Vous êtes sur la page 1sur 205

RESPONSE ESTIMATION OF REINFORCED CONCRETE COLUMNS SUBJECTED

TO LATERAL LOADS

A Thesis

Presented in Partial Fulfillment of the Requirements for

the Degree Master of Science in the

Graduate School of The Ohio State University

By

Muhammad Shadab Lodhi

Graduate Program in Civil Engineering

The Ohio State University

2010

Masters Examination Committee:

Dr. Halil Sezen, Adviser

Dr. Shive K. Chaturvedi

Dr. Ethan Kubatko


Copyright by

Muhammad Shadab Lodhi

2010
ABSTRACT

There is significantly large number of reinforced concrete buildings in the US and many

parts of the world that are not designed according to prevailing seismic code provisions.

These buildings are often characterized by low lateral displacement capacity and rapid

degradation of shear strength and hence are vulnerable to severe damage or even collapse

during strong ground motions. Typically, columns in such buildings have insufficient and

widely spaced transverse reinforcement and lack essential seismic reinforcement details.

These columns exhibit lack of strength and ductility in reverse cyclic loading and are

vulnerable to brittle shear failure and loss of axial load carrying capacity during strong

ground shaking. The need to assess their vulnerability to earthquake damage and hence

suggesting the desired level of retrofit requires evaluation of the expected behavior in

terms of strength and deformation capacity. This can be achieved by estimating the load-

deformation response considering all potential failure mechanisms associated with axial,

flexure and shear behavior. This study evaluates two different macro models to

investigate their capabilities to determine the lateral response. Based on the comparison

of the predicted responses with experimental test data, an analytical procedure is

proposed. In the proposed procedure, flexural and shear deformations are determined

while considering interaction between these mechanisms. Flexural deformations are

calculated by incorporating effect of shear deformation on flexural performance through

ii
compression softening factor. Likewise, shear deformations are calculated by employing

axial strain and shear stresses from axial-flexure model to consider the effect of flexural

deformations on shear behavior. The interaction between axial-flexure and axial-shear

mechanisms allows for accurate response estimation while decoupled flexural analysis

minimizes iterations within the analysis and make the process relatively simple and easy.

In addition, buckling of compression bars is also incorporated by considering separate

stress-strain relationships for reinforcing steel in tension and compression. The

displacement due to reinforcement slip are calculated separately and added to flexural

and shear displacements. Total lateral response is obtained by combining deformation

components due to flexure, shear and reinforcement slip according to a set of rules based

on comparison of the column yield, flexural and shear strength. The comparison of the

predicted responses with experimental test data showed that proposed procedure performs

well in estimating response of reinforced concrete columns to lateral loads.

iii
Dedication

This document is dedicated to my parents, M.N.U.K Lodhi and Hafeez Akhtar

and my wife, Tabassum Shadab

and my children, Muhammad Ibrahim Khan Lodhi and Aliza Lodhi

iv
ACKNOWLEDGMENTS

The completion of this thesis would not have been possible without the help of

many people. First and foremost, I would like to thank my adviser, Dr. Halil Sezen, for

his continuous support and endless patience throughout the years so far. Without his

guidance in our meetings and discussions, I would not have been able to complete my

research. I would like to acknowledge Professors Shive Chaturvedi and Dr. Ethan

Kubatko for giving their time as members of my committee.

I would like to thank my friends, colleagues and family members for their

unconditional support over the years. I also appreciate the support of my friend and

officemate Kyong Yun Yeau who has made the rough times manageable and the good

times memorable.

I am also deeply obliged to Ministry of Defence, Government of Pakistan for

providing me this wonderful opportunity of graduate studies and supporting myself and

my family financially.

v
VITA

June 18, 1971 .. Born Pakistan

1993 B.S. Civil Engineering,


Military College of Engineering Risalpur,
Pakistan

1993 - 2006 .... Commissioned Military Service


Pakistan Army Corps of Engineers

2006 to present. MS/PhD Graduate Student


The Ohio State University, Columbus, Ohio

FIELDS OF STUDY

Major Field: Civil Engineering

Specialization: Structural Engineering

vi
TABLE OF CONTENTS

Page
Abstract .. ii
Dedication... iv
Acknowledgments . v
Vita vi
List of Tables . x
List of Figures xi

Chapters:
1 INTRODUCTION . 1
1.1 Overview 1
1.2 Research Impetus ... 3
1.3 Research Objective 5
1.4 Organization ... 6

2 DISPLACEMENT COMPONENT MODEL...... 11


2.1 Introduction 11
2.2 Deformation Components... 12
2.2.1 Flexural Deformations.. 12
2.2.2 Reinforcement Slip Deformations. 15
2.2.3 Shear Deformations... 18
2.3 Total Response. 20
2.4 Shear Failure Due to High Displacement Ductility.. 24
2.5 Implementation of Displacement Component Model 25
2.5.1 Test Specimens and Material Properties.. 25
2.5.2 Material Constitutive Relationships.. 26
2.5.3 Comparison of Model and Test Results. 28
2.6 Conclusion 29

3 AXIAL SHEAR FLEXURE INTERACTION (ASFI) APPROACH -


THEORY . 39
3.1 Introduction 39
3.2 Background. 41
3.3 Axial-Flexure Model... 42
3.3.1 Effect of Shear Deformations on Flexural Response.. 43

vii
Page
3.3.2 Axial Strain due to Flexure. 43
3.3.3 Flexural Deformations.... 45
3.4 Axial-Shear Model. 46
3.4.1 MCFT Constitutive Laws... 49
3.4.1.1 Concrete in Compression....... 50
3.4.1.2 Concrete in Tension.... 50
3.4.1.3 Reinforcement Stress-Strain Relationship... 51
3.4.2 Consideration of Local Cracks Conditions of Integration.. 53
3.4.3 Material Stiffness Formulations. 56
3.5 Development of Axial Shear Flexure Interaction Approach.. 58
3.5.1 Compatibility Conditions/Relationships. .. 59
3.5.2 Equilibrium Conditions/Relationships 60
3.5.3 Constitutive Laws... 61
3.5.4 Interactions Considered.............. 62
3.5.4.1 Interaction Strain Methodology.. 62
3.5.4.2 Concrete Compression Softening... 64
3.5.5 Stiffness Model in ASFI approach... 64
3.6 Conclusions. 65

4 IMPLEMENTATION OF ASFI APPROACH.. 71


4.1 Introduction.... 71
4.2 Response Estimation of Columns by ASFI approach. 72
4.2.1 Pullout Effect Consideration and Total Deformations... 73
4.2.2 Analytical Procedure for Implementation of ASFI
approach... 75
4.2.3 Solution Technique for Implementing analytical
Procedure 77
4.3 Implementation and Verification of ASFI approach... 80
4.3.1 Details of Test Specimens and Material Properties. 81
4.3.2 Constitutive Laws... 81
4.3.2.1 Constitutive Law for Concrete in Compression... 82
4.3.2.2 Constitutive Law for Concrete in Tension. 83
4.3.2.3 Constitutive Law for Reinforcements. 84
4.3.3 Flexural Deformation Model.. 85
4.3.4 Pullout Deformation Model 85
4.3.5 Buckling of Compression Bars and Post-Peak Analysis 86
4.4 Comparison of the Results...... 87
4.5 Summary and Conclusions. 87

5 COMPARISON OF COMPONENT DISPLACEMENT MODEL AND


ASFI APPROACH..... 99
5.1 Introduction. 99
5.2 Implementation of the Analytical Models.. 100
5.2.1 Test Specimens and Material Properties. 100

viii
Page
5.2.2 Material Constitutive Relationships 101
5.2.2.1 Concrete in Compression 101
5.5.2.2 Concrete in Tension 103
5.2.2.3 Reinforcing Steel ... 103
5.2.3 Flexural Deformation Model...... 104
5.2.4 Pullout Deformation Model.... 104
5.2.5 Buckling of Compression Bars... 105
5.3 Comparison of Predicted and Experimental Response... 105
5.3.1 Flexural Displacements... 106
5.3.2 Reinforcement Slip Displacements 108
5.3.3 Shear Displacements... 108
5.3.4 Total Response 112
5.3 Conclusions. 116

6 DEVELOMPENT AND IMPLEMENTATION OF THE PROPOSED


MODEL.. 130
6.1 Introduction. 130
6.2 Desired Characteristic of Proposed Analytical Procedure.. 131
6.3 Proposed Interaction Methodology. 135
6.3.1 Interaction of Concrete Compression Softening. 135
6.3.2 Interaction of Axial Strains. 139
6.4 Buckling of Compression Bars... 141
6.4.1 Dhakal and Maekawa (2002) Bar Buckling Model... 143
6.4.2 Potger et al. (2001) Bar Model... 144
6.4.3 Implementation of Bar Buckling Models... ... 146
6.4.4 Proposed Model for Buckling of Compression Bars...... 146
6.5 Flexural Analysis 147
6.6 Reinforcement Slip Deformation Model 148
6.7 Shear Deformations.... 148
6.8 Total/Combined Response . 151
6.8.1 Pre-Peak response... ... 152
6.8.2 Post-Peak Response. .. 152
6.9 Analytical Procedure and Solution Technique for Proposed
Model.. 153
6.10 Implementation of the Model and Comparison of Results. 160
6.11 Conclusion.. 161

7 CONCLUSION . 174
7.1 Summary 174
7.1.1 Component Models 175
7.1.2 Combined Model ... 177
7.2 Conclusions 177
7.3 Recommendations for Future Work .. 179

ix
Page
List of References . 181

x
LIST OF TABLES

Table Page

4.1 Geometric and Material Properties of the Test Specimens. 98

xi
LIST OF FIGURES

Figure Page

1.1 Olive View Hospital damaged in the 1971 San Fernando Earthquake
(Steinbrugge, K. V., NISEE).. 8

1.2 Failure of first story columns in the Olive View Hospital (Steinbrugge,
K. V., NISEE). 8

1.3 Imperial County Services Building damaged during the 1979 Imperial
Valley earthquake (Bertero, V. V., NISEE)... 9

1.4 Collapsed reinforced concrete buildings during 2005 Pakistan


earthquake (Nawed, K., eNKay Solutions). 9

1.5 Examples of commonly observed deficiencies and damage in reinforced


concrete construction after 2005 Pakistan earthquake (EEC, Department
of Civil Engineering NWFP UET Peshawar). 10

2.1 Components of lateral deformation in a reinforced concrete column 30

2.2 Plastic Hinge Model for Determining Flexural Deformation 30

2.3 Bar Slip Model ... 31

2.4 Reinforcement slip rotation ... 31

2.5 Shear displacement model . 32

2.6 Spring representation for total response model for fixed ended column 32

2.7 Classification of the column and Lateral load-displacement relationships 33

2.8 Rules governing the combination of deformation components.............. 33

xii
Figure Page
2.9 Constitutive relationships for concrete in compression.. 34

2.10 Constitutive relationship for reinforcing steel 34

2.11 Lateral load flexural deformation relationships. 35

2.12 Lateral load slip deformation relationships.... 36

2.13 Lateral load shear deformation relationships.. 37

2.14 Lateral load total deformation relationships.... 38

3.1 Concrete compression softening... 38

3.2 Axial-Flexural model and determination of axial strain due to flexure 67

3.3 Plastic hinge model for flexural deformations ... 67

3.4 (a) Membrane element subjected to in-plane stresses (b) Average strains
in-plane loading (c) Average strains in cracked concrete (d) Mohrs
circle for average strains. 68

3.5 Average stress-strain relationship for (a) Cracked concrete compression


(b) Cracked concrete in tension (c) Reinforcing Steel..... 68

3.6 (a) Calculated average stresses (b) Local stresses at a crack.. 69

3.7 Parameters influencing crack spacing .... 70

3.8 Coordinate transformation for concrete and steel material stiffness.......... 70

3.9 Spring model for axial-shear-flexural element 70

4.1 Flow chart for implementation of ASFI approach for columns 89

4.2 Constitutive relationships for concrete in compression.. 90

4.3 Constitutive relationship for concrete in tension ... 90

4.4 Constitutive relationship for reinforcing steel ... 91

4.5 Rigid body rotation of column due to reinforcement slip or pullout.... 91

4.6 Pullout deformation model (Okamura and Maekawa, 1991)................. 92

xiii
Figure Page
4.7 Comparison of the results for Specimen-5 (Ousalem et al, 2005).. 93

4.8 Comparison of the results for Specimen-12 (Ousalem et al, 2005)............ 94

4.9 Comparison of the results for Specimen-14 (Ousalem et al, 2005) 95

4.10 Comparison of the results for Specimen-15 (Ousalem et al, 2005). 96

4.11 Comparison of the results for Specimen-16 (Ousalem et al, 2005) 97

5.1 Constitutive relationships concrete in compression 117

5.2 Constitutive relationship concrete in tension.. 117

5.3 Constitutive relationship for reinforcing steel 118

5.4 Plastic Hinge Model for Determining Flexural Deformation. 118

5.5 Lateral load flexural displacement relationship for Specimen-1. 119

5.6 Lateral load flexural displacement relationship for Specimen-2. 119

5.7 Lateral load flexural displacement relationship for Specimen-3. 120

5.8 Lateral load flexural displacement relationship for Specimen-4. 120

5.9 Effect of concrete tensile stress in flexural section analysis... 121

5.10 Effect of concrete compression softening in flexural section analysis... 121

5.11 Effect of compression bars buckling in flexural section analysis... 122

5.12 Lateral load slip displacement relationship for Specimen-1 122

5.13 Lateral load slip displacement relationship for Specimen-2 123

5.14 Lateral load slip displacement relationship for Specimen-3 123

5.15 Lateral load slip displacement relationship for Specimen-4 124

5.16 Lateral load shear displacement relationship for Specimen-1. 124

5.17 Lateral load shear displacement relationship for Specimen-2. 125

xiv
Figure Page
5.18 Lateral load shear displacement relationship for Specimen-3. 125

5.19 Lateral load shear displacement relationship for Specimen-4. 126

5.20 Lateral load displacement relationship for Specimen-1.. 126

5.21 Lateral load displacement relationship for Specimen-2.. 127

5.22 Lateral load displacement relationship for Specimen-3.. 127

5.23 Lateral load displacement relationship for Specimen-4. 128

5.24 Effect of plastic hinge length on total response by ASFI approach


(Specimen-1). 129

6.1 Determination of concrete principal tensile strain. 162

6.2 Dhakal and Maekawa (2002) bar buckling model. 162

6.3 Potger et al. (2001) bar buckling model. 163

6.4 Effect of compression bar buckling Specimen 1. 163

6.5 Effect of compression bar buckling Specimen 2. 164

6.6 Effect of compression bar buckling Specimen 4. 164

6.7 Proposed bar buckling model. 165

6.8 Lateral load flexural displacement relationship for Specimen-1. 165

6.9 Lateral load flexural displacement relationship for Specimen-2. 166

6.10 Lateral load flexural displacement relationship for Specimen-3. 166

6.11 Lateral load flexural displacement relationship for Specimen-4. 167

6.12 Lateral load slip displacement relationship for Specimen-1 167

6.13 Lateral load slip displacement relationship for Specimen-2 168

6.14 Lateral load slip displacement relationship for Specimen-3 168

xv
Figure Page
6.15 Lateral load slip displacement relationship for Specimen-4 169

6.16 Lateral load shear displacement relationship for Specimen-1. 169

6.17 Lateral load shear displacement relationship for Specimen-2. 170

6.18 Lateral load shear displacement relationship for Specimen-3. 170

6.19 Lateral load shear displacement relationship for Specimen-4. 171

6.20 Lateral load displacement relationship for Specimen-1.. 171

6.21 Lateral load displacement relationship for Specimen-2.. 172

6.22 Lateral load displacement relationship for Specimen-3.. 172

6.23 Lateral load displacement relationship for Specimen-4. 173

6.24 Comparison between compressive concrete constitutive laws... 173

xvi
CHAPTER 1

INTRODUCTION

1.1 Overview

There is a large inventory of reinforced concrete buildings in the US and other

parts of the world that are not designed according to modern seismic design provisions.

In the developed countries, these are the buildings which were constructed between

1930s and 1970s according to codes and standards available at the time. Even today, the

many developing countries that lag behind on seismic code developments continue to

design and build structures which do not have essential details deemed vital to withstand

large lateral loads. These buildings are often characterized by low lateral displacement

capacity and rapid degradation of shear strength and hence are vulnerable to severe

damage or even collapse during strong ground motions.

The reconnaissance of damage observed during the past earthquakes suggests that

poorly designed reinforced concrete columns are the most critical elements to sustain

damage leading to a potential building collapse. Typically, these columns have

insufficient and widely spaced transverse reinforcement with poor details such as 90-

degree end hooks and splicing of the longitudinal bars near the column ends, which are

the regions that typically experience the most inelastic deformations. Due to these

1
deficiencies, the columns may not have sufficient shear strength to develop the plastic

hinges at the ends. In addition, wide spacing of the ties does not provide good

confinement for the core concrete. As a result, the columns exhibit non-ductile behavior

and may fail suddenly.

The examples of the damage caused by earthquakes in non-seismic reinforced

concrete construction are shown in Figures 1.1 through 1.5. Figure 1.1 shows Olive View

Hospital building which was damaged in the 1971 San Fernando Earthquake. The

building sustained significant damage in almost all columns in the first story. Close up

views of the two columns in Figure 1.2 show severe shear damage to the columns. The

concrete is entirely crushed and the columns have virtually lost their ability to carry

lateral and axial loads. The Imperial County Services Building, as shown in Figure 1.3,

also suffered significant damage to its first-story columns during the Imperial Valley

earthquake of 1979. Transverse reinforcement did not provide sufficient confinement to

core concrete, resulting in crushing of the concrete and buckling of the longitudinal

reinforcement.

Notice that both buildings in above examples suffered heavy structural damage

but did not collapse. In developing countries, in addition to the absence of requisite

seismic detailing, the quality of construction is also very poor. Resultantly, catastrophic

failure of the structures is observed after every major earthquake. For example, Pakistan

earthquake of 2005 caused widespread damage to reinforced concrete buildings (Figure

1.4). A reconnaissance team of structural engineers that inspected the damage after the

earthquake found that many of the buildings that collapsed or sustained severe damage

suffered from lack of seismic details, poor quality of construction, or both (Amjad et al.,

2
2005). The non-seismic details in collapsed or damaged buildings included insufficient

lap-lengths, improper location of the lap splicing and insufficient concrete confinement

due to large spacing of ties. The team also observed irregular configurations with

soft/weak story or captive columns, constructions without any engineering design and

poor quality of concrete and workmanship. Figures 1.5 shows several examples of these

deficiencies and damaged reinforced concrete buildings from the Pakistan 2005

earthquake.

Seismic design follows strong column-weak beam philosophy where the

columns have sufficient strength and deformation capacity to push inelastic action and

damage to the beams. Damage of the beams affects mainly the immediate surroundings

whereas the failure of a lower-story column can potentially lead to collapse of the

building. Current design codes require buildings to have details that ensure ductile

behavior to withstand high seismic activity. These details include closely spaced ties in

columns, with 135-degree end hooks. The large amount of transverse steel gives the

columns sufficient shear strength to resist lateral earthquake loads, and the closely spaced

ties keep the core concrete confined at high displacements. Similarly, 135-degree end

hooks perform much better and remain closed during cyclic loading as these are

embedded in the core concrete which remains intact due to good confinement.

1.2 Research Impetus

The large numbers of reinforced concrete buildings that are not designed to

withstand earthquake are vulnerable to damage or collapse in a major seismic event. It is

possible to retrofit an existing building to increase its strength and ductility to enhance its

3
performance during an earthquake. The need to assess their vulnerability to earthquake

damage and hence suggesting the desired level of retrofit requires evaluation of the

expected behavior in terms of strength and deformation capacity. This can be achieved by

estimating load-deformation response of the concrete elements, such as beams and

columns, considering all potential failure mechanisms associated with axial, flexure and

shear behavior.

In addition to the retrofitting requirements, there are many situations where the

structures are required to be analyzed accurately to predict their structural responses.

Even important existing buildings and planned future structures may require performance

evaluation in terms of assessing their maximum load capacity, ultimate deformation

capacity, ductility and failure mechanism.

While much is known about the behavior of reinforced concrete components and

systems, there are still areas that require further understanding. For a structure whose

behavior is dominated by flexural mechanism, issues regarding its performance

evaluations have been studied well and design procedures are relatively well established.

However, for the structures whose behavior is affected by shear related mechanisms,

accurate modeling remains elusive with many available approaches and theories.

Understanding the fact that lateral deformation of a structure is mainly comprised of the

flexure and shear components, it must also be realized that these deformations do not

occur independently and these mechanisms interact with each other. Any analytical

procedure that aims to model overall lateral load-displacement relationship must take this

aspect into account. These are few issues that are the motivation for the research reported

here.

4
1.3 Research Objective

As explained earlier that columns tend to be the most critical structural

component in reinforced concrete structures not designed according to modern seismic

design codes. There are a number of studies investigating structural response of non-

ductile reinforced concrete columns following varying approaches. Some of these

approaches can predict structural behavior with good accuracy but employ very

complicated and computation-intensive analytical procedures which may not be

amenable and are difficult to implement. Contrary, many approaches simplify the

analytical process by making simplifying assumption on the structural response but in

most cases this is done at the cost of accuracy. Therefore, it is aimed in this study to

propose a suitable analytical procedure which can address critical issues in structural

performance. The objective is to predict response of the reinforced concrete column

subjected to lateral loading accurately while keeping overall computational process

simple and easy to implement.

In the research reported here, two of the available models will be examined and

analyzed for their applicability in terms of the accuracy of predicted responses and ease

of applications. Both models will be implemented by writing Matlab (MathWorks, 2009)

computer programs and estimated responses for previously tested reinforced concrete

columns will be compared with test data. Based on the observations, an analytical

procedure will be proposed that can accurately predict monotonic lateral load-

displacement response of the reinforced concrete column subjected to lateral loading. The

predicted responses by the proposed procedure will be compared with experimental test

data to evaluate its performance.

5
One of the models analyzed in this study, designated as displacement component

model, was developed by Setzler and Sezen (2005). The macro model was developed

based on the concept that a typical fixed-ended reinforced concrete column, when

subjected to earthquake loading, undergoes lateral deformation which is comprised of the

three components; flexural deformations, reinforcement slip deformations and shear

deformations as illustrated in figure 2.1. In this model, each of three individual

deformation components is estimated separately and then simply added together to get

total pre-peak response. For post-peak analysis, the understudy column is classified into

one of the five categories based on comparison of its predicted shear and flexural strength

and then individual deformation components are combined together according to a set of

rules specified for each category.

The other model evaluated in this study is Axial-flexure-shear interaction (ASFI)

approach by Mostafaei and Kabeyasawa (2007). This model considers interaction

between axial, flexure and shear mechanisms in terms of concrete compression softening

and axial deformation while satisfying conditions on compatibility of average strains and

equilibrium of average stresses. For shear behavior, this model employs Modified

Compression Field theory (MCFT) which is considered very suitable approach for

determining shear deformations of the reinforced concrete elements subjected to shear

and normal stresses.

1.4 Organization

This thesis is organized into seven chapters. The organization is based on the

investigation process implemented in this study. Chapter 2 presents the summary of the

6
displacement component model. The analytical procedure is implemented and predicted

responses for test specimens used in the original study are compared with test data to

verify correct implementation and working of the computer program. Chapter 3

summarizes the theory of axial-flexure-shear interaction approach. As MCFT employed

in this model is a complex approach, analytical procedure and relevant details are

explained in detail. This chapter can serve as a good summary documentation on theory

and implementation of MCFT. In Chapter 4, the complex analytical procedures of ASFI

approach are implemented and predicted responses for the test specimens used in the

original study are presented and compared with reported response to verify correct

implementation. Both understudy approaches are analyzed by comparing their estimated

response for a common set of test specimens to evaluate their capabilities in terms of

predicting peak strength, post peak response and ultimate deformations in Chapter 5. The

observations made in this chapter form the basis of the proposed analytical procedure. In

Chapter 6, the proposed procedure for response estimation is presented. This chapter

explains the procedure to determine flexural and shear deformations considering

interaction between these mechanisms. Two of the available bar buckling models are

analyzed and recommended procedure is presented to incorporate bar buckling

phenomenon in the proposed procedure. Predicted responses by the proposed procedure

are compared with experimental data to verify the model.

A summary of the research completed is included in Chapter 7. This chapter also

discusses conclusions drawn from this study, and suggests several areas of future

research.

7
Figure 1.1: Olive View Hospital damaged in the 1971 San Fernando Earthquake
(Steinbrugge, K. V., NISEE)

Figure 1.2: Failure of first story columns in the Olive View Hospital
(Steinbrugge, K. V., NISEE)
8
Figure 1.3: Imperial County Services Building damaged during the 1979 Imperial Valley
earthquake (Bertero, V. V., NISEE)

(a) Margallah Towers Islamabad

(b) A low rise commercial building in Muzafarabad

Figure 1.4: Collapsed reinforced concrete buildings during 2005 Pakistan earthquake
(Nawed, K., eNKay Solutions)

9
(a) Insufficient and improper lap splicing (b) Widely spaced transverse reinforcement

(c) Poor quality of the concrete (d) Weak column strong beam construction

(e) Shear failure of the column (f) Plastic hinge formation near column end

Figure 1.5: Examples of commonly observed deficiencies and damage in reinforced


concrete construction after 2005 Pakistan earthquake
(EEC, Department of Civil Engineering NWFP UET Peshawar)
10
CHAPTER 2

DISPLACEMENT COMPONENT MODEL

2.1 Introduction

In order to evaluate structural performance of reinforced concrete columns

subjected to lateral loads, Setzler and Sezen (2008) developed a macro model based on

the concept that a typical fixed-ended reinforced concrete column, when subjected to

earthquake loading, undergoes lateral deformation which is comprised of three

components; flexural deformations, reinforcement slip deformations and shear

deformations as illustrated in figure 2.1. In this model, each of three individual

deformation components are estimated separately and then simply added together to get

total pre-peak response. For post-peak analysis, the understudy column is classified into

one of the five categories based on comparison of its predicted shear and flexural strength

and then individual deformation components are combined together according to a set of

rules specified for each category.

In this chapter, the analytical procedure of the above mentioned displacement

component model is briefly summarized. First, deformation component models for

estimating each component of lateral deformation are presented followed by the details of

total response model that outlines the rules for total response. The approach is

11
implemented in a computer program written in MATLAB (MathWorks, 2009) and the

results for response estimation are presented to verify correct implementation of the

model. The simulated lateral load-deformation responses of the test specimens shall be

used for discussion and comparison in subsequent chapters.

2.2 Deformation Components

As mentioned above, total lateral deformation of a reinforced concrete column

can be determined after estimating lateral deformations resulting from flexural, bond slip

and shear mechanisms. In the procedure discussed in this chapter, these deformations are

calculated separately as per the details given below and then added together depending

upon dominant failure mode determined after comparing flexural and shear strength of

the columns.

2.2.1 Flexural deformations

For reinforced concrete elements subjected to bending moment and axial load,

such as beams or columns, flexural deformations can accurately be determined by

performing section analysis on a fiber model in one-dimensional stress field. This is very

handy and reliable approach if realistic material constitutive relationships and actual

stress distribution across the depth of the cross-section are considered. In this approach,

the reinforced concrete cross-section is discritized into finite number of concrete and steel

fibers. Each of the fibers is idealized as a uniaxial element with its unique stress-strain

relationship. Bernoullis principle that plane section before bending remains plane after

bending is the main hypothesis in the analysis and implies that the longitudinal strain in

12
concrete and steel at any point in the cross section is proportional to its distance from

neutral axis resulting in linear strain distribution. Based upon the resulting strain profile,

stress distribution for concrete and reinforcing steel can be determined in accordance with

their respective stress-strain relationships. By satisfying equilibrium equations at the

cross section, the moment capacity of the section is determined. The process is repeated

number of times by incrementing longitudinal stain until either the concrete or steel fails

as per defined failure criterion. The analysis results in moment-curvature relationship,

which is used to determine flexural deformations.

Typically, fiber section analysis takes into account the enhancement in the

strength and ductility of the concrete due to confinement and ignores concrete behavior in

tension. Also, confined core concrete and unconfined cover concrete are configured

separately with their respective stress-strain relationships. It must also be noted that the

accuracy of the analysis is directly related to the ability of the constitutive material

models to simulate real material behaviors in the cross section and the level of

simplifying assumptions during the analysis.

Knowing the moment-curvature relationship for the reinforced concrete section,

lateral force-flexural deformation relationship can be determined either by force-based or

deformation-based approach. Force-based approach integrates curvature over height of

the column to determine lateral deformation due to flexure as

L
f ( x) xdx (2.1)
0

where x is the curvature distribution at distance x measured along column axis, and

L is the height of the column. This approach results into an accurate response as it
13
captures the non-linearity in actual curvature distribution up to the point corresponding to

the maximum moment capacity. However, after the peak has reached, the deformations

will continue to increase and this approach can no longer predict post-peak deflections, as

no direct relation exists between moment and curvature in this range. In the plastic hinge

method for computing flexural deformations, as shown in figure 2.2, a linear curvature

distribution is assumed in the elastic range and inelastic curvatures are lumped at the

column end over the plastic hinge length L p . Flexural displacements are calculated by

taking the moment of the areas of the curvature diagram about the column end. Unlike

force-based approach, plastic hinge model can predict post-peak flexural deformations

but overestimates the deflections in elastic range as it approximates the linear moment-

curvature relationship in this range.

In view of the discrepancies of both of these methods, a hybrid approach is

adopted for determining flexural deformations employing deflection calculations by

curvature integration up to yield point and plastic hinge method after yielding. The yield

point for the section is defined when the longitudinal steel first yields. Following

equation is used to calculate lateral deformations of the column due to flexure.

Lp
f f , y y L p a for y u (2.2)
2

where f , y is the flexural deflection at yielding calculated using the integration method,

is curvature at column end, y is curvature at yield point, a is shear span equal to L

for a cantilever column and L 2 for fixed ended column. The plastic hinge length L p is

14
taken 0.5h per the recommendations of Moehle (1992), whereas h is depth of cross-

section.

From the moments calculated as per the above-mentioned procedure, lateral load

V corresponding to any moment capacity M can be calculated by dividing moment by

shear span a . The maximum lateral force sustainable by the column is

Mp
Vp (2.3)
a

where, M p is maximum moment capacity from flexural section analysis.

2.2.2 Reinforcement Slip deformations

When a reinforcing bar embedded in the concrete is subjected to a tensile force,

strain accumulates in the embedded length of the bar which causes the bar to extend or

slip relative to the concrete in which it is embedded. The same phenomenon is observed

when a reinforced concrete column is subjected to bending moment. The concrete on

tension face of the cross-section, being weak in tension, cracks at the early stage of

loading and becomes ineffective in anchoring the column to the base. Resultantly, the

reinforcing bars carry tensile loads from column to the anchoring concrete. The bond

stresses at concrete-steel interface in the anchoring concrete cause a tensile stress gradient

in the bars over their embedded length. The steel stresses vary from zero at the dead end

of the embedded bar to a maximum value at the face of beam-column or beam-footing

joint. The length of the bar over which these stresses are distributed, and eventually

transferred to concrete by the bond stresses, is called the development length. The

accumulation of the strain over the development length causes the extension of the bars

15
relative to the anchoring concrete. This extension is commonly known as reinforcement

slip and it leads to rigid-body rotation of the column, as shown in figure 2.1. This results

in lateral displacement that can be as large as 25 to 40 % of the total lateral deformations

(Sezen, 2002).

The flexural deformations as determined through conventional fiber section

analysis (moment-curvature analysis) do not account for lateral deformations caused by

reinforcement slip. Reinforcement slip deformations add another component of lateral

deformation to overall response. Therefore, reinforcement slip deformations must be

calculated separately and added to the other deformation components, such as flexure and

shear to accurately model the total drift.

In this study, lateral displacements due to reinforcement slip are calculated

through a model which was proposed Sezen and Moehle (2003) and further developed by

Sezen and Setzler (2008). The model, as shown in figure 2.3, approximates the bond

stress as bi-uniform function with different values for elastic and inelastic steel behavior,

which allows for the efficient computation of the reinforcement slip and eliminates the

need for the nested iteration loops that are required in some of the existing bond stress-

slip models. The value for the bond stress in the elastic range is taken as ub 1 f c

(MPa) based on a study by Sezen (2002) on 12 test columns. For the inelastic range, the

value for bond stress is adopted from the study by Lehman and Moehle (2000) as

ub 0.5 f c (MPa) where f c is concrete compressive strength. Slip at the loaded end of

the reinforcing bar can be calculated by integrating bi-linear strain distribution over the

development length as follows:

16
ld ld

slip ( x)dx
0
(2.4)

where ld and ld' are development lengths for the elastic and inelastic portion of the bar,

respectively. As bond stresses are uniform in each range, strain distribution is bi-linear as

shown in figure 2.3. Integration of equation 2.3 yields following relations for extension or

slip of the reinforcing bars as

s ld S y
slip
2
for (2.5)
y ld ( s y )l d
slip S y
2 2

where s is the strain at loaded end of the bar, and y is steel yield strain. The

development lengths for elastic and inelastic regions can be calculated as

f s db
ld
4ub
(2.6)
( f s f y )db
ld
4ub

where f s is stress at loaded end of the bar, f y is steel yield stress, and d b is diameter of

the longitudinal bar. The reinforcement slip is assumed to occur in tension bars only and

cause the rotation about the neutral axis. Hence, rotation caused due to reinforcement slip

can be calculated from the following equation figure 2.4.

slip
s (2.7)
d c

where d and c are the distances from the extreme compression fiber to the centroid of

the tension steel and the neutral axis, respectively. The lateral displacement due to slip at

17
the free end of a cantilever column can be calculated as the product of slip rotation s

and length of the column L as

s s L (2.8)

2.2.3 Shear Deformations

Shear deformations in reinforced concrete members are have traditionally been

ignored in design and research due to their lack of complete understanding and being

difficult to measure in an experimental set up or a real structure. For a well-designed

reinforced concrete column, shear deformations are small as compared to the flexural

deformations and are often less than 10 percent of total deformations. Contrary, for a

reinforced concrete column not designed according to stricter seismic design provisions,

shear behavior could be the governing failure criterion. Shear deformations in such shear

critical reinforced concrete column could contribute large percentage towards total

deformations and hence cannot be ignored if an accurate analysis of deformation capacity

is desired.

The basis of the shear model used in this analytical approach is the model

developed by Patwardhan (2005), which uses Modified Compression Field Theory

commonly known as MCFT (Vecchio and Collins, 1986). Patwardhan proposed a

piecewise linear model defining key points in the lateral force-shear deformation

envelope through a parametric study implementing MCFT through a computer program

Response-2000 (Bentz, 2000a). In shear model employed in this approach, as shown in

figure 2.5, pre-peak non-linear shear force-shear deformation response is obtained

18
indirectly from Response-2000 by integrating shear strain distribution over the height of

the column for each load step as following:-

L
v ( x)dx (2.9)
0

where x is the average shear strain over the cross-section at each location x along

the height L of the column, and v is the shear displacement. After the peak strength has

reached, the shear strength is assumed to remain constant at its peak value until the onset

of shear strength degradation. By modifying the equation proposed by Gerin and Adebar

(2004), the shear displacement at the onset of shear degradation v ,u can be calculated as

v
v ,u 4 12 n v ,n (2.10)
f c

where, vn is the shear stress at peak strength V peak bd , f c is the concrete

compressive strength, and v ,n is the shear displacement at maximum strength

determined from Response-2000. The peak strength Vpeak is the minimum of the shear

strength of the column Vn and shear force corresponding to the maximum moment

sustainable by the section Vp . After the shear degradation is initiated, shear strength

decreases linearly with increasing shear deformations to the point of axial load failure,

where lateral strength is assumed zero. The shear displacement at axial load failure v, f

is calculated as

v, f ALF f , f s, f v,u (2.11)

19
where ALF is the total displacement at axial load failure, f , f and s , f are the flexural

and slip displacement at the point of axial load failure, respectively. The total

displacement at axial load failure is determined by the expression proposed by Elwood

and Moehle (2005a), which is based on a shear friction model and an idealized shear

failure plane as:-

ALF 4 1 tan 2
(2.12)
L 100 s
tan P
A f d tan
sv yv c

where is the angle of the shear crack, P is the axial load, Asv is the area of transverse

steel with yield strength f yv at spacing s and d c is the depth of the core concrete,

measured to the centerlines of the transverse reinforcement. In the derivation, is

assumed to be 65 degrees. The values of f , f and s , f in equation 2.11 are determined

according to the expected failure mode and classification of the column into categories as

explained in subsequent subsection.

2.3 Total Response

In order to model the response of a structure subjected to lateral loading, all of the

three deformation components must be combined to predict total lateral deformation with

due regard to their interconnectedness. The procedure for total response models each of

flexure, slip and shear deformation by a spring subjected to the same force and the total

response is the sum of the responses of each spring as shown in figure 2.6. In this model,

each of the deformation components is simply added to obtain the total response up to the

peak strength of the column, defined as lesser of the strength from flexural and shear
20
model. However, for post-peak behavior, the column is classified into one of the five

categories based on a comparison of its shear, yield and flexural strength and rules are

specified for the combination of the deformation components for each category. Yield

strength Vy is defined as the lateral load corresponding to the first yielding of the tension

bars in the column whereas flexural strength Vp is the lateral load corresponding to the

peak moment sustainable by the column during flexural analysis. The shear strength of

the column Vn is calculated by the expression developed by Sezen and Moehle (2004) for

lightly reinforced concrete columns as

6 f
Vn k (Vc Vs ) k 0.80 A Asv f yv d
P
c
1 (2.13)
a 6 f c Ag s
g

where Vc is the concrete contribution to shear strength, Vs is the steel contribution to

shear strength, Ag is gross cross-sectional area and k is a factor related to the

displacement ductility which is the ratio of the maximum displacement to the yield

displacement.

Classification of the column based on comparison of shear, yield and flexural

strength determines expected column behavior. The peak response will be limited by the

lesser of the shear strength from shear model and the flexural strength ( Vp ), however post

peak response is assumed to be governed by the limiting mechanism (i.e., flexure or

shear). The classification system and rules governing the post peak response in each

category are described below and are illustrated in figure 2.7 and 2.8, respectively.

21
Category I. Vn V y

The shear strength is less than the yield strength and column fails in shear

while the flexural behavior remains elastic. The deformation at peak strength (i.e.,

shear strength) is the sum of deformations in each spring at the peak strength.

After the peak strength is reached, the shear behavior dominates the response. As

the shear strength degrades, the flexure and slip springs unload along their initial

responses. The post-peak deformation at any lateral load level is the sum of the

post-peak shear deformation and the pre-peak flexural and slip deformations

corresponding to that load.

Category II. V y Vn 0.95V p

The shear strength is less than flexural strength and column fails in shear,

however inelastic flexural deformation occurring prior to shear failure affects the

post-peak behavior. The deformation at peak strength is the sum of the

deformations in each spring at the peak strength. Shear deformations continue to

increase after the peak shear strength is reached, but the flexure and shear springs

are locked at their peak strength values. The post-peak deformation is the sum of

flexural and shear deformations at peak strength and post-peak shear deformation.

Category III. 0.95V p Vn 1.05V p

The shear and flexural strengths are nearly identical. It is not possible to

predict conclusively which mechanism will govern the peak response. Shear and

flexural failure are assumed to occur simultaneously, and both mechanisms

contribute to the post-peak behavior. The post-peak deformation at any lateral

22
load level is the sum of the post-peak flexure, slip, and shear deformations

corresponding to that load.

Category IV. 1.05V p Vn 1.4V p

The shear strength is greater than the flexural strength and the column may

potentially fail in the flexure, however inelastic shear deformations affect the

post-peak behavior and shear failure may occur as the displacements increase.

The deformation at peak strength is the sum of the deformations in each spring at

the peak strength. After the peak strength is reached, flexural and slip

deformations continue to increase according to their models, but the shear spring

is locked at its value at peak strength. The post-peak deformation at any lateral

load level is the sum of the post peak flexural and slip deformations

corresponding to that load and the shear deformation at peak strength.

Category V. Vn 1.4V p

The shear strength is much greater than the flexural strength and column

fails in flexure while shear behavior remains elastic. The peak strength of the

column is the flexural strength calculated from the flexure model. If the column

strength degrades, flexural and slip deformations continue to increase according

to their models, while the shear spring unloads with an unloading stiffness equal

to its initial stiffness. The post-peak deformation at any lateral load level is the

sum of the post-peak flexural and slip deformations and the pre-peak shear

deformation corresponding to that load.

23
For category-I columns, f , f and s , f values to be used in equation 2.11 are

assumed zero. For the category-II columns, shear strength is lesser than flexural strength

and these values are taken as the flexural and slip deformations at the load equal to the

shear strength of the columns. For categories III, IV, and V specimens, f , f and s , f are

the maximum calculated flexural and slip deformations.

2.4 Shear Failure Due to High Displacement Ductility

Shear strength degrades as the displacement ductility increases (Sezen and

Moehle, 2004), which can cause shear failure in the columns that are initially dominated

by flexure. Such behavior, as encountered in category-IV columns where shear failure

can occur at higher displacements, can not be captured by simply combining three

deformation component models as described above, thus necessitating an alternate

approach to predict delayed shear failure. Elwood (2004) proposed that shear failure,

after occurrence of flexural yielding, will occur if the lateral response intersects the shear

failure surface imposed on the lateral load-total displacement behavior of the column.

The shear failure surface is defined by the empirical drift capacity model proposed by

Elwood and Moehle (2005b) as

SF 3 1 v 1 P 1
4 v (2.14)
L 100 500 f co 40 Ag f co 100

where SF is the drift at shear failure, v is the transverse reinforcement ratio, and v is

the nominal shear stress. f co and v have units of psi. The displacement at shear failure is

calculated according to above equation using the peak strength (Vp) in the model to

24
calculate the shear stress. If the total lateral response envelope exceeds the calculated

drift at shear failure, shear failure is assumed to have occurred. The model is modified to

degrade linearly from the point of shear failure to strength of zero at the displacement at

axial load failure.

2.5 Implementation of the Displacement Component Model

In order to implement above mentioned displacement component model, a

computer program is written in MATLAB and lateral load-deformation responses for

flexural, bar slip, shear and total response are calculated for four test columns tested by

Sezen (2002). These are the same test specimens, which were used by Setzler and Sezen

(2008) to develop their model and verify its response predictions. Details of the test

specimens, material constitutive relationships used for response estimation and response

predictions are presented in following sub-sections.

2.5.1 Test Specimens and Material Properties

Sezen (2002) tested four fixed ended reinforced concrete columns

subjected to unidirectional monotonic and cyclic lateral loading. The test specimens,

designated specimens-1 through -4, were designed with the details common in the

buildings that lack current seismic code requirements and represented lightly reinforced

columns that have shear and flexural strengths very close to each other. These tests

rendered very useful data in terms of experimental force-displacement responses for each

25
of the flexure, slip, and shear components individually, as well as for the overall

response.

These are 18 inch square columns with fixed ends at top and bottom having height

of 116 inches. The longitudinal reinforcement consists of eight No. 9 bars with clear

cover of 2 inches. No. 3 column ties with 90-degree end hooks were spaced at 12 in. over

the height of the column. Specimens-1 and -4 were subjected to a constant 150 kip axial

load, Specimen-2 was tested under a constant 600 kip axial load, and Specimen-3 had an

axial load varying from 600 kip in compression to 60 kip in tension to simulate the range

of forces experienced by an exterior building column in an earthquake. The columns were

tested under unidirectional cyclic lateral loading, except for Specimen-4, which was

tested under monotonic loading. Average concrete strength was 3.077 psi with maximum

aggregate size of 1 inch. Yield strength of longitudinal and lateral steel was 59 and 69

ksi, respectively.

2.5.2 Material Constitutive Relationships

Same material constitutive laws for concrete and reinforcing steel are employed in

the analysis which were used by Setzler and Sezen (2008) for their proposed model. In

flexural section analysis, separate stress-strain relationships are employed for confined

core and unconfined cover concrete. For core concrete, constitutive laws in compression

are defined considering the effects of confinement on core concrete as per the

confinement model by Mander et al. (1988). However, in order to represent the expected

post-peak concrete behavior in shear critical columns, the descending branch is modeled

by the relation developed by Roy and Sozen (1964).


26
The pre-peak confined concrete compressive stress-strain relationship is defined

with the help of following equations.

f cc r c cc
fc
r 1 c cc
r

f cc
cc co 1 5 1 for c cc (2.15)
f co
Ec
r
Ec Esec

where f cc is the peak confined concrete strength calculated according to Mander et al.

(1988), c is the concrete strain, cc is the concrete strain at peak stress for confined

concrete, co is the concrete strain at peak stress in unconfined concrete (taken here as

0.002), f co is the concrete compressive cylinder strength, Ec is the modulus of elasticity

of the concrete and is equal to 57,000 fco (in psi units) for normal weight concrete, and

Esec fcc cc is the secant modulus of the concrete.

The post-peak confined concrete, compressive response is calculated from the

slope of the unloading branch and peak strength point ( cc , f cc ). The slope of the

descending leg is calculated from Roy and Sozen 50% strength point having strain as,

3 0.002 f co
50u (2.16)
f co 1000

For unconfined cover concrete, same set of above mentioned equations is used

except that f cc and cc ,wherever they appear, are replaced with f co and co . Stress-

strain relationships for confined and unconfined concrete are illustrated in figure 2.9.

27
The reinforcing steel behavior is modeled considering a linear elastic behavior, a

yield plateau, and a non-linear strain-hardening region, as per following set of equations.

f s Es s s y
f s f y s sh Es for y s sh (2.17)
p

f s f u ( f u f sh ) s sh sh s u
u sh

where Es is the elastic modulus of steel, s is the steel strain, and the subscripts y, sh, and

u refer to the yield point, the onset of strain hardening, and the ultimate stress,

respectively. p is the order of the curve that defines the strain-hardening region, and is

often taken as 2 for a parabolic curve. is a coefficient that defines the slope of the yield

plateau. For the columns tested by Sezen, these parameters were used to define the

longitudinal steel stress-strain model: f y 59 ksi, sh 0.016 , 0.02 , f u 93.5 ksi,

u 0.23 , p 6 , E s 29,000 ksi. Steel stress-strain relationship used for this model is

illustrated in figure 2.10.

2.5.3 Comparison of Model and Test Results

Component and total responses for specimens-1 through -4, obtained after

implementing Setzler and Sezen (2008) approach, are presented in figure 2.11 to 2.14.

Predicted responses for flexural and slip deformations compare well with the

experimental data. Predicted shear responses follow the initial stiffness of the

experimental responses closely but overestimates post-peak deformations as the shear

strength degrades. For total responses also, generally good correlation with the

experimental data is obtained. It can be concluded that the model overall does a good job
28
in predicting lateral behavior of the columns. These results will be used subsequently for

discussion and comparison of the models applicability and suggesting improvements to

the existing procedure.

2.6 Conclusion

Displacement component model by Setzler and Sezen (2008) has briefly been

summarized in this chapter. This approach considers total deformation of a reinforced

concrete column to lateral loading being composed of flexural, bar slip and shear

deformations which can be combined together depending upon dominant failure

mechanism. Structural response of four test specimen is calculated as per the prescribed

model after implementing analytical routine in MATLAB computer program.

Comparison of predicted response and experimental data shows that the understudy

model has correctly been implemented and model performs well in evaluating lateral

load-lateral deformation responses of reinforced concrete columns.

29
Flexure deformation Slip deformation Shear deformation

Figure 2.1: Components of lateral deformation in a reinforced concrete column

y
x x

Lp L - Lp L
y y

Figure 2.2: Plastic Hinge Model for Determining Flexural Deformation

30
F As f s

Reinforcing bar

Beam-column
interface
b' 0.5 fco fs s ld'

fy y
ld
b 1 fco

b fs s

Figure 2.3: Bar Slip Model

Figure 2.4: Reinforcement slip rotation

31
Maximum
Maximum Beginning
Initiation of of shear
shear
strength point
degradation
( v ,n,V
, V peak )
strength point
( degradation
,V )
v,n peak
v,u peak
v,u peak
, V
load
Lateralload
lateral

Response-2000
Response 2000

Axialload
Axial loadfailure
failure
,0
(
,0) v,f
v, f
shear displacement
Shear displacement

Figure 2.5: Shear displacement model

V
Zero-length slip
rotation spring

Zero-length
shear spring

Flexural spring

L
s

Zero-length slip
rotation spring

V
s f v

Figure 2.6: Spring representation for total response model for fixed ended column
32
Cat. I
V Vy (yield strength)
Vp (flexural strength)

Cat. II

Equation 2.13
Eq. 7 (k=1.0)
Vn
Cat. III Cat. IV
Equation
Eq. 7 2.14
0.7*Vn

Cat. V

0
0 2 6
displacement ductility ()

Figure 2.7: Classification of the column and Lateral load displacement relationship

Category
Category V V Category III, IV, V
Category
Category
Category I,I,II,II,
III III
Category
Category I I III, IV, V
Lateral force
lateral force (V)

lateral force (V)

Category
Category IV IV Category II
Category II

shear displacement ( ) flexural or slip displacement ( f, s )


Shear deformations
v Flexural or slip deformations

Figure 2.8: Rules governing the combination of deformation components

33
3500
Confined concrete

3000 Unconfined concrete

2500
Stress (psi)

2000

1500

1000

500

0
0 1 2 3 4 5 6 7 8
Strain x 10
-3

Figure 2.9: Constitutive relationships for concrete in compression

100

90

80

70

60
Stress (ksi)

50

40

30

20

10

0
0 0.025 0.05 0.075 0.1 0.125 0.15
Strain

Figure 2.10: Constitutive relationship for reinforcing steel


34
80
Specimen - 1
60

Lateral force (kip)


40

20

-20

-40

-60 Test data


Model
-80
-2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5

100
Specimen - 2
Lateral force (kip)

50

-50
Test data
Model
-100
-0.9 -0.7 -0.5 -0.3 -0.1 0.1 0.3 0.5 0.7 0.9 1.1
1.12

80
Specimen - 3
60
Lateral force (kip)

40

20

-20

-40 Test data


Model
-60
-3 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2

80
Specimen - 4
60
Lateral force (kip)

40

20

-20

-40

-60 Test data


Model
-80
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.5 3 3.5 4
Flexural Displacement (in)

Figure 2.11: Lateral load flexural deformation relationships


35
80
Specimen - 1
60

Lateral force (kip)


40

20

-20

-40

-60 Test data


Model
-80
-1 -0.75 -0.5 -0.25 0 0.25 0.5 0.75 1 1.25 1.5

100
Specimen - 2
Lateral force (kip)

50

-50
Test data
Model
-100
-0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8

80
Specimen - 3
60
Lateral force (kip)

40

20

-20

-40 Test data


Model
-60
-2.5 -2.25 -2 -1.75 -1.5 -1.25 -1 -0.75 -0.5 -0.25 0 0.25 0.5 0.75 1

80
Specimen - 4
60
Lateral force (kip)

40

20

-20

-40

-60 Test data


Model
-80
-1 0 1 2 3 4
Slip Displacement (in)

Figure 2.12: Lateral load slip deformation relationships


36
80
Specimen - 1
60

Lateral force (kip)


40

20

-20

-40

-60 Test data


Model
-80
-5 -4 -3 -2 -1 0 1 2 3 4

100
Specimen - 2
Lateral force (kip)

50

-50
Test data
Model
-100
-1 -0.75 -0.5 -0.25 0 0.25 0.5 0.75 1

80
Specimen - 3
60
Lateral force (kip)

40

20

-20

-40 Test data


Model
-60
-2 -1.5 -1 -0.5 0 0.5 1 1.5

80
Specimen - 4
60
Lateral force (kip)

40

20

-20

-40

-60 Test data


Model
-80
-4 -3 -2 -1 0 1 2 3 4
Shear displacement (in)

Figure 2.13: Lateral load shear deformation relationships


37
80
Specimen - 1
60

Lateral force (kip)


40

20

-20

-40

-60 Test data


Model
-80
-6 -4 -2 0 2 4 6

100
Specimen - 2
Lateral force (kip)

50

-50
Test data
Model
-100
-3 -2 -1 0 1 2 3

80
Specimen - 3
60
Lateral force (kip)

40

20

-20

-40 Test data


Model
-60
-8 -6 -4 -2 0 2 4

80
Specimen - 4
60
Lateral force (kip)

40

20

-20

-40

-60 Test data


Model
-80
-6 -4 -2 0 2 4 6 8
Displacement (in.)

Figure 2.14: Lateral load total deformation relationships


38
CHAPTER 3

AXIAL SHEAR FLEXURAL INTERACTION (ASFI) APPROACH - THEORY

3.1 Introduction

Displacement-based evaluation approach for estimating structural performance of

reinforced concrete elements requires monitoring of their load-deformation responses in

all failure mechanisms associated with axial, flexural and shear loadings. Traditionally,

the analytical process for response estimation is simplified by splitting the total structural

behavior into individual mechanisms of flexure and shear. The analyses for both

mechanisms are carried out individually and dominant failure mechanism is determined

which controls overall response and design considerations.

Flexural deformations are most commonly estimated by performing flexural

section analysis. This is very convenient and accurate analytical tool for flexural

response, but performance of reinforced concrete elements dominated in shear or shear-

flexure cannot be estimated, as shear behavior is not considered in the approach.

Likewise, most of the available models and theories for shear response are good for

evaluating shear behavior only and can not estimate flexural performance directly.

It must be recognized that flexural and shear deformations in a reinforced

concrete element subjected to combined axial, flexure and shear loading, do not occur

39
independently. Any model that aims to capture total response must consider the

interaction between flexure and shear mechanisms. When the analyses for these

mechanisms are carried out independently, this aspect can not be taken care of due to the

inherent inability of the analytical procedure to perform good for the mechanism it has

not been developed for.

Mostafaei and Kabeyasawa (2007) have recently developed a displacement-based

analysis approach for response estimation of reinforced concrete elements by considering

interaction between axial, shear and flexural mechanisms. The approach is designated as

Axial-Shear-Flexural Interaction (ASFI) Approach and consists of two models that

evaluate flexural and shear response simultaneously. In this approach, axial-flexural

behavior is simulated by employing conventional section analysis or fiber model in one-

dimensional stress field whereas axial-shear response is determined by considering one

integration point in in-plane stress conditions employing modified compression field

theory (MCFT). MCFT is one of the many available shear response approaches such as

truss models and empirical models, and is considered most suitable for determining shear

deformations of the reinforced concrete elements subjected to shear and normal stresses.

In ASFI approach, axial-flexural and axial-shear mechanisms are coupled in stress-strain

field considering axial deformation interaction, softening of concrete compression

strength and satisfying compatibility and equilibrium conditions.

In this chapter, theory of ASFI approach is presented by explaining details of the

component models and development of the approach in generalized form for a reinforced

concrete element between two subsequent flexural sections.

40
3.2 Background

As already mentioned above, structural performance of reinforced concrete

column can accurately be estimated if flexural and shear deformations are evaluated

simultaneously with due regards to their interconnectedness. It must also be recognized

that flexural section analysis and MCFT are powerful assessment tools for the mechanism

originally developed for, and deal only with individual component of flexural and shear

deformations respectively. For MCFT, this statement is true in general, but efforts have

been made to determine flexural-shear response by MCFT. To do so, a reinforced

concrete element must be discritized into large number of biaxially loaded elements and

analyzed by conducting a non-linear finite element analysis. The concept of MCFT has

also been extended to sectional analysis approach elements loaded in combined axial,

shear and flexure loads by employing biaxially stressed elements as concrete fibers

instead of uniaxially stressed elements employed in conventional sectional analysis

approach (Vecchio and Collins 1988). By fixing the longitudinal strain in each concrete

fiber by corresponding section strain, concrete elements are analyzed individually for in-

plane stress field based on MCFT. The most recent implementation of this approach can

be seen in the computer program called Response-2000 (Bentz 2000). The application of

MCFT in finite element approach or sectional analysis approach in the way mentioned

above yields very reliable response but at the same times results in fastidious

computations and is not simple enough for practice.

Some studies have also been carried out to combine individual components of

shear and flexure by modeling them as springs in series. Shirai et al (2001) presented a

macro-element approach by decomposing total deformation of reinforced concrete

41
column into flexural and shear components and simulated these component deformations

by layered element model and in-plane shear element respectively. Setzler and Sezens

displacement component model (chapter 2) is also based on the same concept that total

deformation of a reinforced concrete column is composed of three individual components

of flexure, reinforcement slip and shear. Their study models these component

deformations as spring in series subjected to the same force and total response is

determined by combing the component responses according to a set of rules developed

based on the comparison of the columns yield, flexural and shear strength. Although,

these models are simple for practical purposes and estimate response reasonably, they do

not consider interaction of axial deformation and effect of degrading compression

strength of concrete fibers due to increasing shear deformations.

3.3 Axial-Flexure Model.

Axial-flexure model in ASFI approach evaluates flexural response of reinforced

concrete element by employing standard flexural section analysis techniques. The section

analysis is performed in usual way except that constitutive relationships of diagonally

cracked concrete are employed to consider the effect of shear deformations on flexural

performance of the element. This is done by softening the response of concrete in

uniaxial compression by a factor which decreases as shear deformations increase. This

factor, known as compression softening factor, is function of principal tensile strain in

concrete and is obtained from in-plane shear analysis of the flexural element, thus

coupling axial-flexure model with axial-shear mechanism. Due to interaction of

compression softening factor, axial-flexural analysis in ASFI approach can not be

42
performed independently. This aspect is described in detail in subsequent sections. The

other deviation from usual practice of flexural section analysis is consideration of

concrete tensile behavior in the analysis, which is generally ignored in standard

procedure.

Sometimes, in order to simplify the calculations, section analysis with one stress

block is performed for simplified analysis and design. This is done by replacing actual

concrete stress distribution across the section by an equivalent rectangular stress block,

commonly known as whitney stress block. This allows the analysis to be performed with

much ease and simplicity and eliminates need of any iterative calculations required for

more prcised fiber section analysis approach explained in previous chapter.

Section analysis with equivalent stress block or fiber model in one-dimensional

stress field can be employed as axial-flexure model in ASFI approach, provided that

concrete tensile behavior and concrete compression softening is considered in the

analysis. Choice of the particular method depends on the required level of accuracy and

available amount of computational effort. In the analysis presented here, fiber section

analysis approach is adopted as axial-flexural model in ASFI approach. Details of fiber

section analysis have already been presented in chapter 2.

3.3.1 Effect of Shear Deformations on Flexural Response

Conventionally, the flexural section analysis is performed independently

simulating concrete behavior in uniaxial compression. The compressive stress-strain

relationships for the concrete in flexural analysis are usually derived from its response in

standard cylinder test. It must be recognized that the strain conditions for the concrete in

43
the web of a reinforced concrete beam or column subjected to shear differ significantly

from those in a cylinder test. The concrete in standard cylinder test is subjected to only

small tensile strains primarily due to Poissons effect, whereas, the concrete in diagonally

cracked web is subjected to very substantial tensile strains primarily due to shear. As a

result, the concrete in diagonally cracked web is weaker and softer than the concrete in a

cylinder. To account for theses influences, axial-flexural model in ASFI approach

employs constitutive model for cracked concrete in compression. The degradation in

compressive strength and stiffness of cracked concrete, denoted as concrete compression

softening, has been studied by Vecchio and Collins (1986) in their efforts to investigate

stress-strain characteristics of diagonally cracked concrete. Their study on 30 reinforced

concrete panels subjected to in-plane stress conditions indicates that the principal

compressive stress in the concrete is not only a function of principal compressive strain

but also of the co-existing principal tensile strain, such that compressive strength and

stiffness of the concrete decrease as the tensile strains increase. The concrete

compression softening, as shown in Figure 3.1, is one of the interaction terms from axial-

shear mechanism that is incorporated into axial-flexural model coupling both

mechanisms. In ASFI approach, concrete compression softening is employed in flexural

analysis by softening concrete response in uniaxial compression in the manner that,

fcsof fc (3.1)

where, fcsof is concrete compressive response after softening, f c is concrete

compressive stress given by any of the appropriate stress-strain relationships simulating

concrete behavior in uniaxial compression, and is compression-softening factor

defined as
44
1
1.0 (3.2)

0.8 0.34 c1
co

where, c1 is principal tensile strain and co is concrete strain at maximum concrete

cylinder strength.

3.3.2 Axial Strain Due to Flexure

Axial strain due to flexure is another interaction term in ASFI approach that

couples axial-flexure mechanism to axial shear mechanism. Complete details on axial

strain interaction methodology shall be presented subsequently; however, as axial strain

due to flexure can only be determined from flexural section analysis, it is appropriate

here to mention the way it can be calculated. Average axial strain due to flexure xf

between two flexural sections can be determined as illustrated in Figure 3.2 based on

relative centroidal deformation between two sections assuming linear strain distribution.

1
l12
x
xf
l12
0
o2 o1
l12
dx o 2 o1
2
(3.3)

where o1 and o 2 are centroidal strains of two consecutive flexural sections and l12 is

distance between the sections.

3.3.3 Flexural Deformations.

Flexure deformations in axial-flexure model are determined with the help of

plastic hinge model. In this method, elastic and inelastic curvatures are modeled

separately, and added together. A linear curvature distribution is assumed in the elastic

45
range, and the inelastic curvature is lumped at the column end over the plastic hinge

length. The conceptual illustration of the plastic hinge model for determining the flexural

deformation of a cantilever reinforced concrete column is presented in Figure 3.3. The

flexural drift ratio f can hence be calculated as following:

Lin
1
f
Lin x ( x)dx
0
(3.4)

where, Lin is the distance from column end section to inflection point, L p is plastic hinge

length taken equal to the section depth, ( x ) is curvature function at distance x

measured from inflection point and y is yield curvature.

3.4 Axial-Shear Model

Axial-shear model simulates shear mechanism in ASFI approach employing

modified compression theory. MCFT is a suitable displacement-based evaluation

approach for response estimation of reinforced concrete membrane elements subjected to

normal and shear stresses. Vecchio and Collins (1986) developed the theory after

modifying previously proposed Compression Field Theory (Mithell and Collins 1974)

which includes compression-softening effects, contribution of tensile stresses in cracked

concrete and local stress conditions at crack. MCFT is essentially a smeared rotating

crack model in which cracked concrete is treated as a new orthotropic material with its

unique stress-strain characteristics. The theory consists of compatibility, equilibrium and

stress-strain relationships formulated in terms of average stresses and strains. The critical

aspect of MCFT is the consideration of local stress-strain conditions at cracks ensuring

46
that the tension in the concrete can be transmitted across the crack and shear stress on the

surface of the crack does not exceed maximum shear provided by the aggregate interlock.

Thus, load deformation response of the members loaded in shear can be estimated by

considering compatibility of average strains for concrete and reinforcement, equilibrium

relationships involving average stresses in concrete and reinforcement and appropriate

stress-strain relationships for reinforcement and diagonally cracked concrete. Complete

details of MCFT and its formulation can be found in number literature where it is

employed. In this section, only the aspects relevant to ASFI approach shall be presented.

Following assumptions form the basis of the theory and are adopted as such for

ASFI approach as well.

1. Only one stress state exists corresponding to each strain state

2. Stresses and strains can be considered in terms of average values when taken over

areas or distances large enough to include several cracks.

3. The concrete and reinforcing bars are perfectly bonded at the boundaries of the

element, which implies that there is no overall slip.

4. Orientations of principal strains and principal stresses coincide with each other.

5. The longitudinal and transverse reinforcing bars are uniformly distributed over the

element.

6. The cracks are uniformly distributed and rotating.

7. The normal and shear stresses are applied uniformly.

8. The constitutive relationships for concrete and reinforcement are independent.

9. Shear stresses in the reinforcement are negligible and hence are not considered in

equilibrium relationships.

47
Consider an orthogonally reinforced concrete membrane element as shown in

Figure 3.4a. The element consists of smeared reinforcement in longitudinal (x) and

transverse (y) directions, with the corresponding reinforcement ratios x and y . The

yield strengths of longitudinal and transverse reinforcement are f yx and f yy , respectively.

The concrete is characterized by a cylinder compressive strength f c' , a strain at peak co

and a tensile cracking stress f cr . The elements edge planes are subjected to uniform

normal stresses f x , f y and shear stress xy . The deformation of the element is assumed

to occur such that the edges remain straight and parallel.

Under the applied loads, an equilibrium condition is attained resulting in unique

strain condition defined by two normal strains x and y and the shear strain xy as

shown in Figure 3.4b. From the Mohrs circle of the average strains (Figure 3.4c and

3.4d), average concrete principal tensile strain c1 , average concrete principal

compressive strain c 2 and orientation of principal strain filed can be determined as:-

x y 1
y xy
2 2
c1 x (3.5)
2 2

x y 1
y xy
2 2
c2 x (3.6)
2 2

1 xy
p tan 1 (37)

2 x y

48
The inclination of the principal strain p as given by Equation 3.8 can either be

orientation of principal tensile strain t or orientation of principal compressive strain

(crack angle) c depending upon the magnitude of normal strains x and y . Hence,

p t for x y
(3.8)
p c for x y

The principal tensile plan and principal compressive plane are 90 o apart, i.e.,

t c 90o or c t 90o . In addition, positive orientation of principal planes is

considered to make counterclockwise angle with positive x-axis. Hence, following these

rules, the relationship between t and c can be summarized as,

If p t 0 c t 90o
If p t 0 c t 90o
(3.9)
If p c 0 t c 90o
If p c 0 t c 90o

3.4.1 MCFT Constitutive Laws.

In addition to relevant compatibility and equilibrium conditions, constitutive

relationships are required to link average stresses to average strains for concrete and

reinforcement in MCFT. Derivation of these laws, especially for the concrete, is one of

the important aspects in development of the theory and simulates cracked concrete

behavior in compression and tension.

49
3.4.1.1 Concrete in Compression.

The constitutive laws for concrete in compression in MCFT have been derived by

Vecchio and Collins (1986) after investigating experimental stress-strain behavior of

cracked concrete. These laws consider reduction in compressive strength and stiffness of

concrete due to shear after softening the uniaxial response by compression softening

factor. The relationship for compressive concrete behavior in MCFT was suggested as,

and is shown in Figure 3.5a:-

2
f c 2 f co 2 c 2 c 2 (3.10)
co co

where is compression softening factor defined as per Equation 3.2, c 2 is concrete

compressive strain and f c 2 is corresponding concrete stress. The term multiplied with

in Equation 3.10 is the Hognested Parabolic relationship for concrete in uniaxial

compression, often used for normal strength concrete. Although, this equation is

specifically used in MCFT, any uniaxial concrete compressive stress-strain relationship

can theoretically be used as part of axial-shear model in ASFI approach; however, it must

incorporate compression-softening effect as defined by the Equation 3.2.

3.4.1.2 Concrete in Tension.

The constitutive relationship for the concrete in tension was also developed by

Vecchio and Collins (1986) based on their reinforced concrete panel tests. The concrete

tensile stress-strain relationship, as illustrated in Figure 3.5b, relates principal concrete

tensile stress f c1 to principal concrete tensile strain c1 .

50
The relationship suggested prior to cracking i.e., 0 c1 cr is linearly elastic

and is given by the expression:-

f c1 Ec c1 (3.10)

where, Ec is modulus of elasticity of the concrete, cr is cracking strain corresponding to

uniaxial cracking strength of concrete f cr , and f co uniaxial compressive strength of

concrete in MPa. These quantities are given by following expressions

Ec 2 fc' co ; cr fcr Ec ; f cr 0.33 f c' (3.11)

After cracking, the concrete tensile stresses continue to exist due to bond

interaction between concrete and reinforcement and decrease as the principal concrete

tensile strains increase. This phenomenon is known as tension stiffening. The relationship

suggested after cracking, i.e., c1 > cr is

f cr
f c1 (3.12)
1 200 c1

For large reinforced concrete elements, this relationship is modified slightly to the

following expression

f cr
f c1 (3.13)
1 500 c1

3.4.1.3 Reinforcement Stress-Strain Relationship

MCFT adopts bi-linear stress-strain relationship for the longitudinal and

transverse reinforcement in both tension and compression that consists of initial

ascending linear-elastic branch followed by a yield plateau. In the development of the

51
theory, it is assumed that the axial stress in the reinforcement depends only on one strain

parameter, the axial strain in the reinforcement. The shear stress resisted by the

reinforcement on the plane normal to the reinforcement is zero. The relationship for

average axial stress f s and average strain s used in MCFT is shown in Figure 3.5c and

is described by the following equations

f sx Es sx f sxyield
(3.14)
f sy Es sy f syyield

The subscripts x and y represent x -direction (longitudinal) and y -direction

(transverse) reinforcement, respectively. E s is modulus of the elasticity of the steel, and

f sxyield and f syyield are yield stress of the reinforcement in x and y directions,

respectively.

Compatibility of the average strains in MCFT requires that strain in reinforcement

and concrete must match with the in plane strains in x and y -directions i.e., sx x

and sy y . Hence, the Equation 3.14 takes the form as:-

f sx Es x f sxyield
(3.15)
f sy Es y f syyield

Although MCFT uses conventional bi-linear stress-strain relationship for the

reinforcement, other reinforcement relations allowing strain hardening can be used as part

of axial-shear model in ASFI.

52
3.4.2 Considerations of Local Cracks Conditions

Consideration of the local stress conditions at the cracks is another very important

aspect in development of the modified compression field theory. These considerations

ensure that the stresses can be transmitted across the cracks. The formulations

considering average stresses and strains do not capture local variations that may occur at

the cracks. For example, tensile stresses in the reinforcement will be higher than the

average at the cracks and lower than the average midway between the cracks. On the

other hand, the concrete tensile stresses will be zero at the crack and higher than the

average midway between the cracks. These local variations are important as the response

of bi-axially loaded element may be governed by the reinforcements ability to transmit

tension across the cracks or sliding shear failure along the cracks. To address these

possibilities, MCFT limits the local stresses at the cracks and the average concrete tensile

stress between the cracks.

Figure 3.6ashows average stresses at a section between the cracks perpendicular

to the principal tensile stress direction and Figure 3.6b shows local stresses at the free

surface of the crack. At the free surface of the crack, the average concrete tensile stresses

reduce to zero. This causes reinforcement stresses to increase locally at the crack in order

to transmit tensile stresses across the crack. Hence, average concrete tensile stresses must

be limited to avoid failure of the reinforcement at the crack. Static equivalency of the

average and local stresses in the direction normal to the crack surface results in the

condition that limits the concrete average tensile stress to the following upper limit to

insure its transmission across the crack.

f c1 x f scrx f sx cos 2 nx y f scry f sy cos 2 ny (3.16)


53
where, x and y are reinforcement ratio in x and y -directions respectively, f scrx and f scry

are local reinforcement stresses at crack, and nx and ny are the angles between the

normal to the crack and reinforcement in x and y directions, respectively. The values

for nx and ny are defined positive counterclockwise, as shown in Figure 3.6b, and may

be determined as

nx t
(3.17)
ny c

In MCFT, yielding of the reinforcement is the upper limit for local reinforcement

stresses at the crack and hence, the average concrete tensile stresses are limited to-

f c1 x f sxyield f sx cos 2 nx y f syyield f sy cos 2 ny (3.18)

The terms in the parenthesis in above equation represent the reserve capacity of

the reinforcements to transmit concrete tensile stresses before they fail. Hence, Equation

3.18 may be modified by replacing local reinforcement stresses with specified failure

criteria if it is other than yielding of the reinforcement.

The other consideration of local stresses at the crack deals with the shear

stresses, which are present locally at the crack surface due to reinforcement crossing the

cracks at skew angles (Figure 3.6b). This consideration limits the local shear stresses at

the crack surface by the shear resistance provided by the aggregate interlock mechanism.

Static equivalency of the average and local stresses in the direction tangential to the crack

determines local shear stresses as follows

vci x f scrx f sx cos nx sin nx y f scry f sy cos ny sin ny (3.19)

54
This local shear stress value at the crack must not exceed the shear strength

provided by the interface between the cement paste and the aggregate particles due to

aggregate interlock mechanism.

vci 0.18vci max (3.20)

f co
vci max (3.21)
0.31 24w /(a 16)

where w is crack width in millimeters, a is maximum aggregate size in millimeters and

f co is concrete compressive strength in MPA. The crack width w in Equation 3.21

should be the average crack width over the crack surface and can be taken as product of

the principal tensile strain c1 and the average spacing of diagonal cracks in the direction

normal to the crack S cr .

w c1Scr (3.22)

where,

1
Scr (3.23)
sin cos

S mx S my

S mx and S my in the above equation are the crack spacing that indicate crack control

characteristics of x and y reinforcement respectively. These quantities depend on bond

properties and layout of the reinforcement and can be estimated from CEB-FIP Code as

follows,

Sx d
Smx 2(cx ) 0.25k1 bx
10 x
(3.24)
S d
Smy 2(c y ) 0.25k1 by
10 y
55
where cx , cy , and Sx are the parameters determined from the reinforcement layout as

shown in Figure 3.7. k1 is taken as 0.4 for the deformed bars and 0.8 for plain bars. d bx

and dby are longitudinal and transverse the bar diameters, respectively. S is center-to-

center spacing of the transverse reinforcement. The crack control characteristics of the

longitudinal and transverse reinforcement can, alternatively, be also be taken as

Smx 1.5 maximum distance from x-bars and Smy 1.5 maximum distance from y-bars .

The term in Equation 3.24 is inclination of the crack with respect to longitudinal axis

as shown in Figure 3.6a. Its value can be determined as c such that 0 90 o .

If Equation 3.20 is not satisfied i.e., the local shear stress on the crack exceeds the

shear resistance provided by the aggregate interlock mechanism, the average concrete

tensile stress f c1 must be reduced as

f c1 0.18vci max tan c (3.25)

3.4.3 Material Stiffness Formulations.

Vecchio (1989) presented a procedure that showed that nonlinear analysis of

reinforced concrete membrane elements can be performed accurately by simply

modifying existing linear elastic finite element routines based on secant stiffness

formulation provided realistic constitutive relations for concrete and reinforcement are

employed. In constructing an individual element stiffness matrix K , a material stiffness

matrix D is required to relate stresses to the strains. The material stiffness matrix for a

linear elastic isotropic material in state of plane stress can be modified in a form that

56
depends on type of stiffness moduli used to reflect nonlinear behavior of reinforced

concrete according to appropriate set of constitutive laws. Using the same concept in

ASFI approach, the material stiffness matrix D for axial-shear element is determined

by first defining stiffness matrices for concrete and reinforcements with respect to their

respective principal material directions using secant moduli. The total stiffness matrix is

then obtained by combining the component stiffness matrices, after appropriate

transformations to take into account the directional dependence of the materials. Hence,

D T Dc T T Dsi T
T T
(3.26)
i

where, D c is the concrete material stiffness matrix evaluated with respect to principal 1

and 2 axes system corresponding to the direction of the principal tensile strain and

principal compressive strain, respectively.

Ec 2 0 0
D c 0 Ec1 0

(3.27)
0 Gc
0

where, Ec1 and Ec 2 are secant moduli for the concrete and relate to the stress-strain

behavior in principal directions. These are determined as:-

f c1 fc 2 Ec1 Ec 2
Ec1 , Ec 2 , Gc (3.28)
c1 c2 Ec1 Ec 2

For each reinforcement component, a reinforcement material stiffness matrix D si is


evaluated as follows,

i Esi 0 0

D si 0 0 0 i x, y (3.29)
0 0 0

57
where, E si is the secant moduli for the reinforcing steel and relate to the stress-strain

behavior in two orthogonal directions x and y . For a particular stress/strain state,

f sx f sy
Esx , Esy (3.30)
x y

The transformation matrix T Equation 3.26 will differ for each component and is given

by following expression

cos 2 sin 2 cos sin



T sin 2 cos 2 cos sin (3.31)
2 cos sin 2 cos sin cos 2 sin 2

where, c for the concrete component, 0 and 90o for longitudinal and transverse

reinforcement, respectively (Figure 3.8).

Having determined the material stiffness matrix D , it can then be used to relate

stresses f to strains

f D D f
1
or (3.32)

where,

fx x

f f y and y (3.33)

xy xy

3.5 Development of Axial-Shear-Flexure Interaction Approach

The main objective of the ASFI approach is to couple axial-flexural and axial-

shear mechanisms satisfying compatibility and equilibrium conditions while considering

58
their interaction in terms of axial deformations and concrete compression strength

softening, as explained in the following sections.

3.5.1 Compatibility Conditions/relationships

Figure 3.2 illustrates an element between two flexural sections of a reinforced

concrete column subjected to axial load, bending moment and shear force. Total axial

deformation in a reinforced concrete column can be considered as sum of axial

deformations due to three mechanisms of axial, flexural and shear. In an axial-flexural

mechanism, total axial strain is combination of axial strains caused by axial mechanism

xaf and axial strain due to flexural mechanism xf . Likewise, in an axial-shear

mechanism, axial strain is combination of axial strain due to axial mechanism xas and

axial strain caused by shear mechanism xs . Hence, total axial strain x of the column

between the two sections can be obtained by extracting xf from axial-flexural model and

adding to axial deformations of axial-shear model.

x xas xs xf (3.34)

Compatibility of axial deformations is satisfied when axial deformations due to axial

mechanisms in axial-shear and axial-flexural elements are equal to the axial strain due to

only applied axial load xa , that is

xa xaf xas (3.35)

Hence, Equation 3.34 becomes

x xa xs xf (3.36)

59
Furthermore, the total lateral drift of a reinforced concrete column or beam

between two flexural sections is taken as summation of shear strain s and the flexural

drift ratio f .

s f (3.37)

3.5.2 Equilibrium Conditions/relationships

Equilibrium conditions in ASFI are satisfied in average stress-strain field. For

equilibrium, axial stress in axial-flexural element xf should be equal to axial stress in

axial-shear element xs . Also, the equilibrium of shear stress in axial-flexural element f

and shear stress in axial-shear element s is satisfied simultaneously throughout the

analysis.

xf xs o
(3.38)
f s

where o is axial stress due to applied axial load and is resultant shear stress. Shear

stress in axial-flexural and axial-shear mechanisms are calculated as,

1 M1 M 2
f
bd f l12
(3.39)
V
s
bd s

where M 1 is larger moment on one of the flexural section, M 2 is smaller moment on

other flexural section, b is width of the cross-section, d f is the flexural depth of the

section which can be taken as d f h before concrete cracks in flexure and

60
d f d afterwards. h is overall depth of the section, d is effective depth of the section,

V is applied lateral load, d s is shear depth of the section and can be taken as d s h until

concrete tensile crack due to flexure and then d f d .

Normal stress o due to applied axial load P in both axial-flexural and axial-shear

models is calculated as

P P
o (3.40)
Ai bh
where Ai equals the cross-sectional area of fiber i . Normal stress in the direction

perpendicular to the longitudinal axis of the column, or the clamping stresses

y and z are neglected due to inexistence of lateral external forces along theses

directions.

y z 0 (3.41)

3.5.3 Constitutive Laws

In order to satisfy above mentioned compatibility and equilibrium conditions,

same material constitutive relationships must be employed in both axial-flexural and

axial-shear models. In an axial-flexure model, if constitutive laws of confined core

concrete and unconfined cover concrete are defined differently, constitutive law for the

concrete in axial-shear model must also be modified. Equivalent peak concrete

compression stress fc'as and corresponding strain c' as can approximately be calculated as

f c'as
KAcon Auncon f , c' as
KAcon Auncon (3.42)
co co
A A

61
where A is gross cross-sectional area, Acon is area of confined core, Auncon is area of

unconfined cover and K is confinement parameter is taken as ratio of peak confined

concrete strength and concrete cylinder strength. This modification is not required if

confinement effect is applied only on post peak concrete response. In view of the

simplicity and applicability of secant stiffness method in MCFT, it is applied in axial-

shear model of ASFI approach.

3.5.4 Interactions considered

Axial deformation and concrete compression softening are two main interaction

terms used in ASFI approach interconnecting axial-flexural and axial-shear mechanisms.

3.5.4.1 Axial Strain Interaction Methodology

Axial deformations play a very important role in ASFI approach by

interconnecting axial-flexure and shear mechanisms. Axial deformations due to flexure

increases shear crack width and principal tensile strain in the web resulting in lower shear

capacity for the element. As already mentioned, total axial strain in an element between

two sections is sum of axial strain caused by axial, shear and flexural mechanisms

i.e., x xa xs xf . Axial-shear model gives axial strain caused by axial and shear

mechanism xa xs . Therefore, in order to obtain total axial strain x , axial strain due to

flexure xf from axial-flexure model (determined from Equation 3.3) must be added to the

axial strains of axial-shear model. This can be done by simply adding flexibility

component of axial deformation due to flexure to the corresponding flexibility

62
component of axial-shear model. The flexibility component for axial deformation due to

flexure f xf can be determined as,

xf
f xf (3.43)
x

where, x is applied axial stress in longitudinal direction of the column. In case of the

beams or columns where axial load is zero, in order to avoid having an indefinite value in

Equation 3.43, a very small value must be considered for axial stress.

A stress-strain relationship in terms of flexibility matrix for an in plane shear

element (axial-shear model) can be defined as,

f11 f12 f13 x x xas xs



f 21 f 22 f 23 y y (3.44)
f
f33 xy
31 f32
xy

where, fij i, j 1, 2,3 are flexibility components of in plane shear model,

x xas xs is axial strain in axial-shear element and y is strain in transverse

reinforcement. Axial strain due to flexure xf can be taken into account to axial-shear

model by adding flexibility component obtained from Equation 3.43 into Equation 3.44.

Considering x axis as the main longitudinal axis of the column, stress in

x direction is assumed equal to the applied axial stress, i.e. x o . Stresses in

y direction (clamping stresses) are zero due to inexistence of lateral external force

along the column, i.e. y 0 . Hence, flexibility matrix for axial-shear-flexure element

takes the form

63
f11 f xf f12 f13 o x xa xs xf

f 21 f 22 f 23 0 y (3.45)

f31 f32 f33 xy xy

where xa xs is axial strain in axial-shear model considering the compatibility,

xas xa .

3.5.4.2 Concrete Compression Softening

Concrete compression softening is another interaction term in ASFI approach.

However, unlike interaction of axial deformation from axial-flexural model and axial-

shear model, compression softening of axial-shear model is taken into account in axial-

flexural model. Hence, compression softening factor determined from axial-shear

mechanism (Equation 3.2) must be employed in flexural analysis to soften the uniaxial

compressive behavior of the concrete in axial-flexure model.

3.5.5 Stiffness Model in ASFI

In ASFI approach, stiffnesses of axial-shear-flexure element are modeled as

springs in series as illustrated in Figure 3.9. Shear spring of axial-shear model with

stiffness K s is in series with flexural spring of axial-flexural element of stiffness K f .

Three axial springs, i.e., axial spring in axial-flexural model with stiffness k xf , axial

spring in axial-shear model with stiffness k xs , and axial spring of axial mechanism with

stiffness k xa are also in series. Flexural and shear stiffnesses can be determined based on

following expressions:-

64
f s
Kf , Ks (3.46)
f s

Total stiffness K corresponding to total drift ratio is,

1 1 1
(3.47)
K K f K s

Hence, total shear-total deformation relation can be determined as K where

s f .

As axial strain of axial mechanism is contributed in both axial-shear and axial-

flexural models, in order to determine total axial stiffness k x , axial stiffness from axial-

flexural mechanism is considered into axial-shear mechanism. When axial stiffness is

defined for each axial spring individually based on its deformation, total axial stiffness

can be determined as,

1 1 1 1

k x k xa k xs k xf (3.48)
k x x x where x xs xf xa

where, k xa is axial stiffness of axial mechanism, k xs is axial stiffness of shear mechanism,

and k xf is axial stiffness of flexural mechanism

x
k xa , k xs x , k xf x (3.49)
xa xs xf

3.6 Conclusion

In this chapter, theory and conceptual development of ASFI approach is presented

(Mostafaei and Kabeyasawa, 2007). ASFI approach is a macro-model based approach

65
that consists of two models; axial-flexure and axial-shear model evaluating flexural and

shear responses simultaneously by employing standard flexural section analysis

techniques for axial-flexure response and MCFT for axial-shear response. Total response

at any load step is obtained by adding deformations from component models obtained by

considering interaction of axial strains, compression softening and compatibility and

equilibrium conditions.

66
c2
c2

f co c2

f co c2

c1 c2

c2

co c2

Figure 3.1: Concrete compression softening


P
V M4
M3
Flexural Section 2 M2
Flexural Section 1 M1
Inflection Point Mo
-M1
-M2
-M3
-M4

o2

Flexural Section 2
M2 o2

Centroidal
l12
Axis

Flexural Section 1 h o1
M1
P

o1

Figure 3.2: Axial-Flexural model and determination of axial strain due to flexure
67

y
x x

Lin Lp Lin- Lp
y y

Figure 3.3: Plastic hinge model for flexural deformations

y
y
fy

xy y
c1 y
1
.5
x

xy

fx fx
2
c2

.5
xy x xy

x x

fy
(a) (b) (c)
2

y
y
xy / 2


2 1
2 c

x
x

c2
c1
(d)
Figure 3.4: (a) Membrane element subjected to in-plane stresses (b) Average strains
from in-plane loading (c) Average strains in cracked concrete
(d) Mohrs circle for average strains
68
fc2 f c1

2
f c 2 f co 2 c 2 c 2
f co co co f cr
Ec c1

f cr
1 200 c1

co c2 cr c1
(a) (b)
f si

f si Es si f yi
f yi

yi si
(c)
Figure 3.5: Average stress-strain relationship for
(a) Cracked concrete compression (b) Cracked concrete in tension (c) Reinforcing steel

y y

f sy f sy
ny

f c1 ci
fx 1 1
fx
f sx f sx
nx
2 2
xy x xy x
fy fy
(a) (b)
Figure 3.6: (a) Calculated average stresses (b) Local stresses at a crack

69
Cy

dby


Cx

c
Sx d bx

Figure 3.7: Parameters influencing crack spacing

y y
1

90O

0O
x x
C

2
Figure 3.8: Coordinate transformation for concrete and steel material stiffness matrices

P
M2
V Axial-Shear Ks
Model
k xs
Kf
k xa Ks
Kx
Kf k xf
V
Axial-Flexural
M1
Model
P
Figure 3.9: Spring model for axial-shear-flexural element
70
CHAPTER 4

IMPLEMENTATION OF ASFI APPROACH FOR COLUMNS

4.1 Introduction

The theory and conceptual development of ASFI approach were presented in the

previous chapter by explaining the details of its component models (axial-flexural and

axial-shear) and interaction methodology. Although theoretical aspects of the approach

and the details of the analytical models were derived in generalized form for a reinforced

concrete element between two subsequent flexural sections of the member, the approach

can be extended to evaluate structural performance of the reinforced concrete elements

such as beams and columns. In addition to the models considered for axial, flexural and

shear mechanisms in ASFI approach, the deformations caused by reinforcement slip at

the section of maximum moment must also be incorporated into the analysis for

estimating total response.

In this chapter, the analytical procedure for implementation of ASFI approach for

response estimation of reinforced concrete column subjected to lateral load is outlined.

Solution technique for implementing the analytical procedure is based on ASFI approach

by Mostafaei and Kabeyasawa (2007) as described in Chapter 3. The approach is

implemented in a MATLAB (MathWorks, 2009) computer program written in this study

71
and lateral load-lateral drift responses of previously tested columns are presented to

verify correct implementation of the model. The simulated lateral load-deformation

responses of the test specimens shall be used subsequently for discussion and

comparison.

4.2 Response Estimation of Columns by ASFI Approach

In Chapter 3, it was demonstrated that deformation response of an element

between any two subsequent flexural sections of a reinforced concrete element could be

estimated by considering the interaction of axial deformation and satisfying equilibrium

and compatibility conditions in average stress-average strain environments. This is

theoretically admissible approach as equilibrium of internal and applied forces can be

satisfied at any load level. Hence, this approach can principally be applied for response

estimation of reinforced concrete columns subjected to the lateral loads. However,

instead of considering an arbitrary element between two subsequent flexural sections of

the column, an axial-flexure element and an axial-shear element are considered for

modeling the element from its inflection point to one of the end sections. Conventional

flexural section analysis or fiber model approach, as described in Section 3.3, is applied

for modeling flexural behavior of the axial-flexure element and modified compression

theory as explained in Section 3.4 is employed for modeling axial-shear element. Given

the compatibility and equilibrium relationships, axial-flexure and axial-shear elements are

coupled into one-component model considering axial interaction and material constitutive

relationships in accordance with Section 3.5. In addition, a pullout model as explained

72
below is considered to account for deformation caused by reinforcement slip at the

section adjacent to flexural end section.

4.2.1 Pullout Effect Consideration and Total Deformations

Pullout or reinforcement slip deformations are experienced in fixed ended

reinforced concrete column subjected to bending moment due to rigid body rotation of

the column resulting from slip of longitudinal reinforcing bars out of foundation or beam-

column joint. Unlike flexural and shear deformations that occur along the length of the

column, pullout deformations occur at the end section(s). The mechanics of the pullout

deformations has been explained in detail in Section 2.2.2 and illustrated in Figures 2.1

and 4.5. Axial-flexure model (fiber section analysis) and axial-shear model (MCFT) can

not account for slip or pullout deformations, therefore these must be determined

separately and added to flexural and shear deformations for total response.

To account for pullout effects in total performance of the column, total lateral

drift is taken as sum of the flexural drift ratio, shear strain and pullout rotation Rpul . In

addition, centroidal strain due to pullout pul must also be added to the total axial

deformation of the column. End rotation and centroidal strain due to pullout are

illustrated in Figure 4.5. The pullout element is considered as a rotational spring at

column end, which is in series with springs of flexure, and shear mechanisms. Hence,

total stiffness K corresponding to total drift ratio is,

1 1 1 1

K K f K s K pul (4.1)
K

73
where,

f
K pul (4.2)
pul

s f pul (4.3)

For axial deformation,

1 1 1 1 1

k x k xa k xs k xf k xpul (4.4)
k x x x

where,

x
k xpul (4.5)
xpul

xtot xa xs xf xpul (4.6)

In above equations, K f is flexural stiffness, K s is shear stiffness, is resultant

shear stress and f is shear stress in axial-flexure element, as defined in Section 3.5. In

pullout model, K pul is pullout stiffness, pul is pullout drift that equals Rpul and kxpul is

pullout axial stiffness. In Equation 4.4, k xa , k xs , k xf are axial stiffnesses of axial, shear,

flexural mechanisms, respectively and are defined in Section 3.5.5.

Total axial and lateral deformations in ASFI approach are computed based on

summation of corresponding deformation from four mechanisms of axial, axial-flexure,

axial-shear and pullout. Hence, total lateral drift ratio in Equations 4.3 is sum of shear

deformation s from axial-shear model, lateral flexural deformation f obtained from

axial-flexure model and pullout rotation pul from pullout mode. Similarly, total axial

74
strain x tot in Equation 4.6 is sum of axial strain due to axial xa , shear xs , flexure xf

and pullout xpul pul mechanisms, respectively.

It must be noted that any of the available pullout models such as Otani and Sozen

(1972), Hawkins et al. (1982), Morita and Kaku (1984), Alsiwat and Saatcioglu (1992),

Lehman and Moehle (2000) and Setzler and Sezen (2008) etc can be used to determine

rotation and axial deformation due to reinforcement slip and include these effects in

overall response.

4.2.2 Analytical Procedure for Implementation of ASFI Approach

This section summarizes the procedure for estimating the response of laterally

loaded reinforced concrete columns employing ASFI approach.

1. Consider an axial-flexure element of reinforced concrete column from its

inflection point to one of the end section. Perform flexural section analysis on end section

employing compression-softening factor to soften the compressive stress-strain

relationship of the concrete given by any of the appropriate constitutive law for uniaxial

concrete compressive behavior. Importance of considering confinement effects,

contribution of concrete tensile stresses in the flexural analysis, considering confined

concrete model for core and unconfined concrete model for the cover, and choice of

realistic constitutive material models have already been highlighted in Section 3.3

outlining the details of the axial-flexure model in ASFI approach.

2. Perform axial section analysis at the inflection point and determine axial strain

due to applied axial load only xa considering same material constitutive laws employed

75
for flexural section analysis of the end section. Assuming a linear distribution of average

centroidal axial strain between end section and section at inflection point, axial strain due

to flexure xf can be determined as

lin
1 x
xf
lin
0
o xa
lin
dx 0.5 o xa (4.7)

where, lin is the length of the column between end section and inflection point, o is total

centroidal axial strain determined through flexural section analysis of the end section and

xa is the axial strain due to applied axial load only. It may be noted that xf is also the

average value of axial strains due to flexure only between at end section and section at

inflection point and considered constant over the entire length of the axial-flexure

element.

3. Determine flexural drift and pullout deformation employing any of the

appropriate models and calculate respective stiffnesses.

4. Consider an axial-shear element from inflection point to end section and apply

MCFT as explained in Section 3.4 and determine required material stiffness matrices

considering secant stiffness formulation and relevant material axis transformations.

5. Apply axial strain interaction methodology while satisfying compatibility and

equilibrium conditions as per Section 3.5 and convert axial-flexure and axial-shear

elements into one axial-shear-flexure element.

6. Check convergence of the deformations by considering stiffness model in ASFI

approach as outlined in section 3.5.5 and use of Equation 4.8 relating stresses and strains

of ASFI element through material stiffness matrix. Also, include the effects of pullout

deformation in the ASFI stiffness model as explained in Section 4.2.1.


76
f D D f
1
or (4.8)

4.2.3 Solution Technique for Implementing Analytical Procedure

In the previous section, the analytical procedure for implementation of ASFI

approach is presented. The procedure is iterative and outlines only the major steps

required for response estimation. This section summarizes the solution technique for

implementing above-mentioned analytical procedure for analyzing response of a

reinforced concrete column with ASFI approach.

1. Define or input material properties and geometry. Decide on a sign convention for

tensile and compressive stresses and strains.

2. Input axial load P . If there is no applied axial load, then consider a negligible

small value for P .

3. Select a small value of total drift ratio , such as 0.000001 , as a starting value.

4. Consider the variables of iteration as oi ,i , xi , yi , si and assume some very small

values for each of them for the first iteration. These variables are axial centroidal strain at

the end section in the axial-flexure element ( oi ), curvature of the end section in axial-

flexure element ( i ), average normal strain in x -direction for axial-shear element ( xi ),

average normal strain in y -direction for axial-shear element ( yi ) and average shear

strain of axial-shear element ( si ), respectively.

5. Calculate compression-softening factor with the help of Equation 3.2.

77
6. Perform flexural section analysis at the end section of the axial-flexure element

and determine nominal moment capacity M and corresponding axial centroidal

strain oi 1 .

7. Determine flexural-shear stress f by Equation 3.39 modified for axial-flexure

element of the column as,

M
f (4.9)
Bd f lin

8. Perform axial section analysis at the inflection point and determine axial strain,

xa due to applied axial load only. Calculate axial strain due to flexure xf with Equation

4.7 by replacing o with oi 1 determined in step 6.

9. Determine flexural drift ratio f and pullout deformations by any of the

appropriate models. Pullout deformations include pullout rotations pul and axial strain

due to pullout xpul .

10. Determine o , f xf , K f , and K pul from Equations 3.40, 3.43, 3.46, and 4.2

respectively.

11. Apply MCFT, as explained in Section 3.4 and follow following steps

a. Determine average concrete principal tensile and compressive strains

c1 and c 2 by Equations 3.5 and 3.6, respectively. Also, determine

orientation of principal tensile plane t and principal compressive plane

strains c with the help of Equations 3.7 through 3.9.

78
b. Determine average steel stresses f sx and f sy corresponding to the

strains x and y in accordance with employed constitutive laws for

reinforcement in x and y -directions respectively.

c. Determine average principal tensile and compressive stresses in concrete

by considering same material constitutive laws employed in axial-flexural

model (step 6). However, if in axial-flexure model cover and core concrete

are modeled separately then concrete compressive strength and

corresponding strain must be modified according to the Equation 3.42.

d. Check local stress-strain conditions at cracks ensuring that the tension in

the concrete can be transmitted across the crack and shear stress on the

surface of the crack does not exceed maximum shear provided by the

aggregate interlock by Equations 3.18 through 3.24. If Equation 3.20 is

not satisfied and local shear stress at the surface of the crack exceeds shear

strength provided by aggregate interlock mechanism, then average

concrete principal tensile stress must be reduced as per Equation 3.25.

e. Determine secant moduli for concrete and reinforcements and assemble

material stiffness matrices using Equation 3.27 through 3.30. Determine

total material stiffness D matrix by transforming component material

stiffnesses into global x and y -directions using Equation 3.26 and 2.31.

12. Invert total stiffness matrix D to get flexibility matrix f of Equation 3.44 and

add flexibility component of axial deformation due to flexure f xf to f11 component of f

to determine flexibility matrix of coupled axial-shear-flexure element.

79
13. Consider x o as per Equation 3.40, y 0 , and si from initial assumed

value in the iteration, and obtain s and K s using Equations 4.8 and 3.46. Then determine

K and using Equation 4.1.

14. Assume x o as per Equation 3.40, y 0 , and from step 13, and compute


xi 1 , yi 1 , and si 1 for the integration point using Equation 4.8. Determine f and
Kf

calculate i 1 for the flexural section.

15. Assess the convergence of the acquired variables oi 1 , i 1 , xi 1 , yi 1 , and si 1 . If

satisfactory convergence is achieved, then go to the next step otherwise repeat steps 5 to

15 with improved estimate of the iteration variables.

16. Compute lateral load capacity of the column corresponding to the given total drift

by Equation 3.39 replacing s with obtained in step 13.

17. Compute axial deformation by Equation 4.6.

18. Increment total drift ratio and repeat steps 4 to 17 until required response is

evaluated.

These analytical steps are summarized in a flow chart as shown in Figure 4.1.

4.3 Implementation and Verification of ASFI Approach

Mostafaei and Kabeyasawa (2007) developed and verified ASFI approach by

reporting consistent correlation between estimated analytical response and experimental

results for a number of reinforced concrete test specimens. In order to implement the

ASFI approach, a computer program in MATLAB (MathWorks, 2009) is written and

80
lateral load-deformation responses of previously tested reinforced concrete columns are

estimated. The test specimens modeled in this study are the same which were originally

used by Mostafaei (2006) for verification of ASFI method. Calculated responses are

compared with experimental test data to verify correct implementation of the analytical

procedure of ASFI approach. While modeling test specimens, all efforts are made to use

the same data and models used in the original study and appropriate details are assumed

where sufficient information is not available in the literature.

4.3.1 Details of Test specimens and Material Properties

Five reinforced concrete columns tested by Ousalem et al. (2003) are analyzed in

this study. These columns were loaded under constant axial load of 540 kN and

unidirectional cyclic lateral loads. The columns were designed considering 1/3 scale of

actual sized column, representing columns located in the first story of a building with

moderate height. The columns were designed to fail in different failure mechanisms like

shear-tension, shear-compression, shear-flexure and flexural failures. The test columns

were fixed against rotations at top and the bottom. Geometric and material properties of

test specimens are presented in Table 4.1. The complete details of the tests and discussion

of the analytical results can be found in Mostafaeis PhD dissertation (2006).

4.3.2 Constitutive Laws

Same constitutive laws for concrete and reinforcements are used for both of the

axial-flexure and axial-shear models in ASFI approach. The material constitutive laws

81
used to model behavior of the test columns under consideration are presented in

following sections.

4.3.2.1 Constitutive Law for Concrete in Compression

Stress-strain relationship for concrete in compression employed for both axial-

flexure and axial-shear elements is illustrated in Figure 4.2. Pre-peak stress-strain

relationship i.e., c p is given by following expression

c
2

fc f p 2 c (4.10)
p p

Stress-strain relationship for the post-peak response, i.e., c p is defined based

on concrete confinement as follows

f c f p 1 Z m c p (4.11)

where,

f p K f c' and p K co (4.12)

In Equations 4.11 and 4.12, parameters K and Z m are confinement parameters.

These are determined based on modified Kent and Park model for stress-strain

relationship of concrete confined by rectangular steel hoops (Park et al. 1982).

s f yh
K 1 (4.13)
f c'

0.5
Zm (4.14)
3 0.29 f co 3 h"
K co
145 f c' 1000 4
s
sh

82
where, s is ratio of the volume of rectangular steel hoops to volume of concrete core

measured to outside of the peripheral hoop, f yy is yield strength of the steel hoop, h " is

width of the concrete core measured to the outside peripheral hoop, sh is center to center

spacing of hoop set, and co is strain corresponding to concrete compressive cylinder

strength f co . In Equation 4.12, is compression softening factor as defined in Equation

3.2.

4.3.2.2 Constitutive Law for Concrete in Tension

There is not enough clarification in the literature reported by Mostafaei (2006)

about the final choice of the tensile concrete constitutive material model used for

producing load-deformation responses of the test specimens. The reported literature

discusses two models, one developed by Vecchio and Collins (1986) (the model used in

original MCFT, details of which are given in Section 3.4.1.2) and the other based on

model presented by Izumo (1992). Theoretically, any of the appropriate model can be

used to simulate concrete behavior in tension. However, it is assumed here that

constitutive relationship for average concrete tensile stress-strain used in the originally

reported estimated responses is the one based on model by Izumo. The model is

presented below and illustrated in Figure 4.3.

The relationship suggested prior to cracking i.e., 0 c1 cr is linearly

elastic and is given by the expression:-

f c1 Ec c1 (4.15)

83
where, Ec is modulus of elasticity of the concrete, cr is cracking strain corresponding to

uniaxial cracking strength of concrete f cr , and f co uniaxial compressive strength of

concrete in MPa. These quantities are given by following expressions

Ec 2 fc' co ; cr fcr Ec ; f cr 0.33 f c' (4.16)

After cracking, the concrete tensile stresses remain constant up to strain of

2 cr and then decline in post-peak range as given by following expressions.

f c1 f cr for cr c1 2 cr (4.17)

0.4
2
f c1 f cr cr for c1 2 cr (4.18)
c1

4.3.2.3 Constitutive Law for Reinforcements

A tri-linear stress-strain relationship is considered for the longitudinal and

transverse reinforcements as illustrated in Figure 4.4 and expressed by following set of

expressions.

f s Es s for s y
fs f y for y s sh (4.19)
s sh
fs f y +
u sh
f u fy for sh s u

where, E s modulus of elasticity of steel, f y and fu are yield and ultimate strengths of

steel, respectively, y , sh and u are yield strain, strain at onset of strain hardening and

ultimate strain, respectively.

84
4.3.3 Flexural Deformation Model

Plastic hinge model as illustrated in Figure 3.3 is employed to model flexural

deformations by integrating curvatures along the length of the column. A linear curvature

distribution is assumed in the elastic range, and the inelastic curvature is lumped at the

column end over plastic hinge length. Plastic hinge length L p is taken equal to the height

of the column cross-section. Hence, flexural drift ratio can be determined as,

Lin
1
f
Lin x ( x)dx
0
(4.20)

Using Equation 4.20, flexural drift ratio before yielding, i.e., y can be calculated as,

1
f Lin (4.21)
3

and, after yielding i.e., y

1
Lin L p Lin 1 y Lin Lp
Lin
f x ( x)dx x ( x)dx x 2 dx xdx (4.22)
Lin 0


L Lin L L
in p


Lin L p in
0

In above equations, Lin is the distance from column end section to inflection

point, L p is plastic hinge length, ( x ) is curvature function at distance x measured from

inflection point and y is yield curvature.

4.3.4 Pullout Deformations Model

Pullout model by Okamura and Maekawa (1991) is employed in the analytical

process of ASFI approach to model pullout deformations of test specimens. The pullout

85
model and reinforcement slip-strain relations are illustrated in Figure 4.5 and 4.6.

Important relations used in the model are

23
20 MPa
slip SD ' (4.23)
fc

slip
pul Rpul (4.24)
X

pul
eR 0
pul
(4.25)
Lin

where S is normalized slip as defined in Figure 4.6, D is reinforcing bar diameter in

mm, f c' is concrete compressive strength in MPa, pul is pullout drift ratio which equals

pullout rotation R pul , pul is centroidal strain due to pullout, Lin is the distance from the

end joint to the inflection point. Other parameters like slip , X and e are as shown in

Figure 4.5.

4.3.5 Buckling of Compression Bars and Post-peak Analysis

In ASFI approach, buckling or slip of compression bars is considered by lowering

compressive stresses in longitudinal reinforcing bars when unconfined cover concrete

fibers adjacent to the reinforcement reach 30 % of the maximum concrete strength. After

reaching this stage, steel stresses are reduced linearly according to post-peak compression

stiffness of the confined core.

It must also be noted that in post-peak analysis, pullout and shear stiffnesses are

kept constant at their least value during the analysis. In post-peak stage of the analysis,

only flexural stiffness is the main variable. Also, compression softening factor is used

86
with its calculated value until it start to increase. At this stage, it must then be kept

constant equal to its minimum value for rest of the analysis. In ASFI approach, axial

failure or gravity collapse is defined as the stage when equilibrium in vertical direction in

section analysis cannot be satisfied any more under the applied axial load. At this stage,

the column reaches complete loss of lateral load capacity.

4.4 Comparison of the Results

Lateral load-drift responses for specimens No 5, 12, 14, 15 and 16 are presented

in Figure 4.7 through 4.11. Comparison of the predicted response with experimental data

indicates that ASFI approach performs very well in evaluating structural performance of

all of the test columns. Predicted responses follow initial stiffness very well and peak

loads are estimated quite accurately. Estimated post-peak responses also compare well

with experimental post-peak behaviors in general. Good correlation achieved between

calculated and actual behavior indicates that ASFI approach is effective displacement

based evaluation approach for response estimation of reinforced concrete columns. It can

also be concluded that the ASFI approach has correctly been implemented in the analysis

as per analytical procedure described in Section 4.2 above.

4.5 Summary and Conclusion

Analytical procedure and solution technique for implementation of ASFI

approach for response estimation of reinforced concrete columns is presented. The

procedure in incorporated into a programming routine written in MATLAB (MathWorks,

2009) and response of five columns is estimated based on ASFI approach. Comparison of

87
calculated total lateral load-total lateral drift responses of test specimens with their

experimental responses indicates that ASFI approach is an effective tool for response

estimation of columns to lateral loads. The results also show that the analytical procedure

and solution technique of ASFI approach has correctly been implemented in Matlab

computer program.

88
Input material properties and
geometrical dimensions

Input axial load P . If P 0 , then define P as a negligibly small value.


Specify initial/increment total lateral drift value

Consider variables in iteration i as oi , i , xi , yi , si

Apply axial-flexural model and compute


oi 1 , xa , f , f , pul , pul , f xf , K f , K pul

Apply axial-shear model based on MCFT and construct total


stiffness matrix D of axial-shear element.

Compute flexibility matrix f of axial-shear element by inverting D and obtain total


flexibility matrix of axial-shear-flexure element by adding f xf to f11 term of f

Considering x P bh , y 0 , and si obtain s and K s s si . Get K and .

Considering x P bh , y 0 , and , compute xi 1 , yi 1 , and si 1 for the



integration point and then calculate f and i 1 for the flexural section
Kf

Deformation converged No
oi 1 ,i 1 , xi 1 , yi 1 , si 1

Output shear force and axial strain and calculate total load V bd s .
Increment drift ratio and repeat above steps until failure

Figure 4.1: Flow chart for implementation of ASFI approach for columns

89
35
Confined concrete

30 Unconfined concrete

Stress (Mpa) 25

20

15

10

0
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035
Strain

Figure 4.2: Constitutive relationships for concrete in compression

1.8

1.6

1.4

1.2
Stress (Mpa)

0.8

0.6

0.4

0.2

0
0 0.2 0.4 0.6 0.8 1
Strain x 10
-3

Figure 4.3: Constitutive relationship for concrete in tension


90
600

500

400
Stress (Mpa)

300

200

100

0
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16
Strain

Figure 4.4: Constitutive relationship for reinforcing steel

Rpul
pul slip

slip
X pul R pul
X
e
pul
e 0
pul

Lin

Figure 4.5: Rigid body rotation of column due to reinforcement slip or pullout
91
S Normalized slip

20MPa
slip S D
23

S
Sy

6 7 f u f y s sh
'
fc 2 100

S y 0.2 fu f y s sh

Sy
S s 6 3500 s
s
y sh

Figure 4.6: Pullout deformation model (Okamura and Maekawa, 1991)

92
350
Reported response
Predicted response
300

250
Lateral force (kN)

200

150

100

50

0
0 0.01 0.02 0.03 0.04 0.05 0.06
Drift ratio

Figure 4.7: Comparison of the results for Specimen-5 (Ousalem et al, 2005)

93
300
Reported response
Predicted response
250

200
Lateral force (kN)

150

100

50

0
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045
Drift ratio

Figure 4.8: Comparison of the results for Specimen-12 (Ousalem et al, 2005)
94
350
Reported response
Predicted response

300

250
Lateral force (kN)

200

150

100

50

0
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09
Drift ratio

Figure 4.9: Comparison of the results for Specimen-14 (Ousalem et al, 2005)
95
350
Reported response

300 Predicted response

250
Lateral force (kN)

200

150

100

50

0
0 0.05 0.1 0.15 0.2
Drift ratio

Figure 4.10: Comparison of the results for Specimen-15 (Ousalem et al, 2005)
96
400
Reported response
350 Predicted response

Lateral force (kN) 300

250

200

150

100

50

0
0 0.005 0.01 0.015 0.02 0.025 0.03 0.035 0.04 0.045
Drift ratio

Figure 4.11: Comparison of the results for Specimen-16 (Ousalem et al, 2005)
97
bxh 2Lin sh g w fyx fyy fco
Specimen
(mm) (mm) (mm) (%) (%) (Mpa) (Mpa) (Mpa)

No. 5 300x300 750 50 2.65 0.43 420 410 28.5

No. 12 300x300 900 150 2.26 0.14 415 410 28

No. 14 300x300 900 50 2.26 0.43 415 410 26

No. 15 300x300 900 50 2.26 0.85 415 410 26

No. 16 300x300 600 50 1.80 0.43 415 410 27

b = Width of the column cross section; h = Height of the column cross section;
Lin = Length of the column from inflection point to end section; sh = Hoop spacing;
g = Longitudinal reinforcement ratio; w =Transverse reinforcement ratio;
fyx = Longitudinal reinforcement yield stress; fyy = Transverse reinforcement yield
stress; fco = Concrete compressive cylinder strength

Table 4.1: Geometric and material properties of the test specimens

98
CHAPTER 5

COMPARISON OF DISPLACEMENT COMPONENT MODEL AND ASFI


APPROACH

5.1 Introduction

In previous chapters, two approaches for response estimation of reinforced

concrete columns were presented. In order to verify correct implementation of each

approach, test specimens were modeled with material constitutive relationships and other

component models as proposed in the original study. Comparison of predicted and

experimental responses indicated that both methods perform equally well despite

following quite different approaches for response estimation of respective set of test

columns. In order to evaluate performance of both understudy approaches, it is necessary

to model the same set of specimens employing same material and deformation

component models within each approach.

In this chapter, a comparative study is conducted to evaluate capabilities and

limitations of displacement component model by Setzler and Sezen (2005) and ASFI

approach by Mostafaei and Kabeyasawa (2007). Both approaches are implemented for

same columns employing same set of models for material stress-strain relationships,

flexural and pullout deformations. This comparison forms the basis of the proposed

model presented in Chapter 6.

99
5.2 Implementation of the Analytical Models

In Chapter 2, displacement component model by Setzler and Sezen (2005) is

outlined and response estimated for four columns (Sezen, 2002) is compared with

experimental responses. In Chapter 4, lateral load-deformation responses of specimens

tested by Ousalem et al. (2003) based on ASFI approach are presented. Within

implementation of each of these approaches, same material models are employed which

were originally used in development and verification of the respective approaches. For

comparing estimated responses by both of understudy approaches, it is imperative that an

appropriate set of test specimens and material models are selected and employed for

response estimation by each approach.

5.2.1 Test Specimens and Material Properties

Reinforced concrete columns tested by Sezen (2002) provide data for flexural,

reinforcement slip and shear deformations in addition to total deformation responses.

This provides a unique opportunity to compare and evaluate capabilities of any of the

analytical model intending to predict overall response. Hence, columns by Sezen (2002)

are employed for estimation and comparison of responses by both approaches. Details of

these test specimens were presented in Section 2.5.1 and key features are repeated here

for easy reference.

These columns, designated as Specimen-1 through -4, are fixed ended columns

with square cross-section of 18 inches and length of 116 inches. The columns have eight

No.9 bars as longitudinal reinforcement and transverse reinforcement of No.3 ties with

90- degree hooks at 12 inches center-to-center spacing. Specimen-1, 2 and 4 were tested

100
with constant axial load of 150, 600 and 150 kip, respectively. Specimen-3 was, however,

tested under varying axial load from 60 kip in tension to 600 kips in compression.

Specimen-1 through 3 were subjected to unidirectional cyclic lateral loading whereas,

Specimen-4 was tested under monotonically increasing load after few initial cycles of

elastic loading.

All of the test specimens are modeled with average concrete compressive strength

of 3077 psi and maximum aggregate size of 1 inch. Yield strength of longitudinal and

transverse reinforcement are taken to be 59 and 69 ksi, respectively.

5.2.2 Material Constitutive Relationships

5.2.2.1 Concrete in Compression

In chapter 2, Setzler and Sezens approach was implemented with segmental

stress-strain relationship for concrete in compression derived from Mander et al. (1988)

confined concrete model till peak, and slope of descending branch of Roy & and Sozen

(1964) model after peak (Figure 2.9). In ASFI approach, concrete compressive behavior

was modeled by modified Kent and Park model, 1982 (Figure 4.2). Theoretically, any of

the available constitutive relationships can be used to model compressive concrete

behavior. In this comparison, Mander et al. (1988) model for confined and unconfined

concrete is employed for both analytical approaches. This model is presented below and

is illustrated in Figure 5.1.

According to Mander et al. (1988) model, compressive stress-strain relationship

for confined concrete is defined as

101
f cc r c cc
fc
r 1 c cc
r

f cc
cc co 1 5 1 (5.1)
f co
Ec f
r ; Ec 57, 000 f co ; Esec cc
Ec Esec cc

where f cc is the peak confined concrete strength, c is the concrete strain, cc is the

concrete strain at peak stress for confined concrete, co is the concrete strain at peak

stress in unconfined concrete (taken here as 0.002), f co is the concrete compressive

cylinder strength (in psi units), Ec is the modulus of elasticity of the concrete for normal

weight concrete, and Esec is the secant modulus of the concrete.

In this study, ultimate compressive strain cu for confined concrete is calculated

from Equation 5.2, which is obtained by slightly modifying the maximum strain formula

for spirally confined concrete in Priestley (1996).

cu 0.004 0.14 x y
f yy
(5.2)
f cc

dt2
x y (for square columns) (5.3)
2 sd c

where f yy is yield strength, d t is diameter and s is center-to center spacing of the

transverse reinforcement, respectively.

For unconfined cover concrete, compressive stress-strain relationship is defined

by following equations

102
f co r1 c co
fc for c 2 co
r1 1 c co 1
r

(5.4)
Ec f
r1 ; Ec 57000 f co ; Esec co
Ec Esec co

After reaching the strain of 2 co , cover concrete is assumed to start spalling and

part of falling branch in the region where c 2 co is assumed to be a straight line which

reaches zero stress at spalling strain sp , taken equal to 0.006 in this study (Figure 5.1).

In ASFI approach, in order to apply compression softening effect, compressive

behavior for confined and unconfined concrete obtained from Equations 5.1 through 5.4

must be multiplied by compression softening factor . The compression softening factor

was defined in Equation 3.2.

5.2.2.2 Concrete in Tension

Concrete behavior in tension is not taken into account in flexural analysis by

displacement component model by Setzler and Sezen (2005). However, in ASFI

approach, concrete tensile stresses are considered in both axial-flexure and axial-shear

models. The model by Vecchio and Collins (1986) defines concrete behavior in tension

for ASFI approach. This model will be used in this study and is presented in Section

3.4.1.2 and is illustrated in Figure 5.2.

5.2.2.3 Reinforcing steel

The reinforcing steel behavior in this study is modeled considering a linear elastic

behavior, a yield plateau, and a non-linear strain-hardening region, for both analytical

103
approaches. This model is defined in Section 2.5.2 and is illustrated in Figure 5.3. In

Setzler and Sezens model, same constitutive laws are used for steel behavior in tension

and compression. In ASFI approach, compressive steel behavior is not the same as tensile

steel behavior, and is modified to consider effect of compression bars buckling. After

reaching buckling strain, compressive steel stresses are reduced linearly according to

slope of post-peak confined concrete behavior.

5.2.3 Flexural Deformation Model

Fiber section analysis is conducted for determining moment-curvature

relationship for both of the analytical approaches. For ASFI approach, however, concrete

tensile stresses and compression softening factor are considered in the section analysis.

Flexural deformations are determined by plastic hinge model as defined in Section 4.3.3

and Figure 5.4. This model assumes a linear curvature distribution in elastic range and

inelastic curvatures are lumped at the column end over plastic hinge length L p . In

displacement component model, plastic hinge length is as one half of the section depth,

whereas ASFI approach considers this length equal to section depth. For comparison,

plastic hinge length is taken as one half of the section depth for calculating flexural

deformations by both approaches.

5.2.4 Pullout Deformation Model

Lateral deformations due to pullout or reinforcement slip are calculated by

employing Sezen and Setzler (2008) model. This model has was explained in detail in

Section 2.2.2 and Figures 2.3 and 2.4.


104
5.2.5 Buckling of Compression Bars

ASFI approach considers buckling of compression bars in flexural

analysis after unconfined cover concrete fiber next to compression steel reaches

compressive strength of about 30% of maximum concrete strength. Stresses in

compression bars are then linearly reduced according to slope of post-peak confined

concrete stiffness. In this study, compression steel stresses were reduced as per the model

given in Figure 6.3. According to this model, compression stresses start to decrease when

unconfined cover concrete start to spall. When this happens, corresponding strain in the

relevant steel layer can be calculated from flexural strain distribution across the cross

section depth. This strain is sp as shown in the figure. This point can fall anywhere on

typical stress-strain relationship for steel depending upon the level of flexural strain. Steel

stresses follow their usual constitutive stress-strain relationship until strain reaches this

limit. Then compression stresses in reinforcement follow new path defined by line

joining peak stress point to residual strength point having slope m, which is the same for

descending branch of concrete compression strength. This model is used for modeling

compression steel behavior for in ASFI approach in this study. Displacement component

model by Setzler and Sezen does not consider the bar buckling effect and employs and

same constitutive laws for steel behavior in tension and compression.

5.3 Comparison of Predicted and Experimental Response

Displacement component model and ASFI approach are implemented following

their respective analytical procedures for same set of test specimens with above-

105
mentioned material and deformation component models. The lateral load-displacement

relationships for component and total responses are presented and discussed as follows.

5.3.1 Flexural Displacements

Lateral load-flexural displacement relationships for Specimen-1 through -4 are

presented in Figures 5.5 through 5.8, respectively. For Specimen-1, both approaches

predict identical pre-peak response, which matches very well with the experimental data.

Peak load and deformation at peak load is also estimated very well by both approaches.

For post peak behavior, however, predicted responses are quite different. Post peak

response by ASFI approach closely follows stiffness of the actual response. The

estimated response by displacement component model initially follows post peak

stiffness and then becomes flat.

Lateral load-flexural displacement relationship for Specimen-2 is presented in

Figure 5.6. Initially, both approaches predict similar responses but after reaching the

strains corresponding to start of compression bars buckling, response by ASFI approach

drops gradually till peak load which is slightly underestimated. Predicted response by

displacement component model follows experimental pre peak response well but slightly

overestimates the peak load in positive direction. Again, post peak responses by both

approaches are diverging shortly after peak is reached. However post-peak stiffness of

ASFI approach somewhat matches experimental slope.

Figure 5.7 shows lateral load-flexural displacement relationships for Specimen-3.

Displacement component model overestimates initial stiffness and peak strength; whereas

ASFI approach does a good job in capturing initial response and peak strength. For

106
Specimen-4, ASFI approach predicts actual response that follows experimental pre and

post peak responses very well, as shown in Figure 5.8. With displacement component

model, response until peak load is estimated almost exactly and then diverges with gentle

slope from experimental behavior and response predicted by ASFI approach.

Comparison of the predicted and experimental responses for flexural

deformations indicates that displacement component model and ASFI approach predict

pre peak response almost identically. However, post peak responses differ significantly. It

should be recalled that lateral load-flexural displacement relationships are calculated by

employing common material constitutive laws, fiber section analysis procedures and

flexural deformation model, except that concrete behavior in tension, compression

softening effect and buckling of compression bars are considered in ASFI approach. It

can be seen from Figure 5.9, that difference in moment-curvature relationship with and

without considering concrete tensile behavior in section analysis is not very significant.

Compression softening factor in ASFI approach for all of the specimens modeled was

calculated to be 1.00 for almost all of the loading steps. It means that increasing shear

deformations did not affect flexural performance and both approaches employed same

uniaxial concrete compressive behavior in flexural analysis. The effect of compression

softening is illustrated in Figure 5.10. It can be seen that, this effect is not very significant

for the test specimens modeled in this chapter. If, however, compression-softening factor

has values less than 1.00, then its effect on moment-curvature relationship and softened

response can be significant.

This implies that the only difference in near-peak and post-peak responses

predicted by displacement component model and ASFI approach is due to considering

107
buckling of compression bars in ASFI approach. The effect of buckling of compression

bars is highlighted in Figure 5.1. It can be seen that the response is softened considerably

at peak and post peak stage after considering compression bars buckling effect.

After comparing responses estimated by both approaches with experimental data,

it can be concluded that ASFI approach does a relatively better job than displacement

component model in predicting pre peak response, peak load and post peak response in

flexure.

5.3.2 Reinforcement Slip Displacements

Lateral load-slip displacement relationships for Specimen-1 through -4 are

presented in Figure 5.12 through 5.15, respectively. After comparison of predicted

responses by displacement component model and ASFI approach and experimental data,

same conclusions are drawn as those of flexural deformations. Both approaches produce

almost identical response up until peak load and then diverge at post peak stage. Again,

this highlights the need for considering buckling of compression bars in the flexural

analysis, or another approach to better represent the post-peak response degradation.

5.3.3 Shear Displacements

Lateral load-shear displacement relationships for Specimen-1 through -4 are

presented in Figures 5.16 through 5.19, respectively. For Specimen-1, both approaches

predict identical initial response that matches initial stiffness very well. Near the peak,

predicted response by displacement component model follows the experimental response

slightly better than response by ASFI approach, which continues with same initial

108
stiffness until peak load. Peak load is captured very well by ASFI approach but the

displacement at the peak is underestimated. With displacement component model, peak

load is slightly overestimated and displacement at peak load is relatively closer to the

experimental value. Shear response by ASFI approach terminates at peak load and post

peak response is not captured, as it employs MCFT for estimating shear response. MCFT

being a force based approach has a limitation of predicting response till peak strength

only. The displacement component model does a fair good job in post peak response and

overestimates the displacements as shear strength degrades.

Figure 5.17 shows lateral load-shear displacement relationship for Specimen-2.

Predicted responses by both approaches are identical until observed peak and follows

experimental data perfectly. The peak load is predicted very well by ASFI approach and

overestimated by displacement component model in positive direction. But in the

negative direction, displacement component model captures shear strength very

accurately and ASFI approach underestimates it. This column experienced a brittle

flexural compression failure and did not show degrading shear behavior. Post peak

response by displacement component model showed the same trend. Both approaches

predict initial response for Specimen-3 reasonably that follows initial experimental

stiffness (Figure 5.18). Peak strength is significantly overestimated by displacement

component model, but post peak response captures slope of degrading branch very well.

Peak strength is accurately predicted by ASFI approach however, displacement at peak

load is underestimated. Similar trends are observed for response estimated for Specimen-

4 as shown in Figure 5.19. Initial strength is estimated fairly well by displacement

component model, but peak strength is slightly overestimated. The degrading portion

109
seems to follow the data well, although the deformations are overestimated. Response

estimated by ASFI approach also matches pre peak stiffness well. The peak strength is

estimated very well but the displacement at peak strength is underestimated.

As already described in Chapter 2, pre-peak shear response in displacement

component model is obtained indirectly from Response-2000. It must be noted that

Response-2000 is basically a section analysis routine that incorporates shear effects in the

analysis by implementing MCFT. In this approach, reinforced concrete cross-section is

considered a stack of biaxially loaded concrete elements and longitudinal steel elements.

Each of the concrete elements is analyzed individually for in-plane stress field based on

MCFT and overall section equilibrium conditions are satisfied. This is very rigorous

procedure and can only be implemented in a sophisticated computer program, such as

Response 2000. Nevertheless, it produces realistic response especially for shear critical

elements. In addition to producing total load-deformation response, Response-2000 also

delivers shear strain distribution along the length of the column for each load step. In

displacement component model, this shear strain distribution is copied manually and then

integrated over the length of the column to obtain shear deformation corresponding to

respective load step. As the analytical procedures adopted within Rsponse-2000 are very

deliberate and involved, a good correlation with experimental data is achieved for pre

peak shear response in displacement component model, as seen in the comparison of the

results.

On the other hand, ASFI approach also employs MCFT to estimate shear

deformations, but in a much simplified way than the analytical procedures adopted within

Response 2000. This is done by considering the column length between inflection point

110
and end section as a single shear element subjected to biaxial state of stresses, which are

assumed constant over the length of the element. The response of this element to average

shear and axial stresses is determined through MCFT satisfying compatibility and

equilibrium conditions described in Section 3.5. While evaluating shear response of the

column, average axial strain due to flexure is added to axial strain of axial-shear element.

This is a relatively simple procedure that can be implemented easily in hand calculations

and reasonably good response can be obtained in few steps. In Figures 5.16 through 5.19,

pre peak shear responses estimated by ASFI approach show satisfactory correlation with

experimental data.

Comparison of responses predicted by both approaches with test data shows that

pre peak response estimated by displacement component model follows experimental

curve slightly better than response estimated by ASFI approach. In view of the simplified

procedure employed within ASFI approach, minor loss of accuracy near the peak strength

is acceptable and does not affect overall response much.

As MCFT is a force based approach, shear response in ASFI approach terminates

at peak strength and does not show shear strength degradation with increasing shear

deformations. After reaching peak load, shear deformations in ASFI approach are

calculated from secant stiffness at peak strength, which is kept constant for post peak

behavior. In displacement component model, after estimating peak strength, shear

strength is assumed to remain constant at its peak value until onset of shear degradation.

From this point, the response degrades linearly to the point of axial load failure, where

strength is zero and drift is calculated from the procedure explained in Section 2.2.3.

111
After comparing experimental data with predicted responses, it can be concluded

that shear response envelope of displacement component model is a better choice for post

peak analysis over ASFI approach, which does not give any information on shear strength

degradation.

5.3.4 Total Response

Figure 5.20 shows the comparison of predicted and experimental lateral load-total

displacement relationships for Specimen-1. Shear strength of this specimen by Sezen-

Moehle equation (Equation 2.13) is calculated to be 69.0 kips and flexural strength from

moment-curvature analysis is 70.0 kips. As per displacement component model, the

column is classified as category-III specimen, for which total displacement at any point in

the response is sum of flexural, slip and shear displacement at that load step (Section 2.3).

The initial response is predicted very well up to the peak strength. Peak strength and

deformation at peak load are captured almost exactly. The post peak response initially

follow the experimental response and then becomes flat and deformations are over-

predicted. Response predicted by ASFI approach follows experimental data and response

predicted by displacement component model closely for most of the pre peak response.

The peak strength is predicted well but deformation at peak load and post peak

deformations are greatly underestimated.

Comparison of predicted and experimental lateral load-total displacement

relationships for Specimen-2 is presented in Figure 5.21. As per displacement component

model, this column has shear and flexural strengths of 92.0 and 72.0 kips, respectively,

thus classifying it into category-IV specimen. For this column, shear deformation is

112
frozen at its value at peak strength (flexural strength, 72.0 kips) and added to flexural and

slip displacements for post-peak response. Predicted response by this approach, slightly

overestimates pre peak stiffness and peak load in positive direction and follows post peak

experimental response fairly in both directions. Predicted response by ASFI approach

also slightly overestimates pre peak response in positive direction and underestimates

peak strength in both directions. The deformations at peak load are, however captured

well. Post peak response in negative side follows observed response better than positive

direction. For specimen-3, lateral load-displacement relationships are shown in Figure

5.22. pre peak response predicted by both approaches is identical and overestimates pre

peak stiffness of the experimental behavior. Peak strength is slightly overestimated by

displacement component model and, is underestimated by ASFI approach. Displacement

at peak load is however, well captured by both methods. Post peak response by

displacement component model matches slope of the observed response, but ASFI

approach significantly underestimates post peak response.

Figure 5.23 shows comparisons of predicted and experimental lateral load-total

displacement relationships for Specimen-4. This column has identical to Specimen-1

except that it was tested under monotonically increasing lateral load after few initial

elastic cycles. Comparison of shear and flexural strength classifies this column into

category-III column as per displacement component model. The predicted response by

displacement component model follow the trend in experimental data but slightly

overestimates initial stiffness, peak strength and post peak response. Predicted response

by ASFI approach also overestimates initial stiffness and captures peak strength well.

113
Deformation at peak strength and post peak deformations are significantly

underestimated.

Comparison of the predicted responses by displacement component model and

ASFI approach with each other and experimental test data shows very similar trends for

all of the modeled test specimens. Both approaches produce identical pre peak responses

that generally follow experimental data well. For post peak response, displacement

component model performs relatively better than ASFI approach and deformations are

slightly overestimated. Whereas, predicted post peak deformations by ASFI approach are

significantly underestimated.

It must be noted that the rules for combining deformation components for

response up to the peak are same in both approaches and flexural, bar slip and shear

deformations are simply added together to obtain total response. That is why predicted

pre peak responses by both of the approaches are identical for all test specimens as shown

above. For post peak analysis, the rules for combining deformation components are

however, quite different in both approaches. In displacement component model, post

peak deformations are obtained by combining flexural, bar slip and shear deformations

depending upon predicted failure mode according to set of rules explained in Section 2.3

and Figures 2.7 and 2.8. In ASFI approach, post peak response is calculated with constant

pullout and shear stiffnesses at the values corresponding to peak load and only axial-

flexural model continues until axial load failure. Reinforcement slip and shear

deformations are calculated from their respective constant secant stiffnesses, and are

added to flexural deformations at any load level in post peak state. Estimation of total

response in this way forces slip and shear deformations to decrease with degrading lateral

114
load capacity of the column. This is the reason that total post peak response by ASFI

approach is underestimated for test specimens modeled in this chapter.

In ASFI approach, plastic hinge length in flexural deformation model is taken

equal to section depth while in displacement component model, it is taken as half of the

section depth. In order to compare response estimations by both approaches, same plastic

hinge length was employed and flexural deformations in this chapter were calculated with

plastic hinge length of one-half of the section depth. Figure 5.24 shows a comparison of

lateral load-displacement relationships by ASFI approach for Specimen-1, calculated

with plastic hinge length taken equal to section depth and half of the section depth. It can

clearly be seen that post peak response improves significantly when total deformations

are calculated with plastic hinge length equal to section depth. On the other hand, if

lateral load-flexural displacement relationships for the same specimen is analyzed (Figure

5.25), it is seen that post peak flexural response calculated with plastic hinge length of

one-half of section depth gives very good match with experimental data. This implies that

response estimation in ASFI approach by taking plastic hinge length equal to section

depth does predict a more accurate response and flexural deformations be calculated with

plastic hinge length equal to half of the section depth, as it conforms to behavior. Also, it

is recommended that an effective way to incorporate for combination of deformation

components in ASFI approach must be considered for total response in post peak state.

115
5.4 Conclusion

In this chapter, responses estimated by displacement component model by Setzler

and Sezen (2005) and ASFI approach by Mostafaei and Kabeyasawa (2007) are presented

by modeling the behavior of same set of test specimens. A detailed discussion of the

results is carried out to get an insight into analytical procedures of both approaches.

Conclusions are drawn after comparison of predicted responses for each of the flexural,

slip, shear and total deformations with respective experimental data. Conclusions drawn

in this chapter will form the basis for a model proposed in the next chapter.

116
3500
Confined concrete

3000 Unconfined concrete

Stress (psi) 2500

2000

1500

1000

500

0
0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016
Strain

Figure 5.1: Constitutive relationships concrete in compression

250

200

150
Stress (psi)

100

50

0
0 0.2 0.4 0.6 0.8 1
Strain x 10
-3

Figure 5.2: Constitutive relationship concrete in tension

117
100

90

80

70

60
Stress (ksi)

50

40

30

20

10

0
0 0.025 0.05 0.075 0.1 0.125 0.15
Strain

Figure 5.3: Constitutive relationship for reinforcing steel

y
x x

Lp L - Lp L
y y

Figure 5.4: Plastic Hinge Model for Determining Flexural Deformation

118
80

60

40
Lateral load (kip)
20

-20

-40

Test data
-60
Setzler and Sezen
ASFI approach
-80
-3 -2 -1 0 1 2 3
Flexural displacement (in.)

Figure 5.5: Lateral load flexural displacement relationship for Specimen-1

80

60

40

20
Lateral force (kip)

-20

-40

-60
Test data
-80 Setzler and Sezen
ASFI approach
-100
-1.5 -1 -0.5 0 0.5 1 1.5
Flexural displacement (in.)

Figure 5.6: Lateral load flexural displacement relationship for Specimen-2

119
80

60

40
Lateral force (kip)

20

-20

Test data
-40
Setzler and Sezen
ASFI approach
-60
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
Flexural displacement (in.)

Figure 5.7: Lateral load flexural displacement relationship for Specimen-3

80

60

40
Lateral force (kip)

20

-20

-40

Test data
-60
Setzler and Sezen
ASFI approach
-80
-3 -2 -1 0 1 2 3 4
Flexural displacement (in.)

Figure 5.8: Lateral load flexural displacement relationship for Specimen-4

120
4500

4000

3500

3000

2500
M (k-in.)

2000

1500

1000 Specimen-1 (with concrete tension)


Specimen-1 (without concrete tension)
500 Specimen-2 (with concrete tension)
Specimen-2 (without concrete tension)
0
0 0.5 1 1.5 2 2.5 3
Curvature (1/in.) x 10
-3

Figure 5.9: Effect of concrete tensile stress in flexural section analysis

4000

3500

3000

2500
Moment (k-in.)

2000

1500

1000
Specimen-1 (Setzler and Sezen)
Specimen-1 (ASFI approach)
500
Specimen-2 (Setzler and Sezen)
Specimen-2 (ASFI approach)
0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
Curvature (1/in.) x 10
-3

Figure 5.10: Effect of concrete compression softening in flexural section analysis

121
4500

4000

3500

Moment (k-in.) 3000

2500

2000

1500

1000

500 Specimen-1 (with bar buckling)


Specimen-1 (without bar buckling)
0 Specimen-2 (with bar buckling)
Specimen-2 (without bar buckling)
-500
0 0.5 1 1.5 2 2.5 3
Curvature (1/in.) x 10
-3

Figure 5.11: Effect of compression bars buckling in flexural section analysis

80

60

40
Lateral load (kip)

20

-20

-40

Test data
-60
Setzler and Sezen
ASFI appoach
-80
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
Slip displacement (in.)

Figure 5.12: Lateral load slip displacement relationship for Specimen-1

122
80

60

40

20
Lateral force (kip)

-20

-40

-60
Test data
-80 Setzler and Sezen
ASFI approach
-100
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
Slip displacement (in.)

Figure 5.13: Lateral load slip displacement relationship for Specimen-2

80

60

40
Lateral force (kip)

20

-20

Test data
-40
Setzler and Sezen
ASFI approach
-60
-1 -0.5 0 0.5 1 1.5 2
Slip displacement (in.)

Figure 5.14: Lateral load slip displacement relationship for Specimen-3

123
80

60

40
Lateral force (kip)
20

-20

-40

Test data
-60
Setzler and Sezen
ASFI approach
-80
-2 -1 0 1 2 3 4
Slip displacement (in.)

Figure 5.15: Lateral load slip displacement relationship for Specimen-4

80
Test data
60 Setzler and Sezen
ASFI approach
40
Lateral load (kip)

20

-20

-40

-60

-80
-2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2
Shear displacement (in.)

Figure 5.16: Lateral load shear displacement relationship for Specimen-1

124
100
Test data
80
Setzler and Sezen
60 ASFI approach

Lateral force (kip) 40

20

-20

-40

-60

-80

-100
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.5
Shear displacement (in.)

Figure 5.17: Lateral load shear displacement relationship for Specimen-2

100

80

60
Lateral force (kip)

40

20

-20

Test data
-40
Setzler and Sezen
ASFI approach
-60
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8
Shear displacement (in.)

Figure 5.18: Lateral load shear displacement relationship for Specimen-3

125
80

60

40
Lateral force (kip)
20

-20

-40

Test data
-60
Setzler and Sezen
ASFI approach
-80
-1.5 -1 -0.5 0 0.5 1 1.5
Shear displacement (in.)

Figure 5.19: Lateral load shear displacement relationship for Specimen-4

80

60

40
Lateral load (kip)

20

-20

-40

Test data
-60
Setzler and Sezen
ASFI approach
-80
-6 -4 -2 0 2 4 6
Displacement (in.)

Figure 5.20: Lateral load displacement relationship for Specimen-1

126
80

60

40

20
Lateral force (kip)

-20

-40

-60
Test data
-80 Setzler and Sezen
ASFI approach
-100
-5 -4 -3 -2 -1 0 1 2 3 4 5
Displacement (in.)

Figure 5.21: Lateral load displacement relationship for Specimen-2

80

60

40
Lateral force (kip)

20

-20

-40

Test data
-60
Setzler and Sezen
ASFI approach
-80
-4 -3 -2 -1 0 1 2 3 4 5
Displacement (in.)

Figure 5.22: Lateral load displacement relationship for Specimen-3

127
80

60

40

Lateral force (kip) 20

-20

-40

Test data
-60
Setzler and Sezen
ASFI approach
-80
-6 -4 -2 0 2 4 6 8
Displacement (in.)

Figure 5.23: Lateral load displacement relationship for Specimen-4


80

60

40
Lateral load (kip)

20

-20

-40

Test data
-60
Response with Lp=h/2
Response with Lp=h
-80
-6 -4 -2 0 2 4 6
Displacement (in.)

Figure 5.24: Effect of plastic hinge length on total response by ASFI approach

(Specimen-1)
128
80

60

40
Lateral load (kip)
20

-20

-40

Test data
-60
Response with Lp=h/2
Response with Lp=h
-80
-5 -4 -3 -2 -1 0 1 2 3 4 5
Flexural displacement (in.)

Figure 5.25: Effect of plastic hinge length on flexural response by ASFI approach

(Specimen-1)

129
CHAPTER 6

DEVELOPMENT AND IMPLEMENTATION OF THE PROPOSED MODEL

6.1 Introduction

In Chapter 5, analytical procedures of displacement component model and ASFI

approach were investigated and predicted responses were compared with experimental

test data. Based upon the observations made in the last chapter, a new analytical

procedure is proposed, which is expected to perform better in predicting lateral load-

displacement response of reinforced concrete columns subjected to lateral loads. The

proposed procedure retains features from both of the understudy approaches that help

improve response estimation and make analytical procedure relatively simple and easy to

implement.

In the proposed procedure, flexural and shear deformations are calculated

independently while considering interaction between these mechanisms, and then

combined together depending upon dominant failure mode to obtain total response. The

interaction between axial-flexure and axial-shear mechanisms in terms of compression

softening and axial deformation allows for accurate response estimation while decoupled

flexural analysis minimizes iterations within the analysis and make the process relatively

simple and easy. In addition, buckling of compression bars is also incorporated by

130
considering separate stress-strain relationships for reinforcing steel in tension and

compression.

This chapter presents the development of the proposed analytical procedure by

explaining the details of the component deformation models, interaction methodology

and rules for total response estimation. In order to consider the possibility of compression

bars to buckling under high compressive strains, available steel buckling models are

evaluated and an appropriate model is recommended for use in the proposed model.

6.2 Desired Characteristics of the Proposed Analytical Procedure

It must be acknowledged that in a fixed-ended reinforced concrete column

subjected to lateral loading, flexure and shear mechanisms interact with each other and

corresponding deformations do not occur independently. Previous studies established that

principal compressive stress in the concrete is not only the function of principal

compressive strain but also gets affected by the coexisting principal tensile strain in an

inverse proportion (Vecchio and Collins, 1986). It implies that the concrete subjected to

combined normal compressive and shear stresses is weaker in compression than the

concrete subjected to normal compressive stresses only. In the web of a laterally loaded

reinforced concrete column, the concrete is also subjected to shear stress in addition to

the normal stresses due to axial load and flexure. Due to applied shear stress, the concrete

cracks diagonally and becomes weaker and softer in axial compression. Similarly,

flexural deformations also influence shear behavior of the column. Axial deformation due

to flexure increase width of the shear crack on tension face and hence lower shear

capacity of the column.

131
Any analytical model that aims to capture total response of an element, such as

beam or column, must consider interaction between axial, flexure and shear mechanisms.

In conventional flexure section analysis, the effect of shear stress on degrading

compressive strength of the concrete is ignored and concrete behavior is simulated by its

response in uniaxial compression. The influence of shear mechanism on flexural behavior

can be taken into account by considering compressive stress-strain relationships of

diagonally cracked concrete in flexural section analysis instead of employing constitutive

relationship for concrete in uniaxial compression. Likewise, the effect of flexural

deformations on shear mechanism can also be accounted for by adding axial deformation

component of flexural mechanism to the axial strain of the axial-shear mechanism.

ASFI approach considers interaction between axial, flexure and shear mechanisms

as explained earlier in Chapters 3 and 4. The interaction procedure of the approach is re-

summarized here for easy reference. The ASFI approach employs cracked concrete

behavior in flexural analysis to acknowledge the effect of shear deformations on flexural

performance. This is done by determining compression softening factor from axial-shear

model (through MCFT analysis of shear element) and employing it into axial-flexure

model (through flexural section analysis) to lower uniaxial compressive stress of the

concrete. Similarly, in order to consider the effect of flexural deformation on shear

behavior, ASFI approach incorporates average centroidal axial strain from axial-flexure

model into axial-shear model. While these interactions are taken into consideration,

equilibrium of axial and shear stresses from axial-flexure and axial-shear models are

satisfied for each incremental load step.

132
In previous chapter, the comparison between predicted responses by ASFI

approach and experimental test data was presented. For flexural deformations, predicted

responses (Figures 5.8 through 5.11) show a good correlation with experimental data

indicating that ASFI procedure of considering interaction of compression concrete

softening is effective in capturing effect of shear deformations on flexural performance.

Similarly, predicted shear deformations up to peak load compares well with the observed

shear responses (Figures 5.16 through 5.19), indicating that procedure of axial strain

interaction is also successful in capturing effect of flexural deformations on shear

response. Therefore, it can be concluded that ASFI methodology is effective in capturing

interaction between flexural and shear deformations and must be retained for the

proposed analytical procedure.

It is important to highlight here that identical flexural responses up to peak were

predicted by displacement component model in previous chapter even without

considering interaction between flexural and shear mechanisms. As already explained

that effect of shear deformation on flexural performance is taken into account by

employing compression softening factor from shear model to flexural analysis. For the

columns dominated by flexure or flexure-shear behavior with low shear stresses, it is

most likely that shear strains will not effect flexural deformations. The test specimens

modeled in the Chapter 5 exhibited flexure dominant response up to peak load. The

compression softening factor by ASFI approach was found to be 1.00, eliminating the

need to soften concrete compressive response in flexure. As a result, constitutive law for

concrete in uniaxial compression is employed in ASFI approach, leading to obtain

identical flexural responses by both approaches. It must, however, be noted that if

133
compression softening factor for any column during the analysis by ASFI approach is

found less than 1.00, a significantly different flexural deformations shall be predicted by

ASFI approach and displacement component model.

It must also be noted that the interaction terms in ASFI approach couple axial-

flexure and axial-shear models with each other in such a way that the analysis for any of

the two cannot be performed independently. For example, interaction of shear

deformation with flexural behavior requires determination of compression softening

factor from shear mechanism, coupling axial-shear model with axial-flexure. Similarly,

axial-shear analysis cannot be performed unless axial deformations due to flexure are

determined from axial-flexure mechanism. Thus, the analyses for both mechanisms must

be carried out simultaneously at each loading step. Due to coupling or interlocking of the

mechanisms and requirement on satisfying equilibrium of shear stresses in overall

analysis, the analytical process in ASFI approach is complicated and computationally

intense and solution technique requires a deliberate iteration process.

On the other hand, flexural and shear deformations in displacement component

model are determined independently in a relatively simple analytical procedure. Despite

having limitation on interaction, decoupled analyses methodology of displacement

component model offers the advantage of conducting simple calculations that are easy to

implement. Hence, any analytical process that aims to consider interaction between

flexure and shear can be simplified significantly if analyses for both mechanisms can be

performed independently in decoupled fashion. In view of the need to consider

interaction between flexure and shear deformations in a simplified analytical procedure,

an effort is made in this study to propose a suitable analytical model that performs

134
decoupled flexural and shear analyses offering easy implementation while still

considering interaction between them.

6.3 Proposed Interaction Methodology

As describe above, the proposed procedure considers the interaction between

flexure and shear mechanisms in a manner that flexural section analysis can be performed

independent of shear analysis. Axial strain and shear stress determined from flexure

analysis can then be incorporated to axial-shear model to determine shear deformations.

The proposed interaction methodology is presented as follows.

6.3.1 Interaction of Concrete Compression Softening

One of the objectives of the proposed procedure is to conduct decoupled flexure

and shear analyses while considering interaction between them. If the effect of shear

deformations on flexural performance can be assessed without performing full axial-shear

analysis of the element, it will decouple axial-flexure analysis from axial-shear

mechanism. This will allow flexural deformations to be determined independently

without having the need to perform shear analysis prior to flexure analysis while still

incorporating the effect of shear deformations on flexural performance. Thus, knowing

axial strain due to flexure from flexure analysis, shear analysis can be carried out

considering the interaction of axial deformations. This will make overall analytical

process less complicated and will result in comparatively simple calculations. The effect

of shear deformations on flexural performance is incorporated by lowering concrete

stresses in uniaxial compression by compression softening factor. Hence, compression

135
softening factor must be determined in a simplified way without performing full shear

analysis to represent degradation in concrete strength due to applied shear stresses.

In order to simplify the computation intensive analytical procedure of ASFI

approach, Mostafaei and Vecchio (2008) formulated a new model called Uniaxial Shear

Flexure Model (USFM) by eliminating complex iteration process of the shear model

employed within ASFI approach. In the new approach, single concrete stress block is

employed to represent concrete stress distribution across the cross section depth instead

of fiber section approach. In addition, the axial strain and principal tensile strain of the

element are determined based on average centroidal axial strains and average concrete

strains corresponding to resultant compressive forces of the compression blocks of two

end sections. This simplifies the analytical process significantly and eliminates the

iteration process required for shear analysis in original ASFI approach. The complete

details on formulation, implementation and verification of USFM approach can be found

in Mostafaei and Vecchio (2008).

It must be recalled that the compression softening factor, as defined by Equation

3.2, is a function of concrete principal tensile strain c1 of the element being analyzed. As

USFM approach employs simplified procedure to calculate concrete principal tensile

strain, it is adopted in the proposed procedure with minor modifications. The approach is

based on few fundamental equations of MCFT and two simplifying assumptions on

concrete strains from flexural section analysis and does not require a deliberate shear

analysis (MCFT) of the element. The procedure to determine principal tensile strain and

subsequently compression softening factor is explained below for a fixed ended

reinforced concrete column subjected to in-plane lateral load (Figure 6.1).


136
As already described in Section 3.3.1, compression softening factor , is defined

as

1
1.0 (6.1)
c1
0.8 0.34
co

where c1 is principal tensile strain and co is concrete strain corresponding to the

maximum concrete cylinder strength. For an element considered between inflection point

and one of the end sections of the column, c1 can be determined according to MCFT

equation as follows

c1 x y 2 (6.2)

In above equation, x is average axial strain at the centroid for the element and is

obtained by averaging the values of centroidal axial stain at one of the end section o and

axial strain of the inflection point xa as

x
o xa (6.3)
2

Likewise 2 is concrete principal compressive strain for the element. Its value is

assumed to be the average of the concrete compressive strain corresponding to resultant

compressive force of the stress block at end section c and axial strain at the inflection

point xa . Hence,

2
c xa (6.4)
2

137
The other unknown quantity in Equation 6.2 is strain of the transverse

reinforcement y . Its value is determined from following relationship determined based

on MCFT as

y b2 c b (6.5)

where,

f c1
b 2
2 y Esy 2

c
x 2 fc1 fcx fc1 2
y Esy
f cx f x x f sx

where y is transverse reinforcement ratio, Esy is modulus of elasticity of transverse

reinforcement, f cx is concrete stress in x -direction, f x is applied axial stress, x is

longitudinal reinforcement ratio, f sx is longitudinal steel stress obtained from section

analysis based on average centroidal strain, fc1 0.145 fco is concrete principal tensile

stress, f co is concrete cylinder strength, x is normal strain at the centroid determined

from Equation 6.3 and 2 is concrete principal compressive strains determined from

Equations 6.4.

After calculating concrete principal tensile strain c1 from Equation 6.2,

compression softening factor is determined with the help of Equation 6.1 for a given

curvature. The concrete compressive stresses in flexural section analysis are then lowered

by multiplying concrete stresses given by employed constitutive law with compression

softening factor. The softening of the concrete compressive stress-strain relationship

138
explained in Section 3.3.1 and Figure 3.1. In proposed procedure, compression softening

factor determined with this method is employed till peak flexural load and then a constant

value, corresponding to peak load, is used in post peak analysis.

6.3.2 Interaction of Axial Strains

In order to consider the effect of flexural deformations on shear behavior, the

proposed procedure incorporates axial strain and shear stress due to flexure into in-plane

analysis of the shear element. The proposed procedure is based on the axial strain

interaction methodology of ASFI approach and equilibrium of shear stresses in flexural

and shear mechanisms. In this procedure, interaction of axial strain is taken into account

by adding flexibility component of axial deformation due to flexure to the corresponding

flexibility component of axial-shear model. By employing flexural shear stress to in-

plane stress-strain relationship of the shear element, shear deformations are determined.

For a fixed ended column subjected to lateral load, the procedure for axial deformation

interaction and determination of shear strain is explained as follows.

The length of the column between inflection point and one of the end sections is

considered as a shear element subjected to constant normal stress due to applied axial

load and average shear stresses due to applied lateral load. Performing flexural analysis

on fiber model of the end section and axial model at the inflection point, average axial

strain due to flexure at the centroid of the section for the element can be determined with

following equation. These analyses are carried out incorporating compression softening

factor determined as per the procedure explained above.

139
xf
o xa (6.6)
2

where o and xa are centroidal and axial strain at the column end section and inflection

point, respectively (Figure 6.1). The flexibility component for axial deformation due to

flexure f xf can be determined as

xf
f xf (6.7)
x

where, x is applied axial stress in longitudinal direction of the column and can be

determined by dividing applied axial load P by the area of the cross section

P
x (6.8)
bd

In above equation, b and d are width and depth of the cross section,

respectively. A stress-strain relationship in terms of flexibility matrix for an in plane

shear element (axial-shear model) can be defined as,

f11 f12 f13 x x



f 21 f 22 f 23 y y (6.9)
f
31 f 32 f 33 xy xy

where fij i, j 1, 2,3 are flexibility components of in plane shear model, x is normal

applied stresses in longitudinal direction, y is normal stress in transverse directions, xy

is shear stress, x is normal strain in axial direction, y is normal strain in transverse

direction, and xy is shear strain. In the above equation, longitudinal axis of the shear

element is aligned with horizontal x -axis.

140
Axial strain due to flexure xf can be taken into account in the axial-shear model

by adding flexibility component obtained from Equation 6.7 into Equation 6.9.

f11 f xf f12 f13 x x xf



f 21 f 22 f 23 y y (6.10)

f 31 f 32 f 33 xy xy

In above equation, stresses in transverse direction (clamping stresses) are zero due

to inexistence of lateral external force along the column, i.e., y 0 . In addition, the

applied shear stress xy of the element is taken from flexural section analysis as

M
xy (6.11)
Linbd

where M is the moment obtained from fiber section analysis of the end section and Lin

is length of the element, taken equal to half of the total length for fixed ended column. In

Equation 6.10, knowing the applied stresses, corresponding strains can be calculated. The

flexibility matrix is obtained by inverting material stiffness matrix of the shear element

formulated using secant stiffness methodology as explained in Section 3.4.3.

6.4 Buckling of Compression Bars

The comparison of the responses for flexural deformation in previous chapter

(Figures 5.8 through 5.11) highlighted the need to consider reduction in stresses of the

compression bars due to possibility of their buckling, especially in post peak stage. The

post peak flexural response by ASFI approach compared well with the experimental

behavior for the modeled test specimens, indicating that bar buckling model employed is

effective in capturing the column behavior in flexure. The details of this model were
141
given in Section 5.2.5. In addition to the bar buckling model employed for response

estimation by ASFI approach in previous chapter, two other models are explored to see

their applicability and effectiveness in the analysis. Based upon the analysis results, one

of the models is recommended to be used for response estimation by proposed procedure.

The details on the bar buckling phenomenon, related aspects and proposed bar buckling

model are explained in the following sections.

Reinforced concrete columns in seismic zones are subjected to combined action

mainly due to axial, bending moment and shear forces. Reinforcing bars may experience

inelastic axial compression under severe loading and exhibit lateral deformation known

as buckling. The behavior in the compressive face of a concrete member at overload

depends on a variety of factors such as, size and shape of the cross-section, the amount of

longitudinal compression steel, the amount of transverse reinforcement providing

confinement to the section, thickness of the cover concrete, and stress-strain properties

for the steel and concrete (Potger et al. 2001). The tendency for the compressively loaded

steel bars to buckle and deflect outwards is initially resisted by the lateral restraint

provided by the surrounding cover concrete as well as the transverse steel ties or stirrups.

As the compressive loads increase and approach the section capacity, the concrete

surrounding the compressive bars carries large longitudinal compressive stress, and

eventually becomes prone to longitudinal cracking, and spalling. After the cover concrete

spalls off, stirrups restrain lateral movement and buckling.

The effect of buckling on the load-deformation behavior of reinforcing bars has

been studied under monotonic and cyclic loads (Monti and Nuti (1992), Rodriguez et al.

(1999), Suda et al. (1996), Gomes and Appleton (1997), and Dhakal and Maekawa

142
(2002)). These researchers developed steel constitutive relationships to take into account

the buckling of longitudinal reinforcement, especially focusing on post-buckling load-

deformation behavior.

In this study, two of the available bar buckling models are evaluated to assess

their applicability, accuracy and ease of implementation in the analysis. Flexural

deformations shall be determined by employing these models for the four specimens

tested by Sezen (2002). After comparing predicted responses with test data, appropriate

model shall be employed for analysis.

6.4.1 Dhakal and Maekawa Bar Buckling Model

Dhakal and Maekawa (2002) developed a model to study buckling mechanism of

the reinforcing bars by finite element microanalysis using fiber technique. Their model is

illustrated in Figure 6.2. According to the model, an intermediate point * , * is

established, after which a constant negative stiffness equal to 0.02 Es is assumed until the

average steel stress becomes equal to the residual value of 0.2 f y . The intermediate point

is defined with the help of following set of equations

fy L
* 5.5 2.3 y 7 y
100 db

(6.12)
f L *
* 1.1 0.016 y f s 0.2 f y
100 d
b

where f y is yield stress of longitudinal steel in MPa, L is length of the column which is

taken equal to stirrup spacing, d b is longitudinal bar diameter, y is yield strain of

143
longitudinal bars, f s* is point wise stress on original curve corresponding to * , and is

a factor related to strain hardening region of steel stress-strain relationship. Its value is

1.00 for linear hardening bars and 0.75 for perfectly elastoplastic bars. In modeling test

specimens with non-linearly hardening bars, this factor was assumed 1.00 for

implementing this model in this study. After determining the intermediate point, stress-

strain relationships are then defined as follows

* s y
f sl 1 1 * * f s for y s *

f s y
(6.13)

f sl * 0.02 Es s * 0.2 f y for s *

where f sl is lowered compression steel stress corresponding to steel strain s , f s is point

wise stress corresponding to s , and E s is modulus of elasticity of reinforcing bar.

6.4.2 Potger et al. (2001) Bar Buckling Model

Potger et al. (2001) developed a simplified bar buckling model by using a

predetermined buckling strain, softening slope and stress degradation after modifying

some of the existing models (Kato et al. (1973), Meng et al. (1992), Nakatsuka et al.

(1999), Inoue and Shimizu (1988), Yamada et al. (1993)). The model is illustrated in

Figure 6.3. In this model, buckling and lateral displacement of the bar is assumed to

begin at a critical buckling strain lb indicated by point A in Figure 6.3. After buckling,

the compressive strain in the bar continues to increase and the stress is assumed to fall-off

progressively with a slope of lb Es . Softening continues until a second limiting strain ps

144
is reached at point B. With further increases in compressive strain (past point B), the

stress in the bar is assumed to soften at a less severe rate of 0.005Es .

The critical buckling strain is calculated as follows by the relationships developed

by Nakatsuka et al. (1999). This relationship takes into account the effects on buckling

sh
strain due to lateral reinforcement spacing to confined core diameter ratio , confining
dc

stress, yield strength of lateral reinforcement f yy , shape of reinforcement (circular,

rectangular), and the compressive strength of plain concrete f co .

lb c0 fco 103 f1 f2 f3 f 4 f5 (6.14)

where,

3.6 4.8( sh d c ) 0.1 ( sh d c ) 0.75


f1
0 ( sh d c ) 0.75

f 2 y f yy
2

1.0 for bar in circular column



f3 0.9 for corner bar (6.15)
0.18
for intermediate bar
110
-1 30 MPa f c 0 110 MPa
f4 fc0
0 f c 0 110 MPa

600
f5 0.5 102 f yy 400 MPa
f yy

where co is maximum strain corresponding to peak strength of the plain concrete f co

(MPa), sh is spacing of the transverse reinforcement, d c is smallest side length of

concrete cross section surrounded by lateral reinforcement, y is transverse

reinforcement ratio, and f yy is yield strength of the transverse reinforcement.


145
The post-buckling slope is calculated with following relationship developed by

Inoue and Shimizu (1988). The second post-buckling slope is set at 0.005Es adopted

from the report by Yamada et al. (1993) for the buckling of steel plates.

1
lb 100 yx 1
1 500
2
(6.16)
sh

ir

where yx is yield strain of longitudinal bars, is 1.0 for corner bars and 0.5 for

intermediate bars, and ir is radius of gyration of longitudinal bar.

6.4.3 Implementation of Bar Buckling Models

Both of the above models and bar buckling model proposed in the previous

chapter are incorporated in flexural section analysis and flexural deformations are

calculated for Specimen-1, 2 and 4. The predicted lateral load-flexural displacement

relationships are compared with experimental responses in Figures 6.4 through 6.6.

6.4.4 Proposed Model for Buckling of Compression Bars

Based on the comparison of the predicted and observed responses, the model

explained in Section 5.2.5 is recommended to be used for lowering the compression

stresses of the longitudinal reinforcement to account for bar buckling phenomenon. This

model is also explained Figure 6.7.

It must be noted that diameter of the longitudinal bar and spacing of the transverse

reinforcement are important parameters that affect buckling of the compression bars

146
(Monti and Nuti 1992). Smaller diameter bars contained by widely spaced stirrups are

most likely to undergo lateral deformations and buckling much earlier during loading

history than larger diameter bars confined by closely spaced transverse reinforcement.

Therefore, in the proposed model to account for this fact, for stirrup spacing to bar

diameter ratio sh db of less than 5.00, no buckling is considered and compressive

behavior of the reinforcement is similar to tensile behavior. For sh db ratio above 11.00,

the bars are considered to buckle as soon as reinforcement yields. For sh db ratios

between 5.00 and 11.00, post-buckling softening is considered soon after yielding by

employing the proposed model.

6.5 Flexural Analysis

Conventional fiber section analysis is performed to determine moment-curvature

relationship required to calculate lateral load-flexural displacement response. However,

concrete behavior in tension, compression softening factor determined as per procedure

described in Section 6.3.1 and buckling of compression bars as per procedure explained

in Section 6.4.4 are incorporated in the flexural analysis. Flexural section analysis also

enables to determine axial strain due to flexure xf and shear stress of the axial-flexure

model xy required for shear analysis of the element.

Lateral displacements due to flexure are calculated by plastic hinge model with

the help of following equations. This model is already explained in Section 4.3.3 and

Figure 5.4.

147
1
f a2 for y
3
(6.17)
Lp
f y a 2 y L p a
1
for y u
3 2

where is curvature at column end, y is curvature at yield point, a is shear span equal

to length L of a cantilever column and half the length for a fixed ended column. The

plastic hinge length L p is taken as one-half of the section depth.

The lateral load-flexural displacement relationships for Specimens 1, 2, 3 and 4

are calculated using above mentioned procedure and are presented in Figures 6.8 through

6.11, respectively.

6.6 Reinforcement Slip Deformation Model

In Chapter 5, lateral load-slip displacement relationships for modeled test

specimens were obtained by employing Sezen and Setzler (2008) bar slip model. The

predicted responses showed satisfactory correlation with experimental data (Figure 5.12

through 5.15). Hence, this model is adopted to calculate bar slip deformations in the

proposed procedure. This model is explained in detail in Section 2.2.2 and Figures 2.3,

2.4. The lateral load-slip displacement relationships for Specimens 1, 2, 3 and 4 are

calculated using above mentioned procedure and are presented in Figures 6.12 through

6.15, respectively.

6.7 Shear Deformations

It was shown in the previous chapter that shear response envelope of displacement

component model performs well in predicting shear behavior of the test specimens
148
(Figure 5.16 through 5.19). For pre-peak response, this shear model relies on shear strain

output from Response-2000 computer program. The user is required to input section

configuration, material properties and loading conditions to the program and run full

member analysis. The analysis results in overall lateral load-deformation response up to

the peak load the cross section can sustain. At each load step, the program gives average

shear strain distribution along the length of the column. In order to obtain lateral load-

shear displacement response, the user is required to manually copy this shear strain

distribution at each load step and integrate it over the length of the column to get

corresponding shear displacement. This procedure is repeated for all load steps up to the

peak load, which makes the process very laborious and time consuming. Additionally, as

the load-displacement relationships for shear and flexure/bar slip mechanisms are

calculated separately at different load steps, these responses are required to be

interpolated to a common set of load steps and corresponding displacements to make

their addition possible for total response.

On the other hand, axial-shear model of ASFI approach adopts relatively simple

approach for shear response estimation with reasonably good accuracy. This procedure

enables the interaction of axial strains to capture effect of flexural deformation on shear

behavior. Nevertheless, ASFI approach has a serious limitation of predicting shear

response only up to the peak load and does not predict shear strength degradation with

increasing deformations. This aspect is addressed of by shear response envelope of the

displacement component model that is able to predict post peak behavior up to the point

of axial load failure.

149
The proposed shear displacement model in this study retains the features from

ASFI approach and displacement component model that help estimate response

accurately and make analytical process simple and relatively easy to implement. In the

proposed shear model, shear displacements up to the peak shear load are calculated by in-

plane analysis of the shear element based on direct application of MCFT while adding

axial strain due to flexure to the total axial strain of the shear element. The post-peak

shear behavior is obtained by employing post peak shear response envelope of the

displacement component model.

The procedure for application of MCFT was explained in Section 4.2.3. These

steps along with axial strain interaction methodology explained in Section 6.3.2 are

followed to obtain flexibility matrix in Equation 6.10. This equation is then solved with

applied normal stress due to axial load (Equation 6.8) and shear stress from axial-flexure

model (Equation 6.11) to obtain shear strains up to peak shear load. The shear strength is

then assumed to remain constant at its peak value until the point of shear strength

degradation. From this point, the response degrades linearly to the point of axial load

failure, where strength is zero. The shear failure point and axial load failure point are

calculated according to the procedure explained in Section 2.2.3 and 2.3.

The term peak shear load used above refers to point where pre-peak shear

response by MCFT terminates. This corresponds to the load step at which either shear

failure occurs or peak flexural strength reaches before shear failure. Hence, for the

columns failing in shear prior to experiencing flexural strength, peak shear load will be

the shear strength of the specimen. For the columns having higher shear strength than the

flexural strength, this point will give shear response at peak flexural strength.

150
As the shear analysis in the proposed procedure is performed at the same load

steps employed for flexure and bar slip deformations, lateral load-shear displacement

relationships are not required to be tailored to match the loading intervals of the

flexure/slip load-displacement relationships. This allows for simple addition of all

deformation components for post peak response. The proposed shear model also

eliminates the dependency of displacement component model on Response-2000 and

gives control to the user on choice of material constitutive models and parameters to be

used other than the ones employed in Response-2000.

The lateral load-shear displacement relationships for Specimens 1, 2, 3 and 4 are

calculated using proposed shear procedure and are presented in Figures 6.16 through

6.19, respectively.

6.8 Total/Combined Response

The total lateral response of a reinforced concrete column to the lateral load is

modeled as a set of springs in series (Figure 2.6). The lateral displacement components of

flexure, bar slip and shear are each modeled by a spring. Each spring is subjected to the

same force and total displacement is sum of response of each spring.

The analysis in the proposed procedure is carried out by incrementing load in

steps of appropriate size starting with a small initial value. First, flexural and bar slip

deformations are calculated and then shear response is evaluated at each load step. If at

any load step, shear failure does not occur, the value of shear strain is taken as shear

response at current load level and the shear analysis for the next step is carried out. This

procedure is repeated until either shear failure occurs or peak flexural strength is

151
achieved. The load corresponding to this stage is defined as peak strength. After reaching

the peak, the mechanism limiting the peak strength (flexure or shear) will dominate the

behavior.

6.8.1 Pre-Peak Response

In the proposed procedure, total response up to the peak strength is obtained by

simply adding deformation components due to flexure, bar slip and shear mechanism

determined as per the individual component models explained above.

6.8.2 Post-Peak Response

The rules for combining deformation components for post peak responses in the

proposed procedure are taken from displacement component model. The procedure for

classification of the columns into categories is same as before except that the

determination of shear strength of the specimens is slightly different. It was shown in the

previous chapter that the procedure employed for post peak response in displacement

component model is much realistic and performs better than ASFI approach (Figures 5.20

through 5.23).

The classification of the columns into categories is based on the comparison of

their shear, flexural and yield strength (Setzler, 2005). The yield strength Vy of the

column is defined as the lateral load corresponding to the first yielding of the tension bars

in flexural analysis. The flexural strength V p is the lateral load corresponding to the

peak moment sustainable by the column. The shear strength Vn for the columns failing

152
in shear prior to the reaching flexural strength or failing close to flexural strength is

determined from the proposed shear model in Section 6.7, in which peak shear load is

shear strength of the column. For other columns where peak shear load by proposed shear

model is equal to the flexural strength, Sezen and Moehle (2002) equation is employed to

determine shear strength Vn of the column. The determination of the shear strength

with this procedure minimizes the dependency of the component displacement model on

alternate mean of determining shear strength for classification purpose. The procedure for

combining deformations due to flexure, bar slip and shear are explained in Section 2.3

and are followed as such for post peak behavior.

6.9 Analytical Procedure and Solution Technique for Proposed Procedure

This section summarizes the procedure for estimating the response of laterally

loaded reinforced concrete columns by proposed procedure.

1. Define or input material properties and geometry. Decide on a sign convention for

tensile and compressive stresses and strains.

2. Input axial load P . If there is no applied axial load, then consider a negligible

small value for P .

3. Consider an axial-flexure element of reinforced concrete column from its

inflection point to one of the end section.

4. Select a small value of the curvature , such as 0.00000001 , as a starting

value.

5. Determine an estimate of compression softening factor

153
i. Assume location of the neutral axis c for the first iteration. An initial

h
value of c is good starting point, where h is overall section depth.
2

ii. Apply section analysis procedure at end section and inflection point and

determine centroidal axial stain at end section o , strain corresponding to

resultant compressive force of the stress block at end section c , and axial

strain at the inflection point xa . Refer to Figure 6.1 for better

understanding.

iii. Determine average centroidal strain x and average concrete principal

compression strain 2 for the axial-flexure element using Equations 6.3

and 6.4, respectively.

iv. Determine the average concrete principal tensile strain c1 using

Equations 6.5 and 6.2. The average value of concrete principal tensile

stress fc1 0.145 fco can be used.

v. Calculate compression softening factor using Equation 6.1. Determine

concrete compression stress of the stress block with the help of employed

constitutive law and multiply it with to represent cracked concrete

behavior.

vi. Check for the section equilibrium at the end section. If equilibrium of

axial forces is not achieved, modify the assumed value for depth of the

neutral axis c and repeat steps ii through vi.

154
6. Evaluate axial-flexure mechanism by performing flexural section analysis at the

end section and inflection point.

i. Divide column cross section into appropriate number of concrete fibers.

Apply standard fiber section analysis procedure and obtain strain

distribution across the section depth.

ii. Employ appropriate constitutive laws for compressive behavior of

confined core concrete and unconfined cover concrete, and determine

stress distribution corresponding to the imposed strain profile. Effect of

confinement due to transverse reinforcement on compressive strength and

ductility of the core concrete must be taken into account with the help of

appropriate confinement model.

ii. Multiply compressive stresses of the concrete by compression softening

factor obtained in step 5 above to consider effect of shear deformations

on flexural performance.

iv. Employ appropriate constitutive law to consider concrete behavior in

tension in section analysis.

v. Employ appropriate constitutive law for reinforcing bars and determine

steel stresses corresponding to the section strain at each layer. For

considering effect of compression bars buckling, employ proposed bar

buckling model explained in Section 6.4.4.

vi. Determine forces in each concrete fiber and reinforcing bars by

multiplying stresses by corresponding areas and check for section

155
equilibrium. When section equilibrium is achieved, calculate nominal

moment capacity M and obtain axial centroidal strain o .

vii. Determine flexural-shear stress xy by Equation 6.11.

viii. Perform axial section analysis at the inflection point and determine axial

strain due to applied axial load only xa considering same material

constitutive laws employed for flexural section analysis of the end section.

Calculate axial strain due to flexure xf with Equation 6.6.

ix. Calculate lateral displacements due to flexure f by plastic hinge model

with the help of Equation 6.17. Plastic hinge length is taken equal to one

half of the section depth.

x. Calculate reinforcement slip displacement pull by procedure explained in

Section 6.6. Lateral displacement due to slip is added to flexural and shear

displacements to obtain total response.

xi. Determine x and f xf from Equations 6.8 and 6.7, respectively.

xii. Determine lateral load V corresponding to the nominal moment capacity

M by following equation

M
V (6.18)
Lin

xii. Increment curvature to next load step and repeat all steps until either

concrete crushes or steel ruptures. After reaching peak flexural load Vp ,

defined as load corresponding to the maximum moment capacity of the

156
section, use constant compression softening factor corresponding to its

value at peak load.

7. Evaluate axial-shear mechanism by proposed procedure employing MCFT and

interaction of axial strains for the column element between end section and inflection

point. The procedure for MCFT is explained in Section 3.4.

i. Start with the first load step from axial-flexural analysis explained above.

Consider the variables of iteration as xi , yi , si and assume some very

small values for each of them for the first iteration. These variables are

average normal strain in x -direction ( xi ), average normal strain in y -

direction ( yi ) and average shear strain ( si ) of axial-shear element,

respectively.

ii. Determine average concrete principal tensile and compressive strains

c1 and c 2 by Equations 3.5 and 3.6, respectively. Also, determine

orientation of principal tensile plane t and principal compressive plane

strains c with the help of Equations 3.7 through 3.9.

iii. Calculate compression-softening factor with the help of Equation 6.1

using c1 calculated in above step.

iv. Determine average steel stresses f sx and f sy corresponding to the

strains x and y in accordance with employed constitutive laws for

reinforcement in x and y -directions respectively.

157
v. Determine average principal tensile and compressive stresses in concrete

by considering same material constitutive laws employed in axial-flexural

model (step 6). However, as in axial-flexure model cover and core

concrete are modeled separately, concrete compressive strength and

corresponding strain must be modified according to the Equation 3.42.

vi. Check local stress-strain conditions at cracks ensuring that the tension in

the concrete can be transmitted across the crack and shear stress on the

surface of the crack does not exceed maximum shear provided by the

aggregate interlock by Equations 3.18 through 3.24. If Equation 3.20 is

not satisfied and local shear stress at the surface of the crack exceeds shear

strength provided by aggregate interlock mechanism, then average

concrete principal tensile stress must be reduced as per Equation 3.25.

vii. Determine secant moduli for concrete and reinforcements and assemble

material stiffness matrices using Equation 3.27 through 3.30. Determine

total material stiffness D matrix by transforming component material

stiffnesses into global x and y -directions using Equation 3.26 and 2.31.

viii. Invert total stiffness matrix D to get flexibility matrix f of Equation 6.9

and add flexibility component of axial deformation due to flexure f xf to

f11 component of f to determine flexibility matrix of coupled axial-

shear-flexure element in Equation 6.10.

158
ix. Consider x from Equation 6.8, y 0 , and flexural-shear stress xy from

step 6(vii) and solve Equation 6.10 for xi 1 , yi 1 , and si 1 for the

integration point.

x. Assess the convergence of the acquired variables xi 1 , yi 1 , and si 1 . If

satisfactory convergence is achieved, note the value of si as shear

response at current load level and go to the next load step. Otherwise,

repeat above steps in shear analysis with improved estimate of the

iteration variables. For the next iteration, average of the initial assumed

values and calculated values in last step can be used for fast convergence.

xi. The above mentioned shear analysis steps are repeated for all load steps

until either shear failure is observed or load step corresponding to peak

flexural load Vp is reached. The load corresponding to this stage is defined

as peak shear load.

x. After the peak shear load is reached, shear strength is then assumed to

remain constant at its peak value until the point of shear strength

degradation. From this point, the response degrades linearly to the point of

axial load failure, where strength is zero. The shear failure point and axial

load failure point are calculated according to the procedure explained in

Section 2.2.3 and 2.3.

8. Obtain combined/total response with the help of procedure explained in Section

6.8.

159
6.10 Implementation of the Model and Comparison of the Results

The proposed procedure is implemented to estimate component and total

responses for Specimen-1 through 4. The load-displacement relationships are calculated

by employing same material constitutive laws as explained in the Chapter 5. The pre-

peak flexural and bar slip deformations are generally predicted well (Figures 6.8 through

6.11). The pre-peak shear stiffnesses and peak shear strengths are also captured well.

However, for total response estimation (Figure 6.20 through 6.23), it can be seen that

displacement component model, ASFI approach and proposed procedure, predict almost

identical responses before peak load. After the peak is reached, response predictions by

these models are quite different. ASFI approach generally underestimates post peak

responses for the test specimens. This is mainly due to shear and bar slip displacements

are calculated from constant secant stiffnesses, which follow straight-line path after the

peak and become zero at axial load failure.

The displacement component model generally overestimates the post peak

responses. When the same columns were modeled by this approach in Chapter 2, very

good correlation was obtained between predicted and observed responses. In Chapter 2,

these columns were analyzed by employing constitutive relationship for compressive

concrete that had steep descending slope. This was done to acknowledge the poor

confinement due to widely spaced ties. In this analysis, the concrete compressive

behavior is simulated by Mander et al. (1988) model for both loading and unloading

branches. The constitutive models employed in Chapter 2 and this study are presented in

Figure 6.24.

160
6.11 Conclusion

A procedure is proposed for response estimation of reinforced concrete columns

subjected to lateral loads. The procedure is based on the observations made in the

previous Chapter 5 after evaluating and analyzing displacement component model and

ASFI approach. The proposed procedure determines flexure, bar slip and shear

deformations individually considering interaction between these mechanisms. The axial-

flexure mechanism is decoupled from axial-shear model, allowing relatively simpler

analytical procedure. The flexure model in proposed procedure incorporates concrete

tensile behavior, interaction of compression softening and buckling of compression bars

into the flexural analysis. The shear model evaluates the shear response of the column

while considering effect of flexural deformation on shear behavior. The pre-peak

response is evaluated by employing MCFT, and post peak shear response envelope from

displacement component model is adopted for predicting post peak shear behavior. All

deformation components i.e., flexural, bar slip and shear are added together to get total

response of the column. The total/combined peak response is limited by lesser of the

shear and flexural strength of the column and limiting mechanism governs the post peak

response. The proposed procedure employs relatively simple calculations for overall

response estimation and appears to perform better than displacement component model

and ASFI approach in predicting the component and total response.

161
Concrete stress block
at end section

P NA
End section V o
c
Centroidal
Lin Axis

Inflection Point
xa
x 2
Assumed distribution
h
of x and 2 over
Concrete stress block length of the element
at inflection point

Fixed-ended column
subjected to lateral load
b

h
Column cross section

Figure 6.1: Determination of concrete principal tensile strain

fs

*
s , f s*
fy

,
* *

0.02 Es
0.2 f y

s
y

Figure 6.2: Dhakal and Maekawa (2002) bar buckling model


162
fs

f s ,lb A

f s , ps

s
y

Figure 6.3: Potger et al. (2001) bar buckling model

80

60

40
Lateral force (kip)

20

-20

-40
Test data
Dhakal et al.
-60
Potger et al.
Proposed model
-80
-3 -2 -1 0 1 2 3
Flexural displacement (in.)

Figure 6.4: Effect of compression bar buckling Specimen 1

163
80

60

40

20
Lateral force (kip)

-20

-40

-60 Test data


Dhakal et al.
-80 Potger et al.
Proposed model
-100
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
Flexural displacement (in.)

Figure 6.5: Effect of compression bar buckling Specimen 2

80

60

40
Lateral force (kip)

20

-20

-40
Test data
Dhakal et al.
-60
Potger et al.
Proposed model
-80
-3 -2 -1 0 1 2 3 4
Flexural displacement (in.)

Figure 6.6: Effect of compression bar buckling Specimen 4


164
Figure 6.7: Proposed bar buckling model

80

60

40
Lateral load (kip)

20

-20

-40

Test data
-60
Setzler and Sezen
Proposed model
-80
-3 -2 -1 0 1 2 3
Flexural displacement (in.)

Figure 6.8: Lateral load flexural displacement relationship for Specimen-1

165
80

60

40

20
Lateral force (kip)

-20

-40

-60
Test data
-80 Setzler and Sezen
Proposed model
-100
-1.5 -1 -0.5 0 0.5 1 1.5
Flexural displacement (in.)

Figure 6.9: Lateral load flexural displacement relationship for Specimen-2

80

60

40
Lateral force (kip)

20

-20

Test data
-40
Setzler and Sezen
Proposed model
-60
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
Flexural displacement (in.)

Figure 6.10: Lateral load flexural displacement relationship for Specimen-3

166
80

60

40
Lateral force (kip)
20

-20

-40

Test data
-60
Setzler and Sezen
Proposed model
-80
-3 -2 -1 0 1 2 3 4
Flexural displacement (in.)

Figure 6.11: Lateral load flexural displacement relationship for Specimen-4

80

60

40
Lateral load (kip)

20

-20

-40

Test data
-60
Setzler and Sezen
Proposed model
-80
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
Slip displacement (in.)

Figure 6.12: Lateral load slip displacement relationship for Specimen-1


167
80

60

40

20
Lateral force (kip)

-20

-40

-60
Test data
-80 Setzler and Sezen
Proposed model
-100
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
Slip displacement (in.)

Figure 6.13: Lateral load slip displacement relationship for Specimen-2

80

60

40
Lateral force (kip)

20

-20

Test data
-40
Setzler and Sezen
Proposed model
-60
-1 -0.5 0 0.5 1 1.5 2
Slip displacement (in.)

Figure 6.14: Lateral load slip displacement relationship for Specimen-3

168
80

60

40
Lateral force (kip)
20

-20

-40

Test data
-60
Setzler and Sezen
Proposed model
-80
-2 -1 0 1 2 3 4
Slip displacement (in.)

Figure 6.15: Lateral load slip displacement relationship for Specimen-4

80
Test data
60 Setzler and Sezen
ASFI approach
40 Propose model
Lateral load (kip)

20

-20

-40

-60

-80
-5 -4 -3 -2 -1 0 1 2
Shear displacement (in.)

Figure 6.16: Lateral load shear displacement relationship for Specimen-1

169
100
Test data
80 Setzler and Sezen

60 ASFI approach
Proposed model
Lateral force (kip) 40

20

-20

-40

-60

-80

-100
-1 -0.75 -0.5 -0.25 0 0.25 0.5
Shear displacement (in.)

Figure 6.17: Lateral load shear displacement relationship for Specimen-2

100

80

60

40
Lateral force (kip)

20

-20

-40 Test data


Setzler and Sezen
-60 ASFI approach
Proposed model
-80
-1 -0.8 -0.6 -0.4 -0.2 0 0.2 0.4 0.6 0.8
Shear displacement (in.)

Figure 6.18: Lateral load shear displacement relationship for Specimen-3

170
80

60

Lateral force (kip) 40

20

-20

-40
Test data
Setzler and Sezen
-60
ASFI approach
Proposed model
-80
-2 -1.5 -1 -0.5 0 0.5 1 1.5 2
Shear displacement (in.)

Figure 6.19: Lateral load shear displacement relationship for Specimen-4

80

60

40
Lateral load (kip)

20

-20

-40
Test data
Setzler and Sezen
-60
ASFI approach
Proposed model
-80
-6 -4 -2 0 2 4 6
Displacement (in.)

Figure 6.20: Lateral load displacement relationship for Specimen-1


171
80

60

40

20
Lateral force (kip)

-20

-40

-60 Test data


Setzler and Sezen
-80 ASFI approach
Proposed model
-100
-5 -4.15 -3.3 -2.45 -1.6 -0.75 0.1 0.95 1.8 2.65 3.5 4.35 5
Displacement (in.)

Figure 6.21: Lateral load displacement relationship for Specimen-2

80

60

40
Lateral force (kip)

20

-20

-40
Test data
Setzler and Sezen
-60
ASFI approach
Proposed model
-80
-4 -3 -2 -1 0 1 2 3 4 5
Displacement (in.)

Figure 6.22: Lateral load displacement relationship for Specimen-3

172
80

60

40
Lateral force (kip)
20

-20

-40
Test data
Setzler and Sezen
-60
ASFI approach
Proposed model
-80
-6 -4 -2 0 2 4 6 8
Displacement (in.)

Figure 6.23: Lateral load displacement relationship for Specimen-4

4000
Mander confined concrete model
3500 Mander unconfined concrete model
Sezen confined concrete model
3000
Sezen unconfined concrete model

2500
Stress (psi)

2000

1500

1000

500

0
0 0.005 0.01 0.015 0.02
Srain

Figure 6.24: Comparison between compressive concrete constitutive laws

173
CHAPTER 7

CONCLUSION

7.1 Summary

The focus of this research was to suggest a suitable analytical procedure that can

accurately predict the monotonic lateral force-displacement relationship for reinforced

concrete columns by overcoming limitations of available models. The columns found in

many existing buildings not designed according to modern seismic design standards,

often have limited deformation capacity and are vulnerable to brittle shear failure during

earthquakes. Observed damage pattern after past earthquakes has shown that columns,

especially those in the first story, are often the most critical members in non-ductile

concrete buildings, and failure of columns can lead to structure collapse. The model

proposed here is intended for use in the analysis of existing buildings to determine the

required level of retrofit necessary for satisfactory seismic performance. The proposed

procedure can also be used for the properly designed existing and future planned

buildings for their performance evaluation.

Lateral deformation of a column is composed of three deformation components

due to flexure, reinforcement slip and shear mechanisms. Theses deformations do not

occur independently and axial, flexure and shear mechanisms interact with each other.

174
For example, shear deformations tend to lower the compressive strength of the concrete

in the web of the column that affect flexural behavior. Likewise, flexural deformations

increase shear crack width and lower shear capacity of the column. In the proposed

procedure, flexural and shear deformations are determined while considering these

effects. The flexural deformations are calculated separately by employing cracked

concrete behavior to account for shear deformations on flexural performance. After axial

strains and shear stresses are determined from flexural analysis, shear behavior is then

evaluated. The shear response estimation in the proposed procedure employs MCFT and

post-peak shear response envelope from displacement component model. Deformations

due to bar slip are calculated separately and added to flexural and shear displacements to

obtain total response. The rules for combining deformation components for total response

are the same as employed by displacement component model.

7.1.1 Component Models

The flexural behavior in the proposed procedure was evaluated by fiber section

analysis techniques. However, in order to incorporate the effect of shear deformation on

flexural behavior, cracked concrete behavior was considered in the section analysis. This

was done by determining compression-softening factor from axial-shear mechanism and

employing it in flexural section analysis to lower compressive stress-strain relationship

for the concrete in uniaxial compression. The compression-softening factor was

determined as per the suggested procedure that decouples flexural analysis from shear

mechanism, thus allowing flexural deformations to be determined independently and

easily. The concrete behavior in tension was also incorporated in the flexural analysis. In

175
order to consider the phenomenon of compression bar buckling, few of the bar buckling

models were analyzed and a procedure to incorporate this effect was suggested. After

knowing moment-curvature relationship with above considerations, flexural displacement

were calculated with plastic hinge model by taking plastic hinge length equal to one-half

of the section depth. Comparison of the predicted flexural response with experimental

data showed that the proposed procedure performs well in estimating flexural

displacements.

The shear model in the proposed procedure employs MCFT and considers effect

of flexural deformations on shear strength. This was done by extracting axial strain due to

flexure only from flexural section analysis as per the procedures employed in ASFI

approach and incorporating it to the shear analysis. The normal and shear stress obtained

from axial-flexure analysis are employed as applied load for in-plane shear analysis of

the element. The response up to the peak load is determined with this procedure and then

post peak shear response envelope from displacement component model was employed to

determine points of shear and axial load failures. The shear model in the proposed

procedure eliminated the dependency of the displacement component model on

Response-2000 computer program for obtaining shear strain distribution. The comparison

of the predicted responses with experimental data indicated that axial strain interaction

methodology, shear response estimation by MCFT and post peak shear response envelope

were successful in capturing observed shear behavior.

In addition, lateral displacements due to reinforcement slip were calculated

separately by Setzler and Sezen (2008) bar slip model and were added to flexural and

shear displacements at each load step.

176
7.1.2 Combined Model

The overall lateral deformation of a reinforced concrete column was modeled as

three springs in series, one for each of the deformation components. The shear strength of

the column failing at or prior to reaching flexural strength was calculated using the shear

model of the proposed procedure. For the columns having higher shear strengths than the

flexural strength, shear strength was determined using shear strength equation proposed

by Sezen and Moehle (2004), taking the displacement ductility parameter k as 1.0. This

shear strength was compared to the yield and flexural strengths determined from the

flexural analysis to classify columns into one of five categories. The same rules employed

by displacement component model were used for classification of the columns into

categories. Peak load was defined as lesser of the load predicted by shear model and

flexure model. Up to the peak strength, the three deformation components were simply

added together to obtain total response. After the peak strength was reached, post peak

response was governed by dominant failure mechanism and rules for combining post

peak deformations from displacement component model were employed for obtaining

total response. The comparison of the predicted and observed responses indicated that

proposed procedure performs well in predicting overall response of the column.

7.2 Conclusions

The interaction between flexure and shear deformations must be taken into

account to determine total response of the column to lateral load. The recommended

interaction methodology for considering effect of shear deformation on flexural

177
performance and procedure to calculate compression softening factor appeared to work

well in this research. Decoupled flexural analysis allows calculating flexural

deformations accurately and independently while still considering the interaction with

shear mechanism.

Buckling of compression bars affect post peak behavior and hence must be

considered in the analysis. The procedure adopted in this research for incorporating this

phenomenon worked well and is recommended to be used for the analysis.

Moment-curvature analysis employing fiber model approach while considering

concrete behavior in tension, compression-softening factor, and buckling of compression

bars, and the plastic hinge method of calculating flexural deformations worked well in

this research. These tools appear to be appropriate for use in both well reinforced and

lightly reinforced concrete columns.

The reinforcement slip model by Setzler and Sezen (2008) which utilizes a bi-

uniform bond stress over the embedded length of the bar, represented the experimental

slip behavior well and can be used to model the response with least complexity and good

accuracy.

The shear behavior of columns was predicted well by the shear model employed

in the proposed procedure. The Modified Compression Field Theory proved to be an

effective method for predicting shear displacement, up to the maximum strength of a

member. The procedure to analyze shear behavior of the column while considering

interaction of axial strains and applied shear stresses from flexural model resulted in

fairly accurate response estimation and was simple to implement. Only little experimental

shear displacement data was available. More would be required to determine the

178
effectiveness of the model for a wider range of column types. The post-peak shear

response envelope is essentially required to complete the full load-deformation response.

The classification system and rules to combine deformation component to obtain

total response by displacement component model are very effective in capturing observed

behaviors. Theses rules can be used for wide range of the columns failing in various

modes.

The proposed analytical procedure is relatively easy to implement and employs

simple calculations to estimate component and total deformations. The procedure can be

used to accurately estimate monotonic lateral load-displacement relationships for the

columns exhibiting different failure modes such as flexural failure with very limited or no

shear effects, flexure and/or shear failure following the flexural yielding, and brittle shear

failure prior to flexural yielding etc.

7.3 Recommendations for Future Work

The research reported here presents a model for the monotonic lateral deformation

of lightly reinforced concrete columns subjected to lateral loads. However, there remains

a large amount of work to be completed before the behavior of non-ductile concrete

structures in earthquakes can be fully modeled and understood. Several recommendations

for continued research on this topic follow.

The proposed model was used to predict the behavior of few test columns. While

general good agreement was shown between the test data and model, further comparisons

should be made to complete validation of the model over a wide range of column

properties.

179
Several areas warrant more experimental research to improve the knowledge base.

Particularly, more testing of reinforced concrete columns to axial load failure is

necessary. Most test data available stops before loss of axial capacity. More data would

help improve models for shear capacity and axial capacity; both of these are necessary

components of the proposed model. Additionally, more test specimens should be

instrumented to monitor the individual displacement components, particularly shear

displacements. This would aid in improving the component models, which should lead to

improvements in the combined model as well.

The most significant area of research that remains is the modeling of the

hysteretic behavior of columns. Many hysteretic models for flexural deformations exist,

which should be applicable. Several shear models are available as well, however, these

would need to be evaluated in the context of lightly reinforced columns to determine their

applicability. These hysteretic models would need to be combined through a set of rules

to predict the overall cyclic lateral behavior of lightly reinforced concrete columns.

180
LIST OF REFERENCES

Aboutaha, R. S., Engelhardt, M. D., Jirsa, J. O., and Kreger, M. E., 1999. Rehabilitation
of Shear Critical Concrete Columns by Use of Rectangular Steel Jackets, ACI
Structural Journal. V. 96, No. 1, Jan. 1999. pp. 68-78.

ACI Committee 318, 2002. Building Code Requirements for Structural Concrete and
Commentary, ACI 318-02/318R-02. American Concrete Institute, Farmington
Hills, Mich., 2002. 443 pp.

ACI Committee 408, 1979. Suggested Development, Splice, and Standard Hook
Provisions for Deformed Bars in Tension, ACI 408.1R-79. American Concrete
Institute, Farmington Hills, Mich., 1979. 3 pp.

Alsiwat, J. M. and Saatcioglu, M., 1992. Reinforcement Anchorage Slip under


Monotonic Loading, Journal of Structural Engineering. ASCE, V. 118, No. 9,
Sept. 1992. pp. 2421-2438.

Bentz, E., 2000a. Response-2000. http://www.ecf.utoronto.ca/~bentz/home.shtml.

Bentz, E., 2000b. Sectional Analysis of Reinforced Concrete Members, Ph.D.


Dissertation. University of Toronto, 2000. 310 pp.

Berry, M., Parrish, M., and Eberhard, M., 2004. PEER Structural Performance Database
Users Manual (Version 1.0). Pacific Earthquake Engineering Research Center,
University of California, Berkeley, Jan. 2004. 38 pp. http://nisee.berkeley.edu/
spd/performance_database_manual_1-0.pdf.

Chadwell, C. B. and Imbsen & Associates, 2004. XTRACT Cross Section Analysis
Software for Structural and Earthquake Engineering. http://www.imbsen.com/
xtract.html.

Chadwell, C. B. and Imbsen, R. A., 2004. XTRACT: A Tool for Axial Force-Ultimate
Curvature Interactions, Building on the Past, Securing the Future: Proceedings
of the 2004 Structures Congress. ASCE, Nashville, Tenn., May 2004. 9 pp.

181
Cho, J.-Y. and Pincheira, J. A., 2004. Nonlinear Modeling of RC Columns with Short
Lap Splices, Proceedings of the 13th World Conference on Earthquake
Engineering. Vancouver, B.C., Canada, Aug. 2004. Paper No. 1506.

Ciampi, V., Eligehausen, R., Bertero, V. V., and Popov, E. P., 1982. Analytical Model
for Concrete Anchorages of Reinforcing Bars Under Generalized Excitations,
Report No. UCB/EERC 82/23. Earthquake Engineering Research Center,
University of California, Berkeley, Nov. 1982. 111 pp.

Collins, M. P., and Mitchell, D., Shear and Torsion Design of Prestressed and Non-
Prestressed Concrete Beams, Journal of the Prestressed Concrete Institute, V.
25, No. 5, 1980, pp. 32-100.

Dhakal, P.R., and Maekawa K., Modeling for Postyield Buckling of Reinforcement
,Journal of Structural Engineering. ASCE, V. 129, No. 9, Sept. 2002 pp. 1139-
1147.

Eberhard, M., 2003. Structural Performance Database. Pacific Earthquake


Engineering Research Center, University of Washington, Seattle.
http://nisee.berkeley.edu/spd. Accessed 1 Nov. 2005.

Eligehausen, R., Popov, E. P., and Bertero, V. V., 1983. Local Bond Stress-Slip
Relationship of a Deformed Bar Under Generalized Excitations, Report No.
UCB/EERC 83/23. Earthquake Engineering Research Center, University of
California, Berkeley, Oct. 1983. 169 pp.

Elwood, K. J., 2002. Shake Table Tests and Analytical Studies on the Gravity Load
Collapse of Reinforced Concrete Frames, Ph.D. Dissertation. University of
California, Berkeley, 2002. 419 pp.

Elwood, K. J., 2004. Modeling failures in existing reinforced concrete columns,


Canadian Journal of Civil Engineering. NRC Canada, Vol. 31, No. 5, Oct. 2004.
pp. 846-859.

Elwood, K. J. and Moehle, J. P., 2005a. Axial Capacity Model for Shear-Damaged
Columns, ACI Structural Journal. V. 102, No. 4, July 2005. pp. 578-587.

Elwood, K. J. and Moehle, J. P., 2005b. Drift Capacity of Reinforced Concrete


Columns with Light Transverse Reinforcement, Earthquake Spectra. EERI, V.
21, No. 1, Feb. 2005. pp. 71-89.

Filippou, F. C., Popov, E. P., and Bertero, V. V., 1983. Effects of Bond Deterioration
on Hysteretic Behavior of Reinforced Concrete Joints, Report No. UCB/EERC
83/19. Earthquake Engineering Research Center, University of California,
Berkeley, Aug. 1983. 184 pp.

182
Gerin, M. and Adebar, P., 2004. Accounting for Shear in Seismic Analysis of Concrete
Structures, Proceedings of the 13th World Conference on Earthquake
Engineering. Vancouver, B.C., Canada, Aug. 2004. Paper No. 1747.

Gomes, A. and Appleton, J., (1997), Non-linear cyclic stress-strain relationship of


reinforcing bars including buckling, Engineering Structures, Vol. 19(10) 822-826.

Harajli, M. H., 2004. Comparison of Bond Strength of Steel Bars in Normal- and High-
Strength Concrete, Journal of Materials in Civil Engineering. ASCE, V. 16,
No. 4, Aug. 2004. pp. 365-374.

Hawkins, N. M., Lin, I. J., and Jeang, F. L., 1982. Local Bond Strength of Concrete for
Cyclic Reversed Loadings, Bond in Concrete. Applied Science Publishers,
London, 1982. pp. 151-161.

Inel, M. and Aschheim, M. A., 2002. Models for Estimating the Deformation Capacities
of Reinforced Concrete Columns, ECAS2002 International Symposium on
Structural and Earthquake Engineering. Middle East Technical University,
Ankara, Turkey, Oct. 2002.

Kent, D. C. and Park, R., 1971. Flexural Members with Confined Concrete, Journal of
the Structural Division. ASCE, V. 97, No. ST7, July 1971. pp. 1969-1990.

Lee, D. H., Choi, E., and Zi, G., 2005. Evaluation of earthquake deformation and
performance for RC bridge piers, Engineering Structures. Elsevier, V. 27, No.
10, Aug. 2005. pp. 1451-1464.

Lee, D. H. and Elnashai, A. S., 2001. Seismic Analysis of RC Bridge Columns with
Flexure-Shear Interaction, Journal of Structural Engineering. ASCE, V. 127,
No. 5, May 2001. pp. 546-553.

Lehman, D. E. and Moehle, J. P., 2000. Seismic Performance of Well-Confined


Concrete Bridge Columns, Report No. PEER-1998/01. Pacific Earthquake
Engineering Research Center, University of California, Berkeley, Dec. 2000. 316
pp.

Lin, I. J. and Hawkins, N. M., 1982. Anchorage Characteristics for Reinforcing Bars for
Reversed Cyclic Loading, Report No. SM 82-1. Department of Civil
Engineering, University of Washington, Seattle, June 1982. 326 pp.

Lynn, A. C., 2001. Seismic Evaluation of Existing Reinforced Concrete Building


Columns, Ph.D. Dissertation. University of California, Berkeley, 2001. 308 pp.

183
Lynn, A. C., Moehle, J. P., Mahin, S. A., and Holmes, W. T., 1996. Seismic evaluation
of existing reinforced concrete building columns, Earthquake Spectra. EERI, V.
12, No. 4, Nov. 1996. pp. 715-739.

Mander, J. B., Priestley, J. N., and Park, R., 1988. Theoretical Stress-Strain Model for
Confined Concrete, Journal of Structural Engineering. ASCE, V. 114, No. 8,
Aug. 1988. pp. 1804-1825.

Marquez, J. L. G. and Jirsa, J. O., 1975. A Study of Hooked Bar Anchorages in Beam-
Column Joints, ACI Journal. V. 72, No. 5, May 1975. pp. 198-209.

The MathWorks, Inc., 2005. Matlab. http://www.mathworks.com/products/matlab.

McKenna, F., Fenves, G. L., and Scott, M. H., 2004. Open System for Earthquake
Engineering Simulation. Pacific Earthquake Engineering Research Center,
University of California, Berkeley. http://opensees.berkeley.edu.

Melek, M., Wallace, J. W., and Conte, J. P., 2003. Experimental Assessment of
Columns with Short Lap Splices Subjected to Cyclic Loads, Report No. PEER
2003/04. Pacific Earthquake Engineering Research Center, University of
California, Berkeley, Apr. 2003. 176 pp.

Mo, Y. L. and Wang, S. J., 2000. Seismic Behavior of RC Columns with Various Tie
Configurations, Journal of Structural Engineering, ASCE, V. 126, No. 10, Oct.
2000. pp. 1122-1130.

Moehle, J. P., 1992. Displacement-Based Design of RC Structures Subjected to


Earthquakes, Earthquake Spectra. EERI, V. 8, No. 3, Aug. 1992. pp. 403-428.

Moehle, J. P., Elwood, K. J., and Sezen, H., 2002. Gravity Load Collapse of Building
Frames During Earthquakes, S. M. Uzumeri Symposium: Behavior and Design of
Concrete Structures for Seismic Performance, SP-197. American Concrete
Institute, Farmington Hills, Mich., 2002. pp. 215-238.

Monti, G. & Nuti, C., (1992), Non-linear Cyclic Behaviour of Reinforcing Bars Including
Buckling,Journal of Structural Engineering, ASCE, Vol. 118(12), 3269-3284.

Mostafaei, H., and Kabeyasawa, T., A Simple Approach for Displacement-Based


Assessment of RC Columns, The First NEES/Defense Workshop on Collapse
Simulation of Reinforced Concrete Building Structures, University of California
Berkeley, Berkeley, Calif., July 2005, pp. 47-63.

Mostafaei, H., Axial-Shear-Flexure Interaction Approach for Displacement-Based


Evaluation of Reinforced Concrete Elements. PhD dissertation, Faculty of

184
Engineering, Architrave Department, University of Tokyo, Tokyo, Japan, 2006,
255.

Mostafaei, H., and Kabeyasawa,T., Axial-Shear-Flexure Interaction Approach for


Reinforced Concrete Columns. ACI Structural Journal, vol. 104, No.2, Mar-Apr
2007, pp. 218-226.

Mostafaei, H., and Vecchio, F. J. (2008). Uniaxial Shear-Flexure Model for Reinforced
Concrete Elements. ASCE Journal of Structural Engineering, 134(9), 1538-
1547.

Mostafaei, H., Vecchio, F.J., and Kabeyasawa T. (2009). Deformation Capacity of


Reinforced Concrete Columns. ACI Structural Journal, 106(2), 187-195.

Morita, S. and Kaku, T., 1984. Slippage of Reinforcement in Beam-Column Joint of


Reinforced Concrete Frame, Proceedings of the Eighth World Conference on
Earthquake Engineering. San Francisco, July 1984. pp. 477-484.

National Information Service for Earthquake Engineering (NISEE), 2005. The


Earthquake Engineering Online Archive. Earthquake Engineering Research
Center, University of California, Berkeley. http://nisee.berkeley.edu/elibrary.
Accessed 25 Oct. 2005.

Ousalem, H.; Kabeyasawa, T.; Tasai, A.; and Iwamoto, J., Effect of Hysteretic
Reversals on Lateral and Axial Capacities of Reinforced Concrete Columns,
Proceedings of the Japan Concrete Institute, V. 25, No. 2, 2003, pp. 367-372.

Okamura, H., and Maekawa, K., Nonlinear Analysis and Constitutive Models of
Reinforced Concrete, Gihodo Shuppan Co., Ltd., 1991, 182 pp.

Otani, S. and Sozen, M. A., 1972. Behavior of Multistory Reinforced Concrete Frames
during Earthquakes, Structural Research Series No. 392. University of Illinois,
Urbana, 1972. 551 pp.

Ozcebe, G. and Saatcioglu, M., 1989. Hysteretic Shear Model for Reinforced Concrete
Members, Journal of Structural Engineering. ASCE, V. 115, No. 1, Jan. 1989.
pp. 132-148.

Park, R. and Paulay, T., 1975. Reinforced Concrete Structures. John Wiley & Sons,
Inc., New York, 1975.

Park, R.; Priestley, M. J. N.; and Gill, W. D., Ductility of Square Confined Concrete
Columns, Journal of Structural Division, ASCE, V. 108,No. 4, 1982, pp. 929-
950.

185
Patwardhan, C., 2005. Strength and Deformation Modeling of Reinforced Concrete
Columns, M.S. Thesis. The Ohio State University, Columbus, Ohio, 2005. 166
pp.

Pincheira, J. A., Dotiwala, F. S., and DSouza, J. T., 1999. Seismic Analysis of Older
Reinforced Concrete Columns, Earthquake Spectra. EERI, V. 15, No. 2, May
1999. pp. 245-272.

Pochanart, S. and Harmon, T., 1989. Bond-Slip Model for Generalized Excitation
Including Fatigue, ACI Materials Journal. V. 86, No. 5, Sept. 1989. pp. 465-
474.

Rodriguez, M.E., Botero, J.C. & Villa, J. (1999), Cyclic Stress-Strain Behaviour of
Reinforcing Steel Including Effect of Buckling, Journal of Structural
Engineering, ASCE, Vol. 125(6), 605-612.

Potger, G.M., Kawano A., Griffith M.C., and Warner R.F., Dynamic Analysis of RC
Frames Including Buckling of Longitudinal Steel Reinforcement, Proceedings of
the NZSEE 2001 Conference. Paper No. 4.12.01

Roy, H. E. H., and Sozen, M. A., 1964. Ductility of Concrete, International


Symposium on Flexural Mechanics of Reinforced Concrete Proceedings.
Miami, Fla., Nov. 1964. pp. 213-235.

Saatcioglu, M., Alsiwat, J. M., and Ozcebe, G., 1992. Hysteretic Behavior of
Anchorage Slip in R/C Members, Journal of Structural Engineering. ASCE, V.
118, No. 9, Sept. 1992. pp. 2439-2458.

Saatcioglu, M. and Ozcebe, G., 1989. Response of Reinforced Concrete Columns to


Simulated Seismic Loading, ACI Structural Journal. V. 86, No. 1, Jan. 1989.
pp. 3-12.

Sasani, M., 2004. Shear Strength and Deformation Capacity Models for RC Columns,
Proceedings of the 13th World Conference on Earthquake Engineering.
Vancouver, B.C., Canada, Aug. 2004. Paper No. 1838.

Sezen, H., 2002. Seismic Behavior and Modeling of Reinforced Concrete Building
Columns, Ph.D. Dissertation. University of California, Berkeley, 2002. 324 pp.

Sezen, H., Whittaker, A. S., Elwood, K. J., and Mosalam, K. M., 2003. Performance of
Reinforced Concrete Buildings during the August 17, 1999 Kocaeli, Turkey
Earthquake, and Seismic Design and Construction Practise in Turkey,
Engineering Structures. Elsevier, V. 25, No. 1, Jan. 2003. pp. 103-114.

186
Sezen, H. and Moehle, J. P., 2003. Bond-Slip Behavior of Reinforced Concrete
Members, fib-Symposium: Concrete Structures in Seismic Regions. CEB-FIP,
Athens, Greece, May 2003.

Sezen, H. and Moehle, J. P., 2004. Shear Strength Model for Lightly Reinforced
Concrete Columns, Journal of Structural Engineering. ASCE, V. 130, No. 11,
Nov. 2004. pp. 1692-1703.

Sezen, H., Moehle J.P. Seismic Tests of Concrete Columns with Light Transverse
Reinforcement. ACI Structural Journal 2006; 103 (6) 842-849.

Sezen H., and Setzler E.J. Reinforcement Slip in Reinforced Concrete Columns. ACI
Structural Journal. 2008; 105(3) 280-289

Soroushian, P. and Choi, K.-B., 1991. Analytical Evaluation of Straight Bar Anchorage
Design in Exterior Joints, ACI Structural Journal. V. 88, No. 2, Mar. 1991. pp.
161-168.

Soroushian, P., Obaseki, K., Nagi, M., and Rojas, M. C., 1988. Pullout Behavior of
Hooked Bars in Exterior Beam-Column Connections, ACI Structural Journal.
V. 85, No. 3, May 1988. pp. 269-276.

Setzler, E.J. Modeling the Behavior of Lightly Reinforced Concrete Columns Subjected
to Lateral Loads. M.S. Thesis. The Ohio State University, Columbus, Ohio, 2005.

Setzler, E.J., Sezen, H. Model for the Lateral Behavior of Reinforced Concrete Columns
Including Shear Deformations. Earthquake Spectra. 2008; 24(2) 493-511

Suda, K, Murayama, Y. Ichinomiya, T. & Shimbo, H., (1996), Buckling Behaviour of


Longitudinal Reinforcing Bars in Concrete Columns Subjected to Reverse Lateral
Loading, Eleventh World Conference on Earthquake Engineering, 1996.

Ueda, T., Lin, I. J., and Hawkins, N. M., 1986. Beam Bar Anchorage in Exterior
Column-Beam Connections, ACI Journal. V. 83, No. 3, May 1986. pp. 412-
422.

Umehara, H. and Jirsa, J. O., 1984. Short Rectangular RC Columns under Bidirectional
Loadings, Journal of Structural Engineering. ASCE, V. 110, No. 3, Mar. 1984.
pp. 605-618.

Vecchio, F. J. and Collins, M. P., 1986. The Modified Compression-Field Theory for
Reinforced Concrete Elements Subjected to Shear, ACI Journal. V. 83, No. 2,
Mar. 1986. pp. 219-231.

187
Vecchio, F. J. and Collins, M. P., 1988. Predicting the Response of Reinforced
Concrete Beams Subjected to Shear Using Modified Compression Field Theory,
ACI Structural Journal. V. 85, No. 3, May 1988. pp. 258-268.

Vecchio, F. J., Nonlinear Finite Element Analysis of Reinforced Concrete Membranes,


ACI Structural Journal, V. 86, No. 1, Jan.-Feb. 1989, pp. 26-35.

Vecchio, F. J., and Collins, M. P., Predicting the Response of Reinforced Concrete
Beams Subjected to Shear Using Modified Compression Field Theory, ACI
Structural Journal, V. 85, No. 3, May-June 1988, pp. 258-268.

Wehbe, N. I., Saiidi, M. S., and Sanders, D. H., 1999. Seismic Performance of
Rectangular Bridge Columns with Moderate Confinement, ACI Structural
Journal. Vol. 96, No. 2, Mar. 1999. pp. 248-258.

Wight, J. K. and Sozen, M. A., 1975. Strength Decay of RC Columns under Shear
Reversals, Journal of the Structural Division. ASCE, V. 101, No. ST5, May
1975. pp. 1053-1065.

188