' r~)'.vi\ (I

LIBRARY
Michigan State
University

This is to certify that the
thesis entitled

EXPERIMENTAL EVALUATION AND FIELD MONITORING
OF SELF-CONSOLIDATING CONCRETE PRESTRESSED
BOX BEAMS IMPLEMENTED IN A DEMONSTRATION
BRIDGE

presented by

David A. Bendert

has been accepted towards fulfillment
of the requirements for the

MS. degree in Civil Engineering

 

 

0 5/9 g/2007

Date

MSU is an affirmative-action, equal-opportunity employer

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p:/CIRC/DateDue.indd-pr1

 

 

EXPERIMENTAL EVALUATION AND FIELD MONITORING OF SELF-
CONSOLIDATING CONCRETE PRESTRESSED BOX BEAMS
IMPLEMENTED IN A DEMONSTRATION BRIDGE

By

David A. Bendert

A THESIS

Submitted to
Michigan State University
In partial fulfillment of the requirements
For the degree of

MASTER OF SCIENCE

Department of Civil and Environmental Engineering

2007

ABSTRACT
EXPERIMENTAL EVALUATION AND FIELD MONITORING OF SELF-
CONSOLIDATING PRESTRESSED BOX BEAMS USED IN A
DEMONSTRATION BRIDGE
By

David A. Bendert

Self Consolidating Concrete (SCC) is a tailorable concrete with superior fresh
properties that can lead to reduced production times and costs in precast/prestressed
element production. Concerns from Departments of Transportation about the use of SCC
in highway bridges have prevented these potential benefits from being fully realized in
the production of precast/prestressed bridge beams. A demonstration bridge utilizing
SCC was built to allow the Michigan Department of Transportation (MDOT) to evaluate
the use of SCC in bridge beams including the beam production process and short and
long term performance of SCC beams. Experimental evaluations were used to find the
flexural and shear capacities of the SCC beams, while long-term evaluations included
continuous strain and temperature monitoring of the beams in the demonstration bridge.
Reduced production times of 60% and reduced labor requirements of nearly 50% were
realized during the production of the bridge beams. The flexural and shear capacities of
the SCC beams were very similar to the capacities of the NCC beams with all beams
surpassing design requirements. Similar trends have been realized in the continuous
strain readings of the demonstration bridge. While monitoring of the demonstration
bridge continues, the results of the evaluation have shown that SCC can by successfully

utilized in precast prestressed bridge beams for highway bridges.

This thesis is dedicated to my wife, Allison.
For her unending dedication, unbridled love,
and willingness to stand by me as I pursued my dream.

iii

 

ACKNOWLEDGEMENTS

I first must thank Dr. Rigoberto Burguefio. Without his guidance and teaching
this work would not have been possible. His eagerness to help students learn both in the
classroom and the laboratory has made me a better engineer. I also would like to thank
the members of my thesis defense committee. The teachings and comments provided by
Dr. Parviz Soroushian and Dr. Venkatesh Kodur greatly enhanced my work. I am
grateful to Mr. Siavosh Ravanbakhsh whose guidance and encouragement in and out of
the laboratory helped me complete this research. Also, without the technical support of
Mr. James “JC” Brenton this work would not have been completed. The students of the
MSU Civil Infrastructure lab need special thanks. Mahmood Haq has been a great friend,
office mate, and role model throughout the duration of my work as a graduate student.
Steve Francoviac, Andy Pauly, Ben Pavlich, and Emily Bruce all contributed greatly to
this work. I must also thank my parents Allan and Diana Bendert. They instilled in me a
work ethic that has allowed me to complete this task. Without their love and support this
would not have been possible. I must also thank Mr. Kerry Irons for his help in
reviewing sections of this document. His advice and comments certainly improved this
work. I must also thank my wife, Allison. Her dedication to me through this entire
process not only gave me the strength to finish but also lifted my spirits when times got
tough. Her beautiful smile and constant encouragement made the long nights and early
mornings worth it. Finally, The research described in this thesis was carried out under
funding from the Michigan Department of Transportation through an FHWA-IBRC
project, which is gratefully acknowledged. The assistance advice from the Premarc

Corporation of Grand Rapids, MI, and of Degussa Admixtures, is also appreciated.

iv

TABLE OF CONTENTS

 

 

 

 

 

 

LISTS OF TABLES VIII
LIST OF FIGURES XV
1 PROJECT OVERVIEW AND INTRODUCTION 1
1.1 SUMMARY ............................................................................................................ 1
1.2 RESEARCH OBJECTIVES ....................................................................................... 4
1.3 SCOPE AND ORGANIZATION ................................................................................. 5

2 SELF-COMPACTING CONCRETE TECHNOLOGY 6
2.1 HISTORICAL OVERVIEW ....................................................................................... 6
2.2 THEORETICAL CONSIDERATIONS ......................................................................... 8
2.2.1 High Range Water Reducers (HRWRs) ................................................... 13
2.2.2 Viscosity Modifying Admixtures (VMAs) ............................................... 14
2.2.3 Mineral Admixtures .................................................................................. 14

2.3 STATUS AND RECENT DEVELOPMENTS IN SCC .................................................. 15
2.3.1 Recent SCC Research ............................................................................... 16
2.3.2 Performance of Beams Made With SCC .................................................. 17
2.3.3 Prestressing Strand Concrete Bond ........................................................... 17
2.3.4 Prestressed Beams Made With SCC ......................................................... 18
2.3.5 Standardization of SCC Fresh Property Testing ....................................... 19

2.4 SCC FOR PRECAST CONSTRUCTION ................................................................... 20
2.5 OUTSTANDING ISSUES ........................................................................................ 22

3 MIX DESIGN DEVELOPNIENT AND PERFORMANCE 24
3.1 OVERVIEW ......................................................................................................... 24
3.2 SCC MIX DESIGN DEVELOPMENT AND EVALUATION ........................................ 24
3.2.1 SCC Mix Design Development Methods ................................................. 24
3.2.2 SCC Mix Design Evaluations ................................................................... 26
3.2.3 SCC Mix Design Development ................................................................ 35
3.2.4 Discussion ................................................................................................. 42

3.3 STRAND BOND EVALUATION ............................................................................. 43
3.3.1 Large Block Pull-Out Test (LBPI‘) Overview .......................................... 45
3.3.2 Observations and Results .......................................................................... 52
3.3.3 Discussion ................................................................................................. 55

3 .4 MIX DEVELOPMENT AND PERFORMANCE CONCLUSIONS ................................... 56

4 PRODUCTION OF SCC PRESTRESSED BOX BEAMS 58
4.1 TECHNICAL CONSIDERATIONS ........................................................................... 58
4.1.1 Specimen Mock-up ................................................................................... 66
4.1.2 NCC Observations .................................................................................... 68
4.1.3 SCC Observations ..................................................................................... 69

 

4.1.4 Discussion ................................................................................................. 74
4.2 BEAM PRODUCTION ........................................................................................... 83
4.2.1 Overview ................................................................................................... 83
4.2.2 Beam Geometry ........................................................................................ 85
4.2.3 Casting Process ......................................................................................... 87
4.2.4 First Cast (6/21/05) ................................................................................... 89
4.2.5 Second Cast (6/24/05) ............................................................................... 94
4.2.6 Third Cast (6/29/05) .................................................................................. 98
4.2.7 Fourth Cast (7/1/05) ................................................................................ 102
4.2.8 Beam Production Discussion .................................................................. 103
4.3 MATERIAL PROPERTIES .................................................................................... 105
4.3.1 Compressive Test .................................................................................... 106
4.3.2 Tensile Strength ...................................................................................... 108
4.3.3 Prestressing Strand Properties ................................................................. 109
4.4 PRODUCTION CONCLUSIONS ............................................................................ 109

5 FLEXURAL AND SHEAR PERFORMANCE OF SCC AND NCC

 

 

PRESTRESSED BOX BEAMS 111
5.1 TESTING PROGRAM OVERVIEW ........................................................................ 1 1 1
5.2 THEORETICAL CONSIDERATIONS ..................................................................... 1 12

5.2.1 Flexural Capacity .................................................................................... l 13
5.2.2 Shear Capacity ........................................................................................ 118
5.2.3 Shear-Moment Interaction ...................................................................... 126
5.3 FLEXURAL EXPERIMENTAL EVALUATION ........................................................ 127
5.3.1 Test Setup, Instrumentation and Protocol ............................................... 127
5.3.2 Observations and Results ........................................................................ 130
5.4 SHEAR EXPERIMENTAL EVALUATION .............................................................. 140
5.4.1 Test Setup, Instrumentation and Protocol ............................................... 140
5.4.2 Observations and Results ........................................................................ 145
5.5 EXPERIMENTAL EVALUATION CONCLUSIONS .................................................. 169

6 LONG TERM PERFORMANCE OF SCC BRIDGE BEAMS 172
6.1 OVERVIEW ....................................................................................................... 172
6.2 FIELD MONITORING SYSTEM ........................................................................... 17 3

6.2.1 Instrumentation ....................................................................................... 174
6.2.2 Data Collection System ........................................................................... 183
6.2.3 Observations and Results ........................................................................ 187
6.2.4 Discussion ............................................................................................... 195
6.3 PRESTRESS LOSSES .......................................................................................... 197
6.3.1 Theoretical Considerations ......................... . ...... 199
6.3.2 Observations and Results ........................................................................ 212
6.3.3 Comparison to Measured Data ................................................................ 227
6.4 INFLUENCE OF DIAPHRAGM ON STRAIN IN BEAM ............................................. 234
6.5 EFFECT OF TEMPERATURE ON STRAIN IN BEAM ............................................... 240
6.6 INFLUENCE OF BRIDGE SYSTEM ON STRAIN IN BEAM ...................................... 248
6.7 DATA NORMALIZATION ................................................................................... 251

vi

6.8 LONG TERM PERFORMANCE CONCLUSIONS ..................................................... 253

7 CONCLUSIONS AND RECOMMENDATIONS

APPENDICES

 

APPENDIX A..

APPENDIX B
APPENDIX C
APPENDIX D
APPENDIX E
APPENDIX F
APPENDIX G
APPENDIX H

REFERENCES

ooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooooo

NCC PRODUCTION CAST 1 (6-21-05) ..............................................
SCC 1 PRODUCTION CAST 1 (6-21-05) ............................................
NCC PRODUCTION CAST 2 (6-24-05) ..............................................
SCC 2 PRODUCTION CAST 2 (6-24-05) ............................................
NCC PRODUCTION CAST 3 (6-29-05) ..............................................
SCC 3 PRODUCTION CAST 3 (6-29-05) ............................................
NCC PRODUCTION CAST 4 (7 -1-05) ................................................

 

Vii

255

262
263
277
281
286
290
295
299
304

308

Table 1.

Table 7.

Table 10.
Table 13.
Table 14.
Table 15.
Table 16.
Table 17.
Table 18.
Table 19.
Table 20.
Table 21.
Table 22.
Table 23.
Table 24.
Table 25.
Table 26.
Table 27.
Table 28.
Table 29.
Table 30.

Table 31.

LISTS OF TABLES

Mix Design Experimental Matrix ................................................................. 25
SCC 2 Results ............................................................................................... 40
Project Mix Designs ...................................................................................... 42
Ultimate Loads from LBPT .......................................................................... 53
First Slip Loads for LBPT ............................................................................. 55
Mock-Up Fresh Property Test Results .......................................................... 68
Strand Draw—in Measurements ..................................................................... 83
Untransformed Section Properties ................................................................ 86
Average NCC Mix Design ............................................................................ 90
NCC Beam Casting Time ............................................................................. 90
Average SCC 1 Mix Design ......................................................................... 92
SCC 1 Fresh Properties ................................................................................. 92
SCC l Slump Flow Values ........................................................................... 93
SSCC 1 Beam Casting Times ....................................................................... 94
Average NCC Mix Design ............................................................................ 95
NCC Beam Casting Times ............................................................................ 96
Average SCC 2 Mix Design ......................................................................... 97
SCC 2 Fresh Properties ................................................................................. 97
SCC 2 Slump Flow Values ........................................................................... 98
SCC 2 Beam Casting Times ......................................................................... 98
Average NCC Mix Design ............................................................................ 99
Average SCC 3 Mix Design ....................................................................... 101

viii

Table 32.

Table 33.

Table 34.

Table 35.

Table 36.

Table 37.

Table 38.

Table 39.

Table 40.

Table 41.

Table 42.

Table 43.

Table 44.

Table 45.

Table 46.

Table 47.

Table 48.

Table 49.

Table 50.

Table 51.

Table 52.

Table 53.

Table 54.

SCC 3 Fresh Properties ............................................................................... lOl
SCC 3 Slump Flow Values ......................................................................... 102
Average NCC Mix Design .......................................................................... 103
Production Time Savings Using SCC ......................................................... 104
NCC Compressive Strength ........................................................................ 107
SCC 1 Compressive Strength ...................................................................... 107
SCC 2 Compressive Strength ...................................................................... 107
SCC 3 Compressive Strength ...................................................................... 108
Split Tensile Test Results ............................................................................ 109
Prestressing Steel Properties ....................................................................... 109
Values of 0 and B for Sections With Transverse Reinforcement [3] .......... 125
Flexure Test Concrete Compressive Strengths ........................................... 128
Maximum Achieved Capacities of Flexural Beams ................................... 140
Shear Test Concrete Compressive Strengths .............................................. 141

Maximum Achieved Capacities of Shear Beams at Critical Shear Section 167

NCC VWSG Initial Readings ..................................................................... 179
SCC l VWSG Initial Readings ................................................................... 179
SCC 2 VWSG Initial Readings ................................................................... 179
SCC 3 VWSG Initial Readings ................................................................... 180
Preliminary Strain Measurement Events .................................................... 188
Maximum, Minimum, and Average Temperatures At Top of Deck .......... 193

Maximum, Minimum, and Average Temperatures At Bottom of Beam 193

NCC Transformed Section Properties ........................................................ 213

ix

 

Table 55.

Table 56.

Table 57.

Table 58.

Table 59.

Table 60.

Table 61.

Table 71.

Table 72.

Table 73.

Table 74.

Table 75.

Table 76.

Table A- 1.

Table A-2.

Table A-3.

Table 77.

Table 78.

Table 79.

Table 80.

Table 81.

Table 82.

Table 83.

SCC 1 Transformed Section Properties ...................................................... 214
SCC 2 Transformed Section Properties ...................................................... 214
SCC 3 Transformed Section Properties ...................................................... 214
NCC Transformed Strand and Instrument Eccentricities ........................... 216
SCC 1 Transformed Strand and Instrument Eccentricities ......................... 216
SCC 2 Transformed Strand and Instrument Eccentricities ......................... 216
SCC 3 Transformed Strand and Instrument Eccentricities ......................... 217
NCC Section Properties Without Void ....................................................... 234
NCC Strand and Instrument Eccentricities Without Void .......................... 235
SCC 1 Section Properties Without Void ..................................................... 235
SCC 1 Strand and Instrument Eccentricities Without Void ........................ 235
Recorded Temperatures for NCC Beam (6/21/06) ..................................... 244
Section Properties for Gradient Temperature Analysis .............................. 245
Conceptual Mix Design Matrix .................................................................. 268
Sample 7000 psi Mix Designs from MSU Research .................................. 269
Minimum tests to be used for evaluating SCC performance (see Ref [1]).. 270
NCC Mix Design Batch 1 ........................................................................... 277
NCC Mix Design Batch 2 ........................................................................... 277
NCC Mix Design Batch 3 ........................................................................... 278
NCC Mix Design Batch 4 ........................................................................... 278
NCC Mix Design Batch 5 ........................................................................... 278
NCC Mix Design Batch 6 ........................................................................... 279
NCC Mix Design Batch 7 ........................................................................... 279

 

Table 84.

Table 85.

Table 86.

Table 87.

Table 88.

Table 89.

Table 90.

Table 91.

Table 92.

Table 93.

Table 94.

Table 95.

Table 96.

Table 97.

Table 98.

Table 99.

Table 100.

Table 101.

Table 102.

Table 103.

Table 104.

Table 105.

Table 106.

NCC Mix Design Batch 8 ........................................................................... 279
NCC Mix Design Batch 9 ........................................................................... 280
SCC 1 Mix Design Batch 1 ......................................................................... 281
SCC 1 Mix Design Batch 2 ......................................................................... 281
SCC 1 Mix Design Batch 3 ......................................................................... 282
SCC 1 Mix Design Batch 4 ......................................................................... 282
SCC 1 Mix Design Batch 5 ......................................................................... 282
SCC 1 Mix Design Batch 6 ......................................................................... 283
SCC 1 Mix Design Batch 7 ......................................................................... 283
SCC 1 Mix Design Batch 8 ......................................................................... 283
SCC 1 Mix Design Batch 9 ......................................................................... 284
SCC 1 Mix Design Batch 10 ....................................................................... 284
SCC 1 Mix Design Batch 11 ....................................................................... 284
SCC 1 Mix Design Batch 12 ....................................................................... 285
SCC 1 Mix Design Batch 13 ....................................................................... 285
SCC 1 Mix Design Batch 14 ....................................................................... 285
NCC Mix Design Batch 1 ....................................................................... 286
NCC Mix Design Batch 2 ....................................................................... 286
NCC Mix Design Batch 3 ....................................................................... 287
NCC Mix Design Batch 4 ....................................................................... 287
NCC Mix Design Batch 5 ....................................................................... 287
NCC Mix Design Batch 6 ....................................................................... 288
NCC Mix Design Batch 7 ....................................................................... 288

xi

Table 107.

Table 108.

Table 109.

Table 110.

Table 111.

Table 112.

Table 113.

Table 114.

Table 115.

Table 116.

Table 117.

Table 1 18.

Table 119.

Table 120.

Table 121.

Table 122.

Table 123.

Table 124.

Table 125.

Table 126.

Table 127.

Table 128.

Table 129.

NCC Mix Design Batch 8 ....................................................................... 288
NCC Mix Design Batch 9 ....................................................................... 289
SCC 2 Mix Design Batch 1 ..................................................................... 290
SCC 2 Mix Design Batch 2 ..................................................................... 290
SCC 2 Mix Design Batch 3 ..................................................................... 291
SCC 2 Mix Design Batch 4 ..................................................................... 291
SCC 2 Mix Design Batch 5 ..................................................................... 291
SCC 2 Mix Design Batch 6 ..................................................................... 292
SCC 2 Mix Design Batch 7 ..................................................................... 292
SCC 2 Mix Design Batch 8 ..................................................................... 292
SCC 2 Mix Design Batch 9 ..................................................................... 293
SCC 2 Mix Design Batch 10 ................................................................... 293
SCC 2 Mix Design Batch 11 ................................................................... 293
SCC 2 Mix Design Batch 12 ................................................................... 294
NCC Mix Design Batch 1 ....................................................................... 295
NCC Mix Design Batch 2 ....................................................................... 295
NCC Mix Design Batch 3 ....................................................................... 296
NCC Mix Design Batch 4 ....................................................................... 296
NCC Mix Design Batch 5 ....................................................................... 296
NCC Mix Design Batch 6 ....................................................................... 297
NCC Mix Design Batch 7 ....................................................................... 297
NCC Mix Design Batch 8 ....................................................................... 297
NCC Mix Design Batch 9 ....................................................................... 298

xii

Table 130.

Table 131.

Table 132.

Table 133.

Table 134.

Table 135.

Table 136.

Table 137.

Table 138.

Table 139.

Table 140.

Table 141.

Table 142.

Table 143.

Table 144.

Table 145.

Table 146.

Table 147.

Table 148.

Table 149.

Table 150.

Table 151.

Table 152.

SCC 3 Mix Design Batch 1 ..................................................................... 299

SCC 3 Mix Design Batch 2 ..................................................................... 299
SCC 3 Mix Design Batch 3 ..................................................................... 300
SCC 3 Mix Design Batch 4 ..................................................................... 300
SCC 3 Mix Design Batch 5 ..................................................................... 300
SCC 3 Mix Design Batch 6 ..................................................................... 301
SCC 3 Mix Design Batch 7 ..................................................................... 301
SCC 3 Mix Design Batch 8 ..................................................................... 301
SCC 3 Mix Design Batch 9 ..................................................................... 302
SCC 3 Mix Design Batch 10 ................................................................... 302
SCC 3 Mix Design Batch 11 ................................................................... 302
SCC 3 Mix Design Batch 12 ................................................................... 303
SCC 3 Mix Design Batch 13 ................................................................... 303
SCC 3 Mix Design Batch 14 ................................................................... 303
NCC Mix Design Batch 1 ....................................................................... 304
NCC Mix Design Batch 2 ....................................................................... 304
NCC Mix Design Batch 3 ....................................................................... 305
NCC Mix Design Batch 4 ....................................................................... 305
NCC Mix Design Batch 5 ....................................................................... 305
NCC Mix Design Batch 6 ....................................................................... 306
NCC Mix Design Batch 7 ....................................................................... 306
NCC Mix Design Batch 8 ....................................................................... 306
NCC Mix Design Batch 9 ....................................................................... 307

xiii

Table 153. NCC Mix Design Batch 10 ..................................................................... 307

Table 154. NCC Mix Design Batch 11 ..................................................................... 307

xiv

LIST OF FIGURES

Figure 1. M 50/US 127 Demonstration Bridge using SCC Beams .................................... 2
Figure 2. Schematic of Demonstration Bridge ................................................................. 3
Figure 3. Fluidity, Stability Tradeoff in SCC Mix Designs ............................................ 11
Figure 4. Inverted Slump Flow Test ............................................................................... 27
Figure 5. Concrete Patty from Inverted Slump Flow Test ............................................... 28
Figure 6. J-Ring Test Set Up ............................................................................................ 31
Figure 7. Result of J-Ring Test ........................................................................................ 32
Figure 8. Schematic of J—Ring Test .................................................................................. 32
Figure 9. J-Ring Test Apparatus Profile ......................................................................... 33
Figure 10. L-Box Test Apparatus .................................................................................... 34
Figure 11. L—Box Test ..................................................................................................... 34
Figure 12. Pull-out Test Specimen Prior to Testing ......................................................... 50
Figure 13. Schematic of Pull-Out Test Specimen ............................................................ 51
Figure 14. Test Set-up For Pull Out Test .......................................................................... 51
Figure 15. Ultimate Load of LBPT for all Mix Designs .................................................. 54
Figure 16. 1St Slip Loads for LBPT. ................................................................................ 55
Figure 17. NCC Casting Bed Prior to Casting ................................................................. 60
Figure 18. Casting NCC Beam ....................................................................................... 61
Figure 19. The Bottom Flange of The NCC Beam .......................................................... 61
Figure 20. Installing Top Reinforcement On NCC ......................................................... 62
Figure 21. Casting Top Flange and Side Walls In NCC Beam ...................................... 62
Figure 22. SCC Beam Ready to Be Cast .......................................................................... 63

XV

Figure 23.
Figure 24.
Figure 25.
Figure 26.
Figure 27.
Figure 28.
Figure 29.
Figure 30.
Figure 31.
Figure 32.
Figure 33.
Figure 34.
Figure 35.
Figure 36.
Figure 37.
Figure 38.
Figure 39.
Figure 40.
Figure 41.
Figure 42.
Figure 43.
Figure 44.

Figure 45.

Rising Concrete in SCC Beam ...................................................................... 64
Casting SCC Beam ........................................................................................ 65
Mock-Up Beam Plan ...................................................................................... 67
Mock-up Cross Section ................................................................................. 68
Finished SCC 1 Mock-Up Beam .................................................................... 7O
SCC Flowing to End of Formwork ............................................................... 71
Finished SCC 2 Beam With Exposed Reinforcing Steel .............................. 72
Result of Void Movement in SCC 2 Beam .................................................... 72
Finished SCC 3 Mock-Up Beam .................................................................... 73
Void Restraints Spaced Along Beam Length ................................................. 76
Void Restraint Bolted to Formwork ............................................................... 76
Close up of Void Restraint ............................................................................. 77
Removing Void Restraint From Finished Beam ............................................ 77
Scar On Surface of Beam from Void Restraint .............................................. 78
Finished SCC 2 Beam Cast With Additional Void Restraints ...................... 78
SCC 1 Surface Quality ................................................................................... 79
SCC 2 Surface Quality ................................................................................... 80
NCC Surface Quality ...................................................................................... 80
Casting History .............................................................................................. 85
Beam Plan ....................................................................................................... 87
Typical Beam Cross Section ......................................................................... 87
Concrete Compressive Strength ................................................................... 108
Moment Resistance Variable Definitions .................................................... 115

xvi

Figure 46.
Figure 47.
Figure 48.
Figure 49.
Figure 50.
Figure 51.
Figure 52.
Figure 53.
Figure 54.
Figure 55.
Figure 56.
Figure 57.
Figure 58.
Figure 59.
Figure 60.
Figure 61.
Figure 62.
Figure 63.
Figure 64.
Figure 65.
Figure 66.
Figure 67.

Figure 68.

Flexural Test Setup — Overall Features and Dimensions .............................. 128
Overall View of Flexural Test Setup ............................................................. 129

Plan View of Strain Gage Locations ........................................................... 131

Strain Gage Location Sections A-A and C-C .............................................. 132
Strain Gage Location Section B-B ............................................................... 132
SCCl Flexure Test Beam near Ultimate Capacity ........................................ 133
Failure Mode of SCC] Flexure Beam ........................................................... 134
Applied Load vs. Displacement Response of All Flexure Beams ................. 137
Applied Moment vs. Curvature Response of All Flexure Beams ................. 137
Total Load vs. Displacement Response of All Flexure Test Beams ............. 138
Total Moment vs. Curvature Response of A11 Flexure Test Beams .............. 138
Normalized Total Load-Displacement Response of Flexure Beams ............ 139
Normalized Total Moment-Curvature Response of Flexure Beams ............. 139
Overall View of Shear Test Setup ................................................................. 142
Shear Test Setup for NCC Beam —- Overall Features and Dimensions ........ 143

Shear Test Setup for SCC Beams — Overall Features and Dimensions ...... 143
View of Shear Deformation Panels ............................................................... 145

Strain Gage Instrumentation Location for Shear Test Beams ..................... 147

Strain Gage Locations Sections A-A, B-B, C-C ......................................... 147

Strain Gage Locations Section D-D and E-E .............................................. 148
Response of SCC2 Shear Test Beam near Maximum Response ................... 149
Typical Distress in Shear Span and Critical Shear Region ........................... 149
Flexure-Shear Failure of SCC3 Shear Test Unit ........................................... 150

xvii

Figure 69
Figure 70

Figure 71

Figure 72.

Figure 73
Figure 74

Figure 75

Figure 76.

Figure 77

Figure 78

Figure 79.
Figure 80.
Figure 81.
Figure 82.
Figure 83.
Figure 84.
Figure 85.
Figure 86.
Figure 87.
Figure 88.
Figure 89.
Figure 90.

Figure 91.

. Total Shear Force vs. Center Displacement Response .................................. 152
. Total Moment vs. Curvature Response at Critical Shear Section ................. 152
Moment-Shear Interaction Diagram with AASHTO-LRFD ........................ 153
Shear Deformation Panel Theory ................................................................ 154
Shear Deformation Modes ............................................................................ 155

. Normalized Shear Force vs. Displacement at Critical Shear Section ........... 157
. Normalized Moment vs. Displacement Response at Critical Shear Section. 158
Total Shear Force vs. Shear Strain for NCC Beam ..................................... 158

. Total Shear Force vs. Shear Strain for SCCI Beam ...................................... 159
. Total Shear Force vs. Shear Strain at Critical Section for All Beams .......... 159
Instrumented Shear Stirrup Location ........................................................... 161
Strain Gage Location on Shear Stirrup ........................................................ 161
Strain Profiles on Shear Stirrups at Section B for NCC Beam .................... 164
Strain Profiles on Shear Stirrups at Section A for SCC 1 Beam ................. 165
Strain Profiles on Shear Stirrups at Section B for SCC 2 Beam ................. 166
Strain Profiles on Shear Stirrups at Section B for SCC 3 Beam ................. 166
Field Beam Instrumentation ......................................................................... 175
N aming Convention for VWSG ................................................................... 175
Vibrating Wire Strain Gages Installed in The Top Flange ........................... 176
Schematic of Vibrating Wire Strain Gage .................................................... 177
Measurement End of Thermocouple ............................................................ 180
Installed Thermocouple ............................................................................... 182
Thermocouple Locations ............................................................................. 182

xviii

Figure 92. Full Bridge Picture Showing Solar Panel ..................................................... 184
Figure 93. Wire Routing Scheme ................................................................................... 187
Figure 94. Strains In The Bottom Flange of Section A ................................................. 189
Figure 95. Strains at Top of Beam Flange Section A .................................................... 190
Figure 96. Temperature at Top of Bridge Deck at Section A ....................................... 192
Figure 97. Temperatures in the Bottom Flange at Section A ....................................... 192
Figure 98. Strains in the Top Flange of Section B .......................................................... 193
Figure 99. Strains in the Bottom Flange of Section B ................................................... 194
Figure 100. Strains in the Top Flange of Section C ....................................................... 194
Figure 101. Strains in the Bottom Flange of Section C ................................................. 195
Figure 102. Prestress Strand Geometry ........................................................................ 216
Figure 103. Total Prestress Loss for Different Calculation Methods ............................ 220
Figure 104. Prestress Loss According to AASHTO-LRFD-3 Refined Method [4] ...... 221
Figure 105. Prestress Loss According to Time Step Analysis [34] ............................... 222
Figure 106. Prestress Loss According to AASHTO-LRFD-2 Refined Method [3] ...... 223
Figure 107. Strain Comparison For Bottom VWSG NCC Beam Section A ................. 229
Figure 108. Strain Comparison for Top VWSG NCC Beam Section A ....................... 230
Figure 109. Strain Comparison for Bottom VWSG SCC 1 Beam Section A ................ 231
Figure 110. Strain Comparison for Top VWSG SCC 1 Beam Section A ..................... 231
Figure 111. Strain Comparison for Bottom VWSG SCC 2 Beam Section A ................ 232
Figure 112. Comparison of Strains for Top VWSG SCC 2 Section A ......................... 232
Figure 113. Comparison of Strains for Bottom VWSG SCC 3 Section A ................... 233
Figure 114. Comparison of Strains for Top VWSG SCC 3 Section A ......................... 233

xix

Figure 115.
Figure 116.
Figure 117.
Figure 118.
Figure 119.
Figure 120.
Figure 121.
Figure 122.
Figure 123.
Figure 124.
Figure 125.
Figure 126.

Figure 127.

Figure 128

Figure 129

Figure 130.
Figure 131.

Figure 132.

Bottom Strain Comparison With Diaphragm NCC Section A ................. 236
Top Strain Comparison With Diaphragm NCC Section A ...................... 236
Bottom Strain Comparison With Diaphragm SCC 1 Section A ................ 237
Top Strain Comparison With Diaphragm SCC 1 Section A ..................... 237
Bottom Strain Comparison NCC Section B .............................................. 238
Top Strain Comparison NCC Section B .................................................... 239
Bottom Strain Comparison SCC 1 Section B ............................................ 239
Top Strain Comparison SCC 1 Section B ................................................. 240
Thermocouple Locations and Temperature Variation ............................... 241
Fixed and Expansion Ends of Bridge Beams ............................................ 242
Expansion Due to a Uniform Positive Temperature Change .................... 242
Deformation Caused by Temperature Gradient .................... ' .................... 243
Strain Diagram for Temperature Analysis ................................................ 247
. Assumed Beam System With Bearing Stiffness k ..................................... 248
. Idealized System With Infinite Bearing Stiffness ...................................... 249
Bending Moment Diagram For Idealized System ..................................... 249
Normalized Strain Response for Bottom Flange Section A ...................... 251
Normalized Strain Response for Top Flange at Section A ........................ 252

XX

1 PROJECT OVERVIEW AND INTRODUCTION

1.1 Summary

Self-consolidating concrete (SCC) is a recent development in the concrete
industry that Offers many benefits to both producers and consumers Of concrete
structures, including government Departments of Transportation. The benefits of SCC
are almost entirely related to its fresh properties. Its ability to flow through dense Steel
reinforcing schemes and fill intricate formwork without the aid of vibration make SCC a
popular alternative to normally consolidated concrete (NCC). Properly proportioned
SCC can consolidate under its own weight while remaining homogenous through all
phases of the construction process. When SCC is used effectively, the end result is a high
quality product that performs as well as its NCC counterpart. However, the mix design
modifications that are required to produce the benefits of SCC have led to questions
about the hardened performance of the finished products. Because of this, the Michigan
Department of Transportation (MDOT) was interested in assessing the feasibility of using
SCC in highway bridges. The research project reported in this thesis was developed to
investigate the complete construction process using SCC, including the development of
representative SCC mix designs, evaluation of the bond of the prestressing strand with
SCC, evaluation of the production process, and finally the production and construction of
a demonstration bridge using beams cast with SCC.

The M-SO/US-127 bridge over the Grand River in Summit Township south of
Jackson Michigan was selected as the right candidate for its replacement with a

demonstration bridge utilizing SCC in the bridge beams. The demonstration bridge,

shown in Figure 1, consists of six simply supported 27 in. x 36 in. spread box beams with

a span of 50 ft. A schematic of the bridge is shown in Figure 2.

 

Figure 1. M 50/US 127 Demonstration Bridge using SCC Beams

F———.‘ Effective Span = 50' ————l

 

 

 

 

 

 

 

 

\
\ NCC Uninstrumented ————{-

\ NCC Uninstrumented
Bridge Deck
Width = 49'

NCC Instrumented

/ 5 @ 8'-‘/2"
1 SCC l Instrumented

 

(
\

SCC 2 Instrumented

Q3 Instrumented —L

P—— Beam Length = 52' ————i

Figure 2. Schematic of Demonstration Bridge

 

 

 

 

 

 

Three SCC mix designs, which attempt to bound current SCC mix design
practice, were developed for this project. A fourth mix design, an NCC mix, was used as
an experimental control. Three identical box beams were cast from each of the four mix
designs. Of the three beams from each mix design, one was instrumented for long term
monitoring of strain and temperature and was placed in the demonstration bridge. The
remaining two beams from each mix design were instrumented for experimental
evaluation. The experimental evaluation of the full-scale prestressed beams was done at
Michigan State University’s (MSU) Civil Infrastructure Laboratory prior to the
construction of the demonstration bridge. Two four-point bending tests were conducted
for these evaluations. The two tests evaluated the flexural and shear performance of the
beams. An additional five NCC beams were cast without any instrumentation. Two of
these beams were used in the construction of the demonstration bridge. The remaining

three NCC beams were cast to replace the SCC beams in the event that their laboratory

performance was inadequate. A total of 17 beams, 8 NCC and 9 SCC, were cast for this
project at the Fabricator’s precast plant in Michigan in the months of June and July of

2005.

1.2 Research Objectives

There were three objectives to this research project. Each objective was designed
to help MDOT determine the feasibility of using SCC in highway bridges. The
objectives were:

' To evaluate the short term performance of SCC beams,
' To evaluate the long term performance of SCC beams, and

I TO evaluate the production process and quality control for the

manufacturing of SCC beams.

The Short-term evaluation Of SCC beams through full-scale tests of replica bridge beams
was done to ensure that the capacity demands in the bridge system were safely met by
these members. Both the shear and flexural capacities were experimentally determined
through structural testing to evaluate the short-term performance Of the SCC beams with
respect to the baseline NCC beams. The long-term evaluation of SCC beams includes
monitoring strain and temperature from the demonstration bridge. The collected data is
providing important information about any prestressing force losses in the beams that
could be attributed to different creep and shrinkage behavior of SCC. Finally the
production process evaluation included assessment of the mix design development, and

quality monitoring of the concrete production and finished beams. Also included in this

research effort was an evaluation of the production process of SCC prestressed beams to

define any necessary changes in order to better use SCC.

1.3 Scope and Organization

This thesis presents the research work related to the implementation Of SCC in
bridge beams of a demonstration bridge for MDOT through a Federal Highway
Administration (FHWA) Innovative Bridge Research Construction (IBRC) project. The
thesis is divided into 7 chapters. Chapter 2 presents research on SCC technology
including the historical development of SCC, the effect Of various materials on the fresh
properties of SCC, current status and recent developments using SCC, the use of SCC in
precast facilities, and finally outstanding issues regarding the use of SCC. Chapter 3
presents the methods used to produce the mix designs for this project. This chapter
includes information on fresh property evaluations and results from three separate mix
development processes. Chapter 4 looks at the production and quality control of
prestressed box beams using SCC and provides a discussion on production changes that
were necessary to effectively used SCC in the beam production. The experimental
component to the research, consisting of flexural and shear evaluations of SCC and NCC
beams is presented in Chapter 5. Chapter 6 presents the field—monitoring program and
representative results for the long-term evaluation Of the SCC box beams in the
demonstration bridge. Finally, Chapter 7 presents the conclusions reached from this

research project.

2 SELF -COMPACTING CONCRETE TECHNOLOGY

2.1 Historical Overview

SCC was developed in Japan in the early 1980’s by Okamura and his colleagues
in response to the decline in the skilled labor force in Japan [14]. The inability to find
adequate labor was leading to durability problems in concrete structures caused by either
over or under consolidation. In developing SCC, Okamura wanted to create a durable
concrete that was easy to place and finish independent of the quality of the labor [9] [38].
If successful, Okamura believed SCC would produce products of consistently superior
quality than NCC.

Like all new products, SCC was slow in gaining popularity. SCC was first
successfully produced in 1988 and the first presentation on SCC was at the 2nd East Asia
and Pacific Conference on Structural Engineering and Construction in 1989 [14][23].
However, the first large-scale commercial use of SCC did not occur until 1998. In that
year SCC was used in the anchorages for the Akashi-Kaikyo Bridge in Japan [37]. In this
project, SCC was pumped 600 ft from the on site batching plant to the casting area. The
reduction of the construction period from an estimated 2.5 years with NCC to 2 years
with SCC showed the significant time savings that could be achieved when SCC was
used [14][37]. By the year 2000, 10,800,000 ft3 of SCC was being used annually in
Japan. This volume of concrete was split nearly equally between ready mix applications
and precast production [40].

As knowledge of SCC became more available, its use began spreading from Japan
through Asia and into Europe. Countries such as Korea, Thailand, Sweden, the

Netherlands, and Canada began to research and use SCC in the mid 1990’s [14].

According to Daczko and Vachon [14], the Spread of SCC in Europe did not develop
because of concerns about the quality of conventional concrete, but instead was Spurred
mainly by the economic benefits of SCC recognized by large construction and precast
companies. As the use of SCC became more popular in the mid 1990’s, several
European-wide research projects and committees were formed to help the development
process. One such committee, specifically for the study of SCC, was formed by the
International Union of Laboratories and Experts in Construction Materials, Systems, and
Structures (RILEM from the French name) [18][40]. One project that developed from
these groups was a bridge in Sweden cast completely of SCC. This project differed from
the use of SCC in the bridge in Japan in that the entire bridge was made using SCC. The
use of SCC in this project resulted in a cost savings and a reduction in pollution during
the bridge construction [18][40]. According to Goodier [18] nearly all of Europe was
using or studying SCC by 2003. As the use of SCC became more prevalent in Europe
many countries began adopting guidelines for the use of SCC. By the year 2003
guidelines, or a draft of guidelines, were in place in Austria, Finland, France, Germany,
Italy, Netherlands, Norway, Sweden, and for Europe in general [18].

SCC has developed slower in the United States than in Europe. Beginning around
the year 2000, US precast/prestressed product producers as well as admixture companies
began to push for the ability to use SCC in their projects. According to Daczko and
Vachon [14], and Ouchi et. al. [40] the spread of SCC in the US was motivated by
economic interests similar to those seen in Europe. Research on SCC in the US grew
quickly as was evidenced by the first North American Conference on Self-Consolidating

Concrete, held at Northwestern University in 2002 [18]. Goodier [18] reports in his

survey of the growth of SCC that estimated volumes of 108,000-135,000 ft3 per day of
SCC were being used at precast plants in North America in 2003. Because of the rapid
growth of SCC in the US, a FAST team was created by PCI (Precast/Prestressed
Concrete Institute) with the intention of developing a guideline for the use of SCC.
Interim Guidelines for the Use of SCC in Precast/Prestressed Construction was published

by PCI in 2003 [23].

2.2 Theoretical Considerations

SCC is unique because of its fresh, or plastic, behavior. The ability to fill
formwork and consolidate under its own weight, without external energy, not only
defines SCC, but governs the proportioning of SCC mix designs as well. Concrete in
general can be tailored to achieve performance standards. When developing a traditional
NCC mix design, most of the focus is placed on tailoring the mix design to satisfy the
hardened property requirements. Factors such as the water to cement ratio (w/c), coarse
aggregate content (CAC), sand to past ratio (s/pt), and admixtures (mineral and chemical)
use are adjusted to control the strength, stiffness or the durability of the finished product.
In the case of SCC, these same factors are varied to additionally control the fluidity and
segregation resistance of the fresh concrete. However, tailoring SCC mix designs to
achieve specified hardened properties, while still necessary, is no longer the primary
driving force of an SCC mix design selection.

The widely known benefits of SCC come from the fresh property performance of
the concrete. The fresh properties are governed by three factors: filling ability, passing
ability, and segregation. The filling ability of SCC is its ability to fully fill formwork

without the use of Vibration. This includes flowing around obstacles including dense

reinforcing steel schemes and form block outs. The passing ability of SCC is its ability to
flow around and through obstacles without segregating or experiencing aggregate
blocking. The segregation resistance of SCC is defined as the ability to resist segregation
of its coarse aggregate components through all phases of the construction process
including transportation from the batching plant to casting site, placement in the
formwork, and finishing of the product [23].

The successful development of these three factors can lead to an increase in
productivity and product quality while reducing the cost of production. The high
workability of SCC and the fact that vibration is unnecessary leads to a reduction in labor
required to produce concrete elements as well as a reduction in the time of production.
The self-consolidating nature of SCC can lead to a higher quality finished product than
can be achieved with NCC. While SCC generally has higher material costs, these costs
are more than compensated for by reductions in labor and increased productivity that
result from using SCC [18][23].

The benefits of SCC do not always come easily or immediately. Like with all
new products, there is a learning curve for using SCC [36]. One of the challenges
associated with proportioning SCC is that there are many different ways to achieve the
desired fresh property performance. Compounding this problem is the difference in
behavior that SCC shows depending on the materials used. SCC is said to be extremely
sensitive to small changes in mix design components and quantities [18][36]. When
different methods are used to develop SCC in research, different results often ensue.
Domone [16] reports that research has shown that caution is required when comparing

data from individual mix designs. The practice of reporting results of one SCC mix

design for SCC in general can be misleading, as the data presented in these projects only
accurately represent the specific mix design studied. The results may not be extrapolated
to all SCC mixes in general. To obtain general information on SCC, multiple mix
designs representing multiple proportioning methods need to be analyzed.

Currently there is no commonly accepted procedure to proportion SCC mix
designs [9]. However, according to Khayat [26], the balance that is achieved in creating
SCC is always the same. This balance involves satisfying two Opposing parameters.

On one side of this balance is the need to make concrete fluid enough to flow
easily around Obstacles without becoming obstructed [26]. This high level of fluidity
must be maintained for the duration of the manufacturing process, including placement
and finishing of the concrete. The amount of fluidity required depends on the specific
project being completed. If intricate formwork with a dense reinforcing steel scheme is
used, the fluidity must be relatively high, but if the reinforcing steel is not densely
arranged a lower fluidity can be used.

While maintaining the desired level of fluidity, a quality SCC must also resist
segregation [26]. In order to be effective and useful in practice, the fluid SCC mixture
must remain stable through all phases of the construction process, including transporting
and placing of the concrete. If allowed to segregate, aggregate will settle through the
fluid concrete mixture and collect at the bottom. This non-uniform distribution can lead
to production problems during casting of members, and quality problems with the cured
member.

There is a tradeoff between the fluidity and segregation resistance of the fresh

concrete [26][39]. In general, as the fluidity of concrete is increased, cohesion and the

10

related ability to resist segregation is decreased. At some point both the requirements for
fluidity and segregation resistance can be met. This tradeoff is further illustrated in

Figure 3. Consistently achieving this balance adds to the difficulty of producing high

 

 

 

quality SCC.
Fluidity/Deformability \ / Easy Flow/Low Blockagm
A. Increase paste deformability A. Enhance cohesiveness
* use of HRWR * low We ratio
* balanced w/c ratio * use Of VMA
B. Reduce inter-particle friction B. Compatible flow space and
* low coarse aggregate volume aggregate size
* use cont. graded powder * low coarse aggregate vol

 

 

 

ume
K * low max. size aggregate /
fluidity/Stability flomogenerty/Stabrlrty \

 

 

   

       

Trade-off A. Reduce solids separation
* . .
Low Viscosity. lrmrt aggregate content
g High may * reduce max. srze aggregate
g * increase cohesion & viscosity
g - low w/c ratio
‘5 . . - use of VMA
0 High VISCOSIIY.
Low Flutdity
Viscosity B. Minimize bleeding

* low w/c ratio

* use of high-area powder
* increase VMA

 

 

 

 

 

Figure 3. Fluidity, Stability Tradeoff in SCC Mix Designs

There are many ways to increase the fluidity of the concrete. Perhaps the easiest
way is to add water to the mix [26]. The end result of this process though, is a concrete
with a high w/c ratio and low cohesiveness. Furthermore, a concrete with a high w/c ratio
is often less durable and weaker, which can lead to early deterioration of the concrete
structure.

A second way to increase the fluidity of the concrete is to remove a portion of the
coarse aggregate from the mix. With fewer large particles, less energy is required to

deform the fluid concrete mixture as it flows around obstacles without blocking.

11

However, fewer coarse aggregates also means the ability of the hardened concrete to
resist crack propagation is limited. As coarse aggregate iS generally much stronger than
the cement paste, a crack will develop at the interface between the cement paste and the
aggregate. The random arrangement of the coarse aggregate through the cement paste
creates a tortuous path for a crack to propagate through a concrete element. If fewer
aggregates are present in the concrete, the crack will require less energy to propagate, as
it will have fewer obstacles to avoid.

The segregation resistance of the concrete can be achieved through an increase of
the s/pt ratio. Having more fine aggregate in the concrete mixture raises the viscosity of
the fluid portion of the mixture. This increased viscosity will result in better suspension
of the coarse aggregate. However, a mix using a poor aggregate gradation can lead to
problems with the water demand and workability of the concrete. A mix with a high fine
aggregate content is also more likely to experience drying shrinkage [36].

It can be seen that every change made to a concrete mix design to produce a
desired result may lead to negative consequences as well. To properly proportion a
concrete mix, the effect of altering all of the parameters must be considered together.
SCC generally cannot be efficiently made by merely altering a NCC mix design. To
effectively achieve the desired benefits of SCC, more ingredients need to be considered
than are in the NCC mix design.

Several additional ingredients in SCC make achieving the balance between
fluidity and segregation resistance easier. Among them are chemical admixtures
including high range water reducers (HRWRS) and Viscosity modifying admixtures

(VMAs). Traditional mineral admixtures including silica fume, fly ash, blast furnace

12

slag, and limestone powder also make the balance between fluidity and cohesion possible
[36]. When the effects of these ingredients are considered along with the other
parameters discussed above, high quality concrete that exhibits the benefits of SCC can

be produced consistently.

2.2.1 High Range Water Reducers (HRWRS)

HRWRS are very important to the production of SCC [38][39]. These chemicals
act to increase the fluidity of the concrete while maintaining a reasonable w/c ratio. This
is accomplished by more effectively using the water added to the mixture. HRWRS are
made of organic molecules that adsorb on the solid particle surfaces. Naturally, the
cement particles dispersed in the water have mixed charges causing some particles to be
attracted to each other through electrostatic forces. This attraction causes flocculation of
the particles, which traps a significant portion of the water. Because this water is trapped
it is not utilized in the cement paste, as it cannot come into contact with unhydrated
cement. Over time this water evaporates from the concrete leaving behind pores.
Because this water is not utilized in the cement paste the Viscosity of the paste is higher
than it theoretically could be. The molecules of HRWRS neutralize the surface charge of
the cement particles so that all of the particles have the same charge. This causes the
particles to repel each other throughout the mix water. As the particles are more evenly
dispersed through the water, more of the water is used to hydrate the cement producing a
less viscous paste that can flow much easier than before. The proper use of HRWRS
allows SCC to be developed with a reasonably low w/c ratio leading to a higher strength,

more durable concrete [26][32].

13

2.2.2 Viscosity Modifying Admixtures (VMAS)

VMAs directly affect the mix water of the concrete [36]. These chemicals work
to increase the Viscosity of the cement paste, which in turn helps suspend the coarse
aggregate in the fluid cementitious mixture. Using a VMA results in a more stable and
cohesive concrete mixture. The molecules that make up VMAS are high molecular
weight, water-soluble polymers that act to raise the Viscosity of the water. These
admixtures work to prevent segregation through increased cohesion as well as prevent
bleeding [32]. The use of VMAs achieves the desired increase in viscosity without

leading to an increased susceptibility to drying shrinkage as other methods would [36].

2.2.3 Mineral Admixtures

Partial cement replacement by silica fume, fly ash, blast furnace slag, or limestone
powder is a common practice in NCC, but it can also help to produce SCC as well. The
use of these admixtures influences the fluidity of SCC by reducing inter-particle friction.
The shape of these particles is spherical when compared to the irregular shape of the
cement particles [36]. As mineral admixtures are less reactive than cement, their
inclusion will slow down the hydration reaction. This slower reaction increases the time
that the mixture remains in its fluid state. The longer the reaction takes, the longer the
concrete remains a fluid and preserves the beneficial fresh properties that define SCC.
There is also a cost benefit when using fly ash, as these particles are by-products of other

industries and they are thus widely available and cheap thus reduce the total cost [32].

14

2.3 Status and Recent Developments in SCC

It has been nearly twenty years since the development of SCC. Since its creation
in Japan, the use of SCC has spread around the world and has been used in many
projects. Committees in major industry and research organizations such as the American
Concrete Institute (ACI), the American Society of Testing Materials (ASTM), PCI, and
RILEM continue to evaluate the use of SCC [23].

In the United States the precast industry and admixture companies continue to
push for the widespread use of SCC in many facets of civil engineering construction.
However, some consumers Of concrete technology including state Departments of
Transportation (DOTS) have been slow to accept SCC as a viable alternative to NCC
[35][41]. These groups have concerns about SCC related to quality control and long-term
performance. These concerns keep SCC as a popular research topic in the US, but limit
its implementation in construction projects.

A survey of research and case studies using SCC published in 2006 [16] shows
that the spread of research on SCC has shifted from its center in Asia during 1993-1999
to Europe from 2000-2003, with the USA increasing the number of research and
commercial projects through 2003. In 68 published case studies, 57 were uses of SCC in
ready mix projects and only 11 were in precast facilities. However, Domone noted [16]
that the number of reports Of SCC being used in precast plants was increasing at the time
of the survey. A large majority of the cases were in commercial projects; just 25% of the
case studies were for demonstration or large scale testing projects. A summary of the
mix designs used in these case stridies shows that all mixes follow the same guideline in

general:

15

I Low coarse aggregate content
I Increased paste content

I High powder content

I Low water/powder ratio

I High superplasticizer dose

I Viscosity modifying agent.

Also, 90% of the mixes had a spread flow of 23”-29” [16].

2.3.1 Recent SCC Research

Research on SCC continues to evaluate issues related to the mix design
development, fresh properties, and hardened properties. Many research projects look at
the effect of mix design components including admixtures, on the fresh and hardened
properties of SCC [25][28]. Still others try to develop schemes for developing mix
designs [50]. Many papers have focused on the effect that SCC has on formwork
pressure. These projects are important to the continued advancement of SCC as they
increase the understanding of how SCC works. However, these projects do not address
significant areas of concern regarding the use of SCC. SCC can only be used
successfully if structural elements produced with it perform at a satisfactory level.
Paramount to this is the successful use of SCC in the production of prestressed products.

Limited research on the structural behavior of SCC has been published. The
published projects focus on topics such as the bond of prestressing strand to the concrete
and the shear strength of SCC members. Full scale testing of structural elements cast

with SCC or long term monitoring of SCC structures is rarely reported in the literature.

16

2.3.2 Performance of Beams Made With SCC

Multiple projects investigating the performance of both reinforced and
unreinforced beams made with SCC have produced somewhat conflicting results. One
project looking at the load displacement response reinforced beams cast with SCC reports
that the SCC beams behave similarly to regular concrete, with the only differences in
cracking pattern being attributed to differences in the compressive strengths of the
concretes [45]. Das et. a1. [15] found that reinforced concrete beams cast with SCC had
higher shear strength than those cast with regular concrete. Wilson and Kiousis [53] on
the other hand found that SCC beams had a reduced shear strength when compared to
NCC. Limited information was provided on their SCC mix design. They reported only
that “the tested strength of this mix was 10 ksi and the mix uses a large volume of
smaller aggregates to assist in uniform stratification [53].” They reported that the ACI
Building Code equations for the Shear strength of reinforced concrete beams did not
always yield conservative results when SCC was used to cast beams. Lachemi et a1
tested beams made of SCC without shear reinforcement to analyze the shear strength Of
SCC. They concluded that the reduced coarse aggregate content of SCC lowered the post
shear cracking resistance of the beam [27]. The confusion that follows these results has

made the widespread acceptance of SCC difficult.

2.3.3 Prestressing Strand Concrete Bond

In prestressed concrete applications the bond between the prestressing strand and
the concrete plays a major role in the adequate behavior of the structural element.
Several research projects have studied strand bond performance on SCC beams again

with conflicting results being reported. The bond strength and transfer length of

17

pretensioned SCC bridge girders was investigated by Girgis and Tuan, who found that
SCC has higher bond strength and therefore lower transfer lengths at 28 days than
conventional concrete [33]. Their research did not include full scale testing of the bridge
girders, rather they took measurements for transfer length on full scale prestressed bridge
beams that were to be implemented in a bridge. Hegger and Kommer [21] found that the
ACI Building Code yielded unconservative predictions for the transfer length of
prestressed beams cast with SCC. Larson, Peterman and Esmaeily investigated the
transfer length, and flexural capacity of prestressed SCC beams and found that ACI and
AASHTO codes were conservative in the predictions of both parameters [29]. Burguefio
and Haq [10] reported that ACI equations conservatively predicted transfer lengths in
SCC beams, but the equations for development length did not produce conservative
results for hi gh-fine SCC mix designs. Like the Shear resistance of reinforced beams, the

confusion that results from the reports raises questions about the acceptability of SCC.

2.3.4 Prestressed Beams Made With SCC

While the bond between prestressing steel and concrete is an important factor in
prestressed concrete beams, the absolute capacities of these sections is the ultimate
parameter of importance. Several projects have investigated the flexural and Shear
capacities of prestressed beams cast with SCC through large-scale testing. Choulli and
Mari [12] found that even though the shear capacity of prestressed girders cast with SCC
was lower than NCC, applicable codes conservatively predicted the performance. Naito
[35] analyzed the use of SCC for bridges in Pennsylvania. In doing so he tested 45-inch
deep 35 foot long Bulb-T beams to failure. The response of the SCC beams and the NCC

beams were very similar with both exceeding the AASHTO design capacities. The

18

Virginia DOT investigated using SCC in bridge girders [41] by conducting tests on Bulb-
T beams for transfer and development length. Tests on flexural and shear capacities are
also being conducted but the results reports have not been published yet. Larson et. al.
[30] monitored prestressed I-girders in a bridge for one year to evaluate prestress losses
due to creep and shrinkage. They reported that the SCC beams had a smaller prestress
losses than the NCC beams, however, these losses were within the PCI recommendations.
The results of these studies Show that the capacities of SCC beams are in line with

building code requirements and with the performance of NCC beams.

2.3.5 Standardization of SCC Fresh Property Testing
In 2003 PCI published the “Interim Guidelines for the use of Self-Consolidating
Concrete in Precast/Prestressed Applications” [23]. This report was the first guideline
published in the United States. In the guideline, PCI identified nine common test
methods to evaluate the fresh properties of SCC. These methods were:
I Slump Flow/Inverted Slump Flow, includes Visual Stability Index (VSI)
and T50 test methods
I J -Ring Test method
I V-Funnel Test method
I L-Box Test Method
I U-Box Test Method
I Filling Vessel Test Method
I GTM Screen Stability Test Method
I Orimet Test Method

I Bleeding Test Method (French) [23].

19

The tests were chosen for their common use in both commercial and research
arenas. In the 2006 edition of the ASTM testing methods three tests for SCC were
standardized. These tests include the J-Ring test for passing ability, the Slump Flow test
for flow ability, and the Column technique for segregation resistance [48][49]. The
standardization of these tests by ASTM provides a unified model from which SCC fresh
properties should be measured. A complete discussion of the fresh property tests used for

this project is included in section 3.2.2.

2.4 SCC for Precast Construction

Precast plants have demonstrated successful uses of SCC in Japan, Europe, and
the US [16][18][24][40]. Using SCC in a precast plant presents challenges that are not
seen when using NCC. The mix design development of SCC is time consuming and can
be difficult. Production processes used with NCC may need to be altered to use SCC
effectively. Even with these potential problems, many precast plants successfully use
SCC every day.

In Japan, SCC has been used in several high profile structures. The Higashi-Oozu
Viaduct used precast/prestressed T beams [40]. Cost savings from using SCC in this
project totaled 7%, including a 33% decrease in the labor costs and a 4% increases in the
material costs. Umehara et. al [52] reported that two specific SCC products have been
used to produce a total of 63,000 ft3 of concrete for projects at 35 separate properties.

SCC is a popular alternative to NCC in Europe as well. The Netherlands,
Sweden, Norway, and France are the largest producers of SCC in Europe. The largest
manufacturer of precast concrete in Europe is the Consolis Group [24]. The company has

52 factories and is currently operating in 11 countries. Consolis’ annual use of SCC

20

exceeds 100,000 In3 and accounts for roughly 20% of the total concrete produced. In
2005 the estimated production of SCC was 261,400 cyd [24].

To date the Consolis Group has used SCC in multiple projects including the
manufacture of multiple prestressed beams. Other unique projects, such as concrete sheet
piles, complicated staircase elements, and an artificial reef element have all been cast
using SCC in their plants. While no specifics are given, the company claims benefits
from reduced labor requirements and improved work environment [24].

The precast/prestressed industry in the United States is beginning to use SCC as
an alternative to NCC [40]. The introduction of SCC in precast plants has led to
problems that must be resolved. The increased number of material constituents and
critical reliance on quality control can make consistent production of high quality SCC in
precast plants difficult [17]. Employees who have developed specific skills in working
with NCC need to be retrained to effectively work with SCC. Even with these problems,
SCC is becoming a common product in the US. Two published case studies regarding
the use of SCC in US precast plants are reviewed next to highlight difficulties and
successes with the use of SCC in precast concrete production.

At the Rocky Mountain Prestress Company, Denver, Colorado, a test program
was developed to aid in the successful implementation of SCC in the plant [17]. The
program consisted of three phases, a laboratory phase, a field test phase, and finally an
implementation phase. The laboratory phase consisted of developing a relationship
between mix design, cost, and performance. Mix design parameters were analyzed for
their effect on fresh and hardened properties of the concrete as well as cost to develop an

optimal mix. In the field test phase the flow properties of the concrete were evaluated in

21

conjunction with the plants mixing and placing equipment. Once satisfactory
performance in these preliminary phases was established, the implementation phase was
conducted to assess the performance of the concrete through the production Of several
mock-up of double T elements. The company encountered problems with the
transportation of the concrete from the batching plant to the casting area and with the
slope of the double-T formwork. Working through these problems they were able to
successfully introduce the SCC mixes into their daily operations. The results of tests
show that the SCC mixes used in the plant have decreased the concrete placement time
by 20% compared to the use of NCC [17].

A second example of the use of SCC in the US precast/prestressed industry is at
High Concrete Structures, Inc., Denver, Pennsylvania. At this plant SCC was used to
help cast twice as many Double T beams in one casting bed than was possible with NCC
[43]. To accomplish this, the company had to develop SCC mixes that could obtain 3500
psi in six to seven hours and 7000 psi at 28 days. The use of this new process in one of 4
casting beds led to a 25% increase in production per day for that bed. The company
projected a total increase in production of 75% if all four casting beds were double cycled

in a similar manner [43].

2.5 Outstanding Issues

Concerns about the use of SCC remain, even as it becomes more common in
everyday practice. If SCC is approved for use by state DOTS, use will increase
significantly. However, it is yet to be seen how SCC will be fully utilized by producers
and consumers of SCC technology. Professional organizations globally have begun

producing specifications for the use of SCC. In Europe, the European Concrete Standard

22

EN206-1, is being modified to specifically include SCC. This modification could include
standardization Of test methods for SCC and the development of a specification for the
use of SCC [20].

Further research on long term properties of SCC such as creep and prestress
losses will take time to be conducted and published. Only with continued use and
increased experience will the behavior of structures cast with SCC become better known.
The more SCC is used the better its behavior will be understood.

The use of SCC is currently being hindered by the lack of standards and codes for
the use of SCC. While guidelines have been published, they must be updated so that they
represent the state of the art in the industry. Finally, long term data from research
projects must be collected and presented to fully understand how SCC behaves over long

periods of times.

23

3 MIX DESIGN DEVELOPMENT AND PERFORMANCE

3.1 Overview

The general lack of experience in working with SCC leads to difficulties in
developing SCC mix designs that consistently produce high quality SCC. While many
methods have been discussed for such a development, mixes are heavily dependent on
locally available materials and processes [9][26][39][50]. SCC mix design development
requires trial and error procedures that iterate on desired out comes from a specific fresh
property or a combination of fresh properties. Upon satisfactory achievement of these

properties, the mix must be evaluated and refined for hardened performance.

3.2 SCC Mix Design Development and Evaluation
3.2.1 SCC Mix Design Development Methods

With all of the possible changes that can be made to produce SCC it is often
challenging to develop a quality SCC mix that is reliable and easy to recreate. Many
SCC mix designs have been proposed by researchers. However, the use of a single mix
design in a research project leads to difficulties when interpreting and comparing the
published data. While the results reported from research accurately represent the specific
SCC mix design studied they may not represent all SCC mix designs. This problem
could be alleviated if a range of mix designs are studied in a project. This would allow
researchers to see how mix design modifications affect research results.

Among all of the proposed SCC mix designs, there seems to exist two approaches
that bound most current SCC mix designs. These two limiting concepts have been

identified in the research by Khayat [26] and Bonen and Shah [9]. Bonen and Shah [9]

24

referred to them as powder-type SCC and VMA (Viscosity Modifying Admixture)-type
SCC, while Khayat describes them more generally. The first method described by
Khayat achieves the fluidity requirements through the reduction of the coarse aggregate
volume and the use of HRWR (High Range Water Reducer). The stability comes from a
low w/c ratio with a high s/pt (sand to past) ratio. N o VMA is required in this mix. The
second method uses a high w/c ratio with little to no HRWR to achieve the fluidity
requirements, allowing for a moderate volume of coarse aggregate, while the stability is
achieved through the use of VMA and moderate s/pt ratios. Table 1 details the relative
proportions of the w/c ratio, amount of HRWRS, VMAS, CAC (Coarse Aggregate
Content), and s/pt ratio for the two methods.

In general, powder-type SCC uses HRWRS to increase the fluidity of the cement
paste while maintaining a low w/c ratio and a decreased CAC. Larger amounts of fine
aggregates are necessary to resist segregation in the mix as the VMA level is low. VMA-
type SCC has a higher w/c ratio and uses less HRWR for its fluidity. The Viscosity and
segregation resistance are controlled by the use of VMAs while maintaining moderate to
low s/pt ratio and level of CAC. Either of these methods, or a combination of the two has
been shown to produce high quality SCC [9][26].

Table 1. Mix Desiggxmarimental Matrix

 

 

 

 

Mix Design w/c HRWR VMA CAC S/Pt
Powder SCC Low High Low less More
Combination SCC Moderate Moderate Moderate Moderate Moderate
VMA SCC High Low High Moderate Less

N CC Moderate As needed None Moderate Moderate

 

25

 

 

 

3.2.2 SCC Mix Design Evaluations
As previously discussed, the performance Of SCC is most Often characterized by

its fresh properties. The ability of the concrete to fill formwork, flow around obstacles,
and resist segregation determines the usefulness of the mix for a given application. In the
2006 edition of the ASTM standards for Concrete and Aggregates [48][49], two tests
related to the fresh properties of SCC were included. These tests are the Slump Flow and
the J-Ring. While there are many other tests that commonly used to measure relevant
parameters [9][23], only four were chosen to characterize the SCC mixes for this project.
They are:

A. Inverted Slump Flow Test

B. Visual Stability Index

C. J-Ring

D. L-Box

3.2.2.1 Inverted Slump Flow Test

Slump flow is a measure of the concrete’s ability to flow under its own weight. It
is analogous to a standard concrete slump test for NCC. For this test a standard
Abrahm’s cone is inverted on a flat surface. As shown in Figure 4, concrete is placed in
the cone without any form of Vibration or tamping. Excess water or concrete is removed
from the flat surface around the cone as its presence can affect test results. As the cone is
lifted vertically in a slow and even motion, the SCC flows into a circular concrete patty,
as shown in Figure 5. The average diameter of the patty is obtained by taking two

orthogonal measurements. This average value represents the slump flow for the SCC.

26

An SCC mix that has a high value of slump flow will have an easier time filling
formwork under its own weight [23].

Prior to conducting this test the acceptable range of the SCC slump flow must be
defined. It is important that the acceptable range be relevant to the project for which the
concrete will be used. A larger vale would be required on a project with more intricate
formwork or dense steel reinforcing schemes. Bonen and Shah [9] suggest a minimum
value for the inverted spread of 23 inches. According to the Special Provision for the
Production of Prestressed Beams with Self—Consolidating Concrete [11] written for this
project and included as APPENDIX A, a slump flow diameter of 27 in. plus or minus 1

in. (26 in. — 28 in.) was prescribed as a satisfactory range of slump flow.

 

Figure 4. Inverted Slump Flow Test

27

 

Figure 5. Concrete Patty from Inverted Slump Flow Test

3.2.2.2 Visual Stability Index

While adequate flow is generally the main parameter used to assess SCC, a highly
flowable concrete may not always be acceptable. To be of high quality, SCC must not
only fill formwork and consolidate under its own weight, it must also remain
homogenous throughout the entire construction process. The visual stability index (VSI)
is a qualitative measure of the concretes ability to resist segregation. While no single
property of the concrete can be quantified using the VSI, it can be used as a measure of

the relative quality control of the concrete [23].

28

 

Table 2. VS] Ratings Criteria [23]

 

 

Rating Criteria
0 No evidence of segregation in the slump flow patty or sampling wheelbarrow

 

No mortar halo or aggregate pile in the slump flow patty, but very slight
0.5 evidence of bleed or air popping on the surface of the SCC in the sampling
wheelbarrow

No mortar halo or aggregate pile in the slump flow patty but some slight bleed

 

 

l or air popping on the surface of the concrete in the samplifi wheelbarrow
Just noticeable mortar halo and/or a just noticeable aggregate pile in the slump
1.5 . . . .
flow patty and noticeable bleedrnw the sampling wheelbarrow
2 A slight mortar halo and /or aggregate pile in the slump flow patty, and highly

noticeable bleedinflr the wheelbarrow
Clearly segregating by evidence of a large mortar halo and/or large aggregate

3 pile in the center of the concrete patty and a thick layer of paste on the surface
of the restflg concrete in the wheelbarrow

The VSI can be recorded from the previously described inverted slump flow test.
Once the test has been conducted, a rating is given to the concrete based on criteria such
as the amount of water bleeding from the edge of the concrete patty, the amount of
aggregate piled in the center Of the patty, or the presence of an aggregate free mortar halo
at the edge of the patty. This test is subjective and is best used to relatively compare
several similar SCC mixes.

A perfect VSI rating of 0 would be obtained by a mix that produces a concrete
patty with aggregate dispersed to the edge of the concrete with no visible free water
bleeding beyond the edge of the patty. The ratings increase in increments of 0.5 and vary
from O to 3. Table 2 lists the rating evaluation guide as taken from PCI’s “Interim

Guidelines For the Use of SCC” [23]. For this project a VSI of 1.0 or lower was required

for a qualifying test.

29

3.2.2.3 J -Ring

While the slump flow is an adequate measures of the concrete’s ability to flow, it
does not measure the ability of the concrete to flow around obstacles. As the concrete
flows through obstacles such as reinforcing steel, the entire mixture must travel together.
If the coarse aggregate were to be blocked by the obstacles as the paste continues
through, the mix would have to be reevaluated. The J-Ring test can assess this “passing
ability” of SCC [23].

The J-Ring test is very similar to the inverted Slump flow test. It uses an inverted
Abrahm’s cone placed inside a ring of evenly spaced vertical bars as shown in Figure 6.
The spacing and diameter of the bars in the J-Ring should reflect the reinforcing scheme
of the element being cast and the maximum aggregate size used in the SCC mix design.
The test is executed exactly like the inverted slump flow test with the result shown in
Figure 7.

Important information on the passing ability is not only found in the average
diameters of the concrete patty but in the depth of the concrete inside and outside the
ring. These values are obtained by measuring from the top of the J-Ring to the top of the
concrete and subtracting this value from the full height of the J -Ring. The measurements
are taken at four locations along the two orthogonal diameters as shown in the schematic
drawing of the J—Ring test in Figure 8 and Figure 9. In an ideal situation the concrete
would be able to flow without any resistance from the obstacles. This would be seen by a
small difference between the level of concrete inside the J-ring and the level of the
concrete on the outside of the J-ring. In addition, the coarse aggregate would still remain

evenly dispersed to the edge of the concrete patty.

30

 

Figure 6. J -Ring Test Set Up

Calculating the J-Ring test value is done by taking the median value of the four depth
measurements of the concrete inside and outside the J-ring. The changes in depth from
the center of the concrete patty to the inside of the J-Ring and from the inside of the J-
Ring to the outside of the J-Ring are used to calculate the J-Ring value. Equation (1) is
used to calculate the J-Ring Value:

(1) J — Ring = 2(ham ‘hbml— (hlm ‘haml
where ham is the median inside depth, hbm is the median outside depth and hlm is the depth
at the center of the patty. According to the PCI Interim Guidelines for SCC a value of 0.6
in. shows excellent passing ability and a value of 0.4 in. is acceptable [23]. According to
the Special Provision for the Production of Prestressed Beams with Self-Consolidating
Concrete [11] (APPENDIX A), the state requirements for this project area I Ring value

between 0.5 and 0.6.

31

 

Figure 7. Result of J -Ring Test

Concrete Patty
Diameter 1

_—’.

 

 
        

J-Ring
(IT-Diameter)

Concrete Patty
Diameter 2

 

 

 

 

Figure 8. Schematic of J -Ring Test

32

 

 

Concrete Patty Diameter

l.— 12" J-Ring Diameter T-l

 

 

 

 

 

 

 

 

 

 

 

L l
rh-al hl h-b3
K4 .1. f 1..
L _ I r — I {\LI
h-bl-T T T h-as—T T

Figure 9. J -Ring Test Apparatus Profile

3.2.2.4 L-Box

The L-box test is another measure of the passing ability of SCC. The apparatus
for this test is shown in Figure 10. It consists of two rectangular chambers arranged in
the shape of an L. The vertical chamber is separated from the horizontal chamber by a
vertical sliding gate. To obstruct the flow of the concrete from the vertical camber into
the horizontal chamber, vertical bars are placed just out side the gate. The test is
conducted by filling the vertical chamber and allowing the concrete to rest for one
minute. At the end of one minute, the gate is lifted up to the height of the horizontal
chamber and the concrete flows through the horizontal chamber as shown in Figure 11.
The time the concrete flow takes to reach the halfway point and the end of the horizontal
chamber are recorded. Once the concrete comes to rest the height of the concrete in each
chamber is measured. If the obstacles provided little resistance to the passing ability of
the concrete, the ratio of these heights would equal one [23]. For this test a value from
0.5 to 1.0 is considered acceptable. According to the Special Provision [11] (APPENDIX

A) for this project the state requirement is a L-Box value greater than 0.8.

33

 

Figure 10. L-Box Test Apparatus Figure 11. L-Box Test

3.2.2.5 Fresh Property Test Discussion

It is commonly accepted that the battery of tests just described, while important in
research, may not be practical in a production facility. Results of these tests are
extremely time sensitive, as the plastic properties of SCC decay rapidly as the concrete
curing process progresses. The tests also require more equipment than a standard slump
test, which makes transporting them to and around a job site difficult. Also, the results of
the L—Box and J-Ring test require calculations that cannot be easily done in the field.
However, this does not mean these tests have no value. When developing an SCC mix
design these tests can provide important information about the fresh property behavior of
the concrete. The performances of the concrete can lead directly to changes in the mix
designs. The use of tests with overlapping scopes provides multiple points of reference

on which to base changes to the mix proportioning.

34

As mentioned previously, the 2006 edition of the Annual Book of ASTM
Standards [48][49] includes two standards specifically for the use of SCC. These
standards include the Slump Flow Test Method [49], and the J-Ring Test Method [48].
Some differences exist between the methods published by ASTM and those in PCI’s
“Interim Guidelines for the Use of SCC [23].” The most significant difference is that the
ASTM standard for the J -ring test does not include a calculation based on the depth of the
concrete. Instead the passing ability is calculated by finding the difference between the
Slump Flow and the J-Ring Flow, where both values are the average of two orthogonal
diameters of the concrete patty. According to ASTM a difference of 1 inch or less
represents a concrete with good passing ability while a difference of 2 inches or greater
represents a concrete with poor passing ability [48]. As the ASTM standards were not
available at the time of this project, the test methods presented in the PCI guidelines were

followed.

3.2.3 SCC Mix Design Development

The SCC mix designs used for this project were developed to bound the current
practices of SCC mix proportioning. These mixes were intended to lie at the extremes of
possible SCC mixes. The concept was that any effects produced by these mixes would
bound the results from mixes that would be commonly developed to somewhere in
between these extremes.

All of the mix designs developed for this project used 700 lbs of Type IH cement.
The target 28-day compressive strength of all of the mixes was 5,500 psi with at least
5,000 psi required to release the prestressing strands. Local natural aggregates, which

met MDOT specifications for use in prestressed bridge elements, were used, namely

35

natural coarse aggregate 17A and natural sand 2N S. The level of entrained air was to be
6%. The specific quantities for each SCC mix design and the process used to obtain them
are discussed below.

Three SCC mix designs were developed for this project to bound current practices
of SCC mix design proportioning. One of the mix designs, SCC 1, used a powder based
approach while a second mix, SCC 3, used a VMA-based approach, both similar to the
mix designs described by Khayat [26] and Bonen and Shah [9], to achieve the SCC fresh
property behavior. The final SCC mix design, SCC 2, used a combination of the two
approaches.

The project mix designs were developed by the fabricator based on baseline mix
designs provided by MSU through the project’s special provisions [11]. These baseline
mix designs, presented in Table 3, were previously used in an MSU research project that
studied the bond of prestressing strand and SCC and had an intended compressive
strength of 7000 psi at 28 days. Degussa Admixtures Inc. (now BASF Admixtures)
provided technical assistance in the development of the mix designs.

Development of the final SCC mixes for use in the project required numerous trial
batches and adjustments to the baseline mix designs in Table 3. Performance Of the mix
designs was assessed through the following tests: inverted slump flow, J-Ring, L-Box and
Visual Stability Index [23]. Adjustments were made to the mix designs based on the test

results until satisfactory fresh performance was consistently produced.

36

Table 3. Baseline SCC Mix Designs
NCC SCC 1 SCC 2 SCC 3

 

Constituent (lbs/yd3)

 

 

 

 

 

 

Cement — Type III 700 700 700 700
Fine Aggregate 1216 1519 1426 1275
Coarse Aggregate (17A) 1580 1380 1380 1435
Water 280 245 280 315

Air 6% 6% 6% 6%

w/c ratio 0.4 0.35 0.4 0.45

 

Admixtures (oz/cwt)

 

 

 

 

Air Entraining l 0.5 0.5 0.5
High Range Water Reducer 2 6.29 7 8
Viscosity Modifier 0 0 1 9
Set Retardant 6 6.14 6.69 6

 

3.2.3.1 SCC 1

The SCC 1 mix was intended to imitate a powder-type SCC, as described by
Bonen and Shah [9] and Khayat [26]. This means that the concrete fluidity is achieved
by using a relatively high amount of HRWR and a reduced amount of coarse aggregate,
while the segregation resistance comes from an increased amount of fine aggregate. The
high amount of HRWR allowed the w/c ratio to be kept to a relatively low 0.35.

Table 4 shows the mix design evolution for SCC 1 with the results of the fresh
property and compressive strength tests in Table 5. The first attempt at creating SCC 1
resulted in a slump flow value of 24 in., which was less than the 27 in. requirement
specified in the project special provisions [11]. However, the mix showed good
resistance to segregation with a VSI of 0. To raise the spread value of the concrete,
HRWR was added to the mix for the second trial. Although this change increased the
fluidity of the concrete, it lowered the segregation resistance of the mix. In order to
maintain the fluidity but improve the stability of the concrete, the fine aggregate content

was reduced while the amount of water added to the mix was increased. The result of

37

this was a slightly less fluid mix, with a perfect VSI rating. While the spread did not fall
in the required range, the mix design was adopted as small adjustments to the fluidity
could be made at the time of casting. The final mix design was able to meet the strength
requirement as 5000 psi was obtained in less than one day.

Table 4. SCC 1 Mix Development

 

SCC 1-1 SCC 1-2 SCC 1-3

 

 

 

 

 

 

 

 

 

 

 

Constituents (lliyd3)
Cement 700 700 700
Fine Aggregate 1,612 1,612 1591
Coarse Aggregate (17A) 1350 1350 1350
Water 248 248 256
Air 6.00% 6.00% 6.00%
w/c ratio 0.35 0.35 0.37
Admixtures (oz/ cwt)
Air Entrainin g l 1 l
HRWR 14 15 15
VMA 1 1 l

 

Table 5. SCC 1 Test Results
SCC 1-1 SCC 1-2 SCC 1-3

 

 

 

 

 

 

Slump Flow (in.) . 24 26.5 25
V81 0 .5 O
f’ c (psi)
<1 day 1294 4754 4953
7 day 7630 8677 6265

 

3.2.3.2 SCC 2

The SCC 2 mix represented a balanced combination of a powder and VMA—type
SCC mixes as described Bonen and Shah [9] and Khayat [26]. The target w/c ratio for
this mix was 0.4. As the amount of cement in each mix was fixed at 700 lbs/yd3, the

higher w/c ratio was achieved by using more water than was used in the SCC 1 mix

38

design. Because of the increased amount of water in this mix design a lower amount of
HRWR was used than in the SCC 1 mix. Compared to SCC 1, the SCC 2 mix design had
a larger CAC and a reduced amount of fine aggregates, resulting in a lower s/pt ratio.
Table 6 shows the evolution of the SCC 2 mix design and the results for fresh
property and compressive strength tests are shown in Table 7. The spread value and the
VSI rating for the first SCC 2 mix design attempt fell short of the special provision
requirements [11]. To compensate for this, coarse aggregate was removed from the mix
and the fine aggregate content was increased. HRWR was also added to ensure that good
flow was obtained. While the second mix of SCC 2 met the required slump flow value,
the mix showed some segregation with a VSI rating Of 1. To improve the segregation
resistance, a portion of the fine aggregate was removed and VMA was added. Water was
also added to the mix design, slightly raising the w/c ratio, to maintain the desired spread
value. The test results showed that the final SCC 2 mix met the requirements for both the
slump flow and VSI rating. The flow was within 1 in. of the 27 in. requirement and the
VSI was 0.5. The strength requirement was also met by this mix design as 5000 psi was

obtained in 2 days

39

Table 6. SCC 2 Mix Development

 

SCC 2-1 SCC 2-2 SCC 2-3

 

 

 

 

 

 

 

 

 

 

Constituents (lbs)
Cement 700 700 700
Fine Aggregate 1,457 1,527 1513
Coarse Aggregate (17A) 1420 1350 1350
Water 280 280 285
Air 6.00% 6.00% 6.00%
W/c Ratio 0.4 0.4 0.41
Admixtures (oz/cwt)
Air Entraining 1 1 l
HRWR l l 12 12
VMA 0 0 2

 

Table 7. SCC 2 Results
SCC 2-1 SCC 2-2 SCC 2-3

 

 

 

 

 

 

 

Spread (in.) 25.5 27.5 26
VSI 2 1 .5
f’ c (psi)
<1 days 2873 5012 4294
2 days 5252 - 5614
7 days 6511 6262 6281
14 days 6360 6520 6078

 

3.2.3.3 SCC 3

The SCC 3 mix was modeled after a VMA type SCC mix described by Bonen and Shah
[9] and Khayat [26]. The target w/c ratio of the SCC 3 mix was the highest of the four
project mix designs at 0.45. The larger amount of water used in this mix required that the
amount of HRWR used in the mix design to be the lowest of the SCC mixes. To ensure
adequate segregation resistance of this mix design, a relatively large amount of VMA was
used. The higher amount of VMA meant that segregation could be controlled with fewer

fine aggregates and a higher CAC than in the previous SCC mix designs.

4O

Table 8 shows the mix design evolution for SCC 3. As can be seen from the
results shown in 0, the first SCC 3 mix design did not meet the slump flow requirement,
but had an acceptable segregation resistance with a VSI rating of 1. Both parameters
were improved by increasing the amount of HRWR and VMA. While test results were
not provided by the fabricator, the second batch of the SCC 3 mix design met the
performance standards from the special provisions, and the mix design was adopted

without further alterations.

Table 8. SCC 3 Mix Design Development

 

SCC 3-1 SCC 3-2
Constituents (lbs/yd3)

 

 

 

 

 

 

 

 

 

Cement 700 700
Fine Aggregate 1,320 1,320
Coarse Aggregate (17A) 1450 1450
Water 320 320
Air 6.00% 6.00%
W/c Ratio 0.46 0.46
Admixtures (oz/cwt)
Air Entraining 1 1
HRWR 10 10.7
VMA 5 6

 

Table 9. SCC 3 Results
SCC 3-1 SCC 3-2

 

 

 

 

 

 

Spread (in.) 25 -
VSI 1 -

F’c (psi)
<1 day 5780 -
7 day 6316 -

 

41

3.2.3.4 Final mix Designs

The final mix designs for each concrete type are given in Table 10. The three
SCC mix designs were the result of the mix development process presented in the
previous sections. The NCC mix design was the standard fabricator mix used for bridge
beams. These mixes were chosen for their ability to perform adequately in both the

plastic and hardened states.

Table 10. Prgjgct Mix Designs
SCC 1 SCC 2 SCC 3 NCC

 

 

 

 

 

 

 

 

 

 

 

Constituents (lb/yd3)
Cement 700 700 700 564
Fine Aggregate 1,591 1,513 1,320 1,354
Coarse Aggregate (17A) 1,350 1,350 1,450 1,883
Water 256 285 320 151
Air 6.00% 6.00% 6.00% 6.00%
W/c Ratio 0.35 0.4 0.46 0.27
Admixtures (oz/cwt)
Air Entrainm 1 1 1 1.9
HRWR 15 12 10.7 1 1.3
VMA l 2 6

 

3.2.4 Discussion

From the mix design process just described it can be seen that SCC mixes can be
produced using opposing methods. The mix designs developed for this project were
intended to be at the extremes of the various mix design methods and may not represent a
“good mix” to use in practice. This implies that the mixes used in this project could have
been improved to more easily satisfy production requirements such as finishing and

placing of the concrete. However, as an objective of this project was to investigate a

42

range of SCC mix design by developing and using mixes that lie at the extremes of
current mix design development practice, these mix designs were selected for their
research potential. SCC mixes that use a combination of these procedures and lie in the
middle of the range represented by the project mix designs may produce high quality
concrete with more consistency and ease.

As these mixes were developed, significant collaboration between the three
parties involved was necessary and extremely beneficial. The combined knowledge and
experience of the admixture, precast, and research representatives was highly valuable in
the development of mixes that satisfied both production and research objectives. Without
this collaboration these mix designs would not have been developed.

The ability to economically introduce SCC to local precast plants depends on the
feasibility of using locally available materials and equipment commonly available to the
plants. The mix design development process conducted for this research confirmed that
materials and equipment commonly used in local precast plants is acceptable for use in
the production of SCC. This includes many of the procedures common to local precast
plants, such as methods of concrete mixing and transportation of the concrete from the

mixing equipment to the casting area.

3.3 Strand Bond Evaluation

The ability to effectively use prestressed concrete depends on the quality of the
bond between the concrete and the prestressing strand. The length over which the strand
must be bonded in order for the concrete to achieve the full strand strength depends on
many factors including the arrangement of the strands, the condition of the strands, the

method of release of the strand, and the concrete. The length required to develop full

43

strand stress in the concrete, the transfer length and the development length, are defined
by long-standing equations in the AASHTO Specifications[2][3]. However, research has
shown that this equation may not be conservative for all sources of strand [31] or types of
concrete. Unconservative values lead to longer transfer and development lengths, which
manifest themselves as reduced flexural and shear strength of the prestressed concrete
beam near the ends. Since concrete tensile strength and density impact its ability to bond
to strand, there were concerns that this problem could be further complicated when using
SCC. The concern is that the mix design modifications that lead to the beneficial
properties of SCC may have a negative influence on the bond between the concrete and
the prestressing strand. This uncertainty led MDOT to include a requirement for the
evaluation of the strand bond for the SCC mix designs in the project special provisions
[11]. The PCI guidelines [23] suggest using a flexural development length test or a direct
load test to evaluate the bond between the prestressing stand and an SCC mix. While
several tests have been proposed for the purpose of evaluating the strand bond, the
modified Moustafa test proposed by Logan [23] and recommended by the PCI interim
guidelines on SCC was chosen for this project [31]. The use of direct structural tests to
determine the flexural development length was not considered for this project due to
constraints on production time and cost. The Large Block Pull-Out test (LBPT)
described by Logan [23] was chosen to evaluate strand bond because it requires only
small amounts of material and it’s simple to execute with equipment found at the precast

plant.

44

3.3.1 Large Block Pull-Out Test (LBPT) Overview

Logan analyzed the bond between concrete and prestressing strand from seven
different manufacturers to see if different strand manufacturing processes affected the
bond [31]. A second objective of Logan’s work was to set a benchmark for the pullout
capacity of the strand that could predict the transfer and development lengths. The
testing program included pull out tests conducted on unstressed samples of 0.5-inch
diameter prestressing strand embedded 18 inches in concrete and three-point bending
tests on prestressed beams to measure the required development length [31].

The development lengths Logan found by experiment were compared to the
development lengths predicted by the ACI-318 code equations [1]. It was found that
strands that exceeded an ultimate pullout strength of 36 kips in the large block pullout
test (LBPT) had a development length which was conservatively predicted by the ACI
code. However there were significant differences between the measured and predicted
development length when the pull out force did not exceed 12 kips [31]. From the results
it was determined that the minimum acceptable average pull out capacity for 0.5-in.
strand was 36 kips, the minimum acceptable average for the first slip of the strand during
the pull out test is 16 kips.

All of Logan’s test specimens were cast with a specific NCC mix. Thus, the
focus of his study was on evaluating the prestressing strand, and the concrete was not a
variable in the test. However, the modifications made to the SCC mix designs have
raised concerns about whether proper bond will develop when SCC mixes are used.

Specifically, whether the increased levels of HRWR and VMA in SCC mix designs

45

would lower the ability of the cement paste to develop adequate bond with the strand
[33].

The LBPT described by Logan included pullout tests on embedded unstressed
prestressing strand samples. The 0.5-inch strands were embedded 18 inches into concrete
blocks. The blocks had dimensions of 6 feet 8 inches long, by 2 feet deep, by 2 feet high.
Each block included 18 strands spaced at 8 inches in two rows 12 inches apart. A 2-inch
tube was included around each strand at the top surface of the block to debond the
concrete from the strand. To test the pull out strength, the strand were pulled vertically
from the block using a hydraulic ram bearing against the concrete block. The rate of
loading was 20 kips per minute. The tests were conducted on the morning after casting
[31].

Because of the relative newness of SCC, the issue of prestressing strand is a
widely discussed problem. Research using Logan’s LBPT has been done with SCC.
Morgan-Girgis and Tuan [33] analyzed the bond of 0.6 in. prestressing strand in SCC
using an approach similar to Logan’s. LBPT’s were conducted on the strand and
compared with results from beam tests to assess the transfer and development lengths.
Morgan-Girgis and Tuan found that lower bond between SCC and the strand was seen at
early age. This was shown with longer transfer lengths from beam tests, however the
same trend was not seen in the pull out tests. At 28—days, the SCC showed higher bond
strength than NCC in both the beam tests and the LBPT [33].

For this project the LBPT proposed by Logan was used to evaluate the relative
bond characteristics of the four different concrete mix designs with the prestressing

strand to be used in the bridge beams. Blocks were cast of each concrete mix design: 24

46

in. tall by 36 in. wide and 24 in. deep. Six 0.6 in. diameter unstressed prestressing strand
samples, with an ultimate capacity of 270 ksi, were embedded 18 inches in the blocks and
were arranged as shown in Figure 12, and schematically in Figure 13. The strands were
pulled out of the block 48 hours after casting using a hydraulic ram as shown in Figure
14. Values of first slip and ultimate load were recorded using the dial gage on the ram
control. The accuracy of the gage was limited to 500 lbs. For safety, the upper limit on
the jacking load used on each test was 47,000 lb, or 80% of the ultimate strand capacity.
An additional block was cast using a mix design that was as close as possible to

the one recommended by Logan. The suggested mix design is included in Table 11 while
the modified mix design used for this project, is in Table 12. Other than slight variations
on the mix quantities the modified mix was very similar to the suggested mix design.
This test served to qualify the strand, as it is the main purpose of the test. The rest of the
blocks were cast with the different project SCC and NCC mixes. These tests allowed a
relative evaluation of bond performance on the different types of concrete.

Two rounds of tests were conducted to verify the results. The test used in this project
differed from Logan’s test in the following ways:

I The diameter of the prestressing strands was 0.6 in. instead of 0.5 in.

I The rate of loading was estimated using a stop watch - there is possible error in

this value due to the manual control of the hydraulic ram
I The ultimate load of the test was set at 47,000 lbs or 80% of the ultimate strand

capacity due to safety

47

Table 11. Suggested LBPT Mix Design

 

 

 

 

 

 

 

 

 

 

LBPT Mix Design_
Constituents (per yd3)
Cement (lbs) 660
Fine Aggregate (lbs) 1,100
Coarse Aggregate (lbs) 1,900
Water (lbs) 292
Admixtures (oz/cwt)
Air Entraining 0
Normal Range Water Reducer 26
VMA 0

 

The test results were valuable even with these potential limitations. The PCI
guidelines state, “that the bond with SCC is equivalent or better than with a conventional
concrete of similar design when using similar strand [23]” suggesting a relative
performance rather than a direct comparison to Logan’s benchmark levels. It will be
discussed later, that results from this and other research projects indicate that the direct
pullout strength of strand on SCC is lower than on NCC. Recommendations have been
provided to PCI [19] to modify their recommendations to more realistic expectations as

discussed in Section 3.3.3.

48

Table 12. Modified LBPT Mix Design

 

 

 

 

 

 

 

 

 

 

Modified LBPT
Mix Design
Constituents (per yd3)
Cement (lbs) 660
Fine Aggregate (lbs) 1,031
Coarse Aggregate (lbs) 1,850
Water (lbs) 291
Admixtures (oz/cwt)
Air Entraining 0
Normal Range Water Reducer 0
VMA 0

 

The benchmark described by Logan was deveIOped for 0.5 in. diameter strand.
Logan has suggested a modified benchmark for the use of 0.6 in. diameter strand [31].
The modified benchmark uses a modified block geometry as well. Instead of the 18 in.
embedment used with the 0.5 in. diameter strand, Logan proposed a 20 in. embedment
with the 0.6 in. diameter strand. Maintaining the bond stress for the 0.5-inch. benchmark
at 1.27 ksi, the benchmark for the 0.6 in. strand becomes 48 kips. This benchmark
proposed by Logan for 0.6-inch diameter strands requires the same stress as the 0.5-inch
strand but increases the embedment length to balance the increase in diameter. Logan’s
modified recommendations for 0.6-inch strand were not available until after the tests
were complete, and the embedment length used in this project was 18 inch. The
benchmark level for the 0.6 in. strand with an 18 in. embedment, assuming the same
constant bond stress from the 0.5 in. strand of 1.27 ksi, can be calculated as 43.1 kips.
Similarly, the first slip level was altered such that the bond stress was maintained at a

consistent value. The 16 kip benchmark level represents a stress of 0.565 ksi with 0.5-

49

inch strand embedded in l8-inches of concrete. When this benchmark is modified for the

0.6—inch diameter strand with an embedment of 18 inches, it increases to 19.2 kips.

E
5
g
1::

 

Figure 12. Pull-out Test Specimen Prior to Testing

50

Embedded
/ 0.6" Diameter

Prestressing
Chuck T)// Strand

 

 

 

 

 

 

 

 

 

Hydraulic 3; 2" PVC Pipe
130‘" for Debonding

S 1 Pl Jacking End
tee ate \ F 2

l r— 7" l
11‘6"
Embedment
l]\ i! /ii 4’"
/4" PVC Pipe for
debonding

Figure 13. Schematic of Pull-Out Test Specimen

 

\ \ \\u. .

Figure 14. Test Set-up For Pull Out Test

51

3.3.2 Observations and Results

The ultimate load test is used to show the strength of the bond developed between
the strand and the concrete. It is used in this project to qualify the different concrete mix
designs. The results of both rounds of tests are shown in Table 13 and graphically in
Figure 15. For each test the concrete was 48 hours old. Each test was conducted 48
hours after the casting of the block. Each result is the average of 6 tests for each mix
design. The mix design emulating the one by Logan to verify the strand bond was used
with success here. The bond exceeded the 47,000 lbs limit of the test. This surpasses the
43.1 kip benchmark value that qualifies the strand for adequate bond. Logan’s concrete
mix was not used in the second round of tests (see Table 13). Once the strand is qualified
to have good bond quality, a relative evaluation of the effect of concrete type, i.e., SCC
versus NCC, can be made. The first movement values are shown in Table 14 and
graphically in Figure 16. Each result is the average of 6 tests for each mix design. The
requirements from the Special Provision state that the SCC strands must achieve 67% of
the bond strength of the NCC strands [11] (APPENDIX A).

From Table 13 it can be seen that the mix design used by Logan surpassed the
modified benchmark level of 43.1 kips. This means that the bond between the strand and
the concrete was adequate to achieve the development lengths predicted by the ACI-318
Building Code equations [1]. It is unclear whether this benchmark can be applied to the
other mix designs. Because of this uncertainty, only the relative performances of the four
project mix designs will be discussed without regard to the benchmark load levels

proposed by Logan.

52

In general, it can be seen that the values of ultimate load in the second test were
lower than the values of ultimate load obtained in the first test. Part of the reason for this
may have been inconsistencies in the test procedure in the first test. The hydraulic ram
used to test the Specimen was controlled manually and the loading rate was estimated
using a stopwatch. In the second tests the rate of loading was observed to be more
consistent and lower than in the first test. A higher loading rate may be the cause of the
higher ultimate loads shown in the first round of tests.

Regardless of the value, the relative performance of the four mix designs is
consistent in the first and second series of tests. For the SCC concrete mix designs, the
SCC 1 mix design had the highest ultimate load in both rounds of tests, the SCC 2 mix
design had the second highest load and the SCC 3 mix had the least bond strength of all
of the mixes. Comparing results from the more consistent second round of tests, the SCC
1 mix design performed better than the NCC mix design by 13 percent, while the SCC 2
and SCC 3 mixes showed ultimate loads 3 and 30 percent lower than the NCC mix,
respectively.

Table 13. Ultimate Loads from LBPT

 

 

 

 

 

 

Mix Ultimate Standard % Ultimate Standard %
Design Load Dev. difference Load Dev. difference

Test 1 (lbs) from Test 2 (lbs) from
(lbs) NCC (lbs) ' NCC

NCC 47,000 0 - 34,500 4,212 -
SCC 1 47,000 0 0 39,000 1,530 13.04
SCC 2 37,000 5,562 -21.28 33,500 4,212 -2.90
SCC 3 30,000 2,690 -36.17 24,000 3,216 -30.43

_Logan 47,000 0 - - - -—

 

53

 

50000

- First Test
:1 Second Test

40000‘

30000-

 

20000~

Ultimate Load (lbs)

10000—

 

 

 

 

 

NCC SCC 1 SCC 2 SCC 3

Figure 15. Ultimate Load of LBPT for all Mix Designs

From the data for the first slip load it can be seen that during the first round of
tests the SCC 1 and NCC mix designs had average values above the benchmark level of
19.2 kip. However, during the second round of tests, none of the concretes had average
values that met this benchmark. Again, it is unknown whether these benchmarks can be
used given the modifications made to the testing procedure. What can be seen from this
data is that the relative performance Of the four study mix designs for the first slip load is
similar to the relative performance of the ultimate load. In terms of the load at first slip,
the SCC 1 mix design had the highest average, but was only slightly higher than the NCC
mix design. SCC 1 also had the highest average of the three SCC mix designs. The SCC

3 mix design produced the lowest average value.

54

Table 14. First Slip Loads for LBPT

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Mix 1St Slip Standard % 1St Slip Standard %
Design Load Dev. difference Load Dev. difference
Test 1 (lbs) from Test 2 (lbs) from
(lbs) NCC (lbs) NCC
NCC 30,000 2,683 - 15,500 3,061 -
SCC 1 28,500 2,739 -5 15,000 1,291 -3.23
SCC 2 11,500 895 -61.67 13,500 681 -12.90
SCC 3 11,500 1,441 -61.67 7,500 612 -51.61
Logan 27,500 3,650 - - - -
25M»
- First Test
30000 .. Second Test
:wmm-
E
Z 20000 -
a:
3
E‘ 15000 4
m
“.3...
IMMOE
5000 ~

 

 

 

 

SCC 2

Figure 16. 1st Slip Loads for LBPT.

3.3.3 Discussion

The results of the pullout test may be related to the mix designs. Specifically, the
use of chemical admixtures in the concrete may affect the bond of the concrete with the
prestressing strand. The amount of VMA varies from 0 oz/cwt in the NCC mix to 6
oz/cwt in the SCC 3 mix. As the SCC 3 mix has the lowest bond strength, the VMA may

be responsible for the poor bond. This may be due to the fact that the SCC 1 mix had the

55

least amount of VMA of all three SCC mixes and it showed the highest bond strength of
the three mix designs.

The pull out tests served to confirm the relative performance of the different SCC
mix designs. While the pullout performance of the three SCC mixes was different,
testing that is outside the scope of this project would be required to fully understand why.
However, even with the range of results obtained in these tests, the size of the elements
cast for the bridge are large enough such that slightly increased transfer and development
lengths will not negatively affect the overall performance and safety of the bridge.

Other studies done on the bond performance of SCC using the pull out test have
found that SCC mixes have lower bond strengths than NCC. Burguefio and Haq [10] and
Morgan-Girgis and Tuan [33] have all shown reduced pullout strengths for SCC mix

designs. The reasons for these reduced strengths are not yet fully understood.

3.4 Mix Development and Performance Conclusions

The SCC mix design development and evaluation led to the following conclusions:

I Three SCC mix designs were developed to bound current mix design
development approaches. The mixes represent a range of acceptable SCC mix

designs.

I The mixes were developed through trial and error iterations using fresh property

test evaluations.

I The mixes were developed through the collaboration of representatives from the

admixtures company, the precast company, and the research group.

56

The mix design process Showed that locally available products and equipment

were suitable for producing SCC.

Fresh property tests including the slump flow, VSI, J-Ring, and L-Box were used

to develop the mix designs.

The fresh property tests also helped to assess and determine consistent quality

control of the mixes from batch to batch.

The LBP’T provided values that were used to relatively compare the bond between

the prestressin g strand and the different project concrete types.

SCC 1 had the highest pull-out bond strength, and SCC 3 had the lowest of the
three SCC mixes. However, these values were not considered critical for this

project.

57

4 PRODUCTION OF SCC PRESTRESSED BOX BEAMS

4.1 Technical Considerations

SCC is considered an acceptable alternative to NCC in some projects because of
its superior performance in the fresh or plastic state. Generally, SCC is used in projects
with dense reinforcing schemes or complicated formwork. In these cases it is often
difficult to consolidate the concrete, making SCC an alternative. The use of SCC can
also make some casting processes faster and less complicated.

The production of box beams Often used for highway bridge construction is one
situation where SCC is considered a Viable option. The cross section of the box beam
contains a Styrofoam void that blocks out the center of the beam over specified lengths.
The void is included to block out concrete where it does not add much resistance to the
section and reduce the self—weight of the beam. For practical reasons, the void in the
beam cross-section is Obtained by a stay-in-place Styrofoam block. The creation of the
void through the Styrofoam block makes casting the beam both time and labor intensive,
as the process requires several steps. The use of SCC could reduce these steps.
However, implementing SCC is not as easy as simple as just changing the concrete. One
main challenge is to determine the pressure generated by the flowable concrete onto the
formwork. For the production of box beams it further implied holding the Styrofoam
block in the correct position as the concrete is poured into the formwork. The Styrofoam
block must be restrained against the force of the rising concrete and the restraints should
not affect the finished process.

During the production of the beams for this project these tests offered important

information on the consistency and quality control of the concrete. All three tests were

58

conducted on the first SCC batch of the day. The results of these tests determined
whether that the SCC batch just mixed was adequate for placement in the beams.
Subsequent batches were subjected only to the inverted slump flow test to verify the flow
parameters of the concrete. If substantial differences in the slump flow value were
observed from batch to batch throughout the day, adjustments were made to the mix
designs to bring the flow into an acceptable range.

Current precast/prestressed industry practice for producing box beams requires
four stages, making it both time and labor intensive. This process includes putting the
prestressing strand, bottom stirrups, and metallic straps (ties to secure the Styrofoam
block) in place on the casting bed during the first stage. The top reinforcing steel cage,
including stirrups and longitudinal reinforcement, are assembled and set aside. Figure 17
shows a casting bed that was prepared to cast an NCC beam for this project. In the
second stage the bottom flange of the beam is cast. As the concrete is poured into the
form it is vibrated with mechanical Vibrators. The level Of the concrete in the bottom
flange is checked to ensure that proper thickness of the bottom flange was achieved.
Figure 18 shows the bottom flange of the NCC beam being poured. Once the bottom
flange is poured, consolidated, and leveled, the Styrofoam block and the pre-tied top-
reinforcing cage are secured in place during the third stage of the production process.
Figure 19 shows the concrete poured in the bottom flange with the void in place. This
step must be done as fast as possible to ensure that the concrete in the bottom flange
remains plastic so that a homogeneous joint can be made upon casting the webs of the
box section. Figure 20 shows the amount of activity required for the crew to secure the

pre-tied top portion of the reinforcing steel. In the fourth and final stage of the process

59

the webs, or sidewalls, and top flange are cast. During the fourth stage, the concrete is
poured directly on top of the Styrofoam block and manually distributed throughout the
formwork. This process is shown in Figure 21. Sufficient Vibration is used in the
sidewalls to ensure that a cold joint does not develop between the bottom flange and the

sidewalls.

 

n

 

-_

Figure 17. NCC Casting Bed Prior to Casting

60

 

r ,4

 

Figure 19. The Bottom Flange of The NCC Beam

61

 

Figure 20. Installing Top Reinforcement On NCC

 

 

Top Flange and Side Walls In NCC Beam

Figure 21. Casting

62

One advantage of using SCC to cast the box beams is the elimination some of the
labor-intensive steps that occur during the conventional casting process with NCC. With
SCC the box beam production process can be reduced to two steps. In the first step for
producing the SCC box beams for this project, all of the reinforcing steel and Styrofoam
blocks were put in place. Plastic spacers were used to ensure the correct vertical and
horizontal location of the Styrofoam block, and were left inside the concrete. The void
was tied to the prestressing strands in the bottom flange of the beam with metallic straps
to ensure that it did not rise with the rising level Of concrete. Figure 22 shows an SCC

mock-up beam with all reinforcement and voids in place before casting.

 

Figure 22. SCC Beam Ready to Be Cast

63

 

Figure 23. Rising Concrete in SCC Beam

In the second stage of the production process using SCC, the concrete is placed in
the forms. This is done in one step without any vibration. The high flowability of SCC
allows the concrete to flow around obstacles in the formwork. Figure 23 shows the
concrete as it flows around the voids and reinforcement.

There are multiple ways to place the SCC in the formwork. The location where
the concrete is poured into the form can play a role in the performance of SCC. One
option is to pour the concrete at one end of the form and allow it to flow along the length
of the beam. If the concrete does not have sufficient fluidity it would not be able to
successfully flow along the entire length of the beam and the pouring location would
have to be moved. A second option is to pour the concrete between the void and the side
of the formwork. When added to the form in this manner, the concrete is able to flow

under the void and begin to flow up the other side of the formwork. The discharge point

can be moved along the length of the beam so that the concrete can fill the entire form
evenly. Figure 24 shows the concrete being discharged into the sidewall of the beam
formwork. After the concrete reaches the top level of the formwork, the concrete can be
poured on top of the void, and the concrete can flow along the void to cast the top flange.

A production plan was developed to ensure that SCC was suitable for the
production of box beams. This process included the production of mock-up beams
similar to the beams being cast for the bridge construction. The following sections
discuss this production process, including results and observations from the casting

Operations.

 

Figure 24. Casting SCC Beam

65

4.1.1 Specimen Mock-up

To ensure that box beams could be produced in reduced steps using SCC, a mock
up trial was conducted on May 16th and 18‘h of 2005. The purpose of the mock ups was
two fold: (1) the process of filling the formwork with fewer steps needed to be verified,
ensuring that the Styrofoam void could be properly restrained; and (2) experience needed
to be gained from batching continuous SCC to ensure the ability to consistently produce
high quality SCC.

One mock-up beam was cast for each of the four project mix designs, SCC 1,
SCC 2, SCC 3, and NCC. The plan and cross-section of the beams are Shown in Figure
25 and Figure 26. Each mock—up beam had a length of 20’ and cross section dimensions
of 27” x 36”. The bottom and side flanges were 4.5” thick and the top flange was 5”
thick. Five 0.6” diameter prestressing strands reinforced the bottom flanges of the beams.
Each strand was prestressed to 43.8 kips. Five #4 reinforcing bars spaced at six inches
were included in the top flange of the beam and U-shaped #4 stirrups were spaced evenly
along the length of the beam in both the top and bottom flanges for shear reinforcement.
One 14-inch diaphragm was located at the center of each beam and a 2-foot solid end
block was cast at each end. The design 28-day compressive strength of the concrete was

5,500 psi, with a minimum of 4,000 psi required to release the strands.

66

 

20' Beam Length ——»—1.______-_..
_ 7'-5" Void
Length Typ.

F __________ ] lfi—T'TTT—m—‘T
l l l l

L.________-____ l________._-____l

2' End Block
—- 1'"-2 Dia hra m
l‘——" Typ. *l l" P 8

Figure 25. Mock-Up Beam Plan

 

 

 

 

 

 

 

During the construction Of the mock-up beams, concrete was mixed in 2 yd3
batches. The Inverted Spread Flow, J-Ring, and L-Box tests were conducted on the first
batch from each mix design to verify the quality of the SCC used in the mock-ups.
Subsequent batches were tested only with the Inverted Slump Flow to verify that the SCC
was being produced consistently. The results of the initial tests are presented in Table 15.
The concrete was transported from the batching area to the casting bed using one
discharge bucket carried on a fork truck. The concrete was placed in the forms by
discharging it from the bucket at multiple locations along the forms. The location and
speed of the discharge were varied throughout the production of the mock-up beams to

determine the effect this would have on the filling ability of the SCC.

67

-r<————————————-———— 36" ——————————~——————-—

5 #4 Bars @ 6" I

 

 

 

 

 

 

FF .. . . . .. WWI

/ \

y— #4 Stirrup

/ TandB

\ "4‘4" /
x J
f C U
I I l J
L 5 - 0.6" Diameter Strands .1
6" Spacing = 2'

Figure 26. Mock-up Cross Section

 

 

 

 

27H

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 15. Mock-Up Fresh Property Test Results
SCC 1 SCC 2 SCC 3

 

 

 

 

 

 

 

 

Inverted Spread Flow Test
Spread (in.) 26 28.5 27
VSI 0.5 1.0 0.5
J-Ring Test
Spread (in.) 21.75 27.5 23
J-Ring Value 0.75 0.375 1
L-Box Test
L-Box 0.56 0.56 0.66

 

4.1.2 NCC Observations

The process described above used to cast the NCC mock-up beam took an hour or
more to complete. However, it is important to note that some of this time was spent
waiting for concrete to be delivered from the mixing area to the casting bed. In one

instance a batch of concrete was thrown away as it did not meet quality control standards,

68

significantly adding to the total casting time. Still the casting process with NCC is
cumbersome and difficult, as it requires significant amounts of labor and time. By
reducing the number of steps in the process, the production cost and throughput can be

improved.

4.1.3 SCC Observations

4.1.3.1 SCC 1 Observations

To cast the SCC 1 mock-up beam, the concrete was poured along the side of the
beam between the edge of the form and the Styrofoam block. The fluidity of the concrete
allowed it to flow under the void and rise up on the opposite side. The concrete flowed
along the length of the beam and around all obstacles. The discharge location was moved
to several locations along the beam. Like the NCC mock-up beam, this process took
approximately one hour to complete. Again significant time was spent waiting for
concrete to be transported to the casting area. In addition, two SCC 1 batches were
discarded due to poor test performances. The number of workers helping with the casting
process was reduced by half compared to the NCC beam.

After the casting of this beam was completed, it was discovered that the
Styrofoam void was not adequately restrained. The hydraulic force from the concrete
filling the form pushed the Styrofoam void against the steel straps that were supposed to
hold it in place. The force was large enough to cause the Styrofoam to crush locally at
the straps allowing the void to rise 1” or more along the entire length of the beam. Even
though the void rose significantly, a flat top surface was achieved on the beam as can be

seen in Figure 27 .

69

 

Figure 27. Finished SCC 1 Mock-Up Beam

4.1.3.2 SCC 2 Observations

The process of casting the SCC 2 mock-up beam was altered slightly to assess the
effect of discharge location on the ability of the concrete to fill the formwork. For the
SCC 2 beam, the concrete was discharged into the end block to see if the concrete could
flow the entire length of the beam. The concrete was successful in flowing along the
beam length as is evidenced in Figure 28, where the edge of the concrete flow can be
seen emerging from under the void.

After the first batch of concrete was discharged into the form, the concrete self-
leveled along the length of the beam. This showed that the SCC 2 concrete was able to
flow well along the length of the beam. The second batch of SCC 2 was discharged into

the diaphragm at the center of the beam. The rate of the discharge was increased to see

70

the effect of the discharge rate on the flow properties of the concrete. The resulting
hydraulic force caused the Styrofoam block to rise significantly. As the block rose, the
reinforcing cage was lifted to such a degree that a flat top surface could not be achieved
on the beam. The reinforcing steel can be seen above the top edge of the formwork in
Figure 29 and Figure 30. The number of workers required to cast the SCC 2 beam was
comparable to the number of workers used in the SCC 1 beam. The time that was spent
in casting the SCC 2 beam was also comparable to the SCC l and NCC beams. Again

time was spent waiting for the concrete to be delivered to the casting bed.

   

Edge of w
Concrete Flow ;

  

 

 

     

Figure 28. SCC Flowing to End of Formwork

71

a

 

ing Steel

Inforc

ith Exposed Re

ished SCC 2 Beam W

in

F

gure 29.

F1

 

Figure 30. Result of Void Movement in SCC 2 Beam

72

4.1.3.3 SCC 3 Observations

During the casting of the SCC 3 mock-up beam care was taken to prevent any
movement of the Styrofoam block and the reinforcing steel. The concrete was discharged
along the side of the void as was done previously in the SCC 1 beam. The rate of
discharge was also minimized to reduce the force of the flowing concrete on the block.
In the end, even with the precautionary measures, the Styrofoam block raised similarly to
the SCC l beam. A flat top surface was obtained on the beam as can be seen in Figure
31. However, it is clear that the reinforcement had been vertically shifted due to the
uplift of the Styrofoam block. The casting time of the SCC 3 beam was lower than the

previous beams as the concrete was mixed more efficiently.

 

Figure 31. Finished SCC 3 Mock-Up Beam

73

4.1.4 Discussion

While the production of the mock-up beams confirmed that SCC was able to flow
along the length of the beam and around all of the obstacles in the formwork, the
performance was hampered because of the inability to properly restrain the Styrofoam
block. The reduction in time that was expected from the reduced steps required to cast
the SCC beams was not realized as any time improvements were counteracted by the
increased time to mix and test the SCC. Inefficiencies in the mixing process caused
down time at the casting bed. However, a reduced labor force was used to cast the SCC
beams. Approximately half of the workers used to cast the NCC beams were needed to
cast the SCC beams. Another result observed in the production of the mock-up beams
was the improved surface quality that was evident in the SCC beams when compared to
the NCC beam. The self-consolidating nature of SCC allowed for a better and more
complete compaction of the concrete, which was evident by the presence of fewer and
smaller deformities on the concrete surface, or “bug holes.” A discussion of the

conclusions reached during the mock—up beam production follows.

4.1.4.1 Styrofoam Void Hold Downs

As a result of the failed attempt to properly restrain the Styrofoam void in any of
the three SCC beams an additional SCC 2 mock-up beam was produced two days after
the initial attempt. The concrete used in the additional mock-up beam had a spread flow
of 25 in., and a VSI rating of 0.5. The remaining fresh property values were not provided
by the fabricator. To further restrain the void in this additional mock-up, wood restraints
were constructed and placed along the length of the beam as shown in Figure 32. The

restraints were bolted to the Side metal formwork to keep them in the proper location as

74

shown in Figure 33. A full view of the Styrofoam block restraints is shown in Figure 34.
The restraints consisted of two 2 in. x 4 in. cross beams separated by a 2 in. x 2 in.
spacer. Four 2 in. x 2 in. legs extended from the cross beams to hold the block at the
proper height. A 1/8-inch masonite plate was placed under the restraints to prevent the
legs of the restraint from punching through the Styrofoam block. The masonite plate
remained embedded in the concrete after the concrete cured. Once the concrete was in
place and before it began to set the restraints were removed, as shown in Figure 35.
Figure 36 shows the scar left on the surface of the beam after the restraint was removed.
This scar was smoothed over by traditional finishing of the beams.

The additional restraints were successful in keeping the Styrofoam block in place
during the casting of the final mock-up beam. No movement of the block was observed
after the concrete was placed. Irnportantly, the concrete was able to flow through the legs
of the restraints without blocking. The finished product, seen in Figure 37, shows the

result of the improved production process.

75

 

Figure 32. Void Restraints Spaced Along Beam Length

 

Figure 33. Void Restraint Bolted to Formwork

76

.
w
.w

L
m

..e=~g—>t

 

Figure 34. Close up of Void Restraint

 

Figure 35. Removing Void Restraint From Finished Beam

77

 

t

ain

d Restr

. Scar On Surface of Beam frOm Vo

 

raints

Figure 37. Finished SCC 2 Beam Cast With Additional Void Rest

78

4.1.4.2 Surface Finish Quality

The improved surface quality of SCC is evidenced by a reduction in the size and
number of visible surface imperfections. The ability of the concrete to consolidate under
its own weight removes many of the surface imperfections that are common with NCC.
The SCC mock-up beams showed fewer and smaller “bug holes” on the surface than seen
on the NCC mock-up. Figure 38, Figure 39, and Figure 40 show this clearly. While
some marks were present on the surface of the SCC l specimen they are not as numerous,
or as large, as those on the NCC beam. The SCC 2 beam had better appearance than the
SCC 1 beam, with even fewer surface imperfections. The SCC 3 beam (not pictured)

also had an improved surface quality compared to the NCC beam.

 

Figure 38. SCC 1 Surface Quality

79

 

Figure 40. NCC Surface Quality

80

4.1.4.3 Mixing and Batching Process

Even though the production of a box beam using SCC requires only two steps, the
time to cast the SCC beams was equal to that required for the NCC beams. Much of the
time that it took to cast the beams was spent waiting for concrete to be delivered to the
casting bed from the mixing area because concrete was not being mixed efficiently. A
new batch of concrete was not started until the bucket returned from the casting bed, so
nothing could happen at the casting bed while the concrete was being mixed and tested.
If the fresh performance of the concrete was adequate then the batch would be taken to
the casting bed. If it was not adequate, modifications were made to the concrete and a
second round of tests was conducted. It was learned that the use of two discharge
buckets carried by two fork trucks and continuous mixing of concrete would allow
concrete to be poured into the formwork continuously thus eliminating the time where the
crew at the casting bed was waiting for concrete. As one discharge bucket was being
discharged at the casting bed the second would be being filled at the mixing plant. This
batch could be tested and the fork truck could take the next batch to the casting bed as the
first bucket was returned to receive the next batch. This continuous flow could save time
in the process. Given the experimental nature of the mock-up beams, the time required to
cast the SCC beams was reasonable, however this time did not reflect the reduced cycle
times possible when using SCC.

In the casting of all three SCC mock-up beams, some batches of the SCC had to
be thrown away as they were improperly mixed. To have an efficient flow of concrete
from the mixing area to the casting bed, the mixes need to be of consistent quality.

Significant time was wasted as batches had to be altered to achieve acceptable fresh

81

property performance. On a given day it may take several attempts to produce consistent
concrete as the mix design is modified for factors such as moisture content of the
aggregate and humidity. These same modifications are made to NCC with greater
success as the concrete producer has more experience working with NCC. With time and
experience concrete producers can produce SCC of a consistent quality without

significant added time for adjusting batch properties.

4.1.4.4 Strand Draw —in Measurements

When the prestressing force is transferred to the box beams the prestressing
strands retreat into the beams a small amount. The amount the strands move into the
beam is controlled by the bond strength between the prestressing stand and the concrete.
In an attempt to quantify the bond strength through this “draw-in” value, measurements
of an arbitrary length were taken on selected strands before and after the release of the
prestressing force. The measurement was taken between a reference point on the strand
and the face of the beam using a digital caliper. The reference point was determined by
connecting a 1 in. wide section of a steel channel to the prestressing strand with hose
clamps. The measurement was taken from the leg of the channel farthest from the face of
the beam.

The measurements from the mock-up beams are shown in Table 16. In this table
negative values have been removed. Negative values would imply that the distance
between the face of the beam and the reference point got larger after transferring the
prestressing force to the beam. This is not possible as the strand draws into the beam. It
is more likely that the reference point moved. The process of transferring the

prestressing force to the beam by fire cutting the strand causes a very violent expansion

82

of the strand. This expansion is so great that it sometimes plastically deformed the hose
clamps. After the completion of the mock—up beams the measurement of the strand draw
in using steel channel and hose clamps was deemed unreliable. A second method of
creating a reference point directly on the prestressing strand by means of a gouge or
notch was ruled out for safety reasons. Therefore, these measurements were not taken

during the production of the project beams.

Table 16. Strand Draw-in Measurements

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

End A End B
Mix Strand Draw-in ALE Std Dev Strand Draw-in Avg. Std Dev
1 0.0060 1 0.0538
SCC 1 2 - 0.0060 - 2 0.0332 0.0357 0.017
3 - 3 0.0200
1 0.0762 1 -
SCC 2 2 - 0.0762 - 2 _ - -
3 - 3 -
1 0.0137 1 -
SCC 3 2 0.1465 0.0538 0.0805 2 0.0090 0.0070 0.0028
3 0.0013 3 0.0050
1 - 1 0.0267
NCC 2 - 0.0310 - 2 0.0310 0.0375 0.0151
3 0.0310 3 0.0548
1 - 1 -
SCC 2-2 2 0.0158 0.0172 0.0019 2 0.0032 0.0283 0.0354
3 0.0185 3 0.0533
4.2 Beam Production

4.2.1 Overview
Once mock-up beams were produced with satisfactory results, the actual bridge
beams could be made. Four casts were needed to produce the 17 beams. All of the

beams were cast in late June and early July of 2005. At each of the first three castings,

83

five beams were cast in a 300-foot casting bed. At these times, three of the beams cast
were of one SCC mix design type and two were of the NCC mix design. A fourth cast
was conducted to complete the final two NCC beams.

For each concrete mix design three beams were cast to play a specific role in this
project. One of these beams was instrumented to monitor strain and temperature in the
demonstration bridge; this beam is referred to as the field beam. Two of the beams were
instrumented for laboratory evaluations. The laboratory evaluations consisted of two four
point bending tests to evaluate the shear and flexural performance of the beams. These
beams are referred to as the shear beam and flexure beam respectively. Two additional
bridge beams, were cast using NCC and were included in the bridge but not instrumented.
Finally three extra beams were cast using NCC. These beams were available to replace
any SCC beam in the demonstration bridge that did not perform satisfactorily in

laboratory evaluations. Figure 41 shows the casting sequence for the four casting days.

84

Cast #1 (6-21-05)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

}+——~——— —~~———» —— 300' Casting Bed =11
l SCC 1 SCC 1 SCC 1 NCC NCC
L Flexure Beam Shear Beam Field Beam Flexure Beam Shear Beam
l: 5 Beams @ 52' Tl
Cast #2 (6-24-05)
*—— 300' Casting Bed
SCC 2 SCC 2 SCC 2 NCC NCC
Shear Beam Flexure Beam Field Beam Field Beam Bridge Beam
5 Beams @ 52 +7
Cast #3 (6-29-05)
300' Casting Bed
SCC 3 SCC 3 SCC 3 NCC NCC
Shear Beam Flexure Beam Field Beam Bridge Beam Extra Beam
L 5 Beams @ 52' J
1‘ ’1
Cast #4 (7-1-05)
300' Casting Bed
NCC NCC
Extra Beam Extra Beam

 

 

 

 

 

 

-~— 2 Beams @ 52' .__._,

Figure 41. Casting History

4.2.2 Beam Geometry

85

 

The box beams were 36 in. wide by 27 in. deep. The top flange was 5 in. thick
while the bottom and side flanges were 4.5 in. thick. The Styrofoam void cross section
was 27 in. wide by 17.5 in. deep. Each beam was 52 ft long with a 2-ft solid block on

each end. An 8-in. wide diaphragm was located at the center of each beam. The

Styrofoam void between the end block and the diaphragm is 23 ft - 8 inches long. The
beam plan and cross section are shown in Figure 42 and Figure 43. The beams had a non
composite concrete area of 509.4 square in. and a moment of inertia of 47,223 in4. The
neutral axis for the beams was found to be 13.6 in. from the bottom of the beam, or 13.4
in. from the top of the beam. The composite beam, with the deck considered, had a
concrete area of 1,373 square in. and a moment of inertia of 155,789 in4. The neutral axis
of the beam with the deck moves to 24.86 in. from the bottom of the beam. This
information is summarized in Table 17.

Table 17. Untransformed Section Properties

 

 

 

Time Step Area I (in4) y (in.)
(mi)

Non Composite 509.40 47,223 13.60

Composite 1,373 155,789 24.86

 

The beam’s main reinforcement consisted of 16 - 0.6” diameter prestressing
strands arranged in three rows in the bottom flange of the beam as shown in Figure 43.
Two of the strands in the bottom row were debonded 4 ft from the end of the beam. An
additional two 0.6” diameter prestressing stands were included in the top flange of the
beams. These strands were debonded 13 ft in either direction from the center of the
beam. The prestress jacking force for all 18 of the strands was 43,900 pounds. Other
reinforcing steel consisted of 5 - #4 bars in the top flange with six-inch spacing, and #4
U-shaped shear stirrups in the top and bottom of the beam. The stirrups were spaced in
the following way: two spaces at 2.25 in., 12 spaces at 6 in. then 12 in. spaces. The shear
stirrups are not shown in Figure 43. All 17 beams were identical in size and reinforcing
details. The only exception was that the beams produced for laboratory testing did not

include slab ties

86

 

 

 

 

 

_______________________________________________________________________________________________________________________________

1‘ 52' Beam Length Li
23'-8" 8» 2.
It fl'— Void Length Typ. r Diaphragm End Block Typ. ‘11

________________________________________________________________________________________________________________________________

Figure 42. Beam Plan
-—~ 36" +1

L 4" Typ. 2" Typ- --
- 5 #4 Bars @ 6"

0.6" 5" -‘
Strand Typ. 4.. x 3.. x 3..

 

 

 

 

 

 

 

  

 

 

 

 

 

Each Side A Block Out
in Typ. Each Side
41/2n
. O
\ __ 4V2" /
1’ . o 0
c o O E] I ’ E] ' . .
L 4

 

 

 

 

 

 

 

 

16 - 0.6" Strands

[E Debonded 4' from end
©Debonded 13' in each direction
from the center of the beam

Figure 43. Typical Beam Cross Section

4.2.3 Casting Process

The casting of these beams required extensive collaboration between the MSU
research team and the fabricator’s work crew. This was due in part to the instrumentation
that was installed in the beams and to the use SCC. Two days of work were required

before the actual casting could take place.

87

The work prior to casting included preparing the formwork at the casting bed and
arranging the prestressing strand. An initial force Of 5,000 lbs was placed on the strands
to pull them taught. The MSU research team then applied the foil strain gages to the
strands in the necessary locations. Then the strands were loaded to the required
prestressing force. With the strands in place the fabricator’s crew could begin attaching
the reinforcing cages. In the case of the two NCC beams, only the bottom stirrups and
metallic void ties were put in place. For the three SCC beams the entire reinforcing cage
and voids were placed in the beds. During this process the MSU research team installed
the instruments in the field beams. Once the instruments were installed, the instrument
wires were bundled together in a debonding Strip, run along the length of the beam and
out the south bulkhead of the formwork. Foil strain gage wires from the laboratory test
beams were routed up the nearest stirrup and through a piece of debonding strip at the top
face of the beam.

In all four casts, the NCC beams were cast first. SCC is more sensitive than NCC
to the moisture content of both the mixing drum and aggregate. By mixing NCC in the
mixer before SCC, the mixer was “primed”. The 2 cubic yard batches of concrete were
transported to the casting bed in discharge buckets carried by fork trucks. Instead of
using one bucket and fork truck in the casting process as was done during the production
of mock—up beams, two buckets were used in the production of the project beams. This
was done so that there was no delay at the beds waiting for concrete. As one bucket was
being emptied into the formwork the second bucket was being filled at the mixing plant.

For the SCC beams, the concrete was poured along one edge of the beam

formwork between the Styrofoam void and the formwork. The SCC was able to flow

88

under the void in all directions and began to come up the other side of the formwork.
The dumping location moved down the beam until approximately half of the beam was
poured. Once this was accomplished the concrete was dumped on top of the void starting

at the farthest end.

4.2.4 First Cast (6/21/05)

The beams cast on the first day included the two NCC laboratory evaluation
beams, and the three SCC 1 beams. All of the instruments were put in place before the
morning of the cast. Initial readings from the instruments were recorded on the morning
of the cast. The entire casting process took approximately 5 hours to complete. During
this time the temperature ranged from the 70 °F at the start of the cast to the 85 °F at the
end of the cast. For the entire day the average temperature was 71 0F with a maximum
temperature of 85 0F and an average relative humidity of 71%. The day was mostly

cloudy.

4.2.4.1 NCC Beams

In total, eight — 2 yd3 and one — 1.5 yd3 batches of NCC were produced for a total
of 17.5 yd3. The actual and target average values for the NCC mix design are shown in
Table 18. These values are shown for information only as a representation of all of the
concrete mixes produced during this cast. This is not the mix design that was used to cast
this beam. Any significant deviations from the expected mix design were made by the

fabricator without the knowledge of the researchers. The specific values for each of the 9

NCC batches are listed in APPENDIX B.

89

The first NCC beam casting began at 8:30 am. The bottom flange was finished 15
minutes later at 8:45 am. The top reinforcing steel cage placement began at 8:50 am.
The sides and top flange of the beam began at 9:20 am and the beam was finished at 9:50
am, 1 hour and 20 minutes after the casting began.

The second NCC beam casting began at 9:45 am and the bottom flange was
finished at 10:00 am. The void and top reinforcing steel were in place in the second NCC
beam by 10:30 am and the beam was finished by 11:00 am. The second NCC beam took

1 hour and 15 minutes to cast. These times are summarized in Table 19.

Table 18. Average NCC Mix Design

 

Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 566
Fine Aggregate 1353 1347
Coarse Aggregate 1877 1882
Total water 219 218
w/c 0.39 0.39
Admixture (oz/yd3)
Air Entraining 11 10
VMA 20 20
HRWR 64 63

 

Table 19. NCC Beam Casting Time

 

 

 

 

 

 

Event NCC-1 NCC-2
Start of Casting Bottom Flange 8:30 am 9:45 am
End of Casting bottom flange 8:45 am 10:00 am
End of placing voids and top reinforcement 9:20 am 10:30 am
End of castiflg beam 9:50 am 11:00 am
Total Time 1 hr 20 min 1 hr 15 min

 

90

The average time between NCC batches during the first casting was 18 minutes
19 seconds and the average mixing time was 5 minutes 20 seconds. The time between
batches is representative of how much time was spent without concrete being cast. This
average includes relatively short times when concrete is being poured in the forms, but it
also includes time Spent putting the void and reinforcement in place. From Table 19 it
can be seen that the time between the end of the casting of the bottom flange and the end
of placing the voids and top reinforcement was 35 and 30 minutes. During this time no
concrete was being poured into the formwork. If this time becomes too long the concrete
poured into the bottom flange could begin to set. If this occurs, a cold joint could form
between the sidewall and bottom flange.

During the casting of the NCC beams anywhere from 10 to 18 members of the
fabricator’s work crew were working on the beam. The greatest work force was required
in the second stage of the casting process when the reinforcing cage had to be completed.

This process is shown in Figure 20.

4.2.4.2 SCC 1 Beams

During the casting of the SCC 1 beams, 13 - 2 yd3 and one — l yd3 batches were
produced for a total of 27 yd3. The average actual and target values for the SCC 1 mix
design are shown in Table 20. These values are Shown for information only as a
representation of all of the concrete mixes produced during this cast. This is not the mix
design that was used to cast this beam. Any significant deviations from the expected mix
design were made by the fabricator without the knowledge of the researchers. For the
first SCC 1 batch of the day, the Slump Flow, J-Ring, and L-Box tests were all

conducted. These values, along with the air content, unit weight, and concrete

91

temperature are reported in Table 21. For the remaining batches, only the Slump Flow
test was conducted. The results of these tests are shown in Table 22. The average Slump
Flow value was over 26 inches within the 27 inch plus or minus 1 inch range Specified in
the project special provisions. The mix designs for all of the SCC l batches produced
during the first cast are listed in APPENDIX C.

Table 20. Average SCC 1 Mix Design
Tagget Value Actual Value

 

 

 

 

 

 

 

 

 

 

 

Constituents Obs/yd“)
Cement 700 705
Fine Aggregate 1656 1651
Coarse Aggregate 1350 1357
Total water 260 262
w/c 0.37 0.37
Admixture (oz/yd3)
Air Entrainin g 12 12
VMA 21 21
HRWR 105 105

 

Table 21. SCC 1 Fresh Properties

 

 

 

 

 

 

Test SCC 1
Slump Flow (in.) 27
J-Ring Flow 26
J-Ring Value 0.75
L-Box Ratio 0.79
Air (%) 6.58

 

Conc Temp (F) 77.7
Unit Wt (lb/ydT) 142

 

92

Table 22. SCC l Slump Flow Values

 

Average Slump

 

 

 

 

 

 

 

 

 

 

 

Bate“ Flow (in.)
1 27
2 26.75
3 26.5
4 25.75
5 27.75
6 26.5
7 28.5
8 28.25
9 23.5
10 25.75

Average 26.6

 

The time it took to cast the three SCC 1 beams is shown in Table 23. The beams
took an average of 31 minutes to cast. This is compared to over 1 hour 15 minutes in the
NCC beams. The average time between SCC 1 batches was approximately 8 minutes.
This is much faster than the 18 minutes between the NCC batches. The average mixing
time for the SCC l was batches 6 minutes 51 seconds, which was approximately 1 minute
longer than for the NCC batches.

Not only is the total time to produce the batches significantly reduced, the number
of people required to cast these beams was also significantly reduced. There were only
three people required to cast the SCC 1 beams. Two people were needed to drive the two
fork trucks carrying the discharge buckets from the batching plant to the casting bed. The

third person was required only to discharge the concrete into the formwork.

93

Table 23. SSCC 1 Beam Casting Times
SCC 1-1 SCC 1-2 SCC 1-3

 

 

 

 

Start 11:50 am 12:30 pm 12:54 pm
End 12:30 pm 12:50 pm 1:27pm
Total Time (min) 40 20 33

 

4.2.5 Second Cast (6/24/05)

The second cast occurred on the 24'h of June, three days after the first cast. The
beams cast during this time included two NCC bridge beams including the instrumented
demonstration beam, and the three SCC 2 beams. As in the first cast, all of the
instruments were put in place before the morning of the cast. Initial readings from the
instruments were recorded on the morning of the cast.

The weather for the second cast was very hot. During the approximately 5-hour
casting period the temperature ranged from approximately 75 °F at 6:00 am to
approximately 85 °F at 11 am. The average temperature for the day was 83 °F with a
maximum temperature of 92 °F. The average relative humidity was 58% with sunny

skies throughout the day.

4.2.5.1 NCC

A total of 16.6 yd3 of NCC were produced for the second cast. The average
values from the eight 2 yd3 and one 0.6 yd3 batches are shown in Table 24. These values
are Shown for information only as a representation of all of the concrete mixes produced
during this cast. This is not the mix design that was used to cast this beam. Any
significant deviations from the expected mix design were made by the fabricator without

the knowledge of the researchers. The values for each batch are shown in APPENDD( D.

94

The time it took to cast each beam is shown in Table 25. The total time spent
producing the NCC beams was 2 hours 31 minutes. The average casting time for both of
the NCC beams was approximately 1 hour 15 minutes. The first NCC beams took 59
minutes to cast. It took approximately 20 minutes to complete each of the steps in the
casting process. This included 20 minutes to pour the concrete for the bottom flange, 20
minutes to install and secure the voids and top reinforcing steel, and 20 minutes to pour
the remaining concrete in the top flange and side wall. The second NCC beam took 1
hour 32 minutes to cast due to time spent routing instrument cables in the second beam.
Once the top reinforcing steel was secured in place, the instrument cables had to be
routed from their location near the center of the beam to the end of the beam. This took
additional time that was not included in casting the other NCC beams. The average
mixing time for the NCC batches was 10 minutes 40 seconds with an average of more

than 19 minutes between each batch.

Table 24. Average NCC Mix Design

 

Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/ydj)
Cement 564 57 1
Fine Aggregate 1352 1368
Coarse Aggregate 1883 1889
Total water 221 221
w/c 0.39 0.39
Adnrixture (oz/yd3)
Air Entraining 1 1 10
VMA 14 13
HRWR 64 64

 

95

Table 25. NCC Beam Casting Times

 

 

 

 

 

 

Event 1st NCC Beam 2“(1 NCC Beam
Start casting beam 6: 10 am 7: 10 am
Finish casting bottom flange 6:30 am 7:17 am
Finish placing voids and top reinforcement 6:50 am 7:24 am
Finish casting beam 7:09 am 8:42 am
Total Time 59 min 1 hr 32 min

 

4.2.5.2 SCC 2

A total of 24 yd3 of SCC 2 were produced in 12 — 2 yd3 batches to cast all three of
the SCC 2 beams. The average target and actual values for the SCC 2 mix design are
shown in Table 26 with all of the batches being listed in APPENDIX B. These values are
shown for information only as a representation of all of the concrete mixes produced
during this cast. This is not the mix design that was used to cast this beam. Any
significant deviations from the expected mix design were made by the fabricator without
the knowledge of the researchers. From the mix design values, it can be seen that no
major deviations from the target values occurred. SCC 2 was a combination type SCC,
which means that the mix had a moderate amount of HRWR and VMA, a higher amount

of fine aggregate, and a reduced amount of coarse aggregate when compared to NCC.

96

Table 26. Average SCC 2 Mix Design

 

Target Value Actual Value

Constituents (lbs/yd3)

 

 

 

 

 

 

 

 

Cement 700 709

Fine Aggregate 1488 1483

Coarse Aggregate 1438 1442

Total water 274 276

w/c 0.39 0.39

Admixture (oz/yd3)

Air Entraining 12 11
VMA 21 21
HRWR 84 84

 

The results of the fresh property tests conducted on the first SCC 2 batch are
shown in Table 27. Slump flow values from 9 other batches (as measured by the
fabricator) are shown in Table 28. The average slump flow from these 9 tests was 27
inches. This value shows that SCC could be produced within a reasonable range of the
target slump flow. The 27—inch J—Ring flow indicates that the SCC 2 concrete had
excellent passing ability. There was no change between the obstructed flow and the free

flow. Values of air content, concrete temperature, and unit weight compare well with

NCC.

Table 27. SCC 2 Fresh Properties

 

 

 

 

 

 

 

 

Test SCC 2
Slump Flow (in.) 27
J-Ring Flow 27
J-Ring Value 0.25
L—Box Ratio 0.64
Air (%) 6.8
Conc Temp (F) 77.7
Unit Wt (lb/yd3) 144

 

97

Table 28. SCC 2 Slump Flow Values

 

Average Slump

 

 

 

 

 

 

 

 

 

 

33“" Flow (in.)

l 28.25
2 28
3 27
4 27.“75
5 25.5
6 27
7 27.5
8 26
9 26

Average 27

 

The times required to cast the SCC 2 beams are shown in Table 29. The total
time to cast all three beams was 1 hour 16 minutes. On average it took 25 minutes to cast
each of the three SCC 2 beams. The average time between SCC 2 batches being
produced was 5 minutes 19 seconds. The SCC 2 batches mixed for an average of 3

minutes 42 seconds.

Table 29. SCC 2 Beam Casting Times
SCC 2-1 SCC 2-2 SCC 2-3

 

 

 

Start 9:45 am 10:18 pm 10:36 pm
End 10:18 pm 10:33 pm 11:04 pm
Total Time (min) 33 15 28

 

4.2.6 Third Cast (6/29/05)

The third cast occurred on the 29'h of June, five days after the second cast
(including two weekend days). The beams cast during this time included the final two
uninstrumented NCC bridge beams, and the three SCC 3 beams. As in the first and
second casts, all of the instruments were put in place before the morning of the cast.

Initial readings from the instruments were recorded on the morning of the cast.

98

The weather for the third cast was not as hot as the second cast. The temperature
was between 70 °F and 80 °F during the 5-hour casting period from 5:30 am to 10:30 am.
The average temperature for the entire day was 80 °F and the maximum temperature was
90 °F. The average relative humidity was 69% and the sky was mostly cloudy during the

day.

4.2.6.1 NCC

A total of 18 yd3 of NCC were produced for the third cast. The mix design, which
is nearly identical to the NCC mix designs used in the previous two casts, is Shown in
Table 30. These values are shown for information only as a representation of all of the
concrete mixes produced during this cast. This is not the mix design that was used to cast
this beam. Any significant deviations from the expected mix design were made by the
fabricator without the knowledge of the researchers. . From each mix design listed in
APPENDIX F it is possible to see that some mixes contained no VMA while others had
amounts consistent with that used in the previous casts.

Table 30. Average NCC Mix Design

Targt Value Actual Value
Constituents (lbs/yd3)

 

 

 

 

 

 

 

 

 

 

Cement 564 567
Fine Aggregate 1354 1349
Coarse Aggregate 1885 1888
Total water 220 221
w/c 0.39 0.39
Admixture (oz/yd3)
Air Entrainin g 9 8
VMA 5 5
HRWR 64 64

 

99

The times required to cast each beam were not recorded for the third cast. From
the information provided by fabricator regarding the production of concrete batches it can
be seen that concrete was produced from 5:30 am to 7:45 am. This is consistent with the
two previous casting operations. In cast one and two, the production of the NCC beams
took approximately 2 hours and 30 minutes to complete. The average time between
batches was 17 minutes 43 seconds and the average mixing time was 2 minutes 49

seconds.

4.2.6.2 SCC 3

The total volume of SCC 3 produced for the third cast was 28 yd3. This was done
in 14-2 yd3 batches. The average target and actual mix designs used for all 14 batches are
shown in Table 31. These values are shown for information only as a representation of
all of the concrete mixes produced during this cast. This is not the mix design that was
used to cast this beam. Any significant deviations from the expected mix design were
made by the fabricator without the knowledge of the researchers. The amounts used in
all 14 batches are included in APPENDIX G.

The results of the fresh property tests on the first SCC 3 batch are shown in Table
32. The 23.25 inch slump flow value is below the target of 27 inches. From Table 33 the
average slump flow was 23.45 inches. The fabricator had difficulty in producing
concrete at the high flow value desired. However, the concrete was still able to flow
around the obstacles in the formwork. A 23-inch slump flow was sufficient to fill the
forms for this project. Even though the concrete did not meet the requirements for this
project it was selected because it was able to adequately fill the formwork without

sacrificing the quality of the product.

100

Table 31. Average SCC 3 Mix Design

 

Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 709
Fine Aggregate 1374 1369
Coarse Aggregate 1458 1461
Total water 3 10 3 10
W/c 0.44 0.44
Admixture (oz/yds)
Air Entraining 8 8
VMA 21 21
HRWR 56 55

 

Table 32. SCC 3 Fresh Properties

Test SCC 3
Slump Flow (in.) 23.25
J-Ring Flow 22.25

 

 

 

 

 

 

J-Ring Value 1
L-Box Ratio 0.56
Air (%) 6.3
Conc Temp (F) 80

 

Unit Wt (lb/ydj) 140.72

 

The total time to cast all three SCC beams was 1 hour 40 minutes. The average
time for the three beams was 33 minutes. The average mixing time was 2 minutes 45
seconds and on average 6 minutes 41 seconds elapsed between each batch being

produced.

101

Table 33. SCC 3 Slump Flow Values

 

Slump
Flow (in.)
22.5
23.75
23.25
23
23.25
23.25
23.75
23.75
25
23

Average 23.45

Batch

 

 

 

 

 

 

 

 

 

Exoooqotut-hwto—

4.2.7 Fourth Cast (7/1/05)

The fourth and final cast occurred on the 15' of July. The beams cast during this
time included the two additional NCC bridge beams that served as back ups in case any
of the SCC mix designs beams did not perform satisfactorily in the laboratory
evaluations. No instrumentation was included in these beams.

The weather during the fourth cast was much milder than the first three. The
casting took 3 hours and began at 5:30 am. The temperature did not vary significantly
during the day as the temperature remained in the 70’s for the duration of the cast. The
average temperature on this day was 64 degrees with a maximum temperature of 74
degrees. There were scattered clouds in the sky during the casting and the average

relative humidity for the day was 67%.

4.2.7.1 NCC
A total of 16.6 yd3 of NCC was produced in six - 2 yd3, four - 1yd3, and one - 0.6

yd3 and batches. The average values of the target and actual mix designs are shown in

102

Table 34. These values are shown for information only as a representation of all of the
concrete mixes produced during this cast. This is not the mix design that was used to cast
this beam. Any significant deviations from the expected mix design were made by the
fabricator without the knowledge of the researchers. The target and actual values for

each batch are listed in APPENDIX H.

Table 34. Average NCC Mix Design

 

Target Value Actual Value
Constituents (lbs/yd3)

 

 

 

 

 

 

 

 

Cement 564 568
Fine Aggregate 1354 1349
Coarse Aggregate 1888 1883
Total water 219 219
w/c 0.39 0.39
Admixture (oz/yd3)
Air Entraining 7 7
VMA 6 6
HRWR 64 63

 

While specific times were not recorded regarding the casting of each of the two
NCC beams, concrete was produced from 5:30 am to 8:30 am. This implies that the NCC
beams took 1.5 hours to complete on average. The average mixing time for the concrete
batches was 4 minutes 25 seconds while there was an average of 19 minutes 57 seconds

between batches.

4.2.8 Beam Production Discussion

The beam casting times show that SCC requires significantly less production time

compared to NCC. Table 35 summarizes the production times for all of the casting

103

operations. Time savings of 50 - 70% are seen when using SCC. The average time
savings was 61% for all three SCC mixes. This time savings is a direct result of being
able to cast the SCC beams in one step without any Vibration required to consolidate the
concrete.

Table 35. Production Time Savings Using SCC

 

Total Time SCC Time Average Average Time

Cast Concrete (mi n/beam) Reduction M1x1ng Time Between Batches

 

 

 

 

(min) (min)

1 is. 7;: is
as. :2 ‘39.?

3 as. 6;: :33.
4 N CC 60 - 4.42 19.95

 

The average time between batches was less than 10 minutes for the three SCC
mixes during beam production - SCC was being produced almost constantly during the
production process. The average mixing time results Show that the majority of the time
between pouring batches of SCC was spent mixing the next batch, while for NCC there
would be down time for the production crew while they waited for the batch to be mixed.
This is a more efficient use of labor as the people producing the concrete work more
constantly when producing SCC than when they produce NCC.

Not only does the use of SCC save time, it also resulted in a savings on labor
during the casting operations. For the NCC beams anywhere from 10-18 people were
working on the beam at a given time. The largest number of people was required when
the reinforcing steel and voids were being installed after the bottom flange had been

poured. Casting the SCC beams required only 4 people.

104

The time savings seen when casting with SCC are for the day of casting only.
Some of the time saved during the day of casting is canceled out due to the increased
amount of time spent preparing to cast. However, part of the reason the time between
each casting was so long was because of tasks related to the research project. Time had
to be spent installing instruments as well as taking instrument readings. It is unclear how
much time these research tasks added to the overall preparation. As production
experience is gained in the use of SCC, the actual time savings can be defined with more
precision. These values will vary depending on the project size and type.

The release of the SCC 1 beams was delayed because adequate strength was not
obtained by the time the fabricator’s crew would normally release the strands. It took an
additional hour to reach the required strength compared to the normal practice. This was
the only concrete that did not reach the required strength before the normal time to
release the strands. Adjustments to the SCC 1 mix design could result in a more rapid
early strength gain.

The constant quality control monitoring during the casting process allowed for
changes to be made to the mix design to adjust for the conditions at the time. This Step
was important, as the SCC mixes were more sensitive to small changes in factors such as

the ambient temperature and moisture content of the aggregate.

4.3 Material Properties

The properties of the materials used in this project are important in the testing and
analysis work for this research. Concrete material properties were measured at various
times throughout the project. These were taken at times when they were needed for the

project. The two properties that were measured were the compressive strength and the

105

tensile strength of the concrete. The properties of the prestressing steel were given by the

strand manufacturer.

4.3.1 Compressive Test

The concrete compressive strength, f’c, was measured by testing 4 in. by 8 in.
cylinders in compression according to ASTM Specification C39-05 [46] on four
occasions. The first test was conducted before the prestressing strands were released. A
minimum of 5,000 psi was required to transfer the prestressing force. The next two tests
were conducted in conjunction with the structural tests, one each for the flexural and
shear tests. The final value was taken in April of 2006 in conjunction with other material
property tests not included in this project.

The cylinders tested in compression were moist cured for 24 hours and then
placed in a curing room until they were tested. The cylinders tested in compression were
capped with Sulfur to reduce friction between the steel testing apparatus and the concrete
cylinder. This friction can introduce tension into the test resulting in a reduced
compressive strength.

The results of the tests for the four concrete mix designs are shown in Table 36
through Table 39, and in Figure 44. Each value represents an average of at least three
tests. The standard deviations of the multiple test values are shown in the tables. The
first value in each table was provided by the fabricator with out a standard deviation, thus
none is provided in the table.

The general trends followed by the compressive strength gain are as would be
expected. All mixes, except the SCC 3 mix, showed significant strength gains in the

beginning of the project. The majority of the strength gains for the NCC, SCC 1, and

106

SCC 2 concretes occurred by the time the flexural tests were conducted approximately 40
days after the concretes were produced. The SCC 3 concrete showed a similar rate of
strength gain throughout the entire length of the project. One anomaly is in the reduced
strength seen for the third strength test conducted on the SCC 1 concrete. It is unclear
why this reduction occurred. However, similar results were obtained in all three
cylinders tested as can be seen by the relatively low standard deviation value. By the
time the fourth test was conducted, the SCC 1 concrete showed the highest strength of all

the mixes.

Table 36. NCC Compressive Strength
Age F'c Std. Dev

 

 

 

 

 

 

(days) (psi) (psi) Em“
0.79 6,207 - transfer of prestressing force
41 8,196 437 flexure test
55 8,560 134 shear test
287 9,106 435 April 2006

 

Table 37. SCC 1 Compressive Strength
Age F’c Std. Dev

 

 

 

 

 

 

(days) (psi) (psi) Em“
0.88 6,338 - transfer of prestressing force
43 8,290 300 flexure test
58 7,519 158 shear test
310 10,133 937 April 2006

 

Table 38. SCC 2 Compressive Strength
Age F 'c Std. Dev

 

 

 

 

 

 

(days) (psi) (psi) Event
3 6,572 - transfer of prestressing force
42 6,754 290 flexure test
66 7,71 1 430 shear test
298 7,980 296 April 2006

 

107

Table 39. SCC 3 Compressive Strength
Age F'c Std. Dev

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

. . Event
((12178) (28!) (pm)
0.79 5,062 - transfer of prestressing force
40 6,685 236 flexure test
54 6,953 1 18 shear test
288 7,556 163 April 2006
11000 _
E. 100003
. i
22. .
9 9000 -.
a -(
g 8000 . a
'IT) 1 -
m
s i
Q 7000 n
E i
8 n
'8' i + see
8 5000 ' --— SCCI
O I + SCC 2
j +scca
4000fi...,..fir....,....,....,.........
o 50 100 150 200 250 300 350
Concrete Age (days)

Figure 44. Concrete Compressive Strength

4.3.2 Tensile Strength
Split tensile tests were conducted according to ASTM Standard C496-04 [47] at
the time of the shear test. The results of the test are shown in Table 40. All values are

the results of at least three tests except for NCC, which is the result of only two tests.

108

The results of these tests follow the predicted trend. The strongest concrete in
compression at this point in time was the NCC concrete. It follows logically that it would
also have the highest tensile strength. The SCC 3 compressive strength was the lowest as
is the case for the tensile strength. The SCC 1 and SCC 2 mixes had similar compressive
strengths and also show similar tensile strengths.

Table 40. Split Tensile Test Results

 

Age (days) f’.(‘psi) Std. Dev. (psi)

 

 

 

 

NCC 55 646 105
SCC 1 58 609 56
SCC 2 66 576 38
SCC 3 54 514 73

 

4.3.3 Prestressing Strand Properties

The properties of the prestressing steel are listed in Table 41. Two spools of
strand were used to produce the beams used in this project. The properties listed include
the area of the strand, modulus of elasticity of the strand, and the yield and ultimate

strengths of the strand.

Table 41. Prestressigg Steel Properties

 

 

 

 

 

Spool No. 9003098001 Spool No 9003085202
Strand Area (inz) 0.2268 0.2279
Modulus of Elasticity (psi) 29,030,000 28,940,000
Yield Stress (psi) 242,292.769 245,936.814
Ultimate Stressfi(psi) 266,463.845 265,427.819

 

4.4 Production Conclusions
The production of the box beams for the demonstration project allowed for the
evaluation of several SCC mixes in an actual production setting. From this experience

the following conclusions could be drawn:

109

The use of SCC in the production of box beams reduced the production time

an average of 61% compared to using NCC.

The number of people required to produce the SCC beams was lower than the
number to produce NCC beams. The NCC production crew was 15 workers

compared to a maximum of five for the production of the SCC beams.

In order to successfully produce box beams using SCC some modifications to
the production process were required. Specifically, the use of void hold

downs were required to maintain the desired cross section dimensions.

Due to the improved flow properties of SCC, surface quality of box beams

was improved compared to the NCC beams.

Because the beam casting process was so much faster with SCC, concrete
batches needed to be produced nearly continuously. With NCC, the slower

casting process meant that there was significant down time between batches.

The constant quality control during the production of the SCC batches
allowed for necessary adjustments to maintain consistent fresh property

behavior throughout the beam production process.

110

5 FLEXURAL AND SHEAR PERFORMANCE OF SCC AND NCC
PRESTRESSED BOX BEAMS

5.1 Testing Program Overview

Two test beams were produced for each concrete mix design. One of the beams
was used in an evaluation of the flexural capacity, and the other was evaluated for shear
capacity. The beams were replicas of the beams to be used in the M-50/US-127 Bridge,
(B02 of 38071) specifically: 52-ft long 27 in. by 36 in. prestressed box beams. A typical
cross section of the box beam through the Styrofoam void with reinforcement details was
previously Shown in Figure 43. The test beams were not provided with slab ties since the
beams were not going to be provided with a Slab and the inclusion of the slab ties would
have complicated the application of loads in the test setup.

Evaluation of the beam’s flexural capacity was assessed through a four-point
bending flexural test. For the test, the beams were simply supported with a span of 50-ft
and loaded at two points spaced 8 ft from each other. Shear capacity was also evaluated
through a four-point bending test but with a shortened supported span. Loading
conditions in the shear test were also at two points 8-ft apart and centered about the
beam. For the shear tests, the shear span for the NCC beam was 8 ft — 7 in., while the
shear span for the SCC beams was 6 ft — 7 in.. The change in shear span was made after
testing the NCC beam to increase the shear force demands in the section. The shear test
setups had overhanging cantilevers at both ends of the supported span. Further details on
the test setup and instrumentation are given later. The information herein is part of a
research report on the experimental evaluation of the beams for SCC demonstration

bridge [6].

111

5.2 Theoretical Considerations

The calculations of design shear and flexural capacities of prestressed concrete beams
involved well-known approaches. The formulation of these approaches is thoroughly
described in the literature and building codes, for example Naaman [34] describes the
ACI building code 318 formulations in his textbook, while AASHTO presents two
methods in the 2“d Edition LRFD Specification and in the 17th Edition of the Standard
Specifications [2][3]. Common among these formulations are the simplifying
assumptions that [34]:

I Plane sections remain plane
I Strain is linearly distributed across the cross section
I Concrete and steel experience perfect bond.

The design shear and flexural capacities for the test beams were calculated using
the equations from the 17‘h Edition of the American Association of State Highway and
Transportation Officials (AASHTO) Bridge Standard Specifications [2] since these were
the guidelines used by MDOT for the design of the M-50/US-127 Bridge. The shear
capacities were also analyzed using the 2nd Edition of the AASHTO-LRFD Bridge
Design Specifications [3] since the model for shear resistance in this design code are
more comprehensive and realistic. For the analyses, the voided cross section of the box
beams was simplified to an I-beam whose web thickness was the width of the two box
sidewalls. This section was then analyzed as a flanged section given that the neutral axis

of the cross section lies below the depth of the top flange of the beam.

112

5.2.1 Flexural Capacity

The moment resistance of a simply supported beam loaded in the direction of
gravity is found from the internal couple between the compressive force in the top flange
of the beam and the tensile force in the bottom flange. The compressive force acting at
the top of the beam is primarily carried by the concrete in the top flange, while the tensile
forces at the bottom of the beam is taken by the reinforcing steel, either prestressed or
non prestressed. In the design of concrete structures the tensile strength of concrete is
neglected such that the entire tensile force is carried by the reinforcing steel. If the
section is under-reinforced such that sudden compression failures are avoided, the
flexural capacity of the section is assumed to be reached when the reinforcing steel in the
tension flange yields. Additional assumptions necessary to produce the commonly
accepted simplified formulas for flexural capacity include the use of the Whitney
equivalent stress block. This idealized compressive stress distribution approximates the
compressive stress distribution as a rectangle. Improvements on this assumption could be
made through the use of a parabolic or trapezoidal stress distribution.

The moment resistance can be defined by defining section moment equilibrium
about either the compressive force in the concrete or the tensile force in the steel. As the
flexural capacity of the beam is determined to be the point when the tensile steel yields,
the maximum compressive force in the concrete bust be equal to this tensile force. Due
to the internal couple the maximum compressive force must equal the force in the
reinforcement at yield, this is given in the ACI code for a flanged cross section is given as

Equation (2):

(2) 0.8511(1) - bw)hf + 0.85 fgbwa = Ap,f,,, +A,fy

113

where b is the width of the flange, bw is the width of the web, hf is the thickness of the
flange, a is the depth of the equivalent compressive stress block, Aps is the area of the
prestressing steel, fps is the stress in the prestressing strand at ultimate load, As is the
area of the non prestressed steel in the tension zone, and fy is the yield strength of the non
prestressed steel. The factor of 0.85 in equation (2) is used to correct for the assumed
stress distribution (Whitney stress block). The moment resistance of the section can be

obtained from the compressive force in the section by Equation (3):
(3) M, = 0.85 fc' (b — 1),, )1. f [d — 92) + 0.85fc'b,,,a[d — g) (in- 1b).

Where d is the distance from the extreme compression fiber to the centroid of the
prestressing steel. The moment arm of this couple is the distance between the location of
the compressive force and the tensile force. The compressive force acts at the centroid of
the assumed compressive stress distribution while the tensile force acts at the centroid of
the prestressing steel. In equation (3) this distance is given as the difference between the
depth to the centroid of the prestressing steel, d, and half the depth of the assumed

compressive stress block, a. These variables are shown in Figure 45.

114

f " 7' i
53—- ? / ” —i

 

 

 

 

 

 

 

 

 

 

 

T I 6% l

Figure 45. Moment Resistance Variable Definitions

In equation (3) the values of Mn and a are unknown. The depth a can be found
from equation (2) assuming that the value off}, is known. In the ACI 318-05 Building

Code [1] an estimate for fin, is given for members with bonded tendons as Equation (4):

(4) fp. =fp..[i-0.5pp£j§.—] (psi).

C

with fin, equal to the ultimate stress of the prestressing strand, and pp is given by equation

(5):

 

Thus, the only unknown value, a, can be directly solved for from equation (2) as given by

Equation (6):

115

a _ Apsfps + Asfy TO-85fcl(b _bw)hf

(6) .
085be

 

(in.).

For the case where there is no non-prestressed steel included in the tensile flange of the
cross section the values pertaining to this steel can be set to zero. To find the equations

for a rectangular cross section the value b,,. can be taken equal to 1).

5.2.1.1 17'h Edition AASHTO Standard Specifications F lexural Capacity

The AASHTO Standard [2] presents the following equations given that the beam
designed for this project has no non-prestressed steel in the tensile flange and behaves as
a flanged beam.

The average steel stress at the ultimate load is given by Equation (7):

(7) f3. =fi[1-(7A][p f‘frfl (psi).

where f", is the ultimate stress of the prestressing steel, y" is a factor depending on the
type of prestressing steel, it is taken as 0.28 for the low-relaxation steel used in this
project, ,6] is a factor that depends on concrete strength, and p* is the ratio of prestressing
steel. The value for f,,,* cannot exceed fig. which is the yield stress of the prestressing
steel. ,6, depends on the compressive strength of the concrete and can be taken as 0.85 if
f, is less than 4000 psi or 0.65 if f”, is greater than 8000 psi. If fl. is between 4000 and
8000 psi ,6) is to be obtained by using Equation (8):

_fe'_

8 = 1.05 - 0.05 .
( ) .51 1000

116

The moment resistance of the section is given by Equation (9):

a:
ASI‘ .fSIl

H + 0.8511: (b - bw )1. f (d — 0.51. ) (in-lb),
bwdfc

(9) Mn : Asr f; [I _ 0'6[

where A" is the area of the prestressing steel, and h is the overall height of the member.

5.2.1.2 2“d Edition AASHTO LRFD Specification Flexural Capacity
The AASHTO-LRFD Specifications [3] recommends using a different set of
equations to determine the flexural capacity of prestressed sections. The average stress in

the prestressing steel is estimated using Equation (10):

(10) fps =fp..[1—kdi] (psi).

P

where c is distance from the extreme compression fiber to the neutral axis of the beam

and is given by Equation (1 l):

 

C _ Awfp. + A..):y - Agf; —0.85/i.f.1(b—bw )hf

. f
0.85fcfllbw + kAp, d”“

(11)

(in.).

 

p

and k is given by the Equation (12):

(12) k = 2[1.04— f’” ].

pll

 

The moment resistance is given by Equation (13):

117

.. . . h
(13) Mn = Apsfps[dp —%)—Asfs(ds —%]+0.85f, (1)—1),, ”WE—BL] , in lb

where d,- is the depth of the compressive reinforcement and a is again the depth of the

equivalent stress block given by Equation:

(14) a=cfl1 (in.)

5.2.2 Shear Capacity

Determining the shear resistance of a reinforced or prestressed concrete section is
more complicated than determining flexural resistance. Despite the numerous research
efforts and the many proposed approaches, the problem of shear in concrete beams is still
being debated and no theory has been developed that applies to all types of concrete
structures [34]. Still, through many years of study, applicable (under a set of assumptions
and limitations) theories for the behavior of reinforced concrete beams in shear have been
developed.

Shear stresses cause cracks to develop in concrete in two ways: flexural-shear
cracks and web-shear cracks. Both of these crack types occur when principal tensile
stresses in the concrete exceed its cracking strength. In general, cracks develop
perpendicular to the direction of the principal compressive stresses.

Flexural-shear cracks develop from a combination of flexural and shear loading.
Flexural cracks develop in a concrete beams normal to the beam longitudinal axis. If the
diagonal tension at the crack tip reaches the tensile strength of the concrete, the direction
of the crack will begin to change to an incline. The development of these inclined cracks

could then lead to failure of the beam (shear failure) before the full flexural resistance has

118

been developed. These types of cracks can also lead to secondary cracks that occur along
the longitudinal reinforcement causing a loss of bond that can lead to a failure in the
anchorage zone of the beam [34].

Web-shear cracks occur in beams with narrow webs. These cracks develop when
the principal tensile stresses in a region reach the concrete’s tensile strength before the
flexural stress. The failures associated with web-shear cracking are Similar to those in
beams with flexural-shear cracks.

Prestressed concrete behaves differently in shear than reinforced concrete. The
diagonal tension stress in the concrete beam is reduced by the compressive prestressing
force. This force also leads to a smaller angle of inclination of the cracks that develop
due to diagonal tension. This consequently reduces the required number and size of shear
reinforcing steel [34].

Transverse, or shear, reinforcement is thus provided in reinforced concrete beams
to ensure that the beam fails in flexure. Flexural failures are more ductile then shear
failures and are the preferred method of failure. Once a beam is cracked in shear its
ability to carry tensile stress is significantly reduced. The inclusion of steel transverse to
the longitudinal axis of the beam allows additional tensile stress to be carried by the steel.
A beam is said to be at its shear capacity once the concrete has cracked and the steel has
yielded [34]. Sufficient reinforcing steel is included to ensure that the flexural capacity

of the beam is achieved before the shear capacity.

5.2.2.1 17th Edition AASHTO Standard Specification Shear Capacity Equations
The approach used in the of the AASTHO Standard Specifications [2] to calculate the

shear capacity of a prestressed concrete beam are essentially the same as that

119

recommended by the ACI-318 Building Code [1]. The nominal shear strength of a
prestressed concrete beam is given by Equation (15):

(15) v,, =VC+VS (lbs),

where V,. is the Shear strength provided by the concrete and V, is the shear strength
provided by the steel. The component V. is taken to be the lesser of the force required to
cause flexural-shear cracks or web-shear cracking. The AASHTO Standard
Specifications defines Equation (16) to approximate the shear required to cause web-

shear cracking.

(16) VCW=(3.5 fc'+0.:ifpc)b'al+vp (lbs)

where fin. is the compressive stress in the concrete at the extreme fiber due to the effective
prestress force, b’ is the width of the beam web or flange, d is the depth from the extreme
compressive fiber to the centroid of the prestressing force, and Vp is the vertical
component of the prestressing force. For beams with straight tendon profiles only, as is

the case for the beams in this project, Vp is neglected.

Equation (17) below is used in the AASHTO Standard Specifications [2] to
predict the shear at which flexural-shear cracking will occur. The development of this
equation is complicated by the fact that the shear distribution in a cracked beam is not
known.

. . V-M
(17) Vc,=0.6 fcbd+Vd+fi'—£’- (lbs)

max

120

In Equation (17) Va. is the shear force due to the dead load, V.- iS the factored shear force
due to externally applied loads, Mr, is the cracking moment as calculated with Equation
(18) below, and Mm, is the maximum factored moment at a section due to the externally

applied loads.
1 l - .
(18) Mcr :7[6 fc +fpc—fd) (m‘le)
t

In Equation (18) I is the moment of inertia if the cross section, Y, is the distance from the
centroidal axis of the cross section to the extreme tension fiber, and fd is the stress due to
the unfactored dead load.

The shear strength provided by the steel reinforcement is given by Equation (19)

Av g d
(19) v5 =—L'— (lbs),
S

where A, is the area of shear reinforcement, and s is the spacing of the shear

reinforcement.

The method presented by the Standard Specifications [2] to calculate the shear
resistance of a reinforced concrete beam is often conservative [13]. First, it assumes that
the cracked concrete carries no tensile stress. A second assumption is that the angle of
crack inclination in a reinforced concrete beam is 45 degrees. As cracked concrete
carries a small amount of tensile stress, and cracks in prestressed concrete beams are
inclined at angles less than 45 degrees, the shear resistance of a reinforced concrete beam

is often higher than that predicted by the ACI or Standard Specifications method.

121

5.2.2.2 2“d Edition AASHTO LRF D Specification Shear Capacity Equations

The shear equations presented in the AASHTO-LRFD Specifications [3] are
based on the modified compression field theory (MCFT). This theory improves the shear
resistance calculation for a reinforced concrete beam by accounting for a reduced angle
of crack inclination and the amount of tensile stress carried by the cracked concrete. The
theory was developed by Collins and Vechio and is presented in reference [13].

The procedure required to implement the MCFT is not as straight forward as the
ACI method since it requires iterations to find the answer. The general expression for
shear resistance in Equation (15) still applies. However, the shear strength provided by

the concrete is now given by Equation (20):
(20) v, = 0.03166 fc'bwd, (kips),

where ,6 is a factor that indicates the diagonally cracked concretes ability to transmit
tensile stresses. The shear resistance of the transverse steel reinforcement is given by

Equation (21)

_ Avfydv cott9

S

(21) V.

 

(kips),

where Bis the angle of inclination of diagonal stresses. The values of ,6 and Bare found
analytically. It can be seen, that under the format of Equations (20) and (21) the ACI
method conservatively assumes ,6 and Hto have values of 2 and 45 degrees, respectively.
However, the MCFT provides a way to obtain more realistic values for these parameters.
The AASHTO LRFD specifications adopt a simplified method to determine the values of

,6 and (9 according to the MCFT. The method is still iterative, but it is simplified by using

122

an average axial strain value rather than a more sophisticated sectional analysis. The
simplified procedure begins with an estimated value of the longitudinal strain, 6}, in the
beam element. Using this value and the ratio of the average shear stress (Oh) to the

compressive strength of the concrete (1" c), values of ,6 and Bean be read from Table 42,
which is taken from the AASHTO LRFD Specifications [3]. The value of strain must
then be re-calculated to see if the assumed starting value was correct. The longitudinal
strain can be calculated using Equation (22):

% + 0.5/v“ + 0.5(V,, — vp )eot e — A p, fpo

(22) 8 X = " . ) ,

2(E3A, +EpAp,

 

where M“ is the factored moment, N“ is the factored axial load, and V“ is the factored
shear force, applied by external loads. If the calculated value and the chosen value of the
axial strain do not match, a second estimation of 8,. is made and the process is repeated.

When the estimated and calculated strains are equal, the shear capacity can be calculated

using Equation (15), (20), and (21).

Several assumptions are made to calculate the shear capacity using this method.
First the concrete shear stresses are assumed to be uniformly distributed over the shear
area of the beam. Secondly, the angle of the principal compressive stress is held constant
over the shear area. Finally, that the shear strength of the section can be determined from
one location in the web. Nonetheless, the method provides a considerable improvement
in determining realistic shear capacities while permitting the evaluation of the complete
shear force-deformation response, something that is not possible with the limit-state

equations in the ACI method. Furthermore, the MCFT method explicitly takes into

123

account the interaction of bending and shear forces in the resulting effective strain in the
beam as given in Equation (22). This equation, together with equilibrium of along a
beam section along a flexural crack can be used to formulate the interaction between
flexure and moment capacities of a reinforced concrete section [13]. This is discussed

further in the next section.

124

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MNA MMA OmA MmA DNA MaA NAN MMN MMN QMN oMN OONd W
NAAe OMM M.MM MAM M.NM MAM 0.0M WON AdN O.MN mSN
mMA AmA VOA MBA OMA VAN VMN OVN NVN mvN MmN MNNd W
v.34 M.MM N.MM 044M M.NM MdM 06M M.MN GSN MSN AON
54A AOA MBA VMA VAN NM.N MVN AMN NM.N amN MON OONd W
NA? NdM A.OM WVM MNM 0.0M odN VSN NON A.ON NAN
VOA ABA OoA VAN MN.N 3N NM.N CON mON OO.N MN.N MbAd W
N.Nv hdM M.OM N.MM hNM N.MN o.MN NON M.MN NAN N.MN
AOA NMA MON ANN OMN NM.N OON NWN MN.N MN.N MMN omAd W
M.Nv mdv MSM @AuM ANM M.MN oON OMN NAN M.MN OAN
bOA OQA MAN ONN NVN NON «SN NM.N voN amN MAM MNAd W
N.Mv 0.:a 9.5M véM VAM QHN QMN N.MN M.NN @AN mdA
oOA MmA MAN NM.N OWN MN.N AQN VAM VNM MM.M GNM 236 W
AMAa M.OAa bOM QVM MdM ASN MAN mNN . VAN TON AMA
hOA mmA MN.N MMN amN VON VNM mnM 0A.»V when NM.O Whod W
M.MN Md? ROM N.MM mdM 0.0N MAN MAN OAN eVON MNN
8N w 02 w 2: w who w one w 2.0 w as .o w o w 8.? w 2 .ot w 8.? w .ex
83 x is. 1M.

 

 

 

A2 32:8..853— 8.5.65.5. 5:5 meezoom c8 n 1.8 m no 825» .3 ~35.

125

5.2.3 Shear-Moment Interaction

The shear capacity predicted by the AASHTO Standard Specifications [2] will be
larger than the true shear capacity as evaluated from a flexural test. This is due to the fact
that the development of the equation for the shear capacity assumes that the section
undergoes pure shear loading with out any bending. The inclusion of flexural stresses at
a section reduces the shear capacity of the beam as evidenced by Equation (22). Thus,
there is a trade-off between the shear capacity and the flexural capacity of the beam. If a
test is conducted in pure flexure, with no shear, the flexural capacity can be achieved.
This is what is typically pursued in a typical flexural test, where the section Of interest is
located under constant moment demands. Clearly, if the beam could be tested in pure
Shear the full shear capacity would be achieved. However, the test setup to introduce this
kind of loading is very complicated. Rather, most shear demands are unavoidably
accompanied by flexural demands, as was the case in the shear evaluation experiments in
this project. The critical shear section was thus in a region experiencing both shear and
flexural stresses.

The reduction of the shear capacity by flexural actions is caused by the increase of
the longitudinal strain in the steel reinforcement when shear is applied. This can be seen
from Equation (22). This increase of the strain reduces the amount of strain available to
resist the flexural forces. As the capacity of a section is defined at the yield point of the
steel, additional strain in the longitudinal steel will bring the steel closer to yield.

The MCFT can be used to predict the affects of the interaction between shear and

flexure. From the AASHTO-LRFD [3] Equations (20) and (22) can be used to produce a

flexure-shear interaction diagram. The longitudinal strain a} is changed through a range

126

of values and the shear capacity and moment capacity are found using Equations (20) and
(22). In this project, this evaluation was done using the program Response 2000 [8],
which is a sectional analysis program that incorporates MCFT in its full form [13] rather

than in the simplified approach outlined by the AASHTO-LRFD Specifications.

5.3 Flexural Experimental Evaluation
5.3.1 Test Setup, Instrumentation and Protocol

The flexural capacity of the box beams was assessed through four-point bending
tests. The four-point bending setup was chosen to create a region of constant moment
without Shear to assess pure flexural response. An overview drawing of the test setup is
given in Figure 46. The 52-ft beams were simply supported l-ft from the ends on top of
reinforced concrete support blocks. A 3-inch thick laminated elastomeric pad was used
between the beam and the reinforced concrete block to avoid stress concentrations and
allow for beam end rotations. The beam was symmetrically loaded at midspan at two
locations spaced 8 feet apart. Loading was applied by a pair of servo-controlled
hydraulic actuators attached to a gravity reaction frame. To avoid local loading on top of
the box beam upper flange, the load was distributed by means of a spreader steel beam
that straddled the beam transversely. A l-in. thick glass—reinforced elastomeric pad was
used between the spreader steel beam and the top of the box beam to minimize stress

concentrations. A picture of the test setup is shown in Figure 47.

127

.______r__

ACTUATOR —\

  

I]\ RIGID FRAME

{1 ] TEST BEAM
EL 93.! 0 f
riff")... [  ‘ “ m”ifIf... [ff if __}fj_f If.if.'I'flflffffmf[flffff [fjjf‘

 

 

 

 

m .. ' RIGID; * J, T
if] FLOOR \ l n

F“ 21:0" ——~—+- 80" -a] SlgPInggj
+————*-— —- 52'-0"

 

 

 

Figure 46. Flexural Test Setup — Overall Features and Dimensions

The material properties for the prestressing steel and each of the four project mix
designs are state in section 4.3 in Table 36 through Table 41. The compressive strength
at the time of the flexure test for each mix design is restated in Table 43 for reference.

Table 43. Flexure Test Concrete Compressive Strenglhs

 

 

 

 

 

Age at day of Standard
Mix Design flexure test f’c (psi) Dev.
(daYS) (Esi)
NCC 41 8,196 437
SCC 1 43 8,290 200
SCC 2 42 6,754 290
SCC 3 40 6,685 236

 

The flexural tests were provided with diverse instrumentation to monitor overall

performance and test setup control. The provided instrumentation was:

Displacement transducers to measure vertical motion of the beam at midspan,

point of load application, and support locations.
Rotation transducer at the beam end to measure beam rotation at the support.

Rotation transducer perpendicular to the beam at midspan to evaluate any torsion

effects.

128

I Displacement transducers at 10 prestressing strands on one end of the beam to

monitor any relative slip between the strand and face of the concrete beam.

I Strain gages in selected bottom-row prestressing strands and at selected locations.
A schematic of strain gage locations is given in Figure 48, Figure 49, and Figure
50.

I Strain gages on the concrete surface at the top of the compression flange at beam
midspan. A schematic of strain gage locations is given in Figure 48, Figure 49,

and Figure 50.

 

Figure 47. Overall View of Flexural Test Setup

129

The flexural tests were conducted by monotonically loading the beam up to
failure. The beams were initially loaded in force control in increments of 10 kips at a rate
Of 0.01 kip/sec. The system response was carefully monitored to evaluate the onset Of
significant section cracking. Once the section started to show significant nonlinear
behavior, the loading was applied in displacement control in typical intervals of 0.25 in.

at a rate of 0.0025 in./sec.

5.3.2 Observations and Results

The response of the flexure test beams was very much as expected for all tests.
However, it should be noted that the first two tests (NCC and SCCl) showed
asymmetrical cracking indicative of torsion effects. It was found that this was due to the
uneven top of one of the support blocks. This was corrected for all subsequent tests.
While the torsional stresses have an effect on the first cracking load and cracking
behavior of the beams, the demands on the NCC and SCCl beams are considered
negligible for the ultimate strength. As the ultimate load was the main focus of this study
the effect of the torsion demands were disregarded.

The sectional capacity of the beams was smaller at midspan (center of the
constant moment area) due to the presence of open pockets on the top flange with
approximate dimensions of 4” in length (along the beam length), 3” in width and 3” in
depth, see Figure 43. These pockets were used to flarne-cut the top strands in the section.
The NCC beam was tested with these pockets left un-grouted, which essentially reduced
the width of the section compression flange. The pockets of the SCC2 beam had been
grouted at the producer’s plant and the beam was tested with these pockets filled,

however all other beams were tested with the top flange pockets un-grouted.

130

The appearance of flexural cracks, their paths, and their widths for all beams was
consistent with expected behavior. A photograph of the overall beam response (SCCl
unit) at near maximum capacity is shown in Figure 51. Since the beams were tested
without the deck compression flange, failure of the section was thus dictated by failure of
the compression flange. The failures were explosive since the concrete compressive
strengths were relatively high and also due to the high strength Of the prestressing strand.
However, the failures were technically ductile as the prestressing strands yielded before

failure.

 

 

 

 

________________________________________________________________________________________________________________________________

________________________________________________________________________________________________________________________________

 

 

 

Figure 48. Plan View of Strain Gage Locations

131

 

 

36" , Tl.
«~————— 1'-6" ————. 2" «i -

 

 

 

 

 

 

 

 

 

 

 

O
. >06" 5"“ \
Strand Typ.
Each Side
F
N
4V2"
O O
\ "4V2" /
O O O O
O O O II O II O O O O J
\
1‘L____ 1r-2n _____._‘_ 6n .. lr_4n _"

 

 

 

 

 

Y Strain Gage Location

Figure 49. Strain Gage Location Sections A-A and C-C

 

 

 

 

 

 

 

 

 

 

 

 

36"

._._____ 1"6" ___.~

5 e e e . g
0.6" 5" —
Strand Typ. 4" x 3" X 3"
Each Side Block Out

’5 Typ. Each Side
41/2n
O O

 

 

X _ 4V2" /
L'i. .r. r.r .I'

4" 10" ——fi-+ 6b] 4" le—12"——-
Y Strain Gage Location

 

A

 

 

 

 

 

 

 

 

Figure 50. Strain Gage Location Section B-B

132

A View of the typical failure mode for all of the flexure tests is shown in Figure
52. As it can be seen in this figure, after compression failure of the top flange, the
section essentially folded onto itself as internal equilibrium forces seek a balanced
condition. After failure of the webs, the large rotations spalled off the concrete
underneath the prestressing strands. No deformation was recorded by the transducers

mounted on the prestressing strands at the beam end, indicating no signs of strand slip.

 

Figure 51. SCCl Flexure Test Beam near Ultimate Capacity

133

 

Figure 52. Failure Mode of SCCl Flexure Beam

The results are summarized in Table 44. In this table the maximum total applied

load and displacement are shown as well as the maximum moment that was achieved at
the beam midspan. This maximum value is compared to the midspan moment predicted
using the AASHTO Standard Specification [2]. From this comparison it can be seen that
the NCC beam exceeded the design moment by 10 percent while the SCC 1 beam
exceeded this value by 9 percent. The SCC 2 and SCC 3 beams both exceeded the design
moment by 6 percent.

The applied force per actuator versus the beam maximum (center) displacement
for all flexure beams is shown in Figure 53. The traces for each of the beams are
identified by the use of different symbols. It can be seen that the overall response of all
beams was essentially equal, with the NCC beam reaching a slightly higher load at failure

and the SCCl beam showing the largest deformation at failure. Figure 54 shows the

134

moment versus curvature response of the mid-span section for all beams. The section
curvature was calculated from the strain gage readings on the top surface of the concrete
compression flange and the strain gages in the first row of prestressing strands (see
Figure 48, Figure 49, and Figure 50).

Figure 53 and Figure 54 also shows two response limit lines corresponding to the
design nominal capacity, which is the design capacity for the bridge with the assumed
design concrete compressive strength of 5,500 psi. The sectional capacity was
determined using the simplified approach presented in the AASHTO Standard
Specifications [2] (see Section 5.2.1). This design capacity was calculated for the girder
only, that is, without the deck, in order to compare its value against test results. The
limits are shown for both the full and reduced section. The full cross section refers to the
section capacity neglecting the existence of the before—mentioned top flange pocket. The
reduced section assumes that the compression flange reduces to an effective width of 26
in. upon subtracting the width of both pockets (3 in. each) and the strip of concrete on the
outside (2 in. each), as shown in Figure 43. Clearly, the capacity of the reduced section is
slightly lower than that of the full section.

Comparison of the beam’s load-displacement and applied moment-curvature
response against the design limits in Figure 53 and Figure 54 shows that the design limits
may have not been reached in some cases. However, it should be realized that the
responses in Figure 53 and Figure 54 are due to applied forces only, which is without
considering initial self-weight demands. Self-weight demands for the 50 ft simple span
are significant: a positive moment at midspan of approximately 176.6 kip-ft. Thus, the

capacities achieved for the test beams should consider these initial demands for

135

comparison against the calculated design capacities. This was done for the load-
displacement and moment—curvature traces shown in Figure 55 and Figure 56,
respectively. It can be seen from these figures that the sectional capacities of all beams
clearly exceed the required design capacities.

Comparison of the relative performance of the flexural beams, and particularly,
the performance of the SCC beams compared to the NCC beams, can also be seen in
Figure 55 and Figure 56. It can be observed that the NCC beam reached the highest
section capacity and that the SCC3 had the lowest (see Table 44). As far as deformations,
the SCCl had the highest deformation at failure and the NCC beam had the lowest.
Clearly, the beams had different concrete strengths at the day of test. Thus, in an attempt
to compare performance while taking into account the different concrete strengths, the
responses were normalized with respect to the compressive force in the top beam flange
using equivalent stress block parameters. In normalizing the test values, the compressive
strength from the day of test was used. The normalized force-displacement and moment-
curvature responses are shown in Figure 57 and Figure 58, respectively. It should be
noted that the design limit lines have also been normalized accordingly using the design

concrete compressive strength of 5,500 psi.

136

 

 

 

 

    

 

 

 

 

 

 

 

 

80 q
70: ._ ..................................................... _
605
a 50E
g d
E 401
0 d
-’ . Design-Full’
30 ‘. —- ----- Design-Reduced‘
; P 1 P 'NominaI-AASHT017th Ed.
. i— - . j +SCC1
10: 4e» \~‘._Tg.,///5i +SCC2
. 5 +sccs
0 fiTVI'Tr 'l' '1' I TrI' 'I""I"Tr
0 1 2 4 5 6 7 8 9 10

Displacement (in.)

Figure 53. Applied Load vs. Displacement Response of All Flexure Beams

 

1600 .

 

1400 f

1200 :

Moment (kip-ft)
on
8

 

 

 

 

 

 

 

ne—e-

Design-Full‘
Design-Reduced’

] 'NominaI-AASHTO 17th Ed.

 

 

 

 

 

 

400 f
—O— NCC
- + SCC 1
200 —I-- SCC 2
+ SCC 3
0 . . , . . . .
0.0000 0.0001 0.0002 0.0003 0.0004

Curvature (1/in.)

Figure 54. Applied Moment vs. Curvature Response of All Flexure Beams

137

 

80.

 

Load (kip)

  
 
   

Design-Full'
— ----- Design-Reduced‘
'NominaI-AASHTO 17th Ed.

 

 

 

+NCC

—v—SCC 1
+SCC 2
—e—SCC 3

0 ' V V V V " Tfi' ' V V1 V I V V V f" V V V I U V V V V ' V V V V ' V V V V ' V V V T ' V V V V1

0 1 2 3 4 5 6 7 8 9 10

 

 

 

 

 

Displacement (in.)

Figure 55. Total Load vs. Displacement Response of All Flexure Test Beams

 

 

 

 

 

 

 

 

 

 

 

 

1800 ,
1600-: ___________ ____
1400-:
3 1200-2
.11 :
5 1000-
up
5 800 '
g . P P — — — Design-Full"
E . — -------- Design-Reduced"
600 - , th
I [i 'NomlnaI-AASHTO 17 Ed.
400‘. \ T “IE5” / ’ —e—Ncc
\ // —I—SCC1
200 \l/ —I— see 2
M —0— SCC 3
0,
0.0000 0.0001 0.0002 0.0003 0.0004

Curvature (1fin.)

Figure 56. Total Moment vs. Curvature Response of All Flexure Test Beams

138

 

 

 

 

 

 

 

 

 

0.12
0.11 — ‘T —————————————————
k1=.851081bwhf
0.10 ........................................................... —
A 0.09 I . .
x , a
E 0.08 ’
'o
8 0'07 — — — Normalized-FUN"
-‘ 0.06 / —- -------- Normalized-Reduced*
8 *NominaI-AASHTO 17th Ed.
N /
--_— 0.05
g //
0.04
o // P P
z ..
0.03 //
. _ —e—NCC
001 , ” ‘““—-c})” ‘ +8002
‘ —e— SCC 3
0,00 ...............,.........,....T.-.T,..-.4..”fins
O 1 2 3 4 5 6 7 8 9 10

Displacement (in.)

Figure 57. Normalized Total Load-Displacement Response of Flexure Beams

 

 

 

 

 

 

 

 

 

 

 

 

0.12
x01 ] k2=dv'hf/2 ' J A L, - -
{T- A
g 0.08 J i . ' T
E /V ‘r
“s’
g 0-06 j , , — — — Normalized-Fun"
"c 1 3' -- -------- Normalized-Reduced'
g , .// ' P P ‘NominaI-AASHTO17th Ed.
2 . AT\-\L__ll_,// i)
002 , \ 7 5 , —e—N00
. t \\ / ' SCC 1
+800 2
+800 3
0.00 I i . . I I T i v I I ' ' . . l . fl ' '
0.0000 0.0001 0.0002 0.0003 0.0004

Curvature (1/in.)

Figure 58. Normalized Total Moment-Curvature Response of Flexure Beams

139

Table 44. Maximum Achieved Capacities of Flexural Beams

 

Maximum Maximum Maximum Design Moment Actual to

 

 

 

 

Total Applied Total [AASHTO] * Design
P-Load Center Moment Full/Reduced Ratio
(kip) Disp. (in.) (kip-ft) (kip-ft) (Red.)
NCC 78.5 8.1 1,648.8 1,546/1,499 1.10
SCCl 77.5 9.2 1,628.4 1,546/1,499 1.09
SCC2 75.9 8.5 1,592.8 1,546/1 ,499 1.06
SCC3 75.7 8.2 1,590.0 1,546/1,499 1.06

 

* AASHTO Standard Specifications — l7lh Edition [1]

In spite of the attempt at normalizing the response the comparison does not seem
to be adequate in order to discriminate the beam’s performance against a pre-established
limit. The normalized plots essentially show the efficiency of the section with respect to
the concrete compressive strength. Thus, it can be noted that the beams with the higher
compressive strength (NCC and SCCl) have the lowest normalized response value, while
the beams with the lower concrete strength (SCC2 and SCC3) achieve greater normalized
ratios. This essentially indicates that for the under-reinforced section, the compressive
strength of concrete is not a significant parameter. This is confirmed by the fact that in
spite of having different concrete compressive strengths, the overall load and moment

capacities of the beams as shown in Figure 55 and Figure 56 was essentially equal.

5.4 Shear Experimental Evaluation
5.4.1 Test Setup, Instrumentation and Protocol
The shear capacity of the box beams was also assessed through four-point

bending tests with the same support and loading conditions previously described for the

140

flexure tests, however the shear spans were considerably reduced in order to increase
shear force demands. A picture of the test setup is Shown in Figure 59. The shear span
was chosen with the goal of inducing shear failure in the beam before reaching the
section flexural capacity in the center section. However, reducing the shear spans led to
overhanging cantilevers beyond the simple supports of the setup causing negative
moments to develop in the beam at the supports. To minimize these negative moments,
and their effect on the beam behavior, the reduction of the shear spans was limited. The
setup was thus determined by the need to keep the beam overhangs at a minimum and
minimizing flexure—shear interaction to ensure a shear-dominated failure.

The material properties for the prestressing steel and each of the four project mix
designs are state in section 4.3 in Table 36 through Table 41. The compressive strength

at the time of the shear test for each mix design is restated in Table 43 for reference.

Table 45. Shear Test Concrete Compressive Strenghs

 

 

 

 

 

Age at day of Standard
Mix Design shear test f’c (psi) Dev.
(days) (psi)
NCC 55 8,560 134
SCC 1 58 7,519 158
SCC 2 66 7,711 430
SCC 3 54 6,953 1 18

 

An overview of the first shear test for the NCC beam is shown in Figure 60. The
shear span in this setup was 11 ft. As discussed later, the flexural and shear capacities of
the beam for this configuration were too close and a flexural failure was reached before
inducing significant shear deformations in shear spans. Thus, the tests setup for the

subsequent SCC beams was changed so that the shear span was reduced to 9 ft as shown

141

in Figure 61. Implications of this change in the test observations and in the interpretation

of results are discussed later.

SITES €110“.

 

Figure 59. Overall View of Shear Test Setup

The shear tests were provided with a different instrumentation scheme than the
flexural test beams. Again, instrumentation was placed to monitor performance and
overall test setup control. In order to assess shear behavior, most of the instrumentation
was located in What was determined as the critical shear section. The critical shear
section was determined from sectional analyses using the program Response 2000 [8] to
be located approximately 29 inches into the shear span from the points of load application

(see Figure 60 and Figure 61). Since shear effects are not localized at a section but rather

142

distributed over the region of shear cracks, a shear critical region was thus defined over a

distance of i h/2 from the critical shear section.

 

ACTUATOR X— RIGID FRAME
SHEAR
DEFORMATION

PANEL [ TEST BEAM

 

 

...................................................................................................................

 

 

 

 

 

.......................................................................................................................

PM .. RIGID”. fJLl
] FLOOR _\t\

 

 

 

 

 

~———— 1120" ————-—i.—— 11'-0" ——--—— 8'-O" —1~ ‘L $13013“

 

 

 

52'-0"

 

Figure 60. Shear Test Setup for N CC Beam — Overall Features and Dimensions

 

 

 

 

 

 

 

 

 

ACTUATOR RIGID FRAME
SHEAR
DEFORMATION
PANEL Y o o /“ TEST BEAM
”1': RIGID ............................................... r .................................

 

SUPPORT
BLOCK

-~———— 13'—0" ———«—- 9'-0" ———~~»— 8'-O" —-l-

52'—0"

 

 

 

 

 

 

 

Figure 61. Shear Test Setup for SCC Beams — Overall Features and Dimensions

The instrumentation provided to the shear beams was:
I Displacement transducers to measure vertical motion of the beam at midspan,

point of load application, and support locations.

143

I An arrangement of displacement transducers around the shear critical section
named a shear deformation panel to measure average shear deformations in the

shear critical sections (see Figure 62).
I A rotation transducer on the beam at the support section.

I A rotation transducer perpendicular to the beam at midspan to evaluate any

torsion effects.

I Strain gages in selected bottom-row prestressing strands and at selected locations.
A schematic of strain gage locations is given in Figure 63, Figure 64, and Figure

65.

I Strain gages along the height of selected shear stirrups in one of the shear critical

sections (see Figure 63, Figure 64, and Figure 65

I Strain gages on the concrete surface at the top of the compression flange at beam

midspan and at the critical shear section (see Figure 63, Figure 64, and Figure 65)

The test protocol for the shear beams was the same as that previously described for the

flexural tests (see Section 5.3.1).

144

 

Figure 62. View of Shear Deformation Panels

5.4.2 Observations and Results

Interpretation of results from the shear tests is more intricate than for the flexure
tests. This follows due to complexities from unbonded tendons along the beam length
and the continuity of the beam beyond the supports as cantilever overhangs. The
partially bonded/debonded strands in the top flange modified the shear and flexural
capacity of the beam along its length; while the cantilever overhang continuity over the
supports allowed the beam to redistribute loads beyond the supports (see Figure 66).
Further discussion of these effects follows.

Shortening of the span in the four-point bending setup essentially increases the
force required to reach flexural failure at the beam mid-span due to the reduced lever

arm. Predictive analyses were used to identify the region along the beam expected to

145

have the highest shear deformation demands. For all beams, this section was
approximately 29 inches from the point of loading along the shear span. As previously
mentioned, shear deformation measurement panels were placed centered about this
critical section (see Figure 59 and Figure 62).

As expected, degradation of the system was concentrated through shear
deformations along the shear critical zone as seen in the picture of Figure 67. As seen in
this picture, the inclination of shear cracks was very shallow due to the high levels of
prestressing force in the section. This same prestressing level, which extended beyond
the supports, allowed the beam to redistribute member demands towards the end of the
beams even after the critical shear section reached its capacity. Thus, while the shear
capacity of the critical section was reached, complete shear failure of the beams was not
reached in all cases except for one test unit. The SCC3 beam exhibited a flexure-shear
failure upon crushing of the compression zone in flexure (see Figure 68). This failure is
not a typical shear failure but follows from the combined flexure and shear loading and
the weakening of the beam due to shear cracking. This effect is further discussed next
with reference to the measured data.

The total (corrected for self-weight) shear force versus center displacement for all
of the shear test beams is shown in Figure 69. It can be seen in this figure that the
response of the NCC beam was considerably different than that of the SCC beams. The
reason for this difference was that the shear span for the NCC beam was 11 ft while that
for the SCC beams was 9 ft. Thus, the force—displacement response for the SCC beams

(with the shorter shear span) is much stiffer.

146

:r‘; [3’05

 

 

 

 

 

..............................................................................................................................

L_—.-——_—«_—-~—--->—-_-———-—__._--_—.—.—-——--——_<_-———_—.<—--_--——__4 »_______-___.. __-..____.__-_-__-- _______________________________

 

Figure 63. Strain Gage Instrumentation Location for Shear Test Beams

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

36" *J
‘______ 1"6" _.i 2" j
O O O O O ' \NF
y... - \
Strand Typ.
Each Side
— #4 Stirrup
T d B
an 4%"
\ _ 4V2"
’ J
Lt."—

 

 

 

 

 

 

 

Figure 64. Strain Gage Locations Sections A-A, B-B, C-C

H 6" >
V Strain Gage Location

 

147

 

 

 

 

 

 

 

 

0.6" 5" — \

Strand Typ.
Each Side

— #4 Stirrup
T and B

\ — 41/2" /
{—2 O
t r

. I 1"2"

27" ———-————-1~
j

 

 

 

 

 

 

TL
\
A

g};

 

 

 

 

 

 

 

 

 

 

 

 

 

‘_ 6n ._ lv_4n ____._‘

V Strain Gage Location

Figure 65. Strain Gage Locations Section D-D and EB

While the effect of the shorter shear span is clearly noticeable in the overall beam
force-displacement response, the shear behavior is best evaluated by studying the section
response at the critical shear region. The total moment-curvature response at the critical
shear section for all beams is shown in Figure 70. In this plot it can first be seen that the
response of all beams is very similar. Secondly, it can also be observed that the moment-
curvature response is almost elasto-plastic, with a much sharper shift towards the
inelastic regime than observed in the flexural tests (see Figure 56). This indicates that the
moment capacity at the critical section abruptly reaches a maximum. The additional
section capacity (can think of it as the “hardening” region) is due to the redistribution of

loading on to the cantilever overhang.

148

 

Figure 67 . Typical Distress in Shear Span and Critical Shear Region

149

 

Figure 68. Flexure-Shear Failure of SCC3 Shear Test Unit

Noted in both the force-displacement (Figure 69) and moment-curvature (Figure
70) responses are values of the nominal capacities calculated for the design beam, i.e.,
with a design compressive strength of 5,500 psi. The nominal capacities were calculated
according to both the AASHTO Standard Specifications [2] as well as the simplified
sectional analysis procedures recommended in the AASHTO-LRFD Design
Specifications [3]. The reason to show capacities according to both design codes is that,
as seen in the figures, the values predicted by the 17th Edition Standard Specifications are
much higher than observed in the test, while those obtained with the AASHTO-LRFD
provisions compare much more favorably. In addition, it can be noted that the nominal
capacities of the critical shear section calculated with the AASHTO—LRFD provisions

compare reasonably well to the onset of significant flexural softening of the sections

150

(Figure 70), which is different than the behavior observed for the pure bending sections
in the flexural beam tests (see Figure 56).

The reason for the reduced shear capacity of the tested beams compared to the
design value from the Standard Specifications is due to flexure-shear interaction effects.
The shear and flexural capacities of a section under combined moment and shear
demands will be a compromise between the capacities for pure flexure and pure shear
failure [l3][22]. This effect is not explicitly taken into account in the Standard
Specifications; however, the LRFD Specifications address this issue for the determination
of shear capacities, which is based on the principles of the modified compression field
theory [13]

The presence of shear in a section will also decrease the flexural capacity of a
section due to the additional horizontal tensile forces created by the diagonal compressive
struts in the section web. Neither code provides explicit guidelines to estimate the
reduced flexural capacity of the section due to the presence of shear. However, this
reduced moment capacity can be determined by considering the additional tensile strains
generated in the section due to shear [l3][22] (also see Section 5.2.3). These calculations
were performed by making use of the program Response 2000 [8] for the sectional shear
capacity equations from the AASHTO-LRFD specifications. A shear-moment interaction
diagram computed by Response 2000 for the design beam 0’, = 5,500 psi) is shown in
Figure 71. In this figure, the loading line corresponds to the test setup with a shear span
equal to 9 ft and the extreme ordinate and abscissa values correspond to the pure flexure

and pure shear capacities.

151

 

200

For all SCC Beams: Lv= 9 t1

 

 

 

 

 

 

 

 

 

180 For NCC Beam: Lv= 11 it ‘ ,q I
160 as
r "I
140 4’ I ’
...... ‘7’. . .__._._._._._ _._._._._._._._._._._._.
a 120 _ ll _ _ ___________ _
g j — — — AASHTO LRFD 2" Ed.
13 100 j Design Nominal Capacity
8 ,. [Lv=11 ft]
-‘ 80 ., - ----- AASHTO LRFD 2"“ Ed.
7’ p p ResigrfthominaI Capacity
60 .—
/ l I‘v l Shear Section v-
40 , E x“ X l / f: + NCC
, 5% x .— A ’5 + SCC1
I
+
2° Note: AASHTO 17th Ed.:176 kip . Egg:
0 Design Nominal Capacity
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

Displacement (in.)

Figure 69. Total Shear Force vs. Center Displacement Response

1300*

 

For all SCC Beams: Lv=9 ft
1200 For NCC Beam: Lv=11 ft

 

 

 

 

 

 

 

 

 

 

1100 a. in
1000 ————— , ' ———————————
Afr
A 900 _ __ _ _ ,I ________________________________
3.: 300 4.. nd
f—g- ,, — — — AASHTO LRFD 2 Ed.
V 700 3 Design Nominal Capacity
‘5 . [L -11r1]
0’ 600 l- P P _._._._ v- nd
E , Ly . AASHTO LRFD 2 Ed.
g 500 3’ \ h“ j Shear SectI/on Design Nominal Capacity
400 ,0 FAT? r - ’i <91 '4 j "‘Vzg "1
300 it XX“ —o—NCC
.89 M
200 ' M —v—SCC 1
_ . . . , . _ +3002
100 r Note. Desrgn Nomltrri‘al Capacrty. 1546 kip it ‘ SC C 3
0 [AASTHO 17 Ed.]
0.00000 0.00005 0.00010 0.00015 0.00020

Curvature (1f1n.)

Figure 70. Total Moment vs. Curvature Response at Critical Shear Section

152

 

       

 

 

 

 

 

 

 

: Vu = 130 kips
120 % /
A 1 /
3. : /
E 100 j /
V — /
2:, : /
o 80 ’i /
Ill. : /
L- g / . =
3 60 i / Design Beam. f’c 5,500
5 : / Shear Span: Lv=9 ft
- /
4° ‘3 / Load: AM=8.58 kip-ft, AV=1 kip
a /
20 _j / —— M-V Interaction
a /
1 / ——-— Loading Line
0 FTITIPIIITfifIIIFTPTITIj—rllllll[Tllllllll
O 200 400 600 800 1 000 1 200 1 400 1600
Moment (kip-ft)

Figure 71. Moment-Shear Interaction Diagram with AASHTO-LRF D

As done for the flexural beams, the shear force-displacement and moment-
curvature responses were normalized with respect to the compressive force in the flange
to take into account the different concrete strengths in the beams. The normalized shear
force-displacement and moment-curvature traces are shown in Figure 74 and Figure 75,
respectively. It should be noted that the design limit lines have also been normalized
accordingly using the design concrete compressive strength of 5,500 psi. As before, the
comparison does not seem to be adequate in order to discriminate the beam’s
performance against a pre-established limit. The normalized plots essentially show the
efficiency of the section with respect to the concrete compressive strength. Thus, it can
be noted that the beams with the higher compressive strength (NCC and SCC1) have the
lowest normalized response value, while the beams with the lower concrete strength

(SCC2 and SCC3) achieve greater normalized ratios.

153

lowest normalized response value, while the beams with the lower concrete strength
(SCC2 and SCC3) achieve greater normalized ratios.

The shear performance of the test units is best assessed by studying the shear
force versus shear strain response in the critical section (29” from the loading points).
This was achieved by determining the average shear strains in this region with the before-
mentioned shear deformation panels (Figure 62). The average strain on each edge of the
shear deformation panel as well as the diagonal are shown in Figure 72 as 4;. These
values are derived from the deformations shown in Figure 72 as 14,-, and my. These
deformations can be divided into the five deformation modes shown in Figure 73. Using

Equations (23) and (24) the values of strain can be derived [44].

Uly U4).

tun Cr e+1.14»

" l

 

[By

_L fl t} U311

u2y
u2x C4

 

 

 

 

r
l I
D

Figure 72. Shear Deformation Panel Theory

The resulting shear force vs. shear strain histories in both panels for the NCC and

SCC 1 beams are shown in Figure 76 and Figure 77, respectively. The experimental

154

plots are accompanied by a threshold limit and an analytical shear force vs. shear strain
prediction. Both of these analytical values were computed for the concrete strength of
the beam at the day of the test. The threshold limit shown was calculated according to
the AASHTO-LRFD Specifications [3] simplified sectional analysis procedure. The
measured shear strain plotted in the history is an average value from the five average

strains shown in Figure 72.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

///7 A
.-_/ / ,L/
’ / / \ /
/ \ K
16 / A29_/ \A29_
2 //l / \ \ A9.
2 A \\2
\l
1 2 2 3
F_——7
. j a—
l l
a l l 5
‘2‘ I I?
L. _l 8
L______J7
Figure 73. Shear Deformation Modes
“_L __1_ _L __1_ _ 1 1 1 _L—r“?
“1119“” 2H 20 2H 21) 2H 20 2H 21) u”,
1 1 1 1
— O —— O — 0 ——- O u
A6”) H H H H "2‘
<A60) ___ 0 __l_ O i 0 _i 0 .1. 4 2y}
(4) D D D D u...
5 1 0 1 0 1 0 1 0 u
-— " _ — 3
(8” J 2 2 2 2 u ’
0 —l- 0 —l 0 —i 0 l 4’
(23) _ 2 2 2 2 _ lu“)

155

 

 

 

 

 

 

 

~ '—1 0 0 07 »

l6] 0 l (I) 0 O “1.!

52 O uly

:1 : .1 l

54 0 0 D H 0 “4”

(24) (55) L VDz'l'HZ VDZ‘l’HZ #343,“

As explained earlier, the AASHTO LRFD specifications are used for comparison
with the experimental data from the shear tests since the values estimated by the Standard
Specifications ignore the interaction between shear and flexure, which leads to higher
estimates for shear strength. The analytical response (shown with hollow symbols) was
determined from a sectional analysis program based on the modified compression field
theory (Response 2000) [8]. It can be seen that the calculated capacities according to the
MCFT agree closely with the experimental data, as does the analytical shear force vs.
shear strain response. The deviations seen between the analytical and experimental traces
in the initial part of the response are attributed to limitations on the resolution of the
displacement transducers used in the shear deformation panels (see Figure 62).

From the responses of the NCC and SCC1 beams (Figure 76 and Figure 77,
respectively) it can be seen, first that there is a clear point at which the shear strains in the
critical region increase at a very high rate and second, how this level corresponds very
closely to the predicted shear capacity of the section. It can thus be concluded that the
sections, both in the NCC and SCC1 beams, reached their shear capacity. From the
responses it can also be observed that the “post-elastic” behavior is essentially flat, thus

indicating that the section has essentially no more shear resistance.

156

force vs. shear strain response for all beams is given in Figure 78. In this figure, the
measured strains in the deformation panels for each beam were averaged. Also in Figure
78 are threshold limits that indicate the nominal shear design capacity according to the
AASHTO-LRFD specifications for the design compressive strength of 5,500 psi. Two
threshold lines are shown since one corresponds to the shear span setup of the NCC beam

(11 ft) and the other corresponds to the shear span of the SCC beams (9 ft).

 

 

 

 

 

 

 

 

 

 

0.24 ,
K,=.85rcabwh,
0.22 For all SCC Beams: Lv=9 it
020 FEE—H ”QC—3.2303: .L_V.=_1_1. fl ..... _ ___________ ._ .
A 0.18 _____ _ 4 _. _ __ __ ._ ._
{.—
a_ 0.16
3 0.14 f
3 I; AASHTO LRFD 2"6 Ed
0.12 ' — — — '
8 Normalized Nominal Capacity
E 0-10 [Lv=11 11]
E 008 P P _._._._ AASHTO LRFD 2nd Ed.
5 . L . Normalized Nominal Capacity
2 0.06 l v l Shear Section [Lv=9 fl]
? x \ 1 , ,1 ’1 +Ncc
0.04 KM—J/g —v—SCC1
002 > Note: Normalized Capacity — 0.27 ' -—I— SCC 2
[AASHTO 17m Ed.] + SCC 3
0,00 ”-4.4fi.f.fi.,....,....,-...,........
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

Displacement (in.)

Figure 74. Normalized Shear Force vs. Displacement at Critical Shear Section

157

 

 

 

 

 

 

 

 

 

 

0.075
0.070 K155a5b1h. 121.131.75— Fo—raECFBe—a-mfiilz—Q- 1?,— NE: EJ171— — '—
0.065
”a. 0.060 '
x
g 0050 , g 2 ‘ Note Normalized Nominal Value=0.112
5 0-045 ‘1. " [AASHTO 17"1 Ed.
g 0.040 1 * _._._._ AASHTO LRFD 2nd Ed.
2 0035 a Normalized Nominal Capacity
'0 [Lv=9 it]
I:
g 0'030 ( P lP — — — AASHTO LRFD 2"d Ed.
5 0‘025 ' (_l-y l Shear Section Normalized Nominal Capacity
5 0.020 .’ [\ L‘ ‘ l / ,1 ’1 [Lv=11 it]
' I K . + SCC 2
0.005 ; . \ .89 M + SCC 3
0.000 ....r...M.,....,ffi,,
0.00000 0.00005 0.00010 0.0001 5 0.00020

Curvature (1/in.)

Figure 75. Normalized Moment vs. Displacement Response at Critical Shear Section

 

Ill

5

 

ii

  
 

 

 

 

 

 

 

 

 

 

 

1
.3 10° ‘5 P
e“; r L .
0 _ l v Shear Section
2 80 j _‘ 7 1 ,. ,L /‘1
62 3 L ~ ~— ~J “L
a 50 ‘2
g 3 — ........ Vu - Response 2000
(D 40% ——— Vu-AASHT01999
20 I —0— Experiment - Panel_1
‘3 + Experiment - Panel_2
: Analysis
0 IIIlfIIIIIIIIIIIIIIIIIII'llll[IIIITIFTTIIIIITIIIIITITT

-100 0 100200300400 500600700 8009001000

Average Shear Strain (mlcrostrain)

Figure 76. Total Shear Force vs. Shear Strain for NCC Beam

158

Shear Force (kips)

388

Shear Force (kips)

Figure 78. Total Shear Force vs. Shear Strain at Critical Section for All Beams

dub

8§§§

llllIlLllIllllIllliIlULIlJLlIllllIlllllllll

nub

O
O

O

 

 

 
 
 

P P
l Shear Section

I A /

\K ’j/g/ 7 l

 

 

Vu - Response 2000
vu - AASHTO-LRFD 2nd Ed

00 -oo_co

 

—o— Experiment - Panel_1
—v~— Experiment - Panel_2
Analysis

 

 

 

 

 

IIITTTIFTTI'FFIIITTIIFTIIIIlTIleFTIIITIIIFIIIIIIIT

0 250 500 750 1000 1250

Average Shear Strain (microstrain)

Figure 77. Total Shear Force vs. Shear Strain for SCC1 Beam

asee§§§§§

llLJIJlllilll_1_llllIII]IIILLIJILIJIIJLIIIIIIJ

O

 

 

4|

 

NCC - AASHTO-LRFD 2nd Ed.
[i'c=5500psi, Lv=1 1ft]

. SCC - AASHTO-LRFD 2Ind Ed.
‘4 [f'c=5500psi, Lv=9fi]

) .}

 

—o— NCC Exp_avg

—v— SCC1 Exp_avg

+ SCC2 Exp_avg
‘ $003 Exp_avg

V

 

 

 

 

 

 

IIFTWTTIITIITIIT[TUTIIIIIIIIIIIITIIIIIIUI

0 250 500 750 1000 1 250 1 500 1 750

Average Shear Strain (microstrain)

159

 

1500 1750 2000 2250 2500

2000

While only the SCC 3 beam exhibited a flexure-shear failure, all of the beams
reached their shear capacities at the critical shear section. However, due to the large
amount of prestressing force provided by the prestressing strand the beam was able to
redistribute the shear demand beyond the supports to the overhanging portion of the beam
once the shear capacity was reached. This allowed the beam to carry additional shear
load as the test continued without causing a shear failure of the beam. The beams
eventually failed in flexure, generally the failure was a compression failure in the top
flange of the beam, as the loads were increased. This was previously shown in Figure 68.
Some of the tests were stopped before this flexural failure occurred but only after it was
determined that the shear capacity of the critical section had been reached.

The fact that the beams reached their shear capacities is supported by information
on the shear reinforcement at the critical shear section. By showing that the shear
reinforcement at the critical shear section yielded in all beams, it is possible to show that
the beam was no longer capable of resisting shear loads at this point. The shear
reinforcement consisted of #4 U-shaped stirrups in the top and bottom flanges of the
beams (see Figure 63, Figure 64, and Figure 65). These stirrups were spaced at six
inches throughout the end blocks of the beam and then at twelve inches through the
remaining portion of the beam. Strain gages were installed on four stirrups near the
critical shear section prior to the beams being cast. The four instrumented stirrups were
spaced one foot apart starting five feet from the center line of the beam and continuing to

eight feet from the center line as shown in Figure 79.

160

Section B Section C
Section A \ # Section D
[_— ————————— _' — _ _ll _______________ "—1
L —————————— ”LI — — — 4| x ——————————————— _. |
L4H] spaces @ 1. _..3 _4. I I 5, —’l Beam Centerline j
VI

1‘ 52'

 

 

Figure 79. Instrumented Shear Stirrup Location

 

  
     

  
 

 

 

 

 

 

 

 

 

 

 

 

 

41/ .. i WES In _ EAST SIDE *
‘ 4
J 1
3" ~ Top Stirrup _/ \
Foil Strain 2
/ Gage Locations
Typ. Each Side 3
4
\ /- Bottom Stirrup /5
\ J

 

 

Figure 80. Strain Gage Location on Shear Stirrup

The critical shear section was found through computer analysis to be 6’-5” from
the centerline of the beam or 29” from the point of load application. The gages were
located along the section depth as shown in Figure 80. Five gages were installed on each
side of the stirrups, east and west. The top stirrups had two gages on each leg for a total
of four gages, and the bottom stirrups had three gages per leg for a total of six gages.
With respect to the bottom of the beam, the gages were located at heights of 4.5, 7.25,
13.25, 19.25, and 22.25 in. on each side of the beam (see Figure 80). The gages are

numbered as shown in Figure 80 from top to bottom, 1 through 5. A unique name can be

161

made for each gage using the section name, beam side, and gage number (i.e. A-El
Section A, east side of beam, gage l).

The role of the transverse reinforcement is to resist shear forces once the concrete
in the web of the beam has cracked and to control crack opening. The tension field
created by shear forces is first resisted by the concrete until its tensile strength is
exceeded and cracks perpendicular to the tension stress field are formed. Steel
reinforcement then resists the tensile forces required to provide a “bridge” for tensile
field between concrete cracks. At the crack, increasing shear forces are then carried
solely by the steel. In a region where significant cracking has occurred, shear resistance
is carried by a combination of tensile forces in the transverse steel and friction along the
concrete cracked surfaces. As shear demands increase the deformation demands on the
transverse steel will lead to reaching yield. Once a transverse bar has yielded, the
deformations become plastic and are not recoverable. Further application of load beyond
the point where the steel has yielded will cause the bar to continue to deform and the
shear crack to continue to open. A profile of the strain history from the instrumented
stirrups (Figure 81 - Figure 84) shows that the stirrups in each beam yielded at the
locations of interest. It can be seen that in all tests the load was applied beyond the yield
point of the transverse reinforcement and that the strain in the stirrups increased at a
much higher rate after yield was reached.

Figure 81 shows strain profiles from the second instrumented stirrup in the NCC
beam, located at Section B 3’-6” from the load point along the shear span as shown in
Figure 79 The strains are plotted along the height of the beam for multiple load and

corresponding displacement levels. The development of strain along the beam depth

162

during the test is shown from left to right in the plot figure, with earlier loading events
shown on the left side. From the figure it is possible to see that the stirrup reached the
yield point of (estimated at 2000 microstrains for reinforcement with a yield point of 60
ksi.) at approximately the 132 kip load level. The first gage to record a strain beyond the
yield point was that at 7.25 in. from the beam bottom. As the test progressed strains in
excess of 2000 microstrains were recorded by all but the top gage at this section. The
shape of the plot at the end of the test mimics a parabolic distribution of strain through
the beam depth. Observing the strain profiles at the different load levels, it can be seen
that very little strain was developed in the stirrup during the application of the first 90
kips of the test. However, after the stirrup yielded at 132 kips, small increments of load
(5 and 6 kips) caused large increases in the strain readings. The NCC beam test ended at
a load of 143 kips and a maximum rrridspan displacement of 3.8 in. At this loading level,
a maximum strain of 6500 microstrain was recorded by gage B-E4 on the stirrup at

Section B.

163

 

 

25:

+ 0, 0
+ 90, 0.5
+

 

« Beam Top I e Shear Force (kip). Disp (in.)
|
I

114, 1.5
+ 132, 2.5
+ 137, 3.0
+ 143, 3.8

 

 
   

 

 

 

 

 

Distance Along Beam Height (in.)

0 1 Beam Bottom
0 1000 2000 3000 4000 5 6000 7000

Microstrain

 

 

 

d————

V I t I I r V T T I T U r

Figure 81. Strain Profiles on Shear Stirrups at Section B for NCC Beam

Strain profiles for one side of one stirrup from the SCC 1, SCC 2, and SCC 3
beams are shown in Figure 82, Figure 83, and Figure 84, respectively. Changes made to
the shear test set up after the NCC test to increase the force required to cause flexure
failure changed the response of the SCC beams. The most significant change was a
stiffer response causing higher loads. Therefore, the stirrups in the SCC beams yielded at
higher loads than the NCC beam. The plot of the SCC 1 strain profiles shows that the
stirrup reached yielding before 156 kips. The critical stirrup in the SCC 2 yielded near
145 kips, and the SCC 3 critical stirrup yielded before 156 kips.

Only one side of one stirrup is shown for each beam. These representative values
are presented as they are the most complete profiles from each beam. Other stirrups had

gages that did not record properly through the entire test.

164

 

Beam Top Shear Force (kip), Disp (in.)

 

 

 

 

m
J
111.111

assaso
CONN-ICC
OOU'IOWV

 

 

 

 

 

15 |
l
|
10 I
‘ l
l
5 I
l
o“.mieemaan.nq.n-p.nhsn,fifi.fi
0 2000 4WD 6000 8000 10000 1200014000 16000 18000

Mkmamm
Figure 82. Strain Profiles on Shear Stirrups at Section A for SCC 1 Beam

The strain profiles from the transverse reinforcement in the critical shear region
show that the stirrups were demanded well beyond yield for most of the section depth.
This information, together with the shear force versus shear strain plots from Figure 78,
supports the conclusion that the maximum shear capacity of the beam section was
reached in all tests. Furthermore, the data in Figure 78 shows that the section shear

capacity exceeded the design level.

165

 

 

. Beam Top I
| 81

25-
1 I

    

 

 

N
O
1.11

d
U"
I 1 1

 

 

10

Distance Along Beam Height (in.)

 

 

 

 

 

l
l
l
1

1 Beam Bottom
0 2000 4000 6000 8000

. Microstrain

Shear Force (kip). Disp (in.)

 

T f '

Figure 83. Strain Profiles on Shear Stirrups at Section B for SCC 2 Beam

 

s Shear Force (kip). Disp (in.)

 

 

 

 

 

Distance Along Beam Height (in.)

 

 

 

 

 

0 2000 4000 6000 8000

Microstrain

Figure 84. Strain Profiles on Shear Stirrups at Section B for SCC 3 Beam

166

A summary of the maximum achieved capacities, defined as the point beyond

which shear strains increase at a very rapid rate, or in the “plastic” region (see Figure 78)

are given in Table 46. It should be noted that the nominal design values shown in the

table use the design concrete compressive strength of 5,500 psi. From the values in Table

46 and Figure 78 it can be seen that the SCC1 beam reached the maximum overall shear

capacity and highest ratio (1.22) compared to the nominal strength. While the NCC beam

had the lowest shear capacity, the nominal design capacity for the longer shear span is

also lower due to the larger influence of flexural effects (longer shear span). Thus, the

ratio of actual to nominal capacity for the NCC beam was not the lowest (see Table 46).

The lowest ratio of actual capacity to nominal strength was for the SCC3 beam, which

exceeded the nominal capacity by 8%.

Table 46. Maximum Achieved Capacities of Shear Beams at Critical Shear Section

 

Maximum Maximum

 

 

 

 

Nominal Nominal Actual to
Total Total Design Design Design
Shear Moment Shear “ Moment b Ratio
(kip) (kip-ft) (kip) (kip-ft) (Shear)
NCC 128.4 1102.3 116 994 1.11
SCC1 159.0 1046.6 130 856 1.22
SCC2 145.5 957.8 130 856 1.12
SCC3 140.3 923.3 130 856 1.08

 

8 According to AASHTO LRFD [1] Simplified Section Analysis Method
b Based on moment-shear interaction envelope calculated using Response 2000 [8] for AASHTO-LRFD

criteria.

In spite of the above-noted differences, the shear performance of all SCC beams

was similar and their differences were essentially in the behavior of the “post-elastic”

region. This is consistent with the fact that all mixes had similar tensile strength but their
aggregate content is different. This would imply that cracking behavior would be similar
but that post-cracking response will differ.

The previous discussions have presented data showing that the shear capacity of
the critical section in all beams was exceeded. The ultimate failures of the beams,
however, were not of a pure shear failure in the critical shear region. Further, while the
NCC beam failed in flexure at midspan due to the larger shear span (see Figure 60) the
section response (Figure 76) and stirrup data (Figure 81 - Figure 84) supports that the
shear capacity of the critical shear section was indeed reached. Analysis of this data was
not available during testing thus the shear span was shortened for the SCC beam tests
(see Figure 61). From Figure 78 it can be seen that all the SCC beams were tested past
the point where the beams had exhibited significant “post-elastic” shear deformations but
before any global failure. Thus, the critical shear sections the SCC 1 and SCC 2 beams
reached their capacity even if complete failure of the beams in shear did not occur. A
view of the shear distress in the critical region is shown in Figure 67. The SCC3 beam
had a combined flexure-shear failure initiated in the constant moment region also after
significant “post-elastic” shear deformations in the shear-critical zone (see Figure 78). A
view of the shear-flexure mode for the SCC3 beam was shown in Figure 68.

As explained earlier, the reason the beams were able to continue carrying load
after the shear section exceeded its capacity was due to the redistribution of forces on to
the continuing beam and the large amount of prestressing in the section. Thus, even
though the critical section had reached its shear capacity, the prestressing forces in the

beam were large enough to keep the section prestressed and thus able to redistribute

168

loads. Therefore, in spite of their ultimate global flexure dominated response all beam

tests allowed evaluation of the section shear capacity.

5.5 Experimental Evaluation Conclusions

This chapter presented a summary of the testing program aimed at evaluating the
flexural and shear response of prestressed beams made from self-consolidating-concrete
to determine their suitability for use in the M-50/US-127 demonstration bridge over the
Grand River. Four flexure and four shear tests were conducted on full-scale replicas of
the bridge box beams, one for each of the project mix designs, which were developed to
bound the current approaches to SCC mix design.

The flexural tests showed that the overall behavior of the SCC beams was very
similar to that of the conventional beam (NCC). Cracking patterns, widths, and failure
levels followed predicted responses through conventional prestressed concrete theory.
While the failure of the flexure test units was explosive, the beam was not over-
reinforced and the strands yielded before failure, technically defining the failure mode as
ductile. Consequently, variations in concrete compressive strength did not significantly
modify the beam’s ultimate capacity, which were very similar for all beams. The
absolute capacities of the SCC beams were marginally lower than that of the NCC beam,
specifically, 1.3, 3.4, and 3.5 percent lower for the SCC1, SCC2, and SCC3 beams,
respectively. However, the flexural capacities of the SCC beams exceeded the required
design capacity by 6 to 9 percent according to Table 44.

The shear behavior of the SCC beams was also found to be adequate and very
similar to that of the NCC beam. Cracking paths and widths were consistent in all beams

and the failure levels closely matched analytical predictions. The shear capacities of the

169

SCC beams were comparable to that of the NCC beam. The ratio of capacity to nominal
strength for the SCC1, SCC2 and SCC3 beams as 1.22, 1.12, and 1.08, respectively,
while that same ratio for the NCC beam was 1.11 according to Table 46. The Shear
capacity and response of the SCC beams is thus considered adequate.

Due to the large amount of prestressing and the extension of the beam beyond the
supports, failure of the critical section in shear did not result in overall member failure.
Rather, the beam was able to re-distribute demands beyond the critical shear area thus
allowing the beam to resist increasing loads that in two beams eventually lead to flexure-
induced failures and in two others the tests were terminates. However, even though the
overall member failure was not in a pure shear mode, the shear behavior and capacity
provided by the beams was successfully evaluated. This was determined by studying the
shear force versus shear strain response of the critical shear areas in the beam, which
Showed that the sections reached a peak shear force beyond which shear strains would
increase rapidly in an almost elastic-perfectly-plastic response. This same behavior was
supported by the moment-curvature response of this sections, which had essentially a bi—
linear response with a failure at a load level much below the known flexural capacity.
The reduced flexural capacity of the section is thus a consequence of the shear failure of
the section. In addition, evaluation of strain profiles of the steel stirrups in the critical
shear areas indicated that the transverse steel had yielded through the section depth for all
beams, further supporting that sections shear failure was reached.

Overall, the performance of the SCC prestressed box beams in flexure and shear
was found to be essentially equal to that of the NCC beam and their behavior was well

predicted by conventional prestressed concrete theory. While additional issues pertaining

170

to long-term behavior of SCC in prestressed elements, namely creep effects, are still to be
fully evaluated, the short-term flexural and shear response evaluated through this testing
program indicates that the SCC prestressed beams safely satisfy their prescribed design
requirements. The results of this testing program provided sufficient evidence to believe

the SCC beams could be used in the demonstration bridge as planned.

171

6 LONG TERM PERFORMANCE OF SCC BRIDGE BEAMS

6.1 Overview

The structural testing program described in the previous chapter assessed short-
term properties of bridge members produced with SCC. Results from these tests
indicated that the flexural and shear response of these members is essentially equal to that
of NCC members. Furthermore, it was shown that the capacities of the SCC members
exceeded the design capacity requirements for the demonstration bridge planned for this
project. The experimental program, however, cannot provide answers on questions
regarding differences in the long-term performance of SCC members compared to NCC
members under normal exposures to vehicle loading, weather demands, and long-term
material volume changes.

The performance of a bridge member over its service life is affected by many
external factors. These factors include external loading (including repetitive loading)
from vehicles, exposure to changes in weather, and exposure to deicing chemicals. Other
factors that affect this performance are related to the materials themselves. These include
concrete creep, permeability, and toughness and steel relaxation. Finally, the long-term
performance of the bridge member can be affected by the overall performance of the
bridge system. This includes the ability of the beams to expand and contract freely under
temperature changes. All of these factors will affect the long-term performance of the
beams.

The field-monitoring program for this project was designed to analyze how time

dependent factors affect the structural performance of SCC members. To this end, a

172

demonstration bridge consisting of six spread box beams was built and instrumented to
monitor the performance of both NCC and SCC beams.

Of the six beams composing the demonstration bridge, three were produced with
SCC and three were produced using NCC. The three SCC beams and one of the NCC
beams were instrumented to monitor changes in concrete strain and temperature. Two
types of instruments were installed in each beam to measure these parameters. Eight
Type T thermocouples were installed to record temperature, and eight vibrating wire
strain gages were installed to measure strain in the concrete. The results obtained from
the instruments can be used to assess the performance of the SCC beams.

The field monitoring system, including a description of the instruments and the
datalogging system, are presented followed by the results taken from the system. Finally,
five factors are investigated to explain the data recorded by the field monitoring system.
These include:

I Prestress losses

I Presence of diaphragm

I Temperature

I Bearing stiffness, and

I Data normalization.
6.2 Field Monitoring System

The field monitoring system consists of 64 gages connected to 3 Campbell
Scientific (Logan, Utah) datalogger. The datalogger was programmed to record data 12
times throughout the day (at equal time intervals) and store them until they can be

downloaded to a personal computer. The data was then post-processed to a more useable

173

format. The datalogger system, including the instrumentation, is described in more detail

below.

6.2.] Instrumentation

The purpose of instrumenting the field beams was to obtain long—term information
about the in-service performance of the elements cast with the different concrete mix
designs. To accomplish this, each field beam was provided with eight vibrating wire
strain gages (VWSG) to record changes in concrete strains and eight Type T

thermocouples to record temperatures.

6.2.1.1 Vibrating Wire Strain Gages (VWSG)

The VWSG used in this project were Geokon (Lebanon, New Hampshire) model
number 4200 gages, which are specially designed for embedment in concrete. They have
a range of 3,000 microstrains and a temperature range of 212 °F (-4 0F to 176 °F). VWSG
were selected for this project because of their rugged construction and their long-term
reading stability, or resistance to reading drift. However, the VWSG are slow to respond
and thus not suitable to record rapid changes in strain.

The VWSG were installed in the field beams at three sections along the length of
the beam as shown in Figure 85. Two gages were installed in the top and bottom flanges
at section A, and one gage was installed in each flange at sections B and C. The
nomenclature used to identify the gages is as follows: the gage located in the top flange at
Section C is named C-T, and likewise the gage located in the bottom flange at Section B
is named B-B. To distinguish between the gages at Section A, where two gages are

placed in each flange, a number 1 or 2 follows the gage name. Therefore, A-Tl

174

represents the first gage located in the top flange, while A-B2 is the gage in the bottom
flange as shown in Figure 86. To secure the instruments at the proper location, each gage
was attached to two pieces of #4 reinforcing steel and suspended between the prestressing
strands in the bottom flange and the mild steel reinforcement in the top flange. An

example of the installed VWSG before the concrete was cast is shown in Figure 87.

+—— 3. 7 m (12')
Section A Section 8 Section C
52. !

—-1 |—— .3 m(l')
15.8m(52')

 

 

7f“—
_lL_.____

a) Instrumentation Sections Field Beam

171 ll 1

W © 9

 

 

   

 

  

O Thermocouples
@ Vibrating Wire
Strain Gages

 

 

 

 

 

 

 

Section A Section B Section C
b) Instrument Locations for Field Beam

Figure 85. Field Beam Instrumentation

 

 

P 9
/ “.1 L... \

 

 

\ A-Bl I /’ A-B2 /
a; d

 

 

Figure 86. Naming Convention for VWSG

175

 

Figure 87 . Vibrating Wire Strain Gages Installed in The Top Flange

The VWSG, shown schematically in Figure 88, consists of a wire suspended
between two metal disks. An electromagnetic coil is attached to the gage at the midpoint
of the wire. The coil is used to excite, or “pluck,” the wire by inducing an
electromagnetic field. As the wire is excited it vibrates. The coil can measure the
frequency of the wire vibration. The coil changes the frequency of the wire vibration
through a set range. In this range is the resonate frequency of the wire. The coil
recognizes the resonate frequency and transmits this reading to the datalogger where this

frequency is converted to a strain measure.

176

Protective Instrument
Housrng Thermistorf Cable

   
 

L .L--. __7-___ 6" Gage _____I
7 Length

‘ s'r-j’éw 1:;

 

Figure 88. Schematic of Vibrating Wire Strain Gage

The VWSG measures the change in strain in the concrete by measuring the
change in the Wire’s resonate frequency. As the strain in the concrete changes the two
metal disks change position relative to each other. As this happens the tension in the wire
suspended between the discs changes, causing the wire to resonate at a different
frequency.

The VWSG used in this project also contain a thermistor in the coil housing. The
thermistor records the temperature at the measurement location. It is important to have
an accurate value of temperature-because the difference between the thermal coefficient
of expansion of the embedded steel wire and the concrete affects the accuracy of the
strain reading. In addition, readings are to be taken over time and they need to be
compensated for temperature at the time the measurement is taken.

As the temperature increases both the steel wire in the VWSG and the concrete
expand. However, because the coefficient of thermal expansion of steel is larger than

that of concrete, the wire will expand more than the concrete. This will appear to be a

177

compressive strain in the concrete, when this may not be the case. To account for this the

strain should be corrected for temperature according to Equation (25):

(25) 8 : AT(CTEsteel - CTEconcrete)

where AT represents the change in temperature in °F, CTESW is the coefficient of
thermal expansion of steel, and CTEmnm, is the coefficient of thermal expansion of
concrete both measured in microstrain/°F. The actual strain experienced by the concrete

is then given by Equation (26):

(26) Eactual : (8i — 80 )B + (Ii _ t0 )(CTEsteel — CTEconcrete)

where 8,- and t,- are the i'h reading of the strain and temperature respectively, 80 and to are
the initial readings of the strain and temperature. The constant B is a batch calibration
factor that corrects for wire clamping; which results in a higher reading than the actual
strain. The batch factor for the VWSG used in this project was provided by the

manufacturer as a value of 0.958.

The initial strain and temperature values in the unstressed condition of the beam
were recorded to be used as the reference values for future readings. These values were
taken during the beam production process after the concrete was cast but before the
prestressing strands were cut. To record the values, a Geokon (Lebanon, New
Hampshire) GK-403 Readout Box was used. This box is capable of exciting the wire and
recording the strain as well as the temperature.

The initial temperature and strain values for the gages are shown in Table 47
through Table 50. Two gages did not read properly at the time the strands were cut. The
gages were A-Bl in the NCC beam and GT in the SCC 2 beam. Fortunately, these

instruments began working properly once the beams were in place at the bridge.

178

Table 47. NCC VWSG Initial Readings

 

 

 

 

 

 

 

 

 

Instrument Temperature Strain
(°F) (micro strain)
A-Tl 97.2 2769.3
A-T2 96.3 2613.7
AB 1 96.3 -
A—B2 97.2 2752
B-T 90. l 2875. l
B-B 94.6 2949. l
C-T 90. 1 2779.8
C-B 132.0 2771.2

 

Table 48. SCC 1 VWSG Initial Readings

 

 

 

 

 

 

 

 

 

Instrument Temperature Strain
(°F) (micro strain)

A-Tl 133.9 3364.9
A-T2 133.7 3302.9
A-Bl 146.1 2244.5
A-B2 145.8 2472.3
B-T 120.7 2985.3
B-B 135.7 2436.8
C-T 118.4 2691.4

08 131.4 2740

 

Table 49. SCC 2 VWSG Initial Readings

 

 

 

 

 

 

 

 

 

Instrument Temperature Strain
(°F) (micro strain)
A-Tl 94.1 3421.3
A-T2 94.3 3405
A-Bl 102.0 2702.6
A-B2 102.0 2644.1
B-T 91.2 2815
B-B 99.3 2790.4
C-T 91 .4 -
C-B 98.6 2768

 

179

Table 50. SCC 3 VWSG Initial Readings

 

 

 

 

 

 

 

 

 

 

Instrument Temperature Strain
(°F) (micro strain)

A—Tl 115.0 3214.1
A-T2 116.4 3163.2
A-Bl 130.6 2581.4
A-B2 130.1 2678.2
B-T 94.6 3214.3

B-B 117.5 2746.1

C-T 91.9 2837.7

08 110.5 2672.6

6.2.1.2 Thermocouples

Thermocouples are used to measure temperature in their surrounding
environment. The instrument is made of two wires of dissimilar metallic conductors.
The choice of the materials affects the effectiveness of the thermocouple. The
thermocouples used for this project were Type T from Omega Engineering (Stamford,
Connecticut). Type T thermocouples are made of a positive Copper wire and negative
Constantan wire. The wires are welded together at the measurement end, and left
unconnected at the reference end. Figure 89 shows the measurement end of a Type T

thermocouple used in this project.

Blue — Copper (+)

 

Red — Constantin (-)
Figure 89. Measurement End of Thermocouple

180

The thermocouple is able to measure the difference in the temperature between
the measurement end and the reference end. When the ends of two dissimilar metals are
held at different temperatures a voltage develops between them. This voltage varies
according to the temperature in a known and regular way such that it can be converted to
temperature. The welded end of the wire is embedded into concrete to record the
temperature.

The thermocouples were set in place by attaching them to the reinforcing steel as
shown in Figure 90. The location of the thermocouples throughout the depth of the beam
are shown in Figure 91 Three thermocouples were installed at section A of the beam
(Figure 85). One thermocouple was in the top flange of the beam, the second was in the
sidewall of the beam at the mid height of the beam, and the third was in the bottom flange
of the beam. The remaining five thermocouples were installed once the bridge was being
erected at the bridge site. Three thermocouples were located in the deck of the bridge,
and two thermocouples were placed on the outside of the beam at section A of each of the
four instrumented beams. One of the external thermocouples was placed on the bottom
of the beam and the second was placed at the joint between the deck and the beam of

each of the four instrumented beams.

181

3.. .
as.
a.
:.
3‘:

‘31
. .1.
I"WTP¥\fi‘-' ":—
fl
q
A.
«a.
a
. :1
~31 TL
_ \f " ‘5‘
'4‘} H’s; t..— gib-
., 17,-}. , "‘
.‘h i :1"-
-' ax ‘f
‘3“‘4 wwwwwv'v-V‘M 1P ;-‘w
'3

 

Figure 90. Installed Thermocouple

rzzt'r" I

 

 

 

 

 

 

.1": W

Figure 91. Thermocouple Locations

182

6.2.2 Data Collection System

The automated collection system used to monitor data from the installed
instruments consists of one datalogger, four multiplexers, and one vibrating wire gage
interface. The system was automated to record data at specific times throughout the day.
External power sources provide the necessary voltage to execute the data collection
program.

The datalogger used is a Campbell Scientific (Logan, Utah) CR10x. The
datalogger is powered externally by a l2-volt rechargeable deep cycle marine battery
(model number SRM 29). The battery has a 100 amp-hour capacity and is recharged by a
10 W solar panel. The panel was placed atop a 12-foot pole that is attached to the eastern
return wall of the bridge abutment as shown in Figure 92. The solar panel has
dimensions of 16.5 in. by 10.6 in. and the voltage output from the solar panel at

maximum power is 17.5 V.

183

 

Figure 92. Full Bridge Picture Showing Solar Panel

Two types of multiplexers are used between the instruments and the dataloggers.
The role of the multiplexers is to expand the number of instruments that can be read by
the datalogger. Instruments connected to the multiplexers share a common channel on
the datalogger. The datalogger program advances through the instruments connected to
each multiplexer reading and storing each data point.

The 32 thermocouples are connected to two Campbell Scientific (Logan, Utah)
AM25T solid—state multiplexers that are specifically designed for thermocouples. Each
AM25T can hold 25 thermocouples. The vibrating wire strain gages are connected to
two Campbell Scientific (Logan, Utah) AM 16/32 relay multiplexers capable of handling
16 four-wire gages or 32 two-wire gages. Each of the 32 VWSG used in this project has

four wires thus two AM16/32 multiplexers were required The final component of the

184

datalogging system is a Campbell Scientific (Logan, Utah) AVW4 vibrating wire
interface. This device is placed between the datalogger and the multiplexers to enhance
the reading from the VWSG. These enhancements include expanding the number of
AM16/32 multiplexers that can be connected to the datalogger, removing noise from the
vibrating wire signal, completing the thermistor bridge needed to measure the
temperature with the VWSG, and increasing the peak-to-peak excitation of the VWSG
frequency.

The datalogger functions were automated with a program written using the
Campbell Scientific (Logan, Utah) datalogger support software PC208W program creator
EDLOG. The program is designed to execute every two hours starting at 12:00 am for a
total of 12 executions each day. Two hours is the maximum execution time allowed by
the datalogger. On each execution of the program each gage is read once and its value is
recorded by the datalogger. The vibrating wire strain gages record both strain and
temperature. This allows the strains to be corrected for the difference between the
coefficient of thermal expansion of the steel in the vibrating wire strain gage and the
concrete. Accordingly each time the program executes 3 types of data are recorded: the
strain from the strain gage, the temperature from the strain gage, and the temperature
from the thermocouples. There are 8 strain gages per beam and 4 beams total for a total
of 32 strain gages. There are also 8 thermocouples per beam for a total of 32
thermocouples. That means on each program execution 96 data points are recorded.
Each day 1,152 data points are stored by the datalogger and each month nearly 35,000
data points are stored. The datalogger has a final storage capacity of 62,280 data points.

Given the number of points stored every day, the system will reach its capacity after 53

185

days. Once the capacity of the system is reached, the earliest recorded data will begin to
be overwritten by new data points. Currently the system data is collected every month.
Each time the program runs it takes 1.5 minutes to cycle through all 64 instruments and
record the 96 data points.

The data is downloaded from the datalogger by connecting a laptop computer to
the datalogger using the PC208W software. The connection is made through a Campbell
Scientific (Logan, Utah) SC32B Optically Isolated RS-232 Interface. A cable converting
the 9 pin serial signal to USB is also used to retrieve the data.

The data was post-processed using a custom MATLAB (The Mathworks, Natick,
Massachusetts) program. The program was designed to separate and convert the data into
a more useable form. This includes correcting the strains for temperature and reducing
the data to the desired times. Of the twelve readings taken each day only four are kept
for analysis. These are the 12 am, 6 am, 12 pm, and 6 pm readings. The MATLAB
program plots and stores these values. The rest of the data is stored separately.

To connect the instruments to the datalogger the wires from each instrument were
routed to the north end of the beams and then to the datalogger location as shown in
Figure 93. Once at the end of the beam the bundle of wires crosses through the bridge
backwall. The wires from the NCC and SCC 1 beams and the SCC 2 and SCC 3 beams
were routed to the center of the back wall between each set of two beams where the
bundle of wires cross back through the back wall to the underside of the bridge. Once
under the bridge the wires were routed to the datalogger box, which was mounted on the

back wall underneath the bridge between the NCC and SCC 1 beams.

186

From the time the wires leave the beam they are encased in non-metallic schedule
40 PVC pipe so that the entire system is protected from moisture. The datalogger and
power source were housed in locked weatherproof PVC cabinets. The datalogger box
contains a back plate designed for connecting the system components. All components of
the datalogger system box are secured in this box. Both the datalogger and battery boxes
are connected to the back wall by threaded rods embedded into the concrete with a two-

part construction adhesive.

l.———— Effective s = 50' ———--I

North
Backwall
Bridge Deck

 

49. Datalogger

  

Instrument
ires

 

_lL

 

|~———— Beam Length = 52' El

Figure 93. Wire Routing Scheme

6.2.3 Observations and Results

The data recorded from the instruments in the demonstration bridge could provide
information on the performance of elements cast with SCC. Data showing similar
performances between the NCC and SCC beams would indicate that the SCC does not

behave differently over time when used in a prestressed bridge element. However,

187

differences in the concrete strain in the SCC beams could indicate problems that could

ultimately lead to an early failure of a bridge element cast with SCC.

Before monitoring of the instruments in the beams could begin at the bridge site,
four preliminary strain measurements were taken. Measurements were taken after
transfer of the strand force, after the top strands were cut at the beam centerline, once the
beams were in place at the bridge site (on top of abutment supports), and after the deck
was cast in place. These events, including the date they occurred, are summarized in
Table 51. The bridge was opened to traffic shortly after the deck was cast in the middle

of October 2005. Monitoring of the bridge data began on December 22, 2005.

Table 51. Preliminary Strain Measurement Events

 

 

 

 

 

Event Date
1 After strands were released 6-22-05 — 6-30-05
2 After top strand was cut 8-2-05
3 In place at bridge site 9-21-05
4 After deck was cast 10-2-05

 

The results of the strain measurements in the bottom flange at the midspan of the
beam, section A, at 12:00 pm for all of the beams are shown in Figure 94. AS discussed
previously, the first four points in the plot are the initial measurements taken during the
construction phase of the project. The monitoring of the instruments at the bridge site
began on December 22, 2005. Gaps in the plots, such as the gap in the SCC 3 plot in
Figure 94 during the month of April, are due to loose wire connections. Unfortunately,

these occurred occasionally throughout the year and were corrected as soon as possible.

188

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

200?
-300.

A O

2 . I

'8 -400- v

b ' '

in

9 v 0

.2 ‘ .O

2 -500- V

v .. .

.E

u

b .

‘D -600- o
+NCC
j—v-SC01
: . SCC3 Section A
83388888358888838
E55§§§ESSSSEESE§§

Date (6/22/2005-11/1/2005)

Figure 94. Strains In The Bottom Flange of Section A

The plots shown in Figure 94 show similar behavior over the period of
observation. While each beam has a different starting value the overall shape of each plot
is similar. As will be seen in future figures of the data from the VWSG at other locations
in the beams, the strain readings in all four plots began to become more compressive, i.e.,
more negative, near the end of the month of February 2006. Then during the month of
August 2006 the readings began to become more tensile, or less negative. The maximum
decrease in strain is anywhere from 100 - 200 microstrain during these time periods.

The strain readings in the top flange of Section A are plotted in Figure 95. Again
the first four points represent readings taken during the construction of the bridge. Again

while the overall performance of these beams seems to be similar, with similar shapes to

the responses, the starting point of these plots is different. Because of the different

189

starting points the scale of this plot is larger than the plot for the strain in the bottom
flange. Even though the starting values of the plots are very different, the overall
response is similar for the four beams. The continuous monitoring shows a decrease in
strain starting in the end of February 2005, and an increase in strain beginning in
September 2005. This decrease in strain is comparable to the decrease seen in the bottom

flange results.

 

 

 

 

0
-100 0 °
-200 .
A -300
3 400
- ' o
g 500 .
m -
9 O
3 -600 l :
E O
v -700
.E
2 -800-1 V
('7) :
-900 , '
g ~o— NCC
"1000 ‘ —v— SCC1 ’
+ 3002 C]

 

 

-1100 3

 

 

 

 

-1200

7/1/05 -
8/1/05 -
9/1/05 -
10/1/05
11/1/05 '
2/1/05 1
1/1/06 1
2/1/06 -
3/1/06 -
4/1/06 -
5/1/06 -
6/1/06 '-
7/1/06 1
8/1/06 -
9/1/06 -
10/1/06 -
11/1/06 -

Date (6iZZ/2005-11/1I2006)

Figure 95. Strains at Top of Beam Flange Section A

The decrease in strain shown in both Figure 94 and Figure 95 seems to be related
to temperature as the maximum decrease occurs during the warmest months of the year.
This can be verified by examining the data recorded by the thermocouples. Figure 96 and
Figure 97 show the recorded thermocouple temperatures at the top of the bridge deck and

in the bottom flange of the beams at 12:00 pm each day. As expected, the temperature

190

increased through the summer months. Table 52 gives the maximum, minimum, and
average temperatures recorded in the top of the deck for each beam. The minimum
temperature recorded was approximately 14°F in the middle of February 2006, while the
maximum—recorded temperature was 91 °F in August. The average temperature for all of
the beams was near 56 °F at the top of the deck.

The plot of the temperature in the bottom of the flange at Section A in Figure 97
is very similar to that at the top of the deck. The maximum, minimum and average
temperatures are reported in Table 53. At the bottom of the beam the maximum
temperature occurred on the same day as in the top of the deck, and was found to be 86
°F. The minimum temperature in the bottom of the beam was found to be near 17 °F. to
the temperatures in the top flange. The average temperature at this location was found to

be near 55 °F.

191

 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

35 ‘ .
3 l 90
30:
5 1 80
25 1: h
A : 70 A
0 20 1 IL
L 3 a,
“5’ 15 f _ 60 g
OI. - ii.
9 : ' 9
a 10 - 50 8.
s : E
i- 5 '. . 40 l-
0 .
H— NCC _ 30
+ SCC 1 ;
'5 + $002 -_ 20
+ SCCS :
'10 l I I I I I l I I l
8 3 8 3 8 8 8 8 8 8 8
E- 5 E $ 3 E 5 Es S E E
Date (1 2/21/2005-1 1/1/2006)
Figure 96. Temperature at Top of Bridge Deck at Section A
35 . l-
30 - .
; i I i‘ l :
I 1 .. 1' ‘ " 80
25 A 11“ ' N} . i :
3 l h" 7“ ' "H t
A I t ’ ‘ _
o 20 1 I i t q‘ It 70 E
L : il.‘ . . I 2’
9 151 "1"11; 1'» :60‘5’
a j ' l, 1 I . - a
2 1 l t 2
a 10 .1 ,1 ,1 50 a
r. : ' I .5.
3 t l :
o .1 l .
: l' —-e— NCC :30
‘ + $001 :
Section A + 8C03 i
'10 l fi I I l I I I I I
8 S .8 8 3 8 3 .8 8 S .8
E E1 5 S E S E E 5 E E

Date (1 2/21/2005-1 1/1/2006)

Figure 97 . Temperatures in the Bottom Flange at Section A

192

Table 52. Maximum, Minimum, and Average Temperatures At Top of Deck

 

 

 

NCC SCC 1 SCC 2 SCC 3
Max Temp (°F) 91.6 93.1 89.1 90.6
Minimum TempJ°F) 14.5 17.2 14.8 17.7

 

58.0 55.3 56.9

 

Average Temp (°F) 56.3

Table 53. Maximum, Minimum, and Average Temperatures At Bottom of Beam

 

 

 

 

NCC SCC 1 SCC 2 SCC 3
Max Temp (°F) 86.1 86.4 86.1 86.0
Minimum Temp (°F) 17.1 17.5 17.7 17.7
Average Temp (°F) 54.8 54.9 57.3 54.5

 

Figure 98 through Figure 101 show the plots of the strains in the top and bottom
flanges of sections B and C, which were located 12 inches and 12 feet from Section A,

respectively. These plots are very similar to the plots of the strain at Section A, as similar

trends appear in each case.

 

 

Microstrain

 

 

 

 

 

 

 

 

 

 

.800IIITIITIrijIIfif
38333338383533338
E§§§§§SSSSESES§§§

Date (6122/2005 - 11I1I2006)

Figure 98. Strains in the Top Flange of Section B

193

 

- 8:::

r 8:?

. 8:5

1 8:3

- 8::

. 8:8 Wu

. 8:: m
h

. 8:3 H

.
r 83mm
- 8E~ m

 

u $o>> M

(

 

. 3:3.”
0 D
r 8:::

 

 

 

 

 

 

. . r $9Eo_.

 

1 $9 Em

 

3
C
C
S 1 83m

1 m9 Ch

 

 

 

 

£83222

Figure 99. Strains in the Bottom Flange of Section B

 

 

 

 

 

 

 

 

 

 

59:38::

 

$9 C _. F
$9 (or
$9 Cm
$9 {m
$9 Eh
$9 _.\$
$9 Em
$9 {v

$9 Em
$9 (N

$99?

a (BIZZIZOOS - 11I1I2006)

.828 m

D

$9 9 w w
$9 to _.
m9 Cm
99 Co
$9 In

in the Top Flange of Section C

rains

Figure 100. St

194

 

-400

'500 -1
4 .
1 n . M
l
‘ V
-800 j v V x
‘ ' "1‘”

-1000 - O

Microstrain

 

 

 

 

1+ 8001
"ij+ SCC2

 

 

 

 

 

 

 

 

 

2/1/06 ‘
3/1/06 —
4/1/06 -
5/1/06 -
6/1/06 -
7/1/06 -
8/1/06 -'
9/1/06 ‘
10/1/06 ‘
11/1/06 ‘

1/1/06 "

O
I
10
E?
v—
R
1-

11/1/05 ‘

10/1/05 -

Date (6/22/2005 - 1 1/1/2006)

Figure 101. Strains in the Bottom Flange of Section C

6.2.4 Discussion

In order to verify the plots presented above, as well as to investigate some of the
differences that exist between the responses of the beams for the four concrete mix
designs, several factors that could affect this response were investigated. These include:

I Prestress loss calculations.

0 The expected prestress loss was calculated using multiple
methods. These methods ranged from lump sum analysis to time
dependent calculations. The plot of the strain in the section due
to the prestress loss should match closely to the measured values

as presented above. These calculations can act to verify that the

195

measured strain response is correct. The prestress loss
calculations are discussed in Section 6.3.
I Influence of Diaphragm.

o The influence of an 8 inch diaphragm located at the center of each
beam was analyzed. This analysis was conducted to see how the
presence of this diaphragm could change the strain measured in
the beam. The analysis and results are presented in Section 6.4.

I Influence of Temperature.

0 The increase of temperature in the summer months could affect
the strain measured in the beam. An temperature analysis looking
at both uniform and gradient temperature induced strain was
conducted with the results presented in Section 6.5

I Influence of Bearing Stiffness.

o The stiffness of the bearing, including all components of the
bridge system, could affect the strain measurements in the beam
by preventing the beam from expanding. The bearing stiffness is
investigated to see how it could influence the measured strain
response. The results of this analysis are presented in Section 6.6

I Data Normalization.

o The presented response of the strain in the top flange of the

beams show large differences in the starting values for the four

concrete mix designs. In an attempt to eliminate this discrepancy,

196

the data was normalized to remove the influence of the initial

value. The results of this analysis are presented in Section 6.7.

6.3 Prestress Losses

The performance of the bridge beams could be negatively affected by differences
in the rate of prestress loss, which could lead to larger than expected deflections and the
early onset of crack formation. This could potentially lead to premature failure of the
bridge members. Changes made to achieve the beneficial prOperties of the SCC mix
designs could adversely allow these mitigating factors to produce the negative effects.
Thus, field monitoring of the SCC bridge members could provide information on a
developing problem.

The long-term strain readings taken from the field instrumentation need to be
interpreted for any differences in performance between the SCC beams and the
instrumented NCC beam. While this judgment can be made qualitatively from the
measured data, it is of interest to estimate if the measured strains coincide with the
expected behavior. Further, if differences do exist, it is of interest to determine what
long-term effect may be causing them.

Since this project deals with prestress elements, the strains in the beam element
include not only the effects due to external loading and temperature effects but the
stresses introduced by the prestressing steel. Furthermore, it is well known that the large
forces in the steel strands and the sustained compressive stress state in the concrete will
lead to volume-related material changes in the steel and concrete. These changes will in
turn change the self-equilibrating force state in the prestressed section. The manifestation

of these effects is the loss of prestress in the section. This will clearly affect the readings

197

in the strain instruments and thus it is of interest to estimate prestress losses to better
compare the predicted and measured strain measurements in the instrumented beams.

Over time the prestressing force applied to a structural member will decrease due
to changes in the steel and the concrete. Some losses happen immediately and others
happen over the life of the bridge element. If not properly accounted for in a design, the
loss of prestressing could compromise serviceability or even cause failure of the element.

The losses that occur in prestressed elements can be attributed to four main
sources:

I elastic shortening,

I steel relaxation,

I shrinkage of concrete,
I creep of concrete.

Elastic shortening is an immediate prestress loss that occurs at the time of transfer
of the prestress force to the concrete. When the force is transferred to the concrete the
element instantaneously shortens. Since the concrete is bonded to the strand when this
shortening occurs, the strand also must shorten by the same amount. The shortening of
the prestressing strand causes the force in the prestressing strand to decrease as well. The
amount that this force decreases depends on the size of the force applied to the concrete
and the ratio of the modulus elasticity of the prestressing strand and the concrete.

Steel relaxation, or creep of steel, is a time dependent loss. This loss occurs as a
constant stress is applied to the steel. Two types of relaxation are calculated. An initial
amount of relaxation is generally calculated between the time the strand is stressed and

the time the force is transferred to the concrete. This value is important as it reduces the

198

amount of force transferred to the concrete. This in turn affects the amount of
prestressing force lost due to elastic shortening. Relaxation is then considered from the
time of transfer to the end of the bridge service life.

The prestressing force lost due to the drying shrinkage of the concrete is another
time dependent loss. Concrete shrinkage occurs over time as free water evaporates from
the concrete. The spaces previously occupied by water crush causing the concrete to
shorten [32]. The shortening of the concrete also shortens the bonded prestressing strand,
which causes a reduction on the prestressing force.

The final contribution to the loss of prestressing force is the time dependent loss
due to creep of the concrete. Concrete creep occurs as a compressive stress is applied to
the concrete over a period of time. The strain that occurs in the concrete as a result of
this stress causes the concrete to shorten, again shortening the bonded prestressing strand,
and lowering the prestress force.

Accurately estimating the amount of prestressing force lost over time depends on
the ability to accurately predict the amount of strain produced due to the previously
described prestressing loss sources. Many models have been produced for this purpose.
Each relies on empirical evidence from research data. While differences exist between

the various models, accurate results can be obtained from them.

6.3.1 Theoretical Considerations

Multiple methods have been developed to account for the prestressing losses.
Each method takes into account the same parameters but may look at them with a
different material model. The main differences between the methods are in the

estimation of the time dependent losses. Some methods try to estimate these losses as

199

one lump sum value. Other methods try to predict each time dependent loss separately.
The accuracy of these methods depends on the models used to predict the strains from the
time dependent phenomenon. Five methods were used to analyze the prestressing loss
expected for the members in the demonstration bridge. The first three methods are from
the 2007 AASHTO-LRFD Specification (AASHTO-LRFD-3) [4]. A fourth method is a
time-step analyses using time dependent models described by Naaman [34]. The final
method comes from the 1998 version of the AASHTO-LRFD Specification (AASHTO-

LRFD-Z) [3].

6.3.1.1 AASHTO-LRFD-3 [4] Provisions

Three methods for calculating the total prestressing loss are presented in the
AASHTO—LRFD-3 Specification [4]. These methods include two lump sum estimates
where the total prestress loss is using one equation. The third method, a refined method,
uses separate equations to calculate each of the time dependent losses.

The lump sum estimate presented in the AASHTO—LRFD-3 Specifications [4]
was also previously presented in the AASHTO-LRFD-2 Specification [3]. This method
estimates all of the time dependent values in one lump sum term. The predicted values
rely on the partial prestress ratio (PPR), which is a weighted average of the yield stress
for the prestressed steel to the yield stress of the total steal in the beam. This method
applies to prestressed members that are made with:

I concrete with a strength higher than 3,500 psi,
I normal weight concrete,

I concrete that is either steam or moist cured,

200

I prestressin g strands or bars.

The AASHTO-LRFD-3 Specification [4] states that for this method to be applied,
the member should be constructed in average exposure conditions with average
temperatures. According to the AASHTO-LRFD-3 Specification [4], this method was
developed using a computerized time-step analysis of many bridges and building
members designed and built with an average range of variables including creep
coefficient, concrete shrinkage, relative humidity, moist or steam cured, and, PPR ratio.
When low relaxation steel is used, as is the case for this project, the lump sum values can
be reduced by 4,000 psi.

According to the lump sum method the upper bound of the total time dependent
prestressing loss for a box beam with 270 ksi prestressing strands is given by Equation
(27):

(27) Afp = (21.0 + 4.0PPR - 4.0)x1000 (psi),

where A]? is the time dependent prestress losses from creep, shrinkage, and steel

relaxation. The PPR is the partial prestress ratio given by Equation (28):

APSny
Apsfpy + Asfy

 

(28) PPR =

where AP, is the area of the prestressing strand, fpy is the yield strength of the prestressing
strand, A, is the area of the non-prestressed steel reinforcement, and fy is the yield
strength of the steel reinforcement. The average value of this lump sum estimate is given

by Equation (29):

(29) Afp = (19 + 4.0PPR - 4.0)x1000 (psi).

201

To estimate the total loss of prestress the elastic shortening losses must be added
to this lump sum estimate of the time dependent losses. This loss is given by AASHTO-

LRFD-3 as Equation (30):
E p _
(30) AprS = +fcgp (PSI),
Em 7

where Ep is the modulus of elasticity of the prestressing steel, EC,- is the modulus of
elasticity of the concrete at transfer, and figp is the total compressive stress acting at the
center of gravity of the prestressing tendons at transfer. This stress includes effects from
the prestressing force at transfer and the self-weight of the member. The lump sum

method thus provides a rough estimate of the prestressed loss in a quick manner.

A second lump sum estimate that is given in the AASHTO-LRFD-3 specification
is shown in Equation (31):

fmAmr .
(31) Af1] =10-0—YH 75f +12-07H YST +AfPR (RSI)

c—Dwu
where fin- is the initial stress applied to the strands, AP, is the area of the prestressing strand
and, App,” is the non-composite, transformed area of the beam. )4, is a correction factor

for the humidity at time of casting and is given by Equation (32):

(32) n, =1.7—0.01H

with H = 75 for the region where this bridge was constructed. ygr accounts for the

shrinkage of the concrete and is given by Equation (33):

5
1+fmi

 

(33) J’sr =

202

where f”..- is the strength of concrete at the time of transfer. If this value is unknown it
can be estimated as 80% of the 28-day compressive strength. Finally Apr is the

relaxation of the steel, which is estimated as 2,400 psi for low relaxation steel in the

AASHTO-LRFD-3 specification [4].

The refined method used to predict the amount of prestress lost over time as
presented in the AASHTO-LRFD-3 specification [4] calculates losses over two time
intervals. The first interval is from the time the prestressing force is transferred to the
concrete to the time the deck is cast in place, and the second interval is from the time the
deck is cast in place until the end of the life of the bridge. For simplicity, the end of the
life of the bridge is taken as infinity. This method was first reported by Tadros et. al. in
“NCHRP Report 496: Prestress Losses in Pretensioned High-Strength Concrete Bridge
Girders” [51] before being adopted into the AASHTO—LRFD-3 Specification [4]. In this
method if transformed section properties are used then losses due to elastic shortening is
already accounted for, thus they do not need to be calculated directly. Therefore, in this
method all variables for area, moment of inertia, eccentricity, and neutral axis depth are
assumed to be taken from the transformed section properties.

The total prestress loss calculated using the refined method is found by Equation

(34):

(34) Am = (Afpm + Afpm + MW 1,, + (Afpsp + AprD + Afpm — Afpsp )df (ksi).
In this equation, and throughout this method, the subscript I represents the time at
transfer, the subscript d represents the time at deck casting, and the subscript f represents

the final time or infinity. Thus, the interval id is from the time the prestressing force was

transferred to the strand to the time the deck was cast, and the interval df is from the time

203

the deck was cast to the time infinity. The other variables in this equation represent the
individual losses due to each time dependent factor.
The shrinkage loss from the time of transfer to the time the deck is cast, Afim‘k is
given by Equation (35):
(35) AfpSR = Ebid Epskid (ksi).
8b).; is the shrinkage strain from transfer to the time of deck casting and is given by

Equation (36):
(36) ebid = 480x 10’6 k,(,k,,skskf

where, k“), k;,_,, k,,., and kf, are factors that depend on time, humidity, shape, and concrete
strength respectively. These factors are found using Equations (37) - (40) where all
variables have been defined except for t which is equal to the number of days in the
interval, and V/S which is equal to the volume to surface area ratio.

I

 

 

37 k =
( ) "1 61—4f'a-+t
5
(38) k = '
f l'1'fci

(39) k, = 1.45 4.136%)
(40) khs = 2.00 —0.014H

The final variable in the shrinkage loss equation is kid which is given by Equation (41):

l

 

(41) kid 2 2
E A A 6
ӣ423. 1+-gl_es.[1+o.7.,/,,,.]

ci g g

204

where all variables have been previously defined except for epg, which is the eccentricity
of the prestressing steel, and ijf, which is the creep coefficient for the interval from

transfer to final. This variable is found using Equation (42):

—0.1 18
(42) m = 1-90(ks)(khc)(kf 1km),-
where t,- is the time at the beginning of the interval and k)... is the humidity factor which is
found using Equation (43):

(43) km =1.56—0.008H.

The loss due to concrete creep for the interval from transfer to the time the deck is

cast is given by Equation (44):

E
(44) AprR =fi‘fcgpwbidkid (ksi)

Cl
where flgp is the stress at the centroid of the prestressing strand, and (11M can be found
using Equation (42).

The loss of prestress due to steel relaxation for this first interval is found using

Equation (45):
(45) Afpm = fkfl'[fi—o.ss] (ksi),
L fpy

where f,”- is the initial prestress in the strands, fpy is the yield strength of the prestressing
strands, and kL is equal to 30 for low relaxation steel.

The loss of prestress over the interval from the deck casting to the final time in the
life of the bridge is calculated in a very similar way to the previous interval. The

shrinkage loss is found using Equation (46):

(46) AfpSD = Ebdepskdf (kSl).

205

The difference between this equation and the previous shrinkage loss equation is
that the value of the shrinkage strain, Ebdfis found using Equation (47):
(47) Ebdf = Ebif _gbid
with each variable found using Equation (36). Also, the value of k4; is found using
Equation (48):
1

2
6
”mp ”mm” [How/W]

ci comp comp

 

(48) kdf:
E A A
1+ psi 1+ c

 

 

with the only difference from the calculation of kid is the use of composite transformed
properties.

The loss of prestress due to concrete creep over this interval is found using
Equation (49):

EP EP5

(49) AprD = Eifcgp [Vlbif _Wbid 1kdf + E

 

Afch/fbdf kdf (ksi).

C1 C1

with the only new variable being the Afpd value which is the difference between the stress
caused by the non composite dead loads and the composite dead loads. The loss of
prestress due to steel relaxation over this interval is taken to be equal to the loss of steel

relaxation over the previous interval thus Equation (50):
(50) AprZ 2 Aprl (kSI).

The final component of the prestress loss for the interval from the casting of the
deck to the final time in the life of the bridge is the loss of prestress due to the shrinkage
of the deck concrete. This component accounts for the fact that the deck and beam act

compositely, thus any volume related material changes that occur in the deck should

206

affect the beam as well. The loss of prestress due to the shrinkage of the concrete in the

deck is calculated using Equation (51):

Eps .
(51) AfpSS = E—Afcdf kdf (1 + OJV/bdf) “(51),

(I

where the only new variable, AIM/is found using Equation (52):

 

Edd Ad kE d 1 6 ed '
(52) Afcdf = f ec C _ pg (km)
1 + 0.71/lddf Acomp comp

where Eddy- and VIM- are calculated using the deck properties and ed is the eccentricity of

the deck.

6.3.1.2 AASHTO-LRFD-Z [3] Provisions
The total prestressing loss for a bridge beam according to the AASHTO-LRFD-2

Specification [3] can be estimated by Equation (53):

(53) 11pr = 3pr + AfpSR +23pr (psi),

where Apr is the total prestressing loss, Afim is the prestressing loss due to creep of the
concrete, AfpgR is the prestressing loss due to the shrinkage of the concrete, and 21pr is the
prestressing loss due to the relaxation of the prestressing steel. All of these terms are
time dependent as they occur over time during the life of the bridge. The instantaneous
losses are not considered explicitly in this equation. This is due to the fact that
transformed section properties are used in the calculation of the losses. If gross section
properties were used to calculate this loss, the instantaneous loss due to elastic shortening

would have to be included.

207

The refined method of analysis of prestress losses is described in the AASHTO-
LRFD-2 Specification [3]. This calculation for total prestress losses uses separate
estimates of the time dependent losses. The method only applies to members that have a
span of less than 250 feet, and are constructed with normal density concrete with
compressive strengths greater than 3,500 psi at transfer. While this method was replaced
in the AASHTO-LRFD-2 Specification [4] by the previously discussed refined method it
is discussed here for comparison purposes. For purposes of identification this method
will be labeled AASHTO: Old Refined Method. According to this old refined method
the prestress loss due to concrete shrinkage is given by Equation (54):

(54) AfpSR = (17.0 — 0.150H)x1000 (psi),

where H is the average annual ambient relative humidity. The value of H depends on

location and is given in the AASHTO-LRFD-2 Specification [3] as 75% for Michigan.

The prestress losses caused by creep are calculated using Equation (55):

(55) AprR = (12.0 fcgp - 7.0Afcdp)><1000 2 0 (psi),

where flgp is the stress in the concrete at the level of the prestressing strand, and Afidp is
the change in the concrete stress at the level of the prestressing strand due to the
application of permanent loads except for the loads that are in place at the time of
transfer. For the demonstration bridge in this project the deck and haunch are the only
load that affects the Afidp.

The final time dependent loss, the relaxation of the strands, is given in two parts

as shown in Equation (56):

(56) Apr = Apr1 '1' AprZ ([381).

208

The first part is the relaxation of the strand that occurs from the time the strand is

stressed until it is transferred to the concrete. This value is found using Equation (57):

__ log(24t)

(57> Afpm — [ij

f py

 

—0.55 - si ,
40.0 1pr (p)

where t is the time in days from the stressing of the strand to transfer, and f,,,- is the stress

in the strand at the end of stressing.

The second part of relaxation occurs from transfer throughout the service life of
the bridge. In the AASHTO-LRFD-2 Specification [3] refined analysis this value is

estimated by Equation (58):

(58) Apr2 = (2O.O—.4AprS —0.2(Afp5R +AprR))><1000 (psi),

where all variables have been described previously. Once each of the separate prestress

losses are calculated the total prestress loss can be found using Equation (53).

6.3.1.3 Time Step Analysis Methods

Conducting a time step analysis to calculate the total prestressing loss is valuable
for two reasons: the time step analysis allows for the prediction of the amount of prestress
remaining at different stages of the construction process, and it accounts for the
interdependence of the losses. As the time dependent losses occur at different rates, they
affect the amount of stress remaining to affect the loss due to another source [34]. For
instance, the majority of the shrinkage loss occurs at the beginning of the construction
process as the free water in the concrete evaporates. This causes less stress to be applied
to the concrete member at later ages, causing less creep strain to develop, and

consequently reducing the amount of prestress lost due to creep. A Time step analysis

209

uses time dependent models to calculate the strain in the concrete due to concrete creep,
and concrete shrinkage as well as the relaxation of the steel. Many models exist for this
purpose. Naaman [34] presents models for these calculations in his textbook. The
models he presents are taken from Prestressed Concrete Institute (PCI) and American
Concrete Institute (ACI) recommendations. The model presented by Naaman [34] will
be called Time Step Analysis and abbreviated TSA.

For the time step analysis transformed section properties will be used, as was the
case for the AASHTO Refined Method. However, in this calculation procedure the loss
due to elastic shortening will be calculated explicitly. This will be done with Equation
(30).

The suggested equation for the time dependent loss of prestress due to steel
relaxation comes from the PCI Committee on Prestress Losses. This loss is calculated
using Equation (59):

fs(ti) fsai) t' .
<59) Afpmtpti): ”45 [20y —o.55]log[j].psn

 

where fish.) is the stress in a prestressing strand at the beginning of the time interval
under consideration, tj is time at the end of the time interval, and t,- is the time at the
beginning of the time interval.

For the loss of prestress due to the shrinkage and creep of concrete the TSA
presents an equation that modifies the concrete strain caused by either concrete creep or
concrete shrinkage by a series of correction factors that account for the shape and size of

the element, the average humidity of the region, and the strength of the concrete.

210

The prestressing loss due to shrinkage of concrete according to the TSA is given
by Equation (60):

b(rj—t,-)
(b+qu+g)’

 

(60) AfpS (ti’tj) = EpsgsuKSH KSS pSi,

where 8,.“ is the shrinkage strain, K 5;) is a correction factor that depends on the average
relative humidity, K 55 is a correction factor for the shape and size of the member, and b is
a parameter that is taken to be 35 for moist—cured concrete. According to Naaman [34],
the time function and the value b is suggested by ACI Committee 209. The shrinkage
strain depends on the amount of water in the concrete mix design and is given by

Equation (61):
(61) a -—[2+~33—(w-220)]x10‘4
5“ 230 ’

where w is the water content in the mix design in lb/yd3. The correction factor K5,.
depends on the volume to surface ratio and can be found in the literature. For an average
humidity between 40% and 80% for moist-cured concrete, K5” is calculated using

Equation (62):
(62) KSH = 1.40 — 0.01H .

The loss of prestress due to concrete creep as presented by the TSA is given by

Equation (63):

(63) AprUiJj) = "pCCUKCHKCAKCSfCGS(0)180.” - 8001 (Psi)

where np is the ratio of modulus of elasticity of prestressing strand to the modulus of

elasticity of concrete, CCU is the ultimate creep coefficient, KC” is a correction factor

211

depending on average humidity, KCA is a correction factor depending on the age at
transfer, KCS is a correction factor depending on the size and shape of the member. CCU
depends on the compressive strength of the concrete and varies between 2 and 3.1. Km
for moist-cured concrete depends on the average humidity according to the Equation

(64):
(64) KCH = 1.27 —0.0067H .

where Kcs is the size and shape factor that varies depending on the volume to surface are

ratio. Km, for moist-cured concrete, is given by the Equation (65):

(65) KCA =1.25:;,°-“8

where IA is the time in days to transfer. The time function g(t) is given by Equation (66):

0.6

I
66 t =———
( ) 80 10+z°~6

where t is in days.

As was the case with the previous methods, the total prestress loss in the time step
method is the sum of each of the individual prestress losses. However, in the time step
analysis methods, this total is calculated at various stages throughout the life of the
bridge. Time is considered a variable and the stress is changed depending on the amount

lost in each step.

6.3.2 Observations and Results
To accurately calculate the loss of prestress using the previously discussed

methods the transformed section properties needed to be used. The transformed

212

properties account for the difference in stiffness between the prestressing steel and the
concrete. To calculate the transformed properties the area of steel was increased by the
value n-I. The value of n is found by taking the ratio of the elastic modulus of the
prestressing steel to the elastic modulus of the concrete. As the modulus of the concrete
varies throughout the life of the bridge as well as for the 4 concrete mix designs the
transformed properties were calculated at multiple times. The transformed properties
were taken at three different times. The first time was at the transfer of the prestressing
force. This corresponds to event one in the previously presented Table 51. The second
set of transformed properties were calculated at event two from Table 51. This was the
point at which the top strands were transferred. The final time the transformed properties
were calculated was at event four. This is the point at which the deck is cast making the
composite section.

The transformed properties for the four instrumented beams, which each have a
different concrete mix design, at the three events discussed above are shown in Table 54
through Table 57. The properties shown include cross section area, moment of inertia,
and neutral axis distance from the bottom of the section. These properties account for the

different concrete strengths recorded at each event.

Table 54. NCC Transformed Section Properties

 

Event Area (inz) 1(in4) y (in.)

 

 

 

1 528.57 49,277 13.22
2 526.13 49,024 13.27
4 1,390 163,376 24.61

 

213

areas and centroid of each row of strands.

Table 55. SCC 1 Transformed Section Properties

 

Event Area (inz) I (in4) y (in.)

 

 

 

1 528.92 49,314 13.21
2 526.00 49,012 13.27
4 1,391 163,970 24.59

 

Table 56. SCC 2 Transformed Section Properties
Event Area ($1) I (in‘) y (in.)

 

 

 

 

1 528.50 49,271 13.22
2 528.33 49,253 13.22
4 1,391 164,834 24.59

 

Table 57. SCC 3 Transformed Section Properties

 

Time Step Area (in!) I (in‘) y (in.)

 

 

 

1 528.68 49,289 13.22
2 528.31 49,251 13.22
4 1,392 164,368 24.57

 

In order to accurately estimate the force in the prestressing strands using the
previously described methods, several assumptions had to be made. The prestressing
strands from both the top and bottom flange were assumed to act as one strand with an
equivalent area to the total strand area. The assumed strand is assumed to be located at

the resultant centroid of all of the strands. This centroid was calculated using the strand

demonstration beams: the first row is two inches from the bottom of the beam and has 10
strands, the second row is four inches from the bottom of the beam and has four strands,
the third and fourth rows each have two strands, and are eight inches and 24.5 in. from

the bottom of the beam respectively. At the mid span of these beams, where all of the

214

There are four rows of strands in the

prestress loss analysis take place, the top two strands are debonded and thus do not affect
the calculation of the centroid of the strand at this location.

The presence of debonded strands at various locations in the beams made
calculating the stress throughout the cross section difficult. To account for the debonded
strands the axial load and moments about the neutral axis caused by the force in the
prestressing strands were calculated at several critical locations. The critical locations
were places where the strand bonding condition changed. These locations are shown in
Figure 102. Three prestressing conditions exist throughout the beam. The first condition
is for the top strands where two strands are bonded to the concrete over the first 13 feet of
the beam. This is represented from section A to section C in the figure. At section C the
strands are no longer bonded to the beam. The second condition is for the bottom
strands. It represents the first four feet from the end of the beam, section A to B. In this
section 14 of the 16 strands are bonded to the beam. Two strands in the bottom row of
prestressing strands are not bonded to the beam. The third prestressing condition occurs
from section B to section D, which is the mid point of the beam. Here all 16 of the
strands in the bottom flange are bonded to the beam and no top strands are bonded to the
beam.

The eccentricity changes for each of the four concrete mix designs because the
value relies on the transformed section properties. Thus, these values must be calculated
at the same three events that the transformed properties were calculated. Table 58
through Table 61 gives the eccentricities for the three different bonding conditions. Also
included in these tables are the eccentricities of the VWSG. This value is necessary to

calculate the strain at the location where the strain is being measured in the beams.

215

 

 

 

 

 

 

 

 

TOP PRESTRESS
/ sec. D BEAM
1 :71; _‘1‘ ________ __1i__::i;;i
e bot I ____-____ “"*—* w‘” .15“... J "m“
debond L» 4' 4%: 9' ~—-- 13' ————-i
BOTTOM
PRESTRESS

Figure 102. Prestress Strand Geometry

Table 58. NCC Transformed Strand and Instrument Eccentricities

 

Event Condition Condition Condition Top Inst. Bot. Inst.

 

 

 

1 2 3 (in.) (in.)
l 11.28 9.79 9.97 11.28 11.22
2 11.23 9.84 10.02 11.23 11.27
4 -0.11 21.18 21.36 -0.11 22.61

 

Table 59. SCC 1 Transformed Strand and Instrument Eccentricities

 

Event Condition Condition Condition Top Inst. Bot. Inst.

 

 

 

ll 2 3 (in.) (in.)
1 11.29 9.78 9.96 11.29 11.21
2 11.23 9.84 10.02 11.23 11.27
4 -0.09 21.16 21.29 -0.09 22.59

 

Table 60. SCC 2 Transformed Strand and Instrument Eccentricities

 

Event Condition Condition Condition Top Inst. Bot. Inst.

 

 

 

1 2 3 (in.) (in.)
l 11.28 9.79 9.97 11.28 11.22
2 11.28 9.79 9.97 11.28 11.22
4 -0.09 21.34 21.34 -0.09 22.59

 

216

Table 61. SCC 3 Transformed Strand and Instrument Eccentricities

 

Event Condition Condition Condition Top Inst. Bot. Inst.

 

 

 

1 2 3 (in.) (in.)
1 11.28 9.79 9.97 11.28 11.22
2 11.28 9.79 9.97 11.28 11.22
4 -0.07 21.14 21.32 -0.07 22.57

 

The axial load and moment about the beam neutral axis due to the prestressing
strand were calculated at each of the sections shown in Figure 102. The moments were
found using the eccentricities from the Table 58 through Table 61. These forces were
then summed to find the total axial load and moment about the neutral axis at the
centerline of the beam, which was the location where the analysis was to take place.
Added to this moment was the moment due to the self-weight of the beam, which acts in
the opposite direction than the moment caused by the prestressing strands. Like the beam
self-weight, the weight of the deck caused a moment that opposed the moment caused by
the prestressing force. This moment was considered in the stress calculations after the
deck was cast in place at the bridge site. The total axial load and moment were used to
calculate the stress in the concrete at the level of the strand using Equation (67):

PMC

(67) fcgs = X + 1 (p31)

where P is the total compressive axial load from the prestressing force, A is the area of
concrete in the cross section, M is the total moment acting about the neutral axis of the
beam from the prestressing force, the self-weight and deck loads at the centerline of the
beam, c is the distance from the beam neutral axis to the center of gravity of the effective

prestressing strand, and I is the moment of inertia for the beam.

217

Differences in the four concrete mix designs and the construction procedures for
each day of casting affected the results of the prestressing loss calculations. The
prestressing losses were affected by differences in the rate of concrete compressive
strength gain through the effect that this strength had on the modulus of elasticity of the
concrete. The strengths used to predict the prestress loss at the four preliminary strain
measurements are shown in Table 62. The concrete compressive strength given in the
table for the fourth event is the strength that is used to estimate the prestress loss for all
calculations made after this time.

The difference in the age of the concrete at transfer indirectly affected the losses
due to elastic shortening. The more time that elapsed between the stressing of the strands
and the transfer of the strand stress to the concrete, the stiffer the concrete was and
consequently, the less loss due to elastic shortening. This is evident in the NCC and SCC
2 cases where transfer occurred at 3 days after stressing because casting was done on a
Friday, and transfer did not occur until the following Monday. The age of the concrete at
each of the preliminary measurements are listed in Table 63.

The final effect that the mix designs had on the prestress loss calculations was the
model used to predict the shrinkage strain in the time step analysis method described by
Naaman [34]. This method relies on the amount of water included in the mix. Thus,
mixes with greater amounts of water resulted in higher predictions for shrinkage strains.
The average amount of water used in each cubic yard produced of the mix designs are

listed in Table 64.

218

Table 62. Concrete Compressive Strengths at Construction Events (psi)

Event NCC SCC1 SCC2 SCC3

 

 

 

 

 

l 6537 6338 6572 6471
2 8196 8292 6674 6685
3 8560 7549 7764 6968
4 8560 7549 7764 6968

 

Table 63. Concrete Age at Construction Events (days)
Event NCC SCC 1 SCC 2 SCC 3

 

 

 

 

 

l 3 0.875 3 0.917
2 36 41 36 33
3 86 94 86 83
4 97 102 97 94

 

Table 64. Mix Design Water Content for Shrinkage Loss Estimation (lb/yd3)

NCC SCC1 SCC2 SCC3
218 262 276 310

 

The total prestress loss results from the five methods are given in Table 65 and in
Figure 103. From these results it can be seen that the five methods produce results that
generally agree well with each other. The TSA and AASHTO-LRFD-3 refined methods
produce results that extend the range of values. With all of the methods considered the
prestress loss estimates from the five methods vary only slightly, with a difference

between 6,000 and 11,000 psi for each of the four mix designs.

219

Table 65. Total Prestress Loss (psi)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Method NCC SCC 1 SCC 2 SCC 3
AASHTO, Lump Sum Upper Bound [4] 35,293 35,490 35,256 35,362
AASHTO, Lump Sum Average [4] 33,293 33,490 33,256 33,362
AASHTO, Refined Method [4] 29,008 31,357 28,864 30,545
AASHT 0, Estimate [4] 32,655 33,241 32,544 32,551
Time Step Analysis [34] 31,649 37,497 35,781 41,225
AASHTO, Refined Method (old) [3] 34,373 33,712 34,348 33,736
I [E LRFD Lump Sum: Upper Bound [4]
60000 1
4 m LRFD Lump Sum: Avg [4]
1 [2:52 LRFD Refined [4]
50000 1 :3 LRFD Est [4]
<5 : [1111011 TSA [34]
5 - CED LRFD R f'
01 40000 . e1ned[3]
'0 I H .
3 ‘ is j 3' 1 ' F~ ‘ i 1
asooooa .~El. 41:: ”if: .1“.
2 . r7 _ _ ’\ e L 7x 4 4 wk 1 -
‘6 7 l N ,_ ,_ / N .. C \ 1.. p V w -
e ‘ ,4 N / N j 1’ \ j
a 20000"l VN “ *— ’{N - /\ .. . N H
. /\ 7 7 /\ 7 7 47 C 18 7 7
j ,N Z : ”k : j :7 4 r1 : :
100004 .~__ A.. ,3.. flp-
1 \ V N I N
2 1V” #177 "M p)“
1 N 7 7 / 7 7 ’7 / N r- 71
O . .
NCC SCC 1 SCC 2 SCC 3

Figure 103. Total Prestress Loss for Different Calculation Methods

The difference between the results from various methods can be further analyzed

by looking at the individual components of the prestress loss. As mentioned previously,

the total prestress loss is made up of four components in a prestressed bridge beam.

These components are: elastic shortening, concrete creep, concrete shrinkage, and steel

relaxation. Figure 104 through Figure 106 show how the total prestress loss is divided

into the individual components for each of the three prestress loss calculation methods

that consider the time dependent loss separately.

220

The prestress losses for the AASHTO-LRFD-3 refined method [4] are shown in
Figure 104. From this figure it is possible to see that the largest amount of loss comes
from concrete creep. Due to the fact that elastic shortening is not considered
independently in this method it is assumed that some of this creep loss accounts for this
loss. The amount of creep loss is controlled by the stiffness of the concrete. This
stiffness, the modulus of elasticity, is controlled by the concrete strength at the time the
prestressing force was transferred to the concrete. Thus, the concrete with the highest
strength, NCC, has the least amount of creep loss and the concrete with the least strength,

SCC 1, has the highest amount of creep loss.

 

[E Creep

EB Shrinkage
12525253 Steel Relaxation
- Deck Shrinkage

01
O
O
8

 

.441 IALLL_LI L

 

 

 

 

 

 

 

Prestress Loss (psi)

 

 

 

 

 

 

 

 

 

W7

 

-s\\\87

 

NCC SCC 1 SCC 2 SCC 3

Figure 104. Prestress Loss According to AASHTO-LRFD-3 Refined Method [4]

Figure 105 shows the prestress loss for the TSA method broken down into
separate components. As was the case with the previous method, the loss of prestress due

to the creep of the concrete represents a large portion of the total loss. However, in the

221

TSA method, a significant amount of the total prestress loss is accounted for in the elastic
shortening. However, it is important to note that both the creep and elastic shortening
losses are very similar for all four concrete mix designs. The TSA method produced the
largest range of total prestress loss values. The source for this difference lies in the
calculation of the loss due do shrinkage. The material model used to calculate the
shrinkage strain in the concrete depends on the amount of water used in the concrete mix
design. As the SCC mix designs higher than normal amounts of water this value may not

be reliable. More discussion on this shrinkage strain calculation will follow.

60000

[22 Creep
ESE Shrinkage

50000 m Steel Relaxation
[:1 Elastic Shortening

40000

30000

20000

Prestress Loss (psi)

1 0000

 

0

NCC SCC 1 SCC 2 SCC 3
Figure 105. Prestress Loss According to Time Step Analysis [34]

Figure 106 shows the final prestress loss calculation that considered time separate
losses for each time dependent component. The AASHTO-LRFD-Z refined method [3]

produced the results with the least amount of variability. These results also show similar

222

trends to the previous methods with concrete creep being the dominant factor in the total

prestress loss.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

E 3:51....
j m Steel Relaxation
a 40000-
1% 30000-
:3 20000-2 >\§ >\\ >\§
10000-3 7 // Q
03 /fl fl ///

NCC SCC 1 SCC 2 SCC 3
Figure 106. Prestress Loss According to AASHTO-LRFD-Z Refined Method [3]
In Table 66 the loss due to elastic shortening is shown for all methods. The
development of the AASHTO-LRFD-2 methods assumes the elastic shortening is
included inherently when transformed section properties are used, thus the elastic
shortening is only calculated in the TSA method [51]. This fact is verified by the
relatively small difference that exists in the total calculated prestress loss for the different
methods considered. The main material property that affects the amount of prestress loss
is the elastic modulus of the concrete at the time of transfer. This parameter is directly
affected by the compressive strength of the concrete at this time. Thus, the strongest
concrete, NCC, shows the least amount of elastic shortening loss. The weakest concrete,
SCC 1, shows the most prestress loss due to elastic shortening. The increased strength

that is seen in the NCC and SCC 2 concretes comes from the concrete age at the time of

223

transfer. While the stress in the strands in both the SCC 1 and SCC 3 beams was
transferred less than 24 hours after casting took place, the NCC and SCC 2 beams did not
under go transfer for 3 days. This difference in time has a significant affect on the

amount of prestressing force lost during the life of the bridge element.

Table 66. Prestress Loss Due To Elastic Shortening (psi)

 

 

 

 

 

 

 

Method NCC SCC 1 SCC 2 SCC 3
AASHTO, Lump Sum Upper Bound [4] 0 0 0 0
AASHTO, Lump Sum Average [4] 0 0 0 0
AASHTO, Refined Method [4] 0 0 0 0
AASHTO, Estimate [4] 0 0 0 0

Time Step Analysis [34] 14,292 14,490 14,256 14,362
AASHTO, Refined Method (old) [3] 0 0 0 0

 

Table 67 presents the loss of prestress due to the concrete creep. The three
models that predict concrete creep individually show large differences in the total value.
In general, the AASHTO-LRFD-3 refined method [4] produces the highest value, with
the AASHTO-LRFD-Z refined method [3] and TSA values further behind. The TSA and
LRFD refined method [4] show variability through the four concrete mix designs. The
AASHTO LRFD refined method (old) [3] has very little variation among the four mix
designs. This results from the equations used to calculate the creep. The old refined
method relied on the other loss parameters while the new refined method and the TSA
method both calculate the creep using material models. The main factors in these models
are the compressive strength of the concrete and time elapsed. NCC and SCC 2 have
lower results because more time elapsed between casting and transfer of the strands. This
increase in time also leads to an increased compressive strength. Other factors that affect

this value may have to do with the interdependence of the different losses.

224

Table 67. Prestress Loss Due To Concrete Creep (psi)

 

 

 

 

Method NCC SCC 1 SCC 2 SCC 3

AASHTO, Refined Method [4] 17,823 20,049 17,710 19,356
Time Step Analysis [34] 10,733 13,600 11,285 14,307
AASHTO, Refined Method (old) [3] 16,603 16,584 16,554 16,550

 

Table 68 gives the prestressing loss due to concrete shrinkage. These values do
not show the same range of variance as the creep values. It is also clearer as to what is
governing the results. The AASHTO-LRFD-2 and AASHTO-LRFD-3 refined methods
calculates the prestressing loss due to concrete shrinkage based on the average relative
humidity. The TSA method, also depends on humidity, but uses a shrinkage strain model
that depends on the amount of water used in the concrete. A parametric study on this
equation shows that the amount of the water in the mix hasa large affect on the total
prestress loss through the amount of shrinkage loss. An increase in the amount of water
of 10 pounds per cubic yard of concrete results in an increase in the total prestress loss of
500 psi to 700 psi. For the SCC 3 mix design that has 100 pounds of water per cubic
yard of concrete more than the NCC mix this is a significant amount of loss. It is
believed that the material model used to predict the shrinkage strain in the TSA method is
not correct for SCC. In general a mix design with higher amounts of water will produce
greater amounts of shrinkage. However, in developing SCC mix designs this is
considered. In order to prevent this increase in shrinkage SCC mix designs use VMA.

The affect of the VMA is not considered by this material model. The model proposed by

the AASHTO LRFD refined method [4] does not have this same problem.

225

Table 68. Prestress Loss Due To Concrete Shrinkage (psi)

 

 

 

 

Method NCC SCC 1 SCC 2 SCC 3

AASHTO, Refined Method [4] 7,063 7,177 7,035 7,068
Time Step Analysis [34] 2,601 5,780 6,390 9,091
AASHT O, Refined Method (old) [3] 5,750 5,750 5,750 5,750

 

The losses due to steel relaxation are shown in Table 69. Large differences are
shown between the AASHTO-LRFD-3 refined method [4] and the TSA methods and the
AASHTO-LRFD-2 refined method [3]. This is because the steel relaxation loss
calculated in the AASHTO-LRFD-2 method relies on the three other factors (concrete
creep, elastic shortening, concrete shrinkage) where as the relaxation in the AASHTO-
LRFD—3 refined method [4] and the TSA methods are calculated directly using

essentially the same equation.

Table 69. Prestress Loss Due To Steel Relaxation (psi)

 

 

 

 

Method NCC SCC 1 SCC 2 SCC 3

AASHTO, Refined Method [4] 3,170 3,170 3,170 3,170
Time Step Analysis [34] 4,023 3,628 3,850 3,465
AASHTO, Refined Method (old) [3] 12,020 11,378 12,044 11,436

 

The final factor considered for prestress loss is the shrinkage of the deck concrete.
This is only considered in the AASHTO-LRFD-3 refined method [4]. While the value of
this loss is relatively low it is important to consider this value in bridges where the deck

is cast compositely. The other methods do not explicitly consider this method.

226

Table 70. Prestress Loss Due To Shrinkage of Deck Concrete (psi)

 

 

 

 

Method NCC SCC 1 SCC 2 SCC 3
AASHTO, Refined Method [4] 953 960 949 951
Time Step Analysis [34] 0 0 0 0
AASHTO, Refined Method (old) [3] 0 0 0 0

 

6.3.3 Comparison to Measured Data

While analytical models are useful to predict well-understood phenomenon, their
accuracy may be reduced when new parameters are introduced. This is the case for the
prestressing loss calculations for beams produced with SCC. Even with the different
ways the concrete mix designs are factored into the calculations there is no way to be
certain that these predictions are accurate for beams made with SCC. Much like the case
of bond between the prestressing strand and the concrete, changes made to the mix
designs may affect the loss of prestress in prestressed beams. Verification of these

predictions requires the use of information from the beams themselves.

6.3.3.1 Theoretical Considerations

The data recorded from the instruments in the demonstration bridge could provide
information on the performance of elements cast with SCC. Differences in the concrete
strain in the SCC beams could be from a difference in the prestress loss or deflection of
the beam. The data recorded from the instruments in the demonstration bridge may
provide information on the amount of stress remaining in the prestressing strand. During
the calculation of the prestressing losses the strain in the cross section can be calculated
using the well known Hooke’s law for one dimensional response which is given as

Equation (68):

227

a
68 £=—,
( ) E

where 8represents the strain in the cross section, 0'is the stress in the cross section at the

level of the gage, and E is the elastic modulus of the concrete for the given time step.

6.3.3.2 Discussion

A comparison of the analytical and measured strains for the bottom VWSG in the
NCC beam is shown in Figure 107. From the figure it is possible to see that the
measured first data point does not match the analytical data point. Possible reasons for
this difference could stem from the effect of the unbonded strands in the top flange of the
beam. An analysis to asses the role of these strands on this first point is on going.
Further, the behavior of the measured data during the monitoring period, after December
20th, 2005, does not match the predicted behavior from analysis. The increase in
measured compressive strain seen in the plot could be the result of temperature effects.
From the plots of the temperature presented before it can be seen that the increase in
temperature that occurs near the end of February coincides with the beginning of this
compressive strain growth. Also, the minimum strain point that occurs around the 400th
day after transfer of prestress force, which is in the month of September coincides with
the beginning of the temperature decrease seen in the thermocouples. An analysis of the
affect of temperature gradient across the composite section of these beams is on going

and will be added to this discussion.

228

 

 

Strain (microstrain)

 

 

 

 

 

 

: -0- Analytical Strain C]

. + Measured Strain

'600 v v I v 1 v v 1 v I v w v u l i I i i l . I r t 1 . f 1 ,
O 100 200 300 400 500 600

 

 

 

 

 

 

 

Days (from transfer)

Figure 107. Strain Comparison For Bottom VWSG NCC Beam Section A

Figure 108 shows the comparison of the analytical strains and measured strains
for the VWSG at the top of the NCC beam at section A. These points seem to agree well
for the four initial measurement events. However, once the continuous monitoring
begins, significant differences between the measured values and analytical values exist.
The sudden increase in strain shown in the analytical values is caused by the inclusion of
the deck in the composite beam section. Once the deck is in place on the beam the
section that resists the loads is much larger. The neutral axis of this section moves up
towards the top of the beam. The neutral axis for the composite section is less than 6
inches from the top gage location. Because of this, in the analysis, the stress due to the
moment caused by the prestressing strand at the top location is very small. Consequently

the strain is reduced. The effect of temperature may improve this plot.

229

 

 

Strain (microstrain)

 

 

 

 

,
1 —0- Analytical Strain C]

- + Measured Strain
-500 I T I I I I I I I I I r I f I I I I I I I I I I ' I I l

O 1 00 200 300 400 500 600

 

 

 

 

 

 

 

 

 

Days (from transfer)

Figure 108. Strain Comparison for Top VWSG NCC Beam Section A

Figure 109 through Figure 114 display the comparisons of strains for the SCC
beams. These plots are similar to the plots shown for the NCC beam. The main
difference is in the plots of the comparison of strains from the top flange of the SCC
girders. In the NCC plot, the analytical strains and the measured strains showed good
agreement. In the SCC plots, the plots do not show the same agreement. In these plots,
the shape of the curves is very similar, but the starting value is not. The difference in the
values varies from 600 micro strain. to 700 micro strain. However, the reason for this

difference could not be determined.

230

 

 

 

 

 

 

 

-300 ‘

Strain (microstrain)

-500

L I

 

    

: -0— Analytical Strain
. + Measured Strain

-600 I7 I I I I I T I r 1 I I I I fi I I I I l I I I I '7 f I I T

0 1 00 200 300 400 500 600

 

 

 

 

 

 

Days (from transfer)
Figure 109. Strain Comparison for Bottom VWSG SCC 1 Beam Section A

0.
We t o a

 

 

-200 d
’E 4
- -400 -
g .
w l
O
I- i
.2 -600 -
é :
5 1
a 1
3 -800 -
m .1

 

 

 

 

-1000 l
1 -0— Analytical Strain C]

. + Measured Strain

-1200 rillvrfifthi[lvrfijiurrrfirfi

0 1 00 200 300 400 500 600

 

 

 

 

 

 

 

 

 

Days (from transfer)

Figure 110. Strain Comparison for Top VWSG SCC 1 Beam Section A

231

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

;
.4
-100 j
E '200:
a d
.5- o
8 -300
h
.2
E
V -400
.E
a
b
0’ 6001
45003
1 -0— Analytical Strain C]
j + Measured Strain 1
-700 ,...,r.r.,fi1..rr.fv,.,.,,fir,r
0 100 200 300 400 500 600

Days (from transfer)

Figure 111. Strain Comparison for Bottom VWSG SCC 2 Beam Section A

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0
O
-200
’E
'3 -400
:5
in
O
b
.2 -600
E.
c
'a . ]
:_-, -800 4
m .1
40003
j -0— Analytical Strain
. + Measured Strain
-1200..T,......,..rr..t....,.,....
0 100 200 300 400 500 600

Days (from transfer)

Figure 112. Comparison of Strains for Top VWSG SCC 2 Section A

232

Strain (microstrain)

-1OO

 

-200

-300

 

 

 

 

 

 

-0- Analytical Strain

 

 

 

+ Measured Strain

 

 

 

 

 

l

100

I l T I I I I I I I 1 I I I I I I I I I T

200 300 400 500 600

Days (from transfer)

Figure 113. Comparison of Strains for Bottom VWSG SCC 3 Section A

Strain (microstrain)

 

-600 -
a

 

 

   

 

+ Analytical Strain
+ Measured Strain

 

 

 

 

 

 

 

 

 

 

 

I T l

0 100

200 300 400 500 600

Days (from transfer)

Figure 114. Comparison of Strains for Top VWSG SCC 3 Section A

233

6.4 Influence of Diaphragm on Strain in Beam

The previous plots show the comparison of the analytical or calculated strain to
the measured strain at the midspan of the bridge beams or, section A. The comparisons
shows some significant differences. The first is the discrepancy between the analytical
and measured first point. While none of the points seem to match exactly, this point
seems to have the largest difference. One possible complication that may exist at this
position is the presence of an eight in. diaphragm. This solid concrete portion completely
encases the 6 in. strain gage. The section properties used in the calculation of the
prestress losses were the voided section properties. To see if this solid section caused
inaccuracies in the comparison two methods were carried out, first the section properties
at the diaphragm were used in the prestressing loss analysis, second the prestressing loss
at section B, 12 in. from the midspan were calculated. Each of these methods was done
for both the NCC and SCC 1 beams.

As was the case for the voided section properties, the section properties at the
diaphragm needed to be transposed to account for the area of the prestressin g steel. This
needed to be done at multiple times throughout the bridge life and separately for each
concrete mix design. Table 71 and Table 72 show the properties used to calculate the
strain in the NCC beam while Table 73 and Table 74 show the properties used to

calculate the strain in the SCC 1 beam.

Table 71. NCC Section Properties Without Void
Event Area (inz) I (in‘) y (in.)

 

 

 

 

1 949.60 54,143 12.80
2 983.16 58,609 13.29
4 1,847 212,862 21.81

 

234

Table 72. N CC Strand and Instrument Eccentricities Without Void

 

 

 

 

Event Condition Condition Condition Top Bot.
l 2 3 Inst. Inst.
(in.) (in.)
1 11.70 9.37 9.55 11.70 10.80
2 11.21 9.86 10.04 11.21 11.29
4 2.69 18.38 18.56 2.69 19.81

 

Table 73. SCC 1 Section Properties Without Void

 

Event Area (inz) [(in“) Y (in.)

 

 

 

1 926.79 54,073 13.04
2 962.79 58,581 13.50
4 1,827 212,814 22.01

 

Table 74. SCC 1 Strand and Instrument Eccentricities Without Void

 

Event Condition Condition Condition Top Inst. Bot. Inst.

 

 

 

1 2 3 (in.) (in.)
1 11.46 9.61 9.79 11.46 11.04
2 11.00 10.07 10.25 11.00 11.50
4 2.49 18.58 18.76 2.49 20.01

 

The results of the prestressing loss calculation using the diaphragm section
properties are shown in Figure 115 and Figure 116 for the bottom and top flanges of the
NCC beam. The results for the SCC beam bottom and top flanges are shown in Figure
117 and Figure 118. From these plots for the bottom flange, Figure 115 and Figure 117,
it is possible to see that the first point of the analytical plot seems to be closer to the
measured value, however, the following points including the continuously monitored data
do not show any improvements over the plots with the voided section properties. The
plots of the top flange do not show improvement over the plots with the voided section
properties. From these plots it can be concluded that the use of the section properties

considering the diaphragm does not cause a positive change to the strain comparison.

235

 

-100 _

 

Strain (microstrain)

 

 

 

 

 

 

: + Analytical Strain C]
. + Measured Strain

'600 j—II I I T I T I I r r r r r I I f f r 1 I I I I 1' I I I f

0 1 00 200 300 400 500 600

 

 

 

 

 

 

 

Days (from transfer)
Figure 115. Bottom Strain Comparison With Diaphragm NCC Section A

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

100
5
’5‘
in
b
on
O
L
.2
E.
.E
a
b
(D
2 + Analytical Strain [:1
. + Measured Strain
-500....f.4.rr.r.......T....,....
0 100 200 300 400 500 600

Days (from transfer)
Figure 116. Top Strain Comparison With Diaphragm NCC Section A

236

 

-100

 

 

 

-3OO

 

 

 

-4oo

 

Strain (microstrain)

-500 ‘

 

 

-0— Analytical Strain
. + Measured Strain
-600 I I I I r I I l I l I I I I I I I I I l I I I I I l I I T

0 1 00 200 300 400 500 600

 

 

 

 

 

 

 

Days (from transfer)
Figure 117. Bottom Strain Comparison With Diaphragm SCC 1 Section A

 

 

 

 

 

 

 

Strain (microstrain)

 

   

. + Analytical Strain
I + Measured Strain

 

 

 

 

 

'600 I I I I v v r u l v I r l l r r u v I v r r r l r r r r
0 1 00 200 300 400 500 600

 

Days (from transfer)
Figure 118. Top Strain Comparison With Diaphragm SCC 1 Section A

237

To further investigate the influence of the diaphragm on the strain comparison the
strain was predicted outside the influence of the diaphragm. The results of the
prestressing loss calculations at section B are shown in Figure 119 and Figure 120 for the
bottom and top flanges of the NCC beam and Figure 121 and Figure 122 for the bottom
and top flanges of the SCC l beam. The main difference from in the calculation of the
prestress loss at section A is the calculation of the moment due to the self-weight of the
beam and the deck. Section A represents the maximum moment due to self-weight, thus
at section B, the moment due to self-weight is slightly lower. From these plots it is
possible to see that the analytical values do not match the measured data with any more

accuracy than the values at Section A.

 

-200 -

 

Strain (microstrain)

 

 

 

 

l A l

 

 

 

 

+ Analytical Strain

+ Measured Strain

'800 I I I F 1 I I I I l T f 1 I I I I r I l v r r u l r u u r
0 100 200 300 400 500 600

 

 

 

 

Days (from transfer)
Figure 119. Bottom Strain Comparison N CC Section B

238

 

 

Strain (microstrain)

 

 

 

 

« + Analytical Strain C]

i + Measured Strain

”800 II'IIIIIIT‘IW’Y‘IIII‘fiI‘IIT‘IfiI

0 1 00 200 300 400 500 600

 

 

 

 

 

 

 

 

 

Days (from transfer)
Figure 120. Top Strain Comparison NCC Section B

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

is“

a

b

(D

O

L

.2

1.5.

.E

a

h

:75
Z -0— Analytical Strain C]
« -V— Measured Strain 4

-1OOO....,..rr1....,....,T........

0 100 200 300 400 500 600

Days (from transfer)
Figure 121. Bottom Strain Comparison SCC 1 Section B

239

 

 

Strain (microstrain)

 

 

-500 J L

1 -0— Analytical Strain

. + Measured Strain
-600 I I I I l I I I I l I I I I I I I I I I I I I I I I I I T

0 1 00 200 300 400 500 600

 

 

 

 

 

 

 

 

 

 

Days (from transfer)
Figure 122. Top Strain Comparison SCC 1 Section B

The two methods used to account for the influence of the diaphragm on the strain
comparisons did not increase the accuracy of the results. Differences between the
analytical and measured values existed whether voided section properties were used or
not, and at sections completely outside the diaphragm. The difference that do exist seem
to be consistent through out the four mix designs. This shows that all four of the

instrumented beams behave in a similar manner.

6.5 Effect of Temperature on Strain in Beam

As discussed previously, the increase in compressive strain during the warmest
months of the year seem to indicate that the change in temperature of the beams affects
the strain in the beams. This change in strain is significant as an increase in compressive
strain would imply that the prestressing force is also increasing. This increase could

cause cracking in the concrete if the change was large enough.

240

There are two ways that temperature can affect the strain in a bridge beam. The
first is a temperature change that is uniform throughout the depth of the beam. The
second is when the temperature change throughout the beam depth is not uniform. In

reality a combination of these two events occurs.

 

 

 

 

 

 

 

 

 

 

 

 

Figure 123. Thermocouple Locations and Temperature Variation

Figure 123 shows the location of temperature readings throughout the depth of the
composite beam as black dots in the cross section. The readings taken at these locations
throughout time can be summarized by the assumed temperature gradient shown in the
figure. In this gradient there are two parts. Part A is a uniform temperature that exists
from the top to the bottom of the beam cross section. Part B is the temperature gradient
that is large at the top and reaches zero at the bottom of beam. It is the combination of
these two temperature change that cause the strain in the beam to change.

The bridge in this project is a simply supported beam with one end free to expand,
and the other is fixed against horizontal movements. As is shown in Figure 124, both
ends of the beams for the demonstration bridge rest on elastomeric bearing pads. These
pads allow horizontal movements, or expansions of the beam. One end of the beam is
fixed against these movements through a dowel bar that is grouted into place at the center

of the bearing. This bar extends from the beam into the abutment wall and does not allow

241

the beam to expand. Thus all of the expansion is taken up in one end, the expansion end,

of the bridge beam.

/- Expansion End

M

 

 

Fixed End
Figure 124. Fixed and Expansion Ends of Bridge Beams

The changes in temperature that cause changes to the strain in the beam must be
referenced to the temperature at the time of bridge construction. The temperature when
the beams were placed at the bridge site was approximately 50 degrees Fahrenheit. This
temperature is the reference temperature for all temperature calculations.

The uniform temperature change It is shown in Figure 125. Here a positive
change in temperature causes the free end of the beam to move horizontally. Because the
beam elongates a VWSG located at the center of the beam should record an increase in
strain. This increase should be in a positive direction as the beam lengthens causing the
tension in the concrete. This result comes from the fact that the beams are free to expand
under normal conditions. If the beam was not free to expand, this positive temperature

change would cause an increase in compressive strain in the concrete.

+AT

 

f“—

_ l J

Figure 125. Expansion Due to a Uniform Positive Temperature Change

242

Because the coefficient of thermal expansion for concrete and steel are very
similar the increase in strain seen in the concrete due to a temperature increase should
also cause an increase in strain in the prestressing strand. This will happen because the
prestressing steel and concrete should expand or contract by the same amount. The
coefficient of thermal expansion as given by the AASHTO-LRFD-2 Specification [3] is 6
x 106/°F.

The second way temperature can affect the strain in a beam is through a
temperature gradient. A temperature gradient is formed when the temperature throughout
the depth of the beam is not uniform. It can either be warmer at the top and cooler at the
bottom or the opposite. Of interest for this analysis is the situation during the warmest
months of the year. During these months, the deck, at the top of the section is generally
warmer than the bottom flange of the beams. When this occurs, the top of the beam tries
to expand more than the bottom of the beam. The result of this deformation is an upward
curling of the beam as shown in Figure 126. This curling causes the top of the beam to

be in tension and the bottom of the beam to be in compression.

+AT: Curvature

 

 

 

Figure 126. Deformation Caused by Temperature Gradient

The calculation of the strain induced in the beam due to a uniform temperature
change is done using Equation (69),

(69) 8 uniform : Ct’AT

243

where or is the coefficient of thermal expansion of concrete and AT is the difference in
temperature from the time the beams were set at the bridge until the point of interest. The
point of interest for this calculation is the 365th day of the prestress loss analysis. This
day was chosen as it falls in the middle of the compressive strain increase for the
monitored data. The actual date at this point is June 215‘, 2006. The temperature data
recorded by the thermocouples at 12:00 pm on this day are shown in Table 75. The six
thermocouple recordings are identified by their position in the beam. The remaining
temperatures were chosen to lie on the temperature profile from the thermocouples. The
temperature gradient shown in the last column is the difference between the temperature

at the specific height and the minimum recorded temperature.

Table 75. Recorded Temperatures for NCC Beam (6/21/06)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Location Beam Ht Temperature Gradient
(in) (°F) (°F)
36 76.1 5.1
Top of Deck 34.25 75.2 4.2
32.5 75.2 4.2
Middle of Deck 31.5 76.0 5
30.5 77 6
Bottom Deck 28.75 79.4 8.4
27 77 6
Top Flange 24.5 75.5 4.5
22 73.4 2.4
Middle of Beam 13.5 72.7 1.7
8 71.6 0.6
Bottom Flange 2.25 71.0 0
0 71.0 0

 

The minimum recorded temperature in Table 75 is 71 °F. This temperature is 21

°F higher than the temperature when the beams were set at the bridge site. Thus, this

244

temperature change is used to calculate the strain due to the uniform temperature change
using Equation (69). This strain is a tensile strain equal to 126 microstrain.

The calculation of the strain due to the temperature gradient is more complicated
than that for the uniform temperature change. This calculation requires the cross section
to be separated into multiple sections. The sections for the beams in this project were
chosen such that each of the six thermocouples used in the analysis lie at the centroid of a
section. The section properties for the multiple sections used in the temperature analysis

are shown in Table 76.

Table 76. Section Properties for Gradient Temperature Analysis
Section Area (inz) Centroid (in.) I (in4) Y (in.) d (in.)

 

 

 

 

 

 

 

1 336 34.25 343 9.39 3.5
2 192 31.5 64 6.64 2
3 336 28.75 343 3.89 3.5
4 208.5 23.88 823 0.98 5
5 99 13.5 998 11.36 11
6 202 3.02 786 21.84 4.5

 

There are two components to the gradient temperature induced strain. The first is

an axial temperature strain. This strain is calculated using Equation (70):
a
(70) 8.....1 = ZZTr-Ai , [31

where A is the total area of the composite cross section, T,- is the temperature at the
centroid of a specific section and A,- is the area of a specific section. When the
temperature gradient and section properties for the beams are applied to this equation a
strain value of 27.5 microstain is calculated. This strain is a tensile strain as the beam is

expanding, thus, it will have a positive value.

245

The second component of the temperature induced strain comes from the

curvature caused as the beam curls. This curvature is calculated using Equation (71):

a AT-
(71) r1 =72[7}y.-A.-+7'-1,-].[31

i
where I is the moment inertia of the composite section, 155,789 in4, y; is the distance
from the neutral axis of the total section to the neutral axis of the specific section, AT,- is
the change in temperature from the bottom of a section to the top of that section, I,- is the
moment of inertia for the given section, and d,- is the depth of a given section. The
calculated curvature for these beams is 1.322 in'l.

This curvature can be used to find the strain induced in the beam due to the
temperature gradient. First the moment caused by this curvature is calculated using
Equation (72):

(72) M = Eli/l,
where E is the modulus of elasticity of the concrete. This moment can be used to
calculate the stress at the top or bottom of the beam using Equation (73):

(73) 0:54}.

where y is either the distance from the neutral axis to the top of the beam or to the bottom
of the beam depending on where the stress is to be found. Finally by simplifying the

previous two equations and applying Hooke’s Law, Equations (74) and (75) result:
(74) Ebottom = YbV/ I

(75) Stop = ”W-

246

Thus, the strain at the bottom of the section due to the temperature gradient is a
compressive strain of 32.9 microstrain. While the strain at the top of the beam due to the
temperature gradient is a tensile strain of 14.7 microstrain.

All three strain components, the strain due to the uniform temperature change, and
the axial and curvature based strains from the gradient temperature, can be combined to
find a resultant strain acting at the top and bottom of the section. If temperature is the
reason for the increase in compressive strain seen in the measured plots, the strain from
this analysis should be similar. Figure 127 shows the resulting strain diagrams for the
temperature analysis. The total strain in the section due to the temperature is calculated

to be between 100 and 200 microstrain for this time period.

+1_2_6 +215 +14.7 +168

+ +i=l

+1726 +275 329 +121

 

 

 

 

Figure 127. Strain Diagram for Temperature Analysis

The total strain in the top and bottom flange due to the temperature throughout the
cross section is very close to the value of compressive strain increase seen in the
measured data. However, the sign of this value which is positive for a tensile strain is
opposite of what is seen in the measured data. Thus, it seems that temperature alone

cannot explain the change in strain seen in the measured data.

247

6.6 Influence of Bridge System on Strain in Beam

While the temperature induced strain is the opposite sign of what is expected
based on the measured data it may still be the reason for the compressive strain increase.
The above temperature analysis is based on the assumption that the beam is completely
free to expand. In reality, the elastomeric bearing pad has some stiffness that may
prevent some of the expected temperature based expansion. Furthermore, other factors
could influence the stiffness of the bearing. Some of these factors could include the
presence of a backwall cast between the beams on the abutment wall, or the
interconnectivity of the bridge system through the external diaphragm and the bridge
deck. Figure 128 shows an idealized beam system. In this system the beam rests on a
bearing of height e, with a roller replaced by a spring with stiffness k. This stiffness can
very from zero if nothing obstructed the expansion of the beam to infinity if the beam

was unable to expand at all.

k }/7 Section A *
W by?

Figure 128. Assumed Beam System With Bearing Stiffness k

 

To see if the bearing stiffness of the beam could result in an increase of the
compressive strain 0n the order to that seen in the measured data the worst case scenario
should be analyzed. The worst-case scenario is that where the stiffness of the bearing is

infinity, or the beam is pinned at both ends. This system is shown in Figure 129.

248

/ Section A

WW 3 l
7: ‘ :1 .

Figure 129. Idealized System With Infinite Bearing Stiffness

When a temperature increase on the same order as that seen in the temperature
analysis is applied to the beam the reaction forces, P at the pins develop. This force
creates a bending moment diagram as shown in Figure 130. The maximum moment that
occurs uniformly across the entire length of the beam is equal to the reaction P multiplied

by the height of the bearing e.

P

Fe
\ l/2_
XVW/////////////////.7/7//////////
8
V

 

53

 

 

Figure 130. Bending Moment Diagram For Idealized System

To investigate the affect of the bearing stiffness could induce the increase in
continuous strain that is seen in the measured data two procedures were carried out. First
the idealized system described above was modeled using SAP 2000 (Computers &
Structures Inc, Berkeley, California). The beam in this model was then given a
temperature expansion equal to the uniform temperature expansion seen in the
temperature analysis, 21 °F. The reaction at the pins was then reported by the model.

The second procedure was to back calculate the reaction caused in the idealized

model based on the increase in compressive strain seen in the measured data. This was

249

done by calculating the stress both using Hooke’s Law and the beam flexure equation

shown in Equation (76):

(76) 0:541!“

The moment in Equation (76) is calculated using Equation (77):
(77) M = Pe.
Combining these two equations with Hooke’s law gives Equation (78):

(78) E5 =_P;:_y,

which when simplified for the reaction P is Equation (79):

_5531
8y .

(79) P

Thus, the reaction in the pins can be calculated in two ways, first by using a computer
model and the observed temperature increase of 21°F and secondly by back calculation
from the observed strain increase of 100 microstrain. The results of these two procedures
show that the reaction found using the computer model was 943,380 lbs, while the
reaction from the strain increase was found to be 1,099,900 lbs.

These results would imply that the beam is acting as a pinned beam with no
ability to expand. It is unlikely that the beam is unable to expand as this bridge is built
with a common MDOT bearing detail. Yet, this lack of free expansion most likely plays
a role in the explanation for the compressive strain increase in the measured strain

response.

250

6.7 Data Normalization

The final discrepancy in the data is the large difference that exists between the
values of the NCC and SCC beams. To remove this difference, the data was normalized
by making the reference temperature and reference strain for all readings zero. In doing
this, the data is essentially showing the change in strain, without regard for the starting

value. The result of this for the strain response of the bottom flange is shown in Figure

131.

 

 

 

 

 

 

 

 

 

 

 

26004
«0
J

24ooe
in .

a p. N

.5 « 0'

(02200“ I

.3:- - .

In

g t

._ L .

52000. . v

.E 4 v

.93 : 9

”1800-
I—G—Ncc [ j
j—v—SCC1

,,,,._..sec.. :3
4+SCCS
38383883338338385
ESSEESESSSSSESSEE

Date (6/22/2005-11/1/2006)
Figure 131. Normalized Strain Response for Bottom Flange Section A

From the figure it is possible to see that very little has changed from the
unnormalized plot presented previously. In fact this figure is merely an upward

translation of the other figure. All that has changed is the reference point is now zero

251

instead of what it was previously. The differences between the values of the various
beams is on the order of 100-300 microstrain.

The strain response in the top flange showed a larger difference in the previously
presented figure (Figure 95) than that of the bottom flange. The difference in this plot
was on the order of 600 microstrain. Once the data is normalized Figure 132 is the result.
In this figure the difference between the four plots is reduced to 100-300 microstrain.
Thus, by normalizing the data to a common reference point it is possible to see that the

responses of all four beams are nearly identical.

 

 

 

 

 

 

 

 

 

 

 

 

2800
1
it
j I

2700‘ .

73 I

52600-:

G

b 7 O

W

e .-

32500- X

.2. 1° .

c 1

I. d .

824004

*" i

(D
1+NCC |

230014—3001
.—rsccz
j+sccs

22m‘rjl IIIIIIIIIIIIIII
33338385853858888
ESEEEESESSEEESEEE

Date (6/22/2005-11/1/2006)
Figure 132. Normalized Strain Response for Top Flange at Section A

It is unclear why the data in the unnormalized state shows such large differences
in value. However, the scope of this thesis is limited to the differences between the four
concrete mix designs. From the normalized figures it is possible to see that the responses

of each of the beams is nearly identical and thus no significant differences in the behavior

252

of the different concrete mix designs seems to exist. To investigate further the reasons

for the differences in strain values would be outside the scope of this project.

6.8 Long Term Performance Conclusions

From the results obtained from the field monitoring program it is possible to see
that while the values of strain that are recorded are very different from the analytical
values, the overall behavior of the beams appears to be similar. The beams show similar
behaviors for all of the concrete mix designs. Additional analysis trying to explain the
differences that existed in the beam were conducted but proved to be inconclusive.

Multiple factors were investigated without definitive results.

It is likely that it is not a single factor that causes the differences in the measured
and predicted strains, but rather a combination of many affects. It was shown that the
presence of the diaphragm does not greatly influence the predicted results thus this is not

likely to contribute to the inaccuracies.

The temperature analysis showed, that as expected, temperature does play a role
in the discrepancies seen in the measured and predicted data. However, it can not be the
temperature alone, as the temperature increase seen in the measured data should cause a
tensile strain in the beam. The influence of the bearing system stiffness for the bridge
could introduce a compressive strain increase on the order of what is seen in the
measured data. However, the real behavior of this boundary condition is complicated and

its complete analysis is beyond the scope of this thesis.

Normalization of the data allowed for the direct comparison of the SCC and NCC

plots as significant differences in the values of the strain were removed. The difference

253

in the reference value appeared to cause the difference in the strain values that was
present in the unnormalized plots. From the normalized plots it is possible to see that the

response of the four beams is nearly the same.

In spite of the inconclusive evaluation just presented, the results of the long—term
monitoring system show that the four concrete mix designs behaved similarly over the
measurement period of one year. The strains over time show similar rates of loss of
prestress and affects of temperatures. While differences do exist between the four plots,
they seem to be limited to the overall bridge system and they do not dependent on the

concrete material properties.

254

7 CONCLUSIONS AND RECOMMENDATIONS

The use of SCC to cast elements for highway bridges can lead to reductions in
both the time labor required in production. For this project, casting times for SCC beams
were reduced an average of 61% over NCC beams. Significant reductions in labor led
were also realized. However, implementing SCC in a precast facility is not as simple as
changing the concrete mix design. Extensive work is required to develop a mix design
and certain modifications to current casting operations are required to use SCC
effectively. Further, time is necessary to gain the experience that will allow SCC to be
used most efficiently.

Multiple methodologies can be used to develop SCC mix designs that consistently
produce high quality concrete. These mixes can achieve the desired SCC fresh properties
using commonly available equipment and ingredients. One way to develop and refine an
SCC mix design is by conducting a systematic trial and error process that revaluates the
mix based on fresh property test results after each phase. Fresh property tests, either
preliminary, as those proposed by PCI, or standardized, as those being developed by
ASTM, can provide sufficient information to adjust mix design parameters to achieve
satisfactory performance of the fresh concrete.

The use of chemical admixtures in the SCC mix designs may affect the bond
performance of the concrete. Specifically the use of VMA may lead to reduced bond
strengths. While some reduction in bond strength over the NCC mix was seen, the size of
the structural members used in this project was not bond critical. Slightly reduced bond
strength, or an increased development length would not likely lead to a compromise in

either flexural or shear strength for the bridge beams.

255

The beneficial properties of SCC were realized in this project as the production of
the box beams was reduced from four steps to two through the use of SCC. The most
difficult part of the production process was restraining the Styrofoam block from the
pressure generated by the flowing concrete as it filled the formwork. Top wood restraints
bolted to the formwork were successful in restraining the Styrofoam block while not
impairing the flow of concrete through the formwork.

The self-consolidating nature of SCC allows for a more complete compaction of
the concrete. This, along with the superior flow behavior of SCC, led to beams being
produced with higher surface quality in the SCC beams when compared to the NCC
beams. Fewer and smaller surface imperfections were seen due to trapped air at the
formwork concrete interface. This improved surface quality led to less work being done
to patch the surface of the beams.

The decreased beam production time led to a nearly continuous batching
operation. Greater amounts of concrete were produced in less time during the production
of the SCC beams compared to the NCC beams. This increased production rate leads to
less down time for the labor during beam production.

The constant quality control during the production of the SCC batches allowed for
necessary adjustments to maintain consistent fresh property behavior throughout the
beam production process. Small changes to the SCC mix design required for changing
conditions of weather and material constituents were able to be made. Increased
experience with SCC would lead to more accurate estimates of these necessary

adjustments.

256

From the full-scale flexural tests the overall behavior of the SCC beams was seen
to be very similar to that of the conventional beam (NCC). Cracking patterns, widths, and
failure levels followed predicted responses through conventional prestressed concrete
theory. While the failure of the flexure test units was explosive, the beam was not over-
reinforced and the strands yielded before failure, technically defining the failure mode as
ductile. Consequently, variations in concrete compressive strength did not significantly
modify the beam’s ultimate capacity, which were very similar for all beams. The
absolute capacities of the SCC beams were marginally lower than that of the NCC beam,
specifically, 1.3, 3.4, and 3.5 percent lower for the SCC1, SCC2, and SCC3 beams,
respectively. However, the flexural capacities of the SCC beams exceeded the required
design capacity by 6 to 9 percent.

The shear behavior of the SCC beams was also found to be adequate and very
similar to that of the NCC beam. Cracking paths and widths were consistent in all beams
and the failure levels closely matched analytical predictions. The shear capacities of the
SCC beams were comparable to that of the NCC beam. The ratio of capacity to nominal
strength for the SCC1, SCC2, and SCC3 beams as 1.22, 1.12, and 1.08, respectively,
while that same ratio for the NCC beam was 1.11. The shear capacity and response of
the SCC beams is thus considered adequate.

Due to the large amount of prestressing and the extension of the beam beyond the
supports, failure of the critical section in shear did not result in overall member failure.
Rather, the beam was able to re-distribute demands beyond the critical shear area thus
allowing the beam to resist increasing loads that in two beams eventually lead to flexure-

induced failures and in two others the tests were terminates. However, even though the

257

overall beam failure was not in a pure shear mode, the shear behavior and capacity
provided by the beams was successfully evaluated. This was determined by studying the
shear force versus shear strain response of the critical shear areas in the beam, which
showed that the sections reached a peak shear force beyond which shear strains would
increase rapidly in an almost elastic-perfectly-plastic response. This same behavior was
supported by the moment-curvature response of this sections, which had essentially a bi-
linear response with a failure at a load level much below the known flexural capacity.
The reduced flexural capacity of the section is thus a consequence of the shear failure of
the section. In addition, evaluation of strain profiles of the steel stirrups in the critical
shear areas indicated that the transverse steel had yielded through the section depth for all
beams, further supporting that sections shear failure was reached.

Overall, the performance of the SCC prestressed box beams in flexure and shear
was found to be essentially equal to that of the NCC beam and their behavior was well
predicted by conventional prestressed concrete theory. While additional issues pertaining
to long-term behavior of SCC in prestressed elements, namely creep effects, are still to be
fully evaluated, the short-term flexural and shear response evaluated through this testing
program indicates that the SCC prestressed beams safely satisfy their prescribed design
requirements. The results of this testing program provided sufficient evidence to believe
the SCC beams could be used in the demonstration bridge as planned.

Long-term in—service performance of the SCC beams in comparison of the NCC
beams is being achieved by the field monitoring of 4 beams in the demonstration bridge
(3 SCC beams and one NCC beam). Performance is being measured by means of strains

in selected cross sections and temperature instruments are permitting the compensation in

258

the strain readings as well as evaluation of temperature effects. Results to date indicate
that the SCC beams are performing similarly to the NCC beam. To date, the effects of
temperature on the strain measurements are dominant. An attempt has been made to
compare the measured strains to predicted values by estimating the strain variations due
to prestress losses as well as temperature induced effects. Differences between the
predicted and measured values exist. However, the applicability of time-dependent
losses due to concrete shrinkage and creep to SCC is questionable. Little information on
the applicability of current code prestress loss models to SCC exists. Thus, further work
is required to obtain a better comparison between predicted and measured responses.

The data measured to date indicates that while some variations exist in the initial
strain levels of the girders. The variation of initial readings can be attributed to the
different losses in the SCC beams caused by concrete shrinkage and elastic shortening,
which are controlled by the water content in the mix and the concrete elastic modulus.
Clearly, both of these parameters vary considerably for the SCC beams. Thus, further
work is required to understand the reasons behind the difference in initial readings in the
SCC girders.

Finally, the variation of strain once the bridge was in service seems to be very
similar for all girders. This would indicate that thus far there has been no change in
performance between the different beams. However, it is recognized that the monitoring
process has been in place for slightly over a year. Evaluation of true long-term
performance will require acquisition of data for several more years before more sound

conclusions on the long-term behavior of SCC prestressed elements can be reached.

259

From the results obtained from the field monitoring program it is possible to see
that while the values of strain that are recorded are very different from the analytical
values, the overall behavior of the beams appears to be similar. The beams show similar
behaviors for all of the concrete mix designs. Additional analysis trying to explain the
differences that existed in the beam were conducted but proved to be inconclusive.

Multiple factors were investigated without definitive results.

It is likely that it is not a single factor that causes the differences in the measured
and predicted strains, but rather a combination of many affects. It was shown that the
presence of the diaphragm does not greatly influence the predicted results thus this is not

likely to contribute to the inaccuracies.

The temperature analysis showed, that as expected, temperature does play a role
in the discrepancies seen in the measured and predicted data. However, it can not be the
temperature alone, as the temperature increase seen in the measured data should cause a
tensile strain in the beam. The influence of the bearing system stiffness for the bridge
could introduce a compressive strain increase on the order of what is seen in the
measured data. However, the real behavior of this boundary condition is complicated and

its complete analysis is beyond the scope of this thesis.

Normalization of the data allowed for the direct comparison of the SCC and NCC
plots as significant differences in the values of the strain were removed. The difference
in the reference value appeared to cause the difference in the strain values that was
present in the unnormalized plots. From the normalized plots it is possible to see that the

response of the four beams is nearly the same.

260

In spite of the inconclusive evaluation presented in the long term monitoring
section, the results of the long-term monitoring system show that the four concrete mix
designs behaved similarly over the measurement period of one year. The strains over
time show similar rates of loss of prestress and affects of temperatures. While
differences do exist between the four plots, they seem to be limited to the overall bridge

system and they do not dependent on the concrete material properties.

261

APPENDICES

262

APPENDIX A

Appendix A contains the Special Provisions for Production of Prestressed Beams
With Self Consolidating Concrete. This document was written for the Michigan
Department of Transportation and is included here with MDOT’s permission for
reference purposes only. All references and tables in this document are independent of

the rest of this thesis.

263

MICHIGAN
DEPARTIVIENT OF TRANSPORTATION

SPECIAL PROVISION
FOR
PRODUCTION OF PRESTRESSED BEAMS WITH
SELF-CONSOLIDATING CONCRETE

MSU:RB:C&T:RDT 1 of 7 C&T:APPR:JFS:SJC:Ol-l3-05

a. Description. This specification describes the minimum requirements for the
production of nine prestressed concrete beams with three different self-consolidating
concrete (SCC) mix designs for use in a demonstration bridge. All work shall be
according to Section 708 of the Standard Specifications for Construction, except as
modified herein. The use of SCC for precast/prestressed elements shall follow the
procedures and requirements of the PCI Interim Guidelines for the Use of Self-
Consolidating Concrete in Precast/Prestressed Institute Member Plants [1], except as

modified herein.

In order to place the SCC beams for the bridge, they must perform equally or better
than the conventional beams. Two beams for each SCC mix design and two beams for the
conventional mix design will be tested for flexural characterization. Thus, the number of
conventional beams to be built will be equal to the total number of beams in the bridge
plus the two test units. Similarly, for each SCC beam to be placed, two more beams are
needed for testing. Thus, a total of nine SCC beams are needed (three beams for each of

the three SCC mix designs).

With the above strategy, should the performance of the SCC beams be sub-standard

after structural testing, the bridge project with the conventional beams will proceed

264

without delays. Further, if the SCC beams are placed, the extra conventional beams will
be stored at the Michigan State University (MSU) Civil Infrastructure Laboratory in case
that the field monitoring shows that they have to be replaced.

b. Materials. The materials used shall meet the requirements specified in the

designated section of the Standard Specifications for Construction and as specified

herein:

Portland Cement, Type I or III ..............................................................
901

Fine Aggregate, 2NS ..........................................................................
902

Coarse Aggregate, 6AA or 17A ................................................. 708 and
902

Mineral admixtures shall not be used for the production of SCC for this project.

All conventional chemical admixtures shall conform to Section 903. A trial mixture
program shall be developed in consultation with Dr. Rigoberto Burguefio at MSU
(Phone: 517-353-1743). The fabricator shall demonstrate satisfactory performance of the
admixture relative to fluidity, stability, workability, air content, and strength under the
conditions of use, particularly with respect to temperature and humidity typical of

production conditions in accordance with this specification.

Chemical admixtures for SCC shall be carefully selected in consultation with the
admixture supplier for compatibility with the cement and all admixtures used to ensure

that they perform adequately. Admixture supplier’s recommendations shall be observed.

c. Design and Proportioning of Concrete Mixtures. Three self consolidating
concrete (SCC) mixtures and one normally consolidating concrete (NCC) mixtures shall

be designed and proportioned meeting the following requirements.

265

The tailorable design of SCC mixes, for specific plastic (fresh) and hardened
properties, complicates the selection of representative mix proportioning. In addition,
SCC proportioning is highly dependent on mixing equipment, discharge methods, type of
chemical admixtures, and placement procedures. Thus, the required mix designs shall be
developed in coordination with the chemical admixture supplier and MSU as discussed

below.

While general guidelines for the required mix designs are provided here, the specific
mix designs will vary depending on the preceding variables. The proportioning of the
SCC mixes shall be done by a qualified commercial laboratory or qualified precast
concrete technician according to subsection 605.02D. A qualified concrete technologist
experienced with SCC mix design shall establish the mix design with respect to the
specifications, and concrete mixing and placing conditions. The SCC mix designs for the
project need to be specifically developed and tested by the fabricator in coordination with
MSU and a concrete admixtures company. The fabricator must have experience with

SCC mix design and casting is required.

Three different SCC mix designs are to be developed and used for the project. The
overall concept for proportioning of the SCC mixes is shown in Table l. The key guiding
parameter for the mix designs is the water cement (w/c) ratio. All mixes are to be
designed for a design compressive strength at 28—days of 5500 psi using Type I or III
cement, and aggregates in agreement with section 708. The level of entrained air content
for all mixes shall be 6% +/- 1.5%. For research purposes, the mix designs are bounds to

the approaches for SCC, that is:

266

0 Approach 1: Proportioning with moderate w/c ratios (e.g., 0.45), and use of HRWR
and viscosity-enhancing admixtures (VMA) to provide fluidity and increase
stability, respectively. This approach corresponds to the SCC-3 mix in Table 1 and

Table 2.

0 Approach 2: Mixes without viscosity-enhancing admixture, but with lower w/c
ratios (e.g., 0.33) to reduce free water content and provide stability and use of a
relatively high content of HRWR to provide high-fluidity. This approach

corresponds to the SCC-l mix in Table l and Table 2.

The SCC-2 mix design is the reference SCC mix design for relative “amounts” noted
in Table 1. The SCC-2 mix is to be obtained by using the NCC mix as a basis. Typically,
this requires a decrease of the coarse aggregate content (CAC) with a corresponding
increase of sand, or increase in the sand to paste ratio (SfPt). This approach will require

moderate use of HRWRS and VMAs.

The normally consolidated concrete mix (NCC) used for the conventional beam shall
have a target w/c ratio of that normally used in production (Table l) and use the same

materials, except chemical admixtures, as the SCC mix designs.

The amounts of HRWRS and VMAs will vary depending on the source, mixing

procedure, and delivery needs. Use of a set-retardant admixture may be used, particularly

if using Type III cement, to maintain the SCC plastic properties during placement. The

267

effect of variations of charging the admixtures into the mixer shall be determined from

the recommendations of the admixture supplier, MSU researchers, and by trial mixes.

MSU has experience in developing the proposed SCC mix designs and will assist the
fabricator in the design and calibration of their mixes. Similar mix designs to those
required in this project have been successfully developed and used in MSU research for

design strengths at 28 days of 7000 psi and are listed in Table 2.

The fabricator shall demonstrate the ability to produce the mix designs described
above in a consistent manner and satisfying the quality control requirements outlined in
item (d). A trial batch or data from previous projects shall be used to verify the mix
designs. The trial batches and previous data must use the same materials that will be
used for this project. Trial batches must be repeated, at no cost to the Department, if the

concrete does not meet the requirements of this specification.

 

 

 

 

 

Table A-1. Conce' tual Mix Design Matrix
Mix Design w/c HRWR VMA CAC S/Pt
SCC-l 0.35 More less less or same more
SCC-2 0.40 Moderate moderate conventional conventional
SCC-3 0.45 Less more more less
NCC regular Regular none regular regular

 

 

 

 

 

 

 

 

w/c = water/cement ratio; HRWR = high range water reducer; VMA = viscosity modifying
admixture; CAC = coarse aggregate content; S/Pt = sand to paste ratio

* The guideline of “conventional” in SCC-2 refers to the usual procedure of proportioning an SCC
mix from a standard NCC mix, where the coarse aggregate content is reduced with a
corresponding increase in sand.

** The guidelines of “more,” “less,” or “same” noted in the mix content refers to relative amounts
compared to the SCC-2 mix design.

268

Table A-2. Sample 7000 psi Mix Designs from MSU Research

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Type SCC-1 SCC-2 SCC-3 NCC
Constituents (lbs)
Cement Type-III 700 700 700 700
Sand 2 NS 1519 1426 1275 1216
Coarse Aggregate 6 AA 1380 1380 1435 1580
Water 245 280 315 280
Air 6 % (target) 6% 6% 6% 6%
w/c Ratio 0.35 0.4 0.45 0.4
Admixtures (oz/cwt)
Air Entraining MB-AETM 90 0.5 0.5 9.08 l
HRWR Glenium® 3200 HES 6.29 8 2
VMA Rheomac® VMA 358 0 l 9 0
Set Retardant DELVO® Stabilizer 6.14 6.69 6 6

 

d. Testing and Quality Control. The proportioned SCC mixes shall be tested
according to subsection 604.03.B and to satisfy the SCC performance criterion of having

a highly flowable and stable mix with consideration to temperature and humidity typical

of production conditions.

Testing and quality control shall be done by following the procedures and apparatus
in accordance with the Interim Guidelines for the use of Self-Consolidating Concrete in
Precast/Prestressed Concrete Institute Member Plants by the Precast/Prestressed Concrete
Institute [1]. At a minimum, the tests identified in Table 3 shall be conducted. The
performance of each SCC mix shall be documented as specified in Interim Guidelines for
the use of Self-Consolidating Concrete in Precast/Prestressed Concrete Institute Member
Plants by the Precast/Prestressed Concrete Institute [1]. Acceptability of the mixes for
each of the minimum identified methods is noted in Table 3. Acceptability criteria for

other methods are to be determined by the Engineer in consultation with MSU.

269

 

Table A-3. Minimum tests to be used for evaluatirg SCC performance (see Ref [1])

 

 

 

 

 

 

 

 

 

Test
Method Description Measurement Acceptability
Objective Criteria
The lateral flow (diameter) Evaluate self-
Inverted of the concrete patty is compactibility as it Slump flow equal
Slump measured when executing a relates to yield stress. to
Flow traditional slump test with 27 in. :t 1.0 in.
the slump cone in the
inverted position without
rodding (ASTM C192).
Visual evaluation of the Used to evaluate the
Visual SCC patty prior to relative stability of VSI rating equal
Stability placement and after the batches of the same or to or less than 1
performance of the slump similar SCC mixes. It
flow test. requires considerable
judgment.
Used to force the SCC flow The difference between
through reinforcement in the spread with and J-Ring value
conjunction with the without the ring or the between 0.5 in.
inverted slump flow test. height difference inside and 0.6 in.
J -Ring The 12” ring of vertical and outside the ring is
bars shall be provided with measured.
5/8”-diam. bars spaced at
even intervals along the
ring diameter equal to 3
times the maximum
aggregate size. Other
dimensions are in given in
Reference [1].
Used to simulate the Provides an indication
casting process by forcing a on the static and L-Box ratio
SCC sample to flow dynamic segregation parameter greater
through a removable gate resistance as well as the than 80 (80%)
L- Box with vertical fitted ability to flow through
reinforcement bars under reinforcement.
static pressure. The
sections of reinforcement
bar shall consist of 3 #5
bars spaced at 3 times the
maximum aggregate size.
Other apparatus dimensions
are given in Reference [1].

 

* The reader shall refer to Reference [1] for further details on the test method description, equipment
dimensions, procedure, and interpretation of results.

270

 

e. Mix Design Refinement. Once an SCC mix with the required plastic properties
has been developed, the mix and production process shall be refined to provide an

optimum production mix that meets all of the following requirements:

0 Required compressive strength and air content
0 Required fresh properties (see acceptability criteria in Table 3)

0 Repeatability of acceptable quality.

f. Strand Bond Confirmation Tests. Strand bond tests shall be run for all mixes to
verify that the bond with SCC is equivalent or better than the conventional mix design
with equal strand. Bond characteristics of strand shall be determined for all mixes in
Table 1 by the Moustafa test [2] and using the “standard conventional concrete mix” that
the plant uses for strand bond acceptance testing as the comparison basis. The peak
pullout force (average) for the SCC mix design shall not be less than 67% of the peak

pullout force (average) using the conventional concrete mix design.

g. Use of Mock-Ups in Production Qualification of Mixes. As part of the mix
design qualification/development process, each SCC mix must be subject to actual
production-based confirmation by the production of mock-up elements. Quality

confirmation of elements fabricated with SCC shall include:

0 Visual inspection. Precast concrete elements shall have a post pour inspection
as per PCI MNL-ll6 and the fabricators’ quality systems manual. Special

attention should be paid to any signs of deficiency that would impair the

271

concrete performance such as segregation, sedimentation, cold joints, or other

visual defects.

0 Inspection and testing by saw cutting and coring If a visual inspection of the
product after form removal indicates that problem areas could exist, or if
observations during the placing process identify potential areas and the
element is not to be rejected, then additional verification of the concrete
quality shall be undertaken. Saw cut or cored samples can be visually

inspected or submitted for strength testing and/or petrographic analysis.

0 Strand bond confirmation. Bond confirmation tests shall be performed by
measuring strand pull-in or draw-in after strand release. The average strand

draw-in shall not exceed 0.15 inches.

Test pieces can also help the qualification of placement methods. The following

issues must be evaluated prior to fabrication.

0 Transportation/handling techniques

0 Placement methods/conditions

o Placement distance

0 Free-fall distance

0 Lateral flow distance from charging point

0 Lift height/head pressures.

h. Equipment. The fabricator shall supply the apparatus needed to perform the SCC

plastic property evaluation as specified in section (d). The test apparatus shall follow the

272

requirements outlined in the Interim Guidelines for the use of Self-Consolidating
Concrete in Precast/Prestressed Concrete Institute Member Plants by the

Precast/Prestressed Concrete Institute [1].

i. Beam Instrumentation and Data Collection. The fabricator shall coordinate the
manufacturing of the SCC and NCC beams with Dr. Rigoberto Burguefio at MSU and

allow for the following:

0 Instrumentation of passive and prestressing reinforcement.

0 Place embedded instrumentation on the beam reinforcement before stressing.

0 Installation of additional instruments once the beam reinforcement and strands
are placed, but before stressing.

0 Data collection during stressing, after casting but before prestress release, and
after prestress release.

0 Sampling and testing of the SCC mix designs for information purposes.

The fabricator shall allow the MSU research team to place instrumented

reinforcement and appropriately route and secure instrumentation cabling.

The fabricator shall allow for the instrumentation and monitoring of prestressing

strands with strain gages and load cells during the stressing process.

j. Concrete Placement. The SCC beams shall be cast according to the following

requirements:

273

0 Batching shall be done so that the SCC mix does not loose its plastic
properties.

0 Plastic properties shall be evaluated for every batch.

0 SCC beams shall be placed in a continuous operation from one end of the
beam to the other with all internal components in place.

0 The SCC mix shall not be dropped from a height of more than six feet.

0 The SCC beams shall not be vibrated.

k. Laboratory Testing of Beams. The fabricator shall deliver two beams of each
SCC mix design (six total), along with two beams of the normal mix design to:
Dr. Rigoberto Burguefio
do Mr. Siavosh (Sia) Ravanbakhsh
MSU Civil Infrastructure Laboratory
2851 Jolly Road
Okemos, Michigan 48864
Phone (Campus): 517-353-1743
Fax (Campus): 517—432-1827
Phone (Lab): 517-432-4913
Fax (Lab): 517-432-4915
Two weeks notice prior to delivery is required.
The fabricator and contractor shall allow four weeks for laboratory testing the beams
to determine SCC beam performance and acceptability. If the SCC beams are acceptable,
the fabricator shall ship the three remaining SCC beams to the site for erection at the

locations shown on the plans and ship the remaining three normal mix design beams to

the MSU Civil Laboratory. If the SCC beams are not acceptable, the fabricator shall ship

274

the three normal mix design beams to the site for erection at the locations shown on the

plans and ship the remaining three SCC beams to the MSU Civil Laboratory.

l. Performance Monitoring. The Contractor shall coordinate their construction

activities with MSU researchers and allow for the following:

0 Installation of instrumentation on the demonstration bridge. This includes the
installation of measuring devices at the bridge system level, routing of cables
for the embedded instruments, installation of the data logger, and connection
of cables to the data logger.

0 Static loading through a controlled service-load prior to opening the bridge to
traffic. The contractor shall provide a minimum of a 40,000 pound truck with
known axle weights and axle spacing for loading.

m. Measurement and Payment.

Contract Item (Pay Item)

Pay Unit

Prestressed Concrete Box Beam, SCC, Furnished, 27 Inch .................. Foot
Payment includes furnishing all the necessary materials, labor and equipment
necessary for the production of the SCC beams. All testing, trial batches, coordination
with researchers, and delivery is considered incidental to the pay item. Shipping the SCC
beams and the normal mix beams to the MSU laboratory is considered included in the
pay item for furnishing the prestressed concrete member.

Beams used for testing will also be paid for in addition to those being erected.

275

Erection of the Prestressed Concrete Box Beam, SCC, Furnished, 27 Inch will be paid
according to section 708 of the Standard Specifications for Construction.

References.
[l] Precast/Prestressed Concrete Institute (PCI), “Interim Guidelines for the use of Self-

Consolidating Concrete in Precast/Prestressed Concrete Institute Member Plants,”
TR-6-03, Chicago, IL, 2003.

[2] Logan, D. “Acceptance Criteria for Bond Quality of Strand for Pretensioned
Prestressed Concrete Applications,” PCI Journal, March-April 1997, pp. 52-90.

276

APPENDIX B NCC Production Cast 1 (6-21-05)

The information presented in the following tables is the complete list of NCC batches

 

produced for the first cast on the 21fit of June in 2005.

Table 77. NCC Mix Design Batch 1
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 562.5
Fine Aggregate 1354 1353
Coarse Aggregate 1882.5 1885.5
Total water 219 218.5
W/c 0.39 0.39
Admixture (oz/yd3)
Air Entraining 10.5 10
VMA 28 27.5
HRWR 64 63.5
Table 78. NCC Mix Design Batch 2
Target Value Actual Value
Constituents (lbs/yd3)
Cement 564 574
Fine Aggregate 1353 1356
Coarse Aggregate 1879 1881.5
Total water 219 219.5
W/c 0.38 0.38
Admixture (oz/yd3)
Air Entraining 10.5 1 l
VMA 28 28
HRWR 64 63.5

 

 

277

Table 79. NCC Mix Design Batch 3
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 576.5
Fine Aggregate 1353 1350.5
Coarse Aggregate 1875 1878.5
Total water 219 219
W/c 0.38 0.38
Admixture (oz/yd3)
Air Entraining 11.5 11
VMA 28 28
HRWR 64 63.5

 

Table 80. NCC Mix Design Batch 4
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 562.5
Fine Aggregate 1353 1339
Coarse Aggregate 1875 1880.5
Total water 219.5 219.5
W/c 0.39 0.39
Admixture (oz/yd3)
Air Entraining 1 1.5 11
VMA 28 28
HRWR 64 63.5

 

Table 81. NCC Mix Design Batch 5
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 570.5
Fine Aggregate 1353 1340.5
Coarse Aggregate 1875 1886.5
Total water 218.5 218.5
W/c 0.38 0.38
Admixture (oz/yd3)
Air Entraining 1 1.5 0
VMA 14 14
HRWR 64 63.5

 

278

Table 82. NCC Mix Design Batch 6
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 561.5
Fine Aggregate 1353 1344.5
Coarse Aggregate 1875 1883.5
Total water 219 219
W/c 0.39 0.39
Admixture (oz/yd3)
Air Entraining 11.5 11
VMA 14 13.5
HRWR 64 63.5

 

Table 83. NCC Mix Design Batch 7
Tzflet Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 562
Fine Aggregate 1353 1353.5
Coarse Aggregate 1875 1871.5
Total water 21 l 21 l
W/c 0.38 0.38
Admixture (oz/yd3)
Air Entraining 1 1.5 l 1
VMA 14 14
HRWR 64 63.5

 

Table 84. NCC Mix Desigflatch 8
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 562
Fine Aggregate 1353 1339
Coarse Aggregate 1875 1892.5

Total water 219 219

W/c 0.39 0.39

Admixture (oz/yd3)

Air Entraining 11.5 10.5
VMA 14 13.5

HRWR 64 63.5

 

279

Table 85. NCC Mix Design Batch 9
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 562
Fine Aggegate 1352.67 1348
Coarse Aggregate 1882.67 1882
Total water 220 220
W/c 0.39 0.39
Admixture (oz/yd3)
Air Entrainin g 1 1 10.67
VMA 14 13.33
HRWR 64 63.33

 

280

 

 

APPENDIX C SCC 1 Production Cast 1 (6-21-05)

The information presented in the following tables is the complete list of SCC l batches

produced for the first cast on the 21St of June in 2005.

Table 86. SCC 1 Mix Design Batch 1
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 714
Fine Aggregate 1655.5 1650
Coarse Aggregate 1355.5 1361
Total water 256 235.5
W/c 0.37 0.33
Admixture (oz/yd3)
Air Entraining 16 15.5
VMA 21 20.5
HRWR 105 1 18.5

 

Table 87. SCC 1 Mix Design Batch 2
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)

Cement 700 695

Fine Aggregate 1656 1667

Coarse Aggregate 1350 1350
Total water 264 264

W/c 0.38 0.38

Admixture (oz/yd3)

Air Entraining 12 1 l
VMA 21 20

HRWR 105 104

 

281

Table 88. SCC 1 Mix Desigfliatch 3
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 713
Fine Aggregate 1655.5 1649.5
Coarse Aggregate 1350 1356.5
Total water 265 265
W/c 0.38 0.37
Admixture (oz/yd3)
Air Entraining 12 l 1.5
VMA 21 21
HRWR 105 104.5

 

Table 89. SCC 1 Mix Design Batch 4
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 698.5
Fine Aggregate 1655.5 1646
Coarse Aggregate 1350 1362
Total water 263.5 263.5
W/c 0.38 0.38
Admixture (oz/yd3)
Air Entraining 12 1 1.5
VMA 21 20.5
HRWR 105 104.5

 

Table 90. SCC 1 Mix Design Batch 5
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 700.5
Fine Aggregate 1655.5 1649.5
Coarse Aggregate 1350 1360
Total water 259.5 259
W/c 0.37 0.38
Admixture (oz/yd3)
Air Entraining 12 1 1.5
VMA 2 1 20.5
HRWR 105 104.5

 

282

 

Table 91. SCC 1 Mix Design Batch 6
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/Ld3)

Cement 700 710

Fine Aggregate 1655.5 1645.5
Coarse Aggregate 1350 1360
Total water 263.5 263
W/c 0.38 0.37

Admixture (oz/yd3)

Air Entraining 12 11.5
VMA 21 20.5
HRWR 105 104.5

 

Table 92. SCC 1 Mix Design Batch 7
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 701
Fine Aggregate 1655.5 1646
Coarse Aggregate 1350 ‘ 1359.5
Total water 264 264
W/c 0.38 0.38
Admixture (oz/yd3)
Air Entraining 12 11.5
VMA 21 21
HRWR 105 104.5

 

Table 93. SCC 1 Mix Defln Batch 8
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 712.5
Fine Aggregate 1655.5 1645.5
Coarse Aggregate 1350 1361
Total water 264 264
W/c 0.38 0.37
Admixture (oz/yd3)
Air Entraining 12 11.5
VMA 21 20.5
HRWR 105 104.5

 

283

Table 94. SCC 1 Mix Design Batch 9
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)

Cement 700 700.5

Fine Aggregate 1655.5 1651
Coarse Aggregate 1350 1359
Total water 265.5 265.5

W/c 0.38 0.38

Admixture (oz/yd3)

Air Entraining 12 11.5
VMA 21 20.5
HRWR 105 104.5

 

 

Table 95. SCC 1 Mix Design Batch 10
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)

Cement 700 712.5

Fine Aggregate 1655.5 1648
Coarse Aggregate 1350 1357
Total water 264 263.5

W/c 0.38 0.38

Admixture (oz/yd3)

Air Entraining 12 1 1.5
VMA 21 20.5
HRWR 105 104.5

 

Table 96. SCC 1 Mix Design Batch 11
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/£13)
Cement 700 h 701
Fine Aggregate 1655.5 1659.5
Coarse Aggregate 1350 1345.5
Total water 267.5 260
W/c 0.38 0.37
Admixture (oz/yd3)
Air Entraining 12 1 1.5
VMA 21 20.5
HRWR 105 ‘ 104.5

 

284

 

Table 97. SCC 1 Mix Design_Batch 12
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 715
Fine Aggregate 1655.5 1653.5

Coarse Aggregate 1350 1354
Total water 264 264

W/c 0.38 0.37

Admixture (oz/yd3)

Air Entraining 12 1 1.5
VMA 21 20.5
HRWR 105 104.5

 

Table 98. SCC 1 Mix Dgigg Batch 13
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 700
Fine Aggregate 1655.5 1655.5

Coarse Aggregate 1350 1355
Total water 264 264

W/c 0.38 0.38

Admixture (oz/yd3)

Air Entraining 12 11.5
VMA 21 20.5

HRWR 105 104.5

 

Table 99. SCC 1 Mix Design Batch 14
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 698
Fine Aggregate 1655.5 1651
Coarse Aggregate 1350 1357
Total water 265.5 265.5
W/c 0.38 0.38
Admixture (oz/yd3)
Air Entraining 12 1 1.5
VMA 21 20.5
HRWR 105 104.5

 

285

APPENDIX D N CC Production Cast 2 (6-24-05)
The information presented in the following tables is the complete list of NCC batches

produced for the second cast on the 24th of June in 2005.

Table 100.NCC Mix Design Batch 1
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 570
Fine Aggregate 1351.5 1353
Coarse Aggregate 1886.5 1884.5
Total water 227 227
W/c 0.40 0.40
Admixture (oz/yd3)

Air Entraining 11 10.5
VMA 14 13 . 5

HRWR 64 64

 

Table 101.NCC Mix Design Batch 2
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)

Cement 564 572.5

Fine Aggregate 1351.5 1352
Coarse Aggregate 1875 1884
Total water 219.5 219

W/c 0.39 0.38

Admixture (oz/yd3)

Air Entrainin g 11 10.5
VMA 14 13.5

HRWR 64 63.5

 

286

 

Table 102.NCC Mix Design Batch 3
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 574
Fine Aggregate 1351.5 1342
Coarse Aggregate 1875 1883
Total water 219 218.5
W/c 0.39 0.38
Admixture (oz/yd3)
Air Entraining 1 1 10.5
VMA 14 13.5
HRWR 64 63.5

 

Table 103.NCC Mix Design Batch 4
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 561.5
Fine Aggregate 1351.5 1348
Coarse Aggregate 1886.5 1894
Total water 220 220
W/c 0.39 0.39
Admixture (oz/yd3)
Air Entraining 1 1 10.5
VMA 14 13.5
HRWR 64 63.5

 

Table 104.NCC Mix Design Batch 5
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 572
Fine Aggregate 1351.5 1343.5
Coarse Aggregate 1882.5 1891.5
Total water 219.5 219.5
W/c 0.39 0.38
Admixture (oz/yd3)
Air Entraining 1 1 10.5
VMA 14 13.5
HRWR 64 63.5

 

287

Table 105.NCC Mix Design Batch 6
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 573
Fine Aggregate 1351.5 1348
Coarse Aggregate 1882.5 1865.5
Total water 219.5 219
W/c 0.39 0.38
Admixture (oz/yd3)
Air Entraining 11 10.5
VMA 14 13.5
HRWR 64 63.5

 

Table 106.NCC Mix Design Batch 7
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)

Cement 564 568.5

Fine Aggregate 1351.5 1350

Coarse Aggregate 1886.5 1891.5

Total water 220 220

W/c 0.39 0.39

Admixture (oz/yd3)
Air Entraining 10.5 10

VMA 14 13.5

HRWR 64 63.5

 

Table 107.NCC Mix Design Batch 8
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 566
Fine Aggregate 1351.5 1351
Coarse Aggregate 1886.5 1884.5
Total water 220 220
W/c 0.39 0.39
Admixture (oz/yd3)

Air Entraining 10.5 10
VMA 14 13.5
HRWR 64 63.5

 

288

Table 108.NCC Mix Dfln Batch 9
Taget Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 566
Fine Aggregate 1351.67 1523.33
Coarse Aggregate 1886.67 1920
Total water 225 225
W/c 0.40 0.40
Admixture (oz/yd3)
Air Entraining 10.5 10
VMA 14 13.3333
HRWR 64 63.3333

 

289

APPENDIX E SCC 2 Production Cast 2 (6-24-05)
The information presented in the following tables is the complete list of SCC 2 batches

produced for the second cast on the 24th of June in 2005.

Table 109.SCC 2 Mix Design Batch 1
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 702
Fine Aggregate 1487.5 1485.5
Coarse Aggregate 1435.5 1440.5
Total water 277.5 277.5
W/c 0.40 0.40
Admixture (oz/yd3)
Air Entraining 12 11.5
VMA 21 20.5
HRWR 84 83.5

 

Table 110.SCC 2 Mix Design Batch 2

Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)

Cement 700 724.5
Fine Aggregate 1487.5 1490.5
Coarse Aggregate 1435.5 1433.5
Total water 278.5 278.5

W/c 0.40 0.38

Admixture (oz/yd3)
Air Entraining 12 l 1
VMA 21 20.5
HRWR 84 83.5

 

290

Table 111.SCC 2 Mix Design Batch 3
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 699.5
Fine Aggregate 1487.5 1482.5
Coarse Aggregate 1435.5 1442.5
Total water 279.5 279.5
W/c 0.40 0.40
Admixture (oz/yd3)
Air Entraining 12 11
VMA 21 21
HRWR 84 83.5

 

Table 112.SCC 2 Mix Design Batch 4
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 716
Fine Aggregate 1487.5 1487.5
Coarse Aggregate 1435.5 1435.5
Total water 275 275
W/c 0.39 0.38
Admixture (oz/yd3)
Air Entraining 12 11
VMA 21 20.5
HRWR 84 83.5

 

Table 113.SCC 2 Mix Design Batch 5
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 699.5
Fine Aggregate 1487.5 1490.5
Coarse Aggregate 1435.5 1432
Total water 275 275
W/c 0.39 0.39
Admixture (oz/yd3)
Air Entraining 11.5 11
VMA 21 20.5
HRWR 84 83.5

 

291

 

 

 

Table 114.SCC 2 Mix Design Batch 6
Tget Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 709
Fine Aggrgate 1487.5 1483
Coarse Aggregate 1435.5 1441
Total water 274.5 274.5
W/c 0.39 0.39
Admixture (oz/yd3)
Air Entraining 11.5 11
VMA 21 21
HRWR 84 83.5

 

Table 115.SCC 2 Mix Desigg Batch 7
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 712
Fine Aggregate 1487.5 1487
Coarse Aggregate 1435.5 1438
Total water 275 275
W/c 0.39 0.39
Admixture (oz/yd3)
Air Entraining 11.5 11
VMA 21 21
HRWR 84 83.5

 

Table 116.SCC 2 Mix Design Batch 8
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)

Cement 700 698
Fine Aggregate 1487.5 1477.5
Coarse Aggregate 1435.5 1443.5
Total water 274.5 274.5

W/c 0.39 0.39

Admixture (oz/yd3)
Air Entraining 11.5 11
VMA 21 20.5
HRWR 84 83.5

 

292

Table 117.SCC 2 Mix Design Batch 9
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)

Cement 700 713.5
Fine Aggregate 1487.5 1475.5
Coarse Aggregate 1438 1446
Total water 274.5 274.5

W/c 0.39 0.38

Admixture (oz/yd3)
Air Entraining 11.5 11
VMA 21 20.5
HRWR 84 83.5

 

Table 118.SCC 2 Mix Design Batch 10
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 71 1
Fine Aggregate 1487.5 1483.5
Coarse Aggregate 1438 1446
Total water 274.5 274.5
W/c 0.39 0.39
Admixture (oz/yd3)
Air Entraining 11.5 11
VMA 21 21
HRWR 84 83.5

 

Table 119.SCC 2 Mix Design Batch 11
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)

Cement 700 714

Fine Aggregate 1487.5 1475
Coarse Aggregate 1445.5 1454
Total water 274.5 274.5

W/c 0.39 0.38

Admixture (oz/yd3)
Air Entraining 1 1.5 11

VMA 21 20.5

HRWR 84 83.5

 

293

 

Table 120.SCC 2 Mix Design Batch 12
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 714
Fine Aggregate 1487.5 1476.5
Coarse Aggregate 1445.5 1456
Total water 274.5 274.5
W/c 0.39 0.38
Admixture (oz/yd3)
Air Entraining 1 1.5 11
VMA 21 20.5
HRWR 84 83.5

 

294

APPENDIX F N CC Production Cast 3 (6-29-05)

The information presented in the following tables is the complete list of NCC batches

produced for the third cast on the 29‘h of June in 2005.

Table 121.NCC Mix Design Batch 1
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 569
Fine Aggregate 1354 1346.5
Coarse Aggregate 1884.5 1887
Total water 219.5 226.5
W/c 0.39 0.40
Admixture (oz/yd3)
Air Entraining 9 8.5
VMA 1 l 1 1
HRWR 64 63.5

 

Table 122.NCC Mix Design Batch 2
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 565.5
Fine Aggregate 1354 1344.5
Coarse Aggregate 1884.5 1891.5
Total water 219.5 219
W/c 0.39 0.39
Admixture (oz/yd3)
Air Entraining 8.5 8
VMA 1 1 10.5
HRWR 64 63.5

 

295

Table 123.NCC Mix Design Batch 3
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 561.5
Fine Aggregate 1354 1342.5
Coarse Aggregate 1884.5 1892
Total water 220 220
W/c 0.39 0.39
Admixture (oz/yd3)
Air Entraining 8.5 8
VMA 0 0
HRWR 64 63.5

 

Table 124.NCC Mix Design Batch 4
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 564.5
Fine Aggregate 1354 1344
Coarse Aggregate 1884.5 1895.5
Total water 220 219.5
W/c 0.39 0.39
Admixture (oz/yd3)
Air Entraining 8.5 8
VMA 0 0
HRWR 64 63.5

 

Table 125.NCC Mix Design Batch 5
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 573.5
Fine Aggregate 1354 1344.5
Coarse Aggregate 1884.5 1890
Total water 219.5 219.5
W/c 0.39 0.38
Admixture (oz/yd3)
Air Entraining 8.5 8
VMA l 1 10.5
HRWR 64 63.5

 

296

 

Table 126.NCC Mix Design Batch 6
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 570.5
Fine Aggregate 1354 1347.5
Coarse Aggregate 1884.5 1888
Total water 219.5 219.5
W/c 0.39 0.38
Admixture (oz/yd3)
Air Entraining 8.5 8
VMA 1 1 10.5
HRWR 64 63.5

 

Table 127.N CC Mix Design Batch 7
Target Value Actual Value

 

Constituents (lbs/yd3)

 

 

 

 

 

 

 

 

 

Cement 564 574.5

Fine Aggregate 1354 1359.5
Coarse Aggregate 1884.5 1882
Total water 220.5 220.5
W/c 0.39 0.38

Admixture (oz/yd3)
Air Entraining 8.5 8

VMA 0 0
HRWR 64 63.5

 

Table 128.NCC Mix Design Batch 8
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 563.5
Fine Aggregate 1354 1357
Coarse Aggregate 1884.5 1881
Total water 220.5 220.5
W/c 0.39 0.39
Admixture (oz/yd3)
Air Entraining 8.5 8
VMA 0 0
HRWR 64 63.5

 

297

 

 

Table 129.NCC Mix Design—Batch 9
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 562
Fine Aggregate 1354 1353.5
Coarse Aggregate 1884.5 1889
Total water 220 . 220
W/c 0.39 0.39
Admixture (oz/yd3)
Air Entraining 8.5 8
VMA 0 0
HRWR 64 63.5

 

298

APPENDIX G SCC 3 Production Cast 3 (6-29-05)

The information presented in the following tables is the complete list of SCC 3 batches

produced for the third cast on the 29th of June in 2005.

Table 130.SCC 3 Mix Design Batch 1
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)

Cement 700 705.5

Fine Aggregate 1373.5 1361
Coarse Aggregate 1457.5 1466
Total water 309.5 311.5

W/c 0.44 0.44

Admixture (oz/yd3)
Air Entraining 10 9.5

VMA 21 20.5

HRWR 49 48.5

 

Table 131.SCC 3 Mix Design Batch 2
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 709.5
Fine Aggregate 1373.5 1375.5
Coarse Aggregate 1457.5 1454
Total water 310 312.5
W/c 0.44 0.44
Admixture (oz/yd3)
Air Entraining 8.5 8
VMA 21 21
HRWR 56 55.5

 

299

Table 132. SCC 3 Mix Design Batch 3
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 711.5
Fine Aggregate 1373.5 1367
Coarse Aggregate 1457.5 1466
Total water 309.5 312
W/c 0.44 0.44
Admixture (oz/yd3)
Air Entraining 8 7.5
VMA 21 21
HRWR 56 55.5

 

Table 133. SCC 3 Mix Design Batch 4
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 709
Fine Aggregate 1373.5 1361.5
Coarse Aggregate 1457.5 1465
Total water 309.5 312.5
W/c 0.44 0.44
Admixture (oz/yd3)
Air Entraining 8 7.5
VMA 21 20.5
HRWR 56 55.5

 

Table 134. SCC 3 Mix Design Batch 5
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 708
Fine Aggregate 1373.5 1365
Coarse Aggregate 1457.5 1463
Total water 309.5 312.5
W/c 0.44 0.44
Admixture (oz/yd3)
Air Entraining 8 7.5
VMA 21 20.5
HRWR 56 55.5

 

300

 

Table 135. SCC 3 Mix Design Batch 6
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 712.5
Fine Aggrgate 1373.5 1371
Coarse Aggregate 1457.5 1462.5
Total water 31 1 313.5
W/c 0.44 0.44
Admixture (oz/yd3)
Air Entraining 8 7.5
VMA 21 21
HRWR 56 55.5

 

Table 136. SCC 3 Mix Design Batch 7
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 712
Fine Aggregate 1373.5 1367.5
Coarse Aggregate 1457.5 1461.5
Total water 310.5 313
W/c 0.44 0.44
Admixture (oz/yd3)
Air Entraining 8 7.5
VMA 21 20.5
HRWR 56 55.5

 

Table 137. SCC 3 Mix Design Batch 8
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 713
Fine Aggregate 1373.5 1380.5
Coarse Aggregate 1457.5 1455
Total water 310.5 313
W/c 0.44 0.44
Admixture (oz/yd3)
Air Entraining 8 7.5
VMA 21 20.5
HRWR 56 55.5

 

301

 

 

Table 138. SCC 3 Mix Design Batch 9
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)

Cement 700 709.5
Fine Aggregate 1373.5 1368.5
Coarse Aggregate 1457.5 1464.5

Total water 310.5 313

W/c 0.44 0.44

Admixture (oz/yd3)

Air Entraining 8 7.5

VMA 21 20.5

HRWR 56 55.5

 

Table 139. SCC 3 Mix Design Batch 10
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 709.5
Fine Aggregate 1373.5 1369.5
Coarse Aggregate 1457.5 1460
Total water 310.5 313.5
W/c 0.44 0.44
Admixture (oz/yd3)
Air Entraining 8 7.5
VMA 21 20.5
HRWR 56 55.5

 

Table 140. SCC 3 Mix Design Batch 11
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)

Cement 700 697.5

Fine Aggregate 1373.5 1379
Coarse Aggregate 1457.5 1454.5
Total water 310.5 314.5

W/c 0.44 0.45

Admixture (oz/yd3)
Air Entraining 8 7.5
VMA 21 20.5
HRWR 56 55.5

 

302

Table 141. SCC 3 Mix Design Batch 12
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 699
Fine Aggregate 1373.5 1376.5
Coarse Aggregate 1457.5 1454.5
Total water 31 1 315
W/c 0.44 0.45
Admixture (oz/yd3)
Air Entraining 8 7.5
VMA 21 21
HRWR 56 55.5

 

Table 142. SCC 3 Mix Design Batch 13
TargflValue Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 71 1
Fine Aggregate 1373.5 1363.5
Coarse Aggregate 1457.5 1464.5
Total water 31 1 314
W/c 0.44 0.44
Admixture (oz/yd3)
Air Entraining 8 7.5
VMA 21 20.5
HRWR 56 55.5

 

Table 143. SCC 3 Mix Design Batch 14
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 700 712
Fine Aggregate 1373.5 1363.5
Coarse Aggregate 1457.5 1466.5
Total water 31 1 313
W/c 0.44 0.44
Admixture (oz/313)
Air Entraining 8 7.5
VMA 21 20.5
HRWR 56 55.5

 

303

APPENDIX H NCC Production Cast 4 (7-1-05)

The information presented in the following tables is the complete list of NCC batches

produced for the fourth cast on the lSt of July in 2005.

Table 144. NCC Mix Design Batch 1
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 561.5
Fine Aggregate 1354 1347
Coarse Aggregate 1875 1876
Total water 219.5 219.5
W/c 0.39 0.39
Admixture (oz/yd3)
Air Entraining 7.5 7
VMA l 1 1 1
HRWR 64 63.5

 

Table 145. NCC Mix Design Batch 2
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 562
Fine Aggregate 1354 1333
Coarse Aggregate 1 884 1807
Total water 219 219
W/c 0.39 0.39
Admixture (oz/yd3)
Air Entraining 7.5 7
VMA 11 1 1
HRWR 64 63

 

304

Table 146. NCC Mix Design Batch 3
Targt Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 561.5
Fine Aggregate 1354 1352
Coarse Aggregate 1884.5 1889.5
Total water 219 218.5
W/c 0.39 0.39
Admixture (oz/yd3)
Air Entraining 7.5 7
VMA 0 0
HRWR 64 63.5

 

Table 147. NCC Mix DefingBatch 4
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 562
Fine Agggate 1354 1360
Coarse Aggregate 1884.5 1879.5
Total water 219.5 220
W/c 0.39 0.39
Admixture (oz/yd3)
Air Entraining 7.5 7
VMA 0 0
HRWR 64 63.5

 

Table 148. NCC Mix Design Batch 5
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 562
Fine Aggregate 1353 1339
Coarse Aggregate 1892 1909
Total water 220 220
W/c 0.39 0.39
Admixture (oz/yd3)
Air Entraining 7.5 7
VMA 1 1 1 1
HRWR 64 63

 

305

Table 149. N CC Mix Design Batch 6
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 579
Fine Aggregate 1354 1337.5
Coarse Aggregate 1894 1900
Total water 219 219
W/c 0.39 0.38
Admixture (oz/yd3)
Air Entrainin g 7.5 7
VMA 1 1 10.5
HRWR 64 63.5

 

Table 150. NCC Mix Design Batch 7
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 571
Fine Aggregate 1353 1359
Coarse Aggregate 1894 1896
Total water 220 221
W/c 0.39 0.39
Admixture (oz/yd3)
Air Entrainin g 7 6
VMA 1 1 1 1
HRWR 64 63

 

Table 151. NCC Mix Design Batch 8
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 579
Fine Aggregate 1354 1364
Coarse Aggregate 1894 1889.5

Total water 220 220

W/c 0.39 0.38

Admixture (oz/yd3)
Air Entraining 7 6.5
VMA 0 0
HRWR 64 63.5

 

306

Table 152. NCC Mix Desigp Batch 9
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)

Cement 564 577

Fine Aggregate 1353 1348

Coarse Aggregate 1894 1900.5

Total water 219.5 219

W/c 0.39 0.38

Admixture (oz/yd3)
Air Entraining 7 6.5
VMA 0 0
HRWR 64 63.5

 

Table 153. NCC Mix Design Batch 10
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 564 560
Fine Aggregate 1353 1338
Coarse Aggregate 1894 1914
Total water 219 219
W/c 0.39 0.39
Admixture (oz/yd3)
Air Entrainin g 7 6
VMA 0 0
HRWR 64 63

 

Table 154. NCC Mix Design Batch 11
Target Value Actual Value

 

 

 

 

 

 

 

 

 

 

Constituents (lbs/yd3)
Cement 563 617
Fine Aggregate 1353 1478
Coarse Aggregate 1893 1898
Total water 225 225
W/c 0.40 0.36
Admixture (oz/yd3)
Air Entraining 7 7
VMA 0 0
HRWR 64 63

 

307

[1]

[2]

[3]

[4]

[5]

[6]

[7]

[8]

[9]

[10]

REFERENCES

AC1 Committee 318, “Building Code Requirements for Structural Concrete (AC1
318-05) and Commentary (AC1 318R-05),” American Concrete Institute,
Farmington Hills, MI, 1995.

American Association of Highway Transportation Officials (AASHTO), AASHTO
Bridge Design Specifications — 17th Edition, Washington DC, 2002.

 

American Association of Highway Trans ortation Officials (AASHTO), AASHTO-
LRFD Bridge Design Sp_ecifications - 2" Edititm, Washington DC, 1998.

American Association of Highway Trans ortation Officials (AASHTO), AASHTO-
LRFD Bridge Design Specifications — 3r Edition, Washington DC, 2007.

Bendert, D. A. “Experimental Evaluation and Field Monitoring of Self-
Consolidating Concrete Prestressed Box Beams Implemented in a Demonstration
Bridge.” Masters Thesis, Department of Civil and Environmental Engineering,
Michigan State University, East Lansing, MI, December 2006.

Bendert, D. A., and Burguefio, R., “Report on the Experimental Evaluation of
Prestressed Box beams for SCC Demonstration Bridge.” Research Report CEE-RR-
2006-01, Department of Civil and Environmental Engineering, Michigan State
University, East Lansing, MI, December 2006.

Bendert, D. A., and Burguefio, R., “Report on the Production of Prestressed Box
beams for SCC Demonstration Bridge.” Research Report CEE-RR-2006-02,
Department of Civil and Environmental Engineering, Michigan State University,
East Lansing, MI, December 2006.

Bentz, EC, and Collins, M.P., Response-2000 - Reinforced Concrete Section
Analysis Program using the Modified Compression Field Theory, v. 1.0.5,
University of Toronto, 2000.

Bonen, D, Shah, 8., “Fresh and Hardened Properties of Self-Consolidating

Concrete,” Progress in Structural Engineering and Materials, V. 7 No. 1, January-
March, 2005, pp. 14-26.

Burguei'io, R. and Haq, M., “Transfer and Development Length of Prestressing
Strands in Precast/Prestressed Girders Using Self-Consolidating Concrete.”
Proceedings Second North American Conference on the Design and Use of Self-
Consolidating Concrete and the Fourth International RILEM Symposium on Self-
Compacting Concrete, (CD-ROM), Chicago, Illinois, October, 2005.

308

[11]

[121

[131

[141

[15]

[16]

[171

[18]

[19]

1201

[21]

Burguefio, R., and Till, R., “Special Provision for Production of Prestressed Beams
with Self-Consolidating Concrete,” Internal Report, Michigan Department of
Transportation, Lansing, MI, January 2005.

Choulli, Y. Mari, A.R., “Shear Behaviour of full scale prestressed I beams made
with Self Compacting Concrete,” Proceedings Second North American Conference
on the Design and Use of Self-Consolidating Concrete and the Fourth International
RILEM Symposium on Self-Compacting Concrete, (CD-ROM), Chicago, Illinois,
October, 2005.

Collins, MP, and Mitchell, D., Prestressed Concrete Structures, Response
Publications, Canada, 1997.

 

Daczko, J ., and Vachon, M., “Self Consolidating Concrete (SCC),” Significance of
Tests and Properties of Concrete and Concrete-Making Materials STP 169D.
ASTM International West Conshohocken, PA. May 2006 pp. 637-645.

Das, D, Kaushik, S.K., Gupta V.K., “Shear Resistance of Self—Compacting
concrete,” Proceedings Second North American Conference on the Design and Use
of Self-Consolidating Concrete and the Fourth International RILEM Symposium on
Self-Compacting Concrete, (CD-ROM), Chicago, Illinois, October, 2005.

Domone, P.L., “Self-Compacting Concrete: An analysis of 11 years of case
studies,” Cement and Concrete Composites. V. 28. 2006 pp. 197-208

Fernandez, P. Luciano, J. Constantiner, D., “Successful Implementation of SCC In
a Precast Operation,” Proceedings Second North American Conference on the
Design and Use of Self-Consolidating Concrete and the Fourth Intemationa]
RH_EM Symposium on Self-Compacting Concrete, (CD-ROM), Chicago, Illinois,
October, 2005.

Goodier, C.I., “Development of Self-Compacting Concrete.” Proceedings of the
Institution of Civil Engineers, Structures and Buildings. V 156. November 2003
Issue SB4 pp. 405-414.

Haq, M., and Burgueiio, R., “Effect of Self-Compacting Concrete Mix Design on
Strand Bond Performance Parameters in Prestressed Beams,” Researcb Report
CEE-RR-2006-03, Department of Civil and Environmental Engineering, Michigan
State University, East Lansing, MI, December 2006.

Harrison, T., Concrete. V. 40 n. 7 August 2006 pp. 15.
Hegger, J. Kommer, B., “Pretensioning in Self Consolidating Concrete (SCC),”
Proceedings Second North American Conference on the Design and Use of Self-

Consolidating Concrete and the Fourth International RILEM Symposium on Self-
Compacting Concrete, (CD-ROM), Chicago, Illinois, October, 2005.

309

[241

[25]

[26]

[271

[28]

[29]

[30]

[31]

[32]

Hsu, T.T.C., Unified TheorLof Reinforced Concrete, CRC Press, Boca Raton, FL,
1993.

“Interim Guidelines for the Use of Self-Consolidating Concrete In PCI Member

Plants” TR-6-03 Precast/Prestressed Concrete Institute, Chicago, Illinois, April,
2003, 88 pp.

Juvas, K., ”The European Experience of Working With Self-Compacting Concrete
In the Precast Concrete Industry,” Proceedings Second North American Conference
on the Design and Use of Self-Consolidating Concrete and the Fourth International
RILEM Symposium on Self-Compacting Concrete, (CD-ROM), Chicago, Illinois,
October, 2005.

Khayat, K., “Use of Viscosity-Modifying Admixtures to Reduce Top_bar Effect of
Anchored Bars Cast with Fluid Concrete,” AC1 Materials Journal, March-April
1998 pp. 158-167.

Khayat, K., “Workability, Testing, and Performance of Self-Consolidating
Concrete,” ACI Materials Journal, V. 96 No. 3, May—June, 1999, pp. 346-353.

Lachemi, M., Hossain, K.M.A., Lombros, V., “Shear Resistance of Self-
Consolidating Concrete Beams -— Experimental Investigation,” Canadian Journal of
Civil Engineering. V. 32. no. 6 December, 2005 pp. 1103-1113.

Lachemi, M., Hossain, K.M.A., Labros, V., Nkinamubanzi, P.-C., Bouzoubaa, N.,
“Self-Consolidating Concrete Incorporating New Viscosity Modifying
Admixtures,” Cement and Concrete Research. V 34 2004 pp. 917-926.

Larson, K.H., Peterman, R.J., Esmaeily, A., “Determining the Time-Dependent and
Bond Characteristics of an SCC Mix for Kansas Prestressed Concrete Bridge
Girders.” Proceedings Second North American Conference on the Design and Use
of Self-Consolidating Concrete and the Fourth International RILEM Symposium on
Self-Compacting Concrete, (CD-ROM), Chicago, Illinois, October 2005.

Larson, K.H., Peterman, R.J., Esmaeily, A., “Determining the Time-Dependent
Deformations of Pretensioned Bridge Girders Cast with Self-Consolidating
Concrete,” Proceedings 2006 Concrete Bridge Conference. Dallas Texas, October
2006 (CD-ROM).

Logan, D., “Acceptance Criteria for Bond Quality of Strand for Pretensioned
Prestressed Concrete Applications,” PCI Journal, V. 42, No. 2, March-April, 1997,
pp. 52-90.

Mindess, 8. Young, F. Darwin, D., Concrete Second Edition, Prentice Hall, Upper
Saddle River, NJ, 2003, 644 pp.

310

[33]

[34]

[35]

[36]

[37]

[38]

[39]

[40]

[41]

[42]

[43]

Morgan Girgis, A.F, Tuan,C.Y., “Bond Strength and Transfer Length of
Pretensioned Bridge Girders Cast with Self-Consolidating Concrete,” PCI Journal
V. 50, No. 6, November-December 2005, pp. 72-87.

Naaman, A., Prestressed Concrete Analysis and Desig, McGraw-Hill, Inc. New
York, NY, 1982.

Naito, C. Hoover, M. “Applicability of Self Consolidating Concrete for Use in
Precast Bridge Beam Construction.” Proceedings Second North American
Conference on the Design and Use of Self-Consolidating Concrete and the Fourth
International RILEM Symposium on Self-Compacting Concrete, (CD-ROM),
Chicago, Illinois, October 2005.

Nasvik, J ., “The ABC’s of SCC,” Concrete Construction. March 2003. pp. 40-47.

Okamura, H. “Self-Compacting High-Performance Concrete,” Concrete
International. July 1997. pp. 50-54.

Okamura, H Ouchi, M., “Self-Compacting Concrete. Development, Present Use
and Future,” Proceedings 1" International RILEM Symposium on Self—Compacting
Concrete, Stockholm, Sweden, September 1999, pp. 3-14.

Okamura, H. Ozawa, K., “Mix Design for Self-Compacting Concrete,” Concrete
Library of Japanese Society of Civil Engineers, No. 25, June 1995, pp. 107-120.
Ouchi, M, Nakamura, S, Osterberg, T, Hallberg, S, Lwin, M., “Applications of Self-
Compacting Concrete in Japan, Europe, and The United States,” US Federal
Highway Administration Office of Bridge Technology. 2003 ISHPC. Available
from http://www.fliwa.dot.gov/bridge/scc.pdf.

Ozyildirim, C., “VDOT Early Experience With Self-Consolidating Concrete,”
Transportation Research Record, Journal of the Transportation Research Board.
Transportation Research Board of National Academies, Washington DC, N 1914
2005, pp. 81-84.

Ozyildirim, C., Lane, D.S., “Final Report: Evaluation of Self-Consolidating
Concrete,” Virginia Transportation Research Council, Charlottesville, Virginia,
June 2003.

Ramsburg, P. Ludirdja, D. Masek, O., “Utilizing SCC in Double Cycling
Production of Double Tee Beams,” Proceedings Second North American
Conference on the Design and Use of Self-Consolidating Concrete and the Fourth
International RILEM Symposium on Self-Compacting Concrete, (CD-ROM),
Chicago, Illinois, October 2005.

311

 

[44]

[451

[46]

[471

[48]

[491

[50]

[51]

[52]

[53]

[54]

Seible, F., and Igarashi, A., “Full Scale Testing of Masonry Structures Under
Simulated Seismic Loadings,” Experimental and Numerical Methods in Earthquake
Engineering,J. Donea and RM. Jones (Eds.), ECSC, EEC, EAEC, Brussels and
Luxemburg, Netherlands, 1991, pp. 119-148

Sonebi, M. Taminmi, A.K., Bartos, P.J.M., “Performance of Cracking Behavior of
Reinforced Beams Cast with Self—Consolidating Concrete,” AC1 Materials Journal,
V. 100 n. 6 November-December, 2003, pp. 492-500.

“Standard Testing Method for Compressive Strength of Cylindrical Concrete
Specimen (C39-05),” Annual Book of ASTM Standards, V 4.02 2006.

“Standard Testing Method for Splitting Tensile Strength of Cylindrical Concrete
Specimen (C496-04),” Annual Book of ASTM Standards, V 4.02 2006.

“Standard Testing Method for Passing Ability of Self-Consolidating Concrete by J—
Ring (C 1621-06),” Annual Book of ASTM Standards, V4.02 2006.

“Standard Testing Method of Slump Flow of Self-Consolidating Concrete (C1611-
05),” Annual Book of ASTM Standards, V 4.02 2006.

Su, N., Hsu, K.C., Chai, H.W., “A Simple Mix Design Method for Self-Compacting
Concrete,” Cement and Concrete Research, V 31 2001 pp. 1799-1807.

Tadros, M.K., Al-Omaishi, N., Seguirant, S.J., Gallt, J.G. “NCHRP Report 496:
Prestress Losses in Pretensioned High-Strength Concrete Bridge Girders,”
Transportation Research Board, Washington DC, 2003, 63 pgs.

Umehara, H. Hamada, D., Yamamuro, H., Oka, S. “Development and Usage of
Self-Compacting Concrete in Precast Concrete Field,” Proceedings 1" International
RILEM Symposium on Self-Compacting Concrete, Stockholm, Sweden, September
1999, pp. 705-716.

Wilson, N.D. Kiousis, P., “High-Strength SCC in Shear” Proceedings Second North
American Conference on the Design and Use of Self-Consolidating Concrete and
the Fourth International RILEM Symposium on Self-Compacting Concrete, (CD-
ROM), Chicago, Illinois, October 2005.

Yelton, R., “Flowing smoothly: Superintendent uses SCC to assist Milwaukee
motorists,” The Concrete Producer, V. 23, No. 11, November, 2005, pp. 46-49.

312

ER

tiiiiiiiiitihiiiiu

   

t - -----------il

1/