Thesis Precast Shear Wall With Horizontal Connection

ATTENTION: The Singapore Copyright Act applies to the use of this document.
Nanyang Technological University Library
PRECAST SHEAR WALL WITH

HORIZONTAL CONNECTION
HAN HONGSHENG
School of Civil a n d Environmental Engineering
A thesis submitted to the Nanyang Technological University

in fulfillment of the requirement for the degree of
Doctor of Philosophy
2009
ATTENTION: The Singapore Copyright Act applies to the use of this document. Nanyang Technological University Library
ACKNOWLEDGEMENTS
The work described in this thesis was carried out during my study in the Division of
Construction and Management at Nanyang Technological University under supervisor
A/P Ting Seng Kiong.
The author wishes to express his sincere thanks to A/P Ting Seng Kiong for his
invaluable comments and extensive discussions and suggestions throughout the
research work.
Acknowledgements are also due to A/P Li Bing for his helpful and invaluable support
during all the stages of the work.
The efforts of all colleagues, laboratory and administrative staff members of school of
Civil and Environmental Engineering are greatly appreciated. The author feels
grateful for the help from Wuihui, Rong Haicheng in course of research.
I deepest gratitude goes to my family and friends who have always encouraged and
supported me throughout my education.
Han Hongsheng
Table of Contents
TABLE OF CONTENTS
page
ACKNOWLEDEMENTS i
TABLE OF CONTENSTS iii
ABSTRACT ix
LIST OF TABLES xi
LIST OF FIGURES xiii
LIST OF SYMBOLS xxix
CHAPTER 1 INTRODUCTION 1
1.1 Problem Definition 1

1.2 Objective and Scope of Research 3
1.3 Organization of Thesis 4
CHAPTER 2 LITERATURE REVIEW 6
2.1 Introduction 6
2.2 Reinforced Concrete Shear Walls 7
2.3 Composite Shear Walls 11
2.3.1 Composite Members 11
2.4 Connection 16
2.4.1 Shear Connection 17
2.4.2 Diagonal Reinforcement 17
2.5 Strut-and-tie Model 24
2.5.1 History of Strut-and-Tie models 25
2.5.2 B Region and D Regions 31
2.5.3 Design Procedure of Strut-And-Tie Model 33
2.5.4 Method for Developing Strut-and-Tie Models 36
2.5.5 Model Optimization 39
2.5.6 Dimensioning the Struts, Ties and Nodes 40
2.5.7 Feasible Inclined Angle 0 43
2.5.8 Limiting Compressive Stress in Strut 43
2.5.9 Summary on Strut-and-Tie Model 46
2.6 Summary 47
CHAPTER 3 EXPERIMENT DESIGN 48

3.1 Design of Experiment 49
3.1.1 General Description of Wall Specimens 49
3.1.2 Material Properties 51
3.2 Testing Program 72
3.2.1 Test Set-up 72
3.2.2 Instrumentation 72
3.2.3 Load history 87
3.3 Summary 91
CHAPTER 4 EXPERIMENTAL RESULTS 92

4.1 Loading Procedure 92
4.1.1 Load and Displacement Histories 92
4.2 Experimental Results 93
4.2.1 Load-Displacement Response 99
4.2.2 Backbone Envelop of Load-Displacement Response 104

Table of Contents
4.2.3 Stiffness 106

4.2.4 Cracking Patterns 108
4.2.5 Flexural and Shear Deformations 118
4.2.6 Vertical Elongation of the Walls 131
4.2.7 Curvature 144
4.2.8 Sliding of the Walls 147
4.2.9 Deformations at Horizontal Connections 151
4.2.10 Capacity of Energy Dissipation 154
4.3 Theoretical Study 159
4.3.1 Shear Strength 159
4.3.2 Theoretical Model (Macro Wall Model) 162
4.3.3 Curvature Ductility 183
4.4 summary 186
CHAPTER 5 DISCUSSION OF TEST RESULTS 189

5.1 Behavior of Specimen Wl 189
5.1.1 Strain History of the Gauges 189
5.1.2 Load Transfer 213
5.1.3 Crushing of Concrete and Buckling of Vertical Steel 217
5.2 Behavior of Specimen W2 243
5.2.3 Crushing of Concrete and Buckling of Vertical Steel 258
5.2.4 Strut-and-Tie Model 260
5.3.1 Strain History of Gauges 271
5.3.3 Strut-and-Model 288
Table of Contents

5.4.1 Strain History of Gauges 289
5.4.3 Concrete Crushing and Buckling of I-Beam 299
5.5.3 Concrete Crushing and Breaking of Horizontal Bar 321
5.6.1 Strain History 331
5.6.2 Strut-and-Model 334
5.6.3 Failure Mode 334
5.7 Summary 335
CHAPTER 6 FINITE ELEMENT ANALYSIS OF EXPERIMENT
338
6.1 Introduction of DIANA 339
6.2 Model of Reinforced Concrete Walls for Finite Element Analysis 342
6.2.1 Material Model 342
6.2.2 Meshing 355
6.3 Simulation of Specimen Wl 357
6.3.1 2-D Model 357
6.3.2 2-D Analytical Results 358
6.3.3 3-D Model and Results 363
6.3.4 Summary 365
6.4 Simulation of Specimen W2 366
Table of Contents
6.4.1 2-D Model 366

6.4.1 Analytical Results 368
6.5.1 2-D Model 372
6.6.1 2-D Model 376
6.6.3 3-D Model 379
6.6.4 Summary 389
6.7.1 Analytical Model 390
6.7.2 Summary 395
6.8.1 Analytical Model 395
6.9 Summary 399
CHAPTER 7 CONCLUSIONS AND RECOMMENDATIONS 401

7.1 Conclusions 401
7.2 Recommendations 404
7.3 Future Works 407
REFERENCES 409
APPENDIX 420
PUBLICATION 424
Abstract
SUMMARY
Shear walls are widely used in concrete structures to resist lateral force originated
from wind and earthquake. Currently, shear walls using steel sections as flexural
reinforcement at two ends are successfully used in practice, and this kind structure
also showed good performance in past severe earthquake. Some provisions are
available in the NEHRP (1997) and design recommendations also have been
implemented in AIJ (1985).
In current construction practice of this kind of shear wall, the steel frame must be
erected first, and then concrete is placed. This sequence may cause stability problem
of the steel frame before concrete is cast and vertical formworks are also needed. In
order to make a breakthrough in this field and accelerate the construction speed on
site without influencing the performance of the wall, the concept of "precast shear
wall" with effective connection is introduced into this kind of wall in present
research.
In order to investigate performance of precast walls with horizontal connection, six

shear walls, two monolithic walls acting as prototypes and four precast shear walls
serving as counterparts, have been tested. Two kinds of steel sections, channels and
I-Beams have been used in these walls serving as flexural reinforcement. The
influence of number of shear connectors in the horizontal connection on the
behavior has also been studied.
All specimens have been tested under reversed cyclic load. Similar ultimate strength,
drift ratio were observed between these monolithic walls and their precast
counterparts in the experiment. However, meantime monolithic wall exhibited
different features from those of precast ones during loading course, including
IX
Abstract
cracking patterns, failure modes and failure locations, internal force paths, stiffness
and capacity of energy dissipation.
Theoretical model based on macro model was developed. The precast walls are
divided into macro model for walls and for connection with different material
models. The suggested macro model produced reasonable good prediction for the
precast walls and a parametric study based on the macro model has been done
Based on cracking patterns and records of strain gauges placed horizontal and
vertical reinforcement, strut-and-tie models of each specimen during different
loading stages have been developed.
A nonlinear finite element program, DIANA 8.0, has been used to study the
behavior of specimens under cyclic loads. Two dimensional and three dimensional
models have been built to simulate behavior of these specimens. In two dimensional
models, steel sections, channels and I-Beams, were modeled by truss element model.
It showed that behavior of specimens with embedded channels could be captured by
both two and three dimensional models. However, the ultimate strength of wall with
I-beams is underestimated 15% by two-dimensional model. In three dimensional
models, the steel sections were simulated by shell elements and the interface action
between concrete and steel sections was simulated by interface elements with
frictional material model. The yield strength, ultimate and ductility were predicted
with a high agreement by three dimensional models for all walls. Linear and
nonlinear frictional material models were also investigated.
x
List of Tables
LIST OF TABLES
page
CHAPTER 3 EXPERIMENT DESIGN
Table 3.1.2-1 Average concrete compressive strength at the day of testing 51
Table 3.1.2-2 Properties of reinforcement and steel section 55
Table 3.2.2-1 Number of LVDTs and strain gauges 82
CHAPTER 4 EXPERIMENTAL RESULTS

Table 4.2.4-1 Characteristics of experimental results 119
Table 4.2.6-1 Vertical elongation of specimens 144
Table 4.2.8-1 Ratio of sliding of specimen to total displacement 151
Table 4.3.1-1 Comparison Of Calculated Results Basing On Codes
And Experimental Result 162
Table 4.3.3-1 Comparison Curvature Ductility Based On ACI And
Experimental Result 185
Table 4.3.2-2 Comparison Of Theoretical And Experimental Curvature 188
CHAPTER 5 DISCUSSION OF TEST RESULTS

Table 5.1.4-1 Strain energy in strut-and-tie model of specimen Wl 224
Table 5.2.4-1 Strain energy in strut-and-tie model of specimen W2 262
XI
List of Figures
LIST OF FIGURES
page
CHAPTER 2
Fig.2.4.2-1 External acting forces and internal reactions at shear wall inclined crack
in direction ® 24
Fig.2.5.1-1 Compressive stress-strain curve of concrete 28

Fig.2.5.2-2 Wall shear resisting mechanisms: (a) diagonal; (b) horizontal; (c)
vertical 29
Fig.2.5.2-1 Geometric Discontinuity (ACI 318-02, Appendix) 32
Fig.2.5.2-2 Load Discontinuity (ACI 318-02, Appendix) 33
Fig.2.5.3-1 Design Flow Chart of Reinforced Concrete Structure 35
a) B- and D-Regions (ACI 318-02, Appendix) 35
b) Strut-and-Tie Model for Entire Beam (ACI 318-02, Appendix) 36
Fig.2.5.3-2 Strut-and-Tie Model for Simple Span Beam (ACI 318-02, Appendix) 36
Fig.2.5.4-1 Strut-and-Tie Based on Elastic Stresses (Schlaich, 1987) 37
Fig.2.5.4-2 Strut-and-Tie Based on Elastic Stresses (Schlaich, 1987) 37
Fig.2.5.4-1 Strut-and-Tie Model Based on Load Paths (Schlaich, 1987) 39
Fig.2.5.6-1 Typical Configurations of Compression Field (Wu Hui 2005) 41
Fig.2.5.6-2 Type of Nodes (Schlaich, 1987) 42
Fig.2.5.6-3 Idealized Forces at Nodal Zone (Schlaich, 1987) 42
Fig.2.5.9-1 Limiting Stress for Truss Elements (ACI 318-02, Appendix A) 46
CHAPTER 3
Fig. 3.1.2-1 Stress Versus Strain Relationship of T10 Bar 52
Fig. 3.1.2-2 Stress Versus Strain Relationship of Rl 0 Bar 53
Fig. 3.1.2-3 Coupon design of I-Beam and Channel 53
Fig. 3.1.2-4 Stress Versus Strain Relationship of Channel 54
Xlll
List of Figures
Fig. 3.1.2-5 Stress Versus Strain Relationship of 1-Beam 54

Fig. 3.1.3-1 Specimen Wl and Its Overall Dimensions 62
Fig. 3.1.3-2 Specimen W2 and Its Overall Dimensions 63
Fig. 3.1.3-7 Side View of All Specimens and Their Dimensions 68
Fig. 3.1.3-8 Dimensions of Top Beam and Base Beam 69
Fig. 3.1.3-9 Details of Connection (Cont.) 70
Fig. 3.1.3-10 Details of Connection (Cont.) 71
Fig. 3.1.3-11 Details of Connection 71
Fig. 3.2.1-1 Testing Schematic Arrangement 73
Fig. 3.2.1-2 General View of Set Up of Specimen Wl Fixed To The Lab Floor 74
Fig. 3.2.2-1 Layout of Strain Gauges on Steel Bars Of Wl and W4 76
Fig. 3.2.2-2 Layout of Strain Gauges on Steel Bars of W2 and W5 77
Fig. 3.2.2-3 Layout of Strain Gauges on Steel Bars of W3 and W6 78
Fig. 3.2.2-4 Layout of Strain Gauges on Steel Sections of Wl 79
Fig. 3.2.2-5 Layout of Strain Gauges on Steel Sections of W2, W3, W5 And W6 80
Fig. 3.2.2-6 Layout of Strain Gauges on Steel Sections of W4 81
Fig. 3.2.2-7 Layout of LVDTs on Wl and W4 82
Fig. 3.2.2-8 Layout of LVDTs on W2, W3, W5 and W6 82
Fig. 3.2.2-9 Calculation of Shear Deformation from Rectangle 83
Fig. 3.2.2-10 Flexural Deformation from Verticals at Both Ends of Specimens 86
Fig. 3.2.2- 11 Calculation of Flexural Deformations 86
Fig. 3.2.3-1 Definition of yield displacement (Park, 1989) 88
Fig. 3.2.3-2 Load history of specimen Wl 88
Fig. 3.2.3-3 Load history of specimen W2 89
xiv
List of Figures

CHAPTER4
Fig. 4.1.1-1 Horizontal Displacement History of Specimen W1 93

Fig. 4.1.1 -5 Horizontal Displacement History of Specimen W5 95
Fig. 4.1.1-7 Lateral Load History of Specimen Wl 96
Fig. 4.1.1-8 Lateral Load History of Specimen W2 97
Fig. 4.1.1 -9 Lateral Load History of Specimen W3 97
Fig. 4.2.1-1 Hysteresis Loops of Specimen Wl 101
Fig. 4.2.1 -2 Hysteresis Loops of Specimen W2 102
Fig. 4.2.1-3 Hysteresis Loops of Specimen W3 102
Fig. 4.2.1- 6 Hysteresis Loops of Specimen W6 104
Fig. 4.2.2-1 Backbone Envelopes of Load-Displacement Response of Specimens 105
Fig. 4.2.2-2 Backbone Envelopes of Load-Displacement Response of Specimens 106
Fig. 4.2.3- 1 Typical Relationship for A Reinforced Concrete Element 107
Fig. 4.2.3-2 The relationship of Lateral Stiffness versus Drift Ratio 108
Fig. 4.2.4-1 Cracking Patterns of Specimen Wl 113
Fig. 4.2.4-2 Cracking Patterns of Specimen W2 114
Fig. 4.2.4- 3 Cracking Patterns of Specimen W3 115
xv
List of Figures

Fig. 4.2.5-1 Flexural Deformation of Specimen Wl 121
Fig. 4.2.5-2 Shear Deformation of Specimen Wl 122
Fig. 4.2.5-3 Flexural Deformation of Specimen W2 122
Fig. 4.2.5-4 Shear Deformation of Specimen W2 123
Fig. 4.2.5-11 Lateral Loads versus Deformations of Specimen Wl 126
Fig. 4.2.5-12 Lateral Loads versus Deformations of Specimen W2 127
Fig. 4.2.5- 16 Ratios of Deformations to Total Displacement of Specimen Wl 129
Fig. 4.2.5-17 Ratios of Deformations to Total Displacement of Specimen W2 129
Fig. 4.2.6-1 Local Vertical Elongations of Wl -LVDT2 (cont.) 134
Fig. 4.2.6- 2 Local Vertical Elongations of W1-LVDT3 (cont.) 134
Fig. 4.2.6-3 Local Vertical Elongations of W1-LVDT4 (cont.) 135
Fig. 4.2.6-4 Local Vertical Elongations of W1-LVDT5 135
Fig. 4.2.6-5 Total Vertical Elongations of Wl 136
Fig. 4.2.6-6 Vertical Elongations of Wl versus Scan Number 136
Fig. 4.2.6-7 Local Vertical Elongations of W2 (cont.) 137
xvi
List of Figures

Fig. 4.2.6-12 Vertical Elongation of Specimen W2 139
Fig. 4.2.6- 13 Vertical Elongations of W2 versus Scan Number 140
Fig. 4.2.6-15 Vertical Elongation of Specimen W5 (Cont.) 141
Fig. 4.2.6-19 Vertical Elongations of W5 143
Fig. 4.3.6-20 Vertical Elongations of W5 versus Scan Number 143
Fig. 4.2.7-1 Curvature of Specimen Wl 145
Fig. 4.2.8-1 Sliding Deformation of Specimen Wl 149
Fig. 4.2.8-2 Sliding Deformation of Specimen W2 149
Fig. 4.2.9-1 Records of L23 versus Lateral Load of W2 152
Fig. 4.2.9-2 Records ofL24 versus Lateral Load of W2 152
Fig. 4.2.9-3 Total horizontal Deformation at Horizontal Connection of W2 153
Fig. 4.2.9-4 Flexural Deformation at Horizontal Connection of W2 153
Fig. 4.2.9-5 Shear Deformation at Horizontal Connection of W2 154
Fig. 4.2.10-1 Energy Dissipation per Displacement of Specimen Wl 156
Fig. 4.2.10-2 Energy Dissipation per Displacement of Specimen W2 156
Fig. 4.2.10-6 Comparison of Energy Dissipation per Displacement of All
Specimens 158
Fig. 4.3.2-1 Original Macro Model by Kabeyasawa et al (1982) 163
Fig 4.3.2-2 Suggestion for Improved Macro Models 164
xvn
List of Figures
Fig 4.3.2-3 Macro Models by Linde (1993) 164

Fig. 4.3.2-4 Wall Rotation for Uniform Moment by Linde (1993) 165
Fig. 4.3.2-5 Wall Rotation with Moment Gradient by Linde (1993) 166
Fig. 4.3.2-6 Wall Shear Behavior by Linde (1993) 168
Fig. 4.3.2-7 Wall Shear Behavior by Linde (1993) 168
Fig 4.3.2-8 Suggested Macro Models 169
Fig 4.3.2-9 Suggested Macro Models for Precast Shear Walls W2,3,5,6 170
Fig 4.3.2-10 Load Status in the Horizontal Connection 172
Fig 4.3.2-11 Typical Deformation of the Horizontal Connection 172
Fig 4.3.2-12 Equivalent Stress Distribution 173
Fig 4.3.2-13 Material Model for Outer Springs 173
Fig 4.3.2-14 Forces in Diagonal Reinforcement in the Shear Connection 174
Fig 4.3.2-15 Loads on MMW 175
Fig 4.3.2-16 Flexural Behavior of MMW 176
Fig 4.3.2-17 Property of Flexural Springs 177
Fig 4.3.2-18 Concrete Material Model by Kent & Park (1971) 177
Fig 4.3.2-19 Comparison of Prediction vs. Experimental Data of W2,3 179
Fig 4.3.2-20 Comparison of Prediction vs. Experimental Data of W5,6 179
Fig 4.3.2-21 Comparison of Prediction with HSF Bolts vs. Experimental Data of
W2,3 181
Fig 4.3.2-22 Parametric Study Model for Connection at Bottom 182
Fig 4.3.2-23 Comparison of Prediction with Connection at Bottom 182
vs. Experimental Data of W2,3 182
CHAPTER 5
Fig. 5.1.1-1 Strain History of Vertical Bar SV1 191

Fig. 5.1.1-2 Strain History of Vertical Bar NV2 191
Fig. 5.1.1 -3 Strain History of Vertical Bar S V3 192
xvni
List of Figures

Fig. 5.1.1-7 Strain Profile of Vertical Bar at the Base Line of Specimen Wl for
Negative Load 194
Fig. 5.1.1-8 Strain Profile of Vertical Bar at the Base Line of Specimen Wl for
Positive Load 194
Fig.5.1.1-9 Strain Profiles of along Vertical Bar SV1 for Negative Load 196
Fig.5.1.1-10 Strain Profiles of along Vertical Bar SV1 for Positive Load 196
Fig.5.1.1-11 Strain Profiles of along Vertical Bar NV4 for Negative Load 197
Fig.5.1.1-12 Strain Profiles of along Vertical BarNV4 for Positive Load 197
Fig.5.1.1-13 Strain Profiles of along Vertical Bar NV6 for Negative Load 198
Fig.5.1.1-14 Strain Profiles of along Vertical Bar NV6 for Positive Load 198
Fig. 5.1.1-15 Strain Profile of Channel SSI for Negative Load 200
Fig. 5.1.1 -16 Strain Profile of Channel SS1 for Positive Load 200
Fig. 5.1.1-17 Strain Profile of Channel SS2 for Negative Load 201
Fig. 5.1.1-18 Strain Profile of Channel SS2 for Positive Load 201
Fig. 5.1.1-19 Strain Profile of Channel NN1 for Negative Load 202
Fig. 5.1.1-20 Strain Profile of Channel NN1 for Positive Load 202
Fig. 5.1.1-21 Strain Profile of Channel NN2 for Negative Load 203
Fig. 5.1.1-22 Strain Profile of Channel NN2 for Positive Load 203
Fig.5.1.1-23 History of Strain in Diagonal BarNDl 205
Fig.5.1.1-24 History of Strain in Diagonal Bar SD2 205
Fig.5.1.1-28 Strain History of Horizontal Bar HN2 208
Fig.5.1.1-29 Strain History of Horizontal Bar SH3 209
Fig. 5.1.1 -3 0 Strain History of Horizontal Bar NH4 209
Fig.5.1.1-32 Strain History of Horizontal Bar NH6 210
xix
List of Figures

Fig.5.1.2-1 Diagonal Crack in Left Corner at 0.75^ 215
Fig.5.1.2-2 Diagonal Crack in Left Corner at DF=3 216

Fig.5.1.2-3 Diagonal Crack in Left Corner at DF=6 216
Fig.5.1.3-1 Fracture of Vertical Bar 218
Fig. 5.1.3-2 Buckling of Channel in Specimen W1 219
Fig.5.1.4-1 Simple Strut-and-Tie Model of Specimen Wl 225
Fig.5.1.4-2 Additional Strut-and-Tie Model of Specimen Wl 225
Fig.5.1.4-3 Combined Strut-and-Tie Model of Specimen Wl 226
Fig.5.1.4-4 compression zone (by Yanez 1993) 226
Fig.5.1.4-5 Strut-and-Tie Model I of Specimen Wl for Vertical bars 229
Fig.5.1.4-6 Strut-and-Tie Model II of Specimen Wl for Vertical bars 229
Fig.5.1.4-7 Combined Strut-and-Tie Model of Specimen Wl for Channels 230
Fig.5.1.4-8 Strut-and-Tie Model I of Specimen Wl for Channels 230
Fig.5.1.4-9 Strut-and-Tie Model II of Specimen Wl 231
Fig.5.1.4-11 Strut-and-Tie Model of Specimen Wl 232
Fig. 5.2.1 -1 Strain History of Vertical Bar S V3 243
Fig. 5.2.1-3 Strain History of Vertical Bar NV9 244
Fig.5.2.1-4 Strain Profile of Vertical Bars near the Base for Negative Load 245
xx
List of Figures
Fig.5.2.1-5 Strain Profile of Vertical Bars near the Base for Positive Load 245
Fig.5.2.1-6 Hysteresis Loops of SD1 246
Fig.5.2.1-7 Strain history of SD1 versus Scan Number 247
Fig.5.2.1-8 Hysteresis Loops of ND3 247
Fig.5.2.1 -9 Strain history of ND3 versus Scan Number 248
Fig.5.2.1-10 Hysteresis Loops of SUD5 248
Fig.5.2.1-11 Strain history of SDU5 versus Scan Number 249
Fig.5.2.1-12 Hysteresis Loops of NDL7 249
Fig.5.2.1-13 Strain history of NDL7 versus Scan Num 250
Fig.5.2.1-14 Strain History of Horizontal Bar NH11-2 251
Fig.5.2.1-15 Strain History of Horizontal BarNH10-2 251
Fig.5.2.1-16 Strain History of Horizontal Bar SH9-2 252
Fig.5.2.1 -20 Strain History of Horizontal Bar SS2-1 254
Fig.5.2.1-21 Strain History of Horizontal Bar SS2-2 254
Fig.5.2.1 -22 Strain History of Horizontal Bar NS2-4 255
Fig.5.2.1-23 Strain History of Horizontal Bar NS2-5 255
Fig.5.2.1-24 Strain History of Horizontal Bar NS1-6 256
Fig.5.2.3-1 Crushing of Concrete in Specimen W2 259
Fig.5.2.3-2 Crushing of Concrete in Specimen W2 260
Fig.5.2.4-1 Strut-and-Tie Model for W2 261
Fig.5.2.4-2 Strut-and-Tie Model for Upper Wall Panel and Connection Zone in
Upper Panel 264
Fig.5.2.4-3 First Strut-and-Tie Model of Specimen W2 265
Fig.5.2.4-4 Second Strut-and-Tie Model of Specimen W2 265
Fig.5.2.4-5 First Strut-and-Tie Model of Specimen W2 under Proportioned Load 266
Fig.5.2.4-6 Second Strut-and-Tie Model of Specimen W2 under Proportioned Load 266
Fig.5.2.4-7 Combined Strut-and-Tie Model of Specimen W2 267
Fig.5.2.4-8 Behavior of Horizontal Connection 269
xxi
List of Figures

Fig. 5.3.1 -3 Strain History of Vertical Bar NV6 272
Fig.5.3.1-4 Strain Profile of Vertical Bars near the Base for Negative Load 273
Fig.5.3.1-5 Strain Profile of Vertical Bars near the Base for Positive Load 274
Fig.5.3.1-6 Strain History of Diagonal Bar ND1 275
Fig.5.3.1-7 Relationship Strain versus Scan Number of ND1 275
Fig.5.3.1-8 Strain History of Diagonal Bar ND4 276
Fig.5.3.1-9 Relationship Strain versus Scan Number of ND4 276
Fig.5.3.1-10 Strain History of Diagonal Bar NDU6 277
Fig.5.3.1-11 Relationship Strain versus Scan Number of NDU6 277
Fig.5.3.1-12 Strain History of Diagonal Bar SDU5 278
Fig.5.3.1-13 Relationship Strain versus Scan Number of SDU5 278
Fig.5.3.1-14 Strain History of Diagonal Bar NDL2 279
Fig.5.3.1-15 Relationship Strain versus Scan Number of NDL2 279
Fig.5.3.1-16 Strain History of Diagonal Bar SDU2 280
Fig.5.3.1-17 Relationship Strain versus Scan Number of SDU2 280
Fig.5.3.1-18 Relationship Strain versus Scan Number of SHI-2 281
Fig.5.3.1-19 Relationship Strain versus Scan Number of SH3-2 282
Fig.5.3.1-20 Relationship Strain versus Scan Number of NH5-2 282
Fig.5.3.1-22 Relationship Strain versus Scan Number of SH9-2 283
Fig.5.3.1-24 Relationship Strain versus Scan Number of SHI2-2 284
Fig.5.3.1-25 Strain History of Channel NS1-1 285
Fig.5.3.2-1 Fracture of Channels 288
Fig.5.4.1-1 Strain History of Vertical Bars SV1 290
xxn
List of Figures
Fig.5.4.1-2 Strain History of Vertical Bars NV2 290

Fig.5.4.1-3 Strain History of Vertical Bars SV6 291
Fig.5.4.1-6 Strain Profiles of Vertical Bars for Negative Load 292
Fig.5.4.1-7 Strain Profiles of Vertical Bars for Positive Load 293
Fig.5.4.1 -8 Relationship of Strain versus Scan Number of SD1, ND1 294
Fig.5.4.1-9 Relationship of Strain versus Scan Number of SD3, ND3 294
Fig.5.4.1-10 Relationship of Strain versus Scan Number of NH11 295
Fig.5.4.1-11 Relationship of Strain versus Scan Number of SH9, NH9 296
Fig.5.4.1-14 Relationship of Strain versus Scan Number of SH3 297
Fig.5.4.3-1 Spalling of Concrete Cover in Compression Zone 300
Fig.5.4.3-2 Spalling of Concrete along the Inner Boundaries of I-Beams 300
Fig.5.4.4-1 Center of Gravity of Vertical Bars in Tension 302
Fig.5.4.4-2 Strut-and-Tie Model of W4 303
Fig.5.4.4-4 Combined Strut-and-Tie Model of W4 304
Fig.5.5.1-1 Relationship of strain Versus Scan Number of Vertical Bar SV1 306
Fig.5.5.1-2 Relationship of strain Versus Scan Number of Vertical BarNV2 307
Fig.5.5.1 -3 Relationship of strain Versus Scan Number of Vertical Bar SV3 307
Fig.5.5.1-4 Relationship of strain Versus Scan Number of Vertical Bar NV4 308
Fig.5.5.1-5 Relationship of strain Versus Scan Number of Vertical Bar SV5 308
Fig.5.5.1-6 Strain Profiles of vertical Bar for Positive Load 309
Fig.5.5.1-7 Strain Profiles of vertical Bar for Negative Load 309
Fig.5.5.1-8 Strain History of SD1 311
xxm
J( l/f I IK"' C-d
Fig.5.5.1-9 Relationship of Strain versus Scan Number of SD1 311

Fig.5.5.1-10 Strain History of ND4 312
Fig.5.5.1-11 Relationship of Strain versus Scan Number of ND4 312
Fig.5.5.1-12 Strain History of NDL2 313
Fig.5.5.1-13 Relationship of Strain versus Scan Number of NDL2 313
Fig.5.5.1-14 Strain History of SDL6 314
Fig.5.5.1-15 Relationship of Strain versus Scan Number of SDL6 314
Fig.5.5.1-16 Strain History of SDU5 315
Fig.5.5.1-17 Relationship of Strain versus Scan Number of SDU5 315
Fig.5.5.1-18 Strain History of SDU8 316
Fig.5.5.1-19 Relationship of Strain versus Scan Number of SDU8 316
Fig.5.5.2-1 Cracking direction in lower corner 321
Fig.5.5.4-1 Strut-and-Tie Model 326
Fig.5.5.4-3 Combined Strut-and-Tie Model 327
Fig.5.6.1-1 Strain Profiles of Vertical Bars for Positive Load 332
Fig.5.6.1-2 Strain Profiles of Vertical Bars for Negative Load 332
Fig.5.6.1-3 Strain History of Diagonal Bar SD3 333
Fig.5.6.1-4 Strain History of Diagonal Bar SD4 334
Fig.5.6.3-1 Connection Failure in Specimen W6 335
xxiv
List of Figures
Fig.5.6.3-2 Shear Failure of Bolt in Flexural Connection 335
CHAPTER 6
Fig.6.2.1-1 Nonlinear Tension Softening (Hordijk et al.) 344

Fig.6.2.1-2 (a) Uniaxial Stress-Strain Curve, 346
(b) Uniaxial Stress-Plastic Strain Curve 346
Fig.6.2.1-3 Strain Hardening 347
Fig.6.2.1 -4. Mohr-Coulomb yield condition 349
Fig.6.2.1 -5 Typical Mohr-Coulomb Failure Criterion 349
Fig.6.2.1-6 Derivation of hardening diagram for Mohr-Coulomb yield condition 351
Fig.6.2.1-7 Interlocking model of friction 352
Fig.6.2.1-8. Schematic of the direct interface shear test 352
Fig.6.2.1-9 Coulomb friction model 354
Fig.6.3.1-1 Mesh of Specimen Wl 358
Fig.6.3.1-2 Layout of Reinforcement and Concrete 358
Fig.6.3.2-1 Comparison of Experimental and analytical Results 360
Fig.6.3.2-2 Principal Compressive Stress at -0.25^ 361
Fig.6.3.2-3 Principal Compressive Stress at - ' 361
Fig.6.3.2-4 Principal Compressive Stress at Ductility -8 362

Fig.6.3.2-5 Stress Distribution in Diagonal Reinforcement at Ductility -8 362
Fig.6.3.2-6 Stress Distribution in Vertical and Horizontal Bars at Ductility -8 362
Fig.6.3.2-7 Deformation of Specimen 363
Fig.6.3.3-1 Three Dimensional Model 364
Fig.6.3.3-2 Mesh of Channels 365
Fig.6.3.3-3Comparison of Analytical and Experimental Results 365
Fig.6.4-1 Line Interface Element CL24I 366
Fig.6.4-2 Mesh of Specimen W2 367
XXV
List of Figures
Fig.6.4-3 Stress-Strain Curve for Diagonal Reinforcement in Horizontal Connection368

Fig.6.4.1-1 Comparison between Analytical and Experimental Results 370
Fig.6.4.1 -2 Principal Stress at 0.5 p- 370
Fig.6.4.1-3 Principal Stress at p< 371
Fig.6.4.1-4 Principal Stress at Ductility 3 371

Fig.6.4.1-5 Stress in Horizontal Bars 371
Fig.6.5.1-1 Mesh of Specimen W3 372
Fig.6.5.2-1 Comparison of Analytical and Experimental Result of Specimen W3 374
Fig.6.5.2-2 Principal Stress at -0.5 p< 374
Fig.6.5.2-3 Principal Stress at -p* 375
Fig.6.5.2-4 Principal Stress at ductility -3 375

Fig.6.5.2-5 Principal Stress Horizontal Reinforcement at Ductility -3 375
Fig.6.6.1-1. Cross section of wall W4 376
Fig.6.6.1-2. Simulation of I-Beam in 2-D model 376
Fig.6.6.1-3. Mesh of Specimen W2 with Equivalent I-Beam in 2-D model 377
Fig.6.6.2-1 Comparison of Analytical and Experimental Results of W4 by 2-D
Model 378
Fig.6.6.2-2.Confinement Effect of I-Beam 378
Fig.6.6.2-3 Confining Effect of I-Beam 379
Fig.6.6.3-1 Mesh and layout of confinement 380
Fig.6.6.3-2 Comparison analytical and experimental results 381
Fig.6.6.3-3 Shear stress in I-Beam 381
Fig.6.6.3-4: Curved shell elements, characteristics 383
Fig.6.6.3-5. Models of I-Beam and details of elements 383
Fig.6.6.3-6 Comparison between Analytical and Experimental Results of W4 384
Fig.6.6.3-7 Comparison failure mode 384
Fig.6.6.3-8 Contour of compressive stress 385
Fig.6.6.3-9 Falling of concrete along inner boundary of I-Beam 385
xxvi
List of Figures
Fig.6.6.3-10 3-D model with interface elements 387

Fig.6.6.3-11. Characteristics of CQ48I 387
Fig.6.6.3-12 Variables of three-dimensional structural interfaces 387
Fig.6.6.3-13 Comparison between model with and without interface elements 388
Fig.6.6.3-14 Comparison between experimental results and analytical one with
interface element with linear material model 388
Fig.6.6.3-15 Comparison between Experimental Results and Analytical Ones with
Interface Element of Nonlinear Frictional Model 389
Fig.6.7.1-1 Mesh of Specimen W5 391
Fig.6.7.1-2 Model of I-Beam used in Specimen W5 391
Fig.6.7.2-1 Comparison between Analytical and Experimental Results 393
Fig.6.7.2-2 Principal Stresses of Specimen W5 at -0.5 p> 393
Fig.6.7.2-3 Principal Stresses of Specimen W5 at -p' 394
Fig.6.7.2-4 Principal Stresses of Specimen W5 at ductility -3 394

Fig.6.7.2-5 Stresses in Horizontal Bars of W5 at ductility -3 395
Fig.6.8.2-1 Principal Stresses of Specimen W6 at -0.5 ^ 396
Fig.6.8.2-1 Comparison between Analytical and Experimental Results of W6 397
Fig.6.8.2-2 Principal Stresses of Specimen W5 at -0.5 p< 397
Fig.6.8.2-2 Principal Stresses of Specimen W5 at - p< 398
Fig.6.8.2-2 Principal Stresses of Specimen W5 at Ductility -3 398
Fig. 6.8.2-2 Principal Stresses of Specimen W5 at -0.5 p> 399
xxvn
List of Symbols
LIST OF SYMBOLS
A total area of the element

Aa effective cross sectional area of strut
Ae effective area of steel section
As sectional area
Asd area of the diagonal reinforcement in diagonal direction
Asfl area of edge column longitudinal reinforcement
Asp cross sectional area of outer spring
Astr effective area of diagonal strut
Aw net area of the concrete section bounded by the w e b thickness tw and
the length of the section in the direction of the shear force lw
Av area of the shear reinforcement within distance s
as depth of the diagonal strut
b thickness of wall
bh the width of the hoops
bs width of the diagonal strut that can be taken as the width of the wall
web
bv effective web width
bK thickness of web
C cohesion
Cc compression force taken by concrete
Cs compression force taken b y steel
c depth of neutral axis, cohesion
c^,c2 parameters in definition of tensile behavior of concrete
D matrix of stiffness
d effective compression block
d], d2 distance from centroid of steel section and that of bars in tension
zone to the extreme compression fiber
dm stiffness in normal direction
dss stiffness in tangent direction
dv effective shear depth
d length of lever a r m
E modulus of elasticity
Ec concrete modulus of elasticity
Es steel modulus of elasticity
F horizontal force acting on top beam
Fh sum offerees in horizontal reinforcement
XXIX
List of Symbols
Fv sum of forces in vertical reinforcement in central 0.8 width of wall

Ft force in strut or tie i
Fs force in the bar or group of bars
Fsp force in outer spring,
Fr tension force on steel sections when the specimens reaches its yield
strength
f{<y,K) principal stress space
fh bond stress
fhu design ultimate anchorage bond stress
fc, fc compression strength of concrete cylinder
frJ the design strength of concrete
fni compression strength of concrete cube
fd stress in diagonal reinforcement
fe elastic strength
fsp stress in spring
ft tensile strength of concrete
fu tensile strength of reinforcement
/ steel yield strength
f design strength of this reinforcement
G shear modulus, shear stiffness
G'f tensile energy of concrete element.
Gcr shear stiffness of cracked concrete
g acceleration due to gravity
g{(j,K) plastic potential
H, distance from center of interval of wall to the horizontal axis of the
top beam
Hw the height of the wall measured from base line of wall to the
horizontal axial of the top beam
h wall height, crack bandwidth
hi distance from load point to center of connection
hi distance from load point to top of macro wall model
h distance from central line to horizontal connection, from rotation
center to the top of wall.
hs mean story height
h , hw wall height
h the height of interval covering horizontal connection
/ important factor
j stress block depth factor
K, the axial stiffness
xxx
List of Symbols
K, Ki initial stiffness
Ksp,e Elastic stiffness of outer spring
k internal state variable, empirical factor
l0 the element's length from the point of zero moment, to the point of
maximum moment
/ width of wall, anchorage length of reinforcement, action length of
spring
le elastic action length of bond
/;, plastic action length of bond
ld,sp action length of diagonal reinforcement
lw width of wall
/, distance between the inclined bars' center axes, at the base cross the
section
p
length of the plastic hinge
MSd the design moment
m modular ratio of steel to concrete
N vertical force acting on the wall
Nv the design axial force at the potential sliding plane
n number of bars
P wall axial load, resistance
P. ideal strength
P lateral force acting on top beam when specimens reaches its yielding
strength
Q normal load
S0 overstrength
Su design strength
S yield strength of member
s spacing of shear reinforcement
sh the spacing of the hoops
T tensile strength of brittle material
To tensile limit, a tensile limit is used when the tensile strength of the
material is lower than the strength defined by the failure criterion
Ti tensile strength of vertical web reinforcement and the dowel action
of the horizontal web reinforcement
Tc compressive force at the edge column, concrete tie in tension without
reinforcement
Tcfr aggregate interlock and friction resistance force at the edge column
Td dowel action strength of the horizontal web reinforcement,
Tdg tensile strength of bidiagonal reinforcement
Tft aggregate interlock and friction resistance force, along the crack in
the web,
Th tensile strength of the horizontal web reinforcement and dowel
action of vertical web reinforcement
xxxi
List of Symbols
Ts tensile force in steel, ties in tension with reinforcement

Ts-bars tensile force by bars
T
s-chameh tensile force by channel
thickness of shell element, traction
tolerance between bolts and bolt holes
normal traction
shear traction, tolerance for bolts
web thickness
Ub the bond strength
us horizontal displacement of top beam
K'K' K shear strength of member by concrete
V„ dowel action of vertical bars
<u
V,. the friction resistance
fa
Yu the shear resistance of diagonal bars

Vcd concrete contribution to shear strength again tension failure
VU1 shear strength against diagonal tension failure
Vnil web reinforcement contribution to shear strength against tension
failure
2
effective internal lever arm
a interval length of wall, factors to account for reinforcement
overstrength, angle between diagonal and horizontal
as shear span ratio, depth of equivalent rectangular stress block,
empirical shear reductor
(3 coefficient dependent on the bar type, shear retention factor, dilation
angle, inclined angle of teeth
5h horizontal displacement
Sh horizontal shear deformation
8U roof displacement
Sv vertical displacement
s0 concrete strain at m a x i m u m compressive strength
£•, principal tensile strain
e2 principal compression strain
s50jl strain at 50% strength on the descending branches of the unconfmed
concrete
s50h strain at 50% strength on the descending branches of the confined
concrete
£ £
c max' c„ extreme fiber concrete compression strain
e
s elastic strain
XXXll
List of Symbols
smj mean strain of a member i

an maximum compression strain in connection concrete when specimen
reaches its ultimate strength
s"n ult ultimate crack strain
s"n crack strain.
p
a plastic strain rate
e ,a yielding strain of steel reinforcement in tension
a tensile strain in connection steel when the wall reach its yield
strength
<j> diagonal reinforcement inclined angle, curvature, the diameter of the
largest steel bar, internal frictional angle
<j)o safety factor between shear strength and flexural strength, initial
angle of internal friction
(j)^ steady state frictional angle
<f>t curvature at horizontal connection due to tolerance between bolts
and bolt holes
<j>u friction angle between the two materials
<j)> post-yield curvature
0 yield curvature
y factors to account for reinforcement strain hardening, shear strain
ysy ratio of yield strength of precast walls to that of their monolithic
counterpart
YA ratio of yield displacement of precast walls to that of their
monolithic counterparts
Ysu ratio of ultimate strength of precast walls to that of their monolithic
counterparts
YAu ratio of ultimate displacement of precast walls to that of their
monolithic counterparts
<p angle between tensile strength and horizontal line, dilation angle
<pe the effective bar size which, for a single bar is equal to the bar size
and, for a group of bars in contact is equal to the diameter of a bar
of equal total area
// curvature ductility
/j ductility factor defined as the ratio of AH/AV , friction coefficient
/jr the concrete-to-concrete friction coefficient under cyclic action
v nominal compression stress, Poisson's ratio
9 inclination angle of the crack, rotation of intervals, angle between
diagonal strut and tie
p tension steel reinforcing ratio
xxxin
List of Symbols
p' compression steel reinforcing ratio

p distributed steel reinforcing ratio
p] ratio of tension reinforcement
pt confinement reinforcement ratio

p h , pr ratios of horizontal and vertical web reinforcement, respectively
p shear reinforcement
£, the normalized neutral axis depth
X plastic multiplier
X] the ratio of actual to specified yield strength
X2 the potential increase resulting from strain hardening
CT normal stress
cr, principal compression stress
ad max m a x i m u m compressive stress acting on the nodal zone
O{K) uniaxial yield strength of concrete as a function of the internal state
variable K .
x peak state of shear stress in friction model
zR{li diagonal tension stress,
rA stress contribution b y concrete to zv ,
ca J Rdi "
T ^ stress contribution by web reinforcement to r„„

C, softening coefficient
A deformation
Av ; , vertical deformations
A;, yield displacement
Au ultimate displacement of specimen at failure
Awn opening component of interface cracks
Aut sliding component of interface cracks
m
]T 4,,- e sum of the areas of the vertical bars in the web and of purposely
arranged bars in the boundary elements of the wall
^ As. the sum of the areas of all diagonal bars in both directions
xxxiv
Chapter 1 Introduction
CHAPTER 1
INTRODUCTION
1.1 Problem Definition
Shear walls which have shown good performance over the past forty years
during severe earthquakes are widely used in concrete structures to resist lateral
loads originating from earthquakes and wind. Composite shear walls, namely steel
reinforced concrete (SRC) shear walls, which incorporate steel sections and
concrete, have been successfully used in the First City Tower in Houston recorded
by Taranath (1985). Today, the manner in which steel reinforced concrete is used is
based on the construction sequence, because the steel frame must be erected before
the concrete is placed. This construction manner requires vertical formwork which
creates potential problems with instability of the steel frame without the concrete
placed. As a consequence construction speed can be reduced. However, construction
speed can be increased if the shear wall is cut into several segments in the vertical
direction and prefabricated in a plant, and connected in an effective manner on site.
Extensive studies have been made regarding the seismic response of reinforced
concrete shear walls. Research on steel frame-reinforced concrete infill systems was
conducted by Makino et al. (1980), Makino (1984), and Tong (2001). Yamada (1990)
conducted experiments on composite frame systems including composite shear
walls. Tupper (1999) studied reinforced concrete walls with steel boundary
elements. The bond between the concrete and steel section in the composite
structure was previously assumed to be the same as that of steel reinforcement and
concrete by some researchers. However, because of the difference of surface
conditions between the steel section encased in concrete and the deformed rebar, the
1
bond slip curve of the deformed reinforcement, which is a critical parameter, may
not be directly used in finite element method (FEM) analysis of composite members.
Although design recommendations for seismic design of composite systems have
been implemented in Japan (AIJ, 1985) and initial design provisions for the
composite wall systems have been proposed in NEHRP (1997), this kind of system
is not widely used in practice all over the world due to lack of sufficient research.
Makino et al. (1980) and Makino (1984) showed that two buildings of this type
behaved satisfactorily during the Great Kanto Earthquake of 1923 in Japan, while
other buildings were badly damaged.
When the concept of "precast" is introduced to composite walls, that is, when
the wall is cut into several vertical segments and connected with horizontal
connections between the segments, the force transfer mechanism may change a lot
from the original. Therefore, experimental and theoretical research programs are
needed to investigate the behavior of precast composite wall system and initiate the
development of appropriate design methodology.
The strut-and-tie models first proposed by Ritter (1899) and Morsch (1902)
have been widely used as powerful tools in the analysis of reinforced concrete
structures. The strut-and-tie model is a unified approach that considers all load
effects including moment, axial force, shear and torsion, simultaneously. It has
evolved as one of the most useful design methods for shear critical structures and
for disturbed regions in concrete structures. The strut-and-tie model provides a
rational approach to represent a complex structural member with appropriate
simplified truss models. Now as a design method for deep beams and shear walls,
the strut-and-tie model is included in AASHTO LRFD Specifications (1994) and
the AC1-318 code (2005). Strut-and-tie models are developed for the specimens in
this research for different loading stages.
2
In current seismic design of reinforced concrete structures, performance-based
design is mainly used, in which the seismic load is directly related to the
performance of the structures. Two performance-based designs, capacity design and
displacement-based design, have been proposed and verified as more efficient in the
seismic design of reinforced concrete structures than strength-based design method.
1.2 Objective and Scope of Research
The objective of this research was to investigate the seismic performance of
composite precast shear walls with horizontal connections in which transfer of
forces is simple and the force can be easily quantified, and to develop related design
recommendations.
Different from previous research by US PRESSS and Soudki (1994) on precast
wall panels with reinforcing bars, the precast walls in this study were walls with
embedded steel sections, channels and I-beams used as a solution for reinforcement
congestion. To simplify assemble operation on site and make it practicable at a high
position of the building, where the quality of welding is difficult to be guaranteed,
bolting connection, a kind of the mechanical connection, was applied in this
research. This will ease the erection of the wall panels without additional props.
Shear connectors were widely studied in vertical connection in US PRESSS. In
Soudki (1994) study shear key or friction between panels were used to resist shear
force along the horizontal connection. In present research, mechanical shear
connectors were used in the horizontal connection and the effect of number of shear
connectors on the performance of the whole horizontal connections was also
investigated in case that the shear force was too high to cause local failure of wall
panel.
3
This research consisted of three parts. The first part was the design,
manufacturing and testing six specimens including two monolithic walls and four
precast walls. Characteristics of these walls including strength, stiffness, ductility
and energy displacement were studied. To predict general behavior of precast walls
with horizontal connections, the author developed a wall model based on macro
model for precast walls with different material models for connection zone and wall
panel. The second part was the development of strut-and-tie models for the
specimens at different loading stages, which provided an explanation for the force
transfer and failure modes and failure zones. The third part was the application of
that finite element analysis to study the behavior of these specimens. This part also
involved parametric study to investigate the effect of some parameters on the
behavior of the walls. These parameters included the types of horizontal connection,
shape of the steel sections, and behavior of the interface between the steel section
and concrete.
1.3 Organization of Thesis
This thesis consists seven chapters. Chapter 1 is an introduction to the
background, objective and scope of this research. Chapter 2 provides a literature
review of reinforced concrete shear walls and composite shear walls, shear
connections, and models of reinforced concrete shear walls including strut-and-tie
models. In Chapter 3, the design, manufacturing and testing of six specimens, two
monolithic shear walls acting as prototypes and four precast shear walls with
horizontal connections acting as counterparts were are described. In Chapter 4,
seismic performance of these tested specimens was evaluated through strength,
stiffness, ductility and energy dissipation. A theoretical model based on macro
model is developed for precast walls and parametric study based on the developed
model has been done. In Chapter 5, based on development of cracking patterns in
4
different loading stages and records of strain located on steel sections and
reinforcement, strut-and-tie models were developed. In Chapter 6, the analysis
using commercial finite element software DIANA, finite element analyses is
described. In Chapter 7, conclusions and design recommendations are given.
5
Chapter 2 Literature Review
CHAPTER 2
LITERATURE REVIEW
2.1 Introduction
Over the last forty years, shear walls have exhibited good performance in
severe earthquakes in which modern reinforced concrete buildings have stood the
test of violent shaking (Fintel 1991, 1995). With increasing recognition of the
benefits of composite structures, composite shear walls using steel sections as
longitudinal reinforcement have been widely developed and studied.
In an attempt to break through the construction sequence, and make composite
structures more flexible, the concept of "precast" will be introduced into this field.
Because shear and bending moment exist simultaneously at any horizontal section
of the shear wall, the connections between the two parts must be able to transfer the
shear and moment reliably.
In order to rationally analyze the behavior of the reinforced concrete members,
strut-and-tie models, using concrete struts to resist compression and reinforcement
ties to carry tension, were proposed by Ritter (1899) and Morsch (1902). In this
chapter, past research on the behavior of shear walls and composite shear walls are
reviewed. Research on shear connections including the effect of the diagonal
reinforcement on the shear behavior of shear walls is also reviewed. Models
(strut-and-tie and wall member models) used in analyzing behavior of shear walls
are also presented. At the end of this chapter the philosophy of seismic design is
briefly introduced.
6
2.2 Reinforced Concrete Shear Walls
The evolution of the modern approach to earthquake engineering of buildings

started in the 1950s, at a time of intense construction after the Second World War.
Most of the analytical research in the 1950s and 1960s on the response of structures
to earthquakes has emphasized the importance of using ductile moment-resistant
frames to reduce seismic forces. Presupposing higher seismic forces in more rigid
structures, and assuming brittle response of shear walls to in-plane lateral forces, it
has been concluded that severe damage could be expected in shear wall buildings.
Based on this thinking, shear walls were considered undesirable for earthquake
resistance at that time. In the meantime, it was found that it is better and more
efficient to incorporate shear walls for structural systems of multistory structures of
concrete resisting wind, (Fintel 1991 and 1995). It was found by Paulay (1980, 1986)
shear wall could work in a relative easy and stable way to dissipate energy if a few
special detail requirements were met.
Fintel (1991, 1995) investigated and reported on the behavior of modern

structures in a dozen earthquakes throughout the world since 1963, and was not
aware of single concrete building containing shear walls that had collapsed. The
conclusion was buildings containing shear walls, even if only conventionally
reinforced, withstood severe earthquakes, mostly without damage.
Experiments investigating the behavior of reinforced concrete (RC) walls under

earthquake loading have been conducted mainly in the U.S., Japan, New Zealand,
and Europe. The pioneers in research on RC walls were Benjamin and Williams
(1957), who tested squat walls. Two reports by Regan (1971) to CIRIA included a
significant amount of state-of-the-art research on shear in the U.S. and U.K.
7
Experimental data from Oesterle et al. (1979) demonstrated that increasing the
amount of horizontal web reinforcement was not sufficient to avoid the mode of
shear failure. This experiment showed that increasing the amount of horizontal web
reinforcement by a factor of two had a negligible influence on the hysteretic
response of these walls.
Oesterle (1980) found that in ductile structural walls, the shear deformation in
the plastic region may be significantly larger than in other, predominantly elastic,
regions. Shear resistance of reinforced concrete mainly comes from elastic region of
the walls under cyclic loads suggested by Corley (1981), which gives implication of
the layout the vertical reinforcement.
The flexural response and effect of reinforcement distribution on behavior of
squat walls were studied by Paulay (1975). The mechanism of shear resistance was
studied by Barda (1972). It showed that a significant portion of the shear acting on
the top of the squat wall was directly transmitted to the bottom by diagonal
compression. Failure modes and performance-improving methods including
diagonal reinforcement were widely studied by Priestley and Paulay (1982).
In order to investigate the effect of parameters such as the height-to-width ratio,
the axial load, the concrete strength, and the amount of web horizontal
reinforcement on wall behavior, thirteen large-scale wall models were tested under
the combined action of a constant axial load and a horizontal load monotonically
increasing to failure by Lefas et al. (1990). The experiment included structural walls
with constant thickness and a height-to-width ratio varying between 1 and 2. Such
walls were considered to represent the critical story elements of a structural wall
system with a rectangular cross section. Three levels of constant axial load were
adopted in the testing program: 0.0, 0.1 and 0.2 of the uniaxial compressive strength
of the wall cross section that was equal to 0.85/CMW (where b is the thickness and
/is the width of the wall, and /^ is the compressive strength of the concrete cube),
to represent the amount of axial load at the base of the wall of a single story, a
medium-height, and a high-rise building, respectively. It was found that while axial
compression reduces both vertical and horizontal displacement, it causes an
increase in horizontal load-carrying capacity and secant stiffness. In contrast to
what is widely believed, the horizontal web reinforcement did not appear to have a
significant effect on shear capacity, as suggested by Wood (1990). The strength and
deformation response of the walls were found to be independent of the uniaxial
concrete strength characteristics within a range of 30 MPa to 55 MPa. Shear
resistance appeared to be associated with the development of triaxial compressive
stress conditions within the compression zone near the base of the wall where the
flexural moment attained its maximum value. This conclusion was consistent with
that proposed by Kotsovos (1985) who demonstrated that the behavior of an
element of concrete in a structure can not be described in terms of uniaxial
stress-strain characteristics. Such a description not only unnecessarily complicates
code provisions for ultimate limit state design but also is not theoretically correct.
Design provisions should be based on an understanding of the causes of the
observed structural behavior and this can only be achieved by considering the
triaxial stress conditions that always develop within a structure under load.
In order to investigate the effect of loading history and repair methods on the
structural characteristics of reinforced concrete walls, four identical wall specimens
were constructed at a scaled 1:2.4 with constant thickness and a height-to-width
ratio equal to 2 and tested by Lefas and Kotsovos (1990). It was found that, while
repairing only the damaged regions of the compressive zone was sufficient to fully
restore wall strength, the additional use of epoxy resins to heal major flexural and
9
inclined web cracks led to only a marginal improvement of the structural
characteristics such as stiffness and energy dissipation at the serviceability limit
state, both being distinctly inferior to those of the original walls under cyclic
loading. Such results are in compliance with the concept of the compressive path
and demonstrate that, in contrast to widely held views, the compressive zone is the
main contribution to shear resistance.
To quantify the true ductility supply and energy absorption potential of RC
walls, six RC walls at a 1:2.5 scale were tested under cyclic load by Pilakoutas and
Elnashai (1995). Various types of failure were observed, depending mainly on the
amount and distribution of shear reinforcement. Shear force was partly transmitted
by the concrete in compression and partly by the link reinforcement that enabled
shear stress to be resisted through the concrete in the tensile zone. And the strength
and deformation characteristics of specimens were not affected significantly by
shear reinforcement in excess of the amount required to resist the maximum applied
load. Structures designed to sustain extreme forces inelastically should be analyzed
based on reduced cracked section stiffness by taking into account ductility
considerations. The effect of confinement reinforcement on the ductility was more
profound than on the moment capacity. By decomposition of the total deformation
in flexural and shear components, it can be demonstrated that the major bulk of the
energy dissipation is due to flexure. Net longitudinal extension was expected in
flexural RC elements not subjected to significant axial loads due to cyclic loading.
Most of the extension took place in the lower part of the wall, where moments
exceed the yield capacity. Longitudinal strains on reinforcement near the top of
cantilever shear walls were found to be significantly higher than analytically
expected. This indicates that an arch-and-tie mechanism was also mobilized for
shear resistance.
10
Tasnimi (2000) tested four identical 1:8 scale wall specimens with constant
thickness and a height-to-length ratio equal to 3. These specimens were
experimentally subjected to various slow cyclic horizontal loading regimes. The
crack patterns and failure modes of all specimens were not fully in compliance with
the concept of the compressive force path reported by Lefas and Kotsovos (1990).
And this indicates that the capacity of walls is affected by the flexural as well as
shears strength of the specimen. The strength and deformational responses of
specimens were found to be independent of the cyclic loading sequence.
There has been a great deal of research done on the shear strength of shear
walls, deep beams, and corbels. Formulas and recommended rules for design
abound in the different building codes and research publications. Siao (1994)
attempted to unify the results of all the research efforts to achieve a single
consistent approach in the design and analysis of such important structural members.
Using formula established for top-loaded deep beams and corbels, the predicted
shear capacity of reinforced concrete shear walls showed reasonable good
agreement with the experimental results.
2.3 Composite Shear Walls
2.3.1 Composite Members
Steel reinforced concrete (SRC) composite structures used to resist lateral load
due to earthquake or wind have been widely explored and developed, with
increasing recognition benefits from several different types of composite structures.
Wakabayashi (1986) presented that "buildings of composite construction showed
good earthquake resistant capacity under the Kanto earthquake (1923) as compared
with ordinary reinforced concrete structures".
11
In addition, the 1994 Northridge earthquake revealed some unexpected
problems regarding the seismic response of the steel moment-resisting frames,
especially their connection regions (FEMA, 2000). Cracks were found to occur in
the fully-restrained girder-flange-to-column-flange weld and even to propagate into
the column flange and web, leading to expensive structural repairs. A series of
solutions have been proposed to eliminate or reduce these connection problems
directly, such as the removal of backing bars and the use of high quality welding
materials (FEMA, 2000). However, a final consensus has not been reached in the
earthquake engineering community regarding modifications to the design and
construction procedure for moment-resisting frames. An alternative method is to
reduce the seismic demands on steel frames by avoiding the use of fully-restrained
connections. This philosophy led to the development of the composite steel
reinforced concrete wall system, which originates as reinforced concrete shear walls
with encased flat steel bars, steel trusses and steel plates (Tall Building Committee
A41, 1979), utilizing the high shear stiffness of the reinforced concrete wall to
attract most of the lateral shear forces from earthquake motions.
In composite steel-concrete lateral load resisting system, the underlying
principle in bringing these distinctly different materials together is to benefit the
advantages of both: the inherent stiffness and economy of concrete, the speed of
construction of a conventional steel building. The use of composite construction
resulted in greater ductility with the load-carrying capacity. The composite system
has the desired characteristics of the conventional reinforced concrete system, such
as high strength, stiffness, ductility, and fire resistance, and has been found to be
cost effective for a range of building heights under certain design conditions.
The Composite Structural Steel and Reinforced Concrete Buildings Provisions
12
of AISC (1997) and the National Earthquake Hazards Reduction Program (NEHRP.
1997) provide some design provisions for boundary element shear walls.
In order to make clear the structural characteristics of steel profile encased

reinforced concrete composite structural systems developed widely as high rise
buildings in the half century in Japan, Yamada (1990) studied the steel encased
reinforced concrete composite frames. Tests on steel profile encased reinforced
concrete composite 3-span, multi-storey rigid frame systems with or without shear
wall models were carried out until fracture and compared with ordinary steel or
reinforced concrete rigid frames with or without bracing or shear walls. Initial
stiffness and ultimate horizontal resistance of these composite systems were
between those of the corresponding steel frame or reinforced framework systems.
This may enable to present comfortable states for usages in high-rise buildings.
Tupper (1999) studied the reinforced concrete walls with steel boundary
elements. Three specimens meeting the requirements of standard reinforced
concrete ductile flexural walls were constructed and tested under reversed cyclic
loads. One of the composite walls used rectangular hollow structural sections (HSS)
as boundary elements which were connected to the wall by welding the transverse
bars directly to both HSS elements. The second wall used steel channels connected
to the wall with headed studs welded to the channels. These studs overlapped with
the transverse reinforcing bars, which had headed ends. The details of these three
walls were chosen such that all of the walls had approximately the same flexural
capacity. The reversed cyclic response of the three walls showed that each wall had
similar hysteretic properties. And the design used for ductile flexural walls was
modified to enable comparable design of reinforced concrete walls with steel
boundary elements.
13
Another kind of widely used composite shear wall is steel frame with an infill
of reinforced concrete. Although initial design provisions for the composite steel
frame-RC infill wall systems have been proposed in NEHRP (1997), this system is
rarely used in practice in the United States due to lack of sufficient research. In
Japan, design recommendations for earthquake-resistant design of this type of
composite system have been implemented (AIJ, 1985). Makino et al. (1980) and
Makino (1984) showed that two buildings of this type behaved satisfactorily during
the Great Kanto Earthquake of 1923 in Japan, while other buildings were badly
damaged.
Chrysostomou (1991) pointed out the importance of composite shear
connection between the infilled concrete and steel boundary elements. Without
adequate shear connection, the storey shear was primarily resisted by a compression
strut in the concrete panel. This compressive strut had finite width and was aligned
with the corners of the panel. In contrast, frames with adequate shear connectors
were able to resist the shear with a field of diagonal compression in the concrete
rather than one single strut.
In the composite steel frame-RC infill system, the steel columns and girders
serve as boundary members to resist gravity loads and most of the overturning
moment, while the RC infill walls carry most of the shear forces. The RC infill
walls had the benefit of increasing the lateral stiffness dramatically, thus avoiding
excessive drift and reducing the seismic demands on the steel frames. This provides
the opportunity to use partially-restrained (PR) connections in conjunction with
lighter steel framing instead of more expensive fully-restrained connections, which
are required in the more common steel moment frames (without infill walls).
Adequate interface connections, such as headed shear connectors, can be used to
transfer shear forces uniformly between the steel frames and infill walls to ensure
14
effective composite action. This reduces the probability of unfavorable failure

mechanisms such as the premature formation and crushing of the compression struts
in the RC walls. Experiments conducted by Makino et al. (1980) and Makino (1984)
indicated that this system develops a pattern of closely-spaced cracks with small
widths. Such cracks may be repaired at a reasonable cost after a moderate
earthquake.
Tong (2001) studied the behavior of composite steel frame-reinforced concrete

infill wall structure system. This research conducted a cyclic loading test on one
two-story, one-bay specimen, which represented the bottom two stories of a
six-story prototype structure at approximately one-third scale. Partially-restrained
(PR) connections were used to join the wide flange steel columns and beams. The
infill was cast using regular concrete. Confined headed studs were used as the
interface connectors. The test showed that this system provided sufficient strength
to resist the design lateral loads and ample stiffness to control system drift.
Adequate redundancy was provided to the system by alternate load paths, such as
shear stud-infill interaction and truss action due to steel frame-infill interaction. At
the design lateral load level, 80-90% of the lateral shear force was transferred from
the steel frame to the infill through the headed studs, and approximately 80% of the
overturning moment was carried by the steel columns. The ultimate strength of the
specimen was induced by crushing of the concrete in the corners, yielding and
fracture of the studs, and formation of plastic hinges in the partial rigid connections.
Adequate confining reinforcement was critical to enhance the strength and ductility
of the studs in the infill panels. The failure of the studs was controlled by the low
cycle fatigue.
Realizing the connection between the composite columns and concrete

beam-slab floor as well as the shear-wall edge members and concrete wall are
15
critical, Shirali (2002) studied the connection design at the interface of the different
elements. Ten alternative interface designs, reflecting a one-third scale of the edge
region of the first story wall of an 8-story building, have been developed and tested.
In a nonlinear FEA, two models, one a nonlinear spring model and the other a
truss-like model, have been developed to capture the interface behavior between
edge columns and RC wall panels.
Today, the manner of use of steel reinforced concrete walls is based on

construction sequence, because the steel frame must be erected before the concrete
is placed. The construction of this type of composite system needs a lot of vertical
formwork, costly labor work and also causes instability of the steel frames
exceeding concrete. In order to make such a shear wall more flexible and applicable
and improve construction speed, the concept of "precast" is introduced into this
field in current research. Known from the precast structures, the connection must
have the ability to transfer the force between adjacent members reliably.
2.4 Connection
The design and construction of connections is the most important consideration

in precast concrete structures. Their purpose is to transmit forces between structural
members and/or to provide stability and robustness. There are several different
ways of achieving a satisfactory connection, e.g. bolting, welding, or grouting, but
the method should be simple and must convey unambiguous messages to the site
operative.
In the horizontal connections of shear walls, the shear and flexural moment
exist simultaneously. The flexural moment at the connection can be readily
transferred by connecting vertical steel bars or plates on different parts by several
16
means, for example, welding, coupler, bolting. Since the aim of the shear wall is
to resist the lateral force transferred from other structural members, how to deal
with the shear connection is the focus in design of the horizontal connections of
shear walls.
Ductile connections used precast large panel systems has been investigated in
US PRESSS program. The vertical ductile connectors were deigned weaker than
wall panels and exhibited some energy dissipation characteristics.
2.4.1 Shear Connection
Shear transfer may be achieved locally using mechanical shear connections. In

order to ensure high shear stiffness, the connection is made either by site-welding
embedded plates, or by tightly clamping using friction-grip bolts. When gravity
action exits on the horizontal connection of precast structure, the friction between
two surfaces is an important portion to resist the relative displacement of two parts
of the shear wall. This kind of connection has been studied by Soudki (1994) and is
out of the scope of present research.
Foerster et al. (1989) studied shear connections for precast panels, two of these
connections used continuity reinforcement bars and mechanical shear connectors.
Behavior of the specimen suggested that the presence of the mechanical shear
connectors enhanced the clamping action and overall ductility of the joint, but one
type of connector failed in anchorage of the bars. Only the behavior of the
connection under monotonic load was studied in the research.
2.4.2 Diagonal Reinforcement
17
In order to effectively transfer the shear force between two parts of the wall,
two kinds of reinforcement are usually welded to the embedded steel plates. One is
vertical reinforcement, the other is diagonal reinforcement.
On the experimental side, work on the use of diagonal reinforcement in RC
walls was first studied by Iliya and Bertero (1980) who used grids of diagonal
reinforcement in the web of their specimens. They pointed out that the cost of
constructing diagonally reinforced walls was higher because the bars must be cut at
varying lengths and the placing and anchorage of these bars was more difficult than
placing in vertical and horizontal bars. Nevertheless, their tests indicated that the
45° arrangement of the wall reinforcing bars was more effective in resisting the
effect of shear reversals, i.e., there was less stiffness deterioration with displacement
reversals. In these specimens with the same area of shear reinforcement, in
conventionally reinforced specimens' failure was mainly due to diagonal cracking,
whereas in the specimen with diagonal web reinforcement the failure was largely
flexural.
In order to reduce excessive sliding shear displacements under cyclic reversed
loading, particularly at the wall's base immediately above the base beams, two walls
with diagonal reinforcement were tested by Paulay et al. (1982). These walls
duplicated two other conventionally reinforced walls that had approximately equal
flexural and shear strengths. Tests of walls have shown the detrimental effects of
excessive sliding shear displacements as well as the marked improvement of
response when some diagonal reinforcement crossing the sliding plane was used to
reduce displacements and increase the forces resisting sliding shear. The dramatic
reduction of the strength, particularly in the second cycles of loading to the same
level of ductility was evident in the wall without diagonal reinforcement. The
cross-inclined diagonal reinforcement was effective in both tension and
18
compression and considerably improved seismic response of squat shear walls,
although only 30% of the applied shear was resisted by such reinforcement.
However, when the diagonal bars yielded due to slip displacements, with
cumulative ductility the test unit with diagonal reinforcement dissipated
approximately 70% more energy than its counterpart without diagonal
reinforcement.
Experimental studies by Fiorato et al. (1983) at the Portland Cement
Association included tests of a previously damaged wall that was repaired by
removing the web concrete, adding diagonal reinforcement, and placing new
concrete. Test results indicated that the deformation capacity of this wall which was
susceptible to web crushing was increased by adding the diagonal reinforcement.
The observed degradation in stiffness with cycling was also less in the repaired
specimen.
Because the available experimental data concerning the behavior of the squat RC
walls subjected to high cyclic shear were inconclusive, especially with respect to
the sliding shear failure and the role of inclined bars in preventing it, Salonikios et
al. (1996, 1998, 1999, 2000, 2002) tested 11 specimens which were reinforced
against shear, either conventionally (orthogonal grids of web reinforcement), or
with cross-inclined bars. The effect of web and edge reinforcement ratio, axial load
level, and quality of construction joints were also investigated. Two sets of
specimens were designed: the first consisted of five specimens with an aspect ratio
1.0 (LSW series) and the second set consisted of six specimens with an aspect ratio
1.5 (MSW series). These specimens were designed with respect to the EC8 (1995),
which adopted different equations for design against each possible shear failure
mode, i.e., diagonal tension (truss mechanism), diagonal compression (crushing of
concrete struts), and sliding. LSW2 and LSW3 had the same web reinforcement;
19
LSW1 was similar to LSW2, but had double the web reinforcement. LSW4 and
LSW5 both included bidiagonal reinforcement. LSW5 was similar to LSW2 in
every respect, except three bars inclined by 45 degree were added. In order to yield
a meaningful comparison of the relative effect of the two types of shear
reinforcement, the bidiagonal reinforcement was made in such a way as to achieve
an effective reinforcement ratio for bars crossing the potential (horizontal) sliding
plane almost identical to that of the vertical web reinforcement in LSW1. LSW4
had the same amount of the bidiagonal reinforcement as LSW5, but the intersection
of the 45 degree inclined bars was at the base of the wall. MSW1 to MSW5 were
reinforced in a similar way as LSW1 to LSW5. As for MSW6, the base and main
part were cast at different times and the specimen reinforcement was connected at
the base by lap splicing, without concrete roughening and or additional
reinforcement. This allowed studying the effect of a poorly detailed cold joint. All
specimens failed in a predominantly flexural mode, characterized by concrete
crushing and reinforcement buckling at the confined edges. Pinching of the
hysteretic loops was significant in the conventionally reinforced specimens and
considerably less in the specimens with diagonal reinforcement.
On the analytical side, Sittipunt and Wood (1995) used finite element method to
study the influence of the web reinforcement on the cyclic response of structural
walls. In U. S. building codes, the nominal shear strength of slender walls is
assumed to increase in proportion to the amount of horizontal web reinforcement.
However, many walls subjected to cyclic loading failed due to web crushing after
yielding in flexure. So, in Sittipunt's study, six walls with different web
arrangements were analyzed. Walls in which only the amount of vertical or
horizontal reinforcement was changed exhibited the same overall response. Large
average shear strains were calculated at the base of the walls, which led to severely
pinched hysteretic curves. These calculations indicated that reasonable increases in
20
the amount of the horizontal or vertical web reinforcement are not sufficient to
improve the cyclic response of walls or to reduce the shear distortion at the base.
Walls with diagonal reinforcements exhibited better behavior, lower shear strains in
the lower portions of the walls. Through simulation, the authors found that: for the
walls with diagonal bars, the applied load versus top deflection curves were more
rounded, the shear stiffness did not degrade appreciably with cycling, the shape of
the shear distortion curve was improved by decreasing the region of extremely low
shear stiffness, and the amplitude of the average shear strain in the lower portion of
the walls was reduced. The presence of the diagonal reinforcement did not change
the calculated strength of the walls appreciably. This significant improvement in
performance may be explained by considering that the diagonal reinforcement runs
nearly perpendicular to the cracks that develop in the web of the wall. Therefore,
after cracking, most of the force is carried across the cracks by the diagonal
reinforcing steel in tension. Under uniaxial vertical loading, the reinforcement
displayed well-rounded hysteretic loops and was capable of dissipating energy
efficiently. In the walls with only vertical and horizontal web reinforcement, the
bars were oriented 30 to 45 degrees to the cracks, and most of the load initially must
be carried across the cracks through dowel action. Force transfer through dowel
action was characterized by extremely pinched hysteretic curves and stiffness
degradation. This phenomenon-pinched effect caused by force transferred by dowel
action has been studied by Elliot et al. (1974), Vallenas et al. (1979) and
Jimenez-Perez et al. (1987).
Collins et al. (1996) suggested a general methodology for the design of beams
and columns against shear. The shear strength in a section defined by an inclined
crack is given by the formula:
V„ =4cot0/[l + j5OO£l](JZ)bvdv+AJvdvcot0/s (2-1)

where 6= the inclination angle of the crack,
21
£•, = the principal tensile strain,
fc = the concrete's compression strength,
bv = the effective web width,
dv = the effective shear depth,
Av = the area of the shear reinforcement within distances ,
/ = the steel yield strength,
s = the spacing of shear reinforcement.
Salonikios et al. (1996, 1998, 1999, 2000, 2002) tested 11 specimens and gave
the relevant methodology. During wall deformation, there were three fundamental
mechanisms that resisted the imposed deformations. These were the flexural,
diagonal shear and sliding shear mechanisms. Although these mechanisms were
assumed for simplicity to act independently for the design of the structure, it has
been experimentally shown that they strongly interacted with each other. Especially
the flexural and sliding shear mechanisms were directly coupled, mainly after heavy
flexural cracks. Across the horizontal cracks yield of the sliding shear mechanism
was observed because the other resistance mechanisms became less important while
cracks propagate and cyclic forces were imposed. The mechanisms that gradually
became less important were friction that developed at the crack surface, aggregate
interlock, and dowel action.
For inclined shear cracks as shown in Figure 2.4.2-1:
y
£dML = O=>Tidv+Tddvcot0 + Thdvcot0/2 + T,dv/2 + 2Tdgsmyd/2
= Vudvcot0 + M,l-NdJ2 (2_2)
,
=>r, +7 (/cot6'/2 + r ; / 2 + 2r arg sin^/2
= Vlico\9 + MJdy-NJ2
X Fx =0^V„=Th+ Tfi sin B + Tc/h +Td+ Tdg cos </> (2-3)
22
Y,Fy=0=>Nu= Tfr cos e + Tc-T,-T,- Tdg sin 0 (2-4)
where Ti = the tensile strength of the flexural reinforcement,

Ta = the dowel action strength of the horizontal web reinforcement,
Th= the tensile strength of the horizontal web reinforcement and dowel
action of vertical web reinforcement,
Ti = the tensile strength of vertical web reinforcement and the dowel
action of the horizontal web reinforcement,
Tdg = the tensile strength of bidiagonal reinforcement,
Tfr = the aggregate interlock and friction resistance force, along the crack
in the web,
Tcrv = the aggregate interlock and friction resistance force at the edge
column,
Tc = the compressive force at the edge column,
<f> = diagonal reinforcement inclined angle,
<p= angle between tensile strength and horizontal line.

Considering the influence of reduced axial force and aspect ratio,
Tcfi. = 2.15v(0.75Mmax ldv+Nu 12)1 a, (2-5)
Td=0385(fuAsfl)as (2-6)
where v= NJfcAc,
as=MJ(VJJ,
fu= the tensile strength of reinforcement,
Asjl = the area of edge column longitudinal reinforcement.
In specimens with bidiagonal reinforcement:

Z F< = ° => V«« = T0- + [] + 01/ < /(0.8/„.)](0.27r, + 2Tdg cos <fi)as (2-7)
23
where dg T,=f Asd'•,

J y vsd
n
Asd = the area of the diagonal reinforcement in direction,
/,= the distance between the inclined bars' center axes, at the base cross
the section,
lw= the length of the wall.
0 i
f>^ vr.^ d, cotG
Fig.2.4.2-1 External acting forces and internal reactions at shear wall inclined crack in
direction 0
From the above experimental and analytical experience, the effect of diagonal
reinforcement on reducing shear deformation, stiffness degradation under cyclic
loads, and increasing the shear strength is apparent. So the diagonal reinforcement
was selected in the present study to control the shear deformation (relative
horizontal displacement of two panels) of the specimens.
2.5 Strut-and-tie Model
Rational design of reinforced concrete structures must be based on the

mechanical properties of the materials. Concrete has high strength in compression
but very low strength in tension. When concrete is used in a structure to carry loads,
24
the tensile regions are expected to crack and, therefore, must be reinforced by
materials of high tensile strength, such as steel. This concept of utilizing concrete
struts to resist compression and reinforcing ties to carry tension gave rise to the term
"strut-and-tie model". In this model, the concrete compression struts and the steel
tension ties form a truss that is capable of resisting the applied loads. As such, this
concept is also called the "truss model". This method implies that the structure is
designed according to the lower bound theorem of plasticity. Because concrete
permits only limited plastic deformations, the internal structural system (the
strut-and-tie model) has to be chosen in a way that the deformation limit (capacity
of rotation) is not exceeded at any point before the assumed state of stress is
reached in the rest of the structure.
2.5.1 History of Strut-and-Tie models
Ritter (1899) and Morsch (1902) conceived the truss model concept, the angle
of the concrete struts was conveniently assumed to be 45°. After Mohr's
compatibility condition was established, however, it was realized that 45°
inclination of struts represents the equal strain condition of the longitudinal and
transverse steel.
Shear design procedures for reinforced concrete that, like tension field theory,
determine the angle 0 by considering the deformations of the transverse
reinforcement, the longitudinal reinforcement, and the diagonally stressed concrete
have become know as compression field theory (CFT). Kupfer (1964) and Baumann
(1972) presented approaches for determining the angle 0 assuming that the
cracked concrete and the reinforcement were linearly elastic. Based on the theory of
plasticity, Nielson (1967) and Lampert and Thurlimann (1968) used the membrane
elements to analyze the RC structures. They first established the two-dimensional
25
equilibrium condition of reinforced concrete membrane elements subjected to shear

and normal stresses. Elfgren (1972) developed a truss model which was based on
the equilibrium condition that assumed both the longitudinal and the transverse steel
yield at the ultimate load stage. The term "compression stress field" conveys the
assumption that the direction of the concrete struts coincides with the direction of
the post-cracking principal compression stress. Methods for determining 9
applicable over the full loading range were developed by Mitchell and Collins
(1974) for members in torsion, and were applied to shear design by Collins (1978).
Prior to 1972, the truss model severely overestimated the shear and torsion
strengths when the failure stress of the concrete struts was assumed to be the
uniaxial compression strength of standard concrete cylinders. Robinson and
Demorieux (1968, 1972) observed that a reinforced concrete element under shear is
actually subjected to a two-dimensional stress condition. The strength in the
principal compression direction was found to be softened by the principal tension in
the perpendicular direction. Vecchio and Collins (1981) proposed the first softened
stress strain curve of concrete in compression. This curve incorporated a softened
coefficient, which is a function of the principal tensile strain. The softened
stress-strain curve of concrete was used in combination with Mohr's stress and
strain circle to predict the behavior of reinforced concrete elements. Collins used
the "compression field" to reflect the assumption that the direction of the concrete
struts coincides with the direction of the post-cracking principal compression. In
order to derive the "contribution of concrete", Vecchio and Collins (1986) proposed
the "modified compression field theory (MCFT)".
On the basis of equilibrium, compatibility and constitutive laws of material,

"softened truss model" was developed by Hsu (1988) and Pang and Hsu (1996).
Hsu (1996) proposed a unified nomenclature that covered all four basic actions
26
(bending, axial load, shear and torsion), which was based on the angle of inclination
of the concrete struts, the scope of applications, and three fundamental principles of
the mechanics of materials (stress equilibrium, strain compatibility and constitutive
laws of materials). Six models, including a strut-and-tie model, a Bernoulli
compatibility truss model, an equilibrium (plasticity) truss model, a Mohr's
compatibility truss model, a softened truss model, and a fixed-angle softened-truss
model were provided in this study. The constitutive laws of concrete required in the
rotating angle softened truss model have been formulated as shown in Fig. 2.5.1-1
by Belarbi and Hsu (1995) and Pang and Hsu (1996). The average stress-strain
curve of concrete in compression is given as:
(2-8)
* 2 = &
£2
fro &0
y««
i- # C yr >i (2-9)
° 2 = &
V-l
0.9
C- Vl + 400f, (2-10)
where fc =maximum compressive strength of standard 6><12-in concrete
cylinder,
£0=concrete strain at maximum compressive strength taken as 0.002,
Q =softening coefficient.
27
7^77"-^ _, NONSOFTENED
X
/ / I
Fig.2.5.1-1 Compressive stress-strain curve of concrete
Based on the strut-and-tie concept, equilibrium, compatibility and constitutive

laws of cracked reinforced concrete, Hwang and Lee (1999, 2000) and Hwang et al.
(2000, 2001) proposed a softened strut-and-tie model to predict the strength of the
beam-column joints, squat walls and deep beam; and the effect of the boundary
elements, cyclic loading and vertical loads on the wall and shear strength also were
studied. This model was used to predict the shear strength of deep beams and
squat walls. Each analytical result agreed well with the corresponding experimental
result in the study. After the development of the first cracking pattern in the wall,
the steel bars were subjected to tension and the concrete acted as compressive struts,
thus forming a strut-and-tie action. Three strut-and-tie load paths were proposed to
model the force transfer within the squat wall, and they are diagonal, horizontal, and
vertical mechanism as shown in Fig.2.5.2-2.
28
J^x^,
Steep Strut
Vertical tie
Fig.2.5.2-2 Wall shear resisting mechanisms: (a) diagonal; (b) horizontal; (c) vertical
The diagonal mechanism is a single diagonal compression strut whose angle of

inclination 0is defined as:
# = tan" (2-11)
v I« J
It is assumed that the direction of the diagonal concrete strut coincided with the
direction of the principal compressive stress of the concrete. And the effective area
of the diagonal strut Astr is defined as:
Atr =asXbs (2-12)
29
where as = depth of the diagonal strut,
bs = width of the diagonal strut that can be taken as the width of the wall
web tw.
The depth of the diagonal strut depends on its end condition provided by the
compression zone at the base of the wall. It is intuitively assumed thato^ =aw,
where ctw is depth of the compression zone at the base of the wall. For simplicity,
aM.can be approximated with Paulay and Priestley's (1992) equation for the depth of
the flexural compression zone of an elastic column.
( N }
a,., = 0.25 + 0.85 I (2-13)
V Amfe J
where Aw= net area of the concrete section bounded by the web thickness tw
and the length of the section in the direction of the shear force lw,
fc= compressive strength of the concrete.
N=vertical force acting on the wall.

The horizontal mechanism included one horizontal tie and two flat struts as
shown in Fig.2.5.2-2. The horizontal tie was made up of horizontal shear
reinforcement. When computing the cross sectional area of the horizontal tie, it is
assumed that the horizontal shear reinforcement within the center half of the wall is
fully effective, and the other horizontal steel is included as 50% effective. The
vertical mechanism is composed of one vertical and two steep struts. The vertical tie
includes only the vertical shear reinforcement within the wall web and excludes the
vertical reinforcement of the boundary elements. For a wall without boundary
elements, the vertical shear reinforcement within the central portion of 0.8 lw is
considered effective to constitute the vertical tie. For diagonal compression failure,
30
the shear strength of the squat wall is defined as the concrete compressive stress on
the nodal zone as the concrete reaches its capacity. The boundary of the nodal zone
coincides with the boundary diagonal strut, but the concrete bearing force to be
examined is the summation of the compression from the diagonal, flat and steep
struts.
Because the diagonal compression is mainly transferred in D-direction shown
in Fig.2.5.2-2, the maximum compressive stress acting on the nodal zone:

( H_ 2H
COS 0-tan" COS tan" e
d.max
1
D- Y J) F - h f
jj
F (2-14)
H 2H^
cos tan" sin tan"
Y I
V < J
Where D is depth of diagonal strut,
Fh is sum offerees in horizontal reinforcement;
Fv is sum of force in vertical reinforcement in central 0.8/M
2.5.2 B Region and D Regions
The body of a reinforced concrete structure can be visualized as having two
typical regions: B regions and the D regions.
• B regions are referred to the main regions in which the Bernoulli
hypothesis of plane strain distributions is assumed valid. That is, the
stresses and strains in the main regions are governed by simple equilibrium
compatibility conditions. The internal state of stress of these regions is
easily derived from the sectional forces (bending and torsional moments,
shear and axial forces). It is logical to apply truss model to the main
regions using a regular angle of concrete struts.
• D regions are local regions where the strain distribution is significantly
nonlinear and compatibility conditions are difficult to apply, e.g., near
31
concentrated loads, corners, bends, openings and other discontinuities. The

main types of discontinuity are geometric discontinuity and load
discontinuity as shown in Fig.2.5.2-1 and Fig.2.5.2-2, respectively. In
these regions the stresses and strains are so disturbed and irregular that they
are not easy to get a mathematical solution. In the design of local regions,
the stresses are determined by equilibrium conditions alone, while the
strain compatibility conditions are neglected. So, the strut-and-tie model
with irregular angles is applied to the local regions. The local region is
treated as an isolated free body subjected to external boundary forces.
Boundary forces are calculated from the four actions diagrams including
bending moment and torsion, shear and axial force at the boundary
sections.
H~\
WKM -&
immmm 3
= c-f lllltliii
.&.
Fig.2.5.2-1 Geometric Discontinuity (ACI 318-02, Appendix)
32
ft
i L \ u \ L w w y w w W w w
Fig.2.5.2-2 Load Discontinuity (ACI 318-02, Appendix)
2.5.3 Design Procedure of Strut-And-Tie Model
In designing a reinforced concrete structure, strut-and-tie models can be used to

directly design members and detail. The design procedure is shown in Fig.2.5.3-1
keeping in mind then following requirements:
1) Equilibrium must be maintained
2) Tension in concrete is neglected
3) Forces in struts and ties are uniaxial
4) External forces apply at nodes
5) Prestressing is treated as a load
6) Detailing for adequate anchorage
33
Design Procedure and a flowchart is shown in Fig.2.5.3-2:
1) Determine loads and reactions and estimate dimensions and member sizes.
Considering design requirements of clients and actual surrounding natural
conditions, structural system is selected. The loads and reactions will be
gotten through calculation. According to requirements of design codes,
dimensions and member sizes can be estimated.
2) Define B- and D-regions and isolate each D-region. In using the
strut-and-tie model approach, it is helpful and informative to first subdivide
the structure into its B- and D-regions. The truss model and the design
procedure for the B-regions are then readily available and only the
strut-and-tie models for the D-regions remain to be developed and added.
In defining B- and D- regions, the definition of geometric discontinuity and
loading discontinuity is helpful.
3) Compute resultant forces on each D-region boundary. These resultant
forces include resultant forces from adjacent B-regions and reactions from
supports.
4) Select a truss model to transfer the resultant forces across the D-region.
The axes of the struts and tie, respectively, are chosen to approximately
coincide with the axes of the compression and tension fields. The forces in
the struts and ties are computed.
5) The effective widths of the struts and nodal zones are determined
considering the forces from step 4 and the effective concrete strengths and
reinforcement is provided for the tie considering the steel strength. The
reinforcement should be anchored in the nodal zones.
34
Define Structural System
Es timate Dimensions and Member Sizes
T
Defien B- and D-Regions in the Structure
"
Compute Resultant Forces and Reactions on
Boundaries of D-regions
v
Develop Strut-and-Tie Models for D-Regions
'
T i' <' T
Element Forces and Forces and Tie Details

«-• (-•
Dimensions Stress in * - * •
Stress in for Achorage

Nodes Struts
Fig.2.5.3-1 Design Flow Chart of Reinforced Concrete Structure
1—:
i
1 •
*^WT
-—i?—. U
a) B- and D-Regions (AC1318-02, Appendix)
h
T
35
' .•' • / v '

\
I* • \
T b) Strut-and-Tie Model for Entire Beam (ACI318-02, Appendix)

T
ff" --• jQ — 4
* Compression Strut
*
•• Tmneion Tie
* *
o Node
£-
J c) Detail at End of Beam
Fig.2.5.3-2 Strut-and-Tie Model for Simple Span Beam (ACI 318-02, Appendix)
2.5.4 Method for Developing Strut-and-Tie Models
A. Elastic Analysis Approach

The struts and ties in strut-and-tie models condense the real stress field by
resultant straight lines and concentrate their curvature in nodes, which means that
developing the model is much more simplified if the elastic stresses and principal
stress directions are available. Such an elastic analysis is readily facilitated by the
wide variety of computer programs available today. The direction of struts can be
taken in accordance with the mean direction of principal compression stress or the
more important struts and ties can be located at the center of gravity of the
corresponding stress diagrams.
36
M la li M
T "I 1 I j l l
-™£ -w-j f"'k i^mk wl- nuw i.nni
-j—t—t
strut
lie
Fig.2.5.4-1 Strut-and-Tie Based on Elastic Stresses (Schlaich, 1987)
7—?!
m
i 5
»
n
! !
J L
1UA"T """
Fig.2.5.4-2 Strut-and-Tie Based on Elastic Stresses (Schlaich, 1987)
37
B. The Load Path Method
However, even if no elastic analysis is available and there is no time to prepare
one, it is easy to develop strut-and-tie models using so-called "load paths." The
comparison between load path and strut-and-tie model is shown in Fig.2.5.4-1.
First, it must be ensure that the outer equilibrium of the D-region is satisfied by
determining all the loads and reactions (support forces) acting on it.
The stress diagram is subdivided in such a way, that the loads on one side of the
structure find their counterpart on the other, considering that the load paths
connecting the opposite sides will not cross each other. The load paths begin and
end at the center of the gravity of the corresponding stress diagrams and have the
direction of the applied loads or reactions. They tend to take the shortest possible
streamlined way in between. Curvatures concentrate near stress concentrations
(support reactions or singular loads)
After plotting all load paths with smooth curves and replacing them by
polygons, further struts and ties must be added for transverse equilibrium acting
between the nodes, including those of the U-turn. While doing so, the ties be
arranged with proper consideration of practicality of the reinforcement layout
(generally parallel the concrete surface) and of crack distribution requirements.
A very powerful means of developing new strut-and-tie model for complicated
cases is the combination of an elastic finite element method analysis with the load
path method.
38
apter 2 Literature Review
Load Path Strut-and-Tim Model
*~"-^ -
I pqp^TTB tc^rn
F
3
Fig.2.5.4-1 Strut-and-Tie Model Based on Load Paths (Schlaich, 1987)
2.5.5 Model Optimization
For a structure more than one strut-and-tie models can be developed. Doubts
could arise as to whether the correct model has been chosen out of several possible
ones. In selecting the model, it is helpful to realize that loads try to use the path with
the least forces and deformations. Since reinforced ties are much more deformable
than concrete struts, the model with the least and shortest ties is the best.
The simple criterion for optimizing a model may be formulated as follows:
Z FLs.i i mi = Minimum
39
Where
Ft = force in strut or tie i
/, = length of the member i
emi:= mean strain of member i
This equation is derived from the principle of minimum strain energy for linear
elastic behavior of the struts and ties after cracking. The contribution of the concrete
struts can generally be omitted because the strains of the struts are usually much
smaller than those of the steel ties.
2.5.6 Dimensioning the Struts, Ties and Nodes
Dimensioning not only means sizing and reinforcing the individual struts and
ties for the forces they carry, but also ensuring the load transfer between them by
checking the node regions.
There are basically three types of struts and ties to be dimensioned:
Cc: Concrete struts in compression
Tc: Concrete ties in tension without reinforcement
Ts: Ties in tension with reinforcement
A. Struts and Ties

The struts in the model are resultants of the stress fields. Since by definition the
curvatures or deviations of the forces are concentrated in the nodes, the struts are
straight.
Compression struts serve as the compression chord of the truss mechanism
40
which resists moment and act as the diagonal struts which transfer shear to the
supports. Diagonal struts are generally oriented parallel to the expected axis of
cracking.
Three typical configurations of compression fields:
TnTTfTTf
Fan Bottle Prism
Fig.2.5.6-1 Typical Configurations of Compression Field (Wu Hui 2005)
B. Nodes
Nodes are derived as the intersection points of the three or more straight struts
or ties, which themselves represent either straight or curved stress fields or
reinforcing bars or tendons. Anode implies an abrupt change of direction of forces.
There are essentially four types of nodes depending on the combination of

struts C and ties T as shown in Fig.2.5.6-2, which can be idealized by three kinds of
nodal zones shown in Fig.2.5.6-3.
41
Wufes
a)CCCNotie$
b) CCT Nodes 7v\
e)CTTNo!fcs
dlTTTNodes
Fig.2.5.6-2 Type of Nodes (Schlaich, 1987)
C C
i f
/
I - i €• ->-T
\
\
\
c
c c
W (0) (e)
- -I t ' r t— » - r i f - -- + t
Fig.2.5.6-3 Idealized Forces at Nodal Zone (Schlaich, 1987)
42
2.5.7 Feasible Inclined Angle 0

The maximum length-to-depth ratio of a D-region would be approximately two.
Thus, the smallest angle between the strut and tie in a D-region is arctan2=26.5°.
The feasible inclined angle 9 is different in different codes:
• Swiss Code: 0.5<cot#<2.0 (0 = 26°to 64°)
• European Code: 3/5 < cot#< 5/3(# = 31°to 59°)
• Collin's & Mitchells:
dmin=\0 + \\0(VJWXjd]) deg
0 m , n =9O-0 min deg
• AC12002: #min=25° (0 = 25°to 65°)

mm v /
• If small 0 is assumed in the truss model, the compression strength of the

inclined strut is decreased.
2.5.8 Limiting Compressive Stress in Strut
As mentioned previously, the limiting stresses for truss elements in a

strut-and-tie model is different, as shown in Fig.2.5.9-1, different codes give
different requirements on these stresses as following:
A. AASHTO LRFD 5.6.3.3.3
f = ^ <0.85/
c
" 0.8 + 170*1
Where:
Sj =(e2 + 0.002) cot2 em

fai =the limiting compressive stress
43
6m = the smallest angle between the compressive strut and adjoining tension
ties (deg)
s2 = the tensile strain in the concrete in the direction of the tension tie (in/in)
• Strength of compressive strut AASHTO LRFD 5.6.3.3.3
Pr=FPn (LRFD 5.6.3.2.1)
P„=/aA (LRFD 5.6.3.3.1-1)
Where:
F = 0.7 for compression in strut-and-tie models (LRFD 5.5.4.2.1)
Aa = effective cross-sectional area of strut (LRFD 5.6.3.3.2)
B. Simplified Values for Limiting Compressive Stress in Strut, fcu (Schlaich et al.
1987)
• For an undisturbed and uniaxial state of compressive stress:
/OT=0.85/;
• If tensile strains in the cross direction or tansverse tensile reinforcement
may cause breaking parallel to the strut with normal crack width; this
applies also to node regions where tension steel bars are anchored or
crossing.
/ „ , = 0.8(0.85/ c ) = 0.68/J
• As above for skew cracking or skew reinforcement
/ „ , = 0.6(0.85/) = 0.51/'
• For skew cracks with extraordinary crack width-such as cracks must be
expected if modeling of the struts departs significantly from the theory of
elasticity's flow of internal forces:
/ „ = 0.4(0.85/') = 0.34/'
44
C. ACI 2002 STM Model
Design of struts, ties and nodal zones shall be based on:
<t>F>Fu
The nominal compressive strength of a strut without longitudinal

reinforcement:
F
ns = fcuA
The effective compressive strength of the concrete in a strut is:
/«• = 0.85A/J
• The strength of a longitudinally reinforced strut is:
Fns=fA+Af:
• The nominal strength of a tie shall be taken as:
Fn,=Aslfy + Aps(fse + AAJp)
• The nominal compressive strength of a nodal zone shall be:
nn J at u
45
Element Limiting Stress

1 - CCC Node 0.8S f | j
2 - OCT Node 0.75 4> f'c
3 - CTT or TTT Node 0.65 0 f'c
4 - Strut
f
5 - Tie y or (fpe + fy)
Fig.2.5.9-1 Limiting Stress for Truss Elements (ACI318-02, Appendix A)
2.5.9 Summary on Strut-and-Tie Model
• The strut-and-tie model approach has evolved as one of the most useful design
methods for shear critical structures and for other disturbed regions in concrete
structures.
• The strut-and-tie model formulation that requires the least volume of steel will
be the solution that best models the behavior of a concrete member.
• This approach holds great promise for design offices which could develop or
obtain standard strut-and-tie models for certain commonly encountered
46
situations
• Standard reinforcement details based on a strut-and-tie model could be
developed for common situations.
• The strut-and-tie model then could be reviewed and revised if any parameter
changes.
2.6 Summary
In this chapter, background and significance of present research is addressed.

Past research on performance of shear walls and steel reinforced shear walls has
been done including crack patterns, failure modes, flexural and shear deformations,
horizontal and vertical dilations, pitching effect, stiffness, strength, and energy
dissipation. In these studies, main influential factors including strength of concrete,
ratio of flexural and shear reinforcement, layout of reinforcement, aspect ratio,
gravity and loading history, have been investigated.
As valuable tools in explaining internal forces of concrete members under shear,

strut-and-tie models were applied to analyze behavior of specimens in this research.
47
Chapter 3 Experiment Design
CHAPTER 3
EXPERIMENT DESIGN
In order to investigate seismic performance of precast shear wall with

horizontal connection and study the effect of parameters: horizontal connection,
number of shear connectors and configuration of steel sections, six two-storey
reinforced concrete model walls were tested to failure under reversed cyclic lateral
load. Two monolithic shear walls acting as prototypes and four precast shear walls
with horizontal connection serving as counterparts were tested under cyclic load to
failure.
Capacity of energy dissipation of shear walls was mainly controlled by

plastic deformation of reinforcement before shear failure. And experiment by
Cardenas and Magura (1973) indicated that, under monotonic load, larger ductility
could be attained with concentrated reinforcement at the ends of the walls rather
uniformly distributed reinforcement. Steel sections were used in these specimens
serving as concentrated flexural reinforcement at boundaries of walls, namely,
channels were embedded in specimen Wl, W2 and W3, whereas, I-beams were
utilized in specimen W4, W5 and W6.
Due to concentrated vertical reinforcement placed at two the ends of the

walls and the low aspect ratio of these specimens, a = hw/lw =1.41, these specimens
exhibit possible sliding shear failure under cyclic load. Salonikios et al (1996, 1998,
1999, 2000, 2002) verified that diagonal reinforcement is capable of preventing this
failure mode. In the present study, diagonal reinforcement with 45 deg inclined to
horizontal was used in lower part of the walls with height equal to 0.22hw from the
base.
In precast walls, the horizontal connection consists of shear connectors and

flexural connectors.
48
The effect of confining reinforcement on ductility was more profound than

on the moment capacity, which was supported by experimental evidence by
Pilakoutas and Elnashai (1995). The precast walls are popular in moderate and low
seismic regions, and low ductility is required for these walls. Confinement was not
designed in this research.
To get a simple and clear flow of internal forces, a rectangular cross section
was selected for all walls. No axial load was applied to the specimens, because it is
usually small in structural walls and generally has a favorable effect on their
behavior, suggested by Park and Paulay (1975).
The data acquisition system consisted of a series of internal control and

recording channels and channels for monitoring data from external instruments-
LVDT (linear variable displacement transducers). And a series of strain gauges at
various positions along the reinforcing bars and steel sections at specific positions
were used to record the strain history during cyclic loading.
The foregoing system of measurements made it possible to estimate the

flexural, shear, and sliding components of the wall deformation. Based on strain
gauges and cracking patterns of these walls, strut-and-tie models at different stages
were developed to analyze performance of these specimens.
3.1 Design of Experiment
3.1.1 General Description of Wall Specimens
Total of six walls specimens were tested to failure in the present research.
The dimensions of the walls were 1800mm wide, 2300mm high and 140mm thick
and the aspect ratio isa -hw/lw =1.41. Two Specimens are monolithic walls, Wl
and W4, serving as prototypes, whereas four precast walls, namely, W2 and W3,
W5 and W6, are counterparts of Wl and W4, respectively. Channels were laid at
two ends of wall Wl, W2 and W3, acting as concentrated flexural reinforcement,
49
whereas I-beams were used in wall W4, W5 and W6. The concrete cover of
horizontal reinforcement which was placed outside of vertical bars was 20mm at the
wall end sides and 15mm at wall faces.
The concrete used in all specimens is / u = 4 0 M P a and had 12mm

maximum size aggregate. The vertical reinforcement in the walls was of mild yield
steel bars, as called R bars, with design yield strength 250MPa conforming to
BS4461. And the horizontal reinforcement was of high yield steel bars, called T
bars, with design yield strength 460MPa BS4469. And the steel sections, including
channels and I-Beams were Grade 43. Same reinforcement including size and
layout was used in the six walls except that the vertical reinforcement was cut at the
horizontal connection in precast specimens.
In the horizontal connections, all bolts were Grade 8.8, and diameter of hole
for diameter equal to 12mm bolts is 13mm, and 22mm for diameter equal to 20mm
bolts. The diagonal bars were T bars which welded on steel plate with angle of 25
deg inclined to horizon. Size of steel plates used in specimen W2 and W5 was
400 x 200 x 10mm and 220x 200 x 10mm in specimen W3 and W6.
All specimens were cast vertically in plant. The monolithic ones, Wl and
W4, were cast in one time. But, each of the specimens W2, W3, W5 and 6W was
divided into two segments, and the two segments were cast separately. Upon seven
days of curing, the two segments of each specimen were joined together with bolts,
nuts and concrete with same strength of wall panel.
The reversed cyclic lateral load was applied by a testing rig. On base beam,
there are fourteen PVC tubes with diameter 52mm evenly distributed along the
horizon in two lines. The base beam is firmly fixed on the strong floor of laboratory
with steel rods going through these PVC tubes with nuts. A horizontal beam of the
testing rig was connected to the top beam of the specimens. The dimensions of the
top beam are shown in Fig. 3.1.3-8 and it is a model of the six specimens. There are
four 32mm PVC tubes passing through the top beam, which were used as a duct for
50
loading bars. The loading bars are connected to horizontal loading beam of testing
rig by nuts, which work in pulling stages in the loading cycles. The loading
conditions and boundary conditions are same for all specimens.
3.1.2 Material Properties
A. Concrete
All specimens were cast vertically indoor with a standard mix. Concrete was
supplied by external commercial ready-mix supplier in Singapore. The
characteristic strength /c.Mis 40N/mm2 with 100 ± 25mm slump in accordance with
BS5328. The maximum size of aggregate is equal to 12mm in order to ensure good
compaction of concrete in the specimen walls. A total of eighteen
150x150x150mm cubes and eighteen 150x 150x300mm cylinders were cast of
each batch. Three cubes and three cylinders were cured for each specimen under
identical conditions. Cubes and cylinders were tested at the same day of testing
specimen itself and the average of results was taken as the compressive strength as
shown in Table 3.1.2-1.
Table 3.1.2-1 Average Concrete Compressive Strength at The Day of Testing
Average J^, Average / ,

Specimens Days
(MPa) (MPa)
Wl 60 43.51 35.58
W2 70 44.12 33.53
W3 81 44.78 34.04
W4 87 45.35 35.45
W5 94 43.98 34.02
W6 104 44.65 33.24
B. Reinforcement and steel sections
Five samples of each kind of reinforcing bar were tested in tension. In this
research, reinforcing bar, T10, was used as horizontal shear reinforcement. The
51
stress versus strain relationship is shown in Fig.3.1.2-1. Stress vs strain curve of
vertical reinforcing bar, RIO, is shown in Fig.3.1.2-2. Coupons of web of UB (I-
Beam) and Channel shown in Fig.3.1.2-3 were test and stress against strain curve
was plotted in Fig.3.1.2-4 and Fig.3.1.2-5. The properties of the reinforcement and
steel sections including area of cross section, elasticity modulus, yield strength and
ultimate strength are listed in Table.3.1.2-2 .
700
600
500
Tensile Test TIP Bar:

| 400
ee = 2700
to
10 fe = 486MPa
£ 300
<0 fu =603.68MPa
200
100
10000 20000 30000 40000 50000 60000 70000 80000 90000

Strain (ys)
Fig. 3.1.2-1 Stress Versus Strain Relationship of T10 Bar
52
600
10000 20000 30000 40000 50000 60000 70000 80000

Strain (fjs)
Fig. 3.1.2-2 Stress Versus Strain Relationship of RIO Bar
SO
1 80
s>
^ 0 - - ^
so
55
1
>*5
1
-^~^
230
FFig. 3.1.2-3 Coupon design of I-Beam and Channel
53
500
450
Coupon Test Channel:

£e=1800
fe = 307MPa
fu=463.3MPa
5000 10000 15000 20000 25000 30000 35000 40000 45000

Strain (/JE)
Fig. 3.1.2-4 Stress Versus Strain Relationship of Channel
600
500 •
^ — — •
400
I5
w 300
f Coupon Test I-Beam:
in / se = 2000
200
fe
= 322MPa
/ "
100 1 f„ = 479.8MPo
—1 1 [ 1 1
5000 10000 15000 20000 25000 30000 35000 40000 45000

Strain (\iz)
Fig. 3.1.2-5 Stress Versus Strain Relationship of 1-Beam
54
Table 3.1.2-2 Properties of Reinforcement and Steel Sections

Types RIO R13 T10 T20 Channel I beam
Area of cross section
78.5 132.67 78.5 314 856 2050
As (mm2)
Modulus of Elasticity E
192721.33 193237.58 185553.79 202120.82 192460.35 177279.69
(MPa)
Yield strength f (MPa) 384.63 402.3 485.91 502 307.74 322.16
Yield strain( /US ) 2200 2200 2700 2900 1800 2000
Ultimate strength fu (MPa) 544.5 502.6 603.68 626.4 470.97 487.78
3.1.3 Design of Specimens
A. Design of specimen Wl
The reinforcement layout of specimen Wl is shown in Fig.3.1.2-2. Due to

the limit of capacity of testing rig, reinforcement R10 are selected as vertical
reinforcement for its lower yield strength and ultimate strength comparing with T10
bar. All the horizontal reinforcement is T10 bars which are welded on the channels
at ends of the wall. Every channels embedded in base beam is weld with two bars
with diameter equal to 32mm to ensure anchorage of the channels. For specimens
with embedded I-Beams, same method is applied to reach requirement of anchorage
in base beam.
The initial specimen design is based material properties provided by BS

4449, in which the yield strength of hot rolled mild steel, R10, is 259MPa, and that
of high yield steel, T10, is 460MPa. The yield strength of channel is 275MPa.
Compression strength of concrete cylinder is 30MPa. According to AC1 318-02,
strain in reinforcement and concrete is directly proportional to the distance from the
neutral axis. And the assumed maximum usable strain at extreme concrete
compression fiber is equal to 0.003.
Try the depth of neutral axis, c=170mm. Therefore, two channels and 16
R10 bars are in tension, 1 channel and 2 R10 bars are in compression.
55
a = 0.85c = 144.5mw
C c = 0.85x144.5x140x30 = 516&V
C , = 2 x 78.5x250 + 856 x 275 = 275kN
C= Cs+Cc=19\kN
Ts = 16x78.5x250 + 2x856x275 = 785£N
For 785&V ~ 19\kN, so T=C.
M = Ts_chmls x(di-<x/2) + Ts_hars x (d2 - <*/2)
= 470.8x(1800-101- ] 4 4 -5^) + 3 1 4 x ( 1 8 0 0 - 7 0 0 - 1 4 4 - 5 ^ )
= \0S9kNm
Where dl, d2 are the distances from centroid of channels and that of bars in
tension to the extreme compression fiber, respectively.

Therefore, the horizontal force loading at the top beam will be:
F = M/h = 1089/2.54 = 429kN
Where h is height of wall, measuring from the wall base to horizontal central
line of the top beam.
As shown in Fig.3.1.3-1. T10 bars are used as horizontal reinforcement with

spacing 200mm. According to AC1 318-02,
AJd AJx0.Mw 78.5x2x460x0.8x1800 _„....
Vs = — = = = 520kN
s2 s2 200
^ = 1-21
Yo
429
Due to special layout of vertical reinforcement and low aspect ratio, the
walls may fail in sliding shear near the base under reversed cyclic load. In order to
prevent this kind of failure, four diagonal bars in each direction was placed in
0.22h w (550mm) height from the base, as shown in Fig.3.1.3-1. All diagonal bars
both under tension and under compression contribute to resist the shear force at the
wall base.
Vdla = Adiafy sin45° =4x2x78.5x460x0.707/1000=204kN
B. Design of specimen W2 and W3
56
The specimen W2 and W3 are precast counterparts of monolithic specimen

Wl. The horizontal reinforcement in W2 and W3 are identical with that in Wl. And
the amount of channels and vertical bars used in W2 and W3 is same to that in Wl
except they are cut into two.
In each horizontal section of wall under loading point, flexural and shear
force exist simultaneously. In present study, only the channels in upper and lower
wall panels are connected by a group of bolts with steel plates, which form
mechanism to resist the flexural moment at the horizontal connection level. Also,
shear connectors embedded in upper and lower wall panels, which are composed of
diagonal bars and steel plates and positioned at center of the horizontal connection,
are connected by bolts to withstand the shear force and inhibit sliding along the
horizontal connection.
The dimensions and horizontal and vertical reinforcement of W2 are

identical with those of W3. The difference between specimen W2 and W3 is
different of number of shear connectors. In specimens W2 there is only one shear
connector but in specimen W3.
In design of connection, in order to get a clear load transfer in the horizontal

connection, the flexural moment and shear force are designed separately and
resisted independently. The channels at two ends of the walls were used to take
flexural moment and shear connectors at center of the connection to resistance the
horizontal shear force.
a. Design flexural connectors of specimen W2 and W3
Assuming channels only resist tension and compression force at ends of the
walls which form a moment-resistant mechanism. Grade 8.8 bolts with diameter
equal 12mm are applied to connect upper and lower channels as shown in Fig.3.1.2-
57
2, Fig.3.1.2-3 and Fig.3.1.2-9. And diameter of bolts hole is 13mm and web
thickness of the channel is 5.1mm.
So, the effective area of channels= area of channels-area of hole
= 856x2 -13x4x5.1 = 1446.8/ww2
The yield strength of channel of Grade 43 is 275MPa.
Therefore, the strength of flexural connector is,
1446.8 x 275 = 397.87/tTV « 39UN
The shear capacity of each bolt on channel of material Grade 43 is 12MPa.
398
n= = 15.3 Therefore, 16 bolts are needed to resist the total shear force.
12
And steel plates of Grade 50 are applied to connect the channels in upper
and lower channels with bolts. The yield strength of steel plates of Grade 50 is
355MPa. The dimensions of the steel plates are 10mm thick, 100mm wide and
360mm long as shown in Fig. 3.1.3-9.
355xl0x(100-13x2)x2 = 525&V
Assuming that flexural moment at the level of horizontal connection in W2
and W3 which is equal to ultimate lateral load in Wl can be resist by the connected
channels. Distance from wall edge to center line of two channels at each end of the
walls is 101mm including 20mm concrete cover and half of spacing the channels,
81mm. Height of wall panel is 2300mm and top beam is 480mm height. Location of
the horizontal connection is at center line of wall panel as shown in Fig.3.1.3-2.
The axial compression or tension load in the channel at one end of the wall
is:
w=429x(2540-1150)=374w
(1800-101x2)
Since374kN < 39MN < 525kN, so the flexural connectors can withstand the
compression or tension forces in channels.
58
b. Desifin shear connectors of specimen W2 and W3
Shear connectors consist of steel plates and diagonal reinforcing bars as

shown in Fig.3.1.3-9 and Fig.3.1.3-10. The diagonal reinforcing bars with hook at
one end is 25 deg inclined to horizon and the other end is welded on the steel plates.
It is assumed that the diagonal reinforcing bars in compression and those in

tension resist the horizontal shear force simultaneously.
16x78.5x460cos25° = 524&V >430kN
Assuming leg length of weld is 5mm, then weld length is
:
= 20.2mm
2x0.7x255x5
weld length 20.2+5+5 « 30mm
According to BS8110 Part 1:1997 for concrete with cube strength 40,
anchorage length of diagonal bars is 44 db. In this research, the length of straight
part is 500mm with a end hook.
fb=Fj7apJ = 460x78.5/(3.14x10x500) = 23MPa
f* = PjL = 0-4V40 = 2.5MPa > fb = 2.3.MPa
Where fb is the bond stress;
Fs is the force in the bar or group of bars;
/ is the anchorage length;

<pe is the effective bar size which, for a single bar is equal to the bar size and,
for a group of bars in contact is equal to the diameter of a bar of equal total
area;
fbu is the design ultimate anchorage bond stress;
/? is a coefficient dependent on the bar type.
Three Grade 8.8 bolts with diameter equal to 20mm are used to connect the
steel plated embedded in upper and lower wall panels.
59
Assuming the vertical bars in the shear connectors yield, so the vertical
force at center of the wall is:
4x78.5x460 = 1 4 W
For specimen W2,
"144 Y f430Y
: 75.6JUV
6 J
For specimen W3,
2
\f\AA\
144 ^430?
+\ 8 J 51kN
According to BS5950: Part 1:1985, the capacity of the bolt with diameter
equal to 20mm on steel plate of Grade 50 and thickness equal to 10mm is 1 lOkN.
Since 75.6kN<110kN, and 57kN<110kN so the bolts in shear connection
work.
c. Design of specimen W4
Universal beam is embedded in W4 as main vertical flexural reinforcement.
The select section is UB152x89xl6 as shown in Fig.3.1.3-4. The only difference
between specimen Wl and W4 is different type of steel section. All other variables
keep constant, including vertical and horizontal reinforcing bars, diagonal bars and
strength of concrete. In design of this specimen, it is assumed that the I-beam is
divided into two parts as channels in Wl. The area of cross-section of I-Beam is
2050mm with yield strength 275MPa.
Similar to design of Wl, the results is that the depth of neutral axis is
185mm and lateral load acting on top beam is 486kN.
d. Design of specimen W5 and W6
Specimen W5 and W6 are precast counterparts of monolithic wall W4 as

shown in Fig 3.1.3-5 and Fig.3.1.3-6. W5 and W6 have the identical horizontal
reinforcement with W4. W5 and W6 also have same amount the vertical reinforcing
60
bars and steel section to W4, whereas, the vertical reinforcing bars and steel section
are cut into two. The upper and lower wall panels of W5 and W6 are assembled by
connecting the horizontal connection and fill gap with concrete which has same
strength as wall panel.
Similar to force transfer mechanism in the horizontal connection of W2 and

W3, the I-Beams were connected to provide flexural moment resistance in the
horizontal connection. The axial compression and tension in the I-Beam is
486x(g40-H5<»)
1800-101
The steel plates used to connect the I-Beams embedded in upper and lower
wall panels are shown in Fig 3.1.3-11 is 110 wide and 10mm thick.
(110-2x13) xl0x355x2=596kN>397kN.
The detail design of flexural connectors is shown in Fig 3.1.3-11. The
strength of the cross section of I-Beam is
275x(2050-13x7.7x4) = 454AiV
The number of bolts:

454
n= = 38, So 40 bolts are used in flexural connection to resist the axial
12
force.
Since 454kN>397kN, therefore the flexural connectors can work.
The shear connectors used in W5 and W6 is same to those in W2 and W3,
respectively. For diagonal reinforcement of the shear connector, 524kN>486kN, so
the diagonal is strong enough to resist the horizontal force in the horizontal
connection.
For specimen W2,
For specimen W3,
61
Since 84.5kN<110kN, and 83kN<110kN so the bolts in shear connection
work.
100 3 3 218 .200 .200 .200 .200 .200.218 . / — 100

HI 1 1 if l l TTT
Elevation View
20 20
Channel 76x38x6 Channel 76x38x6
TT It v-,
4 h
>r->
1800
Horizontal Section A-A
Fig. 3.1.3-1 Specimen Wl and Its Overall Dimensions
62
100 z% 218 200 .200 200 200 .200 218 r^~ 100
TTi j I i I i i nn
20 Channel 76x38x6 Elevation View Channel 76x38x6 2L~

IT
i h
1110 {
, 140
1800
Fig. 3.1.3-2 Specimen W2 and Its Overall Dimensions
63
100 7*T\ .218 .200.200 .200 .200.200 218 . r^.lOO

~T1M 1 1 11 1 1 TlTT
20
Channel 76x38x6 Elevation View Channel 76x38x6
~n "->
« -
^
1800
64
100 j j j i 218 200 200 200 200 200.218 r&L Jflfl

Hi i i II i i nn
Elevation View
20 Universal Beam 152x89x16 Universal Beam 152x89x16 20
tr It
H H
.17101.
, 2*0
7<S00
65
100 T~S 218 200 .200 200 200.200 218 r^.li

i t i i t t t irr
Elevation View
20 Universal Beam 152x89x16 20
Universal Beam 152x89x16
ft 11
B H
. 140
1800 >--
66
100 Z I P 218 200 .200.200 200 200.218 M1 100
nn i i ii i i t n
Elevation View
20 20
rr Universal Beam 152x89x16 Universal Beam 152x89x16
J\
H H
[110 [
, 140
in
1800
67
400 400
7^ r^ 3o
r n 3
so o©
^J 2 a
130 ,L_ 130 130 i 730
Fill with concrete after

connecting steel members
I
380 3«0 380 5S0

740 \l40
900 900
Side View ofWl and W4 Side View of W2, W3, W5 and W6
Fig. 3.1.3-7 Side View of All Specimens and Their Dimensions
68
500
i i 9 m
J
^ <J
^ . _ <& <r,
«"1 5J
*••<"» C
'~ ' " • • « i ,©
C>
_ w ••••••' • Ov
I m
•a &a
ui
• r * •**
m—i • *J
1.
CO
480
•a
B
\
:
s
\ s
01
oa
a.
— •
e
I-5a " v
H
1
480
,75 75
f
a,
£
1
I
Si; a
<o ^
f n
I ' Y-
I
\4T20
69
Detail 1
* 1
J
</i
: 00 * <=>
s >~
00 II
oo
oc
00
'•• 53
. "3 •8
a a
:
00 ^~-<
6
>o
00
*r\
- S <»'
! 5 >*
. 00 *-.
(3;
06
II =e
i
00 a
•8 |
00 s^ 9
f 6
•«(
'
-
J
/
762
Grade 50 steel plate
thickness=10mm
§ "4
Detail2
Fig. 3.1.3-9 Details of Connection (Cont.)
70
Grade 8.8 bolt

Detail 3
Fig. 3.1.3-10 Details of Connection (Cont.)
Detail 4
Fig. 3.1.3-11 Details of Connection
71
3.2 Testing Program
3.2.1 Test Set-up

The wall specimens were tested in steel test rig was specially designed and
built for the purpose. A schematic arrangement of the test set-up is shown in
Fig.3.2.1-1. The general view of the test rig is given inFig.3.2.1-2.
The specimen was fixed on the lab strong floor by fourteen high strength
tension bolts with diameter equal to 50mm to prevent it from lifting. The reaction
footings were placed at two ends of the base beam of the specimen to avoid
horizontal sliding during loading course. Therefore, a fixed ends condition for the
base beam will be simulated. There was also a special horizontal beam with a
reversed "U" which was used to restrain the out-of-plane movement of top beam.
The test frame is design to apply 805kN horizontal load on the test walls with a
222mm maximum horizontal travel. The jack was operated by electric hydraulic
pumps. The load was applied at ends of the top beams which were stiff enough to
distribute the load to the whole walls. The specimens were push to left at right end
of the top beams and pull to right through loading bars at the left end of the top
beams.
3.2.2 Instrumentation
The tests were monitored by load vs. displacement curve drawn in x-

y coordinate, which were taken from load cell readings and horizontal displacement
recorded by Linear Variable Differential Transducers (LVDT) at top beam. All
these data were also recorded by data logger. LVDTs measuring horizontal
displacement of top beams were fixed on a steel frame which was seated on the
base beam of the specimens and the rest of LVDTs were mounted on the specimens
which made LVDTs and specimens to be a unit referring to lab ground, that is, the
relative movement of specimens to lab ground did not affect the horizontal
displacement recorded by the LVDTs measuring horizontal displacement.
72
,5 e
IS
si
E
73
Fig. 3.2.1-2 General View of Set Up of Specimen Wl Fixed To The Lab Floor
Electrical resistance strain gauges were mounted on steel bars and channels
or I-Beams to record the strains experience by them during the course of testing.
The electrical resistance strain gauges on the steel were 5mm gauge length,
119.6 ± 0.4 Q resistance, gage factor 2.10, adoptable thermal expansion 11.7
PPM/°C. These strain gauges were glued to steel bars and sections with an ultra-fast
adhesive and cover with silicon acting as water proof, finally a coat with strong glue
was used for insulation and moisture sealing when silicon stiffened. Some of the
strain gauges wires inadvertently broken and some of the gauges were damaged
during the construction process due to the casting operation or vibrating procedures.
Fig. 3.2.2-1 to Fig. 3.2.2-3 show the location of the strain gauges on steel
bars of specimens. Fig. 3.2.2-4 to Fig. 3.2.2-6 show the location of the strain gauges
on steel sections of specimens. In these figures, a "-" is used to indicate position of
the strain gauges. The difference of specimens Wl and W4 was in the location of
74
strain gauges on steel sections. The difference of specimens W2 and W3 is position

of strain gauges on steel bars similar to specimens W5 and W6. The difference
between monolithic walls Wl and W4 and their precast counterparts was the
different position of strain gauges on steel bars and sections. The individual strain
gauge was also tagged for identification. Strain gauges were separated into S-side
and N-side. The strain gauges mounted on vertical bar were also divided into High,
Middle and Low, for example, SVL1 means the strain is on S-side and low position
of vertical bar 1. Strain gauges on horizontal reinforcement were noted by 1,2,3 on
S-side, and 1,2,3,4,5 on N-side, for example, SHI2-1 means that the strain gauge on
S-side, Horizontal bar 12, number 1 counting from left. Strain gauges on diagonal
bars were numerated from left to right. Strain gauges on steel section were
numerated from base to top of the wall.
The positions of LVDTs are indicated in Fig. 3.2.2-7 and Fig.3.2.2-8.
The total number of LVDTs and strain gauges used in each specimen are
listed in Table 3.2.2-1.
75
i—\—1
»
1 1 _ 1 i
"-<
as
1
' ' 1
< > N.
Cv 9) ^c ft
h>
\1 v^ ' / /
t^
3
ve>
>r\
' ,/ 65 T3
/ B
/ / **
. 1 1 1 ft
f\
(/ - ^
'M
o
' " 1 1 i
S
>
:: 1 O
< N " i - ) ' ^ ' « n C t \ 0 0 O , ' — |
Ml
3
\ a
U
«
~^ "Rl
" 1 1 i
ON M.
o V J^
Ov
3
65
I
v\ s \ SO ft oo • •
\ Vv ' / /
K
i3
tv
vo
• •
H
1 ' ' '
« § • •
.ft si
t/ in
5 *-> • •
/ >
/ / \ \ »
• •
ft
Cy ~ \ \ 'n
.a
<») • •
O • ^
'M <N r i
1 1 '
"-1 y
:: 1
76
a
IT;
•o
c
C5
.5 a
N
o
a
g 60
Q
»«( |R| •s
=1
o\ M.
I
00 • •
I K
VO
•
•
•
1 Wl
>
•
•
•
1 <») • •
<N W
•»< y
77
5
c
$
•a
e a
s
I
§
CO
a
o
at
=
0
3
[Hf]
e
c* M
•
°o
l\ • '
H
I \o • '
M
1 «->
"»
•
•
•
l») • '
<N H
y
<a i-M fs
< N " i ' < j - ' o > c ( v = o O v
78
K
Ej
I
l
l c
o
s u
=2
1 480 | 4*0 I jgg [240 [240 [240[ .100
+
!
•I
I I
I
I
i
<
1 430 [ 480 [ 480 1240 [ 240 [ 240 [ 100
79
c
£
•a
8
1? <
3
•5
'5 w
Str
£
^ ri
R
.£
£
«M
••§ O
B 09
a
« ©
| 360 I 360 [ 360 I 360 I 360 [ 360 [
CO
as
c
o
a
3
es
0
C/3
o
1
^R J
a
Pi
in
& «s
e P4
<s o
1 J6fl [ 360 [ 360 [ 360 I 360 [ 360 [
80
s
.g
5
<0
a
IIf
ft
a
to
s I
1s
I
ft
£
J. 480 [ 480 I, 480 [240 [240 [240 [ 100 d
o
s
M
£
3
i
e
J
VD
i
1
ri
8 ex
ft
3
to
i
K
ft
J, 430 J 720 960 -UJOl
81
LI
L2\ JZ.6
li\
LI X\\ / /
L12
\L7
\\ / //
x
/\
/ // \\ s
L4\ Us
L10
15 i
U9
Fig. 3.2.2-7 Layout of LVDTs on Wl and W4
/.;
Ait
Lit 17
0 \ / 0 -L24
L4\ £#
/ s
IS\ L21 L22
L25,
vi i
Fig. 3.2.2-8 Layout of LVDTs on W2, W3, W5 and W6
Table 3.2.2-1 Number of LVDTs and Strain Gauges

Strain Gauges
Specimen LVDTs
Steel Bars Steel Sections
Wl 12 80 28
W2 22 103 24
W3 22 103 24
W4 12 80 22
W5 22 103 24
W6 22 103 24
82
In order to separate the shear and flexural components which contributed to

horizontal displacement of top beam of specimens, as shown in Fig.3.2.2-7 and
Fig.3.2.2-8, LVDTs are aligned with diagonals of specimens to quantify the shear
deformation, and LVDTs are aligned with verticals on both sides of specimens to
evaluate the flexural components.
Fig. 3.2.2-9 Calculation of Shear Deformation from Rectangle
From Fig.3.2.2-9, assuming the original configuration is a rectangle, so

r = K + r2 (3-1)
A, / 2 sin a - A, / 2 sin a
tan 7, = (3-2)
L + A, / 2 cos a + A2/2 cos a
A, /2 cos a - A2 /2 cos a
tan^2 (3-3)
H + A,/2 sin or + A 2 /2sina
H
Where a is defined as tan a = —, y is the shear deformation, A, and A2 are the
Is
deformations of diagonals in different directions and positive in extension,

respectively. L and H are the side length and height of the rectangle.
Since A, sin a - A2 sin a « L, A, cos a - A2 cos a « L , so
Due to the small shear deformation is considered, therefore tan y « y .
83
(A,-A2)(y^+Z2) (3
2(HL)
Therefore, the shear components of total horizontal displacement of top beam is
"s=rHw (3-5)
Where us is component of total horizontal displacement of top beam due to shear
deformation and Hw is the height of wall measured from base of the wall to the
horizontal axis of the top beam.
The scheme of layout of the LVDTs aligned with verticals on both sides of
the specimens is shown in Fig.3.2.2-7 and Fig.3.2.2-8, which is used to catch to
flexural deformation of the walls. The deformation of vertical segments is
represented by a serial of Av,. ., Av, 7 and Av r ; indicate vertical deformations of
these segments on left and right ends of the specimens, respectively. When the top
beam displacement of the wall, Aw , is negative when wall is pushed to the left. In
order to attain the curvature of different segments of the wall, the rotation of these
segments should be available. For example, the rotation of line 2 relative to line 1 is
calculated by
Av,, - Av,,
r
02=^- ± (3-6)
w
where the lw is the distance between both rows of the LVDTs.

The curvature in the interval between linel and line2 on the center can be
calculated as
*=-£- (3-7)
A2
Where # 2 is the relative rotation of linel to line 2, f#2is the curvature segment of
wall between linel and line2. The average gauge length A 2 can be can attained by
(v / , 2 +Av /2; ) + (v,.
: 2 +Av r 2 )
A2 =—• — (3-8)
The formula of rotation and curvature of the different segments of wall can be
achieved by changing the index 2 by 1, 3, 4 and so forth.
84
The rotation angle determined by equation 1 is a relative rotation angle

between two adjacent lines. Therefore, in order to obtain the flexural horizontal
displacement of the top beam, a fixed horizontal line needs to be assumed to be a
reference. Because all these LVDTs were mounted vertically along the ends of the
specimens and no addition fixing frame was used to fix them. Yanez (1993)
discussed the determination of the fixed line. Yanez suggested that the flexural
deformations Av, and Avr]be not correctly evaluated if these deformations was
assumed only occurred above the base line of the wall. So the curvature of the first
interval will be fault. Through checking the gauges adherent below the base line,
horizontal line 50mm below the base line was taken as the fixed line. Therefore, in
present research, the line 50mm below the base line is used as the reference for the
first interval. This accounted for strain penetration into the base beam. The flexural
deformation components of horizontal displacement of top beam can be
approximately evaluated with Q , Q2, <93 and QA.
A = A, + A2 + A3 + A4 = 0XHX + 62H2 + 9lHl + 0AHA (3-9)

Where H.,H2,H3and H4are the distance from center of these intervals to the
horizontal axis of the top beam.
85
V/, 4 +Av / ] 4 V
,,4+Avr,4
v
/,3+AvU V
r.3+Avr,3
V
v
/,2+Av/.2 r,2+Avr,2
v V Av
U+Av/,i r,\ + r,l
/,.
Fig. 3.2.2-10 Flexural Deformation from Verticals at Both Ends of Specimens
Central line of
wallplane
A A A A
4 3. 2. l
* a" * «*
Fig. 3.2.2-11 Calculation of Flexural Deformations
86
3.2.3 Load history

The wall specimens were load with slow cycles of lateral load in two stages.
The first stage involved three reversed load-controlled cycles in the elastic range at
0.25, 0.5 and 0.75 times the nominal maximum load (ideal strength), respectively.
The ideal or nominal strength of a section of a member, St, is derived from the
dimensions, reinforcement content, and details of the section designed, and code-
specified nominal material strength properties, suggested by Paulay (1992). The
second stages involved reversed displacement-controlled cycles of loading at
increasing level of displacement ductility factor (DF) until the load carrying
capacity of the wall was exhausted. As the effect of loading history was not selected
as a test variable, the typical displacement-control load was applied two cycles at
each displacement level. The displacement ductility factor (DF) was defined as the
ratio of the lateral displacement at any step to the displacement at yield. The yield
displacement A was estimated as 1.33 times the average of the positive and
negative displacements corresponding to loads of 0.75 of the ideal strength of the

wall in both directions, respectively, as shown in Fig. 3.2.3-1. The tests were ended
when the strength decreased to less than 80% of the recorded maximum strength.
These criteria for the displacement controlled cycles are based on a reduced
stiffness equivalent elasto-plastic yield and on a significant load capacity after peak
load suggested by Park (1989).
The load history applied for all the specimens is given in Fig. 3.2.3-2 to Fig. 3.2.3-7.
In these figures, scan numbers are used as x-axis and displacement is used for y-
axis. In course of testing, different load stages were shown in photographs including
load-controlled and displacement-controlled stages.
87
lateral load
ideal strength
o. 75 of ideal strength of
first yielding which is less
displacement
A,., + A „ ,
o. 75 of ideal strength of
first yielding which is less
ideal strength
Fig. 3.2.3-1 Definition of yield displacement (Park, 1989)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Cycle
Fig. 3.2.3-2 Load history of specimen W l
88
3.5
3
i
!
.•£> o
a -1
-2
-3
-3.5
Fig. 3.2.3-3 Load history of specimen W2
s
C5
Cycle
89
12 3 4 5 6 7 8 9 10 11 12 13 1415 161718 19 20
Cycle
1 2 3 4 5 6 7 8 9 10 11 12 13
Cycle
90
- Spcimen W6
.A. A A"/\
*s\ {, 1
^ <
V
V V
Sc, in Number 252
V 242
1 2 3 4 5
Cycle
3.3 Summary
In this chapter, total six specimens, two monolithic and four precast walls,
are designed. In order to grasp critical features of specimens during loading, a
number of LVDTs and strain gauges are used in this research and they are laid out
in appropriate locations. Referring to previous studies on shear walls by other
researchers, loading course is given.
91
Chapter 4 Experimental Results
CHAPTER 4
EXPERIMENTAL RESULTS
All the specimens were tested to failure under reversed cyclic load as
previously described. The focus of herein is to summarize the qualitative and
quantitative experimental results. Photographs is taken to record the qualitative
results through the course of testing, such as, cracking patterns, crushing locations,
bar breaking, fracture of steel sections and failure modes. Readings of LVDTs,
strain gauges and crack widths were also used in this research to catch the
quantitative behavior of walls.
Load-displacement response and cracking patterns of the walls were the

mainly features of the results. Other characteristics including stiffness, curvature,
flexural shear and sliding deformation, vertical elongation and failure modes of the
specimens are presented in this chapter.
4.1 Loading Procedure
4.1.1 Load and Displacement Histories
Horizontal displacement versus scan number diagrams as shown in

Fig.4.1.1-1 to Fig.4.1.1-6 show the histories of the horizontal displacement of the
top beam of specimens in the courses of testing. Similarly, Fig.4.1.1-7 to Fig.4.1.1-
12 show the histories of horizontal load affecting on the horizontal axis of the top
beam. As loading diagram schemed before, the loading course was divided into
force-control and displacement-control. In elastic stage, load was applied by force-
control in sequence, 0.25P, 0.5/>, 0.75/) and/), and one cycle was for each load
level in this stage. In the plastic stage, two displacement-control cycles were applied
for each displacement level. Therefore, the horizontal force in the second stage
taken from load cell was results from horizontal displacement at different ductility
levels. Figures of histories of lateral load showed that the maximum strength kept
almost constant when specimen entered the plastic stage and that the maximum
92
strength in the second cycle at each displacement level was a little lower than that
of first cycle. When displacement reached a level, that is, after a plastic plateau the
strength of the specimen started to decrease.
4.2 Experimental Results
All these specimens were designed, ntested to failure following the above-
mentioned loading histories in the heavy structures laboratory at Nanyang
Technological University, Singapore. Reading of strain gauges placed on vertical,
horizontal and diagonal reinforcement and steel sections were collected. And
records of LVDT in the loading course were also applied in this part to analyze
behavior of these specimens.
100 200 300 400 500 600 700 800 900

Scan Number
Fig. 4.1.1-1 Horizontal Displacement History of Specimen Wl
93
200 250 300 500

Scan Number
Fig. 4.1.1-2 Horizontal Displacement History of Specimen W2
200 250 300 500

Scan Number
94
100 200 300 400 500 600 700 800

Scan Number
&J -40 --
50 100 150 200 250 300 350 400 450 500 550 600 650 700
Scan Number
95
SO 100 ISO 200 250 300
Scan Number
100 200 300 400 500 600 700 800 900

Scan Number
Fig. 4.1.1-7 Lateral Load History of Specimen W l
96
50 100 150 200 250 300 350 400 450 500

Scan Nubmer
Fig. 4.1.1-8 Lateral Load History of Specimen W2
s
50 100 150 200 250 300 350 400 450 500
Scan Number
97
I
I
!
100 200 300 400 500 600 700 800

Scan Number
100 200 300 400 500 600 700

Scan Number
98
50 100 150 200 250 300
Scan Number
4.2.1 Load-Displacement Response
Load-displacement responses of all specimens are shown in Fig.4.2.1-1 to

Fig.4.2.1-6. The typical characteristics of the reinforced concrete members
subjected to reversed cyclic load, such as unloading and reloading stiffness
reduction as the cyclic displacement level increases, and pinching of hysteresis
loops, can be clearly seen from these figures. Low height-to-width shear walls
generally produce hysteresis curves that are highly pinched and exhibit significantly
less energy dissipation than would similar walls with a larger height-to-width ratio.
More energy is dissipated through the structure in the post-peak range as the
concrete begin to soften and the reinforcement approaches yielding.
The unloading curves of the second excursion at each displacement level

generally followed the same unloading path as the first. The reloading branches of
the second excursion of displacement followed a similar loading path as the first,
but a lower stiffness, resulting in lower peak strengths. The load-displacement
99
curves indicated that the first excursion of a new displacement level followed the
loading path of the second excursion of the previous displacement level. This
suggests that additional cycles at a specific displacement level would produce
negligible damage in comparison to that experienced by the first unloading-
reloading cycle. The responses demonstrated this trend until the peak load. During
the post-peak ranges, the amount of damage of experienced during subsequent
cycles was significant and similar to that sustained during the second excursion of
the previous displacement. The increased accumulation of damage was a function of
the concrete's inability to contribute significantly the lateral resistance.
The hysteresis loops show that all precast specimens except W6 exhibited
good performance in comparison with their monolithic prototypes including
maximum strength and ultimate displacement which are also indicated by Fig.4.2.1-
1 to Fig.4.2.1-6. However, the ductility of the precast walls is smaller than that of
their monolithic prototypes. Some specimens only resist one cycle at the maximum
displacement level whereas others resist two cycles at the maximum displacement
level when the strength dropped below 75% of the maximum experimental strength.
The horizontal connections in precast walls almost made no difference in

yielding strength and ultimate strength in comparison with their monolithic
prototypes, which can be also seen from Fig.4.2.1-1 and Fig.4.2.1-2. This verified
the idea of connection design, using the embedded steel sections to carry the
bending moment.
The yield displacement of the precast specimens occurred at larger

displacement compared to that in monolithic prototypes which was one source of
the lower ductility of the precast walls. One reason for postponed yield point is due
to the tolerance between the bolt hole and bolt in the flexural and shear connection
which allowed a very small rocking of the upper precast wall panel and a similar
sliding along the horizontal connection. The other was due to the discontinuity of
the vertical reinforcement at the horizontal interface between the precast wall
segments which decreased the stiffness of horizontal connection. From
100
experimental result, the precast walls exhibited lower ultimate displacement. As a

result, the ductility of precast was lower than that of those monolithic walls.
Pinching effect in precast walls was more serious than that in monolithic
ones, which indicated lower capacity of energy dissipation of the precast walls. The
extent of pinching is a hint of crack closing and consequent loss of stiffness during
reloading in the opposite direction. One cause is the closing of the opened crack
under previous loading. For precast wall, due to existence of horizontal connection,
the sliding of the upper wall panel along the horizontal connection during loading is
also a reason and the closing of the horizontal connection is another root.
1% Drift ratio 2%
2% Drift ratio _1% -800 -
Fig. 4.2.1-1 Hysteresis Loops of Specimen W l
101
800 -I Drift ratio 2%

i
Lateral Load (kN)

700 -
600 - i 2 3 3.5 4 =Z)F
1 I«=P/-I
500 - Pi=429kN
400 -
1 300 -
Specimen W2
20jL^
-60 -50 /-40 -20
1 01V,
//-10, 20 30 40 50
Displacement (mm)
60
Pi=-429kN
DF= -4 -3.5 -3 K=39.16kN/mm

-2
Drift ratio -1%

-2%
Fig. 4.2.1-2 Hysteresis Loops of Specimen W2
1% 1.5% Drift ratio
30 40 50 60
Displacement (mm)
Drift ratio -1.5% -1%
102
2% Drift Ratio
-70 -60 40 SO 60 70
Displacement (mm)
K=77.4kN/mm
DF= -8 -7 -6 -5 -4 -3 -2 '800
J L_J 1 L_^ 1
Drift Ratio -2% -1% -1000 J
1000 1% 1.5% DriftRatio

i1 i 'i r
2 3 3.5 4 =DF
Pi=486kN
10 20 30 40 50 60 70
Displacement (mm)
DriftRatio -2% -1% -1000
103
1000 1.59 Drift Ratio

1%
i«-/Y-; 800 -
V—r -v- T
=DF
Pi=486kN
1
Specimen W6 a 400
200
?0 -60 -SO -40 -30 -20 10 20 30 40 SO 60 70

Displacement (mm)
if Mo -
Pi=-486kN
r -6oo -
K=53kN/mm
DF= j-800 -
-2
1 I 1 1 1 1
r -1000 -1
)rift Ratio -2% -1%
Fig. 4.2.1- 6 Hysteresis Loops of Specimen W6
4.2.2 Backbone Envelop of Load-Displacement Response
Fig.4.2.2-1 to Fig.4.2.2-2 show the backbone envelopes of the load-

displacement response of these walls. The envelope response confirmed the
difference between strength and stiffness between the monolithic walls and their
precast counterparts.
The comparison of monolithic wall with embedded channels and its precast
counterparts is shown in Fig. 4.2.2-1. It indicates that the lateral load-resistant
capacity of precast counterparts is almost equal to that of their monolithic
prototypes. However, the initial stiffness of these walls is decreased Wl, W3, W2,
in sequence. This means that the monolithic wall has the highest initial stiffness and
the wall with two shear connectors has a little higher initial stiffness in comparison
with that of wall with one shear connector. However, the difference in strength and
stiffness between the monolithic wall and its precast counterparts is not remarkable.
This hints that the horizontal connection can successfully transfer forces from upper
wall panel to the lower one in precast walls without losing much strength and initial
104
stiffness. Nevertheless, the precast walls showed total different failure mode and
different failure locations, which are discussed in subsequently sections.
Fig.4.2.2-2 shows the backbone envelopes of load-displacement response of

the walls with embedded 1-Beams. The characteristics of these walls exhibit the
similar behaviors as those of walls with embedded channels, except that specimen
W6 experienced connection failure. During load cycles spalling of concrete cover
occurred in walls with embedded I-Beam along inner boundaries of I-Beams, which
did not appear in walls with channels. The monolithic wall showed flexural ductile
behavior whereas precast walls exhibited sudden failure. The causes are discussed
in following sections.
750
600
450
300
150
-2 0
2
-*~ -150
a
-300
-450
-600
-750
-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60
Top displacement (mm)
Fig. 4.2.2-1 Backbone Envelopes of Load-Displacement Response of Specimens
105
-3
-60 -50 -40 -30 -20 -10 0 10 20 SO 40 50 60
Top displacement (mm)
Fig. 4.2.2-2 Backbone Envelopes of Load-Displacement Response of Specimens
4.2.3 Stiffness
A typical nonlinear relationship between induced forces or loads and

displacements, describing the response of a reinforced concrete component
subjected to monotonically increasing displacements, is shown in Fig.4.2.3-1. And
for purpose of routine design, the observed response curve is normally replaced by
two bilinear approximations. The slope of the idealized linear elastic response,
K = Sy/Ay , is used to quantify stiffness. This should be based on the effective
secant stiffness to the real load-displacement curve at a load of about 0.75 S . The
definition of stiffness of reinforcement concrete member under cyclic load is the

initial curved load-displacement characteristic will modify to close to the linear
relationship of the idealized response, Paulay (1992).
As recommended by Park (1988), in order to estimate the ductility level

which specimens fail, the yield displacement was defines as that corresponding to
the point of intersection of the horizontal line passing through the higher points of
the envelope curve and the inclined line passing through the beginning of the axes
106
and point of envelop curve corresponding 75% of the maximum strength. It was
considered that failure of the specimen occurs when its strength is decreased to 75%
of its maximum strength. The final value for displacement ductility is the ratio of
the displacement for a strength decrease up to 75% of the maximum strength,
divided by the yield displacement.
^ displacement
Fig. 4.2.3- 1 Typical Relationship for A Reinforced Concrete Element
(Paulay and Priestley 1992)
The initial stiffness of the precast walls is lower than that of their prototypes.
This is due to the postponed yield points in precast walls. The major reason is the
built-in tolerance between bolt and bolt holes in the horizontal connection. The
stiffness is mainly depends on the number of the shear connectors as compared in
Fig.4.2.2-2. The initial stiffness is calculated by
Where Kt is the initial stiffness of the specimens, S is yield strength of specimen,
approximately equal loSj , in Table.3.2.5-1, S is taken from experimental results
and not equal to St, A is the yield displacement obtain from the experimental
results.
107
Fig.4.2.3-2 shows the relationship of lateral stiffness versus drift ratio,

which calculated from backbone envelope of load-displacement. It shows that in the
elastic range the stiffness is highest and decrease with increasing drift ratio. It also
shows that the fastest decrease of lateral stiffness is in the elastic range. With
increasing drift the slope of lateral stiffness declines which means that the decrease
in stiffness becomes slow. It shows that at same drift ratio the later stiffness of
monolithic walls Wl and W4 exhibit higher stiffness in comparison with their own
precast counterparts.
200 -,
i
ISO
Drift Ratio (%>
Fig. 4.2.3-2 The relationship of Lateral Stiffness versus Drift Ratio
4.2.4 Cracking Patterns
The cracking patterns are important data to internal load paths and failure
modes. General features of the cracking patterns of all specimens are presented in
this section. The particular features of the each specimen are studied in later
sections.
108
Cracking patterns of the specimens are shown in Fig.4.2.4-1 to Fig.4.2.4-6.

The cracking patterns presented herein at DF equal to 1 and cycle number equal to
4 in elastic stage, and DF equal to 2 and cycle number equal to 5 and minimum
ductility factor in plastic stage. The cycle 5 is the first cycle in plastic stage and
ductility increases from 1 to 2, which means the transition from elastic stage to
plastic stage. The increment of number of additional cracks in this cycle is the
largest in any single loading cycle, which corresponds to prominent jump of
readings of the strain gauges on reinforcement shown in later sections, especially in
lower parts of the specimens.
Specimen Wl exhibited a flexural behavior during the test which finally

showed a concrete crushing failure at right lower corner with breaking weld of the
lowest horizontal reinforcement and buckling of the channels. The horizontal cracks
occurred in first cycle at the location of the horizontal reinforcement and propagated
from boundary downward to center of the wall. Near the boundaries the cracks
about 200mm length is nearly horizontal, while further to center they were inclined
to the horizontal. The horizontal cracks are almost evenly distributed, along at the
location of the horizontal reinforcement. The diagonal cracks occurred in the upper
part of the wall, direction of these diagonal cracks were along a diagonal line from
top corner of the wall to the opposite bottom corner and they joined the flexural
cracks at the boundaries. With increasing lateral load, the density of the horizontal
and diagonal crack in central part of the wall became higher. And width of diagonal
concrete bands formed between two adjacent diagonal cracks became smaller and
these bands finally showed an even distribution with a width about 200mm as
shown in Fig.4.2.4-1. When ductility reached 5, vertical cracks also appeared in
lower corners, which indicated spalling of concrete cover at these positions.
Meanwhile, a lower inclined crack along the diagonal reinforcement opened
considerably more than others and after several cycles, concrete at this place was
cracked with dense spacing, which showed as critical region, especially at ends of
the diagonal rebars, but in subsequent cycles, further deterioration did not occur.
Following spalling at the lower corners of the specimen, large permanent
deformation occurred in the vertical bars, which finally caused fracture of the
109
outermost vertical bars at both ends of the wall. This can be considered as the
failure stage, although no abrupt drop in strength was observed. A wide horizontal
crack as shown in cycle 15 occurred along the intersection between wall panel and
base beam which not appeared in its precast counterparts. A wide horizontal crack
as shown in cycle 15 occurred along the intersection between wall panel and base
beam which not appeared in its precast counterparts. Subsequent to this stage,
sliding at the base was observed. Later with a sound of breaking weld of the lowest
horizontal reinforcement, buckling of channels occurred accompanied by concrete
crushing at the lower corner. Length of the horizontal flexural crack at two ends of
the wall almost linearly decreased to zero from bottom to top of the wall with
decreasing bending moment in horizontal sections. In this specimen, in contrast to
the requirement on diagonal reinforcement for avoiding sliding shear failure which
requires diagonal reinforcement should cross all sections within a distance 0.5 lw or
0.5 Hw, which is less, above the base section (Paulay and Priestley, 1992), the
height of the sections with diagonal bars was only 0.31 lw, which also successfully
prevented the sliding shear failure.
Cracking patterns of specimen W2 and W3 which are precast counterparts

of specimen Wl are shown in Fig.4.2.4-2 and Fig.4.2.4-3. Before lateral load
reached half of capacity of the wall, the diagonal cracks mainly occurred in lower
panel of the wall and the inclination of these cracks was lower than those in
specimen Wl. After lateral load reached 0.75 Pt, there was a horizontal crack
passing through the horizontal connection. When ductility reaches 1, density of
diagonal cracks is higher than that in specimen Wl. Almost all main cracks were
produced when ductility increased from 1 to 2, which is a basis of internal load
paths of the wall. The distribution of horizontal cracks is different from those in
specimen Wl, especially, the density of them is high near the connections which
means relative large deformation occurred in this zone. From the cracking patterns
of specimen W2, the diagonal cracks in upper and lower panels look discontinuous,
especially cracks in upper panels. The horizontal crack at the intersection of the
wall panel and base beams which existed in specimen Wl did not appear in
110
specimen W2. The inclination of diagonal in upper and lower panels is lower
compared with that in specimen Wl. In late cycles two considerable wide shear
cracks in opposite directions appeared in along the lower diagonal reinforcement,
these cracks shaped as "V" with an obtuse angle. There are a lot of cracks with high
density, which mainly distributed in the lower panel along the inner channel
embedded in the wall. These dense cracks made concrete in this zone with dense
spacing, and finally caused spalling of concrete cover in last several cycles. The
spalling of concrete is serious and finally the specimen W2 failed in shear with
breaking of horizontal reinforcement passing through the spalling zone and
buckling of channel at this location. Specimen W3 failed with fracture of channel
near the connection with spalling of concrete and breaking of horizontal
reinforcement passing through this region.
Specimen W4 is a wall with I-Beam embedded at two ends acting as main

flexural reinforcement. Before ductility reached 1, the cracking patterns of W4
including distribution of horizontal cracks at ends and diagonal cracks in central
part of the wall was similar to W l , except that some vertical cracks appeared in
lower part of specimen W4 along the inner boundaries of I-Beam. With increasing
lateral load, more diagonal cracks occurred in direction from top corner of wall to
opposite lower corner. The almost vertical cracks along the inner boundaries of I-
Beam extended to higher place. Later these cracks caused spalling of concrete cover
in this location as shown in cycle 17 in Fig.4.2.4-4. Diagonal cracks with lower
inclination occurred in upper part of the wall. These cracks almost passed through
the center of top beam, which did not happen in specimen Wl. Vertical cracks
occurred at lower corner of the wall are similar to that in specimen W l , which
means spalling of the concrete cover at this position. In late several cycles, spalling
of concrete along the inner boundaries of I-Beam is serious from bottom to top of
the wall. When the lateral load reversed, a frictional sound can be heard. Finally, the
wall failed in flexural behavior with crushing concrete at left lower corner.
Compared with crack patterns of specimen W l , the distribution of diagonal cracks
of specimen W4 is more evenly which means the horizontal reinforcement in higher
part of specimen W4 experienced higher strain than those in specimen Wl.
Ill
Specimen W5 and W6 are precast counterparts of specimen W4. Crack

patterns of specimen W5 are shown in Fig.4.2.4-5. In early loading cycles before
the lateral load reached P,, the crack patterns is similar to that of specimen W4 with
diagonal cracks passing through horizontal connection from top of the wall to
bottom. Inclination of diagonal cracks in upper panel is lower than that in specimen
W4. Density of diagonal cracks in lower panel is apparently higher than that in
upper one. Width of the diagonal crack bands in lower wall panel is about 100mm
or less whereas it is about 200mm in upper panel. When ductility is greater than 3,
almost no additional cracks were produced. With increasing ductility, two main
diagonal cracks in each direction appear along the diagonal reinforcement similar to
those in specimen W2 and W3 and spalling of concrete cover occurred at the upper
ends of diagonal bars. In late cycles, spalling of concrete cover along the inner
boundaries of 1-Beam which happened in specimen W4 also occurred, however, the
vertical range is different. In specimen W4 the spalling of concrete occurred almost
in all height of the wall, whereas in specimen W5, spalling is in range of the lower
panel. Under reversed cyclic load, spalling continued to form two triangular zones,
which are more serious than that occurred in specimen W2 and W3. And concrete
crushing occurred at left lower corner of the wall. The specimen failed in shear with
breaking of the horizontal reinforcement passing through the serious spalling zone,
which is different from the fail mode of its prototypes specimen W4 which failed in
flexure.
112
c
K
E
^HHHJiM
u
--If 01
n.
w
i—
IT) e
\ v/S /\ \(\\' :
1)
(J
>^ ts
a
C
V
: i rfmM u
II
pt.
ax
e
3
:p^^ 7 '••
SJD
113
<D
>^
u
in
en
II
Q
a
E
an
o aE
>^ cs
U ft
(N fcJC
II
ft 2au
Q a
0
i
114
ON
<D
O
0
Q
Q.
S
u -
>1 e
U
CN M
II C
Q u
-
U
CI
si
U
T
Q
•ir
IX
O
s
I
o
>-,
o
fN c
II
a
U
o
O
116
a
a
B
••*
u
u
n,
o
>-,
c
a
U a.
ex;
CN e
II
ft!
Q
s
u
i
rr
ox
fa
(U
u
Q
117
DF=1 DF=2
Fig. 4.2.4-6 Cracking Patterns of Specimen W6
A summary of the characteristic of the experimental results is shown in Table
4.2.4-1.
4.2.5 Flexural and Shear Deformations
The total displacement of the walls was monitored by LVDT mounted on

steel frame, which is composed of flexural and shear deformations and sliding of
the walls relative to its base beam. To study the contributions of flexure and shear
components, the total horizontal displacement is decomposed into shear and flexure
and their contribution to total displacement is presented by ratios to total
displacement separately. Flexural deformations is evaluated by the sum of top
horizontal displacements caused by rotations of the intervals and shear deformation
can be obtained through measuring the changes occurred in the two diagonal
LVDTs as discussed in previous sections. All assumptions made in calculation
have been explained in section 3.2.2.
118
Table 4.2.4-1 Characteristic of Experimental Results
Wl W2 W3 W4 W5 W6
(-) (+) (-) (+) (-) (+) (-) (+) (-) (+) (-) (+)
s
> 522.8 522.8 522.8 608.9 608.9 608.9
(kN)
8.57 10.91 12.59 14.11 10.99 11.47 6.61 9.31 12.24 10.07 - -
(mm)
A -aver
9.74 13.35 11.23 7.96 11.16 -
(mm)
570.29 554.28 576.96 541.6 601 547 785.73 728.36 749.4 731.7 723.7 -
(kN)
53.67 52.02 45.99 58.65 53.67 16.8

(mm)
Drift (%) 2.11 2.05 1.81 2.31 2.11 -
M 5.5 3.9 4 7.4 4.8 -
58.33 39.16 46.56 77.4 55.12 -
(kN/mm)
Ysy - 0.99 1.02 - 0.97 -
Yt, - 1.3 1.2 - 1.35 -
1 SU - 0.99 1.02 - 0.97 -
y'A« - 0.97 0.856 - 0.92 -

Failure
Flexural Shear Shear Flexural Shear Connection
Mode
Where S is the yield strength, A v is the horizontal displacement at yielding, Su is the ultimate strength of
the specimen, A ( is the ultimate displacement of specimen at failure, drift is defined as the ratio of Au IH ,
H is the height of the wall measured from base line of wall to the horizontal axis of the top beam, ft is
ductility factor defined as the ratio of A |( / A , K: is the initial stiffness defined as the ratio of S . / A ,
Y is the ratio of yield strength of precast walls to that of their monolithic counterpart, yA, is the ratio of yield
displacement of precast walls to that of their monolithic counterparts, ym ' s the ratio of ultimate strength of
precast walls to that of their monolithic counterparts, yAll is the ratio of ultimate displacement of precast walls
to that of their monolithic counterparts.
Flexural and shear deformations of each specimen under cyclic load are
presented in Fig.4.2.5-1 to Fig.4.2.5-10. The total horizontal displacements of these
specimens are decomposed into shear and flexural deformations which are shown in
Fig.4.2.5-11 to Fig.4.2.5-15. Ratios of flexural deformation and shear deformation
119
to the total horizontal displacements during the loading courses are also given in
Fig.4.2.5-16 to Fig.4.2.5-20, which are used to judge failure mode of each specimen.
The hysteresis loops of flexural deformations versus lateral load of

monolithic walls, as shown in Fig.4.2.5-1 and Fig.4.2.5-7 are much fatter than those
of shear deformations given in Fig.4.2.5-2 and Fig.4.2.5-8, which means that the
energy during test is mainly dissipated through flexural deformations. The
hysteresis loops of flexural deformations against lateral load of precast specimens is
also fatter than their corresponding hysteresis loops of shear deformations. But
these hysteresis loops of flexural deformations of precast walls are narrower in
comparison with those of their monolithic counterparts, which means that less of
energy is dissipated through flexural deformations in precast walls.
Compared with deformations occurred in their monolithic prototypes, in

precast walls, flexural components of total deformation is less at the same
displacement level in plastic stage as shown in Fig.4.2.5-11 to Fig.4.2.5-15 than that
in monolithic prototpes. This is can be also inferred from the different cracking
patterns in monolithic prototypes and their precast counterparts as shown in
Fig.4.2.5-1 to Fig.4.2.5-5. The density and width of horizontal cracks at two ends of
monolithic walls are higher than that in precast walls and there is also a wide
horizontal crack go across along the base beam where bond slip occurred.
In precast specimens, W2, W3 and W5, bigger portion energy dissipated

through shear deformation than that in monolithic walls. The energy dissipated
through shear deformation mainly came from plastic deformation of horizontal
reinforcement and diagonal bars.
The hysteresis loops of shear deformation of monolithic walls, Wl and W4,

exhibit less pinching as compared with the walls without diagonal reinforcement
predicted by FEM in later chapter and this phenomenon is consistent with
prediction by Sittipunt and Wood (1995).
120
The ratio of flexural component to the total deformation increases during the
last cycle in specimens Wl and W4, which indicates that the increment of flexural
deformation is dominating in this stage. This increment indicates that the walls tend
to flexural failure, which is consistent with the observed failure modes of the two
specimens. However, in precast specimens, W2, W3 and W5, ratios of shear
deformation to total horizontal displacement increase while the proportion of
flexural deformation decrease in plastic stage, and in last cycle, a sudden increment
occurred in ratio of shear deformation which means that the specimens experienced
a shear failure. This is consistent with failure modes of these precast specimens
which were observed in experiment.
121
600 -,
-,
?« .
1 -xs 500 -
«
«
400 -
<s
11 11 «>« 300 -
1*4
Wl
-20 -15 ?10 10 15 20

-mo
Shear Deformation (mm)
-600
Fig. 4.2.5-2 Shear Deformation of Specimen Wl
15 20 25 30 35
Flexural Deformation (mm)
Fig. 4.2.5-3 Flexural Deformation of Specimen W2
122
10 15 20 25
-600
Fig. 4.2.5-4 Shear Deformation of Specimen W2
600 i
-40 10 20 30 40
10 H
'•400
-500 -i
-600
123
600
20 30 40 50
-800 -L
124
10 20 30
800 -i
1
•St
«
600 -
1
W5
_l $ 4wr,
-so -40 -30, 20 30 40 50

1600
-800
125
s 800
600
I 400
Lat
200
-800
600
500 1 1
400
300
S> 200 1
>F7 4
^ 100
«
,3 o
1 -100
-300 —•— Total Displacement

M1
-400 —•— Shear Displacement
-500 —*— Flexural Displacement
-600
-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60
Horizontal Displacement (mm)
Fig. 4.2.5-11 Lateral Loads versus Deformations of Specimen W l
126
-20 -10 0 10 20
Fig. 4.2.5-12 Lateral Loads versus Deformations of Specimen W2
600
500 1 I
400 T **
300
200 1 1
100 W3 ,
0
-100
-200
-300
• A —•— Total Displacement
-400 \ 1 —•— Shear Displacement

\ *X T" —A— Flexural Displacement
-500
-600
-700
V. 1 \*A f-
-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60
127
800
600
400
?r 200
ê
teral Load
-200
-400
-600
-800
-1000
-70 -60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60 70
800
1 1
600
400
1
§L 200 W5
e
1 i i i i i
i
| -200 —•— Total Displacement
—•— Shear Displacement
-400 —*— Flexural Displacement
-600
-800 -60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60

128
1
0.9
§
0.8
0.7
0.6
0.5
J
J
u
\
""•^ i m— —• 1
•
0.4
0.3
—* <k *
0.2 **—
/ —A— Shear
0.1
—•— Flexur e
0 1 i 1 1 1 1
-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60
Horizontal Displacement
Fig. 4.2.5- 16 Ratios of Deformations to Total Displacement of Specimen W l
0.9
0.8
0.7
0.6
% 0.5
0.4
0.3
0.2
*—Fle xure
0.1
+-She ar
0 1 1
-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60
Fig. 4.2.5-17 Ratios of Deformations to Total Displacement of Specimen W2
129
0.9
0.8 \J
7-
0.7
0.6
*•
% 0.5
0.4 A -A
0.3 ISV, ^
0.2 ^ \ —*—snear
—t«— Flexure
0.1
0 1 i — •"
-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60
0.9 —w—
0.8 ^
0.7 \
0.6
0.5
0.4
a
0.3
0.2 ^ /
-Shear
^ \
0.1 — Flexu
0
-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60
130
0.9
0.8
0.7
0.6
Ratio
0.5
0.4
0.3
0.2 —*—Shear
—•— Flexure
0.1
0
-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60
4.2.6 Vertical Elongation of the Walls
Different numbers of LVDTs were placed on both ends of these specimens

to measure vertical deformation, that is, vertical elongation at these local positions.
These vertical elongations can be used to determine the extent of the cracking and
yielding of the flexural reinforcement. Cycles beyond yielding lead to a ratchet
effect. This describes the vertical stretching of the wall caused by the irrecoverable
strain that accumulated in the post-yield cycles. Records of LVDTs mounted on left
side of specimens Wl are shown in Fig.4.2.6-1 and Fig.4.2.6-6. Significant
elongations of the end of the wall, while in tension, are evident, due to flexural
cracking and the straining of the reinforcement at the cracks. The elongation of the
ends increased up to the peak load.
Fig.4.2.6-1 to Fig.4.2.6-4 exhibit the local elongations in different vertical

intervals of specimen Wl. The peak values of local elongations show the trend to
increase from top to bottom. The relatively apparent residual deformations
measured by LVDTs on left side of the wall when lateral load is zero indicate that
the yielding of flexural reinforcement at end of the wall was widespread. Fig.4.2.6-6
131
shows a big part of the total elongation of the wall is taken up by elongation in the
lowest interval, which indicates vertical elongation mainly concentrates in lowest
interval. This is also indicated by the cracking patterns of specimen Wl shown in
Fig.4.2.4-1 in which the density of flexural cracks, that is, horizontal cracks at two
ends of the specimen, is higher in lower segment than that in upper part of the wall.
This verified that the accumulated flexural cracks and plastic straining of
reinforcement consists the vertical elongation.
In the reinforced concrete structures, the main capacity of energy dissipation

is contributed by plastic deformation of reinforcement. The records of the LVDTs
can be used to estimate the deformation of the vertical reinforcement, which was
also evaluated from readings of strain gauges placed on reinforcing steels discussed
in detail in subsequent sections. Therefore, it can be concluded that a key part of
energy dissipation that occurred during the reversed cyclic loading test was from
lower 1/4 segment of the specimen Wl where plastic hinge formed.
Vertical deformations of the different vertical intervals of the precast

specimen W2 and total vertical elongation are shown in Fig.4.2.6-7 to Fig.4.2.6-13.
In this specimen, as shown in Fig.4.2.6-9, the local deformation in the interval
covering the horizontal connection comprises the main part of the total elongation,
whereas in specimen Wl the local elongation in the lowest segment does. In
specimen W2, the local elongation in the lowest interval is smaller in comparison to
that in specimen Wl. The cracking patterns of the specimen W2 as shown in
Fig.4.2.4-2 hint these characteristics by the density of the flexural cracks at the two
ends of the wall, which means that the energy dissipation concentrates in the region
of the horizontal connection, whereas the energy dissipation in the lowest segment
of the wall W2 was not so much as that in specimen Wl. Though the total vertical
elongation that occurred in specimen W2 was higher than that of specimen Wl,
considering the capacity of energy dissipation deduced from the configuration of
the hystersis loops, the capacity of specimen W2 exhibited similar or a little lower
energy dissipation than that of specimen Wl. The reason was that not all of the
local vertical deformation in the interval covering the horizontal connection comes
132
from deformation of the steel reinforcement. As mentioned early, the built-in

tolerance between the bolts and bolt-holes in the horizontal connection was a key
variable of the local elongation. In specimen W2, in last two cycles the local
deformation of the second lowest interval was high which corresponded to the
failure location observed in experiment.
Fig.4.2.6-14 to Fig.4.2.6-20 show the characteristics of vertical elongations

of the specimen W5. The deformations of specimenW5 which is a precast wall with
embedded I-Beam exhibit similar characteristics of specimen W2. The local
elongation in the lowest intervalwas smaller compared with that in the interval
covering horizontal connection and also smaller than that in specimen W4 in which
there is a wide horizontal crack occurred along the base line. The elongation in the
interval covering horizontal connection is the main contribution to the total vertical
elongation similar to that in specimen W2.
The maximum values of total vertical elongation of each specimen are listed
in Table 4.2.6-1. And the ratios of maximum vertical elongations to their heights are
also given in this table, which shows that the precast walls experienced higher
vertical deformations than their own monolithic prototypes. This will cause higher
rotation demand on other structure elements connected to the wall. However,
elongation analysis showed that the main vertical elongation of W2 occurred in the
connection. It indicated that the elongation could be reduced by improve the
performance of the connection, for example, using high strength friction bolts to
replace the normal bolts.
133
-600
-0.8 -0.7 -0.6 -0.5 -0.4 -0.3 -0.2 -0.1 0.1
LVDT2 (mm)
Fig. 4.2.6-1 Local Vertical Elongations of W1-LVDT2 (cont.)
600
500
400
300
200
100
-J 0 ~ f - ^ j ^ — ' i ' '" ~~~ i 1 r~ ~
1 -100 1
a -200
a
-300
-400
-500 -
Wl
-600
-0.5 0 0.5 1 1.5 2 2.5
LVDT3 (mm)
Fig. 4.2.6- 2 Local Vertical Elongations of W1-LVDT3 (cont.)
134
Fig. 4.2.6-3 Local Vertical Elongations of W1-LVDT4 (cont.)
10 15 35
LVDTS(mm)
Fig. 4.2.6-4 Local Vertical Elongations of W1-LVDT5
135
600
500
400 -
300
^ 200
^ 100
a
5 °
1
| -100 Iff/ / / s ^-^^^^^ '
3 -200
v
-300 a 16?\^^S^=***^
-400 y^^ r Ff7

1
-500
-600
10 15 20 25 30 35
Sum o>JLeftLVDTs (mm)

Fig. 4.2.6-5 Total Vertical Elongations of W l
35
30
Z1W5
25 Swm O/Z, KDT.S 0/1 le/* S/Vte
i 20 FT;
1
s 75
"5
18
0
-5 H
-70
0 200 200 300 400 500 600 700 800 900
Scan Number
Fig. 4.2.6-6 Vertical Elongations of W l versus Scan Number
136
600
500 -
400
300 /S^"-
200
100
0
-100 J '1 1 ' "~1
-200
-300
/
1/
1
-400
-500 W2
-600
-0.4 -0.2 0 0.2 0.4 0.6 0.8
LVDT2 (mm)
Fig. 4.2.6-7 Local Vertical Elongations of W2 (cont.)
600
500
400
300
200 -
100
0
•Jj^~ 1
% -loo
«3 -200 -
<77? €
-300
-400
-500 W2
-600
1 2 3
LVDT3 (mm)
137
600
500
400 -
300
200
100
0
"1
-100 \lfLr S
>3 -200
-300 \u I I I
-400
-500 W2
-600
-10 10 20 30 40
LVDT4(mm)
10 20 40
LVDT5(mm)
138
600
500
400
300
200
100
! 0
-100
^ ^ ^
1
-3 -200
-300
-400 £
-500
W2
-600
-3 -2 0 1 2 3 4 5 6
LVDT14(mm)
600
500
400
300
200
sI 100
-100
-200
0
U//7 / s / y '
i
-300
\
-400
-500
W2
-600
-700
-10 0 10 20 30 40 50 60 70 80
Sum ofL VDTs on Left Side(mm)
Fig. 4.2.6-12 Vertical Elongation of Specimen W2
139
80
70 Sum ofL VDTs on Left Side
60 -LVDT5
I. LVDT14
SO 100 ISO 200 250 300 350 400

Scan Number
Fig. 4.2.6-13 Vertical Elongations of W2 versus Scan Number
-1.5 -0.5 0 0.5 1.5

LVDT6(mm)
140
800
700 1 1
)
600 II *
II
500
400
300
I53
200
100
ill W5
0 jO\ i i ...
2 -100
-200
-300
-400
-500 jtfVjA-
-600
-700
-800
-5 0 5 10 15
LVDT7(mm)
Fig. 4.2.6-15 Vertical Elongation of Specimen W5 (Cont.)
-5 10 15
LVDT8 (mm)
141
800
1 1
700 -
600
500
%
400
300 ws
s 200 ^\*^s^
100
3 0
2 -100
| -200
^ -300
-400
-500
-600
-700
-800
-3 -2 -1 0 1 2 3 4 5 6 7 8 9 10
LVDT15 (mm)
142
-20 0 20 40 60 80 100 120 140 160
Sum ofLVDTs On Right Side(mm)
Fig. 4.2.6-19 Vertical Elongations of W5
90
80 - Sum ofL VDTson Right Side
70 LVDT8
-LVDT15
60
50
I"3
40
30 W5
20 -
10
0 .^VAA/^/M
-10
100 200 300 400 500 600 700 800
Scan Number
Fig. 4.3.6-20 Vertical Elongations of W5 versus Scan Number
Table 4.2.6-1 Vertical Elongation of Specimens
143
AH
AH/H
Specimen (Max. Vertical
(%)
Displacement mm)
Wl 21.83 0.95
W2 57.72 2.51
W3 52.86 2.3
W4 30.36 1.32
W5 49.77 2.16
Note: H=2300mm,
A//=average of sum of records of LVDTs on left and
right sides when lateral load is zero.
4.2.7 Curvature
Generally rotation is the most favorable inelastic structural deformations in

the reinforced concrete structures. It is based on the tension of reinforcement and
compression of concrete on the other side of the RC member and will not lead to a
sudden failure. The maximum rotation occurs in the plastic hinge when the structure
reaches post-yield stage. In this study, in order to investigate the overall behavior of
the specimens, curvatures along the height of the walls are given. The horizontal
axis is the distance from the base beam. The vertical axis is calculated curvature,
which is plotted in the middle of the intervals determined by the gauge lengths
between adjacent LVDTs.
The curvatures of each specimen are obtained from the records of LVDTs
placed on both sides of the specimens. The calculation of curvature is explained in
previous section. Curvatures at peak loads in first cycle at each displacement level
are adopted.
As shown in Fig.4.2.7-1 and Fig.4.2.7-4, in the monolithic walls the rotation

mainly concentrates in 1/4 height of the wall. This is consistent with experimental
observation. However, as given in Fig.4.2.7-2, Fig.4.2.7-3 and Fig.4.2.7-5, in
precast walls the rotation extends to 2/3 height of the specimens. And the rotation
144
occurred in the intervals covering the horizontal connections is high, but this can
not be considered as the favorable deformation in plastic hinge, because this mainly
comes from the built-in tolerance between bolts and their wholes and discontinuity
of the vertical bars.
J
200 400 600 800 1000 1200 1400 1600 1800 2000
Vertical Distance froom the Base (mm)
Fig. 4.2.7-1 Curvature of Specimen Wl
145
-40
200 400 600 800 1000 1200 1400 1600 1800 2000 2200
Vertical Distance from the Base (mm)
Fig. 4.2.7-2 Curvature of Specimen W2
200 400 600 800 1000 1200 1400 1600 1800 2000 2200
146
I -10
200 400 600 800 1000 1200 1400 1600 1800 2000 2200
Vertical Distance front the Base (mm)
30
20
I 10
a -10
-20
X-DF-4
-30
200 400 600 800 1000 1200 1400 1600 1800 2000 2200
4.2.8 Sliding of the Walls
147
Sliding of each wall during cyclic loading course was monitored by the
LVDT placed above 250mm above the base line. Sliding shear deformations
generally develop at the height at which vertical reinforcement yields. It is proposed
by Penelis (1997) that the length of the plastic hinge is according to the relation
/ = 0.08/0 + 6<j> where /0is the element's length from the point of zero moment, to
the point of maximum moment, and <f> is the diameter of the largest steel bar.
Fig.4.2.8-1 to Fig.4.2.8-4 show that sliding contributed to the total horizontal
displacement, especially in last cycle. The contribution of the sliding increased
particularly when the horizontal crack running through the base line appeared.
The vertical cracks caused spalling of concrete cover at the compression

corner. The spalling finally lead to concrete crushing at the lower corner, readings is
beyond the range of the LVDT or the records are not the true value of sliding. It is
evident as shown in Fig.4.2.8-1. However, it still provides a great deal of
information in middle plastic stages. The malfunction occurred in LVDT to record
sliding of specimen W3, which is not found during the test.
Sliding occurred in the monolithic walls and their precast counterparts are
compared as shown in Table.4.2.8-1. The rate of sliding shows that monolithic
prototypes experienced higher sliding deformation near the base than their precast
counterparts did, which is about 20% of the total horizontal displacement in
monolithic walls, whereas this rate in precast walls is about 10%. This is consistent
with the experimental observations.
148
100 200 300 400 500 600 700 800

Scan Number
Fig. 4.2.8-1 Sliding Deformation of Specimen Wl
50 100 150 200 250 300 350 400 450 500

Scan Number
Fig. 4.2.8-2 Sliding Deformation of Specimen W2
149
0 50 100 ISO 200 250 300 350 400 450 500 550 600 650 700
Scan Number
150
Table 4.2.8-1 Ratio of Sliding of Specimen to Total Displacement
Specimen SN Displacement Sliding S/D (%)

Wl 604 33.6 5.76 17
W2 386 4.81 45.54 10
W4 750 9.3 58.65 20
W5 574 5.35 48.22 12
4.2.9 Deformations at Horizontal Connections
In precast specimen walls, the deformations occurred in the horizontal

connections are important parameters affecting performance of these walls. Two
horizontal LVDTs, two diagonal LVDTs and four vertical LVDTs, as shown in
Fig.3.2.2-8, are used to monitor the horizontal deformations of the connection. The
performance of the horizontal connection in specimen W2 is taken out as an
example which is shown in Fig.4.2.9-1 to Fig.4.2.9-5. Fig.4.2.9-1 and Fig.4.2.9-2
show the records of L23 and L24 respectively, which are located at upper and lower
walls panels near the horizontal connection. The difference between L23 and L24 is
seen as the total horizontal deformation of the connection. The hysteresis loops of
L23 and L24 are not symmetrical, which may be due to the horizontal expansion at
these locations caused by internal forces. This will be explained in later sections.
Fig.4.2.9-3 shows that the total horizontal deformation is almost symmetrical except
in last cycle. The total horizontal deformation of the connection is decomposed into
shear and flexural deformations as shown in Fig.4.2.9-4 and Fig.4.2.9-5. The
flexural deformations exhibit stable behavior. And the shear deformation of the
connection experiences some slippage which was caused by the friction behavior
along the horizontal connection.
151
600 T
, s
o
W2
-30 -25 -20 10 15 20 25 30
Fig. 4.2.9-1 Records of L23 versus Lateral Load of W2
600
H
-+-
IS 20 25 30 35 40
•L24
±
600
Fig. 4.2.9-2 Records of L24 versus Lateral Load of W2
152
^Latera Load (kN)

1 1 500
/MP?
npo
poo
W2
+
-40 -30 10 20 30 40
Horizontal Deformation (mm)
• Total Deformation at Connection
'600
Fig. 4.2.9-3 Total horizontal Deformation at Horizontal Connection of W2
600
i—r *
J
—h 1
-10 4 6 8 10
Fig. 4.2.9-4 Flexural Deformation at Horizontal Connection of W2
153
^ 600 -
"H <tvv y
r W2
i
y
^ t ^
^-^S^^^^^\ 11 JIJ
7
i " ' ' '
8 -fT f/l-4 *Z^^2 4 6 8
~fj\i$h - Shear Deformation (mm)
Aif-
/ \
Ao -
*s
V^ ^y -5oo
•600
Fig. 4.2.9-5 Shear Deformation at Horizontal Connection of W2
4.2.10 Capacity of Energy Dissipation
There are different ways to evaluate the energy dissipation capacity of a

structural member, which is a properly that has long been recognized to be of
importance with respect to the seismic performance of a structure. The total amount
of hysteresis energy, that is, the cumulative area of the hysteresis loops when the
structure fail, is an important parameter, but does not clearly indicate the level of
deformation at which this energy is dissipated, and is also subjected to the
uncertainties involving the definition of failure. Herein, when a 25% drop in
strength (post-peak) occurs in a structure, it is considered to be failed.
Because strength of the walls varies, meaningful comparison can only be

made between walls with similar strength. The energy dissipation also depends on
the maximum displacement attained. In this research, all specimens have the same
height, so they can be assessed directly. The total cumulative energy dissipation is
dependent on the loading regime imposed, and in this study, can not be used on its
own as a comparative measure for assessing performance. Because different loading
regimes for these specimens were applied in this study, the total accumulated
154
energy is not suitable to be used to evaluate the capacity of energy dissipation of

these specimens. Therefore, in order to compare the capacity of energy dissipation,
the energy dissipation per displacement is presented in Fig.4.2.10-1 to Fig.4.2.10-5
and used as a tool to assess the performance of these walls; in each graph the virgin
cycle and second cycle are given. The energy dissipation is drawn against the total
horizontal displacement of the top beam, that is, at each displacement level the
energy dissipation of the specimen in each cycle is divided by the horizontal
displacement. And the energy dissipated in each cycle at every displacement level is
calculated separately.
From the figures of energy dissipation of all these specimens, the

observations show that in prior-yield stage a very small energy is dissipated.
However, in the post-yield stage the virgin cycles dissipate more energy than the
second cycles, especially in precast walls. The walls failed in flexure there is a
sharper rate of dissipation of energy, which is almost constant up to failure. A drop
of this rate indicates shear deterioration, which means significantly decrease in the
total cumulative energy.
Fig.4.2.10-6 shows the comparison of energy dissipation per displacement

of all specimens, a). Specimens with embedded I-Beams exhibits higher energy
dissipation per displacement, which indicates that, with same loading history, the
cumulative dissipated energy of specimens with embedded I-Beams has a higher
capacity of energy dissipation than those walls with embedded channels. For
monolithic walls, W4 has 13.1 percent higher capacity of energy dissipation at the
peak point than Wl does, and for precast walls, W5 has 6.5 percent higher than W2
and 10.2 percent than W3. The major reason is the higher reinforcement ratio due to
bigger sectional area of I-Beam than Channel, which caused higher strength of W4
than Wl at same displacement level, b). In pre-yield stage, the difference in energy
dissipation between monolithic specimens and their relevant precast counterparts is
about 8 percent for walls with embedded Channels, 10 percent for walls with
embedded I-Beams. In post-yield stage, this difference for walls with embedded
channels is 11.9 percent and is 18.8 percent for walls with embedded I-beam.
155
A— First Cycle
• — Second Cycle
20 30 40 50 60
Fig. 4.2.10-1 Energy Dissipation per Displacement of Specimen Wl
4 — First Cycle
• — Second Cycle
10 15 20 25 30 35 40 45 50
Fig. 4.2.10-2 Energy Dissipation per Displacement of Specimen W2
156
First Cycle
Second Cycle
10 15 20 25 30 35 40 45 50
First Cycle
Second Cycle
20 25 30 35 40 60
157
±— First Cycle
m— Second Cycle
15 20 25 SO 50
1000
900
i
i
i
700 / / -•' -<Â^^^^tî-
/ / r'fr^*^ ^ * *
600
/ / /' •-—Wl
<• 500 / /.// • W2
V •'//
5 400 J •'// * W3
--^-W4
300 A^ -•-•--- ws
200
100 ' Ti//
0
10 20 30 40 50 60 70
Displacement (mm)
Fig. 4.2.10-6 Comparison of Energy Dissipation per Displacement of All Specimens
158
4.3 Theoretical Study
4.3.1 Shear Strength
Because the horizontal load was applied gradually in successive loading cycles, the
strength of all specimens was evaluated analytically at any state of behavior,
particularly at yield and ultimate state.
Eurocode 8 is probably the first document of regulatory character that separately

treats the different modes of shear failure in RC walls subjected to seismic loading
by providing a different design equation for each of the following modes:
• Diagonal tension failure
• Diagonal compression failure (web crushing)
• Sliding shear
Resistance VKJ3 against diagonal tension failure is calculated by superimposing a

concrete contribution V. to a web reinforcement contribution V ,, that is,
cd wd ' *
V =V +V (4 3 1-1)
Vcd = 0.06 JJ; (1.2 +40Pl)bz (4.3.1-2)
T, =[ph(a -0.3) + A ( l - 3 - « )]/Az (4.3.1-3)
z
Where /?, is ratio of tension reinforcement, bK is thickness of web; represents
effective internal lever arm, about 0.8 lw (distance between the resultants of tensile
and compression stresses at the section considered); ph and pv are ratios of
horizontal and vertical web reinforcement, respectively; f is design strength of this
reinforcement; a t is the shear-span ratio equal to M/(V7W) where M is moment, V is
shear force, lw is wall length.
Equation (4.3.1-3) implies that for at=\.3, only horizontal web reinforcement is
contributing to shear strength, whereas for at =0.3, only the vertical web
159
reinforcement is resisting shear; both types of web reinforcement are taken into
account for 0.3<at <1.3.
Herein in order to compare theoretical and experimental average shear stress,

formulas (4.3.1-1 ~ 3) are expressed by stress:
^3=r,+rw (4.3.1-4)
*•- = ^ ^ = 0.054^(1.2 + 40/?,) (4.3.1-5)

bwz
*•- = ^ ^ = 0.8[ A (a.-0.3) + A (1.3-a,)]/„ (4.3.1-6)
Where rM, is diagonal tension stress, TCJ is stress contribution by concrete to r„ 3 ,
r w is stress contribution by web reinforcement to r„ 3 .
Eurocode 8 requires an explicit calculation of the resistance of walls against sliding

shear failure; this resistance is assumed to be made up from the dowel action of
vertical reinforcement, the shear resistance of cross-inclined bars, and the frictional
resistance at the sliding plane. For squat walls(Ai //,<2.0), it is suggested that in
Eurocode 8 that at least half of the design shear at the base should be carried by
diagonal bars(at higher levels, the contribution of these bars may be reduced to one-
fourth the total shear). Diagonal bars must be properly anchored beyond the sliding
plane and should cross all section of the wall within a distance of 0.5/^ or 0.5 hw
(which is smaller) above the critical section.
Eurocode 8 gives formula to calculate the resistance of walls against sliding shear
failure
Kd,=VM + V, + V, (4-3.1-7)
Vm =rmn{\3YJAsl{fJJ\Q25fyÂ,} (4.3.1-8)
K=ZAJ,*cos0 (4.3.1-9)
Vf^mm{Mr[YJAJyd+NJ^ + MJz]A25fJltbJ (4.3.1-10)
160
Where Kîs dowel action of vertical bars, Jîs the shear resistance of diagonal bars
(at angle (/>) with assumption that diagonal bars yield simultaneously in tension and
compression., and Vu represents the friction resistance. V A is the sum of the areas
of the vertical bars in the web and of purposely arranged bars in the boundary
elements of the wall (that is, bars not required for flexural strength); Ât. is the
sum of the areas of all diagonal bars in both directions,/^ = f'Jl.S is the design
strength of concrete; pf is the concrete-to-concrete friction coefficient under cyclic
action (usually taken equal to 1.0 for rough concrete); E,=xllw is the normalized
neutral axis depth; MSd is the design moment, and NSd is the design axial force at the
potential sliding plane.
The diagonal compression failure mode (crushing of concrete struts formed in the
web between inclined shear cracks) might be a problem in heavily reinforced walls,
typical of multistory buildings in seismic areas, particularly when strong boundary
elements of flanges are combined with relatively thin webs.
Formulas given by ACI 318-2002: (unit: lbs, psi) are the followings:
Vn=V+V (4.3.1-11)
V=3.3J7hd + ^ - (4.3.1-12)
r~ N
/,. l.25Jfc + 0 . 2 ^
or V. = o.6j/;+- Lh hd (4.3.1-13)
K. 2
Horizontal reinforcement:
A/yd
V = (4.3.1-13)
Vertical reinforcement:
p„ =0.0025 + 0.5 2 . 5 — ^ ( A -0.0025) (4.3.1-14)
161
The calculated Vn =570kN, therefore the average shear stress is 2.26MPa. And in
ACI 318-02 there is no explicit formula to determine the shear strength to prevent
sliding shear failure, but an indirect control, an upper bound of 0.81 y/ £ MPa
(10y/ c ' psi) is provided to prevent this type of failure.
For specimens with embedded channels, Wl, W2 and W3, comparing ultimate
strength from test with results from code (ACI 318-2002) in which influence of
ratio of height to width is not taken into account, it shows that the design strength of
wall can effectively resist the ultimate lateral load, that is, flexural failure controls.
However, horizontal shear failure occurred in W2 and W3.
Table 4.3.1-1 Comparison Of Calculated Results Basing On Codes And Experimental Result
Experimental
Eurocode 8 ACI 318-02 Average Shear
stress
specimen T
cd T
wd T
Rdl
T
Rd,s T
n T
aver
(MPa) (MPa) (MPa) u (MPa) vz (MPa) U (MPa) u

Wl 0.54 2.72 3.26 0.55 2.02 0.34 2.26 0.38 2.23 0.37
W2 0.52 2.72 3.24 0.56 2.02 0.35 2.26 0.39 2.22 0.38
W3 0.53 2.72 3.25 0.56 2.02 0.35 2.26 0.39 2.27 0.39
W4 0.55 2.72 3.27 0.55 2.16 0.36 2.26 0.38 3.00 0.50
W5 0.53 2.72 3.25 0.56 2.16 0.37 2.26 0.39 2.94 0.50
W6 0.52 2.72 3.24 0.56 2.16 0.37 2.26 0.39 - -
4.3.2 Theoretical Model (Macro Wall Model)
A. Introduction
This section is devoted to the development of a theoretical model of precast

shear wall to predict the relationship of lateral load and displacement. To cater for
the configuration of the precast walls and to capture the behavior of the horizontal
connection which is composed of two vertical steel elements at boundary to
resistance moment and one or two shear connectors in the central zone to take the
162
horizontal shear force. Macro model based on the nonlinear springs connected by
rigid beams, as shown in Fig. 4.3.2-1, originally suggested by Kabeyasawa et al
(1982), is the most suitable model. The two outer springs represent boundary
columns while a horizontal spring at center works as web of wall to resist the lateral
shear force, a vertical spring at center represents web of wall to take the vertical
load and a rotation spring at center represents the contribution of web for the
resistance of roation.
To improve the accuracy of simulation on shear walls, Vulcano & Bertero (1987)
suggest a vertical spring model by a set of springs, as seen in Fig. 4.3.2-2a, which
simulate the physical behavior of cracking of concrete and yielding of
reinforcement. The single spring on the top works as uncracked concrete while the
parallel springs below model cracked concrete and steel, respectively. The steel
spring follows a bilinear curve, and the concrete cracking-spring either seizes to act
(crack state) or takes up action (closing of crack).
Vulcano et al (1988) replaced the rotation spring by several additional vertical

springs to simulate the axial behavior of the wall web shown in Fig. 4.3.2-2b. This
method was able to model the gradual yielding of the vertical reinforcement more
smoothly, but it consists of more components and thus leads to a more complicated
model.
Fig. 4.3.2-1 Original Macro Model by Kabeyasawa et al (1982)
163
mz^^T^^zmzzzqzzzzmzzz^zzzzzzzzm
a) Model for Outer Spring b) Multiple Vertical Spring Model by Vulcano at el (1988)
by Vulcano & Bertero
(1987)
Fig 4.3.2-2 Suggestion for Improved Macro Models
A simple model suggest by Linde (1993), as shown in Fig. 4.3.2-3, consists

of three vertical springs, two at boundaries and one at center, and a horizontal
spring at center. The rotation spring in Kabeyasawa model is omitted. A derivation
of the properties for the outer vertical springs was done to satisfy the simulation the
flexural behavior. The horizontal spring, modeling shear behavior, continued its
function.
^ZZ^^Z^^Z^^Z&Z%^Z^Z^ZZ^Z^Z^Z$
V7//////////A
gg-V\A-j
7777/?//////////,
Fig 4.3.2-3 Macro Models by Linde (1993)
The formulas of the wall model were derived by Linde (1993) from a typical
cantilever wall behavior, including elastic flexural behavior, nonlinear flexural
behavior and elastic shear behavior. Different material models for the springs in
164
relevant suggested macro models will be derived and applied to capture the
relationship between lateral force and displacement of a whole precast wall in
subsequent sections.
Elastic Flexural Behavior
To derive the formulas, as shown in Fig. 4.3.2-4 and Fig 4.3.2-5, the
kinematic relations of a simple real wall is compared with those its model. A real
cantilever wall shown in Fig. 4.3.2-4a obeys the basic elastic theory on beams.
Under pure flexure the model in Fig. 4.3.2-4b, would have to simulate the uniform
curvature.
V//////////////////////
] /
-4 \-
a) Elastic Theory b) Model
Fig. 4.3.2-4 Wall Rotation for Uniform Moment by Linde (1993)
From simple beam theory, as shown in Fig. 4.3.2-4a, following formula can be
derived.
Mh
e EJ (4.3.2-1)
Mh2
(4.3.2-2)
2EI
Where Ms height of the wall, / is width,, E is Young's modulus, / is moment inertia.
And for Fig. 4.3.2-4b with F = — , d = — , and 8V = -&- = J^L, we get
P B
I I AE IAE sp
2Mh
0 = ^l E (4.3.2-3)
iK
165
Mhhc (4.3.2-4)
& = eh, =
Where Asp is the spring area, hc is distance from rotation center to the top of the wall,
By setting (4.3.2-1) equal to (4.3.2-3), and (4.3.2-2) equal to (4.3.2-4),

model properties is gotten to represent the outer spring area and relative center of
wall rotation, respectively.
AP - /2
(4.3.2-5)
K-h- (4.3.2-6)
c
2
The elastic stiffness of the outer spring is given by
sp.e h
(4.3.2-7)
/ /
a) Elastic Theory b) Model
Fig. 4.3.2-5 Wall Rotation with Moment Gradient by Linde (1993)
Normally in shear wall structures, the lateral force acts on each story and
causes bending moment gradient along the wall as shown in Fig. 4.3.2-5. As done
above, the properties of model were derived through the equation 4.3.2-5 to 4.3.2-
11.
For real elastic wall shown in Fig. 4.3.2-5a,
Vh (4.3.2-8)
2EI
166
8h (4.3.2-9)
3EI
For macro model shown in'.Fig. 4.3.2-5b
Vh
F
sP (4.3.2-19)
i 2/
F Vh2l
sv _ sPh . 4EI (4.3.2-11)
KE • Vh2
(4.3.2-12)
2EI
i 3
Vh
sh = 9hc = 4EI
(4.3.2-13)
The deflection given by elastic theory (equation 4.3.2-9) differs from that
from model. This deficiency can be reduced by discretization using several
elements verified by Linde (1988).
And if keep h( as parameter in the formula, we can get,
F h Vhh
«y=-E- =-l^- (4.3.2-14)
APE lAspE
e = 2_s}L=mA (432.15)
2Vhh2
S=eh=^^ (4.3.2-16)
' ° I2APE
Through equating (4.3.2-8) to (4.3.2-15) and (4.3.2-9) to (4.3.2-16), we can get,
Asp=-i1L (4.3.2-17)
3/2
hc=Y (4.3.2-18)
Comparing equation (4.3.2-18) with (4.3.2-6), the center of rotation under

only shear force is lower than that in wall under pure bending moment. General
case is that at a wall section bending moment and shear force exist simultaneously.
To cater for the real wall, in this research, a combination of the two situations is
applied according to the proportion of the moment and shear force.
Elastic Shear Behavior
167
A horizontal spring is used to simulate the shear behavior as shown in Fig.

4.3.2-6. From simple elastic theory, the horizontal deflection caused by the shear is
K-Vh
She =
(4.3.2-19)
0.75(1 + u)[\-u2(l-v)]
K-
l-V(l-v)
(4.3.2-20)
rcis a empirical factor suggested Tommii and Osaki (1955) and definition of factors
u and vis shown in Fig. 4.3.2-7.
*$*-*-
V
AspE
Aw.G
W/////////////////////
—' f- b) Model
a) Elastic Theory
Fig. 4.3.2-6 Wall Shear Behavior by Linde (1993)
he
A, t=vb
/'=»/
l=(l+u)L/2
Fig. 4.3.2-7 Wall Shear Behavior by Linde (1993)
168
B. Suggested Model for Procast Walls
This study is intended to develop simple and clear formulations of the macro
model for precast shear walls without vertical load. Because no axial load was
applied in this study and purpose of development of the model is to predict the
relationship between lateral load and displacement, the property of the vertical
spring at center is not a critical factor in this research and has negligible influence
on the accuracy of prediction. The suggested model is shown in Fig. 4.3.2-8, with
two vertical outer springs, one dummy vertical spring and one horizontal spring.
The idea behind the arrangement is to achieve simplicity by using as few springs as
possible. Because the beams connecting the nonlinear springs are flexurally rigid,
the kinematic possibilities are essentially the same as for models with more
complicated spring configurations, and thereby this model is able to provide an
efficient result, with the nonlinear behavior derived properly.
Fig 4.3.2-8 Suggested Macro Models
Because of the totally different features between the horizontal connection

and RC wall panels in the precast walls, the whole precast wall will be simulated by
an assembly of the macro models as shown in Fig 4.3.2-9. The suggested model of
precast wall consists of three components, one macro model of connection (MMC)
and two macro model of wall (MMW). The concept is listed out as following.
In precast walls the horizontal connections were composed of two boundary

steel sections and one or two shear connectors at center of the wall web. At the
169
horizontal connection, the boundary steel members connected by bolts form flexural
a moment resistant system as those in macro model while the shear connectors at
center of the horizontal works the shear spring in the macro model. These features
are consistent with the configuration of the macros model.
For the upper and lower wall panels with concentrated steel at boundaries the major
configurations of the macro model were satisfied.
\ \ ^^SIZSMISZSI^SM^SZIS^SSMSSZ.
\
hi
hi
\
\
^&!2ZE!EMmmW2MZBZ2^BM8Zwfa
Fig 4.3.2-9 Suggested Macro Models for Precast Shear Walls W2,3,5,6
a) Material Models for MMC
Material model for outer springs
At the horizontal connection level, moment and shear force exist

simultaneously, as shown in Fig. 4.3.2-10, which caused rotations as shown in Fig.
4.3.2-11. Because of the existence of the horizontal connection, the tension and
compression forces caused by moment worked like tension out and push in action
170
of rebars. To simplify the bond stress distribution, an equivalent bond stress

distribution is applied in this research, as shown in Fig. 4.3.2-12. An equivalent of
rebar is utilized with same sectional area and perimeter.
For example, in the horizontal connection of W2, two channel were used
with total sectional area 856x2 = 1712mm2 and perimeter
(76 + 38x2)x 2 x 2 = 608mm
The equivalent reinforcement is 17.2 nos diameter 11.26mm rebars as shown Fig.
4.3.2-13.
Bond stress of plain reinforcement has studied by Feldman and Bartlett (2005, 2007)
and found that The average bond strength was 0.98 MPa for as-received bars, and
increased by 124% to 2.2 MPa for bars sandblasted to simulate realistic surface
roughness. Mosley (1999) proposed for plain reinforcement is 0.28^//^ for tension
and 0.35^ff^ for compression, which will be applied in this study.
M _v\ (4.3.2-21)
I
_M
F
sP (4.3.2-22)
I
f (4.3.2-23)
J sp
1
d ~
fj (4.3.2-24)
4uh
Where Fsp is force in outer spring, w* is the bond strength, fsp is stress in spring, /
action length of bond.
When spring in elastic stage,
*,,=2<I V .>4<, (4.3,-25,

2 ' E
When spring goes into plastic stage,
*v, =n\eje+ejp + i ( ^ -ev)lp] (4.3.2-26)
Where sând ^ s t r a i n in yielding and plastic, le is elastic action length of bond, lp
is plastic action length of bond.
171
Because the connection is bolt connection, the tolerance, t,, in between bolts and
bolt holes should be included.
(4.3.2-27)
111
The horizontal displacement at top beam caused by the rotation of the horizontal
connection is
(4.3.2-28)
M
Dummy
spring
i.^v
%
m^v^
Fig 4.3.2-10 Load Status in the Horizontal Connection
.Si,
I r///;ss;->/// >rrrr—\
ASPE
>y////»//»/////»?7>
Fig 4.3.2-11 Typical Deformation of the Horizontal Connection
172
a) Example of Bond Stress

Bond Stress
Bond Stress i
distribution
Strain Equivalent
distribution
b) Strain Distribution c) Assumed Bond Stress
Fig 4.3.2-12 Equivalent Stress Distribution
A V
T V C A
a) Actual Section b) Converted Bar c) Tension Sate d) Compression State
Fig 4.3.2-13 Material Model for Outer Springs
Material model for shear springs
As shown in the Fig. 4.3.2-10, the shear force is applied at center of the
model. To determine the horizontal displacement caused by shear force is to
determine the property of the shear connectors in the horizontal connection. The
connectors were mounted to the wall panels by embedded reinforcement. From Fig.
4.3.2-14, the shear force is resisted by the tension and compression reinforcement.
Here use 0.5-Jf^ for reinforcement in tension and 0.63y/^ for reinforcement
under compression as the working bond stress, suggested by Mosley (1999).
173
From Fig. 4.3.2-14, we can get

V = (T + C)cosa (4.3.2-29)
C 0.63
= 1.26 (4.3.2-30)
0.5
T= (4.3.2-31)
2.26cos«
/,=' (4.3.2-32)
^d,sp
M (4.3.2-33)
4uh
Where a is the inclined angle of diagonal reinforcement, fd is stress in diagonal
reinforcement, U,sp is the action length of diagonal reinforcement.
When spring in elastic stage,
(4.3.2-34)
2 • IE
When spring goes into plastic stage,
8i4 = 2 [ - e v l , + sjp + -{ssp -s sy )l p ] (4.3.2-35)
1 f
#Hc2=2h£Jd)c°sa=-fldcosa + tl (4.3.2-36)
2 E
Fig 4.3.2-14 Forces in Diagonal Reinforcement in the Shear Connection
Therefore, the total lateral deflection caused by deformation of the horizontal

connection is
*fc = * w + * « (4.3.2-37)
Where 8m is the horizontal deformation by moment, Shc] Shc] is the horizontal
deformation by shear and tolerance
174
b) Nonlinear Material Models for MMW
From Fig. 4.3.2-9, the load on the MMW can be gotten as shown in Fig. 4.3.2-15.
Ej^^^^^^^^^w^^^^^^^^^Z
<?////////////
'/,
vzzz y
^//////////////////////////^^//////////////////////////////Â
Fig 4.3.2-15 Loads on MMW
Material model for outer springs
The flexural behavior of the MMW, including deformation by moment and

by shear, is shown in Fig 4.3.2-16. To determine of the model properties beyond the
linear elastic region it is necessary to study the physical behavior of the cross
section of a real wall. The steel reinforcement and embedded steel sections and
concrete worked following their own material constitutive laws, respectively. In
present research, an idealized trilinear material model shown in Fig 4.3.2-17a is
applied for steel. And the concrete model by Kent and Park (1971) as shown Fig
4.3.2-18 is used.
2E \
fc=fc for * < 0.002 (4.3.2-38)
0.002 I 0.002
fc = /c'[l - Z(sc - 0.002)] for sc > 0.002 (4.3.2-39)

Where
0.5
Z = (4.3.2-40)
£
ôu- 50h -0.002
175
3 + 0.002/;
(4.3.2-41)
/J-1000
£
soh = ° - 7 5 A (4.3.2-42)
Where psis the confinement ratio, bh is the width of the hoops, and shis the spacing
of the hoops, and / c is the concrete cylinder strength in psi (1 psi=0.00689 N/mm2),
e50u and e50h repsents the strain at 50% strength on the descending branches of the
unconfmed and confined concrete.
In present study, the stage before cracking of concrete is not considered,

because this will not affect the service limit state of the common precast shear walls.
The moment for cracking can be gotten in RC design codes, and then the cracking
lateral force can be derived through dividing the cracking moment by lever arm.
a) Deformation by Moment b) Deformation by Shear
Fig 4.3.2-16 Flexural Behavior of MMW
176
600
500 • — "Z^"^
| 400 •
8 \
1 300 - — Test EAs E<At
— Idealised, used for model
200 •
~
100 •
|
0 •
10000 20000 30000 40000 50000 60000 70000 80000 90000

Strain (pc)
a) Trilinear Steel Material Model b) Model for Elastic Stiffness
Fig 4.3.2-17 Property of Flexural Springs
O.Sfc' —
0.2fc' •
4 /, 0.002 0.005 0.012
Fig 4.3.2-18 Concrete Material Model by Kent & Park (1971)
To simplify the model, a simple calculation model for the elastic stiffness in
compression is suggested as shown in Fig 4.3.2-17b and the concrete under
compression follow the material model in Fig 4.3.2-18. When the spring in tension,
the concrete lose its strength after flexural cracked appears, only steel takes effect in
the model without considering the tension stiffening effect of surrounding concrete
between cracks.
177
Material model for shear springs
As observed in the experiment, the major shear deformation occurred in the

lower wall panel. After diagonal cracking appeared, the lower panel started to
experience plastic shear deformation. To simplify the model, an empirically-based
ratio a, of cracked to uncracked shear stiffness Wight (1984) is applied in this
research.
as = 0A6p»+0,14 (4.3.2-43)
Jc
Where pwis the horizontal reinforcement ratio. The parametere^, normally ranging
from 0.15 to 0.20, was derived for wall cross sections of different geometry ranging
from rectangular wall to wall with pronounced protruding boundary elements.
Therefore the equation (4.3.2-19) to be
= J
^ 77" (4.3.2-44)
C. Results Comparison
Because the cracking moment is not a critical control in service limit state, and
also to simplify the prediction, present prediction started from lateral load higher
than cracking load onwards. In the course of simulation, a simple Excel sheet is
written to facilitate the prediction, which incorporates structural models of MMC
and MMW, material models for steel and concrete. Random loading increments
were input into the calculation sheet and output was drawn out versus experimental
data as shown in Fig. 4.3.2-19 for precast walls W2 and W3, in Fig. 4.3.2-20 for
precast walls W5 and W6. The comparisons show reasonable good agreement with
experimental measurements in term of yielding strength, yielding displacement and
ultimate strength. Accurately capture the yielding strength and displacement means
the model accurately predicted the secant stiffness when structure reached yielding
point, which is very important in design application. Good prediction on ultimate
load indicates that the suggested model can be used when checking structure design
redundancy for the precast walls with horizontal connection. This verified that the
178
suggested simple model could be applied in actual design of precast wall with
horizontal connections.
20 30 40 SO 60
Displacement (mm)
•— Model
W2
• W3
-700
Fig 4.3.2-19 Comparison of Prediction vs. Experimental Data of W2,3
1000 -
r y*p(-)
1 800
1 1
J 600 - £>^
Specimen W5,6
I
NS
400 -
200 -
(/
60 -SO -40 -30 -20 -10 /A 10 20 30 40 SO 60

-200- Displacement (mm)
Jfioo -
•t ws
-600 - W6
\. -- —*• —»—Model
«— •800 -
-1000 -
Fig 4.3.2-20 Comparison of Prediction vs. Experimental Data of W5,6
179
D. Parametric Study
In course of experimental data analysis and simulation by the suggested model,

it is interesting to compare result under different parameters, such as, bolt type and
location of the horizontal connection. In present study, tolerance between bolt and
bolt hole is 1 mm, which affected the behavior of the connection and thus
influences the performance of the whole precast wall, including yield displacement,
stiffness, ultimate displacement. High strength friction bolt is alternative to control
the behavior of the connection, in which no tolerance is allowed. Location of the
horizontal connection is another factor may affect the whole behavior of the wall,
for example, the connection at bottom of the wall.
It can be seen from Fig. 4.3.2-21, the model with high strength friction bolts
shows a reduction in the yielding displacement with the same yielding strength,
which means higher stiffness for precast wall with high strength friction bolts.
To study effect of the location of the horizontal connection on the behavior of

the precast wall, a typical model is shown Fig. 4.3.2-22. In the model only one
horizontal connection is placed at bottom of the wall. The prediction result on this
model is compared with experimental data W2 and W3 as shown in Fig. 4.3.2-23.
The model shows lower stiffness, higher yielding displacement and lower ultimate
strength. One reason is the less cross section area because when connection at
bottom, no reinforcement surrounding the steel channels is added, which means less
force can be taken by the spring in MMC. Tolerance in the connection causes
additional rotation for the connection, and the additional horizontal displacement
equals to this rotation times whole height of the wall. This is another reason.
As a results parametric study on the type of bolt and location of the connection
the followings is recommended.
1. High strength friction bolts or weld is preferred in precast connection, which
not only reduces the displacement from flexural deformation but also
prevent the sliding along the connection due to tolerance.
180
2. Stronger steel section or section with bigger cross sectional area should be
used in the connection to cater for the decrease in the steel section area
where no reinforcement can be placed.
SftOO
-500
-600
-700
Fig 4.3.2-21 Comparison of Prediction with HSF Bolts vs. Experimental Data of W2,3
181
Fig 4.3.2-22 Parametric Study Model for Connection at Bottom
10 20 30 40 50 60
Displacement (mm)
Model-Connection at bot
W2
Wi
Fig 4.3.2-23 Comparison of Prediction with Connection at Bottom

vs. Experimental Data of W2,3
182
E. Conclusions
1. In this section, a simple model based on analysis on the features of the

precast walls, which is an assembly of macro models, including macro
model of connection (MMC) and macro model of wall (MMW), was
suggested with corresponding constitutive laws.
2. The suggested model did a reasonable good prediction for precast walls on
load-displacement relationship, including yielding displacement and
ultimate load.
3. A parametric study was done on the type of bolt and location of the
horizontal connection. High strength friction bolts improved the stiffness of
the precast wall and thus reduce the structure deformation. The location of
horizontal connection also affected the behavior of the precast walls and
increased the lateral deformation. Stronger section or section with bigger
area should be used in the connection.
4.3.3 Curvature Ductility
There are various types of ductility which influence behavior. Herein, only
strain ductility proposed by Tasnimi (2000) and Chen (2005) is used to evaluate the
behavior of specimens. Therefore assuming linear strain distribution across the
section under consideration and perfect bond between concrete and reinforcement
the value of ductility in all post-yield states could be calculated. Curvature at any
post-yield state is the ratio of concrete compressive strain at any post-yield state to
depth of neutral axis at corresponded state. On the other hand, curvature at the yield
state is the ratio of yield strain of steel reinforcement to the depth of concrete in
tension.
4_iWz£> (4.3.3-1)
d> sc
183
Where p is curvature ductility, <f> post-yield curvature, <j> yield curvature, d length
of lever arm, c depth of neutral axis, en is yielding strain of steel reinforcement in
tension.
It is assumed that A at the top of wall is corresponded to the first yield of the
section and the displacement at ultimate state Au corresponded to the ultimate load.
The prediction of ultimate curvature <f>H was carried out using the ACI method and
value based on text results.
Assuming the strain of first yield of reinforcement in the tension side of the
wall section, and concrete compressive strain at extreme equal to 0.003, the
ductility of the section, according to the reinforcement concrete design described in
ACI 318-05, can be evaluated by:
J fiMzR (4.3.3-2)
Where /?, shall be taken as 0.85 for concrete strength up to and including 30MPa.
For concrete strengths above 30MPa, /?, shall be reduced continuously at a rate of
0.008 for each IMPa of strength in excess of 30MPa. However, /?, shall not be
taken less than 0.65 (ACI 318-05)
The value of k is given as:
k = (p + p')2m2+2(p + p'—)m (p + p')m (4.3.3-3)

a
a=^tLl >ihL (4.3.3-4)

0.856/"
Where en and e are the concrete compressive strain at ultimate state and the yield
strain of reinforcement. The modular ratio is represented by m and b is the width of
the wall, p and p are the tensile and compressive steel percentage, respectively.
184
Compression and tension forces at yield state are represented by A f emdATkfri, d'
and d are distance of center of reinforcement in compression and tension to extreme
compression fiber.
Table 4.3.3-1 Comparison Curvature Ductility Based On ACI And Experimental Result
Calculate curvature ductility based on test
Ductility
results Ratio of calculated
Specimen on ACI
ductility to ACI
<j)r 10'6 <f> 10"6 Mt
P*
6.44 0.84
Wl 6.67 36 5.4
(8.96)* (0.6)*
6.45 0.95
W4 8.03 49.4 6.15
(8.42)* (0.73)*
* Represents curvature ductility basing on code-specified nominal strength of steel and design
strength of concrete.
Curvature of ductility, which indirectly represents flexural deformation of

specimen, is calculated basing on rotation of plastic hinge zone in monolithic
reinforcement concrete shear walls. From table, it is shown that for monolithic walls,
Wl and W4, curvature ductility calculated basing on tested materials exhibits high
agreement. This agrees with curvature figures shown in chapter 4.2.7, in which
main curvature focuses in lower XA height of walls and curvature in upper 3A height
is negligible. However, curvatures in precast walls, W2, W3 and W5 are different
from those in monolithic walls. It can be seen from table that curvature ductility
based on code-specified nominal strength of steel and design strength of concrete is
higher than that based on tested material and that calculated from experimental
results. It hints that yield strength of flexural reinforcement exhibits reverse effect
on curvature ductility of walls.
In order to evaluate curvature occurred at the horizontal connection during

load course, the yielding curvature in the connection is calculated when
reinforcement at bottom of the walls yield, and the ultimate curvature is based on
the ultimate displacement of walls. Results from the suggested macro model and
185
experiment are compared in Table 4.3.3-2. The experimental data shows higher
rotation than the model prediction.
4.4 summary
In this chapter, experimental results have been analyzed to qualify and quantify
seismic performance of these specimens. The author also suggested a macro model
which is an assembly of the macros model with different material models to predict
the lateral-displacement relations for those precast walls.
1. Monolithic prototypes, Wl and W4, exhibit higher stable in plastic stage

than those precast walls, W2, W3, W5 and W6 do. In monolithic walls, ratio
of flexural deformation to total horizontal displacement is higher than that in
precast walls.
2. Pinching effect is severer in precast walls due to higher shear deformation to

total deformation. Through trends of ratios of flexural and shear deformation
to total deformation, failure mode can be accurately predicted.
3. Average energy dissipation shows that the monolithic prototypes experience

higher energy dissipation than their precast counterparts before ductility
reach 2.
4. A macro model is suggested by the author, which produces a reasonable

good prediction on the lateral load-displacement relation for precast walls,
including stiffness, yielding displacement and ultimate strength.
5. Parametric study based on the suggested macro model has been done,
including the type of bolt and location of the connection. High strength
friction is preferred for application connection with bolts. This means
similar stiffness as that of monolithic wall can be achieve with high strength
friction bolts. And stronger section or section with bigger section area is
186
should be used to ensure strength and decrease

deformation.
187
experiment
al ductility
theoretical
Ratio of
1.27
1.09
1.15
to
co 1—4 ^ H
1.9
2.0
t-;
=c
,
1
H a\
• ^."C
*—4
s 2
s
vd <S a
o
vo
&
ID
'
<u
u
^~* t~-
-.^ • * CO
s
OS
"•• 2 co
*^»
U _"** «~i i/i >n
rimental
=*. i—i i—i

r—<
.€" r^ r^
_
*et 2 • * • * o
Is)
2.17 1
caland
^ 2 co CO
v•V
u o O o
©
4) ^ u sa Ov
co
Ov
CO
r—4
ov
CO
»—« 4—4
ff
<M §
retica calculati
e o- 00 oo oo
s Ov Ov Ov
o
« y U"l 1/-I
**"** , 9 >/-> 1—1
ris
*—4 »—4
-'
«
CL r—< oo OO o
499.8
567.4
499.8
Com
ar§
<s
I *8
co
o
o
©'
8
o
o
CO
©
©
<u
S
'i? ©
CN
©
CN
©
w a •—4 i—i •*
I—1
•w
•n 1—;
CO CO
460.
460.
177279
b3"l
CN CN
5- t—4
Ov
1—1
y-^s 00 OO 00
*-^v S ô vb OV
CO
^ 1' 5 >w 4—4

5 o
CM
rt
.§sa CN CO W)
8
A
£ £ £
00
188
Chapter 5 Discussion of Test Results
CHAPTER 5
DISCUSSION OF TEST RESULTS
Strut-and-tie models have become increasingly popular tools for the design of
reinforced concrete structures. They can be especially useful in regions of
discontinuity. As a valuable tool to understand flow of internal load forces,
strut-and-tie models are used to analyze behavior of these walls in this chapter.
5.1 Behavior of Specimen Wl
Specimen W1 is a monolithic wall with embedded channels at two ends acting

as prototype of specimen W2 and W3. This wall finally exhibited flexural failure
with concrete crushing at right lower corner and breaking of weld of horizontal
reinforcement. Characteristics of behavior of specimen Wl were revealed by the
strain gauges mounted on the steel reinforcing bars and embedded steel sections and
cracking patterns. Strut-and-tie models are built to give details of performance. As
the results will verify, the model is a valid tool to display the behavior of the
specimen.
5.1.1 Strain History of the Gauges
A. Strain History of Vertical Bars
Strain histories of vertical bars on both sides of wall are shown in Fig. 5.1.1-1 to
Fig.5.1.1-10. Vertical bars at ends of the walls were broken during the loading
course with sound which can be seen from the strain histories of these vertical bars.
The vertical bars with three strain gauges are selected to illustrate the main
characteristics. And yield strain is also drawn in these figures and is used as
189
standard to evaluate states of these bars.

a. In elastic stages, the gauges recorded negative values. However, in
plastic stages, no negative readings are found except some broken the
gauges. This is due to the permanent deformations at the locations of the
gauges which came from cracks. The cracks crossing these bars open
when in tension and plastic deformation occurred in these bars. When
load reversed, the cracks could not close perfectly.
b. All the vertical reinforcement yielded during test which can be seen
from the strain recorded by the strain gauge named SVL series when
lateral load reach peak values in plastic stages.
c. Strain histories of SVM series at peak loads show plastic deformation of
the bars at this height level. It indicates the plastic deformation of these
bar extend to half height of the wall except 4 bars, such as SVM5,
SVM6, NVM5 and NVM6, in middle of the wall. This means that there
is an elastic core at SVM height level.
d. Strain histories SVH series located at top of wall indicate that no plastic
deformation occurred in any vertical bar at this height level since the
reading of strain at peak loads is in elastic range.
190
-SVL1
SVM1
SVH1
II fi
i
1 \J\Ar\f*
-2000
w\A/W
-4000
-6000
100 200 300 400 500 600 700 800 900
Scan Number
Fig. 5.1.1-1 Strain History of Vertical Bar SV1
8000
100 200 300 400 500 600 700 800 900

Scan Number
Fig. 5.1.1-2 Strain History of Vertical Bar NV2
191
-2000 -
-3000
0 100 200 300 400 500 600 700 800 900
Scan Number
100 200 300 400 500 600 700 800 900

Scan Number
192
-4000
0 100 200 300 400 500 600 700 800 900
Scan Number
14000
12000
10000
8000
J] 6000
•|
3
4000
2000
WfA
0 .-sSfc
-2000
-4000
-6000
0 100 200 300 400 500 600 700 800 900
Scan Number
B. Strain Profile of Vertical Bar at Base Line
The strain profiles of vertical bars along the baseline when load is at peak
values, positive and negative, are shown in Fig.5.1.1-7 to Fig.5.1.1-8. The strain
profiles on N-side in different cycles are presented. It can be seen from these figures
that in early cycles the strain profiles approximately comply with the assumption of
plane section.
193
The intersections of the strain profiles and the horizontal axis are shown in
these figures, which can be approximately seen as the neutral axis. The compression
depth is consistent with the calculation in specimen design. It showed that large vertical
deformation occurred in the field a b o u t 1/3 width of the wall from the extreme tension fiber.
9000
N-side
7000 — cycle 1
— cycle 2
5000 • cycle 3
— - cycle 4
|t Wl
.5 3000
• - cycle 5
a
1000
-1000 -
-3000
0 200 400 600 800 1000 1200 1400 1600 1800 2000
DistancefromLeft Side of Wall (mm)
Fig. 5.1.1-7 Strain Profile of Vertical Bar at the Base Line of Specimen Wl for Negative Load
7000
1 N-side
6000
5000
4000 cycle 3
s
•a. 3000 - ~' -
s 1 / cycle 4
.5 2000 _ v-
2
/ \
/ \
5 1000 . • - : : :
^^=îiii i
. » J
-
• 1 1 1 1 ~ * ~ ~ " ' I"""' * ? T« • » — • ^ - ' — i
0
\
-1000
-2000
-3000
0 200 400 600 800 1000 1200 1400 1600 1800 2000
DistancefromLeft Side of Wall (mm)
Fig. 5.1.1-8 Strain Profile of Vertical Bar at the Base Line of Specimen Wl for Positive Load
C. Strain Profiles along the Vertical Bars
194
Fig.5.1.1-9 to Fig.5.1.1-14 show the strain profiles along the vertical bars in
different cycles. The strain profiles when load is at peak values, positive and
negative, are given. Strain profiles of vertical bar on S-side SV1 are presented in
Fig.5.1.1-9 and Fig.5.1.1-10 for negative and positive load, respectively. Strain
tends to decrease along height of the wall. Under negative load, the strain at top is
very small, whereas under peak positive load this strain tends to a higher value. The
bar NV4 also shows similar tendency of SV1 as shown in shown in Fig. 5.1.1-11 and
Fig.5.1.1-12.
The strain profiles along NV6, which is one of bars nearest to center line of the
wall, are shown in Fig.5.1.1-13 and Fig.5.1.1-14 in different cycles when lateral
load reach peak values, negative and positive. No negative strain value appears in
NV4 and NV6. The peak value of strain on NV6 is much less than those in other
bars. Strain of NV6 at top of wall tends to zero, which is consistent with observed
cracking patterns of this wall as shown in Fig.4.2.4-1 when ductility reaches 2.
Tension force in this vertical bar is transferred to concrete by bond; as a result, this
force is transmitted to the top beam by the concrete in tension.
195
-500
-1000
I^ -1500
-2000
-2500
0 200 400 600 800 1000 1200 1400 1600 1800
Distance from the Base (mm)
Fig.5.1.1-9 Strain Profiles of along Vertical Bar SV1 for Negative Load
5000
SV1
4000 - cycle 1
cycle 2
cycle 3
^ 3000 cycle 4
1s cycle 5
2
« 2000 -
1000
0 200 400 600 800 1000 1200 1400 1600 1800

distance from the base (mm)
Fig.5.1.1-10 Strain Profiles of along Vertical Bar SV1 for Positive Load
196
0 200 400 600 800 1000 1200 1400 1600 1800

Fig.5.1.1-11 Strain Profiles of along Vertical Bar NV4 for Negative Load
1 NV4
4000 -
S
3500 -
X
cycle 3
3000 - i cyde 4
L_ Wl
1 2500 - N
cycle 5
s N
| 2000 -
^-—-_
• - • — . - .
1500 -
* * - . s.
1000 - -"-"
. —"
500 - ----"""
- ^ ^ ^ - 1 ^ „ , . - - -
0 -
0 200 400 600 800 1000 1200 1400 1600 1800
Fig.5.1.1-12 Strain Profiles of along Vertical Bar NV4 for Positive Load
197
2700
2200 - NV&..
cycle 1
1700 - cycle 2
cycle 3
1 1200 - Wl cycle 4
CyCle 5
% 700 \
--> ;••• ? . , :. ^ . ...
-800
0 200 400 600 800 1000 1200 1400 1600 1800
Fig.5.1.1-13 Strain Profiles of along Vertical Bar NV6 for Negative Load
2500
-500
0 200 400 600 800 1000 1200 1400 1600 1800
Fig.5.1.1-14 Strain Profiles of along Vertical Bar NV6 for Positive Load
D. Strain Profile of Steel Sections
198
The strain profiles of all Channels at different height are shown in Fig.5.1.1-15
to Fig.5.1.1-22. The layout of strain gauges on steel channels is shown in
Fig.3.2.2-4. Strain gauges pasted on inner channels on S-side include series of SSI
and SS2, while the other strain gauges placed on outer channels on N-side are
divided into series of NN1 and NN2. In order to catch characteristics of strain of
channels, more strain gauges are placed on each channel than those on vertical
rebars. In the first cycle, the strain profile exhibits linear tendency from base to top,
which is consistent with the moment on each section of wall. With increasing
displacement levels, the deformation of channels near the base beam increases
much faster than upper part and the strain profiles undergo nonlinear trend. The
main deformation occurs in lowest part, which means this part experienced much
vertical elongation. These characters are similar to those of vertical rebars shown in
previous section.
As shown in Fig.5.1.1-15 and Fig.5.1.1-16, the strain profiles of SSI at the
location about 800mm from base beam from cycle 3 in negative and positive
loading courses. However, this sudden increase in strain does not occur in Channel
NN1 as shown in Fig.5.1.1-19 and Fig.5.1.1-20, which is at same end of the wall.
And similar difference happens between SS2 as shown in Fig.5.1.1-17 and
Fig.5.1.1-18, in which the sudden increase of strain locate at 1200mm from base
beam, and NN2 as shown in Fig.5.1.1-21 and Fig.5.1.1-22. All these sudden
increase in strains occurred at these location means regions may experienced a
crack near the gauge.
199
-100
-200
cycle 1
\ -300
cycle 2
cycle 3
I -400
to cycle 4
-500 cycle 5
-600 SSI
-700
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400
Fig. 5.1.1-15 Strain Profile of Channel SSI for Negative Load
5000
4000 i \ =*c SSI

I \
3000
cycle 1
cycle 2
1
.5 2000 -
Wl
cycle 3
I cycle 4
----- cycle 5
1000
-1000
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400
Fig. 5.1.1-16 Strain Profile of Channel SSI for Positive Load
200
5000
4500 1 15=
, SS2
4000 - / \
/ \
3500 - \
_ 3000 s cycie t
Wl ; \ cycle 3
I / .'
s\ N
•
cycle 4
.5 2500 s \ \
2 - ' .... N v cyc/e5
% 2000 • ' ••' '• \ \
1500 _.-''"" -.-•'/,••• X \ \

*, \ "--- -,
1000
1
HI-
500
I i i ? i r r~~" r —
0
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400
Fig. 5.1.1-17 Strain Profile of Channel SS2 for Negative Load
5000
=*[
4000 — cycle 1
— cycle 2
3000 - • • cycle 3
- cycle 4
Wl
I - cycle 5
.s 2000
2 SS2
1000
-1000
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400
Vertical Distancefromthe Base (mm)
Fig. 5.1.1-18 Strain Profile of Channel SS2 for Positive Load
201
400
200
0 H 1-
!? -200
•5 cycle 1
| -400 cycle 2
cycle 3
-600
cycle 4
Wl cycle 5
-800
NN1
-1000
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400
Fig. 5.1.1-19 Strain Profile of Channel NN1 for Negative Load
2500
2000
1500
1000
500 -
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400
Vertical Distancefromthe Base (mm)
Fig. 5.1.1-20 Strain Profile of Channel NN1 for Positive Load
202
3000
NN2
2500
2000
11500
| 1000
500
-500
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400
Fig. 5.1.1-21 Strain Profile of Channel N N 2 for Negative L o a d
1000
-1000 •
•=>[
- cycle 1
\ -2000 - - cycle 2
cycle 3
-3000
- cycle 4
Wl • cycle 5
-4000
NN2
-5000 -
-6000
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400
Fig. 5.1.1-22 Strain Profile of Channel NN2 for Positive Load
E. Strain History of Diagonal Bars
The histories of strain in diagonal bars near the base are shown in Fig.5.1.1-23
to Fig.5.1.1-26. From Fig.5.1.1-23 and Fig.5.1.1-24, when lateral load is negative,
203
there is big jump in strain of ND1 and ND2. This means that diagonal cracks
opened which cross the strain gauges mounted on the diagonal reinforcement. This
is consistent with experimental observation. From reading of strain of strain gauges
on diagonal bars, both positive and negative values are experienced in the cyclic
load course, which means that whether in compression or in tension, diagonal bars
play their roles in resisting lateral load.
As shown in Fig.5.1.1-23 to Fig.5.1.1-26, after ductility reaches 4, the increase

of maximum strain in all diagonal reinforcement in tension when lateral load
reaches peak value at every displacement level, positive and negative, is not
considerable while the total displacement of the specimen increases, which means
that there is not much increment of strain due to plastic deformation and diagonal
rebars did not reach their ultimate strength. There were cracks passing through or
near the gauge. This could be seen from apparent strain increment, from about
500/J£ to 1000 /us . The final maximum strain was about 3500 fie , which was in
the measure range and did fail the gauge. This indicated that the cracks were not so
wide, which was in agreement with the observed cracking patterns. And it is also
that these bars yield when they in tension. The contribution to lateral load can be
evaluated.
204
600
500
400
300
200
1\ I y
*<s
is 100
x
\ / ZZ
e T
0 1 a
I -100
ND1
-200 llVi v\V^^\§^
-300
-400
-500
-600
-2000 -1000 0 1000 2000 3000 4000 5000
Strain(/ts)
Fig.5.1.1-23 History of Strain in Diagonal Bar ND1
600
500
400
300
200
^ ^£
I 100 **> ,V /
'x
0
-J
-100
sm
a -200
-4
vAiW ^ ^ \ ^ & ? ^ w ^ ^
-300
-400
-500
-600
-1000 0 1000 2000 3000 4000 5000
Strain(fie)
Fig.5.1.1-24 History of Strain in Diagonal Bar SD2
205
600 P
500
400
300
200
100
i -100
0
\4W\ — L
N
^
\ ^J
\W -
-3
J>
y.
£*.*-!—1-
*•> -200 y SD3
-300
-400
-500
-600
-1000 0 1000 2000 3000 4000 5000 6000 7000
StrainQis)
600
=
500
400 J /jl/u(fl/lfj /
300
200
8 ~<
h '/
N //
is 100 ^
a \ tf
5 0 'VM-
I
55
-100
SD4
"-" -200
-300 /Mr
-400
-500
-600
11
-2000 2000 4000 6000 8000
Strainfrs)
206
F. Strain History of Horizontal Bars
Strain histories of all horizontal reinforcement are shown in Fig.5.1.1-27 to

Fig.5.1.1-38 from top to base in sequence.
1. During elastic cycles 1 to 3 the strain is negligible. From cycle 4 to ductility
2, considerable increase can be observed in reading of these strain gauges.
2. It can be seen in these figures that a pronounced increase in strain with less
increase in lateral load occurred when a diagonal crack passing through the
strain gauges.
3. In most cases during plastic cycles, the permanent residual strain after
reversed load applied was about 500 /JS to 800 fj,s for most horizontal strain
gauges. This means that these diagonal cracks may not have closed
completely, even though load reversed.
4. The characteristics of distribution of strain of horizontal reinforcement along
height also are obtained from these figures. From NH2 to SH5, the
maximum strain at peak load is about 1500//£, stain of NH6 is about
1900//£ . SH7 and NH8 experience higher strain more than 2700 jus at peak
load, which indicates that they yielded under peak load. In the region with
diagonal bars, the maximum strain decreased at peak load from top end of
diagonal bars to base line, which means that in the lower part of the wall, the
diagonal reinforcement partook of lateral load and decreased strain in
horizontal reinforcement. From peak readings of these strain gauges the
contribution of the horizontal reinforcement to shear can be evaluated.
207
600
.
400 >>
200
-200 SHI •2
-400
-600
-50 -40 -30 -20 -10 0 10 20 30 40 50
Strain (fis)
Fig.5.1.1-27 Strain History of Horizontal Bar SHI
600 j 1 i-^fcia
K ^^^ - AI --
400
r P
1 ^^^^^
^3%z^
_ 200 ^*€s^
^^^ t-
X-
^£^ X-
6)
1 î&r 4-
w \
_^2&
2
>3 -200
m NH2-2
-400
-600
-200 0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200
Strainffjr.)
Fig.5.1.1-28 Strain History of Horizontal Bar HN2
208
600
400
200
6
<1
-3 -200
-400
-600
-500 0 500 1000 1500 2000 2500
Strain(fts)
Fig.5.1.1-29 Strain History of Horizontal Bar SH3
600
it—I—I—I—r
400
•—Y y
rsfs^~~~ JpWV
/£&/
^J&&L/^
14 --
I
200 "
a? / /^z?^'
/ ^£2^/ t-
t X-
1M^^
CWK^^-
4-\
•3 -200 AM NH4-2
-400 -
-600
-500 500 1000 1500 2000 2500
Strain(fis)
Fig.5.1.1-30 Strain History of Horizontal Bar NH4
209
600
\lf~~~ la :::::::
!fl 1/I
400 J^
yjf
In! y^f 4-
_ 200
i 1/ s^r
I ^^^J&' >-
j l T ^ ^ ^ ' '
0
5
S -200 SH
-400
C \ \\l N\N^ltt\
-600
-500 500 1000 1500 2000 2500
Strain(fte)
600
400
1
_ 200
I
4 °
-3 -200
-400
-600
-500 0 500 1000 1500 2000 2500 3000
Strain (fie)
210
600
400
— I
:
200
[
a
I
-200 SH r-2
11 ^ ! ^ W ^ N \
-400
-600
-500 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500
Strain(fie)
600
400 c
- ^ J ^ \ / < ~ ^ /
^ 200 - -y yy//&vZ^JxfZ'1^^^^^ f
// /^XtE/Z^^^^^C-^^ 1
// ^^Za&Zi^S^^^^
I • , is ^âî?- *>* > ' ' j
g
3 -200 MIL
li ^ ^ ^ ^ ^ \
i
NH8 3
-
-400
-600
-500 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000
Strain(fts)
211
600
400
200
.
= =111
<i -200 SH9-2
-400
-600
•500 0 500 1000 1500 2000 2500 3000
Strain(fie)
600
400
_ 200
a
5 0 Wi i <&ft%fflj^S§Ys*, '
•3 -200
^^~^k N ^ ^ ^ ^ ^ ^ ^ ^ ^ N X NHTO-3
-400
-600
-500 0 500 1000 1500 2000 2500
Strain(fie)
212
600
400 -
1
-. 200
1
Xd ,
,5 o 1
' 1
^1V'N & \ \ \ ^. JW/W
-400 f\
-600
-500 0 500 1000 1500 2000
Strain(fie)
600
400
_ 200
•3 -200
-400
-600
-7500 -1000 -500 0 500 1000 1500
Strainffts)
5.1.2 Load Transfer
213
The mechanism of shear transfer consists of aggregate interlock along crack

interface, dowel action of chord reinforcement, shear transfer by concrete in the
flexural compression zones and tensile strength of uncracked concrete.
As shown in Fig.4.2.4-1, there was a wide horizontal crack passing across the
base line of wall, which appeared at DF equal to 5. In late plastic cycles, if in
compression, the crack which previously opened in tension almost closed. In the
loading course, sliding along the interfaces occurred. The friction along the
contacted interfaces took part in resisting lateral load. If this horizontal crack was
tension, no aggregate interlock effect could be expected. The reason is this crack is
flexural crack not a shear crack; its width reached about 20mm at outmost end of
the wall if in tension.
As shown in Fig. 5.1.1-1 to Fig. 5.1.1-6, though all the vertical reinforcement
near the interface of the wall and foundation yielded, dowel action was not observed
because no relatively large sliding along the horizontal crack occurred.
The diagonal reinforcement in tension yield at peak load as shown in

Fig.5.1.1-23 to Fig.5.1.1-26. So the contribution of diagonal reinforcement in
tension to the shear strength of the wall can be easily evaluated.
So it can be concluded that the shear force at the interface of the wall and
foundation is mainly resisted by concrete in compression and diagonal
reinforcement.
On the left lower corner a diagonal crack formed when lateral load reached
0.75 7^ as shown in Fig.5.1.2-1. In following loading cycles, this crack extended
into base beam and width increased gradually With increasing displacement levels
214
as shown in Fig.5.1.2-2, the diagonal cracks coming from upper corner on opposite
end passed through this crack. This means that the diagonal concrete strut was
serving as the first force resistant mechanism. However, the shear strength of the
wall did not decrease when this zone was in compression. And the special thing is
that these diagonal cracks coming from the other end intersected this diagonal crack
with an approximate 90 deg angle. When the diagonal concrete strut was in
compression, this diagonal crack almost closed completely, therefore, the strength
of the wall was not damaged. When concrete crushing occurred in the lower left
corners in following load cycles as shown in Fig.5.1.2-3, the strength of the wall did
not decline. Finally, with fracture of the outermost vertical rebars at the lower right
corner, the negative load value dropped. This phenomenon indicated that crushing
of the concrete cover in a compression zone was not a critical factor in determining
the strength of the wall if the concrete core in confinement is not damaged
seriously.
Fig.5.1.2-1 Diagonal Crack in Left Corner at 0.75 Pi
215
Fig.5.1.2-2 Diagonal Crack in Left Corner at DF=3
Fig.5.1.2-3 Diagonal Crack in Left Corner at DF=6
216
5.1.3 Crushing of Concrete and Buckling of Vertical Steel
When DF reached 6, fracture of vertical reinforcement at the lower right corner

of Wl occurred as shown in Fig.5.1.3-1, and spalling of concrete cover can be seen
from side view is serious, which means the depth of compression zone was
decreased.
In course of testing Wl, fracture of vertical bar occurred at the middle point
between horizontal link and base beam and not at the base line level where, in
theoretical calculation, the highest tension stress happened. This is due to the
buckling of the vertical bar. There are three causes for the buckling of vertical bar.
a. When load reversed, the opened flexural cracks closed gradually and the
vertical bars previously yielded in tension under compression.
b. At same time the lower right corner was under compression which
caused horizontal dilation at this location. The dilation effect pushed the
vertical bar out from the concrete. And the buckling of the vertical bar
occurred in the plane of wall, no evident out-of-plane buckling was
observed.
c. Not enough horizontal restrictions were provided at this place. At the
upper link and base beam, enough restraints were supplied; therefore,
buckling did not occur at these positions.
The strain gauge was pasted near the fracture point; the record of this strain
gauge can describe the strain history the vertical bar at this point as SVL10 shown
in Fig. 5.1.1-6.
Concrete cover contacting channel at right lower corner started to spall when
DF equal to 6 and the channel could be seen from S-side, which reduced the area of
concrete under compression. But the strength of the specimen in positive direction
did not decrease. Spalling of the concrete cover contacting channel was caused by
pushing due to buckling of flange of channel. And the channel started to buckle at
217
right corner. At first cycle of higher displacement level, DF equal to 7, concrete

cover fell off and a decrease of strength in positive direction could be seen from
hysteresis loops of Wl. In second cycle, breaking of weld of horizontal
reinforcement took place, which means that the horizontal restraint was removed.
As a result, the specimen experienced a pronounced decrease in strength.
During positive loading, concrete in the right lower corner was at a state of
multi-axial compression. Horizontal reinforcement provided confining effect on the
channel, when the horizontal bar broke, this effect was removed. As a result,
compression acted on the unconfmed concrete which was already cracked during
previously loading course. And concrete was crushed as shown in Fig.5.1.3-2 with
prominent drop in strength.
Fig.5.1.3-1 Fracture of Vertical Bar
218
5.1.4 Strut-and-Tie Model

The aspect ratio of specimen Wl including top beam is 1.41 and the aspect
ratio was 1.27 if only the wall panel was considered. According to the ACI 318-02
11.10, this kind of wall is a squat wall which frequently exhibits non-ductile
behavior with sudden loss of strength and stiffness.
A.First Stage (DF less than 2)
Strain history of diagonal bars, as shown in Fig.5.1.23 to Fig.5.1.26, indicates

that before ductility reached 2, the tension strain of diagonal bars was low. It means
that the forces in diagonal bars were low. And the horizontal crack crossing the wall
along the base line had not appeared at that stage. It means the contribution of the
diagonal bars to resist lateral load was negligible. So the diagonal bars were not
219
taken into account in the following strut-and-tie models describing the load transfer
in this stage.
In this specimen twenty vertical bars were almost evenly distributed along the
length of the wall, and four channels were placed at boundaries of the specimen
with two channels at each end. This layout of reinforcement indicated that main
vertical steels are concentrated at ends of the specimen. According to the crack
patterns shown in Fig. 4.2.4-1, the horizontal cracks, as hints of flexural
deformations, were limited only in about 300mm from extreme fibers.
After all vertical bars in tension yield, assuming the depth of compression zone
was 200mm at peak load, 16 vertical bars in tension yield, and bending moment due
to these bars is 8 x 2 x 385 x 78.5 x 990 = 479kNm.
If the vertical bars are simply divided into two parts with equal sectional area,
5x2 bars are in tension which located in boundaries of the wall, bending moment
due to these bars is 5 x 2 x 385 x 78.5 x 1600 = 4S4kNm * 479kNm
It hints that the two kinds of the distributions of vertical bars provide wall almost
equal flexural capacity.
a. Simple Strut-and-Tie Model
From previous calculation, it is found that the flexural capacity of the wall with
concentrated bars in boundaries is almost same to the wall with actual evenly
distributed bars. A simple strut-and-tie model is developed as shown in Fig.5.1.4-1,
in which the vertical bars are equally divided into two parts acting as concentrated
bars placed in the boundaries of the wall. These vertical bars work together with
channels located in same positions to resist lateral load.
220
With the assumed 200mm depth of compression zone when the all the vertical
reinforcement in tension yield, the horizontal load is:
(856x2x310 + 5x2x78.5x385)xl600/2540 = 525&V
In order to successfully transfer the lateral load from top beam to the base beam,
enough amount of horizontal reinforcement is needed.

525x1000 2300 ... .
484.5x78.5x2 « 7 and 7 = 328mm so the maximum spacing of
horizontal reinforcement is 320mm.
Vertical reaction force at compression zone is:

525*2540 = m w
1600
Herein, the contribution of steel bars and channels in compression is ignored.
Therefore all the reaction force is resisted by concrete in the compression zone.
C
o *. A * J 833x1000
So the depth of compression zone is a = —= = 197 mm
0.85/ c 6 0.85x35.5x140
C :vertical reaction force in compression zone,
fc: compressive strength of concrete cylinder;
b : thickness of wall;
d :depth of compression zone;
• First mechanism
It is assumed that the diagonal concrete strut DB goes through the entire
compression zone as shown in Fig.5.1.4-4.
The width of diagonal concrete strut:
ws = d' sin 9 = 197 xsin35.8°= 115.2mm
ws: width of diagonal concrete strut;
6: angle between diagonal concrete strut and horizontal.
From Fig.5.1.4-1, it can be seen that the shear capacity of specimen Wl is
determined by the diagonal concrete strut with 35.8 deg to horizontal.
221
= C^= 525x1.23 =40MPa = LUf'

sc
wsb 115.2x140
fsc compression stress in diagonal concrete strut;
Csc compression in diagonal concrete strut;
This stress in the diagonal strut DB shows that this mechanism is incapable to
resist the total lateral load by itself.
Maximum force in strut DB,
CDB = 0.75f'cbws =0.75x35.5xl40xll5.2 = 429.4yW
Lateral load this mechanism can resist is: P = CDB 11.23 = 429.4 /1.23 = 349kN
• Second mechanism
From the cracking patterns, an apparent diagonal concrete strut connecting load
point and opposite lower corner formed between diagonal cracks as shown in
Fig.5.1.4-2. The diagonal strut is a supplement to the first mechanism to resist
lateral force.
Width of the diagonal concrete strut: ws =d sin # = 197 x sin 57.8° =166.7mm
_ • /•("?, 525 x l . o o .~~,*r. ii^,/-1

Compression stress: f = —^- = = 42.3MPa = 1.19 f
wsb 166.7x140
The stress in the diagonal strut shows that this mechanism is unable to resist
total lateral load by itself.
The second mechanism: maximum force in strut CB:
CCB = 0.75 fibw, =0.75x35.5 x 140x166.7 = 621. AkN
Lateral load resisted by this mechanism is:
P = CCB /1.88 = 621.4/1.88 = 330.5&/V
Total shear capacity: 349 + 330.5 = 679.5kN
Safety factor 679.5/525 = 1.29
The two mechanisms work together to resist the lateral load in the early loading
222
stage. With increasing displacement levels, overstrength of the vertical steel should
be considered. The maximum lateral load obtained from experiment was 601kN, the
safety factor is 1.13. In this calculation, the contribution of vertical steel in
compression is not considered to resist vertical compression. If this contribution is
taken into account, the depth of the compression zone will be less. Under cyclic
load, especially under strong earthquake, the concrete cover in compression can to
be damaged, which leads to decrease the depth of compression zone. This indicates
that strong earthquake will cause a sudden failure with sudden loss of strength and
stiffness and the wall will experience a non-ductile behavior when the maximum
flexural capacity of the wall is reached.
Strut-and-tie models shown in Fig.5.1.4-1 and Fig.5.1.4-2 work simultaneously.

A combined strut-and-tie model is developed incorporating the first and second
mechanism according to the minimum energy dissipation theorem. The minimum
energy dissipation theorem is principle of minimum strain energy for linear elastic
behavior of the struts and ties after cracking as shown in following:
£ / % * „ = minimum; (5-1)
Where Ft =force in strut of tie i
I, =length of member i
smi =mean strain of member i
The contribution of the concrete struts can generally be omitted because the
strains of the struts are usually much smaller than those of the steel ties.
Under same lateral load, strain energy dissipated by ties in the models in
Fig.5.1.4.1 and Fig.5.1.4-2 is list in Table.5.1.4-1.
223
Table 5.1.4-1 Strain Energy in Strut-and-Tie of Specimen Wl
Mechanism Member Force Length £ Strain Energy Contribution Rate
CD 87 1390 0.55 Sy 66.5 S,
DA 159 1150 £
y
182.85 ^ ,
First 0.503
DE 100 1600 £
y
160£v
Z« 409.4 ev
Second CA 159 2540 £

y
403.86 Sy 0.497
al=E2l{E2+El) = 403.86/(403.86 + 409.4) = 0.503
a2=El/(E2+E1) = 409.4/(409.4 + 403.86) = 0.497
First mechanism: 525 x 0.503 = 264kN < 349/CJV

Second mechanism: 525 x 0.497 = 26\kN < 330.5&V
Two strut-and-tie models shown in Fig.5.1.4-1 and Fig.5.1.4.2 take effect
together. In order to obtain a mechanism with minimum strain energy, a combined
strut-and-tie model as shown in Fig.5.1.4-3 is built based on models in Fig.5.1.4.1
and Fig.5.1.4-2 with appropriate proportion, in which 50.3 percent of the lateral
force is resisted by the model shown in Fig.5.1.4-1, and the rest, 49.7 percent of the
load, is taken by the model shown in Fig.5.1.4-2.
224
T r
N
IOC N
N
S
])^5
1390
87
100 ff\. E
1 >. 1
X S
* -
^ 1
1150
25 87
"159 72JMv \
1 v
ioo s ^ !
h
>X188
159
Hw
1600 .100
Fig.5.1.4-1 Simple Strut-and-Tie Model of Specimen Wl
c
100
\188
159
159 jjps
159
1600 lion
Fig.5.1.4-2 Additional Strut-and-Tie Model of Specimen Wl
225
100
-K *
122.8 \43.8W66
\ 50.3 \ .
79 \88.4 X
~4<K?\
Si D
50.3\
43 8
159 36.2\*--J>1.9 \ -
50.3
ts
Nt \ y
s
159
1600 100
Fig.5.1.4-3 Combined Strut-and-Tie Model of Specimen W l
Fig.5.1.4-4 compression zone (by Yanez 1993)
226
b. Strut-and-Tie Model II
In the strut-and-tie model II, the contributions of vertical bars and channels to
the flexural capacity are divided and calculated separately. In this situation, sixteen
vertical bars and two channels yield in tension. This is useful to investigate the
effect of the layout of vertical reinforcement on the ductility of the wall.
The distance between center of the bars in tension to the center of assumed
compression was 990mm. Therefore, shear force caused by vertical bars:
(8x2x78.5x385)x990/2540 = 188.5JW
And Shear force aroused by channels:

2x856x310x1600/2540 = 334.3&V
The total shear force: 334.3 + 188.5 = 523kN * 525kN
Reaction force in compression zone:
C = 8x 2 x 78.5 x 38.5 +856x2x310 = 1014.3&V
Depth of compression zone:
C 1014.3x1000 _An
a = —= = 240mm
0.S5 fcb 0.85x35.5x140
It can be seen that the flexural capacity of the wall when contribution of
vertical bars and channels to flexural strength is calculated separately is almost
equal to that of strut-and-tie model I in which the half vertical bars and channel are
seen as concentrated bars placed in boundaries. However, the depth of compression
zone is 240mm is twenty percent higher that of strut-and-tie model I 197mm.
Similar procedures as shown in developing strut-and-tie model I in previous

section are applied to develop the strut-and-tie models, which are built in
accordance with vertical bars and channel, respectively. The strut-and-tie models
only considering the vertical bars are shown in Fig.5.1.4-5 and Fig.5.1.4-6. The
227
strut-and-tie models only pertaining to channels are given in Fig.5.1.4-7 and

Fig.5.1.4-8. Basing on the minimum energy dissipation theorem, the combined
strut-and-tie models are developed similar to that in strut-and-tie model I as shown
in Fig.5.1.4-7 and Fig.5.1.4-10. And the lateral loads act on the two combined
strut-and-tie models according to the percentage of their flexural strength to the
total flexural capacity of the wall, as shown in Fig.5.1.4-10 and Fig.5.1.4-11.
Table 5.1.4-2 Strain Energy in strut-and-tie of specimen Wl
Mechanism Member Force Length S Strain Energy Contribution Rate
EC 140 1390 0.55 Sy 107 Sy
CA 256 1150 S
y 294A£y
First 0.57
CD 100 990 e, 99 Sy
Z« 500.4 £v
Second EA 256 2540 Zy 650.2 £y 0.43
a, =650.2/(650.2+ 500.4) = 0.57
a2 = 500.4/(650.2 + 500.4) = 0.43
228
Fig.5.1.4-5 Strut-and-Tie Model I of Specimen Wl for Vertical bars
Fig.5.1.4-6 Strut-and-Tie Model II of Specimen Wl for Vertical bars
229
100 w
189.8
-V
7\8
-r9§-
\57 >4
110
Fig.5.1.4-7 Combined Strut-and-Tie Model of Specimen W l for Channels
••c
100
>ss132
132.5
87 87\ V
100
100 £x
'159 72 •J23 87
100 N
2S
°0
B
159
1600 100
Fig.5.1.4-8 Strut-and-Tie Model I of Specimen Wl for Channels
230
100
159 \188
\
s
159 7<w\
H
159
lf>0() J4JLM
Fig.5.1.4-9 Strut-and-Tie Model II of Specimen W l
100
"YN
122.8 \ « . * K 6 s 6 - 6
\ 503 \
771 ^8.4
I VXtt
\
\
^9J'\
D 5l)J\
159 36.2>J1.9 \ 43.8

50.3
V \
M
159
1600 - / - / •
100
Fig.5.1.4-10 Combined Strut-and-Tie Model of Specimen Wl
231
E
188.5
w
\ \
357.8 \ \
4- \ 184.7
150>J
k)7.4*\
\118.K
207.4
81 . 1073 i
J
>s \ A
*-)H
Fig.5.1.4-11 Strut-and-Tie Model of Specimen W l
C
334.3
T T
410.5 \46.4\^22-6
264 \88.4
16p\ N \
-~™X U—^
232
c. Strut-and-Tie Model III
In strut-and-tie model II the contributions of vertical bars and channels to total
flexural capacity are calculated separately. And two combined models according to
the flexural strength caused by vertical bars and channels are also drawn out in
Fig.5.1.4-11 and Fig.5.1.4-12. However, this procedure is too tedious. Herein, the
strut-and-tie model III considers the respective contributions of vertical bars and
channels to the flexural capacity of the wall and uses a single vertical tie to replace
vertical bars and channels at the center of the tension zone. The force in the tie is
equal to sum of the vertical force caused by the vertical bars and that caused by the
channels. The flexural capacity caused the tie is same as that caused by that of the
vertical bars and channels. The assumed position of the vertical tie is at the centroid
of the tension zone.
The vertical force caused by vertical bars:

8 x 2 x 7 8 . 5 x 3 8 5 = 483.6&V
The vertical force caused by channels:
856x2x310 = 530.7/W
The total flexural capacity:
483.6 x 0.99 + 530.7 x 1600 = U21.9kN • m
The distance from center of tension zone to the center of compression zone
(assuming depth of compression is 200mm same as that in previous sections):
1327.9/(483.6 + 530.7) = 1.3 \m
Because the vertical tie in this strut-and-tie model is a substitute of vertical bars
and channels. Therefore the strut-and-tie model III will have same flexural capacity
and depth of compression as strut-and-tie model II. The strut-and-tie model III is
simpler compared with strut-ant-tie model II and has the identical shear
reinforcement.
233
Table 5.1.4-3 Strain Energy in strut-and-tie of specimen W l
CD 106 1390 0.55 Sy 81 Sy
DA 194 1150 £
y 223.1 Sy
First 0.53
DE 100 1310 £
y 131 £\.
435.1 Sy
ZW
Second CA 194 2540 £y 492.8 e 0.47
a, =492.8/(435.1 + 492.8) = 0.53
a2= 435.1/(435.1 + 492.8) = 0.47
100
-v
106 106 \137
100 \
2 D 100
s
194 ^133 94,
Jl
V, T
Vx—r
N #
M
\218.2
ip^ ^ 0
7J70 700
Fig.5.1.4-13 Strut-and-Tie Model of Specimen Wl
234
100
\208
\
194 \
\
\
\
B
\218.2
1310 100
-y+
100
\
\ \
147.4 $6.2 ^J2.6
91.\ V 53~^\
\97.6 \,
D 53
- x — *
\105
\
\ 56.2
194 46.6
•<: V
53 ^ v
V. \ t
ST*. B
\218.2
1310 100
235
+C
523
SNr
770.9 294 J03.8
478
\536.6 \.
245~J^s KX
"Si D 277.^2 M_\
-t—<-
1014 245.7
\ i
277.2 X \
j£' B
-±-
, \7J47
>o
52.
7370 700
Fig.5.1.4-16 Combined Strut-and-Tie Model of Specimen W l

The width of the diagonal concrete strut is:
ws =c/'sin<9 = 240xsin41.3o = 158.4mm
C M =0.75/>w, =0.75x35.5 x 140x158.4 = 590.4JHV

The shear force can be resisted by the first mechanism is:
590.4/1.33 = 443.9&V
ws = d' sin a = 240 x sin 62.7" =213.3mm
CDB =0.75/>w s = 0.75x35.5x140x213.3 = 795kN
795/2.18 = 364.7)W
Shear capacity of this model:
P = 443.9 + 364.7 = 808.6/t/Y
Safety factor:
808.6/525 = 1.54
The safety factor of the strut-and-tie model III is higher than that of than
236
strut-and-tie model I in which all the vertical reinforcement is placed in boundaries

of the wall which has a similar flexural capacity to that of strut-and-tie model III. It
indicates that in squat walls the evenly distributed reinforcement is helpful to
improve ductility of the wall. If the concentrated reinforcement is placed in
boundaries of squat wall, the wall will fail under cyclic loads after yielding of the
flexural reinforcement with low ductility. Under a strong earthquake, the walls with
concentrated reinforcement in boundaries will fail suddenly with non-ductile
behavior due to the relative low depth of compression zone.
A. Second Stage (DF more than 4)
After ductility reached 4, the diagonal bars in the lower part of Wl yielded
when they were in tension. And in following cycles, a wide horizontal crack passed
through the interface between the wall and foundation. It indicated that these
diagonal bars involved into resisting lateral load acting at top beam. This is
discussed in later sections. In the strut-and-tie model describing performance in
second stage, the contribution of diagonal bars to resist lateral load should be taken
into account.
Due to the layout of the vertical reinforcement and channels and low aspect
ratio, if no diagonal reinforcement is provided in this research, the specimen will
experience sliding shear failure under cyclic reversed load. Under load reversals,
the intersecting flexural cracks propagate across entire length of the wall at the base
of the wall. As subsequent reversed loads are applied, the concrete near the base is
destroyed by the breaking, abrasion and spalling. With the concrete destroyed, shear
transfer by "truss action" is no possible and the horizontal reinforcement becomes
ineffective. Interface shear transfer is lost. Shear sliding prevents the ductile
behavior after yielding of the main vertical reinforcement. Deterioration of concrete
237
in this region is intensified by the vertical elongation of the wall, caused by the
residual strain in the vertical reinforcement. These strains developed into the
inelastic range with successive load reversals.
As observed experimentally by many researchers, in this case the applied shear

is resisted mainly by friction and aggregate interlock between the cracked surfaces
along the contact zone. For large deformations, dowel action of the longitudinal
reinforcement also takes place. The sliding shear failure is not significantly affected
by the level of confinement provided in ductile design. Diagonal reinforcement is
perhaps the only successful solution for reducing the potential for sliding shear and
enhancing the hysteretic characteristics of squat walls.
Other methods are investigated by other researchers, Saatcioglu and Murat

(1990) experimentally, and by Wood and Sittipunt through FEM (1995). Herein
only contribution of diagonal reinforcement to shear strength is studied.
The contribution of diagonal reinforcement in compression and in tension to the

lateral load is taken into account by a diagonal strut and diagonal tie with a 45 deg
angle to horizontal, respectively The diagonal bars intersected at the mid-point of
the base line. As shown in Fig.5.1.4-17, the acting lateral load resisted by the
diagonal reinforcement was applied at the top beam. Considering the length of the
diagonal reinforcement that extended into the wall from the base beam and the
angle between the diagonal reinforcement and the horizontal, the diagonal strut and
tie were placed in the strut-and-tie model. It is evident in Fig.5.1.4-17 the
contribution of the diagonal reinforcement to resist lateral load. And it is shown in
this figure that the diagonal reinforcement did not participate in resisting moment at
the base level caused by lateral load and only contributed to the resistance of lateral
load, which is the purpose of using diagonal reinforcement. This also verified that
238
vertical reinforcement located in the central part of the squat shear wall (wall web)
was the primary lateral load resisting component in the late load cycles, especially
when the vertical reinforcement in the boundaries of the wall experienced
significant yielding.
The rest of the lateral load is resisted by the strut-and-tie model is shown in
Fig.5.1.4-18. With increasing displacement levels, the elongation of the vertical
steel, vertical rebars and channels, increased when they were in tension with wide
flexural cracks near the base. When the load reversed, these wide flexural cracks
that had opened previously, tended to close. When the lateral load reached zero,
large residual deformations of the vertical reinforcement remained and the flexural
cracks were still open. With increasing load, before the flexural cracks closed, the
vertical force in the compression zone caused by the lateral load was resisted only
this vertical reinforcement. When a limit of vertical load acting on this vertical
reinforcement reached, these vertical steels buckled, which caused serious spalling
of the cover concrete. Therefore, the compression zone became seriously damaged,
which caused pronounced decrease in shear strength and stiffness of the wall. In the
following cycles, the concrete confined in channels was compressed to failure when
the horizontal reinforcement near the base link two channels together failed.
And with increasing displacement levels, the overstrengh of the vertical steel
occurred.
4x 78.5 x 484.5 = 152.1&/V

The lateral force resisted by diagonal reinforcement (the contribution of the
diagonal reinforcement in compression is not considered):
152.1xcos45° =107.6kN
(107.6x2)/601 = 35.8%
239
Shear force resisted by horizontal reinforcement:

525-107.6x2 = 309MN
Number of horizontal bars:
309.8x1000/(2x78.5x484.5) = 4.1
215.2x1000/(2x78.5x484.5) = 2.8
107.6x1000/(2x78.5x484.5) = 1.4
240
309.8
-V
328.7 328.7 ^^424.7

^
N
D 309.8 Jgf N
328.7
600.7 272 ^•-412.3
309.8\
0s-
6oo.7[mo
3B.8
1310 100
5.1.5 Summary
A. The benefits of the evenly distributed vertical reinforcement:

1. When the wall with evenly distributed vertical reinforcement had the
same flexural capacity as that of the wall with concentrated vertical
reinforcement at the ends, the wall with evenly distributed vertical
reinforcement exhibits more ductile behavior before concrete crushing
failure occurs. The main root for this phenomenon is that depth of
compression in wall with evenly distributed reinforcement when the
wall reaches full flexural strength than that in wall with concentrate
vertical reinforcement at boundaries, even thought they have same
amount of vertical reinforcement.
2. For walls with same amount vertical reinforcement and with same
flexural capacity, when the walls reach their flexural capacity, more
vertical reinforcement yields in tension in walls with evenly distributed
vertical reinforcement that that in wall with concentrated vertical
241
reinforcement in boundaries of the wall, which causes different

compression depths, influencing the width of diagonal concrete strut
which is the main element to transfer lateral load to the foundation.
3. Due to the higher depth of compression in the wall with evenly
distributed vertical reinforcement, there are higher safe residual than
that in wall with concentrated reinforcement in boundaries.
4. From this analysis, it is better to place more vertical reinforcement in
the central part of the wall if flexural capacity is satisfied, which can
lead to deeper compression zone, finally produce wider diagonal
concrete strut which means higher safe overplus for concrete crushing
and make the wall exhibit more ductile behavior under cyclic load. This
is useful in unpredictable earthquake.
5. Through analysis of strut-and-tie models, it is found that the shear force
is mainly resisted by compression zone.
B. Benefits of diagonal reinforcement:

1. compared with the wall without diagonal reinforcement near the base,
walls with diagonal reinforcement exhibit higher ductile behavior.
2. the contribution of diagonal reinforcement to resist lateral load took
effect after the ductility DF, especially contribute more after appearance
of the horizontal crack crossing the wall along the baseline in late
cycles.
3. the diagonal reinforcement decrease the force in horizontal
reinforcement, as a result, the force in diagonal concrete strut was
decreased, which means that the concrete crushing failure of the
concrete in the compression zone was postponed.
C. Benefits of confinement:
1. increase ductility of wall, because effect of confinement on improving
242
ductility is more pronounced than on improving strength.

2. confining effects increase strength of the compression zone after
spalling of concrete cover. Concrete in confinement is not easy to be
damaged and postpones compression failure of the compression zone,
as a result, ductility is increased.
5.2 Behavior of Specimen W2
The strain histories of the vertical bars, SV3, SV5 and NV9, is shown in
Fig.5.2.1-1 to Fig.5.2.1-3. These figures show that the records of strain gauges
located at low level and middle level yielded when ductility reached 2. However,
the deformation of the vertical rebars near the boundaries of wall was not so large to
break these bars whereas it occurred in specimen Wl after serious yielding.
-2000
•4000
0 50 100 150 200 250 300 350 400 450 500
Scan Number
243
10000
.£ 4000 -
S3
-2000
0 50 100 ISO 200 250 300 350 400 450 500
Scan Number
-2000
100 200 300 400 500
Scan Number
B. Strain Profile ofVertical Bar at Base Line
The strain profiles of vertical bars near the base under negative and positive
244
load are shown in Fig. 5.2.1-4 and Fig.5.2.1-5, respectively. It can be see from the
two figures that in the first loading cycle, the assumption of plane section can be
used. With increasing displacement levels, the distance from left side of the wall to
the intersection between profiles and the base decreased accompanied with
appearance of flexural cracks, and finally reaches a constant.
0 200 400 600 800 1000 1200 1400 1600 1800 2000
DistancefromLeft Side of the Wall (mm)
Fig.5.2.1-4 Strain Profile of Vertical Bars near the Base for Negative Load
8000
-4000
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Distance from Left Side of the Wall (mm)
Fig.5.2.1-5 Strain Profile of Vertical Bars near the Base for Positive Load
245
C. Strain History of Diagonal Bars
Strain histories of the diagonal bars near the base and in the horizontal
connection region are shown in Fig.5.2.1-6 to Fig.5.2.1-13. Before reaching
ductility of 2 (scan number 137), the strain of diagonal reinforcement was less than
2200 /us , which means during this stage the contribution of these diagonal bars near
the base is negligible. When the ductility reached 3 (scan number 230), the
diagonal rebars near the base yield. It can be also seen from Fig. from ductility
equal to 2, the diagonal reinforcement in the horizontal connection yields. The role
of these rebars in resisting lateral load should be taken into account.
600
500 \ —"—
400 /JnT
300
200
100 :2S .2
•a
a ^N -&
0 v
\ -100
/ "
-2 -200
SDl
-300
•400
-500
-600
-500 500 1000 1500 2000 2500 3000 3500 4000 4500
Strain(fis)
Fig.5.2.1-6 Hysteresis Loops of SDl
246
4500
4000 ]
3500
3000 -
-v 2500
I
.s 2000
2
3 1500 -
1000
500
0
0 100 200 300 400 500
-500
Scan Number
Fig.5.2.1-7 Strain history of SD1 versus Scan Number
600
500
400
300
200
\\ y
100 «5\ uy
«>k
0
-100 ND3
-200
-300
-400
-500
-600
-1000 0 1000 2000 3000 4000 5000
Strain(fie)
Fig.5.2.1-8 Hysteresis Loops of ND3
247
100 200 300 400 500

Scan Number
Fig.5.2.1-9 Strain history of ND3 versus Scan Number
1000 1500 2000 2500 3500

Strain (fie)
Fig.5.2.1-10 Hysteresis Loops of SUD5
248
50 100 150 200 250 300 350 400 450 500

Scan Number
Fig.5.2.1-11 Strain history of SDU5 versus Scan Number
600
500
400
\V^K ^<*~--*4%'
• %?
300 ^V\V\ _2|:
ffvtlN l ZM~- Sfc-
200
WCM\ li^ ^ NDL7
100
\
0
2 -100
M -200 -
-300
-400
-500
-600
-500 500 1000 1500 2000 2500 3000 3500 4000
Strain (fie)
Fig.5.2.1-12 Hysteresis Loops of NDL7
249
0 SO 100 ISO 200 250 300 350 400 450 500

Scan Number
Fig.5.2.1-13 Strain history of NDL7 versus Scan Num
D. Strain History of Horizontal Reinforcement
Strain histories of the horizontal reinforcement are shown in Fig.5.2.1-14 to

Fig.5.2.1-19. From these figures, it can be been that almost all of the horizontal
reinforcement yielded, which is different from the records of strain gauges of
horizontal reinforcement in specimen Wl, in which the horizontal reinforcement in
the upper part of the wall did not experience yielding.
250
-500
0 SO 100 ISO 200 250 300 350 400 450 500
Scan Number
Fig.5.2.1-14 Strain History of Horizontal Bar NH11-2
3500
3000 - \
2500 H
2000 -
1
.5 1500
f
1
I
AJ IAJW
NH10-2 .
1
1000
500 -\
vA/lf
1
-500
0 50 100 150 200 250 300 350 400 450 500
Scan Number
Fig.5.2.1-15 Strain History of Horizontal Bar N H 1 0 - 2
251
4500 -
3500
<£ 2500 H I I ^ = F= ~ S= :=»
S#9-2
1500
500
-500
0 50 iOO 750 200 250 300 350 400 450 500
Scan Number
Fig.5.2.1-16 Strain History of Horizontal Bar SH9-2
-500
0 50 100 150 200 250 300 350 400 450 500
Scan Number
252
5500
-500
0 50 100 150 200 250 300 350 400 450 500
Scan Number
3500
3000 - 7T.
2500 -
NH2-2
2000 -
1
.5 1500
2
5
-500
0 50 100 150 200 250 300 350 400 450 500
Scan Number
E. Strain History of Channel
253
600
500
400
300
200
100
0 V» \ i \ V \ \J\ \ f\\ \ i /
•«
a
s
-J -100
-200
-300
-400
-500
-600
-2000 0 2000 4000 6000 8000 10000 12000
Strain(fie)
Fig.5.2.1-20 Strain History of Horizontal Bar SS2-1
0 2000 4000 6000 8000 10000 12000

Strain(fie)
Fig.5.2.1-21 Strain History of Horizontal Bar SS2-2
254
I
•a
Fig.5.2.1-22 Strain History of Horizontal Bar NS2-4
1500
255
600
500
400
300
200
100
iWjnfcy/^/' NSI-6
0 [$&} Jr^^^ ,
1
<i
-100
-200
-300 A
-400
-500
i
-600
-200 200 400 600 800 1000
Strain(fie)
5.2.2 Load Transfer
In the early loading stages, the cracking patterns were similar to those of
specimen Wl in same stage. It indicates that the internal forces of specimen W2 is
similar to those in specimen Wl before ductility reached 2.
With increasing displacement levels, the cracks patterns were are different from
those of specimen Wl. The cracking patterns of specimen W2 in these stages
exhibited two distinct features compared to those of specimen Wl. First, diagonal
cracks formed in the upper and lower wall panels of W2, which made the wall seem
as two independent wall panels. Second, the range of diagonal cracks at the top
beam level spread over half the width of the wall whereas these kinds of cracks
were limited in a smaller range in specimen Wl. This means that a diagonal
concrete strut DE started approximately from center of the top beam, as shown in
Fig.5.2.4-2, and the tie DC to balance this concrete strut was part of the vertical
reinforcement in the upper wall panel. This means that the existence of the
256
horizontal connection changed the aspect ratio of the wall, especially in upper wall
panel, which finally changed the internal force paths in whole specimen.
Readings of the strain gauges placed on the diagonal bars in the horizontal
connection showed that these diagonal bars yielded after a global ductility of 2.
However, after the appearance of a horizontal crack along the horizontal connection,
no additional deformation or damage of the horizontal connection was observed,
which indicated that the horizontal connection successfully transferred the load
from the upper wall panel to the lower one. In order to simplify the procedure of
developing a strut-and-tie model, the connection zone was defined as the region in
the upper or lower panel containing diagonal reinforcement, and the compression
forces in the concrete struts were assumed to transfer through it. In another words,
the connection zone was assumed to be continuum.
The wide horizontal crack crossing through the wall along the interface with
the foundation, which occurred in specimen Wl was not observed in specimen W2.
However, two diagonal cracks formed just above diagonal reinforcement near the
base intersect at 125 deg angle. The protruding concrete tip and the horizontal bars
crossing it formed the main shear resistant mechanism under cyclic load. In later
loading cycles, the concrete in compression at the ends of the diagonal
reinforcement near the base experienced spalling as shown in Fig.5.2.3-1, which
caused a significant decrease in the compression area of the diagonal concrete struts
at this location. As a result, it decreased the shear resisted by the concrete. Therefore,
shear resistance requirement of the horizontal and vertical reinforcement increased,
which finally caused fracture of the horizontal reinforcement. This indicates that in
shear design for cyclic load, the contribution of concrete to shear should be ignored,
especially in squat shear walls.
In early stages, strut DB shown in Fig.5.2.4-1, was the main shear resistant
257
member, whereas in the later loading stages, strut HI shown in Fig.5.2.4-4, was the
main member to resist lateral load. There was a shift of critical location for resisting
lateral load during the course of loading.
Records of the strain gauges located on diagonal reinforcement near the base
showed that diagonal reinforcement near the base yielded. The contribution of these
diagonal bars to the lateral load separated into forces in tie LK and concrete strut
GL as shown in Fig.5.2.4-4.
5.2.3 Crushing of Concrete and Buckling of Vertical Steel
After reaching a ductility of 2, the critical compression zone shifted from the
lower corners of the wall to the up ends of"diagonal reinforcement near the base.
As shown in Fig.5.2.4-4, after ductility reached 2, a diagonal strut HI from

loading point to end of diagonal reinforcement near the base formed. The density of
the diagonal cracks at this deformation area was high. Under cyclic loading,
concrete at the up end of the diagonal reinforcement near the base deteriorated and
spalled. After several loading cycles, the total area of compression struts at this
location decreased due to spalling. With increasing load, it can be observed that the
diagonal concrete bands in compression at the height level 650mm from the wall
foundation interface were sheared one by one, which caused increment in tensile
forces horizontal reinforcement and finally fracture the horizontal reinforcement.
And the specimen at last experienced shear failure.
When the horizontal reinforcement crossing the protruding concrete tip which
resisted lateral force yielded, the channel embedded in the wall buckled at height
level where horizontal reinforcement broke.
The different crushing location of W2 compared to that of the integral wall Wl
258
was mainly caused by the existence of the horizontal connection, which changed the
internal load paths through which lateral load acting on top panel could be
transferred to the bottom panel.
Fig.5.2.3-1 Crushing of Concrete in Specimen W2
259
Fig.5.2.3-2 Crushing of Concrete in Specimen W2
Before ductility of the wall reached 2, a horizontal crack developed across the
horizontal connection, however, no evidence of sliding was observed along the
connection. The crack patterns were similar to those of specimen Wl at similar
displacement levels. From records of strain gauges placed on the diagonal rebars
near the base as shown in Fig.5.2.1-7 and Fig.5.2.1-9, the contribution of these
diagonal bars to lateral load resistance was negligible before reaching the ductility
of 2. Therefore, in this stage the strut-and-tie mode shown in Fig.5.2.4-1 is
applicable.
260
c
•*h
100
TNT
122.8 \43.8\^6-6
\ 5«\
79 \88.4 X
I rn J* «* ^
Fig.5.2.4-1 Strut-and-Tie Model for W2
After reaching a ductility of 2, the diagonal cracks belonging to the upper and
the lower panels were observed. It indicated changes of the internal load paths of
W2 Due to change of aspect ratio of the upper wall panel, the upper panel was
seemed to be monolithic wall and the connection zone was considered as its
foundation. In this stage, the aspect ratio of the upper panel was 0.59 much lower
than 1. Referring to the cracking patterns in this panel, a concrete strut DE was
added. Due to the changes in force paths in the upper panel, the internal paths were
different from those of specimen Wl.
When the diagonal reinforcement near the base was taken into effect in
resisting lateral load, the contribution of these diagonal rebars was shown in
Fig.5.2.4-4. When the diagonal bars were in tension, they were represented by a
diagonal tie. When they were in compression, the concrete took part in resisting the
compression; therefore, a diagonal concrete strut was used.
261
In Fig.5.2.4-3, lateral load was transferred to the base through the concrete strut
G'F. In Fig.5.2.4-4, in this mechanism, tie KI transferred the lateral load to tie GK
and LK. The internal forces in tie GK and LK were proportional to their total
sectional area.
Table 5.2.4-1 Strain Energy in strut-and-tie of specimen W2

Mechanism Member Force Length e Strain Energy Contribution Rate
AG' 158.8 800 £

y 127 Sv
CD 133.8 1290 £
y 172.6 Sy
BE 100 1600 £
y 160f v
First 0.52
BG' 86.9 650 0.5 Sy 28.2 e
GT 50 1600 0-5 £v 40 £„
>
Z« 527.9 ey
AG 181.4 650 £
y \Yl$ey
GN 118 850 0.65 By 65.2 By
NH 83.9 1040 0.46 Sv 40.1 £v
GK 78 1450 0.78 s 88.2^

Second 0.48
LK 31 919 £
y 28.5*,
KI 100 150 £
y 15 Sy
NM 100 1250 £
y 125 Bv
Z« 479.9 Sy
Based on the minimum strain energy theory, the lateral load was distributed
according to the two mechanisms.
262
E2/(E2+Ey) = 0.4%
E1/(E2+E1) = 0.52
The lateral force resisted by the first mechanism was:

0.48 x 525 = 252kN
The second mechanism resists lateral load:
525-252 = 273&V
The contribution of the two mechanisms to resist lateral load was combined
into one strut-and-tie model, as shown in Fig.5.2.4-7 in which the two mechanisms
took effect simultaneously with appropriate proportion, which is drawn in
Fig.5.2.4-5 and Fig.5.2.4-6. From the combined strut-and-tie model, all the concrete
struts concentrated at point F formed a three-dimensional compression zone. And at
point I a C-C-T node zone formed.
Based on the combined strut-and-tie model, the amount of horizontal

reinforcement needed in upper and lower wall panels was:
For upper panel:
252&V/(78.5 x 2 x 484.5) = 3.1
Safety factor: 5/3.1 = 1.6

For lower panel:
273&V/(78.5 x 2 x 484.5) = 3.59
Safety factor:
5/3.59 = 1.39
For diagonal reinforcement near the base,
84.63JUV/(78.5 x 2 x 484.5) = 1.1
Safety factor:
2/1.1 = 1.8
Comparing these safety factors for reinforcement placed in the specimen, it can
263
be seen that the tie KI was the critical component in the strut-and-tie model.
Considering the overstrength of the vertical reinforcement, the capacity of the wall
was 576.96kN.
576.96/525 = 1.1
This result shows that when the wall reached its capacity, the horizontal tie HI
yielded and initiated plastic deformation, at the same time the contribution of the
concrete to the lateral load decreased. These two factors caused the failure of the
wall.
From the calculations above, the amount of horizontal reinforcement for the
upper wall panel was less than that compared to specimen Wl and more horizontal
reinforcement was demanded in the connection located in the lower panel.
K »
\\ \\
\ \
1
' \\ ' fc
\\ \\
\\ \
\
/ V> M
\ \
connection zone
Fig.5.2.4-2 Strut-and-Tie Model for Upper Wall Panel and Connection Zone in Upper Panel
264
I, 1,100 800 L 800 JJ1W

Fig.5.2.4-3 First Strut-and-Tie Model of Specimen W2
1250
100
I I 100 800 I 650 ) J, L WT

\150
Fig.5.2.4-4 Second Strut-and-Tie Model of Specimen W2
265
Fig.5.2.4-5 First Strut-and-Tie Model of Specimen W2 under Proportioned Load

1250
L [ 100 800 [ 650 [ L I IPT

~" \150
Fig.5.2.4-6 Second Strut-and-Tie Model of Specimen W2 under Proportioned Load
266
1250
Fig.5.2.4-7 Combined Strut-and-Tie Model of Specimen W2
To predict the behavior of the horizontal connection as shown in Fig.5.2.4-8(a),

as studied in section 4.3.2 for design, the flexural deformation of the horizontal
connection was duo to tension and compression deformation, and shear deformation
only comes from deformation of shear connector shown in Fig.5.2.4-8(b). If the
tolerance between bolts and bolt holes is ignored, the behavior of the horizontal
connection can be simulated by the model shown in Fig.5.2.4-8(c).where Kl, K2
and K3 represent the stiffness of the tension, compression and shear springs in the
model, respectively
/
Kl = K2 =
A.E.
267
Where / is the height of connection zorte; A is the effective area of tension tie;
£„ is elastic modulus of steel.
/
^3 =
AdEs sinacona
Where Ad is total area of diagonal reinforcement in tension; a is the incline
angle of diagonal tension bars.
When tolerance between bolts and bolt holes exist and are taken into account, a
reduced value of stiffness should be used.
/N /
\
S S
} \
X s
/ N
V > \
(a)
(b)
Fig.5.2.4-8 Behavior of Horizontal Connection(cont')
268
Kl K2<
K3
(c)
Fig.5.2.4-8 Behavior of Horizontal Connection
5.2.5 Summary:
• In the early stages of loading, when the ductility was less than 2, the
crack patterns of specimen W2 were similar to those of specimen Wl at
the same displacement levels. It means that in the early stage the precast
wall behaved as an integral wall. It also indicates that the internal load
paths were similar to those of specimen Wl at this stage.
• In later loading cycles, spalling of the concrete at the ends of the
diagonal reinforcement near the base caused a decrease in the area of
the compression zone at this location which led to a decrease in the
contribution of the concrete to the lateral load. Simultaneously, the
forces in the horizontal reinforcement increased and finally caused
fracture of these bars at a height level that corresponded to the end of
the diagonal reinforcement near the base.
• In strut-and-tie model developed for Wl, the force in the horizontal tie
could distributed in more horizontal reinforcement with respect to the
number of the involved horizontal bars of the lower wall panels of W2.
It was consistent with the developed strut-and-tie models of W2. It
showed that more reinforcement were in need in between the diagonal
reinforcement in the connection and that near the foundation.to arrest
the failure of specimen W2.
269
• When the ductility was larger than 2, the internal load paths were
different from those in the specimen Wl at the same displacement level.
This situation was due to the existence of the horizontal connection in
the precast wall W2. In specimen W2, the aspect ratio of the upper wall
panel was 0.6. In later loading cycles, the diagonal cracks in the upper
panel spread to the center of the wall at top beam level, which meant
that a diagonal concrete strut started to form from the center of the top
beam.
• Over the whole course of loading, deformation or damage of the
horizontal connection was not observed, which indicates that the lateral
load acting on the top beam was successfully transferred from the upper
wall panel to the lower one.
• Specimen W2 experienced a different failure mode and failure location
than that from those of specimen Wl.
• Contribution of concrete to lateral load resistance under cyclic load
should not be taken into account, especially in a squat wall.
• Compared with specimen Wl, less horizontal reinforcement is
demanded in the upper part of specimen W2. And more horizontal
reinforcement was needed in the connection zone located in the lower
panel.
• Strong base wall is needed for the precast wall due to its high
requirement on shear strength.
Specimen W3 was another precast counterpart of specimen Wl. The only

difference between specimen W2 and W3 was the shear connectors embedded in
wall panels. The rest of the details including reinforcement, channels are same to
270
those in specimen W2.
5.3.1 Strain History of Gauges
Records of the strain gauges placed on the vertical bars are shown in
Fig.5.3.1-1 to Fig.5.3.1-3. It can be seen from these figures that all the readings
noted by L were much bigger than 2700, which means that all the vertical bars
yielded at the baseline level. Some the records label as M yielded during the loading
course while the others did not, such as SVM6. It indicates that at the M level there
was elastic core whereas the two sides of wall yielded. The vertical bars labeled as
H experienced less deformation as shown in these figures, which is a result of
cutting of vertical bars into two at the horizontal connection. All these features of
the vertical bars are similar to those in specimen W2. It hints that the type of shear
connectors made no much difference on the performance of vertical bars.
0 50 100 150 200 250 300 350 400 450

Scan Number
271
50 100 150 200 250 300 350 400 450

Scan Number
50 100 150 200 250 300 350 400 450

Scan Number
B, Strain Profiles of Vertical Bar at Base Line
Profiles of vertical bars near the base are shown in Fig.5.3.1-4 to Fig.5.3.1-5 for
272
negative and positive loading. It can be seen that in the first cycle, namely under a
load of 0.25 Pi, the strain profile exhibited almost a linear trend. With increasing
load this trend changed. It also shows that the depth of compression zone increased
with the increasing load and became stable when the load reached yielding load at
about 190mm from the extreme in compression fiber.
Cycle 1
Cyde 2
Cycle 3
Cycle 4
— • — Cycle 5
200 400 600 800 1000 1200 1400 1600 1800 2000
Distance from Left Side (mm)
Fig.5.3.1-4 Strain Profile of Vertical Bars near the Base for Negative Load
273
7000
6000
5000
4000
3000
I 2000
1000
-1000
-2000
-3000
0 200 400 600 800 1000 1200 1400 1600 1800 2000
DistancefromLeft Side (mm)
Fig.5.3.1-5 Strain Profile of Vertical Bars near the Base for Positive Load
The strain history of diagonal bars is shown in Fig.5.3.1-6 and Fig.5.3.1-8. To

some extent, these figures indicate the energy dissipation capacity of these bars. The
relationships of diagonal bar strain versus scan number are shown in Fig.5.3.1-7 and
Fig.5.3.1-9. It shows that diagonal bar ND1 yielded at scan number 181 and ND4
yielded at scan number 157. These bars showed negligible negative values.
The strain history of the diagonal bars welded on the steel shear connectors and
relationships of strain versus scan number are shown in Fig.5.3.1-10 to Fig.5.3.1-17.
These bars include the diagonal bars embedded in the upper and lower wall panels
and the different loading directions. All of the diagonal bars placed in the
connection zone yielded after the specimen went into plastic stages.
274
700
600
500
400
300
200
$> 100
I o W/^l
\f
/ -^ x^2k*=^-"
§ -100
-200 f\
-300
-400 -\ V\\ \ \V \
-500 ^N ,,*
-600 -^ ND1 — ^ J*
t^Sfc=
-700
-7000 7000 2000 3000 4000 5000 6000
Strain(pis)
Fig.5.3.1-6 Strain History of Diagonal Bar ND1
6000
1 3000
i 2000 H
"5
700 200 300 400 500

5c«« Number
Fig.5.3.1-7 Relationship Strain versus Scan Number of ND1
275
-1000 0 1000 2000 3000 4000 5000 6000 7000 8000 9000
Strain(fie)
Fig.5.3.1-8 Strain History of Diagonal Bar ND4
9000
-1000
50 100 ISO 200 250 300 350 400 450 500
Scan Number
Fig.5.3.1-9 Relationship Strain versus Scan Number of ND4
276
700
600
500 -
^^ *- '-' ^
400 ^fc]
'
W^
300 ;
200 ^> ^wf
3 ^ 3 ;> ^" ^^-~.M)I/6
w\^\
? 100
1
| 0 ' |'
| -100
-200 -\ ^^t^^^
-300
•400 -
-500
-600
-700
-500 500 1000 1500 2000 2500 3000
Strain(fts)
Fig.5.3.1-10 Strain History of Diagonal Bar NDU6
3000
| 1000
-500
100 200 300 400 500
Scan Number
Fig.5.3.1-11 Relationship Strain versus Scan Number of NDU6
277
700
600
500
^ ^s :•• /,
400 ^ Ml.'1
300 . SDU5 ~~
S3fccV
j ? ^ ? F Si : ^
200
£ 100 \
% 0 |
.3 -100
-200
-300
-400
-500
-600
-700
-500 0 500 1000 1500 2000 2500 3000 3500
Strainffis)
Fig.5.3.1-12 Strain History of Diagonal Bar S D U 5
3500
-500
0 SO 100 150 200 250 300 350 400 450 500
Scan Number
Fig.5.3.1-13 Relationship Strain versus Scan N u m b e r of S D U 5
278
-500 500 1000 1500 2000 2500 3000

Strain(fiE)
Fig.5.3.1-14 Strain History of Diagonal Bar NDL2
3000
-500
0 50 100 150 200 250 300 350 400 450 500
Scan Number
Fig.5.3.1-15 Relationship Strain versus Scan Number of NDL2
279
700
600
500 \{%\
400 \m\ ^^^./•a:
M l\
i\»V —• SÎs?
300
200 1 VTO.
\\^\
SDV2 ^fteg^
•? * % *\
100
-a 0
-100
-200
-300
-400
-500
-600
-700
-500 0 500 1000 1500 2000 2500 3000 3500 4000 4500
Strainffie)
Fig.5.3.1-16 Strain History of Diagonal Bar SDU2
-500
50 100 150 200 250 300 350 400 450 500
Scan Number
Fig.5.3.1-17 Relationship Strain versus Scan Number of SDU2
280
D. Strain History of Horizontal Bars
Relationships of strain versus scan number for the horizontal reinforcement are
shown in Fig.5.3.1-18 to Fig.5.3.1-24. It can be seen from these figures that all the
horizontal reinforcement in the upper wall panel did not yield during the loading
course, whereas some horizontal reinforcement in the lower wall panel yielded. It
can also be seen that the horizontal reinforcement near the base near the diagonal
reinforcement crossing the wall-foundation interface experienced lower
deformation in center of the wall compared to that of Wl. One of the main factors
affecting the deformations of these horizontal bars near the foundation was the
diagonal bars which resisted part of the shear force; the other was due to the
constraint of the foundation block. The deformation characteristics of the horizontal
reinforcement placed in the upper and lower wall panels are represented in the
strut-and-tie models shown in Fig.5.2.24-7.
Compared to the horizontal reinforcement in specimen Wl, the shear

deformation of the horizontal reinforcement in the lower wall panel was increased
in the later loading stages while those in the upper wall panel decreased.
3000 -i 1
-1
2500 - V,-.
2000 -
SH1-2
1 1500 -
.s
2 1000 -
as
500 -
0 k
-500
0 50 100 150 200 250 300 350 400 450 500
Scan Number
Fig.5.3.1-18 Relationship Strain versus Scan Number of SH1-2
281
3000 b=j
— == —
2500 *v v
v
2000 - SH3-2
f' 1500
A A A AA
I
to
1000
500 -
JJwwWwW
-500
0 50 100 150 200 250 300 350 400 450 500
Scan Number
-- ===
2500 -
2000 - "~^~-AW5-2
\ 1500 -
s
I 1000 -
/AWMi/U,
co
500 -
n -JM
-Sflft -
0 50 100 150 200 250 300 350 400 450 500

Scan Number
Fig.5.3.1-20 Relationship Strain versus Scan Number of NH5-2
282
6000 -r
5000 -
4000 -
1 3000 -
c
| 2000 -
1000 -
0 -
-1000 -
0 50 100 150 200 250 300 350 400 450 500
Scan Number
Fig.5.3.1-21 Relationship Strain versus Scan Number of NH8-2
3000 -T
2500 -
2000 -
& IOOO -
P
500 -
0 -
-500 -
0 50 100 150 200 250 300 350 400 450 500
Scan Number

6000
5000
4000
| 3000 -\
I 2000
1000
-1000
0 SO 100 150 200 250 300 350 400 450 500
Scan Number
Fig.5.3.1-23 Relationship Strain versus Scan N u m b e r of N H 1 0 - 2
3000
B B S
2500
2000
SH12-2
1 1500 -\
/'
& IOOO -
'7
MX
500
-500
0 50 100 150 200 250 300 350 400 450 500
Scan Number
Fig.5.3.1-24 Relationship Strain versus Scan N u m b e r of SH12-2
284
E. Strain History of Channels
The strain history of the channels that acted as concentrated flexural

reinforcement is shown in Fig.5.3.1-25 to Fig.5.3.1-28. Deformation mainly
occurred in the lower part of the wall and exhibited a decreasing trend from the
bottom to the top of the wall. There was no plastic deformation in the channels
embedded in upper wall panel.
The hysteresis loops of NS were similar to those of the flexural bars at the same
location. The configuration of the loops means that much more energy dissipated
through flexural deformation of the vertical channel in lower wall panel than in
upper wall panel.
700
600
500
400
300 - #^^^^/ II 1
200
100
0
•J
-100
-200
-300
-400
-500
-600 NS1-1 ^^^
-700
-6000 -4000 -2000 0 2000 4000 6000 8000 10000 12000 14000
Strain(fis)
Fig.5.3.1-25 Strain History of Channel NS1-1
285
700
600
500
400
300
^ - ^ ^ ^ F ^ ^ ^ ^
200
100
0
-100
-200
-300
-400
-500
-600 NS1-4 f
-700
-1000 -500 0 500 1000 1500
Strain(fie)
-200 200 400 800 1000

Strain(fig)
286
700
600
500
400
300
200
s> 100
0
a
-100
-200
-300
-400
-500
-600
-700
-100 -50 0 50 100 150 200 250 300
Strain(fis)
5.3.2 Load Transfer
In the early loading stages before the ductility reached 2, though the horizontal
crack along the horizontal connection appeared, this specimen still behaved as an
integral wall, which can be seen from the cracking patterns.
After the ductility reached 2, the cracking pattern changed. They showed that
wall panels behaved as independent squat walls with diagonal cracks crossing the
wall panels from corner to the opposite diagonal corner.
In late loading cycles, shear deformation of the wall mainly occurred in the
lower walls panel, and the density of the flexural and shear cracks increased. The
crack density at the upper ends of the diagonal reinforcement near the base was
much higher than any other location. Under cyclic loading, spalling of the concrete
at this location occurred. The concrete spalling at this location decreased the area of
287
concrete strut in compression, which caused increased stress in horizontal

reinforcement passing through region. At this time, these horizontal bars already
yielded and plastic deformation increased fast. As a result, the reinforcement in the
plastic stage could not to prevent further shear deformation at this height level.
Fig.5.3.2-1 Fracture of Channels
5.3.3 Strut-and-Model
As mentioned before, the only difference between specimen W3 and W2 was

the shear connectors. Through investigating the load transfer of specimen W3, it
exhibited similar behavior to W2. In the early loading stages, specimen W2 behaved
like an integral wall, the strut-and-model shown in Fig.5.2.4-1 was developed to
indicate the flow of the internal forces. With increasing displacement levels, the
upper and lower wall panels behaved independently, and the horizontal connection
transferred the lateral load to the lower wall panel successfully before fracture
occurred in channels. The failure of the horizontal reinforcement occurred near the
288
connection zone caused that the lateral load could not transfer completely by the
shear connectors. As a result, the lateral load was transferred by the channel
connections caused the fracture of the channels. The strut-and-tie models shown in
Fig.5.2.4-5 and Fig.5.2.4-6 were developed for this stage.
Specimen W4 was an integral wall with embedded I-beams which were used as
vertical flexural reinforcement in the boundaries of the wall. The difference
between specimen W4 and Wl was just the different embedded steel shapes.
5.4.1 Strain History of Gauges
The strain history of the vertical bars is shown in Fig.5.4.1-1 to Fig.5.4.1-5. It

can be seen from these figures that all vertical bars yielded at the wall-foundation
interface. Almost all of the vertical reinforcement noted by M yielded which was
located 750mm from the interface.
289
12000
10000
8000
6000
,31
e 4000
1 2000 -
-2000
-4000
100 200 300 400 500 600 700 800 900
Scan Number
Fig.5.4.1-1 Strain History of Vertical Bars SV1
Fig.5.4.1-2 Strain History of Vertical Bars NV2
290
Fig.5.4.1-3 Strain History of Vertical Bars SV6
12000
10000
5
.B
S 4000
-2000
100 200 300 400 500 600 700 S00 900
Scan Number
291
•NVL9
NVM9
NVH9
J "5T « ft ft ft "4 A •> ,T:f.rT . /
^ •' U1
too 200 300 400 500 600 700 800 900

Scan Number
B. The Strain Profiles of the Vertical Bars
Strain profiles of the vertical bars along the wall-foundation interface are
shown in Fig.5.4.1-5 and Fig.5.4.1-7.
Cycle 1
Cycle 2
Cycle 3
Cycie 4
Cycle 5
NSide
500 1000 1500 2000
Fig.5.4.1-6 Strain Profiles of Vertical Bars for Negative Load
292
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Fig.5.4.1-7 Strain Profiles of Vertical Bars for Positive Load
C. Relationship of Strain versus Scan Number of Diagonal Bars

Relationships of strain versus scan number of diagonal bars are shown in
Fig.5.4.1-8 and Fig.5.4.1-9. Strains in diagonal bars early loading stage, that is, at
ductilities less than 2, were lower than 2700, which means that contribution of these
diagonal bars to lateral load was negligible. However, after ductility reached 2, the
strain in diagonal reinforcement increased and yielded. This means that the
deformation of the diagonal bars increased, especially, in later cycles after the
appearance of a wide horizontal crack crossing through wall along the interface
with the foundation.
293
16000
14000
12000
10000 H M>7
1 8000 s\
S fc
^
,/y
•5 ^9<V5F*
8 6000 H =d
3
0 100 200 300 400 500 600 700 800 900

Scan Number
Fig.5.4.1-8 Relationship of Strain versus Scan Number of SD1, ND1
12000
10000 -SD3
ND3
8000
1 6000 N
s^ }s -•*£
//
\^ sy j
JS 4000 -
2000
-2000
0 100 200 300 400 500 600 700 800 900
Scan Number
Fig.5.4.1-9 Relationship of Strain versus Scan Number of SD3, ND3
P. Relationship of Strain versus Scan Number of Horizontal Bars
The strain history of the horizontal reinforcement is shown in Fig.5.4.1-10 to

Fig.5.4.1-15. It can be seen that in range of the diagonal reinforcement, the strain
294
readings of the horizontal reinforcement were lower than 2700, which means that
the lateral load resisted by diagonal reinforcement was low. As a result, in this
region, the diagonal and the horizontal reinforcement worked together to resist the
lateral load. But readings of NH9, NH8 and NH6 were more than 2700, which
indicates these horizontal bars yielded during the loading course. Horizontal
reinforcement in higher levels did not yield.
3000
2500
2000 -NH11-2
NH11-3
1500
tt?
§ 1000
s
a soo
0
-500
-1000
-1500
100 200 300 400 500 600 700 SOO
Scan Number
Fig.5.4.1-10 Relationship of Strain versus Scan Number of NH11
295
7000
-1000
100 200 300 400 500 600 700 800
Scan Number
Fig.5.4.1-11 Relationship of Strain versus Scan Number of SH9, NH9
4000
-1000
0 100 200 300 400 500 600 700 800
Scan Number
296
3000
-1000
100 200 300 400 500 600 700 800
Scan Number
3000
2500 -
2000 -
t 1500
S
I IOOO -
-500
100 200 300 400 500 600 700 800
Scan Number
Fig.5.4.1-14 Relationship of Strain versus Scan Number of SH3
297
0 100 200 300 400 500 600 700 800

Scan Number
5.4.2 Load Transfer
In the early stages, the flexural cracks were limited in a range of 200mm from
extreme fiber of the wall and several cracks crossed the wall from the bottom corner
to opposite the top corner with a concrete band width about 200mm. It meant that in
this stage the lateral load was resist mainly by the diagonal concrete strut.
After reaching a ductility of 1, cracks appeared along the inner boundaries of

I-Beams in the lower parts of W4. With increasing load, the range of this kind of
cracks extended upward and the density of the flexural cracks increased in the lower
part of the wall. When ductility reached 3, the spalling of the concrete occurred
along the inner boundaries of the I-Beams in the lower parts of the wall. With
increasing displacement levels, concrete spalling spread upward and finally almost
extended to the top beam as shown in Fig. 4.2.4-4. This phenomenon indicated that
to some extent, the nodal area of the diagonal concrete bands in compression
decreased.
298
A large horizontal crack occurred which passed through the wall at the interface
with the foundation when ductility reached 5. And at this stage, the readings of the
strain gauges placed on the diagonal bars indicated them already yielded. The
density of the diagonal cracks in the wall increased and the horizontal reinforcement
yielded.
5.4.3 Concrete Crushing and Buckling of I-Beam
When the lateral load reached i>, a number of cracks appeared along the inner
boundaries of the I-Beams in the lower part of W4. After reaching a ductility of 4,
the cracks proliferated and spread from the bottom of the wall to its top. The cracks
intersected the horizontal cracks and the diagonal cracks from the top corner of the
wall. Under cyclic loading, spalling of the concrete occurred along the inner
boundaries of the I-Beams. When ductility reached 5, a vertical crack appeared at
the lower left corner of the wall, which caused spalling of the concrete cover as
shown in Fig.5.4.3-1, to some extent, decreased the lever arm when left lower
corner was in compression.
The spalling of the concrete along the inner boundaries of the I-Beam grew
with the increasing loading cycles and influenced the force transfer between the
I-Beams and the contacting concrete, herein, call it bond-failure as shown in
Fig.5.4.3-2. The bond failure led to a result that the lateral load was transferred to
the foundation mainly by the diagonal concrete struts. Due to crushing of concrete
cover in the lower corner, the nodal areas of the concrete strut in compression
decreased. Finally, concrete nodes in compression were crushed with buckling of
the I-Beam in the lower corner.
299
Fig.5.4.3-1 Spalling of Concrete Cover in Compression Zone
Fig.5.4.3-2 Spalling of Concrete along the Inner Boundaries of I-Beams
In the early loading stages, the diagonal cracks started from the bottom corner
to the opposite top corner of the wall. The strain records of the diagonal bars placed
near the foundation showed that the stress in the diagonal bars was low which
300
meant the contribution of the diagonal bars to resist the lateral load was negligible.
Based on readings of the strain gauges placed on the horizontal reinforcement, two
mechanisms were assumed to simulate the behavior of specimen W4 in this stage.
Fig.5.4.4-2 to Fig.5.4.4-4
In the later loading stages, a large horizontal crack crossed the wall along the
interface with the foundation. The diagonal bars located near the foundation of the
wall passing through the horizontal crack took part in resisting lateral load.
Considering the records of the strain gauges on the diagonal bars, as discussed in
previous section, the contribution of the diagonal reinforcement was taken into
account in developing strut-and-tie model of specimen W4 in the later loading stage.
Strut-and-tie model shown in Fig.5.4.4-5 considered the contribution of the
diagonal bars, in which yield strength of the diagonal bars were used according to
the records of strain gauges placed on them. Lateral load was proportionally
distributed to the individual strut-and-tie model shown in Fig.5.4.4-5 and
Fig.5.4.4-6.
The total vertical force:
7;=322.16x2050 = 660.4faV
r 2 = 8 x 2x78.5x384.5 = 483ftV
r=7;+:r 2 =ii43.4&v
The compression depth:
T 1143.4x1000 .__
a= —= = 287mm
O.S5fcb 0.85x33.53x140
Distance of center of gravity of vertical bars in tension to the center of compression
zone:
1800-710-144 = 946ww
301
w. = d since =316xsin41.5° =209Amm
CDB = 0.75/cWs =0.75x33.5 x 140x209.4 = 763.6ftV

The shear force can be resisted by the first mechanism is:
763.6/1.336 = 551.3ftV
w, = d sin a = 316 x sin 62.7° =280.8ww
CDB =0.75/>w s =0.75x33.5 x 140x280.8 = 978iW
978/2.197 = 449.6yUV
Shear capacity of this model:
P = 551.3 + 449.6 = 100 IkN
Safety factor:
1001/584.4 = 1.7
mid-line of
center of gravity of bars in compression zone
tension ~7
710 946 \ \ U4
-/-
Fig.5.4.4-1 Center of Gravity of Vertical Bars in Tension
The lateral load:

P = {(1800-100-144)x660.4 + 946x483}/2540 = 584.4AW
Distance of the centriod of all of the vertical steels to the mid-line of compress
zone:
584.4x2540
\29Smm
(660.4 + 483)
302

Mechanism Member Force Length s Strain Energy Contribution Rate
CD 107 1390 0.55 Ey 81.8 s,,
DA 195.6 1150 £y 224.9 Sy

First 0.53
DE 100 1298 £y 130*,
Z« 436.7*,
Second CA 195.6 2540 £

y 496.8 Sy 0.47
Lateral load resisted by the first mechanism:
/^ = 0.55P = 0.55 x 584.4 = 321 AkN
Lateral load resisted by the second mechanism:
P2 = 0.47P = 0.47 x 584.4 = 263kN
100
"\T
-M7- 707 J37

100 \
Û~
D 700 -<-4 v-
195A * m Nij.* "*

700 \ \
_£' vi ?
795.6
1298 144
Fig.5.4.4-2 Strut-and-Tie Model of W4
303
100 [\
\ 219.7
195.6
195.6
10
1298 144
100
—V
-v
\ \
^~
-JMJ\ 5*8.9 J5.4
\
U~
D 55 \ ,ion \
\
195.<>' \73.5 \ 58.1
48.7
"53—X~^ 1
^v-
A
195.6
1298 144
Fig.5.4.4-4 Combined Strut-and-Tie Model of W4
304
369.2
395.4 395.4\ 541
J6£2\
D 369.2 V -"» —
i
A
-i—
\S
395.4i =5
722.5 327.1 •-A93.2 !
1—
369.2\
1298 144
305
The strain history of the vertical bars was shown in Fig.5.5.1-1 to Fig.5.5.1-5
by strain vs. scan number. It can be seen from the figures that all of the vertical bars
noted by L and M yielded. However, the vertical bars in the upper panel did not
yield in the early loading stages.
100 200 300 400 500 600 700

Scan Number
Fig.5.5.1-1 Relationship of strain Versus Scan Number of Vertical Bar SV1
306
-1500
0 100 200 300 400 500 600 700
Scan Number
Fig.5.5.1-2 Relationship of strain Versus Scan Number of Vertical Bar NV2
-1000
100 200 300 400 500 600 700
Scan Number
307
9000
8000
7000
6000 -
_ 5000
a.
•5 4000 -
2
2000
1000
0 -£*£?
-1000
0 100 200 300 400 500 600 700
Scan Number
Fig.5.5.1-4 Relationship of strain Versus Scan Number of Vertical Bar NV4
6000
5000
-1000
0 100 200 300 400 500 600 700
Scan Number
B. Strain Profiles of Vertical Bars
The strain profiles of the vertical bars in first 5 cycles were shown in
Fig.5.5.1-6 and Fig.5.51-7 for the positive and the negative load, respectively. In the
308
first cycle, for the positive and the negative load, the strain profiles exhibited linear
trends. With increasing load, the depth of the compression zone increased and the
compression strain at the compressive extreme fiber also increased. In the fifth
cycle, the linear trends of the strain profiles were totally lost.
4000
-3000
200 400 600 800 1000 1200 1400 1600 1800 2000
Fig.5.5.1-6 Strain Profiles of vertical Bar for Positive Load
4000
3000
2000
-3000
0 200 400 600 800 1000 1200 1400 1600 1800 2000
Fig.5.5.1-7 Strain Profiles of vertical Bar for Negative Load
309
Hysteresis loops of the strain of the diagonal bars near the foundation vs. load
were shown in Fig.5.5.1-8 and Fig.5.5.1-10. The diagonal bars near the foundation
exhibited some strain energy dissipation capacity. Relationships of diagonal
reinforcement near the base versus scan number were shown in Fig.5.5.1-9 and
Fig.5.5.1-11. From Fig.5.5.1-9 and Fig.5.5.1-11 it can be seen that the diagonal bar
SD1 yielded at the scan number 310 and ND4 yielded at the scan number 348,
which meant that before reaching a ductility of 3, the diagonal reinforcement near
the foundation did not yield and its contribution to lateral load resistance was
negligible. After ductility reached 3, the diagonal bars near the foundation yielded.
The characteristics of the strain histories of the diagonal bars in the horizontal
connection were shown in Fig.5.5.1-12 to Fig.5.5.1-19. Compared with diagonal
bars near the foundation, the diagonal bars in the horizontal connection yielded
earlier compared to the diagonal bars near the foundation. And from the features of
the hysteresis loops of all of the diagonal bars, the diagonal bars in the horizontal
connection exhibited much lower energy dissipation capacity.
310
-800
-500 0 500 1000 1500 2000 2500 3000 3500 4000
Strain(fts)
Fig.5.5.1-8 Strain History of SD1
100 200 300 400 500 600 700

Scan Number
Fig.5.5.1-9 Relationship of Strain versus Scan Number of SD1
311
I
13
-800
-500 0 500 1000 1500 2000 2500 3000 3500 4000 4500
Strain(fis)
Fig.5.5.1-10 Strain History of ND4
-500
100 200 300 400 500 600 700
Scan Number
Fig.5.5.1-11 Relationship of Strain versus Scan Number of ND4
312
800
600
400
200
S
as
0 f ^~
J^ZZ^
-200 %w.-,<&'
• V?
^§-
-400 M\l ZVF-
-%?+ 3li*
^§
NDLi^
\
-600
yl 'i
-800
-500 500 1500 2500 3500 4500 5500
Strainffie)
Fig.5.5.1-12 Strain History of NDL2
6000
5000 -
4000
^ 3000
•5
| 2000
1000
-1000
100 200 300 400 500 600 700
Scan Number
Fig.5.5.1-13 Relationship of Strain versus Scan Number of NDL2
313
800
600
400
l\ 1 ft \
IV^V
AM\
§ ^ -- r i S !
3g_ 2?
TO\ ^ ^ | " i l ^-S^: ~^^
200 m
u m >*
vW^^V s? ^ " - > ^ SDL6
t 0 ""V^. JSWji^Js.
<5
-J
-200
-400 V^^-S\??S»S<
-600
-500 500 1000 1500 2000 2500
-800
Strainffis)
Fig.5.5.1-14 Strain History of SDL6
3000
I
•5
33 1000
-500
100 200 300 400 500 600 700
Scan Number
Fig.5.5.1-15 Relationship of Strain versus Scan Number of SDL6
314
800
600 -
1
4tHJ -*£
400 ^s
>"
«\ SDUS^ ^ 2 ' S&
2?
200
I« 0
SS
-J
-200
-400
-600
-800
-1000 -500 500 1000 1500 2000 2500 3000 3500
Strain(fiE)
Fig.5.5.1-16 Strain History of S D U 5
100 200 300 400 500 600 700
Scan Number
Fig.5.5.1-17 Relationship of Strain versus Scan Number of SDU5
315
800
600
400
200
fa •
-200
W Biz .::^2t
-400 ss~"l^
.^g-i&
-600
1 ^ ^%
-800
-1000 1000 2000 3000 4000 5000 6000
Strain(fts)
Fig.5.5.1-18 Strain History of SDU8
-1000
100 200 300 400 500 600 700
Scan Number
Fig.5.5.1-19 Relationship of Strain versus Scan Number of SDU8
C. Strain History of Horizontal Bars
The strain histories of the horizontal bars were shown in Fig.5.5.1-20 to

Fig.5.5.1-24. It can be seen from Fig.5.5.1-20 that the horizontal reinforcement in
316
the range of the diagonal bars near the foundation did not yield, which meant that at
any section in the range of the diagonal bars near the foundation, shear force was
resisted by the horizontal reinforcement and the diagonal bars simultaneously.
Fig.5.5.1-21 and Fig.5.5.1-22 indicated that the horizontal bars at this height
level exerted their strength completely. Fig.5.5.1-23 and Fig.5.5.1-24 show that the
horizontal bars in the upper wall panel did not yield, which implied that shear
resistant requirement was smaller in the upper wall panel in precast wall compared
to that in the integral prototype. With increasing displacement level, the critical
horizontal section was just above the diagonal bars near the foundation where
fracture of the horizontal reinforcement occurred.
3000
2500
-NH11-2
NHI1-3
2000
f 1500 -
•5
| 1000
500
-500
100 200 300 400 500 600 700
Scan Number
317
5000
-1000
100 200 300 400 500 600 700
Scan Number
5000
-1000
100 200 300 400 500 600 700
Scan Number
318
5000
-1000
100 200 300 400 500 600 700
Scan Number
5000
-1000
100 200 300 400 500 600 700
Scan Number
5.5.2 Load Transfer
In the early loading stages, the diagonal cracks from the bottom corner to the
319
opposite top comer were the main features of cracking patterns as shown in
Fig.4.2.4-5. The diagonal concrete strut played an important role in resisting the
lateral load. With increasing load, the diagonal cracks from the lower corner to the
opposite higher corner of each wall panel appeared in the upper and the lower
panels, which could be seemed as a sign of the changes of the internal load paths. At
the same time the upper and the lower wall panels performed as independent with
special feature of cracking patterns of squat walls.
Because no evidence of deformation of the horizontal connection was observed,

the lateral load could be transferred successfully through the horizontal connection
to the lower wall panel. Referring to cracking patterns of the wall and the strain in
the horizontal reinforcement at different stages, it can be assumed that in the early
stages, before reaching a ductility of 2, the main lateral resistant mechanism was the
diagonal concrete strut from the lower corner to the opposite upper corner of W5 as
shown in Fig.5.5.2-1. After ductility passed 2, the upper wall panel acted as a squat
wall with a low aspect ratio, a diagonal concrete strut starting from the center of top
beam appeared. The performance of the upper wall panel influenced the load paths
in the whole specimen. As discussed in previous sections, the connection zone can
be seen as a continuum through which internal forces in the upper wall panel can be
transferred without deviation.
320
Fig.5.5.2-1 Cracking direction in lower corner
5.5.3 Concrete Crushing and Breaking of Horizontal Bar
When ductility reached 3, the width of concrete bands between diagonal cracks
was about 100mm. These diagonal cracks were evenly distributed in the upper and
lower wall panels. With increasing displacement levels, two wide cracks just above
the diagonal bars near the base intersected to form a "V" shape with about 120 deg.
At the same time, many short cracks appeared along the inner boundaries of the
I-Beams in the lower wall panel, which intersected with the flexural and diagonal
cracks. These short cracks were not observed in the upper wall panel. In late loading
cycles, the diagonal cracks aligned with diagonal bars near the base became wider
and simultaneously horizontal reinforcement crossing through these cracks yielded.
Diagonal cracks from opposite direction intersected these wide diagonal cracks in
lower wall panel. The short cracks along the inner boundaries of I-Beams cut
321
diagonal cracks and flexural cracks at same positions. The density of the cracks
increased. Under cyclic load, these small concrete pieced fell down, which reduced
nodal area of concrete struts connecting these locations. In the loading course, it can
be observed that the diagonal concrete bands resisting the lateral load be sheared
one by one at the height of the top end level of the diagonal reinforcement near the
foundation. For example, when a positive load was applied, the diagonal concrete
bands at this level were sheared from right to left one by one. It can be seen as the
course of the failure of the section. Two triangular zones formed at near the up ends
of diagonal bars at height level 650mm from base beam. It meant that the
percentage of the contribution of the concrete struts to the lateral load dropped and
the requirements of the horizontal reinforcement increased. Finally two horizontal
bars were fractured with a sudden failure of the specimen.
Fig.5.5.3-1 Features of Diagonal Cracks in Lower Wall Panel
322
Fig.5.5.3-2 Fracture of Horizontal Bar
Fig.5.5.3-3 Failure Mode (Rear View)
323
Specimen W5 performed a similar behavior with specimen W4 in the early

loading stages. The diagonal reinforcement placed near the foundation did not yield,
which implied that the contribution of diagonal bars to lateral load was insignificant.
Therefore strut-and-tie models shown in Fig.5.5.4-1 and Fig5.5.4-2 represented two
individual mechanisms to resist lateral load. Based on theory of minimum strain
energy, a combined strut-and-tie model for specimen W5 is shown in Fig.5.5.5-3.
After reaching a ductility of 2, cracking patterns indicated that the internal load
paths through which the lateral load was transferred to the foundation. The diagonal
cracks from the lower corner to the opposite upper corner in each wall panel
occurred, which indicated that two wall panel behaved as two independent squat
walls. As shown in Fig.5.2.3-2, due to the low aspect ratio of the upper wall panel,
two diagonal concrete struts were assumed. According to the internal load paths
shown in Fig.5.2.3-2, the strut-and-tie model shown in Fig.5.5.4-4 was developed.
The contribution of the diagonal reinforcement located near the foundation was
considered in Fig.5.5.4-5. In the whole course of loading, evidence of deformation
and damage of the horizontal connection was not observed. Therefore, a solid
continuum was assumed to represent the connection zone which was defined in
previous sections. In accordance to the theory of minimum strain energy, the total
lateral load was distributed to the two load-resistant mechanisms as shown in
Fig.5.5.4-6 and Fig.5.5.4-7. The combined strut-and-tie model was shown in
Fig.5.5.4-8 which combined the strut-and-tie models shown in Fig.5.5.4-6 and
Fig.5.5.4-7. In the model shown in Fig.5.5.4-8 , concrete strut BF was not but GL
was the main component which transferred the lateral load to foundation. The
concrete strut HI took an important role in transferring load from upper wall panel
to the lower panel, which caused yielding of the horizontal tie KI. When
overstrength of the vertical steels was considered, requirement of tie KI was
324
increased. Under cyclic loads, due to concrete spalling at upper ends of the diagonal
reinforcement near the foundation, the nodal area of the concrete in compression at
point I was decreased, which enlarged obligation of the horizontal tie KJ to resist
the lateral load. At this moment, KI had come into plastic deformation and lost its
ability to bear the additional lateral load. As a result, a fracture of the horizontal
reinforcement occurred as observed in experiment.
7;= 2050 x 322.16 = 660 AkN
T2 = 384.5 x 2 x 5 x 78.5 = 301MN
T = TX+T2= 962.2kN
d =^ = 962.2x1000 =u%mm
0.85 fcb 0.85x34.02x140
ws =d sin or = 316 x sin 41.5° = 209 Amm
Distance of center of gravity of vertical bars in tension to the center of compression

zone:
1800-710-144 = 946mm
ws =d sin or = 316 x sin 41.5" = 209.4mm
CDB =0.15f'cbws = 0.75x33.5x140x209.4 = 763.6kN
CD 87.9 1390 0.55 £,. S72ey

>
DA 160.6 1150 £
y
184.7 £
First 0.501
DE 100 1581 £
y
158.1 e
T,F,I, 410 ey
Second CA 160.6 2540 £

y
408 ey 0.499
325
100
87.9 87.9Y133
1 0 0 <>
^
D
100 kt \
X
160.6 72.7 ^.123.6

H 87i9
100 N„
160.6
1581 \l!9
I
Fig.5.5.4-1 Strut-and-Tie Model
100
D
160.6 \189.2
^
100 \
160.6
1581
JJW
326
100
^H
124.2 \44 ^666
\ 50.1 \
3-
D
50.>V_£N
^ s
160.6 36.4^-61.9 \ 44
1 X ^
Fig.5.5.4-3 Combined Strut-and-Tie Model
327

AG' 160.7 800 £

y
128.6 Sy
CD 133.6 1290 £y 172.3 ey
BE 100 1581 £
y 158 Sy
First 0.48
BG' 86.8 650 0.5 Sv 2%2ey
y
GT 50 1600 0.5 £y 40 ev
y
Z^ 527.2 ey
AG 181.4 650 £
y 117.9*,
GN 118 850 0.65 ey 65.2 £y
NH 83.9 1040 0.46 £y 40.1 ey
GK 78 1450 0.78 Sy 88.2^

Second 0.52
LK 31 919 £
y 28.5^
KI 100 131 £
y 13.1 ey
NM 100 1250 £
y 125 £
E« 478 6y
328
I 1100 800 I 781 J, 1119~
7250
I I 100 800 J, 650 I I I iff

\l31
329
J. 1100 800 j, 781 J. I iff
1250
S03.9
I I 100 800 I 650 J, t LTTiT

\'iJ7
330
1250
303.9
Specimen W6 is one of the precast counterparts of specimen W4. In specimen

W6, two shear connectors were used to transfer the horizontal force in the
connection.
5.6.1 Strain History
A. Strain Profiles of Vertical Bars
The strain profiles of the vertical bars for the positive and negative loads were
shown in Fig.5.6.1-1 and Fig.5.6.1-2. It can be seen from the figures that in the first
four cycles the distribution of the strain in the vertical bars along the
wall-foundation interface exhibited almost linear feature, especially in first cycle.
331
The increased displacement level caused the specimen to enter the plastic
deformation stage, as nonlinear feature of the strain profiles was observed. With
increasing lateral load, after cracking occurred in the wall, depth of compression
zone increased from Fig. 5.6.1-1 and Fig. 5.6.1-2.
200 400 600 800 1000 1200 1400 1600 1800 2000
Fig.5.6.1-1 Strain Profiles of Vertical Bars for Positive Load
200 400 600 800 1000 1200 1400 1600 1800 2000
Fig.5.6.1-2 Strain Profiles of Vertical Bars for Negative Load
332
B. Strain History of Diagonal Bars
When reaching a ductility of 2, the diagonal reinforcement yielded as shown in

Fig.5.6.1-3 and Fig.5.6.1-4. In the elastic loading cycles, strain readings of the
diagonal bars near the foundation showed that stress in the diagonal bars was low.
Therefore the contribution of the diagonal reinforcement to the lateral load
resistance was insignificant.
800
600 /
400
200 "•*
s ,tt
\ 1 il Usi>
M* H SD3
I o
a
k% \ ^ , ,
o
M
-200
-400 ^~^^^^ \
-600
-800-500 0 500 1000 1500 2000 2500 3000

Strainffis)
Fig.5.6.1-3 Strain History of Diagonal Bar SD3
333
800
I
600
400 \\
\
Is N
200 h Urv *, __ ^ 2
I .
a
kfeJL ^ - ^ 11 11 1 Sfc=!=J=*=B
£
-200
-400
-600
-800
-500 0 500 1000 1500 2000 2500 3000
Strainffte)
Fig.5.6.1-4 Strain History of Diagonal Bar SD4
5.6.2 Strut-and-Model
In the early loading stages, before reaching a ductility of 1, specimen W6

behaved similarly to specimen W5. The cracking patterns were also similar as those
of specimen W5. The diagonal cracks from the bottom corner of the wall to the top
corner. Strain readings of the diagonal reinforcement near the foundation showed
that the diagonal bars did not yield. Similarly to the diagonal reinforcement in
specimen W5, the contribution of the diagonal reinforcement near the foundation to
the lateral load was not taken into account. Therefore the strut-and-tie model shown
in Fig.5.5.4-3 represented the internal forces in the wall in this stage.
5.6.3 Failure Mode
When the displacement level was increased to ductility 2, connection failure

occurred with shear failure of the bolts in flexural connectors at right end of the
334
horizontal connection as shown in Fig.5.6.3-1 and Fig.5.6.3-2. The flexural

connection caused fracture of the diagonal bars welded on shear connection. This
might be due to construction defect.
Fig.5.6.3-1 Connection Failure in Specimen W6
Fig.5.6.3-2 Shear Failure of Bolt in Flexural Connection
5.7 Summary
Six specimens, two integral prototypes and four precast counterparts, were
335
tested to failure at Nanyang Technological University. Records of strain gauges

placed on horizontal, vertical and diagonal reinforcement and steel sections were
analyzed. Considering the cracking patterns and records of strain gauges,
strut-and-tie models were developed to represent the internal force paths and
explain failure causes.
1. In the early loading stages, the diagonal reinforcement near the

wall-foundation interface did not reach their yielding strength in all
specimens. In developing strut-and-tie models for these stages, the
contribution of the diagonal bars to lateral load acting on the top beam is
insignificant, therefore, no strut and tie is used to represent them.
2. In later loading stages, all of the diagonal bars placed near the
wall-foundation interface yielded. Their contribution to lateral load should
be taken into account.
3. In testing integral walls, Wl and W4, in later loading cycles, a wide crack
crossed the wall along the wall-foundation interface. However, such a crack
was not observed in the precast walls.
4. In early loading stages, no evidence of deformation and damage was
observed in the horizontal connections of the precast walls. The cracking
patterns of the precast walls were similar to those of their prototypes.
5. When the displacement level reached a ductility factor of 2, the cracking
patterns in the precast walls differed from those of their prototypes.
Diagonal cracks started from the lower corner to the opposite upper corner
in each wall panel, which made the two wall panels act as independent squat
walls. Due to the low aspect ratio of the upper wall panel, two diagonal
concrete struts were assumed. And force paths in the upper wall panel
influenced force paths in the lower wall panel.
6. Except for the horizontal connection in specimen W6, for which the failure
336
may have been due to construction defect, the other horizontal connection
did not fail during the course of testing. It means that the horizontal
connections in this research were alike to successfully transfer the shear
between two wall panels.
7. Strut-and-tie models were developed for different loading stages. These
strut-and-tie models were used to explain the internal force paths in these
specimens and their failure mode and can be used to direct design.
337

Thesis Precast Shear Wall With Horizontal Connection

Uploaded by

Copyright:

Available Formats

Thesis Precast Shear Wall With Horizontal Connection

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Thesis Precast Shear Wall With Horizontal Connection

Uploaded by

Copyright:

Available Formats

ATTENTION: The Singapore Copyright Act applies to the use of this document.

Nanyang Technological University Library

PRECAST SHEAR WALL WITH

School of Civil a n d Environmental Engineering

A thesis submitted to the Nanyang Technological University

Construction and Management at Nanyang Technological University under supervisor

A/P Ting Seng Kiong.

invaluable comments and extensive discussions and suggestions throughout the

during all the stages of the work.

supported me throughout my education.

TABLE OF CONTENSTS iii

LIST OF FIGURES xiii

LIST OF SYMBOLS xxix

1.1 Problem Definition 1

CHAPTER 2 LITERATURE REVIEW 6

2.5 Strut-and-tie Model 24

2.5.1 History of Strut-and-Tie models 25

2.5.2 B Region and D Regions 31

2.5.3 Design Procedure of Strut-And-Tie Model 33

2.5.4 Method for Developing Strut-and-Tie Models 36

2.5.5 Model Optimization 39

2.5.6 Dimensioning the Struts, Ties and Nodes 40

2.5.7 Feasible Inclined Angle 0 43

2.5.8 Limiting Compressive Stress in Strut 43

2.5.9 Summary on Strut-and-Tie Model 46

CHAPTER 3 EXPERIMENT DESIGN 48

3.1.1 General Description of Wall Specimens 49

3.1.2 Material Properties 51

3.2 Testing Program 72

3.2.1 Test Set-up 72

3.2.3 Load history 87

CHAPTER 4 EXPERIMENTAL RESULTS 92

4.1.1 Load and Displacement Histories 92

4.2 Experimental Results 93

4.2.1 Load-Displacement Response 99

4.2.2 Backbone Envelop of Load-Displacement Response 104

4.2.3 Stiffness 106

CHAPTER 5 DISCUSSION OF TEST RESULTS 189

5.4 Behavior of Specimen W4 289

CHAPTER 6 FINITE ELEMENT ANALYSIS OF EXPERIMENT

6.4.1 2-D Model 366

CHAPTER 7 CONCLUSIONS AND RECOMMENDATIONS 401

In order to investigate performance of precast walls with horizontal connection, six

CHAPTER 4 EXPERIMENTAL RESULTS

CHAPTER 5 DISCUSSION OF TEST RESULTS

Fig.2.5.1-1 Compressive stress-strain curve of concrete 28

Fig. 3.1.2-5 Stress Versus Strain Relationship of 1-Beam 54

Fig. 3.2.3-6 Load history of specimen W5 90

Fig. 4.1.1-1 Horizontal Displacement History of Specimen W1 93

Fig. 4.2.4-5 Cracking Patterns of Specimen W5 117

Fig. 4.2.6-9 Local Vertical Elongations of W2 (cont.) 138

Fig 4.3.2-3 Macro Models by Linde (1993) 164

Fig. 5.1.1-1 Strain History of Vertical Bar SV1 191

Fig. 5.1.1-5 Strain History of Vertical Bar SV5 193

Fig.5.1.1-35 Strain History of Horizontal Bar SH9 212

Fig.5.1.2-1 Diagonal Crack in Left Corner at 0.75^ 215

Fig.5.1.2-2 Diagonal Crack in Left Corner at DF=3 216

Fig. 5.3.1-1 Strain History of Vertical Bar SV1 271

Fig.5.4.1-2 Strain History of Vertical Bars NV2 290

J( l/f I IK"' C-d