
32 - - - - 50
38 Experimental and numerical seismic performance of strong clay masonry infills
(a)
(b)
Figure 4.1. Bare frame specimen instrumentation – in-plane: (a) Front view; (b) Back view
P. Morandi, S. Hak, G. Magenes EUCENTRE 39
Research Report
(a)
(b)
Figure 4.2. Infilled frame specimen instrumentation (TA1_IP, TA2_IP & TA3_IP): (a) Front view; (b) Back
view
40 Experimental and numerical seismic performance of strong clay masonry infills
(a)
(b)
Figure 4.3. Infilled frame specimen instrumentation (TA4_IP): (a) Front view; (b) Back view
Additionally, during the in-plane test on the bare frame (TNT_IP) and on two infilled frame
configurations (TA1_IP, TA2_IP) the deformations of the reinforcement rebars in potential plastic
hinge regions of the RC structural elements have been monitored my means of strain gauges
attached during the construction of the specimens. In particular, four strain gauges were installed
at each of the beam and column ends, resulting in a total of 24 instruments per specimen, as
illustrated in Figure 4.4.
P. Morandi, S. Hak, G. Magenes EUCENTRE 41
Research Report
Figure 4.4. Layout of strain gauges installed on the RC frame reinforcement (TNT, TA1, TA2)
(a) Bare Frame TNT. Following the given loading history scheme (Table 4.1), the bare frame
(specimen TNT) has been loaded at two different force-controlled loading levels (assuming
0.15Fmax ≈ 40.0 kN, 0.30Fmax ≈ 80.0 kN) and subsequently at thirteen different target displacements
(07D, 08D, 09D, 10D, 11D, 12D, 13D, 14D, 15D, 16D, 17D, 18D, 19D), up to a maximum drift of
3.50%. The complete displacement history is shown in Figure 4.5, while the obtained
corresponding in-plane force-displacement response is shown in Figure 4.6.
During the force-controlled and the first two displacement-controlled levels of loading no damage
could be observed on the RC elements. Initial cracks started to be visible on the interface between
the column and the beam in the right corner of the specimen at 09D, while first flexural cracks have
formed at the column bases as well as in the upper part of the column at 10D, while no cracks
were visible yet on the beam. These flexural cracks, above all at the bottom, but also at the top of
the column, continued to propagate progressively during the following loading cycles (11D, 12D,
13D, 14D).
Figure 4.7. Observed crack pattern - Bare frame TNT, 3.50% drift (19D)
P. Morandi, S. Hak, G. Magenes EUCENTRE 43
Research Report
Figure 4.8. Bare frame TNT during the in-plane test at 19D loading amplitude (3.50% drift)
(a) (b)
Figure 4.9. Details of plastic hinge regions at 19D (3.50% drift): (a) RC beam; (b) RC column
First cracks on the RC beam ends were clearly visible at a drift of 1.75% (15D) and propagated
further at of 2.00% drift (16D) and subsequent loading cycles. At a drift of 2.50% (17D) spalling of
the concrete cover could be observed that continued further at 3.00% (18D), in particular at the
bottom of the columns. The concrete cover was found to be completely detached in the plastic
hinge region at the bottom of the columns at a drift of 3.5% (19D). At the same time, concrete
cover was spalling also at the beams ends, indicating after all the formation of a weak beam,
strong column response mechanism, in line with the expected behaviour imposed according to the
design approach followed in modern seismic code provisions. The damage observed at the final
drift equal to 3.5% (19D) is illustrated in Figure 4.7, the corresponding deformed shape of the
specimen during the test at the maximum amplitude is shown in Figure 4.8, while details of the
plastic hinge regions are shown in Figure 4.9.
(b) Infilled Frame TA1_IP. The infilled frame specimen TA1 has been loaded at two different
force-controlled loading levels (assuming 0.15Fmax ≈ 100.0 kN, 0.30Fmax ≈ 200.0kN) and
subsequently at thirteen different target displacements (01D, 02D, 03D, 04D, 05D, 06D, 08D, 09D,
10D, 11D, 12D, 13D, 14D), up to a maximum drift of 1.50%. The complete displacement history is
shown in Figure 4.10. The obtained corresponding in-plane force-displacement response is shown
in Figure 4.11.
During the initial force controlled loading cycles (01F and 02F, reaching drifts below 0.02% and
0.04%, respectively) first cracks have occurred on the plaster covering the joint between the
masonry infill and the surrounding frame, both along the columns and along the beam. On the joint
between the beam and the infill a slight detachment of the plaster could be observed along the
cracks, while the masonry has remained undamaged. In the following first displacement controlled
cycle (01D) the cracks on the plaster have extended further towards the bottom of the columns and
have become more pronounced. In particular, the crack between the beam and the infill has
widened noticeably.
44 Experimental and numerical seismic performance of strong clay masonry infills
At the following loading cycle (2D) first cracks have been observed in the mortar joints of the
masonry infill, primarily in the region of contact between two blocks, predominantly in the central
zone of the panel. The cracks in the plaster were already clearly marked and smaller parts of the
plaster along the beam were falling off or tended to detach. Subsequently, at load case 3D cracks
started to propagate evidently along the masonry mortar joints, in particular in the upper corners of
the infill and in the centre of the panel. Further pieces of plaster detached and fell off at the top
joint. Few bricks in the upper corners were slightly damaged at their borders. At the subsequent
displacement level (4D) first cracks started to form on the blocks in the upper corners of the infill.
Further cracks appeared along the mortar bed joints and existing cracks in the mortar joints
continued to propagate. The primary concentration of cracks on the diagonals of the panel could
already be distinguished (i.e., from the corners towards the centre and at mid height of the panel),
similar to a stepwise crack pattern, since cracks have occurred predominantly in the bed joints,
while the unfilled head joints allowed some extent of deformation without damage of the blocks
themselves. During the 5D cycles further cracks formed in the bed joints following the diagonal
pattern. Some new cracks formed on blocks in the upper corners of the infill and first more
pronounced cracks formed on single blocks, predominantly in the corners of the infill. The cracks
on the joint between the infill and the frame became pronounced, in particular between the top of
the panel and the beam, towards the corners. At a drift of 0.30% (6D) new cracks have formed on
blocks in the corners of the panel. The existing cracks in the joints have continued to propagate
and some became rather pronounced. At the following displacement level (8D) new cracks formed
on other blocks in the corners of the infill and further cracks occurred in the bed joints,
predominantly around the mid height of the panel. Little pieces of the outer shell of some blocks
started to detach. Achieving a drift of 0.50% (9D), as shown in Figure 4.12a, the damage in the
P. Morandi, S. Hak, G. Magenes EUCENTRE 45
Research Report
corners has propagated further, pieces of the outer shell of some blocks in the corners have
detached and fallen off, new cracks have formed on some blocks and the cracks in the bed joints
have continued to extend. Cracks in the bed joints around the centre of the panel have continued
to propagate and started forming also in the mortar joints closer to the top of the infill. At the next
loading level (10D) additional cracks on the blocks in the top corners have formed, further blocks in
the top corners were damaged and pieces of the outer shells have fallen off. Parts of the plaster at
the frame joint have detached. Cracks in the bed joints around the mid height of the panel have
started to extend more markedly also towards the columns and the amount of plaster and masonry
rubble on the ground has become notable. No cracks have been detected in the RC structure.
At the 11D target displacement further cracks of both mortar joints and bricks formed at the centre
of the panel, starting to extend also towards the bottom. The damage of the blocks in the corners
of the panel has propagated further and portions of external shells of several blocks have fallen off.
Diagonal cracks have been observed on the columns. Subsequently, at 1.00% (12D) external
shells have fallen off and further parts on other blocks have detached. Further diagonal cracks
formed in the upper parts of the columns, and a vertical crack opened at both beam ends. Light
horizontal cracks could be observed at the interface between the column and the beam below the
beam-column joint. Subsequently, (at 13D) the external shells of further blocks fell off, resulting in
a considerable amount of rubble on the ground. The cracks on the columns continued to propagate
and the horizontal crack at the interface below the beam-column joint became more pronounced.
At the last target drift equal to 1.50% (14D), illustrated in Figure 4.12b, further cracks developed
and additional portions of masonry have fallen off.
(a)
(b)
Figure 4.12. Observed crack pattern - Infilled frame TA1: (a) 0.50% drift (09D); (b) 1.50% drift (14D)
46 Experimental and numerical seismic performance of strong clay masonry infills
Figure 4.13. Infilled frame TA1 during the in-plane test after 14D loading (1.50% drift)
(a) (b)
Figure 4.14. Details of in-plane damaged TA1 infill corners (back side) after 14D loading (1.50% drift): (a)
Upper left; (b) Upper right
(c) Infilled Frame TA2_IP. The infilled frame specimen TA2 has been loaded at two different
force-controlled loading levels, for 0.15Fmax ≈ 100.0 kN and 0.30Fmax ≈ 200.0kN, and afterwards at
eighteen different target displacements (01D, 02D, 03D, 04D, 05D, 06D, 08D, 09D, 10D, 11D,
12D, 13D, 14D, 15D, 16D, 17D), up to a maximum drift of 2.50%, as illustrated by the loading
history in Figure 4.15. The related in-plane force-displacement response is shown in Figure 4.16.
P. Morandi, S. Hak, G. Magenes EUCENTRE 47
Research Report
During the first force controlled cycles (01F) (reaching a drift around 0.01%) first light vertical
cracks have occurred on the surface of the plaster covering the joint between the masonry infill
and the adjacent columns. These cracks continue to develop during the subsequent loading (02F),
and appeared also along the joint between the infill and the beam, while the masonry has
remained undamaged. In the first displacement controlled cycle (01D) the cracks on the plaster
have extended further and have become more pronounced. Initial, very light cracks have become
visible on the mortar joints of the infill. During subsequent loading (02D) the cracks in the mortar
joints have become more pronounced indicating already the development of a stepwise crack
pattern. The noise of creaking could be observed during the test and therefore slight sliding of the
masonry blocks may be presumed. At the next loading cycles (03D) the cracks in the mortar joints
have extended and the cracks on the joint between the infill and the frame have propagated
further.
Subsequently, during the 04D load case some of the cracks in the bed joint have continued to
propagate emphasizing the stepwise crack pattern (i.e., from the corners diagonally towards the
centre and at mid height of the panel), while the unfilled head joints allowed some extent of
deformation without cracking of the blocks themselves. However, the onset of damage on the
corners of single masonry units could be observed. At a drift of 0.25% (5D) first diagonal cracks on
blocks in both upper corners of the panel have started to develop. The detachment of little pieces
of plaster could be observed, which continued to fall off at the following level of displacement (6D).
At the 08D loading cycles the cracks on the blocks in the upper corners of the infill developed
further. Two evident cracks appeared on blocks in the lower central part of the panel.
48 Experimental and numerical seismic performance of strong clay masonry infills
Subsequently, at target drift 09D pieces of the outer shell of damaged blocks in the upper left
corner have detached and fallen off. New cracks have formed on the blocks in both upper corners
and the existing ones have continued to propagate. At the next loading cycles (10D) further cracks
could be observed in the upper corners, and some new cracks have formed on blocks in the lower
central part of the infill. The stepwise crack pattern was clearly marked, spreading over a
considerable portion of mortar bed joints throughout the panel. During the 11D loading cycles
strong noise of creaking could be noted close to the peak drifts (equal to 0.80%) and pieces of the
outer shell of damaged blocks in both upper corners have detached and fallen off. First diagonal
cracks have been observed in the upper parts of the columns. At a drift of 1.00% (12D), as
illustrated in Figure 4.17a, the remaining outer shells of the damaged blocks in the upper corners
of the infill have fallen off completely and further shear cracks have been observed on the
columns. During the subsequent loading cycles (13D) the outer shell of further blocks in the upper
corners of the infill have fallen of, as well as of several blocks in the lower central part of the panel.
Some new diagonal cracks have formed on the RC columns. After the next target drift (14D) the
majority of masonry blocks in the top course of the infill has lost the outer shell, in particular
towards the corners of the panel. The outer shell of many blocks in the lower part (i.e. the second
course) of the masonry infill has fallen off, mainly in the middle part. The cracks in the upper part of
the columns continue to propagate but seem to be less pronounced than during the test TA1_IP.
During the 15D loading cycles the outer shells of further blocks in the second and in the last course
have continued to detach and fall off. Further shear cracks and a few horizontal cracks have
developed in the upper parts of the columns. Horizontal cracks have formed on the interfaces
between the columns and the beam. At the loading cycles with a target drift equal to 2.00% (16D)
a very intensive damage of the masonry infill could be identified. The outer shells of further blocks
in the second and in the last course, as well as of some blocks in the third and the fourth course
have been lost. The cracks on the RC elements continued to develop and first light flexural cracks
have been observed in the bottom part of the column. After the last cycles of loading (17D),
experiencing a drift of 2.5%, as illustrated in Figure 4.17b, the masonry infill has substantially been
destroyed, even though no strongly pronounced strength degradation of the specimen has been
recorded. The deformed shape of the specimen during the test sustaining the maximum level of
drift is shown in Figure 4.18.
P. Morandi, S. Hak, G. Magenes EUCENTRE 49
Research Report
(a)
(b)
Figure 4.17. Observed crack pattern - Infilled frame TA2: (a) 1.00% drift (12D); (b) 2.50% drift (17D)
Figure 4.18. Infilled frame TA2 during the in-plane test at 17D loading amplitude (2.50% drift)
50 Experimental and numerical seismic performance of strong clay masonry infills
(d) Infilled Frame TA3_IP. Also for the infilled frame specimen TA3 two levels of force-controlled
loading have been accomplished, for 0.15Fmax ≈ 100.0 kN and 0.30Fmax ≈ 200.0 kN. Subsequently,
the specimen has been loaded at six different target displacements (06D, 08D, 09D, 10D, 11D,
12D), up to a maximum drift of 1.00%, as illustrated by the loading history in Figure 4.19. The
related in-plane force-displacement response is shown in Figure 4.20.
During the first force controlled cycles (01F) first light vertical cracks have occurred on the surface
of the plaster covering the joint between the masonry infill and the adjacent columns. These cracks
have developed further during the subsequent loading (02F), reaching a drift of ca. 0.055%. The
reverse cyclic loading has been continued with the displacement level 06D (corresponding to a drift
of 0.30%) associated with the enlargement of the cracks in the bed joints and the opening of light
diagonal cracks in the upper corners of the infill. At the 8D loading cycles new diagonal cracks
have formed on some masonry blocks, in addition to the further development of the cracks in the
mortar joints. During the 9D loading cycles (see Figure 4.21a) the detachment of outer shells of
blocks in the upper left corner of the infill, (which were already damaged during the single larger
displacement imposed) has become more pronounced. An evident diagonal crack has formed on a
masonry block in the upper right corner of the panel. Previously created cracks have continued to
propagate. Subsequently, at 10D, the cracks on the blocks have increased further and some new
cracks have appeared in the mortar bed joints. At the following loading cycle (11D), strong noise of
creaking could be observed close to the peak displacements and further parts of the outer shells of
the blocks in the upper corners have detached and fallen off. A stepwise crack pattern can be
clearly observed. At the final displacement level (12D with a drift equal to 1.00%, see Figure
4.21b), further outer shells of the masonry blocks in the upper corners and of one block in the
P. Morandi, S. Hak, G. Magenes EUCENTRE 51
Research Report
fourth course from above (on the back side of the panel) have fallen off. Further cracks have
developed throughout the infill. Details of the damage in the corners of the panel (on the back side)
of the specimen after having sustained the maximum level of drift are shown in Figure 4.22, and
the front view of the entire damaged specimen is shown in Figure 4.23.
(a)
(b)
Figure 4.21. Observed crack pattern - Infilled frame TA3: (a) 0.50% drift (09D); (b) 1.00% drift (12D)
(a) (b)
Figure 4.22. Details of damaged TA3 infill corners (back side): (a) Upper left; (b) Upper right
52 Experimental and numerical seismic performance of strong clay masonry infills
Figure 4.23. Infilled frame TA3 during the in-plane test after 12D loading (1.00% drift)
(e) Infilled Frame TA4_IP. Prior to the in-plane test on the partially infilled frame specimen TA4
some cracks could be observed on the RC beam and on the foundation, since the same frame has
previously been used for the test on the vertical infill strip (TA5_OP). The horizontal cracks
between the beam and the infill have, however, reclosed as the in-plane setup was placed in
position, probably due to the slight pretension of the bars for vertical load introduction. Light cracks
have remained on the plaster covering the joint between the beam and the masonry, while the
partial infill itself was found to be completely undamaged. Initially, two levels of force-controlled
loading have been carried out, for 0.15Fmax ≈ 60.0 kN and 0.30Fmax ≈ 120.0 kN. Following the
force-controlled loading, the specimen has been loaded at ten different target displacements (02D,
03D, 04D, 05D, 06D, 08D, 09D, 10D, 11D, 12D), up to a maximum drift of 1.00%, as illustrated by
the loading history in Figure 4.24. The related in-plane force-displacement response is shown in
Figure 4.25. For the first level of force loading has been started in the push direction, while for all
subsequent cycles the usual loading direction (first pull, the push) has been adopted.
At 01F light vertical cracks have been observed on the plaster between the columns and the infill.
The opening of some mortar joints could be noticed, in particular on the right panel. During the 02F
loading cycles, reaching a drift of ca. 0.10%, the cracks on the plaster have slightly increased. At
the first displacement-controlled cycles (02D), on the left panel a vertical crack between the infill
and the column along the full masonry height has been observed. On the right panel a similar
crack has developed up to the middle of the second block from the bottom. In addition, a horizontal
crack in the joint above the second course has been observed. Subsequently, at 03D, horizontal
cracks between the beam and the partial infills could be observed, starting from the infill angels
and extending up to two blocks before the central opening. No further cracks have formed in the
joints, or on the blocks themselves. At the following level of displacement (04D), the cracks on the
interface between the infill and the RC column continued to extend and a new crack has formed in
the bottom left corner above the foundation. The crack in the bed joint above the second course of
the right infill panel continued to propagate. During the 05D loading cycles a first crack in the
vertical direction has formed on the block in the upper right infill corner. Additional cracks have
appeared in the first course of the right panel and in the fourth course of the left panel above the
half block next to the column. Subsequently, at 06D, no other cracks have formed on the blocks.
However, further cracks have developed in the horizontal mortar joints, on both sides of the partial
infill. At the next level of displacement, i.e., at 08D (0.4% drift), first noise of creaking could be
observed and during the first cycle of loading in the negative direction (push) an evident diagonal
crack in the left panel has formed, in particular crossing the blocks. The diagonal crack has closed
again during unloading, but at the following cycles notable strength degradation has been
observed. At 09D, reaching a drift of 0.50% (Figure 4.26a), further creaking could be observed
during loading in both directions. The diagonal crack opening in correspondence to negative
loading has extended noticeably; however, the achieved force appears to be almost equal for both
push and pull, although the force-displacement curve provides an unsymmetrical behaviour. On
the bottom left side of the right panel diagonal cracking has been observed, but no full crack along
the entire diagonal of the panel has formed. An increase of the diagonal strut action could be noted
on the left part of the partial infill and first detachment of outer shells in the bottom right corner of
the left infill has been observed. Horizontal cracks of the bed joints have continued to develop
further. During loading in the negative direction (push) the creation of a horizontal crack above the
first course of the right panel has been observed, practically resembling an overturning
mechanism. Subsequently, at the 10D loading cycles, the diagonal crack on the left panel has
widened further and additional damage of the outer shells in the right bottom corners has been
noted. On the right part of the partial infill the crack above the first course has extended further and
a stepwise crack pattern has developed starting from the bottom left corner and reaching up to
mid-height of the right panel. Reaching the level of displacement corresponding to 11D the
diagonal crack on the left panel continued to increase; up to a crack width of 6-8 mm. Significant
54 Experimental and numerical seismic performance of strong clay masonry infills
damage of the outer shells on the compressed diagonal has occurred. Diagonal cracking of the
compressed corners of the right panel has also extended further, but no continuous strut through
the entire diagonal length has formed. The opening of the crack in the mortar bed joint above the
first course of the right panel has increased further.
(a)
(b)
Figure 4.26. Observed crack pattern - Infilled frame TA4 (a) 0.50% drift (09D); (b) 1.00% drift (12D)
At the last level of loading (12D), corresponding to a drift of 1.00% (see Figure 4.26b), the diagonal
crack on the substantially damaged left panel has widened additionally, while in the right panel no
complete diagonal strut has formed. The damaged specimen after having sustained the maximum
level of drift is shown in Figure 4.27. Details of the damage in the bottom corners next to the
opening during the test at the last level of displacement (12D) for the left and the right panel are
shown in Figure 4.28.
P. Morandi, S. Hak, G. Magenes EUCENTRE 55
Research Report
Figure 4.27. Infilled frame TA4 during the in-plane test after 12D loading (1.00% drift)
(a) (b)
Figure 4.28. Details of in-plane damaged TA4 infill corners after 12D loading (1.00% drift) at the bottom (next
to opening): (a) Left panel; (b) Right panel
(f) Remarks. Following the in-plane cyclic test on the bare frame it has been noted that due to the
test conditions, including the precompression and consequently a substantial strength increase of
the RC beam, some cracks started to develop in the upper parts the column elements.
Nevertheless, the clear formation of a weak beam, strong column response mechanism has finally
been obtained, in line with the expected behaviour imposed according to the design approach
followed in modern seismic code provisions.
Based on the experimental test results and the related infill damage observed during the test for
the fully infilled frame configurations, a typical masonry infill failure mechanism for the considered
strong type of infill could be identified. Specifically, the initial damage is characterised by a
diagonal stepwise crack pattern, from the corners towards the centre of the panel, attained through
some extent of shear sliding in the bed joints, which due to slight opening of the dry head joints of
the tongue and groove system can be achieved without significant cracking of the masonry blocks.
With increasing drift demands, such behaviour is followed by the development of diagonal cracks
on the units in the corners of the infill, while subsequently the damage is spreading throughout the
panel. Due to the relatively high resistance of the strong infill, the creation of diagonal shear cracks
on the RC columns was expected, caused by the activation of a diagonal strut action, stressing the
importance to prevent possible shear failure mechanisms through adequate measures in the RC
frame design.
In the case of partial infill, a strongly asymmetric response has been obtained, showing substantial
damage and the creation of a full diagonal strut action in one of the two panels (left-hand side) for
56 Experimental and numerical seismic performance of strong clay masonry infills
the push loading direction, while the right panel after the creation of a significant horizontal crack at
the bottom tended to show the effect of uplift. For the opposite direction of loading the damage was
by far less pronounced and the diagonal strut action in the right panel has not been completely
activated.
During unloading, a minimum force of ca. 5.0 kN has been maintained with the aim to ensure
complete contact between the loading system and the panel. Clearly, at increasing target levels,
during the unloading phase, increasing residual displacements have been obtained. For each level
of loading (target force or displacement) three complete loading cycles have been carried out and
the duration of load application has been kept approximately constant.
(b) Instrumentation. In order to measure the out-of-plane displacements of the tested masonry
panel and to control possible displacements of the surrounding RC frame and the foundation of the
specimen during the out-of-plane test, displacement transducers (linear potentiometers) have been
adopted, as summarised in Table 4.5.
58 Experimental and numerical seismic performance of strong clay masonry infills
(a)
(b)
Figure 4.29. Infilled frame specimen instrumentation (TA1_OP, TA2_OP & TA3_OP): (a) Front view; (b)
Back view
60 Experimental and numerical seismic performance of strong clay masonry infills
(a)
(b)
Figure 4.30. Infilled frame specimen instrumentation (TA4_OP): (a) Front view; (b) Back view
P. Morandi, S. Hak, G. Magenes EUCENTRE 61
Research Report
(a)
(b)
Figure 4.31. Infilled frame specimen instrumentation (TA5_OP): (a) Front view; (b) Back view
In total, 30 potentiometers have been used for the fully infilled frame specimens (TA1, TA2 and
TA3), 38 for the partially infilled configuration (TA4), and 28 for the infill strip (TA5). The typical
layout of the displacement transducers adopted for the different infilled frames is schematically
represented in Figure 4.29, Figure 4.30 and Figure 4.31. The out-of-plane displacements of the
masonry infill have been measured by means of 20 displacement transducers (1, 2, 3, 4, 5, 6, 7, 8,
9, 10, 11, 12, 13, 14, 15, 16, 17, 32, 33, 34) for the full infill, 26 displacement transducers (1, 2, 3,
4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26) for the partial infill,
and 18 displacement transducers (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18) for the
infill strip, distributed symmetrically throughout the panel. The centreline out-of-plane
displacements of the RC frame have been monitored using two potentiometers in the beam-
column centreline intersections (18 and 19, 33 and 34, 26 and 27, respectively). For the fully
infilled frame three additional instruments within the beam span (20, 21, 22) were adopted, while
one potentiometer at the centre of the beam was used for the partial infill (35) and the infill strip
62 Experimental and numerical seismic performance of strong clay masonry infills
(28). One instrument was placed at mid-height of each column for all infill configurations (23 and
24, 36 and 37, 21 and 22, respectively). Possible movements of the specimen foundation have
been measured using three potentiometers for the fully infilled configuration (25, 26, 27) and the
infill strip (23, 24, 25), while for the partial infill one potentiometer (38) was used. Vertical
displacements of the beam were measured in six points for the partially infilled configurations (27,
28, 29, 30, 31, 32) and in two points for the infill strip (19, 20).
Figure 4.34. Infill TA1 during the out-of-plane test at 09D loading amplitude (75 mm displacement)
(a) (b)
Figure 4.35. Details of in-plane & out-of-plane damaged infill TA1 after 09D loading (75 mm displacement):
(a) Diagonal infill crack; (b) Column-infill interface
P. Morandi, S. Hak, G. Magenes EUCENTRE 65
Research Report
Figure 4.36. Crack pattern (in black – out-of-plane damage; in grey – previous in-plane damage) of infill TA1
after the out-of-plane test
(b) Infilled Frame TA2_OP. For the infilled frame specimen TA2 the force-controlled loading
cycles have been omitted and nine different target displacement levels (01D, 02D, 03D, 04D, 05D,
06D, 07D) up to a maximum displacement of 75 mm have been adopted. The complete
displacement history is shown in Figure 4.37. The obtained corresponding out-of-plane force-
displacement response is shown in Figure 4.38.
Figure 4.39. Infill TA2 during the out-of-plane test at 09D loading amplitude (75 mm displacement)
(a) (b)
Figure 4.40. Deformed shape of infill TA2 during the out-of-plane test at 09D loading amplitude (75 mm
displacement): (a) Front view; (b) Back view
68 Experimental and numerical seismic performance of strong clay masonry infills
Figure 4.41. Crack pattern (in black – out-of-plane damage; in grey – previous in-plane damage) of infill TA2
after the out-of-plane test
(c) Infilled Frame TA3_OP. The infilled frame specimen TA3 has been loaded at two different
force-controlled loading levels (assuming 0.10Fmax ≈ 15.0 kN, 0.20Fmax ≈ 30.0 kN) and
subsequently at nine different target displacements (01D, 02D, 03D, 04D, 05D, 06D, 07D, 08D,
09D), up to a maximum displacement of 75 mm. The complete displacement history is shown in
Figure 4.42. The obtained corresponding out-of-plane force-displacement response is shown in
Figure 4.43.
During the first force-controlled level of loading (01F) no new damage on the infill has been
observed, while after the second force-controlled cycle (02F) a small residual displacement has
been noted at the beam-infill interface. This displacement has increased slightly at the following
displacement-controlled level of loading (01D). Subsequently, at 02D some noise has been noted
and slight strength degradation has occurred at the first loading cycle. The drift at the interface
between the infill and the beam continued to increase and a light crack has been observed at the
bottom of the infill above the foundation. At the subsequent level of loading (03D), the out-of-plane
deformation has increased and residual drifts between the infill and the surrounding frame have
become more pronounced. An arching action mechanism has apparently started to form. Some
small cracks have formed around one of the plates used for the load transfer. Horizontal cracks in
the mortar bed joints have become clearly visible, in particular in the joint at mid-height of the
panel. Some shells of masonry blocks previously damaged in-plane have detached and fallen off.
At the 04D loading cycles the deformations have increased notably showing clearly the activation
of the arching action. The crack at the central joint has increased further. The infill has started to
separate from the columns causing also the detachment of small concrete pieces, particularly at
mid-height of the right column.
P. Morandi, S. Hak, G. Magenes EUCENTRE 69
Research Report
has been observed. A significant drop in the resistance of the panel has occurred during and in
particular after the first loading cycle towards the maximum displacement. The deformed shape
during the test at the maximum level of out-of-plane displacement (75 mm) is shown in Figure 4.44
and Figure 4.46, including the damage pattern due to previous in-plane loading, while details of the
damaged specimen are illustrated in Figure 4.45.
Figure 4.44. Infill TA3 during the out-of-plane test at 09D loading amplitude (75 mm displacement)
(a) (b)
Figure 4.45. Details of in-plane & out-of-plane damaged infill TA3 after 09D loading (75 mm displacement):
(a) Diagonal infill crack; (b) Local damage at force introduction
P. Morandi, S. Hak, G. Magenes EUCENTRE 71
Research Report
Figure 4.46. Crack pattern (in black – out-of-plane damage; in grey – previous in-plane damage) of infill TA3
after the out-of-plane test
(d) Infilled Frame TA4_OP. Following the given loading history scheme (Table 4.4), the two
panels of the infilled frame specimen TA4 have been loaded contemporarily through the system for
horizontal load introduction. Firstly, two different force-controlled loading levels (assuming a total
force of 0.10Fmax ≈ 10.0 kN and 0.20Fmax ≈ 20.0 kN) have been imposed on the specimen.
Subsequently the infill has been tested at seven different target displacements (02D, 03D, 04D,
05D, 06D, 07D, 08D), up to a maximum displacement of 40 mm. The complete displacement
history is shown in Figure 4.47. The obtained out-of-plane force-displacement curves are shown in
Figure 4.48 and Figure 4.49 for the left and for the right panel, respectively.
During the force controlled cycles no new damage could be noted on the infill panels previously
damaged in-plane. At the first displacement-controlled level of loading (02D) a horizontal crack in
the first mortar joint below the centre of the infill, between the fifth and the sixth course, has formed
on both panels through about one quarter of the masonry thickness. The creation of an arching
mechanism appeared to be initiated.
Figure 4.48. Out-of-plane test results - Infilled frame TA4_OP, LEFT panel
Figure 4.49. Out-of-plane test results - Infilled frame TA4_OP, RIGHT panel
Subsequently, at 03D, the crack between the fifth and the sixth course has widened, while at 04D
a crack has formed in the joint at mid-height of the right panel. The horizontal crack on the left
panel, which has sustained some residual displacement, has also widened. Remarkably, the left
panel is already reaching significantly larger displacements, also caused by the creation of a
wedge due to the previous in-plane damage, which was significantly more pronounced than on the
right panel. On the back side of the specimen corresponding to the left panel a breaking of the
damaged wedge through the infill wall could actually be observed. At the following loading cycle
(05D) a stepwise crack pattern has formed at the free boundary of the infill, next to the opening,
starting from the crack in the horizontal joint. Previous cracks have continued to increase further
and the breakthrough of the left panel has become more pronounced. During the subsequent level
of loading (06D) the existing damage has developed further. At the 07D displacement level, high
residual displacements have been recorded and a significantly asymmetric response could be
observed; the left panel has sustained much higher displacements than the right panel. Reaching
the final level of loading (08D) corresponding to a displacement of the actuator equal to 40 mm,
significant out-of-plane damage has been reached and a resulting maximum displacement of close
to 70 mm on the left panel has been recorded.The damaged specimen during the test after the
maximum level of out-of-plane loading (40 mm of actuator displacement) is shown in Figure 4.50,
while the corresponding deformed shape at the maximum amplitude is illustrated in Figure 4.51.
P. Morandi, S. Hak, G. Magenes EUCENTRE 73
Research Report
Figure 4.50. Infill TA4 during the out-of-plane test after 09D loading (40 mm actuator displacement)
(a) (b)
Figure 4.51. Deformed shape of infill TA4 during the out-of-plane test at 09D loading amplitude (40 mm
actuator displacement): (a) Front view; (b) Back view
(e) Infilled Frame TA5_OP. Following the given loading history scheme (Table 4.4), the central
vertical infill strip of the infilled frame specimen TA5 has been loaded at two different force-
controlled loading levels (assuming 0.10Fmax ≈ 15.0 kN, 0.15Fmax ≈ 30.0 kN ) and subsequently at
ten different target displacements (01D, 02D, 03D, 04D, 05D, 06D, 07D, 08D, 09D, 10D), up to a
maximum displacement of 75 mm. The complete displacement history is shown in Figure 4.52,
while the obtained corresponding out-of-plane force-displacement response for the vertical infill
strip is shown in Figure 4.53.
74 Experimental and numerical seismic performance of strong clay masonry infills
an inclined sliding surface has formed through the masonry blocks of the bottom course, reaching
from the joint above the first course at the back side of the panel towards the interface between the
infill and the foundation at the front side of the specimen. Subsequently, at 09D the damage
continued to propagate further. The displacement at the bottom of the infill along the crack in the
first course has become even more evident due to notable horizontal sliding at the back side of the
specimen and vertical compression at the front side. The upper part of the panel seemed to be
braced at the interface between the infill and the beam. In fact, first signs of damage could also be
observed on the concrete surface. Reaching the final loading level, corresponding to a
displacement of 75 mm, the outer shells of some blocks in the first course have detached and the
front part of the blocks appeared to be damaged due to compression failure, in correspondence
with the observed creep displacement at the back side of the panel and the damage of concrete in
the region of support. The damaged specimen during the out-of-plane test at the maximum level of
out-of-plane displacement (75 mm) and corresponding details of the region at the bottom and at
the centre of the panel are shown in Figure 4.54.
(c)
Figure 4.54. Infill TA5 during the out-of-plane test at 09D loading amplitude (75 mm displacement): (a)
Deformed shape; (b) Region of hinging at the centre; (c) Region of support at the bottom
(f) Remarks. Based on the out-of-plane cyclic tests on the masonry infill panels previously
damaged in-plane, an initial tendency of the infill to overturn has been observed, followed by the
creation of an evident vertical and horizontal arching action resistance mechanism, once complete
compression has been achieved at the top of the infill. Related to the damage pattern, the
formation of a principal horizontal crack at the central mortar bed joint has been noted in all cases
as well as the creation of diagonal, predominantly stepwise cracks from the corners towards the
central part of the panel. Significant loss of outer masonry shells has been noted, in particular
where considerable in-plane damage had previously been recorded. Even though significant local
damage in the region of horizontal load transfer has occurred at higher levels of imposed
displacements, no local failure of the masonry has been observed and the global response of the
panel could be adequately captured. Some damage has been observed at the corners of the
concrete columns, due to the detachment and rotation of adjacent masonry units. Testing an
undamaged vertical masonry strip in pure vertical bending, the clear formation of a single-bending
arching action mechanism could be obtained, showing well distinguished regions of support at the
contact of the infill with the foundation at the bottom and close to the adjacent beam in the upper
part. Below mid-height between the supports the central region of hinging could be identified.
P. Morandi, S. Hak, G. Magenes EUCENTRE 77
Research Report
(a) (b)
Figure 5.1. Average experimental response: (a) Strong infill without and with opening; (b) Strong and slender
infill without opening
78 Experimental and numerical seismic performance of strong clay masonry infills
The approach adopted for the representation of slender infills by means of a simplified numerical
model for nonlinear static analyses, based on a constitutive rule assigned to a single strut model,
and the corresponding definition of performance levels in function of attained in-plane drift limits
(Hak et al. [2013b]), in principle, appears to be applicable also to the case of strong infill.
Accordingly, taking into account the infill damage propagation observed during the test,
performance levels for a single strong masonry infill without and with opening have been defined
and compared to the case of slender infill, as illustrated in Figure 5.2 and summarised in Table 5.1
.
(a) (b)
Figure 5.2. Simplified masonry infill response: (a) Strong infill without and with opening; (b) Strong and
slender infill without opening
Table 5.1 Comparison of performance levels for single strong and slender masonry infills
Limit State Operational Damage Limitation Ultimate
Reference O–A A–B B–C
Drift
δ ≤ 0.30 0.30 < δ ≤ 0.50 0.50 < δ ≤ 1.75
without opening
[%]
Damage
STRONG INFILL
Drift
δ ≤ 0.20 0.20 < δ ≤ 0.35 0.35 < δ ≤ 1.00
[%]
with opening
Damage
Drift
SLENDER INFILL
[%]
Damage
Specifically, a drift limit equal to 0.50% has been assigned to the attainment of damage limitation
limit state conditions, supported also by the recent findings related to another type of strong infill,
tested at the University of Padova, as reported by da Porto et al. [2012]. Given that for the
P. Morandi, S. Hak, G. Magenes EUCENTRE 79
Research Report
operational limit state a drift limit equal to 2/3 of the value for damage limitation is commonly
assumed in seismic design provisions (e.g. NTC08 [2008]), in accordance with the corresponding
damage pattern, the conservative assumption of a drift equal to 0.30% has been proposed herein.
Furthermore, supported by the observations during the test, a drift of 1.75% may be reached
before exceeding the ultimate limit state conditions. Such drift corresponds exactly to a strength
degradation of 20%, which is commonly assumed also for the bilinear idealisation of the response
of load-bearing masonry walls (see e.g., Morandi et al. [2013a]).
For the strong infill typology investigated at the University of Padova (da Porto et al. [2012]), a
somewhat lower drift limit has been defined for the ultimate limit state (equal to 1.20%); such
conclusion may however be a conservative choice, arising from the fact that no drift levels beyond
1.20% have been achieved during that experimental campaign. Comparing the obtained drift limits
for the strong type of infill to the corresponding values evaluated in the past for the slender clay
masonry infill typology, it can be observed that the ratio of drifts corresponding to damage
limitation and ultimate limit state conditions is approximately preserved. Similarly, for the strong
infill with opening drift values of 0.20%, 0.35% and 1.0% have been assigned to operational,
damage limitation and ultimate limit state conditions, respectively.
(a) (b)
Figure 5.3. Crisafulli [1997] stress-strain model: (a) Monotonic response; (b) Cyclic response
The cyclic deterioration is based on the linear projection of several fixed tangent moduli and a
plastic strain index with an exponent term to describe the rate of cyclic deterioration with the
accumulation of damage. It is assumed that the equation originally proposed for concrete (Sargin
et al. [1971]) can approximately represent the envelope curve for masonry, according to Equation
(5.1), where fm’ is the maximum compressive strength and εm’ the corresponding strain. The
constants k1 and k2 are functions of the mechanical properties of the material.
2
ε ε
k1 m + (k2 − 1) m
εm ' εm '
f (ε m ) = f m ' 2 (5.1)
ε ε
1 + (k2 − 2 ) m + k2 m
εm ' εm '
The descending branch of the strength envelope can be alternatively represented by a parabolic
curve, as assumed also within the scope of this study, according to Equation (5.2), where εu is the
ultimate strain corresponding to zero strength.
ε − ε ' 2
f (ε m ) = f m ' 1 − m m
(5.2)
u ε − ε m'
In line with the recommendations by Carr [2007], to better match the post-peak behaviour, a large
ultimate strain (on the order of 20 times εm’) allows for a smooth decrease of the compressive
stress (Figure 5.4b). Without the strength degradation model contained within cross-sectional area
degradation (Figure 5.4a), the stress of the backbone curve may remain almost constant after the
peak, resembling perfect plasticity and the stress-strain constitutive relation alone may not always
adequately simulate the infill degradation. The conditional statement for the definition of strut area
is shown in Equation (5.3), where d1 and d2 are the displacements which bound the transition
between initial and final values of strut area, while δ(ε) is the inter-storey drift corresponding to a
strain equal to ε and h is the storey height. All hysteresis rules subsequent to the backbone curve
follow a curvilinear path with a prescribed initial slope from the previous step to a projected point
with a prescribed terminal slope.
A1
, δ (ε m )h < d 1
A1 − A2
A (ε m ) = A1 − (d 1 − δ (ε m )h ) , d 1 < δ (ε m )h < d 2 (5.3)
d1 − d 2 , δ (ε m )h > d 2
A2
P. Morandi, S. Hak, G. Magenes EUCENTRE 81
Research Report
(a) (b)
4 0 0
F (ε )
1
A (ε ) k N3 0 0
m
2 Fs( ε ) 2 0 0
0 .5 k N
1 0 0
0 0
0 0 . 0 00 2. 0 00 4. 0 00 6. 0 0 8 0 0 . 0 002. 0 004. 0 006. 0 0
Figure 5.4 (a) Degradation of strut cross section area; (b) Axial force – strain relationship (Oilaee et al.
[2015a])
(c) Evaluation of Model Parameters. As a first step towards the evaluation of the infill model
parameters, the shape of the expected envelope curve assigned to the strong masonry infill
typology without and with opening has been assumed in line with the simplified interpretation of the
obtained average experimental response (for specimens TA2 and TA4, respectively, as illustrated
in Figure 5.2). Accordingly, for the experimentally obtained values of drift assigned to peak infill
resistance, i.e. 0.50% and 0.35%, respectively, the axial strain corresponding to the diagonal has
been evaluated based on simple geometric relations (see Hak et al. [2013b]), resulting in the
characteristic strain parameter εm’, being equal to 0.0023 and 0.0016 for the strong infill and the
strong infill with opening.
In order to obtain the cyclic experimental response of the infill only, the difference between the
available response of the bare frame (TNT) and the infilled frames (TA2 and TA4) has been
evaluated, considering the strength at available coinciding values of target displacement (Oliaee et
al. [2015a]). To verify the adopted numerical model of the bare frame, a comparison of numerical
and experimental results has been accomplished as illustrated in Figure 5.5.
Figure 5.6 Comparison of numerical and experimental cyclic response: Strong infill (TA2, Oliaee et al.
[2015a])
Figure 5.7 Comparison of numerical and experimental cyclic response: Strong infill with opening (TA4,
Oliaee et al. [2015a])
P. Morandi, S. Hak, G. Magenes EUCENTRE 83
Research Report
Figure 5.8 Comparison of numerical and experimental strength envelope: Strong infill (TA2) and strong infill
with opening (TA4; Oliaee et al. [2015a])
Subsequently, to finalise the estimation of the strength envelope assigned to the numerical model,
the obtained experimental infill strength was inspected and the post-yield softening stiffness was
obtained by visually selecting landmark target cycles. Two parameters defining the hysteretic
response have been studied in particular, with the aim to capture the unloading and reloading
behaviour, and to match the total energy dissipation as closely as possible. In particular, for both
types of infill, i.e. for strong infill and strong infill with opening, the unloading stiffness factor γun,
controlling the slope of the unloading branch has been found equal to 5.0, while the reloading
strain factor αre, which defines the point where the reloading curve reaches the strength envelope,
is set equal to 0.3. In addition, the displacement corresponding to the onset of strut area
degradation d1, as well as to the beginning of the second strength plateau d2 has been evaluated,
being equal to approximately 45 mm and 60 mm, respectively, in the case of strong infill without
opening. A ratio between A2 (final degradated cross section area of the strut) and A1 (initial cross
section area) of about 70% has been assumed. Since for the strong infill with opening lower values
of displacement have been achieved during the test and the development of the strength plateau
could not be that clearly distinguished, regarding the degradation of the strut area the same
parameters as in the case of strong infill without opening have been assumed. However, the
reduced deformation capacity could successfully be captured through the adopted value of strain
at maximum strength εm’.
84 Experimental and numerical seismic performance of strong clay masonry infills
Figure 5.9 Comparison of numerical and experimental response: Infilled frame - Strong infill (TA2, Oliaee et
al. [2015a])
Figure 5.10 Comparison of numerical and experimental response: Infilled frame - Strong infill with opening
(TA4, Oliaee et al. [2015a])
P. Morandi, S. Hak, G. Magenes EUCENTRE 85
Research Report
Figure 5.11 Comparison of numerical and experimental strength envelope: Infilled frame - Strong infill (TA2)
and strong infill with opening (TA4, Oliaee et al. [2015a])
The adopted approach clearly relies on the assumption that the response of the infilled frame
ideally corresponds to the summation of the bare frame response and the corresponding response
of the infill only, even though more rigorous considerations should account for the possibility that
the infill presence may modify significantly also the response of the frame. Nevertheless, for the
given infill properties, satisfactory results have been obtained regarding the overall response of the
infilled frame configurations, as illustrated in Figure 5.9 and Figure 5.10, respectively for the strong
infill typology (TA2) and the strong infill with opening (TA4). A corresponding comparison of
experimentally and numerically obtained maximum force displacement envelopes for both infilled
frames is shown in Figure 5.11.
presence of infills. In addition, it has to be stressed that the code requirement given in Eurocode 6
– Part 1 (CEN [2004b]), commonly applied to load-bearing masonry with dry head-joints, to reduce
the initial shear strength under zero compressive stress fv0k by one half, has been neglected in the
approximation of the shear strength, in order to accomplish a conservative shear strength
verification. Accordingly, with the aim to satisfy all code requirements (CEN [2004a]) and exclude
any possibility of column shear failure during the test, the columns of the RC frame specimen were
dimensioned to possess a shear resistance greater than the estimated horizontal infill strength,
supposed to represent the design column shear demand. Assuming characteristic material
properties for steel and concrete, the RC column of the specimen has been found to have a shear
strength equal to about 460.0 kN.
Furthermore, in order to interpret the experimental response of the specimen and to assess the
level of attained concrete shear stresses with respect to acceptable values, based on the
considerations given in Eurocode 2 – Part 1-1 (CEN [2004c]) for plain and lightly reinforced
concrete structures, assuming characteristic values of concrete tensile and compressive strength,
a corresponding “cracking” shear stress equal to 3.20 MPa has been found. For the column such
level of stress, which may be associated with the occurrence of shear cracks, but does not
necessarily point towards the development of shear failure in the RC section, corresponds to a
shear force of about 260.0 kN. Apparently, this level of horizontal strut action in the masonry infill
has been exceeded during the test, imposing the corresponding shear demand on the column, but
the actual shear resistance of the RC section has not been reached. A summary of shear forces
evaluated to verify the resistance and interpret the behaviour of the RC columns is given in Table
5.2.
Table 5.2 Comparison of additional column shear demand and shear resistance
Design column Design column Experimental Cracking column
shear demand shear resistance horizontal infill shear resistance
[kN] [kN] strength [kN] [kN]
428.3 460.0 350.0 260.0
Observations related to the frame response during the tests evidently support the presence of a
substantial level of shear stresses in the RC columns of the infilled frame specimens (TA1, TA2
and TA3), but even so confirm the absence of shear failure. As a matter of fact, referring to
specimen TA2, very first light cracks in the upper part of the columns have been observed at inter-
storey drifts of 0.30% (6D, Figure 5.12a), which according to the response of the infill only (see
Figure 5.6 and Figure 5.8) may correspond to a level of horizontal infill strut action higher than
260.0 kN, imposing on the columns shear stresses above the allowable level for plain concrete. A
more substantial development of shear cracks has been observed starting from target drifts of
0.80% (11D) or 1.00% (12D, Figure 5.12b), which actually correspond to the attainment of the
peak infill strength. Nevertheless, with further increasing levels of drift (e.g. 15D, Figure 5.12c and
17D, Figure 5.12d), no overly significant expansion of shear cracking has been observed and the
response of the RC frame has remained satisfactory, without considerable further signs pointing
towards the development of column shear failure.
P. Morandi, S. Hak, G. Magenes EUCENTRE 87
Research Report
(a) (b)
(c) (d)
Figure 5.12 Shear crack development at increasing drift levels: (a) 0.30%; (b) 1.0%; (c) 1.5%; (d) 2.5%
Figure 5.13. Deflection of fully infilled specimens at 50 mm target displacement: (a) TA2, (b) TA1, (c) TA3
and at 75 mm target displacement: (d) TA2, (e) TA1, (f) TA3
88 Experimental and numerical seismic performance of strong clay masonry infills
Moreover, the envelope of the cyclic force-displacement response obtained for the fully infilled
specimens TA1, TA2 and TA3 is presented in Figure 5.14a, b and c, respectively, while a
comparison of the force-displacement envelopes is shown in Figure 5.14d.
(a) (b)
(c) (d)
Figure 5.14. Force-displ. response: (a) TA1, (b) TA2, (c) TA3, (d) Comparison of TA1, TA2 and TA3
The results demonstrate clearly the significant degradation of stiffness and strength in the out-of-
plane direction for increasing values of previous in-plane drift and related damage, in particular
related to relatively high levels of drift and extensive preceding damage. Specifically, for
specimens TA3 and TA1, which were previously subjected to in-plane drifts of 1.0 % and 1.5 %,
respectively, similar values of peak resistance have been obtained (equal to 163.9 kN and 168.5
kN), but a significant stiffness degradation, equal to approximately 40 %, has occurred. For
specimen TA2, which had sustained an in-plane drift of 2.5 %, the peak resistance has dropped to
102.7 kN, in addition to a reduction of stiffness equal to about 75 %, with respect to the stiffness of
specimen TA3.
As the mass of the infill panel is equal to about 3.9 ton, the values of out-of-plane strength reported
above correspond to equivalent accelerations of 4.4g, 2.7g and 4.3g for specimen TA1, TA2 and
TA3 respectively; therefore, these values of peak strength indicate that for the considered strong
masonry infill typology, adherent to the RC frame, the out-of-plane stability does not present a
critical issue.
(b) Single Bending. Considering that the out-of-plane strength of the undamaged strong infill,
based on the results obtained for specimen TA5, is available for the case of single-bending
conditions, the correlation of single and double-bending response mechanisms may be envisaged
within the scope of future investigations. Expecting that higher values of strength are obtained for
the case of double-bending, a safe-sided estimation of the undamaged overall strength may be
obtained extrapolating the results corresponding to the 1.38 m wide infill strip for the case of the
4.22 m wide infill panel. Accordingly, given the maximum resistance of 65.6 kN obtained for
specimen TA5, the out-of-plane strength of a fully infilled test frame without any previous damage,
conservatively assuming single-bending conditions, may be estimated being equal to 200.5 kN,
which corresponds to an equivalent acceleration of 5.3g. However, it has to be emphasised that for
the actual case of double-bending response due to the constraining boundary conditions on all four
sides of the infill, a higher strength may be expected, resulting in a more pronounced strength
reduction caused by previous in-plane damage.
P. Morandi, S. Hak, G. Magenes EUCENTRE 89
Research Report
The cyclic force-displacement response and the corresponding envelope obtained for specimen
TA5, i.e. a vertical masonry strip under single-bending conditions, is shown in Figure 5.15a.
Corresponding to a number of selected characteristic points on the force-displacement envelope,
the deflection of the specimen along the height, presented in Figure 5.15b, has been evaluated as
the average of the displacements measured by means of the potentiometers installed during the
test (see Figure 4.31a). The related progressive propagation of damage during the test is
illustrated in Figure 5.16a for a cross section view of the specimen, while the damaged specimen
during the test at the maximum level of imposed out-of-plane displacement (75 mm) is shown in
Figure 5.16b, illustrating the creation of the vertical arching response mechanism.
Based on the damage observed during cyclic loading at the top and at the bottom of the infill, in
addition to the creation of a dominant horizontal crack due to the opening of the masonry bed joint
below mid-height, the creation of a vertical arching mechanism could be clearly distinguished. The
response mechanism was apparently achieved through the equilibrium of two rigid bodies, with an
upper support at the interface between the infill and the beam, a bottom support at the crack of the
first course and the intermediate hinge of the arch at the crack in the mortar joint above the fifth
course.
(a) (b)
Figure 5.15. Specimen TA5: (a) Force-displacement response; (b) Deflection along the height
(a) (b)
Figure 5.16. Specimen TA5: (a) Progression of damage; (b) Damage at 75 mm target displacement
As commonly acknowledged (e.g. Angel et al. [1994]), unreinforced masonry panels restrained by
bounding frames can develop a significant out-of-plane resistance due to the formation of an
arching mechanism, depending greatly on the slenderness ratio of the panel, and on the
90 Experimental and numerical seismic performance of strong clay masonry infills
compressive strength of the masonry. Accordingly, for masonry walls built between rigid supports
that restrain outward movement of any part of the wall in the plane of the wall (Figure 5.17a), axial
compressive forces are induced as the wall bends. These in-plane compression forces can delay
cracking and, following cracking, can produce an arching action that in many cases has several
times the capacity of the masonry in pure flexure (Drysdale et al. [1999]). Experimental
investigations (e.g. McDowall et al. [1956]) have shown that, under certain conditions, significantly
larger loads can be sustained than predicted on the basis of conventional bending analysis. The
application of such resistance mechanism for the evaluation of the out-of-plane capacity of
undamaged unreinforced masonry infills appears to be appropriate when the panel is built in full
contact with the surrounding frame. Resistance models based on full vertical arching action, as for
instance described in Eurocode 6 – Part 1-1 (CEN [2004b]), are usually applied to load-bearing
elements subjected to non-seismic actions (wind or gravity loads). Nevertheless, they may be
considered appropriate also for the simplified analysis of undamaged infills, assuming that
masonry walls built solidly between supports are capable of resisting an arch thrust and that an
arch in the relevant direction develops within the thickness of the wall.
(a) (b)
Figure 5.17. (a) Assumption of uniformly distributed out-of-plane action; (b) Formation of out-of-plane arching
mechanism
The calculation may be based on a three-pin arch with a bearing of the arch thrust at the supports
and at the central hinge assumed equal to 10% of the wall thickness tw (CEN [2004b]). The vertical
component of the maximum compressive force Pu acting at each support and at the top of the arch
(see Figure 5.17b) can be calculated from Equation (5.4), where Lw is the length of the wall and fd
= fwv the compressive strength of masonry in the direction of the arch thrust, i.e., the vertical
compressive strength.
Pu = 0.10t w L w f d (5.4)
The corresponding arch rise e, or the lever arm of the couple of forces providing arching action,
can be found from Equation (5.5), where Θ is the inclination angle of each half-height panel and hw
is the height of the panel. For panels having a height to thickness ratio equal to 25 or less, the
deflection of the arch under lateral loads may be assumed close to zero (for Θ ≈ 0, sin Θ ≈ 0 and
cos Θ ≈ 1.0), and the simplified moment of resistance MR can be determined from Equation (5.6),
resulting in a resistance in terms of lateral pressure wR expressed by Equation (5.7).
Figure 5.18. (a), (b) Application of the loads in the single bending out-of-plane test, and c) Forces acting in
the upper part of the panel
In order to check the validity of the simplified expressions reported above, the force-displacement
curve has been turned in terms of moment – displacement response, considering the loads applied
during the test (see Figure 5.18a,b), through the rotational equilibrium of the upper part of the
panel, considered rigid, around the point a (see Figure 5.18c). Applying Equation (5.6), being the
dimensions of the panel tw = 0.35 m and Lw = 1.38 m and the mean compression strength of the
masonry fd = 4.64 MPa, the resisting moment MR results to be equal to 70.6 kNm, rather close to
the maximum moment reached in the experimental test, as shown in Figure 5.19. This outcome
implies that the simplified expression based on the arching mechanism applied on this type of
strong masonry panel in single-bending condition appears to be appropriate, although slightly
conservative.
Figure 5.19. Moment-mid-height displacement response of the panel subjected to single bending
92 Experimental and numerical seismic performance of strong clay masonry infills
Figure 5.20. Out-of-plane strength in function of in-plane drift: (a) Strong infill; (b) Strong infill with opening
P. Morandi, S. Hak, G. Magenes EUCENTRE 93
Research Report
(a) (b)
Figure 5.21. Experimental strength reduction coefficient: (a) Strong infill; (b) Strong infill with opening
P. Morandi, S. Hak, G. Magenes EUCENTRE 95
Research Report
Figure 6.1. Plan and elevation view of 3-story, 6-story and 9-story frame systems
96 Experimental and numerical seismic performance of strong clay masonry infills
The frames of interest are examined as planar structures, and such consideration is justified since,
according to the criteria for structural regularity, being regular in plan and elevation, spatial
structures can be analysed using a planar model. Each frame configuration has been newly
designed for five different levels of seismicity, corresponding to design peak ground accelerations
(PGAs) on ground type A ag, equal to 0.05g, 0.10g, 0.15g, 0.25g and 0.35g, and two different
ductility classes (medium and high) resulting in a total of 30 different bare frame types. In addition
to the bare frames, for the complete sequence of nonlinear analyses, considering firstly infills
without openings, a partially infilled (P) and a fully infilled (F) frame configuration are assumed as
illustrated in Figure 6.2 for the case of a 3-storey frame configuration. The case of partial infill
refers to the vertical distribution of masonry panels only in the middle span of the frame.
(a) (b)
Figure 6.2. (a) Partially infilled (P), and (b) Fully infilled (F) 2-D 3-story frame configurations
In addition, two configurations which include also infills with a full height central opening in the
external bays of 5.0 m span have been considered (see Figure 6.3); in particular, one configuration
with bare central bays and infills with opening (BO) as well as one configuration with infilled central
bays and infills with opening (PO) are assumed. For all prototype structures a constant infill
distribution along the height has been assumed.
(a) (b)
Figure 6.3. (a) Bare central bays & infills with opening (BO), and (b) Infilled central bays & infills with opening
(PO) 2-D 3-story frame configurations
In addition to gravity loads and different levels of seismic actions, evaluated corresponding to the
design PGAs of 0.05g, 0.10g, 0.15g, 0.25g and 0.35g for the 2-dimensional RC frames, the
structures have been designed for a certain level of wind loads, assumed with horizontal pressure
reaching values equal to 0.6, 0.8, and 0.95 kN/m2 at the top of the structure in the case of 3-, 6-,
and 9-story buildings, respectively.
The analysis of each case study structure and the evaluation of the relevant design action effects
have been carried out on simple linear-elastic bare structural models, described in Section 6.3.1,
using a commercial software for structural analysis. The modal response spectrum method is
adopted to account for seismic actions, with the modal characteristics of the structures identified
using eigenvalue analyses of the buildings.
(b) Seismic Actions. In order to account for the different levels of seismicity, performing modal
response spectrum analyses, seismic actions are represented by means of design response
spectral ordinates. To evaluate the possible amplifications of the seismic action due to geological
and topographic parameters, the relevant ground conditions are defined. For the objectives of this
numerical research, it appears suitable to consider the case study buildings founded on ground
type B, which represents a medium stiffness soil condition, following the identification of ground
types given in Eurocode 8 – Part 1 (CEN [2004a]). According to the Italian National Annex, a
topographic configuration of category T1, representing predominantly flat terrain with hills and
isolated peaks of average inclination below 15° is supposed. Since the prototype buildings
represent regular buildings for residential use, importance class II with a corresponding importance
factor γI = 1.0 is considered.
With the aim of defining the response spectra, five sites in Italy exposed to increasing levels of
seismicity have been selected as summarised in Table 6.2, and the corresponding site specific
acceleration response spectra, evaluated according to the Italian national code, have been
adopted for the design. The response spectra corresponding to the damage limitation and the
ultimate limit state have been evaluated using the program Seismic Actions – Response Spectra
ver. 1.03 (Italian: Azioni sismiche - Spettri di risposta ver. 1.03), released by the Italian Board of
Public Works (Italian: Consiglio Superiore dei Lavori Pubblici), accompanying the code. For the
needs of this study, the damage limitation limit state is associated with a seismic action having a
reference return period of TDLS = 50 years corresponding to a probability of exceedance PR = 63%
in 50 years, in accordance to the values set in the Italian national code. Somewhat different
recommendations are given in the general Eurocode 8 – Part 1 (CEN [2004a]); specifically, the
return period TDLS = 95 years corresponding to a probability of exceedance PR = 10% in 10 years is
suggested. To define the seismic action at the ultimate limit state, a reference return period of TULS
= 475 years is adopted, corresponding to a probability of exceedance PR = 10% in 50 years, in
agreement with both the Italian national code and the general recommendations in Eurocode 8.
The specific sites have been selected such that design peak ground acceleration ag = γIagR (agR is
the reference peak ground acceleration), evaluated on ground type A (i.e., rock or other rock-like
geological formation, according to Eurocode 8 – Part 1), equal to the increasing levels of seismicity
(i.e., 0.05g, 0.10g, 0.15g, 0.25g and 0.35g) are obtained at the ultimate limit state. Since the
buildings are assumed to be founded on ground type B, the seismic actions are increased by the
soil factor S (see Table 6.2).
98 Experimental and numerical seismic performance of strong clay masonry infills
Table 6.2. Peak ground accelerations at DLS and ULS for specific sites in Italy
Province Pavia Pavia Brescia L’Aquila Isernia
Municipality Borgo San Siro Brallo di Pergola Agnosine Pacentro Scapoli
Site S1 S2 S3 S4 S5
Ground type B B B B B
DLS
ag 0.023g 0.040g 0.054g 0.097g 0.142g
S 1.20 1.20 1.20 1.20 1.20
agS 0.028g 0.048g 0.065g 0.116g 0.170g
ULS
ag 0.05g 0.10g 0.15g 0.25g 0.35g
S 1.20 1.20 1.20 1.164 1.076
agS 0.060g 0.120g 0.180g 0.291g 0.377g
One exception has been introduced in order to define the highest level of seismicity, i.e., 0.35g on
ground type A at the ultimate limit state, since it exceeds the highest currently defined seismicity
on the Italian territory. Hence, given that there is no site with a peak ground acceleration equal to
0.35g corresponding to a return period of 475 years, for the site of Scapoli, Isernia the return
periods at the damage limitation and the ultimate limit state have been increased to TDLS = 100
years and TULS = 975 years, respectively, in order to achieve the desired level of seismic action.
The elastic acceleration and displacement response spectra for the specific sites normalised to agS
at both limit states are shown in Figure 6.4. The ratio of spectral ordinates of the response spectra
at the two considered limit states is found to vary between 0.35 and 0.45 in the case of ground
type A and between 0.35 and 0.50 in the case of ground type B in the region of dominant periods
of the considered structural configurations, as shown in Figure 6.5.
Figure 6.4. Normalised elastic acceleration and displacement response spectra for specific sites in Italy: (a)
at the DLS; (b) at the ULS
P. Morandi, S. Hak, G. Magenes EUCENTRE 99
Research Report
Figure 6.5. Ratio of spectral ordinates at the DLS and the ULS for specific sites in Italy: (a) on ground type A;
(b) on ground type B
For comparison, note that in the general recommendations of Eurocode 8 – Part 1 (CEN [2004a]),
the elastic spectrum for damage limitation requirements, introduced by means of the reduction
factor νDLS taking into account the lower return period (TDLS = 95 years with respect to TULS = 475),
has the same shape as the design seismic action for the ultimate limit state requirements. The
recommended values of νDLS are 0.40 for importance classes III and IV and 0.50 for importance
classes I and II.
The behaviour factor q for the definition of the corresponding design spectra is determined
following the code provisions, which in this case result in equal values for Eurocode 8 – Part 1
(CEN [2004a]) and NTC08 [2008]. For the frame systems q is equal to 3.90 for ductility class
medium and equal to 5.85 for ductility class high.
(c) Dimensioning, Detailing and Verification. In order to attain satisfactory behaviour under
earthquake induced horizontal loading, due to a controlled development of plastic hinges, the
structural elements are dimensioned for the design action effects evaluated from linear analyses to
comply with the given capacity design principles aimed to achieve structures with strong columns
and weak beams. For each ductility class, the corresponding specific provisions for earthquake
resistant design related to dimensioning and detailing are followed and the relevant verifications of
strength and ductility are performed at the ultimate limit state. Verifications of inter-story drift
limitations at the damage limitation limit state are accomplished with a linear analysis as required
by the code for the case of buildings having non-structural elements of brittle materials attached to
the structure (i.e., δDLS = 0.50%). Possible second-order P-Δ effects due to geometric nonlinearities
are included following the code recommendations based on the calculation of inter-story drift
sensitivity coefficients θP-Δ. The final sizes of structural elements for which all safety verifications
are fulfilled at the ultimate limit state and at the damage limitation state are summarised in Table
6.3 for the 2-dimensional prototype frames.
100 Experimental and numerical seismic performance of strong clay masonry infills
Table 6.3. Beam and column dimensions of the 2-dimensional RC frame systems
In the case of ductility class medium, following common design practise, for all case studies the
beam height was assumed to be equal to the overall slab thickness (i.e., equal to 24 cm).
However, according to the Italian national code, for ductility class high such choice is not permitted
and the relevant beam dimensions were increased, increasing also the overall stiffness of the
structure. The sizes of the elements were selected to satisfy minimum reinforcement ratios of 1.0%
in the columns and of 0.30% for the tension reinforcement in the beams. For frame systems in
regions of high seismicity the critical design requirement was typically the need to control the inter-
story drift limit equal to 0.50% for the damage limitation requirements whereas, for regions of low
seismicity, gravity and wind requirements normally governed. It can be noted that in the case of
pure frame systems when the beam height was assumed equal to the overall slab thickness, larger
column dimensions had to be adopted in the initial phase of the design in order to satisfy the
displacement verifications for higher levels of seismic action.
Further details related to dimensioning, detailing and design verifications of the 2-dimensional case
study frame systems assumed for the parametric study are given in Hak [2010], where however
only the general European code recommendations were followed, without additional reference to
the Italian Nation Code, and therefore somewhat different final design choices were adopted.
being regular in plan and elevation, can be analysed using a planar model. As commonly accepted
for elastic models used for force-based design approaches, rigid beam and column ends are
assumed in the joint panel zone (see Figure 6.6).
curve; for vertical elements the bilinear estimation is based on the moment-curvature relation
corresponding to the axial force occurring in the seismic load combination. As this approach can
lead to very low ratios of cracked to gross stiffness, significantly below current code
recommendations, lower bound values are introduced for the ratio of effective to elastic initial
stiffness: 0.30 for beams and 0.40 for columns and walls. These lower limits were set based on
engineering judgement since there is some uncertainty as to the best means of setting the cracked
stiffness of RC elements due to factors such as tension-stiffening and strain penetration. Finally,
values of effective initial stiffness equal to 25-35% and 35-45% of the elastic values are adopted
for beams and columns, respectively.
Figure 6.7. (a) Concentrated-plasticity frame member; (b) Modified Takeda hysteretic rule; (c) Moment-
curvature bilinear idealisation
Although damage occurring in beam-column joints of existing frame structures, designed for
gravity loads only, can significantly influence the overall structural behaviour and modify the
deformed shape of the structure subjected to lateral loads as reported e.g. by Pampanin et al.
[2002] and Calvi et al. [2002], in the present study, dealing with newly designed buildings,
assuming adequate detailing according to modern seismic codes, beam-column joint regions are
supposed not to cause major modifications of the seismic response. However, following the
recommendations by Priestley et al. [2007], a minimum acceptable joint representation is
introduced in the frame model to account for the contribution of beam-column joint flexibility to total
lateral displacements due to strain penetration and joint shear deformations. Hence, in the region
between the intersection of beam and column centrelines and the member end, where the lumped
plasticity is located, an elastic element of finite length is assumed.
The viscous damping for all nonlinear analyses is represented by an initial stiffness proportional
Rayleigh model verified through a sensitivity study with reference to the results of primary interest
(i.e., global displacement demands), as reported in Hak [2010]. Following the recommendations by
Priestley et al. [2007] the damping corresponding to the fundamental mode of vibration in the
direction of interest is reduced and values of 3.5% and 3.0% are adopted for medium and high
ductility, respectively, while 5.0% damping is assumed in the second mode.
(c) Masonry Infills. Different aspects of possible numerical modelling approaches for the
representation of non-structural masonry infill panels have been investigated in the past and the
suitability of different strategies has been discussed by several authors (e.g., Shing and Mehrabi
[2004]; Asteris [2008]). A wide range of models at different refinement levels, focusing both on
global behaviour and/or local effects is available for research purposes or practical applications.
Founded on early work (Polyakov [1960], Holmes [1961, 1963], Stafford Smith [1966, 1967],
Stafford Smith and Carter [1969], Mainstone [1971]), following the principle that a diagonal strut
with appropriate geometrical and mechanical characteristics is appropriate for the representation of
a masonry infill, a number of refined macro-models have been developed subsequently, as
summarised by Madan et al. [1997]. Further developments have been introduced e.g. by Crisafulli
[1997] and Kappos et al. [1998], while examples of more recent findings can be found in the work
by Puglisi et al. [2009a, 2009b], Rodrigues et al. [2010], Smyrou et al. [2011].
Simultaneously, several micro-modelling approaches have been developed founded on finite-
element methods, such as proposed by Liauw and Kwan [1984], Singh et al. [1998], Ghosh and
Amde [2001], Asteris [2008] and others. Based on extensive experimental and numerical
investigations Hashemi and Mosalam [2007] have introduced a strut-and-tie based masonry infill
model with the aim to account for contemporary in-plane and out-of-plane actions imposed on the
P. Morandi, S. Hak, G. Magenes EUCENTRE 103
Research Report
infill, while subsequent investigations carried out by Kadysiewski and Mosalam [2009] have
resulted in the formulation of a fiber-section interaction model.
Having in mind mostly the main objective of this numerical study, which is closely related to the
evaluation of overall displacement and inter-storey drift demands, but also the extensive number of
analyses to be performed, an adequately simple but yet reliable modelling strategy is envisaged.
Hence, within the scope of this work, an equivalent single-diagonal-strut model (i.e., a single strut
along each diagonal) for the representation of masonry infills is adopted, although clearly more
refined solutions are available. In such a model the compressive diagonal portion of the infill is
replaced with an equivalent compressive strut (see Figure 6.8a) pin-ended at the frame member
centreline intersections (see Figure 6.8b). Such modelling approach for the representation of infills
has been used also in previous numerical studies and has shown to successfully reproduce
experimental results when adopted in the structural analysis program Ruaumoko (Magenes and
Pampanin [2004]). In Figure 6.8 the length of the infill is denoted by Lw, hw stands for the infill
height, h represents the centreline storey height and L the horizontal distance between the
attachment points of the diagonal strut to the structural elements.
(a) (b)
Figure 6.8. (a) Equivalent diagonal single-strut model; (b) Strut element in frame model
The main limitation of the equivalent diagonal strut modelling approach is certainly its inability to
accurately capture possible unfavourable local effects on RC members due to the interaction
between structural frame elements and the masonry infill. However, since the focus here is on
damage to infills, which is essentially governed by inter-storey drift, the single strut model is
considered to be adequate.
Considering the numerous uncertainties related to properties of masonry as a material, the
possibility that different failure modes occur in the infill panels and the apparent simplifications
introduced in the model, relevant mechanical properties have to be assigned to the equivalent strut
with caution in order to ensure a realistic representation of the actual masonry infill behaviour. The
fundamental parameters to be defined are the stiffness and strength characteristics of the
equivalent strut and an axial strain-stress hysteretic rule describing the cyclic behaviour. The
thickness of the strut is typically assumed equal to the thickness of the infill tw. Commonly, the
equivalent strut width bw is expressed in terms of a bw/dw ratio, where dw is the diagonal length of
the infill. In this study for the evaluation of bw, the model proposed by Decanini et al. [1993], given
by Equation (6.1), is adopted.
bw K 1
= + K2 (6.1)
d w λh
The parameters K1 and K2 are defined as functions of the product λh, as summarised in Table 6.4,
with h being the centreline height of the frame and λ the dimensionless relative stiffness
parameter, given by Equation (6.2), introduced by Stafford Smith [1966], Stafford Smith and Carter
[1969] to account for the degree of frame infill interaction. Ewθ and Ec are the elastic moduli of
masonry and concrete, hw is height of the infill panel, Ic is the moment of inertia of the adjacent
column cross section about the normal to the infill panel and θ is the inclination of the diagonal
strut with respect to the horizontal direction.
104 Experimental and numerical seismic performance of strong clay masonry infills
Ewθ t w sin 2θ
λ=4 (6.2)
4 Ec I c hw
The elastic modulus of masonry Ewθ assigned to the inclined direction is evaluated based on a
model for orthotropic elastic materials subjected to bi-axial tensile stresses, defined by Equation
(6.3), where ν is Poisson’s ratio.
−1
cos 4 θ sin 4 θ 1 ν
Ewθ = + + cos 2 θ sin 2 θ − 2 (6.3)
Ewh Ewv G Ewv
One of the important advantages of the masonry infill model proposed by Decanini et al. [1993] is
the ability to account for different failure mechanisms; namely compression at the centre,
compression at the corners, shear sliding and diagonal tension (Figure 6.9). Each failure mode is
associated with a corresponding value of ultimate strength σw as a function of the relevant material
and strut properties allowing the evaluation of the governing condition and correspondingly the
final compressive strength fm’. Hence, as in several previous numerical studies, e.g. by Decanini
[2004a], Magenes and Pampanin [2004], De Sortis et al. [2007], Morandi et al. [2011a, 2011b], the
equivalent strength σw for each mechanism is evaluated through the expressions given in Equation
(6.4) to Equation (6.7). The vertical compression stress due to gravity loads is denoted by σv, while
fwv is the vertical compression strength of the infills, fwu is the sliding resistance of the mortar joints
and fws is the shear resistance under diagonal compression.
(a) (b) (c) (d)
Figure 6.9. (a) Compression at the centre of the infill; (b) Compression at the corners of the infill; (c) Sliding
shear failure; (d) Diagonal tension failure
1.16 f wv tan θ
σw = (Compression at the centre) (6.4)
K 1 + K 2 λh
1.12 f wv sin θ cos θ
σw = (Compression at the corners) (6.5)
K 1 (λh )− 0.12 + K 2 (λh )0.88
σw =
(1.2 sin θ + 0.45 cos θ ) f wu + 0.3σ v (Shear sliding)
bw (6.6)
dw
0.6 f ws + 0.3σ v
σw = (Diagonal tension)
bw (6.7)
dw
In order to include the masonry infill panels in the frame model, two nonlinear spring elements
representing the equivalent struts are defined for each infill, one across every diagonal, and the
relevant strength and stiffness parameters are assigned. Specifically, the strength of the strut is
P. Morandi, S. Hak, G. Magenes EUCENTRE 105
Research Report
defined based on the governing failure mode obtained from Equation (6.4) to Equation (6.7),
resulting in a diagonal strut force Fw,s given by Equation (6.8), while the axial stiffness per unit
length kw of the strut is defined by Equation (6.9).
Fw = f m ' t w bw (6.8)
Ewθ t w b w
kw = (6.9)
dw
The relevant parameters for the definition of the equivalent strut model are reported in Table 6.5
for representative masonry infills without opening in the external span of the RC frame (F -
external), without opening in the internal span (F, P, PO - internal) and with opening in the external
span (PO, BO - external), as reported in Oliaee et al. [2015b]. In the case of infill with opening, the
strut cross section area tw·bw was reduced to take into account the precence of the opening
through the reduction coefficient rp proposed by Dawe and Seah [1988], as specified in Equation
(7.15) and (7.16) of paragraph 7.3.1, therefore, both the stiffness and the strut force are reduced.
Table 6.5. Strength and stiffness properties for masonry infill equivalent struts at the bottom storey of a
representative building configuration (6-storey DCH, 0.25g)
tw Lw hw dw Ewθ tw b w kw fm ’ Fw,s
Typology
[mm] [mm] [mm] [mm] [MPa] [m2] [kN/m] [MPa] [kN]
F / external 350 4550 2775 5329 856 0.579 93005 0.585 339
F, P, PO / internal 350 1550 2775 3179 3379 0.223 237030 1.137 254
PO, BO / external 350 4550 2775 5329 856 0.579·0.49 45572 0.585 166
In the structural models of the infills used for the analyses on the buildings, the values of the cross
section areas of the struts have been slightly modified as respect to the one reported in Table 6.5,
in order to adjust the small difference between the peak force of the experimental responses and
the strut force Fw,s computed according to Equation (6.8), with fm’ evaluated using the strength
parameters fwv, fwu and fws obtained from the tests of characterization on the masonry, as reported
in (Oliaee et al. [2015a]).
Moreover, in order to characterize the stress-strain relationship, reflecting in particular the
deformation capacity of the strong infill typology without and with opening, as well as the
corresponding hysteretic response, the infill model has been established by Oliaee et al. [2015a],
considering the calibration performed based on newly available experimental data for strong infills,
presented in Section 5.1.2.
strong-motion recordings, and the rapid development of digital seismic networks worldwide,
increasing the accessibility to recorded accelerograms.
Masi et al. [2011] have studied the choice of seismic input on structural types representing
reinforced concrete buildings widely present in the Italian and European built environment; i.e. bare
frames, fully infilled frames and infilled frames with open ground storey. Based on a large set of
nonlinear dynamic analyses they have found that the application of artificial accelerograms
generated (using SIMQKE and BELFAGOR procedures) to converge to a code response spectrum
results in very high damage levels, defined in terms of inter-storey drift demands, confirming that
they are overly conservative in comparison to natural accelerograms, assuming equal values of
peak ground accelerations. Hence, following the recommendations from previous studies, in order
to perform the series of nonlinear time-history analyses on general case study frames designed for
different PGAs and to evaluate average response quantities, primarily in terms of inter-storey drifts,
the application of natural records scaled to different levels of seismicity has been accepted as the
most suitable solution for the needs of this study. Even though the design spectra adopted for the
dimensioning of the frames were selected for specific sites in Italy, the case study frames where
meant to represent generic prototype structures designed for a certain PGA, and therefore the
selection of ground motions was not related to specific site characteristics.
For the selection of code compatible earthquake records Eurocode 8 – Part 1 (CEN [2004a])
prescribes the following requirements to be met: a minimum of 3 accelerograms should be used;
the mean of the zero period spectral response acceleration values, calculated from the individual
time-histories, should not be smaller than the value of agS for the site in question, where ag is the
design peak ground acceleration and S is the soil factor; in the range of periods between 0.2T1 and
2T1, where T1 is the fundamental period of the structure in the direction where the accelerogram
will be applied, no value of the mean 5% damping elastic spectrum, calculated from the individual
time-histories, should be less than 90% of the corresponding value of the 5% damping elastic
response spectrum. Additionally, the code prescribes to analyse the structure for at least three
earthquake records, if the most unfavourable response quantities are evaluated, or alternatively for
a minimum of seven records, if average response quantities from all analyses are considered.
For the need of this study, sets of ten earthquake input ground motions are used for the analysis of
each frame model. Besides the prescribed lower bound equal to 90% of the 5% damping elastic
response spectrum to be imposed on the average acceleration spectrum obtained from the
selected records, an upper bound equal to 110% has been adopted for the selected
accelerograms, even though not requested by the code.
The computer software REXEL (Iervolino et al. [2010]; Iervolino and Galasso [2009]) has been
used to select the natural earthquake motion records from the ITalian ACcelerometric Archive
(ITACA; Luzi et al. [2008]; Pacor et al. [2011]). Two sets of records were selected, one
compatible with the elastic response spectrum adopted for design at the ultimate limit state
corresponding to the intermediate peak ground acceleration of 0.15gS and one to the
corresponding elastic response spectrum at the damage limitation limit state with 0.054gS.
For each of the two target spectra the selection has been repeated until in the range between 0.05
s and 3.0 s the deviation of the average of ten earthquake spectra in one set has been found to be
below ±10%, while the spectral value at zero period is set to match the PGA. Only earthquakes
recorded on ground type B have been considered, with magnitudes between 4.0 and 6.0 at the
damage limitation limit state and between 5.5 and 7.0 at the ultimate limit state. The average scale
factor for each set has been kept as low as possible for the available records, resulting in values
below 2.0 in average and below 4.0 for a single record. Subsequently, the compatible records have
been scaled to the remaining PGAs adopted for design, i.e. 0.023gS, 0.040gS, 0.097gS and
0.142gS at the damage limitation limit state and 0.05gS, 0.10gS, 0.25gS and 0.35gS at the
ultimate limit state.
A summary of the selected records is given in Table 6.6 and Table 6.7 along with the main
properties, concerning magnitude, epicentral distance and the adopted total scale factor for
different PGAs. The compatible earthquake spectra compared to the target design spectra at the
damage limitation (with ag = 0.054g) and at the ultimate limit state (with ag = 0.15g) are shown in
P. Morandi, S. Hak, G. Magenes EUCENTRE 107
Research Report
Figure 6.10 and Figure 6.11, respectively. Earthquake time-history plots corresponding to the
considered levels of seismic intensity are displayed in Figure 6.12 for the damage limitation and in
Figure 6.13 for the ultimate limit state.
Table 6.6. Selected spectrum compatible earthquake records at the DLS
Epicentral Scale factor
Ground
Earthquake Date Mw Distance
Type 0.023g 0.040 g 0.054 g 0.097g 0.142 g
[km]
EQ1 Irpinia 24/11/1980 5.0 17.0 B 0.81 1.41 1.90 3.41 5.00
EQ2 L'Aquila 07/04/2009 5.6 9.4 B 0.36 0.63 0.85 1.53 2.24
EQ3 Gran Sasso 09/04/2009 5.4 12.1 B 0.58 1.00 1.35 2.43 3.55
EQ4 Friuli (3rd) 15/09/1976 5.9 41.1 B 0.94 1.63 2.20 3.95 5.79
EQ5 Friuli 18/05/1976 4.1 10.5 B 0.43 0.74 1.00 1.80 2.63
EQ6 L'Aquila 13/04/2009 5.1 17.9 B 1.28 2.22 3.00 5.39 7.89
EQ7 Gran Sasso 09/04/2009 5.4 16.2 B 0.68 1.19 1.60 2.87 4.21
EQ8 East Sicily 13/12/1990 5.6 79.7 B 1.28 2.22 3.00 5.39 7.89
EQ9 Italia Meridionale 16/01/1981 5.2 10.5 B 0.43 0.74 1.00 1.80 2.63
EQ10 Friuli 11/06/1976 4.5 11.9 B 1.58 2.74 3.70 6.65 9.73
AVERAGE 5.2 22.6 0.83 1.45 1.96 3.52 5.15
Figure 6.10. Code compatible earthquake acceleration and displacement spectra at the DLS
P. Morandi, S. Hak, G. Magenes EUCENTRE 109
Research Report
Figure 6.11. Code compatible earthquake acceleration and displacement spectra at the ULS
110 Experimental and numerical seismic performance of strong clay masonry infills
7.1 Introduction
A series of parametric numerical investigations was primarily meant to improve the understanding
of the seismic response of infilled RC structures, in particular regarding the damage behaviour of
different infill configurations for the case of a strong masonry infill typology. Consequently, the
numerical results were thought to provide support for the introduction of supplemental design
criteria for enhanced masonry infill damage control in the design of RC structures. Therefore, to
evaluate the response of 2-dimensional bare and infilled case study RC frame structures,
introduced in Section 6.2.1, firstly, modal analyses have been carried out to assess the
fundamental vibration characteristics for the structural configurations, and subsequently, nonlinear
dynamic time-history analyses has been performed. Using the structural analysis program
Ruaumoko (Carr [2007]), the modelling approach described in Section 6.3.2 has been adopted and
the strut model for strong infill and strong infill with opening, calibrated according to the procedure
described in Section 5.1.2, has been implemented.
The obtained elastic periods of vibration for the first three mode shapes are summarised in Table
7.1 for bare and infilled frame configurations. The periods of vibration clearly reflect the
amplification of stiffness due to the increasing amount of infill present in the building.
Table 7.1. Elastic periods of vibration for bare and infilled 2-D structural models
[s] 3-storey 6-storey 9-storey
DC ag Mode B P BO PO F B P BO PO F B P BO PO F
0.05g 1 0.98 0.37 0.34 0.25 0.21 1.79 0.78 0.65 0.47 0.37 2.42 1.26 0.89 0.67 0.53
- 2 0.30 0.13 0.11 0.09 0.07 0.58 0.21 0.23 0.16 0.13 0.83 0.33 0.32 0.22 0.19
0.25g 3 0.16 0.09 0.07 0.06 0.05 0.31 0.11 0.14 0.09 0.08 0.45 0.16 0.19 0.13 0.11
DCM
1 0.77 0.31 0.28 0.21 0.17 1.61 0.71 0.60 0.43 0.34 2.25 1.18 0.83 0.62 0.49
0.35g 2 0.22 0.11 0.09 0.07 0.06 0.52 0.19 0.21 0.15 0.12 0.76 0.30 0.30 0.21 0.17
3 0.11 0.07 0.05 0.05 0.04 0.27 0.10 0.13 0.08 0.07 0.41 0.15 0.17 0.12 0.10
0.05g 1 0.78 0.36 0.34 0.25 0.21 1.25 0.71 0.62 0.47 0.38 1.56 1.04 0.82 0.66 0.53
- 2 0.25 0.13 0.11 0.09 0.08 0.43 0.21 0.22 0.16 0.13 0.56 0.30 0.30 0.22 0.18
0.25g 3 0.15 0.09 0.07 0.06 0.05 0.25 0.11 0.13 0.09 0.08 0.32 0.16 0.18 0.12 0.10
DCH
1 0.78 0.36 0.34 0.25 0.21 1.25 0.71 0.62 0.47 0.38 1.56 1.04 0.82 0.66 0.53
0.35g 2 0.25 0.13 0.11 0.09 0.08 0.43 0.21 0.22 0.16 0.13 0.56 0.30 0.30 0.22 0.18
3 0.15 0.09 0.07 0.06 0.05 0.25 0.11 0.13 0.09 0.08 0.32 0.16 0.18 0.12 0.10
To carry out the nonlinear time-history analyses, for the numerical integration of the equation of
motion, Newmark’s constant average acceleration scheme has been adopted. The integration time
step set equal to Δt = 0.001 was found to result in satisfactory convergence and reasonable time
consumption. All output quantities have been reported at a frequency of 50 Hz. The influence of
large displacements on the frame response has been accounted for and P-Δ effects have been
included in the analyses of the 9 - storey frames. The viscous damping is represented by an initial
stiffness proportional Rayleigh model and values of 3.5% and 3.0% are adopted in the first
vibration mode for medium and high ductility, respectively, while 5.0% damping is assumed in the
second mode.
An examination of the overall nonlinear response of the prototype frames has shown that plastic
hinges form only at the bottom of the columns in the bottom storey and at the beam ends, while
columns in upper storeys remain elastic, as imposed during design, confirming the satisfactory
achievement of applied capacity design principles and the absence of soft-storey mechanisms.
Since the global structural response was of primary importance for the objectives of the study,
supposing that the frames were properly designed and detailed according to modern seismic
codes, shear deformations were assumed to remain elastic and the possibility of shear failure
114 Experimental and numerical seismic performance of strong clay masonry infills
occurrence has been excluded. Nevertheless, concentrated column shear demands may be
verified a posteriori, as comprehensively discussed in the work by Hak et al. [2013c].
As always the case when dealing with the outcome of time-history analyses, a significant influence
on the results has to be contributed to the selection of the seismic input. The nonlinear analyses
have been carried out representing the seismic action by sets of ten natural earthquake records at
each limit state, with acceleration spectra compatible in average with the code specified elastic
spectrum, as discussed in Section 6.3.3 and the evaluation of response quantities has been
carried out in terms of average values. Inelastic drift demands of the bare and infilled frame
configurations have been assessed in function of the infill typology and compared to results
obtained for slender/weak masonry infill in previous investigations (Hak et al. [2013a]). The
corresponding masonry infill damage has been estimated, considering the performance of single
masonry infills at both the damage limitation and the ultimate limit state, according to the
performance requirements given in Section 5.1.1. Hence, commonly defined limit states for
masonry infills in RC structures (see Hak et al. [2012]) have been verified for each case study
frame, resulting in values of maximum design seismic action for which satisfactory response could
be achieved following the current design approach. Furthermore, relating in-plane drift demands of
bare and infilled structural configurations through a simplified parameter, which accounts for the
contribution of infill strength and stiffness to the global response, the possibility to predict expected
in-plane drift demands of infilled RC structures has been studied, resulting in related implications
for the current design approach.
With the aim to show a representative selection of results, the displacement and drift profiles for 3-
storey, 6-storey and 9-storey buildings designed for a ULS design PGA of 0.35g in ductility class
medium at the ultimate limit state are presented, including the bare structures (B, Figure 7.1) and
the different infilled frame configurations, i.e. the case of infill in the central bays (P, Figure 7.2),
infill with opening in the external bays (BO, Figure 7.3), infill in the central bays and infill with
opening in the external bays (PO, Figure 7.4), as well as infill in all bays (F, Figure 7.5). Analysing
the displaced shapes for different frame types, in the case of the 3-storey frames, maximum drift
values are usually found to occur at the second storey for bare frames and partially infilled
configurations, and at the bottom storey for fully infilled frames. In the case of 6-storey and 9-
storey bare frames maximum drift demands are imposed around or slightly below mid-height, and
are regularly being slightly shifted downwards due to the presence of infills. With increasing peak
ground accelerations, the displacement behaviour of the prototype frames does not change
considerably, besides the displacement and drift amplitudes. The behaviour of infilled frames with
the least amount of infill (configurations BO and P) is found to be the most similar to that of the
bare frames. Observations on the drift profiles confirm that there is no occurrence of any soft-
P. Morandi, S. Hak, G. Magenes EUCENTRE 115
Research Report
storey mechanisms. For frames with varying storey heights and infill typologies, having therefore
different fundamental periods of vibration, different earthquake excitations have been identified as
the most demanding ground motion records, inducing the largest displacement and drift demands
on the frames.
Figure 7.1. Displacement and drift profiles at ULS: 0.35gS DCM frames - Bare B
Figure 7.2. Displacement and drift profiles at ULS: 0.35gS DCM frames - Infill P (Oliaee et al. [2015b])
116 Experimental and numerical seismic performance of strong clay masonry infills
Figure 7.3. Displacement and drift profiles at ULS: 0.35gS DCM frames - Infill BO (Oliaee et al. [2015b])
Figure 7.4. Displacement and drift profiles at ULS: 0.35gS DCM frames - Infill PO (Oliaee et al. [2015b])
P. Morandi, S. Hak, G. Magenes EUCENTRE 117
Research Report
Figure 7.5. Displacement and drift profiles at ULS: 0.35gS DCM frames - Infill F (Oliaee et al. [2015b])
Being the most relevant parameter for the assessment of the infill performance, in order to
summarise maximum average inter-storey drift demands as a function of ULS design PGA,
ductility class and infill configuration at the damage limitation limit state and at the ultimate limit
state, the results for each building typology are presented by means of one bar chart for each bare
or infilled building configuration (denoted by B, BO, P, PO and F). For comparison, the results
corresponding to a previously studied slender/weak clay masonry infill typology (T1, see Hak et al.
[2013a]), for the case of fully infilled configurations, are included. Presented values show the
highest drift demand obtained along the height of the building, evaluated from the average
response induced by ten different earthquake records (referred to as δμ, see Equation (7.2) and
Equation (7.3)).
Accordingly, the plots in Figure 7.6, Figure 7.7 and Figure 7.8 summarise the inter-storey drifts for
the 3-storey, 6-storey and 9-storey buildings, respectively, as a function of ULS design PGA (i.e.,
0.05gS, 0.10gS, 0.15gS, 0.25gS and 0.35gS).
1 10
δµ = ∑ δ max,i
10 i =1
(7.2)
Figure 7.6. Average maximum inter-storey drifts of 3-storey frames – Bare (B) and infilled frames (P, BO,
PO, F & T1): (a) DLS; (b) ULS (Oliaee et al. [2015b])
Figure 7.7. Average maximum inter-storey drifts of 6-storey frames – Bare (B) and infilled frames (P, BO,
PO, F, T1): (a) DLS; (b) ULS (Oliaee et al. [2015b])
P. Morandi, S. Hak, G. Magenes EUCENTRE 119
Research Report
(a) DCM DCH
Figure 7.8. Average maximum inter-storey drifts of 9-storey frames – Bare (B) and infilled frames (P, BO,
PO, F & T1): (a) DLS; (b) ULS (Oliaee et al. [2015b])
The obtained results show a decrease of inter-storey drift demands due to the presence of infills,
with respect to the bare frame. This reduction is more evident for configurations with larger
amounts of infill (e.g. F with respect to P), and for infills without openings (e.g. F with respect to
PO), due to the higher strength and stiffness properties. At lower ground motion intensities the drift
demands of RC frames fully infilled with the slender/weak infill typology (T1) seem to
approximately correspond to values obtained for the case of strong infill with opening in the
external bays and bare central bays (BO). Thus, the comparison reflects the lower strength and
deformation capacity of the slender/weak infill. In addition, with increasing seismic actions, the
inter-storey drifts for the slender/weak infill increase relatively with respect to those for the strong
infill, approaching or exceeding the demand for the case of strong infill in the central bay only (P),
e.g. above 0.15gS at the ultimate limit state. Such behaviour can be attributed to the fact that the
response of slender/weak infill is characterised by rapid strength degradation after the peak
strength, while the strong infill maintains a substantial residual resistance.
Notably, at the damage limitation limit state average inter-storey drifts for the bare frames in some
cases, for regions of high seismicity, exceed the design limit of 0.50%, indicating that the current
code approach based on elastic analyses of a simplified model, safe-sided for the evaluation of
design forces, may not fully control the displacement response, as discussed more extensively in
Hak et al. [2013a].
are obtained at the drift levels of interest, which approximately coincide also with the parameters
corresponding to the dimensions of the tested specimen, as shown in Table 7.3.
Table 7.2. Performance levels for a single masonry infill applied to the considered infill typologies
Damage
Limit State Operational Ultimate Collapse
Limitation
Symbol
Table 7.3. Relation of inter-storey drift and axial strain in the diagonal strut
Damage
Limit State Operational Ultimate
Limitation
without with without with without with
Strong infill opening opening opening opening opening opening
Drift [%] 0.30 0.20 0.50 0.35 1.75 1.00
Configuration L [m] h [m] L/h Axial strain
Test frame 4.57 3.125 1.46 0.0014 0.0009 0.0023 0.0016 0.0081 0.0047
External bay 5.0 3.0 1.67 0.0013 0.0009 0.0022 0.0015 0.0077 0.0044
Internal bay 2.0 3.0 0.67 0.0014 0.0009 0.0023 0.0016 0.0080 0.0046
Hence, Figure 7.9 and Figure 7.10 summarise for all infill configurations (P, BO, PO, F) the
average in-plane damage distribution at the damage limitation and the ultimate limit state,
respectively, obtained for case study frames designed in ductility class medium, for ULS design
PGAs equal to 0.25gS and 0.35gS. The same results for the frames in ductility class high are
shown in Figure 7.11 and Figure 7.12. The reported average damage refers to the assessment of
the infill performance based on average values of axial strain in the equivalent struts obtained from
the analyses for each earthquake record.
(a) (b)
Infill P Infill BO Infill PO Infill F Infill P Infill BO Infill PO Infill F
Figure 7.9. Average damage distribution for RC frames in DCM at the DLS: (a) 0.25gS; (b) 0.35gS (Oliaee et
al. [2015b])
P. Morandi, S. Hak, G. Magenes EUCENTRE 121
Research Report
(a) (b)
Infill P Infill BO Infill PO Infill F Infill P Infill BO Infill PO Infill F
OK OK OK OK
Figure 7.10. Average damage distribution for RC frames in DCM at the ULS: (a) 0.25gS; (b) 0.35gS (Oliaee
et al. [2015b])
(a) (b)
Infill P Infill BO Infill PO Infill F Infill P Infill BO Infill PO Infill F
Figure 7.11. Average damage distribution for RC frames in DCH at the DLS: (a) 0.25gS; (b) 0.35gS (Oliaee
et al. [2015b])
(a) (b)
Infill P Infill BO Infill PO Infill F Infill P Infill BO Infill PO Infill F
Figure 7.12. Average damage distribution for RC frames in DCH at the ULS: (a) 0.25gS; (b) 0.35gS (Oliaee
et al. [2015b])
122 Experimental and numerical seismic performance of strong clay masonry infills
In line with the evaluated inter-storey drift demands, the obtained distribution of damage points to
the significantly less favourable seismic response of partially infilled frame configurations with
respect to fully infilled frames, as well as to the greatly increased vulnerability of infills due to the
presence of openings. Nevertheless, for most frame configurations, the average response reveals
that infills did not experience significant damage at the damage limitation limit state and/or failure
at the ultimate limit state.
Starting from the distribution of damage evaluated for each single masonry infill, the overall
response of the building regarding the fulfilment of corresponding performance criteria can be
assessed. A commonly adopted definition of limit states with reference to masonry infill damage
(Hak et al. [2013a]), supplementing the general description of performance requirements according
to European and Italian code regulations (CEN [2004a]; NTC08 [2008]), is summarised in Table
7.4. The given requirements refer to critical points in the response of a single infill, experimentally
obtained in Section 5.1.1 for the case of strong infill and strong infill with opening (see Figure 5.2),
and expressed as limitations of inter-storey drift (Table 5.1, Table 7.2) or axial strain (Table 7.3).
Table 7.4. Definition of limit states with reference to masonry infill damage
Even though the damage distribution is presented only in terms of average values, the fulfilment of
the performance requirements has been verified for each case study frame exposed to every
earthquake record. Thus, each prototype frame, designed following the current code procedure, is
classified at each specified level of seismicity to either satisfy or fail to meet the damage limitation
and ultimate limit state requirements, for different infill configurations. The frames are assumed to
fulfil the performance criteria if the requirements defined in Table 7.4 are satisfied for more than
50% of the earthquake records (i.e., at least six out of ten) and for the corresponding average
response. The maximum levels of ULS design PGA, for which a satisfactory behaviour has been
obtained at each limit state, are summarised in Table 7.5 for all prototype buildings. The results
indicate that, for almost all infill configurations at both limit states in RC frames for both ductility
classes, a satisfactory behaviour has been achieved up to ULS design seismic PGAs of 0.35gS.
An exception is the BO configuration for the 9th storey frame DCM, which has attained a maximum
ULS design PGA of 0.25gS and the BO configuration for the 6th storey frame DCM, which has
barely satisfied the performance criteria at 0.35gS (in Table 7.5 this value has been asterisked).
Therefore, although the behaviour of strong masonry infills achieved following the current code
provisions can mostly be considered satisfactory with respect to the performance requirements,
exceptions may be expected for scarcely distributed infills and for a predominant use of infills with
openings.
P. Morandi, S. Hak, G. Magenes EUCENTRE 123
Research Report
Table 7.5. Maximum levels of ULS design PGA corresponding to obtained satisfactory masonry infill
performance in RC frames (Oliaee et al. [2015b])
Damage Limitation Ultimate
ag/S [g]
Infill P Infill BO Infill PO Infill F Infill P Infill BO Infill PO Infill F
3-storey frame
DCM 0.35 0.35 0.35 0.35 0.35 0.35 0.35 0.35
DCH 0.35 0.35 0.35 0.35 0.35 0.35 0.35 0.35
6-storey frame
DCM 0.35 0.35 0.35 0.35 0.35 0.35* 0.35 0.35
DCH 0.35 0.35 0.35 0.35 0.35 0.35 0.35 0.35
9-storey frame
DCM 0.35 0.35 0.35 0.35 0.35 0.25 0.35 0.35
DCH 0.35 0.35 0.35 0.35 0.35 0.35 0.35 0.35
Considering that a single criterion for the limitation of damage in the case of rigid masonry infills
corresponding to a maximum drift of 0.50% is commonly adopted, in the case of the configuration
without opening, for strong infills with a relatively high deformation capacity such condition, if it
governs the design choices, may even be overly conservative. Conversely, in the case of scarcely
distributed infills or of a predominant use of infills with openings, even for strong infills, the drift limit
of 0.50 % could not be always suitable to guarantee the fulfilment of the performance criteria and
to limit the in-plane damage of the infills, in particular at a high level of seismic action.
Consequently, the introduction of refinements in the design procedure of RC frame structures in
order to prevent unacceptably extensive damage of masonry infills at higher levels of seismicity
and limit adequately both property loss and risk to human life, without overly penalising the RC
structure, appears to be appropriate for the case of strong infills.
infill performance represent significant properties required to characterise the expected response
of a specific infill typology.
(b) Simplified Evaluation of Relative Structural Stiffness. Values of elastic stiffness assigned
to each storey along the height of a RC structure can be approximated based on the elastic
analysis of a bare frame structural model using the equivalent static force approach. In particular,
according to Equation (7.4), the average elastic structural stiffness KS,j for storey j may be
estimated as the ratio of the horizontal storey shear Vj, evaluated in line with Equation (7.5), ns
being the number of storeys, and the corresponding inter-storey displacement dr,j (representing the
difference of the average lateral displacements ds,j and ds,j-1 at the top and at the bottom of storey j,
respectively), as given in Equation (7.6).
Vj
KS, j = (7.4)
dr, j
ns
V j = ∑ Fi (7.5)
i= j
d r , j = d s , j − d s , j −1 (7.6)
For the evaluation of the elastic structural stiffness KS,j per storey according to Equation (7.4), the
distribution of horizontal forces acting along the building height is needed; hence, the force Fj
acting on storey j can be determined according to Equation (7.7), where Fb is the total seismic
shear force, si, sj are the displacements of masses i and j in the fundamental mode shape, while
mi, mj are the corresponding storey masses, associated with gravity loads, assumed in the seismic
load combination. The evaluation of the seismic base shear Fb is reported in Equation (7.8), where
the fundamental period of vibration T1 and the mass M1 participating in the first mode response can
be estimated from a modal analysis and Se(T1) is the corresponding spectral ordinate of the elastic
response spectrum.
s j ⋅mj
F j = Fb ns
(7.7)
∑ s i mi
i =1
Fb = S e (T1 ) ⋅ M 1 (7.8)
The criterion described above has been applied for each structural configuration considered within
the scope of this study, as thoroughly described in Section 3.1. However, for the evaluation of Fj
other simplified procedures may equally be adopted. For instance, Eurocode 8 – Part 1 (CEN
2004a) proposes a simplified lateral force method of analysis where, instead of calculating the
fundamental mode shapes and periods in the horizontal directions based on a modal analysis of
the structure, linearly increasing horizontal displacements along the height may be assumed. In
this case, the horizontal forces Fj may be taken as given in Equation (7.9), where zi, zj are the
heights of masses mi and mj above the level of seismic force application. The corresponding
seismic base shear force Fb, for each horizontal direction in which the building is analysed, can be
calculated using Equation (7.10), where Sd (T1) is the ordinate of the design spectrum at period T1,
the fundamental period of vibration of the building for lateral motion in the considered direction and
m is the total mass of the building above the foundation or above the top of a rigid basement. The
correction factor λ may be calculated from Equation (7.11a), in function of the corner period of the
spectrum TC, the period T1 and the number of storeys ns. According to the code, the fundamental
period of vibration T1 can be approximated by the expression given in Equation (7.11b) in [s], for
buildings with heights up to 40 m. For spatial RC frames, the coefficient Ct may be assumed to be
equal to 0.075, while H represents the height of the building in [m], from the foundation or from the
top of a rigid basement.
P. Morandi, S. Hak, G. Magenes EUCENTRE 125
Research Report
z j ⋅m j
F j = Fb ns
(7.9)
∑ z i mi
i =1
Fb = S d (T1 ) ⋅ m ⋅ λ (7.10)
(a) (b)
Figure 7.13. (a) Masonry infill secant stiffness; (b) Estimated infill strength reduction due to opening
Fw , i , j Fw , i , j
K I ,i , j = , K I ,i , j = (7.12a,b)
d m ,i , j ' δ m ,i , j ' h j
nb , j
1 nb , j Fw , i , j 1 nb , j
K I , j = ∑ K I ,i , j , K I , j =
hj
∑δ , KI, j =
h jδ m, j
∑ Fw ,i , j (7.13a,b,c)
i =1 i =1 m ,i , j ' 'i =1
nb , j
∑ Fw ,i , j
δ m, j ' = i =1
(7.14)
nb , j
Fw ,i , j
∑δ
i =1 m ,i , j
Fw , i , j = f w , i , j t w , i , j L w , i , j (7.15)
In order to achieve a suitable approximation, the horizontal force Fw,i,j may be taken equal to the
peak horizontal strength of the masonry infill, calculated according to Equation (7.15), where fw,i,j
denotes the governing masonry strength, tw,i,j the thickness and Lw,i,j the length for infill i in storey j.
As discussed e.g. in Hak et al. [2013a], in absence of more specific data, the masonry strength fw,i,j
may be estimated in accordance with Eurocode 8 – Part 1 (CEN [2004a]), on the basis of the
corresponding shear strength of bed joints which may be assumed equal to the initial shear
strength under zero compressive stress fv0,i,j. However, with the aim of being more accurate, in
126 Experimental and numerical seismic performance of strong clay masonry infills
order to evaluate the infill strength fw,i,j , more detailed models accounting for different infill failure
modes, e.g. as discussed in Section 6.3.2 the model proposed by Decanini et al. [1993], should
preferably be adopted when the required data, such as the horizontal and vertical compression
strength as well as the shear strength under diagonal cracking, is available. Nevertheless, such
approach can be directly applied only in the case of unreinforced infills without openings. The
presence of openings may influence the performance of masonry infills significantly, as addressed
also in a number of recent studies (e.g. Kakaletsis and Karayannis [2008]; Stavridis et al. [2012];
Sigmund and Penava [2014]) and shown in Section 4.1.2 based on the obtained experimental
results for strong infills. Therefore, in the case of realistic building configurations, the possible
reduction of strength for perforated masonry infills has to be considered. Assuming the commonly
accepted representation of infills by means of an equivalent diagonal strut model, a simple
reduction factor, herein denoted by rp, can be applied to reduce the width (or cross section area) of
the infill strut in order to account for the presence of openings, as proposed by several authors
(e.g. Dawe and Seah [1988]; Decanini et al. [1994]; Al-Chaar [2002]; Asteris et al. [2011b]).
Specifically, even though more recent proposals are available, the reduction factor introduced by
Dawe and Seah [1988], applied also in a series of newer studies (e.g. Dolšek and Fajfar [2006];
Uva et al. [2012]), being safe-sided and adequately simple, can be considered suitable for a
simplified evaluation of the infill contribution. Hence, within the scope of this study, the presence of
openings is accounted for through the reduction coefficient rp, expressed by Equation (7.16),
defined in function of the ratio Ac of opening width lp and infill length Lw, in Equation (7.17).
r p = 1 − 1.5 Ac ≥ 0 (7.16)
lp 2
Ac = ⋅ 100 ≤ (7.17)
Lw 3
Thus, as illustrated in b, assuming an equivalent opening along the full infill height equal to hw,j, a
portion of the infill length equal to 1.5lp,i,j, where lp,i,j stands for the total width of the openings for bay
i in storey j, is neglected in the calculation of the infill strength. Accordingly, the contribution of a
masonry infill to the overall stiffness is disregarded if the total width of perforations lp,i,j exceeds 2/3
of the infill length Lw,i,j. Such approach, for the single-storey single bay frame which has been
experimentally tested within the scope of this study, as introduced in Section 3.2.1, for the case of
strong infill (specimen TA1, TA2 and TA3) and strong infill with a central opening of a width equal
to one third of the total infill length (specimen TA4), results in a strength reduction coefficient of rp
equal to 0.49. As discussed in Section 5.1.1, comparing the average infill response for the
specimen without opening, which has been tested at the highest drift demand (TA2), and of the
specimen with opening (TA4) a ratio of 0.45 has been obtained (see also Figure 5.1), indicating
that the proposed implementation of the simple reduction coefficient (Dawe and Seah [1988]) can
be considered appropriate for the case of strong infill. Hence, the peak horizontal strength of an
unreinforced infill with openings may be calculated according to the expression given in Equation
(7.17).
(
Fw ,i , j = f w ,i , j t w ,i , j L w ,i , j − 1.5l p ,i , j ) (7.18)
For the specific case of a storey that is fully infilled assuming masonry without openings, having a
strength equal to fw,0, a thickness equal to tw,0 and a value of inter-storey drift at maximum strength
characterising the deformation capacity of the infill equal to δm,0’, the associated infill stiffness KI,0,j
may be evaluated from Equation (7.19). The strength contribution of different infill layouts and
masonry typologies may be compared by means of a relative density parameter wj, as given in
Equation (7.20), representing the ratio of resulting infill strength in storey j of any realistic building
configuration and the corresponding infill strength for the specific reference case of full infill.
Hence, the infill stiffness KI,j for storey j may be expressed in function of the parameters KI,0,j and wj,
as given in Equation (7.21); however, remaining independent of the properties assumed for the
reference infill typology, i.e. the masonry strength fw,0, the infill thickness tw,0, and the characteristic
drift δm,0’.
P. Morandi, S. Hak, G. Magenes EUCENTRE 127
Research Report
nb , j
1 f w , 0t w , 0
K I ,0, j =
h j δ m ,0 '
∑ Li , j (7.19)
i =1
nb , j
∑ f w ,i , j t w ,i , j L w ,i , j
i =1
wj = nb , j (7.20)
f w , 0t w , 0 ∑ L w ,i , j
i =1
δ m ,0 '
K I , j = K I ,0, j w j (7.21)
δm, j '
(d) Evaluation of Average Infill vs. Structural Stiffness Ratio. Subsequently, for each load-
bearing direction of a regular RC structural configuration, given the distribution of infills and all
properties required to characterise the adopted infill typologies, knowing the infill parameter KI,j and
the structural stiffness parameter KS,j for each storey ( j = 1...ns, where ns is the number of stories),
the relative density-stiffness parameter Cj can be defined, representing an average ratio of infill vs.
structural stiffness for each storey along the building height, as given in Equation (7.22).
KI, j
Cj = (7.22)
KS, j
Introducing the simplified density-stiffness coefficient Cj, the combined influence of masonry infill
strength and stiffness with respect to the stiffness of the structural elements on the response of
infilled RC buildings can be assessed. Therefore, the verification of a relation between inter-storey
drift demands for bare and infilled structural configurations in function of the coefficient Cj is
envisaged, aimed to allow a safe-sided estimation of inter-storey drifts for the infilled structure
based on the given structural layout and infill properties, knowing the inter-storey drift demands of
the corresponding bare structural configuration.
(a) (b)
Figure 7.14. Average drift demands, 3-storey, 6-storey & 9-storey frames (P): (a) DLS; (b) ULS
(a) (b)
Figure 7.15. Average drift demands, 3-storey, 6-storey & 9-storey frames (BO): (a) DLS; (b) ULS
(a) (b)
Figure 7.16. Average drift demands, 3-storey, 6-storey & 9-storey frames (PO): (a) DLS; (b) ULS
(a) (b)
Figure 7.17. Average drift demands, 3-storey, 6-storey & 9-storey frames (F): (a) DLS; (b) ULS
Furthermore, for each infill configuration the coefficient KI,j has been evaluated from Equation
(7.13a), based on the infill strength, according to the assumptions adopted in the nonlinear models
(see Section 6.3.2). Importantly, as summarised in
Table 7.7, for each infill a characteristic drift δm,i,j’ reflecting the deformation capacity has been
assumed, being equal to 0.50% and 0.35% for strong infills without and with opening, respectively.
Accordingly, in the calculation of the infill stiffness parameter for configurations with the same type
of infill in all bays of each storey, the characteristic drift δm,j’ assigned to the entire storey is equal to
the corresponding deformation capacity of the infills. Conversely, for the buildings with strong infills
in the central bays and strong infills with opening in the external bays (PO), the equivalent
characteristic drift δm,j’ describing the deformation capacity of the storey has been calculated
according to Equation (7.14), resulting in a value of 0.38%.
P. Morandi, S. Hak, G. Magenes EUCENTRE 129
Research Report
Table 7.6. Summary of equivalent stiffness parameters for two-dimensional frames
Table 7.7. Summary of drifts representing the infill and equivalent storey deformation capacity
Drift Infill P Infill F Infill BO Infill PO
Internal bay i 0.50 0.50 - 0.50
δm,j' [%]
External bay i - 0.50 0.35 0.35
Storey j δm,j' [%] 0.50 0.50 0.35 0.38
The obtained values of KI,j for the building configurations including layouts of strong infills without
opening (P and F) and with opening (BO and PO) are summarised in Table 7.8 and Table 7.9,
respectively. The corresponding density-stiffness coefficients calculated according to Equation
(7.22) are presented in Table 7.10 and Table 7.11.
The presented results indicate that corresponding to a constant value of Cj, for increasing bare
frame drift demands, the average drift demands of the infilled frame increase approximately
linearly up to a certain level of drift. Clearly, changes in the ratio aC,j = δμ,w,j/δμ,j ≤ 1.0, describing the
dependence of infilled frame drifts δμ,w,j as a linear function of bare frame drifts δμ,j, are attributed to
changes in either the structural stiffness coefficient KS,j or the infill strength and stiffness parameter
KI,j. Notably, a value of aC,j equal to one would imply equal drifts of bare and infilled structural
configurations. For buildings designed in ductility class high, having due to the adopted design
choices higher values of initial frame stiffness and higher values of the parameter KS,j, drift
demands of the infilled configurations are closer to those of the bare structures, meaning that aC,j is
closer to one, than in the case of ductility class medium, where lower values of initial frame
stiffness were achieved. On the other hand, for increasing values of the parameter KI,j, quantifying
the presence of infill, smaller values of aC,j have been identified, pointing towards the more
pronounced reduction of drift demands for the infilled configurations with respect to the bare
structures. In conclusion, for smaller values of the structural stiffness coefficient KS,j, or higher
values of the infill strength and stiffness parameter KI,j, a smaller ratio of drift demands
corresponding to infilled and bare structural configurations is obtained. Nevertheless, the presented
results show also that, after exceeding a certain level of drift, the drift demands of infilled
configurations cannot be described in function of bare frame drifts with the same linear relation,
and due to the propagating infill damage, a performance more similar to that of the bare structure
is being approached, in terms of inter-storey drift demands.
In line with the observations discussed based on the results of nonlinear time-history analyses
carried out for bare and infilled RC frames in function of different building parameters, a simplified
relationship can be introduced, allowing the evaluation of the maximum expected average inter-
storey drift demand for masonry infilled structural configurations, in function of the average drift
demand for corresponding bare structures. Given that the design of regular masonry infilled RC
structures, according to current European code regulations (CEN [2004a]) is commonly carried out
on bare configurations, a safe-sided prediction of drift demands for infilled structures is considered
important for design applications.
132 Experimental and numerical seismic performance of strong clay masonry infills
7.3.3 Procedure for the Prediction of In-plane Drifts for Infilled RC Structures
(a) Estimation of Elastic Design Inter-storey Drift Demands. In seismic design procedures
according to Eurocode 8 – Part 1 (CEN [2004a]), the evaluation of displacement demands is
actually only required to account for P-Δ effects and to verify inter-storey drift limits, which are
essential for the control of damage to masonry infills adherent to surrounding structural elements.
As specified in the code, in the case of linear analyses, the displacements at the ultimate limit state
ds,ULS induced by the design seismic action should be calculated on the basis of the elastic
deformations of the structural system, as given in Equation (7.23), where qd is the displacement
global ductility, and de,ULS is the displacement determined by a linear analysis based on the design
response spectrum, obtained from the elastic spectrum introducing the behaviour factor q ≥ 1.5 as
a function of the design ductility class and the structural system. Eurocode 8 - Part 1 (CEN
[2004a]) reports also that the displacement induced by the design seismic action ds,ULS does not
need to be larger than the corresponding value derived from the elastic spectrum de,el,ULS,
corresponding to a behaviour factor q equal to 1.0.
d s ,ULS = qd d e ,ULS (7.23)
Displacements at the damage limitation limit state ds,DLS can be estimated according to Equation
(7.24), introducing the reduction coefficient ν, which accounts for the lower return period of the
seismic action associated with the damage limitation requirement, in line with general Eurocode 8
– Part 1 recommendations (CEN [2004a]), or as the displacement de,el,DLS induced by a site-specific
elastic spectrum defined for verifications at the damage limitation limit state, as given in the Italian
design provisions (NTC08 [2008]).
d s , DLS = ν DLS qd d e ,ULS ≤ ν DLS d e , el ,ULS (7.24)
Allowing a refinement of the displacement evaluation, Eurocode 8 and NTC 2008 note that in
general qd is larger than q, if the fundamental period of the structure T1 is less than the corner
period of the design spectrum TC. Hence, for long-period structures, qd may be evaluated
according to the equal displacement rule (Figure 7.18a), as given in Equation (7.25) and for short-
period structures according to the equal energy rule (Figure 7.18b), as given in Equation (7.26).
=qd q , T1 ≥ TC (7.25)
TC
qd =1 + ( q − 1) , T1 < TC (7.26)
T
Figure 7.18. Force reduction and displacement ductility: (a) Equal displacement rule; (b) Equal energy rule
(b) Inter-storey Drifts as a Function of the Density-stiffness Coefficient. Based on the results
obtained from parametric analyses for the types of structures and infill configurations considered in
this study, as well as for previously investigated slender/weak infill typologies (Hak et al. [2013a]),
it has been found that for a RC structure represented by the stiffness parameter KS,j in storey j, the
inter-storey drift demand δw,j of the infilled configuration, corresponding to the density-stiffness
parameter KI,j in storey j can be estimated in function of the density-stiffness coefficient Cj and the
drift demand δj of the corresponding bare structure. As illustrated in Figure 7.19a for Cj equal to 1.0
and in Figure 7.19b for different values of Cj, such correlation can be expressed using a bilinear
curve, according to Equation (7.27), where δm,j’ denotes the damage limitation drift capacity for the
masonry infills in storey j, while the parameter δC,j defines the relation of drifts between bare and
infilled configurations for Cj equal to 1.0.
P. Morandi, S. Hak, G. Magenes EUCENTRE 133
Research Report
(a) (b)
Figure 7.19. (a) Definition of bilinear curve for Cj = 1.0; (b) Variation of bilinear curve for different values of Cj
δ m , j’ δ j
, δ j ≤ δ m , j’ + δ C , j C j
δw, j = δ m , j’ + δ C , j C j (7.27)
δ −δ C , δ > δ ’ +δ C
j C,j j j m, j C,j j
The relationship has been developed in order to allow a safe-sided approximation of expected
inter-storey drift demands for infilled configurations, ensuring that upper-bound values are
obtained. Considering the values of drift δm,j’ assigned to the addressed infill configurations, the
bilinear curve in function of Cj has been found to describe adequately the obtained average drift
demands, assuming for storey j the value of δC,j, according to Equation (7.28), i.e. equal to 2/5 of
δm,j’, as summarised in Table 7.12.
2
δ C , j = δ m , j’ (7.28)
5
By definition, according to Equation (7.27), for the corner point of the bilinear curve the drift of the
infilled configuration δw,j always corresponds to the inter-storey drift capacity δm,j’, while the related
bare frame drifts δj change depending on the parameter Cj. Notably, a value of Cj equal to zero
actually corresponds to the bare structural configuration, while increasing values of Cj are obtained
for stronger infill typologies, due to higher values of infill strength and/or thickness, as well as for
building layouts with more densely distributed infills. On the contrary, buildings with stiffer load-
bearing structural systems are represented by smaller values of Cj. Accordingly, in the case of very
stiff structures and/or for small values of infill density, the prediction of drifts for the infilled structure
may result in values which are very close to those of the bare configuration. Likewise, for more
flexible structures with substantial amounts of infill, a significant reduction of drifts may be obtained
with respect to the corresponding bare structural system. Hence, the sequence of bilinear curves in
function of Cj effectively accounts for the typology and amount of infill present in the building,
reflecting the corresponding influence on the reduction of drift demands for the infilled configuration
in comparison to that of the bare structure. Moreover, the proposed approach appears to be
equally successful in capturing the evident fact that the same amount of infills with the same
strength and stiffness properties has less effect on the performance of stiffer RC structures and
causes a less pronounced reduction of drift demands with respect to the bare structure, than in the
case of more flexible load-bearing systems.
(c) Validation of Results. The proposed relationship between drift demands of bare and infilled
structural configurations in function of the density-stiffness coefficient Cj, given in Equation (7.27),
has been validated based on a comparison of the average response obtained from time-history
134 Experimental and numerical seismic performance of strong clay masonry infills
analyses for the typical frame structures and infill configurations analysed within the scope of this
study. Therefore, a comparison of numerically obtained average drift demands and drifts predicted
using the proposed bilinear curves has been accomplished. A selection of results obtained at both
limit states for all infill configurations (P, BO, PO and F), corresponding to the representative
frames designed for ductility class high, is presented in Figure 7.20 for the 3-storey frames, in
Figure 7.21 and Figure 7.22 for the 6-storey frames and in Figure 7.23, Figure 7.24 and Figure
7.25 for the 9-storey frames. In the same manner, results obtained for the building typologies in
ductility class medium may be summarised, leading to analogous conclusions.
(a)
(b)
Figure 7.20. Average and predicted drift demands, 3-storey DCH frame (P, BO, PO & F), j = 2: (a) DLS; (b)
ULS
(a)
(b)
Figure 7.21. Average and predicted drift demands, 6-storey DCH frame (P, BO, PO & F), j = 2: (a) DLS; (b)
ULS
P. Morandi, S. Hak, G. Magenes EUCENTRE 135
Research Report
(a)
(b)
Figure 7.22. Average and predicted drift demands, 6-storey DCH frame (P, BO, PO & F), j = 5: (a) DLS; (b)
ULS
(a)
(b)
Figure 7.23. Average and predicted drift demands, 9-storey DCH frame (P, BO, PO & F), j = 3: (a) DLS; (b)
ULS
(a)
(b)
Figure 7.24. Average and predicted drift demands, 9-storey DCH frame (P, BO, PO & F), j = 5: (a) DLS; (b)
ULS
136 Experimental and numerical seismic performance of strong clay masonry infills
(a)
(b)
Figure 7.25. Average and predicted drift demands, 9-storey DCH frame (P, BO, PO & F), j = 7, 8, 9: (a) DLS;
(b) ULS
According to the obtained results, the relationship for the evaluation of inter-storey drifts for
masonry infilled structures, based on drifts for corresponding bare structural configurations,
effectively reproduces the response variation in function of the density-stiffness coefficient Cj.
Apparently, for the considered case studies the average drift demands of the considered infilled
structures obtained from numerical analyses result to be equal or less than the corresponding
values estimated according to Equation (7.27). Hence, for the case of strong infills and strong infills
with openings, the proposed simplified approach can ensure the prediction of safe-sided upper
bound values of expected drift demands for masonry infilled structures depending on properties of
RC structural elements as well as infill layouts and masonry typologies. Based on the results
obtained for realistic three-dimensional frame and wall-frame dual structures in previous
investigations related to the response of slender/weak infills (Hak et al. [2013a]), an exception to
such observation has been found in the bottom storey of some case studies, where because of a
modification in the deformed shape of the structure, due to the presence of infill, the reduction of
drift demands may be less pronounced or even negligible. Consequently, inter-storey drifts at the
first storey of infilled structures may conservatively be assumed to be equal to those of the
corresponding bare configuration, resulting in a density-stiffness coefficient Cj equal to zero. After
all, the proposed approach may be considered appropriate for practical applications in the design
of regular structures, in particular for improvements in the verification of infill damage, commonly
accomplished through the limitation of inter-storey drift demands. However, it has to be
emphasised that possible consequences of irregular infill distribution in plan and elevation have not
been examined within the scope of this study.
P. Morandi, S. Hak, G. Magenes EUCENTRE 137
Research Report
8.1 Introduction
Given the fact that previous investigations (e.g. Dolce and Manfredi [2009]; Hak et al. [2012]) have
pointed to the need for improvements in the design of masonry infilled RC buildings, in particular
regarding the infill damage control based on inter-storey drift verifications, the proposed prediction
of inter-storey drifts for infilled structures may facilitate the introduction of more appropriate inter-
storey drift verifications, accounting in a simplified manner for the influence of masonry infills
present in the building configuration on the structural response. With the aim of ensuring effective
masonry infill damage control for masonry infills built in full contact with the surrounding RC
structure, the verification of inter-storey drifts corresponding to infilled structures, rather than bare
structural configurations, accomplished both at the damage limitation and at the ultimate limit state,
appears to present an adequate solution for the improvement of current provisions related to infill
damage control in the design of masonry infilled RC structures. Nevertheless, analyses carried out
to accomplish the required design verifications can still be performed considering bare
configurations.
In addition to the major aspects considered within the scope of this study, primarily related to the
cyclic in-plane response of strong masonry infills and infilled RC structures as well as the control of
infill damage through more effective inter-storey drift limitations, the estimation of expected in-
plane drift demands for infilled RC structures may induce important developments related to other
design aspects. In particular, estimating the response of the infilled configuration, a more detailed
evaluation of expected shear demands can be accomplished in the verification of local effects due
to the presence of masonry infill, through the assessment of effectively activated infill strut actions,
as discussed in Section 8.3. Furthermore, knowing the expected level of in-plane drift, in the out-
of-plane infill verification, a possible reduction of the out-of-plane resistance due to expected
simultaneous in-plane damage may be accounted for, as discussed in Section 8.4. Hence, the
findings of the presented extensive numerical investigations, which rely on newly available
experimental data for the case of unreinforced strong clay masonry infills, as well as the results of
a previous study for the case of weak/slender infill with and without light reinforcement (Hak et al.
[2013a]), imply the proposed improvements in the design of RC structures with masonry infills,
summarised in Figure 8.1.
δDLS. The inter-storey drift δj,DLS for storey j of storey height hj is calculated as dr,j,DLS/hj, where dr,j,DLS
represents the inter-storey displacement for storey j, evaluated as the difference of displacements
ds,j,DLS and ds,j-1,DLS from Equation (7.24). As introduced in Section 2.4.2, for buildings with brittle
non-structural elements attached to the structure in each storey of the building, δDLS is set equal to
0.50% and the drift verifications are normally carried out on the bare frame structural configuration.
This common choice seems to be additionally supported by the fact that the requirement to apply a
drift limit δDLS equal to 0.50% for infilled structures at the damage limitation limit state would imply
that any typology of rigid infills could remain substantially undamaged up to this level of drift.
However, such implication is in disagreement with major reported observations based on field
surveys pointing to frequent examples of infill failure (e.g. Braga et al. [2011]), as well as the
experimental results of the current study, as discussed in Section 5.1.1, supported also by findings
from earlier investigations (e.g. Calvi and Bolognini [2001]), which indicate that in traditional
unreinforced infills considerable damage may be reached at significantly lower levels of drift.
Commonly, seismic action effects at the damage limitation and the ultimate limit state (i.e.,
displacement and/or force demands), induced by horizontal actions represented by means of
corresponding spectral ordinates, are evaluated from an elastic bare frame structural model,
established to carry out linear analyses of the structure. In order to apply the proposed design
approach, based on the estimated response of the infilled configuration, the elastic stiffness
properties of the load-bearing structure related to assumed dimensions of structural elements to be
verified during design can be expressed through the coefficient KS,j given in Equation (7.4).
Furthermore, for the effective infill damage control, without penalising unnecessarily the RC
structure, the availability of additional relevant data is required. Specifically, the expected
distribution of infills within the plan of each storey, including the presence of significant openings,
should be defined, as commonly obtained from architectural layouts. In addition, in order to assess
the infill strength and stiffness contribution to the structural response through the coefficient KI,j
given in Equation (7.13a), values of masonry thickness tw,i,j and strength fw,i,j need to be known for
each infill. Moreover, characteristic values of inter-storey drift accounting for the deformation
capacity of the adopted infill typologies, i.e. δm,i,j’ equal to δDLS, corresponding to the attainment
of damage limitation limit state performance requirements for each infill, have to be defined. Inter-
storey drift limits for different limit states, in function of corresponding masonry infill performance
requirements, related to the acceptable extent of damage for different infill typologies may be
derived based on experimental data from in-plane cyclic tests, as defined within the scope of this
study for the case of strong infill and strong infill with opening. Previously, masonry infill limit states
have been defined for instance in the work by Hak et al. [2013b] through the interpretation of test
results (Calvi and Bolognini [2001]) for traditional unreinforced and lightly reinforced slender/weak
infill typologies. Similarly, in other recent studies, extensive experimental tests have been carried
out for contemporary strong clay block masonry infills (e.g. da Porto et al. [2012]).
In function of the density-stiffness coefficient Cj evaluated according to Equation (7.22) the bilinear
curve given by equation (7.27) can be defined and inter-storey drift demands in each storey of the
infilled configuration can be predicted. According to the validation of the proposed approach, the
density-stiffness coefficient Cj for the estimation of inter-storey drifts at the first storey should be
taken equal to zero, in order to ensure a safe-sided estimation. Finally, inter-storey drift
verifications at both limit states can be carried out, according to Equation (8.1) and Equation (8.2).
δ w , j , DLS < δ DLS (8.1)
δ w , j ,ULS < δULS (8.2)
In particular, the estimated drift demands for each storey j of the infilled structural configuration
δw,j,DLS and δw,j,ULS should not exceed the drift limits δDLS and δULS assigned to each infill typology,
corresponding to the attainment of damage limitation and ultimate limit state performance
requirements with reference to acceptable infill damage, which can be evaluated directly from in-
plane cyclic tests. If any of the inter-storey drift limitations cannot be satisfied, building properties
need to be modified (e.g., the structural stiffness may be increased, enlarging member sizes, or
infill properties may be modified), resulting in new values of the density-stiffness coefficient Cj.
Hence, the verification of newly obtained infilled frame drifts needs to be repeated. Consequently,
P. Morandi, S. Hak, G. Magenes EUCENTRE 139
Research Report
the force-based procedure that needs to be followed in the design of the RC structure in order to
achieve sufficient damage control for the displacement-sensitive non-structural masonry infills, in
particular at higher levels of seismicity, may result to be iterative. As a matter of fact, if the drift limit
at the ultimate limit state becomes the governing design condition, the ductility of the structure
cannot be fully exploited, implying that the structural displacements imposed by a behaviour factor
q corresponding to ductile bare RC structural configurations are not compatible with the
deformation capacity of typical masonry walls, particularly in the case of traditional unreinforced
typologies.
Figure 8.2. Critical length of columns: (a) Ground floor; (b) Full contact with the infill on one side
The same requirement applies in the case when infills extend over the entire column height, but
only on one side of the column (Figure 8.2b), as well as for columns which are in contact with
adjacent infills along just a part of their total height (Figure 8.3a). In this case the column has to be
verified also for the corresponding shear demand VC,Ed,cl, determined from the provided column
resistance following the capacity design principle applied over the clear height lcl, as given in
Equation (8.3) and illustrated in Figure 8.3b, adopting an overstrength coefficient γRd = 1.1 for
ductility class medium and γRd = 1.3 for ductility class high.
2 MC , Rd
VC , Ed , cl = γ Rd (8.3)
l cl
fs
Figure 8.3. Partial contact of column and infill: (a) Critical length lcr and (b) Decreased column shear span
ratio for clear length lcl; (c) Flexural column capacity at ends of contact length lc
Further existing safety verifications rely upon the common simplification of the masonry infill action
based on the equivalent single-strut model, requiring the verification of increased shear demands
due to horizontal strut forces acting on the column VC,Ed,l, given by Equation (8.4). In particular, the
contact length lc of the column, over which the diagonal strut force of the infill acts on the column
causing a local pressure fs and introducing a concentration of forces, should be verified in shear for
the demand evaluated as the smaller of the following two shear forces:
i) the shear demand VC,Ed,w equal to the horizontal component Fw of the diagonal strut force
Fw,s of the infill, that can be assumed equal to the strength of the panel;
140 Experimental and numerical seismic performance of strong clay masonry infills
ii) the shear demand VC,Ed,M determined following the capacity design principle, assuming that
the flexural column capacity develops at the ends of the contact length lc (Figure 8.3c).
VC , Ed , l = min (VC , Ed , w ,VC , Ed , M ) (8.4)
According to the code, the horizontal component Fw of the diagonal strut force Fw,s can be
assumed to be equal to the horizontal shear strength of the panel, estimated on the basis of the
shear strength of the bed joints, but no explicit recommendations related to the evaluation of such
strength are specified. A reasonable choice, adopted also for the estimation of the horizontal
masonry infill strength corresponding to the experimentally investigated strong infill typology, as
described in Section 5.1.3, could be to use the expression included in Eurocode 6 – Part 1 (CEN
[2004b]), commonly assumed for load-bearing masonry walls, considering the shear strength of
the bed joints equal to the initial shear strength under zero compressive stress fv0. Hence, for a
masonry infill with a thickness tw, length Lw and shear strength fv = fv0 the horizontal strength Fw
may be estimated in a simplified manner according to Equation (8.5), resulting in the
corresponding shear demand VC,Ed,w (Equation (8.6)). The shear demand resulting from the
application of the capacity design principle for the region of contact between the infill strut and the
RC column is given in Equation (8.7).
Fw = f v 0 t w L w (8.5)
VC , Ed ,w = Fw (8.6)
2 M C ,Rd
VC , Ed ,w = γ Rd (8.7)
lc
As discussed in more detail by Hak et al. [2013a, 2013c], a number of more refined models for the
evaluation of the masonry infill strength Fw and corresponding equivalent diagonal strut forces
imposing increased column shear demands is available. Some of these models are able to
account for different failure mechanisms typically observed on masonry panels and usually
supported by experimental studies, e.g. as proposed by Decanini et al. [1993] for compression at
the centre, compression at the corners, shear sliding and diagonal tension. Nevertheless, a
simplified approach assuming that the masonry infill results in the shear demand given by Equation
(8.5) for most practical applications may be considered safe-sided or even overly conservative.
Clearly, the evaluation of local shear design forces can be influenced significantly by the choice of
the contact length lc, over which the diagonal strut force of the infill acts on the column and related
recommendations should be carefully assessed (see e.g. Paulay and Priestley [1992], Decanini et
al. [1993], Fardis [2006]).
of a cautious evaluation of the contact length appears to be additionally pronounced, since the
related capacity design verification may more frequently influence the final design choice.
Furthermore, observations have shown that in some cases relatively high values of column shear
resistance have been achieved due to detailing requirements in order to obtain the needed ductility
in critical regions, or due to the choice of large cross sections required by other design issues,
such as the limitation of in-plane displacement demands. Hence, for slender/weak infill typologies
the additional shear demands due to the presence of infills, given in Equation (8.4), may not
always govern the shear design, in particular at the bottom storeys of taller buildings and/or in the
case of high seismicity. At the same time, in the upper storeys, particularly for buildings of lower
height and/or in the case of lower or intermediate seismicity, shear demands imposed due to the
presence of infills have exceeded considerably the values obtained from other design criteria.
Clearly, for the case of strong infill, such situation can be expected even more frequently, causing
a more pronounced influence on final design choices. Therefore, it is important to stress that,
according to the current design procedure, shear demands induced by local effects due to the
presence of infills are determined independently of the position along the height of the building and
other design conditions, although the assumption that all RC columns adjacent to masonry infills
should be verified for the same maximum strut forces does not always seem rational, since actual
column shear demands vary significantly throughout the structure.
Accordingly, it has been pointed out (Hak et al. [2013c]) that, accepting the assumption that
additional shear is imposed on the columns of the frame due to the activation of a diagonal strut
force in the masonry panel, the full strut action is achieved only after a certain inter-storey demand
has been reached. Hence, in regions of lower seismicity and in particular in upper storeys of
buildings, where displacements imposed by horizontal seismic actions may not be excessive, the
design of columns for the maximum masonry infill strength may not always be feasible. In fact,
leading to uniform demands for all columns in the frame, adjacent to infills of the same strength
and stiffness characteristics, such rigorous verification seems not to be widely applicable in
practise. Therefore, to obtain a strut force activation coefficient aw,j, given in Equation (8.8),
representing the portion of activated infill strength along the building height with respect to the
horizontal resistance of the strut Fw adopted in the model, the average maximum horizontal
component of the achieved strut force Fw,act,j has been evaluated in each storey j from nonlinear
time-history analyses for varying building configurations, peak ground accelerations, ductility
classes and infill typologies, according to Equation (8.9).
Fw ,act , j
aw, j = (8.8)
Fw
1 10
Fw ,act , j = ∑ Fw ,max,i , j
10 i =1
(8.9)
The results have revealed a significant reduction of the strut force activation in the upper quarter of
the building height and a considerable increase of activated strut forces for higher ground motion
intensities, while for increasing infill strength and stiffness properties substantially lower portions of
strut forces have been activated. Furthermore, in order to estimate effective column shear
demands, the average maximum difference between the horizontal component of the activated
strut force and the corresponding shear in the column has also been found from analyses of
selected case studies. The obtained results have shown that the effective column shear demands
are regularly slightly lower than the activated strut forces; hence the verification of the column for
the activated strut force instead of the effective column shear demand can be considered safe-
sided, but not too conservative. Actually, also the code provisions rationally require the verification
of the column for the strut force action neglecting the eventual presence of shear forces in the
columns, considering the strut, however, always fully activated.
Consequently, improvements of the current measures to prevent local effects due to the presence
of infills in full contact along the height of vertical structural elements as a function of design
parameters and infill properties, which are considered to be applicable also to the case of strong
infill, have been proposed (Hak et al. [2013a, 2013c]). In particular, showing that the activation of
142 Experimental and numerical seismic performance of strong clay masonry infills
the strut force aw,j can be expressed as a function of a single parameter, i.e., the corresponding
average inter-storey drift demand δw,j of the infilled frame, following a similar trend for different infill
typologies and building configurations, the simplified relationship given in Equation (8.10), based
on a linear approximation by parts of the results obtained from nonlinear analyses, has been
introduced.
2.1δ w , j δ DLS δ w , j ≤ δ DLS / 3
aw, j = 0.375δ w , j δ DLS + 0.575 δ DLS / 3 < δ w , j ≤ δ DLS (8.10)
0.0.5δ w , j δ DLS + 0.9 δ DLS < δ w , j ≤ 2δ DLS
1.0 δ w , j > δ DLS
Given that the correlation of strut force activation and in-plane drift demands strongly depends on
the assumed stress-strain relationship representing the infill response, and hence, the deformation
capacity of the infill, the coefficient aw,j is expressed in function of the characteristic drift δDLS
assigned to the attainment of damage limitation performance conditions. The strut force activation
coefficient aw,j in function of the in-plane drift demand for the infilled structural configuration δw,j,
assuming values of δDLS equal to 0.50% and 0.30% for strong (see Section 5.1.1) and
slender/weak infill (see Hak et al. [2013b]), respectively, is presented in Figure 8.4a and Figure
8.4b. Assuming that the inter-storey drifts of the infilled structural configuration δw,j can be
predicted, as discussed in Section 7.3.3, the corresponding expected strut force activation aw,j can
be evaluated from a simplified expression, such as according to Equation (8.10).
(a) (b)
Figure 8.4. Strut force activation in function of average inter-storey drift: (a) Strong infill; (b) Slender/weak
infill
Given the conservative estimation that the maximum column shear demand due to infill strut action
is equal to the activated horizontal component of the strut force Fw,act,j, rather than the effective
column shear, the shear demand VC,Ed,w,a may be estimated, according to Equation (8.11), as the
product of the strut force activation coefficient aw,j and the predicted maximum horizontal strut force
Fw.
VC , Ed ,w ,a = a w , j Fw (8.11)
The maximum horizontal strut force Fw should possibly be evaluated based on a model accounting
for different infill failure modes, or may be estimated following other simplified approaches, e.g.
such as given by Equation (8.5). Finally, the capacity design verification has to be accomplished
along the contact length, evaluating VC,Ed,M from Equation (8.7), in order to determine the governing
shear demand VC,Ed,l,a from Equation (8.12).
VC , Ed ,l ,a = min (VC , Ed ,w ,a , VC , Ed , M ) (8.12)
P. Morandi, S. Hak, G. Magenes EUCENTRE 143
Research Report
Figure 8.5. (a) Position of a masonry infill in the RC structure; (b) Normalised seismic coefficient applicable
to non-structural elements (Eurocode 8)
The ratio of the design ground acceleration ag on type A ground, to the acceleration of gravity g is
denoted by α, S is the soil factor, Ta the fundamental vibration period of the non-structural element
and T1 the fundamental vibration period of the building in the relevant direction. The seismic
coefficient Sa should not be taken less than αS. The importance factor for non-structural elements
γa in the case of masonry infills may be taken equal to 1.0, while the upper limit value of the
behaviour factor qa for exterior and interior walls, partitions and façades is defined to be equal to
2.0. The seismic force may be assumed as a distributed load per unit area (kN/m2) on the surface
of a single infill panel of height hw and length Lw, according to Equation (8.15).
Fa
wa = (8.15)
hw L w
144 Experimental and numerical seismic performance of strong clay masonry infills
Note that the seismic coefficient Sa accounts contemporarily for the dynamic amplification of
accelerations along the storey height and for the influence of the initial vibration characteristics of
the non-structural element with respect to the structure. For a very rigid non-structural element,
i.e., Ta ≈ 0, the expression multiplying αS in Equation (8.14) results to be equal to (1+1.5z/H). This
coefficient accounts for the dynamic amplification of accelerations along the height of the building,
which are assumed to increase linearly and result in top floor accelerations being 2.5 times larger,
with respect to those at the bottom. The fundamental period Ta of the masonry infill in the out-of-
plane direction can be calculated according to the expression for the case of single vertical
bending response with hinged ends given in Equation (8.16), where mw is the mass of the infill per
unit height, Ewv is the vertical modulus of elasticity of masonry and Iwy the moment of inertia about
the longitudinal axis of the horizontal cross section of the panel.
2h w 2 m w
Ta = (8.16)
π Ewv I wy
Clearly, this simplified approach for the approximation of the seismic action, in function of the
seismic coefficient Sa depends also on the estimation of the fundamental elastic period of the
structure T1, in the direction orthogonal to the plane of the infill. For the evaluation of T1, the seismic
codes allow the application of expressions based on methods of structural dynamics (e.g. the
Rayleigh method) or simplified empirical formulae, as discussed in Section 2.4.1. The normalised
seismic coefficient Sa/αS as a function of the ratio of periods Ta/T1 is shown in Figure 8.5b,
illustrating that for structures having a higher fundamental period, normally lower force demands
are obtained, since Ta/T1 is expected to be smaller than 1.0. Even though the out-of-plane
verification is performed at the ultimate limit state, when masonry infills are expected to achieve a
certain extent of damage, without questioning requirements related to life safety, the assumption of
a fundamental period corresponding to the bare structure may not be safe-sided, since a reduced
but still important contribution to the structural stiffness is expected to be induced by the infills.
Nevertheless, since the fundamental vibration period Ta of a confined infill in the out-of-plane
direction is expected to be rather low, with exception of very slender infills, for the majority of
practical design situations the seismic coefficient Sa should not substantially be influenced by the
approximate evaluation of the fundamental period of the structure.
For out-of-plane verifications in the design of new RC structures with masonry infills the direct
application of expressions for the calculation of the masonry infill resistance based on the arching
action mechanism, as introduced in Section 5.2.1, entails that the masonry is principally expected
to be undamaged when subjected to out-of-plane actions. For the safety verification of buildings
subjected to seismic actions carried out at the ultimate limit state, commonly required by modern
code regulations, such assumption is however not suitable, since simultaneous actions in two
orthogonal directions are usually imposed on the structure. Consequently, in order to control
efficiently the out-of-plane performance and achieve a satisfactory response in the case of
earthquake actions, the potential reduction of the out-of-plane infill strength in proportion to the
expected damage due to previous in-plane excitations should possibly be accounted for in codified
procedures. In effect, the application of the presented European code approach for the evaluation
of force demands and the simplified method defined above for the estimation of corresponding
levels of resistance to a number of case study configurations (Hak [2010]) has shown that the out-
of-plane verification for undamaged masonry infills does not present a critical issue in the design of
infilled RC structures, with the exception of slender/weak unreinforced typologies. Therefore, a
reasonable approximation of the out-of-plane capacity for masonry infills, previously damaged up
to a certain extent in-plane, is considered to be of major importance, not only for the assessment of
existing buildings, but also in the design of new RC structures with masonry infills.
1.0, δ w , j ≤ δ DLS
βa , j ra ,1 , δ DLS < δ w , j ≤ δULS
= (8.17)
r , δULS < δ w , j
a ,2
r − 1 δ / δ + 1,
( a ,1 ) w , j DLS
δ w , j ≤ δ DLS
=βa , j ra ,1 , δ DLS < δ w , j ≤ δULS (8.18)
(
( ra ,1 − ra ,2 ) δULS − δ w , j + ra ,1 , )δULS < δ w , j
Minimum expected remaining fractions of out-of-plane strength ra,1 and ra,2 assumed to correspond
to the attainment of damage limitation and ultimate limit state performance requirements,
respectively and hence, to the related values of drift δDLS and δULS, have also been estimated
based on experimental observations and the conservative extrapolation of obtained results, as
summarised in Table 8.1. The proposed simplified approximations of strength reduction
coefficients βa,j in comparison to the experimentally obtained results, discussed in Section 5.2.2,
are illustrated in Figure 8.6a and Figure 8.6b, for the case of strong infill and strong infill with
opening, respectively.
Table 8.1. Minimum remaining fractions of out-of-plane strength
Strong infill
Limit state
Without opening With opening
ra,1 Damage limitation 0.40 0.15
ra,2 Ultimate 0.25 0.00
(a) (b)
Figure 8.6. Simplified out-of-plane resistance reduction coefficient βa,j: (a) Strong infill; (b) Strong infill with
opening
A similar interpretation of experimental results has previously been accomplished for the case of
unreinforced and lightly reinforced slender/weak clay masonry infills (Morandi et al. [2013b]).
Nevertheless, due to the limited available data, in particular for the case of strong infill with
146 Experimental and numerical seismic performance of strong clay masonry infills
opening, the proposed strength reduction model may be further investigated, in order to account
more precisely for the related response mechanisms.
As a result, for the out-of-plane safety verification of masonry infills in the design of new RC
structures, complying with European seismic code regulations (CEN [2004a], NTC08 [2008]), the
proposed strength reduction coefficient βa,j may be applied to estimate the reduced out-of-plane
resistance, accounting for a certain level of previous in-plane damage that is likely to be sustained
by the infill. Given that the infill resistance verification is commonly carried out at the ultimate limit
state, the evaluation of an approximate in-plane drift profile of the infilled frame configuration for
the corresponding seismic demand is required. As a matter of fact, assuming for the evaluation of
the out-of-plane infill strength reduction the design drift of the bare frame configuration, the given
procedure may be overly conservative, since reduced drift demands are expected for the
corresponding infilled frame. Hence, adopting the approach introduced in Section 7.3.3, allowing
the correlation of drift demands corresponding to bare and infilled structures, the expected in-plane
response can be estimated starting from a linear analysis of the bare structure, as commonly
accomplished in design practise. Clearly, the in-plane drift demand and the corresponding infill
damage may vary along the height of the building and the resulting out-of-plane strength
degradation, expressed through the strength reduction coefficient βa,j has to be found for each
storey j.
Assuming that the full contact between the infill and the surrounding structure can be preserved
and the arching action remains active, the out-of-plane resistance of a damaged infill wR,β can be
found from Equation (8.19), reducing the strength of the undamaged panel, evaluated according to
Equation (5.7) in terms of lateral pressure wR, applying the corresponding reduction coefficient βa,j.
Thus, in order to satisfy the out-of-plane resistance verification for a masonry infill that is expected
to sustain a certain level of in-plane damage, as given in Equation (8.20), the seismic force Fa,
expressed as equivalent pressure wa in Equation (8.5), acting on the masonry infill, needs to be
smaller than the corresponding out-of-plane resistance wR,β.
w R ,β = β a , j w R (8.19)
wa < wR ,β (8.20)
P. Morandi, S. Hak, G. Magenes EUCENTRE 147
Research Report
9 CONCLUSIONS
9.1 Summary
In this study, the seismic behaviour of strong masonry infills has been investigated based on
experimental test results and the corresponding response of infilled RC building has been
evaluated in function of various design parameters. Particular attention has been devoted to the
infill performance and possible improvements in current European design procedures, relying on
the newly available experimental data, in order to achieve effective infill damage control without
penalising unnecessarily the RC structure.
Following the introductive chapter of this research report, including the motivation and objectives of
the research carried out, as well as the scope and a brief outline, Chapter 2 briefly reviewed the
literature related to the performance of masonry infills in RC structures. Reported field
observations, findings of important experimental and numerical studies, as well as relevant seismic
code provisions currently adopted in Europe were addressed.
In Chapter 3 the experimental campaign carried out in the laboratories of the Department of Civil
Engineering and Architecture of the University of Pavia and EUCENTRE of Pavia was introduced,
the design of the new experimental setup for in-plane and out-of-plane tests on bare and infilled
RC specimens was presented and the design of the specimens was summarised. In addition, a
description of the characterisation tests for the evaluation of mechanical properties for the adopted
material components was provided, including the obtained results.
Subsequently, Chapter 4 the adopted test protocol was presented for both in-plane and out-of-
plane tests on single-storey, single-bay bare and infilled RC frame specimens, the experimental
results were summarised in terms of force-displacement response and observations gathered
during the tests, in particular regarding the damage propagation, were discussed.
Presenting most important issues regarding the interpretation of the obtained experimental results,
Chapter 5 addressed the definition of infill performance levels, the calibration of a simple numerical
model for the representation of infills, possible local effects due to the presence of infills, as well as
out-of-plane response mechanisms and the related out-of-plane strength reduction due to previous
in-plane damage.
In order to introduce the framework of the numerical study of this work, Chapter 6 provided a
description of the considered prototype building configurations, referring to 2-dimensional RC
frame structures analysed in a series of extensive parametric studies. In addition, modelling
assumptions adopted in the elastic structural models used to evaluate design action effects in
terms of forces and displacements, as well as in the inelastic models used for nonlinear time-
history analyses were discussed, and the earthquake records selected to match the design
response spectra were presented.
In Chapter 7 the in-plane response of the 2-dimensional bare and infilled case study RC frame
systems was evaluated. The performance of the masonry infills in terms of damage was estimated
at both the damage limitation and the ultimate limit state in order to assess to which extent the
current design procedure is able to control the infill damage. Moreover, drift demands for bare and
infilled structures were compared and a simplified method for the prediction of drift demands
corresponding to infilled frames based on a relative infill vs. structural stiffness coefficient was
verified for the case of strong infills.
Dealing with possible improvements in the current design approach for new RC structures, in
particular regarding a more effective infill damage control, Chapter 8 discussed the limitation of
inter-storey drift demands in function of infill properties and design conditions, based on the
correlation of drift-demands for bare and infilled structures. Furthermore, possible verifications of
the out-of-plane infill resistance in function of previous in-plane damage were considered with
reference to the obtained experimental results, and the evaluation of increased column shear
demands was discussed.
148 Experimental and numerical seismic performance of strong clay masonry infills
9.2 Conclusions
The presented study has achieved three main research objectives: (i) to increase the
understanding of the seismic behaviour of strong clay block masonry infills based on experimental
test results, (ii) to assess the response of RC structures with strong clay block masonry infills in
function of different design conditions based on nonlinear analyses of prototype buildings, and
consequently (iii) to propose a more effective approach to the design of new RC buildings,
ensuring adequate damage control for masonry infills.
Related to the first research objective, within the framework of the experimental study carried out
with the aim to improve the understanding of the cyclic response of rigid masonry infills, commonly
adopted for building enclosure systems in seismic regions, a strong type of infill constructed using
tongue and groove clay masonry blocks has been considered. For the case of fully infilled frame
configurations, the in-plane response was found to be characterised by a considerable deformation
capacity and the gradual attainment of a relatively high peak resistance, accompanied by slow and
rather low degradation of strength, resulting in a substantial residual resistance, even at higher
levels of drift. The in-plane behaviour of the strong infill with a central opening (equal to one third of
the infill length) was characterised by the somewhat asymmetric response and a substantially
lower strength (equal to approximately 50% with respect to the case of full infill), accompanied by a
considerable reduction of deformation capacity. Regarding characteristic failure modes of the
strong infill, for the in-plane direction the initial development of a diagonal stepwise crack pattern
due to shear sliding in the bed joints and slight opening of the dry head joints has been identified,
followed by the creation of diagonal cracks on the masonry units, firstly in the infill corners and
subsequently throughout the panel. Due to the presence of the central opening, the creation of a
more pronounced strut action has been found in one of the partial panels, while on the opposite
side the response was accompanied by the tendency of uplift. Furthermore, performance levels
with reference to infill damage, in accordance with the design objectives commonly assigned to
damage limitation and ultimate limit states, have been defined in function of attained in-plane drift
values. Based on a simplified representation of the average infill response and observations during
the test, the attainment of operational, damage limitation and ultimate limit state conditions at drifts
equal to 0.30%, 0.50% and 1.75%, respectively, has been found for the case of full infill. For the
strong infill with opening corresponding values have been defined equal to 0.20%, 0.35% and
1.0%, respectively. In order to implement the obtained experimental results in a numerical model to
be used for extensive parametric analyses of RC buildings, a simple single-strut model has been
found appropriate, adopted using a carefully calibrated cyclic stress-strain relationship to represent
as closely as possible the obtained infill response.
For the case of fully infilled frame configurations, observations from the out-of-plane tests,
accomplished under double-bending conditions on specimens previously damaged in-plane,
indicated that for such strong masonry infill typologies, adherent to the RC frame, the out-of-plane
stability does not present a critical issue. The same conclusion clearly does not apply to the strong
infill with opening. Moreover, the formation of a resistance mechanism based on two-directional
arching action has been found for the fully infilled configuration and a typical out-of-plane failure
mechanism has been identified, characterised by the opening of a predominant horizontal crack at
mid-height of the panel and the creation of a stepwise crack pattern, starting from the central crack
and developing diagonally towards the corners of the infill. The out-of-plane response of a
previously undamaged masonry strip tested under vertical single-bending conditions has shown to
comply reasonably with the commonly adopted simplified approach for the evaluation of the
vertical arching action. Based on the results of in-plane and subsequent out-of-plane tests, a
strength reduction model relating the out-of-plane resistance to corresponding in-plane drifts,
accounting for the in-plane deformation capacity of the infill, has been established.
Considering the current force-based design approach for regular RC structures with masonry
infills, commonly adopted in design practise based on the analysis of bare frame structural
configuration, in principle appropriate, findings of previous investigations, which have
demonstrated that the current European design approach for RC structures is not always able to
effectively control the damage to traditional masonry infills, have been extended for the case of
strong infills. Hence, referring to the second objective of the study, a detailed evaluation of the
P. Morandi, S. Hak, G. Magenes EUCENTRE 149
Research Report
seismic response for regular prototype RC frames with different infill layouts has revealed that for
the strong infill typology a satisfactory infill response can be expected up to values of ULS design
PGA equal to 0.35g, although exceptions have been found in the case of BO structural
configurations (external infills with opening and bare central bay). Due to the relatively high
deformation capacity of strong infills without openings, for building configurations with a dense infill
distribution, the adopted design drift limit at the damage limitation limit state (δSLD = 0.50%),
imposed on the bare frame, may even be considered overly conservative, since lower drift
demands could be obtained for the infilled structure. However, in the case of scarcely distributed
infills or of a predominant use of infills with openings, even for strong infills, the drift limit of 0.50 %
could not be always suitable to guarantee the fulfilment of the performance criteria and to limit the
in-plane damage of the infills, in particular at a high level of seismic action.
Based on the comparison of inter-storey drift demands obtained for different bare and infilled
building configurations from nonlinear analyses, a correlation in function of structural stiffness and
infill strength and stiffness properties has been established. Consequently, the possibility to
evaluate a safe-sided approximate prediction of inter-storey drift demands for infilled RC structures
from a simplified relationship, based on analyses which are carried out assuming bare structural
models, has been validated. Such approach requires for each storey j the definition of the density-
stiffness coefficient Cj, which can be evaluated from assumed elastic structural properties, knowing
the expected infill distribution commonly available from architectural layouts and the most relevant
properties of masonry typologies to be adopted. The obtained results have shown that the density-
stiffness coefficient Cj can effectively account for the contribution of different amounts of infill
present in the building, reproducing adequately the reduction of drift demands for the infilled
configuration in comparison to that of the bare structure, which depends also on the deformation
capacity of the infill. Moreover, the influence of structural stiffness properties on the response of
the infilled configuration can be successfully captured, reflecting that a stiffer RC structural system
will be less affected by the presence of infill than a more flexible configuration.
Even though the application of the proposed procedure, considering the amount of infill in each
storey, seems to be a reasonable choice in the attempt to account for the presence of soft-storey
mechanisms, no systematic analyses have been carried out regarding the response of buildings
with irregular distributions of infill along the height. Furthermore, irregular distributions of infill in
plan, possibly inducing a substantial contribution to the torsional response, have not been
considered within the scope of this study; hence, the obtained results are thought to be applicable
to regular building configurations without significant irregularities regarding the distribution of infills
in plan or elevation.
As a result, with respect to the third objective of the study, the evaluation of approximate inter-
storey drift demands for infilled RC structures, obtained without the need to perform more
sophisticated analyses than in common building design procedures, may lead to significant
improvements regarding the inter-storey drift verifications carried out to control the damage of
masonry infills. Importantly, the iterative nature of the verification approach, implying a new value
of the density-stiffness coefficient Cj after each modification of the dimensions for load-bearing
structural elements, resulting in a new relation between drifts of bare and infilled configurations,
limits to some extent the unreasonable increase of structural stiffness due to infill damage control
criteria, imposing instead a sort of optimisation regarding the relation of structural and masonry
infill properties. With respect to current seismic design recommendations, which account for the
limitation of damage through the verification of a single bare frame drift limit at the damage
limitation state, the major advantage of the proposed approach appears to be the possibility to
verify estimated drift demands of infilled structures directly against allowable drift limits at different
performance levels corresponding to considered infill typologies. In addition, the prediction of
structural drifts for infilled configurations can be applied also in other design verifications, in
particular related to combined in-plane and out-of-plane actions and possible local effects on RC
structural elements due to the presence of masonry infills.
150 Experimental and numerical seismic performance of strong clay masonry infills
REFERENCES
Al-Chaar, G. [2002] “Evaluating strength and stiffness of unreinforced masonry infill structures”, ERDC/CERL
TR-02-1 Research Report.
Angel, R., Abrams, D., Shapiro, D., Uzarski, J., Webster, M. [1994] “Behaviour of reinforced concrete frames
with masonry infills”, University of Illinois, Urbana-Campaign, Illinois.
Asteris, P.G. [2008] “Finite element micro-modeling of infilled frames,” Electronic Journal of Structural
Engineering, No. 8, pp. 1-11.
Asteris, P.G., Antoniou, S.T., Sophianopoulos, D.S., Chrysostomou, C.Z. [2011a] “Mathematical
macromodeling of infilled frames: State of the art”, Journal of Structural Engineering, Vol. 137, No. 12,
pp. 1508-1517.
Asteris, P.G., Chrysostomou, C.Z., Giannopoulos, I.P., Smyrou, E. [2011b] “Masonry infilled reinforced
concrete frames with openings”, Proceedings of COMDYN, III ECCOMAS Thematic Conference on
Computational Methods in Structural Dynamics and Earthquake Engineering, Corfu, Greece.
Bergami, A.V. [2007] “Implementation and experimental verification of models for non-linear analysis of
masonry infilled RC frames”, PhD Dissertation, Department of Civil Engineering Sciences, Roma Tre
University, Italy.
Bertero, V.V., Brokken, S.T. [1983] “Infills in seismic resistant buildings,” Journal of Structural Engineering,
Vol. 109, No. 6, pp. 1337–1361.
Beyer, K., Bommer, J.J. [2007] “Selection and scaling of real accelerograms for bi-directional loading: A
review of current practice and code provisions”, Journal of Earthquake Engineering, Vol. 11, pp. 13-45.
Bommer, J.J., Acevedo, A.B. [2004] “The use of real earthquake accelerograms as input to dynamic
analysis”, Journal of Earthquake Engineering, Vol. 8, S. 1, pp. 43-91.
Braga, F., Manfredi, V., Masi, A., Salvatori, A., Vona, M. [2011] “Performance of non-structural elements in
RC buildings during the L’Aquila, 2009 earthquake”, Bulletin of Earthquake Engineering, Vol. 9, pp. 307-
324.
Calvi, G.M., Bolognini D. [1999] “Seismic response of R.C. frames infilled with weakly reinforced hollow
masonry panels”, Research Report, University of Pavia, Department of Structural Mechanics, Pavia,
Italy.
Calvi, G.M., Bolognini D. [2001] “Seismic response of RC frames infilled with weakly reinforced masonry
panels”, Journal of Earthquake Engineering, Vol. 5, No. 2, pp. 153-185.
Calvi, G.M., Magenes, G., Pampanin, S. [2002] “Experimental test on a three storey RC frame designed for
gravity only”, Proceedings of 12th European Conference on Earthquake Engineering, London, UK.
Carr, A.J. [2007] Ruaumoko Manual, University of Canterbury, Cristchurch, New Zealand.
CEN [1998] Methods of test for masonry - Determination of compressive strength, EN1052-1, European
Committee for Standardisation, Brussels, Belgium.
CEN [2000] Methods of test for masonry units – Determination of compressive strength, EN 772-1, European
Committee for Standardisation, Brussels, Belgium.
152 Experimental and numerical seismic performance of strong clay masonry infills
CEN [2002] Eurocode 1 - Actions on structures, Part 1-1: General actions - Densities, self-weight, imposed
loads for buildings, EN 1991-1-1, European Committee for Standardisation, Brussels, Belgium.
CEN [2004a] Eurocode 8 - Design of structures for earthquake resistance, Part 1: General rules, seismic
actions and rules for buildings, EN 1998-1, European Committee for Standardisation, Brussels,
Belgium.
CEN [2004b] Eurocode 6 - Design of masonry structures, Part 1-1: Common rules for reinforced and
unreinforced masonry structures, European Committee for Standardisation, ENV 1996-1-1, Brussels,
Belgium.
CEN [2004c] Eurocode 2 - Design of concrete structures, Part 1-1: General rules and rules for buildings, EN
1992-1-1, European Committee for Standardisation, Brussels, Belgium.
CEN [2005] Specification for masonry units - Part 1: Clay masonry units, EN 771-1/A1, European Committee
for Standardisation, Brussels, Belgium.
CEN [2006] Methods of test for mortar for masonry - Determination of flexural and compressive strength of
hardened mortar, EN 1015-11/A1, European Committee for Standardisation, Brussels, Belgium.
CEN [2007] Methods of test for masonry - Determination of initial shear strength, EN 1052-3/A1, Brussels,
Belgium.
CEB (Comité Euro-International du Béton) [1996] RC frames under earthquake loading – state of the art
report, Thomas Telford, London, UK.
Combescure, D., Pires, F., Cerqueira, P., Pegon, P. [1996] “Test on masonry infilled RC frames and its
numerical interpretation,”Proceedings of 11th World Conference on Earthquake Engineering. Acapulco,
Mexico.
Crisafulli, F.J. [1997] “Seismic behaviour of reinforced concrete structures with masonry infills”, PhD
Dissertation, Department of Civil Engineering, University of Canterbury, New Zealand.
da Porto, F., Guidi, G., Dalla Benetta, M., Verlato, N. [2012] “Sistemi costruttivi e risultati sperimentali”,
Reluis Research Report (in Italian), University of Padova, Padova, Italy.
Dawe, J.L., Seah, C.K. [1988] “Lateral load resistance of masonry panels in flexible steel frames,”
Proceedings of 8th International Brick and Block Masonry Conference, Dublin, Ireland.
Decanini, L.D., Bertoldi, S.H., Gavarini, C. [1993] “Telai tamponati soggetti ad azione sismica, un modello
semplificato: confronto sperimentale e numerico” (in Italian), Proceedings of VI Convegno nazionale
ANIDIS, Perugia, Italy.
Decanini, L.D., Gavarini, C., Bertoldi, S.H., Mollaioli, F. [1994] “Modelo simplificado de Paneles de
Mamposteria con Aberturas incluidos en marcos de concreto reforzado y metalicos. Comparacion y
calibracion con resultados experimentales y numericos” (in Spanish), Proceedings of 9th International
Seminar on Earthquake Prognostics, San Josè, Costa Rica.
Decanini L.D., Liberatore D., Liberatore L., Sorrentino L. [2012] Preliminary Report on the 2012, May 20,
Emilia Earthquake, v.1, http://www.eqclearinghouse.org/2012-05-20-italy-it/
Decanini, L.D., Mollaioli, F., Mura, A. and Saragoni, R. [2004a] “Seismic performance of masonry infilled RC
frames,” Proceedings of the 13th World Conference on Earthquake Engineering, Vancouver, Canada.
P. Morandi, S. Hak, G. Magenes EUCENTRE 153
Research Report
De Sortis, A., Bazzurro, P., Mollaioli, F. Bruno, S. [2007] “Influenza delle tamponature sul rischio sismico
degli edifici in calcestruzzo armato” (in Italian), Proceedings of XII Convegno nazionale ANIDIS, Pisa,
Italy.
Dolce, M., Manfredi, G. [2009] “Research needs in earthquake engineering highlighted by the 2009 l’Aquila
earthquake”, in Manfredi, G., Dolce, M. (eds.) The state of Earthquake Engineering Research in Italy:
the ReLUIS-DPC 2005-2008 Project, pp. 469-480.
Dolšek, M., Fajfar, P. [2006] “Simplified seismic assessment of infilled reinforced concrete frames,”
Proceedings of 1st European Conference on Earthquake Engineering and Seismology, Geneva,
Switzerland.
Drysdale, R.G., Hamid, A.A., Baker, L.R. [1999] Masonry structures: behavior and design, The Masonry
Society, Boulder, Colorado.
Durrani, A.J., Haider, S. [1996] “Seismic response of RC frames with unreinforced masonry infills”,
Proceedings of 14th World Conference on Earthquake Engineering, Acapulco, Mexico.
Emori, K., Schnobrich, W.C. [1978] “Analysis of reinforced concrete frame-wall structures for strong motion
earthquakes”, Civil Engineering Studies, Structural Research Series No. 434, University of Illinois,
Urbana-Campaign, Illinois.
Fardis, M.N. [2000] “Design provisions for masonry-infilled RC frames”, Proceedings of 12th World
Conference on Earthquake Engineering, Auckland, New Zealand.
Fardis, M.N. [2006] “Seismic design issues for masonry-infilled RC frames”, Proceedings of 1st European
Conference on Earthquake Engineering and Seismology. Geneva, Switzerland.
Fardis, M. N. Bousias, S. N., Franchioni, G., Panagiotakos, T.B. [1999] “Seismic response and design of RC
structures with plan-eccentric masonry infills”, Earthquake Engineering and Structural Dynamics, Vol.
28, pp. 173-191.
Fardis, M.N., Panagiotakos, T.B. [1997a] “Seismic design and response of bare and infilled reinforced
concrete buildings. Part I: Bare structures”, Journal of Earthquake Enneering, IC Press, Vol. 1, No. 1,
pp. 219-256.
Fardis, M.N., Panagiotakos, T.B. [1997b] “Seismic design and response of bare and infilled reinforced
concrete buildings. Part II: Infilled structures”, Journal of Earthquake Enneering, IC Press, Vol. 1, No. 3,
pp. 475-503.
Fiorato, A.E., Sozen, M.A., Gamble, W.L. [1970] “An investigation on the interaction of reinforced concrete
frames with masonry filler walls”, Report UILU-ENG-70-100, Department of Civil Engineering, University
of Illinois, Urbana-Campaign, Illinois.
Flanagan, R.D., Bennett, R.M. [1999] “Bidirectional behaviour of structural clay tile infilled frames”, Journal of
Structural Engineering, Vol. 125, No. 3, pp. 236-244.
Franchioni, G. [1997] “Shaking table seismic tests for the assessment of the influence of unsymmetrical infills
in reinforced concrete buildings”, Report to Europen Commission, ISMES, Project STR-8721, Bergamo,
Italy.
Giberson, M.F. [1967] “The response of nonlinear multi-story structures subjected to earthquake excitation”,
EERL Report, California Institute of Technology, Pasadena, California.
154 Experimental and numerical seismic performance of strong clay masonry infills
Ghosh, A.K., Amde, A.M. [2001] “Finite element analysis of infilled frames”, Journal of Structural
Engineering, Vol. 128, pp. 881-889.
Hak, S. [2010] “Seismic behaviour of clay masonry infills in framed structurs”, MSc Thesis, Rose School,
IUSS Pavia, Pavia, Italy.
Hak, S., Morandi, P., Magenes, G., Sullivan, T. [2012] “Damage control for clay masonry infills in the design
of RC frame structures”, Journal of Earthquake Engineering, Vol. 16, S1, pp. 1-35.
Hak, S., Morandi, P., Magenes, G. [2013a] “Damage Control of Masonry Infills in Seismic Design”, Eucentre
Research Report, No. 01.2013, IUSS, Pavia.
Hak, S., Morandi, P., Magenes, G. [2013b] “Evaluation of infill strut properties based on in-plane cyclic tests”,
Građevinar, Vol. 65, No. 6, pp. 509-522.
Hak, S., Morandi, P., Magenes, G. [2013c] “Local Effects in the Seismic Design of RC Frame Structures With
Masonry Infills”, Procediings of 4th ECCOMAS Conference on Computational Methods in Structural
Dynamics and Earthquake Engineering, Kos Island, Greece.
Hak, S., Morandi, P., Magenes, G. [2017] “Prediction of inter-storey drifts for regular RC structures with
masonry infills based on bare frame modelling”, Bulletin of Earthquake Engineering,
https://doi.org/10.1007/s10518-017-0210-y.
Hashemi, A., Mosalam, K. M. [2007] “Seismic evaluation of reinforced concrete buildings including effects of
masonry infill walls”, PEER Report 2007/100, Pacific Earthquake Engineering Research Center,
University of California, Berkeley, California.
Hermanns, L., Fraile, A., Alarcón, E., Álvarez, R. [2014] “Performance of buildings with masonry infill walls
during the 2011 Lorca earthquake”, Bulletin of Earthquake Engineering, Vol. 12, No. 5, pp. 1977-1997.
Holmes, M. [1961] “Steel frames with brickwork and concrete filling”, Proceedings of the Institute of Civil
Engineers, Vol. 19, pp. 473-478.
Holmes, M. [1963] “Combined loading on infilled frames”, Proceedings of the Institute of Civil Engineers, Vol.
25, pp. 31-38.
Iervolino, I., Cornell, C.A. [2005] “Record Selection for Nonlinear Seismic Analysis of Structures”, Earthquake
Spectra, Vol. 21, No. 3, pp. 685-713.
Iervolino, I., Maddaloni, G., Cosenza E. [2008] “Eurocode 8 compliant real record sets for seismic analysis of
structures”, Journal of Earthquake Engineering, Vol. 12, pp. 54-90.
Iervolino, I., Galasso, C., Cosenza, E. [2010] “REXEL: computer aided record selection for code-based
seismic structural analysis”, Bulletin of Earthquake Engineering, Vol. 8, pp. 339-362.
Ioannou, I., Borg, R., Novelli, V., Melo, J., Alexander, D., Kongar, I., Verrucci, E., Cahill, B., Rossetto, T.
[2012] The 29th May 2012 Emilia Romagna Earthquake, EPICentre Field Observation Report,
http://www.eqclearinghouse.org/2012-05-20-italy/reports/
P. Morandi, S. Hak, G. Magenes EUCENTRE 155
Research Report
Kadysiewski, S., Mosalam, K. M. [2009] “Modeling of unreinforced masonry infill walls considering in-plane
and out-of-plane interaction”, PEER Report 2008/102, Pacific Earthquake Engineering Research
Center, University of California, Berkeley, California.
Kakaletsis, D.J., Karayannis, C.G. [2008] “Influence of masonry strength and openings on infilled RC frames
under cyclic loading”, Journal of Earthquake Engineering, Vol. 12, pp. 197-221.
Kappos, A.J., Stylianidis, K.C., Michailidis, C.N. [1998] “Analytical models for brick masonry infilled RC
frames under lateral loading”, Journal of Earthquake Engineering, Vol. 2, No. 1, pp. 59 - 87.
Kaushik, H.B., Rai, D.C., Jain, S.H. [2006] “Code Approaches to seismic design of masonry-infilled
reinforced concrete frames: A state-of-the-art review”, Earthquake Spectra, Vol. 22, No. 4, pp. 961-983.
Klingner, R.E., Bertero, V.V. [1976] “Infilled frames in earthquake resistant construction”, Earthquake
Engineering Research Centre, Report No. EERC 76-32, University of California, Berkeley, USA.
Klingner, R.E., and V.V. Bertero [1978] “Earthquake Resistance of Infilled Frames”, Journal of the Structural
Division, Vol. 104, No. 6, pp. 973-989.
Kose, M.M. [2009] “Parameters affecting the fundamental period of RC buildings with infill walls”,
Engineering Structures, Vol. 31, pp. 93-102.
Koutromanos, I., Stavridis, A., Shing, P.B., Willam, K. [2011] “Numerical modeling of masonry-infilled RC
frames subjected to seismic loads”, Computers and Structures, Vol. 89, pp. 1026–1037.
Liauw, T. C., Kwan, K. [1984] “Nonlinear behaviour of non-integral infilled frames”, Computers and
Structures, Vol. 18, pp. 551 – 560.
Luco, N., Bazzurro, P. [2007] “Does amplitude scaling of ground motion records result in biased nonlinear
structural drift responses?”, Earthquake Engineering and Structural Dynamics, Vol. 36, pp. 1813-1835.
Luzi, L., Hailemikael, S., Bindi, D., Pacor, F., Mele, F., Sabetta, F. [2008] “ITACA (ITalian ACcelerometric
Archive): A Web Portal for the Dissemination of Italian Strong-motion Data”, Seismological Research
Letters, Vol. 79, No. 5, pp. 716–722.
Madan, A., Reinhorn, A.M., Mander, J.B., Valles, R.E. [1997] “Modelling of masonry infill panels for structural
analysis”, Journal of Structural Engineering, No. 123, pp. 1295-1302.
Magenes, G., Bracchi, S., Graziotti, F., Mandirola, M., Manzini, C.F., Morandi, P., Palmieri, M., Penna, A.,
Rosti, A., M. Rota, Tondelli, M. [2012] “Preliminary damage survey to masonry structures after the May
2012 Emilia earthquakes”, v.1, http://www.eqclearinghouse.org/ 2012-05-20-italy/ reports/
Magenes, G., Pampanin, S. [2004] “Seismic reponse of gravity-load design frames with masonry infills”,
Proceedings of 13th World Conference on Earthquake Engineering, Vancouver, Canada.
Mainstone, R.J. [1971] “On the stiffnesses and strengths of infilled frames”, Proceedings of the Institution of
Civil Engineers, Suppl. (IV), pp. 57-90.
Markulak, D., Radić, I., Sigmund, V. [2013] “Cyclic testing of single bay steel frames with various types of
masonry infill”, Engineering Structures, Vol. 51, pp. 267-277.
Masi, A., Vona, M., Mucciarelli, M. [2011] “Selection of Natural and Synthetic Accelerograms for Seismic
Vulnerability Studies on Reinforced Concrete Frames”, Journal of Structural Engineering, Vol. 137, No.
3, pp. 367-378.
156 Experimental and numerical seismic performance of strong clay masonry infills
McDowall, E.L., McKee, K.E., Sevin, E. [1956] “Arching action theory of masonry walls”, Journal of the
Structural Divison, Proceedings of ASCE, pp. 915/1-915/18.
Mehrabi, A.B., Shing, P.B., Schuller, M.P., Noland, J.L. [1996] “Experimental evaluation of masonry infilled
frames”, Journal of Structural Engineering, Vol. 122, No. 2, pp. 228-237.
Mohammadi, M., Akrami, V., Mohammadi-Ghazi, R. [2011] “Methods to Improve Infilled Frame Ductility”,
Journal of Structural Engineering, Vol. 137, No. 6, pp. 646-653.
Morandi, P., Hak, S., Magenes, G. [2011a] “Comportamento sismico delle tamponature in laterizio in telai in
c.a.: definizione dei livelli prestazionali e calibrazione di un modello numerico” (in Italian), Proceedings
of XIV Convegno nazionale ANIDIS, Bari, Italy.
Morandi, P., Hak, S., Magenes, G. [2011b] “Comportamento sismico delle tamponature in laterizio in telai in
c.a.: analisi numeriche su edifici ed implicazioni progettuali” (in Italian), Proceedings of XIV Convegno
nazionale ANIDIS, Bari, Italy.
Morandi, P., Albanesi, L., Magenes, G. [2013a] “In-plane experimental response of masonry walls with thin
shell and web clay units”, Proceedings of Vienna Congress on Recent Advances in
Earthquake Engineering and Structural Dynamics, Vienna, Austria.
Morandi, P., Hak, S., Magenes G. [2013b] “Simplified Out-of-plane Resistance Verification for Slender Clay
Masonry Infills in RC Frames”, Proceedings of XV Convegno nazionale ANIDIS, Padova, Italy.
Negro, P., Pinto, A.V., Verzeletti, G., Magonette, G.E. [1996] “PsD test on four-storey RC building designed
according to Eurocodes”, Journal of Structural Engineering, Vol. 122, pp. 1409-1417.
Negro, P., Verzeletti, G. [1996] “Effect of infills on the global behaviour of RC frames: energy considerations
from pseudodynamic tests”, Earthquake Engineering and Structural Dynamics, Vol. 25, pp. 753-773.
Negro, P., Colombo, A. [1997] “Irregularities induced by non-structural masonry panels in framed buildings”,
Engineering Structures, Vol. 7, No. 7, pp. 576 - 585.
NTC08 [2008] Norme tecniche per le costruzioni (in Italian), D.M. 14 Gennaio 2008, Ministero delle
Infrastrutture, S.O. No. 30 alla G.U. del 4.2.2008, No. 29., Rome, Italy.
NTC08 Circolare esplicativa [2009] Istruzioni per l'applicazione delle Norme tecniche per le costruzioni di cui
al D.M. 14 gennaio 2008 (in Italian), Ministero delle Infrastrutture, S.O. No. 27 alla G.U. del 26.02.2009,
No. 47., Rome, Italy.
NTC – draft [2017] Draft of the new Norme tecniche per le costruzioni (in italian), Ministero delle
Infrastrutture e dei Trasporti, update 2017.
Oliaee M., Morandi P., Magenes G. [2015a] “Macro-model calibration of a strong clay masonry infill to in-
plane cyclic tests”, Proceedings of the 5th International Conference on Computational Methods in
Structural Dynamics and Earthquake Engineering, 25-27 May 2015, Crete Island, Greece.
Oliaee M., Morandi P., Magenes G. [2015b] “In-plane seismic performance of a strong masonry infill”,
Proceedings of the 5th Recent Advances in Mechanics and Materials in Design, 26-30 July 2015, Ponta
Delgada/Azores, Portugal.
Otani, S. [1981] “Hysteresis models of reinforced concrete for earthquake response analysis”, Journal of the
Faculty of Engineering, University of Tokyo, Vol. XXXVI, No. 2, pp. 125-159.
P. Morandi, S. Hak, G. Magenes EUCENTRE 157
Research Report
Pacor, F., Paolucci, R., Luzi, L., Sabetta, F., Spinelli, A., Gorini, A., Nicoletti, M., Marcucci, S., Filippi, L.
Dolce, M. [2011] “Overview of the Italian strong motion database ITACA 1.0”, Bulletin of Earthquake
Engineering, Vol. 9, No. 6, pp. 1723–1739.
Pampanin, S., Calvi, G.M., Moratti, M. [2002] “Seismic behaviour of RC beam-column joints designed for
gravity loads”, Proceedings of 12th European Conference on Earthquake Engineering, London, UK.
Parducci, A., Mezzi, M. [1980] “Repeated horizontal displacement of infilled frames having different stiffness
and connection systems - Experimental analysis”, Proceedings of the 7th World Conference on
Earthquake Engineering, Istanbul, Turkey.
Paulay, T., Priestley, M.J. [1992] Seismic design of reinforced concrete and masonry buildings. John Wiley &
Sons, USA.
Penna, A. [2006] “Campagna sperimentale su telai in calcestruzzo armato con tamponamenti in calcestruzzo
cellulare e diverse soluzioni di rinforzo” (in Italian), Research report, EUCENTRE, Pavia, Italy.
Pinho, R., Crowley, H. [2009] “Revisiting Eurocode 8 formulae for periods of vibration and their employment
in linear seismic analysis”, Eurocode 8 Perspectives from the Italian Standpoint Workshop, Napoli, Italy.
Pinto, A.V., Negro, P., Pegon, P., Arede, A. [1994] “Analysis of the four-storey RC building to be tested in the
ELSA-reaction wall facility”, Proceedings of the 10th European Conference of Earthquake Engineering,
Vienna, Austria.
Polyakov, S.V. [1956] Masonry in framed buildings, Gosudalst-Vennoe Izdatel’stvo Literature po Straitel’stvu
I Arkitecture, Moscow, 1956, Tranl. G.L. Cairns, Building Research Station, Watford, Herts, UK.
Preti, M., Bettini, N., Plizzari, G. [2012] “Infill walls with sliding joints to limit infill-frame seismic interaction:
large-scale experimental test”, Journal of Earthquake Engineering, Vol. 16, pp. 125-141.
Preti, M., Migliorati, L., Giuriani, E. [2014] “Experimental testing of engineered masonry infill walls for post-
earthquake structural damage control”, Bulletin of Earthquake Engineering, pp. 1-21.
Priestley, M.J.N., Calvi, G.M., Kowalsky, M.J. [2007] Displacement-Based Seismic Design of Structures,
IUSS Press, Pavia, Italy.
Puglisi, M., Uzcategui, M., Flórez-López, J. [2009a] “Modeling of masonry of infilled frames, Part I: The
plastic concentrator”, Engineering Structures, Vol. 31, pp. 113-118.
Puglisi, M., Uzcategui, M., Flórez-López, J. [2009b] “Modeling of masonry of infilled frames, Part II: Cracking
and damage”, Engineering Structures, Vol. 31, pp. 119-124.
Ramberg, W., Osgood, W.R. [1943] “Description of stress-strain curves by three parameters”, National
Advisory Committee on Aeronautics, Technical Note 902.
Ricci, P., Manfredi, V., De Luca, F., Verderame, G.M. [2011] “6th April 2009 L’Aquila earthquake, Italy:
reinforced concrete building performance”, Bulletin of Earthquake Engineering, Vol. 9, No. 1, pp. 285-
305.
Rodrigues, H., Varum, H., Costa, A. [2010] “Simplified macro-model for infill masonry panels”, Journal of
Earthquake Engineering, Vol. 14, No. 3, pp. 390-416.
Sargin, M., Ghosh, S.K., Handa, V.K. [1971] “Effects of lateral reinforcement upon the strength and
deformation properties of concrete”, Magazine of Concrete Research, Vol. 23, No. 75-76, pp. 99-110.
158 Experimental and numerical seismic performance of strong clay masonry infills
Schmidt, T. [1989] “Experiments on the nonlinear behaviour of masonry infilled reinforced concrete frames
subjected to seismic loading”, Darmstadt Concrete, Annual Journal on Concrete and Concrete
Structures, Vol. 4, pp. 185-194.
Sextos, A.G., Katsanos, E.I., Georgiou, A., Faraonis, P., Manolis, G.D. [2011] “On the evaluation of EC8-
based record selection procedures for the dynamic analysis of buildings and bridges”, in Papadrakakis,
M., Fragiadakis, M., Lagaros, N.D. (eds.) Computational Methods in Earthquake Engineering,
Computational Methods in Applied Sciences 21, Springer Science+Business Media B.V.
Shing, P.B., Mehrabi, A.B. [2002] “Behaviour and analysis of masonry-infilled frames”, Progress in Structural
Engineering and Materials, Vol. 4, pp. 320-331.
Shapiro, D., Uzarski, J., Webster, M., Angel, R., Abrams, D. [1994] Estimating out-of-plane strength of
cracked masonry infills. University of Illinois, Urbana-Campaign, Illinois.
Sigmund, V., Penava, D. [2014] “Influence of openings, with and without confinement, on cyclic response of
infilled RC frames: an experimental study”, Journal of Earthquake Engineering, Vol. 18, No. 1, pp. 113–
146.
Singh, H., Paul, D.K., Sastry, V.V. [1998] “Inelastic dynamic response of reinforced concrete infilled frames”,
Computers and Structures , Vol. 69, pp. 685-693.
Smyrou, E., Blandon, C., Antoniou S., Pinho, R., Crisafulli, F.J. [2011] “Implementation and verification of a
masonry panel model for nonlinear dynamic analysis of infilled RC frames”, Bulletin of Earthquake
Engineering, Vol. 9, No. 5, pp. 1519-1534.
Stafford Smith, B. [1966] “Behaviour of square infilled frames”, Journal of the Structural Division, Vol. 92, No.
1, pp. 381-403.
Stafford Smith, B. [1967] “Methods for predicting the lateral stiffness and strength of multistorey infilled
frames”, Building Science, Vol. 2, pp. 247-257.
Stafford Smith, B., Carter, C. [1969] “A method of analysis for infilled frames”, Proceedings of the Institution
of Civil Engineers”, pp. 31-48.
Stavridis, A., Koutromanos, I., Shing, P.B. [2012] “Shake-table tests of a three-story reinforced concrete
frame with masonry infill walls”, Earthquake Engineering and Structural Dynamics, Vol. 41, No. 6, pp.
1089–1108.
Tomažević, M. [1999] “Earthquake-resistant design of masonry buildings”, Imperial College Press, London,
UK.
Uva, G., Porco, F., Fiore, A. [2012] “Appraisal of masonry infill wall effects in the seismic response of RC
framed buildings: A case study”, Engineering Structures, Vol. 34, pp. 514-526.
Valiasis, T.N., Stylianidis, K.C. [1989] “Masonry infilled RC frames under horizontal loading – experimental
results”, European Earthquake Engineering, Vol. 3, No. 3, pp. 10-20.
Vicente, R., Rodrigues, H., Costa A., Varum, H., Mendes da Silva, J.A.R [2012] “Performance of masonry
enclosure walls: lessons learned from recent earthquakes”, Earthquake Engineering and Engineering
Vibration, Vol. 11, pp. 23-34.
P. Morandi, S. Hak, G. Magenes EUCENTRE 159
Research Report
Žarnić, R., Tomažević, M. [1985] “Study of the behaviour of masonry infilled reinforced concrete frames
subjected to seismic loading”, Proceedings of the 7th International Brick Masonry Conference,
Melbourne, Australia.
P. Morandi, S. Hak, G. Magenes EUCENTRE 161
Research Report
RC column
Masonry infill
II) Single-leaf wider than the frame
Layer of thermal material
(with thermal coating)
Figure A.2.1.Sketch of the transversal section of the frame/infill systems considered in the present guidelines
Figure A.2.3. Examples of units adopted for the “strong” infill typology
P. Morandi, S. Hak, G. Magenes EUCENTRE 165
Research Report
Figure A.2.4 Example of construction details of the contact joints between frame and masonry infills
166 Experimental and numerical seismic performance of strong clay masonry infills
Appendix A - Guideline proposal for seismic design of masonry infills in RC structures
A.3.1 Introduction
The approach proposed for the design of RC frame or dual frame-wall structural systems with
masonry infills is synthetically illustrated in Figure A.3.1. The method has been implemented and
verified on structural configurations regular in plan and in elevation.
The first step consists in the evaluation of the displacement at DLS and at ULS and of the internal
forces at ULS, with the execution of linear elastic analyses (lateral force method or multimodal
analysis with response spectrum) on a bare model of the buildings; the “bare” structural models are
only constituted by RC structural elements (beams, columns, walls, joints), neglecting the stiffness
and the strength of the infills and internal partitions, which are only considered in terms of mass
and vertical load. This way of operating is common in the current design practise and it complies
with the Italian Norms for Constructions (NTC2008 [2008], [2017]) and with EC8 (CEN [2004a]).
Subsequently, once the displacements are evaluated on the bare structure, the proposed criterion
allows obtaining, in a simplified manner, the corresponding displacements on the infilled structure,
as a function of a parameter, Cj, which represents a simplified ratio between the in-plane
stiffness/strength of the infills and the in-plane stiffness/strength of the bare frame. At this point, it
is sufficient to verify that the displacement demands at each storey of the infilled structure at DLS
and at ULS (for example in terms of inter-storey drifts δ) are lower than the deformation limits at
DLS and ULS for the infill typologies included in the building.
In addition to the usual safety checks in terms of resistance at ULS on the structural elements
(beams, columns, walls, joints), the local effects produced by the thrust of the infill on the adjacent
columns also need to be verified as for example reported in EC8. At this regard, a simplified
procedure has been developed by Hak et al. [2013c] and it allows evaluating these effects in terms
of maximum shear acting on the columns, without being over-conservative.
Finally, the out-of-plane infill verification is carried out at ULS comparing the applied pressure with
the strength of the masonry panels. The Italian Norms and EC8 provide indications for the
evaluation of the out-of-plane actions, however, do not indicate any method for the calculation of
the strength. Therefore, a simplified approach for the estimation of the out-of-plane resistance of
the panels through an “arch mechanism” has been proposed; this method also allows to take in
due consideration the reduction of the out-of-plane resistance as a function of the in-plane damage
of the infill.
In comparison with the current European codes, the proposed design approach defines in a more
appropriate manner the displacement safety checks of the masonry panels with the inclusion of the
contribution of the stiffness/strength of the infills, allows a more precise calculation of the local
effects of the infill on the columns, determines a criterion for the evaluation of the out-of-plane
resistance of the panels and introduces a displacement verification also at ULS, in addition to the
one at DLS, which is the sole included in the European Norms.
P. Morandi, S. Hak, G. Magenes EUCENTRE 167
Research Report
Figure A.3.1 Proposed procedure for the seismic design of RC infilled buildings
1
Structural system in which both the vertical and lateral loads are mainly resisted by spatial frames whose
shear resistance at the building base exceeds 65% of the total shear resistance of the whole structural
system.
2 Structural system in which support for the vertical loads is mainly provided by a spatial frame and
resistance to lateral loads is contributed to in part by the frame system and in part by structural walls,
coupled or uncoupled.
168 Experimental and numerical seismic performance of strong clay masonry infills
Appendix A - Guideline proposal for seismic design of masonry infills in RC structures
where α0 is the prevailing aspect ratio of the walls of the structural system. If the aspect ratios
hwi/lWi of all walls i of a structural system do not significantly differ, the prevailing aspect ratio α0 may
be determined assuming the height as the sum of the height of the single walls and, the length, as
the sum of the length of the sections of the walls.
Finally, KR represents a reduction factor depending on the regularity in elevation of the
construction, being equal to 1.0 for regular structures and 0.8 for irregular structures.
From the linear elastic analyses on the structural model, the inter-storey displacements associated
to the design spectrum at DLS, the internal forces and inter-storey displacements associated to the
design spectrum at ULS are obtained.
Table A.3.1. Effective moment of inertia reduced for cracked elements (adapted from Paulay and Priestley
[1992])
Structural elements Range of Ieff/I0
The values of the inter-storey drift demand computed on the infilled structure allow to carry out the
verification in terms of deformation in order to limit the level of in-plane damage on the panels and
to estimate in a more reliable way the local effects produced by the thrust of the infill on the
adjacent columns and the out-of-plane strength of the walls, in comparison with the case where the
displacement demand is evaluated on a bare frame.
where
μd = q if T1 ≥ TC
(A.3.5)
μd = 1+(q – 1)·TC/T1 if T1 < TC
In any case, μd ≤ 5q – 4. In the expression (A.3.5), q is the behaviour factor of the structure, T1 is
the period of the fundamental mode of vibration and TC is the upper limit of the period of the
constant spectral acceleration branch.
The values of the inter-storey displacement of the j-th storey of the structure dr,j, can be defined,
both at DLS and ULS, as the difference of the lateral displacement at the top and bottom of the
considered storey, whereas the inter-storey drift δj of the j-th storey are equal to the ratio between
the inter-storey displacement dr,j and the storey height of the j-th storey hj:
δj = dr,j/hj (A.3.7)
δ m , j’ δ j
, δ j ≤ δ m , j’ + δ C , j C j
δ w, j = δ m , j’ + δ C , j C j (A.3.8)
δ −δ C , δ > δ ’ +δ C
j C, j j j m, j C, j j
170 Experimental and numerical seismic performance of strong clay masonry infills
Appendix A - Guideline proposal for seismic design of masonry infills in RC structures
where δm,j’ represents the equivalent drift capacity taking into consideration the presence of
different typologies of infills at the j-th storey, as reported in paragraph A.3.3.2.2. For values of
Cj=1, the difference between δw,j and δj is equal to δC,j, that can be defined as 2/5·δm,j’. In order to
compute δm,j’ at the j-th storey, it is necessary to know the deformation capacity of each of the infill
typologies in the building as, for example, the one reported in Table A.3.2. If in a building only one
infill typology is present, the deformation capacity δm,j’ at the j-th storey would result equal to δm’.
(a) (b)
Figure A.3.2 (a) Definition of the bi-linear curve for Cj = 1.0; (b) Variation of the bi-linear curve for different
values of Cj
For each seismic resistant direction of a structural configuration (for example, the longitudinal and
the transversal direction), once the distribution of the infills and all the properties that characterize
the infill typologies are known and once the simplified average parameter of the infill KI,j and the
structural stiffness parameter KS,j are evaluated for each storey (j = 1...ns, where ns is the number
of the storeys), it is possible to define the corresponding density-stiffness parameter Cj, which
represents an average ratio between the infill and the structural stiffness at each storey along the
height of the building, as indicated in equation (A.3.9).
KI , j
Cj = (A.3.9)
KS, j
The drifts at the ground storey of the infilled structures are conservatively taken as the ones of the
bare structure, therefore leading to a coefficient Cj equal to zero.
Vj
KS , j = (A.3.10)
dr , j
P. Morandi, S. Hak, G. Magenes EUCENTRE 171
Research Report
ns
V j = ∑ Fi (A.3.11)
i= j
The force Fj acting on storey j can be determined according to equation (A.3.12), where Fb is the
total seismic shear force, si, sj are the displacements of masses i and j in the fundamental mode
shape, while mi, mj are the corresponding storey masses, associated with gravity loads, assumed
in the seismic load combination. The evaluation of the seismic base shear Fb is reported in
equation (A.3.13), where the fundamental period of vibration T1 and the mass M1 participating in
the first mode response can be estimated from a modal analysis and Sd(T1) is the corresponding
spectral ordinate of the design spectrum.
sj ⋅ mj
F= Fb ⋅
j ns
(A.3.12)
∑s
i =1
i
⋅ mi
The horizontal forces Fj may also be taken in a simplified manner with reference to the force
distribution of a static linear analysis, as given in equation (A.3.14), where zi, zj are the heights of
masses mi and mj above the level of seismic force application. The corresponding seismic base
shear force Fb, for each horizontal direction in which the building is analysed, can be calculated
using equation (A.3.15), where Sd (T1) is the ordinate of the design spectrum at period T1, the
fundamental period of vibration of the building for lateral motion in the considered direction and m
is the total mass of the building above the foundation or above the top of a rigid basement. The
correction factor λ may be calculated from equation (A.3.17), in function of the corner period of the
spectrum TC, the period T1 and the number of storeys ns. According to the code, the fundamental
period of vibration T1 can be approximated by the expression given in (A.3.16) in [s], for buildings
with heights up to 40 m. For spatial RC frames, the coefficient Ct may be assumed equal to 0.075,
while H represents the height of the building in [m], from the foundation or from the top of a rigid
basement.
z j ⋅ mj
F= Fb ⋅
j ns
(A.3.14)
∑z
i =1
i
⋅ mi
Fb = S d (T1 ) ⋅ m ⋅ λ (A.3.15)
T1 = Ct ⋅ H 3 / 4 (A.3.16)
in equations (A.3.19a,b), where nb,j is the number of bays in storey j. If for all infills in the same
storey masonry typologies with equal deformation capacities (δm,i,j’ = δm,j’, i = 1…nb,j) are foreseen,
the stiffness KI,j may be expressed by Equation (A.3.19c). Similarly, if in the same storey the use of
infill types with varying properties in terms of strength and deformation capacity is anticipated, the
equivalent drift capacity δm,j’, resulting in the same stiffness KI,j, may be evaluated as given in
(A.3.19b and c).
(a) (b)
Figure A.3.3 (a) Masonry infill secant stiffness; (b) Estimated infill strength reduction due to opening
Fw,i , j Fw,i , j
=K I ,i , j = , K I ,i , j (A.3.18a,b)
d m ,i , j’ δ m ,i , j’h j
1 n Fw,i , jnb , j
1 n
KI , j ∑∑= ∑ Fw,i , j
b,j b,j
= = K I ,i , j , ,
KI , j
KI , j (A.3.19a,b,c)
=i 1 =h j i 1 δ m ,i , j’= h j δ m , j’ i 1
nb , j
∑F w ,i , j
δ m , j’ = i =1
(A.3.20)
nb , j
Fw,i , j
∑δ
i =1 ’
m ,i , j
In order to achieve a suitable approximation, the horizontal force Fw,i,j may be taken equal to the
peak horizontal strength of the masonry infill, calculated according to equation (A.3.21), where fw,i,j
denotes the governing masonry strength, tw,i,j the thickness and Lw,i,j the length for infill i in storey j.
In absence of more specific data, the masonry strength fw,i,j may be estimated in accordance with
Eurocode 8 – Part 1 (CEN [2004a]), on the basis of the corresponding shear strength of bed joints
which may be assumed equal to the initial shear strength under zero compressive stress fv0,i,j.
Since the presence of openings may influence the performance of masonry infills significantly, in
realistic buildings configurations, a possible reduction of the infill strength due to openings is
envisaged. Although different effective criteria have been developed, the peak horizontal
resistance of an infill with opening can be computed in simplified way, for example with the
expression given in equation (A.3.22), derived from a research conducted by Dawe and Seah
[1988], where lp,i,j is the width of the opening for infilled bay i in storey j.
Table A.3.3 Values of drift limits δSLD and δSLV as a function of the different infill typologies
“Strong” infill “Strong” infill “Weak/slender” infill
Drift
without opening with opening
δSLD [%] 0.50 0.35 0.30
δSLV [%] 1.75 1.00 1.00
If not all the safety checks in terms of inter-storey drift can be fulfilled, the building properties
should be modified, enhancing the structural stiffness, with an increment of the size of the
elements, and/or varying the infill characteristics, leading to new values of the density-stiffness
coefficients Cj. At this point, the drift verification of the infilled frame should be repeated, with the
advantage of optimizing the dimensioning of the RC elements, thus exploiting at the best the
strength and stiffness resources of the masonry infills.
ΔVRw
η = 1 + ≤q
∑V (A.3.25)
Ed
The resistance of masonry infills VRw can be computed as reported in paragraph A.3.5.3.4 (VRw =
VSd,lc,1, equation (A.3.29)).
For values of η smaller than 1.1, the modification of action effects may be neglected.
A.3.5.3.2 Local effects due to infills that extend partially over the column height
In the case of infills that do not extend over the entire column height (Figure A.3.5), the European
Norms recommend to verify the RC members following the indications reported hereafter:
a) consider the total length lt of the columns as critical region lcr and consequently, the relevant
confinement and detailing rules have to be satisfied;
3
The European Norms report that, in the critical regions, the following rules shall be satisfied: the corner
bars should be engaged by hoops; the distance between consecutive longitudinal bars engaged by hoops or
cross-ties does not exceed 150 mm for DCH and 200 mm for DCM; the diameter of hoops and cross-ties
should be of at least 6 mm and their spacing should not exceed the following values:
- 1/3 and 1/2 of the minimum dimension of the concrete core of the column, respectively for DCH and DCM;
- 125 mm and 175 mm, respectively for DCH and DCM;
- 6 and 8 times the longitudinal bars that link, respectively for DCH and DCM.
The mechanical volumetric ratio of the confining hoops within the critical region should fulfil the minimum
requirements according to EC8/Italian Norms.
P. Morandi, S. Hak, G. Magenes EUCENTRE 175
Research Report
b) take precautions in relation to the decrease of the effective span ratio of the columns; the
acting shear to be considered for the part of the column free of the infill, is computed using the
following equation:
s
M C,Rd + M C,Rd
i
V Ed = γ Rd ⋅ (A.3.26)
l cl
where γRd is equal to the over-strength coefficient (1.30 for DCH and 1.10 for DCM), M C,Rd
s
i
and M C,Rd are the design values of the resisting moments at the top column section and at the
column section corresponding to top part of the infill, finally lcl is the clear length of the column
(see Figure A.3.6 and Figure A.3.7).
Figure A.3.5 Critical length of columns with partial contacts with infills
Figure A.3.6 Decreased column shear span ratio for clear length lcl
Figure A.3.7 Bending moment shape in the case of column with partial infill contact (left) and bending
moment shape for columns in absence of infills (right)
c) the transverse reinforcement to resist the shear force calculated at the previous point, should
be placed along the length of the column not in contact with the infills and extend along a
length hc (dimension of the column cross-section in the plane of the infill) into the column part
in contact with the infills;
176 Experimental and numerical seismic performance of strong clay masonry infills
Appendix A - Guideline proposal for seismic design of masonry infills in RC structures
Figure A.3.8 Extension of the transversal reinforcement into the column part in contact with the infills
d) if the length of the column not in contact with the infills is less than 1,5hc, the shear force
should be resisted by diagonal reinforcement.
Figure A.3.9. Diagonal reinforcement in the case of very small part non in contact with the columns
A.3.5.3.3 Local effects due to infills located only on one side of the column
As reported in EC8, where the infills extend to the entire clear length of the adjacent columns, and
there are masonry walls only on one side of the column (e.g. corner columns), the entire length of
the column should be considered as a critical region and be reinforced with the amount and pattern
of stirrups required for critical regions.
Figure A. 3.10. Critical length for columns adjacent to infills only in one side
A.3.5.3.4 Local effects: evaluation of the increased shear acting on the columns due to infills
In the case of infills adherent to the frame, a safety check regarding the increased shear demand
on RC columns due to the horizontal component of the infill thrust is needed. In Figure A.3.11, the
distribution of the internal forces on a column due to the presence of an infill is reported. This
P. Morandi, S. Hak, G. Magenes EUCENTRE 177
Research Report
thrust, resulting as larger as the infill stiffness/strength is larger, can produce brittle failures at the
ends of non-sufficiently resistant columns.
In particular, the contact length lc of the column, over which the diagonal strut force of the infill acts
on the column causing a local pressure fs and introducing a concentration of forces, should be
verified in shear; therefore, it is necessary to check that:
where VC,Rd,lc is the shear strength along the length lc at the ends of the columns and VC,Ed,lc is the
increased shear demand due to the horizontal strut force acting at the ends of the column.
Figure A.3.11. Acting shear on the column due to the horizontal component of the infill thrust
The Italian Norms do not currently contain any information regarding the evaluation of the acting
forces on the columns.
On the other side, EC8 reports generic indications on the evaluation of the contact length and of
the increased shear on the columns. In particular, EC8 recommends that the contact length, lc, of
columns over which the diagonal strut force of the infill is applied, should be verified in shear for
the smaller (VC,Ed,lc) of the following two shear forces VC,Ed,w and VC,Ed,M:
The shear force VC,Ed,w is the horizontal component of the infill strut force, assumed to be equal,
according to EC8, to the horizontal shear strength of the panel, as estimated on the basis of the
shear strength of bed joints:
VC , Ed =
,w F=
w Aw ⋅ f v (A.3.29)
where Aw is the sectional area of the infill (tw·Lw) and fv is the shear strength of the masonry; this
strength can be assumed to be equal to fv0, in relation with the shear strength under zero
compressive stress criterion, as provided by the Italian Norms and by Eurocode 6 (CEN [2004b]).
The shear force VC,Ed,M is calculated in agreement with the following expression:
2 M C , Rd
VC , Ed , M = γ Rd (A.3.30)
lc
178 Experimental and numerical seismic performance of strong clay masonry infills
Appendix A - Guideline proposal for seismic design of masonry infills in RC structures
where γRd is the overstrength coefficient (1.30 for DCH and 1.10 for DCM), assuming that the
flexural bending resistance of the column, MC,Rd, can develop at the two ends of the contact length,
lc .
bw 0.25d w 0.125d w 2
=lc = = (A.3.31)
2cos θ 2cos θ Lw
Figure A.3.13. Definition of the parameters for the evaluation of the contact length lc
the two following figures can be also expressed using the equations (A.3.32), where δw,j is the drift
of the infilled structure at storey j, whereas δSLD is the drift limit at DLS, equal to 0.50% for the
strong infills without openings and equal to 0.30% in the case of slender/weak infills (see Table
A.3.3).
Figure A.3.14. Strut force activation in function of average inter-storey drift aw,j : strong infills
Figure A.3.15. Strut force activation in function of average inter-storey drift aw,j : slender/weak infills
180 Experimental and numerical seismic performance of strong clay masonry infills
Appendix A - Guideline proposal for seismic design of masonry infills in RC structures
S a ⋅ Wa
Fa = (A.3.34)
qa
where Sa is the seismic coefficient applicable to the infill, Wa is the corresponding weight of the
element (pg·tw·Lw·hw, being pg the unit volume weight of masonry, tw, Lw and hw, respectively, the
thickness, the length and the height of the infill) and qa is the behaviour factor of the element
(usually, qa=2 for external infills, internal partitions and façades, as reported in Table A.3.4); in EC8
an important factor, γa, is also introduced and allows to increase the force in case of strategic non-
structural elements (considered always equal to 1 in NTC 2008).
The seismic coefficient Sa may be calculated according to the expression given in (A.3.35), as
reported in EC8 and in NTC2008, where z represents the distance of the centre of mass of the
non-structural element from the level of application of the seismic action (i.e., the top of the
foundation or a rigid basement) and H is the corresponding building height, as shown in Figure
A.3.16.
3(1 + z H )
Sa = α ⋅ S ⋅ − 0.5 (A.3.35)
1 + (1 − Ta T1 )
2
The ratio of the design ground acceleration ag on type A ground, to the acceleration of gravity g is
denoted by α, S is the soil factor, Ta the fundamental vibration period of the non-structural element
and T1 the fundamental vibration period of the building in the relevant direction. The seismic
coefficient Sa should not be taken less than αS.
The fundamental elastic period of the structure T1 can be calculated as T1=Ct·H3/4, where Ct can be
taken conservatively equal to 0.050, in order to consider in a simplified manner the stiffening
effects of the infills, as proposed by Pinho and Crowley [2009]. The fundamental period Ta of the
masonry infill in the out-of-plane direction can be calculated according to the expression for the
case of single vertical bending response with hinged ends given in equation (A.3.36), where hw is
the height of the panel, m is the mass of the infill per unit height, E is the modulus of elasticity of
masonry and I is the out-of-plane moment of inertia of the horizontal cross section of the panel.
2hw2 m (A.3.36)
Ta =
π EI
In any case, for the evaluation of the out-of-plane acceleration, other expressions can also be
used, for example referring to past studies or other seismic codes.
P. Morandi, S. Hak, G. Magenes EUCENTRE 181
Research Report
Table A.3.4 Values of qa for non-structural elements
Dividing the seismic force Fa for the lateral surface of the panel, an out-of-plane pressure on the
infill, wa, is obtained from the following expression:
Fa
wa = (A.3.37)
Lw ⋅ hw
where Lw and hw are, respectively, the length and the height of the infill.
tw
2
wa ≤ wR ,β = βa , j ⋅ 0.72 ⋅ f d ⋅
βa , j ⋅ wR = (A.3.38)
hw
The out-of-plane resistance, wR, has been estimated with the vertical “arching mechanism” for infill
in complete contact with the beam, under the hypothesis of formation of horizontal cracking at mid-
height of the panel, as reported in Figure A.3.18, neglecting second-order effects (acceptable for
hw/tw < 25, being hw and tw respectively the height and the thickness of the infill) and where fd is the
design compression strength of the masonry.
( ra ,1 − 1) δ w, j / δ SLD + 1, δ w, j ≤ δ SLD
=βa, j ra ,1 , δ SLD < δ w, j ≤ δ SLV (A.3.39)
ra ,2 , δ SLV < δ w, j
The values of ra,1 and ra,2, assumed to be corresponding to the attainment of the performance
levels at DLS and ULS and therefore to the related drift values δSLD an δSLV, have been estimated
according to experimental evidence, as resumed in Table A.3.5, in function of the different infill
typology.
P. Morandi, S. Hak, G. Magenes EUCENTRE 183
Research Report
Table A.3.5 Values of ra,1 and ra,2
Strong infills Slender/weak infills
Limit states
Without openings With opening
δSLD [%] Damage Limt State 0.50 0.35 0.30
δSLV [%] Ultimate Limit State 1.75 1.00 1.00
ra,1 Damage Limt State 0.40 0.15 0.20
ra,2 Ultimate Limit State 0.25 0.00 0.00
Similarly, the equations reported above can be represented though the graphs of Figure A.3.19,
Figure A.3.20 and Figure A.3.21 related respectively to strong infills without and with openings, and
to slender/weak infills.
Figure A.3.19 Simplified out-of-plane resistance reduction coefficient βa,j for strong infills without openings
Figure A.3.20 Simplified out-of-plane resistance reduction coefficient βa,j for strong infills with openings
184 Experimental and numerical seismic performance of strong clay masonry infills
Appendix A - Guideline proposal for seismic design of masonry infills in RC structures
Figure A.3.21 Simplified out-of-plane resistance reduction coefficient βa,j for slender/weak infills
A.3.6.2 CASE B1): double-leaf not connected infills with both leaves enchased in the frame
S a ,1 ⋅ Wa ,1
Fa ,1 = (A.3.40)
qa
Sa , 2 ⋅ Wa , 2
Fa , 2 = (A.3.41)
qa
where:
3(1 + z H )
S a ,1 = α ⋅ S ⋅ − 0 .5
1 + (1 − Ta ,1 T1 )
2 (A.3.44)
3(1 + z H )
Sa , 2 = α ⋅ S ⋅ − 0.5
1 + (1 − Ta , 2 T1 )
2 (A.3.45)
Fa ,1
wa ,1 = (A.3.46)
Lw,1 ⋅ hw,1
P. Morandi, S. Hak, G. Magenes EUCENTRE 185
Research Report
Fa ,2
wa ,2 = (A.3.47)
Lw,2 ⋅ hw,2
where the parameters with subscript “1” refer to the external leaf, whereas those with subscript “2”
refer to the internal leaf and have already been defined in the previous paragraph.
Figure A.3.22 Double-leaf infill with both the leaves enchased in the frame and not connected
tw,2
2
wa ,2 ≤ wR ,2,β =
βa ,2, j ⋅ wR ,2 =
βa ,2, j ⋅ 0.72 ⋅ f d ,2 ⋅
h (A.3.49)
w,2
where fd,1 and fd,2 are the design compression strength of the internal and external masonry leaves
and βa,1,j, βa,2,j are the reduction coefficients that take into account the decrease of the out-of-plane
resistance due to possible previous in-plane damage as a function of the effective inter-storey drift
values δj,w at ULS. The parameters βa,1,j and βa,2,j can be obtained as proposed for the case of
single leaf infills, considering implicitly that the 2 leaves are subjected to the same in-plane
deformation.
A.3.6.3 CASE B2): double-leaf connected infills with both leaves enchased in the frame
Sa ⋅ Wa
Fa = (A.3.50)
qa
where:
3 (1 + z H )
Sa = α ⋅ S ⋅ − 0.5 (A.3.52)
1 + (1 − Ta T1 )2
Fa
wa = (A.3.53)
Lw ⋅ hw
where the parameters with subscript “1” refer to the external leaf, whereas those with subscript “2”
refer to the internal leaf; the period Ta can be evaluated computing the out-of-plane stiffness as in
the case of single-leaf wall, having equivalent thickness tw,eq and mass equal to the sum of the
masses of the two leaves.
wa ≤ wR ,β = βa , j ⋅ 0.72 ⋅ f d ⋅
βa , j ⋅ wR = (A.3.54)
hw
where fd is the design compression strength of the masonry leaves and βa,j is the reduction
coefficient that takes into account the decrease of the out-of-plane resistance due to possible
previous in-plane damage, as a function of the effective inter-storey drift values δj,w at ULS. If the
compression strength of the two leaves is different, the use of the lower strength is advised while
the parameter βa,j can be obtained as proposed for the case of single leaf infills, considering the
more conservative value in the case of two different typologies of masonry infills (i.e., strong and
weak infill).
P. Morandi, S. Hak, G. Magenes EUCENTRE 187
Research Report
Figure A.3.23 Double-leaf infill with both the leaves enchased in the frame and connected
Table A.3.6 Admissible values of the maximum slenderness λmax,a for single-leaf clay masonry strong infills
zw/H 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
agS
0.050 27.8 26.4 25.2 24.2 23.2 22.4 21.7 21.0 20.4 19.8
0.075 22.7 21.6 20.6 19.7 19.0 18.3 17.7 17.1 16.6 16.2
0.100 19.6 18.7 17.8 17.1 16.4 15.8 15.3 14.8 14.4 14.0
0.125 17.6 16.7 15.9 15.3 14.7 14.2 13.7 13.3 12.9 12.5
0.150 16.0 15.2 14.6 13.9 13.4 12.9 12.5 12.1 11.8 11.4
0.175 14.8 14.1 13.5 12.9 12.4 12.0 11.6 11.2 10.9 10.6
0.200 13.9 13.2 12.6 12.1 11.6 11.2 10.8 10.5 10.2 9.9
0.225 13.1 12.4 11.9 11.4 11.0 10.6 10.2 9.9 9.6 9.3
0.250 12.4 11.8 11.3 10.8 10.4 10.0 9.7 9.4 9.1 8.9
0.275 11.8 11.3 10.7 10.3 9.9 9.6 9.2 9.0 8.7 8.5
0.300 11.3 10.8 10.3 9.9 9.5 9.1 8.8 8.6 8.3 8.1
0.325 10.9 10.4 9.9 9.5 9.1 8.8 8.5 8.2 8.0 7.8
0.350 10.5 10.0 9.5 9.1 8.8 8.5 8.2 7.9 7.7 7.5
NOTA: for values not provided in the table, linear interpolation is admitted (no extrapolations)
For example, in the case of a RC building with maximum infill height hw = 2.9 m, located in a
seismic region with a peak ground acceleration of ag·S at ULS of 0.20 g, single-leaf infills having
minimum thickness of tw = hw/λmax,a = 2.90/9.9 = 0.29 m can be used in all the storeys of the
building without any other additional calculations. For instance, a 30 cm thick clay unit of group 2
according to EC6 can be selected; this block can have a percentage of void of 50-55%, unit vertical
compressive strength fb ≥ 8 MPa and can be laid with M5 mortar, providing a characteristic
compression strength of the masonry fk equal to 3.13 MPa.
P. Morandi, S. Hak, G. Magenes EUCENTRE 189
Research Report
tw Em fb fm fk fvk0 % of
Infill typology holes
[m] [MPa] [MPa] [MPa] [MPa] [MPa]
“Strong”
0.30 3660 10.0 5.0 3.66 0.20 55
Unit group (EC6): 2
“Slender/weak” 0.08 920 0.92
2.0 5.0 0.20 60
Unit group (EC6): 4 0.10* 1500* 1.50*
* with plaster
0.40 ULS :
0.30 ag = 0.209g, agS = 0.250g
0.20
0.10 DLS:
0.00 ag = 0.070g, agS = 0.084g
0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00
T [s]
Figure A.4.5 Elastic and design spectra (Ruvo del Monte (PZ), Basilicata, Italy)
P. Morandi, S. Hak, G. Magenes EUCENTRE 193
Research Report
Table A.4.5 Periods and percentage of mass participation for the first 7 modes
Mode Period [s] Translational mass x [%] Translational mass y [%]
1 1.27 0.0 71.0
3 0.71 68.0 0.0
4 0.35 0.0 16.0
8 0.18 19.0 0.0
9 0.16 0.0 6.0
12 0.13 2.0 0.0
25 0.12 1.0 0.0
194 Experimental and numerical seismic performance of strong clay masonry infills
Appendix A - Guideline proposal for seismic design of masonry infills in RC structures
Mode 1 Mode 3
A.4.5 Evaluation of the displacement demand of the infilled structure and displacement
checks
As first step, the simplified stiffness parameters of the structural elements KS,j and of the infill KI,j
need to be defined in agreement with the approach discussed in sections A.3.3.2.1 and A.3.3.2.2.
The parameters are reported in Table A.4.6 (KS,j) and in Table A.4.7 (KI,j and density-stiffness
coefficient Cj = KI,j/ KS,j).
For the evaluation of KS,j, the forces Fj have been computed separately for the two directions
(longitudinal and transversal) according to equations (A.3.12) and (A.3.13), and the displacements
dr,j have been calculated at the centres of mass of each storey. For the calculation of KI,j, the
contribution within the structural alignments of the infills highlighted in grey in Figure A.4.1 has
been considered.
Table A.4.6 Evaluation of the simplified parameter for the structural stiffness KS,j
Fj Vj dr,j KS,j
[KN] [KN] [m] [KN/m]
Longitudinal
1 20 914 0.0008 1142655
2 63 894 0.0016 558679
Storey
Longitudinal
1 3.10 0.33 1448.8 140670 0.00
2 3.10 0.33 1448.8 140670 0.25
Storey
From the inter-storey drift demand obtained for the bare structure (δ,j=dr,SLD/hj), the corresponding
drifts of the infilled configuration (δw,j) have been computed for each storey j of the building at the
DLS and ULS, as reported in Table A.4.8. The values have been compared with the limit values of
Table A.4.3, according to the procedure described in sections A.3.3 e A.3.4. At the ground storey,
the coefficient Cj has been taken conservatively equal to 0, in order to obtain, conservatively, an
effective displacement equal to the one of the bare structure. The bilinear curves, that compare the
drift of the bare configuration with the infilled one for different values of Cj, are reported in Figure
A.4.8a and in Figure A.4.8b, for the longitudinal and transversal direction, respectively. The drift
checks for two the directions are also reported in Figure A.4.9a and Figure A.4.9b; in this case, the
deformation limits taken into account are the ones related to "slender/weak" infills.
Table A.4.8. Elastic drift of the bare structure and evaluation of the drift of the infilled structure
(a) (b)
1.00 1.00
Drift δw,j [%]
Figure A.4.8. Bilinear curves for different values of Cj: (a) longitudinal direction; (b) transversal direction
(a)
SLD SLV
(Longitudinale) (Longitudinal)
(Longitudinale)
6 6
5 5
4 4
Storey
Storey
3 3
2 2
1 1
Bare Bare
Infilled Infilled
0 0
0.00 0.10 0.20 0.30 0.40 0.50 0.00 0.20 0.40 0.60 0.80 1.00 1.20
Drift [%] Drift [%]
(b)
SLD SLV
(Trasversale)
(Transversal) (Transversal)
(Trasversale)
6 6
5 5
4 4
Storey
Storey
3 3
2 2
1 1
Bare
Bare
Infilled
0 Infilled
0
0.00 0.10 0.20 0.30 0.40 0.50 0.00 0.20 0.40 0.60 0.80 1.00 1.20
Drift [%] Drift [%]
Figure A.4.9. Drift check: (a) longitudinal direction; (b) transversal direction
A.4.6 Local effects on the RC columns due to the presence of the infills
The checks of the local effects have been carried out on a column located at the corner of the
ground floor in the transversal direction. According to EC8, the critical region within the column is
extended along the whole height of the structural element, and the detailing should be designed
consequently. Figure A.4.10a reports the section of the column. The two beams connected to the
corner column present the reinforcement as shown in Figure A.4.10.b.
P. Morandi, S. Hak, G. Magenes EUCENTRE 197
Research Report
Figure A.4.10. (a) Corner column at the ground floor; (b) beam connected to the corner column
- Alternative shear force VC,Ed,w (horizontal component of the activated strut force):
VC,Ed,w = Fw = aw,j·Aw·fv = 0.845·0.30·4.6·0.2·103 = 233.2 kN
- Activation of the maximum force of the compressive strut aw,j (δw,j = 0.36 %):
=lc =
(
0.125d w 2 0.125 ⋅ 2.825 + 4.6
2 2
= 0.792 m
)
Lw 4.6
- Safety check:
VC,Rd,lc = 473 KN > VC,Ed = 233.2 KN Ok.
T1 =
Ct H 3 / 4 =
0.05 ⋅ 18.63 / 4 =
0.45 s
- Coefficient for the reduction of the out-of-plane strength βa,j (δw,j = 0.35 %):
tw
2
0.1 3
2
wR ,β =
βa , j ⋅ wR =
βa , j ⋅ 0.72 ⋅ f d ⋅ = 0.20 ⋅ 0.72 ⋅ 1.5 ⋅ ⋅ 10 =
0.30 kN / m 2
hw 2.7
- Safety check :
=wa 0.28 kN / m 2 ≤
= wR ,β 0.30 kN / m 2
P. Morandi, S. Hak, G. Magenes EUCENTRE 199
Research Report
- Safety coefficient:
wR ,β / wa = 1.07
T1 =
Ct H 3 / 4 =
0.05 ⋅ 18.63 / 4 =
0.45 s
- Coefficient for the reduction of the out-of-plane strength βa,j (δw,j = 0.35 %):
t
2
0.3 3
2
wR ,β = βa , j ⋅ 0.72 ⋅ f d
βa , j ⋅ wR = ⋅ w =0.40 ⋅ 0.72 ⋅ 3.66 ⋅ ⋅ 10 =
13.01 kN / m 2
hw 2.7
- Safety check:
=wa 0.86 kN / m 2 ≤
= wR ,β 13.01 kN / m 2
- Safety coefficient:
wR ,β / wa = 15.11
200 Experimental and numerical seismic performance of strong clay masonry infills
Appendix A - Guideline proposal for seismic design of masonry infills in RC structures
Figura B.2.1 Vista in sezione trasversale dei sistemi tamponamento/telaio oggetto delle presenti linee guida
Figura B.2.2 Esempi di blocchi per murature non strutturali in soluzione “debole/leggera”
B.3.1 Introduzione
L’approccio proposto per la progettazione di sistemi strutturali in c.a. a telaio o misti telaio–parete
con tamponamenti in laterizio è illustrato sinteticamente in Figura B.3.1. Il metodo è stato
implementato e verificato su configurazioni strutturali regolari in pianta ed in elevazione.
Il primo passo consiste nella valutazione degli spostamenti allo SLD ed allo SLV e delle
sollecitazioni allo SLV con l’esecuzione di un’analisi elastica e lineare (statica equivalente o
dinamica multimodale con spettro di risposta) su un modello strutturale “nudo” dell’edificio,
costituito cioè solo dagli elementi strutturali in c.a. (travi, pilastri, nodi, pareti) e trascurando la
rigidezza e la resistenza dei tamponamenti e delle tramezzature che vengono considerati solo per
il loro contributo in peso e massa; questo modo di operare è comune all’attuale pratica progettuale,
oltre ad essere in conformità alle NTC (NTC08 [2008], [2017]) ed all’EC8 (CEN [2004a]).
Successivamente, noti gli spostamenti sulla struttura “nuda”, il criterio proposto consente di
ricavare, in maniera semplificata, i corrispondenti spostamenti sulla struttura tamponata in funzione
di un parametro, Cj, che corrisponde al rapporto tra le rigidezze/resistenze nel piano dei
tamponamenti rispetto a quelle del telaio nudo. A questo punto è sufficiente verificare che le
richieste di spostamento di tutti i piani della struttura tamponata allo SLD ed allo SLV (per es. in
termini di drift interpiano δ) siano inferiori ai limiti deformativi allo SLD ed allo SLV delle tipologie di
tamponamento presenti nell’edificio.
Oltre alle consuete verifiche in termini di resistenza allo SLV sugli elementi strutturali in c.a. (travi,
pilastri, pareti, nodi), è necessario anche verificare gli effetti locali prodotti dalla spinta della
tamponatura sui pilastri adiacenti ad essa, per esempio come indicato nell’EC8. A tal fine è stata
sviluppata una procedura semplificata (Hak et al. [2013c]) che consente di valutare in modo
efficace tali effetti in termini di massimo taglio agente sulle colonne, senza penalizzare inutilmente
le verifiche.
Infine, per quanto riguarda la verifica fuori piano delle tamponature, essa viene svolta allo SLV
confrontando la pressione applicata con la resistenza dei pannelli murari. Le NTC2008 e l’EC8
forniscono informazioni per valutare le azioni fuori piano, tuttavia non indicano come valutare la
resistenza del pannello. E’ stato dunque proposto un criterio semplificato che consente di stimare
la resistenza di un pannello murario alle azioni fuori piano attraverso un meccanismo ad arco
verticale ed un coefficiente riduttivo della resistenza in funzione del danneggiamento nel piano
della tamponatura.
Rispetto al quadro normativo attuale (NTC e EC8), l’approccio progettuale proposto definisce in
maniera più appropriata le verifiche in spostamento nel piano dei pannelli murari includendo il
contributo di rigidezza/resistenza delle tamponature, consente un calcolo più preciso degli effetti
locali delle tamponature sulle colonne, determina un criterio per la valutazione della resistenza
fuori piano dei pannelli murari ed introduce una verifica in spostamento anche allo SLV in aggiunta
a quella prevista nelle NTC ed EC8 per il solo SLD.
P. Morandi, S. Hak, G. Magenes EUCENTRE 207
Research Report
Figura B.3.1 Procedura di calcolo proposta per la progettazione sismica di edifici in c.a. tamponati
1
Sistemi strutturali nei quali la resistenza alle azioni sia verticali che orizzontali è affidata principalmente a
telai spaziali aventi resistenza al taglio alla base ≥ 65% della resistenza a taglio totale.
2
Sistemi strutturali nei quali la resistenza alle azioni verticali è affidata prevalentemente ai telai, la resistenza
alle azioni orizzontali è affidata in parte ai telai ed in parte alle pareti, singole o accoppiate; se più del 50 %
dell’azione orizzontale è assorbita dai telai si parla di strutture miste equivalenti a telai, altrimenti si parla di
strutture miste equivalenti a pareti.
208 Experimental and numerical seismic performance of strong clay masonry infills
Appendix B – Proposta di linee guida per la progettazione sismica di tamponamenti in muratura in
strutture in c.a.
dove α0 è il valore assunto in prevalenza dal rapporto tra altezze e larghezze delle pareti; nel caso
in cui gli α0 delle pareti non differiscano significativamente tra di loro, il valore di α0 per l’insieme
delle pareti può essere calcolato assumendo come altezza la somma delle altezze delle singole
pareti e come larghezza la somma delle larghezze.
Infine, il coefficiente KR rappresenta un fattore riduttivo che dipende dalle caratteristiche di
regolarità in altezza della costruzione, con valore pari ad 1.0 per costruzioni regolari in altezza e
0.8 per costruzioni non regolari in altezza.
Dalle analisi elastiche condotte sul modello strutturale si ricavano dunque gli spostamenti
interpiano associati ad uno spettro di risposta allo SLD e le sollecitazioni e gli spostamenti
associati ad uno spettro di progetto relativo allo SLV.
dove
μd = q se T1 ≥ TC
(B.3.5)
μd = 1+(q – 1) · TC/T1 se T1 < TC
δj = dr,j/hj (B.3.7)
δ m , j’ δ j
, δ j ≤ δ m , j’ + δ C , j C j
δ w, j = δ m , j’ + δ C , j C j (B.3.8)
δ −δ C , δ > δ ’ +δ C
j C, j j j m, j C, j j
in cui δm,j’ rappresenta la capacità in termini di drift equivalente che considera la presenza di
tipologie differenti di tamponature al piano j, come riportato nel paragrafo B.3.3.2.2. Per valori di
Cj=1, la differenza tra δw,j e δj è pari a δC,j, che può essere definito pari a 2/5·δm,j’. Per calcolare δm,j’
al piano j, è necessario conoscere la capacità deformativa di ognuna delle tipologie di
tamponamento dell’edificio, per esempio quelle riportate in Tabella B.3.2. Se in un edificio fosse
presente una sola tipologia di tamponamento, la capacità deformativa δm,j’ relativa al piano j
risulterebbe pari al corrispondente valore di δm’.
(a) (b)
Figura B.3.2 (a) Definizione della curva bilineare per Cj = 1.0; (b) Variazione della curva bilineare per diversi
valori di Cj
Per ciascuna direzione sismo-resistente di una configurazione strutturale (per es., la direzione
longitudinale e trasversale), note la distribuzione dei tamponamenti e tutte le proprietà richieste per
caratterizzare le tipologie di tamponamento adottate, noto il parametro semplificato medio della
tamponatura KI,j ed il parametro di rigidezza strutturale KS,j per ciascun piano (j = 1...ns, dove ns è il
numero di piani), è possibile definire il relativo parametro densità-rigidezza Cj, che rappresenta un
P. Morandi, S. Hak, G. Magenes EUCENTRE 211
Research Report
rapporto medio tra la rigidezza del tamponamento e quella strutturale per ciascun piano lungo
l’altezza dell’edificio, come indicato nell’Equazione (B.3.9).
KI , j
Cj = (B.3.9)
KS, j
I drift al piano terra di strutture tamponate sono conservativamente assunti pari a quelli della
corrispondente configurazione “nuda”, dando luogo dunque ad un coefficiente Cj pari a zero.
Vj
KS , j = (B.3.10)
dr , j
ns
V j = ∑ Fi (B.3.11)
i= j
La valutazione della forza Fj che agisce al piano j-esimo può essere determinata in base
all’equazione (B.3.12), in cui Fb è il taglio totale alla base, si, sj sono gli spostamenti delle masse i e
j nella forma modale fondamentale, infine mi ed mj sono le corrispondenti masse di piano nella
combinazione sismica. La valutazione di Fb è riportata nell’equazione (B.3.13), in cui il periodo del
modo di vibrare principale nella direzione in esame T1 e la massa partecipante al primo modo M1
possono essere stimati da un’analisi modale, dove Sd(T1) è la corrispondente ordinata spettrale
dello spettro di progetto.
sj ⋅ mj
F= Fb ⋅
j ns
(B.3.12)
∑s
i =1
i
⋅ mi
La forza Fj può anche essere valutata, in modo semplificato, con riferimento alla distribuzione di
forze di un’analisi lineare statica attraverso l’espressione (B.3.14), in cui gli spostamenti nella
prima forma modale si, sj vengono sostituiti dalle quote rispetto al piano di fondazione zi e zj delle
masse i e j, T1 può essere stimato dalla formula (B.3.16) nel caso di edifici che non superino i 40 m
di altezza H (in m) e la cui massa sia approssimativamente distribuita lungo l’altezza, m è la massa
totale dell’edificio e λ è pari all’espressione (B.3.17) in cui TC è il periodo corrispondente all’inizio
del tratto a velocità costante dello spettro.
z j ⋅ mj
F= Fb ⋅
j ns
(B.3.14)
∑z
i =1
i
⋅ mi
Fb = S d (T1 ) ⋅ m ⋅ λ (B.3.15)
T1 = Ct ⋅ H 3 / 4 (B.3.16)
212 Experimental and numerical seismic performance of strong clay masonry infills
Appendix B – Proposta di linee guida per la progettazione sismica di tamponamenti in muratura in
strutture in c.a.
(a) (b)
Figura B.3.3 (a) Rigidezza secante del tamponamento in muratura; (b) Riduzione di resistenza prevista del
tamponamento dovuta all’apertura
Fw,i , j Fw,i , j
=K I ,i , j = , K I ,i , j (B.3.18a,b)
d m ,i , j’ δ m ,i , j’h j
1 n Fw,i , jnb , j
1 n
∑= ∑= ∑ Fw,i , j
b,j b,j
=KI , j K I ,i , j ,
, KI , j
KI , j (B.3.19a,b,c)
=i 1 =h j i 1 δ m ,i , j’= h j δ m , j’ i 1
nb , j
∑F w ,i , j
δ m , j’ = i =1
(B.3.20)
nb , j
Fw,i , j
∑δ
i =1 ’
m ,i , j
P. Morandi, S. Hak, G. Magenes EUCENTRE 213
Research Report
Per raggiungere un’adeguata approssimazione, la forza orizzontale Fw,i,j può essere posta pari alla
resistenza di picco orizzontale del tamponamento, calcolata secondo l’Equazione (B.3.21), dove
fw,i,j indica la resistenza a taglio della muratura, tw,i,j lo spessore e Lw,i,j la lunghezza per il
tamponamento i nel piano j. In assenza di dati più specifici, la resistenza della muratura fw,i,j può
essere stimata secondo l’Eurocodice 8 – Parte 1 (CEN [2004a]), sulla base della corrispondente
resistenza a taglio dei letti di malta che può essere assunta pari alla resistenza iniziale a taglio in
assenza di sforzo di compressione fv0,i,j.
Per quanto riguarda le aperture nei tamponamenti, poiché esse influenzano significativamente il
comportamento dei pannelli murari, nel caso di configurazioni edilizie realistiche, si deve
considerare la possibile riduzione di resistenza per i tamponamenti dovuta alla presenza delle
aperture stesse. Nell’ambito di questo studio, la resistenza di picco orizzontale di un
tamponamento con apertura può essere calcolata secondo l’espressione data nell’Equazione
(B.3.22), derivata da uno studio condotto da Dawe e Seah [1988], in cui lp,i,j è la lunghezza
dell’apertura per il tamponamento i nel piano j.
Tabella B.3.3 Valori limite di drift δSLD e δSLV in funzione delle diverse tipologie di tamponamento
Tamp. “robusto” Tamp. “robusto” Tamp. “leggero”
Drift
senza aperture con aperture
δSLD [%] 0.50 0.35 0.30
δSLV [%] 1.75 1.00 1.00
dove ΔVRw è la riduzione totale della resistenza delle pareti di muratura nel piano considerato,
paragonata al piano con più tamponamenti al di sopra di esso, ΣVEd è la somma delle azioni
sismiche di taglio agenti su tutte le membrature sismiche verticali primarie del piano considerato.
La resistenza delle pareti in muratura VRw può essere stimata come esposto al paragrafo B.3.5.3.4
(VRw = VSd,lc,1, espressione (B.3.29)). Se l’espressione (B.3.25) porta ad un coefficiente di
amplificazione η minore di 1.1, non è necessario modificare gli effetti delle azioni.
B.3.5.3 Verifiche degli effetti locali sugli elementi strutturali in c.a. dovuti ai tamponamenti
Nel caso di tamponature aderenti al telaio, è necessario prestare particolare attenzione agli effetti
sfavorevoli indotti dai tamponamenti sui pilastri in c.a., in particolare nei casi di murature
rigide/resistenti nel proprio piano. Nel seguito, si riportano una serie di indicazioni per la
valutazione degli effetti locali dovuti alla presenza dei tamponamenti sulla struttura in c.a., alcune
delle quali sono tratte dall’EC8 e dalle NTC 2008.
P. Morandi, S. Hak, G. Magenes EUCENTRE 215
Research Report
B.3.5.3.2 Effetti locali dovuti a tamponamenti con altezza minore dell’altezza delle colonne
adiacenti
Nel caso in cui l'altezza dei tamponamenti sia minore dell'altezza libera delle colonne adiacenti, il
punto 5.9(2) dell’EC8 ed i punti 7.4.4.2.1 – 7.4.6.2.2 delle NTC 2008 raccomandano di operare nel
modo seguente:
a) considerare l'intera altezza delle colonne come zona critica e armarla con il quantitativo e la
disposizione di staffe previsti per le zone critiche 3;
3 Le NTC riportano che, nelle zone critiche, devono essere rispettate le condizioni seguenti: le barre disposte
sugli angoli della sezione devono essere contenute dalle staffe; almeno una barra ogni due, di quelle
disposte sui lati, deve essere trattenuta da staffe interne o da legature; le barre non fissate devono trovarsi a
meno di 15 cm e 20 cm da una barra fissata, rispettivamente per CD”A” e CD”B”. Il diametro delle staffe di
contenimento e legature deve essere non inferiore a 6 mm ed il loro passo deve essere non superiore alla
più piccola delle quantità seguenti:
- 1/3 e 1/2 del lato minore della sezione trasversale, rispettivamente per CD”A” e CD”B”;
- 125 mm e 175 mm, rispettivamente per CD”A” e CD”B”;
- 6 e 8 volte il diametro delle barre longitudinali che collegano, rispettivamente per CD”A” e CD”B”.
Si devono disporre staffe in un quantitativo minimo non inferiore a:
f cd ⋅ bst
0.08 f per CD" B"
Ast yd
=
s 0.12 f cd ⋅ bst per CD" A"
f yd
in cui Ast è l’area complessiva dei bracci delle staffe, bst è la distanza tra i bracci più esterni delle staffe, s è il
passo delle staffe, fcd è la resistenza di calcolo a compressione del calcestruzzo e fyd è la resistenza di
calcolo dell’acciaio.
216 Experimental and numerical seismic performance of strong clay masonry infills
Appendix B – Proposta di linee guida per la progettazione sismica di tamponamenti in muratura in
strutture in c.a.
Figura B.3.5 Lunghezza critica per le colonne adiacenti a tamponamenti con altezza minore
b) cautelarsi nei confronti delle conseguenze legate alla diminuzione del rapporto di taglio
effettivo di tali colonne. Le sollecitazioni di taglio da considerare per la parte di pilastro priva di
tamponamento sono calcolate utilizzando la relazione:
s
M C,Rd + M C,Rd
i
VEd = γ Rd ⋅ (B.3.26)
l cl
dove γRd è pari al coefficiente di sovraresistenza (1.30 per CD”A” e 1.10 per CD”B”), M C,Rd e
s
i
M C,Rd sono rispettivamente i valori di progetto dei momenti resistenti della sezione di sommità
del pilastro e della sezione del pilastro in corrispondenza della sommità del muro di
tamponamento, infine lcl è assunta pari all’estensione della parte di pilastro priva di
tamponamento.
Figura B.3.7 Diagramma del momento flettente nel caso di colonne adiacenti a tamponamenti con altezza
minore (a sinistra) e diagramma del momento di colonne nel caso di assenza di tamponamenti (a destra)
P. Morandi, S. Hak, G. Magenes EUCENTRE 217
Research Report
Figura B.3.8 Estensione delle staffe nelle zone non a contatto con la muratura
d) se la lunghezza della colonna che non è a contatto con la parete di tamponamento risulta
essere minore di 1,5hc, si raccomanda che la forza di taglio sia sostenuta da armature
disposte secondo le due direzioni diagonali.
Figura B.3.10. Lunghezza critica per le colonne adiacenti a tamponamenti su un solo lato
218 Experimental and numerical seismic performance of strong clay masonry infills
Appendix B – Proposta di linee guida per la progettazione sismica di tamponamenti in muratura in
strutture in c.a.
B.3.5.3.4 Effetti locali: calcolo della forza di taglio agente sulle colonne dovuta alle tamponature
Nel caso di tamponature aderenti al telaio, è necessario effettuare una verifica del taglio agente
sulle colonne in c.a. dovuto alla componente orizzontale della spinta del puntone dato dalla
presenza delle tamponature. In Figura B.3.11 è riportata la distribuzione delle sollecitazioni su un
pilastro che scaturisce dalla spinta locale del puntone. Tale spinta, che risulta tanto più elevata
all’aumentare della rigidezza e della resistenza del tamponamento, può provocare rotture fragili
alle estremità di colonne non sufficientemente resistenti.
I pilastri in c.a. dovranno dunque essere dimensionati per una forza di taglio prodotta dalla
compressione del puntone diagonale dalla parete di tamponamento agente su una lunghezza, lc,
delle colonne in c.a.. Si deve dunque verificare che:
in cui VC,Rd,lc è la resistenza al taglio all’interno della lunghezza lc di estremità delle colonne e
VC,Ed,lc è la forza di taglio sulla lunghezza lc prodotta dalla componente orizzontale della spinta del
tamponamento.
Figura B.3.11. Taglio agente sulla colonna dovuto alla componente orizzontale della spinta del
tamponamento
Sulla normativa nazionale non vi sono attualmente indicazioni rivolte al calcolo della forza di taglio
sulle colonne innescata dalla reazione di contatto con le tamponature.
L’EC8 riporta invece alcune generiche indicazioni sia per valutare la lunghezza della colonna su
cui è applicata la forza di compressione del puntone diagonale della parete di tamponamento, sia
per valutare l’intensità dell’azione di taglio. In particolare, al punto 5.9 (4), si raccomanda che la
lunghezza, lc, delle colonne su cui è applicata la forza di compressione del puntone diagonale della
parete di tamponamento sia verificata a taglio per la più piccola (VC,Ed,lc) delle seguenti due forze di
taglio VC,Ed,w e VC,Ed,M :
La forza di taglio VC,Ed,w è la componente orizzontale della forza del puntone della parete di
tamponamento, assunta pari alla resistenza a taglio orizzontale del pannello, come stimata sulla
base della resistenza a taglio del letto di malta (in base a quanto riportato nell’EC8):
P. Morandi, S. Hak, G. Magenes EUCENTRE 219
Research Report
VC , Ed =
,w F=
w Aw ⋅ f v (B.3.29)
dove Aw è l’area della sezione della parete di tamponamento (tw·Lw), fv è la resistenza a taglio della
muratura; essa può essere assunta pari al valore fv0 con riferimento al criterio di resistenza a taglio
della muratura in assenza di sforzo normale, fornito dalle NTC e dall’Eurocodice 6 (CEN [2004b]).
La forza di taglio VC,Ed,M è calcolata in conformità alla seguente espressione:
2 M C , Rd
VC , Ed , M = γ Rd (B.3.30)
lc
dove γRd è pari al coefficiente di sovraresistenza (1.30 per CD”A” e 1.10 per CD”B”) ed assumendo
che la resistenza a pressoflessione della colonna MC,Rd, si sviluppi ai due estremi della lunghezza
di contatto, lc.
bw 0.25d w 0.125d w 2
=lc = = (B.3.31)
2cos θ 2cos θ Lw
Figura B.3.13. Definizione delle grandezze per la valutazione della lunghezza di contatto lc
220 Experimental and numerical seismic performance of strong clay masonry infills
Appendix B – Proposta di linee guida per la progettazione sismica di tamponamenti in muratura in
strutture in c.a.
Figura B.3.14. Valutazione della quota di attivazione della forza massima del puntone aw,j per tamponamenti
“robusti”
P. Morandi, S. Hak, G. Magenes EUCENTRE 221
Research Report
Figura B.3.15. Valutazione della quota di attivazione della forza massima del puntone aw,j per tamponamenti
“leggeri”
B.3.6 Verifiche di sicurezza in termini di resistenza (SLV) fuori piano sui tamponamenti
B.3.6.1 CASO A): Tamponamenti monostrato racchiusi nella cornice del telaio.
S a ⋅ Wa
Fa = (B.3.34)
qa
dove Fa è la forza sismica orizzontale agente al centro di massa del tamponamento, Wa è il peso
del tamponamento pari a pg·tw·Lw·hw (pg peso di volume della muratura, tw, Lw e hw sono,
rispettivamente lo spessore, la lunghezza e l’altezza della tamponatura), Sa è l’accelerazione
massima, adimensionalizzata rispetto a quella di gravità che il pannello subisce durante il sisma e
qa è il fattore di struttura del pannello (in assenza di specifiche determinazioni può essere posto
pari a 2, come riportato in Tabella B.3.4 tratta dalle NTC 2008). In mancanza di analisi più accurate
Sa può essere calcolato nel seguente modo:
3(1 + z H )
Sa = α ⋅ S ⋅ − 0.5 (B.3.35)
1 + (1 − Ta T1 )
2
in cui α è il rapporto tra l’accelerazione massima del terreno ag su sottosuolo tipo A da considerare
nello stato limite in esame (v. § 3.2.1 NTC 2008) e l’accelerazione di gravità g, S è il coefficiente
che tiene conto della categoria di sottosuolo e delle condizioni topografiche secondo quanto
riportato nel § 3.2.3.2.1 delle NTC, Ta è il periodo fondamentale di vibrazione della tamponatura, T1
è il periodo fondamentale di vibrazione della costruzione nella direzione considerata, z è la quota
del baricentro della tamponatura misurata a partire dal piano di fondazione, H è l’altezza della
costruzione misurata a partire dal piano di fondazione (Figura B.3.16). Il valore del coefficiente
sismico Sa non può essere assunto minore di α·S. Il periodo del modo di vibrare principale nella
direzione in esame T1 può essere calcolato pari a T1=C1·H3/4 in cui C1 può essere posto
conservativamente pari a 0.050 (in questo modo si tiene conto in maniera semplificata dell’effetto
irrigidente dei tamponamenti, come proposto da Pinho e Crowley [2009]) e H è l’altezza della
costruzione dal piano di fondazione misurata in metri. Il periodo fondamentale di vibrazione della
222 Experimental and numerical seismic performance of strong clay masonry infills
Appendix B – Proposta di linee guida per la progettazione sismica di tamponamenti in muratura in
strutture in c.a.
tamponatura Ta può essere stimato con la seguente espressione che approssima la risposta fuori
piano delle pareti ad elementi trave verticale elastici incernierati alle estremità di altezza hw con
massa distribuita m e rigidezza flessionale di sezione E·I, (E modulo elastico della muratura ed I
momento di inerzia fuori piano della sezione del pannello):
2hw2 m (B.3.36)
Ta =
π EI
In ogni caso, per la valutazione delle accelerazioni fuori piano si possono utilizzare anche altre
espressioni, prendendo per esempio come riferimento studi del passato ed altre normative.
Dividendo la forza sismica Fa per la superficie laterale del pannello si ottiene una pressione agente
fuori piano sulla tamponatura wa, pari a:
Fa
wa = (B.3.37)
Lw ⋅ hw
con Lw e hw rispettivamente pari alla lunghezza ed all’altezza del pannello murario soggetto a
verifica.
tw
2
wa ≤ wR ,β = βa , j ⋅ 0.72 ⋅ f d
βa , j ⋅ wR = ⋅ (B.3.38)
hw
La resistenza wR è stata valutata con un modello ad arco verticale perfettamente aderente alla
trave ipotizzando la formazione di una fessura orizzontale a metà altezza, come riportato in Figura
B.3.18, e trascurando l’inflessione dell’arco (ipotesi accettabile con hw/tw < 25 con hw e tw
rispettivamente l’altezza e lo spessore della tamponatura) e dove fd è la resistenza a compressione
di calcolo della muratura.
224 Experimental and numerical seismic performance of strong clay masonry infills
Appendix B – Proposta di linee guida per la progettazione sismica di tamponamenti in muratura in
strutture in c.a.
( ra ,1 − 1) δ w, j / δ SLD + 1, δ w, j ≤ δ SLD
=βa, j ra ,1 , δ SLD < δ w, j ≤ δ SLV (B.3.39)
ra ,2 , δ SLV < δ w, j
I valori di ra,1 e ra,2 che si assume corrispondano al raggiungimento dei requisiti prestazionali
rispettivamente allo stato limite di danno e allo stato limite ultimo e quindi ai relativi valori di drift
δSLD e δSLV, sono stati stimati sulla base di osservazioni sperimentali, come riassunto in Tabella
B.3.5, in funzione delle diverse tipologie di tamponatura.
Figura B.3.19 Coefficiente semplificato di riduzione di resistenza fuori piano βa,j per tamponamenti “robusti”
senza aperture
Figura B.3.20 Coefficiente semplificato di riduzione di resistenza fuori piano βa,j per tamponamenti “robusti”
con aperture
Figura B.3.21 Coefficiente semplificato di riduzione di resistenza fuori piano βa,j per tamponamenti “leggeri”
226 Experimental and numerical seismic performance of strong clay masonry infills
Appendix B – Proposta di linee guida per la progettazione sismica di tamponamenti in muratura in
strutture in c.a.
B.3.6.2 CASO B1): tamponamenti pluristrato con entrambi i paramenti racchiusi nella
cornice del telaio e non collegati tra di loro.
S a ,1 ⋅ Wa ,1
Fa ,1 = (B.3.40)
qa
Sa , 2 ⋅ Wa , 2
Fa , 2 = (B.3.41)
qa
dove:
3(1 + z H )
S a ,1 = α ⋅ S ⋅ − 0 . 5
1 + (1 − Ta ,1 T1 )
2 (B.3.44)
3(1 + z H )
Sa , 2 = α ⋅ S ⋅ −
1 + (1 − Ta , 2 T1 )
2
0 . 5 (B.3.45)
Fa ,1
wa ,1 = (B.3.46)
Lw,1 ⋅ hw,1
Fa ,2
wa ,2 = (B.3.47)
Lw,2 ⋅ hw,2
dove i parametri con pedice “1” sono riferiti al paramento esterno mentre quelli con pedice “2” sono
riferiti al paramento interno e sono già stati definiti nel paragrafo precedente.
P. Morandi, S. Hak, G. Magenes EUCENTRE 227
Research Report
Figura B.3.22 Tamponamento pluristrato con entrambi i paramenti racchiusi nella cornice del telaio
t
2
wa ,2 ≤ wR ,2,β =
βa ,2, j ⋅ wR ,2 βa ,2, j ⋅ 0.72 ⋅ f d ,2 ⋅ w,2
= (B.3.49)
h
w,2
dove fd,1 e fd,2 sono le resistenze a compressione di calcolo della muratura esterna ed interna e
βa,1,j βa,2,j sono i coefficienti riduttivi che tengono conto della riduzione della resistenza fuori piano
dovuta all’eventuale precedente danneggiamento nel piano in funzione dei valori di drift interpiano
effettivo δj,w allo SLV. I parametri βa,1,j e βa,2,j possono essere ricavati come proposto per il caso di
tamponature monostrato, considerando implicitamente che i 2 paramenti siano soggetti alle
medesime deformazioni nel piano.
B.3.6.3 CASO B2): tamponamenti pluristrato con entrambi i paramenti racchiusi nella
cornice del telaio collegati tra di loro.
Sa ⋅ Wa
Fa = (B.3.50)
qa
dove:
228 Experimental and numerical seismic performance of strong clay masonry infills
Appendix B – Proposta di linee guida per la progettazione sismica di tamponamenti in muratura in
strutture in c.a.
3 (1 + z H )
Sa = α ⋅ S ⋅ − 0.5 (B.3.52)
1 + (1 − Ta T1 )2
Fa
wa = (B.3.53)
Lw ⋅ hw
dove i parametri con pedice “1” sono riferiti al paramento esterno mentre quelli con pedice “2” sono
riferiti al paramento interno; il periodo Ta può essere valutato calcolando la rigidezza fuori piano
come nel caso di una parete monostrato di spessore equivalente tw,eq e la massa come somma
delle masse dei due paramenti.
wa ≤ wR ,β =
βa , j ⋅ wR =
βa , j ⋅ 0.72 ⋅ f d ⋅ (B.3.54)
w
h
dove fd è la resistenza a compressione di calcolo della muratura e βa,j è il coefficiente riduttivo che
tiene conto della riduzione della resistenza fuori piano dovuta all’eventuale precedente
danneggiamento nel piano in funzione dei valori di drift interpiano effettivo δj,w allo SLV. Se la
resistenza a compressione della muratura dei 2 paramenti fosse diversa, si consiglia di utilizzare la
resistenza più bassa mentre, per quanto riguarda il parametro βa,j si può ricavare come nel caso
delle tamponature monostrato andando ad applicare, nel caso di due diverse tipologie di
tamponamento (per es., tamponamento “robusto” esterno e “leggero” interno) il valore più
cautelativo.
Figura B.3.23 Tamponamento pluristrato con entrambi i paramenti racchiusi nella cornice del telaio collegati
tra di loro
P. Morandi, S. Hak, G. Magenes EUCENTRE 229
Research Report
Tabella B.3.6 Valori di snellezza massima λmax,a per pareti di tamponamento monostrato in soluzione
“robusta”
zw/H 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
agS
0.050 27.8 26.4 25.2 24.2 23.2 22.4 21.7 21.0 20.4 19.8
0.075 22.7 21.6 20.6 19.7 19.0 18.3 17.7 17.1 16.6 16.2
0.100 19.6 18.7 17.8 17.1 16.4 15.8 15.3 14.8 14.4 14.0
0.125 17.6 16.7 15.9 15.3 14.7 14.2 13.7 13.3 12.9 12.5
0.150 16.0 15.2 14.6 13.9 13.4 12.9 12.5 12.1 11.8 11.4
0.175 14.8 14.1 13.5 12.9 12.4 12.0 11.6 11.2 10.9 10.6
0.200 13.9 13.2 12.6 12.1 11.6 11.2 10.8 10.5 10.2 9.9
0.225 13.1 12.4 11.9 11.4 11.0 10.6 10.2 9.9 9.6 9.3
0.250 12.4 11.8 11.3 10.8 10.4 10.0 9.7 9.4 9.1 8.9
0.275 11.8 11.3 10.7 10.3 9.9 9.6 9.2 9.0 8.7 8.5
0.300 11.3 10.8 10.3 9.9 9.5 9.1 8.8 8.6 8.3 8.1
0.325 10.9 10.4 9.9 9.5 9.1 8.8 8.5 8.2 8.0 7.8
0.350 10.5 10.0 9.5 9.1 8.8 8.5 8.2 7.9 7.7 7.5
NOTA: per valori non contemplati in tabella è prevista l’interpolazione lineare ma non sono ammesse
estrapolazioni
Nel caso per esempio di un edificio in c.a. con altezza massima delle pareti di tamponamento hw
pari a 2.9 m, sito in una zona sismica con accelerazione massima del terreno ag·S allo SLV pari a
0.20 g, si potrebbero realizzare, per tutta l’altezza dell’edificio, pareti di tamponamento monostrato
di spessore minimo tw pari a hw/λmax,a = 2.90/9.9 = 0.29 m; si potrebbe dunque utilizzare, per
esempio, un blocco di spessore pari a 30 cm del gruppo 2 (in base alla classificazione
dell’Eurocodice 6 (EC6, CEN, 2004b), con percentuale di foratura verticale del 50%-55%, con
resistenza a compressione verticale media fb ≥ 8 MPa e malta M5 che fornisce, in base alle
espressioni del punto 3.6.1.2 dell’EC6, una resistenza caratteristica della muratura fk pari a 3.13
MPa.
P. Morandi, S. Hak, G. Magenes EUCENTRE 231
Research Report
Tipologia di tw Em fb fm fk fvk0 % di
tamponamento [m] [MPa] [MPa] [MPa] [MPa] [MPa] foratura
“Robusta”
0.30 3660 10.0 5.0 3.66 0.20 55
Gruppo blocco (EC6): 2
“Debole/leggera” 0.08 920 0.92
2.0 5.0 0.20 60
Gruppo blocco (EC6): 4 0.10* 1500* 1.50*
* con intonaco
0.80
Elastic SLV
0.70 Categoria stratigrafica B:
Design SLV
0.60 S = 1.198
Elastic SLD
0.50
Sd [g]
0.40 SLV:
0.30 ag = 0.209g, agS = 0.250g
0.20
0.10
SLD:
0.00
ag = 0.070g, agS = 0.084g
0.00 0.50 1.00 1.50 2.00 2.50 3.00 3.50 4.00
T [s]
Figura B.4.5 Spettri elastici e di progetto (Basilicata, Ruvo del Monte (PZ))
In Tabella B.4.5 si riassumono i periodi di vibrazione e la percentuale delle masse partecipante per
i primi 7 modi significativi ottenute dai risultati dell’analisi lineare modale per le due direzioni
principali x e y; inoltre, i primi modi di vibrazione traslazionali sono illustrati in Figura B.4.7.
Modo 1 Modo 3
Fj Vj dr,j KS,j
[KN] [KN] [m] [KN/m]
Longitudinale
1 20 914 0.0008 1142655
2 63 894 0.0016 558679
Piano
Tabella B.4.7 Valutazione del parametro semplificati della rigidezza del tamponamento KI,j
Per ogni piano j dell’edificio, a partire dalle richieste di drift interpiano ottenute sulla struttura “nuda”
(δ,j= dr,SLD/hj), sono stati calcolati i corrispondenti valori drift della configurazione tamponata (δw,j)
sia per lo SLD che per lo SLV, come riportato in Tabella B.4.8, da confrontare con i valori limite di
Tabella B.4.3, sulla base della procedura definita nel paragrafo B.3.3 e B.3.4. Si ricorda che al
piano terra si considera cautelativamente un valore di Cj pari a 0 e dunque lo spostamento effettivo
equivale a quello sul telaio “nudo”. La curva bilineare che mette in relazione i drift della
configurazione “nuda” e di quella tamponata corrispondente ai diversi valori di Cj ed ottenuta per la
direzione longitudinale e per quella trasversale, è mostrata rispettivamente in Figura B.4.8a ed in
Figura B.4.8b. La verifica dei drift per la direzione longitudinale e trasversale è illustrata
rispettivamente in Figura B.4.9a e Figura B.4.9b; in questo caso, come limite deformativo, si è
considerato quello relativo alle tamponature “deboli/leggere”.
238 Experimental and numerical seismic performance of strong clay masonry infills
Appendix B – Proposta di linee guida per la progettazione sismica di tamponamenti in muratura in
strutture in c.a.
Tabella B.4.8. Drift elastici sul telaio “nudo” e stima dei drift del telaio tamponato
(a) (b)
1.00 1.00
Drift δw,j [%]
0.50 0.50
0.00 0.25
Cj 0.39 0.53 0.00 0.84
0.76 1.53 Cj 1.13 1.49
0.00 1.93 3.53
0.00
0.00 0.50 1.00 1.50 0.00 0.50 1.00 1.50
Drift δj [%] Drift δj [%]
Figura B.4.8. Curva bilineare per diversi valori di Cj: (a) direzione longitudinale; (b) direzione trasversale
P. Morandi, S. Hak, G. Magenes EUCENTRE 239
Research Report
a)
SLD SLV
(Longitudinale) (Longitudinale)
6 6
5 5
4 4
Storey
Storey
3 3
2 2
1 1
Bare Bare
Infilled Infilled
0 0
0.00 0.10 0.20 0.30 0.40 0.50 0.00 0.20 0.40 0.60 0.80 1.00 1.20
Drift [%] Drift [%]
(b)
SLD SLV
(Trasversale) (Trasversale)
6 6
5 5
4 4
Storey
Storey
3 3
2 2
1 1
Bare
Bare
Infilled
0 Infilled
0
0.00 0.10 0.20 0.30 0.40 0.50 0.00 0.20 0.40 0.60 0.80 1.00 1.20
Drift [%] Drift [%]
Figura B.4.9. Verifica dei drift: (a) direzione longitudinale; (b) direzione trasversale
B.4.6 Verifiche degli effetti locali sulle colonne in c.a. dovuti ai tamponamenti
La verifica degli effetti locali è stata effettuata su una colonna d’angolo del piano terra nella
direzione trasversale. In base all’EC8, tutta l’altezza della colonna deve essere considerata come
una regione critica ed i dettagli costruttivi realizzati di conseguenza. La sezione della colonna è
riportata in Figura B.4.10.a. Le travi di entrambe le direzioni sostenute dalla colonna d’angolo sono
armate allo stesso modo, come mostrato in Figura B.4.10.b.
(a) (b)
Longitudinale:
Longitudinale:
6φ18
8φ18
Trasversale:
Trasversale:
φ8/9.0 cm (in zona critica)
φ8/8.0 cm (in zona
φ8/22.0 cm (fuori da zona
critica)
critica)
Figura B.4.10. (a) Colonna d’angolo del piano terra; (b) Travi sostenute dalla colonna d’angolo
- Forza di taglio VC,Ed,w alternativa (componente orizzontale della forza del puntone attivata):
VC,Ed,w = Fw = aw,j·Aw·fv = 0.845·0.30·4.6·0.2·103 = 233.2 kN
240 Experimental and numerical seismic performance of strong clay masonry infills
Appendix B – Proposta di linee guida per la progettazione sismica di tamponamenti in muratura in
strutture in c.a.
- Attivazione della forza massima del puntone compresso aw,j (δw,j = 0.36 %):
=lc =
(
0.125d w 2 0.125 ⋅ 2.825 + 4.6
2 2
= 0.792 m
)
Lw 4.6
- Verifica di sicurezza:
VC,Rd,lc = 473 KN > VC,Ed = 233.2 KN Ok.
T1 =
Ct H 3 / 4 =
0.05 ⋅ 18.63 / 4 =
0.45 s
- Parametro della riduzione della resistenza fuori piano βa,j (δw,j = 0.35 %):
t
2
0.1 3
2
wR ,β = βa , j ⋅ 0.72 ⋅ f d
βa , j ⋅ wR = ⋅ w =0.20 ⋅ 0.72 ⋅ 1.5 ⋅ ⋅ 10 =
0.30 kN / m 2
hw 2.7
- Verifica di sicurezza :
=wa 0.28 kN / m 2 ≤=
wR ,β 0.30 kN / m 2
- Coefficiente di sicurezza:
wR ,β / wa = 1.07
T1 =
Ct H 3 / 4 =
0.05 ⋅ 18.63 / 4 =
0.45 s
- Parametro della riduzione della resistenza fuori piano βa,j (δw,j = 0.35 %):
tw
2
0.3 3
2
wR ,β = βa , j ⋅ 0.72 ⋅ f d
βa , j ⋅ wR = ⋅ = 0.40 ⋅ 0.72 ⋅ 3.66 ⋅ ⋅ 10 =
13.01 kN / m 2
w
h 2.7
- Verifica di sicurezza:
=wa 0.86 kN / m 2 ≤=
wR ,β 13.01 kN / m 2
- Coefficiente di sicurezza:
wR ,β / wa = 15.11
P. Morandi, S. Hak, G. Magenes EUCENTRE 243
Research Report