Tension Field

STATE OF CALIFORNIA • DEPARTMENT OF TRANSPORTATION
ADA Notice
For individuals with sensory disabilities, this document is available in alternate
TECHNICAL REPORT DOCUMENTATION PAGE formats. For alternate format information, contact the Forms Management Unit
TR0003 (REV 10/98) at (916) 445-1233, TTY 711, or write to Records and Forms Management,
1120 N Street, MS-89, Sacramento, CA 95814.
1. REPORT NUMBER 2. GOVERNMENT ASSOCIATION NUMBER 3. RECIPIENT'S CATALOG NUMBER
CA15-2316
4. TITLE AND SUBTITLE 5. REPORT DATE
Shear Resistance of End Panels in Steel and Steel-Concrete Composite Girders
March 2015
6. PERFORMING ORGANIZATION CODE
7. AUTHOR 8. PERFORMING ORGANIZATION REPORT NO.
Dong-Won Kim and Chia-Ming Uang SSRP-15/02

9. PERFORMING ORGANIZATION NAME AND ADDRESS 10. WORK UNIT NUMBER
Department of Structural Engineering

Jacobs School of Engineering
11. CONTRACT OR GRANT NUMBER
University of California, San Diego
65A0473
La Jolla, California 92093-0085
12. SPONSORING AGENCY AND ADDRESS 13. TYPE OF REPORT AND PERIOD COVERED
California Department of Transportation Final Report
Division of Research, Innovation, and System Information June 1, 2012 - March 31, 2015
P.O. Box 942873 14. SPONSORING AGENCY CODE
Sacramento, CA 94273-0001
15. SUPPLEMENTARY NOTES
Prepared in cooperation with the State of California Department of Transportation
16. ABSTRACT
Plate girders, usually characterized by having very deep sections, have been widely used for long-span structures and bridges. US design
provisions (AASHTO Specifications for bridge design and AISC Specifications for building design) allow the designer to include the
contribution of tension-field action after web buckling in calculating the ultimate shear strength of interior panels, but not exterior panels,
because it is believed that no effective anchor mechanism exists to resist the horizontal component of the tension-field force. This has had a
negative impact on the evaluation and load rating of existing bridges. This conservatism often requires unnecessary rehabilitation of existing
steel girder bridges which have provided satisfactory service in the past half-century. Testing of two large-size steel plate girders and two
steel-concrete composite plate girders was carried out in Phase 1 of this research to investigate the shear resistance of end panels. Test results
demonstrated a much higher shear resistance than that predicted by code provisions in the steel girders. Tension-field action was observed in
all specimens. The concrete slab in the composite specimens also contributed to the shear resistance, although to a lesser extent. Nonlinear
finite element analyses were conducted to correlate the test results. Results from a parametric study strongly supported the behaviors observed
from testing and confirmed the existence of partial tension-field action in the end panels. Based on the failure mode observed from both
WHsting and finite element simulation, an analytical model was developed to simulate the collapse mechanism. Plastic analysis was used to derive
a predictive shear strength equation. This equation is similar in format to that used in AASHTO Specifications for interior panels but
includes a parameter α to account for the contribution of partial tension-field action. A resistance factor for the proposed shear strength equatioQ
ZDVDOVRGHULYHG%DVHGRQDVWUXWDQGWLHPRGHOWKHFRQWULEXWLRQIURPWKHFRQFUHWHVODEZDVDOVRSUHVHQWHG)RU3KDVHWHVWSURJUDPDQRWKHU
WZRVWHHOSODWHJLUGHUVDQGWZRVWHHOFRQFUHWHFRPSRVLWHSODWHJLUGHUVZLWKODUJHUZLGWKWRGHSWKUDWLRVDQGDVPDOOGHSWKWRWKLFNQHVVUDWLRZHUH
IDEULFDWHGDQGWHVWHG7HVWLQJVKRZHGWKHVDPHIDLOXUHPRGHDVWKDWREVHUYHGIURP3KDVHWHVWSURJUDPDQGFRQILUPHGWKHDFFXUDF\RIWKH
SURSRVHGHTXDWLRQ+RZHYHUWKHFRQFUHWHFRPSRQHQWRIWKHVKHDUHTXDWLRQGHYHORSHGLQ3KDVHXQGHUHVWLPDWHVWKHVKHDUUHVLVWDQFHRI3KDVH
FRPSRVLWHVSHFLPHQVWKDWKDGFORVHO\VSDFHGVKHDUFRQQHFWRUV$PRGLILFDWLRQZDVPDGHWRWKHSURSRVHGHTXDWLRQWRDFFRXQWIRUWKLVVSDFLQJ
HIIHFW5HKDELOLWDWLRQRIHQGSDQHOVE\XVLQJDFRPPRQVFKHPHILUVWSURSRVHGE\%DVOHUDQGKDVEHHQDGRSWHGLQ(XURFRGHZDVIRXQG
XQFRQVHUYDWLYH7ZRUHKDELOLWDWLRQVFKHPHVZHUHSURSRVHGDQGWKHHIIHFWLYHQHVVZDVYHULILHGE\ILQLWHHOHPHQWVLPXODWLRQ$GGLWLRQDOO\
WKHHIIHFWVRIORQJLWXGLQDOVWLIIHQHUVVPDOOIODQJHDUHDVODUJHZLGWKGHSWKUDWLRVQRQFRPSDFWZHEDQGXQHTXDWOWRSDQGERWWRPIODQJHV
ZHUHLQYHVWLJDWHGE\ILQLWHHOHPHQWDQDO\VLVDQGWKHUHVXOWVFRQILUPHGWKDWWKHSURSRVHGHTXDWLRQDOVRJLYHVDUHDVRQDEOHFRUUHODWLRQZLWK
WKHDQDO\VLVUHVXOWV
KEY WORDS
17. 18. DISTRIBUTION STATEMENT
No restrictions. This document is available to the public through the
Shear Strength, End Panels, Steel Girder, Composite Girder,
Tension-Field Action, Post-Buckling Strength, Rehabilitation National Technical Information Service, Springfield, VA, 22161.
19. SECURITY CLASSIFICATION (of this report) 20. NUMBER OF PAGES 21. COST OF REPORT CHARGED
Unclassified 396
Reproduction of completed page authorized.
DISCLAIMER STATEMENT
This document is disseminated in the interest of information exchange. The contents of this
report reflect the views of the authors who are responsible for the facts and accuracy of the data
presented herein. The contents do not necessarily reflect the official views or policies of the State
of California or the Federal Highway Administration. This publication does not constitute a
standard, specification or regulation. This report does not constitute an endorsement by the
Department of any product described herein.
For individuals with sensory disabilities, this document is available in alternate formats. For
information, call (916) 654-8899, TTY 711, or write to California Department of Transportation,
Division of Research, Innovation and System Information, MS-83, P.O. Box 942873,
Sacramento, CA 94273-0001.
STRUCTURAL SYSTEMS
RESEARCH PROJECT
Report No.
SSRP-15/02 Shear Resistance of End Panels in Steel and
Steel-Concrete Composite Plate Girders
by
Dong-Won Kim
Chia-Ming Uang
Final Report to California Department of Transportation

March 2015
Structural Systems Research Project
Report No. SSRP-15/02
SHEAR RESISTANCE OF END PANELS IN STEEL AND

STEEL-CONCRETE COMPOSITE PLATE GIRDERS
by
Dong-Won Kim
Postdoctoral Researcher
Chia-Ming Uang
Professor
Final Report to California Department of Transportation

March 2015
DISCLAIMER
This document is disseminated in the interest of information exchange. The

contents of this report reflect the views of the authors who are responsible for the facts
and accuracy of the data presented herein. The contents do not necessarily reflect the
official views or policies of the State of California or the Federal Highway
Administration. This publication does not constitute a standard, specification or
regulation. This report does not constitute an endorsement by the California Department
of Transportation of any product described herein.
For individuals with sensory disabilities, this document is available in Braille,
large print, audiocassette, or compact disk. To obtain a copy of this document in one of
these alternate formats, please contact: the Division of Research and Innovation, MS-83,
California Department of Transportation, P.O. Box 942873, Sacramento, CA 94273-0001.
i
ACKNOWLEDGEMENTS
The California Department of Transportation (Caltrans) funded this work under

Contract #65A0473. We would like to thank Drs. Charles Sikorsky (project manager),
George Huang, and Lian Duan at Caltrans for their guidance and advice.
The testing was conducted in the Charles Lee Powell Structures Laboratories at
the University of California, San Diego. Technical assistance from Dr. Christopher
Latham and Paul Greco is much appreciated.
ii
ABSTRACT
Plate girders, usually characterized by having very deep sections, have been
widely used for long-span structures and bridges. US design provisions (AASHTO
Specifications for bridge design and AISC Specifications for building design) allow the
designer to include the contribution of tension-field action after web buckling in
calculating the ultimate shear strength of interior panels, but not exterior panels, because
it is believed that no effective anchor mechanism exists to resist the horizontal
component of the tension-field force. This has had a negative impact on the evaluation
and load rating of existing bridges. This conservatism often requires unnecessary
rehabilitation of existing steel girder bridges which have provided satisfactory service in
the past half-century.
Testing of two large-size steel plate girders and two steel-concrete composite
plate girders was carried out in Phase 1 of this research to investigate the shear resistance
of end panels. Test results demonstrated a much higher shear resistance than that
predicted by code provisions in the steel girders. Tension-field action was observed in
all specimens. The concrete slab in the composite specimens also contributed to the
shear resistance, although to a lesser extent.
Nonlinear finite element analyses were conducted to correlate the test results.
Results from a parametric study strongly supported the behaviors observed from testing
and confirmed the existence of partial tension-field action in the end panels. Based on
the failure mode observed from both testing and finite element simulation, an analytical
model was developed to simulate the collapse mechanism. Plastic analysis was used to
derive a predictive shear strength equation. This equation is similar in format to that
used in AASHTO Specifications for interior panels but includes a parameter α to account
for the contribution of partial tension-field action. A resistance factor for the proposed
shear strength equation was also derived. Based on a strut-and-tie model, the
contribution from the concrete slab was also presented.
For Phase 2 test program, another two steel plate girders and two steel-concrete
composite plate girders with larger width-to-depth ratios and a small depth-to-thickness
ratio were fabricated and tested. Testing showed the same failure mode as that observed
iii
from Phase 1 test program and confirmed the accuracy of the proposed equation.
However, the concrete component of the shear equation developed in Phase 1
underestimates the shear resistance of Phase 2 composite specimens that had closely
spaced shear connectors. A modification was made to the proposed equation to account
for this spacing effect.
Rehabilitation of end panels by using a common scheme, first proposed by Basler
and has been adopted in Eurocode 3, was found unconservative. Two rehabilitation
schemes were proposed and the effectiveness was verified by finite element simulation.
Additionally, the effects of longitudinal stiffeners, small flange areas, large width-depth
ratios, noncompact web, and unequatl top and bottom flanges were investigated by finite
element analysis and the results confirmed that the proposed equation also gives a
reasonable correlation with the analysis results.
iv
TABLE OF CONTENTS
DISCLAIMER .................................................................................................................i
ACKNOWLEDGEMENTS............................................................................................ii
ABSTRACT...................................................................................................................iii
TABLE OF CONTENTS................................................................................................ v
LIST OF FIGURES .......................................................................................................ix
LIST OF TABLES.....................................................................................................xviii
1. INTRODUCTION .................................................................................................. 1
1.1 Plate Girders.................................................................................................... 1
1.2 Shear Load-Carrying Mechanism of Stiffened Steel Plate Girder.................. 1
1.3 Tension-Field Action in Interior Panels.......................................................... 2
1.3.1 General ................................................................................................ 2
1.3.2 Basler’s Theory ................................................................................... 2
1.3.3 Brief Review of Various Tension-Field Theories ............................... 5
1.4 US Plate Girder Shear Design Practice........................................................... 7
1.4.1 General ................................................................................................ 7
1.4.2 Interior Panels...................................................................................... 7
1.5 End Panels....................................................................................................... 9
1.6 Statement of Problem...................................................................................... 9
1.7 Shear Resistance in End Panels .................................................................... 10
1.8 Steel-Concrete Composite Plate Girder ........................................................ 12
1.9 Scope and Objectives.................................................................................... 14
1.10 Organization of Report ................................................................................. 14
2. TEST PROGRAM (PHASE 1)............................................................................. 26

2.1 General.......................................................................................................... 26
2.2 Configuration and Fabrication of Test Specimens........................................ 26
2.2.1 Steel Girder Specimens (SG1 and SG2)............................................ 26
2.2.2 Steel-Concrete Composite Girder Specimens (CG1 and CG2)......... 27
2.3 Material Properties........................................................................................ 27
2.3.1 Structural Steel Plates........................................................................ 27
2.3.2 Reinforcement and Concrete ............................................................. 27
2.3.3 Shear Studs ........................................................................................ 28
2.4 Initial Geometrical Imperfections ................................................................. 28
v
2.5 Test Setup and Loading Procedure ............................................................... 28
2.6 Instrumentation ............................................................................................. 29
3. TEST RESULTS (PHASE 1) ............................................................................... 47

3.1 Specimen SG1............................................................................................... 47
3.2 Specimen SG2............................................................................................... 48
3.3 Specimen CG1 .............................................................................................. 48
3.4 Specimen CG2 .............................................................................................. 49
3.5 Recorded Response....................................................................................... 50
3.6 Comparison of Test Results .......................................................................... 51
3.6.1 Failure Mechanism ............................................................................ 51
3.6.2 Comparison with AASHTO Shear Strength ..................................... 51
3.6.3 Contribution of Concrete Slab........................................................... 52
4. NONLINEAR FINITE ELEMENT ANALYSIS OF TEST SPECIMENS

(PHASE 1) ............................................................................................................ 84
4.1 General.......................................................................................................... 84
4.2 Steel Plate Girder Specimens (SG1 and SG2) .............................................. 84
4.2.1 Finite Element Models ...................................................................... 84
4.2.2 Material Stress-Strain Characteristics ............................................... 84
4.2.3 Residual Stresses and Geometric Imperfections ............................... 85
4.2.4 Correlation between Test Results and FEA Predictions ................... 85
4.2.5 Simulated Strains............................................................................... 86
4.3 Steel-Concrete Composite Plate Girder Specimens (CG1 and CG2) ........... 87
4.3.1 Finite Element Models ...................................................................... 87
4.3.2 Material Stress-Strain Characteristics ............................................... 87
4.3.3 Correlation between Test Results and FEA Predictions ................... 89
4.3.4 Simulated Strains............................................................................... 90
5. FACTORS AFFECTING SHEAR RESISTANCE OF END PANELS: A

PARAMETRIC STUDY (PHASE 1) ................................................................. 110
5.1 General........................................................................................................ 110
5.2 Steel Plate Girders....................................................................................... 110
5.2.1 Introduction ..................................................................................... 110
5.2.2 Effect of Flange and Bearing Stiffener Thicknesses ....................... 111
5.2.3 Effect of Panel Depth-to-Thickness and Width-to-Depth Ratios.... 111
5.2.4 Effect of Web End Extension, e ...................................................... 112
5.3 Steel-Concrete Composite Plate Girders .................................................... 113
5.3.1 Introduction ..................................................................................... 113
5.3.2 Effect of Concrete Slab Width ........................................................ 113
vi
5.3.3 Effect of Shear Stud Location ......................................................... 113
6. SHEAR RESISTANCE OF END PANELS: DERIVATION OF PREDICTIVE

EQUATIONS...................................................................................................... 128
6.1 General........................................................................................................ 128
6.2 Failure Modes ............................................................................................. 129
6.3 Tension Band Width, s................................................................................ 130
6.4 Post-Buckling Strength: A Plastic Mechanism Approach .......................... 130
6.5 Plastic Hinge Locations in Boundary Members ......................................... 133
6.6 Simplification.............................................................................................. 135
6.7 Effective Depth of End Panel Boundary Members..................................... 138
6.8 Validation of Shear Strength Formula ........................................................ 140
6.9 Shear Resistance of Steel-Concrete Composite Girder End Panel ............. 142
7. REHABILITATION OF END PANELS............................................................ 174

7.1 General........................................................................................................ 174
7.2 Basler’s Scheme.......................................................................................... 174
7.3 Eurocode 3 Design Provision...................................................................... 176
7.4 Proposed Scheme 1 ..................................................................................... 178
7.5 Proposed Scheme 2 ..................................................................................... 182
8. TEST PROGRAM (PHASE 2)........................................................................... 207

8.1 Test Program............................................................................................... 207
8.1.1 General ............................................................................................ 207
8.1.2 Configuration and Fabrication of Test Specimensl......................... 207
8.1.3 Material Properties .......................................................................... 207
8.1.4 Initial Geometrical Imperfections ................................................... 208
8.1.5 Test Setup, Loading Procedure, and Instrumentation ..................... 208
8.2 Test Results................................................................................................. 209
8.2.1 Specimen SG3 ................................................................................. 209
8.2.2 Specimen SG4 ................................................................................. 209
8.2.3 Specimen CG3................................................................................. 210
8.2.4 Specimen CG4................................................................................. 211
8.2.5 Recorded Local Responses.............................................................. 211
8.2.6 Comparison of Test Results ............................................................ 212
9. ANALYTICAL STUDY OF TEST SPECIMENS (PHASE 2) ......................... 251

9.1 Nonlinear Finite Element Analysis............................................................. 251
9.1.1 General ............................................................................................ 251
9.1.2 Steel Plate Girder Specimens (SG3 and SG4)................................. 251
vii
9.1.3 Composite Plate Girder Specimens (CG3 and CG4) ...................... 252
9.2 Comparison of Test Results with Predictive Equation ............................... 252
9.3 Modification of Predictive Equation for Concrete Shear Strength............. 253
10. NUMERICAL PARAMETRIC STUDY............................................................ 266

10.1 General........................................................................................................ 266
10.2 Effect of Longitudinal Stiffener.................................................................. 266
10.3 Effect of Small Flange Areas...................................................................... 267
10.4 Effect of Noncompact Web ........................................................................ 268
10.5 Effect of Unequal Top and Bottom Flanges ............................................... 269
11. RESISTANCE FACTOR FOR THE PROPOSED SHEAR STRENGTH

EQUATION........................................................................................................ 286
12. SUMMARY AND CONCLUSIONS ................................................................. 289
REFERENCES ........................................................................................................... 293
APPENDIX A. INSTRUMENTATION..................................................................... 298
APPENDIX B. PLOTS OF DEFORMATION AND STRAIN.................................. 316
viii
LIST OF FIGURES
Figure 1.1: Typical Configuration of Steel Plate Girder................................................... 17

Figure 1.2: Typical Plate Girder (Caltrans Br. No. 50-0316) ........................................... 18
Figure 1.3: Plate Girder Shear Load-Carrying Mechanism .............................................. 19
Figure 1.4: Various Tension-Field Models for Interior Panels......................................... 20
Figure 1.5: Free-Body Diagram from Basler (adapted from Basler, 1961a) .................... 22
Figure 1.6: Plane Stress in Pure Shear .............................................................................. 22
Figure 1.7: State of Stresses due to Shear Buckling and Tension-Field Action ............... 23
Figure 1.8: Von Mises Yield Condition (adapted from Basler, 1961a)............................ 23
Figure 1.9: End Panel Tension-Field Model (adapted from Huslid and Rockey, 1979) .. 24
Figure 1.10: Different End Supports for Plate Girders ..................................................... 24
Figure 1.11: Eurocode 3 Shear Resistance Factor (w) of Web........................................ 25
Figure 2.1: Configuration of Steel Girder Test Specimens............................................... 33
Figure 2.2: Specimen CG1: Configuration of Composite Girder Test Specimens........... 34
Figure 2.3: Specimen CG2: Configuration of Composite Girder Test Specimens........... 35
Figure 2.4: Concrete Slab Construction............................................................................ 36
Figure 2.5: Steel Plate Stress versus Strain Curves .......................................................... 37
Figure 2.6: Steel Reinforcement Stress versus Strain Curves........................................... 39
Figure 2.7: Concrete Cylinder Compressive Strength versus Curing Time ..................... 40
Figure 2.8: Measured Locations for Initial Geometrical Imperfections of Steel Girders. 41
Figure 2.9: Initial Geometrical Imperfection Measurements............................................ 41
Figure 2.10: Test Setup: Specimens SG1 and SG2........................................................... 43
Figure 2.11: Test Setup: Specimens CG1 and CG2.......................................................... 44
Figure 2.12: Support Detail............................................................................................... 45
Figure 2.13: Steel Loading Beam ..................................................................................... 45
Figure 2.14: Test Setup: Beam Lateral Bracing................................................................ 46
Figure 3.1: Specimen SG1: Test Setup ............................................................................. 55
Figure 3.2: Specimen SG1: Applied Load versus Mid-Span Deflection.......................... 56
Figure 3.3: Elastic Timoshenko Beam Deformation ........................................................ 56
ix
Figure 3.4: Specimen SG1: Deformed Shape at Event A ................................................. 57
Figure 3.5: Specimen SG1: Web Local Buckling at Event B ........................................... 57
Figure 3.6: Specimen SG1: Yielding Patterns at Event B (Back Side) ............................ 58
Figure 3.7: Specimen SG1: Failure Mode at Event C ...................................................... 58
Figure 3.8: Specimen SG1: Failure Mode after Completion of Test (West End Panel)... 59
Figure 3.9: Specimen SG2: Applied Load versus Mid-Span Deflection.......................... 59
Figure 3.10: Specimen SG2: Minor Web Local Buckling at Event A .............................. 60
Figure 3.11: Specimen SG2: Yielding Pattern at Event B ................................................ 60
Figure 3.12: Specimen SG2: Web Local Buckling at Event B ......................................... 61
Figure 3.13: Specimen SG2: Failure Mode at Event C .................................................... 61
Figure 3.14: Specimen SG2: Failure Mode after Completion of Test (West End Panel). 62
Figure 3.15: Failure Modes of SG1 and SG2 ................................................................... 62
Figure 3.16: Specimen CG1: Applied Load versus Mid-Span Deflection ....................... 63
Figure 3.17: Specimen CG1: Web Yielding and Local Buckling at Δ=0.2 in.................. 64
Figure 3.18: Specimen CG1: Yielding Patterns at Event B (Back Side) .......................... 65
Figure 3.19: Specimen CG1: Propagation Concrete Slab Cracks..................................... 66
Figure 3.20: Specimen CG1: Failure Mode at Event C .................................................... 67
Figure 3.21: Specimen CG1: Failure Mode after Completion of Test ............................. 67
Figure 3.22: Specimen CG2: Applied Load versus Mid-Span Deflection ....................... 69
Figure 3.23: Specimen CG2: Web Local Buckling at Event A......................................... 69
Figure 3.24: Specimen CG2: Yielding Pattern at Event B................................................ 70
Figure 3.25: Specimen CG2: Concrete Slab Crack Propagation ...................................... 71
Figure 3.26: Specimen CG2: Failure Mode at Event C .................................................... 72
Figure 3.27: Specimen CG2: Failure Mode after Completion of Test (West End Panel) 72
Figure 3.28: Typical Locations of Displacement Transducers (Back Side) ..................... 73
Figure 3.29: Specimen SG1: Web Out-of-Plane Deformation ......................................... 74
Figure 3.30: Specimen SG2: Web Out-of-Plane Deformation ......................................... 75
Figure 3.31: Specimen CG1: Web Out-of-Plane Deformation......................................... 76
Figure 3.32: Specimen CG2: Web Out-of-Plane Deformation......................................... 77
Figure 3.33: Typical Locations of Strain Rosettes in End Panels (Back Side)................. 78
Figure 3.34: Principal Strains across Panel Diagonals at Maximum Load....................... 79
x
Figure 3.35: Measured Plastic Hinge Locations ............................................................... 81
Figure 3.36: Specimen CG2: Concrete Shear Crack Propagation .................................... 81
Figure 3.37: Measured Concrete Shear Crack Angles...................................................... 82
Figure 3.38: Comparison between Test Results and AASHTO Nominal Shear Strengths
................................................................................................................................... 83
Figure 4.1: Specimen SG1: Typical Finite Element Model and Mesh ............................. 92
Figure 4.2: Comparison of Deformed Shape .................................................................... 92
Figure 4.3: Correlation between Test Results and FEM Analyses ................................... 94
Figure 4.4: Comparison of Plastic Hinge Locations......................................................... 95
Figure 4.5: SG1: Strain Profiles........................................................................................ 96
Figure 4.6: SG1: Principal Tensile Strains ....................................................................... 98
Figure 4.7: SG2: Strain Profiles........................................................................................ 99
Figure 4.8: SG2: Principal Tensile Strains ..................................................................... 101
Figure 4.9: Specimen CG2: Typical Finite Element Model and Mesh........................... 102
Figure 4.10: Typical Concrete Stress-Strain Curve in Compression.............................. 103
Figure 4.11: Relationship between Post-Failure Stress and Fracture Energy................. 103
Figure 4.12: Correlation between Test Results and FEM Analyses ............................... 104
Figure 4.13: CG1: Principal Tensile Strains (Δ=0.25 in.) .............................................. 105
Figure 4.14: CG1: Principal Strains in Concrete Slab .................................................... 106
Figure 4.15: CG2: Principal Tensile Strains ................................................................... 107
Figure 4.16: CG2: Principal Strains in Concrete Slab .................................................... 108
Figure 4.17: CG2: Concrete Slab Effective Shear Width ............................................... 109
Figure 5.1: Typical Finite Element Model (2 Panels)..................................................... 121
Figure 5.2: Model SG1: Comparison of 4 Panels and 2 Panels...................................... 121
Figure 5.3: Model SG1: Summary of Parametric Study................................................. 123
Figure 5.4: Model SG2: Summary of Parametric Study................................................. 124
Figure 5.5: Effect of d0/D and D/tw on Plastic Hinge Location ...................................... 125
Figure 5.6: Model CG2: Effect of Concrete Slab Width ................................................ 126
Figure 5.7: Model CG2: Effect of Stud End Distance .................................................... 126
Figure 5.8: Specimen CG2: Direction of Compressive Principal Strain ........................ 127
Figure 6.1: Failure Modes............................................................................................... 154
xi
Figure 6.2: Tension-Field Band Width ........................................................................... 155
Figure 6.3: Assumed End Panel Collapse Mechanism ................................................... 156
Figure 6.4: Free-Body Diagram of End Panel with Web Excluded ............................... 157
Figure 6.5: Free-Body Diagram of Boundary Members between Plastic Hinges .......... 158
Figure 6.6: Effect of Shear Ratio Parameters ................................................................. 159
Figure 6.7: Free-Body of End Panel Due to Tension-Field Action ................................ 160
Figure 6.8: Flange Shear Contribution............................................................................ 161
Figure 6.9: Effective Web Depth, de ............................................................................... 161
Figure 6.10: Relationship between C and D/tw ............................................................... 162
Figure 6.11: Effective Sections of Boundary Members.................................................. 162
Figure 6.12: Comparison of Plastic Hinge Location a in Top Flange ............................ 163
Figure 6.13: Comparison of Plastic Hinge Location b in Bearing Stiffeners ................. 164
Figure 6.14: Comparison of Effective Web Depth, de .................................................... 165
Figure 6.15: Comparison of End Panel Shear Strengths (Test Specimens).................... 167
Figure 6.16: Comparison of End Panel Shear Strengths................................................. 167
Figure 6.20: CG1: Shear Crack Comparison in Concrete Slab ...................................... 170
Figure 6.21: CG2: Shear Crack Comparison in Concrete Slab ...................................... 171
Figure 6.22: Specimen CG1: Concrete Crushing ........................................................... 172
Figure 6.23: Analytical Model for Load Transfer Mechanism in Concrete Slab ........... 173
Figure 6.24: Effective Width of Compression Strut ....................................................... 173
Figure 7.1: Configuration of End Stiffeners ................................................................... 192
Figure 7.2: Basler’s Model for End Stiffeners (adapted from Basler 1961a)................. 192
Figure 7.3: Rehabilitated Model SG1: Finite Element Models ...................................... 193
Figure 7.4: Basler’s Scheme: Failure Mode as a Function of End Stiffener Area.......... 194
Figure 7.5: Eurocode 3 Scheme: Failure Mode as a Function of End Stiffener Area .... 195
Figure 7.6: Wing Stiffener Rehabilitation Scheme for End Panel.................................. 196
Figure 7.7: Model SG1R-W1: Applied Load versus Mid-Span Deflection ................... 197
Figure 7.8: Wing Stiffener Rehabilitation Scheme: Failure Modes of SG1 Series ........ 197
xii
Figure 7.9: Wing Stiffener Rehabilitation Scheme: Failure Modes of SG2 Series ........ 198
Figure 7.10: Example Calculation of Mpb ....................................................................... 199
Figure 7.11: Components of Mpb .................................................................................... 199
Figure 7.12: Rehabilitation Scheme with Bolted Angles................................................ 200
Figure 7.13: Subpanels 1 and 2....................................................................................... 200
Figure 7.14: SG1 Series (Approach 1)............................................................................ 201
Figure 7.15: SG2 Series (Approach 1)............................................................................ 202
Figure 7.16: Failure Mode (Approach 2)........................................................................ 203
Figure 7.17: Failure Mode (Approach 2) (continued) .................................................... 204
Figure 7.18: Applied Load versus Mid-Span Deflection (Approach 2) ......................... 205
Figure 7.19: Effect of Width-Depth Ratio on Shear Strength of End Panels ................. 206
Figure 8.1: Configuration of Steel Girder Test Specimens............................................. 220
Figure 8.2: Specimen CG3: Configuration of Composite Girder Test Specimens......... 221
Figure 8.3: Specimen CG4: Configuration of Composite Girder Test Specimens......... 222
Figure 8.4: Specimen SG3 Test Setup ............................................................................ 223
Figure 8.5: Specimen SG3 Applied Load versus Mid-Span Deflection......................... 224
Figure 8.6: Specimen SG3: Web Local Buckling at Event A ......................................... 224
Figure 8.7: Specimen SG3: Web Local Buckling at Event B ......................................... 225
Figure 8.8: Specimen SG3: Failure Mode at Event C .................................................... 226
Figure 8.9: Specimen SG3: Failure Mode after Completion of Test (East End Panel) .. 226
Figure 8.10: Specimen SG4: Applied Load versus Mid-Span Deflection...................... 227
Figure 8.11: Specimen SG4: Web Local Buckling at Event A ....................................... 227
Figure 8.12: Specimen SG4: Web Local Buckling at Event B ....................................... 228
Figure 8.13: Specimen SG4: Deformed Shape at Event C ............................................. 229
Figure 8.14: Specimen SG4: Deformed Shape at Event D............................................. 229
Figure 8.15: Specimen SG4: Failure Mode at Event E................................................... 230
Figure 8.16: Specimen SG4: Failure Mode after Completion of Test (East End Panel) 230
Figure 8.17: Failure Modes of SG1 and SG2 ................................................................. 231
Figure 8.18: Specimen CG3: Applied Load versus Mid-Span Deflection ..................... 231
Figure 8.19: Specimen CG3: Overall View at Event A .................................................. 232
Figure 8.20: Specimen CG3: Web Yielding pattern at Δ = 0.3 in. ................................. 232
xiii
Figure 8.21: Specimen CG3: Deformed Shape at Event B (Maximum Load) ............... 233
Figure 8.22: Specimen CG3: Failure Mode at Event C .................................................. 233
Figure 8.23: Specimen CG3: Propagation Concrete Slab Cracks................................... 234
Figure 8.24: Specimen CG3: Failure Mode after Completion of Test ........................... 235
Figure 8.25: Specimen CG4: Applied Load versus Mid-Span Deflection ..................... 236
Figure 8.26: Specimen CG4: Overall View at Event A .................................................. 236
Figure 8.27: Specimen CG4: Web Local Buckling and Yielding Patterns at Δ = 0.4 in 237
Figure 8.28: Specimen CG4: Yielding Pattern at Event B.............................................. 238
Figure 8.29: Specimen CG4: Concrete Slab Crack Propagation .................................... 239
Figure 8.30: Specimen CG4: Failure Mode at Event C .................................................. 240
Figure 8.31: Specimen SG3: Locations of Displacement Transducers (Back Side) ...... 241
Figure 8.32: Specimen SG3: Web Out-of-Plane Deformation ....................................... 242
Figure 8.33: Specimen SG4: Web Out-of-Plane Deformation ....................................... 243
Figure 8.34: Specimen CG3 Web Out-of-Plane Deformation........................................ 244
Figure 8.35: Specimen CG4 Web Out-of-Plane Deformation........................................ 245
Figure 8.36: Locations of Strain Rosettes in End Panels (Back Side)............................ 246
Figure 8.37: Principal Strains across Panel Diagonals at Maximum Load..................... 247
Figure 8.38: Measured Plastic Hinge Locations ............................................................. 249
................................................................................................................................. 250
Figure 9.1: Specimen SG3: Typical Finite Element Model and Mesh ........................... 257
Figure 9.2 Comparison of Buckling Pattern of Specimen SG3 ...................................... 257
Figure 9.3 Comparison of Buckling Pattern of Specimen SG4 ...................................... 258
Figure 9.4: Correlation between Test Results and FEM Analyses ................................. 259
Figure 9.5: Comparison of Plastic Hinge Locations....................................................... 260
Figure 9.6: Correlation between Test Results and FEM Analyses (CG3 and CG4)....... 261
Figure 9.7 Comparison of Buckling Pattern of Specimen CG3 (at 1.0 in. Disp.) .......... 262
Figure 9.8 Comparison of Buckling Pattern of Specimen CG4 (at 1.0 in. Disp.) .......... 263
Figure 9.9: Assumed Compression Strut (Specimen CG3) ............................................ 264
Figure 9.10: Assumed Compression Strut (Specimen CG4) .......................................... 265
Figure 10.1: Finite Element Models with Different Longitudinal Stiffener Locations .. 277
xiv
Figure 10.2: SG1LS02: Effect of Longitudinal Stiffener ............................................... 278
Figure 10.6: Girder LS2 after Testing (adopted from Cooper, 1966)............................. 282
Figure 10.7: Coimparison of Shear Strengths of Plate Girders with Small Flange Area 283
Figure 10.8: Shear Strength Comparison of Plate Girder End Panel with Noncompact
Web ......................................................................................................................... 284
Figure 10.9: Shear Strength Comparison of Plate Girder End Panel with Unequal Top and
Bottom Flanges ....................................................................................................... 285
Figure A.1: Specimen SG1: Displacement Transducers................................................. 299
Figure A.3: Specimen CG1: Displacement Transducers ................................................ 301
Figure A.9: Specimen SG1: Strain Gage and Rosette Locations.................................... 306
Figure A.10: Specimen SG2: Strain Gage and Rosette Locations.................................. 307
Figure A.11: Specimen CG1: Strain Gage and Rosette Locations ................................. 308
Figure B.1: Specimen SG1: Flange Deformation ........................................................... 316
Figure B.2: Specimen SG1: Bearing Stiffener Deformation .......................................... 317
Figure B.3: Specimen SG2: Flange Deformation ........................................................... 318
Figure B.4: Specimen SG2: Bearing Stiffener Deformation .......................................... 319
Figure B.5: Specimen CG1: Bearing Stiffener Deformation.......................................... 320
xv
Figure B.6: Specimen CG2: Bearing Stiffener Deformation.......................................... 321
Figure B.7: Specimen SG1: Principal Strain .................................................................. 322
Figure B.8: Specimen SG2: Principal Strain .................................................................. 323
Figure B.9: Specimen CG1: Principal Strain.................................................................. 324
Figure B.10: Specimen CG2: Principal Strain................................................................ 325
Figure B.11: Specimen SG3: Principal Strain ................................................................ 326
Figure B.12: Specimen SG4: Principal Strain ................................................................ 327
Figure B.15: Specimen SG1: Strain Profiles in Extended Web...................................... 330
Figure B.16: Specimen SG1: Strain Profiles in Bearing Stiffener.................................. 331
Figure B.17: Specimen SG2: Strain Profiles in Extended Web...................................... 332
Figure B.18: Specimen SG2: Strain Profiles in Bearing Stiffener (East Side)............... 333
Figure B.19: Specimen SG2: Strain Profiles in Bearing Stiffener (West Side).............. 334
Figure B.20: Specimen CG1: Strain Profiles in Extended Web ..................................... 335
Figure B.21: Specimen CG1: Strain Profiles in Bearing Stiffener (East Side) .............. 336
Figure B.22: Specimen CG1: Strain Profiles in Bearing Stiffener (West Side) ............. 337
Figure B.23: Specimen CG2: Strain Profiles in Extended Web ..................................... 338
Figure B.24: Specimen CG2: Strain Profiles in Bearing Stiffener (East Side) .............. 339
Figure B.25: Specimen CG2: Strain Profiles in Bearing Stiffener (West Side) ............. 340
Figure B.28: Specimen CG3: Strain Profiles in Bearing Stiffener ................................. 343
Figure B.29: Specimen CG4: Strain Profiles in Bearing Stiffener ................................. 344
Figure B.30: Specimen SG1: Strain Profiles in Compression Flange (Top Flange) ...... 345
Figure B.31: Specimen SG1: Strain Profiles in Tension Flange (Bottom Flange)......... 346
Figure B.34: Specimen CG1: Strain Profiles in Compression Flange (Top Flange)...... 349
Figure B.35: Specimen CG1: Strain Profiles in Tension Flange (Bottom Flange) ........ 350
xvi
Figure B.46: Specimen CG3 ........................................................................................... 361
Figure B.47: Specimen CG4 ........................................................................................... 362
xvii
LIST OF TABLES
Table 1.1: Eurocode 3 Shear Resistance Reduction Factor (w) of Web ......................... 16
Table 2.1: Sample Caltrans Steel Bridge Girder Dimensions........................................... 30
Table 2.2: Specimen Dimensions ..................................................................................... 30
Table 2.3: Steel Plate Thicknesses.................................................................................... 30
Table 2.4: Average Mechanical Properties of Steel Plates ............................................... 31
Table 2.5: Average Mechanical Properties of Steel Reinforcements ............................... 31
Table 2.6: Average Mechanical Properties of Concrete ................................................... 32
Table 2.7: Measured Initial Web Imperfections (in.) ....................................................... 32
Table 3.1: Comparison of End Panel Shear Strengths...................................................... 53
Table 3.2: Prinicipal Strain Direction at Ultimate Load (Steel Plate Girders) ................. 53
Table 3.3: Prinicipal Strain Direction at Ultimate Load (Composite Plate Girders) ........ 54
Table 4.1: Shear Strength Comparison between FEM and Experiments.......................... 91
Table 5.1: Summary of Model SG1 Parametric Study (D/tw = 210) .............................. 115
Table 5.2: Summary of Model SG2 Parametric Study (D/tw = 280) .............................. 118
Table 5.3: Summary of Model SG1 Parametric Study ................................................... 119
Table 5.4: Summary of Model SG2 Parametric Study ................................................... 119
Table 5.5: Effect of Concrete Slab Width....................................................................... 120
Table 6.1: Plate Girder Test Specimen Dimensions ....................................................... 145
Table 6.2: Material Properties and Identified Plastic Hinge Locations from Tests........ 146
Table 6.3: End Panel Shear Strength Comparison between Prediction and Experiment
Results..................................................................................................................... 147
Table 6.4: Model SG1 Series: End Panel Shear Strength Comparison between Prediction
and Design Codes ................................................................................................... 148
Table 6.5: Model SG2 Series: End Panel Shear Strength Comparison between Prediction
and Design Codes ................................................................................................... 151
Table 6.6: Composite Girder Shear Strength Prediction ................................................ 153
Table 7.1: Effect of End Stiffener Area (Based on Model SG1-7)................................. 185
Table 7.2: Effect of Wing Stiffeners............................................................................... 186
xviii
Table 7.3: Width-Depth Ratios of Panels (Approach 1)................................................. 187
Table 7.4: Comparison of Shear Strength before and after Rehabilitation (Approach 1)
................................................................................................................................. 187
Table 7.5: Comparison of Subpanel 2 Shear Strength (Approach 1) ............................. 188
Table 7.6: Effect of Stiffener Thickness on Shear Strengths (Approach 1) ................... 188
Table 7.7: Width-Depth Ratios of Panels (Approach 2)................................................. 189
Table 7.8: Comparison of Shear Strength (Approach 2) ................................................ 190
Table 7.9: Comparison of Shear Strength Predicted by FEM and Eq. (6.33)................. 191
Table 8.1: Sample Caltrans Steel Bridge Girder Dimensions......................................... 214
Table 8.2: Specimen Dimensions ................................................................................... 214
Table 8.3: Steel Plate Thicknesses.................................................................................. 214
Table 8.4: Average Mechanical Properties of Steel Plates ............................................. 215
Table 8.5: Average Mechanical Properties of Steel Reinforcements ............................. 215
Table 8.6: Average Mechanical Properties of Concrete ................................................. 216
Table 8.7: Measured Initial Web Imperfections (in.) ..................................................... 216
Table 8.8: Prinicipal Strain Direction at Ultimate Load (Steel Plate Girders) ............... 217
Table 8.9: Prinicipal Strain Direction at Ultimate Load (Composite Plate Girders) ...... 218
Table 8.10: Comparison of End Panel Shear Strengths.................................................. 219
Table 9.1: Shear Strength Comparison between FEM and Experiments........................ 255
Table 9.2: Shear Strength Comparison (Steel Girder Specimens).................................. 255
Table 9.3: Shear Strength Comparison (Composite Girder Specimens) ........................ 255
Table 9.4: Comparison of Shear Strength between Modified Prediction and Test Results
................................................................................................................................. 256
Table 10.1: Effect of Longitudinal Stiffener................................................................... 272
Table 10.2: Parameters of FEM Models with Small Flange Areas ................................ 273
Table 10.3: Effect of Small Flange Area on Ultimate Shear Strength............................ 273
Table 10.4: Ranges of Depth-Thickness Ratio for a Noncompact Web......................... 274
Table 10.5: Effect of Noncompact Web ......................................................................... 275
Table 10.6: Effect of Unequal Top and Bottom Flanges ................................................ 276
Table 11.1: Resistance Factor Based on Physical Tests ................................................. 288
xix
1. INTRODUCTION
1.1 Plate Girders
Typically in building construction, hot-rolled W-shape or I-shape members are

used as beams. However, for longer spans in which flexural demands are greater, such
as in bridges, deeper beam sections are required. Since hot-rolled shapes are not
economical in such circumstances, beams built-up from steel plates, which are commonly
referred to as plate girders, have been used for long-span structures.
Since plate girders typically have very deep sections, the web of a plate girder is
very slender. The slender web causes an instability problem under shear loading. To
remediate this issue, the web of the plate girder is typically reinforced with transverse
stiffeners and occasionally longitudinal stiffeners to increase its shear buckling strength.
Figure 1.1 shows the typical structural configuration of a steel plate girder. As shown in
the figure, the web is stiffened with transverse stiffeners. Transverse stiffeners that are
located at the supports are called bearing stiffeners, and should be double-sided. Those
located away from the supports are called intermediate stiffeners, which can be either
one-sided or double-sided. A panel is defined as a region of the girder web that is
surrounded by girder flanges and transverse stiffeners. The panel nearest to the support
is referred to as an end panel or an exterior panel, while the remaining panels are referred
to as interior panels. Also, some portion of the web is typically extended out beyond the
bearing stiffeners and is referred to as the extended web (e in Figure 1.1). Figure 1.2
shows a sample steel girder bridge which was built in 1967. The bridge comprises
welded steel plate girders, which act compositely with a concrete slab through headed
shear studs. Many transverse stiffeners were used to increase the web shear strength.
Therefore, the design strategy for plate girders is to determine an optimal combination of
plate size and thickness, and stiffener spacing.
1.2 Shear Load-Carrying Mechanism of Stiffened Steel Plate Girder
Unlike flexural buckling for columns and lateral-torsional buckling for beams, the
stiffened web panels of steel girders can carry more shear load after exceeding the shear
1
buckling strength. A web in shear is initially in a pure shear stress state. The shear
stresses present are equivalent to two principal stresses (one in tension and one in
compression) that are inclined at 45º to the shear stresses [see Figure 1.3(a)]. This
shear-transfer mechanism prior to web buckling is referred to as “beam action”. When
the elastic shear buckling strength of the panel is reached, the panel buckles along the
panel diagonal while showing out-of-plane deformation.
After elastic shear buckling occurs, the principal compressive stresses do not
increase any further, and the principal tensile stresses continue to increase and approach
the panel yield strength as further buckling occurs along the panel diagonal. This
diagonally buckled portion in the panel functions as a tension tie member, while the
transverse stiffeners serve as compression strut members in a Pratt truss [see Figure 1.3
(b)]. This shear transfer mechanism resembling a truss system provides additional post-
buckling strength. The inclined tensile membrane stress (or tension-field stress) of the
panel is referred to as “tension-field action”.
Therefore, the ultimate shear strength of a plate girder consists of two
components: the “beam action” shear strength (or shear buckling strength) and the
“tension-field action” shear strength (or post-buckling strength).
1.3 Tension-Field Action in Interior Panels
1.3.1 General
The source of the post-buckling strength was first observed by Wilson (1886).
The first tension-field action theory for web panels was developed by Wagner (1931) and
his work was extended by Kuhn (1956) for aircraft design. Basler (1961a) developed
the first mathematical model to quantify the post-buckling shear resistance due to
tension-field action of plate girders for structural engineering applications.
1.3.2 Basler’s Theory

Basler assumed that the tension-field action can develop if the boundaries can
provide enough resistance to the tension-field force. Since tension-field action takes
place in the web diagonally [see Figure 1.3 (b)], the boundaries that comprise the girder
flanges and the transverse stiffeners of neighboring panels should resist the horizontal
2
and vertical component of the diagonal tension-field force. Basler also assumed that the
flanges are too flexible to resist the vertical force induced by the tension-field force, so
tension-field action cannot take place along the flanges. Instead, vertical component of
the tension-field force can be resolved by the transverse stiffeners, while the horizontal
component is resisted by the neighboring panels. Thus, the boundary along the
transverse stiffeners can effectively resist all forces induced by tension-field force.
For an interior panel, the resulting tension band based on this assumption is
shown in Figure 1.4(a). The inclination of the yield band is defined by the angle of  ,
which was chosen to maximize the web shear strength. To evaluate the post-buckling
strength due to tension-field action, Basler first cut a section horizontally at mid-depth
and vertically at the middle of the panels so that the diagonal tensile stress lied over the
free-body (see Figure 1.5). From the free-body diagram, taking the horizontal force
equilibrium and moment equilibrium at point “O” gives
F x  0:
do
F f   t t w d o sin  cos    t t w sin 2 (1.1)
2
M O  0:
do
F f  Vtf (1.2)
D
1
sin 2 in Eq. (1.1) is equal to . The shear strength due to tension-field
1  (d o / D) 2
action can be determined by substituting Eq. (1.2) into Eq. (1.1).
 t Dt w   1 
Vtf    (1.3)
2  1  (d o / D) 2 
Basler also derived the tension-field stress ( t ) by assuming that the
superposition of the buckling stress and the inclined membrane tensile stress satisfies the
yield condition. A two-dimensional view of the element in a state of pure shear is
shown in Figure 1.6(a). The same state of plane stress can be described on any other
coordinate system, such as x  - y  in Figure 1.6 (b). This x  - y  coordinate system is
generated from the x - y coordinate system by a counterclockwise rotation of  . By
3
cutting along the dashed line in Figure 1.6 (a) and drawing the free-body diagram of the
wedge-shape, the stress components in the rotated coordinate system can be determined.
The stresses must first be multiplied by the areas of the sides of the triangular element to
obtain forces. For convenience, the hypotenuse is taken to be of unit length, and the
element thickness is assumed to be unity. The resulting free-body is shown in Figure
1.7(a). Taking horizontal and vertical equilibriums of the free-body gives two
equations:
F x  0;
 y sin    cos    cr cos   0 (1.4)
F y  0;
y cos    sin    cr sin   0 (1.5)
Solving simultaneous Eqs. (1.4) and (1.5) gives

y   cr sin 2
(1.6)
   cr cos 2
By substituting (  +90°) into y , normal stress x becomes   cr sin 2 .
After shear buckling occurs, membrane tension-field stress develops as shown in Figure
1.7 (b). Superposition of the two stress conditions in the direction of y  yields
x   cr sin 2
y   cr sin 2   t (1.7)
   cr cos 2
In a two-dimensional plane stress state, von Mises yield condition can be defined
as
 2M  Fyw2  (x ) 2  (y ) 2  (x )(y )  3() 2 (1.8)
where Fyw is web yield stress. Substituting Eq. (1.7) into Eq. (1.8) gives the tension-
field stress
3 
2 3 
2

 t    cr sin 2  Fyw   cr  sin 2   3
2
(1.9)
2  2  
4
To simplify the solution, Basler assumed a linear approximation of the von Mises
yield condition on the ellipse between points A and B (see Figure 1.8). The equation of
the straight line is 1  Fyw   2  3 1 . Also, assuming   45 ,  'x and  'y in Eq.
(1.7) become principal stresses 1   cr   t , and  2   cr , respectively. Therefore,

as an approximated form of the yield condition, the simple solution for the tension-field
stress is
  
 t  Fyw 1  cr  (1.10)
 y 
 
where  y is shear yield stress (= Fyw / 3 ) Substituting Eq. (1.10) into Eq. (1.3), the
post-buckling strength due to tension-field action is

  cr 
1  
 y 
Fyw Dt w   0.87(1  C ) (1.11)
Vtf   Vp
2 1  (d o / D) 2 1  ( d o / D) 2
Fyw Dt w  cr Fyw
where V p  , C , and  y  . Eq. (1.11) is currently adopted in the
3 y 3
AASHTO Specifications (2014) for calculating the shear strength of interior panels.
The AISC Specifications (2010) also has a similar form, but with  y approximated as
0.6 Fyw .
1.3.3 Brief Review of Various Tension-Field Theories

After Basler’s theory was developed, a number of researchers have proposed
alternate tension-field stress models for predicting the ultimate shear strength of the web.
Most of the works addressing tension-field action were summarized in Ajam (1986) and
SSRC (2010). Figure 1.4 presents the various tension-field models.
Basler (1960, 1961a, 1961b), Takeuchi (1964), Rockey and Skaloud (1972),
Herzog (1974), and Porter et. al. (1975) assumed that a diagonal tension-field develops in
a limited portion of the web. Fujii (1968, 1971), Chern and Ostapenko (1969) and
Komatsu (1971) assumed that a diagonal tension-field develops in a much larger region
5
in the web panel, but the intensity of the diagonal tension varies across the perpendicular
direction of the tension diagonal (Yoo and Lee, 2006).
Takeuchi (1964) first considered the effect of flange out-of-plane stiffness on the
web yield band. He assumed that the tension-field stress extends to the distances c1
and c 2 along the top and bottom flanges, respectively. These distances were also
assumed to be proportional to the moment of inertia of the flanges. Although his model
could not provide a good correlation with the test results, the flange stiffness effects were
considered in all other researchers’ models afterwards. Therefore, all models in Figure
1.4 except Basler’s model show that plastic hinges develop on both flanges, and the
diagonal tension-field action acts on both flanges and sides of the transverse stiffeners.
Chern and Ostapenko (1969), Rockey and Skaloud (1972), and Porter et al.
assumed that a portion of the web interacts with the flanges and was included in
computing the plastic moments of the flanges. Therefore, a T-shaped flange section
including a portion of the web was used to determine the flange plastic moment.
Only Porter et al. (1975) assumed that the tension-field stress angle is different
from the angle of the tension-field zone. In this model, the inclined angle of the tension-
field stress was assumed to be two-third of the angle of the panel diagonal with respect to
the flange. The ultimate shear strength was then determined by iteration.
Hoglund first developed his theory in 1971 (Hoglund, 1971) for the girders with
bearing stiffeners at the supports only, and later (1997) extended his theory to include
girders with intermediate stiffeners. As shown in Figure 1.4(i), Hoglund replaced the
web by a system of tension and compression bars that intersected one another
perpendicularly. The inclination of the bars changed as the load increased. It was
observed that the compression bars carried a constant stress which was equal to the shear
buckling stress, while the stress in the tension bars increased as the angle between the
tension bars and the flanges decreased (Ajam, 1986). He also considered the shear
contribution of the flanges and formulated an empirical solution for the ultimate shear
strength of the panel. His solution was adopted by Eurocode 3 (2006).
Lee and Woo (1998, 1999) conducted a number of finite-element analyses
varying the web depth-to-thickness ratio and the web width-to-depth ratio. Based on the
findings, they suggested a simple empirical solution for the ultimate shear strength.
6
They also demonstrated the accuracy of the empirical solution through an experimental
study.
Note that all studies summarized in this section are for interior panels.
1.4 US Plate Girder Shear Design Practice
1.4.1 General
Since both US design methods [AASHTO Specifications (2014) for bridge design
and AISC Specifications (2010) for building design] adopt Basler’s theory, tension-field
action is only allowed for interior panels whose boundaries comprise the flanges and
intermediate stiffeners of neighboring panels. Since end panels do not have a
neighboring panel adjacent to the bearing stiffener pair, they do not qualify for tension-
field action.
1.4.2 Interior Panels

Based on AASHTO Specifications (2014), the nominal shear resistance ( Vn ) of a
stiffened interior web panel is given by
 0.87(1  C )  2 Dt w
Vn  V p C   for  2.5 (1.12)
 1  d 0 / D 
2  (b fc t fc  b ft t ft )
 
 0.87(1  C )  2 Dt w
Vn  V p C   for  2.5 (1.13)
 1  d 0 / D   d 0 / D  
2 (b fc t fc  b ft t ft )

1
Vp  Fyw Dt w  0.58Fyw Dt w (1.14)
3
where
d0 = transverse stiffener spacing (in.),
D = web depth (clear distance between flanges, in.),
tw = web thickness (in.),
b fc = width of compression flange (in.),
7
b ft = width of tension flange (in.),
t fc = thickness of compression flange (in.),
t ft = thickness of tension flange (in.),
Fyw = web yield stress (ksi),
Vp = plastic shear force or shear yield strength (kips), and
C = ratio of the shear buckling strength to the shear yield strength.
The shear buckling strength to shear yield strength ratio, C , can be determined as
follows:
D Ek
C  1.0 if  1.12 (1.15)
tw Fyw
1.12 Ek Ek D Ek
C if 1.12   1.40 (1.16)
D / tw Fyw Fyw t w Fyw
1.57  Ek  D Ek
C   if  1.40 (1.17)
(D / t w ) 2 F  tw Fyw
 yw 
where
E = modulus of elasticitity of steel
 5 
k = shear buckling coefficient   5  

 d 0 / D 2 
Note that the shear buckling coefficient ( k ) is a simplification of two exact equations
based on simply supported boundary conditions. The exact solutions for k is given by
(Timoshenko and Gere, 1961)
5.34
k exact  4 d0
1
d 0 / D 2 for
D
(1.18)
4 d
5.34  for 0  1
d 0 / D 2 D
8
1.5 End Panels
The first and second terms on the right-hand side of Eq. (1.12) are the shear
strength components due to beam action and tension-field action, respectively. Since an
end panel does not have a neighboring panel beyond the support to anchor the horizontal
component of the tension-field action, the AASHTO Specifications (2014) only considers
the beam action in computing its nominal shear resistance:
Vn  CV p (1.19)
The constant C should be calculated based on Eqs. (1.15) through (1.17). In addition,
AASHTO limits the width-to-depth ratio ( d 0 / D ) of the end panel to 1.5.
1.6 Statement of Problem
Based on Basler’s tension-field action theory, design shear strength equations

were first introduced in the 11th Edition of the AASHTO Standard Specifications (1973).
In the Specifications (11th Edition), shear capacities of both interior and end panels were
the same. Later, the shear capacity of end panels was reduced in AASHTO 12th Edition
(1977). In AASHTO Interim Specifications (1986), the current format of the end panel
shear strength equations was introduced.
Since the maximum shear occurs at the support, the end web panel may become
the controlling component for steel girder design, evaluation and load rating (AASHTO
2011). AASHTO LRFD Specifications (7th Edition, 2014) does not provide any
provisions to allow the end web panel to be designed for tension-field action. As a
result, many steel girder bridges in which the end shear controls have a rating factor
below 1.0. Rating factors less than 1.0 require that the bridges must be strengthened or
vehicular loading on the bridge must be restricted.
However, Basler (1961a), whose equation became the basis of shear design for
plate girders in AASHTO Specifications (2014), pointed out that a partial tension-field
action may occur in the end panel because some amount of horizontal force can be
anchored by the bearing stiffeners and the extended web portion. But since the degree
of post-buckling strength was uncertain, he conservatively ignored it.
9
There may be another source of increase in shear strength at the end panel:
concrete deck shear strength. Plate girders in bridges are almost always used in
conjunction with concrete slabs, which are able to resist some amount of shear.
However, the contribution of concrete deck is not considered for the shear strength design
in AASHTO Specifications (2014).
The current AASHTO Specifications may be too conservative in the shear design
of plate girder end panels. This is especially true for the rating of existing steel plate
girder bridges. Among others, California Department of Transportation (Caltrans) has
identified many existing steel girder bridges that require retrofit based on the current
ASHTO method, although these bridges have performed well in the past few decades.
To avoid unnecessary retrofit, it is necessary to develop more accurate and practical shear
design equations for the end panel that include the potential contributions from both
tension-field action and the concrete slab.
1.7 Shear Resistance in End Panels
Huslid and Rockey (1979) experimentally and analytically studied end panel
behavior. With a total of 8 specimens, they conducted 13 tests by repairing the
specimens. Although the specimen sizes were very small (the ultimate panel shear
strength achieved from the testing varied from 4.8 kips to 13.7 kips), the test results
showed the contribution of tension-field action in the end panels. Figure 1.9 shows their
tension-field model to predict the ultimate shear strength for the end panels. Their
analytical model was extended by a study of the interior panels conducted by Porter et al.
(1975). The study found that the tension-field stress angle differed from the angle of the
tension band, and that the inclined angle of the tension-field stress must be determined to
maximize the shear strength. With their model, an iterative method was necessary to
determine the ultimate shear strength.
Safar (2013) conducted a nonlinear finite element parametric study with 64 plate
girder end panels. The parameters included configurations of bearing stiffeners (e.g.,
having no bearing stiffeners, a one-sided bearing stiffener, one pair of bearing stiffeners,
and two pairs of bearing stiffeners), flange thickness-to-web thickness ratio, web depth-
10
to-thickness ratio, and web width-to-depth ratio. Based on the numerical results, an
empirical solution was suggested for the ultimate end panel shear strength.
In current European design methods, the basic design concept is similar to that of
the US. However, there are two major differences in the Eurocode 3 (CEN 2006)
design procedure. First, Eurocode 3 provides the plate girder shear strength as a
summation of the contributions from the web and flanges. Secondly, Eurocode 3
considers a partial post-buckling strength in the end panel. Figure 1.10(a) shows the
rigid end post at the support, which has an additional pair of stiffeners at the end of the
girder to generate the tension-field action; and the panel with the rigid end post is treated
as an interior panel. Figure 1.10(b) shows the non-rigid end post at the support that is
treated as an end panel as in US design practice. But since the end post and bearing
stiffeners at the support have a limited ability to serve as an anchor for the horizontal
component of the tension-field force, Eurocode 3 accounts for its partial post-buckling
strength which is less than that of the interior panel.
First, consider the shear resistance of the girder web. The non-dimensional web
slenderness ratio,  w , for the end panel is defined in Eurocode 3 as
D
w  (1.20)
37.4t w k exact
where   235 / Fy , Fy is the yield stress (MPa), and kexact is presented in Eq. (1.18).
Based on  w , a reduction factor for the shear resistance of the web  w is determined
(see Table 1.1 and Figure 1.11) for both rigid end posts (interior panels) and non-rigid
end posts (end panels). In the table,  is the coefficient that considers the increase of
shear resistance at smaller web depth-to-thickness ratios. It is recommended that
 =1.2 for steel grades of S235 ( Fy  235 MPa = 34.1 ksi) to S460 ( Fy  460 MPa = 66.7
ksi) and  =1.0 for steel grades over S460. Then, the shear strength contribution from
the web ( Vbw, Rd ) is given by
 w Fyw Dt w
Vbw, Rd  (1.21)
3 M 1
11
where  M 1 is a factor for the partial resistance to instability which may be chosen in the
range 1.0 to 1.10.
Next, consider the shear resistance from the flanges. Eurocode 3 includes the
contribution of the flanges to the shear resistance. When the flange resistance is not
completely utilized in resisting the bending moment ( M Ed  M f , Rd ), the contribution
from the flanges should be obtained as follows:
b f t 2f Fyf   M Ed 
2

Vbf , Rd  1     (1.22)
c M 1   M f , Rd 
 
 
where b f is the effective flange width taken as no greater than 15t f on each side of
the web, t f is the flange thickness, Fyf is the flange yield stress, M Ed is the design
bending moment in a web panel, and M f , Rd is the moment of resistance of the cross
section consisting of the effective area of the flanges only, and c is the width of the
portion of the web between the plastic hinges, which is approximated by
 1.6b f t 2f Fyf 

c  d 0 0.25   (1.23)
 t w D 2 Fyw 
 
By summing up the shear resistances contributed from the web and flanges, the
design shear resistance of a panel Vb , Rd is given by
Fyw Dt w
Vb, Rd  Vbw, Rd  Vbf , Rd  (1.24)
3 M 1
The last term in Eq. (1.24) represents the shear yield strength.
1.8 Steel-Concrete Composite Plate Girder
There were no studies that directly address the shear strength of the end panels of
steel-concrete composite plate girders. All available literature on composite plate girders
addressed the behavior of interior panels in composite plate girders.
Allison et al. (1982) conducted an experimental study to investigate the
contribution of the concrete slab on the post-buckling strength of the steel web under
combined shear and negative bending. One steel plate girder specimen and five steel-
12
concrete composite plate girder specimens were tested. All specimens had transversely
stiffened webs and a web depth-to-thickness ratio of 130. Also, the ratios of steel girder
depth-to-slab thickness were 4.6 and 9.2. Test results showed that the increase in shear
strength of the interior panels due to the concrete slab varied from 2.8% to 6.6%.
Shanmugam and Baskar (2003) experimentally tested a total of six specimens
with a web width-to-depth ratio of 1.5. Among them, two specimens were bare steel
plate girders with web depth-to-thickness ratios of 250 and 150. Among four composite
girders, two composite girders had a reinforced concrete slab, and the other two
composite girders had a reinforced concrete slab with shear links. The ratio of girder
depth-to-slab thickness was 5 for all composite girders. The test results showed that the
increase of shear strength in the interior panels due to the concrete slab without and with
shear links was 76% and 122% for the web depth-to-thickness ratio of 250 and 40% and
68% for the ratio of 150, respectively.
Nie et al. (2004) tested a total of sixteen simply supported steel-concrete
composite specimens and two plain steel girders. Test results shows that the concrete
slab resisted 33% ~ 56% of the total ultimate shear. Based on the test results, an
empirical solution based on an additive formula of the shear strengths of the concrete slab
and steel girder was suggested for calculating the shear strength of the composite girder.
Liang et al. (2005) suggested that the contributions of the concrete slab and
composite action should be incorporated into the vertical shear strength in design codes.
A numerical study was carried out to quantify the contributions of the concrete slab and
composite action. They showed that the ultimate shear strength of the composite girder
is a function of the degree of shear connection and proposed an empirical solution.
Darehshouri et al. (2012) suggested an analytical model to predict the ultimate
shear strength of steel-concrete composite girders. Full composite action was assumed
in this model. The virtual work method was used to determine the post-buckling
strength for the steel girder. The plastic hinges were assumed to form in the flanges.
The top flange and concrete slab were assumed to be a composite flange with an
increased top flange flexural strength, resulting in the plastic hinge in top flange being
located further away from the intersection of the top flange and the transverse stiffeners.
The optimum value of the angle of the tension-field stress was determined by trial and
13
error to give the maximum value of shear strength. The concrete slab was separated into
two regions along its width to determine its shear strength: the first region of the slab had
the same width as the top flange of the steel girder and the second region was the
remaining slab width. A strut-tie model was used to determine the shear strength in the
first region, while the ACI (2005) equation for beam shear strength was used for the
second region. Note that they considered the full slab width as effective in resisting
shear. The ultimate shear strength of the composite girders was an additive formula of
the shear strengths of the concrete slab and steel beam.
1.9 Scope and Objectives
The main objective of this study is to investigate the shear behaviors of end panels
in steel plate girders experimentally and analytically. Specifically, the aim is to explore
the contributions from the tension-field action and the composite action of the concrete
slab; both of which are ignored in the current design specifications. Design equations
that include these effects and are practical for potential code adoption are to be proposed.
1.10 Organization of Report
This report begins with a general introduction and literature review of plate
girders, shear load transfer mechanism, and shear design philosophy for both interior and
end (exterior) panels. Since the test program consisted of two phases, this report
summarizes the research results of each pahse separately.
Phase 1
A total of four specimens (two steel and two steel-concrete composite girders)
were fabricated and tested in Phase 1. Chapters 2 and 3 describe the experimental test
program and results, while Chapters 4 and 5 discuss the finite element analysis results
including a correlation study with the test results, and parametric studies to identify
factors affecting the shear strength of end panels. Chapter 6 presents an analytical work
to develop practical shear design provisions for the end panels. Chapter 7 discusses
rehabilitation schemes of end panels and proposes an effective scheme.
Phase 2
14
Additional four specimens with different panel width-to-depth ratios and depth-
to-thickness ratio were tested in Pahse 2. Chapter 8 presents the experimental test
program and results. Chapter 9 provides the finite element analysis results of the Phase
2 test specimens and compars the test results with the predicted shear strengths to verify
the adequacy of the proposed shear strength equation.
Chapter 10 provides additional finite element analysis results for the effects of
longitudinal stiffeners, small flange areas, noncompact web, and unequal top and bottom
flanges. Chapter 11 presents the development of resistance factor based on the large-
scale test specimens from this research program (Phases 1 and 2) and the small-scale test
specimens from the literature. Chapter 12 summarizes and concludes the findings of
these studies.
15
Table 1.1: Eurocode 3 Shear Resistance Reduction Factor (w) of Web
Rigid End Post Non-Rigid End Post

w
(Interior Panel) (End Panel)
 w  0.83 /   
0.83 0.83
0.83 /    w  1.08
w w
1.37 0.83
 w  1.08
0.7   w w
16
Intermediate
d0 Stiffener
A B
D Interior
End Panel
Bearing Panel
Stiffener
A B
e
Elevation
One Side
Optional
Section A - A Section B - B
Figure 1.1: Typical Configuration of Steel Plate Girder
17
Plate Girder
Figure 1.2: Typical Plate Girder (Caltrans Br. No. 50-0316)
18

σ2 = 
45°
σ1 = 
(a) Before Buckling
Pratt Truss
Plate Girder
(b) After Buckling (adapted from Yen and Basler, 1962)
Figure 1.3: Plate Girder Shear Load-Carrying Mechanism
19
c1

c2
(a) Basler (1961a) (b) Takeuchi (1964)
do/2
do/2
(c) Fujii (d) Chern and Ostapenko (1969)
c c
c c
(e) Komatsu (1971) (f) Rockey and Skaloud (1972)
Figure 1.4: Various Tension-Field Models for Interior Panels
20
c c
D/2

c c
(g) Herzog (1974) (h) Porter et al. (1975)
(i) Hoglund (1997)
Figure 1.4: Various Tension-Field Models for Interior Panels (continued)
21
A B
V D/2
C C
D/2 V
do
Fs
A B
t
C C
O Vtf
Vtf
2 2
Fw Fw
Ff F f  F f
Figure 1.5: Free-Body Diagram from Basler (adapted from Basler, 1961a)
y
y'
( cr ) x'

' 
x
'

( cr )
(a) Original Coordinate System (b) Rotated Coordinate System
Figure 1.6: Plane Stress in Pure Shear
22
y'
x'
y t

cr sin 
 
cr cos 
(a) due to shear buckling stress (b) due to membrane tensile stress
Figure 1.7: State of Stresses due to Shear Buckling and Tension-Field Action
1

1  Fyw   2 3  1 
B
Fyw
A
Fyw / 3
2
 Fyw / 3
Figure 1.8: Von Mises Yield Condition (adapted from Basler, 1961a)
23
c1
c2
Figure 1.9: End Panel Tension-Field Model (adapted from Huslid and Rockey, 1979)
(a) Rigid End Post (b) Non-Rigid End Post
Figure 1.10: Different End Supports for Plate Girders
24
w
Rigid End Post
Non-Rigid End Post

w
0.83 1.08

Figure 1.11: Eurocode 3 Shear Resistance Factor (w) of Web
25
2. TEST PROGRAM (PHASE 1)
2.1 General
The primary objective of the testing program is to determine the ultimate shear
resistance of the end panels of steel-concrete composite plate girders. In Phase 1 test
program, a total of four test specimens were fabricated by a certified commercial
fabricator and were designated as Specimens SG1, SG2, CG1 and CG2. The notation
“SG,” which stands for “Steel Girder,” represents a steel girder without a concrete slab.
The notation “CG,” which stands for “Composite Girder”, represents a steel girder with a
concrete slab. Specimens SG1 and SG2 were tested first, then Specimens CG1 and CG2,
which were nominally identical to the first two specimens, were subsequently tested to
study the effects of the presence of a concrete slab.
2.2 Configuration and Fabrication of Test Specimens
2.2.1 Steel Girder Specimens (SG1 and SG2)

Table 2.1 shows the dimensions of two sample steel bridge girders provided by
the California Department of Transportation (Caltrans) for Phase 1 test program. Table
2.2 shows the specimens dimensions used in Phase 1 testing. Depth-to-thickness ratio
of 210 and 280 were used for Specimens SG1 and SG2, respectively. A panel width-to-
depth ratio of 0.5 was used for both Specimen SG1 and SG2. The actual plate
thicknesses of the specimens were measured and are summarized in Table 2.3.
For testing purpose, the sample girder dimensions from Caltrans were slightly
modified. The width and thickness of the plates used for the top and bottom flanges
of each girder were identical. In consideration for the capacities of the laboratory
actuators, the sample girders were scaled down by 2/3 and 1/2 for Specimens SG1 and
SG2, respectively.
Figure 2.1 shows the configuration of the test specimens. Flexural effects were
minimized by designating a short span in order to study the shear behavior of the
specimens.
26
2.2.2 Steel-Concrete Composite Girder Specimens (CG1 and CG2)
The girders of Specimens CG1 and CG2 were nominally identical to those of
Specimens SG1 and SG2, respectively. Additionally, headed shear studs were welded
to the top flanges of Specimens CG1 and CG2 to incorporate a concrete slab. A total
slab thickness (tc) and haunch depth (dh) are shown in Table 2.2. Figure 2.2 and Figure
2.3 show the size of the shear studs, the geometry of concrete slab, and the steel
reinforcement layout for Specimens CG1 and CG2, respectively.
Figure 2.4 shows the construction procedure of the concrete slab. First,
formwork for the slab was built and reinforcements were placed. Concrete was then
poured for Specimens CG1 and CG2 at the same time. A needle vibrator was used to
compact the concrete during pouring. Then, the concrete was covered by plastic sheets
to reduce the surface evaporation of water during curing.
2.3 Material Properties
2.3.1 Structural Steel Plates

ASTM A709 Grade 50 steel was specified for the flanges and web of the
specimens, while ASTM A36 steel was specified for the bearing and intermediate
transverse stiffeners to match those of the sample girders provided by Caltrans. Tensile
coupons were cut from the same plates for material testing. The standard tensile coupon
test results are shown in Figure 2.5. Average values of the measured mechanical
properties of the component plates are summarized in Table 2.4.
2.3.2 Reinforcement and Concrete

Grade 40 reinforcement (No. 3 and No. 4 for Specimen CG1, and No. 3 for
Specimen CG2) with a minimum yield strength of 40 ksi and a minimum ultimate
strength of 70 ksi was used. Although No. 2 reinforcement bars were specified for the
upper longitudinal direction in the concrete slab of Specimen CG2, No. 2 wires were used
instead due to availability issues. Three samples of each size of reinforcement were
tested. The tensile stress versus strain curves are shown in Figure 2.6. The average
values of the measured mechanical properties are also summarized in Table 2.5.
27
A design concrete compressive strength of 4,000 psi with 3/8 in. river aggregate
was specified for the concrete slab. A slump test was performed before casting the
concrete; the measured slump was 3.5 in. A total of 18 cylinders were cast and cured
under the same conditions as the test specimens to obtain the compressive and split
tensile strengths of the concrete. Each three-cylinder set was tested for compressive
strength on the 7th, 14th, 22nd (day of testing for Specimen CG1), and 30th day (day of
testing for Specimen CG2), while another three-cylinder set was tested for spliting tensile
strength on the day of testing for each composite specimen. Table 2.6 summarizes the
average values of compressive and splitting tensile strengths from testing. Figure 2.7
plots the relationship between concrete compressive strength and curing time.
2.3.3 Shear Studs

Type B headed shear stud connectors made from ASTM A108 material with a
minimum yield strength of 51 ksi and a minimum ultimate strength of 65 ksi was used for
the interconnection of steel girder and concrete slab.
2.4 Initial Geometrical Imperfections
Initial geometrical imperfections of the steel girders were measured before testing.
Figure 2.8 shows the measurement locations, and Figure 2.9 plots the measured
imperfections. The measured web minimum and maximum imperfection values of each
panel are also summarized in Table 2.7.
2.5 Test Setup and Loading Procedure
The overall configuration of the test setup is shown in Figure 2.10 for Specimens
SG1 and SG2, and Figure 2.11 for Specimens CG1 and CG2. Each specimen was
simply supported and monotonically loaded by two 500-kip hydraulic actuators. Figure
2.12 shows a close-up of the support detail. The load was applied to the top flange
through a W36302 loading beam at mid-span of the specimen, as shown in Figure 2.13.
To easily detect yielding and buckling of the specimen, whitewash and 2 in. grid lines
were applied to the web of the specimen. Testing was initially conducted in a load-
control mode up to 90% of the nominal shear strength per AASHTO Specifications
28
(2014). Thereafter, testing was conducted in a displacement-control mode with a small
increment of mid-span deflection until failure occurred.
Figure 2.14 shows the lateral bracing system. To avoid any unexpected failure
mode, L64½ lateral bracing was bolted to one side of the web at two locations along
the girder length as shown in Figure 2.1. The other side of the lateral bracing was
bolted to a WT685 connection piece that was bolted to a W1897 support column. To
accommodate the vertical deflection of the plate girder, the hole on the WT685
connection piece was slotted vertically; two bolts at this location were finger tightened.
2.6 Instrumentation
Test specimens were instrumented with a combination of displacement

transducers, strain gage rosettes, and uni-axial strain gages to measure global and local
responses at specific locations. Most displacement transducers measured the
deformation of the flanges, web, and bearing stiffeners. One displacement transducer
measured vertical displacement at the mid-span of each specimen and was used as the
feedback sensor for actuator control. Strain gage rosettes measured the principal strains
on the web panel, while uni-axial strain gages monitored the strains of flanges and
bearing stiffeners. Load cells mounted on the actuators measured applied load. The
instrumentation plan for each specimen is provided in Appendix A.
29
Table 2.1: Sample Caltrans Steel Bridge Girder Dimensions
Width- Depth-
Section Depth Thickness
Bridge
Ratio, Ratio,
No.
Top Flange Web Bottom Flange
d0/D D/tw
(in.) (in.) (in.)
56-0571 20×1 83×3/8 24×1-1/8 0.58 221
39-0164L 28×1-1/4 106×3/8 28×1-1/4 0.45 283
Table 2.2: Specimen Dimensions

Top and Bottom Total Slab
Width- Depth-
Web Flanges Thickness
Spec. Depth Thickness Scale
Depth, (tc)/ Haunch
No. Ratio, Ratio, Thickness, Width, Factor
Depth (dh)
d0/D D/tw D (in.) tf (in.) bf (in.) (in.)
SG1 No Slab
0.5 210 52.5 0.75 16.0 2/3
CG1 8.0 / 3.0
SG2 No Slab
0.5 280 52.5 0.625 14.0 1/2
CG2 6.0 / 2.0
Table 2.3: Steel Plate Thicknesses

Specimen Nominal Measured
Component
No. Thickness (in.) Thickness (in.)
Flanges 0.75 0.755
SG1 / CG1
Web 0.25 0.247
Flanges 0.625 0.620

SG2 / CG2
Web 0.1875 0.183
Bearing Stiffener 1 1.009

SG1 / CG1
SG2 / CG2
Intermediate Stiffener 0.5 0.496
30
Table 2.4: Average Mechanical Properties of Steel Plates
Tensile
Specimen Steel Component / Yield Stress Elongation
Strength
No. Grade Heat Number (ksi) (%)
(ksi)
Flanges / 39.2b
57.8 (54.0)a 79.7 (75.0)
W9I542 (26)c
SG1, CG1
Web / 32.2b
60.3 (58.5) 76.9 (74.5)
A709 NW7384 (24)c
Gr.50
Flanges / 39.2b
51.5 (54.0) 71.9 (71.5)
NW5341 (25.5)c
SG2, CG2
Web / 29.2b
57.4 (61.0) 80.0 (79.0)
W2K713 (24)c
Bearing
38.4b
Stiffener / 41.1 (41.9) 69.2 (67.5)
(33)b
SG1, CG1, NW8339
A36
SG2, CG2 Intermediate
39.7b
Stiffener / 47.6 (49.0) 68.0 (66.0)
(28)c
E3B184
Note:
a. Values in parentheses are based on Certified Mill Test Reports, others from tensile
coupon tests at UCSD.
b. Elongation is based on 2 in. gage length.
c. Elongation is based on 8 in. gage length.
Table 2.5: Average Mechanical Properties of Steel Reinforcements
Yield Stress Tensile Strength

Bar Size
(ksi) (ksi)
No. 2 60.8 71.0
No. 3 48.5 69.0
No. 4 47.8 69.7
31
Table 2.6: Average Mechanical Properties of Concrete
Curing Period Compressive Strength, Splitting Tensile
(Days) f’c (ksi) Strength, ft (ksi)
7 4.1 NA
14 4.7 NA
22 (Testing of CG1) 5.0 0.47
30 (Testing of CG2) 5.4 0.43
Table 2.7: Measured Initial Web Imperfections (in.)
Specimen No.
Measured Initial
Imperfections
SG1 SG2 CG1 CG2
Min. -0.109 -0.2 -0.203 -0.125

Panel 1
Max. 0.016 0 0.109 0.125
Min. -0.141 -0.25 -0.266 -0.047

Panel 2
Max. 0.078 0.25 0.109 0.266
Min. -0.156 -0.35 -0.156 -0.109

Panel 3
Max. 0.031 0.15 0.094 0.141
Min. -0.078 -0.25 -0.172 -0.125

Panel 4
Max. 0.109 0.15 0.078 0.125
32
434"X1" Gusset Plate for 16"X34" Flange Typ.
Lateral Bracing, Typ.
10"
A A
741"X1" 434"X21"
4'-6" Typ. Typ. 5221"X41" Web
1"
2'-24 1"
2'-24 1' 8" 8" 1' 1"
2'-24 1"
2'-24
4" 4"
12'-9"
1'-4"
View A - A
(a) Specimen SG1
10"
B B
638"X1" 434"X21"
4'-53
4" Typ. Typ. 5221"X16
3
" Web
1"
2'-24 1"
2'-24 1' 8" 8" 1' 1"
2'-24 1"
2'-24
4" 12'-9" 4"
1'-2"
View B - B
(b) Specimen SG2
Figure 2.1: Configuration of Steel Girder Test Specimens
33
5
8" Dia. and 521" Long Headed Shear Stud, Typ.
1"
82 7 Nos. @ 9" 1"
94 1"
92 1"
94 7 Nos. @ 9" 1"
82
10"
434"X1"
Typ.
4'-6" 741"X1" 434"X21" 5221"X41" Web

Typ. Typ.
16"X34" Flange
Typ.
1"
2'-24 1"
2'-24 1' 8" 8" 1' 1"
2'-24 1"
2'-24
4" 4"
12'-9"
No.3, Typ.
No.4 @ 658"
1" Clr.
1'-1" 1'-1"
7" Clr.
5" 8
3"
1'-115
8"
3" No.4, Typ.
3" 18
6'
No.3
(Only at Both Ends)
Reinforcement Layout
Figure 2.2: Specimen CG1: Configuration of Composite Girder Test Specimens
34
1
1"
84 8 Nos. @ 821" 83 83 8 Nos. @ 821" 1"
4" 4" 84
10"
434"X1"
Typ.
4'-53
4"
638"X1" 434"X21" 5221"X16
3
" Web
Typ. Typ.
14"X58" Flange
Typ.
1"
2'-24 1"
2'-24 1' 8" 8" 1' 1"
2'-24 1"
2'-24
4" 12'-9" 4"
No.2, Typ.
No.3 @ 6"
3" Clr.
1"
112 1"
112 4
4"
5" Clr.
8
2"
1"
1'-92
1" No.3, Typ.
6" 12
6'
35
CG1 CG2
(a) Reinforcement Layout
(b) Pouring Concrete (c) Concrete Cylinders
CG1 CG2
(d) Curing
Figure 2.4: Concrete Slab Construction
36
Coupon 1 Coupon 2
100 100
80 80
Stress (ksi)
Stress (ksi)
60 60
Fy = 58.2 ksi Fy = 57.4 ksi
40 40
Fu = 79.5 ksi Fu = 79.9 ksi
20 20
Elongation = 38.4% Elongation = 39.9%
0 0
0.0 0.05 0.10 0.15 0.20 0.0 0.05 0.10 0.15 0.20
Strain (in/in) Strain (in/in)
(a) Specimen SG1 Flange
Coupon 1 Coupon 2
100 100
80 80
Stress (ksi)
60 Stress (ksi) 60
Fy = 59.8 ksi Fy = 60.8 ksi
40 40
Fu = 77.0 ksi Fu = 76.8 ksi
20 20
0 0
0.0 0.05 0.10 0.15 0.20 0.0 0.05 0.10 0.15 0.20
(b) Specimen SG1 Web
Coupon 1 Coupon 2
100 100
80 80
Stress (ksi)
Stress (ksi)
60 60
Fy = 50.5 ksi Fy = 52.5 ksi
40 40
Fu = 71.9 ksi Fu = 71.9 ksi
20 20
0 0
0.0 0.05 0.10 0.15 0.20 0.0 0.05 0.10 0.15 0.20
(c) Specimen SG2 Flange

Figure 2.5: Steel Plate Stress versus Strain Curves
37
Coupon 1 Coupon 2
100 100
80 80
Stress (ksi)
Stress (ksi)
60 60
Fy = 57.4 ksi Fy = 57.4 ksi
40 40
Fu = 79.9 ksi Fu = 80.0 ksi
20 20
0 0
0.0 0.05 0.10 0.15 0.20 0.0 0.05 0.10 0.15 0.20
(d) Specimen SG2 Web
Coupon 1 Coupon 2
80 80
60 60
Stress (ksi)
Stress (ksi)
40 Fy = 41.8 ksi 40 Fy = 40.3 ksi
Fu = 68.8 ksi Fu = 69.5 ksi
20 20
0 0
0.0 0.05 0.10 0.15 0.20 0.0 0.05 0.10 0.15 0.20
(e) Bearing Stiffener
Coupon 1 Coupon 2
80 80
60 60
Stress (ksi)
Stress (ksi)
40 Fy = 48.1 ksi 40 Fy = 46.7 ksi

Fu = 67.3 ksi Fu = 68.6 ksi
20 20
0 0
0.0 0.05 0.10 0.15 0.20 0.0 0.05 0.10 0.15 0.20
(f) Intermediate Stiffener

Figure 2.5: Steel Plate Stress versus Strain Curves (continued)
38
80
60
Stress (ksi)
Coupon Fy Fu
40 No. (ksi) (ksi)
1 60.3 71.1
2 61.6 71.4
20
3 60.6 70.5 Coupon 1
AVG. 60.8 71.0 Coupon 2
Coupon 3
0
0.0 0.05 0.10 0.15 0.20
Strain (in/in)
(a) No. 2 Reinforcement
80
60
Stress (ksi)
Coupon Fy Fu
40 No. (ksi) (ksi)
1 48.9 69.8
2 49.3 69.8
20
3 47.3 70.0 Coupon 1
AVG. 48.5 69.9 Coupon 2
Coupon 3
0
0.0 0.05 0.10 0.15 0.20 0.25 0.30 0.35
Strain (in/in)
(b) No. 3 Reinforcement

Figure 2.6: Steel Reinforcement Stress versus Strain Curves
39
80
60
Stress (ksi)
Coupon Fy Fu
40 No. (ksi) (ksi)
1 47.7 68.7
2 49.1 71.1
20
3 46.6 69.4 Coupon 1
AVG. 47.8 69.7 Coupon 2
Coupon 3
0
0.0 0.05 0.10 0.15 0.20 0.25 0.30 0.35
Strain (in/in)
(c) No. 4 Reinforcement
Figure 2.6: Steel Reinforcement Stress versus Strain Curves (continued)
6
Concrete Compressive Strength (ksi)
0
0 10 20 30 40
Curing Time (day)
Figure 2.7: Concrete Cylinder Compressive Strength versus Curing Time
40
E
Panel 1 Panel 2 Panel 3 Panel 4
Imperfection Measurement Point, Typ.
1"
4'-42
1"
2'-24 1"
2'-24 1' 8" 8" 1' 1"
2'-24 1"
2'-24
4" 12'-9" 4"
Elevation
Figure 2.8: Measured Locations for Initial Geometrical Imperfections of Steel
Girders
l2 l 4
Pane Pane
l1 l 3
Pane Pane
(a) SG1
Figure 2.9: Initial Geometrical Imperfection Measurements
41
l2 l4
Pane Pane
1 l3
Pane
l Pane
(b) SG2
2 l4
Pane
l Pane
l1 l3
Pane Pane
(c) CG1
l2 l4
Pane Pane
l1 l 3
Pane Pane
(d) CG2
Figure 2.9: Initial Geometrical Imperfection Measurements (continued)
42
A
Bracing Bracing
2" Thk. Plate
Steel Loading Beam

(W36X302)
12'-1"
Concrete
Block, Typ.
Strong Floor
A
(a) Elevation
2" Thk. Plate, Typ.

Steel Loading Beam
(W36X302)
500-kip Actuator,Typ.
Steel Mounting
Concrete Plate, Typ.
Block, Typ.
Strong Floor
(b) View A - A
Figure 2.10: Test Setup: Specimens SG1 and SG2
43
A
Bracing Bracing
2" Thk. Plate

Steel Loading Beam
(W36X302)
12'-1"
Concrete
Block, Typ.
Strong Floor
A
(a) Elevation
2" Thk. Plate, Typ.

Steel Loading Beam
(W36X302)
500-kip Actuator,Typ.
Steel Mounting
Concrete Plate, Typ.
Block, Typ.
Strong Floor
(b) View A - A
Figure 2.11: Test Setup: Specimens CG1 and CG2
44
Figure 2.12: Support Detail
Steel Loading Beam
Figure 2.13: Steel Loading Beam
45
W18X97 Guide Column
Lateral Bracing Connection (WT6X85)
Lateral Bracing (L6X4X21 )
Test Specimen
Slotted
Hole
Strong Floor
Lateral Bracing Lateral Bracing
Figure 2.14: Test Setup: Beam Lateral Bracing
46
3. TEST RESULTS (PHASE 1)
3.1 Specimen SG1
Figure 3.1 shows the test setup prior to testing. The relationship between the
applied load and mid-span deflection is shown in Figure 3.2. The theoretical stiffness
was computed by using Timoshenko beam theory. The mid-span deflection of simply
supported beams can be determined by combinating the deformations due to flexure and
shear as follows (see Figure 3.3).
PL3 P L
   f  s     (3.1)
48EI 2GAs  2 
where
P = applied load (kips),
L = beam length (in.),
E = modulus of elasticitity (ksi),
I = moment of inertia (in.4),
G = shear modulus (ksi), and
As = shear area (in.2).

The theoretical stiffness computed from Eq. (3.1) is also plotted in Figure 3.2. It
matches well with the stiffness computed from test data.
Figure 3.4 shows the test specimen at event A, which corresponds with the
AASHTO nominal shear strength for end panels. Minor web buckling was visible at
both end panels. When the maximum applied load was reached (event B), web buckling
became significant and minor yielding at both end panels was observed, as
evidenced by the flaking of the whitewash (see Figure 3.5 and Figure 3.6). After
maximum load was achieved, the strength degraded slowly and plastic hinges formed in
the bearing stiffeners and both flanges of the west end panel. The deformed shape at
event C is shown in Figure 3.7. The front and back sides of the west end panel after the
specimen was unloaded and intrumetation removed are shown in Figure 3.8. The
observed plastic hinges are also marked in the figure.
47
3.2 Specimen SG2
A plot of the applied load versus mid-span deflection is shown in Figure 3.9.
The stiffness from the test started to deviate from the theoretical stiffness when the
AASHTO nominal shear strength (event A) was exceeded.
In this specimen, minor web buckling was observed in all four panels at event A
(see Figure 3.10). At maximum load (event B), yielding in both end panels was
observed (see Figure 3.11). A close-up view of both end panels at event B in Figure
3.12 shows web local buckling. After achieving the maximum load, the strength
degradation was gradual as the displacement was increased. The west end panel failed
in this specimen and the deformed shape with plastic hinges on boundary members at
event C is shown in Figure 3.13. At the completion of testing, failure modes included
plastic hinges on the boundary members and yielding and web buckling on the front and
back sides of west end panel as shown in Figure 3.14. As seen in Figure 3.15, the
failure modes of both Specimens SG1 and SG2 are very similar.
3.3 Specimen CG1
Specimen CG1 was nominally identical to Specimen SG1, except that it

incorporated a concrete slab. Figure 3.16 shows the relationship between applied load
and mid-span deflection. To calculate the theoretical stiffness of the steel-concrete
composite specimen, the moment of inertia (I) in Eq. (3.1) was replaced by the effective
moment of inertia (Ieff) to account for the effects of partial composite beam action (AISC
2010).
I eff  I s 
 Q In
 Is  (3.2)
tr
Cf
where
Q n = summation of the stud strength for partial composite action (= 639 kips),
Cf = horizontal force for full composite action (= 1,807 kips),
I tr = transformed section moment of inertia (= 42,408 in.4), and
Is = moment of inertia of the bare steel beam (= 20,029 in.4).
48
The theoretical stiffness thus calculated is also shown in Figure 3.16.
Specimen CG1 showed similar behavior to Specimens SG1 and SG2. At event
A, minor web buckling developed in the west end panel. Then yielding in the end
panels was observed at a mid-span deflection of 0.2 in. (see Figure 3.17 and Figure 3.18).
At that point, the recorded applied load was 632 kips, which is very close to the
maximum applied load (= 630.5 kips) for Specimen SG1. After that point, the stiffness
softened and further loading resulted in flexural cracks on the concrete haunch at event B
[Figure 3.19(a)]. This flexural crack caused a small drop in the applied load. It was
also observed that a shear crack occurred near bearing stiffeners [Figure 3.19(b)]. As
the mid-span deflection increased, the cracks widened and the number of cracks
increased. The crack patterns are summarized in Figure 3.19. The failure mode of the
specimen at the maximum deflection (= 1.8 in., event C) is shown in Figure 3.20. At the
completion of testing, concrete spalling at the top surface of the concrete slab above the
end panel is shown in Figure 3.21(a). Figure 3.21(b) and Figure 3.21(c) show
significant yielding in the web along the panel diagonal and plastic hinges developed in
the flanges and bearing stiffeners.
3.4 Specimen CG2
Specimen CG2 was nominally identical to Specimen SG2, except that CG2
and mid-span deflection.
At event A, web buckling was visible in all four panels (see Figure 3.23).
Yielding in both end panels was observed when the maximum load was achieved (event
B). The yielding pattern is shown in Figure 3.24. At this point, one flexural crack in
the concrete appeared near the middle of the end panel, and one shear crack appeared
near the bearing stiffeners. Figure 3.25 shows the concrete crack patterns observed
during the test. Figure 3.26 shows the failure mode of the specimen at event C. Figure
3.27 shows the front and back sides of the west end panel at the completion of testing.
Plastic hinges that developed in flanges and bearing stiffeners are also marked in the
figure.
49
3.5 Recorded Response
Figure 3.28 shows the locations of the displacement transducers that measured the
web out-of-plane deformations of each specimen. L2, L3, and L4 were installed on the
back side of the east end panel, while L7, L8, L9 were placed on the back side of the west
end panel. Figure 3.29 shows the web out-of-plane deformation of Specimen SG1.
The expected buckling load predicted per AASHTO Specifications (2014) is also
presented in the figure. In the figure, displacements recorded by L2 in the east end
panel and L7 to L9 in the west end panel start changing in slope near the predicted
buckling strength, and the web out-of-plane deformations rapidly increased. This
indicates that the buckling strength predicted by AASHTO Specifications is reasonable.
In Specimen SG2, similar responses are observed in the east end panel, but not in the
west end panel (see Figure 3.30). Figure 3.31 and Figure 3.32 show the web out-of-
plane deformation of Specimens CG1 and CG2.
Figure 3.33 shows the locations and orientations of the strain gage rosettes for all
specimens except Specimen SG1. Specimen SG1 had the same arrangement of the
rosettes only in the west panel (see Appendix A for the locations of the rosettes in the
east panel). All strain gage rosettes were installed on the back side of the panel web.
From the rosette data, principal strains and directions were computed. Figure 3.34 plots
the principal tensile strains across the panel diagonal of both east and west end panels at
the ultimate load (Event B). The horizontal reference line in the figure represents the
strain value corresponding to the tension-field stress (  t ) calculated by using Eq. (1.10).
The ratio of the shear buckling strength to shear yield strength ( C   cr /  y ) in Eq. (1.10)
is determined by Eqs. (1.15) to (1.17).

The tensile principal strains recorded from five middle strain gage rosettes in each
panel (R05 to R09 in the east end panel and R18 to R22 in the west end panel) are close
to or beyond the strain. Table 3.2 and Table 3.3 summarize the principal tensile
directions of five middle strain gage rosettes of each section at ultimate load (Event B).
The recorded principal angles varied but most of them were within ± 10° of 63.4°, which
is the angle of the panel diagonal with respect to the flange. More principal strain plots
associated with different displacement levels are presented in Appendix B. Recorded
50
response plots from both displacement transducers and uni-axial strain gages are also
shown in Appendix B.
3.6 Comparison of Test Results
3.6.1 Failure Mechanism

Coincidentally, all specimens failed in the west end panel and showed similar
behaviors. Minor web buckling occurred after the shear buckling strength was achieved.
Subsequent yielding in the web appeared. After the ultimate strength was reached,
significant yielding occurred along the diagonal of the web panel, and was accompanied
by the development of plastic hinges in both flanges and bearing stiffeners. Shear
cracks also appeared in the concrete slab for composite specimens.
After completing the tests, plastic hinge locations in the top flange and bearing
stiffeners for each test specimen were measured and are summarized in Figure 3.35. For
the composite specimens, a portion of the concrete slab was cut out to investigate shear
crack propagation (see Figure 3.36). The measured shear crack angles with respect to
the concrete slab were 17° and 14° for CG1 andCG2, respectively (see Figure 3.37).
3.6.2 Comparison with AASHTO Shear Strength

Table 3.1 summarizes the end panel shear strengths based on both AASHTO
Specifications (2014) and experimental results. The shear overstrengths for each
specimen, which is defined as the ratio of the experimental strength to the AASHTO
nominal strength, are listed in the last column of the table. It is clear that AASHTO
Specifications (2014), which ignores tension-field action in end panels, underestimates
the shear strength of the end panels considerably.
Figure 3.38 shows the relationship between the normalized ultimate shear strength
and the web depth-to-thickness ratio. In the plot, based on the actual web yield stress
from the coupon tests, two AASHTO nominal strength curves are presented: one that
excludes tension-field action for end panels, and another that includes tension-field action
for interior panels. Experimental shear strengths were found to be between the
AASHTO strengths with and without tension-field action. Therefore, this concludes
that partial tension-field action develops in the end panels.
51
3.6.3 Contribution of Concrete Slab
In current AASHTO Specifications (2014), the composite action provided by the
concrete slab is ignored in evaluating the shear strength. Specimens CG1 and CG2
showed a higher shear capacity than their counterparts (SG1 and SG2), respectively.
The increase in shear strength due to the contribution from concrete slab is about 12% in
both composite specimens.
52
Table 3.1: Comparison of End Panel Shear Strengths
Web Depth- AASHTO Predicted Experimental
Specimen Shear
Thickness Nominal Strength Shear Strength
No. Overstrength
Ratio (kips) (kips)
SG1 315.3 1.60
210 196.5
CG1 352.6 1.79
SG2 190.3 2.30

280 82.9
CG2 212.1 2.56
Table 3.2: Prinicipal Strain Direction at Ultimate Load (Steel Plate Girders)
Specimen Rosette Panel Diagonal Principal Tensile

No. Number Angle (degree) Strain Angle (degree)
West R18 44
Panel R19 52
SG1 R20 54
R21 47

R22 10
East R09 39
Panel R08 3
R07 63 57
 R06 62
R05 61
SG2
West R18 74
Panel R19 51
R20 60
R21 62

R22 56
53
Table 3.3: Prinicipal Strain Direction at Ultimate Load (Composite Plate Girders)

East R09 51
Panel R08 67
R07 76
 R06 61
R05 58.
CG1
West R17 59
Panel R19 45
R20 57
R21 74

R22 60
63
East R09 74
Panel R08 38
R07 58
 R06 59
R05 55
CG2
West R18 46
Panel R19 59
R20 70
R21 65

R22 79
54
E
Figure 3.1: Specimen SG1: Test Setup
55
700 350
B Predicted Stiffness
End Panel Shear Strength (kips)

600 300
C
Applied Load (kips)
500 250
400 200
A
300 150
AASHTO
200 Nominal Strength 100
100 50
0 0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
Mid-Span Deflection (in.)
Figure 3.2: Specimen SG1: Applied Load versus Mid-Span Deflection
P P P
= f + s

Figure 3.3: Elastic Timoshenko Beam Deformation
56
E
Figure 3.4: Specimen SG1: Deformed Shape at Event A
Figure 3.5: Specimen SG1: Web Local Buckling at Event B
57
Yielding
Yielding
(a) West End Panel (b) East End Panel

Figure 3.6: Specimen SG1: Yielding Patterns at Event B (Back Side)
Figure 3.7: Specimen SG1: Failure Mode at Event C
58
Plastic Plastic
Hinge Hinge
(a) Front Side (b) Back Side

Figure 3.8: Specimen SG1: Failure Mode after Completion of Test (West End Panel)
500 250

Predicted Stiffness
400 200
Applied Load (kips)
B C
300 150
200 100
A
100 50
AASHTO
Nominal Strength
0 0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
59
Figure 3.10: Specimen SG2: Minor Web Local Buckling at Event A
Yielding
West End East End

Panel Panel
Figure 3.11: Specimen SG2: Yielding Pattern at Event B
60
(a) West End Panel (b) East End Panel
61
Plastic Plastic
Hinge Hinge

Figure 3.14: Specimen SG2: Failure Mode after Completion of Test (West End Panel)
SG1 SG2
Figure 3.15: Failure Modes of SG1 and SG2
62
800 400
Predicted Stiffness

700
B
600 300
Applied Load (kips)
C
500
400 200
A
300
AASHTO
200 100
Nominal Strength
100
0 0
0.0 0.5 1.0 1.5 2.0
Figure 3.16: Specimen CG1: Applied Load versus Mid-Span Deflection
63
Figure 3.17: Specimen CG1: Web Yielding and Local Buckling at Δ=0.2 in.
64
E
Yielding
(a) Front View
Yielding
Yielding
(b) West End Panel (c) East End Panel

Figure 3.18: Specimen CG1: Yielding Patterns at Event B (Back Side)
65
Shear Crack
(a) at Δ=0.35 in. (Event B) (b) at Δ=0.5 in.
(c) at Δ=0.7 in. (d) at Δ=1.0 in.
a
(e) at Δ=1.8 in. (f) View a - a
Figure 3.19: Specimen CG1: Propagation Concrete Slab Cracks
66
E
Figure 3.20: Specimen CG1: Failure Mode at Event C
(a) Concrete Spalling

Figure 3.21: Specimen CG1: Failure Mode after Completion of Test
67
Plastic Plastic
Hinge Hinge
(b) Front Side of West End Panel (c) Back Side of West End Panel
Figure 3.21: Specimen CG1: Failure Mode after Completion of Test (continued)
68
500 250
Predicted Stiffness

Applied Load (kips) 400 B 200
C
300 150
200 100
A
100 AASHTO Predicted 50
Nominal Strength
0 0
0.0 0.5 1.0 1.5 2.0
(a) West End Panel

Figure 3.23: Specimen CG2: Web Local Buckling at Event A
69
(b) East End Panel
Figure 3.23: Specimen CG2: Web Local Buckling at Event A (continued)
Figure 3.24: Specimen CG2: Yielding Pattern at Event B
70
(a) at Δ=0.3 in. (Event B) (b) at Δ=0.5 in.
(c) at Δ=0.8 in. (d) at Δ=1.0 in.
(e) at Δ=1.8 in. (f) View a - a

Figure 3.25: Specimen CG2: Concrete Slab Crack Propagation
71
E
Plastic Plastic
Hinge Hinge

Figure 3.27: Specimen CG2: Failure Mode after Completion of Test (West End Panel)
72
1"
1'-14
L2
L2
L3 10"
L3
L4 1'-4"
L4
1"
1'-14
(a) East End Panel
1"
1'-14
L7
L7
L8
10"
L8
1'-4"
L9
L9
1"
1'-14
(b) West End Panel
Figure 3.28: Typical Locations of Displacement Transducers (Back Side)
73
700
600
Applied Load (kips) 500
400
300
AASHTO
200 Nominal
Strength L2
100 L3
L4
0
-0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8
Web Out-of-Plane Deformation (in.)
(a) East End Panel
700
600
500
Applied Load (kips)
400
300
AASHTO
200 Nominal
Strength L7
100 L8
L9
0
-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5
(b) West End Panel
Figure 3.29: Specimen SG1: Web Out-of-Plane Deformation
74
500
400
Applied Load (kips)
AASHTO
300 Nominal
Strength
200
100 L2
L3
L4
0
-0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8
(a) East End Panel
500
400
Applied Load (kips)
AASHTO
300 Nominal
Strength
200
100 L7
L8
L9
0
-2 -1 0 1 2
(b) West End Panel
75
800
700 Note:
L2 and L3 Malfunctioned.
500
400
300
200 AASHTO
Nominal L2
100 Strength L3
L4
0
-0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8
(a) East End Panel
800
L7
L8
700 L9
600
Applied Load (kips)
500
400
300
200 AASHTO
Nominal
100 Strength
0
-3 -2 -1 0 1 2 3
(b) West End Panel
Figure 3.31: Specimen CG1: Web Out-of-Plane Deformation
76
500
Note:
L4 Malfunctioned.
400
Applied Load (kips)
AASHTO
300 Nominal
Strength
200
100 L2
L3
L4
0
-0.8 -0.6 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8
(a) East End Panel
500
400
Applied Load (kips)
AASHTO
300 Nominal
Strength
200
100 L7
L8
L9
0
-2 -1 0 1 2
(b) West End Panel
Figure 3.32: Specimen CG2: Web Out-of-Plane Deformation
77
Rosettes
along This
Section
R11
R9
90 R7
R10
R5 R8
R3
R6
R4
3"
8@ b
r
y
Rosette
Orientation
(a) East End Panel
Rosettes
along This
Section
R16
R18
R20 90
R17
R19 R22
R21 R24
8@ R23
y 3"
r
b
Rosette
Orientation
(b) West End Panel
Figure 3.33: Typical Locations of Strain Rosettes in End Panels (Back Side)
78
0.003
West End Panel
Principal Strain (in./in.) 0.002
0.001
Intermediate Bearing
Stiffener Side Stiffener Side
0.0
R16
R18
R19
R21
R22
R23
R17
R24
R20
West Panel:
Rosette Location
(a) Specimen SG1
0.003
East End Panel.
West End Panel
Principal Strain (in./in.)
0.002
0.001
0.0
R16 R11
R18 R09
R19 R08
R21 R06
R22 R05
R23 R04
R17 R10
R24 R03
R20 R07
East Panel:
West Panel:
Rosette Location
(b) Specimen SG2
Figure 3.34: Principal Strains across Panel Diagonals at Maximum Load
79
0.003
East End Panel.
West End Panel

0.002
0.001
0.0
R16 R11
R18 R09
R19 R08
R21 R06
R22 R05
R23 R04
R17 R10
R24 R03
R20 R07
East Panel:
West Panel:
Rosette Location
(c) Specimen CG1

0.003
East End Panel.
West End Panel

0.002
0.001
0.0
R16 R11
R18 R09
R19 R08
R21 R06
R22 R05
R23 R04
R17 R10
R24 R03
R20 R07
East Panel:
West Panel:
Rosette Location
(d) Specimen CG2
Figure 3.34: Principal Strains across Panel Diagonals at Maximum Load (continued)
80
a
Specimen
a (in.) b (in.)
No.
SG1 10.5 18.125
CG1 10.0 18.0
SG2 9.125 14.75
CG2 9.375 14.75
Figure 3.35: Measured Plastic Hinge Locations
Figure 3.36: Specimen CG2: Concrete Shear Crack Propagation
81
17º
(a) Specimen CG1
14º
(b) Specimen CG2

Figure 3.37: Measured Concrete Shear Crack Angles
82
1.2
1
Vp  Fyw Dt w
(Vn/Vp)
3
Strength
Strength 1.0
0.8 CG1
Shear
SG1
AASHTO Shear
0.6
Shear
Strength w/ TFA
Normalized
0.4
Normalized
AASHTO Shear
Strength w/o TFA
0.2
0.0
0 100 200 300 400 500
Web Panel Depth-Thickness Ratio (D/tw)
(a) Specimens SG1 and CG1
1.2
1
Vp  Fyw Dt w
(Vn/Vp)
3
1.0
Strength
Strength
0.8
Shear
CG2 AASHTO Shear

0.6
Shear
SG2
Strength w/ TFA
Normalized
0.4
Normalized
AASHTO Shear
Strength w/o TFA
0.2
0.0
0 100 200 300 400 500
Web Panel Depth-Thickness Ratio
Web Panel Depth-Thickness Ratio (D/tww))
(D/t
(b) Specimens SG2 and CG2
83
4. NONLINEAR FINITE ELEMENT ANALYSIS OF TEST
SPECIMENS (PHASE 1)
4.1 General
Nonlinear finite element analyses on the shear behavior of plate girders have been
carried out by a number of researchers (Marsh et al. 1988, Baskar et al. 2002, Liang et al.
2005, Real et at. 2007, and Alinia et al. 2009a and 2009b). The analysis results from
these studies have shown reasonable agreement with the test results. In this study, the
commercial finite element analysis software package, ABAQUS 6.11 (2011), was used to
correlate the test results presented in Chapter 3.
4.2 Steel Plate Girder Specimens (SG1 and SG2)
4.2.1 Finite Element Models

Figure 4.1 shows the typical finite element model. Standard 3-D shell elements
(S4R) were used to model both steel plate girder specimens (SG1 and SG2). The S4R
element is a four-node quadrilateral shell element with reduced integration and is suitable
for both thin and thick shells. A simply supported boundary condition was used, and the
lateral bracing was simulated by preventing the lateral movement of the bracing locations
from the experiments. Also, the measured plate thicknesses shown in Table 2.3 were
used in the finite element models.
The load was applied to one node at mid-span of the top flange width, while the
other nodes on the loading surface were constrained to the controlled node by using the
kinematic coupling option for all six degrees of freedom.
4.2.2 Material Stress-Strain Characteristics

Typical steel properties (modulus of elasticity of 29,000 ksi and Poisson’s ratio of
0.3) were used to describe the elastic material characteristics, while the yield stress and
plastic strain extracted from the coupon test results were used to define the plastic
behavior. Since ABAQUS is designed to input true plastic stress and strain, the
84
engineering stress and strain measured from the coupon tests were converted to the true
stress and strain by using Eqs. (4.1) and (4.2).
 T  (1  ) (4.1)
 T  ln(1  ) (4.2)
where  and T are engineering and true stresses, respectively, while  and T are
engineering and true strains, respectively.
4.2.3 Residual Stresses and Geometric Imperfections

Residual stresses were not considered in this study because a number of research
studies revealed that residual stresses had a small or negligible effect on the shear
behavior of I-shaped girders (Horne and Grayson 1983, Stanway et al. 1993).
Initial geometrical imperfections were incorporated in this study. To apply the
initial imperfections, elastic buckling analysis was first conducted. Then, the first
buckling mode shape profile was applied as the initial geometry of the model. It was
noted in Chapter 2 that initial imperfections were measured at several locations prior to
testing; the measured absolute maximum magnitudes of the initial imperfections were
0.16 in. for SG1, 0.35 in. for SG2, and 0.27 in. for CG1 and CG2. Thus, the magnitude
of D/210, which is equal to 0.25 in., was adopted and used in all numerical models.
The chosen value of maximum imperfection was also compared with the
geometric tolerance specified in AWS (2000). AWS Section 5.23.6.2 specifies that the
maximum variation be d0/80 (0.33 in. for the specimens) when the depth-to-thicknes ratio
is greater than 150 (D/tw ≥ 150) and intermediate stiffeners on both sides of the web are
used. Therefore, the selected magnitude of imperfection is within the tolerance set by
AWS and is a reasonable representative magnitude for this study.
4.2.4 Correlation between Test Results and FEA Predictions

Figure 4.2 shows the deformed shape of Specimens SG1 and SG2 from both
testing and finite element analysis. The corresponding load-deflection curves are
compared in Figure 4.3. As shown in Figure 4.2 and Figure 4.3, the finite element
analysis results closely resembled the behavior of the specimens during testing. The
85
ratios of ultimate shear strength from finite element analysis to that from experimental
testing are 1.01 and 1.02 for Specimens SG1 and SG2, respectively (see Table 4.1).
Figure 4.4 compares the plastic hinge locations formed in the top flange and
bearing stiffeners. Plastic hinge locations for SG1 and SG2 from experimental testing
and finite element analysis are summarized in the corresponding table.
4.2.5 Simulated Strains

Figure 4.5(a) marks Events A, B, and C on the plot of the applied load versus mid-
span deflection for discussion of the simulated flexural strains of the flanges and bearing
stiffeners from the finite element analysis of Model SG1. Event A represents the state at
maximum load, and Events B and C are the points when the strength has degraded.
Interestingly, all boundary members (flanges and bearing stiffeners) remained elastic at
maximum load [see Figure 4.5(b) through (d)] for Specimen SG1. Immediately after
maximum load was reached, the top flange near the bearing stiffeners reached the yield
strain, accompanied with significant buckling of the web. Then, the strains at the
expected plastic hinge locations rapidly increased with little loss in shear resistance at
Events B and C. Similar behavior was observed in the bearing stiffeners shown in
Figure 4.5(d). After maximum load was reached, the strains at the expected plastic
hinge locations increased rapidly. Figure 4.5(c) shows the strain profile of the bottom
flange. After maximum load was achieved, the tensile strain at the point of intersection
between the intermediate stiffeners and the bottom flange rapidly increased and reached
the yield strain at Event B. In addition, the compressive strain also increased at a
distance away from the intersection point but still remained in the elastic range.
Figure 4.6 shows the principal tensile strains, which all exceed the strain value
corresponding to the tension-field stress in the web. The principal strain directions are
also shown in the figure. A reference line along the panel diagonal was drawn and two
parallel lines passing through the plastic hinges in the top flange and bearing stiffeners
were also drawn. Interestingly, the principal tensile strains were mostly in the shaded
area. Therefore, the shaded zone can be treated as the tension-field zone. Also, the
figure shows that the directions of principal strains are close to that of the panel diagonal.
86
Similar strain profiles were observed in Model SG2, as shown in Figure 4.7 and
Figure 4.8. However, the strains of the top flange at the intersection with the bearing
stiffeners exceeded the yield strain at maximum load (Event A).
4.3 Steel-Concrete Composite Plate Girder Specimens (CG1 and CG2)
4.3.1 Finite Element Models

Figure 4.9 shows the typical finite element model for Models CG1 and CG2.
Standard 3-D shell elements (S4R) were used for modeling the steel girder. The
concrete slab was modeled with eight-node linear brick elements (C3D8R). The steel
reinforcement and shear studs were modeled with two-node linear beam elements (B31).
These beam elements were bonded with the concrete slab by using the “EMBEDDED”
constraint option in ABAQUS.
Surface-to-surface interaction techniques were also used to properly connect the
surfaces between the two different material surfaces: the steel girder flange and the
concrete slab. The steel girder flange surface was selected as the master surface, while
the concrete slab surface was selected as the slave surface. Then, the hard contact
property was applied for the surface-to-surface interaction to prevent penetration of the
surfaces, while allowing separation between the surfaces. Through these interaction
techniques, slippage between the steel and concrete surfaces was modelled. Boundary
conditions and lateral bracing also simulated the test conditions.
4.3.2 Material Stress-Strain Characteristics

A stress-strain curve of concrete in compression shows an almost linear elastic
behavior up to the initial yield stress, which can be taken as 30~40% of the concrete peak
stress ( f c' ), then the curve becomes nonlinear up to the peak stress. Beyond the peak
stress, strain softening takes place until failure occurs by crushing of the concrete.
Based on these material behaviors, many approaches, including those by Kent and Park
(1971), Popovics (1973), Thorenfeldt et al. (1987), Tsai (1988), and Karthik and Mander
(2011), predict the stress-strain relationship of unconfined (plain) concrete. Among
them, Karthik and Mander’s model (2011) was adopted to simulate concrete behavior in
compression for unconfined concrete. Figure 4.10 shows Karthik and Mander’s
87
compressive stress-strain curve of concrete. As shown in the figure, the constitutive
model consists of three branches; an initial power curve up to the peak stress, followed by
a bilinear relation in the post-peak region. Based on a predicted stress-strain relation of
normal-weight concrete (Collins and Mitchell, 1994), Karthik and Mander expressed the
peak strain as
 c 0  0.0015  f c' (psi) / 10 7 (4.3)
The ultimate stress and strain were defined as f c1  1.74 ksi and  c1  0.0036 ,
respectively. They also specified the failure strain as
 sp  0.012  7  10 7 f c' (psi) (4.4)
With these parameters, stress of the unconfined concrete as a function of strain can be
expressed as follows:
f c  Kf c' 1  1  x 
n
for 0  x 1 (4.5)
 Kf c'  f c1 
f c  Kf  
'
x  1 for 1  x  xu (4.6)
c
 1 
 ux 
 x  xf 
f c  f c1   for xu  x  x f (4.7)
x x 
 u f 
where x is the normalized strain (   c /  c 0 ), xu   c1 /  c 0 , x f   sp /  c 0 , K  1 ,
n  E c  c 0 / f c' , and E c is the Young’s Modulus of concrete.
With Karthik and Mander’s model, concrete behavior in compression can be

inputted into ABAQUS. First, typical concrete material properties (Young’s Modulus,
E c , and Poisson’s ratio of 0.2) were used to describe elastic material characteristics. It
was assumed that first yielding occurs at 40% of the concrete peak stress. Then, the
Concrete Damage Plasticity model available in the material library of ABAQUS was
used to simulate the nonlinear behavior.
Also, tension stiffening behavior was modeled in the analysis. There are three
methods in ABAQUS to model the tension stiffening behaviors: defining cracking strain,
defining cracking displacement, and defining fracture energy. Since the tension
stiffening approach using a stress-strain relationship usually causes an undesirable mesh
sensitivity problem, the fracture energy approach is widely used to mitigate this problem.
88
The fracture energy is the energy required to propagate a tensile crack of unit area.
With the fracture energy ( G f ) approach, the brittle behavior of the concrete is
characterized by a stress-displacement response that is independent of the mesh size

(ABAQUS Manual, 2011) rather than a stress-strain response. Figure 4.11 shows the
relationship between the post-failure stress and fracture energy. As shown in the figure,
it was assumed that the stress decreases linearly after cracking and hence the failure
displacement as a function of fracture energy is
u to  2G f / f to (4.8)
CEB/FIP Model Code (CEB/FIP, 2010) which is the European design code for
concrete structures provides a simple equation to calculate the fracture energy:
G f  73 f cm0.18 (4.9)
where f cm is the mean compressive strength (  f c'  f ) and f  8 MPa. The

fracture energy for ordinary normal weight concrete in Eq. (4.9) has units of N/m. The
mean compressive strengths from the concrete cylinder tests (described in Chapter 2)
conducted for Specimens CG1 and CG2 were used to calculate the fracture energy.
Coupon test results were used to simulate the relationship between the stress and
strain for the steel girders and reinforcement. A bilinear response was assumed for the
shear studs, which have a nominal yield strength of 51 ksi.
4.3.3 Correlation between Test Results and FEA Predictions

The correlation between the predicted and experimental responses for the steel-
concrete composite girder specimens is provided in Figure 4.12. The finite element
analysis was able to predict the maximum load. However, the analysis diverged and
was terminated when it was in the degradation branch. Model CG1, as shown in Figure
4.12(a), only showed a 1-kip load drop after the peak before the analysis was stopped.
This type of convergence issue has been observed and addressed by Baskar et al. (2002)
for finite element analysis of steel-concrete composite girders using ABAQUS. The
convergence issue mainly comes from the concrete, which is a brittle material. Once the
concrete fails in tension, the analysis might have to be aborted. These convergence
issues can be overcome by increasing the fracture energy value or increasing the failure
89
displacement shown in Figure 4.11. However, the increase of fracture energy results in
an increase of the ultimate load. Therefore, it is desirable to avoid modifying the tensile
behavior of concrete. Since the nonlinear finite element analysis successively achieved
the ultimate load in this study, no modification to the tensile behavior of the concrete was
made. The ratios of ultimate shear strengths obtained form finite element analyses to
those of the experiments are 1.03 and 1.06 for Specimens CG1 and CG2, respectively
(see Table 4.1).
4.3.4 Simulated Strains

Figure 4.13 shows the principal tensile strains, which all exceed the strain value
corresponding to the tension-field stress in the web of Model CG1. As with steel girder
specimens like Model SG1, the yield band in the web of the composite girder specimens
can be approximated by the shaded area. The magnitude and directions of the strains
are also similar to those of the steel girders. Figure 4.14(a) shows all principal
compressive strains in the concrete slab. The simulated compressive strains indicate
that the compressive stresses flow from the base of the first stud to the head of the second
stud. Figure 4.14(b) plots the compressive strains that exceed 0.002, which is the peak
strain corresponding to the peak stress ( f c' ). Figure 4.14(c) shows the principal tensile
strains that exceed the cracking strain of concrete. Since most elements between the
shear studs reached the concrete cracking strain, it implies that there is high cracking
potential between the shear studs.
A similar behavior of Model CG2 was observed and principal tensile strains in the
web and principal compressive strains in the concrete slab are shown in Figure 4.15 and
Figure 4.16, respectively. Figure 4.17 shows that the compressive stresses that flow
from the base of the first stud to the head of the second stud are within the flange width
of steel girders. This indicates that the effective width for shear strength evaluation is
much smaller than that for bending.
90
Table 4.1: Shear Strength Comparison between FEM and Experiments
Specimen Panel Shear Strength, V (kips)

VFEM / VTEST
No.
FEM Result (VFEM) Test Result (VTEST)
SG1 318.8 315.3 1.01
SG2 194.5 190.3 1.02
CG1 363.6 352.6 1.03
CG2 225.2 212.1 1.06
91
Loading
Figure 4.1: Specimen SG1: Typical Finite Element Model and Mesh
Experiment
Finite Element Analysis
(a) Specimen SG1
Figure 4.2: Comparison of Deformed Shape
92
Experiment
Finite Element Analysis
(b) Specimen SG2
Figure 4.2: Comparison of Deformed Shape (continued)
93
700 350

600 300
Applied Load (kips)
500 250
400 200
300 150
200 100
Experiment
100 FEM 50
0 0
0.0 0.2 0.4 0.6 0.8 1.0
(a) Specimen SG1
500 250

400 200
Applied Load (kips)
300 150
200 100
100 Experiment 50
FEM
0 0
0.0 0.2 0.4 0.6 0.8 1.0
(b) Specimen SG2
Figure 4.3: Correlation between Test Results and FEM Analyses
94
a
a
b
b
Specimen a (in.) b (in.)

No. Test FEM Test FEM
SG1 10.5 10.1 18.1 18.3
SG2 9.1 8.1 14.8 15.7
Figure 4.4: Comparison of Plastic Hinge Locations
95
700 350
A
B C

600 300
Applied Load (kips)
500 250
400 200
300 150
200 Event A 100

Event B
100 Event C 50
0 0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
(a) Load versus Mid-Span Deflection
0.030
x Event A
0.025 Event B
Event C
Normal Strain (in./in.)
0.020
0.015
0.010
d0
0.005
0.0
-0.005
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
x/d0
x/d0
(b) Normal Strain Profiles on Top Flange

Figure 4.5: SG1: Strain Profiles
96
0.005
0.004
Normal Strain (in./in.) 0.003
0.002
0.001
x
0.0
-0.001
-0.002
-0.003 Event A
Event B
-0.004 Event C
-0.005
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
x/d0
x/d0
(c) Normal Strain Profiles on Bottom Flange
1.2
1.0
0.8
x/D
x/D 0.6
D
0.4 x
Event A
0.2 Event B
Event C
0.0
-0.003 -0.002 -0.001 0.0 0.001 0.002
(d) Normal Strain Profiles along Height of Bearing Stiffener

Figure 4.5: SG1: Strain Profiles (continued)
97
Plastic
Hinge
Panel Diagonal on Panel Diagonal on

Deformed Shape Deformed Shape
(a) at Maximum Load (b) at End of Analysis (Δ=1.2 in.)
Figure 4.6: SG1: Principal Tensile Strains
98
500 250

A B C
400 200
Applied Load (kips)
300 150
200 100
Event A
100 Event B 50
Event C
0 0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
(a) Load versus Displacement
0.030
x Event A
0.025 Event B
Event C
0.020
0.015
0.010
d0
0.005
0.0
-0.005
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
x/d0
x/d0
(b) Normal Strain Profiles on Top Flange

Figure 4.7: SG2: Strain Profiles
99
0.005
0.004
0.002
0.001
x
0.0
-0.001
-0.002
Event A
-0.003 Event B
Event C
-0.004
-0.005
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
x/d0
x/d0
(c) Normal Strain Profiles on Bottom Flange
1.2
1.0
0.8
x/D
x/D 0.6
D
0.4 x
Event A
0.2 Event B
Event C
0.0
-0.003 -0.002 -0.001 0.0 0.001 0.002
(d) Normal Strain Profiles along Height of Bearing Stiffener

Figure 4.7: SG2: Strain Profiles (continued)
100
Plastic
Hinge

Figure 4.8: SG2: Principal Tensile Strains
101
Concrete Slab
(C3D8R Element)
Rebar and Stud

(B31 Element)
Steel Girder
(S4R Element)
Figure 4.9: Specimen CG2: Typical Finite Element Model and Mesh
102
fc
f c : concrete stress
 c : concrete strain
'
f c f c' : peak stress
 c 0 : peak strain
f c1 : ultimate stress
 c1 : ultimate strain
 sp : failure strain
f c1
c
c0  c1  sp
Figure 4.10: Typical Concrete Stress-Strain Curve in Compression
ft
f to : tensile failure stress

f to uto : tensile failure displacement
G f : fracture energy
Gf
ut
uto
Figure 4.11: Relationship between Post-Failure Stress and Fracture Energy
103
800 400

700
600 300
Applied Load (kips)
500
400 200
300
200 100
Experiment
100 FEM
0 0
0.0 0.1 0.2 0.3 0.4 0.5
(a) Specimen CG1
500 250

400 200
Applied Load (kips)
300 150
200 100
100 Experiment 50
FEM
0 0
0.0 0.1 0.2 0.3 0.4 0.5
(b) Specimen CG2
104
Stud 1 Stud 2
Expected Plastic
Hinge Location
Panel Diagonal on
Deformed Shape
Figure 4.13: CG1: Principal Tensile Strains (Δ=0.25 in.)
105
Concrete Slab
Stud 1 Stud 2
Slab Thickness
(a) Principal Compressive Strains (Range from 0 to 0.0058)
Concrete Slab
Stud 1 Stud 2
Slab Thickness
(b) Principal Compressive Strains (Range from 0.002 to 0.0058)

Slab Thickness
(c) Principal Tensile Strains
Figure 4.14: CG1: Principal Strains in Concrete Slab
106
Stud 1 Stud 2
Plastic
Hinge

Figure 4.15: CG2: Principal Tensile Strains
107
Concrete Slab
Stud 1 Stud 2
Slab Thickness
(a) Principal Compressive Strains (Range from 0 to 0.0145)
Concrete Slab
Stud 1 Stud 2
Slab Thickness
(b) Principal Compressive Strains (Range from 0.002 to 0.0145)

Slab Thickness
(c) Principal Tensile Strains
Figure 4.16: CG2: Principal Strains in Concrete Slab
108
Figure 4.17: CG2: Concrete Slab Effective Shear Width
109
5. FACTORS AFFECTING SHEAR RESISTANCE OF END
PANELS: A PARAMETRIC STUDY (PHASE 1)
5.1 General
In Chapter 4, finite element analyses were carried out to study the behavior of the
plate girder end panels, and to correlate the behavior with that observed during testing.
However, the shear resistance of the plate girder end panel may be affected by several
parameters, including flange and bearing stiffener thicknesses, panel depth-to-thickness
ratio, and panel width-to-depth ratio. Finite element analyses were performed in this
chapter in order to study the effects of these parameters. Also, the findings from this
parametric study will be used to develop an analytical model in Chapter 6 for the
prediction of end panel shear strength.
5.2 Steel Plate Girders
5.2.1 Introduction
A total of 99 models (63 variation models from Model SG1 and 36 variation
models from Model SG2) were analyzed. The parameters in this study were flange
thickness ( t f ), bearing stiffener thickness ( t b ), width-to-depth ratio ( d 0 / D ), and depth-
to-thickness ratio ( D / t w ). The main objective of this study was to identify variables
that affect the panel shear strength and plastic hinge locations in the top flange and
bearing stiffeners. Material properties from the test specimens were used in the finite
element models. Also, the nominal plate thicknesses of the test specimens were used for
all finite element models. Figure 5.1 shows a typical finite element model used in this
parametric study. To evaluate the effect of panel width-to-depth ratio, the girder length
had to be increased, and the increased length may cause an unexpected failure mode like
flexural failure. To avoid this failure mode, only two end panels were used in the study.
To confirm similar behaviors between the models with 4 panels and 2 panels, the
behavior of the steel plate girder Model SG1 with 4 panels (see Figure 4.1) and 2 panels
are compared in Figure 5.2. As expected, the initial stiffness of the model with 2 panels
110
is stiffer than the one with 4 panels. However, the maximum shear strengths of two
cases are almost the same. The maximum loads achieved from the finite element
analyses are 637.5 and 638.7 kips for the model with 4 and 2 panels, respectively [see
Figure 5.2(a)]. The plastic hinge locations are also almost the same between the two
models, as shown in Figure 5.2(b). A minor discrepancy between the two might be due
to having slightly different mesh sizes.
5.2.2 Effect of Flange and Bearing Stiffener Thicknesses

Figure 5.3(a-c) and Figure 5.4(a-c) show the effect of the flange and bearing
stiffener thicknesses. The end panel shear capacity increases as the flanges or bearing
stiffeners increase in thickness. Additionally the plastic hinges at the top flange and
bearing stiffeners tend to form further away from the intersection of the top flange and
the bearing stiffeners as thicknesses are increased. This may be due to the fact that the
increased thicknesses of the boundary members (flanges and bearing stiffeners) result in
higher flexural strengths of these members to resist higher tension-field forces before
they fail. According to Basler’s equation [see Eq. (1.10)] to calculate the tension-field
stress, the magnitude of the web tension-field stress remains the same if the same
material and geometry of the web panels are used. Therefore, to increase the tension-
field force, the width of the tension-band should be increased to fail the boundary
members. As a result, the plastic hinges will form further away from the intersection of
the top flange and bearing stiffeners. Table 5.1(a-c) and Table 5.2(a-c) summarize the
plastic hinge locations identified from the analyses.
5.2.3 Effect of Panel Depth-to-Thickness and Width-to-Depth Ratios

Figure 5.3(d) plots the variation of the end panel shear strength with respect to the
width-to-depth ratio and depth-to-thickness ratio. The panel depth, flange size, and
bearing stiffener size corresponded to those in Model SG1, but the panel width and web
thickness were varied. Since AASHTO Specifications (2014) limits the end panel
width-to-depth ratio to 1.5, panel width-to-depth ratios of 0.5, 1.0, and 1.5 were consiered.
The panel depth-to-thickness ratio also varied from 180 to 320.
Table 5.3 and Figure 5.5 show that when the panel width-to-depth ratio is
constant, the depth-to-thickness ratio has a minor effect on the plastic hinge locations.
111
For example, when the depth-to-thickness ratio increased from 180 to 320 with a constant
width-to-depth ratio of 0.5, the plastic hinge locations shifted only by 1 in. Also, for the
same increase in depth-to-thickness ratio with a constant width-to-depth ratio of 1.5, the
plastic hinges formed at the same locations, regardless of the depth-to-thickness ratio.
This observation deserves special attention. In Basler’s equation [see Eq. (1.10)] to
calculate the tension-field stress, the tension-field stress is a function of the ratio of panel
buckling stress to the yield stress. Since both web yield stress, Fyw , and shear yield
stress,  y , are constant, the only variable to determine the tension-field stress is the
buckling stress,  cr . When the panel depth-to-thickness ratio is increased, the buckling
stress of the web panel is decreased, and the tension-field stress increases. With a
higher tension-field stress, boundary members need to resist a higher out-of-plane force
(vertical force for flanges and horizontal force for bearing stiffeners) induced by the
tension-field stress. Therefore, per Basler’s theory, the boundary members should fail
by forming plastic hinges closer to the intersection of the top flange and bearing stiffeners.
However, the results from the parametric study do not support this. Since the
plastic hinge locations remain about the same, it suggests that there must be another
source which increases the out-of-plane flexural strength of the boundary members.
That is, when the tension-field stress increases, flexural strengths of the boundary
members must also increase to be consistent with observations made from the parametric
study. One way to increase the flexural strength of the boundary members is to allocate
a small portion of the web to act together with flanges and bearing stiffeners. This issue
is further discussed in Section 6.7.
5.2.4 Effect of Web End Extension, e

Figure 5.4(d) shows the effect of extending the web beyond the support. The
results, including the ultimate shear strength and plastic hinge locations, are summarized
in Table 5.4. When the web extension, e, increases, the distance to the plastic hinge
location (b) increases, while the plastic hinge remains at the same location in the flange.
Because the web extension together with the bearing stiffeners can resist a higher
horizontal component of the tension-field force, the shear strength also increases. In
this parametric study, at a width-to-depth ratio of 0.5, the shear strength was about 7%
112
greater when the web extension is 8 in. (e = 8 in.) than when there was no web extension.
When the width-to-depth ratios were 1.0 and 1.5, the shear strengths increased by about
8% and 6%, respectively.
5.3 Steel-Concrete Composite Plate Girders
5.3.1 Introduction
To study the contribution of the concrete slab on the ultimate shear strength in the
end panels of steel-concrete composite girders, a parametric study was performed.
Model CG2 was adopted as a base model for finite element analysis. Then two
parameters, concrete slab width and shear stud location, were varied to investigate thier
effects on the shear strength.
5.3.2 Effect of Concrete Slab Width

Both Models CG1 and CG2 had a 6-ft wide concrete slab. In the nonlinear finite
element analysis in Chapter 4, it was observed that the effective width of the concrete
slab for shear resistance was close to the flange width of the steel girder (see Figure 4.17).
This observation contradicted a previous research study (Darehshouri et al. 2012) which
concluded that the entire slab width is effective. To better understand this issue, the
concrete slab width was varied from 2 to 8 ft. The ultimate shear strengths are
summarized in Table 5.5. Figure 5.6 indicates that the concrete slab width has a very
minor effect on the ultimate shear strength of steel-concrete composite girders. This
confirms that, for calculating the shear strength of end panels, the effective concrete slab
width should be close to the steel girder flange width.
5.3.3 Effect of Shear Stud Location

Figure 4.14 and Figure 4.16 show that the compressive stresses in the concrete
flow between the shear studs under shear loading. The most critical section is the
diagonal direction between stud 1 and stud 2. To confirm the load-carrying mechanism
in the concrete slab, stud 2 was placed at varying distances away from stud 1. The total
number of shear studs was maintained to keep the same degree of composite action.
Figure 5.7 plots the effect of the distance between end studs on the shear strength
of end panels. The ultimate shear strength increases slightly as the distance between
113
two consecutive studs decreases. Figure 5.8 confirms the load-transfer mechanism
between two end studs, regardless of the distance between them. This observation will
be used in Section 6.9 for the development of a model that considers the contribution of
the concrete slab to the shear strength of end panels.
114
Table 5.1: Summary of Model SG1 Parametric Study (D/tw = 210)
(a) d0/D = 0.5
tf = 0.5 in. (2tw) tf = 0.75 in. (3tw) tf = 1.0 in. (4tw) tf = 1.5 in. (6tw)
Ultimate Ultimate Ultimate Ultimate

a b a b a b a b
tb Strength Strength Strength Strength
(in.) (in.) (in.) (in.) (in.) (in.) (in.) (in.)
(kips) (kips) (kips) (kips)
0.5 in.
312.2 8.2 15.6 316.9 10.3 17.6 323.0 12.2 18.6 336.5 13.1 19.6
(2tw)
0.75 in.
315.7 8.2 15.6 321.8 10.3 17.6 328.2 13.1 19.6 344.3 13.1 20.6
(3tw)
1.0 in.
316.8 8.2 15.6 324.3* 10.3* 17.6* 331.9 14.1 20.6 362.8 16.2 21.6
(4tw)
1.5 in.
318.2 9.2 17.6 328.1 10.3 19.6 347.0 14.1 21.6 410.1 20.2 26.6
(6tw)
*
Model SG1
115
Table 5.1: Summary of Model SG1 Parametric Study (D/tw = 210, continued)
(b) d0/D = 1.0

a b a b a b a b
0.5 in.
226.3 10.1 9 231.9 12.1 11.8 241.1 14.1 11.8 251.0 16.2 11.8
(2tw)
0.75 in.
228.3 10.1 10 234.7 12.1 12.7 247.1 14.1 14.6 259.2 16.2 14.6
(3tw)
1.0 in.
228.2 10.1 12.7 238.2 12.1 13.6 252.2 14.1 15.6 275.0 18.2 16.6
(4tw)
1.5 in.
230.6 10.1 14.6 243.7 13.1 15.6 259.5 14.1 16.6 305.4 21.2 19.6
(6tw)
116
(c) d0/D = 1.5

a b a b a b a b
0.5 in.
197.9 12.1 10.0 203.3 15.1 11.1 211.6 16.2 11.1 220.4 18.2 11.1
(2tw)
0.75 in.
199.1 12.1 11.1 205.3 15.1 12.1 215.7 16.2 13.1 226.0 18.2 13.1
(3tw)
1.0 in.
199.4 12.1 12.1 207.8 15.1 13.1 219.1 16.2 15.1 232.7 21.2 15.1
(4tw)
1.5 in.
200.9 12.1 14.1 211.6 15.1 14.1 223.7 18.2 16.1 249.1 23.2 17.1
(6tw)
117
Table 5.2: Summary of Model SG2 Parametric Study (D/tw = 280)
(a) d0/D = 0.5
tf = 0.5625 in. (3tw) tf = 0.75 in. (4tw) tf = 1.125 in. (6tw)
Ultimate Ultimate Ultimate

a b a b a b
tb Strength Strength Strength
(in.) (in.) (in.) (in.) (in.) (in.)
(kips) (kips) (kips)
0.5625 in.
195.1 8.2 15.0 202.3 9.2 16.0 216.3 11.3 16.0
(3tw)
0.75 in.
197.2 8.2 16.0 204.9 10.3 17.0 229.6 13.3 18.0
(4tw)
1.125 in.
199.6 9.2 17.0 210.4 10.3 18.0 253.2 15.3 21.0
(6tw)
*
Model SG2
(b) d0/D = 1.0

a b a b a b
tb Strength Strength Strength
(in.) (in.) (in.) (in.) (in.) (in.)
0.5625
in. 141.9 10.1 10.0 150.4. 12.1 11.0 164.6 15.1 12.0
(3tw)
0.75 in.
143.9 10.1 11.0 153.1 12.1 12.0 173.6 16.2 14.0
(4tw)
1.125 in.
147.2 10.1 13.0 157.4 13.1 13.0 187.5 17.2 16.0
(6tw)
118
(c) d0/D = 1.5

tb a b a b a b
Strength Strength Strength
(in.) (in.) (in.) (in.) (in.) (in.)
0.5625 in.
123.1 12.9 10.0 130.2 14.9 11.0 138.0 18.8 11.0
(3tw)
0.75 in.
124.4 12.9 11.0 132.0 14.9 12.0 143.3 18.8 13.0
(4tw)
1.125 in.
126.5 12.9 12.0 134.5 14.9 13.0 151.5 19.8 15.0
(6tw)
Table 5.3: Summary of Model SG1 Parametric Study

(tf = 0.75 in. and tb = 1.0 in.)
d0/D = 0.5 d0/D = 1.0 d0/D = 1.5

a b a b a b
D/tw Strength Strength Strength
(in.) (in.) (in.) (in.) (in.) (in.)
180 407.0 10.3 17.6 302.7 12.1 13.6 266.0 15.1 13.1
210 324.3 10.3 17.6 238.2 12.1 13.6 207.8 15.1 13.1
240 267.1 10.3 17.6 199.8 12.1 13.6 172.9 15.1 13.1
280 219.7 10.3 18.6 164.9 12.1 13.6 141.3 15.1 13.1
320 192.6 11.2 18.6 144.2 13.1 13.6 119.5 15.1 13.1
119
Table 5.4: Summary of Model SG2 Parametric Study
(tf = 0.5625 in. and tb = 1.0 in.)
d0/D = 0.5 d0/D = 1.0 d0/D = 1.5

a b a b a b
e Strength Strength Strength
(in.) (in.) (in.) (in.) (in.) (in.)
0 in. 196.7 8.2 14.0 143.1 10.1 11.0 124.4 12.9 11.0
4 in. 201.4* 8.2* 15.0* 149.2 10.1 12.0 128.4 12.9 12.0
8 in. 210.2 8.2 16.0 154.2 10.1 12.0 132.3 12.9 12.0
Table 5.5: Effect of Concrete Slab Width
Concrete Width Shear Strength

(ft) (kips)
8 227.1
6 225.2
5 223.4
4 221.6
3 219.1
2 218.5
120
Loading
Figure 5.1: Typical Finite Element Model (2 Panels)
700 350

600 300
Applied Load (kips)
500 250
Max. Load
400 200
300 150
200 100
4 Panels, Max. Load = 637.5 kips
2 Panels, Max. Load = 638.7 kips
100 50
0 0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4
(a) Global Response
Figure 5.2: Model SG1: Comparison of 4 Panels and 2 Panels
121
a a
b b
2 Panels 4 Panels
a = 10.25” a = 10.1”
b = 17.625” b = 18.25”
(b) Plastic Hinge Locations
Figure 5.2: Model SG1: Comparison of 4 Panels and 2 Panels (continued)
122
450 450
1.5”

400 400
1.0”
350 0.75” 350
0.5”
1.5”
300 300
1.0”
0.75”
250 250 0.5”
tb
200 200
tb
150 150
100 100
0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0
Flange Thickness (in.) Flange Thickness (in.)
(a) d0/D=0.5
(a) Effectand
of tf D/tw=210
and tb (b) d0/D=1.0
(b) Effectand
of tD/tw=210
f and tb
(d0/D = 0.5 and D/tw = 210) (d0/D = 1.0 and D/tw = 210)
450 500
400
400 D/tw
350 tb
300
300 180
1.5”
250 210
1.0” 200
0.75” 240
0.5”
200 280
320
100
150
100 0
0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0
Flange Thickness (in.) Width-Depth
Width-Depth Ratiod0/D
Ratio,
(c) d0/D=1.5
(c) Effectand
of tfD/tw=210
and tb (d)(d)
tf=0.75
Effectin.
of and
d0/Dtb=1.0
and D/tin.w
(d0/D = 1.5 and D/tw = 210) (tf = 0.75 in. and tb = 1.0 in.)
Figure 5.3: Model SG1: Summary of Parametric Study
123
300 300

1.125”
250 250
0.75” tb
0.563”
200 200
1.125”
0.75”
tb 0.563”
150 150
100 100
0.0 0.5 1.0 1.5 0.0 0.5 1.0 1.5
Flange Thickness (in.) Flange Thickness (in.)
(a) Effectand
(a) d0/D=0.5 of tf D/tw=280
and tb (b) Effectand
(b) d0/D=0.5 of tfD/tw=280
and tb
(d0/D = 0.5 and D/tw = 210) (d0/D = 1.0 and D/tw = 210)
300 300 D/tw

250
250
0.5
200
tb 1.0
200 150
1.5
e
100
1.125”
150 0.75”
0.563” 50
100 0
0.0 0.5 1.0 1.5 0 2 4 6 8
Flange Thickness (in.) Extended
Web End Portion of Web
Extension (in.)
(in.)
(c) Effectand
(c) d0/D=0.5 of tfD/tw=280
and tb (d) Effect of Web End Extension
(d) tf=0.625 and tb=1 (e)
(d0/D = 1.5 and D/tw = 210) (tf = 0.56 in. and tb = 1.0 in.)
Figure 5.4: Model SG2: Summary of Parametric Study
124
400
350 d0/D 1.5 1.0 0.5

Depth-ThicknessRatio, D/tw
Ratio,D/tw a
300
Depth-Thickness
250
D
200
150 d0
100
0.0 0.1 0.2 0.3 0.4 0.5
Plastic Hinge
Normalized PlasticLocation (a/d0) a/d
Hinge Location, 0
(a) Normalized Plastic Hinge Location, a/d0
400
350
D/tw
d0/D 1.5 1.0 0.5

w
Ratio, D/t
a
Depth-Thickness Ratio,
300
Depth-Thickness
b
250
D
200
150
d0
100
0.0 0.1 0.2 0.3 0.4 0.5
Plastic Hinge
Normalized PlasticLocation (b/D) b/D
Hinge Location,
(b) Normalized Plastic Hinge Location, b/D
Figure 5.5: Effect of d0/D and D/tw on Plastic Hinge Location
125
300
End Panel Shear Strength (kips) 250

218.5 kips 227.1 kips
200
150 Concrete Slab Width
100
50
0
0 2 4 6 8 10
Concrete Slab Width (ft)
Figure 5.6: Model CG2: Effect of Concrete Slab Width
300
250 230.3 kips 220.1 kips

225.2 kips
200
150 Stud End Distance
100
50
0
0 5 10 15
Stud End Distance (in.)
Figure 5.7: Model CG2: Effect of Stud End Distance
126
(a) Stud End Distance = 4.25 in.
(b) Stud End Distance = 8.5 in.
(c) Stud End Distance = 12.75 in.
Figure 5.8: Specimen CG2: Direction of Compressive Principal Strain
127
6. SHEAR RESISTANCE OF END PANELS: DERIVATION OF
PREDICTIVE EQUATIONS
6.1 General
Current US design practices (AASHTO and AISC Specifications) ignore the
tension-field action in the plate girder end panels due to the lack of an effective anchoring
mechanism for the horizontal component of the tension-field force. However, Basler
(1961a) indicated that the bearing stiffeners and the web portion extending beyond the
support would provide some amount of bending rigidity to anchor the horizontal
component of tension-field force. This concept is reflected in Eurocode 3 (2006) and
post-buckling strength due to this partial tension-field action is accounted for in the
design strength of end panels.
Test results also showed that ignoring tension-field action in the end panel is too
conservative. In this chapter, an analytical study was conducted to provide a simple
model to calculate the ultimate shear strength of the end panel. In the end panel region,
it is assumed that both the top and bottom flanges have the same dimensions and steel
grade.
The shear strength of a steel plate girder is provided by the following two
contributions: (a) the beam action shear Vcr , and (b) the post-buckling shear due to the
tension-field action shear V ps .
Vs  Vcr  V ps (6.1)
The beam action shear strength is the web panel buckling strength:
Vcr   cr Aw (6.2)
where the elastic shear buckling stress, cr , is
 2 E  tw 2 
 cr  k     (6.3)
12(1   )  D  
2
Aw  Dt w (6.4)
The coefficient k in Eq. (6.3) is the elastic shear buckling coefficient [see Eq. (1.18)].
128
6.2 Failure Modes
Unlike interior panels, end panels do not have an adjacent panel beyond the
support, which results in a deficiency in anchoring the horizontal component of tension-
field force. Therefore, the failure modes are highly dependent on the flexural strength
of the pair of bearing stiffeners to resist this horizontal force. When tension-field action
occurs in the slender web, the post-buckling strength is dependent on the width, s, of the
tension-field zone. Therefore, it is important that this width be determined accurately.
Figure 6.1 shows three possible failure modes of an end panel. The first mode
involves failure of the top and bottom flanges. This occurs when very rigid bearing
stiffeners are used and do not fail. A total of four plastic hinges in this failure mode are
formed: two at the top flange and two at the bottom flange (see Failure Mode 1 in Figure
6.1). This failure mode is the same as for interior panels, because the vertical
boundaries are rigid enough to serve as an anchor like the adjacent web panels of the
interior panel. As a result, the horizontal boundaries (top and bottom flanges) have to
fail. With this type of failure mode, the width of tension band is large to achieve the full
tension-field action as in interior panels.
The second failure mode is the failure of the flanges and bearing stiffeners. This
is most typical of real-life plate girder end panels where non-rigid bearing stiffeners are
used. This failure mode involves four plastic hinges, with one plastic hinge forming in
the bearing stiffeners instead of in the bottom flange. Due to the formation of the plastic
hinge in the bearing stiffeners, the tension band is narrower than the first failure mode
and, hence, only a partial tension-field action can be developed. Referring to Failure
Mode 2 in Figure 6.1(c), a  in the bottom flange has to be less than the dimension a in
Figure 6.1(b) so that a plastic hinge cannot occur in the bottom flange.
The third failure mode in Figure 6.1(d) is a combination of the first and second
failure modes and requires five plastic hinges. However, the tension band width may be
close to that of the first failure mode due to the formation of two plastic hinges at the
bottom flange. Assuming that a plastic hinge in the bottom flange forms at the same
location as in the first failure mode, this failure mode may provide almost the same post-
buckling shear capacity as the first failure mode (i.e., interior panel failure mode) due to
the same tension band width.
129
Since the second failure mode is representative of typical steel end panels, it is the
focus of this study.
6.3 Tension Band Width, s
Figure 6.2 shows an assumed tension band width overlapped on the failure mode
of the end panel of Specimen SG2. The assumed tension band is defined by first
drawing the panel diagonal line (AF) with an inclination angle of  . Then parallel lines
(BE and GJ) are made, where A, B, F, and J correspond to the four plastic hinge locations
in Figure 6.1(c).
Assume that a uniform tension-field stress  t develops across the tension band
width. Given the location of plastic hinges defined by a and b, the tension band width, s,
can be determined by using the panel geometry as:
s  a sin   b cos  (6.5)
where
D
sin   (6.6a)
d  D2
2
0
d0
cos   (6.6b)
d 02  D 2
Considering Eqs. (6.6a) and (6.6b), sin  and cos  can be expressed in one equation.
D
sin   cos  (6.7)
d0
6.4 Post-Buckling Strength: A Plastic Mechanism Approach
Porter et al. (1975), Huslid and Rockey (1979), and Darehshouri et al. (2012) used
plastic mechanism analysis to develop their tension-field theories. Since the plastic
analysis provides an upper bound solution, it requires finding a solution which minimizes
the post-buckling shear strength ( V ps ) of the end panel. Figure 6.3 shows the failure
mechanism of an end panel. The shaded area in the web represents the yielding zone
due to the tension-field action. By excluding the web plate, the tension-field stresses in
the shaded area were considered as external loads that act on the boundary members as
130
shown in Figure 6.4(a). Applying the virtual work principle, a virtual vertical
displacement  is applied at support H, which results in rotations at the four plastic
hinge locations shown in Figure 6.3.
Internal Work
The internal virtual work is
W I  M pA ( B   A )  M pB ( B )  M pJ ( F   A )  M pF ( F ) (6.8)
where M pB and M pF equal the flange plastic moment (= M pf ), M pJ is the plastic
moment of a pair of bearing stiffeners (= M pb ), and M pA is the minimum of M pf and
M pb (= M pm ). Also, the rotations at the plastic hinges can be expressed as a function
of the virtual displacement  .


B 
a

F  (6.9)
d0
( D  b ) F ( D  b )
A  
b bd 0
Substituting Eq. (6.9) into Eq. (6.8), the internal work is rewritten as:
 ( D  b ) 
W I  M pf  M pm   M pb  M pm   M pf  M pb  (6.10)
a bd 0 d0
External Work
Since the web plate is excluded in the free-body in Figure 6.4(a), the tension-field
stresses acting on the boundary members are considered as external loads. The diagonal
tension stresses are decomposed into horizontal and vertical components as shown in
Figure 6.4(b). Vertical shear V ps with an applied vertical displacement  creates
external virtual work.

(1) Member AB
131
Considering a rotation  B at B, member AB with a vertical distributed load
(  t t w sin 2  in Figure 6.5) experiences an average upward movement of a B /2, thus
producing a negative work of  a t t w sin 2 a B / 2 . Since there is no horizontal

movement on member AB, the horizontal component of the tension-field stresses
produces no work.
 a 
WE ( AB )  a t t w sin 2  B  (6.11)
 2 
(2) Member AJ
Similarly, member AJ experiences a vertical movement of  and an average
horizontal movement of b A / 2 . The horizontal and vertical movements together with
horizontal and vertical distributed loads generate the following external work:
 b 
W E ( AJ )  b t t w cos 2  A   b t t w sin  cos   (6.12)
 2 
Member GF
The length of member GF can be calculated by using the panel geometry.
( D  b) ( D  b) bd 0
LGF  d 0   d0   (6.13)
tan  (D / d 0 ) D
The member GF undergoes an average vertical upward movement of (bd 0 / D) F / 2 but
no horizontal movement. This vertical movement results in the following external
work:
 (bd 0 / D) F   t t w b cos 
2 2
bd 0
W E (GF )   t t w sin 2    F (6.14)
D  2  2
Since member EF does not have any virtual movement, the load acting on the member
does not produce work (see Figure 6.3). Therefore, the total external work is
W E  V ps ()  W E ( AB )  WE ( AJ )  WE (GF )
 a t sin 2    b t t w cos 2 ( D  b) 
 V ps ()   t w      b t t w sin  cos  
2   2  (6.15)
   d 0 
 t b cos 
2 2
 t w 
2d 0
Equate Eq. (6.10) with Eq. (6.15) to solve the post-buckling shear strength:
132
M  M pm  ( D  b)M pb  M pm  M  M pb 
V ps   
pf pf
a bd 0 d0
 a t sin 2    b t t w cos 2 ( D  b) 
  t w      b t t w sin  cos  (6.16)
 2   2d 0 
 t b cos 
2 2
 t w
2d 0
6.5 Plastic Hinge Locations in Boundary Members
It is seen from Eq. (6.16) that the post-buckling strength is a function of plastic
hinge locations defined by a and b. These two variables can be computed by using the
static equilibrium of the free-bodies of flange member AB and bearing stiffener member
AJ shown in Figure 6.5. Due to the formation of the plastic hinges, the member end
moments are known. However, the axial and shear forces are still indeterminate.
Therefore, two shear ratio parameters (  for flange shear ratio and  for bearing
stiffener shear ratio) are introduced in the free-bodies of members AB and AJ. Internal
axial forces of these members then can be established by static equilibrium.
By considering the moment equilibrium at A of each member from Figure 6.5(b),
the plastic hinge distances can be determined as follows:
2( M pf  M pm )
a (6.17a)
 t t w (1  2 ) sin 2 
2( M pb  M pm )
b (6.17b)
 t t w (1  2 ) cos 2 
If  or  is greater than 0.5, a or b will have an imaginary number. Therefore, both

 and  must be less than 0.5.
Substituting Eq. (6.17) into Eq. (6.16), the post-buckling strength V ps is
expressed with two independent variables,  and  . Based on the upper bound
theory in plastic analysis (Neal, 1985),  and  should be chosen to minimize the
post-buckling strength V ps :
133
V ps
0 (6.18a)

V ps
0 (6.18b)

Theoretically, the two variables of  and  can be determined by solving the
simultaneous equations in Eq. (6.18). However, it is very complex to derive the closed-
form solution of  and  . Therefore, it was decided to establish the values of  and
 numerically.
Both Models SG1 and SG2 were used for this case study. By varying the 
and  values in Eq. (6.17a) and (6.17b), the values of a and b can be determined.
Substituting these two values into Eq. (6.16) gives the post-buckling strength. Figure
6.6 shows the effect of the shear ratio parameters. It is seen that the post-buckling
strength of the end panel becomes minimal when both  and  approach zero.
Setting both  and  to zero in Eq. (6.17), the plastic hinge location
parameters are
2( M pf  M pm )
a (6.19a)
 t t w sin 2 
2( M pb  M pm )
b (6.19b)
 t t w cos 2 
Substituting Eq. (6.19) into Eq. (6.16), the post buckling strength V ps is reduced to:
M pf  M pb t t wb 2
V ps   t t w sin (a sin   b cos )   cos 2  (6.20)
d0 2d 0
The free-body diagrams of the end panel boundary members due to the tension-field
action are summarized in Figure 6.7.
Summing up the beam action shear buckling strength and post-buckling strength
( Vcr + V ps ) gives the ultimate shear strength of the end panel.
M pf  M pb t t wb 2
Vs   t t w sin (a sin   b cos )   cos 2    cr Dt w (6.21)
d0 2d 0
134
6.6 Simplification
Depending on the relative plastic moment values of the girder flange and bearing
stiffener pair, Eq. (6.21) can be further simplified as follows.
Case 1: M pf  M pb (weak flange)
In this case, M pm is equal to M pf . Then, M pf  M pb  in Eq. (6.19b) becomes
b2
M pf  M pb   t t w cos 2  (6.22)
2
Dividing by d 0 on both sides, Eq. (6.22) can be rewritten as
M  M pb  b2
  t t w cos 2  2
pf
(6.23)
do 2d o
It is readily shown that the second and third terms on right hand side of Eq. (6.21) cancel
out, and the ultimate shear strength is reduced to
Vs   t t w sin (a sin   b cos )   cr Dt w (6.24)
Case 2: M pf  M pb (weak bearing stiffeners)
In this case, M pm is equal to M pb . Then, by using Eq. (6.19a), M pf  M pb  is
a2
M pf  M pb   t t w sin 2  (6.25)
2
Substituting Eq. (6.25) into Eq. (6.21), the ultimate shear strength becomes
 t t w (a 2 sin 2   b 2 cos 2 )
Vs   t t w sin (a sin   b cos )    cr Dt w (6.26)
2d 0
Eq. (6.26) consists of three terms; the first term is the post-buckling strength due to
tension-field action in the web, the second is the shear resistance from the bottom flange
(see Figure 6.7), and the last is the web shear buckling strength.
Generally the flange shear strength is very small compared to the web shear
strength. To evaluate the contribution of the flange to the ultimate shear strength, a total
of 99 cases including Models SG1 and SG2 and the results from a parametric study
which are summarized in Chapter 5 were used. Figure 6.8 shows that the bottom flange
contribution increases with plastic moment of the flange. But the contribution is only a
135
few percent and can be ignored for practical reasons. By ignoring the contribution of
the flange, Eq. (6.26) reduces to (6.24).
Eq. (6.24) can be further simplified. Substituting Eq. (6.6) into Eq. (6.24), the
ultimate shear strength is expressed as:
 a  b( d 0 / D ) 
Vs   t t w     cr Dt w
2  (6.27)
 1  ( d 0 / D ) 
Note that Basler (1961a) suggested a simplified tension-field stress ( t ).
  
 t  Fyw 1  cr   Fyw 1  C  (6.28)
  
 yw 
where C   cr /  yw . Substituting Eq. (6.28) into Eq. (6.19), a and b become:
1 2( M pf  M pm ) d 02  D 2 2( M pf  M pm )
a  (6.29a)
sin  t t w D t t w
1 2( M pb  M pm ) d 02  D 2 2( M pb  M pm )
b  (6.29b)
cos  t t w d0 t t w
Substituting Eq. (6.29) into the numerator in Eq. (6.27),
d  d 02  D 2 2( M pf  M pm ) d 02  D 2 2( M pb  M pm )
a  b 0   
D D t tw D t tw
d 02  D 2  2( M pf  M pm )  2( M pb  M pm ) 
  
D  t tw 
 
1  (d 0 / D) 2 (6.30)
 2( M pf  M pm )  2( M pb  M pm )
t w Fyw (1  C )
Also, the beam-action shear buckling strength can be rewritten as:

 
Vcr   cr Dt w   cr  yw Dt w  0.58Fyw Dt w C (6.31)
 
 yw 
Substituting Eqs. (6.28), (6.30), and (6.31) into Eq. (6.27), the ultimate shear strength is
136
Vs 
Fyw (1  C )t w
1  (d 0 / D) 2
1  (d 0 / D) 2
t w Fyw (1  C )
 2( M pf  M pm )  2( M pb  M pm )  Vcr 


Fyw (1  C )t w D 2( M pf  M pm )  2( M pb  M pm )   0.58F Dt w C
yw
1  (d 0 / D) 2 D t w Fyw (1  C )
 (6.32)
0.87(1  C )  2( M pf  M pm )  2( M pb  M pm ) 
 0.58 Fyw Dt w C 

 1  (d 0 / D) 2  0.58(0.87) D t w Fyw (1  C ) 


 0.58 Fyw Dt w C 

0.87(1  C )  2.8 M pf  M pm  M pb  M pm 

 1  (d 0 / D) 2  D t w Fyw (1  C ) 

The above expression can be rewritten in the following form:
 0.87(1  C ) 
Vs  V p C    (6.33)
 1  (d 0 / D) 2 
where
V p  0.58 Fyw Dt w (6.34)
 cr
C (6.35)
 yw


2.8 M pf  M pm  M pb  M pm   1.0 (6.36)
D t w Fyw (1  C )
Note that Eq. (6.33) has the same format as that used in 2014 AASHTO
Specifications for calculating the shear resistance of interior panels, except that it
contains one additional parameter  to consider the partial tension-field action. The
parameter (  ) accounts for the flexural strength of the boundary members (top flange
and bearing stiffeners). When tension-field action exists, the end panel shear strength
will be generally less than that of an interior panel with the same dimensions. Therefore,
 should be no larger than 1.0. Eq. (6.33) is written in a format which is applicable to
both interior and end panels. At one extreme when both the flange and bearing
stiffeners offer little flexural strength,  approaches zero and only the beam action
contributes to the shear strength. As the other extreme when both the flange and bearing
stiffeners offer large flexural strengths,  approaches 1.0, which is equivalent to the
137
development of full tension-field action as in an interior panel. The value of  will be
less than 1.0 for end panels in typical steel plate girder bridges.
6.7 Effective Depth of End Panel Boundary Members
The post-buckling strength depends on the plastic hinge location parameters, a

and b. These two parameters are a function of the plastic moment capacities of the
flanges ( M pf ) and bearing stiffener pair ( M pb ) as well as the tension-field stress (  t ) as
shown in Eq. (6.29). As observed in Section 5.2.3, varying the depth-to-thickness ratio
(D/tw) while keeping the panel width-to-depth ratio (d0/D) constant had virtually no effect
on the plastic hinge locations in the top flange and bearing stiffener pair. Since the
tension-field stress is a function of D/tw [see Eq. (6.28)], the above observation implies
that some portion of the web would act together with the boundary members (see Section
5.2.3).
Past researchers investigating the shear strength of interior panels (Chern and
Ostapenko 1969, Rockey and Skaloud 1972, and Porter et al. 1975) also assumed that the
flanges act with an effective web depth. Figure 6.9 shows the same concept applied to
the end panel. When the web experiences inelastic buckling in a panel of low depth-to-
thickness ratio, only a small portion of the web area can act with the flange, and this
portion can be ignored conservatively. However, when the web is very slender (high
D/tw ratio) and experiences elastic buckling, a larger portion of the web area will be
effective and should be considered to determine the moment capacities of the flange
( M pf ) and bearing stiffeners ( M pb ). Eqs. (6.37) and (6.38) are the effective depths of
the web suggested by past researchers for interior panels.

  
d eCO  12.5t w  0.8  cr  by Chern and Ostapenko (1969) (6.37)
  yw 

   by Rockey and Skaloud (1972)
d eRS  30t w 1  2 cr  (6.38)
  yw  and Porter et al. (1975)

Both equations assume that the effective web depth is proportional to the shear buckling
stress normalized by the shear yield stress. Since the terms in the parentheses must
remain positive, Eqs. (6.37) and (6.38) imply that inelastic buckling occurs when  cr
138
exceeds 0.8  yw and 0.5  yw , respectively (see Figure 6.10). Note that Basler (1961a)
and US design codes (both 2014 AASHTO Specifications and 2010 AISC Specifications)
specify that inelastic buckling occurs when the plate shear buckling stress  cr exceeds
0.8  yw . Therefore, when elastic buckling occurs, a T-section of the flange and a T- or
cruciform-section (when the web extends beyond the support by a distance e) of the
bearing stiffener pair should be included to compute the plastic section modulus, Z x
(see Figure 6.9 and Figure 6.11). Then the plastic moment capacities are computed as
follows.
M pf  Z xf Fyf (6.39)
M pb  Z xb Fyb (6.40)
where Z xf and Z xb are the effective plastic section moduli of the flange and bearing
stiffener pair, and Fyf and Fyb are the yield strengths of the flange and bearing
stiffeners, respectively.
To evaluate the accuracy of Eqs. (6.37) and (6.38), a parametric study with 99
cases including Models SG1 and SG2 (reported in Chapter 5) was conducted. For each
case, a p and b p are the plastic hinge location parameters predicted by either Eq. (6.37)
or (6.38). These values were then compared with those ( ai and bi ) directly identified
from nonlinear finite element analyses. Figure 6.12(a) and Figure 6.13(a) show the
correlation with the Chern-Ostapenko model, which on average, this model underpredicts
both a and b. Figure 6.12(b) and Figure 6.13(b) show the correlation with the Rockey-
Skaloud model, which tends to overpredict the a and b values. The actual effective
depth of the web falls between these two models.
As a trial to fit the data, two suggested effective depth equations are plotted in
Figure 6.14(a). In the figure, the two straight lines intersect at the point of 0.379 on the
horizontal axis (C = 0.421). As shown in Figure 6.14(b), one straight line, defined by
Eq. (6.37), was drawn from 0 to the intersection point to be consistent with US practice
of inelastic buckling criteria, and another straight line, defined by Eq. (6.38), was drawn
139
beyond the intersection point. The combined lines can be expressed as a power function
using a regression.
2
2   
d e  1t w 0.8  C   1t w  0.8  cr  (6.41)
  yw 

The variables of 1 and  2 in Eq. (6.41) from the regression are 34.1 and 1.6,
respectively. The power curve is also plotted in Figure 6.14(b). Those values ( 1 and
 2 ) are rounded off to the nearest integers of 35 and 2. Figure 6.14(c) shows the power
curve that is calibrated with and lies between the two straight lines defined by Eqs. (6.37)
and (6.38). As a result, it is proposed that Eq. (6.42) be used to determine the effective
web depth.
2
    cr
d e  35t w  0.8  cr  for  0.8 (6.42a)
  yw   yw

 cr
de  0 for  0.8 (6.42b)
 yw
By using the proposed formula for effective depth, the identified and predicted
plastic hinge locations are shown in Figure 6.12(c) for the plastic hinge in the flange and
in Figure 6.13(c) for that in the bearing stiffener pair (see Table 5.1 through Table 5.4 for
the identified plastic hinge locations). From the figures, the proposed equation is
capable of reasonably estimating the plastic hinge locations. In addition, the average
ratios ( a p / a i , and b p / bi ) calculated by the proposed Eq. (6.42) are 0.960 and 1.009,
respectively, with standard deviations of 0.102 and 0.148, respectively.
6.8 Validation of Shear Strength Formula
The end panel shear strengths predicted by Eq. (6.33) are compared with available
test results. Table 6.1 and Table 6.2 provide plate sizes and material properties of the
test specimens, while Table 6.3 summarizes the ultimate shear strength from testing and
nominal shear strengths from Eq. (6.33), AASHTO Specifications (2014), and Eurocode
3. The tables include two tested specimens (SG1 and SG2) in this test program and EPS
series which were tested by Huslid and Rockey (1979). Figure 6.15 compares three
140
predicted shear strengths, which are normalized by the shear strengths from testing. The
predicted-to-test ratio varies from 0.837 to 1.116, and the average ratio is 0.972 with a
standard deviation of 0.078, while the average AASHTO-to-test and Eurocode-to-test
ratios are 0.283 and 0.722 with the standard deviations of 0.122 and 0.147, respectively.
The predictions provided by the proposed model are significantly better than those from
AASHTO Specifications and Eurocode 3.
To further validate Eq. (6.33), the parametric study results from the finite element
analyses described in Chapter 5 are also compared with the predicted values (see Figure
6.16 and Figure 6.17). The shear strengths predicted by Eq. (6.33) correlate very well
with the results from finite element analyses. Table 6.4 and Table 6.5 summarize the
results from nonlinear finite element analyses and predicted shear strengths for SG1 and
SG2 model series, respectively.
Figure 6.18 compares the shear strength for 114 cases (15 from test results and 99
from FEM). The horizontal axis is the predicted shear strength from Eq. (6.33), and the
vertical axis is the shear strength from testing or finite element analysis. The mean ratio
of predicted shear strength to the test (or FEM) is 0.965 with a standard deviation of
0.058. Therefore, it can be concluded that Eq. (6.33) provides a highly accurate
prediction of the ultimate shear strength of end panel.
Figure 6.19 visualizes the results. The black bars correspond to the end panel
shear strength predicted by AASHTO Specifications, which is equal to beam-action
buckling shear strength. The horizontally-hatched bars represent the predicted post-
buckling strength due to the tension-field action in the end panel. Comparison of these
two quantities highlights the contribution to the ultimate shear strength of end panels
predicted by Eq. (6.33), relative to that predicted by AASHTO. Shear strengths
obtained through testing or finite element analysis are also shown in the figure. Eq.
(6.33) very closely matches or slightly underestimates the ultimate shear strengths from
testing and finite element analysis. For comparison purposes, predicted shear strengths
per Eurocode 3 are also provided in the figure. Eurocode 3 generally tends to
underestimate the ultimate shear strengths of the end panel more conservatively than the
ones predicted by Eq. (6.33), but it significantly overestimates the ultimate shear
strengths when thick flanges with small width-to-depth ratios were used.
141
6.9 Shear Resistance of Steel-Concrete Composite Girder End Panel
By providing shear connectors between the concrete slab and steel plate girder,
composite action will increase the flexural strength of the plate girder. The shear
connectors, which are welded to the plate girder top flange and are embedded in the
reinforced concrete slab, accommodate longitudinal (or horizontal) shear force to
minimize the slip. The degree of composite action depends on the strength and the
number of shear stud connectors used.
Experimental test results showed that the plastic hinge locations in the top flange
and bearing stiffeners of the composite girder specimens (CG1 and CG2) are practically
the same as in the steel girder specimens (SG1 and SG2), as shown in Figure 3.35.
Therefore, the ultimate shear strength of the composite plate girder end panel can be
determined by using the superposition of the shear strengths of the steel girder and
concrete slab.
V n  V s  Vc (6.43)
where Vs is computed by using Eq. (6.33). The testing of the two composite girders
exhibited diagonal shear cracks as the primary failure mode in the concrete slab. Based
on the observation of the shear cracks from testing and the findings of the compressive
stress flow in the concrete slab from nonlinear finite element analysis, it is assumed that
one shear crack (upper shear crack) starts from the plastic hinge location A (see Figure
6.20 and Figure 6.21) on the steel girder flange, and the other shear crack (lower shear
crack) starts from the base of the shear stud nearest to the bearing stiffeners (see Figure
6.20 and Figure 6.21). Then, the cracks propagate to the heads of the second and third
nearest studs from the girder end. The assumed crack patterns are compared with the
test results in Figure 6.20 and Figure 6.21 for CG1 and CG2, respectively. The assumed
crack angles are very close to those from the test results. Based on these failure modes,
a strut-tie model was used to evaluate the shear capacity of the concrete slab. The
concrete between the two shear cracks shown in Figure 6.20 and Figure 6.21 serves as a
compression strut. Figure 6.22 shows the failure mode of concrete slab from Specimen
CG1. The compression strut resulted in crushing of the concrete slab.
142
Based on ACI 318 (2008), the nominal compressive strength ( f ns ) of a
compressive strut is determined by multiplying the effective compressive strength of the
concrete ( f ce ) by the cross-sectional area of the strut ( Acs ).
f ns  f ceAcs (6.44)
where
f ce = 0.85 s f c' ,
1.0 for a strut with a uniform cross-section,

s =
0.75 for a bottle shape strut with sufficient transverse reinforcement,
f c' = concrete compressive strength, and
Acs = cross-sectional area perpendicular to the axis of the strut at one end.
Two different angles of the shear cracks will create a bottle shape of the strut and,
therefore, 0.75 equals  s . Figure 6.23 shows an analytical model based on these
assumptions for the load transfer mechanism in concrete. The strut angle can be
determined as
h 
  tan 1  s  (6.45)
 Sh 
The diagonal compressive strut force can be decomposed into the horizontal and
vertical components. The vertical force is resisted by the bearing stiffener pair, and the
horizontal force is anchored by the shear studs which are welded onto the flange.
Therefore, the effective strut width ( bst ) can be assumed as the distance between the
studs as shown in Figure 6.24.
The cross-sectional area of the compressive strut is determined by multiplying the
effective strut width ( bst ) by the compressive strut depth ( d st ) at the end of the strut,
which has the smallest cross-sectional area. The strut depth and area can be determined
as
d st  S h1 sin  (6.46)
Acs  bst d st  bst S h1 sin  (6.47)
143
Taking the vertical component of the inclined compressive strut force, the concrete shear
strength is
Vc  f ns sin   0.85 s f c' Acs sin   0.85(0.75) f c' bst S h1 sin 2  (6.48)
The shear strengths calculated in this manner are 32.7 kips and 17.8 kips for CG1 and
CG2, respectively (see Table 6.6). Adding the shear strength contribution from the steel
girder, the ultimate shear strength of the composite girder end panel can be determined by
using Eq. (6.43).
Table 6.6 compares the predicted shear strengths with test results. The ratios of
the predicted strength to that from testing are 1.026 and 0.975 for CG1 and CG2,
respectively.
144
Table 6.1: Plate Girder Test Specimen Dimensions
Spec. D d0 tw bf tf tb bb e
No. (in.) (in.) (in.) (in.) (in.) (in.) (in.) (in.)
SG1 52.5 26.25 0.247 16 0.75 1.01 14.75 4
SG2 52.5 26.25 0.183 14 0.62 1.01 12.94 4
EPS1-1 18.5 27.76 0.064 4 0.25 0.24 3.94 0
EPS1-2 18.5 27.76 0.065 4 0.25 0.49 3.94 0
EPS4-3 18.5 27.76 0.065 4 0.59 0.59 3.94 0
EPS2-4 18.5 27.76 0.065 4 0.25 0.39 3.94 0
EPS2-5 18.5 27.76 0.065 4 0.25 0.59 3.94 0
EPS3-6 18.5 27.76 0.065 4 0.59 0.24 3.94 0
EPS3-7 18.5 27.76 0.066 4 0.59 0.5 3.94 0
EPS5-8 18.5 27.83 0.063 4 0.24 0.48 3.93 0
EPS5-9 18.5 27.79 0.063 4 0.24 0.48 3.93 0
EPS6-10 18.5 27.81 0.063 4 0.24 0.59 3.94 0
EPS7-11 18.5 18.55 0.039 4 0.24 0.47 3.94 0
EPS7-12 18.5 23.14 0.039 4 0.24 0.48 3.94 0
EPS8-13 18.5 27.64 0.039 4 0.24 0.40 3.91 0

Note: SG series from this study and EPS series from Huslid and Rockey (1979).
145
Table 6.2: Material Properties and Identified Plastic Hinge Locations from Tests
Spec. Fyw (ksi) Fyf (ksi) Fyb (ksi) a (in.) b (in.)

No.
SG1 60.3 57.8 41.1 10.5 18.1
SG2 57.4 51.5 41.1 9.1 14.8
EPS1-1 27.1 45.4 45.0 8.5 3.2
EPS1-2 27.3 45.7 40.6 7.4 6.0
EPS4-3 28.5 34.7 35.1 10.2 7.2
EPS2-4 28.2 46.1 34.3 5.7 5.2
EPS2-5 29.4 45.7 34.3 5.6 6.6
EPS3-6 29.7 34.0 44.7 9.1 5.1
EPS3-7 28.2 34.3 36.1 8.5 8.4
EPS5-8 24.5 31.7 37.7 5.4 5.8
EPS5-9 25.5 31.8 37.7 6.2 5.2
EPS6-10 26.0 30.7 34.4 5.6 6.0
EPS7-11 25.3 31.2 35.8 5.4 7.1
EPS7-12 26.5 31.6 34.6 5.8 6.2
EPS8-13 24.0 29.9 38.4 6.3 6.1
146
Table 6.3: End Panel Shear Strength Comparison between Prediction and Experiment
Results
Shear Strength, Vs (kips)

Spec. Prediction AASHTO Eurocode
No. Predic- AAS- Euro- Test Test Test
Test
tion HTO code
SG1 315.3 328.9 190.2 326.1 1.04 0.60 1.03
SG2 190.3 188.5 76.6 178.6 0.99 0.40 0.94
EPS1-1 8.4 8.4 2.7 5.9 0.99 0.32 0.71
EPS1-2 9.0 9.7 2.9 6.1 1.08 0.32 0.68
EPS4-3 12.9 13.2 2.8 10.1 1.02 0.22 0.78
EPS2-4 8.9 9.2 2.9 6.3 1.04 0.32 0.71
EPS2-5 9.2 10.3 2.9 6.5 1.12 0.31 0.71
EPS3-6 11.6 10.2 2.9 10.5 0.88 0.25 0.91
EPS3-7 13.7 12.3 2.9 10.1 0.89 0.21 0.74
EPS5-8 8.2 7.9 2.6 5.6 0.96 0.31 0.68
EPS5-9 8.6 8.1 2.6 5.8 0.94 0.30 0.68
EPS6-10 8.9 8.5 2.6 5.9 0.954 0.29 0.66
EPS7-11 7.1 5.9 0.8 3.4 0.84 0.12 0.49
EPS7-12 5.6 5.4 0.7 3.2 0.97 0.13 0.57
EPS8-13 4.8 4.3 0.6 2.6 0.88 0.13 0.55
Average 0.97 0.28 0.72
Standard Deviation 0.08 0.12 0.15
Range (min./max.) 0.84/1.12 0.12/0.60 0.49/1.03
147
Table 6.4: Model SG1 Series: End Panel Shear Strength Comparison between Prediction and Design Codes
(Based on Fyw  60.3 ksi, Fyf  57.8 ksi, Fyb  41.1 ksi, b f  16 in., bb  14.75 in., and e  4 in.)

Model D d0 tw tf tb Prediction AASHTO Eurocode
No Preidic AAS- Euro- FEM FEM FEM
(in.) (in.) (in.) (in.) (in.) FEM
-tion HTO code
SG1-1 0.5 312.2 294.9 197.0 284.2 0.95 0.63 0.91
SG1-2 0.75 315.7 300.9 197.0 284.2 0.95 0.62 0.90
52.75 0.5
SG1-3 1.0 316.8 308.4 197.0 284.2 0.97 0.62 0.90
SG1-4 1.5 318.2 325.9 197.0 284.1 1.02 0.62 0.89
SG1-5 0.5 316.9 318.2 196.5 330.9 1.00 0.62 1.04
SG1-6 0.75 321.8 327.2 196.5 330.7 1.02 0.61 1.03
52.5 0.75
SG1-7* 1.0 324.3 333.8 196.5 330.7 1.03 0.61 1.03
SG1-8 1.5 328.1 350.0 196.5 330.6 1.07 0.60 1.01
26.25
SG1-9 0.5 323.0 328.0 195.9 379.7 1.02 0.61 1.18
SG1-10 0.75 328.2 342.8 195.9 379.6 1.04 0.60 1.16
52.25 1.0
SG1-11 0.250 1.0 331.9 360.8 195.9 379.5 1.09 0.59 1.14
SG1-12 1.5 347.0 376.1 195.9 379.1 1.08 0.57 1.09
SG1-13 0.5 336.5 350.3 194.8 493.1 1.04 0.58 1.47
SG1-14 0.75 3443 363.9 194.8 492.9 1.06 0.57 1.43
51.75 1.5
SG1-15 1.0 362.8 380.5 194.8 492.5 1.05 0.54 1.36
SG1-16 1.5 410.1 419.7 194.8 491.4 1.02 0.48 1.20
SG1-17 0.5 226.3 207.4 78.6 168.4 0.92 0.34 0.74
SG1-18 0.75 228.3 212.1 78.6 168.4 0.93 0.34 0.74
52.75 0.5
SG1-19 52.50 1.0 228.2 218.1 78.6 168.4 0.96 0.34 0.74
SG1-20 1.5 230.6 232.8 78.6 168.3 1.01 0.34 0.73
SG1-21 52.5 0.75 0.5 231.9 220.4 78.6 191.2 0.95 0.34 0.83
* Model SG1
148
Table 6.4: Model SG1 Series: End Panel Shear Strength Comparison between Prediction and Design Codes (continued)

(in.) (in.) (in.) (in.) (in.) FEM
-tion HTO code
SG1-22 0.75 234.7 232.3 78.6 191.1 0.99 0.34 0.81
SG1-23 52.5 0.75 1.0 238.2 237.8 78.6 191.0 1.00 0.33 0.80
SG1-24 1.5 243.7 251.5 78.6 190.8 1.03 0.32 0.78
SG1-25 0.5 241.1 228.5 78.6 215.7 0.95 0.33 0.90
SG1-26 0.75 247.1 241.1 78.6 215.5 0.98 0.32 0.87
52.25 1.0
SG1-27 52.50 1.0 252.2 256.6 78.6 215.3 1.02 0.31 0.85
SG1-28 1.5 259.5 273.4 78.6 215.1 1.05 0.30 0.83
SG1-29 0.5 251.0 247.7 78.6 273.0 0.99 0.31 1.09
SG1-30 0.75 259.2 259.5 78.6 272.7 1.00 0.30 1.05
51.75 1.5
SG1-31 1.0 275.0 274.1 78.6 272.2 1.00 0.29 0.99
SG1-32 1.5 305.4 309.0 78.6 271.1 1.01 0.26 0.89
SG1-33 0.250 0.5 197.9 169.4 56.7 143.5 0.86 0.29 0.73
SG1-34 0.75 199.1 173.0 56.7 143.4 0.87 0.29 0.72
52.75 0.5
SG1-35 1.0 199.4 177.6 56.7 143.4 0.89 0.28 0.72
SG1-36 1.5 200.9 189.0 56.7 143.4 0.94 0.28 0.71
SG1-37 0.5 203.3 177.9 56.8 158.2 0.88 0.28 0.78
SG1-38 0.75 205.3 188.0 56.8 158.2 0.92 0.28 0.77
52.5 78.75 0.75
SG1-39 1.0 207.8 192.6 56.8 158.0 0.93 0.27 0.76
SG1-40 1.5 211.6 203.2 56.8 157.9 0.96 0.27 0.75
SG1-41 0.5 211.6 184.1 56.9 174.6 0.87 0.27 0.83
SG1-42 0.75 215.7 193.9 56.9 174.4 0.90 0.26 0.81
52.25 1.0
SG1-43 1.0 219.0 206.0 56.9 174.2 0.94 0.26 0.80
SG1-44 1.5 223.7 220.2 56.9 174.0 0.99 0.25 0.78
149

(in.) (in.) (in.) (in.) (in.) FEM
-tion HTO code
SG1-45 0.5 220.4 199.0 57.1 213.0 0.90 0.26 0.97
SG1-46 0.75 226.0 208.2 57.1 212.7 0.92 0.25 0.94
51.75 78.75 0.250 1.5
SG1-47 1.0 232.7 219.5 57.1 212.5 0.94 0.25 0.91
SG1-48 1.5 249.1 247.0 57.1 211.8 0.99 0.23 0.85
SG1-49 26.25 407.0 435.6 312.0 423.5 1.07 0.77 1.04
SG1-50 52.5 0.292 302.7 293.1 124.8 246.0 0.97 0.41 0.82
SG1-51 78.75 266.0 240.3 90.1 205.1 0.90 0.34 0.77
SG1-52 26.25 324.3 333.8 196.5 330.7 1.03 0.61 1.02
SG1-53 52.5 0.250 238.2 237.8 78.6 191.0 1.00 0.33 0.80
SG1-54 78.75 207.8 192.6 56.8 158.0 0.93 0.27 0.76
SG1-55 26.25 267.1 273.5 131.6 270.0 1.02 0.49 1.01
SG1-56 52.5 52.5 0.219 0.75 1.0 199.8 200.2 52.7 154.7 1.00 0.26 0.77
SG1-57 78.75 172.9 160.5 38.0 126.9 0.93 0.22 0.73
SG1-58 26.25 219.7 223.3 82.9 216.8 1.02 0.38 0.99
SG1-59 52.5 0.188 164.9 166.1 33.2 123.0 1.01 0.20 0.75
SG1-60 78.75 141.3 132.1 23.9 99.5 0.93 0.17 0.70
SG1-61 26.25 192.6 190.6 55.5 181.8 0.99 0.29 0.94
SG1-62 52.5 0.164 144.2 143.1 22.2 102.1 0.99 0.15 0.71
SG1-63 78.75 119.5 113.2 16.0 81.6 0.95 0.13 0.68
Average 0.978 0.38 0.90
Range (min./max.) 0.86/1.09 0.13/0.77 0.68/1.47
150
Table 6.5: Model SG2 Series: End Panel Shear Strength Comparison between Prediction and Design Codes
(Based on Fyw  57.4 ksi, Fyf  51.5 ksi, Fyb  41.1 ksi, b f  14 in., and bb  12.94 in.)

Model D d0 e tw tf tb Prediction AASHTO Eurocode
(in.) (in.) (in.) (in.) (in.) (in.) FEM
-tion HTO code
SG2-64 0.563 195.1 179.9 83.1 179.3 0.92 0.43 0.92
SG2-65 52.69 0.563 0.750 197.2 184.0 83.1 179.3 0.93 0.42 0.91
SG2-66 1.125 199.6 194.3 83.1 179.2 0.97 0.42 0.90
SG2-67 0.563 202.3 195.7 82.9 203.1 0.97 0.41 1.00
SG2-68 52.50 26.25 0.750 0.750 204.9 200.4 82.9 203.0 0.98 0.41 0.99
SG2-69 1.125 210.4 209.9 82.9 202.9 1.00 0.39 0.96
SG2-70 0.563 216.3 208.3 82.5 261.2 0.96 0.38 1.21
SG2-71 52.13 1.125 0.750 229.6 218.6 82.5 260.9 0.95 0.36 1.14
SG2-72 1.125 253.2 243.7 82.5 260.4 0.96 0.33 1.03
SG2-73 0.563 141.9 131.2 34.5 107.2 0.92 0.24 0.76
SG2-74 52.69 0.563 0.750 143.9 134.4 34.5 107.2 0.93 0.24 0.75
4 0.188
SG2-75 1.125 147.2 142.6 34.5 107.1 0.97 0.23 0.73
SG2-76 0.563 150.4 142.0 34.5 119.5 0.94 0.23 0.80
SG2-77 52.50 52.5 0.750 0.750 153.1 147.1 34.5 119.4 0.96 0.23 0.78
SG2-78 1.125 157.4 154.8 34.5 118.7 0.98 0.22 0.76
SG2-79 0.563 164.6 152.1 34.5 149.6 0.92 0.21 0.91
SG2-80 52.13 1.125 0.750 173.6 160.5 34.5 149.3 0.92 0.20 0.86
SG2-81 1.125 187.5 181.1 34.5 148.9 0.97 0.18 0.80
SG2-82 0.563 123.1 104.8 23.9 87.2 0.85 0.19 0.71
SG2-83 52.69 0.563 0.750 124.4 107.3 23.9 87.2 0.86 0.19 0.70
78.75
SG2-84 1.125 126.5 113.6 23.9 87.1 0.90 0.19 0.69
SG2-85 52.50 0.750 0.563 130.2 112.8 23.9 95.0 0.87 0.18 0.73
151

Model D d0 e tw tf tb Prediction AASHTO Eurocode
(in.) (in.) (in.) (in.) (in.) (in.) FEM
-tion HTO code
SG2-86 0.750 132.0 117.1 23.9 95.0 0.89 0.18 0.72
52.50 0.750
SG2-87 1.125 134.5 123.1 23.9 94.9 0.92 0.18 0.71
SG2-88 78.75 4 0.563 138.0 120.6 24.0 114.5 0.87 0.17 0.83
SG2-89 52.13 1.125 0.750 143.3 127.2 24.0 114.4 0.89 0.17 0.80
SG2-90 1.125 151.5 143.4 24.0 114.0 0.95 0.16 0.75
SG2-91 0 196.7 189.0 83.0 189.9 0.96 0.42 0.97
SG2-92* 26.25 4 201.4 195.5 83.0 202.4 0.97 0.41 1.01
0.188
SG2-93 8 210.2 212.1 83.0 213.4 1.01 0.40 1.02
SG2-94 0 143.1 138.3 33.2 144.6 0.97 0.23 1.01
SG2-95 52.63 52.5 4 0.625 1.0 149.2 143.4 33.2 150.3 0.96 0.22 1.01
SG2-96 8 154.2 156.7 33.2 155.9 1.02 0.22 1.01
SG2-97 0 124.4 110.3 23.9 138.2 0.89 0.19 1.11
SG2-98 78.75 4 128.4 114.2 23.9 142.2 0.89 0.19 1.11
SG2-99 8 132.3 124.7 23.9 146.1 0.94 0.18 1.10
* Model SG2 Average 0.94 0.27 0.89
Range (min./max.) 0.85/1.02 0.16/0.43 0.69/1.21
152
Table 6.6: Composite Girder Shear Strength Prediction
Specimen f c' hs Sh S h1  Acs Vc Vs Vn VTEST Vn

No. (ksi) (in.) (in.) (in.) (degree) (in.2) (kips) (kips) (kips) (kips) VTEST
CG1 5.0 5.5 13.5 4.5 22.2 21.6 26.0 328.9 354.9 352.6 1.01
CG2 5.4 4.0 12.75 4.25 17.4 13.4 13.8 188.5 202.3 212.1 0.95
153
Possible Plastic a
Hinge Location at the
Collapse Stage, Typ.
Bearing σt
Stiffener
a
(a) Plate Girder End Panel (b) Failure Mode 1

a a
b
b  b
a
a  a
(c) Failure Mode 2 (d) Failure Mode 3
Figure 6.1: Failure Modes
154
a
Plastic
A B C
Hinge
Location,
Typ.
b
Assumed
Tension s
J
Band Width
D
t t w

H F
G
Front Side Back Side d0
Figure 6.2: Tension-Field Band Width
155
A
B
C
B B
A
F
 a J
A
B C
b
J
(b) Kinematics (Upper Left Corner)
V ps
( D  b) F
D b J
E 
 H
G F
F
bd 0 / D
d0
V ps
(a) Assumed Collapse Mechanism
H F
G
(bd 0 / D) F F
bd 0 / D
(c) Kinematics (Lower Portion)
Figure 6.3: Assumed End Panel Collapse Mechanism
156
a
M pf
A
A B C
B
M pm
t t w sin 
2
t t w sin  cos 
M pm
A
J t t w sin  cos 
V ps
b t t w cos 2 
E
J
M pb
t t w sin 2 
M pf
H
G F
G F
bd 0 / D
V ps
(a) Free-Body of End Panel (b) Free Bodies of Boundary
Member Segments
Figure 6.4: Free-Body Diagram of End Panel with Web Excluded
157
(1   )at t w sin 2  a at t w sin 2 
M pf
A
(1  )bt t w cos 2 
B
M pm t t w sin 
2
at t w sin 2   (1  )bt t w cos 2 
a
(1   )at t w sin 2 
A B (1  )bt t w cos 2 
b M pm t t w sin  cos 
J A
(a) Plastic Hinges

in End Panel J
M pb
bt t w cos 2 
bt t w sin  cos   (1   )at t w sin 2 
(b) Free-Bodies
Figure 6.5: Free-Body Diagram of Boundary Members between Plastic Hinges
158
200
150
180
Vps (kips)
100 170
160
-0.4
-0.2
50 150  0
0.2
140 0.4
-0.10 0.0 0.10
0
-0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5

(a) Model SG1
180
160
140
120
Vps (kips)
100 140
80 130
60 120 -0.4
-0.2
40 110  0
0.2
20 100 0.4
-0.10 0.0 0.10
0
-0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 0.2 0.3 0.4 0.5

(b) Model SG2
Figure 6.6: Effect of Shear Ratio Parameters
159
t t w a sin 2  a M pf M pf
t t wb cos 2 A B
B C
M pm
t t w sin 2 
t t w a sin 2  
t t w a sin  cos   b cos 2  
t t wb cos 2
M pm
A t t w sin  cos 
H t t w sin  cos 
M pb

t t w a sin 2   b sin  cos  
M pb
J
E
aD
t t w cos 2 
d0
F
M pf  M pb  t t wb
2
 bd 
 cos 2  t t w  0  sin 2 
d0 2d 0  D 
t t w sin 2 
M pf
H
G F
F
bd 0
 bd 
D t t w  0  sin  cos 
 D 
V ps
Figure 6.7: Free-Body of End Panel Due to Tension-Field Action
160
5.0
4.5 bearing
stiffeners
Flange Shear Contribution (%) 4.0 flange
weaker
weaker
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Mpf / Mpb
Figure 6.8: Flange Shear Contribution
bf
A
de
A
Section A - A
Effective Web Area
B B
tw
de
e+tb/2+de
tb
Section B - B
Figure 6.9: Effective Web Depth, de
161
cr
C
 yw
Elastic Buckling Zone 1

(AASHTO, AISC, Basler, and Chern and Ostapenko)
1.0
0.8
Elastic Buckling Zone 2
(Rockey and Skaloud, and Porter et al.)
0.5
Ek
1.40
Fyw Based on AASHTO (2010) D
tw
Figure 6.10: Relationship between C and D/tw
tb
bf
tw
tf
de tw
de
(a) Girder Flange (b) Bearing Stiffeners
Figure 6.11: Effective Sections of Boundary Members
162
0.8
a CO
p ai
Mean : 0.849
0.6
STDEV : 0.078
ai
0.4
d0
45o line
0.2
0
0 0.2 0.4 0.6 0.8
CO
a p d0
(a) from Chern and Ostapenko (1969) [Eq. (6.37)]
0.8
a RS
p ai
Mean : 1.082
0.6
STDEV : 0.172
ai
0.4
d0
0.2
0
0 0.2 0.4 0.6 0.8
a RS
p d0
(b) from Rockey and Skaloud (1972) [Eq. (6.38)]
0.8
a p ai
Mean : 0.960
0.6
STDEV : 0.102
ai
0.4
d0
0.2
0
0 0.2 0.4 0.6 0.8
a p d0
(c) from Proposed [Eq. (6.42)]
Figure 6.12: Comparison of Plastic Hinge Location a in Top Flange
163
0.8
bpCO bi
Mean : 0.928
0.6
STDEV : 0.180
bi
0.4
D
45o line
0.2
0
0 0.2 0.4 0.6 0.8
bpCO D
(a) from Chern and Ostapenko (1969) [Eq. (6.37)]
0.8
bpRS bi
Mean : 1.101
0.6
STDEV : 0.133
bi
D 0.4
0.2
0
0 0.2 0.4 0.6 0.8
b pRS D
(b) from Rockey and Skaloud (1972) [Eq. (6.38)]
0.8
b p bi
Mean : 1.009
0.6
STDEV : 0.148
bi
0.4
D
0.2
0
0 0.2 0.4 0.6 0.8
bp D
(c) from Suggestion [Eq. (6.42)]
Figure 6.13: Comparison of Plastic Hinge Location b in Bearing Stiffeners
164
35
30
Rockey and Skaloud
25
d e 20
tw Chern and Ostapenko
15
10 Intersection
0
0.0 0.2 0.4 0.6 0.8 1.0
(0.8  C )
(a) Effective Web Depth from Eqs. (6.37) and (6.38)
35
30 Rockey and Skaloud
25
d e 20
tw
15
be
10  34.1(0.8  C )1.6
tw 2
R  0.94
5
Chern and Ostapenko
0
0.0 0.2 0.4 0.6 0.8 1.0
(0.8  C )
(b) Regression
Figure 6.14: Comparison of Effective Web Depth, de
165
35
30 Rockey and Skaloud
25
d e 20
Proposed
tw Eq. (6.42)
15
10
5
Chern and Ostapenko
0
0.0 0.2 0.4 0.6 0.8 1.0
( 0. 8  C )
(c) Suggested Effective Web Depth
Figure 6.14: Comparison of Effective Web Depth, de (continued)
166
1.6
1.4 SG EPS
Spec. Spec.
Prediction / Test
1.2
Shear Strength Ratio
1.0
0.8
0.6
Eurocode / Test
0.4
0.2
AASHTO / Test
0.0
Test Specimen
Figure 6.15: Comparison of End Panel Shear Strengths (Test Specimens)
1.6
1.4
Prediction / FEM
1.2
1.0
0.8
0.6 Eurocode / FEM
0.4
0.2
AASHTO / FEM
0.0
SG1-63
Model
…………………………………………...
SG1-1
Figure 6.16: Comparison of End Panel Shear Strengths

(SG1 Models from Parametric Study)
167
1.6
1.4
Prediction / FEM
1.2
1.0
0.8
0.6
Eurocode / FEM
0.4 AASHTO / FEM
0.2
0.0
SG2-99
SG2-64
Model
…………………………………………........

(SG2 Models from Parametric Study)
450
45o line
400 Test
FEM
350
VTEST or VFEM (kips)
300
250
200
150
VPrediction / VTEST (or VFEM)
100
Average = 0.965
50 Standard Deviation = 0.058
0
0 50 100 150 200 250 300 350 400 450
VPrediction (kips)
168
500
Predicted Post-Buckling Strength
450
AASHTO Buckling Strength
EuroCode
400
Actual Strength (Test or FEM)
350
300
Vs 250
200
150
100
50
…………………………………………...... ………….................. ………...

SG1
SG1-63
SG2-64
SG2-99
SG1-1
ESP8-13
169
25º
17º
(a) Test Specimen
4.5” 9” 9”
A
17°
22°
10.0”
Plastic
Hinge
Bearing Stiffeners
Intermediate Stiffeners
(b) Assumed Crack Pattern
Figure 6.20: CG1: Shear Crack Comparison in Concrete Slab
170
22º
14º
(a) Test Specimen
4.25” 8.5” 8.5”
A
17° 13°
9.38”
Plastic
Hinge
Bearing Stiffeners
Intermediate Stiffeners
(b) Assumed Crack Pattern
Figure 6.21: CG2: Shear Crack Comparison in Concrete Slab
171
(a) Side View
(b) Top View
Figure 6.22: Specimen CG1: Concrete Crushing
172
dst
hs

Vc Sh1
Sh
Figure 6.23: Analytical Model for Load Transfer Mechanism in Concrete Slab
bst
Figure 6.24: Effective Width of Compression Strut
173
7. REHABILITATION OF END PANELS
7.1 General
The Manual for Bridge Evaluation (AASHTO 2011) is used by bridge engineers
to perform load rating of existing bridges. When the load rating factor is greater than
1.0, the bridges are deemed safe and no further action is required. However, when the
load rating factor is less than 1.0, the bridges must be strengthened to maintain current
vehicular loading. Otherwise, vehicular loading on the bridge must be restricted.
Many bridges have been recently reevaluated. Results showed that some bridges
had a rating factor much smaller than 1.0 due to the deficiency in shear strength in the
end panels, where the shear strength was determined based on the shear buckling strength
per AASHTO Specifications (2014) without taking advantage of any partial tension-field
action and composite slab effect. However, it was shown in Chapter 5 that end panels
could provide a considerable amount of post-buckling strength due to these two factors.
It is believed that using the proposed Eq. (6.33) will eliminate many cases of
unnecessary shear strengthening. If the reevaluated load rating analyses with this
equation show that some bridges still have a rating factor smaller than 1.0 due to the
shear deficiency in the end panels, then these panels need to be reinforced. However,
rehabilitation of the plate girder end panels may be a challenging task to design engineers.
Therefore, guidelines are needed to rehabilitate plate girder end panels.
In this Chapter, available rehabilitation schemes will be reviewed and an alternative
rehabilitation scheme will be proposed.
7.2 Basler’s Scheme
One of the well-known methods to rehabilitating end panels is adding another pair
of stiffeners, or end stiffeners, at the end of the girder (Basler 1961a). Therefore, two
pairs of stiffeners (a pair of bearing stiffeners and a pair of end stiffeners) at or near the
support will resist the horizontal component of the tension-field action force (see Figure
7.1). Basler developed an equation for the required end stiffener area by assuming that
174
the tension-field stress would act at an inclination of 45 degrees and be uniformly
distributed over the entire girder depth (see Figure 7.2). Then, both vertical and
horizontal components induced by the tension-field stresses are (   cr )t w . By
considering the shaded area in Figure 7.2 as a beam with a span of D, the maximum
bending moment due to the horizontal component is (   cr )t w D 2 / 8 . Assuming this
moment is resisted by a force couple in the bearing stiffener and end stiffener pairs, a
maximum compressive force of (   cr )t w D 2 / 8e develops in the end stiffener pair.
Equating the resisting force Ae Fye offered by the end stiffener pair with the above force
leads to Eq. (7.1).

(   cr ) DAw (V  Vcr ) D
Ae   (7.1)
8eFye 8eFye
where Aw (= Dtw) is the web area and Fye is the yield stress of the end stiffeners.
To evaluate the adequacy of Eq. (7.1), nonlinear finite element analyses were
performed again using Model SG1-7 (see Table 6.4). For Eq. (7.1) to be valid, the end
panel should have the same shear strength as if it was an interior panel.
Based on Eq. (6.33) with  =1, which corresponds to the ultimate shear strength
of an interior panel ( V sint ) as specified by AASHTO Specifications (2014), the expected
shear strength of the rehabilitated Model SG1-7 is 401 kips ( Vsint = 401 kips). The
required area of end stiffener pair, Ae , from Eq. (7.1) is 8.2 in.2 by using the same yield
stress as the bearing stiffeners of Model SG1-7 ( Fye  Fyb  41.1 ksi). It was assumed
that the end stiffeners have the same width as bearing stiffeners ( bes = bb =14.75 in.).
Dividing the end stiffener area by the total width of the end stiffeners ( bes ) gives a
thickness ( t es ) of 0.56 in.

Guided by the following criterion for local buckling control of each end stiffener
(AASHTO 2014),
bes  t w E
 0.38 (7.2)
2t es Fy
175
the required thickness t es is about 0.7 in. Although AASHTO requires a thickness of
0.7 in., a thickness of 0.56 in. is assigned to Model SG1R-EP1 for the comparison
purposes. Then, a thickness of 0.75 in. is assigned to Model SG1R-EP2 and the
thickness continues to increase up to 3.0 in. for the remaining models [see Table 7.1(a)].
Figure 7.3(a) shows the finite element model of the rehabilitated Model SG1-7.
Nonlinear finite element analyses were carried out with various areas of end stiffeners.
If the end stiffener pair functions as an effective anchor for the end panel to achieve the
same shear capacity as an interior panel, it was postulated that both end and interior
panels would fail simultaneously.
Table 7.1(a) shows a summary with different end stiffener areas, while Figure 7.4
shows the failure mode associated with each. The end stiffener pair of SG1R-EP1,
which was designed per Basler’s recommendations, did not properly serve as an anchor
and failed to achieve the expected shear strength ( Vsint = 401 kips). Note that the end
panel shear strength without an end stiffener pair was 324 kips (see Model SG1-7 in
Table 6.4), only 11% less than that of SG1R-EP1 with the end stiffener pair. As the end
stiffener area increased, so did the end panel shear strength. Then, when the area of the
end stiffener pair reached 29.50 in.2 (SG1R-EP5 in Table 7.1), the end panel reached a
shear strength of 402 kips. In addition, the interior panel also experienced minor web
buckling as shown in Figure 7.4. SG1R-EP6, which had an end stiffener area of 36.88
in.2, reached an ultimate shear strength of 408 kips and showed the expected behavior
with both end and interior panels failing simultaneously (see Figure 7.4). Increasing the
end stiffener area further did not increase the ultimate shear strength, as the shear strength
and failure mode of SG1R-EP7 were almost the same as in SG1R-EP6. Therefore, it
can be concluded from Table 7.1 and Figure 7.4 that Basler’s recommendation for the
required end stiffener area to have a full tension-field action is non-conservative.
7.3 Eurocode 3 Design Provision
A similar approach has been adopted in Eurocode 3 (2006). Section 9.3 of EN

1993-1-5 (2006) notes that the end stiffener pair can be considered as rigid and the full
tension-field action can develope if the bearing and end stiffener pairs each have a cross
176
sectional area of at least 4 Dt w2 / e , where e is the center-to-center distance between the
stiffener pairs and e  0.1D .

4 Dt w2
Ae and Ab  (7.3a)
e
e  0.1D (7.3b)
where Ae and Ab are the cross section areas of the end stiffener pair and the bearing
stiffener pair, respectively. Following Eq. (7.3), Ae and Ab should be greater than
3.3 in.2 and e should be greater than 5.3 in. to provide full tension-field action for
Model SG1-7. In the parametric study to follow, e was set at 5.5 in. as shown in
Table 7.1(b). For a given total width ( bes in Figure 7.1) of the end stiffener pair of
14.75 in., the required thickness ( t es  0.224 in. ) was too small based on Eq. (7.2).
Following the local buckling criterion of Eq. (7.2), a minimum thickness of 0.75 in. was
assigned to Model SG1-EP8. The remaining four models had increasing stiffener
thicknesses, with SG1R-EP11 having a stiffener area of almost nine times that required
by Eurocode 3. The area of the bearing stiffener pair remained the same as in the
original model (Model SG1-7).
The last column in Table 7.1(b) summarizes the predicted shear strength and
Figure 7.5 shows the associated failure modes. Although SG1R-EP9 and 10 had areas
which were more than 3 times that required by Eurocode 3, both models did not achieve
full tension-field action. The same observation was made by Alinia et al. (2009b) for
the use of end stiffeners, and they commented that the minimum required area of end
stiffener pair specified in Eurocode 3 was non-conservative. SG1R-EP10, which had an
end stiffener area equal to 6.7 times that required by Eurocode 3, successfully achieved
full tension-field action; it was able to achieve the expected shear strength of 401 kips
and showed the expected failure mode.
The parametric study presented above showed that both the Basler and Eurocode
3 schemes of adding end stiffener pair may not be effective in mobilizing the full tension-
field action in the end panels, unless a very large cross sectional area of end stiffener
pairs is used. An alternate scheme is presented in the next section.
177
7.4 Proposed Scheme 1
To develop tension-field action in the end panels, the boundary members must
provide enough flexural strength to serve as an anchor. Also, as presented in Section
6.3, the tension-band width in the web is directly related to the plastic hinge locations in
the boundary members. That is, if the plastic hinge location along the bearing stiffeners
can be shifted downward, the tension-band width in the web and the shear strength in the
end panel will be increased.
One method of increasing the flexural strength of the vertical boundary members
is to weld wing stiffeners to the bearing stiffeners and both flanges. Figure 7.6(a) shows
the rehabilitation scheme using the wing stiffeners and Figure 7.6(b) shows the effective
section of the vertical boundary. Note that when end stiffeners are used for
rehabilitation, the end stiffeners are always in compression and, therefore, local buckling
can be an issue. However, wing stiffeners are mainly in tension when a tension-field
action develops.
Using the proposed scheme, nonlinear finite element analyses were performed.
Using Model SG1-7 as the base model, the width of the wing stiffener was set to 5.5 in.
and the wing stiffener thickness varied from 0.5 in. to 1.0 in. [see Table 7.2(a)]. Figure
7.3(b) shows the typical finite element model with wing stiffeners. Each wing stiffener
was located at mid-width of each bearing stiffener. A plot of the applied load versus
girder mid-span deflection from analysis on Model SG1R-W1 is shown in Figure 7.7.
The maximum load achieved was 394 kips at a mid-span deflection of 1.37 in., as shown
in Table 7.2(a). To compare the ultimate shear strength from the nonlinear finite
element analysis with the shear strength predicted by Eq. (6.33), the parameter  is also
presented in Table 7.2(a). As wing stiffener thickness increased from 0.5 in. to 1.0 in.,
M pb increased from 964 kip-in. to 1,640 kip-in.
As seen in Table 7.2(a), the predicted shear strengths correlate very well with the
results from nonlinear finite element analysis. Model SG1R-W1 with an  value of
0.92 showed a prediction of about 3% lower than the analysis result. Model SG1R-W2
with an  value of 1.0 had a predicted shear strength which is almost the same as that
from nonlinear finite element analysis. Since Model SG1R-W2 reaches the ultimate
178
shear strength of an interior panel, Model SG1R-W3 showed the same result as SG1R-
W2.
Figure 7.8 shows the failure modes. Interestingly, all failure modes were almost
the same and the use of very thick wing stiffeners is unnecessary. Both the end and
interior panels achieved their ultimate shear strengths at about the same time. Therefore,
the addition of wing stiffeners is very effective in allowing the bearing stiffener pair to
function as an anchor.
To further validate the effectiveness of the wing stiffener concept, a parametric
study using Model SG2-92 (see Table 6.5) as the base model was performed. Note that
Model SG2-92 is expected to have an ultimate shear strength of 273 kips with  =1,
which corresponds to that of an interior panel with tension-field action. Table 7.2(b)
provides a summary of the models, while Figure 7.9 shows the corresponding failure
modes. Similar trends can be observed. Compared to the shear strength obtained from
the nonlinear finite element analysis, the predicted strength from Eq. (6.33) was
underestimated by about 7% in Model SG2R-W1. Model SG2R-W1 had already
achieved the ultimate shear strength of the interior panel based on the nonlinear finite
element analysis. Thus, SG2R-W1, 2, 3, and 4 achieved almost the same strength and
the same failure modes as shown in Figure 7.9.
From the results presented above, it is clear that adding wing stiffeners very
effectively mobilizes tension-field action in the end panels. The proposed shear strength
equation [Eq. (6.33)] provides a very accurate and yet slightly conservative estimate of
the actual shear strength.
Design Procedure
Design of wing stiffeners can be performed by using  determined in Chapter 6.
The design procedure follows:
(1) Given the required shear (Vu), solve Eq. (6.33) for the required  . For example,
if the required  equals 0.8, it means 80% of the full tension-field action is
needed.
(2) Determine the required M pb by substituting the required  into Eq. (6.36). At
this stage, one may assume M pm = M pf . If the calculated M pb is greater than
179
M pf , the assumption is correct. If not, let M pm = M pb to solve the required
M pb .
(3) Size the wing stiffeners such that the plastic moment of the effective section shown
in Figure 7.6(b) is at least equal to M pb .
Taking Model SG2-92 as an example, assume the required shear strength is 240
kips. Note that C  0.25 ,   0.59 , (d 0 / D)  0.5 , V p  328 kips, and the nominal
shear strength ( Vn ) per AASHTO is 82.9 kips before rehabilitation (see Table 6.5 for
details). Therefore, based on Eq. (6.33), the shear strength of the end panel before
rehabilitation is
 0.87(1  C ) 
Vs  V p C     195 kips (7.4)
 1  (d 0 / D) 2 
Step (1): Compute required partial tension-field action coefficient,  .

Given the shear demand, solve Eq. (6.33) for the required  .
  0.87(1  0.25) 
3280.25     240 kips (7.5)
  1  (0.5) 2 
  
Solving Eq. (7.5) gives a required  value of 0.83.
Step (2): Determine the required M pb .
Assuming M pm = M pf = 94.8 kips-in. and using Eq. (6.36),
  0.83 

2.8 M pf  M pm  M pb  M pm 
D t w Fyw (1  C )
 
(7.6)
2.8 94.8  94.8  M pb  94.8

52.5 0.1875(57.4)(1  0.25)
Therefore, the required M pb is 832 kips-in.
Step (3): Size wing stiffeners.

Figure 7.10 shows the effective section of the vertical boundary member. For
simplicity, assume that the effective section has the same yield stress as the bearing
180
stiffeners Fy (= 41.1 ksi). Dividing the required M pb by Fy gives the required
plastic section modulus ( Z xr ):
M pb
Z xr   20.2 in.3 (7.7)
Fy
Starting with a trial width of the wing stiffeners ( bws ) as 5.5 in., the required thickness of
the wing stiffeners ( t ws ) can be computed as follows. The total area of the section is
A  0.188(4  0.5  1.96)  1.0(12.94  0.188)  2(5.5)t ws  14  11t ws (7.8)
Assuming the plastic neutral axis is located within the bearing stiffeners, the plastic
neutral axis is
A  t b  14  11t ws  1 
  t w  e     0.188 4  
(7.9)
2  2   2  2 
d PNA    0.49  0.43t ws
bb 12.94
If d PNA is less than 1.0 in., the plastic neutral axis would be in the bearing stiffeners.
In other words, t ws should be less than 1.2 in. to satisfy the assumed plastic neutral axis
location. Then, the plastic section modulus ( Z y ) of the effective section can be
computed by summing up the first moments of area about the plastic neutral axis. The
effective section is divided into 5 segments about the plastic neutral axis shown in Figure
7.10. Therefore, the plastic section modulus is
Z y   Z i   Ai  i (7.10)
i i
where Ai is the area of each segment and  i is the moment arm from the plastic neutral
axis. The plastic section modulus of each segment is
 t  e  t b 2 
Z 1  t w  e  b   d PNA 
 2  2  (7.11a)
 0.188(3.5)(1.75  0.49  0.42t ws )  1.47  0.28t ws
 (0.49  0.42t ws ) 2 
Z 2  (12.94)   1.55  2.66t ws  1.14t ws

2
(7.11b)
 2 
 (1.0  0.49  0.42t ws ) 2 
Z 3  (12.94)   1.68  2.77t ws  1.14t ws

2
(7.11c)
 2 
181
 1.964 
Z 4  (0.188)(1.964)  1.0  0.49  0.42t ws   0.55  0.16t ws (7.11d)
 2 
  5.5 
Z 5  2 5.5(t ws )  1.0  0.49  0.42t ws   35.86t ws  4.62t ws
2
(7.11e)
  2 
Summing up the plastic section modulus of each segment gives
Z y   Z i  5.25  35.87t ws  2.34t ws
2
(7.12)
i
Setting Z y = Z xr = 20.2 in.3, Eq. (7.12) can be rewritten as

2
2.34t ws  35.87t ws  14.95  0 (7.13)
Solving Eq. (7.13) gives the required thickness of the wing stiffeners.
35.87  35.87 2  4(2.34)(14.95)

t ws 
2(2.34)
(7.14)
35.87  33.86
  0.43 in. or 14.90 in.
4.68
Therefore, the required wing stiffener thickness is 0.43 in. Use 0.5 in. thick
wing stiffeners. Note that wing stiffeners are tension members, and therefore, there is
no need to check for local buckling.
7.5 Proposed Scheme 2
Although rehabilitation of plate girder end panels were discussed and one scheme
was proposed in Section 7.4, a more practical rehabilitation scheme is to add another pair
of stiffeners in the end panels. To avoid field welding in rehabilitation, a pair of angles
can be fastened to the web by bolting (see Figure 7.12). The effectiveness of this
scheme was investigated by finite element analysis in this section. To simplify the
modeling, it was assumed that bolting was as effective as welding. Thus, the model
similar to that in Figure 7.13 was analyzed. It is shown in the figure that the original
end panel is divided into two subpanels, with Subpanel 1 serving as the new end panel.
From the rehabilitation design point of view, it is necessary to determine the
location of the added stiffeners. One approach is to locate the stiffeners closer to the
bearing stiffeners to increase the shear resistance (e.g., for C =1.0). An alternative
approach is to add the stiffeners to divide the original end panels into two equal
182
subpanels. For practical design, it will be shown in the following that the latter
approach is more effective and recommended.
Models SG1 and SG2 with three different width-depth ratios (0.5, 1.0, and 1.5)
were considered. For approach 1, stiffeners were added at a distance of 10.5 in. away
from the bearing stiffeners such that the new end panels have a width-depth ratio of 0.2.
Table 7.3 summarize the six models analyzed. Figure 7.14 and Figure 7.15 present the
failure modes and the relationships between the applied load and mid-span deflection of
SG1 and SG2 series, respectively. Table 7.4 summarizes the ultimate shear strengths
before and after the addition of stiffeners. With the addition of stiffeners, the increase in
shear strengths varied from 14% to 40%. However, as shown in the figures, the new
end panel (subpanel 1) experienced buckling. If the new end panel had a plastic shear
capacity (= 0.58Fyw Aw ), the new end panel should have remained undamaged at that
load level and only the new interior panel (subpanel 2) would buckle. In other words, it
was expected with this rehabilitation approach that the ultimate shear strength of each
finite element model should have been governed by the new interior panel.
Based on AASHTO Specifications (2014), the shear strengths ( V AASHTO ) of the
new interior panels were computed and are summarized in Table 7.5. It is obvious from
the table that the ultimate shear strengths from the finite element analysis ( V FEM ) do not
reach those predicted by the AASTHO Specifications. This is due to the fact that new
end panel (subpanel 1) does not have sufficient strength to provide the anchor mechanism.
Increasing the stiffener thickness from 0.5 in. to 1.0 in. only increases the ultimate
strength slightly (see Table 7.6).
Next, consider approach 2. By relocating the additional stiffeners to the mid-
width of the original end panels, the width-depth ratios of the new end panel (subpanel 1)
and the new interior panel (subpanel 2) are the same. Since the shear strength of
subpanel 1 is usually less than that of subpanel 2, Eq. (6.33) can be used to compute the
shear strength of the new end panel. A total of 14 models were investigated to
demonstrate the effectiveness of this approach. Table 7.7 summarizes the details of
these models. Figure 7.16 and Figure 7.17 show the failure modes of Models SG1 and
SG2 series. As expected, new end panel experienced buckling for all finite element
models. Therefore, the proposed equation [Eq. (6.33)] can be used to predict the
183
ultimate shear strength. Figure 7.18 presents the relationship between applied load and
mid-span deflection of the rehabilitated models, while Table 7.8 summarizes the ultimate
shear strengths before and after the rehabilitation. It is observed from Table 7.4 and
Table 7.8 that adding vertical stiffeners in the mid-width of the original end panel is more
effective in increasing the ultimate shear strength.
The predicted shear strengths by using Eq. (6.33) are also compared with the
results from the finite element analyses in Table 7.9. The predicted ultimate shear
strengths generally correlate well with the results from the finite element analyses. The
mean ratios of FEM to the prediction before and after rehabilitation are 1.03 and 0.97,
respectively, with a standard deviation of 0.04 and 0.06. Therefore, adding stiffeners in
the mid-width of the original end panel is recommended for rehabilitation.
The accuracy of using Eq. (6.33) to predict the shear strength is evaluated in
Figure 7.19, where a total of 28 cases (14 from before rehabilitation and 14 from after
rehabilitation of Models SG1 and SG2 series, see Table 7.8) are presented. It is noted
that Eq. (6.33) tends to overestimate the shear strength by up to approximately 15% when
the width-depth ratio is less than 0.5. Therefore, Eq. (6.33) is applicable only when the
width-depth rati is no less than 0.5.
184
Table 7.1: Effect of End Stiffener Area (Based on Model SG1-7)
(a) Basler’s Approach
Model t es bes tb bb Ae Ab e V FEM
No. (in.) (in.) (in.) (in.) (in.2) (in.2) (in) (kips)
SG1R-EP1 0.56 8.26 361
SG1R-EP2 0.75 11.06 368
SG1R-EP3 1.0 14.75 380
SG1R-EP4 1.5 14.75 1.0 14.75 22.13 14.75 4.0 391
SG1R-EP5 2.0 29.50 402
SG1R-EP6 2.5 36.88 408
SG1R-EP7 3.0 44.25 409
(b) Eurocode 3 Approach

Model t es bes tb bb Ae Ab e V FEM
No. (in.) (in.) (in.) (in.) (in.2) (in.2) (in) (kips)
SG1R-EP8 0.75 11.06 389
SG1R-EP9 1.0 14.75 396

14.75 1.0 14.75 14.75 5.5
SG1R-EP10 1.5 22.13 402
SG1R-EP11 2.0 29.50 403
185
Table 7.2: Effect of Wing Stiffeners
(a) Based on Model SG1-7d
Model t ws bws M pf a M pb b V FEM V prediction
c
No. (in.) (in.) (kip-in.) (kip-in.) (kips) (kips)
SG1R-W1 0.5 964 0.92 394 384
SG1R-W2 0.75 5.5 147 1,307 1.00 400 401
SG1R-W3 1.0 1,640 1.08 401 401
(b) Based on Model SG2-92e

Model t ws bws M pf M pb
 V FEM V prediction
No. (in.) (in.) (kip-in.) (kip-in.) (kips) (kips)
SG2R-W1 0.5 929 0.86 265 247
SG2R-W2 0.75 1,268 0.95 267 265

5.5 95
SG2R-W3 0.9 1,465 1.00 268 274
SG2R-W4 1.0 1,594 1.03 267 274

a
from Eq. (6.39)
b
from Eq. (6.40)
c
from Eq. (6.36)
d
see Table 6.4
e
see Table 6.5
186
Table 7.3: Width-Depth Ratios of Panels (Approach 1)
Width-Depth Ratio
before
Model No. after Rehabilitation
Rehabilitation
Subpanel 1 Subpanel 2
End Panel
(New End) (New Interior)
SG1-RS1 0.5 0.3
SG1-RS2 1.0 0.8
SG1-RS3 1.5 1.3

0.2
SG2-RS1 0.5 0.3
SG2-RS2 1.0 0.8
SG2-RS3 1.5 1.3
Table 7.4: Comparison of Shear Strength before and after Rehabilitation (Approach 1)
Ultimate Shear Strength (kips)

Plastic Shear V AR
Model No. Strength (kips), before after
V BR Rehabilitation, Rehabilitation, V BR
V BR V AR
SG1-RS1 324.3 416.3 1.28
SG1-RS2 459.0 238.2 289.5 1.22
SG1-RS3 207.8 237.5 1.14
SG2-RS1 201.6 282.5 1.40
SG2-RS2 327.7 149.1 192.7 1.29
SG2-RS3 128.3 152.4 1.19
187
Table 7.5: Comparison of Subpanel 2 Shear Strength (Approach 1)
V FEM V AASHTO V FEM

Model No.
(kips) (kips) V AASHTO
SG1-RS1 416.3 452.2 0.92
SG1-RS2 289.5 344.1 0.84
SG1-RS3 237.5 272.9 0.87
SG2-RS1 282.5 306.6 0.92
SG2-RS2 192.7 236.8 0.81
SG2-RS3 152.4 186.6 0.82
Table 7.6: Effect of Stiffener Thickness on Shear Strengths (Approach 1)

t 1.0
Stiffener Thickness Stiffener Thickness V AR
Model No.
= 0.5 in. = 1.0 in. t  0.5
V AR
t  0.5 t 1.0
V AR V AR
SG1-RS1 416.3 425.3 1.02
SG1-RS2 289.5 296.1 1.02
SG1-RS3 237.5 241.9 1.02
SG2-RS1 282.5 286.4 1.01
SG2-RS2 192.7 200.7 1.04
SG2-RS3 152.4 157.5 1.03
188
Table 7.7: Width-Depth Ratios of Panels (Approach 2)
Width-Depth Ratio
before
Model No. after Rehabilitation
Rehabilitation
End Panel
(New End) (New Interior)
SG1-RSM1 0.5 0.25 0.25
SG1-RSM2 0.6 0.3 0.3
SG1-RSM3 0.7 0.35 0.35
SG1-RSM4 0.8 0.4 0.4
SG1-RSM5 0.9 0.45 0.45
SG1-RSM6 1.0 0.5 0.5
SG1-RSM7 1.5 0.75 0.75
SG2-RSM1 0.5 0.25 0.25
SG2-RSM2 0.6 0.3 0.3
SG2-RSM3 0.7 0.35 0.35
SG2-RSM4 0.8 0.4 0.4
SG2-RSM5 0.9 0.45 0.45
SG2-RSM6 1.0 0.5 0.5
SG2-RSM7 1.5 0.75 0.75
189
Table 7.8: Comparison of Shear Strength (Approach 2)

Plastic Shear V AR
Model No. Strength (kips), before after
V BR Rehabilitation, Rehabilitation, V BR
V BR V AR
SG1-RSM1 324.3 433.7 1.34
SG1-RSM2 297.5 412.0 1.38
SG1-RSM3 277.6 388.3 1.4
SG1-RSM4 459.0 260.5 364.9 1.4
SG1-RSM5 247.1 336.9 1.36
SG1-RSM6 238.2 323.2 1.36
SG1-RSM7 207.8 270.0 1.30
SG2-RSM1 201.6 291.8 1.45
SG2-RSM2 186.1 267.4 1.44
SG2-RSM3 175.8 251.7 1.43
SG2-RSM4 327.7 164.8 228.9 1.39
SG2-RSM5 156.1 221.7 1.42
SG2-RSM6 149.1 202.2 1.36
SG2-RSM7 128.3 169.8 1.32
190
Table 7.9: Comparison of Shear Strength Predicted by FEM and Eq. (6.33)
before Rehabilitation after Rehabilitation

Model No.
do V FEM V pred VFEM do V FEM V pred VFEM
D (kips) (kips) V pred D (kips) (kips) V pred
0.2
SG1-RSM1 0.5 324.3 333.8 0.97 433.7 459.0 0.94
5
SG1-RSM2 0.6 297.5 300.2 0.99 0.3 412.0 459.0 0.90
0.3
SG1-RSM3 0.7 277.6 278.8 1.00 388.3 441.9 0.88
5
SG1-RSM4 0.8 260.5 262.8 0.99 0.4 364.9 396.0 0.92
0.4
SG1-RSM5 0.9 247.1 249.5 0.99 336.9 359.9 0.94
5
SG1-RSM6 1.0 238.2 237.8 1.00 0.5 323.2 333.8 0.97
0.7
SG1-RSM7 1.5 207.8 192.6 1.08 270.0 270.4 1.00
5
0.2
SG2-RSM1 0.5 201.6 195.5 1.03 291.8 319.0 0.91
5
SG2-RSM2 0.6 186.1 179.9 1.03 0.3 267.4 278.9 0.96
0.3
SG2-RSM3 0.7 175.8 168.6 1.04 251.7 244.3 1.03
5
SG2-RSM4 0.8 164.8 159.2 1.04 0.4 228.9 221.8 1.03
0.4
SG2-RSM5 0.9 156.1 150.9 1.03 221.7 206.6 1.07
5
SG2-RSM6 1.0 149.1 143.4 1.04 0.5 202.2 195.5 1.03
0.7
SG2-RSM7 1.5 128.3 114.2 1.12 169.8 163.7 1.04
5
Average 1.03 0.97
Standard
0.04 0.06
Deviation
191
Bearing
Stiffener
A A
End
bes Stiffener bb
Bearing
tb
Stiffener
pair tes
End
Stiffener
Pair
Section A - A
Elevation
Figure 7.1: Configuration of End Stiffeners
45°
Bearing
Stiffener
Pair = + (  cr )t w
D
End
Stiffener
Pair
e
(  cr )t w D
(   cr )t w D
Figure 7.2: Basler’s Model for End Stiffeners (adapted from Basler 1961a)
192
P, 
(a) Model with End Stiffeners
A P, 
Wing
Stiffener
Section A-A
(b) Model with Wing Stiffeners
Figure 7.3: Rehabilitated Model SG1: Finite Element Models
193
(a)
SG1R-EP1 Ae = 8.26 in.2
(b)
Ae = 11.06 in.2
SG1R-EP2
(c)
SG1R-EP3 Ae = 14.75 in.2
(d)
SG1R-EP5 Ae = 29.50 in.2
(e)
SG1R-EP6 Ae = 36.88 in.2
Figure 7.4: Basler’s Scheme: Failure Mode as a Function of End Stiffener Area
194
(a)
SG1R-EP8 Ae = 11.06 in.2
(b)
SG1R-EP9 Ae = 14.75 in.2
(c)
SG1R-EP10 Ae = 22.13 in.2
(d)
Ae = 29.50 in.2
SG1R-EP11
Figure 7.5: Eurocode 3 Scheme: Failure Mode as a Function of End Stiffener Area
195
A A
Wing
Stiffener
(a) Elevation
Bearing
Stiffener
Wing
Stiffener
Web
tws
de
EQ
tw
EQ bws
(b) Section A-A (Effective Section)

Figure 7.6: Wing Stiffener Rehabilitation Scheme for End Panel
196
1000 500

800 400
Applied Load (kips)
600 300
400 200
200 100
0 0
0.0 0.5 1.0 1.5 2.0 2.5
Figure 7.7: Model SG1R-W1: Applied Load versus Mid-Span Deflection
(a)
SG1R-W1 tew = 0.5 in.
(b)
(c)
Figure 7.8: Wing Stiffener Rehabilitation Scheme: Failure Modes of SG1 Series
197
(a)
SG2R-W1
tew = 0.5 in.
(b)
SG2R-W2
tew = 0.75 in.
(c)
SG2R-W3
tew = 0.9 in.
(d)
Figure 7.9: Wing Stiffener Rehabilitation Scheme: Failure Modes of SG2 Series
198
PNA
dPNA
tb
bb = 12.94”
tw tb = 1.0”
tws e = 4.0”
bb
tw = 0.188”
de
de = 1.96”
bws = 5.5”
bws
Figure 7.10: Example Calculation of Mpb
PNA
Seg. 5
Seg. 1 Seg. 4
Seg. 2 Seg. 3
Figure 7.11: Components of Mpb
199
A
Intermediate Existing
Stiffener Plate Girder
Bearing
Stiffener
Bolt Hole
Angle
(L-Shape)
Angle
(L-Shape)
A
Elevation Section A - A
Figure 7.12: Rehabilitation Scheme with Bolted Angles
10.5”
Added Stiffeners
(New End Panel) (New Interior Panel)
Note: The shaded area represents original end panel.
Figure 7.13: Subpanels 1 and 2
200
SG1-RS1
Added
Stiffener,
TYP.
SG1-RS2
SG1-RS3
(a) Failure Mode
1000 500
SG1-RS1

800 400
Applied Load (kips)
SG1-RS2
600 300
400 200
SG1-RS3
200 100
0 0
0.0 0.5 1.0 1.5 2.0
(b) Applied Load versus Mid-Span Deflection
Figure 7.14: SG1 Series (Approach 1)
201
SG2-RS1
Added
Stiffener,
TYP.
SG2-RS2
SG2-RS3
(a) Failure Mode
600 300

SG2-RS1 250
SG2-RS2
Applied Load (kips)
400 200
150
200 SG2-RS3 100
50
0 0
0.0 0.5 1.0 1.5 2.0
(b) Applied Load versus Mid-Span Deflection

Figure 7.15: SG2 Series (Approach 1)
202
SG1-RSM1
SG1-RSM2
Added
Intermediate
Stiffener, TYP.
SG1-RSM3
SG1-RSM4
SG1-RSM5
SG1-RSM1
SG1-RSM1
(a) SG1 Series (after Rehabilitation)
Figure 7.16: Failure Mode (Approach 2)
203
SG2-RSM1
SG2-RSM2
Added
Intermediate
Stiffener, TYP.
SG2-RSM3
SG2-RSM4
SG2-RSM5
SG2-RSM6
SG2-RSM7
(b) SG2 Series (after Rehabilitation)
Figure 7.17: Failure Mode (Approach 2) (continued)
204
1000 500
SG1-RSM1

Applied Load (kips) 800 400
600 300
400 200
SG1-RSM7
200 100
0 0
0.0 0.2 0.4 0.6 0.8 1.0
(a) SG1 Series
600 300
SG2-RSM1

250
Applied Load (kips)
400 200
150
SG2-RSM7
200 100
50
0 0
0.0 0.2 0.4 0.6 0.8 1.0
(b) SG2 Series

Figure 7.18: Applied Load versus Mid-Span Deflection (Approach 2)
205
1.4
1.2
1.0
VFEM 0.8
Avg.: 1.02
V pred
Stdev.: 0.04
0.6
Avg.: 0.96
0.4 Stdev.: 0.06
SG1 Series
SG2 Series
0.2 d o / D = 0.5
0.0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
Width-Depth Ratio, d o / D
Figure 7.19: Effect of Width-Depth Ratio on Shear Strength of End Panels
206
8. TEST PROGRAM (PHASE 2)
8.1 Test Program
8.1.1 General
Continuing the Phase 1 tests, another four test specimens were fabricated and
tested. The specimens were designated as Specimens SG3, SG4, CG3, and CG4.
Specimens CG3 and CG4 were nominally identical to Specimens SG3 and SG4, except
that the former incorporated concrete slabs.
8.1.2 Configuration and Fabrication of Test Specimensl

Table 8.1 shows the dimensions of three sample steel bridge girders provided by
the California Department of Transportation (Caltrans) for the Phase 2 test program.
Table 8.2 shows the specimen dimensions used in testing to represent the sample bridge
girders. Depth-to-thickness ratio of 138 was used for all specimens, while the panel
width-to-depth ratios of 1.5 and 0.75 were used for Specimen SG3 and SG4, respectively.
The actual plate thicknesses of the specimens were measured and are summarized in
Table 8.3.
The girders of Specimens CG3 and CG4 were nominally identical to those of
Specimens SG3 and SG4, respectively, with the exception that the composite girder
specimens included the headed shear studs welded to the top flanges of steel girder
specimens to incorporate a concrete slab. The total slab thickness (tc) and haunch depth
(dh) are shown in Table 8.2. Figure 8.2 and Figure 8.3 show the size of the shear studs,
the geometry of concrete slab, and the steel reinforcement layout for these two composite
specimens.
8.1.3 Material Properties

ASTM A36 steel was specified for all steel plates including the flanges, web, and
intermediate and bearing stiffeners of the specimens. Tensile coupons were cut from the
same plates for material testing. Average values of the measured mechanical properties
of the component plates are summarized in Table 8.4.
207
Grade 40 longitudinal reinforcement (No. 4 and No. 5 for Specimens CG3 and
CG4) with a minimum yield strength of 40 ksi and a minimum ultimate strength of 70 ksi
was used. Due to the availiabity issues, Grade 60 transverse reinforcement was used for
No. 6 rebars. Three samples of each size of reinforcement were tested. The average
values of the measured mechanical properties are also summarized in Table 8.5.
A design concrete compressive strength of 4,000 psi was specified for the
concrete slab. A total of 21 cylinders were cast and cured under the same conditions as
the test specimens to obtain the compressive and split tensile strengths of the concrete.
Each three-cylinder set was tested for compressive strength on the 7th, 14th, 21st, 28th, and
32nd day (day of testing for CG4), while another three-cylinder set was tested for splitting
tensile strength on the day of testing for each composite specimen. Although Specimen
CG3 was tested on the 25th day of the concrete cast, the concrete cylinder was tested on
the 28th day due to the laboratory schedule. However, it is believed that the difference
should be minor. Table 8.6 summarizes the average values of compressive and splitting
tensile strengths from testing.
Type B headed shear stud connectors made from ASTM A108 material with a
minimum yield strength of 51 ksi and a minimum ultimate strength of 65 ksi was used for
the interconnection of steel girder to concrete slab.
8.1.4 Initial Geometrical Imperfections

Initial geometrical imperfections of the steel girders were measured before testing.
The measured web minimum and maximum imperfection values of each panel are also
summarized in Table 8.7.
8.1.5 Test Setup, Loading Procedure, and Instrumentation

The configuration of steel girder specimens (Specimens SG3 and SG4) is shown
in Figure 8.1 for Specimens SG3 and SG4. Specimen SG3 has a width-depth ratio of
1.5, which is the upper limit per AASHTO Specifications (2014). Speimen SG4 is the
same as Specimen SG3 except that additional intermediate stiffeners were incorporated to
make the width-depth ratio be half of Specimen SG3. Figure 8.2 and Figure 8.3 show
the overall configuration of Specimens CG3 and CG4. The figures also provide the size
of the shear studs, the geometry of the concrete slab, and the steel reinforcement layout.
208
Since Phase 2 test specimens had the same overall specimen length, the same test
setup scheme of Phase 1 test program was used. That is, each specimen was simply
supported and monotonically loaded to the top flange through a W36302 loading beam
at mid-span of the specimen. See Chapter 2 for the details of the test setup.
Test specimens were instrumented with a combination of displacement
transducers, strain gage rosettes, and uni-axial strain gages to measure global and local
responses at specific locations. One displacement transducer measured vertical
displacement at the mid-span of each specimen and was used as the feedback sensor for
actuator control. Strain gage rosettes measured the principal strains on the web panel,
while uni-axial strain gages monitored the strains of flanges and bearing stiffeners.
Load cells mounted on the actuators measured applied load. The instrumentation plan
for each specimen is provided in Appendix A.
8.2 Test Results
8.2.1 Specimen SG3

Figure 8.4 shows the test setup prior to testing. The relationship between the
applied load and mid-span deflection is shown in Figure 8.5. The theoretical stiffness
computed from Eq. (2.1) is also shown in the figure.
Figure 8.6 shows the test specimen at event A, which corresponds to the
AASHTO nominal shear strength for end panels. Minor web buckling was observed at
this event. When the maximum applied load was achieved (event B), web buckling
became significant (see Figure 8.7). After event B, the strength degraded slowly and
plastic hinges formed in the bearing stiffeners and both flanges as observed in Phase 1
test specimens. The deformed shape at event C is shown in Figure 8.8. The front and
back sides of the east end panel after the specimen was unloaded and intrumetation
removed are shown in Figure 8.9. The observed plastic hinge locations are also marked
in the figure.
8.2.2 Specimen SG4

A plot of the applied load versus mid-span deflection is shown in Figure 8.10.
The theoretical stiffness is well matched with the inistiall stiffeness from the test.
209
Figure 8.11 and Figure 8.12 show the deformed shape of the specime at events A
and B, respectively. Minor web buckling was first observed at event A and became
significant at event B. A close-up view of both end panels at event B in Figure 8.12
shows that the east end panel has more significant web local buckling. The load
dropped with the significant web buckling of the east end panel, but the load started
increasing again with the significant buckling of the west end panel (see Figure 8.13).
Then, the applied load dropped again before the specimen resisted more load and
achieved the maximum load. After achieving the maximum load at event D, the load
slowly degraded with the deformation in the flanges and bearing stiffeners. Figure 8.14
shows the deformed shape at event D. It is obvious from the figure that the boundary
members also experienced a large deformation. Eventually, the east end panel failed
with plastic hinges formed in the flanges and bearing stiffeners. The deformed shape at
event E is shown in Figure 8.15.
The failure modes included plastic hinge formation on the boundary members and
web buckling on the front and back sides of the east end panel as shown in Figure 8.16.
Figure 8.17 shows the failure modes of both Specimens SG3 and SG4. The failure
modes are very similar to those of Phase 1 Specimens (SG1 and SG2) as shown in Figure
3.15.
8.2.3 Specimen CG3

Specimen CG3 was nominally identical to Specimen SG3, except that it
and mid-span deflection.
Specimen CG3 showed similar behavior to Specimen SG3. Figure 8.19 shows
the deformed shape at event A when minor web buckling developed in both end panels.
Then minor yielding in the end panels was observed at a mid-span deflection of 0.2 in.
Figure 8.20 shows the web buckling pattern at a mid-span defection of 0.3 in. Damage
in the concrete slab was not visible and the specimen was able to resist more loads.
Eventually, the specimen reached the maximum applied load (= 697 kips) at a mid-span
deflection of 1.14 in. Figure 8.21 shows the buckling pattern at event B. The shear
crack on the concrete slab caused a loss of load resistance. As the mid-span deflection
210
increased, the cracks widened. The failure mode of the specimen at event C (Δ = 2.5
in.) is shown in Figure 8.22. The crack patterns are also summarized in Figure 8.23.
At the completion of testing, concrete crushing at the top surface of the concrete slab is
shown in Figure 8.24(a). Figure 8.24 (b) shows the yielding pattern as a result of
tension-field formation along the panel diagonal in the web and plastic hinges developed
in the flanges and bearing stiffeners.
8.2.4 Specimen CG4

Specimen CG4 was nominally identical to Specimen SG4, except that CG4
and mid-span deflection. Since actuators reached its capacity during the test, it was
decided to add steel mounting plates (11.64 kips) on the top of the concrete slab near the
mid-span of the specimen to fail the specimen. Considering this additional weight, the
expected global response is also plotted in the figure.
At event A, web buckling was very minor in both end panels. The measured
web out-of-plane deformation was about 0.063 in. in both end panels. Figure 8.26
shows the deformed shape at event A. The amplitude of web buckling grew rapidly and
the measured out-of-plane deformation was 0.375 in. in both end panels at a mid-span
deflection of 0.2 in. After this point, the stiffness softened, web buckling became
significant, and yielding due to the tension-field formation occurred in both end panels.
Also note that two interior web panels experienced buckling at a mid-span deflection of
0.35 in. Figure 8.27 shows the web local buckling and the associated yielding pattern at
a mid-span deflection of 0.4 in. At this event, flexural concrete cracks were also
observed. The number of cracks increased as the mid-span deflection increased. The
buckling pattern at the maximum load (= 890 kips) is shown in Figure 8.28. Figure 8.29
shows the concrete crack patterns observed during the test. Figure 8.30 shows the
failure mode of the specimen at event C. Plastic hinges that developed in flanges and
bearing stiffeners are also marked in the figure.
8.2.5 Recorded Local Responses

Figure 8.31 shows the locations of the displacement transducers that measured the
web out-of-plane deformations. L2, L3, and L4 were installed on the back side of the
211
east end panel, while L5, L6, L7 were placed on the back side of the west end panel.
All displacement transducers were located at the mid-width of the panel. L2 and L5
were located at the distance of 11 in. away from the top flange, while L4 and L7 were
placed at the same distance from the bottom flange. L3 and L6 were installed near the
mid-height of the panels. Figure 8.32 to Figure 8.35 show the web out-of-plane
deformation of each test specimen. The expected buckling load predicted per the
AASHTO Specifications (2014) is also presented in the figure.
Figure 8.36 shows the locations and orientations of the strain gage rosettes for all
specimens. All strain gage rosettes were installed on the back side of the panel web.
From the rosette data, principal strains and directions were computed. Figure 8.37 plots
the principal tensile strains across the panel diagonal of both east and west end panels at
the ultimate load (Event B). The horizontal reference line in the figure represents the
strain value corresponding to the tension-field stress (  t ) proposed by Basler (1961a).
The principal tensile strains recorded at five middle strain gage rosettes in each
panel are beyond the strain corresponding to the tension-field stress. Table 8.8 and
Table 8.9 summarize the principal tensile directions of five middle strain gage rosettes of
each section at ultimate load (Event B). The recorded principal angles varied but most
of them were within ± 10° of the angle of the panel diagonal (34° for Specimens SG3 and
CG3, and 53° for Specimens SG4 and CG4) with respect to the flange.
8.2.6 Comparison of Test Results

Table 8.10 summarizes the end panel shear strengths based on both AASHTO
Specifications (2014) and experimental results. The shear overstrengths for each
specimen, computed as the ratio of the experimental strength to the AASHTO nominal
strength, are listed in the last column of the table. As observed from Phase 1 test results,
the AASHTO Specifications (2014) are very conservative in calculating the shear
strength of the end panels. Figure 8.38 summarizes the measured plastic hinge locations
in the top flange and bearing stiffeners for each test specimens after completing the tests.
Figure 8.39 shows the relationship between the normalized ultimate shear strength
and the web depth-to-thickness ratio. The curves were constructed based on the actual
web yield stress from the coupon tests. It is found that the actual shear strengths of
212
Phase 2 test specimens were slightly higher than theAASHTO-predicted strengths of the
interior panel, which consider full tension-field action. It should be noted that the
boundary members (flanges and bearing stiffeners) in these specimens provided high
flexural strength such that the coefficient, α, in the proposed Eq. (6.33) is greater than 1.0.
Therefore, the shear strengths of Specimens SG3 and SG4 are very close to the shear
strengths of the panels considering full tension-field action.
In Phase 1 tests, the increase in shear strength due to the contribution from
concrete slab was about 12% for composite Specimens CG1 and CG2. In Phase 2 tests,
however, the increase in shear strength due to the contribution from concrete slab was
significantly higher and was about 61% in Specimen CG3 and 63% in CG4. This is
expected as, for a given concrete slab, the composite action effect is more significant on
shallower than deeper steel girders.
213
Table 8.1: Sample Caltrans Steel Bridge Girder Dimensions
Width- Depth-
Section Depth Thickness
Bridge No. Ratio, Ratio,
Top Flange Bottom Flange
Web (in.) d0/D D/tw
(in.) (in.)
42-0246R 10×5/8 42×5/16 18×1 0.86 134
50-0316
12×1 45×5/16 20×1-1/8 0.71 144
(Spans 2&3)
50-0316
12×5/8 45×5/16 20×7/8 0.80 144
(Spans 1&4)
Table 8.2: Specimen Dimensions

Top and Bottom Total Slab
Width- Depth-
Web Flanges Thickness
Spec. Depth Thickness Scale
Depth, (tc)/ Haunch
No. Ratio, Ratio, Thickness, Width, Factor
Depth (dh)
d0/D D/tw D (in.) tf (in.) bf (in.) (in.)
SG3 No Slab
1.5 138 43 1.0 18.0 1
CG3 8.25 / 2.75
SG4 No Slab
0.75 138 43 1.0 18.0 1
CG4 8.25 / 2.75
Table 8.3: Steel Plate Thicknesses

Specimen Nominal Measured
Component
No. Thickness (in.) Thickness (in.)
Flanges 1.00 1.006
SG3 / CG3
SG4 / CG4
Web 0.3125 0.305
Bearing Stiffener 1.00 1.006

SG3 / CG3
SG4 / CG4
Intermediate Stiffener 0.5 0.486
214
Table 8.4: Average Mechanical Properties of Steel Plates
Tensile
Specimen Steel Component / Yield Stress Elongation
Strength
No. Grade Heat Number (ksi) (%)
(ksi)
Flanges /
45.6 (47.7)a 71.5 (63.5) 27.2b (28)c
7N615
Web /
36.9 (45.1) 60.7 (64.5) 38.1b (24)c
919696
SG3, CG3,
A36 Bearing
SG4, CG4
Stiffener / 45.6 (47.7) 71.5 (63.5) 27.2b (28)c
7N615r
Intermediate
Stiffener / 40.1 (46.6) 64.2 (65.0) 30.4b (24)c
918573
Note:
a. Values in parentheses are based on Certified Mill Test Reports, others from tensile
coupon tests at UCSD.
b. Elongation is based on 2 in. gage length.
c. Elongation is based on 8 in. gage length.
Table 8.5: Average Mechanical Properties of Steel Reinforcements

Yield Stress Tensile Strength
Bar Size
(ksi) (ksi)
No. 4 45.9 67.4
No. 5 47.2 71.6
No. 6 64.6 93.3
215
Table 8.6: Average Mechanical Properties of Concrete
Curing Period Compressive Strength, Splitting Tensile
(Days) f’c (ksi) Strength, ft (ksi)
7 3.45 NA
14 4.19 NA
21 4.48 NA
28 4.80 0.47
32 5.07 0.43
Table 8.7: Measured Initial Web Imperfections (in.)
Specimen No.
Measured Initial
Imperfections
SG3 SG4 CG3 CG4
Min. -0.0625 -0.0372 0.0000 0.0000

Panel 1
Max. 0.1875 0.2123 0.2377 0.2311
Min. NA 0.0000 NA -0.0495

Panel 2
Max. NA 0.2188 NA 0.2064
Min. NA -0.0551 NA -0.0625

Panel 3
Max. NA 0.1624 NA 0.1624
Min. 0.0000 0.0000 0.0000 -0.1123

Panel 4
Max. 0.3125 0.2500 0.3125 0.2874
216
Table 8.8: Prinicipal Strain Direction at Ultimate Load (Steel Plate Girders)

R05 40
East R04 37
Panel
R03 56
 R02 67
R01 56
SG3 34
R06 37
West R07 45
Panel R08 50
 R09 52
R10 49
R05 37
East R04 48
Panel
R03 NA
 R02 53
R01 57
SG4 53
R08 48
West R09 20
Panel R10 NA
 R11 50
R12 54
217
Table 8.9: Prinicipal Strain Direction at Ultimate Load (Composite Plate Girders)

R05 38
East R04 40
Panel
R03 22
 R02 36
R01 32
CG3 34
R06 33
West R07 34
Panel R08 16
 R09 31
R10 31
R05 48
East R04 19
Panel
R03 49
 R02 48
R01 52
CG4 53
R08 51
West R09 44
Panel R10 47
 R11 51
R12 53
218
Table 8.10: Comparison of End Panel Shear Strengths
AASHTO Predicted Experimental

Specimen Shear
Nominal Strength Shear Strength
No. Overstrength
(kips) (kips)
SG3 212.1 1.57
135.4
CG3 348.6 2.57
SG4 279.0 1.15

241.6
CG4 445.4 1.84
219
10"
A A
841"X1"
3'-9" Typ.
5
43"X16 " Web
1"
5'-42 8" 8" 1"
5'-42
4" 4"
12'-9"
1'-6"
View A - A
(a) Specimen SG3

10"
B B
841"X1" 434"X21"
3'-9" Typ. Typ.
5
43"X16 " Web
1"
2'-84 1"
2'-84 8" 8" 1"
2'-84 1"
2'-84
4" 4"
12'-9"
1'-6"
View B - B
(b) Specimen SG4

Figure 8.1: Configuration of Steel Girder Test Specimens
220
7
1"
62 14 Nos. @ 5" 5" 5" 14 Nos. @ 5" 1"
62
10"
434"X1"
841"X1" Typ.
3'-9" Typ. 5
43"X16 " Web
18"X1" Flange Typ.
1"
5'-42 8" 8" 1"
5'-42
4" 4"
12'-9"
No.6 @ 11" 1'-10" 1'-10" No.4, Typ.

1" Clr.
12
1" 1" Clr.
14
84
23
4"
1"
1'-104 No.5, Typ.
23
4"
2"
6'
221
7
1"
62 14 Nos. @ 5" 5" 5" 14 Nos. @ 5" 1"
62
10"
434"X1"
841"X1" 434"X21" Typ.
3'-9" Typ. Typ. 5
43"X16 " Web
18"X1" Flange Typ.
1"
5'-42 8" 8" 1"
5'-42
4" 4"
12'-9"
No.6 @ 11" 1'-10" 1'-10" No.4, Typ.

1" Clr.
12
1" 1" Clr.
14
84
23
4"
1"
1'-104 No.5, Typ.
23
4"
2"
6'
222
E
(a) Front View
(b) Side View

Figure 8.4: Specimen SG3 Test Setup
223
500 250
Predicted Stiffness

B
400 200
C
Applied Load (kips)
300 150
A
200 100
AASHTO
Nominal Strength
100 50
0 0
0.0 0.5 1.0 1.5 2.0
Figure 8.5: Specimen SG3 Applied Load versus Mid-Span Deflection
Figure 8.6: Specimen SG3: Web Local Buckling at Event A
224
E
(a) Overall View
(b) West End Panel
(c) East End Panel

225
E
Plastic
Hinge

Figure 8.9: Specimen SG3: Failure Mode after Completion of Test (East End Panel)
226
600 Predicted Stiffness 300
B E

D
500 C 250
A
Applied Load (kips)
400 200
AASHTO
Nominal Strength
300 150
200 100
100 50
0 0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
Figure 8.11: Specimen SG4: Web Local Buckling at Event A
227
E
(a) Overall View
(b) West End Panel
(c) East End Panel

228
E
(a) Overall View

Figure 8.13: Specimen SG4: Deformed Shape at Event C
Figure 8.14: Specimen SG4: Deformed Shape at Event D
229
E
Figure 8.15: Specimen SG4: Failure Mode at Event E
Plastic
Hinge

Figure 8.16: Specimen SG4: Failure Mode after Completion of Test (East End Panel)
230
SG4
SG3
Figure 8.17: Failure Modes of SG1 and SG2
800 400
Predicted Stiffness

700
B
600 300
Applied Load (kips)
C
500
400 200
300
A
200 100
100 AASHTO
Nominal Strength
0 0
0.0 0.5 1.0 1.5 2.0 2.5 3.0
231
E
Figure 8.19: Specimen CG3: Overall View at Event A
Figure 8.20: Specimen CG3: Web Yielding pattern at Δ = 0.3 in.
232
E
Figure 8.21: Specimen CG3: Deformed Shape at Event B (Maximum Load)
233
(a) at Δ = 0.6 in. (Event B) (b) at Δ = 1.2 in.
(c) at Δ = 1.5 in. (d) at Δ = 2.0 in.
(e) at Δ=2.5 in. (f) End View A - A

Figure 8.23: Specimen CG3: Propagation Concrete Slab Cracks
234
(a) Concrete Crushing
Plastic
Hinge
(b) Front Side of West End Panel

Figure 8.24: Specimen CG3: Failure Mode after Completion of Test
235
1000 500
Predicted Stiffness B

Applied Load (kips) 800 400
C
600 With Added Weight 300
A
400 200
AASHTO
200 Nominal Strength 100
0 0
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Figure 8.26: Specimen CG4: Overall View at Event A
236
E
(a) Overall View

Figure 8.27: Specimen CG4: Web Local Buckling and Yielding Patterns at Δ = 0.4 in
237
E
Figure 8.28: Specimen CG4: Yielding Pattern at Event B
238
(a) at Δ = 0.4 in. (b) at Δ = 0.6 in.
(c) at Δ = 1.0 in. (d) at Δ = 1.75 in.
(e) at Δ = 2.5 in. (f) End View A - A

Figure 8.29: Specimen CG4: Concrete Slab Crack Propagation
239
E
(a) Overall View
Plastic
Hinge
(b) Front Side of East End Panel

240
L2 L5
L3 L6
L4 L7
(a) East End Panel (b) West End Panel
CL West CL
11" 11"
L02 L05
L03 2" L06 2"

CL
L04 L07
11" 11"
(c) Locations of Displacement Transducers
Figure 8.31: Specimen SG3: Locations of Displacement Transducers (Back Side)
241
500
400
Applied Load (kips)
300
200
AASHTO
Nominal
100 L2
Strength L3
L4
0
-3 -2 -1 0 1 2 3
(a) East End Panel
500
400
Applied Load (kips)
300
200
AASHTO
Nominal
100 Strength L5
L6
L7
0
-0.4 -0.2 0.0 0.2 0.4
(b) West End Panel
242
600
500
Applied Load (kips)
400
AASHTO
300 Nominal
Strength
200
L2
100 L3
L4
0
-3 -2 -1 0 1 2 3
(a) East End Panel
600
500
Applied Load (kips)
400
AASHTO
Nominal
300
Strength
200
100 L5
L6
L7
0
-3 -2 -1 0 1 2 3
(b) West End Panel
243
800
700

AASHTO
500
Nominal
400 Strength
300
200
L2
L3
100 L4
0
-3 -2 -1 0 1 2 3
(a) East End Panel
800
700
600
Applied Load (kips)
500 AASHTO
Nominal
400 Strength
300
200
L5
100 L6
L7
0
-3 -2 -1 0 1 2 3
(b) West End Panel
Figure 8.34: Specimen CG3 Web Out-of-Plane Deformation
244
1000
800
Applied Load (kips)
600
400
AASHTO
200 Nominal L2
L3
Strength L4
0
-3 -2 -1 0 1 2 3
(a) East End Panel
1000
800
AASHTO
Applied Load (kips)
Nominal
600 Strength
400
200 L5
L6
L7
0
-2 -1 0 1 2
(b) West End Panel
Figure 8.35: Specimen CG4 Web Out-of-Plane Deformation
245
West
R05 R06
R04 R07
R02 R03 5 Rosettes @ 3" R08
b R09 y
R10
R01
r r
y b
Rosette Rosette
Orientation 90° Orientation
West
5 Rosettes @ 3"
R08
R04
R05 R10
R02 R09
R03 90° R12
R11
R01
b y
r Rosette Rosette r
y Orientation Orientation b
Figure 8.36: Locations of Strain Rosettes in End Panels (Back Side)
246
0.0020
East End Panel.
West End Panel
0.0015
0.0010
0.0005
0.0
R07 R04
R06 R05
R08 R03
R09 R02
R10 R01
East Panel:
West Panel:
Rosette Location
(a) Specimen SG3
0.03
R03 and R10 East End Panel.
Malfunctioned West End Panel
0.02
0.01
Bearing
Intermediate Stiffener Side
Stiffener Side
0.0
R09 R04
R08 R05
R10 R03
R11 R02
R12 R01
East Panel:
West Panel:
Rosette Location
(b) Specimen SG4
Figure 8.37: Principal Strains across Panel Diagonals at Maximum Load
247
0.020
East End Panel.
West End Panel
0.015
0.010
0.005
0.0
R07 R04
R06 R05
R08 R03
R09 R02
R10 R01
East Panel:
West Panel:
Rosette Location
(c) Specimen CG3
0.020
East End Panel.
West End Panel
0.015
Bearing
Stiffener Side
0.010
Intermediate
Stiffener Side
0.005
0.0
R09 R04
R08 R05
R10 R03
R11 R02
R12 R01
East Panel:
West Panel:
Rosette Location
(d) Specimen CG4
Figure 3.34: Principal Strains across Panel Diagonals at Maximum Load (continued)
248
a
Specimen
a (in.) b (in.)
No.
SG3 20.13 15.25
CG3 17.5 15.0
SG4 15.5 18.5
CG4 14.25 17.5
Figure 8.38: Measured Plastic Hinge Locations
249
1.6
(Vn/Vp)
1.4
Strength
1.2 1 CG3
Strength
Fyw Dtw
3
1.0
Shear
AASHTO Shear
Shear
0.8
SG3 Strength with TFA
Normalized
0.6
Normalized
0.4
AASHTO Shear
Strength without TFA
0.2
0.0
0 100 200 300 400 500
Web Panel
Web Depth-Thickness
Panel Ratio
Depth-Thickness (D/tw)
Ratio
1.6
CG4
(Vn/Vp)
1.4
Strength
1.2 1
Strength
Fyw Dtw
3
1.0 SG4 AASHTO Shear
Shear
Strength with TFA

0.8
Shear
Normalized
0.6
Normalized
AASHTO Shear
0.4 Strength without TFA
0.2
0.0
0 100 200 300 400 500
Web Panel
Web Depth-Thickness
Panel Ratio
Depth-Thickness (D/tw)
Ratio

250
9. ANALYTICAL STUDY OF TEST SPECIMENS (PHASE 2)
9.1 Nonlinear Finite Element Analysis
9.1.1 General
Standard 3-D shell elements (S4R) were used to model steel plate girder
specimens (Specimens SG3 and SG4), while the eight-node linear brick elements
(C3D8R) were used to model the concrete slab for Specimens CG3 and CG4. The steel
reinforcement and shear studs were modeled with two-node linear beam elements (B31).
True stress and strain were calculated from the engineering stress and strain from the
coupon test results and used to describe the nonlinear behavior of the materials (see
Chapter 4). A simply supported boundary condition was used, and the lateral bracing
was simulated by preventing the lateral movement of the bracing locations. Also, the
measured plate thicknesses shown in Table 8.3 were used in the finite element models.
Residual stresses were not considered, but initial geometrical imperfections were
incorporated in the analysis.
9.1.2 Steel Plate Girder Specimens (SG3 and SG4)

Figure 9.1 shows the typical finite element model. Figure 9.2 and Figure 9.3
show the buckling patterns of Specimens SG3 and SG4 from both testing and finite
element analysis. The corresponding load-deflection curves are compared in Figure 9.4.
As shown in the figures, the finite element analysis could reasonablely simulate the
behavior of both specimens. The buckling patterns of the test specimens and finite
element models were very similar, and the load-deflection curves between the two were
also similar. Although the maximum load was achieved at a larger mid-span deflection,
the value of the maximum load from the finite element analysis is very close to the one
from the test results. The ratios of ultimate shear strength from experimental testing to
that from finite element analysis are 1.00 for both specimens (see Table 9.1).
Figure 9.5 compares the plastic hinge locations formed in the top flange and
bearing stiffeners. Plastic hinge locations (a and b) from experimental testing and finite
251
element analysis are summarized in the corresponding table. As shown in the figure and
table, the plastic hinges formed in similar locations.
9.1.3 Composite Plate Girder Specimens (CG3 and CG4)

The correlation between the predicted and experimental responses for the steel-
concrete composite girder specimens is provided in Figure 9.6. The response from
Model CG3 is very close to that from Specimen CG3 up to the maximum load. After
reaching the maximum load, Model CG3 showed a load drop of about 1-kip and the
analysis was stopped due to the convergence issues [see Figure 9.6(a)]. Model CG4
showed a little soft behavior in the second branch (nonlinear region) and the analysis was
stopped at a deflection of about 1.0 in. due to the convergence issues. For this reason,
the buckling patterns of Specimens CG3 and CG4 from both testing and finite element
analysis were compared at were compared at this displacement level as shown in Figure
9.7 and Figure 9.8. The ratios of ultimate shear strengths obtained form the experiments
to those of finite element analyses are 0.99 and 1.08 for Specimens CG3 and CG4,
respectively (see Table 9.1).
9.2 Comparison of Test Results with Predictive Equation
The predicted shear strengths of the specimens were calculated by using the
proposed Eq. (6.33) for steel girder specimens (Specimens SG3 and SG4). Table 9.2
compares the shear strengths from tests, finite element analysis, and prediction from Eq.
(6.33). Since the calculated  is greater than 1.0 for both specimens, the coefficient
 of 1.0 is adopted for both calculations. Therefore, the calculated shear strength from
Eq. (6.33) provides the same prediction for the case of interior panels with tension-field
action per AASHTO Specifications (2014). The ratios of the shear strengths from the
predictive equation to those from the tests are 0.95 for both Specimens SG3 and SG4.
Table 9.3 summarizes the shear strengths for composite girder specimens
(Specimens CG3 and CG4). The predicted shear strengths for both specimens were still
too conservative compared to the results from both tests and finite element analyses.
The ratios of the shear strengths from the predictive equation to those from the tests are
252
0.71 for both Specimens CG3 and CG4. The conservatism is mainly from Eq. (6.48) for
predicting the contribution from concrete slab.
9.3 Modification of Predictive Equation for Concrete Shear Strength
In developing Eq. (6.48), it was assumed that only one compression strut
developed as shown in Figure 6.23. Because shear connectors were much more closely
space in Phase 2 specimens, multiple compression struts could be developed. Figure
9.9(b) and Figure 9.10(b) show that the major cracks occurred between the two observed
plastic hinges.
For CG3, Figure 9.9(c) shows four rows of shear connectors existed between
plastic hinges. The “effective” compression strut can be assumed to run from points A
and B. The strut angle is
h 
  tan 1  s  (9.1)
 Sh 
where hs is the stud height and S h is the distance between the bearing stiffeners and
the first row of shear connectors to the right of the interior plastic hinge as shown in
Figure 9.9(c). The horizontal component of the compression force in the strut is limited
by the shear strength of the shear connectors, which can be computed as nQn , with n =
number of shear connectors between the plastic hinges and Qn = nominal shear
resistance of a single shear connector. According to AISC Specificatios (2010), Qn is

determined as:
Qn  0.5 Asa f c' E c  R g R p Asa Fu (9.2)
where Asa = cross-sectional area of the shear stud, f c' = concrete compressive strength,
E c = modulus of elasticity of concrete, Fu = the specified minimum tensile strength,
and the values of R g and R p are 1.0 and 0.75, respectively, when the studs are welded
directly to the steel girder. Then, the vertical component of the effective compression
strut force is the vertical shear resisted by the concrete slab:
Vc  nQn tan  (9.3)
253
By applying the above procedure to CG4, Figure 9.10(c) shows the effective
compression strut. This procedure results in a  value of 17.3° and 21.7° for CG3 and
CG4, respectively, which correlate well with the observed crack angles in Figure 9.9(a)
and Figure 9.10(a). The calculated Vc and the total shear resistances are suumarized in
Table 9.4. The ratios of the shear strengths from the modified prediction to those from
the tests are 0.99 and 0.91 for Specimens CG3 and CG4, respectively.
Note that Eq. (6.48) was developed on the basis that compression failure of the
concrete strut is the limit state, while Eq. (9.3) for a composite slab with closely spaced
shear connectors used the shear strength of the connectors as the limit state. Further
study is needed to establish a criterion in judging which limit state would govern the
shear strength of concrete slab ( Vc ).
254
Table 9.1: Shear Strength Comparison between FEM and Experiments
Specimen Panel Shear Strength, V (kips)

VTEST / VFEM
No.
FEM Result, VFEM Test Result, VTEST
SG3 211.3 212.1 1.00
SG4 277.9 279.0 1.00
CG3 351.4 348.6 0.99
CG4 411.5 445.4 1.08
Table 9.2: Shear Strength Comparison (Steel Girder Specimens)

Specimen
α
No.. Prediction
Test FEM
[Eq. (6.33)]
SG3 1.0 212.1 211.3 200.6
SG4 1.0 279.0 277.9 266.2
Table 9.3: Shear Strength Comparison (Composite Girder Specimens)
Prediction (kips)
Specimen Test FEM
No. (kips) (kips) Vs Vc Vn (= Vs + Vc )
[Eq. (6.33)] [Eq. (6.48)] [Eq. (6.43)]
CG3 348.6 351.4 200.6 48.1 248.7
CG4 445.4 411.5 266.2 50.1 316.3
255
Table 9.4: Comparison of Shear Strength between Modified Prediction and Test Results
Prediction (kips) Prediction

Specimen Test
No. (kips) Vs Vc Vn (= Vs + Vc ) Test
[Eq. (6.33)] [Eq. (9.3)] [Eq. (6.43)]
CG3 348.6 200.6 146.0 346.6 0.99
CG4 445.4 266.2 139.9 406.1 0.91
256
Loading
Figure 9.1: Specimen SG3: Typical Finite Element Model and Mesh
(a) Experiment
(b) Finite Element Analysis
Figure 9.2 Comparison of Buckling Pattern of Specimen SG3
257
(a) Experiment
Figure 9.3 Comparison of Buckling Pattern of Specimen SG4
258
500 250

400 200
Applied Load (kips)
300 150
200 100
100 Experiment 50
FEM
0 0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
(a) Specimen SG3
600 300

500 250
Applied Load (kips)
400 200
300 150
200 100
Experiment
100 FEM 50
0 0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
(b) Specimen SG4
259
a
a
b b
Specimen a (in.) b (in.)

No. Test FEM Test FEM
SG3 20.1 19.2 15.3 15.0
SG4 15.5 14.1 18.5 17.0
Figure 9.5: Comparison of Plastic Hinge Locations
260
800 400

700
600 300
Applied Load (kips)
500
400 200
300
200 Experiment 100

FEM
100
0 0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
(a) Specimen CG3
1000 500

800 400
Applied Load (kips)
600 300
400 200
200 Experiment 100

FEM
0 0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8
(b) Specimen CG4
Figure 9.6: Correlation between Test Results and FEM Analyses (CG3 and CG4)
261
(a) Experiment
Figure 9.7 Comparison of Buckling Pattern of Specimen CG3 (at 1.0 in. Disp.)
262
(a) Experiment
Figure 9.8 Comparison of Buckling Pattern of Specimen CG4 (at 1.0 in. Disp.)
263
20°
Crack1
Crack2
(a) Crack Angle
1"
22 13@5"
Possible Compression Strut, Typ.
Crack 1
3"
Observed Plastic
12" Hinge, Typ.
Crack 2
Bearing Stiffener
Intermediate Stiffener
(b) Possible Formation of Compression Struts
1"
22 13@5"
B
hs
A
Sh Plastic Hinge, Typ.
Bearing Stiffener
Intermediate Stiffener
(c) Effective Compression Strut

Figure 9.9: Assumed Compression Strut (Specimen CG3)
264
19°
Crack1
Crack2 Crack3
(a) Crack Angle
13@5" 1"
22
Possible Compression Strut, Typ.
Crack 1
Crack 2 3"
Observed
Crack 3 10" Plastic Hinge,
12" Typ.
Intermediate Bearing Stiffener

Stiffener
(b) Possible Formation of Compression Struts
13@5" 1"
22
B
hs
A
Sh Plastic Hinge,
Typ.
Intermediate Bearing Stiffener

Stiffener
(c) Effective Compression Strut

Figure 9.10: Assumed Compression Strut (Specimen CG4)
265
10. NUMERICAL PARAMETRIC STUDY
10.1 General
Since the results of nonlinear finite element analyses correlated well with the test
results, additional numerical study with ABAQUS was carried out to further investigate
the behavior of plate girder end panels. In this chapter, the effects of longitudinal
stiffener, small flange areas, large width-depth ratios, noncompact web, and unequal top
and bottom flanges are examined.
10.2 Effect of Longitudinal Stiffener
Longitudinal stiffeners are frequently used in existing highway bridge girders.

The primary function of longitudinal stiffeners is to control lateral web deflections, and
hence the longitudinal stiffeners can serve as a simple support. As a result, two
subpanels are simply supported and each subpanel would develop its own shear strength.
By summing up the shear strengths from each subpanel, the ultimate shear strength of the
entire panel with longitudinal stiffener can be determined. Cooper (1966) also assumed
that each subpanel develops its own tension field after buckling based on the observation
from the experimental testing conducted by Fielding and Cooper (1965).
A study by Rockey and Leggett (1962) showed that the optimum location of a
longitudinal stiffener under pure bending is at about 0.2 of the web depth (0.2D) from the
compressive flange, while its optimum location under pure shear is at mid-depth (0.5D),
respectively. For combined shear and bending, the longitudinal stiffener should be
located between 0.2D and 0.5D (Salmon and Johnson, 1996). With those two
longitudinal stiffener locations, nonlinear finite element analyses with longitudinal
stiffeners added in Models SG1 and SG2 were carried out to investigate the effect of the
longitudinal stiffener.
Figure 10.1 shows the finite element models with two different placements of
longitudinal stiffeners. The longitudinal stiffeners were added to both sides of the web
and extended to the ends of the girder. The same material properties and thicknesses
from the flanges of each specimen were used for the longitudinal stiffeners. Table 10.1
266
summarizes the shear strength of each subpanel and the entire shear strength of the end
panels. For example, subpanel 1 in Model SG1LS02 has a depth-thickness ratio of 42,
which falls in the shear yield zone with a shear yield strength (Vy = 0.58AwFyw) of 91.8
kips. However, subpanel 2 falls in the elastic buckling zone where the proposed
equation [Eq. (6.33)] was used to calculate the shear strength. The calculated  value
from Eq. (6.36) was 0.862 and the shear strength from Eq. (6.33) was 297.4 kips.
Summing up the shear strength of each subpanel gives an ultimate shear strength
(Vprediction) of 362 kips. This calculation slightly overestimates the shear strengthresult
from FEM (see Figure 10.2). However, the ultimate shear strengths predicted by this
procedure for the other models (Models SG2LS02, SG1LS05, and SG2LS05) correlate
well but with a slight conservatism as shown in Figure 10.3 through Figure 10.5.
Therefore, the proposed Eq. (6.36) provides either close or somewhat conservative shear
strength when longitudinal stiffeners are used.
Note from the buckling patterns (Figure 10.2 through Figure 10.5) that subpanels
1 and 2 tend to develop the same tension-field angle which is the diagonal direction of the
larger subpanel (subpanel 2 in this case). This trend was also observed in Cooper’s
testing (see Figure 10.6), although he assumed the tension field angle would always be
the diagonal in each subpanel.
10.3 Effect of Small Flange Areas
For an interior panel to be qualified for a tension-field action, AASHTO

Specifications (2014) require that the flange area be sufficiently large:
 0.87(1  C )  2 Dt w
Vn  V p C   for  2.5 (10.1)
 1  d 0 / D 
2  (b fc t fc  b ft t ft )
 
When Eq. (10.1) is violated, the AASHOTO Specifications provide the following
equation to account for the partial tension-field action:
 0.87(1  C )  2 Dt w
Vn  V p C   for  2.5 (10.2)
 1  d / D  2
 d / D   (b fc t fc  b ft t ft )
 0 0 
That is, the AASHTO Specifications provide two shear strength equations for the interior
panels, depending on whether sufficient flange areas are provided. For end panels,
267
however, the proposed Eq. (6.33) explicitly considers the effect of small flanges in the
form of Mpf in Eq. (6.36). To check the adequacy of Eq. (6.33) when small flanges are
used, an additional six ABAQUS models, which were variations of Specimens SG1 and
SG2, were created and analyzed (see Table 10.2). From Specimen SG1, the flange
thickness and the width were taken as 0.5 in. and 8 in., respectively, such that the
following web-to-flange area ratio is 3.28 which is larger than 2.5. This ratio is applied
to three different width-depth ratios (0.5, 1.0, and 1.5) as shown in Models SG1-1, -2, and
-3 in Table 10.2. Similarly, Models SG2-1, -2, and -3 were modified from Specimen
SG2. The flange thickness and width used for the models are also presented in Table
10.2. The ultimate shear strengths of FEM and prediction [i.e., Eq. (6.33)] are
compared in Table 10.3. The average shear strength ratio of prediction to FEM is 0.9
with a standard deviation of 0.03. Figure 10.7 plots the results to compare the ultimate
shear strengths of prediction and FEM. The proposed equation still provides an accurate
and slightly conservative estimate of the actual shear strength.
10.4 Effect of Noncompact Web
A noncompact web is a web that experiences inelastic buckling. Since

Specimen SG4 had the noncompact web, the plastic hinge locations per the proposed
equation were different than the actual hinge locations, although the predicted ultimate
shear strength is still close to the test result due to the limit of  ( 1.0) . To further
investigate the effect of a noncompact web, three finite element models per each test
specimen were modeled by varing the web thicknesses. For Specimens SG1 with a web
yield stress of 60.3 ksi and a width-depth ratio of 0.5, the depth-thickness ratios should be
between 123 and 154 to be classified as a noncompact web. Table 10.4 summarizes the
range of the depth-thickness ratios required for a noncomapc web in each test specimen.
The depth-thickness ratios for SG1 series of models were selected to vary from 130 to
150. See Table 10.5 for the selected depth-thickness ratios for the other finite element
models. The ultimate shear strengths of finite element analyses, predictions per Eq.
(6.33), and AASHTO Specifications are also summarized in the table.
Figure 10.8 presents the predicted shear strengths per Eq. (6.33) and AASHTO
Specifications, normalized by FEM results. It is noted from Figure 10.8(a) that the
268
proposed equation overestimates the ultimate shear strength by approximately 16%.
However, Figure 10.8(b) shows that the proposed equation provides a very accurate
estimation of the ultimate shear strength. The average ratio of the prediction to FEM is
1.00 with a standard deviation of 0.02. Since the proposed equation provides either a
close or slightly nonconservative estimation when a noncompact web is used, the
resistance factor (  ) may need to be determined separately.
10.5 Effect of Unequal Top and Bottom Flanges
Highway steel plate girder bridges often have top and bottom flanges of different
dimensions. In Chapter 6, the ultimate shear strength of plate girder end panels was
derived, assuming that both the top and bottom flanges have the same dimensions and
steel grade.
Considering different top and bottom flanges, Eqs. (6.16) and (6.19) can be
rewritten as:
V ps 
M t
pf  M pm   ( D  b)M pb  M pm 

M b
pf  M pb 
a bd 0 d0
 a t t w sin 2    b t t w cos 2 ( D  b) 
  
 
  b t t w sin  cos 
 (10.3)
 2   2 d 0 
 t b cos 
2 2
 t w
2d 0
2( M tpf  M pm )
a (10.4a)
 t t w sin 2 
2( M pb  M pm )
b (10.4b)
 t t w cos 2 
where M tpf and M bpf are the plastic moments of top and bottom flanges, respectively.
Substituting Eq. (10.4) into Eq. (10.3), the post buckling strength ( V ps ) is:
M bpf  M pb t t wb 2
V ps   t t w sin (a sin   b cos )   cos 2  (10.5)
d0 2d 0
269
Therefore, summing up the beam action shear buckling strength ( Vcr ) and post-buckling
strength ( V ps ) gives the ultimate shear strength of an end panel with unequal top and
bottom flanges.
M bpf  M pb  t t wb 2
Vs   t t w sin (a sin   b cos )   cos 2    cr Dt w (10.6)
d0 2d 0
Substituting Eq. (10.4b) into the third term of Eq. (10.6) gives
M bpf  M pm
d0
The second term on right-hand side of Eq. (10.7) is the additional shear capacitiy when
the difference between top and bottom flange sizes is considered.
Consider two cases as in Section 6.6: weak flange and weak bearing stiffeners.
Case 1: M tpf  M pb (weak flange)
In this case, M pm equals to M tpf . Then, Eq. (10.7) becomes
M bpf  M tpf
d0
It is readily shown that the second term on right-hand side of Eq. (10.8) vanishes when
top and bottom flanges are identical and, thus, Eq. (10.8) reduces to Eq. (6.24).
Case 2: M pf  M pb (weak bearing stiffeners)
In this case, M pm equals to M pb and Eq. (10.7) becomes
M bpf  M pb
d0
The additional shear capacity terms, M bpf  M tpf d 0 for Case 1 or M bpf  M pb  d 0
for Case 2 is generally small when compared with the remaining two terms in either Eq.
(10.8) or Eq. (10.9). Therefore, by ignoring the additional shear terms, the proposed Eq.
(6.33) is still applicable for cases with unequal top and bottom flanges. That is, the
proposed equation essentially considers the size of the top flange only.
To validate that the proposed equation still can be used for unequal top and
bottom flanges, additional finite element analyses were carried out based on Models SG1
and SG2. Table 10.6 summarizes the top and bottom flange dimensions as well as the
270
ultimate shear strengths from both finite element analyses and the prediction per Eq.
(6.33). The ultimate shear strengths are compared in Figure 10.9. The prediction
provides a good correlation with the finite element analysis results and, therefore, it is
concluded that the proposed equation is valid for both equal and unequal top and bottom
flanges.
271
Table 10.1: Effect of Longitudinal Stiffener
(a) Longitudinal Stiffener at 0.2D
Model Subpanel Shear V prediction VFEM

Subpanel
No. Strength (kip) (kip)
1 91.8*
SG1LS02 389.2 371.0
2 297.4
1 65.5*
SG2LS02 242.2 260.1
2 176.7
* Based on shear yielding (= 0.58AwFyw); others based on Eq. (6.33).
(b) Longitudinal Stiffener at 0.5D
Model Subpanel Shear V prediction VFEM

Subpanel
No. Strength (kip) (kip)
1 201.7
SG1LS05 403.4 414.6
2 201.7
1 66.3
SG2LS05 252.6 286.8
2 66.3
272
Table 10.2: Parameters of FEM Models with Small Flange Areas
Web Depth- Width- Beam Flanges

Web
Yield Thk. Depth 2 Dt w
Model Depth,
Stress, Ratio, Ratio, Thickness, Width, (b fc t fc  b ft t ft )
No.
Fy (ksi) D/tw d0/D D (in.) tf (in.) bf (in.)
SG1-1 0.5
SG1-2 60.3 210 1.0 0.5 8 3.28
SG1-3 1.5
52.5
SG2-1 0.5
SG2-2 57.4 280 1.0 0.56 6 2.93
SG2-3 1.5
Table 10.3: Effect of Small Flange Area on Ultimate Shear Strength
FEM Eq. (6.33) Eq. (6.33)

Model No.
(kips) (kips) FEM
SG1-1 308.8 287.1 0.93
SG1-2 222.0 203.2 0.92
SG1-3 191.0 166.5 0.87
SG2-1 191.8 170.3 0.89
SG2-2 134.4 124.7 0.93
SG2-3 115.5 99.8 0.86
Average ̶ 0.90
STDEV ̶ 0.03
273
Table 10.4: Ranges of Depth-Thickness Ratio for a Noncompact Web
Range of Depth-
Specimen Web Yield Stress Width-Depth
Thickness Ratio for
No. (ksi) Ratio
Noncompact Web
SG1 60.3 0.5 123 ̶ 154
SG2 57.4 0.5 126 ̶ 157
SG3 36.9 1.5 84 ̶ 106
SG4 36.9 0.75 117 ̶ 146
274
Table 10.5: Effect of Noncompact Web
(a) Based on Models SG1 and SG2

Depth- Shear Strength (kips) Eq. (6.33)
Model
Thickness
No. FEM Eq. (6.33) AASHTO FEM
Ratio
SG1SR1 130 638.6 732.4 700.5 1.15
SG1SR2 140 578.5 669.8 604.0 1.16
SG1SR3 150 527.0 615.2 526.2 1.17
SG2SR1 130 591.6 700.9 683.5 1.18
SG2SR2 140 546.8 640.8 589.3 1.17
SG2SR3 150 500.9 580.7 513.3 1.16
Average ̶ 1.16
STDEV ̶ 0.01
(b) Based on Models SG3 and SG4

Depth- Shear Strength (kips) Eq. (6.33)
Model
Thickness
No. FEM Eq. (6.33) AASHTO FEM
Ratio
SG3SR1 85 449.0 463.4 462.2 1.03
SG3SR2 95 392.7 392.5 370.0 1.00
SG3SR3 105 339.7 338.6 302.9 1.00
SG4SR1 120 322.0 327.3 321.6 1.02
SG4SR2 130 298.2 295.2 274.0 0.99
SG4SR3 140 278.6 268.6 236.3 0.96
Average ̶ 1.00
STDEV ̶ 0.02
275
Table 10.6: Effect of Unequal Top and Bottom Flanges
(a) Based on Model SG1
Flange Width Top and Shear Strength

(in.) Bottom (kips) Eq. (6.33)
Model No. Flange
Top Bottom Thickness FEM
FEM Eq. (6.33)
Flange Flange (in.)
SG1UTB1 16 324.3 333.8 1.03
SG1UTB2 14 322.0 327.9 1.02

16 0.75
SG1UTB3 12 320.6 320.4 1.00
SG1UTB4 10 318.9 312.4 0.98
Average ̶ 1.01
STDEV ̶ 0.02
(b) Based on Model SG2
Flange Width Top and Shear Strength

(in.) Bottom (kips) Eq. (6.33)
Model No. Flange
Top Bottom Thickness FEM
FEM Eq. (6.33)
Flange Flange (in.)
SG2UTB1 14 201.0 195.5 0.97
SG2UTB2 12 14 0.625 200.2 190.8 0.95
SG2UTB3 10 198.4 184.6 0.93
Average ̶ 0.95
STDEV ̶ 0.02
276
P, 
Subpanel 1
tf
tf
0.2D
6” D
Subpanel 2
Section
(a) Longitudinal Stiffeners at 0.2D
P, 
Subpanel 1
tf
0.5D
tf
D
6”
Subpanel 2 Section
(b) Longitudinal Stiffeners at 0.5D

Figure 10.1: Finite Element Models with Different Longitudinal Stiffener Locations
277
800 400
700 B
A

600 Predicted Strength = 389.2 kips 300
Applied Load (kips)
500
400 200
300
200 100
at A: VFEM = 364.3 kips
100 at B: VFEM = 371.0 kips
0 0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
(a) Global Response
Tension-Field Direction
(b) Buckling Pattern

Figure 10.2: SG1LS02: Effect of Longitudinal Stiffener
278
600 300
B

500 250
A
Applied Load (kips)
400 200
Predicted Strength = 242.2 kips
300 150
200 100

100 50
at B: VFEM = 260.1 kips
0 0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
(a) Global Response

279
1000 500

800 400
A
Applied Load (kips)
600 Predicted Load = 403.4 kips 300
400 200
200 at A: VFEM = 397.8 kips 100

0 0
0.0 0.5 1.0 1.5 2.0 2.5 3.0
(a) Global Response

280
600 300
B
A

500 250
Applied Load (kips)
400 Predicted Load = 252.6 kips 200
300 150
200 100
100 50
0 0
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6
(a) Global Response

281
Detail “A”
(a) Overall Deformed Shape
(b) Detail “A”

Figure 10.6: Girder LS2 after Testing (adopted from Cooper, 1966)
282
350
300 45o line
250
VFEM (kips)
200
150
100
50
0
0 50 100 150 200 250 300 350
VPrediction (kips)
Figure 10.7: Coimparison of Shear Strengths of Plate Girders with Small Flange Area
283
2.0
Prediction per Eq. (6.33)
Prediction per AASHTO
1.5
Prediction/FEM
1.0
0.5
0.0
R1
R2
R3
R3
R1
R2
1S
1S
2S
1S
2S
2S
SG
SG
SG
SG
SG
SG
Model No.
(a) Model Series of SG1 and SG2
2.0
Prediction per Eq. (6.33)
Prediction per AASHTO
1.5
Prediction/FEM
1.0
0.5
0.0
R1
R2
R3
R3
R1
R2
3S
3S
4S
3S
4S
4S
SG
SG
SG
SG
SG
SG
Model No.
(b) Model Series of SG3 and SG4

Figure 10.8: Shear Strength Comparison of Plate Girder End Panel with
Noncompact Web
284
350
45o line
300
VFEM (kips)
Set of SG1
250
200
Set of SG2
150
150 200 250 300 350
VPrediction (kips)
Figure 10.9: Shear Strength Comparison of Plate Girder End Panel with Unequal
Top and Bottom Flanges
285
11. RESISTANCE FACTOR FOR THE PROPOSED SHEAR
STRENGTH EQUATION
The resistance factor,  , for the proposed shear strength in Eq. (6.33) is
computed as follows (SSRC 2010):
   R e  RVR (11.1)
where
 R  G  M  P (11.2)
 R is the bias coefficient for the resistance, i.e., the mean value of the ratio of the
measured resistance to the nominal resistance, and the remaining terms are defined as
follows:
G : bias coefficient for the cross-sectional geometry
M : bias coefficient for the material property
P : bias coefficient for the design equation
VR : coefficient of variation associated with  R  VGVM VP 

 : Reliability index taken as 3.0
R : coefficient of separation taken as 0.55
Kennedy and Gad Aly (1980) statistically determined the coefficient of the
geometry (  G ) and the associated coefficient of variation ( VG ) which are 1.015 and 0.013,
respectively. Those values were also used in White and Barker (2008) to develop the
resistance factor. Cooper et al. (1978), Kennedy and Gad Aly (1980), and White and
Barker (2008) used a bias coefficient for the material properties (  M ) of 1.10 with an
associated coefficient of variation ( VM ) of 0.110. Thus, Eq. (11.1) can be re-written as:
 ( 3)( 0.55 ) ( 0.013 )( 0.110 )VP
  1.015(1.10) P e (11.3)
Where
 Measured Strength 
 P  mean   (11.4)
 Predicted Strength per Eq. (6.33) 
286
standard deviation ( p )
VP  (11.5)
mean value ( p )
 P and VP are the bias coefficient for the design equation and the associated
coefficient of variation.
Based on the available test results of steel girder specimens (SG series: SG1 to
SG4, EPS series: EPS1-1 to EPS 8-13 in Table 6.3), the values of  P and VP are
computed and listed in Table 11.1. From Eq. (11.3), the calculated  value is 1.136,
which is larger than 1.0. It implies that Eq. (6.33) is somewhat conservative. For
design purposes, a value of 1.0 can be used.
287
Table 11.1: Resistance Factor Based on Physical Tests
Shear Strength (kips)

Specimen No. Test/Prediction
Test Prediction
SG1 315.3 328.9 0.959
SG2 190.3 188.5 1.010
SG3 212.1 200.6 1.057
SG4 279.0 266.2 1.048
EPS1-1 8.4 8.4 1.000
EPS1-2 9.0 9.7 0.928
EPS4-3 12.9 13.2 0.977
EPS2-4 8.9 9.2 0.967
EPS2-5 9.2 10.3 0.893
EPS3-6 11.6 10.2 1.137
EPS3-7 13.7 12.3 1.114
EPS5-8 8.2 7.9 1.038
EPS5-9 8.6 8.1 1.062
EPS6-10 8.9 8.5 1.047
EPS7-11 7.1 5.9 1.203
EPS7-12 5.6 5.4 1.037
EPS8-13 4.8 4.3 1.116
Avg. (  P ) 1.035
Stdev. (  P ) 0.079
VP 0.077
288
12. SUMMARY AND CONCLUSIONS
Welded steel plate girders, usually characterized by having very deep sections,
have been widely used for bridge construction. This economically efficient structural
form typically comprises a very slender web which can easily buckle in shear.
Therefore, the slender web is usually stiffened with transverse stiffeners. In early 1960s,
Basler studied the shear capacity of stiffened web panels and quantified their post-
buckling strength due to the tension-field action. This concept was adopted in
AASHTO Specifications in 1973, which allowed post-buckling strength due to the
tension-field action for both interior and end panels. When the 1986 AASHTO Interim
Specifications were released, however, tension-field action for end panels was not
permitted. This also applies to the AISC Specification (2014) for building construction.
This change negatively impacts the load rating of many steel plate girder bridges built
before 1973. The main objective of this research was to evaluate the actual shear
strength of end panels through experimental and analytical investigations and to develop
reliable shear strength equations.
The testing program consisted of two phases. In Phase 1, two large-size steel
plate girders and two steel-concrete composite girders were fabricated and each specimen
consisted of two end panels and two interior panels with a panel width-to-depth ratio of
0.5. Two different depth-to-thickness ratios (210 and 280) were considered. During
testing of two steel girder specimens, out-of-plane deformations
of the web rapidly increased once the shear buckling load predicted by AASHTO
Specifications (2014) was achieved. But the end panels were able to resist much higher
shear force after the shear buckling load was reached, indicating that a tension-field
action took place in the web.
Testing showed that the collapse mechanism in the end panels involved the
development of four plastic hinges (see Figure 3.8): one in the top flange, one in the
bearing stiffener pair, one at the junction of the top flange and bearing stiffener pair, and
one at the junction of the bottom flange and first intermediate stiffener(s). The recorded
principal tensile strains at ultimate load were close to or beyond the strain value
corresponding to the tension-field stress proposed by Basler [Eq. (1.10)]. In addition,
289
the angles of the principal tensile strains were generally within ±10° of the panel diagonal
angle with respect to the flange. The addition of a concrete slab further increased the
shear capacity of the end panels, but to a lesser extent. Two composite girder specimens
showed about 12% higher shear strength than their steel counterparts.
Finite element analyses predicted very well the actual behaviors of the test
specimens. Then a series of parametric studies was conducted to evaluate the effects of
the following parameters: flange and bearing stiffener thicknesses, panel depth-to-
thickness ratio, panel width-to-depth ratio, concrete slab width, and end shear stud
spacing. The plastic hinge locations in the top flange and bearing stiffeners directly
affected the width of the tension-field band and, thus, the post-buckling strength. The
analyses also confirmed that some portion of the web needed to be included in the
calculation of plastic moments for the top flange and bearing stiffener pair; these two
flexural strengths affected the plastic hinge locations. The effect of concrete slab width
was insignificant; testing also showed that the effective width of the concrete slab was
close to the top flange width of the steel girder.
Analytical studies were then conducted to evaluate the shear strength of the plate
girder end panels. A plastic analysis with virtual work was used to derive the post-
buckling shear strength. This model led to a predictive equation [Eq. (6.33)], which
included a partial tension-field action factor α [Eq. (6.36)]. To determine α, plastic
moments of the girder top flange and the bearing stiffener pair were needed; an effective
depth of the web was proposed for this purpose. Results from 15 experimental tests (13
small-scale specimens and 2 large-size specimens tested in Phase 1 of this program) and
99 finite element analyses correlated very well with the proposed equation. In addition,
a strut-and-tie model (see Section 6.9) was developed to include the shear strength
contribution from the concrete slab.
When the end panels of an existing steel plate girder bridge require retrofit due to
a deficiency in shear resistance, one scheme, originally proposed by Basler (1961a) that
adds another pair of end stiffeners next to the bearing stiffeners (see Figure 7.2), was
investigated. Finite element simulations of two design procedures, one proposed by
Basler and another one that has been adopted in Eurocode 3 (2006), were found to be
non-conservative and not effective. Alternate rehabilitation schemes that either (1) add
290
a pair of wing stiffeners (see Figure 7.6) welded to the bearing stiffeners and flanges and
flanges to increase the flexural strengths of these boundary elements, or (2) add another
pair of stiffeners in the mid-width of existing end panels panels (see Figure 7.12) were
proposed. For the first approach, finite element analysis showed that the proposed
retrofit scheme effectively shifts the plastic hinge in the bearing stiffener pair downward,
widens the tension-field band and, thus, increases the shear strength. A design
procedure and an example are provided (see Section 7.5). The other approach is using a
pair of angles fastened to the web by bolting in the mid-width of the existing end panels.
The effectiveness of this approach is also verified by finite element analysis. For this
approach, finite element analysis showed that placing the new stiffeners at the midspan of
the original end panel, not placing them very close to the bearing stiffeners, was more
effective. The proposed Eq. (6.33) can be used to predict the shear strength of the
rehabilitated end panel. The proposed equation is valid when the width-to-depth ratio is
not less than 0.5.
Phase 2 test specimens (two steel and two composite plate girders) had a panel
width-to-depth ratio of 0.75 and 1.5, and had a much lower depth-to-thickness ratio (=
138). The behaviors and failure modes of these shallower test specimens were very
similar to those of Phase 1. Tension-field action could be developed in these specimens
and the actual shear strengths were much higher than those predicted by the AASHTO
Specifications (2014). The ultimate shear strengths from testing were also compared
with those predicted by the proposed shear strength equations developed in the Phase 1
study. The predicted shear strengths of the steel specimens correlated well with those
from testing. However, the concrete slab resisted more shear than that predicted by the
proposed equation. It was observed that using more closely spaced shear connectors, as
was the case in Phase 2 spcimens, required the strut-and-tie model to be adjusted to
account for the different failure mechanism (see Section 9.3).
Nonlinear finite element analyses were also conducted to investigate the effects of
longitudinal stiffeners, small flange areas, noncompact web, and unequal top and bottom
flanges on the shear strength of end panels. When longtidunial stiffeners are used, the
total shear strength of the end panel can be calculated by summing up the shear strengths
of subpanels with the proposed shear strength equation. To calculate the shear strength
291
with tension-field action in the interior panels, the AASHTO Specifications (2014)
provide two equations based on the flange area relative to the toal cross-sectional area.
For end panel applications, the proposed shear strength equation automatically considers
the effect of flange area and, thus, a separate shear strength equation is not needed. The
proposed shear strength equation also applies to the case of equal or unequal top and
bottom flanges. When a noncompact web is used, the proposed shear strength equation
provides either close or slightly nonconservative estimate of the shear strength.
Based on 17 test results (4 large-size specimens and 13 small-scale specimens),
the resistance factor was statistically determined. The calculated resistance factor was
larger than 1.0, which implies that the proposed equation is somewhat conservative. For
design purposes, it is recommended that a value of 1.0 be used.
292
REFERENCES
(1) AASHTO. (1973). Standard Specifications for Highway Bridges, 11th Ed.,
American Association of State Highway Officials, Washington, D.C.
(2) AASHTO. (1977). Standard Specifications for Highway Bridges, 12th Ed.,
American Association of State Highway Officials, Washington, D.C.
(3) AASHTO. (2014). AASHTO LRFD Bridge Design Specifications, American

Association of State Highway and Transportation Officials, Washington, D.C.
(4) AASHTO. (2011). Manual for Bridge Evaluation, 2nd Ed., American Association
of State Highway and Transportation Officials, Washington, D.C.
(5) ABAQUS. (2011). ABAQUS Standard User’s Manual, Version 6.11, Dassault
Systèmes. Providence, RI.
(6) ACI. (2005). Building Code Requirements for Structural Concrete and
Commentary. ACI 318-05, American Concrete Institute, Farmington Hills,
Michigan.
(7) ACI. (2008). Building Code Requirements for Structural Concrete and
Commentary. ACI 318-08, American Concrete Institute, Farmington Hills,
Michigan.
(8) AISC. (2010). Specification for Structural Steel Buildings. ANSI/AISC 341-10,
American Institute of Steel Construction, Chicago, IL.
(9) Ajam, W. M. A. (1986). “Finite element study of the post-buckling behaviour of

plate girder panels under shear load.” Ph.D Thesis, Concordia Univ., Montreal,
Quebec, Canada.
(10) Alinia, M. M., Hahashi, H. R., and Khorram, A. (2009a). “Nonlinearity in the
postbuckling behaviour of thin steel shear panels.” Thin-Walled Structures, 47, pp.
412-420.
(11) Alinia, M. M., Shakiba, M., and Hahashi, H. R. (2009b). “Shear failure
characteristics of steel plate girders.” Thin-Walled Structures, 47, pp. 1498-1506.
(12) Allison, R. W., Johnson, R. P., and May, I. M. (1982). “Tension field action in
composite plate girdersr.” Proc., Inst. Civ. Eng., 73(2), pp. 255-276.
(13) AWS D1.1. (2000). Structural Welding Code-Steel, American Welding Society,
Miami, Florida.
293
(14) Baskar, K., Shanmugam, N. E., and Thevendran, V. (2002). “Finite-element
analysis of steel-concrete composite plate girder.” J. Struct. Eng., ASCE, 128(9),
pp. 1158-1168.
(15) Basler, K. (1961a). “Strength of plate girders in shear.” J. Struct. Div., ASCE,
87(7), pp. 151-181.
(16) Basler, K. (1961b). “Strength of plate girders under combined bending and
shear.” J. Struct. Div., ASCE, 87(2), pp. 181-197.
(17) Basler, K., Yen, B. T., Muller, J. A., and Thürlimann, B. (1960). “Web buckling
tests on welded plate girders.” Welding Research Council Bulletin, No. 64.
(18) Chern, C., and Ostapenko, A. (1969). “Ultimate strength of plate girder under
shear” Friz Engineering Laboratory Report No. 328.7, Lehigh Univ. Bethehem,
PA.
(19) Cooper, P. B. (1966). “Static strength of longitudinally stiffened plate girders”

Friz Engineering Laboratory Report No. 304.9, Lehigh Univ. Bethehem, PA.
(20) Cooper, P. B., Galambos, T. V., and Ravindra, M. K. _1978_. “LRFD criteria for
plate girders.” J. Struct. Div., 104(9), pp. 1389–1407.
(21) Darehshouri, S. F., Shanmugam, N. E., and Osman, S. A. (2012). “Collapse

behavior of composite plate girders loaded in shear.” J. Struct. Eng., ASCE,
138(3), pp. 318-326.
(22) CEN. (2006). Eurocode 3: Design of Steel Structures - Part 1-5: General Rules -
Plated Structural Elements., EN 1993-1-5, European Committee for
Standarization, Brussels, Belgium.
(23) Fujii, T. (1971) “A comparison between theoretical values and experimental

results for the ultimate shear strength of plate girders.” IABSE Volume-Band 11,
Colloquium, Design of Plate and Box Girders for Ultimate Strength, London, pp.
161-171.
(24) Herzog, M. (1974). “Ultimate strength of plate girders from tests.” J. Strut. Div.,
ASCE, 100(5), pp. 849-864.
(25) Hӧglund, T. (1971). “Simply supported thin plate plate I-girders without web
stiffeners subjected to distributed transverse load.” IABSE, Colloquium, Design of
Plate and Box Girders for Ultimate Strength, London
(26) Hӧglund, T. (1997). “Shear buckling resistance of steel and aluminum plate
girders.” Thin-Walled Structures, 29(1-4), pp. 13-30.
294
(27) Horne, M. R., and Grayson, W. R (1983). “Parametric finite element study of
transverse stiffeners for webs in shear.” Proc., Michael R. Horne Conf., L. J.
Morris, ed., Granada Publishing, London, pp. 329-341.
(28) Huslid, J. M., and Rockey, K. C. (1979). “The influence of end post rigidity on
the collapse behaviour of plate girders.” Proc., Inst. Civ. Eng., 67(2), pp. 285-312.
(29) Karthik, M. M., and Mander, J. B. (2011). “Stress-block parameters for

unconfined and confined concrete based on a unified stress-strain model.” J.
Struct. Eng., ASCE, 137(2), pp. 270-273.
(30) Kennedy, D. J. L., and Gad Aly, M. (1980). “Limit states design of steel
structures—Performance factors.” Can. J. Civ. Eng., 7(1), pp. 45–77.
(31) Kent, D. C., and Park, R. (1971). “Flexural members with confined concrete.” J.
Struct. Div., Proc., ASCE, 97(ST7), pp. 1969-1990.
(32) Komatsu, S. (1971) “Ultimate strength of stiffened plate girders subjected to

shear.” IABSE Volume-Band 11, Colloquium, Design of Plate and Box Girders for
Ultimate Strength, London, pp. 49-65.
(33) Kuhn, P. (1956). Stresses in Aircraft and Shell Structures, McGraw-Hill, New
York.
(34) Lee, S. C., and Yoo, C. H. (1998). “Strength of plate girder web panels under pure
shear.” J. Struct. Eng., ASCE, 124(2), pp. 184-194.
(35) Lee, S. C., and Yoo, C. H. (1999). “Experimental study on ultimate shear strength
of web panels.” J. Struct. Eng., ASCE, 125(8), pp. 838-846.
(36) Liang, Q. Q., Uy, B., Bradford, M. A., and Ronagh, H. (2005). “Strength analysis
of steel-concrete composite beams in combined bending and shear.” J. Struct.
Eng., ASCE, 131(10), pp. 1593-1600.
(37) Marsh, C., Ajam, W., and Ha, H. K. (1988). “Finite Element Analysis of
postbuckled shear webs.” J. Struct. Eng., ASCE, 114(7), pp. 1571-1587.
(38) Neal B. G. (1985). The plastic methods of structural analysis, 3rd Ed., Chapman
and Hall, New York.
(39) Nie, J., Xiao, Y., and Chen, L. (2004). “Experimental studies on shear strength of
steel-concrete composite beams.” J. Struct. Eng., ASCE, 130(8), pp. 1206-1213.
(40) Popovics, S. (1973). “A numerical approach to the complete stress-strain curves

of concrete.” Cement and Concrete Research, 3(5), pp. 583-599.
295
(41) Porter, D. M., Rockey, K. C., and Evans, H. R. (1975). “The collapse behaviour
of plate girders loaded in shear.” The Structural Engineer, 53(8), pp. 313-325.
(42) Real, E., Mirambell, E., and Estrada, I. (2007). “Shear response of stainless steel
plate girders.” Engineering Structures, 29(7), pp. 1626-1640.
(43) Rockey, K. C., and Leggett, D. M. A. (1962). “The buckling of a plate girder web
under pure bending when reinforced by a single longitudinal stiffener.” Proc Inst.
Civ. Eng., Vol. 21
(44) Rockey, K. C., and Skaloud, M. (1972). “The ultimate load behaviour of plate
girders loaded in shear.” The Structural Engineer, 50(1), pp. 29-48.
(45) Salmon, C. G., Johnson, J. E. (1996). Steel Structures-Design and Behavior, 4th
Ed., Prentice Hall, New Jersey.
(46) Safar, S. S. (2013). “Shear strength of end web panels.” Thin-Walled Structures,
67, pp. 101-109.
(47) Shanmugam, N. E., and Baskar, K. (2003). “Steel-concrete composite plate

girders subject to shear loading.” J. Struct. Eng., ASCE, 129(9), pp. 1230-1242.
(48) SSRC. (2010). Guide to Stability Design Criteria for Metal Structures, 6th Ed.,
Edited by Ronald D. Ziemian, Wiley, New York.
(49) Stanway, G. S., Champman, J. C., and Dowling, P. J (1993). “Behaviour of a web
plate in shear with an intermediate stiffener.” Proc., Inst. Civ. Eng., 116(1), pp.
54-68.
(50) Thorenfeldt, E., Tomaszewicz, A., and Jensen, J. J. (1987). “Mechanical

properties of high-strength concrete and application in design.” Proc. of the
Symposium on Utilization of High-Strength Concrete, Tapir, Trondheim, Norway,
pp. 149-159.
(51) Timoshenko, S. P., Gere, J. M. (1961). Theory of Elastic Stability, 2nd Ed.,
McGraw-Hill, New York.
(52) Tsai, W. T. (1988). “Uniaxial compression stress-strain relation of concrete.” J.

Struct. Eng., 114(9), pp. 2133-2136.
(53) Wagner, H. (1931). “Flat sheet metal girder with very thin metal web.” Tech.
Notes. 604-606, National Advisory Committee on Aeronautics (NACA),
Washington, D.C.
(54) White, D. W., Barker, M. G. (2008). “Shear resistance of transversely stiffened

steel I-girders.” J. Struct. Eng., ASCE, 134(9), pp. 1425-1436.
296
(55) Wilson, J. M. (1886). “On specifications for strength of iron bridges.” Trans.
ASCE, Vol. 15, Part I, pp. 401-403, 489-490.
(56) Yen, B. T., and Basler, K. (1962). “Results of an investigation of plate girders”
Friz Engineering Laboratory Report No. 251.25, Lehigh Univ. Bethehem, PA.
(57) Yoo, C. H., and Lee, S. C. (2006). “Mechanics of web panel post-buckling
behavior in shear.” J. Struct. Eng., ASCE, 132(10), pp. 1580-1589.
297
APPENDIX A. INSTRUMENTATION
A1 Displacement Transducers
A1.1 Specimens SG1 and SG2

The locations of displacement transducers for Specimens SG1 and SG2 are shown
in Figures A.1 and A.2, respectively. Displacement transducer L1 measured the girder
mid-span vertical deflection and was used as the feedback sensor for actuator control.
L2 through L9 measured web out-of-plane deformation. L10 through L21 measured top
flange deformation along half of the end panel width in the region of plastic hinge
formation. L22 through L31 measured deformations along half of the length of the
bearing stiffeners. L32 through L41 measured bottom flange deformation. L42
through L49 measured the average shear deformation of the panels.
A1.2 Specimens CG1 and CG2

The locations of displacement transducers for Specimens CG1 and CG2 are
shown in Figures A.3 and A.4, respectively. These specimens had an arrangement of
displacement transducers similar to their steel girder counterparts, except that the
transducers on top and bottom flanges of the steel girder were removed and two
displacement transducers (L28 and L29) where added to measure the slip between the
concrete slab and steel girder (see Figures A.3 and A.4).
A1.3 Specimens SG3 and SG4

The locations of displacement transducers for Specimens SG3 and SG4 are shown
in Figures A.5 and A.6, respectively. These specimens had a similar arrangement of
displacement transducers to Specimens SG1 and SG2. However, the displacement
transducers on bearing stiffeners and flanges were removed.
A1.4 Specimens CG3 and CG4

Figures A.7 and A.8 show the locations of displacement transducers for
Specimens SG3 and SG4, respectively.
298
A West B
L10~L15 L16~L21
(L to R) (R to L)
1"
1'-14 1"
1'-14
L2 L7
L22~L26 L27~L31
(T to B) (T to B)
L3 L5 3" 3"
L6 L8
L42 L43 L44 L45 L46 L47 L48 L49

L4 L9
1"
1'-14 1"
1'-14
L32~L36 L37~L41
(L to R) (L to R)
A L1 B
Elevation
C
L
West
C
L
L10~L15 L16~L21
5" (L to R) (R to L) 5"
5@3" 5@3"
1" Top Face of Top Flange 1"
1" 4@341" 4@341" 1"

C
L
1'-54 1'-54
West
L32~L36 5" L37~L41
C
L
(L to R) (L to R)
Bottom Face of Bottom Flange
13
4" 13
4"
2" 2"
6" 6"
L22~L26
6" (T to B) 6"
L27~L31
6" (T to B) 6"
6" 6"
South North North South
Section A -A Section B - B
Figure A.1: Specimen SG1: Displacement Transducers
299
A West B
L10~L15 L16~L21
(L to R) (L to R)
1"
1'-14 1"
1'-14
L2 L7
L22~L26 L27~L31
(T to B) (T to B)
L3 L5 L6 L8
3" 3"
L42 L43 L44 L45 L46 L47 L48 L49

L4 L9
1"
1'-14 1"
1'-14
L32~L36 L37~L41
(L to R) (L to R)
A L1 B
Elevation
C
L
West
C
L
L10~L15 L16~L21
4" (L to R) (L to R) 4"
5@3" 5@3"
1" Top Face of Top Flange 1"
1" 1"
C
L
1'-54 4@341" 4@341" 1'-54

West
L32~L36 4" 4" L37~L41
C
L
(L to R) (L to R)
17
8" 17
8"
2" 2"
6" 6"
L22~L26
6" (T to B) 6"
L27~L31
6" (T to B) 6"
6" 6"
300
A West B
L28 L29
1"
1'-14 1"
1'-14
L2 L7
L10~L14 L15~L19
(T to B) (T to B)
L3 L5 L6 L8
3" 3"
L20 L21 L22 L23 L24 L25 L26 L27

L4 L9
1"
1'-14 1"
1'-14
A L1 B
Elevation
13
4" 13
4"
2" 2"
6" 6"
L10~L14
6" (T to B) 6"
L15~L19
6" (T to B) 6"
6" 6"
Figure A.3: Specimen CG1: Displacement Transducers
301
A West B
L28 L29
1"
1'-14 1"
1'-14
L2 L7
L10~L14 L15~L19
(T to B) (T to B)
L3 L5 3" 3"
L6 L8
L20 L21 L22 L23 L24 L25 L26 L27

L4 L9
1"
1'-14 1"
1'-14
A L1 B
Elevation
17
8" 17
8"
2" 2"
6" 6"
6" L10~L14 L15~L19 6"

(T to B) (T to B)
6" 6"
6" 6"
302
West
CL CL
11"
11"
L02 L05
2"
2"
L03 L06
CL
L04 L07
11"
11"
L1
Elevation
West
CL CL CL CL
11"
11"
L02 L05
2"
2"
2"
2"
L03 L08 L09 L06
CL
L04 L07
11"
11"
L1
Elevation
303
CL West CL
L08 L09
11"
11"
L02 L05
2"
2"
L03 L06
CL
11" L04 L07
11"
L1
Elevation
CL CL West CL CL
L10 L11
11"
11"
L02 L05
2"
2"
2"
2"
L03 L08 L09 L06
CL
L04 L07
11"
11"
L1
Elevation
304
A2. Strain Gages
A number of strain rosettes and uniaxial strain gages were installed on the
specimens. The layouts of the gages are shown in Figure A.9 to Figure A.16. The
strain gages labeled in parentheses represent gages on back side of the specimens.
The strain gage rosettes consist of three individual strain gage elements oriented
precisely 45° apart. The strain gage rosettes were installed on the web panels to
measure the principal strains and directions. In all end panels excluding the West end
panel of Specimen SG1, a group of strain gage rosettes were placed 3 in. apart along a
line perpendicular to and intersecting the midpoint of the panel diagonal (see Figures A.9
through A.16).
Uniaxial strain gages were used to measure the strains at the specific locations.
Most uniaxial strain gages were installed on the flanges and bearing stiffeners near plastic
hinge locations. In Specimens SG2, CG1, and CG2, uniaxial strain gages were placed
back-to-back on the bearing stiffeners to measure both flexural and axial strains.
305
A West B
13
4" 2" 13
4"
53 3
8" 58" 2"
2" 2" 2"
6" S1 R1 R2 R3 R12 R14
S6 6"
1'-01"
S2 8 S7
6" 6"
1"
2'-04
6" S3 R4 R16
S8 6"
1"
1'-08 R18
S4 R20 S9
6" 90 6"
R17 R22
S5 R13 R15 R19 R24 S10
R5 R6 R7
1"
1'-08 R21
(R25)
9@3" (2') R23
b b y
R8 1"
2'-24
r r r y
y 1'-01" y b
8
R9 R10 R11 r
b
2"
53 3
8" 58"
A 2"
Elevation B
1" 1" 33
43
4"
32 32 4"
C
L
53" 53" 62
1"
8 8
West
C
L
S23 S24 S26 S28 S29 S30
S25 S27
13
4"
Top Face of Top Flange
C
L
West
C
L
S31 S32 S33 S34 S35 S36
1" 53" 53" 53" 53" 1"

1'-54
8 8 Bottom Face of Bottom Flange 8 8
1'-54
1"
34 1"
34
2" 2"
S12 6" 6" S18 1"
1"
34 34
S13 6" 6" S19
S11 S14 6" 6" S20 S17

S15 6" 6" S21
S16 S22
Figure A.9: Specimen SG1: Strain Gage and Rosette Locations
306
A West B
13
4" 2" 2" 3"
14
2" 2"
2" 2"
6" S1 R1 R14 S6 6"
S2 S7
6" 1" 6"
1"
2'-04 2'-24
6" S3 R11 R16 S8 6"
R9 R18
1" S4 90 R7 R20 90 S9 1"
64 64
R5 R10 R17 R22
S5 R3 R8 R12 R13 R19 R24 S10
R6 R21
R4 R23
8@3" (2') 8@3" (2')
1" b y
2'-04 1"
2'-24
b r r y
y b
r r
R2 R15
y b
2" 2"
A 2"
Elevation 2"
B
1" 1" 3" 33 1"

43
4"
32 32 34 4"
1"
32 32 43
4"
C
L West
C
L
S47 S48 S50 S52 S54 S56
S49 S51 S53 S55
13
4"
3"
14
C
L
West
C
L
S57 S58 S59 S60 S61 S62
1" 53 3 53 3 1"
1'-54 8" 58" Bottom Face of Bottom Flange 8" 58" 1'-54
33" 33"
8 8
2" 2"
S13 (S14) 6" 6" S31 (S32)
33" 33"
8 S15 (S16) 6" 8
6" S33 (S34)
S11 S17 (S18) 6" 6" S35 (S36) S29

(S12) (S30)
S19 (S20) 61" 1" S37 (S38)
64
4
S21 (S22) 1 1" S39 (S40)
8 "4 84
South North South North
S23 (S24) S41 (S42)
8" 8"
S25 (S26) S43 (S44)

8" 8"
2" 2"
S27 (S28) S45 (S46)
Note: Gages in parentheses
Section A -A are on back side Section B - B
307
A West B
13
4" 2" 2" 13
4"
2" 2"
2" 2"
6" S1 R1 R14
S6 6"
S2 S7
6" 1" 6"
1"
2'-04 2'-24
6" S3 R11 R16
S8 6"
R9 R18
S4 R7 R20 S9
6" 90 90 6"
R5 R10 R17 R22
S5 R3 R8 R12 R13 R19 R24 S10
R6 R21
R4 R23
8@3" (2') 8@3" (2')
1" b y
2'-04 1"
2'-24
b r r y
y b
r r
R2 R15
y b
2" 2"
A 2"
Elevation 2"
B
1" 1" 33 33 1"

43
4"
32 32 4" 4"
1"
32 32 43
4"
C
L West
C
L
S47 S48 S50 S52 S54 S56
S49 S51 S53 S55
13
4" 13
4"
C
L
West
C
L
S57 S58 S59 S60 S61 S62
1"
1'-54 3" 53"
58 3" 53"
58 1"
1'-54
8 Bottom Face of Bottom Flange 8
1"
34 1"
34
2" 2"
S13 (S14) 6" 6" S31 (S32)
1"
34 1"
34
S15 (S16) 6" 6" S33 (S34)
S11 S17 (S18) 6" 6" S35 (S36) S29

(S12) (S30)
S19 (S20) 61" 1" S37 (S38)
64
4
S21 (S22) 1 1" S39(S40)
8 " 4 84
S23 (S24) S41 (S42)
8" 8"
S25 (S26) S43 (S44)

8" 8"
2" 2"
S27 (S28) S45 (S46)
Figure A.11: Specimen CG1: Strain Gage and Rosette Locations
308
A West B
13
4" 2" 2" 13
4"
2" 2"
2" 2"
6" S1 R1 R14 S6 6"
S2 S7 6"
6" 1"
1"
2'-04 2'-24
6" S3 R11 R16 S8 6"
R9 R18
1" S4
64 90 R7 R20 90 S9 61"
4
R5 R10 R17 R22
S5 R3 R8 R12 R13 R19 R24 S10
R6 R21
R4 R23
8@3" (2') 8@3" (2')
1" b y
2'-04 1"
2'-24
b r r y
y b
r r
R2 R15
y b
2" 2"
A 2" Elevation 2"

B
1" 1" 33 33 1"

43
4"
32 32 4" 4"
1"
32 32 43
4"
C
L
West
C
L
S47 S48 S50 S52 S54 S56
S49 S51 S53 S55
13
4" 13
4"
C
L
West
C
L
S57 S58 S59 S60 S61 S62
1" 53 3 53 3 1"
1'-54 8" 58" Bottom Face of Bottom Flange 8" 58" 1'-54
33
8" 33
8"
2" 2"
S13 (S14) 6" 6" S31 (S32)
33
8" 33
8"
S15 (S16) 6" 6" S33 (S34)
S11 S17 (S18) 6" 6" S35 (S36) S29

(S12) (S30)
S19 (S20) 61" 1" S37 (S38)
64
4
S21 (S22) 1 1" S39 (S40)
8 " 4 84
S23 (S24) S41 (S42)
8" 8"
S25 (S26) S43 (S44)

8" 8"
2" 2"
S27 (S28) S45 (S46)
309
A West B
3"
@
es
ett
R05 R06
R04 R07
os
5R
R02 R03 R08
R09
R10
b R01 y
r
r
b
y
90°
A Elevation B
5" 6" 6" 6" 6"

CL West 6" 6" 6" 6" 5"
CL
S13 S14 S15 S16 S17 S18 S19 S20 S21S22
CL
West
S23 S24 S25 S26

CL
1 1
2" 2"
1"
4'-14 1"
1'-62 1"
1'-62 1"
4'-14
1" 1"
42
42
1"
1"
S07
4" 4" 4" 4"
S01
4" 4" 4" 4"
S02 S08
S03 S09
S04 (S05) S10 (S11)
S06 S12

310
A West B
3"
@
t tes
o se
5R
R08
R04
R05 R10
R02 R09
90°
R03 R06 R07 R12
R11
R01
y
b
r
r
b
y
A Elevation B
1" 41" 41" 41"

44 1" 41" 41" 41"
44
4 4 4 4 4 4
5"
CL West 5"
S14 S16 S19 S21

CL
S13 S15 S17 S18 S20 S22
CL
West
S23 S24 S25 S26

CL
1 1
2" 2"
1"
2'-02 11" 11" 1"
2'-02
1" 1"
42
42
1"
1"
S01 S07
5"
5"
S02 S08
5"
5"
S03 S09
5"
5"
S04 S10
5"
5"
S05 S11
6"
6"
S06 S12
311
A West B
3"
@
es
ett
R05 R06
R04 R07
os
5R
R02 R03 R08
R09
R10
b R01 y
r
r
b
y
90°
A Elevation B
5" 6" 6" 6" 6"

CL West 6" 6" 6" 6" 5"
CL
S13 S14 S15 S16 S17 S18 S19 S20 S21S22
CL
West
S23 S24 S25 S26

CL
1 1
2" 2"
1"
4'-14 1"
1'-62 1"
1'-62 1"
4'-14
1" 1"
42
42
1"
1"
S07
4" 4" 4" 4"
S01
4" 4" 4" 4"
S02 S08
S03 S09
S04 (S05) S10 (S11)
S06 S12

312
Steel Steel
Plate Plate
West
43°
S27, S28, S29, S30
S31~S34
C C
CL CL
Elevation
CL
11"
S31
S32
Shear Stud, TYP.
S27 S29
?
S28 S30
S33
S34
11"
View C - C
Figure A.15: Specimen CG3: Strain Gage and Rosette Locations (continued)
313
A West B
3"
@
t tes
o se
5R
R08
R04
R05 R10
R02 R09
90°
R03 R06 R07 R12
R11
R01
y
b
r
r
b
y
A Elevation B
1" 41" 41" 41"

44 1" 41" 41" 41"
44
4 4 4 4 4 4
5"
CL West 5"
S14 S16 S19 S21

CL
S13 S15 S17 S18 S20 S22
CL
West
S23 S24 S25 S26

CL
1 1
2" 2"
1"
2'-02 11" 11" 1"
2'-02
1" 1"
42
42
1"
1"
S01 S07
5"
5"
S02 S08
5"
5"
S03 S09
5"
5"
S04 S10
5"
5"
S05 S11
6"
6"
S06 S12
314
Steel Steel
Plate Plate
West
43°
S27, S28, S29, S30
S31~S34
C C
CL CL
Elevation
CL
11"
S31
S32
Shear Stud, TYP.
S27 S29
CL
S28 S30
S33
S34
11"
View C - C
Figure A.16: Specimen CG4: Strain Gage and Rosette Locations (continued)
315
APPENDIX B. PLOTS OF DEFORMATION AND STRAIN
B1. Deformation Plots
L16…………………….….…..L1……………………………L10
0.0
East
-0.2
Flange Deformation (in.)
-0.4
-0.6
at 0.15-in. disp.
-0.8 at 0.25-in. disp. (max. load)
at 0.3-in. disp
at 0.35-in. disp.
-1.0 at 0.4-in. disp.
at 0.6-in. disp.
-1.2
0 20 40 60 80 100 120 140 160
Location(in.)
(a) Top Flange Deformation (in.)
L41…….……….……..L1………….….………L32
0.0
East
-0.2
-0.4
-0.6
at 0.15-in. disp.
at 0.3-in. disp
at 0.35-in. disp.
-1.0 at 0.4-in. disp.
at 0.6-in. disp.
-1.2
0 20 40 60 80 100 120 140 160
Location(in.)
(b) Bottom Flange Deformation (in.)
Figure B.1: Specimen SG1: Flange Deformation
316
60
50 L22
……...........
40
Location (in.)
30
L26
20 at 0.15-in. disp.
at 0.25-in. disp. (max. load)
at 0.3-in. disp
at 0.4-in. disp.
at 0.6-in. disp.
0
-1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0
Bearing Stiffener Deformation (in.)
(a) East Side
60
50 L27
……...........
40
Location (in.)
30
L31
at 0.3-in. disp
at 0.4-in. disp.
at 0.6-in. disp.
0
0.0 0.2 0.4 0.6 0.8 1.0 1.2
(b) West Side
Figure B.2: Specimen SG1: Bearing Stiffener Deformation
317
L21…………………….…...…..L1……………………………L10
0.0
East
-0.2
-0.4
-0.6
at 0.15-in. disp.
at 0.3-in. disp
at 0.35-in. disp.
-1.0 at 0.4-in. disp.
at 0.6-in. disp.
-1.2
0 20 40 60 80 100 120 140 160
Location(in.)
(a) Top Flange Deformation (in.)
L41…….……….…..…..L1……………….……L32
0.0
East
-0.2
-0.4
-0.6
at 0.15-in. disp.
at 0.3-in. disp
at 0.35-in. disp.
-1.0 at 0.4-in. disp.
at 0.6-in. disp.
-1.2
0 20 40 60 80 100 120 140 160
Location(in.)
(b) Bottom Flange Deformation (in.)
Figure B.3: Specimen SG2: Flange Deformation
318
60
50 L22
……...........
40
Location (in.)
30
L26
at 0.3-in. disp
at 0.4-in. disp.
at 0.6-in. disp.
0
-1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0
(a) East Side
60
50 L27
……...........
40
Location (in.)
30
L31
at 0.3-in. disp
at 0.4-in. disp.
at 0.6-in. disp.
0
0.0 0.2 0.4 0.6 0.8 1.0 1.2
(b) West Side
Figure B.4: Specimen SG2: Bearing Stiffener Deformation
319
60
50 L10
……...........
40
Location (in.)
30
L14
at 0.4-in. disp
at 0.5-in. disp.
at 0.6-in. disp.
0
-1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0
(a) East Side
60
50 L15
…...........
40
Location (in.)
30
L19
at 0.4-in. disp
at 0.5-in. disp.
at 0.6-in. disp.
0
0.0 0.2 0.4 0.6 0.8 1.0 1.2
(b) West Side
Figure B.5: Specimen CG1: Bearing Stiffener Deformation
320
60
50 L10
……...........
40
Location (in.)
30
L14
at 0.3-in. disp
at 0.4-in. disp.
at 0.6-in. disp.
0
-1.2 -1.0 -0.8 -0.6 -0.4 -0.2 0.0
(a) East Side
60
50 L15
…...........
40
Location (in.)
30
L19
at 0.3-in. disp
at 0.4-in. disp.
at 0.6-in. disp.
0
0.0 0.2 0.4 0.6 0.8 1.0 1.2
(b) West Side
Figure B.6: Specimen CG2: Bearing Stiffener Deformation
321
B2. Strain Plots
B2. 1 Principal Strain Plots in Web Panels
0.005
at 0.15-in. disp.
at 0.3-in. disp
0.004 at 0.35-in. disp.
at 0.4-in. disp.
0.003
0.002
0.001
0.0
0 5 10 15
R16…………………….…...…..……………….………R24 20 25 30
Distance across tension field (in.)
West End Panel
Figure B.7: Specimen SG1: Principal Strain
322
0.004
at 0.15-in. disp.
at 0.3-in. disp
0.003 at 0.35-in. disp.
at 0.4-in. disp.
0.002
0.001
0.0
0 5 10 15 20
R03…………………….…...…..………...……….………R11 25 30
(a) East End Panel
0.004
at 0.15-in. disp.
at 0.3-in. disp
0.003
at 0.35-in. disp.
at 0.4-in. disp.
0.002
0.001
0.0
0 5 10 15 20
R16…………………….…...…..………...……….………R24 25 30
(b) West End Panel
323
0.005
at 0.15-in. disp.
0.004 at 0.4-in. disp
at 0.45-in. disp.
at 0.5-in. disp.
0.003
0.002
0.001
0.0
0 5 10 15 20
R03…………………….…...…..………...……….………R11 25 30
(a) East End Panel
0.005
at 0.15-in. disp.
0.004 at 0.4-in. disp
at 0.45-in. disp.
at 0.5-in. disp.
0.003
0.002
0.001
0.0
0 5 10 15 20
R16…………………….…...…..………...……….………R24 25 30
(b) West End Panel
Figure B.9: Specimen CG1: Principal Strain
324
0.004
at 0.15-in. disp.
0.003 at 0.3-in. disp
at 0.35-in. disp.
at 0.4-in. disp.
0.002
0.001
0.0
-0.001
0 5 10 15 20
R03…………………….…...…..………...……….………R11 25 30
(a) East End Panel
0.004
at 0.15-in. disp.
0.003 at 0.3-in. disp
at 0.35-in. disp.
at 0.4-in. disp.
0.002
0.001
0.0
-0.001
0 5 10 15 20
R16…………………….…...…..………...……….………R24 25 30
(b) West End Panel
325
0.010
at 0.15-in. disp.
0.008 at 0.30-in. disp
at 0.45-in. disp.
at 0.60-in. disp.
0.006
0.004
0.002
0.0
0 5 10
R05…………………….…...…..……………R01 15 20
(a) East End Panel
0.010
at 0.15-in. disp.
0.008 at 0.26-in. disp. (max. load)
at 0.30-in. disp
at 0.45-in. disp.
at 0.60-in. disp.
0.006
0.004
0.002
0.0
0 5 10
R06…………………….…...…..……………R10 15 20
(b) West End Panel
326
0.03
at 0.15-in. disp.
at 0.25-in. disp.
at 0.75-in. disp
at 1.50-in. disp.

0.02
0.01
0.0
0 5 10
R05…………………….…...…..……………R01 15 20
(a) East End Panel
0.03
at 0.15-in. disp.
at 0.25-in. disp.
at 0.75-in. disp
at 1.50-in. disp.

0.02
0.01
0.0
0 5 10
R08…………………….…...…..……………R12 15 20
(b) West End Panel
327
0.015
at 0.15-in. disp.
at 0.25-in. disp.
at 0.50-in. disp
at 0.75-in. disp.
0.005
0.0
0 5 10
R05…………………….…...…..……………R01 15 20
(a) East End Panel
0.015
at 0.15-in. disp.
at 0.25-in. disp.
at 0.50-in. disp
at 0.75-in. disp.
0.005
0.0
0 5 10
R06…………………….…...…..……….…R10 15 20
(b) West End Panel
328
0.015
at 0.15-in. disp.
at 0.25-in. disp.
at 0.50-in. disp
at 0.75-in. disp.
0.010 at 1.00-in. disp.
0.005
0.0
0 5 10
R05…………………….…...…..……………R01 15 20
(a) East End Panel
0.015
at 0.15-in. disp.
at 0.25-in. disp.
at 0.50-in. disp
at 0.75-in. disp.
0.010 at 1.00-in. disp.
0.005
0.0
0 5 10
R08…………………….…...…..……………R12 15 20
(b) West End Panel
329
B2. 2 Strain Plots in Extended Webs and Bearing Stiffeners
60
Top Flange
50 S01
……...........
40
Location (in.)
30
S05
at 0.15-in. disp.
20 at 0.25-in. disp. (max. load)
at 0.3-in. disp
at 0.35-in. disp.
10 at 0.4-in. disp.
at 0.6-in. disp.
Bottom Flange
0
-0.004 -0.002 0.0 0.002 0.004
(a) East Side
60
Top Flange
50 S06
……...........
40
Location (in.)
30
S10
at 0.15-in. disp.
at 0.3-in. disp
at 0.35-in. disp.
10 at 0.4-in. disp.
Bottom Flange
at 0.6-in. disp.
0
-0.03 -0.02 -0.01 0.0 0.01 0.02 0.03
(b) West Side
Figure B.15: Specimen SG1: Strain Profiles in Extended Web
330
60
Top Flange
50 S12
……...........
40
Location (in.)
30
S16
at 0.15-in. disp.
at 0.3-in. disp
at 0.35-in. disp.
10 at 0.4-in. disp.
at 0.6-in. disp.
Bottom Flange
0
-0.005 -0.003 -0.001 0.001 0.003 0.005
(a) East Side
60
Top Flange
50 S18
…….......
40
Location (in.)
S21
30
at 0.15-in. disp.
at 0.3-in. disp
at 0.35-in. disp.
10 at 0.4-in. disp.
Bottom Flange at 0.6-in. disp.
0
-0.005 -0.003 -0.001 0.001 0.003 0.005
(b) West Side
Figure B.16: Specimen SG1: Strain Profiles in Bearing Stiffener
331
60
Top Flange
50 S01
……...........
40
Location (in.)
30
S05 at 0.15-in. disp.
at 0.3-in. disp
at 0.35-in. disp.
10 at 0.4-in. disp.
at 0.6-in. disp.
Bottom Flange
0
-0.004 -0.002 0.0 0.002 0.004
(a) East Side
60
Top Flange
50 S06
……...........
40
Location (in.)
30
S10
at 0.15-in. disp.
at 0.3-in. disp
at 0.35-in. disp.
10 at 0.4-in. disp.
at 0.6-in. disp.
Bottom Flange
0
-0.03 -0.02 -0.01 0.0 0.01 0.02 0.03
(b) West Side
Figure B.17: Specimen SG2: Strain Profiles in Extended Web
332
60
Top Flange
50
40
Location (in.)
30
at 0.15-in. disp.
at 0.3-in. disp
at 0.35-in. disp.
10 at 0.4-in. disp.
at 0.6-in. disp.
Bottom Flange
0
-0.005 -0.003 -0.001 0.001 0.003 0.005
Bending Strain (in./in.)
(a) Bending Strain Profiles
60
Top Flange
50
40
Location (in.)
30
at 0.15-in. disp.
at 0.3-in. disp
at 0.35-in. disp.
10 at 0.4-in. disp.
at 0.6-in. disp.
Bottom Flange
0
-0.005 -0.003 -0.001 0.001 0.003 0.005
Axial Strain (in./in.)
(b) Axial Strain Profiles
Figure B.18: Specimen SG2: Strain Profiles in Bearing Stiffener (East Side)
333
60
Top Flange
50
40
Location (in.)
30
at 0.15-in. disp.
at 0.3-in. disp
at 0.35-in. disp.
10 at 0.4-in. disp.
at 0.6-in. disp.
Bottom Flange
0
-0.005 -0.003 -0.001 0.001 0.003 0.005
60
Top Flange
50
40
Location (in.)
30
at 0.15-in. disp.
at 0.3-in. disp
at 0.35-in. disp.
10 at 0.4-in. disp.
at 0.6-in. disp.
Bottom Flange
0
-0.005 -0.003 -0.001 0.001 0.003 0.005
Figure B.19: Specimen SG2: Strain Profiles in Bearing Stiffener (West Side)
334
60
Top Flange
50 S01
……...........
40
Location (in.)
30
S05
at 0.15-in. disp.
at 0.4-in. disp
at 0.45-in. disp.
10 at 0.5-in. disp.
at 0.6-in. disp.
Bottom Flange
0
-0.004 -0.002 0.0 0.002 0.004
(a) East Side
60
Top Flange
50 S06
……...........
40
Location (in.)
30
S10
at 0.15-in. disp.
at 0.4-in. disp
at 0.45-in. disp.
10 at 0.5-in. disp.
at 0.6-in. disp.
Bottom Flange
0
-0.03 -0.02 -0.01 0.0 0.01 0.02 0.03
(b) West Side
Figure B.20: Specimen CG1: Strain Profiles in Extended Web
335
60
Top Flange
50
40
Location (in.)
30
at 0.15-in. disp.
at 0.3-in. disp
at 0.35-in. disp.
10 at 0.4-in. disp.
at 0.6-in. disp.
Bottom Flange
0
-0.005 -0.003 -0.001 0.001 0.003 0.005
60
Top Flange
50
40
Location (in.)
30
at 0.15-in. disp.
at 0.4-in. disp
at 0.45-in. disp.
10 at 0.5-in. disp.
at 0.6-in. disp.
Bottom Flange
0
-0.005 -0.003 -0.001 0.001 0.003 0.005
Figure B.21: Specimen CG1: Strain Profiles in Bearing Stiffener (East Side)
336
60
Top Flange
50
40
Location (in.)
30
at 0.15-in. disp.
at 0.4-in. disp
at 0.45-in. disp.
10 at 0.5-in. disp.
at 0.6-in. disp.
Bottom Flange
0
-0.005 -0.003 -0.001 0.001 0.003 0.005
60
Top Flange
50
40
Location (in.)
30
at 0.15-in. disp.
at 0.4-in. disp
at 0.45-in. disp.
10 at 0.5-in. disp.
at 0.6-in. disp.
Bottom Flange
0
-0.005 -0.003 -0.001 0.001 0.003 0.005
Figure B.22: Specimen CG1: Strain Profiles in Bearing Stiffener (West Side)
337
60
Top Flange
S01
50
……...........
40
Location (in.)
30
S05
at 0.15-in. disp.
at 0.3-in. disp
at 0.35-in. disp.
10 at 0.4-in. disp.
at 0.6-in. disp.
Bottom Flange
0
-0.004 -0.002 0.0 0.002 0.004
(a) East Side
60
Top Flange
50 S06
……...........
40
Location (in.)
30
S10
at 0.15-in. disp.
at 0.3-in. disp
at 0.35-in. disp.
10 at 0.4-in. disp.
at 0.6-in. disp.
Bottom Flange
0
-0.03 -0.02 -0.01 0.0 0.01 0.02 0.03
(b) West Side
Figure B.23: Specimen CG2: Strain Profiles in Extended Web
338
60
Top Flange
50
40
Location (in.)
30
at 0.15-in. disp.
at 0.3-in. disp
at 0.35-in. disp.
10 at 0.4-in. disp.
at 0.6-in. disp.
Bottom Flange
0
-0.005 -0.003 -0.001 0.001 0.003 0.005
60
Top Flange
50
40
Location (in.)
30
at 0.15-in. disp.
at 0.3-in. disp
at 0.35-in. disp.
10 at 0.4-in. disp.
at 0.6-in. disp.
Bottom Flange
0
-0.005 -0.003 -0.001 0.001 0.003 0.005
Figure B.24: Specimen CG2: Strain Profiles in Bearing Stiffener (East Side)
339
60
Top Flange
50
40
Location (in.)
30
at 0.15-in. disp.
at 0.3-in. disp
at 0.35-in. disp.
10 at 0.4-in. disp.
at 0.6-in. disp.
Bottom Flange
0
-0.005 -0.003 -0.001 0.001 0.003 0.005
60
Top Flange
50
40
Location (in.)
30
at 0.15-in. disp.
at 0.3-in. disp
at 0.35-in. disp.
10 at 0.4-in. disp.
at 0.6-in. disp.
Bottom Flange
0
-0.005 -0.003 -0.001 0.001 0.003 0.005
Figure B.25: Specimen CG2: Strain Profiles in Bearing Stiffener (West Side)
340
45
Top Flange S01
40
……...........
35
30
Location (in.)
S06
25
15 at 0.30-in. disp
at 0.45-in. disp.
at 1.00-in. disp.
5
Bottom Flange
0
-0.0020 -0.0010 0.0 0.0005 0.0015
(a) East Side
45
Top Flange S07
40
……...........
35
30
Location (in.)
S12
25
15 at 0.30-in. disp
at 0.45-in. disp.
at 1.00-in. disp.
5
Bottom Flange
0
-0.0020 -0.0010 0.0 0.0005 0.0015
(b) West Side
341
45
Top Flange S01
40
…….............................
35
30
Location (in.)
25
15 at 0.75-in. disp
at 1.00-in. disp.
5
Bottom Flange
0
-0.010 -0.005 0.0 0.005 0.010
(a) East Side
45
Top Flange S07
40
……........... ...............
35
30
Location (in.)
25
15 at 0.75-in. disp
at 1.00-in. disp.
5
Bottom Flange
0
-0.010 -0.005 0.0 0.005 0.010
(b) West Side
342
45
Top Flange S01
40
……...........
35
30
Location (in.)
S06
25
at 0.25-in. disp.
15 at 0.50-in. disp
at 0.75-in. disp.
5 Bottom Flange
0
-0.006 -0.004 -0.002 0.0 0.002 0.004
(a) East Side
45
Top Flange S07
40
……...........
35
30
Location (in.)
S12
25
at 0.25-in. disp.
15 at 0.50-in. disp
at 0.75-in. disp.
5
Bottom Flange
0
-0.006 -0.004 -0.002 0.0 0.002 0.004
(b) West Side
Figure B.28: Specimen CG3: Strain Profiles in Bearing Stiffener
343
45
Top Flange S01
40
…….............................
35
30
Location (in.)
25
15 at 0.50-in. disp
at 0.75-in. disp.
5 Bottom Flange
0
-0.010 -0.005 0.0 0.005 0.010
(a) East Side
45
Top Flange S07
40
……........... ...............
35
30
Location (in.)
25
15 at 0.50-in. disp
at 0.75-in. disp.
5
Bottom Flange
0
-0.010 -0.005 0.0 0.005 0.010
(b) West Side
Figure B.29: Specimen CG4: Strain Profiles in Bearing Stiffener
344
B2. 3 Strain Plots in Steel Girder Flanges
0.004
0.003
0.002 S23…………………………........…S27
0.001
0.0
-0.001 at 0.15-in. disp.

-0.002 at 0.3-in. disp
at 0.35-in. disp.
at 0.4-in. disp.
-0.003 at 0.6-in. disp.
-0.004
0 5 10 15 20 25 30
Location (in.)
(a) East Side
0.004
S28…………………....…S30
0.003 at 0.15-in. disp.
0.002 at 0.3-in. disp
at 0.35-in. disp.
at 0.4-in. disp.
0.001 at 0.6-in. disp.
0.0
-0.001
-0.002
-0.003
-0.004
0 5 10 15 20 25 30
Location (in.)
(b) West Side
Figure B.30: Specimen SG1: Strain Profiles in Compression Flange (Top Flange)
345
0.004
0.003
0.002
S31…………………....…S33
0.001
0.0
-0.001 at 0.15-in. disp.

-0.002 at 0.3-in. disp
at 0.35-in. disp.
at 0.4-in. disp.
-0.003 at 0.6-in. disp.
-0.004
0 5 10 15 20 25 30
Location (in.)
(a) East Side
0.004
S34……………........…S36
0.003
0.002
0.001
0.0
-0.001 at 0.15-in. disp.

-0.002 at 0.3-in. disp
at 0.35-in. disp.
at 0.4-in. disp.
-0.003 at 0.6-in. disp.
-0.004
0 5 10 15 20 25 30
Location (in.)
(b) West Side
Figure B.31: Specimen SG1: Strain Profiles in Tension Flange (Bottom Flange)
346
0.005
0.004
0.002 S47…………………………........…S51
0.001
0.0
-0.001 at 0.15-in. disp.
at 0.3-in. disp
-0.003 at 0.35-in. disp.
at 0.4-in. disp.
-0.004 at 0.6-in. disp.
-0.005
0 5 10 15 20 25 30
Location (in.)
(a) East Side
0.005
0.004 at 0.15-in. disp.
0.003 at 0.3-in. disp
at 0.35-in. disp. S52……………….………..…....…S56
0.002 at 0.4-in. disp.

at 0.6-in. disp.
0.001
0.0
-0.001
-0.002
-0.003
-0.004
-0.005
0 5 10 15 20 25 30
Location (in.)
(b) West Side
347
0.005
0.004
0.002 S57…………………....…S59
0.001
0.0
-0.001 at 0.15-in. disp.
at 0.3-in. disp
-0.003 at 0.35-in. disp.
at 0.4-in. disp.
-0.004 at 0.6-in. disp.
-0.005
0 5 10 15 20 25 30
Location (in.)
(a) East Side
0.005
0.004
S60……………........…S62
0.003
0.002
0.001
0.0
-0.001 at 0.15-in. disp.
at 0.3-in. disp
-0.003 at 0.35-in. disp.
at 0.4-in. disp.
-0.004 at 0.6-in. disp.
-0.005
0 5 10 15 20 25 30
Location (in.)
(b) West Side
348
0.004
0.003
S47…………………………........…S51
0.002
0.001
0.0
-0.001 at 0.15-in. disp.

-0.002 at 0.4-in. disp
at 0.45-in. disp.
at 0.5-in. disp.
-0.003 at 0.6-in. disp.
-0.004
0 5 10 15 20 25 30
Location (in.)
(a) East Side
0.004
0.003 at 0.15-in. disp.

0.002 at 0.4-in. disp
at 0.45-in. disp.
0.001
0.0
-0.001
-0.002
S52……………….………..…....…S56
-0.003
-0.004
0 5 10 15 20 25 30
Location (in.)
(b) West Side
Figure B.34: Specimen CG1: Strain Profiles in Compression Flange (Top Flange)
349
0.004
0.003
0.002
S57…………………....…S59
0.001
0.0
-0.001 at 0.15-in. disp.

-0.002 at 0.4-in. disp
at 0.45-in. disp.
at 0.5-in. disp.
-0.003 at 0.6-in. disp.
-0.004
0 5 10 15 20 25 30
Location (in.)
(a) East Side
0.004
0.003
S60……………........…S62
0.002
0.001
0.0
-0.001 at 0.15-in. disp.

-0.002 at 0.4-in. disp
at 0.45-in. disp.
at 0.5-in. disp.
-0.003 at 0.6-in. disp.
-0.004
0 5 10 15 20 25 30
Location (in.)
(b) West Side
Figure B.35: Specimen CG1: Strain Profiles in Tension Flange (Bottom Flange)
350
0.004
0.003 S47…………………………........…S51
0.002
0.001
0.0
-0.001 at 0.15-in. disp.

-0.002 at 0.3-in. disp
at 0.35-in. disp.
at 0.4-in. disp.
-0.003 at 0.6-in. disp.
-0.004
0 5 10 15 20 25 30
Location (in.)
(a) East Side
0.004
0.003
0.002
MALFUNCTIONED
0.001
0.0
-0.001
-0.002
-0.003
-0.004
0 5 10 15 20 25 30
Location (in.)
(b) West Side
351
0.004
0.003
0.002
S57…………………....…S59
0.001
0.0
-0.001 at 0.15-in. disp.

-0.002 at 0.3-in. disp
at 0.35-in. disp.
at 0.4-in. disp.
-0.003 at 0.6-in. disp.
-0.004
0 5 10 15 20 25 30
Location (in.)
(a) East Side
0.004
0.003
S60……………........…S62
0.002
0.001
0.0
-0.001 at 0.15-in. disp.

-0.002 at 0.3-in. disp
at 0.35-in. disp.
at 0.4-in. disp.
-0.003 at 0.6-in. disp.
-0.004
0 5 10 15 20 25 30
Location (in.)
(b) West Side
352
0.010
0.008 at 0.15-in. disp.
0.006 at 0.30-in. disp
at 0.45-in. disp.
0.004 at 0.60-in. disp.

at 1.00-in. disp.
0.002
0.0
-0.002
-0.004
Bearing Intermediate
-0.006
Stiffener Stiffener
-0.008
-0.010
0 10 20 30 40 50 60 70
Location (in.)
(a) East Side
0.005
0.004 at 0.15-in. disp.
0.003 at 0.30-in. disp
at 0.45-in. disp.
0.002 at 0.60-in. disp.

at 1.00-in. disp.
0.001
0.0
-0.001
-0.002
-0.003
Stiffener Stiffener
-0.004
-0.005
0 10 20 30 40 50 60 70
Location (in.)
(b) West Side
353
0.0010 0.020
0.0005 0.015

0.0 0.010
-0.0005 0.005
-0.0010 0.0
0 500 1500 2500 0 500 1500 2500
Step Step
(a) S23 (b) S24
0.0010 0.0010
0.0008 0.0008
0.0006 0.0006
0.0004 0.0004
0.0002 0.0002
0.0 0.0
0 500 1500 2500 0 500 1500 2500
Step Step
(c) S25 (d) S26
354
0.020
at 0.15-in. disp.
0.015 at 0.25-in. disp.
at 0.75-in. disp
0.010 at 1.00-in. disp.
at 1.50-in. disp.
0.0
-0.005
-0.010 Bearing Intermediate

Stiffener Stiffener
-0.015
-0.020
0 10 20 30 40 50 60 70
Location (in.)
(a) East Side
0.020
at 0.15-in. disp.
0.015 at 0.25-in. disp.
at 0.75-in. disp
0.010 at 1.00-in. disp.
at 1.50-in. disp.
0.0
-0.005
-0.010 Intermediate Bearing

Stiffener Stiffener
-0.015
-0.020
0 10 20 30 40 50 60 70
Location (in.)
(b) West Side
355
0.002 0.04
0.001 0.03

0.0 0.02
-0.001 0.01
-0.002 0.0
0 500 1000 1500 2000 0 500 1000 1500 2000
Step Step
(a) S23 (b) S24
0.04 0.002
0.03 0.001
0.02 0.0
0.01 -0.001
0.0 -0.002
0 500 1000 1500 2000 0 500 1000 1500 2000
Step Step
(c) S25 (d) S26
356
0.010
0.008 at 0.15-in. disp.
at 0.25-in. disp.
0.006 at 0.50-in. disp
at 0.75-in. disp.
0.004 at 1.00-in. disp.

0.002
0.0
-0.002
-0.004
Bearing Intermediate
-0.006
Stiffener Stiffener
-0.008
-0.010
0 10 20 30 40 50 60 70
Location (in.)
(a) East Side
0.005
0.004 at 0.15-in. disp.
at 0.25-in. disp.
0.003 at 0.50-in. disp
at 0.75-in. disp.
0.002 at 1.00-in. disp.

0.001
0.0
-0.001
-0.002
-0.003
Stiffener Stiffener
-0.004
-0.005
0 10 20 30 40 50 60 70
Location (in.)
(b) West Side
357
0.0010 0.020
0.0005 0.015

0.0 0.010
-0.0005 0.005
-0.0010 0.0
0 400 800 1200 1600 0 400 800 1200 1600
Step Step
(a) S23 (b) S24
0.020 0.0010
0.015 0.0005
0.010 0.0
0.005 -0.0005
0.0 -0.0010
0 400 800 1200 1600 0 400 800 1200 1600
Step Step
(c) S25 (d) S26
358
0.020
at 0.15-in. disp.
0.015 at 0.25-in. disp.
at 0.50-in. disp
0.010 at 0.75-in. disp.
at 1.00-in. disp.
0.0
-0.005
-0.010 Bearing Intermediate

Stiffener Stiffener
-0.015
-0.020
0 10 20 30 40 50 60 70
Location (in.)
(a) East Side
0.020
at 0.15-in. disp.
0.015 at 0.25-in. disp.
at 0.50-in. disp
0.010 at 0.75-in. disp.
at 1.00-in. disp.
0.0
-0.005
-0.010 Intermediate Bearing

Stiffener Stiffener
-0.015
-0.020
0 10 20 30 40 50 60 70
Location (in.)
(b) West Side
359
0.002 0.04
0.001 0.03

0.0 0.02
-0.001 0.01
-0.002 0.0
0 500 1500 2500 0 500 1500 2500
Step Step
(a) S23 (b) S24
0.04 0.002
0.03 0.001
0.02 0.0
0.01 -0.001
0.0 -0.002
0 500 1500 2500 0 500 1500 2500
Step Step
(c) S25 (d) S26
360
B2. 4 Strain Plots in Concrete Deck
0.0002
S27.
S28
S29
S30
0.0001
Strain (in./in.)
0.0
-0.0001
-0.0002
0 200 400 600 800 1000 1200 1400 1600
Step
(a) S27 through S30
0.0002
S31.
S32 Malfunctioned S33
S34
0.0001
Strain (in./in.)
0.0
-0.0001
-0.0002
0 200 400 600 800 1000 1200 1400 1600
Step
(b) S31 through S34
Figure B.46: Specimen CG3
361
0.0002
S27.
S28
S29
S30
0.0001
Strain (in./in.)
0.0
-0.0001
-0.0002
0 500 1000 1500 2000 2500 3000
Step
(a) S27 through S30
0.0002
S31
S32
S33
S34
0.0001
Strain (in./in.)
0.0
-0.0001
-0.0002
0 500 1000 1500 2000 2500 3000
Step
(b) S31 through S34
Figure B.47: Specimen CG4
362

Tension Field

Uploaded by

Copyright:

Available Formats

Tension Field

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

Tension Field

Uploaded by

Copyright:

Available Formats

STATE OF CALIFORNIA • DEPARTMENT OF TRANSPORTATION

1. REPORT NUMBER 2. GOVERNMENT ASSOCIATION NUMBER 3. RECIPIENT'S CATALOG NUMBER

7. AUTHOR 8. PERFORMING ORGANIZATION REPORT NO.

Dong-Won Kim and Chia-Ming Uang SSRP-15/02

Department of Structural Engineering

Final Report to California Department of Transportation

Department of Structural Engineering

Report No. SSRP-15/02

SHEAR RESISTANCE OF END PANELS IN STEEL AND

Final Report to California Department of Transportation

Department of Structural Engineering

This document is disseminated in the interest of information exchange. The

The California Department of Transportation (Caltrans) funded this work under

LIST OF FIGURES .......................................................................................................ix

2. TEST PROGRAM (PHASE 1)............................................................................. 26

3. TEST RESULTS (PHASE 1) ............................................................................... 47

4. NONLINEAR FINITE ELEMENT ANALYSIS OF TEST SPECIMENS

5. FACTORS AFFECTING SHEAR RESISTANCE OF END PANELS: A

6. SHEAR RESISTANCE OF END PANELS: DERIVATION OF PREDICTIVE

7. REHABILITATION OF END PANELS............................................................ 174

8. TEST PROGRAM (PHASE 2)........................................................................... 207

9. ANALYTICAL STUDY OF TEST SPECIMENS (PHASE 2) ......................... 251

10. NUMERICAL PARAMETRIC STUDY............................................................ 266

11. RESISTANCE FACTOR FOR THE PROPOSED SHEAR STRENGTH

12. SUMMARY AND CONCLUSIONS ................................................................. 289

REFERENCES ........................................................................................................... 293

APPENDIX A. INSTRUMENTATION..................................................................... 298

APPENDIX B. PLOTS OF DEFORMATION AND STRAIN.................................. 316

Figure 1.1: Typical Configuration of Steel Plate Girder................................................... 17

1.1 Plate Girders

Typically in building construction, hot-rolled W-shape or I-shape members are

1.2 Shear Load-Carrying Mechanism of Stiffened Steel Plate Girder

1.3 Tension-Field Action in Interior Panels

1.3.2 Basler’s Theory

action can be determined by substituting Eq. (1.2) into Eq. (1.1).

Basler also derived the tension-field stress ( t ) by assuming that the

 y sin    cos    cr cos   0 (1.4)

y cos    sin    cr sin   0 (1.5)

Solving simultaneous Eqs. (1.4) and (1.5) gives

By substituting (  +90°) into y , normal stress x becomes   cr sin 2 .

y   cr sin 2   t (1.7)

(1.7) become principal stresses 1   cr   t , and  2   cr , respectively. Therefore,

post-buckling strength due to tension-field action is

1.3.3 Brief Review of Various Tension-Field Theories

1.4 US Plate Girder Shear Design Practice

1.4.2 Interior Panels

D = web depth (clear distance between flanges, in.),

tw = web thickness (in.),

b fc = width of compression flange (in.),

t fc = thickness of compression flange (in.),

t ft = thickness of tension flange (in.),

Fyw = web yield stress (ksi),

Vp = plastic shear force or shear yield strength (kips), and

C = ratio of the shear buckling strength to the shear yield strength.

E = modulus of elasticitity of steel

1.6 Statement of Problem

Based on Basler’s tension-field action theory, design shear strength equations

1.7 Shear Resistance in End Panels

from the flanges should be obtained as follows:

1.8 Steel-Concrete Composite Plate Girder