NOTICE

The United States Government and the State of Texas do not endorse products or
manufacturers. Trade or manufacturers’ names appear herein solely because they are
considered essential to the object of this report
 TECHNICAL REPORT DOCUMENTATION PAGE
1. Report No.: 2. Government Accession No.: 3. Recipient’s Catalog No.:
FHWA/TX-10/0-5849-1
4. Title and Subtitle: 5. Report Date: November 2010
Evaluating Existing Culverts for Load Capacity Allowing
for Soil Structure Interaction 6. Performing Organization Code:

8. Performing Organization Report

7. Author(s): William D. Lawson, Timothy A. Wood,
No. 0-5849-1
Charles D. Newhouse, Priyantha W. Jayawickrama
9. Performing Organization Name and Address: 10. Work Unit No. (TRAIS):
Texas Tech University
College of Engineering
11. Contract or Grant No. :
Box 41023
Lubbock, Texas 79409-1023 0-5849
12. Sponsoring Agency Name and Address 13. Type of Report and Period Cover:
Texas Department of Transportation Technical Report September 2007-
Research and Technology Implementation Office August 2009
P. O. Box 5080 Sponsoring Agency Code:
Austin, TX 78763-5080
15. Supplementary Notes:
This study was conducted in cooperation with the Texas Department of Transportation and the
Federal Highway Administration.
16. Abstract:
This study explores culvert load rating practices and procedures as applied to TxDOT’s archive of 1477 culvert
designs and their inventory of more than 13,000 in-service reinforced concrete box culverts. The problem is that
when older culverts are load-rated based on current AASHTO policy, many competent, serviceable culverts are
shown to be deficient, requiring load posting, retrofit or replacement. A disconnect exists between culvert
structural analysis practices and actual culvert performance. To address this challenge, the research focused on
development of a clear, repeatable and reliable procedure for TxDOT engineers and their consultants to use for
load rating culverts in the TxDOT roadway system. Articulated in TxDOT’s Culvert Rating Guide, the new
load rating procedure uses three increasingly-sophisticated analysis approaches, ranging from a direct stiffness
frame model to a production-oriented finite element model which accounts for soil-structure interaction.
Validation of the Culvert Rating Guide involved three major tasks. First, the researchers load-rated a
statistically representative sample of 100 of TxDOT’s culvert designs. Second, a parametric study was
performed to evaluate six independent variables associated with culvert load rating. Third, instrumented load
tests on three in-service culverts were conducted to compare measured demands with predicted values. This
work showed that the analytical methods in the Culvert Rating Guide produce conservative load ratings yet still
allow for reduction in excess over-conservatism in the load rating process.

18. Distribution Statement

No restrictions. This document is available to
17. Key Words
the public through the National Technical
load rating, culvert, soil-structure interaction
Information Service, Springfield, Virginia
22161 www.ntis.gov
Security Classif. (of 20. Security Classif.
this report) (of this page) No. of Pages 310 22. Price
Unclassified Unclassified
Form DOT F 1700.7 (8-72)
Evaluating Existing Culverts for Load Capacity Allowing for Soil Structure Interaction

by

William D. Lawson, P.E., Ph.D.

Timothy A. Wood, MSCE
Charles D. Newhouse, P.E., Ph.D.
Priyantha W. Jayawickrama, Ph.D.

Report Number
0-5849

Conducted for

Texas Department of Transportation

by the
Center for Multidisciplinary Research in Transportation
AUTHOR’S DISCLAIMER
The contents of this report reflect the views of the authors who are responsible for the
facts and the accuracy of the data presented herein. The contents do not necessarily reflect
the official view of policies of the Texas Department of Transportation or the Federal
Highway Administration. This report does not constitute a standard, specification, or
regulation.

1. PATENT DISCLAIMER

There was no invention or discovery conceived or first actually reduced to practice in the
course of or under this contract, including any art, method, process, machine, manufacture,
design or composition of matter, or any new useful improvement thereof, or any variety of
plant which is or may be patentable under the patent laws of the United States of America or
any foreign country.

2. ENGINEERING DISCLAIMER

Not intended for construction, bidding, or permit purposes.

3. TRADE NAMES AND MANUFACTURERS’ NAMES

The United States Government and the State of Texas do not endorse products or
manufacturers. Trade or manufacturers’ names appear herein solely because they are
considered essential to the object of this report.

Prepared in cooperation with the Texas Department of Transportation and the U.S.
Department of Transportation, Federal Highway Administration.

iv
TABLE OF CONTENTS
Author’s Disclaimer ................................................................................................................. iv
1. Patent Disclaimer ................................................................................................... iv
2. Engineering disclaimer .......................................................................................... iv
3. Trade Names and Manufacturers’ Names ............................................................. iv
Table of Contents ...................................................................................................................... v
List of Figures ......................................................................................................................... xii
List of Tables ......................................................................................................................... xvi
List of Appendix Tables....................................................................................................... xviii
List of Equations .................................................................................................................... xix
1. Introduction ................................................................................................................... 1
1. The Research Problem ............................................................................................ 1
2. The Culvert Load Rating Equation ......................................................................... 1
3. Culvert Load Rating Practices at TxDOT .............................................................. 3
4. Modified Research Direction – Development of a TxDOT Culvert Rating Guide 5
5. Research Focus: Validation of the Culvert Rating Guide ...................................... 6
6. Organization of this Research Report ..................................................................... 7
2. Culvert Load Rating Literature Review........................................................................ 9
1. Overview ................................................................................................................ 9
2. Academic Literature ............................................................................................... 9
1. Box Culvert Definitions ...................................................................................... 9
2. Soil Loads ......................................................................................................... 11
3. Vehicle Loads ................................................................................................... 14
3. Culvert Analysis Tools ......................................................................................... 15
1. Frame Models ................................................................................................... 15
2. Finite Element Models ...................................................................................... 16
4. AASHTO Policy Guidance .................................................................................. 17
5. State Department of Transportation Survey ......................................................... 18
1. Introduction ....................................................................................................... 18
2. State DOT Contacts .......................................................................................... 18
3. Interview Summaries ........................................................................................ 20
4. Results from the State DOT Survey.................................................................. 31

v
 5. Conclusions ....................................................................................................... 32
6. TxDOT District Survey ........................................................................................ 33
1. Purpose.............................................................................................................. 33
2. District Notes .................................................................................................... 33
3. Summary ........................................................................................................... 35
7. Conclusions .......................................................................................................... 36
3. Development of the Culvert Rating Guide ................................................................. 37
1. Overview .............................................................................................................. 37
2. A Brief Summary of the Culvert Load Rating Process ........................................ 38
3. Primary Development Goal: Reliability ............................................................ 42
1. Repeatability ..................................................................................................... 42
2. Capacity Calculation Approach ........................................................................ 42
3. Model Input Variables ...................................................................................... 42
4. Model Sophistication ........................................................................................ 43
5. Reliability Further Discussed............................................................................ 44
4. Selection of Analytical Models ............................................................................ 46
1. Analytical Models and Programs for Determining Demand ............................ 46
2. Value Analysis of Models ................................................................................. 53
5. Development of Level 1 Analysis: Two-Dimensional, Simply-Supported
Structural Frame Model ...................................................................................... 58
1. Model Construction ......................................................................................... 58
2. Loads ................................................................................................................ 58
3. Analytical Program – CULV-5 ....................................................................... 60
6. Development of Level 2 Analysis: Two Dimensional Structural Frame Model
with Soil Springs ................................................................................................. 60
7. Development of Level 3 Analysis: Two Dimensional Finite Element Soil-
Structure Interaction Model .................................................................................. 61
1. Live Load Distribution...................................................................................... 61
2. Production Load Rating Considerations ........................................................... 63
3. A Simple Beam Model Comparison Between CANDE and RISA .................. 65
4. Full Culvert Model Comparison Between CANDE and RISA .................... 68
5. RISA-2D as the Level 3 Analysis Program of Choice ................................. 70
8. Commentary on the Culvert Rating Guide ........................................................... 71
1. Chapter I: Introduction .................................................................................... 71
2. Chapter II: Policy Requirements .................................................................... 71

vi
 3. Chapter III: Culvert Load Rating Procedure ..................................................... 72
4. Chapter IV: Culvert Details ............................................................................ 72
5. Chapter V: Culvert Capacity Calculations .................................................... 73
6. Chapter VI: Analytical Modeling for Demand Loads .................................. 73
7. Chapter VII: The General Analytical Model for Culvert Load Rating ............. 74
8. Chapter VIII: Limitations ................................................................................. 74
9. Validation of the Culvert Rating Guide ................................................................ 75
4. Load Rating 100 TxDOT Culvert Designs ................................................................. 78
1. Overview .............................................................................................................. 78
2. Method .................................................................................................................. 79
1. Sample of 100 Representative Culvert Designs................................................ 79
2. Application of Load Rating Method to Domains ............................................. 87
3. Load Rating Results for 100 Culvert Designs ...................................................... 88
1. Operational Statement and Hypotheses ............................................................ 88
2. Presentation of the Research Findings .............................................................. 88
4. 1938 Era Results ................................................................................................... 91
1. Undifferentiated Inventory Ratings .................................................................. 91
2. Identification of Statistically-Significant Culvert Variables............................. 93
3. Differentiated Inventory Ratings by Significant Culvert Variables ................. 94
4. Controlling Failure Mode ................................................................................. 96
5. Controlling Critical Section .............................................................................. 97
5. 1946 Era Results ................................................................................................... 99
1. Undifferentiated Inventory Ratings .................................................................. 99
2. Identification of Statistically-Significant Culvert Variables........................... 100
3. Differentiated Inventory Ratings by Significant Culvert Variables ............... 101
4. Controlling Failure Mode ............................................................................... 103
5. Controlling Critical Section ............................................................................ 103
6. 1958 Era Structural Grade Steel (36 ksi) ............................................................ 105
1. Undifferentiated Inventory Ratings ................................................................ 105
2. Identification of Statistically-Significant Culvert Variables........................... 106
3. Differentiated Inventory Ratings by Significant Culvert Variables ............... 107
4. Controlling Failure Mode ............................................................................... 109
5. Controlling Critical Section ............................................................................ 109
7. 1958 Era Grade 60 steel (60 ksi) ........................................................................ 111

vii
 1. Undifferentiated Inventory Ratings ................................................................ 111
2. Identification of Statistically-Significant Culvert Variables........................... 112
3. Differentiated Inventory Ratings by Significant Culvert Variables ............... 113
4. Controlling Failure Mode ............................................................................... 114
5. Controlling Critical Section ............................................................................ 115
8. 2003 Era .............................................................................................................. 117
1. Undifferentiated Inventory Ratings ................................................................ 117
2. Identification of Statistically-Significant Culvert Variables........................... 118
3. Differentiated Inventory Ratings by Significant Culvert Variables ............... 119
4. Controlling Failure Mode ............................................................................... 120
5. Controlling Critical Section ............................................................................ 121
9. Discussion ........................................................................................................... 123
1. Evaluation of Hypotheses ............................................................................... 123
2. Evaluations of Failure Modes ......................................................................... 128
3. Evaluation of Controlling Critical Sections .................................................... 129
10. Conclusions ........................................................................................................ 131
1. Reliability and Modeling Sophistication ........................................................ 131
2. Principle Findings ........................................................................................... 131
3. Limitations of Research .................................................................................. 132
5. Parametric Analysis of Six Culvert Variables .......................................................... 133
1. Overview ............................................................................................................ 133
2. Sample Culvert Designs ..................................................................................... 134
3. Results ................................................................................................................ 135
1. Modulus of Subgrade Reaction, k................................................................... 135
2. Poisson’s Ratio................................................................................................ 140
3. Multibarrel Effects .......................................................................................... 143
4. Lateral Earth Pressure, LEP ............................................................................ 147
5. Depth of Fill .................................................................................................... 157
4. Conclusions ........................................................................................................ 164
6. Instrumented Load Tests on Three In-service Culverts ............................................ 166
1. Overview ............................................................................................................ 166
1. Purpose and Scope .......................................................................................... 166
2. Selection of Culverts for Instrumentation and Load Testing .......................... 166
2. Load Test Design ................................................................................................ 167

viii
 1. Load Rating Equation ..................................................................................... 167
2. Measuring Live Load Demands – Instrumentation Design ............................ 168
3. Application of Live Load – Field Test Procedure .......................................... 168
4. Predictive Analyses – Modeling the Load-Culvert-Soil System .................... 170
5. Obtaining Test Parameters .............................................................................. 171
3. Geotechnical Explorations at the Culvert Sites .................................................. 172
1. Description ...................................................................................................... 172
2. Field Drilling and Sampling............................................................................ 172
3. Laboratory Testing .......................................................................................... 172
4. Falling Weight Deflectometer Testing............................................................ 172
5. The Subsurface Profiles for Each Culvert Site ............................................... 173
6. Determination of Modulus of Subgrade Reaction Parameters ....................... 173
7. Determination of Soil Modulus Parameters .................................................... 174
4. The Shallowater Culvert Instrumented Load Test .............................................. 179
1. Culvert Condition............................................................................................ 179
2. Summary of Parameters .................................................................................. 180
3. Instrumentation Plan ....................................................................................... 181
4. Results for Shallowater Culvert with 2’ of Fill ............................................... 182
5. Results for Shallowater Culvert with 4’ of Fill ............................................... 193
5. The Plainview Culvert Instrumented Load Test ................................................. 201
1. Culvert Condition............................................................................................ 201
2. Summary of Parameters .................................................................................. 202
3. Instrumentation Plan ....................................................................................... 203
4. Load Rating..................................................................................................... 204
5. Live Load Moment – Predicted and Measured ............................................... 205
6. Dead Load Moment – Predicted ..................................................................... 206
7. Live Load “Goodness” .................................................................................... 207
8. Load Rating “Goodness” ................................................................................ 208
9. Critical Section Analysis................................................................................. 208
10. Deflections ...................................................................................................... 211
11. Summary of Findings for the Plainview Culvert ............................................ 212
6. The Tulia Culvert Instrumented Load Test ........................................................ 213
1. Culvert Condition............................................................................................ 213
2. Summary of Parameters .................................................................................. 214

ix
 3. Instrumentation Plan ....................................................................................... 215
4. Load Rating..................................................................................................... 216
5. Live Load Moment – Predicted and Measured ............................................... 217
6. Dead Load Moment – Predicted ..................................................................... 218
7. Live Load “Goodness” .................................................................................... 219
8. Load Rating “Goodness” ................................................................................ 220
9. Critical Section Analysis................................................................................. 220
10. Deflections ...................................................................................................... 222
11. Summary of Findings for the Tulia Culvert .................................................... 222
7. Limited Dead Load Investigation ....................................................................... 223
1. Dead Load Moment – Predicted and Measured.............................................. 223
2. Dead Load “Goodness”................................................................................... 224
3. Critical Section Analysis................................................................................. 224
4. Deflections ...................................................................................................... 227
5. Summary of Findings for the Limited Dead Load Investigation .................... 228
8. Summary and Conclusions ................................................................................. 229
7. Conclusions and Research Limitations ..................................................................... 230
1. Conclusions ........................................................................................................ 230
1. Evaluation of TxDOT’s Culvert Load Rating Practices and Procedures ....... 230
2. Development of the Culvert Rating Guide ..................................................... 230
3. Evaluation of TxDOT Culvert Designs and Analysis Methods...................... 230
4. Parametric Analysis ........................................................................................ 231
5. Instrumented Load Tests on Three In-Service Culverts ................................. 231
2. Limitations of Research ...................................................................................... 232
1. Culvert Type ................................................................................................... 232
2. Culvert Drainage Condition ............................................................................ 232
3. Soil Elastic Modulus ....................................................................................... 232
4. Depth of Fill .................................................................................................... 232
5. Live Load Demand Measurements ................................................................. 232
6. Dead Load Demand Measurements ................................................................ 233
3. Recommendations .............................................................................................. 233
8. Bibliography ............................................................................................................. 235
Appendix A 1938 Era Load Ratings ..................................................................................... 241
Appendix B 1946 Era Load Ratings ..................................................................................... 245

x
Appendix C 1958 Era 36ksi Steel Load Ratings .................................................................. 248
Appendix D 1958 Era 60ksi Steel Load Ratings .................................................................. 255
Appendix E 2003 Era Load Ratings ..................................................................................... 262
Appendix F Soil Boring Logs ............................................................................................... 266
Appendix G FWD Data ........................................................................................................ 288
Appendix H Maximum and Minimum Strains From Load Testing ..................................... 297

xi
 LIST OF FIGURES
Figure 2.1. Embankment culvert............................................................................................... 9
Figure 2.2. Trench culvert....................................................................................................... 10
Figure 2.3. Imperfect trench culvert ....................................................................................... 10
Figure 2.4. Negative soil arching ............................................................................................ 11
Figure 2.5. Postive soil arching............................................................................................... 12
Figure 2.6. Soil pressure distributions due to culvert deformation ......................................... 13
Figure 2.7. Culvert deformation due to lateral load only ........................................................ 13
Figure 3.1. Load Posting Guidelines....................................................................................... 39
Figure 3.2. Flow chart depicting the TxDOT culvert load rating process. ............................. 40
Figure 3.3. Prepackaged, two-dimensional culvert models loading diagram ........................ .47
Figure 3.4. General, two-dimensional, linear-elastic structural frame
models loading diagram. ............................................................................................. 48
Figure 3.5. General, two-dimensional, linear-elastic soil-structure interaction
models loading ........................................................................................................... 50
Figure 3.6. Two dimensional Frame Load Geometry ............................................................. 59
Figure 3.7. Two dimensional frame analysis comparisions of live loads ............................... 59
Figure 3.8. Two dimensional finite element AASHTO LRFD live load geometry ................ 62
Figure 3.9. Two-dimensional FEA Comparisons of Live Loads ............................................ 63
Figure 3.10. Demands from simple beam model comparison. ............................................... 67
Figure 3.11. Load rating Comparison of full scale culvert ..................................................... 68
Figure 3.12. Moment rating of full scale culvert .................................................................... 69
Figure 4.1. TxDOT’s AS-BUILT Culvert Population ............................................................ 79
Figure 4.2. Culvert designs vs built vs sample by design era ................................................. 80
Figure 4.3. Distribution of the number of spans for designs, built and sample populations .. 83
Figure 4.4. Distribution of the length of spans for designs, built and sample populations..... 84
Figure 4.5. Distribution of box heights for designs and sample populations ......................... 85
Figure 4.6. Distribution of maximum depth of fill for designs and sample populations ........ 86
Figure 4.7. Moment critical sections for culverts without haunches. ..................................... 90
Figure 4.8. 1938 era inventory ratings: undifferentiated ........................................................ 91
Figure 4.9. 1938 era statistical analysis for significant independent variables ....................... 93
Figure 4.10. 1938 era inventory ratings: direct traffic ............................................................ 94
Figure 4.11. 1938 era inventory ratings: 3' and 4' of fill ......................................................... 95
Figure 4.12. 1938 era inventory ratings: 6' of fill ................................................................... 95
Figure 4.13. 1938 era inventory ratings: 8' to 18' of fill ......................................................... 96
Figure 4.14. 1938 era controlling failure modes ..................................................................... 96
Figure 4.15. 1938 era controLling critical sections: 0 to 6' of fill .......................................... 97
Figure 4.16. 1938 era controlLing critical sections: 8' to 18' of fill ........................................ 98
Figure 4.17. 1946 era inventory ratings: undifferentiated ...................................................... 99

xii
Figure 4.18. 1946 era statistical analysis for significant independent variables ................... 100
Figure 4.19. 1946 era inventory ratings: MBC-3.................................................................. 101
Figure 4.20. 1946 era inventory ratings: MC10-3-45 ........................................................... 102
Figure 4.21. 1946 era inventory ratings: FM-MBC-3-26 ..................................................... 102
Figure 4.22. 1946 era controlling failure modes ................................................................... 103
Figure 4.23. 1946 era controlling critical sections: MBC-3 ................................................. 103
Figure 4.24. 1946 era controlling critical sections: MC-10-3-45 ......................................... 104
Figure 4.25. 1946 era controlling critical sections: FM-MBC-3-26 ..................................... 104
Figure 4.26. 1958 era 36ksi steel inventory ratings: undifferentiated .................................. 105
Figure 4.27. 1958 era 36ksi steel statistical analysis for significant independent variables . 106
Figure 4.28. 1958 era 36ksi steel inventory ratings: 2' of fill ............................................... 107
Figure 4.29. 1958 era 36ksi steel inventory ratings: 4' of fill ............................................... 108
Figure 4.30. 1958 era 36ksi steel inventory rating: 6' of fill ................................................. 108
Figure 4.31. 1958 era 36ksi steel controlling failure modes ................................................. 109
Figure 4.32. 1958 era 36ksi steel controlling critical sections: 2' of fill ............................... 109
Figure 4.33. 1958 era 36ksi steel controlling critical sections: 4' of fill ............................... 110
Figure 4.34. 1958 era 36ksi steel controlling critical sections: 6’ of fill .............................. 110
Figure 4.35. 1958 era 60ksi steel inventory ratings: undifferentiated .................................. 111
Figure 4.36. 1958 era 60ksi steel statistical analysis for significant indepedent variables ... 112
Figure 4.37. 1958 era 60ksi steel inventory ratings: 2' of fill ............................................... 113
Figure 4.38. 1958 era 60ksi steel inventory rating: 4' of fill ................................................. 113
Figure 4.39. 1958 era 60ksi steel inventory ratings: 6' of fill ............................................... 114
Figure 4.40. 1958 era 60ksi steel controlling failure modes ................................................. 114
Figure 4.41. 1958 era 60ksi steel controlling critical sections: 2’ of fill .............................. 115
Figure 4.42. 1958 era 60ksi steel controlling critical sections: 4' of fill ............................... 115
Figure 4.43. 1958 era 60ksi steel controlling critical sections: 6' of fill ............................... 116
Figure 4.44. 2003 era inventory ratings: undifferentiated .................................................... 117
Figure 4.45. 2003 era statistical analysis for significant indepednet variables ..................... 118
Figure 4.46. 2003 era inventory ratings: 7' to 16' of fill ....................................................... 119
Figure 4.47. 2003 era inventory ratings: 20' of fill ............................................................... 119
Figure 4.48. 2003 era inventory ratings: 23' of fill ............................................................... 120
Figure 4.49. 2003 era controlling failure mode .................................................................... 120
Figure 4.50. 2003 era controlling critical sections: 7' to 16' of fill ....................................... 121
Figure 4.51. 2003 era controlling critical sections: 20' of fill ............................................... 121
Figure 4.52. 2003 era controlling crital sections: 23' of fill .................................................. 122
Figure 5.1. Modulus of Subgrade Reaction vs. the Inventory Rating for Seven Representative
Culvert Designs............................................................................................................ 137
Figure 5.2. Poisson’s Ratio vs. the Inventory Rating for Seven Representative Culvert
Designs.......................................................................................................................... 142

xiii
Figure 5.3. Number of Barrels vs. the Inventory Rating for Seven Representative Culvert
Designs......................................................................................................... ................. 146
Figure 5.4. Lateral Earth Pressure vs. the Inventory Rating for Seven Representative Culvert
Designs using CULV-5. .................................................................................................. 150
Figure 5.5. Lateral Earth Pressure vs. Inventory Rating for Seven Representative Culvert
Designs using RISA-2D.Modulus of Elasticity, E .......................................................... 151
Figure 5.6. Modulus of Elasticity vs. the Inventory Rating for Seven Representative Culvert
Designs............................................................................................................................ 156
Figure 5.7. Inventory Rating vs. Depth of Fill for MC-7-16 4 7x4w16. .............................. 160
Figure 5.8. Inventory Rating vs. Depth of Fill for MC-10-20 5 10x10w20. ........................ 160
Figure 5.9. Inventory Rating vs. Depth of Fill for MC7-1 3 7x6w2..................................... 161
Figure 5.10. Inventory Rating vs. Depth of Fill for MC9-2 2 9x9w4................................... 161
Figure 5.11. Inventory Rating vs. Depth of Fill for MC6-2 6 6x3w6................................... 161
Figure 5.12. Inventory Rating vs. Depth of Fill for MBC-5-34 5 9x8w0............................. 162
Figure 5.13. Inventory Rating vs. Depth of Fill for MBC1-44-F 2 5x4w6. ......................... 162
Figure 6.1. Critical sections .................................................................................................. 168
Figure 6.2. Gage line consisting of two foot grid marks to determining truck positions. .... 169
Figure 6.3. Live load configurations: a) truck straddling gage line b) wheel on gage line
c) Two trucks straddling gage line. .................................................................................. 169
Figure 6.4. Idealized soil profile and raw field and laboratory soil strength data for the three
culvert sites. .................................................................................................................... 176
Figure 6.5. Idealized soil profile and derived modulus of subgrade reaction values for the
three culvert sites. ............................................................................................................ 177
Figure 6.6. Idealized soil profile and derived elastic modulus values for the three culvert
sites. ................................................................................................................................. 178
Figure 6.7. Shallowater location. .......................................................................................... 179
Figure 6.8. Shallowater instrumentation plan. ...................................................................... 181
Figure 6.9. Load rating for Shallowater culvert: 2’ of fill .................................................... 182
Figure 6.10. Live load moments predicted and measured for Shallowater
culvert: 2’ of fill.
. .................................................................................................................................... 183
Figure 6.11. Dead load predicted for Shallowater culvert: 2’ of fill..................................... 184
Figure 6.12. Live load goodness plot for Shallowater culvert: 2’ of fill .............................. 186
Figure 6.13. Load rating goodness plot for Shallowater culvert: 2’ of fill ........................... 188
Figure 6.14. Measured deflections in millimeters for the Shallowater culvert: 2’ of fill ..... 191
Figure 6.15. Load rating for Shallowater culvert: 4’ of fill .................................................. 193
Figure 6.16. Live load moments predicted and measured for Shallowater culvert: 4’ of fill.
............................................................................................................................................... 194
Figure 6.17. Dead load predicted for Shallowater culvert: 4’ of fill..................................... 195
Figure 6.18. Live load goodness plot for Shallowater culvert: 4’ of fill .............................. 196
Figure 6.19. Load rating goodness plot for Shallowater culvert: 4’ of fill ........................... 197

xiv
Figure 6.20. Measured deflections in millimeters for the Shallowater culvert: 4’ of fill ..... 199
Figure 6.21. Plainview location. ........................................................................................... 201
Figure 6.22. Plainview instrumentation plan. ....................................................................... 203
Figure 6.23. Load rating for Plainview culvert: 3.5’ of fill .................................................. 204
Figure 6.24. Live load moments predicted and measured for Plainview culvert:3.5’ fill. ... 205
Figure 6.25. Dead load predicted for Plainview culvert: 3.5’ of fill. .................................... 206
Figure 6.26. Live load goodness plot for Plainview culvert: 3.5’ of fill............................... 207
Figure 6.27. Load rating goodness plot for Plainview culvert: 3.5’ of fill ........................... 208
Figure 6.28. Measured deflections in millimeters for the Plainview culvert: 3.5’ of fill ..... 211
Figure 6.29. Tulia location. ................................................................................................... 213
Figure 6.30. Tulia instrumentation plan................................................................................ 215
Figure 6.31. Load rating for Tulia culvert: 1.5’ of fill .......................................................... 216
Figure 6.32. Live load moments predicted and measured for Tulia culvert: 1.5’ of fill....... 217
Figure 6.33. Dead load predicted for Tulia culvert: 1.5’ of fill. ........................................... 218
Figure 6.34. Live load goodness plot for Tulia culvert: 1.5’ of fill ...................................... 219
Figure 6.35. Load rating goodness plot for Tulia culvert: 1.5’ of fill ................................... 220
Figure 6.36. Measured deflections in millimeters for the Tulia culvert: 1.5’ of fill ............. 222
Figure 6.37. The change in dead load moment, predicted and measured, for the Shallowater
culvert from 2' to 4' of fill. .................................................................................................... 223
Figure 6.38. Change in dead load goodness for Shallowater culvert from 2' to 4' of fill. .... 224
Figure 6.39. Measured deflections in millimeters for Shallowater culvert from 2' to 4' fill. 227

xv
 LIST OF TABLES
Table 1.1. Documented design/rating considerations from codes, policy and TxDOT
examples. .................................................................................................................................. 3
Table 2.1. State DOT response ............................................................................................... 19
Table 2.2. Load types .............................................................................................................. 31
Table 2.3. Load application .................................................................................................... 31
Table 2.4. Analytical programs ............................................................................................... 31
Table 3.1. Summary of models ............................................................................................... 46
Table 3.2. Summary of culvert applicable models ................................................................. 52
Table 3.3. Qualitative Evaluation of Model sophistication and reliabilty .............................. 53
Table 3.4. Qualitative Evaluation of Model input variables ................................................... 54
Table 3.5. Qualitative Evaluation of ease of model generation .............................................. 55
Table 3.6. Qualitative Evaluation of Model load rating suitability ........................................ 56
Table 3.7. Summary of Qualitative Evaluation of Models for production load rating ........... 57
Table 4.1. Evaluation of hypotheses by era. ......................................................................... 123
Table 4.2. Evaluation of hypotheses for low fill culverts ..................................................... 124
Table 4.3. Evaluation of hypotheses for medium fill culverts .............................................. 125
Table 4.4. Evaluation of hypotheses for high fill culverts .................................................... 126
Table 4.5. Prominent controLling critical sections ............................................................... 129
Table 5.1. Parametric Analysis Test Matrix ......................................................................... 133
Table 5.2. Parametric Sample ............................................................................................... 134
Table 5.3. Percent Difference in Inventory Ratings. ............................................................ 135
Table 5.4. The Load Rating for 7 Representative Culvert Designs with Three Varying
Modulus of Subgrade Reaction, k (psi). ............................................................................... 136
Table 5.5. Number of Culvert Designs Out of a Sample of 100 Representative Designs which
Meet Specific Criteria for the Maximum Percent Difference in the Inventory Rating. ....... 139
Table 5.6. Percent Difference in Inventory Ratings. ............................................................ 140
Table 5.7. The Load Rating for 7 Representative Culvert Designs with Three Poisson’s
Ratios, ν. ............................................................................................................................... 141
Table 5.8. Percent difference in Inventory Ratings. ............................................................. 143
Table 5.9.The Load Rating for 7 Representative Culvert Designs with Increasing Number of
Barrels. .................................................................................................................................. 145
Table 5.10. Change in Inventory Rating and Percent Difference in Inventory Rating. ........ 147
Table 5.11. The Load Rating for 7 Representative Culvert Designs with Increasing Lateral
Earth Pressure. ...................................................................................................................... 149
Table 5.12. Required confidence for modulus of elasticity, E.............................................. 153
Table 5.13. The Load Rating for 7 Representative Culvert Designs with Increasing Modulus
of Elasticity. .......................................................................................................................... 155

xvi
Table 5.14. The Load Rating for 7 Representative Culvert Designs with Increasing Depth of
Fill. ........................................................................................................................................ 158
Table 5.15. Summary of Parametric Study Conclusions. ..................................................... 164
Table 6.1. Live load predicted to actual ratio thresholds ...................................................... 185
Table 6.2. Operating rating goodness thresholds .................................................................. 187
Table 6.3. Live load demand problem critical sections for Shallowater culvert: 2’ of fill ... 190
Table 6.4. Operating rating problem critical sections for Shallowater culvert: 2’ of fill ..... 190
Table 6.5. Live load demand problem critical sections for Shallowater culvert: 4’ of fill ... 198
Table 6.6. Operating rating problem critical sections for Shallowater culvert: 4’ of fill ..... 198
Table 6.7. Live load demand problem critical sections for Plainview culvert: 3.5’ of fill ... 210
Table 6.8. Operating rating problem critical sections for Plainview culvert: 3.5’ of fill...... 210
Table 6.9. Live load demand problem critical sections for Tulia culvert: 1.5’ of fill........... 221
Table 6.10. Operating rating problem critical sections for Tulia culvert: 1.5’ of fill ........... 221
Table 6.11. Live load demand problem critical sections for the change in dead load in the
Shallowater culvert from 2’ to 4’ of fill ................................................................................ 225
Table 6.12. Operating rating problem critical sections for the change in dead load in the
Shallowater culvert from 2’ to 4’ of fill ................................................................................ 226

xvii
LIST OF APPENDIX TABLES
Appendix Table 1. 1938 Era Sample Details and Culv-5 Load ratings ................................ 242
Appendix Table 2. 1938 era RISA-2D with springs load ratings ......................................... 243
Appendix Table 3. 1938 era RISA-2D with LEFE load ratings ........................................... 244
Appendix Table 4. 1946 era sample details and CULV-5 load ratings ................................ 246
Appendix Table 5. 1946 era RISA-2D with springs load ratings ......................................... 246
Appendix Table 6. 1946 era RISA-2D with LEFE load ratings ........................................... 247
Appendix Table 7. 1958 era 36ksi steel sample details and CULV-5 load ratings .............. 249
Appendix Table 8. 1958 era 36ksi steel RISA-2D with springs load ratings ....................... 251
Appendix Table 9. 1958 era 36ksi steel RISA-2D with LEFE load ratings ......................... 253
Appendix Table 10. 1958 era 60ksi steel sample details and CULV-5 load ratings ............ 256
Appendix Table 11. 1958 era 60ksi steel RISA-2D with springs load ratings ..................... 258
Appendix Table 12. 1958 era 60ksi steel RISA-2D with LEFE load ratings ....................... 260
Appendix Table 13. 2003 era sample details and CULV-5 load ratings .............................. 263
Appendix Table 14. 2003 era RISA-2D with springs load ratings ....................................... 264
Appendix Table 15. 2003 era RISA-2D with LEFE load ratings ......................................... 265
Appendix Table 16. Maximum and minimum strain (με) from Live Load test on Shallowater
Culvert under 2' of fill. .......................................................................................................... 298
Appendix Table 17. Maximum and minimum strain (με) from Live Load test on Shallowater
Culvert under 4' of fill. .......................................................................................................... 299
Appendix Table 18. strain (με) from Dead Load test on Shallowater Culvert do to a change in
fill depth from 2’ to 4'. .......................................................................................................... 300
Appendix Table 19. Maximum and minimum strain (με) from Live Load test on Plainview
Culvert under 3.5' of fill. ....................................................................................................... 301
Appendix Table 20. Maximum and minimum strain (με) from Live Load test on Tulia
Culvert under 1.5' of fill. ....................................................................................................... 302

xviii
 LIST OF EQUATIONS
Equation 1.1. The Rating Factor Equation from AASHTO MCEB 6.5.1.1-A. ........................ 2
Equation 6.1. Curvature from inside and outside strain........................................................ 170
Equation 6.2. Curvature from inside strain. .......................................................................... 170
Equation 6.3. Measured moment from strain........................................................................ 170

xix
TxDOT 0-5849
 1. INTRODUCTION

1. THE RESEARCH PROBLEM

This report presents findings of a two-year research study of culvert load rating practices
and procedures as applied to reinforced concrete box culverts that have been designed, built
and maintained by the Texas Department of Transportation (TxDOT). The problem facing
TxDOT is that when roads or rights-of-way are widened and/or raised, culverts which pass
under them need to be reassessed for the extension and/or increased soil loads. These in-
service culverts, which may have performed satisfactorily for many years, must be
reanalyzed using current American Association of State Highway and Transportation
Officials (AASHTO) methods. AASHTO has revised their culvert rating guidelines upward
over the years to impose larger loads on buried structures, so this means that many older
culverts, reanalyzed in the process of designing extensions, are seen as deficient, requiring
either retrofit or replacement.
The problem, therefore, is that using today’s culvert analysis methods on older culverts
does not appear to accurately predict performance and structural capacity. A disconnect
exists between current culvert structural analysis methods and actual culvert performance.
This calls for research into the actual performance of the culvert-soil system which can be
used to develop rational performance-based load rating guidelines. The objective is to satisfy
current AASHTO policy yet not require unnecessary replacement or retrofit of older,
competently-performing in-service culverts.

2. THE CULVERT LOAD RATING EQUATION

The American Association of State Highway and Transportation Officials (AASHTO)

defines “load rating” as the maximum truck tractor tonnage, expressed in terms of HS load
designation, permitted across a culvert. The load rating is expressed in terms of two separate
ratings – an Inventory Rating and an Operating Rating. The Inventory Rating (IR) is the
maximum truck load that can safely utilize the culvert for an indefinite period of time
(AASHTO, 2003; TxDOT, 2002). The Operating Rating (OR) is the absolute maximum
permissible truck load that may use the culvert. Load ratings are based on the culvert’s
current condition and are determined through analysis and engineering judgment by
comparing the culvert structure’s capacity and dead load demand to live load demand.

TxDOT 0-5849 1
 The AASHTO Manual for Condition Evaluation of Bridges (MCEB)(AASHTO, 2003),
and the AASHTO Standard Specification for Highway Bridges (SSHB) (AASHTO, 2002)
provide governing policy for load rating culverts. The rating factor identified in Equation 1.1
is the mathematical expression of the culvert load rating process, and as such, is the focus of
this research study. This rating factor is the ratio of the structural capacity minus the dead
load demand to the live load demand.
EQUATION 1.1. THE RATING FACTOR EQUATION FROM AASHTO MCEB 6.5.1.1-A.
𝐶 − 𝐴1 𝐷
𝑅𝐹 =
𝐴2 𝐿(1 + 𝐼)
where: RF = the rating factor
C = the structural capacity of the member
D = the dead load effect on the member
L = the live load effect on the member
I = the impact factor, IM from SSHB 3.8.2.3
A1 = 1.3 = factor for dead loads, from MCEB 6.5.3
A2 = 2.17 for Inventory Level = factor for live loads, from MCEB 6.5.3
= 1.3 for Operating Level = factor for live loads, from MCEB 6.5.3
The load rating equation is deceptively simple; whereas, the load rating process is
arguably complex. This becomes apparent when one considers that Equation 1.1 must be
used to determine the rating factor for each critical section of the culvert (corners, midspans,
top and bottom slabs, and interior and exterior walls), for each demand type (moment, shear
and thrust), for different load envelopes (maximum, minimum) at each rating level (inventory
and operating). The lowest inventory rating factor and the lowest operating rating factor
control the load rating for the culvert.
The culvert load rating factors directly depend on how culvert capacity, dead load
demands, and live load demands are established. Typical practice is to determine culvert
capacity based on the details found on the original construction documents in combination
with historical material property assumptions which are correlated by visual inspection of the
culvert condition. The demand calculation process is accomplished through analytical
modeling. Here, the AASHTO SSHB specifies a soil unit weight, equivalent fluid weight for
lateral loads, and live load distributions through the soil, but it does not specify the type of
demand model that should be used or specifically how to apply the loads to the model. This
means that engineering decisions about modeling practices and procedures are necessary, and
the assumptions, simplifications and mathematical structures of demand modeling tools can
have a significant effect on the culvert load rating analysis.

TxDOT 0-5849 2
 3. CULVERT LOAD RATING PRACTICES AT TXDOT

The design and analysis of culverts at TxDOT has for many years incorporated TxDOT’s
own in-house computer analysis program, CULV-5. CULV-5 is an MS-DOS program
developed and distributed by TxDOT. The heart of the program is a two-dimensional direct
stiffness frame model. Documentation supporting CULV-5 includes the Version 1.71
Readme file (TxDOT, 2004), Input Guide (TxDOT, 2003), and CULV5 – Concrete Box
Analysis Program (TxDOT, 2003).
Upon initiation of the project, TxDOT engineers discussed their culvert load rating
practices and procedures, including several load rating models. These included CULV-5
which was widely considered to be overly conservative. TxDOT engineers also noted their
practice of ignoring bottom slab failures identified from CULV-5 load rating analyses. RISA
frame models were mentioned as the preferred tool for TxDOT consultants. In general RISA
frame models are considered to be slightly more accurate than CULV-5 models; however, it
was noted that consultants refuse to ignore bottom slab failures regardless of the analysis
tool.
Relative to articulation of their culvert load rating procedures, TxDOT engineers cited a
DRAFT culvert rating guide developed for summer interns. To facilitate work on this
research project, TechMRT requested of TxDOT their current culvert load rating procedure
and some examples. In response, TxDOT provided three documents:
1. An example hand calculation with selected culvert detail sheets
2. An informal discussion of TxDOT’s culvert load rating procedure
3. The DRAFT culvert rating guide corresponding to a RISA-3D analytical model
The original research goals focused around modification of these culvert load rating
procedures. However, as the TechMRT researchers began their work of reviewing these and
other TxDOT documents that have to do with culvert load rating, four different sets of initial
assumptions emerged. Table 1.1 summarizes these initial assumptions.
TABLE 1.1. DOCUMENTED DESIGN/RATING CONSIDERATIONS FROM CODES, POLICY AND TXDOT EXAMPLES.
Txdot Draft Culvert Rating Guide
Current Aashto Pre-1948 Txdot Current Txdot
Design Considerations Example
Manual Policy Policy

(1) (2) (3) (4) (5)

Vertical Earth Pressure 120 pcf (.7)(120pcf) 120 pcf 125 pcf
Horizontal Earth Pressure 60 or 30 pcf 30 pcf 40 pcf 41 pcf
Live Load HS-20 12k wheel 16k wheel HS-20
Live Load Distribution 1.75*D Westeguard’s 1.7*D 1.75*D
2’ surcharge 2’ surcharge + 2’ surcharge
Corner Moments Full HEP
Full HEP Full HEP Full HEP
Positive Moments Half HEP No HEP Half HEP Half HEP
Shear Φ = .85 Slab Design
Thrust Φ = .9
Soil-Structure Interaction Empirical
Moment Dist. RISA-3D
Analytical Model
By Hand Plate Model

TxDOT 0-5849 3
 The first set of design parameters (Table 1.1, Column 2) depicts guidance from the
current AASHTO MCEB (AASHTO, 2003)and SSHB (AASHTO, 2002). The AASHTO
MCEB contains the load rating equations, the recommended material property assumptions
and structural capacity equations. The AASHTO SSHB provides guidance for determining
the dead and live loads used in the MCEB’s load rating equations. It provides details for
determining vehicle live loads, culvert specific live load distribution, culvert specific soil
related dead loads and reinforced cast-in-place box culvert empirical soil structure interaction
equations.
The second set of parameters (Table 1.1, Column 3) is from the old TxDOT culvert load
rating approach used before 1948. These values are published in the TxDOT Bridge Design
Manual (TxDOT, 2001) which provides loading guidance and historical and recommended
design parameters for reinforced concrete box culverts.
The third set of parameters (Table 1.1, Column 4) is from the current TxDOT approach
as represented by the TxDOT Bridge Design Manual, the TxDOT Bridge Inspection Manual
and the AASHTO MCEB. The TxDOT Bridge Inspection Manual (TxDOT, 2001) points out
some limitations of historic design parameters, but generally directs load rating engineers to
the AASHTO MCEB (AASHTO, 2003).
The fourth set of parameters (Table 1.1, Column 5) is from the TxDOT DRAFT culvert
rating guide corresponding to the RISA 3D analytical model. The parameters embodied in
this example differ from the parameters outlined in the AASHTO codes and TxDOT
Manuals.
The existence of four culvert load rating approaches caused confusion relative to the load
rating process. Selection of governing policy is a key issue, because a clear, reliable, and
repeatable culvert load rating procedure needs to be solidly built on authoritative policy. A
meeting of the Project Monitoring Committee was convened to resolve this issue. At that
point a decision was made to base all work on current AASHTO policy requirements.
Another issue of concern relative to establishing a clear, repeatable procedure for culvert
load rating is the diversity of analytical tools available to the culvert load rater. At the first
project meeting, TxDOT’s in-house culvert analysis program (CULV-5) and RISA frame
models were discussed. In TxDOT’s DRAFT culvert rating guide, a three dimensional plate
model expressed in terms of RISA-3D was used. Other analytical models and computer
programs are also available, each with their advantages and disadvantages. The problem is
that each model will produce a different load rating. Practically this means that until a model
or tool is specified in the load rating procedure, even given a consistent set of input
parameters, no repeatable load rating can be determined.
The net outcome of this work was the realization that no one consistent, reliable
procedure for load rating culverts existed within TxDOT. This called for a significant change
in the research direction.

TxDOT 0-5849 4
 4. MODIFIED RESEARCH DIRECTION – DEVELOPMENT OF A TXDOT CULVERT RATING
GUIDE

In TechMRT’s quest for the unified TxDOT culvert load rating procedure the following
things became apparent:
1. TxDOT’s current culvert load rating procedure was not well-articulated; that is, it was
not clear, consistent with AASHTO policy, and repeatable.
2. TechMRT would not be able to modify the current TxDOT culvert load rating
procedure until that procedure became well-articulated.
It was determined that TxDOT’s needs would be better served by redirecting research effort
toward the development, refinement, and validation of a new and improved TxDOT culvert
rating procedure.
The new procedure should meet several requirements. The first is that it must be based
upon authoritative AASHTO code. This resolves any policy issues for TxDOT’s culvert
rating process. Any engineer experienced in load rating should be able to understand how the
assumptions and decisions in the procedure stem from the code.
A second requirement is that the procedure be conceptually clear. This helps the engineer
to see and understand exactly what physical conditions are being modeled in the load-rating
process. Distinguishing between the model for analysis (the conceptual plan) and the method
of analysis (the computer program) is the issue here. Clarifying this relationship would help
to reduce confusion and error.
Third, the new procedure should be general enough to be used with many analytical
methods. The engineer should have the freedom to use the analysis method with which
he/she is most comfortable. For example, if the engineer is not familiar with RISA-2D, it
should be acceptable to use another frame analysis program.
Fourth, the new procedure should incorporate escalating levels of analytical rigor, the
goal being to balance load rating effort with reliability of the findings. This is particularly
important for production load rating of culverts. The first levels of analysis would be
relatively simple, quick and easy to use, and built around a two-dimensional frame model or
moment distribution. Higher level models would incorporate soil-structure interaction. The
highest levels would allow for enhanced (project-specific) input values, and the use of a
three-dimensional finite-element solution with soil-structure interaction. These highest levels
would be considered research oriented rather than production oriented.
Finally, the new culvert rating procedure would need to be validated through application
of the procedure to a diverse sample of TxDOT culvert designs, by parametric analyses
which compare results among the different modeling approaches, and by field observation
and testing. The culvert rating procedure should express a valid relationship between
predictions based upon the analytical models and actual culvert stresses determined from in-
service performance.

TxDOT 0-5849 5
 To reiterate, the focus for this project was redirected toward developing a new, improved
culvert rating procedure that incorporates the following features:
1. It is based on authoritative AASHTO code
2. It is conceptually clear
3. It is generally applicable to many analytical methods
4. It incorporates escalating levels of analytical rigor
5. It is validated through research activities focusing on breadth of application,
sensitivity to parameters, and correspondence with actual culvert performance
To satisfy these objectives, the research team developed and published TxDOT’s Culvert
Rating Guide (TXDOT, 2009). The Culvert Rating Guide first appeared in DRAFT form and
went through multiple iterations of review and refinement. This work comprised the bulk of
the first year of effort for this research study.

5. RESEARCH FOCUS: VALIDATION OF THE CULVERT RATING GUIDE

Having articulated TxDOT’s recommended practices and procedures for culvert load
rating in the Culvert Rating Guide, the second year of research effort was directed toward
validating these practices and procedures. The approach for validating the culvert load rating
practices and procedures in the Culvert Rating Guide involved three research thrusts.
First, the Culvert Rating Guide was applied to a statistically representative sample of 100
of TxDOT’s 1477 unique culvert designs. Rating these 100 culvert designs provided
assurance the Culvert Rating Guide could be used for the full population of TxDOT’s
culverts, and not just a few select cases.
Second, a parametric study was performed to evaluate six independent variables
associated with culvert load rating. The parametric study explored the influence of each
parameter on the inventory rating of the culvert for a set of seven culvert designs. This
helped determine the sensitivity of the culvert rating process to the different variables.
Third, the research team instrumented and load tested three in-service culverts in the
field. This work facilitated a comparison of measured demand moments to predicted values
obtained through analytical modeling as per the Culvert Rating Guide.
Taken together, these three research tasks provided validation of the Culvert Rating
Guide through breadth of application, sensitivity of expression, depth of correlation of the
culvert rating practices and procedures relative to the full population of TxDOT’s reinforced
concrete box culverts.

TxDOT 0-5849 6
 6. ORGANIZATION OF THIS RESEARCH REPORT

This report is organized into seven chapters. Chapter 1 is this introduction which presents
the research problem and the context for this study.
Chapter 2 presents a discussion of the literature on culvert load rating. This includes
academic literature about various culvert rating issues, as well as results from a survey of
culvert rating practices and procedures for the 50 State DOTs and the 25 TxDOT Districts.
Chapter 3 describes the development of the Culvert Rating Guide. This includes a brief
summary of the culvert load rating process, a detailed discussion of several load rating
considerations addressed during development of the Culvert Rating Guide, and a chapter-by-
chapter commentary on the Culvert Rating Guide.
Chapter 4 represents of shift from development of the Culvert Rating Guide to its
validation. This chapter presents results from load rating a statistically-representative sample
of 100 TxDOT culvert designs. This includes a discussion of how TxDOT’s culvert designs
are stratified by design era, the operational research problem statement and hypotheses which
define three different levels of analytical modeling, and the load rating results by era – the
1938 era, the 1946 era, the 1958 era, and the 2003 era. This chapter also includes a
discussion of the results and a summary of conclusions for this research task.
The parametric study of six independent variables associated with culvert load rating is
the topic of Chapter 5. The variables include modulus of subgrade reaction, Poisson’s ratio,
multibarrel effects, lateral earth pressures, soil modulus of elasticity, and depth of fill. The
summary of conclusions for this chapter points out how culvert load rating is not very
sensitive to the first four of these variables, but culvert load rating is sensitive to the last two.
Chapter 6 presents the findings from instrumented load tests on three in-service culverts.
This includes a discussion of the load test design, geotechnical studies associated with the
load testing, and presentation of the instrumented load test results by culvert site.
The report closes (Chapter 7) with a summary of key research findings and with
recommendations both for additional research and for implementation.
This report includes three appendices. Appendix A through E presents data from load
rating the 100 culverts in Chapter 4. Appendix F presents the geotechnical boring logs
obtained from each culvert site in Chapter 6. Appendix G presents the falling weight
deflectometer data obtained from each culvert site. Appendix H presents the maximum and
minimum strain data collected during live load testing.

TxDOT 0-5849 7
TxDOT 0-5849 8
2. CULVERT LOAD RATING LITERATURE REVIEW

1. OVERVIEW

This chapter briefly summarizes conceptual information, loading issues, analytical

modeling tools, and policy guidance associated with culvert load rating which are discussed
in the academic literature. Also included is a summary of culvert load rating practices and
procedures as determined based on interviews with representatives from the 50 State DOTs
and the 25 TxDOT Districts. This information informed our evaluation of TxDOT’s culvert
load rating practices and procedures and provided insight about modeling and analysis issues
encountered during development of the Culvert Rating Guide.

2. ACADEMIC LITERATURE

1. Box Culvert Definitions

In order to better explain the findings from the academic literature survey, it is important
to define several types of box culverts. Box culverts are differentiated by the type of
installation. According to the literature there are three popular reinforced concrete box
culvert installation procedures.
The first and most often modeled in early finite element analyses is the embankment
culvert. These culverts are installed by placing the culvert on existing or built-up soil and
then burying them beneath back fill. Figure 2.1 shows and embankment culvert. It is
important to point out that the surrounding soil mass, even if well consolidated, is less stiff
than the combined culvert and soil column portion. Therefore, the surrounding material will
tend to settle more than the soil directly above the culvert.

backfill layers

culvert

existing ground
FIGURE 2.1. EMBANKMENT CULVERT

The second and most often constructed culvert is installed in a trench. Figure 2.2
illustrates the trench installation. In this situation the backfilled soil will tend to be less stiff
than the surrounding in-situ soil and will experience more settlement than the in-situ soil.

TxDOT 0-5849 9
 backfill layers

existing ground

culvert

FIGURE 2.2. TRENCH CULVERT

The third and experimental method is called the imperfect trench method culvert. Figure
2.3 shows this culvert type. This culvert is installed by first placing the culvert on the in-
place or built-up soils and backfilling around the culvert. Then a layer of compressible
material such as polystyrene, straw or compressive soil is placed directly above the culvert.
Then the rest of the material is backfilled to final grade and compacted. This installation
causes the column of culvert, compressible material and backfill to be less stiff than the
surrounding backfill. The corresponding relative settlement is similar to the trench culvert
and shares the same load reduction characteristic (Kim & Yoo, 2005). These will be
discussed in greater detail in the second section of this literature review.

backfill layers

comp. material

culvert

existing ground
FIGURE 2.3. IMPERFECT TRENCH CULVERT

TxDOT 0-5849 10
 2. Soil Loads

Two primary actions determine the magnitude and distribution of soil loads on a culvert.
The first is called soil arching, and the second is culvert deformation. In this discussion, it is
important to keep the indeterminate nature of the soil structure interaction problem in mind.
Most commonly-used design programs take into consideration the indeterminacy of the
frame structure of box culverts; this is why matrix analysis methods are used so commonly to
determine moments, thrust and shear. However, beyond just the structural indeterminacy, the
loads are indeterminate in nature. Because soil has some strength on its own, it can carry a
portion of its own weight or even some of the surrounding soil weight. Elastic theory has
difficulty accounting for the non-linear, unpredictable nature of this behavior. This explains
why there is so much interest in applying finite element methods to solve the soil-structure
interaction problems.
One indeterminate effect on soil load is soil arching. Soil arching occurs when
differential settlement occurs in soil. As one section of soil settles more than a neighboring
section, shear stresses develop to resist the settlement. The application of soil arching to
culverts is primarily dependent on the type of culvert installation.
In embankment installation culverts, soil arching creates a negative arching effect. As
discussed in Box Culvert Definitions, the combined column of culvert and soil is stiffer than
the surrounding soil. As the surrounding soil settles more than the soil above the culvert,
shear planes develop along the interface. These shear forces transfer some of the
neighboring soil weight onto the culvert. The net result is that the structure is required to
carry the weight of the soil column as well as some of the surrounding soil weight. Figure
2.4 shows this effect. As the soil continues to settle over time the load will continue to
increase. Some studies suggest that the increased load may be as much as twice the weight
of the in-situ soil column. (Tadros, 1986; Yang, 1997; Yang, 1999)

backfill layers

shear forces

culvert

existing ground
FIGURE 2.4. NEGATIVE SOIL ARCHING

In trench culverts and imperfect trench culverts, positive arching occurs. Here the culvert
and soil column are less stiff and experience greater settlement than the surrounding soil.
Therefore the shear stress and load changes are in the opposite direction. The resulting load
reductions can be less than half the weight of the soil column. Figure 5 shows this effect.
(Dasgupta & Sengupta, 1991; Dasgupta & Sengupta, 1991; Vaslestad, Johansen, & Holm,
1993)

TxDOT 0-5849 11
 backfill layers
shear forces

existing ground

culvert

FIGURE 2.5. POSTIVE SOIL ARCHING

Some attempt has been made to take the increase in load into consideration for
embankment installations. The earliest AASHO specification included a highly empirical
equation for determining the increase in load (Gardner & Jeyapalan, 1982). These
formulations were later developed and refined into soil-structure interaction factors by
Spangler (Bennett, 2005). The Portland Cement Association developed an empirical design
chart to determine the loads as well (PCA, 1975). Later AASHTO codes formalized another
soil-structure interaction factor, Fe. However, field test research suggests this soil-structure
interaction factor still underestimates the effective weight of structures (Bennett, 2005; Kim
& Yoo, 2005; Tadros & Benak, 1989; Yang, 1999).
In trenched installations, the outlook is a little better. Research suggests that AASHTO,
though still unconservative, may more closely match test results. Bennett suggests that the
AASHTO specification, though unconservative, still meets the reliability demands of the new
LRFD code (Bennett, 2005). In imperfect trench installations, though still fairly
experimental and difficult to design, the actual loads may be far less than those predicted by
AASHTO (Vaslestad, Johansen, & Holm, 1993).
The way the culvert deflects also affects the amount of load on the structure. This adds
yet another level of indeterminacy to the soil-culvert system. One place that culvert
deflections affect load is at the center of the top and bottom slabs. As load causes the culvert
to deflect downward in the center of the span, the soil begins to transfer the load away from
the center of the span to the outside of the culvert. This results in a decreased load in the
center of the span and an increased load at the supports. Such culvert deformation helps
reduce the moment in the top slab. The same deformation and moment reduction occurs in
the bottom slab. Several papers indicated that the actual pressure distribution is parabolic
instead of uniform (Dasgupta & Sengupta, 1991; Katona & Vittes, 1982). At least one author
indicated that the load redistribution might continue due to creep (Oswald, 1996). See Figure
2.6.

TxDOT 0-5849 12
 soil pressure distribution

culvert boundary

FIGURE 2.6. SOIL PRESSURE DISTRIBUTIONS DUE TO CULVERT DEFORMATION(DASGUPTA & SENGUPTA, 1991)

Deflection due to lateral loads may also affect loads elsewhere on the structure. Lateral
deflection tends to cause the box culvert to deflect in an opposite manner and negate the
deflections from the vertical pressure (see Figure 2.7). This effect causes another decrease in
moment in the top and bottom slab (Awwad, 2000).

FIGURE 2.7. CULVERT DEFORMATION DUE TO LATERAL LOAD ONLY

This review of soil loads suggests several conclusions related to culvert load rating. First,
in the case of embankment culverts, AASHTO specifications are already unconservative. It
is doubtful that any analysis will allow for reduction of excessive overconservatism in load
ratings for existing culverts designed by this method. For trench installed culverts, more
accurate modeling of positive soil arching, culvert deflections, and creep considerations may
allow for reduction of excessive overconservatism in load ratings. The analytical model used
for predicting these effects must be capable of dealing with all the indeterminacies of the
soil-structure system.

TxDOT 0-5849 13
 3. Vehicle Loads

This section explores the live load conditions and modeling that affect culvert load
ratings. The earliest research in vehicle loads on culverts was done by Spangler, Manson and
Winfrey in 1926. Their study indicated that the Boussinesq stress distribution for point loads
represents the maximum load distributed to culverts from vehicle loads. Spangler finds it
very interesting that Boussinesq equations for continuous materials model granular
substances so well. Their testing also suggested that impact loads increased the loads by
50% to 100%.
AASHTO codes simplify the distribution of soil loads to the culvert by expanding the
contact dimensions by 1.75 times the depth of fill for fill depths ranging from 2 ft to 8 ft.
Below 2 ft, the culvert is treated as a direct traffic culvert with no load distribution. Above 8
ft, live load may be ignored. Several papers indicate that this approximation is safe and
reasonable (Abdel-Karim, Tadros, & Benak, 1990; Tadros & Benak, 1989). The AASHTO
LRFD design is far more conservative. It allows a spread of only 1.15H in good soil cases
and 1H in all others. It also specifies an impact factor which varies with fill depth and ranges
from 0.33 at 2 ft to 0 at 8ft (Rund & McGrath, 2000).
More recent studies indicate that the Boussinesq stress distribution, two dimensional
finite element analysis, three-dimensional finite element analysis and field testing all
correlate reasonably well. Seed and Raines provide an equivalent line load equation to
determine the axle load for a two dimensional finite element analysis (Seed & Raines, 1988).
Most of the studies approve of AASHTO’s square area distribution as conservatively
reflecting the load. They also agree that beyond 10 ft of fill the truck load becomes
negligible compared to earth pressure loads (Abdel-Karim, Tadros, & Benak, 1990; Awwad,
2000; Tadros & Benak, 1989). One study suggested including the distributive effect of the
road bed. In the case of flexible pavements (asphalt paving) the suggestion is just to treat the
pavement as additional fill depth. For rigid pavement structures (concrete) the load can be
distributed through the pavement according to Boussinesq’s equations. Another option
would be to develop an equivalent depth for rigid pavements (Abdel-Karim, Tadros, &
Benak, 1990).
Possible methods for reduction of excessive overconservatism in load ratings include
more accurate modeling of distribution of the applied loads through finite element analysis or
Boussinesq’s equations, and by considering the load-distribution effects of pavement
stiffness.

TxDOT 0-5849 14
 3. CULVERT ANALYSIS TOOLS

1. Frame Models

Several modeling programs are available to analyze culverts. The simplest of these are
two dimensional frame models. Two dimensional frame models have many advantages.
They are simple to construct with often fewer than a dozen nodes; some even construct the
model automatically from a few culvert geometry properties. Their structural stiffness
matrices are smaller and therefore require less computation time and introduce fewer errors.
They can deal with the behavior of reinforced concrete by using beam elements with
transformed moments of inertia. The beam elements themselves are built around a proven
and well understood mechanics of materials model.
Several companies, Departments of Transportation, and academic institutions have
developed design charts and computer programs built upon the AASHTO specification for
the simplified design of culverts. All of these programs use the AASHTO specifications and
some form of two dimensional frame analysis. The programs included the Portland Cement
Association (PCA) design manual (PCA, 1975), the American Society for Testing and
Materials (ASTM) C789, C850 and C1433 specifications (Garg, Abolmaali, & Fernandez,
2007), BOXCAR by the Federal Highway Administration (FHWA), SPIDA from the
American Concrete Pipe Association (ACPA) (Tadros, 1986). Several Departments of
Transportation have developed their own design programs including Virginia (Latona, Heger,
& Bealey, 1973), Alabama (Lakmazaheri & Edwards, 1996) and Texas (TxDOT, 2003).
Again none of these programs consider soil-structure interaction. Basically these programs
make the structural design of culverts simpler and less time consuming. These programs are
not sophisticated or precise enough for the refinement of the current load rating procedures.
Other programs are general structural programs. The most popular in common use are
the RISA-2D and RISA-3D programs. Programs like RISA are built upon the same basic
modeling principles as the culvert-specific programs listed above. The advantage to general
structural programs over culvert-specific programs is their ability to model intermediate
boundary conditions using springs. This allows the bottom slab to be supported more
realistically. The slight disadvantage to a general structural program is the need to determine
loads by hand and create the model manually.
Relative to TxDOT’s approach to load-rating culverts, TxDOT engineers like to use their
in-house culvert-specific program, CULV-5, for load rating culverts. CULV-5 is simple and
easy to use. It takes culvert geometry inputs and produces design moments, shears, and
thrusts. These can then be used to design appropriate culvert sections with adequate capacity.
CULV-5 incorporates, at some level, all the appropriate code loading requirements including
the live load distribution, and it is a terrific design tool. However, CULV-5 does have one
significant drawback. It is generally overly conservative in how it applies live load. It
determines the live load demands by developing influence lines for the structure. The
influence lines are actually determined by not only applying the moving unit load to the top
slab, but also by applying one-twentieth loads at twentieth points across the bottom slab.
This results in overly conservative results, particularly in the bottom slab (TxDOT, 2003).

TxDOT 0-5849 15
 2. Finite Element Models

Culvert load rating literature indicates that the finite element analysis (FEA) method
offers superior capabilities for predicting culvert and soil-culvert behavior. Finite element
codes allow for “modeling phenomena not described by the culvert specific codes” and for
graphical investigations of the results (Duane, Robinson, & Moore, 1986). The most popular
soil models can be integrated in the FEA code. Such models include linear elastic models,
elasto-plastic with Mohr-Coulomb failure, soil hardening with stress dependent stiffness and
Mohr-Coulomb failure, Hardin, Duncan, and bilinear. Duncan is the most popular (Kim &
Yoo, 2005; Kitane & McGrath, 2006). Though it is clear that FEA is the analytical tool of
choice for analyzing culvert structures, the particular implementation of FEA must be
determined.
The number of general and soil-specific FEA programs available is impressive. Culvert
analysis research has used SSTIPN (Gardner & Jeyapalan, 1982; Duane, Robinson, &
Moore, 1986; Sharma & Hardcastle, 1993), REA (Roschke & Davis, 1986), STUDL
(Frederick, 1988), FINLIN (Duane, Robinson, & Moore, 1986; Sharma & Hardcastle, 1993),
NLSSIP (Sharma & Hardcastle, 1993), SUPERB (Duane, Robinson, & Moore, 1986),
MARC (Duane, Robinson, & Moore, 1986), ISBILD (Kim & Yoo, 2005), Plaxis (Kitane &
McGrath, 2006), ABAQUS (Yang, 1997; Kim & Yoo, 2005; Kitane & McGrath, 2006) and
CANDE (Katona & McGrath, 2007; Katona & Smith, 1976; Katona & Vittes, 1982; Katona
M. G., 1976; Katona M. G., 1979)(Oswald, 1996; Kim & Yoo, 2005). Of all these, two FEA
programs stand out from the rest.
ABAQUS has the distinct advantage of three dimensional modeling. It incorporates a
general-purpose FEA code, but has been successfully programmed to incorporate soil
models. ABAQUS’ well developed graphical interface makes the program easy to use and
the results easy to interpret.
However, Culvert ANalysis and DEsign, or CANDE, appears to be the de facto standard
for soil-culvert interaction analysis. It was commissioned by the FHWA and developed by
Katona in 1976. Even from its earliest punch-card/FORTRAN versions, CANDE included a
great deal of sophistication including three soil models, a crack conscious concrete model,
and time dependent construction sequences. It has also been adapted by researchers to
consider concrete creep and shrinkage and interface slippage. CANDE’s primary advantage
is the amount of validation that has been done. Over the last 30 years test data has been
compared to CANDE’s predictions and has showed error of around 10% for service loads
and less than 1% for ultimate loads (Katona 1976). That degree of accuracy is within the
normal tolerances for structural design and far better than expected for geotechnical
engineering. The development of CANDE-2007 to allow for graphical output has only made
the program more user-friendly.

TxDOT 0-5849 16
 4. AASHTO POLICY GUIDANCE

The development of the American Association of State Highway and Transportation

Officials’ (AASHTO) provisions for culvert design actually begins with the American
Association of State Highway Officials’ (AASHO) Standard Specifications for Highway
Bridges (AASHO, 1949). In Article 3.2.2, AASHO defines the unit weight of “compacted
sand, earth, gravel or ballast” as 120 pcf. It then proceeds to tell the designer exactly how
this earth load is to be applied to culverts.
Per AASHO, earth pressures or load on culverts could be computed ordinarily as the
weight of earth directly above the slab. In order to have the effect of increasing the allowable
design dead load stresses 40 per cent more than allowed for live load, AASHO allowed the
effective weight of earth backfill to be taken as 70 per cent of its actual weight (AASHO,
1949). This provision is the source of the culvert load rating problem currently experienced
by TxDOT.
The very next paragraph in the AASHO code instructs the design engineer to use the
principles of soil mechanics when designing rigid culverts. It then provides recommended
equations for two classes of culvert installation: trench and embankment.
The only further code direction generally applicable to reinforced culvert design is the
definition of the allowable stress. For concrete the allowable stress is one third of the
compressive strength (AASHO, 1949).
In 1977, the 70% provision disappeared from the AASHTO Bridge Specification (Kim &
Yoo, 2005). In 1983, the code still used the 120 pcf value for the vertical unit weight of soil,
and it defined the effective horizontal unit weight as 30 pcf. The specific equations for
determining the load were no longer provided though it was generally expected that vertical
soil loads would be calculated based on 100% of the soil unit weight.
The 1983 AASHTO code also prescribed a method for distributing the live load from an
HS truck load through the cover soil. It instructed the design to increase the dimensions of
the tire prints by 1.75 H for covers of more than 2 ft of fill. When these areas overlap, the
total load should be redistributed over the area defined by the outside boundaries. This
increase in footprint reduced the live load pressure dramatically with depth. The code
indicates that the live load can be ignored for cover depths of more than 8 ft. The live load
assumptions are the consistent throughout latter revisions of the code.
In 1987, AASHTO released an interim report that updated the 1983 code in two ways.
First, it increased the horizontal soil unit weight in some cases from 30 pcf to 60 pcf. The
1987 code is unclear about when these values should be used. The second update was the
addition of a crack control stress limit.
In 1990, the AASHTO code received another update for reinforced concrete box culverts.
This consisted of a revised equation for calculating the earth pressure on the structure which
included a soil-structure interaction factor for both embankment installations and trench
installations.

TxDOT 0-5849 17
 The inclusion of the soil-structure interaction factors was the first and only coded
application of soil-structure interaction in reinforced concrete box culverts. This method for
determining the soil pressure is used throughout the rest of the AASHTO Standard Bridge
Specifications. Later editions increased the complexity of the crack control provision only.
AASHTO specifications suggest that for culverts with less than 2 ft of fill, the soil does
very little to distribute the load and that the culvert should be designed as a direct slab.
Several researchers have expressed concern about the inconsistencies that this assumption
creates. Many authors have indicated that the AASHTO provisions greatly underestimate
actual soil pressures (Abdel-Karim, Tadros, & Benak, 1990; Tadros, Benak, & Gilliland,
1989; Yang, 1999).

5. STATE DEPARTMENT OF TRANSPORTATION SURVEY

1. Introduction

This research study included a survey of State Departments of Transportation (DOTs) in

order to broaden our understanding of the research problem and to gain information about
how other State DOTs address culvert load rating issues.
One primary focus of the State DOT survey was to identify other transportation agencies
that were having problems similar to TxDOT relative to the load rating and evaluation of
their existing concrete box culverts. The survey also explored how these agencies addressed
this problem and what procedures they used to load rate their culverts.
Forty-nine out of the 50 State DOTs, excluding Texas, were contacted by phone. We
obtained data from Texas directly. Of the 50 State DOTs, 39 DOTs responded via phone, in
person, or by email, which is a 78% response rate.

2. State DOT Contacts

Contact information for each State DOT was obtained from the directory on each State
DOT website. The target contact was generally found to reside within the bridge design or
the bridge maintenance/inspection division. Once this person was contacted via phone they
were given the choice as to whether they completed the questionnaire over the phone or
through email.
Table 2.1 provides the response summary for all 50 states. It also identifies the type of
response received including those states where no response was obtained. Totals show the
number of states in each response category and the percentage relative to the total number of
states contacted.

TxDOT 0-5849 18
 TABLE 2.1. STATE DOT RESPONSE

State DOT No Contact Telephone Interview Emailed Survey Completed Survey

Alabama  
Alaska  
Arizona  
Arkansas  
California  
Colorado  
Connecticut  
Delaware  
Florida  
Georgia  
Hawaii  
Idaho 
Illinois  
Indiana  
Iowa  
Kansas  
Kentucky 
Louisiana  
Maine 
Maryland 
Massachusetts 
Michigan  
Minnesota  
Mississippi 
Wyoming  
Missouri  
Montana  
Nebraska  
Nevada  
New Hampshire  
New Jersey  
New Mexico  
New York  
North Carolina  
North Dakota  
Ohio  
Oklahoma 
Oregon  
Pennsylvania 
Rhode Island 
South Carolina  
South Dakota  
Tennessee  
Texas  
Utah  
Vermont 
Virginia  
Washington  
West Virginia  
Wisconsin 
50 TOTAL (100%) 6 (12%) 3 (6%) 41 (82%) 39 (78%)

TxDOT 0-5849 19
 The formal questionnaire provided a general description of the research project including
background information describing TxDOT’s experiences relative to load-rating their in-
service culverts. This provided a frame of reference for the respondent so that s/he could
relate their answers to be as helpful as possible in our search for information. The survey
inquired about State DOT culvert rating procedures, techniques and software used. If their
culvert rating procedure accounted for soil-structure interaction, the survey requested specific
information about this.

3. Interview Summaries

This section summarizes the results of the email/phone interviews for each of the State
DOT that responded to the research inquiry.
1. Alabama

 Uses HS20 loading

 Applies loads according to AASHTO specs.
 Does account for soil structure interaction
 Conducts lab testing for variable soil conditions
 Software is BRASS Culvert and GTSTRUDL
 They do not have a similar problem because their culverts are evaluated based on
the process under which they were designed
 Have sponsored research for soil interaction for high fill culverts
 When the circumstance arises that neither software program provides an adequate
response, they physically go to the culvert and attach strain gages to the culvert
and run known load vehicles over the culvert to verify the load that the culvert
can support.
2. Alaska

 Uses HS20 loading

 Load rates according to AASHTO specs 17th Edition, and applies dead loads as
required per section 6.2 and live loads as required per section 6.4
 They makes the assumption that the AASHTO methods account for soil-structure
interaction
 No accounting for varying soil conditions
 Software used is RISA 3D
 Does not have a similar problem
3. Arizona

 Does not load rate or evaluate their existing culverts at this time
4. Arkansas

 Uses HS20 loading

 Load rates according to AASHTO specs, and applies dead loads as required per
section 6.2 and live loads as required per section 6.4

TxDOT 0-5849 20
  Accounts for soil-structure interaction in section 16.6.4.2 in AASHTO specs
 Software used is BRASS Culvert Version 2.2
 Does not have a similar problem
5. California

 Uses HS20-44 loading

 The loads are applied to the culvert through an equivalent fluid weight system,
with the application varying based on depth of fill over culvert
 Soil-structure interaction is accounted for in the equivalent fluid weight system
 Concrete box culverts are designed as “rigid” structures based on the equivalent
fluid method above and therefore varying soil conditions are assumed to have no
influence
 LRFD design is implemented by the CANDE-2007, SAP2000, and CTBC
software
 Does not have a similar problem
 Does have a culvert evaluation procedure for culverts > 20 ft span
 Sponsored research includes: NCHRP 15-28: CANDE-2007
 NCHRP 15-29: Live load distribution thru soil
6. Colorado

 Does not load rate or evaluate their existing culverts at this time
7. Connecticut

 Uses HS20 loading

 If the culvert is in good condition a judgment rating is applied, but in special
instances they do analysis following the procedure outlined in AASHTO.
 Does not account for soil-structure interaction
 Does not account for varying soil conditions
 The culvert is modeled using STAAD or analyzed with PennDOT’s Box 5
program.
 They do not evaluate existing culverts for pavement rehab. If the culvert is
functioning properly they apply a judgment rating.
8. Delaware

 Uses HS20 and Delaware legal loading

 Use Section 6 of AASHTO specs to apply loads
 Does not account for soil-structure interaction
 Does not account for varying soil conditions
 BRASS Culvert software is used for analysis
 They do have a similar problem but in relation to culverts with less than 2 feet of
fill. They are currently working with the University of Delaware to load test these
structures in order to determine a better live load distribution.
 They do not have a formal procedure, but plan to use the modified live load
distribution factors and adjust the BRASS analysis results

TxDOT 0-5849 21
 9. Florida

 LFR (HS-20) prior to 2005; LRFR 2005 and after

 Applies loads according to AASHTO specs
 LRFD program includes a beam on elastic foundation analysis to model the
bottom slab soil interaction
 Does not account for varying soil conditions
 Generally use a LFD program, initially written by the North Carolina DOT, a
LRFD Mathcad program, written in-house, and BARS
 Does not have a similar problem
10. Georgia

 Does not load rate existing culverts at this time

11. Hawaii

 Uses HS20 loading

 The application of the loading is that the full projected dead load of the soil over
the culvert is conservatively used and the live load is applied per AASHTO specs
 Does not account for soil-structure interaction
 Does not account for varying soil conditions
 SAP2000 or BRASS culvert software is used for analysis
 Does not have a similar problem
 They do state that their design procedure is conservative and that the
consideration of soil-structure interaction would reduce the demand capacity,
especially for deeper buried culverts
12. Illinois

 Uses HS20 loading, Load Factor Rating Method as described in the AASHTO
Manual for Condition Evaluation of Bridges (MCEB), Sections 6.5 &6.6.3
 Load application is based on section 6 of AASHTO specs
 Does not account for soil-structure interaction
 Does not account for varying soil conditions
 They have an in-house program that calculates the capacity of culverts and also
determine the load effects per section 6 of AASHTO specs
 They do not have a similar problem. Typically, culverts are rated and replaced
based on its observed condition, but have recently noticed an issue of strength
with structures that have been in service 50-60 years
 They note that in MCEB section 7.4.1 there is an allowed condition for the
determined capacity for older concrete structures
 They said that if the structure has been working under normal conditions for an
appreciable length of time and shows no distress then no rating should be
required, but frequent inspections are suggested

TxDOT 0-5849 22
 13. Indiana

 Uses HS20 loading, for the few culverts they have that can be load rated
 In past years when they did rate their culverts they would add the soil weight as a
dead load and if the height of the fill was greater than 8 ft they would ignore the
live load, if fill was more than 2 ft but less than 8ft they would apply the live load
according the AASHTO specs at the current time
 They use a software called Virtis but it does not work on culverts, they also have
the BRASS program but have not used it thus far
 Does not account for soil-structure interaction
 Does not account for varying soil conditions
 They do not have a similar problem
 They rate their existing culverts based on levels of observed stress and provide a
procedure for the regular biennial inspection to follow
14. Iowa

 Uses HS20 loading

 They only apply the soil as a vertical pressure and there is no consideration of a
laterally applied load. The live load is applied as described is AASHTO section 6
 Does not account for soil-structure interaction
 Does not account for varying soil conditions
 They do not consider the frame of action in the design; they only use in-house
programs to analyze the loading of the culvert. They only calculate simple span or
continuous span reactions without regard to end restraint. They assume the deck
slab controls the rating.
 Provided a culvert design criteria
 Provided a culvert rating example
15. Kansas

 They use H, Type 3, HS20, 3S2, 3-3 loading

 Loads are distributed according to AASHTO specs
 Does not account for soil-structure interaction
 Does not account for varying soil conditions
 They use BRASS Girder LFD for RCB’s and BRASS Culvert for RFB’s software
 They do not have a similar problem, when culverts are functioning well
engineering judgment is used in the replacement decision
 They feel that pavement 6 inches or more in thickness is giving better live load
distribution than AASHTO is allowing but they have not researched this yet
16. Louisiana

 Does not load rate existing culverts at this time

17. Michigan

 Rates for Michigan legal loads

TxDOT 0-5849 23
  For the load application, the soil weight is treated as a design pressure, and the
live load is distributed according to AASHTO spec 6.4
 Does not account for soil-structure interaction
 Does not account for varying soil conditions
 They use BARS or hand calculation as analytical procedures
 They do not have a similar problem
18. Minnesota

 Uses HS20 loading

 Loads are distributed according to AASHTO specs
 They do account for soil-structure interaction, but they allow AASHTO
exceptions for older culverts
 They do not directly account for varying soil conditions but state that their load
combinations account for this indirectly
 They use BRASS Culvert, STAAD, CANDE, and BOXCAR as analysis software
 They do have a similar problem and based on their experience with older culverts
they applied the AASHTO provisions. If the physical inspection shows no
distress then the culvert is not replaced.
 They are currently looking into developing new guidance to rating older culverts,
in which they are considering the procedure mentioned in the AASHTO
provisions.
19. Missouri

 They only perform load rating analysis when there is less than 2 ft of fill. When
this occurs they load rate the culvert for the vehicles shown the vehicle load sheet.
 Loads are distributed and soil-structure interaction is based on the Load Factor
Design Bridge Manual, Section 3.2 Box Culverts
 Does not account for varying soil conditions, they feel that they have used
suitable fill material that will adequately transfer the load to the top to the box
culvert
 They generally use the BRASS Culvert software
 They do not have a similar problem. They use a visual inspection looking for
signs of distress based on existing loading conditions and will use the existing
culvert to widen/extend if feasible
20. Montana

 Uses HS25, but this is a recent development

 For loading they use DL: unit weight x load factor , and
 LL: fill height x 1.75 x load factor, they then state that this is what they will use
until something comes along specifically for load rating
 They do account for soil-structure interaction by 1.15 max (soil interaction factor
in the design code)
 They do not account for varying soil conditions, they use 60pcf max and 30 pcf
min for lateral earth pressures
 They use BRASS Culvert Version 2.2.5

TxDOT 0-5849 24
  They do not have a similar problem, but appear to not replace culverts based on
loading, but on hydraulic conditions
21. Nebraska

 Uses HS20 loading

 For loading purposes the soil pressure is uniformly distributed under all load cases
 Does not account for soil-structure interaction
 Does not account for varying soil conditions
 They use FRAME ACTION for analysis procedures
 They do not have a similar problem
22. Nevada

 They use HS20 and P13 truck

 They use AASHTO specs from 1977 with modification and the reference the
attached plan sheet
 For soil-structure interaction and varying soil conditions they reference AASHTO
and/or BRASS Culvert
 They use BRASS Culvert for design and reference design example
 They do not have a similar problem
23. New Hampshire

 They officially use HS20

 Loads are distributed according to AASHTO specs, usually 120 pcf DL and
distributing through a pyramid where the base is 1.75 times the depth of the fill
 Their procedure accounts for soil-structure interaction, but not well, they have
some computer programs that claim to consider soil-structure interaction but they
are not always used
 Does not account for varying soil conditions
 They do not use a program for analysis purposes. All calculations are done by
hand, or if precast, the calculations are done by the manufacturer
 They do not have a similar problem. They will leave a culvert in place if it has
enough capacity to carry highway loads and still maintains the required flow.
They note that this may be different from the design load
 They have a system where they look at the longitudinal effective span length and
compare capacity to the demand for their legal loads.
24. New Jersey

 They use HS20 AASHTO Type 3 & 3-3, and NJDOT Type 3S2 (40T)
 Loads are distributed according to AASHTO specs
 Does not account for soil-structure interaction
 Does not account for varying soil conditions
 They use PennDOT’s program for analysis purposes

TxDOT 0-5849 25
  They do not have a similar problem because they do not require existing culverts
to meet new design criteria. They do not replace existing culverts that are
functioning well and are in good condition.
25. New Mexico

 Uses HS20 loading

 Loads are applied through their concrete box culverts standard drawing specs for
2007, which uses LRFD code specs
 They do account for soil-structure interaction. A new standard drawing was
designed for the first series with 0-10ft of cover so the influence of the live load
on the design is diminished
 Does not account for varying soil conditions
 They do not often re-rate their existing culverts
 They do not have a similar problem. They only time they re-rate a culvert is
when addition fill is needed. Then they compare the ultimate strength vs. their
original design in working stress.
26. New York

 Uses HS20 and HS25 loading, with the standard Military Load option
 The dead load on the top slab consists of soil weight plus the weight of the
concrete slab. The program they use is capable of analyzing additional uniform
dead load as well as accepting three extra concentrated dead loads. When the
culvert has traffic running directly on the top slab, wheel loads are distributed as
in ordinary bridges. This is also done when the height of fill on the culvert is less
than 2 feet. The program will distribute wheel loads over a slab width, E, equal to
4 + 0.06S, where S is the perpendicular distance between wheel centerlines.
When the culvert is skewed relative to the over roadway, they magnify the
intensity of the live load by reducing the distribution width. In no case shall the
distribution width exceed 7 feet nor the section length of precast units. When the
height of fill is greater than or equal to 2 feet, wheel loads are distributed over
areas having sides equal to 1.75 times the depth of fill. When these areas overlap
the wheel loads are evenly distributed over the gross area, but the total width of
distribution shall not exceed the total width of the supporting slab. Their program
considers two, three and four adjacent vehicle lanes, as appropriate, and selects
the critical case. Appropriate AASHTO lane reduction percentages are used for
the three and four lane loading cases.
 Their program is in agreement with AASHTO Articles 17.6.4.2 and 17.7.4.2,
Modification of Earth Loads for Soil Structure Interaction, of the Standard
Specifications, for embankment installations only. The soil-structure interaction
factor, Fe , is not applicable if the Service Load Design Method is used.
 A soil unit weight is selected, 120 pcf being the default, and is used for the entire
height of the soil envelope. Two cases of lateral wall earth pressure are
investigated, 60 pcf maximum and 30 pcf minimum.
 The program they use was obtained in 1995 from the North Carolina Department
of Transportation and modified for New York State use by NYSDOT staff and the

TxDOT 0-5849 26
 Precast Concrete Association of New York (PCANY). The program’s method of
analysis is the stiffness method.
 It is assumed that loading applied to the top slab will be uniformly distributed
over the whole bottom slab. They feel this is a reasonable assumption since a mat
of granular material is placed beneath the bottom slab of the culvert, and there is
usually a lack of precise soil information for each site.
 Typically if the culvert is in good condition, and functioning from a hydraulic
standpoint, they would not necessarily replace an existing culvert that does not
meet current design requirements.
27. North Carolina

 Does not load rate existing culverts at this time

 The wheel loads are distributed over squares having sides equal to 1.75 times the
depth of the fill. When the squares overlap, the wheel loads are spread evenly
over the gross area
 Does not account for soil-structure interaction
 Does not account for varying soil conditions
 They use an in-house program to perform a frame analysis on the “strip” of the
culvert, using the stiffness method. The analysis determines the maximum load
effects and completes a standard concrete design
 They do not have a similar problem
28. North Dakota

 Uses HS20 loading

 To load the culverts they follow section 6 Culverts in the 2002 Interim Revisions
to the Standard Specs for Highways and Bridges. Dead loads per section 6.2 and
live loads per section 6.4
 Does not account for soil-structure interaction
 Does not account for varying soil conditions
 They use an in-house program that determines the required reinforcing for a
particular loading
 They do not have a similar problem, but do suggest the use of the Manual for
Condition and Evaluation of Bridges as the method they use to evaluate concrete
slabs that have little known information
29. Ohio

 They load rate existing RC Box Culverts for AASHTO HS20 and four legal Ohio
loads
 They use BRASS Culvert which is based on AASHTO specs
 This process does account for soil-structure interaction
 Does not account for varying soil conditions
 They do not have a similar problem
 They have sponsored some research on concrete arches and also some research in
Corrugated Metal Pipe Culverts.

TxDOT 0-5849 27
  They note that several short span culverts designed using ASTM tables do not rate
very well using the BRASS Culvert program
30. Oregon

 Uses HL-93 Truck/Tandem LL, and states that the truck can have up to two axles
on the culvert at one time
 Unit weights of in-place, compacted backfill beside and over box culverts are in
accordance with AASHTO LRFD Table 3.5.1-1 (120pcf typically assumed). A
design depth of live load surcharge (HLS) of 2ft is used in accordance with
AASHTO LRFD. Load factors for horizontal and vertical earth pressures are in
accordance with AASHTO LRFD Table 3.4.1-1 & 2: 0.90min to 1.35max for
horizontal earth pressure, at-rest; and 0.90min to 1.30max for vertical earth
pressures, rigid buried structures. If the backfill depth over the top of the box
culvert (H) is less than 2ft, AASHTO LRFD Eq. 4.6.2.10.2-1 is used – otherwise
the width is determined as the maximum of Eq. 4.6.2.10.2-1 or from the
provisions in AASHTO LRFD 3.6.1.2.6 as illustrated in Figure 3. The 1.15 factor
is not applied to the depth term in Eq. 3.6.1.2.6. In the direction parallel to the
span, the axle loads are modeled as point loads. The allowable increase in load
length is neglected as allowed in LRFD C4.6.2.10.2.
 They do not directly account for soil-structure interaction. ODOT used the GT-
Strudl finite element structural analysis software to determine force effects. The
soil response to box culvert loading is represented by a “soft spring” (vertical
coefficient of subgrade reaction).
 They do not directly account for varying soil conditions; however, ODOT is
looking closely at the box culvert differential settlement countermeasures being
developed by Florida DOT
 ODOT used the GT-Strudl finite element structural analysis software
 They do not have a similar problem
 They are currently re-doing the calculations for all their standard drawings. The
new calculations and updated drawings will be in LRFD. They did a comparison
between their old standard drawings that were developed using LFD and
determined the designs shown on the drawings only needed minor modifications
to meet the new design standard.
 ODOT is in the process of updating existing standard drawings for
gravity/cantilever retaining walls and box culverts from ASD\LFD to AASHTO
LRFD Specifications.
31. South Carolina

 They evaluate the culverts under the criteria in which it was designed
 They apply the loads according to AASHTO Standard Specs for Highways and
Bridges or AASHTO LRFD Bridge Design Specs.
 They do account for soil-structure interaction
 They consider existing information for the soil conditions, and for widening they
conduct a subsurface investigation to design the foundation.

TxDOT 0-5849 28
  They use BRASS Culvert version 2.2.6 for culverts that are cast-in-place and
BOXCAR version 2.03 for precast culverts
 They do not have a similar problem
 Culverts of bridge length 20ft. or greater are checked on a regular basis, but
culverts less than 20ft. are checked by maintenance offices before the culverts are
extended.
32. South Dakota

 Uses HS25 loading for new design but do not currently rate their box culverts
 For design purposed loads are distributed according to AASHTO specs
 Does not account for soil-structure interaction
 Does not account for varying soil conditions
 They have their own software, South Dakota Box, which uses ASD, but are now
transitioning to BRASS Culvert which uses LRFD
 They do not have a similar problem
33. Tennessee

 Does not load rate existing culverts at this time

 Notes a quote from the MCEB, “A concrete bridge need not be posted for
restricted loading when it has been carrying normal traffic for an appreciable
amount of time.”
34. Utah

 Depending on what design standard was used, is what controls the loading
condition they use for the rating. For ASD they use HS-20 and for LRFD they
use HL-93
 They apply the dead load as the unit weight of the soil times the thickness and
applied uniformly, and the live load thru the fill effects are taken with a slope of
1:2
 Does not account for soil-structure interaction
 Does not account for varying soil conditions
 They use the PennDOT’s software which uses the direct stiffness method
 They replace culverts based on existing condition rather than load rating
35. Virginia

 Uses HS25 loading

 Loads are distributed according to AASHTO specs
 Does not account for soil-structure interaction
 They do not account for varying soil conditions when loading, but they do
consider the variability when designing the foundation to support the culvert
 They previously used an in-house program which was a modification of the old
North Carolina box program, but have now adopted the BRASS Culvert software

TxDOT 0-5849 29
  They do not have a similar problem. Culverts that are functioning well do not get
replaced unless the new required fill height is too much of a load for the old
culvert to handle
 They typically see that culverts are replaced due to other criteria besides loading;
such as relocation or hydraulic deficiencies
36. Washington

 Uses HS20 loading, three legal and two overload trucks (Types 1,2 & 3, OL-1 &
OL-2 found in Bridge Design Manual 13.1.1.G)
 They apply the soil as a uniform surcharge and reference their calculations, Sheet
1 for distribution factor or truck lane multiplier.
 They do account for soil structure interaction and earth pressure, which is
significant in the Br. 169/30C example.
 Does not account for varying soil conditions
 For analysis they use Concrete Bridge, a program owned by WSDOT
 They do not have a similar problem
 They seldom have plans for culverts, but when they do they rely on inspection
results and perform administrative ratings
37. West Virginia

 Does not load rate existing culverts at this time

 They do not replace culverts that are functioning well and are in good condition
38. Wyoming

 They use Type 3, HS20, 3S2, 3-3 loading

 Loads are distributed according to AASHTO specs
 Does not account for soil-structure interaction
 Does not account for varying soil conditions
 They use the BRASS Culvert software
 They do have a similar problem, they address it in various ways, including
replacing the culvert if the grade is to be changed or just extend the culvert if the
grade is not being changed.

Additional information is presented in the detailed Logs of Activity (documents) archived

in the project research file. Information regarding the State DOT respondents and their
contact information is also included in these documents. Reference materials that were
provided by any of the State DOTs are also included in these documents.

TxDOT 0-5849 30
 4. Results from the State DOT Survey

Table 2.2 shows the different types of loads used to rate the culverts. More load types
exist than there are responding states that rate culverts, because most responses included
more than one type of loading. Six of the responding states do not load rate their culverts.
TABLE 2.2. LOAD TYPES

HS20 HS25 HS15 STATE LEGAL OTHER

23 4 1 7 7

Table 2.3 shows the how each of the responding states that evaluate culverts apply the
load to the concrete box culvert. Out of these 32 loading scenarios, only 15 incorporate soil-
structure interaction and only 7 consider the effects of varying soil conditions.
TABLE 2.3. LOAD APPLICATION

AASHTO CUSTOM OTHER

21 9 2

The following Table 2.4 illustrates the breakdown of analytical software and methods
used by the responding states to calculate and analyze the effects of the load application
directly on the concrete box culvert.
TABLE 2.4. ANALYTICAL PROGRAMS

BRASS 14 BOXCAR 2
GTSTRUDL 2 BARS 2
RISA 3D 1 FRAME-ACTION 1
CANDE 2 PENNDOT 2
SAP2000 2 IN-HOUSE 8
CTBC 1 HAND CALC. 3
STAAD 1

The interviews revealed that only three of the responding states had a problem similar to
that which TxDOT is experiencing, and only four states have sponsored research that could
be considered relevant to this subject.

TxDOT 0-5849 31
 5. Conclusions

In conclusion, only three states (Delaware, Illinois, and Minnesota) have culvert load
rating problems that can be considered similar to that which TxDOT is currently
experiencing. Delaware’s problem is in relation to culverts with less than 2 feet of fill and
they are currently working with the University of Delaware to research possible solutions.
Illinois has recently noticed a strength issue with structures that have been in service for 50-
60 years. They are looking to provisions in the Manual for Condition Evaluation of Bridges
for a possible solution. Minnesota has developed some exceptions to AASHTO
specifications to apply when rating older culverts. They are also looking into developing
new guidance for rating older culverts.
It should be noted that the responses do not directly correlate to what is actually being
done in the load rating procedure. For example, two states that claimed to use AASHTO
specifications to apply the load gave different answers to the questions about whether their
procedure accounted for soil-structure interaction and varying soil conditions. This suggests
that there is definite confusion as to what the AASHTO specifications require and what
conditions are accounted for.
It seems that most of the responding states do not replace their culverts on the basis of
load rating, but on the basis of hydraulic functionality. These states all seem to go by the
adage, “If it’s not broke, don’t fix it.” The justification of this adage comes from a quote in
the MCEB, referenced several times by the responding states, “A concrete bridge need not be
posted for restricted loading when it has been carrying normal traffic for an appreciable
amount of time and shows no distress.”

TxDOT 0-5849 32
 6. TXDOT DISTRICT SURVEY

1. Purpose

The intent of this survey was to obtain responses from each of the 25 TxDOT District
Bridge Engineers regarding how they approach the issue of load rating concrete box culverts.
The survey questions focused on who in the district is doing the load ratings, how they are
doing them, and the reasons that bring about the need to load rate the concrete box culverts.

2. District Notes

This section outlines the responses provided by the Bridge Engineer within the TxDOT
districts.
Abilene: Left messages, no response.
Amarillo: Relies on the Bridge Division in Austin or the consultants that do the bridge
inspection (SDW and Associates) for culverts that need to be rated. No load rating is done in
house. The load rating is deemed necessary after the failing of a visual inspection and when
the posted operating load is less that HS 20.
Atlanta: Relies on the Bridge Division in Austin or the consultants that do the bridge
inspection (Klotz and Associates, SDW and Associates) for culverts that need to be rated.
No load rating is done in house. Consultants seldom load rate culverts because when they do
and want to change to operating rating of a culvert TxDOT will not let them without doing an
extensive investigation before-hand, so the load rating is typically not pursued any further. If
the load rating is necessary it is performed after the failure of a visual inspection.
Austin: Left messages, no response.
Beaumont: Left messages, no response.
Brownwood: Left messages, no response.
Bryan: Relies on the Bridge Division in Austin for culverts that need to be rated. No
load rating is done in house. They send culvert rating request to Austin when deteriorating
conditions warrant, after failure of a visual inspection, and before widening, but these
situations do not occur frequently. They usually replace culverts due to hydraulic capacity
demands or deteriorating structural conditions. They state that the difference of construction
cost is minimal between replacing a culvert versus lengthening a culvert.
Childress: Relies on the Bridge Division in Austin or the consultants that do the bridge
inspection (JPH Consulting, Inc., Edwards and Kelsey, and Structural Diagnostics Inc.) for
culverts that need to be rated. No load rating is done in house. The consultants conduct a
load rating when the culvert scores a 4 or below on a 0-9 scale during a visual inspection and
then make a recommendation to Austin that the posted load should be changed. They only
provide the consultants with the guidelines from Austin and the structure’s history file.
Corpus Christi: Left messages, no response.

TxDOT 0-5849 33
 Dallas: Culverts are load-rated very rarely, practically never, in this office. Any
culvert design or analysis is done using CULV5. If culvert problems are encountered, they
contact the Bridge Division.
El Paso: No one in this office does load ratings or has any knowledge of how or when
they would be done. They believe that if load rating is being done it is being done by Austin.
Fort Worth: If a culvert fails a visual inspection, this information is passed on to
Austin so the culvert can be load rated.
Houston: They do not do load rating in house, but the consultant (Structural Diagnostics
Inc.) that does the bi-annual bridge inspections does load rate the culverts that fail the visual
inspection. The Houston District provides the consultants with the construction details of the
structure and the structure’s history file.
Laredo: They just started a bridge division in this district and have not load rated any
structures as of this time. Anything prior was done by Austin.
Lubbock: Consultants do the load ratings after a culvert fails its visual bi-annual
inspection, but these inspections only include bridge class culverts, which are culverts that
span twenty feet or more.
Lufkin: Left messages, no response.
Odessa: They rely on Austin for their load ratings. If a culvert fails a visual inspection
from the bi-annual bridge inspection conducted by the TxDOT-approved consultants, the
consultants will load rate the culvert in question. For culverts that are on roads that are to be
widened they send it to Austin to be rated.
Paris: Relies on the consultants that do the bi-annual bridge inspections. For off-
system structures they use Klotz and Associates and Clear Span Engineering. For on-system
structures they use Maverick Engineering Inc. and Edwards and Kelsey. There is no load
rating done in-house, but when it is done by the consultants, it is done after a structure has
failed a visual inspection. They provide the consultants with the structure’s history file and
its construction details.
Pharr: They do not load rate culverts or seem to have consultants load rate culverts.
They use CULV-5 and do a redesign with the new proposed fill height to see if the culvert
can hold the load. If it cannot, they replace the culvert.
San Angelo: Left messages, no response.
San Antonio: Not contacted.
Tyler: They rely on the Austin Bridge Division and consultants (Howell
Engineering) to load rate their culverts. The consultants usually end up rating one culvert
during the bi-annual bridge inspection because it has failed a visual inspection. There is
rarely a need for the district to request that Austin load rate a culvert, outside of the bi-annual
inspection.

TxDOT 0-5849 34
 Waco: They very seldom load rate culverts in-house and it has been a few years since
they have done a load rating. Otherwise the consultants that conduct the bi-annual bridge
inspections do the load rating for bridge class culverts that fail the visual inspection.
Wichita Falls: They do not do any load rating in-house. If load rating is done it is
done by the consultants that conduct the bi-annual bridge inspections, and only after the
structure fails a visual inspection.
Yoakum: They do not rate culverts, but did note that their files are in need of being
updated.

3. Summary

The results of the TxDOT District survey show that nearly all culvert load rating for
TxDOT is done by the Bridge Division in Austin, or by consultants that conduct TxDOT’s
bi-annual bridge inspections. Even when the consultants do the load ratings, the calculations
are still sent to Austin for review to ensure that the proper procedure was used, and for
approval of changes. It is very seldom that load rating is done inside a TxDOT district
office. Since TxDOT provides no specific guidance to the consultants about load rating, the
load rating procedures these consultants are following are unknown. The only culverts that
are inspected are bridge class; that is, culverts that are 20 feet in length or longer. This leaves
a large number of culverts that receive no regular attention.

TxDOT 0-5849 35
 7. CONCLUSIONS

The following methods, approaches and conclusions can be drawn from the academic
literature that addresses the culvert load rating.
1. For embankment installed culverts, it is doubtful that a more accurate analysis will
allow for reduction of excessive overconservatism in load ratings for existing
culverts.
2. For trench and imperfect trench culverts, it is possible that a more refined analysis
will allow for reduction of excessive overconservatism in load ratings while
maintaining an acceptable factor of safety.
3. For culverts which have been in service for 50-plus years, it seems reasonable to
assume that subsurface stresses are stable and it is not clear whether the
embankment/trench dichotomy affects soil stresses.
4. By considering more sophisticated methods of live load distribution, a reasonable
decrease in culvert load can be predicted.
5. By considering the effect of pavement stiffness, the live load effect may be further
reduced.
6. Finite element analysis techniques will be required to solve this problem.
7. CANDE and ABAQUS are among the more powerful finite element programs for this
application.
The State DOT survey indicated that only three states (Delaware, Illinois, and Minnesota)
have culvert load rating problems that can be considered similar to that which TxDOT is
currently experiencing. Most of the responding states do not replace their culverts on the
basis of load rating, but on the basis of hydraulic functionality. These states prominently cite
a quote in the MCEB: “A concrete bridge need not be posted for restricted loading when it
has been carrying normal traffic for an appreciable amount of time and shows no distress.”
Results from the TxDOT District survey show that nearly all culvert load rating for
TxDOT is done by the Bridge Division in Austin, or by consultants that conduct TxDOT’s
bi-annual bridge inspections. It is very seldom that load rating is done inside a TxDOT
district office.
This information was used to inform and address TxDOT’s culvert load rating questions.
Subsequent chapters of this report discuss the development of the Culvert Rating Guide, and
the research by which the Culvert Rating Guide was validated.

TxDOT 0-5849 36
 3. DEVELOPMENT OF THE CULVERT RATING GUIDE

1. OVERVIEW

This chapter describes the development of the TxDOT Culvert Rating Guide. The stated
purpose of the Culvert Rating Guide is “to present a clear, repeatable and valid procedure
for TxDOT engineers and their consultants to use for load rating culverts in the TxDOT
roadway system.”
It is appropriate to begin with a brief summary of the culvert load rating process.
This is followed by detailed discussion of key decisions addressed during development
of the Culvert Rating Guide including reliability considerations, selection of appropriate
analytical models for production load rating of culverts, and articulation of three levels
of analysis; namely, Level 1 (two -dimensional, simply-supported structural frame
model), Level 2 (two dimensional structural frame model with soil springs), and Level 3
(two dimensional finite element soil-structure interaction model). This is followed by a
chapter-by-chapter commentary about the Culvert Rating Guide.

TxDOT 0-5849 37
 2. A BRIEF SUMMARY OF THE CULVERT LOAD RATING PROCESS

The culvert load rating process is one aspect of the inspection process and consists of
determining the safe load-carrying capacity of the culvert structure, determining whether
specific legal or overweight vehicles can safely cross the culvert, and determining if the
culvert needs to be restricted and if so, what level of load posting is required.
Load posting consists of placing signage by the structure indicating the largest truck that
may be permitted across the structure. The flow chart from the TxDOT Bridge Inspection
Manual (Figure 3.1) defines the culvert load posting process (TxDOT 2002). Culverts may
be load posted at the operating rating if the culvert condition rating is greater than that
defined in the flow chart. Otherwise the culvert must be load posted at the inventory rating.
Load posting, then, directly interacts with culvert load rating.
The basic culvert load rating procedure is as follows. Per Equation 1.1, the main variables
are culvert capacity, the dead load demand, and live load demand. Culvert capacity is
established from equations set forth in AASHTO policy, whereas dead load and live load
demands must be determined by structural modeling (computer analyses). While this seems
simple enough, the challenge is to obtain reliable values for each of these variables.
A “road map” of the culvert rating process helps avoid confusion. Figure 3.2 presents the
load-rating process in terms of a flow chart. The first step is to identify the culvert that will
be load rated. As noted, this might be because the culvert failed a scheduled inspection or for
some other reason. Either way, a visual inspection of the culvert is necessary. For all intents
and purposes, the culvert load rating process begins here.
The load rating factor calculations require determination of both culvert capacity and
dead and live load demands. It is helpful, therefore, to think of culvert capacity and demand
calculations as separate and distinct aspects of the load rating process.

TxDOT 0-5849 38
 FIGURE 3.1. LOAD POSTING GUIDELINES (TXDOT BRIDGE INSPECTION MANUAL FIG.5-3).

TxDOT 0-5849 39
 FIGURE 3.2. FLOW CHART DEPICTING THE TXDOT CULVERT LOAD RATING PROCESS.

TxDOT 0-5849 40
 Capacity calculations are based on equations established in AASHTO policy. These do
not require a computer model and are independent of the level of analysis selected for
demand calculations. Inputs for capacity calculations are obtained from the construction
drawings, visual inspection, and AASHTO policy and consist of strength properties for
concrete and steel, culvert section dimensions, and the location and amount of reinforcing
steel. The calculations determine moment, shear and thrust capacity for each critical section
of the culvert structure.
Determination of dead and live load demands do require computer modeling. Thus the
first decision to be made is to select the type of analytical model for the load rating process.
The Culvert Rating Guide describes three levels of analysis, each with increasing analytical
sophistication. A trade-off exists between sophistication of analysis and required work effort.
The advanced models require more work but typically yield more accurate results.
Once the level of analysis is chosen, it is necessary to gather data to facilitate creation of
the analytical model. Modeling parameters include but are not limited to culvert dimensions,
properties of the concrete and reinforcing steel, soil parameters, the location and amount of
reinforcing steel, and culvert installation details.
With this information, the load rater can create the analytical (computer) model from
which s/he will obtain demand moments, shears and thrusts. This involves laying out the
model, specifying boundary conditions, identifying critical sections, applying loads, and
defining load cases.
Determining the inventory and operating load rating factors requires multiple sets of
calculations from the computer model. This is because demand loads and their corresponding
capacity must be determined for each critical section, for each failure mode, and for multiple
load cases. From these sets of calculations, the load rater selects the controlling (minimum)
operating and inventory rating factor for each critical section, for each load case. The
minimum operating and inventory rating factors from the critical sections are the rating
factors for the culvert.
A decision is required at this point. If the inventory and operating rating factors are
greater than 1.0, the culvert will not require load posting. It is unrestricted. This means that
the culvert load rating can be calculated by multiplying the rating factors by the tractor
tonnage (for HS-20 trucks, this is 20 tons) to determine the operating (OR) and inventory
(IR) load ratings. However, if either the inventory rating factor or the operating rating factor
is less than 1.0, the culvert may require load posting. As an alternative to posting, the load
rater may elect to perform the calculations again, using a higher level (more sophisticated)
modeling approach.
This is the basic load rating procedure articulated in the Culvert Rating Guide.
Development of this procedure required that several key design issues be addressed, and
the next sections of this report discuss those decisions.

TxDOT 0-5849 41
 3. PRIMARY DEVELOPMENT GOAL: RELIABILITY

Reliability is one of the most important qualities of any sort of predictive engineering
analysis and was therefore one of the primary concerns in developing the Culvert Rating
Guide. For this discussion, reliability means the coherence between the predicted outcome
and the measured event. The predicted value may be above or below the measured value, but
the closer the predicted value is to the measured value, the more reliable the prediction
method is. Several factors affect reliability for culvert load rating contexts. These include
repeatability, capacity calculation approach, model input variables and demand model
sophistication.

1. Repeatability

Repeatability of the modeling procedure has an immediate and direct impact on

reliability. If several engineers load rate a structure in the same way and produce several
different load ratings, very little confidence should be placed in the results. There is only one
real load rating. The most reliable procedure will identify only this load rating. Ideally, a
procedure would be clearly articulated to remove inconsistencies, identify unique levels of
reliability and do so repeatedly. A repeatable load rating procedure should, for a given
structure, produce a single inventory rating and a single operating rating. Engineers should
be able to, within reason, agree upon and reproduce those ratings for a given culvert.
Without a repeatable load rating procedure, the load rating for a structure holds no real or
certain meaning. The decision to develop the Culvert Rating Guide derived in part from a
lack of repeatability in TxDOT’s current state of practice.
2. Capacity Calculation Approach

The capacity calculations also affect reliability. The capacity of a section is determined
by considering the concrete and steel properties and using these to calculate the moment,
shear and thrust capacity. Each of these capacities can be determined several ways. Ultimate
moment capacity can be calculated by models such as Whitney’s stress block, the modified
Hognestad model, the Kupfer and Gerstle model or the Kent and Park model. Shear can be
determined using the ACI/AASHTO semi-empirical method, modified compression field
theory or truss methods. Axial load can be considered uniaxial or combined with moment
and shear. The Culvert Rating Guide must provide capacity calculation guidance that is
widely accepted and generally understood by the majority of structural engineers, and that is
considered sufficiently reliable for load rating analysis.

3. Model Input Variables

Another factor that affects reliability is the input variables for the various models; that is,
the geometric, material, and loading parameters used to calculate the load rating. Model
input variables have perhaps the strongest effect on reliability. These variables can be
obtained several different ways. They may be determined from construction plans, manuals,
and published references. They can be determined by correlated tests. They can also be
determined by field investigations and lab tests. While it is recognized that the more case-
specific methods increase the reliability of the resulting answer, the Culvert Rating Guide

TxDOT 0-5849 42
must provide guidance for obtaining input parameters, either from published sources or
through site specific data-gathering approaches.

4. Model Sophistication

It has been noted that the reliability of a culvert load rating (i.e. the load rating value) also
depends on the degree of sophistication of the demand model. The load rating method,
capacity determination, and input variables are often defined by policy. However, model
sophistication is rarely controlled by policy. Instead, this is generally left to the discretion of
the engineer. As a general statement, the assumptions, simplifications and mathematical
structures of various demand modeling tools can have a significant effect on modeling
reliability for culvert load rating analysis. Therefore, in order for the Culvert Rating Guide to
be reliable, the issue of model sophistication must be bounded and directed. Model
sophistication reliability effects primarily occur in two classes: soil-structure interaction and
live load distribution. These will be discussed as they are resolved in the Culvert Rating
Guide.
1. Soil-Structure Interaction

For culvert applications, soil-structure interaction is expressed primarily in terms of soil

arching. Soil arching, in turn, can be both global – which has to do with culvert installation
method – and local – which has to do with soil response to culvert deflection. Because this
research project is primarily focused on the load rating of in-service culverts, the effects of
global soil arching are assumed to have dissipated. This assumption would not necessary be
valid for newly-constructed culverts, however. Local soil arching does affect culvert load
ratings.
Soil-structure interaction, by way of soil arching, is a very real phenomenon that must be
considered in order to accurately predict culvert demands. Models which consider the full
effect of soil arching should produce the most reliable demands. However, such modeling
can be time-consuming and labor intensive. Because of this, many models incorporate
assumptions to simplify the determination of demands. To say it another way, simplifying
assumptions are used to more easily calculate less reliable demands that are, ostensibly,
higher than actual values, which in turn results in lower or more conservative load ratings.
This is a form of over-conservatism associated with model sophistication. These
simplifications will be discussed at length in the following section on analytical models and
programs.
2. Live Load Distributions

The other area where model sophistication affects reliability is live load distribution.
Vehicles produce complicated load paths. Typically the loads pass from the vehicle axle to
the tire, through a patch of tire and into the pavement. In the pavement, the load is
distributed by the pavement stiffness and then passes as a diffused pressure into the bedding
and supporting soil. In the supporting soil, the pressure is further diffused throughout the soil
mass. If a culvert happens to reside in the soil mass, the load is further distributed into the
culvert top slab and walls and back into the soil mass through the bottom slab. To further

TxDOT 0-5849 43
complicate matters, vehicles produce dynamic pulse loading which behaves differently than
static and quasi-static loadings. Attempting to model this complex behavior is very difficult.
Much research has been done to explore how live loads from vehicles affect buried
structures. Section 2.2.3 of this report discusses the previous research in detail.
Modeling sophistication strongly affects the overall reliability of demand calculations and
load ratings. To the extent that less sophisticated models produce excess conservatism, an
increase in the model sophistication is anticipated to reduce over-conservatism. As the
sophistication of the models increases, some of the excess conservatism can be removed to
produce more reliable, that is, higher, load ratings.
Though the operational hypothesis of this present research study assumes a correlation
between modeling sophistication and reliability, it must be noted that model sophistication
does not guarantee load rating reliability. Without a repeatable load rating procedure and
representative input parameters for demand and capacity models, the load rating will not be
reliable. In fact, with non-representative input parameters and an unreliable procedure, the
resulting load rating could be higher than reality, unconservative and possibly unsafe. This is
the reason for the Culvert Rating Guide.
5. Reliability Further Discussed

For culvert load ratings, reliability can be considered as a function of three key concepts:
predicted behavior, accepted conservatism and load-soil-structure interaction.
1. Predicted Behavior

In an ideal load-rating procedure, the actual behavior of the soil-structure system would
be perfectly predicted. For an engineer whose primary concern is accurate predictions (not
safety), this is the only objective. Therefore, the engineer would use application-specific soil
properties, highly-directed modeling techniques and capacity calculations designed to
accurately predict soil-structure behavior as it happens in the real world. This is a basic
requirement for a reliable load rating.
2. Accepted Conservatism

The second aspect of reliability is accepted conservatism which acknowledges

uncertainties in the design and construction process. Engineers are concerned about safety as
well as accuracy; therefore, the prediction of actual soil-structure behavior may be
intentionally adjusted to ensure safe predictions. That is, the material properties, capacity
calculations and modeling techniques may be modified or simplified to ensure conservative
predictions. The load factors in the load rating equations are primarily concerned with
predicting behavior in the worst case scenario, not the actual real world behavior. The
capacity equations have an accepted and intentional amount of conservatism. Even the
concrete and steel material properties are assumed to be weaker than reality. Accepted
engineering conservatism for soil properties is less defined, but also exists. The net effect is
that an accepted degree of conservatism will exist in the culvert load rating process.

TxDOT 0-5849 44
 3. Load-Soil-Structure Interaction

The last element of reliability affecting culvert load ratings is load-soil-structure

interaction. The most basic models include over-conservatism to acknowledge their lack of
refinement relative to the load-soil-structure interaction phenomenon. More sophisticated
models consider load-soil-structure interaction. For this reason, more sophisticated models
should produce higher load ratings as the over-conservatism associated with load-soil-
structure interaction is removed.
The philosophy used in the Culvert Rating Guide is built upon this assumption. When
the different analytical models are discussed, no net change is assumed to occur in the
balance of actual behavior and accepted conservatism. Rather, the investigated concept is
load-soil-structure interaction. Modeling sophistication is what allows the influence of load-
soil-structure interaction (excess capacity) to be identified and the associated over-
conservatism to be removed. A corollary assumption is that the load ratings produced by
more sophisticated models will increase by reducing over-conservatism but will not produce
load ratings above those which exist in reality.

TxDOT 0-5849 45
 4. SELECTION OF ANALYTICAL MODELS

1. Analytical Models and Programs for Determining Demand

Many models are available to determine live and dead load demands for culvert load
rating problems. Determining which of the models to use can be a daunting and difficult
task. Discussing the types of available models and discussing their similarities and
differences will help identify those models which are most appropriate for culvert load rating.
Table 3.1 shows the nine available model types and their distinguishing characteristics.
TABLE 3.1. SUMMARY OF MODELS
GENERAL (GEN) / TWO DIMENSIONS (2D) / THREE LINEAR-ELASTIC (LE) / STRUCTURAL (STRUC) /
MODEL #
CULVERT (CULV) DIMENSIONS (3D) NON-LINEAR (NL) SOIL-STRUCTURAL (S-S)
1 CULV 2D LE STRUC
2 GEN 2D LE STRUC
3 GEN 2D NL STRUC
4 GEN 2D LE S-S
5 GEN 2D NL S-S
6 GEN 3D LE STRUC
7 GEN 3D NL STRUC
8 GEN 3D LE S-S
9 GEN 3D NL S-S

TxDOT 0-5849 46
 1. Prepackaged, Two-Dimensional, Culvert Models

The simplest models are two-dimensional frame models which rely on static loading
(loads are balanced between top and bottom slabs) and frame analysis matrix methods. Two-
dimensional frame models have many advantages. They are simple to construct with often
fewer than a dozen nodes; some programs even construct the model automatically from a few
culvert geometry properties. Their structural stiffness matrices are smaller and therefore
require less computation time and introduce fewer rounding errors. The beam elements
themselves are built around a proven and well understood mechanics of materials model.
Basically, these models take the most conservative interpretation of AASHTO policy
requirements and apply them to a simply-supported frame model. Figure 3.3 illustrates a half
space model where member self-weight is automatically determined. Loads include vertical
dead and live load, lateral (or horizontal) dead and live load and balanced upward dead and
live load forces on the base of the culvert including self-weight.

LL vT

DL v

DL hT

Y
X

LLh DL hB

SW

DL v

LL vB
FIGURE 3.3. PREPACKAGED, TWO-DIMENSIONAL CULVERT MODELS LOADING DIAGRAM. (TXDOT, 2009).

This sort of model tends to produce very conservative demands. Moment, particularly in
the bottom slab, tends to be overly conservative. For design purposes this is very acceptable;
culverts are designed to be stouter than they need to be. For load rating purposes, however,
such over-conservatism is less desirable, particularly if sufficient conservatism exist
elsewhere. What this sort of model does very poorly is account for “real world” behavior
such as soil arching and the effects of differential settlement in the foundation. This model is
the least sophisticated model and therefore should produce the most conservative load rating.

TxDOT 0-5849 47
 2. General, Two-Dimensional, Linear-Elastic, Structural Models

The second model type is the general, two-dimensional, linear-elastic structural model.
This model is very similar to the first: it uses AASHTO prescribed loads, it is based on
matrix methods, and the structural elements are modeled as beams. The difference is that
this model is designed for general structural purposes, so the engineer has greater control
over model generation. The real improvement is that intermediate spring supports may be
used to support the culvert instead of balanced upward forces. Figure 3.4 illustrates the
loading diagram for this model.

LL vT

DL v

DL hT

Y
X

SM= model span (in.)

scorn = SM/20 (in.)

sint = SM /10 (in.)

LLh DL hB

FIGURE 3.4. GENERAL, TWO-DIMENSIONAL, LINEAR-ELASTIC STRUCTURAL FRAME MODELS LOADING DIAGRAM.(TXDOT, 2009)

This means that instead of applying loads to the bottom of the culvert, springs are spaced
uniformly across the bottom slab to provide support. Though the model still does not model
soil-structure interactions, it does account for the effect of differential settlement in the
foundation and allows for more natural distributions of the live load across the bottom slab.
The result, ostensibly, is improved reliability in the bottom slab demands.

TxDOT 0-5849 48
 3. General, Two-Dimensional, Non-Linear, Structural Models

The general, two-dimensional, non-linear, structural model is another improvement upon

the linear-elastic model. The key change is that the material models for the culvert (steel,
concrete, aluminum, etc.) have been improved to consider non-linear behavior. For example,
as reinforced concrete deforms under load, it goes through periods of decreased stiffness as
the concrete cracks and varying stiffness as the reinforcing steel begins to yield. The result is
a non-linear stress-strain curve.
Models with these capabilities are popular for dynamic, fatigue or post-ultimate loading
such as earthquake or dynamic wind design. Culverts however are generally not prone to
extremely dynamic or fatigue loading, and any post-ultimate behavior in a culvert would
result in very poor visual inspection performance. So while this is a more sophisticated
model, reinforced box culvert behavior does not typically venture into the ranges where this
increase in sophistication provides more reliability. This model will not be discussed any
further.

TxDOT 0-5849 49
 4. General, Two-Dimensional, Linear-Elastic, Soil-Structural Models

The fourth model to consider is the two-dimensional, linear-elastic, soil-structural model.

This is the most basic of the finite-element models. In general, finite-element programs
allow for “modeling phenomena not described by the culvert specific codes” and for
graphical investigations of the results (Katona M. G., CANDE-a Modern Approach for the
Structural Design and Analysis of Buried Culverts, 1976). This particular finite-element
model builds upon the concepts of the two-dimensional, linear-elastic structural frame model.
However, instead of applying AASHTO loads directly to the structural elements, the
surrounding soil mass is modeled using linear-elastic finite-elements. Figure 3.5 shows the
load condition for this model.
LL vT(P=16k) LL vT(P=16k)

LL vT(P=4k)

14' 14' to 30'

FIGURE 3.5. GENERAL, TWO-DIMENSIONAL, LINEAR-ELASTIC SOIL-STRUCTURE INTERACTION MODELS LOADING

DIAGRAM.(TXDOT, 2009)

This model allows the culvert and soil self-weight to be automatically distributed through
body forces and the live load to be distributed automatically in one plane. One limitation is
that this model allows for soil tension and shear at large displacements. However, it does
model soil-structure interaction and differential settlement in the foundation. For the
increased complexity associated with moving from a frame model to a finite-element model,
reliability is gained by accounting for soil-structure interaction.

TxDOT 0-5849 50
 5. General, Two-Dimensional, Non-Linear, Soil-Structural Models

The two-dimensional, non-linear, soil-structure model is an improvement over the linear-

elastic finite-element model. Again, only the material models are improved. Like the non-
linear structural model, the culvert material (concrete, steel, etc.) can be modeled non-
linearly. The big difference is that the soil can also be modeled using non-linear models.
Such models include elasto-plastic with Mohr-Coulomb failure, soil hardening with stress
dependent stiffness and Mohr-Coulomb failure, Hardin, Duncan, and bilinear, of which
Duncan is the most popular (Katona M. G., CANDE-a Modern Approach for the Structural
Design and Analysis of Buried Culverts, 1976). Improved soil models can have a very large
effect on the reliability of the culvert load rating.
In the linear-elastic soil-structural model, a single modulus of elasticity and Poisson’s
ratio is used throughout the soil mass. At the very best, the modulus may be stratified with
depth. In the non-linear soil-structure model, the soil stiffness can be increased not only with
depth but also varied with the stress history. The result is a more reliable picture of the soil
behavior. For the increase in sophistication, namely a shift from linear-elastic to non-linear
soil models, the reliability for culvert load rating should improve.
6. General, Three-Dimensional, Linear-elastic, Structural Models

Sophistication of models increases as these models move from the two-dimensional

realm, where the culvert slabs are modeled as beam-like slices, to the three-dimensional
realm, where the culvert slabs are modeled as plates. By nature, the three-dimensional
models are finite-element models. The simplest of the three-dimensional models is,
unsurprisingly, the linear-elastic structural model. This model uses linear-elastic finite plate
elements to model the culvert, springs to model foundation support and AASHTO equations
to directly load the culvert. It is basically a three-dimensional expansion of the two-
dimensional, linear-elastic, structural model.
Though there is a significant increase in sophistication, this model is still based on
AASHTO loading guidelines. Therefore all the conservatism inherent in the AASHTO
assumptions remains. The only real improvement is the way slabs carry loads differently
from beams. The one situation where this makes a difference is in direct contact culverts. In
this case the model would allow for consideration of some edge effects. But for deeper fill
culverts this improvement is lost in the AASHTO assumptions.
Second, the mechanics of plate behavior are far less understood than beam behavior,
especially for the typical structural engineer. Therefore, moving into a plate model rather
than a beam model introduces uncertainties, both real and perceived, that decrease the
suitability of this model.
Third, finite-element plate models require convergence studies to ensure that the proper
mesh has been applied to predict the desired value. This highlights a situation where an
increase in model sophistication can actually result in a decrease in confidence and
reliability. For this reason, this model is not well suited for the purposes of culvert load
rating.

TxDOT 0-5849 51
 7. General, Three-Dimensional, Non-Linear, Structural Models

The three-dimensional, structural model with non-linear material models is the next step
in modeling sophistication. If the linear-elastic equivalent was considered appropriate for
culvert load rating, the non-linear model would only be a marginal and unutilized
improvement over the linear model. As it is, the three-dimensional, non-linear, structural
model suffers from all the same pitfalls as its linear-elastic cousin. The increase in
sophistication can lead to a decrease in reliability.
8. General, Three-Dimensional, Linear-elastic, Soil-Structural Models

The eighth model is a generalized, three-dimensional, linear-elastic, soil-structural model

which uses plate elements to model the culvert and linear-elastic continuum elements to
model the surrounding soil. This provides excellent soil-structure interaction and three-
dimensional live load distributions. The increased sophistication of this generalized model
can achieve significant increases in reliability over less sophisticated models.
9. General, Three-Dimensional, Non-Linear, Soil-Structural Models

The general, three-dimensional, non-linear, soil-structure model is the most sophisticated

finite-element model available. It improves upon the three-dimensional, linear-elastic, soil-
structure model by upgrading the material models from linear-elastic to non-linear. This
takes into consideration soil-structure interaction, soil and structural behavior as it varies
with stress history, edge effects, three-dimensional live load distribution and slab behavior.
As the most sophisticated model, it should, given the proper inputs, closely predict the
behavior of a load test. It is expected to give the most reliable results.
10. Models Appropriate for Culvert Load Rating

Of the nine classes of models available for soil-structure modeling, Table 3.2 summarizes
the six models that are worthy of further consideration for culvert modeling and load rating.
Note that models three, six and seven have been excluded because they do not yield an
appreciable increase in reliability despite an increase in sophistication and in the effort
needed to specify and apply the model.
TABLE 3.2. SUMMARY OF CULVERT APPLICABLE MODELS
GENERAL (GEN) / TWO DIMENSIONS (2D) / LINEAR-ELASTIC (LE) / STRUCTURAL (STRUC) /
MODEL #
CULVERT (CULV) THREE DIMENSIONS (3D) NON-LINEAR (NL) SOIL-STRUCTURAL (S-S)
1 CULV 2D LE STRUC
2 GEN 2D LE STRUC
4 GEN 2D LE S-S
5 GEN 2D NL S-S
8 GEN 3D LE S-S
9 GEN 3D NL S-S
The models in Table 3.2 are listed in increasing rank by sophistication and anticipated
reliabilty. In the next section, a value analysis will be made to determine which of these
models are appropriate for production load rating and which lend themselves to research
oriented analysis.

TxDOT 0-5849 52
 2. Value Analysis of Models

Though several types of models are available to load rate culverts; not all models may be
useful to the engineer for production load rating. A production engineer’s primary concern is
achieving the necessary reliability with the least effort. Therefore an exploration of the
balance between reliability and ease of use would be helpful. The categories of evaluation
include reliability, inputs, general use and load rating specific use. In this section, each
model will be qualitatively assessed in each category and then analyzed to determine which
models are appropriate for load rating.
1. Degree of Sophistication and Assumed Reliability

The engineer would prefer the most reliable model available if all other things are equal.
Qualitatively, the three-dimensional, soil-structure models will have the highest
sophistication and highest anticipated reliability (H). Two-dimensional soil-structure models
will have average or medium reliability (M). Two-dimensional structural models will have
lowest anticipated reliability and sophistication (L). Table 3.3 shows the qualitative ratings.
TABLE 3.3. QUALITATIVE EVALUATION OF MODEL SOPHISTICATION AND RELIABILTY
MODEL # GEN / CULV 2D / 3D LE / NL STRUC / S-S QUALTIATIVE SOPHISTICATION / RELIABILTY
1 CULV 2D LE STRUC L
2 GEN 2D LE STRUC L
4 GEN 2D LE S-S M
5 GEN 2D NL S-S M
8 GEN 3D LE S-S H
9 GEN 3D NL S-S H

2. Identification of Input Variables

The identification of input variables is important because it directly influences production

readiness of a particular model. This category will be broken into three sub-categories: the
number of variables needed, the ease or difficulty associated with identifying those variables,
and the required confidence in the variables indentified.
1. Number of Variables

All models use the same sort of structural material properties such as thickness, strength,
modulus of elasticity, reinforcing, etc. Even the variables used in non-linear structural
models are generally well agreed upon in the structural world and thus require no further
mention.
The primary issue is the soil parameters. The least sophisticated models require no soil
parameters while higher models may require as many as a dozen variables associated with
three-dimensional, non-linear soil models. The fewer the number of required variables, the
more desirable the model is. Therefore, qualitatively, models with very few soil parameters
have the highest desirability (H). A model with an average number of soil parameters, such
as in linear-elastic finite-element models, is moderately desirable (M). A model with many
soil variables will be the least desirable (L).

TxDOT 0-5849 53
 2. Ease of Variable Determination

Beyond the number of variables required, the models should also be scored based upon
the ease or difficulty of determining the variables. Reasonable values for some soil variables
can be found in reference books. Other variables require expensive and time-consuming
laboratory or field tests to represent the materials.
Again, this variable can be qualitatively rated in terms of high (H), moderate (M) and low
(L) desirability. Models requiring less expensive and less time consuming preparatory test
are the most desirable. The highest desirability models (H) have variables whose values can
be easily determined from reference materials. Models which require some sort of correlated
in-situ testing are only moderately desirable (M). The least desirable models (L) require
costly and time-consuming field and lab testing to determine variable properties.
3. Required Confidence in Variables

It is also important to evaluate the degree of confidence associated with each variable.
This is really an issue of model sensitivity. Structural models are generally well understood
and, relative to demand calculations, these models are unaffected by changes in individual
variables. This sort of insensitivity to variable inputs, relative to model ease of use, is highly
desirable (H). However, finite-element programs can be very sensitive to various input
parameters, and therefore require greater confidence in those parameters. This means that in
terms of variable confidence, finite-element models are less desirable.
Table 3.4 summarizes the anticipated qualitative performance of the culvert applicable
models in the realm of variable identification.
TABLE 3.4. QUALITATIVE EVALUATION OF MODEL INPUT VARIABLES
GEN / STRUC NUMBER EASE CONFIDENCE VARIABLE
MODEL # 2D / 3D LE / NL
CULV / S-S SCORE SCORE SCORE TOTAL
1 CULV 2D LE STRUC H H H H
2 GEN 2D LE STRUC H H H H
4 GEN 2D LE S-S M M L M
5 GEN 2D NL S-S L L L L
8 GEN 3D LE S-S M L L L
9 GEN 3D NL S-S L L L L

TxDOT 0-5849 54
 3. Model Generation Ease of Use

The next category is the ease with which the model can be built in individual programs.
The most highly desirable models (H) in terms of ease of model generation will undoubtedly
be the culvert specific models. These models take the most basic geometric properties and
automatically construct the model. The moderately desirable models (M) allow for graphical
model generation. The least desirable models (L) require extensive input file writing to build
the model. Model generation in non-graphical models can be extremely difficult and time
consuming.
Table 3.5 shows the ease of model generation for the various models.
TABLE 3.5. QUALITATIVE EVALUATION OF EASE OF MODEL GENERATION
MODEL # GEN / CULV 2D / 3D LE / NL STRUC / S-S MODEL GENERATION
1 CULV 2D LE STRUC H
2 GEN 2D LE STRUC M
4 GEN 2D LE S-S M
5 GEN 2D NL S-S L
8 GEN 3D LE S-S L
9 GEN 3D NL S-S L

4. Load Rating Specific Ease of Use

The final issue for consideration is the ease with which a particular model can be used for
load rating. Load rating issues include the identification of critical sections, separation of
dead and live load and the application of moving live loads.
1. Identification of Critical Sections

When load rating, it is important to be able to accurately determine the demands at the
critical sections. In many cases this is as simple as placing a node at the critical section.
Some models then allow the user to filter the demands to select just those nodes; others
require the user to identify the critical section demands from the output manually.
The highly desirable models (H) allow the user to identify and filter out the critical
section demands automatically. Moderately desirable models (M) allow the user to identify
the critical sections but require the user to isolate the demands manually from the output.
The least desirable models have fixed outputs which require interpolation to identify the
critical section demands.
2. Separation of Dead and Live Loads

Another important step in the load rating process is identifying the dead and live loads
separately. Some models provide totally independent dead and live load runs. These models
are highly desirable (H). Other models require a dead load run, and then a dead plus live
load run. The live load demand is then isolated by subtracting the dead load demand from
the dead plus live load demand. Such models are only moderately desirable (M).
Conceivably, a program could require two totally independent models for determining dead

TxDOT 0-5849 55
and live load demands. In other words, the model would have to be built twice. This sort of
model would be the least desirable (L). However, almost all modern modeling programs
would not require this level of redundancy.
3. Moving Live Load Applications

The last load rating requirement is the ease with which a moving live load envelope
solution can be determined. Some models automatically determine the moving live load
solution by simply inputting a load pattern and path. As the ideal solution, these models are
the most desirable (H). Other models require programming a set of load cases manually for
each live load position and then determining the envelope solution manually. These models
are only moderately desirable. The worst case would be a model which required not only
separate dead and live load models, but also separate live load models for each live load
position. Most programs do not require this much work, but if they did, they would be the
least desirable models (L).
Table 3.6 qualitatively summarizes the load rating specific evaluation.
TABLE 3.6. QUALITATIVE EVALUATION OF MODEL LOAD RATING SUITABILITY
GEN 2D LE STRUC DEAD LOAD
CRITICAL MOVING
MODEL # / / / / VS RATING
SECTIONS LIVE LOAD
CULV 3D NL S-S LIVE TOTAL
1 CULV 2D LE STRUC L H H H
2 GEN 2D LE STRUC H H H H
4 GEN 2D LE S-S H H H H
5 GEN 2D NL S-S M M M M
8 GEN 3D LE S-S M H M M
9 GEN 3D NL S-S M M M M

TxDOT 0-5849 56
 5. Overall Suitability for Production Use

The ideal model for culvert load rating would perfectly represent reality (highest
reliability), would require very little effort to identify the input variables, would
automatically generate the model, and would provide output ideally suited for load rating.
The most highly desirable models come close to this goal. However, engineers interested in
production culvert load rating are often willing to sacrifice some reliability for faster, easier
ways to obtain results. Any model which requires significantly more work to gain marginal
increases in reliability is of very little interest to the production load rater. Such models
might find use in specialized research applications, but are not ideally suited to culvert load
rating on a day-to-day basis. Table 3.7 summarizes the qualitative assessment of the six
models under consideration.
TABLE 3.7. SUMMARY OF QUALITATIVE EVALUATION OF MODELS FOR PRODUCTION LOAD RATING
GEN 2D LE STRUC SOPHISTICATION
MODEL VARIABLE MODEL LOAD
/ / / / / TOTAL
# INDENTIFICATION GENERATION RATING
CULV 3D NL S-S RELIABILTY
1 CULV 2D LE STRUC L H H H H
2 GEN 2D LE STRUC L H M H M
4 GEN 2D LE S-S M M M H M
5 GEN 2D NL S-S M L L M L
8 GEN 3D LE S-S H L L M L
9 GEN 3D NL S-S H L L M L

Table 3.7 shows the more desirable models are two-dimensional and use linear-elastic
material models. Two-dimensional models are easier to use than their three-dimensional
counterparts and require far fewer variables with lower required variable confidence.
In contrast, the higher order models, despite their increase in sophistication and
reliability, are less attractive for production load rating due to their need for more input
variables, more complex model generation, and lower suitability to culvert load rating
applications. Accordingly, the models that are well suited for production culvert load rating
are:
1. Prepackaged, Two-Dimensional, Culvert Models
2. General, Two-Dimensional, Linear-Elastic Structural Models
3. General, Two-Dimensional, Linear-Elastic, Soil-Structural Models
These production-ready models are suitable for culvert load rating as defined by a
measured balance of analytical effort and reliability. These are the models used in the
Culvert Rating Guide.

TxDOT 0-5849 57
 5. DEVELOPMENT OF LEVEL 1 ANALYSIS: TWO-DIMENSIONAL, SIMPLY-
SUPPORTED STRUCTURAL FRAME MODEL

The purpose of this analysis level is to perform a quick and easy demand calculation
using a prepackaged, two-dimensional, culvert specific model. The main reason for this
analysis level is to include TxDOT’s culvert analysis program CULV-5 in the Culvert
Rating Guide.
1. Model Construction

Model dimensions are not needed to construct a model in CULV-5; rather this
information is used to determine the location of the critical sections. Though AASHTO
requires the identification of the worst-case critical midspan section, throughout the
Culvert Rating Guide, the midpoint is used instead. This is a safe assumption due to the
relative flatness of the moment diagram at this location and the already included
conservatism of the production-ready models.

2. Loads

Applied culvert loads are dictated by AASHTO. CULV-5 uses influence lines to
determine the application of live load. The magnitudes of the loads are calculated
automatically.
The dead loads vary linearly with depth. However, the live loads vary non-linearly
with depth. The live load distribution is based on AASHTO Standard Specifications for
Highway Bridges Section 6 for culverts. The basic premise is that the load from a single
wheel should be uniformly applied over an area equal to 1.75*D. When the wheel areas for
multiple wheels overlap, the total load is to be uniformly applied over the encompassed area.
AASHTO specifies multiple presence factors when 3 and 4 or more trucks affect the load.
Equations were derived for 1, 2, 3 and 4 or more trucks. For culverts with less than 2’ of fill,
the load is treated as a direct load with the appropriate impact factor according to an
equivalent line load equation.
The AASHTO load geometries are shown in the following figure.

TxDOT 0-5849 58
 One Truck Load Geometry
P = 16k or 4k P = 16k P = 16k
P = 4k
6' 14' 14'
3.4' D
8.0'

Two Trucks Load Geometry

P = 16k or 4k P = 16k P = 16k
P = 4k
6' 4' 6' 14' 14'
2.3' 3.4' D
8.0'

Three Trucks Load Geometry

P = 16k or 4k P = 16k P = 16k
P = 4k
6' 4' 6' 4' 6' 14' 14'
2.3' 3.4' D
8.0'

Four or More Trucks Load Geometry

P = 16k or 4k P = 16k P = 16k
P = 4k
6' 4' 6' 4' 6' 4' 6' 14' 14'
2.3' 3.4' D
8.0'

FIGURE 3.6. TWO DIMENSIONAL FRAME LOAD GEOMETRY

The live load pressures were developed from the load geometries. The worst case
scenario was determined for each range by plotting the load factor as a function of trucks and
depth. This plot is shown in Figure 3.7 This resulted in the step function outlined in the
Culvert Rating Guide.

load factor
0.000 0.050 0.100 0.150 0.200 0.250 0.300 0.350
0
1
2
depth of fill, D (ft)

3
4
5
6
7
8
1 truck 2 trucks 3 trucks 4 or more trucks

FIGURE 3.7. TWO DIMENSIONAL FRAME ANALYSIS COMPARISIONS OF LIVE LOADS

TxDOT 0-5849 59
 3. Analytical Program – CULV-5

The Culvert Rating Guide provides a step by step procedure for calculating the
demand loads using CULV-5. The strengths and limitations are summarized from both
the CULV-5 documentation and literature review. The step-by-step instructions guide
the user through the process of creating the model and running the program, applying the
load cases and determining demands, to calculating the rating factors and selecting the
final inventory and operating load ratings.

6. DEVELOPMENT OF LEVEL 2 ANALYSIS: TWO DIMENSIONAL STRUCTURAL

FRAME MODEL WITH SOIL SPRINGS

Level 2 represents a slight modification compared to the Level 1 analysis. The

general modeling assumptions, model dimensions, loads and load cases are the same
between the Level 1 and Level 2 models. The main difference is the boundary
conditions. Instead of a simply supported structure with balanced loads (Level 1),
compression soil springs are used to model foundation support (Level 2). This level was
introduced to take advantage of the capabilities of the more general frame models.
CULV-5 is incapable of this level of analysis. Level 2 is the recommended first
modeling step for those not using CULV-5.
RISA-2D is the non-CULV-5 tool of choice for TxDOT. Therefore, though the
modeling instructions are written for any modeling program, specific instructions for
RISA-2D are provided. These instructions include an overview, a summary of strengths
and limitations, and a step-by-step procedure for creating the model and load cases in
RISA-2D. This includes calculating the demands, load rating factors and inventory and
operating load rating factors.

TxDOT 0-5849 60
 7. DEVELOPMENT OF LEVEL 3 ANALYSIS: TWO DIMENSIONAL FINITE ELEMENT SOIL-
STRUCTURE INTERACTION MODEL

The Level 3 analysis is a significant change from the Level 1 and Level 2 analyses.
In this model, AASHTO standard loadings are discarded in favor of a finite element
method to model the behavior of the soil mass. In this way, the self-weight of the soil
applies the dead load and the live load is applied as a point load on the soil surface and
is distributed through the soil. The Guide provides an overview of assumptions and
model dimensions. The boundary conditions include the size of the modeled soil mass.
The size of this soil mass was selected from the guidelines provided by the defacto
standard in two-dimensional culvert finite element modeling program, CANDE (Katona
M. G., CANDE-a Modern Approach for the Structural Design and Analysis of Buried
Culverts, 1976). The boundary conditions are set to model an infinite continuous soil
mass.

1. Live Load Distribution

When using two dimensional finite element models, it is not necessary to distribute the
live load through the soil to a pressure applied to the culvert frame elements as it is in the
frame model. In the finite element case, the live load is applied to the surface of the soil and
the finite elements translate and distribute the load in the in-plane direction by design.
However, it is necessary when using a two dimensional finite element model to distribute the
load in the out-of-plane direction. This can be done in several ways.
The first is related to the AASHTO Standard Specification for Highway Bridges. The
method would be to take a point load and distribute it over a line that is 1.75*D long. Like in
the frame approach overlapping loads would be averaged over the total overlapping length.
The second is described in the CANDE-2007 Solution Methods manual and has been
adopted by AASHTO LRFD. In this case the point load is distributed over a line which is
1.15*D+H long where H is the length of the tire contact patch (20 in.).
The third approach is also described in the in the CANDE-2007 Solution Methods
manual. It is a modified elastic approach. The length of the line is obtained from a
complicated formula. Unlike the other two methods, this method does not take into
consideration the presence of multiple wheel loads.
Like when determining the live load for the frame analysis, it is necessary to derive the
load with respect to depth and number of trucks and then determine the worst case loading
conditions. The derivation of the AASHTO SSHB method uses the same load geometries
used in the Level 1 and 2 analyses shown in Figure 3.6. The AASHTO LRFD method
uses a slightly different load geometry. This is shown in Figure 3.8. The elastic
approach uses the Boussinesq stress distribution. Once the load functions were derived
from the load geometries, the worst case loading was determined for each by plotting the
load factor as a function of trucks and depth.

TxDOT 0-5849 61
 One Truck Load Geometry
P = 4k or 16k

6'
20"
3.8'

Two Trucks Load Geometry

P = 4k or 16k

6' 4' 6'

2.0' 20"
3.8'

Three Trucks Load Geometry

P = 4k or 16k

6' 4' 6' 4' 6'

2.0' 20"
3.8'

Four Trucks Load Geometry

P = 4k or 16k

6' 4' 6' 4' 6' 4' 6'

2.0' 20"
3.8'

FIGURE 3.8. TWO DIMENSIONAL FINITE ELEMENT AASHTO LRFD LIVE LOAD GEOMETRY
Figure 3.9 shows a comparison of load distribution approaches. It is clear that for depths
of less than eight feet, the elastic method is far and away the most conservative. It is also
clear that except for a narrow band from two feet of fill to around three feet of fill, the
AASHTO SSHB approach is the least conservative. Expert opinion and soon-to-be released

TxDOT 0-5849 62
NCHRP reports indicate the AASHTO LRFD approach best represents “reality” with a
comfortable balance between conservatism and accuracy. It is also convenient that the
LRFD approach removes the discontinuity between the code equation of direct traffic loads
(D<2’) and greater fill depths. Therefore, the AASHTO LRFD approach was adopted and
included in the Culvert Rating Guide.

load factor
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35
0
2
4
depth of fill, D (ft)

6
8
10
12
14
16
18
20

AASHTO SSHB AASHTO LRFD Elastic

FIGURE 3.9. TWO-DIMENSIONAL FEA COMPARISONS OF LIVE LOADS

It is also worth noting that for soil-structure FEA models there is only one load case. The
reduced lateral loading case specified by AASHTO only applies to direct stiffness models
and has no meaning in a finite element model.

2. Production Load Rating Considerations

Two common programs are available for the Level 3 analysis. CANDE is the standard
culvert finite element model. It is widely accepted as the culvert modeling tool of choice,
particularly for dead loads. RISA-2D is a general frame analysis program capable of linear
elastic plate modeling.
The advantages of CANDE-2007 are mainly in the inclusion of complex soil models, the
use of improved concrete models that consider the effect of reinforcing steel, and in the
provision of “canned” finite element meshes that facilitate specification of the analytical
model in a user-friendly way. Further, CANDE has been extensively documented and as
noted in Chapter 2, CANDE’s primary advantage is the 30 years of validation that has been
done on the program.
Notwithstanding these impressive benefits, CANDE is not particularly user-friendly for
production culvert load rating. This is primarily due to the asymmetrical nature of the wheel
loads which must be applied for rating purposes, and the fact that these wheel loads must be
manually programmed to “travel” across the culvert cross section in order to establish load
envelopes for the culvert structure. The canned culvert meshes in CANDE cannot be used for
this application. This means that the finite element mesh and wheel loads must be hand-

TxDOT 0-5849 63
generated, a process that is very tedious and error-prone for users who are not familiar with
the CANDE program.
Another challenge to using CANDE for culvert load rating arises from CANDE’s use of a
sophisticated nonlinear reinforced concrete model. This model establishes the structural
capacity of the culvert slabs and walls, and when applied loads exceed the calculated member
capacity, the culvert structure “fails.” Culvert load rating, however, typically applies a
standard load to the culvert per AASHTO policy, that is, an HS-20 truck. If the culvert is
stout enough to support the HS-20 load, a properly-identified CANDE model will determine
the culvert load rating. However, if the culvert does not have adequate capacity to support the
HS-20 load, the culvert structure will fail. While this conclusively establishes one load that
the culvert cannot carry, culvert load rating is about determining the load the structure can
support. To achieve this objective, the load-rater must backtrack and resort to a trial-and-
error process of manually reducing the applied load until the (weak) culvert does not fail.
These challenges for CANDE only exist when the objective is production culvert load
rating. The effort necessary to create the model, generate the load conditions, and achieve a
load rating with CANDE would not be at issue for other culvert analysis and design
applications. But for production load rating of culverts by engineers who are not expert
CANDE users, a more user-friendly approach would be desirable. RISA-2D is such a model.
RISA incorporates a graphical user interface that can readily generate the finite element mesh
for both the culvert structure and surrounding soil. The program also includes a feature that
automatically applies a moving load across the culvert model to facilitate determination of
the load envelope. In addition to ease of input, graphical output features are also excellent.
Notwithstanding the user-friendly nature of the program, the pertinent question becomes,
“Can RISA-2D provide valid results?” Stated another way, would the validity of results
from RISA-2D approach the validity of results from CANDE?
Initial inquiry into this question revealed the following items:

• RISA-2D has very few of the powerful features of CANDE-2007. It can however
model linear-elastic finite elements and it can create meshes and moving live loads
very easily via the graphical interface. It does not have an improved concrete model
like CANDE, but it does use a linear-elastic gross section property approach typical
of the constitutive models still widely used for culvert load rating.

• In the case of load rating analyses for culverts which have been in the ground for a
number of years and installation details may have been lost, it is appropriate to use a
linear elastic soil model (Katona M. G., 2008).

• Some concern exists about the validity of RISA-2D, specifically, how it handles the
finite elements. Soil-structure interaction problems like this, even with a linear-elastic
soil model, approach the limit of RISA-2D’s capabilities. One concern is that RISA-
2D would not accurately model the interface between the soil elements and the beam
elements.

• Exploration into this issue revealed that CANDE-2007 includes interface elements,
but generally does not use them, but rather assumes a bonded condition between soil

TxDOT 0-5849 64
 and structure. Though this may not be appropriate for many soil structure systems,
this modeling condition is usually conservative compared to going through the extra
work of including interface elements. This holds particularly true considering the
linear-elastic soil model that is being considered in this case (Katona M. G., 2008).

• The Whitney’s ultimate capacity concrete model used in RISA-2D is very accurate.
The typical cases were more complex concrete models are more valuable are when
determining deflection or when confining steel is present. In the case of culvert load
rating, neither of these factors comes into play. The more complex models also set an
upper limit to the rating range. That is to say that the only valid load ratings are those
which use a test truck which is smaller than the rating truck.
These observations show no a priori reasons why RISA-2D would not be an appropriate
program for culvert load rating. Relative to validity, a test was developed to establish
whether results from RISA-2D appropriately correspond to those of CANDE. Assuming that
all other things are equal, if results from RISA-2D could be matched to CANDE to within
10% for moment, the research team felt it would be appropriate to use RISA for a Level 3
culvert analysis. CANDE could be used for correlating other models or for research-oriented
load rating studies.

3. A Simple Beam Model Comparison Between CANDE and RISA

In order to determine if CANDE and RISA behave the same mathematically, a simple
model was designed. This model was a simply supported beam, four feet in length, carrying
two feet of “soil.” In order to produce demands with similar precision (three significant
digits), the beam was specified using a modulus of elasticity related to concrete with one inch
by one inch dimensions. The “soil” elements were modeled using the low stiffness soil
elasticity modulus, one inch wide but 12 kcf in self weight. Two meshes were developed,
one using one foot square soil elements, and the other using six inch square elements.
Several lessons were learned from this experiment. First it was determined that eight or
more finite elements are needed in between supports to correctly predict even the shape of
the shear and moment diagrams. This arises from the way both programs translate element
loads through the nodes only. Therefore, instead of translating the weight of the soil
elements into a relatively uniform pressure along the beam, it resolves the load into an
equivalent number of point loads at the node locations. For example, if only two elements
are used (two foot square elements in this simple model), the shear and moment diagrams
appear to behave as though they were loaded with a single point load at the center of the
beam. This is clearly incorrect, and for this reason, it is appropriate to focus on the data from
the model which used six inch square elements.
The next discovery is that CANDE’s beam output is dependent upon the number of
elements that make up the beam. This suggests that CANDE uses a true finite element model
to determine the beam reactions. It was also found that the number of beam elements need
not be greater that the number of attached soil elements.
The way CANDE deals with beam elements is different from the way RISA deals with
them. RISA analyses are not dependent upon the number of subdivisions in the element. In

TxDOT 0-5849 65
fact one beam element was all that was required. This is due to the fact that RISA uses a
frame modeling approach which converts intermediate loads into equivalent fixed end
reactions which are then input into the solution matrix. This means that while the CANDE
model is sensitive to both the number of beam elements and the number of adjacent soil
elements, the RISA model is only sensitive to the number of adjacent soil elements.
Another modeling consideration is the culvert slice width. Originally the comparison
assumed both models to be one inch wide. This is a requirement in CANDE. However in
RISA, any width is possible. This is helpful because for 2D culvert rating analyses, the width
of the culvert slice is typically assumed to be one foot. Comparison of a one-inch wide RISA
model with a one-foot wide model converted into similar units showed that both models
produced the same results.
When comparing the predicted moments, shears and thrust between the RISA and
CANDE programs, it was found that they matched very closely (see Figure 3.10). For
moments, the differences were less than 10%. For axial loads, differences were less than
25%. For shear in the mid-span, the differences were acceptably slight. However, near the
supports, shear did not match as closely as desired. The shape of the shear diagrams for the
two models were similar. Both approaches seem valid and they are reasonably close.

TxDOT 0-5849 66
 Moment
0.600
moment (kip-ft/in.)

0.500
0.400
0.300
0.200
0.100
0.000
0 1 2 3 4 5 6 7 8 9 10

RISA - 8 FEA - 1" CANDE - 8 FEA - 1"

Shear
3.000
2.000
shear (kip/in.)

1.000
0.000
-1.000 0 1 2 3 4 5 6 7 8 9 10
-2.000
-3.000

RISA - 8 FEA - 1" CANDE - 8 FEA - 1"

Thrust
-3.000
-2.500
thrust (kip/in.)

-2.000
-1.500
-1.000
-0.500 0 1 2 3 4 5 6 7 8 9 10
0.000

RISA - 8 FEA - 1" CANDE - 8 FEA - 1"

FIGURE 3.10. DEMANDS FROM SIMPLE BEAM MODEL COMPARISON.

TxDOT 0-5849 67
 4. Full Culvert Model Comparison Between CANDE and RISA
To further evaluate consistency between the CANDE and RISA models, it was necessary
to analyze a full culvert using both programs and compare the results. The culvert selected
for this analysis was a two barrel culvert from sheet MC5-2, five feet wide and three feet tall
per span.
This culvert-model test revealed more nuances concerning the similarities between RISA-
2D and CANDE-2007. The most remarkable and important note is the sensitivity of these
models to the number of soil elements adjacent to the culvert beams. The CANDE model
only used six elements, while the initial RISA model used eight. This resulted in a 25 percent
higher rating using RISA, highlighting the importance of this issue.
For comparison purposes, CANDE was first modeled using a linear elastic beam model.
This is referred to as CANDE LEFE in the graphs. However, CANDE’s advantage over
RISA is its advanced concrete model capabilities. When this was used, another problem
emerged. CANDE refused to converge upon a solution. As seen in the linear elastic models,
the test culvert for most soil qualities and models does not rate for an HS-20 truck. Therefore
when CANDE attempts to “accurately” model the concrete, it finds that the concrete fails and
“yields” without convergence. To work around this problem, the load was reduced to the
truck weight represented by the CANDE LEFE model. This also leads to a more precise
rating because CANDE does not follow a linear relationship and therefore, the linear
assumptions in the load rating equations do not really hold. To properly rate a culvert using
CANDE, the maximum size truck should be used to result in a load rating factor of 1.0. The
resulting load rating factors are included in the following graphs as CANDE Advanced (uses
advanced concrete model).

25.0

20.0
HS load rating

15.0

10.0

5.0

0.0
Level 1: CULV- Level 1:RISA Level 2: RISA- Level 3: RISA- Level 3: Level 4:
5 2D Springs 2D LEFE CANDE LEFE CANDE ADV

AASHTO Load Low Soils Medium Soil High Soils

FIGURE 3.11. LOAD RATING COMPARISON OF FULL SCALE CULVERT

Figure 3.11 shows the comparison of load ratings determined by the different analytical
models. For purposes of this analysis, it should be noted that the RISA and CANDE models

TxDOT 0-5849 68
produce similar rating trends. Results from CANDE using the full array of its concrete
models are lower than results from RISA because of CANDE’s nonlinear analysis approach.
When trying to determine whether RISA and CANDE yield similar results, it is more
helpful to compare the moment ratings from similarly meshed models (Figure 3.12). This
figure shows that a RISA model with essentially the same mesh as the CANDE model
produces essentially the same load rating.

1.4
1.2
1
load rating factor

0.8
0.6
0.4
0.2
0
Level 1: CULV-5 Level 1:RISA Level 2: RISA-2D Level 3: RISA-2D Level 3: CANDE
Springs LEFE LEFE

AASHTO Load Low Soils Medium Soils High Soils

FIGURE 3.12. MOMENT RATING OF FULL SCALE CULVERT

TxDOT 0-5849 69
 5. RISA-2D as the Level 3 Analysis Program of Choice

RISA-2D and CANDE handle linear elastic finite element modeling of moment and
deflection in essentially the same ways. Some discrepancy occurs in the calculations of shear
and thrust, but these rarely control in culvert applications. The lessons learned during this
comparison between RISA and CANDE are:

• Eight of more “soil” elements must be used along each culvert slab or wall to get
similar results
• RISA-2D may be under-conservative when calculating shear
• For problem culverts which rate at less than HS-20, CANDE may not provide a
solution when using the advanced concrete models
• Culvert analyses using CANDE’s advanced concrete models must use a truck
load equal to the rating to produce an accurate rating.
For TxDOT’s culvert load rating applications, this research indicates that RISA-2D and
CANDE-07 do provide similar results for moment ratings. By virtue of its production-ready
capabilities and in particular, its user-friendly graphical user interface, RISA-2D emerges as
the program of choice for Level 3 analytical modeling.
Therefore, the Culvert Rating Guide identifies RISA-2D as the Level 3 model, and
provides step-by-step instructions to construct a model, set boundary conditions and load
cases, and determine the demands using RISA-2D with linear elastic finite elements. The
Culvert Rating Guide provides a straightforward overview of the issues associated with using
finite element modeling. It also discusses the strengths and limitations of the RISA-2D
program for modeling a linear elastic finite element soil-structure system.

TxDOT 0-5849 70
 8. COMMENTARY ON THE CULVERT RATING GUIDE

The following paragraphs provide a commentary on development of each chapter of

the Culvert Rating Guide.

1. Chapter I: Introduction

The introduction to the Culvert Rating Guide discusses the purpose, history, scope
and philosophy of the Guide. The purpose of the Culvert Rating Guide is the purpose of
this research endeavor; that is, to articulate a clear, reliable procedure for the load rating
of culverts. This section also introduces the AASHTO definition of a load rating as the
largest truck load permissible on a structure either indefinitely (the inventory rating) or
limited use (operating rating) (AASHTO, 2003).
The history of culvert design in TxDOT highlights design eras and unique design
information, in particular, design philosophies and the problematic Texas Highway
Department Supplement No. 1. This information derives from a thorough investigation of
TxDOT reinforced box culvert designs archived by the TxDOT Bridge Division.
The scope of the Culvert Rating Guide was limited by the Project Monitoring
Committee to include the load rating of in-service, cast-in-place, reinforced concrete box
culverts with drained backfill and adequate visual inspection condition. The principles
in the Culvert Rating Guide have limited applicability to other culvert types and
conditions but must be applied with caution and engineering judgment.
The philosophy of the Culvert Rating Guide summarizes the load rating process as a
comparison of capacity, dead load demand and live load demand. The capacity is
determined using AASHTO specifications. The dead and live load demands are
determined using an analytical modeling approach which uses analytical tools of
escalating complexity, difficulty and accuracy.
2. Chapter II: Policy Requirements

It has been noted that the Project Monitoring Committee established that the policy
which must be followed for culvert load rating is found in the AASHTO Manual for
Condition Evaluation of Bridges (AASHTO, 2003) and Standard Specifications for Highway
Bridges (AASHTO, 2002). Whereas historic TxDOT practices for culvert load rating
sometimes differs from AASHTO policy, the Culvert Rating Guide requires that AASHTO
policy must be satisfied.
The AASHTO SSHB defines three failure modes, critical sections, and total and reduced
lateral load cases. A unique interpretation of these policy requirements is defined explicitly
in the Guide in order to avoid confusion. An analysis approach which considers the cross
sectional slice of the culvert as a two dimensional model is also explicitly described and is
typical in structural analysis practice. The rating variables as defined by AASHTO SSHB
and MCEB are also explicitly defined and interpreted. The MCEB rating equations are also
reiterated and defined.

TxDOT 0-5849 71
 3. Chapter III: Culvert Load Rating Procedure

Load rating is one portion of the larger culvert inspection process. The third chapter
of Culvert Rating Guide puts load rating into context and outlines the overall load rating
procedure. The load posting and visual inspection context of the culvert load rating
process are governed by TxDOT’s Bridge Inspection Manual, AASHTO’s MCEB, and
the Federal Highway Administration’s (FHWA) Culvert Inspection Manual (AASHTO,
2003; FHWA, 1986; TxDOT, 2002).
The complexity of the culvert load rating process is illustrated using the flow chart
reprinted as Figure 3.2 in this report. The Culvert Rating Guide also emphasizes the
concept of economy of work versus sophistication of analysis when selecting an
analytical model and the importance of quality control, review and checking of load
rating calculations and the oversight of a Licensed Professional Engineer.
4. Chapter IV: Culvert Details

Chapter four of the Culvert Rating Guide leads the engineer through the process of
collecting the pertinent data needed to load rate a culvert. The section on units discusses
the typical units used to describe culverts. Culvert dimension variables are defined and
given symbolic nomenclature. The structural material properties, such a steel and
concrete, are discussed. These data can be taken from the standards or plan sheets,
collected from tests on field samples or from steel quality control tickets, or assumed to
be the values provided by AASHTO MCEB.
The section dealing with soil properties addresses soil unit weight, modulus of
subgrade reaction, Poisson’s ratio, and modulus of elasticity. The simplest of the soil
parameters is the unit weight, with this value controlled by the AASHTO SSHB.
The modulus of subgrade reaction used in the second level analysis can be
determined by correlation based on soil classification of the bearing soil for the culvert.
The Culvert Rating Guide presents a table of typical values selected from pavement and
beam-on-elastic-foundations texts. Alternatively, this parameter can be established
through field tests.
Soil input parameters for the third level of analysis include soil modulus and
Poisson’s ratio. The parametric analysis discussed in Chapter 5 in this report indicates
that the culvert load rating is not particularly sensitive to Poisson’s ratio.
The soil modulus of elasticity is a highly important variable for the Level 3 analysis.
The load rating varies widely based on this parameter. A table showing typical modulus
ranges and recommended values based on soil classification is provided. This table was
derived from a collection of published geotechnical and pavement resources. The
Culvert Rating Guide recommends that the modulus be verified in the field. Based on
the field test portion of this project, the falling weight deflectometer is recommended for
validating the soil modulus.

TxDOT 0-5849 72
 The next section of chapter four outlines the nomenclature for discussing the steel
reinforcing schedule. This is fairly normal for the discussion of doubly reinforced
concrete beams and slabs.
The last section of chapter four discusses the culvert installation method. AASHTO
manuals discuss this variable at length, though not articulately. In general, it is believed
that the installation method only affects the early age behavior of the culvert. For load
rating of older, in-service culverts, the installation method can be ignored.
5. Chapter V: Culvert Capacity Calculations

Calculation of the culvert capacity is dictated by policy. The first section of this
chapter identifies the applicable policy sections from the AASHTO SSHB.
The second section articulates the unified sign convention for the Culvert Rating
Guide. The bending sign convention was developed according to typical structural
standard for the top midspan. This convention was applied throughout the culvert in
relationship to the inside and outside of the culvert, rather than the typical top and
bottom tensile faces. The sign convention in the Culvert Rating Guide states that for
positive bending, the tension face is inside the culvert (i.e., for the top midspan the
bottom surface is in tension). When the tension face is outside the culvert, bending is
negative (i.e., for the top midspan the top surface is in tension).
The third section outlines a generalized procedure for determining the capacities for
the critical sections in the culvert. Though not all the equations shown in this section
are included in the SSHB they are all derived from the code.

6. Chapter VI: Analytical Modeling for Demand Loads

The sixth chapter is the heart and soul of the Culvert Rating Guide. The first section
provides an overview for the rest of the chapter. The analytical modeling philosophy
starts with the models which are the easiest to use and have the most conservatism, and
then moves to more complex models with less conservatism. The demand calculation
process for culvert load rating is significantly less defined than the capacity calculations.
Again, guidance exists in the AASHTO policy, in particular Section 6 of the SSHB, but it
gives the engineer much more leeway. The AASHTO SSHB specifies a soil unit weight,
equivalent fluid weight and live load distributions through the soil. It does not specify the
type of model that should be used or specifically how to apply the loads to the models.
Because of this, great care must be taken when constructing a model, and every model-
specific assumption should be noted.
The second section of this chapter provides a generalized process for determining
demands. This procedure determines the demands and then uses the previously calculated
demands to determine load rating factors and select the controlling load rating.
The third section of the chapter provides special guidance for when shear controls. The
generalized approach to culvert load rating assumes that moment controls, and this typically
is the case. However, if shear controls the code allows for a less conservative shear critical

TxDOT 0-5849 73
section, as well as a demand-dependent shear capacity. This section provides a step-by-step
procedure for applying these more complicated shear provisions.
The remaining sections of this chapter provide specific guidance for performing demand
calculations using three different models of increasing model sophistication. When
developing the Culvert Rating Guide, all possible models were considered and evaluated to
determine production readiness.
7. Chapter VII: The General Analytical Model for Culvert Load Rating

In this chapter, a Level 4 analysis is defined. More sophisticated models, or models

which are not production oriented, may be used to further refine the culvert load rating. The
Culvert Rating Guide makes it clear that a Level 4 analysis is used only in specific cases.
Usually such a load rating will not be economically feasible for production rating of culverts.
This chapter also discusses the differences and advantages between two-dimensional and
three-dimensional models. CANDE-07 is the recommended program for a Level 4 two-
dimensional analysis. Other generalized programs can be used for three-dimensional
analysis.

8. Chapter VIII: Limitations

Chapter eight of the Culvert Rating Guide includes specifically articulated limitations for
using the Guide. These limitations are generally bounded by the limitations of the validation
process. These include the culvert type, fill depths, backfill drainage conditions, soil
parameters and analytical models. The guiding force in determining the scope of
applicability was driven by the perceived needs of TxDOT as articulated by the project
monitoring committee.

TxDOT 0-5849 74
 9. VALIDATION OF THE CULVERT RATING GUIDE

The Culvert Rating Guide was written in several drafts with critical reviews, both
informal and formal, conducted on each version. Having articulated TxDOT’s recommended
practices and procedures for culvert load rating, the research focus was directed toward
validating these practices and procedures. This involved three distinct research activities.
First, the Culvert Rating Guide was applied to a statistically representative sample of 100
of TxDOT’s 1477 unique culvert designs. Chapter 4 of this report presents this research
effort.
Second, a parametric study was performed to evaluate six independent variables
associated with culvert load rating. Chapter 5 of this report presents the parametric analysis.
Third, the research team instrumented and load tested three in-service culverts in the
field. This work facilitated a comparison of measured demand moments to predicted values
obtained through analytical modeling as per the Culvert Rating Guide. Chapter 6 of this
report presents this research effort.
Taken together, these three research tasks explore the breadth of application, sensitivity
of expression, depth of correlation of the culvert rating practices and procedures in the
Culvert Rating Guide relative to the full population of TxDOT’s reinforced concrete box
culverts.

TxDOT 0-5849 75
 4. LOAD RATING 100 TXDOT CULVERT DESIGNS

1. OVERVIEW

This chapter reports research findings from load rating a statistically representative
sample of 100 TxDOT culvert designs. The research team performed this work with several
objectives in mind, including:

• Confirm that the load rating procedures in the Culvert Rating Guide are clear and
repeatable.

• Verify that the load rating procedures in the Culvert Rating Guide can be applied
to the full population of TxDOT culvert designs.

• Determine whether load rating results for a broad sample of TxDOT’s culvert
designs cohere with the intuitions and experiences of TxDOT engineers who
work with culverts on a daily basis.

• Identify characteristics of “problem” culverts

• Test the hypotheses that analytical models with increased sophistication will
produce increased load ratings.
Load-rating 100 culvert designs provided a broad evaluation of the efficacy of the
Culvert Rating Guide relative to production load rating of TxDOT culverts. The following
sections of this chapter present the findings.

TxDOT 0-5849 78
 2. METHOD

1. Sample of 100 Representative Culvert Designs

TxDOT’s culvert inventory includes 13,192 in-service culverts, constructed between

1905 and 2008, as shown in Figure 4.1. The TxDOT culvert design archives reflect 1477
unique designs on 73 standard box culvert design sheets. Culvert design has evolved
throughout TxDOT’s history. TxDOT archives reveal four eras of culvert design, each
representing substantively different design approaches. These are the 1938 era, the 1946 era,
the 1958 era, and the 2003 era.
Culvert designs from the 1938 era were designed using slightly unconservative earth
loads, lower truck loads, but overly-conservative concrete construction that resulted in very
durable culverts. The 1938 collection consists of 428 different culvert designs representing a
diverse range of span lengths, number of spans, and barrel heights. Fill depths typically
range from 0 to 6 feet.
During the mid-1940s, principally when the Farm-to-Market road system was being
constructed, new culvert designs (59 total) were added to the body of 1938 designs. The
1946 era culverts were issued under the less conservative structural codes of the Texas
Highway Department Supplement No. 1. These designs resulted in culverts which generally
perform well, but which are not as robust as culverts designed per current AASHTO
standards.

TxDOT Culverts, Constructed by Year

450
number of constructed culverts

400
350
300
250
200
150
100
50
0
1905
1916
1920
1923
1926
1929
1932
1935
1938
1941
1944
1947
1950
1953
1956
1959
1962
1965
1968
1971
1974
1977
1980
1983
1986
1989
1992
1995
1998
2001
2004
2007

contructed year

FIGURE 4.1. TXDOT’S AS-BUILT CULVERT POPULATION

In 1958, coincident with the advent of the Interstate Highway System, TxDOT
redesigned and reissued their full set of culvert construction drawings. The 1958 set consists
of 380 designs representing a diverse range of span lengths, number of spans, and barrel
heights, with fill depths from 0 to 6 feet. The 1958-era designs use slightly less conservative
soil loads but more conservative structural considerations and HS-20 truck loads.

TxDOT 0-5849 79
 The most recent era of culvert designs dates from 2003. Once again TxDOT redesigned,
expanded, and reissued their complete set of culvert construction drawings. The 2003 set
consists of 610 culvert designs, including new designs for deep fill culverts with fill heights
up to 23 feet. Culvert designs for the 2003 era are based on current AASHTO policy.
For this study, 100 culvert designs were selected to statistically represent the full
population of TxDOT culverts. In sample selection, consideration was given both to the
diversity of culvert designs and the characteristics of the as-built culvert population.
1. Weighted by Era

Because of the importance of era, the 100 sampled culvert designs were selected with
consideration to the number of designs in an era and the number of culverts actually
constructed. Figure 4.2 shows the percentage of designs, the percentage of constructed
culverts and the percentage of the 100 culvert sample corresponding to each era.

80%
59%
60%
designs (%)

39% 39% 41%

40% 29% 30% 26% 21%
20% 10%
4% 0% 3%
0%
1938 1946 1958 2003
design era (yrs)

Designs (%) Built (%) Sample (%)

FIGURE 4.2. CULVERT DESIGNS VS BUILT VS SAMPLE BY DESIGN ERA

Allocation of the 100 culvert sample to the different design eras was accomplished as
follows:

• 1938 era – 30 culverts. This number is representative of both the percentage of

culverts built and the percentage of designs. These culverts typically are not
problematic from a load-rating perspective.

• 1946 era – 10 culverts. As-built culverts of the 1946 era are included with the
1938 era statistic. TxDOT personnel identified culverts from this era as
problematic relative to load rating. This era was oversampled to reflect their
concerns.

• 1958 era – 39 culverts. The selected number of culverts is roughly the average of
the percentage of designs and the percentage of constructed culverts. Because the
majority of Texas’ in-service culverts were built in this era, it was felt that this era
needed to be more heavily sampled.

• 2003 era – 21 culverts. Though very few culverts have been built in the 2003 era,
this is the era with the most designs. It is also unclear when these designs will be

TxDOT 0-5849 80
 updated, so this design era may continue for the next twenty years or more.
Therefore, it was appropriate to analyze several designs from this era in an effort
to consider not only culverts that have been constructed, but those that will be
built.

Having allocated the 100 culvert sample to the four design eras, this set the stage for
stratified sampling relative to other key culvert variables.
2. Representative by Variable

The full population of culverts was sorted by independent variable, and the sample was
randomly selected to represent the population. Besides design era, there are four
distinguishing and important culvert design variables. These are the number of spans, the
box length, the box height and the depth of fill. A stratified sort by these variables achieved
the sample of 100 culverts which is representative for all variables.
1. Number of Spans

It is important to consider the number of spans in a multi-barrel box culvert. It is

anticipated that single span culverts will experience greater stress than equivalent multi-span
culverts. Multi-span culverts should have reduced moment, at least in the top and bottom
slabs, because of the effect of continuous slab support. Figure 4.3 shows the percentage of
culvert designs, constructed culverts and sampled culvert designs for each era.
2. Length of Spans

Another unique culvert variable is the length of the individual spans. This variable
should be well represented from short spans to very long spans. Figure 4.4 shows the
distributions for the length of span for each era.
3. Height of Box

The height of the culverts is another important variable. Perhaps, only the tallest boxes
will experience significant bending in the wall mid-span, but the whole range of heights
should be considered. Figure 4.5 illustrates that the sample distribution, though not perfectly
matching the culvert design distribution, does cover the whole range of box heights.
4. Depth of Fill

The last culvert variable is the maximum design depth of fill. Depth of fill is important
because of its direct effect on the ratio of the live load demand to the dead load demand (this
ratio is the essence of load rating). When the depth of fill is low, the dead load demand is
low while the live load demand is high. When the depth of fill is high, the live load becomes
nearly negligible and dead load dominates. At intermediate fill depths, the effect of dead and
live load is split. Therefore, it is crucial to match ranges of depth of fill in the sample to the
ranges in the culvert design population. Figure 4.6 shows the distributions in each design
era.

TxDOT 0-5849 81
 Figure 4.2 through Figure 4.6 show that the sample of 100 culvert designs is an accurate
representation of the total population of TxDOT culvert designs by era. Results and trends
detected for the sample of 100 culverts will therefore indicate trends in the whole culvert
population.

TxDOT 0-5849 82
 1938
40%
designs (%) 29%
26%
30% 19% 20% 19% 20%
17% 17% 17% 16%17%
20% 11%
13% 12% 10%
8%10% 7%
10% 0%
3% 4%
0% 2% 0% 2% 2% 0%
0%
1 2 3 4 5 6 7 8 9
spans (#)

1946
30%
designs (%)

20% 19% 20% 20% 20% 19%

20% 14% 15%
10% 10% 10% 10%
10% 7% 5%
2% 0%
0%
2 3 4 5 6 7 8 9
spans (#)

1958
40% 33%
designs (%)

30% 20% 21% 20% 21% 20%

25%
21% 20% 21% 20% 18%
20% 15% 13%
9%
10%
0%
2 3 4 5 6
spans (#)

2003
28%
30% 24% 24%
designs (%)

20% 20% 19% 20% 19% 20% 19% 20% 19%

20% 13% 14% 13%
9%
10%
0% 0%
0%
2 3 4 5 6 7+
spans (#)

Designs (%) Built (%) Sample (%)

FIGURE 4.3. DISTRIBUTION OF THE NUMBER OF SPANS FOR DESIGNS, BUILT AND SAMPLE POPULATIONS

TxDOT 0-5849 83
 1938
designs (%) 30% 23%
27%
22% 23%
27%
20% 20%
20% 13% 14%
16%
12%
15% 13% 13%
17%
10% 10%
10% 5%

0%
5 6 7 8 9 10
length of span (ft)

1946
60%
42%
designs (%)

40% 40%
40% 25%

20% 8% 10% 8% 8% 10% 7%

0% 0%
0%
3 4 5 6 7 10
length of span (ft)

1958
40% 34%
28% 28%
designs (%)

30% 20% 20% 20% 20% 20%

16%
20% 11% 8% 11% 11%
15%
9%
15%
8%
10% 5%

0%
5 6 7 8 9 10
length of span (ft)

2003
40% 30%
34% 33%
designs (%)

30% 24%
19% 20%
16% 14% 16%
20% 9% 10%
14% 12%
9%
14%
7%
10% 5%
0%
0%
5 6 7 8 9 10
length of span (ft)

Designs (%) Built (%) Sample (%)

FIGURE 4.4. DISTRIBUTION OF THE LENGTH OF SPANS FOR DESIGNS, BUILT AND SAMPLE POPULATIONS

TxDOT 0-5849 84
 1938
designs (%) 30%
20% 20%
20% 11%
17%
13% 14%
17%
11% 10% 10%
9% 10% 10%
10% 6%
3%
7% 5% 3% 3%
0% 0% 0%
0%
2 3 4 5 6 7 8 9 10 11 12
height of box (ft)

1946
60%
designs (%)

40%
40% 25%
15% 17% 20% 20%
20% 8% 10% 8% 10% 8% 8% 8%
0% 0% 0%
0%
3 4 5 6 7 8 9 10
height of box (ft)

1958
28%
30%
designs (%)

17% 18%
20% 13% 13% 13% 13%
16%
12% 13%
9% 8% 8% 8%
10% 3% 3% 4% 3%

0%
2 3 4 5 6 7 8 9 10
height of box (ft)

2003
30% 24%
designs (%)

20% 18% 19%

20% 11%
14% 14% 16% 14%
12%
7% 9%
10% 2% 0%
5% 5% 5% 5%

0%
2 3 4 5 6 7 8 9 10
height of box (ft)

Designs (%) Sample (%)

FIGURE 4.5. DISTRIBUTION OF BOX HEIGHTS FOR DESIGNS AND SAMPLE POPULATIONS

TxDOT 0-5849 85
 1938
designs (%) 60% 46%43%

40% 32%30%

20% 8% 10%
4% 3% 1% 3% 1% 0% 1% 3% 1% 3% 1% 0% 1% 3% 1% 0%
0%
0 3 4 6 8 10 12 14 16 18 20
maximum depth of fill (ft)

1946
100% 80%
designs (%)

66%

50% 34%
20%

0%
0 6
maximum depth of fill (ft)

1956
60% 44%
designs (%)

37% 37%
40% 26% 28% 28%

20%
0%
2 4 6
maximum depth of fill (ft)

2003
40% 29%
33%
designs (%)

30% 19% 20%

24% 25%

20% 10%
13% 13%
10%
10% 5%
0%
0%
7 10 13 16 20 23
maximum depth of fill (ft)

Designs (%) Sample (%)

FIGURE 4.6. DISTRIBUTION OF MAXIMUM DEPTH OF FILL FOR DESIGNS AND SAMPLE POPULATIONS

TxDOT 0-5849 86
 2. Application of Load Rating Method to Domains

Each culvert design in the sample of 100 TxDOT culvert designs was load rated seven
times following the procedures in the Culvert Rating Guide. The first load rating used
TxDOT’s prepackaged culvert analysis program, CULV-5, to calculate demands (Level 1
analysis). The next three load ratings used RISA-2D as a two-dimensional structural
program. Each rating used a different value for the spring constant from the Culvert Rating
Guide, hence the three ratings (Level 2 analysis). The final three load ratings were
determined using RISA-2D as a two-dimensional, soil-structural program. Each rating used a
different modulus of elasticity from the Culvert Rating Guide (Level 3 analysis).
The culvert load rating process yields inventory and operating ratings. To fully describe
the nature of the ratings, the controlling critical section, failure mode (moment, shear or
thrust) and load case (total or reduced lateral) must also be defined. This means that for a
single culvert load rating, a total of 32 data points is produced: seven inventory load ratings,
seven operating load ratings, seven controlling critical sections, seven controlling failure
modes, and four controlling load cases.

TxDOT 0-5849 87
 3. LOAD RATING RESULTS FOR 100 CULVERT DESIGNS

1. Operational Statement and Hypotheses

As noted in Chapter 1, it is expected that more sophisticated, production-ready analytical

models produce more reliable load ratings for culverts. To test this concept, our research
compared the results of seven load ratings per culvert, performed upon 100 representative
designs for TxDOT cast-in-place, reinforced concrete box culverts, using CULV-5, RISA-2D
with springs (structural model) and RISA-2D with linear-elastic finite elements (LEFE) (soil-
structural model). The objective was to determine whether the two-dimensional soil-
structural model produced a higher load rating than the two-dimensional structural model,
which in turn should produce a higher load rating than the prepackaged culvert-specific
model.
Modeling sophistication is what allows the load-soil-structure interaction portion of
reliability to be identified and analyzed. The CULV-5 program (Level 1 analysis) does not
model soil-structure interaction. Therefore, a high degree of conservatism must be included
in this model to account for the uncertainties associated with load-soil-structure interaction.
The RISA-2D with spring model slightly reduces some of this over-conservatism by
considering the interaction between the foundation and the bottom slab (Level 2 analysis).
RISA-2D with LEFE removes a great deal of uncertainty by modeling the whole load-soil-
structure system in two dimensions (Level 3 analysis). Presumably, some over-conservatism
still exists in the out-of-plane direction.
The hypotheses are designed to test the portion of reliability associated with load-soil-
structure interaction as detected by model sophistication. Because uncertainty and over-
conservatism are removed with each increase in modeling sophistication, it was expected that
the load ratings would generally increase with increases in modeling sophistication.
The hypotheses for this study are:
H-1. RISA-2D with spring supports (Level 2 analysis) will produce a higher rating
than CULV-5 (Level 1 analysis).
H-2. RISA-2D with LEFE (Level 3 analysis)will produce a higher rating than
RISA-2D with springs (Level 2 analysis).
H-3. RISA-2D with LEFE (Level 3 analysis)will produce a higher rating than
CULV-5 (Level 1 analysis).
2. Presentation of the Research Findings

The results of the seven-fold load rating of the 100 culverts produced more than 3000
data points. To simplify presentation of this information, results will be considered by design
era. This means that four sets of results will be presented, one per culvert design era. The
direct relationship between inventory and operating ratings also simplifies the problem.
These two ratings show exactly the same trends with only a coefficient difference between
them. Therefore, only one plot for the dependent variable, in this case inventory rating, is
needed. Appendices A through E include all the data in tabulated form by era; however, to

TxDOT 0-5849 88
make trends easier to identify, this chapter presents the results in chart form. Five different
classes of data are presented: (1) undifferentiated inventory ratings, (2) identification of
statistically-significant culvert variables, (3) differentiated inventory ratings by significant
culvert variable, (4) controlling failure mode, and (5) controlling critical section.
1. Undifferentiated Inventory Ratings

The first plot will show the inventory ratings for all rated designs in the era. Results are
sorted by analytical method ordered according to the expected increase in load rating: i.e.
CULV-5, followed by RISA-2D with Springs using low, then medium, then high quality soil
parameters, followed by RISA-2D with LEFE using low, then medium, then high quality soil
parameters. Also included in this plot is the average of all the positive load ratings. A
negative load rating means that the culvert fails under dead load, but the magnitude of
negative load rating is essentially meaningless. The nature of the load rating equation makes
it extremely sensitive to small live loads. A large magnitude negative rating may actually be
closer to rating positively than a small magnitude negative rating. For this reason, negative
live loads are neglected in the average calculation. The average allows trends to be identified
more easily.
2. Identification of Statistically-Significant Culvert Variables

The second set of plots show the inventory ratings compared to the various independent
variables: depth of fill, box span, number of spans and box height. The coefficient of
determination for each relationship is used to identify which independent variables directly
affect load rating values. Those which are statistically significant are further evaluated.
3. Differentiated Inventory Ratings by Significant Culvert Variables

The third set of plots presents the inventory ratings in terms of the analytical method but
this time the plots are differentiated by significant variable. An average trend line is included
in these graphs as well. This will allow for the identification of trends between models and
the significant independent culvert variables.
4. Controlling Failure Mode

The fourth plot presents the controlling failure modes: moment, shear or thrust. If
multiple failure modes control, the plot differentiates between models. From these plots the
most probable failure mode can be identified.
5. Controlling Critical Section

The last set of plots shows the controlling critical section differentiated by significant
independent variable. The controlling critical sections (Figure 4.7) are expected to change
with analytical method. By comparing the critical sections to intuitive expectations,
interesting trends may become obvious.

TxDOT 0-5849 89
 TEC TEM TIC

WTEC WTIC

WEM WIM

WBEC WBIC

BEC BEM BIC

FIGURE 4.7. MOMENT CRITICAL SECTIONS FOR CULVERTS WITHOUT HAUNCHES.
Abbreviations for the typical critical sections shown in Figure 4.7, listed clockwise, are:
top exterior corner (TEC), top exterior mid-span (TEM), top interior corner (TIC), top
interior mid-span (TIM), wall top interior corner (WTIC), wall interior mid-span (WIM),
wall bottom interior corner (WBIC), bottom interior mid-span (BIM), bottom interior corner
(BIC), bottom exterior mid-span (BEM), bottom exterior corner (BEC), wall bottom exterior
corner (WBEC), wall exterior mid-span (WEM), and wall top exterior corner (WTEC). For
multiple-span box culverts, the sections are designated as per the culvert span; e.g., TIC1,
TIC2, BIC1, BIC2, etc.

TxDOT 0-5849 90
 4. 1938 ERA RESULTS

The 1938 era is the first complete set of TxDOT’s standard reinforced box culvert design
sheets. The designs are typified by larger gross slab dimensions, lower strength materials and
haunches. Though many culverts were originally designed for 15 ton trucks, because of the
inherent conservatism in the older allowable stress design and material assumptions, many of
these culverts perform very well, both analytically and in the real world.

1. Undifferentiated Inventory Ratings

For the 1938 era, Figure 4.8 shows all the inventory ratings and an average trend line.

150.0

100.0
inventory rating, IR (HS-tons)

50.0

0.0

-50.0

-100.0

-150.0
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
SPG SPG SPG LEFE LEFE LEFE
k=75pci k=150pci k=250pci E=8ksi E=20ksi E=36ksi
analytical models by increasing sophistication

FIGURE 4.8. 1938 ERA INVENTORY RATINGS: UNDIFFERENTIATED

It appears that on average, the first hypothesis is supported. The load ratings from RISA-
2D with spring-supports are just slightly higher than the CULV-5 ratings. Interestingly, the
increase from CULV-5 to RISA-2D with springs is of the same order of magnitude as the
increase from soft springs to medium springs and medium springs to stiff springs. The
increase is small.
The second hypothesis is only partially supported by the whole of the 1938 sample. If
the modulus of elasticity of the soil is high enough, an increase of load rating can be expected
from RISA-2D with springs to RISA-2D of LEFE. However, for low stiffness soil, the
calculated load rating using RISA-2D with LEFE tends to be less than the RISA-2D with
spring analysis. The hypothesis is supported by the high and medium stiffness soils, but
unsupported by low stiffness soils.

TxDOT 0-5849 91
 The third hypothesis is also only partially supported. For very stiff soils, the LEFE
model produces higher ratings, but for soft soils, the load rating is lower than even CULV-5.
RISA-2D with LEFE appears to be highly sensitive to the modulus value.

TxDOT 0-5849 92
 2. Identification of Statistically-Significant Culvert Variables

It is reasonable to ask, “Which independent variables affect the load ratings for a given
sample of culverts?” This question can be answered by performing linear regression of the
relationship between the independent variables and the actual inventory ratings. Figure 4.9
presents these results for the 1938 era along with the coefficient of determination for each
variable.

160.0 160.0
RISA-2D LEFE - H - IR (HS-)

RISA-2D LEFE - H - IR (HS-)

140.0 140.0
120.0 120.0
100.0 100.0
80.0 80.0
60.0 60.0
40.0 40.0
20.0 20.0
0.0 0.0
0 2 4 6 8 0.0 5.0 10.0 15.0
R² = 0.0944 Number of Spans R² = 0.0257 Span Length (ft)
P-value > 0.05 P-value > 0.05
160.0 160.0
RISA-2D LEFE - H - IR (HS-)

RISA-2D LEFE - H - IR (HS-)

140.0 140.0
120.0 120.0
100.0 100.0
80.0 80.0
60.0 60.0
40.0 40.0
20.0 20.0
0.0 0.0
0.0 5.0 10.0 15.0 0.0 5.0 10.0 15.0 20.0
R² = 0.0106 Barrel Height (ft) R² = 0.828 Max Fill Depth (ft)
P-value > 0.05 P-value = 2.25E-10
FIGURE 4.9. 1938 ERA STATISTICAL ANALYSIS FOR SIGNIFICANT INDEPENDENT VARIABLES
According to the linear regression analysis, the only significant variable for the 1938 era
designs is the depth of fill. No statistically significant relationships exist between number of
spans, barrel height or span length and the load rating. However, the relationship between
depth of fill and load rating tends toward higher load ratings for higher depths of fill.

TxDOT 0-5849 93
 3. Differentiated Inventory Ratings by Significant Culvert Variables

Figure 4.10 shows the inventory ratings for direct traffic culverts – that is, those with no
fill – in the 1938 era. In this subset, the first hypothesis is not supported. In this case,
CULV-5 produces a higher rating than the all but the stiffest soil springs in the RISA-2D
model. The second and third hypotheses are again only partially supported. For stiff soils,
RISA-2D with LEFE produces a higher rating than RISA-2D with springs and CULV-5.
However for low quality soils, the rating is lower.

20.0
inventory rating, IR (HS-tons)

15.0

10.0

5.0

0.0

-5.0
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
SPG SPG SPG LEFE LEFE LEFE
k=75pci k=150pci k=250pci E=8ksi E=20ksi E=36ksi
analytical models by increasing sophistication

FIGURE 4.10. 1938 ERA INVENTORY RATINGS: DIRECT TRAFFIC

Figure 4.11 and Figure 4.12 show the inventory ratings for medium fill culverts. This
subset supports the first hypothesis; RISA-2D with springs produces noticeably higher load
ratings than CULV-5. Again the second and third hypotheses are only partial supported. For
low soil stiffness the RISA-2D with LEFE produces lower ratings than the CULV-5 and
RISA-2D with spring models. However, if the soil is stiff enough, the second and third
hypotheses are supported.

TxDOT 0-5849 94
 50.0

inventory rating, IR (HS-tons)

40.0
30.0
20.0
10.0
0.0
-10.0
-20.0
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
SPG SPG SPG LEFE LEFE LEFE
k=75pci k=150pci k=250pci E=8ksi E=20ksi E=36ksi
analytical models by increasing sophistication

FIGURE 4.11. 1938 ERA INVENTORY RATINGS: 3' AND 4' OF FILL

100.0
inventory rating, IR (HS-tons)

80.0

60.0

40.0

20.0

0.0

-20.0
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
SPG SPG SPG LEFE LEFE LEFE
k=75pci k=150pci k=250pci E=8ksi E=20ksi E=36ksi
analytical models by increasing sophistication

FIGURE 4.12. 1938 ERA INVENTORY RATINGS: 6' OF FILL

Figure 4.13 shows the inventory ratings for four deep-fill culverts. These are from a
single sheet showing single barrel culvert designs. For these culverts, CULV-5 and RISA-
2D with springs provide very similar ratings. The results are so close, in fact, that they
cannot be said to support or not support the first hypothesis. The second and third
hypotheses are supported for these deep fill culverts for stiff soils. They are neither
supported nor unsupported for the lowest quality soil.

TxDOT 0-5849 95
 150.0

inventory rating, IR (HS-tons)

100.0

50.0

0.0

-50.0

-100.0

-150.0
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
SPG SPG SPG LEFE LEFE LEFE
k=75pci k=150pci k=250pci E=8ksi E=20ksi E=36ksi
analytical models by increasing sophistication

FIGURE 4.13. 1938 ERA INVENTORY RATINGS: 8' TO 18' OF FILL

4. Controlling Failure Mode

Figure 4.14 shows that in the 1938 era, the load ratings are controlled by moment alone.
This is to be expected due to the haunches characteristic of this design era.

moment
shear
axial thrust

200

FIGURE 4.14. 1938 ERA CONTROLLING FAILURE MODES

TxDOT 0-5849 96
 5. Controlling Critical Section

Figure 4.15 shows the critical sections for the 1938 era stratified by depth of fill. The
number of designs show along the horizontal axis differs for CULV-5 because CULV-5 can
only load rate culverts with four or fewer barrels.

Direct Traffic
RISA-2D LEFE E=36ksi 0 1 0 12
RISA-2D LEFE E=20ksi 5 0 8
RISA-2D LEFE E=8ksi 8 0 5 WBEC
RISA-2D SPG k=250pci 12 0 1 WTEC
RISA-2D SPG k=150pci 11 1 0 1
WBIC
RISA-2D SPG k=75pci 12 0 1
CULV-5 3 1 5 0 WTIC

0 2 4 6 8 10 12 14

3' to 4' of Fill

RISA-2D LEFE E=36ksi 1 0 3
RISA-2D LEFE E=20ksi 3 0 1
RISA-2D LEFE E=8ksi 4 0 WBEC
RISA-2D SPG k=250pci 1 3 0
RISA-2D SPG k=150pci 2 2 0 WTEC
RISA-2D SPG k=75pci 3 1 0 WTIC
CULV-5 2 2 0

0 1 2 3 4

6' of Fill
RISA-2D LEFE E=36ksi 7 1 1
RISA-2D LEFE E=20ksi 8 0 1
RISA-2D LEFE E=8ksi 8 0 1 WBEC
RISA-2D SPG k=250pci 7 1 1
RISA-2D SPG k=150pci 7 1 1 WTEC
RISA-2D SPG k=75pci 7 1 1 WTIC
CULV-5 3 1 0

0 2 4 6 8 10

FIGURE 4.15. 1938 ERA CONTROLLING CRITICAL SECTIONS: 0 TO 6' OF FILL

For the direct traffic to medium fill culverts, the critical sections show the same
tendencies. First, all the critical sections are in wall corners. Second, softer soils tend to
push the critical section to the bottom corner, while stiffer soils are more likely to fail in the
top interior corners.

TxDOT 0-5849 97
 Figure 4.16 shows the critical section for the high fill culverts in the 1938 era. All of
these culverts also happen to be single barrel culverts. This may be partially responsible for
the mid-span critical sections controlling so often.

RISA-2D LEFE E=36ksi 3 0 1

RISA-2D LEFE E=20ksi 3 0 1
RISA-2D LEFE E=8ksi 3 0 1 0 WBEC
RISA-2D SPG k=250pci 0 3 0 1 WEM
RISA-2D SPG k=150pci 0 3 0 1 TEM
RISA-2D SPG k=75pci 0 2 0 2 BEM
CULV-5 2 1 0 1

0 1 2 3 4

FIGURE 4.16. 1938 ERA CONTROLLING CRITICAL SECTIONS: 8' TO 18' OF FILL

TxDOT 0-5849 98
 5. 1946 ERA RESULTS

The 1946 era is the smallest of the design eras with only 53 designs on three sheets. The
distinctive characteristics for this era must be broken down by sheet. Originally designed in
1946, the FM-MBC-3-26 sheet appears to be the last and pinnacle design using the 1938 era
design philosophy. It is a zero to six foot of fill design with haunches. The MBC-3 sheet
was also designed in 1946. This design has more in common with the 1958 era design
philosophy. The sheet was designed for direct traffic using slightly thinner slabs without
haunches. The MC-10-3-45 is the only sheet designed with a 45 degree skew angle. This
sheet was also designed under the Texas Highway Department Supplement Number 1. The
TxDOT Bridge Inspection Manual specifically notes that the THD supplement produced
drastically unconservative designs. The MC-10-3-45 was designed for four to six feet of fill.

1. Undifferentiated Inventory Ratings

Figure 4.17 shows all the inventory ratings for the 1946 era. This plot indicates that the
first hypothesis is generally supported by this era. The RISA-2D with spring model produces
higher ratings than the CULV-5 model. Again the second and third hypotheses are partially
supported. For medium and high quality soils, the RISA-2D with LEFE produces higher
ratings than RISA-2D with springs and CULV-5, but for low quality soil, RISA-2D with
LEFE produces the lowest ratings.

100
80
60
inventory rating, IR (HS-tons)

40
20
0
-20
-40
-60
-80
-100
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
SPG SPG SPG LEFE LEFE LEFE
k=75pci k=150pci k=250pci E=8ksi E=20ksi E=36ksi
analytical models by increasing sophistication

FIGURE 4.17. 1946 ERA INVENTORY RATINGS: UNDIFFERENTIATED

TxDOT 0-5849 99
 2. Identification of Statistically-Significant Culvert Variables

Figure 4.18 shows the linear-regression plots for the 1946 era. However, because of the
transitional nature of the design philosophies in this era, the load rating is most significantly
affected by the design sheet rather than these variables.

90.0 90.0
RISA-2D LEFE - H - IR (HS-)

RISA-2D LEFE - H - IR (HS-)

80.0 80.0
70.0 70.0
60.0 60.0
50.0 50.0
40.0 40.0
30.0 30.0
20.0 20.0
10.0 10.0
0.0 0.0
0 5 10 0.0 5.0 10.0 15.0
R² = 0.0861 Number of Spans R² = 0.001 Span Length (ft)
P-value > 0.05 P-value > 0.05
90.0 90.0
RISA-2D LEFE - H - IR (HS-)

RISA-2D LEFE - H - IR (HS-)

80.0 80.0
70.0 70.0
60.0 60.0
50.0 50.0
40.0 40.0
30.0 30.0
20.0 20.0
10.0 10.0
0.0 0.0
0.0 5.0 10.0 15.0 0.0 2.0 4.0 6.0 8.0
R² = 0.0185 Barrel Height (ft) R² = 0.4235 Max Fill Depth (ft)
P-value > 0.05 P-value > 0.05
FIGURE 4.18. 1946 ERA STATISTICAL ANALYSIS FOR SIGNIFICANT INDEPENDENT VARIABLES

TxDOT 0-5849 100

 3. Differentiated Inventory Ratings by Significant Culvert Variables

Figure 4.19 shows the inventory ratings for the direct traffic designs sampled from the
MBC-3 sheet. These designs produced very low ratings. The first and third hypotheses are
unconfirmed for this subset, because the sampled culvert designs had more than four barrels
making them unratable using CULV-5. The second hypothesis is confirmed only for the
higher level soil stiffness. For low soil stiffness, RISA-2D with LEFE produces comparable
ratings to RISA-2D with springs.

20
18
inventory rating, IR (HS-tons)

16
14
12
10
8
6
4
2
0
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
SPG SPG SPG LEFE LEFE LEFE
k=75pci k=150pci k=250pci E=8ksi E=20ksi E=36ksi
analytical models by increasing sophistication
FIGURE 4.19. 1946 ERA INVENTORY RATINGS: MBC-3
Figure 4.20 shows that half the designs sampled from the sheet which were designed
using THD Supplement Number 1 fail to rate positively for all but the stiffest RISA-2D with
LEFE model. For the designs that do rate positively, the first hypothesis is only slightly
supported by marginally higher RISA-2D with spring ratings than CULV-5 ratings. The
second and third hypotheses are unsupported by the lowest soil stiffness in the RISA-2D with
LEFE model, indefinite for the medium soil stiffness in the RISA-2D with LEFE model and
supported for the high soil stiffness in the RISA-2D with LEFE model.

TxDOT 0-5849 101

 40

inventory rating, IR (HS-tons)

20

-20

-40

-60

-80
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
SPG SPG SPG LEFE LEFE LEFE
k=75pci k=150pci k=250pci E=8ksi E=20ksi E=36ksi
analytical models by increasing sophistication

FIGURE 4.20. 1946 ERA INVENTORY RATINGS: MC10-3-45

Figure 4.21 applies to sheet FM-MBC-3-26 which is most closely related to the 1938 era.
The newest haunch designs perform very well. The low CULV-5 rating supports the first
hypothesis. The second and third hypotheses are supported by the medium and high grade
soil in the RISA-2D with LEFE but are unsupported by the low grade soil stiffness.

100
inventory rating, IR (HS-tons)

80

60

40

20

0
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
SPG SPG SPG LEFE LEFE LEFE
k=75pci k=150pci k=250pci E=8ksi E=20ksi E=36ksi
analytical models by increasing sophistication

FIGURE 4.21. 1946 ERA INVENTORY RATINGS: FM-MBC-3-26

TxDOT 0-5849 102

 4. Controlling Failure Mode

Figure 4.22 shows that the only controlling failure mode is moment.

moment
shear
axial thrust

67

FIGURE 4.22. 1946 ERA CONTROLLING FAILURE MODES

5. Controlling Critical Section

Figure 4.23 shows that the earliest of the 1946 design philosophies tended to create weak
points in the top of the walls.

RISA-2D LEFE E=36ksi 0 2

RISA-2D LEFE E=20ksi 0 2
RISA-2D LEFE E=8ksi 0 2
RISA-2D SPG k=250pci 1 1 WTEC
RISA-2D SPG k=150pci 1 1 WTIC
RISA-2D SPG k=75pci 1 1
CULV-5 0

0 1 2

FIGURE 4.23. 1946 ERA CONTROLLING CRITICAL SECTIONS: MBC-3

Figure 4.24 shows that while the bottom wall exterior corner is the most likely place for
failure, the THD Supplement Number 1 may have lead to under designing the mid-spans.

TxDOT 0-5849 103

 RISA-2D LEFE E=36ksi 4 0
RISA-2D LEFE E=20ksi 4 0
WBEC
RISA-2D LEFE E=8ksi 4 0
WEM
RISA-2D SPG k=250pci 0 2 1 1 0
WTEC
RISA-2D SPG k=150pci 0 2 1 1 0
RISA-2D SPG k=75pci TEM
2 2 0
CULV-5 1 2 0 1 BEM

0 1 2 3 4

FIGURE 4.24. 1946 ERA CONTROLLING CRITICAL SECTIONS: MC-10-3-45

Figure 4.25 shows that for the sheet most related to the 1938 era, the trends are the same
as for the 1938 era. Stiffer soil models tend to move the critical section from the base of the
exterior wall to the top of the exterior wall.

RISA-2D LEFE E=36ksi 3 1

RISA-2D LEFE E=20ksi 3 1
RISA-2D LEFE E=8ksi 4 0
RISA-2D SPG k=250pci 1 3 WBEC
RISA-2D SPG k=150pci 2 2 WTEC
RISA-2D SPG k=75pci 2 2
CULV-5 1 2

0 1 2 3 4

FIGURE 4.25. 1946 ERA CONTROLLING CRITICAL SECTIONS: FM-MBC-3-26

TxDOT 0-5849 104

 6. 1958 ERA STRUCTURAL GRADE STEEL (36 KSI)

The 1958 era represents the design era under which most of TxDOT’s culverts have been
built. Designs from this era have no haunches. The designs were originally developed for
HS-20 loading but with a reduced dead load, and this resulted in increasing the allowable
stress in the structure. Under the current load rating requirements, these designs show some
of the greatest variety in load rating and critical section. Available data suggest that this
design era was updated in 1977 to require grade 60 steel instead of the assumed structural
grade steel required by the original design. This has the affect of basically splitting this era
into two sub-eras: the 1958-1977 era (which uses structural grade steel) and the 1977-2003
era (which uses grade 60 steel). Therefore the sample culverts in this era were analyzed
twice. The first round of analysis assumed structural grade steel with a yield stress of 36ksi.

1. Undifferentiated Inventory Ratings

Figure 4.26 shows all the load rating for the 1958 era using 36ksi steel. From this chart is
difficult to say if the hypotheses are supported. CULV-5, RISA-2D with springs and the
lowest quality RISA-2D with LEFE all produce equivalent results on average. The second
and third hypotheses are partially supported by the medium and high quality soils in the
RISA-2D with LEFE. These two methods provide higher ratings than all the others. Also of
interest is the fact that almost all of the less sophisticated models produce load ratings below
the required HS-20. It is only with decent soils and the LEFE model that the culverts begin
to load rate adequately.

60

40
inventory rating, IR (HS-tons)

20

-20

-40

-60
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
SPG SPG SPG LEFE LEFE LEFE
k=75pci k=150pci k=250pci E=8ksi E=20ksi E=36ksi
analytical models by increasing sophistication

FIGURE 4.26. 1958 ERA 36KSI STEEL INVENTORY RATINGS: UNDIFFERENTIATED

TxDOT 0-5849 105

 2. Identification of Statistically-Significant Culvert Variables

Figure 4.27 makes is clear that depth of fill is the only variable which is directly related
to the inventory rating.

70.0 70.0
RISA-2D LEFE - H - IR (HS-)

RISA-2D LEFE - H - IR (HS-)

60.0 60.0
50.0 50.0
40.0 40.0
30.0 30.0
20.0 20.0
10.0 10.0
0.0 0.0
0 2 4 6 8 0.0 5.0 10.0 15.0
R² = 0.0015 Number of Spans R² = 0.1067 Span Length (ft)
P-value > 0.05 P-value > 0.05
70.0 70.0
RISA-2D LEFE - H - IR (HS-)

RISA-2D LEFE - H - IR (HS-)

60.0 60.0
50.0 50.0
40.0 40.0
30.0 30.0
20.0 20.0
10.0 10.0
0.0 0.0
0.0 5.0 10.0 15.0 0.0 2.0 4.0 6.0 8.0
R² = 0.1593 Barrel Height (ft) R² = 0.652 Max Fill Depth (ft)
P-value = 6.99E-3 P-value = 1.37E-6
FIGURE 4.27. 1958 ERA 36KSI STEEL STATISTICAL ANALYSIS FOR SIGNIFICANT INDEPENDENT VARIABLES

TxDOT 0-5849 106

 3. Differentiated Inventory Ratings by Significant Culvert Variables

Figure 4.28 shows what is quickly becoming the normative trend. The CULV-5 rating is
lower than the RISA-2D with spring ratings, thus supporting the first hypothesis. The second
and third hypotheses are supported because the medium and high quality soils in the RISA-
2D with LEFE model produce higher ratings than RISA-2D with springs and CULV-5.
However, the second and third hypotheses are not supported totally because the RISA-2D
with LEFE model using low stiffness soils produces lower ratings than CULV-5 and RISA-
2D with springs.

20
inventory rating, IR (HS-tons)

15

10

-5
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
SPG SPG SPG LEFE LEFE LEFE
k=75pci k=150pci k=250pci E=8ksi E=20ksi E=36ksi
analytical models by increasing sophistication

FIGURE 4.28. 1958 ERA 36KSI STEEL INVENTORY RATINGS: 2' OF FILL
It is also worth noting that no culvert designed for two feet of fill rates high enough to not
require load posting. This suggests a weakness in the TxDOT culvert designs in this sub-era.
Figure 4.29 and Figure 4.30 show the inventory ratings for four foot and six foot of fill.
For these medium depth culverts, the ratings using CULV-5, RISA-2D with spring and the
low quality soil in the RISA-2D with LEFE model are too close together to make a judgment
call. It is clear that for higher stiffness soil, RISA-2D with LEFE rates higher than other
methods.

TxDOT 0-5849 107

 50

inventory rating, IR (HS-tons)

40
30
20
10
0
-10
-20
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
SPG SPG SPG LEFE LEFE LEFE
k=75pci k=150pci k=250pci E=8ksi E=20ksi E=36ksi
analytical models by increasing sophistication

FIGURE 4.29. 1958 ERA 36KSI STEEL INVENTORY RATINGS: 4' OF FILL

60
inventory rating, IR (HS-tons)

40
20
0
-20
-40
-60
-80
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
SPG SPG SPG LEFE LEFE LEFE
k=75pci k=150pci k=250pci E=8ksi E=20ksi E=36ksi
analytical models by increasing sophistication

FIGURE 4.30. 1958 ERA 36KSI STEEL INVENTORY RATING: 6' OF FILL

TxDOT 0-5849 108

 4. Controlling Failure Mode

Figure 4.31 shows that for this sub-era, moment is the predominant controlling failure
mode. A single rating failed in shear for the highest quality soil in the RISA-2D with LEFE.
This would be disturbing if that single rating were not nearly three times greater than the
design load (HS-59 compared to HS-20).

10

moment
shear
axial thrust

257

FIGURE 4.31. 1958 ERA 36KSI STEEL CONTROLLING FAILURE MODES

5. Controlling Critical Section

The controlling critical sections showed a great deal of variety. For the low fill culverts,
Figure 4.32 shows that CULV-5 produced failing critical sections in the bottom slab only.
TxDOT has always felt CULV-5 was overly conservative in the bottom slab. The other
models showed failure modes throughout the corners of the structures. Interestingly for the
higher level RISA-2D with LEFE models, the controlling critical sections occurred at interior
corners. These are not typically thought to be critical in culvert design.

RISA-2D LEFE E=36ksi 0 4 0 1 6 0 WBEC

RISA-2D LEFE E=20ksi 0 1 0 1 9 0 WEM
RISA-2D LEFE E=8ksi 3 0 1 1 6 0 WTEC
RISA-2D SPG k=250pci 0 1 5 1 1 3 0 TEC
RISA-2D SPG k=150pci 1 1 3 1 1 2 0 2 TIC
RISA-2D SPG k=75pci 4 1 0 1 1 2 2 WTIC
CULV-5 0 6 BEC
0 2 4 6 8 10 12 BEM

FIGURE 4.32. 1958 ERA 36KSI STEEL CONTROLLING CRITICAL SECTIONS: 2' OF FILL

TxDOT 0-5849 109

 Figure 4.33 shows that for the culverts designed for four feet of fill, the results are only
slightly different. CULV-5 tended to identify the bottom slab. In the RISA-2D with springs
models, the top slab was often the controlling section. For the RISA-2D with LEFE the
exterior wall corners were the weakest points.

RISA-2D LEFE E=36ksi 3 0 9 0 1 0

RISA-2D LEFE E=20ksi 7 0 5 0 1 0 WBEC
RISA-2D LEFE E=8ksi 11 0 2 0 WEM
RISA-2D SPG k=250pci 0 2 3 5 1 2 WTEC
RISA-2D SPG k=150pci 0 2 3 5 1 2 TEM
RISA-2D SPG k=75pci 4 2 0 4 1 2 BEC
CULV-5 0 2 0 4 BEM
0 2 4 6 8 10 12 14

FIGURE 4.33. 1958 ERA 36KSI STEEL CONTROLLING CRITICAL SECTIONS: 4' OF FILL
Figure 4.34 shows that for the six foot of fill culverts, the CULV-5 and RISA-2D with
springs identified similar critical sections in the wall and top slab mid-spans. For RISA-2D
with LEFE the bottom exterior wall corner tended to controlled again.

RISA-2D LEFE E=36ksi 11 0 2 1 10 WBEC

RISA-2D LEFE E=20ksi 13 0 2 0 WEM
RISA-2D LEFE E=8ksi 14 010 WTEC
RISA-2D SPG k=250pci 0 6 4 5 0 TEM
RISA-2D SPG k=150pci 1 6 2 4 0 2 0 TIC
RISA-2D SPG k=75pci 5 4 0 2 01 3 0 BEC
CULV-5 4 3 0 4 1 BEM
0 2 4 6 8 10 12 14 16 BIC

FIGURE 4.34. 1958 ERA 36KSI STEEL CONTROLLING CRITICAL SECTIONS: 6’ OF FILL

TxDOT 0-5849 110

 7. 1958 ERA GRADE 60 STEEL (60 KSI)

The second round of analysis for this era assumed a steel yield strength of 60ksi. The
demands were unchanged, but the capacity was increased. This resulted in different results
altogether.

1. Undifferentiated Inventory Ratings

In Figure 4.35, the most obvious improvement in the 1958 era using 60ksi steel is the
overall increase in the ratings. Even CULV-5 averages above HS-20. This is a significant
increase in load rating over the 36ksi equivalent. Much like the 36ksi sub-era, the values
neither support or deny the first hypothesis. CULV-5 and RISA-2D with springs simply
provide solutions which are too similar. What is interesting is that the second and third
hypotheses are for the first time, fully supported. All levels of RISA-2D with LEFE produce
higher ratings than CULV-5 or RISA-2D with springs.

80

60
inventory rating, IR (HS-tons)

40

20

-20

-40
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
SPG SPG SPG LEFE LEFE LEFE
k=75pci k=150pci k=250pci E=8ksi E=20ksi E=36ksi
analytical models by increasing sophistication

FIGURE 4.35. 1958 ERA 60KSI STEEL INVENTORY RATINGS: UNDIFFERENTIATED

TxDOT 0-5849 111

 2. Identification of Statistically-Significant Culvert Variables

For the 1958 era with 60ksi steel, Figure 4.36 shows that the only independent variable
that statistically impacts the inventory rating is the depth of fill.

80.0 80.0
RISA-2D LEFE - H - IR (HS-)

RISA-2D LEFE - H - IR (HS-)

70.0 70.0
60.0 60.0
50.0 50.0
40.0 40.0
30.0 30.0
20.0 20.0
10.0 10.0
0.0 0.0
0 2 4 6 8 0.0 5.0 10.0 15.0
R² = 0.0533 Number of Spans R² = 9E-05 Span Length (ft)
P-value > 0.05 P-value > 0.05
80.0 80.0
RISA-2D LEFE - H - IR (HS-)

RISA-2D LEFE - H - IR (HS-)

70.0 70.0
60.0 60.0
50.0 50.0
40.0 40.0
30.0 30.0
20.0 20.0
10.0 10.0
0.0 0.0
0.0 5.0 10.0 15.0 0.0 2.0 4.0 6.0 8.0
R² = 0.0002 Barrel Height (ft) R² = 0.9192 Max Fill Depth (ft)
P-value > 0.05 P-value = 1.56E-6
FIGURE 4.36. 1958 ERA 60KSI STEEL STATISTICAL ANALYSIS FOR SIGNIFICANT INDEPEDENT VARIABLES

TxDOT 0-5849 112

 3. Differentiated Inventory Ratings by Significant Culvert Variables

For this sub-era with two feet of fill, Figure 4.37 shows that all hypotheses are supported.
RISA-2D with LEFE produces higher ratings that RISA-2D with springs which is higher
than CULV-5.

25
inventory rating, IR (HS-tons)

20

15

10

0
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
SPG SPG SPG LEFE LEFE LEFE
k=75pci k=150pci k=250pci E=8ksi E=20ksi E=36ksi
analytical models by increasing sophistication

FIGURE 4.37. 1958 ERA 60KSI STEEL INVENTORY RATINGS: 2' OF FILL
For the four foot of fill culverts, Figure 4.38 shows that the case is not as conclusive for
the first hypothesis. The CULV-5 and RISA-2D with springs models all produce
approximately the same load ratings. The second and third hypotheses are fully supported.
The RISA-2D with LEFE models produce far greater ratings than the CULV-5 or RISA-2D
with spring models.

60
inventory rating, IR (HS-tons)

50

40

30

20

10

0
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
SPG SPG SPG LEFE LEFE LEFE
k=75pci k=150pci k=250pci E=8ksi E=20ksi E=36ksi
analytical models by increasing sophistication

FIGURE 4.38. 1958 ERA 60KSI STEEL INVENTORY RATING: 4' OF FILL
The six foot of fill culvert ratings in Figure 4.39 show inconclusive changes in the load
rating between the CULV-5, RISA-2D with springs and the lowest quality RISA-2D with

TxDOT 0-5849 113

LEFE models. That means that the second and third hypotheses are only partly supported by
the higher ratings of the RISA-2D with LEFE models using the medium and high stiffness
soil modulus.

80
inventory rating, IR (HS-tons)

60

40

20

-20

-40
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
SPG SPG SPG LEFE LEFE LEFE
k=75pci k=150pci k=250pci E=8ksi E=20ksi E=36ksi
analytical models by increasing sophistication

FIGURE 4.39. 1958 ERA 60KSI STEEL INVENTORY RATINGS: 6' OF FILL

4. Controlling Failure Mode

Figure 4.40 shows the controlling failure modes differentiated by model for the 1958 era
culverts using 60ksi steel. In this case shear controls nearly a third of the time. The increase
in steel strength is directly related to an increase in moment capacity, but shear capacity is
unaffected by reinforcing steel. Shear controls more often in the stiffer RISA-2D with LEFE
models.

RISA-2D LEFE E=36ksi 5 34

RISA-2D LEFE E=20ksi 16 23

RISA-2D LEFE E=8ksi 34 5

moment
RISA-2D SPG k=250pci 32 7
shear
RISA-2D SPG k=150pci 31 8 axial thrust

RISA-2D SPG k=75pci 30 9

CULV-5 18 6

0 5 10 15 20 25 30 35 40 45

FIGURE 4.40. 1958 ERA 60KSI STEEL CONTROLLING FAILURE MODES

TxDOT 0-5849 114

 5. Controlling Critical Section

Figure 4.41 shows that for this 1958 sub-era, the most popular controlling critical section
in the top of the interior wall corner. For CULV-5, the bottom mid-span controls. As the
soil becomes stiffer, the RISA-2D with LEFE model identifies critical sections in the top slab
at the interior corners.

RISA-2D LEFE E=36ksi 0 7 4 0

RISA-2D LEFE E=20ksi 0 3 8 0 WBEC
RISA-2D LEFE E=8ksi 0 1 10 0 WEM
RISA-2D SPG k=250pci 0 1 2 1 7 0 WTEC
RISA-2D SPG k=150pci 0 1 1 1 8 0 TIC
RISA-2D SPG k=75pci 1 1 0 1 8 0 WTIC
CULV-5 0 6 BEM
0 2 4 6 8 10 12

FIGURE 4.41. 1958 ERA 60KSI STEEL CONTROLLING CRITICAL SECTIONS: 2’ OF FILL
Figure 4.42 shows the critical sections for those culverts with four feet of fill. The less
sophisticated models show no defined trends. In the higher stiffness RISA-2D with LEFE,
the top span interior corners control almost exclusively.

RISA-2D LEFE E=36ksi 0 13 0 WBEC

RISA-2D LEFE E=20ksi 2 0 10 1 0 WEM
RISA-2D LEFE E=8ksi 5 0 3 4 0 WTEC
RISA-2D SPG k=250pci 0 2 2 4 2 3 0 TEM
RISA-2D SPG k=150pci 0 2 1 5 2 3 0 TIC
RISA-2D SPG k=75pci 0 2 1 4 2 3 0 WTIC
CULV-5 0 2 0 2 0 1 1 BEC
0 2 4 6 8 10 12 14 BEM

FIGURE 4.42. 1958 ERA 60KSI STEEL CONTROLLING CRITICAL SECTIONS: 4' OF FILL
In the culverts designed for six foot of fill, Figure 4.43 shows that the top interior corner
is the most likely critical section. For the less sophisticated models the exterior wall mid-
span also has a high probability of controlling the load rating.

TxDOT 0-5849 115

RISA-2D LEFE E=36ksi 0 15 0 WBEC

RISA-2D LEFE E=20ksi 4 0 9 0 0 WEM

WTEC
RISA-2D LEFE E=8ksi 11 01 2 10
TEM
RISA-2D SPG k=250pci 0 7 1 2 5 0
TIC
RISA-2D SPG k=150pci 0 7 0 2 6 0
WTIC
RISA-2D SPG k=75pci 0 6 0 7 0 2 0
BEC
CULV-5 0 4 0 3 5
BEM
0 2 4 6 8 10 12 14 16 BIC

FIGURE 4.43. 1958 ERA 60KSI STEEL CONTROLLING CRITICAL SECTIONS: 6' OF FILL

TxDOT 0-5849 116

 8. 2003 ERA

The most recent full set of culverts sheets were designed in 2003. These culverts feature
high strength steel and concrete, have no haunches and are relatively thin slabbed. This set
contains the largest number of designs and the fewest number of constructed culverts. The
most notable design characteristic is the depth of fill. Previous designs were developed for
no more than six feet of fill and usually for only a two foot range. The 2003 culverts are
designed for maximum fill between seven and twenty-three feet. For all designs the
minimum fill is two feet. This makes the whole set of culverts significantly different from all
other design eras.

1. Undifferentiated Inventory Ratings

Figure 4.44 shows the trends for all the designs sampled from this era. Clearly, the first
hypothesis is not supported. CULV-5 produces slightly higher ratings than the RISA-2D
with springs models. The second and third hypotheses are partially supported for the
medium and high stiffness soils in the RISA-2D with LEFE models. The low stiffness
modulus in the RISA-2D with LEFE produce lower load ratings than the CULV-5 and RISA-
2D with springs models. Also, on average the ratings are higher than HS-20.

240
220
200
180
inventory rating, IR (HS-tons)

160
140
120
100
80
60
40
20
0
-20
-40
-60
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
SPG SPG SPG LEFE LEFE LEFE
k=75pci k=150pci k=250pci E=8ksi E=20ksi E=36ksi
analytical models by increasing sophistication

FIGURE 4.44. 2003 ERA INVENTORY RATINGS: UNDIFFERENTIATED

TxDOT 0-5849 117

 2. Identification of Statistically-Significant Culvert Variables

Once again, Figure 4.45 shows that the only variable that significantly impacts the
inventory rating is the depth of fill.

250.0 250.0
RISA-2D LEFE - H - IR (HS-)

RISA-2D LEFE - H - IR (HS-)

200.0 200.0

150.0 150.0

100.0 100.0

50.0 50.0

0.0 0.0
0 2 4 6 8 0.0 5.0 10.0 15.0
R² = 0.0679 Number of Spans R² = 0.0042 Span Length (ft)
P-value > 0.05 P-value = 7.81E-3
250.0 250.0
RISA-2D LEFE - H - IR (HS-)

RISA-2D LEFE - H - IR (HS-)

200.0 200.0

150.0 150.0

100.0 100.0

50.0 50.0

0.0 0.0
0.0 5.0 10.0 15.0 0.0 10.0 20.0 30.0
R² = 0.0059 Barrel Height (ft) R² = 0.5926 Max Fill Depth (ft)
P-value > 0.05 P-value = 2.29E-8
FIGURE 4.45. 2003 ERA STATISTICAL ANALYSIS FOR SIGNIFICANT INDEPEDNET VARIABLES

TxDOT 0-5849 118

 3. Differentiated Inventory Ratings by Significant Culvert Variables

The lowest fill culverts in this era have more fill than the high fill culverts in the previous
design eras. For culverts with between seven feet and sixteen feet of fill, Figure 4.46 shows
that it is too close to decide if the first hypothesis is supported. The CULV-5 model may
produce slightly higher load ratings than the RISA-2D with springs. The lowest stiffness
RISA-2D with LEFE model produces even lower ratings. Never the less, the second and
third hypotheses are partially supported by higher ratings in the medium and high quality
RISA-2D with LEFE analyses.

240
220
inventory rating, IR (HS-tons)

200
180
160
140
120
100
80
60
40
20
0
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
SPG SPG SPG LEFE LEFE LEFE
k=75pci k=150pci k=250pci E=8ksi E=20ksi E=36ksi
analytical models by increasing sophistication

FIGURE 4.46. 2003 ERA INVENTORY RATINGS: 7' TO 16' OF FILL

For twenty feet of fill, the trend is much the same. Figure 4.47 shows that CULV-5
produces marginally higher ratings than RISA-2D with springs. RISA-2D with LEFE and
low soil stiffness produces the lowest ratings, but RISA-2D with LEFE and the medium and
high soil stiffnesses produce ratings higher than the other models. The first hypothesis is
barely unsupported, while the second and third hypotheses are partially supported.

220
200
inventory rating, IR (HS-tons)

180
160
140
120
100
80
60
40
20
0
-20
-40
-60
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
SPG SPG SPG LEFE LEFE LEFE
k=75pci k=150pci k=250pci E=8ksi E=20ksi E=36ksi
analytical models by increasing sophistication

FIGURE 4.47. 2003 ERA INVENTORY RATINGS: 20' OF FILL

TxDOT 0-5849 119

 For the deepest fill culverts, Figure 4.48 shows that the first hypothesis is decidedly
unsupported. The second and third hypotheses are again partially supported with higher load
ratings for the medium and high stiffness soils.

220
200
inventory rating, IR (HS-tons)

180
160
140
120
100
80
60
40
20
0
-20
-40
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
SPG SPG SPG LEFE LEFE LEFE
k=75pci k=150pci k=250pci E=8ksi E=20ksi E=36ksi
analytical models by increasing sophistication

FIGURE 4.48. 2003 ERA INVENTORY RATINGS: 23' OF FILL

4. Controlling Failure Mode

Figure 4.49 shows that the split between shear controlled failure and moment controlled
failure is nearly even. Having shear control so often is unnerving.

RISA-2D LEFE E=36ksi 11 10

RISA-2D LEFE E=20ksi 12 9

RISA-2D LEFE E=8ksi 15 6

moment
RISA-2D SPG k=250pci 10 11
shear
RISA-2D SPG k=150pci 9 12 axial thrust

RISA-2D SPG k=75pci 9 12

CULV-5 4 9

0 5 10 15 20 25

FIGURE 4.49. 2003 ERA CONTROLLING FAILURE MODE

TxDOT 0-5849 120

 5. Controlling Critical Section

Figure 4.50 shows that for the seven foot to sixteen foot subset of the 2003 culverts,
CULV-5 and RISA-2D with LEFE favor the bottom slab interior corner more than half the
time. The RISA-2D with spring model fails either in the wall mid-span or in the top slab
interior corners.

RISA-2D LEFE E=36ksi 1 0 2 0 5

RISA-2D LEFE E=20ksi 1 0 2 1 1 3 WBEC
RISA-2D LEFE E=8ksi 4 0 1 1 0 2 WEM
RISA-2D SPG k=250pci 0 3 5 0 TIC
RISA-2D SPG k=150pci 0 3 5 0 WTIC
RISA-2D SPG k=75pci 0 3 5 0 BEC
CULV-5 1 1 0 4 BIC
0 2 4 6 8

FIGURE 4.50. 2003 ERA CONTROLLING CRITICAL SECTIONS: 7' TO 16' OF FILL
For twenty feet of fill, Figure 4.51 shows that the CULV-5 model indentifies the bottom
slab interior corners. RISA-2D with springs evenly splits the controlling section between the
wall mid-spans and the top slab interior corners. The RISA-2D with LEFE models tend to
fail around the top interior corners, either in the wall or the top slab.

RISA-2D LEFE E=36ksi 2 0 1 2 1

RISA-2D LEFE E=20ksi 2 0 2 2 0
WBEC
RISA-2D LEFE E=8ksi 0 2 4 0
WEM
RISA-2D SPG k=250pci 0 3 3 0
TIC
RISA-2D SPG k=150pci 0 3 3 0
RISA-2D SPG k=75pci 0 WTIC
3 3 0
CULV-5 0 2 BIC

0 2 4 6

FIGURE 4.51. 2003 ERA CONTROLLING CRITICAL SECTIONS: 20' OF FILL

Figure 4.52 indicates that for the deepest fill culverts, the top interior corners are the
consistently weak points, though the wall mid-span often controls in the lower order models.

TxDOT 0-5849 121

RISA-2D LEFE E=36ksi 2 0 1 4 0
RISA-2D LEFE E=20ksi 1 0 1 5 0
WBEC
RISA-2D LEFE E=8ksi 1 0 1 5 0
WEM
RISA-2D SPG k=250pci 0 4 2 0 1
TIC
RISA-2D SPG k=150pci 0 3 3 0 1
RISA-2D SPG k=75pci 0 WTIC
3 3 0 1
CULV-5 0 2 0 3 BIC

0 2 4 6 8

FIGURE 4.52. 2003 ERA CONTROLLING CRITAL SECTIONS: 23' OF FILL

TxDOT 0-5849 122

 9. DISCUSSION

1. Evaluation of Hypotheses

1. By Era

In Table 4.1, the hypotheses are shown as supported (S), unsupported (US) or
indeterminate (I) relative to each era and the soil parameters selected for analysis. From this
table, several generalizations can be made about the hypotheses.
TABLE 4.1. EVALUATION OF HYPOTHESES BY ERA.
HYPOTHESIS 2:
HYPOTHESIS 1: HYPOTHESIS 3:
RISA-2D SPRG <
CULV-5 < RISA-2D SPRG CULV-5 < RISA-2D LEFE
ERA INVENTORY RATING FIGURE RISA-2D LEFE
SPRG SPRG SPRG LEFE LEFE LEFE LEFE LEFE LEFE
LOW MEDIUM HIGH LOW MEDIUM HIGH LOW MEDIUM HIGH

1938 FIGURE 4.8 S S S US S S US S S

1946 FIGURE 4.17 S S S US S S US S S

1958
FIGURE 4.26 I I I I S S I S S
36KSI

1958
FIGURE 4.35 I I I S S S S S S
60KSI

2003 FIGURE 4.44 I I I US S S US S S

The first hypothesis stated that RISA-2D with springs would produce higher ratings than
CULV-5. In practice, this appears to be supported only by the oldest culverts designed in the
1938 and 1946 era. For more recent designs, no definitive trend exist to suggest that RISA-
2D with springs would load rate a culvert higher or lower than CULV-5. To summarize, for
culverts designed before 1958, RISA-2D with springs will probably be a slight improvement
over CULV-5, but for culverts designed in 1958 or later, both programs produce similar
ratings.
It is also interesting to note that the varying levels of soil stiffness in the RISA-2D with
springs models do not differ greatly in the load rating they produce. A more in-depth
parametric analysis would be helpful in confirming this, but it appears that generally
speaking the spring stiffness does not appreciably influence the load rating. In essence, the
results suggest that CULV-5 and RISA-2D with varying soil spring stiffnesses really
represent a single level of reliability.
The second and third hypotheses stated that RISA-2D with LEFE would produce higher
load ratings than RISA-2D with springs and CULV-5 respectively. Clearly, for medium to
high modulus of soil elasticity, the hypothesis is supported. However, for low modulus
values, the hypotheses are unsupported.
For RISA-2D with LEFE, the load rating is highly sensitive to the modulus of elasticity
of the soil. On one hand, this can be very helpful. If a culvert fails to rate using simpler
models and the culvert has excellent quality backfill, the RISA-2D with LEFE load rating

TxDOT 0-5849 123

may show that the culvert rates acceptably. But it also means that for low quality fill, the
RISA-2D with LEFE model may show far worse performance than the other models. The
sensitivity to modulus also means that the soil modulus should be determined using field
testing. A textbook value will not be sufficient to accurately load rate the culvert.
The bottom line when looking at model performance by era is simple. First, RISA-2D
with springs does not produce higher load ratings often enough to be preferred over CULV-5.
Second, RISA-2D with LEFE can greatly improve the load rating, if and only if the soil
modulus of elasticity is high enough.
2. By Depth of Fill

Depth of fill clearly impacts load ratings, so this must be taken into account when
evaluating trends. When load ratings are differentiated by depth of fill, the support or lack of
support for the first hypothesis may vary. The relationships between CULV-5 and RISA-2D
with spring differ from the general conclusions.
1. Low Depth of Fill: Direct Traffic to Two Feet

For culverts with two feet or less of fill, Table 4.2 shows the extent to which the
hypotheses were supported, unsupported or indefinite.
TABLE 4.2. EVALUATION OF HYPOTHESES FOR LOW FILL CULVERTS
HYPOTHESIS 2:
HYPOTHESIS 1: HYPOTHESIS 3:
RISA-2D SPRG <
INVENTORY CULV-5 < RISA-2D SPRG CULV-5 < RISA-2D LEFE
ERA RISA-2D LEFE
RATING FIGURE
SPRG SPRG SPRG LEFE LEFE LEFE LEFE LEFE LEFE
LOW MEDIUM HIGH LOW MEDIUM HIGH LOW MEDIUM HIGH

1938 Figure 4.10 US US I US S S US S S

1946 Figure 4.19 I I I US S S I I I

1958
Figure 4.28 S S S US S S I S S
36KSI

1958
Figure 4.37 S S S S S S S S S
60KSI

By era, it has already been seen that the difference between load ratings calculated using
CULV-5 were not predictably different for the ratings determined using RISA-2D with
springs. However, for low fill heights, RISA-2D with springs can produce higher load
ratings than CULV-5. For low fill heights, the first hypothesis is supported.
The second and third hypotheses are supported in the same manner in the low depth of
fills as they are in the population at large. If the modulus of elasticity is high enough, RISA-
2D with LEFE produces higher ratings than RISA-2D with springs or CULV-5.

TxDOT 0-5849 124

 2. Medium Depth of Fill Culverts: Three to Six Feet

Table 4.3 shows the evaluation of the hypotheses for medium fill culverts.
TABLE 4.3. EVALUATION OF HYPOTHESES FOR MEDIUM FILL CULVERTS
HYPOTHESIS 2:
HYPOTHESIS 1: HYPOTHESIS 3:
RISA-2D SPRG <
INVENTORY CULV-5 < RISA-2D SPRG CULV-5 < RISA-2D LEFE
ERA RISA-2D LEFE
RATING FIGURE
SPRG SPRG SPRG LEFE LEFE LEFE LEFE LEFE LEFE
LOW MEDIUM HIGH LOW MEDIUM HIGH LOW MEDIUM HIGH

1938 Figure 4.11 S S S US S S US S S

1938 Figure 4.12 S S S US S S US S S

1946 Figure 4.20 I S S US I S US I S

1946 Figure 4.21 S S S US S S US S S

1958 36 Figure 4.33 I I S I S S I S S

1958 36 Figure 4.34 I I I I S S I S S

1958 60 Figure 4.38 I I I S S S S S S

1958 60 Figure 4.39 I I I I S S I S S

Medium fill culverts support the hypotheses in exactly the same manner as the population
at large. The first hypothesis is a split almost evenly along the era division. For the older
culverts, RISA-2D with springs produces higher load ratings than CULV-5. For newer
culverts RISA-2D and CULV-5 produce approximately the same load ratings. The second
and third hypotheses are again only supported for medium to high soil stiffnesses. With low
soil stiffness, RISA-2D with LEFE will not produce load ratings higher than RISA-2D with
springs or CULV-5.

TxDOT 0-5849 125

 3. High Depth of Fill Culverts: Seven to Twenty-Three Feet

Table 4.4 shows the hypotheses as supported, unsupported or indefinitely supported for
the high fill depth culverts.
TABLE 4.4. EVALUATION OF HYPOTHESES FOR HIGH FILL CULVERTS
HYPOTHESIS 2:
HYPOTHESIS 1: HYPOTHESIS 3:
RISA-2D SPRG <
INVENTORY CULV-5 < RISA-2D SPRG CULV-5 < RISA-2D LEFE
ERA RISA-2D LEFE
RATING FIGURE
SPRG SPRG SPRG LEFE LEFE LEFE LEFE LEFE LEFE
LOW MEDIUM HIGH LOW MEDIUM HIGH LOW MEDIUM HIGH

1938 Figure 4.13 I I I I S S I S S

2003 Figure 4.50 I I I US S S US S S

2003 Figure 4.51 I I I US S S US S S

2003 Figure 4.52 US US US US S S US S S

In this case, the first hypothesis is neither supported nor denied in the high fill culverts.
Generally speaking CULV-5 and RISA-2D with springs produce the same ratings. However,
for the deepest fill culverts, it is more likely that CULV-5 will produce the higher, less
conservative load rating than RISA-2D with springs. The second and third hypotheses react
in exactly the same way as before: the higher quality soil modulus produce higher ratings
than CULV-5 and RISA-2D with springs, but the lowest soil modulus fails to raise the load
rating.
It should be noted that the hypothesis evaluations based on depth of fill and those based
on era may be identifying the same trend. Era and depth of fill are not truly independent of
each other. Most of the high fill culverts are also 2003 design era. The question arises as to
whether the models are providing different ratings due to design philosophy or to depth of
fill. This study is not able to accurately answer that question.

TxDOT 0-5849 126

 3. Concerns about Findings

An issue of great concern arises from the findings of this study. Namely, if a more
sophisticated model, which is assumed to be more reliable, produces lower load ratings than
a less-sophisticated, less-reliable model, are the less sophisticated models unconservative? It
is the responsibility of every structural engineer to design and maintain safe structures.
Typically, this is accomplished by incorporating an intentional bias to overdesigning the
structure. But what if the tools used are not as conservative as originally believed?
The most disturbing finding of this research is that soils with low stiffness in a finite-
element, soil-structure model create worse loading conditions than the AASHTO loadings.
The soil stiffness must be reliable. Though the analytical model suggests this might be the
case, there are several factors not considered by the scope of this thesis. For example, how
many culverts are actually backfilled with extremely low grade material? Most culverts
probably have fair to excellent quality backfill, otherwise more culverts would be found to be
experiencing structural distress. There is also the potential that a uniform stiffness soil mass
does not accurately model soil behavior. Perhaps soil stiffness that increases with depth, as it
does in reality, would actually produce the amount of reliability that is assumed within a two-
dimensional, linear-elastic, soil-structural model. Needless to say, there are several concerns
that will need to be addressed in future study on this issue.

TxDOT 0-5849 127

 2. Evaluations of Failure Modes

Three different failure modes – moment, shear or thrust – could have controlled each load
rating. Typically in culvert and slab design, moment is assumed to be the primary concern.
This study shows that more than three-fourths of the time, this is the case. However, shear
does control occasionally.
Shear controlled failures are unexpected and, in many ways, undesired. Concrete box
culverts, like most concrete slab structures, are designed without shear reinforcing. Shear is
only resisted by concrete strength. While moment failures in reinforced concrete are, by
design, ductile failures, shear failures in concrete without shear reinforcement tend to be
brittle, rapidly forming failures. This means that a culvert that fails in shear will fail
suddenly and without warning. Though this is unnerving, it is comforting to realize that in
the minority of culverts that are controlled by shear in the sampled designs, the load ratings
are well above the required HS-20 level. So, at least for the sample of TxDOT culvert
designs, shear may control from time to time, but it should not be a problem in actuality.
Also noteworthy are the classes of culverts that fail in shear. The culverts which are most
likely to fail in shear are the 2003 design era and the 1958 design era with 60ksi steel. Both
of the design eras use higher strength reinforcing steel. High strength reinforcing steel
creates high moment capacity in relatively thin slabs. Meanwhile, the reinforcing steel does
nothing for shear capacity. The only way to increase shear capacity in a slab structure is to
increase the slab thickness. The primary reason that the newer culverts have a tendency to
fail in shear is because the moment capacity created by better quality materials is much
higher than the required capacity, but the shear capacity is relatively unimproved. It is a
matter of design philosophy and material qualities.
The models used have only a slight effect on failure mode. Generally, all models have
the same probability of finding shear as the controlling mode as they do of finding moment
as the controlling mode. The controlling mode is much more sensitive to culvert slab design.
However, in the case of the 1958 era with 60ksi steel, a trend does appear where the higher
stiffness soil in the RISA-2D with LEFE models are more likely to show shear as the
controlling mechanism. This is because the RISA-2D with LEFE model automatically
predicts the effect called soil arching.
The soil arching effect is automatically determined in the RISA-2D with LEFE model.
This phenomenon may account for shear controlling the load rating more often. The increase
shear controlled failures with soil stiffness is primarily due to the fact that the stiffer soils
redistribute the load more. The mid-span moment directly decreases the mid-span moment
load rating, causing the relatively unaffected corner shear ratings to control.

TxDOT 0-5849 128

 3. Evaluation of Controlling Critical Sections

Table 4.5 identifies the most popular controlling critical section for each model, divided
by era and depth of fill.
TABLE 4.5. PROMINENT CONTROLLING CRITICAL SECTIONS
CRITICAL RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
ERA SECTION CULV-5 SPRG SPRG SPRG LEFE LEFE LEFE
FIGURE LOW MEDIUM HIGH LOW MEDIUM HIGH

1938 Figure 4.15 WBIC WBEC WBEC WBEC WBEC WTIC WTIC

WTEC/ WTEC/
1938 Figure 4.15 WBEC WTEC WBEC WBEC WTIC
WBEC WBEC

1938 Figure 4.15 WBEC WBEC WBEC WBEC WBEC WBEC WBEC

WEM
1938 Figure 4.16 WBEC WEM WEM WBEC WBEC WBEC
BEM

WTIC/ WTIC/ WTIC/

1946 Figure 4.23 WTIC WTIC WTIC
WTEC WTEC WTEC

WEM/
1946 Figure 4.24 WEM WEM WEM WBEC WBEC WBEC
WBEC

WTEC/ WTEC/
1946 Figure 4.25 WTEC WTEC WBEC WBEC WBEC
WBEC WBEC

1958 36 Figure 4.32 BEM WBEC WTEC WTEC BEC WTIC WTIC

WBEC/
1958 36 Figure 4.33 BEM TEM TEM WBEC WBEC WTEC
TEM
BEM/
1958 36 Figure 4.34 WBEC WEM WEM WBEC WBEC WBEC
WBEC

1958 60 Figure 4.41 BEM WTIC WTIC WTIC WTIC WTIC TIC

1958 60 Figure 4.42 TEM TEM TEM WBEC TIC TIC

1958 60 Figure 4.43 BIC TIC WEM WEM WBEC TIC TIC

2003 Figure 4.50 BIC TIC TIC TIC WBEC BIC BIC

TIC/ TIC/ TIC/

2003 Figure 4.51 BIC WTIC
WEM WEM WEM

TIC/ TIC/
2003 Figure 4.52 BIC WEM WTIC WTIC WTIC
WEM WEM

Some trends exist. The easiest trend to identify is that almost every time the controlling
critical section changes from model to model. In real culverts, only one critical section
controls. The actual controlling critical section should match the controlling critical section
from the most reliable model. This could act as a qualitative reliability check on analytical
models against actual field behavior. Comparing the controlling critical sections from field
testing to the critical sections from each model would provide one check of reliability of the
type of structural response, beyond simply comparing the magnitudes of the demands or
ratings.

TxDOT 0-5849 129

 Another clear trend shows that the most popular critical sections are located at the wall
corners. Particularly for RISA-2D with LEFE, the corners control almost exclusively. This
is most likely due to soil arching which decreases the center span moments at the expense of
corner moments and shears.
Another noteworthy trend occurs in tall, deep fill culverts. In these cases, in the CULV-5
and RISA-2D with springs models, the wall mid-span has an increased probability of
controlling. Moment failures in the wall mid-spans are most strongly affected by lateral
earth pressure assumptions. The sections may only control because of TxDOT’s less
conservative lateral earth pressure design assumptions.

TxDOT 0-5849 130

 10. CONCLUSIONS

1. Reliability and Modeling Sophistication

In this study, reliability has been defined as the coherence between predicted and
observed behavior. For culvert load rating, reliability is a function of the actual behavior of
the soil-structural system, accepted conservatism to account for design and construction
uncertainties, and model sophistication related to load-soil-structure interaction. Within the
results, the balance of actual behavior and accepted conservatism is assumed to be constant.
The conclusions focus on model sophistication as a way to identify load-soil-structure
interaction.
The hypotheses are intended to test the portion of reliability associated with load-soil-
structure interaction as detected by model sophistication. Because uncertainty and over-
conservatism are removed with each increase in modeling sophistication, it was expected that
the load ratings would generally increase with increases in modeling sophistication. This
research shows that though this is often the case, there are occasions where the hypotheses
are not entirely supported.
2. Principle Findings

1. RISA-2D with Spring Supports will produce a higher rating than CULV-5.

The first hypothesis predicted that RISA-2D with springs would produce higher load
ratings than CULV-5. Generally speaking, this hypothesis was neither supported nor denied.
RISA-2D with springs did not significantly change when the modulus of subgrade reaction
increased. This suggests that CULV-5 and RISA-2D with all three spring stiffness represent
the same level of reliability.
However, the hypothesis was found to be supported for older and shallow fill depth
culverts. For these depths, spring supports are thought to provide more reliable and higher
load ratings.
2. RISA-2D with Linear-Elastic Finite-elements (LEFE) will produce a higher
rating than RISA-2D with Springs.

The second hypothesis stated that RISA-2D with LEFE would produce higher load
ratings than RISA-2D with springs. This hypothesis was only partially supported. RISA-2D
with LEFE proved to be highly sensitive to the modulus of elasticity used to model the soil.
For medium to high modulus values, i.e. stiffer soils, the hypothesis was easily supported.
However, for the least stiff soil model with the lowest modulus values, RISA-2D with LEFE
rarely produced load ratings at or above the level of RISA-2D with springs.
3. RISA-2D with LEFE will produce a higher rating than CULV-5.

The third hypothesis claimed that RISA-2D with LEFE would produce higher load
ratings than CULV-5. Because CULV-5 and RISA-2D were found to produce similar load
ratings, it should be unsurprising to find that the third hypothesis was partially supported in

TxDOT 0-5849 131

the same manner as the second hypothesis. For medium and high modulus values, RISA-2D
with LEFE easily outclassed CULV-5. However, the low stiffness RISA-2D with LEFE
model rarely produced load ratings as high as CULV-5.

3. Limitations of Research

Though this portion of the report clearly met its operational statement, one limitation
should be noted. Though design era and depth of fill were identified as significant variables,
the relationship between these variables and the analytical models were not completely
investigated.
For example, a general trend showed that deep fill and new culverts tend to produce the
same load ratings between RISA-2D with springs and CULV-5. However, depth of fill and
design era are not totally independent. Are the trends really applicable to depth of fill or
design era? The data from this study does not answer that question.
Another example arises when the controlling critical section and failure mode change and
move between models. Is this a function of modeling methodology or design methodology?
Is the model placing more stress in certain parts of the culvert or were those parts under-
designed? In all likelihood the effect is created by a combination of the two, but this study
fails to indentify the distinction.

TxDOT 0-5849 132

 5. PARAMETRIC ANALYSIS OF SIX CULVERT VARIABLES

1. OVERVIEW

This chapter presents research findings from a parametric study designed to evaluate the
influence of selected culvert variables on inventory rating values for a sample of seven
culvert designs taken from the 100 TxDOT culvert designs discussed in Chapter 4. The
research team performed this work with the following objectives in mind:

• Evaluate the sensitivity of inventory load ratings over an expected range of

parameter values.
• Establish the desired level of precision necessary for the various parameters.
• Further explore trends identified when load rating the 100 culvert designs.

Table 5.1 identifies the variables of interest, the models to which they apply, the range of
values considered in the parametric analyses, the untested variable assumptions, and the
desired results.
TABLE 5.1. PARAMETRIC ANALYSIS TEST MATRIX

Variable Model Range Assumptions Conclusive result

Modulus of
LEP = 60/30 psi;
subgrade 2 75 pci 150 pci 250 pci ΔIR < 10%
DOF = max
reaction, k
Poisson's ratio, Uniform E = 20ksi;
3 0.1 0.3 0.5 ΔIR < 10%
ν DOF = max

Multibarrel Uniform E = 20ksi;

3 5 vs 4 6 vs 4 7 vs 4 8 vs 4 ΔIR < 10%
effects DOF = max; ν = 0.3

Lateral earth Typ. K;

1&2 40 pcf 60 pcf 80 pcf 100 pcf Record ΔIR
pressures, lep DOF = max

Modulus of Typ. Ν;
3 8 ksi 20 ksi 36 ksi Ranges where ΔIR < 10%
elasticity, e DOF = max

Average typ. values for

Depth of fill, dof 1, 2 & 3 Min Mid Max Max + Record ΔIR
each level
Abbreviations note:
pci – pounds per square inch/inch (cubic inch)
LEP – lateral earth pressure
DOF – depth of fill
ΔIR – change in inventory rating
ksi – kips per square inch
vs – versus
pcf – pounds per cubic foot
Typ – typical
min – minimum
mid – mid-range
max – maximum

TxDOT 0-5849 133

 2. SAMPLE CULVERT DESIGNS

The parametric study was tested on seven culvert designs selected from the 100 culvert
designs previously tested. These designs represent a “good” and a “bad” design from the
1938 era, a culvert from each fill range from the 1958 era, and two deep-fill culverts from the
2003 era. Table 5.2 shows the distribution of culvert variables for the sample.
TABLE 5.2. PARAMETRIC SAMPLE

MAX FILL HEIGHT, FT.

BARREL LENGTH, FT.

BARREL HEIGHT, FT.

FILL RANGE, FT.

# OF SPANS

SKEW (S/H)

ORG. YEAR
SHEET

ID
MBC-5-34 5 9.0 8.0 1.13 0.0 0.0 1934 MBC-5-34 5 9X8W0
MBC-1-44-F 2 5.0 4.0 1.25 0.0 - 6.0 6.0 1935 MBC-1-44-F 2 5X4W6
MC9-2 2 9.0 9.0 1.00 2.08 - 4.0 4.0 1958 MC9-2 2 9X9W4
MC6-2 6 6.0 3.0 2.00 4.08 - 6.0 6.0 1958 MC6-2 6 6X3W6
MC7-1 3 7.0 6.0 1.17 0.0 - 2.0 2.0 1958 MC7-1 3 7X6W2
MC-10-20 5 10.0 10.0 1.00 2.0 - 20.0 20.0 2003 MC-10-20 5 10X10W20
MC-7-16 4 7.0 4.0 1.75 2.0 - 16.0 16.0 2003 MC-7-16 4 7X4W16

TxDOT 0-5849 134

 3. RESULTS

1. Modulus of Subgrade Reaction, k

1. Parametric Sample

The modulus of subgrade reaction (k) estimates the support of the soil layer below a rigid
concrete slab; e.g., the bottom slab of the culvert.
For a sample of 7 culvert designs, the inventory ratings were determined using three
different values of the modulus of subgrade reaction. These values were chosen as
representative for low quality (75 pci), medium quality (150 pci) and high quality (250 pci)
soils.
2. Results

Table 5.4 presents the calculated values of the inventory rating and Figure 5.1 presents
plot of inventory rating with respect the modulus of subgrade reaction. As can be clearly seen
in Figure 5.1, the change in rating with respect to subgrade reaction is very small. The
percent difference between the inventory ratings is generally small as seen in Table 5.3.
Some outliers occur when the magnitude of the rating is small, but this represents sensitivity
to the percent difference calculation (smaller magnitudes result in larger percent differences)
rather than sensitivity in the load rating process. For these cases the inventory rating does not
change by more than HS-2 between the selected subgrade modulus values. Clearly there is
very little sensitivity in the load rating to modulus of subgrade reaction as a parameter.
TABLE 5.3. PERCENT DIFFERENCE IN INVENTORY RATINGS.
ID ΔIR(K) L TO H ΔIR(K) L TO M ΔIR(K) M TO H
MC-7-16 4 7X4W16 1% 0% 1%
MC6-2 6 6X3W6 2% 1% 1%
MBC-1-44-F 2 5X4W6 8% 3% 5%
MC-10-20 5 10X10W20 9% 4% 5%
MC7-1 3 7X6W2 21% 15% 7%
MC9-2 2 9X9W4 21% 10% 11%
MBC-5-34 5 9X8W0 27% 14% 15%

TxDOT 0-5849 135

 TABLE 5.4. THE LOAD RATING FOR 7 REPRESENTATIVE CULVERT DESIGNS WITH THREE VARYING MODULUS OF SUBGRADE REACTION, K (PSI).
ID RISA-2D SPG – L – K = 75 PSI RISA-2D SPG – M – K = 150 PSI RISA-2D SPG - H– K = 250 PSI
MBC-5-34 5 9X8W0 7.1 11.8 WBEC M MIN RLL 8.2 13.7 WBEC M MIN RLL 9.6 16.1 WBEC M MIN RLL
MBC-1-44-F 2 5X4W6 39.4 65.7 WBEC M MIN RLL 40.6 67.8 WBEC M MIN RLL 42.6 71.1 WBEC M MIN RLL
MC9-2 2 9X9W4 -3.7 -6.2 WEM M MAX TL -4.1 -6.9 WEM M MAX TL -4.7 -7.8 WEM M MAX TL
MC6-2 6 6X3W6 15.8 26.3 TEM M MAX RLL 15.9 26.6 TEM M MAX RLL 16.1 26.9 TEM M MAX RLL
MC7-1 3 7X6W2 5.8 9.7 BEM M MAX RLL 6.8 11.3 BEM M MAX RLL 7.3 12.2 WTIC M MIN RLL
MC-10-20 5 10X10W20 -30.1 -50.3 WEM M MIN TL -31.4 -52.5 WEM M MIN TL -33.1 -55.2 WEM M MIN TL
MC-7-16 4 7X4W16 58.3 97.4 TIC V MIN RLL 58.3 97.3 TIC V MIN RLL 58.7 98.0 TIC V MIN RLL

TxDOT 0-5849 136

 80.0

60.0
Inventory Rating, IR (HS-)

40.0 MC-7-16 4 7x4w16

MC6-2 6 6x3w6
20.0 MBC-1-44-F 2 5x4w6
MC-10-20 5 10x10w20
0.0 MC7-1 3 7x6w2
MC9-2 2 9x9w4
-20.0 MBC-5-34 5 9x8w0

-40.0
50 100 150 200 250
modulus of subgrade reaction, k (psi)

FIGURE 5.1. MODULUS OF SUBGRADE REACTION VS. THE INVENTORY RATING FOR SEVEN REPRESENTATIVE CULVERT DESIGNS.

TxDOT 0-5849 137

 3. The Sample of 100 Culvert Designs:

Subgrade modulus data are available from the full sample of 100 culvert designs because
that task determined load ratings for each of the three subgrade modulus values. For this
larger sample, the percent differences in the inventory rating values found that for the
expected variation of the modulus of subgrade reaction, the percent difference between the
inventory rating for the lowest k and the highest k may be larger than desired. Table 5.5
shows the number of culvert designs that satisfy the shown criteria. Again, this analysis is
sensitive to the magnitude of the ratings with smaller magnitudes resulting in larger percent
differences.
TABLE 5.5. NUMBER OF CULVERT DESIGNS OUT OF A SAMPLE OF 100 REPRESENTATIVE DESIGNS WHICH MEET SPECIFIC CRITERIA FOR
THE MAXIMUM PERCENT DIFFERENCE IN THE INVENTORY RATING.
CRITERIA ΔIR(K) L TO H ΔIR(K) L TO M ΔIR(K) M TO H
ΔIR(K) <10% 43 66 80
ΔIR(K) <20% 76 89 94
ΔIR(K) <30% 89 95 96
ΔIR(K) <40% 93 96 98

4. Summary and Conclusion

The sensitivity of the inventory rating to the modulus of subgrade reaction in the Level 2
model is small. The slope of rating versus the modulus of subgrade reaction is less than 0.02
HS-tons/pci. The percent difference between the lowest k (75 pci) and the highest k (250
pci) is less than 20% for 76 of the 100 sample culverts, and in most of the remaining culverts
the difference was less than HS-2. This indicates that while the sensitivity of the inventory
rating to the modulus of subgrade reaction is low, the three values for the modulus are
analytically appropriate for keeping the error under control.

TxDOT 0-5849 139

 2. Poisson’s Ratio

1. Parametric Sample

Poisson's ratio (ν) is the ratio, when a sample object is stretched, of the contraction or
transverse strain (perpendicular to the applied load), to the extension or axial strain (in the
direction of the applied load).
For a sample of 7 culvert designs, inventory ratings were determined using three different
values of the Poisson’s ratio and an “average” soil modulus of elasticity of 20 ksi. Poisson’s
ratio values of 0.5, 0.3, and 0.1 were chosen for analysis.
2. Results

Table 5.7 shows the calculated inventory rating values and Figure 5.2 presents a plot of
the inventory rating with respect the Poisson’s ratio. As can be clearly seen in Figure 5.2, the
slope of the change in rating with respect to Poisson’s ratio is typically very small. The
percent difference between the inventory ratings is generally small as seen in Table 5.6.
TABLE 5.6. PERCENT DIFFERENCE IN INVENTORY RATINGS.
ID ΔIR(Ν) L TO H ΔIR(Ν) L TO M ΔIR(Ν) M TO H
MC-10-20 5 10X10W20 232% 0% 232%
MC6-2 6 6X3W6 12% 7% 5%
MC-7-16 4 7X4W16 5% 1% 4%
MBC-1-44-F 2 5X4W6 6% 1% 5%
MC7-1 3 7X6W2 1% 2% 3%
MBC-5-34 5 9X8W0 12% 5% 8%
MC9-2 2 9X9W4 568% 203% 255%

One outlier, MC-10-20 5 10x10w20, occurs because for very high Poisson’s ratio (low
quality soil) in this tall, deeply buried culvert, the critical section moves to the midspan of the
exterior wall. The rating then becomes negative.
The other outlier, MC9-2 2 9x9w4, occurs because the magnitude of the rating is small.
Again, this is sensitivity to the percent difference calculation (smaller magnitudes result in
larger percent differences) rather than sensitivity in the load rating process.
3. Conclusion

The inventory rating is not very sensitive to the Poisson’s ratio, generally exhibiting less
than a 10% change across the range of Poisson’s ratios. For most cases, a typical value for
Poisson’s ratio of 0.3 provides suitable results. One exception occurs for tall culverts under
significant depth of fill and which are backfilled with very poor materials; i.e., highly plastic
clays. For this case, it is appropriate to use a Poisson’s Ratio of 0.5. But unless the soil is
very poor and the culvert is tall and deeply buried, an average value for the Poisson’s ratio of
0.3 is appropriate

TxDOT 0-5849 140

 TABLE 5.7. THE LOAD RATING FOR 7 REPRESENTATIVE CULVERT DESIGNS WITH THREE POISSON’S RATIOS, Ν.
ID RISA-2D LEFE - Ν = .1 RISA-2D LEFE - Ν = .3 RISA-2D LEFE - Ν = .5
MC-10-20 5 71. 119. MA 71. 119. MA - - MA
WTIC M WTIC M WEM M
10X10W20 5 4 X 7 6 X 54.2 90.5 X
50. MA 47. MA MA
MC6-2 6 6X3W6 84.7 TEM M 79.3 TEM M 45.4 75.8 TEM M
7 X 5 X X
36. MA 36. MA
MC-7-16 4 7X4W16 60.6 BIC V 61.0 BIC V 38.1 63.6 BIC V MIN
3 X 5 X
28. 28. WBE WBE
MBC-1-44-F 2 5X4W6 46.7 WBEC M MIN 47.1 M MIN 29.7 49.6 M MIN
0 2 C C
12. WTIC MA 12. MA
MC7-1 3 7X6W2 20.5 M 20.8 WTIC M 12.1 20.2 BEC M MIN
3 1 X 5 X
MA MA MA
MBC-5-34 5 9X8W0 6.9 11.6 WTIC M 7.3 12.2 WTIC M 7.9 13.2 WTIC M
X X X
WBE WBE
MC9-2 2 9X9W4 5.2 8.7 WBEC M MIN 1.7 2.9 M MIN -1.1 -1.9 M MIN
C C

TxDOT 0-5849 141

 80.0

60.0

40.0
Inventory Rating, IR (HS-)

MC-10-20 5 10x10w20
20.0
MC6-2 6 6x3w6
0.0 MC-7-16 4 7x4w16
MBC-1-44-F 2 5x4w6
-20.0
MC7-1 3 7x6w2
-40.0 MBC-5-34 5 9x8w0
MC9-2 2 9x9w4
-60.0

-80.0
0 0.1 0.2 0.3 0.4 0.5 0.6
Poisson's Ratio, ν

Figure 5.2. Poisson’s Ratio vs. the Inventory Rating for Seven Representative Culvert Designs.

TxDOT 0-5849 142

 3. Multibarrel Effects

1. Parametric Sample

One of the limitations in TxDOT’s CULV-5 program is that it cannot model culverts with
more than four barrels. When TxDOT engineers encounter a culvert with more than four
barrels, they analyze the culvert as if it only had four barrels, assuming this is a safe
assumption. The question being addressed for this aspect of the parametric study was
whether it is acceptable to model a 5+ barrel culvert as a having only 4 barrels.
For a sample of 7 culvert designs, the inventory ratings were determined using RISA-2D
with LEFE for culverts having 4, 5, 6, and 7 barrels. Analyses were based on the number of
barrels designed for on the appropriate design sheet. For all designs, no changes were made
to dimensions or reinforcing to account for additional boxes. Analyses used a Poisson’s ratio
of 0.3 and a modulus of elasticity of 20 ksi which are representative of medium quality soil.
2. Results

Table 5.9 presents the calculated inventory rating values and Figure 5.3 is a plot of the
inventory rating with respect the number of barrels. As can be clearly seen in Figure 5.3, the
slope of the change in rating with respect to the number of barrels is very small. The percent
difference between the inventory ratings is small and typically less than the 10% structural
tolerance, as seen in Table 5.3.
The trends show that the 4 barrel model generally produces the lowest and most
conservative rating. The only place where this trend does not hold true is for the culvert with
a load rating much larger than necessary.
The other outlier, MC9-2 2 9x9w4, occurs because the magnitude of the rating is small.
This represents sensitivity to the percent difference calculation (smaller magnitudes result in
larger percent differences) rather than sensitivity in the load rating process.
TABLE 5.8. PERCENT DIFFERENCE IN INVENTORY RATINGS.
ID 4 BARREL 5 BARREL 6 BARREL 7 BARREL ΔIR(K) 4 TO 5, 6, OR 7
MC-10-20 5 10X10W20 77.5 71.7 70.7 NA -10%
MC6-2 6 6X3W6 46.8 47.1 47.5 NA 2%
MC-7-16 4 7X4W16 36.5 38.3 38.3 NA 5%
MBC-1-44-F 2 5X4W6 26.7 27.5 28.6 29.1 8%
MC7-1 3 7X6W2 12.4 12.6 12.9 NA 4%
MBC-5-34 5 9X8W0 7.4 7.3 NA NA -2%
MC9-2 2 9X9W4 1.1 3.3 3.1 NA 64%

TxDOT 0-5849 143

 3. Conclusion

The sensitivity of the inventory rating to the number of culvert barrels included in the
model is small, less than a 10% change for 5, 6 and 7 barrel culverts modeled as a 4 barrel
culvert. Generally this is conservative assumption. The only exceptions occur when the
rating is very near zero or much greater than HS-20. This suggests it is acceptable to model
5+ barrel culverts with only 4 barrels. However, whenever possible it is preferable to model
all the barrels in the culvert.

TxDOT 0-5849 144

 TABLE 5.9.THE LOAD RATING FOR 7 REPRESENTATIVE CULVERT DESIGNS WITH INCREASING NUMBER OF BARRELS.
ID 4 BARREL 5 BARREL 6 BARREL 7 BARREL
MC-10-20 5 10X10W20 77.5 129.4 WTIC M 71.7 119.6 WTIC M 70.7 118.1 WTIC M NA NA NA NA
MC6-2 6 6X3W6 46.8 78.1 TEM M 47.1 78.6 TEM M 47.5 79.3 TEM M NA NA NA NA
MC-7-16 4 7X4W16 36.5 61.0 BIC V 38.3 63.9 BIC V 38.3 64.0 BIC V NA NA NA NA
MBC-1-44-F 2 5X4W6 26.7 44.5 WBEC M 27.5 46.0 WBEC M 28.6 47.7 WBEC M 29.1 48.5 WBEC M
MC7-1 3 7X6W2 12.4 20.6 WTIC M 12.6 21.1 WTIC M 12.9 21.6 WTIC M NA NA NA NA
MBC-5-34 5 9X8W0 7.4 12.4 WTIC M 7.3 12.2 WTIC M NA NA NA NA NA NA NA NA
MC9-2 2 9X9W4 1.1 1.9 WBEC M 3.3 5.5 WBEC M 3.1 5.2 WBEC M NA NA NA NA

TxDOT 0-5849 145

 90.0

80.0

70.0
Inventory Rating, IR (HS-)

60.0 MC-10-20 5 10x10w20

50.0 MC6-2 6 6x3w6

MC-7-16 4 7x4w16
40.0
MBC-1-44-F 2 5x4w6
30.0 MC7-1 3 7x6w2

20.0 MBC-5-34 5 9x8w0

MC9-2 2 9x9w4
10.0

0.0
3 4 5 6 7 8
number of barrels (#)

FIGURE 5.3. NUMBER OF BARRELS VS. THE INVENTORY RATING FORSEVEN REPRESENTATIVE CULVERT DESIGNS.

TxDOT 0-5849 146

 4. Lateral Earth Pressure, LEP

1. Parametric Sample

Lateral earth pressure (LEP) is the pressure soil exerts in the horizontal plane. When
calculating lateral earth pressures, engineers commonly assume the pressure distribution to
be triangular and calculate the magnitude of the pressure as if the soil were an “equivalent
fluid.” This calculation requires the equivalent fluid unit weight for the soil which in turn
depends on soil properties, the stress history of the soil, and the characteristics of the
application.
For a sample of 7 culvert designs, the inventory ratings were determined using lateral
earth pressure (equivalent fluid unit weight) values ranging from 40pcf to 100pcf, varied at
20pcf increments. CULV-5 and RISA-2D with springs apply the lateral earth pressures to
the culvert sidewalls, and these were the methods tested for this parameter. RISA-2D with
LEFE does not apply lateral earth pressures to the culvert as loads but instead calculates soil
stresses around the entire culvert subsurface regime.
2. Results

Table 5.11 shows the calculated inventory rating values. Figure 5.4 and Figure 5.5 are
plots of the inventory rating with respect the lateral earth pressure (equivalent fluid unit
weight) values. These plots show that unless the inventory rating becomes negative, the load
rating is not particularly sensitive to the lateral earth pressure value. Negative ratings occur
in tall culverts because the critical section moves to the exterior wall mid-spans.
Table 5.10 shows the change in the inventory rating and the percent change in the
inventory rating. It is clear that unless the inventory rating is very small or negative, the
lateral earth pressure does not significantly affect the load rating.
TABLE 5.10. CHANGE IN INVENTORY RATING AND PERCENT DIFFERENCE IN INVENTORY RATING.
RATING DATA CULV-5 RISA-2D
DIFFERENCE PERCENT DIFFERENCE PERCENT
ID
(HS-TONS) DIFFERENCE (HS-TONS) DIFFERENCE
MC-7-16 4 7X4W16 -5.1 -9% -5.0 -9%
MBC-1-44-F 2 5X4W6 1.2 4% 3.8 10%
MC6-2 6 6X3W6 -0.7 -6% -0.7 -5%
MBC-5-34 5 9X8W0 -0.5 -9% -0.6 -8%
MC7-1 3 7X6W2 1.0 25% 2.7 41%
MC-10-20 5 10X10W20 123.1 334% 78.3 -1015%
MC9-2 2 9X9W4 638.0 8119% 139.2 1222%

TxDOT 0-5849 147

 3. Conclusion

The sensitivity of the load rating to the lateral earth pressure is surprisingly small.
Lateral earth pressures appear to only affect tall sidewall culverts, and as such produce
negative load ratings where the magnitude of the rating does not have numerical significance
(failure under dead load). Because lateral earth pressures matter very little when determining
load ratings, it is logical to use the AASHTO requirement of 60pcf. The AASHTO values are
reasonable, approved, and provide liability protection.

TxDOT 0-5849 148

 TABLE 5.11. THE LOAD RATING FOR
SEVEN REPRESENTATIVE CULVERT DESIGNS WITH INCREASING LATERAL EARTH PRESSURE.
ID CULV-5 LEP = 40PSF CULV-5 LEP = 60PSF CULV-5 LEP = 80PSF CULV-5 LEP = 100PSF
MBC-5-34 5 9X8W0 5.7 9.6 BIC M RLL 5.9 9.9 BIC M RLL 6.1 10.2 BIC M RLL 6.3 10.5 BIC M RLL
MBC-1-44-F 2 5X4W6 28.4 47.4 WBEC M TL 27.9 46.6 WBEC M TL 27.1 45.3 WBEC M TL 27.1 45.3 WBEC M TL
MC9-2 2 9X9W4 7.9 13.1 BEM M RLL -2.9 -4.8 WEM M TL -13.3 -22.2 WEM M TL -630.2 -1051.9 WEM M TL
MC6-2 6 6X3W6 13.2 22.0 BEM M RLL 13.4 22.4 BEM M RLL 13.7 22.8 BEM M RLL 13.9 23.2 BEM M RLL
MC7-1 3 7X6W2 4.2 7.0 BEM M RLL 4.3 7.2 BEM M RLL 4.5 7.5 BEM M RLL 3.2 5.3 WEM M TL
MC-10-20 5 10X10W20 36.9 61.5 BIC V RLL -29.2 -48.7 WEM M TL -66.0 -110.1 WEM M TL -86.2 -143.9 WEM M TL
MC-7-16 4 7X4W16 53.5 89.3 BIC V RLL 55.2 92.1 BIC V TL 56.8 94.9 BIC V RLL 58.5 97.7 BIC V RLL

ID RISA-2D LEP = 40PSF RISA-2D LEP = 60PSF RISA-2D LEP = 80PSF RISA-2D LEP = 100PSF
MBC-5-34 5 9X8W0 7.6 12.7 WBEC M RLL 8.2 13.7 WBEC M RLL 8.0 13.3 WBEC M RLL 8.2 13.6 WBEC M RLL
MBC-1-44-F 2 5X4W6 39.4 65.8 WBEC M RLL 40.6 67.8 WBEC M RLL 41.8 69.8 WBEC M RLL 35.6 59.4 WEM M TL
MC9-2 2 9X9W4 11.4 19.0 TEM M RLL -4.1 -6.9 WEM M TL -14.7 -24.6 WEM M TL -127.8 -213.3 WEM M TL
MC6-2 6 6X3W6 15.7 26.2 TEM M RLL 15.9 26.6 TEM M RLL 16.2 27.0 TEM M RLL 16.4 27.4 TEM M RLL
MC7-1 3 7X6W2 6.6 11.1 BEM M RLL 6.8 11.3 BEM M RLL 6.3 10.6 BEC M TL 3.9 6.6 WEM M TL
MC-10-20 5 10X10W20 -7.7 -12.9 WBEC V TL -31.4 -52.5 WEM M TL -66.5 -111.0 WEM M TL -86.0 -143.6 WEM M TL
MC-7-16 4 7X4W16 56.7 94.6 TIC V RLL 58.3 97.3 TIC V RLL 60.0 100.1 TIC V RLL 61.7 102.9 TIC V RLL

TxDOT 0-5849 149

 80.0

60.0

40.0

20.0
Inventory Rating, IR (HS-)

MC-7-16 4 7x4w16
0.0
MBC-1-44-F 2 5x4w6
MC6-2 6 6x3w6
-20.0 MBC-5-34 5 9x8w0
MC7-1 3 7x6w2

-40.0 MC-10-20 5 10x10w20

MC9-2 2 9x9w4

-60.0

-80.0

-100.0
30 40 50 60 70 80 90 100 110
lateral earth pressure (psf)

FIGURE 5.4. LATERAL EARTH PRESSURE VS. THE INVENTORY RATING FOR SEVEN REPRESENTATIVE CULVERT DESIGNS USING CULV-5.

TxDOT 0-5849 150

 80.0

60.0

40.0

20.0
Inventory Rating, IR (HS-)

MC-7-16 4 7x4w16
0.0
MBC-1-44-F 2 5x4w6
MC6-2 6 6x3w6
-20.0 MBC-5-34 5 9x8w0
MC7-1 3 7x6w2

-40.0 MC-10-20 5 10x10w20

MC9-2 2 9x9w4

-60.0

-80.0

-100.0
30 40 50 60 70 80 90 100 110
lateral earth pressure (psf)

FIGURE 5.5. LATERAL EARTH PRESSURE VS. INVENTORY RATING FOR SEVEN REPRESENTATIVE CULVERT DESIGNS USING RISA-2D.MODULUS OF ELASTICITY, E

TxDOT 0-5849 151

 1. Parametric Sample

The modulus of elasticity for a material is basically the slope of its stress-strain plot
within the elastic range. Sometimes called Young's modulus, the elastic modulus (E) can be
determined for any solid material and represents a constant ratio of stress and strain (a
stiffness).
For a sample of 7 culvert designs, the inventory ratings were determined using a modulus
of elasticity ranging from 4 ksi to 40 ksi by intervals of 4 ksi. This parameter only applies to
the RISA-2D with LEFE model, as it uses a linear-elastic constitutive soil model. Poisson’s
ratio of 0.3 was selected for all analyses.
2. Results

Table 5.13 presents the calculated inventory rating values. Figure 5.6 presents a plot of
the inventory rating with respect the modulus of elasticity.
The clearest trend observed is that greater depths of fill show greater sensitivity to
modulus of elasticity. This means that for deeper culverts it is more important to accurately
identify the actual modulus.
Figure 5.6 shows that the relationship between modulus and inventory rating is not linear.
However, linear approximations of the slopes between measured points can be used to
estimate the required precision in modulus to determine the inventory rating to within HS-2
(10% of the design load, HS-20). Using this approach, Table 5.12 shows the average and
maximum slope values and the average and worst-case tolerances required for each of the
analyzed culverts. This analysis indicates that for inventory rating calculations to be reliable,
the modulus of elasticity must be known with a high degree of precision.

TxDOT 0-5849 152

 TABLE 5.12. REQUIRED CONFIDENCE FOR MODULUS OF ELASTICITY, E.
AVERAGE SLOPE MAX SLOPE AVERAGE REQUIRED WORST-CASE
ID
(HS-TONS/KSI) (HS-TONS/KSI) CONFIDENCE (KSI) REQUIRED CONFIDENCE (KSI)
MC-7-16 4 7X4W16 5.9 9.3 0.34 0.21
MC-10-20 5 10X10W20 7.4 11.4 0.27 0.18
MC6-2 6 6X3W6 1.3 3.1 1.49 0.64
MBC-1-44-F 2 5X4W6 1.2 1.3 1.70 1.54
MC9-2 2 9X9W4 1.2 1.7 1.60 1.17
MC7-1 3 7X6W2 0.5 1.2 3.94 1.71
MBC-5-34 5 9X8W0 0.4 1.2 5.65 1.62

More specifically, Table 5.12 suggests it would not be unreasonable to say that the soil
modulus must be identified to within ±200 psi for fill depths greater than 6 ft and ±1000 psi
for fill depths less than 6 ft. This is very high precision for geotechnical work.

TxDOT 0-5849 153

 3. Conclusion

The modulus of elasticity greatly and directly affects the inventory rating. For higher fill
depths the influence of soil modulus on load rating is even more pronounced. To control
error in load rating calculations to within 10% according to structural tolerance (±HS-2), the
modulus of elasticity should be identified to within ±200 psi for high fill depths (more than 6
ft of fill) and ±1000 psi for low fill depths (less than 6 ft of fill).
Much can be said about the challenge of achieving a reasonable degree of precision for
the soil elastic modulus value. Soils are highly variable, their strength properties are stress-
dependent, and these properties can vary over time. All of the uncertainties and errors
associated with geotechnical sampling and testing come into play. Further, the soil modulus
parameter can be obtained by multiple methods, ranging from the selection of tabulated
“textbook” values to site-specific determination through in-situ tests. These factors suggest
that the selection of soil modulus values for culvert load rating purposes can introduce
significant uncertainty into the calculation.

TxDOT 0-5849 154

 TABLE 5.13. THE LOAD RATING FOR SEVEN REPRESENTATIVE CULVERT DESIGNS WITH INCREASING MODULUS OF ELASTICITY.
ID E = 4 KSI E = 8 KSI E = 12 KSI E = 16 KSI E = 20 KSI

MC-7-16 4 7X4W16 14.4 24.0 WTIC M 51.7 86.3 WTIC M 81.4 135.9 WTIC M 107.0 178.6 WTIC M 130.9 218.6 WTIC M
MC-10-20 5 10X10W20 -69.9 -116.8 WTIC M -24.3 -40.6 WTIC M 11.5 19.2 WTIC M 42.8 71.5 WTIC M 71.7 119.6 WTIC M
MC6-2 6 6X3W6 12.3 20.6 WBEC M 24.9 41.5 WBEC M 32.7 54.6 TEM M 40.1 66.9 TEM M 47.5 79.3 TEM M
MBC-1-44-F 2 5X4W6 8.7 14.4 WBEC M 13.8 23.1 WBEC M 18.8 31.3 WTIC M 23.6 39.3 WBEC M 28.2 47.1 WBEC M
MC9-2 2 9X9W4 -21.4 -35.8 WBEC M -14.6 -24.3 WBEC M -8.7 -14.4 WTIC M -3.3 -5.5 WBEC M 1.7 2.9 WBEC M
MC7-1 3 7X6W2 -1.5 -2.5 BEC M 3.1 5.3 BEC M 6.9 11.5 WBEC M 10.0 16.7 WBEC M 12.5 20.8 WTIC M
MBC-5-34 5 9X8W0 -2.5 -4.2 WTIC M 2.4 4.0 WTIC M 5.1 8.5 WTIC M 6.6 11.1 WTIC M 7.3 12.2 WTIC M

ID E = 24 KSI E = 28 KSI E = 32 KSI E = 36 KSI E = 40 KSI

MC-7-16 4 7X4W16 153.0 255.5 WTIC M 173.5 289.6 WTIC M 193.5 323.1 WTIC M 211.9 353.7 BIC V 226.5 378.1 BIC V
MC-10-20 5 10X10W20 98.1 163.7 WEM M 121.0 201.9 WEM M 146.1 243.9 WEM M 173.5 289.6 WTIC M 196.8 328.5 WTIC M
MC6-2 6 6X3W6 52.9 88.3 TIC V 55.0 91.9 TIC V 57.1 95.3 TIC V 59.0 98.4 TIC V 60.6 101.2 TIC V
MBC-1-44-F 2 5X4W6 32.9 54.9 WBEC M 37.5 62.6 WBEC M 41.9 70.0 WBEC M 46.6 77.7 WBEC M 51.0 85.1 WBEC M
MC9-2 2 9X9W4 6.5 10.8 WBEC M 11.0 18.3 WBEC M 15.3 25.6 WBEC M 19.5 32.6 WBEC M 23.6 39.4 WBEC M
MC7-1 3 7X6W2 13.7 22.9 WTIC M 14.9 24.9 WTIC M 15.9 26.6 WTEC M 16.4 27.4 WTEC M 16.8 28.0 WTEC M
MBC-5-34 5 9X8W0 8.5 14.1 WTIC M 9.0 15.1 WTIC M 9.5 15.9 WTIC M 9.9 16.5 WTIC M 10.2 17.0 WTIC M

TxDOT 0-5849 155

 250.0

200.0

150.0
Inventory Rating, IR (HS-)

MC-7-16 4 7x4w16
100.0
MC-10-20 5 10x10w20
MC6-2 6 6x3w6

50.0 MBC-1-44-F 2 5x4w6

MC9-2 2 9x9w4
MC7-1 3 7x6w2

0.0 MBC-5-34 5 9x8w0

-50.0

-100.0
0 5 10 15 20 25 30 35 40 45
modulus of elasticity, E (ksi)

FIGURE 5.6. MODULUS OF ELASTICITY VS. THE INVENTORY RATING FOR SEVEN REPRESENTATIVE CULVERT DESIGNS.

TxDOT 0-5849 156

 5. Depth of Fill

1. Parametric Sample

The depth of fill refers to the amount of overburden soil above the culvert top slab.
Direct traffic culverts are those with less than 2 feet of fill, and deep-fill culverts are those
with more than six feet of fill. Culvert designs typically are specified relative to a maximum
fill depth. However, situations arise, such as in roadway rehabilitation projects, where it
becomes desirable to increase the depth of fill above the design amount, or to lessen the
depth of fill. The question being addressed with this parameter is, what influence does the
depth of fill have on the culvert inventory rating.
For a sample of seven culvert designs, the inventory ratings were determined five times
for each culvert at four different fill depths. Four depths of fill were chosen for each culvert
to represent the minimum and maximum design depth, an average between the two and an
overload depth. No increments of less than 2ft were used. Analytical models included
CULV-5 with AASHTO lateral earth pressures, RISA-2D with springs with a modulus of
subgrade reaction of 150psi and AASHTO lateral earth pressures, and RISA-2D with LEFE
using a Poisson’s ratio of 0.3 and soil modulus values of 8ksi, 20ksi and 30ksi. These were
chosen to portray the range of model abilities and soil qualities.
2. Results

Table 5.14 presents the calculated inventory rating values for different depths of fill.
Figure 5.7 through Figure 5.13 are plots of the inventory rating with respect to the depth of
fill.
Several observations may be made from these data. First, the data emphasize the
sensitivity of the relationship between modulus of elasticity and depth of fill. As the depth of
fill increases, the sensitivity to changes in the modulus of elasticity also increases.
Second, typically the highest rating occurs when the depth of fill is at the maximum
design depth. At this depth, the culvert is precisely designed for the dead load, and the live
load is most dissipated.
Third, it is clear that just because a culvert rates well for maximum fill, it may not rate as
well for the minimum fill. Though the dead load is less for the minimum fill, the live load
may be exponentially larger. There is less fill to dissipate the live loads.
Fourth, typically, the intermediate fill depths produce load ratings in between the
maximum and minimum values.
Fifth, it appears that culverts may have some capacity for overload fill depths, though this
may be highly dependent upon soil properties.

TxDOT 0-5849 157

 TABLE 5.14. THE LOAD RATING FOR SEVEN REPRESENTATIVE CULVERT DESIGNS WITH INCREASING DEPTH OF FILL.
ID LVL 1: MIN LVL 1: AVERAGE LVL 1: MAX LVL 1: OVERLOAD
MC-7-16 4 7X4W16 16.1 26.8 WTIC M RLL 81.1 135.3 WTIC M RLL 55.2 92.1 BIC V RLL -242.3 -404.5 BIC V TL
MC-10-20 5 10X10W20 9.9 16.6 WEM M TL -4.2 -7.1 WEM M TL -29.2 -48.7 WEM M RLL -7761.4 -12955.6 BIC V TL
MC6-2 6 6X3W6 14.5 24.3 BEM M RLL 13.4 22.4 BEM M RLL 4.6 7.7 BEM M RLL -11.9 -19.8 BEM M RLL
MBC-1-44-F 2 5X4W6 9.9 16.5 WBEC M TL 16.3 27.2 WBEC M TL 27.9 46.6 WBEC M TL 23.6 39.4 WBEC M TL
MC9-2 2 9X9W4 -0.9 -1.4 WEM M TL -2.9 -4.8 WEM M TL -16.3 -27.2 WBEC M TL -41.4 -69.1 WBEC M TL
MC7-1 3 7X6W2 6.2 10.4 BEM M RLL 4.3 7.2 BEM M RLL 3.0 5.0 BEM M RLL -4.8 -8.1 BEM M RLL
MBC-5-34 5 9X8W0 5.9 9.9 BIC M RLL 1.7 2.9 BIC M RLL -5.4 -9.1 BIC M RLL -19.8 -33.1 BIC M RLL

ID LVL 2: MIN LVL 2: AVERAGE LVL 2: MAX LVL 2: OVERLOAD

MC-7-16 4 7X4W16 10.1 16.9 WTIC M RLL 77.2 128.9 WTIC M RLL 58.3 97.3 TIC V RLL -134.2 -224.0 TIC V TL
MC-10-20 5 10X10W20 12.8 21.4 WEM M TL -4.9 -8.2 WEM M TL -31.4 -52.5 WEM M TL -4788.3 -7992.8 TIC V TL
MC6-2 6 6X3W6 15.8 26.3 TEM M RLL 15.9 26.6 TEM M RLL 8.8 14.7 TEM M RLL -10.5 -17.6 TEM M RLL
MBC-1-44-F 2 5X4W6 16.7 27.9 WBEC M RLL 20.7 34.6 WBEC M RLL 40.6 67.8 WBEC M RLL 53.3 88.9 WBEC M RLL
MC9-2 2 9X9W4 -1.5 -2.6 WEM M TL -4.1 -6.9 WEM M TL -81.8 -136.5 WEM M TL -49.1 -81.9 WBEC M TL
MC7-1 3 7X6W2 7.5 12.5 WTIC M TL 6.5 10.9 BEM M RLL 6.9 11.6 BEM M RLL -0.9 -1.5 BEM M RLL
MBC-5-34 5 9X8W0 7.8 13.0 WBEC M RLL 6.5 10.8 WBEC M RLL 9.2 15.3 WBEC M RLL 0.6 0.9 BIC M RLL

TxDOT 0-5849 158

 TABLE 5.14. THE LOAD RATING FOR SEVEN REPRESENTATIVE CULVERT DESIGNS WITH INCREASING DEPTH OF FILL. (CONT).
ID LVL 3: E=8KSI: MIN LVL 3: E=8KSI: AVERAGE LVL 3: E=8KSI: MAX LVL 3: E=8KSI: OVERLOAD
MC-7-16 4 7X4W16 11.3 18.9 WTIC M -35.8 -59.7 WTIC M 51.7 86.3 WTIC M -8.6 -14.3 TIC V
MC-10-20 5 10X10W20 0.5 0.8 WTIC M -10.7 -17.9 WTIC M -24.3 -40.6 WTIC M -2982.5 -4978.5 BEC V
MC6-2 6 6X3W6 21.1 35.3 TEM M 24.9 41.5 WBEC M 16.5 27.6 WBEC M 4.7 7.8 WBEC M
MBC-1-44-F 2 5X4W6 8.9 14.9 WBEC M 12.3 20.6 WBEC M 13.8 23.1 WBEC M 5.9 9.8 WBEC M
MC9-2 2 9X9W4 -3.1 -5.2 WBEC M -14.6 -24.3 WBEC M -27.1 -45.2 WBEC M -1435.7 -2396.6 BEM M
MC7-1 3 7X6W2 6.1 10.2 WTIC M 3.1 5.3 BEC M -2.5 -4.1 WBEC M -12.1 -20.1 WBEC M
MBC-5-34 5 9X8W0 2.5 4.2 WTIC M -1.5 -2.5 WBEC M -11.5 -19.2 WBEC M -24.1 -40.3 WBEC M

ID LVL 3: E=20KSI: MIN LVL 3: E=20KSI: AVERAGE LVL 3: E=20KSI: MAX LVL 3: E=20KSI: OVERLOAD
MC-7-16 4 7X4W16 16.0 26.7 WTIC M 32.2 53.7 WTIC M 130.9 218.6 WTIC M 68.9 115.0 TIC V
MC-10-20 5 10X10W20 11.8 19.7 WTIC M 41.5 69.3 WTIC M 71.7 119.6 WTIC M -72.6 -121.2 WEM M
MC6-2 6 6X3W6 33.6 56.0 TEM M 47.5 79.3 TEM M 46.9 78.3 WBEC M 35.9 60.0 WBEC M
MBC-1-44-F 2 5X4W6 12.4 20.7 WBEC M 20.2 33.7 WBEC M 28.2 47.1 WBEC M 21.4 35.8 WBEC M
MC9-2 2 9X9W4 7.3 12.1 WBEC M 1.7 2.9 WBEC M -10.0 -16.8 WBEC M -25.2 -42.1 WBEC M
MC7-1 3 7X6W2 8.3 13.9 WTIC M 12.5 20.8 WTIC M 12.8 21.4 WBEC M 4.3 7.3 WBEC M
MBC-5-34 5 9X8W0 7.7 12.8 WTIC M 9.3 15.5 WTIC M 12.4 20.7 WBEC M 3.2 5.3 WBEC M

ID LVL 3: E=36KSI: MIN LVL 3: E=36KSI: AVERAGE LVL 3: E=36KSI: MAX LVL 3: E=36KSI: OVERLOAD
MC-7-16 4 7X4W16 18.9 31.5 WTIC M 89.7 149.7 WTIC M 211.9 353.7 BIC V 156.2 260.8 BIC V
MC-10-20 5 10X10W20 18.4 30.7 WTIC M 92.8 155.0 WTIC M 173.5 289.6 WTIC M 14.5 24.3 WEM M
MC6-2 6 6X3W6 48.6 81.2 TIC V 59.0 98.4 TIC V 58.6 97.9 BIC V 49.8 83.1 BIC V
MBC-1-44-F 2 5X4W6 14.8 24.7 TIC M 30.0 50.0 WBEC M 46.6 77.7 WBEC M 40.7 67.9 WBEC M
MC9-2 2 9X9W4 15.5 25.8 TEM M 19.5 32.6 WBEC M 10.0 16.6 WBEC M -3.7 -6.1 WBEC M
MC7-1 3 7X6W2 8.9 14.9 WTIC M 16.4 27.4 WTEC M 28.7 47.8 WBEC M 21.9 36.6 WBEC M
MBC-5-34 5 9X8W0 9.9 16.5 WTIC M 13.6 22.6 WTIC M 32.0 53.4 WTIC M 37.0 61.8 WBEC M

TxDOT 0-5849 159

 FIGURE 5.7. INVENTORY RATING VS. DEPTH OF FILL FOR MC-7-16 4 7X4W16.

275.0

225.0
Inventory Rating, IR (HS-)

175.0

125.0 Lvl.1: CULV-5

Lvl. 2: RISA-2D SPG
75.0
Lvl. 3: RISA-2D LEFE: E=8ksi

25.0 Lvl. 3: RISA-2D LEFE: E=20ksi

Lvl. 3: RISA-2D LEFE: E=36ksi
-25.0

-75.0
0 5 10 15 20 25 30
Depth of Fill, DOF (ft)

FIGURE 5.8. INVENTORY RATING VS. DEPTH OF FILL FOR MC-10-20 5 10X10W20.

225.0

175.0
Inventory Rating, IR (HS-)

125.0
Lvl.1: CULV-5
75.0 Lvl. 2: RISA-2D SPG
Lvl. 3: RISA-2D LEFE: E=8ksi
25.0
Lvl. 3: RISA-2D LEFE: E=20ksi
Lvl. 3: RISA-2D LEFE: E=36ksi
-25.0

-75.0
0 5 10 15 20 25 30 35
Depth of Fill, DOF (ft)

TxDOT 0-5849 160

 FIGURE 5.9. INVENTORY RATING VS. DEPTH OF FILL FOR MC7-1 3 7X6W2.

35.0
Inventory Rating, IR (HS-) 30.0
25.0
20.0
15.0 Lvl.1: CULV-5
10.0 Lvl. 2: RISA-2D SPG
5.0
0.0 Lvl. 3: RISA-2D LEFE: E=8ksi
-5.0 Lvl. 3: RISA-2D LEFE: E=20ksi
-10.0
-15.0 Lvl. 3: RISA-2D LEFE: E=36ksi
0 1 2 3 4 5 6 7
Depth of Fill, DOF (ft)

FIGURE 5.10. INVENTORY RATING VS. DEPTH OF FILL FOR MC9-2 2 9X9W4.

40.0
Inventory Rating, IR (HS-)

20.0
0.0
Lvl.1: CULV-5
-20.0
-40.0 Lvl. 2: RISA-2D SPG

-60.0 Lvl. 3: RISA-2D LEFE: E=8ksi

-80.0 Lvl. 3: RISA-2D LEFE: E=20ksi
-100.0 Lvl. 3: RISA-2D LEFE: E=36ksi
1 3 5 7 9
Depth of Fill, DOF (ft)

FIGURE 5.11. INVENTORY RATING VS. DEPTH OF FILL FOR MC6-2 6 6X3W6.

70.0
Inventory Rating, IR (HS-)

60.0
50.0
40.0 Lvl.1: CULV-5
30.0
20.0 Lvl. 2: RISA-2D SPG
10.0 Lvl. 3: RISA-2D LEFE: E=8ksi
0.0
Lvl. 3: RISA-2D LEFE: E=20ksi
-10.0
-20.0 Lvl. 3: RISA-2D LEFE: E=36ksi
3 5 7 9 11
Depth of Fill, DOF (ft)

TxDOT 0-5849 161

 FIGURE 5.12. INVENTORY RATING VS. DEPTH OF FILL FOR MBC-5-34 5 9X8W0.

50.0

40.0

30.0
Inventory Rating, IR (HS-)

20.0
Lvl.1: CULV-5
10.0 Lvl. 2: RISA-2D SPG

0.0 Lvl. 3: RISA-2D LEFE: E=8ksi

Lvl. 3: RISA-2D LEFE: E=20ksi
-10.0
Lvl. 3: RISA-2D LEFE: E=36ksi
-20.0

-30.0
0 1 2 3 4 5 6
Depth of Fill, DOF (ft)

FIGURE 5.13. INVENTORY RATING VS. DEPTH OF FILL FOR MBC1-44-F 2 5X4W6.

MBC-1-44-F 2 5x4w6
60.0

50.0
Inventory Rating, IR (HS-)

40.0
Lvl.1: CULV-5
30.0 Lvl. 2: RISA-2D SPG
Lvl. 3: RISA-2D LEFE: E=8ksi
20.0
Lvl. 3: RISA-2D LEFE: E=20ksi
10.0 Lvl. 3: RISA-2D LEFE: E=36ksi

0.0
0 2 4 6 8
Depth of Fill, DOF (ft)

TxDOT 0-5849 162

 3. Conclusion

The sensitivity of inventory rating to the depth of fill is very great and highly related to
the structural design, particularly the maximum design fill depth. As expected, the load
rating is heavily and non-linearly dependent upon depth of fill. As the fill depth increases, it
becomes more important to identify the modulus of elasticity more precisely.
When designing culverts it is appropriate to calculate the load ratings for the maximum
and minimum design fill depths, but intermediate fill depths may be ignored. When real
culverts are rated, the rating should be calculated for the site-specific fill condition. If the
soil properties reflect high quality backfill, it may be possible to exceed the maximum design
fill depth, but soil properties must be validated.

TxDOT 0-5849 163

 4. CONCLUSIONS

Table 5.15 summarizes results for the six variables analyzed in the parametric study.
Four of the six variables – modulus of subgrade reaction, Poisson’s ratio, multibarrel effects,
and lateral earth pressures – do not greatly influence the calculated inventory rating values.
TABLE 5.15. SUMMARY OF PARAMETRIC STUDY CONCLUSIONS.

VARIABLE MODEL SENSITIVITY OF INVENTORY RATING TO SELECTED PARAMETER

Not sensitive;
Modulus of subgrade reaction, k 2
keep Culvert Rating Guide values

Not sensitive;
Poisson's ratio, ν 3 Culvert Rating Guide value of 0.3 is appropriate;
0.5 is more appropriate for deep fill clay soils

Not sensitive;
Multibarrel effects 3 model 5+ barrels as 4 barrels for culv-5;
model all barrels in higher order analyses

Lateral earth pressures, lep Not sensitive;

1&2
(equivalent fluid unit weight) use AASHTO recommended value (60pcf)

Very sensitive;
use factor of safety
Modulus of elasticity, e 3 OR
±200 psi for high fill depths (more than 6ft)
±1000 psi for low fill depths (less than 6 ft)

Very sensitive;
Depth of fill 1, 2 & 3 check at actual depth of fill for as-built culvert;
check at maximum and minimum depth of fill for designs

Calculated inventory ratings are highly sensitive to the soil modulus of elasticity, which
is the key parameter in the soil constitutive model for the Level 3 analysis (RISA 2D with
LEFE). Given the complexity and variability of soil, it will be difficult to determine this
parameter with the desired level of precision, and this will introduce uncertainty into the load
rating process.
Calculated inventory ratings are also highly sensitive to the depth of fill above the culvert
top slab. The effect is more pronounced for deep-fill culverts, compounding the sensitivity to
soil modulus values. Culvert load rating analyses should model actual site conditions;
whereas, culvert design should evaluate both maximum and minimum fill depths. A culvert’s
capacity for overload fill depths will largely depend on the quality of the backfill.
Relative to validation of the Culvert Rating Guide, the findings from the parametric study
are incorporated into the discussions for the respective parameters.

TxDOT 0-5849 164

TxDOT 0-5849 165
 6. INSTRUMENTED LOAD TESTS ON THREE IN-SERVICE CULVERTS

1. OVERVIEW

1. Purpose and Scope

This chapter presents findings from field instrumentation and load testing of three in-
service reinforced concrete box culverts evaluated as part of this research study. The research
team performed this work with the following objectives in mind:

• Obtain project-specific model input parameters for selected culverts.

• Measure actual strains and deflections under known load conditions to determine
actual demand moments at critical sections for the culvert structure.
• Through site-specific geotechnical explorations, evaluate alternative methods for
determining soil modulus values.
• Compare measured versus predicted demand moments to evaluate the reliability
of the analytical modeling procedures presented in the Culvert Rating Guide.
• Further explore factors which influence production load rating of culverts.
This work continues the process of validation of the Culvert Rating Guide from breadth
(load rating 100 culverts discussed in Chapter 4) to sensitivity (parametric analysis of 7
culverts discussed in Chapter 5) to an in depth study of the structural response of the culvert
load-soil-structure system (presented herein).

2. Selection of Culverts for Instrumentation and Load Testing

This project required that three in-service culverts be instrumented and load tested in the
field. These three culverts were selected to be representative of the primary design eras of
concern, namely 1946 era, the 1958 era with structural grade reinforcing, and the 1958 era
with Grade 60 reinforcing. In addition to requiring three culverts from three different eras,
the potential test culverts were evaluated for safety, ease of access, and distribution of other
design parameters including depth of fill, box geometry, and number of spans.
The first test culvert is located 5 miles northwest of Shallowater, Texas. This culvert was
built in 1963 making it an early 1958 era design (structural grade reinforcing). It consists of
four 10’ by 8’ boxes. The location is a divided highway with a wide median; therefore,
maneuvering room was available to perform load tests over two foot of fill and over four foot
of fill.
The second test culvert is located in Plainview, Texas. This culvert was built in 1991
making is a late 1958 era design (Grade 60 reinforcing). It consists of four 10’ by 6’ boxes
with 3.5 feet of fill and roadway over the culvert.
The third test culvert is located 8 miles east of Tulia Texas. This culvert was built in 1951
making it a 1946 transitional culvert design. It consists of five 6’ by 6’ barrels beneath 1.5
feet of fill and roadway.

TxDOT 0-5849 166

 2. LOAD TEST DESIGN

1. Load Rating Equation

The load test design for this study focuses on measuring and/or determining the three
independent variables in load rating equation. These are the capacity, the dead load demand
and the live load demand. Throughout this research, the capacity portion of the load rating
equation has been assumed to be clearly defined per AASHTO policy and understood with a
reasonable degree of precision. Apart from obtaining project-specific culvert parameters,
capacity has not been specifically investigated in this task.
The dead and live load demands from the load rating equation, however, are determined
through analytical modeling, and this research does investigate those terms. The field
instrumentation plan and the load test procedure were designed to specifically measure and
validate the live load moment demands relative to the culvert load rating process. This means
that measuring live load demands is where most of the effort for this task was expended.
The initial research design restricted the load-testing portion of the study to consider only
in-service culverts. One of the consequences of this a priori decision is that it places
limitations on acquisition of dead load data. For in-service culverts, the initial stress
condition of the as-built culvert is taken as a baseline for live load strain measurements.
Those strains associated with placement of the culvert backfill, which amount to the dead
load demand in the load rating equation, cannot be measured because the fill is already there.
Notwithstanding this limitation, the research team did attempt to perform limited dead load
measurements by placing overload fill at the Shallowater test site. Otherwise, the focus for
the load test design was on measuring strains and deflections under known load conditions to
calculate live load demands.

TxDOT 0-5849 167

 2. Measuring Live Load Demands – Instrumentation Design

Load ratings are determined for each critical section on the structure. Figure 6.1 shows
the critical sections for a culvert without haunches. Two dimensional models are typically
used to analyze culverts, therefore a unit width strip is typically considered. Because the
load rating models consider a thin strip of the culvert and only analyze the demands at critical
sections, the instrumentation plan was designed to consider a single cross section of the
culvert at the critical sections. This was accomplished by establishing a gage line at a known
location, normal to the culvert barrels, and placing various types of instrumentation along
this gage line at or near the critical sections.
TEC TEM TIC

WTEC WTIC

WEM WIM

WBEC WBIC

BEC BEM BIC

FIGURE 6.1. CRITICAL SECTIONS

The typical instrumentation plan consisted of placing strain gages at all interior critical
sections for both corners and walls. Exterior critical sections were not readily accessible
because the culvert is buried; however, critical sections along the top slab could typically be
instrumented. In addition to the strain gages, the instrumentation plan included deflection
measurements at midspans of the top and bottom slabs.

3. Application of Live Load – Field Test Procedure

The load test design was limited to in-service culverts in an effort to minimize field
testing costs. The live load consisted of loaded TxDOT 10CY dump trucks. The live load
procedure involved driving the loaded truck(s) across the top of the culvert, stopping the
truck at two-foot increments along the pre-established gage line, and taking strain and
deflection measurements at each stop. Once the truck(s) passed over the culvert, they were
turned around and driven back over the culvert in the same manner but in the opposite
direction, again stopping every 2 feet for a reading. Figure 6.1 shows a typical gage line used
to establish the location of instrumentation as well as the dump truck’s position above the
culvert structure.

TxDOT 0-5849 168

 FIGURE 6.2. GAGE LINE CONSISTING OF TWO FOOT GRID MARKS TO DETERMINING TRUCK POSITIONS.

Three truck configurations were used for each load test. The first configuration consisted
of a single truck driving over the culvert with the gage line between the wheels, i.e.
straddling the gage line. The second configuration used a single truck driving over the culvert
with one set of wheels directly on the gage line. The last configuration used two trucks
driving over the culvert, in tandem, four feet apart with the gage line between them. Figure
6.3 shows the various truck positions over the gage line.

FIGURE 6.3. LIVE LOAD CONFIGURATIONS: A) TRUCK STRADDLING GAGE LINE B) WHEEL ON GAGE LINE C) 2 TRUCKS STRADDLING GAGE
LINE.

The three truck configurations with the trucks traveling back and forth across the culvert
were designed to facilitate creation of moment envelopes for the applied load combinations.
The goal was to measure the worst case maximum and minimum moment demands for each
critical section of the culvert structure.

TxDOT 0-5849 169

 4. Predictive Analyses – Modeling the Load-Culvert-Soil System

The load test measured strains at the culvert critical sections; whereas, the load rating
equation requires demand moments at each critical section. This means that the measured
strains must be transformed into moment demands, and this is done through consideration of
the strain profile in the culvert members (slabs and walls).
Curvature is calculated from strain. Where the strain is measured on both sides of the
culvert slab or wall, the curvature is calculated directly using Equation 6.1.
EQUATION 6.1. CURVATURE FROM INSIDE AND OUTSIDE STRAIN.
𝜀𝐼 − 𝜀𝐸
𝜑=
ℎ
Where: φ = curvature
εI = strain on the interior of the wall
εE = strain on the exterior of the wall
h = the wall thickness (in.)
Where the strain is only known on the inside face of the slab or wall, curvature is calculated
assuming the centroid lies at the gross neutral axis according to Equation 6.2
EQUATION 6.2. CURVATURE FROM INSIDE STRAIN.
2𝜀𝐼
𝜑=
ℎ
Where: φ = curvature
εI = strain on the interior of the wall
h = the wall thickness (in.)
The cracked moment of inertia is established based on ACI approximations. For slabs,
the cracked moment of inertia is 0.3 times the gross moment of inertia. For walls, the cracked
moment of inertia is 0.5 times the gross moment of inertia. The measured moment is then
calculated according to Equation 6.3.
EQUATION 6.3. MEASURED MOMENT FROM STRAIN.
𝑀𝑚 = 𝜑𝐸𝑐 𝐼
Where: Mm = the measured moment
φ = curvature
Ec = modulus of elasticity of concrete
I = the moment of inertia of the wall
This process is consistent with standard mechanics of materials formulations and was used to
convert the measured strains into demand moments at each critical section for each load test.

TxDOT 0-5849 170

 5. Obtaining Test Parameters

In addition to measuring live load demands, the field test procedure required acquisition
of several other culvert parameters including dimensional data, structural strength data, soil
parameters, and truck weights.
1. Culvert dimensional data

The number of barrels, span, height and wall thicknesses were determined from the
culvert plan sheets and spot checked by field measurements for the actual culverts. Slab
thicknesses and steel reinforcing schedules were determined from the plan sheets, and
verified where practicable.
2. Structural strength data

The reinforcing steel yield strength was determined based on AASHTO Manual for
Condition Evaluation of Bridges (MCEB) guidelines as per the culvert plan sheets. This
research did not perform project-specific testing of the reinforcing steel.
The Level 2 and 3 analytical models used concrete compressive strengths determined by
lab tests on concrete core specimens obtained from the bottom slab of the culverts. For the
Level 1 model, the concrete compressive strength was assumed as per the MCEB values. The
MCEB value was also used for the Level 2 model for comparison purposes. The modulus of
elasticity of the concrete was calculated using the normal weight concrete relationship
between compressive strength and modulus of elasticity.
3. Soil parameters

The soil unit weight was assumed as 120pcf according to guidance from the AASHTO
Standard Specifications for Highway Bridges (SSHB). This assumption was confirmed for
reasonableness based on unit weight tests for undisturbed soil samples and nuclear density
tests taken in the field. Soil strength data was determined from exploratory geotechnical
borings, as described in the next section of this report.
4. Truck weights

The truck weights were measured independently by the TxDOT maintenance crews at
local scales. Weights were measured for the whole truck, the front axle, and the rear tandem
axle.

TxDOT 0-5849 171

 3. GEOTECHNICAL EXPLORATIONS AT THE CULVERT SITES

1. Description

The geotechnical exploration for this study consisted of drilling three geotechnical
borings, approximately 15 feet deep, at each of the three culvert sites. TxDOT provided
traffic control, and the boring locations were cleared for underground utilities (Dig-TESS)
prior to drilling. The drill crew was required to follow TxDOT safety protocol while working
within the TxDOT right-of-way including wearing reflective vests, hard hats, steel toe boots,
etc.

2. Field Drilling and Sampling

The test borings were drilled in the backfill zone of the culvert structures, two borings on
one side and one boring on the other. Field drilling and sampling operations were designed to
collect multiple sets of data from each boring including thin-walled tube samples (TWT),
standard penetration tests (SPT), dynamic cone penetration (DCP) tests, pressuremeter tests
(PMT), and Texas cone penetration (TCP) tests. Sampling and testing operations were
performed in substantial accordance with applicable standard test methods. Appendix F
presents the logs of borings for each culvert site.

3. Laboratory Testing

The laboratory study consisted of performing tests on recovered samples to aid in

identification and classification of soils and determination of soil properties. TxDOT
personnel conducted these tests including Atterberg limits, minus-200 mesh sieve, sieve
analysis, moisture content, unit weight, and unconfined compression. Tests were performed
in substantial accordance with applicable TxDOT test methods. Data are summarized on the
boring logs.
4. Falling Weight Deflectometer Testing

In addition to the geotechnical sampling and testing, TxDOT conducted Falling Weight
Deflectometer testing at each of the three culvert locations.
The falling weight deflectometer (FWD) is a trailer-mounted device that places an 11.8
inch (300 mm) diameter load plate in contact with the highway at each test location. A load
column above the load plate carries a stack of weights that are dropped to impart a load to the
pavement similar to that imparted by a passing dual truck tire set. A series of seven
geophones spaced away from the load plate at 12 inch increments measures the surface
deflection, generating a “deflection bowl.” (TxDOT, 2008) The testing interval was set at 30
locations per project, with data obtained for both approach and departure at each culvert.
FWD data including the raw deflection file, pavement layer thicknesses, layer Poisson
ratios, probable layer moduli ranges, and asphalt temperatures at the time of testing are used
to perform backcalculation of modulus values for the pavement subgrade. TxDOT currently
uses version 6.0 of their MODULUS software for backcalculation of deflection data.
Appendix G presents the backcalculated FWD modulus data for the each culvert.

TxDOT 0-5849 172

 5. The Subsurface Profiles for Each Culvert Site

The logs of borings (Appendix F) provide detailed information about subsurface

conditions for each culvert site. These data were synthesized to create an idealized subsurface
profile, and it is this idealized profile which is used for subsequent culvert modeling. Figure
6.4 summarizes the idealized profiles for each culvert site, along with a tabulation of field
and laboratory soil strength tests.
Review of Figure 6.4 shows the following:

• Subsurface conditions for each project site can be idealized as two strata,
Stratum I being from the top of ground to the base of the culvert, and Stratum II
being below the base of the culvert.

• Culvert backfill soils are similar to the soils of the surrounding area. The culvert
backfills explored for this study did not consist of “better material.”

• Significant variability exists in the raw soil data (SPT, TCP, DCP, UCS) for each
stratum, with certain parameters easily varying more than an order of magnitude.

• Soil strength values used for analysis typically are the mean values with
adjustments based on judgment.
The variability inherent in the raw soil data becomes significantly filtered in subsequent
analyses. For example, SPT N-values for Stratum II of the Shallowater culvert range from 3
to 53 blows per foot, but subsequent analyses are based on a single value of 8 blows per foot.
This means that geotechnical interpretations about basic soil strength properties do not carry
forward the range of values, but the selected values do incorporate significant uncertainty.

6. Determination of Modulus of Subgrade Reaction Parameters

As has been noted, Level 2 analyses are based on the RISA-2D structural model and use
soil springs for support of the culvert base. The stiffness of these springs is the modulus of
subgrade reaction, k.
Figure 6.5 summarizes the idealized profiles for each culvert site, along with a tabulation
of calculated modulus of subgrade reaction values for the lower soil stratum. Five methods
were used to calculate the modulus of subgrade reaction values. These include:

• Table IV-2 of the Culvert Rating Guide (TxDOT, 2009)

• Portland Cement Association relationship chart (PCA, 1992)

• Figure 6, p. 7.1-219 from NAVFAC DM 7.1 (NAVFAC, 1986)

• Table 9.1, page 505 of Foundation Analysis and Design (Bowles, Foundation
Analysis and Design, 1996)

• Equation 9.9, page 503, Foundation Analysis and Design (Bowles, Foundation
Analysis and Design, 1996)

TxDOT 0-5849 173

 Figure 6.5 presents the modulus of subgrade reaction value for each method along with a
brief statistical summary. Subgrade modulus values as determined by the different methods
vary 200 to 350 percent. Recall that these calculations for modulus of subgrade reaction
reflect idealized soil strength values, so this variation represents inherent scatter in the
parameter, not variation in the soil. Of course, variation in the soil strength also exists.

7. Determination of Soil Modulus Parameters

Level 3 analyses are based on the RISA-2D structural model but use a linear-elastic
constitutive model for the soil surrounding the culvert. The parameter that defines the
stiffness of the linear-elastic soil model is the elastic modulus.
Figure 6.6 summarizes the idealized profiles for each culvert site, along with a tabulation
of elastic modulus values determined for each soil stratum. Seven methods were used to
calculate the elastic modulus values. These include:

• Tabular values from the Culvert Rating Guide

• Tabular values from McCarthy and Bowles texts (McCarthy D. F., 2007; Bowles,
Foundation Analysis and Design, 1996)

• Derived values based on SPT data for sandy soils and UCS data for clayey soils
(Bowles, Foundation Analysis and Design, 1996; Coduto, 2001)

• Correlated TCP data, related to SPT values (TxDOT, 2006)

• Correlated DCP data, related to SPT values (DGSI, 2005)

• Measured PMT data (Briaud J.-L. , 2005)

• Backcalculated values obtained from FWD data (TxDOT, 2008)

Review of Figure 6.6 shows the following:

• Elastic modulus values within a given stratum as determined by the different

methods vary as much as 2500 percent or more. These calculations reflect
idealized soil strength values, so this variation represents inherent scatter in the
elastic modulus parameter, not variation in the soil. Of course, variation in the soil
strength also exists.

• Tabulated elastic modulus values from the Culvert Rating Guide are typically at
the high end.

• Falling weight deflectometer values are consistent and are typically second
highest.

• Tabulated elastic modulus values from textbooks (beams on elastic foundations)

are consistent and approximate the average of all values shown.

TxDOT 0-5849 174

 • Measured elastic modulus values based on the pressuremeter test are consistent
and are the lowest values.

• Derived and correlated values based on field penetration tests yield the greatest
variability.
As noted in Chapter 5, the inventory ratings for culverts are highly sensitive to the soil
elastic modulus value. Therefore, rather than take an average of the different methods, this
study determined the culvert load rating based on each of the different modulus values, the
goal being to determine which method for determining soil modulus yields the most reliable
results.

TxDOT 0-5849 175

 FIGURE 6.4. IDEALIZED SOIL PROFILE AND RAW FIELD AND LABORATORY SOIL STRENGTH DATA FOR THE THREE CULVERT SITES.

TxDOT 0-5849 176

 FIGURE 6.5. IDEALIZED SOIL PROFILE AND DERIVED MODULUS OF SUBGRADE REACTION VALUES FOR THE THREE CULVERT SITES.

TxDOT 0-5849 177

 FIGURE 6.6. IDEALIZED SOIL PROFILE AND DERIVED ELASTIC MODULUS VALUES FOR THE THREE CULVERT SITES.

TxDOT 0-5849 178

 4. THE SHALLOWATER CULVERT INSTRUMENTED LOAD TEST

1. Culvert Condition

The Shallowater culvert is located in Lubbock County approximately 5 miles

northwest of Shallowater, Texas. Figure 6.7 shows the location of the Shallowater test
culvert. This culvert was built in 1963. The design is a Pre-1977 1958 era culvert
designed under the THD Supplement No. 1. It consists of four 10’ wide by 8’ tall barrels.
A three foot thick layer of silt had to be removed from inside this culvert before testing.
The Shallowater culvert crosses US Highway 84, which is a divided four-lane road with a
45ft wide median. The culvert has 45° skew to the roadway. The culvert showed some
distress with longitudinal cracking and efflorescence with rust. The culvert also suffered
from very limited spalling.

FIGURE 6.7. SHALLOWATER LOCATION.

TxDOT 0-5849 179

 2. Summary of Parameters

1. Culvert Test Parameters

Culvert dimensional data from the plan sheets showed 4 barrels, 10ft spans, 8ft
height, 7.5in. thick top and bottom slabs, 8in. thick walls and 36ksi steel yield strength.
According to the MCEB, 3ksi concrete was assumed for Level 1 and 2 analyses. For
Level 2 and 3 analyses, the tested compressive strength of 6ksi from the concrete cores
was used. The trucks weighed 54kips. The front single axles weighed 14kips, while the
rear tandem axles weighed 40kips.
2. Soil Test Parameters

A soil unit weight of 120pcf was selected from the Culvert Rating Guide. A modulus
of subgrade reaction, k, of 150pci from Culvert Rating Guide was used for Level 2
analysis. The modulus of elasticity for the soil, Es; was determined using seven different
methods with values as follows:
1. 20 ksi from Culvert Rating Guide
2. 6 ksi from Bowles and McCarthy textbooks
3. 2 ksi from Standard Penetration Test (SPT)
4. 2 ksi from Texas Cone Penetration Test (TCP)
5. 3 ksi from Dynamic Cone Penetrometer (DCP)
6. 1 ksi from Pressuremeter (PMT)
7. 12 ksi from Falling Weight Deflectometer (FWD)
Testing was performed under two foot and four foot of fill.

TxDOT 0-5849 180

 3. Instrumentation Plan

The instrumentation plan involved placing 4in. electrical resistance strain gages at
every critical section on the inside of the two westernmost barrels. Strain gages were also
placed on the exterior of the top slab and at the top corner exterior walls opposite the
corresponding interior gages. This allowed for direct measurement of changes in
curvature in the strain profile. Linear displacement gages were placed at the centerline of
the top and bottom slabs. Figure 6.8 shows the location of the strain gages (in purple)
and the linear displacement gages (in blue).
8" 4'-0" 1'-0" 1'-0" 4'-0" 8" 4'-0" 1'-0" 1'-0" 4'-0"
ETEC ETEM ETIC1 ETIC2 ETIM ETIC3

7 1" ITEC ITEM ITIC1 ITIC2 ITIM ITIC3

2
EWTEC
IWTEC IWTIC1 IWTIC2 IWTIC3 IWTIC4

LVDT TES LVDT TIS

4'-0"

MATCH LINE
IWEM IWIM1 IWIM2 IWIM3 IWIM4

4'-0"
LVDT BES LVDT BIS

IWBEC IWBIC1 IWBIC2 IWBIC3 IWBIC4

IBEC IBEM IBIC1 IBIC2 IBIM IBIC3 IBIC4

FIGURE 6.8. SHALLOWATER INSTRUMENTATION PLAN.

TxDOT 0-5849 181

 4. Results for Shallowater Culvert with 2’ of Fill

1. Load Rating

Figure 6.9 shows the normal load rating for the Shallowater culvert under two feet of
fill using all the different models and soil properties. Only four methods produced
positive ratings and no methods produced a load rating, inventory or operating, above
HS-20. The Level 2 analysis and the Level 3 analysis for both the Culvert Rating Guide
and the falling weight deflectometer approaches show that the culvert should fail under
live load. Level 1 and the other Level 3 analyses show that the culvert should fail under
dead load.

25.0
20.0
Inventory Rating, IR (HS-)

15.0
10.0
5.0
0.0
-5.0
-10.0
-15.0
-20.0
-25.0
Lvl 1; Lvl 2; Lvl 2; Lvl 3; Lvl 3; Lvl 3; Lvl 3; Lvl 3; Lvl 3; Lvl 3;
TEXT TEXT; STRUC; CRG; Bowles; SPT; 2ksi TCP; 2ksi DCP; 3ksi PMT; FWD;
k=150pci k=150pci 20ksi 6ksi 1ksi 12ksi
Depth of Fill, DOF (ft)

Inventory Rating Operating Rating

FIGURE 6.9. LOAD RATING FOR SHALLOWATER CULVERT: 2’ OF FILL

Notwithstanding anticipated (poor) culvert performance as predicted by the load

rating process, the Shallowater culvert remains in service and clearly has not failed.
Various explanations can be offered for this. First, the assumed structural material
properties may be conservative. Second, it has already been noted that the soil around the
culvert is a variable material and attempts to quantify soil strength relative to culvert
support have produced a wide range of values. Third, the culvert rating calculation is
sensitive to the constitutive model used for both the soil and the structure. The load
ratings identified in Figure 6.9 reflect less-sophisticated constitutive models for both the
reinforced concrete and the soil. And finally, environmental factors at this site are
favorable, e.g., the site is typically dry and well-drained with low relative humidity.
Taken together, these factors can account for the fact that the as-built culvert is
serviceable, whereas the load rating prediction would suggest otherwise. The culvert has
“stood the test of time.”
TxDOT 0-5849 182
 2. Live Load Moment – Predicted and Measured

Figure 6.10 shows the live load moment envelope for the Shallowater culvert beneath
two feet of fill. The blue and red lines show the measured moment as well as error bars
which show a reasonable range due to variations in concrete modulus, cracked moment of
inertia and nature of four inch strain gages which measure average strain over the gage
length and not the actual strain at the critical section. The measured moment is less than
the predicted moment for all models.

8.000 CULV-5 MAX

CULV-5 MIN

6.000 RISA-2D TEXT MAX

RISA-2D TEXT MIN

4.000 RISA-2D STRUC MAX

RISA-2D STRUC MIN

RISA-2D LEFE CRG MAX

2.000
RISA-2D LEFE CRG MIN

RISA-2D LEFE Bowles

0.000 MAX
RISA-2D LEFE Bowles
MIN
RISA-2D LEFE SPT/TCP
-2.000 MAX
RISA-2D LEFE SPT/TCP
MIN
RISA-2D LEFE DCP MAX
-4.000 RISA-2D LEFE DCP MIN

RISA-2D LEFE PMT MAX

-6.000 RISA-2D LEFE PMT MIN

RISA-2D LEFE FWD MAX

-8.000 RISA-2D LEFE FWD MIN

ACTUAL MAX

ACTUAL MIN
-10.000

FIGURE 6.10. LIVE LOAD MOMENTS PREDICTED AND MEASURED FOR SHALLOWATER CULVERT: 2’ OF FILL.

TxDOT 0-5849 183

 3. Dead Load Moment – Predicted

Figure 6.11 shows the predicted dead load distributions for the Shallowater culvert
beneath two feet of fill for all models. The models predict similar directions for dead
load effect, except for exterior corner locations.

8.000

6.000 CULV-5 MAX

4.000 CULV-5 MIN

2.000 RISA-2D TEXT MAX

RISA-2D TEXT MIN
0.000
RISA-2D STRUC MAX
-2.000
RISA-2D STRUC MIN
-4.000
RISA-2D LEFE CRG
-6.000 RISA-2D LEFE Bowles
-8.000 RISA-2D LEFE SPT/TCP

-10.000 RISA-2D LEFE DCP

RISA-2D LEFE PMT
-12.000
RISA-2D LEFE FWD
BEC
BEM
WEM

WBIC1

WBIC2
WBEC

TEC
TEM

WTIC1

WTIC2
WTEC

WIM1

WIM2
BIC1

BIC2
TIM1
TIC1

TIC2

TIC3

BIM1
BIC3
FIGURE 6.11. DEAD LOAD PREDICTED FOR SHALLOWATER CULVERT: 2’ OF FILL.

TxDOT 0-5849 184

 4. Live Load “Goodness”

In order to determine how well the model predicts live load (dead load cannot be
measured on in-service culverts), the ratio of the predicted live load moment was
compared to the actual live load moments. The ideal model would show a ratio of live
load predicted to live load actual of 1.0. Excellent models would produce live load ratios
only slightly greater than 1.0. Overly conservative models would produce live load ratios
far greater than 1.0. Models which predict the appropriate direction but fail to predict
significant magnitude would have a ratio of less than 1.0. A model which fails to even
predict the appropriate direction would have a negative ratio. Table 6.1 shows the
thresholds and color scheme used in the plots showing the goodness of the live load
predictions.
From an economic view, critical sections which result in blue load ratings indicate a
very safe but expensive model. Models which produce mostly green critical section
ratios produce economical designs that maximize safety while minimizing cost. Yellow
and red critical sections mean that the model makes unsafe predictions.
TABLE 6.1. LIVE LOAD PREDICTED TO ACTUAL RATIO THRESHOLDS
threshold color At that critical section, the model:
10 < MP/MM dark blue Is overly conservative.
5 < MP/MM < 10 blue Is very conservative.
2 < MP/MM < 5 light green Is conservative.
1 < MP/MM < 2 green Is reasonably accurate.
0 < MP/MM < 1 yellow Predicts the correct sign but an unconservative magnitude.
MP/MM < 0 red Fails to predict the correct sign or magnitude.

Figure 6.12 shows the goodness plot for the Shallowater culvert beneath two feet of
fill. The plot shows the number of critical sections which fall into each threshold.
According to the plot, CULV-5 model produces the most conservative predictions. The
least conservative model is the RISA-2D with LEFE using the Culvert Rating Guide and
the Falling Weight Deflectometer soil moduli.

TxDOT 0-5849 185

 100%
90%
12 14
80% 34 35 17 18 18 19
70%
58
60% 10 10 < Mp/Mm
9
50% 5 < Mp/Mm < 10
20 20 11
12 11 2 < Mp/Mm < 5
40% 12
10
30% 1 < Mp/Mm < 2
13
11 21 20 0 < Mp/Mm < 1
20% 10 10
7 9 8
Mp/Mm < 0
10% 15 4
8 8 2 2 1 2
3 2 1 2 1 2
0% 0 1
0 1
0 0 0 0 0 0 0
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
(σ=53kft) TEXT STRUC LEFE CRG LEFE LEFE LEFE DCP LEFE PMT LEFE FWD
(σ=20kft) (σ=22kft) (σ=8kft) Bowles SPT/TCP (σ=24kft) (σ=38kft) (σ=11kft)
(σ=17kft) (σ=29kft)

FIGURE 6.12. LIVE LOAD GOODNESS PLOT FOR SHALLOWATER CULVERT: 2’ OF FILL

Also included in Figure 6.12 is the average standard deviation of the ratios for each of
the models. The smaller the average standard deviation, the better the model fit. By this
measure, CULV-5 is the least appropriate for accurately predicting the live load
moments. RISA-2D with LEFE using the Culvert Rating Guide soil modulus values is
the most appropriate model with the least amount of scatter and the most accurate
prediction of the live load moment envelope shape.

TxDOT 0-5849 186

 5. Load Rating “Goodness”

Live load testing is a one-time load condition. Therefore, when evaluating the
goodness of the load rating, the one-time load rating, or operating rating, was analyzed.
The ratings calculated for the field test were determined using the field truck weights.
Again the ideal model and design would produce a load rating of 1.0. Table 6.2 shows
the thresholds and meaning of the various load ratings.
In this case, blue operating ratings means that the design as analyzed by that model at
a given critical section is very adequate to carry the combined dead and live load at the
cost of too stout a critical section. Green operating ratings indicate that the culvert was
precisely adequate at that location to carry the dead load and truck weight. A yellow
rating indicates that that critical section should have been broken under the truck load and
the culvert was therefore unsafe under the truck load. A red rating indicated that that
critical section should have failed under dead load alone and should not be standing under
the dead load. Note that the color spectrum is the same for both Table 1.1 and Table 1.2,
but these tables relate different concepts.
TABLE 6.2. OPERATING RATING GOODNESS THRESHOLDS
threshold color The critical section:
10 < (C-1.3D)/2.17L dark blue Has an overly high load rating.
5 < (C-1.3D)/2.17L < 10 blue Has a very high load rating.
2 < (C-1.3D)/2.17L < 5 light green Has an acceptably high load rating.
1 < (C-1.3D)/2.17L < 2 green Has an optimized load rating.
0 < (C-1.3D)/2.17L < 1 yellow Should fail under dead and live load.
(C-1.3D)/2.17L < 0 red Should fail under dead load alone.

Figure 6.13 shows the number of critical sections in each threshold for each model.
According to this plot, CULV-5 indicates that in at least one critical section the structure
should be failed under dead load, and in almost 60% of the critical sections, the culvert
should have failed under the truck load. RISA-2D with LEFE however shows that 50%
of the culvert critical sections are well designed to handle the load and only 12% of the
culvert critical sections should have failed under the truck load.

TxDOT 0-5849 187

 100% 1 2 2
9 5
90% 14 14 3
11 7 9
8 9
80% 6
10 10
10 4
70% 3 15
11
8 9 8 10 < (C-1.3D)/2.17L
60% 20 20
11 11 5 < (C-1.3D)/2.17L < 10
50%
8 6 2 < (C-1.3D)/2.17L < 5
40% 9
17 15 11 1 < (C-1.3D)/2.17L < 2
30% 48 10
11 9 0 < (C-1.3D)/2.17L < 1
9 9
20% (C-1.3D)/2.17L < 0
23 25 8
10% 7 7
5 5 4
0% 1 0 0 0 2 0
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
TEXT STRUC LEFE CRG LEFE LEFE LEFE DCP LEFE LEFE
Bowles SPT/TCP PMT FWD

FIGURE 6.13. LOAD RATING GOODNESS PLOT FOR SHALLOWATER CULVERT: 2’ OF FILL

6. Critical Section Analysis

It is important to note where the problem areas occur for the live load prediction and
the operating rating. Problem sections for the live load predictions under-predict the live
load moment. Problem sections for the load rating indicate that the in-service culvert
should have failed under dead or live load. Qualitatively, the relationship between live
load demand and load rating is inverse. This means that when the live load demand is
low, the load rating will be high, and when the live load demand is high, the load rating
will be low. Further, this means that the problem sections for the live load predictions,
i.e. those sections where the live load demands are underestimated, will not produce the
controlling, problematic, lowest load rating. In fact, the load rating should be controlled
by the most conservative live load prediction.
To illustrate this point, the following tables show all the critical sections in every
model for the Shallowater Culvert. Table 6.43 highlights the problem critical sections for
the predicted live load demand. Table 6.4 shows the problem critical sections for the
load rating. Careful examination of

TxDOT 0-5849 188

 Table 6.3 and Table 6.4 reveals that there is no overlap between the two sets of
problem critical sections for the RISA-2D with spring models and most of the RISA-2D
with LEFE models. Critical sections for live load are the same as critical sections
predicted by the CULV-5 model at the wall top exterior corner WTEC and RISA-2D
LEFE model with the pressuremeter soil values at the bottom interior midspan. These are
the least accurate and most conservative models at non-critical sections.

TxDOT 0-5849 189

 TABLE 6.3. LIVE LOAD DEMAND PROBLEM CRITICAL SECTIONS FOR SHALLOWATER CULVERT: 2’ OF FILL
CULV-5 RISA-2D TEXT RISA-2D STRUC RISA-2D LEFE CRG RISA-2D LEFE Bowles RISA-2D LEFE SPT/TCP RISA-2D LEFE DCP RISA-2D LEFE PMT RISA-2D LEFE FWD
TL RLL TL RLL TL RLL
max min max min max min max min max min max min max min max min max min max min max min max min
WBEC 6.9 13.7 26.8 11.2 8.6 11.1 16.1 10.1 9.3 12.3 16.7 11.2 10.0 5.5 29.5 8.7 59.0 12.2 47.4 10.9 77.8 14.4 16.2 6.7
WEM 24.5 5.5 11.8 6.9 16.0 3.4 3.5 4.7 15.9 3.5 3.3 4.9 5.1 0.8 6.9 2.1 7.7 2.9 7.5 2.7 7.9 3.2 6.0 1.3
WTEC 11.1 2.3 28.1 2.0 12.4 1.6 19.0 1.5 14.5 1.6 21.1 1.5 12.4 1.3 28.7 1.4 50.8 1.6 42.2 1.6 65.0 1.8 18.1 1.4
TEC 4.0 8.6 14.6 8.1 9.2 4.0 18.6 3.5 11.2 4.0 20.6 3.5 11.1 3.2 26.1 3.5 47.6 4.2 38.9 3.9 61.9 4.7 16.3 3.3
TEM 15.1 26.1 15.4 23.3 9.3 15.9 9.6 13.2 9.4 15.8 9.6 13.1 6.3 9.0 7.8 10.9 8.6 11.9 8.4 11.6 8.8 12.2 7.0 10.0
TIC1 -13.0 5.9 -9.0 6.0 -14.0 2.8 -10.3 2.8 -14.8 2.8 -11.2 2.9 -2.7 1.9 -10.5 2.2 -31.9 2.5 -21.6 2.4 -48.6 2.8 -4.7 2.1
WBIC1 4.6 4.1 4.7 4.0 2.8 1.2 2.8 1.3 2.8 1.1 2.9 1.1 1.8 1.8 2.9 2.5 5.2 3.7 4.2 3.2 7.0 4.8 2.2 2.1
WIM1 15.9 14.4 16.3 14.0 7.7 6.6 8.0 6.3 8.0 6.6 8.3 6.3 4.2 4.3 5.5 5.3 6.2 5.6 6.0 5.6 6.4 5.6 4.8 4.8
WTIC1 2.3 2.2 2.3 2.2 1.7 2.0 1.7 2.0 1.6 2.0 1.7 2.0 1.4 1.5 1.9 2.1 2.3 2.8 2.1 2.5 2.5 3.3 1.6 1.7
BEC 17.6 31.6 30.6 28.9 4.5 13.5 15.7 11.0 5.4 15.3 16.4 12.9 10.7 7.7 31.2 12.7 61.3 18.5 49.6 16.3 80.7 21.9 17.4 9.6
BEM 192.0 10.5 197.4 9.2 92.5 1.8 95.0 0.7 104.2 1.9 107.3 0.7 25.0 0.1 44.7 0.5 57.2 1.3 53.5 1.0 61.3 2.0 33.5 0.2
BIC1 70.3 71.8 23.9 24.5 25.9 26.5 20.1 28.9 38.0 34.6 43.3 23.7
TIC2 63.3 18.1 58.1 18.3 38.2 8.0 33.8 8.1 42.7 8.1 38.4 8.3 5.9 5.4 15.5 6.3 50.2 7.0 32.2 6.6 75.9 7.6 9.1 5.8
TIM1 10.4 16.2 10.4 16.4 6.7 6.1 6.7 6.3 6.7 6.1 6.7 6.3 4.5 3.6 5.5 4.2 6.0 4.8 5.9 4.6 6.3 5.2 5.0 3.8
TIC3 105.3 14.1 107.8 14.0 60.9 6.3 63.1 6.2 64.8 6.2 67.0 6.2 6.3 4.7 22.3 5.2 62.6 5.5 42.7 5.4 94.5 5.7 12.2 5.0
WBIC2 10.2 8.4 10.2 8.4 4.4 3.5 4.4 3.5 4.5 3.6 4.4 3.5 4.0 3.1 5.9 4.7 9.3 7.3 7.9 6.3 11.4 9.0 4.6 3.7
WIM2 15.5 20.7 15.5 20.7 6.1 7.5 6.1 7.5 6.1 7.5 6.1 7.6 4.1 5.1 5.0 6.2 5.3 6.6 5.3 6.5 5.5 6.7 4.6 5.7
WTIC2 3.6 3.3 3.6 3.3 2.9 2.5 2.9 2.5 2.9 2.5 2.9 2.5 2.3 2.0 3.0 2.6 3.7 3.1 3.4 2.9 4.1 3.5 2.6 2.2
BIC2 33.0 33.3 10.5 10.8 11.4 11.6 9.1 12.9 15.9 14.9 17.2 10.7
BIM1 223.9 10.0 222.9 10.1 49.7 2.4 50.1 2.5 53.2 2.6 53.4 2.7 31.8 0.0 43.0 0.0 38.8 0.0 41.7 0.0 36.8 0.1 38.1 0.0
BIC3 185.4 35.1 190.4 34.9 60.7 12.5 64.2 12.4 61.7 13.3 65.2 13.3 31.8 10.3 84.1 14.6 151.0 18.4 126.7 17.1 186.2 20.4 49.3 12.1

TABLE 6.4. OPERATING RATING PROBLEM CRITICAL SECTIONS FOR SHALLOWATER CULVERT: 2’ OF FILL
CULV-5 RISA-2D TEXT RISA-2D STRUC RISA-2D LEFE CRG RISA-2D LEFE Bowles RISA-2D LEFE SPT/TCP RISA-2D LEFE DCP RISA-2D LEFE PMT RISA-2D LEFE FWD
TL RLL TL RLL TL RLL
max min max min max min max min max min max min max min max min max min max min max min max min
WBEC 26.12 0.05 5.41 0.47 14.94 0.65 7.21 0.88 14.35 0.65 7.22 0.88 15.65 0.75 7.36 -0.39 4.82 -0.96 5.46 -0.79 4.19 -1.18 11.12 0.18
WEM 0.67 1.88 3.19 1.17 0.91 3.16 9.84 1.78 0.97 3.16 10.75 1.78 3.90 13.17 3.50 4.64 3.74 3.15 3.67 3.50 3.80 2.83 3.51 7.49
WTEC 11.22 0.26 3.54 0.51 7.62 0.68 4.32 0.86 6.44 0.68 3.83 0.86 5.35 1.27 1.20 1.51 -0.38 1.88 0.06 1.75 -0.87 2.06 3.13 1.33
TEC 41.47 0.07 10.17 0.18 13.34 0.71 5.73 1.02 10.88 0.71 5.14 1.02 8.61 1.35 2.45 1.69 0.25 2.05 0.84 1.93 -0.38 2.20 5.27 1.45
TEM 0.20 3.21 0.17 3.79 0.33 5.22 0.28 6.61 0.34 5.22 0.29 6.61 0.64 10.37 0.43 9.14 0.36 8.65 0.38 8.79 0.34 8.50 0.53 9.66
TIC1 13.50 0.24 20.26 0.21 9.86 0.92 13.91 0.83 9.45 0.92 12.96 0.83 52.93 1.36 16.95 0.70 7.11 0.06 9.61 0.29 5.34 -0.28 33.09 1.06
WBIC1 0.94 1.16 0.87 1.25 1.65 3.59 1.66 3.47 1.64 3.59 1.64 3.47 2.91 2.10 2.48 0.78 2.19 -0.60 2.27 -0.13 2.12 -1.17 2.71 1.59
WIM1 0.83 1.04 0.78 1.11 1.73 2.27 1.62 2.43 1.68 2.27 1.57 2.43 3.14 3.54 2.29 3.01 2.00 2.89 2.06 2.91 1.95 2.88 2.69 3.23
WTIC1 0.60 0.70 0.57 0.74 0.78 0.82 0.70 0.90 0.79 0.82 0.71 0.90 0.74 1.27 0.18 1.32 -0.45 1.50 -0.21 1.44 -0.81 1.60 0.52 1.25
BEC 15.05 -0.12 7.92 0.03 44.77 0.70 11.50 1.25 38.11 0.70 11.34 1.25 20.95 0.72 9.03 -0.48 5.65 -1.06 6.49 -0.88 4.82 -1.29 14.17 0.10
BEM 0.41 4.11 0.35 4.94 1.08 21.49 1.00 53.56 0.93 21.49 0.85 53.56 4.23 524.94 1.33 122.09 0.65 47.47 0.79 64.01 0.55 32.15 2.52 308.44
BIC1 13.86 0.08 18.94 0.04 22.89 1.31 38.88 1.25 23.95 1.31 41.30 1.25 39.88 1.92 11.54 1.38 4.11 1.34 5.94 1.33 2.29 1.38 23.41 1.60
TIC2 7.05 0.24 7.82 0.23 8.99 0.97 10.36 0.93 8.12 0.97 9.19 0.93 56.46 1.60 22.74 1.27 6.86 1.19 10.91 1.22 4.27 1.18 37.86 1.43
TIM1 0.24 1.75 0.25 1.72 0.38 4.59 0.39 4.43 0.39 4.59 0.40 4.43 0.70 9.01 0.56 7.82 0.56 6.47 0.56 6.94 0.59 5.65 0.61 8.51
TIC3 3.61 0.29 3.49 0.30 4.77 1.10 4.58 1.12 4.52 1.10 4.35 1.12 44.49 1.66 13.61 1.33 5.61 0.98 7.74 1.13 4.16 0.72 23.58 1.50
WBIC2 0.94 0.90 0.94 0.90 2.15 2.15 2.18 2.18 2.13 2.15 2.20 2.18 2.42 2.42 1.62 1.62 1.04 1.04 1.22 1.22 0.85 0.85 2.08 2.08
WIM2 0.90 0.83 0.90 0.83 2.29 2.29 2.28 2.28 2.31 2.29 2.29 2.28 3.41 3.41 2.80 2.80 2.63 2.63 2.67 2.67 2.58 2.58 3.07 3.07
WTIC2 0.63 0.59 0.63 0.59 0.78 0.78 0.79 0.79 0.78 0.78 0.79 0.79 1.01 1.01 0.76 0.76 0.63 0.63 0.67 0.67 0.56 0.56 0.89 0.89
BIC2 8.33 0.09 9.15 0.08 10.42 1.31 12.00 1.24 10.40 1.31 11.94 1.24 20.60 1.73 8.76 0.97 5.25 0.58 6.11 0.70 4.41 0.39 13.91 1.34
BIM1 0.50 2.50 0.51 2.45 2.75 9.22 2.72 8.97 2.54 9.22 2.52 8.97 4.95 NA 2.80 NA 2.17 NA 2.36 NA 1.63 587.61 3.71 NA
BIC3 5.38 0.13 5.19 0.14 11.63 1.31 10.95 1.33 11.73 1.31 11.04 1.33 20.95 1.88 8.28 1.24 4.40 1.05 5.39 1.09 3.36 1.03 13.92 1.53

TxDOT 0-5849 190

 The meaningful conclusion from this analysis is that CULV-5 should be used for
design purposes. In general, the CULV-5 model overestimates the demands and
produces safe structures. However, for load rating, a much less conservative model,
even an unconservative model, is desirable, because it will reduce over-conservatism in
the controlling critical sections without actually introducing unconservatism into the load
rating.
7. Deflections

Deflection data were collected as a backup measure. Because the strain data provided
such consistent and accurate results, only a cursory look was given at the measured
deflections under live load. Figure 6.14 shows the measured deflection in the culvert at
the midspans. In the top spans the maximum deflection inward (positive) was between
0.84 and 0.64 millimeters. Some very slight outward deflection (negative) was also
recorded. Deflections in the bottom slabs were essentially negligible.

1.000

0.800
midspan deflection (mm)

0.600

0.400

0.200

0.000

-0.200
TES BES TIS BIS
Maximum Deflection 0.839 0.022 0.637 0.016
Minimum Deflection -0.044 -0.002 -0.082 -0.028
FIGURE 6.14. MEASURED DEFLECTIONS IN MILLIMETERS FOR THE SHALLOWATER CULVERT: 2’ OF FILL

TxDOT 0-5849 191

 8. Summary of Findings for the Shallowater Culvert with 2’ of Fill

From this case study the following findings are noted:

1. Predicted sign for live load moment corresponds to actual moment sign indicating
excellent agreement between reality and the models.
2. Predicted magnitude for live load moment typically higher than actual moment
magnitude and therefore conservative.
3. Scatter for predicted dead load moment is higher than for predicted live load
moment (stdev = 0.98/0.56).
4. Critical sections for live load are not the same as critical sections for the
Operating Rating.
5. CULV-5 yields most conservative moments and among the worst Operating
Rating.
6. RISA-LEFE (Es per Culvert Rating Guide) yields least conservative moments and
highest Operating Rating.

TxDOT 0-5849 192

 5. Results for Shallowater Culvert with 4’ of Fill

1. Load Rating

Figure 6.15 shows the normal load rating for the Shallowater culvert under four feet
of fill using all the different models and soil properties. Only one method produced
positive ratings and no method produced a load rating, inventory or operating, above HS-
20. The Level 3 analysis using the Culvert Rating Guide soil modulus shows that the
culvert should fail under the slightest live load. All other models show that the culvert
should fail under four feet of fill.

25.0
20.0
15.0
Inventory Rating, IR (HS-)

10.0
5.0
0.0
-5.0
-10.0
-15.0
-20.0
-25.0
-30.0
-35.0
-40.0
-45.0
-50.0
Lvl 1; Lvl 2; Lvl 2; Lvl 3; Lvl 3; Lvl 3; Lvl 3; Lvl 3; Lvl 3; Lvl 3;
TEXT TEXT; STRUC; CRG; Bowles; SPT; 2ksi TCP; 2ksi DCP; 3ksi PMT; FWD
k=150pci k=150pci 20ksi 6ksi 1ksi
Depth of Fill, DOF (ft)

Series1 Series2

FIGURE 6.15. LOAD RATING FOR SHALLOWATER CULVERT: 4’ OF FILL

As already noted, the Shallowater culvert remains in service and clearly has not
failed, this despite the fact that the load rating process would predict otherwise. Favorable
environmental conditions, conservative structural properties, soil properties, and
constitutive models for both the reinforced concrete and the soil can reasonably explain
why this culvert has “stood the test of time.”

TxDOT 0-5849 193

 2. Live Load Moment – Predicted and Measured

Figure 6.16 shows the live load moment envelope for the Shallowater culvert beneath
four feet of fill. The blue and red lines show the measured moment. The error bars show
a reasonable range due to variations in concrete modulus, cracked moment of inertia, and
the nature of four inch strain gages which measure average strain over the gage length
and not the actual strain at the critical section. The measured moment is less than the
predicted moment for all models.

5.000 CULV-5 MAX

4.000 CULV-5 MIN
3.000
RISA-2D TEXT MAX
2.000
1.000 RISA-2D TEXT MIN

0.000 RISA-2D STRUC MAX

-1.000
RISA-2D STRUC MIN
-2.000
RISA-2D LEFE CRG MAX
-3.000
-4.000 RISA-2D LEFE CRG MIN
-5.000 RISA-2D LEFE Bowles
-6.000 MAX
RISA-2D LEFE Bowles
MIN

FIGURE 6.16. LIVE LOAD MOMENTS PREDICTED AND MEASURED FOR SHALLOWATER CULVERT: 4’ OF FILL.

TxDOT 0-5849 194

 3. Dead Load Moment – Predicted

Figure 6.17 shows the predicted dead load distributions for the Shallowater culvert
beneath four feet of fill for all models. The models predict similar directions for dead
load effect, except for exterior corner locations.

8.000
6.000
CULV-5 MAX
4.000 CULV-5 MIN
2.000 RISA-2D TEXT MAX
0.000 RISA-2D TEXT MIN
-2.000 RISA-2D STRUC MAX
-4.000 RISA-2D STRUC MIN
-6.000 RISA-2D LEFE CRG

-8.000 RISA-2D LEFE Bowles

-10.000 RISA-2D LEFE SPT/TCP

-12.000 RISA-2D LEFE DCP

RISA-2D LEFE PMT
-14.000
RISA-2D LEFE FWD
BEC
BEM
WEM

WBIC1

WBIC2
WBEC

TEC
TEM

WTIC1

WTIC2
WTEC

WIM1

WIM2
BIC1

BIC2
TIM1
TIC1

TIC2

TIC3

BIM1
BIC3
FIGURE 6.17. DEAD LOAD PREDICTED FOR SHALLOWATER CULVERT: 4’ OF FILL.

TxDOT 0-5849 195

 4. Live Load “Goodness”

Figure 6.18 shows the goodness plot for the Shallowater culvert beneath four feet of
fill. The plot shows the number of critical sections which fall into each threshold.
According to the plot, the CULV-5 model produces the most conservative predictions.
The least conservative model is the RISA-2D with LEFE using the Culvert Rating Guide
soil modulus.

100%
90% 6 8
17 17
14
80%
7 20 20 20
70% 50
23 23 12
60% 10 < Mp/Mm
11 5 < Mp/Mm < 10
50%
18
40% 10 10 2 < Mp/Mm < 5
25 25 13 14
30% 15 1 < Mp/Mm < 2
12
0 < Mp/Mm < 1
20% 9 7
12 11 9 6 Mp/Mm < 0
16 6
10% 3
8 2 4 2
2 7 2 2 2
0% 1
0 0 0 0 0 1
0 1
0 1
0 0
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
(σ=18kft) TEXT STRUC LEFE CRG LEFE LEFE LEFE DCP LEFE PMT LEFE FWD
(σ=6kft) (σ=7kft) (σ=5kft) Bowles SPT/TCP (σ=13kft) (σ=21kft) (σ=6kft)
(σ=9kft) (σ=16kft)

FIGURE 6.18. LIVE LOAD GOODNESS PLOT FOR SHALLOWATER CULVERT: 4’ OF FILL

Also included in Figure 6.18 is the average standard deviation of the ratios for each of
the models. The smaller the average standard deviation, the better the model fit. By this
measure, CULV-5 and RISA-2D with LEFE using the pressurmeter (PMT) soil modulus
are the least appropriate for accurately predicting the live load moments. RISA-2D with
LEFE using the Culvert Rating Guide soil modulus values is the most appropriate model
with the least amount of scatter and the most accurate prediction of the live load moment
envelope shape.

TxDOT 0-5849 196

 5. Load Rating “Goodness”

Figure 6.19 shows the number of critical sections in each threshold for each model.
According to this plot, CULV-5 indicates that 25% of the critical sections in the structure
should fail under dead load, and in 10% of the critical sections, the culvert should have
failed under the truck load. RISA-2D with LEFE using the Culvert Rating Guide
however shows that nearly 50% of the culvert is well designed to handle the load and
only two critical sections should have failed under the trucks.

100%
7 5
90% 17 10
11 13
29 29 14 6
80%
8 6
70% 4 4
6
3 3 10 10 < (C-1.3D)/2.17L
60% 6
7 9 8 5 < (C-1.3D)/2.17L < 10
50% 23 19 19 10
4 8 2 < (C-1.3D)/2.17L < 5
40% 5
12 9 1 < (C-1.3D)/2.17L < 2
6
30% 9 8
22 21 9 0 < (C-1.3D)/2.17L < 1
9
20% 4 (C-1.3D)/2.17L < 0
21 6 6 11
10% 11 3
10 6 6
0% 1 1 2
0 2 2
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
TEXT STRUC LEFE CRG LEFE LEFE LEFE DCP LEFE LEFE
Bowles SPT/TCP PMT FWD

FIGURE 6.19. LOAD RATING GOODNESS PLOT FOR SHALLOWATER CULVERT: 4’ OF FILL

6. Critical Section Analysis

The load rating should be controlled by the most conservative live load prediction.
The following tables show all the critical sections in every model. highlights the problem
critical section for the predicted live load demand. Table 6.6 shows the problem critical
section for the load rating. Careful inspection of Table 6.5 and Table 6.6 reveals that
there is no overlap between the two sets of problem critical sections.

TxDOT 0-5849 197

 TABLE 6.5. LIVE LOAD DEMAND PROBLEM CRITICAL SECTIONS FOR SHALLOWATER CULVERT: 4’ OF FILL
CULV-5 RISA-2D TEXT RISA-2D STRUC RISA-2D LEFE CRG RISA-2D LEFE Bowles RISA-2D LEFE SPT/TCP RISA-2D LEFE DCP RISA-2D LEFE PMT RISA-2D LEFE FWD
TL RLL TL RLL TL RLL
max min max min max min max min max min max min max min max min max min max min max min max min
WBEC 1.9 22.1 5.8 16.2 0.2 10.3 3.3 7.7 0.4 11.1 3.5 8.6 2.4 8.8 8.1 14.1 17.5 20.2 13.8 17.8 23.8 23.9 4.2 10.8
WEM 7.4 7.3 2.6 11.2 5.5 2.8 0.7 6.6 5.5 3.0 0.7 6.9 1.1 1.2 1.6 3.8 1.8 5.7 1.7 5.1 1.9 6.3 1.4 2.3
WTEC 0.1 2.6 7.6 2.0 0.2 1.6 0.4 1.4 0.2 1.6 0.4 1.4 3.2 1.3 9.1 1.7 17.8 2.1 14.4 1.9 23.8 2.4 5.1 1.4
TEC 5.9 9.9 31.7 8.8 0.9 4.2 1.5 3.3 0.9 4.2 1.5 3.2 13.3 3.3 38.0 4.4 74.6 5.7 60.2 5.2 99.0 6.7 21.4 3.7
TEM 13.9 43.0 14.5 35.3 8.0 24.3 8.5 16.9 8.0 24.5 8.5 17.0 6.0 6.7 9.1 14.1 10.9 19.1 10.4 17.6 11.5 21.0 7.4 9.9
TIC1 5130.6 7.2 3349.4 7.5 1830.0 2.8 590.0 3.0 1790.0 2.9 550.0 3.1 580.0 2.2 920.0 2.9 7070.0 3.6 3830.0 3.3 12540.0 4.0 550.0 2.5
WBIC1 5.0 3.4 5.2 3.2 2.2 0.6 2.2 0.6 2.2 0.5 2.3 0.5 1.9 1.4 3.9 2.4 7.3 3.9 5.9 3.3 10.0 5.1 2.6 1.8
WIM1 13.4 16.6 13.9 15.9 5.7 6.7 6.1 6.1 5.9 6.7 6.3 6.2 2.9 4.0 4.6 6.0 5.7 6.9 5.4 6.7 6.1 7.0 3.7 4.9
WTIC1 2.1 2.3 2.2 2.2 1.2 1.9 1.3 1.8 1.2 2.0 1.3 1.8 1.2 1.5 2.0 2.5 2.7 3.7 2.4 3.2 3.1 4.5 1.6 1.9
BEC 1.8 43.5 8.8 37.8 1.5 11.7 4.1 6.4 1.2 13.0 4.4 7.7 3.4 10.9 11.1 18.1 23.1 26.7 18.3 23.4 31.2 32.1 5.8 13.6
BEM 53.5 15.0 56.2 12.0 22.7 3.7 24.1 1.0 25.5 3.9 27.2 1.0 9.1 0.2 15.9 0.9 20.3 2.2 19.0 1.7 21.7 3.3 12.0 0.3
BIC1 32.0 68.2 19.7 70.7 14.3 23.8 5.9 24.8 14.2 25.6 5.7 26.7 5.4 22.1 21.7 33.0 53.9 44.5 40.6 40.2 77.2 51.4 9.5 26.4
TIC2 26.4 14.8 23.4 15.1 4.0 5.8 1.6 6.0 4.2 5.9 1.8 6.2 2.6 4.4 6.7 5.8 13.5 7.0 10.7 6.5 21.6 7.9 3.9 5.1
TIM1 7.8 46.3 7.7 47.1 4.4 12.5 4.4 13.2 4.5 12.7 4.4 13.4 3.5 4.9 5.2 10.7 6.2 15.7 5.9 14.0 6.6 18.2 4.3 7.2
TIC3 49.4 3.9 51.6 3.9 6.5 1.5 7.8 1.5 7.4 1.5 8.7 1.5 3.6 1.2 6.9 1.5 27.4 1.8 14.6 1.7 48.6 1.9 4.9 1.4
WBIC2 12.5 5.9 12.5 5.9 5.0 2.2 4.9 2.2 4.9 2.2 4.8 2.2 5.0 2.3 9.4 4.2 15.7 7.1 13.2 6.0 19.5 8.8 6.6 3.0
WIM2 21.2 13.6 21.2 13.6 7.0 4.3 7.1 4.4 7.1 4.4 7.1 4.4 4.5 2.8 6.8 4.2 7.9 4.9 7.6 4.7 8.1 5.0 5.6 3.5
WTIC2 3.0 3.5 3.0 3.5 2.0 2.4 2.0 2.3 2.1 2.4 2.1 2.4 1.8 2.1 2.9 3.3 3.9 4.6 3.6 4.1 4.5 5.2 2.2 2.6
BIC2 14.0 54.1 12.1 55.1 7.4 18.3 5.8 18.9 7.6 19.5 6.0 20.2 3.6 17.6 11.1 25.9 22.5 32.7 18.1 30.4 29.8 35.9 6.0 21.1
BIM1 34.8 13.2 34.5 13.5 8.5 2.5 8.6 2.6 9.2 2.7 9.3 2.9 6.2 0.0 8.3 0.0 7.3 0.0 8.0 0.0 6.7 0.0 7.4 0.0
BIC3 27.8 45.8 29.3 45.3 7.9 16.6 8.9 16.4 8.1 17.7 9.1 17.5 4.8 15.1 14.6 22.1 28.7 28.5 23.3 26.2 36.6 32.0 7.9 18.0

TABLE 6.6. OPERATING RATING PROBLEM CRITICAL SECTIONS FOR SHALLOWATER CULVERT: 4’ OF FILL
CULV-5 RISA-2D TEXT RISA-2D STRUC RISA-2D LEFE CRG RISA-2D LEFE Bowles RISA-2D LEFE SPT/TCP RISA-2D LEFE DCP RISA-2D LEFE PMT RISA-2D LEFE FWD
TL RLL TL RLL TL RLL
max min max min max min max min max min max min max min max min max min max min max min max min
WBEC -7.00 -0.61 13.77 0.03 384.38 0.76 18.55 1.69 174.02 0.76 18.36 1.69 32.87 0.12 13.58 -1.43 8.23 -2.17 9.53 -1.95 6.95 -2.45 21.83 -0.68
WEM 1.15 3.87 7.46 1.74 1.28 10.61 25.31 3.17 1.37 10.61 26.87 3.17 5.74 24.29 5.96 7.30 6.88 4.50 6.62 5.14 7.14 3.91 5.49 12.88
WTEC -1.99 -0.10 10.34 0.41 5.50 0.65 174.45 1.09 6.31 0.65 165.41 1.09 15.72 1.95 3.85 1.99 0.13 2.44 1.11 2.25 -0.93 2.68 9.13 1.85
TEC 89.97 -0.42 15.04 -0.22 3.40 0.73 213.50 1.58 3.85 0.73 212.11 1.58 20.44 2.18 5.35 2.28 0.92 2.74 2.07 2.57 -0.29 2.95 11.88 2.12
TEM 0.00 7.20 -0.08 9.23 0.03 12.58 -0.09 19.02 0.05 12.58 -0.08 19.02 1.03 40.92 0.31 22.91 0.10 18.21 0.15 19.43 0.04 16.98 0.61 30.20
TIC1 21.90 -0.25 34.65 -0.31 45.03 0.86 145.33 0.65 46.55 0.86 157.56 0.65 128.50 1.78 105.08 0.28 17.42 -0.80 29.46 -0.44 11.20 -1.31 151.01 1.08
WBIC1 1.60 2.18 1.44 2.41 4.16 11.25 4.18 10.11 4.14 11.25 4.08 10.11 5.49 4.05 3.76 1.13 3.22 -1.04 3.34 -0.33 3.12 -1.92 4.55 2.72
WIM1 1.39 1.88 1.26 2.07 3.37 4.60 3.01 5.20 3.24 4.60 2.89 5.20 6.78 7.74 3.88 5.45 3.03 4.90 3.23 5.02 2.87 4.82 5.18 6.42
WTIC1 1.09 1.42 1.00 1.54 1.71 1.81 1.41 2.09 1.71 1.81 1.41 2.09 1.50 2.65 0.18 2.26 -0.86 2.38 -0.49 2.31 -1.42 2.49 0.89 2.38
BEC 100.85 -0.83 18.44 -0.62 5.77 0.73 26.24 3.17 6.47 0.73 25.24 3.17 39.37 0.05 15.38 -1.55 9.12 -2.29 10.65 -2.06 7.61 -2.57 25.65 -0.78
BEM 0.20 8.82 0.08 11.59 1.30 29.73 1.11 109.59 1.02 29.73 0.83 109.59 4.33 770.64 0.44 183.35 -0.47 80.18 -0.27 103.64 -0.60 54.82 2.04 436.46
BIC1 29.14 -0.56 48.78 -0.62 41.61 1.22 102.98 1.10 43.29 1.22 108.82 1.10 99.98 2.03 25.15 1.30 8.43 1.38 12.26 1.31 4.76 1.53 58.13 1.57
TIC2 14.13 -0.19 16.21 -0.23 66.28 1.03 174.34 0.93 64.09 1.03 155.38 0.93 89.80 2.25 38.68 1.29 19.46 1.05 24.65 1.09 11.71 1.04 63.60 1.71
TIM1 0.16 4.22 0.18 4.11 0.30 15.59 0.33 14.58 0.33 15.59 0.36 14.58 1.14 38.48 0.57 19.07 0.50 12.88 0.50 14.61 0.53 10.72 0.79 27.32
TIC3 9.63 -0.10 9.13 -0.08 52.77 1.25 43.89 1.29 46.84 1.25 39.52 1.29 82.51 2.38 49.21 1.38 14.37 0.70 25.52 0.96 8.98 0.27 64.20 1.85
WBIC2 1.81 1.73 1.81 1.73 4.56 4.55 4.62 4.61 4.72 4.55 4.76 4.61 4.55 4.55 2.43 2.43 1.46 1.46 1.73 1.73 1.17 1.17 3.47 3.47
WIM2 1.56 1.51 1.56 1.51 4.74 4.74 4.69 4.69 4.73 4.74 4.69 4.69 7.41 7.41 4.92 4.92 4.25 4.25 4.41 4.40 4.13 4.12 5.98 5.98
WTIC2 1.21 1.20 1.21 1.20 1.76 1.76 1.77 1.77 1.73 1.76 1.74 1.77 2.04 2.04 1.25 1.25 0.92 0.92 1.01 1.01 0.80 0.80 1.62 1.62
BIC2 19.45 -0.49 22.76 -0.52 24.52 1.20 31.77 1.07 24.40 1.20 31.33 1.07 47.73 1.75 17.55 0.67 9.70 0.17 11.58 0.33 7.83 -0.07 30.16 1.18
BIM1 0.43 5.40 0.45 5.25 2.83 25.27 2.78 23.89 2.47 25.27 2.45 23.89 5.36 NA 2.32 NA 1.02 545.22 1.49 NA -0.03 -4.93 3.62 NA
BIC3 13.40 -0.39 12.63 -0.37 31.15 1.32 27.57 1.36 31.32 1.32 27.67 1.36 47.44 2.05 16.67 1.17 8.28 0.98 10.37 1.01 6.18 0.99 30.01 1.55

TxDOT 0-5849 198

 The meaningful conclusion from this analysis is that CULV-5 should be used for
design purposes. In general, they overestimate the demands and produce safe structures.
However, for load rating, a much less conservative model, even an unconservative
model, is desirable, because it will reduce the conservatism in the controlling critical
sections without actually introducing unconservatism into the load rating.
7. Deflections

Deflection data were collected as a backup measure. Figure 6.20 shows the measured
deflection in the culvert at the midspans. The live load deflections for four feet of fill are
roughly half the magnitude of the live load deflections for two feet of fill. This
corresponds with a conceptual understanding of increased load distribution through more
fill.

0.500

0.400
midspan deflection (mm)

0.300

0.200

0.100

0.000

-0.100
TES BES TIS BIS
Maximum Deflection 0.438 0.021 0.407 0.012
Minimum Deflection -0.041 -0.003 -0.031 -0.034
FIGURE 6.20. MEASURED DEFLECTIONS IN MILLIMETERS FOR THE SHALLOWATER CULVERT: 4’ OF FILL

TxDOT 0-5849 199

 8. Summary of Findings for the Shallowater Culvert with 4’ of Fill

From this case study the following findings are noted:

1. Predicted sign for live load moment corresponds to actual moment sign indicating
excellent agreement between reality and the models.
2. Predicted magnitude for live load moment typically higher than actual moment
magnitude and therefore conservative.
3. Scatter for predicted dead load moment is higher than for predicted live load
moment (stdev = 1.15/0.44).
4. Critical sections for live load are not the same as critical sections for Operating
Rating for all but CULV-5 WTEC and RISA-2D LEFE PMT BIM and these cases
are not the controlling Operating Rating.
5. CULV-5 yields most conservative moments and among the worst Operating
Rating.
6. RISA-LEFE (Es per Culvert Rating Guide) yields least conservative moments and
highest Operating Rating.

TxDOT 0-5849 200

 5. THE PLAINVIEW CULVERT INSTRUMENTED LOAD TEST

1. Culvert Condition

The Plainview culvert is located in Hale County on SH-194 approximately 100yd

south of FM-70 in Plainview, Texas. Figure 6.21 shows the location of the Plainview test
culvert. This culvert was built in 1991. The design is a Post-1977, 1958 era culvert
design. It consists of four 10’ wide by 6’ tall barrels. The culvert is oriented
perpendicular to the roadway underneath five lanes of traffic. The culvert showed very
little distress.

FIGURE 6.21. PLAINVIEW LOCATION.

TxDOT 0-5849 201

 2. Summary of Parameters

1. Culvert Test Parameters

Culvert dimensional data from the plan sheets showed 4 barrels, 10ft spans, 6ft
height, 8.5in. thick top and bottom slabs, 7in. thick walls and 60ksi steel yield strength.
According to the MCEB, 3ksi concrete was assumed for Level 1 and 2 analyses. For
Level 2 and 3 analyses the tested compressive strength of 8ksi from the concrete cores
was used. The truck weighed 47kips. The front single axles weighed 11.5kips, while the
rear tandem axles weighed 35.5kips.
2. Soil Test Parameters

A soil unit weight of 120pcf was selected from the Culvert Rating Guide. A modulus
of subgrade reaction, k, of 150pci from Culvert Rating Guide was used for Level 2
analysis. The modulus of elasticity for the soil, Es; was determined for a two layer soil
system using seven different methods. The first modulus value represents the layer from
the ground surface to slightly below the bottom of the culvert. The second modulus value
is used for the native soil beneath the culvert.
1. 8 / 20 ksi from Culvert Rating Guide
2. 6 / 9 ksi from Bowles and McCarthy textbooks
3. 23 / 5 ksi from Unconfined Compressive Strength (UCS) and Standard
Penetration Test (SPT)
4. 25 / 10 ksi from Texas Cone Penetration Test (TCP)
5. 23 / 4 ksi from Dynamic Cone Penetrometer (DCP)
6. 0.9 / 0.8 ksi from Pressuremeter (PMT)
7. 8 / 8 ksi from Falling Weight Deflectometer (FWD)
Testing was performed under three and a half foot of fill including the pavement surface.

TxDOT 0-5849 202

 3. Instrumentation Plan

The instrumentation plan involved placing 4in. electrical resistance strain gages at
every critical section on the inside of the two northernmost barrels. An attempt was made
to place strain gages on the exterior of the top slab and at the top corner exterior walls
opposite the corresponding interior gages. However, there was greater fill depth than
expected making gage placement impossible without causing significant damage to the
pavement surface. Linear displacement gages were placed at the centerline of the top and
bottom slabs. Figure 6.22 shows the location of the strain gages (in purple) and the linear
displacement gages (in blue).

7" 4'-0" 1'-0" 1'-0" 4'-0" 7" 4'-0" 1'-0" 1'-0" 4'-0"

1"
82 ITEC ITEM ITIC1 ITIC2 ITIM ITIC3
IWTEC IWTIC1 IWTIC2 IWTIC3 IWTIC4
LVDT TES LVDT TIS
3'-0"

MATCH LINE
IWEM IWIM1 IWIM2 IWIM3 IWIM4

3'-0"
LVDT BES LVDT BIS
IWBEC IWBIC1 IWBIC2 IWBIC3 IWBIC4

IBEC IBEM IBIC1 IBIC2 IBIM IBIC3

FIGURE 6.22. PLAINVIEW INSTRUMENTATION PLAN.

TxDOT 0-5849 203

 4. Load Rating

Figure 6.23 shows the normal load rating for the Plainview culvert under three and a
half feet of fill using all the different models and soil properties. All but one method
produced positive ratings. Seven of the analysis methods produced Operating Ratings
great than HS-20. Four of the methods produced inventory ratings above HS-20.

40

30
Inventory Rating, IR (HS-)

20

10

-10

-20
Lvl1; Lvl 2; Lvl 2; Lvl 3; Lvl 3; Lvl 3; Lvl 3; Lvl 3; Lvl 3; Lvl 3;
TEXT TEXT; STRUC; CRG; Bowles; UCS/SPT; TCP; DCP; PMT; FWD; 8
k=150pci k=150pci 8/20 6/9 23/5 25/10 23/4 .9/.8
Depth of Fill, DOF (ft)

Inventory Rating Operating Rating

FIGURE 6.23. LOAD RATING FOR PLAINVIEW CULVERT: 3.5’ OF FILL

TxDOT 0-5849 204

 5. Live Load Moment – Predicted and Measured

Figure 6.24 shows the live load moment envelope for the Plainview culvert beneath
three and a half feet of fill. The blue and red lines show the measured moment as well as
error bars which show a reasonable range due to variations in concrete modulus, cracked
moment of inertia, and the nature of four inch strain gages which measure average strain
over the gage length and not the actual strain at the critical section. The measured
moment is less than the predicted moment for all models.

8.000 CULV-5 MAX

6.000 CULV-5 MIN

4.000 RISA-2D TEXT MAX

2.000 RISA-2D TEXT MIN

RISA-2D STRUC MAX

0.000
RISA-2D STRUC MIN
-2.000
RISA-2D LEFE CRG MAX
-4.000
RISA-2D LEFE CRG MIN
-6.000
RISA-2D LEFE Bowles
-8.000 MAX
RISA-2D LEFE Bowles
MIN

FIGURE 6.24. LIVE LOAD MOMENTS PREDICTED AND MEASURED FOR PLAINVIEW CULVERT: 3.5’ OF FILL.

TxDOT 0-5849 205

 6. Dead Load Moment – Predicted

Figure 6.25 shows the predicted dead load distributions for the Plainview culvert
beneath three and a half feet of fill for all models. The models predict similar directions
for the dead load effect, except for exterior corner locations.

6.000
CULV-5 MAX
4.000
CULV-5 MIN
2.000 RISA-2D TEXT MAX
0.000 RISA-2D TEXT MIN
RISA-2D STRUC MAX
-2.000
RISA-2D STRUC MIN
-4.000
RISA-2D LEFE CRG
-6.000 RISA-2D LEFE Bowles

-8.000 RISA-2D LEFE UCS/SPT

RISA-2D LEFE TCP
-10.000
RISA-2D LEFE DCP
-12.000 RISA-2D LEFE PMT
WEM
WBEC

TEC
TEM

WBIC1

BEC
BEM

WBIC2
WTEC

WTIC1

WTIC2
WIM1

WIM2

BIM1
BIC1
TIC1

TIC2
TIM1

BIC2

BIC3
TIC3

RISA-2D LEFE FWD

FIGURE 6.25. DEAD LOAD PREDICTED FOR PLAINVIEW CULVERT: 3.5’ OF FILL.

TxDOT 0-5849 206

 7. Live Load “Goodness”

Figure 6.26 shows the goodness plot for the Plainview culvert beneath three and a
half feet of fill. The plot shows the number of critical sections which fall into each
threshold. According to the plot, CULV-5 model produces the most conservative
predictions. The least conservative model is the RISA-2D with LEFE using the Culvert
Rating Guide and the Texas Cone Penetrometer soil moduli.

100%
9 9 4 6 6
90% 9 10 8
80% 10
20 20 9 19
45
70% 14 12
13 13 10 < Mp/Mm
60%
50% 5 < Mp/Mm < 10
31 33
40% 20 19 2 < Mp/Mm < 5
15 13
14 12 15
30% 12 1 < Mp/Mm < 2

12 0 < Mp/Mm < 1

20% 15 14
4 6
5 5 6 4 Mp/Mm < 0
10% 8 5
9 8 4 2
4 3 3 2 2 2 3
0% 0 0 0 0 0 0 0 0 0 0
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
(σ=14kft) TEXT STRUC LEFE CRG LEFE LEFE LEFE TCP LEFE DCP LEFE LEFE
(σ=3kft) (σ=4kft) (σ=3kft) Bowles USC/SPT (σ=4kft) (σ=5kft) PMT FWD
(σ=4kft) (σ=5kft) (σ=14kft) (σ=4kft)

FIGURE 6.26. LIVE LOAD GOODNESS PLOT FOR PLAINVIEW CULVERT: 3.5’ OF FILL

Also included in Figure 6.26 is the average standard deviation of the ratios for each of
the models. The smaller the average standard deviation, the better the model fit. By this
measure, CULV-5 and RISA-2D with LEFE using the Pressuremeter data are the least
appropriate for accurately predicting the live load moments. RISA-2D with LEFE using
the Culvert Rating Guide soil modulus values and RISA-2D with Springs using textbook
assumptions are the most appropriate models with the least amount of scatter and the
most accurate prediction of the live load moment envelope shape.

TxDOT 0-5849 207

 8. Load Rating “Goodness”

Figure 6.27 shows the number of critical sections in each threshold for each model.
Overall, the models produce very good load ratings. This is more likely due to the
adequate or even over-conservative structural design.

100%
90% 19 9
80% 33 33 15 15
18 18 19 19
6
70%
10
10 < (C-1.3D)/2.17L
60%
5 < (C-1.3D)/2.17L < 10
50% 35 12 12
8 7 2 < (C-1.3D)/2.17L < 5
40% 34 34 15 11
14 1 < (C-1.3D)/2.17L < 2
30%
11 0 < (C-1.3D)/2.17L < 1
20% 13 13 13
22 11 4 (C-1.3D)/2.17L < 0
10% 17 17 8 2
3
2 1 2 1
1 1 2 3 2
0% 0 0 0 0 0 0 0 0 0
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
TEXT STRUC LEFE LEFE LEFE LEFE LEFE LEFE LEFE
CRG Bowles UCS/SPT TCP DCP PMT FWD

FIGURE 6.27. LOAD RATING GOODNESS PLOT FOR PLAINVIEW CULVERT: 3.5’ OF FILL

9. Critical Section Analysis

IT IS IMPORTANT TO NOTE THE CRITICAL SECTIONS WHERE PROBLEM AREAS OCCUR FOR THE LIVE LOAD PREDICTION AND THE
OPERATING RATING. THE FOLLOWING TABLES SHOW ALL THE CRITICAL SECTIONS IN EVERY MODEL. TABLE 6.7 HIGHLIGHTS THE
PROBLEM CRITICAL SECTIONS FOR THE PREDICTED LIVE LOAD DEMAND. TABLE 6.8 SHOWS THE PROBLEM CRITICAL SECTIONS
FOR THE LOAD RATING. CAREFUL INSPECTION OF

TxDOT 0-5849 208

 Table 6.7 and Table 6.8 reveals that there is no overlap between the two sets of
problem critical sections. It is also meaningful to note that far fewer sections are problem
sections than for the older Shallowater culvert.

TxDOT 0-5849 209

 TABLE 6.7. LIVE LOAD DEMAND PROBLEM CRITICAL SECTIONS FOR PLAINVIEW CULVERT: 3.5’ OF FILL
CULV-5 RISA-2D TEXT RISA-2D STRUC RISA-2D LEFE CRG RISA-2D LEFE Bowles RISA-2D LEFE UCS/SPT RISA-2D LEFE TCP RISA-2D LEFE DCP RISA-2D LEFE PMT RISA-2D LEFE FWD
TL RLL TL RLL TL RLL
max min max min max min max min max min max min max min max min max min max min max min max min max min
WBEC 0.1 8.0 6.8 6.2 0.3 4.0 3.3 3.2 1.7 4.5 4.5 3.8 4.2 4.3 8.7 5.7 12.0 5.6 6.9 4.6 14.0 5.9 32.3 11.0 9.5 5.7
WEM 7.2 8.8 3.4 11.5 4.5 3.6 0.8 6.2 4.5 3.7 0.8 6.4 1.8 4.0 1.6 4.7 1.4 3.0 0.9 2.3 1.6 3.2 1.4 6.8 1.4 4.4
WTEC 1.1 0.9 9.6 0.8 0.5 0.5 4.0 0.4 1.0 0.5 4.6 0.4 6.3 0.6 11.4 0.6 14.9 0.6 9.6 0.5 16.8 0.6 36.1 1.0 12.0 0.6
TEC 1.0 44.2 1.9 40.1 1.1 13.9 1.6 10.2 0.7 14.2 1.8 10.5 2.7 16.8 5.0 18.4 6.4 19.1 4.2 16.6 7.2 20.2 15.1 31.4 5.2 18.7
TEM 2.1 5.5 2.1 4.9 1.1 2.8 1.2 2.2 1.1 2.8 1.2 2.2 1.3 2.2 1.4 2.4 1.1 1.6 1.0 1.4 1.1 1.6 1.7 2.9 1.4 2.2
TIC1 25.0 11.2 19.2 11.5 4.7 4.4 0.5 4.7 4.8 4.6 0.5 4.8 2.3 4.8 2.7 5.1 6.7 3.8 3.5 3.8 12.3 3.8 65.3 6.7 3.1 4.9
WBIC1 5.7 4.4 5.9 4.3 2.3 1.6 2.3 1.5 2.6 1.8 2.7 1.7 3.2 2.8 4.5 3.4 5.8 3.9 3.7 2.9 6.6 4.4 14.6 8.6 4.7 3.5
WIM1 17.5 8.3 17.9 8.0 6.6 3.1 7.0 2.9 6.9 3.2 7.3 2.9 6.0 3.0 6.7 3.3 5.2 2.6 4.8 2.5 5.3 2.6 8.2 3.5 6.5 3.2
WTIC1 3.2 3.0 3.2 2.9 2.3 2.4 2.4 2.3 2.3 2.6 2.4 2.4 2.8 2.8 3.2 3.4 2.9 3.4 2.4 2.7 3.1 3.7 5.3 7.0 3.1 3.4
BEC 0.6 30.2 1.9 26.7 0.8 8.2 1.5 5.0 0.1 9.6 2.1 6.5 2.1 9.8 4.2 13.3 5.9 13.3 3.5 10.9 6.9 14.1 14.6 27.1 4.6 13.3
BEM 44.6 28.2 46.0 24.8 16.5 4.6 17.3 1.4 18.6 4.9 19.5 1.8 10.1 0.1 13.3 0.7 11.3 1.6 10.4 0.5 11.3 2.1 17.2 6.7 13.0 0.8
BIC1 5415.7 20.9 4136.0 21.4 2800.0 6.9 1970.0 7.2 2610.0 7.7 1780.0 7.9 1530.0 8.1 2990.0 10.0 2960.0 10.3 1550.0 9.0 3530.0 10.8 13210.0 16.0 3050.0 10.1
TIC2 60.5 15.1 55.3 15.3 11.2 5.4 7.2 5.6 12.8 5.6 8.7 5.8 3.4 6.2 8.1 6.5 18.2 5.1 10.1 5.0 21.5 5.2 52.1 7.8 10.1 6.3
TIM1 8.9 15.8 8.9 15.9 4.8 5.2 4.8 5.4 4.9 5.3 4.8 5.5 5.8 4.3 6.2 4.8 5.0 2.9 4.6 2.6 5.1 3.0 7.8 7.6 6.1 4.5
TIC3 22.0 17.0 22.6 16.9 1.8 6.3 2.1 6.2 2.1 6.3 2.4 6.3 1.0 7.2 1.7 7.3 3.3 6.0 2.3 5.9 4.3 5.9 20.9 7.8 2.0 7.1
WBIC2 6.9 3.7 6.9 3.7 2.9 1.5 2.9 1.5 3.3 1.7 3.3 1.7 3.9 2.0 5.3 2.7 6.8 3.5 4.6 2.4 7.7 4.0 13.1 6.8 5.6 2.9
WIM2 11.7 10.9 11.7 10.9 3.6 3.3 3.6 3.3 3.6 3.3 3.6 3.3 3.7 3.4 4.0 3.6 3.2 2.9 3.0 2.8 3.2 2.9 4.4 4.0 3.9 3.5
WTIC2 1.5 1.8 1.5 1.8 1.1 1.3 1.1 1.3 1.1 1.3 1.1 1.3 1.3 1.5 1.5 1.8 1.5 1.8 1.2 1.5 1.6 2.0 2.5 3.0 1.5 1.8
BIC2 8387.2 18.7 7624.8 19.0 3660.0 6.2 2980.0 6.4 4070.0 6.8 3390.0 7.0 2910.0 7.4 5350.0 8.9 6650.0 9.1 4230.0 8.0 7540.0 9.4 15470.0 12.6 5710.0 9.0
BIM1 40.7 34.0 40.4 34.5 8.6 0.9 8.6 0.9 9.1 0.8 9.1 0.9 8.1 0.1 9.0 0.2 7.9 0.0 8.6 0.1 7.3 0.0 9.0 0.7 8.9 0.1
BIC3 15.5 19.3 16.0 19.1 4.1 6.5 4.6 6.5 4.6 7.1 5.0 7.0 3.8 7.6 6.7 9.2 8.0 9.6 5.2 8.4 8.9 10.0 17.4 13.9 7.1 9.3

TABLE 6.8. OPERATING RATING PROBLEM CRITICAL SECTIONS FOR PLAINVIEW CULVERT: 3.5’ OF FILL
CULV-5 RISA-2D TEXT RISA-2D STRUC RISA-2D LEFE CRG RISA-2D LEFE Bowles RISA-2D LEFE UCS/SPT RISA-2D LEFE TCP RISA-2D LEFE DCP RISA-2D LEFE PMT RISA-2D LEFE FWD
TL RLL TL RLL TL RLL
max min max min max min max min max min max min max min max min max min max min max min max min max min
WBEC 171.69 1.26 11.37 2.16 233.95 3.78 20.26 5.07 54.91 3.46 19.47 4.46 24.28 3.08 13.72 1.56 11.35 0.87 17.21 2.03 10.19 0.54 5.69 -0.64 12.95 1.45
WEM 3.66 4.54 9.72 3.05 5.53 11.47 39.36 5.83 5.70 11.43 40.97 5.94 14.19 10.76 17.00 8.75 16.52 15.08 23.83 19.35 14.34 14.23 21.90 5.69 18.93 9.64
WTEC 86.39 1.22 8.71 1.81 169.70 3.10 17.99 3.93 98.24 3.41 20.65 4.29 13.61 3.19 6.34 3.39 3.85 4.04 7.64 3.98 3.00 4.08 0.04 3.58 5.86 3.47
TEC 41.94 0.99 57.58 1.26 68.28 5.44 55.76 8.04 109.19 5.63 50.51 8.24 32.59 5.36 16.77 5.38 12.25 5.65 20.41 5.85 10.48 5.56 3.60 4.67 15.89 5.36
TEM 1.75 6.49 1.67 7.51 3.30 12.48 3.09 16.17 3.49 16.10 3.26 20.83 3.10 19.67 2.86 18.46 4.07 25.01 4.36 27.54 4.02 24.53 2.29 15.57 3.00 19.34
TIC1 33.32 1.38 43.93 1.30 149.55 4.80 1559.42 4.49 152.91 5.18 1462.94 4.87 313.63 4.88 284.12 4.35 112.42 5.85 203.90 6.35 62.83 5.72 14.45 2.08 245.99 4.54
WBIC1 2.96 3.77 2.82 3.96 8.15 9.59 8.17 10.18 7.29 8.72 7.22 9.21 6.36 5.20 5.00 3.79 4.40 2.66 5.80 4.63 4.09 2.06 2.94 -0.50 4.86 3.57
WIM1 2.50 3.44 2.40 3.60 6.72 9.08 6.32 9.75 6.57 9.16 6.16 9.91 7.55 9.47 6.61 8.98 8.60 11.32 9.45 11.68 8.48 11.33 5.32 8.64 6.85 9.24
WTIC1 1.90 2.49 1.83 2.58 2.42 3.28 2.24 3.54 2.42 3.21 2.25 3.45 1.90 3.05 1.34 2.86 1.14 3.15 1.95 3.43 0.87 3.10 -0.59 2.54 1.31 2.90
BEC 49.15 1.00 54.15 1.37 71.01 7.24 56.26 12.97 662.62 6.32 40.88 10.24 43.68 5.79 23.82 3.44 18.48 2.43 29.16 4.07 16.34 1.95 8.73 0.16 22.38 3.25
BEM 2.10 7.74 1.98 9.09 6.32 41.43 5.95 135.73 5.73 53.60 5.36 148.86 11.01 3643.43 7.37 432.31 8.29 185.17 9.77 528.52 8.11 141.92 4.87 47.46 7.48 344.95
BIC1 35.54 1.33 47.12 1.26 53.08 6.41 75.87 6.18 59.40 6.41 87.64 6.17 95.64 6.48 49.48 5.21 49.64 5.06 96.14 5.75 41.16 4.91 9.44 3.79 48.50 5.12
TIC2 20.36 1.49 22.43 1.44 92.09 5.72 145.30 5.47 83.69 6.14 123.69 5.89 307.01 5.63 132.20 5.27 56.46 7.05 100.78 7.20 47.87 6.98 20.61 4.39 105.45 5.48
TIM1 2.13 3.40 2.15 3.35 3.95 10.30 4.00 9.84 4.12 13.44 4.17 12.88 3.56 16.12 3.36 14.06 4.47 21.10 4.73 24.11 4.43 20.21 2.88 7.94 3.49 14.94
TIC3 11.79 1.73 11.42 1.76 123.88 6.10 103.15 6.18 109.00 6.72 92.45 6.80 220.57 6.07 130.22 5.87 67.28 7.27 95.14 7.55 52.14 7.21 12.27 4.57 109.70 6.03
WBIC2 3.17 3.06 3.17 3.06 7.63 7.63 7.62 7.62 6.78 6.78 6.92 6.92 5.75 5.75 4.24 4.24 3.32 3.32 4.89 4.89 2.95 2.95 1.72 1.72 4.05 4.05
WIM2 2.84 2.75 2.84 2.75 9.26 9.26 9.20 9.20 9.38 9.38 9.33 9.33 9.10 9.10 8.48 8.48 10.73 10.73 11.20 11.20 10.73 10.73 7.66 7.66 8.74 8.74
WTIC2 2.12 2.09 2.12 2.09 2.94 2.94 2.95 2.95 2.89 2.89 2.90 2.90 2.56 2.56 2.15 2.15 2.13 2.12 2.64 2.64 1.99 1.99 1.30 1.30 2.14 2.14
BIC2 22.40 1.39 24.83 1.34 40.90 6.24 50.60 5.99 38.36 6.30 46.40 6.06 51.47 6.11 29.09 4.82 24.05 4.55 36.67 5.37 21.42 4.35 11.03 3.02 27.38 4.73
BIM1 2.63 4.34 2.66 4.26 13.03 163.05 12.98 158.05 12.78 247.57 12.79 223.54 15.15 1614.96 12.91 1156.31 14.02 NA 13.47 2297.95 14.81 3947.17 10.49 353.91 12.93 1400.77
BIC3 14.29 1.62 13.77 1.65 42.96 6.70 38.89 6.79 40.62 6.79 37.07 6.90 46.47 6.67 26.87 5.41 22.64 5.19 34.75 5.93 20.15 5.01 9.67 3.83 25.47 5.31

TxDOT 0-5849 210

 The culmination of this data suggests that CULV-5 or RISA-2D with spring analyses
are appropriate to be used to design culverts. However, safe, less conservative load
ratings can be determined using RISA-2D with LEFE. In general, all the models
overestimate the demands and produce safe structures. However, for load rating, a much
less conservative model, even an unconservative model, is desirable, because it will
reduce the conservatism in the controlling critical sections without actually introducing
unconservatism into the load rating.

10. Deflections

Deflection data were collected as a backup measure. Figure 6.28 shows the measured
deflection in the culvert at the midspans. In the top spans the maximum deflection
inward (positive) was around 0.25 millimeter. Some very slight outward deflection
(negative) was also recorded. Deflections in the bottom slabs were essentially negligible.

0.300
0.250
0.200
midspan deflection (mm)

0.150
0.100
0.050
0.000
-0.050
-0.100
-0.150
TES BES TIS BIS
Maximum Deflection 0.259 0.018 0.240 0.034
Minimum Deflection -0.102 -0.063 -0.065 0.000
FIGURE 6.28. MEASURED DEFLECTIONS IN MILLIMETERS FOR THE PLAINVIEW CULVERT: 3.5’ OF FILL

TxDOT 0-5849 211

 11. Summary of Findings for the Plainview Culvert

From this case study the following findings are noted:

1. Predicted sign for live load moment corresponds to actual moment sign indicating
excellent agreement between reality and the models.
2. Predicted magnitude for live load moment typically higher than actual moment
magnitude and therefore conservative.
3. Scatter for predicted dead load moment is higher than for predicted live load
moment (stdev = 0.77/0.36).
4. Critical sections for live load are not the same as critical sections for Operating
Rating.
5. CULV-5 yields most conservative moments and among the worst Operating
Rating.
6. RISA-LEFE (Es per Culvert Rating Guide) yields least conservative moments and
highest Operating Rating.

TxDOT 0-5849 212

 6. THE TULIA CULVERT INSTRUMENTED LOAD TEST

1. Culvert Condition

The Tulia culvert is located in Swisher County approximately 8 miles east of Tulia,
Texas on FM-1318. Figure 6.29 shows the location of the Tulia test culvert. This culvert
was built in 1951. The design is a 1948 transition era culvert designed under the THD
Supplement No. 1. It consists of five 6ft wide by 6ft tall barrels. This culvert is oriented
perpendicular to the two lane FM road. It is in excellent condition and shows evidence of
board form work.

FIGURE 6.29. TULIA LOCATION.

TxDOT 0-5849 213

 2. Summary of Parameters

1. Culvert Test Parameters

Culvert dimensional data from the plan sheets show 5 barrels, 6ft spans, 6ft height,
6in. thick top and bottom slabs, 7in. thick walls and 36ksi steel yield strength. According
to the MCEB, 3ksi concrete was assumed for Level 1 and 2 analyses. For Level 2 and 3
analyses the tested compressive strength of 9.75ksi from the concrete cores was used.
The truck weighed 51kips. The front single axles weighed 12.3kips, while the rear
tandem axles weighed 38.7kips.
2. Soil Test Parameters

A soil unit weight of 120pcf was selected from the Culvert Rating Guide. A modulus
of subgrade reaction, k, of 150pci from Culvert Rating Guide was used for Level 2
analysis. The modulus of elasticity for the soil, Es; was determined for a two layer soil
system using seven different methods. The first modulus value represents the layer from
the ground surface to slightly below the bottom of the culvert. The second modulus value
is used for the native soil beneath the culvert.
1. 8 / 20 ksi from Culvert Rating Guide
2. 6 / 13 ksi from Bowles and McCarthy textbooks
3. 18 / 4 ksi from Unconfined Compressive Strength (UCS) and Standard
Penetration Test (SPT)
4. 18 / 6 ksi from Texas Cone Penetration Test (TCP)
5. 54 / 3 ksi from Dynamic Cone Penetrometer (DCP)
6. 0.3 / 1.9 ksi from Pressuremeter (PMT)
7. 9 / 9 ksi from Falling Weight Deflectometer (FWD)
Testing was performed under one and a half foot of fill including the pavement structure.

TxDOT 0-5849 214

 3. Instrumentation Plan

The instrumentation plan involved placing 4in. electrical resistance strain gages at
every critical section on the inside of the two westernmost barrels. Strain gages were also
placed on the exterior of the top slab and at the top corner exterior walls opposite the
corresponding interior gages. However, during the backfill process, the gages on the
middle spans were lost. Linear displacement gages were placed at the centerline of the
top and bottom slabs. Figure 6.30 shows the location of the strain gages (in purple) and
the linear displacement gages (in blue).

7" 2'-5" 7" 7" 2'-5" 7" 2'-5" 7" 7" 2'-5" 7" 3'-0"
ETEC ETEM ETIC1
6" ITEC ITEM ITIC1 ITIC2 ITIM1 ITIC3 ITIC4 ITIM2
EWTEC
IWTEC IWTIC1 IWTIC2 IWTIC3 IWTIC4

3'-0" LVDT TES LVDT TIS

MATCH LINE
IWEM

3'-0"
LVDT BES LVDT BIS
IWBEC IWBIC1 IWBIC2 IWBIC3 IWBIC4

IBEC IBEM IBIC1 IBIC2 IBIM1 IBIC3 IBIC4 IBIM2

FIGURE 6.30. TULIA INSTRUMENTATION PLAN.

TxDOT 0-5849 215

 4. Load Rating

Figure 6.31 shows the normal load rating for the Tulia culvert under one and a half
feet of fill using all the different models and soil properties. Five methods produced
positive ratings but no method produced a load rating, inventory or operating, above HS-
20. Those models outlined in the Culvert Rating Guide, CULV-5, RISA-2D with
Springs, and RISA-2D with LEFE, shows that the culvert should fail under live load. All
other models show that the culvert should fail under the 1.5 feet of fill.

25
20
Inventory Rating, IR (HS-)

15
10
5
0
-5
-10
-15
Lvl 1; Lvl 2; Lvl 2; Lvl 3; Lvl 3; Lvl 3; Lvl 3; Lvl 3; Lvl 3; Lvl 3;
TEXT TEXT; STRUC; CRG; Bowles; UCS/SPT; TCP; DCP; PMT; FWD; 9
k=150pci k=150pci 8/20 6/13 18/4 18/6 54/3 .3/1.9
Depth of Fill, DOF (ft)

Inventory Rating Operating Rating

FIGURE 6.31. LOAD RATING FOR TULIA CULVERT: 1.5’ OF FILL

Like the Shallowater culvert, the Tulia remains in service and clearly has not failed,
this despite the fact that the load rating process would predict otherwise. Favorable
environmental conditions, conservative structural properties, soil properties, and
constitutive models for both the reinforced concrete and the soil can reasonably explain
why this culvert has “stood the test of time.”

TxDOT 0-5849 216

 5. Live Load Moment – Predicted and Measured

Figure 6.32 shows the live load moment envelope for the Tulia culvert beneath one
and a half feet of fill. The blue and red lines show the measured moment as well as error
bars which show a reasonable range due to variations in concrete modulus, cracked
moment of inertia, and the nature of four inch strain gages which measure average strain
over the gage length and not the actual strain at the critical section. The measured
moment is less than the predicted moment for all models.

8.000 CULV-5 MAX

6.000 CULV-5 MIN

4.000 RISA-2D TEXT MAX

2.000 RISA-2D TEXT MIN

RISA-2D STRUC MAX

0.000
RISA-2D STRUC MIN
-2.000
RISA-2D LEFE CRG MAX
-4.000
RISA-2D LEFE CRG MIN
-6.000
RISA-2D LEFE Bowles
-8.000 MAX
RISA-2D LEFE Bowles
MIN

FIGURE 6.32. LIVE LOAD MOMENTS PREDICTED AND MEASURED FOR TULIA CULVERT: 1.5’ OF FILL.

TxDOT 0-5849 217

 6. Dead Load Moment – Predicted

Figure 6.33 shows the predicted dead load distributions for the Tulia culvert beneath
1.5 feet of fill for all models. The models predict similar directions for dead load effect,
except for some corner locations.

4.000
CULV-5 MAX
3.000
CULV-5 MIN
2.000
RISA-2D TEXT MAX
1.000 RISA-2D TEXT MIN
0.000 RISA-2D STRUC MAX
-1.000 RISA-2D STRUC MIN

-2.000 RISA-2D LEFE CRG

RISA-2D LEFE Bowles
-3.000
RISA-2D LEFE UCS/SPT
-4.000
RISA-2D LEFE TCP
-5.000
RISA-2D LEFE DCP
WEM

WBIC1

BEC
WBEC

TEC
TEM

WTIC1

BEM

WBIC2
WTEC

WIM1

WTIC2
WIM2

BIM1
TIC1

BIC1

BIC2

BIC3
TIC2
TIM1
TIC3

RISA-2D LEFE PMT

FIGURE 6.33. DEAD LOAD PREDICTED FOR TULIA CULVERT: 1.5’ OF FILL.

TxDOT 0-5849 218

 7. Live Load “Goodness”

Figure 6.34 shows the goodness plot for the Tulia culvert beneath one and a half feet
of fill. The plot shows the number of critical sections which fall into each threshold. In
this case, CULV-5 has the most grossly over-conservative predictions, but it also shows
some predictions that are in the wrong direction. However, in the general sense, the least
conservative model is the RISA-2D with LEFE using the Culvert Rating Guide soil
modulus.

100%
90% 11
24 29 12 14
80% 17 19 19
20
39
70% 7
21 12 10 < Mp/Mm
60% 23 12
10 7 5 < Mp/Mm < 10
50% 7 11
15
40% 14 2 < Mp/Mm < 5
28 10
25 10 1 < Mp/Mm < 2
30% 13 11 14
14 7 12 0 < Mp/Mm < 1
20% 8 6
12 9
6 5 7 5 Mp/Mm < 0
10% 8 2
3 11 10 6 6
4 3 3 3 4
0% 0 0 0 0 0 0 0 0 0
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
(σ=14kft) TEXT STRUC LEFE CRG LEFE LEFE LEFE TCP LEFE DCP LEFE LEFE
(σ=14kft) (σ=14kft) (σ=11kft) Bowles USC/SPT (σ=11kft) (σ=24kft) PMT FWD
(σ=11kft) (σ=13kft) (σ=15kft) (σ=11kft)

FIGURE 6.34. LIVE LOAD GOODNESS PLOT FOR TULIA CULVERT: 1.5’ OF FILL
Also included in Figure 6.34 is the average standard deviation of the ratios for each of
the models. The smaller the average standard deviation, the better the model fit. By this
measure, the DCP based model is the least appropriate for accurately predicting the live
load moments. RISA-2D with LEFE using the Culvert Rating Guide, textbook values,
TCP and FWD soil moduli values all model the culvert with the same level of
appropriateness with the least amount of scatter and the most accurate prediction of the
live load moment envelope shape.

TxDOT 0-5849 219

 8. Load Rating “Goodness”

Figure 6.35 shows the number of critical sections in each threshold for each model.
According to this plot, all the field test versions of the RISA-2D with LEFE model
produce one or more critical sections in the structure that should fail under dead load. All
the models show that the culvert should fail under live load, although CULV-5 indicates
that in almost 60 percent of the critical sections, the culvert should have failed under the
truck load.

100% 4 2 2
12 8 5 6 4 4 5
90% 8 3
3 5 3 8
13
80% 16 11 10
17
70% 20
10 < (C-1.3D)/2.17L
60% 6 38 24 22
23
32 23 5 < (C-1.3D)/2.17L < 10
50% 17
19 2 < (C-1.3D)/2.17L < 5
40%
14 1 < (C-1.3D)/2.17L < 2
30% 49 10
28 29 11 0 < (C-1.3D)/2.17L < 1
11 12 11
20% 9
7 7 (C-1.3D)/2.17L < 0
10% 5 3
12 12 4 5 4
0% 3 3 3 3 1
0 0 0 0 0
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
TEXT STRUC LEFE LEFE LEFE LEFE LEFE LEFE LEFE
CRG Bowles UCS/SPT TCP DCP PMT FWD

FIGURE 6.35. LOAD RATING GOODNESS PLOT FOR TULIA CULVERT: 1.5’ OF FILL

9. Critical Section Analysis

For the most part, the critical sections which are problematic in the live load
predictions do not produce problematic load ratings. Table 6.9 highlights the problem
critical sections for the predicted live load demand. Table 6.10 shows the problem
critical sections for the load rating. Careful inspection of Table 6.9 and Table 6.10
reveals that the only overlap is at the top wall exterior corner (WTEC) in CULV-5 and
RISA-2D with springs. Even then this critical section does not control the load ratings.

TxDOT 0-5849 220

 TABLE 6.9. LIVE LOAD DEMAND PROBLEM CRITICAL SECTIONS FOR TULIA CULVERT: 1.5’ OF FILL
RISA-2D LEFE
CULV-5 RISA-2D TEXT RISA-2D STRUC RISA-2D LEFE CRG RISA-2D LEFE Bowles RISA-2D LEFE UCS/SPT RISA-2D LEFE TCP RISA-2D LEFE PMT RISA-2D LEFE FWD
DCP
TL RLL TL RLL TL RLL
max min max min max min max min max min max min max min max min max min max min max min max min max min
WBEC 1.7 4.4 5.0 3.8 2.5 4.6 3.7 4.3 4.1 5.2 5.2 4.9 4.9 2.7 6.6 3.1 14.0 3.6 11.4 3.3 14.9 3.5 14.2 5.3 8.7 3.3
WEM 74.6 5.3 37.1 6.5 43.9 2.7 7.0 3.9 41.8 2.7 4.6 3.9 38.6 1.7 39.1 1.9 23.4 1.6 24.2 1.5 22.7 1.0 49.1 2.8 32.7 1.8
WTEC 0.2 1.1 0.4 1.0 0.7 0.5 0.8 0.5 0.8 0.5 0.9 0.5 0.4 0.5 0.6 0.6 0.9 0.7 0.8 0.7 1.0 0.7 1.3 0.7 0.7 0.6
TEC -0.7 33.7 -0.4 31.8 1.3 10.1 1.6 8.4 1.7 11.0 2.0 9.4 0.9 11.0 1.1 11.8 2.0 16.7 1.7 15.0 2.2 17.3 2.8 16.0 1.4 13.2
TEM 14.9 1.7 15.1 1.6 6.0 0.9 6.2 0.7 6.0 0.9 6.2 0.7 4.9 0.7 5.0 0.7 4.8 0.7 4.7 0.7 4.1 0.7 5.5 0.8 5.0 0.7
TIC1 -1.1 4.3 -1.3 4.3 0.6 1.9 0.5 2.0 0.8 2.2 0.7 2.3 0.2 1.4 0.6 1.5 2.1 1.5 1.5 1.4 2.4 1.4 2.3 2.2 1.0 1.5
WBIC1 7.3 3.0 7.4 2.9 3.2 1.0 3.2 1.0 4.7 1.3 4.8 1.2 4.2 1.7 5.3 2.0 10.5 3.3 8.5 2.8 11.9 3.6 12.5 4.3 6.6 2.4
WIM1 6.0 4.2 6.1 4.1 2.7 1.6 2.8 1.5 2.7 1.6 2.8 1.5 1.8 1.3 1.9 1.3 1.8 1.2 1.8 1.2 1.5 1.1 2.2 1.2 1.9 1.3
WTIC1 2.2 3.2 2.2 3.1 1.7 2.5 1.8 2.4 1.9 2.9 2.0 2.8 1.7 2.4 1.9 2.7 2.3 3.9 2.1 3.4 2.3 4.2 2.9 4.5 2.0 3.0
BEC 7.6 20.1 12.2 18.6 5.0 12.0 8.9 10.6 8.8 14.2 12.7 12.9 11.1 8.1 15.1 9.5 31.3 11.2 25.4 10.3 34.3 11.2 31.9 16.9 19.4 10.1
BEM 24.8 30.1 25.4 26.4 15.1 4.8 15.6 1.2 15.7 5.1 16.3 1.4 5.8 0.5 6.5 1.0 5.7 4.2 5.6 2.9 4.9 5.2 9.1 8.3 6.2 1.7
BIC1 4686.7 16.1 3729.6 16.4 2980.0 8.4 2370.0 8.5 5030.0 9.4 4520.0 9.6 3660.0 7.8 5120.0 8.8 9080.0 11.5 7440.0 10.5 9300.0 11.3 14360.0 12.7 6340.0 9.6
TIC2
TIM1 14.9 8.4 14.9 8.5 6.8 4.7 6.8 4.8 7.0 5.2 6.9 5.3 4.9 3.2 5.1 3.4 5.2 3.1 5.0 3.0 4.6 2.3 5.7 4.5 5.1 3.3
TIC3 32.8 19.9 33.8 19.8 22.4 6.8 22.8 6.7 37.4 7.1 37.7 7.1 5.9 6.1 10.1 6.5 43.0 7.0 31.4 6.6 50.0 6.6 59.2 7.6 19.7 6.6
WBIC2 8.9 9.9 8.9 9.9 4.2 3.5 4.2 3.5 5.7 4.6 5.7 4.6 3.7 3.9 4.6 4.9 8.5 8.7 7.0 7.3 9.6 9.7 9.9 9.6 5.6 5.9
WIM2 13.1 13.5 13.1 13.5 4.5 6.1 4.5 6.1 4.5 6.0 4.4 6.1 3.7 5.0 3.8 5.1 3.6 4.9 3.7 4.9 3.3 4.4 3.8 5.2 3.8 5.1
WTIC2 4.8 1.7 4.8 1.7 4.3 6.0 4.2 6.1 4.7 7.0 4.7 7.0 4.0 5.0 4.5 5.5 5.8 7.8 5.2 6.9 6.1 8.1 6.3 8.9 4.8 6.1
BIC2 16.3 15.5 15.1 15.7 12.9 9.3 12.0 9.5 19.7 10.3 18.7 10.5 12.0 8.4 16.2 9.4 31.7 11.4 26.0 10.6 34.4 11.2 38.0 12.6 20.6 10.1
BIM1 26220.0 10.5 26110.0 10.7 7130.0 0.2 7110.0 0.3 6580.0 0.6 6530.0 0.7 4940.0 0.0 4910.0 0.0 3240.0 60.0 3970.0 30.0 3200.0 150.0 4470.0 0.0 4600.0 0.0
BIC3 33.5 54.6 34.3 54.3 31.1 31.1 13.2 31.0 18.3 35.3 18.7 35.2 14.1 27.8 18.7 31.2 33.4 40.4 28.1 37.0 35.3 40.6 38.6 45.3 23.2 34.0
TIC4 6.7 9.9 6.8 9.9 9.6 10.9 9.8 10.9 1.3 8.5 3.2 8.9 12.5 10.1 9.3 9.5 14.4 9.5 16.3 11.6 6.0 9.2
TIM2 4.7 3.2 4.7 3.2 4.8 3.4 4.8 3.3 4.1 2.7 4.2 2.8 4.3 2.5 4.2 2.6 3.9 1.8 4.8 2.9 4.3 2.8
BIC4 13.7 16.1 13.9 16.0 19.2 18.2 19.4 18.1 10.3 13.8 14.1 15.4 25.9 19.2 21.5 17.9 27.2 19.0 31.4 21.1 17.6 16.7
BIM2 88.4 1.1 88.5 1.5 81.2 1.6 81.6 2.3 61.3 0.0 61.0 0.1 39.3 2.4 48.9 1.3 34.8 4.4 50.8 0.9 57.0 0.5

TABLE 6.10. OPERATING RATING PROBLEM CRITICAL SECTIONS FOR TULIA CULVERT: 1.5’ OF FILL
CULV-5 RISA-2D TEXT RISA-2D STRUC RISA-2D LEFE CRG RISA-2D LEFE Bowles RISA-2D LEFE UCS/SPT RISA-2D LEFE TCP RISA-2D LEFE DCP RISA-2D LEFE PMT RISA-2D LEFE FWD
TL RLL TL RLL TL RLL
max min max min max min max min max min max min max min max min max min max min max min max min max min
WBEC 18.44 0.19 5.24 0.46 10.17 0.42 6.52 0.50 10.12 0.38 7.59 0.50 9.85 0.22 7.62 0.01 4.40 -0.61 5.06 -0.40 4.35 -0.80 4.43 -0.46 6.16 -0.16
WEM 0.59 1.41 2.45 0.93 1.01 2.79 12.97 1.58 1.10 2.85 20.15 1.58 2.09 3.82 2.13 3.41 2.98 4.26 3.00 4.62 2.92 7.33 1.89 2.27 2.45 3.80
WTEC 11.55 0.27 4.05 0.43 2.40 0.94 1.93 1.15 3.18 0.99 2.72 1.15 5.04 1.34 3.55 1.40 1.33 1.57 1.96 1.49 0.95 1.75 0.81 1.77 2.72 1.43
TEC 2.97 0.34 6.75 0.43 5.77 1.62 4.41 2.19 4.60 1.57 3.68 2.19 7.75 2.13 5.56 2.16 2.37 2.02 3.22 2.02 1.85 2.13 1.50 2.28 4.33 2.08
TEM 0.73 2.29 0.71 2.67 1.82 4.58 1.73 5.96 1.89 7.81 1.80 5.96 2.34 9.86 2.27 9.55 2.36 10.32 2.41 10.28 2.76 10.42 2.03 8.66 2.30 9.71
TIC1 4.79 0.56 4.02 0.53 9.38 1.40 12.56 1.32 6.93 1.28 8.44 1.32 28.14 1.76 13.03 1.50 4.30 0.98 5.32 1.24 3.90 0.97 4.06 0.53 7.70 1.37
WBIC1 0.92 0.99 0.87 1.04 2.18 2.69 2.16 2.84 1.57 2.15 1.54 2.84 2.23 1.18 1.99 0.73 1.62 -0.40 1.69 -0.06 1.66 -0.68 1.57 -0.55 1.81 0.35
WIM1 0.84 0.93 0.80 0.98 1.85 2.46 1.73 2.66 1.87 2.55 1.74 2.66 2.70 3.36 2.55 3.33 2.69 3.52 2.73 3.50 3.03 3.87 2.23 3.42 2.59 3.37
WTIC1 0.62 0.60 0.59 0.63 0.73 0.81 0.68 0.88 0.65 0.75 0.60 0.88 0.46 1.18 0.27 1.19 -0.34 1.30 -0.12 1.26 -0.60 1.42 -0.44 1.30 0.11 1.20
BEC 10.10 0.81 5.80 0.99 13.19 1.67 6.78 2.04 7.73 1.47 4.92 2.04 6.95 2.20 5.52 1.67 3.46 0.66 3.90 1.03 3.42 0.40 3.42 0.43 4.60 1.36
BEM 0.62 3.19 0.57 3.81 1.03 20.12 0.95 86.18 1.00 31.19 0.92 86.18 2.41 372.44 2.01 172.60 2.14 42.36 2.18 62.07 2.47 34.20 1.35 21.28 2.02 103.00
BIC1 8.82 0.62 11.42 0.58 10.80 1.64 13.78 1.58 6.47 1.54 7.32 1.58 7.94 2.05 5.41 1.88 2.18 1.71 3.11 1.75 1.90 1.82 1.18 1.64 4.09 1.79
TIC2 8.07 0.58 7.04 0.57 5.55 2.09 5.96 2.03 5.07 2.03 5.41 2.03 29.67 2.26 21.20 2.15 3.57 2.08 5.12 2.13 2.60 2.31 3.32 1.77 9.31 2.08
TIM1 0.76 1.74 0.76 1.71 1.66 3.11 1.67 3.00 1.71 4.80 1.72 3.00 2.45 7.67 2.39 7.18 2.44 7.45 2.48 7.63 2.80 9.23 2.20 5.12 2.40 7.25
TIC3 4.57 0.59 4.40 0.60 5.71 2.05 5.60 2.06 3.49 2.04 3.46 2.06 22.61 2.33 13.59 2.16 3.72 1.70 4.76 1.94 3.34 1.69 2.94 1.38 7.23 2.05
WBIC2 0.69 0.70 0.69 0.70 1.47 2.01 1.47 2.02 1.12 1.56 1.13 2.02 1.81 1.72 1.53 1.32 1.06 0.48 1.15 0.71 1.05 0.30 1.00 0.32 1.31 1.01
WIM2 0.88 1.13 0.88 1.13 2.58 2.50 2.61 2.47 2.65 2.59 2.69 2.47 3.21 3.09 3.14 3.02 3.25 3.16 3.25 3.15 3.57 3.55 3.10 2.98 3.16 3.04
WTIC2 0.71 2.37 0.71 2.37 0.80 0.67 0.80 0.67 0.73 0.60 0.73 0.67 0.82 0.88 0.70 0.83 0.34 0.77 0.48 0.78 0.22 0.83 0.24 0.73 0.60 0.79
BIC2 6.28 0.72 6.88 0.70 6.31 1.61 6.94 1.55 4.25 1.52 4.57 1.55 7.13 1.82 5.47 1.58 3.24 1.08 3.73 1.25 3.24 0.96 2.86 0.89 4.46 1.41
BIM1 0.67 2.15 0.67 2.10 2.56 130.08 2.57 67.38 2.80 62.65 2.83 67.38 3.79 NA 3.66 NA 4.69 253.45 4.09 540.89 4.45 94.98 3.22 NA 3.74 NA
BIC3 3.82 0.68 3.70 0.69 1.59 1.59 7.66 1.60 5.65 1.47 5.52 1.60 6.96 1.97 5.18 1.77 2.53 1.51 3.22 1.57 2.20 1.58 2.04 1.40 4.08 1.65
TIC4 6.06 1.90 5.94 1.90 4.29 1.83 4.22 1.90 30.77 2.34 13.23 2.21 3.40 1.92 4.53 2.06 2.90 2.08 2.72 1.62 7.02 2.15
TIM2 2.01 4.29 2.01 4.32 2.06 7.04 2.05 4.32 2.42 8.52 2.35 8.17 2.40 8.68 2.44 8.54 2.79 11.24 2.15 7.52 2.36 8.14
BIC4 5.39 1.54 5.30 1.54 3.94 1.43 3.89 1.54 7.12 1.95 5.24 1.72 2.93 1.35 3.49 1.46 2.81 1.36 2.43 1.22 4.24 1.58
BIM2 2.59 203.31 2.59 152.48 2.86 142.59 2.84 152.48 3.92 NA 3.80 1852.67 5.08 83.98 4.34 169.87 5.32 42.28 3.65 211.67 3.93 447.17

TxDOT 0-5849 221

 10. Deflections

Deflection data were collected as a backup measure. Figure 6.36 shows the measured
deflection in the culvert at the midspans.

0.250

0.200
midspan deflection (mm)

0.150

0.100

0.050

0.000

-0.050

-0.100

-0.150
TES BES TIS BIS
Maximum Deflection 0.209 0.018 0.006 -0.002
Minimum Deflection -0.025 -0.091 -0.001 -0.101
FIGURE 6.36. MEASURED DEFLECTIONS IN MILLIMETERS FOR THE TULIA CULVERT: 1.5’ OF FILL

11. Summary of Findings for the Tulia Culvert

From this case study the following findings are noted:

1. Predicted sign for live load moment corresponds to actual moment sign indicating
excellent agreement between reality and the models except for the top slab
corners in CULV-5.
2. Predicted magnitude for live load moment typically higher than actual moment
magnitude and therefore conservative.
3. Scatter for predicted dead load moment is slightly higher than for predicted live
load moment (stdev = 0.38/0.31).
4. Critical sections for live load are not the same as critical sections for Operating
Rating except for the exterior wall top corner for CULV-5 and RISA-2D with
springs, and this location is not the controlling location.
5. CULV-5 yields most conservative moments and among the worst Operating
Rating.

RISA-LEFE (Es per Culvert Rating Guide) yields least conservative

6.
moments and highest Operating Rating.
TxDOT 0-5849 222
 7. LIMITED DEAD LOAD INVESTIGATION

The standard deviation for dead load as determined by the various models is much
higher than for live load. In the case studies the live load was investigated at length and
has been found to almost always predict the appropriate bending direction and usually
predict a conservative magnitude for moment. However, because the instrumentation and
load testing was done using in-service culverts, no consistent or truly comparative
measure of the dead load could be made.
1. Dead Load Moment – Predicted and Measured

One data point dealing with dead load exists. Instrumentation was able to measure the
change in dead load when additional fill was added to the Shallowater culvert to increase
the fill thickness from two feet to four feet. The change in moment was predicted using
the various models. These results are compared in Figure 6.37. The dark blue line shows
the measured moment.

2.000
CULV-5 MAX
1.000 CULV-5 MIN
RISA-2D TEXT MAX
0.000
RISA-2D TEXT MIN
RISA-2D STRUC MAX
-1.000
RISA-2D STRUC MIN
RISA-2D LEFE CRG
-2.000
RISA-2D LEFE Bowles

-3.000 RISA-2D LEFE SPT/TCP

RISA-2D LEFE DCP
-4.000 RISA-2D LEFE PMT
WEM

BEC
BEM
WBEC

WBIC1

WBIC2
WTEC
TEC
TEM

WTIC1
WIM1

WTIC2

BIM
TIM

WIM2
BIC1

BIC2

BIC3
TIC1

TIC2

TIC3

RISA-2D LEFE FWD

FIGURE 6.37. THE CHANGE IN DEAD LOAD MOMENT, PREDICTED AND MEASURED, FOR THE SHALLOWATER CULVERT FROM 2' TO
4' OF FILL.

TxDOT 0-5849 223

 2. Dead Load “Goodness”

For all but the exterior wall corner and the center wall, the measure dead load
moments are much smaller than the predicted values. Figure 6.38 shows the dead load
goodness.

100%
90% 16 16 6 8 8 6
20 10 12
80% 4 2 4
8 8 6
70% 6 6
16 6 10
60% 20 14 10 < Mp/Mm
22
8 6 8 10
50% 6 5 < Mp/Mm < 10
4
40% 8 12 2 2 2 2 2 < Mp/Mm < 5
10
12 10 1 < Mp/Mm < 2
30% 16
16 16 10 10 10 10
0 < Mp/Mm < 1
20%
Mp/Mm < 0
10% 16 8 8
12 12 6 6 6 6
0%
CULV-5 RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D RISA-2D
(σ=31kft) TEXT STRUC LEFE CRG LEFE LEFE LEFE DCP LEFE PMT LEFE FWD
(σ=22kft) (σ=25kft) (σ=17kft) Bowles SPT/TCP (σ=28kft) (σ=31kft) (σ=21kft)
(σ=25kft) (σ=29kft)

FIGURE 6.38. CHANGE IN DEAD LOAD GOODNESS FOR THE SHALLOWATER CULVERT FROM 2' TO 4' OF FILL.

Amazingly, the dead load seems to be model with approximately the same degree of
inaccuracy by all models. Across the models, about 15% of the critical sections are not
predicted in the correct direction. For all but the Culvert Rating Guide version of RISA-
2D with LEFE about 25% of the critical sections do not predict adequate magnitude. The
RISA-2D with LEFE using the soil modulus from the Culvert Rating Guide shows that
nearly 50% of the critical sections are unconservatively modeled. However, based on the
standard deviation of the difference between measured and predicted, the Culvert Rating
Guide produces the best moment shape compared to the measure moment envelope.

3. Critical Section Analysis

Table 6.11 shows that the problem critical sections for dead load prediction are actually
very uniform across the models. The wall midspans, the top slab interior corner, and the
centermost bottom midspans all have trouble. Table 6.12 shows the combined problem
critical sections for the operating ratings at both two feet of fill and four feet of fill. A
close examination of these two tables shows some overlap but no overlapped critical
section controls the load ratings.

TxDOT 0-5849 224

 TABLE 6.11. LIVE LOAD DEMAND PROBLEM CRITICAL SECTIONS FOR THE CHANGE IN DEAD LOAD IN THE SHALLOWATER CULVERT FROM 2’ TO 4’ OF FILL
CULV-5 RISA-2D TEXT RISA-2D STRUC RISA-2D LEFE CRG RISA-2D LEFE Bowles RISA-2D LEFE SPT/TCP RISA-2D LEFE DCP RISA-2D LEFE PMT RISA-2D LEFE FWD
TL RLL TL RLL TL RLL
max min max min max min max min max min max min max min max min max min max min max min max min
WBEC 4.9 3.9 4.9 3.9 3.1 2.7 3.1 2.7 3.2 2.8 3.2 2.8 2.9 2.9 3.9 3.9 5.0 5.0 4.6 4.6 5.8 5.8 3.3 3.3
WEM 0.5 1.1 0.5 1.1 0.3 0.9 0.3 0.9 0.4 0.9 0.4 0.9 -0.3 -0.3 0.0 0.0 0.3 0.3 0.2 0.2 0.4 0.4 -0.2 -0.2
WTEC 1.5 1.2 1.5 1.2 1.1 1.0 1.1 1.0 1.1 1.0 1.1 1.0 0.7 0.7 0.9 0.9 0.8 0.8 0.9 0.9 0.6 0.6 0.8 0.8
TEC 5.7 5.1 5.7 5.1 2.8 2.2 2.8 2.2 2.7 2.2 2.7 2.2 1.8 1.8 2.1 2.1 1.7 1.7 2.0 2.0 1.2 1.2 2.0 2.0
TEM 6.7 7.0 6.7 7.0 6.6 6.9 6.6 6.9 6.6 6.9 6.6 6.9 1.9 1.9 4.3 4.3 5.7 5.7 5.3 5.3 6.1 6.1 3.0 3.0
TIC1 -34789.4 -35680.0 -34789.4 -35680.0 -21560.0 -22330.0 -21560.0 -22330.0 -21590.0 -22360.0 -21590.0 -22360.0 -11540.0 -11540.0 -17700.0 -17700.0 -22810.0 -22810.0 -20960.0 -20960.0 -25840.0 -25840.0 -14230.0 -14230.0
WBIC1 -5.6 -6.9 -5.6 -6.9 1.7 2.1 1.7 2.1 1.3 1.5 1.3 1.5 0.4 0.4 3.0 3.0 11.9 11.9 7.5 7.5 21.8 21.8 1.0 1.0
WIM1 -4.5 -5.6 -4.5 -5.6 -2.8 -3.6 -2.8 -3.6 -3.1 -3.9 -3.1 -3.9 -0.6 -0.6 -2.5 -2.5 -4.0 -4.0 -3.6 -3.6 -4.4 -4.4 -1.3 -1.3
WTIC1 0.8 1.0 0.8 1.0 1.4 1.7 1.4 1.7 1.4 1.7 1.4 1.7 0.3 0.3 1.4 1.4 3.4 3.4 2.6 2.6 5.2 5.2 0.7 0.7
BEC 50.0 44.9 50.0 44.9 20.3 15.6 20.3 15.6 21.5 16.9 21.5 16.9 19.8 19.8 26.8 26.8 34.7 34.7 31.5 31.5 40.5 40.5 22.8 22.8
BEM 41.2 43.2 41.2 43.2 29.0 29.9 29.0 29.9 31.8 32.9 31.8 32.9 19.6 19.6 32.3 32.3 40.0 40.0 37.7 37.7 42.5 42.5 25.3 25.3
BIC1 37.5 38.5 37.5 38.5 18.4 18.7 18.4 18.7 19.3 19.7 19.3 19.7 15.4 15.4 18.8 18.8 18.4 18.4 18.9 18.9 16.8 16.8 17.3 17.3
TIC2 8.1 8.3 8.1 8.3 4.9 5.0 4.9 5.0 4.9 5.0 4.9 5.0 2.8 2.8 3.9 3.9 4.5 4.5 4.3 4.3 4.6 4.6 3.3 3.3
TIM1 3.6 3.6 3.6 3.6 3.6 3.6 3.6 3.6 3.6 3.6 3.6 3.6 1.3 1.3 2.6 2.6 3.1 3.1 3.0 3.0 3.1 3.1 2.0 2.0
TIC3 7.0 6.9 7.0 6.9 4.2 4.2 4.2 4.2 4.2 4.2 4.2 4.2 2.6 2.6 3.6 3.6 4.1 4.1 3.9 3.9 4.4 4.4 3.1 3.1
WBIC2 0.0 0.0 0.0 0.0 0.1 0.0 0.1 0.0 0.1 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
WIM2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
WTIC2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0
BIC2 16.5 16.7 16.5 16.7 8.7 8.9 8.7 8.9 9.1 9.3 9.1 9.3 7.2 7.2 9.1 9.1 10.0 10.0 9.7 9.7 10.4 10.4 8.1 8.1
BIM1 -111.6 -110.6 -111.6 -110.6 -92.4 -92.7 -92.4 -92.7 -98.9 -99.1 -98.9 -99.1 -66.8 -66.8 -97.5 -97.5 -113.4 -113.4 -108.3 -108.3 -121.3 -121.3 -81.9 -81.9
BIC3 19.8 19.7 19.8 19.7 10.4 10.3 10.4 10.3 10.7 10.6 10.7 10.6 8.5 8.5 10.2 10.2 10.7 10.7 10.6 10.6 10.6 10.6 9.4 9.4

TxDOT 0-5849 225

 TABLE 6.12. OPERATING RATING PROBLEM CRITICAL SECTIONS FOR THE CHANGE IN DEAD LOAD IN THE SHALLOWATER CULVERT FROM 2’ TO 4’ OF FILL
Operating Ratings Under 2' of Fill
CULV-5 RISA-2D TEXT RISA-2D STRUC RISA-2D LEFE CRG RISA-2D LEFE Bowles RISA-2D LEFE SPT/TCP RISA-2D LEFE DCP RISA-2D LEFE PMT RISA-2D LEFE FWD
TL RLL TL RLL TL RLL
max min max min max min max min max min max min max min max min max min max min max min max min
WBEC 26.12 0.05 5.41 0.47 14.94 0.65 7.21 0.88 14.35 0.65 7.22 0.88 15.65 0.75 7.36 -0.39 4.82 -0.96 5.46 -0.79 4.19 -1.18 11.12 0.18
WEM 0.67 1.88 3.19 1.17 0.91 3.16 9.84 1.78 0.97 3.16 10.75 1.78 3.90 13.17 3.50 4.64 3.74 3.15 3.67 3.50 3.80 2.83 3.51 7.49
WTEC 11.22 0.26 3.54 0.51 7.62 0.68 4.32 0.86 6.44 0.68 3.83 0.86 5.35 1.27 1.20 1.51 -0.38 1.88 0.06 1.75 -0.87 2.06 3.13 1.33
TEC 41.47 0.07 10.17 0.18 13.34 0.71 5.73 1.02 10.88 0.71 5.14 1.02 8.61 1.35 2.45 1.69 0.25 2.05 0.84 1.93 -0.38 2.20 5.27 1.45
TEM 0.20 3.21 0.17 3.79 0.33 5.22 0.28 6.61 0.34 5.22 0.29 6.61 0.64 10.37 0.43 9.14 0.36 8.65 0.38 8.79 0.34 8.50 0.53 9.66
TIC1 13.50 0.24 20.26 0.21 9.86 0.92 13.91 0.83 9.45 0.92 12.96 0.83 52.93 1.36 16.95 0.70 7.11 0.06 9.61 0.29 5.34 -0.28 33.09 1.06
WBIC1 0.94 1.16 0.87 1.25 1.65 3.59 1.66 3.47 1.64 3.59 1.64 3.47 2.91 2.10 2.48 0.78 2.19 -0.60 2.27 -0.13 2.12 -1.17 2.71 1.59
WIM1 0.83 1.04 0.78 1.11 1.73 2.27 1.62 2.43 1.68 2.27 1.57 2.43 3.14 3.54 2.29 3.01 2.00 2.89 2.06 2.91 1.95 2.88 2.69 3.23
WTIC1 0.60 0.70 0.57 0.74 0.78 0.82 0.70 0.90 0.79 0.82 0.71 0.90 0.74 1.27 0.18 1.32 -0.45 1.50 -0.21 1.44 -0.81 1.60 0.52 1.25
BEC 15.05 -0.12 7.92 0.03 44.77 0.70 11.50 1.25 38.11 0.70 11.34 1.25 20.95 0.72 9.03 -0.48 5.65 -1.06 6.49 -0.88 4.82 -1.29 14.17 0.10
BEM 0.41 4.11 0.35 4.94 1.08 21.49 1.00 53.56 0.93 21.49 0.85 53.56 4.23 524.94 1.33 122.09 0.65 47.47 0.79 64.01 0.55 32.15 2.52 308.44
BIC1 13.86 0.08 18.94 0.04 22.89 1.31 38.88 1.25 23.95 1.31 41.30 1.25 39.88 1.92 11.54 1.38 4.11 1.34 5.94 1.33 2.29 1.38 23.41 1.60
TIC2 7.05 0.24 7.82 0.23 8.99 0.97 10.36 0.93 8.12 0.97 9.19 0.93 56.46 1.60 22.74 1.27 6.86 1.19 10.91 1.22 4.27 1.18 37.86 1.43
TIM1 0.24 1.75 0.25 1.72 0.38 4.59 0.39 4.43 0.39 4.59 0.40 4.43 0.70 9.01 0.56 7.82 0.56 6.47 0.56 6.94 0.59 5.65 0.61 8.51
TIC3 3.61 0.29 3.49 0.30 4.77 1.10 4.58 1.12 4.52 1.10 4.35 1.12 44.49 1.66 13.61 1.33 5.61 0.98 7.74 1.13 4.16 0.72 23.58 1.50
WBIC2 0.94 0.90 0.94 0.90 2.15 2.15 2.18 2.18 2.13 2.15 2.20 2.18 2.42 2.42 1.62 1.62 1.04 1.04 1.22 1.22 0.85 0.85 2.08 2.08
WIM2 0.90 0.83 0.90 0.83 2.29 2.29 2.28 2.28 2.31 2.29 2.29 2.28 3.41 3.41 2.80 2.80 2.63 2.63 2.67 2.67 2.58 2.58 3.07 3.07
WTIC2 0.63 0.59 0.63 0.59 0.78 0.78 0.79 0.79 0.78 0.78 0.79 0.79 1.01 1.01 0.76 0.76 0.63 0.63 0.67 0.67 0.56 0.56 0.89 0.89
BIC2 8.33 0.09 9.15 0.08 10.42 1.31 12.00 1.24 10.40 1.31 11.94 1.24 20.60 1.73 8.76 0.97 5.25 0.58 6.11 0.70 4.41 0.39 13.91 1.34
BIM1 0.50 2.50 0.51 2.45 2.75 9.22 2.72 8.97 2.54 9.22 2.52 8.97 4.95 NA 2.80 NA 2.17 NA 2.36 NA 1.63 587.61 3.71 NA
BIC3 5.38 0.13 5.19 0.14 11.63 1.31 10.95 1.33 11.73 1.31 11.04 1.33 20.95 1.88 8.28 1.24 4.40 1.05 5.39 1.09 3.36 1.03 13.92 1.53
Operating Ratings Under 4' of Fill
CULV-5 RISA-2D TEXT RISA-2D STRUC RISA-2D LEFE CRG RISA-2D LEFE Bowles RISA-2D LEFE SPT/TCP RISA-2D LEFE DCP RISA-2D LEFE PMT RISA-2D LEFE FWD
TL RLL TL RLL TL RLL
max min max min max min max min max min max min max min max min max min max min max min max min
WBEC -7.00 -0.61 13.77 0.03 384.38 0.76 18.55 1.69 174.02 0.76 18.36 1.69 32.87 0.12 13.58 -1.43 8.23 -2.17 9.53 -1.95 6.95 -2.45 21.83 -0.68
WEM 1.15 3.87 7.46 1.74 1.28 10.61 25.31 3.17 1.37 10.61 26.87 3.17 5.74 24.29 5.96 7.30 6.88 4.50 6.62 5.14 7.14 3.91 5.49 12.88
WTEC -1.99 -0.10 10.34 0.41 5.50 0.65 174.45 1.09 6.31 0.65 165.41 1.09 15.72 1.95 3.85 1.99 0.13 2.44 1.11 2.25 -0.93 2.68 9.13 1.85
TEC 89.97 -0.42 15.04 -0.22 3.40 0.73 213.50 1.58 3.85 0.73 212.11 1.58 20.44 2.18 5.35 2.28 0.92 2.74 2.07 2.57 -0.29 2.95 11.88 2.12
TEM 0.00 7.20 -0.08 9.23 0.03 12.58 -0.09 19.02 0.05 12.58 -0.08 19.02 1.03 40.92 0.31 22.91 0.10 18.21 0.15 19.43 0.04 16.98 0.61 30.20
TIC1 21.90 -0.25 34.65 -0.31 45.03 0.86 145.33 0.65 46.55 0.86 157.56 0.65 128.50 1.78 105.08 0.28 17.42 -0.80 29.46 -0.44 11.20 -1.31 151.01 1.08
WBIC1 1.60 2.18 1.44 2.41 4.16 11.25 4.18 10.11 4.14 11.25 4.08 10.11 5.49 4.05 3.76 1.13 3.22 -1.04 3.34 -0.33 3.12 -1.92 4.55 2.72
WIM1 1.39 1.88 1.26 2.07 3.37 4.60 3.01 5.20 3.24 4.60 2.89 5.20 6.78 7.74 3.88 5.45 3.03 4.90 3.23 5.02 2.87 4.82 5.18 6.42
WTIC1 1.09 1.42 1.00 1.54 1.71 1.81 1.41 2.09 1.71 1.81 1.41 2.09 1.50 2.65 0.18 2.26 -0.86 2.38 -0.49 2.31 -1.42 2.49 0.89 2.38
BEC 100.85 -0.83 18.44 -0.62 5.77 0.73 26.24 3.17 6.47 0.73 25.24 3.17 39.37 0.05 15.38 -1.55 9.12 -2.29 10.65 -2.06 7.61 -2.57 25.65 -0.78
BEM 0.20 8.82 0.08 11.59 1.30 29.73 1.11 109.59 1.02 29.73 0.83 109.59 4.33 770.64 0.44 183.35 -0.47 80.18 -0.27 103.64 -0.60 54.82 2.04 436.46
BIC1 29.14 -0.56 48.78 -0.62 41.61 1.22 102.98 1.10 43.29 1.22 108.82 1.10 99.98 2.03 25.15 1.30 8.43 1.38 12.26 1.31 4.76 1.53 58.13 1.57
TIC2 14.13 -0.19 16.21 -0.23 66.28 1.03 174.34 0.93 64.09 1.03 155.38 0.93 89.80 2.25 38.68 1.29 19.46 1.05 24.65 1.09 11.71 1.04 63.60 1.71
TIM1 0.16 4.22 0.18 4.11 0.30 15.59 0.33 14.58 0.33 15.59 0.36 14.58 1.14 38.48 0.57 19.07 0.50 12.88 0.50 14.61 0.53 10.72 0.79 27.32
TIC3 9.63 -0.10 9.13 -0.08 52.77 1.25 43.89 1.29 46.84 1.25 39.52 1.29 82.51 2.38 49.21 1.38 14.37 0.70 25.52 0.96 8.98 0.27 64.20 1.85
WBIC2 1.81 1.73 1.81 1.73 4.56 4.55 4.62 4.61 4.72 4.55 4.76 4.61 4.55 4.55 2.43 2.43 1.46 1.46 1.73 1.73 1.17 1.17 3.47 3.47
WIM2 1.56 1.51 1.56 1.51 4.74 4.74 4.69 4.69 4.73 4.74 4.69 4.69 7.41 7.41 4.92 4.92 4.25 4.25 4.41 4.40 4.13 4.12 5.98 5.98
WTIC2 1.21 1.20 1.21 1.20 1.76 1.76 1.77 1.77 1.73 1.76 1.74 1.77 2.04 2.04 1.25 1.25 0.92 0.92 1.01 1.01 0.80 0.80 1.62 1.62
BIC2 19.45 -0.49 22.76 -0.52 24.52 1.20 31.77 1.07 24.40 1.20 31.33 1.07 47.73 1.75 17.55 0.67 9.70 0.17 11.58 0.33 7.83 -0.07 30.16 1.18
BIM1 0.43 5.40 0.45 5.25 2.83 25.27 2.78 23.89 2.47 25.27 2.45 23.89 5.36 NA 2.32 NA 1.02 545.22 1.49 NA -0.03 -4.93 3.62 NA
BIC3 13.40 -0.39 12.63 -0.37 31.15 1.32 27.57 1.36 31.32 1.32 27.67 1.36 47.44 2.05 16.67 1.17 8.28 0.98 10.37 1.01 6.18 0.99 30.01 1.55

TxDOT 0-5849 226

 4. Deflections

Figure 6.39 shows the deflections. Because the moment measurement at the critical
sections was so successful, no further analysis was made of the dead load deflections. It
is worth noting that the dead load deflections are nearly an order of magnitude smaller
than the live load deflections.

0.120
0.100
0.080
midspan deflection (mm)

0.060
0.040
0.020
0.000
-0.020
-0.040
-0.060
-0.080
-0.100
TES BES TIS BIS
Deflection 0.100 0.052 0.060 -0.081
FIGURE 6.39. MEASURED DEFLECTIONS IN MILLIMETERS FOR THE SHALLOWATER CULVERT FROM 2' TO 4' OF FILL

TxDOT 0-5849 227

 5. Summary of Findings for the Limited Dead Load Investigation

The summary of the limited dead load investigation are as follows:

1. For the one case, sign for predicted live load moment does NOT correspond to
sign for actual moment for 10-20% of the critical sections
2. For the one case, predicted magnitude for dead load moment is higher than actual
moment magnitude for ~60% of the critical sections
3. Some overlap occurs between Operating Rating problem critical sections and
Dead Load ratio problem critical sections, though none of these sections are the
controlling critical sections.
1. CULV-5 has problems in the interior walls, TIC1 and BIM.
2. RISA-2D with Springs has problems at WEM, TIC1, and WTIC2
3. RISA-2D with LEFE is dependent on the modulus of elasticity:
1. For the Culvert Rating Guide value, only the WTIC1 has a
problem. The Culvert Rating Guide is the best model for dead
load.
2. For the textbook value, TIC1 and WTIC2 have problems.
3. For the SPT, TCP and DCP values, WTEC, TIC1, and WTIC2
have problems.
4. For the pressuremeter values, WTEC, TIC1, WBIC2, WTIC2 and
BIM have problems.
5. For the FWD values, the top wall interior corners have problems.
4. Predicted dead load moment is more variable than predicted live load moment and
increases with fill depth.

TxDOT 0-5849 228

 8. SUMMARY AND CONCLUSIONS

The field instrumentation portion of this research study was successful in meeting the
stated objectives.
The exercise of collecting site specific data was quite illuminating. The limited
findings from this study suggest that TxDOT does not use high quality backfill around
culverts. Rather, native materials are used to bury reinforced box culverts. It was also
illuminating to see the quality of the concrete used in construction. The concrete was
stronger, in some cases much stronger, than the values suggested by AASHTO. This
suggests that the culvert rating process contains significant conservatism in constructed
materials, and is often appropriately conservative in relation to soils.
Measuring the actual strains under known loads confirmed that the models do
appropriately predict culvert behavior. Under live loads, the shape of the moment
envelope is modeled correctly. The magnitudes of the moments are also conservative,
particularly for lower level models.
The process of determining site specific soil testing only highlighted the order of
magnitude difficulty associated with obtaining soil parameters for analytical modeling
purposes. Though a great many methods are available for determining soil stiffness,
many labor-intensive methods produce very conservative results, while the less labor-
intensive methods produce results that still produce conservative live load predictions.
Soil remains a complex and difficult material to quantify.
The instrumented load tests, by design, did not address several key factors concerning
culvert load rating. Most notably the effect of dead load on culvert behavior was not
explored. Though some data were available, the ways in which soil acts to support itself
and supply load to a buried structure is still not well understood.
The most important conclusion from this portion of the study is that all the Culvert
Rating Guide analytical methods produce conservative load ratings. Even though the
higher order models with less conservative soil values may produce slightly
unconservative live load moment predictions in some critical sections, the load rating is
always controlled by a conservative, and often over-conservative, critical section demand
prediction. Therefore it is safe to say that any presented method is safe for culvert load
rating.
However, it is important to note that this is not the case for culvert design. Because
demands for some critical sections are under predicted by the higher order models, the
higher order models should not be used to design reinforced concrete box culverts.
Rather, the faithful, lower-order models, e.g., CULV-5, should continue to be used to
design safe, serviceable culverts.

TxDOT 0-5849 229

 7. CONCLUSIONS AND RESEARCH LIMITATIONS

1. CONCLUSIONS

1. Evaluation of TxDOT’s Culvert Load Rating Practices and Procedures

At the beginning of this research study, the assumption was that TxDOT had a
repeatable load rating procedure created around CULV-5. However, upon investigation
this was not found to be the case. Rather, TxDOT had a definite need for such a
procedure. Therefore this project initially focused on articulating a clear and repeatable,
production-oriented load rating procedure that would yield reliable load ratings. More
sophisticated analyses could conceivably reduce excess overconservatism in the load
ratings by considering the effects of soil-structure interaction. To this end the Culvert
Rating Guide was developed as the project deliverable.

2. Development of the Culvert Rating Guide

The Culvert Rating Guide is the main deliverable for this project. The guide
articulates a clear and repeatable load rating procedure designed to satisfy current
AASHTO specifications and provide for four levels of increasing demand modeling
sophistication. These four levels are: Level 1, culvert specific frame analysis programs
typified by CULV-5; Level 2, two-dimensional general frame analysis programs as
typified by RISA-2D with spring subgrade support; Level 3, two-dimensional finite
element soil-structure interaction programs as typified by RISA-2D with linear elastic
finite elements; and Level 4, higher order generalized programs including non-linear two-
dimensional models and three dimensional models. The Culvert Rating Guide provides
specific direction for load rating using the first three methods.

3. Evaluation of TxDOT Culvert Designs and Analysis Methods

Validation of the Culvert Rating Guide was accomplished with a breadth and depth
approach. In the initial validation task, one hundred TxDOT culvert designs
representative of the full population of TxDOT’s culvert inventory were load rated using
the first three analysis levels. The results showed that in general, the Level 2 analysis
produces marginally higher load ratings than the Level 1 analysis. It also showed that the
Level 3 method can produce much higher load ratings if the soil is sufficiently stiff.
However, if culvert backfill is of poor quality, the higher-level load rating may be less
than that determined by CULV-5.
This work also revealed that the presenting problem upon which TxDOT
commissioned this research study may in fact be real. That is, for cases where in-service
culverts must be lengthened or reconfigured, unless the culvert backfill soil is sufficiently
stiff, the culvert may require load posting or replacement. Generally, the newer the
culvert is, the more likely that the culvert will load rate acceptably.

TxDOT 0-5849 230

 4. Parametric Analysis

The findings of the parametric analysis were incorporated into the Culvert Rating
Guide. The recommended values for the modulus of subgrade reaction were found to be
acceptable. The Level 3 analysis was found to be relatively insensitive to Poisson’s ratio
and a typical value of 0.3 is appropriate for all but deep fill culverts beneath clay soils.
The parametric study also showed that CULV-5 can be used conservatively to load rate
culverts with five or more four barrels by modeling the culvert with only four barrels.
The load rating is not very sensitive to the lateral earth pressure, therefore AASHTO’s
equivalent fluid weight values are recommended.
Culvert load ratings were found to be highly sensitive to the modulus of elasticity for
the soil in the Level 3 analysis. The depth of fill is also a highly sensitive parameter;
therefore, culverts should be load rated at their actual depth of fill and culvert designs
should be evaluated at both their maximum and minimum depths.

5. Instrumented Load Tests on Three In-Service Culverts

Field instrumentation and load tests were limited to three in-service TxDOT culverts,
the key objective being a comparison of measured versus predicted live load moment
demands. This work primarily evaluated the reliability of analytical modeling approaches
recommended in the Culvert Rating Guide to predict live load demands.
The instrumented load test data indicated that the culvert load ratings for each model
were conservative. The higher level models yielded slightly unconservative results at
some critical sections. However, these are not the controlling critical sections for the
load rating. Therefore, the most important finding from the field study is that all models
may be conservatively used for load rating.
Relative to culvert design, however, only the lowest order model, i.e. CULV-5,
should be used.
The very limited dead load evaluation indicated that the distribution of moment
demands due to dead load is not well understood. An appropriate way to further explore
this would be to instrument a newly constructed culvert.
Site-specific soil testing performed as part of this study highlighted the order of
magnitude difficulty associated with obtaining soil elastic modulus values for Level 3
analytical modeling purposes. Several methods are available to determine soil elastic
modulus, but values determined by these methods vary widely within a given soil
stratum. Soil elastic modulus remains a complex and difficult material to quantify.

TxDOT 0-5849 231

 2. LIMITATIONS OF RESEARCH

1. Culvert Type

This research only considered cast-in-place reinforced concrete box culverts. Though
the principles outlined in the Culvert Rating Guide could be used to load rate other types
of culverts, the analytical demand models and capacity calculations described in the
Guide are for cast-in-place reinforced concrete box culverts.

2. Culvert Drainage Condition

Another significant limitation is the backfill drainage condition. Throughout this

project, the culvert backfill has been assumed to be fully drained. This is often the case
in the dryer portions of the state of Texas where culverts primarily serve to control flash
flood conditions. However, in the wetter portions of the State, in particular toward the
east and along the Gulf Coast, culverts often bridge over creeks and drainage ditches
which are continually filled with water. The Culvert Rating Guide has not considered the
effects of pore water pressure, buoyancy or saturation of backfill on the load rating.

3. Soil Elastic Modulus

The parametric study indicated that culvert load rating is highly sensitive to the soil
modulus value for Level 3 analyses. Field testing explored several methods for
determining soil elastic modulus, however, the soil elastic modulus vary widely within a
given stratum. The modulus values provided in the Culvert Rating Guide can be used
with significant engineering judgment. More work is necessary on this aspect of culvert
load rating.
4. Depth of Fill

The field work was also limited relative to the depth of fill that was evaluated. All of
the evaluated culverts had relatively low fill depths. Many in-service culverts in Texas
are low fill culverts; however the Culvert Rating Guide should apply with equal
confidence to all types of culverts in the TxDOT inventory. This means that more field
work should be done to evaluate the Guide’s modeling capabilities relative to deep fill
culverts.

5. Live Load Demand Measurements

The field instrumentation portion of this study was limited to three in-service
culverts. Though the findings from the instrumented load tests were reasonable, this is a
limited validation of the live load demand predictions using the various models.
Additional study would be appropriate.

TxDOT 0-5849 232

 6. Dead Load Demand Measurements

The field instrumentation portion of this study considered dead load for in a limited
way for only one culvert project. This very limited effort did not provide conclusive
validation of the dead load demand predictions. Additional work is necessary to increase
confidence in the dead load predictions, particularly for the higher order models.

3. RECOMMENDATIONS

The Culvert Rating Guide developed as part of this research study represents a
significant improvement in TxDOT’s culvert load rating capabilities. It provides clear
guidance for repeatable load rating, including the ability to reduce excess
overconservatism in load ratings by taking into account soil-structure interaction effects.
The practices and procedures in the Culvert Rating Guide could be automated.
Implementation might consist of pre-programmed worksheets that facilitate capacity
calculations and model generation. A more sophisticated approach would input culvert
details as outlined in Chapter 4 of the Guide, automatically calculate culvert capacity and
demands per Chapter 5 and Chapter 6 of the Guide, and output the culvert load rating,
critical section and failure mode.
Implementation of the Culvert Rating Guide could be further enhanced through
development and dissemination of training materials, workshops, tutorials, other
educational aids.
Notwithstanding the advance in culvert load rating procedure and practice embodied
in the Culvert Rating Guide, the limitations identified above suggest the need to further
explore several aspects of the culvert load rating problem. These include:

• Culvert Drainage Condition

• Soil Elastic Modulus

• Depth of Fill

• Live Load Demand Measurements

• Dead Load Demand Measurements

TxDOT culvert load rating analyses will continue to indicate that certain culverts
should be load-posted or retrofitted when, in fact, many of these structures are probably
serviceable. One type of response in such cases is to quote Section 7.4.1 of the MCEB:
“A concrete [culvert] need not be posted for restricted loading when it has been carrying
normal traffic for an appreciable amount of time and shows no distress” (AASHTO,
2003). A better response is further study of the load rating concepts identified above in
order to illuminate the issues. Ideally such work will enable further refinement in

TxDOT 0-5849 233

TxDOT’s culvert load rating practices and procedures, the goal being to provide safe,
effective and efficient movement of people and goods in Texas’ roadway system.

TxDOT 0-5849 234

 8. BIBLIOGRAPHY
AASHO. (1949). Standard Specifications for Highway Bridges 5th Ed. Washington D.C.:
American Association of State Highway Officials.
AASHTO. (2007). LRFD Bridge Design Specifications Customary Units 4th Ed.
Washington D.C.: American Asssociation of State Highway and Transportation
Officials.
AASHTO. (2003). Manual for Condition Evaluation of Bridges 2nd Ed. Washington,
D.C.: American Association of State Highway and Transportation Officials.
AASHTO. (1977). Standard Specifications for Highway Bridges 12th Ed. Washington
D.C.: American Association of State Highway and Transportation Officials.
AASHTO. (1983). Standard Specifications for Highway Bridges 13th Ed. Washington
D.C.: American Association of State Highway and Transporation Officials.
AASHTO. (1989). Standard Specifications for Highway Bridges 14th Ed. Washington
D.C.: American Association of State Highway and Transportation Officials.
AASHTO. (1992). Standard Specifications for Highway Bridges 15th Ed. Washington
D.C.: American Association of State Highway and Transportation Officials.
AASHTO. (2002). Standard Specifications for Highway Bridges 17th Ed. Washington,
D.C.: American Association of State Highway and Transportation Officials.
Abdel-Karim, A. M., Tadros, M. K., & Benak, A. J. (1993). Structural Response of Full-
Scale Concrete Box Culvert. Struc. Eng. , 119 (11), 3238-3254.
Abdel-Karim, A. M., Tadros, M. K., & Benak, J. V. (1990). Live Load Distribution on
Concrete Box Culverts. Washington DC: TRB.
Abu-Farsakh, M. Y., Almohd, I., & Farrag, K. (2007). Comparison of Field and
Laboratory Pullout Test on Geosynthetics in Marginal Soils. Transportation
Research Board No. 1975 , 124-136.
Abu-Hejleh, N. M., Hanneman, D., Wang, T., & Ksouri, I. (2008). Evaluation and
Recommendations for Flowfill and Mechanically Stabilized Earth Bridge
Approaches. Transporation Research Record No. 2045 , 51-61.
Allen, T. (2001). Development of the Simplified Method for Internal Stability Design of
Mechanically Stabilized Earth Walls. Olympia, Washington: Washing State
Department of Transportation.
Awwad, E. e. (2000). Finite element analysis of concrete box culverts. Computing in
Civil and Building Engineering (pp. 1051-1053). Proc. of the Eight International
Conference ASCE.
TxDOT 0-5849 235
Bennett, R. M. (2005). Vertical Loads on Concrete Box Culverts under High
Embankments. Bridge Eng. , 10 (6), 643-649.
Bowles, J. E. (1996). Foundation Analysis and Design. New York: McGraw Hill.
Bowles, J. E. (1996). Foundation Analysis and Design (5th ed.). New York: McGraw-
Hill.
Briaud, J., Meriwether, M., & Porwoll, H. (1983). Pressuremeter Design of Retaining
Walls. Arlington, TX: Texas Transportation Institute.
Briaud, J.-L. (2005). The Pressuremeter. London: Taylor & Francis Group.
Carrasco, B. (2008, October 7). Culvert Histogram by year built. (T. A. Wood,
Interviewer)
Chang, D. T., Sun, T. S., & Hung, F. Y. (1995). Pullout Mechanism of Geogrid under
Confinement by Sandy and Clayey Soils. Transportation Research Record No.
1474 , 64-72.
Chang, J. C., Hannon, J. B., & Forsyth, R. A. (1977). Pull Resistance and Interaction of
Earthwork Reinforcement and Soil. Transportation Research Board 56th Annual
Meeting.
Coduto, D. P. (2001). Foundation Design Principles and Practices (3rd ed.). Upper
Saddle River, New Jersey: Prentice-Hall.
Dasgupta, A., & Sengupta, B. (1991). Large-Scale Model Test on Square Box Culvert
Backfilled with Sand. Geotechnical Eng. , 117 (1), 156-161.
DGSI. (2005). S-200 & S-205 Dynamin Cone Penetrometer Use and Operating
Guidelines. Stone Mountain, GA: Durham Geo Slope Indicator.
Duane, J., Robinson, R., & Moore, C. A. (1986). Culvert-Soil Interaction Finite Element
Analysis. Transportation Eng. , 112 (3), 250-263.
Elton, D. J., & Patawaran, M. A. (2007). Mechanically Stabilized Earth Reinforcement
Tensile Strength for Tests of Geotectile-Reinforced Soil. Transporation Research
Record No. 1868 , 81-88.
FHWA. (1986). Culvert Inspection Manual. Washington D.C.: Federal Highway
Administration.
Frederick, G. R. (1988). Investigation of the Structural Adequacy of C 850 Box Culverts.
Washington DC: TRB.
Gardner, M. P., & Jeyapalan, J. K. (1982). Preliminary Analysis of the Behavior of
Reinforced Concrete Box Culverts. Arlington, Tx: Texas Transportation Institute
(TTI).
TxDOT 0-5849 236
Garg, A. K., Abolmaali, A., & Fernandez, R. (2007). Experimental Investigation of Shear
Capacity of Precast Reinforced Concrete Box Culverts. Bridge Eng. , 12, 511.
Gladstone, R. A., Anderson, P. L., Fishman, K. L., & Withiam, J. L. (2007). Durability of
Galvanized Soil Reinforcement: More than 30 Years of Experience with
Mechanically Stabilized Earth. Transporation Research Record No. 1975 , 49-59.
Jayawickrama, P. W., & Lawson, W. D. (2008). GEO102: Geotechnical Applications for
Transportation Projects. Austin, Tx.
Katona, M. G. (2008, May 21). (T. A. Wood, & W. D. Lawson, Interviewers)
Katona, M. G. (1976). CANDE-a Modern Approach for the Structural Design and
Analysis of Buried Culverts. Port Hueneme, Ca.: Naval Construction Battalion
Center.
Katona, M. G. (2007, July). Live Load Distributions for use in CANDE. (T. A. Wood, &
W. D. Lawson, Interviewers)
Katona, M. G. (1979). Structural evaluation of new concepts for long-span culverts and
culvert installations : interim report. Springfield, Va.: Federal Highway
Administration.
Katona, M. G., & McGrath, T. J. (2007). A Guideline for Interpreting AASHTO LRFD
Specifications to Design or Evaluate Buried Structures with Comprehensive
Solution Methods. Transportation Research Board 86th Annual Meeting.
Washington DC: TRB.
Katona, M. G., & Smith, J. M. (1976). CANDE User Manual. Port Hueneme, Ca.: Naval
Construction Battalion Center.
Katona, M. G., & Vittes, P. D. (1982). Soil-Structure Analysis and Evaluation of Buried
Box-Culvert Designs. Washington DC: TRB.
Katona, M. G., Mlynarski, M., & McGrath, T. J. (2008). CANDE-2007 Culvert Analysis
and Design Solution Methods and Formulations. National Cooperative Highway
Research Project NCHRP 15-28.
Kilgore, J. (2007, January 4). Evaluating Existing Culverts foir Load Capacity Allowing
for Soil Structure Interaction. RTI Research Project Statement . Austin, Texas:
TxDOT.
Kim, K., & Yoo, C. H. (2005). Design Loading on Deeply Buried Box Culverts.
Geotechnical & Geoenvironmental Eng. , 131 (1), 20-27.
Kitane, Y., & McGrath, T. J. (2006). Three-Dimensional Modeling of Live Loads on
Culverts. ASCE Conf. Proc. 211, (p. 82).

TxDOT 0-5849 237

Koerner, R. M. (1992). GRI Test Methods and Standards: Geogrid Pullout. Philadelphia,
Pa.: Geosynthetic Research Institute, Drexel University.
Lakmazaheri, S., & Edwards, P. (1996). Integrated Culvert Design and Detailing in a
CAD Environment. Auburn, Al.: Auburn University.
Latona, R. W., Heger, F. J., & Bealey, M. (1973). Computerized Design of Precast
Reinforced Concrete Box Culverts. Washington DC: HRB.
McCarthy, D. F. (2002). Essentials of Soil Mechanics and Foundations. Upper Saddle
River, New Jersey: Prentice-Hall.
McCarthy, D. F. (2007). Essentials of Soil Mechanics and Foundations Basic
Geotechnics. Upper Saddle River, NJ: Pearson Prentice Hall.
Mitchell, J. K., & Villet, C. V. (1987). Reinforcement of Earth Slopes and Embankments.
Washington, D.C.: Transportation Research Board National Research Council.
Mlynarski, M., Katona, M. G., & McGrath, T. J. (2008). CANDE-2007 Culvert Analysis
and Design User Manual and Guideline. Moon Township, PA: National
Cooperative Highway Research Project NCHRP 15-28.
Morris, D., & Delphia, J. (1999). Specifications for Backfill of Reinforced-Earth
Retaining Walls. College Station, Tx: Texas A&M University, Texas
Transportation Institute (TTI).
NAVFAC. (1986). Soil Mechanics: NAVFAC DM 7.01. Alexandria, VA: Naval Facilities
Engineering Command.
Nielsen, M. R., & Anderson, L. R. (1984). Pullout Resistance of Wire Mats Embedded in
Soil for Hilfiker Company. Logan, Utah: Utah State University.
Oswald, C. J. (1996). Analysis of reinforced concrete culvert considering concrete creep
and shrinkage. Washington DC: TRB.
PCA. (1975). Concrete Culverts and Conduits. Skokie, Il.: Portland Cement Association.
PCA. (1992). PCA Soil Primer. Skokie, IL: Portland Cement Association.
Peterson, L. M. (1980). Pullout Resistance of Welded Wire Mesh Embedded in Soil.
Logan, Utah: Utah State University.
Rathje, E., Rauch, A. F., Trejo, D., Folliard, K. J., Viyanant, C., Esfeller, M., et al.
(2006). Evaluation of Crushed Concrete and Recycled Asphalt Pavement As
Backfill for Mechanically Stabilized Earth Walls. Austin, TX: University of Texas
at Austin, Center for Transporation Research (CTR).

TxDOT 0-5849 238

Rathje, E., Trejo, D., & Folliard, K. (2006). Potential Use of Crushed Concrete and
Recycled Asphalt Pavement as Backfill for MSE Walls. Austin, Tx: University of
Texas at Austin, Center for Transportation Research (CTR).
Reese, L. C., & Wang, S. T. (2006, August 16). Verification of Computer Program
LPILE as a Valid Tool for Design of a Single Pile under Lateral Loading.
Retrieved March 25, 2009, from http://www.ensoftin.com.
RISA Technologies. (2006). RISA-3D (Version 6). Foothill Ranch, CA: RISA
Technologies.
Roschke, P. N., & Davis, R. E. (1986). Rigid Culvert Finite Element Analysis.
Geotechnical Eng. , 112 (8), 749-767.
Rund, R. E., & McGrath, T. J. (2000). Comparison of AASHTO standard and LRFD code
provisions for buried concrete box culverts. ASTM Special Technical Publication.
Seed, R., & Raines, J. R. (1988). Failure of Flexible Long-Span Culverts Under
Exceptional Live Loads. Washington DC: TRB.
Seki, G. Y., Pepe, F. J., & Chen, H.-P. (2002). Performance of a Drilled Shaft Retaining
Wall and Abutment in Hawaii. Deep Foundations 2002, (pp. 728-742).
Sharma, S., & Hardcastle, J. H. (1993). Evaluation of culvert deformations using the
finite element method. Washington DC: TRB.
Sisson, R. C., Harris, C. J., & Mokwa, R. L. (2004). Design of Cantilever Soildier Pile
Retaining Walls in Stiff Clays and Claystones. GeoSupport 2004: Innovation and
Cooperation in the Geo-Industry (pp. 309-321). Orlando, Florida: American
Society od Civil Engineers.
Spangler, M. G., Mason, C., & Winfrey, R. (1926). Experimental determinations of static
and impact loads transmitted to culverts. Ames, Ia.: Iowa State College.
Tadros, M. K. (1986). Cost-Effective Concrete Box Culvert Design. Lincoln, Ne: FHWA
and Nebraska University.
Tadros, M. K., & Benak, J. V. (1989). Load Distribution On Box Culverts. Lincoln, Ne.:
FHWA and Nebraska University.
Tadros, M. K., Benak, J. V., & Gilliland, M. K. (1989). Soil Pressure on Box Culverts.
ACI Structural Journal , 86 (4), 439-450.
Tarhini, K., Frederick, G. R., & Mabsout, M. (1996). Finite element analysis of precast
concrete box culverts. Congress on Computing in Civil Engineering, (pp. 677-
682).
TxDOT. (2001). Bridge Design Manual. Austin, Tx: Texas Department of
Transportation.
TxDOT 0-5849 239
TxDOT. (2002). Bridge Inspection Manual. Austin, Tx: Texas Department of
Transportation.
TxDOT. (2003). Concrete Box Culvert Analysis Program - CULV5 Input Guide. Austin:
Texas Department of Transportation.
TxDOT. (2003). CULV5 - Concret Box Culvert Analysis Program. Austin: Texas
Department of Transportation.
TxDOT. (2004). CULV5 (Version 1.71) Readme.pdf. Austin, Tx: Texas Department of
Transportation.
TxDOT. (2009). Culvert Rating Guide. Austin, TX: Texas Department of Transportation.
TxDOT. (2006). Geotechnical Manual. Austin, TX: Texas Department of Transportation.
TxDOT. (2008). Pavement Design Guide. Austin, Tx: Texas Department of
Transportation.
VanTil, e. (1972). PCA Soil Primer. Skokie, IL: Portland Cement Association.
Vaslestad, J., Johansen, T. H., & Holm, W. (1993). Load Reduction on Rigid Culverts
Beneath High Fills: Long-Term Behavior. Washington DC: TRB.
Vidal, H. (1969). The Principle of Reinforced Earth. Highway Research Record No 282,
Washington, D. C.
Wang, S. T., & Reese, L. C. (1986). Study of Design Method for Vertical Drilled Shaft
Retaining Walls. Austin, TX: Center for Tranporation Research.
Wyoming DOT. (2008). Analysis, Design and Rating of Reinforced Concrete Box
Culverts BRASS-CULVERT. Cheyenne, WY: Wyoming Department of
Transportation.
Yang, M. Z. (1997). Influence of Compaction Effort on Earth Pressure on a Box Culvert.
9th International Conference on the Assoc. for Computer Methods and Advances
in Geomechanics, (pp. 2021-2026). Wuhan, China.
Yang, M. Z. (1999). Measurement of earth pressures on concrete box culverts under
highway embankments. Proceedings of the 1998 Field Instrumentation for Soil
and Rock, (pp. 87-100).

TxDOT 0-5849 240

 APPENDIX A
1938 ERA LOAD RATINGS

TxDOT 0-5849 241

 APPENDIX TABLE 1. 1938 ERA SAMPLE DETAILS AND CULV-5 LOAD RATINGS
NO. BARREL BARREL FILL MAX FILL ORIGINAL LAST CULV-5
ID SHEET OF LENGTH HEIGHT RANGE DEPTH DESIGN REVISION IR OR
CCS CFM LC
SPANS (FT) (FT) (FT) (FT) YEAR DATE (HS-) (HS-)
MBC-4-34 2 7X6W0 MBC-4-34 2 7.0 6.0 0.0 0.0 1934 NOV. 1938 7.5 12.5 WBIC M RLL
MBC-5-34 5 9X8W0 MBC-5-34 5 9.0 8.0 0.0 0.0 1934 NOV. 1938 0.0 0.0 0.0 0.0 0.0
MBC-2-34 4 5X5W0 MBC-2-34 4 5.0 5.0 0.0 0.0 1935 NOV. 1938 8.9 14.9 WBEC M TL
MBC-2-34 7 5X5W0 MBC-2-34 7 5.0 5.0 0.0 0.0 1935 NOV. 1938 0.0 0.0 0.0 0.0 0.0
MBC-3-34 3 6X4W0 MBC-3-34 3 6.0 4.0 0.0 0.0 1935 NOV. 1938 8.1 13.5 WBIC M RLL
MBC-6-40 3 10X10W0 MBC-6-40 3 10.0 10.0 0.0 0.0 1935 NOV. 1938 8.4 14.0 WTEC M TL
MBC-11-36 3 5X3W0 MBC-11-36 3 5.0 3.0 0.0 0.0 1936 JAN. 1944 8.1 13.5 WBIC M TL
MBC-12-36 4 5X5W0 MBC-12-36 4 5.0 5.0 0.0 0.0 1936 JAN. 1944 8.9 14.9 WBEC M TL
MBC-12-36 7 5X5W0 MBC-12-36 7 5.0 5.0 0.0 0.0 1936 JAN. 1944 0.0 0.0 0.0 0.0 0.0
MBC-14-36 2 7X7W0 MBC-14-36 2 7.0 7.0 0.0 0.0 1936 FEB. 1944 8.2 13.6 WBIC M RLL
MBC-14-36 3 8X7W0 MBC-14-36 3 8.0 7.0 0.0 0.0 1936 FEB. 1944 7.4 12.3 WBIC M RLL
MBC-15-36 3 8X8W0 MBC-15-36 3 8.0 8.0 0.0 0.0 1936 FEB. 1944 7.9 13.2 WBEC M RLL
MBC-16-36 5 10X8W0 MBC-16-36 5 10.0 8.0 0.0 0.0 1937 FEB. 1944 0.0 0.0 0.0 0.0 0.0
MBC-6-42-F 4 10X9W3 MBC-6-42-F 4 10.0 9.0 0.0 - 3.0 3.0 1935 NOV. 1938 11.3 18.8 WTEC M TL
MBC-16-44-F 2 10X10W3 MBC-16-44-F 2 10.0 10.0 0.0 - 3.0 3.0 1937 NOV. 1938 9.7 16.2 WTEC M TL
MBC-5-34-F 2 9X8W3 MBC-5-34-F 2 9.0 8.0 0.0 - 3.0 3.0 1937 NOV. 1938 2.2 3.7 WBEC M TL
MBC-4-34-F 4 8X7W4 MBC-4-34-F 4 8.0 7.0 0.0 - 4.0 4.0 1938 NOV. 1938 15.7 26.2 WBEC M TL
MBC-1-44-F 2 5X4W6 MBC-1-44-F 2 5.0 4.0 0.0 - 6.0 6.0 1935 OCT. 1938 27.9 46.6 WBEC M TL
MBC-11-40-F 7 5X4W6 MBC-11-40-F 7 5.0 4.0 0.0 - 6.0 6.0 1936 NOV. 1938 0.0 0.0 0.0 0.0 0.0
MBC-12-44-F 5 6X5W6 MBC-12-44-F 5 6.0 5.0 0.0 - 6.0 6.0 1936 NOV. 1938 0.0 0.0 0.0 0.0 0.0
MBC-12-44-F 6 6X6W6 MBC-12-44-F 6 6.0 6.0 0.0 - 6.0 6.0 1936 NOV. 1938 0.0 0.0 0.0 0.0 0.0
MBC-13-44-F 5 6X4W6 MBC-13-44-F 5 6.0 4.0 0.0 - 6.0 6.0 1936 NOV. 1938 0.0 0.0 0.0 0.0 0.0
MBC13-52-F 5 6X4W6 MBC13-52-F 5 6.0 4.0 0.0 - 6.0 6.0 1936 NOV. 1956 0.0 0.0 0.0 0.0 0.0
MBC-14-44-F 3 7X6W6 MBC-14-44-F 3 7.0 6.0 0.0 - 6.0 6.0 1936 NOV. 1938 8.4 14.1 WTEC M RL
MBC-2-34-F 5 5X5W6 MBC-2-34-F 5 5.0 5.0 0.0 - 6.0 6.0 1938 NOV. 1938 0.0 0.0 0.0 0.0 0.0
MBC-3-34-F 4 6X4W6 MBC-3-34-F 4 6.0 4.0 0.0 - 6.0 6.0 1938 JUN. 1939 17.9 29.8 WBEC M RLL
BC-4 1 8X7W8 BC-4 1 8.0 7.0 6.08 - 8.0 8.0 1934 JUL. 1938 18.1 30.2 WBEC M RLL
BC-4 1 9X8W12 BC-4 1 9.0 8.0 10.08 - 12.0 12.0 1934 JUL. 1938 18.8 31.4 WBEC M RLL
BC-4 1 9X9W14 BC-4 1 9.0 9.0 12.08 - 14.0 14.0 1934 JUL. 1938 -49.9 -83.2 BEM M RLL
BC-4 1 10X10W18 BC-4 1 10.0 10.0 16.08 - 18.0 18.0 1934 JUL. 1938 40.1 66.9 WEM M TL

TxDOT 0-5849 242

 APPENDIX TABLE 2. 1938 ERA RISA-2D WITH SPRINGS LOAD RATINGS
RISA-2D SPG K=75PCI RISA-2D SPG K=150PCI RISA-2D SPG K=250PCI
ID IR OR IR OR IR OR
CCS CFM LC CCS CFM LC CCS CFM LC
(HS-) (HS-) (HS-) (HS-) (HS-) (HS-)
MBC-4-34 2 7X6W0 7.5 12.6 WBEC M RLL 8.3 13.8 WBEC M RLL 9.1 15.2 WBEC M RLL
MBC-5-34 5 9X8W0 7.1 11.8 WBEC M RLL 8.2 13.7 WBEC M RLL 9.6 16.1 WBEC M RLL
MBC-2-34 4 5X5W0 5.2 8.8 WBEC M RLL 5.7 9.5 WBEC M RLL 6.2 10.4 WBEC M RLL
MBC-2-34 7 5X5W0 4.9 8.2 WBEC M RLL 5.4 9.0 WBEC M RLL 5.9 9.9 WBEC M RLL
MBC-3-34 3 6X4W0 7.3 12.1 WTIC M RLL 8.0 13.3 WTIC M RLL 8.7 14.5 WTIC M RLL
MBC-6-40 3 10X10W0 6.7 11.3 WBEC M RLL 8.1 13.6 WTEC M RLL 9.7 16.1 WBEC M RLL
MBC-11-36 3 5X3W0 5.8 9.7 WBEC M RLL 6.1 10.2 WBEC M RLL 6.5 10.8 WBEC M RLL
MBC-12-36 4 5X5W0 5.2 8.8 WBEC M RLL 5.7 9.5 WBEC M RLL 6.2 10.4 WBEC M RLL
MBC-12-36 7 5X5W0 4.9 8.2 WBEC M RLL 5.4 9.0 WBEC M RLL 5.8 9.7 WBEC M RLL
MBC-14-36 2 7X7W0 7.4 12.3 WBEC M RLL 8.2 13.6 WBEC M RLL 9.1 15.1 WBEC M RLL
MBC-14-36 3 8X7W0 7.3 12.1 WBEC M RLL 8.0 13.4 WBEC M RLL 9.0 15.1 WBEC M RLL
MBC-15-36 3 8X8W0 7.6 12.7 WBEC M RLL 8.5 14.1 WBEC M RLL 9.5 15.8 WBEC M RLL
MBC-16-36 5 10X8W0 7.2 12.0 WBEC M RLL 8.8 14.6 WBEC M RLL 10.4 17.4 WBEC M RLL
MBC-6-42-F 4 10X9W3 19.7 33.0 WBEC M RLL 20.1 33.6 WTEC M TL 19.5 32.5 WTEC M TL
MBC-16-44-F 2 10X10W3 19.0 31.6 WTEC M TL 18.1 30.2 WTEC M TL 17.3 28.8 WTEC M TL
MBC-5-34-F 2 9X8W3 5.0 8.3 WBEC M RLL 6.4 10.6 WBEC M RLL 8.2 13.6 WBEC M RLL
MBC-4-34-F 4 8X7W4 20.5 34.3 WBEC M RLL 24.2 40.4 WBEC M RLL 26.1 43.6 WTEC M TL
MBC-1-44-F 2 5X4W6 39.4 65.7 WBEC M RLL 40.6 67.8 WBEC M RLL 42.6 71.1 WBEC M RLL
MBC-11-40-F 7 5X4W6 38.3 64.0 WTIC M RLL 48.9 81.6 WTIC M TL 58.4 97.6 WTIC M TL
MBC-12-44-F 5 6X5W6 25.0 41.7 WBEC M RLL 26.7 44.5 WBEC M RLL 29.4 49.1 WBEC M RLL
MBC-12-44-F 6 6X6W6 22.7 37.8 WBEC M RLL 24.7 41.3 WBEC M RLL 27.3 45.6 WBEC M RLL
MBC-13-44-F 5 6X4W6 23.2 38.7 WBEC M RLL 24.5 41.0 WBEC M RLL 26.7 44.6 WBEC M RLL
MBC13-52-F 5 6X4W6 23.2 38.7 WBEC M RLL 24.5 41.0 WBEC M RLL 26.7 44.6 WBEC M RLL
MBC-14-44-F 3 7X6W6 23.6 39.4 WTEC M RLL 24.7 41.2 WTEC M RLL 24.4 40.8 WTEC M RLL
MBC-2-34-F 5 5X5W6 32.0 53.4 WBEC M RLL 33.5 55.9 WBEC M RLL 35.8 59.7 WBEC M RLL
MBC-3-34-F 4 6X4W6 26.5 44.2 WBEC M RLL 28.4 47.3 WBEC M RLL 31.0 51.7 WBEC M RLL
BC-4 1 8X7W8 28.9 48.3 BEM M RLL 33.1 55.2 WEM M TL 31.5 52.6 WEM M TL
BC-4 1 9X8W12 43.7 73.0 WEM M TL 42.3 70.7 WEM M TL 40.6 67.7 WEM M TL
BC-4 1 9X9W14 -111.8 -186.7 BEM M TL -76.2 ##### BEM M TL -52.4 -87.4 BEM M RLL
BC-4 1 10X10W18 4.7 7.8 WEM M TL 3.7 6.1 WEM M TL 2.6 4.4 WEM M TL

TxDOT 0-5849 243

 APPENDIX TABLE 3. 1938 ERA RISA-2D WITH LEFE LOAD RATINGS
RISA-2D LEFE E=8KSI RISA-2D LEFE E=20KSI RISA-2D LEFE E=36KSI
ID IR OR IR OR IR OR
CCS CFM CCS CFM CCS CFM
(HS-) (HS-) (HS-) (HS-) (HS-) (HS-)
MBC-4-34 2 7X6W0 10.3 17.2 WBEC M 11.5 19.1 WTIC M 11.8 19.7 WTIC M
MBC-5-34 5 9X8W0 2.4 4.0 WTIC M 7.3 12.2 WTIC M 9.4 15.6 WTIC M
MBC-2-34 4 5X5W0 5.3 8.9 WBEC M 10.1 16.8 WBEC M 13.9 23.2 WTIC M
MBC-2-34 7 5X5W0 5.4 9.0 WBEC M 10.4 17.4 WBEC M 14.9 24.9 WTIC M
MBC-3-34 3 6X4W0 10.4 17.3 WTIC M 12.5 20.8 WTIC M 13.5 22.6 WTIC M
MBC-6-40 3 10X10W0 -2.7 -4.5 WBEC M 9.2 15.3 WBEC M 14.9 24.9 WTIC M
MBC-11-36 3 5X3W0 8.7 14.5 WTIC M 10.9 18.1 WTIC M 12.6 21.0 WTIC M
MBC-12-36 4 5X5W0 5.3 8.8 WBEC M 10.1 16.8 WBEC M 13.9 23.2 WTIC M
MBC-12-36 7 5X5W0 5.4 9.0 WBEC M 10.4 17.4 WBEC M 14.8 24.8 WTIC M
MBC-14-36 2 7X7W0 7.8 13.0 WBEC M 12.1 20.2 WTIC M 12.6 21.1 WTIC M
MBC-14-36 3 8X7W0 4.2 7.0 WTIC M 8.0 13.3 WTIC M 9.7 16.3 WTIC M
MBC-15-36 3 8X8W0 2.8 4.6 WTIC M 7.0 11.7 WTIC M 9.2 15.3 WTIC M
MBC-16-36 5 10X8W0 4.3 7.2 WBEC M 15.0 25.1 WTIC M 16.4 27.4 WTEC M
MBC-6-42-F 4 10X9W3 2.6 4.3 WBEC M 22.6 37.7 WTIC M 33.1 55.3 WTIC M
MBC-16-44-F 2 10X10W3 -0.3 -0.5 WBEC M 21.8 36.5 WBEC M 39.1 65.2 WTIC M
MBC-5-34-F 2 9X8W3 -11.0 -18.4 WBEC M -0.2 -0.4 WBEC M 13.3 22.2 WBEC M
MBC-4-34-F 4 8X7W4 1.4 2.3 WBEC M 26.6 44.4 WBEC M 43.5 72.5 WTIC M
MBC-1-44-F 2 5X4W6 13.8 23.1 WBEC M 28.2 47.1 WBEC M 46.6 77.7 WBEC M
MBC-11-40-F 7 5X4W6 24.9 41.5 WTIC M 53.8 89.8 WTIC M 81.1 135.4 WTIC M
MBC-12-44-F 5 6X5W6 -0.3 -0.6 WBEC M 21.8 36.5 WBEC M 46.7 77.9 WBEC M
MBC-12-44-F 6 6X6W6 -8.5 -14.1 WBEC M 10.8 18.0 WBEC M 32.0 53.4 WBEC M
MBC-13-44-F 5 6X4W6 9.6 16.1 WBEC M 34.2 57.1 WBEC M 62.4 104.1 WBEC M
MBC13-52-F 5 6X4W6 9.6 16.1 WBEC M 34.2 57.1 WBEC M 62.4 104.1 WBEC M
MBC-14-44-F 3 7X6W6 9.6 15.9 WBEC M 40.1 67.0 WBEC M 54.5 91.0 WTEC M
MBC-2-34-F 5 5X5W6 -0.3 -0.6 WBEC M 21.1 35.3 WBEC M 45.5 76.0 WBEC M
MBC-3-34-F 4 6X4W6 9.3 15.5 WBEC M 32.7 54.6 WBEC M 59.6 99.5 WBEC M
BC-4 1 8X7W8 26.5 44.2 WBEC M 68.3 114.0 WBEC M 124.9 208.5 WBEC M
BC-4 1 9X8W12 24.0 40.0 WBEC M 70.7 118.0 WBEC M 136.1 227.1 WBEC M
BC-4 1 9X9W14 -11517.5 -19225.4 TEM M -890.0 -1485.7 BEM M 19.0 31.6 BEM M
BC-4 1 10X10W18 23.7 39.5 WBEC M 64.6 107.8 WBEC M 126.1 210.4 WBEC M

TxDOT 0-5849 244

 APPENDIX B
1946 ERA LOAD RATINGS

TxDOT 0-5849 245

 APPENDIX TABLE 4. 1946 ERA SAMPLE DETAILS AND CULV-5 LOAD RATINGS
NO. BARREL BARREL FILL MAX FILL ORIGINAL LAST CULV-5
ID SHEET OF LENGTH HEIGHT RANGE DEPTH DESIGN REVISION IR OR
CCS CFM LC
SPANS (FT) (FT) (FT) (FT) YEAR DATE (HS-) (HS-)
MBC-3 7 3X3W0 MBC-3 7 3.0 3.0 0.0 0.0 1946 MAR. 1946 0.0 0.0 0.0 0.0 0.0
MBC-3 8 5X5W0 MBC-3 8 5.0 5.0 0.0 0.0 1946 MAR. 1946 0.0 0.0 0.0 0.0 0.0
MC10-3-45 4 10X10W6 MC10-3-45 4 10.0 10.0 4.08 - 6.0 6.0 THD#1 1949 DEC. 1949 -119.3 -199.1 WEM M TL
MC10-3-45 2 10X6W6 MC10-3-45 2 10.0 6.0 4.08 - 6.0 6.0 THD#1 1949 DEC. 1949 7.4 12.3 WBEC M TL
MC10-3-45 5 10X10W6 MC10-3-45 5 10.0 10.0 4.08 - 6.0 6.0 THD#1 1949 DEC. 1949 0.0 0.0 WEM M TL
MC10-3-45 3 10X7W6 MC10-3-45 3 10.0 7.0 4.08 - 6.0 6.0 THD#1 1949 DEC. 1949 5.1 8.5 BEM M RLL
FM-MBC-3-26 4 6X6W6 FM-MBC-3-26 4 6.0 6.0 0.0 - 6.0 6.0 1946 DEC. 1948 20.2 33.7 WTEC M TL
FM-MBC-3-26 6 6X6W6 FM-MBC-3-26 6 6.0 6.0 0.0 - 6.0 6.0 1946 DEC. 1948 0.0 0.0 WTEC M TL
FM-MBC-3-26 3 6X5W6 FM-MBC-3-26 3 6.0 5.0 0.0 - 6.0 6.0 1946 DEC. 1948 21.0 35.1 WBEC M RLL
FM-MBC-3-26 5 7X4W6 FM-MBC-3-26 5 7.0 4.0 0.0 - 6.0 6.0 1946 DEC. 1948 0.0 0.0 0.0 0.0 0.0

APPENDIX TABLE 5. 1946 ERA RISA-2D WITH SPRINGS LOAD RATINGS

RISA-2D SPG K=75PCI RISA-2D SPG K=150PCI RISA-2D SPG K=250PCI
ID IR OR IR OR IR OR
CCS CFM LC CCS CFM LC CCS CFM LC
(HS-) (HS-) (HS-) (HS-) (HS-) (HS-)
MBC-3 7 3X3W0 2.9 4.9 WTIC M RLL 3.4 5.8 WTIC M RLL 3.9 6.5 WTEC M RLL
MBC-3 8 5X5W0 9.6 16.0 WTEC M RLL 11.3 18.9 WTEC M RLL 12.8 21.3 WTIC M RLL
MC10-3-45 4 10X10W6 -71.0 -118.5 WEM M TL -68.1 -113.6 WEM M TL -65.9 -109.9 WEM M TL
MC10-3-45 2 10X6W6 7.8 13.0 WBEC M TL 11.0 18.3 WTEC M TL 10.0 16.7 WTEC M TL
MC10-3-45 5 10X10W6 -73.8 -123.2 WEM M TL -70.3 -117.4 WEM M TL -67.7 -113.0 WEM M TL
MC10-3-45 3 10X7W6 5.5 9.1 WBEC M TL 9.3 15.6 TEM M RLL 9.5 15.9 TEM M RLL
FM-MBC-3-26 4 6X6W6 34.6 57.7 WTEC M RLL 35.3 58.9 WTEC M RLL 34.5 57.5 WTEC M RLL
FM-MBC-3-26 6 6X6W6 34.4 57.4 WTEC M RLL 35.0 58.4 WTEC M RLL 34.2 57.0 WTEC M RLL
FM-MBC-3-26 3 6X5W6 27.6 46.1 WBEC M RLL 31.1 51.9 WBEC M RLL 31.7 52.9 WTEC M RLL
FM-MBC-3-26 5 7X4W6 17.8 29.7 WBEC M RLL 20.8 34.8 WBEC M RLL 23.0 38.4 WBEC M RLL

TxDOT 0-5849 246

 APPENDIX TABLE 6. 1946 ERA RISA-2D WITH LEFE LOAD RATINGS
RISA-2D LEFE E=8KSI RISA-2D LEFE E=20KSI RISA-2D LEFE E=36KSI
ID IR OR IR OR IR OR
CCS CFM CCS CFM CCS CFM
(HS-) (HS-) (HS-) (HS-) (HS-) (HS-)
MBC-3 7 3X3W0 3.1 5.2 WTIC M 4.9 8.2 WTIC M 5.8 9.6 WTIC M
MBC-3 8 5X5W0 10.6 17.6 WTIC M 15.8 26.4 WTIC M 18.2 30.5 WTIC M
MC10-3-45 4 10X10W6 -26.9 -44.9 WBEC M 0.7 1.2 WBEC M 30.4 50.7 WBEC M
MC10-3-45 2 10X6W6 -9.1 -15.2 WBEC M 9.7 16.3 WBEC M 30.8 51.3 WBEC M
MC10-3-45 5 10X10W6 -26.0 -43.4 WBEC M 2.4 4.0 WBEC M 33.2 55.4 WBEC M
MC10-3-45 3 10X7W6 -14.9 -24.9 WBEC M 8.9 14.9 WBEC M 35.2 58.8 WBEC M
FM-MBC-3-26 4 6X6W6 6.2 10.3 WBEC M 35.6 59.4 WBEC M 69.0 115.2 WBEC M
FM-MBC-3-26 6 6X6W6 6.8 11.4 WBEC M 37.4 62.5 WBEC M 72.1 120.4 WBEC M
FM-MBC-3-26 3 6X5W6 11.3 18.8 WBEC M 39.7 66.2 WBEC M 72.5 120.9 WBEC M
FM-MBC-3-26 5 7X4W6 20.6 34.4 WBEC M 56.5 94.2 WTEC M 83.3 139.1 WTEC M

TxDOT 0-5849 247

 APPENDIX C
1958 ERA 36KSI STEEL LOAD RATINGS

TxDOT 0-5849 248

 APPENDIX TABLE 7. 1958 ERA 36KSI STEEL SAMPLE DETAILS AND CULV-5 LOAD RATINGS
NO. BARREL BARREL FILL MAX FILL ORIGINAL LAST CULV-5
ID SHEET OF LENGTH HEIGHT RANGE DEPTH DESIGN REVISION IR OR
CCS CFM LC
SPANS (FT) (FT) (FT) (FT) YEAR DATE (HS-) (HS-)
MC10-1 5 10X7W2 MC10-1 5 10.0 7.0 0.0 - 2.0 2.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC10-1 6 10X9W2 MC10-1 6 10.0 9.0 0.0 - 2.0 2.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC9-1 5 9X8W2 MC9-1 5 9.0 8.0 0.0 - 2.0 2.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC9-1 4 9X6W2 MC9-1 4 9.0 6.0 0.0 - 2.0 2.0 1958 NOV. 1964 4.2 7.0 BEM M RLL
MC7-1 3 7X6W2 MC7-1 3 7.0 6.0 0.0 - 2.0 2.0 1958 NOV. 1964 4.3 7.2 BEM M RLL
MC7-1 2 7X4W2 MC7-1 2 7.0 4.0 0.0 - 2.0 2.0 1958 NOV. 1964 4.7 7.8 BEM M RLL
MC8-1 4 8X7W2 MC8-1 4 8.0 7.0 0.0 - 2.0 2.0 1958 NOV. 1964 3.9 6.4 BEM M RLL
MC8-1 3 8X5W2 MC8-1 3 8.0 5.0 0.0 - 2.0 2.0 1958 NOV. 1964 4.2 7.0 BEM M RLL
MC10-1 6 10X8W2 MC10-1 6 10.0 8.0 0.0 - 2.0 2.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC8-1 4 8X6W2 MC8-1 4 8.0 6.0 0.0 - 2.0 2.0 1958 NOV. 1964 4.0 6.7 BEM M RLL
MC7-1 6 7X7W2 MC7-1 6 7.0 7.0 0.0 - 2.0 2.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC6-1 5 6X5W4 MC6-1 5 6.0 5.0 0.0 - 4.0 4.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC7-2 5 7X5W4 MC7-2 5 7.0 5.0 2.08 - 4.0 4.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC7-2 6 7X7W4 MC7-2 6 7.0 7.0 2.08 - 4.0 4.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC8-2 5 8X4W4 MC8-2 5 8.0 4.0 2.08 - 4.0 4.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC8-2 6 8X6W4 MC8-2 6 8.0 6.0 2.08 - 4.0 4.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC9-2 6 9X5W4 MC9-2 6 9.0 5.0 2.08 - 4.0 4.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC10-2 2 10X9W4 MC10-2 2 10.0 9.0 2.08 - 4.0 4.0 1958 NOV. 1964 0.4 0.7 WEM M TL
MC6-1 4 6X3W4 MC6-1 4 6.0 3.0 0.0 - 4.0 4.0 1958 NOV. 1964 8.7 14.5 BEM M RLL
MC7-2 3 7X3W4 MC7-2 3 7.0 3.0 2.08 - 4.0 4.0 1958 NOV. 1964 10.7 17.8 BEM M RLL
MC7-2 4 7X3W4 MC7-2 4 7.0 3.0 2.08 - 4.0 4.0 1958 NOV. 1964 10.9 18.2 BEM M RLL
MC7-2 2 7X3W4 MC7-2 2 7.0 3.0 2.08 - 4.0 4.0 1958 NOV. 1964 11.7 19.5 BEM M RLL
MC8-2 5 8X6W4 MC8-2 5 8.0 6.0 2.08 - 4.0 4.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC9-2 2 9X9W4 MC9-2 2 9.0 9.0 2.08 - 4.0 4.0 1958 NOV. 1964 -2.9 -4.8 WEM M TL

TxDOT 0-5849 249

 APPENDIX TABLE 7. 1958 ERA 36KSI STEEL SAMPLE DETAILS AND CULV-5 LOAD RATINGS (CONT.)
NO. BARREL BARREL FILL MAX FILL ORIGINAL LAST CULV-5
ID SHEET OF LENGTH HEIGHT RANGE DEPTH DESIGN REVISION IR OR
CCS CFM LC
SPANS (FT) (FT) (FT) (FT) YEAR DATE (HS-) (HS-)
MC9-3 4 9X9W6 MC9-3 4 9.0 9.0 4.08 - 6.0 6.0 1958 NOV. 1964 -186.0 -310.5 WEM M TL
MC10-3 5 10X10W6 MC10-3 5 10.0 10.0 4.08 - 6.0 6.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC6-2 6 6X3W6 MC6-2 6 6.0 3.0 4.08 - 6.0 6.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC8-3 3 8X8W6 MC8-3 3 8.0 8.0 4.08 - 6.0 6.0 1958 NOV. 1964 4.7 7.9 WEM M TL
MC9-3 3 9X7W6 MC9-3 3 9.0 7.0 4.08 - 6.0 6.0 1958 NOV. 1964 5.2 8.6 WBEC M TL
MC5-2 3 5X4W6 MC5-2 3 5.0 4.0 4.08 - 6.0 6.0 1958 NOV. 1964 7.7 12.9 BEM M RLL
MC10-3 4 10X8W6 MC10-3 4 10.0 8.0 4.08 - 6.0 6.0 1958 NOV. 1964 8.8 14.7 WBEC M TL
MC10-3 3 10X6W6 MC10-3 3 10.0 6.0 4.08 - 6.0 6.0 1958 NOV. 1964 9.2 15.4 BEM M RLL
MC8-3 2 8X6W6 MC8-3 2 8.0 6.0 4.08 - 6.0 6.0 1958 NOV. 1964 11.1 18.5 BEM M RLL
MC9-3 2 9X5W6 MC9-3 2 9.0 5.0 4.08 - 6.0 6.0 1958 NOV. 1964 11.4 19.0 WBEC M TL
MC5-2 2 5X2W6 MC5-2 2 5.0 2.0 4.08 - 6.0 6.0 1958 NOV. 1964 12.7 21.3 BEM M RLL
MC7-3 2 7X7W6 MC7-3 2 7.0 7.0 4.08 - 6.0 6.0 1958 NOV. 1964 8.6 14.3 BIC M TL
MC7-3 4 7X4W6 MC7-3 4 7.0 4.0 4.08 - 6.0 6.0 1958 NOV. 1964 12.8 21.4 WBEC M TL
MC8-3 6 8X6W6 MC8-3 6 8.0 6.0 4.08 - 6.0 6.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC9-3 3 9X9W6 MC9-3 3 9.0 9.0 4.08 - 6.0 6.0 1958 NOV. 1964 -150.0 -250.4 WEM M TL

TxDOT 0-5849 250

 APPENDIX TABLE 8. 1958 ERA 36KSI STEEL RISA-2D WITH SPRINGS LOAD RATINGS
RISA-2D SPG K=75PCI RISA-2D SPG K=150PCI RISA-2D SPG K=250PCI
ID IR OR IR OR IR OR
CCS CFM LC CCS CFM LC CCS CFM LC
(HS-) (HS-) (HS-) (HS-) (HS-) (HS-)
MC10-1 5 10X7W2 5.2 8.6 WBEC M TL 5.9 9.9 WTEC M TL 5.6 9.4 WTEC M TL
MC10-1 6 10X9W2 1.3 2.1 WEM M TL 0.7 1.2 WEM M TL 0.2 0.3 WEM M TL
MC9-1 5 9X8W2 5.3 8.8 BEC M TL 7.4 12.4 TEC M TL 7.1 11.9 TEC M TL
MC9-1 4 9X6W2 5.5 9.2 WBEC M TL 6.4 10.7 WTEC M TL 6.2 10.3 WTEC M TL
MC7-1 3 7X6W2 5.8 9.7 BEM M RLL 6.8 11.3 BEM M RLL 7.3 12.2 WTIC M RLL
MC7-1 2 7X4W2 6.2 10.3 BEM M RLL 7.2 12.1 BEM M RLL 7.6 12.6 WTEC M TL
MC8-1 4 8X7W2 4.8 7.9 BEC M TL 6.3 10.6 WBEC M TL 6.9 11.5 WTEC M TL
MC8-1 3 8X5W2 6.1 10.2 WTIC M RLL 6.0 10.1 WTIC M RLL 6.0 10.1 WTIC M RLL
MC10-1 6 10X8W2 5.5 9.3 WBEC M TL 6.9 11.5 WTEC M TL 6.6 11.0 WTEC M TL
MC8-1 4 8X6W2 6.2 10.4 WBEC M TL 6.7 11.1 WTIC M RLL 6.6 11.1 WTIC M RLL
MC7-1 6 7X7W2 1.8 3.0 TIC M RLL 2.1 3.5 TIC M RLL 2.4 3.9 TIC M RLL
MC6-1 5 6X5W4 10.3 17.3 BEM M RLL 12.0 20.1 BEM M RLL 14.5 24.2 BEM M RLL
MC7-2 5 7X5W4 10.6 17.6 WBEC M TL 11.8 19.7 TEM M RLL 11.9 19.9 TEM M RLL
MC7-2 6 7X7W4 3.8 6.3 BEC M TL 5.6 9.4 BEC M TL 7.9 13.1 BEC M TL
MC8-2 5 8X4W4 8.6 14.4 TEM M RLL 8.7 14.6 TEM M RLL 8.9 14.8 TEM M RLL
MC8-2 6 8X6W4 9.0 15.0 TEM M RLL 9.1 15.2 TEM M RLL 9.2 15.4 TEM M RLL
MC9-2 6 9X5W4 8.7 14.6 TEM M RLL 8.8 14.8 TEM M RLL 9.0 15.0 TEM M RLL
MC10-2 2 10X9W4 -0.5 -0.8 WEM M TL -1.1 -1.8 WEM M TL -1.7 -2.9 WEM M TL
MC6-1 4 6X3W4 11.8 19.7 BEM M RLL 13.6 22.7 BEM M RLL 16.4 27.4 BEM M RLL
MC7-2 3 7X3W4 9.6 16.1 WBEC M TL 11.5 19.2 WTEC M TL 11.2 18.7 WTEC M TL
MC7-2 4 7X3W4 9.9 16.6 WBEC M TL 11.4 19.0 WTEC M TL 11.2 18.6 WTEC M TL
MC7-2 2 7X3W4 10.9 18.1 WBEC M TL 11.6 19.3 WTEC M TL 11.4 19.0 WTEC M TL
MC8-2 5 8X6W4 9.0 15.0 TEM M RLL 9.1 15.2 TEM M RLL 9.2 15.4 TEM M RLL
MC9-2 2 9X9W4 -3.7 -6.2 WEM M TL -4.1 -6.9 WEM M TL -4.7 -7.8 WEM M TL

TxDOT 0-5849 251

 APPENDIX TABLE 8. 1958 ERA 36KSI STEEL RISA-2D WITH SPRINGS LOAD RATINGS (CONT.)
RISA-2D SPG K=75PCI RISA-2D SPG K=150PCI RISA-2D SPG K=250PCI
ID IR OR IR OR IR OR
CCS CFM LC CCS CFM LC CCS CFM LC
(HS-) (HS-) (HS-) (HS-) (HS-) (HS-)
MC9-3 4 9X9W6 -47.5 -79.4 WEM M TL -44.9 -74.9 WEM M TL -42.9 -71.7 WEM M TL
MC10-3 5 10X10W6 -67.3 -112.4 WEM M TL -64.3 -107.4 WEM M TL -62.0 -103.6 WEM M TL
MC6-2 6 6X3W6 15.8 26.3 TEM M RLL 15.9 26.6 TEM M RLL 16.1 26.9 TEM M RLL
MC8-3 3 8X8W6 3.9 6.5 WEM M TL 3.3 5.5 WEM M TL 2.5 4.2 WEM M TL
MC9-3 3 9X7W6 4.9 8.2 WBEC M TL 8.6 14.4 WBEC M TL 9.7 16.2 WTEC M TL
MC5-2 3 5X4W6 8.9 14.9 BEM M RLL 10.3 17.2 BEM M RLL 11.4 19.0 TEM M RLL
MC10-3 4 10X8W6 8.4 13.9 WBEC M TL 11.6 19.3 WEM M TL 10.3 17.2 WEM M TL
MC10-3 3 10X6W6 11.3 18.9 WBEC M TL 13.5 22.5 TEM M RLL 13.3 22.2 WTEC M TL
MC8-3 2 8X6W6 15.6 26.1 BEM M RLL 15.7 26.3 TEM M RLL 15.9 26.5 TEM M RLL
MC9-3 2 9X5W6 13.0 21.7 WBEC M TL 14.4 24.0 WTEC M TL 13.7 22.9 WTEC M TL
MC5-2 2 5X2W6 13.7 22.9 BEM M RLL 14.9 25.0 BEM M RLL 15.4 25.7 TEM M RLL
MC7-3 2 7X7W6 6.9 11.5 BEC M TL 7.3 12.1 WEM M TL 6.7 11.2 WEM M TL
MC7-3 4 7X4W6 14.0 23.3 WBEC M TL 16.6 27.7 WTEC M TL 16.0 26.7 WTEC M TL
MC8-3 6 8X6W6 13.8 23.0 TEM M RLL 14.0 23.4 TEM M RLL 14.3 23.9 TEM M RLL
MC9-3 3 9X9W6 -57.6 -96.1 WEM M TL -51.9 -86.6 WEM M TL -48.5 -80.9 WEM M TL

TxDOT 0-5849 252

 APPENDIX TABLE 9. 1958 ERA 36KSI STEEL RISA-2D WITH LEFE LOAD RATINGS
RISA-2D LEFE E=8KSI RISA-2D LEFE E=20KSI RISA-2D LEFE E=36KSI
ID IR OR IR OR IR OR
CCS CFM CCS CFM CCS CFM
(HS-) (HS-) (HS-) (HS-) (HS-) (HS-)
MC10-1 5 10X7W2 1.4 2.3 WBEC M 9.9 16.6 WTIC M 12.6 21.0 WTEC M
MC10-1 6 10X9W2 -3.8 -6.3 BEC M 9.5 15.8 WTIC M 13.3 22.3 WTIC M
MC9-1 5 9X8W2 -1.4 -2.4 BEC M 9.9 16.5 WTIC M 13.8 23.0 WTIC M
MC9-1 4 9X6W2 4.0 6.6 WBEC M 10.1 16.8 WTIC M 13.0 21.8 WTEC M
MC7-1 3 7X6W2 3.1 5.3 BEC M 12.5 20.8 WTIC M 16.4 27.4 WTEC M
MC7-1 2 7X4W2 9.5 15.8 WBEC M 13.2 22.0 WTEC M 14.5 24.3 WTEC M
MC8-1 4 8X7W2 0.4 0.6 BEC M 11.0 18.4 WTIC M 15.2 25.5 WTIC M
MC8-1 3 8X5W2 7.0 11.6 WTIC M 10.5 17.6 WTIC M 13.8 23.1 WTIC M
MC10-1 6 10X8W2 -0.5 -0.9 BEC M 9.6 15.9 WTIC M 13.0 21.7 WTIC M
MC8-1 4 8X6W2 5.0 8.3 BEC M 11.0 18.4 WTIC M 14.8 24.6 WTIC M
MC7-1 6 7X7W2 -0.6 -1.0 TIC M 4.0 6.7 TIC M 7.3 12.1 TIC M
MC6-1 5 6X5W4 7.0 11.7 BEC M 26.2 43.7 BEC M 45.0 75.1 BEC M
MC7-2 5 7X5W4 5.2 8.6 WBEC M 22.4 37.5 WBEC M 38.1 63.5 WTEC M
MC7-2 6 7X7W4 -8.8 -14.7 BEC M 6.3 10.6 WBEC M 22.2 37.0 WBEC M
MC8-2 5 8X4W4 11.8 19.8 WBEC M 27.5 45.9 WTEC M 35.6 59.3 WTEC M
MC8-2 6 8X6W4 0.6 0.9 WBEC M 19.5 32.5 WBEC M 37.4 62.5 WTEC M
MC9-2 6 9X5W4 5.8 9.6 WBEC M 25.6 42.8 WBEC M 32.9 55.0 WTEC M
MC10-2 2 10X9W4 -10.9 -18.2 WBEC M 9.0 15.0 WBEC M 30.5 50.9 WBEC M
MC6-1 4 6X3W4 18.9 31.6 WBEC M 32.7 54.6 WTEC M 41.5 69.2 WTEC M
MC7-2 3 7X3W4 14.7 24.6 WBEC M 25.9 43.2 WTEC M 34.3 57.2 WTEC M
MC7-2 4 7X3W4 14.7 24.5 WBEC M 25.9 43.2 WTEC M 34.7 57.9 WTEC M
MC7-2 2 7X3W4 15.4 25.7 WBEC M 25.7 43.0 WTEC M 34.0 56.8 WTEC M
MC8-2 5 8X6W4 0.2 0.4 WBEC M 19.2 32.0 WBEC M 37.6 62.8 WTEC M
MC9-2 2 9X9W4 -14.6 -24.3 WBEC M 1.7 2.9 WBEC M 19.5 32.6 WBEC M

TxDOT 0-5849 253

 APPENDIX TABLE 9. 1958 ERA 36KSI STEEL RISA-2D WITH LEFE LOAD RATINGS (CONT.)
RISA-2D LEFE E=8KSI RISA-2D LEFE E=20KSI RISA-2D LEFE E=36KSI
ID IR OR IR OR IR OR
CCS CFM CCS CFM CCS CFM
(HS-) (HS-) (HS-) (HS-) (HS-) (HS-)
MC9-3 4 9X9W6 -21.6 -36.1 WBEC M 3.4 5.6 WBEC M 30.4 50.8 WBEC M
MC10-3 5 10X10W6 -19.2 -32.0 WBEC M 12.8 21.3 WBEC M 47.0 78.4 WBEC M
MC6-2 6 6X3W6 24.9 41.5 WBEC M 47.5 79.3 TEM M 59.0 98.4 TIC V
MC8-3 3 8X8W6 -12.1 -20.3 WBEC M 10.5 17.5 WBEC M 34.8 58.0 WBEC M
MC9-3 3 9X7W6 -16.4 -27.3 WBEC M 4.8 8.0 WBEC M 27.7 46.2 WBEC M
MC5-2 3 5X4W6 4.6 7.6 WBEC M 21.9 36.6 WBEC M 40.0 66.7 WBEC M
MC10-3 4 10X8W6 -17.7 -29.5 WBEC M 7.1 11.9 WBEC M 33.7 56.3 WBEC M
MC10-3 3 10X6W6 -6.2 -10.4 WBEC M 17.6 29.4 WBEC M 43.0 71.7 WBEC M
MC8-3 2 8X6W6 -0.1 -0.2 WBEC M 20.6 34.3 WBEC M 43.5 72.5 WBEC M
MC9-3 2 9X5W6 1.1 1.9 WBEC M 22.5 37.5 WBEC M 43.5 72.7 WTEC M
MC5-2 2 5X2W6 17.3 28.9 TEM M 28.8 48.1 TEM M 45.4 75.7 TEM M
MC7-3 2 7X7W6 -7.0 -11.7 WBEC M 9.9 16.5 WBEC M 29.2 48.8 WBEC M
MC7-3 4 7X4W6 7.5 12.5 WBEC M 31.3 52.3 WBEC M 55.8 93.2 WTEC M
MC8-3 6 8X6W6 -1.1 -1.8 WBEC M 23.4 39.1 WBEC M 49.9 83.3 WBEC M
MC9-3 3 9X9W6 -21.3 -35.6 WBEC M 2.8 4.6 WBEC M 28.9 48.2 WBEC M

TxDOT 0-5849 254

 APPENDIX D
1958 ERA 60KSI STEEL LOAD RATINGS

TxDOT 0-5849 255

 APPENDIX TABLE 10. 1958 ERA 60KSI STEEL SAMPLE DETAILS AND CULV-5 LOAD RATINGS
NO. BARREL BARREL FILL MAX FILL ORIGINAL LAST CULV-5
ID SHEET OF LENGTH HEIGHT RANGE DEPTH DESIGN REVISION IR OR
CCS CFM LC
SPANS (FT) (FT) (FT) (FT) YEAR DATE (HS-) (HS-)
MC10-1 5 10X7W2 MC10-1 5 10.0 7.0 0.0 - 2.0 2.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC10-1 6 10X9W2 MC10-1 6 10.0 9.0 0.0 - 2.0 2.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC9-1 5 9X8W2 MC9-1 5 9.0 8.0 0.0 - 2.0 2.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC9-1 4 9X6W2 MC9-1 4 9.0 6.0 0.0 - 2.0 2.0 1958 NOV. 1964 10.2 17.0 BEM M RLL
MC7-1 3 7X6W2 MC7-1 3 7.0 6.0 0.0 - 2.0 2.0 1958 NOV. 1964 10.1 16.9 BEM M RLL
MC7-1 2 7X4W2 MC7-1 2 7.0 4.0 0.0 - 2.0 2.0 1958 NOV. 1964 10.4 17.4 BEM M RLL
MC8-1 4 8X7W2 MC8-1 4 8.0 7.0 0.0 - 2.0 2.0 1958 NOV. 1964 9.4 15.6 BEM M RLL
MC8-1 3 8X5W2 MC8-1 3 8.0 5.0 0.0 - 2.0 2.0 1958 NOV. 1964 10.0 16.7 BEM M RLL
MC10-1 6 10X8W2 MC10-1 6 10.0 8.0 0.0 - 2.0 2.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC8-1 4 8X6W2 MC8-1 4 8.0 6.0 0.0 - 2.0 2.0 1958 NOV. 1964 9.7 16.2 BEM M RLL
MC7-1 6 7X7W2 MC7-1 6 7.0 7.0 0.0 - 2.0 2.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC6-1 5 6X5W4 MC6-1 5 6.0 5.0 0.0 - 4.0 4.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC7-2 5 7X5W4 MC7-2 5 7.0 5.0 2.08 - 4.0 4.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC7-2 6 7X7W4 MC7-2 6 7.0 7.0 2.08 - 4.0 4.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC8-2 5 8X4W4 MC8-2 5 8.0 4.0 2.08 - 4.0 4.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC8-2 6 8X6W4 MC8-2 6 8.0 6.0 2.08 - 4.0 4.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC9-2 6 9X5W4 MC9-2 6 9.0 5.0 2.08 - 4.0 4.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC10-2 2 10X9W4 MC10-2 2 10.0 9.0 2.08 - 4.0 4.0 1958 NOV. 1964 12.2 20.4 WEM M TL
MC6-1 4 6X3W4 MC6-1 4 6.0 3.0 0.0 - 4.0 4.0 1958 NOV. 1964 24.0 40.1 BEM M RLL
MC7-2 3 7X3W4 MC7-2 3 7.0 3.0 2.08 - 4.0 4.0 1958 NOV. 1964 26.9 44.9 TEM M RLL
MC7-2 4 7X3W4 MC7-2 4 7.0 3.0 2.08 - 4.0 4.0 1958 NOV. 1964 27.1 45.2 TEM M RLL
MC7-2 2 7X3W4 MC7-2 2 7.0 3.0 2.08 - 4.0 4.0 1958 NOV. 1964 28.5 47.5 BIC V RLL
MC8-2 5 8X6W4 MC8-2 5 8.0 6.0 2.08 - 4.0 4.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC9-2 2 9X9W4 MC9-2 2 9.0 9.0 2.08 - 4.0 4.0 1958 NOV. 1964 9.3 15.5 WEM M TL

TxDOT 0-5849 256

 APPENDIX TABLE 10. 1958 ERA 60KSI STEEL SAMPLE DETAILS AND CULV-5 LOAD RATINGS (CONT.)
NO. BARREL BARREL FILL MAX FILL ORIGINAL LAST CULV-5
ID SHEET OF LENGTH HEIGHT RANGE DEPTH DESIGN REVISION IR OR
CCS CFM LC
SPANS (FT) (FT) (FT) (FT) YEAR DATE (HS-) (HS-)
MC9-3 4 9X9W6 MC9-3 4 9.0 9.0 4.08 - 6.0 6.0 1958 NOV. 1964 7.3 12.1 WEM M TL
MC10-3 5 10X10W6 MC10-3 5 10.0 10.0 4.08 - 6.0 6.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC6-2 6 6X3W6 MC6-2 6 6.0 3.0 4.08 - 6.0 6.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC8-3 3 8X8W6 MC8-3 3 8.0 8.0 4.08 - 6.0 6.0 1958 NOV. 1964 22.3 37.1 WEM M TL
MC9-3 3 9X7W6 MC9-3 3 9.0 7.0 4.08 - 6.0 6.0 1958 NOV. 1964 35.6 59.5 BIC V RLL
MC5-2 3 5X4W6 MC5-2 3 5.0 4.0 4.08 - 6.0 6.0 1958 NOV. 1964 34.7 57.9 BEM M RLL
MC10-3 4 10X8W6 MC10-3 4 10.0 8.0 4.08 - 6.0 6.0 1958 NOV. 1964 28.2 47.0 WEM M TL
MC10-3 3 10X6W6 MC10-3 3 10.0 6.0 4.08 - 6.0 6.0 1958 NOV. 1964 37.0 61.8 BEM M RLL
MC8-3 2 8X6W6 MC8-3 2 8.0 6.0 4.08 - 6.0 6.0 1958 NOV. 1964 31.7 52.8 BIC V RLL
MC9-3 2 9X5W6 MC9-3 2 9.0 5.0 4.08 - 6.0 6.0 1958 NOV. 1964 35.4 59.1 BIC V RLL
MC5-2 2 5X2W6 MC5-2 2 5.0 2.0 4.08 - 6.0 6.0 1958 NOV. 1964 42.8 71.4 BEM M RLL
MC7-3 2 7X7W6 MC7-3 2 7.0 7.0 4.08 - 6.0 6.0 1958 NOV. 1964 26.4 44.1 BIC V RLL
MC7-3 4 7X4W6 MC7-3 4 7.0 4.0 4.08 - 6.0 6.0 1958 NOV. 1964 30.6 51.1 BIC V RLL
MC8-3 6 8X6W6 MC8-3 6 8.0 6.0 4.08 - 6.0 6.0 1958 NOV. 1964 0.0 0.0 0.0 0.0 0.0
MC9-3 3 9X9W6 MC9-3 3 9.0 9.0 4.08 - 6.0 6.0 1958 NOV. 1964 7.9 13.1 WEM M TL

TxDOT 0-5849 257

 APPENDIX TABLE 11. 1958 ERA 60KSI STEEL RISA-2D WITH SPRINGS LOAD RATINGS
RISA-2D SPG K=75PCI RISA-2D SPG K=150PCI RISA-2D SPG K=250PCI
ID IR OR IR OR IR OR
CCS CFM LC CCS CFM LC CCS CFM LC
(HS-) (HS-) (HS-) (HS-) (HS-) (HS-)
MC10-1 5 10X7W2 12.2 20.4 WTIC M TL 12.7 21.2 WTIC M RLL 12.4 20.7 WTEC M TL
MC10-1 6 10X9W2 13.3 22.2 WEM M TL 12.5 20.8 WEM M TL 11.8 19.7 WEM M TL
MC9-1 5 9X8W2 13.0 21.7 WTIC M RLL 13.6 22.7 WTIC M RLL 13.6 22.7 WTIC M RLL
MC9-1 4 9X6W2 11.9 19.9 WTIC M RLL 11.7 19.5 WTIC M RLL 11.5 19.2 WTIC M RLL
MC7-1 3 7X6W2 11.8 19.6 WTIC M RLL 12.3 20.5 WTIC M RLL 12.6 21.0 WTIC M RLL
MC7-1 2 7X4W2 12.8 21.4 WBEC M TL 14.3 23.8 WTEC M TL 14.0 23.4 WTEC M TL
MC8-1 4 8X7W2 12.0 20.0 WTIC M RLL 12.3 20.5 WTIC M RLL 12.1 20.2 WTIC M RLL
MC8-1 3 8X5W2 10.9 18.2 WTIC M RLL 10.6 17.8 WTIC M RLL 10.5 17.6 WTIC M RLL
MC10-1 6 10X8W2 12.7 21.2 WTIC M RLL 13.4 22.3 WTIC M RLL 13.6 22.7 WTIC M RLL
MC8-1 4 8X6W2 11.9 19.9 WTIC M RLL 11.6 19.4 WTIC M RLL 11.5 19.1 WTIC M RLL
MC7-1 6 7X7W2 5.9 9.8 TIC M RLL 6.5 10.9 TIC M RLL 7.0 11.6 TIC M RLL
MC6-1 5 6X5W4 27.6 46.0 TIC V RLL 28.0 46.8 TIC V RLL 28.4 47.5 TIC V RLL
MC7-2 5 7X5W4 26.5 44.3 TIC V RLL 26.9 44.9 TIC V RLL 27.2 45.4 TIC V RLL
MC7-2 6 7X7W4 22.6 37.7 BEC M TL 26.6 44.5 TEM M RLL 26.5 44.3 WTEC M TL
MC8-2 5 8X4W4 23.0 38.3 TEM M RLL 23.2 38.7 TEM M RLL 23.4 39.0 TEM M RLL
MC8-2 6 8X6W4 23.4 39.0 TEM M RLL 23.5 39.3 TEM M RLL 23.8 39.7 TEM M RLL
MC9-2 6 9X5W4 23.6 39.5 TEM M RLL 23.8 39.8 TEM M RLL 24.0 40.1 TEM M RLL
MC10-2 2 10X9W4 12.3 20.6 WEM M TL 12.0 20.0 WEM M TL 11.5 19.2 WEM M TL
MC6-1 4 6X3W4 23.9 39.9 WTIC M RLL 26.0 43.5 WTIC M RLL 27.8 46.4 WTIC M RLL
MC7-2 3 7X3W4 20.0 33.3 WTIC M RLL 21.1 35.2 WTIC M RLL 22.0 36.7 WTIC M RLL
MC7-2 4 7X3W4 23.2 38.8 WTIC M RLL 24.3 40.6 WTIC M TL 25.1 41.9 WTIC M TL
MC7-2 2 7X3W4 26.0 43.4 WTEC M TL 25.8 43.0 WTEC M TL 25.5 42.6 WTEC M TL
MC8-2 5 8X6W4 23.3 38.9 TEM M RLL 23.5 39.3 TEM M RLL 23.7 39.6 TEM M RLL
MC9-2 2 9X9W4 9.1 15.2 WEM M TL 8.9 14.9 WEM M TL 8.6 14.3 WEM M TL

TxDOT 0-5849 258

 APPENDIX TABLE 11. 1958 ERA 60KSI STEEL RISA-2D WITH SPRINGS LOAD RATINGS (CONT.)
RISA-2D SPG K=75PCI RISA-2D SPG K=150PCI RISA-2D SPG K=250PCI
ID IR OR IR OR IR OR
CCS CFM LC CCS CFM LC CCS CFM LC
(HS-) (HS-) (HS-) (HS-) (HS-) (HS-)
MC9-3 4 9X9W6 7.7 12.8 WEM M TL 7.1 11.8 WEM M TL 6.4 10.7 WEM M TL
MC10-3 5 10X10W6 -17.4 -29.0 WEM M TL -17.9 -29.8 WEM M TL -18.7 -31.1 WEM M TL
MC6-2 6 6X3W6 35.3 58.9 TIC V RLL 35.5 59.3 TIC V RLL 35.9 59.9 TIC V RLL
MC8-3 3 8X8W6 21.5 35.9 WEM M TL 20.8 34.7 WEM M TL 20.0 33.3 WEM M TL
MC9-3 3 9X7W6 38.4 64.0 TIC V RLL 37.6 62.8 WEM M TL 36.3 60.7 WEM M TL
MC5-2 3 5X4W6 37.3 62.3 BEM M RLL 38.9 64.9 TEM M RLL 39.1 65.3 TEM M RLL
MC10-3 4 10X8W6 31.7 52.9 WEM M TL 30.4 50.7 WEM M TL 29.1 48.5 WEM M TL
MC10-3 3 10X6W6 40.4 67.5 TIC V RLL 40.6 67.8 TIC V RLL 40.6 67.7 WTEC M TL
MC8-3 2 8X6W6 34.4 57.3 TIC V RLL 34.8 58.1 TIC V RLL 35.2 58.8 TIC V RLL
MC9-3 2 9X5W6 37.6 62.7 TIC V RLL 38.1 63.5 TIC V RLL 38.5 64.2 TIC V RLL
MC5-2 2 5X2W6 45.2 75.5 BEM M RLL 46.1 76.9 TEM M RLL 46.3 77.3 TEM M RLL
MC7-3 2 7X7W6 25.1 41.9 WEM M TL 24.9 41.6 WEM M TL 24.5 41.0 WEM M TL
MC7-3 4 7X4W6 33.5 55.9 TIC V RLL 33.8 56.4 TIC V RLL 34.1 56.9 TIC V RLL
MC8-3 6 8X6W6 36.4 60.8 TIC V RLL 36.8 61.4 TIC V RLL 37.2 62.1 TIC V RLL
MC9-3 3 9X9W6 7.7 12.8 WEM M TL 7.1 11.8 WEM M TL 6.4 10.6 WEM M TL

TxDOT 0-5849 259

 APPENDIX TABLE 12. 1958 ERA 60KSI STEEL RISA-2D WITH LEFE LOAD RATINGS
RISA-2D LEFE E=8KSI RISA-2D LEFE E=20KSI RISA-2D LEFE E=36KSI
ID IR OR IR OR IR OR
CCS CFM CCS CFM CCS CFM
(HS-) (HS-) (HS-) (HS-) (HS-) (HS-)
MC10-1 5 10X7W2 11.5 19.2 WTIC M 17.5 29.2 WTIC M 22.8 38.1 WTIC M
MC10-1 6 10X9W2 10.3 17.1 WTIC M 17.5 29.2 WTIC M 23.3 38.8 WTIC M
MC9-1 5 9X8W2 11.0 18.4 WTIC M 18.1 30.2 WTIC M 23.1 38.6 TIC V
MC9-1 4 9X6W2 12.1 20.1 WTIC M 17.7 29.6 WTIC M 23.0 38.3 WTIC M
MC7-1 3 7X6W2 14.9 24.9 WTIC M 21.4 35.7 TIC V 23.4 39.1 TIC V
MC7-1 2 7X4W2 18.0 30.0 WTIC M 21.3 35.6 TIC V 23.3 38.9 TIC V
MC8-1 4 8X7W2 12.3 20.6 WTIC M 19.9 33.1 WTIC M 22.3 37.3 TIC V
MC8-1 3 8X5W2 13.2 22.1 WTIC M 18.4 30.7 WTIC M 22.2 37.0 TIC V
MC10-1 6 10X8W2 10.9 18.2 WTIC M 17.2 28.7 WTIC M 22.4 37.3 WTIC M
MC8-1 4 8X6W2 12.9 21.6 WTIC M 19.4 32.4 WTIC M 22.3 37.1 TIC V
MC7-1 6 7X7W2 6.8 11.3 TIC M 12.8 21.4 TIC M 17.7 29.5 TIC M
MC6-1 5 6X5W4 34.2 57.0 TIC V 41.8 69.8 TIC V 48.3 80.6 TIC V
MC7-2 5 7X5W4 30.6 51.0 WBEC M 43.0 71.7 TIC V 49.9 83.2 TIC V
MC7-2 6 7X7W4 13.1 21.9 BEC M 39.0 65.1 WBEC M 49.8 83.2 TIC V
MC8-2 5 8X4W4 30.5 50.8 WTIC M 44.4 74.2 TIC V 51.6 86.1 TIC V
MC8-2 6 8X6W4 27.2 45.3 WBEC M 44.4 74.2 TIC V 51.7 86.4 TIC V
MC9-2 6 9X5W4 28.0 46.7 WTIC M 47.8 79.7 WTIC M 56.2 93.7 TIC V
MC10-2 2 10X9W4 19.2 32.1 WBEC M 47.3 78.9 TIC V 54.9 91.7 TIC V
MC6-1 4 6X3W4 35.3 58.9 TIC V 42.1 70.3 TIC V 48.0 80.1 TIC V
MC7-2 3 7X3W4 34.1 57.0 WTIC M 42.9 71.5 TIC V 49.2 82.1 TIC V
MC7-2 4 7X3W4 33.7 56.2 WTIC M 42.7 71.2 TIC V 49.3 82.3 TIC V
MC7-2 2 7X3W4 34.1 56.9 TIC V 40.8 68.1 TIC V 47.6 79.4 TIC V
MC8-2 5 8X6W4 27.0 45.0 WBEC M 44.0 73.4 TIC V 51.1 85.3 TIC V
MC9-2 2 9X9W4 11.9 19.8 WBEC M 39.7 66.3 WBEC M 54.4 90.9 TIC V

TxDOT 0-5849 260

 APPENDIX TABLE 12. 1958 ERA 60KSI STEEL RISA-2D WITH LEFE LOAD RATINGS (CONT.)
RISA-2D LEFE E=8KSI RISA-2D LEFE E=20KSI RISA-2D LEFE E=36KSI
ID IR OR IR OR IR OR
CCS CFM CCS CFM CCS CFM
(HS-) (HS-) (HS-) (HS-) (HS-) (HS-)
MC9-3 4 9X9W6 15.4 25.7 WBEC M 58.1 97.0 WBEC M 69.5 116.0 TIC V
MC10-3 5 10X10W6 21.1 35.3 WTIC M 63.2 105.5 TIC V 73.4 122.6 TIC V
MC6-2 6 6X3W6 40.8 68.0 TIC V 50.1 83.6 TIC V 58.5 97.6 TIC V
MC8-3 3 8X8W6 24.0 40.1 WBEC M 49.9 83.3 BEC V 63.5 106.0 TIC V
MC9-3 3 9X7W6 17.4 29.1 WBEC M 54.0 90.2 WBEC M 70.6 117.9 TIC V
MC5-2 3 5X4W6 39.4 65.7 WBEC M 56.0 93.5 TIC V 64.9 108.3 TIC V
MC10-3 4 10X8W6 20.7 34.5 WBEC M 63.3 105.7 WBEC M 73.2 122.1 TIC V
MC10-3 3 10X6W6 31.1 51.9 WBEC M 64.3 107.4 TIC V 73.9 123.3 TIC V
MC8-3 2 8X6W6 36.5 60.9 WBEC M 52.2 87.2 TIC V 62.3 104.0 TIC V
MC9-3 2 9X5W6 38.1 63.6 WBEC M 57.1 95.3 TIC V 67.5 112.6 TIC V
MC5-2 2 5X2W6 46.3 77.2 TEM M 54.0 90.1 TIC V 61.7 103.0 TIC V
MC7-3 2 7X7W6 23.3 39.0 WBEC M 44.0 73.5 BEC V 54.8 91.5 TIC V
MC7-3 4 7X4W6 39.7 66.2 TIC V 49.5 82.7 TIC V 57.8 96.5 TIC V
MC8-3 6 8X6W6 35.5 59.3 WBEC M 56.2 93.8 TIC V 65.6 109.5 TIC V
MC9-3 3 9X9W6 15.4 25.7 WBEC M 56.4 94.1 WBEC M 69.4 115.8 TIC V

TxDOT 0-5849 261

 APPENDIX E
2003 ERA LOAD RATINGS

TxDOT 0-5849 262

 APPENDIX TABLE 13. 2003 ERA SAMPLE DETAILS AND CULV-5 LOAD RATINGS
NO. BARREL BARREL FILL MAX FILL ORIGINAL LAST CULV-5
ID SHEET OF LENGTH HEIGHT RANGE DEPTH DESIGN REVISION IR OR
CCS CFM LC
SPANS (FT) (FT) (FT) (FT) YEAR DATE (HS-) (HS-)
MC-10-7 3 10X8W7 MC-10-7 3 10.0 8.0 0.0 - 7.0 7.0 2003 DEC. 2003 25.6 42.7 WBEC M TL
MC-10-7 4 10X6W7 MC-10-7 4 10.0 6.0 0.0 - 7.0 7.0 2003 DEC. 2003 38.2 63.8 WEM M TL
MC-9-10 2 9X6W10 MC-9-10 2 9.0 6.0 0.0 - 10.0 10.0 2003 DEC. 2003 36.2 60.4 BIC V RLL
MC-7-10 5 7X7W10 MC-7-10 5 7.0 7.0 0.0 - 10.0 10.0 2003 DEC. 2003 0.0 0.0 0.0 0.0 0.0
MC-7-10 6 7X5W10 MC-7-10 6 7.0 5.0 0.0 - 10.0 10.0 2003 DEC. 2003 0.0 0.0 0.0 0.0 0.0
MC-10-10 2 10X5W10 MC-10-10 2 10.0 5.0 2.0 - 10.0 10.0 2003 DEC. 2003 45.6 76.2 BIC V RLL
MC-7-16 3 7X6W16 MC-7-16 3 7.0 6.0 2.0 - 16.0 16.0 2003 DEC. 2003 52.2 87.2 BIC V RLL
MC-7-16 4 7X4W16 MC-7-16 4 7.0 4.0 2.0 - 16.0 16.0 2003 DEC. 2003 55.2 92.1 BIC V RLL
MC-5-20 2 5X3W20 MC-5-20 2 5.0 3.0 0.0 - 20.0 20.0 2003 DEC. 2003 69.0 115.2 BIC V RLL
MC-8-20 6 8X6W20 MC-8-20 6 8.0 6.0 2.0 - 20.0 20.0 2003 DEC. 2003 0.0 0.0 0.0 0.0 0.0
MC-8-20 5 8X8W20 MC-8-20 5 8.0 8.0 2.0 - 20.0 20.0 2003 DEC. 2003 0.0 0.0 0.0 0.0 0.0
MC-10-20 5 10X10W20 MC-10-20 5 10.0 10.0 2.0 - 20.0 20.0 2003 DEC. 2003 0.0 0.0 0.0 0.0 0.0
MC-10-20 6 10X8W20 MC-10-20 6 10.0 8.0 2.0 - 20.0 20.0 2003 DEC. 2003 0.0 0.0 0.0 0.0 0.0
MC-7-20 2 7X4W20 MC-7-20 2 7.0 4.0 2.0 - 20.0 20.0 2003 DEC. 2003 51.8 86.5 BIC V RLL
MC-8-23 3 8X7W23 MC-8-23 3 8.0 7.0 2.0 - 23.0 23.0 2003 DEC. 2003 39.8 66.5 WEM M TL
MC-4-23 2 4X4W23 MC-4-23 2 4.0 4.0 0.0 - 23.0 23.0 2003 DEC. 2003 117.5 196.1 WEM M TL
MC-9-23 5 9X9W23 MC-9-23 5 9.0 9.0 2.0 - 23.0 23.0 2003 DEC. 2003 0.0 0.0 0.0 0.0 0.0
MC-10-23 3 10X8W23 MC-10-23 3 10.0 8.0 2.0 - 23.0 23.0 2003 DEC. 2003 48.6 81.2 BIC V TL
MC-9-23 6 9X7W23 MC-9-23 6 9.0 7.0 2.0 - 23.0 23.0 2003 DEC. 2003 0.0 0.0 0.0 0.0 0.0
MC-8-23 4 8X5W23 MC-8-23 4 8.0 5.0 2.0 - 23.0 23.0 2003 DEC. 2003 78.9 131.7 BIC V RLL
MC-10-23 4 10X6W23 MC-10-23 4 10.0 6.0 2.0 - 23.0 23.0 2003 DEC. 2003 51.5 86.0 BIC V TL

TxDOT 0-5849 263

 APPENDIX TABLE 14. 2003 ERA RISA-2D WITH SPRINGS LOAD RATINGS
RISA-2D SPG K=75PCI RISA-2D SPG K=150PCI RISA-2D SPG K=250PCI
ID IR OR IR OR IR OR
CCS CFM LC CCS CFM LC CCS CFM LC
(HS-) (HS-) (HS-) (HS-) (HS-) (HS-)
MC-10-7 3 10X8W7 25.5 42.5 WEM M TL 23.6 39.5 WEM M TL 21.9 36.5 WEM M TL
MC-10-7 4 10X6W7 31.3 52.2 TIC V RLL 31.2 52.1 TIC V RLL 31.3 52.2 TIC V RLL
MC-9-10 2 9X6W10 33.7 56.2 TIC V RLL 34.1 57.0 TIC V RLL 34.7 57.8 TIC V RLL
MC-7-10 5 7X7W10 24.3 40.6 WEM M TL 23.5 39.2 WEM M TL 22.5 37.6 WEM M TL
MC-7-10 6 7X5W10 52.6 87.7 TIC V RLL 52.3 87.2 TIC V RLL 52.3 87.4 TIC V RLL
MC-10-10 2 10X5W10 42.9 71.7 TIC V RLL 43.2 72.2 TIC V RLL 43.7 72.9 TIC V RLL
MC-7-16 3 7X6W16 55.1 92.0 WEM M TL 53.8 89.9 WEM M TL 52.2 87.1 WEM M TL
MC-7-16 4 7X4W16 58.3 97.4 TIC V RLL 58.3 97.3 TIC V RLL 58.7 98.0 TIC V RLL
MC-5-20 2 5X3W20 63.6 106.2 TIC V RLL 63.9 106.7 TIC V RLL 64.5 107.7 TIC V RLL
MC-8-20 6 8X6W20 63.1 105.3 TIC V RLL 63.3 105.7 TIC V RLL 63.9 106.7 TIC V RLL
MC-8-20 5 8X8W20 -18.2 -30.4 WEM M TL -19.1 -32.0 WEM M TL -20.3 -33.9 WEM M TL
MC-10-20 5 10X10W20 -30.1 -50.3 WEM M TL -31.4 -52.5 WEM M TL -33.1 -55.2 WEM M TL
MC-10-20 6 10X8W20 51.7 86.3 WEM M TL 49.1 82.0 WEM M TL 46.9 78.2 WEM M TL
MC-7-20 2 7X4W20 55.9 93.3 TIC V RLL 56.7 94.7 TIC V RLL 57.3 95.6 TIC V RLL
MC-8-23 3 8X7W23 36.8 61.5 WEM M TL 35.6 59.4 WEM M TL 34.1 56.9 WEM M TL
MC-4-23 2 4X4W23 21.1 35.2 BIC V RLL 22.6 37.6 BIC V RLL 24.2 40.4 BIC V RLL
MC-9-23 5 9X9W23 -21.6 -36.0 WEM M TL -22.7 -37.8 WEM M TL -24.0 -40.1 WEM M TL
MC-10-23 3 10X8W23 63.2 105.5 WEM M TL 60.9 101.7 WEM M TL 58.3 97.4 WEM M TL
MC-9-23 6 9X7W23 78.3 130.7 TIC V RLL 78.0 130.3 TIC V RLL 77.2 128.9 WEM M TL
MC-8-23 4 8X5W23 81.5 136.0 TIC V RLL 82.4 137.6 TIC V RLL 83.0 138.6 TIC V RLL
MC-10-23 4 10X6W23 62.8 104.8 TIC V RLL 64.0 106.8 TIC V RLL 65.3 109.0 TIC V RLL

TxDOT 0-5849 264

 APPENDIX TABLE 15. 2003 ERA RISA-2D WITH LEFE LOAD RATINGS
RISA-2D LEFE E=8KSI RISA-2D LEFE E=20KSI RISA-2D LEFE E=36KSI
ID IR OR IR OR IR OR
CCS CFM CCS CFM CCS CFM
(HS-) (HS-) (HS-) (HS-) (HS-) (HS-)
MC-10-7 3 10X8W7 12.1 20.2 WBEC M 57.9 96.6 TIC V 68.9 115.0 TIC V
MC-10-7 4 10X6W7 41.3 68.9 WBEC M 58.7 98.0 TIC V 69.4 115.9 TIC V
MC-9-10 2 9X6W10 30.4 50.7 BIC V 48.0 80.2 BIC V 64.5 107.6 BIC V
MC-7-10 5 7X7W10 13.7 22.8 WBEC M 51.6 86.2 BEC V 68.4 114.2 BIC V
MC-7-10 6 7X5W10 44.3 74.0 TIC V 61.4 102.5 BIC V 73.4 122.6 BIC V
MC-10-10 2 10X5W10 40.9 68.3 BIC V 58.1 96.9 BIC V 74.2 123.8 BIC V
MC-7-16 3 7X6W16 35.4 59.1 WBEC M 84.6 141.3 WBEC M 138.1 230.5 WBEC M
MC-7-16 4 7X4W16 51.7 86.3 WTIC M 130.9 218.6 WTIC M 211.9 353.7 BIC V
MC-5-20 2 5X3W20 41.9 70.0 TIC V 79.7 133.1 TIC V 127.6 213.0 TIC V
MC-8-20 6 8X6W20 18.7 31.2 WTIC M 80.0 133.6 WBEC M 144.3 240.9 WBEC M
MC-8-20 5 8X8W20 8.5 14.1 WTIC M 91.4 152.6 WBEC M 166.2 277.5 WBEC M
MC-10-20 5 10X10W20 -24.3 -40.6 WTIC M 71.7 119.6 WTIC M 173.5 289.6 WTIC M
MC-10-20 6 10X8W20 -13.2 -22.0 WTIC M 87.6 146.3 WTIC M 190.4 317.8 WTIC M
MC-7-20 2 7X4W20 57.6 96.1 TIC V 118.4 197.7 TIC V 192.0 320.5 BIC V
MC-8-23 3 8X7W23 0.2 0.4 WBEC M 49.8 83.1 WBEC M 105.7 176.4 WBEC M
MC-4-23 2 4X4W23 49.4 82.4 TIC V 78.1 130.3 TIC V 113.9 190.1 TIC V
MC-9-23 5 9X9W23 -16.5 -27.6 WTIC M 70.8 118.2 WTIC M 162.4 271.1 WTIC M
MC-10-23 3 10X8W23 11.1 18.5 WTIC M 90.3 150.7 WTIC M 176.8 295.1 WTIC M
MC-9-23 6 9X7W23 -8.8 -14.7 WTIC M 89.6 149.6 WTIC M 189.2 315.9 WTIC M
MC-8-23 4 8X5W23 30.1 50.3 WTIC M 117.6 196.4 WTIC M 201.0 335.6 WBEC M
MC-10-23 4 10X6W23 3.1 5.2 WTIC M 106.1 177.2 WTIC M 206.9 345.4 WTIC M

TxDOT 0-5849 265

 APPENDIX F
SOIL BORING LOGS

TxDOT 0-5849 266

 BORING NUMBER B-1
Texas Tech University - Civil Engineering PAGE 1 OF 1

CLIENT Texas Department of Transportation PROJECT NAME Culvert 0052-07-010, US Hwy 84

PROJECT NUMBER 0-5849 (culvert loading) PROJECT LOCATION Shallowater, Lubbock County, TX
DATE STARTED 5/12/09 COMPLETED 5/12/09 GROUND ELEVATION HOLE SIZE 4 inches
DRILLING CONTRACTOR W.E.S.T Drilling GROUND WATER LEVELS:
DRILLING METHOD Hollow Stem Auger AT TIME OF DRILLING ---
LOGGED BY PWJ CHECKED BY WDL AT END OF DRILLING ---
NOTES AFTER DRILLING ---
SAMPLE TYPE

RECOVERY %

(N VALUE)

GRAPHIC
COUNTS

U.S.C.S.
DEPTH

BLOW

LOG
(ft)

REMARKS MATERIAL DESCRIPTION

0.0
(SC) FILL: CLAYEY SAND,reddish tan, w/ calcareus inclusions

ST 44

ST 33 PP = 4.5+ tsf SC
2.5

5-3-4
SPT 0
(7)
GENERAL BH / TP / WELL - GINT STD US LAB.GDT - 9/22/09 15:14 - E:\TXDOT CULVERT\CULVERT BORING LOGS\SWUS84.GPJ

4.5
(SC) FILL: CLAYEY SAND, brown, w/ calcareous gravel, asphalt
5.0
ST 83 PP = 3.5 tsf

ST 75 PP = 4.5+ tsf
(-)200 = 27%

ST 75
7.5
SC

ST 75 PP = 4.5+ tsf

3-3-2
SPT 100 PP = 4.5+ tsf
(5)

10.0
PP = 4.5+ tsf
10.5
MC = 14%
ST 78 LL = 34 (SC) CLAYEY SAND: reddish tan, silty
PL = 12
(-)200 = 34% SC

12.0
ST 56 (SC) CLAYEY SAND: light grey
12.5

SC
PP = 4.5+ tsf
12-23-30
SPT 100 MC = 14%
(53)
(-)200 = 35%
14.5
Bottom of borehole at 14.5 feet.
 BORING NUMBER B-2
Texas Tech University - Civil Engineering PAGE 1 OF 1

CLIENT Texas Department of Transportation PROJECT NAME Culvert 0052-07-010, US Hwy 84

PROJECT NUMBER 0-5849 (culvert loading) PROJECT LOCATION Shallowater, Lubbock County, TX
DATE STARTED 5/12/09 COMPLETED 5/12/09 GROUND ELEVATION HOLE SIZE 4 inches
DRILLING CONTRACTOR W.E.S.T Drilling GROUND WATER LEVELS:
DRILLING METHOD Hollow Stem Auger AT TIME OF DRILLING ---
LOGGED BY PWJ CHECKED BY WDL AT END OF DRILLING ---
NOTES AFTER DRILLING ---
SAMPLE TYPE

RECOVERY %

(N VALUE)

GRAPHIC
COUNTS

U.S.C.S.
DEPTH

TESTS
BLOW

LOG
(ft)

AND MATERIAL DESCRIPTION

REMARKS

0.0
(SC) FILL: CLAYEY SAND; reddish brown, w/ calcareous gravel pieces
PP = 4.5+ tsf
ST 33 MC = 6%
(-200) = 31%

MC = 7%
LL = 29
ST 33 SC
2.5 PL = 10
(-200) = 34%

6-6-3
SPT 100 PP = 4.5+ tsf
(9)
GENERAL BH / TP / WELL - GINT STD US LAB.GDT - 9/22/09 15:14 - E:\TXDOT CULVERT\CULVERT BORING LOGS\SWUS84.GPJ

4.5
(SC) FILL: CLAYEY SAND ; Reddish- brown, w/ calcareous gravel and asphalt
5.0 pieces
ST 44

PP = 4.5+ tsf
MC = 13%
ST 61 LL = 27
PL = 11
(-200) = 34%
7.5
SC
ST 50 PP = 4.5+ tsf

3-5-5
SPT 100 PP = 3.5 tsf
(10)

10.0

ST 44 PP = 3.25 tsf
11.0
(SC) CLAYEY SAND: brown

PP = 2.75 tsf
ST 56 MC = 12%
12.5
(-200) = 24%
SC

1-2-1
SPT 100
(3)

14.5
Bottom of borehole at 14.5 feet.
 BORING NUMBER B-3
Texas Tech University - Civil Engineering PAGE 1 OF 1

CLIENT Texas Department of Transportation PROJECT NAME Culvert 0052-07-010, US Hwy 84

PROJECT NUMBER 0-5849 (culvert loading) PROJECT LOCATION Shallowater, Lubbock County, TX
DATE STARTED 5/12/09 COMPLETED 5/12/09 GROUND ELEVATION HOLE SIZE 4 inches
DRILLING CONTRACTOR W.E.S.T Drilling GROUND WATER LEVELS:
DRILLING METHOD Hollow Stem Auger AT TIME OF DRILLING ---
LOGGED BY PWJ CHECKED BY WDL AT END OF DRILLING ---
NOTES AFTER DRILLING ---
SAMPLE TYPE

(N VALUE)

GRAPHIC
COUNTS

U.S.C.S.
DEPTH

BLOW

LOG
(ft)

MATERIAL DESCRIPTION

0.0
(SC) FILL: CLAYEY SAND; reddish brown, w/ calcareous gravel pieces

DCP 20-30/0"

SC
2.5

TCP 8-6
GENERAL BH / TP / WELL - GINT STD US LAB.GDT - 9/22/09 15:14 - E:\TXDOT CULVERT\CULVERT BORING LOGS\SWUS84.GPJ

4.5
(SC) FILL: CLAYEY SAND ; Reddish- brown, w/ calcareous gravel and asphalt pieces
5.0

DCP 10-22-
20/0"

7.5
SC

TCP 9-10

10.0

11.0
(SC) CLAYEY SAND: brown

DCP 7-12-11/0"

12.5
SC

TCP 6-5

14.5
Bottom of borehole at 14.5 feet.
 BORING NUMBER B-1
Texas Tech University - Civil Engineering PAGE 1 OF 1

CLIENT Texas Department of Transportation PROJECT NAME Culvert 0439-05-025 SH194

PROJECT NUMBER 0-5849 ( culvert loading) PROJECT LOCATION Plainview, Hale County, TX
DATE STARTED 5/13/09 COMPLETED 5/13/09 GROUND ELEVATION HOLE SIZE 4 inches
DRILLING CONTRACTOR W.E.S.T Drilling GROUND WATER LEVELS:
DRILLING METHOD Hollow Stem Auger AT TIME OF DRILLING ---
LOGGED BY PWJ CHECKED BY WDL AT END OF DRILLING ---
NOTES AFTER DRILLING ---
SAMPLE TYPE

RECOVERY %

(N VALUE)

GRAPHIC
COUNTS

U.S.C.S.
DEPTH

BLOW

LOG
(ft)

REMARKS MATERIAL DESCRIPTION

0.0
NR PAVEMENT: 4" asphalt over base

1.0
(CL-CH) FILL: LEAN/ FAT SANDY CLAY, Dark brown

ST 83
2.5

3-5-7
SPT 100 PP = 3.75 tsf
(12)
GENERAL BH / TP / WELL - GINT STD US LAB.GDT - 9/22/09 15:27 - E:\TXDOT CULVERT\CULVERT BORING LOGS\PVSH194.GPJ

5.0
ST 61 PP = 1.0 tsf
CL-
CH

PP = 2.0 tsf
ST 78
MC = 20%

7.5

ST 75 PP = 1.0 tsf

3-2-3
SPT 100
(5)

10.0

10.5
ST 78 PP = 1.5 tsf (SP) SAND, Brown& Tan, silty

SP

PP = 1.75 tsf
ST 56 MC = 15%
12.5 12.5
(-200) = 31%
(GC) CLAYEY GRAVEL: light tan, calcareous
GC
13.5
6-8-8 (SW) SAND: light tan, w/ calcareous gravel
SPT 100
(16)
SW
14.5
Bottom of borehole at 14.5 feet.
 BORING NUMBER B-2
Texas Tech University - Civil Engineering PAGE 1 OF 1

CLIENT Texas Department of Transportation PROJECT NAME Culvert 0439-05-025 SH194

PROJECT NUMBER 0-5849 ( culvert loading) PROJECT LOCATION Plainview, Hale County, TX
DATE STARTED 5/13/09 COMPLETED 5/13/09 GROUND ELEVATION HOLE SIZE 4 inches
DRILLING CONTRACTOR W.E.S.T Drilling GROUND WATER LEVELS:
DRILLING METHOD Hollow Stem Auger AT TIME OF DRILLING ---
LOGGED BY PWJ CHECKED BY WDL AT END OF DRILLING ---
NOTES AFTER DRILLING ---
SAMPLE TYPE

RECOVERY %

(N VALUE)

GRAPHIC
COUNTS

U.S.C.S.
DEPTH

BLOW

LOG
(ft)

REMARKS MATERIAL DESCRIPTION

0.0
NR PAVEMENT: 4" asphalt over base

1.0
(CL-CH) FILL: LEAN/FAT SANDY CLAY; Dark Brown, w/ occasional sand
seams, occasional calcareous gravel

ST 67 PP = 4.5+ tsf
2.5

5-5-9
SPT 100
(14)
GENERAL BH / TP / WELL - GINT STD US LAB.GDT - 9/22/09 15:27 - E:\TXDOT CULVERT\CULVERT BORING LOGS\PVSH194.GPJ

5.0
ST 61 PP = 4.5+ tsf
CL-
CH
PP = 4.5+ tsf
MC = 16%
ST 61 LL = 42
PL = 13
(-200) = 60%
7.5

ST 83

3-4-4
SPT 100 PP = 4.5+ tsf
(8)

10.0
PP = 3.5 tsf
10.5
MC = 17%
ST 89 LL = 34 (CL) SANDY LEAN CLAY: Brown, silty
PL = 12
(-200) = 66%

ST 78 PP = 4.5+ tsf CL
12.5

8-12-15
SPT 100
(27) 14.0
SW (SW) SAND : Tan
14.5
Bottom of borehole at 14.5 feet.
 BORING NUMBER B-3
Texas Tech University - Civil Engineering PAGE 1 OF 1

CLIENT Texas Department of Transportation PROJECT NAME Culvert 0439-05-025 SH194

PROJECT NUMBER 0-5849 ( culvert loading) PROJECT LOCATION Plainview, Hale County, TX
DATE STARTED 5/13/09 COMPLETED 5/13/09 GROUND ELEVATION HOLE SIZE 4 inches
DRILLING CONTRACTOR W.E.S.T Drilling GROUND WATER LEVELS:
DRILLING METHOD Hollow Stem Auger AT TIME OF DRILLING ---
LOGGED BY PWJ CHECKED BY WDL AT END OF DRILLING ---
NOTES AFTER DRILLING ---
SAMPLE TYPE

(N VALUE)

GRAPHIC
COUNTS

U.S.C.S.
DEPTH

BLOW

LOG
(ft)

MATERIAL DESCRIPTION

0.0
PAVEMENT: 4" asphalt over base

1.0
(CL-CH) FILL: LEAN/FAT SANDY CLAY; Dark Brown, w/ occasional sand seams, occasional calcareous
gravel
DCP 4-4-8/0"

2.5

TCP 9-9
GENERAL BH / TP / WELL - GINT STD US LAB.GDT - 9/22/09 15:27 - E:\TXDOT CULVERT\CULVERT BORING LOGS\PVSH194.GPJ

5.0

CL-
CH

DCP 9-24-27/0"

7.5

TCP 7-6

10.0

10.5
(CL) SANDY CLAY: Brown, silty

DCP 6-21-25/0"

CL
12.5

TCP 32-4
14.0
SW (SW) SAND : Tan
14.5
Bottom of borehole at 14.5 feet.
 BORING NUMBER B-1
Texas Tech University - Civil Engineering PAGE 1 OF 1

CLIENT Texas Department of Transportation PROJECT NAME Culvert 1345-01-002 FM 1318

PROJECT NUMBER 0-5849 (culvert loading) PROJECT LOCATION Tulia, Swisher county, TX
DATE STARTED 5/14/09 COMPLETED 5/14/09 GROUND ELEVATION HOLE SIZE 4 inches
DRILLING CONTRACTOR W.E.S.T Drilling GROUND WATER LEVELS:
DRILLING METHOD AT TIME OF DRILLING 9.50 ft
LOGGED BY TAW CHECKED BY WDL AT END OF DRILLING ---
NOTES AFTER DRILLING ---
SAMPLE TYPE

RECOVERY %

(N VALUE)

GRAPHIC
COUNTS

U.S.C.S.
DEPTH

BLOW

LOG
(ft)

REMARKS MATERIAL DESCRIPTION

0.0
NR PAVEMENT: Multiple seal coats over gravel base

1.0
(SC/CL) FILL: CLAYEY SAND/ SANDY CLAY; brown, w/ gravel

ST 33 PP = 4.5+ tsf
2.5

3-2-2
SPT 100
(4)
GENERAL BH / TP / WELL - GINT STD US LAB.GDT - 9/22/09 15:56 - E:\TXDOT CULVERT\CULVERT BORING LOGS\TUFM1318.GPJ

SC/CL
5.0
ST 44 PP = 4.5+ tsf

MC = 15%
LL = 29
ST 39
PL = 12
(-)200 = 38%
7.5

8.0
ST 100
(SC) CLAYEY SAND: Brown
SC
9.0
1-4-4 (SP) POORLY GRADED (FINE) SAND: Tan, silty
SPT 100
(8)

10.0

ST 92

ST 100

SP
MC = 7%
DS 100
12.5 (-)200 = 11%

7-11-11
SPT 100
(22)

14.5
Bottom of borehole at 14.5 feet.
 BORING NUMBER B-2
Texas Tech University - Civil Engineering PAGE 1 OF 1

CLIENT Texas Department of Transportation PROJECT NAME Culvert 1345-01-002 FM 1318

PROJECT NUMBER 0-5849 (culvert loading) PROJECT LOCATION Tulia, Swisher county, TX
DATE STARTED 5/14/09 COMPLETED 5/14/09 GROUND ELEVATION HOLE SIZE 4 inches
DRILLING CONTRACTOR W.E.S.T Drilling GROUND WATER LEVELS:
DRILLING METHOD AT TIME OF DRILLING 9.00 ft
LOGGED BY TAW CHECKED BY WDL AT END OF DRILLING ---
NOTES AFTER DRILLING ---
SAMPLE TYPE

RECOVERY %

(N VALUE)

GRAPHIC
COUNTS

U.S.C.S.
DEPTH

BLOW

LOG
(ft)

REMARKS MATERIAL DESCRIPTION

0.0
NR PAVEMENT: Multiple seal coats over gravel base

1.0
(SC/CL) FILL: CLAYEY SAND/ SANDY CLAY, Brown, w/ gravel

ST 67
2.5

2-1-2
SPT 100 PP = 4.5+ tsf
(3)
SC/CL
GENERAL BH / TP / WELL - GINT STD US LAB.GDT - 9/22/09 15:56 - E:\TXDOT CULVERT\CULVERT BORING LOGS\TUFM1318.GPJ

PP = 0.5 tsf
5.0
MC = 16%
ST 56 LL = 31
PL = 11
(-)200 = 36%

ST 67 PP = 1.5 tsf
7.0
(SC) CLAYEY SAND; Grayish tan
7.5
SC
8.0
ST 75
(SW) SAND: Tan, fine

PP = 0.5 tsf
3-4-8
SPT 100 MC = 9%
(12)
(-)200 = 11%
SW
10.0
ST 100

DS 100
11.5
(GW/SW) GRAVEL& SAND: Tan

MC = 8%
DS 100
12.5 (-)200 = 16%

GW/SW

23-16-13
SPT 100
(29)

14.5
Bottom of borehole at 14.5 feet.
 BORING NUMBER B-3
Texas Tech University - Civil Engineering PAGE 1 OF 1

CLIENT Texas Department of Transportation PROJECT NAME Culvert 1345-01-002 FM 1318

PROJECT NUMBER 0-5849 (culvert loading) PROJECT LOCATION Tulia, Swisher county, TX
DATE STARTED 5/14/09 COMPLETED 5/14/09 GROUND ELEVATION HOLE SIZE 4 inches
DRILLING CONTRACTOR W.E.S.T Drilling GROUND WATER LEVELS:
DRILLING METHOD AT TIME OF DRILLING ---
LOGGED BY TAW CHECKED BY WDL AT END OF DRILLING ---
NOTES AFTER DRILLING ---
SAMPLE TYPE

(N VALUE)

GRAPHIC
COUNTS

U.S.C.S.
DEPTH

BLOW

LOG
(ft)

MATERIAL DESCRIPTION

0.0
PAVEMENT: Multiple seal coats over gravel base

1.0
(SC/CL) FILL:CLAYEY SAND/ SANDY CLAY, Brown, w/ gravel

DCP 8-21-15/0"

2.5

SC/CL
GENERAL BH / TP / WELL - GINT STD US LAB.GDT - 9/22/09 15:56 - E:\TXDOT CULVERT\CULVERT BORING LOGS\TUFM1318.GPJ

5.0
TCP 2-4

DCP 7-11-10/0"
7.0
(SC) CLAYEY SAND; Grayish tan
7.5
SC
8.0
(SW) SAND: Tan, fine

TCP 2-11

SW
10.0

11.5
DCP 14-30/0" (GW/SW) GRAVEL& SAND: Tan

12.5

GW/SW

TCP 41-50

14.5
Bottom of borehole at 14.5 feet.
 APPENDIX G
FWD DATA

TxDOT 0-5849 288

 US84 EB Approach.asc
----------------------------------------------------------------------------------------------------------------------------------------
TTI MODULUS ANALYSIS SYSTEM (SUMMARY REPORT) (Version 6.0)
----------------------------------------------------------------------------------------------------------------------------------------
District:5 (Lubbock) MODULI RANGE(psi)
County :152 (LUBBOCK) Thickness(in) Minimum Maximum Poisson Ratio Values
Highway/Road: US0084 Pavement: 6.00 50,000 300,000 H1: v = 0.38
Base: 10.00 10,000 1,000,000 H2: v = 0.35
Subbase: 0.00 H3: v = 0.00
Subgrade: 67.24(by DB) 15,000 H4: v = 0.40
----------------------------------------------------------------------------------------------------------------------------------------
Load Measured Deflection (mils): Calculated Moduli values (ksi): Absolute Dpth to
Station (lbs) R1 R2 R3 R4 R5 R6 R7 SURF(E1) BASE(E2) SUBB(E3) SUBG(E4) ERR/Sens Bedrock
----------------------------------------------------------------------------------------------------------------------------------------
0.000 8,933 8.73 5.09 3.43 2.44 1.60 1.39 1.11 112.8 729.2 0.0 11.3 8.85 86.9
0.000 8,901 8.64 5.06 3.42 2.43 1.58 1.39 1.09 115.1 720.5 0.0 11.3 8.96 84.7
11.000 8,862 8.54 4.81 3.28 2.35 1.51 1.34 1.11 106.9 828.4 0.0 11.8 8.92 81.1
21.000 8,925 8.68 4.96 3.31 2.34 1.56 1.37 1.12 107.5 793.9 0.0 11.7 9.41 91.8
32.000 8,874 9.04 4.82 3.20 2.30 1.48 1.32 1.04 91.8 874.1 0.0 12.4 9.66 80.9
41.000 8,886 8.86 4.59 3.09 2.20 1.44 1.26 0.98 90.3 957.5 0.0 12.9 9.26 85.4
54.000 8,854 8.81 5.01 3.43 2.42 1.56 1.36 1.07 106.9 742.6 0.0 11.5 8.63 83.4
63.000 8,850 8.27 4.72 3.19 2.30 1.49 1.31 1.03 111.9 842.1 0.0 12.0 8.91 82.1
72.000 8,901 7.96 4.66 3.11 2.21 1.44 1.27 1.01 124.9 766.5 0.0 12.5 9.25 83.5
82.000 8,894 8.33 4.64 3.09 2.22 1.44 1.27 1.01 107.3 861.9 0.0 12.7 9.34 82.0
92.000 8,894 8.42 4.74 3.20 2.30 1.48 1.35 1.06 107.7 863.7 0.0 12.1 9.56 80.7
102.000 8,945 8.34 4.91 3.39 2.45 1.62 1.44 1.15 116.7 869.7 0.0 11.0 8.69 88.7
111.000 8,838 8.67 4.88 3.28 2.37 1.57 1.39 1.12 102.1 885.2 0.0 11.5 9.27 88.7
121.000 8,945 8.85 4.85 3.19 2.30 1.52 1.35 1.07 97.2 889.1 0.0 12.2 9.81 87.2
131.000 8,862 7.93 4.61 3.05 2.18 1.43 1.29 1.06 120.2 828.9 0.0 12.5 9.82 84.6
143.000 8,905 9.30 5.00 3.27 2.33 1.49 1.32 1.08 92.1 767.5 0.0 12.4 9.84 79.9
154.000 8,866 9.34 5.00 3.17 2.23 1.43 1.26 1.01 92.0 690.1 0.0 13.1 10.41 80.3
162.000 8,830 8.91 4.89 3.16 2.22 1.43 1.31 0.99 97.8 755.0 0.0 12.7 10.68 81.5
173.000 8,882 8.37 5.00 3.16 2.25 1.48 1.27 0.93 126.6 604.2 0.0 12.5 9.79 85.2
181.000 8,933 8.67 5.13 3.36 2.33 1.50 1.30 1.02 124.5 565.8 0.0 12.2 9.32 82.4
193.000 8,981 7.87 5.06 3.30 2.30 1.52 1.34 1.07 172.4 507.8 0.0 11.8 9.50 88.3
202.000 8,941 8.54 5.02 3.28 2.30 1.46 1.27 1.03 125.0 583.0 0.0 12.5 9.32 78.5
215.000 8,901 10.20 5.43 3.53 2.41 1.56 1.32 1.01 86.0 599.8 0.0 12.1 9.52 85.9
221.000 8,977 10.13 5.36 3.41 2.37 1.55 1.31 1.02 84.6 649.4 0.0 12.4 9.88 86.8
230.000 8,957 9.63 5.13 3.30 2.30 1.44 1.26 1.04 91.5 637.9 0.0 13.0 9.96 76.5
240.000 9,040 8.39 4.74 3.08 2.15 1.32 1.15 0.93 121.9 600.5 0.0 14.0 9.40 72.1
252.000 8,929 8.87 5.20 3.46 2.39 1.53 1.33 1.11 119.2 576.4 0.0 11.8 9.11 82.0
265.000 8,874 8.47 5.13 3.53 2.55 1.66 1.46 1.19 123.9 728.7 0.0 10.5 8.42 84.8
276.000 8,878 8.57 4.98 3.41 2.47 1.62 1.45 1.17 110.2 851.2 0.0 10.9 8.95 86.3
285.000 8,949 9.30 5.47 3.43 2.40 1.56 1.37 1.11 113.6 523.8 0.0 11.9 10.36 84.0
295.000 8,965 8.02 4.78 3.19 2.33 1.53 1.38 1.12 124.5 841.6 0.0 11.7 9.47 84.3
312.000 8,909 11.80 6.87 4.27 2.91 1.81 1.54 1.17 98.5 307.3 0.0 10.3 9.78 78.4
323.000 8,909 13.31 8.00 4.90 3.29 2.00 1.66 1.28 114.3 179.6 0.0 9.2 9.08 76.0
333.000 8,929 11.19 6.78 4.23 2.89 1.81 1.53 1.14 121.8 273.4 0.0 10.2 9.38 79.3
355.000 8,957 9.64 5.50 3.60 2.53 1.63 1.38 1.02 102.4 585.7 0.0 11.4 9.02 83.2
----------------------------------------------------------------------------------------------------------------------------------------
Mean: 9.05 5.17 3.39 2.39 1.54 1.35 1.07 110.3 693.8 0.0 11.9 9.41 83.2
Std. Dev: 1.13 0.70 0.37 0.23 0.13 0.09 0.07 16.6 183.0 0.0 0.9 0.51 4.1
Var Coeff(%): 12.50 13.50 11.02 9.44 8.24 7.02 6.89 15.0 26.4 0.0 7.8 5.37 5.0
----------------------------------------------------------------------------------------------------------------------------------------

Page 1
 US84 WB Approach.asc
----------------------------------------------------------------------------------------------------------------------------------------
TTI MODULUS ANALYSIS SYSTEM (SUMMARY REPORT) (Version 6.0)
----------------------------------------------------------------------------------------------------------------------------------------
District:5 (Lubbock) MODULI RANGE(psi)
County :152 (LUBBOCK) Thickness(in) Minimum Maximum Poisson Ratio Values
Highway/Road: US0084 Pavement: 6.00 50,000 300,000 H1: v = 0.38
Base: 10.00 10,000 1,000,000 H2: v = 0.35
Subbase: 0.00 H3: v = 0.00
Subgrade: 66.49(by DB) 15,000 H4: v = 0.40
----------------------------------------------------------------------------------------------------------------------------------------
Load Measured Deflection (mils): Calculated Moduli values (ksi): Absolute Dpth to
Station (lbs) R1 R2 R3 R4 R5 R6 R7 SURF(E1) BASE(E2) SUBB(E3) SUBG(E4) ERR/Sens Bedrock
----------------------------------------------------------------------------------------------------------------------------------------
0.000 9,247 14.15 8.35 4.51 2.93 1.81 1.57 1.27 140.4 106.3 0.0 10.6 11.12 76.0
13.000 9,259 13.57 7.76 4.07 2.57 1.59 1.37 1.10 136.3 107.1 0.0 12.2 11.89 75.9
25.000 9,144 10.42 6.02 3.36 2.21 1.38 1.20 0.98 156.1 185.1 0.0 14.1 11.02 75.5
37.000 9,203 9.26 5.26 3.42 2.45 1.59 1.43 1.16 102.8 748.8 0.0 11.7 10.13 82.8
47.000 9,144 11.15 6.19 3.75 2.65 1.74 1.55 1.17 82.9 549.8 0.0 11.1 11.46 87.1
52.000 9,076 14.95 8.47 4.73 3.16 1.93 1.73 1.36 97.2 144.2 0.0 9.9 11.46 74.9
61.000 9,080 15.29 7.91 4.29 2.89 1.86 1.64 1.32 64.1 232.9 0.0 10.9 12.92 84.1
74.000 9,033 14.31 8.40 4.76 3.26 2.09 1.82 1.43 92.5 195.5 0.0 9.3 11.51 83.5
82.000 9,025 14.97 8.52 4.71 3.17 1.99 1.76 1.41 92.9 153.5 0.0 9.8 11.68 79.2
94.000 9,080 14.74 8.18 4.41 2.88 1.84 1.60 1.26 100.1 130.7 0.0 10.8 12.53 83.0
102.000 9,017 15.22 8.27 4.14 2.63 1.63 1.41 1.11 107.1 91.9 0.0 11.9 13.31 75.9
112.000 9,092 13.76 7.22 3.54 2.27 1.41 1.26 1.03 109.8 106.7 0.0 14.0 14.32 74.5
123.000 9,033 13.98 7.78 3.83 2.41 1.51 1.34 1.11 129.9 89.8 0.0 12.8 13.91 76.9
133.000 9,005 15.85 8.35 4.07 2.52 1.59 1.37 1.09 102.1 80.9 0.0 12.4 14.31 79.5
149.000 8,973 14.48 7.87 3.89 2.50 1.59 1.38 1.11 109.3 99.5 0.0 12.4 14.23 79.6
157.000 8,945 15.17 8.23 4.00 2.53 1.61 1.38 1.12 110.6 84.7 0.0 12.2 14.41 80.2
163.000 8,973 15.07 8.46 4.07 2.52 1.60 1.32 1.06 131.2 72.2 0.0 12.2 13.91 79.9
178.000 8,957 13.82 7.66 3.81 2.40 1.45 1.33 1.23 130.9 89.4 0.0 12.9 13.17 71.3
186.000 8,973 14.00 7.83 3.87 2.54 1.71 1.50 1.19 112.0 114.0 0.0 11.9 15.49 94.7
193.000 8,977 14.90 8.09 3.98 2.63 1.75 1.62 1.30 95.7 115.9 0.0 11.6 15.99 186.6
205.000 8,945 15.12 8.06 3.98 2.65 1.74 1.60 1.23 90.9 116.4 0.0 11.6 15.72 197.6
210.000 9,017 10.57 5.99 3.58 2.55 1.61 1.49 1.34 92.5 491.5 0.0 11.6 11.96 76.7
225.000 9,005 10.26 5.88 3.43 2.38 1.57 1.44 1.22 97.6 469.2 0.0 12.2 12.55 87.2
230.000 8,993 10.95 6.30 3.74 2.64 1.74 1.57 1.30 90.0 476.9 0.0 10.9 11.94 87.8
244.000 8,949 11.63 6.66 3.61 2.44 1.61 1.48 1.28 109.2 227.5 0.0 12.3 13.35 87.8
251.000 8,969 11.67 6.26 3.51 2.38 1.56 1.42 1.23 80.6 373.6 0.0 12.6 13.13 86.6
264.000 8,981 10.51 6.08 3.35 2.27 1.53 1.40 1.19 109.1 320.6 0.0 12.9 13.25 94.1
271.000 8,929 11.75 6.48 3.44 2.29 1.50 1.39 1.17 109.6 194.8 0.0 13.3 14.06 85.6
281.000 8,921 12.20 6.40 3.39 2.28 1.49 1.38 1.15 82.1 266.5 0.0 13.4 14.20 84.8
291.000 8,866 12.19 6.61 3.52 2.33 1.55 1.44 1.20 94.6 215.6 0.0 12.9 14.43 91.3
301.000 8,909 12.49 6.80 3.52 2.36 1.58 1.47 1.24 96.4 194.2 0.0 12.9 15.22 92.1
315.000 8,917 13.69 7.40 3.98 2.48 1.63 1.48 1.24 106.6 127.2 0.0 12.0 13.69 91.8
321.000 8,937 13.59 7.61 3.96 2.52 1.66 1.48 1.26 117.3 118.2 0.0 11.8 14.01 90.2
334.000 8,929 14.22 8.05 4.04 2.55 1.65 1.53 1.29 123.1 96.7 0.0 11.7 14.54 84.6
----------------------------------------------------------------------------------------------------------------------------------------
Mean: 13.23 7.34 3.89 2.57 1.65 1.48 1.21 106.0 211.4 0.0 12.0 13.26 82.5
Std. Dev: 1.83 0.96 0.40 0.26 0.16 0.14 0.10 19.1 162.8 0.0 1.1 1.46 13.1
Var Coeff(%): 13.80 13.11 10.40 10.27 9.65 9.38 8.65 18.0 77.0 0.0 9.3 11.05 15.8
----------------------------------------------------------------------------------------------------------------------------------------

Page 1
 US84 EB Departure.asc
----------------------------------------------------------------------------------------------------------------------------------------
TTI MODULUS ANALYSIS SYSTEM (SUMMARY REPORT) (Version 6.0)
----------------------------------------------------------------------------------------------------------------------------------------
District:5 (Lubbock) MODULI RANGE(psi)
County :152 (LUBBOCK) Thickness(in) Minimum Maximum Poisson Ratio Values
Highway/Road: US0084 Pavement: 6.00 50,000 300,000 H1: v = 0.38
Base: 10.00 10,000 1,500,000 H2: v = 0.35
Subbase: 0.00 H3: v = 0.00
Subgrade: 75.06(by DB) 15,000 H4: v = 0.40
----------------------------------------------------------------------------------------------------------------------------------------
Load Measured Deflection (mils): Calculated Moduli values (ksi): Absolute Dpth to
Station (lbs) R1 R2 R3 R4 R5 R6 R7 SURF(E1) BASE(E2) SUBB(E3) SUBG(E4) ERR/Sens Bedrock
----------------------------------------------------------------------------------------------------------------------------------------
0.000 8,945 9.26 5.12 3.33 2.33 1.52 1.43 1.22 99.5 666.0 0.0 13.5 10.64 85.5
11.000 8,941 9.19 4.72 3.16 2.27 1.52 1.41 1.18 85.7 984.5 0.0 13.7 10.15 92.3
21.000 8,890 9.76 5.35 3.44 2.41 1.57 1.42 1.18 94.1 593.0 0.0 13.2 10.17 85.6
32.000 8,870 9.20 5.03 3.28 2.31 1.53 1.40 1.15 96.4 713.5 0.0 13.5 10.25 89.7
41.000 8,901 9.16 5.03 3.26 2.34 1.52 1.38 1.17 98.1 704.9 0.0 13.6 10.01 82.9
50.000 8,866 8.89 5.22 3.36 2.41 1.60 1.44 1.19 113.7 626.8 0.0 12.8 9.96 89.1
62.000 8,886 9.41 5.46 3.57 2.53 1.66 1.48 1.21 107.9 574.7 0.0 12.3 9.50 87.6
70.000 8,882 8.87 5.03 3.45 2.50 1.69 1.50 1.25 102.9 859.8 0.0 12.0 8.74 97.2
80.000 8,878 8.37 4.93 3.36 2.41 1.61 1.47 1.22 118.5 782.2 0.0 12.3 9.21 92.3
91.000 8,882 7.91 4.97 3.41 2.42 1.60 1.44 1.17 152.7 621.8 0.0 12.1 8.61 89.4
103.000 8,874 8.04 4.98 3.37 2.43 1.61 1.50 1.26 138.2 701.9 0.0 12.0 9.45 88.8
113.000 8,842 7.83 4.80 3.22 2.32 1.57 1.45 1.23 135.0 775.0 0.0 12.5 9.66 95.5
124.000 8,882 8.52 4.32 2.98 2.24 1.53 1.44 1.22 89.1 1338.6 0.0 13.4 9.68 96.5 *
132.000 8,905 7.66 4.26 2.96 2.24 1.53 1.43 1.24 111.9 1304.4 0.0 13.0 9.20 95.2 *
144.000 8,878 7.76 4.30 3.01 2.26 1.51 1.43 1.28 109.7 1266.9 0.0 13.0 9.26 88.1
152.000 8,878 7.33 4.37 3.03 2.25 1.58 1.47 1.23 128.0 1219.5 0.0 12.4 9.31 111.9
162.000 8,890 7.67 4.26 3.01 2.28 1.57 1.48 1.25 114.5 1270.1 0.0 12.7 9.00 99.5 *
173.000 8,850 7.50 4.09 2.98 2.28 1.58 1.50 1.26 119.6 1267.0 0.0 12.7 8.38 102.4 *
180.000 8,854 7.85 4.25 3.10 2.36 1.62 1.54 1.31 113.7 1222.0 0.0 12.2 8.57 99.2 *
193.000 8,854 7.91 4.47 3.22 2.44 1.67 1.57 1.32 115.4 1173.1 0.0 11.7 8.50 98.6 *
204.000 8,901 8.63 4.52 3.20 2.39 1.64 1.55 1.32 93.1 1242.3 0.0 12.4 9.14 101.1 *
215.000 8,866 7.60 4.40 3.10 2.29 1.58 1.46 1.23 119.4 1199.1 0.0 12.4 8.89 103.1
222.000 8,901 7.33 4.18 2.95 2.17 1.48 1.46 1.22 120.6 1312.3 0.0 13.1 9.90 96.8 *
232.000 8,894 7.91 4.15 2.86 2.11 1.43 1.33 1.13 100.0 1303.8 0.0 14.2 9.54 93.4
242.000 8,897 7.87 4.24 2.89 2.09 1.41 1.31 1.09 105.7 1125.8 0.0 14.4 9.75 92.9
251.000 8,897 8.57 4.34 2.94 2.13 1.46 1.33 1.10 88.9 1177.6 0.0 14.3 9.75 100.3
264.000 8,838 8.52 4.27 2.94 2.16 1.44 1.31 1.07 88.4 1199.1 0.0 14.2 9.21 88.8
277.000 8,866 8.61 4.45 3.02 2.20 1.46 1.35 1.10 91.3 1077.1 0.0 14.0 9.70 87.9
281.000 8,874 8.63 4.52 3.11 2.26 1.48 1.35 1.10 94.1 1012.9 0.0 13.7 9.13 84.8
292.000 8,858 8.32 4.57 3.12 2.26 1.49 1.32 1.09 105.3 914.6 0.0 13.6 8.74 86.6
304.000 8,854 10.35 5.82 3.65 2.50 1.59 1.39 1.15 102.6 384.0 0.0 13.1 9.75 81.1
316.000 8,894 8.24 4.86 3.31 2.35 1.52 1.36 1.12 127.8 668.0 0.0 13.1 8.70 82.9
327.000 8,858 8.24 4.74 3.12 2.16 1.38 1.22 1.01 128.5 559.3 0.0 14.7 9.13 80.0
343.000 8,878 9.51 5.44 3.61 2.44 1.50 1.30 1.04 118.2 406.7 0.0 13.5 8.49 75.8
356.000 8,842 8.74 5.36 3.61 2.47 1.59 1.31 0.99 151.2 400.0 0.0 12.7 7.36 84.5
----------------------------------------------------------------------------------------------------------------------------------------
Mean: 8.43 4.71 3.20 2.31 1.54 1.42 1.18 110.8 932.8 0.0 13.1 9.30 91.1
Std. Dev: 0.73 0.45 0.22 0.12 0.07 0.08 0.09 17.3 311.2 0.0 0.8 0.66 7.5
Var Coeff(%): 8.69 9.64 6.99 5.08 4.85 5.73 7.23 15.6 33.4 0.0 5.9 7.12 8.2
----------------------------------------------------------------------------------------------------------------------------------------

Page 1
 US84 WB Departure.asc
----------------------------------------------------------------------------------------------------------------------------------------
TTI MODULUS ANALYSIS SYSTEM (SUMMARY REPORT) (Version 6.0)
----------------------------------------------------------------------------------------------------------------------------------------
District:5 (Lubbock) MODULI RANGE(psi)
County :152 (LUBBOCK) Thickness(in) Minimum Maximum Poisson Ratio Values
Highway/Road: US0084 Pavement: 6.00 50,000 300,000 H1: v = 0.38
Base: 10.00 10,000 1,000,000 H2: v = 0.35
Subbase: 0.00 H3: v = 0.00
Subgrade: 74.37(by DB) 15,000 H4: v = 0.40
----------------------------------------------------------------------------------------------------------------------------------------
Load Measured Deflection (mils): Calculated Moduli values (ksi): Absolute Dpth to
Station (lbs) R1 R2 R3 R4 R5 R6 R7 SURF(E1) BASE(E2) SUBB(E3) SUBG(E4) ERR/Sens Bedrock
----------------------------------------------------------------------------------------------------------------------------------------
0.000 8,909 11.65 6.60 3.66 2.52 1.68 1.50 1.20 106.7 229.0 0.0 13.1 12.06 91.1
10.000 8,981 12.87 7.11 4.11 2.85 1.91 1.71 1.34 78.5 320.5 0.0 11.5 11.74 95.7
20.000 8,973 11.43 6.98 4.21 3.02 2.01 1.81 1.39 106.0 345.1 0.0 10.4 10.84 91.7
31.000 8,937 13.38 7.53 4.57 3.27 2.11 1.87 1.46 74.3 356.9 0.0 10.0 10.57 84.8
40.000 8,894 10.77 6.59 4.07 2.89 1.89 1.72 1.37 113.6 356.6 0.0 10.8 10.52 87.0
50.000 8,878 11.30 6.29 3.88 2.74 1.84 1.61 1.26 82.4 488.2 0.0 11.5 10.51 96.3
60.000 8,886 11.63 6.51 3.98 2.84 1.89 1.67 1.33 81.2 462.2 0.0 11.2 10.69 92.4
70.000 8,897 11.47 6.45 3.92 2.75 1.84 1.59 1.28 86.1 415.0 0.0 11.6 10.46 94.9
82.000 8,905 10.92 6.15 3.89 2.82 1.91 1.64 1.44 83.9 588.8 0.0 11.0 9.71 98.6
92.000 8,894 10.57 5.81 3.86 2.85 1.93 1.74 1.42 80.3 807.4 0.0 10.5 9.62 98.6
100.000 8,882 9.74 6.03 3.94 2.85 1.91 1.72 1.41 115.3 538.2 0.0 10.4 9.59 94.7
129.000 8,937 12.33 7.18 4.52 3.22 2.13 1.86 1.48 84.2 405.5 0.0 9.8 9.84 92.9
130.000 8,846 12.41 7.33 4.58 3.20 2.06 1.81 1.46 93.6 305.8 0.0 9.9 9.80 85.8
141.000 8,925 11.65 6.98 4.46 3.21 2.05 1.80 1.41 97.7 380.1 0.0 9.9 9.30 82.4
152.000 8,925 12.41 7.06 4.36 3.02 1.99 1.68 1.32 85.6 337.4 0.0 10.7 9.65 92.3
161.000 8,941 11.48 6.82 4.14 2.93 1.90 1.65 1.30 104.4 316.6 0.0 11.0 10.03 84.8
171.000 8,886 12.17 6.80 4.14 2.90 1.90 1.68 1.32 81.1 381.9 0.0 11.1 10.66 89.0
181.000 8,894 12.09 6.78 4.10 2.87 1.87 1.63 1.28 84.3 355.5 0.0 11.3 10.46 87.0
193.000 8,941 12.40 6.80 4.13 2.95 1.93 1.73 1.37 74.3 441.4 0.0 11.0 10.98 87.5
202.000 8,933 14.26 7.56 4.31 3.00 2.00 1.77 1.38 64.7 310.7 0.0 11.0 11.87 94.5
214.000 8,913 13.57 8.05 4.49 3.10 2.01 1.78 1.41 119.3 145.0 0.0 10.6 10.94 85.0
221.000 8,878 13.84 7.82 4.39 3.01 1.99 1.76 1.37 87.9 198.1 0.0 11.0 11.50 91.5
231.000 8,894 14.42 7.85 4.35 3.00 1.97 1.76 1.39 75.7 211.5 0.0 11.2 11.92 89.0
242.000 8,894 13.46 7.79 4.34 2.99 1.98 1.76 1.37 100.1 180.9 0.0 11.0 11.64 91.1
251.000 8,854 14.51 8.32 4.65 3.21 2.08 1.84 1.44 93.2 160.9 0.0 10.3 11.10 85.7
263.000 8,858 13.80 7.91 4.40 3.05 2.00 1.80 1.39 92.8 186.1 0.0 10.8 11.61 88.0
275.000 8,866 14.32 8.30 4.47 3.07 2.02 1.78 1.43 111.5 125.7 0.0 10.7 12.13 88.6
282.000 8,894 14.59 8.37 4.50 3.02 2.00 1.80 1.47 112.4 115.3 0.0 10.8 12.50 91.9
292.000 8,886 14.42 8.36 4.65 3.17 2.06 1.86 1.44 104.5 140.9 0.0 10.4 11.38 86.7
301.000 8,838 14.74 8.59 4.63 3.11 2.05 1.81 1.43 117.3 108.2 0.0 10.4 12.04 90.6
314.000 8,901 13.94 7.91 4.59 3.12 2.05 1.83 1.46 83.9 221.9 0.0 10.5 11.22 91.4
324.000 8,862 12.52 7.24 4.21 2.91 1.91 1.68 1.34 95.7 247.3 0.0 11.2 10.92 89.0
336.000 8,929 12.31 7.37 4.31 3.00 1.97 1.74 1.38 108.6 233.7 0.0 10.8 10.77 88.8
----------------------------------------------------------------------------------------------------------------------------------------
Mean: 12.65 7.25 4.27 2.98 1.96 1.74 1.38 93.4 315.7 0.0 10.8 10.86 90.4
Std. Dev: 1.36 0.75 0.27 0.16 0.09 0.09 0.07 14.7 153.7 0.0 0.6 0.87 4.0
Var Coeff(%): 10.75 10.34 6.29 5.48 4.67 4.97 4.86 15.7 48.7 0.0 5.8 7.97 4.4
----------------------------------------------------------------------------------------------------------------------------------------

Page 1
 SH194 Approach mod.asc
----------------------------------------------------------------------------------------------------------------------------------------
TTI MODULUS ANALYSIS SYSTEM (SUMMARY REPORT) (Version 6.0)
----------------------------------------------------------------------------------------------------------------------------------------
District:5 (Lubbock) MODULI RANGE(psi)
County :96 (HALE) Thickness(in) Minimum Maximum Poisson Ratio Values
Highway/Road: SH0194 Pavement: 1.50 693,200 693,200 H1: v = 0.38
Base: 3.50 50,000 1,000,000 H2: v = 0.35
Subbase: 6.00 4,000 200,000 H3: v = 0.35
Subgrade: 167.78(by DB) 15,000 H4: v = 0.40
----------------------------------------------------------------------------------------------------------------------------------------
Load Measured Deflection (mils): Calculated Moduli values (ksi): Absolute Dpth to
Station (lbs) R1 R2 R3 R4 R5 R6 R7 SURF(E1) BASE(E2) SUBB(E3) SUBG(E4) ERR/Sens Bedrock
----------------------------------------------------------------------------------------------------------------------------------------
0.000 9,223 29.16 22.08 13.14 8.20 5.16 4.11 3.35 693.2 396.3 19.6 6.6 4.63 163.3
10.000 9,215 34.96 24.54 14.47 8.66 5.33 4.19 3.25 693.2 218.9 15.4 6.2 3.70 151.8
20.000 9,148 34.77 22.19 12.54 7.74 4.73 3.96 3.38 693.2 115.9 21.2 6.9 4.26 140.6
30.000 9,112 31.94 22.85 13.56 8.19 4.97 4.15 3.39 693.2 261.2 17.5 6.5 4.79 139.9
42.000 9,040 34.06 23.54 12.94 7.76 4.92 3.95 3.08 693.2 177.5 15.7 6.6 4.87 173.6
51.000 9,096 33.86 23.60 13.57 8.00 4.91 3.94 3.24 693.2 220.9 13.9 6.6 4.23 148.6
62.000 9,029 40.87 25.54 12.56 7.21 4.50 3.88 3.38 693.2 96.6 10.1 6.9 5.85 160.3
71.000 9,072 39.54 24.33 11.85 6.61 4.17 3.54 2.97 693.2 102.8 9.4 7.5 5.53 136.2
83.000 9,112 31.94 20.22 10.27 6.09 3.90 3.26 2.71 693.2 124.8 17.5 8.3 5.51 174.4
92.000 9,172 28.45 18.03 9.39 5.77 3.78 3.17 2.68 693.2 124.6 26.7 8.9 5.65 196.3
101.000 9,148 27.53 17.88 9.85 6.11 3.96 3.28 2.69 693.2 155.2 28.5 8.6 4.95 183.6
115.000 9,172 22.20 15.27 9.11 6.04 4.06 3.33 2.70 693.2 148.4 74.6 8.7 5.49 232.1
121.000 9,279 21.12 15.14 8.79 5.90 3.98 3.33 2.69 693.2 205.1 72.7 9.0 6.32 235.7
131.000 9,160 19.17 13.44 8.22 5.74 4.00 3.32 2.67 693.2 103.4 164.8 9.0 6.05 300.0
142.000 9,128 21.43 13.56 7.91 5.48 3.79 3.13 2.67 693.2 55.6 162.1 9.5 6.13 288.3
151.000 9,187 20.53 14.07 8.22 5.53 3.71 3.21 2.69 693.2 144.5 87.7 9.5 6.69 222.2
161.000 9,096 20.49 13.95 8.43 5.66 3.72 3.26 2.72 693.2 130.3 97.1 9.2 6.42 191.2
172.000 9,140 22.82 14.91 8.83 6.09 4.05 3.37 2.62 693.2 72.4 111.0 8.7 5.86 208.0
180.000 9,084 24.18 16.37 9.24 5.99 3.89 3.18 2.56 693.2 173.8 42.8 8.8 5.28 181.2
199.000 9,044 31.31 17.82 9.96 6.25 4.08 3.35 2.65 693.2 51.0 41.3 8.3 4.65 193.0
202.000 9,021 31.32 18.61 10.03 6.57 4.11 3.37 2.82 693.2 74.6 30.9 8.1 4.58 148.0
211.000 9,044 27.99 17.27 9.54 6.13 3.95 3.28 2.66 693.2 92.3 37.7 8.5 5.07 174.0
220.000 9,120 20.08 13.81 8.16 5.60 3.77 3.19 2.64 693.2 124.7 106.8 9.4 6.36 227.0
231.000 9,076 22.59 15.38 9.13 6.24 4.12 3.47 2.84 693.2 116.4 84.0 8.4 5.93 199.1
240.000 9,076 26.45 18.50 10.54 6.63 4.50 3.64 2.80 693.2 198.5 34.9 7.7 5.47 272.0
251.000 9,088 30.38 21.51 12.17 7.67 4.81 3.86 3.12 693.2 232.6 20.8 6.9 4.61 157.3
261.000 9,056 34.52 23.84 12.45 6.20 4.01 3.24 2.59 693.2 225.4 5.9 8.0 5.96 88.3
270.000 9,080 23.24 16.02 9.21 5.77 3.64 2.94 2.45 693.2 275.2 32.7 9.2 4.41 154.2
281.000 9,056 25.07 16.98 9.72 5.83 3.50 2.94 2.50 693.2 256.7 24.3 9.0 4.64 126.2
293.000 8,965 30.72 18.97 9.97 6.20 3.95 3.22 2.67 693.2 99.4 24.7 8.2 4.61 166.0
301.000 9,076 21.50 15.02 9.06 5.91 3.83 3.14 2.55 693.2 205.6 62.8 8.9 5.03 177.7
----------------------------------------------------------------------------------------------------------------------------------------
Mean: 27.88 18.56 10.41 6.51 4.19 3.46 2.83 693.2 160.7 48.9 8.1 5.28 178.8
Std. Dev: 6.05 3.84 1.91 0.92 0.49 0.36 0.29 0.0 76.8 43.1 1.0 0.76 47.0
Var Coeff(%): 21.70 20.68 18.37 14.14 11.66 10.44 10.21 0.0 47.8 88.3 12.6 14.43 26.3
----------------------------------------------------------------------------------------------------------------------------------------

Page 1
 SH194 Departure mod.asc
----------------------------------------------------------------------------------------------------------------------------------------
TTI MODULUS ANALYSIS SYSTEM (SUMMARY REPORT) (Version 6.0)
----------------------------------------------------------------------------------------------------------------------------------------
District:5 (Lubbock) MODULI RANGE(psi)
County :96 (HALE) Thickness(in) Minimum Maximum Poisson Ratio Values
Highway/Road: SH0194 Pavement: 1.50 606,600 606,600 H1: v = 0.38
Base: 3.50 50,000 1,000,000 H2: v = 0.35
Subbase: 6.00 4,000 200,000 H3: v = 0.35
Subgrade: 123.58(by DB) 15,000 H4: v = 0.40
----------------------------------------------------------------------------------------------------------------------------------------
Load Measured Deflection (mils): Calculated Moduli values (ksi): Absolute Dpth to
Station (lbs) R1 R2 R3 R4 R5 R6 R7 SURF(E1) BASE(E2) SUBB(E3) SUBG(E4) ERR/Sens Bedrock
----------------------------------------------------------------------------------------------------------------------------------------
0.000 8,703 51.02 30.96 13.81 7.28 4.51 3.89 3.48 606.6 63.1 7.0 5.6 7.81 108.7
10.000 8,834 49.92 29.24 13.32 7.24 4.62 3.81 3.32 606.6 55.4 9.2 5.7 7.46 122.7
21.000 8,766 49.59 27.09 11.70 6.01 3.70 3.25 2.89 606.6 51.8 7.2 6.6 7.15 95.9
33.000 8,850 42.51 24.48 11.40 6.63 4.09 3.47 2.97 606.6 55.3 14.7 6.5 6.91 147.6
43.000 8,826 40.81 23.89 11.87 6.98 4.30 3.61 3.00 606.6 56.9 18.6 6.2 6.48 147.6
52.000 8,691 45.68 22.35 10.15 5.72 3.66 3.12 2.65 606.6 50.0 9.8 7.3 8.70 141.4 *
64.000 8,953 33.15 20.02 9.88 5.50 3.59 2.87 2.37 606.6 89.9 20.3 7.7 7.08 130.2
74.000 8,909 33.65 20.64 9.65 5.48 3.44 3.02 2.64 606.6 94.8 17.3 7.8 8.12 146.3
82.000 8,830 37.31 21.87 10.78 6.26 4.05 3.50 3.13 606.6 63.8 20.4 6.8 7.78 176.6
103.000 8,735 43.60 25.40 11.22 5.85 3.81 3.15 2.75 606.6 68.3 8.9 6.8 8.14 100.1
112.000 8,878 31.76 20.71 10.50 6.03 3.62 3.18 2.79 606.6 132.0 19.7 7.3 7.82 129.6
121.000 8,909 31.69 21.41 10.98 6.60 4.17 3.59 3.06 606.6 163.4 20.3 6.8 7.79 163.6
131.000 8,822 38.41 24.79 11.92 7.20 4.61 3.98 3.36 606.6 84.4 18.1 6.0 8.66 181.0
142.000 8,854 34.15 21.96 11.12 6.47 4.08 3.46 2.90 606.6 107.9 20.1 6.7 7.38 164.4
151.000 8,985 33.96 24.54 12.02 6.78 4.06 3.24 2.70 606.6 229.7 9.7 6.6 7.56 132.2
160.000 8,921 35.50 22.77 9.81 5.52 3.27 2.62 2.30 606.6 131.9 8.7 8.0 7.35 133.9
171.000 8,973 27.94 17.94 9.61 5.67 3.59 2.90 2.31 606.6 140.9 29.4 7.9 6.22 161.6
180.000 8,933 30.08 18.77 9.66 5.69 3.35 2.86 2.50 606.6 118.8 23.7 7.9 6.62 118.5
191.000 8,770 34.27 25.04 11.09 6.02 3.64 2.96 2.48 606.6 202.7 6.9 7.1 9.16 117.4
201.000 8,921 31.09 18.76 9.03 5.00 3.13 2.60 2.21 606.6 106.0 18.5 8.5 6.86 124.6
211.000 8,909 32.54 18.59 8.43 4.85 3.25 2.53 2.23 606.6 77.1 19.1 8.7 8.39 155.9
221.000 8,866 30.88 21.58 8.61 5.19 3.43 2.93 2.47 606.6 158.4 12.2 8.3 12.84 82.0
231.000 9,005 24.02 17.28 9.17 5.00 3.47 2.93 2.50 606.6 320.0 25.4 8.5 9.40 115.5
242.000 8,913 26.47 16.62 8.89 5.35 3.39 2.87 2.42 606.6 113.4 37.9 8.3 7.15 159.2
251.000 8,945 27.04 17.44 8.98 5.39 3.54 2.98 2.49 606.6 135.6 31.5 8.2 8.34 200.7
261.000 8,822 31.21 19.99 10.00 5.59 3.72 3.02 2.47 606.6 116.5 21.4 7.5 7.99 133.9
271.000 8,838 36.74 24.62 11.97 6.71 6.27 4.54 3.44 606.6 75.0 26.7 5.8 15.92 141.5
282.000 8,655 52.42 27.89 12.65 6.86 4.24 3.83 3.39 606.6 50.0 7.5 6.0 7.53 120.5 *
291.000 8,842 39.66 26.20 14.23 7.85 4.69 3.59 2.79 606.6 127.3 13.1 5.6 5.02 134.7
301.000 8,683 49.11 28.34 13.79 7.37 4.17 3.33 2.77 606.6 60.2 8.6 5.7 4.85 115.0
----------------------------------------------------------------------------------------------------------------------------------------
Mean: 36.87 22.71 10.87 6.14 3.92 3.25 2.76 606.6 110.0 17.1 7.1 7.95 134.6
Std. Dev: 7.93 3.81 1.62 0.81 0.63 0.46 0.38 0.0 60.7 8.1 1.0 2.06 27.0
Var Coeff(%): 21.50 16.79 14.92 13.27 15.99 14.18 13.74 0.0 55.2 47.3 14.0 25.94 20.0
----------------------------------------------------------------------------------------------------------------------------------------

Page 1
 FM1318 Approach.asc
----------------------------------------------------------------------------------------------------------------------------------------
TTI MODULUS ANALYSIS SYSTEM (SUMMARY REPORT) (Version 6.0)
----------------------------------------------------------------------------------------------------------------------------------------
District:5 (Lubbock) MODULI RANGE(psi)
County :219 (SWISHER) Thickness(in) Minimum Maximum Poisson Ratio Values
Highway/Road: FM1318 Pavement: 1.50 315,000 315,000 H1: v = 0.38
Base: 6.00 10,000 150,000 H2: v = 0.35
Subbase: 0.00 H3: v = 0.00
Subgrade: 89.33(by DB) 15,000 H4: v = 0.40
----------------------------------------------------------------------------------------------------------------------------------------
Load Measured Deflection (mils): Calculated Moduli values (ksi): Absolute Dpth to
Station (lbs) R1 R2 R3 R4 R5 R6 R7 SURF(E1) BASE(E2) SUBB(E3) SUBG(E4) ERR/Sens Bedrock
----------------------------------------------------------------------------------------------------------------------------------------
0.000 8,770 43.88 16.52 6.39 4.23 3.00 2.69 2.39 315.0 25.4 0.0 9.3 23.17 71.0
11.000 8,814 59.83 20.78 7.52 4.84 3.27 3.14 2.69 315.0 14.6 0.0 7.8 21.87 60.9
21.000 8,707 69.93 22.89 8.70 5.70 3.89 3.48 2.91 315.0 11.4 0.0 6.9 23.38 68.6
31.000 8,754 45.35 18.08 8.13 5.67 4.08 3.50 2.81 315.0 30.8 0.0 7.8 26.49 185.3
43.000 8,802 52.28 21.46 9.09 6.00 4.18 3.88 3.13 315.0 25.1 0.0 6.9 23.89 112.5
53.000 8,818 49.36 22.01 8.71 5.88 4.16 3.72 3.08 315.0 29.1 0.0 6.9 23.30 78.1
60.000 8,766 66.35 25.00 9.65 6.13 4.30 3.73 2.93 315.0 15.4 0.0 6.2 21.94 72.8
71.000 8,878 42.06 17.07 7.69 5.39 3.80 3.35 2.76 315.0 35.4 0.0 8.4 26.40 186.9
83.000 8,874 45.74 17.98 7.59 5.04 3.47 3.02 2.47 315.0 27.2 0.0 8.3 23.49 107.8
93.000 8,715 52.59 20.71 7.46 4.83 3.38 3.11 2.58 315.0 20.1 0.0 7.7 21.43 60.4
105.000 8,778 65.22 22.79 8.18 4.88 3.06 2.92 2.64 315.0 12.6 0.0 7.3 17.97 60.8
113.000 8,723 66.51 23.94 7.89 4.72 3.37 2.94 2.41 315.0 11.9 0.0 7.1 18.56 55.4
122.000 8,766 72.04 26.05 8.15 4.83 3.21 3.13 2.71 315.0 10.3 0.0 6.8 17.50 54.8
131.000 8,782 73.23 30.87 9.30 5.08 3.61 3.06 2.67 315.0 11.6 0.0 6.0 17.87 56.0
141.000 8,727 71.81 29.04 8.43 4.80 3.33 2.91 2.42 315.0 11.0 0.0 6.4 17.89 56.4
152.000 8,822 58.42 25.29 8.65 5.00 3.37 3.00 2.52 315.0 18.5 0.0 6.7 16.70 57.5
161.000 8,917 30.37 16.17 7.84 5.09 3.44 2.96 2.51 315.0 98.6 0.0 8.1 20.65 300.0
174.000 8,774 53.16 22.60 9.20 5.56 3.66 3.52 2.89 315.0 23.8 0.0 6.8 19.61 90.6
181.000 8,854 51.26 22.17 9.50 5.71 3.83 3.39 2.80 315.0 27.5 0.0 6.8 19.84 125.7
192.000 8,806 52.60 22.20 9.32 5.80 3.89 3.28 2.77 315.0 25.1 0.0 6.8 20.53 107.6
201.000 8,747 47.15 19.46 8.28 5.33 3.72 3.31 2.81 315.0 28.2 0.0 7.6 22.91 115.6
213.000 8,747 73.94 29.82 11.15 6.50 4.60 3.73 2.89 315.0 13.8 0.0 5.3 18.03 68.1
220.000 8,794 57.02 25.67 9.91 6.12 4.03 3.44 2.84 315.0 22.8 0.0 6.1 18.52 73.2
230.000 8,917 32.46 16.19 8.69 5.82 3.95 3.38 2.75 315.0 101.3 0.0 7.4 23.09 250.7
241.000 8,814 46.13 19.88 9.75 6.48 4.22 3.77 3.16 315.0 39.5 0.0 6.7 24.12 300.0
250.000 8,790 50.07 22.08 10.76 7.01 4.87 4.10 3.26 315.0 37.4 0.0 6.0 23.84 300.0
261.000 8,862 36.33 18.64 9.61 6.41 4.30 3.65 2.97 315.0 84.9 0.0 6.6 21.82 232.0
271.000 9,005 31.27 18.16 9.90 6.51 4.25 3.45 2.70 315.0 147.1 0.0 6.4 18.06 189.6
281.000 8,953 35.02 17.66 9.41 6.19 4.05 3.18 2.58 315.0 89.8 0.0 6.9 21.03 190.8
293.000 8,627 54.03 23.77 10.50 6.54 4.29 3.58 2.92 315.0 27.4 0.0 6.0 20.45 167.4
304.000 8,790 46.27 20.46 9.90 6.65 4.44 3.69 3.00 315.0 41.1 0.0 6.5 23.94 300.0
----------------------------------------------------------------------------------------------------------------------------------------
Mean: 52.63 21.79 8.88 5.64 3.84 3.36 2.77 315.0 36.1 0.0 7.0 21.24 96.8
Std. Dev: 12.81 3.92 1.08 0.71 0.47 0.34 0.23 0.0 32.9 0.0 0.8 2.69 48.1
Var Coeff(%): 24.34 18.01 12.17 12.61 12.27 10.07 8.19 0.0 91.1 0.0 12.0 12.69 50.6
----------------------------------------------------------------------------------------------------------------------------------------

Page 1
 FM1318 Departure.asc
----------------------------------------------------------------------------------------------------------------------------------------
TTI MODULUS ANALYSIS SYSTEM (SUMMARY REPORT) (Version 6.0)
----------------------------------------------------------------------------------------------------------------------------------------
District:5 (Lubbock) MODULI RANGE(psi)
County :219 (SWISHER) Thickness(in) Minimum Maximum Poisson Ratio Values
Highway/Road: FM1318 Pavement: 1.50 267,000 267,000 H1: v = 0.38
Base: 6.00 10,000 150,000 H2: v = 0.35
Subbase: 0.00 H3: v = 0.00
Subgrade: 189.30(by DB) 15,000 H4: v = 0.40
----------------------------------------------------------------------------------------------------------------------------------------
Load Measured Deflection (mils): Calculated Moduli values (ksi): Absolute Dpth to
Station (lbs) R1 R2 R3 R4 R5 R6 R7 SURF(E1) BASE(E2) SUBB(E3) SUBG(E4) ERR/Sens Bedrock
----------------------------------------------------------------------------------------------------------------------------------------
0.000 8,870 40.61 20.34 9.04 5.56 3.55 3.11 2.59 267.0 37.2 0.0 9.2 6.00 172.8
12.000 9,013 23.03 11.63 6.33 4.66 3.43 3.11 2.62 267.0 122.5 0.0 13.2 17.27 300.0
23.000 9,100 23.77 13.28 7.85 5.61 4.06 3.48 2.79 267.0 150.0 0.0 11.0 15.16 300.0 *
32.000 9,084 22.26 12.14 7.08 5.09 3.68 3.21 2.59 267.0 150.0 0.0 12.3 15.77 300.0 *
43.000 8,949 31.32 15.38 7.75 5.33 3.78 3.31 2.79 267.0 64.8 0.0 10.8 12.85 300.0
52.000 8,957 34.63 17.18 7.85 5.07 3.53 3.13 2.59 267.0 48.7 0.0 10.5 8.52 227.7
60.000 8,965 28.74 15.31 7.80 5.25 3.79 3.37 2.67 267.0 85.5 0.0 10.7 12.05 300.0
70.000 8,909 35.09 17.62 8.57 5.55 3.70 3.30 2.83 267.0 52.5 0.0 9.8 8.58 217.2
82.000 8,941 36.25 19.93 10.27 6.34 4.07 3.44 2.63 267.0 61.8 0.0 8.5 5.25 175.3
92.000 8,854 41.26 23.46 11.27 6.67 4.19 3.62 2.89 267.0 48.0 0.0 7.6 6.10 160.5
101.000 8,937 39.80 21.62 9.41 5.64 3.88 3.47 2.93 267.0 42.8 0.0 8.8 8.29 142.9
110.000 8,894 42.55 23.45 10.85 5.98 4.27 3.56 2.88 267.0 41.3 0.0 7.9 7.25 125.7
121.000 8,933 39.60 22.10 9.03 5.16 3.48 3.00 2.51 267.0 39.9 0.0 9.1 9.22 92.8
133.000 9,033 30.34 14.90 6.72 4.27 2.95 2.57 2.26 267.0 54.6 0.0 12.5 7.45 190.5
143.000 9,064 24.02 13.32 6.59 4.33 2.93 2.70 2.31 267.0 105.3 0.0 13.0 9.13 232.3
155.000 8,977 23.61 13.69 7.75 5.17 3.41 2.87 2.37 267.0 150.0 0.0 11.1 9.34 194.2 *
167.000 9,052 20.91 13.22 7.88 5.26 3.36 2.92 2.45 267.0 150.0 0.0 11.8 11.53 158.6 *
174.000 8,969 32.07 18.43 9.33 5.54 3.45 2.94 2.49 267.0 72.1 0.0 9.5 4.92 147.1
181.000 8,997 30.73 17.72 8.53 5.16 3.51 2.85 2.34 267.0 74.0 0.0 10.1 6.20 242.4
190.000 9,029 32.49 17.04 7.48 4.41 2.90 2.52 2.13 267.0 51.2 0.0 11.4 6.96 149.2
200.000 8,993 29.10 13.42 6.10 3.70 2.44 2.19 1.89 267.0 50.2 0.0 14.0 5.26 207.9
211.000 9,076 16.19 9.32 5.16 3.60 2.42 2.12 1.78 267.0 150.0 0.0 18.6 14.54 209.7 *
222.000 9,096 20.01 10.17 5.22 3.43 2.44 2.14 1.88 267.0 117.2 0.0 16.9 10.95 300.0
232.000 8,913 36.55 17.27 6.72 4.03 2.68 2.38 2.15 267.0 33.4 0.0 11.7 8.74 73.2
243.000 9,060 24.19 13.22 6.71 4.56 3.14 2.83 2.43 267.0 110.0 0.0 12.7 11.10 300.0
251.000 9,033 22.90 15.08 8.52 5.80 3.63 2.99 2.34 267.0 150.0 0.0 10.5 9.12 143.4 *
263.000 9,009 30.63 17.97 9.87 6.18 3.95 3.38 2.87 267.0 98.0 0.0 8.9 5.82 168.3
271.000 8,854 37.55 20.00 8.36 4.77 3.13 2.69 2.22 267.0 40.1 0.0 9.9 8.30 104.9
281.000 8,977 29.09 15.59 7.07 4.41 3.07 2.66 2.19 267.0 65.9 0.0 11.9 7.89 205.8
292.000 9,048 26.14 14.99 7.38 4.39 2.86 2.48 2.04 267.0 90.6 0.0 12.0 5.52 182.4
301.000 8,997 26.50 15.13 7.68 4.68 3.08 2.70 2.35 267.0 94.6 0.0 11.4 5.61 196.1
----------------------------------------------------------------------------------------------------------------------------------------
Mean: 30.06 16.26 7.94 5.02 3.38 2.94 2.45 267.0 83.9 0.0 11.2 9.05 196.8
Std. Dev: 7.06 3.68 1.48 0.80 0.52 0.43 0.32 0.0 41.0 0.0 2.4 3.36 78.2
Var Coeff(%): 23.49 22.66 18.69 15.92 15.29 14.56 12.93 0.0 48.8 0.0 21.2 37.07 40.7
----------------------------------------------------------------------------------------------------------------------------------------

Page 1
 APPENDIX H
MAXIMUM AND MINIMUM STRAINS FROM LOAD TESTING

TxDOT 0-5849 297

 APPENDIX TABLE 16. MAXIMUM AND MINIMUM STRAIN (με) FROM LIVE LOAD TEST ON SHALLOWATER CULVERT UNDER 2' OF FILL.
M1 - IWBEC M1 - IWEM M1 - IWTEC M1 - EWTEC M2 - ITEC M2 - ETEC M2 - ITEM M2 - ETEM O M2 - ITIC1 M2 - ETIC1
One Truck max 0.000 -1.896 -4.739 37.465 0.000 36.627 22.290 2.847 -13.281 781.525
Gage Line Beneath Wheel min -4.737 -10.425 -90.980 -1.423 -70.399 -0.951 -2.845 -27.047 -105.766 246.089
One Truck max 1.421 2.369 1.896 42.682 4.281 39.957 28.455 3.322 -4.269 563.280
Straddling Gage Line min -4.737 -10.425 -94.297 -2.371 -74.204 -3.805 -3.320 -39.383 -100.075 3.807
Two Trucks max -1.421 -4.265 -7.582 59.281 0.000 55.655 36.518 1.898 -21.345 6402.179
Straddling Gage Line min -11.370 -21.798 -136.464 -2.845 -103.692 -4.281 -7.114 -55.041 -147.971 410.375
M3 - IWBIC1 M3 - IWBIC2 M3 - IWIM1 M3 - IWIM2 M3 - IWTIC1 M3 - IWTIC2 M4 - IBEC M4 - IBEM M4 - IBIC1 M5 - ITIC2
One Truck max 13.754 5.203 0.000 -0.949 28.112 27.458 -0.947 -1.890 -2.368 -1.422
Gage Line Beneath Wheel min -20.394 -10.878 -9.463 -9.014 -70.989 -67.218 -7.575 -5.671 -6.629 -41.248
One Truck max 16.600 8.040 0.000 0.949 39.548 37.874 3.314 2.363 0.000 0.948
Straddling Gage Line min -21.816 -12.770 -9.936 -10.437 -61.937 -62.485 -6.628 -4.253 -6.156 -39.351
Two Trucks max 22.292 8.040 -0.473 -1.423 51.938 44.502 -2.367 -0.945 -2.841 -2.371
Straddling Gage Line min -38.415 -23.175 -18.452 -18.028 -91.474 -96.091 -16.097 -9.924 -11.838 -59.737
M5 - ETIC2 M5 - ITIM M5 - ETIM M5 - ITIC3 M5 - ETIC3 M6 - IWBIC3 M6 - IWBIC4 M6 - IWIM3 M6 - IWIM4 M6 - IWTIC3
One Truck max 29.888 40.230 3.785 -3.776 40.755 4.731 5.222 -1.421 -0.473 26.459
Gage Line Beneath Wheel min -1.898 -11.358 -32.647 -38.230 -1.422 -12.773 -9.494 -8.999 -11.341 -43.938
One Truck max 29.414 52.063 3.785 2.360 42.177 5.677 4.272 0.947 1.890 30.712
Straddling Gage Line min -2.372 -10.885 -42.582 -39.646 -5.213 -11.354 -10.918 -8.999 -10.869 -42.048
Two Trucks max 41.749 66.262 3.785 -6.136 58.764 5.677 8.545 -2.368 -3.308 41.107
Straddling Gage Line min -3.321 -21.297 -60.560 -60.883 -5.687 -18.923 -16.140 -16.577 -18.902 -64.725
M6 - IWTIC4 M7 - IBIC2 M7 - IBIM M7 - IBIC3
One Truck max 26.981 -3.784 -3.318 -1.897
Gage Line Beneath Wheel min -38.812 -12.770 -7.583 -12.804
One Truck max 30.768 0.000 1.896 0.948
Straddling Gage Line min -33.132 -12.297 -6.162 -11.855
Two Trucks max 40.235 -6.622 -5.214 -3.320
Straddling Gage Line min -48.751 -26.486 -16.115 -23.236

TxDOT 0-5849 298

 APPENDIX TABLE 17. MAXIMUM AND MINIMUM STRAIN (με) FROM LIVE LOAD TEST ON SHALLOWATER CULVERT UNDER 4' OF FILL.
M1 - IWBEC M1 - IWEM M1 - IWTEC M1 - EWTEC M2 - ITEC M2 - ETEC M2 - ITEM M2 - ETEM O M2 - ITIC1 M2 - ETIC1
One Truck max 2.369 2.369 0.000 20.393 0.476 20.931 12.805 1.898 -1.423 555.181
Gage Line Beneath Wheel min -4.737 -6.634 -47.388 0.000 -29.964 0.476 -0.474 -14.234 -43.153 0.000
One Truck max 2.842 6.160 6.633 25.135 4.281 23.785 16.599 3.321 6.165 0.000
Straddling Gage Line min -3.316 -6.160 -45.484 -0.948 -32.818 -0.951 -0.948 -18.504 -40.307 0.000
Two Trucks max 3.316 4.738 -1.421 36.518 0.476 34.250 23.239 4.270 -4.742 0.000
Straddling Gage Line min -4.737 -7.581 -66.329 0.948 -45.184 1.903 0.474 -25.147 -68.758 0.000
M3 - IWBIC1 M3 - IWBIC2 M3 - IWIM1 M3 - IWIM2 M3 - IWTIC1 M3 - IWTIC2 M4 - IBEC M4 - IBEM M4 - IBIC1 M5 - ITIC2
One Truck max 8.537 5.202 1.419 2.372 12.864 11.835 3.788 2.836 0.000 0.000
Gage Line Beneath Wheel min -11.382 -5.202 -4.258 -6.167 -42.402 -36.449 -7.575 -4.253 -3.788 -23.229
One Truck max 13.280 6.621 3.785 2.846 21.917 17.989 6.155 4.726 1.421 4.741
Straddling Gage Line min -10.908 -7.094 -5.677 -4.744 -37.162 -36.449 -2.841 -0.473 -3.315 -22.755
Two Trucks max 16.125 9.459 3.312 1.423 30.494 21.776 5.681 4.253 0.000 -1.422
Straddling Gage Line min -20.867 -8.040 -7.097 -8.065 -49.072 -54.909 -3.314 -2.363 -7.103 -34.607
M5 - ETIC2 M5 - ITIM M5 - ETIM M5 - ITIC3 M5 - ETIC3 M6 - IWBIC3 M6 - IWBIC4 M6 - IWIM3 M6 - IWIM4 M6 - IWTIC3
One Truck max 17.554 27.925 3.785 0.472 27.014 4.731 5.222 1.421 3.780 14.647
Gage Line Beneath Wheel min 0.000 -5.679 -16.086 -22.181 0.000 -5.204 -3.323 -4.736 -5.670 -24.568
One Truck max 17.554 34.551 6.624 3.776 27.962 6.150 6.171 3.315 6.615 18.899
Straddling Gage Line min -2.372 -1.893 -19.870 -22.181 -2.843 -3.311 -3.798 -4.736 -3.780 -25.040
Two Trucks max 32.263 52.537 4.731 -1.416 200.506 7.096 11.867 3.789 6.143 27.876
Straddling Gage Line min 2.372 -0.947 -29.332 -47.664 2.370 -8.515 -2.373 -4.736 -4.253 -29.764
M6 - IWTIC4 M7 - IBIC2 M7 - IBIM M7 - IBIC3
One Truck max 15.620 1.892 4.266 1.423
Gage Line Beneath Wheel min -18.459 -5.676 -7.109 -6.639
One Truck max 18.460 4.730 4.740 3.319
Straddling Gage Line min -17.986 -5.676 0.000 -5.216
Two Trucks max 21.773 4.730 7.110 3.319
Straddling Gage Line min -30.292 -8.986 -0.474 -10.432

TxDOT 0-5849 299

 APPENDIX TABLE 18. STRAIN (με) FROM DEAD LOAD TEST ON SHALLOWATER CULVERT DO TO A CHANGE IN FILL DEPTH FROM 2’ TO 4'.
M1 - IWBEC M1 - IWEM M1 - IWTEC M1 - EWTEC M2 - ITEC M2 - ETEC M2 - ITEM M2 - ETEM O M2 - ITIC1 M2 - ETIC1 M3 - IWBIC1
Dead Load
Strain -14.685 -27.010 -77.711 18.969 -54.702 20.930 6.165 -34.638 -93.431 0.000 -18.971
Difference
M3 - IWBIC2 M3 - IWIM1 M3 - IWIM2 M3 - IWTIC1 M3 - IWTIC2 M4 - IBEC M4 - IBEM M4 - IBIC1 M5 - ITIC2 M5 - ETIC2 M5 - ITIM
Dead Load
Strain -17.026 -15.613 -13.284 -10.482 -22.723 -4.261 3.308 -8.997 -45.988 32.261 19.404
Difference
M5 - ETIM M5 - ITIC3 M5 - ETIC3 M6 - IWBIC3 M6 - IWBIC4 M6 - IWIM3 M6 - IWIM4 M6 - IWTIC3 M6 - IWTIC4 M7 - IBIC2 M7 - IBIM M7 - IBIC3
Dead Load
Strain -39.270 -46.252 37.438 -14.192 -12.342 -10.893 -21.264 -5.670 -11.833 -19.392 4.266 -14.700
Difference

TxDOT 0-5849 300

 APPENDIX TABLE 19. MAXIMUM AND MINIMUM STRAIN (με) FROM LIVE LOAD TEST ON PLAINVIEW CULVERT UNDER 3.5' OF FILL.
M1 - IWBEC M1 - IWEM M1 - IWTEC M2 - ITEC M2 - ITEM M2 - ITIC1 M2 - ETIC1 M3 - IWBIC1 M3 - IWBIC2 M3 - IWIM1

One Truck max 2.851 4.742 0.947 8.593 77.860 1.425 0.000 10.899 7.099 2.365
Gage Line Beneath Wheel min -4.751 -0.948 -94.659 -1.432 -5.697 -22.328 0.000 -1.895 -2.840 -4.256
One Truck max 1.425 2.371 -0.473 7.161 84.507 0.475 0.000 8.055 4.733 0.000
Straddling Gage Line min -6.651 -4.742 -97.025 -2.864 -6.646 -23.278 0.000 -8.529 -6.152 -7.094

Two Trucks max 1.900 4.268 2.367 7.638 128.191 0.950 0.000 8.529 5.679 1.419
Straddling Gage Line min -11.402 -6.639 -127.312 -5.251 -10.918 -30.879 0.000 -14.689 -9.938 -10.404
M3 - IWIM2 M3 - IWTIC1 M3 - IWTIC2 M4 - IBEC M4 - IBEM M4 - IBIC1 M5 - ITIC2 M5 - ITIM M5 - ITIC3 M6 - IWBIC3

One Truck max 3.800 16.100 10.424 9.485 4.270 -3.795 0.945 26.482 3.791 5.709
Gage Line Beneath Wheel min -1.900 -24.623 -10.897 5.691 0.000 -13.283 -16.058 6.148 -10.426 -5.709

One Truck max 2.375 18.941 9.476 4.268 3.321 -0.474 0.472 26.482 4.265 2.379
Straddling Gage Line min -4.275 -22.729 -13.266 -1.423 -1.423 -12.808 -17.475 -1.892 -12.322 -9.039

Two Trucks max 2.375 26.992 12.319 5.691 4.744 -0.474 -0.472 27.901 2.844 7.136
Straddling Gage Line min -7.600 -32.672 -22.268 -6.165 -1.898 -17.078 -22.198 -6.148 -18.956 -13.796
M6 - IWBIC4 M6 - IWIM3 M6 - IWIM4 M6 - IWTIC3 M6 - IWTIC4 M7 - IBIC2 M7 - IBIM M7 - IBIC3

One Truck max -2019.493 3.320 9.004 33.152 23.649 -0.474 4.265 5.252
Gage Line Beneath Wheel min -2044.761 -3.320 0.948 -43.095 -27.431 -11.379 1.422 -6.207
One Truck max -14.095 4.269 8.056 38.836 23.649 -0.948 2.843 4.774
Straddling Gage Line min -2032.595 -4.269 -0.948 -42.621 -29.323 -15.646 -1.422 -9.548

Two Trucks max -23.022 3.795 8.530 55.887 29.325 -0.474 4.739 4.774
Straddling Gage Line min -2030.723 -9.487 -9.004 -56.827 -43.984 -19.439 -2.370 -18.142

TxDOT 0-5849 301

 APPENDIX TABLE 20. MAXIMUM AND MINIMUM STRAIN (με) FROM LIVE LOAD TEST ON TULIA CULVERT UNDER 1.5' OF FILL.
M1 - IWBEC M1 - IWEM M1 - IWTEC M1 - EWTEC M2 - ITEC M2 - ETEC M2 - ITEM M2 - ETEM M2 - ITIC1 M2 - ETIC1

One Truck max 3.337 0.475 13.760 30.372 10.898 35.146 47.071 98.382 1.896 278.825
Gage Line Beneath Wheel min -10.963 -9.502 -235.761 -87.782 -7.581 -60.787 -7.607 4.277 -30.339 -98.931
One Truck max 4.767 0.000 12.811 30.846 5.212 9.974 45.645 69.388 2.844 30.920
Straddling Gage Line min -10.963 -9.502 -203.511 -6.169 -6.634 -71.709 -8.082 -15.682 -21.332 -139.830

Two Trucks max 4.290 -1.900 35.587 26.575 11.372 -9.024 49.924 99.333 1.422 196.018
Straddling Gage Line min -23.832 -15.202 -256.154 -91.578 -8.055 -160.974 -10.460 -41.343 -26.072 -106.065
M3 - IWBIC1 M3 - IWBIC2 M3 - IWTIC1 M3 - IWTIC2 M10 - IBIC4 M10 - IBIM2 M4 - IBEC M4 - IBEM M4 - IBIC1 M8 - ITIC4

One Truck max 4.264 8.082 31.380 4.731 3.318 0.000 3.832 6.630 -0.474 8.978
Gage Line Beneath Wheel min -16.580 -5.705 -59.426 -19.398 -9.005 -1.892 -8.143 -1.421 -16.101 -22.209

One Truck max 3.790 9.984 46.595 5.678 3.318 0.946 6.227 7.577 0.000 3.308
Straddling Gage Line min -19.422 -5.229 -43.738 -16.087 -9.479 -1.419 -8.143 -2.368 -17.522 -18.429

Two Trucks max 9.948 15.213 67.992 6.151 5.213 0.946 4.790 13.260 -0.947 8.033
Straddling Gage Line min -28.896 -8.557 -58.951 -25.549 -16.114 -2.365 -18.680 -2.368 -29.361 -19.374
M5 - ITIM M5 - ITIC3 M6 - IWBIC3 M6 - IWBIC4 M8 - ITIM2 M6 - IWTIC3 M6 - IWTIC4 M7 - IBIC2 M7 - IBIM M7 - IBIC3

One Truck max 39.817 1.421 4.259 4.262 52.880 2.846 9.945 1.897 -1.896 0.949
Gage Line Beneath Wheel min -8.532 -30.305 -5.679 -7.104 -8.498 -20.396 -24.150 -15.177 -7.109 -5.222

One Truck max 35.551 2.368 5.679 6.630 38.243 3.320 9.471 3.794 -0.948 1.424
Straddling Gage Line min -9.006 -20.835 -5.679 -7.577 -9.442 -17.075 -22.256 -14.703 -6.161 -6.171
Two Trucks max 44.083 2.841 10.885 8.998 51.464 3.320 8.998 4.743 -1.422 3.798
Straddling Gage Line min -13.746 -26.044 -8.992 -9.472 -14.636 -23.716 -34.094 -26.560 -9.953 -8.070

TxDOT 0-5849 302

TxDOT 303