construction and load tests of a segmental precast box...
TRANSCRIPT
1. Report No. 2. Goyernment Accession No.
CFHR 3-5-69-121-5
4. Title and Subtitle
CONSTRUCTION AND LOAD TESTS OF A SEGMENTAL PRECAST BOX GIRDER BRIDGE MODEL
7. Author")
S. Kashima and J. E. Breen
9. Performing Organization Name and Addre ..
Center for Highway Research The University of Texas at Austin Austin, Texas 78712
~~----------~--~~--------------------------~ 12. Sponsoring Agency Name and Addre ..
Texas State Department of Highways and Public Transportation
Transportation Planning Division P. O. Box 5051 Austin, Texas 78763
1 S. Supplementary Not ..
TECHNICAL REPORT STANDARD TITLE PAGE
3. Recipient's Catalog No.
5. Report Date
February 1975 6. Performing Organi zation Code
8. Performing Organization Report No.
Research Report 121-5
10. Work Unit No.
11. Contract or Grant No.
Research Study 3-5-69-121 13. Type 01 Report and Period Coyered
Interim
14. Sponsoring Agency Code
Study conducted in cooperation with the U.S. Department of Transportation, Federal Highway Administration. Research Study Title: tTIesign Procedures for Long Span Prestressed Concrete Bridges of Segmental Construction"
16. Abstract
The cantilever construction of the first segmental precast prestressed concrete box girder bridge in the United States has been recently completed on the John F. Kennedy Memorial Causeway, Corpus Christi, Texas. The segments were precast, transported to the site, and erected by the balanced cantilever method of post-tensioned construction, using epoxy resin as a jointing material.
In order to check the applicability and accuracy of the design criteria, analytical methods, construction techniques, and the shear performance of the epoxy joints, an accurate one-sixth scale model of the three-span continuous bridge was built at the Civil Engineering Structures Research Laboratory of The University of Texas Ba1cones Research Center.
This report documents the construction and load testing of the bridge. Experimental results are compared with analytical values for the various stages of construction, service loadings, ultimate proof loadings, and final failure tests. During the cantilever construction and under service level loadings after completion, experimental results generally agreed with the computerized theoretical analyses. Because of the general absence of warping, a beam theory analysis reasonably predicted behavior of the bridge during the cantilever construction and under uniform service level loading. However, a folded plate theory analysis was required to predict distribution for nonuniform loadings and transverse moment distribution for wheel loadings. Ultimate load theories correctly indicated the load capacity of the structure when all loading and structural configurations were considered.
17. Key Word.
segmental, precast, box girder, bridge model, construction, load tests, cantilever
18. Di.tribution Siolement
No restrictions. This document is avai 1-able to the public through the National Technical Information Service, Springfield, Virginia 22161.
19. Securily Clonif. (of Ihi. rep or I) 20. Security Clanif. (of Ihi. page) 21. No. of Page. 22. Price
Unclassified Unclassified 284
Form DOT F 1700.7 18-69)
CONSTRUCTION AND LOAD TESTS OF A SEGMENTAL PRECAST
BOX GIRDER BRIDGE MODEL
by
S. Kashima
and
J. E. Breen
Research Report No. 121-5
Research Project No. 3-5-69-121
"Design Procedures for Long Span Prestressed Concrete Bridges of Segmental Construction"
Conduc ted for
The Texas Highway Department In Cooperation with the
U. S. Department of Transportation Federal Highway Administration
by
CENTER FOR HIGHWAY RESEARCH THE UNIVERSITY OF TEXAS AT AUSTIN
February 1975
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 views or policies of the Federal Highway Administration. This report does not constitute a standard, specification, or regulation.
i1
PREFACE
This report is the fifth in a series which summarizes the detailed
investigation of the various problems associated with design and construc
tion of long span prestressed concrete bridges of precast segmental con
struction. An initial report in this series summarized the general state
of the art for design and construction of this type bridge as of 1969.
The second report outlined requirements for and reported test results of
epoxy resin materials for joining large precast segments. The third
report summarized design criteria and procedures for bridges of this
type and included two design examples. One of these examples was the
three-span segmental bridge constructed in Corpus Christi, Texas, during
1972-73. The fourth report summarized the development of an incremental
analysis procedure and computer program to analyze segmentally erected
box girder bridges. This report summarizes structural performance data
obtained from a realistic one-sixth scale model of the structure and
compares these data with analytic results from several computer analyses.
This work is a part of Research Project 3-5-69-121, entitled
"Design Procedures for Long Span Prestressed Concrete Bridges of Seg
mental Construction." The studies described were conducted as a part of
the overall research program at The University of Texas at Austin,
Center for Highway Research. The work was sponsored jointly by the
Texas Highway Department and the Federal Highway Administration under an
agreement with The University of Texas at Austin and the Texas Highway
Depar tmen t.
Liaison with the Texas Highway Department was maintained through
the contact representative, Mr. Robert L. Reed, and the State Bridge
Engineer, Mr. Wayne Henneberger; Mr. D. E. Hartley and Mr. Robert E.
Stanford were the contact representatives for the Federal Highway Adminis
tration. Special thanks are due to Messrs. Thomas Gallaway,
iii
Lawrence G. Griffis, and John T. Wall, all assistant research engineers
at the Civil Engineering Structures Research Laboratory at The University
of Texas at Austin's Balcones Research Center. These gentlemen played
key roles in the development of model techniques, fabrication procedures,
erection methods, and instrumentation systems, and were responsible for
various stages of construction operations. The Laboratory staff all
contributed significantly to this project with their untiring willingness
to work long hours and make an extra effort throughout this project.
The overall study was directed by Dr. John E. Breen, Professor
of Civil Engineering. Dr. Ned H. Burns, Professor of Civil Engineering,
acted as an advisor on many questions concerning prestressing systems.
The erecting and testing was under the immediate supervision of
Dr. Satoshi Kashima, research engineer, Center for Highway Research.
iv
SUM MAR Y
The cantilever construction of the first segmental precast
prestressed concrete box girder bridge in the United States has been
recently completed on the John F. Kennedy Memorial Causeway, Corpus
Christi, Texas. The segments were precast, transported to the site, and
erected by the balanced cantilever method of post-tensioned construction,
using epoxy resin as a jointing material.
In order to check the applicability and accuracy of the design
criteria, analytical methods, construction techniques, and the shear
performance of the epoxy joints, an accurate one-sixth scale model of the
three-span continuous bridge was built previously at the Civil Engi
neering Structures Research Laboratory of The University of Texas
Balcones Research Center.
This report documents the construction and load testing of the
bridge. Experimental results are compared with analytical values for
the various stages of construction, service loadings, ultimate proof
loadings, and final failure tests. During the cantilever construction
and under service level loadings after completion, experimental results
gen~rally agreed with the computerized theoretical analyses. Because of
the general absence of warping, a beam theory analysis reasonably pre
dicted behavior of the bridge during the cantilever construction and
under uniform service level loading. However, a folded plate theory
analysis was required to predict distribution for nonuniform loadings and
transverse moment distribution for wheel loadings. Ultimate load theories
correctly indicated the load capacity of the structure when all loading
and structural configurations were considered.
The model bridge carried the ultimate proof loads [1.35 dead load
+ 2.25(live load + impact load)] specified by the 1969 Bureau of Public
Roads' criteria for all critical conditions. During tests to failure, the
epoxy joints performed very well and there was no evidence of epoxy
v
separation at the joints. The theoretical calculation for the failure
load agreed very well with the experimental results and indicated the
necessity for change in the conventional procedure for computing ultimate
design load for this type of bridge. Since the structural configuration
changes from a simple cantilever to a multispan continuous structure
during construction of the bridge, the ultimate design load for computa
tion of shears and bending moments for the bridge should be specified as
two values, each which must be satisfied as follows:
and
where
U U2
+ U3
Ul = 1.35.DLl + 2.25(LLl + 1Ll )--to be computed for a balanced cant~lever
1.35 DL, for negative moments, and 0.90 DL for positive moments, to be computed for a balanced cantilever
U3 = 1.35 DL3 + 2.25(LL3 + 1L3 ) + SL --to be computed for the completed continuous structure
dead load during cantilevering
dead load added after completion of closure (roadway surfacing, hand rails, etc.)
live load due to construction operations
design live load
impact load of construction operations
design impact load
resultant reactions due to prestressing of tendons and seating forces at outer supports
vi
IMP L E MEN TAT ION
This report presents the details of a comprehensive model test
program of a segmental prestressed concrete box girder bridge. This type
of construction is becoming increasingly popular in the United States
and this test represents a comprehensive check of the design procedures
used for proportioning of such a structure and documents the performance
of the structure at various stages of development during the cantilever
erection procedure, as well as under test loads at the completion of the
entire structure.
The test program validates the general design procedures and
is particularly important in verifying the adequacy of the epoxy jointing
procedures used in this type of construction. The model used was a
one-sixth scale representation of the box girder bridge erected over the
Intracoastal Waterway at Corpus Christi, Texas, and was valuable in
acquainting design, construction, and contractor personnel with the
type of problems which are encountered in this type of construction. In
itself the model had a high educational benefit.
Results of the test program indicated that this type of construc
tion is both safe and dependable. A very adequate factor of safety was
demonstrated and minor difficulties encountered in the field were satis
factorily explained as a result of additional model tests. As a result
of the tests several specific recommendations are made for development of
more accurate design procedures and criteria. Implementation of these
recommendations should result in economic cost savings in future projects
of this type and a more realistic assessment of the strength of such
bridges. A number of construction procedures high-lighted during the
model tests were carried out in the prototype erection and additional
changes are being made in future bridges of this type, based on the test
program.
vii
Chapter
1
CON TEN T S
INTRODUCTION .
1.1 1.2 1.3 1.4
General . Related Research Objective and Scope Report Contents . .
of the Model Study
2 SCALE FACTORS FOR THE BRIDGE MODEL
3
4
2.1 2.2
2.3
Factors Affecting Scale Selection • . . . Dimensions of Prototype and Model Bridge 2.2.1 Longitudinal Dimensions 2.2.2 Transverse Dimensions Choice of Materials 2.3.1 Concrete .•... 2.3.2 Prestressing Steel .•••••. 2.3.3 Reinforcing Bars
MATERIAL PROPERTIES
3.1
3.2
3.3
Concrete 3.1.1 Concrete Mix Design 3.1.2 Strength of Concrete 3.1.3 Modulus of Elasticity 3.1.4 Poisson's Ratio S tee 1 . . . . . . . . . . . . 3.2.1 Prestressing Strands and Wire
3.2.1.1 Ultimate Strength •. 3.2.1.2 Modulus of Elasticity.
3.2.2 Reinforcing Wire ..••.... 3.2.3 Strength of Bolts at Main Piers Epoxy Resin as a Jointing Material 3.3.1 General .•.•
SEGMENT DETAILS
4.1 Reinforcing Cage for Precast Segments • 4.1.1 Reinforcement 4.1.2 Anchorages ••. 4.1.3 Tendon Duct
viii
Page
1
1 4 5 6
9 9 9 9 9 9
11 11
15
15 15 15 17 17 18 18 18 18 18 20 20 20
23
23 23 23 29
Chapter
5
6
7
4.2 Forms . . . . . . . . . . . . . . . . . . . . 4.3 Strain Instrumentation in Segment Reinforcing
Cages . . . . . . 4.4 Casting Procedure . 4.5 Curing of Segments 4.6 Surface Preparation of Segments 4.7 Compensation Dead Load for Model Bridge 4.8 Actual Properties of the Model Bridge
INSTRUMENTATION AND DATA REDUCTION
5.1 5.2 5.3 5.4
General . . . . . . Load Cells and Pressure Gages Strain Gage Instrumentation and Data Reduction Surveyor's Level and Dial Gages
CONSTRUCTION OPERATIONS AND OBSERVATIONS
6.1 Construction Procedure 6.1.1 General .... 6.1.2 Construction of Piers 6.1.3 Erection of Pier Segments 6.1.4 Erection of Segments during Cantilever
Stages . . .. ....... . 6.1.4.1 General .......... . 6.1.4.2 Details of Segment Erection.
6.1.5 Correction of Horizontal and Vertical Al ignment . . . . . . . . . . . .
6.1.6 Casting of the Closure Segment ... 6.1.7 Prestressing of the Positive Tendons 6.1.8 Grouting of Tendons •...•... 6.1.9 Casting of the Midstrip Closure between
the Two Box Units ....... . 6.2 Behavior of the Bridge during Construction
6.2.1 Rotation Stiffness at the Main Pier 6.2.2 Prestressing Force 6.2.3 Deflections .... 6.2.4 Strains . . . . 6.2.5 Reaction at Outer Supports during
Positive Tendon Operations
DESIGN SERVICE AND ULTIMATE LOAD TESTS
7.1 7.2
General • . • Test Procedures • . • . • . 7.2.1 Simulation of Loading
7.2.1.1 Truck Loading. 7.2.1.2 Lane Loading
ix
Page
31
36 37 40 41 41 43
45
45 45 45 46
49
49 49 50 52
52 52 53
61 65 67 69
73 73 73 78 80 84
95
99
99 100 100 100 100
Chapter
8
9
7.3
7.4
7.2.2 Loading System and Instrumentation .. . 7.2.2.1 Loading System ..... . 7.2.2.2 Instrumentation and Observation
Test Results and Interpretations 7.3.1 General ••........ 7.3.2 Service Loading ...•..
7.3.2.1 Four Lane Loadings ••.. 7.3.2.2 Two Lane Loadings.
7.3.3 Ultimate Design Loading Specified by BPR 7.3.3.1 Additional 0.35 DL .... 7.3.3.2 Maximum Positive Moment, Main Span 7.3.3.3 Maximum Negative Moment at the
Main Pier . . . . . . . . . . . • 7.3.3.4 Maximum Shear Loading Adjacent to
the Main Pier • . • . . . • . 7.3.3.5 Maximum Positive Moment in the
Side Span . . . 7.3.4 Study of Transverse Moment. Summary . • • .
Page
103 103 103 108 108 109 109 117 123 123 123
134
139
140 140 148
FAILURE LOAD TESTS 149
8.1 General 149 8.2 Side Span Failure Load Test . 149
8.2.1 General 149 8.2.2 Test Results. 151 8.2.3 Calculation of Side Span Live Load
Capacity. 159 8.3 Failure in the Main Span 180
8.3.1 General 180 8.3.2 Test Results 181 8.3.3 Determination of the Main Span Live Load
Capacity. . . 192 8.3.3.1 Flexure. 192 8.3.3.2 Shear. 204
8.4 Punching Shear Tests 206
SPECIAL MEASUREMENTS . 209
9.1 General . • . . 209 9.2 Prototype Instrumentation. 209 9.3 Measurements during Construction 211
9.3.1 Strains during Balanced Cantilever Construction . . . . . . 211
9.3.2 Closure Operations 213 9.4 Prototype Load Testing 216 9.5 U1 tima te Shear Tes t . . . 223
x
Chapter
10 CONCLUSIONS AND RECOMMENDATIONS
10.1 Conclusions ...... . 10.1.1 Primary Conclusions 10.1.2 Secondary Conclusions
10.2 Recommendations •..•.•.•. 10.2.1 Design Recommendations 10.2.2 Construction Recommendations
REFERENCES
APPENDIX A PROTOTYPE BRIDGE PLANS
APPENDIX B SPECIFICATION FOR EPOXY BONDING AGENT
APPENDIX C NOTATION. . . . . . . . . . . . . . .
xi
Page
229
229 229 230 232 232 233
235
239
251
255
Table
2.1
2.1
3.1
3.2
3.3
3.4
3.5
3.6
3.7
3.8
3.9
3.10
4.1
4.2
6.1
6.2
6.3
6.4
7.1
7.2
LIS T o F TAB L E S
Choice of Strands and Wire for the Model Bridge
Choice of Reinforcing Bars for the Model Bridge
Concrete Mix Design
Compressive Strength of Concrete
Splitting Tensile Strength of Concrete .
Modulus of Elasticity of Concrete
Ultimate Strength of Prestressing Strands and Wire.
Modulus of Elasticity of Wire and Strands
Test Results of Reinforcing Wire of Precast Segments ••
Tensile Strength of Bolts at Main Piers
Epoxy E Flexure Test Results.
Epoxy E Shear Test Results .•
Dimensions of Anchorage Bearing Plates
Material Properties and Section Properties of the Model Bridge . . . . . • .
Friction Loss Test Results •
Elongation of Prestressing Cables in Several Stages
Details at Each Stage of Construction
Reaction at Outer Support during Prestressing in Main Span . . . • . . . . . . . . . . • • . . . •
Critical Loading Conditions in Longitudinal Direction
Moment or Shear for Each Loading Case
xii
Page
12
12
15
16
16
17
18
19
19
20
21
21
29
44
79
80
85
96
99
105
Table Page
7.3 Comparison of Experimental and Theoretical Transverse Strains ........ ..... . 146
8.1 LF of (LL + IL) for the First Plastic Hinge at Each Joint 198
8.2 Punching Shear Test Results 208
xiii
Figure
1.1
2.1
2.2
2.3
2.4
3.1
4.1
4.2
4.3
4.4
4.5
4.6
4.7
4.8
4.9
4.10
4.11
4.12
4.13
4.14
LIS T o F FIG U RES
Typical balanced cantilever construction
Longitudinal dimensions of bridge
Cross section of bridge
Details of prestressing strands and wire
Details of reinforcing bars
Testing of reinforcing wire
Individual mats for a cage
Cage for a segment ..
Details of reinforcing mats
in segment .
Details of web reinforcement modification
Reinforcement for the main pier segment model
Details of anchorage and spiral reinforcement
Typical arrangement of tendon duct near anchorage
Tendon duct profiles in main span
Tendon duct profiles in side span
Details of tendon profiles .
General view of soffit forms
Details of forms for segment
General view of assembled forms
Position of strain gages in segments
4.15. Sequence of casting segments ..••
xiv
Page
2
10
10
13
14
19
24
24
25
27
28
30
31
32
33
34
35
35
37
38
39
Figure Page
4.16 Position of rams for separation of segment 40
4.17 Tools used for surface preparation 42
4.18 Compensating dead loads for the model bridge 42
5.1 Strain gage instrumentation 46
5.2 VIDAR system and teletype 47
6.1 Reinforcement of the main pier 50
6.2 Connection of pier to the testing floor 51
6.3 Detail of support at main pier. 51
6.4 General view of the pier segment on the main pier 53
6.5 Typical independent crane used for erection of segments 54
6.6 Mechanical device for temporary joining 55
6.7 Typical structure supported cranes used in some cases 55
6.8 Insertion of prestressing cable 57
6.9 Adjusbnent of level 58
6.10 Application of the epoxy resin on the joining surface 58
6.11 General view of prestressing system for negative tendons 59
6.12 Details of prestressing system 60
6.13 Errors of horizontal alignment 61
6.14 Adjustment of horizontal alignment 63
6.15 Accumulation of joining error 64
6.16 Correction of the vertical alignment 65
6.17 Connection of tendon duct at the closure 66
6.18 Simulation of the closure 66
6.19 Setup for prestressing of positive tendons • 68
xv
Figure Page
6.20 Order of grouting for each tendon group 70
6.21 Examples of tendon arrangement . . . . . 70
6.22 Connection of injection pipes and inlets of tendons 72
6.23 Deflection at unbalanced loading 74
6.24 Individual factors affecting the deflection 76
6.25 Suggested improvement for temporary support 78
6.26 Deflection predicted by SIMPLA2 for cantilever erection 82
6.27 Typical measured deflection during cantilever erection • 82
6.28 Deflection relative to the center of 6th segment .•.• 83
6.29 Deflection at the center in main span during positive tendon operations 83
6.30 Strain in Ml and Sl segments during balanced cantilever construction 86
6.31 Strains in Ml and Sl segments during erection of half segments . . . . . . 87
6.32 Strains in Ml segments during closure operation 88
6.33 Strains in the top slab of Ml segments during construction • 89
6.34 Strains in the bottom slab of Ml segments during construction . 90
6.35 Strains in the M6 and S6 segments at erection of the 6 th segment. . . . . . . . . . . . . . 91
6.36 Strains in the M9 segment during closure operations 92
6.37 Strains in the top slab around the center of the main span . . . . . . . . . . 93
6.38 Strains in the bottom slab around the center of the main span . . . . . . . . . . . 93
6.39 Calculation of end reaction due to positive tendons 96
xvi
Figure
7.1
7.2
7.3
7.4
7.5
7.6
7.7
Dimensions of full size and 1/6 scale model AASHO HS20-Sl6 truck . . . . . . . . .
Scaled wheels and AASHO HS20-Sl6 truck
Application of concentrated loads above the webs (in each lane) ........ . . . .
Lane loads and equivalent concentrated loads for four lanes with impact allowances •
Application of lane load ..
Locations of dial gage readings
Position of truck loading
7.8 Deflections for truck loadings (four lanes) in side
7.9
7.10
7.11
7.12
7.13
span ....
Reactions at outer supports for truck loadings (four lanes) in side span
Longitudinal strains along SS7 and SS6 for truck loadings (four lanes) in side span
Deflections for lane loadings (four lanes) in main span
Reactions at outer supports for lane loadings (four lanes) in main span . . . . . . .
Longitudinal strains along NM9 for lane loadings (four lanes) in main span . . . .
7.14 Deflections for lane loadings (four lanes) in main and
7.15
7.16
7.17
7.18
7.19
one side span
Reactions at outer supports for lane loadings in main and one sid e span . . . . . .
Longitudinal strains for lane loadings (four lanes) in main and one side span . . . . . . . . . . .
Deflections for lane loadings (two lanes) in main span .
Longitudinal strains along NM9 for lane loadings (two lanes) in main span . . . . . .
Reactions at outer supports for lane loadings (two lanes) in main span . . . . • . . . . . .
xvii
Page
101
102
103
104
106
107
109
110
111
112
113
113
114
115
115
116
118
119
119
Figure Page
7.20 Deflections for lane loadings (two lanes) in main and
7.21
7.22
7.23
7.24
7.25
7.26
7.27
7.28
7.29
7.30
7.31
7.32
7.33
7.34
7.35
7.36
7.37
one side span
Longitudinal strains for lane loadings (two lanes) in main and one side span . . . . . .. ..... .
Reactions at outer supports for lane loadings (two lanes) in main and one side span
Concrete blocks for 0.35 DL
Deflections for 0.35 DL
Reactions at outer supports for 0.35 DL
Longitudinal strains along NM9, NMl and NSl for 0.35 DL
Deflections at SM10 for lane loadings (four lanes) in main span (design ul tima te) .......... .
Deflections for lane loadings (four lanes) in main span (design ultima te) ........ . . . .
Reactions at outer supports for lane loadings (four lanes) in main span (design ultimate) .....
Longitudinal strains along NM9 for lane loadings (four lanes) in main span (design ultimate) .....
Development of cracks for lane loadings (four lanes) in main span (design ultimate) ....... .
Relation between the end reaction and cracking moment
Deflections at SM10 for lane loadings (four lanes) in the main and one side span (design ultimate) ...
Deflections for lane loadings (four lanes) in the main and one side span (design ultimate) .....
Longitudinal strains at NM9 for lane loadings (four lanes) in the main and one side span (design ultimate)
Longitudinal strains at NMl for lane loadings (four lanes) in the main and one side span (design ultimate)
Reactions at outer supports for lane loadings (four lanes) in the main and one side span (design ultimate)
xviii
120
121
122
124
125
125
126
127
129
130
130
131
133
133
135
136
136
137
Figure
7.38
7.39
7.40
7.41
7.42
7.43
7.44
7.45
7.46
7.47
7.48
8.1
8.2
Development of cracks for lane loadings (four lanes) in the main and one side span (design ultimate) •
Arrangement of slip gage at the first joint
Position of truck loadings (design ultimate)
Deflections at SS7 for truck loadings (four lanes) in side span (design u1 tima te) •......•.
Deflections for truck loadings (four lanes) in side span (design ultimate) . • . . . • . • .
Longitudinal s trains at SS7 for truck loadings (four lanes) in side span (design u1 timate) .
Longitudinal strains at SS6 for truck loadings (four lanes) in side span (design u1 tima te) . . .
Reactions at outer supports for truck loadings (four lanes) in side span (design u1 tima te) . . . Transverse strain at SS7 R for truck loadings (four lanes) in side span (design ultimate) .....•
Transverse slab moment (M ) diagram for different loading cases by MUPD1 . ~ . . . . . . • . • . .
Typical strain readings in punching shear tests
Bridge failure loading .
Truck loadings at 1.0(LL + 1L) for maximum positive moment in the side span . • . .
. .
8.3 Deflections at the center of the SS7 segment in side span • . . . •
8.4 Longitudinal strains at the center of the SS6L moment
8.5 Longitudinal strains at the center of the SS7R segment
8.6 Development of cracks during loading . . . . .
8.7 Longitudinal strains at the center of the SSl segment
8.8 Reactions at the outer supports
8.9 Failure mechanism for truck loadings in the side span
xix
Page
138
139
141
141
142
143
143
144
144
145
147
150
150
152
153
153
154
155
156
157
Figure
8.10 Horizontal force on the top of the outer pier
8.11 Deflections along the center of the SS7 segment in side span • • . . • . . .• ........ .
8.12 Transverse strains at the center of the SS7 segment
8.13 Forces and moments acting on the bridge for a typical cantilever stage . . • • .. .....•... . . •
8.14 Horizontal forces and moments at completion of cantilever erection
8.15 Horizontal forces and moments due to prestressing in side span
8.16 Horizontal forces and moments due to prestressing in
8.17
8.18
8.19
8.20
8.21
8.22
8.23
8.24
8.25
8.26
8.27
8.28
8.29
8.30
main span
Horizontal forces and moments at completion of all prestressing operations .•..
Moment diagram for 0.35 DL (ME3) •
Moment diagram for truck loads in side span (~4)
Ultimate force at a certain section
Stress-strain curves of prestressing cables
Calculation of T. for different strain profiles 1.
Calculation of ultimate internal moment
Calculation of ultimate internal moment at pier section
Plastic hinges and moment diagram
1. O(LL + IL) loading for maximum shear at the NM pier
Reaction at outer supports during loading to failure
Longitudinal strains at SS7R and SS6L during loading to failure . • • . . • . . . . . . . . . • . . •
Longitudinal strains at SSI during loading to failure
Longitudinal strains at SMI during loading to failure
xx
Page
158
160
161
162
163
165
166
167
170
170
171
172
173
174
177
178
181
182
182
183
183
Figure
8.31
8.32
8.33
8.34
8.35
8.36
8.37
8.38
8.39
8.40
8.41
8.43
8.44
8.45
9.1
9.2
Longitudinal strains at NS6 during loading to failure
Longitudinal strains at NM6 during loading to failure
Deflections at SM10 during loading to failure
Development of cracks around NM pier during loading to failure
Development of cracks around the center of main span during loading to failure ...
Typical crack around the joint (after failure)
Longitudinal strains at NM9 during loading to failure
Longitudinal strains at NM1 during loading to failure
Longitudinal strains at NS1 during loading to failure
General view of the bridge after failure .
Deflections along the bridge during loading to failure
Moment diagram for 0.35 DL for three different conditions (ME3) . . . . . . ....
Moment diagram for 1.0(LL + IL) for three different conditions (ME4)' ..... . ....
Failure loading at the center of the main span.
Location of strain gages in prototype bridge
Measured strain in M1 and Sl segments during several stages of balanced cantilever construction .
9.3 Measured strains in M1 segments during closure operation
9.4 Measured strains in the model M9 segment and prototype M10 during closure operations .... . . . . . •
9.5 (a) Truck loading tests (b) Average truck properties
9.6 Influence line for bottom flange strain, Station SSl
xxi
Page
184
184
186
187
188
189
190
191
191
193
194
196
197
202
210
212
214
215
217
219
Figure Page
9.7 Influence line for bottom flange strain, Station MS1 220
9.8 Influence line for bottom flange strain, Station SS7 221
9.9 Influence line for bottom flange strain, Station MS10 222
9.10 Test arrangement--u1timate shear test . . . . . . . . 225
xxii
C HAP T E R 1
INTRODUCTION
1.1 General
Construction of longer span bridges is increasing in the United
States to satisfy requirements of function, economics, safety, and aes
thetics. The long span potential of prestressed concrete cannot be fully
developed in pretensioned I-girder and composite slab systems. These
systems have practical limits in the 120 ft. span range. However, sub
stantially longer span prestressed concrete bridges have been built in
several countries by utilizing precast and cast-in-place box girder 32
bridges erected using cantilevering techniques: This study will treat
only the segmental precast prestressed concrete box girder bridge erected
by the cantilever method.
In this construction method, precast segments [Fig. l.l(a)] are
cast and transported to the bridge site. The precast segments are
erected, as shown in Fig. l.l(b), as balanced cantilevers from the pier
segment which is rigidly connected to the pier either temporarily or
permanently. In some applications temporary props are used to provide
the cantilever moment capacity. In the first applications of this con
struction technique, concrete or mortar was used as a jointing material
between segments. However, the French used epoxy resin successfully as a
jointing material in 1964. 30 Because of the rapid setting of the epoxy
resin, this type of jointing shortened the construction period appreciably
and became widely accepted. As each pair of segments is positioned at the
ends of the balanced cantilever, negative moment tendons are inserted and
tensioned. These tendons must provide moment capacity for the full canti
lever moment. Erection continues until the last cantilevered sections are
placed at the center of the span and at the end supports, as shown in
1
2
(0) TYPICAL CROSS SECTION OF BOX GI RDER
R"!¢ 't::::.::::]'~ B ••
(b) CANTILEVER ERECTION OF PRECAST SEGMENTS
JOINT
MAIN PIER
..........
PRESTRESSING CABLE TO BALANCE THE WEIGHT OF THE SEGMENT AND RESIST --+ NEGATIVE
MOMENT
REGULAR SEGMENT
SEGMENT
BOLTS OR PRESTRESSING CABLES
(c) COMPLETION OF CANTILEVER CONSTRUCTION
CAST-IN-PLACE CLOSURE SEGMENT
PRESTRESSING CABLES TO RESIST LIVE LOAD IN POSITIVE MOMENT REGION
Fig. 1.1. Typical balanced cantilever construction
3
Fig. l.l(c). The positive moment tendons in the end span are prestressed
and the end segments are seated on their supports prior to or during
stressing of prestressing cables in the main span positive moment region.
At midspan, the gap between the two cantilever arms is closed with cast
in-place concrete. Prestressing cables to resist live load in the central
span positive moment region are inserted and stressed, Reactions in the
end spans are adjusted as required.
The use of such precasting and cantilevering techniques has the 16 30 following advantages: '
(1) Maintenance of navigational or traffic clearance during construction.
(2) High quality control of segments and control of deflection.
(3) Flexible choice of the segment length depending on the capacity of transportation and lifting equipment.
(4) Simultaneous start of segment casting and pier construction.
(5) Significantly reduced erection time at the site.
(6) Highly efficient use of forms.
Because this type of bridge had never been built in the United
States, a cooperative research project with the Texas Highway Department
and the Federal Highway Administration to investigate the various prob
lems associated with design and construction procedures for long span
precast prestressed concrete box girder bridges of segmental construction
was undertaken by The University of Texas at Austin, Center for Highway
Research, in 1968. 9
The Texas Highway Department utilized a preliminary design devel
oped as part of the project by The University of Texas at Austin researchers
in developing plans for a long span bridge on the John F. Kennedy Memorial
Causeway, Park Road 22, Corpus Christi, Texas. The requirement to main
tain navigational clearance during construction as well as the highly
corrosive environment on the Texas coast led to the choice of a precast
prestressed concrete box girder bridge built in cantilever.
In order to study the applicability and accuracy of the design
criteria, analytical methods, construction techniques, and shear
4
performance of the epoxy resin joints, an accurate one-sixth scale model
of the three-span continuous bridge was built and tested at the Civil
Engineering Structures Research Laboratory of The University of Texas at
Austin's Ba1cones Research Center. This report summarizes the construc
tion and testing of the model. Complete details are available in Ref. 18.
1.2 Related Research
This model test was only one element in a comprehensive investi
gation. Related research which has been completed includes:
20 22 (1) State-of-the-Art-Survey.' In the initial stages of the
research program, a comprehensive literature survey was completed.
(2) Design and Optimization. 21 ,23 Criteria were developed for
design procedures and preliminary designs of several example structures
were made. The Texas Highway Department adopted one of these prelimi
nary designs for the Corpus Christi structure and developed final plans.
The bridge was largely designed by the Ultimate Strength Design method
assuming that beam theory was applicable. Allowable stresses during
construction and under service loads were also checked using beam theory.
Service load behavior of the completed structure was then checked using
folded plate theory to determine the effect of warping. Optimization of
the cross section was studied by unconstrained nonlinear programming,
although the optimal cross section was not used in the final design.
. 10 11 (3) Segmental Ana1ys~s. ' In order to investigate the various
critical stages during the cantilevering procedures, an incremental box
girder analysis computer program utilizing the Finite Segment Method was
developed by Brown. This program (SIMPLA2) treats staged construction
with realistic prestressing forces and determines effects in both longi
tudinal and transverse directions.
(4) E . 1 S d fER . J' . 17,19 Wh'l xper~menta tu y 0 poxy es~n o~nt~ng. ~ e epoxy
resins have been widely used with concrete as coating or patching materials
and although a guide4 existed for the general use of epoxy resin with con
crete, at the inception of the study there was no U.S. specification for
5
the specific use of epoxy resins in jointing a segmental precast
prestressed concrete box girder bridge. A program of evaluations of
various epoxy resin properties and requirements was undertaken which
resulted in development of a Texas Highway Department tentative specifi
cation. Prior to the model bridge construction, a number of different
epoxy resins were evaluated and the epoxy resin which came closest to
meeting the specifications was selected for construction of the model.
A revised specification has been suggested in Ref. 19.
(5) Field Study. The Corpus Christi bridge was instrumented
with strain gages which were mounted on the reinforcement in various
segments. Readings were taken at the time of cantilever erection.
Structural behavior under actual service loads was observed for various
loading conditions upon completion of the prototype construction. A
summary of the field observations and a comparison with the model test
resul ts will be given in this report.
1.3 Objective and Scope of the Model Study
Tests of model structures can be very powerful tools to check the
adaptability or accuracy of design procedures and to verify analytical
assumptions and procedures.
In utilizing precast segments in construction of a bridge, continuity
of the structure at the joints is a very important problem. Several con
crete box girder bridge models have been tested in various labora
tories~2,16,35,36 All of the model bridges, except for a two-span (72 ft.
long) four-cell reinforced concrete box girder model tested in California,35
were precast prestressed microconcrete box segments joined with mortar. 12
One of the models had continuous reinforcement across all the joints.
The 1/12th scale Mancunian Highway microconcrete model 16 (16 ft., 10 in. 36 long) and the 1/16th scale three-cell box beam tested by Swann proved the
strength adequacy of the mortar joint in failure tests.
16 Some French tests have been reported which demonstrated the
structural continuity of epoxy joints in continuous slabs (1/4.64th scale,
50 ft. long). An investigation of the shear capacity of precast segments 16 joined with epoxy resin has been reported in Japan.
6
However, to the authors' knowledge there has not been another
test program in which a precast prestressed concrete segmental bridge
model with epoxy resin joints has been built with accurate simulation of
the construction procedure and then loaded to failure. Since this model
of a three-span precast prestressed concrete segmental twin box girder 1 29 bridge is a one-sixth scale "direct" mal. e1,' and was built in full con-
formance with prototype construction procedures, it closely simulates the
behavior of the prototype both in the elastic and inelastic range.
The objectives of this model study included the following:
(1) Determination of strain distribution due to prestressing and of deflections during cantilever construction and during closure operations.
(2) Documentation of bridge behavior under service level loading for the various design loading conditions.
(3) Comparison of analytical results from beam theory and folded plate theory with the construction and service level loading experimental results.
(4) Determination of bridge behavior under ultimate proof loading (1.35 DL + 2.25 (LL + 1L)) for the various design loading conditions.
(5) Determination of final failure mechanisms with special attention to any adverse effect of the epoxy resin on the shear or flexural capacity of the bridge. Determination of the punching shear capacity of the top slab and evaluation of any adverse effect of the epoxy resin on such capacity.
(6) Assessment of the applicability of the Ultimate Strength Design criteria proposed for this type bridge.
(7) Determination of any improvements of design details which might minimize field construction problems prior to the prototype bridge construction.
(8) Provision of a meaningful demonstration to prospective contractors to assist them in the visualization of the construction technique so as to reduce uncertainty and encourage bidding for the prototype construction.
1.4 Report Contents
Details of modeling procedures used are given in Chapter 2. Test
results of materials used for the model are discussed in Chapter 3.
Details of the segments, casting procedures, and segment preparation are
7
illustrated in Chapter 4. The instrumentation used in the model bridge
tests and the procedures for data reduction are briefly explained in
Chapter 5. Chapter 6 illustrates all construction procedures for the
model bridge and discusses the behavior of the bridge during construction.
After completion of construction, a wide variety of service level loadings
were applied to the bridge and then various design ultimate loads speci
fied by the 1969 Bureau of Public Roads criteria were applied to prove
the safety of the structure. Results from these tests are discussed in
Chapter 7.
Finally, the bridge model was loaded to complete failure. The
behavior during failure loadings is documented and comparisons of the
theoretical and experimental values for the ultimate strength of the
bridge are discussed in Chapter 8. A brief comparison of model test
results with field measurements and description of a special test of
the shear capacity of a reduced model consisting of several segments is
given in Chapter 9. Chapter 10 gives the conclusions determined from
the model study and recommendations for improvements in design and con
struction criteria.
!!!!!!!!!!!!!!!!!!!"#$%!&'()!*)&+',)%!'-!$-.)-.$/-'++0!1+'-2!&'()!$-!.#)!/*$($-'+3!
44!5"6!7$1*'*0!8$($.$9'.$/-!")':!
C HAP T E R 2
SCALE FACTORS FOR THE BRIDGE MODEL
2.1 Factors Affecting Scale Selection
Selection of the scale factor for the model bridge was primarily
governed by availability of materials and loading facilities. Consid
eration of dependability of results, costs, and construction times were
additional important factors.
The one-sixth scale factor was dictated by availability of model
materials (mainly prestressing cable). 14 At this scale, all prestressing
strands and reinforcing bars could be modeled very closely with a maximum
9 percent deviation. Although deformed bars were not available for these
small sizes, bonding of the rebar wires was not a critical problem in
this study. After tentative selection of the one-sixth scale factor,
testing facility and loading equipment availability were not found to be
controlling factors.
2.2 Dimensions of Prototype and Model Bridge
2.2.1 Longitudinal Dimensions. The 200 ft. main span and
balancing 100 ft. side spans of the prototype bridge were modeled in
one-sixth scale, as shown in Fig. 2.1.
2.2.2 Transverse Dimensions. Details of the transverse cross
section are shown in Fig. 2.2.
2.3 Choice of Materials
2.3.1 Concrete. At The University of Texas at Austin, a 1/5.5
scale microconcrete24
(reduced aggregate which has a gradation similar to
ordinary aggregate gradation) has been used for a number of model studies.
This microconcrete is designed to closely match the critical properties
(Ec and tc ) of the prototype concrete. It was decided to use this type
9
10
A 8 C D A~------~&-----------------A---------b
SPAN A8 8C CD
PROTOTYPE 100' 200' 100'
MODEL 16.67' 33.33' 16.67' i
Fig. 2.1. Longitudinal dimensions of bridge
8
81 85 87 CAST IN PLACE STRIP -1'~
T~b:=d~~F=~~~~~~~~~~==d T
82 ~
T3', AT MAX. SECTION ~'1"-JV1 FILLET DETAIL
B 81 82 83 85 87 T1 T2 T3 T3' T4 T6 0 Hi Vi H2 V2 H3
PROTO- 1671' " 1\ 1\ II 1\ II 1/ II II /I 1\ II " 1\ II /I " TYPE 171.5 14 156 80 24 8 7 6 10 12 6 96 8 6 8 6 8
" " II " /I II 1/ 1/ " " " " 1/ 1/ 1/ /I II /I
MODEL 112 11.9 2.3 26 13.3 4 1.33 LI7 1 1.61 2 I 16 I~ I 1.3a I 1.3~
Fig. 2.2. Cross section of bridge
V3 H4 V4
1/
6 4" 4"
1/ .. II
I P.67 ~.67
concrete after trial mixes indicated it possible to attain the minimum
f' = 6000 psi. Details of the scaled microconcrete are described in c
Sec. 3.1.
11
2.3.2 Prestressing Steel. Four groups of strands were used in
the original design of the prototype. Minor changes were made in the
tendon groups actually used in the prototype consturction. The scale
reduction for the model tendons is shown in Table 2.1. All tendons were
scaled down to match the ultimate strength requirements. Therefore, the
type and grades of steel for the prototype and the model were not always
the same. Figure 2.3 shows the position of each tendon.
2.3.3 Reinforcing Bars. Three different reinforcing bar sizes
were used in each segment and another bar size was used for the longi
tudinal cast-in-place joint. These bars were specified as Grade 60 in
the prototype. The reduction to one-sixth scale for these bars is
described in Table 2.2. The reinforcing bars used for the model were
C-1018 wire. These wires were not deformed but the bond strength of wire
to microconcrete has been shown to be adequate in other tests. l In
order to obtain a distinct yield plateau in the wires to match the rein
forcing bars of the prototype, wires for the cages were annealed by a
comrnerical firm in a controlled temperature oven. Then the reinforcing
cages were assembled. Details of the reinforcement are shown in Fig. 2.4.
All details matched those of the prototype units.
Details of individual bars and some modifications (such as spirals
and anchorages) are shown in Sec. 4.1.
12
TABLE 2.1. CHOICE OF STRANDS AND WIRE FOR THE MODEL BRIDGE
Prototype Model
No. of 1/2" <P Ultimate Required
Selected Wires Strength of
270K Strands Strength Ultimate or Strands Selected
(kip) Strength (kip) Material (kip)
20 825 22.9 3/8 in. Strands 23.0
13 537 14.9 7 mm Wire 14.7
8 329 9.1 1/4 in. Strands 9.0
6 248 6.9 6 gao Wire 7.2
TABLE 2.2. CHOICE OF REINFORCING BARS FOR THE MODEL BRIDGE
Prototype Model
Bar Area Required Area Chosen Wire Area of Chosen (sq. in.) (sq. in.) Wire (sq. in.)
118 0.79 0.0220 8 gao Wire 0.0206
116 0.44 0.0122 1/8 in. Wire 0.0122
115 0.31 0.0086 12 gao Wire 0.0087
114 0.20 0.0056 14 gao Wire 0.0050
SIDE SPAN MAIN SPAN
PROTOTYPE" MODEL PROTOTYPE MODEL
81 13 STRANDS 7 MM. WIRE A1 8 STRANDS Y4 IN. STRANDS 82 13 STRANDS 7 MM. WIRE
83 20 STRANDS :Va IN. STRANDS A2 8 STRANDS ¥4 IN. STRANDS A3 8 STRANDS Y4 IN. STRANDS
84 20 STRANDS 3fa IN STRANDS A4 8 STRANDS Y4 IN. STRANDS 85 13 STRANDS 7 MM. WIRE A5 8 STRANDS Y4 IN. STRANDS 86 13 STRANDS 7 MM. WIRE A6 8 STRANDS V4 IN. STRANDS 87 13 STRANDS 7 MM. WIRE 88 6 STRANDS 6 GA. WIRE C1 6 STRANDS 6 GA. WIRE 89 6 STRANDS 6 GA. WIRE C2 6 STRANDS 6 GA. WIRE 810 6 STRANDS 6 GA. WIRE C3 6 STRANDS 6 GA. WIRE 811 6 STRANDS 6 GA. WIRE C4 6 STRANDS 6 GA. WIRE
" AS ORIGINALLY DESIGNED
Fig. 2.3. Details of prestressing strands and wire
14
I
I- 8 SP. AT 02
t.B RIDGE
It) a t-oe(
0: (/)
U)
REINFORCE~
R1 R2
R3
R4 R5 RS R7
R8
R9 RiO
R11 R12
{
-~ \fl· ~
1\ "
I· ,.
I .. I
RI2 ~
~
I
32 SP. AT 01
I 19 SP. AT 02
I I
I I
I
7 SP. AT 03
8 SP. AT 03 15 SP, AT 04
'\
15 SP. AT 04
~R1
R2' .~7 V1 8 SP. AT 02
R3l ~R3
jR5 ) ,
~ -I "'R4 -I f RS
r""'-R7
~ .....r-. R8
-~b I.-si -f- R11
- f-h '1 1
'" ~ R9 R10
PROTOTYPE MODEL SPACING PROTOTYPE MODEL
#8 8 GA. WIRE #8 8 GA. WIRE #8 8 GA. WIRE #4 14 GA. WIRE #4 14 GA. WIRE
01 10 IN. I.S7 IN. 02 7 IN. 1.17 IN. 03 18 IN. 3 IN.
04 7Y2 1N. 1.25 IN.
05 12 IN. 2 IN.
#4 14 GA. WIRE #S va IN. WIRE #4 14 GA. WIRE
*4 14 GA. WIRE *4 14 GA. WIRE :11:4 14 GA. WIRE #3 IS GA. WIRE
Fig. 2.4. Details of reinforcing bars in segment
J I
C HAP T E R 3
MATERIAL PROPERTIES
3.1 Concre te
3.1.1 Concrete Mix Design. After several trial mixes, the mix
of Table 3.1 was used for segment casting. Consistency of the micro
concrete was judged by visual inspection. The workability was good even
though the mixes appeared somewhat dry and harsh.
TABLE 3.1. CONCRETE MIX DESIGN
(2.4 cu. ft.)
Design strength
W/C (lb/lb)
Water
Cement (Type III portland cement)
TCM 154 (Aggregate)
Ottawa Sand
Blast sand No. 1
Blast sand No. 2
Colorado River red sand
Airsene L (Retarder)
6000 psi
51.6%
41.5 lbs.
80.5 lbs.
61.2 lbs.
70.9 Ibs.
65.4 lbs.
18.9 lbs.
18.9 lbs.
94.8 cc
3.1.2 Strength of Concrete. One or two cylinders (3 X 6 in. or
2 X 4 in.) were cast with each segment. Several cylinders were tested
at the time of erection and the remainder were tested at the time of
loading tests. Compressive strength of the closure segment was tested
one week after casting but prior to positive tendon operation. Test
results are summarized in Table 3.2 and Table 3.3.
15
16
TABLE 3.2. COMPRESSIVE STRENGTH OF CONCRETE
Type of No. of Cylinders Average Compressive Standard Deviation (8) of Compressive
Cylinder Tested Strength (psi) Strength (psi)
3 x 6 in. 49 7090 523
2 x 4 in. 24 7430 810
TABLE 3.3. SPLITTING TENSILE STRENGTH OF CONCRETE
Average Tensile Standard Deviation
Type of No. of Cylinders (8) of Tensile Cylinder Tested Strength (psi) Strength (psi)
3 x 6 in. 35 597 31
2 x 4 in. 20 633 71
A universal hydraulic testing machine (max. 120 kip) was used for
compression and splitting tensile tests of cylinders. Scaled loading
heads and caps were necessary to minimize scatter of test results. l ,3l
For the compression test, cylinders were capped on both ends with sulphur
capping compound in a scaled capping device and tested by using scaled
adjustable spherical load heads on the top of cylinders. For the splitting
tensile test, loading bars and wooden strips were scaled down for each
size of cylinder. Also, the loading rate was reduced by the appropriate
scale.
No cylinder tested had a compressive strength lower than the
specified 6000 psi. The strengths of the 2 x 4 in. cylinders were slightly
greater and showed higher variations than the 3 x 6 in. cylinders. The
values obtained with the 3 x 6 in. cylinders are more reliable than those
17
of the 2 x. 4 in. cylinders because the standail deviation for the 3 X 6 in.
cylinders is much less. than for the 2 X 4 in. cylinders. Therefore, the
values for f"c of 7090 psi and spli tting tensile strength of 597 psi are
used in later calculations. F for the 3 X 6 in. cylinders is sp 597/J7090 ~ 7.09.
3.1.3 Modulus of Elasticity. Four 0.4 in. paper strain gages
were mounted vertically in series at the middle of several cylinders.
The cylinders were preloaded to about l/lOth of the estimated failure
load two or three times prior to taking any strain reading. The strain
readings of all four gages were averaged and stress-strain curves were
drawn to find E. E was determined by the slope of the chord up to c c 13
about 0.5f' (the secant modulus of elasticity). Eight cylinders c (3 X 6 in.), including one from the closure, were tested and the results
are shown in Table 3.4. Although the value obtained from the 3 X 6 in.
cylinders is used in all calculations, Ec of the 2 X 4 in. cylinders was
also checked by the same procedure.
TABLE 3.4. MODULUS OF ELASTICITY OF CONCRETE
Type of No. of Average E Standard Deviation Average f' Cylinders (psi) C
(8) of E C Cylinder Tested (psi)
C (psi)
3 x 6 in. S 4.56 x 106 0.37 x 106 7550
2 x 4 in. 2 4.46x106 0.24 x 10
6 7190
The ACI2 formula gives E ~ wl . 5 X 3~ . c c
Since the unit weight
of microconcrete is 133 lb/cu.ft.,1,24 E = 1331 . 5 c X J7 550 == 4. 39 X 106 ps i
for t ~ 7550 psi. This is in close agreement with the measured values. c
3.1.4 Poisson's Ratio. Six 0.4 in. paper strain gages were
mounted vertically and horizontally in series at the middle of several
test cylinders. Three vertical strain readings and three horizontal
strain readings were averaged and Poisson's ratio was calculated. Three
18
3 x 6 in. cylinders were tested to find Poisson's ratio. The average
Poisson's ratio was 0.184 and standard deviation was 0.016.
3.2 Steel
3.2.1 Prestressing Strands and Wire
3.2.1.1 Ultimate Strength. Six to seven specimens of each size
were tested. Results are given in Table 3.5.
TABLE 3.5. ULTIMATE STRENGTH OF PRESTRESSING STRANDS AND WIRE
Type of Cable No. of Average F' Area Average f' and Specimens (kip) s (sq. in.)
s (8) (ksi)
6 gao Wire 6 8.22 0.029 280 (13)
1/4 in. Strands 6 9.15 0.0356 257 (8)
7 nun Wire 6 15.3 0.0594 258 (8)
3/8 in. Strands 7 22.0 0.085 259 (4)
3.2.1.2 Modulus of Elasticity. Two strain gages were mounted
on opposite faces of the wire and E was calculated from stress-strain s curves. In an attempt to measure the modulus of strands, epoxy resin
coatings were applied to get a smooth surface for strain gages. Two
strain gages were then mounted using the same procedure as that for the
wire. Since these strains were not measured successfully, values for
strands were taken from Ref. 26. Es values are shown in Table 3.6.
3.2.2 Reinforcing Wire. Samples of all reinforcing wire sizes
used in the segments were tested, as shown in Fig. 3.1, without breaking
the spot-welded cross wires. Some dif ference in f and f' was noticed y s
between sizes of wires. Test results are shown in Table 3.7. All segment
reinforcing wires had yield points above Grade 60 minimums. The 12 gage
wire which was used in the midstrip closure had f ,,45.1 ksi (8 ,1 3.1) and y
t' = 53.6 ksi (9 = 1.8). s
19
Fig. 3.1. Testing of reinforcing wire
TABLE 3.6. MODULUS OF ELASTICITY OF WIRE AND STRANDS
Type of Cable No. of Specimens Average E and (8) (psi) s
6 gao Wire 4 30.9 )( 106 (0.11 )( 10
6)
1/4 in. Strands 27.0 )( 106 ( )
7 !DID. Wire 5 30.5 )( 106 (0.17 x 106
)
3/8 in. Strands 27.0 x 106
( )
TABLE 3.7. TEST RESULTS OF REINFORCING WIRE OF PRECAST SEGMENTS
Type of Wire No. of Specimens Average f and Average f' and - y _ s
(6) (ksi) (5) (ksi)
8 gao Wire 6 70.6 (3.4) 75.1 (4.2)
1/8 in. Wire 6 79.0 (4.3) 80.4 (3.0)
14 gao Wire 15 72.2 (12.1) 74.6 (11.4)
20
3.2.3 Strength of Bolts at Main Piers. Texas Highway Department
plans specify that twelve 3-in. diameter threaded rods were to be used for
moment connections at the main piers during construction. For the model
bridge, twelve 1/2-in. diameter high strength bolts were used at the main
piers to hold segments temporarily during construction. Test results of
those bolts are shown in Table 3.8. Experimental fl was in the range s
(60-100 ksi) specified by ASTM for the grade bolt originally called for
in the bridge plans. Subsequent to the model test, the grade of the bolt
material specified was changed to obtain a higher f in the prototype. y
TABLE 3.8. TENSILE STRENGTH OF BOLTS AT MAIN PIERS
Type of Bolt
1/2-in. diameter
No. of Specimens
3
3.3 Epoxy Resin as a Jointing Material
Average f and (s) (ksi)
75.1 (0.85)
Average ~ ~ and (s) (ks1)
78.8 (0.9)
3.3.1 General. The advantage of the epoxy resin as a jointing
material as compared to concrete or mortar joints is its quick hardening.
The epoxy should seal the joint against corrosion and the epoxy joint
should be as strong as concrete in flexure and shear. The epoxy resin
joints should increase the cracking moment at joints or shift the cracks
to some other position under overloading.
A complete report on the epoxy evaluation program has been pre
sented in Ref. 19. The epoxy used in construction of the model was
Epoxy E in that report. It had flexural and shear strengths when jointing
hardened concrete, as shown in Tables 3.9 and 3.10.
Epoxy Resin
(E)
Epoxy Resin
(E)
21
TABLE 3.9. EPOXY E FLEXURE TEST RESULTS
Flexural Percentage of
Condi tions of Strength Type of Failure Strength of
Specimen (psi)
Monoli thic SEecimen (%)
Dry, No Oil 733 Concrete Adjacent 100 to Joint
Dry, Oil 742 Concrete 102
Sa tura ted, No Oil 467 Concrete Adjacent 64 to Joint
Compressive Strength of Cylinder: 6760 psi(s = 842)
Splitting Tensile Strength: 538 psi(s '" 73)
Flexural Strength of Monolithic Specimen: 729 psi(s 43)
TABLE 3.10. EPOXY E SHEAR TEST RESULTS
Shear Strength
(psi) Type of Failure
542 Epoxy Separation
504 Concrete Adjacent to Joint
666 Concrete Adjacent to Joint
Average Shear Streng th and
(8) (psi)
571 (69)
Compressive Strength of Cylinder: 6760
Splitting Tensile Strength: 538
Shear Strength of Monolithic Specimen: 753
Percentage of Strength of Monolithic
Specimen (%)
76
psi(s 842)
psi(8 73)
psi(; 103)
!!!!!!!!!!!!!!!!!!!"#$%!&'()!*)&+',)%!'-!$-.)-.$/-'++0!1+'-2!&'()!$-!.#)!/*$($-'+3!
44!5"6!7$1*'*0!8$($.$9'.$/-!")':!
C HAP T E R 4
SEGMENT DETAILS
4.1 Reinforcing Cage for Precast Segments
4.1.1 Reinforcement. Hand assemblage of each reinforcement cage
with individually tied wires would have required excessive time to complete
84 segments. Therefore, the five different wire mats which were required
for each segment were lightly spot-welded and were then annealed. These
mats were then cut and bent to proper length and shape, as shown in
Fig. 4.1. All mats of a segment were assembled as shown in Fig. 4.2, by
using a jig. Details of each mat are shown in Fig. 4.3. Development
length of web reinforcement spliced into the top slab was longer than
that specified by the prototype plan to compensate for the use of plain
bars in the model instead of deformed bars.
Modification of the cages was required at the shear key and
anchorage points as shown in Fig. 4.4. Whenever a wire was cut, replace
ment bars were spliced in, or wires were rotated or rerouted as necessary.
No bars called for on the plans were ever omitted.
The reinforcement of the pier segments was scaled as shown in
Fig. 4.5.
4.1.2 Anchorages. Since it was impossible to exactly model the
multiple strand commercial tendons and anchorages used in the prototype,
prestressing cables were modeled by selecting for each tendon an equivalent
single cable which could produce the correctly scaled tendon force. Normal
strand chucks were selected as the basic anchorage device for the single
cables. The model was constructed before the prototype bidding and selection
of the contractor. At the time of anchorage selection for the model, the
prototype plans allowed the contractor to choose bearing or wedge types of
anchorage. For the model, the bearing type was chosen.* The dimensions
*The contractor later chose the wedge type for prototype construction.
23
24
...... " .. ~'" /.,8, ~~.,~IRE MAT (1)
14 GA. WIRE
• u ••• ••
/ V8 IN. WIRE • • • ••• • \...-J ....«......«..... '---' ... · . I «
"""'-- e GA. WIRE MAT (2)
.---8 GA. WIRE
MAT (3)
MAT (4)
MAT (5)
Fig. 4.1. Individual mats for a cage
Fig. 4.2. Cage for a segment
8 GA WIRE j' . 14 GA WIRE ; .
14 AT US": 21" 7.25"
8 GA. WIRE '18" WIRE ~ ~-,.
•• 0" 18 AT L2S": 20"
25
=~ UPPER cD REINFORCEMEN T
"
MAT (1)
:
= ~ LOWER !i REINFORCEMENT II
= ~ MAT (2)
: I'·~I. 1.0" .1. 8.5".1. 21.0" .1. 7.25~. 9.0" • Ha!" -Ir- • • · · · · · · . · · · · · · · · · · _UPPER REINFORCEMENT -2L -_ ..................... ""'\.-...I ••••••••• «....... ,---,'" ...... LOWER REINFORCE MENT
14d17.5H Util 200" IILJ!1 7,5" LJ
eO 1-' .F1 -t . · '''M i· · p~ aIS" 1.7"" 2.t.s" OS" lo"
REINFORCEMENT IN TOP SLAB
Fig. 4.3(a). Details of reinforcing mats
26 8 GA. WIRE j
14 GA.WIRE
~I
ll, 7 AT U", ,7..' -1--1 , " \ 318 1~8'
WEB REINFORCEMENT
14 GA.WIRE , j
Iyl. 8 AT 2.75":: 22.0" I,~ 1.7S" 1.75"
..
i ' ~" M~ ~ C: . : . : :: ~.~
8 eA. WIRE
MAT (3)
UPPER REINFORCEMENT
MAT (4)
LOWER RE IN FORCEMENT
MAT (5)
REINFORCEMENT IN BOTTOM SLAB
Fig. 4.3 (b) • ~ontinued)
8 GA. WIRE
8 GA. WIRE
-1----4--1--1.14 GA. WIRE
STEP(t)-ORIGINAL CAGE
)
/
~ 14 GA. WIRE
STEP (2)-CUT SOME RE1NFORCEMENT
8 Gil.. WIRE
8 Gil.. WIRE
14 Gil.. WIRE
8 GA.WIR E-
16 GA. WIRE
1
J
1
• HEAVY LINE INDICATES THE ADDITIONAL REINFORCEMENT
14 GA.WIRE
STEP (3)-ADD REINFORCEMENT" FOR SHEAR KEY AND ANCHORAGE
14 GA. WIRE 16 GA. WIRE H---l--I-"'I
----.=1 . 8 GA. WIRE
\--+-+--t--"!14 GA. WIRE
I-+~N~::J"",ANCHOR PLATE
CHUCK 800'1'
STEP (4)- FINAL. SET CHUCK BODY AND TENDON DUCT
Fig, 4.4, Details of web reinforcement modification
28
1.83" 19.4"
I
20"
12.5" 3" 9"
-+ ~ rr- 4 I--
-~ r:.67"SLAB 1.67" SLAB ~
~ L - -
TOP VIEW
54.0"
,II 2" 206"
K10 r~ =(0 o
10.7" 1<'10 4"
26"
AT TOP OF UNIT ONLY
3" .2.!i1AT TOP OF UNIT ONLY
I ~
1== H
1" CAST HOLES (TYP)
- ~BARS Kl-6 ... 1-1.67" SLAB
~ BARS K7
t::::: ~
I ... ~ 0.25" COVER (TYP)
..., n II
. ~K9 --', J/K1 -10
I i f--K2
11.3"
TYPE OF REINFORCING BAR
K',K2,K3
K4,K5
K6,K8
K7, K9 ,K10
1/8 IN. WIRE
12 GA. WIRE
Fig. 4.5. Reinforcement for the main pier segment model
n
29
of the anchorage bearing plates shown in Table 4.1 were sized to satisfy
the then current ACI 2 and AASH06 specifications in a preparatory study.14
The commercial anchorage chuck was welded to the bearing plate as shown
in Fig. 4.6. When double tendons in each web were used in the first
segments adjacent to the main piers, the area of the bearing plate was
doubled and both commercial anchorages were welded to the same plate.
TABLE 4.1. DIMENSIONS OF ANCHORAGE BEARING PLATES
Cable Bearing Plate Bearing Stress Allowable
(in. x in. x in.) at 0.8f' (psi) f (psi) s cp
3/8 in. Strands 1-1/2 x 1/4 x 3 4280 4390
7 mm Wire 1-1/2 x 1/4 x 2 4200 4410
1/4 in. Strands 1-1/2 x 1/4 x 1-1/2 3370 4390
6 gao Wire 1-1/2 x 1/4 x 1-1/2 2700 4390
14 During preliminary tests, a tendency for splitting along the
tendon was observed, which was restricted by use of a spiral. Even
though the bearing plates for the model were conservatively sized and
the tendon curvature was not as great as in the preliminary tests, the
use of the spiral shown in Fig. 4.6 was adopted as a good detailing
practice. Each duct and its spiral were always completely contained
inside the reinforcement cages.
4.1.3 Tendon Duct. Polyethelene tubes were used to form the
curved ducts in the segments containing the duct anchorages. Steel tubes
were used as duct formers in the other straight portions, as shown in
Fig. 4.7. Pipes for grouting access were welded to the tendon ducts near
the anchorage.
Two different sizes of ducts were used. 7/l6-in. diameter (1.0.)
was used for the 7 mm wire, 1/4-in. strand and 6 gao wire tendons, while
9/l6-in. diameter (1.0.) was used for the 3/8-in. strand tendons.
30
SPIRAL CD
SPIRAL ®
BEARING PLATE
* (COMMERCIAL II SUPERIOR II
SINGLE USE)
I RAL ® SPIRAL CD
r=:z=_=::::::==~...L yi' T
~I~8' -lll- 13 GA. WIRE
3/a'
13 GA. WIRE
POLYETHYLENE TUBE
Fig. 4.6. Details of anchorage and spiral reinforcement
31
STEEL TUBE POLYETHELENE TUBE
NEGATIVE TENDON GROUTING HOLE
ANCHORAGE
POLYETHElENE TUBE
POSITIVE TENDON
Fig. 4.7. Typical arrangement of tendon duct near anchorage
Figures 4.8 and 4.9 show the profile of the tendon ducts in the
main and side spans while Fig. 4.10 shows the detailed profiles of tendon
ducts in the curved portions.
4.2 Forms
The accuracy of vertical and horizontal alignment of the segments
and the need for segments to fit exactly are key considerations dictating
well-controlled match casting. To simplify control. continuous soffit
forms were used for the base form and segments were match cast against the
segment to which they were going to be joined at the time of construction.
Two continuous soffit forms (half the length of the total span of the
bridge) were mounted on the test floor, as shown in Fig. 4.11. These
soffits were built straight and level within 1/16-in. accuracy.
A_
81.", -183
MP M1 M2 A....-
Mt M2
MAIN PIER
A ,
.... 84 -. 8S ... 86 -::!
I
M3 M4 ' M5 M6
PLAN
M3 M4 M5 M6
ELEVATION
WEB
SECTION A-A
WEB
6RADUAL TRANSI TION OF CURVATURE
B........-
I ---tae---. 87 - 89 ..-t - ----...-
M7 Me M9 B -+-
M7 Me M9
SECTION B-B
A2 BOTTOM SLAB
810 811 ... ,..
MIO
CLOSURE I
MIO
Fig. 4.8. Tendon duct profiles in main span
GRADUAL TRANSITION OF CURVATURE
~
~ ,
~ 810 ~~ L B1 ~ 86" as 84 ... 83 III""'" Bt.B2 r-----
~ SIO S9 S8 S7 S6 S5 , S4 S3 S2 S1
PLAN
SIO C4 CI C2 Ct
S9 S8 S7 S6 S5 S4 S3 S2 S1
ELEVATION
MAIN PIER
Fig. 4.9. Tendon duct profiles in side span
34
IS.O ,.
5.62"
-:-:::
16.d'
-'L---
TENDON BI 83-B10 15.4"
TENDON B2 14.8"
20"
4.38" 5"
...---
IS.O •
20"
5" 5"
.... ~ r-- .... , "". "'~"
14.9" 13.5"
13.1" 11.1"
4.38"
~'"
"" ILl"
9.10"
..........
GENERAL NOTE: DIMENSION AT THE BOTTOM INDICATES DISTANCE FROM BOTTOM OF PRECAST UNIT TO <t.. OF TENDON •
"-TENDON 81 , B3-B10
TENDON 82
8.7
6.7
7" 7"
20"
5" 5.62" 4.38·~, 5" _1 10.62"
110.27,,1 1 1~
V·"""" '1
,/ TENDON C3. C4 a 811
..... ~ . -'""'--
!7" 7" 7.21" 8.21" 10.5" 12.9 15.5"
2 " 0 20 " 20 " 10.S" 5" 4.4" 5.&2" 5" 5" 4.38" 5.62" 5" 9.38"
'- .......
~.,
~ ~., "
.............. I"- - • -- . .
C1,C2 TENDON Ai A2 15.3" 12.S" 10.3" 7.61" 5.19" 3.2'" 2.15" LZI5" I" ," TENDON A3 A4 15.8" 13.1" 10.8" 8.11" 5.69" 1).19" 2.65" 1.75" U~ .. 1.5"
TENDON AS A6 13.6" 11.3" 8.67" S.19" 4.Z'" 3.15" 2.25" 2" 2"
Fig. 4.10. Details of tendon profiles
Fig. 4.11. General view of soffit forms
PLYWOOD
FORM~
TESTING FLOOR
FORM CD - PLEXIGLAS
FORM ~ .. PLYWOOD FORM" .. PLYWOOD
Fig. 4.12. Details of forms for segment
35
36
Three materials (steel, plywood, and plexiglas) were considered
for segment forms during the planning stage. Steel was abandoned because
of high estimated costs. After several trial castings using plexiglas
and plywood, stiffened plexiglas (1/4 in. thick) and plywood (1/2 in.
thick) were used in combination, as shown in Fig. 4.12. Plywood costs
were lowest but the use of plexiglas was desirable for either the interior
or exterior side forms in order to observe the flow of concrete in the
webs so as to minimize honeycombing in the congested areas around the
anchorages. Plywood surfaces were coated with lacquer to simplify
stripping and protect the surfaces for repeated use. The end bulkheads
were made of plywood faced with plexiglas. These forms were set at the
end of the segment and secured rigidly to the side forms with bolts.
Figure 4.13 is a general view of the assembled forms.
All forms were stripped one day after casting. End bulkhead and
outer sideforms were stripped first and then the inner top form [Form (3)
in Fig. 4.12] was stripped by folding the joint at the center. Then the
interior side forms were stripped. These forms were moved ahead as each
segment was cast against the previous segment. Most segments were not
separated until all segments on the soffit were cast in order to simplify
maintenance of overall tolerances. A few of the initial segments cast
were separated early to check the adequacy of the bond breaking compound.
4.3 Strain Instrumentation in Segment Reinforcing Cages
Since the main purpose of this study was to document the behavior
of critical sections during various stages of construction as well as
during the loading tests, strain gages were primarily placed in the first
segments next to the main piers and at the center of the main span.
Although it was possible to mount strain gages directly on the
wires of the cage, because of the difficulty of protecting the gages
during cage preparation and handling it was decided to mount strain gages
on "extensometers" consisting of separa te 10 in. lengths of O.l-in.
diameter wire and to connect these to the cages at certain positions.
37
Fig. 4.13. General view of assembled forms
The bars with strain gages were positioned longitudinally underneath the
top mat in the upper slab and on top of the bottom mat in the lower slab.
Strain gages could not be successfully placed on the transverse bars of
the cage to measure transverse behavior in the segments due to the thin
cover and difficulty of casting concrete to proper thickness. Transverse
paper gages were applied to the concrete surface of the segments at
required positions before the loading tests.
Four types of gage installation patterns were used with the
segments. The position of strain gages for each patter~ and the pattern
used in each particular segment location is shown in Fig. 4.14.
4.4 Casting Procedure
Usually four segments were cast per day, with two segments cast
on each continuous soffit form. The order of casting on the continuous
soffit form is shown in Fig. 4.15.
38
• \ : J PATTERN A
• \ : 1 PATTERNB
\ : 1 PATTERN C
1 : I PATTERN 0
t BRIDGE
.1 I
I t BRIDGE
I
t BRIDGE
I I
I t BRIDGE
I I
B -----
B ---..
.~ o A
C
o A
--....
::::::::
x-- POSITION Of GAGE :----+ C
JOINT JOINT
-MAIN PIER
{I A,B,C
:-
1-
C
B
c -
NSIOR NSIOL NS9R NS9L
NS8R NS8L NS7R NS7L NS6R NS6L 4--
NS5R NS5L
NS4R NS4L
NS3R NS3L NS2R NS2L NS1R NS1L +--
NMP (W) NMP (El NM1L NM1R I::::-NM2L NM2R NM3L NM3R NM4L NM4R
NM5L NM5R
NM6L NM6R '+-NM7L NM7R NM8L NM8R
NM9L NM9R +--NMIOL NMIOR CLOSURE CLOSURE SMIOR SMIOL f+-SM9R SM9L SM8R SM8L SM7R SM7L
SM6R SM6L
SM5R SM5L SM4R SM4L
SM3R SM3L
SM2R SM2L
SM1R SM1L t'4-SMP (W) SMP (E)
SS1L SS1R ~ SS2L SS2R
SS3L SS3R
SS4L SS4R SS5L SS~R
SS6L SS6R
-.S..l7L SS7R ~
SS8L SS8R SS9L SS9R
~IOL SSIOR
Fig. 4.14. Position of strain gages in segments
z ~ (I')
C
c
C
B
B
C
B
C
39
810 89 88 87 86 85 84 83 82 81 MP M1 M2 M3 M4 M5 M6 M7 M8 M9 M10
110 19 I 81 7 I 6 I 5 I 4 I 3 I 2 11 11111 12 I 3 14 I 5 I 6 I 7 I 8 [ 9 110 I
Fig. 4.15. 8equence of casting segments
For the first segment, the two bulkheads were set at carefully
determined distances. Subsequently the forms were moved ahead, and each
segment was cast against the previously cast segment.
The preparation for casting segments after the first segments
was as follows:
(1) Remove all forms.
(2) Clean the surface of the previous segment and grind off the excess tendon duct. Seal the anchorage holes of previous segment with styrofoam.
(3) Clean all parts of forms and put on coating of lacquer.
(4) Put bond breaking compound on the outer face of the hardened segment one day before the next casting.
(5) Set reinforcing cage on the soffit form adjacent to the previous segments.
(6) Set chucks and polyethelene duct in curved portion, and set steel duct in straight portion.
(7) Set the end bulkhead into position and adjust all tendons in the proper holes of the bulkhead.
(8) Put caulking compound along the joints of forms and set both inner and outser side forms into position. Position the interior form for the top slab. Bolt all side forms to the bulkhead and soffit forms.
(9) Place polyethelene blockouts for pass through openings for attachment of dead load blocks and for web openings for use in separation and temporary prestressing operations.
(10) Install bracing for interior side forms.
All web concrete was placed through the top of the web and the side
form was filled from bottom to top and consolidated to eliminate any honey
comb effect. A vibrator was used carefully for this consolidation while
40
the web was observed through the interior plexiglas form. Although many
ways of vibrating the web concrete were tried, it was concluded that the
most effective way was to vibrate directly on the cage of the top slab
over the web. Before casting the web portion, blocks were set between
the webs and the bottom slab to prevent the flow of concrete from the web
portion to the bottom slab. The bottom slab was cast after completing
the casting of the webs and then the blocks between the webs and bottom
slab were removed. The top slab was cast last. After casting the entire
portion, the top slab was coated with a membrance coating compound to
prevent shrinkage cracks.
After completing all casting of 21 segments on a soffit form, each
segment was separated by using four small rams. Figure 4.16 shows the
position of rams. Although it would be possible to develop the forces to
separate the segments by hand, usage of rams to equalize forces minimized
local concrete damage at the shear keys on the webs and the guide keys of
the top and bottom slabs. In most cases, the force actually applied on
each ram was less than 0.4 kips .
. " :.
4 TON RAM
Fig. 4.16. Position of rams for separation of segment
4.5 Curing of Segments
Because the sequence used in casting and construction was the same,
all segments except the pier segments were erected five to six months after
41
casting. The segments were air cured at 75 ±30
F until they were steam
cleaned.
4.6 Surface Preparation of Segments
Surface preparation of the joint faces is necessary to ensure a
good epoxy joint between segments. The order of surface preparation was
as follows:
(1) The segments were checked to make sure that the tendon ducts did not extend beyond the joining surface of the segments.
(2) Excess concrete and sealing material were taken out using a hand grinder (No.4 in Fig. 4.17).
(3) Light grinding of the joint faces was accomplished using a small grinder connected to an electric drill (No.3 in Fig. 4.17) in lieu of sandblasting.
(4) If the stressing jack seating attachment would not set on the chuck body smoothly, the holes for the seating attachment at the anchorages were enlarged by drilling, using No.5 shown in Fig. 4.17.
(5) Chucks were cleaned using a wire brush connected to an electric drill (No.2 in Fig. 4.17).
(6) Segments were steam cleaned and rust was removed from chucks and steel ducts. Steam cleaning was used because it was found that form oil had been used in error on the bulkhead before casting for a few of the segments. The effect of oil was eliminated by steam cleaning, as reported in Ref. 19.
4.7 Compensation Dead Load for Model Bridge
To satisfy similitude requirements and to obtain the same dead
load stress conditions as the prototype bridge, it is necessary that the
density of a one-sixth scale model material be six times that of the
prototype bridge. This is impractical to implement.
Compensating dead loads have been added in various ways for model 1 12 structures.' In this case, five times the weight of the model segment
was added to the segments using concrete blocks, as shown in Fig. 4.18.
All dead load blocks were distributed to represent the weight of
each portion and to give reasonable transverse as well as longitudinal
42
Fig. 4.17. Tools used for surface preparation
CONCRET~
BLOCK
Fig. 4.18. Compensating dead loads for the model bridge
43
distribution. Two 140 lb. blocks were used for the top slab cantilever
portions, a 310 lb. block was used for the interior top slab, and two
355 lb. blocks compensated for the webs and bottom slab, as shown in
Fig. 4.18.
Four points were used to support the 310 and 355 lb. dead loads
and two points were used for the 140 lb. blocks, since the additional
dead weight could not be placed uniformly. The maximum weight on any
loading point was less than 90 1bs. This distribution closely simulates
the uniform dead load.
Because of the heavier weight of the thicker slabs in the segments
adjacent to the main piers, it was theoretically necessary to hang 70 1bs.
more weight on them. However, this minor effect was ignored and the same
compensating dead weights were hung uniformly on all segments.
4.8 Actual Properties of the Model Bridge
All experimental material properties and measured as built
section properties of the model bridge are summarized in Table 4.2. Those
values were used in the theoretical calculations.
TABLE 4.2. MATERIAL PROPERTIES AND SECTION PROPERTIES OF THE MODEL BRIDGE
HALF SECTION FULL SECTION MAX. SECT. MIN. SECT. MAX. SECT. MIN. SECT.
AREA 2
(IN. ) 179 165 363 335
SECTION PROPERTIES DIST. FROM TOP TO CENTROID (IN.) 6.86 6.20 6.77 6.12
OF BOX SECTION SECOND MOMENT OF AREA (If~~) 6970 6060 14100 12300 SECTION MODULUS AT TOP (lN~ 1020 977 2090 20~0
SECTION MODULUS AT BOT. (lN~) 752 611 1510 1250
EC = 4.56 x 106 PSI (i = 0.37 x 106 PSI ) I
fc = 7090 PSI (i = 523 PSI) PROPERTIES OF I
CONCRETE fsp = 597 PSI (i = 31 PSI)
f = 0.184 (; = 0.016)
REQUIRED !EXPERIMNTL. A EXPERIMNTL. Es 0.6 fiat (KIPS) fiat (KIPS) Itt/., f~ (KSI) (PSI)
PROPERTIES OF PRESTRESSING 3/8 IN. STRANDS 13.75 22 •• 0.015 251 27.0 X 106
CABLES 7 MM. WIRE 8.94 15.3 0.05" 258 30.5 X 106
1/4 IN. STRANDS 5.46 9.IB 0.0356 257 27.0 X 106
6 GA. WIRE 4.13 1.11 O.OR. 280 30.1 X 101
C HAP T E R 5
INSTRUMENTATION AND DATA REDUCTION
5.1 General
Load cells, pressure gages, strain gages, surveyor's level, and
dial gages comprised the instrumentation used for the model study.
5.2 Load Cells and Pressure Gages
Prestressing forces during construction and live loads for the
completed structure were applied by hydraulic rams. Strain indicators
connected to load cells were the primary control for forces applied.
Calibrated hydraulic pressure gages were used for checks of the applied
forces.
5.3 Strain Gage Instrumentation and Data Reduction
Two types of strain gages were used for the model bridge: Foil
strain gages were used on the steel wires and paper gages were used on
the surface of the concrete (see Sec. 4.3).
One-quarter in. long foil strain gages were mounted on 0.1 in.
diameter ~ 10 in. long steel wires, as shown in Fig. 5.1(a). These strain
gages were waterproofed and rarely failed even if they were kept in a
segment more than one year. Eight-tenths in. long paper gages were put
on the smooth surface of the concrete prior to loading tests [Fig. 5.1(b)].
Two short wires were connected to the strain gages in the segment. Then
three conductor wires were connected to two conductor wires, as shown in
Fig. 5.1(c). Because of the use of three conductors wires, the effect of
temperature change in the long 1eadwires was eliminated. These three
conductor wires were connected to the VIDAR digital data acquisition 37 system. The VIDAR system is able to scan rapidly and autorange with
45
46
SOFT RUBBER
L ~ ~IO IN. WIRE ~IO'N'T .
LEAD WIRE LEAD WIRE
GAGE
10 I CONCRETE
(A) FOI L STRAIN GAGE ON WI R E (B) PAPER GAGE ON CONCRETE
STRAIN GAGE
(B)
WIRE (A),(B) - TWO CONDUCTOR WIRE WIRE (C) ,(0),8 (E) - THREE CONDUCTOR WIRE
(D)
(E)
__ ... CONNECT TO VIDAR SYSTEM
RESISTOR (PROVIDES COMPENSATING ARM OF BRIDGE)
(C) CONNECTION BETWEEN TWO AND THREE CONDUCTOR WIRE
Fig. 5.1. Strain gage instrumentation
1 microvolt resolution. Output from the system can be either teletype,
punch paper tape, or magnetic tape. Figure 5.2 shows the VIDAR system
and teletype. Data outputs were put into permanent file in the computa
tion center and were reduced by the data reduction program SPEED. 8
5.4 Surveyor's Level and Dial Gages
A surveyor's level and dial gages were used to measure deflections
during construction and loading tests. Accuracy of the level readings was
0.01 in. and of dial gages within 0.0005 in.
47
Fi g . 5. 2. vrDAR system and teletype
•
!!!!!!!!!!!!!!!!!!!"#$%!&'()!*)&+',)%!'-!$-.)-.$/-'++0!1+'-2!&'()!$-!.#)!/*$($-'+3!
44!5"6!7$1*'*0!8$($.$9'.$/-!")':!
C HAP T E R 6
CONSTRUCTION OPERATIONS AND OBSERVATIONS
6.1 Construction Procedure
6.1.1 General. The following is an outline of the steps
followed in the model bridge erection and closure:
(1) Pier segments were temporarily fixed to the piers using bolts
tightened to a predetermined torque.
(2) The precast segments were sequentially erected using the canti
lever construction method with epoxy joints.
(3) The bolts at the pier were temporarily slackened and the vertical
and horizontal alignment was adjusted after completion of the
erection of precast segments 1 through 9. The bolts were then
retorqued.
(4) The outer pier segments (SlO) were erected and the positive moment
tendons in the side spans were prestressed.
(5) The half segments (MlO) in the main span were erected. The longi
tudinal reinforcement extending across the midspan gap from each of
the half segments was joined and the concrete closure segment was
cast.
(6) The positive tendons in the main span were inserted and tensioned
after 7 days of curing of the closure segment. The bolts temporarily
fixing the segments to the main piers were released during the posi
tive tendon stressing operation.
(7) The bridge was lowered to final position on neoprene pads on the
piers.
(8) The correct reactions were jacked into the outer piers.
49
50 , 6.1.2 Construction of Piers. Because the pier height of the
prototype bridge was 90 ft. and complete pier similtude was not a require
ment of the study, the full height of the pier was not scaled in the model.
The reduced pier height for the model bridge was set at 55 in. to allow
adequate space below the model for insertion of compensating dead load
blocks. However, the general cross section of the prototype piers (see
Appendix A) was used at one-sixth scale. Figure 6.1 shows the reinforcing
bars for an inner pier model. The piers were carefully set in correct
position prior to casting of concrete. In order to prevent overturning of
the piers under unba1an~ed loading, I-beams were welded to the pier base
and tied down to the testing floor, as shown in Fig. 6.2.
Fig. 6.1. Reinforcement of the main pier
The completed pier with the bolts for temporary connection of pier
segments and the neoprene pads for final bearing is shown in Fig. 6.3. The
bolts and neoprene pads were scaled down to one-sixth actual size. The
diameter of the bolts in the model was 1/2 in. and the size of the neoprene
pad was 4-1/6 X 7 X 1/4 in. Dimensions of the bearing plates used to
restrain or support the pier segments at the top and bottom faces for each
6-bolt group were 4 X 16 X 1/2 in.
51
Fig. 6.2. Connection of pier to the testing floor
NUT
• YZ" BOLT
NEOPRENE PAD
STEEL PLATE TO SUPPORT PIER ~1I2D':Ia:r:L...., C:::;;:;;:;:,...".. SEGME NT TE MPO RA R I LV
,......k:~ II (4.16xI/2 IN.)
II
" II II U
Fig. 6.3. Detail of support at main pier
52
6.1.3 Erection of Pier Segments. It is very important to set the
initial horizontal and vertical alignment of the pier segments as correctly
as possible in order to minimize adjustments of the two separate cantilever
sections prior to the closure operation. Adjustment of the vertical align
ment was made prior to setting the pier segment by turning the nuts under
the steel plates shown in Fig. 6.3. For erection of the first half of the
bridge, the horizontal alignment of the pier segments was carefully judged
by eye and was later considered inaccurate. In erection of the second half,
the horizontal alignment of each pier segment was carefully adjusted using
a positive reference line formed by a piano wire stretched between each
pier segment. This procedure was far more satisfactory.
The mechanism for anchoring each pier segment consisted of 12-1/2 in.
diameter bolts and 1/2 in. thick steel bearing plates. The pipes which
formed vertical ducts in the pier segments to pass over the 1/2 in. diameter
bolts had a 1 in. inner diameter, so that there was adequate play to allow
adjustment of the horizontal alignment. After initial alignment of the
pier segments, the anchor bolts were tensioned to approximately 1 kip each,
using a torque wrench. The gap between the segment and the neoprene pad on
the main pier was set at 1/4 in. for the first half of the bridge. However,
a gap of 3/4 in. had to be used for the second half to minimize laterverti
cal adjustment in order to match the first half of the bridge. Figure 6.4
shows a general view of the pier segment on the pier
6.1.4 Erection of Segments during the Cantilever Stages.
6.1.4.1 General. Possible erection methods for precast concrete cantilever
construction over water may be classified as follows:30
(1) Segments floated in on barges with barge-mounted crane erection.
(2) Segments floated in on barges but erected using lifting equipment
supported on the previously erected cantilever segments.
(3) Transportation of segments on the structure itself with a large
launching gantry used to place segments.
The cost of erection equipment for method (3) was considered too
expensive and not necessary for the prototype bridge. Erection methods
•
" '_ 6.~. c-."a[ .I ... of ..... Ior .. __ , 00 tho _I. P'o<
"
(1) ..... (1 ) _.0 .... '" .... Iduool .... U . al l o r <1010 pro"""". Sf .. ,
.... "o<t.o<o ho4 t~ •• "OCI to .0. 01"' .. _tb ..... 1H>tt. ... TO .1_1 . . ... In
"'" _.1 ) , 14 .... ooll .... 1><0<1". tho _, •• It, o f .ho _.1 "\"1' .. «-
Uotl, ._to ~ ... 11ft..:! "" •• d .... ",..,ho.d " '.", ••• h""" 'n 'I,. 6. S. Th ... « ..... ",TO ~o t ""r,.rid 011 .' ........ 1 •••• ct"! ,_".
TIo. u"It • • on <.""0« .. '_,nlly b, _<IIon,ool 40.\000 ( .... 6.61 ••
"',UTO Torr.« .... , ... "" ... ro "' .. p............ 'IlI1o _.,," •••• 1 .... , ..
wo. d •• I~'" '0 ., .... , 0., .. ~ •• t tho '01 ..... ,' """'" ..... .. .... oooid .." .ff~ ..... ,<OHr""1I& fo. <o op'pU'" ~, III ... b\ ••. _ • .c.
La. _r " I ..... , 11f .... "",,_, _ ••• 04 *It, ....... , ...... H , h,
. ..... _.ted <>II "' ................. 001 •• MII •••• '''u''u,. (H I . 6 . 1).
til, .. , .... of ...... _ ....... J1U lb •. <ll,' kip. ,. tho p.DtDtn O).
6.1.~.2 ..... U. of , ••••• ' n , """ . lYo "A_M. v« o . ,.oll,
., .. t o<! pu doy. HOW .... , It .... no< dHfi ... u <b u,« r ... _.....,'" p ••
d.y " UII P«'I''' p.op.n,l ........ doq ... ' . I ........ !l.U!!',. AU .... t!""
... ,~ .... d ... 1. th. or .. ...- '" ai_lai .. <1> •• rr ... of ,_."uo" <11.0", •
... I~ .. __ ",u_. T., . ............ ~U,. ... ,1 .. tt.. <1_ o ... 0."""
54
Fig. 6.5. Typical independent c rane used for erection of segments
Fig. 6.6 . Mechan i ca l device for temporary joining
Fig. 6 . 7. Typical struc ture supported cranes u sed in some cases
55
56
varied between 75 - 900 F and 50 - 70 percent RH, respectively. All segments
of one longitudinal box girder were erected completely before cantilevering
started on the other box girder.
Details of the erection pro~edure were as follows:
(1) Lead wires of any implanted strain gages in the unit to be erected
were connected to the VIDAR system.
(2) Two wide flanges (4 X 4 X 50 in.) were connected to the bolts used
to support the dead load blocks which pierced the top slab of each
segment, as shown in Fig. 6.5.
(3) Preweighed epoxy resin and hardener sufficient for joining two
segments was mixed.
(4) Prestressing wire or strands were inserted in the straight portion
of the tendon ducts (Stage 1 in Fig. 6.8).
(5) Segment (A) was lifted, the height adjusted and leveled by turn
buckles, as shown in Fig. 6.9 (Stage 2 in Fig. 6.8).
(6) Segment (A) was separated after this adjustment. The joining sur
faces were cleaned with acetone and the epoxy resin was spread on
both joining surfaces,as shown in Fig. 6.10.
(7) Segment (A) was clamped temporarily near the lower flange as shown
(8)
(9)
(10)
in Fig. 6.6 and also at the top of the segments (Stage 3 in Fig. 6.8).
The force applied on the mechanical clamping device was approximately
1 kip each and was checked with a torque wrench. Lead wires connected
to the prestressing cable were pulled south until the end of the
straight portion. Lead wires were not necessary for the strands as
they could be pushed through by hand.
Same as (5) for Segment (B).
Same as (6) for Segment (B) •
Segment (B) was joined temporarily. The lead wire was pulled north
until the prestressing cable was extended just enough for seating
(Stage 4 in Fig. 6.8).
57
STAGE 1
I SOUTH I <i ---~~~---~~~ .. ~~~~~ I NORTH I
'" ,
@~LEAD WIRE
,
I~IWIDE FLANGE
STAGE 2
TEMPORARY PRESTR. STAGE 3
TEMPORARY PRESTRESSING
PIER
"
AfiI> LEAD WIRE
®kLEAD WIRE
STAGE 4 TEMPORARY PRESTRESSING
F!:~=-="I!If==-=:P"=-1=t=1mr-=-==r-~pi:=-.;~~ PRE STRE SS I NG , CABLE
~~-L~ __ ~~-L~~~~ ® TEMPORARY PRESTRESSI NG
@ ... @, @ ... @): LEAD WIRE
(II) ... (G): PRESTRESSING CABLE
Fig. 6.8. Insertion of prestressing cable
r' •. ~.IO. 'ppUeoUon of th. '1'0Il' ... ,n "" tho )olnl"3 I U.rlC<
59
(11) The prestressing equipment was se t on one end (Fig . 6.11) . Details
of the prestressing sy stem are shown in Fig . 6.12. Prestressing
force was applied using 30 ton hydraulic rams . The amount of
(12)
force applied was con trolled by a load cell. In addition. the
prestressing force applied was checked approximately with a
pressure gage.
A prestressing force of O . 8~ was applied in order to over come the 5
friction loss and then was dropped to O.6SCs
for seating . Although
the prestressing operations must be performed at both ends in the
prototype, prestressing force wa s applied only from one end
for the model because of the shorter lengths involved. However.
prestressing operation s were done alternately at each end.
Friction loss tests we re performed at various stages, as reported
in Sec. 6 . 2.2.
Fig. 6.11. General view of prestressing system for negative tendons
60
. . .' '. I,. . ~ :
," . :".' .: .. : . ... :, ... " TRESSING SYSTEM (B)
EL PLATE T. RAM STRESSING SYSTEM (A) ~~~--~----~
ETE SEGMENT
PRESTRESSING PROCEDURES:
(1) INSERT PRESTRESSING CABLE.
(2) PUT JAW INTO CHUCK BODY.
PUMP
INDICAT.
(3) SET SEATING ATTACHMENT, LOAD CELL AND 30 TON RAM.
UCK (TEMPo)
(4) INSERT TEMPORARY CHUCK AT THE END OF RAM AND HOLD IN POSITION.
(5) EXTEND STRESSING SYSTEM (A) UNTIL 0.8 f~ IS REACHED IN
PRESTRESSING CABLE.
(6) DROP THE PRESTRESSING FORCE FROM 0.8t; TO 0.651.'.
(7) SET SMALL RAMS AT STRESSING SYSTEM (B) AND APPLY 2 TO 3 KIPS
OF LOAD TO PUSH THE JAWS INTO CHUCK BODY IN ORDER TO SEAT THE
PRESTRESSING CABLE. RELEASE SYSTEM (B).
(8) RELEASE SYSTEM (A), AND REMOVE SEATING ATTACHMENT, LOAD CELL
AND 30 TON RAM.
Fig. 6.12. Details of prestressing system
61
(13) Excess prestressing wire or strands protruding from anchors were
cut with an electric grinder after seating of the jaws into the
chuck bodies. Anchorage holes were filled with epoxy mortar
[(Epoxy Resin): (Sand + Aggregate) = 1:1]. These surfaces were
ground smooth on the next day.
6.1.5 Correction of Horizontal and Vertical Alignment. Correction
of horizontal and vertical alignment was performed after the 9th segment
out from each pier was erected because the balanced cantilever sections
were still symmetrical and easy to adjust.
The following are the correction procedure steps used:
(a) Horizontal Alignment. There are two possible errors in hori
zontal alignment which may occur in (twin box) precast segmental cantilever
construction, as shown in Fig. 6.13. Points a, b, c, and d in Fig. 6.13
are the center points of the pier segments.
CASE 1
CLOSURE CENTER OF PIER SEGMENT
CENTER OF CASE 2
ab II cd
PIER SEGMEN1 ---+--L..-.I -__ - ~, ~=_ -_ ---+ ___ , b -- -±at ---- -)
[ - .......... 11-- Is I +-1 -
ab jI.. cd
Fig. 6.13. Errors of horizontal alignment
62
In case (1), lines ab and cd are parallel, but the cantilever
sections co and do extending from piers c and d are not in correct
horizontal alignment.
In case (2), piers are not correctly positioned and the distances
between ac and bd are not equal so that ab and cd are not parallel.
The anchor bolts in the piers were carefully set so that there
was no error of the case (2) type in the model.
There was about 3/4 in. of correction for case (1) required at
the closure of the two cantilever sections during erection of the first
half of the bridge. During the construction of the second half, a piano
wire was stretched between the points c and d (in Fig. 6.13) in order to
provide a base line for setting the two pier sections. This provided
much finer alignment. Less than 1/4 in. of correction was required for
the second half of the bridge.
Each cantilever section was adjusted after erection of nine seg
ments so that the center line of each cantilever tip would meet on the
base line. In order to adjust the cantilever sections with equipment
which could be practically scaled up for the prototype construction, a
small channel was connected to the webs of the 8th and 9th segments and
was pulled laterally by a ratchet hoist attached to the end pier, as
shown in Fig. 6.14. The maximum capacity of the ratchet hoist was 4 kips.
(b) Vertical Alignment. There are three possible errors in
vertical alignment at the closure of the two cantilever sections. They
are:
(1) Errors due to incorrect initial adjustment of the nuts supporting the pier segments.
(2) Errors due to variations in height of the soffit at the time of casting.
(3) Errors in joint widths at the time of cantilever erection.
It had been predicted in the design2l that a small tensile stress
would occur on the bottom fiber at a distance of 30 in. from the pier
center (at the joint between the 1st and 2nd cantilevered segments), when
•
til. <.~tI\ ... ", ... Io"l'h .... '0 'n, • .0 10 ' n. 1. til ...... t ... of <II.
fl ....... ' f 01 'h. bttd&o, d.H.it. J o,"~, .no •• o".nto! ' «'u" of
_.,t"" .f tIl. '_0 •• ..,. ,,~ •• '" ....... ,,_., <II. b ..... o f .ho •• _6< ... """~''''<I til ••• uaoU. ( ....... , ........ , .... of th. f t,,, ......... .. _t..o ..... '.0 .......... ,""'tI .. ,.""...tu ••• , .... 16 ....... tha, <lI ••• to
00 .......... "" oar _ Jo . ... d ••• ". .h. ' .... 110. ... ""'00. "III ...
..... "tho< , ...... , •• In • • Of<OU tft 'M fI .. , I ... Jototo bd o," ...... pett-
."". woo " ... no<l ..... 11 •• po .. lb •••• ,.'" ~" ..... ""ot''''' In tho ho 'ih.
of <b •• oftlt. n. ... f.," , .. b .... tt ..... <l eol .dJ~ ............ to bo d .....
0' """ .... ~I 1tI. <.~t(I .. Y" .......... .
.IolaUn •• " .... 10 <II_ .... , ...... 01 'n'Il .... ... e""
... _taco ...... . _.nU ...... 'h ..... of ",. "otll.vo ..... h""" 'n
"" b . \).
AlthO«lh tho .. <-' ... 01 <II. b"4,. " .. < .... «ve' ... v. '" "_0< ..... ~tI .. 1 ocro" ... J ....... , 01 "''' ' eol .11,_, ~" ... e ..... ' <. , . ..... d •• ~., , ..... ",. 1_1 ..... 1 .... cI ... ," v, ''' til. '.0.1 ..... '. ,,,,, •
... ,,<tto! ......
64
h E2
-fr-E1
E2 = .It... x E1 h
where,
h = Depth of segment
)E = Distance from the joint with the erection error
E 1 = Thickness of error in joint at bottom
E2 = Accumulated error in height at the distance )E
Fig. 6.15. Accumulation of joining error
Using procedures possible in the field, to correct the vertical
alignment the nuts at D in Fig. 6.16 were loosened slightly and about
600 lbs. of unbalanced weight was put on the top of the M9 segment. The
nuts at B were then backed down in small increments. If the nuts at B
were turned down one turn, then the outer end of the 59 segment dropped
1-1/4 in. [(1/13) X (196/12)] when the weight was removed from M9.
In the erection of the first box of the bridge, the two cantilver
sections were adjusted to the correct vertical alignment prior to the
closure operation, but were not lowered to the neoprene pads until after
65
19611
18411
600 LBS.
B A
Fig. 6.16. Correction of the vertical alignment
the positive tendons had been stressed. However, for the second half
(box) of the bridge, the two cantilever sections were separately lowered
to their final position on shims on the neoprene pads prior to the
closure operation. It was easier to lower the balanced sections onto
the neoprene pads prior to the closure operation and it simplified
the positive tendon operation. All lowering was done by unbalancing
the overhangs and backing the plate support nuts down in small increments.
6.1.6 Casting of the Closure Segment. After adj~ ting the
vertical and horizontal alignments, the half segments with diaphragms
were erected over the end piers and the positive tendons in the side
span were prestressed (details are in Sec. 6.1.7). Half segments (8.5 in.
long) were erected for each cantilever in the main span, and the over
lapping reinforcement extending from each half segment was joined by tie
wires. In addition, the tendon ducts extending from each midspan half
segment were connected by polyethelene tubes at the closure and sealed by
epoxy resin, as shown in Fig. 6.17. The positive tendons for the main
span were inser ted immediately after the above opera tion was comple ted
and before concreting of the closure.
Figure 6.18 shows the simulation of the closure prior to the
construction of the half segments. Since there was no space to work
66
Fig. 6.17. Connection of tendon duct a t the closure
Fig. 6.18. Simulation of the closure
67
inside the model segments, interior side and top slab forms were set in
place prior to joining the reinforcement at the closure and the top slab
forms were supported using the holes made for dead load blocks. Plexi
glas windows were provided in the bottom of the exterior side forms, in
order to observe any tendency for honeycombing around the tendon ducts.
Also, temporary bracing channels were rigidly attached to the top and
bottom slabs of the half segments, as shown in Fig. 6.18. These channels
were intended to provide temporary flexural stiffness across the closure
equivalent to half of the flexural stiffness of the finished concrete
box section and were kept in place until the concrete had cured.
In order to cast concrete in the bottom slab at the closure, the
concrete was placed through the gap between the top slabs and vibrated
from outside the bottom form. The form for the top slab was then set in
place prior to casting the web and top slab. Concrete in the closure was
cured for a week until the positive tendon stressing operation was
performed.
6.1.7 Prestressing of the Positive Tendons. In order to minimize
the effects of the secondary moments due to prestressing which occur in
indeterminate structures, the prestressing operation for the positive
moment tendons in the side spans was completed prior to the closure opera
tion. The C3 and C4 tendons, as shown in Fig. 2.3, were also intended to
prevent tensile cracking in the top slab due to prestressing of the C1 and
C2 tendons. The order of the prestressing operation was tendons C4, C3, C2,
and CI, as designated in Fig. 2.3. All tendons were prestressed from the
anchoragffiin the top slab, as shown in Fig. 6.19. Details of the prestress
ing equipment were the same as reported in Sec. 6.1.4.2.
The procedure for the positive tendon operations in the main span
was specified in the Texas Highway Department plans as follows:
(1) Set jacks at end supports.
(2) Stress tendons A1, A2, A3, and A4, in that order.
(3) Lower restraining nuts at pier units PS and transfer load to bearing pads.
(4) Adjust jacks at outer supports for a reaction of 15 kips (0.417 kips for the model) for each segment at each support.
68
Fig. 6.19. Setup for prestressing of positive tendons
(5) Stress tendon AS, then A6.
(6) Increase jacking reactions at outer supports to 92 kips (2.56 kips for the model) at each segment, at each support. Set bearing pads and shim to maintain this reaction.
This procedure was used as a guide for the model bridge construc
tion. For the model box girder bridge construction, the procedure indi
cated in the Texas Highway Department plans was followed, except for
step (3) and the amount of reaction in step (4). Step (3) was done after
stressing tendons AS and A6. The reaction of 15 kips in step (4)
(0.417 kips for the model) was very small and was already produced by
stressing AI, A2, A3, and A4 tendons. Therefore, the reaction was increased
by 0.28 kips (10.1 kips for the prototype) according to Lacey's preliminary
d . d . 21 es~gn recommen at~on.
The reaction at the ends was measured using load cells, while the
deflections at the ends and at the center of the main span were measured
with dial gages. Detailed results are given in Sec. 6.2.3 and 6.2.5.
Details of the prestressing system were the same as in Sec. 6.1.4.2.
Prestressing was performed from each end alternately.
In view of the experience gained in this erection sequence,
recommended procedures for the prototype bridge construction were as
follows:
(1) Lower the S9-M9 completed sections to their final position over the main pier at the symmetrical cantilever stage.
(2) Attach the end segments and prestress positive tendons in the side spans.
(3) Erect the midspan half segments and insert the positive
69
tendons in the main span, prior to casting the closure segment.
(4) Cast the closure segment.
(5) Set jacks at outer supports and stress tendon A1.
(6) Release restraining nuts at the main piers and temporary stiffness at the closure.
(7)-(9) Stress tendons A2-A4.
(10) Increase jacking reaction by 10 kips.
(11)-(12) Stress tendons A5-A6.
(13) Adjust elevation of the end segments to the optimum amount of reaction (see Sec. 7.3.3.2).
6.1.8 Grouting of Tendons. There were two options for grouting
the tendons of the model bridge.
(1) After prestressing each tendon.
(2) After prestressing all tendons.
Procedure (1) was requrred in the prototype structure to minimize corro
sion hazards. However, grouting after each stressing was considered too
time-consuming and too prone to the hazard of possible leakage of grout
into the adjacent tendon ducts due to possible honeycombing in the con
crete or if sufficient epoxy resin was not spread between the tendon ducts
at the joints. Therefore, procedure (2) was used in the model bridge.
Grouting was done separately for five groups of tendons, as shown in
Fig. 6.20. The grout pump and hoses were washed with clean water after
grouting each group of tendons.
Tendons in each group were located close to each other, as shown
in Fig. 6.21. When the B9 tendon was grouted, grout came out from the
outlets of both the B9 and B7 tendons simultaneously. This happened
70
TENDON 8 (SOUTH) t BRIDGE
TENDON 8 (NORTH)
TENDON C (SOUTH) TENDON A TENDON C (NORTH)
ORDER OF GROUTING: I) TENDON 8 (NORTH)
2} TENDON 8 (SOUTH)
3} TENDON c (NORTH)
4) TENDON C (SOUTH)
5) TENDON A
Fig. 6.20. Order of grouting for each tendon group
TENDON 8 81
89 87 85 86 sa 810 t
AT THE JOINT BETWEEN MAIN PIER
AND FIRST SEGMENTS
TENDON A
A5 A3
Ai
A6
A4
I
I AT THE JOINT BETWEEN M9 AND MtO
SEGMENTS
Fig. 6.21. Examples of tendon arrangement
71
because the distance between the B7 and B9 tendons was very small and
epoxy resin was apparently not spread fully between the tendon ducts at
some joint. However, it was not critical since grouting of the B7 tendon
was completed immediately after grouting the B9 tendon. It could have had
very serious consequences in the prototype with sequenced grouting. Grout
also came out from a joint which had marked damage on the concrete face
near a tendon duct.
Grouting mix and grouting procedures were in accord with Ref. 33.
The grouting mix used was:
Type III portland cement:
Water:
Admixture (INTRAPLAST C, SIKA):
94 lbs.
42 lbs.
1 lb.
In the first half of the bridge all tendon ducts were flushed
with clean water under pressure immediately before grouting. For the
second half, all tendons were blown out with air pressure immediately
prior to grouting and not flushed.
The following grouting procedure was used:
(1) Grouting injection pipes were screwed into the inlets of tendons as shown in Fig. 6.22
(2) Grout was pumped through the ducts and permitted to flow from the outlet until no visible slag of water or air was ejected.
(3) The valves of the injection pipes were closed while maintaining pressure and the outlets were closed.
(4) The injection pipes were removed from the inlets of the tendons and the inlets were closed.
The B series tendons were grouted completely in both the first
and second half of the bridge. It was initially felt that the A and C
series tendons were only grouted with about 70 percent effectiveness
because of some signs that grout might have leaked at some joints due to
honeycombing of concrete around the joints. However, by cutting segments at
several points after the failure of the bridge, it was determined that the
grouting for A and C series tendons was also virtually fully effective.
72
Fig . 6.22. Connection of injection pipes and inlets of tendons
73
6.1.9 Casting of the Midstrip Closure between the Two Box Units.
It was necessary to connect the two box girder bridges along the longi
tudinal midstrip closure after completing all prestressing and grouting
work for both boxes. If the reinforcement splices in the midstrip closure
had followed the prototype plan, the transverse reinforcement would have
extended past the concrete 1.8 in. and been spliced by welding a 2 in.
long bar to both protruding transverse bars (see Appendix A). Because
space limitations made it extremely difficult to weld splice bars on the
model, the transverse reinforcement in the top slab was extended about 3 in.
at the time of casting so that the bars overlapped. Thus the connection
was developed by lap splicing of the transverse reinforcement. Twelve gao
wires were inserted between and tied to the transverse reinforcement at
each connection to serve as longitudinal and distribution reinforcement
in the closure strip.
A continuous form was set under the top slab of the cantilever
portion and sealed to prevent the flow of concrete. Since the vertical
surface of the top slab at the midstrip closure was not rough, water was
sprayed on the surface, and a light coating of cement mixed with water was
placed on the surfaces just prior to casting. Concrete (same mix as used
in the precast segments) was cast and the surface was coated to prevent
shrinkage cracks in the midstrip closure.
6.2 Behavior of the Bridge during Construction
6.2.1 Rotational Stiffness at the Main Pier. Unbalanced loading
tests were performed at various stages of construction in order to check
the performance of the temporary anchor bolts. These results are shown
in Fig. 6.23.
After each unbalanced loading test was completed, the balancing
segment was erected. Although the bolts behaved elastically during the
test of the 7th segment, there was noticeable residual deflection after
the test of the 9th segment. When the structure was brought to a balanced
configuration after the unbalanced loading experiment with the 9th segment,
74
- TO z -z 0
0.:: 0 ; j:: z u 31: IJJ 8 0.5 ..J LI.
~
1.0
9 TH. SEGMENT
~~9~~~8~1_7_1~6~_~~_4~3~_2~_1~P~~t~ 6 DIAL SAGE
EXPERIMENT:
<D TOTAL DEFLECTION
® DEFLECTION EXCEPTOfL€lGUM'\:~TS @ DEFLECTION AT BALANCE
THEORY:
m TOTAL DEFLECTION
00 DEFLECTION EXCEPT FLEXURE OF SEGMENTS
Fig. 6.23. Deflection at unbalanced loading
75
the bolts under the pier segments were carefully examined and it was found
that the bolts on the compression side had deformed and showed evidence
of yielding. Although the calculated direct compressive stress was less
than the bolt yield strength, yielding was apparently influenced by the
large gap between the pier segments and the pier, with consequent local
bending and accentuated by the stress concentration in the threads.
Experimental and theoretical values for deflection under unbalanced
loading are compared in Fig. 6.23. In both cases, the experimental deflec
tion was appreciably larger than the theoretical value. Several factors
which affect the deflection under unbalanced loading are:
(1) Flexure of segments Isee Fig. 6.24, Case (1) ] •
(2) Elongation of the pier bolts [see Fig. 6.24, Case (2)]
(3) Bending of the pier [see Fig. 6.24, Case (3)]
(4) Moment connection of the pier to the test floor.
(5) Slippage of bolts with respect to the pier concrete.
All cases except (4) and (5) were considered in the calculation of the
theoretical deflection.
For the 7 th 'segment, deflections were ca1cu1a ted as follows:
Deflection due to flexure of segments IFig. 6.24 (1)]:
°1 = P~/3EI ~ 1.56 X 1403/(3 X 4.56 X 10
3 X 6060)
Deflection due to elongation of bolts [Fig. 6.24 (2)]:
Ll1 = SIR. /AE c S = (19.0 x 1.5)/(6 x 0.142 x 29 x 10 3
)
= 1.15 X 10- 3 in.
Ll2 = S2
R. /AE = (17.4 x 19.0)/(6 x 0.142 x 29 x 10 3)
t S
= 13.4 X 10- 3 in.
82 = (1.15 + 13.4) x 10- 3 /12 = 1.21 x 10- 3 radian.
°2 = 1.21 x 10- 3 x 125 = 0.151 in.
0.052 in.
76
1.56 K
!
146"
CASE (1) - FLEXURE OF SEGMENTS
1.56K
.L + ~ 51 ....I-~-----=-,....,----.....,~
T I. 140" ~
CASE (2) - ELONGATION OF SOLTS
REACTION S1 = 19.0 K
SI = -17. 4K
L _= n-------, f " s,~
-S-0-LT-(-2-)...... ~ p~r BOLT (1) •
CASE (3) - BENDING OF PIER
M
55"
\ \
n ::::::::c 93 ,
PIER
1:: ----.L ----=----1[ 03
Ec OF PIER:
= W1.5 x 33ft; = 1451.5x 33xJ3770
= 3.54 x 103
KSI
Ie OF PIER:
= 8870 IN.4
Fig. 6.24. Individual factors affecting the deflection
77
Deflection due to bending of pier [Fig. 6.24 (3)]:
83
- Mh/EcI - 218 x 55/(3.51 x 10 3 x 8870)
- 0.382 x 10- 3 radian.
03 - 0.382 x 10- 3 x 130 = 0.050 in.
Therefore
02 + 03 - 0.151 + 0.050 = 0.201 in.
While it is not possible to quantify the exact length of bolt
embedment required for full development of bond and prevention of any
slip with reference to the concrete, if this length is as much as 33 per
cent of the development length specified for an equivalent deformed bar
under the 1971 ACI Code,3 the calculated deflection would increase by
almost 33 percent and be in much better agreement with the observed
values. A substantial part of the deflection discrepancy is believed
due to the uncertainty of effective bolt imbedment.
Deflection for the 9th segment (unbalanced) was calculated in
the same manner as follows:
01 = 0.110 in. 02 = 0.259 in. 03 - 0.084 in.
Although there are many ways to connect the pier segment to the
pier temporarily or permanently, it is not recommended that the same
method used in this model bridge be used in future bridges, because of
the high compressive force applied to the bolts and the lack of stiffness
in the connection. If the same general type of moment connection is to
78
be used, as shown in Fig. 6.25, it should provide for temporary
compression support blocks to carry the compression and require the
bolts to be prestressed so that the large deflection component (02) due
to bolt elongation would be eliminated.
REDUCE GAP TO ABSOLUTE MJNIMUM
PRESTRESSED
II
" L-------~ TEMPORARY SUPPORT
BLOCKS
----t,-t-'---- NEOPRENE PAD
....--- BOLTS
Fig. 6.25. Suggested improvement for temporary support
6.2.2 Prestressing Force. In Brown's incremental ana1ysis,10
he recommended that an initial 0.8f' overstress be applied and then be s
reduced prior to seating to
tendon force would be 0.6f' s
cables. In the analysis of
about 0.65f' , so that after losses the s
in the straight portion of the prestressing
the bridge (using SIMPLA2 as explained in
Sec. 6.2.3), the friction coefficient (~) and the wobble coefficient (A)
were assumed as 0.25/radian and 0.000017/in., respectively.
Friction loss tests were performed during stressing of three
different cables (B6, A1, and A4 in the second half of construction).
Prestressing force was applied at one end and the resulting force at
the far end was measured using a load cell at that point. Prestressing
force was applied up to the specified 0.8f' and the elongation of the s
prestressing cables was also measured.
79
The generally accepted equation26
to find prestressing force at
any location after friction and wobble loss is given as follows:
where
P ij = prestressing force at a certain point (kip)
P. = pres tressing force applied at the end (kip) ~o
j.l = friction coefficient (per radian)
a angle change of tendon (radian)
A wobble coefficient (per in.)
tp = length of cables from the end to the point considered (in.)
Theoretical total elongation may be calculated by the follow equation:
where
t.t = 2:(P .. /E A )dt p ~J s P P
Pij = prestressing force at the certain point (kip)
E = modulus of elasticity of prestressing cable (ksi) s
A area of prestressing cable (in~) p tp length of prestressing cable (in.)
Experimental results were compared with the theoretical values,
as shown in Tables 6.1 and 6.2. Tendon profiles used in the theoretical
calculations were taken from the plans.
TABLE 6.1. FRICTION LOSS TEST RESULTS
Tendon Force (0.8f') Force at the Far End ~k2 (Experiment/Theory) AEElied ~kiE~ EXEeriment Theory
B6 11.9 8.94 9.26 0.966
A1 7.28 5.30 5.58 0.950
A4 7.28 5.26 5.59 0.942
80
TABLE 6.2. ELONGATION OF PRESTRESSING CABLES IN SEVERAL STAGES
Experimental Theoretical Elongation Tendon Elongation E (ksi) M; (in.) Exp. /Theo.
~in. 2 s p
B6 1.67 27.0 X 103
1.65 1.01 7 nun wire
B7 1.99 1.91 1.04
B9 2.28 X 103
2.31 0.986 6 gao wire 27.0
B10 2.41 2.55 0.945
A1 2.46 2.64 0.931
A2 2.19 27.0 X 103
2.37 0.924
A3 1/4 in.
2.06 2.11 0.975 strands A4 1. 78 1.85 0.961
Table 6.1 indicates that the measured force at the far end was less
than the theoretical value by 5 percent. Five percent difference at the far
end indicates that there would be about 3 percent difference between the
theoretical and experimental values in the straight portion. So a pre
stressing force of 0.6f' was assumed in the straight portion in all s
theoretical calculations.
Table 6.2 lists some measured elongations during various stages of
construction, including the friction loss tests. Experimental results are
compared with the theoretical calculations. The ratio of measured (experi
ment) to calculated (theory) elongation of all strands showed the same good
accuracy as the results for forces shown in Table 6.1.
When the applied prestressing force was lowered to 0.65f' for s
seating, some change in elongation was measured, but no change was noted
in the force at the far end.
6.2.3 Deflections. It was difficult to accurately measure the
vertical deflection during construction because the deflection was so sna11
in comparison to the span (L/1800). Casting and construction of the model
81
segments required six times the accuracy of the prototype in order to
maintain the same relative tolerance as the prototype. This was not
completely feasible, so deflections were measured only to see the general
trend of cantilever section behavior. Theoretical deflections were
calculated using the computer program SIMPLA2ll which provides an analysis
at each stage of erection using the finite segment technique.
The theoretical and experimental deflection profiles are shown
in Figs. 6.26 and 6.27, respectively, for all stages of erection. Relative
deflections for one typical case are compared in Fig. 6.28. There are
many factors which affect the vertical deflections, as mentioned in
Sec. 6.l.5(b). Therefore, it is very hard to determine the cause of
errors in deflection. This is especially difficult since the component
of deflection due to dead load is almost completely balanced by the
deflection due to the prestressing. Figures 6.26 and 6.27 indicate that
the overall trend of the theoretical and experimental results agreed
fairly well except that the measured results show a pronounced upward
skew. Although the experimental deflection at construction of the 7th
segment was still upward at the tip of the last unit, it decreased upon
the addition of the 8th and 9th segments, as did the theoretical results.
Experimental deflection measuremen~s shown in Fig. 6.27 are
averages at each corresponding point to eliminate the effect of unbalance
or bolt bending. Figure 6.27 readily indicates the jointing errors in
the initial stages of cantilever erection when temporary tensile stresses
at the bottom of initial joints were not controlled and joints widened
at the base, causing upward deflections.
During the positive tendon operations in the main span, Fig. 6.29
indicates experimental and theoretical deflections agreed well. The
additional theoretical procedure used to calculate these deflections using
program BMCOL5028
is explained in Sec. 6.2.5.
Superimposing results from Figs. 6.26 and 6.29 indicates that the
relative displacement at the center in the main span should be almost zero
upon prestressing of all positive tendons (Al-A6). The center of the main
82
o G:
~ • O. a..
~ :+ 0 l-~;;a._._~.~. _~~._~.~=-~-~.~_;_~.~_+_:.=_::::.=_:_=_=._:==_:_--@-8+------1 ••• G: ._ ~ c . __ -J :. ._ u. z -0. ._ I.&J ~ • __
O :. '--
o ._ o . __ '-'--'---<ID -0.10
Fig. 6.26. Deflection predicted by SIMPLA2 for cantilever erection
'i z 0 ~ 0 I.&J -J U. I.&J 0
/
6 /
0.2 r- ____ ---@
~/
0.10 0 G:
.F"---@ "~ ~:/--- -----"""'--@ y ~.~ ~
v .. ---.. ~ /' . ,,-
; a..
y-.~ ./ .' ---- .. --"
~
1 0
Fig. 6.27. Typical measured deflection during cantilever erection
""'"': Z -Z 0 ~ U 1.&.1 ...J LL LaJ 0
0 a::: ~ Q.
-: 01
z ci+ - j z z Q ~ ...... 8 ~ 0.10 ...J LL 1.&.1 o
ASSUMES THE CENTER OF 6TH SEGMENT IS ZERO WHEN 6 TH SEGMENTS ARE ERECTED.
~SEGMENT ERECTED
EXPERIMENT
83
Fig. 6.28. Deflection relative to the center of 6th segment
0.15
0.10
0.
\ \ \ \ \ . \ . \
~ EXPERIMENT
--- BMCOL 50 _ .. - SIMPLA 2
\ \ \
L Ai A2 A3 A4 A5 A6 RAISE 26/100"
AT BOTH END SUPPORTS
Fig. 6.29. Deflection at the center in main span during positive tendon operations
84
span was subsequently lowered 0.08 in. (0.48 in. in the prototype) by the
jacking at the outer supports
6.2.4 Strains. The sequence of construction stages is listed in
Table 6.3. Measured and computed strains are shown in Figs. 6.30 to 6.38.
Longitudinal strains plotted are usually averages of duplicate
readings at identical positions in similar segments.
Figure 6.30 shows that the experimental strains in the top slab of
the Ml and Sl segments varied widely across the cross section during
erection of several segments. The measured results are in good agreement
with the theoretical calculations for stage 1 and after erection of the
5th segment. Experimental strains were uniform across the cross section
and close to the beam analysis values for the erection of the 6th and
subsequent segments. Although the longitudinal strains in the top slab
of the Ml and Sl segments varied across the cross sections in early
erection stages, it is not a serious problem. All strains are in com
pression across the cross section and are well below the strains which
would accompany the maximum allowable compressive stress. The non
uniformity of strain in the top slab over the web was probably greatly
influenced by the local concentrated tendon forces. As additional tendons
were stressed during the erection of the second through fifth segments,
these local effects died out, as shown in Fig. 6.30.
For strains in the Ml and Sl segments, SIMPLA2 gave reasonable
predictions of the longitudinal strain distribution at the erection of the
first segment, but then showed poorer agreement with the experimental
results until the erection of the 6th segments. Since beam theory does
not predict any variation in longitudinal strain distribution across a
transverse section, the deviation from the experimental values indicated
for the first two stages were somewhat expected. For erection of the 3rd
through 6th segments, calculation by beam theory was closer to the experi
mental values. In subsequent stages, beam theory showed excellent agree
ment with measured values. Deviation of the experimental and beam theory
results in the initial stages greatly affects the accumulated values of
85
TABLE 6.3. DETAILS AT EACH STAGE OF CONSTRUCTION
Segment Prestressing Amount of Force Amount of Force Stage Per Cable at Per Cable at Erected Cable 0.8f' (kip) O. 6f' (kip) s s
1 M1, Sl B1, B2 11. 9 8.94
2 M2, S2 B3 18.3 13.8
3 M3, S3 B4 18.3 13.8
4 M4, S4 B5 11.9 8.94
5 MS, S5 B6 11.9 8.94
6 M6, S6 B7 11.9 8.94
7 M7, S7 B8 5.52 4.13
8 M8, S8 B9 5.52 4.13
9 M9, S9 B10 5.52 4.13
10 S10 C4 5.52 4.13
11 ------ C3 5.52 4.13
12 ------ C2 5.52 4.13
13 ------ C1 5.52 4.13
14 M10 B11 5.52 4.13
Closure segment was cast and supported at ends (3 span continuous beam) •
15 ------ Al 7.28 5.46
16 ------ A2 7.28 5.46
17 ------ A3 7.28 5.46
18 ------ A4 7.28 5.46
19 ------ A5 7.28 5.46
20 ------ A6 7.28 5.46
21 Raise 0.26 in. at outer supports.
86
ERECT M1,S1 -I
.... -50 z
" c..>
z t 0 ~ I-z
0. A
r----41
IlBRIDGE ERECT M5, S5 ~
o ~ t -2~ .. _ 34""-~--Li~--...L4Il""" :::l. I-
~ ! ~~ i BRIDGE t-----------------
ERECT M6,S6
- 52
~l"~~~-~~ ~ I-
~ ! -2:~ -40
I-~ i -~~ r-----------~-G-~~--------r------------------------------~ ERECT M2,S2 A ERECT M7, S7
ERECT M3, S3
~ " c..> ~ t 0 3-z I-
~ r' I- 2~r
ERECT M4,S4
-50
~ c..> ....... t z 0
3- I-z < c..>
-~~ a::
f I-en
25
- - --- ---f!~ A
-45
ERECT M8,S8
-83
o L _.1 ... 8 __ ..... __ --i1--l
2~ ~-~i&--
I-
ERECT M9, S9
c..>
i t " Z
-42 ~ I-
Fig. 6.30. Strain in M1 and Sl segments during balanced cantilever construction
STAGE
(14)
S1 .
SIDE SPAN- B j
AT SECTION 8-8
· Mi
LA -MAIN SPAN
o - EXPERIMENT 8. - SIMPLA 2
8M THEORY
AT SECTION A-A
87
PRESTRESS C4,C3,C2,C1 It BRIDGE PRESTRESS C4,C3,C2,C1 It BRIDGE
~ r1~. ~ r~~·5.6 -2 - to- 25~ 0 to- 25 .3- ~ -~ U-~l ~ l-~L t 2* t; >- 25~ - 1 -3
ERECT Mi0 ERECT Mi0 t BRIDGE ...,
2 Ul z
r~~ .13 " to' 't ~ -6 ! 3- to- 2 ~ to- 25
lBRIDGE
z fl z
~ < r:~ D: 0::
to- t; (f)
t- 25 -38 -15
Fig. 6.31. Strains in M1 and Sl segments during erection of half segments
88
1
A'-' STRAIN READINGS AT A-A M1! AND A'-A' WERE AVERAGED
~IJ I 1 I
PRESTRESS Ai i BRIDGE PRESTRESS AS tBRIDGE
2
~ 3-z
~ r:~~ PRESTRESS A2 PRESTRESS A6 fl BRIDGE
~ O-l A..- ~ (,)
"- t 0 -13 ~ t Z 14 :::t I- 25
I 3- I-
Z
r!~ ~ f -~ ex
+20 ex
ex: ex: l- t; (f)
PRESTRESS A3 q,. BRIDGE RAISE 26/100 IN. AT END SUPPORTS iBRIDGE ..., O~l~ -ll- -&-
z
~ to· ~ (,) -25 . , -12 ~ t 0 -43 - 46
~ I- 25 z =l.
-.44 ex
l~~ ..... I- 25
I ex: z
f -~f t; ex
ex: +46 ~ 25 t;
r' PRESTRESS A4 i BRIDGE
2
t~~ ~ 3- I- 25 o - EXPERIMENT z ~ rl &. - SIMPLA2 I-(f) _. - 8M THEORY
I- 25
Fig. 6.32. Strains in HI segments during closure operation
SEGMENT TENDON 0 - COMPRESSION
STRAIN (}J IN/IN.) -100 -200 -300
M 1, S1 B1, B2
M2,S2 B3
M3,S3 B4
M4,S4 B5 A EXPERL •• • A A SIMP. 2 )( e EI
MS,SS B6 i.BRIDGE • • )( A eo M6,S6 B7 •• • )( A eo M7,S7 B8 ", • )( A- CD
M8,S8 B9 • • X A eEl M9,S9 B10 • )( A e9
M10 B11
A1
A2
A3 • • )( 0 A e A4
A5
A6 RAISE END SUPPORT
(26/100 IN.) co \D
Fig. 6.33. Strains in the top slab of Ml segments during construction
SEGMENT TENDON ----+ COMPRESSION o
STRAIN ()' IN/IN'>
200 -300 100 - -I I 1
M1,St B1, B2 ,... - 1_ v "7 __ rc;J ,
M2,S2 B3 _.III
III ~--. "III '"'U
M3,S3 B4 -1- ..- ..... -. 1- ., L.J
M4,S4 B5 -- ~ ...... .... -- "'- L.J
M5,S5 86 __ V
,... r-. -- ..., ---.::r
M6,S6 87 -~
_ .... - _I00.I
M7,S7 B8 - "'" .... -,.. ..., L.I
M8,S8 B9 ct BRIDGE .~ 0S
M9,S9 B10
Mi0 B11
Ai
A2
A3
\ <t OF Mt i • .)( eo
.x B e
•• B e .1 i
~ERI. ~XPERI •• S 0
• • B e • SIMPLA 2 • SIMP. 2'
A4 X BM THEORY •• >e e A5 .- --... ,... __ --u:II\ -A6 ..... --~"
-. ---"" RAISE END SUPPOR T ~ "" (26/100 IN.) --- "
Fig. 6.34. Strains in the bottom slab of Ml segments during construction
~ -25
......, (,) z
t ::::(. 0 -z Ci I-
a: ! -i l-(/)
I- .. I
D
D STRAIN READING
t BRIDGE
--++ ,
o - EXPERI MENT
A- SIMPLA 2
_. - 8M THEORY
Fig. 6.35. Strains in the M6 and S6 segments at erection of the 6th segment
91
92
1 220"
I 1 180" .,
PRESTRESS Ai t BRIDGE PRESTRESS A5 ~ BRIDGE
~r~t~~~~ 3- I- 25~ z
~ t~l I- 25~ -30
PRESTRESS A3
- 36
PRESTRESS A4
t BRIDGE: PRESTRESS A6 ~ BRIDGE
~ + ~t .... _1 ~...---..I.a...r 3- I- 25~ z
~ r~t I- 2~~
RAISE 26/100 IN.AT END SUPPORT ~ BRIDGE iB lOGE
~ 0 ~i -A---i--
- 40
~ t 0 -32 -45 3- I- 2 z
- 37
~ +-0 I- 25
5
+60 + 45
~-- ____ H
lBRIDGE
~ r~t .a-_6
~--.........., 3- I- 2~r
75
z
~ r~t I- 2~~
o - EXPERIMENT
8. - SIMPLA 2 ___ 8M THEORY
Fig. 6.36. Strains in the M9 segment during closure operations
93
STRAIN (,u IN.lINJ
SEGMENT TENDON 0---. COMPRESSION -100 -200 ~----------------~~--------------~~---
Ai
A2
A3
A4
AS
A6 RAISE END SUPPORT
o EXPERI 8. EXPERI. • SIMPLA2 SIMPLA 2
~ BRIDGE
EI EXPER. ct OF NM$ • SIMP. 2 !
(26/100 IN.) t-----II.---.aJ.--X 8M THEORY
Fig. 6.37. Strains in the top slab around the center of the main span
SEGMENT TENDON
A1
A2
A3
A4
AS
A6
RAISE END SUPPORT (26/100 IN.)
~COMPRESSION o
STRAI N (jJ IN./IN.)
-100
• 0 a
-200
Fig. 6.38. Strains in the bottom slab around the center of the main span
94
strain in the top slab of Ml and Sl, as shown in Fig. 6.33, even though
the subsequent longitudinal strain increments agreed during cantilever
erection, as shown in Figs. 6.30 and 6.31. Figure 6.33 gives a distorted
feel for the accuracy. The large discrepancies at each stage are almost
completely due to the lack of agreement in the first two stages.
Because Figs. 6.30 through 6.34 indicated so much deviation in
the longitudinal strains across the cross section for the Ml and Sl
segments during erection of the first groups of segments, the longitudinal
strains in the 6th segment during its erection are shown in Fig. 6.35.
At this stage the experimental strains measured in the top slab were uni
form although SIMPLA2 predicted some deviation. The magnitude of this
deviation was small in comparison to the first segment because of the
lower prestressing force as the cantilever sections extended. Apparently
little local compression effect existed.
In the bottom slab the experimental and theoretical results agreed
very well, as shown in Figs. 6.30, 6.31, and 6.34. The strain produced
was almost uniform in the bottom slab, except at the erection of the first
three segments. Figure 6.34 shows tensile strains developed in the bottom
of the Ml-Sl segments at the erection of the first two segments as
predicted.
During the positive tendon stressing operations after casting the
closure segment, Figs. 6.33 through 6.38 show that the values predicted
by both SIMPLA2 and beam theory deviated from the experimental results
in the top slab of the Ml and M9 segments. Although the experimental
results were consistently greater than the theoretical calculations, no
ready explanation for the fifference is known. The small absolute values
of the strain might be one of the factors for the discrepancy due to
inherent difficulties in accurately measuring these very small strains.
Figure 6.33 shows that the maximum strains in the Ml and Sl segments
occurred in the top slab at the erection of the 6th segments as predicted
and at the completion of construction. Figure 6.34 shows the maximum
strain in the bottom slab to occur at the erection of the MIO segment.
95
The magnitude of the maximum compressive strain in the M1 and Sl segments
was approximately 330 ~in./in. in both the top and bottom slab. There
fore, the maximum compressive stress was about 1500 psi. This value was
only 21 percent of the ultimate compressive strength of the concrete
(7090 psi) and about half of the maximum allowable stress. The maximum
strain in the bottom slab in the closure strip at the center of the main
span was about 200 ~in./in. during construction of the bridge.
All readings were taken before and after the application of load
and not cumulatively. Creep effects appeared very small, although the
experimental strain values were generally larger than the values given by
beam theory. Since the beam theory agreed fairly well with the experi
mental results, the BMCOL50 program can be modified for theoretical calcula
tions for construction stages for this type of bridge with similar transverse
and longitudinal stiffness.
In general, both theoretical solutions and the experimental
results were in reasonable agreement when the change of strain in each
stage was reasonably large, except for the local strains in the top slab
during initial stages when the large local compressive forces from the
tendons seemed to affect the strain distributions.
6.2.5 Reaction at Outer Supports during Positive Tendon Operations.
Prior to positive tendon prestressing operations in the main span, the end
supports were adjusted to just bear on the underside of each web at the
edge of the end pier segments. Reactions at the end supports were measured
by sensitive load cells during each stage. Comparison between theoretical
and experimental results is shown in Table 6.4.
Figure 6.39 shows the simplified procedure used to calculate the
reaction at the end. Prestressing forces were replaced by the vertical
forces which produced the moment diagram due to prestressing. These
reactions were calculated using the BMCOL50 program.28
Experimental reactions agreed very well with the theoretical
values for large values such as A1 and A2 tendons. As the reaction
increment became smaller, the experimental readings became much smaller
96
TABLE 6.4. REACTION AT OUTER SUPPORT DURING PRESTRESSING IN MAIN SPAN
Tendon Experiment
Al
A2
A3
A4
AS
A6
Raise 0.26 in. at outer supports
Total
(A) PRESTRESSING FORCE AND ECCENTRICITY OF CABLE.
M,= Pe
0.245
0.200
0.115
0.074
0.012
0.012
1.04
1.698
(B) MOMENT DUE TO PRESTRESSING.
A
(kip) BMCOL 50 (kip)
0.251
0.213
0.169
0.131
0.091
0.055
0.700
1.610
(D) POSITION OF LOAD FOR EACH CABLE.
A1 6
A2 ~M ~ o
A3 A
A4
. ..,. . 2 481K
Q t '20" Ji '20"Jt'20"j' (C) EQUIVALENT VERTICAL FORCE FOR M,.. " 20
470K 20
A
A
A
! v;. L A5 f' i a t t A a a---"""III,r--14-0,,-.l.rr.:;.;t-"~-1-40-""""4~-~A ~ .~z4 ~ l120
" A6
6 t 160" :W'1. 160" j 2.d' 20"
A
Fig. 6.39. Calculation of end reaction due to positive tendons
97
than the theoretical values. However, the total experimental reaction
(1.698 kip) agreed very well with the total theoretical value (1.610 kip).
!!!!!!!!!!!!!!!!!!!"#$%!&'()!*)&+',)%!'-!$-.)-.$/-'++0!1+'-2!&'()!$-!.#)!/*$($-'+3!
44!5"6!7$1*'*0!8$($.$9'.$/-!")':!
C HAP T E R 7
DESIGN SERVICE AND ULTIMATE LOAD TESTS
7.1 General
The completed bridge was load tes ted for the governing MSHO
design loading conditions21 shown in Table 7.1.
TABLE 7.1. CRITICAL LOADING CONDITIONS IN LONGITUDINAL DIRECTION
CASE CRITICAL CONDITION LOADING CONDITION
(1)
(2)
(3)
(4 )
MAXIMUM POSITIVE MOM·· LANE
ENT AT THE CENTER OF THE MAIN SPAN.
b NE
MAXIMUM POSITIVE MOM-TRUCK LOADING:
ENT IN THE SIDE SPAN. A a SE SM
MAXIMUM NEGATIVE MOM- LANE
ENT AT THE MAIN PIER.
I = 0.222
A NM
A NE
MAXIMUM SHEAR ADJA- LANE LOADING: r w-4 k,zzzzzz"ZZZZillMz,'1
CENT TO THE MAIN PIER. SE SM NM NE I=0.182
• I = 1M PACT FACTOR
99
100
The order of testing was as follows:
(1) Service loading {1. 0 DL + 1. 0 (LL + IL)].
(a) Case (2) (b) Case (1) (c) Cage (3) (d) Case (4)
(2) Ultimate loading. The ultimate load criteria used in the actual
design of the bridge was the Bureau of Public Roads 1969 Ultimate Design
Criteria7 that U = 1.35 DL + 2.25 (LL + IL). Prior to applying any of
the ultimate design live loads, supplementary concrete blocks corresponding
to 0.35 dead load were added to the structure. Then, design ultimate
live load was applied in the following sequence:
(a) Case (1) (b) Case (3) (c) Case (4) (d) Case (2)
The AASHO reduction factors for load intensity were not used in
any tests. These factors (reflecting improbable coincident maximum
loading in all four lanes) would have allowed a 25 percent reduction in
live load. Special loads were applied to study transverse moment dis
tribution for the different types of truck loadings and lane loading (two
lanes). The weight of the asphalt topping, which could be about 8 percent
of the dead load, was not included in the model bridge test.
7.2 Test Procedures
7.2.1 Simulation of Loading
7.2.1.1 Truck Loading. AASH06 HS20-S16-44 truck loadings were
scaled as shown in Fig. 7.1. Tire pressure was assumed as 80 psi in sizing
the loading pads, as shown in Fig. 7.2.
7.2.1.2 Lane Loading. Uniform lane loads were closely simulated
by applying a series of concentrated loads at 4 ft. intervals. Since each
lane is roughly centered over a web, as shown in Fig. 7.3, load was applied
above each web in the main test series. Transverse distribution was
checked in another series.
ITEM FULL SIZE TRUCK MOOEL TRUCK
W (Iba.) 40,000 1111
d, (ft.) 6.0 1.0
d2(ft.) 14.0 2~3
~(ft.) 14.0-30.0 2 ~3-!S.0
Fig. 7.1. Dimensions of full size and 1/6 scale model AASHO HS20-S16 truck
101
102
LOAD LOAD
~ :::I I" (STEEL PLATE)
~ I V4"(NEOPRENE PAD)
~
(a) FRONT WHEEL (b) REAR WHEEL
(c) 4 LANE AASHO HS20-16 TRUCK
Fig. 7.2. Scaled wheels and AASHO HS20-S16 truck
eo PSI
103
6" 24" 24" 6"
1 LANE 1 LANE • 1 LANE 1 LANE
LOAD
, , 1 l!TEE
II V NEO L PLATE PRENE PAD
• • I • • •
0 ~D 13" 29" 28" 29" J
.-
Fig. 7.3. Application of concentrated loads above the webs (in each lane)
13" _
Moments and shears for the uniform loads and the equivalent
concentrated loads of Fig. 7.4 are compared in Table 7.2. Although
extremely close agreement is shown for bending moment, there are some
minor differences in shears. These are generally on the conservative
side. The least conservative case represents a difference of only
4 percent.
7.2.2 Loading System and Instrumentation
7.2.2.1 Loading System. Loading reactions were furnished by
steel beams attached to the structural test floor as shown in Fig. 7.5.
Hydraulic rams were connected to pumps with hoses and the number of pumps
was minimized by using many manifolds. Load was applied by electric or
manual pumps. Two load cells were generally used in each loading system.
One load cell was used to regulate pump pressure while the other load
cell was used to check the load applied. In addition, hydraulic system
pressure gage readings were also recorded for check purposes. Load control
procedures worked well.
7.2.2.2 Instrumentation and Observation. Strain gage measurements
were taken as shown in Fig. 4.14. Numerous dial gages measured deflections
of the top slab, as shown in Fig. 7.6. Although nine gages were used at
104
t BRIDGE
I 1231
K
!' ./ 0.0410 KIlN.
SE NMO, ~INE 100" 400" !100M
~--~~--~~--------~~K--------WT----~~---Jl
SM
LOADING CASE (1) I( I( Ie' 2.311( K
1.97 1.97 1.17 1.97 1.97 1.97
SE SM 52" .. NM 3 AT 4'"
200" ~OOU 200"
t BRIDGE
SE j. '00'
I 'r '40' eo·'t 0.0420 KIlN. 1 4 t· gll,l, , tt II , I "Ill' t, t Z , Z z1 t Z I I I Z I , I!
LOADING CASE (3)
SE 6 AT 4."
200· 400· 200"
LOADING CASE
SE SM
7 AT 4'" I a AT48· <40"
200" 400" 200"
Fig. 7.4. Lane loads and equivalent concentrated loads for four lanes with impact allowances
NE
NE
TABLE 7.2. MOMENT OR SHEAR FOR EACH LOADING CASE
SE (I) (2) (3) (4) SM (5) (6) (7) (8) (9) (10) (II) (12) (13) NM (14) (/5) (16) (17) NE
, , II: o -99 -219 -038 457 -497 - 219 .121 ..... 515 .555 .51' <36 •• 121 -219 -497 -457 -3E -99 0
MOVENT FOR CASE 1 is •
is H U~, U' 0 - 100 -Z20 - 3110 -460 -!500 -221 .129 .313 .. 524 .. 552 .. 524 +373 .129 -221 -500 -460 -340 0 -100 0 (K-IN.)
a -z: :II.
t 1 ,
i l 0 -89 -197 -297 412 -447 -188 +120 .. 332 .. 441 .0467 .. 464 +356 +68 -31 -630 -494 -154 +13 +46 0
~ENT Fffi CA!:£3 & lIE
JJ HHlilldu ~ 0 -89 -195 -301 -407 -44~ -176 +128 +336 +446 +453 +459 +316 .84 -307 - -133 .. 5 +46 0
(K-IN.) 21: :a.
~ 0 -80 -176 _2"7' -368 -400 -158 +122 +305 +392 .399 ~ -229 -524 -24 +30 0 MOMENT FOR CASE 4 i 1 is
, , 1 , , , , ,1. , , , 0 -80 -176 -272 -368 400 -150 +130 +314 +401 .396 .39C .l89 +81 1-2211-52E -412 -166 -16 .36 0
(K-INJ - ... - -l -2.00 _~nt'
-2.00 +6.8S .4.83 +0.80 -o.ZI -1211-3.24r26~7.29 -12.0
2nt'! -200 200 02.81 +6.15 4.13 +2.11 +0.10 -1.58 SHEAR FOR CASE 4 41 Ii I.. - +8.19 +6.8:1
a' 1" Hi lUI H~ -2:00 +l85 .~B4 .3.82 +1.81 -0.21 -222 -4.23 -6.25 -11.7 .7.14 +5.13 +3.11 +1./0
& 2.00 -2.OC -2DO -2.00 ..z.OC ..QZI -0.92 (KIPS) .7.85 +5.84 .3,82 .. L81 -0.21 -222 -423 -6.25 ..s.zs +7.14 +5.13 +3.11 .1.10 -0.92
£ BRIDGE
SEA I I I 1- I I I I I I I I I -I I I I
( I ) (Z) (3) (4JsM (5) (6) (7) (8) (9) (10) (II) (IZ) (13)NM(t4) (15) (16) (17) ONE
40" 3 AT 48" !6i "3Z" 3 AT 48" 2t( 24 3 AT 48" 32" 16" 3 AT 48" 40"
200" 400" ZOO"
-- -'-'----II II
" U
--_ ...... --II II U
FLOOR LEVEL
W 12
BEARING PLATE AND NEOPRENE PAD
---'1.---1..1_ _ __
(a) LOADING SETUP 1
FLOOR LEVEL
II II II U
------- --~~
(b) LOADING SETUP 2
II II II U
W 12
BEARING PLATE AND NEOPRENE ~D
(e) ARRANGEMENT FOR LANE
LOADING
Fig. 7.5. Application of lane load
107
DIAL GAGE
4 8
E w
CASE (1) - 4 DIAL GAGES
DIAL GAGE
4
CASE (2) - 9 DIAL GAGES
SE SS4 SM NM
~NM 100" 100" 77.3"
200"
Fig. 7.6. Locations of dial gage readings
108
some stations, only the readings at the four webs'were averaged for
longitudinal deflections. Also, slip gages were provided to check relative
movement across the critical joint in the maximum shear loading case.
Reactions were measured by load cells under two webs of one box
at the outer piers. Load cells could not be used at the main piers
because of the anchor bolts. Reaction readings from one box were gen
erally sufficient because most of the loadings were applied uniformly
(four lane loads). In the case of nonsymmetric (two lane) loadings, the
loadings were repeated on the opposite side to allow reactions for four
webs to be determined, taking advantage of symmetry.
Prior to any loading test, 1/4 of the service load was pre10aded
in order to get stable initial readings.
7.3 Test Results and Interpretations
7.3.1 General. The results of deflection, strain, and reaction
measurements are compared with solutions of BMCOL5028 and MUPDI34 programs.
BMCOL50 is a beam-type analysis program which solves the linearly
elastic beam or column by a discrete element analysis procedure. This
program takes into account variable loads and nonlinear supports. Load
deflection relations for each neoprene pad were measured and these values
were input to BMCOL50 as the spring constant at the supports.
MUPDI is a versatile generalized elastic analysis program which
can analyze folded plate or box structures with interior rigid diaphragms
or supports using folded plate theories which consider cross section
warping.
Although BMCOL50 can treat variable sections, the section for
MUPDI has to be uniform. This MUPDI limitation should not be serious in
this case because of the small variations in the cross sections (thickened
bottom slabs only at main pier and adjacent segments). BMCOL50 was used
to analyze uniform transverse loadings (four lane loadings) while MUPDI
was used primarily for nonuniform transverse loading as in the two-lane
loading or in the transverse moment study. For a check and direct comparison
of MUPDI and BMCOL50, both were run for one uniform loading case.
109
All external and thickness dimensions for each member of typical
box sections were measured for several segments and averaged section prop-21 erties were calculated by the BOX2 program. These lias buil til section
properties and the measured properties of the materials used in the
theoretical calculations were listed in Table 4.2.
7.3.2 Service Loading
7.3.2.1 Four Lane Loadings. (a) Truck Loading. Four lanes of
truck loadings were applied in the south span, as shown in Fig. 7.7, to
produce the maximum positive moment. Theoretically, additional truck
loads should also have been applied simultaneously in the north side span
in order to get the maximum moment at the critical section. The calculated
effect of loadings in the north span was so small that they were disregarded.
0.888 X 1.222 X 4 = 4.34 K
4.34/'
11l085K
SEA ~ ASM ANM ANE
I ~OIlI32IL~ 80" I I I ~::~_·_··-=~~OO~-_~ ____ .:~:.~ __________ ~4~OO~II __________ ~.~~. _____ 2~0~0_" _____ .~
1.085K FOR REAR WHEELS 0.271 K FOR FRONT WHEELS
Fig. 7.7. Position of truck loading
110
As shown in Fig. 7.8(a), experimental and computed longitudinal
vertical deflections agreed very well. However, transversely, Fig. 7.8(b)
shows that the deflections of the cantilever slabs and the center of the
midstrip closure at the SS7 segment were larger than predicted by BMCOL50,
which does not consider transverse behavior. The deflection/span ratio
under full design live load was approximately 1/7200 in the side span,
which is much smaller than 1/300, which is generally considered as
acceptab 1e.
LOADING CONDITION: (+1.0 oL)
Ht A & • A
SE SM NM NE
<:> EXPERIMENT
SE SS7 EXPERIMENT 0.0153 0.0280
BMeOl 50 0.0160 0.0286
o
; 0.01
z 90.0 l-t) I.&J
~ 0.0
~
SM SMIO 0.0005 -0.0153
0.0007 -0.0175
(a) Longitudinal
5
0.0265 0.0283
_._.- BMeOl50
0-. _. . =&=. - -E>
NM o
0.0001
8
120" 80"
NS6 NE 0.0016 0.0006
0.0024 0.0009
9
BMeOl ~
0.0237 0.0328 0.029'
(b) Transverse
Fig. 7.8. Deflections for truck loadings (four lanes) in side span
Experimental and BMCOL50 reaction results, shown in Fig. 7.9,
agreed fairly well at the outer support of the loaded span and showed
reasonable agreement in the unloaded span considering the very small
magnitude of that reaction.
\ w , w I
111
SE <D (i) TOTAL NE @ ® TOTAL EXPERIMENT (I(J 1.43 1.26 2.69 EXPERIMENT (K) 0.078 0.082
BMeOL 50 - - 2.54 BMeOL 50 - --EXPERIMENT I &MeOL 50 (%) 106 EXPERIMENT IBMCOL 50 (%)
(a) Loaded span (b) Unloaded span
Fig. 7.9. Reactions at outer supports for truck loadings (four lanes) in side span
0.160
0.125
128
Strains shown in Fig. 7.10 showed larger values in the top slab
above the webs than at midspan of the top slab. The bottom slab strains
were almost uniform. The experimental and BMCOL50 results agreed very
well except for the inability of BMCOL50 to treat section warping.
(b) Lane Loadings. Lane loadings were applied for maximum posi
tive moment in the main span (see Figs. 7.11 to 7.13) and maximum negative
moment at an interior pier (see Figs. 7.14 to 7.16).
Comparisons of experimental and theoretical data for service
level four lane loadings show excellent agreement especially for longi
tudinal deflection.
Tensile strains due to the service level live loadings were smaller
than the compressive strains induced by prestressing during construction
so that the sections remained completely in compression.
ll2
LOADING CONDITION: U.
A B a SE SM NM
A NE
-so z ~ _0 __ 0 -0--0-_0_0_0 __ . __ 0 _0_ C/')
w ~o
; t \ S~:R r \ SS7L I ~ ~0~ _____ +~60~~+~5~9 ______________ _
z o
z 0 c;; C/') w a::: a.. ~ u
f
~ z 0 en z w ...
-50
~o ....... :it
2--.. z Ci a::: ... C/')
0
t50
_0_00
0"0
_o __ o __ o __ o __ o_o---'i}o -0--· ---
\ SS6R I
SE 400"
20"
-38 -29
~o--G)o __
NM 200"
o EXPERI MENT
_0- SMeOl 50
NE
Figo 70 .10. Longitudinal strains along SS7 and SS6 for truck loadings (four lanes) in side span
LOADING CON TtON :
l A
SE 1 NM
Q
NE
113
\ E t \ W I
~.- .L~::'::::::.:. ~, e-:::' .. ~ ~ ~ o
z o ~ 0.05 u ILl ...J IL.
~ 0.10
i' \
.r o EXPERIMENT /
_.- BMCOL 50
\_ .. - MUPDl '\, ~.
122.5" 77. 5" .:.~/
193" .I 107" 100" 77.3" 122.7"
EXPERIMEN BMCOL 50 MUPDl
o
~ 0.05
z o ~ 0.10 u ILl ...J IL.
SMlOL
SM10 0.127 0.124 0.129
NM5 NM NS4 0.0781 0.00/6 - .0236 0.0813 0.0016 -0.0256 0.083' 0 -0.0226
l:!:: 0.15 0./260.128 0.127 OJ25 0.126 0.128 0.127 0.127 0.128
Fig. 7.11. Deflections for lane loadings (four lanes) in main span
SE CD EXPERIMENT (10 -0.583
BMCOL 50 -EXPERIMENT / BMCOl 50
Fig. 7. 12.
W
® TOTAL -0.670 -1.253
- -1.24
(0/0) 101
NE
I E
t~
EXPERIMENT (K)
BMCOL 50
(j) -0.658
-EXPERIMENT /BMCOL 50
® TOmL -0.672 -1.33
- -1.24
(Ofo) 107
Reactions at outer supports for lane loadings (four lanes) in main span
NE
114
LOADING CONDITION:
S't s'f!""'ZZt:"'~
z G Q "'tr" .. .... .. , e
(J) -50 (J)
w Q: Q...
~
,. t!l
h NE
~o 0 -50 -59 -56 -6'3 (,)
~ I ..
\ NM:R I 3-z C( Q: l-
t (1)0 0 +83 +85
z 0 en .50 z w I-
_.---f>.-C:> .. _ .. -- .. -+100
WLJ
li) B c:> -" II .--a
-61 -e8 -69 -48
l NM:L I ..
..as +16 +91
e ~'~'-=e-.. . .
«:> EXPERIMENT
_.- BMeOl 50 _ ... - MUPDI
Fig. 7.13. Longitudinal strains along NM9 for lane loadings
(four lan~s) in main span
115
o
~ 0.05 i= o I.IJ ...J ~ 0.10
LOADING CONDITION:
A kIZZttUU!" ltV ,J,lI l SE SM SM~O NM NE ww
. , ..... _.-' --e-. (!J" ......
7'.
" ,
122.5" 77.'t/' 100"
SE SS4 SM
e EXPERIMENT
BMCOL 50
'4i>
" / ....... ""'E>-'" .
93" I 107"
SM5 SMJO
.....--..,-. ....-:"
/ / .
f}1
100" 77.3" 122.7"
NM5 NM NS4 EXPERIMENT -0.0099
BMCOL 50 -0.0106
-0.0207 .0.0016 .0.0699 .0.107
-0.0268 .0.0014 .0.0703 .0.106
.0.0641
.. 0.0620
.0.0025 ·0.005&
.0.0027 ·0.0051
8
o
z =- 0.05 SM10L SM10R
z o ~ ~ 0.10 ...J II.. I.IJ o
-&-'--c!r-'---&' e· e ._.-e--.~ ......... ___ 0.105 0.108 OJ07 0.107 0.106 0.106 O.ro1 0.101 0.108
0.15
Fig. 7.14. Deflections for lane loadings (four lanes) in main and one side span
I SE I []I] , E I \ w f
, E I
, w 1 t(1) 1@ f@ f®
SE <D ~ TOTAL NE (J) (i) TOTAL EXPERIMENT (10 -0.550 - 0.570 -1.120 EXPEFtIMENT O<l 0.!12 0.272 0.584 BMCOL 50 - - -1.08 BMCOL 50 - - 0.516
EXPERIMENT I BMCOL 50 (0/0) 104 EXPEFtIMENT I BMCOL 50 (%) II!
Fig. 7.15. Reactions at outer supports for lane loadings in main and one side span
.~
NE
116
LOADING CONDITION:
s't st" " ~,,i,, '7Ni"'~ NM1 NS1
WLJ -50 "l:T"' ---4). _. ~ . ~. -' -- . ...t:l- . ---L:l.. ~ . -
z . 0 0 en V) w "i 0 a: -42 -49 -46 -54 -51 -50 -55 -38 a..
~ ..
\ N:9R J \ N:9L I • ~
0 (,) 3-
l z Ci a: l-
t V)o D .73 .. 73 .71 .. .75
z 0 en +50 z 0 w 9 . -Ii) a.._._.-Q I-
+50 z ".-.&-.-.~. _._. -e. ~.-.--.~.-& 0 V) z zO w
T ~ .41 .43 .48 D .42 .37 D .43 .43
! 3-z ~ z
Q l-V) V)o V)
NM1R NM1L I .. ..
-63 -54 -60 -~ -1(3 -~ 6
w a: a.. ~ -50 (,) e-. -~-.---0 if)'-'e-'~
+50 -.-.-.~.---e. -'-' -_. __ .---(i) z o 0 0 en z w -'0 I-
~ .48 .48 .. 41 .«> .. 47 D
1 \ NS~L I \ N~lR I
It
3-
l ~ cs: a:
z I-0 V) -60 -65 -60 -58 D en 0 V) w a: a.. ~ -50
~. --<::t._ (,) _. ---11). -' e
G EXPERIMENT
-.- BMCOL 50
Fig. 7.16. Longitudinal strains for lane loadings (four lanes) in main and one side span
117
The deflection/span ratio under full design live load was
approximately 1/3200 in the main span. This value is much smaller than
1/300 which is generally considered as acceptable.
It is very difficult to determine possible warping or shear lag
from the experimental data, because the low magnitude strain readings are
very sensitive to the exact position of strain gages and dimensional
placement tolerances. There was somemdication of shear lag in the top
slab, but it was negligible for design purposes. The twin box sections
were generally acting as a beam and little warping occurred. Comparison
of the predicted strain using BMCOLSO and MUPDI in Fig. 7.13 show this
effect would be expected to be negligible.
Although no special instrumentation was included for the critical
shear stress locations, the maximum shear loading case was also applied.
There was no visible diagonal tension cracking around the NM pier and no
slip at the joints.
7.3.2.2 Two Lane Loadings (Loads on One Box). Two lanes on one
side of the bridge were loaded with lane loadings to produce the critical
moment at the midspan of the main span and then at the main pier. Test
results for deflections and strains were compared with the results of
MUPDI, since BMCOLSO cannot treat unsymmetrical loading across the section.
Comparison of the theoretical and experimental deflection and
strain diagrams (Figs. 7.17, 7.18, 7.20, and 7.21) show that experimental
results and the MUPDI anaylsis agreed extremely well in general. Approxi
mately one-third of the load was distributed to the unloaded box section
at midspan of the main span. Loading only in the main span was more
critical transversely than was loading of both main and side spans. At
the main supports the webs farthest from the loaded box deflected upward
in both loading cases.
Strains across the bottom slab of each box were fairly uniform
because of high torsional rigidity of the box section. However, strains
across the top slab definitely decreased from one side to the other side.
The reactions at the outer supports shown in Fig. 7.19 indicated that the
118
LOADING CONDITION:
W "","t'PP''''k I w I 4 A SE SM S 0 NM NE
., -=-~' .. ··.::::I;~I" 0
Z .~ ..
® .' '. "-"
z \'<. .. ~., .eG0 EXPERIMENT 0 j: o. _ .. - MUPDI u
\~ 1.1.1 ...J 1.1.. '\ ' . 1.1.1 Q
DIAL
0.0 122.5" 77. 5" 193" 107" 100" 77.3" 122.7"
SE
~ ® -0.0068 z &AI ® -0.0072 2 i: @ -0.0013 &AI A. ® -0.0039 )( &AI
® 0
.... ® 0 ~ @ 0 ~ 2 ® 0
0
~
~ 0.05
j: u 1.1.1 ...J LL 1.1.1 Q
0.10
EXPERIMENT
MUPDl
SS4 SM
-0.0140 .0.0029
-0.0150 .0.0007
-0.0067 .0.0002
-0.0089 -0.0006
-0.0137 0
-0.0143 0
-0.0083 0
-0.0090 0
SM 10L
0.0931 0.0913 0.0875
0.0926 0.0895 0.0858
SM10 NM5 NM NS4
.0.0913 .0.OM6 .0.0019 -0.0125
.0.0834 .0.0465 .0.0004 -0.0118
.0.0463 .0.0239 .0.0013 -0.0089
.0.0363 .0.0227 -0.0005 -0.0091
.0.0895 .0.0588 0 -0.0137
.0.0811 .0.0524 0 -0.0143
.0.0479 .0.0316 0 -0.0083
.0.0394 .0.0251 0 -0.0090
/ .'
,./ ~ EXPERIMENT
.' _ .... - MUPDI
0.0834 0.0ID3 0.0463 0.0408 0.0363 0.0339
0.0811 0.0146 0.0473 0.0430 0.0589 0.0368
N£
-0.0057
-0.0064
-0.0033
-0.0036
0
0
0
0
Fig. 7.17. Deflections for lane loadings (two lanes) in main span
WADING CONDITIO,:
15 RZ ZZlZZI{lZZ Z? l SE SM NM9 NM
-50 -z 0
0
A NE w
0 en
____ 0
-" "--a-.. __
"---"-en I&J a:: ci--&-
119
a.. 2 0
\l)"~
0
1
j Z 0 in z I&J ~
~o "-
-40 -44 -34 -40 -20 -18 -14 -10 z 1 ~ T 3-z NM9R NM9L « a:: ~ en o
\ j .5-8 .62 • f4 .21 +~ 8
0 <.::L.- .. - .. -.(i)
0 E)cPER IMENT +50 ~
MUPDI _ .. --" ....e- .. --t!)
Fig. 7.18. Longitudinal strains along NM9 for lane loadings (two lanes) in main span
rnJ
1 E , l WI I f(D t@ f@, t(D'
co (2) C~I (1)' EXPER. (K) -0.328 -0.412 -0.240 -0.239
(%) 26.9 33.8 19.7 19.6
~
I E I , w I
f@ t@ t@1 t@,
@ ® (9' @' EXPER (K) -0.392 -0.402 -0.267 -0.268
(4Yo) 29.5 30.2 20.1 20.2
Fig. 7.lQ. Reactions at outer supports for lane loadings (two lanes) in main span
TOTAL
-1.219
100
TOTAL
-1.329
100
120
LOADING CONDITION: w s1 st"'~t;L,~t L'~E
z o i= (,)
o
~ 0.05 LL. LLI o
~ ® A::~ .. ·j/·:.;::r>; •• 80 EX ERIMENT .. ~ J -. - MU D[
.~~ .. -\§)
122.5" 77.5" 100" 93" 107" 100" 77.3" 122.7"
I- ® z ® w
:E i: ® If @ )( w
® - ® 0 G.. ® :,:, :::tI @
SE -O.()()33
-0.0025
-0.0022
-0.0050
z o i= (,)
LLI
0
0
0
0
o
~ 0.05 LLI o
EXPERIMENT
MUJtOI
SS4 SM SM5 SM10 NM5 NM NS4 NE -0.0080 -0.0005 .0.0193 .0.0298 .0.0115 -0.0010 -0.0046 -0.0029
-0.0060 .0.00 16 .0.0234 .0.0388 +0.0232 .0.0010 -0.0032 -0.0001
-0.0138 .0.0020 .0.0455 .0.0681 ..0.0386 .0.0010 -0.0005 .0.0029
-0.0121 .0.0018 .0.0513 .0.0115 .o.oaol .0.0038 .0.0031 .0.0061
-0.0080 0 .0.021' .0.0339 .0.0209 0 -0.0063 0
·0.0013 0 .0.0211 .0.0411 .0.0210 0 ·0.0048 0
-0.0129 0 .0.0461 .0.0113 .0.0461 0 ·0.0014 0
-0.0123 0 .0.0526 .0.0191 .0.0521 0 0 0
8 9
-!L ...... 'L..---A __ ....I
.. --.. - . <:> EXPERIMENT
_.,- MUPDI ................
.......... .
0.02160.0296 0.0321 0.0368 0.0522 0.0698 0.0139 0.0115 0.0804
0.0315 0.0339 0.0312 0.0411 0.0566 0.0113 0.0156 0.0191 0.0819
Fig. 7.20. Deflections for lane loadings (two lanes) in main and one side span
z o Cii
LOADING CONDITION;
Ii SE
-50
+50
sti"'" '~~':'~ II~E
--" -e- .. --£)
CD
e ~ .. -.c.:>
_--- .. .l:i:.QlIo---G> z ..a.-. .. _.. ---1!)"
~1 z 0 1----l~==..;7---:-=--_::_---__:_=--l._:__----...%----" .4.6 .13 D +18 +19 D .. 36 +38
~ i · t N~1R I · \ N:1L J · ~ tn 0 1-___ 6..:,16.::....-_---...:-1:.:..1 __ -...::9 ______ -:.:5:..:.7_--.:-:.::5:::.2_-=-::::53~ __ _
~ c:l- .. ~ .. ~ ~ o o
+
-e- -B-- .. -.....- •. ....(i)- .• -" -e---" ..
z o Cii z T ~ 0 1---:;;;;;;..;.....-.-. ----.. -.-:'12=----.1:-::7-----.-2..1.2---.. -28---+38~--D--
~ f \ N:L I l NS~R I · ~ CI) 0 1--______ ....;-1:.::3 __ -...:..:13:..-_____ .::...::5::..:.1 _____ -.::.4.::.4_----!D~ __ _
~ c.:> e If -"-"-" ~
8 -50 e ~~ . ......--.. .. -
121
e EXPERIMENT
MUPDI
Fig. 7.21,. Longitudinal strains for lane loadings (two lanes) in main and one side span
122
LOADING CONDITION:
A 1''''''' ""1,,,,, i 1 ""1 SE SM NM NE
\ w t@,
SE <D ® ®' (J)' TOTAL EXPERIMENT (K) -0.2:59 -0.197 -0.:547 -0.:514 -1.097
(%) 21.8 18.0 :51.6 28.6 100
\ E
t@ NE ® ® ®' (I)' TOTAL
EXPERIMENT (K) -0.170 -0,067 0.:554 0.486 .. 0.158:5 C%) -29.1 -II.S 57.:5 8:5. :5 100
Fig. 7.22. Reactions at outer supports for lane loadings (two l~nes) in main and one side span
123
loaded box is experiencing little twist because reactions under the two
webs of the box section were almost equal when loading was only in the
main span. When the loads were applied in both the main and side spans,
Fig. 7.22 showed the end reactions were not uniform under each box of the
loaded side span and showed substantial twist effects. However, the
reaction under each box was reasonably uniform at the end support of the
unloaded side span.
Although the support conditions and section properties input into
MUPOI differed slightly from the actual model, the MUPOI results agreed
very well with the experimental results. Therefore, MUPOI can be used in
design to predict the longitudinal load distribution for nonuniform
loading. Transverse moments from these loadings are discussed in
Sec. 7.3.4 and are compared with MUPDI results.
7.3.3 Ultimate Design Loading Specified by BPR Criteria
7.3.3.1 Additional 0.35 OL. Ultimate design dead load is speci
fied as 1.35 OL by the BPR criteria.7
The additional 0.35 DL was applied
to the completed structure, as shown in Fig. 7.23, by using additional
concrete blocks to supplement the previously adjusted 1.0 DL segment
weight. These blocks were permanently added to the bridge, so 1.35 OL
was the effective dead load in all later cases.
The experimental results are generally slightly larger than the
theoretical values for both deflection and reactions, as shown in
Figs. 7.24 and 7.25. Strain readings show considerable scatter and do
not agree with the theoretical values, as shown in Fig. 7.26. The maximum
moment due to this loading is less than that due to the service live load
producing maximum positive moment at midspan of the main span. Thus, the
bridge should still be in the elastic range. The large deviation of the
experimental strain results from the theoretical values is probably due to
temperature and time variations in the electrical data record instruments
(it took about 10 hours to hang all concrete blocks).
7.3.3.2 Maximum Positive Moment, Main Span. Live loading, as
shown in Fig. 7.27, was applied and incrementally increased to design
124
0.0151 K
0.025 K
CONCRETE BLOCK
SERVICE OEAD LOAD BlOCKS
V77J ULTIMATE DEAD IliL1 LOAD BlOCKS
Fig. 7.23. Concrete blocks for 0.35 DL
"
125
z 52 O. ~ o w -'
o
~ 0.10 o
LOADING CONDITION:
is: SM
_.- ....... I,. '""
,0.36 ilL.
i NM
~. , " ~.
A NE
" ...... .; ......... ..."., .
. .~
/
<:) EXPERIMENT
BMCOll5O
. -'"1 1)- .
" ~
/ /
62.5" 137.5" 100" 93" r 107" 100" 17.3" 122.1"
SE SS7
EXPERIMENT 0.0058 0.0024
BMeOl 50 0.0048 0.0000
SM SM5 0.0024 0.0736
0.0021 0.0682
SM10
0.115
0.107
NM5 NM NS4 0.0749 0.0018 -0.00& 1
0.0632 0.0021 -0.0064
Fig. 7.24. Deflections for 0.35 DL
I \ w I t~
SE <D (i) TOTAL HE <I> ® TOTAL EXPERIMENT (K) 0.700 0.636 1.336 EXPENMENT(Kl 0.D78 0.623 1.199
BMeOl5O - - 1.09 BMCOl. DO - - 1.09
(EXPERIMENT/SMeOL 5O)XfOO (%) 123 (EXfllfftMENT/8MCOl5O)X1OO (%) 110
Fig. 7.25. Reactions at outer supports for 0.35 DL
NE 0.0056
0.0048
126
z 0 U) U) I.aJ
f ~ (.)
1
I z Q U) Z I.aJ I-
~ (i) z I.aJ I-
1 l z 0 (i) U) I.aJ a:: Q..
~ (.)
LOADING CONDI TION:
-50
~o .::; z
3-z <{ 0::
tio
+50
+1
+50 r-
~o .::; ~
3-~ <{ 0:: I-U)o
-50 '-
50
0
0
50
0.3S Ql.
gli I 'iii'"~ "!i/'li"~""P,! SE SM NM9 NM1NMNS1 NE
• e EXPERIMENT
BMeOl ~
,(D..._._. __ ._._._. _. _.G)-. --e'--0 e
e e -so -81 -15 -26 -26 -49 -44 -15
1(
\ NM:R J \ N~9L 1 x
0 .84 +90 .97 +105 +116
_.--,.,-'-e
• • _. _._._._. __ .---. __ ._._.-
• • • +71 +47 +11 0 0 +49 +16 0 +22 +58
1 n
NM1R NM1L
I. -27 -40 -45 .; -65 -75
• • • _._ .. _-- _.-.--• ~
0 -' -' .-_._- . __ . __ . __ . __ .-
0 e G
+52 +59 +36 +1 +26 0 n
NS1L NS1R
0 -; 4 .54 -64 0
0 e _ ._._- _.c:L... _
Fig. 7.26. Longitudinal strains along NM9, NMl and NSl for 0.35 DL
I 225
- 2.0 0
9 I-0
f :E 1.5 +
lLJ > ..J -u.. 1.0 0 I-Z lLJ :E lLJ It: 0 0.5 z
/" /'"
/'"
/'" lOADING CONDITION: ( .. 1.35 Dl)
"a /" /'" st sttllll~f~ till! ~M 0. NE
1.350LO 1C.-__ /" _____ --L ________ -1-________ ....l-________ ..l..-___ _
o 0.1 0.2 0.3 0.4
DEFLECTION (lNJ - .............
Fig. 7.27. Deflections at SMIO for lane loadings (four lanes) in main span (design ultimate)
...... N -.J
128
ultimate load levels. As shown by the deflection diagram of Fig. 7.27,
the measured deflections at midspan in the main span were larger than
the theoretical values calculated by BMCOL50. The load-deflection
relationship at midspan of the main span was linear up to the [1.35 DL +
0.75 (LL + 1L)] increment and then started to change slope as the load
increased. An appreciable change in slope was noted with formation of
the first crack at [1.35 DL + 1.75(LL + 1L)]. The increase of deflection
at the increment of [1.35 DL + 2.25 (LL + 1L)] was very large and the
outer supports raised up suddenly, as shown in Fig. 7.28. The deflection
at the midspan of the main span did not immediately come back to its
original value after releasing the load. However, it recovered almost
completely after several hours.
Figure 7.29 shows the increase of end reactions was linear up to
the {1.35 DL + 1.25 (LL + 1L)] increment and agreed with the theoretical
values very well up to that point. The actual reactions become zero at
about 2.13 (LL + 1L), which is greater than the actual design ultimate
if advantage is taken of the 25 percent design load reduction for a four
lane bridge.
From the strain diagrams shown in Fig. 7.30, it can be seen that
the rate of strain increase did not change very much up to the [1.35 DL + 1.5 (LL + 1L)] level. Strain at the outer edges of the top slab was con
sistently less than the other positions. Strains along the bottom slab
in the service loading test (Fig. 7.13) were almost uniform, but there
was some deviation in this test even when the amount of live load was
small. Strains in the bottom slab (tension side) started to deviate sub
stantially at the [1.35 DL + 1.50 (LL + 1L)] increment and varied widely
at the [1.35 DL + 1.75 (LL + 1L)] level.
As shown in Fig. 7.31, a crack appeared at the center of the
closure segment (OE side) at the fl.35 DL + 1.75 (LL + 1L)] increment
and developed along the joint of the closure segment. On the OW side,
the first crack appeared at the fl.35 DL + 1.88 (LL + 1L)] increment.
At the [1.35 DL + 2.25 (LL + 1L)] increment, cracks extended to the
midheight of the webs.
-0.10
o
0.10
Q
-z z 0 ~ 0.30 u UJ ...J LL. UJ 0
0.40
(LL. IL)
1.00
1.50
2.00
2.125
2.2~
LOADING CONDITION: (+ 1.35 OLl I
6. K"ZiZl"""'i SE SMW hJ A
NE
.IL)
129
EXPE IMENT
12Z.5" 77.5" 100" 93" 107" 100" 77.3" 122.7"
SE SS4 SM SM5 SM10 NM5 NM NS4 NE
-0.0041 -0.0229 .0.0013 .0.086 .0.131 .0.083 .0.0014 -0.0239 -0.0049
-0.0019 -0.0361 .0.002~ .0.136 .0.224 .0.134 .0.0020 -0.0381 -0.0090
-O.OISS -0.0&~3 .0.0048 .0.203 .0.348 .0.202 .0.0032 -0.0:563 -0.011S
-0.0229 -0.0611 .0.00152 .0.221 .0.382 .0.221 .0.0038 -0.0631 -0.0321
-0.101 -0.0948 .0.00~9 .0.271 .0.469 .0.272 .0.0040 -0.1038 -0.128
Fig. 7 •. 2& Deflections for lane loadings (four lanes) in main span (design u1 tima te)
130
LOADING CONDITION'(.1.35 oLl
Il: SE
.. SM
EXPERIMENT .5" «D.®I
1
"
Hlill liJ <D <I>
(ll. -3.0
1.35Dl+(ll,rLl 0.19
\
A NM
" "
-2.0
1.19
"
A NE
'\
"
1.0
219
REACTION (KIPS)
;:::! . ..J ~
2.0
1.5
1.0
0.5
0
319
~~ :t\\ \ ·~BMCOl ~O +2.0
\ .
\
" EXPERIMENT>
I@'@l " " " " " LiJ[@liJ " " ~ t@ t@
(LL. -1.0 0
1.35 Dl • 005 105 2.05 305 ILL • ILl
REACTION (KIPS)
Fig. 7.29. Reactions at outer supports for lane loadings (four lanes) in main span (design ultimate)
IN) __ -
(0) Ie) Ie) (AI O()(I'II (IJ (JI (K)
"S "S" (<<) IF) (EI ILl (M)(NI
o TENSION--
-100 -COMPRESSION
STRAIN If IN.lIN.l
Fig. 7.30. Longitudinal strains along NM9 for lane loadings (four lanes) in main span (design u1 Uma te)
.
<C~
LOADING CONDITION: (+ 1.35 DL
Jl "'" vi "'" "1 SE SM S~IO NM
rlL--E---J1 OE
Z\ NE ww
, W
* NUMBERS ALONG THE CRACK ARE MULTIPLES OF (LL .. III
SM9 SMIO NMIO NM9
225 2.0
\~. 1./15 12.25
ON THE SURFACE OF OE
NM9 NMIO SMIO SM9
Z25
ON THE SURFACE OF OW
131
<C
Fig. 7.31. Development of cracks for lane loadings (four lanes) in main span (design ultimate)
:..
132
The theoretical cracking moment at the closure segment, as 2
predicted by the 1963 ACI Building Code, is calcula ted as follows:
where
M Wb ( f + 6 JT') - M cr pe c s
1230 (1.34 + 0.505) - 740
1529 k-in.
Wb = section modulus at bottom
f = compressive stress of concrete due to prestressing only pe at bottom fiber
f' = compressive strength of concrete c
Ms moment due to end reaction caused by prestressing of the positive tendons in the main span and seating force at the ou ter suppor ts
Therefore, the LF (load factor) of (LL + IL) for the load at which
cracks might appear can be calculated as:
M cr
1529
1.35 ~L + LF X M(LL + IL)
= 481 + LF X 552
LF = 1048/552 1.90
Since initial cracks appeared on one web at the LF = 1.75 level and on
the other web at the 1.88 increment, the computed value [1.35 DL +
1.90(LL + IL)] was very accurate.
This method of calculation shows that the cracking moment is
greatly affected by the adjusting force at the end supports. If the
reaction force provided at the end supports is large, midspan cracks will
appear at lower increments of (LL + IL). If the adjusting force provided
at the end is small, the end segments will raise up from the neoprene
pads under very small increments of (LL + IL). Therefore, where possible
the end reactions for the prototype bridge should be selected at an optimum
point which balances these two factors. Figure 7.32 shows the relation
between the initial end reaction applied, the LF of (LL + IL) at which the
end segments will raise up from the neoprene pads and the LF of (LL + IL) at
which first cracks will appear. Therefore, 4.9 kips (176 kips in the
'::i ...... +
....I
....I ~
4.0
3.0
\ I \ I ~
TOTAL
133
'0 2.0
IJ... ....I
1.0
o o 1.0
ZERO REACTION AT END SUPPORTS
(1.0DL+LF(LL+IU)
2.0 3.0 4.0 5.0 6.0 7.0
INITIAL REACTION DUE TO PRESTRESSING IN MAIN SPAN AND JACKING UP FORCE AT END SUPPORTS (KIPS)
Fig. 7.32. Relation between the end reaction and cracking moment
....I ...... ..
....I ....I -
2., 2.0 ....
1.5
ID~ I
0.5
/ /
/
/
1//// -7
7
SMCOl50' /
~// "'AD'"' CON"'TlDN' .1.3 Dl
J/ /
/ / / SE st" " ~J:~ L. I!~A~! 1 It ~"E
LdhJ
o~/ _________ ~ __________ ~ ________ -L'~ ____ ___
o 0.1 0.2 0.3
DEFLECTION (INJ
Fig. 7.33. Deflections at SM10 for lane loadings (four lanes) in the main and one side span (design ultimate)
134
prototype bridge) total for two boxes is the optimum initial reaction
at each outer support. Although the weight of the asphalt topping,
which is about 8 percent of the weight of the segment, was not included
in the model bridge test, the effect of the asphalt is included in the
dead load calculations on which Fig. 7.32 is based.
7.3.3.3 Maximum Negative Moment at the Main Pier. Under this
loading condition the experimental deflection at midspan in the main
span was slightly larger than the theoretical values (BMCOL50), as
shown in Fig. 7.33. Increase in deflection was linear up to the
[1.35 DL + 0.75 (LL + IL)] increment. Increases of deflection at the
unloaded span outer support (SE) became rapid at the [1.35 DL +
2.12 (LL + IL)] increment, as shown in Fig. 7.34.
The trend of strains at the NM9 segment, as shown in Fig. 7.35,
was the same as noted in Sec. 7.3.3.2. Strains in the top slab increased
almost linearly up to the I1.35 DL + 2.25 (LL + IL)] increment, but the
strains in the bottom slab were linear only up to the [1.35 DL + 0.5(LL +
1L)] increment and started to deviate after that increment. The strains
in the bottom slab showed considerable difference between the webs and
the middle of the bottom slab. In the NMl segment, strains in the top
and bottom slabs increased linearly until [1.35 DL + 1.50(LL + IL)] and
[1.35 DL + 1.75(LL + IL)] increments, respective, as shown in Fig. 7.36.
Reaction at the NE support increased linearly until the {1.35 DL +
1.00 (LL + IL)] increment and that at the SE support decreased linearly
up to the [1.35 DL + 1.50(LL + IL)] increment, as shown in Fig. 7.37.
Besides reopening positive moment zone cracks developed in the
earlier test (Sec. 7.3.3.2), cracks appeared along the trajectories of
the negative moment prestressing cables in the 2nd and 3rd segments from
the NM pier, as shown in Fig. 7.38. Appearance of these cracks along the
tendons was probably due to a combination of high bending moments and
shears around the NM pier and relatively thin covers in relation to the
large diameter of the tendons (3/8 in. diameter) which were used to con
struct the 2nd and 3rd segments. It is believed that NMl and NS1, which
~
Z
z 0 j: (.) I&J ...J u.. I&J Q
-0.10
o
0.1
O.
o.
IS. SE
122.~ II
135
Ii' , " , , ,.t , , , i' J, "! WLJ SM NM NE
EXPE IMENT
77.~" 100" 77.3" 122.7"
(LL+IU SE SS4 SM SM5 SMIO NM5 NM NS4 NE
1.00 -0.0035 -0.0203 .0.0007 .0.0727 .0.113 .0.0641 .0.0023 -0.0032 .0.0014
1.50 -0.0073 -0.0311 .0.0011 .0.1150 .0.184 .0.1060 .0.0038 -0.0063 .0.0022
2.00 -0.0212 -0.0471 .0.0019 .0.1620 .0.266 .0.1510 .0.0050 -0.0098 .0.0031
2.125 -0.0464 -0.0!S79 .0.0027 .0.1790 .0.293 .0.1660 .0.0~3 -0.0117 .0.0034
2.25 -0.0921 -0.0756 .0.0028 .0.2020 .0.327 .0.1840 .0.0058 -0.0136 .0.0035
Fig. 7. 3~ Deflections for lane loadings (four lanes) in the main and one side span (design ultimate)
136
2.25
2.13
2.00
,p='lItllJ."II lI" \" u. lOADlNG CONOITIOW(+1.35 OU N.E_{d.·_··~··.· ....... _.~:TJ
SE SM NM
(M)(F) (E)
()()(H) (I) (J) !I<)
· El" M) (F)(E) (!.l(M) IN)
---.IL6oc-------~-~- --..... -----.......... .':.~ ... -_. __ . --2-~00~---····----··----COMPRESSION TENSION----
STRAIN I~ IN/INJ !
Fig. 7.35. Longitudinal strains at NM9 for lane loadings (four lanes) in the main and one side span (design ultimate)
~ ..J
::::!
225
2.13
2.00
L75
1.50
(Mel tS) OOOOU>)(H) (J) (.11 (K) 1.25
·S .. S" {(;) (F)(EI (Ll {M)(NI
1.00
STRAIN If IN/IN,) -COMPRESSiON TENSION-
Fig. 7.36. Longitudinal strains at NMI for lane loadings (four lanes) in the main and one side s?an (design ul tima te)
(pI ".
(LL.IL) -2.0
1.3& DL. 1.19 (LL.IL)
LOADING CONDITION: (+ 1.35 DL) i ,
i' , , , , , ' , it 'v " 1 tE SM NM NE
-L5
1.&9
-1.0
2.19
REACTION (KIPS)
-0.5
2'&9
LdW -~ .... + ~ ~ -
Z.Z& 2.13 2.00
1.88
1.75
I. !SO
1.25
1.00
0.75
0.50
o 5.19
-~ .... + ~ ~ -
EXPERIMENT ~BMCOL 50 (@+®) 'f
f
\ E
t(j) w
O~ ________ ~ ____________ -L ______________ -L ________ ____
o 3.05
0.& 3.5&
1.0
4.0&
REACTION (KIPS]
1.5 (LL. IL)
4.&& 1.58 DL. (LL.IL)
Fig. 7.37. Reactions at outer supports for lane loadings (four lanes) in the main and one side span (design ultimate)
LOADING CONDITIONS: (+ 1.35 OL)
... ~
.... >
IS. SE st" "" .t ~~,t~sa ~
NM2 NS!
NM3 NM2
~O ~ 2.0
I- ... .c ..
ON THE SURFACE OF OE
NS3 NS2 ......, -
~ 2 2.0
..c"
ON THE SUR FACE OF OW
NS2
NM2
M NUMBERS ALONG THE CRACKS ARE NUL TIPLES OF (LL. IL)
NS3
~~ ..c -
NM3
2.25
Fig. 7.38. Development of cracks for lane loadings (four lanes) in the main and one side span (design ultimate)
139
were subjected to higher moments and shears, did not have such cracks
because of double tendons and the smaller size of tendon ducts. Because of
high compressive stress due to negative tendons in the top slab around
the main pier, no flexural cracks appeared around the main pier. There
fore, this loading condition was not critical.
7.3.3.4 Maximum Shear Loading Adjacent to the Main Pier. No
special strain instrumentation was provided for shear loadings. Design
loadings for maximum shear were applied until the [1.35 DL + 2.25 LL(LL +
IL)] ultimate design increment.
The moment diagram for this loading is very similar to the maximum
negative moment loading case in Sec. 7.3.3.3 and the magnitudes of the
maximum positive and negative moments for this loading are only about
17 percent less than this previous loading case. Therefore, the experi
mental results were very similar to those of Sec. 7.3.3.3, except slightly
reduced at each level of load. Because one of the major questions was the
dependability of the joints under high shear, this type of loading was
applied to the bridge in a later test to failure.
No additional flexural or diagonal tension cracks were observed
in this test. Slip gages set across the critical first joint in the main
span (as shown in Fig. 7.39) showed zero movement during the loading test.
DIAL GAGE (NO CHANGE IN READING DURING THE DESIGN ULTIMATE TEST)
JOINT NM1 NS1
MAIN SPAN MAIN PIER
Fig. 7.39. Arrangement of slip gage at the first joint
140
7.3.3.5 Maximum Positive Moment in Side Span. Truck loads
(four lanes) were applied at the same longitudinal position as shown
in Fig. 7.7, but the transverse positions were changed to those shown
in Fig. 7.40.
Deflection at the center of the SS7 segment, as shown by Fig. 7.41,
was exactly the same as the theoretical values at the [1.35 DL +
0.5(LL + IL)] increment. The experimental deflections became larger
than the theoretical values after this point. The experimental deflec
tion was about 14 percent larger than the theoretical deflection at the
[1.35 DL + 2.25(LL + IL)] increment. Also, the transverse deflection at
the center of the box section (3), in Fig. 7.42, was increasing as the
load increased. Details of transverse moment distribution are discussed
in Sec. 7.3.4. Observed strains and reactions, as shown in Figs. 7.43 to
7.45, were generally linear. No cracking was observed during this loading
test.
Transverse strain was measured at the center of the top slab of
SS7R and compared very favorably with MUPDI results, as shown in
Fig. 7.46. Transverse strain increased almost linearly up to the
2.25(LL + IL) increment and agreed with MUPDI.
7.3.4 Study of Transverse Moment. It is very hard to accurately
simulate the behavior of the prototype bridge transversely because of the
increased difficulty in reducing the Rcale correctly for the very shallow
slab sections used.
Transverse strain gages were put on the surface of the segments
at some points and these strain readings were compared with MUPDI analysis
results.
The loading cases considered in this section are shown in Fig. 7.47.
Since the transverse effect of dead load is small,2l (LL + IL) only was
considered at the service load level. Experimental transverse deflections
agreed very well with MUPDI for case (3), as shown in Fig. 7.17. Transverse
gage locations on the SS7 segment are shown in Fig. 7.47. Experimental
strain readings agreed very well with MUPDI results, as shown in Table 7.3.
...J ... •
....I
....I
2.25
2.0
U5
K K 4.34 434
1 r j~.085K SEdl.--l~.J .J ------oA~SM ~~_NM 6L~."INE L 60" 321~8j 80" I
~ 200" :.. 400" 200"
Fig. 7.40. Position of truck loadings (design ul tima te)
LOADING CONDITION: (. 1.35 DL
.. ~. A
SE SS7 SM A
NM
•• COC'~/// V
/
b / ~EXPERIMENT
f// . /
1.0 /
0.5 ;-
/
/ /
/ . /
./' /
14Jl
O~~L-~ ______ -....I~ ____ ~~ ____ ~~ ____ ~~ __ __
o 0.01 0.02 0.03 0.04 0.05
DEFLECTION (IN,)
Fig. 7.41. Deflections at SS7 for truck loadings (four lanes) in side span (design ultimate)
142 -0.080
-z
z o .... o L&J ...J I.L L&J o
(LL+ IL)
1.00
1.&0
2.00
2.12&
2.25
-Z
z o .... o L&J ...J I.L L&J o
SE SS7
.0.0082 .0.0228
.0.0130 .0.0362
+0.0184 .0.0495
.0.0198 +0.0530
.0.0209 .0.0668
o
0.01
--e- EXPERIMENT
LOADING CONDITION: (+ 1.35 DL
Ht st S\1 a
NM
100" 93" 107" 100" 77.3" 122.7"
SM SM5 SMIO NM5 NM NS4 NE
.0.0003 -0.0121 -C.0119 -0.0020 .0.0002 .0.0010 .0.0002
.0.0006 -0.018& -0.0181 -0.0031 .0.0001 .0.0017 • 0.0005
.0.0012 -0.0247 -0.024& -0.0041 0.0000 .0.0025 .0.0010
.0.0016 -0.0262 -0.0257 -0.0043 -0.0001 .0.0026 .0.0010
.0.0017 -0.0278 -0.0274 -0.0047 .0.0001 .0.0028 .0.0010
SS7L
EXPE IMENT
Fig. 7.42. Deflections for truck loadings (four lanes) in side span (design u1 tima te)
.
(H)
143
= . ...I d
,L_OA_1_1N_1G_CoN_O_IT_10_N_' _('_I'_35_0_L_) ___ ,_R.'_'-E_9~_-"_ DwA I _ s'E SM N"M N"E
(B)(A) Z.Z5
(El
2.00
1.75
1.50
1.25 (Sl (AI
1.00 §] (Fj(E)
-50 -COMPRESSION TENSION-
STRAIN It IN! IN.!
Fig. 7.43. Longitudinal strains at SS7 for truck loadings (four lanes) in side span (design ultimate)
d . ...I :::!
(II 2.25
2.13
2.00
1.75
1.50
(HI m 1.25 8
(L1 (M) 1.00
0.75
-ao o -COMPRESSION TENSION-
STRAIN l.f IN/IN.!
Fig. 7.44. Longitudinal strains at SS6 for truck loadings (four lanes) in side span (design ultimate)
(F)
144
:; ... + ::l
2.25
2.13
2.00
1.75
LOAOING CONOITION: (+1.35 OLl
II. moo
. ;/
BMCOl.50R/' . ,....
1.50 /'
. ;/e--' EXPERIMENT (<D+@) /
EXPERIMENT (@+@l
I.U
1.00
0.75
0.50
/ /
/
",/
/if /
/ /
/ / ,
U ~....L_--:-=-___ :-':-:::--__ -::-':-::--__ ----':,-__ ----'L ___ --'-ILL. ILl 0 1.00 2.00 5.00 4.00 &.00 6.00
1.2&
1.00
0.
o 1.35DL. 3.18 4" • II
ILL. ILl . ~. 8.1' 7. " 8.11 9.1' 5.05
l_
1.00
4.05
---L-_ 2.00
5.05
REACTION (KIPS) REACTION (KIPS)
Fig. 7.45. Reactions at outer supports for truck loadings (four lanes) in side span (design ultimate)
;::::! . ....J ....J
2.2
2.
1.75
1.50
1.25
1.00
0.75
O.
LOADING CONDITION: (.1.35 DL
HI
o -100 COMPRESSION -
moo POSITION OF WHEELS
14" 12" 8" 12" ) I I I I 'I i. BRIDGE
++ ++ !
EXPERIMENT
SS7
08.1 "r .. R~TlON OF PAPER GAGE ( TRANSVERSELY)
275"
!
SEi~:_0'_'1£2Q~~L~_"~:+[~SM ____ ~4~0~0'_' _____ ~~I._~2~00~ .. ~]NE
-200 -300
STRAIN 't IN.lIN.1
Fig. 7.46. Transverse strain at 887 R for truck loadings (four lanes) in side span (design ultimate)
CASE (1) • DETAILS OF LOADING ARE IN SEC. 7. 3.2.1,
HI 4 A li ..
CASE {21 • DETAILS OF LOADING ARE IN SEC. 7.3.3.5.
2S b
CASE (31- DETAILS OF LOADING ARE IN SEC. 7. 3.2.2.
J 8 A
CASE (4) • NOT TESTED. , A ti:J
E ( It OF SS7R)
Fig. 7.47,
CASE (11
CASE (2)
CASE (~)
CASE (4)
M.' LONGITUDINAL SLAB MOMENT PER UNIT WIDTH
M,=TRANSVERSE SLAB MOMENT PER UNIT WIDTH
~O.026
Transverse slab moment eN) diagram for different loading cases as computed by HUPD!
I t8RIDGE
'1. BRIDGE
I
'1. BRIDGE
I
C1. BRIDGE
I
UNIT: K -IN.IIN.
146
Loading Case
(1)
(2)
TABLE 7.3. COMPARISON OF EXPERIMENTAL AND THEORETICAL TRANSVERSE STRAINS
Strain (~ in. lin.) Qlge No. Experiment/MUPDI
Experiment MUPDI
(2) 64 63 1.02
(1) 122 107 1.14
Because MUPDI showed good agreement with the experimental results,
transverse moment diagrams for each case were drawn (Fig. 7.47) using
the MUPDI results. The computed longitudinal slab moment (M ) at the x
location of the maximum transverse moment is also shown in Fig. 7.47.
Among these four cases, case (2) gave the largest values in transverse
positive and negative moments. In order to judge the transverse strength
of the section, several strain gages were put at these critical positions
in later punching shear tests (punching shear test results are given in
Sec. 8.4) and strain readings were taken almost to failure. Transverse
and longitudinal slab moments at the service load level were calculated
from the experimental results for the loading cases shown in Fig. 7.48.
These moments are much larger than the values calculated for the above
four cases and strains increased linearly almost to the punching shear
failure loads which were about 18 and 7 times (LL + IL) at the middle of
the twin boxes and the edge cantilever, respectively. Also, loading
case (1) in Fig. 7.47 was applied at the 5.25(LL + IL) increment in the
failure test of Sec. 8.2 and no visible crack was observed in the top
slab. Although it is not possible to relate these results directly to
the prototype because of casting tolerances at this section being exceeded
in the model (the thickness of the top slab was about 15 percent thicker
than thickness specified), the above results indicate that there is ample
safety in transverse bending for the top slab.
..... .... .... .. .... .... ...... 9 II..
\ 0 II! -...i Go
iC 2Q ......
0
9 10.
15.0
10.0
5.0
5.0
o
P.0.578 KIPS AT 1.0 (LL+lL)
1 ! , 1
LOADING PAD kGAGECtI l I GA8E (21
LONGITU DI NAL
0.001
AT '.0 (LL.IL) .... 0.282 K-IN/IN. ",.0.212 K-IN./IN.
0.002 0.003
STRAIN (IN/IN]
-.... ... .. .... .... .... g !; II! ...i
15.0
10.0
-Go iC ...... 0 c 9
5.0
... : LON81TUDINAL MOMENT K-IN./IN. II,: TRANSVERSE .. OMENT K-IN./IN.
P.O.f578 KIPS AT 1.0 (LL.IL)
! ~1\ 1
GAGE
AT '.O(LL.IL) 11,1: - 0.354 K-IN./IN.
0.001 Q.OO2
STRAIN (1N.l1 NJ
Fig. 7.48. Typical strain readings in punching shear tests
148
7.4 Summary
The excellent agreement obtained between analytical and experimental
results at the service live load level indicates that this structure can
be analyzed very accurately for transversely uniform loading by an elastic
beam-type analysis, such as BMCOL50. When loaded nonuniformly in the
transverse direction, the beam-type analysis is not applicable. For this
case, the elastic folded plate type analysis such as MUPOI showed excellent
agreement. The effect of shear lag in this section was small.
The structure behaved in a most acceptable manner under design
ultimate load conditions [1.35 OL + 2.25(LL + IL)]. If advantage is taken
of,the 25 percent load reduction in a four lane bridge, this load corre
sponds to [1.35 OL + 3.0(LL + IL)]. Only minor cracking occurred at
midspan under positive moment loading and near the piers along the tendon
paths in several segments under negative moment loading. No sign of
joint slip or diagonal tension cracking was noted under design ultimate
shear loading.
:
C HAP T E R 8
FAILURE LOAD TESTS
8.1 General
The bridge model readily carried the BPR design ultimate load for
all critical flexural and shear loading conditions, as shown in the
previous chapter. To obtain maximum return from the model, several failure
load tests were planned to determine ultimate capacity. In the first
test to failure, factored AASHO truck loads were applied in one side span
to produce a bending failure, as shown in Fig. 8.l(b). This loading was
selected even though the calculated failure live load factor for this
case was larger than that which was calculated for maximum moment loading
in the main span. Since both of these type failures would be flexural,
it was felt that a failure test in the side span could verify flexural
ultimate calculations. Such a test would leave the structure with two
relatively undamaged spans so that a shear test to failure could also
be run by applying lane loadings to the main and opposite side span. This
loading, as shown in Fig. 8.l(c), would produce maximum shear at the main
pier and be an effective test of epoxy joint performance. Truck loading
on the side span [Fig. 8.l(b)] was stopped after distinct yielding had
occurred in the side span and at support SM, as judged by the deflection
and strain readings, but before complete collapse of the side span.
Although loads were applied in this test after formation of a plastic
hinge, the effect of this loading on the ultimate bending and shear
strength of spans (B) and (C) during the second failure loading test
shown in Fig. 8.l(c) was judged negligible. No live load would be
applied on span (A) and the end segment at SE would deflect upward.
8.2 Side Span Failure Load Test
8.2.1 General. The scaled AASHO HS20-Sl6 truck loads shown in
Fig. 8.2 were applied to all four lanes of the side span to produce
149
150
(A) SPAN AND SUPPORT NOTATION
SPAN CA) 6 a SPAN (8) SPAN (C)
SE SM
(8) ULTIMATE LOADING CONFIGURATION FOR MAXIMUM MOMENT IN SIDE SPAN.
& SE
& NM NE
(C) ULTIMATE LOADING CONFIGURATION FOR MAXIMUM SHEAR ADJACENT.
TO NM IN THE MAIN SPAN.
4 i SE SM NM
Fig. 8.1. Bridge failure loading
1.0 (ll + Il) a 1.222 II FOR 4 lANES .. 4.34K
4.34Kil.OasK
J I I
400" 200"
0.271 K - FRONT WHEIL
1.085K
- REAR WHEEL
NE
r
Fig. 8.2. Truck loadings at 1.0 (LL + IL) for maximum positive moment in the side span
'.
151
maximum moment. The allowable load reduction for a four lane bridge was
ignored. The live loads applied were in addition to the already applied
1.35 dead load. Live and impact loads were increased to the 5.25(LL + 1L)
level. Load increments of 0.25(LL + 1L) or 0.125(LL + 1L) were used after
the 0.75(LL + 1L) level was reached.
8.2.2 Test Results. At the 2.88(LL + 1L) increment, appreciable
deviations from the generally linear load vs. deflection diagram (Fig. 8.3)
were noted. Strains in the bottom slab of the SS6L segment changed rapidly
at the 2.75(LL + lL) increment, as shown in Fig; 8.4. It appears as though
cracking may have started to tevelop in the inner webs, although no cracking
was visible in the outer webs. At the 3.25(LL + lL) increment, a flexural
crack on the outer web around the center of the SS7R segment was visible
almost up to midheight and the strain gages in the bottom slab showed a
large increase in strain, as shown in Fig. 8.5.
At the 4.25(LL + 1L) increment, a wide crack (more than liB in.
visually) developed suddenly at the SS6-7 joint in the outer web of the
west side (OW), as shown in Fig. B.6. At the 4.38(LL + 1L) increment a
major crack formed near the SS6-7 joint in the outer web of the east side
(OE), as shown in Fig. 8.6. After these cracks developed, Figs. 8.3 to
8.5 indicate changes in strain at the center of the SS7R and SS6L segments
stopped, and deformations were concentrated in the vicinity of these large
cracks. From the strain diagram for the SSI segment shown in Fig. 8.7,
the strains are seen to increase rapidly at the 4.38(LL + 1L) increment
and to increase less rapidly after that increment. The rapid increase of
the strain at the SSI segment, and in the reaction at the NE support (as
shown in Fig. 8.8) around the 4.25 or 4.3B(LL + 1L) increment means that a
plastic hinge was formed at the SS6-7 joint. After forming the plastic
hinge the loads were redistributed and more load was carried at the SM
pier region. Since the bridge is a three-span continuous beam, plastic
hinges have to be formed at (A) and (B) in Fig. 8.9 to have a complete
failure mechanism for loading in the side span. The average reaction at
the SE support increased fairly linearly, as shown in Fig. B.8. Changes in
the individual web reactions at the SE support can be seen at the 2.88 and
4.38(LL + 1L) increments.
':J -+ ...J ...J -5.0
4.0
3.0
2D
1.0
o o
LOADING CONDITION: (+ 1.35 DL)
1h SE~114011 rSM NMA ANE
SS7
2 4 6 8
DEFLECTION = AVERAGE OF 4 READINGS ON WEBS
0.2 0.3 0.4 DEFLECTION (IN.)
0.1
Fig. 8.3. Deflections at the center of the SS7 segment in side span
..
LOADING CONDITION:
SE~SM 6.0
4 " NM NE (I) (H) SS6
n rl '16' 5.0
\.!.J
4.0
t B"IIDGE
(H) (I)
a {Ll (M)
-250 - COMPRESSION
:i ... + ..J ..J
/ /.'
/' /1
/ I tI I
I I I.
(l) (M)
I I I
I I I I
I I ~ I
I I I I ,/
!I J'
TENSION-
153
.750 .1000
STRAIN If' IN./IN.!
Fig. 8.4. Longitudinal strains at the center of the SS6L segment
lOADING CONOI TION: (. I. 35 DLl
S~SM b
/ /
/ I"
/ /
/ /
I / /
/ ~
/ /
/ /
/
-250 +250 +500 .75(J
-COMPRESSION TENSION-
h
NE
(F) IE)
----
i BRIDGE
(B) (AI I a , I
(F) (E)
.1000 .1250 .1500
STR~IN {f IN./IN.!
Fig. 8.5. Longitudinal strains at the center of the SS7R segment
154 rl E I I w Fl OE OW
LOADING CONDITION: (+ 1.35 DL)
lh n rl~ s1: l 1M t1M a NE LD W
SS7
557R 556R 555R
4.15 ON THE 5URFACE OF OE
* NUMBERS ALONG THE CRACKS ARE MULTIPLES OF (LL+IL)
556L 557L S58L
4.38 4.25
.. .... 4.25 4.0 00:: ....
T ~27 4.0 5.15
~ 2& 3.11
ON THE SURFACE OF OW
Fig. 8.6. Development of cracks during loading
(N)
(D) 00 (8) (A) 00 (I) (J) (K)
-250 -200 -150 -100 -50 STRAIN Y.1N.lIN') -COMPRESSION
o
LOADING CONDITION: (+ 1.35 DL) Uj
50 100 TENSION-
150
A NM
Fig. 8.7. Longitudinal strains at the center of the SSl segment
a NE
..... VI VI
156
'J 8.0 .... + ...J
2 5.0
4.0
3.0
2.0
1.0
0 0
'J 6.0 .... +
...J
...J
5.0
4.0
/ /
1.0
LeADING CONDITION: (+ 1.35 DLl
11 • l> NE
./ /'
./ _/
¥' ---.-/" -
/" /"
/" /"
y
,/' -,/'- ...----/
/-
-- -:;:::/ --,/'/
/ ,I
/
0.1
/
2.0
0.2 0.3
3.0 4.0
0.4
(I)
(O) (II
0.5 0.6 0.7 REACTiON (KIPS)
(I). (0) (0) -2-
rf @!]
(0) (I)
I I 5.0 6.0 1.0
REACTION (KIPS)
Fig. 8.8. Reactions at the outer supports
(O)
/'
I
( I)
7' /"
/"
0.8
8.0
( A) LOADING CONDITION: (+ 1.35 DL)
! e
SE (A) K
(8) SM
2S NM
(B) FIRST PLASTIC HINGE AND MOMENT DIAGRAM.
K (A) (8)
€!I- t st PLASTIC HINGE
157
! NE
- - - - - - - - - - - - - - M M 'PLASTIC MOMENT 01' 02'
(C) SECOND PLASTIC HINGE AND MOMENT DIAGRAM AT COLLAPSE.
&:::... ~2nd PLASTIC ;NGE
~ (8) ___ L __
(8)
Fig. 8.9. Failure mechanism for truck loadings in the side span
158
Due to the extreme widening of the crack at the SS6-7 joint,
an unexpected horizontal force occurred on the top of the SE pier. It
could be visually observed at high load levels that the SE pier was
tilting and inclining after the large cracks opened around the SS6-7
joint. Apparently a significant horizontal force due to the deformation
of the bridge was induced at the top of the SE pier, as shown in Fig. 8.10.
The moment connection between the end pier and the test floor was not
strong enough to keep the pier from tilting under this force.
r SS9
, I I 'SE I J ,
SS8 SS7 Sse SS~
I~ CRACK
://HORIZONTAL FORCE t.;.
,
""
Fig. 8.10. Horizontal force on the top of the outer pier
Because it was obvious that a plastic hinge had formed near the
4.38(LL + IL) increment and because of the inclination of the end pier,
it was decided to stop loading and release all live load at the
5.25(LL + IL) increment. This represented practical failure of the side
span, although total collapse did not occur. In this way further load
testing could be completed in the other two spans.
The maximum width of the crack at the 666-7 joint was 1/4 in. on
the outer west web (OW) and about 1/8 in. on the outer east web (OE) under
the maximum load increment. Cracks were distributed more in the outer
east web than in the outer west web. This might be due to a somewhat
weaker joint in the west side of the box because there was not much
apparent difference in the grouting effectiveness.
159
In the transverse direction, as seen in Fig. 8.11, deflections
were reasonably uniform across the SS7 segment at small loading increments.
Relative deflections increased in the cantilever slabs and at the middle
of the midstrip closure as loading increased. Transverse strains on the
bottom face of the top slab were measured by paper gages, as shown in
Fig. 8.12. Transverse strains also increased rapidly at the 4.38(LL + IL)
increment.
8.2.3 Calculation of Side Span Live Load Capacity. In order to
find the actual live load capacity factor (LF) for the truck loading in
the side span, it is necessary to consider all forces and moments acting
on the bridge.
Each time precast segments were erected and joined in symmetrical
cantilever construction, negative tendons were prestressed in order to
hold the precast segments in the proper balanced position, as shown in
Fig. 8.l3(a). For a typical construction stage, if a section is cut at
A-A, the prestressing force ~ is acting as shown in Fig. 8.l3(b) and
produces a moment M = [(~) X e]. These forces and moments can be calcu-p
lated at each section and a force diagram for r.P and a moment diagram for
M can be drawn as in Fig. 8.l3(c) and (d), respectively. The opposing p
dead load moment due to the weight of the segments is shown in Fig. 8.l3(e).
Although vertical forces are acting along the curved portions of the tendon
ducts, no vertical forces are included, since all negative tendons are
horizontal at the joint. At the completion of the balanced cantilever
construction scheme, the diagrams of the total force and the total moment
due to prestressing cables (negative tendons) and the total moment due to
dead load (weight of segments) are shown in Fig. 8.14. The moment due to
the external load (weight) and prestressing cables were designated as the
external moment (M'El) and the internal moment (MIl)' respectively. The
force due to the prestressing cables was designaled as the internal force
160 LOADING CONDITION: (+ 1.35 DL)
U, ., rl
M St: L 1M ~M rtE 0 SS7
2 3 4 5 6 7 8 9
•
\ •
1 Jl
\ •
1 •
SS7R SS7L
245 6 789 O.O~-r-....,...---r-----,r---....,...---r---r-----r--r--
-~ z o ~
0.1
0.2
~ 0.3 ...J LI.. LU o
0.4
0.5
1.0 (LL + IL)
2.25 (LL + IL) 2.75 (LL + IL)
4.0 (LL + XL)
4.5 (LL + ILl
5.25 (LL + IL)
Fig. 8.11. Deflections along the center of the SS7 segment in side span
161
&.0
(TA)
4.0
3.0
2.0
"
(TAI-- TRANSVERSE GAUGE (PAPER GAUGEI
to
__ ~~ ____ ~~ ____ ~~ ____ ~~ __ ~~ ____ ~~ ____ -=~ ____ ~~ ____ ~O -aoo -1'00 -600 -lIOO -400 -!CO -200 -100
STRAIN (f IN.IINl - COMPRESSION
Fig. 8.12. Transverse strains at the center of the SS7 segment
162
(A) CANTILEVER CONSTRUCTION
(B) PRESTRESSING FORCES
AND MOMENTS AT A-A
. . -......... -, ........ :.,:::s: II -II/::::::: .. .. II
-C.G .
" AJ " " II
II II II II II II
(C) PRESTRESSING FORCES ALONG SEGMENTS
(D) MOMENTS DUE TO PRESTRESSING FORCES
(E) MOMENTS DUE TO WEIGHT OF SEGMENTS
151 K-IN. , t I Z l 'lZZLlJIIAlfIf.lllZZ II ' t
Fig. 8.13. Forces and moments acting on the bridge for a typical cantilever stage
"
.'
163
(A) COMPLETION OF CANTILEVER CONSTRUCTION
1 EYE-fTYEn;R·tttt9 ]1"1-111 EIT E tTl' E"1::tr'1"1 "ttl"tt'll I' I I
(8) HORIZONTAL FORCES DUE TO PRESTRESSING (I),)
(C) MOMENTS DUE TO PRESTRESSING (Mn> 2100 K-IN.
~., -=-(D) MOMENTS DUE TO WEIGHT OF SEGMENTS (M~
~~
?~'t
2930 K-IN.
~'l
830 K-IN.
Fig. 8.14. Horizontal forces and moments at completion of cantilever erection
164
Positive tendons in the side spans were inserted and prestressed
after completing the cantilever construction. Prestressing of positive
tendons in the side spans cause an internal force (FI2
) and moment (MI2
) ,
as shown in Fig. 8.15.
The closure segment at the center of the main span was then cast.
At this stage there were external moments due to 1.0 DL, (~l)' and
internal force (F Il ) and moment due to negative tendons (MIl) in the main
span. However, the moment at the closure segment was zero. The positive
moment cables at midspan essentially need to resist only live loads. Thus,
the amount of center span positive prestressing cables should be less in
a bridge constructed by cantilever construction than in a bridge con
structed on falsework. Details of these differences are referred to in 30 Muller's paper.
After the required strength of concrete in the closure was devel
oped, the positive tendons in the main span were prestressed. Outer
supports were set in position touching the girder ends prior to pre
stressing the positive tendons in the main span. Thus, resultant forces
were produced at the outer supports when the positive tendons were pre
stressed, since the side span outer ends tried to deflect downward. In
addition, to ensure that no uplift at the outer supports occurs at service
level loading, specified additive vertical reactions were jacked into the
outer supports at completion of the stressing (see Sec. 7.3.3.2). Internal
forces (FI3
) and moments (MI3
) due to positive tendons in the main span
are shown in Fig. 8.l6(b) and (c). The moment (~2) caused by the resultant
forces due to prestressing and the applied forces used to adjust the
reactions and elevations at the outer supports are shown in Fig. 8.l6(d).
FIl , FI2 , FI3 , and MIl' ~2' MI3 , ~l' and ~2 are the forces
and moments acting on the bridge at the time of completion of construction.
Figure 8.17 shows the total or net horizontal forces and moments at each
joint at the completion of all prestressing operations.
In order to compute the ultimate capacity of the bridge under
factored dead and live load, the construction history must be considered.
165
(A) PRESTRESSING OF POSITIVE TENDONS IN SIDE SPAN
(8) PRESTRESSING FORCE AND MOMENT AT 8-8
__ c.~.
(C) HORIZONTAL FORCES DUE TO PRESTRESSING IN SIDE SPAN (FI2)
66.1 KIPS
(D) MOMENTS DUE TO PRESTRESSING IN SIDE SPAN (MI2)
Fig. 8.15. Horizontal forces and moments due to prestressing in side span
166
(A) PRESTRESSING OF POSITIVE TENDONS IN MAIN SPAN
(8) HORIZONTAL FORCES DUE TO PRESTRESSING IN MAIN SPAN (Fn)
~ (C) MOMENTS DUE TO PRESTRESSING IN MAIN SPAN (Mra)
1220 K-IN.
(0) MOMENTS DUE TO RESULTANT FORCE OF PRESTRESSING IN MAIN SPAN AND JACKING FORCES AT OUTER SUPPORTS (ME2)
740 K-IN.
Fig. 8.16. Horizontal forces and moments due to prestressing in main span
~
CI) Q..
i: ILl (.) 0: f2 -...J -2 !CI: ~ -100
2 0
~ :I:
(A) HORIZONTAL FORCES DUE TO PRESTRESSING
(B) MOMENTS DUE TO PRESTRESSING AND WEIGHT OF SEGMENTS
Fig. 8.17. Horizontal forces and moments at completion of all prestressing operations
168
The basic ultimate design guide used, the 1969 BPR "Strength and
Serviceability Criteria, Reinforced Concrete Bridge Members, Ultimate
Design,,,7 specifies 1.35 ~L as the ultimate dead load moment. This value 21 (1.35 MuL) was used in the design of the bridge by Lacey assuming that
the basic structural configuration would be the same for both the 1.0 DL
and 0.35 DL. Actually the 1.0 DL was applied in the cantilevering stage
and the 0.35 DL was applied to the three span continuous structure in the
model test. In designing the positive tendons of the prototype, the
external design ultimate moment was calculated for an ideal three span
continuous beam, as follows:
where
M u = 1.35 MuL + 2.25 M(LL + IL) + Msl + Ms2 + Ms3
~L = moment due to dyd load
M(LL + IL) = moment due to (live + impact) load
Msl = secondary moment due to prestressing tendons
Ms2 = secondary moment due to prestressing tendons in the main span
Ms3 = secondary moment due to prestressing tendons in the side span
of negative
of positive
of positive
Secondary moments induced by the prestressing of all negative and positive
tendons were considered in the calculation of the design ultimate moment,
as shown in the above equation. 2l The additional factored 0.35 DL may
represent additional weight due to heavier sections which would be caused
by casting errors, or allow for later dead load changes in the bridge such
as resurfacing, or simply be a safety margin. Since these types of seg
ments are precast under close control, there seems to be a low probability
of section overweight occurrence. It would seem logical to apply some of
the 0.35 DL on the balanced cantilever and the remainder of the 0.35 DL
on the completed continuous structure. However, such a division of the
factored load was not simulated in the model test. The additional 0.35 DL
was applied as if a live load on the three span structure after completion
of construction. Thus, it is necessary to calculate the moment due to
1.0 DL and 0.35 DL separately. The moment for 1.0 DL should be calculated
for the determinate structure (balanced cantilever) and that for the
169
0.35 DL (which was applied to the completed structure after service load
testing) should be calculated for the indeterminate three span continuous
beam structure. The moment due to 0.35 DL (ME3
) is shown in Fig. 8.18.
Figure 8.19 shows the critical design truck loading in the side span and
the corresponding moment diagram at service load of 1.0 (LL + IL). When
the service (live + impact) load is increased to ultimate, a flexural
failure of the bridge will occur by either rupture of the concrete or of
the prestressing cables. The positive tendons in the positive moment
region and the negative tendons in the negative moment region as well as
the concrete compression zones present primary flexural resistance to live
load.
The following equations for the ultimate external moment were
used to calculate the LF of (LL + IL) in the model tests.
where
M = Mul + Mu2 u
Mul = 1.0 ~L to be computed as a balanced cantilever (= ~l)
Mu2 0.35 ~L + LF X M(LL + IL) + Ms to be computed as an
ideal three span continuous beam
0.35 ~L = ~3
M(LL + IL) ~4 M
s
In very underreinforced sections with bonded tendons, as in the
positive moment region, the ultimate compressive force (C) and the tensile
force (~T.) may be calculated by assuming the bottom layer of prestressing 1-
cables as its ultimate strain (f'). If joint 6-7 in the side span is taken s
for the free body, prestressing forces (p) due to negative tendons remain
relatively constant and add to C and T, as shown in Fig. 8.20. These P
forces have to be taken into account in computing the ultimate moment
capacity, except at the closure segment where P = o.
Since the positive tendons were placed in mUltiple layers in some
sections, it is necessary to use the stress-strain curve for each tendon
in order to find T. Although several specimens were tested in an attempt
170
is Sf
K K 4.34 4.34 K
11 i,085
e '
SM
664 K-IN.
«0.35 D.L. , , , 2 ,
481 K-IN.
li' NM
Fig. 8.18. Moment diagram for 0.35 DL (ME3
)
400"
, 2 # 2
200"
~~ ~4_vZZ7ZZII?I'Zr'
39.8 K-IN. ,(+)
119 K-IN.
Fig. 8.19. Moment diagram for truck loads in side span (ME4)
, , ~ NE
ft.)
.'
NEGATIVE TENDON
POSITIVE TENDON
171
T.
C: ULTIMATE COMPRESSiVe: FORCE
p: FORCE DUE TO NEGATIVE TENDON
T : ULTIMATE TENSILE FORCE IN POSITIVE TENDON
Fig. 8.20. Ultimate force at a certain section
to obtain actual stress-strain curves for samples of the wires and strands,
a complete stress-strain curve up to failure was not obtained. The speci
mens usually failed at the grips in the testing machine outside the strain
gaged length. Based on stress-strain curves for a 270k grade 7-wire strand
furnished by a manufacturer, typical stress-strain curves were developed,
as shown in Fig. 8.21. These curves were based on known ultimate strengths
for all of the prestressing tendons and measured E and strains from the s
incomplete material tests. The ultimate strain of the prestressing cables
d 0 06 ' I' 26 was assume as . ~n. ~n.
In order to find the ultimate forces at joint 6-7 in the side span,
T. values were found from strain compatability by using the stress-strain ~
curves of Fig. 8.21. Compressive strains for the concrete in the top fiber
were assumed for three different cases «( = 0.003, 0.002, and 0.0015 c
in./in.) and then ~T. were found, as shown in Fig. 8.22. There was no . ~
difference in ~T. computed whether ( at the compression fiber was assumed ~ '. c
as 0.003, 0.002, or 0.0015 in./in., so that it was not necessary to make
further iterations for T. in order to find the internal ultima~ moment ~
capacity. By assuming (c : 0.003 in. lin. at the extreme compression fiber
and a rectangular stress block, as shown in Fig. 8.23(a), C and P were
calculated as follows:
-en ~
300
250
200
;150 en LLJ 0:: t-en 100
50
----------
-:~~~::::~~~~~~::~~~~~~==~====~~~~==~~==~==~ ------------------- -------..".,..-- ------------------
-- 270 K GRADE 7 WIRE STRAND
--- 6GA. WIRE
---Y4IN.STRANDS, ~8IN.STRANDS _00- 7mm WIRE
O+-----~------~----~~----~----~_=----~----~------~----~------~----~~-----L 0.005 0,010 0.015 0.020 0,025 0.030 0.035 0.040 0.045 0,050 0.055 0.060
STRAIN (irvin,)
Fig. 8.21. Stress-strain curves of prestressing cables
173
CASE (1) - eS3 = 0.06 IN./IN., E.c = 0.003 IN./IN., c = 0.72 IN.
€'s ( IN.lIN,) O! (KSI) TI (KIPS) lj for 2 boxes (KIPSl llj (KIPS)
£5. = 0.019 261 7.57 30.3 (T1)
t 52= 0.035 269 7.80 31. 2 (T2) 127
£5.= 0.06 280 8.11 65.0 (T3)
. CASE (2) - En = 0.06 IN./IN., t.c = 0.002INJIN. t C = 0.49 IN.
es( IN.lIN.) ,
~(KSI) Tj (KIPS) .lj for 2 boxes (KI PS flj (KIPS)
£5. = 0.020 262 7.60 30.4 (11)
£52= 0.035 269 7.80 31. 2 (T2) 127
t n = 0.06 280 8.11 65.0 (T3)
CASE (3) - t.s3=0.06 1N./IN' f €.c=0.0015 1N1IN., C = 0.37 IN.
£'S (I N.II N.) ~(KSI) 1i (KIPS) lj for 2 boxes (KIPS ~1j (KIPS)
£51 = 0.020 262 7.60 30.4 (11)
E52 = 0.036 269 7.80 31. 2 (T2) 127
C.5.= 0.06 280 8.11 65.0 (T3)
Fig. 8.22. Calculation of Ti for different strain profiles -
174
(A) RECTANGULAR STRESS BLOCK
ossf
a C
c
~c = CONCRETE STRAIN
c = DISTANCE FROM EXTREME COMPRESSION
FIBER TO NEUTRAL AXIS AT
ULTIMATE STRENGTH
1----+ P a = DEPTH OF EQUIVALENT RECTANGULAR
(B) PARABOLIC STRESS BLOCK
£c f r--r-..,....lr,....,...,
c C
'-NA
STRESS BLOCK = ktc
I
fc = COMPRESSI VE STRENGTH OF CONCRETE
b = WIDTH OF COMPRESSION FACE. OF
FLEXURAL MEMBER
f..p = STRAI N AT LEVEL OF PRESTRESSING CABLES (NEGATIVE TENDONS)
A, = AREA OF PRESTRESSING CABLES
C = ULTIMATE COMPRESSIVE FORCE
P = FORCE DUE TO NEGATIVE TENDON
(C) CALCULATION OF ULTIMATE INTERNAL MOMENT
+ .. ------r -r. P
PLASTIC CENTER
T = ULTIMATE TENSILE FORCE
dqd" dti = DISTANCE FROM PLASTIC CENTER TO EACH FORCE
Fig. 8.23. Calculation of ultimate internal moment
'.
0.06 in. lin.
0.003 in. lin.
c ",O.72in.
a ~ 0.70 X 0.72 = 0.504 in.
C = 0.85f' X a X b 0.85 X 7.09 X 0.504 X 112 = 340 kips c
p 49.6 - A X E X € = 49.6 - 0.348 X 30 X 103 X 0.00025 p s p
C
49.6 - 2.7 46.9 kips
340 » P + T. = 46.9 + 127 ~
174 kips
This indicated that the compressive strain at the top fiber had not
reached 0.003 in. lin. Then, by assuming the strain at the bottom of
175
steel as 0.06 in./in. and the parabolic stress block shown in Fig. 8.23(b),
C and P were found after several iterations in which various neutral
axes were assumed until C = ~T. + P. The internal ultimate moment was I~ ~
then calculated as follows [see Fig. 8.23(c)]:
, €c 0.0015 in. lin.
c 0.37 in.
f 0.95 X f' = 0.95 X 7.09= 6.74 ksi c c €p = 0.0012 in. lin.
P 49.6 + 0.348 X 30.9 X 103 X 0.0012 62.5 kips
C 187 kips
C 187 = )'T. ~
+ P = 62.5 + 127 189.5 kips
~I C X d - P c X d p Tl X dtl + T2 X d t2 + T3 X d t3
= 187 X 5.90 62.5 X 5.46 30.4 X 0.8 + 31.2 X 3.0 + 65 X 8.98
1103 - 341 24.3 + 93.6 + 584 = 1415 k-in.
At th k . t th ft' t 25 . d1 h e pea compress~ve s reng 0 concre e, exper~men s ~n cate t at
strains reach about 0.002 in. lin. for various strengths of concrete. If
the strain at the peak stress is assumed as 0.002 in./in., about 95 percent
of f' will develop at the extreme fiber for 0.0015 in./in. strain. 27 c
Therefore, f was assumed as 0.95 X f' in the above calculation. c c
In all calculations of the ultimate moment and shear capacity,
(r'I = 1.0 was used, since all as built dimensions and material strengths
were known for the model.
176
In overreinforced sections with bonded tendons (around the main
pier), the concrete strain at failure at the bottom fiber was assumed as
0.003 in. lin. according to the ACI Code 2 and strains were calculated at
the level of the prestressing cables. The compressive stress block was
assumed as rectangular. Several trials were made until C equalled T by
varying assumed values of c in Fig. 8.24(c). T was determined from the
stress-strain curves of Fig. 8.21 for Es = Esp + Est' Although some com
pressive strain existed at the bottom fiber before applying any live load,
the magnitude (about 0.00015 in./in.) was very small compared to
€ = 0.003. Therefore, this effect was neglected and the strain was c
simply assumed as 0.003 in. lin. at the bottom fiber in order to calculate
the ultimate moment at the pier. After determining the proper c, the
internal ultimate moment was calculated as follows:
c = 2.46 in.
Ec 0.003 in. lin.
a = 0.7c = 1.72 in.
b 52 in.
C = 0.85f' X a X b 0.85 X 7.09 X 1.72 X 52 539 kips c
Area Esp Est - E + F' IJ s T T for 2
(in. 2) Es- spst boxes
(in./in.)(in./in.) (in./in.) (ksi) (kip) (k)
3/8 in. 0.085 0.00600 0.0159 0.0219 243 20.7 166 strands 7nnn 0.0593 0.00495 0.0159 0.0199 242 14.4 288 wire 6ga. 0.029 wire
0.00462 0.0159 0.0205 242 7.0 84
Total T 538
jd = 15.46 - a/2 = 15.46 - 0.86 14.6 in.
Ultimate Internal Moment:
~I = 538 X 14.6 = 7855 k-in.
Since the completed bridge is a three span continuous beam, it
has to form two plastic hinges for complete failure in the side span,
(A) POSITION OF TENDONS AT MAIN PIER 177
1.31311
" __ "-+---+-_=C.G._
1.78
U) C\I ~ It) cD
(B) PROPERTIES OF EACH TENDON
TENDONS AREA
(IN~
7MM.WIRE B1,B2,B5,86,B7 0.0594
j'STRANDS B3,B4 0.085
6 GA. WIRE B8,B9,B10 0.029
(C) ULTIMATE INTERNAL
FORCES T
jd
L C
~\~ O.85f ' c
EXPERIMENTAL VALUE APPLI ED VALU E
Fs' fs' E, 0.6 F,' a.6f; espAT Q6f~ (KIPS) (KS·I) (KSI) (KIPS) (KSI) (INJINJ
15.31 258 30.5x103 8.94 151 0.00495
22.05 259 27.0X103 13.75 162 0.00600
8.13 280 30.9x103 4.14 143 0.00462
E.sp= STRAIN DUE TO PRESTRESSING (IN./IN.) t-s, = STRAIN DUE TO EXTERNAL LOAD(lN.lIN.) ts = ULTIMATE STRAIN OF PRESTRESSING CABLE
= E..SP + e,SJ (I N.l1 N.) f.,F = 0.003 IN./IN. fc = COMPRESSIVE STRENGTH OF CONCRETE T = TENSILE FORCE AT ULTIMATE (KIPS) C = ULTIMATE COMPRESSIVE FORCE c = DISTANCE FROM EXTREME COMPRES. FIBER
TO NEUTRAL AXI S AT UL T. STRENGTH (IN.) jd = DISTANCE BETWEEN T a C (IN.) a = DEPTH OF EQUIV. RECT. STRESS BLOCK (IN.) b = WIDTH OF COMPRESSION FACE OF
FLEXURAL MEMBER (IN.)
Fig. 8.24. Calculation of ultimate internal moment at pier section
178
as shown in Fig. 8.25(b). The LF for (LL + IL) to produce the first
plastic hinge was calculated as follows.
equated.
(A) LOADING CONDITION
SM NM NE II
(8) FORMATION OF PLASTIC HINGE
~ICHIN'E SE 8M
70- I"
tC) PLASTIC MOMENT
Fig. 8.25. Plastic hinges and moment diagram
Internal moment and external moment at the ultimate capacity are
MuI = ~1 + ~2 + ~3 + LF X ME4
~I - MEl - ~2 - ~3 LF for (LL + IL) = --=-=----==-----=~--=~
ME4 1415 + 319 - 259 - 34
322
= 1441/322 = 4.48
'.
Therefore, a first plastic hinge should form at the 1.35 DL +
4.48 (LL + IL) increment at the SS6-7 joint. Since the plastic hinge
at the SS6-7 joint was observed to form between 4.25 to 4.38 (LL + IL)
in the experiment, the calculated value (4.48) is very accurate.
Tensile strength of the only available 6 gage wire, which was
used for the positive tendons in the side span was about 18 percent
higher than the specified minimum, although all other prestress wires
179
or strands had tensile strengths very close to the minimum values
specified. It would therefore be expected that the first plastic hinge
in a prototype with strands having exactly the specified minimum tensile
strength would form at a loading somewhat less than the test 4.48(LL + IL)
increment fat about 3.8(LL + IL)J. However, if the 25 percent live load
reduction is used for the four lane bridge, a LF of 5.1 would be calculated
which is ample.
If a second plastic hinge is assumed to form at the SM pier
segment as shown in Fig. 8.25(b), the LF of (LL + IL) for complete
failure can be calculated as follows:
Assume
reaction at the SE support.
= plastic moment at the SS6-7 joint (MVI at the SS6-7 joint).
= plastic moment at the SM pier segment (MVI at the SM pier segment).
x LF of (LL + IL).
From equilibrium, at the SS6-7 joint
MOl = - 4.34 X lOX + 70Y
and at the center of the SM pier segment
-M02 = - (4.34 X 140 + 4.34 X 108 + 1.085 X 80)X
+ 200 Y
From Eq. (1)
1415 - 43.4 X + 70 Y ... (1)'
(1)
(2)
180
From Eq. (2)
- 7855 -1164 X + 200 Y ... (2)'
[(2)' X 0.35] - 2750 -408 X + 70 Y . . (3)
r(O' - (3)] 4165 = 364 X :. X = 11.4
Therefore, the second plastic hinge would form around the[1.35 DL +
l1.4(LL + IL)] increment and the SM pier segment would fail in compression,
if the initial plastic hinge had sufficient ductility. It is thus con
servative to assume that the side span is fully capable of carrying
[1.35 DL + 4.48(LL + IL)] considering four lanes fully loaded. Using
AASHO load reduction factors for a four lane bridge, this would become
[1.35 DL + 6.0(LL + IL)].
8.3 Failure in the Main Span
8.3.1 General. Four AASHO lane loads were applied to the main
span and one adjacent side span to produce the critical shear condition at
the first joint in the main span. It was anticipated from computations
that with full development of shear strength the bridge would fail in
flexure even though under a maximum shear loading. However, it was
decided to check the shear capacity since basic information about flexural
capacity was obtained by applying the truck loads to the side span. Lack
of published information made it very desirable to check the performance
of the epoxy joints under realistic high shear loadings.
In addition to the 1.35 dead load, live and impact loadings, shown
in Fig. 8.26 were applied by rams and increased until failure. The position
of the heavy concentrated load could greatly affect the shear strength of
the bridge. It was considered that a direct shear failure might occur as
the effective depth decreased due to flexural cracks, so concentrated
loads were applied outside but adjacent to the first joint in the main
span. Loading increments of 0.5(LL + IL) were used up to a loading of
2.0(LL + IL), after which the increments were reduced to 0.25(LL + IL) up
to failure.
'.
"
181
K K 2.014 3.42K 2.014
K K K/ KKK 2.014 2.014 2.014 2.014 2.014 2.014
SE SM NE 3'i 48" 48" 48" 48" 48" 4811
~~~~~~~~I~~~~~~~~~
100" 200"
Fig. 8.26. 1.0(LL + 1L) loading for maximum shear at the NM pier
8.3.2 Test Results. After the load reached 2.25(LL + 1L), the
reaction at the north end (NE) started to decrease and at higher loads the
north reaction was unloading the dead load effects, as shown in Fig. 8.27.
At the 2.63(LL + 1L) increment, the south end segment (SE) raised com
pletely from the neoprene pad support. At this load level the crack
which had previously developed at the joint of the main span closure
segment during the positive moment ultimate design load test (see
Sec. 7.3.3.2) started to reopen. Figures 8.28 to 8.30 show that strains
in segments SS7, SS6, SSl, and SMI increased almost linearly up to
2.63(LL + 1L), but remained constant after that increment because the
south end reaction became zero and no load was applied to the unloaded
side span. Strains at NS6 were very low until the 2.5(LL + 1L) increment,
then increased steadily until failure, as shown in Fig. 8.31. This change
was caused by the alteration in structural configuration when the south
side span became a free cantilever.
At the 3.25(LL + 1L) increment, the strain increas at the NM6
~egment stopped, as shown in Fig. 8.32. This was due to the concentration
182
lOADING CONDITION: (.1.35 Dl)
SM ~M ~E
M (0) (II
(0). (I) (0) --2- (I)
-1.5 KIPS
'-
-\.0
ww
UPWARD
6.02
...J d )(
5.0
4.0
3.0
2.0
t.O
-\.5
(I)
M (0) (I)
-1.0 -- .0.5 KIPS
UPWARD DOWNWARD
Fig. 8.27. Reaction at outer supports during loading to failure
LOADING CONDITION: + 1.35 Dl)
(E) (F)
(8) (A)
13 (F) (E)
-100 -50
s\x s'C SS7R SS6l
...J
'": 6.0 ...J ...J
(A) (8)
-COMPRESSION
.50
TENSION-STRAIN <t IN.IIN.)
) .. LJLJ (lW.t)
i8RIDGE
I (H) III
E0 (Ll (M)
-100 -50 -COMPRESSION
STRAIN 't IN.IIN.)
(1)
.50
TENSION-
Fig. 8.28. Longitudinal strains at SS7R and SS6L during loading to failure
.'
(GHNlIF)(M)
LOADING CONDITION: (. L35 DL
NM
{OIOO (al lAl
"8 (Gl (F) 00
! NE
()() (I) (JI (K)
S" 00 (M) INI
-200 -150
STRAIN 'f IN.lIN)
6,0
-100 -!l0 -COMPRESSION
183
(D)
(AI (a)IJ~KI II)
• TENSION-
Fig. 8.29. Longitudinal strains at 551 during loading to failure
(al (A) (H) (I)
-150 STRAIN {f IN.lINJ
-100 -50 ...-COMPRESSION
,'') ~II (H) (8)
.50 .100 TENSION---
Fig. 8.30. Longitudinal strains at 5Ml during loading to failure
184
(N) (M)
(B) 00 00 (I) IJ) (K)
EJ a" IF) IE) IL) (M)IN)
-200 -150 STRAIN (fiN/IN.)
...J .... ~ 6.0
...J
d
1.0
(J) (I) (K) (B)
-100 -50 o .50 .100 -COMPRESSION TENSION-
Fig. 8.31. Longitudinal strains at NS6 during loading to failure
LOADING CONDITION' (.1.35 oLl
Ii " SE SM /.w:JNM
NM6
(8) (AI
a (F) (EI
(A)
A NE
-200 -150
STRAIN 't IN./IN.!
(H) (B)
-100 -50 o --COM PRE SSION
IF)
.lIO
TENSION-
(E)
+150
Fig. 8.32. Longitudinal strains at NM6 during loading to failure
. '
of deformation in the crack around the center of the main span. By
observing the deflection diagram for the SM10 segment in Fig. 8.33, it
is seen that the rate of deflection increase changed substantially at
3.25(LL + 1L). Also, a diagonal tension crack started to develop at
the first segment in the main span (outer web on the west side) as
shown in Fig. 8.34.
At the 3.75(LL + 1L) increment, a flexural crack at the joint
of the closure segment extended to near the top of the web and many
cracks started to develop in the region of segments SM6 to SM9, as
shown in Fig. 8.35.
185
At the 4.25(LL + 1L) increment, the diagonal tension crack and
the flexural crack around the NM pier joined (see Fig. 8.34) and a wide
flexural crack developed about 1 in. away from the first joint in the
main span. At this stage, the flexural crack on the top slab was only
in the outer cantilever portion. At this loading the south end segment
raised up about 1 in. from the surface of the neoprene pad support .
At the 4.25(LL + 1L) increment in the east side and the
4.75(LL + 1L) increment in the west side, very wide flexural cracks
developed at the SM6-7, SM7-8, and SM8-9 joints, as shown in Fig. 8.35.
These cracks developed near the epoxy joints in the web portion (in the
flexural tension zone) and about 1 in. away from the joint in the bottom
slab (in the pure tension zone), as shown in Fig. 8.36. The cracks at
these joints went nearly to the top of the web with an increase of one
increment of loading. The increase of strain around the 4.00 to 4.75(LL + 1L) increment range at NM9, NMl, and NSl stopped because of concentration
of deformations at these joints and at the first joints from the main
pier. Figures 8.37 to 8.39 show the effect of the concentration of
deformation on strain at higher load.
At the 5.0(LL + 1L) increment, the crack on the top slab of
segment NSl and the NM pier segment extended the full wid th of the slab
(on the east side of the box).
,..... CP 0"-
- 7.0 ...J ..... +
...J
...J 6.0 -lL. 0 a:: 5.0 0 I-0
Lf 0
4.0 « 9
3.0 LOADING CONDITION :(+ 1.35 DL)
Ii' ....... '~M • "a., ww 2.0 A
SE
o L-______ ~ ________ ~ ______ ~ ________ ~ ________ ~ ______ ~ ________ ~ ________ L_ __
o 1.0 2.0 3.0 4.0 DEFLECTION AT THE CENTER OF MAINSPAN (IN.)
Fig. 8.33. Deflections at SMIO during loading to failure
.'
NM1
NSf
~~ ~2&
NM 4.10
.. 21
AT THE SURFACE OF OE
NM -.050 5.5
AT THE SURFACE OF OW
* NUMBERS ALONG THE CRACKS ARE MULTIPLES OF (LL.IL)
NS1
NM1
Fig. 8.34. Development of cracks around NM pier during loading to failure
, I
LOADING CONDITION: C+I.35DL)
NM8 NM9
Fig. 8.35.
LJ L07 • NUMBERS ALONG THE CRACKS ARE MULTIPLES OF (LL.ILl
OW
SM8 SM9 SM~ CLOSURE NMIO
ON THE SURFACE OF OE
NM~ CLOSURE SMtO SM9 SMa SM7 SM6
ON THE SURFACE OF OW
Development of cracks around the center of main span during loading to failure
NM9
..
,..... 00 00
189
J
\J c..:
\
t;
!,..,
" ,. a v..
Fig. 8.36. Typical crack around the joint (after failure)
- (M) (F) (l) (E) (N) ...J .... 6.0 +
...J
...J -5.0
4.0
lOADING CONDITION: • 1.35 Dl
4 Ii III,nUjh,\,JeIIiIl1i1 SE SM J- 180 JNM NE
NM
LdbJ 00 (F) (E) (l) (M) (N)
-400 -200 0 .200 .400 .600 .800 .1000 .1200 "-"--COMPRESSION TENSION'---'" STRAIN ~ IN.lINJ
Fig. 8.37. Longitudinal strains at NM9 during loading to failure
IG)(EI IFl(Ll
'N) - --- - - ----.._,..,.-
(Dile) Ie) ()O O(KP)(H) CII IJIIK)
"E1"a-(G) (F) IE) (LlIM)("')
-~
STRAIN 'f IN.lINJ
-500 -250 -COMPRESSION
191
(HI (0)
o .2e.o TENSION-
Fig. 8.38. Longitudinal strains at NMl during loading to fa ilure
1M) (L1 (A) Ie)
\ 6.0 :;] .
-' .::!
!j,0
4,0
IE)
LOADING CONDITION: (+1.35 Oll
a N~~O" 1 (B) (AI (HI (Il (J) 00 4 SE Silo! HE 8 E1 •
NSI
[)(lIE) III (M)(~)
1 I I
-750 -e.oo -250 .e.oO .750 .1000 -COMPRESSION STRAIN r IN./INJ
Fig. 8.39. Longitudinal strains at NSl during loading to failure
192
At the 5.75(LL + IL) increment, the flexural cracks at the top
~lab near the NM pier extended the full width of the slab (two boxes).
The flexural cracks at the SM6-7 and SM7-8 joints were getting wider, but
the width of cracks in some other portions were small. At this stage, the
bridge looked straight from the SE pier to the SM6-7 joint and all major
deformation was concentrated at the SM6-7 joint. The NM pier segment on
the neoprene pad support started to crush on the east side, due to the
high compression force.
At the 6.0(LL + IL) increment, the width of the flexural crack
at the SM6-7 joint was about 1/8 in. and the SE segment raised up about
4.5 in.
After taking the instrument readings at the 6.25(LL + IL) incre
ment, the loads were being increased to the 6.50(LL + IL) increment when
a sudden rupture of the positive moment prestressing cables occurred at
joint SM6-7 on the west side. This failure occurred before applying less
than half of the planned increment and the load dropped immediately after
the failure. The load was then brought back to the 6.25(LL + IL) incre
ment and rupture of the positive moment prestressing cables in the east
side box occurred after a small increase in load. Figure 8.40 shows the
general view of the failed bridge. Even after rupture of the positive
moment cables, the bridge exhibited great toughness and continued to
carry 1.35 DL as balanced mntilevers.
The strain at the various positions in the same cross section
varied as the load increased, although the tendency of strain change was
similar at each position. However, the increase of deflection at the
same station in each cross section was uniform as the load increased.
Figure 8.41 shows the deflection along the bridge during the failure
loading test.
8.3.3 Determination of the Main Span Live Load Capacity
8.3.3.1 Flexure. One major difference between this bridge and
the generally assumed three-span continuous bridge is the lack of upward
vertical restraint as usually assumed for pin supports. The completed
bridge rested on neoprene pads at all four piers with no hold-down
193
Fig. 8.40. General view of the bridge after failure
2
\ \ \ \ \ LOADING CONDITION: ( ... 1.35 DL)
1 \ \\ ~ ld W 1 \ \ \ .. R t1 ~E 0 SE SM NM 0:: \\\ j \\\ a.. ::::> ~\
SM SM5 CLOSURE NM 0
SE NS4 NE 0 0:: \, j
\\ z ~ 1 \\ 0
1 \ \ \ \ \ \
2 \ \ 2
- \ \ 0 EXPERIMENT z \ \ z °3 \ :3 r <..> IJ.I \ ...J ~ IJ.I \ 0
4 4 ......... .,..G.O.35 DL ... 6.25 (LL ... IL)
Fig. 8.41. Deflections along the bridge during loading to failure
195
devices. The bridge behavior would be that of a three-span continuous
beam only if the tendency for uplift at the outer supports was restricted
at higher loads. However, to match the prototype support conditions,
no upward motion restrictions at the outer supports were added. There
fore, under high levels of loading in the center span, the bridge started
to act as a two-span continuous beam with an overhang or a simple beam
with two overhangs.
It is, therefore, necessary to calculate ~I' MEl' ME2 , ~3'
and ~4 for three different support conditions (MuI and ~l are the
same for the three cases). The latter are calculated using the same
procedure as explained in Sec. 8.2.3. Moment diagrams for ~2' ~3' and
~4 for the three different support conditions are shown in Figs. 8.42,
8.43, and 8.44.
By examining Fig. 8.44, it can be seen that the value of positive
moment for 1.0(LL + IL) in the main span increased substantially in the
critical region due to the change of structural configuration. Also,
the shift in position of the maximum moment is clearly shown in Fig. 8.44.
There is not much difference in moment between the two-span continuous
beam with an overhang and the simple span beam with two overhangs for
this loading case (Fig. 8.44).
In contrast, Fig. 8.43 indicates that the positive moment caused
by the additional 0.35 DL is erased by the change of structural configura
tion from three continuous spans or two continuous spans with an over
hang to a simple beam with overhangs. However, the negative moment
around the NM pier increased.
The LF or level of (LL + IL) which would form the first plastic
hinge for this loading was calculated for each type structure (Table 8.1).
The first plastic hinge would form at the joint of the closure segment
at an increment of [1.35 DL + 5.2l(LL + IL)] if the structure was ideally
supported by pins and there was no uplift possible at the end supports.
However, it is not proper to calculate the LF for a three-span continuous
beam since the SE support raised off its support pad at the
196
(A) THREE SPAN CONTINUOUS BEAM 740 K-IN.
SE SM NM
f 4 x
(B) TWO SPAN CONTINUOUS BEAM WITH AN ADDITIONAL OVERHANG SPAN
A A 1 ~. SE SM NM NE
(C) SIMPLE BEAM WITH TWO OVERHANG SPANS
SE SM 2i
NE MOMENT DUE TO END REACTION IS ZERO
BECAUSE OF NO RESTRAINT AT SE AND NE
Fig. 8.42. Moment diagram due to resultant force of prestressing and jacking force at end supports (M
E2)
(A) 481 K-IN.
(B)
(C)
Fig. 8.43.
X'" ""1'" oIl?;~,~ .~~, Mil "! SE SM NM NE
-664 K-IN.
0.35 DL lr",t"lilli ,c'~I'iaa*1l
SE SM NM NE
-1100 K-IN.
0.35 DL pon ,ll1(,izzzz II III Ii 'ftlllltl"
SE SM NM NE
-1100 K-IN. -1100 K-IN.
Moment diagram for 0.35 DL for three different conditions (ME3 )
SE S
SE
SE
SE
li SM
i. SM
i BRIDGE I
~,..,.,.....3-.96 K -IN.
NM
<t.. BRIDGE
~ NM
t BRIDGE 543 K-IN.I
/
! ~ 2i
SM NM
NE -526 K-IN.
1 NE
-661 K-IN.
NE
-701 K-IN.
Fig. 8.44. Moment diagram for 1.0 (LL + IL) for three different conditions (ME4)
197
NE
NE
......
'" 00
TABLE 8.l. LF OF (LL + IL) FOR THE FIRST PLASTIC HINGE AT EACH JOINT
LF (LL + IL) Case (1) 3 span continuous beam ~; ??, R: , , , , 'f)\fS»))Vn)))))UB)hu)}))b)))) )))))8%\
SE SM NM NE
Ultimate Moment Moment due EM = ~I - ~1 Moment due LF of Internal due to to end Moment due to 1.0 (LL+ IL)
Joint Moment 1.0 DL support to 0.35 DL -~2 - ~3 (LL + IL) EM MUI MEl forces ~2 ~3 (k-in. ) ~4 (k-in. ) ~4 ~k-in. ~ ~k-in. ~ ~k-in·l
SM 5 - 6 2342 -603 740 255 1950 245 7.96
6 - 7 2873 -361 740 346 2148 318 6.75
7 - 8 3395 -181 740 415 2421 354 6.84
8 - 9 3558 - 62 740 458 2422 390 6.21
9 - 10 3558 - 6 740 479 2345 398 5.89
closure 3285 0 740 481 2064 396 5.21
NM10 - 9 3558 - 6 740 479 2345 394 5.95
9 - 8 3558 - 62 740 458 2422 378 6.41
8 - 7 3395 -181 740 415 2421 333 7.27
7 - 6 2873 -361 740 346 2148 289 7.43
6 - 5 2342 -603 740 255 1950 208 9.38
Joint
SM 5 - 6
6 - 7
7 - 8
8 - 9
9 - 10
closure
NM 10 - 9
9 - 8
8 - 7
7 - 6
6 - 5
TABLE 8.1 (Continued)
Case (2) 2 span continuous beam with an overhang
Ultimate Moment Moment due Internal due to to end Moment 1.0 DL support
Mol MEl forces ~2 (k-in. ) (k-in. ) (k-in. )
2342 -603 204
2873 -361 241
3395 -181 278
3558 - 62 315
3558 - 6 352
3285 0 370
3558 - 6 389
3558 - 62 426
3395 -181 463
2873 -361 500
2342 -603 537
LF (LL + IL)
1. 35 DL _. . " _ - '" (" , J~ , ..... " '! ill I II tl I I 'it» »))) )} > Ii J'? ih u v UJj
SE SM NM NE
EM - MuI - ~1 Moment due LF of Moment due to 1.0 (LL+IL) to 0.35 DL -~2 - ~3 (LL + IL) EM
~3 (k-in. ) ~4 (k-in.) ~4
- 22 2763 498 5.55
99 2894 544 5.32
196 3102 554 5.60
270 3035 563 5.39
320 2892 544 5.32
337 2578 529 4.87
350 2825 513 5.51
358 2836 470 6.03
344 2769 399 6.94
307 2427 328 7.40
245 2163 220 9.83 I-' \0 \0
Joint
SM 5 - 6
6 - 7
7 - 8
8 - 9
9 - 10
closure
NM 10 - 9
9 - 8
8 - 7
7 - 6
6 - 5
TABLE 8.1 (Continued)
LF (LL + IL)
Case (3) Simple beam with overhangs 1.35 DL ) ~ S \ Pi S. , FS S S S. S \ CiS rFi S S \ S S. < s (
IlZ1ZZlZZ1'2U2UJ2lZZUOYY22 I}J IIIDJ)!
Ultimate Internal Moment Hul
(k-in. )
2342
2873
3395
3558
3558
3285
3558
3558
3395
2873
2342
Moment due to 1.0 DL
MEl (k-in. )
-603
-361
-181
- 62
- 6
o - 6
- 62
-181
-361
-603
Moment due to end support
forces ME2 (k-in. )
o o o o o o o o o o o
SE
Moment due to 0.35 DL
~3
-207
-120
- 57
- 16
o o o
- 16
- 57
-120
-207
SM
EM = ~l - ~1
-~2 - ~3
(k-in. )
3152
3354
3633
3636
3564
3285
3564
3636
3633
3354
3152
NM
Moment due to 1.0
(LL + IL) ~4 (k-in.)
485
530
536
543
523
507
487
445
372
295
188
NE
LF of (LL+ IL)
EM
~4 6.50
6.33
6.78
6.70
6.81
6.48
7.32
8.17
9.77
11.4
16.8
N o o
201
[1.35 DL + 2.63(LL + IL)] increment. Since the south end support raised
completely from the neoprene pad supports, all forces applied at the time
of construction (such as end reaction due to positive tendon prestressing
in the main span or the jacking force at the end supports to adjust the
reaction) were erased and the structure became a two-span continuous beam
with an overhang. If the structure were an ideal two-span continuous
beam with an overhang, the reaction at the NE support would have to
increase as the load increased. But the reaction at the NE support
decreased after the [1.35 DL + 2.25(LL + IL)] increment due to the
appearance of cracks and concentration of deformation around the center
of the main span. Observations indicated that a plastic hinge was not
formed at the closure segment at [1.35 DL + 2.25(LL + IL)] as would be
indicated by Table 8.1(2) for a two-span continuous beam with an overhang.
Since the calculation for case (3) in Table 8.1 indicated that the minimum
LF of 6.33(LL + IL) for the first plastic hinge was at the SM6-7 joint
for a simple beam with overhangs, it will be proper to calculate the LF
of (LL + IL) for that case and then take into account the reaction left
at the NE support. If the structure is an ideal simple beam with over
hangs, the first plastic hinge would form at the 6.33(LL + IL) increment.
The effect of the reaction left at the NE support was small and 5.88(LL + IL) is the calculated increment to form the first plastic hinge when
taking into account the end reaction at the NE support. This value agreed
well with the 6.25(LL + IL) experimental value. If the AASHO allowance
for load reduction on a four lane bridge was considered, the bridge would
withstand [1.35 DL + 8.33(LL + IL)] in this load configuration.
All section properties used to calculate LF of (LL + IL) in
Table 8.1 were based on the measured values.
After demolishing the bridge, the joints where failure occurred
were carefully examined and it was found that the five positive tendons
in each web were completely broken through.
Although the side span positive tendons were adequately propor
tioned by the design procedure which assumed ideal three-span continuous
beam action. The positive moment reserve was reduced in the main span
because the design did not consider the upward unrestrained end support
202
condition. Therefore, it is also necessary to check the LF for the
loading condition which would produce maximum moment at the center of the
main span, as shown in Fig. 8.45 In the loading case of Fig. 8.45 it is
certain that the end supports (SE and NE) would raise up at failure
(because the supports raised up in the previous test in Sec. 7.3.3.2).
The structure at failure will be a simply supported beam with overhangs
and the calculated LF is [1.35 DL + 3.l3(LL + lL)], as shown in the
following calculation:
~4 at 1.0(LL + lL) E 1051 k-in.
rM ~ MUl - (~l + ME2 + ~3) • 3285 [case (3) in Table 8.1]
LF of (LL + lL) 3285/1051 = 3.13
SE
2.81 Ie "~AT l.otLL..f.IL)al 051 Ie-IN.
--------------------~a:E:!2ii5B~:z:::c~1.;8;5 D;;L~./ ,REE
ME 'N 20 • 200·
400·
Fig. 8.45. Failure loading at the center of the main span
Therefore, the support condition does not unduly affect the safety of the
bridge, although the main span ~ximum positive moment flexural capacity
is reduced to [1.35 DL + 3.13(LL + 1L)] if the AASHO load redu~tion for
mUltiple lanes is ignored. This would be [1.35 DL + 4.l7(LL + lL)] if
the normal design specifications are used for a four lane bridge. While
this load case was not tested to failure, the good agreement of other
flexural test results and calculations indicated that this value would
undoubtedly have been attained.
203
In order to match the test loading conditions, the model's
external dead load moment was computed with 1.0 OL acting on a balanced
cantilever and 0.35 OL acting on the completed continuous structure. It
has been shown that because of the construction sequence it is not logical
to base the analysis of the completed structure on fully continuous beam
dead load moments for 1.35 OL. A more rational load factor procedure
for computation of the ultimate design moments in the completed structure
should consider possible uncertainty in the dead load at various stages
of construction. Based on experience in this program, the following
factors are suggested for analysis of the completed structure to check the
negative moment and shear capacity:
Load factors are chosen to conform to the BPR general load factor
philosophy
v = 1.35 OL + 2.25(LL + IL)
For a segmental bridge erected in cantilever, during the construc
tion phase Mu ~ Mul based on
Vl = 1.35 OLI + 2.25(LLI
+ ILl) to be computed for a balanced can til ever
Also, upon completion Mu ~ Mu2 + Mu3 ' where
where
V2 =
V3
OL I OL3
1.35 OLI
to be computed for a balanced cantilever
1.35 OL3 + 2.2~(LL3 + IL3
) + SL to be computed for the completed cont~nuous structure
dead load during cantilevering
dead load applied after completion of closure (topping, r ail ing, e tc . )
LLI live load due to construction operations
LL3 design live load
ILl = impact load of construction operations
IL3 = design impact load
SL resultant reactions due to prestressing of tendons and seating forces at outer supports
204
Negative tendons can be designed by WSD or USD to balance the dead
load of segments and the weight of construction equipment on the segments
during the balanced cantilever stages. However, the ultimate negative
moment capacity of the cantilever structure should be checked for Ul. The
ultimate negative moment capacity of the completed structure should be
checked for U = U2
+ U3
.
In determining positive moment tendons it will be unconservative
to use U2 = 1.35 DL1 . A highly conservative approach would be to use
U2 = 0.90 DLl computed for the balanced cantilever and the preceding
equations.
8.3.3.2 Shear. Initial formation of shear cracking is practically
independent of the amount of web reinforcement,26 so this design appears
to have an adequate safety factor for shear withouttndue reliance on shear
reinforcement, since the initial diagonal tension crack appeared at the
[1.35 DL + 3.25(LL + IL)] increment. This exceeds the specified design
ul tima te load.
Observation of cracks around the main pier showed there was no
shear weakness due to the epoxy joint. The flexural cracks which formed
on the top slab at the first joint in the main span did not extend
straight along the joint and these flexural cracks connected to the
diagonal tension cracks in the web. eracking was frequent along the
webs and the extremely well-distributed crack patterns of Figs. 8.34 and
8.35 indicate that both the epoxy joints and the grouting worked well.
The truss analogy is widely accepted as a simple and safe design
procedure for shear. Shear reinforcement at ultimate strength can be
checked by ignoring the effect of the concrete, but a portion of the
shear is carried by concrete at the ultimate.26
The ACI Code2 specifies
that shear reinforcement should be not less than A = (V - ~V )/(~f ). v u c y
The ACI Code also specifies another equation for shear reinforcement (not
less than Av = (A 180) X (f'/f ) X (sid) xJd/b'. p s Y
It is suggested by Lin26 that the critical section for shear
computation be taken at a distance d (= 15.4 in.) away from the theoreti
cal point of maximum shear. However, the critical section for shear was
considered to be at the first joint from the pier in this bridge because
the flexural cracks occurred at the top slab of the first joint and
.
205
extended into the diagonal tension cracks. Many of the diagonal tension
cracks appeared around the first joint in the main span. The LF of
(LL + IL) for shear capacity was calculated by using the prestressed con
crete equations in the 1963 ACI Code. 2 Since the bridge at failure was a
simply supported beam with overhangs, shear and moment due to dead
and live load were calculated for a simple beam with overhangs, as
follows:
(a) Shear capacity carried by concrete and web shear reinforcement Shear carried by concrete:
M V I r;:r cr ci = 0.6b d ¥fc + M/V _ d/2 + Vd
= 0.6 X 8.81 X 15.4 X 0.0842 + 1490/36.0 + (10.9 + 28.6)
= 6.85 + 41.4 + 39.5 = 87.75 kips
where b ' = 8.81
where
d = 15.4
~ = J7090/l000 '" 84.2/1000 = 0.0842 ksi c
fpe = -PIA - M/WT = -338/(2 X 179)- 3i8XX1~2~0
:: -0.944 - 1.027 = -1.971 ksi
fd (= due to 1.35 DL) = M/WT = 2630/2040 + 990/2090
= 1.29 + 0.474 = 1.764 ksi
M '" 1. (6 Jf' + f - f ) cr y c pe d 2090(6 X 0.0842 + 1.97 - 1.76)
= 1494 k-in. d M/V - 2 = 572/13.1 - 15.4/2
= 43.7 - 7.7 = 36.0
Vd(due to 1.35 DL) 10.9 + 28.6
39.5 kips
V = b'd(3.5/f' + 0.3f ) + V cw c pc P 8.81 X 15.4(3.5 X 0.0842 + 0.3 X 0.944) + 0
= 78.5 kips
b' = 8.81 in.
d = 15.4
Jf:. = 0.0842 ksi c f 0.944 ksi pc V o kip p
Therefore,
V .. 78.') kin!': (= V )
206
where
Shear carried by the shear reinforcement:
V A d f /s s v y = 0.165 X 15.4 X 70/2.5 = 71.1 kips
A v 0.0206 X 8 = 0.1648 in. 2
f y = 70 ksi
d = 15.4 in.
s = 2.5 in.
Total shear carried by concrete and shear reinforcement:
Vu = ~(Vc + Vs ) = 1.0(78.5 + 71.1) = 149.6 kips
(b) Shear due to dead and live load
V 1.3s(DL shear) + LF(LL + IL shear) u
39.5 + LF X 13.1
(c) LF of (LL + IL) for shear capacity at the first joint in the main span
LF of (LL + IL) = (150 - 39.5)/13.1 = 8.44
(d) Shear developed in test at flexural failure
Vu = 1.3s(DL shear) + 6.2s(LL + IL shear) = 121.4 kips
Therefore, the shear capacity was not critical at the time of
flexural failure. The test developed 81 percent of the calculated shear
capacity prior to the flexural failure, and indicates successful jointing.
As explained in Sec. 7.2.1.2, simulation for shear was not as
accurate as that for bending moment. The LF of (LL + IL) for the ideal
loading condition would be slightly higher than the value obtained in
the model test.
8.4 Punching Shear Tests
After failure of the bridge in the longitudinal tests, a series
of punching shear tests were performed by using a rear wheel (HS20)
loading pad in the north side span which did not get any appreciable
damage. Since the positive tendons in the main span were broken and wide
flexural cracks appeared around the main pier segments in the overall
failure test, no estimate of the amount of residual compressive stress in
207
the top slab is possible. Compressive stress of 150 to 250 psi would
exist in the longitudinal direction in the top slab where punching shear
tests were performed, if the grouting were still effective. Ultimate
two-way shear stress V was assumed as 4 ~ for the calculations, as used u2 c
in reinforced concrete. If longitudinal compressive stress in the top
slab is taken into account, the principal stresses will increase 25 to
45 percent.
Most of the experimental values of punching shear were approxi
mately twice the ACI value in the normal punching shear test [case (A) in
Table 8.2]. Even if the increase in shear stress due to the longitudinal
compressive stress is considered it appears that the value of the ultimate
shear stress as specified by the 1963 ACI Code is extremely conservative.
Punching shear tests on the cantilever overhangs of case (B) in Table 8.2
were an abnormal condition, ~ince wheel loads would not be able to be
applied this close to the edge in the prototype. Several flexural cracks
appeared at 60 to 70 percent of the punching shear load and ringed the
loading pad. Shear failure still occurred in a small area around the
loading pad. Even under this abnormal condition, failure loads were more
than eight times that of a service load level rear wheel (HS20), and 20
to 30 percent more conservative than the Code values. Thus, the bridge
should have no problem with punching shear. Punching tests across the
joints show there was no weakness of the epoxy joints under punching
shear in the top slab.
TRANSVERSE CASE POSITION OF
LOADING
P (l.
~ I \ I I
A
P
~ \ I 0
P i 1 I ~}.- \ I B I
TABLE 8.2. PUNCHING SHEAR TEST RESULTS
LONGITUDINAL POSITION ULTIMATE AVERAGE UL T.
OF LOADING LOAD (KIPS) LOAD AND 'S'
(KIPS)
SJOINT
9.81
11.8
t ; {JOINT 9.30 10.3 (1=0.9)
tdJOINT 10.3 JOINT
1 ~ 4JOINT 10.9 10.9
5.63 FllEE ~JOINT END
4.76 5.12
(!=0.37) FllEEQJOINT END· JOINT 4.97
L.F. OF (LL+IL)
= LF. OF 1.3 LL
17.8
18.9
8.85
CALCULA. BY EQUA. IN ACI
(KIPS)
5.09
4.11-
* ASSUME ONE FACE IS OPEN
N o 00
C HAP T E R 9
SPECIAL MEASUREMENTS
9.1 General
The unique nature of this type of construction and the extreme
efforts to maintain close similitude between the model construction and
loadings and that to be expected in the prototype structure at Corpus
Christi presented an opportunity for comparison between laboratory and
field measurements. A modest program of field measurements was under
taken on the prototype structure to obtain information similar to that
found in the model. Measurements were limited to strains in a few
segments at the time of construction operations and during a load test
using maximum weight limit truck vehicles upon completion of the bridge.
A brief summary of the results of these tests is included in this chap
ter, along with the report of a special model test undertaken to verify
the shear capacity of the segments as part of a study of the effect of
minor anchorage cracking in the webs. This latter study was undertaken
as a result of minor cracking which occurred in the prototype structure
during stressing and was not related to the main model tests described
in the earlier chapters.
9.2 Prototype Instrumentation
Four segments were chosen for a limited field instrumentation
program. Figure 9.1 indicates the location of strain gage stations. Two
of the segments chosen were immediately adjacent to a main pier unit, one
segment was at the third point of a side span, and the other segment was
near the midpoint of the main span. Sixteen inch lengths of #3 deformed
reinforcing bars were instrumented with strain gages at the Ba1cones
Research Center Laboratories. These bars were used as a form of extensom
eter and were buried in the flanges of the segments, being wired to
209
210
13 -9/18/72
-10/26/72
2 -10/26/72
3 -10/12/72
- 10/11/72
x -- LONGITUDINAL STRAIN GAGE 0-- TRANSVERSE STRAIN GAGE
Fig. 9.1. Location of strain gages in prototype bridge
211
reinforcing bars in the assembled cages in the forms in the precast yard.
Lead wires were run to junction boxes, which were cast into the curb units
of the segments. All cages were thoroughly waterproofed and protected
from concrete placement and vibration hazards. The gaging followed the
same general pattern as described for the model tests.
At the time of erection of a segment, the gages were connected to
a strain indicator and measurements were made of the strains during
placement. Results on similar gages were averaged, as in the data reduc
tion in the model tests. The general strain instrumentation program was
successful, with over thirty of the gages giving readings during various
construction stages.
9.3 Measurements during Construction
9.3.1 Strains during Balanced cantilever Construction. Figure 9.2
shows the strains measured in the Ml and Sl segments immediately adjacent
to the main pier segment during two stages of the balanced cantilever con
struction. For comparison the values of strains measured in the model at
the same stages of construction are shown, as well as the strains predicted
for the model by beam theory. These predicted strains should be of the
same general magnitude for the prototype construction, since the only differ
ence in the method of calculation would be a slight variation in modulus of
elasticity between the model and the prototype concrete. The prototype con
crete would have a 5 to 10 percent higher value of modulus of elasticity,
which would result in a reduction of 5 to 10 percent in the magnitude of
the calculated strains using beam theory for a given loading condition.
Values of measured strain are shown both at the time of erection of
the Ml and Sl segments and at the time of erection of the M7 and S7 segments.
Strains measured in the prototype at time of erection of Ml and Sl show sig
nificantly less scatter in the top flange. This is undoubtedly due to the
much greater distance of the strain gage from the anchorage. The compressive
stresses noted are only approximately half the magnitude of those found in
the model, while slightly higher tension was noted in the prototype lower
flange than in the model. The existence of this tension confirmed the
212
ERECT MI,SI -100 t BRIDGE
...... Z "- -50
o MODEL ~f • 0
• PROTOTYPE « .... a: ....
--- 8M THEORY C/) u
-2~~ + .... 25
ERECT M7, S7 <t BRIDGE
. u -205~ ~ t -.:;:---~---I--I - .... 25 ~-::i.
~ r:~~ Fig. 9.2. Measured strain in Ml and Sl segments during
several stages of balanced cantilever construction
213
tendency of the lower flange to open under initial erection stresses. The
inability of the fresh epoxy joint to resist such tensile stresses was
compensated for by provision of temporary stress across the joints.
Strains measured in the prototype adjacent to the piers at time
of erection of the 7th segments indicate substantially less compressive
stress at the piers and suggest substantial friction losses in the prototype
construction. On numerous occasions, difficulty was encountered in
stressing cables in the prototype because of suspected strictures in the
tendon conduit and apparently high friction.
Because of the low magnitude of strains in these units and the
sensitivity of the field equipment, the data measurement cannot be con
sidered highly accurate. Substantial scatter was noted between individual
data points in several cases. The authors feel that the field measure
ments during erection can only be regarded as a general indication of the
actual strains experienced in the prototype structure and do not neces
sarily contradict the results of the model tests. Unfortunately, they
do not necessarily confirm the results either.
9.3.2 Closure Operations. A greatly improved opportunity for
making a number of strain measurements in the prototype was presented by
the closure operations. In contrast to the balanced cantilever erection
techniques which were scattered throughout several months, final closure
stressing was accomplished in one day. After all of the main span positive
tendon cables were inserted, tendons A1 through A4 were sequentially ten
sioned, the end reactions were jacked to attain proper elevations, and
tendons AS and A6 were prestressed. This was a slightly different sequence
than that used in the model, wherein tendons A1 through A6 were stressed
and then the end supports were raised. During the tensioning of all of
these cables, strain readings were taken in segmental units M1, 51, 57,
and M10. Typical comparison of data from the model and the prototype con
struction during closure are shown in Figs. 9.3 and 9.4. The slight
difference in sequence involved in jacking the end supports to the final
elevation is ignored in these comparisons. Figure 9.3 shows that the
values measured in the main span segment immediately adjacent to the pier
214
I
o MODEL
• PROTOTYPE
- -- 8M THEORY
PRESTRESS AI • • (.,.) -25~ " ~ t 0 ~----
:Z I- 25
~ (.,.) -25~ 4 t 0 ~ 25 CI) I-
PRESTRESS A2
• PRESTRESS A4 Cl. BRIDGE
Il BRIDG PRESTRESS A5
~ r:Ilt:.:.:::I::-=-:=tj ...... I- 2!5~ Z
! f:~~ t;;
PRESTRESS A3 It BRIDGE
~ (.,.) -2!5~ ~ t 0 ~ I- 215 CI)
PREST RESS A6
_ (.,.) -2!5~ !> • ~ t 0 ----
~ I- 215
1
It BRIDGE
It. BRIDGE.
t BRIDGE
Fig. 9.3. Measured strains in Ml segments during closure operation
rc IM9
1 { " 220 I~[ " 180 J o MODEL
• PROTOTYPE
--- 8M THEORY
PRESTRESS AI PRESTRESS A4
PRESTRESS A2 It BRIDGE PRESTRESS A5
PRESTRESS A6 PRESTRESS A3 It BRIDGE
Fig. 9.4. Measured strains in the model M9 segment and prototype MlO during closure operations
215
1
Ii.. BRIDGE
ct.. BRIDGE
It BRIDGE
216
are in very good agreement in the model and the prototype tests, except
for the stressing of the initial cable. In this case the prototype com
pressive stresses substantially exceed the model and the theoretical
stresses. Substantial difficul ty was reported in the prototype stressing
in obtaining the required elongations. During stressing of tendons AI,
the rams were bled off, then brought back to required pressure several
times before the proper elongation was obtained. Greatly improved agree
ment was noted in the strains near midspan. The comparison shown in
Fig. 9.4 is not completely valid, since the model was instrumented in
segment M9 while the prototype was instrumented in the adjacent segment
MlO. However, these two units are close enough together to give a good
general indication of the nature of strains. The beam theory line shown cor
responds to the model M9 segment. It is almost exactly the same as
the values that would be calculated for the MlO segment.
The generally good agreement shown during the complex closure
stage indicates the applicability of the elastic calculations for this
type of structure.
9.4 Prototype Load Testing
Upon completion of all construction and immediately prior to
opening the structure to traffic, a modest load test was performed to
check on design and calculation procedures. Since it was impractical to
impose the design lane loadings used in service load level tests of the
model structure and in the actual design of the bridge, two heavily loaded
trucks were used to obtain "influence lines" for bending strains at
the various locations in the structure where strain gages had been
implanted in segments. The time available for load testing was very
limited and an extremely modest program was undertaken. However, it is
felt that the results are significant and further indicate the validity
of general calculation procedures and the completely continuous nature of
the actual structure. Two trucks approximating HS20 loading were used as
shown in Fig. 9.5. Because the single axle allowable load limitation in
Texas does not allow the HS20 vehicle to be operated on the highways, the
trucks had tandem rear axles. They were loaded with sand and carefully
.,'
•
218
weighed. The final weight of the two trucks was approximately 2 percent
higher than that of an HS20 truck. Actual axle weights are shown in
Fig. 9.5(b) as averages for the two trucks. No axle value varied more
than 5 percent between the two vehicles.
During the load testing an individual truck was first positioned
in various locations on the bridge. In general, the measured strain values
were so low as to be unmeaningfu1 within the general accuracy of the
equipment. In order to obtain relatively measurable strains, the two
trucks were placed side by side on one-half of the bridge with the centers
of the trucks being essentially above the box girder webs. The trucks
were moved to an initial position with the centroid of the load at
Station 0 plus 25 ft. from the side pier. A complete set of readings
was taken of all gages in each of the four segments which were instrumented.
The pair of trucks were then advanced 25 ft., stopped, and a new set of
readings was taken. This sequence was continued for the length of the bridge.
Results of typical strain measurements are shown in Figs. 9.6
through 9.9. The measured strains at a given station as the truck
advanced along the structure represent a strain influence line. On each
figure, the measured strain values are compared to a theoretical strain
influence line. This influence line was calculated in the following
manner. Using elastic theory and assuming no variation in EI along the
member, influence lines were obtained for bending moment at stations
corresponding to the middle of the segments indicated. These values were
obtained in terms of nondimensiona1 ordinates for a nondimensiona1 load
and length. Multiplying by the combined weight of the trucks and the
length of the span, influence lines were obtained for a bending moment at
the indicated measurement stations. The values of bending moment are then
divided by the average section modulus for the structure and by an assumed
modulus of elasticity of 4.62 .~ 106 psi for f' = 6000 psi concrete. In c
this way the influence line ordinates were reduced to equivalent total
strain values for the whole bridge. However, since the loading was applied
only above one box of the section, and since the loading was resisted by
x p o \"""-e--I. 7
[Measured strain
+2
Strain Ot-====~::~!:::~~~~~~----------------------------------------~~---3~~--__ ======~~~ Micro in.! • • in.
-25 • • •
• , Measurement station
~eoretical strain
Fig. 9.6. Influence line for bottom flange strain, Station SSI
Truck Truck
+50 p
\ 1I }I
/ X \ / Measured strain
+25
Strain !~===!:::::i:::::)r=~~~~~.---------------------------------~.L~~~~------------------~ Micro in. / d~.'
sta tion
-25 Theoretical strain • •
-50
Fig. 9.7. Influence line for bottom flange strain, Station MSI
N N o
Truck Truck
r ~ ~ ~ .l
+50 , 7 " 7 X
Measured strain
+25 •
Strain 0 ~==~~===-==~-==-.. --~~--------------------------------~~~----------~----~tmicro. in. lin.
-25
-50
• Ie • Measurement station
Theoretical strain
Fig. 9.8. Influence line for bottom flange strain, Station SS7
N N .....
N N N
Truck Truck
I , , t
+50 r \ 7 \ . 7 X
! • • Measured strain
• +25
• _____ Theoretical strain
Strain ••• • Micro in./in.O~~--==============~~~~----------------------------------------~~.-=========---~--~~
t Measurement station
-25
-50
Fig. 9.9. Influence line for bottom flange strain, Station MSlO
223
two boxes in an unsynnnetrical fashion, this value of strain would be
unrealistically high. Using the results reported in Sec. 7.3.2.2 for
loads on one box only, the measured value of strains on the loaded box
as a function of total strain was computed. It was found that for almost
all load cases, th is ra tio was 70 percent. Accordingly, Figs. 9.6 through
9.9 were plotted assuming that the same proportion (70 percent) would hold
in the prototype.
Comparison of the measured strain values with the theoretical
influence lines in Figs. 9.6 through 9.9 indicates generally good agree
ment. In all cases, the general shape of the influence line is very near
that of the distribution of the measured points. At the maximum value the
measured strains run 10 to 30 percent higher than the theoretical strains.
This could be due to calculation inaccuracies, since the calculation pro
cedures assumed a uniform cross section and station MS10 in particular is
somewhat less rigid than the pier sections, or to errors in the assumption
that the strain in the loaded boxes would be approximately 70 percent of
the total strain. It is particularly interesting that the strains noted
in segment MS10 seem to be most in error. Since this unit is farthest
away from the rigid pier section with its diaphragms, the load distribu
tion between the two boxes is probably the least effective.
In general, the agreement between measured and computed points is
well within the state of the art at this low level of load. It is parti
cularly reassuring that even under these heavily loaded trucks the maximum
change in stress noted would correspond to less than 1500 psi steel stress,
or 250 psi concrete stress. These tests confirm the adequacy of the
design procedure.
9.5 Ultimate Shear Test
During the actual construction of the prototype structure in
Corpus Christi, a series of inclined cracks appeared in the webs of units
being stressed. These fine cracks generally became visible immediately
ahead of the anchorage assembly and propagated along the inclined tendon
path into the flange. No similar cracking had ever been noted in the model
224
construction. Careful examination of details indicated that the manufacturer
of the post-tensioning system and the contractor had not utilized the
spiral confinement along the tendon path, as shown in Fig. 4.6. In order
to investigate various factors which may have caused or contributed to
this cracking and to determine if the inclined cracking reduced the
anchorage or shear strength of the units, a special test was programmed
using three model segments. These segments were constructed to model the
"wors t probable cons truc tion conditions" in the prototype. Any error or
omission which could be attributed to prototype construction was modeled
in the units. Several segments had transverse bars cut or relatively
small portions of the web reinforcement omitted. In general, the units
had a much reduced spiral reinforcement along the tendon path to model the
exact anchoring used in the prototype.
Three segments were erected using the epoxy used in the prototype.
The level of prestressing used corresponded to the values for segments M1,
M2, and M3. The units were not erected in a balanced cantilever fashion,
but rather cantilevering off of a pier and steel frame, as shown in Fig. 9.10.
During the stressing of each segment, careful observation was made
of the webs and very definite cracking similar to that in the prototype
was noted in the webs of segment 1. Later visual examination of the
interior of the webs indicated the same type of cracking inside the webs
of segment 2, but no visible cracking was noted in segment 3. A positive
moment tendon similar to tendon A1 was then inserted in segment 1 (near
the pier) and the tendon was taken up to the design value and then a 50 per
cent overload with no apparent damage in the positive moment anchorage
vicinity. All indications were that the inclined cracks were slightly con
trolled by the stressing of the positive moment tendon.
In order to determine the behavior of the cracked units under heavy
shear loads, hydraulic rams were then used to apply shear loads to the
outer or third segment. Initially the rams were brought up to the level
corresponding to the dead load shear for the structure. Loading was then
increased to 1.35 dead load with nothing dramatic happening. Load was
then increased in increments of one-half the design live plus impact loading,
Fig. 9.10. Test arrangement--ultimate shear test
226
based on the maximum shear loading and the heavy concentrated load which
accompanies shear at the support. Initial inclined shear cracks occurred
at a level of 1.35 dead plus 2.25 (live plus impact). These cracks were
generally in the direction opposite to those formed along the tendon paths. Due
to flexural capacity difficulties, additional rams were inserted at the
midpoint of the second segment and the loading was increased by 0.25 (live
plus impact) values.
Loading was increased to 1.35 dead plus 3.0 (live plus impact) and
a very large, flat inclined crack appeared extending almost the full length
along segment 1. The crack started about 2 in. from the support and
extended to within about 2-1/2 in. of the joint. It was a very low crack
going only up about to the height of the specimen. Live load was increased
in increments to 1.35 dead plus 5.0 (live plus impact). The inclined
shear crack was quite wide and extended through the epoxy joint up towards
the load point. Several substantial inclined cracks passed right through
the tendon anchorage near the top of the shear key and extended into the
web of the adjacent segment. Because of the high moment, the jacks on the
outer segment were dropped down to 1.35 dead load and they were maintained
at that level for the remainder of the test. The jacks on the second seg
ment were incremented up to 5.0 (live plus impact) and then increased in
small increments at a level of 6.0 (live plus impact). The shear crack,
which extended from the top of the shear key in the first segment down
through the lower flange, was extremely wide. Additional shear cracks began
to appear below the shear key toward the bottom of the web. In the bottom
flange it was obvious that the web reinforcement began to yield in the
shear key area. Some evidence of cracking on the bottom flange was visible.
Loading was advanced but at 1.35 dead plus 6.1 (live plus impact) a massive
shear crack opened below the shear key and a compression crack extended
well back along the web. The entire joint ripped apart at the top flange
level and final failure occurred.
Using the calculation procedure for ultimate shear strength outlined
in Sec. 8.3.3.2, the theoretical shear strength can be compared to the actual
shear applied to the model. The actual applied shear at ultimate during
227
the test on the joint between the first and second segments where
principal failure occurred was 61.4 kips. Using the calculation procedure
of Sec. 8.3.3.2, based on a ff = 5400 psi for these segments, 120 kip pre-c
stressing force, and 60 ksi web reinforcement, the calculated ultima te
shear strength would be 63.9 kips. Thus, the test specimen developed
96 percent of the theoretical shear strength. Considering the large
variations inherent in shear tests, this must be considered as an extremely
satisfactory test. This is especially so since the units were made to
represent the worst possible construction and deliberately had some rein
forcement misplaced, altered, or omitted. The results of this test can
be taken as a conclusive indication of the efficiency of properly applied
epoxy joints in segmental construction. Provision of these joints did
not significantly lower the shear strength of the unit.
!!!!!!!!!!!!!!!!!!!"#$%!&'()!*)&+',)%!'-!$-.)-.$/-'++0!1+'-2!&'()!$-!.#)!/*$($-'+3!
44!5"6!7$1*'*0!8$($.$9'.$/-!")':!
C HAP T E RIO
CONCLUSIONS AND RECOMMENDATIONS
10.1 Conclusions
10.1.1 Primary Conclusions. (1) The segmental bridge model
safely carried the ultimate design loads for all critical moment and shear
loading configurations on which its design had been based, as specified
by the 1969 Bureau of Public Roads Ultimate Strength Design Criteria.
Detail results were shown in Sec. 7.3.3.
(2) The deflection under design live load in four lanes (only three
lanes required by live load reduction factors) was approximately L/3200 in
the main span. This is much smaller than L/300 which is generally consid
ered as acceptable.
(3) Positive tendons in the main span were designed as if an
ideal three-span continuous beam. Since the completed bridge was supported
on neoprene pads which have no vertical restraint against uplift, the outer
ends were able to rise off their supports so that the structure did not
act continuously at ultimate conditions under main span positive moment
loading. Even so, there was sufficient reserve strength in the main span
to carry design ultimate load as shown in Sec. 8.3.
(4) Under very high combined moment and shear loading (see Sec. 8.3),
flexural cracks appeared near the epoxy joints in the top slab near the
main pier. However, they joined the diagonal tension cracks and did not
extend along the joints. There was no sign of any direct shear failure at
the joints. In tests of the full bridge model, approximately 75 percent of
the theoretical ultimate shear load was applied in the maximum shear loading
test prior to failure of the bridge during that test by flexure. No sign
of shear distress was evident. Subsequent tests of a three-segment model
under severe shear loading as a cantilever section indicated that full
229
230
shear strength of the unit was developed. Hence, the epoxy joint technique
used did not reduce the design shear strength.
(5) During erection of the first few segments, tensile stress
occurred in the bottom slab as predicted in the design (Sec. 6.l.5(b) and
Sec. 6.2.3). Temporary prestress devices successfully controlled the
effects of these stresses.
(6) Theoretical calculation of the load factor for live and impact
loads required to form the first plastic hinge agreed very well with the
experimental results, as shown in Sec. 8.2 and 8.3. These tests proved
the accuracy and applicability of the ultimate load calculation procedure
of Sec. 8.2.3.
10.1.2 Secondary Conclusions. (1) Near failure, major cracks con
centrated near the epoxy joints which had no continuous conventional
reinforcement (Secs. 8.2 and 8.3). However, throughout the loading
sequence cracks were generally well-distributed because of the effective
grouting and the strength of the epoxy joints (Secs. 7.3.3, 8.2, and 8.3).
(2) While behavior of the epoxy joints was quite satisfactory, it
should be considered that the model segments were joined in a dry condition.
Ref. 19 indicates that while most of the epoxy resins performed adequately
for joining dry specimens, the strengths developed by most of the epoxy
joints were very weak when joined with concrete segments in a saturated
condition.
(3) Transverse moment capacity of the bridge cross section was very
adequate, as shown by the punching shear load test results of Secs. 7.3.4
and 8.4.
(4) There was no adverse effect of the epoxy joints on the slab
punching shear strengths, as shown in Sec. 8.4.
(5) Bolts used for the temporary connection of the pier segments
to the main piers yielded locally under the most critical unbalanced loading,
although the calculated direct compressive stress was less than the actual
yield strength. The bolts used in the model were also below the yield
strength specified for the bolts in the prototype. Yielding was apparently
231
caused by the large gap between the pier segments and the pier, with
consequent local bending, and was accentuated by the stress concentrations
in the threads (Chapter 6).
(6) Most of the theoretical calculations were in good agreement
with the experimental results, although there were some appreciable devia
tions between the experimental and theoretical values of strain in the
top slab in some stages of cantilever construction (Chapters 6 and 7).
The BMCOL50 program was very useful in predicting the behavior of
the bridge during construction and for uniform loading tests. The BMCOL50
results agreed very well with the experimental results for longitudinal
strains and deflections. The relatively simple data input for BMCOL50 is
another advantage when compared to the folded plate theory programs.
The SIMPLA2 program reasonably predicted the variation of the
longitudinal strain under very high stress levels across the top slabs
of the newly erected segments.
The MUPDI program, which can be used only for a constant cross
section, agreed very well with t~ experimental results at the service
load level. The variation of cross section along this bridge was very
small. MUPDI can be used to determine the transverse moments and shears
under unsymmetrical loading and can be used effectively in designing the
transverse reinforcement.
(7) The initial overstressing to 0.8f' with release to 0.65f' 10 s s
before seating, suggested by Brown, worked well. The friction factor
and wobble coefficient used with SIMPLA2 were reasonable, as confirmed by
the tests (Sec. 6.2.2).
(8) Separation of the match cast segments was smoothly carried out
without any damage to the segments, by careful application of uniform
force using hydraulic rams (Sec. 4.4).
232
10.2 Recommendations
10.2.1 Design Recommendations. (1) Since the structural
configuration changes from a statically determinate cantilever structure
to a mU1tispan continuous structure during construction of this type of
bridge, the ultimate design load for effects (moment, shears, etc.) for
this bridge should be specified as two values which must each be satisfied,
as follows:
and
where U1 = 1.35 DL1+ 2.25 (LL1 + ILl) to be computed for a balanced cantilever
U2
= 1.35 DL1 for negative moments and 0.9 DL1 for positive moments to be computed for a balanced cantilever
1.35 DL3 + 2.25 (LL3
+ IL3) + SL to be computed for the completed continuous structure
= dead load during cantilevering
dead load applied after completion of closure (topping, railing, etc.)
= live load due to construction operations LL1
LL3 = ILl = IL3 -
SL
design live load
impact load of construction operations
design impact load
= resultant reaction due to prestressing of the tendons* and seating force at outer supports
(2) Negative tendons should be designed so that no tensile stress
is developed across any joint during erection. Otherwise some temporary
erection procedure must be required to keep the joint in compression until
erection stresses change from tension to compression.
(3) In calculating the internal ultimate moment in the positive
moment region, possible contributions of the negative tendons present at
that section should be considered.
*Resu1tant reaction is zero for the tendons stressed while a determinate structure.
233
(4) Negative moment and shear capacity must be checked for both
cantilever erection stages and for the completed structure under design
and ultimate loads, as shown in the design of this bridge.2l
(5) If the outer support details provide no restraint to upward
vertical movement, this should be considered in designing positive moment
tendons. Alternate solutions should be examined to decide whether the
positive moment prestress should be increased, vertical restraints provided
at outer supports, or outer spans lengthened. The designer must be aware
of the effects of change of structure configuration if the side spans
rise up from their supports under ultimate loading in the central span, as
shown in Sec. 8.3.3.1.
(6) The jacking forces required to adjust the end reaction or
elevation of the bridge should be carefully calculated to prevent prema
ture cracks at service load levels, as shown in Sec. 7.3.3.2.
(7) The sequence of positive moment tendon stressing operations
should be specified to minimize or preferably eliminate tension in the
top slab, especially at the closure segment.
(8) Although the model bridge was safely supported by the bolt
details used during cantilever erection, Sec. 6.2.1 and Fig. 6.25 suggest
an improved temporary support system to reduce the high compressive stresses
on the bolts in the unbalanced loading condition and stiffen the connection
and thus reduce unbalanced deflections.
(9) Although the effects of creep and shrinkage were minimal in
this study, Muller30
points out for this class of structure:
The effect of steel and concrete creep must be considered with regard to moment distribution, together with the possible effect of moment reversal. Final adjustment and compensation for shrinkage and concrete creep may help the structure to reach the optimum equilibrium.
10.2.2 Construction Recommendations. (1) The pier segments should
be carefully placed on the piers to close vertical and horizontal alignments,
in order to minimize the final closure adjustments.
234
(2) If practical, positive tendons in the main span should be
inserted before casting the closure segment, in order to make sure that
ducts are clear, since concrete may penetrate into the tendon duct near
the joint during casting of the closure joint.
(3) If there is any small damage on the surface of the segment,
it will be better to fill it with the epoxy material at the time of jointing
than to pa tch it earlier and risk the danger of the uni ts not ma ting due
to an excess of patching material.
REFERENCES
1. Aldridge, W. W., and Breen, J. E., IIUsefu1 Techniques in Direct Modeling of Reinforced Concrete Structures, II Models for Concrete Structures, ACI Publication SP24, 1970, pp. 125-140.
2. American Concrete Institute, Building Code Requirements for Reinforced Concrete (ACI 318-63), Detroit, June 1963.
3. American Concrete Institute, Building Code Requirements for Reinforced Concrete (ACI 318-71), Detroit, 1971.
4. American Concrete Institute, IIGuide for Use of Epoxy Compound with Cbncrete (ACI 403),11 Journal of the American Concrete Institute, August 1962.
5. American Society for Testing and Materials, IIStandard Method of Test for Working Life of Liquid or Paste by Consistency and Bond Strength (ASTM D 1338-56),11 ASTM Standard, Part 16, June 1969, pp. 498-500.
6. American Association of State Highway Officials, Standard Specifications for Highway Bridges, Tenth Edition, American Association of State. Highway Officials, Washington, D.C., 1969.
7. Bureau of Public Roads, Strength and Serviceability Criteria, Reinforced Concrete Bridge Members, Ultimate Design, Second Edition, Bureau of Public Roads, U. S. Department of Transportation, Washington, D.C., 1969.
8. Breen, J. E., and Orr, D. M. F., "A Rapid Data Acquisition and Process Sys tern for Structural Model Usage," Proceedings of Struc tura1 Model Symposium, Sydney, Australia, 1972.
9. Breen, J. E., and·Burns, N. H., "Design Procedures for Long-Span Prestressed Concrete Bridges of Segmental Construction," Research Study Proposal, Center for Highway Research, The University of Texas at Austin, 1968.
10. Brown, R. C., IIComputer Analysis of Segmentally Constructed Prestressed Box Girders,1I unpublished Ph.D. dissertation, The University of Texas at Austin, 1972.
11. Brown, R. C., Burns, N. H., and Breen, J. E., "Computer Analysis of Segmentally Erected Precast Prestressed Box Girder Bridges," Research Report No. 121-4, Center for Highway Research, The University of Texas at Austin, November 1974.
235
236
12. Corley, W. G., Carpenter, J. E., Russell, H. G., Hanson, N. W., Cardenas, A. E., He1gaston, T., Hanson, J. M., and Hognestad, E., "Design Ultimate Load Test of 1/10-Sca1e Micro-Concrete Model of New Potomac River Crossing, 1-266," Journal of the Prestressed Concrete Institute, Vol. 16, No.6, pp. 70-84.
13. Ferguson, P. M., Reinforced Concrete Fundamentals, Second Edition, John Wiley & Sons, Inc., New York, 1965.
14. Gallaway, T. M., "Industrialization and Modeling of Segmentally Precast Box Girder Bridges," unpublished Master's thesis, The University of Texas at Austin, 1971.
15. Henneberger, Wayne, and Breen, John E., "First Segmental Bridge in the U.S.," Civil Engineering, Vol. 44, No.6 (June 1974), pp. 54-57.
16. Ikeda, T., Precast Segmental Construction Method, Nikkan Kogyo News Paper Inc. (JAPAN), 1969.
17. Kashima, S., "Experimental Study of Epoxy Resins as a Jointing Material for Precast Concrete Segments," unpublished Master's thesis, The University of Texas at Austin, 1971.
18. Kashima, S., "Construction and Load Tests of a Segmental Precast Box Girder Bridge Model," unpublished Ph.D. dissertation, The University of Texas at Austin, January 1974.
19. Kashima S., and Breen, J. E., "Epoxy Resins for Jointing Segmentally Constructed Prestressed Concrete Bridges, Center for Highway Research, Research Report No. 121-2, Center for Highway Research, The University of Texas at Austin, August 1974.
20. Lacey, G. C., and Breen, J. E., "State-of-the-Art--Long Span Prestressed Concrete Bridges of Segmental Construc"tion," Research Report 121-1, Center for Highway Research, The University of Texas at Austin, May 1969.
21. Lacey, G. C., "The Design and Optimization of Long Span, Segmentally Precast, Box Girder Bridges," unpublished Ph.D. dissertation, The University of Texas at Austin, May 1969.
22. Lacey, G. C., Breen, J. E., and Burns, N. H., "Long Span Prestressed Concre te Bridges of Segmental Cons truc tion--Sta te-of- the-Ar t," Journa 1 of the Prestressed Concrete Institute, Vol. 16, No.5, October 1971.
23. Lacey, G. C., and Breen, :I. E., "The Design and Optimization of Segmentally Precast Prestressed Box Girder Bridges," Research Report 121-3, Center for Highway Research, The University of Texas at Austin, July 1975.
237
24. Leyendecker, E. V., and Breen, J. E., "Structural Modeling Techniques for Concrete Slab and Girder Bridges," Research Report 94-1, Center for Highway Research, The University of Texas at Austin, August 1968, pp. 26-28.
25. Leyendecker, E. V., "Behavior of Pan Formed Concrete Slab and Girder Bridges," unpublished Ph.D. dissertation, The University of Texas at Austin, 1969.
26. Lin, T. Y., Design of Prestressed Concrete Structures, Second Edition, John Wiley & Sons, Inc., New York, 1963.
27. Litle, W. A., Cohen, E., and Somerville, G., "Accuracy of Structural Models," Models for Concrete Structures, ACI Publication SP-24, 1970, pp. 65-124.
28. Matlock, H., and Haliburton, T. A., "A Program for Finite Element Solution of Beam-Columns on Nonlinear Supports," unpublished report, The University of Texas at Austin, June 1964.
29. Mattock, A. H., "Structural Model Testing--Theory and Applications," Journal of the PCA Research and Development Laboratories, Vol. 4, No.3, September 1962, pp. 12-22.
30. Muller, J., "Long Span Precas t Pres tressed Concre te Bridge Buil t in Cantilever," Concrete Bridge Design, ACI Publication SP-23, 1969, pp. 705-740.
31. Neville, A. M., Properties of Concrete, John Wiley & Sons, Inc., New York, 1963.
32. Prestressed Concrete Institute (Japan), Construction Manual for Precast Concrete Segmental Construction, December 1968.
33. Prestressed Concrete Institute Post-Tensioning Commi~tee, Recommended Practice for Grouting of Post-Tensioned Prestressed Concrete, PCl Specification Draft, 1970.
34. Scordelis, A. C., "Analysis of Simply Supported Box Girder Bridges," Report No. SESM 66-17, Department of Civil Engineering, University of California, Berkeley, October 1966.
35. Scordelis, A. C., Bouwkamp, J. G., and Wasti, S. T., "Structural Behavior of a Two-Span Reinforced Concrete Box Girder Bridge Model, Vol. II," Report No. US SESM 71-16, Department of Civil Engineering, University of California, Berkeley, October 1971.
36. Swann, R. A., liThe Construction and Testing of a One-Sixteenth Scale Model for the Prestressed Concrete Superstructure of Section 5, Western Avenue Extension," Technical Report, Cement and Concrete Association, London, 1970.
238
37. VIDAR Corporation, "VIDAR 5404 D-Das System Information,1I Technical Manual, California, May 1971.
A P PEN D I X A
PROTOTYPE BRIDGE PLANS
[239]
!!!!!!!!!!!!!!!!!!!"#$%!&'()!*)&+',)%!'-!$-.)-.$/-'++0!1+'-2!&'()!$-!.#)!/*$($-'+3!
44!5"6!7$1*'*0!8$($.$9'.$/-!")':!
~~r~~~~·~'O~'·~U«r~~I-----------
I
, E
ek6ot. I.X)·
NOTE', T~"t P,II.,S ..,ee." .. lO;PI 'U'lOl' 1°w,rn "0" t..1E .. tl> '9."'UZI.
I, I
/:xJ~O"
I ~'a;nul :":;b4·nlJ
--------"" .. ' "',
~~ ___ -_--'''~"''~~- _J?O!~_~~==]~~'Q~~~==~
~TimbQrpJ.r"9
~\
E'
-~':J~~~ :!!'
LAYOUT
INTRACOASTAL :!!p
CANAL BRIDGE PARK ROAD 22 I
.1I~":r7nJ
16 Ni.«: .. ~ 611 I 7 17
SECTION A-A
-----~.
SECTION C-C
HALP ELeVATION
B , F> 5,-7:""-B!""·
51-r."~!J'.B • 5,......-'~'·_.e.-'I· 51-C'.II"'·-f?~ 55-2~II"JO'..,!)' ~ .56-7"_-'l'_t!':~' 51-7:'1'_-'~ I/. •
..5e-- a&'-o..>'" •
.5~-5.·IJJ-7!-'A·
8=,.-..5 51 "f")r. .• 5~
: Bor-3 EJ~E~ I I ~'lI'
~
.Iit-~~~~~
I--_---'e""'o"-, __ I
SECTION D-D
Cb- N= .51zr? ~ f." AU> M
"" 2. 'I. 50'-2' A, 2. ~ ~ ... Z"
~'-2" ~ !!L .~-
!t~ ::: ~,; ------c2 16 ~- 95.6'
:, _ I !=r-~-;;::: E. 10 e :' ~ ----" F, IOC
F, I!) __
F. .'0 J
5, i6: 5, 156 5. <;()
5, 27 5. 27 5, 27 S. S.
I2l:?mf ~~/ Cone cl 'C" Con:- cI 'E"
"10-'10
-' , ' 5
" ' 5 , 5
" ' 5 '5 '5 '5
{J. I srrvc+&n::Y
-tJ3 .:' -.n~:::'''j
-~-~~ z;:-q1 £'_Cr ~i _~
.n·.~ 54' . .:1' 2'~~~ ~':.I;·
Z7~ :y
16 17105 21152 21841 I ~22 11689 ~~92
'""Q;".:. t"
_'~L_
:: ,"';2
':">:~ 6"~ :".<1
~7~~:-o~.:iQ 5Z'-S 1 13'; 1530
I
65 112 I A
12 f7Q7
Ad 9(Jorrtde:s str:-,n arc> b- cYl!!" p.r: "IIU I'll
-. II h,. I • C • I .a, h "'I~" " IL' •• ~" •• ~, •• ,I I.u"."u .,11 'hi ••• tsll~Ch'" ..... , '~I .. ~, ,It tUn' •• '.UI, 1., .,1" .. Hili .... , .. ID I" U' .~ .. I .. 12U " UI ......... 11K ,
"'''IU " ,"'O'U"", "III I. I" e ..... fl •• "" .. "'<1 I •. e .. ( •• , U hnOlh, ••• n p •• " ..... ' ••• 11
I. '0" hit n.h .... 11. Ito ... , .. t.,., ... ·".·.' .... 1111 ........ , .......... '" .... ' ..
......... 1.11
II [.~.". "U", .•• , ........ ",.. U,. '''11 It ," ... t.'or~I:'-"O'I':!opI.Of'~
'"'1' I .. , '''''''-1 •• ", It " tI, ••• "" ,.! "'nil ~, .. ,,' ,"I. I ''''I lUll.,,,, h ... "'1 0",60 .. ' .... , ... , •• '., 111' .. , lou ..... _
... , "." n.~" .',. "'11,\,<""., ."., It •• 1,., on, 1~' ,t .. '. II. "'on." " •• , on. to" .uc.-u. I.,,, \. .... • ..... ' <II II. , ••• ~I' t .... , lin
• p tc' '"., ••• , .. I .... , ... ~ '" , .. ,."' ,., ~II"
S, 'Ut .... II'""'" ""1 h ,III. ," "".11"" ""~ 1111 ....... ..
t.I, .... "." .,,1 •• 'U ........... tit ...... tl,a_ •• , ~IU'I.IU " ........... th C ........ . II 1.,.<. ,I, •• U .,'1 II .... I" •• I,e •••• t "'" .... n.1I II ,u,,,,,"1 II ""~ " ........ t ...... ', .. ,.
TEXAS HIGHWAY DEPARTII£JrfT BRIDGE DIVISIC»I
INTERIOR PIERS
20 & 21
INTRACOASTAL CANAL BRIDGE
N ~ N
t:r:Hs :!rrl M?rflCO/-+--fI-'fI-'tt-----iHHI
TYPICAL 5:;CTION
1=1=~--===-·-'-11~
nJiY BeARING seAT per.
PLAT!! WK. It
NO~ ;:-'~ k<t 8. (IJG'OcrP O"'r! 3"nut$ ~~h) fo be- LI:!JOa bJYT~1"I
~ ~:g;;::eG~U~f o:::corC7CY'C~ wdfJ A 5rM-"15J
hl a .. "" ••••• ,. ~'HII'"". U I ....... 'J~_ ... ht "un! •• ,l , 1 H, .. , 0,11", ,~t. "Jul v~,1 I. U'''''~n·,~ "'uU •• u n 1_ ....... ",'" 1.1 .• 1 ".1 Unl
) 1, ... u ~I' ~.h_ !~P , _ .. ". '" I •• ,1, •• 1f."H.1I ,I IUHla fat'l .
i"" U"'II""I'~A ~fI ~a"'I'"' " '"'" U'''' .1., !h 'H. 'I U I. III""." •• ,. '"' ~n""1 ,IU u .... 11 "~H'u',n .. ~ .HU ......... 8<, U "01' • " I.
't .. ** tiP ~.H I ,I'~H 1, ... ,,' J h! ~,d<u"t ... , '11 .. 11 •• ', U ~I ........ HI
UHI"" "It .. I, ... , «IP" , ....... Hilt I. U.Io,.
J '''''''. luC. -n .'11;'.' I, "1.1 'I. ' •• 1 t.l' , •• ,." U, .i~" ... ',HI'ff ,,~" .... It 'our '1, hI.
I'. '~!1: It ' .... ,1 .~q I. ,,,' '~IH"I .. "'''~' '~II h,. nOli .... ,~u u ~."".n ,. ,., ~,., II .. ~,.,
• 1>11 •• r~1 UlI, l'fH II I .. or ".,t' ,HI ... ~. U 1ft
,- " '''Hl .. 01, ~U.oI '''' '" .. l 'tIU. ,,~" <. " U tlHU •• ~~ .,"~" t~. al"
TEXAS HIGHWAl' DEPARTMENT B~IOGE DIVISION
PIER DETAILS
PI E RS 20 s\.i.l I NTRACOepTAL CANAL BRIDGE.
_ .......... _ ... ;0(:
... ....,.-~ .....
'~~~----------------------------------~--------------------------~---------~i~"~~~.~,~.~~.~·o~,~~~~~ ____________________ ~ _________________ . ___________________ ~J~'~O' __ __
I I i
l,md of area Ibr
1: ..
TYPICAL SeCTION
100-'0'
eLeVATION ~ T .... don ",cArds"
•. '.11 u .. ,_1t , ........ 11 .......... lI.nU., In 1,.01'" ,. ", ...... Ir .. II • f." •• " " II' "'-, II ••• ' .n ..... , t, ......... U.',_I •• 11 •• ,.
I .... ,",.1'" .... UI ",11 ,,".tU' 1 1110 .......... 11 .... _ .: ••.• ,1 ...
,"" .... I .. I "~ ."" .. FlU .. ""_ .... """ t. II , .... CU .... O. U II."!' hI .... .
T'h.e da~.<;n .... ,11 aan"""'Clldar. 0 l~" ~ of..... • ...... t_ .. I,"I .. _ ... I.IUl •• I"
.,na of t .... carTt,a.-r .... G!IXS.i,on 'to ... ~ • 0' '" -.qrr __ t Th •• _"4' ... 11 tJII ........ a • , ..... 11"' ..... "11'" •• 11.1 ... """"
II.' I ...... 1, .. d. h., II 'tU.I. ,_ ,II, m qu,da '" OI'...:.I ... o'+'nq ....... ~,.GC'ts dlCll'ly"";on .o~ ..or ~'''q apPDn:2'tu'S attac.hd'd +0 ..,. 0:0,.,1".101"_.- • '.,.0' ,I .• II a.H"t '1' n h.
O't~, ... .-- .,~;::. ... fC"'::~~'" ~ CCJ""'~'~~ .ro,-"d..x:I .ne ..".n,m"rn ~qn f~~ ~.,c .... " on .. ,.,.,. '"', ...... o.,.on ora ICIl"IOv.d..:i
tr~u:·ho,., Todor .. Fr.r ca/c""laf,,,,. ~Ff"~cf'..-r: p~ •• f~ ....... 111 b# II. Tallow.
R'9,d ConalJl1"· J,: •• 0002 ",... •. 25 FI •• ,bl. CorJd",,';·k".OOI5 ,.4. ·::S
.1 I. I ••• '" 1,.1 "I.,a! a\ .. ,loUI
I ~"'.'U" , ......... " •••• , " ... 11 II. n "n •••• , •. : ••••• " , ,. Uf' ~H' loT ""'ftl ,.". H,.' " •• ,"." nil .. , •• t.
PloCoa 2',0 'CY"Iq"ud,nal cICSUr"D .,..,-..-,p
011 IaoTrs ~ orr lind E'fff!cf, .. pt'r3Wp;SJ ~ ofttN 0Uu
T--; .... ___ ""_=:" "T_""'","--" __ :::,'::..-::"-::,,-_-"""-='-"~,,,,";7."1 1Jl---'»~4· AZ------/!J/U· lJJ ------'-'110 .... • A.J------t:JtJ."" "" ---- #!J6.0·- M -------1~4· M ------ 2ZiI. A.!J -----/!J44' tJ6---.»r#" "7 ------ 322.4'
""---1#8'" "'----14""· MJ----- IJB,,' &/------ J4J~ ...
AG----/!J/U· ~AIDON ~ AIw '-:3r
£/ ------- /.148' a -------14M' " ---------. (A-------IeB·
.~~nPqdn~, ··Opflonal.? Jt!on:::t:Y15.?.er
51\-' ItIII
TEXAS HIGHWAY O[PARTM[NT BRIOG[ DIVISION
PLAN AND ELEVATION
400 FT. SEGMeNTAL POST TENSIONED UNIT
i NTRACOA5TAL CANAL BRIOOE
:. H I!AR K!y I! LI! VAT I ON
eLEVATION
D I!TAIL 'C' ':.-:In ,Itt:.. • ~JJ(HJlFXN71 hor
~ W I £WU ...... AtA 1JN:d~<!I.nQ#"'" JJ511l1€1D:Ji.llill
lU! 1* ... 1 ..... , lOtH'" , .... U .~"i ... " ..••• ... uh"l~' U J'nU' '.'U .J .nU-C;l U U,n, ••• :, If It ••• u, II! .. ill nu' ••••• '.I"al. tIl •••• "~. r •• ~I"~"1 "I ~"\ •• <~.,, H .... p '5~1I'·. '0' ~ •• ,nn".,I.HU u .,,, .. "" It; '~""'" ... ,~" I",,~. I'M 'Ut 'h ... t ....... ft .. ',: _ ...... 'tt., "If .,.,..,,,1 U ,~, frl"H' " Hop ~UI'I •• '''~I'I , ~;-F=::::t=:"t f •• t ...... ,.
~ t "ttl" ""II Holtl .. ,0,1 .t~'I.l.H.t'1 1, , .... ,
.q::r---------',Cl-:··'-:::4:!il :;:!:: •• ~~~~~::~ .:1 D:1; .:~'!~~::::'~'~'; .:;:::_ ,~;':~, I '.t l"""'Ut •• q, hl",1 II I"." ..... n '''''''1 .'tU'-'''! ~',n' I., fl.!n<l'
• '11 ""-' •• "."1., '.~\I ." ; .. '" ~lU 1. t.t fI'- . ..... ,,' ,tttl .t' J'-l" ~n\." I. Ulr .. I"at .. _._ . • I,~" .. " •••••• I""~jf 4.""'1 ,l.u •• ~\ "I ••.. ' .... ' •• U.,,,I, I!II' If' "",; 'I , .. \ ...... u .,l! .. ~hn.t I, '''''';' : ... 1 •• 'U"i' nn c~(., .. , •• ~". _.n " I ... " .. U1'.ct~ •••
I ." ,Po •• ~ut u ..... It .~'U u elf~~ ., ..... 0"1f· ..... H (f. f._t ··".!,~.f h ... ,D, n~
r • .,.,.iOrHI'lW Sfnr,n,fh .. (;CQOp.a.i. Ha,.,dI1r,9 srrw"t;!fh" JlOOOp .• ,i.
TEXAS H1GHWAY D£PART"E~T
BR1DGE DIVIS10N
TYPICAL PRECAST UNIT 400 FT. SEGMeNTAL
POST TENSIONED UNIT
I NTRACOASTAL CANAL BRIOOE ............ " '>'1 0 ~.:,::,;:.'=' .......... ., ,. if7,7 . ."
DeTAIL A-A ,"""""-J5j ?RECAST VN1T Mit MD
L CH>"
KJfCY
DETAIL a-I ,_ ~ !.!t\fI
...... <;:,,l(:.
" ~~
u.
&v A, • ..
'4 C, C '" F -4 - 0 (j, rr -" Gr 3B '6
;:;Nf'fforo C::Jnt:t'{n·e·~Ja~ 1)
~ .. ~~.~ ....
TEXAS HtGHWAY DEPARTMENT BRIDGE DIVISION
PRECAST UNITS CS-PS-SIQ
400 FT. SEGM!NTAL POST TENSIONED UNIT
INTRACOASTAL CANAL BRIDGE
eL.EYA TION
SECTION """A SECTION IS-IS SECTION c-c SECTION D-D
I
1-'
SECTION e-e
5h(;!l(?idOA'l':.-___________ -I TE XAS HIGHWAY OEPARTM£NT
8RIOGE OIVISION
TENDON PROFILES CENTER SPAN
.00 PT. SEGMENTAL. POST TENSIONED UNIT
I NTRACOA;,TAL CANAL BRIDGE .-...:: ._ ... _-
I!LI!YATION
----;- I ; .. , I ~ -., . "I "I ~' j i
SECTION ~
TEXAS HIGHWAY DEPARTMENT BRIDGE OIVISION
TENDON PROFILES SIDE SPANS
400 FT. SEGMENTAL POST TENSIONED UNIT
INTRACOA5TAL CANAL BRIDGE
N +'" 00
TENDON PROPILE5 TENDON$ ... I\ AI THRUM - CI &C2
TENDON PROFILES TENOONs-IttII?\ c.J - C"'DII
TENDON ANCHORAGE
itS&~~:~~@A64 Cill'4ryC!
TENDON ANCHORAGE BEARING PLATE TYPE ANQ10R TYPICAl pEtA.IL
TENDON PflOFIL!!S T'ENtX:"'" My., 81 I. BIZ
TENOOH ANCHORAGE
~N~TY~ ~.A!2
I a9y a:JI' I
TENDON PROFILes TE'NOONS Mt\. &3 1MRV alO
.5~"6ofl
TEXAS HIGHWAy DEPART"ENT BRIDGE DIVISION
TENPON PROFILES AND DETAILS
400 FT. SEGMENTAL POST TeNSIONED UNIT
I NTRACOAST,/\L CANAL BRIDGE
8LOCKINCi DIACiRAM Dt.oc1Iaad ~ pc8f·~~(IJf'/nphar
"-- ·tJ,I',)"bp t:ar bor-A
SECTION THRU LON~ITUOIN"L JOINT
EL£V l"R.4.NSVEASE JOINT
... ~ C ¥ of 0 J.I CU"'OC Dna U7·~' ..... ,"''''c:wT'"'''9 lOr frc,!toHPr_.,o,n'Ot1P1't::tt>., !lIIC/lI'f7rdll"tb-d prlr:~ 'or ~~ntol Ptw.t,.. ••• d unn"
---oj 4 8
8aL L OF REI,*" .. EST. C!U!\Nt. CLOSURE Sll'IP 'fL~/tud!nQO
"0 Slle II I IJIJi ." /'.0· 3156
-, _ ~~_ ~~ i'!J6I
LD 6110 (y 11
TEXAS HIGHWAY DEPARTMENT BRIDGE Oll/ISION
BLOCK lNG DETAILS
400 FJ. SEGMENTAL POST TENSIONED UNIT
INTRACOASTAL CANll.L BRIDGE
N \J1 o
A P PEN D I X B
SPECIFICATION FOR EPOXY BONDING AGENT
!!!!!!!!!!!!!!!!!!!"#$%!&'()!*)&+',)%!'-!$-.)-.$/-'++0!1+'-2!&'()!$-!.#)!/*$($-'+3!
44!5"6!7$1*'*0!8$($.$9'.$/-!")':!
TEXAS HIGHWAY DEPARTMENT
SPECIAL SPECIFICATION
ITEM 2131
EPOXY BONDING AGE~~
1. DESCRIPTION. This item shall govern for the furnishing and application of epoxy material for use in che j?int between precast concrete units, as required by the plans.
2. MATERIALS. The epoxy material shall be of two components, a resin and a hardener (1 to 1 ratio), meeting the following requjrementE:
a. Pot Life Min. 90 minutes at 68 F (ASTM D1338) b. Compressive Strength 6,000 p.s.i. min. c. Tensile Strength (Direct or Bending) 2,000 p.s.i. min. d. Specific Gravity' 70 to 120 lbs./cu.ft. e. Viscosity at 68 F 10,000 to 50,000 cps f. Coefficient of Thermal Expansion Within 10% of that
for concrete
The material shall have a rate of absorption, rate of shrinkage, chemical resistance and weather resistance compatible with concrete and a consistency such that it will not flow appreciably when applied to a vertical concrete surface. The color shall be concrete gray.
The Contractor shall furnish the Engineer a sample of the material for testing, and a certification from a reputable laboratory indicuting that the material complies with the above requir~~ents.
The sample of the material submitted will be tested additionally for the following,
a. Ability to join test specimen under the following conditions:
Temperature Range Surface Conditions (Moist is defined as saturation' .)
50 F to 100 F Dry to Moist
'one hour drying after complete
213) .001) 3-71 .
b. The joint material shall be able to develop 95 percent of the flexural tensile strength and 70 percent of the shear strength of a wDnolithic test specimen.
The test specimen shall be made of concrete having a minimum compressive strength of 6.000 p.s.i. The specimen will be tested with both dry and n.oist surface conditions.
3. CONSTRUCTION METHODS. Surfaces to which the epoxy material is to be applied shall be free from all oil, laitance or any other material that would prevent the material fram bonding to the concrete surface. All laitance shall be removed by sanding or by washing and wire brushing.
Mixing of the resin and hardener components shall be in accordance with the manufacturer's instruction. Use of a proper sized mechanical mixer will be required.
The epoxy material shall be applied to all surfaces to be joined within the first half of the pot life as shown on the containers.
The coating shall be smooth and uniform and shall cover the entire surfaces to be joined with a maximum thickness of 1/16 inch. The u~its shall be joined within 45 minutes after application of the epoxy material.
No jointing operations shall be performed when the ambient temperature is below 50 F or above 100 F. When the temperature is above 85 F the epoxy coated surfaces shall be shaded from direct sunlight."
If the jointing is not completed within 45 minutes after application of the epoxy material the operation shall be stopped and the epoxy material shall be completely removed from the surfaces. Fresh material shall be applied to the surfaces before resuming jointing operations.
4. MEASUREMENT AND PAYMENT. No direct measurement or payment will be made for the materials, work to be done or equipment to be furnished under this item, but it shall be considered subsidiary to the particular items required by the plans and the contract.
N Ln W
!!!!!!!!!!!!!!!!!!!"#$%!&'()!*)&+',)%!'-!$-.)-.$/-'++0!1+'-2!&'()!$-!.#)!/*$($-'+3!
44!5"6!7$1*'*0!8$($.$9'.$/-!")':!
A P PEN D I X C
NOTATION
[255]
!!!!!!!!!!!!!!!!!!!"#$%!&'()!*)&+',)%!'-!$-.)-.$/-'++0!1+'-2!&'()!$-!.#)!/*$($-'+3!
44!5"6!7$1*'*0!8$($.$9'.$/-!")':!
A
A P
A v
a
b
b'
C
c
d
d c
d P
d t1 ,dt2
d t3
DL1
DL3
E c
E s
e
F sp
= average area of cross sections of specimens [Chapter 3]; area of bolt rChapter 6]
= area of prestressing cable
= area of web reinforcement placed perpendicular to the axis of the member
= depth of equivalent rectangular stress block
257
= average width of specimen [Chapter 3]; width of compression face of flexural member [Chapter 8]
minimum width of a flanged member
= ultimate compressive force
= distance from extreme compression fiber to neutral axis
~ average depth of specimen [Chapter 3]; distance from extreme compression fiber to centroid of prestressing force [Chapter 8]
= distance from plastic center to centroid of compressive force
distance from plastic center to centroid of negative tendon
= distance from plastic center to centroid of each positive tendon
= dead load of the cantilevered section
dead load added after completion of closure
= modulus of elasticity of concrete
modulus of elasticity of steel
thickness of error in joint at bottom
accumulated error in height at the distance tE
= distance from centroid of prestressing cable to C.G.
= distance from centroid of each prestressing cable to C.G.
compressive strength of concrete
ratio of splitting tensile strength to the square root of compressive strength
258
F s
t s
f Y
f cp
F
f sp
f pe
f pc
h
h p
I
IL
jd
= ultimate load of steel
: ultimate strength of steel
= yield strength of reinforcement
= permissible compressive concrete stress on bearing area under anchor plate of post-tensioning steel
internal force
= internal force due to prestressing cable
= fahrenheit degree
= splitting tensile strength of concrete
compressive stress of concrete due to prestressing only at bottom fiber
= stress due to dead load at the extreme fiber of a section at which tensile stress is caused by applied load
= compressive stress in concrete, after all prestress losses have occurred, at the centroid of the cross section resisting the applied load
~ depth of segment
= height of pier
moment of inertia [Chapter 6]; impact factor [Chapter 7]
= impact load
distance between ultimate compressive force (C) and ultimate tensile force (T)
span length
= distance from the joint with the erection errors
length of bolt in compression side
= length of bolt in tension side
= length of prestressing cable
LL
LF
M
M P
M cr
M s
M u
~L
259
= live load
-. load £ac tor
moment due to externally applied load
= moment due to prestressing force
= net flexural cracking moment
moment due to end reaction caused by prestressing of positive tendons and seating forces at outer supports
= secondary moment due to prestressing of negative tendons
= secondary moment due to prestressing positive tendons in main span
= secondary moment due to prestressing of positive tendons in side span
ultimate external moment
= moment due to dead load
M(LL+IL)= moment due to (live + impact) load
M = longitudinal slab moment per unit width x
M transverse slab moment per unit width y
MOl
,M02
= plastic moment
~ ::::: external moment
MI internal moment
~l moments due to weigh t of segment (1. 0 DL)
~2 moments due to resultant force of prestressing in main span and jacking force at outer supports (=M )
s
~3 ::::: moments due to 0.35 DL
~4 = moment due to live load
~I = ultimate internal moment
~l moments due to negative tendons
MI2 = moments due to positive tendons in side span
260
M13 = moments due to positive tendons in main span
P = applied load {Chapters 3, 6, 7]; prestressing force [Chapters 6, 8J
P = ultimate applied load u
s
s
T
= prestressing force at a certain point
= prestressing force applied at the end
= prestressing force of each tendon
= rela tive humidity
= standard deviation
= force on the bolt
= resultant reactions due to prestressing of tendons and seating forces at outer supports
longitudinal spacing of web reinforcement
= ultimate tensile force
Tl ,T2,T3= tensile force of each prestressing cable
u
v
v p
V c
V • C1
V cw
Vd
V u
w
W
Wb
= ultimate design load
shear force due to externally applied load
equivalent vertical load of prestressing force
~ shear carried by concrete
= shear at diagonal cracking due to all loads, when such cracking is the result of combined shear and moment
= shear force at diagonal cracking due to all loads, when such cracking is the result of excessive principal tensile stress in the web
= shear due to dead load
shear due to specified ultimate load
= unit weight of concrete
weight of truck
- section modulus at bottom
y
a
(P
~u
Est
°1 °2 °3
~1'~2
92
93
~
261
= section modulus at top
~ load factor of (LL + IL)
= reaction at support at complete collapse of the structure
= angle change of tendon
= diameter rChapter 2]; capacity reduction factor rChapter 8]
= ultimate shear strength
= ultimate flexural strength
stress of prestressing cable
= ultimate steel strain
= concrete strain
= strain of each prestressing cable (positive tendon)
= strain at level of prestressing cable (negative tendon)
steel strain
= steel strain due to prestressing
= steel strain due to external load
cantilever tip deflection to to segment flexure
= cantilever tip deflection due to elongation of bolts
= cantilever tip deflection due to bending of pier
= elongation of bolts
= angle change due to elongation of bolts
= angle change of pier at the top due to bending of pier
= Poisson's ratio [Chapter 4]; friction coefficient [Chapter 6]
= wobble coefficient