Inverted T Beams

1. Report No.
FHWA/TX-13/0-6416-1
Technical Report Documentation Page

2. Government
3. Recipients Catalog No.
Accession No.
4. Title and Subtitle

Strength and Serviceability Design of Reinforced Concrete
Inverted-T Beams
7. Author(s)
Nancy Larson, Eulalio Fernndez Gmez, David Garber,
Oguzhan Bayrak, and Wassim Ghannoum
5. Report Date
October 2012, Rev. March 2013, Published
June 2013
6. Performing Organization Code
8. Performing Organization Report No.
0-6416-1
9. Performing Organization Name and Address

Center for Transportation Research
The University of Texas at Austin
1616 Guadalupe, Suite 4.202
Austin, TX 78701
10. Work Unit No. (TRAIS)

11. Contract or Grant No.
0-6416
12. Sponsoring Agency Name and Address

Texas Department of Transportation
Research and Technology Implementation Office
P.O. Box 5080
Austin, TX 78763-5080
13. Type of Report and Period Covered

Technical Report 9/1/2009 8/31/2012
14. Sponsoring Agency Code
15. Supplementary Notes

Project performed in cooperation with the Texas Department of Transportation and the Federal Highway
Administration.
16. Abstract
Significant diagonal cracking in reinforced concrete inverted-T straddle bent caps has been reported
throughout the State of Texas. Many of the distressed structures were recently constructed and have generally
been in service for less than two decades. The unique nature of the problem prompted a closer look into the
design and behavior of such structural components. An experimental study was conducted in which 33
reinforced concrete inverted-T beam specimens were tested. The effects of the following variables were
evaluated: ledge depth and length, quantity of web reinforcement, number of point loads, member depth, and
a/d ratio. A strut-and-tie design method proposed by TxDOT Project 0-5253, initially calibrated for
compression-chord loaded deep beams, was investigated. It was concluded that the strut-and-tie method was a
simpler and accurate design method and was recommended for use in inverted-T beam design. A
recommendation was also made on the amount of minimum web reinforcement needed for strength and
serviceability considerations. A simple service-load check was proposed for the purpose of limiting diagonal
cracking under service loads. Lastly, a chart was created to aid in the distress evaluation of a diagonallycracked inverted-T bent cap in the field.
17. Key Words
18. Distribution Statement
Inverted-T Bent Caps, Strut-and-Tie Modeling, Crack
No restrictions. This document is available to the
Widths, Diagonal Cracking, Serviceability, Full-scale
public through the National Technical Information
Service, Springfield, Virginia 22161; www.ntis.gov.
19. Security Classif. (of report) 20. Security Classif. (of this page) 21. No. of pages
Unclassified
Unclassified
234
Form DOT F 1700.7 (8-72) Reproduction of completed page authorized
22. Price
ii
Strength and Serviceability Design of Reinforced Concrete

Inverted-T Beams
Nancy Larson
Eulalio Fernndez Gmez
David Garber
Oguzhan Bayrak
Wassim Ghannoum
CTR Technical Report:

Report Date:
Project:
Project Title:
Sponsoring Agency:
Performing Agency:
0-6416-1
October 2012, Rev. March 2013, Published June 2013
0-6416
Shear Cracking in Inverted-T Straddle Bents
Texas Department of Transportation
Center for Transportation Research at The University of Texas at Austin
Project performed in cooperation with the Texas Department of Transportation and the Federal Highway
Administration.
iii

1616 Guadalupe, Suite 4.202
Austin, TX 78701
www.utexas.edu/research/ctr
Copyright (c) 2013
All rights reserved
Printed in the United States of America
iv
Disclaimers
Author's Disclaimer: The contents of this report reflect the views of the authors, who
are responsible for the facts and the accuracy of the data presented herein. The contents do not
necessarily reflect the official view or policies of the Federal Highway Administration or the
Texas Department of Transportation (TxDOT). This report does not constitute a standard,
specification, or regulation.
Patent Disclaimer: There was no invention or discovery conceived or first actually
reduced to practice in the course of or under this contract, including any art, method, process,
machine manufacture, design or composition of matter, or any new useful improvement thereof,
or any variety of plant, which is or may be patentable under the patent laws of the United States
of America or any foreign country.
Engineering Disclaimer
NOT INTENDED FOR CONSTRUCTION, BIDDING, OR PERMIT PURPOSES.
Project Engineer: Oguzhan Bayrak
Professional Engineer License State and Number: Texas No. 106598
P. E. Designation: Research Supervisor
Acknowledgments
The authors are sincerely grateful to the Texas Department of Transportation (TxDOT) for
providing the funds to conduct this research study. The contributions of the project director
Jamie Farris (BRG) and the TxDOT project advisors including Courtney Holle (BRG), Dean
Van Landuyt (BRG), Glenn Yowell (ATL), Mike Stroope (LBB), Nicholas Nemec (BRG), and
Roger Lopez (HOU) are greatly appreciated.
vi
Table of Contents
CHAPTER 1. Introduction ...........................................................................................................1
1.1
Overview ..........................................................................................................................1
1.2
Project Objective ..............................................................................................................2
1.3
Project Scope ...................................................................................................................2
1.4
Organization .....................................................................................................................3
CHAPTER 2. Background of Inverted-T Straddle Bent Caps..................................................5
2.1
Overview ..........................................................................................................................5
2.2
Field Observations ...........................................................................................................5
2.3
Background on Inverted-T Bent Caps .............................................................................7
2.3.1 Tension-Chord vs. Compression-Chord Loaded Beams .............................................7
2.3.2 Components of an Inverted-T Beam ............................................................................8
2.4
Discontinuity Regions of Beams .....................................................................................9
2.5
Theoretical Background of Strut-and-Tie Modeling .....................................................10
2.5.1 Struts ..........................................................................................................................11
2.5.2 Ties.............................................................................................................................12
2.5.3 Nodal Zones ...............................................................................................................12
2.6
Inverted-T Design Provisions ........................................................................................12
2.6.1 AASHTO LRFD Bridge Design Specifications (2012).............................................13
2.6.2 TxDOT Project 0-5253 STM Provisions ...................................................................14
2.7
Inverted-T Deep Beam Database ...................................................................................26
2.7.1 Literature Review.......................................................................................................26
2.7.2 Filtered Database .......................................................................................................30
2.7.3 Evaluation Database...................................................................................................30
2.8
Summary ........................................................................................................................32
CHAPTER 3 Experimental Program ....................................................................................... 33
3.1
Overview ........................................................................................................................33
3.2
Testing Program .............................................................................................................33
3.2.1 Overview of Test Specimens .....................................................................................33
3.2.2 Series I: Ledge Length ...............................................................................................39
3.2.3 Series II: Ledge Depth ...............................................................................................40
3.2.4 Series III: Web Reinforcement Ratio .........................................................................42
3.2.5 Series IV: Number of Point Loads .............................................................................43
3.2.6 Series V: Web Depth .................................................................................................44
3.2.7 Series VI: Loaded Chord ...........................................................................................45
3.2.8 Summary of Test Specimen Details ...........................................................................46
3.3
Fabrication of Specimens ...............................................................................................48
3.3.1 Construction of Specimens ........................................................................................50
3.4
Test Setup.......................................................................................................................51
3.5
Instrumentation ..............................................................................................................52
3.5.1 Strain Measurements ..................................................................................................52
3.5.2 Load and Displacement Measurements .....................................................................54
3.5.3 Crack Width Measurements .......................................................................................56
3.6
Test Procedure ...............................................................................................................56
vii
3.7
Summary ........................................................................................................................57
CHAPTER 4 Experimental Results ...........................................................................................59

4.1
Overview ........................................................................................................................59
4.2
Summary of Experimental Results ................................................................................59
4.2.1 Evaluation of Strength Data .......................................................................................63
4.2.2 Evaluation of Serviceability Data ..............................................................................64
4.3
Series I: Ledge Length ...................................................................................................66
4.3.1 Experimental Results .................................................................................................66
4.3.2 Strength Results .........................................................................................................67
4.3.3 Serviceability Results.................................................................................................70
4.3.4 Summary- Series I: Ledge Length .............................................................................77
4.4
Series II: Ledge Depth ...................................................................................................77
4.4.4 Summary: Series II: Ledge Depth..............................................................................84
4.5
Series III: Web Reinforcement ......................................................................................85
4.5.4 Summary ....................................................................................................................94
4.6
Series IV: Number of Point Loads .................................................................................95
4.6.4 Summary ..................................................................................................................100
4.7
Series V: Web Depth ...................................................................................................101
4.7.1 Strength Results .......................................................................................................101
4.7.2 Serviceability Results...............................................................................................102
4.7.3 Summary ..................................................................................................................107
4.8
Series VI: Loaded Chord .............................................................................................107
4.8.1 Strength Results .......................................................................................................107
4.8.2 Serviceability Results...............................................................................................110
4.8.3 Summary ..................................................................................................................113
4.9
Summary of Experimental Results ..............................................................................113
CHAPTER 5 Analysis of Results ............................................................................................. 115
5.1
Overview ......................................................................................................................115
5.2
Strength Analysis .........................................................................................................115
5.2.1 Failure Modes ..........................................................................................................117
5.2.2 Ledge Design ...........................................................................................................119
5.2.3 Ultimate Strength .....................................................................................................121
5.2.4 Summary of Strength Results ..................................................................................140
5.3
Diagonal Cracking under Service Loads .....................................................................141
5.3.1 Background ..............................................................................................................142
5.3.2 Diagonal Cracking in Inverted-T Beams .................................................................143
5.3.3 Design Implications .................................................................................................151
viii
5.3.4 Summary and Conclusions ......................................................................................152

5.4
Correlation of Maximum In-Service Diagonal Crack Width with Ultimate Shear
Strength ....................................................................................................................................153
5.4.1 Background ..............................................................................................................153
5.4.2 Diagonal Crack Widths in Inverted-T Beams ..........................................................154
5.4.3 Summary and Conclusions ......................................................................................163
CHAPTER 6 In-Service Inverted-T Bridge Bents ................................................................. 165
6.1
Background ..................................................................................................................165
6.2
Inspection Reports .......................................................................................................165
6.2.1 Austin (TX-290 and I-35) ........................................................................................165
6.2.2 San Antonio (I-35 S) ................................................................................................170
6.2.3 El Paso (I-10 E Geronimo Drive Exit) .....................................................................172
6.2.4 Waco (TX-6 E and I-35N) .......................................................................................176
6.2.5 Findings from the Field Inspection ..........................................................................181
6.3
Serviceability Behavior ................................................................................................182
6.3.1 Diagonal Cracking under Service Loads .................................................................182
6.3.2 Maximum Diagonal Crack Widths ..........................................................................184
6.4
Summary ......................................................................................................................185
CHAPTER 7 Design Recommendations ................................................................................. 187
7.1
Introduction ..................................................................................................................187
7.2
Proposed Changes to TxDOT Design Specifications ..................................................187
7.2.1 General Design Recommendations ..........................................................................187
7.2.2 Strut-and-Tie Model Design Procedure ...................................................................188
7.2.3 TxDOT Report 5-5253 Design Examples................................................................190
CHAPTER 8 Summary and Conclusions ............................................................................... 193
8.1
Summary ......................................................................................................................193
8.2
Conclusions ..................................................................................................................194
8.2.1 Effects of Ledge Length...........................................................................................194
8.2.2 Effects of Ledge Depth ............................................................................................195
8.2.3 Effects of Web Reinforcement Ratio .......................................................................195
8.2.4 Effects of Multiple Loading Points ..........................................................................196
8.2.5 Effects of Web Depth...............................................................................................196
8.2.6 Effects of Tension Chord Loading ...........................................................................197
8.2.7 Proposed STM Design Provisions ...........................................................................197
8.2.8 Limiting Diagonal Cracking under Service Loads ..................................................198
8.2.9 Correlation of Maximum Diagonal Crack Width to Web-Shear Strength ..............198
8.3
Concluding Remarks ....................................................................................................198
References ..................................................................................................................................201
APPENDIX A. Design Example .............................................................................................. 205
A.1 Overview ...........................................................................................................................205
A.2 Defining the Beam ............................................................................................................207
A.3 Longitudinal Strut-and-Tie Model ....................................................................................208
A.4 Cross-Sectional Strut-and-Tie Model ...............................................................................215
ix
List of Figures
Figure 2-1: Severely distressed inverted-T bent cap in El Paso. .....................................................6
Figure 2-2: Left (a) rectangular bent cap, (b) inverted-T bent cap; right: flow path of forces
in strut-and-tie models: (c) Compression-chord loaded beam, (d) tension-chord
loaded beam (Fernandez 2012). ...........................................................................................8
Figure 2-3: Longitudinal elevation of an inverted-T bent cap with discontinuous ledges. .............8
Figure 2-4: Typical details of inverted-T bent caps. ........................................................................9
Figure 2-5: Stress trajectories within B- and D-regions (adapted from Birrcher, et al. 2009) ........9
Figure 2-6: Strut-and-tie model: Simply supported beam supporting a concentrated load
(adapted from Birrcher et al., 2009)...................................................................................10
Figure 2-7: Prismatic and bottle-shaped struts (adapted from Birrcher et al., 2009).....................11
Figure 2-8: Node designations. ......................................................................................................12
Figure 2-9: Node designations in an inverted-T beams .................................................................14
Figure 2-10: Geometry of a CCT node (adapted from Birrcher at al., 2009). ...............................15
Figure 2-11: CCT hanger node- (a) Original geometry of the STM; (b) adjacent struts
resolved together; (c) node divided into two parts; (d) final node geometry (adapted
from Williams et al., 2011). ...............................................................................................16
Figure 2-12: Geometry of a hanger CCT node (adapted from Birrcher et al., 2009) ....................17
Figure 2-13: Geometry of CTT node(adapted from Birrcher et al., 2009) ....................................18
Figure 2-14: Determination of CTT vertical tie .............................................................................18
Figure 2-15: Determination of A2 for stepped or sloped supports (from ACI 318-11) .................20
Figure 2-16: Available development length for ties (adapted from Birrcher et al., 2009) ............21
Figure 2-17: Strut-and-tie model of an inverted-T bent cap; top: tri-dimensional model,
center: cross-sectional models, bottom: longitudinal model (from Fernandez 2102) .......22
Figure 2-18: Hanger tie widths for beams with short and cut-off ledges ......................................23
Figure 2-19: Widths of compression and tension chords...............................................................23
Figure 2-20: Development of strut and tie model ..........................................................................24
Figure 2-21: Forces in cross-sectional models...............................................................................24
Figure 2-22: Load spread area for ledge reinforcement.................................................................25
Figure 2-23: STM for inverted-T test specimen ............................................................................26
Figure 2-24: Scaled cross-sections of literature review specimens and in-service bent caps. .......29
Figure 2-25: Sources of the inverted-T database. ..........................................................................31
Figure 3-1: Simplified strut-and-tie model showing elements limiting shear capacity .................34
Figure 3-2: Free body and shear diagram for a specimen subjected to three point loads ..............35
xi
Figure 3-3: Specimen nomenclature. .............................................................................................35

Figure 3-4: Typical specimen geometries ......................................................................................37
Figure 3-5: Typical reinforcement details......................................................................................38
Figure 3-6: Ledge lengths ..............................................................................................................39
Figure 3-7: Load spread in specimens with short and cut-off ledges ............................................40
Figure 3-8: Ledge length effect on support region. .......................................................................40
Figure 3-9: Ledge depths ...............................................................................................................41
Figure 3-10: Ledge depth to beam height ratios of cracked in-service bent caps..........................41
Figure 3-11: Load spread in specimens with deep and shallow ledges .........................................42
Figure 3-12: Web reinforcement ratios ..........................................................................................42
Figure 3-13: Inverted-T number of point loads .............................................................................43
Figure 3-14: Inclination angle of ledge strut in cross-sectional STM ...........................................44
Figure 3-15: Cross-section loading ................................................................................................45
Figure 3-16: Scaled comparison of actual bent caps and IT beams included in current and
past research programs. ......................................................................................................46
Figure 3-17: Fabrication of specimens ..........................................................................................50
Figure 3-18: Test setup ..................................................................................................................52
Figure 3-19: Typical location of strain gauges in longitudinal section .........................................53
Figure 3-20: Strain gauges on the hanger and ledge reinforcement ..............................................54
Figure 3-21: Steel strain gauge installation ...................................................................................54
Figure 3-22: Load cell arrangement at supports ............................................................................55
Figure 3-23: Deflection measurement locations ............................................................................55
Figure 3-24: Measuring crack widths with a comparator card ......................................................56
Figure 3-25: Three point loads, first and second test .....................................................................57
Figure 4-1: Load and shear force diagram for typical beam test. ..................................................63
Figure 4-2: Visual and experimental determination of first cracking load. ...................................64
Figure 4-3: Service load level estimation (Birrcher, Tuchscherer, et al., Strength and
Serviceability Design of Reinforced Concrete Deep Beams 2008) ...................................65
Figure 4-4: Typical crack width progression plot ..........................................................................66
Figure 4-5: Series I: Ledge Length- Direct comparisons of Vtest normalized by f'cbwd .............68
Figure 4-6: Series I: Ledge Length- Direct comparisons of Vtest normalized by
Figure 4-7: Series I: Ledge Length- Direct comparisons of Vcrack normalized by
bwd .............69
bwd ..........71
Figure 4-8: First diagonal crack comparison for ledge length specimens .....................................72
xii
Figure 4-9: Series I: Ledge Length- Direct comparisons of crack width progression ...................73
Figure 4-10: Crack patterns for D_3-42-1.85-03 (ledge length varies) .........................................75
Figure 4-11: Crack patterns and strain gauges for S_3-42-1.85-03 (ledge length varies) .............76
Figure 4-12: Series II: Ledge Depth- Direct comparisons of Vtest normalized by f'cbwd. ..........79
Figure 4-13: Series II: Ledge Depth- Direct comparisons of Vtest normalized by
Figure 4-14: Series II: Ledge Depth- Direct comparisons of Vcrack normalized by
bwd. ..........80
bwd. .......81
Figure 4-15: First diagonal crack comparison for ledge depth specimens ....................................82
Figure 4-16: Series II: Ledge Depth- Direct comparisons of crack width progression. ................83
Figure 4-17: Crack patterns and strain gauges for _C3-42-1.85-03 (ledge depth varies)..............84
Figure 4-18: Series III: Web Reinforcement- Direct comparisons of Vtest normalized by
f'cbwd. ................................................................................................................................87
Figure 4-19: Series III: Web Reinforcement- Comparisons of Vtest normalized by
bwd. .......88
Figure 4-20: Series III: Web Reinforcement- Direct comparisons of normalized Vcrack. ...........89
Figure 4-21: First diagonal crack comparison for reinforcement ratio specimens ........................90
Figure 4-22: Series III: Web Reinforcement- Direct comparisons of crack width
progression. ........................................................................................................................91
Figure 4-23: Diagonal crack widths for specimens tested at a/d of 1.85 .......................................92
Figure 4-24: Diagonal crack widths for specimens tested at a/d of 2.50 .......................................93
Figure 4-25: Crack patterns and strain gauges for SC3-42-1.85-_ (reinforcement varies) ...........94
Figure 4-26: Series IV: Number of Point Loads- Direct comparisons of Vtest normalized by
f'cbwd. ................................................................................................................................96
bwd. ................................................................................................................................97
Figure 4-28: Series IV: Number of Point Loads- Direct comparisons of Vcrack. ........................98
Figure 4-29: First diagonal crack comparison for number of point load specimens .....................98
Figure 4-30: Series IV: Number of Point Loads- Direct comparisons of crack width
progression. ........................................................................................................................99
Figure 4-31: Crack patterns for SS_-42-1.85-03 (number of point loads varies) ........................100
Figure 4-32: Series VI: Web Depth- Direct comparisons of Vtest normalized by f'cbwd. .........102
Figure 4-33: Series VI: Web Depth- Direct comparisons of Vtest normalized by
Figure 4-34: Series VI: Web Depth- Direct comparisons of Vcrack normalized by
bwd. .........102
bwd. .....103
Figure 4-35: First diagonal crack comparison for web depth specimens ....................................103
Figure 4-36: Series VI: Web Depth- Direct comparisons of crack width progression. ...............104
Figure 4-37: Diagonal crack widths for specimens with 0.3% reinforcement both directions....105
xiii
Figure 4-38: Crack patterns for SS1-_-1.85-03 (web depth varies).............................................106

Figure 4-39: Series VI: Loaded Chord- Direct comparisons of Vtest normalized by f'cbwd. ....109
Figure 4-40: Series VI: Loaded Chord- Direct comparisons of Vtest normalized by
Figure 4-41: Series VI: Loaded Chord Comparison of Vcrack normalized by
bwd. ....110
bwd. ...........111
Figure 4-42: Series V: Loaded Chord- Direct comparisons of crack width progression.............112
Figure 4-43: Series V: Loaded Chord- General comparisons of crack width progression. .........113
Figure 5-1: Failure modes- (A) SS3-42-2.50-06 flexure, (B) DL3-42-1.85-03 flexure, (C)
SC1-42-2.50-03 shear friction, (D) SC1-42-1.85-03 ledge tie yielding, and (E) SS175-2.50-03 punching shear...............................................................................................117
Figure 5-2: Shear failure: concrete crushing for a/d = 1.85 and stirrup yielding for a/d =2.50 ...119
Figure 5-3: Inverted-T Cross Section STM and Strain Gauges ...................................................119
Figure 5-4: Load Spread for Cross Section Design .....................................................................120
Figure 5-5: Typical Hanger Stains ...............................................................................................120
Figure 5-6: Series I: Ledge Length- Direct comparisons of experimental capacity with
TxDOT Project 0-5253 STM calculations .......................................................................123
Figure 5-7: Series I: Ledge Length- STM strength predictions. ..................................................125
Figure 5-8: Series II: Ledge Depth- Direct comparisons of experimental capacity with
Figure 5-9: Series II: Ledge Depth- STM and LRFD strength predictions .................................128
Figure 5-10: Series III: Web Reinforcement- Direct comparisons of experimental capacity
with TxDOT Project 0-5253 STM calculations ...............................................................130
Figure 5-11: Series III: We Reinforcement Ratio- STM capacity results ...................................132
Figure 5-12: Series IV: Number of Point Loads- Direct comparisons of experimental
capacity with TxDOT Project 0-5253 STM calculations. ...............................................134
Figure 5-13: Series IV: Number of Point Loads- STM strength predictions ...............................135
Figure 5-14: Series V: Web Depth- Direct comparisons of experimental capacity with
Figure 5-15: Series VI: Loaded Chord- Direct comparisons of experimental capacity with
TxDOT Project 0-5253 STM calculations. ......................................................................138
Figure 5-16: Series VI: Loaded Chord- STM conservatism ........................................................140
Figure 5-17: Conservatism of STM provisions as applied to inverted-T beams .........................141
Figure 5-18: Types of cracks in reinforced concrete inverted-T deep beams..............................142
Figure 5-19: Effect of section size on diagonal cracking load of inverted-T beams. ..................144
Figure 5-20: Effect of tensile strength on diagonal cracking load of inverted-T beams. ............145
Figure 5-21: Effect of a/d ratio on diagonal cracking load of inverted-T beams. .......................146
xiv
Figure 5-22: Effect of depth on the diagonal cracking load of beams of inverted-T beams. ......146
Figure 5-23: Effect of web reinforcement on the diagonal cracking load of directly
comparable inverted-T specimens. ..................................................................................147
Figure 5-24: Effect of web reinforcement on diagonal cracking load of inverted-T beams........147
Figure 5-25: Effect of ledge length on the diagonal cracking load of inverted-T beams. ...........148
Figure 5-26: Effect of ledge depth on the diagonal cracking load of inverted-T beams. ............149
Figure 5-27: Effect of multiple point loads on the diagonal cracking load of inverted-T
beams. ..............................................................................................................................149
Figure 5-28: Assessment of proposed equation for estimate of diagonal cracking. ....................151
Figure 5-29: Effect of web reinforcement on diagonal crack widths of test specimens. .............157
Figure 5-30: All crack width data at an a/d ratio of 1.85 used in this task with trend lines. .......159
Figure 5-31: All crack width data at an a/d ratio of 2.50 used in this task with trend lines. .......160
Figure 5-32: All crack width data at an a/d ratio of 1.85 with straight line approximations. ......160
Figure 5-33: All crack width data at an a/d ratio of 2.50 with straight line approximations. ......161
Figure 5-34: All crack width data with straight line approximations. .........................................161
Figure 5-35: Proposed chart that links diagonal crack width to percent of ultimate capacity
of inverted-T bent caps. ...................................................................................................162
Figure 6-1: Location of inverted-T straddle bent caps in Austin (Mapquest) .............................165
Figure 6-2: Plan View of Austin Bent 6K ...................................................................................167
Figure 6-3 Photograph and Sketch of Northwest Corner of Austin Bent 6K ..............................167
Figure 6-4: Plan View of Austin Bent 3M ...................................................................................168
Figure 6-5: Photograph and Sketch of Southwest corner of Austin Bent 3M .............................168
Figure 6-6: Plan View of Austin Bent 28K .................................................................................169
Figure 6-7 Photograph and Sketch of Northwest Corner of Austin Bent 28K ............................169
Figure 6-8: Crack size and location on the northeast corner of Austin bent 28...........................170
Figure 6-9: Location of inverted-T straddle bent cap in San Antonio (Mapquest) .....................170
Figure 6-10: Plan View of San Antonio Bent Cap ......................................................................171
Figure 6-11: Crack Size and Location on the Northwest Corner of the San Antonio Bent .........171
Figure 6-12: Crack Size and Location on the Southwest Corner of the San Antonio Bent .........172
Figure 6-13: Crack size and location on the southeast corner of the San Antonio bent ..............172
Figure 6-14: Location of inverted-T straddle bent caps in El Paso (Mapquest) ..........................173
Figure 6-15: Plan View of El Paso Bent 4 ...................................................................................174
Figure 6-16: Crack Size and Location on the Southwest Corner of El Paso Bent 4 ....................174
xv
Figure 6-17: Crack Size and Location on the Southeast Corner of El Paso Bent 4 .....................175
Figure 6-18: Plan View of El Paso Bent 5 ...................................................................................175
Figure 6-19: Crack Size and Location on the Northeast Corner of El Paso Bent 5 .....................176
Figure 6-20: Crack Size and Location on the Northwest Corner of El Paso Bent 5 ....................176
Figure 6-21: Location of inverted-T Straddle Bent Caps in Waco (Mapquest) ..........................177
Figure 6-22: Plan View of Waco Bent 17 ....................................................................................177
Figure 6-23: Scraping off Efflorescence and Measuring Diagonal Cracks .................................178
Figure 6-24: Crack Size and Location on the Southwest Corner of Waco Bent 17 ....................178
Figure 6-25: Crack Size and Location on the Northwest Corner of Waco Bent 17 ....................179
Figure 6-26: Plan View of Waco Bent 19 ....................................................................................179
Figure 6-27: Crack Size and Location on the Northwest Corner of Waco Bent 19 ....................179
Figure 6-28: Crack Size and Location on the Southwest Corner of Waco Bent 19 ....................180
Figure 6-29: Crack Size and Location on the Southeast Corner of Waco Bent 19 .....................180
Figure 6-30: Crack Size and Location on the Northeast Corner of Waco Bent 19 .....................180
Figure 6-31: Assessment of in-service bent caps with the proposed equation for diagonal
cracking estimation ..........................................................................................................183
Figure 7-1: D-regions in inverted-T bent caps ............................................................................ 188
Figure 7-2: Strut-and-tie model design procedure (adapted from TxDOT Report 5-5253) ....... 190
Figure 7-3: Illustration of struts and nodes within the moment frame inverted-T bent cap
(Williams 2011). ............................................................................................................. 191
Figure 7-4: US-59/ N W. Little York Bent #4, Houston Texas .................................................. 192
Figure A-1: Beam DL1-42-1.85-03 ............................................................................................ 205
Figure A-2: Elevation and cross-sectional details of DL1-42-1.85-03. ...................................... 206
Figure A-3: Strut-and-tie model for DS1-42-1.85-03 ................................................................. 209
Figure A-4: Geometry of CCT Node B ...................................................................................... 210
Figure A-5: Strut and node notation ........................................................................................... 211
Figure A-6: Cross-Sectional Strut-and-Tie Model ..................................................................... 215
xvi
List of Tables
Table 2-1: Crack width summary of in-service bent caps. ..............................................................7
Table 2-2: TxDOT Project 5-5253-01 concrete efficiency factors, v ............................................20
Table 2-3: Inverted-T database assembly. .....................................................................................32
Table 3-1: Variables in Testing Program. ......................................................................................36
Table 3-2: Summary of beam details .............................................................................................47
Table 3-3: Typical concrete mixture properties .............................................................................48
Table 3-4: Specimen material strengths .........................................................................................49
Table 4-1: Summary of experimental results .................................................................................61
Table 4-2: Series I: Ledge Length- Experimental results. .............................................................67
Table 4-3: Series II: Ledge Depth- Experimental results. .............................................................78
Table 4-4: Series III: Web Reinforcement- Experimental results. ................................................86
Table 4-5: Series IV: Number of Point Loads- Experimental results. ...........................................95
Table 4-6: Series VI: Web Depth- Experimental results. ............................................................101
Table 4-7: Series VI: Loaded Chord- Experimental results.........................................................108
Table 5-1: Vtest/Vcalc results for 5253 STM provisions. ...........................................................116
Table 5-2: Overall accuracy of TxDOT Project 0-5253 STM provisions ...................................121
Table 5-3: Series I: Ledge Length ...............................................................................................122
Table 5-4: Series I: Ledge Length- STM summary by a/d ratio. .................................................124
Table 5-5: Series II: Ledge Depth ................................................................................................126
Table 5-6: Series II: Ledge Depth- STM summary by a/d ratio. .................................................128
Table 5-7: Series III: Web Reinforcement Ratio .........................................................................129
Table 5-8: Series III: Web Reinforcement Ratio- STM summary...............................................131
Table 5-9: Series IV: Number of Point Loads .............................................................................133
Table 5-10: Series IV: Number of Point Loads- STM summary. ................................................134
Table 5-11: Series V: Web Depth ................................................................................................136
Table 5-12: Series VI: Loaded Chord ..........................................................................................137
Table 5-13: Series VI: Loaded Chord STM summary. ................................................................139
Table 5-14: Summary of experimental/calculated shear capacity ...............................................141
Table 5-15: Specimens used in correlating crack width to capacity. ...........................................156
Table 6-1: Important characteristics of Austin straddle bents (6K, 3M, 28K) ............................166
Table 6-2: Important Characteristics of San Antonio Straddle Bents ..........................................171
xvii
Table 6-3: Important Characteristics of El Paso Straddle Bents..................................................173

Table 6-4: Important Characteristics of Waco Straddle Bents ....................................................177
Table 6-5: Inverted-T Crack Width Summary .............................................................................181
Table 6-6: Estimated percent of ultimate capacity for in-service bent caps ................................184
xviii
CHAPTER 1
Introduction
1.1
OVERVIEW
Diagonal web cracking of recently built inverted-T straddle bent caps has been reported
with increasing frequency in Texas, triggering concerns about current design procedures
for such elements. To address these concerns, the Texas Department of Transportation
(TxDOT) funded Project 0-6416 with the objectives of obtaining a better understanding
of the behavior of inverted-T beams and developing strength and serviceability design
criteria that will minimize such cracking in the future.
Inverted-T straddle bent caps are beam elements with girder loads applied to ledges at the
bottom of the section (bottom- or tension-chord loading). The loads are transferred in the
transverse direction from the ledges to the bottom of the web, then vertically to the
compression chord, and finally in the longitudinal direction to the supports. This threedimensional flow of forces, in addition to the deep beam loading conditions commonly
encountered in bent caps, generate regions of stress discontinuities that have been
traditionally designed using empirical equations. In the past two decades, US structural
design codes have adopted strut-and-tie modeling as a more rational option for the design
of deep beams and other structures with discontinuities like those present in the invertedT bent caps.
Most inverted-T bent caps in Texas are designed using the traditional empirical
procedures outlined in the TxDOT Bridge Design Manual LRFD (2011) that follows
closely the AASHTO LRFD Bridge Design Specifications (2012). Given the observed
cracking in inverted-T bent caps, it was the intent of this study to investigate the
applicability of STM procedures developed for rectangular deep beams to improve the
design of inverted-T bent caps. The TxDOT Project 0-5253 provisions (Strength and
Serviceability Design of Reinforced Concrete Deep Beams) provided several
improvements on the AASHTO (2012) STM procedures that, if found to be accurate and
conservative, could be readily implemented to the design of inverted-T bent caps.
Due to scarcity of experimental investigations on inverted-T beams, a comprehensive
experimental program was undertaken to examine the behavior of such structural
elements and assess the accuracy and validity of implementing the STM design. Thirty
three specimens were fabricated and tested as part of the current research program.
Unlike those found in the literature, the test specimens in the current project were
considered more representative of inverted-T beams designed in practice in terms of their
size and reinforcement details.
1.2
PROJECT OBJECTIVE
This research report was funded by the Texas Department of Transportation (TxDOT).
Since the inclusion of strut-and-tie modeling (STM) provisions in the AASHTO LRFD
specifications in 1994, TxDOT engineers have been examining the impact that the
provisions have on the design of bent caps. A recently completed research project,
TxDOT Project 0-5253, examined the application of strut-and-tie models to the design of
rectangular bent caps. From that project, recommendations were made to the AASHTO
LRFD specifications to improve the strength and serviceability behavior of bent caps. A
subsequent investigation, TxDOT Report 5-5253, provided additional recommendations
and STM design examples for several structural bridge components.
The current project aims to evaluate the applicability of the recommended provisions,
which were calibrated for compression-chord loaded beams, to inverted-T, or tensionchord loaded beams. An extensive experimental program was conducted to reveal what,
if any, changes will be required for the proposed strut-and-tie modeling provisions for
inverted-T bent caps. Field inspections of distressed inverted-T bent caps provided the
basis for the test variables investigated in the experimental program.
1.3
PROJECT SCOPE
In order to accomplish the objectives mentioned above, the following tasks are addressed
in TxDOT Project 0-6416:
1. Conduct a comprehensive literature review to expose the current state of
knowledge on inverted-T beams.
2. Conduct a detailed assessment of the condition of distressed in-service bent
caps
3. Conduct experimental investigations to uncover the main factors affecting the
web diagonal cracking behavior of inverted-T beams:
i.
Determine the influence that the length of the ledge has on the strength
and serviceability behavior of an inverted-T beam (Section 4.3).
ii.
Determine the influence that the depth of the ledge has on the strength
iii.
Determine the influence that the web reinforcement ratio has on the
strength and serviceability behavior of an inverted-T beam (Section
4.5).
iv.
Determine the influence that the number of point loads has on the
strength and serviceability behavior of an inverted-T beam (Section
4.6).
v.
Determine the influence that the depth of the web has on the strength
vi.
Determine the effect of tension-chord loading on the strength and

serviceability of deep beams (Section 4.8).
4. Make a recommendation on the application of TxDOT Project 0-5253 STM

provisions for the design of inverted-T bent caps (Section 5.2).
5. Make a recommendation on the feasibility of limiting diagonal cracking under
service loads (Section 5.3).
6. Make a recommendation for relating the maximum diagonal crack width of an
inverted-T beam to its residual capacity (Section 5.4).
Assembly of an inverted-T database produced 130 test results from the literature and the
current project. However, most of the tests were either not applicable to the inclined
cracking focus of this project or conducted on beams drastically smaller than the bent
caps in service in Texas. Moreover, very limited serviceability information regarding
diagonal crack widths was available in the literature. It was therefore deemed necessary
to conduct a comprehensive experimental program of full-scale inverted-T beam
specimens to achieve project goals.
Thirty three full-scale tests were conducted with several specimens measuring among the
largest reinforced concrete deep beams ever tested to determine shear capacity. Based on
the results of the experimental series treated in this project, design recommendations for
strength of inverted-T beams were developed and are presented in this report.
Serviceability criteria for minimizing diagonal cracking in inverted-T beams under
service loads were developed based on the test results. A minimum reinforcement ratio
required to restrain diagonal crack widths is also recommended. Finally a chart designed
to correlate the crack width with inverted-T beam residual capacity was developed.
1.4
ORGANIZATION
A general background of the design and behavior of inverted-T bent caps is presented in
Chapter 2. An overview of the distressed bent caps in service is also presented along with
the assembly of the inverted-T database from the literature.
In Chapter 3, the experimental program is described in detail. An overview of the
specimens, a description of the six experimental series, and the fabrication of the
specimens is presented. The test setup and instrumentation are described and the overall
test procedure is outlined.
Experimental results are presented in Chapter 4. Criteria for strength and serviceability
evaluation are detailed. Comparisons of strength, diagonal cracking load, and the crack
width progression are presented for the experimental variables covered in this project.
Analysis of the experimental results is provided in Chapter 5. A discussion of the failure

modes is provided along with an assessment of the accuracy of the STM design method.
The serviceability design considerations are evaluated in terms of the sizing the member
to prevent or limit diagonal cracking and providing minimum reinforcement to restrain
crack widths.
Descriptions of the in-service inverted-T bent caps inspected as part of this project are
summarized in Chapter 6. An overview of the design recommendations are presented in
Chapter 7. All of the findings and conclusions of the research program are summarized in
Chapter 8.
CHAPTER 2
Background of Inverted-T Straddle Bent Caps
2.1
OVERVIEW
Included in this Chapter is an overview of the theoretical background of inverted-T beam

design and behavior. A summary of the distressed in-service inverted-T bent caps are
presented to frame the problem observed in Texas bridges. All cap inspections are
described in detail in Chapter 6. Background information on inverted-T beams and strutand-tie modeling is provided to familiarize the reader. The AASHTO LRFD Bridge
Design Specifications and TxDOT Bridge Design Manual are discussed and the proposed
TxDOT 0-5253 STM provisions are summarized for the design of inverted-T beams.
Lastly, the inverted-T deep beam database is presented along with the filtering process
implemented to remove specimens that were not considered to be representative of the
field structures.
2.2
FIELD OBSERVATIONS
Several recently built Texas inverted-T bent caps have shown significant inclined
cracking triggering concern about current design procedures for such structures. For this
reason TxDOT funded Project 0-6416 to obtain a better understanding of the structural
behavior of inverted-T bent caps and develop new design criteria to minimize/eliminate
such cracking in the future.
One of the tasks of the current project was to conduct a thorough inspection of eight inservice bent caps that had evidence of significant diagonal cracking. In general, the
measured crack widths were small ( 0.016 in.) posing only aesthetic and durability
concerns. Several of the bents were much more severely distressed with diagonal crack
widths measuring up to 0.040 in. (Figure 2-1). In all cases, the observed cracking patterns
on both faces of the distressed bent caps were symmetric about the longitudinal axis of
the beams, indicating shear issues rather than torsional problems. While cracking is
expected in reinforced concrete, the crack widths observed in some caps suggest
structural deficiencies or overloading, prompting the field inspections.
Figure 2-1: Severely distressed inverted-T bent cap in El Paso.
A summary of the results obtained from inspections of the bent caps is provided in Table
2-1. The vertical reinforcement ratio, v, is evaluated as the cross-sectional area of the
vertical reinforcement (stirrups) divided by the width of the section times the spacing of
the stirrups. Similarly, the horizontal reinforcement ratio, h, is the ratio of the horizontal
skin reinforcement to the web width times the skin spacing. The shear span-to-depth
ratio, a/d, is evaluated as the distance from the center of the support to the closest load
point divided by the effective depth of the section. The ledge length was determined by
examining the amount the ledge was continued past the exterior bearing pad; with cutoffs terminated right at the bearing, short lengths extended a distance at least equal to the
depth of the ledge, and long lengths extending to the support. More information on these
variables is provided in Chapter 3. As shown in the table, the maximum diagonal crack
widths of the inspected bent caps varied between 0.010 and 0.040 in. The field
inspections are discussed in depth in Chapter 6.
Table 2-1: Crack width summary of in-service bent caps.
Bent
Austin IH-35 / Tx-290
Bent 3M
Bent 6K
Bent 28K
San Antonio IH-35 S
Exit 165
El Paso IH-10 /
Geronimo Bent 4
El Paso IH-10 /
Geronimo Bent 5
Waco IH-35 / LP340
Bent 17
Waco IH-35 / LP340
Bent 19
a/d
Ledge Ledge Height / Max Diagonal

Length Cap Height
Crack Width
0.43% 0.37%
1.4
Short
35%
0.02 in.
0.43% 0.37%
1.7
Short
35%
0.016 in.
0.43% 0.37%
1.4
Short
35%
0.03 in.
Long
33%*
0.015 in.
Not available 1.85*

0.57% 0.19%
1.7
Cut-Off
29%
0.04 in.
0.57% 0.19%
3.4
Cut-Off
29%
0.02 in.
0.46% 0.30%
2.5
Short
36%
0.01 in.
0.46% 0.30%
2.5
Short
36%
0.015 in.
* Approximated based on field observations

2.3
BACKGROUND ON INVERTED-T BENT CAPS
In this section, a background on inverted-T beam behavior is given. The difference

between tension-chord loaded beams and typical reinforced concrete deep beams is
examined and the unique characteristics of inverted-T beams are discussed.
2.3.1 Tension-Chord vs. Compression-Chord Loaded Beams
Inverted-T bent caps are often used in bridge construction to reduce the elevation of
bridges and/or to improve available clearance beneath the beams. The bridge girders are
supported on ledges near the bottom of the beam, effectively loading the cap along its
tension chord as shown in Figure 2-2. This arrangement generates a tension field in the
web at the loading points as forces are hung from the tension chord to the top of the
beam. In contrast, compression-chord loaded beams are not subjected to such
concentrated tension fields in the web as the load is applied to the top of the beam.
Elevation
reduction
Clear
height
a)
b)
c)
d)
Figure 2-2: Left (a) rectangular bent cap, (b) inverted-T bent cap; right: flow path of forces
in strut-and-tie models: (c) Compression-chord loaded beam, (d) tension-chord loaded
beam (Fernandez 2012).
2.3.2 Components of an Inverted-T Beam

Inverted-T beams have two main components, as shown in Figure 2-3. Load is applied to
the ledges of the beam, and is then transferred into the stem or web. The web then carries
the shear and flexure forces to the supports.
Ledge
Web (Stem)
Shear Span
Inverted-T
Cross-Section
Rectangular
Cross-Section
Figure 2-3: Longitudinal elevation of an inverted-T bent cap with discontinuous ledges.
Two additional types of reinforcement are required in an inverted-T beam compared to

that of the typical reinforcement layout of a rectangular beam. Ledge reinforcement is
necessary to resist flexural tension forces in the cantilevered ledge and allow the applied
load to flow into the web of the beam. Additional vertical bars, or hanger reinforcement,
are then used to hang the load transferred from the ledge at the bottom of the beam to
the compression-chord at the top. The reinforcement layout for inverted-T beams is
illustrated in Figure 2-4.
Ledge
Reinforcement
Compression
Reinforcement
Web
Web (Stem)
Hanger
Reinforcement
Stirrups
Skin
Reinforcement
Tension
Reinforcement
Ledge
Figure 2-4: Typical details of inverted-T bent caps.
2.4
DISCONTINUITY REGIONS OF BEAMS
Typically, reinforced concrete beams are designed based on the assumption that plane
sections remain plane; referred to as the Bernoulli hypothesis or beam theory. Within that
theory, the strains in the beam are presumed to vary linearly at a section and the beam is
said to be dominated by sectional behavior. As shown in Figure 2-5, these regions of
linear stress (or strain) are referred to as B-regions (with the B standing for beam or
Bernoulli).
P
0.29P
d
D-region
d
d
d
D-region D-region D-region
3d
B-region
0.71P
Figure 2-5: Stress trajectories within B- and D-regions (adapted from Birrcher, et al. 2009)
A D-region (with D standing for discontinuity or disturbed) can typically be found on

either side of a B-region. These regions cannot be designed using the sectional
procedures because the assumptions used to derive the beam theory are no longer valid.
As illustrated in Figure 2-5, the strains are distributed nonlinearly throughout the Dregions. These disturbances are caused either by abrupt changes in geometry or loading.
Frame corners, dapped ends, openings, and ledges are examples of geometric
discontinuities. Point loads such as girder bearings or support reactions also result in
nonlinear strains. According to St. Venants Principle, an elastic stress analysis would
indicate that a disturbance dissipates at about one member depth away from the
discontinuity. In other words, a D-region is assumed to extend one member depth, d,

from the load or geometric discontinuity.
A deep beam is one in which the entire span is assumed to be dominated by nonlinear
behavior. For this to be true, the shear span, a, must be less than about 2 to 2.5 times the
member depth, d. The right shear span in Figure 2-5 is entirely composed of D-regions
and is thus considered a deep beam. A beam with a greater a/d ratio, as shown in the left
span, is assumed to behave according to beam theory and can be designed using the
sectional procedures. Deep beams, however, require the use of non-sectional design
procedure such as the strut-and-tie modeling procedure discussed below.
2.5
THEORETICAL BACKGROUND OF STRUT-AND-TIE MODELING
Strut-and-tie modeling (STM) offers a rational approach for obtaining lower-bound

solutions for the strength design of deep beams. In the models, the complex state of
stresses in a member is idealized as a system of uniaxial force elements acting as a truss
within the concrete member as shown in Figure 2-6. This system will yield a conservative
design if the resulting truss model is in equilibrium with the external forces and the
concrete has enough deformation capacity to accommodate the assumed distribution of
forces (Schlaich et al. 1987). The compressive forces must not exceed the factored
concrete strengths and, likewise, the tensile forces must not exceed the factored tie
capacities. It is also crucial that proper anchorage of the reinforcement is achieved.
Node
Tie
Strut
Figure 2-6: Strut-and-tie model: Simply supported beam supporting a concentrated load
(adapted from Birrcher et al., 2009).
Strut-and-tie models consist of three components: struts, ties, and nodes assembled
together to represent the flow of forces through a structure, as shown in the simply
supported beam illustrated in Figure 2-6. After calculating the external reactions and
defining the geometry of the STM, the individual member forces of the truss are
determined through statics. The compression members are referred to as struts, the
tension members as ties, and the regions in which they intersect are nodes.
10
The versatility of strut-and-tie modeling allows it to be used to design of any D-region

and accommodate various load transfer mechanisms. Its adaptability is a significant
advantage but can also pose major design challenges as no one correct STM exists for
any particular structure. As long as the principles required to achieve a lower-bound
solution are met, any model can be considered a safe design. A model should, to the best
extent possible, follow the actual stress field as determined by an elastic stress analysis. If
the model varies substantially from the structures stress field, the structure will undergo
substantial deformations leading to an increased chance of cracking. Many examples of
strut-and-tie modeling can be found in the literature, including TxDOT Report 5-5253
(Williams et al., 2011) which aides engineers in developing strut-and-tie models for
structural components of highway bridges. For further explanation of the theoretical
background of STM, the reader is encouraged to reference the TxDOT Project 0-5253
report (Birrcher et al., 2009).
2.5.1 Struts
Struts are compression elements that vary in shape depending on their location within a
structure. Idealized as dashed lines, struts can be bottle-shaped if allowed to spread along
their length, or prismatic in regions of uniform stresses such as the compression zone of a
beams flexural region. Often bottle-shaped struts are idealized as prismatic struts as
shown in Figure 2-7. It is important to provide transverse reinforcement in the vicinity of
these diagonal bottle-shaped struts to carry the tensile forces, strengthen the strut, and
control the bursting cracks that develop with the spreading.
Prismatic Strut
Bottle-Shaped Strut
Tension
Develops
Idealized Prismatic Strut

Figure 2-7: Prismatic and bottle-shaped struts (adapted from Birrcher et al., 2009)
11
2.5.2 Ties
STM tension elements, or ties, are generally made up of reinforcing steel and denoted by
solid lines in Figure 2-7 and throughout this report. Enough reinforcement must be
provided to carry the tensile demand of the tie and should be distributed so that its
centroid coincides with the tie location. Details such as bar spacing and anchorage need
to be considered for proper strut-and-tie modeling.
2.5.3 Nodal Zones
Due to the concentration of stresses from intersecting truss members, the nodes are the
most highly stressed regions of a structural member. Three types of nodes can exist
within an STM and are named based on the elements framing into them, as shown in
Figure 2-8. Within the nodal designations, C refers to a compression element, such as a
strut or an externally applied load, and T stands for tension, such as a tie.
CCC: nodes where only struts intersect

CCT: nodes where tie(s) intersect in only one direction
CTT: nodes where ties intersect in two different direction
Struts are often resolved together to reduce the number of members intersecting at a
node, such as the CCC node illustrated in Figure 2-8. The type of node governs its
behavior and thus its strength. Additional information on strut-and-tie modeling for
inverted-T design is provided in Section 2.6.2.
CCC Node
CCT Node
CTT Node
Figure 2-8: Node designations.
2.6
INVERTED-T DESIGN PROVISIONS
In this section, the current design provisions for inverted-T beams in AASHTO LRFD
Bridge Design Specifications (2012) and the TxDOT Bridge Design Manual-LRFD
(2011) are discussed. The design procedure used in the experimental portion of this
project, the TxDOT Project 0-5253 STM provisions, is also presented.
12
2.6.1 AASHTO LRFD Bridge Design Specifications (2012)

AASHTO LRFD specifies two design methods for inverted-T beams: Strut-and-tie
modeling of the beam or a series of separate design provisions for the web portion and
the ledge portion.
The TxDOT Design Manual utilized a slightly modified version of the second method.
For the web portion of inverted-T beams, sectional design provisions apply. To design
the ledge portion, a set of checks outlined in the provisions of Articles 5.13.2.5.2 through
5.13.2.5.5 are used. These provisions focus on designing beam ledges to resist forces in
various location of the ledge, including flexure, shear and horizontal forces at the ledge to
web interface, tension force in the top of the ledge, punching shear through the ledge, and
bearing forces.
If the shear span-to-depth ratio of a beam is less than about 2.0, the AASHTO Code
specifies that strut-and-tie modeling should be considered. AASHTO (2012) Clause
5.6.3.1 specifies: The strut-and-tie model should be considered for the design of deep
footings and pile caps or other situations in which the distance between the centers of
applied load and the supporting reactions is less than about twice the member
thickness. This definition is not clear and can easily be taken to mean two things. Either
the shear span is defined as the distance between the center of the column (support) and
the center of the first load (girder) or the distance between the column and the center of
all the applied loads (girders). Based on the second assumption, most if not all inverted-T
bent caps would be considered slender and designed using sectional shear provisions as
the center of the applied load (bridge girders) would be near the center of the bent itself.
With large loads applied at distinct locations along the ledge, the first assumption, with
the shear span defined as the distance from the support to the first load point is likely to
be more appropriate when the actual behavior of the beam is considered.
In cases where the ledge of an inverted-T beam is within twice the web depth for the
support regardless of where the loads are applied, the web can be considered as a Dregion and STM may be appropriate. AASHTO LRFD is silent on the treatment of
geometric discontinuities such as ledges in inverted-T beams.
Based on the geometric and loading discontinuities, strut-and-tie modeling is considered
a more appropriate design procedure than sectional shear for inverted-T beams and will
be investigated in the experimental program. A detailed overview of the STM provisions
of AASHTO (2008) can be found in Birrcher (2009) and will not be covered here. Note
that these provisions changed little in AASHTO (2012).
13
2.6.2 TxDOT Project 0-5253 STM Provisions

The proposed changes to the STM provisions in the AASHTO LRFD Bridge Design
Specifications (Project 0-5253) and their application to the inverted-T beams in the
experimental program are summarized herein. Further discussion on the
recommendations derived from this project for STM of inverted-T caps is provided in
Chapter 7.
2.6.2.1 Recommended Changes to AASHTO LRFD STM Provisions
TxDOT Project 0-5253 and subsequently Project 5-5253 improved upon the current strutand-tie (STM) modeling provisions and recommended modifications to both the ACI 318
and AASHTO LRFD codes. These recommendations are presented in their entirety in
Birrcher et al. (2009) and Williams et al.(2011). The most significant modifications that
affect the design of inverted-T beams focused on the strength of the nodes, particularly
the strut-to-node interfaces.
Three types of nodes exist in strut-and-tie modeling. For inverted-T beams, two of these
nodes are present as shown in Figure 2-9. Hydrostatic nodes are proportioned to have
equal stresses applied to each face. Such nodal constraints result in unrealistic nodal
geometries and impractical reinforcement layouts. Only non-hydrostatic nodes are
therefore recommended in the design of inverted-T beams, as they more closely
correspond to the actual stress concentrations at the nodal regions. For additional details
on hydrostatic and non-hydrostatic nodes refer to Birrcher et al, (2009).
CCT Node
CCT Node
CTT Node
Figure 2-9: Node designations in an inverted-T beams
The CCT node at the support in Figure 2-9 is shown in detail in Figure 2-10. The length
of the bearing face, lb, corresponds to the dimension of the bearing plate. The length of
the back face, wt, is defined by the width of the tie corresponding to the longitudinal
reinforcement. It is taken as twice the distance from the bottom of the beam to the
centroid of the longitudinal steel.
14
Strut-to-Node
wt cos Interface
ws
lb sin
wt
0.5wt
Back Face
lb
Bearing Face
Figure 2-10: Geometry of a CCT node (adapted from Birrcher et al., 2009).
The strut-to-node interface is the face at which the diagonal strut frames into the node.
This surface is perpendicular to the angle, , of the diagonal strut, and thus the length, ws,
depends on the angle. From the geometry of the node, the following equation for ws can
be derived with the previously defined node variables:
(2-1)
For CCC nodes, such as the one under the load of the compression-chord loaded beam in
Figure 2-8, adjacent struts are often resolved together to reduce the number of forces
acting on the node. The node is then divided into two parts since the diagonal struts enter
the node from the right and the left. In the design of inverted-T beams, however, no CCC
nodes exist as the loads are applied to the ledge near the tension chord and hung up to
the top chord of the beams. The resulting node at the compression chord is thus a CCT
node, as shown in Figure 2-9. It is proportioned in a manner similar to that of a CCC at a
load point in a typical rectangular beam.
The struts that intersect the vertical tie at the CCT hanger node are shown in Figure
2-11(a). To simplify the nodal geometry, the adjacent struts are resolved together,
resulting in the diagonal struts presented in Figure 2-11(b). The compressive forces F1
and F2 have combined together to form the force FR. A similar process is repeated but not
shown for the struts on the right side.
15
F1
2 < 1
F2
FR
(a)
(b)
Right Portion
Left Portion
Beam Surface
FR
FR
)P
)P
Tie Width
(c)
(d)
Figure 2-11: CCT hanger node- (a) Original geometry of the STM; (b) adjacent struts
resolved together; (c) node divided into two parts; (d) final node geometry (adapted from
Williams et al., 2011).
The node is divided into two parts when diagonal struts enter from both sides (Figure
2-11(c)). Each section is dimensioned proportionally to the fraction of the applied load P
that the adjacent strut transfers to each support. From statics, if P flows to the left
support, the load acting on the left portion of the node will be P and (1- )P will act
on the right portion.
The geometry of the CCT hanger node, shown in Figure 2-11(d)and Figure 2-12 can now
be defined. The length of the node in the longitudinal direction, lh, is determined by the
width of the hanger tie. Since of the applied load P acts on the left portion of the
node, it occupies of the total node length, lh. The length of the face of the left portion
of the node is therefore taken as lh.
16
lh
lh
(1-
a
lh sin
) lh
Back Face
ws
Strut-to-Node
Interface
(1-
)P
a cos
Figure 2-12: Geometry of a hanger CCT node (adapted from Birrcher et al., 2009)
The length, a, of the back face is taken as the depth of the compression stress block as
defined in a typical flexural analysis. Using this assumption, the value of a for a
rectangular web is determined using the following equation:
(2-2)
where:
As
fy
As
fy
bw
fc
=
=
=
=
=
=
area of longitudinal tension steel (in.2)

yield strength longitudinal tension steel (psi)
area of longitudinal compression steel (in.2)
yield strength longitudinal compression steel (psi)
web width (in.)
specified compressive strength of concrete (psi)
Although a traditional flexural analysis is not valid in a D-region due to the nonlinear
distribution of strains, defining a using the equation above is conservative according to
Williams et al., (2011). The assumption is well-established in practice and will be used to
design the inverted-T specimens in the current experimental program.
The length of the strut-to-node interface, ws, is determined from the same equation as the
support CCT node, Equation 2-1, except the variable wt is replaced with a.
CTT nodes, such as the node denoted in the strut-and-tie model in Figure 2-9, are often
interior nodes and can be classified as smeared nodes. Smeared nodes do not have a
geometry that is restricted by a bearing plate or geometric boundaries and thus cannot be
fully defined. The CTT node shown in Figure 2-13 is similar to the CCT node illustrated
17
in Figure 2-10, but instead of being defined by a bearing surface, the size of the exterior
face, la, is dependent on the width of the vertical tie.
wt cos
la sin
ws
Back Face
Strut-to-Node
Interface
wt
Exterior Face
la
Figure 2-13: Geometry of CTT node(adapted from Birrcher et al., 2009)
In order to proportion the exterior face la, TxDOT Project 0-5253 utilizes a method
recommended by Wight and Parra-Montesinos (2003). They propose that any stirrup that
intersects an adjacent strut at an angle greater than 25 degree can be engaged as part of
the vertical tie in the CCT node. The tie width or the exterior face, la, can then be
conservatively estimated as the entire distance between the outermost stirrups included in
the vertical tie. For the inverted-T specimens in the experimental program, the vertical tie
was considered to be bounded by the 25 degree line on one side and the end of the hanger
load spread, lh, on the other as illustrated in Figure 2-14.
25
lh
la
Figure 2-14: Determination of CTT vertical tie
18
The back face, wt of the node is calculated the same way as the CCT node at the support
and is equal to twice the distance to the centroid of the tension steel, measured from the
tension surface of the beam. The width of the strut-to-node interface, ws, is determined as
in the previous node descriptions, with the angle of the strut framing into the strut-tonode interface is based on the truss geometries.
After the geometry of a node is defined, the design strength of each face is calculated and
compared to the applied force. The design strength of the nodes was governed by the
limiting compressive stresses at the faces of the node, fcu, multiplied by the effective
cross sectional area of the faces, Acn. The cross-sectional area is obtained by multiplying
the length of the face as described in Figure 2-10, Figure 2-12, and Figure 2-13 by the
width of the node perpendicular to the plane of the page. If the node is defined by a
bearing plate, the width of the node is taken as the width of the plate. In other cases, such
as the CCT node above the ledge in an inverted-T beam, the width of the node is the
same as the width of the member, bw. The limiting compressive stress at the face of the
node is calculated in design using the following equation:
(2-3)
Where m is the triaxial confinement factor, v is the concrete efficiency factor, and fc is
the specified compressive strength of the concrete. When evaluating test specimens, the
measured concrete compressive strength was used in this study; as determined from
cylinders tested in accordance with ASTM C39.
If the node is defined by a bearing area with a width smaller than that of the structural
member, the concrete strength for all the faces in that node can be increased due to
triaxial confinement. The triaxial confinement factor, m, is calculated using the following
equation.

(2-4)
Where A1 is the loaded area and A2 is measured on the plane defined by the location at
which a line with a 2 to 1 slope extended from the loading area meets the edge of the
member as shown in Figure 2-15. This modification factor is found in Article 5.7.5 of
AASHTO LRFD (2011) and 10.14.1 of ACI 318-11.
19
45 deg
Loaded Area A1
45 deg
Loaded Area
45 deg
A1
2
1
A2 is measured on this plane
Elevation
Plan
Figure 2-15: Determination of A2 for stepped or sloped supports (from ACI 318-11)
The concrete efficiency factors, , are used to reduce the compressive strength of the
concrete in the node depending on the type of node (CCC, CCT, or CTT) and face
(bearing face, back face, strut-to-node interface) under consideration. The factors
developed in TxDOT Project 0-5253 are summarized in Table 2-2. These factors
correspond directly to the strength of the node and thus were used to determine the
nominal capacity of the specimens in the experimental program.
Table 2-2: TxDOT Project 5-5253-01 concrete efficiency factors, v
Face
Bearing Face
Back Face
CCC
Node Type
CCT
0.85
0.70
CTT
Strut-to-Node
Interface*
* If crack control reinforcement requirement of AASHTO Art. 5.6.3.5 is not satisfied,
use v = 0.45 for the strut-to-node interface
It can be noted from Table 2-2 that the efficiency factor at a strut-to-node interface is
given as the same for both CCC and CCT nodes. Current recommendations therefore do
not reduce the nodal strength due to the presence of a tension field in an inverted-T beam.
In compression-chord loaded members, the node below the applied load is a CCC node.
20
However, the same node in a tension-chord loaded inverted-T member is a CCT node.
This report aims to explore potential differences between tension- and compression-chord
loaded members that may affect efficiency factors of CCT nodes.
The design of struts is simplified by focusing on the strut-to-node interfaces, which
implicitly accounts for the strut capacity and eliminates trivial checks.
Ties are simple in that their design capacity is the total cross-sectional area multiplied by
the yield strength of the bars. As before, no strength reduction factors were used in this
report in order to investigate the nominal strength of the specimens. The ties must be
properly anchored to ensure that the structure can achieve the stress distribution assumed
by the STM. STM procedures require that the yield strength of the tie be developed at the
critical point where the centroid of the tie meets the end of the extended nodal zone;
otherwise known as the edge of the diagonal strut as shown in Figure 2-16.
Assumed
Prismatic Strut
Extended
Nodal Zone
Nodal Zone
Critical Section for

Development of Tie
Available Length
Figure 2-16: Available development length for ties (adapted from Birrcher et al., 2009)
2.6.2.2 Outline of Strut-and-tie Modeling of Inverted-T Bent Specimens

TxDOT Project 0-5253 demonstrated the effectiveness of the modifications proposed to
the AASHTO LRFD STM design procedures for rectangular deep beams. TxDOT Project
5-5253 provided several additional modifications to the provisions. Due to the geometry
of inverted-T beams, certain assumptions not treated in the available STM procedures
had to be made in designing the inverted-T specimens. The design procedures used to
design test specimens are summarized below. The validity of the proposed application of
STM provisions of TxDOT Project 0-5253 and 5-5253 to inverted-T beams is
investigated in Chapter 5.
Inverted-T bent caps transfer the loads in multiple dimensions: from the ledges to the
web, from the tension- to the compression-chord, and from the loading points to the
supports. In order to properly model this behavior it is necessary to consider a three-
21
dimensional strut-and-tie model as the one shown in Figure 2-17. The model can be
divided into two 2-dimensional models to simplify the analysis, provided that the
interaction between them is considered as follows: first, the external loads are applied to
the longitudinal model and forces are calculated for the hanger ties, then, these calculated
hanger forces are applied to the cross-sectional models.
Figure 2-17: Strut-and-tie model of an inverted-T bent cap; top: tri-dimensional model,
center: cross-sectional models, bottom: longitudinal model (from Fernandez 2102)
22
Define loads and solve statics
As the majority of an inverted-T beam is often adjacent to sections with point loads and
geometric discontinuities, it is therefore appropriate to use the STM procedure to design
the beam in its entirety. Once the cross-sectional area of inverted-T specimen is decided
upon, the loads were defined and the reactions determined.
2
Define geometry of the longitudinal strut-and-tie model
In general, each tie must be aligned with the centroid of the reinforcing bars provided to
carry the load. Vertical hanger reinforcement is placed at each load point with the
location of the tie corresponding to the center of the bearing pad. A 45-degree spread
under the loading plates is assumed to define width of hanger ties; as shown in Figure
2-18. For cut-off ledges, such as at the leftmost load point in Figure 2-18, load spread is
limited by the length of the ledge.
45
df
45
c+0.5W+df
W+2df
Figure 2-18: Hanger tie widths for beams with short and cut-off ledges
The horizontal strut along the top of the beam is assumed to be prismatic. The depth, a, of
the compression stress block is defined from a typical flexural analysis as discussed in the
proportioning of hanger CCT nodes in Section 2.6.2.1 (Figure 2-19).
The horizontal tie along the bottom of the beam is aligned with the centroid of the bottom
flexural reinforcement. The width of that tie is assumed to be twice the distance from the
extreme tension fiber to the centroid of the steel (Figure 2-19).
As
Strut Width, a
Compression Chord
Hanger Reinforcement
As
Tension Chord
Tie Width
Figure 2-19: Widths of compression and tension chords
The location of the intermediate tie, Tie BC in Figure 2-20, for the two panel models in
longer shear spans is determined using the technique proposed by Wight and ParraMontesinos (2003). A line is projected at a 25 degree angle from the edge of the support
23
plate at Node A to the top of the beam to define the limit of the tie. Tie BC is then
centered half way between the 45-degree projection from the loading plate at DE and the
25-degree projection from support plate at Node A (Figure 2-20).
B
25
Stirrup
Tie BC
Hanger
Tie DE
Hanger
Tie FG
Hanger
Tie HI
Figure 2-20: Development of strut and tie model
Diagonal bottle-shaped struts, idealized as dashed lines, are then connected to complete
the flow of forces in the longitudinal strut-and-tie model. The angles between the strut
and ties should be checked to ensure that they are greater than or equal to 25 degrees.
Once the longitudinal model is completed, the forces in each element can be calculated
using statics.
3
Define geometry of the cross-sectional strut-and-tie model
Along with the longitudinal model, a strut-and-tie model for the cross section at each load
point is required to design the ledge of the inverted-T beam. The external loads are
applied equally to both sides of the web as shown in Figure 2-21. For unequal spans, it is
conservative to design for the longer span and the resulting greater load. The hanger ties
discussed in the longitudinal STM are located at the center of the vertical reinforcement.
The closed loop ledge reinforcement is positioned next to the hanger reinforcement, with
the top of the loop corresponding to the horizontal ledge tie. The centroid of the
horizontal compression strut is positioned at the depth of the flexural reinforcement from
the longitudinal model. A diagonal strut transfers the applied load from the loading plate
to the bottom of the hanger reinforcement, as shown in Figure 2-21.
Hanger Tie
Hanger
Reinforcement
Diagonal Strut
Ledge Tie
25
Figure 2-21: Forces in cross-sectional models
24
Ledge
Reinforcement
The angle between the diagonal strut and hanger tie must be checked to ensure it is
greater than or equal to 25 degrees but less than 65 degrees. This limit is enforced to
prevent an incompatability of strains. A smaller angle would result in the tension tie
overlapping more of the diagonal strut, decreasing its effectiveness. If the angle is too
small, a wider or shallower ledge is required.
4
Perform nodal strength checks
As the test specimens were designed using TxDOT Project 0-5253 STM provisions, the
nodes were sized and detailed as specified in Section 2.6.2.1. For specimens tested at an
a/d ratio of 1.85, the strut-to-node interface at either the support node or the node above
the closest load point was found to govern the strength of the strut-and-tie model. Thus
the capacity of the specimens tested at the shorter shear span-to-depth ratios were
determined by the strength of the nodes.
5
Determined required steel area for each tie
Once the forces in the truss members were calculated and the nodes checked, the required
steel area was determined to satisfy the tensile forces in the ties. The strength of each tie
was taken as the total area of the reinforcement multiplied by the yield strength of the
steel. After testing the coupon bars obtained with each order of rebar, the yield strength
would be recalculated with the actual material properties. Proper anchorage of the ties
was also ensured within the extended nodal regions.
The bars required for the hanger reinforcement were uniformly distributed within the load
spread area shown in Figure 2-18. The ledge reinforcement was typically paired with the
hanger bars. The effective load spread area was considered to be W+5af or 2c as defined
in AASHTO 5.13.2.5.3 and illustrated in Figure 2-22. If additional tie capacity was
required, the size of the reinforcing bars was increased from No. 5 to No. 6 bars.
af
2c
W+5af
W+5af
Figure 2-22: Load spread area for ledge reinforcement
The size and spacing of the stirrups in the intermediate tie was predetermined by the two
reinforcement ratios investigated in the experimental program. A reinforcement ratio of
0.3% was achieved through spacing No. 4 bars at 6.5 in. on center and 0.6% was
25
achieved with No. 5 bars at 5 in. on center. Additional discussion on the reinforcement
ratio is provided in Section 3.2.4.
a/d = 2.50
J
a/d = 1.85
Figure 2-23: STM for inverted-T test specimen
A final strut-and-tie model for a beam loaded with three point loads is illustrated in
Figure 2-23. Shear spans of 1.85d and 2.50d were examined in this specimen with
shallow, short ledges. Resistance and load factors are required for STM design but were
neglected for the purpose of this report as nominal strengths were computed and
compared with experimental strengths. In general, no serviceability checks were made
before testing the specimens. Rather, the cracking data obtained from loading the beams
was used to validate current serviceability equations and/or make recommendations for
application to inverted-T beams as discussed in Chapter 5.
For complete strut-and-tie design examples for inverted-T bent caps, refer to TxDOT
Report 5-5253 (Williams et al., 2011). A sample specimen design is provided in
Appendix A.
2.7
INVERTED-T DEEP BEAM DATABASE
In addition to the experimental and field investigation portions of the current research
program, a database of deep inverted-T beam shear tests (a/d 2.50) was assembled. The
purpose of this database is to provide a concise collection of specimens that, along with
the results from the experimental program, can be used to verify the accuracy of proposed
design provisions. The database and its filtration criteria are modeled after the deep beam
database compiled as part of TxDOT Project 0-5253.
2.7.1 Literature Review
The first stage in constructing the database consisted of an exhaustive gathering of all the
inverted-T specimens in the literature and collecting all the pertaining information
regarding geometry, reinforcement, boundary conditions, strength, and serviceability.
Many of the specimens reported in the literature were found to be unrepresentative of the
bent caps in Texas due to size alone. The initial survey was expanded and the primary
outcomes of the full literature review are highlighted herein.
26
Initially, the literature review considered experimental studies on reinforced concrete

corbels. While similarities were encountered with respect to the ledge failure
mechanisms, the corbel specimens did not provide insights into the shear strength and
serviceability of inverted-T members. Corbel test data were therefore not included in the
database.
A total of 97 tension-chord loaded specimens reported within 13 unique sources were
ultimately compiled in the literature review. Many of the tests were not applicable to the
current study due to missing plate size information, irregular support conditions, irregular
reinforcement details, impractical cross-section size, and loading conditions that were not
comparable to the in-service bent caps under investigation. The unique aspects and
shortcomings of each reference are noted here. The scaled cross sections of the test
specimens are illustrated in Figure 2-24.
Graf, O., Brenner, E., & Bay, H. (1943): One tension-chord-loaded specimen was
tested in this study. The specimen had a tapered web at the support locations, with
rectangular section in between measuring 118.1 x 3.9 in. (depth x width). Plate
sizes were not reported. The specimen was loaded at five points to simulate a
uniformly distributed load.
Ferguson, P. M. (1956): One tension-chord-loaded specimen was tested in this

study. The specimen was indirectly supported on one end of the beam and loaded
with two concentrated loads. The specimen had a web measuring 12 x 4 in. (depth
x width). Plate sizes were not reported.
Schtt, H. (1956): Six tension-chord-loaded specimens were tested in this study.

The specimens had a tapered web at the support locations and a rectangular
section in between measuring 39.4 x 2.8 in. or 21.7 x 2.8 in. (depth x width). Plate
sizes were not reported.
Taylor, R. (1960): Five tension-chord-loaded specimens were tested in this study.

The specimens had no shear reinforcement. Specimens were indirectly supported
on both ends of the beam and loaded at two points. Specimens measured 12 x 6
in. (depth x width). Plate sizes were not reported.
Leonhardt, F., & Walther, R. (1966): Four tension-chord-loaded specimens were

reported in this study. Specimens measured 63 x 3.9 in. (depth x width).
Specimens had six point loads simulating a uniformly distributed load.
Furlong, Ferguson, & Ma, (1971): Twenty-four inverted-T specimens were tested
in this study. Specimens measured 62 x 24 in., 59 x 24in., and 21 x 8in. (depth x
width). The larger specimens exceeded the test setup capacity and could not reach
failure; except for one specimen that observed local failure of the ledge. The
smaller specimens (21 x 8 in.) were subjected to failures in shear, flexure,
punching shear, and shear friction. Some specimens were loaded on both chords
simultaneously or in cantilever portions of the beam to generate negative bending.
27
Open stirrups were used in some of the specimens. Support plate sizes were not
reported.
Furlong, R. W., & Mirza, S. A. (1974): Twenty seven inverted-T specimens were
tested in this study. Specimens measured 21 x 8 in. (depth x width). Most of the
specimens were post-tensioned and were loaded in positive and negative flexure
with torsional couples. Support plate sizes were not reported.
Smith, K. N., & Fereig, S. M. (1974): Four tension-chord-loaded specimens were

reported in this study. Specimens measured 24 x 3 in. and 12 x 3 in. (depth x
width). Specimens were indirectly supported on both ends of the beam. Plate sizes
were not reported. Some specimens had no shear reinforcement.
Fereig, S. M., & Smith, K. N. (1977): Two tension-chord-loaded specimens were

tested in this study. Specimens measured 12 x 6 in. (depth x width). Specimens
were indirectly supported on both ends of the beam. Plate sizes were not reported.
One specimen did not have shear reinforcement.
Cussens, A. R., & Besser, I. I. (1985): Five tension-chord-loaded specimens were

reported in this study. Specimens measured 39.4 x 2.8in. (depth x width).
Specimens had six point loads simulating a uniformly distributed load. Cracking
loads are available in the report.
Tan, K. H., Kong, F. K., & Weng, L. W. (1997): Six tension-chord-loaded

specimens were reported in this study. Specimens measured 19.7 x 4.3in. (depth x
width). Specimens had two point loads. Cracking loads are available in the report.
Zhu, R. R.-H., Dhonde, H., & Hsu, T. T. (2003): Four inverted-T specimens were
tested in this study. Specimens measured 26 x 21 in. (depth x width). Specimens
had negative and positive bending generated by four point loads, two on the main
span, and one on each cantilever end of the beam. Diagonal reinforcement was
used at the re-entrant corners in the cross-section to control the crack forming
between the ledge and the web. Failure modes were not reported.
Galal, K., & Sekar, M. (2007): Four tension-chord-loaded specimens were

constructed in this study. Specimens measured 24 x 7.1in. (depth x width).
Specimens were loaded to failure, then repaired using CRFP and re-tested.
Specimens had two point loads. Cracking loads are available in the report.
A scaled comparison of the cross-sections of the specimens from the literature, the
inverted-T beams tested in the current project, and distressed in-service bent caps within
the State of Texas is presented in Figure 2-24. A notable difference in size between the
in-service bent caps (hatched blue) and the specimens found in the literature (shaded
black) can be observed.
28
118
Graft & Brenner

(1943)
IH-45 N. Freeway
Bent #5
IH-35 / LP 340
Bent # 17 & 19
Leonhardt & Furlong & Ferguson

Walther (1966)
(1971)
Furlong & Ferguson

(1971)
Cusens &
Besser (1985)
Freigh &
Smith (1974)
IH-35/290
Bent 6K, 28K, &3M
IH-10/ Geronimo
Bent # 4 & 5
Sht
(1956)
Furlong &
Ferguson (1971)
Zhu et al.
(2008)
Furlong &
Mirza (1974)
TxDOT
0-6416
US-59 / N W. Little
York, Bent #4 & 5
TxDOT
0-6416
TxDOT
0-6416
Galal &
Sekar (2008)
Freigh &
Smith (1974)
Furlong &
Mirza (1974)
Tan, et al.
(1997)
12
Taylor
(1960)
Freigh &
Smith (1977)
Freigh &
Smith (1977)
Freigh &
Smith (1974)
Freigh &
Smith (1974)
Ferguson
(1956)
Figure 2-24: Scaled cross-sections of literature review specimens and in-service bent caps.
The sources identified in Figure 2-24 form a comprehensive image of the currently
available literature on inverted-T members. Due to the lack of relevant, complete test
date, the literature review reinforced the need for an extensive experimental program to
evaluate the strength and serviceability effects of tension chord loading. The collection
database was compiled based on the research papers described above and supplemented
with results from the current experimental program. A total of 130 specimens from 14
different sources compose the collection database; including thirty three tests conducted
within Project 0-6416.
29
2.7.2 Filtered Database

As complete test results were required in the inverted-T database, many of the specimens
collected from the literature could not be used. This stage of the database construction
consisted in removing forty one specimens for the following reasons.
In ten of the tests, the specimens were not loaded to failure. Ultimate capacity
information is essential to evaluate the performance of the specimens and calibrate the
new design provisions for inverted-T beams and is thus required for the database.
Filtering based on the actual mode of failure was not performed as it is the intent of the
project to perform a comprehensive assessment of all design provisions for inverted-T
beams (not just those applicable to web shear). As such some beams in the filtered
database were subjected to ledge or flexural failures.
The reports of ten additional specimens were lacking plate size information essential for
the construction of strut-and-tie models to evaluate the performance of the specimens.
Two specimens had no shear reinforcement; an unrealistic condition, as in-service beams
generally have a minimum amount of transverse reinforcement.
Nineteen specimens were found to have complicated support conditions, geometry, or
reinforcement details- conditions were unrealistic when compared to in-service beams.
In summary, the filtered database was reduced to results from 89 specimens that have
adequate details necessary to perform reliable strut-and-tie analyses.
2.7.3 Evaluation Database
The majority of the specimens found in the literature were unrepresentative of the bent
caps in service in Texas, and thus another set of filters were required. Most of the
inverted-T specimens presented in Figure 2-24 have shear areas of less than 200 in.2.
Texas bent caps typically have shear areas of 1,200 in.2 or greater. Also, a significant
number of specimens in the literature review have an aspect ratio greater than 4; some
have a depth over 12 times greater than their width. Such a high aspect ratio is unrealistic
for inverted-T bent caps. Conventional beams have an aspect ratio of approximately one
to three.
In order to construct a useful inverted-T database for the purposes of this project, further
refinement was required to remove specimens that were unrepresentative of the distressed
field members. In this stage 56 tests were filtered due to the following reasons.
Eleven specimens with a web depth-to-web width aspect ratio greater than four were
removed due to the fact that specimens under this condition resemble walls. Their
30
behavior is different from that of bent caps with typical aspect ratios on the order of one
to three.
Nine specimens with web widths smaller than 4.5 in. were filtered out of the database.
The minimum limit was selected as the required width to accommodate two number five
longitudinal bars with 1 in. of clear space between them, with a No. 3 stirrup and a clear
cover of in.
Nine additional specimens were tested with a combined tension- and compression-chord
loading. This condition is unrepresentative of the field specimens that are not subjected to
significant loads on both chords and thus the specimens were removed.
Twenty seven specimens were subjected to torsional loads. These specimens were filtered
out as the distressed field members showed no signs of torsional problems, only web
shear deficiencies (in all cases the observed cracking pattern is consistent with web shear
distress).
In summary, a total of one hundred thirty specimens from fourteen different sources were
included in the collection database (Figure 2-25). The database filtering record is
provided in Table 2-3. Thirty three specimens remained in the evaluation database, all of
them conducted within the TxDOT Project 0-6416. This fact highlighted the importance
of the current experimental program and reinforced the need for a large number of test
specimens to fully evaluate the strength and serviceability behavior of inverted-T bent
caps.
Collection Database = 130 Specimens
Graf & Brenner
1%
Ferguson
1%
Shtt
5%
Taylor
4%
Inverted-T Database = 33 Specimens
Leonhardt &
Walther
3%
TxDOT 0-6416
24%
Furlong &
Ferguson
19%
Galal & Sekar

6%
ZhuWanichakornHsu-Vogel
3%
Tan, Kong,
Cusens &
Weng
Besser
5%
Fereig &
4%
Smith
1%
TxDOT 0-6416
100%
Furlong &
Mirza
21%
Fereig & Smith
3%
Figure 2-25: Sources of the inverted-T database.
31
Table 2-3: Inverted-T database assembly.
2.8
Stage 1
Filtering
Stage 2
Filtering
Collection Database
specimen did not fail
incomplete plate size information
no shear reinforcement
complicated supports/geometry/reinforcement
Filtered Database
h / bw > 4
bw < 4.5in.
tension- and compression-chord loaded
torsional loads
Inverted-T Database
130
10
10
2
19
89
11
9
9
27
33
tests
tests
tests
tests
tests
tests
tests
tests
tests
tests
tests
SUMMARY
Four topics were reviewed in this chapter. First, an overview of distressed inverted-T
bent caps in service in Texas were presented including diagonal crack width information.
Next, background information on strut-and-tie modeling design and behavior of invertedT beams was presented. Then, design provisions for inverted-T beams from the
AASHTO LRFD code, TxDOT bridge design manual, and TxDOT project 5253 were
summarized. Finally, assembly of the inverted-T deep beam database was presented.
32
CHAPTER 3
Experimental Program
3.1
OVERVIEW
In this chapter, details of the design, fabrication, and load testing of the specimens in the
experimental program are provided. Thirty three tests were conducted on beams with
various web reinforcement details, ledge geometries, section sizes, number of point loads,
and a/d ratios.
3.2
TESTING PROGRAM
To accomplish the objectives of the current project, it was necessary to develop an

extensive testing program. Data in the literature was generally insufficient to address the
tasks of the project due to the scarcity of research related to tension-chord loaded
specimens. One of the major objectives of this project was to examine the serviceability
performance of inverted-T deep beams and very little relevant information, including
diagonal crack width data, was available in the literature. Furthermore, the cross-sectional
dimensions of inverted-T beams tested in the past, especially those that led to the
development of shear provisions in the TxDOT bridge design manual, are drastically
smaller than those of in-service members. In order to properly address both the strength
and serviceability objectives of this study it was deemed necessary to fabricate full-scale
specimens within the experimental program.
The experimental program comprised several large-scale tests that encompassed the
variables found in the in-service bridge bents exhibiting problems in the field. The
objective of the tests was to investigate the influence of the following variables on the
strength and serviceability of inverted-T bent caps:
1.
2.
3.
4.
5.
6.
The length of the ledge beyond the bearing of the exterior stringer
The depth of the ledge with respect to the total member depth
The amount of web reinforcement (transverse and longitudinal)
The number of point loads (stringers) on the ledge
The overall depth of the member
Chord loading (tension or compression)
3.2.1 Overview of Test Specimens

An overview of the thirty three specimens is presented in this section. Load testing was
conducted as described in Section 3.6. The test specimens were designed such that they
fail by web shear. Final designs for all test specimens was achieved through use of the
strut-and-tie modeling provisions of TxDOT Project 5253 as implemented for inverted-T
33
members in Section 2.6.3. Six possible failure modes can be expected for inverted-T
beams (Furlong, et al. (1974)):
1.
2.
3.
4.
5.
6.
Flexure.
Torsion.
Web Shear (This failure mode is the focus of the current project).
Yielding of hanger reinforcement.
Punching shear in ledge.
Shear friction in ledge.
Strut-and-tie models inherently consider all failure modes identified above. The
specimens were tested at a shear span-to-depth (a/d) ratio of 1.85 or 2.50. Specimens with
an a/d ratio of 1.85 capture the deep beam behavior of transferring shear from the load
point to the support through a direct compression strut. Specimens with an a/d ratio of
2.50 transfer shear forces through a double strut (or double panel) system and are
considered to be at the limit of sectional shear behavior (Birrcher et al., 2009). The two
a/d ratios were also selected to be directly comparable with the compression-chord loaded
specimens of TxDOT Project 0-5253.
In order to ensure web shear failures, the models were adjusted such that specimen
capacities were controlled by the elements carrying the web shear; i.e. strut-to-node
interfaces for beams with an a/d ratio of 1.85 and the intermediate web tie for beams with
an a/d ratio of 2.50; as shown in Figure 3-1.
a/d = 2.50
a/d = 1.85
Figure 3-1: Simplified strut-and-tie model showing elements limiting shear capacity
The shear span-to-depth ratio is defined within the context of this report as the ratio of the
distance from the center of the support to the center of the nearest loading point, a, with
respect to the effective depth of the specimen d, measured from the centroid of the
longitudinal tension steel to the extreme compressive fiber of the web (Figure 3-2). Such
a definition of shear span is thought to be more appropriate for inverted-T beams than
that provided by AASHTO LRFD (Art. 5.6.3.1) since inverted-T beams are essentially
comprised of D-regions adjacent to large point loads (girders) and geometric
discontinuities (ledges).
34
a
Test Region
Figure 3-2: Free body and shear diagram for a specimen subjected to three point loads
In order to distinguish the specimens from one another and provide a quick insight into
their defining characteristics, the nomenclature presented in Figure 3-3 was developed.
Each numeral is a variable within the testing program.
Web reinforcement ratio (0.3%, 0.6%)
Shear span-to-depth ratio
DS1-42-1.85-03
Web Height (in)
No. of Point Loads
Ledge Length (C = Cut-off, S = Short, L = Long)

Ledge Depth (D = Deep, S = Shallow)
Figure 3-3: Specimen nomenclature.
The testing program was divided into five series based on the parameters under
investigation. A summary of the variables used in each series is provided in Table 3-1. A
brief description of each testing series and the details of the variables under investigation
are provided in Sections 3.2.2 to 3.2.6.
35
Table 3-1: Variables in Testing Program.
Series I
Series II
Series III
Series IV
Series V
Ledge Length Ledge Depth Web Reinforcement Number of Point Loads Web Depth
Cut-off
Shallow (h/3) v = 0.3% h = 0.3%
1
42 in.
Short
Deep (h/2) v = 0.6% h = 0.6%
3
75 in.
v = 0.6% h = 0.3%
Long
The width of the beams was proportioned in order to maximize the cross-sectional area of
the specimen, while keeping it narrow enough to ease installation and removal from the
test setup. A constant web width of 21 in. was used for each beam in the experimental
program. The width of the ledge was also the same, 10.5 in. on each side, for each of the
beams. All other dimensions varied between the beams. Typical specimen geometries and
reinforcing details are shown in Figure 3-4 and Figure 3-5.
36
Figure 3-4: Typical specimen geometries
37
Figure 3-5: Typical reinforcement details.
38
3.2.2 Series I: Ledge Length

The purpose of the Series I specimens was to evaluate the influence of the ledge length
on the strength and serviceability of inverted-T beams. The results from twenty tests were
organized into eight groups of two or three directly comparable specimens, in which
every parameter was kept constant except the length of the ledge.
The varying ledge lengths of inverted-T bent caps have been simplified to three types as
shown in Figure 3-6. A cut-off ledge is one in which the ledge was interrupted just past
the edge of the bearing pad of the outermost stringer. If the ledge runs continuously from
support to support it was considered a long ledge. In a bent cap with a short ledge,
the ledge is allowed to continue a distance equal to the depth of the ledge past the
outermost stringer. Additional details of the Series I specimens are provided in Section
3.2.8 in Table 3-2. Inspected bent caps in the field contained a variety of ledge length
ranging from cut-off to long.
Cut-off Ledge
Short Ledge
Long Ledge
Figure 3-6: Ledge lengths
As discussed in Section 2.3, inverted-T bent caps are tension chord loaded structures in
which the loads from bridge stringers have to be hung up to the compression chord
before being transferred to the support. This induces a tension field in the web, the size of
which is determined by the ledge length as shown in Figure 3-7. In the case of short and
long ledges, the tension field can spread to engage as many hanger bars as possible.
The width of this field is estimated to be equal to the width of the bearing pad plus twice
the ledge height. In the case of cut-off ledges, the force can only spread on one side of the
bearing plate thus concentrating the load in a smaller area and increasing the tensile
stresses in the beam.
39
Engaged Reinforcement
Short Ledge
Cut-off Ledge
Figure 3-7: Load spread in specimens with short and cut-off ledges
Furthermore, by extending a ledge the entire length of a beam, the capacity of the support
node may increase. The additional cross-sectional area provided by the ledge can provide
confinement in the nodal region and increase the bearing width at the support as
compared to beams with short and cut-off ledges as illustrated in Figure 3-8.
Confinement of
support node
Bearing Width
Bearing Width
Long Ledge
Short Ledge
Figure 3-8: Ledge length effect on support region.
3.2.3 Series II: Ledge Depth

The specimens in Series II were evaluated to investigate the effect of the depth of the
ledge on the strength and serviceability performance of inverted-T deep beams. Twenty
tests were organized into ten pairs of directly comparable specimens, in which every
parameter was kept constant except the ledge depth. This allowed for a thorough
examination of the results, which are provided in Section 4.4.
In order to better capture the effect of ledge geometry, the specimens in this series were
designed with two ledge depths. Shallow ledge specimens had depth equal to one-third
and deep ledge specimens were one-half the total height of the beam (Figure 3-9). Ledges
40
shallower than one third of member depth could not be used due to concerns of local
ledge failures. Additional details of the Series II specimens are provided in Section 3.2.8
in Table 3-2.
Deep Ledge
Shallow Ledge
Figure 3-9: Ledge depths
0.5
Deep Ledge hle/h = 0.5
0.4
Shallow Ledge hle/h = 0.3
0.3
0.2
0.1
IH-10
(El Paso)
US 59
(Houston)
IH-35/340
(Waco)
IH-35/290
(Austin)
0
IH-45
(Houston)
Ledge / Beam Height (hle/h)
The specimen ledge depths were chosen to give and adequate range of the configurations
used in practice. The distressed in-service bent caps had ledge depths, hle, between 0.28
and 0.42 times the total height of the beam, h. A comparison between the ledge depth to
beam depth ratio of the in-service beams and those included in the Series III investigation
is shown in Figure 3-10.
Figure 3-10: Ledge depth to beam height ratios of cracked in-service bent caps
As mentioned in Section 3.2.2, loading on the ledge of an inverted-T induces a tension

field in the web, as the forces have to be hung up to the compression chord. As seen
with the ledge length, the ledge depth also has a direct effect on the width over which this
tension field spreads. Deeper ledges allow the forces to spread over a wider area,
41
consequently decreasing the tensile stress in the web whereas a shallow ledge would
theoretically lead to a more concentrated stress (Figure 3-11).
Deep Ledge
Shallow Ledge
Figure 3-11: Load spread in specimens with deep and shallow ledges
3.2.4 Series III: Web Reinforcement Ratio

The effects of web reinforcement on the strength and serviceability of inverted-T deep
beams was evaluated in Series III. Two amounts of web reinforcement were investigated
with areas of steel equal to 0.3% and 0.6% of the effective web area as illustrated in
Figure 3-12. Fourteen tests were organized into six groups of two or three directly
comparable specimens, in which every parameter was kept constant except the web
reinforcement ratio.
sv
bw
Ah
sh
Av
Section A-A
No. 5
6.5
v = h = 0.6%
v = h = 0.3%
No. 4
6.5
Figure 3-12: Web reinforcement ratios
An orthogonal grid of reinforcement is required at each face such that the ratio of total
reinforcement to the gross concrete area is equal to the desired reinforcement ratio. In
42
most tests, the amount of vertical and horizontal web reinforcement was kept equal. Two
specimens had 0.3% in the horizontal direction (skin reinforcement) and 0.6% in the
vertical direction (shear stirrups). The reinforcement ratio of 0.003 (0.3%) is achieved
through spacing No. 4 bars at 6.5 in. on center at each face of the beam. Likewise a 0.006
(0.6%) ratio corresponds to No. 5 bars at 5 in. on center. Additional details of the Series
III specimens are provided in Section 3.2.8 in Table 3-2.
The lower limit of 0.3% corresponds to the minimum required skin reinforcement
according to the TxDOT Bridge Design Manual LRFD (2011), the AASHTO LRFD
Bridge Design Specification 2012, and the findings of TxDOT Project 0-5253. The upper
limit of 0.006 (0.6%) was selected to encompass the maximum reinforcement ratio
(0.57%) found in the in-service distressed bents. The size and spacing of the bars were
chosen to provide crack control. According to Project 0-5253, adequate crack control was
ensured for web bar spacing less than 12 in. or d/4.
3.2.5 Series IV: Number of Point Loads
The purpose of Series IV was to examine the effect of single and multiple loading points
on the strength and serviceability of inverted-T beams. Twelve tests were conducted in
six pairs of directly comparable specimens, in which every parameter was kept constant
except the number of point loads. The specimens in this series were loaded with either
one or three point loads, as shown in Figure 3-13. Specimens with a single load point
were used for direct comparison with compression-chord loaded specimens from TxDOT
Project 0-5253, all of which were tested with a single load point.
One Point Load
Three Point Loads

Figure 3-13: Inverted-T number of point loads
Specimens with three point loads allowed for shallower ledges by distributing the total
force to multiple locations and helping prevent local failure of the ledge. They were also
43
used to help quantify the effect of multiple girders on bridge bent caps. The number of
girders on each side of the bent caps inspected as part of this program ranged from two
U-beams to six box beams.
Three point loads were also chosen because the resulting load spacing was representative
of that of in-service girders. Laboratory testing constraints also factored in the decision.
Additional details of the Series IV specimens are provided in Section 3.2.8 in Table 3-2
and the test results are discussed in Section 4.6.
3.2.6 Series V: Web Depth
Series V was developed to compare the effect of web depth on the strength and
serviceability of the inverted-T beams. A review of the literature revealed a significant
difference in the size of the in-service bent caps when compared to the specimens used to
calibrate the shear provisions in the current design code (TxDOT Bridge Design Manual
2011). Full-scale specimens with web depths of 42 and 75 in were constructed and
tested for the experimental program. This series contains two groups of two directly
comparable specimens, in which every parameter was kept constant except the web
depth.
As with the 42 in. specimens, the web width for the larger beams was limited to 21 in..
By keeping the same web width, the effect of the depth of the beam could be directly
investigated. Each 75-in. beam was constructed with a short, shallow ledge. Providing
enough ledge and hanger reinforcement was not feasible in a cut-off ledge and crane
weight restrictions prevented the construction of a long ledge.
deep
deep
shallow
shallow
75-in. Specimens
42-in. Specimens
Figure 3-14: Inclination angle of ledge strut in cross-sectional STM
44
The design of deep ledges in the 75-in. specimens were restricted due to geometric and
strut-and-tie intersection limitations. The ledge depth defines the inclination of the ledge
strut in the cross-sectional STM as shown in Figure 3-14. The angle of the strut impacts
the strength of the ledge and may lead to the incompatibility of strains in the associated
nodes when the strut intersects a tie at an angle less than 25 degrees as discussed in
Section 2.6.2.2. This incompatibility was seen when 75-in. deep specimen was designed
with a deep ledge and thus only shallow ledges were constructed and tested. Additional
details of the Series V specimens are provided in Section 3.2.8 in Table 3-2 and the test
results are discussed in Section 4.7.
3.2.7 Series VI: Loaded Chord
A final series was added to evaluate the effects of tension chord loading on the strength
and serviceability of reinforced concrete deep beams. As discussed in Section 2.4, ledge
loading introduces additional tensile forces into inverted-T beams that are not present in
rectangular bent caps. It is likely that varying the location at which the load is applied on
the cross-section (Figure 3-15) could result in significant changes in the behavior of the
beam.
Compression Chord Loading
Tension Chord Loading
Figure 3-15: Cross-section loading
Twenty two tests were examined in four groups of directly comparable specimens, in
which every parameter was kept constant except the loaded chord. Results from fourteen
inverted-T and one compression-chord loaded specimens tested in this project are
compared to six compression-chord loaded specimens from TxDOT Project 0-5253 in
Section 4.8. A compression-chord loaded beam, C1-42-1.85-06, was designed and
constructed as part of the current inverted-T experimental program to provide a direct
comparison for reinforcement ratio of 0.6% in each orthogonal direction. Additional
details of the loaded chord specimens are provided in Section 3.2.8 in Table 3-2 and the
test results are discussed in Section 4.7.
45
3.2.8 Summary of Test Specimen Details

Thirty three inverted-T beam tests were conducted in the current experimental program.
The inverted-T deep beams tested in this research project represent some of the largest
deep beam shear test available in the literature. As shown in Figure 3-16, the specimens
from the current study populate the upper bound of the inverted-T deep beam data in the
literature as measured by the shear area of the beam (bwd). As previously noted, the
objectives of the current study necessitated the testing of specimens of comparable size to
field members. A comparison between the investigated in-service bent caps, the beams in
the current study, and beams from previous research projects is provided in Figure 3-16.
Present Study
Previous Research
that led to Code
Development
Figure 3-16: Scaled comparison of actual bent caps and IT beams included in current and
past research programs.
A summary of the details for the thirty three inverted-T tests in the experimental program
is presented in Table 3-2. Two specimens were specially designed with ledge lengths
equal to the length of the bearing plate and are designated with b at the end of the
specimen notation. Because of the large number of variables, each specimen was used in
at least one series. The experimental results for the test specimens are provided and
discussed in Chapter 4.
46
Table 3-2: Summary of beam details

Ledge Ledge Load d a / d Support Load
Specimen
Depth Length Points (in) ratio Plate
Plate
DS1-42-1.85-03
h/2
Short
1
37.6 1.96 16" x 20" 26" x 9"
DS1-42-2.50-03
h/2
Short
1
37.6 2.65 16" x 20" 26" x 9"
DS1-42-1.85-06
h/2
Short
1
37.6 1.85 16" x 20" 26" x 9"
DS1-42-2.50-06
h/2
Short
1
37.6 2.50 16" x 20" 26" x 9"
DL1-42-1.85-06
h/2
Long
1
37.6 1.85 16" x 20" 26" x 9"
DL1-42-2.50-06
h/2
Long
1
37.6 2.50 16" x 20" 26" x 9"
SS3-42-1.85-03
h/3
Short
3
37.6 1.85 16" x 20" 18" x 9"
SS3-42-2.50-03
h/3
Short
3
37.6 2.50 16" x 20" 18" x 9"
SS3-42-2.50-06
h/3
Short
3
37.6 2.50 16" x 20" 18" x 9"
SC3-42-2.50-03
h/3 Cut-off
3
37.6 2.50 16" x 20" 18" x 9"
SC3-42-1.85-03
h/3 Cut-off
3
37.6 1.85 16" x 20" 18" x 9"
DS3-42-2.50-03
h/2
Short
3
37.6 2.50 16" x 20" 18" x 9"
DL1-42-1.85-03
h/2
Long
1
37.6 1.85 16" x 20" 26" x 9"
DL1-42-2.50-03
h/2
Long
1
37.6 2.50 16" x 20" 26" x 9"
SL3-42-1.85-03
h/3
Long
3
37.6 1.85 16" x 20" 18" x 9"
SL3-42-1.85-06
h/3
Long
3
37.6 1.85 16" x 20" 18" x 9"
C1-42-1.85-06
1
37.6 1.85 16" x 20" 30" x 20"
DC1-42-1.85-06
h/2 Cut-off
1
37.6 1.85 30" x 21" 30" x 10"
SS1-75-1.85-03
h/3
Short
1
68.2 1.87 16" x 20" 30" x 10"
DC3-42-1.85-03
h/2 Cut-off
3
37.6 1.85 16" x 20" 18" x 9"
DS3-42-1.85-03
h/2
Short
3
37.6 1.85 16" x 20" 18" x 9"
SS1-42-2.50-03
h/3
Short
1
37.6 2.50 16" x 20" 26" x 9"
SS1-42-1.85-03
h/3
Short
1
37.6 1.85 16" x 20" 26" x 9"
DC1-42-2.50-03
h/2 Cut-off
1
37.6 2.50 16" x 20" 18" x 9"
DL3-42-1.85-03
h/2
Long
3
37.6 1.85 16" x 20" 18" x 9"
SL1-42-2.50-03
h/3
Long
1
37.6 2.50 16" x 20" 26" x 9"
SC1-42-2.50-03
h/3 Cut-off
1
37.6 2.50 16" x 20" 26" x 9"
DS1-42-1.85-06/03 h/2
Short
1
37.6 1.85 16" x 20" 26" x 9"
DS1-42-2.50-06/03 h/2
Short
1
37.6 2.50 16" x 20" 26" x 9"
SC1-42-1.85-03
h/3 Cut-off
1
37.6 1.85 30" x 20" 26" x 9"
DC1-42-1.85-03
h/2 Cut-off
1
37.6 1.85 30" x 20" 26" x 9"
SC1-42-1.85-03*
h/3 Cut-off
1
37.6 1.85 30" x 20" 30" x 10"
DC1-42-1.85-03*
h/2 Cut-off
1
37.6 1.85 30" x 20" 30" x 10"
SS1-75-2.50-03
h/3
Short
1
68.2 2.50 16" x 20" 30" x 10"
Plate dimensions: [in direction of span] x [transverse to direction of span]
* Ledge length set equal to bearing pad length
47
rv
rh
0.3%
0.3%
0.6%
0.6%
0.6%
0.6%
0.3%
0.3%
0.6%
0.3%
0.3%
0.3%
0.3%
0.3%
0.3%
0.6%
0.6%
0.6%
0.3%
0.3%
0.3%
0.3%
0.3%
0.3%
0.3%
0.3%
0.3%
0.6%
0.6%
0.3%
0.3%
0.3%
0.3%
0.3%
0.3%
0.3%
0.6%
0.6%
0.6%
0.6%
0.3%
0.3%
0.6%
0.3%
0.3%
0.3%
0.3%
0.3%
0.3%
0.6%
0.6%
0.6%
0.3%
0.3%
0.3%
0.3%
0.3%
0.3%
0.3%
0.3%
0.3%
0.3%
0.3%
0.3%
0.3%
0.3%
0.3%
0.3%
3.3
FABRICATION OF SPECIMENS
The specimens in the current project were constructed using conventional materials and
methods. Steel formwork was used to expedite the fabrication process and to ensure
dimensional accuracy. In general, the assembly of the reinforcement cage, installation of
strain gauges, placement of concrete, and removal of formwork took about three weeks to
complete per specimen. Beams were tested approximately 28 days after concrete
placement. The following sections describe in detail the construction process and
materials used.
Concrete Mixture Design
Typically, TxDOT engineers specify the compressive strength of concrete used in
inverted-T bent caps to be between 3600 and 5000-psi. Proportions of the typical 5000psi concrete mix are presented in Table 3-3. Actual compressive strengths measure the
same day of testing varied from 3000 to 6400-psi as shown in Table 3-4. During the
placement of concrete for each beam, standard 4 in. x 8 in. cylinders were prepared in
accordance with ASTM C31 and tested in accordance with ASTM C39.
Table 3-3: Typical concrete mixture properties
Material
Quantity
Type I Portland Cement

Flys Ash
CA: 3/4" Limestone
FA: Sand
Water
HRWR Admixture
Set Retardant Admixture
Water/Cement Ratio
Slump
388 lb/cy
94 lb/cy
1650 lb/cy
1528 lb/cy
24 gallons/cy
24 oz/cy
5 oz/cy
0.52
6 2 inches
*HRWR: High Range Water Reducing (i.e. Superplasticizer)

Reinforcement Properties
Domestic Grade 60 deformed bars satisfying the requirements of ASTM A 615 were used
for all steel reinforcement (rebar). Cross sectional dimensions of the bars complied with
the nominal sizes given in ASTM A615.
Each order delivered to the Ferguson Laboratory was accompanied by a set of four
coupons for each bar size. The tensile strength of the coupons was measured in
accordance with ASTM A370 and is provided in the material properties in Table 3-4.
48
Table 3-4: Specimen material strengths
# 11 Bars # 6 Bars # 5 Bars # 4 Bars

f y (ksi)
f y (ksi) f y (ksi) f y (ksi)
DS1-42-1.85-03
69.24
63.38
64.69
63.14
DS1-42-2.50-03
69.24
63.38
64.69
63.14
DS1-42-1.85-06
64.13
63.38
60.68
N/A
DS1-42-2.50-06
64.13
63.38
60.68
N/A
DL1-42-1.85-06
67.90
63.38
64.69
N/A
DL1-42-2.50-06
67.90
63.38
64.69
N/A
SS3-42-1.85-03
68.60
64.68
62.75
67.25
SS3-42-2.50-03
68.60
64.68
62.75
67.25
SS3-42-2.50-06
69.50
61.83
60.90
N/A
SC3-42-2.50-03
66.20
63.50
60.25
64.27
SC3-42-1.85-03
66.20
63.50
60.25
64.27
DS3-42-2.50-03
63.60
62.63
60.22
64.58
DL1-42-1.85-03
71.01
61.90
64.29
64.43
DL1-42-2.50-03
71.01
61.90
64.29
64.43
SL3-42-1.85-03
75.18
60.62
63.58
65.57
SL3-42-1.85-06
70.38
63.26
64.80
62.62
C1-42-1.85-06
73.30
63.98
60.81
N/A
DC1-42-1.85-06
73.30
63.98
60.81
N/A
SS1-75-1.85-03
66.10
61.97
64.69
65.08
DC3-42-1.85-03
63.63
66.00
63.09
63.16
DS3-42-1.85-03
63.63
66.00
63.09
63.16
SS1-42-2.50-03
65.44
69.57
77.76
66.58
SS1-42-1.85-03
65.44
69.57
77.76
66.58
DC1-42-2.50-03
70.06
64.13
69.77
62.44
DL3-42-1.85-03
70.06
64.13
69.77
62.44
SL1-42-2.50-03
68.70
71.41
N/A
64.47
SC1-42-2.50-03
68.70
71.41
N/A
64.47
DS1-42-1.85-06/03
65.80
70.92
64.94
65.18
DS1-42-2.50-06/03
65.80
70.92
64.94
65.18
SC1-42-1.85-03
66.36
64.04
N/A
67.28
DC1-42-1.85-03
66.36
64.04
N/A
67.28
SC1-42-1.85-03b
70.47
63.12
N/A
68.56
DC1-42-1.85-03b
70.47
63.12
N/A
68.56
SS1-75-2.50-03
65.22
63.85
63.62
63.76
Specimen
49
f 'c
(psi)
5258
5389
5024
5088
4830
4986
5891
5891
6255
5873
5873
5687
4929
4929
5037
5250
3727
3727
3127
4568
4568
5703
5721
4035
4202
4281
4281
4173
4173
4330
4303
3013
2996
5158
3.3.1 Construction of Specimens

The steel reinforcement was supplied by a local steel manufacturer and delivered in the
specified lengths and with the appropriate bends. The reinforcement cages were
assembled in the laboratory and, upon completion and instrumentation, were moved into
the formwork in the casting area. The specimens were cast in the same orientation that
they were tested. As they were loaded from beneath, the primary flexural reinforcement
was located at the top of the section as shown in Figure 3-17.
The concrete used to fabricate the specimens was provided from a local ready-mix
supplier. Upon the arrival of the concrete truck, a slump test was performed according to
ASTM C143. If necessary, additional water was added to increase the slump to
approximately 6 2 in. The additional amount of water added did not exceed the amount
of water that was held back at the batch plant as indicated on the ticket. 4-in. diameter
cylinders were prepared in accordance with ASTM C31. They were covered and cured
under the same ambient conditions as the beam.
Figure 3-17: Fabrication of specimens
50
Two trucks were ordered for each 75-in. deep beam and one truck was required for each
42-in. deep beam. For the larger pours, each truck was filled with the same mixture
design from the same batch plant. The second truck was scheduled to arrive 30 minutes
after the first truck to minimize idling and prevent the formation of a cold joint. The
compressive strength from one truck generally differed from the other by less than 20
percent.
Concrete was placed via a one-cubic yard bucket lifted by an overhead crane. Proper
consolidation was achieved through the use of external pneumatic vibrators that moved
along the length of the formwork and internal rod vibrators, or stingers, that were placed
in the concrete from the top as shown in Figure 3-17. After initial setting, the top surface
of the beam was finished and the beam was covered with a plastic sheet and cured under
the ambient laboratory conditions.
Seven days after casting, the specimens were uncovered and stripped of their forms. The
cylinders were likewise uncovered and removed from their molds. The specimens were
then moved into the test setup with an overhead crane and allowed to reach an age of at
least 28 days before testing.
3.4
TEST SETUP
The specimens were tested at the Phil M. Ferguson Structural Engineering Laboratory at
the University of Texas at Austin. The test setup was originally constructed for TxDOT
Project 0-5253 and was designed for an upside-down simply-supported beam test. Minor
modifications were required for use in the current project. The load was applied via a 5
million pound capacity, double-acting hydraulic ram for single point load tests, and three
2 million pound capacity rams for multiple point load specimens. U-shaped frames were
introduced to spread loads around the web and apply load evenly to the ledges on both
faces of the test specimens. The strong floor consisted of a 96,000-lb steel platen. At each
support, six 3-in. diameter threaded rods reacted against a 7,000-lb transfer girder to
resist the applied load. In the current configuration shown in Figure 3-18, the test set
setup can resist a shear force of approximately 1.5 million pounds or an applied load at
midspan of approximately 3 million pounds. For additional details of the test setup design
and construction, refer to Huizinga (2007).
51
Transfer Beam
Load Cells
Load plates
Support Plates
Loading U Frame
3 Diameter Rods
Steel Platen
Hydraulic Ram
Figure 3-18: Test setup
At each support, pin connections were achieved by the use of a 2-in. diameter steel bar
between two 2-in. steel plates. The steel bar was welded to the bottom plate which was
placed on top of a thin layer of hydrostone that provided a planar bearing surface on top
of the beam. Horizontal movement was permitted by the flexibility of the six threaded
rods. Similarly, at the applied load, rotation was permitted by a 3-in. diameter roller
which was allowed to roll freely between a 5-in. or 7-in. thick plate and the top of the Uframe. A smaller -in. thick steel plate was placed on top of the larger plate to achieve
the desired size of the load plate. A -in. neoprene bearing pad was placed between the
concrete ledge and load plate to ensure a uniform load distribution.
3.5
INSTRUMENTATION
The instruments used during testing to obtain and record data included steel electrical
strain gauges, linear potentiometers, load cells, and crack comparator cards. Further
details regarding each of these are provided in the following sections.
3.5.1 Strain Measurements
Strain gauges were affixed to the steel reinforcement to measure the change in strain in
areas of particular interest including flexural reinforcement, hanger and ledge bars, skin
reinforcement, and shear stirrups. The gauge type was model FLA-3-11-5LT
52
manufactured by Tokyo Sokki Kenkyujo Co. The width and length of the gauges were
1.5 and 3 mm, respectively, with a resistance of 120-ohms ( 0.5).
Typical locations of strain gauges are illustrated in Figure 3-19. Strain gauges were
attached to the stirrups and horizontal skin reinforcement along the assumed centerline of
the inclined strut in all of the test specimens. Beams with a shear span-to-depth ratio of
1.85 were instrumented along the axis of the direct strut that spans from the support to the
first loading point. Beams with the longer shear span-to-depth ratio of 2.50 were
instrumented along the axis of the first strut in the multiple panel model as well as the
line between the load and the support. The purpose of installing a gauge along the
proposed strut centerline was to measure steel strains at or close to the primary diagonal
splitting crack. Strain gauges were also affixed to the primary flexural reinforcement in
several locations including the load point to monitor the maximum strain in the
reinforcement as the beam was loaded to failure.
Two Panel Model (a/d = 2.5)
Direct Strut (a/d = 1.85)
Gauges on Stirrups
Gauges on Stirrups
Gauges on Longitudinal Bars
Gauges on Ledge and Hanger Bars
Figure 3-19: Typical location of strain gauges in longitudinal section
Strain gauges were used to monitor the hanger and ledge reinforcement in the regions of
the load plates as shown in Figure 3-20. They served to verify the assumed 45 degree
load-spread and the associated number of hanger bars that transferred the applied loads to
the compression chord.
53
Assumed Load Spread
IT Cross Section
Figure 3-20: Strain gauges on the hanger and ledge reinforcement
The installation procedure of the gauges is shown in Figure 3-21. The deformations on
the bars were removed with a grinder and polished to provide a smooth planar surface for
the application of the gauges. Care was taken to not significantly reduce the crosssectional area of the bar. The surface was cleaned and the strain gauges were glued to the
reinforcement. A butyl rubber tape was applied to protect and water proof the gauges,
which were then wrapped in foil tape to further isolate them from the concrete.
Figure 3-21: Steel strain gauge installation
3.5.2 Load and Displacement Measurements

The force applied to the beam during testing was measured by 500-kip capacity load cells
placed between the transfer beam and the reaction nut at each of the twelve rods, six at
each support as shown in Figure 3-22. Because the load cells were located at either end,
the shear in the test region could be accurately measured throughout the loading history.
54
Care was taken to balance the reaction at each side of the supports to prevent torsion in
the test specimen and overloading of a single rod.
500-kip
Load Cell
Transfer Girder
Support Rods
Test Specimen
Figure 3-22: Load cell arrangement at supports
Beam deflection and rigid body motion were measured using an arrangement of five 6-in.
linear potentiometers with one located at each support, one at mid-span, and two at the
load point closest to the support as shown in Figure 3-23. The purpose of the linear
potentiometers was to measure the deflections of the beam throughout the test. The
specimens underwent rigid body motion as they were lifted off their supports at the start
of the test and as the reaction rods elongated. The displacement measured at the support
was used to subtract this motion from the beam midspan deflections. The two linear
potentiometers at the load point allowed for observation of rotation of the beam along the
longitudinal axis.
CL
Linear Potentiometers
Figure 3-23: Deflection measurement locations
55
3.5.3 Crack Width Measurements

Diagonal crack widths were measured and recorded for each test specimen as part of the
serviceability considerations of the experimental program. At each load increment, the
maximum width of any diagonal crack was collected for each face of the shear span
under investigation. Several cracks were also arbitrarily selected to be monitored at the
same location throughout the entire test. Independent measurements were obtained by
two students with the use of a crack comparator card as shown in Figure 3-24. The crack
widths were averaged to produce the diagonal crack width data. No distinction was made
between web-shear or flexure-shear cracks. As long as the crack formed at a significant
angle from the beam longitudinal axis it was considered a diagonal crack. In general, the
location of the maximum width of the diagonal cracks was near or slightly above the
mid-height of the member.
Figure 3-24: Measuring crack widths with a comparator card
3.6
TEST PROCEDURE
Test specimens were monotonically loaded in 50-kip increments up the appearance of the
first diagonal crack, then in 100-kip increments up to failure. After each load increment,
cracks were marked and the widths of the widest diagonal shear cracks were measured.
Photographs were taken after each load increment and the failure was recorded with a
video camera.
Two tests were conducted on each beam. Specimens with a single point load were loaded
a distance from one support corresponding to the desired a/d ratio. After a shear failure
was achieved, the load was removed and post-tensioning clamps were installed to
strengthen the first test region. The hydraulic ram and U-frame were then moved to the
opposite end of the beam and the test procedure was repeated. Both test regions cracked
56
during the first test on each beam, therefore the cracking load was not able to be recorded
for the second test region for beams with single point loads.
Failure Crack
Test 1
Failure Crack
External PT Clamps
Test 2
Figure 3-25: Three point loads, first and second test
Specimens with three loading points were designed such that both ends were tested
simultaneously as shown in Figure 3-25. For those specimens, the cracking load was
obtained for the two test regions, both of which were monitored until a shear failure was
achieved at one end of the beam. At that point the load was removed, post-tensioned
clamps were installed in the failed test region, and the load was reapplied at the same
location until the opposite end of the beam failed in shear. Two tests were also conducted
simultaneously for the 75-in. specimens. In this case, the size of the specimen was chosen
such that the resulting a/d ratios on each side of the beam matched the appropriate a/d
ratios of the experimental program.
Although the beams in the experimental program were tested in an inverted, simplysupported manner, the photographs and drawings of the test regions in the rest of the
report will be shown upside down. This is to maintain consistency with the conventional
manner of the in-service inverted-T beams (i.e. with tension steel at the bottom of the
beam).
3.7
SUMMARY
The details of the experimental program were provided in this chapter. The experimental
variables studied in this project were: (1) ledge length, (2) ledge depth, (3) web
reinforcement, (4) number of point loads, and (5) web depth. The test specimens were
designed to evaluate the effect of these variables on the strength and serviceability of
inverted-T beams, thus requiring specimens of comparable size to field members. Thirty
three tests were conducted on beams with widths of 21 in. and depths of 42 in. and 75 in.
at shear span-to-depth ratios of 1.85 and 2.50.
57
In addition, the details of the fabrication and testing procedure were provided in this
chapter. Two tests were performed on each beam with the help of external post-tensioned
clamps for strengthening the off-span. Instrumentation that measured steel strains,
applied loads, and beam deflations were monitored throughout the test. Crack width
measurements were also taken between each load increment. The results for the
experimental program are provided in Chapter 4. The analysis of the experimental results
and the additional design objectives of the current project are presented in Chapter 5.
58
CHAPTER 4
Experimental Results
4.1
OVERVIEW
In this chapter, the experimental results of the testing program are presented in detail.
Thirty three tests were conducted on full-scale specimens with the following six
parameters investigated:
length of the ledge (Section 4.3)

depth of the ledge (Section 4.4)
amount of web reinforcement (Section 4.5)
number of point loads (Section 4.6)
member depth (Section 4.7)
tension- versus compression-chord loading (Section 4.8)
Prior to discussing the effects of the various parameters, a summary of the experimental
results for all of the test specimens is provided. In addition, pertinent information
regarding the evaluation of the strength and serviceability data is given to facilitate the
interpretation of results.
4.2
SUMMARY OF EXPERIMENTAL RESULTS
The strength and serviceability results of all tests conducted in the experimental program
are summarized in Table 4-1. Other important details were provided previously in Table
3.2. The variables presented in Table 4-1 are defined as follows:
bw =
beam web width, in.
d=
distance from extreme compression fiber to centroid of tensile

reinforcement, in.
fc =
compressive strength of concrete at the time of testing measured in

accordance with ASTM C39, psi.
fyl =
yield strength of longitudinal reinforcement measured in accordance

with ASTM A370, ksi.
fyv =
yield strength of transverse reinforcement measured in accordance

with ASTM A370, ksi.
fyh =
yield strength of skin reinforcement measured in accordance with

ASTM A370, ksi.
59
fyha =
yield strength of hanger reinforcement measured in accordance with

ASTM A370, ksi.
a/d ratio = shear span-to-depth ratio

Vcrack =
shear carried in the test region at the formation of the first diagonal
crack, kip.
Specific details regarding the determination of the diagonal cracking
load are presented in Section 4.2.2.
Vtest =
maximum shear carried in the critical section of the test region,

including self-weight of the specimen and test setup.
Specific details regarding the determination of the applied shear are
presented in Section 4.2.1.
It should be noted that the majority of the specimens sustained web shear failures, but in
a few cases flexural failure, diagonal strut failure in the cross section, punching shear, or
ledge-to-web shear friction failures were observed. The value reported for V test is the
maximum shear carried at the critical section at the onset of strength loss, regardless of
the failure mode. A note was added in Table 4-1 to the specimens that experienced a
failure mode different than web shear.
60
Table 4-1: Summary of experimental results
Specimen
DS1-42-1.85-03
DS1-42-2.50-03
DS1-42-1.85-06
DS1-42-2.50-06
DL1-42-1.85-06
DL1-42-2.50-06
SS3-42-1.85-03
SS3-42-2.50-03
SS3-42-2.50-06 (f)
SC3-42-2.50-03
SC3-42-1.85-03
DS3-42-2.50-03
DL1-42-1.85-03
DL1-42-2.50-03
SL3-42-1.85-03
SL3-42-1.85-06
DC1-42-1.85-06
C1-42-1.85-06
bw
f'c fyl fyv fyh fyha a/d Vcrack
d (in.)
(psi) (ksi) (ksi) (ksi) (ksi) ratio (kip)
(in.)
21
21
21
21
21
21
21
21
21
21
21
21
21
21
21
21
21
21
37.64
37.64
37.64
37.64
37.64
37.64
37.64
37.64
37.64
37.64
37.64
37.64
37.64
37.64
37.64
37.64
37.64
37.64
5258
5389
5024
5088
4830
4986
5891
5891
6255
5873
5873
5687
4929
4929
5037
5250
3727
3727
69
69
64
64
68
68
69
69
70
66
66
64
71
71
75
70
69
69
63
63
61
61
61
61
67
67
61
64
64
65
64
64
66
65
61
61
63
64 1.96
63
64 2.65
61
64 1.85
61
64 2.50
61
64 1.85
61
64 2.50
67
65 1.85
67
65 2.50
61
62 2.50
64
64 2.50
64
64 1.85
65
63 2.50
64
62 1.85
64
62 2.50
66
64 1.85
65
63 1.85
61
64 1.85
61
64 1.85
(f) Flexural failure
61
172
N/A
188
N/A
168
N/A
126
140
115
113
90
143
242
N/A
172
154
107
N/A
Vtest
(kip)
2.99
N/A
3.35
N/A
3.06
N/A
2.08
2.31
1.84
1.87
1.48
2.40
4.36
N/A
3.06
2.69
2.23
N/A
0.24
N/A
0.30
N/A
0.23
N/A
0.24
0.31
0.22
0.34
0.19
0.33
0.39
N/A
0.30
0.21
0.21
N/A
712
406
621
503
741
622
523
447
516
329
483
430
626
510
571
744
519
637
0.17
0.10
0.16
0.13
0.19
0.16
0.11
0.10
0.10
0.07
0.10
0.10
0.16
0.13
0.14
0.18
0.18
0.22
12.42
6.99
11.09
8.93
13.48
11.15
8.62
7.38
8.25
5.44
7.98
7.21
11.28
9.19
10.17
13.00
10.76
13.21
Table 4-1(Cont.d): Summary of experimental results
bw
(in.)
SS1-75-1.85-03 21
DC3-42-1.85-03 21
DS3-42-1.85-03 21
SS1-42-2.50-03 21
SS1-42-1.85-03 21
DC1-42-2.50-03 21
DL3-42-1.85-03 (f) 21
SL1-42-2.50-03 21
SC1-42-2.50-03 (r) 21
DS1-42-2.50-06/03 21
DS1-42-1.85-06/03 21
SC1-42-1.85-03 (le) 21
DC1-42-1.85-03 21
SC1-42-1.85-03b 21
DC1-42-1.85-03b 21
SS1-75-2.50-03 (p) 21
Specimen
d
(in.)
68.20
37.64
37.64
37.64
37.64
37.64
37.64
37.64
37.64
37.64
37.64
37.64
37.64
37.64
37.64
68.20
f'c fyl fyv fyh fyha a/d Vcrack

(psi) (ksi) (ksi) (ksi) (ksi) ratio (kip)
3127 66 65
65
62 1.87 346
4.51
4568 64 63
63
66 1.85 152
2.84
4568 64 63
63
66 1.85 164
3.07
5703 65 67
67
70 2.50 157
2.63
5721 65 67
67
70 1.85 N/A
N/A
4035 70 62
62
64 2.50
70
1.40
4202 70 62
62
64 1.85 276
5.39
4281 69 64
64
71 2.50 167
3.24
4281 69 64
64
71 2.50 N/A
N/A
4173 66 65
65
71 2.50 115
2.25
4173 66 65
65
71 1.85 N/A
N/A
4330 66 67
67
64 1.85 N/A
N/A
4303 66 67
67
64 1.85 127
2.45
2996 70 69
69
63 1.85 N/A
N/A
3013 70 69
69
63 1.85 N/A
N/A
5158 65 64
64
64 2.50 225
2.19
(r) Shear friction failure of the web-to-ledge interface
(le) Ledge tie failure
(p) Punching shear failure
62
0.46
0.38
0.36
0.39
N/A
0.19
0.44
0.34
N/A
0.21
N/A
N/A
0.25
N/A
N/A
0.35
Vtest
(kip)
745
395
454
398
583
365
629
498
319
539
739
463
517
456
424
649
0.17
0.11
0.13
0.09
0.13
0.11
0.19
0.15
0.09
0.16
0.22
0.13
0.15
0.19
0.18
0.09
9.31
7.39
8.49
6.67
9.75
7.28
12.27
9.62
6.18
10.59
14.49
8.67
9.97
10.54
9.77
6.31
4.2.1 Evaluation of Strength Data

The shear strength of the test specimens, Vtest, was defined as the maximum shear carried
at the critical section. The critical section was considered to be the point halfway between
the support and the nearest load in the test region of interest. Vtest was determined by
summing the reactions measured by the load cells at the support, RA or RB, the
corresponding portion of the self-weight of the specimen, SW, and the weight of a
transfer girders, PTR, as shown in Figure 4-1. The self-weight was assumed to be uniform
along the entire length of the beam
RA
PTR
RB
PTR
critical section
PL+PD+2PTR
a
L-a
LH
LH
Vtest = SW(LH+a/2)+RB+PTR
Where:
PL = RA + RB
PTR = 7.8 kip
PD = SW(2LH + L)
a/2
L = 255.25 in.
LH = 38.375 in.
SW = Specimen self-weight, kip/ft
Figure 4-1: Load and shear force diagram for typical beam test.
The shear capacity of the test specimens presented in this chapter is normalized by the
cross-sectional area and the strength of the concrete. Depending on the a/d ratio the
failure mechanism can depend on the tension and/or compressions strengths of the
concrete. At higher a/d ratios (> 2.0), sectional shear (diagonal tension) generally controls
the strength of a member. The shear capacity in these beams is therefore related to the
tensile strength of concrete or its proxy . The capacity of deep beams (a/d < 2.0) is
related to the compression strength and to some extent the tensile strength of concrete as
failure occurs in such beams due to crushing and splitting of the concrete struts and
nodes. The ultimate capacity of the test specimens were therefore normalized by both
bwd and
bwd so that more direct comparisons between the various beams
63
capacities can be made In Table 4-1, the diagonal cracking loads of the test specimens,
which depend on the tensile strength of the concrete, are normalized by
bwd.
4.2.2 Evaluation of Serviceability Data

In order to evaluate the serviceability performance of the specimens, two parameters
were considered: the first diagonal cracking load, and the progression of maximum crack
width with increasing shear.
The load corresponding to the first shear crack was recorded during testing by visual
inspection of the test region. Upon examination of the test data, the first cracking load,
Vcrack, was determined at the load when a sudden increase in strain was measured by the
gauges affixed to the stirrups. This load was confirmed with the observations made
during testing. As noted in Section 3.6, the diagonal cracking loads were only obtained
for the first test on each specimen, unless both spans were tested simultaneously. An
SSV in Figure 4-2.
illustration of the data used to determine Vcrack is presented
400
SL1-42-2.50-03
350
NSV3
300
NSV5
Shear (kip)
NSV2
250
SSV4
SSV6
200
SSV8
150
SSV9
Vcrack
100
(168 kip)
3
Observed
First Crack
50
2
4
7
5
0
0
0.0002
0.0004
0.0006
0.0008
0.001
Stirrup Strain (in./in.)

Figure 4-2: Visual and experimental determination of first cracking load.
At each load increment, the beam was inspected and cracks were marked. The maximum
diagonal crack width was measured with a crack comparator card and recorded along
with the corresponding applied load. An example of crack width progression is given in
Figure 4-4.
In order to characterize the cracking performance of test specimens at service load levels,
a benchmark crack width of 0.016 in. was selected. Maximum crack widths recorded
below that threshold were deemed acceptable for long-term serviceability considerations.
64
The selected value is consistent with the tolerable service crack widths listed in ACI
224R-01 and fib-1999 for dry exposure, as well as with TxDOT Project 0-5253. ACI
224R-01 reports that crack width limits are expected to be exceeded by a significant
portion of the cracks thus the values are only meant as general guidelines to be used in
conjunction with sound engineering judgment. Thus even though bent caps may be
exposed to wet and dry cycles, the dry exposure crack limit was deemed acceptable for
the evaluation of test specimens for which the actual maximum crack widths were
recorded at every loading increment.
Along with the limit on maximum crack width, a service load level corresponding to 33%
of the maximum applied load was selected as an approximate service load level for test
specimens. This value is consistent with the value used in TxDOT Project 0-5253.
Assumptions leading to the 33% value are detailed in Figure 4-3. Maximum diagonal
crack width progressions of four typical tests are presented in Figure 4-4 in conjunction
with the load and crack width serviceability criteria. In that figure, specimens with crack
progression outside of the bottom right quadrant drawn by the selected limits are deemed
to have acceptable detailing to limit crack widths at service loads. Crack width
evaluations at service loads for all specimens tested are presented in Section 4.5.3.
Nominal Capacity Service Load
Assumptions:
2/3
Service Load
Nominal Capacity
1). Load Case: 1.25DL + 1.75LL

2). DL = 75% of Service Load
LL = 25% of Service Load
3). Nominal = 2/3 Experimental
0.70
1.4
= 0.33
= 1.4
Service Loads
Experimental Capacity
= strength reduction factor, 0.70

= load factor
DL = dead load
LL = live load
Figure 4-3: Service load level estimation (Birrcher, Tuchscherer, et al., Strength and
Serviceability Design of Reinforced Concrete Deep Beams 2008)
65
% of Maximum applied load
100
90
80
70
60
50
40
33%
a/d = 1.85 r = 0.003
30
a/d = 1.85 r = 0.006
20
a/d = 2.50 r = 0.003
10
0
0.000
a/d = 2.50 r = 0.006
0.016
0.020
0.040
0.060
Maximum Diagonal Crack Width, in.
Figure 4-4: Typical crack width progression plot
4.3
SERIES I: LEDGE LENGTH
Different ledge lengths were investigated in the current study to evaluate the effect on the
strength and serviceability of inverted-T straddle bent caps. The results of Series I will be
used to develop design recommendations in regards to the length of ledges in bridge bent
caps.
Three different ledge lengths were investigated. Beams with ledges that were interrupted
right after the outermost stringer were cut-off ledge specimens. Bent caps in which the
ledge ran from end to end of the beam were considered to be long ledge specimens.
Short ledge specimens were the in-between case where the ledge was extended enough
to allow a 45-degree force spread from the last bearing pad to the bottom of the beam, but
did not continue all the way to the support. Section 3.2.2 provides additional background
information on the ledge length series.
4.3.1 Experimental Results
Twenty tests produced eight groups of two or three directly comparable specimens in
which every parameter including ledge depth, web reinforcement, number of point loads,
and web depth was kept constant except the ledge length. A summary of the experimental
results from this series is provided in Table 4-2. All variables are defined in Section 4.2.
Strength results are normalized as discussed in Section 4.2.1.
All specimens in this series failed in web shear except DL3-42-1.85-03 and SC1-42-2.5003. The test of Specimen DL3-42-1.85-03 was halted upon the onset of yielding of the
flexural tensile reinforcement and crushing of the concrete in the flexural compression
region. Specimen SC1-42-2.50-03 suffered a shear friction failure and the test was halted
66
upon the bending of the ledge longitudinal bars and the separation of the ledge from the
web of the beam. The maximum shear values reported in Table 4-2 is the amount of shear
carried at the critical section of the beam at the onset failure, regardless of the failure
mode.
Table 4-2: Series I: Ledge Length- Experimental results.
f 'c
Vtest
(psi) (kip)
5258 712
DS1-42-1.85-03
D_1-42-1.85-03
DL1-42-1.85-03
4929 626
DC3-42-1.85-03
4568 395
D_3-42-1.85-03 DS3-42-1.85-03
4568 454
DL3-42-1.85-03 (f) 4202 629
SC3-42-1.85-03
5873 483
S_3-42-1.85-03 SS3-42-1.85-03
5891 523
SL3-42-1.85-03
5037 571
DS1-42-1.85-06
5024 621
D_1-42-1.85-06
DL1-42-1.85-06
4830 741
DC1-42-2.50-03
4035 365
D_1-42-2.50-03 DS1-42-2.50-03
5389 406
DL1-42-2.50-03
4929 510
SC1-42-2.50-03 (r) 4281 319
S_1-42-2.50-03 SS1-42-2.50-03
5703 398
SL1-42-2.50-03
4281 498
SC3-42-2.50-03
5873 329
S_3-42-2.50-03
SS3-42-2.50-03
5891 447
DS1-42-2.50-06
5088 503
D_1-42-2.50-06
DL1-42-2.50-06
4986 622
Comparison
Vcrack
(kip)
Specimen
0.17
0.16
0.11
0.13
0.19
0.10
0.11
0.14
0.16
0.19
0.11
0.10
0.13
0.09
0.09
0.15
0.07
0.10
0.13
0.16
12.42
11.28
7.39
8.49
12.27
7.98
8.62
10.17
11.09
13.48
7.28
6.99
9.19
6.18
6.67
9.62
5.44
7.38
8.93
11.15
172
242
152
164
276
90
126
172
188
168
70
N/A
N/A
N/A
157
167
113
140
N/A
N/A
2.99
4.36
2.84
3.07
5.39
1.48
2.08
3.06
3.35
3.06
1.40
N/A
N/A
N/A
2.63
3.24
1.87
2.31
N/A
N/A
4.3.2 Strength Results

The strength results of the twenty ledge depth specimens are directly compared in Figure
4-5 and Figure 4-6. Each of the eight plots represents a group in which every parameter
was kept constant except the ledge length. Vtest is normalized by
bwd in Figure 4-5 and
by bwd in Figure 4-6. The first figure is often considered more appropriate for
shorter shear spans (a/d < 2) and the second for longer (a/d > 2) but both are included for
evaluation purposes.
67
D_1-42-1.85-03
D_3-42-1.85-03
0.25
0.25
0.20
0.20
0.15
0.15
0.10
0.10
0.05
0.05
0.00
Cut-off
Short
0.00
Long
D_1-42-2.50-03
S_3-42-1.85-03
0.25
0.20
0.2
(f)
D_1-42-1.85-03
Cut-off
0.15
0.15
0.10
0.1
0.05
0.00
Long
Short
D_1-42-1.85-06
0.25
D_3-42-1.85-03
0.05
Short
Long
Cut-off
0.25
0.25
0.25
0.20
0.20
0.20
0.2
0.15
0.15
0.15
0.15
0.10
0.10
0.05
0.00
Cut-off
Short
0.00
Long
0.10
Cut-off
0.05
0.00
Long
Short
Long
0.1
S_1-42-2.50-03
0.05
S_3-42-2.50-03
D_1-42-2.50-06
Cut-off
Short
Long
Cut-off

Figure 4-5: Series I: Ledge Length- Direct comparisons of Vtest normalized by f'cbwd
68
Short
D_1-42-2.50-06
0.25
0.05
D_1-42-1.85-06
0
Cut-off
S_3-42-2.50-03
S_1-42-2.50-03
(r) D_1-42-2.50-03
S_3-42-1.85-03
Short
Long
D_1-42-1.85-03
D_3-42-1.85-03
S_3-42-1.85-03
D_1-42-1.85-06
15
15.00
15.00
15
12
12.00
12.00
12
9.00
9.00
6.00
6.00
3.00
0
Cut-off
Short
0.00
Long
D_1-42-2.50-03
(f)
D_1-42-1.85-03
Cut-off
3.00
0.00
Long
Short
D_3-42-1.85-03
3
0
Cut-off
Short
Cut-off
Long
15
15.00
15.00
15
12
12.00
12.00
12
9.00
9.00
6.00
3.00
0
Cut-off
Short
0.00
Long
6.00
(r) D_1-42-2.50-03
Cut-off
3.00
0.00
Long
Short
Long
6
S_1-42-2.50-03
S_3-42-2.50-03
D_1-42-2.50-06
0
Cut-off
Short
Long
Cut-off

Figure 4-6: Series I: Ledge Length- Direct comparisons of Vtest normalized by bwd
69
Short
D_1-42-2.50-06
S_3-42-2.50-03
S_1-42-2.50-03
D_1-42-1.85-06
S_3-42-1.85-03
Short
Long
The nomenclature for the comparison groups is similar to that of the specimens
themselves. For example, D_1-42-1.85-03 is comprised of two specimens with deep
ledges (D), tested with one point load (1), a total depth of 42 in. (42), a shear span-todepth ratio of 1.85 (1.85) and 0.3% reinforcement in each direction (03). The blank (_)
denotes the length of the ledge, the only variable that differs between the specimens. A
similar nomenclature is used in each series with the blank (_) representing the variable
currently under investigation.
The same trend was observed in both figures. The normalized shear strength appears to
increase with the length of the ledge regardless of the a/d ratio, web reinforcement ratio,
and ledge depth. This would suggest that increasing the length of the ledge of an
inverted-T straddle bent cap would lead to an increase in shear capacity. It should also be
noted that the two specimens that did not fail in web shear fell within the range of Vtest
values in their respective ledge length and a/d ratios- i.e. their capacity did not define a
maximum or minimum normalized Vtest.
4.3.3 Serviceability Results
In addition to the strength results, serviceability data was collected for each of the tests
compared within Series I, including the first cracking load and crack widths at each load
increment. As the first cracking load is associated with the concrete tensile strength, the
values presented in Table 4-2 are normalized with respect to the square root of the
compressive strength of the concrete.
Because the first cracking load could not be measured for the second test on most beams,
only fourteen of the twenty tests are available in six groups of two or three directly
comparable specimens A delay in cracking with respect to the normalized shear load is
observed with increasing ledge lengths (Figure 4-7). The trend was observed for both a/d
ratios and ledge depths.
70
D_1-42-1.85-03
D_3-42-1.85-03
S_3-42-1.85-03
6.00
6.00
5.00
5.00
4.00
4.00
3.00
3.00
2.00
2.00
1.00
1.00
Cut-off
Short
0.00
Long
0.00
Cut-off
Short
Cut-off
Long
5.00
S_1-42-2.50-03
D_1-42-1.856.00
03
5.00
4.00
4.00
3.00
3.00
2.00
2.00
1.00
1.00
D_1-42-1.85-06
0
Cut-off
Short
6.00
0.00
Long
Short
Long
S_3-42-2.50-03
D_3-42-1.8503
S_3-42-1.8503
0.00
Cut-off
Short
Long
Cut-off
Short
Long
D_1-42-1.85-
Figure 4-7: Series I: Ledge Length- Direct

by bwd
06 comparisons of Vcrack normalized
S_1-42-
The location and angle of first diagonal crack is shown for one group of directly
comparable specimens in Figure 4-8. As with the normalized diagonal cracking load, the
percent of ultimate capacity at which the beam first cracked increased with the ledge
length. This is expected as the longer ledge slightly increases the size of the beam, thus
delaying the first diagonal crack. The location and angle of the initial diagonal crack was
similar for the cut-off and short ledge specimen, but appeared more shallow and further
from the load point for the specimen with the deep ledge.
71
S_3-42-2.5003
SC3-42-1.85-03: 17% ultimate load
SS3-42-1.85-03: 24% ultimate load
SL3-42-1.85-03: first crack at 30% ultimate load
Figure 4-8: First diagonal crack comparison for ledge length specimens
Crack width progressions are shown in Figure 4-9 with twenty specimens presented in
eight groups of two or three directly comparable specimens. Contrary to the observations
of the first cracking load, no clear trend can be distinguished in the crack width
progression plots. In some cases, specimens with longer ledges showed a more
accelerated crack widening, whereas in other cases the widening of cracks in specimens
with cut-off ledges were more accelerated. Therefore it was concluded that the length has
no appreciable effect on the width of the diagonal cracks.
72
60
40
20
0
0.000
0.033
100
80
60
40
20
0
0.00
0.03
D_1-42-1.85-03
D_3-42-1.85-03
S_3-42-1.85-03
D_1-42-1.85-06
100
80
60
40
0.067
20
% of Maximum Applied Load
80
100
0.100
0.000
0
0.00
0.033
0.067
DS1-42-1.85-03
0.000 DC3-42-1.85-03
0.033
0.067
0.100
DS3-42-1.85-03
0.07
0.10
0.00
0.03
0.07
Maximum
DS1-42-1.8503
D_1-42-2.50-03
SL3-42-1.85-03
DL3-42-1.85-03
DL1-42-1.85-03
0.03
0.000 SC3-42-1.85-03
0.033
0.067
0.100
SS3-42-1.85-03
0.10
0.00
0.03
0.07
0.10
Diagonal Crack Width, in.

DC3-
0.100
DS1-42-1.85-06
DL1-42-1.85-06
SC3-42-1.850.00
0.03
0.07
0.10
03
SS3-42-1.8503
DS3-
S_3-42-2.50-03
S_1-42-2.50-03
DS1-42-1.8506
D_1-42-2.50-06
100
80
60
40
0.07
20
DC1-42-2.50-03
0.03
0.07
0.10 0.00
DS1-42-2.50-03
0
0.00
DL1-42-2.50-03
0.03
0.07
0.10
SC1-42-2.50-03
0.03
0.07
0.100.00
SS1-42-2.50-03
0.03
0.07
0.10 0.00
SC3-42-2.50-03
SL1-42-2.50-03
SS3-42-2.50-03
SC1-42-2.50DC1-42-2.500.00
0.03
0.07
0.10
0.00
0.03
0.07
0.10
03
03
Maximum Diagonal SS1-42-2.50Crack Width, in.
DS1-42-2.5003
03
0.10
DS1-42-2.50-06
DL1-42-2.50-06
0.00
SC3-42-2.5003
Figure 4-9: Series I: Ledge Length- Direct comparisons of crack width progression
73
0.03
0.07
0.10
DS1-42-2.5006
In addition to widths, the crack patterns were also recorded by means of photographs
taken at each loading increment. This enabled the formation of new cracks and the
extension of existing cracks to be monitored. The location of the cracks at various
percentages of ultimate applied load are shown in Figure 4-10 for a direct comparison,
with deep ledges, tested at an a/d ratio of 1.85, and 0.3% reinforcement in each direction
and three load points. The sketches have been oriented to represent beams in the field and
it should be noted that no cracking could be observed directly above each load point as
that was the location of the loading frame.
It is more difficult to observe cracking along the direct strut in the beams with long
ledges due to the fact that the cracks did not propagate onto the ledge. The ledge length
appears to have some effect on the angle of the crack between the load and support. The
cut-off ledge results in a much steeper crack when compared to what can be observed for
the beam with the long ledge, especially at lower levels of load. As the load increases the
crack angle appears to decrease, tending more towards a straight line between the inner
corner of the bearing plate and the center of the node above the load point. A similar
trend can be observed in the beam with the short ledge, but the cracking pattern appears
steeper at higher loads due to the fact that the cracks extend to the top of the beam above
the load point. Overall the angle of the cracks observed on the beam with the deep, long
ledge were shallower than those of the short and cut-off ledge. It should also be noted
that the angle did not appear to change with higher levels of ultimate load.
A three point load direct comparison with shallow ledges is presented in Figure 4-11. The
cracking pattern is very similar to the deep ledge specimens and indications of
reinforcement stress, either yield or half of yield, is shown in this figure to demonstrate
the effectiveness of the strain gauges. The gauges were installed along the assumed direct
strut for these specimens tested at an a/d ratio of 1.85 as discussed in Section 3.5.1. The
measured strain increases in the gauges as additional load is applied and the cracks form
and widen. As with the effect of ledge length on the crack location, the reinforcement is
shown to yield earlier in the specimen with the longer ledge when the general crack angle
more closely resembles the angle of the assumed direct strut. At higher levels of load the
cracking pattern and reinforcement stresses are almost indistinguishable.
74
DC3-42-1.85-03
DS3-42-1.85-03
30%
Percent of Ultimate Load
50%
60%
80%
90%
100%
100%
Figure 4-10: Crack patterns for D_3-42-1.85-03 (ledge length varies)
75
DL3-42-1.85-03
SC3-42-1.85-03
30%
50%
50%
30%
60%
60%
80%
80%
90%
90%
100%
100%
SS3-42-1.85-03
fs= fy
Reinforcement Key:
SL3-42-1.85-03
fs= fy
Figure 4-11: Crack patterns and strain gauges for S_3-42-1.85-03 (ledge length varies)
76
4.3.4 Summary- Series I: Ledge Length

Direct comparisons were presented in this section to evaluate the influence of the length
of the ledge on the strength of inverted-T straddle bent caps, the appearance of the first
diagonal crack, and the crack width progression.
The results from twenty two tests have shown that increasing the ledge length increases
the strength of the inverted-T beam and delays the appearance of the first diagonal crack.
Specimens with cut-off ledges were found to have the lowest normalized shear strength
while long ledges generally resulted in greater strength as well as diagonal cracking
loads. No significant difference was observed in the correlation between the maximum
crack width and percent of maximum applied load for the three ledge lengths.
4.4
SERIES II: LEDGE DEPTH
This series was designed to evaluate the effects of the depth of the ledge on the strength
and serviceability of inverted-T straddle bent caps. The results of Series II will be used to
develop design recommendations in regards to the depth of ledges in similar bent caps.
In order to represent the various ledge geometry found in the inspected in-service bent
caps, two different ledge depths were investigated. A ledge depth of half the total section
height was considered a deep ledge and represented the upper bound of the range of
ledge depth seen in the field. Similarly, a ledge depth of a third the total section height
was a shallow ledge and represented the upper bound of the range of ledge depth seen
in the field. Section 3.2.3 provides additional background information on the ledge depth
series.
Twenty of the tests produced ten pairs of directly comparable specimens in which every
parameter was kept constant except the ledge depth. A summary of the experimental
results from the ledge depth series is provided in Table 4-3. All variables are defined in
Section 4.2.
77
Table 4-3: Series II: Ledge Depth- Experimental results.
f ' c Vtest
(psi) (kip)
SC1-42-1.85-03 (le) 4303 463
_C1-42-1.85-03
DC1-42-1.85-03
4330 517
SC1-42-1.85-03b
2996 456
_C1-42-1.85-03b
DC1-42-1.85-03b
3013 424
SS1-42-1.85-03
5721 583
_S1-42-1.85-03
DS1-42-1.85-03
5258 712
SC3-42-1.85-03
5873 483
_C3-42-1.85-03
DC3-42-1.85-03
4568 395
SS3-42-1.85-03
5891 523
_S3-42-1.85-03
DS3-42-1.85-03
4568 454
SL3-42-1.85-03
5037 571
_L3-42-1.85-03
DL3-42-1.85-03 (f) 4202 629
SC1-42-2.50-03 (r) 4281 319
_C1-42-2.50-03
DC1-42-2.50-03
4035 365
SS1-42-2.50-03
5703 398
_S1-42-2.50-03
DS1-42-2.50-03
5389 406
SL1-42-2.50-03
4281 498
_L1-42-2.50-03
DL1-42-2.50-03
4929 510
SS3-42-2.50-03
5891 447
_S3-42-2.50-03
DS3-42-2.50-03
5687 430
(le) Horizontal ledge tie failure in cross section model
Comparison
Vcrack
(kip)
Specimen
0.15
0.14
0.19
0.18
0.13
0.17
0.10
0.11
0.11
0.13
0.14
0.19
0.09
0.11
0.09
0.10
0.15
0.13
0.10
0.10
9.97
8.90
10.54
9.77
9.75
12.42
7.98
7.39
8.62
8.49
10.17
12.27
6.18
7.28
6.67
6.99
9.62
9.19
7.38
7.21
127
NA
NA
NA
NA
172
90
152
126
164
172
276
0
70
157
0
167
0
140
143
2.45
NA
NA
NA
NA
2.99
1.48
2.84
2.08
3.07
3.06
5.39
0.00
1.40
2.63
0.00
3.24
0.00
2.31
2.40
All specimens failed in web shear except for three: DL3-42-1.85-03 failed in flexure, and
SC1-42-1.85-03 and SC1-42-2.50-03 experienced local ledge failures. The maximum
shear value, Vtest, reported in Table 4-2, is the amount of shear carried at the critical
section of the beam at the onset failure, regardless of the failure mode.
Strength results are compared in Figure 4-12 and Figure 4-13. Each plot is a direct
comparison of two specimens in which the ledge depth is the only varying parameter. As
discussed in Section 4.2.1, Vtest is normalized by f'cbwd in Figure 4-12 and by bwd in
Figure 4-13.
78
_C1-42-1.85-03
_C1-42-1.85-03b
_S1-42-1.85-03
_C3-42-1.85-03
0.25
0.25
0.25
0.25
0.20
0.20
0.20
0.20
0.15
0.15
0.15
0.15
0.15
0.10
0.10
0.25
0.20
(le)
0.05
0.05
_C1-421.85-03
0.00
0.00
Shallow
Shallow
Deep
_L3-42-1.85-03
0.25
0.20
Shallow
Shallow
0.00
Deep
_S1-42-2.50-03
Deep
Shallow
0.20
0.20
0.20
0.15
0.15
0.15
0.00
Deep
1.85-03
Shallow
0.05
0.00
Deep
0.10
_C1-422.50-03
0.05
0.10
_S1-422.50-03
Deep
0.00
Shallow
Deep
Shallow

Figure 4-12: Series II: Ledge Depth- Direct comparisons of Vtest normalized by f'cbwd.
79
_S3-422.50-03
_L1-422.50-03
0.05
0.00
Shallow
Deep
_S3-42-2.50-03
_L1-42-2.50-03
0.20
0.10
_S3-421.85-03
0.00
Shallow
0.25
_L3-42(r)
_C3-421.85-03
0.05
0.25
0.05
0.00
0.00
Deep
0.05
0.10
_S1-421.85-03
0.25
0.10
0.05
0.05
_C1-42-2.5-03
0.15
0.10
0.10
_C1-42-
0.25
(f)
0.15
0.10
_S3-42-1.85-03
Deep
_C1-42-1.85-03
_C1-42-1.85-03b
_S1-42-1.85-03
_S3-42-1.85-03
_C3-42-1.85-03
15.00
15.00
15.00
15
12.00
12.00
12.00
12
9.00
9.00
9.00
6.00
3.00
15
12
(le)
_C1-421.85-03
0.00
0
Shallow
Shallow
Deep
_L3-42-1.85-03
15
12
6.00
3.00
0.00
Deep
3.00
0.00
Deep
_S1-42-2.50-03
_C1-42-2.5-03
_S1-421.85-03
3
Deep
Shallow
_L1-42-2.50-03
15.00
15.00
12.00
12
12.00
12.00
9.00
9.00
6.00
3.00
_L3-42(r)
1.85-03
0.00
Deep
_C1-422.50-03
Shallow
Deep
Shallow
6.00
3.00
0.00
Deep
_C3-421.85-03
6.00
_S1-422.50-03
3.00
Deep
_S3-42-2.50-03
_L1-422.50-03
_S3-422.50-03
0.00
Shallow
Deep
Shallow

Figure 4-13: Series II: Ledge Depth- Direct comparisons of Vtest normalized by bwd.
80
_S3-421.85-03
0
Shallow
15
9.00
Shallow
Shallow
6.00
15.00
(f)
_C1-421.85-03b
Deep
Two sets of 42-in. specimens with cut-off ledges, one point load, an a/d of 1.85 and 0.3%
reinforcement were tested in the current experimental program. The second set, denoted
by _C1-42-1.85-03b, resembled corbels in that the beams were designed with ledges with
lengths equal to the length of the load plate. These specimens were used to directly
examine the node above the load point when the load was confined to the width of the
plate and not allowed to spread along the ledge in either direction.
The results shown in Figure 4-12 and Figure 4-13 indicate that the ledge depth does not
influence shear strength. Only in two comparisons, _S1-42-1.85-03 and _L3-42-1.85-03
did the specimens with deep ledges exhibited significantly higher strengths than
specimens with shallow ledges. Considering that about half of the comparisons revealed a
slight decrease in the shear strength of deeper ledges and the inherent variability in shear
tests, one can conclude that ledge depth has no significant effect in the shear strength.
The first cracking loads, Vcrack, for the ledge depth series are presented in Figure 4-14.
Vcrack was normalized by bwd due to the fact that the first cracking load is associated
with the tensile strength of the concrete.
_C3-42-1.85-03
_S3-42-1.85-03
2
_C3-42-1.85-03
1
1
0
_S3-42-1.85-03
Shallow
Deep
Shallow
Deep
_S3-42-2.50-03
_L3-42-1.85-03
6
2
_L3-42-1.85-06
1
_S3-42-2.50-03
0
Shallow
Shallow
Deep
Deep
Figure 4-14: Series II: Ledge Depth- Direct comparisons of Vcrack normalized by bwd.
81
Because the first test on a specimen resulted in cracks on the opposite span, the first
cracking load was not recorded for many of the tests in this series. Only eight tests were
available for direct comparisons in four groups of two specimens. A trend can be
observed in Figure 4-14 in which the specimens with shallow ledges crack at a lower load
than comparable specimens with deep ledges. In other words, increasing the depth of the
ledge delays the appearance of diagonal cracks.
This was confirmed when examining the percentage of ultimate capacity at the
appearance of the first diagonal crack for the direct comparison shown in Figure 4-15. An
increase in diagonal cracking load was observed but the overall location and angle of the
diagonal crack appears very similar. The crack in the deep ledge specimen does start
slightly higher on the web of the beam.
SC3-42-1.85-03: 17% ultimate load
DC3-42-1.85-03: 38% ultimate load
Figure 4-15: First diagonal crack comparison for ledge depth specimens
The progression of crack widths with respect to the applied load for the ledge depth series
is illustrated in Figure 4-16. Twenty specimens are presented in ten groups of two
directly comparable specimens in which the ledge depth was the only parameter varied.
Regarding the crack width progressions, no clear trend can be distinguished in Figure
4-16. In a few cases, the specimens with deeper ledges appeared to show a more
accelerated crack widening but for the most part the maximum crack widths of the two
specimen types were indistinguishable. Ultimately, it can be concluded that the depth of
the ledge does not have a significant effect on the width of diagonal cracks when
compared to the percentage of applied load.
82
_C1-42-1.85-03
_C1-42-1.85-03b
_S1-42-1.85-03
_S3-42-1.85-03
_C3-42-1.85-03
100
80
60
40
20
0
0
0.00
0.03
0.07
0.1
0
0.10 0.00
_L3-42-1.85-03
100
80 100
0.03
0.07
_S1-42-2.50-03
SC1-42-1.8503
0.1
0
0.10 0.00
0.03
0.07
0.1
0
0.10 0.00
_L1-42-2.50-03
SS1-42-1.8503
0.03
0.07
0.10
0.1
_S3-42-2.50-03
SC3-42-1.8503
SS3-42-1.8503
40
20
20
0 0
0 0.000.03
0.00
_C1-42-2.5-03
SC1-42-1.8503
100
0.1
0
0.10 0.00
60
60
40
20
0.07
0.07
80
60 80
40
0.03
0.03
0.07
0.03
0
0.00
0.03
0.07
0.10
0.00
0.10 0.00
0.1
0
0.03
0.07
0.1
0
0.03
0.07 0.1
0
0.03
0.00
0.10
0.10
0.07
0.10
SC1-42-1.85-03
DC1-42-1.85-03
Deep Ledge Depth
Shallow LedgeSC1-42-2.50Depth
SS1-42-2.50SL3-42-1.85-SC1-42-1.85-03
DC1-42-1.85-0303
03
03
0.07
0.07
0.10 0.00
0.1
0
SL1-42-2.5003
Figure 4-16: Series II: Ledge Depth- Direct comparisons of crack width progression.
83
0.03
0.07
0.10
0.1
SS3-42-2.5003
Examining the crack patterns and reinforcement strains shown in Figure 4-17, there
appears to be little difference between the shallow and deep ledge specimen. The only
chnage being the development of cracks above the approximate strut along the web strain
gauges for the shallow ledge specimen. The ledge depth did not have an influence on the
crack pattern or yielding of the reinforcement for the other direct comparison specimens.
DC3-42-1.85-03
SC3-42-1.85-03
30%
30%
50%
50%
50%
60%
60%
60%
80%
80%
80%
90%
90%
90%
100%
100%
100%
30%
fs= fy
Reinforcement Key:
fs= fy
Figure 4-17: Crack patterns and strain gauges for _C3-42-1.85-03 (ledge depth varies)
4.4.4 Summary: Series II: Ledge Depth

Direct comparisons were presented in this section to evaluate the influence of the depth
of the ledge on the shear strength of inverted-T straddle bent caps, the appearance of first
diagonal crack, and the crack width progression.
84
The results from twenty tests have shown that increasing the ledge depth does not have a
significant effect on the strength of the inverted-T beams. There was also no significant
effect on the correlation between the maximum crack width and the percent of maximum
applied load. It was, however, observed that increasing the depth of the ledge can delay
the appearance of the first diagonal crack.
4.5
SERIES III: WEB REINFORCEMENT
This series was designed to evaluate the effects web reinforcement on the strength and
serviceability behavior of the reinforced concrete inverted-T straddle bent caps. The
results of Series III will be used to evaluate recommendations of minimum horizontal and
vertical reinforcement.
Two different web reinforcement ratios were investigated. The minimum horizontal and
vertical reinforcement requirement for deep beams designed using strut-and-tie models
according to AASHTO LRFD 2012 is an orthogonal grid of reinforcement at each face
such that the ratio of total reinforcement to the gross concrete area is equal to 0.003
(0.3%). The spacing of the reinforcement is limited to 12 in. The purpose of this
requirement is to control the width of cracks and allow for the redistribution of internal
stresses. For additional discussion on the minimum reinforcement requirements, refer to
Birrcher (2008). By constructing and testing specimens with this reinforcement ratio, the
performance of the minimum reinforcement ratio in inverted-T beams can be analyzed.
The reinforcement ratio of 0.006 (0.6%) was chosen to provide an upper bound on the
maximum reinforcement ratio (0.57%) found in the in-service distressed bent caps. Two
specimens were also constructed with a reinforcement ratio of 0.6% in the vertical
direction and 0.3% in the horizontal. The purpose of these tests was to examine the
effects of mixed reinforcement in inverted-T straddle bent caps; as is common in current
design practices when sectional shear design is used. Section 3.2.4 includes additional
background information on this series.
Fourteen tests produced six groups of two or three directly comparable specimens in
which every parameter was kept constant except the reinforcement ratio. A summary of
the experimental results for Series III is provided in Table 4-4. All variables are defined
in Section 4.2. Strength results are normalized as discussed in Section 4.2.1
85
Table 4-4: Series III: Web Reinforcement- Experimental results.
f 'c
(psi)
5258
DS1-42-1.85-03
DS1-42-1.85-_ DS1-42-1.85-06/03 4173
DS1-42-1.85-06
5024
DL1-42-1.85-03
4929
DL1-42-1.85-_
DL1-42-1.85-06
4830
SL3-42-1.85-03
5037
SL3-42-1.85-_
SL3-42-1.85-06
5250
DS1-42-2.50-03
5389
DS1-42-2.50-_ DS1-42-2.50-06/03 4173
DS1-42-2.50-06
5088
DL1-42-2.50-03
4929
DL1-42-2.50-_
DL1-42-2.50-06
4986
SS3-42-2.50-03
5891
SS3-42-2.50-_
SS3-42-2.50-06 (f) 6255
Comparison
Specimen
Vtest
(kip)
712
739
621
626
741
571
744
406
539
503
510
622
447
516
Vcrack
(kip)
0.17
0.22
0.16
0.16
0.19
0.14
0.18
0.10
0.16
0.13
0.13
0.16
0.10
0.10
12.42
14.49
11.09
11.28
13.48
10.17
13.00
6.99
10.59
8.93
9.19
11.15
7.38
8.25
172
NA
188
242
168
172
154
NA
115
NA
NA
NA
140
115
3.00
NA
3.35
4.36
3.06
3.06
2.69
NA
2.25
NA
NA
NA
2.31
1.84

Direct comparisons of the strength results are provided in Figure 4-18 and Figure 4-19.
Each plot is a direct comparison of two or three specimens. Vtest is normalized by f'cbwd in
Figure 4-18 and by bwd in Figure 4-19.
86
0.25
DS1-42-1.85-_
DL1-42-1.85-_
SL3-42-1.85-_
0.25
0.25
0.20
0.20
0.20
0.15
0.15
0.15
0.10
0.10
0.10
0.05
0.05
0.05
0.00
0.6%
0.6%
0.00
= 0.6%
v0.3%
= 0.3% v0.6%
h = 0.6%
DS1-42-1.85h = 0.3%
_
0.00
v = 0.3% 0.6%
h = 0.3% 0.3%
0.25
DS1-42-2.50-_
0.25
SS3-42-2.50-_
DL1-42-2.50-_
0.20
0.20
0.15
0.15
0.15
0.10
0.10
0.10
0.05
0.05
0.05
0.00
0.00
0.6%
0.6%
0.00
0.3%
0.6%
v = 0.3% v = 0.6%
h = 0.3% h = 0.6%
DS1-42-2.50_
SL3-42-1.85_
0.25
0.20
v = 0.3% 0.6%
h = 0.3% 0.3%
= 0.6%
v0.3%
= 0.3% v0.6%
h = 0.3% h = 0.6%
DL1-42-1.85_
(f)
= 0.6%
v0.3%
= 0.3% v0.6%
h = 0.3% h = 0.6%
DL1-42-2.50_
Figure 4-18: Series III: Web Reinforcement- Direct comparisons of Vtest normalized by
f'cbwd.
As seen in both Figure 4-18 and Figure 4-19, an increase in reinforcement ratio appears
to lead to an increase in shear strength, even for beams with a/d ratios of 1.85. In
specimens with a/d of 1.85, shear capacity is governed by the compression strut and
nodes spanning from load to support. The purpose of the web reinforcement in singlepanel beams (a/d=1.85) is to resist the transverse tensile forces that develop in a bottleshaped strut. Increasing the amount of web reinforcement above a minimum amount
required to resist transverse stresses was not found to increase strut capacity in the beams
tested with reinforcement ratios of 0.2% and 0.3% in Project 0-5253. Although no
difference was observed in the rectangular beams, increasing the reinforcement ratio from
0.3% to 0.6% appears to increase the strength of the direct strut in inverted-T beams. In
specimens with a/d of 2.5 however, the tension tie in the shear span limits the shear
capacity. It would therefore be expected that a larger increase in shear strength with a
larger amount of web reinforcement would be recorded for the beams with a/d ratios of
87
SS3-42-2.50_
2.50. However, this was not the case, which indicates that the concrete may play as large
a role in the shear strength of longer beams as in shorter beams.
15
DS1-42-1.85-_
DL1-42-1.85-_
SL3-42-1.85-_
15.00
15.00
12
12.00
12.00
9.00
9.00
6.00
6.00
3.00
3.00
v = 0.3%
h = 0.3%
0.6%
0.3%
0.00
0.6%
0.6%
DS1-42-2.50-_
= 0.6%
v 0.3%
= 0.3% v 0.6%
h = 0.3% h = 0.6%
DS1-42-1.85_
0.00
15.00
15.00
12
12.00
12.00
9.00
9.00
6.00
6.00
3.00
3.00
v = 0.3%
h = 0.3%
0.6%
0.3%
0.00
0.6%
0.6%
0.6%
v 0.3%
= 0.3% v =
0.6%
=
0.6%
DS1-42-2.50=
0.3%
h
h
0.00
(f)
= 0.6%
v 0.3%
= 0.3% v 0.6%
=
0.3%
h = 0.6%
h
DL1-42-2.50-
Figure 4-19: Series III: Web Reinforcement- Comparisons of Vtest normalized by bwd.
All but two beams tested in the experimental program had equal reinforcement ratios in
the horizontal and vertical direction. A significant increase in shear capacity can be seen
in Figure 4-18 and Figure 4-19 between specimens with v = 0.6% to h = 0.3% and those
with both ratios at 0.3%. In fact the specimens with v = 0.6% to h = 0.3% showed
slightly higher shear strength than those with both ratios at 0.6%. This indicates that the
vertical reinforcement ratio likely has a greater effect on the shear strength of inverted-T
beams. In previous research of reinforced concrete deep beams, horizontal reinforcement
above and beyond 0.3% was found to have little effect on the shear strength (Birrcher
(2009)).
88
SL3-42-1.85_
SS3-42-2.50-_
DL1-42-2.50-_
15
= 0.6%
v 0.3%
= 0.3% v 0.6%
h = 0.3% h = 0.6%
DL1-42-1.85_
SS3-42-2.50_

The cracking loads and the width of diagonal cracks were used to assess the effect of web
reinforcement on the serviceability performance of the specimens in this series. As with
the strength results, directly comparable specimens were used to evaluate the
serviceability behavior on inverted-T deep beams. This allowed all other variables to be
kept constant to isolate any effects that could be attributed directly to the amount of web
reinforcement.
The diagonal cracking loads for the web reinforcement series are presented in Figure
4-20. Eight tests were available for comparisons in four groups. As seen in Figure 4-20,
the results revealed a slight decrease in the normalized cracking load with an increase in
the quantity of reinforcement for most of the beams. Such a result was not expected as
significant strains are not typically experienced by the reinforcement prior to cracking .
However, as the observed trend is slight, it may be concluded that the web reinforcement
has no appreciable effect on the diagonal cracking load.
DS1-42-1.85-_
DL1-42-1.85-_
6.00
5.00
4.00
3.00
2.00
1.00
v 0.3%
= 0.3%
h = 0.3%
v 0.6%
= 0.6%
h = 0.6%
0.00
6.00
5.00
4.00
3.00
2.00
1.00
v 0.3%
= 0.3%
h = 0.3%
v 0.3%
= 0.3%
h = 0.3%
v 0.6%
= 0.6%
h = 0.6%
SS3-42-2.50-_
SL3-42-1.85-_
6
DL1-42-1.85-_
DS1-42-1.85-_
v 0.6%
= 0.6%
h = 0.6%
0.00
SL3-42-1.85-_
v 0.3%
= 0.3%
h = 0.3%
SS3-42-2.50-_
v 0.6%
= 0.6%
h = 0.6%
Figure 4-20: Series III: Web Reinforcement- Direct comparisons of normalized Vcrack.
An example of the location and angle of the first diagonal crack for two directly
comparable specimens is provided in Figure 4-21. The beam with 0.3% web
89
reinforcement in both directions cracked at a slightly lower ultimate load and had a
slightly steeper diagonal crack.
DS1-42-1.85-03: 24% ultimate load
Figure 4-21: First diagonal crack comparison for reinforcement ratio specimens
The maximum diagonal crack widths were examined using all fourteen directly
comparable beams as shown in Figure 4-22. It is interesting to note that in half of the
comparisons, the web reinforcement ratio did not appear to greatly affect the maximum
crack widths in terms of the percentage of the maximum applied load. In the other half of
the groups, there was a significant difference. The specimens with a web reinforcement
ratio of 0.3% were subjected to a given crack width at a smaller percentage of maximum
applied than those with 0.6%. This indicated that inverted-T bent caps with larger
amounts of web reinforcement are closer to their ultimate load at a given crack width.
Conversely, at a given percent of ultimate load, higher reinforcement ratios result in
narrower cracks. The two specimens with mixed reinforcement ratios showed crack
widths between those of the two other reinforcement layouts. However, the mixed
reinforcement specimen cracks were closer to those of the beams reinforced with 0.3% in
both directions.
90
DS1-42-1.85-_
100
80
80
60
60
0.0040 0.03
20
0
0.00
DL1-42-1.85-_
100
100
80
60
0.0040
0.03
0.07
0.10
0.0040 0.03
DS1-42-1.85-03
20
DS1-42-1.85-06/03
100
DS1-42-1.85-06
0
80
0.00
0.03
0.07
0.10
DS1-42-1.85-03
60
SL3-42-1.85-_
DS1-42-1.85-06/03
100
40
DS1-42-1.85-06
80
20
60
0
0.0040 0.03
0.07
0.10
SL3-42-1.85-03
SL3-42-1.85-06
20
0
0.00
0.03
0.07
DL1-42-2.50-_
100
0
0.00
0.10
60
60
0.03
0.07
0
0.00
0.10
0.00
40
DL1-42-2.50-03
DL1-42-2.50-06
20
0.03
0.07
0.03
0.07
DS1-42-2.50-_
0.07
0.03
0.03
0.07
0.07
0
0.00
0.10
DS1-42-2.50-03
DS1-42-2.50-06/03
DS1-42-2.50-06
SS3-42-2.50-03
SS3-42-2.50-06
0.03
MaximumDL1-42-2.50-03
DL1-42-2.50-06
DL1-42-1.85-03
DL1-42-1.85-06
0.10
20
0.10
0.10
0.10
DS1-42-2.50-03
DS1-42-2.50-06/03
DS1-42-2.50-06
SS3-42-2.50-_
SL3-42-1.85-03
100
SL3-42-1.85-06
80
0.10
DL1-42-1.85-03
DL1-42-1.85-06
20
80
0.0040
0.07
0.07
0.10
SS3-42-2.50-03
SS3-42-2.50-06
Figure 4-22: Series III: Web Reinforcement- Direct comparisons of crack width
progression.
Similar trends can be observed when all test specimen maximum crack widths are
investigated. The maximum diagonal crack widths for nineteen specimens from the
experimental program that were tested at an a/d ratio of 1.85 are plotted versus the
percentage of maximum applied load in Figure 4-23.
91
Percent of Maximum Applied Load (%)
Although there is greater scatter in the data for beams with a reinforcement ratio of 0.3%
in both orthogonal directions, the general trend clearly shows that quantity of web
reinforcement has a direct impact on the width of diagonal cracks. As can be seen in the
figure, a larger reinforcement ratio results narrower cracks for a given load ratio.
Therefore a beam with 0.6% in each direction is closer to its ultimate capacity at a given
crack width than a beam with a reinforcement ratio of 0.3%. The beam with 0.6% in the
vertical direction and 0.3% in the horizontal direction had crack widths at lower loads
comparable to those of beams reinforced with 0.3% in both directions. As the percentage
of maximum applied load increased past 50% however, cracks width of the mixed
reinforcement ratio specimens were in-between those of the two uniformly reinforced
specimens.
100 100
100100
v = 0.3%/0.3%
0.3/0.3
0.3%/0.3%
0.3%/0.3%
h = 0.3%
0.6/0.3
0.6%/0.3%
0.6%/0.6%
0.6%/0.3%
h = 0.3%0.6%/0.6%
v = h =0.6/0.6
0.6%
0.6%/0.3%
0.6%/0.6%
v = 0.6%,
90
80 80
80 80
70
60 60
60 60
50
40 40
40 40
30
33%
20 20
20 20
a/d = 1.85
19 tests
10
0.016
0 00 0
0.04
0.00 0.00 0.02
0.02 0.02 0.04
0.040.04
0 0.00
0.02
0.06
0.06 0.06
0.08 0.08 0.08
0.10
0.10 0.10
0.060.08
0.1

Figure 4-23: Diagonal crack widths for specimens tested at a/d of 1.85
The crack width data in Figure 4-23 can be compared to the crack width limit of 0.016 in.
and the estimated service load of 33% ultimate capacity as discussed in Section 4.2.2. At
the estimated service load, the maximum diagonal crack widths for the specimens with
0.3% web reinforcement in each direction were at or below the tolerable crack width
limit. Near first cracking and at the expected service load, the specimens with 0.6%
reinforcement ratio saw a limited reduction in the maximum diagonal crack widths
compared with those of the more lightly reinforced specimens. The effects of higher
reinforcement ratios only became more significant at load levels higher than service. One
can therefore conclude that providing at least 0.3% reinforcement in each orthogonal
92
direction can adequately restrain the widths of diagonal cracks at service loads.
Additional reinforcement beyond the recommended minimum would only have limited
effect on crack widths at service level loads. It should be noted that the crack width of
0.016 in. and the estimated service load should not be treated as hard limits, but rather as
a general benchmark to gauge the effectiveness of the reinforcement.
Crack width data from fourteen tests with a/d=2.5 are presented in Figure 4-24. Results
for the longer shear spans are consistent with observations made for beams tested at an
a/d of 1.85; but results showed less scatter.
100 100
100100
v = 0.3%/0.3%
0.3/0.3
0.3%/0.3%
0.3%/0.3%
h = 0.3%
0.6/0.3
0.6%/0.3%
0.6%/0.6%
0.6%/0.3%
h = 0.3%0.6%/0.6%
v = h =0.6/0.6
0.6%
0.6%/0.3%
0.6%/0.6%
v = 0.6%,
90
80 80
80 80
70
60 60
60 60
50
40 40
40 40
30
33%
20 20
20 20
a/d = 2.50
14 tests
10
0.016
0 00 0
0.00 0.00 0.02
0.02 0.02
0 0.00
0.02
0.04
0.04
0.040.04
0.06
0.06 0.06
0.08 0.08 0.08
0.10
0.10 0.10
0.060.08
0.1

Figure 4-24: Diagonal crack widths for specimens tested at a/d of 2.50
As shown in Figure 4-25, increased amounts of web reinforcement was also shown to
affect the crack distribution. It is understood that as the reinforcement ratio increased, the
width of shear cracks is reduced (Figure 4-22) and the distribution becomes more spread
out. This is confirmed for inverted-T beams as well by the increased amount of cracking
in specimen SC3-42-1.85-06 shown in Figure 4-25. Not only did the additional stirrups
and skin reinforcement increase the shear capacity of the specimens, it also delayed first
yield of the web reinforcement slightly.
93
SC3-42-1.85-06
SC3-42-1.85-03
30%
50%
60%
80%
90%
100%
fs= fy
Reinforcement Key:
fs= fy
Figure 4-25: Crack patterns and strain gauges for SC3-42-1.85-_ (reinforcement varies)
4.5.4 Summary
The influence of the amount of web reinforcement on the strength, appearance of first
diagonal crack, and the progression of the maximum crack width was investigated.
Increasing the web reinforcement ratio was found to increase shear strength of inverted-T
beams, both at a/d ratios of 1.85 and 2.50. The reinforcement ratio had little effect on the
diagonal cracking load. Larger amounts of web reinforcement were found to better
restrain the widths of diagonal cracks. Consequently, inverted-T beams with higher levels
of web reinforcement are more likely to be closer to their ultimate capacity at a given
crack width than those with a lower reinforcement ratio.
94
The minimum reinforcement ratio of 0.3% in each orthogonal direction recommended by

TxDOT Project 0-5253 for adequate crack width restraint in deep beams was also found
to adequately restrain cracks at service loads in inverted-T beams.
4.6
SERIES IV: NUMBER OF POINT LOADS
In this section, the effect of multiple point loads on the strength and serviceability
behavior of inverted-T beams is investigated. In order to quantify the effect of multiple
girders framing into bridge bent caps, twelve specimens were tested with either one load
point or three load points along the length of the beam.
Twelve tests were grouped into six pairs of directly comparisons in which the number of
point loads, either one or three, was the only parameter that was varied. A summary of
the experimental results from the number of point loads series is provided in Table 4-5.
The strength results are normalized as discussed in Section 4.2.1.
Table 4-5: Series IV: Number of Point Loads- Experimental results.
f 'c
(psi)
5721
SS1-42-1.85-03
SS_-42-1.85-03
SS3-42-1.85-03
5891
DS1-42-1.85-03
5258
DS_-42-1.85-03
DS3-42-1.85-03
4568
DL1-42-1.85-03
4929
DL_-42-1.85-03
DL3-42-1.85-03 (f) 4202
SC1-42-2.50-03 (r) 4281
SC_-42-2.50-03
SC3-42-2.50-03
5873
SS1-42-2.50-03
5703
SS_-42-2.50-03
SS3-42-2.50-03
5891
DS1-42-2.50-03
5389
DS_-42-2.50-03
DS3-42-2.50-03
5687
(r) Shear friction failure
Comparison
Specimen
Vtest
(kip)
583
523
712
454
626
629
319
329
398
447
406
430
Vcrack
(kip)
0.13
0.11
0.17
0.13
0.16
0.19
0.09
0.07
0.09
0.10
0.10
0.10
9.75
8.62
12.42
8.49
11.28
12.27
6.18
5.44
6.67
7.38
6.99
7.21
NA
126
172
164
242
276
NA
113
157
140
NA
143
NA
2.08
2.99
3.07
4.36
5.39
NA
1.87
2.63
2.31
NA
2.40
It is important to note that all specimens in this series failed in web shear except DL3-421.85-03 and SC1-42-2.50-03, which failed in flexure and shear friction respectively. The
value reported for Vtest is the maximum shear carried at the critical section at the onset of
failure, regardless of the failure mode.
95

Direct comparisons of the strength results are provided in Figure 4-26 and Figure 4-27.
Vtest is normalized by f'cbwd in Figure 4-26 and by bwd in Figure 4-27. In each subplot the results are compared for specimens in which every parameter was kept constant
except the number of point loads.
SS_-42-1.85-03
DS_-42-1.85-03
DL_-42-1.85-03
0.25
0.25
0.25
0.20
0.20
0.20
0.15
0.15
0.15
0.10
0.10
0.10
0.05
0.05
0.05
0.00
0.00
One
Three
0.00
One
Three
One
0.25
SS_-42-2.50-03
SS_-42-1.850.25
03
0.20
0.20
0.20
0.15
0.15
0.15
0.10
0.10
0.05
0.05
0.25
SC_-42-2.5-03
0.10
(f)
Three
DS_-42-2.50-03
DS_-42-1.8503
DL_-42-1.8503
(r)
0.05
0.00
0.00
One
Three
0.00
One
Three

SC_-42-2.50(r) Shear friction failure of the web-to-ledge
interface
03
One
Three
SS_-42-2.5003
f'cbwd.
Only two comparisons, DS_-42-1.85-03 and DL_-42-1.85-03, in Figure 4-26 and Figure
4-27 show a significant difference between the strength of the two directly comparable
specimens. These specimens, however, reveal contradictory trends. The remaining plots
suggest similar strengths for specimens with one and three point loads. Observations
from the results indicated that the number of point loads does not have an appreciable
effect on the strength of the inverted-T specimens within the range of parameters studied.
This is a significant observation as experimental specimens are often tested with one
point load due to laboratory restrictions, whereas in-service bent caps support multiple
bridge girders. Since the number of loads did not affect the shear strength of the
96
DS_-42-2.5003
specimens, test results from single point load specimens can be confidently used to
develop design recommendations for beams subjected to multiple concentrated loads.
SS_-42-1.85-03
DS_-42-1.85-03
15
15.00
12
12.00
12.00
9.00
9.00
6.00
6.00
3.00
3.00
0.00
0
One
Three
One
15
15.00
SS_-42-2.50-03
SS_-42-1.8515.00
03
12
12.00
12.00
9.00
9.00
6.00
6.00
3.00
3.00
SC_-42-2.5-03
(r)
3
0.00
0
One
Three
(f)
0.00
One
Three
DL_-42-1.85-03
15.00
Three
DS_-42-2.50-03
DS_-42-1.8503
DL_-42-1.8503
0.00
One
Three

SC_-42-2.50(r) Shear friction failure of the web-to-ledge
interface
03
One
Three
SS_-42-
bwd.
Not only was the normalized shear strength unaffected by the addition of load points, the
failure mode was also unchanged. For beams tested at an a/d ratio of 2.50, yielding of the
stirrups, consistent with a two panel tie failure, was observed to be the dominate failure
mechanism. For the specimens tested at an a/d ratio of 1.85, crushing of the direct strut
and the corresponding strut-to-node interfaces occurred for one and three point load tests.
Therefore it can be concluded that shifting the center of the applied load further from the
support for specimens with three loading points did not change the behavior of the
inverted-T specimens.
97
DS_-42-

First cracking loads, normalized by
for the series are presented in Figure 4-28.
The first cracking load of six tests were available for three pairs of comparable specimens
in which every parameter was kept constant except the number of loading points.
DS_-42-1.85-03
SS_-42-2.50-03
DL_-42-1.85-03
0
One
Three
One
Three
DL_-42-1.85DS_-42-1.85Figure 4-28: Series IV: Number of Point Loads- Direct comparisons of Vcrack.
03
03
0
One
Three
No clear trend is indicated by the results in Figure 4-28. One comparison shows a similar
normalized cracking load, while another shows an increase, and the third indicated a
slight decrease when the inverted-T beam was subjected to multiple point loads. Similar
crack angles, with the beam tested at one point being slightly steeper are shown in Figure
4-29. It does, however show an increase in the percentage of ultimate load at which the
first diagonal crack occurred. Based on these limited results it appears that no trend exists
between the number of point loads and the load at which an inverted-T beam is likely to
crack due to shear forces.
Figure 4-29: First diagonal crack comparison for number of point load specimens
The serviceability performance of the test specimens was also evaluated by plotting the
progression of the maximum width of the diagonal cracks (Figure 4-30). No trend
relating crack width progression and number of point loads can be observed in Figure
4-30. Therefore it can be concluded that multiple point loads do not have a significant
98
SS_-42-2.5003
100
80
effect on the progression of crack widths with respected to the percentage of ultimate
capacity. This confirms the applicability of experimental serviceability results from
single point loaded specimens to in-service members subjected to multiple loads.
60
SS_-42-1.85-03
40
20
0
0.00
0.03
100
80
80
60
60
40 0.07
0.10
0
0.00
0.03
0.00
0.03
0.03
80
0.07
40
80
40 0.07
0.10
0
0.00
0.03
0.03
40 0.07
0.07
0
0.00
0.10
60
60
20
0
0.00
0.00
0.03
400.07
SS3-42-2.50-03
0.07
SC1-42-2.50-03
SC3-42-2.50-03
0.03
0.07
0.10
0
0.00
0.10
SC1-42-2.50-03
SC3-42-2.50-03
0.10
DS1-42-2.50-03
DS3-42-2.50-03
20
SS1-42-2.50-03
0.03
DS1-42-1.85-03
DS3-42-1.85-03
DL1-42-1.85-03DS_-42-2.50-03
100
DL3-42-1.85-03
80
0.10
0.10
0.10
20
80
40 0.07
0.07
60
0
0.00
DL1-42-1.85-03
DL3-42-1.85-03
20
0.03
80
20
60
DS1-42-1.85-03
DS3-42-1.85-03
SS1-42-1.85-03SC_-42-2.5-03
100
SS3-42-1.85-03
SS_-42-2.50-03
0.03
0
0.00
0.10
60
DL_-42-1.85-03
100
0.10
20
100
0.00
40 0.07
SS1-42-1.85-03
100
SS3-42-1.85-03
20
0.00
DS_-42-1.85-03
100
0.03
0.07
0.10
Maximum Diagonal
Crack Width, in.
SS1-42-2.50-03
DS1-42-2.50-03
SS3-42-2.50-03
Figure 4-30: Series IV: Number of Point Loads- Direct comparisons of crack widthDS3-42-2.50-03
progression.
As with the crack width progression, the location and development of cracks at various
load stages did not indicated an effect due to the number of point loads as shown in
99
Figure 4-31. In general the cracks have the same angle of inclination and distribution
between the two specimens compared in the figure. The arching action, or formation of
the bottle-shaped strut, can be seen in the crack patterns of both figures, with a clearer
picture in SS1-42-1.85-03. It should be noted that this was the second test on this beam so
the one point loaded specimen did have a few preexisting cracks (marked in lighter gray).
SS3-42-1.85-03
SS1-42-1.85-03
30%
50%
60%
80%
90%
100%
Figure 4-31: Crack patterns for SS_-42-1.85-03 (number of point loads varies)
4.6.4 Summary
The direct comparisons presented in this section have shown that the number of point
loads has a negligible effect on the strength and the appearance of the first diagonal
crack. No effect was observed on the crack width progression or crack distribution of
inverted-T beams with the application of increasing loads.
100
4.7
SERIES V: WEB DEPTH
The effects of web depth on the strength and serviceability behavior of inverted-T beams,
are presented in this section. The purpose of this series is to extend the conclusions drawn
for the 42 in. deep specimens to larger inverted-T bent caps. Most of the specimens in the
experimental program were constructed with a total depth of 42 in. Two beams had total
depth of 75 in. and are discussed in in light of results from tests on the 42 in. deep
specimens.
Although no comparable large-scale inverted-T tests were available in the literature, an
investigation on the effect of web depth on compression-chord loaded beams was
performed by TxDOT Project 0-5253. Based on previous research studies, the length of
the bearing plates appeared to affect the strength results of deep beams as the effective
depth increased. Rather than link the size of the bearing plates to the depth of the
member, the plate dimensions were selected to create similar size nodal regions for each
section depth.
The four tests, two of each depth, discussed in this section are summarized in
The strength results are normalized as discussed in Section 4.2.1.
f ' c Vtest
Vcrack
Comparison TableSpecimen
4-6: Series VI: Web Depth- Experimental results.
(kip)
(psi) (kip)
f 'c V
5721
583
0.13
9.75 VNA
test
crack
Comparison SS1-42-1.85-03
Specimen
(kip)
(psi) (kip)
SS1-_-1.85-03 SS1-75-1.85-03 (p) 5925
913 0.11
8.28
260
SS1-75-1.85-03b
2867
9.72
346
5721 745
583 0.18
0.13
9.75
NA
SS1-42-1.85-03
SS1-_-1.85-03
SS1-42-2.50-03
5703
6.67
157
SS1-75-1.85-03
2867 398
745 0.09
0.18
9.72
346
SS1-_-2.50-03
SS1-75-2.50-03
SS1-42-2.50-03
5703 398 0.09
6.67
157
SS1-_-2.50-03
SS1-_-2.50-06 SS1-75-2.50-06
688 0.08
6.01
232
SS1-75-2.50-03 (p) 6404
5158 649*
0.09
6.31
225
Table 4-1.
NA
2.36
4.51
NA
2.63
4.51
2.63
2.02
2.19
*Although this beam failed due to punching shear, it was determined to be within 10% of
its ultimate capacity and the maximum measured shear is reported herein.
Direct comparisons of the strength results of two pairs of specimens are provided in
Figure 4-32 and Figure 4-33. As discussed in Section 4.2.1, Vtest is normalized by f'cbwd
in Figure 4-26 and by bwd in Figure 4-27. It is interesting to note that for the
specimens tested at an a/d of 1.85, there was an increase in shear strength when
normalized by f'cbwd in the first figure, but the shear strength is essentially equal when
they are normalized by bwd. For the specimens tested at an a/d ratio of 2.50, there was
essentially no difference between the normalized strength for the 42 in. and 75 in. deep
101
beams. Given the limited data set, a definitive conclusion cannot be made regarding beam
size effects. However, it appears that there are no significant size effects within the size
range tested. This indicates that strength results from 42 in. deep beams are applicable to
larger beams.
SS1-_-1.85-03
SS1-_-2.50-03
0.25
0.25
0.20
0.20
0.15
0.15
0.10
0.10
0.05
0.05
0.00
SS1-_-1.85-03
(p)
SS1-_-2.50-03
0.00
42 in.
75 in.
42 in.
75 in.

Web Depth, h (in.)
Figure 4-32: Series VI: Web Depth- Direct comparisons of Vtest normalized by f'cbwd.
SS1-_-1.85-03
SS1-_-2.50-03
15
15.00
12
12.00
9.00
6.00
3.00
SS1-_-1.85-03
(p)
SS1-_-2.50-03
0.00
42 in.
75 in.
42 in.
75 in.

Web Depth, h (in.)
Figure 4-33: Series VI: Web Depth- Direct comparisons of Vtest normalized by bwd.

The diagonal cracking loads for the web depth series is presented in Figure 4-34. The first
cracking load, Vcrack, was not available for specimen SS1-42-1.85-03 so only one
comparison can be made for the first diagonal cracking with respect to different web
depths. As shown in Figure 4-34, there is a slight decrease in the normalized first
cracking load with an increase in web depth.
102
SS1-_-2.50-03
6
5
4
3
2
SS1-_-2.50-03
1
0
42 in.
75 in.
Web Depth, h (in.)

Figure 4-34: Series VI: Web Depth- Direct comparisons of Vcrack normalized by bwd.
Comparing the angle and location of the first diagonal crack shows a very similar result
in Figure 4-35. The percentage of ultimate load at the appearance of the first crack were
also almost identical. Based on the limited data and conclusions from the compressionchord study, web depth does not appear to be a factor in the diagonal cracking load of an
inverted-T beam.
SS1-42-2.50-03: first crack at 39% ultimate load
SS1-75-2.50-03: first crack at 35% ultimate load
Figure 4-35: First diagonal crack comparison for web depth specimens
103
100
80
The maximum diagonal crack widths are presented for the directly comparable specimens
in Figure 4-36. It is interesting to note that regardless of the a/d ratio, there is no
noticeable difference between the crack width progression of the 42-in. and 75-in. beam.
This is an important observation as the relationship between the maximum diagonal crack
width and the reserve strength does not depend on the depth of the inverted-T beam. Thus
a correlation between the diagonal crack width and the ultimate capacity of a member can
be made regardless of the height of the bent cap.
60
SS1-_-2.50-03
SS1-_-1.85-03
100
20
0
0.00
% of Maximum
Applied Load
40
0.03
80
60
40 0.07
0.10
20
0
0.00
0.03
0.00
0.03
0.07
0.10
SS1-42-1.85-03
SS1-42-2.50-03
SS1-75-1.85-03b
SS1-75-1.20-03
0.07
0.10
0.00
0.03
0.07
0.10

SS1-42-1.85-03
Figure 4-36: Series VI: Web Depth- Direct comparisons

of crack width progression.
SS1-75-1.85-03b
SS1-42-2.50-03
SS1-75-2.50-03
The plot of the maximum crack widths of the four specimens discussed in this section
versus the corresponding percent of maximum applied load are presented in Figure 4-37.
All four specimens had reinforcement ratios of 0.3% in each orthogonal direction. No
trend could be discerned in crack width progression for the two web depths. The results
readily suggest that the width of diagonal cracks are not dependent on web depth at a
given percentage of maximum applied load as the data is pretty much indistinguishable
for the four beams.
104
42-in. Deep Beam
75-in. Deep Beam
100
80
60
40
20
v = h = 0.3%
4 tests
0
0.00
0.02
0.04
0.06
0.08
0.10

Figure 4-37: Diagonal crack widths for specimens with 0.3% reinforcement both directions
The distribution of cracks at various load stages is shown in Figure 4-38. Although a
larger number of cracks form on the 75-in. deep specimen, the spacing between cracks
and the overall pattern is very similar between the two beams. The cracks begin forming
in a direct strut between the corner of the support plate and the area above the load plate.
As the load is increased and the strut begins to spread, additional cracks form in an
arching shape similar to the assumed bottle-shape of the strut.
105
SS1-75-1.85-03
SS1-42-1.85-03
30%
50%
60%
80%
90%
100%
Figure 4-38: Crack patterns for SS1-_-1.85-03 (web depth varies)
106
4.7.3 Summary
From the limited direct comparisons between 75-in. and 42-in. deep beam tests
performed in the experimental program, no notable effect of web depth was discovered
on the strength and serviceability behavior of the inverted-T beams. Increasing the web
depth appeared to increase the shear strength of the beam tested at an a/d ratio of 1.85,
but only when normalized by bwd. No difference was noted when it was normalized by
bwd or for the beam tested at an a/d ratio of 2.50 regardless of the normalization
technique; which suggests that size effect was not an issue for these large-scale beams.
A slight but otherwise insignificant decrease in the first cracking load was observed for
the 75-in. beam. There was also no effect on the correlation between the maximum crack
width and the percent of maximum applied load. The larger beam did experience more
cracking than the 42-in. beam but the overall spacing and pattern were the same.
4.8
SERIES VI: LOADED CHORD
The loaded chord series was created to examine the important strength and serviceability
differences in behavior between tension- and compression-chord loaded beams. As
previously discussed, loading on the ledge of an inverted-T beam induces additional
tensile forces into the web that could result in changes of the behavior of the beam.
In the TxDOT Project 0-5253, different variables were examined in order to develop the
strut-and-tie modeling provisions. This resulted in beams with various depths, web
widths, reinforcement ratios, number of stirrup legs, bearing and loading plate sizes, and
a/d ratios. Because some of these objectives differ greatly from the current project, which
focuses more on ledge geometry, reinforcement ratio, and number of point loads, many of
the tests from Project 0-5253 cannot be used in this comparison. Only rectangular beams
with 21 in. web width, 42 in. or 75 in. web depth, 0.3% web reinforcement, and a/d ratios
of 1.85 and 2.50 are included in this series.
Due to a lack of rectangular beams with 0.6% web reinforcement in both directions, a 42
in. specimen, C1-42-1.85-06, was constructed and loaded on its compression chord at an
a/d ratio of 1.85. Section 3.2.7 provides additional background information on the loaded
chord series.
The experimental strength of the twenty one beams examined in this test series is
summarized in Table 4-7. Fourteen inverted-T beams were compared to seven
compression-chord loaded specimens in four similar groups in Figure 4-39 and Figure
4-40. All variables presented in the table are defined in Section 4.2. The strength results
were normalized as discussed in Section 4.2.1
107
Table 4-7: Series VI: Loaded Chord- Experimental results.
f 'c
(psi)
DC1-42-1.85-03 4303
DS1-42-1.85-03 5258
DL1-42-1.85-03 4929
SC1-42-1.85-03 (le) 4330
_1-42-1.85-03 SS1-42-1.85-03 5721
III-1.85-03
4990
III-1.85-03b
3300
I-03-2
5240
I-03-4
5330
SS1-75-1.85-03b 2867
_1-75-1.85-03
IV-2175-1.85-03 4930
DC1-42-1.85-06 3727
DS1-42-1.85-06 5024
_1-42-1.85-06
DL1-42-1.85-06 4830
C1-42-1.85-06 3727
DC1-42-2.50-03 4035
DL1-42-2.50-03 4929
SC1-42-2.50-03 (r) 4281
_1-42-2.50-03
SS1-42-2.50-03 5703
SL1-42-2.50-03 4281
III-2.5-03
5030
Comparison
Specimen
Vtest
(kip)
517
712
626
463
583
412
471
569
657
745
842
519
621
741
637
365
510
319
398
498
516
0.15
0.17
0.16
0.14
0.13
0.10
0.18
0.13
0.15
0.18
0.12
0.18
0.16
0.19
0.22
0.11
0.13
0.09
0.09
0.15
0.13
9.97
12.42
11.28
8.90
9.75
7.20
10.12
9.73
11.14
9.72
8.29
10.76
11.09
13.48
13.21
7.28
9.19
6.18
6.67
9.62
8.98
Vcrack
(kip)
127
172
242
N/A
N/A
137
114
144
N/A
346
218
107
188
168
N/A
70
N/A
N/A
157
167
N/A
2.45
2.99
4.36
N/A
N/A
2.39
2.45
2.45
N/A
4.51
2.15
2.23
3.35
3.06
N/A
1.40
N/A
N/A
2.63
3.24
N/A
Compression-chord loaded rectangular beam from TxDOT Project 0-5253

Compression-chord loaded rectangular beam from TxDOT Project 0-6416
Web shear failure was not achieved in two beams in this series due to inadequate ledge
capacity. SC1-42-1.85-03 (le) failed due to ledge tie yielding and SC1-42-2.50-03 (r)
failed due to shear friction failure. Both of these beams exceeded the predicted shear
capacity and showed signs of being very close to web shear failure at their ultimate load.
108
_1-75-1.85-03
0.15
0.15
0.00
Tension Chord
0.05
0.00
Compression Chord
0.15
0.00
Tension Chord
C1-42-1.85-06
0.15
DL1-42-1.85-06
0.20
DS1-42-1.85-06
0.20
DC1-42-1.85-06
0.25
0.05
Compression Chord
_1-42-2.5-03
_1-42-1.85-06
0.25
0.10
Tension Chord
0.10
0.05
0.00
Compression Chord
III-2.5-03*
0.05
0.10
DC1-42-2.50-03
DL1-42-2.50-03
SC1-42-2.50-03(r)
SS1-42-2.50-03
SL1-42-2.50-03
0.10
IV-2175-1.85-03*
0.20
III-1.85-03*
III-1.85-03b*
I-03-2*
I-03-4*
0.20
DC1-42-1.85-03
DS1-42-1.85-03
DL1-42-1.85-03
SC1-42-1.85-03 (le)
SS1-42-1.85-03
0.25
SS1-75-1.85-03b
_1-42-1.85-03
0.25
Tension Chord
Compression Chord
Figure 4-39: Series VI: Loaded Chord- Direct comparisons of Vtest normalized by f'cbwd.
No trend is seen amongst the shear strength comparisons normalized by f'cbwd in Figure
4-39. A similar lack in trend is shown in Figure 4-40 where the shear strength results
were normalized by bwd.
109
10
10
4
2
Tension Chord
Compression Chord
_1-42-1.85-06
10
10
C1-42-1.85-06
DL1-42-1.85-06
12
DS1-42-1.85-06
12
DC1-42-1.85-06
14
Compression Chord
_1-42-2.5-03
14
Tension Chord
III-2.5-03*
DC1-42-2.50-03
DL1-42-2.50-03
SC1-42-2.50-03(r)
SS1-42-2.50-03
SL1-42-2.50-03
III-1.85-03*
III-1.85-03b*
I-03-2*
I-03-4*
12
DC1-42-1.85-03
DS1-42-1.85-03
DL1-42-1.85-03
SC1-42-1.85-03(le)
SS1-42-1.85-03
12
IV-2175-1.85-03*
_1-75-1.85-03
14
SS1-75-1.85-03b
_1-42-1.85-03
14
Tension Chord
Compression Chord
Tension Chord
Compression Chord
Figure 4-40: Series VI: Loaded Chord- Direct comparisons of Vtest normalized by bwd.

The diagonal cracking loads and the maximum width of the cracks were used to assess
the effect of web reinforcement on the serviceability performance of ledge-loaded
inverted-T beams versus rectangular beams loaded on their compression chord.
The available diagonal crack loads for the beams tested in this series were normalized by
bwd and plotted in Figure 4-41 by loaded chord. No first cracking loads were
available for the compression-chord loaded beams in the _1-42-1.85-06 or _1-42-2.5-03
comparisons, so the corresponding inverted-T beams are omitted from the figure.
110
5
4
IV-2175-1.85-03*
I-03-02*
III-1.85-03b*
III-1.85-03*
SS1-75-1.85-03b
DL1-42-1.85-03
DS1-42-1.85-03
DC1-42-1.85-03
0
Tension Chord
Compression Chord
Figure 4-41: Series VI: Loaded Chord Comparison of Vcrack normalized by bwd.
As shown in Figure 4-41, the location of the applied load on the cross section of a
member has an effect on the amount of load at which the member cracks. For 42 in. deep
beams with short and cutoff ledges, the cracking load is similar to that of compressionchord loaded specimens. Beams with long ledges on the other hand seem to have a larger
first cracking load compared to rectangular beams. The 75 in. deep inverted-T beam
shown in Figure 4-41 has a larger cracking load than rectangular beam even though it has
a short ledge. Overall through, inverted-T beams appear to crack at a higher normalized
load than traditional compression-chord loaded beams. Hence rectangular-beam cracking
loads can be used as conservative reasonable and conservative estimates of inverted-T
beam cracking loads, regardless of ledge geometry.
The maximum diagonal crack widths are plotted versus the corresponding maximum
applied load in Figure 4-42. There are several general trends that can be observed in the
plots. First, tension-chord loaded beams have more scatter with respect to the maximum
diagonal crack width as indicated by the R2 values from the power trend lines. Secondly,
the tension-chord loaded beams tend to have smaller crack widths at a given percentage
of ultimate load. This suggests that an inverted-T beam is closer to its ultimate capacity
than a compression-chord loaded beam with the same width diagonal crack. This
difference is not as clear for the first comparison, _1-42-1.85-03, but the trend is still
present.
111
_1-42-1.85-03
100
Tension Chord
R = 0.77
_1-75-1.85-03
Compression Chord
R = 0.86
Tension Chord
R = 0.97
100
80
80
60
60
40
40
20
20
0.02
0.04
0.06
0.08
0.1
0.02
Tension Chord
R = 0.83
Compression Chord
R = 0.89
100
80
80
60
60
40
40
20
20
0
0
0.02
0.04
0.06
0.04
0.06
0.08
0.1
_1-42-2.5-03
_1-42-1.85-06
100
Compression Chord
R = 0.96
0.08
0.1
Tension Chord
R = 0.88
Compression Chord
R = 0.98
0
0
0.02
0.04
0.06
0.08
0.1

Figure 4-42: Series V: Loaded Chord- Direct comparisons of crack width progression.
The crack widths of the beams discussed in this series are plotted collectively in Figure
4-43. No distinction was made for a/d ratio, reinforcement ratio, or web height. Although
there is a significant amount of scatter, there is also a noticeable difference between
compression and tension cord loaded beams. This difference varies up to 10% and will be
considered in the correlation of maximum diagonal crack width to capacity discussion in
Section 5.4.
112
Tension Chord Loaded

R = 0.78
100
90
80
70
60
50
40
30
20
10
0
Compression Chord Loaded

R = 0.87
20 tests
0.02
0.04
0.06
0.08
0.1

Figure 4-43: Series V: Loaded Chord- General comparisons of crack width progression.
4.8.3 Summary
General observations and direct comparisons have been presented in this section to
illustrate the influence of the loaded chord on the strength, appearance of first diagonal
crack, and the progression of the maximum crack width with respect to applied load.
It has been shown the applying load closer to the tension chord of a reinforced concrete
member does not have a significant effect on its web shear strength. However, when the
serviceability behavior is considered there is a noticeable difference. Inverted-T beams
with long ledges tend to crack at a higher normalized load than comparable compressionchord loaded members. Inverted-T beams with short or cut-off ledges tend to crack at a
similar normalized load than comparable compression-chord loaded members. The
progression of diagonal crack widths was also affected by chord loading. The tensionchord loaded beams tended to have greater scatter with respect to the maximum diagonal
crack widths but the general trend indicated that inverted-T beams have narrower crack
widths at a given percentage of ultimate capacity than comparable compression-chord
loaded members. This observation will be investigated further in Section 5.4- correlation
of maximum diagonal crack width to member capacity.
4.9
SUMMARY OF EXPERIMENTAL RESULTS
Experimental results of specimens tested within TxDOT Project 0-6416 were included in
this chapter. General conclusions regarding the evaluation of strength and serviceability
criteria were made based on the normalization of strength results, the shear force at first
inclined cracking, and the progression of the maximum diagonal crack widths as
113
compared to the percent of ultimate applied load. Effects of ledge length, ledge depth,
web reinforcement ratio, number of point loads, and web depth were presented in detail.
Increasing the ledge length was observed to increase the shear strength of IT beams and
delay the appearance of the first diagonal cracking. Ledge length was found to have no
significant effect on crack width progression.
Ledge depth has no significant effect on the strength or crack width progression of
inverted-T beams. However, it was observed that increasing the ledge depth delayed the
appearance of the first diagonal cracking slightly.
An increased amount of web reinforcement was shown to increase the shear strength of
the test specimens and slightly decrease the diagonal cracking load. Similar levels of
shear strength increases with increasing reinforcement ratio were observed for specimens
with both a/d ratios (i.e., 1.85 and 2.50). A significant effect on the maximum crack
width progression was observed, with a greater degree of crack width restraint for
specimens with higher web reinforcement ratios. The recommended minimum
reinforcement ratio of 0.3% was also shown to adequately control crack widths at service
level loads.
Loading inverted-T beams at multiple points resulted in a negligible effect on the shear
strength of the member, the diagonal cracking load, and the crack width progression with
applied load.
Increasing the web depth from 42 in. to 75 in. resulted in no notable effect on the strength
of the inverted-T beams. Similarly the serviceability behavior was unchanged between
the beams with different web depths.
When comparing compression- and tension-chord loaded beam, no notable trend was
observed on the strength of the beams. A slight increase in first cracking load was
observed for the inverted-T beams (dependent on ledge geometry) and the crack width
progression suggests narrower cracks in inverted-T beams at the same percent of ultimate
load.
114
CHAPTER 5
Analysis of Results
5.1
OVERVIEW
In this chapter, analyses of the data from the experimental program are presented to
address the following tasks:
Evaluating the applicability of existing strut-and-tie modeling provisions for design of

inverted-T beams
Providing recommendations for limiting diagonal cracking under service loads
Correlating the maximum inclined crack width to member ultimate shear strength
5.2
STRENGTH ANALYSIS
A summary of the experimental versus calculated shear strengths (Vtest/Vcalc) for all tests
conducted within this project is provided in Table 5-1. The calculations of Vcalc for the
specimens were discussed in Chapter 4.
Vtest =
Vcalc =
maximum shear carried in the critical section of the test region, including
self-weight of the specimen and test setup.
shear capacity calculated using the measured material properties presented
in Section 3.4 and TxDOT Project 0-5253 STM provisions as
implemented for inverted-T beams in Section 2.5.1.
As can be seen in Table 5-1, all values of Vtest/Vcalc are higher than 1.0, which indicates
that the STM provisions proposed by Project 0-5253 as implemented for inverted-T
beams in Section 2.5.1 are conservative for all specimens tested. Values of Vtest/Vcalc
lower than 1.2 are shaded in the table to highlight tests where the STM provisions had a
relatively low safety margin. One should note however that strength reduction factors
were not used in calculating Vcalc and hence conservatism would be greater in design.
The calculated and observed failure modes are also summarized for all specimens in
Table 5-1. From the test observations it was often difficult to distinguish between node
and strut crushing, therefore both failure modes were referred to as concrete crushing.
All specimens were designed using the STM provisions to fail in web shear, either
through yielding of the intermediate tie or through crushing of the strut-to-node interface.
The calculated mode of failure was observed in all tests with the exception of the five
shaded in Table 5-1. For those five tests, although the expected failure mode did not
occur, each of these specimens failed well above the web shear capacity predicted by the
strut-and-tie model. Thus the strength estimates were still conservative.
115
Table 5-1: Vtest/Vcalc results for 5253 STM provisions.
Vtest
Specimen
DS1-42-1.85-03
DS1-42-2.50-03
DS1-42-1.85-06
DS1-42-2.50-06
DL1-42-1.85-06
DL1-42-2.50-06
SS3-42-1.85-03
SS3-42-2.50-03
SS3-42-2.50-06
SC3-42-2.50-03
SC3-42-1.85-03
DS3-42-2.50-03
DL1-42-1.85-03
DL1-42-2.50-03
SL3-42-1.85-03
SL3-42-1.85-06
C1-42-1.85-06
DC1-42-1.85-06
SS1-75-1.85-03
DC3-42-1.85-03
DS3-42-1.85-03
SS1-42-2.50-03
SS1-42-1.85-03
DC1-42-2.50-03
DL3-42-1.85-03
SL1-42-2.50-03
SC1-42-2.50-03
DS1-42-1.85-06/03
DS1-42-2.50-06/03
SC1-42-1.85-03
DC1-42-1.85-03
SC1-42-1.85-03b
DC1-42-1.85-03b
SS1-75-2.50-03
Vcalc
kip
Observed
Failure Mode
712
406
621
503
741
622
523
447
516
329
483
430
626
510
571
744
637
519
745
395
454
398
583
365
629
498
319
739
539
463
517
456
424
649
Concrete Crushing
Sectional Shear
Concrete Crushing
Sectional Shear
Concrete Crushing
Sectional Shear
Concrete Crushing
Sectional Shear
Flexure Failure
Sectional Shear
Concrete Crushing
Sectional Shear
Concrete Crushing
Sectional Shear
Concrete Crushing
Concrete Crushing
Concrete Crushing
Concrete Crushing
Concrete Crushing
Concrete Crushing
Concrete Crushing
Sectional Shear
Concrete Crushing
Sectional Shear
Flexure Failure
Sectional Shear
Shear Friction
Concrete Crushing
Sectional Shear
Ledge Tie
Concrete Crushing
Concrete Crushing
Concrete Crushing
Punching Shear
463
202
479
339
464
353
456
215
415
257
427
236
468
235
409
424
428
428
389
370
370
205
501
259
359
261
259
416
362
443
474
362
362
357
116
STM Controlling
Element
kip
1.54
2.01
1.30
1.48
1.60
1.76
1.15
2.08
1.24
1.28
1.13
1.82
1.34
2.17
1.39
1.76
1.49
1.21
1.92
1.07
1.23
1.94
1.16
1.41
1.75
1.91
1.23
1.78
1.49
1.05
1.09
1.26
1.17
1.82
Node at support
Intermediate tie
Node at support
Intermediate tie
Node at support
Intermediate tie
Node at support
Intermediate tie
Intermediate tie
Intermediate tie
Node at support
Intermediate tie
Node at support
Intermediate tie
Node at support
Node at support
Node at load
Node at load
Node at support
Node at support
Node at support
Intermediate tie
Node at support
Intermediate tie
Node at support
Intermediate tie
Intermediate tie
Node at support
Intermediate tie
Node at load
Node at load
Node at load
Node at load
Intermediate tie
5.2.1 Failure Modes

Web shear failure was observed in all specimens except for five in which flexure, shear
friction, ledge tie failures, or punching shear occurred (SS3-42-2.50-06, DL3-42-1.85-03,
SC1-42-2.50-03, SC1-42-1.85-03, and SS1-75-2.50-03) as shown in Figure 5-1.
E
Figure 5-1: Failure modes- (A) SS3-42-2.50-06 flexure, (B) DL3-42-1.85-03 flexure, (C) SC142-2.50-03 shear friction, (D) SC1-42-1.85-03 ledge tie yielding, and (E) SS1-75-2.50-03
punching shear.
117
For specimen SS3-42-2.50-06, a flexural mode of failure was observed (Figure 5-1A).
During specimen design, web shear failure was predicted using a strut-and-tie model and
specified material properties. However, when measured material strengths were
considered the flexural capacity was estimated to be only 6% higher than the web shear
strength; which partly explains the flexural failure mode. Nevertheless the specimen was
able to exceed the capacity calculated using the TxDOT Project 0-5253 STM provisions.
Specimen DL3-42-1.85-03 also failed in flexure (Figure 5-1B). The strut-and-tie model
was controlled by web shear with a predicted overstrength of 38% over the flexural
strength. As with the previous test, the ultimate strength of the specimen was well above
the estimated strengths with Vtest/Vcalc ratio of 1.75.
Shear friction failure of the ledge was observed in SC1-42-2.50-03 (Figure 5-1C). The
STM design was controlled by web shear, with overstrengths of 83% and 135% over
ledge tie and strut strengths respectively. No indication of local failure was anticipated in
the design phase. It is interesting to note that this specimen had a shallow, cut-off ledge
and a single loading point; which indicates that these parameters may have influenced the
weaker than anticipated ledge strength. The Vtest/Vcalc ratio was still conservative at 1.23.
Specimen SC1-42-1.85-03 sustained a local failure in the ledge (Figure 5-1D). As with
the previous specimen with ledge failure, this beam also had a shallow, cut-off ledge and
a single loading point. The STM design was governed by web shear; however, when the
actual material strengths were considered the ledge strut and ledge tie were calculated to
be only 10% stronger than the strut-to-node interface in the web. Hanger reinforcements
had an overstrength of 20%. Since the concrete strength was higher than anticipated in
the original design, web shear-strength was greater than anticipated and the calculated
safety against local ledge failure was compromised.
Specimen SS1-75-2.50-03 failed due to punching shear in the ledge (Figure 5-1E). This
beam had a short, shallow ledge that was not able to carry the higher than anticipated
load. At yielding of the hanger bars, crack widths were recorded and it was determined
that the beam was within approximately 10% of its failure load. Retesting through
application of load to the web of the beam confirmed this observation and it was
determined that the test provided valid information on the shear strength and behavior of
a 75-in. deep beam.
For the remaining twenty nine specimens, all of the observed web shear failures similar
to those shown in Figure 5-2 and were correctly predicted using the STM provisions of
TxDOT Project 0-5253. Overall the STM provisions resulted in accurate predictions of
the actual failure modes. In the 12% of the cases where web shear was not the limiting
factor, the actual failure mode involved the next weakest element in the STM.
118
Concrete Crushing Failure
Stirrup Yielding Failure
Figure 5-2: Shear failure: concrete crushing for a/d = 1.85 and stirrup yielding for a/d =2.50
5.2.2 Ledge Design

Although the ledge length and depth were variables considered in this report, the factors
affecting the strength of the ledge itself were not directly investigated. Each specimen
was designed such that its shear strength would govern its capacity. In order for that to
occur, the ledge and cross sectional model was required to have greater strength. This is
another advantage to strut-and-tie modeling; the desired failure mode can be chosen by
identifying the weakest link and strengthening other components. For example, in
specimens tested at an a/d ratio of 2.50 it was necessary for the shear stirrups to yield
before the ledge or hanger reinforcement (shown in Figure 5-3), and the reinforcement
was proportioned accordingly. If crushing of the diagonal strut in the cross section was
governing the model, the concrete strength can be increased.
Hanger
Reinforcement
Ledge
Reinforcement
Strain
Gauges
Diagonal
Strut
Figure 5-3: Inverted-T Cross Section STM and Strain Gauges
As discussed in Section 2.6.2.2, certain assumptions were required to apply the STM
provisions that were developed for rectangular beams to the design of inverted-T beams.
In order to determine the number of required hanger bars, the applied load was assumed
to spread at a 45-degree. Therefore, hanger ties were given a width equal to the length of
the bearing plate (W) plus twice the depth of the ledge (df) for short and long ledges. In
cut-off ledges, the hanger tie was restricted and could only spread in one direction, as
119
shown in Figure 5-4. The same assumptions are found in AASHTO Eq. 5.13.2.5.5-3 for
calculating the strength of hanger reinforcement.
Cut-off ledge
c
45
c+0.5W+df
df
45
W+2df
Figure 5-4: Load Spread for Cross Section Design
The hanger tie width assumptions were validated by measuring strains using electrical
strain gauges described in Section 3.5.1, the locations of which are shown in the cross
section Figure 5-3. Yield was determined for each bar size and specimen as described in
Section 3.3. Typical measured strains normalized by yielding strains for the hanger
reinforcement are shown in Figure 5-5.
One Point Load, Short Shallow Ledge (SS1-42-1.85-03)
Three Point Loads, Cut-off and Short, Deep Ledges (DC1- & DS1-42-1.85-03)
Reinforcement Key:
s y
s y
s y
s y
Figure 5-5: Typical Hanger Stains
In the above figure, higher strains can be seen to concentrate within the assumed load
spread length, with a tendency for the load to spread towards the closest support. Similar
distributions observed in most specimens. Strain gauge measurements thus indicate that
the 45-degree load spread assumption is reasonable and conservative.
Also, as discussed in Section 5.2.1, only three of the thirty three inverted-T specimens
had ledge-related failure modes. The actual failure mode was typically the second
weakest element in the STM and occurred only after the calculated shear capacity was
surpassed. It is important to note that these failures also occurred in specimens loaded at
120
one point, which is not typical in field applications with multiple bridge girders. It is
therefore recommended to calculate the hanger tie widths as shown in Figure 5-4.
5.2.3 Ultimate Strength
The statistical results for the strength ratios of the thirty three tests in the experimental
program are summarized in Table 5-2. As shown in the table, all ratio values fall above
1.0, indicating that the design method yielded conservative estimations of strength. The
mean strength ratio for all tests is 1.50, which indicates fairly accurate strength
calculations.
Table 5-2: Overall accuracy of TxDOT Project 0-5253 STM provisions
33 tests
Min
Max
Mean
Vtest/Vcalc
1.05
2.17
1.50
Unconservative*
0%
Standard deviation
0.33
Coefficient of Variation**
0.22
* Unconservative = % of tests for which Vtest /Vcalc < 1.0
** Coefficient of Variation = Standard Deviation / Mean
In order to evaluate more closely the TxDOT Project 0-5253 STM design provisions for
inverted-T beams, direct comparisons of the experimental shear strength, Vtest, and the
nominal capacity, Vcalc, calculated per the strut-and-tie models are made in the following
sections for the individual series.
5.2.3.1 Effects of Ledge Length
Specimens were constructed with three different ledge lengths: cut-off, short and long
ledges. The Vtest/Vcalc values from the ledge length series specimens are summarized in
Table 5-3. The ratio of test to calculated strengths are organized into eight groups of two
or three directly comparable specimens as illustrated in Figure 5-6. Presenting the data in
this manner allows for observations to be made on specimens in which all variables are
kept constant except the length of the ledge.
121
Table 5-3: Series I: Ledge Length
f ' c Vtest Vcalc

(psi) (kip) (kip)
DS1-42-1.85-03
5258 712 463
D_1-42-1.85-03
DL1-42-1.85-03
4929 626 468
DC3-42-1.85-03
4568 395 370
D_3-42-1.85-03 DS3-42-1.85-03
4568 454 370
DL3-42-1.85-03 (f) 4202 629 359
SC3-42-1.85-03
5873 483 427
5891 523 456
S_3-42-1.85-03 SS3-42-1.85-03
SL3-42-1.85-03
5037 571 409
DS1-42-1.85-06
5024 621 479
D_1-42-1.85-06
DL1-42-1.85-06
4830 741 464
DC1-42-2.50-03
4035 365 259
D_1-42-2.50-03 DS1-42-2.50-03
5389 406 202
DL1-42-2.50-03
4929 510 235
SC1-42-2.50-03 (r) 4281 319 259
S_1-42-2.50-03 SS1-42-2.50-03
5703 398 205
SL1-42-2.50-03
4281 498 261
SC3-42-2.50-03
5873 329 257
S_3-42-2.50-03
SS3-42-2.50-03
5891 447 215
DS1-42-2.50-06
5088 503 339
D_1-42-2.50-06
DL1-42-2.50-06
4986 622 353
Comparison
Specimen
1.54
1.34
1.07
1.23
1.75
1.13
1.15
1.39
1.30
1.60
1.41
2.01
2.17
1.23
1.94
1.91
1.28
2.08
1.48
1.76
The difference in ledge length is accounted for in the strut-and-tie models by the size of
the node above the load point. When the ledge length is shortened, the hanger load is
concentrated and the resulting node size is decreased as illustrated in Figure 3-7. As long
as this node controls the strength of the model, the strut-and-tie method would capture the
effect of the ledge length on the web-shear strength of the inverted-T beam. For the
specimens tested at an a/d ratio of 1.85, the node at the support typically controlled the
predicted capacity of the beam. For the specimens tested at an a/d ratio of 2.50, the
vertical tension tie was designed to govern the beams shear strength. Therefore varying
the ledge length was not expected to have an effect on the calculated capacity of the
specimens according to the strut-and-tie models.
122
D_1-42-1.85-03
D_3-42-1.85-03
S_3-42-1.85-03
D_1-42-1.85-06
2.50
2.50
2.50
2.50
2.00
2.00
2.00
2.00
1.50
1.50
1.50
1.50
1.00
1.00
1.00
0.50
(f)
D_1-42-1.85-03
0.50
0.00
Cut-off
Short
0.00
Long
Cut-off
D_1-42-2.50-03
Short
0.50
0.00
Long
1.00
D_3-42-1.85-03
Cut-off
Short
0.50
0.00
Long
2.50
2.50
2.50
2.50
2.00
2.00
2.00
2.00
1.50
1.50
1.50
1.50
1.00
1.00
0.50
0.50
0.00
Cut-off
Short
0.00
Long
(r)
Cut-off
Short
0.50
0.00
Long
Short
Long
1.00
1.00
D_1-42-2.50-03
Cut-off
D_1-42-1.85-06
D_1-42-2.50-06
S_3-42-2.50-03
S_1-42-2.50-03
S_3-42-1.85-03
S_1-42-2.50-03
Cut-off
Short
0.50
0.00
Long
D_1-42-2.50-06
S_3-42-2.50-03
Cut-off
Short
Long

Figure 5-6: Series I: Ledge Length- Direct comparisons of experimental capacity with TxDOT Project 0-5253 STM calculations
123
As shown in Figure 5-6, the ratio of experimental shear strength to nominal capacity
values varied between 1.07 and 2.17. A clear trend can be observed in the directly
comparable specimens where the conservatism increases as the ledge length increases.
There are a couple of cases that did not follow this trend, but that can be attributed to
variability in shear tests. Considering the large number of tests presented in this series, it
is evident that longer ledges provide additional strength not captured by the STM
provisions.
This observation is confirmed when a more general comparison of thirty three tests from
the experimental program is made. The ratio of the experimental to calculated shear
strengths are summarized by ledge length and a/d ratio in Table 5-4 and Figure 5-7. The
minimum, maximum, and average Vtest/Vcalc values increase with the ledge length
suggesting that the STM provisions do not fully take into account the beneficial effect of
longer ledges on the strength of inverted-T beams. It can be seen in Table 5-4 that the
conservatism increases with ledge length for both a/d ratios; with strength estimates for
beam tested at an a/d ratio of 2.50 being more conservative than those tested at 1.85. It is
also important to note that the highest strength ratio (2.17) was found for a long ledge
specimen and the lowest (1.05) corresponded to a cut-off ledge specimen. The cut-off
ledges also had the lowest average Vtest/Vcalc value of 1.14.
Table 5-4: Series I: Ledge Length- STM summary by a/d ratio.
Cut-off Ledge Length

Short Ledge Length
Long Ledge Length
a/d 1.85 a/d 2.50 Avg a/d 1.85 a/d 2.50 Avg a/d 1.85 a/d 2.50 Avg
7 tests 3 tests 10 tests 7 tests 8 tests 15 tests 5 tests 3 tests 8 tests
Min
1.05
1.23
1.05
1.15
1.24
1.15
Max
1.26
1.41
1.41
1.92
2.08
2.08
Mean
1.14
1.31
1.19
1.44
1.74
1.60
STDV 0.08
0.09
0.11
0.31
0.30
0.33
COV* 0.07
0.07
0.09
0.22
0.17
0.16
* COV = Coefficient of Variation (Standard Deviation / Mean)
124
1.34
1.76
1.57
0.20
0.12
1.76
2.17
1.95
0.21
0.11
1.34
2.17
1.71
0.27
0.16
a/d = 1.85
a/d = 2.5
2.50
2.00
1.50
1.00
0.50
15 tests
10 tests
8 tests
0.00
Cut-off Ledge
Short Ledge
Long Ledge
Figure 5-7: Series I: Ledge Length- STM strength predictions.
Results in this section indicate that using cut-off ledges reduces the conservatism of the
STM provisions, nevertheless each model resulted in a safe estimate of the capacity of
the specimens. Given the relatively low conservatism for cut-off ledges, it may be
preferable to avoid cut-off ledges in practice.
5.2.3.2 Effects of Ledge Depth
Two ledge depths were tested: a shallow depth equal to one third the total height of the
member and a deep ledge depth equal to one half of the members height. The depths
were chosen to encompass the range observed in the field and still ensure a web-shear
failure without punching through the ledge. A summary of the experimental shear
strength, Vtest, and the nominal capacity, Vcalc, calculated per the TxDOT Project 0-5253
STM provisions for the beams in the ledge depth series is provided in Table 5-5.
125
Table 5-5: Series II: Ledge Depth
Comparison
Specimen
f 'c
(psi)
Vtest Vcalc
(kip) (kip)
SC1-42-1.85-03 (le) 4303 463

DC1-42-1.85-03
4330 517
SC1-42-1.85-03b
2996 456
_C1-42-1.85-03b
DC1-42-1.85-03b
3013 424
SS1-42-1.85-03
5721 583
_S1-42-1.85-03
DS1-42-1.85-03
5258 712
_C3-42-1.85-03 SC3-42-1.85-03
5873 483
DC3-42-1.85-03
4568 395
_S3-42-1.85-03 SS3-42-1.85-03
5891 523
DS3-42-1.85-03
4568 454
_L3-42-1.85-03 SL3-42-1.85-03
5037 571
DL3-42-1.85-03 (f) 4202 629
_C1-42-2.50-03 SC1-42-2.50-03 (r) 4281 319
DC1-42-2.50-03
4035 365
_S1-42-2.50-03 SS1-42-2.50-03
5703 398
DS1-42-2.50-03
5389 406
_L1-42-2.50-03 SL1-42-2.50-03
4281 498
DL1-42-2.50-03
4929 510
_S3-42-2.50-03 SS3-42-2.50-03
5891 447
DS3-42-2.50-03
5687 430
_C1-42-1.85-03
443
474
362
362
501
463
427
370
456
370
409
359
259
259
205
202
261
235
215
236
1.05
1.09
1.26
1.17
1.16
1.54
1.13
1.07
1.15
1.23
1.39
1.75
1.23
1.41
1.94
2.01
1.91
2.17
2.08
1.82
The Vtest/Vcalc values from twenty tests are plotted in ten pairs of directly comparable
specimens to investigate the effect of the ledge height in Figure 5-8. Every variable in
each group is held constant except for the depth of the ledge. No clear trend could be
distinguished when examining the capacity ratio plots in Figure 5-8.
126
_C1-42-1.85-03
_C1-42-1.85-03b
_S1-42-1.85-03
_C3-42-1.85-03
2.50
2.50
2.50
2.50
2.50
2.00
2.00
2.00
2.00
2.00
1.50
1.50
1.50
1.50
1.50
_S3-42-1.85-03
(le)
1.00
1.00
0.50
0.50
0.00
Shallow
0.00
Deep
_L3-42-1.85-03
2.50
2.00
1.00
_C1-421.85-03
Shallow
_C1-42-2.5-03
Deep
_S1-42-2.50-03
Deep
_L1-42-2.50-03
2.00
2.00
2.00
2.00
1.50
1.50
1.50
(r)
_L3-421.85-03
0.00
Shallow
Shallow
2.50
0.50
0.00
Deep
Shallow
1.00
0.50
0.00
Deep
1.00
_C1-422.50-03
Shallow
0.50
0.00
Deep
_S3-421.85-03
_C3-421.85-03
0.00
0.00
Shallow
2.50
1.00
0.50
0.50
2.50
1.50
1.00
0.00
Deep
0.50
1.00
_S1-421.85-03
2.50
(f)
1.50
0.50
1.00
_C1-421.85-03b
1.00
_S1-422.50-03
0.50
Shallow
Deep
_S3-42-2.50-03
_L1-422.50-03
_S3-422.50-03
0.00
Shallow
Deep
Shallow
Deep
Ledge Depth
Figure 5-8: Series II: Ledge Depth- Direct comparisons of experimental capacity with TxDOT Project 0-5253 STM calculations
127
The ratio of the experimental to calculated shear strengths are summarized by ledge depth
and a/d ratio in Table 5-6 and Figure 5-9. When examining the collection of all the tests
in the experimental program, it is apparent that the ledge depth has no appreciable effect
on the accuracy or conservatism of the strut-and-tie models. Only a slight difference is
noted between the average conservatism for the two a/d ratios. One can conclude then
that the effect of the ledge depth on web-shear strength of inverted-T beams is adequately
captured by the strut-and-tie model; i.e., no bias is introduced by the design procedure.
Table 5-6: Series II: Ledge Depth- STM summary by a/d ratio.
Shallow Ledge Depth

a/d = 1.85 a/d = 2.50 Avg
8 tests
7 tests
Deep Ledge Depth

a/d = 1.85 a/d = 2.50 Avg
15 tests 11 tests
7 tests
18 tests
1.41
2.17
1.74
0.29
0.17
1.07
2.17
1.51
0.32
0.21
Min
1.13
1.23
1.13
1.07
Max
1.92
2.08
2.08
1.78
Mean
1.38
1.64
1.50
1.37
Standard deviation
0.30
0.37
0.35
0.25
COV**
0.22
0.23
0.23
0.19
** COV = Coefficient of Variation (Standard Deviation / Mean)
a/d = 1.85
a/d = 2.5
2.50
2.00
1.50
1.00
0.50
18 tests
15 tests
0.00
Deep Ledge
Shallow Ledge
Figure 5-9: Series II: Ledge Depth- STM and LRFD strength predictions
The STM provisions mandate a minimum angle of 25 degrees between a strut and a tie
and thus a minimum ledge depth is implicit in the provisions. Given the lack of bias in
the strength estimates when ledge depth is considered, no further recommendations are
128
made about ledge depth by the current study. The few beams that sustained ledge failures
all had shallow ledges and were loaded with a single point load. These beams did
however fail at a substantially higher load than the ledge strength estimated by the STM
provisions.
5.2.3.3 Effects of Web Reinforcement
It was shown in Section 4.5 that increasing the web reinforcement ratio from 0.3% to
0.6% in both orthogonal directions increased the strength of inverted-T beams. Larger
amounts of web reinforcement were also found to affect the serviceability behavior of the
beam as additional reinforcement better restrained the widths of the diagonal cracks.
A summary of the experimental shear strength, Vtest, and the nominal capacity, Vcalc,
calculated per the TxDOT Project 0-5253 STM provisions for the fourteen directly
comparable specimens investigated in Section 4.5 is presented in Table 5-7.
Table 5-7: Series III: Web Reinforcement Ratio
Comparison
DS1-42-1.85-_
DL1-42-1.85-_
SL3-42-1.85-_
DS1-42-2.50-_
DL1-42-2.50-_
SS3-42-2.50-_
Specimen
f 'c
(psi)
Vtest
(kip)
Vcalc
(kip)
DS1-42-1.85-03
DS1-42-1.85-06/03
DS1-42-1.85-06
DL1-42-1.85-03
DL1-42-1.85-06
SL3-42-1.85-03
SL3-42-1.85-06
DS1-42-2.50-03
DS1-42-2.50-06/03
DS1-42-2.50-06
DL1-42-2.50-03
DL1-42-2.50-06
SS3-42-2.50-03
SS3-42-2.50-06 (f)
5258
4173
5024
4929
4830
5037
5250
5389
4173
5088
4929
4986
5891
6255
712
739
621
626
741
571
744
406
539
503
510
622
447
516
463
416
479
468
464
409
424
202
362
339
235
353
215
415
1.54
1.78
1.30
1.34
1.60
1.39
1.76
2.01
1.49
1.48
2.17
1.76
2.08
1.24

The ratio of test shear capacity to that calculated using the strut-and-tie models is shown
in Figure 5-10 for the beams in this series. Each of the six groups is a direct comparison
of two or three specimens in which every parameter was kept constant except the
reinforcement ratio.
129
DS1-42-1.85-_
DL1-42-1.85-_
SL3-42-1.85-_
2.50
2.50
2.50
2.00
2.00
2.00
1.50
1.50
1.50
1.00
1.00
1.00
0.50
0.50
0.50
0.00
0.00
v = 0.3% 0.6%
h = 0.3% 0.3%
0.00
= 0.6%
v 0.3%
= 0.3% v 0.6%
h = 0.3% h = 0.6%
DL1-42-1.85_
= 0.6%
v 0.3%
= 0.3% v 0.6%
h = 0.3%
DS1- h = 0.6%
0.6%
0.6%
42-
DS1-42-2.50-_
SS3-42-2.50-_
DL1-42-2.50-_
2.50
2.50
2.50
2.00
2.00
2.00
1.50
1.50
1.50
1.00
1.00
1.00
0.50
0.50
0.50
0.00
0.6%
0.6%
(f)
0.00
0.00
v = 0.3% 0.6%
h = 0.3% 0.3%
SL3-42-1.85_
= 0.6%
v 0.3%
= 0.3% v 0.6%
DS1-42-2.50h = 0.6%
h = 0.3%
_
= 0.6%
v 0.3%
= 0.3% v 0.6%
h = 0.3%
h = 0.6%
DL1-42-2.50_
Figure 5-10: Series III: Web Reinforcement- Direct comparisons of experimental capacity
with TxDOT Project 0-5253 STM calculations
The Vtest/Vcalc ratios varied between 1.24 and 2.17 for the beams in the reinforcement ratio
series. Except for one point, an increase in conservatism was observed as the
reinforcement ratio is increased for the specimens with an a/d ratio of 1.85. For the beams
with the larger a/d ratio of 2.50, the opposite was true with decreased conservatism
observed with increasing reinforcement ratio.
Using the TxDOT Project 0-5253 STM provisions, the specimens tested at an a/d ratio of
1.85 were estimated to fail by crushing of the direct strut or the strut-to-node interface.
The STM provisions do not increase the strength of struts or nodes based on the amount
of distributed reinforcement. Since higher strengths were observed for higher
reinforcement ratio in specimens with an a/d of 1.85, and increase in conservatism in the
STM provisions is observed here with increasing reinforcement ratio. The observed
increase in shear strength with increasing amounts of web reinforcement could be due to
additional crack control and confinement provided to the node by the additional
130
SS3-42-2.50_
reinforcement. By contrast, the specimens with an a/d ratio of 2.50 failed due to yielding
of the transverse reinforcement (intermediate tie) in the shear span. The STM provisions
relate web shear capacity to the transverse reinforcement ratio; i.e., doubling the amount
of reinforcement doubles the predicted contribution of that steel to shear capacity. Test
results however have shown that doubling the amount of transverse reinforcement less
than doubled the percentage of the steel contribution to the shear capacity; as evidenced
by the lower conservatism of STM provision for the higher web reinforcement ratio. It is
possible that the considered tie width is dependent on the amount of steel and may be
narrower than estimated.
Similar observations can be made when results from all test specimens are considered
(Table 5-8, Figure 5-11).
Table 5-8: Series III: Web Reinforcement Ratio- STM summary
v = h = 0.3%
a/d = 1.85 a/d = 2.50 Avg
14 tests
v = h = 0.6%
a/d = 1.85 a/d = 2.50 Avg
10 tests 24 tests 4 tests
Min
1.05
1.23
1.05
1.21
Max
1.92
2.17
2.17
1.76
Mean
1.30
1.77
1.50
1.47
Standard deviation
0.26
0.34
0.37
0.25
COV**
0.20
0.19
0.25
0.17
131
3 tests
7 tests
1.24
1.49
1.39
0.13
0.10
1.21
1.76
1.43
0.20
0.14
a/d = 1.85
a/d = 2.5
2.50
2.00
1.50
1.00
0.50
24 tests
0.00
2 tests
v = 0.6%
h = 0.3%
v = 0.3%
h = 0.3%
7 tests
v = 0.6%
h = 0.6%
Figure 5-11: Series III: We Reinforcement Ratio- STM capacity results
Overall though, the strut-and-tie provisions produced accurate and reasonably

conservative estimates of strength for all the inverted-T specimens. It is therefore not
recommended to introduce adjustments to the provisions to account for the effects of
reinforcement ratios.
5.2.3.4 Effects of Number of Point Loads
As discussed in Section 4.6, specimens with different number of point loads were
investigated to determine what effect the loading condition had on the strength and
serviceability of inverted-T beams. A slight but otherwise insignificant decrease in
strength and first cracking load was observed for beams tested with multiple point loads.
Results thus indicate that the conclusions from one-point load tests may be extended to
multiple point loads; which is the case for the vast majority of bent caps in the field.
In order to examine the applicability of TxDOT Project 0-5253 STM provisions for
inverted-T beams subjected to multiple loads, results from specimens that were tested
with one and three loading points are compared. The capacities from twelve directly
comparable specimens, in which every variable (ledge geometry, reinforcement ratio, a/d
ratio) was kept constant, are summarized in Table 5-9.
132
Table 5-9: Series IV: Number of Point Loads
Comparison
Specimen
SS1-42-1.85-03
SS3-42-1.85-03
DS1-42-1.85-03
DS_-42-1.85-03
DS3-42-1.85-03
DL1-42-1.85-03
DL_-42-1.85-03
DL3-42-1.85-03 (f)
SS_-42-1.85-03
SC1-42-2.50-03 (r)
SC3-42-2.50-03
SS1-42-2.50-03
SS_-42-2.50-03
SS3-42-2.50-03
DS1-42-2.50-03
DS_-42-2.50-03
DS3-42-2.50-03
SC_-42-2.50-03
f 'c
(psi)
5721
5891
5258
4568
4929
4202
4281
5873
5703
5891
5389
5687
Vtest
(kip)
583
523
712
454
626
629
319
329
398
447
406
430
Vcalc
(kip)
501
456
463
370
468
359
259
257
205
215
202
236
1.16
1.15
1.54
1.23
1.34
1.75
1.23
1.28
1.94
2.08
2.01
1.82
It can be inferred from Table 5-9 that many of the Vtest/Vcalc values are similar amongst
the number of point load comparisons. The ratios are plotted in Figure 5-12 in six pairs of
directly comparable specimens in order to better illustrate this point. The Vtest/Vcalc ratios
varied between 1.15 and 2.08. No clear trend resulting from different number of loads is
seen in the figure. Such observations suggest that no bias is introduced in the STM design
provisions when multiple point loads are considered.
133
SS_-42-1.85-03
DS_-42-1.85-03
2.50
2.50
2.50
2.00
2.00
2.00
1.50
1.50
1.50
1.00
1.00
0.50
One
One
Three
SC_-42-2.5-03
One
Three
SS_-42-2.50-03
2.50
2.50
2.00
2.00
2.00
1.50
1.50
(r)
1.00
1.00
0.50
0.50
1.00
SC_422.50-03
0.50
0.00
0.00
One
Three
Three
DS_-42-2.50-03
2.50
1.50
DL_421.85-03
DS_421.85-03
0.00
0.00
0.00
(f)
1.00
SS_421.85-03
0.50
0.50
DL_-42-1.85-03
SS_422.50-03
DS_422.50-03
0.00
One
Three
One
Three

Figure 5-12: Series IV: Number of Point Loads- Direct comparisons of experimental
capacity with TxDOT Project 0-5253 STM calculations.
Similar observations can be made when results from all test specimens are considered
(Table 5-10 and Figure 5-13).
Table 5-10: Series IV: Number of Point Loads- STM summary.
One Point Load

Three Point Loads
a/d = 1.85 a/d = 2.50 Avg a/d = 1.85 a/d = 2.50 Avg
12 tests 10 tests 22 tests 7 tests
4 tests 11 tests
Min
1.05
1.23
1.05
1.07
Max
1.92
2.17
2.17
1.76
Mean
1.37
1.72
1.53
1.35
Standard deviation
0.28
0.30
0.34
0.29
COV**
0.20
0.18
0.22
0.22
134
1.24
2.08
1.61
0.41
0.26
1.07
2.08
1.44
0.34
0.24
STM a/d = 1.85
STM a/d = 2.5
2.50
2.00
1.50
1.00
0.50
22 tests
0.00
One Point Load
11 tests
Three Point Load
Figure 5-13: Series IV: Number of Point Loads- STM strength predictions
The STM design provisions of TxDOT Project 0-5253 thus provided accurate and
acceptably conservative estimates of strength for the specimens evaluated in this series
and adequately captured the behavior regardless of the number of point loads used in
testing.
5.2.3.5 Effects of Web Depth
In order to ensure that the TxDOT Project 0-5253 STM provisions are applicable to
larger inverted-T beams, a series investigating different web depths was developed. Most
of the specimens in the experimental program were constructed with a total web height of
42 in. Two beams were constructed with a total depth of 75 in. and are investigated in
this section. As mentioned previously, laboratory testing limitations prevented testing
additional beams with web depths of 75 in. The web height is the only variable that is
changed between the direct comparisons provided in Table 5-11. All beams were loaded
on short, shallow ledges and their estimated strengths are presented in Table 5-11.
135
Table 5-11: Series V: Web Depth
f 'c
(psi)
SS1-42-1.85-03
5721
SS1-_-1.85-03
SS1-75-1.85-03
2867
SS1-42-2.50-03
5703
SS1-_-2.50-03
SS1-75-2.50-03 (p) 5158
Comparison
Specimen
Vtest
(kip)
583
745
398
649
Vcalc
(kip)
501
389
205
357
1.16
1.92
1.94
1.82
The ratio of the shear capacity obtained from testing to that predicted by the STM is
shown in Figure 5-14 for the two pairs of comparisons in this series. Although there are
not enough comparisons to draw firm conclusions on the effect of web depth, it is
important to note that strut-and-tie models were used to estimate conservative results for
specimens of both depths. An increase in conservatism was noted for larger beams tested
at an a/d of 1.85 and a negligible decrease was observed for larger beams tested at an a/d
ratio of 2.50.
SS1-_-1.85-03
SS1-_-2.50-03
2.50
2.50
2.00
2.00
1.50
1.50
1.00
1.00
0.50
0.50
(p)
SS1-_-2.50-03
SS1-_-1.85-03
0.00
0.00
42 in.
75 in.
42 in.
75 in.
Web Depth, h (in.)

Figure 5-14: Series V: Web Depth- Direct comparisons of experimental capacity with
TxDOT Project 0-5253 STM calculations
5.2.3.6 Effects of Compression- and Tension-Chord Loading

The difference in STM strength estimates between tension- and compression-chord
loaded beams is investigated here. For direct comparison with specimens tested in this
project, only the compression-chord loaded rectangular beams with 21 in. web width, 42
in. or 75 in. web depth, 0.3% web reinforcement, and a/d ratios of 1.85 and 2.50 are
included in this series from Project 0-5253. As no beams were tested with 0.6% web
136
reinforcement in the previous project, a compression-chord loaded beam, C1-42-1.85-06,

was tested in the current experimental program to fill that gap.
Table 5-12: Series VI: Loaded Chord
f ' c Vtest Vpred

(psi) (kip) (kip)
DC1-42-1.85-03 4303 517 474
DS1-42-1.85-03
5258 712 463
DL1-42-1.85-03
4929 626 468
SC1-42-1.85-03 (le) 4330 463 443
_1-42-1.85-03 SS1-42-1.85-03
5721 583 501
III-1.85-03
4990 412 374
III-1.85-03b
3300 471 258
I-03-2
5240 569 381
I-03-4
5330 657 382
SS1-75-1.85-03
2867 745 389
_1-75-1.85-03
IV-2175-1.85-03 4930 842 501
DC1-42-1.85-06 3727 519 428
DS1-42-1.85-06
5024 621 479
_1-42-1.85-06
DL1-42-1.85-06
4830 741 464
C1-42-1.85-06
3727 637 428
DC1-42-2.50-03 4035 365 259
DL1-42-2.50-03
4929 510 235
SC1-42-2.50-03 (r) 4281 319 259
_1-42-2.50-03
SS1-42-2.50-03
5703 398 205
SL1-42-2.50-03
4281 498 261
III-2.5-03
5030 516 282
Comparison
Specimen
1.09
1.54
1.34
1.05
1.16
1.10
1.83
1.49
1.72
1.92
1.68
1.21
1.30
1.60
1.49
1.41
2.17
1.23
1.94
1.91
1.83
Compression-chord loaded beam from TxDOT Project 0-5253

Compression-chord loaded beam from TxDOT Project 0-6416
Estimated shear strength ratios Vtest/Vcalc are summarized in The difference in STM
strength estimates between tension- and compression-chord loaded beams is investigated
here. For direct comparison with specimens tested in this project, only the compressionchord loaded rectangular beams with 21 in. web width, 42 in. or 75 in. web depth, 0.3%
web reinforcement, and a/d ratios of 1.85 and 2.50 are included in this series from Project
0-5253. As no beams were tested with 0.6% web reinforcement in the previous project, a
137
compression-chord loaded beam, C1-42-1.85-06, was tested in the current experimental

program to fill that gap.
Table 5-12 for this series. The strength ratios are also plotted in Figure 5-15 for each of
the four comparison groups. Horizontal lines indicate the mean value for each group in
the figure.
1.50
1.50
0.00
Tension Chord
0.50
0.00
Compression Chord
_1-42-1.85-06
1.50
0.00
Tension Chord
C1-42-1.85-06
1.50
DL1-42-1.85-06
2.00
DS1-42-1.85-06
2.00
DC1-42-1.85-06
2.50
0.50
Compression Chord
_1-42-2.5-03
2.50
1.00
Tension Chord
1.0
1.0
1.00
0.50
0.00
Compression Chord
III-2.5-03
0.50
1.0
1.00
DC1-42-2.50-03
DL1-42-2.50-03
SC1-42-2.50-03(r)
SS1-42-2.50-03
SL1-42-2.50-03
1.00
III-1.85-03
III-1.85-03b
I-03-2
I-03-4
2.00
DC1-42-1.85-03
DS1-42-1.85-03
DL1-42-1.85-03
SC1-42-1.85-03(le)
SS1-42-1.85-03
2.00
IV-2175-1.85-03
_1-75-1.85-03
2.50
SS1-75-1.85-03b
_1-42-1.85-03
2.50
Tension Chord
Compression Chord
1.0
Figure 5-15: Series VI: Loaded Chord- Direct comparisons of experimental capacity with
TxDOT Project 0-5253 STM calculations.
The Vtest/Vcalc ratios varied between 1.05 and 2.17 for tension-chord loaded (inverted-T)
beams and 1.10 to 1.83 for compression-chord loaded beams. It is important to note that
all capacity ratios were above 1.0, indicating a conservative strength estimation for all
specimens. As shown in Figure 5-15, the conservatism with respected to the loaded chord
varied depending on the shear span to depth ratio, web depth, and reinforcement ratio.
For 42 in. deep beams tested at an a/d of 1.85, the STM provisions were slightly more
138
conservative (the capacity ratio was higher) for compression-chord loaded specimens
with either reinforcement ratio. The opposite was true for 75 in. deep beams with a
reinforcement ratio of 0.3% in both directions; however with such few shear tests at that
web depth, a definitive conclusion cannot be made. For the 42 in. deep beams tested at an
a/d ratio of 2.50 and 0.3% reinforcement in both directions, the strut-and-tie models for
the tension-chord loaded beams were slightly more conservative. Overall the STM
models appear to estimate relatively accurately shear strength with little bias toward
loading chord and with only two tests having Vtest/Vcalc values over 2.0.
When comparing all of the specimens in this series without grouping, a slight decrease in
the conservatism in the models is seen with the tension-chord loaded specimens
(Vtest/Vcalc = 1.50) versus the compression-chord loaded specimens (Vtest/Vcalc = 1.59) as
summarized in Table 5-13 and Figure 5-16. This trend is true for beams tested at an a/d
ratio of 1.85, as the average Vtest/Vcalc is 1.37 for tension-chord loaded beams and 1.83 for
compression-chord loaded beams. The opposite is true for specimens tested at an a/d ratio
of 2.50 as the conservatism of the inverted-T beams is higher than that of the rectangular
beams, but there are fewer results to compare. Given the conflicting trends for various
parameter groups and only a slight overall difference observed between compression- and
tension-chord loaded specimens, it can be concluded that the STM provisions show little
bias towards compression- or tension-chord loading.
Table 5-13: Series VI: Loaded Chord STM summary.
Tension Chord
Compression Chord
all
a/d = 1.85 a/d = 2.50
all
a/d = 1.85 a/d = 2.50
14 tests 9 tests
5 tests 7 tests 6 tests
1 tests
Min
1.05
1.05
1.23
1.10
1.10
1.83
Max
2.17
1.92
2.17
1.83
1.83
1.83
Mean
1.49
1.35
1.73
1.59
1.55
1.83
STDV
0.36
0.28
0.39
0.26
0.26
COV*
0.24
0.21
0.23
0.16
0.17
* COV = Coefficient of Variation (Standard Deviation / Mean)
Vtest / Vcalc
139
2.50
21 tests
1.00
0.50
III-1.85-03
III-1.85-03b
I-03-2
I-03-4
IV-2175-1.85-03
C1-42-1.85-06
III-2.5-03
1.50
DC1-42-1.85-03
DS1-42-1.85-03
DL1-42-1.85-03
SC1-42-1.85-03(le)
SS1-42-1.85-03
SS1-75-1.85-03b
DC1-42-1.85-06
DS1-42-1.85-06
DL1-42-1.85-06
DC1-42-2.50-03
DL1-42-2.50-03
SC1-42-2.50-03
SS1-42-2.50-03
SL1-42-2.50-03
2.00
1.0
0.00
Tension Chord
Compression Chord
Figure 5-16: Series VI: Loaded Chord- STM conservatism
5.2.4 Summary of Strength Results

The strut-and-tie provisions developed by TxDOT Project 0-5253 and implemented for
inverted-T beams as described in Chapter 2 produced reasonably conservative estimates
of the strength of all the inverted-T specimens. The STM procedures offer a more rational
approach to designing inverted-T deep beams than sectional design, as they inherently
consider all failure modes for the ledges, web, and bearing points. A summary of the
comparisons of Vtest/Vcalc from the strut-and-tie models for all specimens tested is
provided in Figure 5-17 and Table 5-14.
The STM provisions showed no bias to ledge depth, number of point loads, beam depth,
or chord loading; i.e., they captured the effects of these variables adequately. The
provisions did however show limited bias to ledge length and reinforcement ratio, but
produced conservative results in all cases with reasonable safety margins. The strength
results from the thirty three tests in the experimental program thus support a
recommendation to use TxDOT Project 0-5253 STM provisions for the strength design of
inverted-T beams. It is also recommended based on strength observations to avoid cut-off
ledges because the specimens were more prone to ledge failures and strengths estimates
with lower safety margins.
140
a/d = 1.85
a/d = 2.5
2.5
1.5
0.5
33 tests
Inverted-T Beams (TxDOT Project 0-6416)

Figure 5-17: Conservatism of STM provisions as applied to inverted-T beams
Table 5-14: Summary of experimental/calculated shear capacity
Inverted-T Beams Evaluation Database

(0-6416)
(0-5253)
33 tests
179 tests
Min
1.05
0.73
Max
2.17
4.14
Mean
1.50
1.54
*
0.0%
0.6%
Unconservative
Standard deviation
0.33
0.43
Coefficient of Variation**
0.22
0.28
* Unconservative = % of tests for which Vtest /Vcalc < 1.0
** Coefficient of Variation = Standard Deviation / Mean
5.3
DIAGONAL CRACKING UNDER SERVICE LOADS
In TxDOT Project 0-5253, a recommendation to limit shear stresses in deep beams at

service loads was given. Such a limit is useful for the prevention and control of diagonal
cracks. The limit was given through a simple and reasonably conservative equation. This
equation and its applicability to inverted-T beams is discussed herein.
141
5.3.1 Background
For durability considerations, it is necessary to limit or prevent diagonal cracking under
service loads in reinforced concrete members. In this section, trends between the shear
force at first diagonal cracking and pertinent variables are investigated.
The main types of cracks found in inverted-T beams are depicted in Figure 5-18. The
focus of the current project is diagonal cracking, including both flexure-shear cracks and
web-shear cracks. Flexure-shear cracks form after or simultaneously with the flexural
cracks and extend from the tip of the flexural crack towards the load or, as in the case of
inverted-T beams, to the compression chord directly above the load. Web-shear cracks
occur independently from flexural cracking and form when the principal tension stress in
the web of the member exceeds the tensile strength of the concrete. In deep beams,
transverse tensile stresses develop due to the spreading of compressive stresses in bottleshaped struts and also contribute to the width of flexure-shear cracks (MacGregor and
Wight, 2005).
Web-shear crack
Punching-shear crack
Flexure-shear crack
Flexural crack
Figure 5-18: Types of cracks in reinforced concrete inverted-T deep beams.
With regard to the current project, no distinction was made between flexure-shear or
web-shear cracks when evaluating the serviceability behavior of the test specimens. Both
were treated as diagonal cracks.
ACI-ASCE Committee 326 report (1962) identified the major variables that affect the
diagonal cracking load of reinforced concrete beams. These variables are: section size
(bwd), tensile strength of concrete (that can be taken as a function of ( )), longitudinal
reinforcement ratio (l), and moment to shear ratio at the critical section (M/V). Since
M/V is constant in the main shear span of beams loaded with concentrated loads, the
shear span-to-depth ratio (a/d) can be used in lieu of M/V.
In TxDOT Project 0-5253, the effects of these variables (bwd, , a/d, l, d) on the
diagonal cracking load of deep beams was assessed with data from the experimental
program and the literature. The resulting information was then used to recommend an
142
equation that estimates a load below which service loads must remain to limit diagonal
cracking.
As with all cracking data, considerable scatter was observed in diagonal cracking loads of
deep beams. The deep beam project showed that the primary variables affecting the
diagonal cracking load were the section size, the tensile strength of the concrete, and the
a/d ratio. The longitudinal reinforcement ratio was not explicitly examined as part of the
experimental program and although it may contribute to the diagonal cracking load to
some degree, there was insufficient data to fully evaluate the variable. The effect of the
section depth was found to be small and erratic based on test results and data from the
literature. The quantity of web reinforcement was also not found to have a significant
effect on the load at first diagonal cracking as the steel is not typically engaged until after
the section cracks.
Two approaches to limit diagonal cracking under service loads were considered in
TxDOT Project 0-5253. The first was associated with the strut-and-tie models and
involved limiting the force generated in the diagonal strut by service loads to a specified
percentage of its unfactored capacity. This was, in effect, a separate service-load strutand-tie analysis that could be integrated easily with the ultimate-strength since the same
model would be used with different applied loads and efficiency factors.
It was decided not to recommend this approach however due to a fundamental flaw.
Strut-and-tie modeling is a lower-bound approach and therefore is intended to be used in
design to provide a safe estimate for ultimate strength. In order for a member to reach the
ultimate capacity estimated by the STM, redistribution of forces and cracking is expected
to occur. The analysis procedure is not intended to accurately estimate service level
stresses or limit cracking. Thus the researchers of TxDOT Project 0-5253 concluded that
a STM-based approach to limit diagonal cracking under service loads would not be
appropriate.
The resulting recommendation was a separate service load check. A reasonably
conservative equation was developed to estimate the diagonal cracking load of deep
beams based on data from the experimental program and the deep beam database. This
check was simple and more theoretically justified than the STM approach. In this task,
the applicability of the deep beam diagonal cracking load estimate is evaluated for use
with inverted-T beam design.
5.3.2 Diagonal Cracking in Inverted-T Beams
Several experimental variables were evaluated to determine the effect on the diagonal
cracking load of inverted-T deep beams. Along with the section size (bwd), the tensile
143
strength of the concrete ( ), and the a/d ratio, the effect of the ledge geometry,
reinforcement ratio, and number of point loads were assessed.
Diagonal Cracking Load, Vcrack (kip)
The effect of section size on the diagonal cracking load of the specimens in the
experimental program is shown in Figure 5-19. As expected, the diagonal cracking load
increases as the shear area (bwd) increases. Prior to diagonal cracking, the member
behaves elastically and thus the entire section contributes to the diagonal cracking
strength. As the specimen cracks the layout and material strength influence the ultimate
capacity of the inverted-T beams.
400
N = 25
350
300
250
200
150
100
50
0
200
400
600
800
1000
1200
1400
1600
Shear Area, bwd (in2)
Figure 5-19: Effect of section size on diagonal cracking load of inverted-T beams.
There is scatter within each section size in Figure 5-19, which is likely a result of the
other variables that contribute to the diagonal cracking load of deep beams. To determine
the effect of the tensile strength of the concrete, the diagonal cracking load was
normalized by the shear area and plotted versus
144
in Figure 5-20.
0.4
N = 25
Vcrack / bwd (ksi)
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
0
10
20
30
40
50
60
70
80
90
(psi)
Figure 5-20: Effect of tensile strength on diagonal cracking load of inverted-T beams.
It has been widely accepted that the diagonal cracking load is a function of the square
root of the compressive strength of concrete; the cracking load is expected to increase
with the tensile strength of the concrete that is a function of the square root of the
compressive strength of concrete. The data plotted in Figure 5-20 shows considerable
scatter without a clear trend between the diagonal cracking load and the square root of the
compressive strength of the concrete. It is important to note here that concrete strengths
did not vary substantially across test specimens and therefore the variable could not be
adequately assessed using the inverted-T specimens.
Based on past research and TxDOT Project 0-5253, the diagonal cracking loads are
normalized by the shear area times the square root of the compressive strength of
concrete; i.e., the influence of section size and concrete strength on cracking were
included in the analyses.
The effect of the shear span-to-depth ratio on the diagonal cracking load of inverted-T
beams is shown in Figure 5-21. In the figure, the normalized diagonal cracking load can
be seen to decrease as the a/d ratio increases. This observed trend is similar to that
observed for compression-chord loaded beams in the TxDOT Project 0-5253. As
discussed in Birrcher et al. (2008), the downward trend is related to the change in
principle tensile stress distribution that occurs when the a/d ratio changes. Lower a/d
ratios are subjected to a more complex state of stress due to the proximity of the applied
load to the support. As a result there is considerable scatter at an a/d ratio of 1.85. As the
a/d ratio exceeds 2, the beam tends more towards slender beam behavior and diagonal
cracking strengths are concentrated at 2 bwd (in psi units).
145
N = 25
bw d
Vcrack /
3
2
1
0
0
0.5
1.5
2.5
Shear Span to Depth Ratio (a/d)

Figure 5-21: Effect of a/d ratio on diagonal cracking load of inverted-T beams.
As the longitudinal reinforcement ratio in the current study was not varied significantly,
the effect of that parameter could not be assessed with respect to crack width. The effect
of beam depth on the diagonal cracking load was evaluated in both projects. The results
from TxDOT Project 0-5253 indicate that the diagonal cracking load decreased with
increasing beam depth, but the effect was small and erratic. Other deep beam literature
sources revealed conflicting results. No clear trend is revealed by plotting the data from
the current inverted-T beam experimental program in Figure 5-22. No depth effect was
apparent at the different a/d ratios.
a/d = 1.85
a/d = 2.50
N = 25
bw d
5
4
Vcrack /
3
2
1
0
0
10
20
30
40
50
60
70
80
Effective Depth, d (in)

Figure 5-22: Effect of depth on the diagonal cracking load of beams of inverted-T beams.
The effect of web reinforcement ratio was also evaluated in the experimental program of
the current project. As discussed in Section 4.5, a slight decrease in diagonal cracking
146
load was observed for the directly comparable inverted-T beams; illustrated in Figure
5-23. Three out the four comparisons show this trend, however, due to the limited
number of direct comparisons and the weak trend observed it can be concluded that web
reinforcement ratio has a negligible effect on the load at first diagonal cracking. A
similar observation can be made when the normalized diagonal cracking loads for all test
specimens are examined (Figure 5-24). Likewise, no significant effect of web
reinforcement ratio on the diagonal cracking load of deep beams was observed in Project
0-5253.
0.3%/0.3%
0.6%/0.6%
4.5
bw d
4
3.5
3
SS3-42-2.50-06
SS3-42-2.50-03
SL3-42-1.85-06
SL3-42-1.85-03
0.5
DL1-42-1.85-06
DL1-42-1.85-03
1.5
DS1-42-1.85-06
DS1-42-1.85-03
Vcrack /
2.5
Reinforcement Ratio, /
v h
Figure 5-23: Effect of web reinforcement on the diagonal
cracking load of directly
comparable inverted-T specimens.
a/d = 1.85
a/d = 2.50
N = 25
bw d
5
4
Vcrack /
3
2
1
0
0.0%
0.1%
0.2%
0.3%
0.4%
0.5%
0.6%
0.7%
Reinforcement Ratio, v = h
Figure 5-24: Effect of web reinforcement on diagonal cracking load of inverted-T beams.
147
The variables unique to the current experimental program, including ledge length, ledge
depth, and the number of point loads, were also investigated to determine their effect on
the diagonal cracking load of inverted-T beams.
As discussed in Section 4.3, the length of the ledge on the inverted-T specimens was
found to have a positive effect on the diagonal cracking load. An increase in the
normalized cracking load was observed with increasing ledge lengths between directly
comparable beams. As shown in Figure 5-25, the trend can also be seen when all
cracking loads in the experimental program are considered. Even with the scatter in the
data there is a clear increase in diagonal cracking load with longer ledge lengths for both
a/d ratios. Thus the cracking load results reinforce the strength-based recommendation for
avoiding the use of cut-off ledges in practice. Not only do cut-off ledges generate
relatively low web-shear strengths compared to other ledge lengths, but they also cause
adjacent webs to crack at lower shear stresses than with longer ledges.
Thus given the recommendation not to use cut-off ledges and the inherent scatter in test
results, it was decided to consider a conservative limit on shear stress at diagonal
cracking that encompasses short and long ledges without explicitly accounting for ledge
length.
a/d = 1.85
a/d = 2.50
N = 25
bwd
5
4
Vcrack /
2
1
0
Cut-off
Short
Long
Figure 5-25: Effect of ledge length on the diagonal cracking load of inverted-T beams.
A weaker trend between diagonal cracking load and ledge depths was noted in Section
4.4. According to the results of the direct comparisons, deeper ledges tended to
experience a slightly higher normalized cracking load than shallow ledges. The general
comparison shown in Figure 5-26 supports this observation for beams tested at an a/d
ratio of 1.85, but there is no apparent trend for the beams tested at an a/d ratio of 2.50.
The observed trends however are relatively weak and the depth of the ledge is not
148
considered to be a primary variable affecting the diagonal cracking load of inverted-T

beams.
a/d = 1.85
a/d = 2.50
N = 25
bw d
5
4
Vcrack /
2
1
0
Shallow
Deep
Figure 5-26: Effect of ledge depth on the diagonal cracking load of inverted-T beams.
A third variable unique to this study is the number of point loads. The direct comparisons
relating to the number of point loads did not reveal an effect on the diagonal cracking
load as discussed in Section 4.6. As shown in Figure 5-27, no significant trend can be
observed from the general comparison of data from all tests of the experimental program.
a/d = 1.85
a/d = 2.50
N = 25
bw d
5
4
Vcrack /
3
2
1
0
1 Point Load
3 Point Loads
Figure 5-27: Effect of multiple point loads on the diagonal cracking load of inverted-T
beams.
In summary, the primary variables found to affect the diagonal cracking loads were the
shear area, bwd, and the shear span-to-depth ratio, a/d. The tensile strength of the
concrete, represented by , was also considered for normalization purposes although
149
no clear trend was observed in this data set. Ledge length showed some influence on the
load at first diagonal cracking; the influence was however not large enough to warrant
including it in limits on service-load shear stresses. Cut-off ledges tended to crack at
lower loads than beams with other ledge lengths. Such an observation reinforces the
strength-based recommendation for not using cut-off ledges in practice.
Estimating Diagonal Cracking Loads
TxDOT Project 0-5253 examined empirical equations from the deep beam literature and
determined that they did not have the level of accuracy and conservatism required for the
current task. Many of these equations only applied to slender beams or those of
considerably smaller shear area and thus were not applicable to the compression-chord
loaded bent caps under investigation in that project. Very limited data was available prior
to this project of crack width progression for inverted-T beams.
In TxDOT project 0-5253, a simple and reasonably conservative equation was selected
for determining the diagonal cracking load as is was deemed the most reasonable
approach given the scatter in the data. The equation proposed in TxDOT Project 0-5253
is compared to diagonal cracking data of a deep beam database (from Project 0-5253) and
results for this projects experimental program (Project 0-6416) in Figure 5-28.
As shown in Figure 5-28, the TxDOT Project 0-5253 equation estimates a lower bound
on shear stress at first diagonal cracking not only for compression-chord loaded beams
but also for inverted-T beams. The four inverted-T beams that showed diagonal cracking
at shear stresses lower than estimated by the equation had shallow and/or cut-off ledges
(SC1-42-1.85-03, SL1-42-1.85-03, SC3-42-1.85-03, SC3-42-2.50-03). Such an
observation strengthens the recommendation for eliminating cut-off ledges in practice
based on strength. If cut-off ledges were not used, the given equation would be
conservative for all but one of the inverted-T specimens. Overall the performance was
comparable to the compression-chord loaded beams for which it was developed. It is
therefore concluded that the existing equation for limiting shear stress at service loads in
rectangular beams is equally applicable to inverted-T beams.
150
0-6416
0-5253
Other Compression-Chord Loaded Studies
10
9
8
7
6
5
4
3
2
1
0
N = 86
0.0
0.5
1.0
1.5
2.0
2.5
3.0
0-6416 Inverted-T Beams

10
9
8
7
6
5
4
3
2
1
0
N = 24
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Figure 5-28: Assessment of proposed equation for estimate of diagonal cracking.
5.3.3 Design Implications

To limit diagonal cracking in inverted-T bent caps under service loads, the following
approach should be taken. The shear in the member due to unfactored service loads
should be computed and compared to the estimated diagonal cracking load given by the
following equation:
151

but not greater than 5
with
(5.1)
nor less than 2
a = shear span (in.) (Refer to Figure 3-2)

d = effective depth of the member (in.)
fc = compressive strength of concrete (psi)
bw = web width of the member (in.)
A member is not expected to crack under normal loading conditions if the service level
shear is less than the estimated diagonal cracking load. If the service-load shear is greater,
the design of the member can be altered to increase the value of the diagonal cracking
load. The section size (bwd) of the member can be increased, but a greater depth will
reduce the a/d ratio for a member with a fixed span length. Theoretically, a higher
concrete strength can also be specified to increase the tensile capacity of the concrete but
based on the lack of trend observed in Figure 5-20, this is likely more beneficial for
compression-chord loaded beams than inverted-T beams. If these options are not
practical, additional web reinforcement can be provided to help restrain the diagonal
crack widths under service loads. Note that, as discussed in Section 4.5.3, there are lesser
benefits of supplying additional web reinforcement for the purpose of crack width
control, especially around service level loads.
The service load check provides an indication of the likelihood of the formation of
diagonal cracks in service. As discussed in Birrcher et al. (2009), the proposed equation
can also be used to determine at what load an inverted-T or deep beam is expected to
crack, if the service load shear exceeds the expected diagonal cracking load. This will
indicate if a bent cap can be expected to crack under dead loads alone; a scenario that
should be avoided.
5.3.4 Summary and Conclusions
The variables that affect the diagonal cracking load of inverted-T beams were
investigated with data from the current experimental program. The primary variables
were found to be the shear area, bwd, and the a/d ratio of the member. Web reinforcement
ratio and ledge length were found to have limited influence with the exception of several
beams with cut-off ledges. The diagonal cracking load recorded in inverted-T beam tests
compared well with the lower-bound empirical equation developed for rectangular
beams. Only four of the inverted-T beams showed lower diagonal cracking loads than
estimated by the equation but their values were well within the acceptable range of
scatter. Those beams had mostly cut-off ledges. Given the relatively low diagonal
cracking load recorded for specimens with cut-off ledges, it is not recommended to
design inverted-T beams with cut-off ledges.
152
It is recommended to use the same diagonal cracking equation for inverted-T beams and
rectangular beams. Limiting the shear stresses at service loads to below values estimated
by the equation will reduce the likelihood of diagonal cracking at service loads but will
not guarantee that a reinforced concrete inverted-T beam will remain uncracked under
service loads. Inconsistencies in design assumptions and actual field conditions can result
in stresses higher than those accounted for in this equation. However, the process is
simple and should significantly reduce diagonal cracking in service. It also draws the
attention of the designer to the serviceability aspect of design for inverted-T beams,
which is lacking in the AASHTO LRFD Design Specifications and TxDOT Bridge
Design Manual as both of these methods only limit the stress in the hanger reinforcement
at service loads.
5.4
CORRELATION OF MAXIMUM IN-SERVICE DIAGONAL CRACK WIDTH

ULTIMATE SHEAR STRENGTH
WITH
As discussed in Chapter 1, diagonal cracking has been observed in multiple inverted-T

straddle bent caps in existing structures. During field inspections, engineers are tasked
with assessing the amount of damage in the members. Currently there is little information
in the literature that presents a reliable method for relating the width of diagonal cracks to
the amount of distress in an inverted-T beam. In this section, a method for comparing the
maximum width of an in-service diagonal crack with the residual capacity of an invertedT straddle bent caps is developed.
TxDOT Project 0-5253 created a simple chart that related the maximum diagonal crack
width on a deep beam to the load on the member. Unlike an analytical approach which
would require complex models and multiple assumptions to estimate diagonal crack
widths, the empirically-based chart can easily be used by engineers in the field. That
chart is currently not approved for use with inverted-T beams.
5.4.1 Background
Limited crack width data was available in the literature, none of which could be used to
relate crack widths to load levels in inverted-T beams. Only results from this study were
applicable to inverted-T beams. The following presents an overview of crack width data
for deep rectangular beams to frame the issue for inverted-T beams.
Birrcher et al. (2009) reviewed the few studies on deep beam shear that included the
serviceability information to determine what factors affected the crack widths. Although
none of the specimens were loaded on the tension chord, the conclusions were made on
deep beams and have some relevance to the current task.
153
Effect of web reinforcement on the diagonal crack widths of rectangular deep beams
The review of deep beam literature by Birrcher et al. (2009) revealed that distributed web
reinforcement was reported as being the most important variable in controlling diagonal
crack widths. From limited test data in which crack widths were reported, additional
vertical reinforcement beyond a certain minimum was found to have little effect on
further restraining the diagonal crack widths in the service load range. Above 50% of the
load capacity however, Bracci et al., (2000) and Young et al., (2002) found that
specimens with 0.6% reinforcement in the vertical direction had narrower crack widths
than those with 0.3% reinforcement. These results were obtained from full-scale (33 in. x
36 in.) specimens and agree well with the findings for inverted-T specimens. It was
observed for the inverted-T specimens that increasing the amount of web reinforcement
from 0.3% to 0.6% had limited effect on the diagonal crack widths at typical service load
levels, but was found to make a significant difference at higher loads as discussed in
Chapter 4.
TxDOT Project 0-5253 came to a similar conclusion when evaluating the diagonal crack
widths for deep beams. In general the amount of transverse reinforcement was found to
have limited effect on the maximum width of the diagonal crack at first cracking and
more significantly at higher loads. Additionally, the spacing of the stirrups was found in
that study not to significantly affect the width of diagonal shear cracks as long as the
existing spacing limit of d/4 or 12 in. in Section 5.13.2.3 of AASHTO LRFD 2012 was
upheld.
Effect of a/d ratio on the diagonal crack widths of rectangular deep beams
A review of the literature on reinforced concrete deep beams resulted in mixed
observations on the effect of the a/d ratio on diagonal crack widths. A few studies
discussed in Birrcher et al. (2009) observed some increases in diagonal crack widths as
the a/d ratio increased, but the data was not conclusive. Others saw no trend between the
diagonal crack widths and a/d ratio. In their own work, Birrcher et al. saw a trend when
examining some sets of comparable beams and no trend in others. When all of the
specimens were plotted collectively, a slight trend with the a/d ratio was observed but it
was attributed more to the scatter in diagonal crack width data. It was therefore
concluded that the a/d ratio did not have a significant effect on the diagonal crack widths
of rectangular deep beams.
5.4.2 Diagonal Crack Widths in Inverted-T Beams
All of the test specimens of the current project are listed in Table 5-15. The specimens
that failed in punching shear or flexure were also included as they had reached their
predicated shear capacity at failure. The results of thirty three inverted-T beam shear tests
154
were used in the current task. Nineteen specimens were tested at an a/d ratio of 1.85 and
fourteen were tested at an a/d ratio of 2.50. The overall height of the specimens was
either 42 or 75 in. and the width was constant at 21 in. The ledge depth and length varied
as discussed in Section 3.2. The web reinforcement was either 0.3% or 0.6% in both
orthogonal directions. Two specimens were tested with 0.6% reinforcement in the
vertical and 0.3% in the horizontal direction. Several different bearing plate sizes were
used as shown in Table 5-15.
The measured maximum diagonal crack widths were plotted versus the percent of
maximum applied load in the figures included in this section. Plotting the crack data from
the experimental program in this manner enabled specimens of different size, ledge
geometries, and concrete strength to be compared. It also enabled a correlation to be
made between the maximum diagonal crack width and the percent of the ultimate
capacity. The size of the bearing plates were assumed to have no effect on the width of
the diagonal cracks.
155
Table 5-15: Specimens used in correlating crack width to capacity.

Specimen
b h
in. in.
d
in.
Ledge
Depth
Ledge
Length
Point
Loads
Support
Plate in.
Load
Plate in.
DS1-42-1.85-03
21 42
37.6
h/2
Short
16 x 20
26 x 9
0.3% 0.3%
1.85
DS1-42-2.50-03
21 42
37.6
h/2
Short
16 x 20
26 x 9
0.3% 0.3%
2.50
DS1-42-1.85-06
21 42
37.6
h/2
Short
16 x 20
26 x 9
0.6% 0.6%
1.85
DS1-42-2.50-06
21 42
37.6
h/2
Short
16 x 20
26 x 9
0.6% 0.6%
2.50
DL1-42-1.85-06
21 42
37.6
h/2
Long
16 x 20
26 x 9
0.6% 0.6%
1.85
DL1-42-2.50-06
21 42
37.6
h/2
Long
16 x 20
26 x 9
0.6% 0.6%
2.50
SS3-42-1.85-03
21 42
37.6
h/3
Short
16 x 20
18 x 9
0.3% 0.3%
1.85
SS3-42-2.50-03
21 42
37.6
h/3
Short
16 x 20
18 x 9
0.3% 0.3%
2.50
SS3-42-2.50-06 (f)
21 42
37.6
h/3
Short
16 x 20
18 x 9
0.6% 0.6%
2.50
SC3-42-1.85-03
21 42
37.6
h/3
Cut-off
16 x 20
18 x 9
0.3% 0.3%
1.85
SC3-42-2.50-03
21 42
37.6
h/3
Cut-off
16 x 20
18 x 9
0.3% 0.3%
2.50
DS3-42-2.50-03
21 42
37.6
h/2
Short
16 x 20
18 x 9
0.3% 0.3%
2.50
DL1-42-1.85-03
21 42
37.6
h/2
Long
16 x 20
26 x 9
0.3% 0.3%
1.85
DL1-42-2.50-03
21 42
37.6
h/2
Long
16 x 20
26 x 9
0.3% 0.3%
2.50
SL3-42-1.85-03
21 42
37.6
h/3
Long
16 x 20
18 x 9
0.3% 0.3%
1.85
SL3-42-1.85-06
21 42
37.6
h/3
Long
16 x 20
18 x 9
0.6% 0.6%
1.85
DC1-42-1.85-06
21 42
37.6
h/2
Cut-off
30 x 21
30 x 10
0.6% 0.6%
1.85
SS1-75-1.85-03
21 75
68.2
h/3
Short
16 x 20
30 x 10
0.3% 0.3%
1.85
DC3-42-1.85-03
21 42
37.6
h/2
Cut-off
16 x 20
18 x 9
0.3% 0.3%
1.85
DS3-42-1.85-03
21 42
37.6
h/2
Short
16 x 20
18 x 9
0.3% 0.3%
1.85
SS1-42-1.85-03
21 42
37.6
h/3
Short
16 x 20
26 x 9
0.3% 0.3%
1.85
SS1-42-2.50-03
21 42
37.6
h/3
Short
16 x 20
26 x 9
0.3% 0.3%
2.50
DC1-42-2.50-03
21 42
37.6
h/2
Cut-off
16 x 20
26 x 9
0.6% 0.6%
1.85
DL3-42-1.85-03 (f) 21 42
a/d ratio
37.6
h/2
Long
16 x 20
18 x 9
0.3% 0.3%
1.85
21 42
37.6
h/3
Long
16 x 20
26 x 9
0.3% 0.3%
2.50
SC1-42-2.50-03 (r) 21 42
37.6
h/3
Cut-off
16 x 20
26 x 9
0.3% 0.3%
2.50
Short
16 x 20
26 x 9
0.6% 0.3%
1.85
SL1-42-2.50-03
DS1-42-1.85-6/3
21 42
37.6
h/2
DS1-42-2.50-6/3
21 42
37.6
h/2
Short
16 x 20
26 x 9
0.6% 0.3%
2.50
SC1-42-1.85-03 (le) 21 42
37.6
h/3
Cut-off
30 x 21
18 x 9
0.3% 0.3%
1.85
DC1-42-1.85-03
21 42
37.6
h/2
Cut-off
30 x 21
18 x 9
0.3% 0.3%
1.85
SC1-42-1.85-03b
21 42
37.6
h/3
Cut-off
30 x 21
30 x 10
0.3% 0.3%
1.85
DC1-42-1.85-03b
21 42
37.6
h/2
Cut-off
30 x 21
30 x 10
0.3% 0.3%
1.85
SS1-75-2.50-03
21 75
68.2
h/3
Short
16 x 20
30 x 10
0.3% 0.3%
2.50

156
5.4.2.1 Effect of test variables on the diagonal crack widths of inverted-T beams
As discussed in Section 5.3.1, two different types of diagonal cracks exist in reinforced
concrete deep beams: flexure-shear cracks and web-shear cracks. In the analysis of the
crack width data, no distinction was made between the two types of diagonal cracks. The
variables that are suspected of affecting the width of the diagonal cracks in the inverted-T
beams were explored using test results of this experimental program.
The maximum diagonal crack widths were examined using directly comparable beam
sets in Chapter 4. In general, the amount of web reinforcement was found to affect the
maximum width of diagonal cracks throughout the loading history; as illustrated in the
two sets of comparisons in Figure 5-29. Specimens subjected to a given percent of
ultimate load were found to have wider crack widths when reinforced with 0.3% web
reinforcement ratio than those with 0.6%. In other words, the larger reinforcement ratio
resulted in greater diagonal crack width restraint.
100
90
80
70
60
50
40
30
20
10
0
DS1-42-1.85-03
0.3%/0.3%
0.3%/0.3%
0.3%/0.3%
SL3-42-1.85-03
DS1-42-1.85-06
0.6%/0.6%
0.6%/0.6%
0.6%/0.6%
SL3-42-1.85-06
100 100
90 90
80 80
70 70
60 60
50 50
40 40
30 30
20 20
a/d = 1.85
10 10
4 tests
0 0
0 0 0 0.02 0.020.02 0.04 0.040.04 0.06 0.060.06 0.08 0.080.08 0.1 0.1 0.1
Figure 5-29: Effect of web reinforcement on diagonal crack widths of test specimens.
The serviceability effects of the other variables in the experimental program were also
considered in Chapter 4. Except for the web reinforcement ratio, the direct comparisons
did not reveal any appreciable trends between the inverted-T variables and the crack
width progression. To some extent, the ledge length contributed to the width of diagonal
cracks. However, the effect was small in relation to the scatter associated with the crack
widths of similar specimens. The other variables investigated in the experimental
157
program, including a/d ratio, ledge depth, number of point loads, and web depth, did not
have a significant effect on the width of the diagonal cracks. Thus, a chart correlating the
diagonal crack width to the residual capacity of the inverted-T beams was developed
considering the quantity of web reinforcement as the primary variable; as was done in
TxDOT Project 0-5253.
5.4.2.2 Correlation of Crack Width to Residual Capacity
The crack width data from the thirty three specimens used in the current task are plotted
in the following figures. Each of the nineteen beams included in Figure 5-30 were tested
at an a/d ratio of 1.85. Figure 5-31 contains the crack width data from fourteen beams
tested at an a/d ratio of 2.5. In both plots the data was separated into three groups by the
quantity of web reinforcement: 0.3% reinforcement in each direction, 0.6% vertical
reinforcement and 0.3% horizontal reinforcement, and 0.6% reinforcement in each
direction.
As shown in Figure 5-30, there is a consistent trend between the maximum diagonal
crack width and the amount of web reinforcement for an a/d ratio of 1.85. A power
function trend line was fitted through the data for beams with equal reinforcement ratios
in each direction . The square of the correlation coefficient, R2, was provided next to each
tend line. This value quantifies the error between the trend line and the data points, with
an R2 value of 1.0 representing a perfect fit. In Figure 5-30, it is shown that the R2 value
increases as the quantity of web reinforcement increases. This indicated that with a
smaller amount of reinforcement there was more scatter in the diagonal crack width data
for beams tested at an a/d ratio of 1.85. As the amount of web reinforcement was
increased from 0.3% to 0.6% in each direction, the maximum width of diagonal cracks
was more consistent. This observation was in agreement with the crack width findings for
rectangular reinforced concrete deep beams in TxDOT Project 0-5253.
One specimen was constructed with 0.6% in the vertical and 0.3% in the horizontal
direction with an a/d ratio of 1.85. At lower loads, below 60% of the maximum applied
load, its crack widths resembled those of beams with 0.3% reinforcement in both
directions. As the load increased past 60% of maximum, the diagonal crack widths
tended more towards an average of the crack widths of specimens with 0.3% or 0.6%
reinforcement ratios exclusively. Such findings suggest that a beam constructed with
mixed reinforcement would mimic the behavior of a beam with equal reinforcement in
both directions corresponding to approximately the average of the vertical and horizontal
reinforcement ratios.
158
0.3/0.3
0.6/0.3
R = 0.85
100
0.6/0.6
R = 0.83
90
80
70
60
50
40
30
20
a/d = 1.85
19 tests
10
0
0
0.02
0.04
0.06
0.08
0.1

Figure 5-30: All crack width data at an a/d ratio of 1.85 used in this task with trend lines.
Crack width data from fourteen tests with an a/d ratio of 2.5 was also examined. As
shown in Figure 5-31, the trend for the reinforcement ratios is consistent with that shown
for an a/d ratio of 1.85. In fact, there is even less scatter in the beams tested at an a/d of
2.50 than those at 1.85 for specimens with 0.3% reinforcement in each direction, as
indicated by the higher R2 value. Higher amounts of web reinforcement showed lower
scatter in the diagonal crack width data for a/d = 2.5 as well. One specimen with a
vertical reinforcement ratio of 0.6% and horizontal reinforcement ratio of 0.3% was
tested at an a/d ratio of 2.50. Similarly to the specimen tested at an a/d=1.85 with unequal
reinforcement ratios, the crack width progression of this beam corresponded roughly to
that of a beam that would be tested with equal reinforcement ratios in both directions
equal to the average of both reinforcement ratios.
159
0.3/0.3
0.6/0.3
0.6/0.6
R = 0.90
R = 0.88
100
90
80
70
60
50
40
30
20
a/d = 2.50
14 tests
10
0
0
0.02
0.04
0.06
0.08
0.1

Figure 5-31: All crack width data at an a/d ratio of 2.50 used in this task with trend lines.
In Figure 5-32 and Figure 5-33, the power function trend lines were replaced with
straight line segments for a/d=1.85 and a/d=2.5 respectively. Combining the two figures
in Figure 5-34 showed minimal differences between trend lines for beams tested at the
two a/d ratios. The diagonal crack widths were found to vary primarily due to the amount
of web reinforcement present in the beam, and a/d ratios have little effect.
0.3/0.3
0.6/0.3
0.6/0.6
100
90
80
70
60
50
40
30
20
a/d = 1.85
19 tests
10
0
0
0.02
0.04
0.06
0.08
0.1

Figure 5-32: All crack width data at an a/d ratio of 1.85 with straight line approximations.
160
0.3/0.3
0.6/0.3
0.6/0.6
100
90
80
70
60
50
40
30
20
a/d = 2.50
14 tests
10
0
0
0.02
0.04
0.06
0.08
0.1

Figure 5-33: All crack width data at an a/d ratio of 2.50 with straight line approximations.
Several percentages of ultimate capacity were tabulated along with their corresponding
maximum diagonal crack width values. For each percentage of ultimate load selected, a
range of scatter based on the variance of the crack widths in the plot was also assigned.
The resulting chart comparing the percent of ultimate capacity of a member to a
measured crack width with respect to the members reinforcement ratio is included in
Figure 5-35.
a/d = 1.85
0.3/0.3
0.6/0.3
a/d = 2.50
0.6/0.6
0.3/0.3
0.6/0.3
0.6/0.6
100
90
80
70
60
50
40
30
20
10
33 tests
0
0
0.02
0.04
0.06
0.08
Figure 5-34: All crack width data with straight line approximations.
161
0.1
Load on the Member, Quantified as a Percent of Ultimate Capacity on Average ( scatter)

wmax (in.)
0.01
0.02
0.03
0.04
0.05
Reinforcement
30 ( 10)
50 ( 15)
65 ( 15)
75 ( 15)
80 ( 15)
v = 0.003 h = 0.003
40 ( 10)
65 ( 10)
85 ( 10)
~ Ultimate
~ Ultimate
v = 0.006 h = 0.006
Notation:
wmax = maximum measured diagonal crack width (in.)
v = reinforcement ratio in vertical direction (v = Av / bsv)
h = reinforcement ratio in horizontal direction (h = Av / bsh)
Av & Ah = total area of stirrups or horizontal bars in one spacing (in.2)
sv & sh = spacing of stirrups or horizontal (skin reinf.) bars (in.)
b = width of web (in.)
0.06
90 ( 10)
~ Ultimate
Directions:
1). Determine v and h for bent cap
2). Measure maximum diagonal crack width, wmax, in inches
3). Use chart with wmax, v, and h to estimate % of capacity.
Interpolate for intermediate values v and h. For unequal v and
h use the average of the two when reading off the chart.
Important Notes:
In this chart, the maximum width of the primary diagonal crack in a shear-critical member is linked to the load on the member, quantified
as a percent of its ultimate capacity. The intent of this chart is to aid field engineers in evaluating residual capacity in diagonally-cracked,
reinforced-concrete bent caps subject to concentrated loads at a/d ratios between 1.0 and 2.5. This chart was developed from crack width
data from 33 tests of simply supported reinforced concrete inverted-T beams with overall heights between 42 and 75. The testing was
conducted at a/d ratios of 1.85 and 2.5.
This chart should be used in conjunction with sound engineering judgment with consideration of the following limitations:
-variability in crack widths in general ( scatter)
-differences between field and laboratory conditions
-members loaded at a/d < 1.85 may be at slightly higher % of capacity
-implications of an unconservative estimate of capacity
This chart is intended to be used for inverted-T bent caps. Not applicable with reinforcement ratios above 0.6%
Figure 5-35: Proposed chart that links diagonal crack width to percent of ultimate capacity of inverted-T bent caps.
162
The inverted-T diagonal crack width chart indicates a larger percent maximum load for a
given crack width than that developed by TxDOT Project 0-5253 for compression-chord
loaded deep beams with reinforcement ratios of 0.3%. It appears that the presence of
tension in the web of the inverted-T members affects the width of the diagonal cracks.
The chart was developed with data from specimens constructed with the minimum
reinforcement ratio of 0.3% and an upper bound of 0.6% in both directions. The
percentage of ultimate capacity for inverted-T bent caps with less than 0.3%
reinforcement can be conservatively estimated using the crack width data for the lower
bound. This chart should be used with caution and sound engineering judgment for
inverted-T bent caps that are constructed with reinforcement ratios greater than 0.6%. It
is assumed that the crack widths will continue to decrease with increased amounts of
reinforcement at a given percent of ultimate capacity, but specimens with greater than
0.6% reinforcement in either direction were not tested as part of the current project.
Likewise, no bent cap inspected in this project was constructed with a web reinforcement
ratio greater than 0.6%.
For the distressed in-service inverted-T bent caps that fall within the reinforcement ratio
range given in Figure 5-35, it is acceptable to interpolate between the tabulated values. It
is also acceptable to take an average of the reinforcement ratios in each orthogonal
direction when using the chart. Using the chart with an assumed reinforcement ratio
greater than the actual cap ratio will lead to a conservative estimate of the percent of
ultimate capacity.
It should be emphasized that this chart should always be used in conjunction with sound
engineering judgment. The conditions in the field can vary drastically from those seen in
the laboratory. The chart is an estimate and its accuracy can change due to the variability
in crack width data, the limited variables in the chart, and the differences between field
and laboratory conditions. Thus the chart should only be viewed as a useful guide to
making informed decisions regarding the level of distress in a diagonally cracked
inverted-T bent cap.
5.4.3 Summary and Conclusions
In this section, variables that affect the maximum width of diagonal cracks in shearcritical inverted-T deep beams were discussed. Test results indicate that the amount of
web reinforcement crossing the diagonal crack is the primary variable influencing the
crack widths examined in this study. The a/d ratio was found to have little effect on the
maximum width of the diagonal cracks. From the experimental data, a chart was
produced that correlates the maximum width of a diagonal crack to the load acting on a
member in terms of a percent of ultimate capacity. Results from thirty three full-scale
163
tests on specimens with varying ledge lengths, ledge depths, reinforcement ratios, number
of point loads, and web depths were used to develop the chart. The chart should be used
with sound engineering judgment because crack widths generally have high variability,
the chart only accounts for a limited set of variables, and differences may exist between
field and laboratory conditions.
164
CHAPTER 6
In-Service Inverted-T Bridge Bents
6.1
BACKGROUND
In this chapter, critical design characteristics and conditions of the investigated invertedT straddle bent caps is presented. The residual capacity of each bent cap is then estimated
with comparisons to the measured crack widths obtained during the field inspection.
6.2
INSPECTION REPORTS
The following sections include the reports from the field inspections of eight reinforced
concrete inverted-T straddle bent caps at four cities in the state of Texas. The location of
each straddle bent, date of inspection, key design characteristics, and crack widths are
provided. Additional figures are included to document the condition of the structures.
6.2.1 Austin (TX-290 and I-35)
Three reinforced concrete inverted-T straddle bent caps in the I-35/TX-290 interchange
on the north side of Austin were found to have significant diagonal cracks. The location
of the straddle bents are shown in Figure 6-1. Completed in November of 2001, the
bridges were just over 8 years old when they were inspected by the TxDOT Project 06416 research team. Shear cracks were observed on both faces of the bent caps near the
supports but due to the high traffic in the area not all sides could be accessed for
inspection.
N
Bent 6K
Bent 3M
Bent 28K
Figure 6-1: Location of inverted-T straddle bent caps in Austin (Mapquest)
165
A summary of the data from the bridge drawings that is of interest to this project is
provided in Table 6-1. The same parameters were collected for each bent cap inspected to
allow comparisons among inverted-T bent caps in the field and the specimens tested in
the experimental program. The individual bent caps are discussed in detail in the
following sections.
Table 6-1: Important characteristics of Austin straddle bents (6K, 3M, 28K)
Bent
6K
3M
28K
Location
(30.326788,
-97.706456)
(30.322401,
-97.702838)
(30.321920,
-97.703878)
Ledge Ledge Height / No. of

Length Cap Height Girders
a/d
0.43%
0.37%
1.7
Short
35%
0.43%
0.37%
1.4
Short
35%
0.43%
0.37%
1.4
Short
35%
The individual bent caps are identified by their GPS location (latitude, longitude). The
reinforcement ratios, v and h, were calculated at the location of the shear cracks
(between the column and the exterior girder). The shear span to depth ratio, a/d, was
taken as the distance from the center of the column to the center of the bearing of the
exterior girder divided by the effective depth of the member as calculated from the bridge
drawings. The ledge length was determined by the distance the ledge extended past the
exterior girder. If the ledge was continuous across the bottom of the beam to the column
it was considered long. If the ledge terminated a short distance from the edge of the
girder, but enough to allow the 45 load spread, then it was considered short. A short
ledge is the typical detail of most bent caps in this inspection. If the ledge stopped
immediately after the exterior girder it was considered cut-off. The ledge height was
determined by comparing the height of the ledge to the total height of the bent cap. A
ledge height of 25-38% of the total height of the cap was considered shallow and 3950% was deep. These characteristics reflect variables considered in the experimental
portion of the current project.
6.2.1.1 Austin Bent 6K
Austin Bent 6K, located along the southbound feeder of I-35, provided access to
eastbound TX-290 over an I-35 on-ramp. Two U-beams were supported by the bent cap.
Shear cracks were observed between the exterior girders and columns on both faces but
only the northwest corner, outlined in Figure 6-2, could be accessed for crack
measurements due to traffic and site restrictions.
166
NW
NE
SW
SE
Figure 6-2: Plan View of Austin Bent 6K
A picture of the investigated shear span along with a sketch showing the size and location
of observed cracks is provided in Figure 6-3. The cracks were highlighted in the
photographs to better show their position. The largest shear crack, which measured 0.016
in. at its widest point, ran from mid face of the U-beam to the end of the ledge as shown
in the sketch. Other cracks, with widths up to 0.013 in., were observed along the top and
side of the web-ledge interface. A crack was also observed to extend under the beam.
0.01
0.016
0.013
Figure 6-3 Photograph and Sketch of Northwest Corner of Austin Bent 6K
6.2.1.2 Austin Bent 3M

Bent 3M, located along the westbound feeder of TX-290 on the north side of Austin,
provided access to southbound I-35 over a TX-290 on-ramp. Three U-beams were
supported by the bent caps. Shear cracks, similar to those present on Bent 6K, were
observed on both faces of Bent 3M but only the southwest corner could be accessed as
shown in Figure 6-4.
167
NE
SE
NW
SW
Figure 6-4: Plan View of Austin Bent 3M
A photograph and a sketch of the cracks on southwest corner of Bent 3M is provided in

Figure 6-5. The largest shear crack, measuring 0.02 in., ran from above the end of the
ledge to the column near the bottom of the beam. Only hairline cracking was observed
along the web-ledge interface.
The northeast corner of Bent 3M appeared to have much larger cracks but access for the
purpose of crack measurements was not possible. Crack sizes of up to 0.03 in. were
approximated from ground observation by The University of Texas at Austin researchers.
0.005
0.02
Figure 6-5: Photograph and Sketch of Southwest corner of Austin Bent 3M
6.2.1.3 Austin Bent 28K

Bent 28K provided access to eastbound TX-290 from northbound I-35 over the TX-290
feeder road. Three U-beams were supported on each side of the inverted-T bent cap.
Shear cracks were also observed on both faces of Bent 28K. The northwest and northeast
corners, shown in Figure 6-6, were investigated
168
NE
SE
NW
SW
Figure 6-6: Plan View of Austin Bent 28K
A photograph and sketch of the location and widths of the cracks on the northwest corner
of Bent 28K are provided in Figure 6-7. Two shear cracks were observed in the web of
the bent with one extending from mid face of the U-beam to the ledge and measuring
0.016 in. The second crack was larger, 0.03 in., and ran parallel to and above the first.
This crack was recorded as measuring 0.025 in. during a previous inspection. Cracking
measuring up to 0.03 in. was also observed along the web-ledge interface and on the top
side of the ledge. This was considerably larger than the 0.016 in. that was measured
previously.
0.012
0.016
0.03
0.03 (top of ledge)

0.016
Figure 6-7 Photograph and Sketch of Northwest Corner of Austin Bent 28K
The northeast corner of the beam had a similar cracking pattern, as shown in Figure 6-8,
with the two cracks measuring 0.012 and 0.02 in. Cracks of comparable width to the
previous inspection (0.012 in.) were also observed along the web-ledge interface. Also,
torsion did not appear to be an issue because the cracks ran in the same direction on each
side of the bent cap. If the cracks were to spiral around the bent cap, angled one way on
the east side of the beam and the opposite direction on the west, then torsional effects
would need to be considered.
169
0.02
0.016
0.012
Figure 6-8: Crack size and location on the northeast corner of Austin bent 28
6.2.2 San Antonio (I-35 S)

Shear cracks were observed on one reinforced concrete inverted-T straddle bent cap on
the I-35 southbound frontage road just north of San Antonio. The location of the straddle
bent is shown in Figure 6-9. Five TxDOT Type C I-girders were supported by the bent.
The bridge was inspected by TxDOT Project 0-6416 members on July 26th, 2010.
N
SA Bent
Figure 6-9: Location of inverted-T straddle bent cap in San Antonio (Mapquest)
Table 6-2 summarizes the data collected from the bridge drawings that is of interest to
this project. The reinforcement ratio, shear span, and ledge data were unknown or
approximated from the visual inspection. No bridge plans were available to the
researchers at The University of Texas at Austin.
170
Table 6-2: Important Characteristics of San Antonio Straddle Bents
Bent
Location
SA
(29.512478,
-98.397567)
a/d
Unavailable
1.85

Long
33%
Shear cracks were observed on both faces of the bent caps near the columns but due to
site restrictions only three corners, shown in the plan view in Figure 6-10, could be
accessed by the bucket truck. Cracks associated with a moment connection between the
column and bent cap were observed at the frame corners and an impact mark was found
on the north side of the bent.
NW
NE
SW
SE
Figure 6-10: Plan View of San Antonio Bent Cap
The northwest corner of the bent had a single shear crack that measured 0.01 in. The
cracks extended from the top of the exterior I-girder to the corner of the web-ledge
interface as shown in Figure 6-11. A flexure crack measuring 0.01 in. was also observed
at the end of the bent cap (i.e. the frame corner). Evidence of spalling was present near
the beam-deck interface on the northwest corner of the bent.
Spalling
0.01
0.01
0.005
Figure 6-11: Crack Size and Location on the Northwest Corner of the San Antonio Bent
The southwest corner of the bent was nearly identical to the northwest with a shear crack
measuring 0.01 in. that extended from the end of the ledge to the top of the I-girder. A
similar width flexure crack was observed in the corner of the beam, suggesting the
presence of a moment connection, and is shown in Figure 6-12.
171
0.01
0.01
0.01
0.01
Figure 6-12: Crack Size and Location on the Southwest Corner of the San Antonio Bent
The southeast corner had a slightly different cracking pattern as shown in Figure 6-13.
The cracks were larger than the west side of the beam with a maximum width of 0.015 in
and no significant flexure cracking was observed in the corner of the bent. Due to the
lack of bridge drawings it was not possible to draw conclusions on the effects of
reinforcement as it was unknown and the shear span-to-depth ratios appeared very similar
under a visual observation.
0.01
0.015
0.015
0.01
Figure 6-13: Crack size and location on the southeast corner of the San Antonio bent
6.2.3 El Paso (I-10 E Geronimo Drive Exit)

Two inverted-T straddle bent caps located on the feeder road of the I-10 East to Gateway
Blvd East connector in El Paso were identified as having significant diagonal shear
cracks. The off-ramp is located just west of Geronimo Drive as shown in the map in
Figure 6-14.
172
Bent 4
Bent 5
Figure 6-14: Location of inverted-T straddle bent caps in El Paso (Mapquest)
The bridge was inspected by TxDOT Project 0-6416 members from the University of
Texas on August 17th, 2010. Before the site visit, relevant geometric and design
information was gathered from the original bridge drawings and recorded in Table 6-3.
Table 6-3: Important Characteristics of El Paso Straddle Bents
Bent
4
5
Location
(31.78047,
-106.41612)
(31.78055,
-106.41599)
a/d

0.57%
0.19%
1.7
Cut Off
29%
5/6
0.57%
0.19%
3.4
Cut Off
29%
The cut off ledge in these straddle bent caps raised some concern in the current project
because it does not allow the load entering the ledge to spread as much as a longer ledge
(Section 3.2.2). The resulting concentrated tensile zone could have resulted in the larger
crack widths seen in the El Paso structures. Further discussion on the characteristics and
crack widths for various ledge lengths can be found in Chapter 4.
6.2.3.1 El Paso Bent 4
A plan view of Bent 4 is shown in Figure 6-15 with the inspected regions highlighted in
red. The north end of the straddle bent was not investigated because it was built directly
into a retaining wall and thus had no measurable shear span.
173
SE
NE
NW
SW
Figure 6-15: Plan View of El Paso Bent 4
The southwest corner Bent 4 had a large shear crack that was highly visible to passing
traffic. The crack extended from mid height of the exterior girder to the lower corner of
the cap-to-column connection and measured 0.04 in. at its widest point. Minor spalling
was observed at two locations along the length of the crack. The concrete directly next to
the problematic areas was labeled 11-8-95 suggesting that it had been a concern during
a previous investigation. A picture and sketch of the cracks observed during the field
inspection are provided in Figure 6-16.
Unlike the bent cap investigated in Austin, the El Paso inverted-T caps were cast
monolithically with the columns and the reinforcement was detailed to create a moment
connection. This was confirmed by the existence of fairly large (0.02 in. wide) cracks
shown at the frame corners as depicted in Figure 6-16 and Figure 6-17. These cracks
were horizontal on the columns, vertical on the bent cap, and ran from the top outside
corner to the bottom inside corner- opposite to the direction of the shear cracks at the
beam-column connection. Similar, yet less significant cracking patterns were observed in
the San Antonio inverted-T straddle bent cap.
0.02
0.04
0.02
Spalling
Figure 6-16: Crack Size and Location on the Southwest Corner of El Paso Bent 4
The southeast corner appeared to be slightly less distressed with a maximum crack width
of 0.03 in. as shown in Figure 6-17. Overall the cracking pattern was very similar with a
large shear crack and multiple smaller moment cracks in the corner of the beam.
174
Five prestressed concrete box beams framed into the west face of the Bent 4 and were
approximately 83 ft. long. Six framed into the east side and were 61 ft. The difference in
span lengths, and thus applied load, was likely a factor for the variance in crack widths
measured on the two sides of the bent cap.
0.02
Spalling
0.03
Figure 6-17: Crack Size and Location on the Southeast Corner of El Paso Bent 4
6.2.3.2 El Paso Bent 5

The plan view of Bent 5 from the Geronimo exit off of I-10 in El Paso is given in Figure
6-18 with the inspected northeast and northwest corners highlighted in red. The south
side of the bent was not inspected due to the fact that the box beams extended all the way
to the column on the south side of the straddle bent.
SE
NE
NW
SW
Figure 6-18: Plan View of El Paso Bent 5
A photograph and sketch of the cracks on the northeast corner of Bent 5 is shown in
Figure 6-19. Due to the longer shear span, the cracks were more typical of a bent cap
exhibiting sectional shear behavior. The cracks in this inverted-T straddle bent were also
fewer in number and narrower in width than those observed on Bent 4. This was due in
part to the longer shear span (a/d of 3.4 vs 1.7 seen in Bent 4) and the fact that most of
the box beams in this bridge were closer to the south support, thus reducing the load on
the north side of the straddle bent. The widths of the two largest shear cracks on the
northeast corner of the bent measured 0.01 in.
175
0.01
Moment
Cracks
Figure 6-19: Crack Size and Location on the Northeast Corner of El Paso Bent 5
The northwest corner, shown in Figure 6-20, appeared to experience slightly more
distress. The shear crack closer to the exterior girder was twice the width of that seen on
the opposite side of the beam. Nevertheless with a maximum shear crack width of 0.02
in., Bent 5 was found to be subjected to less distress than the El Paso Bent 4.
0.01
0.02
Figure 6-20: Crack Size and Location on the Northwest Corner of El Paso Bent 5
6.2.4 Waco (TX-6 E and I-35N)

Two inverted-T straddle bent caps in the recently completed TX-6 East to I-35 North
connector in Waco (refer to Figure 6-21) were found to have developed diagonal cracks.
The 6-year-old flyover was inspected by TxDOT Project 0-6416 researchers on January
11th, 2010. Shear cracks were observed on both faces of the two bent caps in the region of
the shear span. All cracks were inspected except for those on the east side of Bent 17 due
to traffic and site restrictions.
176
Bent 19
Bent 17
Figure 6-21: Location of inverted-T Straddle Bent Caps in Waco (Mapquest)
The location, reinforcement ratio, shear span, ledge data and number of girders supported
on the ledges located at each side of the straddle bents are presented in Table 6-4.
Table 6-4: Important Characteristics of Waco Straddle Bents
Bent
17
19
Location
(31.496031,
-97.148663)
(31.496476,
-97.148489)

Length Cap Height U-Beams
a/d
0.46%
0.30%
2.5
Short
36%
0.46%
0.30%
2.5
Short
36%
6.2.4.1 Waco Bent 17

The Waco Bent 17 bent crossed an off-ramp feeding onto the I-35 North frontage road.
Only the northwest and southwest corners of the bent, highlighted in red in Figure 6-22,
were inspected due to site restrictions.
NW
NE
SW
SE
Figure 6-22: Plan View of Waco Bent 17
The diagonal cracks on Bent 17 appeared larger in the photographs due to efflorescence.
The buildup was scraped off the bent cap before crack widths were measured as shown in
177
Figure 6-23. It was important to ensure accuracy and not allow the stained concrete to
affect the results of the field inspection. The scraping and measuring process was
repeated for various cracks to obtain a good record of the condition of the straddle bent.
Figure 6-23: Scraping off Efflorescence and Measuring Diagonal Cracks
A photograph and a sketch showing the diagonal cracking on the southwest corner of
Bent 17 is provided in Figure 6-24. All of the cracks extended from the column support
to the top of the cap behind the exterior U-beam. This pattern was typical of all bent caps
inspected in this report and of the beams tested in the experimental portion of the current
project. No flexureal cracking was observed near the supports- confirming the simply
supported bearing condition in the bridge drawings. Overall the shear cracks on this bent
were relatively small with the largest measuring 0.01 in.
0.01
0.005
Figure 6-24: Crack Size and Location on the Southwest Corner of Waco Bent 17
The northwest corner of Bent 17, shown in detail in Figure 6-25, had similar diagonal
crack sizes and patterns. Cracks ranging from hairline to 0.005 and 0.01 in. extended
from the column to the top of the bent cap next to the U-beam. Overall the distress in this
inverted-T straddle bent cap was found to be moderate, despite the large amount of
staining and efflorescence.
178
0.01
0.005
0.005
Figure 6-25: Crack Size and Location on the Northwest Corner of Waco Bent 17
6.2.4.2 Waco Bent 19

All four corners of Bent 19 could be accessed and were inspected as shown in the Figure
6-26 plan view. The straddle bent cap was built on relatively level ground without any
roadway or traffic interference as at the time of the inspection the exit road had yet to be
constructed under the bent. Unlike the previous straddle bent, there was very little
staining or efflorescence overemphasizing the size and extent of the diagonal cracking as
shown in the figures in this section.
NW
NE
SW
SE
Figure 6-26: Plan View of Waco Bent 19
The cracking pattern was similar to that seen in the previous Waco bent but the cracks
themselves were slightly wider. The largest crack measured on Bent 17 was 0.01 in.
whereas a crack on the northwest corner of Bent 19 was found to have a width of 0.015
in. as shown in Figure 6-27.
0.005
0.01
0.015
0.005
Figure 6-27: Crack Size and Location on the Northwest Corner of Waco Bent 19
The southwest corner of Bent 19, shown in Figure 6-28, was subjected to a similar
cracking pattern with a series of cracks that extended from the column to the exterior U-
179
beam. However, the largest cracks on the southwest corner were found to have a smaller
width of 0.01 in.
0.01
0.005
Figure 6-28: Crack Size and Location on the Southwest Corner of Waco Bent 19
A similar pattern was observed when comparing the southeast and northeast corners. The
southeast shear span had cracks ranging from hairline to widths of 0.005 and 0.01 in. as
shown in Figure 6-29. The northeast corner, shown in Figure 6-30, had a maximum crack
width of only 0.005 in. It appeared that a difference in span lengths on either side of the
bent cap could be responsible for the varying crack widths between the north and south
sides of the inverted-T straddle bent cap.
0.005
0.01
0.005
Figure 6-29: Crack Size and Location on the Southeast Corner of Waco Bent 19
0.005
Figure 6-30: Crack Size and Location on the Northeast Corner of Waco Bent 19
Comparing the west corners of the bent cap to the east also suggests that one side has
incurred more distress. The measured cracks on the west corners in Figure 6-27 and
180
Figure 6-28 range from 0.005 to 0.015 in. The east corner, however, has only a few
cracks over 0.005 in. as shown in Figure 6-29 and Figure 6-30.
6.2.5 Findings from the Field Inspection
A visual inspection could not reveal all factors contributing to shear cracking in and
inverted-T straddle bent cap. No coring or destructive testing was performed to assess the
strength of the structures. One of the goals of the current project was to construct and test
comparable inverted-T beams in a laboratory setting. The observations made from the
results of these beams will aid inspectors in determining the significance of cracking
through the use of the diagonal crack width and ultimate capacity comparison chart
presented in Figure 5-31.
By varying different parameters found in the field, including shear reinforcement ratios,
shear span-to-depth ratios, and ledge geometry, this report aimed to determine what
factors influence the size and extent of the shear cracks. Summarized in Table 6-5 are the
variables from the different bent caps in this inspection report accompanied with the
corresponding maximum crack size. As discussed in Chapter 5, the reinforcement ratio
had the most significant effect on the width of the diagonal cracks.
Table 6-5: Inverted-T Crack Width Summary
Bent
a/d
Ledge Ledge Height / Max

Length Cap Height
Crack

0.43% 0.37% 1.7 Short
Bent 6K
0.43% 0.37% 1.4 Short
Bent 3M
0.43% 0.37% 1.4 Short
Bent 28K
San Antonio IH-35 S
Unavailable 1.85* Long
Exit 165
El Paso IH-10 /
0.57% 0.19% 1.7 Cut-Off
Geronimo Bent 4
El Paso IH-10 /
0.57% 0.19% 3.4 Cut-Off
Geronimo Bent 5
Waco IH-35 / LP340
0.46% 0.30% 2.5 Short
Bent 17
Waco IH-35 / LP340
0.46% 0.30% 2.5 Short
Bent 19
* Approximated based on field observations
181
35%
0.016 in.
35%
0.02 in.
35%
0.03 in.
33%*
0.015 in.
29%
0.04 in.
29%
0.02 in.
36%
0.01 in.
36%
0.015 in.
A few observations could be made on the geometry of the inverted-T straddle bent caps
and their respective maximum crack widths. The largest crack width was measured on
Bent 4 in El Paso. Two of the factors contributing to the size of this crack, the ledge
height and ledge length, are recorded in Table 6-5. Compared to every other bent cap in
this investigation, El Paso Bent 4 had the shortest and shallowest ledge along with the
smallest horizontal web reinforcement ratio. Bent 5 had an identical ledge geometry but a
larger shear span-to-depth ratio (a/d) and was arranged so that most of the load from the
box beams was transferred directly to the opposite support (refer to Figure 6-19).
Nevertheless the maximum crack size measured on Bent 5 was also larger than many of
those observed in this survey.
The San Antonio bent had the longest ledge and some of the smallest shear cracks, 0.015
in., of any cap inspected in this report. Similarly, the Waco straddle bent experienced
some of the smallest shear cracks. The largest shear cracks measured on the Waco Bent
17 and Bent 19 were 0.010 and 0.015 in., respectively. The reinforcement ratios and
shear span-to-depth ratios were comparable to the other beams investigated in this study.
6.3
SERVICEABILITY BEHAVIOR
6.3.1 Diagonal Cracking under Service Loads

One of the tasks of the current project was to develop a means to prevent or limit
diagonal cracking for inverted-T beams subjected to service level loads. A previously
derived equation from the deep beam study. TxDOT Project 0-5253, was evaluated with
diagonal cracking loads from the inverted-T experimental program. It was concluded that
the existing recommendation to estimate the diagonal cracking load based on a simple
equation was valid for inverted-T beams as well as rectangular bent caps.
The diagonal cracking load was found to be a function of the shear area, b wd, the tensile
strength of the concrete, , and the a/d ratio of the member as shown in Equation 6-1.
but not greater than 5
(6.1)
nor less than 2
where a = shear span (in.)

d = effective depth of the member (in.)
fc = compressive strength of concrete (psi)
bw = web width of the member (in.)
In order to evaluate the performance of the in-service beams inspected as part of the
current project, the service load of the inspected bents was compared to the calculated
182
diagonal cracking load as shown in Figure 6-31. CAP18, A TxDOT Bent Cap Analysis
Program, was used to determine the maximum bending and shear force envelopes on the
simply supported bridge bent caps. Loads assigned to the bent caps were calculated in
accordance with the bridge drawings and AASHTO LRFD Bridge Design Specifications,
4th Ed. (2007) and the 2008 Interim Revisions as prescribed by TxDOT LRFD Bridge
Design Manual (July 2008). A design example provided by the TxDOT bridge division
aided in the determination of loads and proper use of the analysis program.
Austin 3M
Waco 19
10
Austin 6K
El Paso 4
Austin 28K
El Paso 5
Waco 17
High Risk of
Diagonal Cracking
6
4
2
0
Low Risk of
Diagonal Cracking
0
0.5
1.5
2.5
3.5

Figure 6-31: Assessment of in-service bent caps with the proposed equation for diagonal
cracking estimation
In this figure, any point below the diagonal cracking line would indicate a bent cap with a
low risk of diagonal cracking- i.e., the service level shear is below the cracking load. Any
point above this line would represent a bent cap that is very likely to crack when
subjected to its corresponding service level load. According to the cracking equation,
increasing the section size or the concrete compressive strength would decrease the risk
of diagonal cracking.
Based on the proposed equation, the service level shear was greater than the estimated
diagonal cracking load for each in-service bent cap inspected as part of the current
project. Therefore the inverted-T bents had a high risk of diagonal cracking when
subjected to service level loads. Overloads and other unexpected conditions would only
increase this chance.
For future designs it is recommended that the bent caps be sized according to Equation 61 to limit the risk of diagonal cracking. If those options are not available due to geometry
183
restrictions the designer is encouraged to increase the compressive strength of the

concrete and provide additional web reinforcement to control crack widths.
6.3.2 Maximum Diagonal Crack Widths
Another task for the current project was to develop a chart that could be used to correlate
the maximum diagonal crack width on an inverted-T beam to the load on the member,
quantified as a percent of the ultimate capacity. A similar chart was developed in TxDOT
Project 0-5253 for use with rectangular deep beams but was not approved for use with
tension-chord loaded members. The inverted-T chard and its development are presented
in Section 5.4.
The diagonal crack width to capacity chart provided in Figure 5-31 was applied to the inservice members inspected in the current project. Based on the measured maximum crack
width and the web reinforcement ratio, the capacities in the following table were
estimated. It should be noted that conditions in the field can vary from those seen in the
laboratory and chart results should only be viewed as a useful guide to determining the
level of distress in a diagonally cracked inverted-T bent cap.
Table 6-6: Estimated percent of ultimate capacity for in-service bent caps
Maximum Ultimate
Crack Width Capacity
Austin IH-35 / Tx-290 Bent 6K
0.43% 0.37%
0.016 in. 45 (15)%
Austin IH-35 / Tx-290 Bent 3M
0.43% 0.37%
0.02 in.
55 (15)%
Austin IH-35 / Tx-290 Bent 28K 0.43% 0.37%
0.03 in.
70 (15)%
San Antonio IH-35 S Exit 165
Unavailable
0.015 in.
40-50%*
El Paso IH-10 / Geronimo Bent 4 0.57% 0.19%
0.04 in.
85 (15)%
El Paso IH-10 / Geronimo Bent 5 0.57% 0.19%
0.02 in.
55 (15)%
Waco IH-35 / LP340 Bent 17
0.46% 0.30%
0.01 in.
35 (15)%
Waco IH-35 / LP340 Bent 19
0.46% 0.30%
0.015 in. 45 (15)%
*Unknown reinforcement ratio, the range represents v=h = 0.3% to v=h = 0.6%
Bent
The chart presented in Chapter 5 provides capacity estimations based on select crack
widths and reinforcement ratios of 0.3% or 0.6% in each orthogonal direction. As shown
in Table 6-6, inverted-T bent caps typically fall within this range so the capacity values
need to be interpolated. For the in-service inverted-T beams the vertical and horizontal
reinforcement ratios were averaged and then the percent of ultimate capacity was
calculated by linear interpolating values from the chart based on the maximum crack
width. For example, the average reinforcement ratio of Austin Bent 6k was 0.4% which is
the difference between the 0.3% and 0.6% values. The maximum crack width of 0.016
in. is located roughly halfway in-between the provided values of 0.01 and 0.02 in.
184
Therefore the percent of ultimate capacity should be about the difference between 40%
and 53% resulting in an estimated load of 45% of the ultimate capacity of the member.
For the San Antonio bent cap, no information was available to determine the web
reinforcement ratio. Therefore the provided range corresponds to the values obtained for
0.3% to 0.6% in each orthogonal direction. The 15% scatter should also be considered.
6.4
SUMMARY
Field inspections revealed inverted-T bent caps subjected to varying levels of distress
across the state of Texas. Pertinent information including ledge geometry, a/d ratio,
reinforcement ratio, and number of girders was collected for each structure. Photographs
were taken and crack width measured to document the condition of each bent.
The in-service structures were then compared to the diagonal cracking load estimate
discussed in Section 5.3. It was found that based on the section size, a/d ratio, and
concrete strength, each of the inverted-T bent caps inspected as part of the current study
had a high risk of cracking under service level loads.
Using the data gathered in the field inspections, the load on the bent caps, quantified as
the percent of ultimate capacity, was determined with the diagonal crack width chart
provided in Figure 5-31. It was then concluded that several of the existing structures had
been subjected to loads up to 70-85% of their ultimate capacity.
Based on the analysis of results from the inverted-T experimental program and the
conditions of the in-service bent caps, it is recommended to design such structures in the
future based on the recommendations of this project that are summarized in Chapter 7.
185
186
CHAPTER 7
Design Recommendations
7.1
INTRODUCTION
The design recommendations presented in this chapter reflect the conclusions drawn as a
result of the extensive experimental program on inverted-T beams and assessing the
accuracy and applicability of currently available STM procedures that were calibrated
with compression-chord loaded beam test results, as implemented in Chapter 2 of this
report.
7.2
PROPOSED CHANGES TO TXDOT DESIGN SPECIFICATIONS

7.2.1
General Design Recommendations
The following recommendations for designing inverted-T beams resulted from the
experimental work performed in this project. Examples of inverted-T design using strutand-tie modeling are presented in Williams et al. (2011).
1- Use STM design procedures for designing inverted-T bent caps. Almost all
inverted-T beams should be classified as deep beams because they are mostly
made up of disturbed regions. These areas of nonlinear strain are the result of
point loads (girders or column supports) and geometric discontinuities (ledges) as
illustrated in Figure 7-1. It is therefore inappropriate to use the sectional shear
approach to design inverted-T beams and thus strut-and-tie models should be
used.
2- Determine the following tie widths when implementing the STM provisions for
inverted-T beams. The load spread under the loading points defines the hanger tie
width and should be taken as the width of the bearing pad plus two times the
effective depth of the ledge. The width of the horizontal ledge tie at the top of the
ledge is smaller and should be taken as the bearing pad width plus five times the
distance from the center of the pad to the face of the web. This is consistent with
current AASHTO recommendations for ledge design. These recommendations are
discussed in Chapter 2 and highlighted in examples in TxDOT Project 5-5253.
3- Avoid using cut-off ledges. Cut-off ledges were shown to result in lower web
shear strengths than longer ledges. Cut-off ledges were also found to reduce the
load at first diagonal cracking in the web. Moreover, cut-off ledges increased the
risk of ledge failures.
4- Size member to limit web shear stresses at service loads. Based on the shear span
to depth ratio, the concrete compressive strength, and the web width, estimate the
187
diagonal cracking load using Equation 5-1. This provision is essential to limit
diagonal cracking at service load levels.
5- Provide a minimum of 0.3% web reinforcement ratio in each direction. This
minimum ratio should be used to limit diagonal crack widths at service loads. As
demonstrated by experimental results, should a beam sustain diagonal cracking at
service load levels, providing the recommended minimum reinforcement ratio
will ensure narrow crack widths. Furthermore, this minimum ratio is needed to
fully mobilize strut and node strengths.
Figure 7-1: D-regions in inverted-T bent caps
7.2.2 Strut-and-Tie Model Design Procedure

The following outlines a series of steps typically followed when designing a deep
structural component using the STM procedure developed as part of Project 0-5253. The
procedure can be adapted to the particular design scenarios as necessary and it should be
noted that the steps are sometimes performed simultaneously. The STM procedure is
presented in a flow-chart format in Figure 7-2.
1. Separate B- and D-regions Determine which regions of the structural component are
expected to exhibit deep beam behavior or if the entire component should be designed
using STM. For inverted-T bent caps, D-regions typically account for most if not all of
the beam.
2. Define load case Calculate the factored loads acting on the structural component, and
if necessary, make simplifying assumptions to develop a load case that can be applied to
a reasonable STM.
3. Analyze structural component Solve for the structural components support reactions
assuming linear elastic behavior.
188
4. Size structural component using the shear serviceability check Determine the initial
geometry of the structural component by using the shear serviceability check presented in
Section 5.3 to prevent or limit diagonal cracking at service loads.
5. Develop strut-and-tie model Position struts and ties to represent the actual flow of
forces within the structural component, and determine the forces in the struts and ties.
6. Proportion ties Specify the reinforcement needed to carry the force in each tie.
7. Perform nodal strength checks Define the geometries of the critical nodes, and ensure
the strength of each face is adequate to resist the applied forces determined from the
analysis of the STM.
8. Proportion crack control reinforcement Specify the required crack control
reinforcement to restrain diagonal cracks formed by the transverse tensile stresses of
bottle-shaped struts. As discussed in Section 4.5, 0.3% reinforcement orthogonal in both
directions is the minimum recommended for inverted-T bent caps.
9. Provide necessary anchorage for ties Ensure reinforcement is properly anchored at
the nodal regions.
189
Separate B- and D-Regions
Define Load Case
Analyze Structural Component
Size Structural Component
Develop Strut-and-Tie Model
Proportion Ties
Perform Nodal Strength Checks
Proportion Crack Control Reinforcement
Provide Necessary Anchorage for Ties

Figure 7-2: Strut-and-tie model design procedure (adapted from TxDOT Report 5-5253)
For the sake of brevity, additional details for each step in the strut-and-tie model design
procedure can be found in TxDOT Report 5-5253. The proposed revisions to the
AASHTO LRFD Bridge Design Specifications for strut-and-tie modeling are also
presented before the design examples for ease of reference in the report.
7.2.3 TxDOT Report 5-5253 Design Examples
Detailed design examples for inverted-T straddle bent caps are provided in TxDOT
Report 5-5253. The design examples explicitly note the difference between the design of
an inverted-T and a rectangular beam. These differences were highlighted in Chapter 2 of
190
this report. As the girder loads are applied to the ledge of inverted-T beams, vertical ties
(stirrups) are necessary to transfer the loads through the web to the compression chord.
This results in tension across the width of the beam that must be resisted by transverse
ledge reinforcement. In order to account for the flow of forces through the cross section
and along the beam length, a three-dimensional STM must be developed for the design of
an inverted-T bent cap.
A moment frame and a simply supported bent cap were investigated in Project 5-5253 to
compare the influence of boundary condition assumptions. The struts and nodes within
the bent cap are illustrated in Figure 7-3. For the moment frame example, the bent was
modeled to allow the forces to turn around the frame corners. This required the use of a
curved-bar node at the top corners of the bent, which occurs at a frame corner where a
diagonal strut is equilibrated by two ties that represent curved continuous reinforcing bars
(Klien, 2008). This type of node is not unique to inverted-T beams and as moment frame
specimens were not tested, it was not investigated in the current experimental program.
Figure 7-3: Illustration of struts and nodes within the moment frame inverted-T bent cap
(Williams 2011).
For the simply supported design example, the basic principles were the same but the
geometry of the STM, the member forces, and the resulting reinforcement layout were
different from those of the moment frame structure. The reinforcement details for the
moment frame design allowed for the flow of forces around the frame corners and
permitted moment to be transferred between the columns and the bent cap. For the simply
supported case, only vertical reactions were assumed to be transferred between the cap
and the columns. Due to the fixity of the structure, the moment at the midspan of the
moment frame bent cap was significantly smaller than the simply supported bent cap
which required more reinforcement in the bottom chord. Additional stirrups were also
required for the simply supported member due to the reduced truss depth. These design
differences are expected between simply supported and moment frame structures.
The design examples were modeled after a moment frame inverted-T bent cap that was
originally designed using TxDOT sectional procedure. Although not inspected as part of
the current project, the in-service bent cap was reported to have significant diagonal
191
cracking in the shear spans as shown in Figure 7-4. The shear serviceability check
developed by TxDOT Project 0-5253 and evaluated in the current project indicated that
diagonal cracking was likely for the structure. An increase in cross-sectional dimensions
and/or concrete strength would have been necessary to limit cracking. Flexural cracking
was also observed at the outside of the frame corners, which would indicate that the bend
radius of the longitudinal bars was too small. In the moment frame example, it was
discovered that a larger bend radius was required for the longitudinal bars in the corners
of the frame, once again verifying the applicability of the recommended STM design
procedure.
Figure 7-4: US-59/ N W. Little York Bent #4, Houston Texas
192
CHAPTER 8
Summary and Conclusions
8.1
SUMMARY
Diagonal cracking in the shear span of inverted-T bent caps prompted an investigation
into the design of such structures from a strength and serviceability standpoint. A
previous research project, TxDOT Project 0-5253, investigated rectangular deep beams
loaded on the compression chord. In this project recommendations were made for the
design of such structures through strut-and-tie modeling provisions. Serviceability
recommendations were also made to limit diagonal cracking and relate maximum crack
width to deep beam residual capacity. These provisions proved useful to TxDOT
engineers but were not approved for tension-chord loaded structures such as inverted-T
bent caps.
Due to scarcity of experimental investigations on inverted-T beams, a comprehensive
experimental program was undertaken to examine the behavior of such structural
elements and assess the accuracy and validity of implementing the TxDOT Project 05253 STM design provisions. In order to accomplish the aforementioned goal, the scope
of the project was divided into the following tasks:
1. Conduct a comprehensive literature review to expose the current state of knowledge
on inverted-T beams.
2. Conduct a detailed assessment of the condition of distressed in-service bent caps
3. Conduct experimental investigations to uncover the main factors affecting the web
diagonal cracking behavior of inverted-T beams:
i.
Determine the influence that the length of the ledge has on the strength and
serviceability behavior of an inverted-T beam (Section 4.3).
ii.
Determine the influence that the depth of the ledge has on the strength and
iii.
Determine the influence that the web reinforcement ratio has on the strength
iv.
Determine the influence that the number of point loads has on the strength
v.
Determine the influence that the depth of the web has on the strength and
vi.
Determine the effect of tension-chord loading on the strength and

serviceability of deep beams (Section 4.8).
193
4. Make a recommendation on the application of TxDOT Project 0-5253 STM

provisions for the design of inverted-T bent caps (Section 5.2).
5. Make a recommendation on the feasibility of limiting diagonal cracking under
service loads (Section 5.3).
6. Make a recommendation for relating the maximum diagonal crack width of an
inverted-T beam to its residual capacity (Section 5.4).
An exhaustive database was assembled to evaluate all relevant results on inverted-T

beams test from the literature and the current experimental program. It was discovered
that research in this field is scarce. Of the 130 beam tests compiled, most were of beams
not comparable to in-service structures. Complicated support conditions, unrealistic
geometry, impractical reinforcement layouts, and unrealistically small dimensions
precluded the use of all inverted-T specimens in the literature.
An extensive experimental program was therefore undertaken to achieve the project
objectives. Thirty three full-scale inverted-T specimens were tested as part of this project.
The cross-sectional dimensions of the specimens included: 21 in. web widths and 42 in.
or 75 in. web heights. Ledge width for all specimens was 10.5 in. on each side. Length
and depth of the ledges were varied as per the projects tasks. During the tests,
measurements of the applied load, deflection along the beam, the strain at various
locations, and diagonal crack widths were recorded. With the data from these tests the
eight primary objectives of the current project were addressed.
8.2
CONCLUSIONS
The conclusions of the current study, based on data from the experimental program, are
presented in this section.
8.2.1 Effects of Ledge Length
The effects of the length of the ledge on the shear strength and serviceability of invertedT beams were investigated experimentally. Specimens in the experimental program were
constructed with one of three different ledge lengths: cut-off, where the ledge was
terminated directly after the bearing plate, short, where the ledge was extended a distance
equal to the ledge depth from the edge of the bearing plate, and long ledges that extended
all of the way to the support. Twenty tests were conducted on eight groups of directly
comparable specimens in which every variable was kept constant except for the length of
the ledge. The following could be concluded from test results:
Increasing the length of the ledge increases the web shear strength of inverted-T
beams and delays the appearance of first diagonal cracking in the web. Based on the
experimental results specimens with cut-off ledges were found to have the lowest
normalized shear strength and were more likely to crack at lower loads. It is
194
recommended to extend the ledge at least a distance equal to its depth past the bearing
plate of the exterior girder.
The length of the ledge has no appreciable effect on the width of diagonal cracks..
Once the inverted-T specimen cracked, no further trend was observed between the length
of the ledge and the diagonal crack width as crack width expansion due to increasing
load was not a function of ledge length.
8.2.2 Effects of Ledge Depth

The effects of the depth of the ledge on the strength and serviceability of inverted-T beam
webs were investigated experimentally. Specimens in the experimental program were
constructed with either deep ledges or shallow ledges; with depths equal to one half or
one third of the height of the beam. Twenty tests were conducted on ten groups of
directly comparable specimens. The following could be concluded from test results:
Increasing the depth of the ledge had no appreciable effect on the strength or
serviceability behavior of the webs of inverted-T specimens. No trends were observed
between the depth of the ledge and the normalized shear strength of the webs. Although
deeper ledges did increase the diagonal cracking load slightly, they had no effect on the
progression of crack widths with applied load.
8.2.3 Effects of Web Reinforcement Ratio

The effects of the web reinforcement ratio on the strength and serviceability behavior of
reinforced concrete inverted-T beams were investigated experimentally. The minimum
reinforcement ratio recommended by TxDOT Project 0-5253 was also evaluated to
ensure adequate performance. Several distributions of web reinforcement were
investigated on 42-in. deep beams. The majority of test specimens had either 0.3% or
0.6% reinforcement in each orthogonal direction. Two specimens were constructed with
0.6% in the vertical direction (shear stirrups) and 0.3% in the horizontal (skin
reinforcement). Fourteen tests were conducted on six groups of directly comparable
specimens. The following could be concluded from test results:
Increasing the horizontal and vertical web reinforcement ratio from 0.3% to 0.6%
increased the shear strength for specimens tested at a/d ratios of 1.85 and 2.50. The
specimens tested at an a/d ratio of 1.85 failed in a manner consistent with a single-panel,
direct-strut mechanism. While this mechanism depends mostly on the strength of the
concrete it appeared that additional reinforcement served to reinforce the strut and the
corresponding strut-to-node interface. The specimens tested at an a/d ratio of 2.50
generally failed in a manner that was consistent with a sectional-shear model, or a
multiple panel STM. At this longer shear span, increasing the vertical reinforcement was
expected to increase the strength of the member as the vertical web tie was the weakest
element.
195
Increasing the horizontal and vertical web reinforcement ratio from 0.3% to 0.6%
had little influence on the shear at first cracking of the webs but did reduce diagonal
crack widths at larger levels of load.
To adequately restrain the maximum diagonal crack widths in the webs at service
loads, a minimum of 0.3% web reinforcement ratio should be provided and spaced
evenly in each orthogonal direction within the effective strut area. The maximum
diagonal crack width of specimens with 0.3% reinforcement in each direction were found
to satisfy the 0.016 in. limit at estimated service level loads. The 0.3% web reinforcement
ratio is consistent with the recommendations in TxDOT Project 0-5253 and with the
current AASHTO LRFD provision (Article 5.6.3.6, 2012).
8.2.4 Effects of Multiple Loading Points

The effects of multiple point loads on the strength and serviceability behavior of
inverted-T beams were investigated experimentally. Specimens were loaded on the ledge
at either one or three points. Twelve tests were conducted on six pairs of directly
comparable specimens in which every parameter was kept constant except the number of
load points. The following could be concluded from test results:
The number of point loads did not influence the strength or serviceability behavior
of the inverted-T specimens. No appreciable trends were observed when the shear
strengths were compared between specimens loaded at one or three points and the
diagonal cracking load was essentially unchanged. Multiple loading points did not
significantly influence the progression of diagonal crack widths with applied load. The
results therefore lend support to extending experimental findings from single-point
loaded specimens to multi-point loaded bent caps in the field.
8.2.5 Effects of Web Depth

The effects of web depth on the strength and serviceability performance of reinforced
concrete inverted-T beams were investigated experimentally. Tests were conducted at a/d
ratios of 1.85 and 2.50 on specimens with 42 in. and 75 in. deep webs and 0.3% web
reinforcement in each direction. A single load was applied to short, shallow ledges for all
directly comparable specimens. The following could be concluded from test results:
Within the full-scale web depth range tested, web depth did not influence the
strength or serviceability behavior of the inverted-T specimens. No appreciable
trends were observed when the calculated and experimental shear strengths were
compared between specimens of different depths and the diagonal cracking load was
essentially unchanged. Web depth did not significantly influence the progression of
diagonal crack widths with applied load.
196
8.2.6 Effects of Tension Chord Loading

The effects of tension chord loading on the strength and serviceability performance of
reinforced concrete deep beams were investigated by comparing inverted-T beams from
the current study with rectangular bent caps from TxDOT Project 0-5253.
No strength effects were observed due to tension-chord loading, but the tension field
does affect crack width. The inverted-T beams were observed to crack at higher levels
of load and sustain smaller crack widths throughout loading. Thus an inverted-T beam
with the same crack width as a compression-chord loaded beam would be closer to its
ultimate capacity.
8.2.7 Proposed STM Design Provisions

The STM design provisions developed by TxDOT Project 0-5253 were evaluated for the
design of inverted-T bent caps. The following recommendations were issued for the
design of inverted-T beams:
TxDOT Project 0-5253 STM provisions are recommended for the design of
inverted-T bent caps. A comparison between the ultimate shear capacity obtained from
the test result and the nominal shear capacity from the calculations revealed conservative
strength estimates for every specimen. Furthermore, the strut-and-tie models accurately
predicted the ultimate failure mode for twenty eight of the thirty three specimens. For the
five that did not fail in shear, the calculated shear capacity was exceeded and the actual
failure mode was the second weakest element in the model. Within these provisions, a
minimum web reinforcement ratio is given as 0.3% in each orthogonal direction. That
minimum worked well in inverted-T beams both in terms of developing strut/node
strengths and in terms of limiting diagonal crack widths at service levels. It is
recommended to provide that minimum ratio in inverted-T beams.
Recommendations were given in Chapter 2 for implementing the STM provisions

for inverted-T beams. The geometry of inverted-T beams requires the use of a threedimensional STM model or two complimentary two-dimensional models.
Recommendations were given to aid in developing these models. The width over which
the load is assumed spreads under the loading points defines the hanger tie width and was
defined. The width of the tie at the top of the ledge was also defined based on current
AASHTO design procedures for ledges.
Cut-off ledges are not recommended. Cut-off ledges were shown to result in lower web
shear strengths than longer ledges. Cut-off ledges were also found to reduce the load at
first diagonal cracking in the web. Moreover, cut-off ledges increased the risk of ledge
failures.
197
Application of the STM provision to the design of one of the inverted-T beam
specimens is given in Appendix A. Two detailed design examples of inverted-T
beams are given in the final report of Project 5-5253 (Williams et al. (2011)).
8.2.8 Limiting Diagonal Cracking under Service Loads
Along with strength design provisions, a service-load shear check to limit diagonal
cracking was given.
The simple and reasonably conservative equation to estimate the diagonal cracking
load of deep beams is recommended for use with inverted-T beams. As in the deep
beam project, the diagonal cracking load of inverted-T beams was found to be mainly a
function of the shear area, the square root of the compressive strength of concrete, and
the a/d ratio. With this equation the service level shear in the member can be sized to
limit diagonal cracking. If the service level shear (full dead load plus live load) exceeds
the estimated diagonal cracking load, the design of the section can be modified to
increase section size and/or additional web reinforcement can be provided to control
diagonal crack widths. This check encourages the designer to consider the likelihood of
diagonal cracking in service.
Longitudinal post tensioning could be investigated for structures that do not pass
service checks. It has been shown in rectangular beams that post-tensioning reduces
crack widths and may increase shear strength. It is likely it would have a similar result for
inverted-T beams, however, this has not been directly investigated.
8.2.9 Correlation of Maximum Diagonal Crack Width to Web-Shear Strength

Diagonal crack widths were measured and compared to applied load to help field
engineers evaluate the residual capacity of a diagonally-cracked inverted-T bent caps.
8.3
A simple chart was developed to correlate the maximum diagonal crack width in a
deep beam to the load acting on the member, quantified as a percent of its ultimate
capacity. As the other variables investigated in the experimental program had limited
effects on the width of the diagonal cracks, the chart is only a function of the amount of
web reinforcement in the member. It is similar to the one produced for the compressionchord loaded deep beams, but for a given percentage of resulting residual strength,
cracks in inverted-T beams were found to be smaller than those in rectangular beams.
This chart is a simple means to make an informed decision regarding the amount of
distress in a diagonally-cracked inverted-T bent cap when a more sophisticated means of
evaluation is unavailable.
CONCLUDING REMARKS
In this research project, the behavior of inverted-T webs was studied through a
comprehensive experimental program. Thirty three tests were conducted on large-scale
198
specimens, some of which are amongst the largest deep beams, and certainly inverted-T
beams, ever tested in the history of shear research. From the analysis of this data, the
effect of the following parameters on the strength and serviceability behavior of invertedT beam webs was determined: the length of the ledge, the depth of the ledge, the quantity
of web reinforcement, the number of point loads, the depth of the member, and the shear
span-to-depth ratio.
With the results obtained from examining the effect of these parameters, the strut-and-tie
provisions proposed by TxDOT Project 0-5253 were assessed for applicability to
inverted-T beams. The design provisions were found to yield accurate and reasonably
conservative results for tension-chord loaded beams. The recommended minimum web
reinforcement ratio adequately restrained the width of diagonal cracks and the serviceload check for limiting the formation of diagonal cracks under service loads was suitable
for inverted-T beams as well as the deep beams for which they were developed. The
knowledge obtained from this study was also used to improve the evaluation of bent caps
in the field as diagonal crack widths of inverted-T beams were found to correspond to a
greater percentage of ultimate capacity than their compression-chord loaded counterpart.
This report provides a comprehensive summary of the parameters affecting the behavior
of inverted-T beams with which the condition of in-service bent caps can be properly
assessed.
199
200
References
AASHTO LRFD, Bridge Design Specifications, American Association of State Highway and
Transportation Officials, Washington, D.C., 2012.
ACI Committee 224, Control of Cracking in Concrete Structures (ACI 224R-01), ACI Manual
of Concrete Practice, American Concrete Institute, Farmington Hills, MI, 2008.
ACI Committee 318-11, Building Code Requirements for Reinforced Concrete (ACI 318-11),
American Concrete Institute, Farmington Hills, MI, 2011.
ACI-ASCE Committee 326, Shear and Diagonal Tension, American Concrete Institute,
Farmington Hills, MI, 1962.
ACI-ASCE Joint Committee 426-73, The Shear Strength of Reinforced Concrete Members (ACIASCE 426-72), American Concrete Institute, Detroit, MI, 1973.
ACI-ASCE Joint Committee 445, Recent Approaches to Shear Design of Structural Concrete
(ACI 445R-99), American Concrete Institute, Farmington Hills, MI, 1999.
ASTM A 370 08a, Standard Test Methods and Definitions for Mechanical Testing of Steel
Products, American Society for Testing and Materials, West Conshohocken, PA, May
2008.
ASTM A 615/A 615M 08, Standard Specification for Deformed and Plain Carbon-Steel Bars
for Concrete Reinforcement, American Society for Testing and Materials, West
Conshohocken, PA, March 2008.
ASTM C 143/C 143M 08, Standard Test Method for the Slump of Hydraulic-Cement Concrete,
American Society for Testing and Materials, West Conshohocken, PA, March 2008.
ASTM C 31/C 31M 08a, Standard Practice for Making and Curing Concrete Test Specimens
in the Field, American Society for Testing and Materials, West Conshohocken, PA, April
2008.
ASTM C 39/C 39M 05, Standard Test Method for Compressive Strength of Cylindrical
Concrete Specimens, American Society for Testing and Materials, West Conshohocken,
PA, November 2005.
Bergmeister, K.; Breen, J. E.; Jirsa, J. O.; and Kreger, M. E., Detailing for Structural Concrete,
Report No. 1127-3F, Center for Transportation Research, University of Texas at Austin,
Austin, Texas, May 1993.
Birrcher, D., Tuchscherer, R., Huizinga, M., Bayrak, O., Wood, S., & Jirsa, J. Strength and
Serviceability Design of Reinforced Concrete Deep Beams. Austin, TX: Center for
Transportation Research, The University of Texas at Austin. 2009.
201
Bracci, J. M., Keating, P. B., and Hueste, M. B. D., Cracking in RC Bent Caps, Research Report
1851-1, Texas Transportation Institute, The Texas A&M University System, College
Station, Texas, Oct. 2000, 257 pp.
Brown, M. D.; Sankovich, C. L.; Bayrak, O.; Jirsa, J. O.; Breen, J. E.; and Wood, S. L., Design
for Shear in Reinforced Concrete Using Strut-and-Tie Models, Report No. 0-4371-2,
Center for Transportation Research, University of Texas at Austin, Austin, Texas, Apr.
2006.
Cussens, A. R., & Besser, I. I. Shear strength of reinforced concrete wall-beams under combined
top and bottom loads. The Structural Engineer, 63B(3), Steptember 1985. 50-56 pp.
Fereig, S. M., & Smith, K. N. Indirect Loading on Beams with Short Shear Spans. ACI Journal,
74(5), May 1, 1977, pp. 220-222.
Ferguson, P. M. Some Implications of Recent Diagonal Tension Tests. Journal of the American
Concrete Institute, 53(8), 1956, pp. 157-172.
Fernndez-Gmez, E., Larson, N., Garber, D., Ghannoum, W., & Bayrak, O. Strength and
Serviceability of Reinforced Concrete Inverted-T Straddle Bent Caps. PCI/NBC
Proceedings. 2011
FIB, Structural Concrete, Textbook on Behaviour, Design, and Performance, Volume 3,
International Federation for Structural Concrete, Lausanne, Switzerland, 1999, 269 pp.
Furlong, R. W., & Mirza, S. A. 153-1F - Strength and Serviceability of Inverted T-Beam Bent
Caps Subject to Combined Flexure, Shear, and Torsion. Austin: Center for Highway
Research, University of Texas at Austin.1974.
Furlong, R. W., Ferguson, P. M., & Ma, J. S. 113-4 - Shear and anchorage study of
reinforcement in inverted-T beam bent cap girders. Austin TX: Center for highway
research at The University of Texas at Austin. 1971.
Galal, K., & Sekar, M. Rehabilitation of RC inverted-T girders using anchored CFRP sheets. (S.
Direct, Ed.) Composites: Part B - Engineering, 39(4), June 1, 2007, pp. 604-617.
Garber, D. B. Shear Cracking in Inverted-T Straddle Bents. Masters Thesis, University of Texas
at Austin. August 2011.
Graf, O., Brenner, E., & Bay, H. Versuche mit einem wandartigen Trager aus Stahlbeton.
Deutscher Ausschuss fur Stahlbeton, 99, 1943, pp. 41-54.
Huizinga, M. R., Strength and Serviceability Performance of Large-Scale Deep Beams: Effect of
Transverse Reinforcement, Masters Thesis, University of Texas at Austin, August 2007,
232 pp.
Leonhardt, F., & Walther, R. (1966). Wandartige Trger. Deutscher Ausschuss furStahlbeton,
178.
202
Ma, J. S. PhD Dissertation: Behavior of reinforced concrete inverted T-beams. Austin, TX:
University of Texas at Austin. 1971.
MacGregor, J. G. and Wight, J. K., Reinforced Concrete, Mechanics and Design, 4th Edition,
Pearson Prentice Hall, New Jersey, 2005, 1132 pp.
MacGregor, J. G., Derivation of Strut and Tie Models for the 2002 ACI Code, ACI SP-208
Examples for the Design of Structural Concrete with Strut-and-Tie Models, American
Concrete Institute, Michigan, 2002, 242 pp.
Mitchell, D., and Collins, M.P., Diagonal Compression Field Theory A Rational Model for
Structural Concrete in Pure Torsion, ACI Journal, Vol. 71, No. 8, August 1974, pp. 396408.
Ramirez, J.A., and Breen, J.E., Proposed Design Procedures for Shear and Torsion in
Reinforced and Prestressed Concrete, Report No. 248-4F, Center for Transportation
Research, University of Texas at Austin, Austin, Texas, 1983.
Schlaich, J., Schfer, K. and Jennewein, M., Toward a Consistent Design of Structural
Concrete, PCI Journal, Vol. 32, No. 3, May-June 1987, pp.74-150.
Schtt, H. ber das Tragvermgen wandartiger Stahlbetontrger. Beton und Stahlbetonbau, 10,
October 1956, pp. 220-224.
Smith, K. N., & Fereig, S. M. Effect Of Loading And Supporting Condidtions On The Shear
Strength Of Deep Beams. ACI, SP 42, January 1, 1974, pp. 441-460.
Tan, K. H., Kong, F. K., & Weng, L. W. (1997, June 3). High strength concrete deep beams
subjected to combined top-and bottom-loading. The Structural Engineer, 75(11), 191197.
Taylor, R. Some shear tests on reinforced concrete beams without shear reinforcement.
Magazine of Concrete Research, 12(36), November 1960, pp. 145-154.
TxDOT. Bridge Design Manual - LRFD. 2011
TxDOT, Texas Department of Transportation Standard Bridge Drawings, 2008.
Wight, J.K., and Parra-Montesinos, G., Use of Strut-and-Tie Model for Deep Beam Design as
per ACI 318 Code, ACI Concrete International, Vol. 25, No. 5, May 2003, pp. 63-70
Williams, C. S. Strut-and-Tie Model Design Examples for Bridges. Masters Thesis, University
of Texas at Austin. 2011
Young, B. S., Bracci, J. M., Keating, P. B., and Hueste, M. B., Cracking in Reinforced Concrete
Bent Caps, ACI Structural Journal, Vol. 99, No. 4, July-August 2002, pp. 488-498.
203
Zhang, N. and Tan, K. H., Size effect in RC deep beams: Experimental investigation and STM
verification, Engineering Structures, Vol. 29, 2007, pp. 3241-3254.
Zhu, R. R.-H., Dhonde, H., & Hsu, T. T. TxDOT Project 0-1854: Crack Control for Ledges in
Inverted "T" Bent Caps. University of Houston. 2003.
204
APPENDIX A. Design Example

A.1 Overview
A detailed strut-and-tie design example of one of the specimens in the experimental program is
provided in this appendix. The STM design provisions recommended in this project for invertedT beams were used.
Figure A-1: Beam DL1-42-1.85-03

Beam 10a: DL1-42-1.85-03, shown in Figure A-1, was designed to fail in shear. The shear-spanto-depth ratio was 1.85 and the predicted failure mode was crushing of the strut-to-node interface
at the support closest to the load point. Therefore the objective of this design example was to
ensure that all other elements in the strut-and-tie model do not govern the capacity of the beam.
This is thus not a typical design where a beam is dimensioned to resist a certain combination of
loads. Rather this beam was designed to fail in shear while its strength remained within
laboratory testing capacities.
As mentioned in Chapter 3, all of the beams in the experimental program were constructed with
the same web width of 21 in. and ledge width of 10.5 in. The total height of this beam was 42in.
As this was to be a long, deep ledge specimen, the ledge was constructed with a depth of 21 in.
and continued from the edge of the bearing plate to the support. The ledge dimensions and the
final reinforcement layout are shown in Figure A-2.
Originally the beam was designed with a specified concrete compressive strength of fc = 4 ksi
and steel tensile strength of fy = 60 ksi. The values shown in the following example are the actual
measured material strengths, as presented in Chapter 3. Examining the strut-and-tie model using
these values enables a direct comparison between the calculated strength of the specimen and the
actual strength obtained through testing; as discussed in Chapter 5.
205
7'-1081"
5'-985"
P2
P1
Beam 010a
DL1-42-1.85-03
Beam 10b
DL1-42-2.50-03
20'-914"
21'-314"
27'-8"
3"
3'-283"
C
912"
STIRRUP
SPACING
[email protected]" = 7'-7"
8@8" = 5'-4"
No. 4 @ 6.5"
3"
3@9" = 2'-3"
No. 6 @ 8"
No. 6 @ 9"
C
21"
13@5" = 5'-5"
1'-9"
112"
6 No. 11
3'-185" 3'-6"
6 No. 11
8 No. 4
No. 6 @ 5"
1012"
No. 6 @ 8"
No. 5 @ 5"
6 No. 9
6 No. 11
1012"
No. 4 @ 6.5"
8 No. 4
112"
2"
C
[email protected]" = 5'-5"
No. 6 @ 5"
3'-283"
8 No. 4 @ 6.5"
No. 4 @ 6.5"
No. 5 @ 8"
6 No. 9
No. 5 @ 6.5"
6 No. 9
1'-9"
2"
3'-6"
Cross Section
12 No. 11
4 No. 11
12 No. 11
Section A-A
4 No. 11
Section B-B
Figure A-2: Elevation and cross-sectional details of DL1-42-1.85-03.
206
12 No. 11
Section C-C
4 No. 11
A.2 Defining the Beam

The following calculations outline the steps taken to develop a strut-and-tie model to determine
the capacity of DL1-42-1.85-03.
Gross Section Properties
First the gross dimensions of the beam are determined as presented in Figure A-2.
1
L 255 in
Length of the beam between supports
4
5
L1 69 in
8
5
Near support to load
L2 185 in
8
Far support to load
b 21in
Web width
h l 21in
Ledge height
Load and Support Plate Properties

The next step is to define the load and support plate sizes as these determine the size of their
respective nodes and influence the strength of the strut-and-tie model. Once the size of the load
plate is known, the distance the load will spread over the length of the ledge can also be
calculated.
wl 9in
Load plate width
ll 26in
Load plate length
ws 20in
Support plate width
ls 16in
Support plate length
d l h l 1.5in 0.5d 5 19
3
lsp ll 2 d l 64 in
8
3
16
in
45
Distance from the top of ledge
to compression reinforcement
ll
dl
ll+2dl
Resulting load spread
Material Properties
The beam was originally designed with specified material strengths of fc = 4 ksi and fy = 60 ksi.
This example, however, will consider the measured material strengths to get a better estimate of
the actual strength of the beam.
207
f' c 4.93ksi
2
fy_11 71.0075ksi A11 1.56in
d 11 1.41in
d 6 0.75in
d 5 0.625in
d 4 0.5in
fy_6 61.9ksi
A6 0.44in
fy_5 64.285ksi
A5 0.31in
fy_4 64.425ksi
A4 0.20in
STM Nodal Factors

As discussed in Chapter 2, the strength of the node depends on the elements (struts and/or ties)
framing into it as well as the particular face under consideration. A concrete efficiency factor, ,
is applied to the concrete strength, fc, to limit the compressive stress at the face of each node.
fcu m f' c
CCC_b 0.85
Bearing and back face of CCC node
CCT_b 0.7
Bearing and back face of CCT node
0.85
0.65 if 0.85
CCT_stn 0.45 if
0.85
f' c
20ksi
0.45 0.604
20ksi
f' c
0.65
20ksi
f' c
CCC and CCT strut-to-node interface
otherwise
A confinement modification factor, m, is also used to increase the allowable bearing stress in a
node to account for the effects of confinement provided by the concrete when the loaded area,
A1, is smaller than the supported area, A2. In this case the support plate is only 20 in. wide
whereas the beam is 21 in. wide.
2
A 1 ws ls 320 in
A 2 ws 1in ls 1in 357 in

m
A2
A1
1.056
A.3 Longitudinal Strut-and-Tie Model

The next step is to determine the location of the struts and ties as shown in Figure A-3. The
horizontal flexure tie, Tie AH, is positioned at the center of the longitudinal reinforcement. The
208
horizontal strut at the top of the beam, Strut BF, is located at the center of the compression stress
block, determined from a standard flexural analysis. Tie BC is located at the center of the hanger
reinforcement at the load point. Tie DE and Tie FG were then located at third points between the
edge of the hanger bar spread and the line projected at a 25 degree angle from the edge of the
support plate at Node H to the top of the beam as shown in Figure 2-14.
F
L2
L1
Figure A-3: Strut-and-tie model for DS1-42-1.85-03

2
A' s 6 A 11 9.36 in
Compression Steel
A s 12A 11 18.72in
Flexure Steel
wflex 2 4.365in 8.73 in
Flexure Tie Width
A s fy_11 A' s fy_11

0.85 b f' c
d 42in
h d
ad
L1
d
a
2
wflex
2
7.553in
Compression Stress Block
37.635in
Top of beam to center of flexural reinforcement
33.859in
Moment arm
1.85
Shear span-to-depth ratio
With the locations of the struts and ties determined, the size of the nodes can be defined. The
length of Node B in the longitudinal direction is determined by the spread length of the hanger
tie, lsp. It is then divided proportionally to the amount of the applied load that acts on the left and
right side of the node to obtain le and la respectively. The depth of the node is taken as the depth
of the compression stress block, a, determined above.
209
le
la
B
a
L2
lsp
L1
Figure A-4: Geometry of CCT Node B

L2
la lsp 46.813in
Portion of the load spread length, lsp, on the left side of Node B
L2
le lsp 1
17.562in
Portion of the load spread length, lsp, on the right side of Node B
To determine the angle of each of the struts, the arctangent is taken of the distance from the
center of the flexural steel to the center of the compression stress block divided by the horizontal
length of the node.
lab 0.5la le 0.5lsp 8.781in
lbe la 0.5le 0.5lsp 23.407in
l2.5
L2 lbe
3
A atan
54.07in
0.508
L1 lab
E atan
0.559
l2.5

G atan
H atan
0.559
l2.5

0.559
l2.5

Distance from center of length la to location of Tie B C

Distance from center of length le to location of Tie B C
Distance from the edge of the load spread, lsp, to Support H
A A
180
E E
180
G G
180
H H
180
29.092
Angle between Strut AB and Tie AC
32.055
Angle between Strut B E and T ie DE
32.055
Angle between Strut DG and Tie EG
32.055
Angle between Strut FH and T ie GH
Node Capacities
The strength of each node is dependent on the area of the face of the node and the limited
compressive stress, fcu. The area of each node face is equal to its length, calculated according to
Figure 2-12 or Figure 2-13, multiplied by the beam width, b, or the plate width, if applicable. For
the nodes defined by the support or bearing plates the concrete modification factor is applied.
Only the critical nodes, the support Nodes A and H, and the hanger Node B are checked in the
longitudinal model. For each critical node, the capacity of the strut-to-node interfaces (STNI)
210
and bearing surfaces are determined. For Node A, the notation RA refers to the bearing strength
at the support at A as shown in Figure A-4. ABA refers to the capacity of the strut-to-node
interface of Strut AB at Node A. Likewise ABB is the capacity of the strut-to-node interface of
strut AB at Node B, BDB is the strength of the strut-to-node interface of strut BD, etc. PL is the
total bearing capacity at the load point, considering both ledges. The factors in the following
equations reflect the respective type of node face.
Node
F F
Node D
B
ABB
BEB
Strut BE
FHH
H
RH
Node B
Strut BD
Strut FH
Node H
BDB
Node
G G
L2
NodeEE
Strut AB
PL
Node CC
ABA
L1
Node A
RA
A
Figure A-5: Strut and node notation

Node A
RA m CC T_b f' c ws ls 1166.418kip
ABA m CC T_stn ws ls sin A wflex cos A f' c 968.409kip
Node B
BDB CC T_stn b a f' c A' s fy_11 1136.518kip
a
BEB CC T_stn b le a tan E sin E
cos
f' 182.417kip
ABB CC T_stn b la sin A a cos A f' c 1834.463kip
Node H
FHH m CC T_stn ws ls sin H wflex cos H f' c 998.751kip
RH m CC T_b f' c ws ls 1166.418kip
Load
PL 2 m CC T_b f' c wl ll 1705.886kip
Member Forces
The purpose of the experimental program was to examine the shear capacity and behavior of the
inverted-T beams. Therefore each beam was designed to fail in shear. For an a/d ratio of 1.85,
the expected failure mechanism is crushing of the nodes at the ends of the direct strut. Therefore,
the beam was detailed with that as its weakest element. For DS1-24-1.85-03, the strut-to-node
interface of Strut AB at Node A controlled with the smallest capacity of 968.4 kips. The low
capacity value of the STNI of Strut BE at Node B is misleading as it can be combined with the
STNI of BD (discussed later on). The design check is then performed with the applied load, P,
that would result in a force of 968.4 kip in Strut AB. The resulting force in each member is
donated as Felement such as FAB for the force in Strut AB or FAC for the force in Tie AC. For the
critical nodes, the ratio of capacity to force is highlighted in the following calculations. As long
as that ratio is greater than 1.0, the element is not expected to control the strength of the beam
211
L
P ABA sin A 647.49kip
L2
P 647 kip
Design Beam to Fail at Strut-to-Node Interface at Node H

Node A
L2
FRA P 470.848kip
L
FRA
FAB
968.409kip
sin A
RA
FRA
ABA
FAB
ABB
FAB
2.477
1.894
FAC FAB cos A 846.238kip
Node H
L1
FRH P 176.639kip
RH
FRH
FRH
FFH
332.821kip
sin H
FHH
FFH
FGH FFH cos H 282.079kip
Node F

FFG FFH sin H 176.639kip
FDF FFH cos H 282.079kip
Node G
FFG
FDG
332.821kip
sin G
FEG FDG cos G FGH 564.159kip
212
6.603
3.001
Node D
FBD FDG cos G FDF 564.159kip

FDE FDG sin G 176.639kip
Node E
FDE
FBE
332.821kip
sin E
FCE FBE cos E FEG 846.238kip
BDB
Node B
FBC FBE sin E FAB sin A 647.487kip
FBD
BEB
2.015
FBE
0.548
Try combining Strut BE and Strut BD (shallow angle)

Checks
FRH FRA P 0 kip
FAC FCE 0 kip
FBD FBE cos E FAB cos A 0 kip
Two comparisons are highlighted red in the above calculations. The first is the ratio of the
strength of the strut-to-node interface at Node A to the force in Strut AB. That ratio is equal to
1.0, confirming that the applied load was determined based on the capacity of that element. The
second red highlight is for the strut-to-node interface of Strut BE at Node B. Based on the above
calculations it would appear that the node does not have the capacity to carry the applied node.
This is not true, however because of the shallow angle between the two, Strut BE and Strut BD
can be combined when they intersect Node B. As they move away from each other, Strut BE can
spread and thus will not be critical to the capacity of the beam.
Modified Node face at Strut BE and BD

0.206
FBD FBE cos E
Fx FBD FBE cos E 846.238kip
mod atan
FBE sin E
mod mod
180
11.79
Fy FBE sin E 176.639kip
Fmod Fx Fy 864.477kip
modF CC T_stn b la sin mod a cos mod f' c A' s fy_11 cos mod 1710.185kip
modF
Fmod
1.978
ok
213
Tie Requirements
In order to ensure a shear failure due to node crushing, enough reinforcement must be provided
in each tie to carry the applied load without yielding. For each of the 42 in. beams in the
experimental program, twelve No. 11s were provided as the flexural reinforcement. If this
proved insufficient steps were taken to ensure a shear failure, such as decreasing the compressive
strength of the concrete. From the truss forces calculated above, the remaining ties were detailed
as follows. The number of stirrups (No. 4 bars) and spacing was predetermined based on the
reinforcement ratio for each beam. The hanger ties, however, were designed specifically to resist
the force in the member, F, with T representing the tensile capacity of the strut based on the
total area and yield strength of the steel. For example, the force in hanger Tie BC from the
applied load P would be FBC determined in the previous section, and the capacity of the tie is TBC
determined by the steel provided to resist that force.
Flexural Reinforcement- #11 bars
Fflex max FAC FCEFEGFGH 846.238kip
Tflex A s fy_11 1329.26kip
Hanger B ars- #6 stirrups

TBC 28 A 6 fy_6 762.608kip
Tflex
Fflex
TBC
FBC
1.571
1.178
Tie FG- #5 stirrups and #6 hangers

TFG 14A 4 fy_4 6A 6 fy_6 343.806kip
TFG
FFG
1.946
Tie DE- #6 hangers

TDE 16A 6 fy_6 435.776kip
214
TDE
FDE
2.467
A.4 Cross-Sectional Strut-and-Tie Model

A similar detailing process is repeated for the cross-sectional strut-and-tie model. The overall
dimensions are set by the type of inverted-T beam being designed. DS1-42-1.85-03 is a 42 in.
deep beam with 21 in. deep ledges as show in Figure A-2. The resulting strut-and-tie model is
detailed in Figure A-6. Strut BC is located at the depth of the longitudinal flexure tie. Hanger Tie
BF and EC are aligned with the vertical hanger bars and Tie AD is aligned with the horizontal
ledge bars. Struts AB and DC allow the load to be transferred from the ledge to the bottom of the
web.
F
A
l4
D
l5
l3
hl
B
l1
C
l2
l1
Figure A-6: Cross-Sectional Strut-and-Tie Model

Truss Geometry
3
l4 1.5 in
8
l1 l4 1in 0.5wl 7.375in
l2 b 2l4 17.25in
5
l5 2
in 2.313in
16
l3 h l 0.5wflex l5 14.322in
ln lsp wflex
l3
1.095
l1

atan
180
62.755
Because the cross section and loading is symmetric, only one side needs to be considered. The
capacity of Node A and Node B are calculated in the same manner as those in the longitudinal
strut-and-tie model with Ra referring to the bearing capacity at Node A and aba to the capacity of
the STNI of Strut AB at Node A.
215
Node a
Ra CCT_b f' c wl ll 807.534kip
ab a CCT_stn ll wl sin ( ) 2l5 cos ( ) f' c 782.758kip
Node b
ab b CCT_stn lsp wflex cos ( ) 2l4 sin ( ) f' c 1404.031kip
bcb CCT_stn ln wflex f' c 1445.322kip
Next the forces in the cross-sectional strut-and-tie model are determined by applying half of the
load, P, from the longitudinal model to the bearing pads on each ledge. It is important to ensure
that the ratio of the capacity of each node face is greater than the applied force so that a ledge
failure does not occur. Once again, the F denotes the force in the member due to the applied
load, P.
Ra
P
FRa 323.771kip
FRa
FRa
Fab
364.174kip
sin ( )
ab b
Fab
2.494
3.855
Fdc Fab 364.174kip
Fbc Fab cos ( ) 166.717kip
bcb
Fbc
8.669
Fad Fab cos ( ) 166.717kip
Fbf Fab sin ( ) 323.771kip
Fce 0.5FBC 323.771kip
Fbf Fce 0 kip
Load check
The final step in the design process is to detail the steel in the ledge. The hanger bars were
checked in the longitudinal strut-and-tie model and do not need to be examined here. Enough
horizontal ledge bars need to be provided to prevent a ledge failure and ensure the performance
of the beam such that Tad > Fad.
Tad 12 A5 fy_5 239.14kip
216
Tad
Fad
1.434

Inverted T Beams

Uploaded by

Copyright:

Available Formats

Inverted T Beams

Uploaded by

Document Information

Original Description:

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

Inverted T Beams

Uploaded by

Copyright:

Available Formats

1. Report No.

Technical Report Documentation Page

4. Title and Subtitle

9. Performing Organization Name and Address

10. Work Unit No. (TRAIS)

12. Sponsoring Agency Name and Address

13. Type of Report and Period Covered

15. Supplementary Notes

Strength and Serviceability Design of Reinforced Concrete

CTR Technical Report:

Center for Transportation Research

CHAPTER 4 Experimental Results ...........................................................................................59

5.3.4 Summary and Conclusions ......................................................................................152

Figure 3-3: Specimen nomenclature. .............................................................................................35

Figure 4-38: Crack patterns for SS1-_-1.85-03 (web depth varies).............................................106

Table 6-3: Important Characteristics of El Paso Straddle Bents..................................................173

Determine the effect of tension-chord loading on the strength and

4. Make a recommendation on the application of TxDOT Project 0-5253 STM

Analysis of the experimental results is provided in Chapter 5. A discussion of the failure

Included in this Chapter is an overview of the theoretical background of inverted-T beam

Figure 2-1: Severely distressed inverted-T bent cap in El Paso.

Table 2-1: Crack width summary of in-service bent caps.

Ledge Ledge Height / Max Diagonal

Not available 1.85*

* Approximated based on field observations

BACKGROUND ON INVERTED-T BENT CAPS

In this section, a background on inverted-T beam behavior is given. The difference

2.3.2 Components of an Inverted-T Beam

Two additional types of reinforcement are required in an inverted-T beam compared to

Figure 2-4: Typical details of inverted-T bent caps.

DISCONTINUITY REGIONS OF BEAMS

A D-region (with D standing for discontinuity or disturbed) can typically be found on

discontinuity. In other words, a D-region is assumed to extend one member depth, d,

THEORETICAL BACKGROUND OF STRUT-AND-TIE MODELING

Strut-and-tie modeling (STM) offers a rational approach for obtaining lower-bound

The versatility of strut-and-tie modeling allows it to be used to design of any D-region

Idealized Prismatic Strut

CCC: nodes where only struts intersect

Figure 2-8: Node designations.

INVERTED-T DESIGN PROVISIONS

2.6.1 AASHTO LRFD Bridge Design Specifications (2012)

2.6.2 TxDOT Project 0-5253 STM Provisions

Figure 2-9: Node designations in an inverted-T beams

area of longitudinal tension steel (in.2)

Figure 2-13: Geometry of CTT node(adapted from Birrcher et al., 2009)

Figure 2-14: Determination of CTT vertical tie

Critical Section for

2.6.2.2 Outline of Strut-and-tie Modeling of Inverted-T Bent Specimens

Define loads and solve statics

Define geometry of the longitudinal strut-and-tie model

Figure 2-19: Widths of compression and tension chords

Figure 2-20: Development of strut and tie model

Define geometry of the cross-sectional strut-and-tie model

Figure 2-21: Forces in cross-sectional models

Perform nodal strength checks

Determined required steel area for each tie

Figure 2-22: Load spread area for ledge reinforcement

Figure 2-23: STM for inverted-T test specimen

INVERTED-T DEEP BEAM DATABASE