Manual de Diseño de Cimentaciones de Puentes PDF
Manual de Diseño de Cimentaciones de Puentes PDF
Manual de Diseño de Cimentaciones de Puentes PDF
Department
of Transportation
Federal Highway
Administration
))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))))
)))))))))))))))
PREFACE
For many years, engineers have designed foundations, walls and culverts for highway and other
applications using allowable stress design (ASD) methods. In ASD, all uncertainty in loads and
material resistance is combined in a factor of safety or allowable stress. For most highway
design, the American Association of State Highway and Transportation Officials (AASHTO)
Standard Specifications for Highway Bridges represents the primary source document for ASD
of substructures. In 1994, AASHTO approved Load and Resistance Factor Design (LRFD) in
the LRFD Highway Bridge Design Specifications. In LRFD, the uncertainty in load is
represented by a load factor and the uncertainty in material resistance is represented by a
resistance factor. Due to the fundamental differences between the substructure design process by
ASD and LRFD, this course has been developed to present the fundamentals of LRFD for the
geotechnical design of highway bridge substructures. Because this document was prepared for
engineers and others who are already familiar with the design of substructures using ASD, it is
intended for use in conjunction with other documents describing standard geotechnical design
procedures.
The objectives of this reference manual are to provide the basis for an understanding of the:
The authors also wish to acknowledge contributions of the following members of the Technical
Working Group listed in alphabetical order:
Lastly, the authors wish to thank the following employees from DAppolonia for their vital
contributions and significant effort in preparing this document: Ms. Tara Heyer, Mr. Robert
Whitaker, Mr. David Mudry, Mr. Christopher Lewis, Mr. James Wang and Dr. Yasser A.
Hegazy.
ii
TABLE OF CONTENTS
LIST OF TABLES
LIST OF FIGURES
LIST OF SYMBOLS
xii
xix
xxiii
CHAPTER 1 - INTRODUCTION
1.1 Manual Objectives .............................................................................................................. 1-1
1.2 Manual Outline ................................................................................................................... 1-2
1.3 General References ............................................................................................................. 1-3
CHAPTER 2 - TRANSITION TO LRFD FOR SUSBSTRUCTURE DESIGN
2.1 Introduction......................................................................................................................... 2-1
2.2 Allowable Stress Design ..................................................................................................... 2-1
2.3 Load and Resistance Factor Design.................................................................................... 2-2
2.4 LRFD Calibration ............................................................................................................... 2-8
2.5 Summary ............................................................................................................................. 2-9
CHAPTER 3 PRINCIPLES OF LIMIT STATES DESIGN
3.1 Introduction......................................................................................................................... 3-1
3.2 Design Procedures .............................................................................................................. 3-2
3.2.1 Allowable Stress Design (ASD) ............................................................................ 3-2
3.2.2 Shortcomings of ASD ............................................................................................. 3-3
3.2.3 Load and Resistance Factor Design (LRFD) .......................................................... 3-4
3.2.4 Advantages and Limitations of LRFD.................................................................... 3-5
3.3 Calibration........................................................................................................................... 3-6
3.4 Review of Statistics and Probability Concepts ................................................................. 3-10
3.4.1 Statistical Descriptors ........................................................................................... 3-10
3.4.2 Probability Density Functions............................................................................... 3-12
3.4.3 Probability of Failure ............................................................................................ 3-16
3.4.4 Reliability Index, $ ............................................................................................... 3-17
3.4.5 Resistance Statistics .............................................................................................. 3-20
3.4.6 Load Statistics....................................................................................................... 3-23
3.4.7 Calibration by a Combined Approach .................................................................. 3-29
3.5 Student Exercise................................................................................................................ 3-31
CHAPTER 4 - LOADS
4.1 Introduction......................................................................................................................... 4-1
4.2 Load Considerations for Geotechnical Design of Substructures ........................................ 4-2
4.2.1 Permanent Loads..................................................................................................... 4-3
4.2.1.1 Dead Loads - DC, DW and EV .............................................................. 4-3
4.2.1.2 Downdrag Load - DD............................................................................. 4-3
4.2.1.3 Lateral Earth Pressure Load- EH............................................................ 4-4
4.2.1.4 Earth Surcharge Load - ES..................................................................... 4-8
4.2.2 Transient Loads....................................................................................................... 4-9
iii
TABLE OF CONTENTS
(continued)
4.3
4.4
4.5
4.6
4.7
4.8
iv
TABLE OF CONTENTS
(continued)
5.5 Local Geologic Problem Conditions................................................................................. 5-21
5.6 Example Problem: Develop Site Specific Resistance Factors from Field and
Laboratory Test Data ........................................................................................................ 5-23
CHAPTER 6 - GEOTECHNICAL DESIGN PARAMETER SELECTION
6.1 Introduction......................................................................................................................... 6-1
6.2 Test Method Selection ........................................................................................................ 6-1
6.2.1 Field Test Methods ................................................................................................. 6-2
6.2.1.1 Shear Strength.......................................................................................... 6-2
6.2.1.2 Compressibility ........................................................................................ 6-4
6.2.2 Laboratory Tests ..................................................................................................... 6-5
6.2.2.1 Shear Strength.......................................................................................... 6-5
6.2.2.2 Compressibility ........................................................................................ 6-5
6.2.2.3 Index Tests ............................................................................................... 6-5
6.2.3 Parameter Evaluation .............................................................................................. 6-6
6.3 Reliability of Tests for Estimating Design Parameters....................................................... 6-7
6.3.1 Field Test Methods ................................................................................................. 6-7
6.3.1.1 Shear Strength.......................................................................................... 6-7
6.3.1.1.1 Effective Stress Strength........................................................ 6-7
6.3.1.1.2 Total Stress Strength .............................................................. 6-9
6.3.1.2 Compressibility ...................................................................................... 6-11
6.3.2 Laboratory Tests ................................................................................................... 6-14
6.4 Design Parameter Selection and Resistance Factors ........................................................ 6-16
6.5 Example Problem: Characterization Planning .................................................................. 6-18
CHAPTER 7 - CALIBRATION AS PART OF THE DESIGN PROCESS
7.1 Introduction......................................................................................................................... 7-1
7.2 Reliability Theory or Fitting with ASD? ............................................................................ 7-1
7.3 Evaluating Resistance Factors Using Reliability Method .................................................. 7-2
7.3.1 Estimating Resistance Bias 8R and Coefficient of Variation COVR....................... 7-3
7.3.2 Example Problem - Estimating 8R and COVR for Friction Angle .......................... 7-5
7.3.3 Example Problem - Estimating 8R and COVR for Undrained Strength.................. 7-5
7.3.4 Values of Dead Load Bias 8QD and Coefficient of Variation COVQD .................... 7-5
7.3.5 Example Problem - Estimating 8QD and COVQD for a Bridge................................ 7-6
7.3.6 Values of Load Factors (P for Permanent (Dead) Loads........................................ 7-6
7.3.7 Values of QD/QL ...................................................................................................... 7-6
7.3.8 Values of $T ............................................................................................................ 7-8
7.3.9 Example Problem - Resistance Factors for the Nordlund Method ......................... 7-8
7.4 Simplified Reliability Method to Evaluate Resistance Factors ........................................ 7-10
7.4.1 Example Problem - Estimating Resistance Factors for the Nordlund Method..... 7-10
7.5 Fitting with ASD to Evaluate Resistance Factors............................................................. 7-11
7.5.1 Example Problem - Estimating Resistance Factors for the Nordlund Method..... 7-11
7.6 Summary and Closing Remarks........................................................................................ 7-12
TABLE OF CONTENTS
(continued)
7.7 Student Exercise................................................................................................................ 7-14
Appendix 7A - Derivation of Simplified Reliability Calibration of Resistance Factors ........... 7-15
Appendix 7B - Derivation for Fitting with ASD ....................................................................... 7-17
CHAPTER 8 - SPREAD FOOTING DESIGN
8.1 Introduction......................................................................................................................... 8-1
8.2 Design Methods .................................................................................................................. 8-1
8.2.1 ASD Summary ........................................................................................................ 8-1
8.2.2 LRFD Summary...................................................................................................... 8-5
8.2.2.1 Limit States .............................................................................................. 8-7
8.2.2.2 Resistance Factors.................................................................................... 8-8
8.2.2.3 Comparison of Spread Footing Design Using LRFD and ASD ............ 8-10
8.2.2.4 Modification of Resistance Factors ....................................................... 8-13
8.2.3 Summarized Comparison of ASD and LRFD ...................................................... 8-14
8.3 Performance Limits........................................................................................................... 8-14
8.3.1 Displacements and Tolerable Movement Criteria ................................................ 8-16
8.3.2 Bearing Resistance................................................................................................ 8-17
8.3.3 Sliding Resistance................................................................................................. 8-19
8.3.4 Load Eccentricity (Overturning)........................................................................... 8-19
8.3.5 Overall Stability .................................................................................................... 8-20
8.3.6 Structural Resistance............................................................................................. 8-21
8.3.7 Other Considerations ............................................................................................ 8-21
8.3.7.1 Footing Embedment............................................................................ 8-21
8.3.7.2 Buoyancy and Uplift ........................................................................... 8-21
8.4 Student Exercise: Bearing Resistance of Spread Footings on Sand ................................. 8-23
8.5 Student Problem: Footing Design on Soil Using ASD and LRFD................................... 8-25
CHAPTER 9 - DRIVEN PILE DESIGN
9.1 Introduction......................................................................................................................... 9-1
9.2 Design Methods .................................................................................................................. 9-1
9.2.1 ASD Summary ........................................................................................................ 9-1
9.2.2 LRFD Summary...................................................................................................... 9-4
9.2.2.1 Limit States .............................................................................................. 9-6
9.2.2.2 Resistance Factors.................................................................................... 9-7
9.2.2.3 Comparison of Driven Pile Design Using LRFD and ASD .................. 9-12
9.2.2.3.1 Geotechnical Design ............................................................ 9-12
9.2.2.3.2 Structural Design ................................................................. 9-14
9.2.2.4 Modification of Resistance Factors ....................................................... 9-14
9.2.2.4.1 Geotechnical Design ............................................................ 9-14
9.2.2.4.2 Structural Design ................................................................. 9-15
9.2.3 Summarized Comparison of LRFD and ASD ...................................................... 9-16
9.3 Performance Limits........................................................................................................... 9-16
9.3.1 Displacements and Tolerable Movement Criteria ................................................ 9-18
vi
TABLE OF CONTENTS
(continued)
9.3.2
vii
TABLE OF CONTENTS
(continued)
11.2 Design Methods ................................................................................................................ 11-1
11.2.1 ASD Summary ................................................................................................... 11-1
11.2.2 LRFD Summary................................................................................................. 11-2
11.2.2.1 Limit States ......................................................................................... 11-2
11.2.2.2 Resistance Factors ............................................................................. 11-3
11.2.2.3 Comparison of Wall Design Using LRFD and ASD ........................ 11-5
11.2.2.4 Modification of Resistance Factors ................................................... 11-5
11.2.3 Summarized Comparison of ASD and LRFD ................................................... 11-5
11.3 Performance Limits........................................................................................................... 11-5
11.3.1 Displacements and Tolerable Movement Criteria ............................................. 11-6
11.3.2 Geotechnical Resistance .................................................................................... 11-7
11.3.3 Structural Resistance.......................................................................................... 11-7
11.3.4 Other Considerations ......................................................................................... 11-7
11.3.4.1 Loss of Passive Resistance ................................................................ 11-7
11.3.4.2 Drainage ............................................................................................ 11-7
11.4 Design Example: Cantilever Retaining Wall on Spread Footing by LRFD and ASD ..... 11-8
CHAPTER 12 - PREFABRICATED MODULAR WALL DESIGN
12.1 Introduction ..................................................................................................................... 12-1
12.2 Design Methods............................................................................................................... 12-1
12.2.1 ASD Summary ................................................................................................... 12-1
12.2.2 LRFD Summary................................................................................................. 12-2
12.2.2.1 Limit States........................................................................................ 12-2
12.2.2.2 Resistance Factors ............................................................................. 12-3
12.2.2.3 Comparison of Wall Design Using LRFD and ASD ........................ 12-3
12.2.2.4 Modification of Resistance Factors ................................................... 12-3
12.2.3 Summarized Comparison of ASD and LRFD ................................................... 12-4
12.3 Performance Limits ......................................................................................................... 12-4
12.3.1 Displacements and Tolerable Movement Criteria ............................................. 12-5
12.3.2 Geotechnical Resistance .................................................................................... 12-5
12.3.3 Structural Resistance.......................................................................................... 12-5
12.3.4 Other Considerations ......................................................................................... 12-6
12.3.4.1 Loss of Passive Resistance ................................................................ 12-6
12.3.4.2 Drainage ............................................................................................ 12-6
12.4 Design Example: Geotechnical Design of Prefabricated Modular Retaining Wall ........ 12-7
CHAPTER 13 - ANCHORED WALL DESIGN
13.1 Introduction ..................................................................................................................... 13-1
13.2 Design Methods............................................................................................................... 13-1
13.2.1 General Design Considerations.......................................................................... 13-1
13.2.2 ASD Summary ................................................................................................... 13-3
13.2.3 LRFD Summary................................................................................................. 13-5
13.2.3.1 Limit States........................................................................................ 13-7
viii
TABLE OF CONTENTS
(continued)
13.2.3.2 Resistance Factors ............................................................................. 13-8
13.2.3.3 Comparison of Anchored Wall Design Using LRFD and ASD........ 13-8
13.2.3.3.1 Geotechnical Design ................................................... 13-8
13.2.3.3.2 Structural Design ...................................................... 13-10
13.2.3.4 Modification of Resistance Factors ................................................. 13-11
13.2.3.4.1 Geotechnical Design ................................................. 13-11
13.2.3.4.2 Structural Design ...................................................... 13-13
13.2.4 Summarized Comparison of ASD and LRFD ................................................. 13-13
13.3 Performance Limits ....................................................................................................... 13-13
13.3.1 Displacements and Tolerable Movement Criteria ........................................... 13-13
13.3.2 Anchor Pullout ................................................................................................. 13-16
13.3.3 Bearing Resistance (Vertical Wall Elements).................................................. 13-17
13.3.4 Passive Resistance (Vertical Wall Elements) .................................................. 13-18
13.3.5 Overall Stability ............................................................................................... 13-18
13.3.6 Structural Resistance........................................................................................ 13-18
13.3.6.1 Vertical Wall Elements ................................................................... 13-18
13.3.6.2 Permanent Facing ............................................................................ 13-19
13.3.7 Other Considerations ....................................................................................... 13-20
13.3.7.1 Corrosion ......................................................................................... 13-20
13.3.7.2 Drainage .......................................................................................... 13-20
13.4 Student Problem: Anchored Soldier Pile Wall Design by LRFD ................................. 13-21
CHAPTER 14 - MECHANICALLY-STABILIZED EARTH (MSE) WALL DESIGN
14.1 Introduction ..................................................................................................................... 14-1
14.2 Design Methods............................................................................................................... 14-1
14.2.1 General Design Considerations.......................................................................... 14-1
14.2.2 ASD Summary ................................................................................................... 14-2
14.2.3 LRFD Summary................................................................................................. 14-6
14.2.3.1 Limit States ...................................................................................... 14-8
14.2.3.2 Resistance Factors............................................................................ 14-9
14.2.3.3 Comparison of Wall Design Using LRFD and ASD ..................... 14-12
14.2.3.4 Modification of Resistance Factors ............................................... 14-12
14.2.4 Summarized Comparison of ASD and LRFD ................................................. 14-13
14.3 Performance Limits ....................................................................................................... 14-14
14.3.1 Displacements and Tolerable Movements ....................................................... 14-14
14.3.2 Bearing Resistance........................................................................................... 14-15
14.3.3 Sliding Resistance............................................................................................ 14-15
14.3.4 Overall Stability ............................................................................................... 14-15
14.3.5 Pullout of Reinforcing Elements...................................................................... 14-15
14.3.6 Rupture of Reinforcing Elements .................................................................... 14-18
14.3.7 Horizontal Forces for Internal Stability Calculations ...................................... 14-18
14.3.8 Design Life....................................................................................................... 14-20
14.4 Design Example 1: MSE Wall Design by LRFD.......................................................... 14-22
ix
TABLE OF CONTENTS
(continued)
14.5 Design Example 2: Design of MSE Wall with Broken Backslope and
Traffic Surcharge by LRFD .......................................................................................... 14-39
CHAPTER 15 - FLEXIBLE CULVERT DESIGN
15.1 Introduction ..................................................................................................................... 15-1
15.2 Design Methods............................................................................................................... 15-2
15.2.1 ASD Summary ................................................................................................... 15-2
15.2.2 LRFD Summary................................................................................................. 15-4
15.2.2.1 Limit States ..................................................................................... 15-5
15.2.2.2 Loads and Load Factors ................................................................... 15-6
15.2.2.3 Resistance Factors............................................................................ 15-7
15.2.2.4 Comparison of Flexible Culvert Design Using ASD and LRFD..... 15-8
15.2.2.5 Modification of Resistance Factors ................................................. 15-9
15.2.3 Summarized Comparison of ASD and LRFD ................................................. 15-10
15.3 Performance Limits ....................................................................................................... 15-10
15.3.1 Displacements and Tolerable Movements ....................................................... 15-10
15.3.2 Wall Compression (Thrust).............................................................................. 15-13
15.3.3 Wall Buckling .................................................................................................. 15-15
15.3.4 Seam Strength .................................................................................................. 15-17
15.3.5 Handling Flexibility ......................................................................................... 15-18
15.3.6 Other Considerations ....................................................................................... 15-19
15.4 Design Example ............................................................................................................ 15-20
CHAPTER 16 - RIGID CULVERT DESIGN
16.1 Introduction ..................................................................................................................... 16-1
16.2 Design Methods............................................................................................................... 16-1
16.2.1 Indirect Design Method ..................................................................................... 16-2
16.2.2 Direct Design Method........................................................................................ 16-4
16.2.2.1 Limit States ...................................................................................... 16-7
16.2.2.2 Resistance Factors............................................................................ 16-7
16.2.2.3 Comparison of Rigid Culvert Design Using ASD and LRFD......... 16-8
16.3 Performance Limits ......................................................................................................... 16-8
16.3.1 Elastic Load Models .......................................................................................... 16-8
16.3.2 Flexure and Thrust ........................................................................................... 16-10
16.3.3 Shear ................................................................................................................ 16-11
16.3.4 Radial Tension ................................................................................................. 16-11
16.3.5 Crack Width ..................................................................................................... 16-11
CHAPTER 17 - SUMMARY
17.1 Summary and Overview.................................................................................................. 17-1
REFERENCES
APPENDIX A - METRIC STANDARDS AND UNIT CONVERSIONS
TABLE OF CONTENTS
(continued)
A.1 Introduction........................................................................................................................ A-1
A.2 Standard Metric Units and Conversions with English Units ............................................. A-1
A.2.1 Base Units .............................................................................................................. A-1
A.2.1.1 Length ..................................................................................................... A-1
A.2.1.2 Mass ........................................................................................................ A-1
A.2.1.3 Time ........................................................................................................ A-2
A.2.1.4 Temperature ............................................................................................ A-2
A.2.2 Derived Units ......................................................................................................... A-2
A.2.2.1 Force ....................................................................................................... A-2
A.2.2.2 Pressure ................................................................................................... A-2
A.2.2.3 Mass Density and Unit Weight ............................................................... A-2
A.2.2.4 Temperature ............................................................................................ A-2
A.2.3 Unit Prefixes .......................................................................................................... A-2
A.2.4 Summary of Standard and Preferred Metric Units ................................................ A-3
A.2.5 English to SI Unit Conversions ............................................................................. A-3
A.2.6 SI Units in Engineering Documents ...................................................................... A-5
A.2.6.1 Variable Symbols.................................................................................... A-5
A.2.6.2 Base Dimensions..................................................................................... A-6
A.2.6.3 Unit Symbols .......................................................................................... A-6
A.2.6.4 Combined Units and Symbols ................................................................ A-8
A.3 Examples.......................................................................................................................... A-10
APPENDIX B - SOLUTIONS TO STUDENT EXERCISES AND PROBLEMS
B.1 Chapter 3 Student Exercise ...............................................................................................B-1
B.2 Chapter 4 Student Exercise ...............................................................................................B-5
B.3 Chapter 4 Student Problem: Load Combinations for Retaining Wall..............................B-7
B.4 Chapter 7 Student Exercise .............................................................................................B-15
B.5 Chapter 8 Student Exercise: Bearing Resistance of Spread Footings on Sand ..............B-17
B.6 Chapter 8 Student Problem: Footing Design on Soil Using ASD and LRFD................B-21
B.7 Chapter 9 Student Problem: Comparison of Pile Designs by ASD and LRFD ..............B-33
B.8 Chapter 9 Student Exercise: Pile Capacity Evaluation by Nordlund Method.................B-45
B.9
Chapter 10 Student Problem: Development of Resistance Factor for Design of
Drilled Shafts in Sand
B-48
B.10 Chapter 13 Student Problem: Anchored Soldier Pile Wall Design by LRFD.................B-49
xi
LIST OF TABLES
Table
No.
2-1
3-1
3-2
3-3
3-4
3-5
3-6
3A-1
4-1
4-2
4-3
4-4
4-5
4-6
4-7
4-8
4-9
4-10
4-11
4-12
4-13
4-14
Title
Relationship Between pf and $ for Lognormal Distribution
Values of Resistance Factors Corresponding by Fitting for Different Values of
Safety Factor and Dead to Live Load Ratios for (D = 1.25 and (L = 1.75
Summary of Measured and Predicted Driven Pile Axial Soil Resistance Using
Meyerhof Procedure
Relationship Between Probability of Failure and Reliability Index for Lognormal
Distribution
Summary of Resistance Statistics
Statistics for Bridge Load Components
Resistance Factors for Axially Loaded Driven Piles in Sand, In-Situ Methods
Relationship Between Probability of Failure and Reliability Index for Normal
Distribution
Relative Movements Needed to Achieve Active of Passive Earth Pressure
Conditions
Equivalent Height of Soil for Vehicular Loading
Values of V0 and Z0 for Various Upstream Surface Conditions
Base Pressure, PB, for Va = 160 km/hr
Seismic Performance Zones
Site Coefficients
Classification of Soil Profile Types
Response Modification Factors for Substructure Design
Response Modification Factors for Connector Design
Load Combinations and Load Factors
Load Factors for Permanent Loads, (P
Minimum Soil Cover for Culverts
Multiple Presence Factors
Unfactored Vertical Loads and Resisting Moments
4-15
4-16
Load Factors
4-17
4-18
4-19
xii
LIST OF TABLES
(continued)
4-20
4-21
4-22
4-23
4-24
4-25
4-27
4-28
4-29
5-1
5-2
5-3
5-4
5-5
5-6
5-7
5-8
Summary of Inherent Soil Variability and Measurement Variability for Index Tests
5-9
5-11
5-12
5-13
Field and Laboratory Shear Strength Test Data for Example Problem
6-1
6-2
6-3
6-4
6-5
6-6
6-7
4-26
5-10
xiii
LIST OF TABLES
(continued)
7-1
7-2
7-3
7-4
Values of QD/QL
7-5
7-6
8-2
8-3
8-4
8-10
Suggested Minimum Factors of Safety for Overall Slope Stability from FHWA
Required Minimum Factors of Safety for Overall Slope Stability from AASHTO
ASD
Limit States for Design of Spread Footing Foundations
Resistance Factors for Semi-Empirical Evaluation of Bearing Capacity for Spread
Footings on Sand Using Reliability-Based Calibration
Resistance Factors for Geotechnical Strength Limit State for Shallow Foundations
Comparison of ASD Factor of Safety with LRFD Equivalent Factor of Safety for
Spread Footing Foundations
Unfactored and Factored Vertical Loads
8-11
8-12
8-13
8-14
8-15
8-16
8-17
8-18
9-2
9-3
Strength and Service Limit States for Design of Driven Pile Foundations
7-7
7-8
8-1
8-5
8-6
8-7
8-8
8-9
9-1
xiv
LIST OF TABLES
(continued)
9-4
9-5
9-6
Resistance Factors for Driven Piles for Estimating the Axial Geotechnical Pile
Capacity Using Reliability-Based Calibration
Resistance Factors for Geotechnical Strength Limit State for Axially Loaded Piles
9-8
9-9
9-10
Load Factors
9-11
9-12
9-13
9-14
9-15
10-1
10-9
11-1
11-2
12-1
Strength and Service Limit States for Design of Prefabricated Modular Walls
12-2
12-3
12-4
12-5
Load Factors
12-6
9-7
10-2
10-3
10-4
10-5
10-6
10-7
10-8
xv
LIST OF TABLES
(continued)
12-7
12-8
12-9
12-10
12-11
12-12
12-13
13-1
13-2
13-3
13-4
13-5
13-7
13-8
14-1
14-2
ASD Criteria for Internal Stability Evaluation of Reinforcements for MSE Walls
14-3
14-4
14-5
14-6
Relationship Between Joint Width and Limiting Distortion of MSE Wall Facing
14-7
14-8
14-9
14-10
14-11
14-12
14-13
14-14
14-15
14-16
13-6
xvi
LIST OF TABLES
(continued)
14-17
14-22
14-23
14-24
14-25
14-26
14-27
14-28
14-29
14-30
14-31
14-32
14-33
14-35
15-1
15-2
15-3
15-4
15-5
14-18
14-19
14-20
14-21
14-34
15-6
15-7
15-8
15-9
15-10
LIST OF TABLES
(continued)
16-1
16-2
16-3
16-4
A-1
A-2
A-3
xviii
LIST OF FIGURES
Figure
No.
2-1
2-2
2-3
2-4
2-5
2-6
2-7
2-8
3-1
3-2
3-3
3-4
3-5
3-6
3-7
3-8
4-1
4-2
4-3
4-4
4-5
4-6
4-7
4-8
4-9
4-10
4-11
4-12
4-13
Title
Factor of Safety for ASD
LRFD Design Approach
Variation of Load and Resistance
Variation of Load and Resistance
Reliability Index, $
Empirical Rates of Failure for Civil Works Facilities
Example Condition for Service Limit State Evaluation
Example Condition for Strength Limit State Evaluation
Histogram of Measured to Predicted Axial Driven Pile Resistance for Example
Problem 3.2
Lognormal Probability Density Function
Standard Normal Density Function
Lognormal Density Function
Probability Density Functions for Normally Distributed Load and Resistance
Definition of Reliability Index, $ for Lognormal Distributions of
R and Q
Reliability Index for Meyerhofs SPT Method
Resistance Factors for Meyerhofs SPT Method for $T = 2.0 and 2.5
Typical Loading for Substructure Design
Downdrag Loading
Location of Resultant Lateral Earth Pressure
Earth Pressure Distribution for Anchored Wall Design
Earth Pressure Distribution for MSE Wall Design with Level Backfill Surface
Earth Pressure Distribution for MSE Wall Design with Sloping Backfill Surface
HS20 Truck Configuration
Horizontal Pressure on Wall Due to Uniformly-Loaded Strip Pressure
Earthquake Loads on a Bridge
Definition Sketch for Monobe-Okobe Analysis
Load Combination for Evaluation of Sliding Resistance of a Cantilever Retaining
Wall Supported on a Spread Footing
Load Combination for Evaluation of Bearing Resistance and Eccentricity of a
Cantilever Retaining Wall Supported on a Spread Footing
Schematic Representation of Factored Loads on Driven Piles Subjected to
Downdrag
xix
LIST OF FIGURES
(continued)
Figure No.
4-14
4-15
4-16
4-17
4-18
4-19
4-20
4-21
4-22
5-1
5-2
5-3
5-4
5-5
5-6
5-7
5-8
5-9
5-10
6-1
6-2
6-3
6-4
6-5
6-6
6-7
6-8
6-9
6-10
Title
Soil Blocks Mobilized by Arching Over Culverts
Design Truck Pressure Distribution From Elastic Theory
Comparison of AASHTO Live Load Pressures Through Earth Fills
Surface Loading Pattern for Design Truck and Design Tandem
Distributed Area at Depth for Design Truck and Design Tandem
Live-Load Distribution Through Soil for Design Truck
Live-Load Distribution Through Soil for Design Tandem
Schematic of Example Problem
Retaining Wall Area Designations for Weight of Concrete
Reliability Variation with Sample Size for Indirect Testing for Mobilized
Undrained Shear Strength
Optimization of Foundation and Exploration Cost
Illustrative Variation of Resistance Factor with Resistance Coefficient of
Variability
Split-Barrel Sampler
Electrical Friction-Cone Penetrometer Tip
Field Vane Geometries and Sizes
Menard Pressuremeter Equipment
Thin-Walled Tube Sampler
Double-Tube Core Barrel
Histograms of Intact Rock Strength Properties
Uncertainty in Soil Property Estimates
Comparison of Field and Laboratory Shearing Modes
Correlation of CPT Tip Resistance, SPT N-Value and Mean Particle Size
Correlation of CPT Tip Resistance, SPT N-Value and Fines Content
Correlation of Triaxial Compression Effective Stress Friction Angle with Cone
Penetration Tip Resistance
Field Vane Shear Correction Factor
Pressuremeter Modulus versus N Value for Cohesionless Soil
Pressuremeter Modulus versus N Value for Cohesive Soils
Correlation of CPT Tip Resistance with Constrained Modulus and Relative
Density for NC Sands
Correlation of CPT Resistance with Constrained Modulus and Relative Density
for OC Sands
xx
LIST OF FIGURES
(continued)
Figure No.
6-11
6-12
6-13
6-14
8-1
8-2
8-3
8-4
8-5
8-6
8-7
9-1
9-2
9-3
9-4
9-5
9-6
9-7
9-8
9-9
9-10
10-1
10-2
10-3
10-4
11-1
12-1
12-2
12-3
Title
Correlation Between Normalized Undrained Shear Strength and Liquidity Index
for NC Clays
Compression and Unload-Reload Indices versus Plasticity Index
Normalized Undrained Shear Strength Versus Undrained Strength Ratio from
CIUC and UU Tests
Example Problem, Site Exploration Plan for Western Approach
Circular Arc Slope Stability Failure Surface Showing Typical Soil Slice Forces
Generalized Flow Chart for Spread Footing Design by ASD and LRFD
Comparison of Resultant Force Location for Footing on Soil Using ASD and
LRFD
Comparison of Resultant Force Location for Footings on Rock Using ASD and
LRFD
Schematic of Example Problem
Reduced Footing Dimensions
Procedure for Estimating Sliding Resistance of Footings on Clay
Generalized Flow Chart for Driven Pile Design by LRFD and ASD
Pile Group Loading
Generalized Problem Geometry and SPT Design Envelope
Representative Pile and CPT Tip Resistance Profile Used in Example
Representative Pile and CPT Side Resistance Profile Used in Example
Pile Cap and Pile Layout
Axial Load, Pu, Summary
Strong Axis Maximum Bending Moment, Mux, Summary
Weak Axis Maximum Bending Moment, Muy, Summary
Pile Settlements
Generalized Flow Chart for Drilled Shaft Design by LRFD and ASD
Soil Profile for Example
Shaft Group Layout for Example
Preliminary Longitudinal Shaft Reinforcement for Example
Generalized Flow Chart for Conventional Retaining Wall and Abutment Design
by ASD and LRFD
Generalized Flow Chart for Modular Retaining Wall Design
Schematic of Example Problem
Retaining Wall Area Designation for Weight of Concrete and Soil
xxi
LIST OF FIGURES
(continued)
Figure No.
12-4
13-1
13-2
13-3
13-4
13-5
14-1
14-2
14-3
14-4
14-5
14-6
14-7
14-8
14-9
14-10
15-1
15-2
15-3
15-4
15-5
16-1
16-2
16-3
16-4
Title
Bearing Pads at Rear and Front of Modules
Anchored Wall Nomenclature
Generalized Flow Chart for Anchored Wall Design by ASD and LRFD
Settlement Profiles Behind Anchored Walls
Schematic of Example Problem
Lateral Earth Pressure Diagram for Soldier Pile Embedment into Cohesionless
Soil
Generalized Flow Chart for MSE Wall Design by ASD and LRFD
Pullout Factors for Inextensible Mesh and Grid Reinforcement
Determination of Failure Plane and Earth Pressure Coefficients for MSE Wall
with Inextensible Reinforcements
Problem Geometry
Pressure Diagram for Computation of quniform for Reinforcement Tensile
Resistance Evaluation of Layer 1
Pressure Diagram for Computation of quniform for Reinforcement Pullout
Resistance Evaluation of Layer 1
Problem Geometry
Earth Pressure Distribution for MSE Wall with Broken Back Backfill Surface
Geometry of Assumed Failure Surface for Example Problem
Pressure Diagram for Computation of quniform for Internal Stability Evaluation of
Layer 1
Generalized Flow Chart for Flexible Culvert Design by ASD and LRFD
Comparison of Flexible Culvert Deflection Modes in Loose and Dense Soils
Problem Definition
Pressure Distribution at Pipe Crown for Design Truck
Pressure Distribution at Pipe Crown for Design Tandem
D-Load Test Setup
Comparison of AASHTO Live Load Pressures Through Earth Fills
Paris (a) and Olander (b) Pressure Distributions
Heger Pressure Distribution with Arching Factors
xxii
LIST OF SYMBOLS
A
A
A
A
A
a
AN
aB
Ab
=
=
=
=
=
=
=
=
=
xxiii
LIST OF SYMBOLS
(continued)
COVQL =
COVR =
COVRA =
COVRN =
COV8 =
COV8i =
Cn
=
CR
=
CRF =
Csm
=
CT
=
Cur
=
CV
=
Cv
=
D
=
D
=
D
=
DC
DD
DE
DLP
DTP
Dr
DW
DWT
D10
D50
E
e
eB , eL
Ec
ecc
ED
EH
Em
emax
En
EPMT
ES
Es
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
=
EQ
Eu
=
=
xxiv
LIST OF SYMBOLS
(continued)
EV
F
Fb
FC
=
=
=
=
xxv
LIST OF SYMBOLS
(continued)
k
=
k
=
k
=
=
ka
Kae
=
KE
=
kh
=
kh
=
ko
=
Kpe
=
ks
=
kv
=
L
=
L
=
L
=
l
=
l = Le =
LN
=
R
=
LD
=
Leff = Le
LF
=
LI
=
LL
=
LS
=
M
=
M
=
m
=
Mdt
=
Mmax =
=
Ms
N
=
N
=
N
=
n
=
n
=
=
Nk
Np
=
OCR
P
p
p
p
PA
Pa
=
=
=
=
=
=
=
xxvi
LIST OF SYMBOLS
(continued)
pa
=
Lateral earth pressure (kPa)
pa
=
Atmospheric pressure (kPa)
Pae
=
Seismic active earth pressure (kN/m)
=
Allowable axial, tensile or flexural structural capacity (kN)
Pall
Pall(LRFD)
=
LRFD equivalent allowable tensile load (kN)
PB
=
Base wind pressure (kPa) (4.2.2.8)
=
Equivalent static vessel impact force (kN)
PB
PC
=
Probability of bridge collapse due to a collision with an aberrant vessel (dim)
PD
=
Design wind pressure (kPa)
pf
=
Probability of failure (dim)
Pfg
=
Nominal pullout resistance of steel grid reinforcement (kN)
Pfs
=
Nominal pullout capacity of ribbed or smooth steel reinforcing strips (kN)
PH
=
Horizontal load (kN)
Ph
=
Horizontal component of lateral earth pressure per unit length of wall (kN/m)
ph
=
Horizontal stress (kPa)
PI
=
Plasticity Index (dim)
PL
=
Pedestrian live load (kN)
PL
=
Factored crown pressure (kPa)
PN
=
Wind pressure normal to structure component (kPa)
Pn=Pult =
Ultimate (nominal) structural resistance (kN)
Pn
=
Ultimate anchor tensile resistance or GUTS (kN)(13.2.2)
=
Factored structural resistance of reinforcements, facing and connections (kN)
Pr
PP
=
Wind pressure parallel to structure (kPa)
Ppe
=
Seismic passive earth pressure (kN/m)
Pr
=
Factored structural resistance (kN)
ps
=
Probability of survival (dim)
=
Vertical component of lateral earth pressure resultant per unit wall length (kN/m)
Pv
Q
=
Load or design load (kN)
q
=
Surcharge pressure (kPa)
q
=
Average pressure of an assumed rectangular distribution (kPa)
func {overline Q}
=
Mean load (kN)
func {overline q}
=
Magnitude of uniform distribution of soil pressure (kPa)
Qa
=
Ultimate unit resistance between grout/soil/rock in anchor bond zone per unit
length (kN/m)
=
Allowable design load (kN)
Qall
=
Cone penetrometer tip resistance (kPa)
qc
QD
=
Dead load (kN)
Qep
=
Nominal passive resistance of foundation material (kN)
Qi
=
Force effect, stress or stress resultant (kN or kPa)
QL
=
Live load (kN)
Maximum unit bearing stress (kPa)
qmax =
qmin =
Minimum unit bearing stress (kPa)
Qn
=
Nominal sliding resistance of footing (kN)
=
Ultimate pile tip resistance (kN)
Qp
qp
=
Unit tip resistance (kPa)
=
Factored geotechnical resistance (kN)
QR
xxvii
LIST OF SYMBOLS
(continued)
qR
=
Factored unit bearing resistance (kPa)
Qs
=
Ultimate pile side resistance (kN)
qs
=
Uniform surcharge applied to upper surface of active earth wedge (kPa)
=
Vehicular live load surcharge (kPa)
qs
qs
=
Unit shear resistance (kPa)
qT
=
Corrected cone tip resistance (kPa)
=
Ultimate geotechnical resistance (kN)
Qult
qult
=
Ultimate unit bearing resistance (kPa)
quniform =
Rectangular distribution of soil pressure (kPa)
=
Nominal shear resistance between footing and foundation material (kN)
QJ
R
=
Resistance (kN)
R
=
Radius of curvature of traffic lane (m) (4.2.2.3)
R
=
Earthquake response modification factor (dim) (4.2.2.11)
R
=
ADTT reduction factor (dim) (4.5)
R
=
Schmidt Hammer Rebound (dim)
R
=
Radius of culvert cross section at neutral axis (m)
r
=
Radius of gyration of culvert wall (m/m)
r2
=
Coefficient of determination (dim)
func {overline R}
=
Mean resistance (kN)
Rm
=
Measured resistance (kN)
Rn
=
Nominal (ultimate) resistance (kN)
Rr
=
Factored resistance (kN or kPa)
S
=
Site coefficient for earthquake load evaluation (dim) (4.2.2.11)
S
=
Diameter or span of culvert (m) (4.5)
S
=
Soil spreading factor (dim) (4.5)
Sc
=
Estimated total consolidation settlement (mm)
S.D. =
Standard deviation
SE
=
Force effects due to settlement (kN)
SH
=
Force effects due to shrinkage (kN)
Si
=
Internal diameter of culvert (m)
SS
=
Seam strength (kN/m)
=
Undrained shear strength (kPa)
Su
T
=
Limit State tensile capacity from creep tests (kN)
T
=
Tangential force (kN)
t
=
Ice thickness (m)
TL
=
Factored thrust in culvert wall (kN/m)
Tn
=
Ultimate wide width tensile yield strength (kN)
To
=
Tensile Strength (kPa)
=
Serviceability State tensile capacity at which total strain is not expected to exceed
Tw
5% based on wide-width tensile test design life of structure (kN)
TR
=
Highest load level at which log time-creep-strain rate continues to decrease with
time within required lifetime without either brittle or ductile failure (kN)
T5
=
Tension level at 5% strain based on wide width tensile test (kN)
ti
=
Initial thickness of reinforcing (m)
=
Thickness of reinforcing after 100 years (m)
t100
Tm
=
Period of vibration in mth mode (sec)
xxviii
LIST OF SYMBOLS
(continued)
TG
=
Force effects due to temperature gradient deformation (kN)
TU
=
Force effects due to uniform temperature deformation (kN)
ubt
=
Pore Pressure measured behind cone tip (kPa)
v
=
Highway design speed (m/s)
V
=
Vessel impact velocity (m/s)
VAF =
Vertical Arching Factor (dim) (4.5)
VB
=
Base wind velocity of 160 km/hr
VDZ =
Design wind velocity at design elevation (km/hr)
Vest
=
Most likely value of property
Vmax =
Largest conceivable value of property
Vmin =
Lowest conceivable value of property
VP
=
Compression Wave Velocity (m/s)
V10
=
Wind velocity at 10 m above low ground or design water level (km/hr)
V0
=
Friction velocity (km/hr)
w
=
Width perpendicular to direction of travel of surface tire contact area (m)
w
=
Width of grid reinforcement mat (m)
wi
=
Initial width of reinforcing (m)
wL
=
Liquid limit (%)
wn
=
Natural water content (%)
wP
=
Plastic limit (%)
w100 =
Width of reinforcing after 100 years (m)
WA =
Water load and stream pressure (kN)
=
Width perpendicular to direction of travel for distributed area (m)
WD
WE
=
Total unfactored earth load per unit length (kN/m)
WL
=
Total unfactored live load per unit length (kN/m)
WT
=
Weight of soil slice in slope stability analysis (kN)
=
WE + WL (kN/m)
WT
WL =
Horizontal wind pressure effects on vehicles (kN)
WS
=
Wind load on structure (kN)
x
=
Height of section for vertical element being considered (m)
func {overline x}
=
Mean value of data
xi
=
Data set value
Xo
=
Location of resultant from toe of wall (m)
Z
=
Structure height above low ground or design water level (m)
Z
=
Depth below effective top of wall to reinforcement (m)
z
=
Depth from ground surface (m) (4.2.1.3)
=
Friction length of upstream fetch (m)
Zo
"
=
Inclination of pier nose to vertical (deg)
"
=
Constant (dim)
"
=
Adhesion (dim)
$
=
Reliability index,
func {overline g~/._g} (dim)
$
=
Slope of wall backface (deg)
$
=
Coefficient for determining downdrag force by $ method (dim)
$
=
Load combination coefficient (dim)
$T
=
Target reliability index (dim)
xxix
LIST OF SYMBOLS
(continued)
(
=
Load factor (dim)
(
=
Total unit weight (kN/m3)
func {overline (}
=
Average load factor (dim)
(D
=
Load factor applied to dead load (dim)
=
Dry unit weight (kN/m3)
(d
(EH
=
Load factor applied to horizontal earth pressure (dim)
(EV
=
Load factor applied to vertical earth pressure (dim)
=
Load factor applied to live surcharge load (dim)
(LS
(s
=
Soil density (kN/m3)
(i
=
Load factor (dim)
=
Load factor applied to live load (dim)
(L
(p
=
Load factor for permanent load (dim)
=
Unit weight of soil (kN/m3)
(s
(rs
=
Effective unit weight of soil (kN/m3)
)
=
Lateral movement needed to develop active or passive earth pressure (m)
*
=
Angle of friction between soil and wall (deg)
*i
=
Estimated displacement or differential displacement (mm)
*max =
Maximum tolerable settlement (mm)
*n
=
Tolerable or differential displacement or movement (mm)
)p
=
Constant horizontal earth pressure due to uniform surcharge (kPa)
)P
=
Constant horizontal earth pressure due to uniform surcharge (kPa)
)ph =
Horizontal earth pressure due to surcharge (kPa)
*tol
=
Tolerable displacement or differential displacement (mm)
.
=
Standard deviation of a lognormally distributed data set
2
=
Angle between wind direction and normal to structure component (deg.) (4.2.2.8)
2
=
Skew angle (deg) (4.4)
0
=
Load modifier to account for effects of ductility, redundancy and operational
importance (dim)
0
=
Efficiency factor for piles (dim)
=
Load modifier account for effects of ductility (dim)
0D
0I
=
Load modifier account for effects of operational importance (dim)
0R
=
Load modifier account for effects of redundancy (dim)
8
=
Bias factor (dim)
8
=
Combined bias factor (dim)
8i
=
Individual bias factor (dim)
Overall bias for dead load (kN)
8QD =
8QL
=
Overall bias for live load (kN)
=
Overall bias for resistance (kN)
8R
8RA
=
Bias factor for soil internal friction angle using Nordlund Method (dim)
=
Bias factor for pile capacity using Nordlund Method (dim)
8RN
=
Correction factor (dim)
<
=
Poissons Ratio (dim)
>m
=
Lognormal mean
D
=
Average elastic settlement (mm)
xxx
LIST OF SYMBOLS
(continued)
Dtol
=
F
=
Fall
=
Fall
=
Fall(LRFD)
Fg
=
FH
=
=
Fn
Fn
=
Fp
=
=
Fult
FV
=
=
Frv
Fvo
=
Frvo =
Jav
=
N
=
N
=
N
=
Nr
=
Nep
=
Nf
=
NNf
=
Nm
=
Nqp
=
Nqs
=
NR
=
Nrtc
=
NJ
=
Q
=
xxxi
CHAPTER 1
INTRODUCTION
1.1 Manual Objectives
In the US, the design of foundations, walls and culverts has traditionally been performed using
allowable stress design (ASD) in which all uncertainty in loads and material resistance is combined
in a factor of safety. For most highway engineering applications, a source document for ASD of
substructures has been Division I of the AASHTO "Standard Specifications for Highway Bridges"
(1996). In 1989, work began on an entirely new specification in which the uncertainty in load(s) is
represented by a load factor(s) which generally has a value greater than one, and the uncertainty in
material resistance(s) is represented by a resistance factor(s) which generally has a value less than
one. This effort involved a team of about 50 consultants and contractors under the direction of
Modjeski and Masters, Inc., and review by a project panel, AASHTO technical subcommittees, state
highway departments, transportation authorities, and industry representatives. The AASHTO Load
and Resistance Factor Design (LRFD) Specification was approved for use in 1994.
Due to the differences between the substructure design process by ASD and LRFD, FHWA
sponsored development of this training course to present the fundamentals of LRFD to bridge design
engineers, geotechnical engineers, engineering geologists and others who are responsible for design
of bridge substructures using the AASHTO LRFD Specification. This manual focuses on the
geotechnical aspects of substructure design.
The objectives of this manual are to provide the basis for an understanding of the:
y
Process for selection and application of load factors and load combinations
Although the design procedures differ between ASD and LRFD, the methods used to estimate
1-1
design loads and ultimate resistance remain essentially unchanged. Therefore, verification of the
quality of design and construction is equally important for ASD and LRFD.
1.2 Manual Outline
This manual consists of 17 chapters which describe the background, development and use of LRFD
for highway bridge substructure design, and which present worked example problems for the
majority of substructure types. The manual is intended as a reference document to supplement the
Participant=s Workbook prepared for a National Highway Institute (NHI) two-day short course on
the design of substructures using LRFD. It is hoped that this reference manual and associated short
course will facilitate the implementation of LRFD for highway bridge substructures and provide
both the basis and impetus for further development and refinement of the AASHTO and state
transportation department specifications for design by LRFD.
In addition to this Introduction and a Summary at the end of the manual, the content of the chapters
includes the following topics:
y
y
y
y
y
y
y
y
y
y
y
y
y
y
y
Chapter
Chapter
Chapter
Chapter
Chapter
Chapter
Chapter
Chapter
Chapter
Chapter
Chapter
Chapter
Chapter
Chapter
Chapter
Chapters 2, 3 and 4 focus on limit state design principles and development of loads and load
combinations needed for substructure design. Chapters 5 and 6 cover aspects of geotechnical site
characterization and selection of geotechnical parameters for design. Chapters 5 and 6, which
address aspects of material variability, are included in this manual because the development and
selection of soil and rock properties are an integral part of both Load and Resistance Factor Design
(LRFD) and Allowable Stress Design (ASD) methods for substructure design. Calibration methods
used to develop the AASHTO LRFD Specification (1997a) are presented in Chapter 7 as an
introduction to the design chapters and to provide a reference to agencies for future additions to the
specification. Chapters 8 through 16 cover the design of foundations, retaining walls and culverts by
LRFD. In addition, these chapters provide guidance for transforming current ASD methods to
LRFD. Chapters 8 through 16 are not intended to provided instruction in applying the LRFD
Specification, but instead highlight differences in substructure design between LRFD and ASD.
Each design chapter provides a brief discussion and comparison of the general elements of
1-2
substructure design by ASD and LRFD, and identifies the performance limits important for
substructure design (e.g., settlement, bearing capacity, sliding and overturning for spread footing
foundations). Application of LRFD principles is illustrated in each chapter by example problems.
In addition, several short student exercises are included to test your understanding of the information
presented in the manual.
1.3 General References
Copies of the following references are available for use during the class:
AASHTO, 1994, LRFD Bridge Design Specifications, American Association for Transportation and
Highway Officials, Washington, DC, 1st Ed.
AASHTO, 1997a, 1997 Interims to LRFD Highway Bridge Design Specifications, SI Units,
American Association of State Highway and Transportation Officials, Washington, D.C., First
Edition (1997 LRFD Interims)
AASHTO, 1996, Standard Specifications for Highway Bridges, American Association of State
Highway and Transportation Officials, Washington, D.C., 16th Edition (1996 ASD)
AASHTO, 1997b, 1997 Interims to Standard Specifications for Highway Bridges, American
Association of State Highway and Transportation Officials, Washington, D.C., 16th Edition (1997
ASD Interims)
Barker, R.M., J.M. Duncan, K.B. Rojiani, P.S.K. Ooi, C.K. Tan and S.G. Kim, 1991, Manuals for
the Design of Bridge Foundations, National Cooperative Highway Research Program Report 343,
Transportation Research Board, Washington, D.C., 308p.
Barker, et al. (1991) was published as a result of research conducted in the Department of Civil
Engineering at Virginia Polytechnic and State University in Blacksburg, Virginia. That work
formed for basis for development of Section 10 - Foundations and Section 11 - Abutments, Piers and
Walls found in the AASHTO LRFD Specification.
1-3
1-4
CHAPTER 2
TRANSITION TO LRFD FOR SUBSTRUCTURE DESIGN
2.1 Introduction
Highway substructures (i.e., foundations and abutments) have traditionally been designed using
Allowable Stress Design (ASD) methods whereas superstructure components have been designed
used Load Factor Design (LFD) methods. This application of ASD for substructure design and LFD
for superstructure design leads to uncertain and incompatible safety margins in the design of
structure components. Solution of this problem requires a coherent method of design for the
structure system. As described in subsequent chapters, Load and Resistance Factor Design (LRFD)
represents an approach in which applicable failure and serviceability conditions can be evaluated
considering the uncertainties associated with loads and material resistances. This chapter:
y
Identifies the limit states that must be evaluated for structure design
Existing practice for the geotechnical design of substructures follows the ASD approach, wherein all
uncertainty in the variation of applied loads transferred to the foundation and the ultimate
geotechnical capacity of the soil and rock to support the loads are incorporated in a factor of safety,
FS. The factor of safety is an empirical, but arbitrary, measure used to reduce the potential for
adverse performance (e.g., sliding failure of a footing and bearing failure of driven pile). The
general relationship used in applying ASD takes the general form:
Rn
Q
FS
(Eq. 2-1)
where:
Rn = Nominal (ultimate) resistance
FS = Factor of safety
Q = Summation of force effects
2-1
In practice, FS can range from values of about 1.2 to 6 depending on factors such as the type of
problem being evaluated, the model used to estimate resistance and the experience of the designer.
Graphically, the ASD process can be illustrated as shown in Figure 2-1.
Figure 2-1
Factor of Safety
Figure 2-1 illustrates one of the principal limitations of ASD, wherein the values of Q and Rn are
assumed to be unique such that they have a probability of occurrence of unity. In addition, selection
of FS is subjective, depends on the design models used and material parameters chosen, and is not
inherently related to the probability of component failure. For geotechnical engineering, a rational
design approach should consider the uncertainties associated with:
y
y
y
y
y
y
Variability of engineering properties with aerial and vertical extent, and with
time
Reliability and applicability of property measurements
Sufficiency and applicability of sampling and testing methods
Errors in prediction models used
Errors in measuring material resistance
Variability in load prediction estimates
Although many or all of these sources of variability are usually considered by the designer using
ASD, their consideration is generally qualitative rather than quantitative, leading to a wide range of
failure probabilities among designs.
2.3 Load and Resistance Factor Design
Load and resistance factor design represents a more rational approach by which the more significant
uncertainties listed above (i.e., load and material resistance) can be incorporated quantitatively into
the design process. As used in the AASHTO LRFD Specification (AASHTO, 1997a), the basic
LRFD equation or relationship is defined by:
2-2
i Qi R n
(Eq. 2-2)
where:
i =
Qi =
Rn =
=
Figure 2-2
LRFD Design Approach
When applying LRFD, the estimated magnitudes of the various types of load effects are multiplied
by appropriate load factors to determine the factored load effects, and the estimated nominal
(ultimate) resistance is multiplied by a resistance factor prescribed for the model used to estimate
material resistance and the field and/or laboratory test methods used to develop the material
properties. For simplicity, the nominal and factored loads and the nominal and factored resistances
are shown in Figure 2-2 as unique values. Of course, the load and material resistances vary, such as
shown in Figure 2-3.
Figure 2-3 shows a possible variation of load and resistance as a function of the frequency or
probability of occurrence. If the peak values of load, Q, and resistance, R, are defined by their
respective mean values, Q and R , the equivalent ASD factor of safety is the ratio of R to Q as
shown in the figure, and the margin of safety is the difference between R and Q . However, the
figure also points out that some potential for failure exists in the area where the distributions of Q
and R overlap. Therefore, unless very high factors of safety are used, some probability of failure
will always exist.
2-3
Figure 2-3
Variation of Load and Resistance
Figure 2-3 shows a possible variation of load and resistance as a function of the frequency or
probability of occurrence. If the peak values of load, Q, and resistance, R, are defined by their
respective mean values, Q and R , the equivalent ASD factor of safety is the ratio of R to Q as
shown in the figure, and the margin of safety is the difference between R and Q . However, the
figure also points out that some potential for failure exists in the area where the distributions of Q
and R overlap. Therefore, unless very high factors of safety are used, some probability of failure
will always exist.
Another factor which must be considered is the distribution of load and resistance. Figure 2-3
represents only one pair of distributions. Figure 2-4 shows that using the ratio of R to Q to define
safety can be misleading. Two pair of load and resistance distributions are presented which have
identical values of Q and R . The upper distribution for resistance is relatively small with steep
flanks about R , whereas the lower distribution for resistance is broad with flat flanks, also about R .
As a result, the area of overlap between the upper distributions of R and Q is small, representative
of a small probability of failure. Conversely, the area of overlap between the lower distributions is
large, representative of a greater probability of failure.
For substructure design, the majority of loads which must be supported are prescribed by the
structural or bridge designer in the form of vehicle and other types of transient load types. Thus,
with the exception of the type of soils used for backfill behind walls and abutments and around
culverts, geotechnical engineers have only limited control over the load side of the relationship.
However, on the resistance side, geotechnical engineers have the opportunity to control the extent
and type of sampling and testing used to characterize a site, the procedures or models used for
design, and the measures employed to monitor the construction processes.
2-4
Figure 2-4
Distribution of Load and Resistance
Reliability-based design (or LRFD) is the process by which the risks and uncertainties associated
with the safety of a system are defined in mathematical terms. In applying the process, the
uncertainties (or distributions) of load and resistance are assumed to be independent, random
variables, and the design risk is defined by the probability of failure, pf. To evaluate pf, a single
probability density function is developed, as shown in Figure 2-5, which represent the combined
distributions and uncertainties of Q and R.
The left side of Figure 2-5 shows typical distributions of load and resistance and the right side shows
the combined distribution function for Q and R. For the pair of distributions to the left, failure is
possible in the shaded area where the distributions overlap and the margin of safety is represented by
the difference between R and Q . For the combined distribution, failure is possible in the shaded
area to the left of the ordinate and the margin of safety is represented by the number of standard
deviations, , of the mean value of the combined distribution to the right of the ordinate. In LRFD
parlance, is referred to as the reliability index. Depending on the shape of original distributions
(i.e., both normal, both lognormal or one normal and one lognormal), the probability of failure can
be mathematically related to the reliability index. For lognormal distributions, the relationship
between pf and is presented in Table 2-1.
2-5
Figure 2-5
Reliability Index,
Table 2-1
Relationship Between pf and for Lognormal Distribution
Reliability
Index
1.96
2.50
3.03
3.57
4.10
4.64
Probability of
Failure
pf
1:10
1:100
1:1 000
1:10 000
1:100 000
1:1 000 000
To implement the LRFD concept in a design code, it is necessary to select a minimum level of safety
which must be achieved. One process is to conduct cost-benefit analyses to determine the total costs
(i.e., sum of initial construction cost, maintenance costs and estimated costs of failures) as a function
of the probability of failure. Using this approach, the target probability of failure would be the pf for
which costs are minimized. Alternatively, the target probability of failure could be established based
on the failure rates estimated from actual case histories. However, the probability of failure for
constructed facilities is not solely a function of uncertainties associated with the design process, and
probably is at least an order of magnitude lower than the theoretical probability of failure. For
general guidance, information such as that presented in Figure 2-6 regarding empirical rates of
failure for civil engineering facilities can be used. For structure foundations, the annual probability
of failure ranges from about 1:100 to 1:1000. In calibrating the LRFD Specification, values of pf
range approximately within these limits, depending on the foundation type and the level of
redundancy available in the system (e.g., pile foundations are usually constructed in a group such
that failure of a single pile does not imply failure of the group).
2-6
Figure 2-6
Empirical Rates of Failure for Civil Works Facilities
(Kulhawy, et al., 1995)
A limit state is a condition beyond which a structural component, such as a foundation or other
bridge component, ceases to fulfill the function for which it is designed. The limit states which must
be evaluated in the AASHTO LRFD Specification (AASHTO, 1997a) include:
y
y
y
y
The Service Limit State represents structure performance under service load conditions. Examples
for substructure design include settlement of a foundation or lateral displacement of a retaining wall.
Another example of a Service Limit State condition is presented in Figure 2-7 which shows the
rotation of a rocker bearing on an abutment caused by instability of the earth slope which supported
the abutment.
Strength Limit States involve the total or partial collapse of the structure. Examples of Strength
Limit States in geotechnical engineering include bearing capacity failure, sliding, and overall slope
instability such as shown in Figure 2-8.
2-7
Figure 2-7
Example Condition for Service Limit State Evaluation
Figure 2-8
Example Condition for Strength Limit State Evaluation
2.4 LRFD Calibration
Calibration of the load and resistance factors is required to achieve the desired results when applying
LRFD. Calibration procedures for selection of resistance factors are described fully in Chapters 3
and 7, but generally involve:
y
y
y
y
Engineering judgment
Fitting to ASD
Reliability theory
A combination of approaches
2-8
For the LRFD Specification, a combination of approaches was used in selecting resistance factors, ,
for design. In general, the resistance factors selected for most of the methods used for foundation
design were developed principally using reliability-based calibration procedures where sufficient
performance data were available (e.g., bearing resistance of footings and individual deep foundation
elements). The value of chosen for a particular design procedure and limit state from a reliabilitybased calibration can take into account the:
y
y
y
y
y
On the other hand, LRFD does present a challenge to the practicing engineer, in that:
y
Resistance factors vary with design methods and are not constant
Subsequent chapters will provide additional information regarding the development and
application of LRFD for substructure design.
2-9
2-10
CHAPTER 3
PRINCIPLES OF LIMIT STATES DESIGN
3.1 Introduction
In this chapter, comparisons between Allowable Stress Design (ASD) and Load and Resistance
Factor Design (LRFD) are presented. Emphasis is placed on reliability concepts on which LRFD is
based and how these concepts relate to a predefined limit state. Procedures are described for
calibrating and selecting resistance factors for geotechnical components which target an acceptable
probability of survival. Because resistance factors are based on statistics, a brief review of statistics
and probability concepts is presented.
Simply stated, safety in engineering design is assumed when the cross-sections and materials
supplied exceed the demands put on them by applied loads so that:
Supply Demand
Another way of stating this same principle is that the resistance of the materials must exceed the
effect of the loads, so that:
Resistance Effect of Loads
(Eq. 3-1)
When applying this simple principle to design, it is essential that both sides of the inequality be
evaluated for the same conditions. For example, if the effect of applied loads produces compressive
stress on a soil, it is obvious that the load should be compared to the bearing resistance of the soil,
and not some other aspect of the material (e.g., soil).
When a particular loading condition reaches its limit, failure will result. Such a condition is referred
to as a limit state, and is defined as:
A limit state is a condition beyond which a structural component, such as a foundation
or other bridge component, ceases to fulfill the function for which it is designed.
Strength Limit States involve the total or partial collapse of the structure. Examples of Strength
Limit States in geotechnical engineering include bearing capacity failure, sliding, and overall
instability.
Service Limit States affect the function of the structure under regular service conditions. Service
Limit States may be reached in foundations through excessive settlement, excessive lateral
deflection, structural deterioration of the foundation or excessive vibration.
In this definition of a limit state, both the resistance and load sides of Eq. 3-1 are included. For
example, if adequacy of bearing strength of a soil under a footing is being investigated, more than
one load combination must be evaluated, especially if the footing is subjected to eccentric or
3-1
inclined loads. When the bearing pressure due to the loads exceeds the bearing strength, a limit state
(i.e., a Strength Limit State) is reached and failure results. Similarly, if the footing movements due
to the loads exceeds the tolerable settlement, the Service Limit State is reached. In the chapters on
component design (i.e., Chapters 8 -16), the various resistance or deformation conditions (e.g.,
bearing capacity or settlement) investigated for each component are defined as Performance Limits.
An important goal, but not the only goal of the designer is to prevent a limit state from being
reached. Other goals that must be considered and balanced in the overall design are function,
appearance, and economy. Because it is not economical to design a bridge so that none of its
components could ever fail under the most improbable loads, an acceptable level of risk or
probability of failure must be determined.
The determination of an acceptable margin of safety (i.e., how much greater the predicted magnitude
of resistance should be compared to the magnitude of the loads) is not based on the opinion of one
individual. The acceptable margin of safety must combine the collective experience and judgment
of a qualified group of engineers and officials. Many public owners apply their experience and
judgment to their own unique methods of design and they need to know what level of risk their
design entails. This concern is addressed in Section 3.4 and is more fully developed in Chapter 7.
In the sections that follow, a review of current design procedures in geotechnical engineering will be
addressed. A discussion of the shortcomings of current design procedures will show why a new
approach, the LRFD method, is desirable.
3.2 Design Procedures
Over the years, design procedures have been developed by engineers to provide satisfactory margins
of safety. These procedures were based on the engineer's confidence in predicting the magnitude of
the load and the effect of the load on the strength of the materials being provided. In the next
section, the existing procedure for design of foundations is discussed.
3.2.1 Allowable Stress Design (ASD)
The design of foundations has traditionally been based on ASD. ASD is different for the Service
Limit State and the Strength Limit State. For the Strength Limit State, safety is achieved in the
foundation element by restricting the estimated loads (or stresses) to values less than the ultimate
resistance divided by a factor of safety, FS using the relationship:
Rn
Qi
FS
(Eq. 3-2)
where:
Rn =
Nominal resistance (e.g., ultimate bearing resistance of foundation soils)
Q i = Qn = Nominal load effect (e.g., moment produced by design vehicle)
Load effects consist of dead, live and environmental load components. Environmental loads include
wind, water and earthquake forces, for example. In ASD all of these loads are assumed to have the
3-2
same variability. As a result, load factors are not applied on the load combinations considered for
either the strength or service limit states.
The factor of safety is a number greater than unity. The FS provides reserve strength in the event
that an unusually high load occurs or in the event that the resistance is less than expected.
For the Service Limit State, unfactored loads are used to calculate deformations, and these
deformations are compared to the maximum tolerable values.
The advantage of ASD is its simplicity; however, there are shortcomings of ASD.
3.2.2 Shortcomings of ASD
In ASD, no consideration is given to the fact that various types of loads have different levels of
uncertainty. For example, the dead load of a bridge can be estimated with a high degree of accuracy.
However, earthquake loads acting on bridge abutments and piers cannot be estimated with the same
degree of accuracy and confidence. Nevertheless, dead, live, and environmental loads are all treated
equally in ASD. In ASD, fixed values of design loads are selected, usually from a specification or
design code. The factor of safety is applied to the resistance side of the design inequality, and the
load side of the inequality is not factored.
Factors of safety in geotechnical engineering vary considerably depending on the type of problem.
y
y
y
y
Because the factor of safety chosen is based on experience and judgment, quantitative measures of
risk cannot be determined for ASD.
Limitations of ASD:
Does not adequately account for variability of loads and resistances. The FS is applied only to
resistance. Loads are considered to be without variation (i.e., deterministic).
Does not embody a reasonable measure of strength, which is a more fundamental measure of
resistance than is allowable stress.
Considers variability not only in the resistance, but also in the effect of loads
Uses the strength of the material (e.g., soil and rock) as a basis of resistance
Provides a measure of safety related to probability of failure
3-3
where:
=
Rn =
i =
i =
Qi =
Because Eq. 3-3 involves both load factors and resistance factors, the design method is called Load
and Resistance Factor Design (LRFD). For a satisfactory design, the factored nominal resistance
should equal or exceed the sum of the factored load effects for a particular limit state. Load and
resistance factors are chosen so that in the highly improbable event that the nominal resistance of the
foundation material is overestimated, and at the same time the loads are underestimated, there is a
reasonably high probability that the actual resistance of the foundation material should still be large
to support the loads.
The value of chosen for a particular limit state can take into account the:
3-4
y
y
y
y
y
In the AASHTO LRFD Specifications (1997a), not all of these features have been implemented.
However, they all can be included in the LRFD format once research has been completed and the
experience base has been established. The LRFD Specifications will continue to improve and the
resistance factors will be adjusted as more field performance measurements are evaluated. It is
important that the experience of geotechnical engineers with LRFD be shared and that the quality
of the geotechnical data base be improved (e.g., load-deformation response of spread footing
foundations) through the use of well planned and instrumented testing programs. As a result, our
understanding of design methods and the safety margins needed for their effective and economic
use in reducing the risk of failure can be improved.
Some methods of predicting the nominal resistance are empirical whereas others are based on
classical theories of mechanics. Also, different methods of predicting resistances employ the use of
different soil parameters [e.g., friction angle of cohesionless soils based on Standard Penetration
Test (SPT) blow counts or Cone Penetration Test (CPT) tip resistance, and undrained shear strength
of cohesive soils based on CPT sleeve friction]. It is generally true in ASD that a higher FS is used
for empirically-based methods (e.g., bearing resistance estimated using SPT blow counts), as
opposed to methods based on classical bearing resistance theories. Because different methods of
predicting resistance have different degrees of reliability, different values of resistance factors are
required for each method.
The load factor, i, chosen for a particular load type must consider the uncertainties in the:
y
y
y
(1985). Limit states concepts for foundation design are also being considered in Japan and other
countries in Europe. Some of the advantages and limitations of the LRFD method are:
Advantages of LRFD:
Achieves relatively uniform levels of safety based on the strength of soil and
rock for different limit states and foundation types.
Limitations of LRFD:
The most rigorous method for developing and adjusting resistance factors to
meet individual situations requires availability of statistical data and
probabilistic design algorithms.
Resistance factors vary with design methods and are not constant.
Due to the advantages of LRFD and the prospect that the method will eventually supersede ASD in
geotechnical engineering, a discussion on the basis for the derivation of resistance factors is
presented in Section 3.3. This derivation is presented for background information and for
understanding the process. In using the AASHTO LRFD Specifications, resistance factors are
provided and no probability or statistical analysis is required.
3.3 Calibration
The process of assigning values to resistance factors and load factors is called calibration. A design
code may be calibrated by use of (1) judgment, (2) fitting to other codes, (3) reliability theory, or
(4) a combination of approaches.
Calibration by judgment requires experience. For example, poor past performance of foundations
may force a code authority to adjust the code until satisfactory results are achieved. Code
parameters for structures that perform satisfactorily were accepted as correct, although this may be
excessively conservative. A fundamental disadvantage of this method of calibration is that it results
in non-uniform levels of conservatism.
Calibration by fitting to other codes, or simply fitting, involves using parameters (i.e., resistance
factors) that would result in the same minimum permissible physical dimensions of a foundation as
3-6
by ASD. Calibration by fitting does not achieve more uniform margins of safety than the ASD
procedures it replaces. It does, however, make it possible to use the same loads for superstructure
and foundation, and it ensures that the new code will not lead to radically different designs from the
old code. Calibration by fitting with ASD can be used where there is insufficient statistical data to
perform a more formal process of calibration by reliability theory.
A code can be calibrated by fitting to ASD as follows:
y
Divide the LRFD equation (Eq. 3-3) with i = 1.0 by the ASD equation (Eq.
3-2):
R n i i Q i
Rn
FS Q i
from which:
Q
FS Q
i
(Eq. 3-4)
i
If the loads consist only of dead load QD and live load QL, then Eq. 3-4
becomes:
DQD + LQL
FS(Q D + Q L )
(Eq. 3-5)
(Eq. 3-6)
From Eq. 3-6, and equivalent LRFD resistance factor can be calculated from a given ASD safety
factor for known load factors and the unfactored dead to live load ratio.
Example Problem 3-1
Calculate the resistance factor, , that is equivalent to an ASD safety factor FS = 2.5 if the dead load
factor D = 1.25, the live load factor L = 1.75, and the dead to live load ratio QD/QL = 3.0.
Substituting values into Eq. 3-6:
=
D (Q D Q L ) + L 1.25(3.0) + 1.75
=
= 0.55
FS (Q D Q L + 1)
2.5 (3.0 + 1)
(Eq. 3-6)
The calculated resistance factor indicates that to obtain an LRFD resistance equivalent to that
3-7
obtained using ASD with a safety factor of 2.5, the resistance predicted by the LRFD limit state
equations must be reduced by multiplying by a factor of 0.55. Values of resistance factor obtained
from Eq. 3-6 for a range of safety factors and dead to live load ratios are shown in Table 3-1 for D =
1.25 and L = 1.75. The ratio of dead to live load depends on the construction material (steel,
concrete or wood) and the span length of the bridge (the longer the span, the larger the dead to live
load ratio; QD/QL . span (m)/20). In general, Table 3-1 shows that the resistance factor decreases
with increasing safety factor and the influence of QD/QL is small.
Note in Eq. 3-6 that the values of resistance factors vary with values of load factors. The load
factors and load combinations for highway bridge design using the AASHTO LRFD Specification
(1997a) are presented and discussed in Chapter 4.
Table 3-1
Values of Resistance Factors by Fitting
for Different Values of Safety Factor and
Dead to Live Load Ratios for D = 1.25 and L = 1.75
Safety
Factor
1.5
2.0
2.5
3.0
3.5
4.0
QD/QL = 1
1.00
0.75
0.60
0.50
0.43
0.38
Resistance Factor,
QD/QL = 2
QD/QL = 3
0.94
0.92
0.71
0.69
0.57
0.55
0.47
0.46
0.40
0.39
0.35
0.34
QD/QL =4
0.90
0.68
0.54
0.45
0.39
0.34
3-8
is like the design point on the failure surface in the Level II method. The Level I method says the
midspan value is sufficiently close and does not examine other points on either side of the midspan
for the absolute moment. For the AFOSM method, an iterative procedure must be used in which an
initial value of reliability index, , is assumed and the process is repeated until the difference in
calculated values of on successive iterations is within a small tolerance. This iterative procedure is
based on normal approximations to nonnormal distributions at the design point developed by
Rackwitz and Fiessler (1978). The load and resistance factors used in the AASHTO LRFD
Specification for foundation design were developed using the nonlinear Level II FOSM procedures
(Barker, et al., 1991b). Implementing the Level II method requires the use of a computer program to
efficiently perform the iterations.
The Level I method is referred to as the mean value FOSM (MVFOSM) method because the
linearization occurs at the mean values of Q and R, rather than at the design point. The values of the
reliability index, , determined by the MVFOSM method are the least accurate of the methods, but
have the advantage that explicit equations can be written for . The equations for in Chapter 3 are
based on Level I approximations.
The basic procedure adopted for calibration of the AASHTO LRFD Specifications (Level I or Level
II) by reliability theory employed the following steps:
Step 1
Step 2
Observe the variation of reliability levels with different span lengths, load
ratios (e.g., dead to live load and other load combinations), geometry of the
foundations and methods of predicting resistance.
Step 3
Step 4
A simple example is provided to illustrate the various steps involved in the calibration process in
Section 3.4. To help understand the calibration process using reliability theory, a brief primer on the
basic concepts of reliability theory is given in the next section.
Use of reliability analyses is not necessary to apply the LRFD method in practice. For design
situations or methods that are not encompassed by the AASHTO LRFD Specifications, reliability
analysis may be used by the engineer in the calibration process to obtain appropriate resistance
3-9
factors for design. Alternatively, appropriate values of the resistance factor can be established by
fitting. Fitting is always used, even when reliability analyses are used, because fitting provides an
effective means of checking to ensure that the results of the more complex reliability analyses are
reasonable.
3.4 Review of Statistics and Probability Concepts
3.4.1 Statistical Descriptors
Statistical uncertainties can be described by the mean, x , standard deviation, , and coefficient of
variation, COV. These terms are defined in the following paragraphs.
The mean value, x , of a given set of data, x = (x1, x2, ..., xN), is calculated as:
x=
(Eq. 3-7)
where
N = Number of data values (dim).
The mean value is also called the expected value or the average of the data set. The standard
deviation, , is a measure of dispersion of the data in the same units as the data, xi, and is defined as:
=
(x
x)
(N 1)
(Eq. 3-8)
The coefficient of variation, COV, is a dimensionless measure of the variability of the data. The
COV is defined as the standard deviation () divided by the mean value (x
):
COV = x
(Eq. 3-9)
The COV expresses the magnitude of the variability as a percentage or fraction of the mean value.
Knowledge of these statistical descriptions is needed when compiling the load and resistance
statistics required for calibration by reliability analyses. An example is provided below.
Example Problem 3-2:
This example illustrates the calculation of the mean, x, standard deviation, , and coefficient of
variation, COV, using Eqs. 3-7, 3-8 and 3-9, respectively. The bias factor is also defined and
calculated. Analysis of the data shows that a larger sample size is needed so that a few outlying
points do not distort the distribution of the data.
Several case histories of good quality static pile load tests in sands were compiled, together with the
boring information available from these sites. Using the SPT blow counts from the borings, the
driven pile axial soil resistance for each case was predicted using Meyerhof's (1976) procedure.
3-10
Table 3-2 summarizes the results of the predicted resistances, the measured resistances and the ratio
of the measured to the predicted resistances.
Table 3-2
Summary of Measured and Predicted Driven Pile Axial Soil Resistance
Using Meyerhof Procedure
Measured Resistance
(tons)
60
67
39
65
34
44
232
197
71
55
145
120
25
49
229
54
272
158
98
14
27
85
21
20
Predicted Resistance
(tons)
89
89
52
72
37
48
240
199
72
55
145
120
25
48
221
48
220
129
73
10
19
56
10
5
Measured Resistance/
Predicted Resistance
0.67
0.75
0.77
0.90
0.90
0.91
0.97
0.99
0.99
0.99
1.00
1.00
1.01
1.03
1.04
1.13
1.23
1.23
1.34
1.46
1.47
1.52
2.13
3.96
The mean, x, standard deviation, , and coefficient of variation, COV, of the ratio of measured to
predicted capacities are 1.22, 0.66 and 0.54, respectively. Because x, of the ratio of the measured to
the predicted resistance is 1.22, this implies for this example that Meyerhof's method tends to
underpredict, on average, the pile resistance. Thus, there is a difference between what is predicted
and what is measured. This difference is referred to as the "bias." The bias factor, , of Meyerhof's
SPT method is defined as the ratio of the measured resistance to the predicted resistance as shown in
Eq. 3-10:
= Rm Rn
(Eq. 3-10)
3-11
where:
Rm =
Rn =
For the example given in Table 3-2, the average bias factor, , is 1.22 with a COV = 0.54. The COV
is relatively high because one standard deviation represents a variation of 54 percent from the mean
compared to typical values less than 20 percent. The high variation is caused by the large bias in the
last two data entries in Table 3-2. If these two data points are dropped, the values of x, , and COV
are changed to 1.06, 0.23 and 0.22, respectively. These values are more reasonable. Because the
data set is relatively small, dropping the number of data points from 24 to 22 by removing the two
outlying points changes the statistics dramatically. The data base of load tests should be large
enough and should contain high quality data, so that the statistics derived from the data base will be
representative of the loads and prediction practice.
3.4.2 Probability Density Functions
A histogram showing the frequency of occurrence of the measured to predicted axial driven pile
resistances from Table 3-2 is presented in Figure 3-1. The histogram was constructed by counting
the number of ratios in each of the equal intervals of 0.25 (e.g., in the interval from 0.51 to 0.75
there are 2 values and in the interval from 0.76 to 1.00 there are 10 values). It is apparent by
examining Figure 3-1 that the distribution of measured to predicted resistance is not symmetrical
about the mean value of 1.22. Further, the histogram shows how extreme the value near 4.0 is and
brings into question its validity.
Figure 3-1
Histogram of Measured to Predicted Axial Driven Pile Resistance
for Example Problem 3.2
The histogram also provides an approximate estimate of the probability of Meyerhof=s method to
equal or overpredict the driven pile resistance as 12/24 = 0.50 (the estimate is approximate because
the data set includes only 24 values out of possibly thousands of load tests that could be used for the
3-12
comparison). This ratio is analogous to flipping a coin to obtain either a heads or tails outcome; but
that is not the whole story. The histogram shows that most of the data points are clustered around
the mean value (1.22), and that is good. In this case, 16 of the 24 values (i.e., 67%) are within the
range 0.76 to 1.25, and 22 of 24 (92%) are within the range 0.51 to 1.75. The histogram shows that
there is more to interpreting data than knowing the mean value. To have confidence in measured to
predicted resistance, the scatter or dispersion of the data must also be known.
To ensure that the predicted resistance is less than or equal to the measured resistance for all cases in
Figure 3-1, the predicted resistance can be multiplied by 0.67, where 0.67 represents the minimum
ratio of measured to predicted resistance in Table 3-2. The result of this multiplication is to shift the
entire histogram to the right so that no value of measured to predicted resistance will be less than 1.
The value of 0.67 is similar to a reduction or resistance factor applied to Meyerhof=s method for
predicting the axial resistance of driven piles.
For years, histograms have been plotted to show distributions of natural phenomena. A recurring
pattern was noticed in these distributions that could be described mathematically. The derived
mathematical function is referred to as a probability density function. This function is similar to
drawing a smooth curve that approximately passes through the values of the histogram in Figure 3-1.
A sketch of a function f(x) is shown in Figure 3-2.
Figure 3-2
Lognormal Probability Density Function
The function f(x) in Figure 3-2 is standardized by dividing its ordinates by the total area under the
curve (includes all of the data points). The total area under this normalized curve is unity or a
probability of one because it includes all the data points in the theoretical total population of all
possible outcomes. Thus, the area under the curve, or probability of occurrence, P(x), over a small
interval, dx, between x and x + dx is equal to the product of the function at x and dx (i.e., P(x) =
f(x)dx).
3-13
The mathematical expression for the standard or normal form of f(x) is derived as an exponential
function symmetrical about the mean value. Values of f(x) and areas (probabilities) between fixed
intervals are available in textbooks. The shape of the normal distribution curve is like a bell and just
two parameters can describe the function: the mean, x (m in Figure 3-3), and the standard deviation,
. Changes in the shape of f(x) with variations in these two parameters are shown in Figure 3-3. If
m changes and remains constant, the curve shifts to the right, but its shape does not change (i.e.,
Figure 3-3b). If m remains constant and changes, the position of the curve does not change, but its
shape does. If decreases (i.e., less scatter of data), the shape becomes more compact (i.e., Figure
3-3c). If increases (i.e., more scatter), the shape spreads out (i.e., Figure 3-3d). In all of the cases
shown in Figure 3-3, the area under the curve is unity.
Figure 3-3
Standard Normal Density Function (Benjamin and Cornell, 1970)
When the data distribution is unsymmetric, a logarithmic normal (or simply lognormal) probability
density function is often suitable. Stated mathematically, if y = ln(x) is normally distributed, then x
is said to be lognormal. In calibrating the AASHTO LRFD Specification, the lognormal function
was used because it appeared to better represent the observed distribution of resistance data for
predicting the moment strength of bridge girders. The distribution is probably skewed because the
materials supplied usually have strengths greater than the nominal values assumed in the prediction
equations. Based on the distribution of data collected from weigh-in-motion studies, a normal
3-14
probability density function was used to represent the observed distribution of load data. Other
design methods and data bases could result in the use of different distribution functions for load and
resistance (e.g., both distributions for load and resistance could be lognormal).
Lognormal probability density functions are shown in Figure 3-4 for different values of its standard
deviation, . Notice that as increases, the lack of symmetry becomes more pronounced and the
smooth curve more closely represents the histogram of Figure 3-1.
Figure 3-4
Lognormal Density Function
The lognormal mean, m, and lognormal standard deviation, , can be determined using Eq. 3-11 and
3-12, respectively.
= ln x
(1 + COV )]
2
(Eq. 3-11)
and
= ln 1 + COV 2
(Eq. 3-12)
where:
x =
Mean value defined by Eq. 3-7
COV = Coefficient of variation defined by Eq. 3-9
ln ( ) = Natural logarithm of the expression in parentheses
For values of COV less than 0.2, Eqs. 3-11 and 3-12 are approximately equal to the following
simplified relationships:
= ln x
(Eq. 3-13)
3-15
and
2 = ln COV 2
(Eq. 3-14)
Thus, the lognormal mean and lognormal standard deviation can be calculated from the statistics
obtained from the standard normal function. In Example Problem 3-2, using Eqs. 3-11 and 3-12, the
lognormal mean and lognormal standard deviation are 0.071 and 0.50, respectively. The COV is too
large to use the approximate equations. Taking the anti-log of 0.071 gives a comparable normal
mean of 1.07 which indicates that the skew is to the left of the standard normal mean value of 1.22.
3.4.3 Probability of Failure
In the context of reliability analysis, failure is defined as the conditions where a predefined limit
state is reached. Load and resistance factors are selected to insure that each possible limit state has
an acceptably small probability of occurrence. The probability of failure can be determined if the
mean and standard deviation of the resistance and load are known.
To illustrate the procedure, consider the probability density functions for the load, Q, and resistance,
R, shown in Figure 3-5 for a particular limit state. Provided the resistance R is greater than the load
Q, there is a margin of safety for the limit state under consideration.
Figure 3-5
Probability Density Functions for Normally Distributed Load and Resistance
(Eq. 3-15)
where the right side of Eq. 3-15 represents the probability, P, that R is greater than Q. Because the
values of both R and Q vary, load and resistance factors are chosen so that there is a small
3-16
(Eq. 3-16)
where the right hand side of Eq. 3-16 represents the probability, P, that R is less than Q.
The probability density functions for R and Q in Figure 3-5 have been drawn to represent different
coefficients of variation, COVR and COVQ. The areas under the two curves are both equal to unity,
but the resistance, R, is shown with greater dispersion than the load, Q. This pattern typically occurs
because the design loads are usually known with more certainty than the resistance (e.g., axial pile
geotechnical resistance). The shaded area where the curves overlap in Figure 3-5 indicates the
region of failure, but the shaded area is not equal to the probability of failure because it is a mixture
of areas from two distributions with different ratios of standard deviation to mean value.
To evaluate the probability of failure, pf, a single combined probability density function, g(R,Q),
should be used that represents the margin of safety. This limit state function has its own unique
statistics. Use of a combined probability density function, g(R,Q) is examined below to show how
the probability of failure can be estimated.
If R and Q are normally distributed, the limit state function g(R,Q) can be expressed as:
g (R , Q ) = R Q
(Eq. 3-17)
For lognormally distributed R and Q, the limit state function g(R,Q) shown in Figure 3-6 can be
written as:
g (R , Q ) = ln (R ) ln (Q ) = ln (R Q )
(Eq. 3-18)
In both cases, the limit state is reached when R = Q and failure occurs when g(R,Q) < 0.
To determine the probability of failure, pf, it is not necessary to construct the function g(R,Q).
All that is required are the mean values, R and Q , and the coefficients of variation, COVR and
COVQ of the resistance, R, and load, Q, determined separately. This procedure is described in
Section 3.4.4.
3.4.4 Reliability Index,
A simple method of expressing the probability of failure is to use the reliability index, .
Assuming the function g(R,Q) defined by Eq. 3-18 has a lognormal distribution, the frequency
distribution of the function would have the shape similar to the curve shown in Figure 3-6. This
curve is a single frequency distribution curve which combines the uncertainties of both R and Q.
The probability of attaining a limit state (i.e., R < Q) is equal to the probability that ln(R/Q) < 0, and
3-17
Figure 3-6
Definition of Reliability Index, for Lognormal Distributions of R and Q
The probability of failure would be smaller, and thus safety increased, if the grouping of data about
the mean, g , was smaller, or if g was located further to the right in Figure 3-6. These two
possibilities can be combined into one if the position of the mean from the origin is specified in
terms of the standard deviation, g, of g(R,Q). Thus, the distance g from the origin to the mean in
Figure 3-6 provides a measure of safety. The number of standard deviations in this measure is
known as the reliability index, .
Reliability index, , is defined as the number of standard deviations, g, between the
mean value, g , and the origin (i.e., = g /g).
If the resistance, R, and load, Q, are both lognormally distributed random variables and are
statistically independent, it can be shown that the mean value of g(R,Q) is:
R 1 + COV 2
Q
g = ln
2
Q 1 + COVR
(Eq. 3-19)
[(
)(
g = ln 1 + COVQ2 1 + COVR2
)]
(Eq. 3-20)
where:
R, Q =
COVR, COVQ =
Mean values
Coefficients of variation of R and Q
3-18
Using g = g
, and substituting for g and g using Eqs. 3-19 and 3-20, respectively, the relationship
for the reliability index, , can be expressed as:
=
ln (R Q ) 1 + COVQ2 1 + COVR2
ln[(1 + COV )(1 + COV )]
2
Q
)]
(Eq. 3-21)
2
R
Eq. 3-21 is convenient because does not depend on the distribution of the combined function
g(R,Q), but only on the statistics of R and Q individually. The reliability index, , is important
because it is the basic parameter for code calibration using reliability theory.
Eq. 3-21 shows the effect of the variables that influence the value of . While values of
determined from Eq. 3-21 (Level I probability theory) are sufficiently accurate to be useful for most
purposes, more accurate procedures are available to estimate values of (Level II probability
theory). These techniques are iterative and are beyond the scope of this chapter. A description of
this advanced method is presented in Appendix A of Barker, et al. (1991b). Using Level I theory
gives values within about 10 percent of those obtained using Level II theory for lognormal
distributions of load and resistance. The same is true for other combinations of load and resistance
distributions.
A commonly accepted relationship between the reliability index, , and the probability of failure, pf,
has been developed by Rosenblueth and Esteva (1972) for lognormally distributed values of R and Q
using the relationship:
p f = 460 exp( 4.3 )
2<<6
(Eq. 3-22)
ln (460 p f )
10 1 < p f < 10 9
4 .3
(Eq. 3-23)
Values for both of these relationships are given in Table 3-3. In general, a change of 0.5 in results
in an order of magnitude change in pf (e.g., = 2.5 compares to 1 chance in 100, = 3.0 compares to
1 chance in 1000). As mentioned previously, there is no comparable relationship between the factor
of safety used in ASD and probability of failure. Inclusion of the probability of failure is a major
advantage of LRFD compared to ASD.
If the resistance, R, and the load, Q, are both normally distributed random variables, equations for
the mean value, x, standard deviation, , reliability index, , and probability of failure, pf, are
presented in Appendix 3A. Equations are also presented in the appendix for a lognormally
distributed resistance, R, and a normally distributed load, Q, which is the combination used in
calibration of the AASHTO LRFD Specification.
For a given value of , the probability of failure, pf, for the lognormal distributions (Table 3-3) is
generally less than for the normal distributions (Table 3A-1 in Appendix A). For example, for =
3-19
3.5, pf from Table 3-3 is 1.34 H 10-4 while from Table 3A-1, pf = 2.33 H 10-4. If the distributions for
R and Q are mixed with one lognormal and the other normal, pf will be between the values presented
in Tables 3-3 and 3A-1.
Table 3-3
Relationship Between Probability of Failure
and Reliability Index for Lognormal Distribution
Reliability
Index
2.5
3.0
3.5
4.0
4.5
5.0
5.5
Probability of
Failure
pf
0.99 H 10-2
1.15 H 10-3
1.34 H 10-4
1.56 H 10-5
1.82 H 10-6
2.12 H 10-7
2.46 H 10-8
Probability of
Failure
pf
1 H 10-1
1 H 10-2
1 H 10-3
1 H 10-4
1 H 10-5
1 H 10-6
1 H 10-7
Reliability
Index
1.96
2.50
3.03
3.57
4.10
4.64
5.17
The calculated nominal resistance, Rn, can differ from the mean value of the measured resistance,
Rm. To account for the discrepancy between Rm and Rn, a bias or correction factor is introduced as
follows:
Rm = Rn
(Eq. 3-24)
where represents the bias factor. The bias factor includes the net effect of various sources of error
such as the tendency of a method to underpredict pile resistance, energy losses in the equipment in
obtaining SPT blow counts, and soil borings in strata not representative of the site.
The uncertainty associated with may be represented by its mean and COV. Assuming that the
values of are statistically independent, the mean of the overall bias factor may be written as the
product of the individual bias factors:
= 1 2 KK N
(Eq. 3-25)
3-20
and the COV of the overall bias factor is the square root of the sum of the squares of the individual
coefficients of variation:
COV = COV21 + COV22 + KKK + COV2N
(Eq. 3-26)
Sources of error in predicting axial geotechnical pile resistance by the SPT method include:
1. Model error: Error due to an overall bias in the prediction method. For example,
the tendency of Meyerhof=s method to underpredict, on average, the axial
geotechnical resistance of driven piles.
2. Systematic error: Error in measurement of SPT blow count due to equipment or
procedure. For example, use of an older drill rig may provide systematically higher
blow counts compared to a new rig using a safety hammer.
3. Inherent spatial variability. The uncertainty associated with the value of SPT
blow count at a point is larger than the uncertainty associated with an average value
measured over a distance or volume, because of the averaging effect. For example,
only a limited number of soil borings are advanced at a site to characterize the
subsurface conditions for a large volume of soil and rock. If more borings are
advanced, the uncertainty of spatial variations across a site will be reduced.
4. Statistical uncertainty due to an insufficient number of tests: The uncertainty in
the predicted value of SPT blow count is inversely proportional to the square root of
the number of tests performed.
5. Error in measuring actual pile capacities in load tests: The error due to friction
in jacking systems, inaccuracy in calibration or data recording and method used to
define the failure load.
The error associated with Item 4 is site specific and the error associated with item 5 is difficult to
quantify. In our example, only the first three sources of error are considered. The model error
associated with Meyerhof's SPT method for predicting axial pile capacity in sands was discussed in
3.4.1. (The values of bias and COV from Example 3-2 for this method are 1.22 and 0.54,
respectively.) Because Items 4 and 5 have not been included in the analysis, as well as other
unknown factors, engineering judgment must be used in the final selection of resistance factors
applied to resistance prediction methods.
The SPT is the most commonly used method for in situ testing of soils. However, many variations
of the test equipment give rise to different energy levels imparted to the soil. Such factors include
the type of hammer (which affects the amount of energy delivered to the system), length of drill
rods, diameter of borehole, nature of drilling fluid, type of drill bit, type of sampling spoon, rate of
blow count and type of drill rods (Seed and DeAlba, 1986). Most of these effects can be minimized
by standardizing the test for 60 percent efficiency of the hammer blows. Nevertheless, variability in
the equipment and procedure cannot be eliminated. Orchant, et al. (1988) found that systematic
3-21
error due to equipment, procedure and random effects cause variation in the measured values of N.
They proposed the COV values for these effects ranging from 0.15 to 0.45. A bias factor
approaching unity may be used due to averaging effects, if statistics are not available.
Vanmarke (1977) showed that the COV of a soil property is reduced when the property is averaged
over a length or volume. The amount of reduction depends on the variance function V(dz), which is
defined as the ratio of the COV of a soil property averaged over a length dz, (COVSPT)dz, to the point
COV of that soil property measured at a point in the soil, COVSPT:
V (dz) =
(COVSPT )dz
(Eq. 3-27)
COVSPT
Vanmarke (1977) defined the scale of fluctuation, S, as the distance over which the soil property
shows strong correlation from point to point. He further showed that the variance function may be
related to the scale of fluctuation as follows:
V (dz) =
S
dz
(Eq. 3-28)
Equating Eq. 3-27 and 3-28, the COV of the soil property (in this example the SPT blow counts)
over a depth dz can be written as:
S
dz
(Eq. 3-29)
where:
S=
Scale of fluctuation of the soil property
COVSPT = COV of the soil property at a point
Studies by Vanmarke (1977) showed that the scale of fluctuation for SPT is approximately 2.5 m.
Briaud and Tucker (1984) showed that the COVSPT is about 0.42. Therefore, with S = 2.5 m and
COVSPT = 0.42, over a depth equal to L, the COV due to inherent spatial variability is:
(COVSPT )dz
= COVSPT
S
2.5 0.44
= 0.42
=
dz
L
L
0.5
A summary of the mean and COV values for this example is presented in Table 3-4.
3-22
Table 3-4
Summary of Resistance Statistics
Correction
1. Model Error
2. Equipment/Procedure Used in SPT Test
3. Inherent Spatial Variability
Note: L = Length of pile (m)
= 1 2 3
R = 1.3 1.0 1.0 = 1.3
Using Eq. 3-27, the COV of the overall bias on the resistance is:
COV = COV21 + COV22 + COV23
The results of statistical analysis of highway dead and live loads are summarized in Table 3-5
(Nowak, 1993). The largest variation is the weight of the wearing surface placed on bridge decks.
Also of interest, as indicated by the bias factor, is that the observed actual loads are greater than the
specified nominal values.
3-23
Table 3-5
Statistics for Bridge Load Components
Load Component
Dead Load
y Factory-made
y Cast-in-place (CIP)
y Asphaltic wearing surface
Live Load (w. dynamic load allowance)
Bias,
COV
1.03
1.05
1.00
1.10 - 1.20
0.08
0.10
0.25
0.18
In this example, the bridge is assumed to be constructed using a CIP concrete deck ( = 1.05, COV =
0.10) on steel girders (i.e., = 1.03, COV = 0.08) without an asphaltic wearing surface, and subject
to live load ( = 1.15 and COV = 0.18).
Using Eq. 3-25, the overall bias for the dead load components is:
= 1 2
QD = 1.03 1.05 = 1.08
Using Eq. 3-26, the COV of the overall bias on the dead load is:
COV = COV21 + COV22
where:
Q, R =
3-24
Qn, Rn =
Q, R =
R 1 + COV 2
Q
ln R n
2
Q Q n 1 + COVR
ln 1 + COVR2 1 + COVQ2
[(
)(
)]
Rn can now be expressed in terms of Qn from Eq. 3-2 as Rn = FS H Qn, where FS is the factor of
safety, and Qn is composed of both dead load, QD, and live load, QL, each with their own separate
bias factors, QD and QL so that:
R n = FS(Q D + Q L )
n Q n = QD Q D + QL Q L
Further, it is assumed that the square of the COV of a function of multiple variables is equal to the
sum of the squares of the individual COV=s (Eq. 3-26), so that:
FS(Q + Q ) 1 + COV 2 + COV 2
QD
QL
D
L
ln R
2
1 + COVR
QD Q D + QL Q L
=
2
2
ln 1 + COVR2 1 + COVQD
+ COVQL
[(
)]
)(
Finally, dividing the numerator and denominator of the ln [ ] term in the numerator of the
relationship above by QL, we obtain:
FS(Q Q + 1) 1 + COV 2 + COV 2
QD
QL
D
L
ln R
2
1 + COVR
QD Q D Q L + QL
=
2
2
ln 1 + COVR2 1 + COVQD
+ COVQL
[(
)]
)(
(Eq. 3-30)
where:
QD, QL =
FS =
R, QD, QL =
COVR, COVQD, COVQL =
y
y
y
y
FS
QD/QL
R, QD, QL
COVR, COVQD, COVQL
The safety factor used with Meyerhof's SPT method is usually about 3.5.
We can study the variation of the reliability index with the dead to live load ratio. The ratio of dead
to live load is related to the span length of a bridge. The dead to live load ratio is smaller for bridges
with shorter spans, and is approximately equal to the span in meters divided by 20.
Step 2: Observe the Variation of the Reliability Indices
A plot of reliability index, , versus pile length can now be generated by substituting bias factors: R
= 1.3, QD = 1.08, QL = 1.15; and coefficients of variation: COVR = (0.34 + 0.44/L)0.5, COVQD =
0.13, COVQL = 0.18 from Sections 3.4.5 and 3.4.6 into Eq. 3-30 for several values of dead to live
load ratio and pile length. The result of such a substitution is shown in Figure 3-7. Note that the
reliability index varies between 2.4 and 2.6 and is not very sensitive to pile length (a measure of
spatial variability) or dead to live load ratio (a measure of bridge span). For an abutment supported
on piles, this lack of sensitivity means that the reliability appears to be independent of the size of the
abutment and the span of the bridge.
Figure 3-7
Reliability Index for Meyerhof's SPT Method
Step 3. Select a Target Reliability Index T
Based on the computed values of reliability index, desired or target reliability indices can now be
selected. Values of reported by others can provide guidance in this selection. Meyerhof (1970)
suggested that pf of foundations should be between 1 H 10-3 and 1 H 10-4, corresponding to
between 3 and 3.6 (Table 3-3). Wu, et al. (1989) calculated that for pile systems is approximately
4, corresponding to pf = 5 H 10-5. Tang, et al. (1990) reported that for offshore piles ranges from
1.4 to 3.0, corresponding to pf between 1 10-1 and 1 10-3. The value calculated for Meyerhof=s
SPT method of . 2.5 is well within the range determined by others.
3-26
In addition to the SPT method, values of reliability indices for other commonly used methods (e.g.,
, , and CPT) of predicting the axial resistance of driven piles were calculated and were found to
be between about 1.5 and 3.0 (Barker, et al., 1991a). Thus, a target reliability index between 2.5 and
3.0, corresponding to a pf of about 1 10-3, appears to be about the average value determined for
single piles based all of the previous studies. However, because piles are often used in groups,
failure of one pile does not necessarily imply that the pile group will fail. Due to redundancy in pile
groups, it is reasonable to use a lower value of for a single pile. If a single pile in a group has the
smallest resistance and begins to fail, the load is transferred to other piles in the group with greater
resistance and the foundation does not fail. Thus, a reasonable value of target reliability index, T,
for single driven piles appears to be in the range of 2.0 to 2.5, corresponding to pf between 1 10-1
and 1 10-2.
Step 4: Calculate the Resistance Factor
The selected values of reliability index, T, provide the basis for determining values of resistance
factors. An expression for the resistance factor can be derived from the LRFD safety criterion in Eq.
3-3.
R n i Q i
(Eq. 3-31)
from which:
iQi R n
(Eq. 3-32)
/R, gives:
Replacing the nominal resistance, Rn, by the mean value divided by the bias factor, R
R iQi R
(Eq. 3-33)
R=
[(
)(
(1 + COV ) (1 + COV )
2
Q
)]}
2
R
(Eq. 3-34)
Substituting R
from Eq. 3-34 into Eq. 3-33, and replacing with the target (desired) reliability
index T, gives the following expression for the resistance factor:
=
2
2
R ( i Qi ) (1 + COVQ) /(1 + COV R )
Q exp { T
When dead and live loads are considered, Eq. 3-35 can be written as:
3-27
(Eq. 3-35)
R (D
=
2
2
1 + COVQD
+ COVQL
QD
+ L)
1 + COV 2R
QL
Q
2
2
(QD D + QL) exp [T ln(1 + COV 2R )(1 + COVQD
+ COVQL
)]
QL
(Eq. 3-36)
where D and L are the specified load factors for dead and live load. Even though the asphaltic
wearing surface of the bridge has a different load factor than the dead load, it constitutes only a
small fraction of the bridge load compared to the dead load of the bridge. In our example, the load
factor for the asphaltic wearing surface can be taken equal to the load factor for the dead load
without significant error.
It can be seen from Eq. 3-36 that the resistance factor, , depends on the dead to live load ratio,
QD/QL, load factors for dead and live loads: D = 1.25 and L = 1.75; bias factors: R = 1.3, QD =
1.08, QL = 1.15; and coefficients of variation: COVR = (0.34 + 0.44/L)0.5, COVQD = 0.13, COVQL =
0.18 from Sections 3.4.5 and 3.4.6; and the target reliability index, T. Values of resistance factors
for Meyerhof's SPT method corresponding to T of 2.0 and 2.5 are shown in Figure 3-8, for different
dead to live load ratios (measure of bridge span) and pile lengths (measure of spatial variability).
The values of resistance factors vary from 0.39 to 0.46 for T = 2. For T = 2.5, the resistance factor
varies between 0.28 and 0.33. The resistance factors, similar to the reliability index, are relatively
insensitive to pile length and dead to live load ratios. These results, tempered by judgment and
fitting with ASD experience, can be used to select an appropriate value of for use with Meyerhof's
SPT method of estimating pile resistance. A discussion of the issues involved in selecting for use
in a design specification is provided in Section 3.4.7.
Figure 3-8
Resistance Factors for Meyerhof's SPT Method for T = 2.0 and 2.5
3-28
D (Q D / Q L ) + L
QD (QD / QL) + QL
1.3(1.0) + 2.17
= 1.74
1.0(1.0) + 1.0
avg =
1.25(1.0) + 1.75
= 1.50
1.0(1.0) + 1.0
to:
the resistance factor will be lower by the ratio of 1.50/1.74 = 0.86 or a reduction of 14 percent. For a
longer span of 74 m (QD/QL . 3.7), avg, drops from 1.49 to 1.36 representing a reduction of 9
percent. Applying a ratio of 0.89, which is between 0.86 and 0.91, to the reliability determined
resistance factors in Table 3-6, the SPT values become 0.43 and 0.45 while the CPT values become
0.53 and 0.55. This relatively small drop in the calculated resistance factor does not suggest a
need to revise the corresponding values prescribed in the AASHTO LRFD Specifications (1997a)
of = 0.45 for SPT and = 0.55 for CPT.
Table 3-6
Resistance Factors for Axially Loaded
Driven Piles in Sand, In-Situ Methods
Pile
Length
(m)
SPT data
10
SPT data
30
CPT data
10
CPT data
30
(1)
T for a single pile
Soil Test
Factor of
Safety
3.5
3.5
2.5
2.5
Target
Reliability
Index T(1)
2.0
2.0
2.0
2.0
3-29
Resistance Factor,
Reliability Fitting with
Selected
Analysis(2)
ASD(3)
0.48
0.39
0.45
0.51
0.39
0.45
0.59
0.54
0.55
0.62
0.54
0.55
(2)
(3)
Level II reliability analysis with QD/QL = 3.7 for D = 1.3 and L = 2.17
Eq. 3-6 with QD/QL = 3.7 for D = 1.25 and L = 1.75
When the resistance factors determined by reliability theory and fitting with ASD are close to one
another, such as with the CPT data in Table 3-6, it is not difficult to select the final value of = 0.55.
However, when they are quite different, as with the SPT data, engineering judgment is required to
select an appropriate value. In this case, the reliability based value was given more weight because
the ASD safety factor for SPT data was judged to be larger than necessary, and the selected = 0.45.
In a similar manner, values of resistance factors were selected by Barker, et al. (1991a) for
bearing resistance of footings and sliding of footings on sand and clay, bearing resistance of
axially loaded single piles and pile groups, and bearing resistance of axially loaded single drilled
shafts and groups of drilled shafts. Recognizing the uncertainties in the selection of the factors,
it was decided to select values that varied by increments of 0.05, because it was felt that use of
additional significant figures was not justified.
Application of these resistance factors to the design of bridge foundations are illustrated in the
design manuals prepared by Barker, et al. (1991b). Tables summarizing the resistance factors
from the AASHTO LRFD Specifications are presented in the design chapters.
3-30
4. What are the advantages and disadvantages of both LRFD and ASD? (Refer to Sections
3.2.4 and 3.2.2)
5. What are the methods by which resistance factors can be calibrated? (Refer to Section 3.3)
3-31
6. Using Eq. 3-6 in Section 3.3, determine the appropriate resistance factor through calibration
with ASD for the following parameters:
FS
D
L
QD/QL
= 3.0
= 1.25
= 1.75
= 2.0
7. Define the mean, standard deviation and coefficient of variation of a set of normally distributed
data. (Refer to Section 3.4.1)
9. Define the lognormal mean and lognormal standard deviation. (Refer to Section 3.4.2)
3-32
11. For lognormal distributions of load and resistance, recall from Section 3.4.4 that the reliability
index, , can be expressed as follows:
=
(Eq. 3-21)
2<<6
(Eq. 3-22)
A subsurface exploration and laboratory testing program including field vane shear tests (VST) and
unconfined compression (UC) tests results in a mean ultimate pile capacity prediction, R
= 1000 kN
as shown below. The mean unfactored load on this pile, Q
= 333 kN. The reliability of each test
method is somewhat different, with COVR = 0.25 for the VST and a COVR = 0.35 for the UC tests.
For these load and resistance characteristics, determine the ASD factor of safety and the LRFDbased probability of failure based on VST and UC testing.
R
= 1000 kN
Q
= 333 kN
COVQ = 0.15
COVR (VST) = 0.25
COVR (UC) = 0.35
3-33
3.6 Appendix 3A
Both R and Q Normally Distributed
Mean Value, g:
g =R -Q
(Eq. 3A-1)
Standard Distribution, g:
g =
2
2
R + Q
(Eq. 3A-2)
where R
and Q
are mean values, g and g are standard deviations of the resistance, R, and the
load, Q.
Reliability Index, :
R - Q
(Eq. 3A-3)
2
Q
2
+ R
(Eq. 3A-4)
where Fu( ) is the standard normal cumulative distribution function. There is no simple
expression for Fu, but it has been evaluated numerically and tabulated. Relating the reliability
index, to the probability of failure, pf:
= F-1u (1 - pf )
(Eq. 3A-5)
where Fu-1 ( ) is the inverse standard cumulative distribution function. Values for the
relationships in Eq. 3A-4 and Eq. 3A-5 are presented in Table 3A-1.
3-34
Table 3A-1
Relationship Between Probability of Failure
and Reliability Index for Normal Distribution
Reliability
Index
2.5
3.0
3.5
4.0
4.5
5.0
5.5
Probability of
Failure
pf
0.62 10-2
1.35 10-3
2.33 10-4
3.17 10-5
3.40 10-6
2.90 10-7
1.90 10-8
Probability
of Failure
pf
1 10-2
1 10-3
1 10-4
1 10-5
1 10-6
1 10-7
1 10-8
Reliability
Index
2.32
3.09
3.72
4.27
4.75
5.20
5.61
Resistance Factor,
=
R ( i Qi )
Q + T
(Eq. 3A-6)
2
2
R + Q
R ( D QD + L QL )
(QD QD + QL QL) + Tsqrt 2R + Q2
(Eq. 3A-7)
Mean Value g
g R [1 - k(COVR )] {1 - l n [1 - k(COV R )]} Q
(Eq. 3A-8)
Standard Deviation,
{R [1 - k(COVR )]}2 + Q2
(Eq. 3A-9)
where R and Q are mean values, R and Q are standard deviations, COVR and COVQ are
coefficients of variation of the resistance R and the load Q. The parameter k is comparable to the
number of standard deviations from the mean value. As an initial guess, k is often taken as 2.
Reliability Index,
3-35
(Eq. 3A-10)
{R [1 - k(COVR )]}2 + Q2
Probability of Failure
For this mixed case, pf cannot be explicitly defined. However, for a given value of , it is reasonable
to assume that pf is between the values given in Table 3-3 and those in Table A-1.
Resistance Factor,
Q + T
(Eq. 3A-11)
{R [1 - k(COVR )]}2 + Q2
( QDQ D + QL QL) + T
{R [1 - k(COVR )]}
3-36
2
Q
(Eq. 3A-12)
CHAPTER 4
LOADS
4.1 Introduction
The fundamental principals of LRFD as embodied in the AASHTO LRFD Specification (AASHTO,
1997a), and comparisons between working stress and limit state design methods were presented in
Chapter 3. Recall the LRFD equation:
i i Qi R n = R r
where:
i =
i =
Qi =
=
Rn =
Rr =
In using this approach, the designer must (1) estimate the loads and load combinations which could
be imposed on the structure, and (2) estimate the nominal (ultimate) resistance of the structure
component or ground. This chapter describes general methods which can be used to estimate loads,
Qi, and load factors, i, required for the design of foundations, walls and culverts using LRFD.
Chapters 8 through 16 will present general methods which can be used to estimate nominal
resistance, Rn, and resistance factors, , needed for substructure design.
LRFD requires that loads on and from structures be estimated for the geotechnical design of
substructures. This process includes estimating permanent loads resulting from the dead weight of
structure components and burial below grade, and transient loads such as those due to traffic, water,
wind, ice, and vessel impact. With the exception of earth loads, the loads required for substructure
design are usually provided by the structural engineer. In this regard, there is no difference between
ASD and LRFD. The development of loads and factored loads is covered in Section 3 of the
AASHTO LRFD Specification.
This chapter:
The chapter does not describe in detail methods prescribed in the AASHTO LRFD Specifications to
estimate loads familiar to geotechnical engineers (e.g., earth loads).
4.2 Load Considerations for Geotechnical Design of Substructures
Substructures must be designed to support load combinations which include permanent and transient
loads from a bridge superstructure and from externally applied forces (e.g., impact loading from a
vessel, earth loading by structure backfills and downdrag loading on deep foundations extending
through a compressible fill). Figure 4-1 illustrates the types of loads which need to be considered
for the design of a bridge pier and a cantilevered retaining wall.
Figure 4-1
Typical Loading for Substructure Design
With the possible exception of earth, surcharge and downdrag loads which are often developed by
the foundation engineer, the selection of applicable loads and load combinations for structure design
are made by the structural engineer for superstructure design. Although the foundation engineer
may never have to develop the loads needed for substructure design, some general familiarity with
the types and methods used to develop design loads is useful in understanding the selection of limit
4-2
states and load combinations used in LRFD. The following sections describe the permanent and
transient loads which may be required for substructure design.
4.2.1 Permanent Loads
Permanent loads on substructures include dead loads from structure components, earth loads from
embedment of walls and burial of culverts, and downdrag loads on piles or drilled shafts.
4.2.1.1 Dead Loads - DC, DW and EV (A3.5.1)
Dead loads include the weight of factory-made (e.g., structural steel and prestressed concrete) and
cast-in-place (e.g., deck slabs, abutments and footings) structure components, asphalt wearing
surfaces, future overlays and planned widenings, miscellaneous items (e.g., scuppers, railings and
supported utility services), and the weight of earth cover.
4.2.1.2 Downdrag Load - DD (A3.11.8)
When a driven pile or drilled shaft penetrates a soft layer subject to settlement (e.g., where an
overlying embankment may cause settlement of the layer), the force effects of downdrag or negative
loading on the foundations must be evaluated. These force effects are illustrated in Figure 4-2, and
are fully mobilized at relative movements of about 3 to 12 mm (Reese and O'Neill, 1988).
Downdrag acts as a permanent additional axial load on the pile or shaft. If the force is of sufficient
magnitude, structural failure of the foundation element or a bearing capacity failure at the tip is
possible. At smaller magnitudes of downdrag, the foundation may experience additional settlement.
Figure 4-2
Downdrag Loading
(Hunt, 1986)
For driven piles or drilled shafts that derive their resistance mostly from end bearing, the structural
resistance of the foundation section must be adequate to support the factored structure and downdrag
loads. Battered piles should be avoided at sites where downdrag loading is possible due to the
potential for bending of the pile section. Details and example problems showing the analysis of deep
foundations subjected to downdrag loading are presented in GRL (1996), Reese and O'Neill (1988)
and Briaud and Tucker (1994).
4-3
The stiffness of the structure and the characteristics of the retained earth are the most significant
factors in the development of lateral earth pressure distributions. Structures which can tilt, move
laterally or deflect structurally away from the retained soil (i.e., most retaining walls and abutments)
can mobilize an active state of stress in the retained soil mass. These structures should be designed
using an active (i.e., minimum) earth pressure distribution. Structures which are restrained against
movement (e.g., integral abutments or walls for which lateral ground movement of the backfill could
adversely affect nearby facilities within a distance of less than about one-half the wall height behind
the wall) should be designed to resist an at-rest earth pressure distribution. Walls which are forced
to deflect laterally toward the retained soil should be designed to resist the passive earth pressure.
For practical purposes, the passive state of stress occurs most commonly as a result of lateral
deflection of the embedded portions of retaining walls into the supporting soil. In the AASHTO
LRFD Specification, passive earth pressure is treated as a resistance rather than a load.
The basic earth pressure (p) is defined by:
p = k h s z
where:
p=
kh =
s =
z=
The value of k depends on the stress history of the soil (i.e., whether the soil is normally
consolidated [NC] or over consolidated [OC]) and the displacement of the structure (i.e., whether the
structure is flexible or stiff and whether soil loading is active or passive). The initial or at-rest value
of kh (i.e., ko) ranges between about 0.4 and 0.6 for NC soils, and can exceed 1.0 for heavily OC
soils. Structure movement will increase or decrease the value of kh from ko such that movement
away from the soil will cause the value of k to decrease below ko and movement toward the soil will
cause the value of k to increase above ko. Minimum and maximum values of kh (i.e., ka, and kp) are
achieved when the shear strength of the soil is completely mobilized. For conventional walls (i.e.,
gravity, semi-gravity and cantilever), the lateral movement required to develop the minimum active
earth pressure or maximum passive earth pressure is a function of the type of soil retained as shown
in Table 4-1.
Table 4-1
Relative Movements Needed to Achieve Active or Passive Earth Pressure Conditions
Clough and O'Rourke (1991)
Values of /H
Active
Passive
0.001
0.01
0.002
0.02
0.004
0.04
Backfill Type
Dense Sand
Medium Dense Sand
Loose Sand
where:
=
H=
Nearly all cantilever retaining walls of typical proportions deflect sufficiently to permit mobilization
of active earth pressures. Gravity and semi-gravity walls designed with a sufficient mass to support
only active earth pressures will tilt and/or translate in response to more severe loading conditions
(e.g., at-rest earth pressures) until stresses in the retained soil are relieved sufficiently to permit
development of an active stress state in the retained soil. However, at-rest earth pressures could
develop on the stem of cantilevered retaining walls where a rigid stem-to-base connection may
prevent lateral deflection of the stem with respect to the base. For such a condition, excessive lateral
earth pressures on the stem could conceivably cause structural failure of the stem or stem-to-base
connection.
As shown in Figure 4-3a, the resultant force from a linearly increasing pressure distribution is
4-5
located at the centroid of the pressure diagram at H/3 from the base of a wall which tilts about its
base. However, if the wall tilts about its top or translates laterally as shown in Figures 4-3b and 43c, the location of the resultant force is higher than traditionally assumed for design. Location of the
resultant force above the centroid of the pressure diagram occurs because as a wall deflects in
response to lateral earth loading, the backfill must slide down along the back of the wall for the
retained soil mass to achieve an active state of stress. This movement causes arching of the backfill
against the upper portion of the wall which causes an upward shift in the location at which the
resultant of the lateral earth load is transferred to the wall. In the AASHTO LRFD Specification
(A3.11.5.1), it is assumed that the resultant earth pressure for active loading is located at 0.4H from
the base of the unsupported wall section for conventional concrete gravity and reinforced concrete
cantilever walls.
where:
pa =
ka =
's =
H=
The lateral earth pressure on a mechanically-stabilized earth (MSE) retaining wall is assumed to
increase linearly with depth, as shown in Figure 4-5 and 4-6, for level and sloping backfill surfaces,
respectively. The earth pressure resultant is assumed to act at a height of h/3 above the base of the
wall for evaluation of external wall stability. The magnitude of the lateral earth pressure resultant is
taken as:
Pa = 0.5 k a 's h
where:
Pa =
ka =
's =
h=
4-7
p = k s qs
(4-5) (A3.11.6.1-1)
where:
p =
ks =
qs =
For active earth pressure conditions, ks, is taken as ka; for at-rest conditions, ks is taken as ko.
4.2.2 Transient Loads
The transient loads that may need to be developed for structure design include vehicular traffic,
earthquake, water and stream pressure, ice, wind, temperature, and vehicle impact. The methods
used in the AASHTO LRFD Specification to estimate these transient loads are described below.
4.2.2.1 Vehicular Live Load - LL (A3.6.1)
The force effects of vehicle loads are modeled using a design vehicle to represent typical variations
in axle loads and spacing. The live load model in the AASHTO LRFD Specification is based on
evaluation of various vehicle configurations (NCHRP, 1990). The group of vehicles and loadings
considered in the report were analyzed to determine their force effect on simply-supported and twospan continuous structures. The results of the analyses demonstrated that the combination of the
design tandem and uniform lane load and the combination of the HS20 design truck with the design
lane load provide an adequate basis for design. The configuration and loading from the HS20 truck
in the AASHTO LRFD Specification is shown in Figure 4-7. The design tandem consists of a pair
of 110 kN axles spaced at 1.20 m apart with a transverse spacing of 1.80 m. The design lane load
consists of a load, uniformly distributed in the longitudinal direction equal to 9.3 kN/m.
4-9
Additional guidance regarding the application and distribution of vehicle loads is presented in
Section 4.4 where vehicle load effects are discussed for culvert design.
4.2.2.2 Vehicular Dynamic Load Allowance - IM (A3.6.2)
The force effects of dynamic vehicle loading on structures is reflected by applying a dynamic
allowance factor to the design load. This factor accounts for the effects of:
For foundations and abutments supporting bridges, these force effects are incorporated in the loads
used for superstructure design. For buried substructures (e.g., culverts and embedded wall
elements), the effects of dynamic loading dissipate with depth such that below a depth of about 3 m,
the effects are negligible.
4.2.2.3 Vehicular Centrifugal Force - CE (A3.6.3)
Vehicles traveling on a superstructure location on a horizontal curve generate a centrifugal force
effect that must be considered in design. In the AASHTO LRFD Specification, the centrifugal force
is estimated as the product of the design axle load of the design truck or tandem and the factor C
which is determined as:
C=
4 v2
3g R
where:
C=
v=
g=
R=
In the AASHTO LRFD Specification, centrifugal forces are applied horizontally at a distance of 1.8
m above the roadway surface. In computing the centrifugal force, the lane load is neglected because
the spacing of vehicles at high speed is assumed to be large so that the density of vehicles is small.
For substructure design, centrifugal forces would be developed by the structural engineer and would
represent a horizontal force effect.
4.2.2.4 Vehicular Braking Force - BR (A3.6.4)
Vehicle braking represents a horizontal force effect along the length of a bridge which must be
resisted by the structure foundations. In the AASHTO LRFD Specification, these forces are taken as
25 percent of the design axle load of the design truck or tandem in each lane of traffic headed in the
4-10
same direction. As with centrifugal force effects, braking forces are applied at a distance of 1.8 m
above the roadway surface.
4.2.2.5 Pedestrian Live Load - PL (A3.6.1.6)
Vehicular bridges supporting pedestrian sidewalks and bridges used for pedestrian and/or bicycle
traffic must be designed for the loads imposed by this use. The AASHTO LRFD Specification
requires a pedestrian load of 3.61 kPa for bridges with sidewalks wider than 0.6 m and a load of 4.1
kPa for bridges used solely for pedestrian and bicycle traffic.
4.2.2.6 Live Load Surcharge - LS (A3.11.6.2)
The force effects of surcharge and traffic loads on backfills must be considered for the design of
walls and abutments. As shown in Figure 4-8, live load surcharge effects produce a lateral pressure
component on a wall in addition to lateral earth loads. The magnitude of the lateral pressure due to
the surcharge load is a function of the type (i.e., point or uniformly distributed), magnitude and
proximity of surface loading, the strength of the backfill, and the stiffness of the wall relative to wall
displacement. The distribution of horizontal pressure produced by surcharge loads is usually
estimated using a Boussinesq distribution and assuming the soil backfill is elastic.
If traffic is expected within a distance behind a wall equal to about the wall height, the live load
traffic surcharge is assumed to act on the retained earth surface. The uniform increase in lateral
earth pressure due to live load surcharge is typically estimated as:
p = k s s h eq
where:
p=
's =
ks =
heq =
For active earth pressure conditions, ks is taken as ka, and for at-rest conditions, ks is taken as ko.
Traditionally, heq has been assumed to be constant for walls of any height (e.g., 0.61 m in AASHTO
ASD). However, in the AASHTO LRFD Specification, heq is currently applied as a function of wall
height as shown in Table 4-2.
Table 4-2 (A3.11.6.2-1)
Equivalent Height of Soil for Vehicular Loading
(AASHTO, 1997a)
Wall Height
(m)
< 1.5
3.0
6.0
> 9.0
heq
(m)
1.70
1.20
0.76
0.61
The values of heq in the table were determined based on evaluation of horizontal pressure
distributions produced on retaining walls by the vehicular live loads prescribed in AASHTO LRFD
Specification using a Boussinesq elastic half-space solution and a Poisson's Ratio of 0.5. For other
live load distributions, the live load surcharge, ph, can be determined as shown in Figure 4-8 using:
ph = 2 p/ ( - sin cos ( + 2))
where:
p=
4-12
the effects of stream pressure include floating debris, waves and stream currents, and scour.
4.2.2.8 Wind Load - WS and WL (A3.8)
The principal wind loads which need to be considered by the structural engineer include horizontal
wind pressure effects on the structure, WS, and vehicles, WL. The effects of vertical wind pressure
on the underside of bridges due to an interruption of the horizontal flow of air, and the effects of
aeroelastic instability represent special load conditions for long-span bridges. For small and/or low
structures, wind loading does not usually govern the design. However, for large and/or tall bridges,
noise walls, overhead signs and light standards, wind loading can govern the design and should be
investigated. Where wind loading is important, the wind pressure should be evaluated from two or
more different directions to determine the windward (facing the wind), leeward (facing away from
the wind) and side pressures which produce the most critical loads on the structure.
The design wind velocity is a function of the height of the structure and the aerodynamic
characteristics of the near-surface environment in the upstream wind direction. In the AASHTO
LRFD Specification, the wind velocity for bridges more than 10 m above the water or low ground,
VDZ, can be determined using:
V10 Z
V DZ = 2.5 V0
ln
V B Z0
where:
VDZ =
V10 =
VB =
Z=
V0 =
Z0 =
Condition
V0 (km/hr)
Z0 (m)
Open Country
13.2
0.07
Suburban
15.2
0.30
City
19.4
0.80
Using the design velocity VDZ, the wind pressure on structures, WS is determined as:
2
V DZ
V 2DZ
=
PD = PB
P
B
25 600
VB
where:
PB =
Structural Component
Trusses, Columns and Arches
Beams
Large Flat Surfaces
In the AASHTO LRFD Specification, the wind pressure on vehicles, WL, is applied as a moving
force of 1.46 kN/m acting normal to the structure at a height of 1.8 m above the roadway surface.
This load is based on a long row of randomly sequenced passenger cars, commercial vans and
trucks exposed to a 90 km/hr design wind.
When the wind direction is not normal to the exposed surface area of the structure, the
components of wind pressure are:
PN = PD cos 2
and
PP = PD cos sin
where:
PN =
Pp =
PD =
=
4-14
Thus, consideration of force effects due to superimposed deformations are of limited concern
during most substructure design.
4.2.2.11 Earthquake Force - EQ (A3.10)
Earthquake force effects are predominately horizontal and act through the center of mass of the
structure. Because most of the weight of a bridge is in the superstructure, seismic loads are assumed
to act through the bridge deck as shown in Figure 4-9. These loads are due to inertial effects and
therefore are proportional to the weight and acceleration of the superstructure. The effects of
vertical components of earthquake ground motions are typically small and are usually neglected
except for complex bridges.
Figure 4-9
Earthquake Loads on a Bridge
(Buckle, 1994)
4-15
The inertial loads in the superstructure must be adequately transferred to the ground through the
connections and bearings to the columns, piers and abutments, and then through the foundations.
Because there is usually little redundancy in this load path, failure of any one can lead to partial or
complete collapse of a bridge. The importance of load-path integrity has been demonstrated
repeatedly from recent earthquakes. In some cases, the failure of a small component (e.g., a bolt)
has led to progressive failures of other components and the collapse of major bridge components.
The effects of soil liquefaction and slope movements resulting from earthquake events must be
evaluated separately from the effects of seismic loading on structure components.
Earthquake design loads are horizontal forces determined from the product of the elastic seismic
response coefficient, Csm, the equivalent weight of the superstructure, and a response modification
factor. For most sites and bridges with span lengths less than 150 m, the elastic seismic response
coefficient can be determined as:
Csm =
1.2 AS
T 2m/ 3
2.5 A
where:
Csm =
A=
S=
Tm =
For Soil Profiles III and IV (Table 4-7) and in areas where A $ 0.3, Csm
need not exceed 2.0A
For Soil Profiles III and IV (Table 4-7), and for modes other than the
fundamental mode which have periods less than 0.3 sec, the value of Csm
can be determined as:
(Eq. 4-14) (A3.10.6.2-1)
Csm = A (0.8 + 4 T m)
Csm = 3 AST3m/ 4
Except where sites are located near an active fault, the acceleration coefficient, A, can be determined
using the contour maps presented in the AASHTO LRFD Specification (A3.10.2). In areas of active
faulting, the value of A should be determined by an experienced seismologist.
4-16
The AASHTO LRFD Specification uses Seismic Performance Zones to define the minimum
requirements which must be followed with regard to methods of analysis, minimum support lengths,
column design details and foundation and abutment design procedures for various seismic regions.
The Seismic Performance Zones are presented in Table 4-5.
Table 4-5 (A3.10.4-1)
Seismic Performance Zones
(modified after AASHTO, 1997a)
Acceleration
Coefficient
(dim)
A # 0.09
0.09 < A # 0.19
0.19 < A # 0.29
0.29 < A
Seismic Zone
1
2
3
4
The effect of subsurface conditions at a site on the response of a bridge or other structure to
earthquake loading is accounted for using a site coefficient, S. As used in the AASHTO LRFD
Specification, the value of S depends on the soil profile at the site as shown in Table 4-6. The soil
conditions corresponding to each soil profile type are presented in Table 4-7. In general, sites that
have softer soils have a higher site coefficient to account for the effects of amplification of bedrock
ground motions through soil.
Table 4-6 (A3.10.5-1)
Site Coefficients
(AASHTO, 1997a)
Site Coefficient
S
I
1.0
IV
2.0
Because it is uneconomical to design a bridge to resist large earthquake forces elastically (i.e.,
without damage), most bridges are designed to perform in a ductile manner during large
earthquakes. The seismic force effects on substructures and connections between parts of structures
are, therefore, limited due to displacement and/or yielding. Accordingly, the force for seismic
design can be determined by dividing the force from an elastic analysis by a response modification
factor, R, from either Table 4-8 or 4-9.
4-17
Table 4-7
Classification of Soil Profile Types
(after AASHTO, 1997a)
Soil Profile
Type
I
II
III
IV
Classification
Shale-like or crystalline rock; stable deposits of gravel, sand or stiff
clays with the depth to rock < 60 m; or soils with shear wave velocity >
750 m/sec.
Stiff cohesive or deep cohesionless soils comprised of stable deposits of
gravel, sand or stiff clays with depth to rock > 60 m.
Soft to medium-stiff clays with or without intervening layers of sand or
other cohesionless soils with thickness $ 9 m.
Soft clays or silts > 12 m in depth; or loose natural deposits or manmade, non-engineered fill; or materials with shear wave velocity < 150
m/sec
Table 4-8 (A3.10.7.1-1)
Response Modification Factors for Substructure Design
(AASHTO, 1997a)
Importance Category
Critical
Essential
Other
1.5
1.5
2.0
Substructure
Wall-type piers - larger dimension
Reinforced concrete pile bents
Vertical piles only
With batter piles
Single columns
Steel or composite steel and concrete
pile bents
Vertical piles only
With batter piles
Multiple column bents
1.5
1.5
1.5
2.0
1.5
2.0
3.0
2.0
3.0
1.5
1.5
1.5
3.5
2.0
3.5
5.0
3.0
5.0
In the AASHTO LRFD Specification, essential bridges are structures which must remain open to
emergency vehicles for security or defense purposes after the design earthquake (i.e., event with a
475-year return period). Critical bridges are defined as structures which must remain open to
emergency vehicles for security or defense purposes after a large earthquake (i.e., event with a
minimum return period of 2,500 years).
4-18
Connection
Superstructure to abutment
Expansion joints within a span of the superstructure
Columns, piers, or pile bents to cap beam or superstructure
Columns or piers to foundations
All Importance
Categories
0.8
0.8
1.0
1.0
R factors for connections are smaller than those for substructure members to preserve the integrity of
the bridge under extreme loads.
Lateral earth pressures on retaining structures are amplified during earthquake loading due to
horizontal acceleration of the retained earth mass. The Mononobe-Okabe method of analysis is a
pseudo-static method which is often used to develop an equivalent static fluid pressure to model
seismic earth pressure on walls. The method is applicable when:
The failure surface defining the active wedge of soil loading the wall is
planar
4-19
Figure 4-10
Definition Sketch for Mononobe-Okabe Analysis
(Barker, et al., 1991)
where the seismic active earth pressure coefficient, Kae (AA.11.1.1.1-2) is:
2
sin( + ) sin ( - - i)
cos ( - - )
1
+
K ae =
cos( + + ) cos(i - )
cos cos2 cos( + + )
-2
(Eq. 4-17)
where:
s =
H=
kv =
kh =
=
=
=
=
I=
The equivalent expression for the passive force mobilized if a wall is pushed into soil is:
Ppe = 0.5 s H 2 (1 - k v) K pe
where the seismic passive earth pressure coefficient, Kpe (AA.11.1.1.1-4) is:
2
sin( + ) sin ( - + i)
cos ( - + )
=
1
+
K pe
cos( - + ) cos(i - )
cos cos2 cos( - + )
-2
(Eq. 4-19)
To estimate lateral earth pressures for earthquake loading, the vertical acceleration coefficient, kv, is
often assumed equal to zero and the horizontal acceleration coefficient, kh, is taken as:
k h = 0.5
(Eq. 4-20)
for walls permitted to move horizontally up to about 250 mm (e.g., gravity, cantilever and MSE
walls, and
k h = 1.5
(Eq. 4-21)
for walls for which essentially zero horizontal displacement is permitted (e.g., integral abutments
and anchored walls).
where:
=
A=
A/100
Horizontal earthquake acceleration coefficient (% of g)
The resultant of the dynamic component of the earth pressure, Pae, is located significantly higher on
the wall than the static active earth pressure resultant, Pa, as shown in Figure 4-10. Typically, the
combined static and seismic lateral earth pressure are assumed to be uniformly distributed with a
resultant, Pae, acting at the midheight of the wall.
4.2.2.12 Ice Load - IC (A3.9)
Ice force effects on piers are a function of the size of ice flow, the strength and thickness of ice, and
geometry of the pier. In determining the ice force, site conditions and the modes of ice flow must be
considered. These factors include:
Static uplift or vertical load from adhering ice in waters or fluctuating level
4-21
The effective ice crushing strength, p, can range from about 380 kPa for ice near melting and
substantially disintegrated to about 1530 kPa where ice break up occurs where the ice temperature is
measurably below the melting point.
For piers of standard dimensions on larger water bodies, crushing and bending ice failure usually
control the dynamic ice force for design. On smaller streams which cannot transport large ice flows,
impact failure usually controls the dynamic ice force for design.
For ice failure by crushing or bending, the horizontal force acting on a pier resulting from moving
ice, F, is affected by the ratio of the pier width to ice thickness, w/t, such that:
(Eq. 4-22)
F = Fc or Fb , whichever is less
F = Fc
Fb = C n p t
5 t
Ca = + 1
w
Cn =
0.5
tan ( - 15)
where:
t=
=
p=
4-22
w=
Fc =
Fb =
Ca =
Cn =
The ice force used for design can be modified additionally to account for the effects of stream size,
pier orientation and shape, and ice adhesion to the pier.
4.2.2.13 Vehicular Collision Force - CT (A3.6.5)
Structures not protected from collisions by roadway and rail vehicles must be designed for the force
effects of vehicle collisions. The AASHTO LRFD Specification requires that abutments and piers
located within a distance of 9 m to the edge of a roadway, or within a distance of 15 m to the
centerline of a railway track be designed to resist an equivalent static force of 1800 kN acting in any
direction in a horizontal plane at a height of 1.2 m above ground. This equivalent static force was
developed from analysis of full-scale crash tests. For column piers, the force should be applied as a
point load. For other structures, the load can be applied as a point force or distributed over an area
deemed suitable for the size of structure and anticipated impacting vehicle. Measures deemed
suitable for structure protection from vehicle collisions include embankments, and structurallyindependent, crashworthy, ground-mounted barriers of suitable height.
4.2.2.14 Vessel Collision Force - CV (A3.14)
Due to their proximity to navigation channels, bridge substructures may be subjected to vessel
collision by ships and barges. The principal factors affecting the risk and consequences of vessel
collisions with waterway substructures are related to vessel, waterway and bridge characteristics.
The consequences of vessels colliding with a structure are a function of the size, type, loading
condition and direction of the vessel. The width and depth of the navigation channel, stream
velocity, channel alignment, cross-section geometry, water elevation and hydraulic conditions
represent the primary characteristics of waterways. The bridge characteristics which need to be
considered include bridge layout, geometry and strength. While bridge piers should be located
outside the waterway, economic and engineering constraints limit the span of bridges and their
vertical clearance. In general, bridge piers can be designed or protected from vessel collisions.
However, it is usually not economically feasible to design bridge superstructures for vessel
collisions.
The impact force from a head-on vessel collision with a pier is:
Ps = 120 V
DWT
where:
Ps =
DWT =
V=
The impact force from a head-on collision of a barge with a pier is:
P B = 60 000 a B
if aB $ 0.1 m, where:
PB =
aB =
Equivalent static impact force for a standard inland hopper barge (kN)
Barge bow damage length (m)
The barge bow damage length for a standard hopper barge is:
a B = 3.1 ( 1 + 1.3 x 10-7 KE - 1)
where:
aB = Barge bow damage length (m)
KE = Vessel collision energy (joule)
The energy of a moving vessel to be absorbed during a noneccentric collision with a bridge pier is:
KE = 500 CH M V 2
where:
KE = Vessel collision energy (joule)
M = Vessel displacement tonnage (metric ton)
CH = Hydrodynamic mass coefficient equal to 1.05 if the underkeel clearance
exceeds 0.5 times the draft, and 1.25 if the underkeel clearance is less than
0.1 times the draft (dim)
V = Vessel impact velocity (m/s)
4.3 Load Factors and Load Combinations
When using ASD, there is no distinction between the loads (i.e., either in magnitude or combination
of load types) used to evaluate the ultimate load capacity or the deformation potential of the ground.
Thus, the same load magnitudes are used to estimate the ultimate bearing capacity and the
4-24
settlement of a foundation for ASD, despite the inherent uncertainty in loads applied from a structure
to its substructure components. LRFD incorporates this uncertainty by application of load factors
for various permanent and transient load types using the LRFD equation:
i i Qi R n = R r
where:
i =
i =
Qi =
=
Rn =
Rr =
Selection of the load factor(s) to be used is a function of the type of load and limit state being
evaluated. (Recall from Chapter 3 that a limit state is a condition beyond which a foundation or
structure component ceases to fulfill its intended function.) This section addresses load factor and
load modifier selection. The selection of a resistance factor(s) and estimation of the nominal
resistance is presented in Chapters 8 through 10 for design of foundations, Chapters 11 through 14
for design of walls, and in Chapters 15 and 16 for design of culverts.
For the AASHTO LRFD Specification, the limit states, load factors and load combinations which
must be investigated are presented in Tables 4-10 and 4-11.
Structure and foundation design by LRFD requires that:
4-25
LOAD
COMBINATION
LIMIT STATE
DC
DD
DW
EH
EV
ES
STRENGTH-I
(unless noted)
STRENGTH-II
STRENGTH-III
STRENGTH-IV
EH, EV, ES, DW
DC ONLY
STRENGTH-V
EXTREME
EVENT-I
EXTREME
EVENT-II
1.00
0.50/
1.20
0.50/
1.20
1.00
1.50
SE
EQ
IC
CT
CV
SE
SE
SE
TG SE
EQ 1.00
1.00
1.00
p 0.50 1.00
1.00
1.00/
1.20
1.00/
1.00
1.20
1.00/
1.00
1.20
SERVICE-I
SERVICE-II
SERVICE-III
FATIGUE-LL, IM
& CE ONLY
CONSTRUCTION
0.75
TG SE
-
TG SE
-
(1)
The reduced values of are used when calculating force effects other than displacements of joints
and bearings (A14.4.1).
4-26
Load Factor
Type of Load
DC: Component and Attachments
DD: Downdrag
DW: Wearing Surfaces and Utilities
EH: Horizontal Earth Pressure
Active
At-Rest
EV: Vertical Earth Pressure
Retaining Structure
Rigid Buried Structure
Rigid Frames
Flexible Buried Structures
Flexible Metal Box Culverts
ES: Earth Surcharge
Maximum
1.25
1.80
1.50
Minimum
0.90
0.45
0.65
1.50
1.35
0.90
0.90
1.35
1.30
1.35
1.95
1.50
1.50
1.00
0.90
0.90
0.90
0.90
0.75
From the Table 4-10, the limit states which must be investigated include:
The limit states are further subdivided based on consideration of applicable load combinations as
follows:
Strength I - Basic load combination related to the normal vehicular use of the
bridge without wind.
Strength II - Load combination relating to the use of the bridge by Ownerspecified special design vehicles and/or evaluation permit vehicles, without
wind.
Strength IV - Load combination relating to very high dead load to live load
force effect ratios exceeding about 7.0 (e.g., for spans greater than 75 m).
4-27
As will become evident in this and later chapters, most substructure designs will require evaluation
of foundation and structure performance at the Strength I and Service I Limit States. These limit
states are generally analogous to evaluations of ultimate capacity and deformation behavior in ASD,
respectively.
Load factors used in the AASHTO LRFD Specification for permanent loads and labeled as p in
Table 4-10, are presented in Table 4-11 as maximum and minimum values.
In reviewing Tables 4-10 and 4-11, the load factors vary for different load categories and limit states
to reflect either the certainty with which the load can be estimated or the importance of each load
category for a particular limit state. Some general comments about magnitude and relationship
between various load factors are highlighted below:
A load factor of 1.00 is used for all permanent and most transient loads for
Service I.
The live load factor for Strength I is greater than that for Strength II (i.e.,
1.75 versus 1.35) because variability of live load is greater for normal
vehicular traffic than for a permit vehicle.
4-28
The live load factor for Strength I is greater than that for Strength V (i.e.,
1.75 versus 1.35) because variability of live load is greater for normal
vehicular use without wind than for a bridge subjected to a wind of 90 km/hr,
and because less traffic is anticipated during design wind conditions.
The load factor for wind load on structures for Strength III is greater than for
Strength V (i.e., 1.40 versus 0.40) because the wind load represents the
primary load for Strength III where structures are subjected to a wind
velocity greater than 90 km/hr compared to Strength V where wind velocity
of 90 km/hr represents just a component of all loads on the structure.
The live load factor for Strength III is zero because vehicular traffic is
considered unstable and, therefore unlikely, under extreme wind conditions.
The load factors for wind load for Strength V are less than 1.00 (i.e., 0.40)
consistent with the common practice in ASD of allowing somewhat lower
factors of safety for structures subjected to both normal vehicle and wind
loads.
The maximum and minimum load factor for downdrag loads represent the
extreme values of pmax and pmin due to the uncertainty in accurately
estimating downdrag loads on piles.
The AASHTO LRFD Specification requires that certain permanent loads, including earth loads, be
factored using maximum and minimum load factors as shown in Table 4-11. Criteria for their
application require that:
Load factors be selected to produce the total extreme factored force effect,
and for each combination, both maximum and minimum extremes be
investigated
For load combinations where one force effect decreases the effect of another
force, the minimum value shall be applied to the load that reduces the force
effect;
For permanent force effects, the load factor which produces the more critical
combination shall be selected from Table 4-11; and
In the AASHTO LRFD Specification, each factored load is adjusted by a load modifier, i, to
4-29
account for the combined effects of ductility, D, redundancy, R, and operational importance, l.
The i factors (A1.3.2.1) represent a first attempt at codifying the influence of ductility, redundancy
and operational importance on structure performance. For loads for which a maximum value of i is
appropriate:
i = D R I 0.95
1
1.00
D R I
Due to a lack of precise information, the effect of these modifiers has been judged to range between
" 5 percent when accumulated geometrically. With time, it is hoped that improved quantification of
ductility, redundancy and operational importance, and their interaction and system synergy, can be
attained to better account for these factors in design.
For design at the Strength Limit State, values of i range as follows:
Ductility - D
Redundancy - R
Operational Importance - I
4-30
The methods used to estimate loads for the design of foundations and walls using LRFD are
fundamentally no different than procedures used in the past for ASD. Thus, procedures commonly
used by geotechnical engineers to estimate earth pressures, negative loads on piles, seepage
pressures due to hydraulic gradients, or surcharge load effects on earth pressure, for example, can be
used for LRFD. What has changed however is the manner in which loads are considered for
evaluations of foundation/structure stability (e.g., bearing and sliding resistance of spread footing
foundations) and foundation movement (e.g., axial and lateral displacement of a pile-supported pier).
This section highlights some of the areas in which considerations of force effects for the design of
foundations and walls differ between LRFD and ASD.
The design of foundations supporting bridge piers or abutments should consider all limit states
loading conditions applicable to the structure being designed. From a review of the Limit States
loading descriptions in Section 4.3 and the load factors in Tables 4-10 and 4-11, it can be seen that a
variety of Strength Limit States may control the design of a bridge pier or abutment foundation.
The Strength I Limit State will control for very high live to dead load ratios
The Strength III or V Limit States may control for structures subjected to
high wind loads
The Strength IV Limit State will control for high dead to live load ratios
The Extreme Limit States may control the design of foundations in a seismically active area (i.e.,
Extreme Limit I) or for foundations for piers which may be exposed to vehicle or vessel impact (i.e.,
Extreme Limit II). With respect to deformation, (i.e., lateral deflection or settlement), the Service I
Limit State will control. The Service II and Service III Limit States are used to evaluate specific
critical structural components and are not generally applicable to foundation design.
For a typical retaining wall, dead weight, earth pressure and live load surcharge will be the
predominant loads influencing wall stability. From a review of the Limit States loading
descriptions in Section 4.3 and Tables 4-10 and 4-11, it can be seen that the Strength I and IV
Limit States, for which the largest dead, earth and live load factors apply, will control the design
of a retaining wall with respect to stability. With respect to deformation, the Service II and
Service III Limit States only apply to special structures such that only the Service I Limit State
applies to retaining wall design.
Applying the criteria in Section 4.3 to evaluate the resistance of walls and foundations at the
Strength Limit State, some general observations can be made. The stability of conventional (i.e.,
gravity, semi-gravity and cantilever) retaining walls must be evaluated for sliding, bearing and
overturning. When applying the AASHTO LRFD Specification, these stability cases are discussed
below for a typical analysis of a cantilever wall at the Strength I Limit State.
4-31
Figure 4-11 shows that the vertical earth load, EV, on the rear of a wall is factored by pmin (1.00)
and the structure weight, DC, is factored by pmin (0.90) because these forces result in an increase in
the contact stress (and shear strength) at the base of the foundation. The horizontal earth load, EH,
is factored by pmax (1.50) for an active earth pressure distribution because the force results in a more
critical sliding force at the base of the wall. If EH has both horizontal and vertical components (e.g.,
a wall supporting a sloping backfill) both components are factored by 1.50.
Figure 4-11
Load Combination for Evaluation of Sliding Resistance
of a Cantilever Retaining Wall Supported on a Spread Footing
Figure 4-12a shows that the critical load combination for an analysis of bearing resistance of the
wall foundation often, but not always, occurs when the structure weight is factored by pmax (1.25),
the vertical earth load is factored by pmax (1.35), and the horizontal active earth pressure is factored
by pmax (1.50). For the case of a wall supporting sloping backfill, both the horizontal and vertical
components of EH are factored by 1.50.
Figure 4-12b shows that the critical load combination for an analysis of eccentricity (or overturning)
of the wall foundation occurs when the structure weight is factored by pmin (0.90), the vertical earth
load is factored by pmin (1.00), and the horizontal active earth pressure is factored by pmax (1.50).
For a wall supporting a sloping backfill, both the horizontal and vertical components of EH are
factored by 1.50. (This case may also control for analysis of bearing resistance due to the decrease in
effective foundation bearing width associated with increasing eccentricity.)
4-32
4-33
Culverts must be designed for the force effects from vertical earth and unbalanced horizontal earth
pressures, the weight of pavement layers, and vehicle live loads. For large flexible structures, loads
due to compaction of soil backfill may be important. In addition, water buoyancy loads should be
evaluated for structures with inverts below the water table. Because culverts are constructed below
grade, loads considered important for other types of structures (e.g., wind, temperature, vehicle
braking and structure dead load), are insignificant compared to the force effects due to vertical earth
pressure. Likewise, fatigue has not been shown to be a factor in long-term performance of culverts.
Similarly, the effects of earthquakes can generally be neglected for typical culverts, except where the
potential for ground instability or for unbalanced loading exists. This section focuses on the
determination of force effects on culverts due to vertical earth pressures and vehicular live loads
transmitted through earth fill. Design aspects for flexible and rigid culverts are presented in
Chapters 15 and 16, respectively.
The force effects of external loads on culverts depend on a number of factors including cover height,
backfill density, culvert size and shape, culvert stiffness relative to the adjacent side fill and
foundation stiffness. Moreover, the side fill soil adjacent to the culvert can represent either a load or
resistance, depending on its mode of interaction (i.e., either active or passive). Therefore, the force
effects on culverts can be variable and their determination depends on the accuracy of the
assumptions used in their development.
4.5.1 Earth Loads
Both the AASHTO LRFD Specification (1997a) and the AASHTO ASD Specification (1997b)
calculate the force effects of vertical earth pressure on a culvert on the basis of a prism load. The
prism load is the weight of the rectangular prism of soil above the culvert, or the free field geostatic
stress at the crown of the culvert multiplied by the culvert diameter. If a culvert deflects vertically,
contracts circumferentially or settles into the foundation more than the adjacent side fill, the load
4-34
supported by the culvert is reduced (i.e., positive arching occurs). Conversely, negative arching and
an increase in the load supported by a culvert occurs when the side fill settles less than the culvert
can deflect vertically, or when the culvert contracts circumferentially or settles more than the
foundation material can settle. The condition of equal settlement between the culvert and adjacent
side fill is termed neutral arching. The approximate boundaries of soil backfill mobilized by
positive, neutral and negative arching are illustrated in Figure 4-14.
Figure 4-14
Soil Blocks Mobilized by Arching Over Culverts
The prism load is represented in Figure 4-14(b) for the neutral arching case. Positive and negative
arching can be accounted for by modifying the prism load for soil-structure interaction using the
following relationship:
WE = Fe s B c H
where:
WE =
Unit load on culvert due to vertical earth pressure (kN/m)
Soil-structure interaction factor = VAF (dim)
Fe =
VAF = Arching factor (dim)
<1.0 for positive arching
= 1.0 for neutral arching
> 1.0 for negative arching
Unit weight of soil above culvert (kN/m3)
s =
H=
Height of backfill above culvert crown (m)
Diameter of culvert, outside diameter for wall profiles with a smooth
Bc =
outside surface, or wall centerline diameter for profiles with a corrugated
outside surface (m)
4-35
Current design practice limits the arching factors to a value of 1.0 for flexible culverts and 1.5 for
rigid culverts. Different methodologies for determining arching factors have evolved between
flexible and rigid culverts. A more complete treatment of these various approximate methods
specific to culvert type is given in Chapter 15 for flexible culverts and Chapter 16 for rigid culverts.
4.5.2 Vehicular Live Loads
The force effects due to vehicular live loads on culverts can be estimated by closed-form elastic
solutions, numerical finite element methods, and approximate methods as used in the AASHTO
specifications. The AASHTO ASD Specification (1997b) specifies use of a surface point load that
is spread through the underlying soil over an area having sides equal to 1.75 times the depth of
cover. This distribution is applicable only for cover depths greater than 0.6 m.
The AASHTO LRFD Specification (1997a) assumes that the contact pressure from a prescribed
tire footprint is distributed through the soil backfill in a manner similar to the 60 degree (from
horizontal) spreading rule found in many textbooks on soil mechanics. In this case, the spreading
is to take effect for cover depths greater than 0.6 m. For depths less than 0.6 m, the area of the
tire footprint itself is to be used to determine pressure below the surface due to live loads. At 1 m
of cover, this modification has the effect of increasing design pressures by 70% over the method
in the AASHTO ASD Specification.
As with other approximate methods for determining vertical earth pressures, the AASHTO
procedures for spreading live loads through earth fills are intended to obtain force effects averaged
across the culvert diameter. These procedures are used to calculate the average wall thrust due to
vehicle live loads, but are not appropriate for determining concentrated force effects from live loads,
such as bending stresses or localized deflections, because actual wheel loads do not distribute
uniformly through the soil. Rather, wheel loads distribute more as predicted by elastic theory as
shown in Figure 4-15. As indicated on Figure 4-15, live loads produce much higher localized effects
under shallow covers than predicted by the approximate average pressure models. However, the
peak pressures attenuate rapidly with depth. Thus, consideration of localized bending effects due to
live loads generally are of concern only for culverts under shallow covers or those subjected to
larger than typical concentrated live loads.
4-36
Figure 4-15
Design Truck Pressure Distribution From Elastic Theory
4-37
Type
Corrugated Metal Pipe
Spiral Rib Metal Pipe
Condition
Steel Conduit
Aluminum pipe for S # 1.2 m
Aluminum pipe for S > 1.2 m
Unpaved areas and under
flexible pavement
Compacted granular fill under
rigid pavement
-
Minimum Cover
S/8 $ 0.3 m
S/4 $ 0.3 m
S/2 $ 0.3 m
S/2.75 $ 0.6 m
Bc/8 or B=c/ 8, whichever
is greater, $ 0.3 m
0.23 m
ID/8 $ 0.3 m
Other factors regarding vehicle loads that have been modified in the AASHTO LRFD Specification
include load factors, live load impact factors (now termed dynamic load allowance factors), multiple
presence factors and the design live load model itself. The AASHTO LRFD Specification has
reduced the live load factor from 2.0 to 1.75, a reduction of 12%. However, both the magnitude and
the effective depth of live load impacts have been increased. At 1 m of cover, the modification to
the impact factor increases design pressures by over 125%. Also, the multiple lane presence factor
has been increased by 20% for one lane contributions, although it was left unchanged for two lane
contributions.
The net effect on design due to all of the factors described above is illustrated in Figure 4-16 for
one lane=s contribution from the design truck. As can be seen, the changes significantly increase
the conservatism of the specification for live load design. At 1 m of cover the required design
pressure is about 100% greater under the AASHTO LRFD Specification than under the AASHTO
ASD Specification.
While considerable effort was expended in developing the live load model in the AASHTO LRFD
Specification, the appropriateness of these changes for below-ground structures was not evaluated.
Further study is needed to refine the new LRFD live load model for the design of culverts.
Currently, utilization of the AASHTO LRFD provisions for distributing vehicle live loads to culverts
involves:
4-38
Figure 4-16
Comparison of AASHTO Live Load Pressures Through Earth Fills
Calculating the distributed pressure at the crown of the culvert due to tire
contact pressures at the road surface from alternative design vehicles
To perform these steps, the following independent and dependent variables, and geometric constants
must be determined or evaluated:
Independent Variables - Depth of cover (DE), group axle load (AL), load
factor (), soil spreading factor (SE), number of lanes (N) and capacity
reduction factor (R)
Geometric Constants - Tire footprint width, w, (0.51 m), truck width (1.80
m), tandem spacing (1.20 m), lane width (3.60 m) and lane clearance (0.60
m)
The general relationship for distributing design vehicle live loads to culverts is:
4-39
IM
m R AL 1 +
100
DTP =
AD
(Eq. 4-36)
where:
DTP =
m=
R=
ADTT =
=
AL =
IM =
AD =
The general relationship for calculating the distributed lane load, DLP, on a culvert is calculated as:
DLP =
9.3mR
3.00
(Eq. 4-37)
where:
DLP =
The vehicle lane load is distributed at the surface and includes the multiple presence factor, m,
and ADTT reduction factor, R (from A3.6.11.2), based on the number of lanes that govern the
design vehicle, but does not include the dynamic load allowance factor, IM.
Characteristics of the design truck and design tandem were described in Section 4.2.2.1. The
extreme effect of this load combination (designated HL-93 in the AASHTO LRFD Specification), is
determined by considering each possible combination of loaded lanes, the potential for multiple
presence of vehicles in each loaded lane, and the average one-way daily truck traffic. Other factors
which can affect determination of the design live load include:
For the design vehicle, the surface tire contact area equals the width of each wheel (i.e., 0.51 m)
times the wheel length given by:
IM
l = 0.00228 1 +
P
100
where:
R=
=
P=
IM =
(Eq. 4-39).(A3.6.2.2-2)
where:
DE =
Depth below road surface (m); DE = 0 for calculation of tire contact area at
road surface.
For calculation of pressures due to dynamic loads, DE = depth of cover from surface to culvert
crown. The effects of dynamic load allowance can be safely neglected for depths of cover greater
than about 2.4 m.
The maximum live load force effects are to be determined by considering each possible combination
of number of loaded lanes multiplied by the multiple presence factors given in Table 4-13. The
factors in Table 4-13 are based on average one-way daily truck traffic (ADTT) of 5000 vehicles.
For sites having significantly less truck traffic, a reduction factor (R) may be applied to the multiple
presence factors of Table 4-13. Values for (R) are based on ADTT as follows:
The dimensions of the distributed area at depth is equal to the dimensions of the tire footprint area at
the surface plus the depth, DE, times the soil spreading factor, SE. In the AASHTO LRFD
Specification, the distribution of surface live load pressure at depth is spread through the soil cover
as follows (A3.6.1.2.6):
Number of
Loaded Lanes
1
2
3
>3
Multiple Presence
Factor, m
1.20
1.00
0.85
0.65
The distributed area calculation is more complicated than merely adding the product (SE DE) to the
surface footprint dimensions because as depth increases, distributed areas begin to overlap, either
from adjacent wheel loads, adjacent axle loads, or at greater depths, from adjacent lanes. The
concept of overlapping distributed area is shown in Figures 4-17 and 4-18. Figure 4-17 illustrates
the tire footprint areas at the surface for both the design truck and the design tandem vehicles,
whereas Figure 4-18 illustrates the distributed area at depth when both wheel and axle loads have
overlapped.
4-42
The transition depths at which overlap occurs varies with load factor, , magnitude of wheel load, P,
dynamic load allowance, IM, and the soil spreading factor, SE. Several transition depths occur due
to varying geometry between truck width (1.80 m), tandem spacing (1.20 m), and lane width (3.60
m). Figure 4-18 provides dimensions of the distributed areas for the case of a single lane with full
overlap for both the design truck and the design tandem vehicles.
The distributed length, LD, due to the single lane contribution is calculated as:
For the design truck:
LD = l + SE DE
(Eq. 4-40)
where:
LD =
1.20 - l
SE
:
(Eq. 4-41)
LD = 2 (l + SE DE )
1.20 - l
LD = l + SE DE + 1.20
SE
:
(Eq. 4-42)
4-43
The distributed width, WD, due to the single lane contribution for both the design truck and
design tandem is:
For DE
1.80 - w
SE
:
(Eq. 4-43)
W D = 2 (w + SE DE )
where:
WD = Width dimension perpendicular to the direction of travel for distributed
area at depth DE (m)
w=
For DE >
1.80 - w
SE
:
(Eq. 4-44)
W D = w + SE DE + 1.80
A D = LD W D
The calculation of distributed area and length can be simplified with reasonable accuracy using
only one transition depth using the following relationships:
A D = 4.056 (SE DE )
(Eq. 4-46)
LD = 0.231 + SE DE
3
2
A D = 0.675 (SE DE ) - 3.53 (SE DE ) + 13.40 (SE DE ) - 3.18
For DE
(Eq. 4-47)
1.20 - 0.176
SE
:
4-44
(Eq. 4-48)
LD = 0.176 + SE DE
For DE >
1.20 - 0.176
SE
(Eq. 4-49)
LD = 0.176 + SE DE + 1.20
(Eq. 4-50)
The effect of multiple lanes also can be investigated using Eq. 4-36. The results are presented in
Figure 4-19 for the design truck and Figure 4-20 for the design tandem. Note that the single lane
contribution could be used conservatively in lieu of considering multiple lane effects for depths of
cover less than 2.4 m. Also, at a depth of 2.4 m, the distributed pressures from the design truck and
design tandem are nearly equal. At shallower depths, the design truck appears to govern the design
for all multiple lane loading combinations.
Figure 4-19
Live Load Distribution Through Soil for Design Truck
4-45
Figure 4-20
Live Load Distribution Through Soil for Design Tandem
This evaluation demonstrates that accounting for multiple presence factors for buried culverts
can lead to unrealistic and unconservative results for more than two lanes. At 1 m of cover, a 35
percent reduction in pressure is permitted for structures loaded by 4 or more lanes (Table 4-13).
Clearly, vehicles that are three lanes away should not be considered either as adding to or
subtracting from the pressure distribution under shallow cover. Further modification of the
AASHTO LRFD Specification is needed to better define the influence of multiple lane loadings
for culvert design.
4-46
Strength I
Sliding
4-47
Bearing
Service I
Overturning
Settlement
4-48
Figure 4-21
Schematic of Student Problem
4-49
Figure 4-22
Retaining Wall Area Designation for Weight of Concrete
(B) Vertical Earth Pressure (EV)
Unit Weight of Soil (1 = 18.835 kN/m3
Weight of Soil on Footing
PEV = W4 = B3 H1 (1 = (2.0 m)(4.5 m)(18.835 kN/m3) = 169.5 kN/m
(C) Live Load Surcharge (LS)
A live load surcharge is applied when vehicle loads will be supported on the backfill within a
distance equal to H. The live load surcharge is applied as an equivalent height of soil for the
design vehicle loading (heq) using Table 4-2, and a wall height of 5 m.
By interpolation, heq = 0.907 m
4-50
Using the unit weight of the soil backfill (i.e., (1 = 18.835 kN/m3), the unit vertical live load
surcharge, LS, over the heel of the wall is:
pLSV = (1 heq = (18.835 kN/m3)(0.907 m) = 17.1 kPa (kN/m2)
For a heel width B3 of 2 m:
PLSV = pLSV B3 = (17.1 kN/m2)(2 m) = 34.2 kN/m of wall length
The active earth pressure coefficient ka for a wall friction angle, * = Nf = 31o, and a horizontal
backslope is, from AASHTO (1997a) Table 3.11.5.3-1:
k = ka = 0.29
From Eq. 4.7, the lateral earth pressure due to the live load surcharge is:
)p = ks (fs heq = ka (f1 heq= (0.29)(18.835 kN/m2)(0.907 m) = 4.95 kPa
Using a rectangular distribution, the live load lateral earth pressure resultant is:
PLS = )p H = (4.95 kN/m2) (5 m) = 24.8 kN/m of wall length
The horizontal and vertical components of the live load lateral earth pressure are:
PLSh = PLS cos = (24.8 kN/m)(cos 31o) = 21.3 kN/m
(Eq. 4-2)
where kh = ka = 0.29
At the base of the footing (i.e., @ z = H):
p = (0.29)(18.835 kN/m3)(5.00 m) = 27.3 kPa (kN/m2)
The resultant of the basic lateral earth pressure (triangular distribution) acting on the wall is:
PEH = Pa = 0.5 pH = (0.5)(27.3 kN/m2)(5.00 m) = 68.3 kN/m length of wall
The horizontal and vertical components of the lateral earth pressure are:
4-51
V
kN/m
Moment Arm
About Toe (m)
Moment About
Toe (kN-m/m)
W1
W2
W3
PEV
PLSV
)PLSv
Pav
TOTAL
Table 4-15
Unfactored Horizontal Loads and Overturning Moments
Item
H
(kN/m)
Moment Arm
About Toe (m)
Moment About
Toe (kN-m/m)
)PLSh
Pah
Step 2: Determine the Appropriate Load Factors
In theory, structures could be evaluated for each of the limit states identified in Section 4.3.
However, depending on the particular loading conditions and performance characteristics of a
structure, only certain limit states need to be evaluated. For the classroom example problem,
each limit state will be qualitatively assessed below relative to that limit state is applicable for
the design problem:
Strength II - Load combination relating to the use of the bridge by Ownerspecified special design vehicles and/or evaluation permit vehicles,
without wind. (Not applicable because special vehicle loading is not
specified).
4-52
Consequently, only the Strength I, Strength IV and Service I Limit States apply to the retaining
wall design. Therefore, from Tables 4-10 and 4-11, select the applicable load factors and
combinations and present them in Table 4-16. (Note: Strength I-a and I-b represent the
Strength I Limit State using minimum and maximum load factors, respectively, from Table 411.)
4-53
Table 4-16
Load Factors
Group
(EV
(DC
(EH
(active)
(LS
Probable Use
Strength I-a
Strength I-b
Strength IV
Service I
Notes: BC - Bearing Capacity; EC - Eccentricity; SL - Sliding
By inspection:
Strength I-a (minimum vertical loads and maximum horizontal loads) will
govern for the case of sliding and eccentricity (overturning)
For the case of bearing capacity, maximum vertical loads will govern, and
the factored loads must be compared for Strength I-b and Strength IV
W1
kN/m
W2
kN/m
W3
kN/m
PEV
kN/m
PLSV
kN/m
)PLSv
kN/m
Pav
kN/m
VTOT
kN/m
V (Unf.)
31.8
10.6
35.3
169.5
34.2
12.8
35.2
329.4
Strength I-a
Strength I-b
Strength IV
Service I
Table 4-18
Factored Horizontal Loads
Group/Item
Units
)PLSh
kN/m
Pah
kN/m
HTOT
kN/m
H (Unf.)
21.3
58.6
79.9
Strength I-a
Strength I-b
Strength IV
Service I
4-54
Table 4-19
Factored Moments from Vertical Forces (Mv)
Group/
Item Units
Mv (Unf.)
W1
W2
W3
PLSV
Pav
PEV
MvTOT
)PLSv
kN-m/m kN-m/m kN-m/m kN-m/m kN-m/m kN-m/m kN-m/m kN-m/m
27.0
6.7
53.0
339.0
68.3
38.4
105.6
638.0
Strength I-a
Strength I-b
Strength IV
Service I
Table 4-20
Factored Moments from Horizontal Forces (Mh)
Group/Item
Units
)PLSh
kN-m/m
Pah
kN-m/m
MhTOT
kN-m/m
Mh (Unf.)
53.3
97.9
151.2
Strength I-a
Strength I-b
Strength IV
Service I
Summary
This example illustrates:
For cantilever retaining walls, dead, earth and live load surcharge are
the predominate loads
Because the load factor for active horizontal (or lateral) earth pressure
in Table 4-11 is the same for all limit states, the controlling limit states
for a typical cantilever retaining wall are generally those for which dead
load and live load surcharge load factors in Table 4-11 are the greatest
(i.e., Strength IV and Strength I, respectively)
4-55
The geotechnical (foundation) design for the retaining wall in this example is presented in the
Classroom Example in Chapter 8, Section 8.4.
4-56
The loads and resulting moments developed by the structural engineer for
design of the bridge pier are reviewed and tabulated
The factored loads and moments for limit states applicable to design of the
pier, as determined by the structural engineer, are tabulated
The critical limit states and load combinations for design of the pier
foundation are established based on a review of the tabulation of factored
design loads and moments from the pier design, and the factored loads and
moments are adjusted to account for the footing or pier cap and overlying
soil
The geotechnical design of spread footing-, pile- and drilled shaft-supported foundations for the
pier are presented in Chapters 8, 9 and 10, respectively.
Step 1: Summarize Unfactored Loads and Moments
The loads which the structural engineer has determined to be applicable to the pier design are
described briefly below, followed by the structural engineer's tabulation of unfactored loads and
moments:
(DC)
(DW)
(LL) (IM)
4-57
LL1
Two lanes fully loaded on both spans with traffic offset to one side
of bridge (transverse loading)
LL2
Two lanes fully loaded on one span and offset to one side of bridge
(biaxial loading)
LL3
One lane fully loaded on both spans and offset to outside of bridge
(transverse loading)
LL4
One lane fully loaded on one span and offset to outside of bridge
(biaxial loading)
Whereas transverse loading results only in a transverse moment and axial load on
the pier stem, biaxial loading also results in a longitudinal moment on the stem.
(BR)
(CE)
Horizontal Thrust
The structural engineer has incorporated an additional horizontal thrust force and
moment on top of the pier associated with loads on the pier cap related to thermal
expansion of the bridge deck support beams. Due to the skew of the pier, this thrust
results in longitudinal and transverse loads and moments.
The unfactored loads and moments acting at the centroid of the base of the pier stem are
summarized in Table 4-21.
4-58
Table 4-21
Unfactored Pier Loads and Moments
Load
Designation
Dead Loads:
DC
DW
Braking
Forces:
BR1
BR2
Wind Case 1:
WSH
WSV
WL
Wind Case 2:
WSH
WL
Thrust:
Hu
Mu
Live Load
LL1
LL2
LL3
LL4
Horizontal Load
(kN)
Moment
(kN-m)
Axial
Load
(kN)
Long.
Trans.
Long.
Trans.
6129
450
0
0
0
0
167
16
0
0
0
0
95
158
24
39
1279
2131
319
531
0
-294
0
32
0
7
180
0
41
304
0
98
1723
772
554
0
0
150
24
105
17
1266
326
886
228
0
0
50
0
9
0
477
53
85
10
1412
1044
848
627
0
0
0
0
0
0
0
0
0
345
0
207
2504
1851
3027
2238
Step 2: Summarize Factored Loads and Moments from Structural Design of Pier Stem
The structural designer has determined that the following limit states are applicable for design of
the bridge pier:
The structural designer has also defined eight applicable load combinations corresponding to the
various possible combinations of live load, braking force and wind load as follows:
Load Case A:
Load Case B:
Load Case C:
Load Case D:
Load Case E:
Load Case F:
Load Case G:
From Eq. 4-1, the factored load effect for each limit state and load combination is determined
using:
Q = i i Q i
(Eq. 4-1)
where:
0i =
(i =
DC
DD
DW
EH
EV
ES
LL
IM
CE
WA WS WL FR
BR
PL
LS
STRENGTH-I
(unless noted)
1.75 1.00
STRENGTH-II
1.35 1.00
STRENGTH-III
STRENGTH-IV
EH, EV, ES, DW
DC ONLY
1.00 1.40
1.00
0.50/
1.20
0.50/
1.00
1.20
0.50/
1.00
1.20
1.50
STRENGTH-V
EXTREME EVENT-I
EXTREME EVENT-II
p
p
EQ 1.00
0.50 1.00
1.00
1.00
SERVICE-I
SERVICE-II
1.00
SERVICE-III
1.00
0.75
TU
CR
SH
EL
1.00
1.00
(1)
0.50/
1.20
0.50/
1.20
1.00/
1.20
1.00/
1.20
1.00/
1.20
-
TG
SE
EQ
IC
CT
CV
TG
SE
TG
SE
TG
SE
TG
SE
TG
SE
TG
SE
The reduced values of ( are used when calculating force effects other than displacements.
Shaded areas represent Limit States evaluated in example problem.
4-60
Limit
State
Max
Min
Max
Min
Max
Min
Max
Min
Max
Min
Max
Min
Max
Min
Max
Min
STR. I
1.07
1.05
1.07
1.05
1.06
1.05
1.06
1.05
1.06
1.05
1.06
1.05
1.06
1.04
1.06
1.04
STR. II
1.05
1.03
1.05
1.04
1.05
1.03
1.05
1.04
1.05
1.03
1.05
1.04
1.05
1.03
1.05
1.04
STR. V
1.06
1.05
1.06
1.05
1.06
1.04
1.06
1.04
1.06
1.04
1.06
1.04
1.06
1.04
1.06
1.04
4-61
The factored load and moment effects for the various load combinations and limit states
applicable to design of the pier stem are summarized in Table 4-25. Example calculations for the
Load Case A Strength I and Service I Limit States follow.
Table 4-25
Summary of Factored Loads and Moments in Pier Stem
MAXIMUM LOADS
LOAD CASE
(1)
STR I
STR III
STR V
SER I
STR I
STR III
STR V
SER I
STR I
STR III
STR V
SER I
STR I
STR III
STR V
SER I
STR I
STR III
STR V
SER I
STR I
STR III
STR V
SER I
STR I
STR III
STR V
SER I
STR I
STR III
STR V
SER I
Moments
Long. Trans.
(kN-m) (kN-m)
4772
5632
1046
3723
3982
5314
2912
3803
4772
5632
2537
1358
4514
4825
3269
3454
5427
4432
1046
3723
4488
4388
3257
3150
5427
4432
2537
1358
5019
3899
3614
2801
3080
6203
1046
3723
2683
5754
2060
4114
3080
6203
2537
1358
3212
5266
2417
3765
3475
4754
1046
3723
2988
4636
2267
3325
3475
4754
2537
1358
3517
4148
2624
2976
Axial
Load
(kN)
11 348
8321
10 755
7991
11 348
8753
10 755
7991
10 671
8321
10 233
7623
10 671
8753
10 233
7623
10 311
8321
9955
7427
10 311
8753
9955
7427
9905
8321
9642
7206
9905
8753
9642
7206
MINIMUM LOADS
Horiz. Load
Long.
(kN)
317
74
267
220
317
247
324
261
317
74
267
220
317
247
324
261
201
74
178
157
201
247
234
198
201
74
178
157
201
247
234
198
Trans.
(kN)
77
270
153
115
77
159
112
85
77
270
153
115
77
159
112
85
49
270
132
100
49
159
90
70
49
270
132
100
49
159
90
70
Moment
Long. Trans.
(kN-m) (kN-m)
4171
5096
862
3369
3475
4808
2912
3803
4177
5097
2189
1229
3948
4366
3269
3454
4761
4010
862
3369
3926
3970
3257
3150
4761
4010
2189
1229
4398
3528
3614
2801
2671
5613
862
3369
2318
5214
2060
4114
2671
5613
2189
1229
2790
4765
2417
3765
3024
4302
862
3369
2591
4195
2267
3325
3024
4302
2189
1229
3061
3753
2624
2976
Axial
Load
(kN)
7866
5128
7330
7991
7866
5519
7330
7991
7254
5128
6858
7623
7254
5519
6858
7623
6928
5128
6606
7427
6928
5519
6606
7427
6561
5128
6323
7206
6561
5519
6323
7206
Note: Tabulated and calculated values may differ slightly due to round-off.
4-62
Horiz. Load
Long.
(kN)
287
67
242
220
287
224
294
261
287
67
242
220
287
224
294
261
182
67
161
157
182
224
212
198
182
67
161
157
182
224
212
198
Trans.
(kN)
70
245
139
115
70
144
102
85
70
245
139
115
70
144
102
85
45
245
120
100
45
144
82
70
45
245
120
100
45
144
82
70
Note: WSv neglected due to its negative value. WSv would only be included if investigating
overturning of the bridge.
4-63
q = Q/Ar
where the net effective bearing area, Ar, is computed as:
Ar = [(B - 2 eB)(L - 2 eL)] = [(1.5 m - 2 eB)(3.5 m - 2 eL)]
Values of the factored axial load and factored average bearing pressure at the base of the pier
stem (for B=1.5 m and L=3.5 m) are summarized in Table 4-26. The eccentricities are computed
as:
eB = Longitudinal Moment/Axial Load
eL = Transverse Moment/Axial Load
Table 4-26
Summary of Factored Axial Loads and Average
Bearing Pressures at Base of Pier Stem
MAXIMUM LOAD
LOAD CASE
F
G
STR. I
STR. III
STR. V
SER. I
STR. I
STR. III
STR. V
SER. I
STR. I
STR. III
STR. V
SER. I
STR. I
STR. III
STR. V
SER. I
STR. I
STR. III
STR. V
SER. I
STR. I
STR. III
STR. V
SER. I
STR. I
Axial
Load
(kN)
11 348
8321
10 755
7991
11 348
8753
10 755
7991
10 671
8321
10 233
7623
10 671
8753
10 233
7623
10 311
8321
9955
7427
10 311
8753
9955
7427
9905
eB
(m)
eL
(m)
Ar
(m2)
q
(kPa)
0.421
0.126
0.370
0.364
0.421
0.290
0.420
0.409
0.509
0.126
0.439
0.427
0.509
0.290
0.490
0.474
0.299
0.126
0.270
0.277
0.299
0.290
0.323
0.325
0.351
0.496
0447
0.494
0.476
0.496
0.155
0.449
0.432
0.415
0.447
0.429
0.413
0.415
0.155
0.381
0.367
0.602
0.447
0.578
0.554
0.602
0.155
0.529
0.507
0.480
1.650
3.252
1.909
1.967
1.650
2.935
1.717
1.798
1.287
3.252
1.643
1.727
1.287
2.935
1.424
1.527
2.071
3.252
2.250
2.263
2.071
2.935
2.085
2.113
2.027
6876
2559
5633
4062
6876
2982
6263
4445
8292
2559
6227
4413
8292
2982
7187
4993
4979
2559
4424
3282
4979
2982
4774
3515
4887
4-64
MINIMUM LOAD
Axial
Load
(kN)
7866
5128
7330
7991
7866
5519
7330
7991
7254
5128
6858
7623
7254
5519
6858
7623
6928
5128
6606
7427
6928
5519
6606
7427
6561
eB
(m)
eL
(m)
Ar
(m2)
q
(kPa)
0.531
0.168
0.474
0.364
0.531
0.397
0.539
0.409
0.656
0.168
0.573
0.427
0.656
0.397
0.641
0.474
0.386
0.168
0.351
0.277
0.386
0.397
0.422
0.325
0.461
0.648
0.657
0.656
0.476
0.648
0.223
0.596
0.432
0.553
0.657
0.579
0.413
0.553
0.223
0.514
0.367
0.810
0.657
0.788
0.554
0.810
0.223
0.721
0.507
0.656
0.965
2.545
1.208
1.967
0.965
2.156
0.974
1.798
0.450
2.544
0.829
1.727
0.450
2.156
0.539
1.527
1.369
2.545
1.535
2.263
1.369
2.156
1.350
2.113
1.265
8148
2015
6068
4062
8148
2560
6808
4445
16 117
2015
8271
4413
16 117
2560
12 725
4993
5062
2015
4302
3282
5062
2560
4893
3515
5188
STR. III
STR. V
SER. I
STR. I
STR. III
STR. V
SER. I
8321
9642
7206
9905
8753
9642
7206
0.126
0.310
0.315
0.351
0.290
0.365
0.364
0.447
0.481
0.461
0.480
0.155
0.430
0.413
3.252
2.233
2.243
2.027
2.935
2.033
2.064
2559
4317
3213
4887
2982
4743
3491
5128
6323
7206
6561
5519
6323
7206
0.168
0.410
0.315
0.461
0.397
0.484
0.364
0.657
0.663
0.461
0.656
0.223
0.594
0.413
2.545
1.478
2.243
1.265
2.156
1.230
2.064
2015
4277
3213
5188
2560
5140
3491
(I)
Deep Foundations:
The critical loading conditions for evaluation of the bearing resistance and settlement of a
group of piles or drilled shafts are combinations of axial load and moment which produce the
maximum axial stresses in the piles or shafts.
For design at the Strength Limit State, the critical combinations from Table 4-25 are:
For design at the Service Limit State, the critical combinations from Table 4-25 are:
Case A, Service I Limit State & Case E, Service I Limit State produce
maximum combinations of axial load and moment in the transverse
direction, respectively:
Case A, Service I: Q = 7991 kN; MT = 3803 kN-m
Case E, Service I: Q = 7427 kN; MT = 4114 kN-m
Case B, Service I Limit State & Case D, Service I Limit State produce
maximum combinations of axial load and moment in the longitudinal
direction, respectively:
Case B, Service I: Q = 7991 kN; ML = 3269 kN-m
Case E, Service I: Q = 7623 kN; ML = 3614 kN-m
(II)
Spread Footing Foundations:
The critical loading conditions for evaluation of bearing resistance and settlement of spread
footing foundations are those which produce the maximum factored and unfactored bearing
pressures.
For design at the Strength Limit State, the critical load combination from Table 4-26 is:
For design at the Service Limit State, the critical load combination from Table 4-26 is:
(B)
Overturning:
(I)
Deep Foundations:
Overturning failure is not typically evaluated for deep foundation systems for bridge piers.
Overturning would be considered only for structures subjected to extreme uplift and/or
horizontal loads which could result in net tension loading of deep foundation elements.
(II)
Spread Footing Foundations
The critical loading condition for spread footing foundations are those which produce the
maximum base pressure resultant eccentricity. This failure mode is checked only for the
Strength Limit State.
The critical loading conditions, from Table 4-26, are:
(C)
(I)
Deep Foundations
The critical loading conditions for lateral loading of deep foundation groups are generally
combinations of maximum horizontal load, moment and axial load which produce the greatest
foundation element stresses and lateral deflections.
For Strength Limit State Design, the critical load combinations from Table 4-25 are:
Cases A/C, Strength III, Maximum Loads - produce most severe loading
in the transverse direction:
Horizontal load = 270 kN
For Service Limit State Design, the critical load combinations from Table 4-25 are:
4-66
(II)
Spread Footing Foundations
The critical loading conditions for sliding of spread footing foundations are those which produce
the greatest horizontal loads and horizontal to vertical load ratios. This failure mode is checked
only for the Strength Limit State. The critical load combinations for sliding of spread footing
foundations, from Table 4-25, are:
(E)
Evaluation Criteria
Bearing Resistance
Lateral Load Resistance
Settlement and Lateral
Deflection
4-67
Table 4-28
Critical Load Combinations for Spread Footing Foundations
Critical Load Combinations
Evaluation Criteria
Bearing Resistance
Overturning (Eccentricity)
Sliding
Settlement
EV = 206 kN
The factored earth pressure on the footing is:
QS = 0 ((EV EV)
For an importance factor 0I = 1.05 and a maximum earth load factor, (EV = 1.35:
QS = 1.05(1.35)206 kN = 292 kN
For an importance factor (0) of 0.95 and a minimum load factor of 0.90:
QS = 0.95(0.90)206 kN = 176 kN
For design at the Service Limit State:
QS = 206 kN
(III) Summary of Factored Loads for Possible Critical Loading Combinations
For this example, it is assumed that the pier stem is centered on the footing. Therefore, the
footing and overlying soil weight impart no unbalanced loading such that they only increase the
vertical load at the base of the footing. The factored loads from Table 4-25 for the possible
critical loading combinations in Tables 4-27 and 4-28, adjusted for footing weight and
embedment, are summarized in Table 4-29.
Table 4-29
Summary of Factored Loads for Critical Foundation Design Load Combinations
MAXIMUM LOADS
LOAD CASE
A
B
C
D
E
F
STR I
STR III
SER I
STR I
SER I
STR I
STR III
STR I
STR V
SER I
STR I
SER I
STR I
Moments
Long. Trans.
(kN-m) (kN-m)
4772
5632
1046
3723
2912
3803
4772
5632
3269
3454
5427
4432
1046
3723
5427
4432
----3614
2801
----2060
4114
-----
Axial
Load
(kN)
12 644
9617
8962
12 644
8960
11 967
9617
11 967
--8594
--8398
---
MINIMUM LOADS
Horiz. Load
Long.
(kN)
317
74
220
317
261
317
74
317
--261
--157
---
Trans.
(kN)
77
270
115
77
85
77
270
77
--85
--100
---
Moment
Long. Trans.
(kN-m) (kN-m)
----862
3369
------------4761
4010
862
3369
4761
4010
4398
3528
----2671
5613
----2671
5613
Axial
Load
(kN)
--5958
------8084
5958
8084
7688
--7758
--7758
Horiz. Load
Long.
(kN)
--67
------287
67
287
294
--182
--182
Trans.
(kN)
--245
-----70
245
70
102
--45
--45
The design of a pile foundation to support the pier in this example is described in Chapter 9,
Design Example 1, Section 9.6.
4-70
CHAPTER 5
GEOTECHNICAL SITE CHARACTERIZATION
5.1 Introduction
Geotechnical characterization of sites for bridge and other highway structures is an integral part of
the design process. Proper site characterization has always been needed for the design of highway
substructures, and the evaluation of prospective structure sites should be performed as a
collaborative effort between the structure designer, geotechnical engineer, and for structures near
rivers and streams, the hydraulic engineer. The LRFD concepts presented herein for the design of
foundations and retaining walls were developed and calibrated based on the use of specific types of
information to define the subsurface conditions and engineering properties of soil and rock materials
and the use of resistance factors consistent with the reliability of the information. This chapter
provides some of the information essential to any refinement of this calibration process along with a
brief review of the exploration and testing methods which can be used to develop the geotechnical
information needed for design.
The following sections of this chapter:
y
Describe the process and basis for planning exploration and testing programs
for substructure design using LRFD
5.2.1 General
Subsurface exploration and testing are required for highway structures to:
y
Aid in the preliminary selection of substructure types which are viable for a
particular site, and for bridges, the type of superstructure to be supported;
Provide a basis for selecting soil and rock properties needed for substructure
design; and
Examples of the latter include the presence of elevated ground water levels which could affect the
stability of excavations during construction, and the presence of corrosive materials which could
affect the service life of a structure.
For most highway projects, subsurface data are acquired during the planning stages when the
5-1
feasibility of particular routes and structure types are being evaluated, and during the design stage
when localized site-specific data are required. Additional data regarding subsurface conditions are
not usually required after the design stage unless the need arises to resolve questions and/or install
construction-phase instrumentation and monitoring. For most projects, the following type of
subsurface information is needed for the selection, design and construction of highway
substructures:
y
y
y
y
y
y
y
y
Because it is not possible to develop criteria for selecting the type, location and frequency of
subsurface exploration and testing applicable for all sites and types of planned construction, it is
important that the project team include geotechnical engineers, engineering geologists and geologists
so that their perspective is available to the team.
Geotechnical load and resistance factor design requires an initial "calibration" of the relationship
between the applied loads and available resistance. The predictive model(s) and accompanying insitu or laboratory tests used to estimate the resistance of soil and rock are inherent in the
determination of an appropriate resistance factor in this calibration process. For the present state of
practice, this calibration process has used available data bases covering the types of exploration and
testing programs usually performed. Consequently, the resistance factors recommended by the
AASHTO LRFD Specifications are not related in a precise manner to variables such as the number
of borings, exploration depth, or boring spacing, but instead reflect the standard of care
representative of each data set considered. To a large extent it will never be possible to rationally
specify a single set of requirements for exploration programs due to the range of subsurface
conditions which may be encountered. In the future, additional research utilizing comprehensive
exploration and testing programs may allow evaluations of the best means to optimize and quantify
exploration program requirements.
5.2.2 Soil and Rock Variability
In developing exploration and testing programs, the geotechnical engineer should qualitatively
assess the effects of variables such as the expected type and importance of the structure, magnitude
and rate of loading, and viability of foundation alternatives relative to technical, economic and
constructibility considerations. From the planning stage, information regarding land use and
topographic, surficial soil and geologic conditions can be used to define guidelines for developing
subsurface exploration and testing programs. Sources of information available for many sites
include:
y
Well Drilling Logs - General description of soil and rock, and ground water
levels at the time of drilling
For example, given the project information that a heavily-loaded bridge pier is to be located adjacent
to an old river with low velocity might suggest the need to use some type of deep foundation
systems due to the likely presence of relatively soft, fine-grained alluvial soils. With this
information, the geotechnical engineer can determine the types and number of in-situ and laboratory
tests needed for the expected geology, foundation type(s) to be used, load conditions to be evaluated
(i.e., undrained versus drained loading) and stress history of the foundation soils. A drained analysis
should be based on effective stress parameters and an undrained analysis should be based on total
stress parameters. The stress history of the foundation soils is also an important factor in the design
process. For example, whereas the stability of an embankment constructed on a lightly
overconsolidated clay will be controlled by undrained behavior representative of the short-term case,
a similar embankment built over a deposit of heavily over consolidated clay will probably be
controlled by drained behavior representative of the long term case. Accordingly, selection of an
appropriate in-situ test method must reflect these considerations so that the test program will yield
the necessary information about the subsurface conditions. For example, whereas a SPT can be
expected to provide reasonable information about the effective stress friction angle for a clean sand,
it has almost no value in evaluating the undrained shear strength of a clay. Likewise, a CPT will not
provide meaningful information in a deposit of coarse gravel.
At present, the variability of subsurface conditions and the level of subsurface exploration (i.e.,
number of borings) or testing (i.e., number of in-situ or laboratory tests) are not explicitly related to
resistance factors used in the AASHTO LRFD Specification. Instead, planning subsurface
exploration and testing programs are based on guidelines such as those recommended by FHWA
(1988) in Table 5-1, the availability of information from previous explorations in the vicinity of the
site, and/or engineering experience and judgment.
A simple quantitative measure of the variability of data or an engineering property is the coefficient
of variation, COV, which is defined as:
COV = / x
where:
COV = Coefficient of variation (dim);
=
Standard deviation of the data; and
x =
Mean value of data.
5-3
(Eq. 5-1)
Structure Foundation
Retaining Walls
Culverts
Bridge Approach
Embankments Over Soft
Ground
Borings alternatively spaced every 30 to 60 m in Extend borings to depth of 2 times wall height or a minimum of 3 m
front of and behind wall.
into bedrock.
Two borings depending on length.
The greater the value of COV, the less reliable the data. An example of how the COV can be used
as a guide in planning a subsurface exploration program is illustrated in Figure 5-1.
Figure 5-1
Reliability Variation with Sample Size for Indirect Testing
for Mobilized Undrained Shear Strength
(after Teng, et al., 1992)
Figure 5-1 relates the COV of the mobilized undrained shear strength, Su, estimated using cone
penetrometer (CPT) and field vane shear (VST) in-situ tests with the number of samples tested for a
slope stability problem (Teng, et al., 1992). These in-situ tests are compared with a correlation of Su
with the preconsolidation stress, 'p, as determined by laboratory consolidation tests. As expected,
the COV decreases with increasing numbers of samples until a limiting value is reached. The figure
also shows that the field VST is a more reliable method for estimating Su than the CPT, and that the
field VST and laboratory test methods provide comparable results. In the future, it is conceivable
that relationships such as those shown in Figure 5-1 could be used in designing site exploration
programs with the type and number of tests selected to achieve a specified T.
In addition to considerations regarding the type and extent of exploration, another important factor
in planning any subsurface exploration is the cost-benefit relationship of the exploration program
relative to construction cost. In general, as the amount and quality of subsurface exploration
increases, the uncertainties and resulting conservatisms in the design process decrease. Conversely,
an inadequate geotechnical exploration program can result in significant project cost overruns during
construction as shown in the Table 5-2 which summarizes a recent study of 58 highway projects in
the United Kingdom.
5-4
Table 5-2
Comparison of Ground Investigation Cost to Project Cost Overruns
(Modified after Whyte, 1995)
Ground Investigation Cost/
Mean Cost Increase Due to
Total Project Cost
Geotechnical Origins
(%)
(% of Total Project Cost)
# 1.5
14
1.5 - 2.0
8
> 2.0
4
The process of optimization of subsurface exploration with respect to the project foundation costs is
illustrated from a conceptual perspective in Figure 5-2.
Figure 5-2
Optimization of Foundation and Exploration Cost
(after Kulhawy, et al., 1983)
5-5
Figure 5-2a shows the efficiencies of scale which may be realized by an expanded exploration
program, and Figure 5-2b shows the expected trend of decreasing foundation cost with the
increasing reliability associated with greater exploration. These two concepts are combined in
Figure 5-2c which shows the optimization of the combined costs of exploration and foundations.
While it may never be possible to quantify the combined costs to establish an optimal level of
exploration, the concepts of reliability-based design provide a more rational framework for
developing economical subsurface exploration programs than is currently practiced.
Optimization of the exploration program should also consider the reliability of the different methods
available for engineering property assessment of soil and rock. The three primary sources of error
which contribute to the uncertainty of material resistance estimates were identified in Chapter 3 as:
y
To develop a resistance factor, for a particular design approach (e.g., bearing resistance of a spread
footing on sand using SPT) or subsurface conditions, these sources of uncertainty must be combined
with uncertainties in load and the level of safety required. As described in Section 3.4.6 of Chapter
3, these factors can be combined to develop a -factor for design using:
R ( D
=
2
2
1 + COVQD
+ COVQL
QD
+ L)
1 + COV 2R
QL
Q
2
2
(QD D + QL) exp [T ln(1 + COV 2R )(1 + COVQD
+ COVQL
)]
QL
(Eq. 5-2)
R =
COVR =
(Eq. 5-3)
where:
COVMODEL =
Coefficient of variation of predictive model (dim)
COVMEASUREMENT = Coefficient of variation of property measurement (dim)
Coefficient of variation of inherent soil variability (dim)
COVINHERENT =
Figure 5-3 shows the relationship between and COVR for QD/QL = 2.0 (i.e., characteristic of a
medium-span structure) using Eq. 5-2. Representative values of the other variables are presented on
the figure and were developed from information presented in Chapter 3.
Figure 5-3
Relationship Between and COVR for QD/QL = 2.0 Using Eq. 5-2
5-7
The following section provides information about the inherent variability of important soil and rock
properties and measurement errors resulting from various in-situ and laboratory test methods.
Model error is discussed in Chapter 6.
5.3 Field Test Methods
A variety of methods are available for the design of foundations and walls based on the results of
field tests. The results of these tests can be used to estimate engineering properties (e.g., strength or
compressibility) needed for analysis or design, or used directly in semi-empirical design methods for
design (e.g., ultimate axial bearing resistance of a driven pile). The most commonly used in-situ test
methods for geotechnical site characterization and foundation design include:
y
y
y
y
This section summarizes the devices and procedures used for each method and discusses the
relative reliability and sources of error in each.
The resistance factors presented in the chapters for foundation design take into account different
in-situ test methods and reflect the inherent reliability of each test type. Table 5-3 shows the
coefficient of variation (COV) of the three sources of error for each test method, total COV
range for the test method and the most likely range in COV for each test method when
reasonable care is exercised in performance of the test. As shown in this table, the test results of
the SPT are likely to be less reliable than those of the VST or ECPT due to the larger inherent
variability in the SPT. This expectation is well reflected in the resistance factors presented in
later sections on foundations, which recommend using = 0.35 for bearing capacity based on
SPT derived friction angles and = 0.45 for CPT derived friction angles. The data shown in
Table 5-3 can be used to calibrate models not presented herein.
Table 5-3
Estimates of In-Situ Test Variability
(after Orchant, et al., 1988)
Test
SPT
MCPT
ECPT
VST
PMT
COV
Equipment
0.05 (3) - 0.75 (4)
0.05
0.03
0.05
0.05
COV
Procedure
0.05 - 0.075
0.10 (5) - 0.15 (6)
0.05
0.08
0.12
COV
Random
0.12 - 0.15
0.10 - 0.15
0.05 (5) - 0.10 (6)
0.10
0.10
5-8
COV (1)
Total
0.14 - 1.00
0.15 - 0.22
0.07 - 0.12
0.14
0.16
COV (2)
Range
0.15 - 0.45
0.15 - 0.25
0.05 - 0.15
0.10 - 0.20
0.10 - 0.20 (7)
(1)
COV (Total) = [COV (Equipment)2 + COV (Procedure)2 + COV (Random)2]1/2; (2) Due to limited data
and judgment used to estimate COV, ranges represent probable magnitudes of test measurement error
when reasonable care is exercised in performing test; (3) Best case scenario for SPT test conditions; (4)
Worst case scenario for SPT test conditions; (5) Tip resistance CPT measurements; (6) Side resistance CPT
measurements; (7) Results may differ for po, pf, and pL, but data are insufficient to clarify this issue
The following sections present a brief description of each of these test methods, as well as a more
detailed discussion of the potential sources of measurement error and variability for each method.
This discussion of each test is prefaced by more general comments regarding the applicability of the
tests.
5.3.1 Split-Barrel Sampling
In the U.S., split-barrel (or split-spoon) sampling is the most common method of in-situ testing and
disturbed sampling of soils. As shown in Figure 5-4, the steel split-barrel sampler consists of a 51
mm O.D. (38 mm I.D.) by 457 mm long tube which is split longitudinally to permit opening of the
sampler and removal of soil retained within the sampler.
A = 25 to 50 mm; B = 0.457 to 0.762 m; C = 34.93 " 0.13 mm; D = 38.1 " 1.3 - 0.0 mm;
E = 2.54 " 0.25 mm; F =50.8 " 1.3 - 0.0 mm; G = 16.0E to 23.0E
Figure 5-4
Split-Barrel Sampler
(ASTM, 1997)
The sampler is advanced into the ground by impacting the drill rods with a 620 Newton (N) weight
falling from a height of 760 mm. Samples can be obtained continuously, although samples taken at
intervals of 0.9 to 1.5 m are common. At shallow depths (e.g., 1.5 to 3 m), samples can be obtained
by driving the sampler into the open borehole created by the sampler. For greater depths, samples
are usually obtained by advancing the borehole using hollow-stem flight augers which provide
access for the sampler and drill rods through the hole in the augers. In some parts of the U.S., the
borehole is advanced by solid-flight augers and drilling mud to maintain the stability of the borehole.
5-9
When samples are obtained, the sampler is driven into the ground and the number of blows required
to drive the sampler are recorded in three 150 mm intervals. The sum of blow counts for the last two
150 mm intervals is referred to as the Standard Penetration Test (SPT) blow count or "N"-value.
The blow count is used to estimate the in-situ relative density of cohesionless soils and as a direct
input to various design methods (e.g., for estimating footing settlement on sandy soils and axial
capacity of driven piles). The standard method for the SPT is described in ASTM D 1586 (1997).
Because the soil retained in the sampler is disturbed, the samples are used only for index laboratory
tests such as moisture contents, Atterberg limits, and particle-size analyses. The soil samples are not
for tests to estimate in-situ soil shear strength or compressibility.
Several of the major sources of error that contribute to the variability in results due to equipment,
procedure, and random variability are listed in Table 5-4. The ASTM standard for Standard
Penetration Testing suggests that variations in N-values greater than 100 percent have been reported
for separate drillers and different equipment working closely in the same soil layer, but current
opinion indicates a COV of about 10 percent for a single apparatus and driller.
Table 5-4
Selected Major Sources of Error in the Standard Penetration Test
(after Kulhawy and Mayne, 1990)
Cause
Effect
Influence on N Value
Failure to maintain adequate
Borehole bottom may become quick
Decreases
head of water in borehole
Hammer energy varies (5 to 7%
Hammer weight inaccurate
Increases or Decreases
common)
Lack of hammer free fall
Hammer energy reduced
Increases
Sampler driven above casing
Sample driven in disturbed,
Increases greatly
bottom
artificially densified soil
Careless blow count
Inaccurate results
Increases or Decreases
Coarse gravel, cobbles, shells Sampler becomes clogged or
Increases
or cemented sands
impeded
5.3.2 Cone Penetration Testing
The cone penetration test (CPT) involves pushing a cone-tipped device similar to that shown in
Figure 5-5 into soil at a rate of 10 to 20 mm per second. Due to the need to push the device at a
continuous rate, a special load apparatus is required, although it is possible to use a drill rig adapted
with special attachments and operated by a skilled work crew. Cone penetration development began
with a mechanical device which has largely been replaced with electrical devices such as that shown
in Figure 5-5. For electrical cones, the maximum force is measured using a digital gage or strip
chart attached by leads to strain-gaged sensors located in the tip and friction sleeve. Recent
developments in CPT equipment permits measurement of pore water pressure, temperature, shear
wave velocity, and some types of soil and ground water contaminants. The standard method for the
CPT is fully described in ASTM D 3441 (1997). Additional details regarding application of CPT
results for foundation design are presented in Riaund and Miran (1991).
5-10
5 Adjustment ring
6 Waterproof bushing
7 Cable
8 Connection with rods
Figure 5-5
Electric Friction-Cone Penetrometer Tip
(ASTM, 1997)
The results of the CPT can be used to continuously log and classify soil strata, estimate soil strength
and compressibility, and to directly design shallow and deep foundations. Because no sample is
retrieved from a CPT, some engineers do not use the test. Although the device is relatively rugged,
CPT testing is difficult in very dense cohesionless soils, and in very stiff to hard cohesive soils due
to the large reaction load needed to advance the cone.
The ASTM standard for cone penetration testing indicates that a standard deviation of 5 percent in
tip resistance and 10 percent in side resistance results for electrical cones is typical when CPTs are
conducted by experienced personnel. Some of the major sources of variability associated with the
cone penetration test are shown in Table 5-5.
At present, the AASHTO LRFD Specification resistance factors have not been refined to the state
where effects on testing such as those listed in Table 5-5 can be quantified in selection of an
appropriate resistance factor. Some of the resistance factors for cone penetration testing in the
AASHTO LRFD Specification are statistically based and include consideration of a range of
potential reliability of the CPT, but other resistance factors are still simply a product of correlations
with current ASD practice. In the future, compilation and analysis of large, detailed data sets may
allow more rational selection of resistance factors and may incorporate the effects of more variables
on testing procedure.
5-11
Table 5-5
Major Sources of Error in the Cone Penetration Test
(after Kulhawy and Mayne, 1990)
Cause
Gravel, cobbles, shells or
cemented sands
Worn penetrometer tip
Leaky water seal
Improper calibration
Partial saturation of porous
stone
Effect
Impedes penetration of cone tip
Causes deviations in instrument
verticality
Tip may become dull and/or
surface roughness may become
greater or lesser than standard
Electrical transducers may
become corroded
Inaccurate measurements
Inaccurate measurements
Influence on Results
Increases qc greatly
Increase or decrease qc and fs
Increases or decreases qc and fs
slightly.
Increases or decreases qc and fs
Increases or decreases qc and fs
Decreases pore pressure
magnitude
Figure 5-6
Field Vane Geometries and Sizes
(ASTM, 1997)
When performed correctly, the VST can provide one of the best measures of the in-situ undrained
shear strength for soft or sensitive clays which are difficult to sample. While ASTM does not
provide any commentary about the reliability of this test procedure, Orchant, et al. (1988) suggests a
total COV of 0.14 for the VST. Table 5-6 presents some of the major sources of error in the test.
5-12
Table 5-6
Major Sources of Error in the Vane Shear Test
(after Kulhawy and Mayne, 1990)
Cause
Friction between torque rods and
soil or casing
Poorly calibrated torque
measurement
Vane rotated too quickly
Test performed in disturbed soil
Damaged vane
Unknown sand/silt/shell lenses
Isolated gravel/cemented nodules
Effect
Measured torque includes spurious
component of resistance
Inaccurate torque
Soil sheared too rapidly
Soil structure broken down
Disturb soil excessively or shear
only partial surface
Drainage during test
Measured torque includes spurious
component of resistance
Influence on Strength
Measurement
Increases
Increases or decreases
Increases
Decreases
Decreases
Increases
Increases
For tests performed in a cased hole, it is very important to limit the stress relief at the bottom of the
hole by maintaining an adequate head of water or drilling mud to prevent disturbance near the hole
bottom.
Again, many of the AASHTO LRFD resistance factors applicable to the vane shear test do not have
a firm statistical basis. However, it is likely that sufficient data on the test itself exists to provide
this basis and simply needs to be collected and analyzed.
5.3.4 Pressuremeter Test
The pressuremeter test is conducted by advancing a cylindrical probe into the ground and expanding
the probe radially while measuring the changing probe volume and the pressure necessary for
expansion. The probe can be inserted into a pre-bored hole as shown in Figure 5-7, or advanced by a
self-boring mechanism which is an integral part of the device. ASTM D 4719 (1997) describes the
testing process in detail. Once the probe is installed to the test depth, the PMT is performed by
expanding the guard cells against the sides of the borehole and incrementally increasing the
pressure in the measuring cell and measuring the change in volume or lateral displacement of the
cell. Although numerous measurements of pressure and volume change are made during the test,
key measurements are made at the beginning and end of the linear portion of the volume versus
pressure curve as well as the limit pressure asymptote which is approached at the end of the test.
5-13
Figure 5-7
Menard Pressuremeter Equipment
(after Kulhawy and Mayne, 1990)
The PMT gives information which can be used to estimate the in-situ stress state, strength and
stiffness of the soil, as well as provide information used in direct correlations with foundation
capacity and settlement. Table 5-7 lists some of the major sources of error associated with the PMT.
While the ASTM standard does not provide an evaluation of the test precision, it does emphasize
the importance and difficulty of preparing a good borehole for each test.
Table 5-7
Major Sources of Error for the Pressuremeter Test
(after Kulhawy and Mayne, 1990)
Variable
Relative Effect on Test Results
Expansion of Tubing
Minor to moderate
Probe dimensions
Minor to moderate
Membrane aging
Minor
Method of drilling and borehole
Significant
preparation
Rate of probe inflation
Minor to moderate
5.3.5 Other Methods
Other methods available for in-situ characterization include undisturbed sampling, rock coring and
ground water level measurement. Although there are no quantitative measures of the reliability of
these methods, they are discussed herein because the reliability of laboratory methods rely on the
quality of soil and rock samples used for testing and accurate measurement of the ground water level
is an important aspect of substructure design.
5-14
Figure 5-8
Thin-Walled Tube Sampler
(Modified after ASTM, 1997)
Thin-walled tube sampling does not produce a numerical result, and therefore cannot be described
statistically. However, the quality of the sample taken can have a significant effect on the results of
any subsequent testing of the sample. For soils particularly sensitive to disturbance effects, it may
be desirable to evaluate the sample quality (e.g., by X-raying) of each tube.
5.3.5.2 Rock Coring
Rock core drilling using single-, double-, or triple-tube core barrels with diamond end bits is used to
advance borings in rock when the resistance to driving a split-barrel sampler is greater than
about 50 blows per 150 mm. A variety of core barrel sizes is available, but the most common size
used is the NW-size core barrel which has an O.D. of 75.4 mm and an I.D. of 54.8 mm. The length
of a core barrel is either 1.5 or 3.0 m. Double-tube core barrels such as shown in Figure 5-9 are
most commonly used, although triple-tube core barrels are sometimes used when weak rock is
drilled and high quality rock core is required for testing. A single-tube core barrel should not be
used due to the poor quality of rock that is often recovered.
5-15
Figure 5-9
Double-Tube Core Barrel
(Fang, 1991)
The core barrel is attached to a string of drill rods, and the rods are attached to the kelly bar of a drill
rig. Rock cutting is accomplished by a shoe or bit which is attached to the end of the core barrel and
containing embedded industrial diamonds. Rock cuttings are removed from the borehole and the
diamond bit is cooled by water which is flushed through the drill string, into the core barrel and up
the sides of the borehole to the ground surface. Torque and thrust to the core barrel are provided by
a drill rig. Coring is accomplished by inserting the core barrel and drill string to the rock surface
within hollow-stem augers or casing sealed into the top of rock surface. Core is retained in the core
barrel using a spring-like retainer inside the sampler. Special core barrels known as wireline
samplers permit retrieval of the rock core from the barrel without removal of the drill rods from the
boring by use of an inner liner of the core barrel which is lifted through the special drill rod using a
steel wire line. When the full or partial length of coring is completed, the barrel and drill string are
removed from the borehole and the sample is removed and usually placed in a core box. The
standard method for diamond core drilling is fully described in ASTM D 2113 (1997).
5.3.5.3 Ground Water Location
Definition of ground water highs and lows over time is critical for the design process due to the
importance of ground water during construction and its effect on the shear strength of the foundation
medium. The design ground water level should reflect any possible seasonal variations and its
selection should include consideration of direct ground water readings taken from borings and
piezometers, as well as information inferred from soil coloration changes (e.g. mottling) observed
during drilling and any other available records (e.g., adjacent river level fluctuations).
5.4 Laboratory Test Methods
5.4.1 Soil Index Testing
Soil classification is accomplished by a combination of visual identification and selective laboratory
index testing to confirm the visual classification. The most commonly used method for classifying
soils is the Unified Soil Classification System (USCS) as presented in ASTM D 2488 (1997) for
visual identification and ASTM D 2487 (1997) for classification based on the results of laboratory
index testing. To develop a soil classification using the USCS with laboratory tests, the particle-size
distribution and Atterberg (i.e., liquid and plastic) limits of the soil are needed. Particle-size
analyses provide the maximum size of particle and the size distribution of soil particles in the
sample, and the liquid and plastic limits are used to classify the character of the fine soil fraction
which is less than 0.425 mm in size. In addition to their usefulness in classifying soils, these tests
can be used with the results of other index tests (i.e., moisture content, specific gravity and unit
weight) to provide preliminary estimates of the possible engineering behavior of soils (e.g., Carter
5-16
and Bentley, 1991). For example, the Atterberg limits and natural moisture content can be used as a
guide to the consolidation and undrained shear strength behavior of cohesive soils, and the
particle-size distribution of cohesionless soils can be used to estimate permeability. Because index
tests are inexpensive and can be completed relatively quickly, it is useful to obtain and evaluate the
results of these tests at the initial stages of the testing program to aid in selecting samples for
engineering property testing.
Table 5-8 presents the inherent variability and measurement variability for several of the most
commonly performed index tests for soil.
5.4.2 Rock Index Testing
The most commonly used index tests for characterizing rock behavior include the compressional
wave velocity, Vp, Schmidt hammer rebound, R, and the point load test, Is. These measurements are
commonly used in correlations with mechanical properties such as tensile strength, To, unconfined
compressive strength, Co, and the Young's modulus, E. Due to the need to consider discrete
lithological units, anisotropy, the sufficiency of the available data set for statistical analysis and
distribution best suited to model the data, few studies present useful statistically summaries of the
variability of these index tests. Grasso, et al. (1992) present the results of a study on a
homogeneous, slightly anisotropic, moderately strong calcareous mudstone with more than 50
measurements each of Vp, R and Is. The results of this study are summarized in Table 5-9 and serve
as an illustration of the magnitudes of variability that might be expected for the relatively
homogeneous material tested.
Table 5-8
Summary of Inherent Soil Variability and Measurement Variability for Index Tests
(after Phoon, et al, 1995)
Inherent Soil Measurement ASTM Precision
Soil Type
Variability
Variability
Estimate
Property(1)
Mean COV
Mean COV
COV
wn
Fine-grained
0.18
0.08
---wl
Fine-grained
0.18
0.07
0.05
wp
Fine-grained
0.16
0.1
0.17
PI
Fine-grained
0.29
0.24
---LI
Clay, Silt
0.74
------
Fine-grained
0.09
0.01
---d
Fine-grained
0.07
------Sand
0.19(2)
------Dr
(3)
------Sand
0.61
(1)
5-17
Table 5-9
Index Test and Inherent Variability of Calcareous Mudstone
(after Grasso, et al., 1992)
Index Test
Approximate COV
Compression Wave Velocity, Vp
0.09
Schmidt Hammer Rebound, R
0.19
Point Load Strength, Is
0.36
While resistance factors are presented in the following sections for use in foundation design in rock,
the high degree of potential variability between rock types suggests that it is prudent and may also
be economically attractive to modify those resistance factors for large projects to reflect the site
specific rock parameters.
5.4.3 Engineering Property Testing of Soil
Engineering property testing of soils provides a basis for directly defining the soil material
properties needed for design and analysis. This type of laboratory testing is performed using
undisturbed soil samples, and the duration and cost of testing is greater than for index tests. It is
important to carefully plan the subsurface exploration program to include an adequate number and
quality of samples for testing. Engineering property testing of soils includes:
y
In addition to these two major categories of laboratory testing, tests may be conducted to assess
swell and collapse potential, permeability, moisture-density relationships and dynamic soil
properties and behavior.
At the strength limit state, the shear strength of the soil is typically the most important parameter in
the evaluation of the resistance. Table 5-10 presents a summary of the inherent soil variability and
measurement variability for some of the various strength tests described previously. Note that no
information is currently available concerning the statistical reliability of consolidation test data. In
contrast to some of the relatively large COV values in this table, Nowak (1991) reports a bias factor
of 1.095 and a COV = 0.075 for the moment capacity of steel, non-compact sections, and a bias
factor of 1.12 and COV = 0.12 for lightly reinforced concrete T-beams.
5.4.4 Engineering Property Testing of Rock
Engineering property testing of rocks provides a partial basis for directly defining the rock material
5-18
properties needed for design and analysis. In addition to the intact rock properties, it is essential to
incorporate the effect of discontinuities and structure on the overall behavior of the rock mass. This
section focuses on laboratory testing of intact rock specimens. For weak, moisture-sensitive rock
types (e.g., clay shales), it is important that testing only be conducted using rock core which has
been carefully preserved to minimize loss of moisture.
Table 5-10
Summary of Inherent Soil Variability
and Measurement Variability for Strength Tests
(after Phoon, et al., 1995)
(1)
Property
Su (UC)
Su (UU)
Su (CIUC)
f 2)
tan f (TC)
tan f (DS)
tan f (2)
Soil Type
Fine-grained
Clay, Silt
Clay
Sand
Clay, Silt
Clay, Silt
Sand, Silt
Clay, Silt
Clay
Sand
Inherent Soil
Variability
Mean COV
0.33
0.22
0.32
0.09
0.21
0.20
---0.23
---0.09
Measurement
Variability
Mean COV
------0.19
---------0.08
---0.14
----
(1)
The engineering behavior and properties of intact rock can be highly variable, and the rock mass
perhaps even more so. Table 5-11 presents the variability of Young's modulus for a number of
different rock groups, with the COV for any given group ranging from a low of 0.14 to a high of
1.02 (in this table different rock types are from different sites, e.g., Pikes Peak Granite and Barre
Granite). For the example loading conditions and load variability described in Section 5.2, this
range of COV would correspond to resistance factors ranging from 0.82 to 0.05. This extreme
variability clearly illustrates the necessity for careful, site specific consideration of the intact rock
5-19
Figure 5-10 shows histograms of COV for both the compressive strength and tensile strength for a
number of rock types. Again, it is important to distinguish between the measured properties of the
intact rock as shown in this figure and the actual properties of the entire rock matrix including
discontinuites, anisotropy and differential weathering.
Figure 5-10
Histograms of Intact Rock Strength Properties
(after Kulhawy, et al., 1991)
5-20
No specification for substructure design can envision all possible geologic conditions or provide
sufficient guidance regarding their evaluation in the design process. The variability of soil and rock
materials and the reliability of field and laboratory test methods to estimate their engineering
behavior as described in the previous sections are representative of typical subsurface conditions.
Accordingly, it might be necessary to modify resistance factors used for LRFD when unusual or
highly variable soil and rock conditions are encountered, or when very uniform or well defined
soil and rock conditions exist. Modification of resistance factors would be analogous to modifying
the factor of safety in ASD when critical situations or geologic conditions are present (e.g.,
proximity to deformation-sensitive structures or heterogeneous foundation soils) and which require
special attention to design and construction details (e.g., presence of moisture-sensitive soils).
Special geologic conditions which can be encountered in the U.S. are summarized in Table 5-12.
5-21
Table 5-12
Problem Conditions Requiring Special Consideration
(AASHTO, 1991)
Problem
Soil
Description
Organic Soil; Highly Plastic Clay
Sensitive Clay
Micaceous Soil
Expansive Clay/Silt; Expansive Slag
Liquefiable Soil
Collapsible Soil
Pyritic Soil
Laminated Rock
Expansive Shale
Pyritic Shale
Soluble Rock
Rock
Cretaceous Shale
Weak Claystone (Red Beds)
Gneissic and Schistose Rock
Subsidence
Sinkholes/Solutioning
Condition
Comments
Low strength and high compressibility
Potentially large strength loss upon large
straining
Potentially highly compressible (saprolite)
Potentially large expansion upon wetting
Complete strength loss and high deformations
due to earthquake loading
Potentially large deformations upon wetting
(Caliche; Loess)
Potentially large expansion upon oxidation
Low strength when loaded parallel to bedding
Negative Friction/Expansion
Loading
Corrosive Environments
Permafrost/Frost
Capillary Water
5-22
The first phase of exploration and preliminary design included field vane shear tests which are
identified herein as Series 1. As shown in Table 5-13, the average Su for the Series 1 testing was
13.8 kPa, which led several designers to preliminarily conclude that an anchored sheetpile wall was
not a feasible solution to the problem. However, it was suspected that, a full head of water was not
always maintained in the augers during the Series 1 field vane testing and that the data was,
therefore, suspect.
Subsequently, a contractor and designer, believing that the anchored sheetpile wall was a viable
solution, performed the Series 2 field vane shear testing as well as the CIUC tests, the results of
which are also presented in Table 5-13. The Series 2 field vane shear testing was performed using
the same drilling rig and testing equipment as the Series 1 testing, but was performed using a full
head of water in the hollow stem augers for each test. Both the Series 1 and 2 field vane testing
were distributed relatively evenly over the thickness of the soft clay. The CIUC testing data was
subsequently performed on undisturbed samples primarily to assess the effective stress strength
parameters of the soft clay, but also as another characterization of the total stress strength parameters
for the clay. The CIUC testing was performed at effective confining pressures representing the
range of in-situ conditions.
Objective: Demonstrate statistical interpretation of testing as related to LRFD and development of
site specific resistance factors.
Solution: First complete the data tabulation by calculating the statistical parameters at the bottom of
the table. Recall that the standard deviation, , is calculated from Chapter 3 as:
= ( x i - x ) 2 / (N - 1)
(Eq. 3-8)
where:
xi = Individual data point
x = Average of all data points
N = Total number of data points
and the COV from Chapter 3 is:
5-23
COV = /x
(Eq. 3-9)
Table 5-13
Field and Laboratory Shear Strength Test Data for Example Problem
FVT
FVT
Series 2
CIUC Series
Series 1
N
Corrected
Su, (kPa)
Corrected
Su, (kPa)
Su, (kPa)
1
3.7
11.8
22.4
2
10.1
42.7
37.6
3
11.2
30.1
65.9
4
12.0
54.1
30.4
5
28.0
17.5
61.4
6
13.1
17.0
17.6
7
5.6
33.5
29.6
8
16.4
38.2
41.4
9
26.2
22.2
10
13.4
25.9
11
7.5
26.0
12
18.7
17.7
13
19.1
19.1
14
14.9
41.7
15
13.1
30.4
16
7.5
17.3
17
29.5
18
13.6
19
20.4
20
31.2
21
22.3
22
11.9
23
29.7
N
16
23
8
x
13.8
26.2
38.3
6.8
10.6
17.4
COV
0.49
0.40
0.45
R
1.86
1.00
0.67
Step 1: Select the nominal (ultimate) resistance strength value for use in design
1.
The Series 1 FVTs may be biased due to a lack of water head in the borehole
or other reasons. Therefore, this data is discarded.
2.
3.
The FVT has been shown to provide one of the best measurements of
undrained shear strength for use in global stability types of problems.
Step 2: Estimate the Bias Factor, r, For Each Data Series From Chapter 3 as:
R = Rm/Rn
(Eq. 3-10)
where:
Rm = Measured resistance
Rn = Predicted resistance
The values of R in Table 5-13 represent a "relative bias" calculated as the ratio of the average
results of the Series 1 and CIUC testing (predicted values) to the average from the Series 2 testing
(assumed measured value). It is assumed that the Series 2 testing provides the correct, unbiased
undrained shear strength. During the course of construction, this assumption was largely validated
when a contractor placed too much fill behind the wall in one area and nearly precipitated a failure
of the system. An inclinometer installed directly behind the wall at this location showed rapidly
increasing deformations before the load was removed and stability analyses including the excess fill
suggested that the Series 2 strength characterization was consistent with the imminent failure
observed with the inclinometer.
Step 3: Develop Resistance Factors For Each Data Set
As discussed in Section 5.2.2, it is theoretically conceivable to develop a site specific or material
specific resistance factor using Eq. 5-2 and Eq. 5-3 if the target reliability index, T, can be satisfied.
However, Eq. 5-2 requires that the COV and bias of the loads also be known. Chapter 7 presents an
approximate method to evaluate resistance factors which assumes that the standard deviations for the
load and resistance do not differ by more than a factor of 3. Using Eq. 7-6, the resistance factor for
lognormally distributed R and Q is approximated as:
= R exp (- T COVR)
(Eq. 7-6)
where:
5-25
R =
=
COVR =
T =
T
3.5
3.0
2.5
2.0
FVT Series 1
R = 1.86
COVR = 0.49
0.42
0.51
0.64
0.79
Resistance Factors,
FVT Series 2
CIUC Tests
R = 1.00
R = 0.67
COVR = 0.40
COVR = 0.45
0.30
0.17
0.35
0.21
0.42
0.25
0.50
0.31
Summary:
This example illustrates the development of site specific resistance factors using field and laboratory
test data. As discussed in Section 5.2.2, the development of resistance factors, , must consider
uncertainty related to:
y
y
y
5-26
indicated by a high bias factor) will tend to have a higher resistance factor
than less conservative test methods for similar coefficients of variation and
target reliability indexes
(2)
Any applicable test method may be used to predict resistance as long as the
statistical parameters (x
, , COVR and R) describing the inherent spatial
variability, measurement error and model error are accurately defined
5-27
5-28
CHAPTER 6
GEOTECHNICAL DESIGN PARAMETER SELECTION
6.1 Introduction
Chapter 3 introduced inherent soil variability, measurement uncertainty and transformation model
uncertainty as the three primary factors which contribute to the overall uncertainty in estimating soil
properties. The contribution of these factors in the property estimation process is illustrated in
Figure 6-1. Details regarding inherent soil variability and measurement uncertainty for commonly
used index and strength properties developed from various in-situ and laboratory measurement tests
were presented in Chapter 5. This chapter addresses the transformation model used for estimating
engineering properties of soils from in-situ and laboratory tests.
Figure 6-1
Uncertainty in Soil Property Estimates
(after Kulhawy, 1990)
In applying LRFD for substructure design, it is important to use resistance factors which have been
calibrated to account for each of these three components of uncertainty, and the uncertainty involved
in using the estimated soil properties in a given predictive model. For example, when using the method to determine skin friction capacity of a pile or drilled shaft in cohesive soils, it is important
that the undrained shear strength be characterized in a manner consistent with the resistance factor
calibration. The following sections of this chapter:
Discuss the suitability of various field and laboratory test methods for
estimating soil properties or foundation performance
Selection of a particular field or laboratory test method for estimating a geotechnical parameter for
design depends on the technical acceptability and relative cost and benefit of the method. To be
technically acceptable, the test must be directly applicable to the design requirements for the project.
6-1
For example, using standard penetration tests (SPTs) to estimate the shear strength of granular soils
is appropriate, whereas using SPTs to estimate the shear strength of cohesive soils is not appropriate.
LRFD concepts can be applied directly to evaluate the cost effectiveness of various exploration and
testing programs. Although this process can be demonstrated by many example problems presented
in later chapters, a problem in Chapter 9 provides a particularly good example of the approach. The
problem involves the design of a pile group in medium dense sand where a subsurface exploration
was conducted by SPT and cone penetration test (CPT) methods. In the problem, the SPT-based
design requires 21 piles whereas the CPT-based design requires 16 piles. Although significant
regional variability exists in construction costs, this 5-pile difference represents an approximate
savings of at least $7,500 (Means, 1996) using the CPT versus SPT approach. Regional variations in
construction costs and project specific variables may either increase or decrease this cost difference.
The remainder of this chapter focuses on the technical applicability of a given test type to estimate a
particular soil property and to address specific design needs.
6.2.1 Field Test Methods
As will be highlighted in later chapters, the results of in-situ testing can be used in one of two ways
for substructure design. One approach is to use in-situ measurements (e.g., N, qc or fs) directly to
estimate foundation performance such as bearing resistance or settlement. Tables 6-1 and 6-2
identify some of these relationships for cohesive and cohesionless soils, respectively. Note that a set
of calibrated resistance factors has been prepared (AASHTO, 1997a) for several of the methods
listed in these tables.
Alternatively, in-situ measurements can be used to estimate engineering properties (e.g., 'f, Su or E)
which can then be incorporated into a predictive model to evaluate foundation performance. Some
of the correlations which are available to estimate engineering properties are presented in Table 6-3
for cohesive soils and in Table 6-4 for cohesionless soils.
Based on the combined expectations of subsurface conditions and foundation types, it is possible to
use tables like those shown above in the preliminary selection of the types of in-situ tests which are
applicable to a given project.
6.2.1.1 Shear Strength
The effective stress shear strength of cohesionless soils, 'f, can be estimated from the results of
SPTs, CPTs and pressuremeter tests (PMTs). Selection of a particular test method should be based
on regional availability and expertise, cost and the reliability of each test in predicting the effective
stress friction angle, 'f. The reliability of these various in-situ test methods is described in Section
6.3.1.1.1.
6-2
Table 6-1
Correlation of In-Situ Test with Foundation Performance for Cohesive Soils
In-Situ Test Method
Foundation
Behavior
Type
Component
SPT
CPT
PMT
Side
--U
Resistance
Driven
Tip
Pile
--U
Resistance
Settlement
--U
Side
--U
Resistance
Drilled
Tip
Shaft
--U
Resistance
Settlement
--U
Bearing
-U(1)
Spread
U
Capacity
Footing
Settlement
--U
(1)
Resistance factors have been developed for at least one procedure.
Table 6-2
Correlation of In-Situ Test with Foundation Performance for Cohesionless Soils
Foundation
Behavior
In-Situ Test Method
Type
Component
SPT
CPT
PMT
Side
U(1)
U(1)
U
Resistance
Driven
Tip
Pile
U(1)
U(1)
U
Resistance
Settlement
U
U
U
Side
U(1)
-U
Resistance
Drilled
Tip
Shaft
U(1)
-U
Resistance
Settlement
--U
Bearing
U(1)
U(1)
Spread
U
Capacity
Footing
Settlement
--U
(1)
Resistance factors have been developed for at least one procedure.
6-3
Table 6-3
Correlations of In-Situ Tests with Engineering Behavior for Cohesive Soils
Field Test Correlation
Property
Soil
Property
Category
SPT CPT CPTU PMT
Preconsolidation Stress, p
U
U
U
U
Overconsolidation Ratio, OCR
--U
U
In-Situ Stress Coefficient of Horizontal Stress, Ko
U
U
U
U
Effective Stress Friction Angle, 'f
----Undrained Shear Strength, Su
Strength
U(1)
U
U
U
Poisson's Ratio, v
----Young's Modulus, E
-U
--Constrained Modulus, M
--U
U
Deformability Coefficient of Consolidation, cv
---U
Permeability Coefficient of Permeability, k
---U
(1)
VST
U
U
--U(1)
------
Table 6-4
Correlations of In-Situ Tests with Engineering Behavior for Cohesionless Soils
Property
Field Test Correlation
Soil Property
Category
SPT
CPT CPTU PMT
In-Situ Stress Coef. of Horizontal Stress, Ko
--U
U
(1)
(1)
Strength
U
-U
U
Effective Stress Friction Angle, =f
Poisson's Ratio, v
----Deformability Young's Modulus, E
-U
U
---Constrained Modulus, M
-U
Permeability Coefficient of Permeability, k
----Liquefaction
(1)
--
--
The total stress (i.e., undrained) shear strength of cohesive soils, Su, can be estimated most reliably
by the vane shear test (VST); however where a site-specific calibration is available based on
previous testing, PMTs, CPTs and SPTs can also be used. The reliability of these tests for
estimating Su is discussed in Section 6.3.1.1.2. In general, the SPT should not be used to estimate Su
of cohesive soils. Again, selection of a particular test should be based on local availability and
experience, cost and the reliability of the test method.
6.2.1.2 Compressibility
The drained modulus of cohesionless soils can be estimated using SPTs, CPTs, CPTs with pore
pressure measurements (CPTUs) or PMTs. The undrained modulus for cohesive soils can be
estimated using the CPTU and SPT. The CPTU has a unique capability in that it is the only in-situ
test which can provide information on pore pressure dissipation for estimating the rate of
consolidation of cohesive soils. This engineering property can be developed when cone
6-4
advancement is stopped for a period of time to permit excess pore pressures around the cone tip to
dissipate.
6.2.2 Laboratory Tests
6.2.2.1 Shear Strength
When selecting a laboratory strength test for substructure design, the type and rate of structure
loading and the drainage characteristics of the foundation soils must be defined as they dictate the
type of test to be used and whether total and/or effective stress strength parameters will control
design. For example, a consolidated-undrained triaxial compression test with pore pressure
measurements will provide both the total and effective stress properties of a cohesive soil whereas a
drained test will provide only effective stress parameters. Also, the boundary conditions which
characterize the test and those expected to represent actual field conditions must also be recognized.
Figure 6-2 shows some typical failure modes for common engineering problems and the
corresponding laboratory tests which most closely model the shearing modes in these problems.
While it is important to consider these shearing modes in both total and effective stress conditions,
the total stress strength parameter selection is especially sensitive to the type and rate of loading.
6.2.2.2 Compressibility
Soil compressibility depends on the rate of load application and the drainage characteristics of the
soil. Due to the high permeability of cohesionless soils, deformation is nearly immediate as load is
applied, such as the rate of loading effects are not significant (except for dynamic loads) and the
drained soil modulus controls compressibility. With the possible exception of very loose soils, the
time-dependent deformation of cohesionless soils is insignificant. For cohesive soils however, soil
compressibility can occur during both undrained and drained loading, with immediate settlements
occurring during undrained loading, and time-dependant settlements (e.g., consolidation and
secondary compression) occurring during drained loading. The compressibility characteristics of
soils can be determined in the laboratory using oedometer or consolidation tests in which the sample
is laterally confined within a rigid ring, or by unconfined or triaxial compression tests.
6.2.2.3 Index Tests
Correlations between engineering properties and index tests (e.g., particle-size analyses, unit weight,
water content and Atterberg limits) provide a relatively inexpensive means for making broad initial
estimates of soil properties during the preliminary phases of a project, and serve as indicators of
general soil behavior. Some possible correlations of test data with soil properties are identified in
Table 6-5. Because most laboratory test programs include water content, particle-size analyses and
Atterberg limits determinations, the results of these tests are generally available for use as a guide in
estimating material properties such as shear strength, compressibility and permeability.
6-5
Figure 6-2
Comparison of Field and Laboratory Shearing Modes
(after Kulhawy and Mayne, 1990)
Table 6-5
Soil Property Correlations with Index Test Data
Index Test
Possible Correlation
SPT and CPT
Shear strength, relative density, and compressibility
Water Content
Shear strength and compression index of clay
Particle-Size Distribution Permeability, strength and drainability of cohesionless soils
Atterberg Limits
Compressibility, shrink-swell, drained strength of cohesive soils
Void Ratio, Density
Compressibility and shear strength
Relative Density
Strength and compressibility of cohesionless soil
6.2.3 Parameter Evaluation
Following the completion of field and laboratory testing programs, the test results should be
evaluated for reasonableness and variability in selecting material parameters for use in design.
When possible, test results should be compared with data from prior testing in the vicinity of the
6-6
planned construction site, and performance of adjacent structure foundations should be considered
using information from foundation submissions and maintenance records. When highway
construction occurs in areas not previously developed or when records on past construction are not
available or are incomplete, selection of soil and rock properties for design can be aided using
project test results and reported correlations between index parameters and soil and rock properties
for strength, compressibility and permeability (e.g., Barker, et al., 1991a; Carter and Bentley, 1991;
Kulhawy and Mayne, 1990; and NAVFAC, 1986).
6.3 Reliability of Tests in Estimating Design Parameters
As discussed in Section 6.1 and illustrated in Figure 6-1, a transformation model must be used to
manipulate the test results to arrive at either a direct prediction of foundation performance or an
engineering property estimate. This section focuses on the uncertainty of a variety of transformation
models which have been developed based on the results of field and laboratory tests. The
uncertainty associated with the measurement process and inherent soil variability was discussed in
Chapter 5.
6.3.1 Field Test Methods
Several field test methods were identified in Section 6.2.1 for estimating the drained strength of
cohesionless and cohesive soils, and the undrained strength of cohesive soils. These tests include:
SPT
CPT using either a mechanical (MCPT) or electrical (ECPT) device
Field VST
PMT
The reliability of these tests for estimating the strength and compressibility of soils is discussed in
Sections 6.3.1.1 and 6.3.1.2, respectively.
6.3.1.1 Shear Strength
6.3.1.1.1 Effective Stress Strength
Standard Penetration Test
The SPT has been and remains the mainstay of in-situ testing in the United States, and has perhaps
been used most often to estimate the effective stress friction angle of cohesionless soils. Numerous
relationships exist between SPT blow count, N, and effective stress angle of friction, 'f (e.g.,
Meyerhof, 1956; Peck, Hanson, Thornburn, 1974); however, there is no direct, statistically-based
correlation between these parameters.
6-7
Figure 6-3
Correlation of CPT Tip Resistance, SPT N-Value and Mean Particle Size
(after Kulhawy and Mayne, 1990)
Figure 6-4
Correlation of CPT Tip Resistance, SPT N-Value and Fines Content
(after Kulhawy and Mayne, 1990)
6-8
Figure 6-5 shows a statistical correlation of the normalized CPT tip resistance, (qc/pa)/('vo/pa) with
the triaxial compression effective stress friction angle, 'tc. The CPT tip resistance, qc, is normalized
with respect to atmospheric pressure, pa and the in-situ vertical effective stress, 'vo. The correlation
is based on 633 tests. The data set from which Figure 6-5 is derived consists of 24 sets of calibration
chamber data on reconstituted, recently-deposited sands with a percentage of fines from 0 to 6
percent, D50 from 0.16 to 1.0 mm, D10 from 0.10 to 0.70 mm and a uniformity coefficient from 1.10
to 2.60. For the range of friction angles reported in Figure 6-5, 0.05 < COV < 0.10 with a much
tighter spread of about 0.07 < COV < 0.08 for 35o < 'tc < 40o.
6.3.1.1.2 Total Stress Strength
Standard Penetration Test
Although correlations have been developed between N and the undrained shear strength, Su, of
cohesive soils, the correlations do not account for an N-value standardized to a given energy value,
any particular reference strength for Su, or the sensitivity of the clay. Recall that different drilling
rigs, hammer arrangements and general test practices can impart significantly different amounts of
energy to the split spoon and result in very different N-values for the same soil. Therefore, unless
calibrated for a given geologic setting, energy level and reference strength, the SPT does not provide
a useful quantitative correlation with undrained shear strength.
Figure 6-5
Correlation of Triaxial Compression Effective Stress Friction Angle
with Cone Penetration Tip Resistance
(after Kulhawy and Mayne, 1990)
Cone Penetration Test
The cone tip resistance has been correlated with undrained shear strength, Su, using the relationship:
6-9
qc = N k Su + vo
(Eq. 6-1)
where:
qc = Cone tip resistance (kPa)
vo = Total overburden stress (kPa)
Nk = Cone bearing factor (dim)
Theoretical predictions of Nk range from about 7 to 18, and empirical correlation for specific sites
range from about 5 to 75. The reliability of Eq. 6-1 can be improved by substituting a corrected tip
resistance accounting for pore pressure effects when using a piezocone (CPTU) by:
q t = qc + (1 - a) u bt
(Eq. 6-2)
where:
qt =
a=
6-10
Figure 6-6
Field Vane Shear Correction Factor
(after Azzouz, et al., 1983)
6.3.1.2 Compressibility
Pressuremeter Test
The PMT provides a direct measure of a modulus value which is typically assumed to be equal to an
undrained modulus, Eu, in cohesive soils and Young's modulus, E, in cohesionless soils.
Standard Penetration Test
Prior to the introduction of probability based design concepts, numerous non-statistical correlations
between the SPT N-value and the drained modulus for cohesionless soils were proposed (e.g.
D'Appolonia, et al., 1970, Mitchell and Gardner, 1975). Due to the lack of a quantitative estimate of
the reliability associated with these early methods, only some more recent correlations are described
herein. Figure 6-7 shows a correlation between the uncorrected SPT N value and a normalized
pressuremeter modulus, EPMT/pa.
The wide spread in data about the trend line in Figure 6-7 illustrates the high degree of uncertainty
associated with trying to predict the deformation properties of a cohesionless soil based on SPTs.
The COV for this model ranges from a low of about 1.5 to greater than 2.0. A similar relationship is
shown in Figure 6-8 for cohesive soils, for which the COV ranges from approximately 1.5 to 2.5.
Considering the uncertainty associated with the relationships in Figures 6-6 and 6-7, SPTs should
only be used for preliminary estimation of the modulus of soils.
6-11
Figure 6-7
Pressuremeter Modulus versus N Value for Cohesionless Soil
(after Kulhawy and Mayne, 1990)
Figure 6-8
Pressuremeter Modulus versus N Value for Cohesive Soils
(after Kulhawy and Mayne, 1990)
6-12
Figure 6-9
Correlation of CPT Tip Resistance with Constrained Modulus
and Relative Density for NC Sands
(after Kulhawy and Mayne, 1990)
6-13
Figure 6-10
Correlation of CPT Tip Resistance with Constrained Modulus
and Relative Density for OC Sands
(after Kulhawy and Mayne, 1990)
6.3.2 Laboratory Tests
The primary laboratory strength and compressibility tests consist of the triaxial compression test and
the consolidation test. Results from each of these tests exhibit some measurement error and error
due to inherent variability of the sample tested. As each test provides a direct measure of a
particular engineering property, a transformation model is generally not required to utilize the test
results. However, test results may be correlated to other parameters where the potential failure
modes in the field do not match the laboratory boundary conditions in the triaxial test (e.g. plane
strain vs. triaxial failure modes).
Laboratory index tests can be used to provide cost-effective preliminary estimates of some common
engineering properties. For cohesive soils, Atterberg limits and their relationship to the in-situ water
content are most commonly used for such correlations. Figure 6-11 shows one correlation between
the Liquidity Index, LI, and the ratio of undrained shear strength to vertical effective stress level in
triaxial compression, Su/'vo.
6-14
Figure 6-11
Correlation Between Normalized Undrained Shear Strength and
Liquidity Index for NC Clays
(after Kulhawy and Mayne, 1990)
Correlations of index tests with some deformation properties are also available, as shown in Figure
6-12 which depicts relationships between the Compression Index, Cc, and the Plasticity Index, PI,
and between the Recompression Index, Cur, and the PI, for cohesive soils.
Figure 6-12
Compression and Unload-Reload Indices versus Plasticity Index
(after Kulhawy and Mayne, 1990)
6-15
6-16
Table 6-6
Approximate Guidelines for Design Property Variability
(after Phoon, et al., 1995)
Design
Point COV
Spatial Avg.
Test
Soil Type
Property
(%)
COV (%)
Su (UC)
Direct (lab)
Clay
20-55
10-40
Su (UU)
Direct (lab)
Clay
10-35
7-25
Su (CIUC)
Direct (lab)
Clay
20-45
10-30
Su (field)
VST
Clay
15-50
15-50
Su (UU)
qT
Clay
30-40
30-35
Su (CIUC)
qT
Clay
35-50
35-40
Su (UU)
N
Clay
40-60
40-55
Su (field)
PI
Clay
30-55
-Direct (lab)
Clay, Sand
7-20
6-20
f'
Sand
10-15
10
qT
f' (TC)
EPMT
Direct (PMT)
Sand
20-70
15-70
EPMT
N
Clay
85-95
85-95
ED
N
Silt
40-60
35-55
Su = Undrained shear strength; UC = Unconfined compression test; UU = Unconsolidated-undrained
triaxial compression test; CIUC = Consolidated isotropic undrained triaxial compression test; Su (field)
= Corrected Su from vane shear test; f' = Effective stress friction angle; TC = Triaxial compression
test; qT = Corrected cone tip resistance; EPMT = Pressuremeter modulus; ED = Dilatometer modulus
averaged over a depth interval of 5 m
Figure 6-13
Normalized Undrained Shear Strength
Versus Undrained Strength Ratio from CIUC and UU Tests
(Chen and Kulhawy, 1993)
6-17
It is important to note that the resistance factors presented in the AASHTO LRFD Specifications
are developed using data bases which include a variety of property characterization methods.
Further development and/or refinement of resistance factors should explicitly consider both the
suitability and variability of the property characterization. Specifically, resistance factor
development should consider the overall bias of the resistance and the coefficient of variation of the
bias. For example, if this development is done in a rational manner it would be possible to use a test
which overpredicts the available soil strength because this bias would be accounted for in the
resistance factor determination. However, in general it is preferable to utilize a testing procedure
which provides the best measurement of the property of interest without a significant bias.
6.5 Example Problem:
Characterization Planning
Problem: Design a single-phase, cost-effective field exploration and laboratory testing program
using LRFD concepts to provide information for design of the pier foundation, wall and
embankment systems which might be considered for the site shown in Figure 6-14. Table 6-7
provides some of the regional costs associated with typical laboratory and field tests, and provides
spaces for developing the scope and cost of the exploration program. The selected borings may be
located directly on Figure 6-14 and Table 6-7.
Objectives: Demonstrate general LRFD concepts as they pertain to:
Background: Figure 6-14 shows a site plan for one side of a river crossing. The proposed grading
and construction plans consist of a single pier, an anchored sheet pile wall for shoreline protection
and the embankment approach to the bridge. A preliminary review of the site topography, geologic
setting and local knowledge suggest that the soil profile at the pier site consists of a 12-m thick
deposit of medium stiff clay underlain by a deep deposit of medium dense to dense sand. The sand
deposit is known to become very dense at depths exceeding about 25 m. At other sites along this
river valley, past exploration programs have disclosed the existence of sand seams within the upper
clay layer. The pre-construction elevation along the alignment of the anchored wall anchored wall is
El. 417 m. Therefore, 4 m of fill will be needed behind the wall.
From a preliminary design of the bridge, the Strength I Limit State factored loads at the base of the
pier are estimated to be 17 500 kN. The combined effects of this relatively large load and the
medium stiff clay layer underlain by a dense sand suggest a pile foundation system as a likely
support for the pier. Because a pile foundation would be expected to achieve most of its capacity
through tip resistance, information concerning the engineering behavior of the sand deposit is of
primary interest for design of the pier foundations. For the estimated pier foundation load, a simple
preliminary design estimate might suggest the need for about 20 steel piles at an installed unit cost of
approximately $1,250 (Means, 1996). The wall construction and the approach embankment
6-18
construction raise issues of slope stability and deformation, which will be dictated by the strength of
the upper clay layer as well as the potential magnitude and rate of consolidation settlement beneath
the approach embankment . The exploration program should provide sufficient data to permit
consideration of viable foundation options for the pier, evaluation of the stability and settlement of
the embankment, and design of the wall.
Key geotechnical issues which must be addressed for the site include definition of the water table for
all structures and the following specific issues for various aspects of the proposed construction:
Approach Embankment:
Total and effective stress strength of the top 12 m of medium stiff clay
Magnitude of consolidation settlement of clay
Rate of consolidation of clay and possible presence and impact of sand seams
within clay
West Pier:
6-19
Figure 6-14
Example Problem
Site Exploration Plan for Western Approach
6-20
Table 6-7
Example Problem Work Sheet
Unit Price
($/Unit)
Number
of Tests
Total Price
($)
Laboratory Tests
Water Content
$5/ea
12
$60
Atterberg Limits
$75/ea
$450
Unit Weight
Grain Size
Distribution
3 Pt. CIUC w/ Pore
Pressure
Measurement
Consolidation
$35/ea
$70
$60/ea
$360
$1000/ea
$1,000
$350/ea
2
Subtotal
$700
$2,640
$55/m
25 m
$1,375
$50/ea
$200
$400
$400
Subtotal
$1,975
CPTU w/ Dissipation
$40/m
133
$5,320
$750
FVT
FVT Mob. / Demob
$30/ea
$250
1
Subtotal
12
1
Subtotal
Total
$750
$6,070
$360
$250
$610
$11,295
6-21
Reason
Consolidation of clay
Perform field vane shear testing at 1 m intervals in the clay layer. At the
location of each FVT, collect a grab sample with the split-spoon for
laboratory index testing
Collect 4 undisturbed samples (i.e. shelby tubes) in the clay between FVT
locations for strength and consolidation testing
CPTU Soundings at C-1 and C-2 (West Pier), C-3 (Embankment Toe) and C-5, C-6 and C-7 (Sheet
Pile Wall)
Perform CPTUs to the top of the sand deposit to evaluate Su of clay and to
identify presence of sand seams in clay.
parameters will be evaluated with the 3-point, CIUC with pore pressure measurements.
The consolidation test will provide data allowing an estimate of the magnitude and rate of
consolidation settlement for the approach embankment. In addition, the CPTUs can provide an
estimate of the coefficient of consolidation by measuring pore pressure dissipation. More
importantly, the CPTU soundings will provide a nearly continuous profile which will assist in
identifying the presence of sand seams in the clay. Because the rate of consolidation is proportional
to the square of the drainage path length, the presence of continuous sand seams within this clay
layer might conceivably decrease the predicted time for consolidation by an order of magnitude or
more.
West Pier:
As suggested previously, a pile foundation system might be a good foundation solution for the
proposed West Pier due to its heavy loading and the combined presence of a medium stiff clay in the
top 12 meters and the medium dense to dense sand below the clay. End bearing steel pipe piles
should provide a cost effective foundation scheme for these conditions. For the expected subsurface
conditions, it is likely that the LRFD design of a pile foundation system may be economically
performed using in-situ data comprised of SPT or CPT testing, or may be based on the results of a
dynamic load test. While the design could also be based on the results of a static load test, for this
problem it is unlikely that this will be economical and is therefore not considered herein. The
resistance factors associated with each of these methods from the AASHTO LRFD Specification
(1996) (A10.5.4) are:
AASHTO LRFD
Specification Resistance
Factor
0.45
0.55
0.70
Whereas the pile driving analysis (PDA) method has the highest -factor for this situation, the cost
of performing a single test of this type during the design phase may be $10,000 to $15,000. A PDA
will provide additional information about hammer selection and drivability. In contrast, the CPT
method provides a resistance -factor about 20 percent better than the SPT method and costs less
than the SPT. The CPT therefore appears to be the most cost effective exploration tool from the
standpoint of minimizing the combined exploration and foundation construction costs for this single
pier. However, depending on the number of additional piers in similar soil conditions which might
be needed to span the river in this problem, it may well be cost effective to mobilize a pile driving rig
for pile driving analysis in the design phase of the problem.
6-23
The FVT was chosen as a reliable means to develop base total stress
strength parameters, but the CPT was selected as a cost effective,
reliable tool to gather data on the variation of total stress strength
parameters across the site.
The resistance factors and relative costs of SPT- CPT- and PDAbased pile design were evaluated and the CPT was judged to provide
the best combination of reliability and economy for this example.
Note: As discussed in Chapter 9, the 1997 Interims of the AASHTO LRFD Specifications
(1997a) include an initial attempt to incorporate the specified method and extent of field pile
construction control and capacity verification into the determination of the resistance factors
for axially loaded piles (Table A10.5.5-2). Additional consideration and future revision of
these requirements are anticipated. The resistance factors cited in this example (from
AASHTO, 1996) were obtained from the original reliability-based calibration (Barker, et al.,
1991a) and do not consider construction control or field capacity verification. The resistance
factors in this example, therefore, reflect only the relative reliability of the SPT and CPT based
static capacity prediction methods. The resistance factor for PDA is based on direct
calibration to ASD.
6-24
CHAPTER 7
CALIBRATION AS PART OF THE DESIGN PROCESS
7.1 Introduction
Development of the AASHTO LRFD Specifications (1997a) involved evaluating resistance factors
for the various design methods included in the Specifications. Even though the Specifications
contain values of resistance factors for a considerable number of design methods and conditions,
designers may need to develop resistance factors for methods or conditions that are not included in
the AASHTO LRFD Specifications (1997a). For example:
y
A designer may want to use a method that is not covered in the LRFD
Specifications, because its previous use under ASD has shown that it is an
effective method, with a successful record of use in the soil conditions in the
designer's area.
A designer's experience with local soils may indicate that the resistance
factors in the LRFD Specifications are unsuitable for some of the soils in the
designer's area.
In either of these cases, it is appropriate for the designer to evaluate the resistance factors needed to
apply the LRFD design procedures. Fortunately, this can be done simply, and without great
expenditure of time, using common sense, engineering judgment, and a few simple formulas.
7.2 Reliability Theory or Fitting with ASD?
Two approaches for calibration are available:
(1)
(2)
The information regarding the reliability of ASD methods needed for the approach in (1) has been
determined through research studies by Barker, et al. (1991a) and others.
Neither of these approaches need be cumbersome. The best procedure is to use both because they
serve to check each other. Small differences in values determined by reliability and by fitting with
ASD are to be expected. In such cases, the designer is justified in concluding that either value, or
one between them, is suitable.
7-1
Large differences between values of resistance factors determined by these two approaches may
indicate (a) a blunder in the computations, (b) use of inappropriate statistics on strength, or (c) a
mismatch between the probability of failure used in the reliability method and the factor of safety
used in the ASD fitting computations.
The statistics for strength that are needed for reliability calculations are the average value and the
standard deviation.
When the standard deviation for strength is not known and cannot be estimated with confidence, the
only alternative is fitting with ASD. In most cases, however, the designer will be able to use simple
rules-of-thumb to estimate the statistics required for the reliability method.
7.3 Evaluating Resistance Factors Using Reliability Method
Although the reliability method has an extensive theoretical background, in application the method
can be accomplished by evaluating a single equation. The equation is complicated, but the
computation can be automated (for example, in a spreadsheet), and an engineer need only input the
values of the parameters involved.
Most methods of estimating foundation capacity involve calculating the products and ratios of a
number of variables. The results of these types of calculations are fitted more closely by lognormal
distributions than by normal distributions (Barker et al., 1991b). In this chapter, only methods of
calculating resistance factor that are based on lognormal distributions will be discussed, because this
is the more practical choice.
The expression for the resistance factor, , based on lognormal distributions of load and resistance is
given by Eq. 3-35 in Chapter 3, which for convenience is repeated as Eq. 7-1.
2
2
R ( D QD / QL + L) 1 + COVQD + COVQL
1 + COV2R
QD QD / QL + QL
2
2
exp T ln[(1 + COVQD
+ COVQL
)(1 + COV2R )]
(Eq. 7-1)
QD =
Bias factor for dead load; QD varies with the material (e.g., steel,
concrete, asphalt as shown in Table 7-2 and equals the product of the bias
factors for the materials making up the dead load (dim)
QL =
Bias factor for live load; QL = 1.15 from field studies used to develop
live load model in AASHTO LRFD Specifications (dim)
7-2
COVR =
COVQD = Coefficient of variation for dead load; COVQD varies with material type
(See Table 7-2) and equals the value of the square root of the sum of the
squares of the COV=s for the materials making up the dead load (dim);
COVQL = Coefficient of variation for live load; COVQL = 0.18 field studies used to
develop live load mode in AASHTO LRFD Specifications (dim);
D =
Load factor for materials making up the dead load; D = 1.25 in the
AASHTO LRFD Specifications; (dim)
L =
QD/QL =
Ratio of dead load divided by live load; QD/QL varies depending on the
structure, and can be computed when the loads are known or estimated
based on the span length of the bridge (dim)
T =
The bias factor, R, is a measure of the inherent conservatism of the method being calibrated. A
value of R larger than 1.00 indicates that the method tends to result in conservatively low values of
calculated resistance; a value smaller than 1.00 indicates that the method is inherently
unconservative (Barker et al. 1991a).
The reliability index, T, is related to the probability of failure that underlies the design. Probability
of failure means the probability that the actual resistance may be smaller than the actual load.
Studies by Barker et al. (1991a) have shown that ASD methods, on which foundation designs have
traditionally been based, typically result in probabilities of failure in the range from 0.01 to 0.0001.
These values of probability of failure correspond to values of T ranging from 2.5 to 3.5. The same
studies have shown that probabilities of failure of piles designed by ASD methods are larger,
ranging from 0.1 to 0.01. These higher probabilities of failure of piles, which correspond to values
of T ranging from 2.0 to 2.5, are acceptable because piles are used in groups. Failure of one pile in
a group would not necessarily cause the entire group to fail. The probability of failure of the group
is therefore smaller than the probability of failure of a single pile within the group.
Simple methods of estimating R, QD, COVR, COVQD, and QD/QL are discussed in the following
sections, together with examples of values of T for various types of structures.
7.3.1 Estimating Resistance Bias R and Coefficient of Variation COVR
As discussed in Chapter 3, the value of the bias factor on resistance (R) represents the net effect of
various sources of error, and COVR represents the usual variability of the results. Values of R and
COVR for various types of tests on soil are presented in Table 7-1.
7-3
Table 7-1
Values of R and COVR
Test
SPT (Standard Penetration Test)
CPT (Cone Penetration Test)
Angle of internal friction ( or )
Cohesion (c or c)
Wall friction angle ()
Earth pressure coefficient (k)
R
1.3
1.0
1.0
1.0
1.0
1.0
COVR
0.60 to 0.80
0.40
0.10
0.40
0.20
0.15
For measures of soil or rock resistance or strength not listed, the designer can use judgment to
estimate a value of R, and a simple rule-of-thumb to estimate a value of COVR (Demsky, 1996).
Estimates of values of R can be based on the observation as illustrated by the values in Table 7-1.
The table shows that R is likely to be 1.00 for methods of measurement that are based on principles
of mechanics, while values of R may differ significantly from 1.00 for methods that rely on
empirical rules to relate test results to strength values.
It can be seen in Table 7-1 that the CPT, the triaxial test, and the direct shear test, which are
interpreted using principles of mechanics, all have values of R = 1.00. For any other method that is
based on the principles of mechanics and rational interpretation of results, the value of R can
reasonably be estimated to be equal to 1.00.
The SPT, which can only be interpreted using empirical rules, has a value of R = 1.3. The value of
R for other methods that are interpreted using empirical rules can only be evaluated accurately using
data from field load tests, which provide a body of data for comparing calculated and measured load
capacities.
The value of COVR can be estimated using a rule-of-thumb, known as the "six sigma" rule. Use
of the Asix sigma@ rule involves three simple steps using strength values that can be estimated by any
experienced engineer:
(1)
Estimate the most likely value of the property (Vest), the lowest conceivable
value (Vmin), and the largest conceivable value (Vmax).
(2)
Use the "six sigma rule" (the following empirical formula) to estimate the
value of the standard deviation ():
=
(3)
Vmax - Vmin
6
(Eq. 7-2)
7-4
COVR =
(Eq. 7-3)
Vest
COVQD
0.08
0.10
0.25
When the dead load has more than one component, the bias factor and COV are determined from
equations presented in Chapter 3, that is, the combined bias factor is the product of the individual
bias factors,
= 1 2 K N
(Eq. 7-4)
7-5
and the combined COV is the square root of the sum of the squares of the individual COVs:
COV = COV12 + COV 22 + K + COV 2N
(Eq. 7-5)
Minimum and maximum values of load factors are given in Table 7-3. The value to be used in a
given calculation is the one that leads to the more conservative result. For example, P = 1.80 would
be used in calculating downdrag loads in compression piles, and P = 0.45 would be used in
calculating uplift capacity of friction piles. Similarly, P = 1.30 would be used to calculate earth
loads on buried rigid structures where bending moment increases with increasing earth pressure.
However, if the bending moment of interest decreases with increasing earth pressure, P = 0.45
would be used.
7.3.7 Values of QD/QL
Hansell and Viest (1971) showed that the dead load-to-live load ratio (QD/QL) varies with span
length, L. Their study used the AASHTO ASD Specifications (i.e., Article 3.8.2.1), and the
corresponding Impact Factor, I, which varies with L as shown in Table 7-4. The Impact Factor is an
allowance for the increased live load effect that results from vehicle impact loads on the bridge deck.
Using the Impact Factors for ASD shown in Table 7-4, Hansell and Viest determined that the ratio
QD/QL could be expressed by the empirical formula QD/QL = (1 + I)(0.0132L), where L is expressed
in feet.
7-6
Table 7-3
Load Factors for Permanent Loads, P
(AASHTO, 1994)
Load Factor
Type of Load
DC: Component and Attachments
DD: Downdrag
DW: Wearing Surfaces and Utilities
EH: Horizontal Earth Pressure
y Active
y At-Rest
EV: Vertical Earth Pressure
y Overall Stability
y Retaining Structure
y Rigid Buried Structure
y Rigid Frames
y Flexible Buried Structures
y Flexible Metal Box Culverts
ES: Earth Surcharge
Maximum
1.25
1.80
1.50
Minimum
0.90
0.45
0.65
1.50
1.35
0.90
0.90
1.35
1.35
1.30
1.35
1.95
1.50
1.50
N/A
1.00
0.90
0.90
0.90
0.90
0.75
In the AASHTO LRFD Specifications, the factor corresponding to I is called the Dynamic Load
Allowance Factor, IM. The value of IM is independent of span length. As shown in Table 7-4, IM
= 0.33 for all span lengths. The values of QD/QL shown in rightmost column of Table 7-4 were
calculated using Hansell and Viest=s empirical formula, with IM replacing I. The values in this
column provide reasonable estimates of QD/QL as a function of span, consistent with the provisions
of the AASHTO LRFD Specifications (1997a).
Table 7-4
Values of QD/QL
ASD Values
Span Length, L
(m)
Impact, I
QD/QL
9
18
27
36
45
60
75
0.30
0.27
0.23
0.20
0.18
0.15
0.13
0.50
0.99
1.44
1.90
2.30
2.99
3.67
7-7
LRFD Values
Dynamic Load
QD/QL
Allowance, IM
0.33
0.52
0.33
1.04
0.33
1.56
0.33
2.07
0.33
2.59
0.33
3.46
0.33
4.32
7.3.8 Values of T
Values of target reliability index, T, appropriate for geotechnical design of various types of
foundation elements have been determined by Barker, et al. (1991a) through analysis of
conventional design practices. These values are listed in Table 7-5. The values of T for piles are
smaller than those used for other types of foundations because piles are used in groups, and there is
considerable redundancy in the capacity of pile groups that contain several piles. The probability of
failure of a group of piles - that is, the probability that all of the piles in the group will fail - is
smaller than the probability of failure of one pile in the group. As a result, acceptable values of T
for capacity of a single pile are smaller than for other types of foundations, because it is presumed
here that a single pile would not be used to support structural loads. If a single pile was to be used to
support a structure load, the target value of should be higher, in the range from 2.5 to 3.5, such as
for drilled shafts and spread footings which are often used singly.
Table 7-5
Values of Target Reliability Index, T
(Barker, et al., 1991a)
Foundation Type
Spread Footings
Drilled Shafts
Driven Piles (group)
T
3.0 to 3.5
2.5 to 3.0
2.0 to 2.5
Resistance factors for the Nordlund method were not included in the AASHTO LRFD Specifications
for design of bridge foundations. However, Eq. 7-1 can be used to calculate resistance factors for
this method.
Nordlund (1963) presents data comparing measured pile capacities with values calculated using his
proposed method. These data can be used to establish values of the bias factor, RN = average value
of measured capacity divided by computed capacity and the coefficient of variation, COVRN for the
method. Using the data in Table 2 of Nordlund's paper, RN = 1.04, and COVRN = 0.17. These
measures of the uncertainty in Nordlund's method represent a simplification of a more complex
relationship. The complexity occurs because Nordlund's prediction equation for capacity has two
independent terms: one for point bearing and one for skin friction. Each of these terms will have
their own statistics which could be combined according to Eq. 7-4 and Eq. 7-5 to obtain better values
of the bias and COV if the necessary data are available. However, for the purpose of this example,
the combined measures of uncertainty in the prediction equation will be used because data showing
the separate contributions of end bearing and skin friction are not available.
7-8
In applying Nordlund's method a value must be selected for the angle of internal friction of the
supporting soil. Statistics for the uncertainty of determining the angle of internal friction from SPT
data have been tabulated by Barker, et al. (1991a) as RA = 1.0 and COVRA = 0.25.
Combining the resistance statistics using Eq. 7-4 and Eq. 7-5, RA and COVRA can be determined as
follows:
R = RN + RA = 1.04 1.00 = 1.04
and
2
2
COVR = COV2RN + COV2RA = (0.17 ) + (0.25 ) = 0.30
Other values needed to evaluate the resistance factor for Nordlund's method are:
QD =
QL =
COVQD =
COVQL =
D =
L =
As shown by Table 7-4, the value of QD/QL varies with span length. Because QD/QL appears on the
right side of Eq. 7-1, different values of resistance factor, , will be applicable to different span
lengths. Two approaches are possible:
(1)
Determine values of for a range of span lengths and select a single value to
be used for any span length based on the results. This is the approach used in
developing resistance factors for the AASHTO LRFD Specifications for
bridge foundations.
(2)
Use a specified span length and corresponding QD/QL ratio for a specific
project in Eq. 7-1, to determine a value of for use on a specific project.
For purposes of illustration in this example, values of will be determined for values of QD/QL = 1.0
and QD/QL = 3.0, corresponding to span lengths in the range from 15 m to 50 m, as shown in Table
7-4.
It is also necessary to select a value of target reliability index to calculate a value of using Eq. 7-1.
In keeping with conventional practice for driven piles, values of T = 2.0 and T = 2.5 will be used
in this example. These values of T correspond to values of probability of failure of a single pile
ranging approximately from 0.1 to 0.01.
7-9
The value of for the Nordlund method can be calculated by substituting values of these parameters
into Eq. 7-1.
2
2
R ( D QD / QL + L) 1 + COVQD + COVQL
1 + COV 2R
QD QD / QL + QL
2
2
exp T ln[(1 + COVQD
+ COVQL
)(1 + COV 2R )]
(Eq. 7-1)
Substituting in Eq. 7-1 for QD/QL = 1.0, T = 2.0, and the values of QD, QL, COVQL, D, and L
listed above, the value of is:
= 0.66
Using Eq. 7-1, values of for the Nordlund method were calculated for QD/QL = 1.0 and 3.0, and T
= 2.0 and 2.5. The results are shown in Table 7-6. Based on these results and the desired level of
conservatism, values of ranging from about 0.50 to 0.65 would be appropriate. The average value
is about 0.58.
Table 7-6
Values of for Nordlund's Method Calculated Using Reliability Theory
Value of QD/QL
T = 2.0
T = 2.5
1.0
0.66
0.55
3.0
0.61
0.51
As shown in the preceding sections, the value of depends on the variability of loads as well as the
variability of strength. However, it is possible to derive approximate expressions for that are
independent of the variability of the loads, as described in Appendix 7A. The required assumption is
that the values of the standard deviations for load and resistance do not differ by more than a factor
of 3, which is usually the case. The other approximations are numerical. The resulting expression
for is accurate within about " 10%.
As derived in Appendix 7A, the resistance factor for lognormally distributed R and Q is
approximated by:
= R exp(- T COV R )
(Eq. 7-6)
7.4.1 Example Problem - Estimating Resistance Factors for the Nordlund Method
7-10
By using the resistance statistics for the Nordlund method determined previously (R = 1.04, COVR
= 0.30) and setting = 0.87 in Eq. 7-6, approximate values of can be calculated for T = 2.0 and
2.5. These values of are presented in Table 7-7.
Table 7-7
Values of for Nordlund's Method
Calculated Using the Simplified Approximate Reliability Method
Resistance Factor,
0.62
0.54
It is interesting to note the close agreement between the resistance factors in Table 7-6 with the
values in Table 7-7 which are calculated independent of the load statistics. Based on the results
shown in Table 7-7, a reasonable value of would be about 0.55 to 0.60. The average value is 0.58.
This agreement is depends on using = 0.87 as a means of adjusting for the effects of the
approximations involved in the method.
7.5 Fitting with ASD to Evaluate Resistance Factors
Values of can also be determined by "fitting" the value of to the conventional factor of safety
that would be used in allowable stress design. The procedure for fitting is derived in Appendix 7B
from the basic forms of the design equations used in allowable stress design (ASD) and load and
resistance factor design (LRFD).
As derived in Appendix 7B, the resistance factor estimated by fitting with ASD can be calculated
from
=
D (Q D / QL ) + L
FS(1 + QD /QsubL)
(Eq. 7-7)
(Eq. 7-8)
7.5.1 Example Problem - Estimating Resistance Factors for the Nordlund Method
The value of calculated using this fitting equation depends on the ratio of QD/QL, and the factor of
safety that would be used in ASD. Using values of QD/QL = 1.0 and 3.0 as used in Example 7.4, and
values of FS = 2.25 and 2.75, which are reasonable values for the Nordlund method, the following
results are obtained:
7-11
It can be seen that these values do not differ much from the values of calculated using reliability
theory, which are listed in Tables 7-6 and 7-7. This indicates that FS = 2.25 and FS = 2.75
correspond closely with values of T = 2.0 and T = 2.5 in this case.
Based on the results shown in Table 7-8, it would be logical to select a value of in the range from
0.55 to 0.60. The average value of by the fitting method is 0.58, the same as the average value
calculated using reliability theory.
Table 7-8
Values of for Nordlund's Method
Calculated Using Fitting with ASD
Values of QD/QL
1.0
3.0
FS = 2.25
0.67
0.61
FS = 2.75
0.55
0.50
In this chapter, straightforward relationships have been presented for calculating LRFD resistance
factors. These relationships are expressed by equations that can be used in three different situations.
y
If both load and resistance statistics (bias factor and coefficient of variation
COV) are known, or can be estimated, Eq. 7-1 should be used
If no statistics on resistance are available, but the factor of safety FS used for
ASD design is known, Eq. 7-7 can be used
In the examples on calculating for Nordlund's method, all three methods -- the basic and the
simplified reliability calculations, and the method of fitting with ASD -- suggest that a reasonable
value of would be 0.55 to 0.60. Thus, in this particular case, there is no conflict between the
results of the three methods of evaluating , because the values of T (2.0 and 2.5) are compatible
with the values of safety factor (2.25 and 2.75) that were used in the calculations.
Even though all three methods result in essentially the same values of , it is important to consider
whether the results of the calculations are reasonable. One way of judging the reasonableness of the
result is by comparing the -factor calculated for the Nordlund method with the -factors for other
7-12
methods for evaluating the capacities of piles in cohesionless soils -- SPT-method, = 0.45, CPTmethod, = 0.55. As compared with these values, does a value of = 0.55 to 0.60 for the Nordlund
method seem reasonable? A more reliable method should have a larger value of . Is Nordlund's
method more reliable than the SPT and CPT methods? Does Nordlund's provide a more reliable
estimate of pile capacity? Does the calculated -factor seem appropriate? Would it be more
appropriate if it was increased or decreased? These questions need to be answered before a final
selection of the resistance factor for the use of Nordlund's method in a particular region or for a
particular site is made.
In this chapter, a number of trends relating to the resistance factor have been observed. When T or
FS is increased, the value of decreases. This trend is expected because both T and FS are
measures of the safety margin. If the safety margin is to be increased, a smaller fraction of the
calculated resistance must be used for design, which would require use of a smaller value of .
Another trend observed is that an increase in QD/QL results in a decrease in the value of . This
trend occurs because the larger proportions of dead load with a smaller load factor than live load
results in a smaller effective weighted load factor. In the basic design equation -- Eq. 7A-6 -- if the
load factor is smaller, the -factor must be smaller to preserve the equality.
Another observation is the direct relationship between R and . This can be seen in Eq. 7-6 and
also in Eq. 7-1. If R increases by 10 percent, increases by 10 percent. This one-to-one increase
makes sense, because if a resistance model consistently under predicts the resistance, the -factor
should increase. This increase in the value of reflects increased confidence that this model will
underestimate the capacity.
Finally, because of the way bias factors and coefficients of variation are combined in Eq. 7-4 and
Eq. 7-5, the more we know about the resistance, the less our confidence in it. This is ironic, but true.
If we assume that only two variables influence the reliability of a resistance prediction, we combine
their uncertainties using Eq. 7-5 and calculate a value for COVR. If we discover that a third variable
also contributes uncertainty and put its COV into Eq. 7-5, the resulting COVR is larger. The increase
in knowledge has increased our estimate of uncertainty. As we add more and more variables, the
calculated -factor continues to decrease to reflect the increased uncertainty. Some sense can be
brought to this dilemma by always calculating a -factor based on fitting to ASD and selecting a
value that does not differ greatly from the value determined by fitting.
7-13
1.
Name and briefly describe the two general methods of calibrating resistance factors (Refer to
Section 7.2)?
2.
The Asix sigma rule@ can be used to provide rule-of-thumb estimates of the standard
deviation, , and coefficient of variation, COVR. Describe the Asix sigma rule@ (Refer to
Section 7.3.1).
7-14
Appendix 7A
Derivation of Simplified Reliability Calculations of Resistance Factors
It is assumed R and Q are lognormally distributed and statistically independent, the safety index , is
given by Eq. 3-20 in Chapter 3 as:
=
) (1 + COV )
)(1 + COV )]
ln (R Q ) 1 + COVQ2
[(
ln 1 + COVR2
2
R
(Eq. 7A-1)
2
Q
where R and Q are the mean values, COVR and COVQ are the coefficients of variation of
resistance, R, and load, Q, respectively.
If it is assumed that COVR and COVQ are less than about 0.30, Eq. 7A-1 can be greatly simplified.
The quotient under the radical in the numerator will be close to unity, and the function under the
radical in the denominator can be expressed as the sum of two logarithmic functions, that is:
ln [(1 + COV Q2 ) (1 + COV 2R )] = ln(1 + COV Q2 ) + ln(1 + COV 2R )
(Eq. 7A-2)
1 4
1
1
x + x 6 - x8 + K
4
2
3
(Eq. 7A-3)
For an infinite series with alternating signs, the error is no more than the first neglected term. For
COVs less than 0.30, the first term of the infinite series represents the function in Eq. 7A-3 within
five percent, so that Eq. 7A-2 can be written as
ln [(1 + COV Q2 ) (1 + COV 2R )] COV Q2 + COV 2R
(Eq. 7A-4)
Because there are two terms in Eq. 7A-4, each with an error of approximation of about five percent,
the combined error is about ten percent. The overall inaccuracy in this approximate method can be
minimized by judicious selection of the value of the factor , which is discussed below in
connection with Eq. 7A-6.
Using Eq. 7A-4, the expression for given by Eq. 7A-1 can be expressed as:
ln (R / Q)
2
COVQ
+ COVsub R 2
ln R - ln Q
2
COVQ
+ COV 2R
(Eq. 7A-5)
To separate the effects of R and Q, the linear approximation for the square root term suggested by
Lind (1971) is used, that is:
7-15
(Eq. 7A-6)
where is a constant whose value can be used to achieve the best agreement with the results of Eq.
7-1. Trial and error has shown that a value of = 0.87 is best for the ranges of values of statistics
for load and resistance involved in foundation design.
Using the relationship given by Eq. 7-6, Eq. 7A-5 can be expressed as:
ln (R / Q) = (COVQ + COV R )
(Eq. 7A-7)
(Eq. 7A-8)
(Eq. 7A-9)
(Eq. 7A-10)
By comparing Eq. 7A-10 with the basic design equation Eq. 7A-11:
R n = Qn
(Eq. 7A-11)
(Eq. 7A-12)
This approximate resistance factor is expressed only in terms of its own statistics and some fraction
of a target safety index. It does not involve the load statistics. The optimum value of the fitting
factor has been found to be about 0.87, as noted previously.
7-16
Appendix 7B
Derivation for Fitting with ASD
The procedure for fitting is derived from the basic forms of the design equations used in ASD and
LRFD.
For ASD:
R
Qi
FS
(Eq. 7B-1)
where:
R=
FS =
Qi =
For LRFD:
R i Qi
(Eq. 7B-2)
where:
=
i =
Resistance factor
Load factor for load type I
If both equations are used to derive expressions for resistance, these are:
R (FS) Qi
(Eq. 7B-3)
and
R
i Qi
(Eq. 7B-4)
When these expressions for R are set equal, the result is:
(FS) Qi =
i Qi
(Eq. 7B-5)
7-17
i Qi
(Eq. 7B-6)
(FS) Qi
D QD + L QL
(FS)(QD + QsubL)
(Eq. 7B-7)
D (Q D / Q L ) + L
(FS)(1 + QsubD / QL)
(Eq. 7B-8)
7-18
CHAPTER 8
SPREAD FOOTING DESIGN
8.1 Introduction
For both Allowable Stress Design (ASD) and Load and Resistance Factor Design (LRFD), the
design of spread footing foundations requires consideration of geotechnical capacity, overall
stability and deformation limits. The design processes therefore require both establishment of
criteria for acceptable stress and deformation levels, and comparison of these criteria with stress and
deformation levels estimated from the design. This chapter:
y
With few exceptions, the design procedure for spread footing foundations using LRFD (A10.6) is
identical to the procedure followed using ASD. Generally, the ultimate bearing capacity, overall
stability and settlement are checked for spread footings subjected to vertical loads. Where spread
footings are subjected to lateral or inclined loads, the effect of inclined loading on the ultimate
bearing capacity is checked, and the foundation is checked for resistance to sliding. The following
sections summarize the general design processes for spread footing foundation design using the ASD
and LRFD approaches.
8.2.1 ASD Summary
Existing practice for geotechnical design of spread footing foundations follows the ASD approach,
wherein all uncertainty in the variation of applied loads transferred to the foundation(s) and the
ultimate geotechnical capacity of the soil and rock to support the loads are incorporated in a factor of
safety, FS. As a result, loads used for design, Q, consist of those actual forces estimated to be
applied directly to the structure. In LRFD terminology, this process is equivalent to applying a load
factor of 1.0 to the estimated forces. In ASD, six primary performance and failure conditions are
evaluated in the design of a spread footing foundation:
y
y
Settlement
Bearing Capacity
8-1
y
y
y
y
Sliding Resistance
Load Eccentricity (Overturning)
Overall Stability
Structural Capacity
Settlement
The design of spread footings by ASD is often controlled by deformation or settlement
considerations. Thus, the design of footings by ASD requires estimation of foundation settlements
under the applied loads, and comparison of estimated settlements with deformation criteria using the
following:
i n
(Eq. 8-1)
where:
i =
n =
Tolerable movement criteria are usually a function of the type of structure, and for bridges depend
primarily on the span length and whether the superstructure has fixed- or simply-supported spans. In
the case of abutments and other earth retaining structures, designs may be governed by criteria for
lateral deflection or tilt, as described in Chapter 11.
Bearing Capacity and Sliding
Following selection of a preliminary footing size based on settlement considerations, the ultimate
bearing capacity and sliding resistance, Rn, are estimated by available theoretical or semi-empirical
procedures. The suitability of the design with respect to bearing capacity and sliding are then
evaluated by determining the allowable vertical and horizontal load components, Qall, using:
Q Qall = R n /FS
(Eq. 8-2)
where:
Q=
Qall =
Rn =
FS =
The required FS with respect to bearing capacity and sliding are generally specified by the
governing agency, and may be constant or variable. The AASHTO ASD FS (AASHTO, 1997b) for
bearing capacity and sliding are constant, as shown in Table 8-1. The AASHTO FS values assume
that designs are based on soil and rock properties determined through appropriate field and
laboratory testing and not presumptive properties. The relatively high FS for bearing capacity
reflects the acknowledgment that significant downward footing movements are required to
completely mobilize the shearing resistance within the soil below the footing.
8-2
Table 8-1
Factors of Safety on Ultimate Geotechnical Capacity
of Spread Footings for Bearing Capacity and Sliding Failure
(AASHTO, 1997b)
Required Minimum
Failure Condition
Factor of Safety (FS)
Bearing Capacity of Footing on
3.0
Soil or Rock
Sliding Resistance of Footing
1.5
on Soil or Rock
Other sources provide guidelines for variable FS values for bearing capacity which consider the
likelihood that maximum design loads will occur, the extent of subsurface exploration and the
consequences of failure. FHWA (Cheney and Chassie, 1993) suggest a variable factor of safety for
bearing capacity for footings on soil (Table 8-2) depending on the basis of soil strength estimates.
Another example of a suggested system of variable FS values, with a FS ranging from 2.0 to 4.0, is
shown in Table 8-3.
Table 8-2
Factors of Safety on Ultimate Bearing Capacity of Spread Footings on Soil
(Cheney and Chassie, 1993)
Suggested Minimum
Basis for Soil Strength Estimate
Factor of Safety (FS)
Standard Penetration Tests
3.0
Laboratory/Field Strength Tests
2.5
Table 8-3
Variable Factors of Safety on Ultimate Bearing Capacity of Spread Footings
(after Vesic, 1975)
Category
Typical
Structures
Category
Characteristics
Railway bridges
Warehouses
Blast furnaces
Hydraulic
Retaining walls
Silos
Highway bridges
Light industrial
and public
buildings
Apartment and
office buildings
8-3
3.0
4.0
2.3
3.0
2.5
3.5
2.0
2.6
2.0
3.0
2.0
2.3
(Eq. 8-3)
(Eq. 8-4)
The minimum required FS with respect to overall slope stability is generally specified by the
governing agency and should reflect uncertainty related to:
y
y
y
y
y
Method of analysis
Method of soil shear strength determination
Reliability and extent of subsurface information
Potential for variation over time in subsurface conditions from those
analyzed
Consequences of a failure
Required or recommended FS values vary significantly. FHWA (Cheney and Chassie, 1993)
suggests the minimum FS values presented in Table 8-4, modified for individual projects based on
the considerations listed above.
8-4
Table 8-4
Suggested Minimum Factors of Safety
for Overall Slope Stability From FHWA
(Cheney and Chassie, 1988)
Recommended Minimum
Condition
Factor of Safety (FS)
Highway Embankment Side Slopes
1.25
Slopes Affecting Significant Structures
1.30
(e.g., bridge abutments, major walls)
AASHTO ASD (AASHTO, 1997b) provides specific requirements for minimum FS values
considering the detail of subsurface exploration and potential consequences of failure, as indicated in
Table 8-5.
Table 8-5
Required Minimum Factors of Safety
for Overall Slope Stability From AASHTO ASD
(after AASHTO, 1997b)
Required Minimum Factor of Safety (FS)
Condition
Limited Exploration
Detailed Exploration(1)
Highway Embankment Slopes and
1.3
1.5
Retaining Walls
Slopes Supporting Abutments or
1.5
1.8
Abutments Above Retaining Walls
(1)
Soil and rock parameters and ground water levels based on in-situ and/or laboratory tests.
Structural Resistance
Finally, after all geotechnical deformation and capacity criteria are met and a suitable spread footing
size is determined, the structural design of the footing is performed using service loads and
allowable stresses (AASHTO, 1997b). Because the structural design of the footing is somewhat
independent of the geotechnical design, the structural design is sometimes performed by LFD or
LRFD to be consistent with the superstructure design.
8.2.2 LRFD Summary
Whereas ASD considers all uncertainty in the applied loads and ultimate geotechnical or structural
capacity in factors of safety or allowable stresses, LRFD separates the variability of these design
components by applying load and resistance factors to the load and material capacity, respectively.
When properly developed and applied, the LRFD approach provides a consistent level of safety for
the design of all structure components. Thus, the probability that a structure component will fail or
perform unacceptably is no different than any other component. As described in Section 4.3, the
resistance and deformation of supporting soil and rock materials and structure components must
satisfy the LRFD equations below. For the Strength Limit States:
i i Qi R n = R r
(Eq. 8-5)
8-5
(Eq. 8-6)
where:
i =
i =
Qi =
=
Rn =
Rr =
i =
n =
Relative to bearing capacity and sliding of a spread footing, the suitability of a spread footing with
respect to the geotechnical resistance can be obtained using Eq. 8-6, rewritten as:
i i Qi Qult = QR
(Eq. 8-7)
where:
i i Qi
=
Qult =
QR =
The load factors and load factor combinations used for design were presented in Chapter 4. In
general, values of i > 1.0 are used to evaluate ultimate ground or structure capacity at the Strength
Limit States, whereas the deformation performance of structures is evaluated at the Service I Limit
State using i = 1.0 (or i = 0.3 for wind loads). In ASD (AASHTO, 1997b), values of i = 1.0 (or i
= 0.3 for wind loads) are used to evaluate structures for both strength (allowable stress) and
serviceability (deflection). Therefore, analysis of foundation deformations (e.g., settlement or
lateral displacement) by LRFD and ASD are identical.
When using Eq. 8-5 for spread footing foundation design at the Strength Limit States, the following
values of can normally be used:
y
y
D =R = 1.00
I = 1.05 for structures deemed operationally important, 1.00 for typical
structures and 0.95 for relatively less important structures.
8-6
Determination of the operational importance of a structure (such as a bridge) is made by the facility
owner, as described in Chapter 4. The appropriate value of I, is then applied throughout the
superstructure design by the bridge engineer. The value of I selected by the superstructure
designer should then be applied in the foundation design. For the purpose of this chapter, the value
of I is assumed equal to 1.0.
When using Eq. 8-6 to evaluate a spread footing at any Service Limit State, D, R, and I = 1.0.
Values of load factor and load factor combinations for each applicable limit state must be developed
using the guidelines described in Chapters 3 and 4 and Section 8.2.2.1, and loads should be
developed as described in Chapter 4. The ultimate resistance, Rn, should be determined for each
type of resistance (e.g., bearing criteria, sliding, or overall stability) and the footing should be
checked for overturning based on load eccentricity criteria as described in Section 8.3. Values of
# 1.0 are applied when evaluating spread footing resistance for any strength limit state using Eq. 8-5.
Currently, the value of = 1.0 is applied when evaluating a spread footing for any service limit
state using Eq. 8-6. Selection and modification of resistance factors, , are described in Sections
8.2.2.2 through 8.2.2.4.
Structural design of footings is somewhat independent of the geotechnical design. Structural design
is performed using existing LRFD procedures using factored loads and the factored bearing pressure
distribution.
8.2.2.1 Limit States (A10.5)
In general, the design of spread footing foundations using LRFD requires evaluation of footing
suitability at various Limit States (i.e., applicable Strength Limit States and the Service I Limit
State). The selection of a Strength Limit State(s) depends on the type of applied loading (e.g.,
Strength I for design vehicle loading without wind or Strength II for permit vehicle loading). The
design considerations or performance limits which must be evaluated for footings designed at the
Strength and Service I Limit States are summarized in Table 8-6. As conditions warrant, it may also
be necessary to evaluate foundation performance at other limit states (e.g., Extreme Event I for
loading from earthquakes).
Table 8-6
Limit States for Design of Spread Footing Foundations
Performance Limit
Strength Limit State(s)
Service I Limit State
Settlement
U
Bearing Resistance
U
Sliding Resistance
U
Overturning (Eccentricity of
U
Base Pressure Resultant)
Overall Stability
U
Structural Capacity
U
More detailed discussion of procedures used to evaluate the various Performance Limit States is
provided in Section 8.3.
8-7
As an example of this process, Table 8-7 summarizes the results of analyses performed to develop
values for estimating the bearing capacity of spread footings on sand. For these analyses, the
variables included:
y
y
y
y
Although the CPT is known to be less variable than the SPT, the lower values associated with the
CPT-based procedure reflect the use of a lower factor of safety in the CPT-based procedure
compared to that used in the procedure based on SPT data. The lower factor of safety is employed
partly because the CPT data is obtained continuously through the boring depth, whereas SPTs are
performed intermittently (typically at a 1 meter center to center spacing) through the boring depth.
The selected values of resistance factor reflect adjustment of the reliability-based values to be more
consistent with ASD experience.
Table 8-7
Resistance Factors for Semi-Empirical Evaluation of Bearing Capacity
for Spread Footings on Sand Using Reliability-Based Calibration
(modified after Barker, et al., 1991b)
Resistance Factor
Factor
Average
Target
Estimation
of
Reliability Reliability Span
Selected
Fitting with Reliability
(m)
Method
Safety
Index
Index
ASD
Based
FS
T
10
0.37
0.49
0.45
SPT
4.0
4.2
3.5
50
0.37
0.53
0.45
10
0.60
0.52
0.55
CPT
2.5
3.2
3.5
50
0.60
0.57
0.55
8-8
Similar to ASD, LRFD requires an evaluation of footing safety against overturning based on the
location of the factored bearing pressure resultant on the base of the footing with respect to the
footing centroid. As described in Chapter 11, the eccentricity criteria for AASHTO LRFD (A10.6.3)
were calibrated directly with existing AASHTO ASD criteria through analysis of cantilever and
gravity retaining walls and abutments covering a wide range of load conditions, design heights and
dead to live load ratios (Barker, et al., 1991b). Variations in these parameters were found to have
little influence on the resultant location. Based on the calibration study, the location of the factored
bearing pressure resultant in AASHTO LRFD must be maintained:
y
y
Overall stability of slopes containing spread footing foundations is performed under AASHTO
LRFD at the Service I Limit State (A3.4.1, A10.6.2.2.4). The resistance factors for overall slope
stability analysis using LRFD were developed through direct calibration with ASD using the
following:
=
FS
(Eq. 8-8)
where:
=
=
FS =
Current common methods of analysis and particularly computer codes for evaluation of overall slope
stability do not lend themselves readily to the LRFD concept for the Strength Limit State, which
employs maximum and minimum load factors for permanent earth and dead loads. Consider, for
example, a simple slope composed of dry soil and analyzed using a limiting equilibrium method of
analysis circular arc failure surfaces. The loads and resisting forces on a typical slice of soil are
shown in Figure 8-1. As indicated, the driving force and frictional component of resistance along
the failure surface are a function of the soil weight, whereas the cohesive component of resistance
along the failure surface is independent of the soil weight. For the typical condition discussed in
Chapter 4, different load factors are applied to earth loads and surcharges for Strength Limit State
evaluations depending on whether the loads have a stabilizing or destabilizing effect. Therefore, the
load factors on the soil weight and any surcharge loads (and therefore the factored load effects)
should vary depending on their location along the failure surface (i.e., T = WT sin may either drive
of resist movement). Because no current method of analysis readily accommodates variable soil
weights, the LRFD Specification (AASHTO, 1997a) specifies evaluation of overall stability at the
Service Limit State (i.e., = 1.0). Further, because no reliability-based calibration of typical slope
stability methods has yet been performed, the LRFD Specification applies a single resistance factor
to both the frictional and cohesive soil shear resistance, independent of the method of soil shear
strength determination. The resistance factor specified in AASHTO (1997a) for a slope supporting
or containing a foundation is 0.65 (Table 8-8), which results in an equivalent ASD safety factor of
approximately 1.5 (from Eq. 8-8).
8-9
Figure 8-1
Circular Arc Slope Stability Failure Surface Showing Typical Soil Slice Forces
Resistance factors for the geotechnical design of spread footings are presented in Table 8-8. A
majority of the resistance factors in the table are based on design procedures which are commonly
used for ASD of spread footings. The resistance factors presented in the table were developed
mostly by calibration with reliability theory where sufficient statistical information was available
regarding particular design procedures or soil and soil-structure properties, tempered with
engineering judgment for some cases. Where statistical information was insufficient, resistance
factors were chosen by calibration with ASD (fitting) so that LRFD and current design would result
in footings having similar dimensions.
8.2.2.3 Comparison of Spread Footing Design Using LRFD and ASD
To illustrate the relative differences between LRFD and ASD, the equivalent LRFD factor of safety
(FSLRFD) has been determined for each of the methods presented in Table 8-8 for estimating the
geotechnical capacity of a spread footing with respect to bearing capacity, sliding and overall slope
stability. In the table, FSLRFD was determined as:
FSLRFD = /
(Eq. 8-9)
where:
=
=
Average load factor (assumed = 1.45 for bearing capacity and sliding and
taken = 1.00 for overall stability)
Resistance factor from Table 8-8
The applicable ASD factor of safety (FSASD) from Tables 8-1, 8-2 and 8-5 for each category is
also presented in Table 8-9.
8-10
Bearing
Capacity
Sand
Semi-empirical procedure using SPT data
Semi-empirical procedure using CPT data
Rational Method using f from:
SPT data
CPT data
Clay
Semi-empirical procedure using CPT data
Rational Method using Su from:
CPT data
Rock: Semi-empirical procedure
Plate Load Test
Precast concrete on sand using f from:
SPT data
CPT data
Concrete cast-in-place on sand using f from:
SPT data
CPT data
Precast or cast-in-place concrete on 150-mm thick
sand subbase on clay where Su < 0.5 n
Sliding
and Passive
Pressure
Using shear resistance (Su) measured from:
CPT data
Precast or cast-in-place concrete on 150-mm thick
sand subbase on clay where Su > 0.5 n
Soil on soil
Passive pressure component of sliding resistance
Geotechnical parameters are well defined and slope
Overall
does not support/contain a structure
Geotechnical parameters are based on limited data,
Stability
or slope supports/contains a structure
(1)
(2)
RESISTANCE(1)
FACTOR
0.45
0.55
0.35
0.45
0.50
0.60
0.60
0.50
0.60
0.55
0.90
0.90
0.80
0.80
0.85
0.85
0.80
0.85
1.00
0.50
0.85(2)
0.65(2)
Refer to Section 8.3 for description of design procedures for which =s have been calibrated.
From AASHTO LRFD A3.4.1
8-11
Table 8-9
Comparison of ASD Factor of Safety
with LRFD Equivalent Factor of Safety for Spread Footing Foundations
METHOD/SOIL/CONDITION
Bearing
Capacity
Sliding and
Passive
Resistance
RESISTANCE
FACTOR,
Sand
Semi-empirical procedure using
SPT data
Semi-empirical procedure using
CPT data
Rational Method using f from:
SPT data
CPT data
Clay
Semi-empirical procedure using
CPT data
Rational Method using Su from:
lab tests (UU Triaxial)
field Vane Shear tests
CPT data
Rock: Semi-empirical procedure
Plate Load Test
Precast concrete on sand using f
from:
SPT data
CPT data
Concrete cast-in-place on sand
using f from:
SPT data
CPT data
Precast or cast-in-place concrete
on 150-mm thick sand subbase
on clay where Su < 0.5 n
LRFD
EQUIVALENT
FACTOR OF
SAFETY, /(1)
ASD
FACTOR OF
SAFETY
0.45
3.2
3.0
0.55
2.6
2.5(2)
0.35
0.45
4.1
3.2
2.5(2)
2.5(2)
0.50
2.9
2.5(2)
0.60
0.60
0.50
2.4
2.4
2.9
2.5(2)
2.5(2)
2.5(2)
0.60
0.55
2.4
2.6
3.0
3.0
0.90
0.90
1.6
1.6
1.5
1.5
0.80
0.80
1.8
1.8
1.5
1.5
8-12
(3)
In ASD, passive resistance in front of a footing is often neglected in evaluating sliding resistance due to the relatively
large displacement required to fully mobilize the passive resistance.
(4)
A minimum factor of safety of 1.5 is specified for slopes supporting abutments or retaining walls by FHWA and
AASHTO ASD where a detailed subsurface exploration and/or laboratory testing program has been performed to define
subsurface conditions and soil properties.
The value of FSLRFD in Table 8-9 was determined assuming = 1.45. Actually, can range from
1.25 for structure component loads to 1.75 for live loads so that FSLRFD could vary from the value
shown in the table depending on the relative proportion of live to dead load for a particular structure.
However, assuming that = 1.45 (which corresponds to a dead to live load ratio of 3 to 2)
represents a reasonable approximation, values of FSLRFD are generally similar to FSASD, with FSLRFD
typically ranging from about 20% lower to 40% higher than FSASD for predictive methods generally
considered more and less reliable, respectively. Thus, while general agreement exists between
LRFD and ASD, the comparison depends on the load factor and method of capacity analysis used
for design. Accordingly, whereas spread footing foundation designs performed using LRFD will be
comparable to those using ASD, a precise or very close approximation between the two should not
typically be expected.
8.2.2.4 Modification of Resistance Factors
As indicated in Section 8.2.2.2, the LRFD resistance factors for geotechnical design of spread
footings in Table 8-8 were generally developed using a reliability-based calibration procedure,
except for the resistance factor for overall stability, which is based on calibration with ASD. As
described in Section 8.2.2.3, Table 8-9, application of these resistance factors in conjunction with
= 1.45 results in an "equivalent" factor of safety ranging from 2.4 to 4.1 for bearing capacity and 1.6
to 2.9 for sliding, depending on the soil conditions, loading conditions and method of geotechnical
capacity prediction.
In ASD, the designer or owner might decide to increase or decrease required factors of safety or
allowable design stresses in consideration of a number of factors, such as:
y
y
y
y
y
When using LRFD, similar flexibility to vary the required level of safety should also be available
and, in some cases, is inherent in the load and resistance factors used. Whereas the same factor of
safety is generally used in ASD regardless of the source of loading, the equivalent factor of safety in
LRFD (defined by Eq. 8-10) varies for a given resistance factor depending on the source of loading
as a function of the available load factors. As stated previously, an average load factor of 1.45 was
generally used to identify the equivalent safety factors in Table 8-9.
Although an average load factor of 1.45 is generally reasonable for a typical bridge abutment, pier or
8-13
a retaining wall, the average load factor could theoretically range from a low of 1.25 (i.e., all dead
load) to nearly 1.75 (i.e., extremely high live load). To modify the resistance factors for
geotechnical design of footings to account for average load factors other than 1.45 and equivalent
factors of safety other than those identified in Table 8-9, the following equation may be used:
FS
m = T x T T
FSD D
(Eq. 8-10)
where:
m =
T =
FST =
FSD =
T =
D =
Modifying resistance factors may seem reasonable, but such modification may not be consistent with
the goal of LRFD to achieve equal reliability against failure of structure components, unless the
factor of safety accurately models the reliability of the footing capacity predictive method used.
The resistance factors may be more appropriately modified through application of the probabilistic
procedures described in Chapter 3 to achieve the desired level of reliability if sufficient data are
available.
8.2.3 Summarized Comparison of ASD and LRFD
As noted before, the process used to develop a spread footing foundation design using LRFD differs
very little from the process used for ASD. The similarity is illustrated in the parallel flow charts in
Figure 8-2. Specific differences between the methods and other important issues are highlighted in
the following section.
Other aspects of the spread footing design such as identifying special considerations (e.g., potential
for loss of support through scour), developing a design foundation profile and determining
requirements for construction control are inherent aspects of the design process required for both
LRFD and ASD.
8.3 Performance Limits
Design of a spread footing foundation by either LRFD or ASD must provide adequate resistance
against geotechnical and structural failure and limit deformations to within tolerable limits. In
determining the footing size and details, and in establishing a suitable bearing level to meet the
criteria for vertical, inclined and/or moment loading, the design of spread footing foundations
requires consideration of many factors which can affect the spread footing foundation performance,
including:
8-14
Figure 8-2
Generalized Flow Chart for Spread Footing Design by ASD and LRFD
y
y
y
y
y
y
y
For these performance limits, there is no difference between LRFD and ASD analysis procedures.
The following sections highlight differences between LRFD and ASD in the performance criteria
and application of design procedures.
8.3.1 Displacements and Tolerable Movement Criteria (A10.6.2.2)
The vertical and lateral displacements of spread footings must be evaluated for all applicable dead
and live load combinations, and compared with tolerable movement criteria. Because evaluations of
structure displacements by LRFD are made at the Service I Limit State where = 1.0 and = 1.0,
methods used to estimate settlement by LRFD are identical to those used for ASD. Consequently,
footing settlements can be computed by conventional methods using semi-empirical correlations
with in-situ test results or measurements by in-situ or laboratory test methods to estimate
engineering soil properties. The potential for lateral displacement of footings should be estimated
where the footing is:
y
y
y
y
Vertical settlement can be a combination of the elastic, consolidation and secondary compression
movements. In general, the settlement of footings on cohesionless soils, very stiff to hard cohesive
soils, and rock with tight, unfilled joints will be elastic and will occur as load is applied. For
footings on very soft to stiff cohesive soils, the potential for consolidation and secondary
compression settlement components should be evaluated in addition to elastic settlement.
Methods for estimating settlement of footings are presented in the AASHTO LRFD Specification
(1997a) (A10.6.2.2.3) and are described in Gifford, et al. (1986), Cheney and Chassie (1993) and
Barker, et al. (1991a). The elastic settlements are estimated using elastic theory and a value of
elastic modulus based on the results of in-situ or laboratory testing. The consolidation and
secondary compression settlement on cohesive soils is estimated using consolidation theory and the
results of laboratory consolidation tests.
The tolerable movement of spread footing foundations (A10.6.2.2) depends on structural criteria
such as the type and size of the supported superstructure, as well as factors such as the cost and
difficulty of implementing repairs in the future, rideability, aesthetics and safety. Based on such
factors, limits on foundation movements have been set arbitrarily or based on empirical assumptions
(e.g., 25 mm maximum vertical movement) without consideration of actual structural performance.
However, due to the effects of creep, relaxation and redistribution of forces in bridge
superstructures, bridges can accommodate substantially more settlement than traditionally allowed
or anticipated in design. This conclusion is supported by the results of FHWA research (Moulton, et
8-16
al., 1985) which involved a performance survey of more than 200 bridges supported on shallow
foundations. The study showed that angular distortions (i.e., relative settlement of adjacent
foundations divided by the span length) of 0.008 or less for simple-span structures and 0.004 or less
for continuous-span structures are acceptable. Using these relationships, the maximum tolerable
settlement between foundations can be much greater than normally assumed. The final selection of a
tolerable movement should be made by the bridge designer in coordination with the geotechnical
engineer.
8.3.2 Bearing Resistance (A10.6.3.1 and A10.6.3.2)
Spread footing foundations must be designed to resist vertical loads without bearing failure of the
foundation soil or rock, and structural failure of the footing. Methods used for ASD to estimate the
ultimate geotechnical bearing resistance of a footing can be used for LRFD. Therefore, the factored
unit bearing resistance, qR, can be determined from:
q R = q n = q ult
where:
qR =
qn =
=
qult =
As the load on a footing is rarely concentric, the bearing pressure varies across the base of the
footing. The variation in bearing pressure is generally assumed to be linear. For foundations on
soil, compression of soil beneath the footing results in a redistribution of bearing stress to a more
uniform value. Therefore, for foundations on soil, the factored bearing resistance is compared to the
factored uniform unit bearing stress acting on the base of a concentrically-loaded footing area as
follows:
qR q =
i i Qi
(Eq. 8-12)
where:
q =
Factored uniform unit bearing stress (kPa)
i i Q i = Factored vertical load effect (kN)
A' =
Effective footing area (m2)
and A' is determined (A10.6.3.1.5) as:
A = (B - 2 eB )(L - 2 eL )
(Eq. 8-13)
where:
8-17
B=
L=
eB, eL =
For foundations on rock, the factored bearing resistance is compared to the factored maximum unit
bearing stress on the base of the effective footing area as follows (A11.6.3.2):
q R q max
(Eq. 8-14)
where:
qmax =
Various procedures are available for estimating the ultimate geotechnical bearing resistance of
spread footings supported on soil or rock using rational or semi-empirical methods. When using the
AASHTO LRFD Specification, only those methods referred to in Table 8-8 for which calibrated
-factors have been developed can be used without developing other method-specific resistance
factors. These methods include:
y
The resistance factors in Table 8-8 were developed for the specific design procedures referenced in
the AASHTO LRFD Specification (1997a). Other procedures can be used for LRFD provided
calibrated factors are developed using the methods described in Chapters 3 and 7. In lieu of more
detailed calibration, Table 8-9 and Section 8.2.2.4 can be used as a guide in selecting and modifying
resistance factors for various procedures to be consistent with ASD.
8-18
QR = Qn = Q + ep Qep
where:
Qn = Nominal sliding resistance of footing (kN)
= Resistance factor for shear between footing and foundation material
from Table 8-8 (dim)
Q = Nominal shear resistance between footing and foundation material
(kN)
ep = Resistance factor for passive earth pressure component of sliding
resistance from Table 8-8 (dim)
Qep = Nominal passive resistance of foundation material available during
the design life of the footing (kN)
The sliding resistance of a footing on cohesionless soil is based on the normal stress and the
interface friction between the foundation and the soil. In the AASHTO LRFD Specification, it is
assumed that at least 150 mm of cohesionless soil is placed beneath a footing bearing on a cohesive
soil subgrade. For this case, the sliding resistance is assumed to be a combination of shear through
the cohesive soil subgrade and the cohesionless subbase depending on the contact stress distribution
along the base of the footing.
Often, the passive resistance component of sliding resistance, Qep, is ignored or reduced because it is
difficult to assure that loss of ground (e.g., temporary excavation) or loss of contact by shrinkage
will not occur in the future.
The resistance factors in Table 8-8 were developed for the specific design procedures referenced in
the AASHTO LRFD Specification. Other procedures can be used for LRFD provided calibrated
factors are developed using the methods described in Chapters 3 and 7. In lieu of more detailed
calibration, Table 8-9 and Section 8.2.2.4 can be used as a guide in selecting and modifying
resistance factors for various procedures to be consistent with ASD.
8.3.4 Load Eccentricity (Overturning) (A10.6.3.1.5 and A10.6.3.2.5)
Spread footing foundations must be designed to resist overturning which results from lateral and
eccentric vertical loads. The criteria for evaluating overturning in ASD requires that the resultant
force be maintained within B/6 and/or L/6 of the foundation centroid for footings on soil
(A10.6.3.1.5), and within B/4 and/or L/4 for footings on rock (A10.6.3.2.5). For LRFD, the criteria
were revised to reflect the factoring of loads in LRFD. As a result, the eccentricity of footings for
factored loads must be less than B/4 and/or L/4 for footings on soil, and less than 3B/8 and/or 3L/8
for footings on rock. These limits were developed by direct calibration with ASD (Barker, et al.,
8-19
1991a; Barker, et al., 1991b). The criteria for ASD and LRFD are illustrated below in Figure 8-3 for
footings on soil and in Figure 8-4 for footings on rock.
Figure 8-3
Comparison of Resultant Force Location
for Footings on Soil Using ASD and LRFD
Figure 8-4
Comparison of Resultant Force Location
for Footings on Rock Using ASD and LRFD
As illustrated in the figures, the effect of factoring loads is to increase the eccentricity of the load
resultant such that the permissible eccentricity is increased for LRFD.
8.3.5 Overall Stability (A10.6.2.2.4)
Slopes containing spread footing foundations must be designed to resist loads associated with soil,
ground water, and footing and other surcharge pressures without a slope failure. Limiting
equilibrium methods of analysis which employ the Modified Bishop, Simplified Janbu, Spencer or
other generally accepted slope stability analysis methods can be used for LRFD as well as ASD. For
LRFD and currently available methods of analysis, the analysis must first be performed at the
Service I Limit State using unfactored loads and factored soil shear strengths. Thereafter, the
suitability of the slope and foundation with respect to overall stability can be determined as follows:
R R n iQi
(Eq. 8-16)
8-20
where:
Rn =
Qi =
R =
i =
The load factor of 1.00 and resistance factor of 0.65 in Table 8-8 for overall slope stability were
developed to provide for an equivalent LRFD factor of safety of approximately 1.5 for slopes
supporting or containing foundations. Tables 8-4, 8-5 and 8-8 and can be used as a guide in
selecting and modifying resistance factors for overall slope stability to achieve other levels of safety
consistent with ASD.
8.3.6 Structural Resistance (A10.6.4)
For structural design, the bearing pressure at the base of the footing is usually assumed to be uniform
or to vary linearly across the bottom of the footing for concentric and eccentric loading, respectively
(A10.6.3.1.5). The structural design is performed using the factored bearing pressure distribution
and the factored resistance of the reinforced concrete footing.
8.3.7 Other Considerations
8.3.7.1 Footing Embedment (A10.6.1.2)
If undermining by scour is a potential problem, footings must be constructed below the depth of
scour. Requirements for evaluation of scour are prescribed in the AASHTO LRFD Specification
and procedures for evaluating scour at bridges are described in FHWA (1991). In regions of the
U.S. where ground freezing is possible, foundations should be placed below the maximum depth of
frost penetration to prevent damage from frost heave.
8.3.7.2 Buoyancy and Uplift (A10.6.1.4 and A10.6.1.5)
Where footings are constructed below the anticipated ground water table, the effects of buoyancy
and uplift must be considered in evaluating the various performance limits. For sliding, the highest
ground water table will be critical because buoyancy will reduce the sliding resistance. For bearing,
the lowest water table will produce the maximum bearing pressure but the highest water table will
produce the minimum bearing resistance. Where differential water levels exist on either side of a
footing (e.g., for a retaining wall along a stream channel which is backfilled with poorly-draining
soil), the effects of differential water loads on sliding and bearing resistance, and the potential for
piping of soil fines below the footing should be evaluated.
8-21
8-22
The ultimate unit bearing resistance of a spread footing foundation in dense sand, qult, has been
estimated at 1000 kPa. The load on the footing at the Strength I Limit State is composed of a dead
load, QD, of 2000 kN (for which D = 1.25) and a live load, QL, of 1000 kN (for which L = 1.75).
Assuming a typical structure (i = 1.0), determine the minimum square footing size needed to satisfy
bearing resistance requirements for the following:
1. LRFD (using Eq. 8-12), if the ultimate resistance was estimated by:
a. Semi-empirical procedure using SPT data ( = 0.45)
b. Semi-empirical procedure using CPT data ( = 0.55)
c. Rational method using shear resistance (f) estimated from SPT data ( = 0.35)
qR =
i i Qi
A
(Eq. 8-12)
2. ASD (using Eq. 8-2), if the minimum required factor of safety against a bearing capacity failure(1)
is:
a.
b.
(Eq. 8-2)
(1)
Minimum factors of safety using FHWA criteria (Cheney and Chassie, 1993); AASHTO ASD minimum
factor of safety is 3.0 for all cases.
8-23
D QD
(kN)
2000
2500
2000
2500
2000
2500
2000
---
2000
---
LRFD
ASD
qall=qult/FS
qR=qult
(kPa)
(kPa)
LRFD: SEMI-EMPIRICAL PROCEDURE USING SPT DATA (=0.45)
1000
1750
--1000
--LRFD: SEMI-EMPIRICAL PROCEDURE USING CPT DATA (=0.55)
1000
1750
--1000
--LRFD: RATIONAL METHOD USING f ESTIMATED FROM SPT (=0.35)
1000
1750
--1000
--ASD: MINIMUM FACTOR OF SAFETY = 3.0
1000
----1000
--ASD: MINIMUM FACTOR OF SAFETY = 2.5
C
1000
----1000
QL
(kN)
L QL
(kN)
ASD
Q
(kN)
LRFD
i Qi
(kN)
8-24
qult
(kPa)
Minimum Required
Footing Size, Areq
(m x m)
You recall from Chapter 4 that the cantilever retaining wall in Figure 8-5 is being considered for a
grade separation between roadway lanes. The wall will be backfilled with a free draining granular
fill such that the seasonal high water table will be below the bottom of the footing. The vehicular
live load surcharge (LS) on the backfill is applied as shown in the figure.
Figure 8-5
Schematic of Student Problem
During the subsurface exploration, you determined that the foundation soils are predominantly clay
to a depth of 6 m below the proposed bottom of footing. Therefore, you have decided to place a
0.15-m thick blanket of compacted granular material below the footing to provide a uniform base for
foundation construction and improve sliding resistance. In performing the wall design, you assume
the following:
y
Dense sand and gravel underlies the clayey foundation soils so that the elastic
settlement of the dense sand and gravel will be negligible
8-25
The proposed wall backfill will consist of a free draining granular fill
The seasonal high water table will be at the bottom of the footing
Objective: Conduct a geotechnical design of a spread footing by LRFD and compare the results to
those with a design prepared by ASD.
Approach: You will perform the evaluation using the following steps:
y
Select footing length, L, and width, B, and determine unfactored and factored
bearing pressure distribution
Sliding: For ASD, ensure that unfactored ultimate geotechnical lateral load
capacity, Qn, of the footing divided by the factor of safety, FS, is greater than
or equal to the design load due to lateral load components, Q, and for LRFD,
ensure that the sum of the factored lateral load components, i i Q i , is
less than or equal to the factored geotechnical lateral load resistance, QR =
Qn
Overturning: For ASD, ensure that the resultant unfactored vertical load
component is located within L/6 and B/6 of the footing centroid, and for
LRFD ensure that the factored resultant vertical load component is located
within L/4 and B/4 of the footing centroid.
For convenience, the factored and unfactored loads and moments for critical load combinations from
the example problem in Section 4.6 of Chapter 4 are presented in Tables 8-10, 8-11, 8-12 and 8-13.
8-26
Group/
Item Units
W1
kN/m
W2
kN/m
W3
kN/m
PEV
kN/m
PLSV
kN/m
PLSv
kN/m
Pav
kN/m
VTOT
kN/m
V (Unf.)
31.8
10.6
35.3
169.5
34.2
12.8
35.2
329.4
Strength I-a
28.6
9.5
31.8
169.5
59.8
22.4
52.8
374.4
Strength I-b
39.7
13.3
44.2
228.8
59.8
22.4
52.8
461.0
Strength IV
47.7
15.9
53.0
228.8
52.8
398.2
Service I
31.8
10.6
35.3
169.5
34.2
12.8
35.2
329.4
Group/Item
Units
PLSh
kN/m
Pah
kN/m
HTOT
kN/m
H (Unf.)
21.3
58.6
79.9
Strength I-a
37.3
87.9
125.2
Strength I-b
37.3
87.9
125.2
Strength IV
87.9
87.9
21.3
58.6
79.9
Service I
Group/
Item Units
W1
W2
W3
PEVkNPLSV
PLSv
Pav
MvTOT
kN-m/m kN-m/m kN-m/m
m/m
kN-m/m kN-m/m kN-m/m kN-m/m
Mv (Unf.)
27.0
6.7
53.0
339.0
68.3
38.4
105.6
638.0
Strength I-a
24.3
6.0
47.7
339.0
119.6
67.2
158.4
762.2
Strength I-b
33.8
8.4
66.2
457.7
119.6
67.2
158.4
911.3
Strength IV
40.5
10.1
79.5
457.7
158.4
746.2
Service I
27.0
6.7
53.0
339.0
68.3
38.4
105.6
638.0
8-27
Group/Item
Units
PLSh
kN-m/m
Pah
kN-m/m
MhTOT
kN-m/m
Mh (Unf.)
53.3
97.9
151.2
Strength I-a
93.3
146.9
240.2
Strength I-b
93.3
146.9
240.2
Strength IV
146.9
146.9
53.3
97.9
151.2
Service I
Step 1: Calculate the Settlement of the Retaining Wall on Cohesive Soil Foundation
Assume embankment construction has been performed earlier and that consolidation settlement from
the embankment loading has already occurred. The original ground surface is located 1 m above the
footing level.
Divide the 6-m thickness of cohesive soil below the wall foundation into four layers as follows:
y
y
y
y
H1 = 1.0 m
H2 = 1.0 m
H3 = 2.0 m
H4 = 2.0 m
The depth of the footing below the existing ground surface is 1 m, so the depth to the center of each
layer from the final ground surface in front of the wall is:
y
y
y
y
d1 = 1.5 m
d2 = 2.5 m
d3 = 4.0 m
d4 = 6.0 m
(A) Calculate the effective overburden pressure at the center of each layer before wall construction
The depth of footing below the existing ground surface is 1 m so that the footing is located at the
annual high water level. The effective overburden stress at the center of each layer is:
!oi = di !2
!o1 = (1.0 m)(17.265 kN/m3) +(0.5 m)(7.455 kN/m3) = 21.0 kPa
!o2 = (1.0 m)(17.265 kN/m3) +(1.5 m)(7.455 kN/m3) = 28.4 kPa
8-28
329.4 kN/m
= 109.8 kPa
3m
Because the cohesive foundation soil is assumed to have consolidated under embankment loading
(i.e., !p = !o), the only consolidation settlement that will occur will be due to recompression under
the foundation loading, after removal of excess embankment soil and wall construction. The
recompression settlement can be computed from AASHTO LRFD A10.6.2.2.3c as:
_ +
_
Hc
1
+
1
+
_
_
e
e
o
o
o
o
(A10.6.2.2.3c-1)
where:
Ccr = 0.012 and eo = 0.6
Complete Table 8-14 and answer the conclusions below the table.
Table 8-14
Consolidation Settlement Calculation
Layer
i
(di - 1)/B
!/qo
!
(kPa)
!oi
(kPa)
Hi
(m)
1
2
3
4
0.17
0.50
1.00
1.67
0.95
0.80
0.55
0.34
104.3
87.8
60.4
37.3
21.0
28.4
39.6
54.5
1.0
1.0
2.0
2.0
8-29
oi +
Hi log
oi
0.78 m
0.61 m
0.80 m
0.45 m
= 2.64 m
where:
di - 1 = Depth to the center of layer I below the footing from finished grade (m)
!/qo = Boussinesq stress contour at center of a continuous foundation (dim) (A10.6.2.2.3a)
From Table 8-14, the estimated total consolidation settlement is:
Sc _
0.012
(2.64 m) = 0.020 m = 20 mm
1 + 0.6
Assuming the maximum tolerable settlement, Smax = 25 mm, the total estimated settlement is
acceptable.
Step 2: Eccentricity
The eccentricity of the retaining wall is checked in Table 8-15 as described in Section 8.3.4 by
comparing the calculated eccentricity, e, for each loading group to the maximum allowed
eccentricity (emax) using the relationship:
eB = B/2 - Xo
where:
B/2 =
Xo =
emax =
m
Location of the resultant from the toe = (MvTOT - MhTOT)/VTOT
B/4 =
m/4 =
m
For each load group, the total vertical forces (VTOT), horizontal forces (HTOT), moments due to
vertical forces (MvTOT) and moments due to horizontal forces (MhTOT) are obtained from Tables 810, 8-11, 8-12 and 8-13, respectively. Complete Table 8-15 using the information from these
tables. Note that the force and moment due to live load surcharge over the heel (PLSV) are not
included in the eccentricity (i.e., overturning) evaluation (i.e., VTOT = VTOT(Table 8-10) - PLSV(Table 8-10) and
MvTOT = MvTOT(Table 8-12) - PLSV(Table 8-12).
Group/Item
Units
VTOT
(kN/m)
Table 8-15
Summary of Eccentricity Check
HTOT
MvTOT
MhTOT
Xo
(kN/m) (kN-m/m) (kN-m/m)
(m)
Strength I-a
125.2
240.2
Strength I-b
125.2
240.2
Strength IV
87.9
146.9
Service I
79.9
151.2
8-30
eB
(m)
emax
(m)
For all cases, eB is#, = or $ emax (underline correct answer); therefore, the design is
adequate/inadequate (underline correct answer) in regard to eccentricity.
Step 3: Bearing Resistance
(A) Estimate the Bearing Pressures
From Section 8.3.2, the adequacy for bearing capacity is developed based on a rectangular
distribution of soil pressure ( q ) over the reduced effective area as indicated in Figure 8-6. For a
rectangular distribution:
L' =
B' =
eB =
Xo =
q =
Note that the force and moment due to live load surcharge over the heel (PLSV) are included in
the bearing resistance evaluation.
Complete Table 8-16 using information from Tables 8-10, 8-12 and 8-13 for each applicable limit
state.
8-31
Table 8-16
Summary of Factored Bearing Pressures
Group/Item
Units
VTOT
(kN/m)
MvTOT
(kN-m/m)
MhTOT
(kN-m/m)
Strength I-a
240.2
Strength I-b
240.2
Strength IV
146.9
Service I
151.2
Xo
(m)
q
(kPa)
[(
)(
kPa)(3.08) + (
kPa) =
kN/m3)(
m)(0.45)]
kPa > q from Table 8-16
Because the factored bearing resistance, qR, is less than/exceeds (underline correct answer) the
maximum factored uniform bearing stress, q = 158 kPa, the bearing resistance is
adequate/inadequate (underline correct answer).
Step 4: Sliding
Sliding of walls on clay is checked under AASHTO LRFD (1997a) using Figure 8-7.
From Section 8.3.3, the factored resistance, QR, against failure by sliding is:
QR = Q + ep Qep (Eq. 8-15)
where:
=
from Table 8-8
Q = Nominal shear resistance between footing and foundation material (kN)
Qep = Nominal passive resistance of foundation material available
throughout the design life of the footing (kN)
8-32
where the cohesion of the foundation soil (c2) = 150.0 kPa and one-half the factored normal stress
(qmax) is given in Table 8-17. As for the eccentricity check, the force and moment due to the
live load surcharge over the heel (PLSV) are not included in the sliding evaluation.
The actual base pressure (i.e., normal stress at the foundation/soil contact) will have a trapezoidal
shape except when the eccentricity is greater than B/6 (i.e., 0.5 m), at which point the base pressure
distribution becomes triangular and acts over a reduced base width. From Table 8-15, eB < 0.5 m for
all limit states. The values of qmax and qmin in Table 8-17 are calculated; therefore, for a
trapezoidal base pressure distribution as follows:
qmax = (VTOT/B) + [(6)(VTOT)(eB)/B2]
qmin = (VTOT/B) - [(6)(VTOT)(eB)/B2]
where VTOT and eB are obtained from Table 8-15.
Because c2 > qmax/2 in all cases, the normal stress at the footing/soil interface is used in the
calculation of Q.
8-33
For comparison with the total factored horizontal forces (HTOT) from Table 8-11, Q can be
computed as follows:
Q =
Using the relationships for qmax and qmin, the equation above for Q, from Table 8-8, and HTOT
from Table 8-11, complete Table 8-17 for each applicable limit state.
Table 8-17
Summary of Sliding Resistance
Group/Item
Units
qmax
(kPa)
qmin
(kPa)
qmax/2
(kPa)
Q
(kN/m)
Q
(kN/m)
HTOT
(kN/m)
Strength I-a
Strength I-b
Strength IV
Service I
Because the factored sliding resistance ( Q) calculated is greater/less (underline correct answer)
than the factored horizontal loading for all Strength Limit States, the sliding resistance is
satisfactory/unsatisfactory.
Summary
Table 8-18
Summary of Spread Footing Design by LRFD and ASD
LRFD
Performance Limit
Factored
Resistance/
Eccentricity
Limit
ASD
Factored
Required FS/
Actual FS/
Load/
U Eccentricity
U
Eccentricity
Eccentricity
Limit
Eccentricity
0.25 B
0.07 B
0.167 B
0.03 B
Bearing Resistance
282 kPa
158 kPa
3.0
4.2
Sliding Resistance
134 kN/m
125 kN/m
1.5
1.8
8-34
As summarized in Table 8-18, a comparable design is achieved by LRFD and ASD. Whereas the
ASD factors of safety for bearing resistance and sliding are fixed, however, the LRFD resistance
factors could possibly be increased with additional data accumulation and reliability-based
calibration for similar soils and loading conditions. As such, the LRFD provisions reflect the
reliability of the soil strength estimates and capacity prediction models and provide a more
rational basis for design than the ASD provisions. Therefore, with further data accumulation and
calibration, more reliable and economical designs might be achieved using LRFD.
8-35
8-36
CHAPTER 9
DRIVEN PILE DESIGN
9.1 Introduction
For both ASD and LRFD, the design of driven pile foundations requires consideration of
geotechnical and structural capacity and deformation limits. The design processes therefore require
both establishment of criteria for acceptable stress and deformation levels, and comparison of these
criteria with stress and deformation levels estimated from the design. This chapter:
y
Describes primary differences between driven pile design by LRFD and ASD
with respect to the comparison of loads and deformations to resistance and
tolerable deformations
With few exceptions, the design procedure for driven pile foundations using LRFD (A10.7) is
identical to the procedure followed using ASD. Generally, the ultimate bearing capacity and
settlement are checked for axially-loaded piles. Although design of laterally-loaded piles may be
governed either by lateral capacity or deflection criteria, mobilization of the ultimate lateral capacity
of the soil requires such large displacements that lateral capacity does not represent a realistic basis
for laterally-loaded pile design. The following sections summarize the general design processes for
driven pile foundation design using the ASD and LRFD approaches.
9.2.1 ASD Summary
Geotechnical Resistance
Existing practice for design of pile foundations follows the ASD approach. For ASD (AASHTO,
1997b), all uncertainty in the variation of applied loads transferred to the pile(s) and the ultimate
geotechnical capacity of the soil and rock to support the load are incorporated in a factor of safety,
FS. As a result, loads used for design, Q, consist of those actual forces estimated to be applied
directly to the structure. In LRFD terminology, this is equivalent to applying a load factor of 1.0 to
the estimated forces. Relative to axial loading of a pile, the ultimate axial geotechnical load
capacity, Rn = Qult, is estimated using any number of available methods. Then the suitability of the
design is evaluated by determining the allowable axial design load, Qall, using:
Q Q all = R n FS = Q ult FS
(Eq. 9-1)
9-1
where:
Q=
Qall =
Rn =
FS =
The required FS is generally specified by the governing agency, and may be constant or variable.
The AASHTO ASD FS for the axial geotechnical capacity of a single pile is presented in Table 9-1.
Table 9-1
Factor of Safety on Ultimate Axial Geotechnical Capacity
Based on Level of Construction Control
(AASHTO, 1997b)
Basis for Design and Type of
Increasing Design/Construction Control
Construction Control
U
U
U
U
U
Subsurface Exploration
U
U
U
U
U
Static Calculation
U
Dynamic Formula
U
U
U
U
Wave Equation
U
U
CAPWAP Analysis
U
U
Static Load Test
(1)
Factor of Safety (FS)
3.50
2.75
2.25
2.00
1.90
(1)
For any combination of construction control that includes a static load test, FS = 2.0.
As Table 9-1 shows, AASHTO ASD for pile design permits the FS to vary depending on the level of
control exercised during the design and construction phases. On this basis if Rn is developed using
information from a site-specific subsurface exploration based on a static calculation and wave
equation analysis, FS 2.75. However, if capacities are estimated using more reliable procedures
and/or are confirmed through field testing procedures(e.g., by performing dynamic measurements
during driving and wave equation analysis of pile driving operations or by conducting a static load
test, a reduced factor of safety could be used to reflect the greater confidence level achieved by these
improved methods for estimating and confirming the axial capacity of a pile.
Structural Resistance
The reduction of the structural capacity of a pile from an ultimate value (based on the yield or
ultimate strength depending on material type) to an allowable value accounts for all uncertainty in
the variation of applied loads and the ultimate structural capacity of the pile. Therefore, as with
geotechnical design, the structural design of a pile or pile group is performed by ASD using actual
estimated pile loads.
For structural design based on allowable stress, Eq. 9-1 can be written as:
all
A = Pall
(Eq. 9-2)
9-2
where:
all = Allowable axial stress in pile (kPa)
Pall = Allowable axial structural capacity (kN)
A = Cross sectional area of pile (m2)
In ASD, the ultimate unit stress for each material type is reduced to an allowable value by
application of a modifier to account for uncertainty in resistance and applied load. The modifiers are
based on a comprehensive study by Davisson, et al. (1983) which evaluated the in-service stress
level in piles based on structural analysis and case history data. The study used load and resistance
factor concepts to evaluate permissible stress levels in piles, and incorporated the effects of driving
stress levels, mislocation of piles, eccentricity of applied load, damage to piles due to driving and
inherent variability in pile material properties of steel, concrete and timber piles. As an example, the
maximum allowable stress, all, for a steel pile is determined in ASD by the following:
()(ecc)(HDF)
all =
Fy
LF
(Eq. 9-3)
where:
all =
=
ecc =
HDF =
Fy =
LF =
The maximum allowable stresses recommended for various pile types are presented in Table 9-2.
The allowable axial load, Pall, on a pile is:
Pall = all A
(Eq. 9-4)
where:
Pall = Allowable axial load on pile (kN)
A = Cross-sectional area of the structural component for which all is applicable
(m2)
9-3
Table 9-2
Allowable Stresses in Piles (GRL, 1996; AASHTO, 1997b)
Pile Type
Steel
Driving Damage Likely
0.25 Fy
0.33 Fy
Driving Damage Unlikely
Concrete-Filled Steel Pipe
0.25 Fy + 0.40 fc'(1)
Prestressed Concrete
0.33 fc' - 0.27 fpe(2)
Round Timber
Douglas-Fir - Coast
8.3
Douglas Fir - Interior
7.6
Lodgepole Pine
5.5
Red Oak
7.6
Southern Pine
8.3
6.9
Western Hemlock
(1)
Applied over cross-sectional area of steel pipe and cross-sectional area of concrete;
(2)
Applied over gross cross-sectional area of concrete; Fv = Yield strength of steel
(kPa); fc' = Concrete compressive strength (kPa); and fpe = Concrete compressive
stress due to prestressing (kPa).
Deformation
In addition to axial geotechnical and structural capacity evaluation, the design of driven piles by
ASD requires evaluations of pile deflections and comparisons with deformation criteria using the
following.
i n
(Eq. 9-5)
where:
i =
n =
Tolerable movement criteria are usually a function of the type of structure, and for bridges depend
primarily on whether the superstructure has fixed- or simply-supported spans. The design of
laterally-loaded piles by ASD is usually governed by criteria for deflection under actual applied
loads, except for pile groups containing batter piles for which the axial and/or bending capacity of
the pile can control.
9.2.2 LRFD Summary
Whereas ASD considers all uncertainty in the applied loads and ultimate geotechnical or structural
capacity in factors of safety or allowable stresses, LRFD separates the variability of these design
components by applying load and resistance factors to the load and material capacity, respectively.
When properly developed and applied, the LRFD approach provides a consistent level of safety for
the design of all structure components. As described in Section 4.3, the probability that a structure
9-4
component will fail or perform unacceptably is no different than any other component. For design,
the resistance and deformation of supporting soil and rock materials and structure components must
satisfy the LRFD equations below. For the Strength Limit States:
Q
i
Rn = Rr
(Eq. 9-7)
where:
i =
Relative to axial loading of a pile, the suitability of a pile with respect to the geotechnical and
structural resistance can be obtained using Eq. 9-4, rewritten as:
Q ult = Q r
(Eq. 9-8)
Pn = Pr
(Eq. 9-9)
=
Qult =
QR =
Pn =
Pr =
and
i
where:
i
The load factors and load factor combinations used for design were presented in Chapter 4. In
general, values of i > 1.0 are used to evaluate ultimate ground or structure capacity at the Strength
Limit States, whereas the deformation performance of structures is evaluated at the Service I Limit
State using i = 1.0 (or i = 0.3 for wind loads). In ASD (AASHTO, 1997b), values of i = 1.0 (or i
9-5
= 0.3 for wind loads) are used to evaluate structures for both strength (allowable stress) and
serviceability (deflection). Therefore, analysis of pile deformations (e.g., settlement or lateral
displacement) by LRFD and ASD are identical.
When using Eq. 9-6 for driven pile foundation design at the Strength Limit States, the following
values of can normally be used:
y
D =R = 1.00; and
Determination of the operational importance of a structure (such as a bridge) is made by the facility
owner as described in Chapter 4. The appropriate value of I is then applied throughout the
superstructure design by the structural engineer. The value of I selected by the superstructure
designer should then be applied in the foundation design. For the purpose of this chapter, the value
of I is assumed equal to 1.0. When using Eq. 9-7 to evaluate a pile foundation at any Service Limit
State, D, R, and I = 1.0.
Values of load factor and load factor combinations for each applicable limit state must be developed
using the guidelines described in Chapters 3 and 4 and in Section 9.2.2.1, and loads should be
developed as described in Chapter 4. The ultimate resistance, Rn, should be determined for each
type of resistance (e.g., axial, side, or group resistance) as described in Section 9.3.
Values of # 1.0 are applied when evaluating pile foundation resistance for any strength limit state
using Eq. 9-6. Currently, the value of = 1.0 is applied when evaluating a pile foundation for any
service limit state using Eq. 9-7. Selection and modification of resistance factors, , are described in
Sections 9.2.2.2 through 9.2.2.4.
9.2.2.1 Limit States (A10.5)
In general, the design of driven pile foundations using LRFD requires evaluation of pile suitability at
various Strength Limit States and the Service I Limit State. The selection of a Strength Limit
State(s) depends on the type of applied loading (e.g., Strength I for design vehicle loading without
wind or Strength II for permit vehicle loading). The design considerations which must be evaluated
for piles designed at the Strength and Service I Limit States are summarized in Table 9-3. As
conditions warrant, it may also be necessary to evaluate pile performance at other limit states (e.g.,
Extreme Event I for loading from earthquakes or Extreme Event II for vessel impact and ice
loading).
9-6
Table 9-3
Strength and Service Limit States
for Design of Driven Pile Foundations
Performance Limit
Strength Limit
State(s)
Service I Limit
State
U
U
U
U
U
U
Methods of evaluating piles at these various limit states are described in Section 9.3.
9.2.2.2 Resistance Factors (A10.5.5)
Resistance factors for geotechnical design of friction piles in soil based on static and in-situ test
methods of analysis were developed using the reliability-based calibration procedure described in
Chapter 3. Resistance factors for geotechnical design of bearing piles in clay and rock and of
friction and/or end bearing piles based on field testing of piles by Pile Driving Analyzer (PDA) and
static load tests are based on direct calibration with ASD. The reliability-based calibration
procedure involved:
y
Observing the variation in reliability levels with different span lengths, deadto live-load ratios, foundation geometry and methods used to estimate
ultimate resistance and load combinations
Selecting a target reliability, T, based on the margin of safety used for ASD
This process was used to develop resistance factors for the design of driven pile foundations. Table
9-4 summarizes the results of analyses performed to develop values for estimating the axial
resistance of piles. For these analyses, the variables included:
y
A dead to live load ratio of 3.69 which corresponds to a bridge with a span
9-7
length of 75 m
y
CPT
Type I
Type II
Type I
Type II
10
2.0
0.78
0.92
0.79
0.53
0.65
0.59
30
2.0
0.84
0.96
0.79
0.55
0.71
0.62
10
2.5
0.65
0.69
0.68
0.41
0.56
0.48
30
2.5
0.71
0.73
0.68
0.44
0.62
0.51
0.78
0.74
0.56
0.55
Average
0.70
0.50
0.55
0.55
Selected
Note: Type I refers to soils with Su < 50 kPa; Type II refers to soils with Su > 50 kPa
Pile
Length
(m)
SPT
0.48
0.51
0.36
0.38
0.43
0.45
Except for axial pile capacity estimates using the method, the selected
values of are generally equal to the average value of
A factor for the method much less than the average value was selected for use in the LRFD
Specification based on the limited number of reported cases where the method had been used in
conjunction with performance testing, and the engineering judgment of the code developers. The
reliability-based procedure for determining using the Nordlund method for estimating the ultimate
axial geotechnical capacity of a pile is presented in Chapter 7.
Resistance factors for the geotechnical design of driven pile foundations using the , and
methods for axial capacity and for other design methods and considerations suggested in the
AASHTO LRFD Specifications (AASHTO, 1996) are presented in Table 9-5.
9-8
METHOD/SOIL/CONDITION
RESISTANCE(1)
FACTOR
Ultimate
Bearing
Resistance of
Single Piles
Block Failure
Uplift
Resistance of
Single Piles
Group Uplift
Resistance
(1)
-method
-method
-method
End Bearing: Clay and Rock
0.70
0.50
0.55
Clay
Rock
Skin Friction and End Bearing: Sand
0.70
0.50
SPT-method
CPT-method
Skin Friction and End Bearing: All Soils
0.45
0.55
Load Test
Pile Driving Analyzer
Clay
-method
-method
-method
SPT-method
CPT-method
Load Test
Sand
Clay
0.80
0.70
0.65
0.60
0.40
0.45
0.35
0.45
0.80
0.55
0.55
Refer to Section 9.3 for description of design procedures for which resistance
factors have been calibrated.
The variation in -factors for determination of skin friction in clay provides a good indication of the
concepts inherent in LRFD. The -method, which is a total stress analysis method based on relating
the undrained shear strength of the soil to the adhesion between the pile and soil, has been shown to
give reasonable results for a wide range of soil conditions and therefore has a relatively high
resistance factor. In contrast, the -method, which is an effective stress method, has a lower
resistance factor, due in part to the fact that this method lumps the effects of such factors as in-situ
stress state, effective stress friction angle and pile-soil interface friction angle, all into a single
parameter, . Because the results of pile load tests or pile driving analyzer measurements during
9-9
installation provide more reliable estimates of axial capacity of a pile compared to other methods
(i.e., based on performance of an actual pile rather than estimates of soil properties), values of for
these methods are the highest. Values of for uplift and group loading are proportionally lower than
for compression loading to account for Poisson effects for uplift loading and interaction effects for
group loading.
The resistance factors presented in Table 9-5 (AASHTO, 1996) assume that the indicated capacity
prediction method is the only method used to estimate and/or field verify pile capacities. In
reality, static methods (e.g., and SPT methods) are often used to predict a capacity which is
subsequently field verified (i.e., by PDA and/or load tests). The resistance factor used during the
design stage should, therefore, reflect the reliability of the final method and extent of pile capacity
verification.
One relatively simple method to recalibrate static design method resistance factors for specified field
verification methods based on experience with specific pile and soil conditions is to equate the
factored static capacity prediction with the field-verified capacity prediction as follows:
Q R = Q RSTATIC = Q FIELD
(Eq. 9-10)
(Eq. 9-11)
or:
where:
QRSTATIC =
QRFIELD =
STATIC =
FIELD =
Subsequent field capacity verification can then be performed by verifying a nominal (or ultimate)
capacity identified as:
Q ULT = Q R FIELD
(Eq. 9-12)
In this manner, pile lengths estimated using static methods during the design stage would prove
adequate based on field capacities verified during construction. Also, the value for the final pile
installation would be equal to the target reliability index, T, for the final (and most reliable) method
of capacity verification. This procedure does, however, require accumulation and evaluation of pile
capacity prediction and verification data for similar pile installations and subsurface conditions.
The 1997 Interims of the AASHTO LRFD Specifications (AASHTO, 1997a) include an initial
9-10
attempt to incorporate the specified method and extent of field construction control and capacity
verification into the determination of resistance factors for axially loaded piles. However, the
modified resistance factors are not based on any specific calibration, and in many cases may
result in significantly more conservative designs than current ASD practice. Therefore,
additional consideration and future revision of the 1997 AASHTO LRFD requirements are
anticipated.
Resistance factors for structural design of piles using LRFD were developed through direct
calibration with ASD using the following:
= all n
(Eq. 9-13)
where:
=
=
all =
r =
For example, the ultimate capacity of a steel H-pile is defined as the yield strength, Fy, and the
allowable stress in ASD for a steel H-pile for conditions where pile damage is likely during driving
is 0.25 Fy. Therefore, the factor for steel H-piles for severe driving is:
= all n
= 1.45 (0.25 Fy/Fy) = 0.36
Values of for structural design of various pile types (rounded to the nearest 0.05) are listed in
Table 9-6. The determination of the ultimate structural capacities of various pile types is discussed
in Section 9.3.2.2.
The resistance factors for structural design of piles in the AASHTO LRFD Specifications (1997a)
are applied to nominal structural resistance, Pn, which are reduced for inadvertent load eccentricity
(A6.9.4.1) and reductions due to pile damage related to driving difficulty (AC10.5.5). The
Aeffective@ resistance factors in Table 9-6 include the base resistance factors for axial compression of
steel piles (0.90 in A6.5.4.2), concrete piles (0.75 in A5.5.4.2.1) and timber piles (0.90 in A8.5.2.2)
along with the reduction factor on nominal resistance, Pn, of 0.78 for H-piles and 0.87 for pipe piles
to account for inadvertent load eccentricity (A6.9.4.1) and suggested reduction factors of 0.875 and
0.75 for moderately difficult and difficult driving, respectively (AC10.5.5). As indicated by Table 96, the effective resistance factors implied in the AASHTO LRFD Specification (1997a) for
structural design of axially loaded piles are 35 to 40 percent to resistance factors derived through
calibration to the AASHTO ASD Specification (1997b). The difference is due to the use of a load
factor of 2.0 by Davisson (1983) in developing the ASD allowable stress levels for driven piles.
9-11
Table 9-6
Resistance Factors for Structural Design
of Axially-Loaded Piles Calibrated to ASD
Pile Type
Steel H-Piles:
Severe Driving Conditions
Good Driving Conditions
Steel Pipe Piles:
Severe Driving Conditions
Good Driving Conditions
Prestressed Concrete
Concrete-Filled Pipe
Steel Pipe
0.35
0.55
Concrete
Timber
0.55
0.75
0.65
(1)
Resistance factors provide stresses approximately equivalent to allowable stresses in
AASHTO (1997b) (Table 9-2) for = 1.45; (2) Resistance factors include AASHTO
LRFD (1997a) reductions in nominal resistance to account for unintended eccentricity
(A6.9.4.1) and driving difficulty (AC10.5.5), and assume = 1.45.
9.2.2.3 Comparison of Driven Pile Design Using LRFD and ASD
9.2.2.3.1 Geotechnical Design
To illustrate the relative differences between LRFD and ASD, the equivalent LRFD factor of safety
(FSLRFD) has been determined for each of the methods presented in Table 9-5 for estimating the axial
geotechnical capacity of a single pile or pile group. As presented in Table 9-7, FSLRFD was
determined as:
FS LRFD =
(Eq. 9-12)
where:
=
=
The ASD factor of safety (FSASD) from Table 9-1 for each category is also presented in Table 9-7.
9-12
Table 9-7
Comparison of ASD Factor of Safety
with LRFD Equivalent Factor of Safety
Geotechnical Strength Limit State for Axially-Loaded Piles
Method/Soil Condition
Resistance
Factor,
LRFD
Equivalent
Factor of Safety,
(for =1.45)(1)
ASD
Factor of
Safety(2)
value shown in the table depending on the relative proportion of live to dead load for a particular
structure. Assuming that = 1.45 represents a reasonable approximation, FSLRFD . FSASD
depending on the method of capacity estimation. While general agreement exists between LRFD
and ASD, the comparison depends on the actual average load factor and method of capacity
analysis. Therefore, while driven pile designs performed using LRFD will be comparable to those
using ASD, a precise or very close approximation between the two should not typically be expected.
9.2.2.3.2 Structural Design
In LRFD (AASHTO, 1997a), the resistance factors and capacity reduction factors account for the
same capacity reductions cited in Section 9.2.1 for ASD, and uncertainty in loading is accounted for
separately by a load factor. The factored axial resistance, Pr, is related to the allowable axial load,
Pall, by the following:
Pall =
Pn Pr
=
(Eq. 9-14)
where:
=
Pn =
Pr =
=
The provisions for structural design of axially loaded concrete, steel and timber piles are contained
in Sections 5, 6 and 8, respectively, of the AASHTO LRFD Specification (AASHTO, 1997a). The
provisions in Section 6 (A6.9.4.1) specify reduced nominal capacities, Pn, for steel H-piles
(0.78FyAs) to account for unintended eccentricities. Additional reduction factors are suggested in
Section 10 (AC10.5.5) of the AASHTO LRFD Specification (AASHTO, 1997a) to reflect moderately
difficult (0.875) and difficult (0.75) driving conditions. The Aeffective@ AASHTO LRFD resistance
factors for structural design presented in Table 9-6 include the reduction factors recommended and
suggested in AASHTO. The resistance factors in Table 9-6 which are calibrated to ASD are
calibrated to the maximum allowable stresses in Table 9-2. All resistance factors in Table 9-2
assume = 1.45.
9.2.2.4 Modification of Resistance Factors
9.2.2.4.1 Geotechnical Design
As stated in Section 9.2.2.2, the LRFD resistance factors for geotechnical design of piles in Table 95 were developed using a reliability-based calibration procedure. Application of these resistance
factors in conjunction with = 1.45 results in an "equivalent" factor of safety ranging from 1.8 to
4.2, depending on the soil conditions, loading conditions and method of geotechnical capacity
prediction, as described in Section 9.2.2.3.1.
In ASD, the designer or owner might decide to increase or decrease required factors of safety or
allowable design stresses in consideration of a number of factors, such as:
9-14
When using LRFD, similar flexibility to vary the required level of safety should also be available.
Additionally, whereas the same factor of safety is generally used in ASD regardless of the source of
loading, the equivalent factor of safety in LRFD (defined by Eq. 9-13) varies for a given resistance
factor depending on the source of loading. As stated previously, an average load factor of 1.45 was
used to identify the equivalent safety factors in Table 9-7 and to develop the resistance factors in
Table 9-5. Although = 1.45 is generally reasonable for a typical bridge abutment or pier, or a
retaining wall, the average load factor could theoretically range from a low of 1.25 (i.e., all dead
load) to nearly 1.75 (i.e., extremely high live load).
To modify the resistance factors for geotechnical design of piles to account for average load factors
other than 1.45 and equivalent factors of safety other than those identified in Table 9-7, the
following equation may be used:
FS
m = T T
FS D
T
D
(Eq. 9-15)
where:
m =
T =
FST =
FSD =
T =
D =
9-15
To modify the resistance factors for structural design of piles to account for other than 1.45 and
allowable stresses other than those presented in Table 9-2, the following equation may be used:
m = T allA
allT
T
D
(Eq. 9-16)
where:
T = Tabulated resistance factor from Table 9-6 (dim)
allA = Actual allowable stress (kPa)
allT = Tabulated allowable stress from Table 9-2 (kPa)
Modifying resistance factors may seem reasonable, but such modification may not be consistent with
the goal of LRFD to achieve equal reliability against failure of structure components, unless the
factor of safety accurately models the reliability of the pile capacity predictive method used. The
resistance factors may be more appropriately modified through application of the probabilistic
procedures described in Chapter 3 to achieve the desired level of reliability if a sufficient amount of
data is available.
9.2.3 Summarized Comparison of LRFD and ASD
As noted before, the process used to develop a pile foundation design using LRFD differs very little
from the process used for ASD. The similarity is illustrated in the parallel flow charts in Figure 9-1.
Specific differences between the methods and other important issues are highlighted in the
following section. Other aspects of pile design such as identifying special considerations (e.g.,
potential for loss of lateral support or negative loading), developing design foundation profile and
determining requirements for construction control (e.g., dynamic monitoring or load testing) are
inherent aspects of the design process required for both LRFD and ASD.
9.3 Performance Limits
Design of a driven pile foundation by either LRFD or ASD must provide adequate resistance against
geotechnical and structural failure and limit deformations to within tolerable limits. In determining
the pile section and details to meet these criteria for axial, lateral and/or moment loading, the design
of driven pile foundations requires consideration of many factors which can affect pile performance,
including:
y
y
y
y
y
y
y
y
y
Figure 9-1
Generalized Flow Chart for Driven Pile Design by LRFD and ASD
9-17
For these design factors, there is no difference between LRFD and ASD analysis procedures. The
following sections highlight differences between LRFD and ASD in the performance criteria and
application of design procedures.
9.3.1 Displacements and Tolerable Movement Criteria (A10.7.2)
The vertical and lateral displacement of driven piles must be evaluated for all applicable dead and
live load combinations, and compared with tolerable movement criteria. Because evaluations of
structure displacements by LRFD are made at the Service I Limit State where = 1.0 and = 1.0,
methods used to estimate settlement and lateral displacement by LRFD are identical to those used
for ASD. Consequently, axial compression (or tension) of the pile section can be computed by
elastic methods and axial and lateral ground displacements can be determined by conventional
methods using empirical correlations with in-situ test results or measurements by in-situ or
laboratory test methods to estimate engineering soil properties. Lateral displacements need to be
evaluated if:
y
y
y
For pile groups in cohesionless soils, settlements will occur immediately as the pile group is loaded.
Pile groups in cohesive soils, however may experience long-term consolidation settlement as well as
immediate elastic settlement. Elastic settlement generally predominates in overconsolidated clays,
whereas consolidation settlement generally predominates in normally consolidated clays. The
consolidation settlement of pile groups in cohesive soil is typically estimated using the equivalent
footing procedure (GRL, 1996) (A10.7.2.1). During the past decade, computer programs have
become available for axial and lateral load-deflection analysis of deep foundations based on the t-z
and p-y curve methods of analysis.
The tolerable axial and lateral movement of driven piles should be based on criteria developed by
the structural engineer for the superstructure, or by consideration of the effects of foundation
movements on adjacent structures. The AASHTO LRFD Specification (AASHTO, 1997a) restricts
lateral movements to 38 mm (A10.7.2.2). In some cases, the tolerable lateral movement is fixed at a
small displacement (e.g., 12 mm based on observed acceptable performance and engineering
judgment) to ensure acceptable structure performance.
9.3.2 Axial Resistance
9.3.2.1 Geotechnical Resistance (A10.7.3)
Pile foundations must be designed to resist axial loads without structural failure of the pile, and
without excessive deflection. Methods used for ASD to estimate the ultimate axial geotechnical
resistance of a single pile can be used for LRFD. Therefore, the ultimate axial geotechnical
resistance of piles subjected to axial loading, Rn, can be determined as:
R n = Q ult = Q p + Q s
(Eq. 9-17)
9-18
and the factored axial geotechnical resistance, QR, can be determined as:
R r = Q R = Q ult = qp Q p + qs Q s
for which:
Qp = q p A p
Qs = q s A s
and
where:
qp, qs =
Qp, Qs =
qp, qs =
Ap, As =
Various procedures are available for estimating the ultimate axial geotechnical capacity of driven
piles in soil using semi-empirical methods and in-situ testing, and for piles bearing on or in rock
using semi-empirical methods. When using the AASHTO LRFD Specification, only those methods
referred to in Table 9-5 for which calibrated -factors have been developed can be used without
developing other method-specific resistance factors. These methods for a single pile include:
y
Static Load Test - A method for estimating total load capacity based
on tests (ASTM, 1996) representative of pile, load and subsurface
conditions expected for the prototype piles (AASHTO, 1997a)
(A10.7.3.6).
With exception of the -method, these methods are used by FHWA for evaluating the geotechnical
axial capacity of driven piles (GRL, 1996). In addition, FHWA recommends (GRL, 1996) use of a
semi-empirical procedure by Nordlund (1963) for estimating qs and qp for piles in cohesionless soils,
and recognizes (GRL, 1996) LPC-method for estimating pile capacity based on CPT results
(Bustamante and Gianeselli, 1983). FHWA also recognizes use of wave equation methods of
analysis (without PDA) for estimating static pile capacity in soil (GRL, 1996). These and other
methods can be used for LRFD provided calibrated -factors are developed using the methods
described in Chapters 3 and 7.
9.3.2.2 Structural Resistance (A10.7.4)
In addition to meeting the requirements for adequate geotechnical resistance of piles subjected to
axial and lateral loading, the axial and flexural resistance of the pile section must also be adequate
for all potential loading cases. The ultimate unit structural capacity, n, of a pile is based on the
yield strength, Fy, for steel, the unconfined compressive strength, fc', for concrete, Fy and fc' for
reinforced concrete and composite sections, and the crushing strength parallel to the grain, Fco, for
timber. To evaluate an axially-loaded pile for structural resistance, the maximum factored stress,
max, is compared to the factored unit resistance, r, which is determined as follows:
r = n
(Eq. 9-21)
where:
=
n =
9-20
Relative to axial loading of a pile, Eq. 9-20 can be multiplied by the cross-sectional area of the pile
and written as:
Pr = Pn
(Eq. 9-22)
where:
Pn =
Pr =
The factor used in Eq. 9-21 and presented in the left column of Table 9-6 was developed by direct
calibration with ASD to include the effects of manufacturing variability, accidental load eccentricity
and hidden defects due to damage during pile driving as described in Section 9.2.2.2.
9.3.3 Lateral Resistance (A10.7.3.8, A10.7.3.11)
Pile foundations must be designed to resist lateral loads without structural failure of the pile, and
without excessive deflection (i.e., without exceeding the tolerable lateral movement). Although
passive failure of the soil into which the pile is embedded is a potential failure mode, such failure
occurs only at relatively large deflections which generally exceed tolerable movements. Therefore,
design of piles subjected to lateral loads is commonly based on structural capacity and loaddeflection behavior considerations. For piles subjected to both axial and lateral loading, the
structural resistance of the pile as a beam-column must be checked using appropriate interaction
equations calibrated with consideration of the load factors being utilized.
Methods used for ASD to estimate the lateral resistance of a single pile or pile group can also be
used for LRFD. For homogeneous foundation materials and a fixed- or free-head condition at the
pile cap, simplified methods of analysis (e.g., GRL, 1996; and Barker, et al., 1991a) can be used.
For more complex foundation and loading conditions and intermediate levels of fixity at the pile cap,
the analysis of laterally-loaded piles and pile groups is usually accomplished using computer
programs based on the p-y curve method of lateral load-deflection analysis. These programs include
COM624P (Wang and Reese. 1993) for individual vertical piles in level or sloping ground and
GROUP (Reese, et al., 1994) for a group of vertical and battered piles in level ground. These
methods of analysis incorporate p-y curves to model the nonlinear lateral load-deflection behavior
for each soil type along the length of the pile. When using these methods at the LRFD Strength
Limit State, the model pile or pile group is subjected to the factored lateral and axial loads and
factored moment, and the resulting factored axial and bending stresses are compared with the
factored axial and bending capacities of the pile.
Determine the maximum lateral ground line deflection at the Service Limit
State and the maximum moment at the Strength Limit State for an individual
pile considering installation method (e.g., jetting or preaugering) for the
selected pile section.
9-21
2.
Repeat Step 1 if the lateral ground line deflection exceeds the tolerable
deformation or results in axial or combined stresses which exceed the
maximum factored axial resistance of the pile or do not meet the appropriate
interaction criteria.
3.
The pile section is acceptable for the design loads if neither the lateral
groundline deflection nor the stress criteria from Step 2 is exceeded.
In practice, the piles in a group are tied together with a pile cap which is typically embedded below
the ground surface. In the analyses of the lateral load resistance of this type of system, the
contribution of passive resistance of the embedded portion of the pile cap is typically neglected.
Neglecting passive resistance of the embedded pile cap can provide a significant degree of additional
conservatism at both service and ultimate load levels.
FHWA promotes the use of vertical piles to resist lateral loads. However, if the lateral loads acting
on a foundation are large, batter piles can be used. Batter piles transmit lateral loads predominantly
as an axial force into the soil, and should be avoided when:
y
Methods such as GROUP (Reese, et al., 1994) used for ASD to estimate the lateral resistance of a
pile group containing batter piles can also be used for LRFD.
9.3.4 Other Considerations
9.3.4.1 Group Effects (A10.7.3.10)
Adjustments used in ASD to account for the effects of group action can also be used for LRFD. For
axial loading, the effect of closely-spaced piles can be considered in analyses by modifying the
nominal resistance of each pile using an efficiency factor, , to account for the:
y
y
y
In cohesive soils, the ultimate axial resistance of a pile group may be less than the cumulative
resistance of the individual piles due to overlapping of zones of shear deformation in the soil
between the piles (A10.7.3.10.2). In stiff cohesive soils, there is no loss in resistance due to group
effects. However, in soft cohesive soils where the pile cap is not in contact with the underlying soil,
the resistance of a pile group will be less than the cumulative resistance of an equal number of
individual piles if the center to center (CTC) spacing between piles is less than 6 D. For closer pile
9-22
spacings, the resistance of each pile can be estimated by multiplying the nominal capacity by ,
where:
y
y
y
The ultimate axial capacity of a pile group in cohesive soil is then the lesser of:
y
The sum of the individual resistances of each pile in the group, modified by
The resistance of an equivalent pier consisting of the piles and the block of
soil bounded within the piles
For a pile group in cohesionless soil (A10.7.3.10.3), = 1.0 regardless of the spacing of piles or
contact between the pile cap and the ground. No reduction in capacity is used for a pile group in
sand due to the increase in soil density that usually occurs when piles are driven into sand.
If a pile group is embedded in a stiff soil deposit overlying a weaker deposit (A10.7.3.10.4), the
potential exists for a punching failure of the pile group through the stiff soil into the soft soil. In
addition, the potential for settlements in the soft layer must be evaluated.
9.3.4.2 Negative Loading (A10.7.1.4)
Methods in ASD to estimate the magnitude and location of the maximum downdrag load can also by
used in LRFD. General comments regarding downdrag load estimation are presented in Chapter 4.
Briaud and Tucker (1994a) indicate that the potential for downdrag loading should be considered in
design when the indicators in Table 9-8 are present. In terms of performance limits, downdrag
generally poses a foundation settlement concern for friction piles and for end-bearing piles bearing
on a very stiff layer such as very dense sand or rock.
Table 9-8
Conditions When Downdrag Should be Considered in Design
(after Briaud and Tucker, 1994a)
1
Total settlement of the ground surface > 10 mm
2
Settlement of ground surface after pile driving > 1 mm
3
Height of embankment filling on ground surface > 2 m
4
Thickness of soft compressible layer >10 m
5
Water table drawn down >4 m
6
Piles length > 25 m
WARNING: Downdrag can occur even if the above conditions are not met.
9.3.4.3 Uplift Loading (A10.7.1.9, A10.7.3.7)
Uplift loading on piles can be caused by lateral loads from supported structures, buoyancy effects or
expansive soils. Evaluation of the ultimate uplift resistance of a single pile or pile group by ASD
9-23
and LRFD are conducted using the same methods used to estimate ultimate axial side resistance, QS,
of the pile or group to axial compression loading, but with a 20 percent reduction in QS. In the
AASHTO LRFD Specification (AASHTO, 1997a), the 20 percent reduction in QS for uplift is
incorporated in the resistance factors for uplift in Table 9-5. In addition, the structural resistance of
the pile and the pile/pile cap connection must also be checked.
9.3.4.4 Driving Stresses and Driveability (A10.7.1.16)
The effects of the pile driving process must be evaluated during design to ensure that the pile can be
installed without damage that could affect the capacity of the pile to perform within acceptable
limits. A driveability evaluation is needed because the highest pile stresses are usually developed
during driving to facilitate penetration of the pile to the depth needed to mobilize the required
resistance. However, the high strain rate and temporary nature of loading during pile driving allow a
substantially higher stress level to be used during installation than for service. The driveability of
candidate pile-hammer-system combinations can be evaluated using wave equation analyses (GRL,
1996).
The LRFD Specification (AASHTO, 1997a) provides performance limits for pile stresses during
driving consistent with the current AASHTO ASD Specification (AASHTO, 1997b). These driving
stress criteria are summarized in Table 9-9, and include limitations of (unfactored) driving stresses
in piles based on:
Table 9-9
Permissible Stresses During Pile Driving
(AASHTO, 1997a)
Pile Type
Steel
Compression:
Tension:
Concrete:
Compression
Tension
(1)
Prestressed Concrete:
Normal environments
- Compression
- Tension
Severe corrosive
environments
- Tension
Timber:
Compression
Tension
fc and fpe in MPa.
Stress Level
(MPa)
Resistance
Factor
AASHTO
LRFD
Provision
0.90 Fy
0.90 Fy
1.00
1.00
(A6.5.4.2)
(A6.5.4.2)
0.85 fc
0.70 Fy of Steel
Reinforcement
1.00
1.00
(A5.5.4.2.1)
(A5.5.4.2.1)
1.00
1.00
(A5.5.4.2.1)
(A5.5.4.2.1)
fpe
1.00
(A5.5.4.2.1)
Fco
Fco
1.15
1.15
(A8.5.2.2)
(A8.5.2.2)
9-24
Ultimate compressive strength of the gross concrete section " the effective
prestress after losses for prestressed concrete piles loaded in tension or
compression, respectively
Base resistance for the selected species of timber piles loaded in compression
parallel to the grain
Based on the results of full-scale load tests (e.g., Castilla, et al., 1984; Shahawy and Issa, 1992), a
depth of embedment of 2D to 3D into the pile cap provides full fixity for most service load
conditions. Because the degree of pile-cap fixity is usually established empirically, there is no
difference in the approach taken for pile foundations designed by either ASD or LRFD.
9-25
Determine appropriate load factors and load combinations and calculate the
total factored load effects
Estimate the unfactored and factored axial resistance of a single pile based on
correlation with the results of CPTs
Estimate the unfactored and factored axial resistance of a single pile based on
the results of full-scale load tests
The unfactored vertical loads on the pile group shown in Figure 9-2 are:
DC = Dead load of structural components and non-structural attachments = 4600 kN
DW = Dead load of wearing surfaces and utilities = 3900 kN
LL = Vehicular live load = 3450 kN
9-26
Figure 9-2
Pile Group Loading
kN +
kN
For the Strength I and Service I Limit States, Eq. 9-6 is used to compute the factored load effect:
Q
i
9-27
Table 9-10
Load Factors
Limit State
Strength I
Service I
DW
DC
LL
The total factored load effects are then calculated as follows. For the Strength I Limit State:
Q
Q
Q
= i [DC DC + DW DW + LL LL]
[(
)(
kN) + (
)(
kN) + (
)(
kN)]
)(
kN) + (
)(
kN)]
kN
Q
Q
Q
i
[(
)(
kN) + (
kN
(Eq. 9-19)
(Eq. 9-20)
and
where:
Qult, Qp, Qs = Ultimate total, tip and side resistance (kN)
qp, qs = Unit tip and side resistance (kPa)
Ap, As =
Area of pile tip and side (m2)
9-28
Figure 9-3
Generalized Problem Geometry and SPT Design Envelope
Using Eq. 9-17, the ultimate axial resistance of a single pile is:
Qult = Qp + Qs =
Qult =
kN
kN +
kN
9-29
(Eq. 9-17)
From Eq. 9-18, the factored axial resistance of a single pile is:
QR = Qult = qp Qp +qs Qs
(Eq. 9-18)
From Table 9-5, the resistance factors for the SPT method are:
y
y
qp =
qs =
kN) +
kN) =
kN
Using Nottingham and Schmertmann (1975) (A10.7.3.4.3), estimate ultimate axial capacity of the
pile driven into the soil profile in Figures 9-4 and 9-5.
Qp = qp Ap = (11 300 kPa) (0.46 m/2)2 = 1880 kN
(Eq. 9-19)
The ultimate unit side resistance of the pile is estimated using the CPT sleeve resistance values from
Figure 9-4 and the following relationship:
N2
N1 Li
(A10.7.3.4.3c-1)
+
=
Qs K s
f si a si h i f si a si h i
8
D
i
i
=
i
=
1
Ni
For a steel pile with a length to diameter (z/D) ratio of 11.00 m/0.46 m = 23.9, the correction factor,
Ks for a mechanical cone (from Figure A10.7.3.4.3c-1) is 0.39. The incremental values of Qsi along
the pile are presented in Table 9-11.
9-30
Figure 9-4
Representative Pile and CPT Tip Resistance Profile Used in Example
9-31
Figure 9-5
Representative Pile and CPT Side Resistance Profile Used in Example
9-32
Table 9-11
Tabulation of Side Resistance Along Pile Length
Pile
Segment, I
1
2
3
4
5
6
7
8
Li
(m)
0.643
1.885
2.970
3.523
NA
NA
NA
NA
Di
(m)
0.46
0.46
0.46
0.46
NA
NA
NA
NA
fsi
(kPa)
345
173
249
403
403
441
403
307
asi
(m2/m)
1445
1445
1445
1445
1445
1445
1445
1445
Qsi
(kN)
4
7
9
7
11
25
60
55
hi
(m)
1.105
1.380
0.790
0.315
0.480
1.000
2.650
3.190
TOTAL
Qs:
178
kN +
kN =
kN
(Eq. 9-17)
The factored axial resistance of a single pile is determined from Eq. 9-18 as:
QR = Qult = qp Qp +qs Qs
(Eq. 9-18)
From Table 9-5, the resistance factors for the CPT method are:
y
y
qp =
qs =
kN) +
kN) =
kN
Step 5: Estimate Axial Capacity of Single Pile from Full-Scale Load Tests
From a full-scale axial load test on a pile at the site (Vesic, 1970), the ultimate tip and side resistance
of a pile driven to a depth of 12 m at the site are:
y
y
Qs = 1180 kN
Qp = 1900 kN
from which the ultimate axial resistance based on Eq. 9-17 is:
9-33
Qult = Qp + Qs =
kN +
kN =
kN
(Eq. 9-17)
The factored axial resistance of a single pile from Eq. 9-18 is:
QR = qp Qp + qs Qs
(Eq. 9-18)
From Table 9-5, the resistance factors for bearing resistance of a single pile based on a load test are
as follows:
qp =
qs =
kN) +
kN) =
kN
For ASTM A709M, Grade 250, Fy = 250 000 kPa, the allowable axial stress is:
y
kPa)
kPa)
kN
kN
9-34
For LRFD, the factored axial resistance, Pr, of an unfilled steel pipe pile from Eq. 9-21 and Eq. 9-22
is:
Pr = Pn = n As = Fy As
From Table 9-6, the resistance factors for axial compression loading of an unfilled steel pipe pile
are:
y
kN
kN
kPa)(0.0176 m2) =
Q
Q
; x
(Qult /FS)
Pall
where:
Qult/FS = Allowable geotechnical resistance (kN)
Pall =
Allowable structural resistance (kN)
From Table 9-1, the required factors of safety (FS) for axial load resistance based on various
predictive methods are:
y
y
Based on these safety factors and the results of Steps 1 through 6, complete Table 9-12 to determine
the number of piles need to resist the design load based on geotechnical and structural resistance by
9-35
ASD.
Table 9-12
Determination of Required Pile Group Size Based on ASD
Resistance
Geotechnical:
SPT Method
CPT Method
Static Load Test
Structural:
Severe Driving
Good Driving
FS
or
all
Design
Group
Load
Q
(kN)
3.5
3.5
2.0
11 950
11 950
11 950
0.25 Fy
0.33 Fy
11 950
11 950
Single Pile
Ultimate
Resistance
Qult
(kN)
Single Pile
Allowable
Resistance
Qult/FS or Pall
(kN)
Required
Number of
Piles,
x
--C
For LRFD, the number of piles, x, required to support the factored design load, Q, can be computed
based on Eqs. 9-7 and 9-8 from the following:
i i Qi
QR
; x
i i Qi
Pr
where:
i i Qi = Sum of the factored axial loads (kN)
QR = Factored geotechnical resistance (kN)
Pr = Factored structural resistance (kN)
Based on the results of Steps 1 through 6, complete Table 9-13 to determine the number of piles
need to resist the design load based on geotechnical and structural resistance by LRFD.
9-36
Table 9-13
Determination of Required Pile Group Size Based on LRFD
Single Pile Resistance
Factored(1)
Resistance Design Group
Factored
Ultimate
Factor
Resistance
Load
QR or Pr
Q
ult
i i Qi
(kN)
(kN)
(kN)
Geotechnical:
SPT Method
0.45
17 638
CPT Method
0.55
17 638
Static Load Test
0.80
17 638
Structural:
Severe Driving
0.35
17 638
Good Driving
0.45
17 638
(1)
For the Strength I Limit State
Required
Number
of Piles,
x
Summary
This example illustrates the design of axially loaded pile groups by LRFD and compares the results
to design by ASD. As shown on Tables 9-12 and 9-13:
y
The geotechnical resistance controls the design for static methods of axial
capacity prediction and the structural resistance controls the design based
on static load testing for both the ASD and LRFD procedures
Comparison with Table 9-12 shows that the LRFD procedures based on the
reliability-based calibration (i.e., SPT and CPT) are somewhat less
conservative than the ASD procedure for the conditions analyzed
LRFD design procedures based on direct calibration with ASD (i.e., static
load test and structural resistance) provide essentially identical results with
ASD (minor differences occur due to differences in the actual average load
factor and the load factor assumed in the resistance factor calibration)
As indicated in Section 9.2.2.2, the AASHTO LRFD Specifications do not yet satisfactorily
incorporate the impact of field capacity verification (e.g., PDA or load tests) on the reliability of
pile installations for which design capacities are initially predicted using static design methods.
Therefore, this example assumes that the piles are designed on the basis of static methods without
field capacity verification.
9-37
The development of a resistance factor for evaluation of the axial geotechnical resistance of a pile by
Nordlund=s method was presented in Chapter 7, Sections 7.3 and 7.4. This exercise compares LRFD
and ASD for the Nordlund pile capacity prediction method for the Classroom Example in Section
9.4, and compares the results with the methods used in the Classroom Example. Nordlund's method
for estimation of pile side and tip resistance in cohesionless soil (GRL, 1996) utilizes the following
equation:
d=D
Qult =
CF pd
d=0
sin ( + )
Cd d + t N q A t pt
cos
where:
d=
D=
K =
CF =
pd =
=
=
f =
Cd =
d =
t =
Nq =
At =
pt =
Depth (m)
Embedded pile length (m)
Coefficient of lateral earth pressure at depth d (dim)
Correction factor for K when f (dim)
Effective overburden pressure at the center of depth increment d (kPa)
Friction angle between pile and soil (deg)
Angle of taper from vertical (deg)
Soil friction angle estimated from SPT (deg)
Pile perimeter at depth d (m)
Length of pile segment (m)
Dimensionless factor (dependent on pile depth-width relationship) (dim)
Bearing capacity factor (dim)
Pile toe area (m2)
Effective overburden pressure at the pile toe (kPa)
To focus solely on the differences between LRFD and ASD, the details of Nordlund's method as
applied to this problem are not presented herein. The results of Nordlund's method are tabulated
below. For the steel pipe piles and soil conditions in the Example in Section 9.4:
=
t =
Nq =
At =
Pt =
0
0.65
55
0.166 m2
139.7 kPa
and values of other parameters in the Nordlund equation are tabulated below:
9-38
Depth
Interval (m)
1-2
2-7
7 11.25
11.25 - 12
pd
(kPa)
28.4
63.3
110.4
139.7
K
(dim)
0.92
0.91
0.90
0.90
f
(deg)
29
31
33
34
CF
(dim)
0.92
0.91
0.90
0.90
(deg)
23
24
26
27
Cd
(m)
1.45
1.45
1.45
1.45
d
(m)
1
5
4.25
0.75
For ASD:
x
y
Q
(Qult /FS)
For LRFD:
Q
i
QR
9-39
Resistance
Geotechnical:
SPT Method
CPT Method
Static Load Test
Nordlund Method
Structural:
Severe Driving
Good Driving
FS
or
all
Design
Group
Load
Q
(kN)
Single Pile
Ultimate
Resistance
Qult
(kN)
Single Pile
Allowable
Resistance
Qult/FS or Pall
(kN)
Required
Number of
Piles,
x
3.5
3.5
2.0
3.5
11 950
11 950
11 950
11 950
1900
2057
3080
1385
542.9
587.7
1540
_______
22
21
8
______
0.25 Fy
0.33 Fy
11 950
11 950
-----
1098
1450
11
9
Resistance
Resistance
Factor
Geotechnical:
0.45
SPT Method
0.55
CPT Method
0.80
Static Load Test
0.60
Nordlund Method
Structural:
0.35
0.45
Severe Driving
Good Driving
(1)
For the Strength I Limit State
Factored(1)
Design Load
i i Qi
(kN)
17 638
17 638
17 638
17 638
17 638
17 638
9-40
855
1131
2464
______
1538
1977
Required
Number
of Piles,
x
21
16
8
____
12
9
The pile group is subjected to biaxial bending and bi-directional lateral forces
in addition to a vertical load
Objective: To illustrate the procedure for driven pile design using LRFD for an end-bearing pile
group subjected to axial, lateral and moment loading.
Approach: To perform the evaluation, the following steps are taken:
y
Determine the factored structural moment resistances about the strong and
weak axes for a single pile
Estimate the number of piles required to support the factored loads and
determine the geometric arrangement of piles in the group
Determine the factored axial loads and bending moments, about both axes,
for each pile in the pile group
For the most heavily loaded piles, compare the factored axial and bending
resistances with the factored axial, bending and lateral loads
Determine whether the axial and lateral displacements of the pile group are
acceptable
Determine whether any load case could result in uplift forces on any of the
piles
If for any factored load combination the pile group is incapable of adequately
supporting the factored loading, either due to inadequate factored resistance
or excessive deformation, increase the number of piles and repeat the
9-41
LOAD
CASE
STR I
A STR III
SER I
STR I
B
SER I
STR I
C
STR III
STR I
D STR V
SER I
STR I
E
SER I
F STR I
MAXIMUM LOADS
MINIMUM LOADS
Moments
Horiz. Load
Moment
Horiz. Load
Axial
Axial
Long.
Long. Trans. Load Long. Trans.
Trans. Load Long. Trans.
(kN(kN-m) (kN-m) (kN) (kN) (kN)
(kN-m) (kN) (kN) (kN)
m)
4772
5632
12 644 317
77
----------1046
3723
9617 74
270
952
3723
6585 74
270
2912
3803
8962 220
115
----------4772
5632
12 644 317
77
---------3269
3454
8960 261
85
----------5427
4432
11 967 317
77
5262 4432
8935 317
77
1046
3723
9617 74
270
952
3723
6585 74
270
5427
4432
11 967 317
77
5262 4432
8935 317
77
----------4861 3899
8497 324
112
3614
2801
8594 261
85
--------------------2952 6203
8575 201
49
2060
4114
8398 157
100
--------------------2952 6203
8575 201
49
y
y
y
Pu 8.0 M ux M uy
+
+
1.0
Pri 9.0 M rx M ry
(A6.9.2.2-2)
where:
Pu =
Pri =
Mux, Muy =
Mrx, Mry =
and
Pri =
Mrx =
Mry =
ci =
fi =
ci Pn = 0.60 Fy As
fi Mnx = 0.85 Fy Sx
fi Mnx = 0.85 Fy Sy
Axial resistance factor for combined axial load and flexure (dim)
Flexural resistance factor for combined axial load and flexure (dim)
[The interactive resistance factors, ci and fi, used in this problem were selected based on strength
reductions for pile material properties and unanticipated load eccentricity as recommended by
Davisson, et al. (1983), and differ somewhat from values prescribed in AASHTO (1997a) and AISC
(1986) as follows:
AISC:
ci = 0.85; fi = 0.90
AASHTO LRFD:
ci = 0.70; fi = 0.78
This problem:
ci = 0.60; fi = 0.85]
At some depth, the moments will decrease to a negligible level, and only axial forces will be present.
Each pile must be capable of resisting the compressive stresses due to the axial loading. In this
9-43
c Pn
0.35 for hard driving (Table 9-6)
0.45 for easy driving (Table 9-6)
Step 4: Establish the Factored Structural Axial Resistance for a Single Pile
For an HP 310 110 section, As = 0.0141 m2, and Fy = 250 000 kPa (Grade 250). Because the piles
will be driven a relatively short distance through a medium stiff clay, little driving damage should
occur. Therefore, assume a = 0.45.
The pile cap plan dimensions should be at least as large as the pier, which
measures 3.5 m 1.5 m.
As the bending moments and lateral forces for the subject pier can act in
either the positive or negative directions, the pile cap should be made
concentric to the pier. (Note: For a structure, such as an abutment, for which
the direction of overturning moments and horizontal loads cannot be
reversed, piles are usually offset to resist the imposed load eccentricity).
The pile bending stresses will accrue primarily due to the lateral forces. The
9-44
maximum lateral longitudinal and transverse forces, 317 and 270 kN,
respectively, are similar. Therefore, orient one-half, i.e. 8, of the piles with
their strong axes parallel to the longitudinal direction, and the other 8 parallel
to the transverse direction.
The pile layout geometry shown in Figure 9-6 satisfies the above-discussed considerations.
Step 7: Determine Factored Axial Loads and Bending Moments for Each Pile
The factored loads, presented in Step 1, the pile properties, presented in Steps 4 and 5, and the pile
layout geometry, presented in Step 6, were input into the computer program GROUP. The results of
this analysis are presented below. The pile axial loads (in kN) are shown in Figure 9-7.
The average pile axial load due to the vertical load is:
= 12 644 kN/16 piles
= 790 kN/pile
Half of the piles (i.e. 8) shown in Figure 9-7 carry less than the average load, and 8 piles carry more.
The axial load imbalance is caused primarily by the biaxial overturning moments.
The maximum pile bending moments about both the strong and weak axes (in kN-m) are shown in
Figures 9-8 and 9-9, respectively.
The pile bending moments are primarily caused by the lateral forces. The analysis indicates that
piles oriented with their strong axes parallel to the direction of applied lateral load carry a high
proportion of the resulting bending moment.
In performing the GROUP analysis, the degree of pile-cap fixity must be specified (i.e., fixed,
pinned or some intermediate level). The degree of fixity is related to the bending moment at the pile
to cap interface, with the degree of fixity decreasing with increasing moment. The results presented
in Figures 9-7, 9-8 and 9-9 were based on an assumed fixed condition. The results of the analysis
indicate that the pile bending moments are a relatively small fraction of the bending capacity of the
piles. Therefore, the assumed fixed condition is reasonable.
9-45
Figure 9-6
Pile Cap and Pile Layout
9-46
Figure 9-7
Axial Load, Pu, Summary
9-47
Figure 9-8
Strong Axis Maximum Bending Moment, Mux, Summary
9-48
Figure 9-9
Weak Axis Maximum Bending Moment, Muy, Summary
Step 8: Determine Structural Resistance Adequacy
Piles A4, A5 and B5 are the most heavily loaded. Their ability to resist the loads and moments
presented in Figures 9-7, 9-8 and 9-9 will be evaluated by both the interactive axial and bending
stress equation,
Pu 8.0 M ux M uy
+
+
1.0
Pri 9.0 M rx M ry
(A6.9.2.2-2)
9-49
Pu
1 .0
Pra
A4:
= 0.578 +
2115 9.0 327 106
9.0
8.0
1357 8.0 5.1 15.6
+
+
(0.016 + 0.147) = 0.79 < 1.0 O . K
= 0.642 +
2115 9.0 327 106
9.0
1208 8.0 27.7 2.2
8.0
B5 :
+
+
(0.085 + 0.021) = 0.67 < 1.0 O . K
= 0.571 +
2115 9.0 327 106
9.0
A5:
For the axial load only equation, only pile A5 needs to be evaluated since it carries the highest axial
load.
1357
= 0.86 < 1.0 O . K
1586
The above-shown analysis indicates that the selected pile group is capable of resisting the Cases
A/B, Strength I factored loads.
Step 9: Determine Whether Deformations Are Acceptable
The Service I Case loads are to be used to determine whether deformations are acceptable.
However, a separate deformation analysis need not be conducted if the deformations resulting from
the critical load application are small, as the Service I case deformations would be smaller.
The GROUP analysis for the critical load case, presented in Step 7, yielded deformation data. The
pile cap settlements at each of the pile locations (in mm) is shown in Figure 9-10.
9-50
Figure 9-10
Pile Settlements
The lateral displacements in the longitudinal and transverse directions were determined by GROUP
to be 1.8 and 0.5 mm, respectively.
These displacements, even though they are based on the critical loading, are small. Therefore, there
is no need to evaluate deformations due to the smaller Service I loads.
Step 10: Examine Whether Any Other Load Conditions Are More Critical
In Step 1, Load Cases A/B, Strength 1 were identified as probably being most critical. The
succeeding analyses were conducted using these loads. Step 1, however, also pointed out that
possibly Load Cases C/D, Strength 1 could be more critical. The latter loading was subsequently
analyzed via GROUP and was found to be non-critical.
Step 11: Check for Pile Uplift
The load data need to be evaluated to determine whether there are any load cases which result in
uplift on any of the piles. A high tensile pile force could overcome the piles' uplift resistance, and
cause load to be transferred to other piles, possibly causing overstressing. Pile uplift would occur if
the uplift effect of the sum of the longitudinal and transverse moments could overcome the axial
9-51
compressive load.
All of the load cases were checked to determine whether any tensile pile forces could exist. This
evaluation showed that, for all the load cases, all the piles were in compression.
Step 12: Examine Whether Design Is Overconservative
In Step 8, it was determined that Pile A5 was the most heavily stressed. The axial load only criterion
was the more critical. The analysis showed that Pile A5 was loaded to 86.2 percent of its capacity.
Therefore, some excess capacity exists. A reduced pile quantity can be estimated as follows:
9-52
Negative skin friction is similar to a pile's resistance to settlement, except that it acts on the pile in
the opposite direction. It can therefore be estimated using the or methods.
method (A10.7.3.3.2a)
qs = Su
where:
Su = 37.5 kPa
= adhesion . 1 for clays with Su = 37.5 kPa
Therefore:
qs = 1.0 x 37.5 kPa = 37.5 kPa
method (A10.7.3.3.2b)
qs = v
9-53
(Fig. A10.7.3.3.2a-1)
1.7 m + 11.7 m
x 8.5 kN/ m3 = 57.0 kN/ m2 = 57.0 kPa
v _ =
2
where:
= 0.3 for NC clays
(Fig. A10.7.3.3.2b-1)
Therefore:
qs = (0.3)(57.0 kPa) = 17.1 kPa
qs =
Because the factored downdrag force per pile is 605 kN, the factored loads exceed the factored
resistance by 376 kN. Therefore, the designed pile group has insufficient factored resistance to carry
the factored loads and must be redesigned.
9-54
Prior to proceeding with redesign, the sources of the axial load acting on Pile A5 must be quantified.
The total factored axial force is equal to the force determined in Design Example 1, 1357 kN, plus
the 605 kN factored downdrag force, or 1962 kN.
This total factored axial load accrues due to the following loads:
y
y
y
y
9-55
vertical force =
(4772 kN- m) (2 m)
= 251 kN
38 m2
(5632 kN- m) (2 m)
= 296 kN
38 m2
The sum effect of the two overturning moments is a vertical force of 547 kN for pile A5.
Lateral Force Component
The vertical force caused by the lateral forces can be determined by subtracting the vertical force,
790 kN, and the moment force, 547 kN, effects from the total vertical force of 1357 kN as
determined by Step 7 in Design Example 1. The resulting calculation shows the lateral forces to
cause a relatively minor vertical load of 20 kN on pile A5.
In summary, the total factored vertical load acting on Pile A5 accrues from the following
components:
Vertical Load
Overturning Moments
Lateral Force
Soil Downdrag
TOTAL
= 790 kN
= 547 kN
= 20 kN
= 605 kN
=1962 kN
Pu 1.0
Pra
from Step 3, Section 9.6
1962 kN
1.0
F y As
where:
c = 0.45; and
Fy = 250 000 kPa
Therefore,
9-56
1962 kN
Pu =
1.0
Pra (112 500 kPa)(As)
and
As 0.0175 m2
A HP 380 152 section, having a cross-sectional area of 0.0194 m2, would be adequate. (Note: As
this member has a 19 percent larger perimeter than the 310 x 110 pile for which downdrag was
calculated, the pile downdrag would be increased by a similar proportion, i.e. 0.19 605 kN, or 115
kN. This increase in factored load would require a member with a cross-sectional area of 0.0185 m2,
which the HP 380 152 member satisfies.)
Step 6 - Option B: Increase the Pile Spacing
Increasing the pile spacing decreases the vertical force component due to overturning moment,
shown to be equal to 547 kN in Step 4, by increasing d 2 . In Step 3, it was shown that Pile A5
would be overstressed by 376 kN upon the application of the downdrag force. If the vertical load
due to the overturning moment were to be reduced by 376 kN down to 171 kN, 16 HP 310 110
piles would have adequate factored resistance, although a greater pile spacing, and therefore larger
pile cap, would be required.
As shown in Figure 9-6, the 16 piles are arranged on a 1 m 1 m grid. Increasing the grid size
decreases the Pile A5 factored vertical load due to overturning moments as shown in Table 9-15
below:
Table 9-15
Pile Spacing Effect
Grid Size
(m m)
1.0 1.0
1.5 1.5
2.0 2.0
2.5 2.5
3.0 3.0
38.0
85.5
168
238
342
Factored Vertical
Load
(kN)
547
365
248
219
182
(m2)
Table 9-15 shows that increasing pile spacing alone is not a feasible option. Even at the 3 m 3 m
grid size, which would require a very large 13.2 m 13.2 m pile cap, the factored vertical load due
to overturning moments is insufficiently reduced.
9-57
The vertical pier load will be spread out over more members
The vertical load due to the overturning moments will be decreased due to an
increase in dx2 and dy2
As shown in Figure 9-6, the original design consisted of H piles occupying 16 of the available 25
locations on the 5 x 5 grid. Now, add an additional 9 piles, so that a pile occupies each grid
intersection.
The factored vertical load imposed on Pile A5 due to the vertical pier load will now be:
12 644 kN
kN
= 506
25 piles
pile
The factored vertical load due to overturning moments is calculated as follows:
d 2x = 10(2 m )2 + 10(1 m )2 = 50 m2
d 2y = d 2y = 50 m2
vertical load =
The factored soil downdrag force will remain unchanged at 605 kN. Assume that the small vertical
load due to the horizontal forces remains unchanged at 20 kN. In summary, the following vertical
forces act on Pile A5.
Vertical load effect
Overturning moment effect
Lateral force effect
Soil downdrag effect
TOTAL
= 506 kN
= 416 kN
= 20 kN
= 605 kN
= 1547 kN
Therefore,
Pu 1.0
Pra
1586 kN
1.0
1574 kN
0.98 < 1.0 O . K .
9-58
where:
N 0 = Factor ranging from 0.01 to 0.05
po = Mean horizontal effective stress
1.7 m x 11.7 m
qs = 0.05 x 0.5
x 8.5 kN/ m3 = 1.4 kPa
2
Note that this is a significant reduction from the uncoated pile qs value of 27.3 kPa calculated in Step
1. Applying this stress to the 12.32 m2 pile contact area results in an unfactored downdrag load of
17 kN. Applying a load factor of 1.8 results in a factored downdrag load equal to 31 kN.
In Step 3, it was shown that the 16-member pile group designed in Design Example 1, utilizing HP
310 110 members, had some excess load resistance capacity prior to considering soil downdrag.
Pile A5 was most heavily loaded at 1357 kN compared to its factored resistance of 1586 kN.
Adding the factored 31 kN downdrag force acting on the coated pile increases the factored load
transmitted to Pile A5 to 1388 kN, still somewhat less than its factored resistance.
Step 8 - Summarize Redesign Options
In summary, four options were analyzed. Three were found to be feasible as follows:
9-59
9-60
CHAPTER 10
DRILLED SHAFT DESIGN
10.1 Introduction
For both ASD and LRFD, the design of drilled shaft foundations requires consideration of
geotechnical and structural capacity and deformation limits. As with pile design, the design
processes require both establishment of criteria for acceptable stress and deformation levels, and
comparison of these criteria with stress and deformation levels estimated from the design. This
chapter:
With few exceptions, the procedure for design of drilled shaft foundations using LRFD (A10.8) is
identical to the procedure followed using ASD. Generally, the ultimate bearing capacity and
settlement are checked for axially-loaded shafts. As with driven piles, design of laterally-loaded
shafts may be governed either by lateral capacity or deflection criteria. Again, however,
mobilization of the ultimate lateral capacity of the soil requires such large displacements that lateral
capacity does not represent a realistic basis for laterally-loaded shaft design. The following sections
summarize the general design processes for drilled shaft foundation design using the ASD and
LRFD approaches.
10.2.1 ASD Summary
Existing practice for design of drilled shaft foundations follows the ASD approach. For ASD
(AASHTO, 1997b), all uncertainty in the variation of applied loads transferred to the shaft(s) and the
ultimate geotechnical capacity of the soil and rock to support the load are incorporated in a factor of
safety, FS. As a result, loads used for design, Q, consist of those actual forces estimated to be
applied directly to the structure. In LRFD terminology, this is equivalent to applying a load factor of
1.0 to the estimated forces. Relative to axial loading of a shaft, the ultimate axial geotechnical load
capacity, Qult, is estimated using any number of available methods. Then the suitability of the design
is evaluated by determining the allowable axial design load, Qall, using:
10-1
Q = Q all =
R n Q ult
=
FS
FS
(Eq. 10-1)
where:
Q=
Qall =
Rn =
FS =
The required FS is generally specified by the governing agency, and may be constant or variable.
The FS used for AASHTO ASD for the axial geotechnical capacity of a single shaft is presented in
Table 10-1. If Qult is developed using information from a site-specific subsurface exploration based
on a semi-empirical calculation, FS 2.50. However, if capacities are estimated using a static field
load test, FS = 2.0 could be used to reflect the greater confidence level achieved by the improved
method for estimating and confirming the axial capacity of a shaft. As the table shows, ASD for
shaft design assumes that a reasonable level of quality control will be exercised during the design
and construction phases. If normal levels of field construction quality control cannot be assured
however, AASHTO ASD requires the use of higher (but unspecified) safety factors.
Table 10-1
Factor of Safety on Ultimate Axial Geotechnical Capacity
(AASHTO, 1997b)
Required Minimum Factor of
Basis for Design
Safety (FS)(1)
Semi-Empirical
2.5
Calculation
Load Test
2.0
(1)
FHWA (Reese and O'Neill, 1988 and Cheney and Chassie, 1993) suggests a typical range for safety
factors of 2.0 to 3.0 (where good-quality geotechnical data are available), and provide guidance for a
variable FS considering the level of construction control and structure design life, as shown in Table
10-2.
10-2
Table 10-2
Factors of Safety on Ultimate Geotechnical Capacity
Based on Design Life and Level of Construction Control
(Reese and O'Neill, 1988)
Required Minimum Factor of Safety (FS)(1)
Design Life (Type of
Poor Quality
Normal Quality
Good Quality
Structure)
Control
Control
Control
200 to 500 years (large
bridges & monumental
3.5
2.3
1.7
structures)
75 to 100 years (typical rail &
road bridges & large
2.8
1.9
1.5
buildings)
25 to 50 years (industrial
2.3
1.7
1.4
buildings)
(1)
Assumes good-quality geotechnical information and reliable model.
The structural design of drilled shafts is described in detail in Reese and Allen (1977) and is covered
briefly by Reese and O'Neill (1988). Although geotechnical design of drilled shafts may follow
ASD procedures, structural design is currently performed in accordance with the American Concrete
Institute (ACI) code which uses the LRFD method.
In addition to axial geotechnical and structural capacity evaluation, the design of drilled shafts by
ASD requires evaluations of shaft deflections and comparisons with deformation criteria using the
following:
i n
(Eq. 10-2)
where:
i = Estimated displacement (mm)
n = Tolerable displacement established by designer (mm)
Tolerable movement criteria are usually a function of the type of structure, and for bridges depend
primarily on the span length and whether the superstructure has fixed- or simply-supported spans.
The design of laterally-loaded shafts by ASD is often governed by criteria for deflection under actual
applied loads, although design of large diameter shafts and shaft groups containing batter shafts may
be controlled by the axial and/or bending capacity of the shaft(s).
10.2.2 LRFD Summary
Whereas ASD considers all uncertainty in the applied loads and ultimate geotechnical capacity in
factors of safety, LRFD separates the variability of these design components by applying load and
resistance factors to the load and material capacity, respectively. When properly developed and
applied, the LRFD approach provides a generally consistent level of safety for the design of all
structure components. Thus, the probability that any one structure component will fail or perform
10-3
unacceptably is no different than any other component. As described in Section 4.3, the resistance
and deformation of supporting soil and rock materials and structure components must satisfy the
LRFD equations below. For the Strength Limit States:
i i Qi R n = R r
i i i n
(Eq. 10-4)
where:
i =
i =
Qi =
=
Rn =
Rr =
i =
n =
Relative to axial loading of a shaft, the suitability of a shaft with respect to the geotechnical and
structural resistance can be obtained using Eq. 10-3, rewritten as:
i i Qi Qult = QR
(Eq. 10-5)
(Eq. 10-6)
and
i
Pn = Pr
where:
i i Q i
=
Qult =
QR =
Pn =
Pr =
The load factors and load factor combinations used for design were presented in Chapter 4. In
general, values of (i > 1.0 are used to evaluate ultimate ground or structure capacity at the Strength
Limit States, whereas the deformation performance of structures is evaluated at the Service I Limit
State using (i = 1.0 (or (i = 0.3 for wind loads). In ASD (AASHTO, 1997b), values of (i = 1.0 (or
10-4
(i = 0.3 for wind loads) are used to evaluate structures for both strength (allowable stress) and
serviceability (deflection). Therefore, analysis of shaft deformations (e.g., settlement or lateral
displacement) by LRFD and ASD are identical.
When using Eq. 10-3 for drilled shaft foundation design at the Strength Limit States, the following
values of 0 can normally be used:
0D =0R = 1.00
0I = 1.05 for structures deemed operationally important, 1.00 for typical
structures and 0.95 for relatively less important structures.
Determination of the operational importance of a structure (such as a bridge) is made by the facility
Owner as described in Chapter 4. The appropriate value of 0I is then applied throughout the
superstructure design by the structural engineer. The value of 0I selected by the superstructure
designer should then be applied in the foundation design. For the purpose of this chapter, the value
of 0I is assumed equal to 1.0.
When using Eq. 10-4 to evaluate a drilled shaft foundation at any Service Limit State, 0D, 0R, and 0I
= 1.0.
Values of load factor and load factor combinations for each applicable limit state must be developed
using the guidelines described in Chapters 3 and 4 and in Section 10.2.2.1, and loads should be
developed as described in Chapter 4. The ultimate resistance, Rn, should be determined for each
type of resistance (e.g., axial, side, or group resistance) as described in Section 10.3. Values of
1.0 are applied when evaluating drilled shaft foundation resistance for any strength limit state using
Eq. 10-3. Currently, the value of = 1.0 is applied when evaluating a drilled shaft foundation for
any service limit state using Eq. 10-4. Selection and modification of resistance factors, , are
described in Sections 10.2.2.2 through 10.2.2.4.
10.2.2.1 Limit States (A10.5)
In general, the design of drilled shaft foundations using LRFD parallels that of pile foundations, and
requires evaluation of shaft suitability at various Strength Limit States and the Service I Limit State.
The selection of a Strength Limit State(s) depends on the type of applied loading (e.g., Strength I for
design vehicle loading without wind or Strength II for permit vehicle loading). The design
considerations which must be evaluated for shafts designed at the Strength and Service I Limit States
are summarized in Table 10-3. As conditions warrant, it may also be necessary to evaluate shaft
performance at other limit states (e.g., Extreme Event I for loading from earthquakes or Extreme
Event II for vessel impact and ice loading). Methods of evaluating shafts at the various limit states
are described in Section 10.3.
10-5
Table 10-3
Limit States for Design of Drilled Shaft Foundations
Performance
Limit
Strength Limit
State(s)
Service I Limit
State
U
U
U
U
U
U
Observing the variation in reliability levels with different span lengths, deadto live-load ratios, foundation geometry and methods used to estimate
ultimate resistance and load combinations
Selecting a target reliability, $T, based on the margin of safety used for ASD
This process was used to develop resistance factors for the design of drilled shaft foundations. As an
example of this process, Table 10-4 summarizes the results of analyses performed to develop
values for estimating the axial resistance of shafts. For these analyses, the variables included:
A dead to live load ratio of 3.69 which corresponds to a bridge with a span
length of 75 m
10-6
The Horvath and Kenney (1979), Carter and Kulhawy (1987) and Reese and
O'Neill (1988) methods for estimating ultimate shaft resistance, Qs, as
described in Section 10.3.2.1
Selected values of for the Reese & O'Neill and Horvath & Kenney methods
are less than the average value of , whereas the selected value of for the
Carter & Kulhawy method is greater than the average value
Table 10-4
Resistance Factors for Drilled Shafts for Estimating the Ultimate Axial Shaft Capacity
Using Reliability-Based Calibration
(modified Barker, et al., 1991b)
Value by Method of Axial Shaft Capacity Estimation
Shaft
Length
(m)
2.5
10
2.5
0.72
0.73
0.56
30
2.5
0.80
---
---
3.0
---
0.56
0.37
10
3.0
0.72
0.59
0.43
30
3.0
0.71
---
---
Average
0.74
0.65
0.46
Selected
0.65
0.65
0.55
The relative factor values for shafts in rock were selected for use in the LRFD Specification based
on the limited number of reported cases where the methods had been used in conjunction with
performance testing, and the engineering judgment of the code developers. Values of for the
geotechnical design of drilled shaft foundations using these methods for axial capacity and for other
design considerations are presented in Table 10-5.
As indicated in Table 10-5, resistance factors have not been developed for drilled shafts in
cohesionless soils due to a lack of field load test data. Shaft side resistance in cohesionless soils is
highly dependent upon installation practice and quality control. Because large displacements are
10-7
generally required to mobilize shaft base resistance in cohesionless soils, the base resistance is
generally limited on the basis of service (settlement) criteria. The LRFD Specification (AASHTO,
1997a) recommends estimation of shaft and base resistance in cohesionless soils using all applicable
methods, and selection of a factored capacity based on judgment and experience with similar
conditions (A10.8.3.4.3).
The value of for structural design of axially-loaded, reinforced concrete drilled shafts (A5.5.4.2.1)
is shown in Table 10-6. The calculation of the ultimate structural capacities of drilled shafts is
described in Section 10.3.2.2.
Table 10-5 (A10.5.5-3)
Resistance Factors for Geotechnical Strength Limit States
for Axially Loaded Drilled Shafts
(AASHTO, 1997a)
Method/Soil/Condition
Resistance (1)
Factor
0.65
0.55
Various
---(2)
Various
---(2)
0.55
0.65
0.50
CGS (1992)
Pressuremeter Method (CGS, 1992)
Load Test
Clay
Clay
0.50
0.80
0.65
"-method
Belled Shafts (Reese & O'Neill, 1988)
Various
0.55
0.50
0.45
0.55
---(2)
0.80
Group Uplift
Sand
0.55
Resistance
Clay
0.55
(1)
Refer to Section 10.3 for description of design procedures for which N factors have been calibrated.
(2)
factors not been developed for drilled shafts in cohesionless soils due to a lack of adequate field data.
10-8
Table 10-6
Resistance Factor for Structural Design of Axially-Loaded Drilled Shafts
(AASHTO, 1997a)
Shaft Type
Resistance Factor
Reinforced Concrete with Spiral Reinforcement
0.75
0.75
(Eq. 10-7)
where:
= Average load factor (assumed = 1.45)
= Resistance factor from Table 10-5
The ASD factor of safety (FSASD) from Table 10-1 for each category is also presented in Table 10-7.
The values of FSLRFD in Table 10-7 were determined assuming = 1.45. Actually, (i can range
from 1.25 for structure component loads to about 1.75 for live loads so that FSLRFD could vary from
the value shown in the table depending on the relative proportion of live to dead load for a particular
structure. Assuming that = 1.45 represents a reasonable approximation, FSLRFD . FSASD
depending on the method of capacity estimation. Thus, while general agreement exists between
LRFD and ASD, the comparison depends on the load factor and method of capacity analysis.
Therefore, while drilled shaft designs performed using LRFD will be comparable to those using
ASD, a precise or very close approximation between the two should not necessarily be expected.
10.2.2.4 Modification of Resistance Factors
As stated in Section 10.2.2.2, each of the LRFD resistance factors for geotechnical design of shafts
in Table 10-5 was developed using either a reliability-based calibration procedure or direct
calibration with ASD, tempered with engineering judgment and experience. Application of these
resistance factors in conjunction with = 1.45 results in an "equivalent" factor of safety ranging
from 1.8 to 3.2, depending on the soil conditions, loading conditions and method of geotechnical
capacity prediction, as described in Section 10.2.2.3.
10-9
Table 10-7
Comparison of ASD Factor of Safety
with LRFD Equivalent Factor of Safety
Geotechnical Strength Limit State for Axially-Loaded Shafts
LRFD Equivalent
Resistance Factor of Safety ASD Factor
/, for
of Safety(2)
Factor,
= 1.45(1)
Method/Soil/Condition
Side Resist. - Clay
"-method
0.65
2.2
2.5
0.55
2.6
2.5
Various
---(3)
---(3)
---(3)
Various
---(3)
---(3)
---(3)
0.55
0.65
0.50
0.50
2.6
2.2
2.9
2.9
2.5
2.5
2.5
2.5
0.80
1.8
2.0
0.65
2.2
2.5
"-method
Belled Shafts
Various
0.55
0.50
2.6
2.9
2.5
2.5
---(3)
---(3)
---(3)
0.45
0.55
3.2
2.6
2.5
2.5
0.80
1.8
2.0
0.55
2.6
NA(4)
An average load factor, = 1.45, is assumed for a typical FSLRFD. (2) From AASHTO ASD, a range of
factors of safety from 1.4 to 3.5 is suggested by FHWA, depending on structure type, design life and the
degree of "construction control exercised. A FS of 2.5 is consistent with FHWA guidelines for a typical
road bridge having a design life of 50 to 100 years and constructed using "poor" quality control, and is
chosen for comparison purposes only herein. (3) See Table 10-5. (4) The procedures used in LRFD are
not equivalent to those in ASD. Therefore, the equivalent factors of safety for these methods cannot be
directly compared.
In ASD, the designer or owner might decide to increase or decrease required factors of safety in
consideration of a number of factors, such as:
The likelihood that the design loading conditions will be realized (often
dictated by the structure design life).
When using LRFD, similar flexibility to vary the required level of safety should also be available.
Additionally, whereas the same factor of safety is generally used in ASD regardless of the source of
loading, the equivalent factor of safety in LRFD (defined by Eq. 10-7) varies for a given resistance
factor depending on the source of loading. As stated previously, an average load factor of 1.45 was
used to identify the equivalent safety factors in Table 10-7. Although an average load factor of 1.45
is reasonable for a typical bridge abutment or pier , or a retaining wall, the average load factor could
theoretically range from a low of 1.25 (i.e., all dead load) to nearly 1.75 (i.e., extremely high live
load).
To modify the resistance factors for geotechnical design of shafts to account for average load factors
other than 1.45 and equivalent factors of safety other than those identified in Table 10-7, the
following equation may be used:
FS
m = T T
FS D
T
D
(Eq. 10-8)
where:
m =
T =
FST =
FSD =
T =
D =
Modifying resistance factors may seem reasonable, but such modification may not be consistent with
the goal of LRFD to achieve near equal reliability against failure of structure components, unless the
factor of safety accurately models the reliability of the shaft capacity predictive method used. The
resistance factors may be more appropriately modified through application of the probabilistic
procedures described in Chapters 3 and 5 to achieve the desired level of reliability if sufficient data
are available.
10.2.3 Summarized Comparison of LRFD and ASD
As noted before, the process used to develop a drilled shaft foundation design using LRFD differs
very little from the process used for ASD. The similarity is illustrated in the parallel flow charts in
Figure 10-1. Specific differences between the methods and other important issues are highlighted in
the following section.
10-11
Figure 10-1
Generalized Flow Chart for Drilled Shaft Design
by LRFD and ASD
Other aspects of the drilled shaft design such as identifying special considerations (e.g., potential for
loss of lateral support or negative loading), developing the design foundation profile and
determining requirements for construction control (e.g., the load testing) are inherent aspects of the
design process required for both LRFD and ASD.
10-12
For these design factors, there is no difference between LRFD and ASD analysis procedures. The
following sections highlight differences between LRFD and ASD in the performance criteria and
application of design procedures.
10.3.1 Displacements and Tolerable Movement Criteria (A10.8.2)
The vertical and lateral displacement of drilled shafts must be evaluated for all applicable dead and
live load combinations, and compared with tolerable movement criteria. Because evaluations of
structure displacements by LRFD are made at the Service I Limit State where ( = 1.0 and = 1.0,
methods used to estimate settlement and lateral displacement by LRFD are identical to those used
for ASD. Consequently, axial compression (or tension) of the shaft section can be computed by
elastic methods and axial and lateral ground displacements can be determined by conventional
methods using empirical correlations with in-situ test results or measurements by in-situ or
laboratory test methods to estimate engineering soil properties. Lateral displacements need to be
evaluated if:
For drilled shaft groups in cohesionless soils or rock, settlements will generally occur immediately
as the group is loaded. Drilled shaft groups in cohesive soils, however may experience long-term
consolidation settlement as well as immediate elastic settlement. Elastic settlement generally
predominates in overconsolidated clays, whereas consolidation settlement generally predominates in
normally consolidated clays. The consolidation settlement of shaft groups in cohesive soil is
typically estimated using the equivalent footing procedure described in Chapter 9 for pile groups.
10-13
During the past decade, computer programs have become available for axial and lateral loaddeflection analysis of deep foundations based on the axial load-axial displacement (i.e., t-z) and
lateral load-lateral displacement (i.e., p-y) curve methods of analysis. The tolerable axial and lateral
movement of drilled shafts should be based on criteria developed by the structural engineer for the
superstructure, or by consideration of the effects of foundation movements on adjacent structures. In
some cases, the tolerable lateral movement is fixed at a small displacement (e.g., # 12 mm based on
observed acceptable performance and engineering judgment) to ensure acceptable structure
performance.
10.3.2 Axial Resistance
10.3.2.1 Geotechnical Resistance (A10.8.3)
Drilled shaft foundations must be designed to resist axial loads without structural failure of the shaft,
and without excessive deflection. Methods used for ASD to estimate the ultimate axial geotechnical
resistance of a single shaft can be used for LRFD. Therefore, the ultimate axial geotechnical
resistance of shafts subjected to axial loading, Rn, can be determined as:
R n = Q ult = Q p + Q s
(Eq. 10-9)
and the factored axial geotechnical resistance, QR, can be determined as:
R r = Q R = Q ult = qp Q p + qs Q s
(Eq. 10-10)
for which:
Qp = q p A p
(Eq. 10-11)
Qs = q s A s
(Eq. 10-12)
and
where:
qp, qs =
Qp, Qs =
qp, qs =
Ap, As =
Various procedures are available for estimating the ultimate axial geotechnical capacity of drilled
shafts in soil using semi-empirical methods and in-situ testing, and for shafts bearing on or in rock
using semi-empirical methods. When using the LRFD Specification, only those methods referred
to in Table 10-5 for which calibrated N-factors have been developed can be used without
developing other method-specific resistance factors. Calibrated design methods for shafts in
cohesive soil (clay) and rock, and uncalibrated design methods for shafts in cohesionless soil
10-14
10-15
Static Load Test - A method for estimating total load capacity based
on tests (ASTM, 1997) representative of shaft, load and subsurface
conditions expected for the prototype shafts (AASHTO, 1997a).
With exception of the methods for estimating the side resistance of shafts in cohesionless soils based
on Standard Penetration Resistance and for the Touma and Reese, Meyerhof and Reese and Wright
methods for estimating the tip resistance of shafts in cohesionless soil, the methods described above
are used by FHWA (Reese and O'Neill, 1988) for evaluating the axial geotechnical capacity of
drilled shafts. Other methods can be used for LRFD provided calibrated -factors are developed
using the methods described in Chapters 3 and 7.
10.3.2.2 Structural Resistance (A10.8.4)
In addition to meeting the requirements for adequate geotechnical resistance of shafts subjected to
axial and lateral loading, the axial and flexural resistance of the shaft section must also be adequate
for all potential loading cases. The ultimate structural capacity of a reinforced concrete shaft is based
on the yield strength, Fy, of the reinforcing steel and the unconfined compressive strength, fc', of the
concrete. To evaluate an axially-loaded shaft for structural resistance, the factored axial load, Pu, is
compared to the factored axial resistance, Pr, which is determined as follows:
Pr = Pn
(Eq. 10-13)
where:
=
Pn =
Pr =
The nominal structural resistance of a reinforced concrete drilled shaft subjected to an axial load
can be computed as follows (AASHTO, 1997a):
Pn = 0.85 0.85 f c A g A st + A st Fy
P n = 0 . 8 0 . 85 f c A g A st + A st F y
Pn =
fcr =
Ag =
Ast =
Fy =
10-16
The -factors used in Eq. 10-13 in combination with Eq. 10-14 and Eq. 10-15 and presented in Table
10-6 are generally consistent with the existing ACI Code and include the effects of manufacturing
variability and small accidental load eccentricity.
The current ACI code specifies load factors for dead and live loads of 1.4 and 1.7, respectively,
whereas the LRFD Specification specifies maximum dead load factors ranging from 1.25 to 1.50 and
maximum live load factors ranging from 1.35 to 1.75. Further, the ACI code specifies resistance
factors, , of 0.75 and 0.70 for spirally reinforced and tied shafts, respectively, whereas the LRFD
Specification specifies a value of 0.75 for both types of reinforcement. Based on a review of the
load factors in Chapter 4 and values described above, and as summarized in Table 10-8, the
level of safety provided for structural design of drilled shafts in the AASHTO LRFD
Specifications (1997a), as measured by an "equivalent safety factor," is typically about 10 to 15
percent less than that provided under the ACI code, except for high live to dead load ratios.
Shaft
Reinforcement
Type
Spiral
Tied
Table 10-8
Comparison of ACI and AASHTO LRFD
Load and Resistance Factors for Drilled Shafts
ACI LRFD
AASHTO LRFD
Resistance
Load
Equivalent Resistance
Load
Equivalent
Factor
Factor
Safety
Factor
Factor
Safety
Factor
Factor
/
/
1.4
1.87
1.25
1.67
0.75
0.75
1.7
2.27
1.75
2.33
1.4
2.0
1.25
1.67
0.70
0.75
1.7
2.43
1.75
2.33
10-17
methods of analysis incorporate p-y curves to model the nonlinear lateral load-deflection behavior
for each soil type along the length of the shaft. When using these methods at the LRFD Strength
Limit State, the model shaft or drilled shaft group is subjected to the factored lateral and axial loads
and factored moment, and the resulting factored axial and bending stresses are compared with the
factored axial and bending resistance of the shaft.
In general, the design of laterally-loaded shafts involves:
1.
Determine the maximum lateral ground line deflection at the Service Limit
State and the maximum moment at the Strength Limit State for an individual
shaft considering installation method for the selected shaft section.
2.
Repeat Step 1 if the lateral ground line deflection exceeds the tolerable
deformation or results in axial or combined stresses which exceed the
maximum factored axial resistance of the shaft or do not meet the appropriate
interaction criteria.
3.
The shaft section is acceptable for the design loads if neither the lateral
ground line deflection nor the stress criteria from Step 2 is exceeded.
In practice, the shafts in a group are tied together with a cap which is typically embedded below the
ground surface. In the analyses of the lateral load resistance of this type of system, the contribution
of passive resistance of the embedded portion of the cap is typically neglected. Neglecting passive
resistance of the embedded cap can provide a significant degree of additional conservatism at both
service and ultimate load levels.
FHWA promotes the use of vertical shafts to resist lateral loads. However, if the lateral loads acting
on a foundation are large, batter shafts can be used. Batter shafts transmit lateral loads
predominantly as an axial force into the soil, and should be avoided when:
Methods such as GROUP (Reese, et al., 1994) used for ASD to estimate the lateral resistance of a
drilled shaft group containing batter shafts can also be used for LRFD.
10.3.4 Other Considerations
10.3.4.1 Group Effects (A10.8.3.9)
Adjustments used in ASD to account for the effects of group action can also be used for LRFD. The
effect of closely-spaced shafts on axial load resistance can be considered in analyses by modifying
the p-y relationships using an efficiency factor, , to account for the:
10-18
In cohesive soils, the ultimate axial resistance of a drilled shaft group may be less than the
cumulative resistance of the individual shafts due to overlapping of zones of shear deformation in
the soil between the shafts (A10.8.3.9.2). In stiff cohesive soils, there is no loss in resistance due to
group effects. However, in soft cohesive soils where the cap is not in contact with the underlying
soil and in cohesionless soils (A10.8.3.9.3), the resistance of a group will be less than the cumulative
resistance of an equal number of individual shafts if the center-to-center (CTC) spacing between
shafts is less than 6D, and the group efficiency factors, , in Section 9.3.4.1 apply.
If a drilled shaft group is embedded in a stiff soil deposit overlying a weaker deposit (A10.8.3.9.3),
the potential exists for a punching failure of the group through the stiff soil into the soft soil. In
addition, the potential for settlements in the soft layer must be evaluated.
10.3.4.2 Negative Loading (A10.8.1.5)
Methods in ASD to estimate the magnitude and location of the maximum downdrag load can also by
used in LRFD. General comments regarding downdrag load estimation are presented in Chapter 4.
Briaud and Tucker (1994a) indicate that the potential for downdrag loading should be considered in
design when the indicators in Table 10-9 are present. In terms of performance limits, downdrag
generally poses a foundation settlement concern for friction shafts and for end-bearing shafts bearing
on a very stiff layer such as very dense sand or rock.
10.3.4.3 Uplift Loading (A10.8.3.7)
Uplift loading on shafts can be caused by lateral loads from supported structures, buoyancy effects
or expansive soils. Evaluation of the ultimate uplift resistance of a single shaft or shaft group by
ASD and LRFD are conducted using the same methods used to estimate ultimate axial side
resistance, QS, of the shaft or group to axial compression loading, but with a 20 percent reduction in
QS. The resistance factors for uplift in Table 10-5 incorporate a 20 percent reduction in QS for uplift
loading. In addition, the structural resistance of the shaft and the shaft/cap connection must also be
checked.
Table 10-9
Conditions When Downdrag Should be Considered in Design
(after Briaud and Tucker, 1994a)
5
6.
WARNING: Downdrag can occur even if the above conditions are not met.
10-19
Based on the results of full-scale load tests (e.g., Castilla, et al., 1984; Shahawy and Issa, 1992), a
depth of embedment of 2D to 3D provides full fixity for most service load conditions. Because the
degree of shaft-cap fixity is usually established empirically, there is no difference in the approach
taken for drilled shaft foundations designed by either ASD or LRFD.
10-20
Figure 10-2
Soil Profile for Example
Objective: To demonstrate the relationship between ASD and LRFD using an example drilled shaft
design by both methods.
Approach: To perform the evaluation, the following steps are taken:
y
Determine appropriate load factors and load combinations and calculate the
total factored load effects
Determine the required number and length of shafts in the group based on
geotechnical and structural criteria
Check the allowable (ASD) and factored (LRFD) axial structural resistance
of a single shaft
Using software program GROUP, Version 3.0 (Ensoft, Inc., Austin, Texas,
1994) determine the maximum factored bending moment in an individual
shaft and the maximum deformations for the shaft group
Determine whether the combined shaft stresses and shaft group deformations
are acceptable
kN-m
Longitudinal moment =
kN-m
Transverse moment =
kN-m
Vertical (axial) load =
kN
Longitudinal horizontal load =
kN
Transverse horizontal load =
kN
Step 3: Determine Ultimate Axial and Factored Geotechnical (Bearing) Resistance of a Single
Shaft
In this example, shaft tip resistance and side resistance in soil overburden and tip resistance in rock
were neglected. Side resistance in the underlying sandstone layer is assumed to provide the entire
10-22
axial load resistance. For shafts embedded into rock with uniaxial compressive strength greater than
1900 kPa, the Horvath and Kenney (1979) method is recommended for determining geotechnical
side resistance.
From the AASHTO LRFD Specification (AC10.8.3.5), the ultimate unit side resistance of a drilled
shaft in rock is according to Horvath and Kenney (1979):
qs (MPa) = 0.21 q u o r f c (MPa), whichever is lower
where:
qs =
qu =
fc' =
Therefore:
qs = 0.21 14 MPa = 0.786 MPa = 786 kPa
(Eq. 10-12)
where:
Qult = Ultimate axial resistance (kN)
Qs = Ultimate side resistance (kN)
As = Side wall area (m2)
For a shaft diameter = 0.9144 m,
As = 0.9144 Hs
and the ultimate shaft resistance, Qult, in kN is:
Qult = Qs = qs As =
Hs
where:
Hs =
Hs = __________ Hs
10-23
where:
qs = 0.65 from Table 10-5 for side resistance based on Horvath and Kenney (1979)
Step 4: Determine Required Shaft Length for ASD:
Based on consideration of the pier geometry, the shaft group layout in Figure 10-3 was selected.
Figure 10-3
Shaft Group Layout for Example
Based on Eq. 10-1, the required rock socket embedment depth can be computed using the following
relationship:
3 Qult 3 [( _______ kN/m) Hs]
=
Q = _______ kN
FS
FS
Using FS = 2.5,
Hs
7991 kN x 2.5
= 2.95 m
2258 kN/m x 3
m=
m.
10-24
3 QR = 3 (qs Qs) = 3 (
Hs
kN/m) H s i i Q i =
kN
m=
m.
(0.9144 m )2
d2
=
= 0.657 m2
4
4
Figure 10-4
Preliminary Longitudinal Shaft Reinforcement for Example
10-25
_______ kN
OK or NG
3
Mn = 2570 kN-m
From AASHTO (1997a), the factored moment resistance, Mr, is then:
Mr = Mn = (0.9) 2570 kN-m = 2313 kN-m
(A5.7.3.2.1-1)
where:
=
Using the software program GROUP, Version 3.0 (Reese, et al., 1994) and factored loads, the
maximum factored moments on any shaft in the transverse and longitudinal directions were
determined to be:
MUT = 20 kN-m
MUL = 207 kN-m
10-26
where:
MUT = Maximum factored transverse moment (kN-m)
MUL = Maximum factored longitudinal moment (kN-m)
Because, Mr >> MUT and MUL and the shafts are satisfactory with respect to bending resistance.
Note that the structural resistance of the shafts must also be checked for combined axial compression
and biaxial flexure interaction and for shear.
Step 9: Displacements and Tolerable Movement Criteria
Because evaluations of structure displacements by LRFD are made at the Service I Limit state where
= 1.0 and = 1.0, methods used to estimate settlement and lateral displacement by LRFD are
identical to those used for ASD, and are expressed by Eq. 10-2 as follows:
i # n
(Eq. 10-2)
Using GROUP program (Reese, et al., 1994) for lateral loaded pile/shaft group analysis, the shaft
cap deflections, i, were estimated as follows:
Vertical Settlement
Longitudinal Horizontal Deflection
Longitudinal Rotation
Transverse Horizontal Deflection
Transverse Rotation
9.7 mm
0.68 mm
-0.000187 rad
0.089 mm
-0.000159 rad
10-27
D Q D / QL + L
FS (1 + QD / QL)
(Eq. 7-7)
From Section 10.2.1, the AASHTO ASD factor of safety for these type of predictive methods is 2.5.
1.0
Dead Load
Factor,
D
1.25
2.0
1.25
1.75
3.0
1.25
1.75
QD/QL
Live Load
Factor, L
1.75
10-28
Resistance
Factor,
Average Load
Factor,
CHAPTER 11
CONVENTIONAL RETAINING WALL AND ABUTMENT DESIGN
11.1 Introduction
For both Allowable Stress Design (ASD) and Load and Resistance Factor Design (LRFD), the
design of conventional retaining walls and abutments requires consideration of geotechnical
capacity, overall stability, structural capacity and deformation limits. The design processes therefore
require both establishment of criteria for acceptable stress and deformation levels, and comparison
of these criteria with stress and deformation levels estimated from the design. This chapter:
With the exception of differences in foundation design noted in Chapters 8, 9 and 10, the
procedure for design of conventional gravity and cantilever retaining walls and abutments using
LRFD (A11.6) is identical to that followed using ASD. Generally, the ultimate bearing capacity,
resistance to sliding, overall stability and settlement of the wall foundation are checked, and the wall
is checked for lateral deflection. The following sections summarize the general design processes for
conventional retaining wall and abutment design using the ASD and LRFD approaches.
11.2.1 ASD Summary
Existing practice for geotechnical design of conventional retaining walls and abutments typically
follows the ASD approach for design of spread footing foundations presented in Chapter 8. Where a
retaining wall or abutment is supported with piles or drilled shafts, the design of the wall foundation
follows the ASD approach outlined in Chapter 9 or 10, as appropriate.
The design of conventional retaining walls and abutments may often be governed by criteria for
lateral deflection or tilt. Therefore, the design of a retaining wall or abutment by ASD requires an
estimation of foundation settlements and lateral displacements (geotechnical) and deflections
(structural) associated with mobilization of design earth pressures, and comparison of estimated
deflections with deformation criteria using the following:
i n
(Eq. 11-1)
where:
i =
n =
After all geotechnical deformation and capacity criteria are met and a suitable foundation type, size
and geometry is selected, the structural design of the wall or abutment is performed using service
loads and allowable stresses (AASHTO, 1997b), or by LFD or LRFD (AASHTO, 1997b or ACI),
consistent with the superstructure design.
11.2.2 LRFD Summary
Whereas ASD considers all uncertainty in the applied loads and ultimate geotechnical or structural
capacity in factors of safety or allowable stresses, LRFD separates the variability of these design
components by applying load and resistance factors to the load and material capacity, respectively.
The geotechnical design of conventional retaining walls and abutments typically follows the LRFD
approach for design of spread footing foundations presented in Chapter 8. Where a retaining wall or
abutment is supported with piles or drilled shafts, the design of the wall foundation follows the ASD
approach outlined in Chapter 9 or 10, as appropriate. Structural design of the wall is then performed
by LRFD procedures using factored loads and resistances.
Using the guidelines described in Chapters 3 and 4, load magnitudes, values of load factor and load
factor combinations for each applicable limit state must be developed. The key issues in the design
of retaining walls and abutments by LRFD is the application of maximum and minimum load
factors for dead, earth and surcharge loads as described in Chapter 4 and illustrated in the
example problem in Section 4.6.
11.2.2.1 Limit States (A11.5)
The design of conventional walls and abutments using LRFD requires evaluation of foundation
suitability at various Limit States (i.e., applicable Strength Limit States and the Service I Limit
State). The selection of a Strength Limit State(s) depends on the type of applied loading (e.g.,
Strength I for design vehicle loading without wind or Strength II for permit vehicle loading). The
design considerations which must be evaluated for a typical retaining wall or abutment supported on
a spread footing designed at the Strength and Service I Limit States are summarized in Table 11-1.
As conditions warrant, it may also be necessary to evaluate foundation performance at other limit
states (e.g., Extreme Event I for loading from earthquakes).
More detailed discussion of procedures used to evaluate the various Performance Limits is provided
in Section 11.3, and in Chapters 8, 9 and 10.
11-2
Table 11-1
Limit States for Design of Gravity Retaining Walls and Abutments
Strength Limit
Service Limit
Performance Limit
State(s)
State
Location of Base Resultant Force
U
Bearing Resistance
U
Sliding Resistance
U
Overall Stability
U
Settlement and Horizontal Movement
U
Structural Capacity of Wall Elements
U
11.2.2.2 Resistance Factors (A11.5.6)
Resistance factors for geotechnical design of spread footings with respect to bearing capacity and
sliding resistance were discussed in Chapter 8. Resistance factors for geotechnical and structural
design of driven pile and drilled shaft foundations were discussed in Chapters 9 and 10, respectively.
As indicated in Section 8.2.2.2, the eccentricity criteria for design of spread footings under AASHTO
LRFD were calibrated directly with existing AASHTO ASD criteria through analysis of cantilever
and gravity retaining walls and abutments. The calibration study (Barker, et al., 1991b) considered
a wide range of parameters.
In general, the calibration was performed by developing ASD wall designs resulting in critical base
pressure eccentricities (eB = 0.167 B or B/6 for footings on soil and eB = 0.25 B or B/4 for footings
on rock), factoring the loads and recomputing the eccentricity for LRFD. The computation process
considered the following parameter variations:
Varying wall heights from 1.5 m to 7.5 m and abutment heights from 4.5 m
to 7.6 m while maintaining constant base widths and other dimensions
Varying wall and abutment base widths while maintaining constant heights
and other dimensions
Varying stem location while maintaining constant height and base width
Varying the friction angle, f, of the retained earth and foundation soil
11-3
Varying the ratio of dead to live load applied to abutments from 1.0 to 3.0
A summary of the results from the calibration process are presented in Table 11-2. The tabulation
indicates the parameter which was varied in a given analysis, and the LRFD eccentricity resulting
for the parameter value which produced the critical ASD eccentricity. As indicated in Table 11-2,
the selected LRFD eccentricity limitation for retaining walls is approximately equal to the average
value from the calibration, whereas the selected LRFD eccentricity limitation for abutments is
somewhat greater (or less conservative) than the average value from the calibration.
Table 11-2
Summary of Parametric Study Results
from LRFD Eccentricity Calibration
(after Barker, et al., 1991b)
Structure
Type
Abutments
Retaining
Walls
eB/B
Footing on Soil
Footing on Rock
ASD
LRFD
ASD
LRFD
0.167
0.22
0.250
0.31
0.167
0.22
0.250
0.32
0.167
0.23
0.250
0.33
Varied Parameter
Stem Location
Soil Friction Angle
Wall Height
Base Width with:
QD = 35 kN
QL = 35 kN
0.167
0.21
0.250
0.30
QD = 54 kN
QL = 18 Kn
0.167
0.22
0.250
0.33
QD = 93 kN
QL = 31 kN
0.167
0.21
0.250
0.31
0.167
0.167
0.167
0.25
0.25
0.24
0.250
0.250
0.250
0.42
0.36
0.35
0.167
0.167
0.167
0.218
0.247
0.250
0.250
0.250
0.250
0.316
0.376
0.375
Stem Location
Wall Height
Base Width (w/ QD = QL = 0)
Average eB/B
Abutments
Walls
Selected eB/B
The calibration was performed in 1991 using the AASHTO LFD (AASHTO, 1997b) load factors
before the LRFD load factors listed in Chapter 4 had been finalized. Subsequent analyses by the
original calibrators for a limited number of retaining wall and abutment design cases indicated that
11-4
the differences in load factors does not have a significant effect on the calibration.
11.2.2.3 Comparison of Wall Design Using LRFD and ASD
The differences between the geotechnical design of a conventional retaining wall or abutment by
ASD and LRFD relate to differences in the foundation design as discussed in Chapters 8, 9 and 10.
11.2.2.4 Modification of Resistance Factors
Resistance factors for retaining wall and abutment foundation design can be modified as described in
Chapters 8, 9 and 10. Due to the variation in load factors for various load effects applicable for a
wall or abutment design, and the range of load combinations which may apply depending on the
problem geometry, modification of resistance factors or eccentricity criteria for these designs should
be based on a parametric study covering expected potential variations in loading conditions and
problem geometry. As a minimum, such a parametric study should include the anticipated range of
wall heights and dead to live load combinations to define the applicable range of the average load
factor for net destabilizing forces.
11.2.3 Summarized Comparison of ASD and LRFD
As noted before, the process used to develop a conventional retaining wall or abutment design using
LRFD differs very little from the process used for ASD. The similarity is illustrated in the parallel
flow charts in Figure 11-1. Specific differences between the methods and other important issues are
highlighted in the following section.
Other aspects of retaining wall or abutment design such as identifying special considerations (e.g.,
potential for loss of support through scour), developing a design foundation and retained soil profile
and determining requirements for construction control are inherent aspects of the design process
required for both LRFD and ASD.
11.3 Performance Limits
Design of a conventional retaining wall or abutment by either LRFD or ASD must provide adequate
resistance against geotechnical and structural failure and limit deformations to within tolerable
limits. In determining the wall geometry and details, and in establishing a suitable bearing level to
meet the criteria for vertical, inclined and/or moment loading, the design of these structures requires
consideration of many factors which can affect wall or abutment performance, including:
11-5
(1)
Figure 11-1
Generalized Flow Chart for Conventional Retaining Wall
and Abutment Design by ASD and LRFD
11.3.1 Displacements and Tolerable Movement Criteria (A11.6.2)
The vertical and lateral displacements of conventional retaining walls and abutments must be
evaluated for all applicable dead and live load combinations and compared with tolerable movement
criteria. Vertical and lateral movements of wall foundations can be estimated and evaluated using
the procedures and criteria described in Chapters 8, 9 and 10. At a minimum, conventional retaining
walls and abutments must deflect (either by foundation movement or structural deformation) a
sufficient magnitude to permit mobilization of the shear strength of the backfill soil and
development of the design (usually active) earth pressure on the wall stem. In general, lateral
movements of walls on shallow foundations can be estimated assuming the wall rotates or translates
as a rigid body due to the effects of earth loads and differential settlements along the base of the
wall.
11-6
The tolerable movement criteria for abutments should generally follow the guidance provided in
Section 8.3.1 for spread footing foundations. For other retaining walls, tolerable movement criteria
should be developed with consideration of the function and type of wall, anticipated service life, and
consequences of unacceptable movements (e.g., affect of wall movements on adjacent facilities).
11.3.2 Geotechnical Resistance (A11.6.3)
The geotechnical resistance of retaining walls and abutments can be evaluated for the selected
foundation type as described in Chapters 8, 9 and 10. The overall stability of an abutment or
retaining wall, including the retained ground and foundation, should be evaluated for all walls using
limit equilibrium methods of slope stability analysis at the Service Limit State as described in
Chapter 8.
11.3.3 Structural Resistance (A11.6.4)
The design of conventional walls and abutments must meet the requirements for the structural design
of concrete structures. The structural design of footings, driven piles and drilled shafts is discussed
briefly in Chapters 8, 9 and 10, respectively. Structural design of the wall stem and base slab is
performed using the factored loads developed for the geotechnical design of the wall foundation and
the factored structural resistance of the reinforced and unreinforced concrete wall or abutment.
11.3.4 Other Considerations
11.3.4.1 Loss of Passive Resistance (A11.6.3.6)
The sliding resistance of a retaining wall is controlled by the (1) shear resistance between the wall
foundation and the foundation subgrade, and by the (2) passive resistance in front of the wall due to
embedment. The component of sliding resistance provided by embedment is often ignored during
design due to the possibility that permanent or temporary excavations in front of the wall could
occur during the service life of the structure and lead to partial or complete loss of passive
resistance. Additional shear resistance can be developed by keying when the foundation materials
are rock or very stiff soils, or by increasing the width of the foundation.
11.3.4.2 Drainage (A11.6.6)
The lateral load on walls is affected by (1) the shear strength and unit weight of the soil backfill, and
by (2) the presence of water in the backfill. For these reasons, free-draining, granular soils are
usually specified for backfill to minimize the lateral earth pressure and to ensure that a permanent
water level cannot be maintained above the level of weep holes in the wall. The use of freedraining, granular backfills is also beneficial in limiting the development of uplift pressures along
the base of the wall due to seepage from the backfill soil.
11-7
2.
Unfactored and factored loads and moments for design of a cantilever retaining wall were developed
in the Classroom Example in Chapter 4. The geotechnical design of a spread footing to support the
wall is presented in the Classroom Example in Chapter 8.
11-8
CHAPTER 12
PREFABRICATED MODULAR WALL DESIGN
12.1 Introduction
Similar to the requirements described in Chapter 11 for conventional retaining walls and abutments,
both Allowable Stress Design (ASD) and Load and Resistance Factor Design (LRFD) of
prefabricated modular retaining walls requires consideration of geotechnical capacity, overall
stability and deformation limits. The design processes therefore require both establishment of
criteria for acceptable stress and deformation levels, and comparison of these criteria with stress and
deformation levels estimated from the design. This chapter:
y
With the exception of differences in foundation design noted in Chapter 8, the procedure for
design of prefabricated modular retaining walls using LRFD (A11.10) is identical to that followed
using ASD. Generally, the ultimate bearing capacity, resistance to sliding, base pressure resultant
eccentricity, overall stability and settlement of the wall foundation are checked, and the wall is
checked for lateral deflection. (Note: Sliding, eccentricity and overall stability should be checked at
the base of each module.) The following sections summarize the general design processes for
modular retaining wall design using the ASD and LRFD approaches.
12.2.1 ASD Summary
With the exception of differences in the contribution of soil weight in the evaluation of overturning
stability and eccentricity, the procedures for design of modular retaining walls by ASD are identical
to those described in Chapter 11 for conventional retaining walls and abutments. In this regard,
ASD design of modular retaining walls differs from the design of conventional retaining walls as
follows (AASHTO, 1997b):
y
The design of prefabricated modular walls may be governed by criteria for lateral deflection or tilt.
Therefore, the design of a modular retaining wall or abutment by ASD requires an estimation of
foundation settlements and lateral displacements associated with mobilization of design earth
pressures, and comparison of estimated deflections with deformation criteria using the following:
i n
(Eq. 12-1)
where:
i =
n =
After all geotechnical deformation and capacity criteria are met, the structural design of the wall is
performed using service loads and allowable stresses or by LFD (AASHTO, 1997b), or by LRFD
(AASHTO, 1997a or ACI).
12.2.2 LRFD Summary
Whereas ASD considers all uncertainty in the applied loads and ultimate geotechnical or structural
capacity in factors of safety or allowable stresses, LRFD separates the variability of these design
components by applying load and resistance factors to the load and material capacity, respectively.
The geotechnical design of prefabricated modular retaining walls typically follows the LRFD
approach for design of spread footing foundations presented in Chapter 8. Structural design of the
wall is then performed by LRFD procedures using factored loads and resistances.
Using the guidelines described in Chapters 3 and 4, load magnitudes, values of load factor and load
factor combinations for each applicable limit state must be developed. The key issue in the design
of modular walls by LRFD is the application of maximum and minimum load factors for dead,
earth and surcharge loads as described in Chapter 4 and illustrated in the example problem in
Section 12.4.
12.2.2.1 Limit States (A11.5)
The design of prefabricated modular walls using LRFD requires evaluation of foundation suitability
at various Performance Limit States (i.e., applicable Strength Limit States and the Service I Limit
12-2
State). The selection of a Strength Limit State(s) depends on the type of applied loading (e.g.,
Strength I for design vehicle loading without wind or Strength II for permit vehicle loading). The
design considerations which must be evaluated for a modular wall designed at the Strength and
Service I Limit States are summarized in Table 12-1. As conditions warrant, it may also be
necessary to evaluate foundation performance at other limit states (e.g., Extreme Event I for loading
from earthquakes). More detailed discussion of procedures used to evaluate the various Strength and
Service Limit States is provided in Section 12.3 and in Chapter 8.
Table 12-1
Strength and Service Limit States
for Design of Prefabricated Modular Walls
Strength Limit
Service I Limit
Design Consideration
State(s)
State
Location of Base Resultant Force
U
Bearing Resistance
Sliding Resistance
Overall Stability
U
U
(1)
Refer to Chapter 8.
Figure 12-1
Generalized Flow Chart for Modular Retaining Wall Design
12.3 Performance Limits
Design of a prefabricated modular retaining wall by either LRFD or ASD must provide adequate
resistance against geotechnical and structural failure and limit deformations to within tolerable
12-4
limits. In determining the wall geometry and details, and in establishing a suitable bearing level to
meet the criteria for vertical, inclined and/or moment loading, the design of these structures requires
consideration of many factors which can affect wall or abutment performance, including:
y
y
y
y
y
y
of concrete structures. The structural design of footings is discussed briefly in Chapter 8. Structural
design of the module members, and base slab if any, is performed using the factored loads developed
for the geotechnical design of the wall foundation and the factored structural resistance of reinforced
concrete for concrete systems, and the factored structural resistance of steel for steel systems.
12.3.4 Other Considerations
12.3.4.1 Loss of Passive Resistance (A11.10.3.7)
The sliding resistance of a retaining wall is controlled by (1) the shear resistance between the wall
foundation and the foundation subgrade, and (2) the passive resistance in front of the wall due to
embedment. The component of sliding resistance provided by embedment is often ignored during
design due to the possibility that permanent or temporary excavations in front of the wall could
occur during the service life of the structure and lead to partial or complete loss of passive
resistance. Additional shear resistance can be developed for modular walls by inclining the base of
the wall or by increasing the base width of the wall.
12.3.4.2 Drainage (A11.10.6)
The lateral load on and within the elements of modular walls is affected by (1) the shear strength and
unit weight of the soil backfill, and (2) the presence of water in the backfill. For these reasons, freedraining, granular soils should specified for backfill to minimize the lateral earth pressure and to
ensure that a permanent water level cannot be maintained behind the wall. The use of free-draining,
granular backfills is also beneficial in limiting the development uplift pressures along the base of the
wall due to seepage from the backfill soil.
12-6
Figure 12-2
Schematic of Example Problem
During the subsurface exploration, it was determined, based on CPT results, that the foundation soils
are predominantly a dense sand to a depth of more than 6 m below the proposed bottom of footing.
12-7
Dense sand and gravel underlies the foundation so that the elastic settlement
of the dense sand and gravel will be negligible
Wall and wall modules will be backfilled with free-draining granular fill
The seasonal high water table is 3 m below the base of the wall
The base of the wall will be supported at the rear and front of the modules by
concrete bearing pads
Objective: Demonstrate the procedure for geotechnical design of a prefabricated modular retaining
wall by LRFD and compare the results with those obtained using ASD.
Approach: To perform the evaluation, the following steps are taken:
y
The loads and resulting moments due to wall components, earth pressures
and live load surcharge are calculated
The appropriate load factors and combinations are determined and applied to
the unfactored loads and moments to determine the factored loading
conditions
Based on the factored vertical loads and resisting and overturning moments,
adequacy against overturning will be determined by checking the eccentricity
The sum of the factored horizontal forces for each strength limit loading case
are compared to the factored horizontal resistance to determine the adequacy
against sliding
12-8
Figure 12-3
Retaining Wall Area Designation for Weight of Concrete and Soil
The weights of the concrete modules and retained fill are summarized in Table 12-2.
Table 12-2
Unfactored Unit Properties of 2.44-m Long Wall Modules
(1)
Unit
No.
Height/Width
(m)
1.22/1.22
Weight of
Concrete,
WC
(kN)
32.2
1.22/1.83
1.83/2.44
Total
Density of fill = 17.3 kN/m3
WC/m
(kN/m)
Volume of
Fill, FV
(m3/m)
Weight of
Fill/m, WF(1)
(kN/m)
13.2
0.89
15.4
38.1
15.6
1.53
26.5
65.8
27.0
3.27
56.6
136.1
55.8
5.69
98.5
12-9
sin 2 ( + )
sin( + ) sin( - i)
sin sin( - ) 1 +
sin( - ) sin( + i)
from which:
ka =
= 0.315
(Eq. 4-7)
Using a rectangular distribution inclined as shown in Figure 12-3, the horizontal and vertical
components of live load surcharge and vertical earth pressure acting on the wall are:
PLSH = p H cos (90o - + ) = (5.3 kPa) (4.6 m) cos (6.5o + 24.8o) = 20.7 kN/m
PLSV = p H sin (90o - + ) = (5.3 kPa) (4.6 m) sin (6.5o + 24.8o) = 12.6 kN/m
12-10
(A3.11.5.1-1)
At the base of the footing (i.e., @ z = H), p = 0.315 (17.3 kN/m3) (4.61 m) = 25.1 kPa
The resultant of the active earth pressure is inclined at the wall friction angle, with respect to plane
surface along the back of the modular wall. The horizontal and vertical components of the earth
pressure (triangular distribution) acting on the wall are:
PaH = 2 pH cos( + 90o - ) = 0.5 (25.1 kPa)(4.61 m) cos(24.8o + 6.5o) = 49.5 kN/m
PaV = 2 pH sin( + 90o - ) = 0.5 (25.1 kPa)(4.61 m) sin(24.8o + 6.5o) = 30.0 kN/m
(E) Summary of Unfactored Loads
Table 12-3
Unfactored Vertical Loads and Resisting Moments
V
Moment Arm
Moment About
Item
(kN/m)
About Toe (m)
Toe (kN-m/m)
WC1
13.2
1.20
15.9
WC2
15.6
1.30
20.4
WC3
27.0
1.35
36.5
WF1
15.4
1.20
18.5
WF2
26.5
1.30
34.6
WF3
56.6
1.35
76.6
PEV1
3.7
1.97
7.3
PaV
30.0
2.24
67.3
PLSV
12.6
2.15
27.0
VTOT
200.6
Table 12-4
Unfactored Horizontal Loads and Overturning Moments
H
Moment Arm
Moment About
Item
(kN/m)
About Toe (m)
Toe (kN-m/m)
PLSH
20.7
1.91
39.6
PaH
49.5
TOTAL
70.2
1.14
56.4
96.0
12-11
DC
EV
LSV
LSH
Strength I-a
0.90
1.00
1.75
1.75
EH
(active)
1.50
Strength I-b
1.25
1.35
1.75
1.75
1.50
BC (max. value)
Strength IV
1.50
1.35
1.50
BC (max. value)
Probable Use
BC/EC/SL
Service I
1.00
1.00
1.00
1.00
1.00
Settlement
Notes: BC - Bearing Capacity; EC - Eccentricity; SL - Sliding; EH applies to both PaV and PaH
By inspection:
y
Strength I-a (minimum vertical and maximum horizontal loads) will govern
for the case of sliding and overturning (eccentricity)
For the case of bearing capacity (maximum), both Strength I-b and Strength
IV must be evaluated
12-12
Strength I-b(1)
(kN/m)
Strength IV(1)
(kN/m)
Service I(1)
(kN/m)
WC1
V
(unfactored)
(kN/m)
13.2
11.9
16.5
19.8
13.2
WC2
15.6
14.0
19.5
23.4
15.6
WC3
Item
27.0
24.3
33.7
40.5
27.0
(1)
15.4
12.3
16.6
16.6
12.3
WF2(1)
26.5
21.2
28.6
28.6
21.2
(1)
56.6
45.3
61.1
61.1
45.3
WF
98.5
78.8
106.3
106.3
78.8
PEV1
3.7
3.7
5.0
5.0
3.7
PLSV
12.6
22.0
22.0
12.6
PaV
30.0
45.0
45.0
45.0
30.0
WF1
WF3
(1)
VTOT
200.6
199.7
248.0
240.0
180.9
80% of the soil weight inside the modules is considered for the factored vertical loads.
Table 12-7
Factored Horizontal Loads
PLSH
H
(unfactored)
(kN/m)
20.7
PaH
49.5
74.3
74.3
74.3
49.5
HTOT
70.2
110.5
110.5
74.3
70.2
Item
Strength I-a
(kN/m)
Strength I-b
(kN/m)
Strength IV
(kN/m)
Service I
(kN/m)
36.2
36.2
20.7
12-13
Table 12-8
Factored Moments from Vertical Forces (Mv)
(1)
Item
M
(unfactored)
(kN-m/m)
Strength I-a(1)
(kN-m/m)
Strength I-b(1)
(kN-m/m)
Strength IV(1)
(kN-m/m)
Service I(1)
(kN-m/m)
WC1
15.9
14.3
19.9
23.8
15.9
WC2
20.3
18.3
25.4
30.5
20.3
WC3
36.5
32.9
45.6
54.8
36.5
WF1
18.5
14.8
20.0
20.0
14.8
WF2
34.6
27.7
37.4
37.4
27.7
WF3
76.6
61.7
82.7
82.7
61.3
PEV1
7.3
7.3
9.8
9.8
7.3
PLSV
27.0
47.3
47.3
27.0
PaV
67.3
100.9
100.9
100.9
67.3
MvTOT
304.0
324.7
389.0
359.9
278.1
80% of the soil weight inside the modules is considered for the factored vertical loads.
Table 12-9
Factored Moments from Horizontal Forces (Mh)
PLSH
M
(unfactored)
(kN-m/m)
39.6
PaH
56.4
84.7
84.7
84.7
56.4
MhTOT
96.0
153.9
153.9
84.7
96.0
Item
Strength I-a
(kN-m/m)
Strength I-b
(kN-m/m)
Strength IV
(kN-m/m)
Service I
(kN-m/m)
69.2
69.2
39.6
Step 4: Eccentricity
The eccentricity of the retaining wall is checked in Table 12-10 by comparing the calculated
eccentricity, e, for each loading group to the maximum allowed eccentricity (emax) of B/4 specified
in AASHTO LRFD (1997a) (A10.6.3.1.5):
e = B/2 - Xo
where:
Xo = Location of the resultant from the toe=(MvTOT - MhTOT)/VTOT
B = 2.44 m (cos 9.462o) = 2.41 m
12-14
Group
Strength I-a
Table 12-10
Summary of Eccentricity Check
VTOT
MvTOT
MhTOT
Xo
(kN/m) (kN-m/m) (kN-m/m)
(m)
199.7
324.7
153.9
0.86
e
(m)
0.34
emax
(m)
0.60
Strength I-b
248.0
389.0
153.9
0.95
0.26
0.60
Strength IV
240.0
359.9
84.7
1.15
0.06
0.60
Service I
180.9
278.1
96.0
1.01
0.20
0.60
For all cases, e is less than or equal to emax; therefore, the design is adequate with regard to
eccentricity.
Note that for ASD, emax = B/6, or 0.40 m for a 2.41 m wide footing. Because ASD is equivalent to
the Service I Limit State in Table 12-10, the design also meets the ASD eccentricity requirements.
Step 5: Bearing Resistance
(A) Estimate the Bearing Pressures
The adequacy for bearing resistance is developed based on a rectangular distribution of soil pressure
over point supports (i.e., bearing pads) at the rear and front of the modules as indicated in Figure 124. In accordance with AASHTO (1997a) (A11.10.3.3) 80 percent of the soil weight within the
modules is considered to be transferred to the front and rear support points or bearing pads.
Therefore, the critical bearing case will be for the individual bearing pads rather than for the full
base width of the wall.
For a rectangular distribution with the resultant bearing pressure located near the toe and heal
respectively:
B - Xo
Q1 = VTOT
B
Q2 = VTOT - Q1 = VTOT Xo/B
The uniform pressure will be distributed over the bearing pad width, BBP = 0.6 m:
12-15
qmax = q1 = Q1/BBP
qmin = q2 = Q2/BBP
Figure 12-4
Bearing Pads at Rear and Front of Modules
Table 12-11
Summary of Maximum Bearing Pressures
Strength I-a
VTOT
(kN/m)
199.7
Xo
(m)
0.86
Q1
(kN/m)
128
Q2
(kN/m)
71
qmax
(kPa)
213
qmin
(kPa)
118
Strength I-b
248.0
0.95
150
98
250
163
Strength IV
240.0
1.15
125
115
208
192
Service I
180.9
1.01
105
76
175
127
Group
qR = qn = qult
qult = 0.5 s B CW1 Nm + s CW2 Df Nqm
(A10.6.3.1.1-1)
(A10.6.3.1.2c-1)
Values of CW1, CW2, Nm and Nqm can be obtained using the equations and tables in the AASHTO
LRFD Specification (A10.6.3.1.2). The resistance factor, , from Table 8-8 in Chapter 8 for bearing
capacity of a spread footing on sand determined by the rational method using f estimated from CPT
results is 0.45. Therefore:
12-16
kN
m
)(0.60 m)(1.0)(25.0)+(18.0
kN
m
)(1.0)(1.0 m)(25.1)]
qR = 0.45 [135 kPa + 451.8 kPa] = 0.45 [586.8 kPa] = 264 kPa
Because the factored bearing resistance, qR, exceeds the maximum factored bearing stress, qmax =
250 kPa, the bearing resistance is adequate.
Relative to ASD, as represented by the Service I Limit State, the factor of safety against bearing
capacity failure is qult/qmax = 586.8 kPa/175 kPa = 3.3. Therefore, the design is also acceptable with
respect to bearing capacity by ASD.
Step 6: Sliding
For sliding at the bearing pad level, sliding of the pads and shear of the soil between the pads is
checked. As per Step 5, 80 percent of the soil fill weight is assumed to be transferred to the bearing
pads (A11.10.3.3). The remaining soil fill weight (20 percent) is assumed to be transferred directly
to the foundation soil. Using this assumption, the sliding resistance is calculated using Eq. 8-15 as
follows:
QR = Q + ep Qep
where:
=
=
ep =
Q =
Qep =
Q will have units of force per unit length of wall when using the vertical forces from Table 12-6.
For individual bearing pads, the sliding resistance is computed as follows:
Q = V tan
(A10.6.3.3-2)
Strength I-a
Table 12-12
Summary of Sliding Resistance
EV (0.2 WF)
VTOT
EV
(kN/m)
(kN/m)
1.00
199.7
19.7
Strength I-b
1.35
248.0
26.6
157.6
110.5
Strength IV
1.35
240.0
26.6
153.1
74.3
Service I
1.00
180.9
19.7
140.4
70.2
Group
12-18
Q
(kN/m)
125.6
HTOT
(kN/m)
110.5
Because the factored sliding resistance ( Q) calculated is greater than the factored horizontal
loading for all strength limit states, the design is acceptable by LRFD. By ASD (as represented by
the Service I Limit State), the factor of safety against sliding = Q/HTOT = 140.4 kN/m / 70.2 kN/m =
2.0, which is acceptable ($ 1.5 as indicated in Table 8-1 of Chapter 8).
Summary
Table 12-13
Summary of Spread Footing Design by LRFD and ASD
LRFD
ASD
Factored
Required
Factored
Performance Limit
Resistance/
FS/
Actual FS/
Load/
Eccentricity
Eccentricity Eccentricity
Eccentricity
Limit
Limit
U
Eccentricity
0.25 B
0.14 B
0.167 B
0.08 B
Bearing Resistance
264 kPa
250 kPa
3.0
3.3
Sliding Resistance
158 kN/m
110 kN/m
1.5
2.0
The design of a prefabricated modular wall, as illustrated in this example problem, essentially
involves the design of a geometry (i.e., base width) and bearing pads which satisfy the eccentricity,
bearing resistance and sliding resistance criteria described in Chapter 8 for spread footings. As
discussed in chapter 8, the resistance factors for bearing and sliding were developed from a
reliability-based calibration, whereas the eccentricity were calibrated directly to ASD.
As summarized in Table 12-13, an acceptable design is achieved in this example by both LRFD and
ASD. Whereas the ASD factors of safety, FS, for bearing and sliding resistance are fixed, however,
the LRFD resistance factors could possibly be increased with additional data accumulation and
reliability-based calibration for similar soils and loading conditions. It is possible therefore, that
with further calibration, more reliable and more economical designs may be achieved using
LRFD.
12-19
12-20
CHAPTER 13
ANCHORED WALL DESIGN
13.1 Introduction
For both Allowable Stress Design (ASD) and Load and Resistance Factor Design (LRFD), the
design of anchored retaining walls requires consideration of geotechnical capacity of anchorages and
embedment of vertical wall elements, structural capacity of anchorages, vertical wall elements and
facings, overall stability and deformation limits. The design processes therefore require both
establishment of criteria for acceptable stress and deformation levels, and comparison of these
criteria with stress and deformation levels estimated from the design. This chapter:
y
With the exception of differences in foundation design noted in Chapters 8, 9 and 10, the
procedure for design of anchored retaining walls using LRFD (A11.8) is identical to that followed
using ASD. The geotechnical aspects of anchored wall design include evaluation of anchor pullout
resistance, the bearing and passive resistance of the vertical wall elements, overall stability and
foundation and wall movements. In addition, the structural capacity of the anchors, vertical wall
elements and facing (temporary and permanent) must also be evaluated. The following sections
present general design considerations and summarize the general design processes for anchored wall
design using the ASD and LRFD approaches.
13.2.1 General Design Considerations
As shown in Figure 13-1, anchored walls are comprised of anchors constructed using prestressing
strand or steel bars, discrete vertical wall elements (i.e., usually rolled steel sections) spaced
approximately 2 to 3 m along the length of the wall, and facing (i.e., timber or precast concrete
lagging or shotcrete, and cast-in-place concrete for some permanent walls). In some instances,
typically in urban environments, the vertical wall element consists of continuous reinforced concrete
panels constructed using diaphragm wall methods which rely on a bentonite or polymer slurry to
provide temporary support during excavation of the panel. As the excavation is made, anchors are
installed into drilled holes and extend through the wall face to a bearing plate supported on the face
of the vertical wall element or a wale between adjacent vertical wall elements.
13-1
Drilling the anchor hole at a typical angle of about 10o to 30o to the
horizontal, then installing and grouting the anchor in the bond length
Stressing and testing each anchor, locking off the anchor at the design load,
and if the anchor has grease and sheathing corrosion protection, grouting the
unbonded length
13-2
The geometry of wall construction shown in Figure 13-1 is typical for most anchored walls. The
length of anchor is designed to provide an adequate bond length to resist loads on the wall and to
extend a reasonable distance beyond the critical failure surface. Although the inclination of anchors
is usually limited to a maximum of about 30o to the horizontal to minimize axial loads on the vertical
wall elements, steeper inclinations can be used to avoid interference with buried structures or to
reach more suitable materials which provide a higher bond capacity. The vertical spacing of anchors
can be optimized to provide a balanced design between the number of anchors and bending of the
vertical wall element.
Additional detail relative to the design, installation and field testing of anchors and anchored walls is
provided in Cheney (1988).
13.2.2 ASD Summary
Existing practice for geotechnical design of anchored walls follows the ASD approach (AASHTO,
1997b), wherein all uncertainty in the variation of applied loads transferred to the wall elements and
the ultimate geotechnical capacity of the soil and rock to support the loads are incorporated in a
factor of safety, FS. Structural design of anchors, vertical wall elements and facing is based on
allowable stresses, all, which account for uncertainty in the variation in applied loads and structural
capacity. As a result, loads used for design, Q, consist of those actual forces estimated to be applied
directly to the structure. In LRFD terminology, this process is equivalent to applying a load factor
of 1.0 to the estimated forces. In ASD, six primary performance and failure conditions are evaluated
in the design of an anchored wall:
y
y
y
y
y
y
(Eq. 13-1)
where:
i =
n =
13-3
The actual wall movements will be controlled primarily by the stiffness of wall and anchorages, the
engineering characteristics of the supported ground, and the quality control exercised during
construction. Because highway structures are not typically supported by anchored walls, tolerable
movement criteria are usually a function of the proximity of the wall to deformation-sensitive
structures. However for walls supporting an abutment, designs may be governed by criteria for
lateral deflection or tilt as described in Chapter 11. The basis for estimation of ground movements
behind anchored walls and criteria for evaluating their performance are discussed in Section 13.3.1.
Following the preliminary selection of the spacing of anchors to support the apparent earth pressure
distribution assumed for design, the ultimate pullout capacity of anchors, Rn, can be estimated as
described in Section 13.3, and the bearing and lateral capacity of vertical wall elements, Rn, can be
estimated by available theoretical or semi-empirical procedures as described in Chapters 8 and 9.
The suitability of the design with respect to anchor pullout and to bearing and lateral capacity are
then evaluated by determining the allowable load components, Qall, using:
Q Q all = R n FS = Q ult FS
(Eq. 13-2)
where:
Q=
Qall =
Rn =
FS =
The required FS with respect to bearing capacity and sliding are generally specified by the
governing agency, and may be constant or variable. The AASHTO ASD (AASHTO, 1997b) and
FHWA (Cheney, 1988) FS criteria for anchor pullout, bearing capacity and passive lateral capacity
of vertical wall elements, and overall stability are shown in Table 13-1. The AASHTO and FHWA
FS values assume that designs are based on soil and rock properties determined through appropriate
field and laboratory testing and not presumptive properties.
Table 13-1
Factors of Safety for Anchored Walls
(AASHTO, 1997b; Cheney, 1988)
Required Minimum Factor of
Failure Condition
Safety (FS)
AASHTO
FHWA
Anchor Pullout
y Soil
2.5
2.5
y Rock
3.0
3.0
Bearing Capacity
Refer to Chapters 8, 9 and 10
Lateral (Passive) Capacity
1.5
1.5
Overall Stability
Refer to Chapter 8
The overall stability of the retained earth and foundation must also be evaluated using procedures
13-4
(Eq. 13-3)
where:
all = Allowable tensile or flexural stress (kPa)
Pall = Allowable tensile or flexural structural capacity (kN)
A=
AASHTO (1997b) and FHWA (Cheney, 1988) indicate maximum allowable anchor loads, Pall, as
indicated in Table 13-2.
Table 13-2
Maximum Allowable Tensile Loads for Ground Anchors
(AASHTO, 1997b; Cheney, 1988)
(1)
(2)
(3)
Anchor Type
Allowable Load (1)
Temporary(2)
0.80 Pn(2)(3)
Permanent
0.60 Pn(2)
Pn = Guaranteed Ultimate Tensile resistance (GUTS) of anchor (kN)
Implied by Division II anchor stressing and load testing requirements (i.e., to
1.33Q) and assuming that temporary stressing to 80% GUTS is permissible.
For acceptance testing only.
Alternatively, the structural design of the vertical wall elements and facing may be performed by
LFD (AASHTO, 1997b) or LRFD (AASHTO, 1997a) or ACI.
13.2.3 LRFD Summary
Whereas ASD considers all uncertainty in the applied loads and ultimate geotechnical or structural
capacity in factors of safety or allowable stresses, LRFD separates the variability of these design
components by applying load and resistance factors to the load and material capacity, respectively.
When properly developed and applied, this approach can provide a consistent level of safety for
the design of all structure components. Thus, the probability that a structure component will fail or
perform unacceptably is no different than any other component. As described on Section 4.3, the
resistance and deformation of supporting soil and rock materials and structure components must
satisfy the LRFD equations below. For the Strength Limit States:
i i Qi
Rn = Rr
i i i
(Eq. 13-5)
where:
i =
i =
Qi =
=
Rn =
Rr =
i =
n =
Relative to the external stability of the wall and the structural adequacy of system components,
the suitability of an anchored wall with respect to the geotechnical and structural resistance can
be obtained using Eq. 13-4, rewritten as:
i i Qi Qult = QR
(Eq. 13-6)
i i Q i Q i
(Eq. 13-7)
and
= Qi
where:
i i Q i
=
Resistance factor (dim)
Qult = Ultimate geotechnical resistance (kN)
QR = Factored geotechnical resistance (kN)
Pn = Pult = Nominal (ultimate) structural resistance of vertical wall elements,
anchors, facing and connections (kN)
Pr = Factored structural resistance of vertical wall elements, anchors, facing and
connections(kN)
The load factors and load factor combinations used for design were presented in Chapter 4. In
general, values of i > 1.0 are used to evaluate ultimate ground or structure capacity at the Strength
Limit States, whereas the deformation performance of structures is evaluated at the Service I Limit
State using i = 1.0 (or i = 0.3 for wind loads). In ASD (AASHTO, 1997b), values of i = 1.0 (or i
= 0.3 for wind loads) are used to evaluate structures for both strength (allowable stress) and
serviceability (deflection). Because the AASHTO LRFD Specification for anchored walls was
developed based on direct calibration with AASHTO ASD, the results of analyses by LRFD and
ASD to evaluate external wall stability and the structural adequacy of wall components are
similar, and for foundation and wall deformations are identical.
13-6
When using Eq. 13-4 for anchored wall design at the Strength Limit States, the following values of
can normally be used:
D = R = 1.00
I = 1.05 for structures deemed operationally important, 1.00 for typical
structures and 0.95 for relatively less important structures.
For the purpose of this chapter, the value of I is assumed equal to 1.0.
When using Eq. 13-5 to evaluate an anchored retaining wall at any Service Limit State, D, R, and
I = 1.0.
Values of load factor and load factor combinations for each applicable limit state must be developed
using the guidelines described in Chapters 3 and 4, and loads should be developed as described in
Chapter 4. The ultimate resistance, Rn, should be determined for each type of resistance (e.g.,
bearing criteria, sliding, overall stability, and anchor pullout and structural adequacy) as described in
Section 13.3.
Values of # 1.0 are applied when evaluating geotechnical or structural resistance for any strength
limit state using Eq. 13.4. Currently the value of = 1.0 is applied when evaluating an anchored
retaining wall for any service limit state using Eq. 13-5. Selection and modification of resistance
factors, , are described in Sections 13.2.3.2 and 13.2.3.4.
13.2.3.1 Limit States (A11.5)
In general, the design of anchored retaining walls using LRFD requires evaluation of foundation
suitability at various Performance Limit States (i.e., applicable Strength Limit States and the Service
I Limit State). The selection of a Strength Limit State(s) depends on the type of applied loading
(e.g., Strength I for design vehicle loading without wind or Strength II for permit vehicle loading).
The design considerations which must be evaluated for a typical anchored retaining wall designed at
the Strength and Service I Limit States are summarized in Table 13-3. As conditions warrant, it may
also be necessary to evaluate wall performance at other limit states (e.g., Extreme Event I for loading
from earthquakes).
13-7
Table 13-3
Strength and Service Limit States for Design of Anchored Walls
Strength Limit
Service I
Design Consideration
State(s)
Limit State
Anchor Pullout
U
Structural Resistance of Anchor
U
Structural Resistance of Vertical Wall Elements
U
Passive and Bearing Resistance of Vertical Wall Elements
U
Structural Resistance of Facing
U
Overall Stability
U
Foundation and Wall Movement
U
Additional guidance for the design of anchored walls is provided in Cheney (1988).
13.2.3.2 Resistance Factors (A11.5.6)
Resistance factors for the design of anchored walls are presented in Table 13-4. The resistance
factors for passive resistance of vertical wall elements, anchor pullout and structural design of
anchors were developed based on engineering judgment and calibration with current allowable stress
design procedures for anchored walls. The provisions of Sections 5, 6 and/or 8 of the AASHTO
LRFD Specification (1997a) should be followed in evaluating the structural resistance of wall
components.
The variation of -factors for geotechnical resistance represents the judged relative reliability of the
different methods of estimating resistance.
13.2.3.3 Comparison of Anchored Wall Design Using LRFD and ASD
13.2.3.3.1 Geotechnical Design
To illustrate the relative differences between LRFD and ASD, the equivalent LRFD factor of safety
(FSLRFD) has been determined for each of the methods presented in Table 13-4 for estimating the
geotechnical capacity of passive resistance of embedded vertical wall elements and the pullout
capacity of anchors. As presented in Table 13-5, FSLRFD was determined as:
FSLRFD = /
(Eq. 13-8)
where:
=
=
The ASD factor of safety (FSASD) from Table 13-1 for each category is also presented in Table 13-5.
13-8
Resistance
Factor (1)
Refer to Chapter
8, 9 or 10
13-9
Table 13-5
Comparison of FSASD and FSLRFD for Anchored Walls
Resistance
Load
FSLRFD(1
Factor
Method/Soil/Condition
Factor(1)
FSASD(2)
)
Passive
1.35
2.2
1.5
0.60
Resistance Soil
1.50
2.5
1.5
of Vertical
1.35
2.2
1.5
Wall
Rock
0.60
1.50
2.5
1.5
Elements
1.35
2.1
2.5
Sand:
0.65
1.50
2.3
2.5
Correlation w/ SPT
1.35
1.9
2.5
Pullout load tests
0.70
1.50
2.1
2.5
Pullout
1.35
2.1
2.5
Clay:
Resistance
0.65
1.50
2.3
2.5
Correlation w/ Su from lab tests
of Anchors
1.35
2.1
2.5
Correlation w/ Su from vane tests
0.65
1.50
2.3
2.5
1.35
1.9
2.5
Pullout load tests
0.70
1.50
2.1
2.5
1.35
2.5
3.0
Rock:
0.55
1.50
2.7
3.0
Correlation with rock type only
1.35
2.3
3.0
Pullout
Co from lab tests - soft rock only
0.60
1.50
2.5
3.0
Resistance
1.35
1.8
3.0
of Anchors Laboratory rock-grout bond tests
0.75
1.50
2.0
3.0
1.35
1.7
3.0
Pullout load tests
0.80
1.50
1.9
3.0
(1)
(2)
= 1.35 for at-rest earth pressure and 1.50 for active earth pressure; From Table 13-1
13.2.3.3.2 Structural Design
In ASD, the ultimate unit stress or ultimate load capacity for each material type is reduced to an
allowable value by application of a modifier to account for uncertainty in resistance and applied
load. The allowable tensile loads for ground anchors recommended by AASHTO (1996b) and
FHWA (Cheney, 1988) are listed in Table 13-2.
To illustrate the relative differences between LRFD and ASD with respect to anchor tensile capacity,
the equivalent LRFD allowable load has been determined for each of the loading conditions
identified in Table 13-4. As presented in Table 13-6, Pall(LRFD) was determined as:
Pall(LRFD) =
Pn
(Eq. 13-9)
where:
13-10
Pall(LRFD) =
=
=
Pn =
The FHWA allowable anchor tensile loads for each loading condition from Table 13-1 are also
presented in Table 13-6, as is the ASD allowable stress for flexural resistance of steel beams
(AASHTO, 1997b).
Table 13-6
Comparison of ASD Allowable Load/Stress with
the LRFD Equivalent Load/Stress for
Structural Design of Ground Anchors
ASD Allowable
LRFD
Load/Stress
Equivalent
Resistance
Load
Pall/all
Allowable
Factor
Loading Condition
Factor (1)
Load/Stress
1.35
0.74 Pn
1.0
Tensile
Temporary
0.67 Pn
1.50
Resistance
0.67 Pn
1.35
0.9
0.60 Pn
of Anchors Permanent
1.50
(1)
= 1.35 for at-rest earth pressure and 1.50 for active earth pressure.
0.80 Pn
0.80 Pn
0.60 Pn
0.60 Pn
0.80 Pn
0.80 Pn
0.60 Pn
0.60 Pn
As described in Section 13.2.3.3.1 with respect to geotechnical design, the actual average load
factor will typically range between 1.50 (for active earth pressure) to somewhat less than 1.75 (for a
large live load surcharge), such that the actual LRFD equivalent allowable loads and stresses will
be somewhat lower (i.e., more conservative) than the values listed in Table 13-6 for active earth
pressures.
Other aspects of anchored wall design such as identifying special considerations (e.g., settlement of
soil behind the wall, sequence of construction, developing a design foundation and retained soil
profile and determining requirements for construction control) are inherent aspects of the design
process required for both LRFD and ASD.
13.2.3.4 Modification of Resistance Factors
Resistance factors for anchored wall design using LRFD were developed based on direct calibration
with ASD. Therefore, resistance factors for various aspects of anchored wall design can be modified
using the procedures summarized in Chapter 7.
13.2.3.4.1 Geotechnical Design
As stated in Section 13.2.3.2, the LRFD resistance factors for geotechnical design of anchored walls
in Table 13-4 were developed based on direct calibration with ASD. As shown in Table 13-5,
13-11
application of the resistance factors results in "equivalent" safety factors of 1.9 to 2.7 for anchor
pullout 2.5 and for passive resistance of embedded vertical wall elements for walls supporting active
soil pressure (i.e., = 1.50). For walls subjected to live load surcharge, the "equivalent" factor of
safety would be somewhat higher.
In ASD, the designer or owner might decide to increase or decrease required factors of safety or
allowable design stresses in consideration of a number of factors, such as:
When using LRFD, similar flexibility to vary the required level of safety should also be available.
Additionally, whereas the same factor of safety is generally used in ASD regardless of the source of
loading, the equivalent factor of safety in LRFD (defined by Eq. 13-8) varies for a given resistance
factor depending on the source of loading, such as described previously with respect to live load
surcharge.
To modify the resistance factors for geotechnical design of piles to account for average load factors
and equivalent factors of safety other than those identified in Table 13-5, the following equation may
be used:
FS
m = T x T x T
FSD D
(Eq. 13-10)
where:
m =
T =
FST =
FSD =
T =
D=
13-12
m = T x allA x T
allT D
(Eq. 13-11)
P
m = T x allA x T
PallT D
(Eq. 13-12)
or
where:
T =
allA =
allT =
PallA =
PallT =
Modifying resistance factors may seem reasonable, but such modification may not be consistent with
the goal of LRFD to achieve equal reliability against failure of structure components, unless the
factor of safety accurately models the reliability of the predictive method used.
The resistance factors may be more appropriately modified through application of the probabilistic
procedures described in Chapter 3 to achieve the desired level of reliability if a sufficient amount of
data is available.
13.2.4 Summarized Comparison of ASD and LRFD
As noted before, the process used to develop an anchored wall design using LRFD differs very little
from the process used for ASD. The similarity is illustrated in the parallel flow charts in Figure 132. Specific differences between the methods and other important issues are highlighted in Section
13.3.
13.3 Performance Limits
13.3.1 Displacements and Tolerable Movement Criteria (A11.8.3, A3.11.5.6)
The displacement of anchored walls must be evaluated at the Service I Limit State for all applic-
13-13
Figure 13-2
Generalized Flow Chart for Anchored Wall Design
by ASD and LRFD
13-14
able load combinations. Because evaluations of structure displacements by LRFD are made at the
Service I Limit State where = 1.0 and = 1.0, methods used to estimate settlement and lateral
displacement by LRFD are identical to those used for ASD. The vertical and lateral displacement
of anchored walls is a complex soil-structure interaction problem, and deformation analyses can be
performed using modified forms of beam on elastic foundation theory or finite element analyses.
For many projects, however, such analyses are not warranted unless deformation-sensitive structures
are in close proximity to the wall. Alternatively, for anchored walls built using proven wall
construction technology, Figure 13-3 can be used as a guide to estimate ground surface settlement
behind the walls. The corresponding maximum lateral movement for walls supporting sand and stiff
to hard clay soils usually does not exceed about 0.3 percent of the depth of excavation. Figure 13-3
does not consider the effects of other construction activities (e.g., dewatering, ground heave at the
wall base, or poor construction quality).
Because anchored walls are not typically used to support highway superstructures, the tolerable
movement of the wall should be developed based on consideration of the effects of possible wall
movements on other structures and facilities near the wall.
13-15
L i Q i Q a
(Eq. 13-13)
where:
L=
Bond Length (m)
iQi = Summation of factored anchor loads from contributory area (kN)
=
Qa =
The factored anchor load is developed using an apparent earth pressure distribution, which depends
on the soils to be supported as described in Chapter 4, and various assumptions regarding the
distribution of load from the tributary area of wall face. In the proportional method, the top anchor
row is assumed to support the tributary area of pressure between the top of the wall and the midpoint
between the upper two anchor levels; and the bottom anchor row is assumed to support the pressure
between the base of the wall and the midpoint between the two lowest anchor rows. Alternately, the
embedded portion of the vertical wall element can be assumed to support the pressure between the
base of the exposed wall and the midpoint between the base and the lowest anchor row.
Intermediate anchor rows are assumed to support the pressure from the midway points between
vertically adjacent anchor rows and pairs of vertical wall elements.
As described in FHWA research reports (Cheney, 1988; Weatherby, 1982), various procedures are
available for estimating the ultimate unit geotechnical resistance of ground anchors in soil and rock
using semi-empirical correlations or in-situ testing. The LRFD Specification, for which resistance
factors in Table 13-4 were developed, uses the following methods:
Cohesive Soils - Based on soil type and stiffness with reference to undrained
shear strength (Cheney, 1988) or based on field pullout load tests.
Rock - Based on correlation with rock type (Cheney, 1988) or rock shear
strength testing (Weatherby, 1982), or based on laboratory rock-grout bond
tests or field pullout load tests.
Tables 13-7 and 13-8 provide conservative values of Qa for anchors in soil and rock, respectively.
These values are intended for preliminary design or evaluation of the feasibility of straight shaft
anchors installed in small-diameter holes which are grouted using low pressures. Pressure-grouted
anchors usually achieve higher capacities. Because anchor capacity in soil and rock can be strongly
influenced by the method of anchor hole advancement, hole diameter, anchor type, type of grout and
13-16
grouting pressure, selection of the anchor type and determination of Qa for final design should be
made by the specialty geotechnical contractor selected for wall construction.
Table 13-7 (A11.8.4.2-1)
Ultimate Unit Resistance of Anchors in Soil
(AASHTO, 1997a; AASHTO, 1997b; Cheney, 1988)
Soil Type
Sand and
Gravel
Sand
Sand and
Silt
Soil Type
Silt-Clay
Mixture
Ultimate Unit
Anchor Resistance,
Qa (kN/m)
145
220
290
100
145
190
75
100
130
Ultimate Unit
Anchor Resistance
Qa (kN/m)
30
60
Rock Type
Granite or Basalt
Dolomitic Limestone
Soft Limestone/Sandstone
Slate and Hard Shale
Soft Shales
13-17
Unless beam on elastic foundation, finite element or other methods of soil-structure interaction
analysis are used, the maximum bending moment, Mmax, in the vertical wall element may be
determined using the proportional method to compute anchor forces and wall element stresses. For
the wall section above the top anchor row, the vertical wall element is designed as a cantilever and
Mmax is determined as:
M max = 0.5 pLx
For wall sections between rows, the vertical wall element is designed as a simply-supported beam
and Mmax is determined as:
M max = 0.125 pLx
If the wall is not embedded or the lateral resistance of the embedded length is neglected, the vertical
wall element below the lowest anchor level can be designed as a cantilever and Mmax can be
determined using Eq. 13-14. If the embedded portion of the vertical wall element is assumed to
provide support and embedment is not required to resist unbalanced forces due to base instability,
Mmax can be determined for the section of wall below the lowest anchor row of anchors using:
M max = 0.333 pLx
For sections of vertical wall elements sections of vertical wall elements spanning three or more
equally-spaced anchor levels, the vertical wall element can be designed as continuous beams and
13-18
where:
Mmax = Factored maximum flexural moment (kN-m)
p=
Average factored lateral pressure carried by vertical wall facing (kPa)
L=
Horizontal spacing between vertical wall elements (m)
x=
Height of vertical element between anchors or supports (m)
Mmax is considered to be the positive design moment between anchors and the negative design
moment at the anchors. If the variation in lateral pressure with depth is large, moment diagrams
should be constructed to provide improved accuracy.
13.3.6.2 Permanent Facing (A11.8.5.3)
The maximum spacing between vertical wall elements should be determined based on the relative
stiffness of the vertical elements and the type and condition of soil to be supported. The horizontal
spacing between vertical elements typically varies between 2 to 3 meters. If timber facing is
specified, the timber should be stress-grade pressure treated.
Facing can be designed assuming simple support between elements, with or without soil arching, or
assuming the facing behaves as a continuous support across several elements. Based on these
assumptions, the value of the maximum factored flexural moment, Mmax, on a unit (1 meter) width or
height of facing may be determined as:
M max = 0.125 pL
M max = 0.083 pL
Continuous:
M max = 0.100 pL
where:
Mmax = Factored maximum flexural moment on facing (kN-m/m)
p=
Average factored lateral pressure acting on wall facing (kPa)
L=
Horizontal spacing between vertical wall elements (m)
13-19
If the variation in lateral pressure with depth is large, moment diagrams should be constructed to
provide improved accuracy.
Eq. 13-18 is applicable for simply supported facing behind which the soil will not arch between
vertical supports (e.g., in soft cohesive soils or for rigid concrete facing placed tightly against the
in-place soil). Eq. 13-19 is applicable for simply supported facing behind which the soil will arch
between vertical supports (e.g., in granular or stiff cohesive soils with flexible facing or rigid facing
behind which there is sufficient space to permit the in-place soil to arch). Eq. 13-20 is applicable for
facing which is continuous over several vertical supports (e.g., reinforced shotcrete).
Temporary timber lagging thickness is generally based on empirical relationships with excavation
depth (Goldberg, et al., 1975) rather than on a rigorous structural design.
13.2.7 Other Considerations
13.3.7.1 Corrosion (A11.8.7)
Except for temporary walls, the designer must provide adequate protection for anchor and anchor
hardware against aggressive ground or the presence of stray currents which can corrode and lead to
loss of service life for the primary resistance element of anchored walls. Typically, corrosion
resistance for anchors is provided by encapsulation by electrostatically-applied, resin-bonded epoxy
coatings, grease and grout. For anchor hardware, protection is provided by electrostatically-applied,
resin-bonded epoxy coatings or by use of a sacrificial thickness.
13.3.7.2 Drainage (A11.8.9)
As with other wall types, the lateral load on an anchored wall is affected by the shear strength and
unit weight of the retained ground, and by the presence of water in the retained ground. Because
anchored walls are typically constructed from the top down and support in-situ soil and rock
materials however, the designer should consider installing drainage features between the ground and
the facing to minimize permanent water loads on the wall and to preclude uncontrolled seepage
through the permanent wall face. For anchored soldier pile and lagging walls, geocomposite drains
can be installed between the ground the facing to intercept and convey seepage to drainage control
structures. For anchored diaphragm walls, seepage can be controlled by water stops between panels
and continuous pours of high-quality, flowable concrete.
13-20
During the subsurface exploration, it was determined that the foundation soils consist of medium
dense sand to a depth of 13 m below the ground surface, underlain by hard, sandstone bedrock. In
performing the wall design, you can assume the following:
13-21
The anchor derive their capacity wholly within the medium dense sand
Only the end of excavation stage needs to be checked for this problem (Note:
For a complete design, each stage of the excavation should be checked)
Objective: To demonstrate the procedure for anchored retaining wall design by LRFD
Approach: To perform the anchor wall design, you should take the following steps:
Compute and tabulate the unfactored loads and moments required for design
at the applicable limits states
Determine and tabulate the factored loads and moments required for design at
the applicable limits states
Estimate wall movements at the Service I Limit State
Evaluate the global stability of the excavation
Evaluate the earth pressures on the excavation support system
Estimate the required anchor bond zone length
Determine the required soldier pile section using the applicable factored
loads and simplified structural analysis procedures
Determine the required section for the timber lagging
Evaluate the vertical stability of the wall
Solution:
Step 1: Calculate Loads
The first step in determining the loading conditions is to assess the earth pressures on the wall. The
apparent earth pressure diagram shown in Figure 13-4 is selected in accordance with
recommendations made by Peck, et. al (1974) and AASHTO (1997a, 1997b) for wall systems with
multiple levels of support in cohesionless soils (A3.11.5.6). A live load surcharge of 12 kPa
(equivalent to a uniform soil surcharge of about 0.63 m) is used for design. The calculations for
development of the lateral earth pressure and lateral pressure due to vehicular live load surcharge are
shown below.
Unfactored Apparent Earth Pressure (EH):
Pa = 0.65 ka s H = 0.65 tan2(45E - f/2) s H (A3.11.5.6-1)
E/2) (
kN/m3) (
m)
Pa = 0.65 tan2 (45E kPa
Pa =
Unfactored Surcharge Pressure (LS):
p = ka qs = tan2 (45E -
E/2) (
) kPa =
13-22
kPa
(A3.11.6.1-1)
The apparent earth pressure and surcharge pressure diagrams are then distributed to the various
members of the system assuming tributary areas extending equidistant between the support levels as
shown below.
The unfactored horizontal force to be supported at Level I is the sum of the apparent earth and
vehicular surcharge lateral pressures from the ground surface down to a depth of 3.5 m (zI = 3.5 m)
as shown below:
Unfactored horizontal load from apparent earth pressure diagram:
kPa)(
m) =
kN/m
P1 = Pa zI = (
Unfactored horizontal load from vehicular surcharge lateral pressure:
kPa)(
m) =
kN/m
P2 = p zI = (
Total unfactored horizontal load:
kN/m +
PI = P1 + P2 =
kN/m =
kN/m
The load factors from Chapter 4 (Tables 4-10 and 4-11) for horizontal earth pressure, EH, and live
load surcharge, LS, are:
EH =
LS =
Therefore, the total factored horizontal load is:
PI = EH P1+LS P2 =
kN/m) +
kN/m) =
kN/m
The unfactored horizontal force to be supported at Level II is the sum of the apparent earth and
vehicular surcharge lateral pressures from a depth of 3.5 m to a depth of 6.5 m (zII = 3.0 m) as shown
below:
Unfactored horizontal load from apparent earth pressure diagram:
P1 = Pa zII = (
kPa)(
m) =
kN/m
Unfactored horizontal load from vehicular surcharge lateral pressure:
kPa)(
m) =
kN/m
P2 = p zII = (
Total unfactored horizontal load:
kN/m +
PII = P1 + P2 =
kN/m =
kN/m
Applying the load factors for EH and LS, the total factored horizontal load is:
PII = EH P1+LS P2 =
kN/m) +
kN/m) =
kN/m
The unfactored horizontal force to be supported at the base of the wall for design of the soldier pile
embedment is the sum of the apparent earth and vehicular surcharge lateral pressures from a depth of
6.5 m to the base of the wall (zIII = 1.5 m) as shown below:
13-23
kN/m =
kN/m
Applying the load factors for EH and LS, the total factored horizontal load at the wall base is:
PIII = EH P1+LS P2 =
kN/m) +
kN/m) =
kN/m
Load
Level I Anchor
Level II Anchor
Excavation Base
Unfactored
EH
(kN/m)
Unfactored
LS
(kN/m)
Unfactored
Horizontal
Load
(kN/m)
Factored
Horizontal
Load
(kN/m)
In addition to anchor design and soldier pile embedment, it is necessary to consider load distribution
and load factoring for structural design of the soldier piles and timber lagging. The distribution and
factoring of the apparent earth pressure and vehicular surcharge lateral pressures to the anchors
provides the necessary loading information for analysis of the vertical stability of each pile.
Step 2: Estimate Lateral Wall Deflection and Settlement Profile Behind Wall
Figure 13-3 shows the typical settlement profiles behind a wall of this type assuming good
construction practices are followed. For this problem, it is evident from Curve I in Figure 13-3 that
the maximum settlement at the wall will be about 0.3 percent of the excavation depth or 0.024 m.
The lateral wall deformations should be similar in magnitude to this settlement. It is assumed that
this level of deformation is acceptable for this problem.
13-24
An analysis of the global stability of the excavation is not presented herein, but would be checked
using a limit equilibrium analysis (A11.8.4.3) and is assumed to be adequate for this example.
Step 4: Preliminary Geotechnical Design of Anchors
As with ASD, design of an anchored wall includes preliminary estimation of anchor capacity to
permit selection of a reasonable anchor spacing and design of facing and other structural
components of the wall. For this problem, you can assume that the anchors will develop their
capacity within the medium dense sand deposit. In accordance with AASHTO (1997a) and FHWA
(Cheney, 1988) recommendations presumptive ultimate anchor capacity, Qa, as provided in Table
13-7 for estimating the length of the bond zone in sand for each anchor. The bond zone lengths are
estimated as shown in the following calculations. The adequacy of anchor capacity is confirmed
during construction in the same manner as for ASD, by means of proof and performance tests. For
the purposes of this example, the structural capacity of the anchor strands is assumed to be adequate.
Table 13-7 (A11.8.4.2-1 Excerpt)
Ultimate Unit Resistance of Anchors in Soil
13-25
Q = PaI s/(cos i) = (
Q = PI s/(cos i) = (
kN/m)(
kN/m)(
m)/(cos
E) =
m)/(cos
E) =
kN
kN
From Table 13-4, the resistance factor for anchor pullout in sand based on correlation with Standard
Penetration Test is:
= _______
For preliminary design, the ultimate anchor capacity, Qa, for estimation of anchor bond length is:
Qa =
kN/m
(Eq. 13-13)
)(
kN/m) =
Anchor Level II
Using the anchor inclination, i, of 35E from the horizontal, the unfactored and factored horizontal
loads for Anchor Level II from Step 1, the total unfactored and factored anchor loads for a spacing,
s, of 2.4 meters along the length of the wall between anchors are:
Q = PaII s/(cos i) = (
Q = PII s/(cos i) = (
kN/m)(
kN/m)(
m)/(cos
E) =
m)/(cos
E) =
Then using Eq. 13-13, the estimated anchor bond length, L, is:
L $ Q/ Qa (Eq. 13-13)
L=
kN/(
)(
kN/m) =
13-26
kN
kN
The table below summarizes the bond zone lengths for Anchors at Levels I and II.
Anchor Level
LRFD
Bond Zone Length
(m)
Level I
Level II
Step 5: Soldier Pile Embedment
The passive resistance of the embedded section of soldier pile must resist the lateral earth load at the
excavation base. From Step 1, the unfactored and factored horizontal loads at the excavation base
are 42.9 kN/m and 65.5 kN/m. For a pile spacing, s, of 2.4 m, the total unfactored and factored loads
on each embedded pile section are:
Q = PaIII s = (42.9 kN/m)(2.4 m) = 103.0 kN
Q = PIII s = (65.5 kN/m)(2.4 m) = 157.2 kN
As depicted in Figure 13-5, the net passive resistance of each embedded pile is obtained assuming
that passive pressure on the front of the pile acts over a width equal to three times the pile flange
width and that the active pressure on the back of the pile acts over the actual pile flange width.
From Step 1, the active earth pressure coefficient is:
ka = tan2 (45E - /2) = tan2 (45E - 36E/2) = 0.26
(Note that ka = 0.26 is also obtained using AASHTO Eq. A3.11.5.3-1 and A3.11.5.3-2 for a vertical
wall, level backfill and no wall friction.)
The passive earth pressure coefficient, kp, for a vertical wall, horizontal ground surface and no wall
friction ( = 0E) and f = 36E is obtained from AASHTO (1997a) Figure A3.11.5.4-1 as:
kp = (0.36)(11.6) = 4.18
From Table 13-4, the resistance factor, , for passive resistance of embedded vertical wall elements
in soil is 0.60.
13-27
Figure 13-5
Lateral Earth Pressure Diagram
for Soldier Pile Embedment into Cohesionless Soil
From Step 1, the load factor for active earth pressure, EH, is 1.50 and for live load surcharge, LS, is
1.75. For a 310 mm HP soldier pile (b = 0.31 m), the factored active pressure from Figure 13-5 is:
Pa2 = EH 0.5 ka s D2 b + (EH ka s H + LS p) D b
Pa2 = 1.5(0.5)(0.26)(18.865 kN/m3)D2(0.31 m)
+ [1.5(0.26)(18.865 kN/m3)(8 m) + 1.75(3.1 kPa)]D(0.31 m)
Pa2 = 1.14 D2 + 19.9 D
and the factored passive pressure from Figure 13-5 is:
Pp = (1.5 kp s D2 b)
Pp = 0.60 (1.5)(4.18)(18.865 kN/m3)(D2)(0.31 m)
Pp = 22.0 D2
Each soldier pile must resist the factored lateral load, Q, computed as the unit factored horizontal
load at the excavation base from Step 1 times the pile spacing, s, of 2.4 m so that:
Q = (65.5 kN/m)(2.4 m) = 157.2 kN
13-28
Equating the net factored passive resistance of the embedded vertical element to the load at the
excavation base:
Q = Pp - Pa2
157.2 kN = 22.0 D2 - 1.14 D2 - 19.9 D
or
0 = D2 - 0.95 D - 7.54
The soldier piles must support the vertical component of anchor load by side and base resistance of
the soldier pile within the depth of embedment.
Tip Resistance
The nominal unit tip resistance, qp, of a pile in cohesionless soil is:
qp =
38 Ncorr Db
< ql
D
(A10.7.3.4.2a-1)
for which:
1920
N
Ncorr = 0.77 log10
v
(A10.7.3.4.2b-1)
and:
Ncorr =
N=
D=
Db =
qR =
N v =
Representative SPT blow count near the pile tip corrected for Nv (blows/0.3 m)
Measured SPT blow count (blows/0.3 m)
Pile width or diameter (m)
Depth of penetration in bearing stratum (m)
Limiting tip resistance taken as 400 Ncorr for sands and 300 Ncorr for silt (kPa)
Effective overburden pressure at pile tip (kPa)
For a pile tip at 3.3 m below the excavation base (or 3.3 m + 8.0 = 11.3 m) below the original ground
surface:
1920
20 = 14.7
Ncorr = 0.77 log10
2
(18.865
kN/
)(11.3
m)
m
13-29
and
qp =
(38)(14.7)(3.3 m)
= 5946 kPa < ql = (400)(14.7) = 5880 kPa
0.31 m
kN
Side Resistance
The nominal unit side resistance, qs, of an H-pile in cohesionless soil is:
qs = 0.96 N
where:
qs = Unit skin friction for driven piles (kPa)
N = Average (uncorrected) SPT blow count along pile shaft (blows/0.3 m)
from which:
qs = 0.96 (20) = 19.2 kPa
For a typical pile perimeter area of 2 (0.30 m) + 2 (0.31 m) = 1.22 m2/m for an HP 310 pile section,
the nominal side resistance for a 3.3 m embedment is:
Qs = qs As
Qs = (
(Eq. 9-20)
kPa)(
m 2/m)(
m) =
kN
QR = Qn = qp Qp + qs Qs
(Eq. 9-18)
The resistance factor for a single pile for capacity estimation based directly on SPT results is
obtained from Table 9-5 in Chapter 9 as:
qp = qs = =
from which:
13-30
QR =
QR =
kN) +
kN)
kN
Factored Load
The total factored vertical load on an individual soldier pile is calculated as follows:
kN
kN
kN
For the final LRFD check, the factored resistance effect, QR, must be greater than or equal to the
factored load effect, Q using Eq. 13-7, as shown below.
Q # QR
kN # or >
kN OK or NG (Select One)
Therefore, a soldier pile embedment of 3.3 m is insufficient to adequately resist the vertical
components of the anchor forces. By inspection, the additional skin friction available through the
remaining 1.7 m depth of soil would not increase the resistance sufficiently to resist the anchor
forces. Therefore, drive soldier piles to the top of rock at a depth of 5 m below the excavation base.
For the purpose of this example, it is presumed that the pile bearing resistance on the sandstone
bedrock is adequate.
Step 7: Soldier Pile Structural Design
The soldier piles considered in this problem are subjected to flexure due to the imposed earth
pressures and to axial load from the two levels of anchors. Therefore, the structural analysis of these
soldier piles must consider the interaction effects of combined axial load and flexure. As specified
in the LRFD Specification (AASHTO, 1997a) Article A6.9.2.2, Eq. A6.9.2.2-2 applies for Pu/Pr $
0.2:
Pu + 8 M ux + M uy 1
Pr 9 M rx M ry
(A6.9.2.2-2)
where:
Pu =
Pr =
Mux =
Mrx =
Muy =
Mry =
13-31
The AASHTO LRFD Specification (1997a) provides simplified design equations for flexural
analysis of the vertical supporting elements dependent upon the average factored lateral pressure,
distance between vertical elements and height of the section of vertical element being considered.
Because the load and resistance factors in LRFD are calibrated for this approach, this example
problem uses these simplified design equations. Using these simplified equations, the maximum
moment, Mmax, will occur below the bottommost row of anchors as:
Mmax = 0.333 p L x2
(AC11.8.5.2-3)
where:
p=
L=
x=
In the AASHTO LRFD Specification (1997a), the average lateral pressure is taken as the average
factored lateral pressure and the maximum flexural moment is the maximum factored flexural
moment. The factored maximum moment below the Level II anchors is:
Mmax = 0.333 p L x2
(AC11.8.5.2-3)
where:
p = (1.5)(25.5 kPa) + (1.75)(3.1 kPa) = 43.7 kPa
Mmax = 0.333 (43.7 kPa)(2.4 m)(3 m)2 = 314.3 kN-m
From Section 6 of the AASHTO LRFD Specification (Steel Structures), structural design of the
soldier piles must also include consideration of the interaction effects from combined axial load and
flexure. From the calculations in Step 6, the factored axial load is 477.0 kN.
From the AASHTO LRFD Specification (1997a), Article 6.5.4.2, the resistance factors for flexure
and compression in LRFD are:
f = 1.00 (flexure)
c = 0.90 (compression)
First trying an HP 310 H 94 pile, the nominal moment capacity, Mn, of the HP soldier piles is
determined from the AASHTO LRFD Specification (1997a) Articles 6.12.2.2.1, 6.10.6.2 and
6.10.5.2.1-1 and Metric Properties of Structural Shapes (AISC, 1992) as:
Mn = Mp
Mn = Zx Fy
Mn = (14.5 H 10-4 m3)(250 000 kPa)
Mn = 362.5 kN-m
where:
13-32
K l Fy
=
= 0.045
rs E
which is applicable for # 2.25, where is the normalized column slenderness factor defined as:
for:
As =
Fy =
E=
K=
R=
rs =
Through substitution back into the equation for the nominal axial capacity we find that:
Pn = 0.660.045 Fy As = 0.98 Fy As = 0.98 (250 000 kPa)(11.9 H 10-3 m2) = 2916 kN
Then the factored axial capacity, Pr is:
Pr = c Pn = 0.90 Pn = 0.90 (2916 kN) = 2624 kN
At this point, the axial resistance component of the interaction equation can be evaluated using:
Pu = 477.0 kN = 0.18 < 0.2
Pr 2624 kN
For this problem, it is reasonable to expect that there will be no significant bending in the plane of
the wall and therefore we can assume that Muy is zero. The factored flexural resistance about the x
axis is:
Mrx = Mn = (1.00) 362.5 kN-m = 362.5 kN-m
13-33
(A6.9.2.2-2)
Therefore, this soldier pile provides a section which is adequate for the design loading conditions for
LRFD. In addition to this combined stress check, the slenderness of the soldier beam should be
checked to assure that buckling will not control design. For this example, slenderness is not a
controlling factor.
Step 8: Timber Lagging Design
As for the structural design of the soldier piles, in this problem you can use the simplified design
equations to select a timber lagging section. The maximum factored moment is:
Mmax = 0.083 p L2
(AC11.8.5.3-2)
where:
p = Average factored lateral pressure = (1.50)(25.5 kPa) + (1.75)(3.1 kPa) = 43.7 kPa
L = Lagging span =2.4 m - 0.306 m = 2.094 m (Note: Pile flange width = 306 mm)
Substituting into Eq. AC11.8.5.3-2, the maximum factored moment is:
Mmax = 0.083 (43.7 kPa)(2.094 m)2 = (15.9 kN-m/m)
The factored flexural capacity, Mr, of the lagging is determined by multiplying the resistance factor
of 0.85 for flexure of wood from Article 8.5.2.2 of the AASHTO LRFD Specification (1997a) by the
nominal flexural resistance, Mn, as follows:
Mr = Mn = 0.85 Mn
(A8.6.1-1)
where the nominal flexural resistance of wood, Mn, from Article 8.6.2 of the AASHTO LRFD
Specification (1997a) is:
Mn = Fb S Cs
(A8.6.2-1)
and:
13-34
Fb =
S=
Cs =
Evaluation of the size effect factor and the specified resistance in flexure both require calculations
which are not presented herein, although the results shown are consistent with the material and
geometry specified. For this problem, use Select Structural Southern Pine 102 mm thick and 254
mm wide for the timber lagging. For this material, Fb = 29 000 kPa, S = 4.4 H 10-4 m3, Cs = 1.0 and
the nominal flexural resistance of the lagging is:.
Mn = (29 000 kPa)(4.4 H 10-4 m3)(1.0) = 12.8 kN-m
The factored flexural resistance, Mr, is then:
Mr = 0.85 (12.8 kN-m) = 10.8 kN-m
The maximum factored moment on a 254 mm wide lagging section is:
Mmax = (15.9 kN-m/m) (0.254 m) = 4.0 kN-m
Therefore, the lagging is acceptable.
As notted in Section 13.3.6.2, the thickness of timber lagging for temporary support is generally
selected based on empirical relationships with excavation depth (Goldberg, et al., 1975). For an
excavation depth of 8 m, a lagging thickness of 76 mm to 102 mm is prescribed for soldier piles
spaced at 2.5 m in medium dense sand. It is apparent from the above calculation that thinner lagging
would be acceptable.
Summary
This example problem illustrates the design of an anchored soldier pile and
lagging retaining wall by LRFD
The wall consists of driven steel H-piles supported by two rows of anchors
and spanned using timber lagging
13-35
13-36
CHAPTER 14
MECHANICALLY-STABILIZED EARTH (MSE) WALL DESIGN
14.1 Introduction
For both Allowable Stress Design (ASD) and Load and Resistance Factor Design (LRFD), the
design of mechanically-stabilized earth (MSE) walls requires consideration of geotechnical capacity,
overall stability, structural capacity and deformation limits. The design processes therefore require
both establishment of criteria for acceptable stress and deformation levels, and comparison of these
criteria with stress and deformation levels estimated from the design. This chapter:
y
Describes primary differences between MSE wall design by LRFD and ASD
Briefly summarizes methods commonly used for evaluating the external and
internal stability of MSE walls
gravity walls
y
Requirements for structural resistance within the reinforced soil mass and the
panel units, and for the development of reinforcement extending beyond
assumed failure zones
The 1997 Interims to the AASHTO ASD Bridge Design Specifications (AASHTO, 1997b) incorporate
a number of revisions to the design of MSE walls. The revisions are based primarily on the results
of FHWA Demonstration Project 82, AMechanically Stabilized Earth Walls and Reinforced Slopes
Design and Construction Guidelines@ (Elias and Christopher, 1996) and generally include the
following:
1.
2.
3.
4.
5.
6.
The use of minimum average roll values (MARV) of ultimate strength from
wide width tests for Tult.
7.
8.
The current AASHTO LRFD Specification for MSE walls (AASHTO, 1997a) is based on and
calibrated to the former AASHTO ASD Specification (AASHTO, 1996). Therefore, discussions
related to ASD criteria for MSE wall design in this chapter correspond to the AASHTO ASD
Specifications and the FHWA recommendations on which the specifications are based. Efforts
are currently underway as part of NCHRP 20-7, Task 88, to revise the AASHTO LRFD
Specifications for MSE walls to reflect the changes in the 1997 AASHTO ASD Interim
Specifications (AASHTO, 1997b) as developed based on the results of FHWA Demonstration
Project 82 (Elias and Christopher, 1996), and to develop resistance factors for MSE wall design
using a reliability-based calibration and the revised design procedures.
Deformation
The design of MSE walls by ASD can be controlled by deformation or settlement considerations.
Thus, the design of MSE walls by ASD requires estimation of foundation settlement and wall
movements under the applied loads, and comparison of estimated settlement and wall movement
with deformation criteria using the following:
i n
(Eq. 14-1)
where:
i =
n =
The actual wall settlement will be controlled primarily by the wall height and foundation
compressibility. Tolerable movement criteria are usually a function of the type of structure, and for
14-3
abutments and other earth retaining structures, criteria for lateral deflection or tilt may govern the
design, as described in Chapter 11. Acceptable vertical and horizontal wall displacement criteria for
MSE walls are discussed in Section 14.3.1.
Geotechnical/Structural Resistance
Following selection of a preliminary width of the reinforced zone based on empirical relationships
with the height of wall, the ultimate bearing capacity and sliding resistance, Rn, are estimated by
available theoretical or semi-empirical procedures as described in Chapter 8. The suitability of the
design with respect to bearing capacity and sliding are then evaluated by determining the allowable
vertical and horizontal load components, Qall, using:
Q Q all =
Rn
FS
(Eq. 14-2)
where:
Q=
Qall =
Rn =
FS =
The required FS with respect to bearing capacity and sliding are generally specified by the
governing agency, and may be constant or variable. The AASHTO ASD (AASHTO, 1996) FS
criteria for bearing capacity, sliding, overturning and overall stability are shown in Table 14-1. The
AASHTO ASD FS values assume that designs are based on soil and rock properties determined
through appropriate field and laboratory testing and not presumptive properties.
In addition to bearing capacity, sliding failure and overturning failure, the location of the bearing
pressure resultant of MSE walls is checked with respect to the centroid of the reinforced soil zone.
In ASD, the location of the bearing pressure resultant must be maintained within B/6 of the center of
the reinforced soil zone.
The overall stability of the retained earth and foundation must also be evaluated using procedures
described in Section 8.2.1.
Table 14-1
Factors of Safety for External Stability of MSE Walls
(AASHTO, 1996)
Required Minimum
Failure Condition
Factor of Safety
(FS)
Bearing Capacity
2.0 to 2.5
Sliding Resistance
1.5
Overturning
2.0
Overall Stability
Refer to Chapter 8
14-4
The internal stability or structural capacity of the reinforcements and connections with the wall
facing against rupture, and the pullout resistance of the reinforcements must also be evaluated. The
structural capacity of the reinforcement is checked as follows:
all
A = Pall
(Eq. 14-3)
where:
all = Allowable tensile stress in soil reinforcement (kPa)
Pall = Allowable tensile structural capacity of reinforcement (kN)
A = Cross-sectional area of soil reinforcement less any sacrificial thickness (m2)
Table 14-2 summarizes the maximum allowable tensile force criteria in AASHTO (1996) for
evaluating the internal stability of reinforcements and facing connections. Because polymeric
materials exhibit time- and temperature-dependent creep behavior, these reinforcements are
evaluated at a maximum load level obtained from creep tests (i.e., Limit State) and at a load
representative of a total strain level of 5 percent (i.e., Serviceability State). (As noted previously, the
report on Demonstration Project 82 (Elias and Christopher, 1996) recommends elimination of the
Serviceability Limit State tension check for polymeric reinforcements.) In addition, the maximum
allowable tensile load of polymeric reinforcements at both the Limit and Serviceability Limit States
are adjusted to account for the effects of environmental and aging losses, FD, and for the effects of
construction damage, FC. The overall factor of safety, FS, accounts for uncertainties in structure
geometry, engineering properties of backfill soils, variations in reinforcement manufacturing, and
externally applied loads.
Table 14-2
ASD Criteria for Internal Stability Evaluation of Reinforcements for MSE Walls
(AASHTO, 1996)
Reinforcement Type
Steel
y Strip
y Grid
Polymeric
y Limit State
y Serviceability State (2)
(1)
Maximum Allowable
Tensile Load (1)
(kN)
0.55 As Fy
0.48 As Fy
TR/FC FD FS
Tw/FC FD
(2)
FD = Durability FS ranging from 1.1 to 2.0, and taken as 2.0 w/o tests (dim)
FS = FS to account for load, geometry, and material property uncertainty (dim)
Recent FHWA Demonstration Project 82 (Elias and Christopher, 1996) recommends elimination
of the Serviceability Limit State check on polymeric reinforcement tension.
The value of FS is taken as 1.78 for the most recent version of AASHTO (1996) and 1.50 for earlier
versions of AASHTO (1991) and FHWA Demonstration Project 82.
Pullout resistance is evaluated using Eq. 14-2. The minimum factor of safety against pullout of
reinforcements is taken as 1.5.
Finally, after all geotechnical deformation and capacity criteria are met, the structural design of the
facing is performed using service loads and allowable stresses (AASHTO, 1996), or by LFD or
LRFD (AASHTO, 1996) or ACI.
14.2.3 LRFD Summary
Whereas ASD considers all uncertainty in the applied loads and ultimate geotechnical or structural
capacity in factors of safety or allowable stresses, LRFD separates the variability of these design
components by applying load and resistance factors to the load and material capacity, respectively.
When properly developed and applied, the LRFD approach can provide a consistent level of safety
for the design of all structure components. Thus, the probability that a structure component will
fail or perform unacceptably is no different than any other component. For design, the resistance
and deformation of supporting soil and rock materials and structure components must satisfy the
LRFD equations below. For the Strength and Service Limit States:
Q
i
R n = R r
(Eq. 14-5)
where:
i =
i =
Qi =
=
Rr =
Rn =
i =
n =
Relative to the external stability of the wall and the internal stability of system components, the
suitability of an MSE wall with respect to the geotechnical and structural resistance can be obtained
14-6
Q ult = Q R
(Eq. 14-6)
Pn = PR
(Eq. 14-7)
and
i
where:
i
=
Qult =
QR =
Pn =
Pr =
The load factors and load factor combinations used for design were presented in Chapter 4. In
general, values of i > 1.0 are used to evaluate ultimate ground or structure capacity at the Strength
Limit States, whereas the deformation performance of structures is evaluated at the Service I Limit
State using i = 1.0 (or i = 0.3 for wind loads). In ASD (AASHTO, 1996), values of i = 1.0 (or i =
0.3 for wind loads) are used to evaluate structures for both strength (allowable stress) and
serviceability (deflection). Because the AASHTO LRFD Specification for MSE walls was
developed based on direct calibration with ASD, the results of analyses by LRFD and ASD to
evaluate external wall stability and the internal stability of wall components are similar. Because
the calibration used some discretion in the selection of safety factors and resulted in the selection
of average resistance factors for the range of possible average load factors, some minor
differences in walls designed by ASD and LRFD will occur due to variations in actual design
loading conditions. For foundation and wall deformations, ASD and LRFD are identical.
When using Eq. 14-4 for MSE wall design at the Strength Limit States, the following values of can
normally be used:
y
D = R = 1.00
Determination of the operational importance of a structure (such as a bridge) is made by the facility
owner as described in Chapter 4. Where an MSE wall supports a structure, the appropriate value of
I is then applied throughout the superstructure design by the structural engineer. The value of I
selected by the superstructure designer should then be applied in the MSE wall design. For the
purpose of this chapter, the value of I is assumed equal to 1.0.
14-7
When using Eq. 14-5 to evaluate an MSE wall at any Service Limit State, D, R, and I = 1.0.
Values of load factor and load factor combinations for each applicable limit state must be developed
using the guidelines described in Chapters 3 and 4 and Section 14.2.3.1, and loads should be
developed as described in Chapter 4. The geotechnical design of MSE retaining walls with respect
to external stability generally follows the LRFD approach for design of conventional retaining walls
described in Chapter 11. The geotechnical and structural design of MSE walls with respect to
internal design are described in Section 14.3. The ultimate resistance, Rn, should be determined for
each type of resistance (e.g., bearing criteria, sliding, overall stability, and pullout and rupture of
reinforcements) and the reinforced soil zone should be checked for overturning based on load
eccentricity criteria as described in Section 8.3.4.
Values of # 1.0 are applied when evaluating geotechnical resistance for any strength limit state
using Eq. 14-4. Currently the value of = 1.0 is applied when evaluating an MSE wall for any
service limit state using Eq. 14-5, except for the evaluation of tensile resistance of polymeric
reinforcement, for which a < 1.0 is considered and Eq. 14-4 is applied.
14.2.3.1 Limit States (A11.5)
The design of MSE walls using LRFD requires evaluation of the suitability of the external stability
of the wall, the internal stability of wall components and wall movements at various Performance
Limit States (i.e., applicable Strength Limit States and the Service I Limit State). The selection of a
Strength Limit State(s) depends on the type of applied loading (e.g., Strength I for design vehicle
loading without wind or Strength II for permit vehicle loading). The design considerations which
must be evaluated for MSE walls designed at the Strength and Service I Limit States are summarized
in Table 14-3. As conditions warrant, it may also be necessary to evaluate wall and foundation
performance at other limit states (e.g., Extreme Event I for loading from earthquakes).
Table 14-3
Strength and Service Limit States for Design of MSE Walls
Performance Limit
Bearing Resistance
Sliding Resistance
Location of Base Resultant Force
Overall Stability
Rupture of Reinforcing Elements
Pullout of Reinforcing Elements
Structural Resistance of Face Elements
Structural Resistance of Wall Connections
Settlement and Horizontal Movement
(1)
Checked for polymeric reinforcements only.
14-8
Strength
Limit State(s)
U
U
U
U
U
U
U
U
Service I
Limit State
U(1)
all
n
(Eq. 14-8)
where:
n = Ultimate tensile capacity, Fy, (kPa)
all = Allowable tensile capacity from ASD (kPa)
= Assumed average load factor (dim)
The resistance factors in Table 14-4 were developed assuming an allowable stress of 0.55 Fy for strip
reinforcements and 0.48 Fy for grid reinforcements and connectors for yielding of the gross section
less sacrificial area. The reduction in resistance factors for fracture is consistent with a reduction of
about 15 to 20 percent in the allowable tensile stress as required by AASHTO (1996) for steel
members with holes for bolts or rivets. The average load factor, , used to develop these resistance
factors was approximately 1.6.
The resistance factors for tensile resistance of geosynthetic or polymeric reinforcements for tensile
resistance determined by a wide-width tensile test were developed using the following relationships:
14-9
Table 14-4
Resistance Factors for Strength Limit State of MSE Walls
(Modified after AASHTO, 1997a)
Condition
Bearing Resistance
Sliding (Geosynthetic on soil)
Sliding (Soil on soil)
Strip reinforcements
y Yielding of gross section less sacrificial area
y Fracture of net section less sacrificial area
Tensile Resistance Grid reinforcements
of Metallic
y Yield of gross section less sacrificial area
Reinforcement
y Fracture of net section less sacrificial area
Strength Limit
State Tensile
Resistance of
Polymeric
Reinforcements
Resistance
Factor(1)
See Chapter 8
1.00
1.00
0.85
0.70
0.75
0.60
Connectors
y Yielding of gross section less sacrificial area
y Fracture of net section less sacrificial area
Applied to TR(2)
0.75
0.60
0.25
Applied to Tn(3)
y Polyethylene
y Polypropylene
y Polyester
y Polyamide
y High Density Polyethylene
Applied to Tw(2)
0.05
0.05
0.11
0.09
0.09
0.25
Tall
T
all
Tn
T5
(Eq. 14-9)
and
14-10
Tall =
T CRF
Tn CRF
5
FD FC FS FD FC
(Eq. 14-10)
where:
Tall =
Tn =
T5 =
FS =
CRF
0.20
0.20
0.40
0.35
0.35
The long-term tensile strengths are estimated from a wide-width tensile test as TR = (CRF) Tn and Tw
= (CRF) T5.
Combining Eq. 14-9 and 14-10 for application to Tn:
=
CRF
FC FD FS
(Eq. 14-11)
CRF
FC FD
(Eq. 14-12)
The average load factor, , used to develop the resistance factors for polymeric reinforcements was
approximately 1.6 for the strength limit and 1.0 for the service limit state.
Where the Strength Limit State Tensile resistance, TR, of polymeric reinforcements is obtained from
laboratory creep tests, no reduction is made in the laboratory strength for creep (i.e., CRF = 1.0), and
14-11
the resistance factor, , applied to TR is equal to 0.25 for all reinforcement types.
The resistance factors for the Serviceability Limit State tensile resistance of polymeric
reinforcements obtained from laboratory creep tests is obtained from Eq. 14-11 and Eq. 14-12
without reductions for creep (CRF).
The resistance factor, , for reinforcement pullout in Table 14-4 provides an equivalent factor of
safety of 1.5 for an average load factor, , of approximately 1.35 (because the live load surcharge is
not included over the reinforced fill for pullout evaluation), based on the following:
=
FS
(Eq.14-13)
The resistance factors for tensile strength of polymeric reinforcements incorporate factors for the
effects of creep, aging, environmental losses and construction damage. The resistance factors for the
Strength Limit State in the table include default values of FC = 2.0 for the effects of construction
damage and FD = 2.0 for the effects of post-construction environmental and aging strength losses,
generally consistent with Table 14-2 for ASD. The LRFD Specification permits use of higher
multipliers to increase the resistance factor if product-specific data is available that indicates that the
effects of construction damage and/or environmental deterioration and aging are less severe than
assumed in developing the table.
FHWA Demonstration Project 82 (Elias and Christopher, 1996) recommend a total default value of
(FC FD)/CRF equal to 7 for preliminary design or applications not having severe consequences in
the event of poor performance or failure. This default value would result in reinforcement tensile
strength resistance factors of approximately 0.15 on Tn and T5 for all polymeric reinforcement types,
compared to a range of 0.05 to 0.11 in Table 14-4. Demonstration Project 82 also recommends
elimination of the Service Limit State tensile resistance evaluation for polymeric reinforcements.
14.2.3.3 Comparison of Wall Design Using LRFD and ASD
The differences between the design of an MSE retaining wall by ASD and LRFD relate to
differences in the foundation design as discussed in Chapter 8, and minor differences related to
selection of average safety factors and resistance factors as described in Section 14.2.
Other aspects of MSE wall design such as identifying special considerations (e.g., potential for loss
of support through scour), developing a design foundation and retained soil profile and determining
requirements for construction control are inherent aspects of the design process required for both
LRFD and ASD.
14.2.3.4 Modification of Resistance Factors
Resistance factors for MSE wall design using LRFD were developed based on direct calibration with
ASD. In general, resistance factors for geosynthetic reinforcements were developed using default
values for CRF, FD and FC. More appropriate resistance factors can be developed using productspecific data with regard to creep characteristics, durability and construction damage. Resistance
factors for various aspects of MSE wall design can be modified for product-specific data, differing
14-12
load factors and/or different safety factors using Eq. 14-11 and 14-12, and the procedures
summarized in Chapters 7 and 9.
14.2.4 Summarized Comparison of ASD and LRFD
The process used to develop an MSE wall design using LRFD differs very little from the process
used for ASD. The similarity is illustrated in the parallel flow charts in Figure 14-1. Specific
differences between the methods and other important issues are highlighted in Section 14.3.
Figure 14-1
Generalized Flow Chart for MSE Wall Design by ASD and LRFD
14-13
The tolerable settlement of MSE walls is limited by the longitudinal deformability of the facing and
the ultimate purpose of the structure. Limiting tolerable differential settlement to prevent open
spaces in the joints for systems with panels less than 2.8 m2 in area and a maximum joint width of 19
mm are presented in Table 14-6.
Table 14-6
Relationship Between Joint Width and Limiting Distortion of MSE Wall Facing
(AASHTO, 1996; AASHTO, 1997a)
Joint Width
(mm)
19.0
12.7
6.4
Limiting Vertical
Distortion
1/100
1/200
1/300
Where foundation conditions indicate the potential for large differential settlements over a short
horizontal distance, a vertical full-height slip joint should be used. Tolerable horizontal wall
displacement/alignment criteria are recommended by FHWA (Christopher, et al., 1990), as follows:
y
y
For ribbed or smooth steel reinforcing strips, the ultimate pullout resistance, Pfs, is determined as:
Pfs = f ( s ZA s )
where:
Pfs =
ft =
s =
Z=
As =
In the absence of pullout data for ribbed reinforcement strips in backfill materials conforming to
Division II for MSE backfills, the LRFD Specification imposes the following limitations on the
value of ft:
14-15
ft should not exceed 2.0 at ground level, and can be assumed to decrease
linearly to tanf at (and below) a depth of 6.00 m
For smooth steel reinforcing strips, ft is assumed to be constant at all depths and is determined
using:
f * = tan 0.4
where:
= Soil-reinforcement angle of friction (deg)
For steel grid reinforcing systems with transverse bar spacings of 0.15 m or greater, the generalized
relationship for the nominal pullout capacity, Pfg, is:
Pfg = N p s Z n A b
where:
Np =
s =
Z=
n=
Ab =
14-16
For grid reinforcements with transverse spacing less than 0.15 m, the nominal pullout capacity, Pfg,
is calculated using:
Pfg = 2 w l s Z f d tan f
where:
Pfg =
w=
l=
s =
Z=
fd =
f =
The range of values of fd applicable for various types of steel reinforcements are presented in Table
14-7.
Table 14-7
Coefficient of Resistance for Steel Reinforcements
(AASHTO, 1997a)
Reinforcement Type
fd
Continuous sheets
0.45
Bar mats (transverse spacing = 0.15 m)
0.80
14-17
The appropriate values of fd must be determined experimentally for each grid geometry. For
polymeric reinforcements, Eq. 14-17 is applicable where fd is developed for a range of normal
stresses in accordance with GRI-GG-5 (GRI, 1994). Experimental values of fd may be limited by the
Strength Limit State Tensile Resistance at ultimate per Table 14-4.
For evaluation of pullout, the factored pullout resistance, Pfs or Pfg, is compared to the factored
horizontal force acting on the reinforcement at any level, Pi:
Pi = H H i
where:
Pi =
H =
hi =
The value of hi is the vertical distance from the mid-point between layer i and the next overlying
layer to the mid-point between layer i and the next underlying layer.
FHWA Demonstration Project 82 (Elias and Christopher, 1996) recommends a unified normalized
approach to evaluation of pullout resistance which would result in a modification to Eq. 14-14
through 14-17, but would not appear to substantially affect the magnitude of estimated ultimate
pullout resistance.
14.3.6 Rupture of Reinforcing Elements (A11.9.5.1)
MSE walls must also be evaluated for internal failure by rupture of the reinforcements at any level.
The factored tensile resistance of the reinforcement at each level is compared to the factored
horizontal force, Pi, from Eq. 14-18 using Eq. 14-7.
14.3.7 Horizontal Forces for Internal Stability Calculations (A11.9.5.2)
Traditionally, the computation of horizontal forces on inextensible (metallic) and extensible
(polymeric) reinforcements are treated somewhat differently as follows.
FHWA Demonstration Project 82 (Elias and Christopher, 1996) recommends a modified approach
to the determination of internal forces applicable to walls reinforced with either inextensible or
extensible reinforcements. The simplified approach may result in some differences in the internal
forces from the procedures described below.
Inextensible Reinforcements (A11.9.5.2.2)
The internal stability of structures constructed with metallic strip or grid (i.e., inextensible)
reinforcements is analyzed by considering that the reinforced zone is divided into active and
resistant zones by assuming a bilinear failure surface as shown in Figure 14-3.
14-18
H = p v k
where:
P =
k=
V =
Load factor for vertical earth loads from Table 4-11 in Chapter 4 (dim)
Horizontal pressure coefficient (dim)
Pressure due to the resultant vertical forces at reinforcement level (kPa)
The resultant vertical forces at each reinforcement level are determined considering only the forces
acting at that level and by assuming a uniform pressure distribution over the effective width (B - 2
eB) of the reinforced soil zone.
In the LRFD Specification, MSE wall structures are designed assuming k = ko at H1 above the top of
the leveling pad, and that the value of k decreases linearly to k = ka at a depth of 6.00 m as shown in
Figure 14-3. Below a depth of 6.00 m, k = ka. The earth pressure coefficients ka and ko remain the
same regardless of the external loading conditions. FHWA Demonstration Project 82 recommends a
revised distribution of k with depth in which k is a function of ka and depth, and for which somewhat
greater stresses are predicted.
The value of f used to determine the horizontal force within the reinforced soil zone should not
exceed 34o unless the specific project backfill is tested for frictional strength by triaxial or direct
14-19
shear testing. Live loads must be positioned for extreme force effects within the zone subjected to
live loads following the requirements for surcharge loads described in Chapter 4.
Extensible Reinforcements (A11.9.5.2.3)
The internal stability for MSE structures constructed with polymeric (i.e., extensible) reinforcements
are analyzed using a tie-back wedge approach which assumes that the shear strength of the
reinforced fill is fully mobilized and active lateral earth pressures based on ka, are developed.
Therefore, the assumed failure plane for both horizontal and sloping backfill conditions can be
represented by the Rankine active earth pressure zone defined by a straight line passing through the
wall toe and oriented at an angle of 45o + f'/2 from the horizontal.
The tensile force at each level of reinforcement is a function of the vertical stress induced by gravity,
uniform normal surcharge and active thrust multiplied by ka. Tension in the reinforcement induced
by vertical or horizontal line or point loads should be added by superposition to the tensile forces
induced by the reinforced soil mass and the retained backfill.
The value of ka in the reinforced soil mass is assumed to be independent of all external loads except
sloping fills. If testing of the site-specific select backfill is not available, the LRFD Specification
prescribes that the value of f used to compute the horizontal stress within the reinforced soil mass
not exceed 34o. Where site-specific tests are performed, the soil strength can be evaluated at
residual stress levels.
14.3.8 Design Life (A11.9.8)
The long-term durability of steel and polymeric reinforcements must be considered in the design of
MSE walls to ensure suitable performance throughout the design life of the structure. The structural
design of galvanized steel soil reinforcements and connections is made on the basis of a thickness,
Ec, defined as:
Ec = En Es
where:
Ec =
En =
Es =
Based on the results of performance testing and monitoring of constructed MSE walls, the following
sacrificial thicknesses should be computed for each exposed surface:
y
Loss of galvanizing = 0.015 mm/year for the first 2 years and 0.004 mm/year
thereafter
14-20
Alloys such as aluminum and stainless steel should not be used for
reinforcements
TR - The highest load level at which the log time-creep-strain rate continues
to decrease with time within the required lifetime without either brittle or
ductile failure; or
The effects of aging, chemical and microbiological attack, environmental stress cracking, stress
relaxation, hydrolysis and variations in the manufacturing process, as well as the effects of
construction damage, must be evaluated and extrapolated to the required design life.
14-21
During the subsurface exploration, it was determined that the foundation soils consist predominantly
of hard clay to a depth of 10 m beneath the base of the proposed MSE wall. Bedrock underlies the
hard clay. Field vane shear tests indicate that the hard clay has an average undrained shear strength,
Su, equal to 230 kPa and consolidated undrained triaxial compression testing indicates that the clay
has no effective cohesion intercept, c!f, and an effective stress friction angle, !f, equal to 28 degrees.
Similar triaxial testing of the reinforced backfill and embankment fill material indicates an effective
stress friction angle of 38 degrees. The seasonal high groundwater table is located at a depth of 6 m
below the bottom of the MSE wall.
The leveling pad will be constructed using precast concrete. The wall will be constructed using a
polyester geogrid reinforcement.
Approach: To perform the design of the wall, the following steps are taken:
14-22
Compute and tabulate the unfactored loads and moments required for design
at the applicable limit states
Determine and tabulate the factored loads and moments required for design at
the applicable limit states
Solution:
Step 1: Unfactored Loads
(A) Length of Soil Reinforcement
The minimum soil reinforcement length, L, is the greater of either 70% of the wall height, H,
measured from the leveling pad (i.e., 0.7 x 7.0 m = 4.9 m) or 2.40 m (A11.9.5.1.4). However, for
many problems, longer reinforced lengths are needed. As a preliminary estimate, use L = 5.0 m.
(B) Vertical Earth Pressure (EV)
The weight of the reinforced soil backfill is (A3.5.1):
PEV = H L r
PEV = (7 m)(5.0 m)(17.3 kN/m3)
PEV = 605.5 kN/m length of wall
(C) Live Load Surcharge (LS)
The live load surcharge, LS, is applied where vehicular load is expected to act on the backfill within
a distance equal to the wall height behind the wall. Because LS is applied above the reinforcing
strips, both the vertical and horizontal forces will be considered.
Interpolating from Table 14-8, the equivalent height, heq, of soil for the effect of vehicular loading
from the AASHTO LRFD Specification (1997a) equals 0.71 m for a wall height of 7.0 m.
14-23
Wall Height
(m)
1.5
heq
(m)
1.70
3.0
1.20
6.0
0.76
9.0
0.61
ka =
sin 2 ( + )
sin 2 sin ( )
(A3.11.5.3-1)
in which:
sin ( + )sin ( )
= 1 +
sin ( )sin ( + )
(A3.11.5.3-2)
where:
=
Nominal slope of backfill behind wall (deg) = 0o
=
Angle of wall friction (deg) = 0o
= b =
38
o
=
90 for vertical wall
from which:
ka = 0.238
14-24
The uniform increase in horizontal earth pressure due to a live load surcharge is:
p = (k)(s)(heq) = (ka)(b)(heq)
p = (0.238)(17.3 kN/m3)(0.71 m) = 2.92 kPa
(A3.11.6.2-1)
The resultant of the live load surcharge horizontal (lateral) earth pressure, PLSH, acting on the
reinforced soil mass (assuming a uniform or rectangular distribution) is:
PLSH = (p)(H)
PLSH = 2.92 kPa x 7.00 m = 20.4 kN/m length of wall
The basic earth pressure is assumed to vary linearly with depth in a triangular distribution. The
resultant, Pa, of the triangular distribution can be determined (A3.11.5.1) using:
PEH = PaH = 0.5 (b)(H2)ka
PEH = (0.5)(17.3 kN/m3)(7 m)2(0.238) = 100.9 kN/m length of wall
(E) Summary of Unfactored Loads
Table 14-9
Unfactored Vertical Loads/Moments
Item
V
(kN/m)
Moment Arm
About Toe
(m)
PEV
PLSV
Total
Table 14-10
Unfactored Horizontal Loads/Moments
Item
H
kN/m
Moment Arm
About Toe
(m)
PEH = PaH
PLSH
Total
Step 2: Load Factors
From Tables 4-10 and 4-11, the applicable load combinations and load factors are:
14-25
Table 14-11
Load Factors and Load Combinations
EH (Active)
LS
GROUP
EV
Strength I-a
1.00
1.50
1.75
Strength I-b
1.35
1.50
1.75
Service I
1.00
1.00
1.00
GROUP
PLSV
(kN/m)
PEV
(kN/m)
VTOT =
PEV1 + PLSV
(kN/m)
Unfactored
Strength I-a
Strength I-b
Service I
Table 14-13
Factored Horizontal Loads
GROUP
PLSH
(kN/m)
PEH
(kN/m)
HTOT =
PEH + PLSH
(kN/m)
Unfactored
Strength I-a
Strength I-b
Service I
Table 14-14
Factored Moments from Vertical Loads
MvTOT =
MLSV
MEV
MEV1 + MLSV
GROUP
(kN-m/m)
(kN-m/m)
(kN-m/m)
Unfactored
Strength I-a
Strength I-b
Service I
14-26
Table 14-15
Factored Moments from Horizontal Loads
MhTOT =
MLSH
MEH
MEH + MLS
GROUP
(kN-m/m)
(kN-m/m)
(kN-m/m)
Unfactored
Strength I-a
Strength I-b
Service I
Step 4: MSE Wall Settlement/Lateral Displacement
For this example, it is assumed that the MSE wall will not experience unacceptable settlements or
lateral displacements due to the relative stiffness of the hard clay foundation soil, adequate
construction control and sufficient reinforcement length.
Step 5: Eccentricity
Group/Item
Units
Strength I-a
Table 14-16
Summary for Eccentricity Check
MEV
MhTOT
Xo
PEV
(kN/m)
(kN-m/m)
(kN-m/m)
m
605.5
1513.8
478.8
1.71
eB
m
0.79
quniform
(kPa)
177
Strength I-b
817.4
2043.6
478.8
1.91
0.59
214
Service I
605.5
1513.8
307.2
1.99
0.51
152
where:
Xo = Location of the resultant from toe of wall = (MEV - MhTOT)/PEV
B = Base width = Length of reinforcement strips
eB = Eccentricity = B/2 - Xo = 2.5 m - Xo
quniform = PEV/(B - 2 eB) = PEV/2X0
The location of the resultant must be in the middle half of the base (A11.9.4.3, A11.6.3.3).
emax= B/4 = 5.0 m/4 = 1.25 m
For all cases, eB < emax; therefore, the design is adequate with regard to eccentricity.
Step 6: Sliding
From Chapter 8, the factored resistance against failure by sliding, QR, is:
QR = Q + ep Qep
(Eq. 8-15)
14-27
Because of the potential for loss of soil in front of the wall, neglect the passive resistance of the
foundation material so that ep Qep = 0 (A10.6.3.3). Then from Table 14-4, = 1.0 for sliding of soil
against soil.
Nominal shear resistance is:
Q = V tan
(A10.6.3.3-2)
where:
tan = tan f
V = PEV
Therefore:
Q = PEV tan f = PEV tan
PEV
)(
kN/m) =
)(
kN/m) =
kN/m) sliding
(Eq. 8-11)
(A10.6.3.1.2b-1)
where:
c = Su = 230 kPa
s = f = 20.5 kN/m3
Df = 0 m (i.e., embedment neglected)
Therefore:
14-28
qult = c Ncm
The bearing capacity factor, Ncm, for Df/B # 2.5, B/L # 1.0 and H/V # 0.4 is obtained from
A10.6.3.1.2b:
Ncm = 3.8
and
qult =(3.8)(230 kPa) = 874 kPa
The resistance factor, , for bearing resistance of clay using the rational method and shear resistance
measured in laboratory tests is obtained from Table 8-8, from which = 0.60. The factored bearing
resistance is then:
qR = qult = 0.60 (874 kPa) = 524 kPa
Group/Item
Units
Strength I-a
Table 14-17
Summary for Bearing Check
MvTOT
MhTOT
Xo
VTOT
(kN/m)
(kN-m/m)
(kN-m/m)
m
713.0
1782.4
478.8
1.83
eB
m
0.67
quniform
(kPa)
194
Strength I-b
924.9
2132.2
478.8
1.79
0.71
258
Service I
666.9
1667.3
307.2
2.04
0.46
163
Because e < B/6 = 0.83 m, for Strength I-a and Strength I-b, the actual base pressure distribution for
these cases will be trapezoidal. For design of MSE walls, an equivalent uniform base pressure is
considered (A11.9.4.2). The uniform pressure, quniform, is taken as the average pressure of an
assumed rectangular distribution.
quniform = VTOT/(B - 2 eB)= VTOT/2 Xo
From Table 14-17, the maximum value of quniform = 258 kPa for Strength I-b. Because qR > quniform
(i.e., 524 kPa > 258 kPa), the bearing resistance is adequate.
Step 8: Overall Stability
Verify overall stability against a deep-seated soil failure using a limit equilibrium method of
analyses. For this example, it is assumed that the overall stability is adequate.
Step 9: Internal Stability - Tensile Resistance Reinforcement
(A) Factored Horizontal Force Acting on the Reinforcement
The factored horizontal force acting on any single layer of reinforcement is:
14-29
where:
hi =
H =
p =
The appropriate load factors are applied to the components comprising PVT used to calculate v and
the moments used to calculate Xo (and eB).
pv =
pv =
PVT =
k=
e=
B
B M - Mh L
- Xo = - v
= - M v - Mh overP VT
2
2
2
P VT
14-30
Layer
(1)
Table 14-18
Factored Vertical Load and Moment at Reinforcement Levels
for Reinforcement Tensile Resistance Evaluation
Mv
Mh
e(1)
quniform
Z
PVT
(kN/m)
(kN-m/m)
(kN-m/m)
(m)
(kPa)
(m)
0.33
146.0
365.0
0.3
0.002
29.2
1.00
224.3
560.8
3.6
0.02
45.2
1.67
302.5
756.2
11.9
0.04
61.5
2.33
379.5
948.8
26.9
0.07
78.1
3.00
457.8
1144.5
50.8
0.11
95.8
3.67
536.1
1340.2
85.3
0.16
114.6
4.33
613.1
1532.8
131.5
0.21
133.9
5.00
691.4
1728.5
192.5
0.28
155.7
5.67
769.6
1924.0
269.8
0.35
179.0
10
6.33
846.7
2116.8
363.5
0.43
204.5
For determination of quniform, eccentricity is assumed as zero for calculated e < 0.
Sample Calculations for Layer 1 follow (See Figure 14-5 for pressure diagram at Layer 1).
Figure 14-5
Pressure Diagram for Computation of quniform
for Reinforcement Tensile Resistance Evaluation of Layer 1
PEV1 = 1.35 (17.3 kN/m3) (0.33 m) (5.0 m) = 38.54 kN/m length of wall
PLSV = 107.5 kN/m length of wall (from Table 14-12)
PVT = PEV1 + PLSV = 38.5 kN/m + 107.5 kN/m = 146.0 kN/m
Moments About Front Face of Wall at Connection to Strip, Mv and Mh
Mv = (146.0 kN/m) (5.0 m/2) = 365.0 kN-m/m length of wall
Mh = 1.5 (17.3 kN/m3) (0.238) (0.33 m)3/6 + 1.75 (2.92 kPa) (0.33 m)2/2
Mh = 0.3 kN-m/m length of wall
Uniform Pressure on Effective Base Width
e=
L/2 - [(Mv - Mh)/PVT]
e=
5.0 m /2 - [(365.0 kN-m/m - 0.3 kN-m/m)/146.0 kN-m/m] = 0.002 m
quniform = PVT/(L - 2 e) = (146.0 kN-m/m)/[5.0 m - 2(0.002 m)] = 29.2 kPa
Values of H can be calculated by multiplying the factored pv values (quniform from Table 14-18) by
the active earth pressure coefficient, ka = 0.238. Pi can then be calculated by multiplying H by hi
(which was set at about 0.66 m). The AASHTO LRFD procedure (AASHTO, 1997a) requires a
check of the Strength Limit State and Service Limit State tensile resistance for polymeric
reinforcement, with different resistance factors given for each condition. Therefore, Table 14-19
includes the factored horizontal force acting on each layer for both the Strength I-b Limit State, PiA,
and the Service I Limit State, PiB.
Z
(m)
0.33
Table 14-19
Summary of Factored Horizontal Loads
for Reinforcement Tensile Resistance Evaluation
HA
HB
hi
PiA
p V
(kPa)
(kPa)
(kPa)
(m)
(kN/m)
29.2
7.0
4.3
0.66
4.6
1.00
45.2
10.8
7.1
0.66
7.1
4.7
1.67
61.5
14.6
9.9
0.66
9.6
6.5
2.33
78.1
18.6
12.8
0.66
12.3
8.4
3.00
95.8
22.8
15.9
0.66
15.0
10.5
3.67
114.6
27.3
19.1
0.66
18.0
12.6
4.33
133.9
31.9
22.5
0.66
21.0
14.8
5.00
155.7
37.1
26.1
0.66
24.5
17.2
5.67
179.0
42.6
30.0
0.66
28.1
19.8
10
6.33
204.5
48.7
34.2
0.66
32.1
22.6
Layer
14-32
PiB
(kN/m)
2.8
The tensile loading at the level of each grid must then be compared to the geogrid tensile resistance.
(B) Reinforcement Tensile Resistance
The vertical geogrid spacing is 0.66 m. From Table 14-19, for the strength limit state the maximum
value of HA = 48.7 kPa. Therefore, the maximum load per m length of wall, PiA, is:
Pi =
For this problem, laboratory creep tests have been performed which indicate a minimum limit state
reinforcement tensile strength, TR, of 130 kN/m and a tensile strength at 5 percent strain, Tw, of 91
kN/m. Applying = 0.25 from Table 14-4 to TR, results in a factored resistance of 33.6 kN/m, which
is adequate for this example from Eq. 14-8 as follows:
i Qi # Pn
PiA # TR
32.1 kN/m # 0.25 (130 kN/m) = 32.5 kN/m
For the service limit state, the maximum value of HB = 34.2 kPa. Therefore, the maximum load per
m length of wall, PiB, is:
PiB = HB hi (1 m) = (34.2 kPa)(0.66 m)(1 m) = 22.6 kN/m
The resistance factor of 0.25 from Table 14-4 for the service limit state will result in an acceptable
design from Eq. 14-8 for this example, as follows:
Qi # Pn
PiB # Tn
22.6 kN/m # 0.25 (91 kN/m) = 22.8 kN/m
Step 10: Internal Stability - Facing Connection
For this problem, it is assumed that the geogrid to wall facing connection is adequate.
Step 11: Internal Stability - Reinforcement Pullout
(A) Factored Horizontal Force Acting on the Reinforcement
The factored horizontal force acting on any single layer of reinforcement is:
Pi =
H hi = (p v k) hi = [EV (PEVi)] k hi
hi =
H =
p =
where:
14-33
The appropriate load factors are applied to the components comprising PVT used to calculate v and
the moments used to calculate Xo (and e).
pv =
pv =
PVT =
k=
Layer
1
Table 14-20
Factored Vertical Load and Moment at Reinforcement Levels
for Pullout Resistance Evaluation
Mv
Mh
e(1)
quniform
Z
PVT
(kN/m)
(kN-m/m)
(kN-m/m)
(m)
(kPa)
(m)
0.33
38.5
96.3
0.3
0.006
7.7
2
3
1.67
195.0
487.5
11.9
0.06
40.0
2.33
272.0
680.2
26.9
0.10
56.7
3.00
350.3
875.8
50.8
0.14
74.4
3.67
428.6
1071.4
85.3
0.20
93.1
4.33
505.6
1264.1
131.5
0.26
112.9
5.00
583.9
1459.7
192.5
0.33
134.5
5.67
662.1
1655.3
269.8
0.41
158.2
10
For determination of quniform, eccentricity is assumed as zero for calculated e < 0.
The values in Table 14-20 are:
(1)
PVT = PEV - Calculated based on the weight of the soil overlying strip using EV = 1.35
Mv - Calculated considering the same loads and load factors used to calculate PVT
Mh - Calculated based on the horizontal earth pressures with a load factor of EH = 1.5
14-34
e=
B M Mh L Mv Mh
B
X0 = v
=
2
PVT
2
PVT
2
Sample Calculations for Layer 1 follow (See Figure 14-6 for pressure diagram at Layer 1).
Figure 14-6
Pressure Diagram for Computation of quniform
for Reinforcement Pullout Resistance Evaluation of Layer 1
Layer
Z
(m)
p V
(kPa)
H
(kPa)
hi
(m)
Pi
(kN/m)
0.33
7.7
1.84
0.66
1.2
1.67
40.0
9.52
0.66
6.3
2.33
56.7
13.49
0.66
8.9
3.00
74.4
17.71
0.66
11.7
3.67
93.1
22.16
0.66
14.6
4.33
112.9
26.87
0.66
17.7
5.00
134.5
32.01
0.66
21.1
5.67
158.2
37.65
0.66
24.8
10
The tension loading at the level of each grid must then be compared to the geogrid pullout
resistance.
(B) Reinforcement Pullout Resistance
The pullout resistance of the geogrid reinforcement, Pfg, is:
(Eq. 14-17)
where:
fd =
From Table 14-4, = 0.90 for reinforcement pullout resistance in soil. The factored pullout
resistance Pfg for each layer is calculated in Table 14-20 and sample calculations for Layer 1 are
presented below.
Sample Calculation for Layer 1:
The pullout resistance, Pfg, at any layer is:
14-36
Pfg = 2w R fd s Z tanf
(Eq. 14-17)
0.33 m
Leff = L - (H - Z)/tan 64o = 5.0 m - (7.0 m - 0.33 m)/tan 64o = 1.75 m
0.80 (assumed typical value)
tan b = tan 38o = 0.78
17.3 kN/m3
1m
Substituting:
Pfg = 2wRfd s Z tanf = 2(1 m)(1.75 m)(0.8) (17.3 kN/m3)(0.33 m)(0.78) = 12.47 kN/m
and the factored pullout resistance is:
Pfg = 0.90 (12.47 kN/m) = 11.2 kN/m
Comparing the factored pullout resistance to the maximum factored horizontal stress at Level 1, Pi =
1.2 kN/m, (from Table 14-18) using Eq. 14-7:
i Qi # Qult
Pi # Pfg
1.2 kN/m # 11.2 kN/m
From a comparison of Pi with Pfg in Table 14-22, it is apparent that pullout resistance is adequate at
all reinforcement levels.
14-37
Table 14-22
Pullout Resistance in Reinforcement Levels
Layer
Z
(m)
Leff
(m)
Pfg
(kN/m)
Pi
(kN/m)
0.33
1.75
11.2
1.2
1.67
2.40
78.0
6.3
2.33
2.72
123.4
8.9
3.00
3.05
178.0
11.7
3.67
3.38
241.1
14.6
4.33
3.70
311.5
17.7
5.00
4.02
391.5
21.1
5.67
4.35
480.0
24.8
10
Summary
This example illustrates the design of an MSE retaining wall by LRFD. The AASHTO LRFD
Specification (pre 2002) for MSE walls was calibrated directly to ASD procedures in effect at the
time the calibration was performed (AASHTO, 1996; Christopher, et al., 1990). A reliabilitybased calibration for MSE wall design incorporating the procedures specified in the current
AASHTO ASD Interims has been completed based on the results of FHWA Demonstration
Project 82 (Elias and Christopher, 1996)
14-38
14-39
Approach: To perform the geotechnical design of the wall, the following steps are taken:
Compute and tabulate the unfactored loads and moments required for
design at the applicable limit states
Determine and tabulate the factored loads and moments required for
design at the applicable limit states
Solution:
Step 1: Unfactored Loads
(A) Length of Soil Reinforcement
The minimum soil reinforcement length, L, is the greater of either 70% of the wall height, H,
measured from the leveling pad (i.e., 0.7 6.0 m = 4.2 m) or 2.40 m (A11.9.5.1.4). However, for
many problems, longer reinforced lengths are needed. As a preliminary estimate, use L = 5.4 m.
(B) Vertical Earth Pressure (EV)
The weight of the reinforced soil backfill is (A3.5.1):
PEV1 = H L (r = (6 m)(5.4 m)(18.835 kN/m3) = 610.3 kN/m length of wall
The weight of the fill above the reinforced soil backfill is:
PEV2 = [ R (ht - H) + (L - R)(ht - H)] (b
PEV2 = [()(4 m)(2 m) + (5.4 m - 4 m)(2 m)](20.6 kN/m3) = 140.1 kN/m length of wall
14-40
heq
(m)
#1.5
1.70
3.0
1.20
6.0
0.76
$9.0
0.61
It is assumed for this problem that no live load surcharge acts over the reinforced soil mass due
to the backfill slope above the reinforcement.
(D) Horizontal Earth Pressure (EH)
The active earth pressure coefficient, ka for the broken back slope as shown on Figure 14-8 is:
sin 2 ( + )
ka =
sin 2 sin ( )
(A3.11.5.3-1)
in which:
sin ( + )sin ( )
= 1 +
sin ( )sin ( + )
(A3.11.5.3-2)
where:
= B = Nominal slope of backfill behind wall (deg)
= Angle of wall friction (deg) = 0
= b = 36
= 90 for vertical wall
14-41
By geometry (referring to Figures 14-7 and 14-8), the slope of the backfill, B, is:
B = tan 1
8m 6m
h1 H
= 9.5
= tan 1
2H
2(6 m )
Substituting:
' = 2.31; and
ka = 0.28
The uniform increase in horizontal earth pressure due to a live load surcharge is:
)p = (k)((s)(heq) = (ka)((b)(heq)
)p = (0.28)(20.6 kN/m3)(0.76 m) = 4.38 kPa
(A3.11.6.2-1)
The sum of the live load surcharge horizontal earth pressure, PLSH, acting over the entire wall
(assuming a rectangular distribution) is:
PLSH = ()p)(h) = (4.38 kPa)(8.00 m) = 35.0 kN/m length of wall
14-42
The basic earth pressure is assumed to vary linearly with depth in a triangular distribution. The
resultant, Pa, and the horizontal, PaH, and vertical, PaV, force components of the triangular
distribution can be determined (A3.11.5.1) using:
Pa =
Pa =
PaH =
PaV =
0.5 (b)(h2)ka
(0.5)(20.6 kN/m3)(8 m)2(0.28) = 184.6 kN/m length of wall
Pa cos B = (184.6)(0.99) = 182.7 kN/m length of wall
Pa sin B (184.6)(0.16) = 29.5 kN/m length of wall
Item
V
(kN/m)
Moment Arm
About Toe (m)
PEV1
610.3
2.7
1648
PEV2
140.1
3.5
490
PaV
29.5
5.4
159
Total
779.9
Table 14-25
Unfactored Horizontal Loads/Moments
Item
H
kN/m
Moment Arm
About Toe (m)
PEH = PaH
182.7
2.7
493
PLSH
35.0
4.0
140
Total
217.7
Strength I-a
1.00
1.50
1.75
Strength I-b
1.35
1.50
1.75
Strength IV
1.35
1.50
Service I
1.00
1.00
1.00
14-43
GROUP
PEV1
(kN/m)
PEV2
(kN/m)
PaV
(kN/m)
Unfactored
610.3
140.1
25.5
VTOT =
PEV1 + PEV2 + PaV
(kN/m)
779.9
Strength I-a
610.3
140.1
38.2
788.6
Strength I-b
823.9
189.1
38.2
1051.2
Strength IV
823.9
189.1
38.2
1051.2
Service I
610.3
140.1
29.5
779.9
Table 14-28
Factored Horizontal Loads
GROUP
PEH
(kN/m)
PLS
(kN/m)
Unfactored
182.7
35.0
HTOT =
PEH + PLS
(kN/m)
217.7
Strength I-a
274.0
61.2
335.2
Strength I-b
274.0
61.2
335.2
Strength IV
274.0
274.0
Service I
182.7
35.0
217.7
Table 14-29
Factored Moments from Vertical Loads
GROUP
MEV1
(kN-m/m)
MEV2
(kN-m/m)
MaV
(kN-m/m)
Unfactored
1648
490
159
MvTOT =
MEV1 + MEV2 + MaV
(kN-m/m)
2297
Strength I-a
1648
490
238
2376
Strength I-b
2225
662
238
3125
Strength IV
2225
662
238
3125
Service I
1648
490
159
2297
14-44
Table 14-30
Factored Moments from Horizontal Loads
MhTOT =
MEH
MLS
MEH + MLS
GROUP
(kN-m/m)
(kN-m/m)
(kN-m/m)
Unfactored
493
140
633
Strength I-a
740
245
985
Strength I-b
740
245
985
Strength IV
740
740
Service I
493
140
633
The average elastic settlement can be estimated (A10.6.2.2.3a, A10.6.2.2.3b) using the
DAppolonia Method (Gifford, et al., 1987) as:
D = q B o 1/M
where:
q = VTOT /LE = (779.9 kN/m)(5.4 m) - 144.4 kPa
B = LE = 5.4 m
Estimating the modulus of compressibility of sand, M, based on the average SPT blow count:
M = 33.5 MPa = 33 500 kPa
o and i are correction factors for the settlement equations based on D, H and B.
for D/B = 0.19; o = 0.9
for H/B = 1.9 and L/B >10; 1 = 0.7
Substituting:
D = (144.4 kPa)(5.4)(0.9)(0.7)/(33 500 kPa) = 0.015 m
D < Dtol = 0.022 m; settlement is acceptable.
14-45
Lateral Displacment
Using empirical procedures developed by FHWA (Christopher, et al., 1990), the lateral wall
facing displacment during construction is estimated at 5.2 mm/m of height which is within an
acceptable displacement limit of 6.2 mm/m from Section 14.3.1.
Step 5: Eccentricity
Table 14-31
Summary for Eccentricity Check
Group/Item
Units
VTOT
(kN/m)
MvTOT
(kN-m/m)
MhTOT
(kN-m/m)
Xo
m
eB
m
quniform
(kPa)
Strength I-a
788.6
2376
985
1.76
0.94
224.0
Strength I-b
1051.2
3125
985
2.04
0.66
257.6
Strength IV
1051.2
3125
740
2.27
0.43
231.5
Service I
779.9
2297
633
2.14
0.56
182.2
where:
Xo = Location of the resultant from toe of wall = (MvTOT - MhTOT)/VTOT
eB = Eccentricity = B/2 - Xo = 2.7 - Xo
The location of the resultant must be in the middle half of the base.
emax= B/4 = 5.4 m/4 = 1.35 m
For all cases, e < emax; therefore, the design is adequate with regard to eccentricity (A11.9.4.3,
A11.6.3.3).
Step 6: Sliding
From Chapter 8, the factored resistance against failure by sliding, QR, is:
QR = NJ QJ + Nep Qep (Eq. 8-15)
Because of the potential for loss of soil in front of the wall, neglect the passive resistance of the
foundation material so that Nep Qep = 0 (A10.6.3.3). Then from Table 14-4, NJ = 1.0 for sliding
of soil against soil.
Nominal shear resistance is:
Q = V tan * Qep
(A10.6.3.3-2)
where:
14-46
(Eq. 8-11)
(A10.6.3.1.2c-1)
Using values of Cw1, Cw2, Nqm and N(m from A10.6.3.1.2c, and substituting (f = 20.6 kN/m3, B =
5.4 m and Df =1 m:
kN
kN
q ult = (0.5) 20.6 3 (5.4 m )(0.62)(19.4 ) + 20.6 3 (1.0)(1.0 m )(19.3) = 1066 kPa
m
m
The resistance factor, , for bearing resistance of cohesionless soil using the rational method and
shear resistance based on SPT data is obtained from Table 8-8, from which = 0.35. The
factored resistance is then:
qR = qult = (0.35)(1066 kPa) = 373 kPa
Because e < B/6 = 0.9 m, for Strength I-b and Strength IV, the actual base pressure distribution
for these cases will be trapezoidal. Because e > B/6 for Strength I-a, the actual base pressure
distribution for this case will be triangular. For design of MSE walls, an equivalent uniform base
pressure is considered. The uniform pressure, quniform, is taken as the average pressure of an
assumed rectangular distribution.
quniform =
14-47
From Step 5, the maximum value of quniform = 257.6 kPa for Strength I-b. Because qR > quniform
(i.e., 373 kPa > 258 kPa), the bearing resistance is adequate.
Step 8: Overall Stability
Verify overall stability against a deep-seated soil failure using a limit equilibrium method of
analyses.
Step 9: Internal Stability - Reinforcement Tensile and Pullout Resistance
To evaluate internal stability, the slip or the rupture of the reinforcement strips must be
considered.
(A) Factored Horizontal Force Acting on Reinforcement
The factored horizontal force acting on any single layer of reinforcement is:
Pi = FH hi = ((p FV k) hi
where:
hi =
FH =
(p =
The appropriate load factors are applied to the components comprising PVT to calculate (p and Xo
(and e).
p v = Pressure due to resultant of factored vertical forces at Level i (kPa)
p v = quniform = PVT/(B - 2 eB) (1 m) = PVT/2 Xo (kPa)
PVT = Resultant of factored vertical forces at the base of Layer i (kN/m)
k=
Lateral earth pressure coefficient from Figure 14-9 (dim)
14-48
Figure 14-9
Geometry of Assumed Failure Surface for Example Problem
Use the geometry shown in Figure 14-9 to establish the distribution of the lateral earth pressure
coefficient, k, with depth. From Figure 14-9, k = ko at 1.1 m (i.e., H1 - H) above the top of the
wall, and k = ka at 4.9 m below the top of the wall and lower (i.e., at 6.0 m below point where ko
is applied). From Eq. A3.11.5.7-5:
ko = 1 - sin (Nr) = 1 - sin (33o) = 0.46
From Eq. A3.11.5.3-1 and A3.11.5.3-2 with 2 = 90o, * = 0o and $ taken as 0o, ka = 0.29.
Based on Table 14-31, the load case which produces the largest eccentricity and the
corresponding largest uniform pressure is Strength Ib (EV = 1.35 and EH = 1.5). Note that the
live load surcharge is not included in the calculation of the vertical load resultant, PVT, for
internal stability, regardless of position, and is considered for vertical load calculations only for
bearing capacity and global stability if the surcharge is positioned over the reinforced soil zone.
The factored uniform pressure distribution is then calculated at each layer of reinforcing. For
this example, reinforcing strips are placed at a vertical spacing of 0.75 m beginning at 0.375 m
below the top of the wall. The uniform pressure for each layer is shown in Table 14-32 and
sample calculations for layers 1 and 7 follow.
14-49
Table 14-32
Factored Vertical Load and Moment at Reinforcement Levels
(1)
Layer
z
(m)
PVT
(kN/m)
Mv
(kN-m/m)
Mh
(kN-m/m)
e(1)
(m)
quniform
(kPa)
0.38
206
566
40.7
0.15
40.3
1.12
312
860
80.8
0.20
62.3
1.88
418
1158
140.3
0.27
86.0
2.62
526
1461
222.7
0.35
111.7
3.38
634
1768
331.6
0.44
140.0
4.12
743
2080
470.6
0.54
171.8
4.88
853
2395
643.3
0.65
207.8
5.62
964
2715
853.3
0.77
249.6
where:
PVT = Calculated based on the weight of the soil overlying the strip and the
vertical component of the lateral earth pressure, and includes ( = 1.35 and
1.50 as appropriate
Mv = Calculated in the same manner as PVT
Mh = Calculated based on the horizontal earth pressures with a load factor of 1.5
e=
B M Mh
B
X0 = v
2
PVT
2
14-50
Figure 14-10
Pressure Diagram for Computation of quniform
for Internal Stability Evaluation of Layer 1
Layer 1
Factored weight of soil overlying strip above wall (kN). For broken-back
slopes, the weight of the soil wedge above the top of the reinforced soil
zone is conventionally replaced in internal stability analysis by a uniform
surcharge equal to 0.5 (b hS where hS is the height of the slope at the back
of the wall.
P2 =
14-51
At Layer 1:
PVT = P1 + P2 + P3 = 150.2 + 51.5 + 4.0 = 205.7 kN/m length of wall
Moments About Front Face of Wall at Connection to Strip, Mr and Mh
Mv =
Mv =
Mh =
1.5 cos (9.5o) (20.6 kN/m3) (0.28) (2.0 m + 0.375 m)3/6 + 1.75 (4.38
kPa) (2 m + 0.375 m)2/2
40.7 kN-m/m length of wall
Mh =
Layer 7
P1 =
P2 =
P3 =
PVT =
Mv =
Mv =
Mh =
Mh =
e=
quniform =
By calculating the lateral earth pressure coefficients, k, at specific locations over the height of
the wall, the value of k can be determined by interpolation at the levels of the various strips as
indicated on Table 14-33. Values of FH can be calculated by multiplying these k values by the
factored Fv (quniform from Table 14-32). Pi can then be calculated by multiplying FH by hi (which
was set at 0.75 m).
14-52
Table 14-33
Summary of Factored Horizontal Loads
Layer
z
(m)
FV
(kPa)
FH
(kPa)
hi
(m)
Pi
(kN/m)
0.375
0.42
40.3
16.9
0.75
12.7
1.125
0.40
62.3
24.9
0.75
18.7
1.875
0.38
86.0
32.7
0.75
24.5
2.625
0.36
111.7
40.2
0.75
30.2
3.375
0.33
140.0
46.2
0.75
34.6
4.125
0.31
171.8
53.3
0.75
39.9
4.875
0.29
207.8
60.3
0.75
45.2
8*
5.625
0.29
249.6
72.4
0.75
54.3
Having determined the loading at the level of the strip the loading on the strips must then be
determined and compared to the allowable load. To determine the load on the strip, the area of
the strip at the end of the design life (the reduced strip area to include corrosion loss) will be
used.
(B) Corrosion Losses
The initial width, wi, and thickness, ti, of the reinforcing are:
w = 50 mm
ti = 4 mm
Use a galvanizing thickness of 0.086 mm/side.
For structural design, requirements for sacrificial thickness for each exposed surface are
(A11.9.8.1):
Loss of galvanizing =0.015 mm/yr for first 2 years, and 0.004 mm/yr thereafter
Loss of carbon steel = 0.012 mm/yr after zinc depletion
Pmax
38.0 kN
=
= 384 000 kPa
w t 100 0.05 m 0.00198 m
The horizontal force used to design the connection to the panels may be taken as not less than
85% of the maximum calculated force as determined above, except for the lower one-half of the
structure (below Layer 4) where the connections must be designed for 100% of the maximum
calculated force (A11.9.5.1.2).
Connection design forces are calculated by multiplying the factored horizontal force from Table
14-33 by the tributary width for the strip (= c/c spacing). For the upper half of the wall the
horizontal force is reduced by 15%. The design force for the strips is summarized in Table 14-34
and sample calculations for the strips follow.
Sample Calculation for Layer 1:
Pi = 12.7 kN/m
Tributary Width (c/c spacing) = 0.7 m
Force/Strip = (12.7 kN/m) (0.7 m) = 8.9 kN
Connection Design Force =(0.85) (8.9 kN) = 7.6 kN
14-54
18.7
13.1
11.1
24.5
17.2
14.6
30.2
21.1
17.9
34.6
24.2
24.2
39.9
27.9
27.9
45.2
31.6
31.6
54.3
38.0
38.0
(Eq. 14-14)
where:
f* =
s =
Z=
As =
From Table 14-4, = 0.90 for soil pullout. The factored pullout resistance Pfs for each layer is
calculated in Table 14-35 and sample calculations for Layer 1 follow.
14-55
Table 14-35
Pullout Resistance in Reinforcement Levels
Layer
Z
(m)
ft
Leff
(m)
As
(mm2)
Pfs
(kN)
N Pfs
(kN)
1.43
1.68
3.28
0.3282
14.9
13.4
2.18
1.51
3.28
0.3282
20.4
18.3
2.93
1.34
3.28
0.3282
24.3
21.9
3.68
1.17
3.38
0.3375
27.4
24.7
4.43
1.00
3.82
0.3825
32.0
28.8
5.18
0.83
4.28
0.4275
34.8
31.3
5.93
0.67
4.72
0.4725
35.1
31.6
6.68
0.65
5.18
0.5175
42.3
38.1
By inspection, Leff minimum length of reinforcement beyond the assumed failure surface of 0.9
m, and the minimum total required length of reinforcement of 2.4 m for all levels (A11.9.5.1.4).
Sample Calculation for Layer 1:
he pullout resistance, Pfs, at any layer is:
Pfs = ft (s Z As
(Eq. 14-14)
14-56
CHAPTER 15
FLEXIBLE CULVERT DESIGN
15.1 Introduction
Culverts are produced with large variety of material properties, geometric wall sections, sizes and
shapes. For highway applications, culverts are concrete, steel, aluminum, thermoplastic and
composites of these materials, and can have diameters as small as 0.30 m and spans as large as 15 m.
The scope of this training manual does not permit treatment of all the available culvert systems.
Therefore, discussions herein are limited to small-diameter, round culverts.
Culverts are generally divided into two major classes, flexible and rigid. Flexible culverts respond
to vertical loads by depending on a large strain capacity and interaction with the surrounding soil to
hold their shape. If the backfill envelope is not constructed to develop adequate passive resistance
and stiffness, the culvert will deflect beyond a tolerable limit. Conversely, the strain capacity of
rigid culverts is limited. Therefore, these structures must develop significant ring stiffness and
strength to support the vertical pressures imposed upon them. This is the basis of the D-Load (i.e.,
Three-Edge Bearing) design method.
The differentiation between flexible and rigid culverts is becoming blurred. New culvert products
being advanced that employ a flexible structural shell with a rigid liner to resist abrasion and
corrosion. The strain capacity of such a composite is more limited than the flexible shell.
Alternatively, the latest direct design method being advanced for rigid culverts (ASCE, 1993)
provides for consideration of side and haunch soil support in excess of the active or at-rest cases.
Hence, rigid culvert designs are becoming more dependent upon soil support than designs based
solely upon the rigid body three-edge bearing model. This developing trend is not yet recognized in
the current AASHTO specifications. Consistent with AASHTO and common current practice, the
treatment of culverts in this training manual is divided into two parts. This chapter focuses with
comparisons between Allowable Stress Design (ASD) and Load and Resistance Factor Design
(LRFD) for flexible culverts. Rigid culverts are addressed in Chapter 16.
Culverts, like other transportation systems, require input from multiple engineering disciplines
during design. Successful performance of the total system requires application of the principles of
hydraulics, hydrology, economics, roadway geometric design, traffic and safety. All but the
structural and geotechnical aspects of culvert design are beyond the scope of this manual. Relative
to geotechnical and structural design of flexible culverts, this chapter:
y
y
y
15-1
Relative to loading of a flexible culvert, the ultimate load capacity, Qult, is estimated based on the
culvert material strength and the suitability of the design is evaluated by determining the
allowable axial design load, Qall, using:
Q Q all =
Q all
(Eq. 15-1)
FS
where:
Q = Design load (kN)
Qall = Allowable design load (kN)
15-2
ult
A = Pall
FS
(Eq. 15-2)
where:
all =
Pall =
ult =
A=
The required values of FS for wall compression or buckling and for longitudinal seam resistance
to thrust are commonly taken as indicated in Table 15-1.
Table 15-1
Factor of Safety on Ultimate Structural Capacity of Flexible Culverts
(AASHTO, 1997b)
Performance Limit
Wall Compression (Thrust)
Wall Buckling
Seam Strength
Required Minimum
Factor of Safety (FS)
2.0
2.0
3.0
The reduction of the structural capacity of a culvert from an ultimate value (based on the yield or
buckling strength of the material) to an allowable value based on a Factory of Safety (FS) accounts
for all uncertainty in the variation of applied loads and the ultimate structural capacity of the pipe.
Therefore, the structural design of the pipe is performed by ASD using actual estimated loads.
In addition to structural capacity evaluation, the design of flexible culverts by ASD establishes
limitations of culvert flexibility to avoid damage due to handling and construction-related
deformations using:
FF FFmax
(Eq. 15-3)
where:
FF = Flexibility Factor (m/kN)
FFmax =
Suggested maximum permissible Flexibility Factor (m/kN)
15-3
i i Q i R n
= Rr
(Eq. 15-3)
i i i n
(Eq. 15-4)
where:
i =
i =
Qi =
=
Rn =
Rr =
i =
n =
For flexible culvert design at the Strength Limit States, the following values of in Eq. 15-4
normally can be used:
D = R = 1.0
I =
1.05 for structures deemed operationally important, 1.00 for typical structures
and 0.95 for relatively less important structures.
Assessments of the relative ductility and redundancy of flexible culverts and of the roles that
ductility and redundancy play in culvert performance have not yet been made. Therefore, D = R =
1.0 is assumed for flexible culverts in this chapter.
Determination of the operational importance of a structure is made by the facility owner as described
in Chapter 4. For the purpose of this chapter, the value of I is assumed equal to 1.0.
At any Service or Construction Limit State, D, R and I = 1.0.
15-4
Relative to loading of a flexible culvert, the suitability of a culvert section with respect to structural
resistance can be obtained using Eq. 15-4, rewritten as:
i i Q i Pn
= Pr
(Eq. 15-6)
where:
i i Q i
Pn=
Pr =
Values of load factor and load factor combinations for each applicable limit state must be developed
using the guidelines described in Chapters 3 and 4 and in Section 15.2.2.1, and loads should be
developed as described in Chapter 4. The ultimate resistance, Rn, should be determined for each
type of resistance described in Section 15.3. Selection and modification of resistance factors, , are
described in Section 15.2.2.3 through Section 15.2.2.5.
15.2.2.1 Limit States (A12.5)
Limit states, as implemented in the AASHTO LRFD Specification (AASHTO, 1997a), represent a
more refined version of the concept of critical load combinations that have been part of the
AASHTO specifications for nearly 30 years (AASHO, 1969). However, before limit state design
concepts were introduced into the specification, various percentages of allowable stresses were
specified for different critical load combinations.
The AASHTO LRFD Specification requires that Strength and Service I Limit States be considered
for culvert design. In addition, the LRFD specification requires that construction loads be
considered in the design. For buried culverts, construction loads may govern design under shallow
covers, particularly when heavy earth moving equipment is considered. The Service I Limit State
represents performance at working levels of the AASHTO design vehicle live load.
The performance limits and limit states that must be considered for design of small-diameter, round
flexible culverts are listed in Tables 15-2 and 15-3.
Table 15-2
Limit States Applicable for Design of Metal Pipes and Arches
Performance Limits
Wall Compression (Thrust)
Strength I
Limit State
U
Service I
Limit State
Construction
Limit State
U
Wall Buckling
Seam Strength
Handling Flexibility
15-5
Table 15-3
Limit States Applicable for Design of Thermoplastic Pipe
Performance Limits
Wall Compression (Thrust)
Wall Buckling
Strength I
Limit State
U
Service I
Limit State
Construction
Limit State
U
U
Tensile Strain
Handling Flexibility
Although handling flexibility is not actually evaluated under Service loads, the consideration of
flexibility is in effect a Service Limit State in that the intent is to limit deformations during handling
and installation. Excessive deformations could induce bending stresses of unquantifiable magnitude
in flexible culverts, and could negatively impact the structural capacity of a culvert.
From Tables 15-2 and 15-3 it is evident that deflection criteria (i.e., tensile strain) under service
loads are applied only to thermoplastic pipe. This is not meant to imply that deflection limits are
inherently more applicable to one type of flexible pipe than another. Rather, it recognizes that
thermoplastics may experience limitations of ductile strain capacity while undergoing deflection.
Some thermoplastic product designs must be limited to between 3.5 and 5 percent strain to avoid
local buckling of the cross section and material rupture.
15.2.2.2 Loads and Load Factors (A12.5.3, A12.5.4)
Culvert design is influenced primarily by vertical earth pressures and, under shallow cover, by
vehicular live loads. Chapter 4 points out that loads resulting from unbalanced horizontal earth
pressures, compaction and water buoyancy can be important in the design of flexible culverts.
However, loads from wind, temperature fluctuations, vehicle braking, and structure dead weight are
insignificant compared to the vertical earth pressures and can be safely neglected. Also, historical
earthquake records appear to support the fact that buried culverts are highly resistant to seismic
damage, except in the vicinity of faulting or where ground instability or unbalanced loading occurs.
The magnitude of vertical earth pressure on a culvert can vary significantly, being not only
dependent upon the height of cover and density of material above the culvert, but also upon the ring
stiffness of the culvert relative to its adjacent side fill. The magnitude of vertical pressure exerted on
flexible culverts is integrally associated with the passive pressure that the culvert develops in the
adjacent side fill. Also, foundation stiffness can be a factor. If the culvert settles more than its
adjacent side fill, the load carried by the culvert wall is reduced. When the opposite is true, the load
on the culvert wall is increased. As in other structural systems, the stiffest element attracts the
greatest load.
Because of the numerous factors affecting its magnitude, vertical earth pressure is highly variable
relative to other loads on culverts. The load factors specified in the LRFD AASHTO Specification
and given in Table 15-4 reflect this variability.
15-6
Table 15-4
Load Factors for Flexible Culvert Design
(after AASHTO, 1997a)
Load Type
Strength I
Earth Loads:
y Max. Vertical Earth Press (EV)
y Min. Vertical Earth Press. (EV)
Vehicular Live Loads:
y Design Truck or Tandem (LL)
y Dynamic Load Allowance (IM)
Load Factor,
Limit State
Service I
Construction
1.95
0.90
1.00
1.00
1.00
1.00
1.75
1.75
1.00
1.00
1.50
1.50
As seen in the table, = 1.95 is assigned for vertical earth pressure, EV, under the Strength I Limit
State, whereas for vehicular live load or impact load, = 1.75 for the same limit state. This
represents a reverse ratio compared to other structural systems, where live loads are generally
considered to have more uncertainty than dead loads.
15.2.2.3 Resistance Factors (A12.5.5)
Resistance factors for culverts are summarized in Table 15-5. The factors are based on direct
calibration with ASD.
Table 15-5
Resistance Factors for Design of Flexible Culverts
(after AASHTO, 1997a)
Structure Type and Design Consideration
Resistance
Factor
15-7
1.00
0.67
0.67
0.67
0.67
1.00
i i
(Eq. 15-7)
where:
i =
i =
=
Resistance factor (dim)
FSLRFD = LRFD equivalent factor of safety (dim)
Table 15-6 suggests that the LRFD load and resistance factors generally result in slightly less
conservative design than ASD when earth loads predominate and a less conservative design than
ASD when live loads predominate for a structure designed for the Strength I Limit State. However,
due to the greater live loads incorporated in the LRFD Specification (AASHTO, 1994) as noted in
Chapter 4, the LRFD method is in fact more conservative than ASD, even where live loads
predominate.
Table 15-6
Comparison of Effective Factors of Safety for Flexible Culverts
Strength I Limit State
Loading Combinations
Load
Factor
i
EV:
y Thrust & Buckling (no seam)
1.95
y Seam Strength
1.95
LL, IM:
y Thrust & Buckling (no seam)
1.75
y Seam Strength
1.75
(1)
Assumes i = 1.0 for typical structure.
Maximum
Resistance
Factor
LRFD Equivalent
Factor of Safety,
FSLRFD(1)
ASD Factor
of Safety
1.00
0.67
2.0
2.9
2.0
3.0
1.00
0.67
1.8
2.6
2.0
3.0
15-8
When using LRFD, similar flexibility to vary the required level of safety should also be available
and, in some cases, is inherent in the load and resistance factors used. Whereas the same factor of
safety is generally used in ASD regardless of the source of loading, the equivalent factor of safety in
LRFD (defined by Eq. 15-7) varies for a given resistance factor depending on the source of loading
as a function of the available load factors.
To modify the resistance factors for design of flexible culverts to account for average load factors
and equivalent factors of safety other than those identified in Table 15-6, the following equation may
be used:
FS
m = T T
FS D
(i i )T
( i i Q i ) Q i
(Eq. 15-8)
where:
m = Modified resistance factor (dim)
T = Tabulated resistance factor (dim)
FST = Tabulated factor of safety (dim)
FSD = Desired factor of safety (dim)
(ii)T =
Product of tabulated load factors (dim)
i i Qi
= Actual average load factor (dim)
Qi
Modifying resistance factors may seem reasonable, but such modification may not be consistent with
the goal of LRFD to achieve equal reliability against failure of structure components, unless the
factor of safety accurately models the reliability of the capacity predictive method used. The
15-9
resistance factors may be more appropriately modified through application of the probabilistic
procedures described in Chapter 3 to achieve the desired level of reliability if sufficient data are
available.
15.2.3 Summarized Comparison of ASD and LRFD
As noted before, the process used to develop a flexible culvert LRFD differs very little from the
process used for ASD. The similarity is illustrated in the parallel flow charts in Figure 15-1.
Specific differences between the methods and other important issues are highlighted in the following
section.
Other aspects of the culvert design such as identifying special considerations (e.g., potential for loss
of support through scour), developing a design foundation profile and determining requirements for
construction control are inherent aspects of the design process required for both LRFD and ASD.
15.3 Performance Limits
Design of flexible culverts by either LRFD or ASD must provide adequate resistance against
geotechnical and structural failure and limit deformations to within tolerable limits. Following
selection of a culvert based on hydraulic considerations, the design of a flexible culvert requires
consideration of many factors which can affect the culvert performance, including:
y
y
y
y
For these design factors, there is no difference between LRFD and ASD analysis procedures. The
following sections highlight differences between LRFD and ASD in the performance criteria and
application of design procedures.
15.3.1 Displacements and Tolerable Movements (A12.6.2)
Wall displacements are usually limited based on controls required during construction. However.
short- or long- term deflection limits for small diameter round culverts are generally not specified
explicitly.
Control of ring deformation of flexible culverts must be considered both during and after culvert
installation. During handling and placement, the culvert wall stiffness and strength, as well as
specified installation practices, serve to limit ring deformation. The handling flexibility factor
described in Section 15.3.5 is intended to quantify this performance limit.
After culvert installation, the stiffness of the surrounding soil envelope totally controls long term
ring deformations. This is true for any practical wall stiffness or strength associated with flexible
culverts. These facts were established by early investigators such as Spangler (1941, 1950) and
Watkins (1958) who found the long-term culvert deflections were a function not only of the ring
stiffness of the culvert but also of the modulus of the surrounding soil envelope. This soil-structure
15-10
ASD
LRFD
Figure 15-1
Generalized Flow Chart for Flexible Culvert Design
by ASD and LRFD
15-11
interaction phenomena has been studied extensively using both empirical and analytical techniques
in an effort to develop a generalized theory of performance. While great strides have been achieved
in the understanding and quantification of culvert performance, displacements remain the most
difficult performance limit to predict due to the lack of a consistent and reliable predictive model for
the backfill stiffness modulus of soil surrounding a flexible culvert.
It is instructional to note that round flexible culverts generally show two types of ring deformation
modes, depending upon the relative density of the adjacent side fill, as illustrated in Figure 15-2.
For loosely placed backfill, the deformation mode is one of ovaling. As load is increased, the top arc
eventually becomes so flat that a reverse curvature (snap-through buckling potential) occurs. On the
other hand, for densely placed backfill, ovaling is not predominant. Rather, a local instability, either
wall crushing or inward cusping (local bending/buckling) occurs at some point around the pipe
periphery, generally over the top third.
Figure 15-2
Comparison of Flexible Culvert Deflection Modes in Loose and Dense Soils
(After Watkins and Moser, 1969)
Tests on over 100 corrugated steel pipe specimens ranging in diameter from 0.90 m to 1.50 m
15-12
performed in the Utah Test Cell produced the ovaling mode of deformation for soil that was either
loosely placed or compacted to less than about 85 percent relative compaction (AASHTO T-180;
Watkins and Moser, 1969). For this range of relative compaction levels, reverse curvature was
reached at vertical deflections of between 10 and 20 percent. For relative compaction levels
between 85 and 90 percent, a mix of ovaling and inward cusping was reached at vertical deflections
of between 5 and 10 percent. For relative compaction levels greater than 90 percent, the
performance limit was generally wall crushing and inward cusping, which occurred at vertical
deflections of less than 5 percent. As the compactive effort and soil density increased, loss of
deflection tolerance was offset by greater loads than could be supported.
This finding led to the conclusion that, for small-diameter, round culverts, adequate density of the
backfill envelope eliminates the potential for catastrophic, snap-through buckling due to excessive
ovaling and reverse curvature. Because snap-through buckling has been analytically and empirically
challenging to quantify for soil-structure interaction systems, the focus became field control of
deflections through a combination of backfill material quality and construction procedures.
While this approach is not fully rigorous from a theoretical viewpoint, its implementation has been
highly successful statistically. However, the approach is not intended to be applied to large diameter
and multi-radius culverts and long spans that can experience more complicated failure mechanisms.
15.3.2 Wall Compression (Thrust) (A12.7.2.2, A12.12.3.4)
Flexible culvert walls must be designed to resist earth and live loads without a compression failure.
The factored resistance of a culvert wall to compression failure can be determined as:
R r = R n = Af L
S
R r TL = PL
2
where:
Rn =
Rr =
=
S=
PL =
TL =
A=
fL =
Mechanical properties for common flexible culvert materials are presented in Tables 15-7 and 15-8.
Note that the material limit stress for aluminum and steel is the yield point stress. For
thermoplastics, the limit stress is the 50-year ultimate tensile stress. Wall section properties for
15-13
culvert types can be found in Section 12, Appendix A of the LRFD Specification.
Table 15-7 (A.A12-9)
Strength and Elastic Properties of Aluminum and Steel Culverts
(after AASHTO, 1997a)
Aluminum
fL = fy
(kPa)
165 000
Em
(kPa)
69 000 000
Steel
228 000
Culvert Material
Culvert Material
Material Cell
Class
fL(3) = fy
(kPa)
Em(3)
(kPa)
PE:(1)
152 000
9 930
335434C
Solid Wall
152 000
6 210
315420C
Corrugated Wall
138 000
7 720
334433C
Profile Wall
152 000
9 930
335434C
Profile Wall
PVC:(2)
965 000
25 500
12454C
Solid Wall
1 090 000
17 900
12364C
Solid Wall
965 000
25 500
12454C
Profile Wall
1 090 000
17 900
12364C
Profile Wall
(1)
ASTM D3350 Cell Class Specification; (2) ASTM D1784 Cell Class Specification
(3)
50-year properties
Several sources of variability exist on both the load and resistance sides of Eq. 15-10. First, the
design pressure is dependent upon assumptions of live and dead load magnitude and distribution
through the soil. Undoubtedly, load estimation is the source of greatest uncertainty in this
relationship. Typically, provisions for load and load distribution have been structured with very
conservative assumptions for most applications. Moreover, loads and load distribution are
considered to be the same for all flexible culvert types. AASHTO LRFD specifies use of the prism
load described in Chapter 4. Load relief through positive arching is not currently recognized.
Another significant source of uncertainty relative to the determination of resistance in Eq. 15-10 is
the calculation of wall area. For linear elastic wall sections that are fully effective, such as for
corrugated metal, the variability is low. On the other hand, the metal spiral rib wall section is not
fully effective. Whereas it is known that the degree of cross-section effectiveness is load dependent,
current design criteria only recognize the effective area at ultimate load, based on tests. Hence, at
working loads an unquantified "safety factor" or reliability index is present, which is greater than
that which exists for the corrugated profile at the same percentage of load.
15-14
The variation in diameter due to manufacturing processes is very small. However, flexible culverts
may deflect during or after installation, which increases their horizontal diameter and, hence, the
load they must support. Failure to account for this increase is unconservative. But, for small
diameter culverts, the effects are small and readily quantifiable. A five percent deflection generally
results in less than a five percent increase in load for circular culverts with adequate backfill density.
However, this simplicity does not apply to larger multi-radius culverts and long spans, where shape
changes can have major nonlinear effects on both load and resistance.
Thermoplastic pipe exhibits viscoelastic behavior. Therefore, its material limit stress is both load
and time dependent. Because of this complex performance and the relatively recent introduction of
thermoplastics as a culvert material, provisions base the thermoplastic limit stress on a 50-year,
continuous tensile loading. This is generally believed to be very conservative for burial depths
greater than two diameters (assuming a stable wall section under compression). More research is
currently being conducted to better define the compressive strength of thermoplastic culvert wall
sections.
15.3.3 Wall Buckling (A12.7.2.4, A12.12.3.6)
Flexible culvert walls must be designed to resist earth and live loads without a buckling failure. The
resistance of a culvert wall section to buckling failure can be determined as:
R r = R n = Af cr
(Eq. 15-11)
where:
Rn =
Rr =
=
S=
PL =
TL =
A=
fcr =
Two wall buckling relationships have commonly been adopted for flexible culverts. The first,
applicable to metal culverts, has been credited to Watkins, although details of its development have
never been published. This relationship was first published as a design criterion in the "Handbook
of Steel Drainage and Highway Construction Products (AISI, 1967), and was later adopted by
AASHTO (1969). The relationship is based on the theoretical buckling of thin circular tubes under
hydrostatic pressure (Timoshenko and Gere 1961), but was modified for application to buried
15-15
12 E m
kS
r
where:
fcr =
Em =
k=
S=
r=
An interaction equation is also used to check metal culverts for behavior in a transition between the
above critical buckling stress and yield stress. However, this interactive check applies only to small
diameter culverts and is not discussed herein.
The thermoplastic culvert industry chose to use the buckling equation originated by Luscher (1966)
and adopted by the American Water Works Association. Luscher assumed an elastically supported
thin plate with an empirical correction for soil stiffness based on his tests with model tubes, and also
recognized that shallow cover buckling was less sensitive to side fill stiffness than to the relationship
between cover height ring stiffness. This equation was further modified by Glascock (1980), who
performed further calibrations based on additional soil box tests and some full scale observations.
The resulting equation was adopted by AASHTO (1997a, 1997b):
1/ 2
f cr = 0.77
R B Ms E m I
A 0.149 R 3
where:
fcr =
R=
A=
B=
Ms =
15-16
Em =
I=
hw =
h=
A complete treatment of culvert buckling is beyond the scope of this chapter. In addition to the
criteria described, the buckling of buried conduits has been addressed by many other investigators.
Some other studies were summarized by Leonards and Stetkar (1978), which may serve as a further
reference to the workshop participant.
15.3.4 Seam Strength (A12.7.2.5)
Specification provisions for seam strength are applicable only to culverts with bolted, riveted or spot
welded seams. Bolted seams are found in both aluminum and steel versions of field-assembled
corrugated plate, known as structural plate. These plates are fabricated in a wide range of sizes and
shapes, with diameters or spans ranging from 1.50 m to over 15 m. The corrugations of structural
plate are typically deeper and the metal thickness greater than used for factory-made pipe. Two
primary reasons for using-field assembled pipe are when size restrictions preclude shipping of
factory made pipe and for applications requiring heavier wall sections than can be fabricated from
the spiral wound coil stock used in factory made pipe.
Some small diameter factory made pipe is still being fabricated from sheets that are riveted or spot
welded together. However, this practice is decreasing dramatically in the United States, being
supplanted by the more efficient spiral corrugation process. In addition to most factory made
corrugated profiles, all aluminum and steel box rib profiles are spiral wound. Also, all thermoplastic
pipes employed as culverts are fabricated without seams.
Longitudinal seams carry the primary ring compression forces and are the focus of the design
provisions. Circumferential seams are secondary. They serve to clamp the plates together but
generally do not carry significant forces.
The resistance of longitudinal culvert wall seams, SS, to the applied load is checked using:
S
SS T L = PL
(Eq. 15-15)
where:
PL =
=
S=
TL =
SS =
Values of SS for bolted, riveted and spot-welded corrugated steel and aluminum pipe are provided in
15-17
the Section 12, Appendix of the AASHTO LRFD Specification. The values for SS tabulated in
AASHTO are based on the results of short column compression tests using average rather than
minimum material properties. These numbers represent two very different failure modes. The most
catastrophic mode is brittle shear failure of the fasteners, which could result in collapse of the
culvert. This mode is generally associated with high wall stiffness relative to fastener stiffness and
strength.
A less catastrophic failure mode is ductile local yielding and buckling of the wall material at the
fastener location. This results from high fastener stiffness and strength relative to wall stiffness.
The local deformations associated with this type of failure results in clamping of the wall section, at
which point the fasteners are partially or fully unloaded. Also, the local buckling that occurs at the
fastener location shortens the culvert wall axially, which induces positive arching, further reducing
the load on the culvert wall under deep burial conditions (greater than one diameter of cover).
The specifications have long recognized the greater uncertainty associated with seam capacity and
have assigned a higher safety factor for wall compression of culverts with seams compared to those
without. However, wall sections having ductile failure modes have grater levels of safety against
collapse in extreme events than those with brittle failure modes. Nevertheless, this difference is not
currently recognized in the specification provisions.
15.3.5 Handling Flexibility (A12.7.2.5, A12.12.3.7)
Loads applied during construction often represent the most significant source of bending stress in
flexible culverts. However, these loads are extremely difficult to quantify. As a result, empirical
measures are employed to define minimum wall sectional properties that offer adequate bending
resistance to handling and construction-related deformations. These measures are formulated in
terms of ring flexibility (the inverse of ring stiffness) using:
FF =
S
Em I
where:
FF =
S=
Em =
I=
FF is then compared to the maximum permissible flexibility factor, FFmax, using Eq. 15-3. Values of
FFmax for various culvert materials are provided in Table 15-9.
Factors other than the parameters identified in Eq. 15-16 (e.g., material type, corrugation profile,
cross-sectional shape and backfill characteristics) are known to contribute to the ability of culverts to
resist handling and construction distress. Therefore, the AASHTO LRFD Specifications provide a
number of tables that list maximum flexibility factors for various parameters.
15-18
Material Type
Steel Pipe
Aluminum Pipe
Steel Plate
Aluminum Plate
Thermoplastic Pipe
These tabulated flexibility limit values have been developed and promulgated by the industries that
manufacture culvert structures. As experience factors, they implicitly include economic
considerations as well. For instance, a contractor installing a flexible culvert may be willing to
tolerate additional handling care as a tradeoff for lighter handling weight. Therefore, flexibility
factors represent gross measures of successful installation experience. However, they are not
sufficiently precise or rigorous to be used as a basis for comparison between culverts of different
material types or wall sectional profiles.
Also note from Eq. 15-16 that no load combinations are explicitly specified and no resistance factors
(or safety factors for ASD) are applied. Hence, flexibility factors do not relate to either Strength or
Serviceability Limit States, but are simply empirical factors that are applied in the same form to
both ASD and LRFD methodologies.
15.3.6 Other Considerations
To assure successful performance, a number of other factors must be considered as well. These
include backfill material and density, differential foundation movement, construction procedures,
end conditions and unsymmetric loading. Design procedures for these factors are generally
independent of either ASD or LRFD methodologies.
15-19
Problem Definition
Objective: Develop of permanent and live loads for structural design of the culvert, and select
structural plate CMP section to meet performance limits for LRFD.
Approach: To perform the culvert design, the following steps are taken:
y
Determine the surface tire contact area for the design truck and design
tandem.
Determine the distribution areas of live load at the culvert crown for the
design truck and design tandem.
Determine the unfactored distributed live load pressure for the design truck,
design tandem and design lane.
Determine whether the design truck load or design tandem load is critical for
design of the culvert.
Calculate the factored vertical earth pressure and total factored vertical
15-20
Select the pipe wall based on seam strength and check for wall buckling and
handling flexibility.
Solution:
Step 1: Develop Design Parameters
From Table 15-4 for the Strength I Limit State, the load factors are:
LL = IM = 1.75
EV = 1.95
As described in Section 4.5.2, a single lane contribution can be conservatively assumed for buried
structures. From Table 4-13, the Multiple Presence Factor, m (A3.6.1.1.2), is:
m = 1.2
Due to the low volume of truck traffic (i.e., ADTT = 1000), the reduction factor, R = 0.95
(AC3.6.1.1.2).
Because the structure will be backfilled and covered using a compacted granular fill material, the
spreading factor, SE = 1.15 (AC3.6.1.2.6).
Using a prism load as a conservative estimate of the vertical earth pressure, the arching factor, Fe =
1.0.
From Table 15-5 for a structural plate alternative, the resistance factor, = 0.67
Step 2: Determine Surface Contact Area for Design Truck and Design Tandem
From Eq. 4-39, the dynamic load allowance, IM, at the ground surface (DE = 0) is:
l = 0.00228 1 +
72.5
P = 0.00228 (1.75) 1 +
100
100
R= 0.405 m
15-21
40
l = 0.00228 (1.75) 1 +
55 = 0.307 m
100
Note the width dimension, w, of the tire contact area is constant for both design vehicles and equals:
w = 0.51 m
Step 3: Determine Dimensions and Distributed Area of Live Load at Crown (DE = 1.0 m)
For the design truck, from Eq. 4-40 and from Steps 1 and 2:
15-22
m R LL AL 1 +
10
(1.2) (0.95) (1.75) (145 kN) (1.24)
100
DTP =
=
= 69.2 kPa
5.18 m2
AD
DTP = 69.2 kPa
Using Eq. 4-36 and AL = 220 kN, the pressure distribution for the Design Tandem is:
IM 3
m R LL AL 1 +
10
(1.2) (0.95) (1.75) (220 kN) (1.24)
100
DTP =
=
= 61.6 kPa
8.83 m2
AD
Using Eq. 4-37, the factored pressure distribution due to the Uniform Lane Load is:
DLP =
The distributed live load pressure distributions at the pipe crown for the Design Truck is presented in
Figure 15-4. The corresponding figure for the Design Tandem is presented in Figure 15-5.
Figure 15-4
Pressure Distribution at Pipe Crown for Design Truck
15-23
Figure 15-5
Pressure Distribution at Pipe Crown for Design Tandem
Step 5: Determine Whether Design Truck and Design Tandem Controls Design
Using Steps 3 and 4, the load for the design truck for LD # Bc is:
(DTP) (LD) = (69.2 kPa) (1.56 m) = 108 kN/m
Using Steps 3 and 4, the load for the design tandem for LD # Bc) is:
(DTP) (LD) = (61.6 kPa) (2.66 m) = 164 kN/m
Load for design tandem > load for design truck; therefore, the design tandem governs.
Step 6: Calculate the Factored Load Due to Vertical Earth Pressure
Using Eq. 4-35, calculate the prism load, assuming an arching factor Fe = 1.0, which is conservative
for flexible culverts under yielding foundations.
EV WE = EV [Fe (S Bc H)] = 1.95 [(1.0) (21.4 kPa) (1 m) (3.6 m)]
EV WE = 1.95 [77.0 kN/m] = 150.2 kN/m
Step 7: Calculate Total Factored Equivalent Pressure on Culvert
Using Steps 4, 5 and 6:
15-24
= 251 kN / m
2 0.67 2
From Table 15-10, a 2.77 mm thick steel plate section with a 152 x 50 mm corrugation is selected
(i.e., SS = 251 kN/m < 628 kN/m).
Table 15-10
Minimum Longitudinal Seam Strength
for Steel and Aluminum Structural Plate Pipe
(AASHTO, 1997a)
6 Bolts/mm
(kN/m)
2630
8 Bolts/mm
(kN/m)
2830
From AASHTO LRFD (1997a) Tables A12-3 and A12-10, the following properties are obtained:
A=
r=
I=
Em =
Fy =
Fu =
f cr =
12 E m
kS
r
=
2
12 (2 x 108 kPa)
0.22 x 3.6 m
0.0173 m
15-26
CHAPTER 16
RIGID CULVERT DESIGN
16.1 Introduction
The design and analysis of rigid culverts represent complex soil-structure interaction problems in
which the strength and deformation limits of the structure must be balanced with the reaction
characteristics of the soil. Therefore, the development of methods for the design and analysis of
these structures has involved efforts to quantify various aspects of the soil-structure interaction
behavior controlling their performance.
Beginning with Marston and Anderson (1913), rational methods have been advanced to account for
the important effects of positive and negative arching loads on conduits, trench and embankment
construction methods, and bedding conditions (e.g., Schlick, 1920; Marston, 1930; Schlick, 1932;
Spangler, 1933; Spangler, 1950). In addition, evolving refinements in design and analysis
techniques for reinforced-concrete pipe have enabled greater precision and reliability in the
placement of reinforcement elements to optimize both load capacity and economy (e.g., Paris, 1921;
Olander, 1950; Heger, 1963 and 1982; and ASCE, 1993).
The focus of this chapter is limited to a comparison of ASD (AASHTO, 1997b) versus LRFD
(AASHTO, 1997a) aspects of the design process of rigid (i.e., reinforced concrete) pipe. As such,
the presentation assumes that the participants have some familiarity with the principals of soilstructure interaction and reinforced concrete design. This chapter:
y
16-1
In the Direct Design Method, the loads and installation effects on the culvert are determined in the
same way as for the indirect method. However, the structural responses of a culvert are evaluated
using the principals of reinforced concrete design which take into account the geometry and
constitutive material properties of the structural cross section. Direct Design represents a relatively
recent innovation for RCP and was developed around modern limit state design concepts consistent
with LFD and LRFD. Additionally, direct design is the only approach applicable for the design of
special shapes such as reinforced concrete box culverts.
16.2.1 Indirect Design Method (A12.10.4.3)
In both ASD (AASHTO, 1997b) and LRFD (AASHTO, 1997a), the indirect design methodology
assumes loads at a magnitude consistent with the service limit state (i.e., working loads). Likewise,
structural capacities are taken at classical working stress performance limits. The Indirect Design
methodology does not accommodate different levels of uncertainties for either the load or
resistance components of a structural system.
As for flexible culvert design, rigid culvert design practice specifies the maximum and minimum
cover height and the design live load. Also, the cross sectional size and shape are usually controlled
by hydraulic or other end use considerations. Thus the design process is directed toward selecting
the structural wall section of the culvert. For reinforced concrete culverts, this process requires
evaluation of:
y
y
y
y
Flexure
Thrust
Shear
Radial tension
The Indirect Design Method for reinforced concrete culverts is based on a comparison of the
laboratory determined strength of a given pipe design, known as the Three Edge Bearing Strength
(or D-Load), with an empirically derived field equivalent strength, which is calculated by
multiplying the laboratory strength by a Bedding Factor. Thus, flexure, thrust, shear and radial
tension are accounted for indirectly.
During development of the Indirect Design Method for rigid culverts early in this century, a number
of laboratory test protocols were tried. Eventually, the Three-Edge Bearing Test (ASTM, 1997)
became the standard. An illustration of the test setup is given in Figure 16-1. The result of a ThreeEdge Bearing Test is reported as the D-Load. D-Load is calculated by dividing the total measured
strength of the test specimen by its inside diameter as follows:
D=
TEB
Si
(Eq. 16-1)
16-2
Figure 16-1
D-Load Test Setup
(ASTM, 1997)
where:
D=
TEB =
Si =
The laboratory tests generally measure two D-Load strengths. The first is the load required to
produce a 0.25-mm wide crack which represents the limit of desirable performance based on
considerations of corrosion and longevity. The second is the ultimate load capacity of the section,
which for RCP is between 1.25 and 1.50 higher than the D-Load. The AASHTO ASD and LRFD
Specifications for indirect design of rigid culverts are based on the D-Load strength representing the
0.25 mm crack.
For the Indirect Design Method, the nominal design load carrying capacity of the pipe, as
represented by the D-Load, must equal or exceed the Service Limit State design load on the pipe, or:
D = iQi
(Eq. 16-2)
16-3
1 W + WF
D = E
S i B FE
where:
BFE =
BFLL =
Si =
WE =
WF =
WL =
Because the design of rigid (RCP) culverts involves a structural rather than geotechnical design
procedure, the Direct Design Method in ASD follows the Load Factor Design (LFD) methodology
used for structural design of all reinforced concrete structures in the AASHTO ASD Specification.
In LFD, like LRFD, loads are modified by load factors and ultimate strengths are modified by
strength-reduction factors to account for uncertainty.
In ASD, the load and resistance elements of RCP culverts must satisfy the general LFD relationship
given by:
i Qi R n
(Eq. 16-4)
where:
Rn =
=
=
i =
Qi =
Design of rigid culverts is influenced primarily by vertical earth pressures and, for the case of
shallow cover, by vehicular live loads. The magnitude of vertical earth pressure on rigid culverts
16-4
can vary significantly, depending on the degree of soil-structure interaction taking place. In general,
rigid culverts are much stiffer than the surrounding soil and attract greater load than flexible
culverts.
For rigid culvert design by ASD, the applicable load factor is = 1.3 and the applicable load
coefficients are summarized in Table 16-1.
Table 16-1
Load Coefficients/Factors for LFD of Rigid Culverts
(after AASHTO, 1997b)
Load Type
Vertical Earth
Pressure
Vehicular Live
Load and Impact
Load Coefficient
Effective Load
Factor,
1.00
1.30
1.67
2.17
In LRFD, the load and resistance elements of rigid culverts must satisfy the general LRFD
relationship given by:
Q
i
R n
(Eq. 16-5)
where:
i =
i =
Qi =
=
Rn =
The load factors and load combinations used in the AASHTO LRFD Specification for design of
rigid culverts were presented in Chapter 4 and are summarized in Table 16-2.
For rigid culvert design at the Strength Limit States, the following values of normally can be used:
D = R = 1.0
I =
1.05 for structures deemed operationally important, 1.00 for typical structures
and 0.95 for relatively less important structures
16-5
Strength I
Load Factor,
Limit State
Service I
Construction
1.30
0.90
1.00
1.00
1.00
1.00
1.75
1.75
1.00
1.00
1.50
1.50
Load Type
Earth Loads:
y Max. Vertical Earth Press (EV)
y Min. Vertical Earth Press. (EV)
Vehicular Live Loads:
y Design Truck or Tandem (LL)
y Dynamic Load Allowance (IM)
Determination of the operational importance of a structure is made by the facility owner as described
in Chapter 4. For the purpose of this chapter, the value of I is assumed equal to 1.0.
At any Service or Construction Limit State, D, R and I = 1.0.
In addition to the difference between load factors used in LFD (Table 16-1) and LRFD (Table 16-2),
there is also a significant difference between the methods to compute live load pressures on a culvert
due to revision of the live load model as discussed in Section 4.5.2. As discussed therein and shown
on Figure 16-2, the net effect of the revised load factors and live load model is to increase the design
live load by as much as 100 percent.
250
200
150
2000
3500
5000
16-6
The limit states that must be considered for design of small-diameter, round reinforced concrete pipe
are listed in Table 16-3.
Table 16-3
Limit States Applicable for Design of Reinforced Concrete Pipe
Strength I
Service I
Performance Limits
Limit State
Limit State
Wall Flexure
U
Wall Thrust
U
Wall Shear
U
Radial Tension
U
Wall Cracking
U
16.2.2.2 Resistance Factors (A12.5.5)
Resistance/strength-reduction factors specified in the AASHTO ASD and LRFD Specifications for
the design of rigid culverts are presented in Table 16-4. For the most part, resistance factors for
rigid culvert design have been established at the same values used for other reinforced concrete
members. For flexure, the AASHTO LRFD Specification assumes the same level material
variability for cast-in-place construction and in-plant manufacturing processes. Therefore, = 0.90
is specified for both.
Table 16-4 (excerpt from A12.5.5-1)
Resistance/Strength-Reduction Factors for Precast Reinforced Concrete Culverts
(after AASHTO, 1997a; AASHTO, 1997b)
Resistance/Strength
Performance Limit
Reduction Factor
AASHTO Type I Installation
0.90
y Flexure
0.82
y Shear
0.82
y Radial Tension
Other Type Installation
1.00
y Flexure
0.90
y Shear
0.90
y Radial Tension
16-7
The following sections review the basic design aspects of the LRFD Direct Design Method and
some of the variability associated with the method. However, to assure successful performance of
rigid culverts, many factors must be considered beyond the design basis presented above for ASD
and LRFD methodologies. Some of these include backfill material type and density, differential
settlement, foundation consolidation, construction practices, and end treatments. Design procedures
and specifications for these factors are generally independent of the design method.
16.3.1 Elastic Load Models (A12.10.4.2.1, A12.10.2)
The design load depends upon the assumptions used to develop the live and dead load contributions
and their distribution through the soil. This load intensity can vary significantly due to a number of
factors which apply to both flexible and rigid culverts. However, for rigid culverts, many more
factors must be considered in the design including:
y
y
y
y
y
These factors significantly increase the generality and precision of the model. However, its accuracy
is only as good as the individual assumptions associated with each of the parameters listed above. In
addition to load intensity, some of the other significant uncertainties include installation related
factors such as bedding type, trench width and relative soil stiffness. On the other hand, the
variation in culvert diameter due to manufacturing processes is very small. Specifications limit
variations in diameter to 1.5 percent. Also, variations in minimum concrete strength are limited to
10 percent.
Under the direct design methodology employed in the LRFD specification for rigid culverts,
structural responses of the RCP are evaluated from an elastic analysis on an assumed culvert wall
section. Once the moment, thrust and shear structural responses are obtained, the area of
reinforcement steel is determined. Then, the assumed cross section is checked for adequacy. This
process is iterated until convergence is reached.
16-8
The elastic analysis of the pipe ring is generally performed using an assumed pressure distribution
such as proposed by Paris (1921) or Olander (1950). These are illustrated in Figure 16-3. It is
noteworthy that between the two, Paris= distribution produces somewhat higher moments. However,
Olander=s distribution produces significantly higher thrusts and heaver designs for many cases.
These distributions are only applicable to round shapes.
Recent work with finite element analysis of soil-structure interaction has resulted in the development
of generalized computer models that are applicable to any shape and relative stiffness and do not
require an assumed pressure distribution. Heger has utilized such a computer model to derive a new
assumed earth pressure distribution for round pipe (ASCE, 1993) shown in Figure 16-4.
As RCP culvert walls approach their ultimate limit state, they will try to shed their load into the
surrounding soil. The surrounding soil picks up this load in a proportionate share to its relative
stiffness. And, its stiffness decreases significantly as shear stress in the soil increases. This load
transfer manifests itself by greater nonlinearity in wall bending moments than it does in wall thrusts.
Therefore, failure to account for the ultimate load level and the nonlinearity in the analysis can
result in at least two shortcomings, less accuracy and unknown and potentially unconservative
factors of safety. This uncertainty is mitigated when the nonlinear effects of ultimate limit states and
soil-structure interaction are carefully accounted for in the analysis, such as with nonlinear finite
element models.
16-9
Once the pipe wall responses are determined, the reinforced concrete wall section can then be
designed using techniques that are well documented in other sources (AASHTO, 1997a and 1997b).
However, these techniques and the force effects of the load distributions shown in Figures 16-3 and
16-4 cannot be evaluated readily by manual calculations. Hence, computer programs have been
developed that perform both the load distribution and the wall section design (McGrath, et al.,
1988).
A brief description of the reinforced concrete pipe design process is presented in the following
sections.
16.3.2 Flexure and Thrust (A12.10.4.2.4)
The direct design equations developed by Heger (1982) utilized ultimate strength methodology. The
primary criterion was flexure induced by the design loads, based on the Whitney stress block. To
assure some level of ductility in pure bending, the section is proportioned such that the steel
reinforcement reaches yield before the concrete reaches ultimate compression. However, under deep
burial where compressive thrust is predominate, the section must be designed with compression
reinforcement and radial ties, such as for a column that resists combined bending and axial loads.
16-10
Like flexible culverts, the governing load scenario for rigid culvert is sometimes based on
installation stresses rather than permanent loads. This circumstance is accounted for by empirically
derived relationships for minimum reinforcement that are based only on diameter and wall thickness.
In addition, a lower limit is applied to these minimum reinforcement relationships to represent
practical limits for manufacturing.
16.3.3 Shear (A12.10.2.5)
Heger and McGrath (1982) found that conventional ACI relationships for beam shear in pipe, box
sections and slabs (also known as diagonal tension) were unconservative under concentrated loads
and excessively conservative under distributed loads. They proposed new provisions to more
accurately represent the shear capacity of such members. These provisions take into account such
factors as:
y
y
y
Crack depth
Curvature
Wall thrust
Wall sections that are found to be inadequate to resist beam shear must be increased in thickness, or
radial ties must be utilized.
16.3.4 Radial Tension (A12.10.4.2.6)
Bending moments that produce tension on the inside of a pipe also produce radial tension (also
known as bowstring tension) that tries to straighten out a curved section of reinforcement. The
result of excessive radial tension is the delamination of the reinforced section, which is manifested
by spalling of concrete to the depth of the inner layer of reinforcement. Based on curved slab and DLoad tests, Heger and McGrath (1982) developed design provisions to account for radial tension in
the design. When radial tension exceeds the capacity of the section, the wall thickness must be
increased or radial ties must be employed.
16.3.5 Crack Width (A12.10.4.2.4d)
The control of tension crack formation under service loads is required in both ASD and LRFD
versions of the AASHTO specifications. Also, both versions limit the width to 0.25 mm extending
over a maximum length of 305 mm. However, this control is accomplished in fundamentally
different ways. Under ASD, the laboratory applied load that produces a 0.25 mm crack is
experimentally determined. Under LRFD, sufficient reinforcement is specified to limit the crack
width. An analytical check of crack width is an iterative process. The factors affecting the amount
of reinforcement to control cracking include:
y
y
y
y
y
y
Bending moment
Axial thrust
Cross section geometry
Pipe diameter
Reinforcement type and placement
Concrete compressive strength
16-11
16-12
CHAPTER 17
SUMMARY
17.1 Summary and Overview
In Chapter 1, the stated objectives of this manual were to provide a basis for developing an
understanding of:
y
Process for selection and application of load factors and load combinations
Except for the manner in which uncertainty is embodied in a factor of safety in ASD and by load
and resistance factors in LRFD, very little difference exists between ASD and LRFD.
Where reliability-based calibrations exist, the level of safety in structure and substructure
components is comparable, which should result in more efficient, and possibly cost effective
designs. Where calibrations are based on direct correlation with ASD, LRFD should result in
designs generally comparable to existing ASD practice.
The importance of site characterization and selection of geotechnical design parameters is critical
regardless of whether substructure design is accomplished by ASD or LRFD. In Chapters 5 and 6,
the reliability of various methods for developing geotechnical properties for design is demonstrated
for various in-situ and laboratory test methods. The reliability of various semi-empirical methods
used directly for design was commented upon in certain of the design chapters. In ASD, the
reliability of the basis for development and selection of geotechnical properties is sometimes
reflected in a variable factor of safety. In LRFD, each method of property estimation is assigned a
resistance factor which reflects the relative reliability of the estimation method.
Methods used to calibrate the LRFD Specification were discussed in many of the chapters. The
information provided in these chapters should help facilitate modification and development of
17-1
resistance factors by user agencies to reflect local design practices and to incorporate new design
approaches.
Processes for design of substructure elements using LRFD principles and the AASHTO LRFD
Specification in particular have been presented. Repeating an earlier point, very little difference
exists between ASD and LRFD except for the manner in which uncertainty is embodied in a factor
or safety in ASD and by load and resistance factors in LRFD.
Lastly, what does the future hold for LRFD? LRFD is used or under development in other parts of
the world. Two examples include the Ontario Bridge Code and EuroCode 7. Also, the LRFD
Specification should be considered as a "Work in Progress". As with any design specification,
articles are revised and new are added to address mistakes, changes in design philosophy, and to
introduce new technologies and design approaches. For the past two years, AASHTO has supported
maintenance of the specification (i.e., NCHRP 12-42 - LRFD Bridge Design Specifications Support)
to recommend changes and develop modifications to the LRFD Specification to assist the efforts of
AASHTO Technical Committee T-15 on Foundations and Substructures in their charge to oversee
efforts to update and revise the AASHTO specifications in this subject area. Other areas of
sponsored research and study of LRFD include:
y
NCHRP 20-5, Topic 28-02 - This project, to be completed by late 1998, will
result in publication of a report synthesizing a survey of limit state design
practice in Canada and Europe
In addition, reliability-based resistance factors may be revised as more data regarding substructure
design and performance are gathered. Another area which could be revised in the future is
application of load and resistance factors for the Service Limit State (i.e., other than = 1.0 and =
1.0) to reflect the reliability of methods used to estimate soil and rock deformation characteristics
and substructure movements. However, the most useful input in this process is input from you, as a
user of the document. By directing your constructive comments to the Chair of the AASHTO T-15
Technical Committee, necessary modifications and updates can be accomplished and incorporated
into future editions of the LRFD Specifications.
17-2
APPENDIX A
METRIC STANDARDS AND UNIT CONVERSIONS
A.1 Introduction
Equations and calculations presented in this publication use the metric or International System of
Units (SI). The following sections of this appendix provide:
y
y
y
This appendix describes the metric units considered standard by FHWA (1995) and ASTM (1993).
In some cases, these standard units differ from standard units used in the AASHTO LRFD Bridge
Design Specifications - SI Units (1997a).
A.2 Standard Metric Units and Conversions with English Units
Because a variety of metric units can be used in practice, care should be taken to assure that all
calculations and equations are dimensionally consistent. Metric dimensions consist of base and
derived units. The base and derived metric units used throughout this course are described in the
following sections.
A.2.1 Base Units
Base units commonly used in geotechnical engineering consist of those dimensions used to define
length, mass, time and temperature. Base units are defined as dimensionally independent units used
to derive all other measurement units.
A.2.1.1 Length
The standard metric unit for length is the meter, m. Other commonly used metric units of length
include:
y
y
y
y
FHWA, ASTM and AASHTO discourage use of the term "metric ton" for engineering documents,
but agree that the metric ton (t) is acceptable for common use.
A.2.1.3 Time
The standard metric unit for time is the second (s).
A.2.1.4 Temperature
The standard metric unit for thermodynamic temperature is the degree Kelvin (K). The degree (o)
symbol is not used with K.
A.2.2 Derived Units
Derived units commonly used in geotechnical engineering consist of those dimensions used to define
force, pressure, mass density, unit weight, bending or overturning moment, and temperature.
Derived units are defined as units formed by combining base units and/or other derived units by
means of multiplication, division or the use of exponents.
A.2.2.1 Force
The standard metric unit for force, or mass (M) times acceleration of gravity (g), is the newton (N =
kg-m/s2). Other metric units commonly used include:
y
y
A.2.2.2 Pressure
The standard metric unit for pressure is the pascal (Pa = N/m2). Other metric units commonly used
include:
y
y
Symbol
G
M
k
m
Power
109
106
103
10-3
10-6
6 in x
1 ft
0.3048 m
x
_ 0.1524 m
12 in
1 ft
A "hard" conversion is typically used when a convenient rounded metric equivalent is desired, such
as for specifying requirements for minimum foundation embedment or structure dimensions. As an
example of a "hard" conversion:
6 in = 0.1524 m = 0.15 m
English to metric conversions commonly required for geotechnical problems are provided in Table
A-3.
A-3
Table A-2
Metric Units for Common Geotechnical Problems
Variable
Symbol
Length
Width
Height or Thickness
Depth
Diameter
Area
Volume
Time
Velocity
Acceleration
Acceleration Due to Gravity (g
= 9.81 m/s2)
Mass
Mass Density
Unit Weight
Settlement
Pressure or Stress
Elastic Modulus
Grain Size
Coefficient of Consolidation
Cohesion or Shear Strength
Hydraulic Conductivity
Force or Resistance
Moment of Force
L, R
B, b
H, h
D, z
D, d
A
V
t
v
a
Base Unit
Dimensions
L
L
L
L
L
L2
L3
T
LT-1
LT-2
LT-2
m/s2
varies
varies
D, d
cv
c,
k
varies
varies
M
ML-3
ML-2T-2
L
ML-1T-2
ML-1T-2
L
L2T-1
ML-1T-2
LT-1
MLT-2
ML-2T-2
kg
kg/m3
kN/m3
mm
kPa
kPa
mm
m2/s
kPa
m/s
kN
kN-m
A-4
SI Unit Symbol
m
m
m
m
m
m2
m3
s
m/s
m/s2
Table A-3
Common English to Metric Unit Conversions
English Unit
Metric Equivalent(1)
1 in
25.40 mm*
Length
1 ft
0.3048 m*
2
1 in
645.2 mm2
Area
1 ft2
0.092 m2
0.028 m3
1 ft3
Volume
1 cy
0.764 m3
Area/Unit Length
1 ft2/ft
0.3048 m2/m*
Mass
1 lbm
0.4536 kg
3
Mass Density
1 lbm/ft
16.02 kg/m3
Acceleration Due to Gravity
1 g = 32.174 ft/sec2
9.807 m/s2
1 lb/ft3
157.1 N/m3
Unit Weight
3
1 kip/ft
157.1 kN/m3
1 lb
4.448 N
Force
1 kip
4.448 kN
Moment
1 kip-ft
1.356 kN-m
Force/Unit Length
1 kip/ft
14.59 kN/m
2
6.895 kPa
1 lb/in
6.895 MPa
1 kip/in2
Pressure, Stress
1 lb/ft2
0.0479 kPa
47.88 kPa
1 kip/ft2
2
0.1571 kPa/m
1 lb/ft /ft
Pressure/Unit Length
2
157.1 kPa/m
1 kip/ft /ft
(1)
Asterisk indicates exact conversion; all other conversions are rounded to four significant figures.
Quantity
L is used as the symbol for length to avoid confusion with the abbreviation
for the unit symbol for liter AL@
m is used as the symbol for mass to avoid confusion with the length unit
symbol for meter Am@
A-5
Velocity (v)
Acceleration (a)
Mass density ()
Unit weight ()
LT-1
LT-2
ML-3
ML-2T-2
Text Style
Unit symbols are always presented as upright text, regardless of the type style used in the
surrounding text. For example:
Incorrect:
Correct:
Plural Expression
SI symbols are unaltered in the plural. Therefore, SI unit symbols should be presented as singular
and unit names written as plural. For example:
Incorrect:
Correct:
Correct:
Periods
SI symbols and prefixes are symbols and not abbreviations and, therefore, should not have periods
that follow except when they fall at the end of a sentence. For example:
Incorrect:
Correct:
km
kg
kilometer
kilogram
A-6
Unit symbols derived from a proper name, in which case the unit symbol is
presented in upper case and the unit name is written in lower case.
Incorrect
n
j
w
Correct
Newton
Joule
Watt
N
J
W
newton
joule
watt
The unit name "Celsius". Unlike the other units derived from proper names,
the unit name Celsius must always be capitalized, Also, the degree symbol
"o" must always accompany the unit symbol for Celsius (oC).
Correct:
C degree Celsius
The symbol for the liter unit. Because the letter l can be easily confused for
the numeral 1, an upper case L for liter is recommended. The same rule for
symbols derived from proper names applies to the symbol for liter.
Correct:
L liter
Correct
35m
250kg
20oC
35 m
250 kg
20 oC
Correct
k Pa
km
kilopascal
kilometer
kPa
km
Symbol prefixes should be presented in lower case except when they represent a unit quantity
A-7
giga = 109
mega = 106
kilo = 103
Incorrect
Mega Newton
Kilo Newton
Kilo Pascal
Mega Pascal
mn
Kn
Kpa
mpa
unity
= 1.0
m
milli = 10-3
micro = 10-6
Correct
meganewton
kilonewton
kilopascal
megapascal
MN
kN
kPa
MPa
Product Units
Product unit names are written with a space or hyphen between the individual factor unit names.
The product unit symbols are presented as dot products of the individual factor units. For example:
Incorrect
Correct
newtonmeter
N*m
kilo newton meter kN*m
kilogrammeter
kg*m
Quotient Units
Quotient units should only be written with the word "per" between each factor unit (i.e., unit per
unit). Quotient unit symbols can be correctly presented with either a divisor (slash or straight bar)
between individual factor unit symbols or as a dot product with the denominator unit to a negative
power. For example:
Incorrect
Correct
meters/seconds, m*s
newtons-meter, N-m
m
s
The numbers in the "international" column have been aligned with the location of the decimal point.
US
26,345.0
3.141596
2,123,987.23
34.34523
24,246,680.0
International (SI)
26 345.0
3.141 596
2 123 987.23
34.345 23
24 246 680.0
The exceptions to this practice are on engineering drawings and financial statements. If the numbers
contain a maximum of four digits on either side of the decimal point, a space is not necessary except
for uniformity in tables. Examples are given below:
Incorrect
0.133 5
2 345
8976.345 6
6 234.5678
Correct
0.1335
2345
8976.3456
6234.5678
Correct
2 345
+56 234.567 8
+6 056.123 345
64 635.691 145
Correct
.1234
0.1234
Fractions
Fractions are not used in the SI system. Non-integer numbers should be expressed as decimals.
Incorrect
Correct
22
3c
2.5
3.125
A-9
A.3 Examples
1. Which soil has a higher unit weight, Soil A with = 110 lb/ft3 or Soil B with = 17.3 kN/m3 ?
= 110 lb/ft3 = 110 lb/ft3 x 0.157 kN/m3 = 17.3 kN/m3
the unit weight of Soil A and Soil B are equal
2. If the total unit weight of a soil submerged below the water table is 20 kN/m3, what is the vertical
effective soil stress at a depth of 5 m?
' = 'z = 20 kN/m3 x 5 m = 100 kPa
3. What is the ultimate load capacity of a spread footing with an area, A, equal to 2.5 m2 and a
nominal bearing resistance, qN equal to 420 kPa?
QR = qR x A = 420 kPa x 2.5 m2 = 1050 kN = 1.050 MN
A-10
APPENDIX B
SOLUTIONS TO STUDENT EXERCISES AND PROBLEMS
B.1 Chapter 3 Student Exercise
1.
2.
Rn
Qi
FS
(Eq. 3-2)
where:
Rn = Nominal resistance
Qi = Load effect
FS = Factor of safety (dim)
3.
(Eq. 3-3)
where:
=
Rn =
i =
Qi =
=
4.
What are the advantages and disadvantages of both LRFD and ASD?
ASD Advantages:
ASD Disadvantages:
1)
ASD does not adequately account for variability of loads and resistances.
Dead, live and environmental loads are all considered to be have no
variability (i.e., deterministic). The factor of safety is applied only to
resistance.
B-1
2)
3)
LRFD Advantages:
1)
2)
Achieve relatively uniform levels of safety based on the strength of soil and
rock for different limit states and foundation types.
3)
LRFD Disadvantages:
5.
1)
2)
Resistance factors vary with design methods and are not constant.
3)
6.
Determine the appropriate resistance factor through calibration with ASD for the following
parameters:
FS =
D =
L =
QD/QL =
3.0
1.25
1.75
2.0
7.
(Eq. 3-6)
Define the mean, standard deviation and coefficient of variation of a set of normally
B-2
distributed data.
1) mean, x = xi/N
where:
xi = data set
N = Number of data values (dim)
2) Standard deviation, = ( x i - x )2 / (N - 1)
3) Coefficient of variation, COV = /x
8.
9.
Rm/Rn
Bias factor (dim)
Measured capacity
Predicted (nominal) capacity
(Eq. 3-11)
(Eq. 3-12)
where:
x =
Mean value defined by Eq. 3-7
COV = Coefficient of variation defined by Eq. 3-9
ln ( ) = Natural logarithm of the expression in parentheses
10.
Reliability index, , is defined as the number of standard deviations g between the mean value, g,
and the origin (i.e., = g/g.}
where:
g (R, Q) = R - Q
11.
For lognormal distributions of load and resistance, recall that the reliability index, , can be
expressed as follows:
B-3
(Eq. 3-21)
2<<6
(Eq. 3-22)
A subsurface exploration and laboratory testing program including field vane shear tests (VST) and
unconfined compression (UC) tests results in a mean ultimate pile capacity prediction, R = 1000 kN
as shown below. The mean unfactored load on the pile, Q = 333 kN. The reliability of each test
method is somewhat different, with a COVR = 0.25 for the VST and a COVR = 0.35 for the UC tests.
For these load and resistance characteristics, determine the ASD factor of safety and the LRFDbased probability of failure based on VST and UC testing.
R =
Q =
COVQ =
COVR (VST) =
COVR (UC) =
1000 kN
333 kN
0.15
0.25
0.35
1)
2)
LRFD: VST: =
R
Q
1000 kN
= 3.0
333 kN
= 3.74
2 < <6
pf = 460 exp(-4.3)
pf = 4.6 x 10-5 . 1 in 22,000
UC: =
= 2.83
pf = 460 exp (-4.3)
pf = 2.4 x 10-3 . 1 in 400
B-4
The maximum load factors are used for all components of load for bearing.
LS is included over the heel of the wall for evaluation of bearing.
All the applicable loads have a load factor of 1.00 and the analysis results
are compared to a project specific settlement tolerance.
The appropriate load factors for the critical load combinations for the performance limits indicated
are tabulated below.
Limit State and Performance Limit
Strength I
Sliding
Bearing
Overturning
1.75
1.75
1.75
Service I
Settlemen
t
1.00
1.75
1.75
1.75
1.00
1.50
1.50
1.50
1.00
1.00
1.35
1.00
1.00
0.90
1.25
0.90
1.00
Load Effect
Summary:
Consistent with the example in Section 4.6, this exercise illustrates the selection of load factors for
geotechnical design of a reinforced concrete cantilever wall. For such walls, dead earth and
concrete loads, lateral earth pressures and vertical and horizontal live load surcharge
predominate. Load factor combinations resulting in the maximum vertical load generally control
for bearing pressure. Load factor combinations which include minimum vertical loads and
maximum horizontal loads result in the greatest net overturning moment (and therefore greatest
base pressure resultant eccentricity) and the lowest vertical stress (and therefore sliding
resistance) at the bearing level.
B-5
B-6
Figure 4-21
Schematic of Student Problem
B-7
Figure 4-22
Retaining Wall Area Designation for Weight of Concrete
(B) Vertical Earth Pressure (EV)
Unit Weight of Soil 1 = 18.835 kN/m3
B-8
The active earth pressure coefficient ka for a wall friction angle, = f = 31o, and a horizontal
backslope is, from AASHTO (1997a) Table 3.11.5.3-1:
k = ka = 0.29
From Eq. 4.7, the lateral earth pressure due to the live load surcharge is:
p = ks fs heq = ka f1 heq= (0.29)(18.835 kN/m2)(0.907 m) = 4.95 kPa
Using a rectangular distribution, the live load lateral earth pressure resultant is:
PLS = p H = (4.95 kN/m2) (5 m) = 24.8 kN/m of wall length
The horizontal and vertical components of the live load lateral earth pressure are:
PLSh = PLS cos = (24.8 kN/m)(cos 31o) = 21.3 kN/m
PLSv = PLS sin = (24.8 kN/m)(sin 31o) = 12.8 kN/m
(D) Lateral Earth Pressure (EH)
The lateral earth pressure is assumed to vary linearly with the depth of soil backfill as given by:
p = kh s z = ka s z = 0.29 s z
(Eq. 4-2)
B-9
The resultant of the basic lateral earth pressure (triangular distribution) acting on the wall is:
PEH = Pa = 0.5 pH = (0.5)(27.3 kN/m2)(5.00 m) = 68.3 kN/m length of wall
The horizontal and vertical components of the lateral earth pressure are:
Pah = Pa cos = (68.3 kN/m)(cos 31o) = 58.6 kN/m
Pav = Pa sin = (68.3 kN/m)(sin 31o) = 35.2 kN/m
(E) Summary of Unfactored Loads
Table 4-14
Unfactored Vertical Loads and Resisting Moments
Item
V
kN/m
Moment Arm
About Toe (m)
Moment About
Toe (kN-m/m)
W1
31.8
0.85
27.0
W2
10.6
0.63
6.7
W3
35.3
1.50
53.0
PEV
169.5
2.00
339.0
PLSV
34.2
2.00
68.3
PLSv
12.8
3.00
38.4
Pav
35.2
3.00
105.6
TOTAL
329.4
Table 4-15
Unfactored Horizontal Loads and Overturning Moments
Item
H
(kN/m)
Moment Arm
About Toe (m)
Moment About
Toe (kN-m/m)
PLSh
21.3
2.50
53.3
Pah
58.6
1.67
97.9
B-10
Strength I - Basic load combination related to the normal vehicular use of the
bridge without wind. (Applicable as a standard load case).
Strength II - Load combination relating to the use of the bridge by Ownerspecified special design vehicles and/or evaluation permit vehicles, without
wind. (Not applicable because special vehicle loading is not specified).
Strength IV - Load combination relating to very high dead load to live load
force effect ratios exceeding about 7.0 (e.g., for spans greater than 75 m).
(Applicable because dead loads predominate).
Consequently, only the Strength I, Strength IV and Service I Limit States apply to the
retaining wall design. Therefore, from Tables 4-10 and 4-11, select the applicable load
factors and combinations and present them in Table 4-16. (Note: Strength I-a and I-b
B-11
represent the Strength I Limit State using minimum and maximum load factors,
respectively, from Table 4-11.)
Table 4-16
Load Factors
Group
DC
EV
LS
EH
(active)
Probable Use
Strength I-a
0.90
1.00
1.75
1.50
EC/SL
Strength I-b
1.25
1.35
1.75
1.50
BC (max. value)
Strength IV
1.50
1.35
1.50
BC (max. value)
Service I
1.00
1.00
1.00
1.00
Settlement
Strength I-a (minimum vertical loads and maximum horizontal loads) will
govern for the case of sliding and eccentricity (overturning)
For the case of bearing capacity, maximum vertical loads will govern, and
the factored loads must be compared for Strength I-b and Strength IV
Group/
Item Units
W1
kN/m
W2
kN/m
W3
kN/m
PEV
kN/m
PLSV
kN/m
PLSv
kN/m
Pav
kN/m
VTOT
kN/m
V (Unf.)
31.8
10.6
35.3
169.5
34.2
12.8
35.2
329.4
Strength I-a
28.6
9.5
31.8
169.5
59.8
22.4
52.8
374.4
Strength I-b
39.7
13.3
44.2
228.8
59.8
22.4
52.8
461.0
Strength IV
47.7
15.9
53.0
228.8
52.8
398.2
Service I
31.8
10.6
35.3
169.5
34.2
12.8
35.2
329.4
B-12
Table 4-18
Factored Horizontal Loads
Group/Item
Units
PLSh
kN/m
Pah
kN/m
HTOT
kN/m
H (Unf.)
21.3
58.6
79.9
Strength I-a
37.3
87.9
125.2
Strength I-b
37.3
87.9
125.2
Strength IV
87.9
87.9
Service I
21.3
58.6
79.9
Table 4-19
Factored Moments from Vertical Forces (Mv)
Group/
Item Units
MvTOT
W2
W3
PEV
PLSV
PLSv
Pav
W1
kN-m/m kN-m/m kN-m/m kN-m/m kN-m/m kN-m/m kN-m/m kN-m/m
Mv (Unf.)
27.0
6.7
53.0
339.0
68.3
38.4
105.6
638.0
Strength I-a
24.3
6.0
47.7
339.0
119.6
67.2
158.4
762.2
Strength I-b
33.8
8.4
66.2
457.7
119.6
67.2
158.4
911.3
Strength IV
40.5
10.1
79.5
457.7
158.4
746.2
Service I
27.0
6.7
53.0
339.0
68.3
38.4
105.6
638.0
Table 4-20
Factored Moments from Horizontal Forces (Mh)
Group/Item
Units
PLSh
kN-m/m
Pah
kN-m/m
MhTOT
kN-m/m
Mh (Unf.)
53.3
117.2
170.5
Strength I-a
93.3
175.8
269.1
Strength I-b
93.3
175.8
269.1
Strength IV
175.8
175.8
Service I
53.3
117.2
170.5
Summary
This example illustrates:
B-13
For cantilever retaining walls, dead, earth and live load surcharge are the
predominate loads
Because the load factor for active horizontal (or lateral) earth pressure in
Table 4-11 is the same for all limit states, the controlling limit states for a
typical cantilever retaining wall are generally those for which dead load
and live load surcharge load factors in Table 4-11 are the greatest (i.e.,
Strength IV and Strength I, respectively)
Minimum load factors typically control for sliding and eccentricity criteria,
because the lower factored soil and concrete dead weights provide less
resistance to sliding and overturning
Maximum load factors typically control for bearing as the higher factored
soil and concrete dead weights exert a higher bearing pressure
The geotechnical (foundation) design for the retaining wall in this example is presented in the
Classroom Example in Chapter 8, Section 8.4.
B-14
1.
2.
Name and briefly describe the two general methods of calibrating resistance factors (Refer to
Section 7.2)?
1)
Use of reliability equations with statistics in soil strengths and loads; that is,
computing resistance factors that will, on average, result in the same reliability (or
probability of satisfactory performance) as achieved under ASD.
2)
Fitting to ASD Specifications; that is, choosing resistance factors that will, on
average, give the same size foundations as would result from design using ASD.
The Asix sigma rule@ can be used to provide rule-of-thumb estimates of the standard
deviation, , and coefficient of variation, COVR. Describe the Asix sigma rule@ (Refer to
Section 7.3.1).
The "six sigma" rule involves three steps using strength values that can be estimated by
an experienced engineer to estimate the statistical parameters and COVR.
1)
Estimate the most likely value of the property (Vest), the lowest conceivable value
(Vmin), and the largest conceivable value (Vmax)
2)
Use the following equation to estimate the value of the standard deviation ():
=
3)
V max - V min
6
COV R =
Vest
B-15
B-16
The ultimate unit bearing resistance of a spread footing foundation in dense sand, qult, has been
estimated at 1000 kPa. The load on the footing at the Strength I Limit State is composed of a dead
load, QD, of 2000 kN (for which D = 1.25) and a live load, QL, of 1000 kN (for which L = 1.75).
Assuming a typical structure (i = 1.0), determine the minimum square footing size needed to satisfy
bearing resistance requirements for the following:
1. LRFD (using Eq. 8-12), if the ultimate resistance was estimated:
a.
b.
c.
qR =
(Eq. 8-12)
2. ASD (using Eq. 8-2), if the minimum required factor of safety against a bearing capacity failure(1)
is:
a. 3.0 [Soil Strength (f) based on SPT]
b.
2.5 [Soil Strength (f) based on Laboratory/Field Strength Tests]
Q Qall = R n /FS _ Q (q ult A)/FS
(Eq. 8-2)
(1)
Minimum factors of safety using FHWA criteria (Cheney and Chassie, 1993); AASHTO ASD
minimum factor of safety is 3.0 for all cases.
B-17
QD
(kN)
D QD
(kN)
QL
(kN)
L QL
(kN)
ASD
Q
(kN)
LRFD
iQi
(kN)
qult
(kPa)
LRFD qR=
qult
(kPa)
ASD
qall=qult/FS
(kPa)
Minimum Required
Footing Size, Areq
(m m)
2500
1000
1750
---
4250
1000
450
---
3.07 3.07
2500
1000
1750
---
4250
1000
550
---
2.78 2.78
2500
1000
1750
---
4250
1000
350
---
3.48 3.48
330
3.03 3.03
400
2.74 2.74
---
1000
---
3000
---
1000
---
---
1000
LRFD: Areq =
ASD: Areq =
---
3000
---
1000
i Qi
qR
Q
qall
B-18
---
Summary
This exercise illustrates the similarities and differences in spread footing design by LRFD and ASD.
Whereas LRFD and ASD result in similar footing dimensions, LRFD considers and accounts for the
reliability of the bearing resistance prediction models. Of note is the greater reliability of semiempirical procedures (which estimate bearing resistance directly from field test results) as compared
to the rational method (which requires first an estimate of f from a field test and then an estimate of
resistance from f).
As evident from the example, use of a more reliable design procedure has the same effect as
reducing the required factor of safety (e.g., from 3.0 specified in AASHTO ASD to about 2.5).
Although reduced factors of safety can be and often are used to reflect greater reliability of soil
strength date (e.g., Cheney and Chassie, 1993), the reduction is generally subjective. Conversely,
the LRFD procedure using reliability-based calibration of resistance factors provides a rational
means for considering reliability in design.
B-19
B-20
You recall from Chapter 4 that the cantilever retaining wall in Figure 8-5 is being considered for a
grade separation between roadway lanes. The wall will be backfilled with a free draining granular
fill such that the seasonal high water table will be below the bottom of the footing. The vehicular
live load surcharge (LS) on the backfill is applied as shown in the figure.
Figure 8-5
During the subsurface exploration, it was determined that the foundation soils are predominantly
clay to a depth of 6 m below the proposed bottom of footing; therefore, a 0.15-m thick blanket of
compacted granular material will be placed below the footing to provide a uniform base for
foundation construction and improve sliding resistance. In performing the wall design, the following
are assumed:
B-21
Dense sand and gravel underlies the clayey foundation soils so that the elastic
settlement of the dense sand and gravel will be negligible
The proposed wall backfill will consist of a free draining granular fill
The seasonal high water table will be at the bottom of the footing
Objective: Demonstrate the procedure for geotechnical design of a spread footing by LRFD and
compare the results to those obtained using ASD.
Approach: To perform the evaluation, the following steps are taken:
Select footing length, L, and width, B, and determine unfactored and factored
bearing pressure distribution
Sliding: For ASD, ensure that unfactored ultimate geotechnical lateral load
capacity, Qn, of the footing divided by the factor of safety, FS, is greater than
or equal to the design load due to lateral load components, Q, and for LRFD,
ensure that the sum of the factored lateral load components, i i Q i , is
less than or equal to the factored geotechnical lateral load resistance, QR =
Qn
Overturning: For ASD, ensure that the resultant unfactored vertical load
component is located within L/6 and B/6 of the footing centroid, and for
LRFD ensure that the factored resultant vertical load component is located
within L/4 and B/4 of the footing centroid.
The factored and unfactored loads and moments for critical load combinations from the example
problem in Section 4.6 of Chapter 4 are presented in Tables 8-10, 8-11, 8-12 and 8-13.
B-22
Group/
Item Units
W1
kN/m
W2
kN/m
W3
kN/m
PEV
kN/m
PLSV
kN/m
PLSv
kN/m
Pav
kN/m
VTOT
kN/m
V (Unf.)
31.8
10.6
35.3
169.5
34.2
12.8
35.2
329.4
Strength I-a
28.6
9.5
31.8
169.5
59.8
22.4
52.8
374.4
Strength I-b
39.7
13.3
44.2
228.8
59.8
22.4
52.8
461.0
Strength IV
47.7
15.9
53.0
228.8
52.8
398.2
Service I
31.8
10.6
35.3
169.5
34.2
12.8
35.2
329.4
Group/Item
Units
PLSh
kN/m
Pah
kN/m
HTOT
kN/m
H (Unf.)
21.3
58.6
79.9
Strength I-a
37.3
87.9
125.2
Strength I-b
37.3
87.9
125.2
Strength IV
87.9
87.9
Service I
21.3
58.6
79.9
Group/
Item Units
W1
W2
W3
PEV
PLSV
PLSv
Pav
MvTOT
kN-m/m kN-m/m kN-m/m kN-m/m kN-m/m kN-m/m kN-m/m kN-m/m
Mv (Unf.)
27.0
6.7
53.0
339.0
68.3
38.4
105.6
638.0
Strength I-a
24.3
6.0
47.7
339.0
119.6
67.2
158.4
762.2
Strength I-b
33.8
8.4
66.2
457.7
119.6
67.2
158.4
911.3
Strength IV
40.5
10.1
79.5
457.7
158.4
746.2
Service I
27.0
6.7
53.0
339.0
68.3
38.4
105.6
638.0
B-23
Group/Item
Units
PLSh
kN-m/m
Pah
kN-m/m
MhTOT
kN-m/m
Mh (Unf.)
53.3
117.2
170.5
Strength I-a
93.3
175.8
269.1
Strength I-b
93.3
175.8
269.1
Strength IV
175.8
175.8
Service I
53.3
117.2
170.5
Step 1: Calculate the Settlement of the Retaining Wall on its Cohesive Foundation
Assume embankment construction has been performed earlier and that consolidation settlement from
the embankment loading has already occurred. The original ground surface is located 1 m above the
footing level.
Divide the 6-m thickness of cohesive soil below the wall foundation into four layers as follows:
H1 = 1.0 m
H2 = 1.0 m
H3 = 2.0 m
H4 = 2.0 m
The depth of the footing below the existing ground surface is 1 m, so the depth to the center of each
layer from the final ground surface in front of the wall is:
d1 = 1.5 m
d2 = 2.5 m
d3 = 4.0 m
d4 = 6.0 m
(A) Calculate the effective overburden pressure at the center of each layer before wall construction
The depth of footing below the existing ground surface is 1 m so that the footing is located at the
annual high water level. The effective overburden stress at the center of each layer is:
oi = d i 2
o1 = (1.0 m)(17.265 kN/m3) +(0.5 m)(7.455 kN/m3) = 21.0 kPa
o 2 = (1.0 m)(17.265 kN/m3) +(1.5 m)(7.455 kN/m3) = 28.4 kPa
B-24
Using Service I loading from Table 8-10, the total vertical unfactored load (VTOT) for calculation of
settlement is 329.4 kN/m length of wall. Assuming the vertical load is uniformly distributed over
the base width of the wall foundation (i.e., B = 3 m), the increase in pressure at the base of the
footing per unit length of wall is:
qo = (VTOT ) /(B) =
329.4 kN/m
= 109.8 kPa
3m
_ +
_ C
Hc
_o 1 + eo
1 + eo
_o
(A10.6.2.2.3c-1)
where:
Ccr = 0.012 and eo = 0.6
Table 8-14
Consolidation Settlement Calculation
Layer
I
(di - 1)/B
!/qo
!
(kPa)
!oi
(kPa)
Hi
(m)
1
2
3
4
0.17
0.50
1.00
1.67
0.95
0.80
0.55
0.34
104.3
87.8
60.4
37.3
21.0
28.4
39.6
54.5
1.0
1.0
2.0
2.0
oi +
Hi log
oi
0.78 m
0.61 m
0.80 m
0.45 m
= 2.64 m
where:
di - 1 = Depth to the center of layer i below the footing from finished grade (m)
!/qo = Boussinesq stress contour at center of a continuous foundation (dim) (A10.6.2.2.3a)
B-25
0.012
(2.64 m) = 0.020 m = 20 mm
1 + 0.6
Assuming the maximum tolerable settlement, Smax = 25 mm, the total estimated settlement is
acceptable.
Step 2: Eccentricity
The eccentricity of the retaining wall is checked in Table 8-15 as described in Section 8.3.4 by
comparing the calculated eccentricity, e, for each loading group to the maximum allowed
eccentricity (emax) using the relationship:
eB =
B/2 - Xo
where:
B/2 = 1.5 m
Xo = Location of the resultant from the toe = (MvTOT - MhTOT)/VTOT
emax = B/4 = 3.0 m/4 = 0.75 m
For each load group, the total vertical forces (VTOT), horizontal forces (HTOT), moments due to
vertical forces (MvTOT) and moments due to horizontal forces (MhTOT) are obtained from Tables 8-10,
8-11, 8-12 and 8-13, respectively. Note that the force and moment due to live load surcharge
over the heel (PLSV) are not included in the eccentricity (i.e., overturning) evaluation (i.e., VTOT
= VTOT(Table 8-10) - PLSV(Table 8-10) and MvTOT = MvTOT(Table 8-12) - PLSV(Table 8-12)).
Table 8-15
Summary of Eccentricity Check
Group/Item
Units
VTOT
(kN/m)
HTOT
(kN/m)
Strength I-a
314.6
125.2
642.6
Strength I-b
401.2
125.2
Strength IV
398.2
Service I
295.2
MvTOT
MhTOT
(kN-m/m) (kN-m/m)
Xo
(m)
eB
(m)
emax
(m)
240.2
1.28
0.22
0.75
791.7
240.2
1.37
0.13
0.75
87.9
746.2
146.9
1.50
0.00
0.75
79.9
569.7
151.2
1.42
0.08
0.50
For all cases, eB is #, = or $ emax (underline correct answer); therefore, the design is
adequate/inadequate (underline correct answer) in regard to eccentricity.
Note that for ASD, emax = B/6, or 0.5 m for a 3.0-m wide footing. Because ASD is equivalent to
the Service I Limit State in table 8-15, the design also meets the ASD eccentricity
requirements.
B-26
L' =
B' =
eB =
Xo =
q =
Note that the force and moment due to live load surcharge over the heel (PLSV) are included in
the bearing resistance evaluation.
Complete Table 8-16 using information from Tables 8-10, 8-12 and 8-13 for each applicable limit
state.
Table 8-16
Summary of Factored Bearing Pressures
Group/Item
Units
VTOT
(kN/m)
MvTOT
(kN-m/m)
MhTOT
(kN-m/m)
Xo
(m)
q
(kPa)
Strength I-a
374.4
762.2
240.2
1.39
135
Strength I-b
461.0
911.3
240.2
1.46
158
Strength IV
398.2
746.2
146.9
1.50
133
Service I
329.4
638.0
151.2
1.48
112
B-27
Sliding of walls on clay is checked under AASHTO LRFD (1997a) using Figure 8-7.
From Section 8.3.3, the factored resistance, QR, against failure by sliding is:
QR = Q + ep Qep (Eq. 8-15)
where:
= 0.85 from Table 8-8
Q = Nominal shear resistance between footing and foundation material (kN)
Qep = Nominal passive resistance of foundation material available
throughout the design life of the footing (kN)
where the cohesion of the foundation soil (c2) = 150.0 kPa and one-half the factored normal stress
(qmax) is given in Table 8-17. As for the eccentricity check, the force and moment due to the
live load surcharge over the heel (PLSV) are not included in the sliding evaluation.
B-28
The actual base pressure (i.e., normal stress at the foundation/soil contact) will have a trapezoidal
shape except when the eccentricity is greater than B/6 (i.e., 0.5 m), at which point the base pressure
distribution becomes triangular and acts over a reduced base width. From Table 8-15, eB < 0.5 m for
all limit states. The values of qmax and qmin in Table 8-17 are calculated, therefore, for a
trapezoidal base pressure distribution as follows:
qmax = (VTOT/B) + [(6)(VTOT)(eB)/B2]
qmin = (VTOT/B) - [(6)(VTOT)(eB)/B2]
where VTOT and eB are obtained from Table 8-15.
Because c2 > qmax/2 in all cases, the normal stress at the footing/soil interface is used in the
calculation of Q.
B-29
For comparison with the total factored horizontal forces (HTOT) from Table 8-11, Q can be
computed as follows:
Q =
Using the relationships for qmax and qmin, the equation above for Q, from Table 8-8, and HTOT
from Table 8-11, complete Table 8-16 for each applicable limit state.
Table 8-17
Summary of Sliding Resistance
Group/Item
Units
qmax
(kPa)
qmin
(kPa)
qmax/2
(kPa)
Q
(kN/m)
Q
(kN/m)
HTOT
(kN/m)
Strength I-a
151.0
58.8
75.5
157.4
133.8
125.2
Strength I-b
168.5
98.9
84.2
200.6
170.5
125.2
Strength IV
132.7
132.7
66.4
199.0
169.2
87.9
Service I
114.1
82.7
57.0
147.6
147.6
79.9
Because the factored sliding resistance ( Q) calculated is greater/less (underline correct answer)
than the factored horizontal loading for all Strength Limit States, the sliding resistance is
satisfactory/unsatisfactory. By ASD (as represented by the Service I Limit State), the factor of
safety against sliding = Q/HTOT = 147.5 kN/m ) 79.9 kN/m = 1.85, which is acceptable as indicated
in Table 8-1.
Summary
Table 8-18
Summary of Spread Footing Design by LRFD and ASD
LRFD
ASD
Factored
Factored
Required FS/
Performance Limit
Actual FS/
Resistance/
Load/
U Eccentricity
U
Eccentricity
Eccentricity
Eccentricity
Limit
Limit
0.07 B
0.25 B
0.03 B
Eccentricity
0.167 B
U
U
Bearing Resistance
282 kPa
158 kPa
3.0
4.2
Sliding Resistance
134 kN/m
125 kN/m
1.5
1.8
As summarized in Table 8-18, a comparable design is achieved by LRFD and ASD. Whereas the
ASD factors of safety for bearing resistance and sliding are fixed, however, the LRFD resistance
factors could possibly be increased with additional data accumulation and reliability-based
B-30
calibration for similar soils and loading conditions. As such, the LRFD provisions reflect the
reliability of the soil strength estimates and capacity prediction models and provide a more rational
basis for design than the ASD provisions. Therefore, with further data accumulation and
calibration, more reliable and economical designs might be achieved using LRFD.
B-31
B-32
Determine appropriate load factors and load combinations and calculate the
total factored load effects
Estimate the unfactored and factored axial resistance of a single pile based on
correlation with the results of CPTs
Estimate the unfactored and factored axial resistance of a single pile based on
the results of full-scale load tests
B-33
Figure 9-2
Pile Group Loading
For the Strength I and Service I Limit States, Eq. 9-6 is used to compute the factored load effect:
i i Q i
Assume a typical structure such that i = 1.0.
Complete Table 9-10 by selecting load factors for the Strength I and Service I Limit States from
Table 4-10 in Chapter 4:
B-34
Limit State
Table 9-10
Load Factors
DC
DW
LL
Strength I
1.25
1.50
1.75
Service I
1.00
1.00
1.00
The total factored load effects are then calculated as follows. For the Strength I Limit State:
i i Q i = i [ DC DC + DW DW + LL LL]
i i Q i = 1.00 [(1.25)(4600 kN) + (1.50)(3900 kN) + (1.75)(3450 kN)]
i i Q i = 17 638 kN
For the Service I Limit State:
i i Q i = i [ DC DC + DW DW + LL LL]
i i Q i = 1.00 [(1.00)(4600 kN) + (1.00)(3900 kN) + (1.00)(3450 kN)]
i i Q i = 11 950 kN
Step 3: Estimate Axial Capacity of Single Pile from SPTs
Figure 9-3 shows the generalized soil profile, pile geometry and idealized SPT blow count profile
for use in design based on several borings performed near the pier location. Using Meyerhof (1976)
(A10.7.3.4.2), the estimated ultimate axial capacity of the pile driven into the soil profile in Figure 93 is:
(Eq. 9-19)
(Eq. 9-20)
and
where:
Qult, Qp, Qs = Ultimate total, tip and side resistance (kN)
qp, qs = Unit tip and side resistance (kPa)
Ap, As =
Area of pile tip and side (m2)
B-35
Figure 9-3
Generalized Problem Geometry and SPT Design Envelope
B-36
Using Eq. 9-17, the ultimate axial resistance of a single pile is:
Qult = Qp + Qs = 1460 kN + 445 kN
Qult = 1905 kN
(Eq. 9-17)
From Eq. 9-18, the factored axial resistance of a single pile is:
QR = Qult = qp Qp + qs Qs
(Eq. 9-18)
From Table 9-5, the resistance factors for the SPT method are:
qp = 0.45
qs = 0.45
Using Nottingham and Schmertmann (1975) (A10.7.3.4.3), estimate ultimate axial capacity of the
pile driven into the soil profile in Figures 9-4 and 9-5.
Qp = qp Ap = (11 300 kPa) (0.46 m/2)2 = 1880 kN
(Eq. 9-19)
The ultimate unit side resistance of the pile is estimated using the CPT sleeve resistance values from
Figure 9-4 and the following relationship:
N2
N1 L
Qs = Ks i f si a si h i + f si a si hi
i =1 8 Di
i = Ni
(A10.7.3.4.3c-1)
For a steel pile with a length to diameter (z/D) ratio of 11.00 m/0.46 m = 23.9, the correction factor,
Ks for a mechanical cone (from Figure A10.7.3.4.3c-1) is 0.39. The incremental values of Qsi along
the pile are presented in Table 9-11.
B-37
Figure 9-4
Representative Pile and CPT Tip Resistance Profile Used in Example
B-38
Figure 9-5
Representative Pile and CPT Side Resistance Profile Used in Example
B-39
Table 9-11
Tabulation of Side Resistance Along Pile Length
Li
Di
fsi
asi
hi
(m)
(m)
(kPa)
(m2/m)
(m)
Pile
Segment, I
Qsi
(kN)
0.643
0.46
345
1445
1.105
1.885
0.46
173
1445
1.380
2.970
0.46
249
1445
0.790
3.523
0.46
403
1445
0.315
NA
NA
403
1445
0.480
11
NA
NA
441
1445
1.000
25
NA
NA
403
1445
2.650
60
NA
NA
307
1445
3.190
TOTAL
Qs:
55
178
(Eq. 9-17)
The factored axial resistance of a single pile is determined from Eq. 9-18 as:
QR = Qult = qp Qp + qs Qs
(Eq. 9-18)
From Table 9-5, the resistance factors for the CPT method are:
qp = 0.55
qs = 0.55
Qs = 1180 kN
Qp = 1900 kN
from which the ultimate axial resistance based on Eq. 9-17 is:
B-40
(Eq. 9-17)
From Eq. 9-18, the factored axial resistance of a single pile is:
QR = qp Qp + qs Qs (Eq. 9-18)
From Table 9-5, the resistance factors for bearing resistance of a single pile based on a load test are
as follows:
qp = 0.80
qs = 0.80
For ASTM A709M, Grade 250, Fy = 250 000 kPa, the allowable axial stress is:
B-41
For LRFD, the factored axial resistance, Pr, of an unfilled steel pipe pile from Eq. 9-21 and Eq. 9-22
is:
Pr = Pn = n As = Fy As
From Table 9-6, the resistance factors for axial compression loading of an unfilled steel pipe pile
are:
For ASD, the number of piles, , required to support the total (unfactored) design load, Q, can be
computed based on Eqs. 9-1 and 9-2 from the following:
Q
Q
;
(Qult /FS)
Pall
where:
Qult/FS =
Pall =
From Table 9-1, the required factors of safety (FS) for axial load resistance based on various
predictive methods are:
Based on these safety factors and the results of Steps 1 through 6, complete Table 9-12 to determine
the number of piles need to resist the design load based on geotechnical and structural resistance by
ASD.
Table 9-12
Determination of Required Pile Group Size Based on ASD
Single Pile
Design
Single Pile
Ultimate
Group
Allowable
FS
Resistance
Load
Resistance
Resistance
or
Qult
Q
Qult/FS or Pall
all
(kN)
(kN)
(kN)
Required
Number of
Piles,
x
Geotechnical:
SPT Method
3.5
11 950
1900
542.9
22
CPT Method
3.5
11 950
2057
587.7
21
2.0
11 950
3080
1540
Structural:
Severe Driving
0.25 Fy
11 950
---
1098
11
Good Driving
0.33 Fy
11 950
---
1450
As shown in Table 9-12, a geotechnical design based on the SPT or CPT methods controls the size
of the pile group, while a pile group size based on the results of a pile load test would not be as
critical as the structural design.
For LRFD, the number of piles, , required to support the factored design load, Q, can be computed
based on Eqs. 9-7 and 9-8 from the following:
i i Qi
QR
i i Qi
Pr
where:
i i Q i
QR =
Pr =
Based on the results of Steps 1 through 6, complete Table 9-13 to determine the number of piles
need to resist the design load based on geotechnical and structural resistance by LRFD.
B-43
Table 9-13
Determination of Required Pile Group Size Based on LRFD
Single Pile Resistance
Factored(1)
Resistance Design Group
Factored
Ultimate
Load
Factor
Resistance
QR or Pr
Q
ult
i i Q i
(kN)
(kN)
(kN)
Required
Number
of Piles,
Geotechnical:
SPT Method
0.45
17 638
1900
855
21
CPT Method
0.55
17 638
2057
1131
16
0.80
Static Load Test
Structural:
0.35
Severe Driving
0.45
Good Driving
(1)
For the Strength I Limit State
17 638
3080
2464
1538
1977
12
9
17 638
17 638
Summary
This example illustrates the design of axially loaded pile groups by LRFD and compares the results
to design by ASD. As shown on Tables 9-12 and 9-13:
The geotechnical resistance controls the design for static methods of axial
capacity prediction and the structural resistance controls the design based
on static load testing for both the ASD and LRFD procedures
Comparison with Table 9-12 shows that the LRFD procedures based on the
reliability-based calibration (i.e., SPT and CPT) are somewhat less
conservative than the ASD procedure for the conditions analyzed
LRFD design procedures based on direct calibration with ASD (i.e., static
load test and structural resistance) provide essentially identical results with
ASD (minor differences occur due to differences in the actual average load
factor and the load factor assumed in the resistance factor calibration)
As indicated in Section 9.2.2.2, the AASHTO LRFD Specifications do not yet satisfactorily
incorporate the impact of field capacity verification (e.g., PDA or load tests) on the reliability of
pile installations for which design capacities are initially predicted using static design methods.
Therefore, this example assumes that the piles are designed on the basis of static methods without
field capacity verification.
B-44
Required
Number of
Piles,
22
CPT Method
3.5
11 950
2057
587.7
21
2.0
11 950
3080
1540
Nordlund Method
3.5
11 950
1385
395.7
31
Severe Driving
0.25 Fy
11 950
---
1098
11
Good Driving
0.33 Fy
11 950
---
1450
Structural:
ult
(kN)
(kN)
(kN)
Required
Number
of Piles,
Geotechnical:
SPT Method
0.45
17 638
1900
855
21
CPT Method
0.55
17 638
2057
1131
16
0.80
17 638
3080
2464
0.60
Nordlund Method
Structural:
0.35
Severe Driving
0.45
Good Driving
(1)
For the Strength I Limit State
17 638
1385
831
22
1538
1977
12
9
17 638
17 638
Summary
This exercise illustrates the application of LRFD in the design of an axially loaded pile group, and
compares the results of LRFD with ASD. From a review of the completed tables above, the
following can be concluded:
B-45
1. The required number of piles is the same by ASD and LRFD for capacity
estimation based on a Static Load Test, as the resistance factor for this method is
based on direct calibration with ASD.
2. For the static (SPT, CPT and Nordlund=s) methods of capacity prediction (for
which resistance factors were developed using a reliability-based calibration),
LRFD results in fewer required piles (as much as 1/3 less for Nordlund=s method)
than ASD. The resistance factors for these methods considers both their variability
and bias, where the bias is a measure of the inherent conservatism, or lack thereof,
of the method. As such, the AEquivalent LRFD Factor of Safety@ for each method
varies somewhat from the typical ASD Factor of Safety (equal to 3.5) as follows:
Method
SPT Method
CPT Method
Nordlund=s Method
FSLRFD
3.33
2.75
2.55
where:
FSLRFD =
Qult
Q
SPT Method
CPT Method
Nordlund=s Method
R
1.62
1.50
1.00
2.22
where:
=
As described in Chapter 7, a bias factor greater than 1.00 indicates that the
predictive method tends to result in conservatively low values of predicted
resistance. (Note that the bias factors tabulated above are applicable only for this
B-46
example.)
4. The required number of piles is essentially the same by ASD and LRFD based on
structural requirements, as the resistance factors for structural design in Table 9-6
are based on direct calibration with ASD. (Note that minor differences do occur
due to a difference between the average load factor used in the exercise of 17 638/11
950 = 1.48, and the average load factor of 1.45 used for calibration of the resistance
factors.)
B-47
D Q D / QL + L
FS (1 + QD / QL)
(Eq. 7-7)
From Section 10.2.1, the AASHTO ASD factor of safety for these type of predictive methods is 2.5.
Solution: The table below shows the variation in the calibrated resistance factor with variations
in the dead and live load factors as well as the ratio of dead to live load.
From Chapter 4, Table 4-11, it can be seen that, for typical bridge structure and retaining wall
foundations, dead load (including earth load) load factors range from 1.25 to 1.50. From the
table below, it appears reasonable to select a resistance factor of about 0.60 for bearing resistance
of drilled shafts in sand for design of most structures.
1.00
Average Dead
Load Factor,
D
1.25
2.00
1.25
1.75
0.57
1.42
3.00
1.25
1.75
0.55
1.38
1.00
1.50
1.75
0.65
1.63
2.00
1.50
1.75
0.63
1.58
3.00
1.50
1.75
0.63
1.56
QD/QL
Live Load
Factor, L
Resistance
Factor,
Average Load
Factor,
1.75
0.60
1.50
B-48
=
qs =
pa =
p =
During the subsurface exploration, it was determined that the foundation soils consist of medium
dense sand to a depth of 13 m below the ground surface, underlain by hard, sandstone bedrock. In
performing the wall design, you can assume the following:
B-49
The anchor derive their capacity wholly within the medium dense sand
Only the end of excavation stage needs to be checked for this problem (Note:
For a complete design, each stage of the excavation should be checked)
Objective: To demonstrate the procedure for anchored retaining wall design by LRFD
Approach: To perform the anchor wall design, you should take the following steps:
Compute and tabulate the unfactored loads and moments required for design
at the applicable limits states
Determine and tabulate the factored loads and moments required for design at
the applicable limits states
Estimate wall movements at the Service I Limit State
Evaluate the global stability of the excavation
Evaluate the earth pressures on the excavation support system
Estimate the required anchor bond zone length
Determine the required soldier pile section using the applicable factored
loads and simplified structural analysis procedures
Determine the required section for the timber lagging
Evaluate the vertical stability of the wall
Solution:
Step 1: Calculate Loads
The first step in determining the loading conditions is to assess the earth pressures on the wall. The
apparent earth pressure diagram shown in Figure 13-4 is selected in accordance with
recommendations made by Peck, et. al (1974) and AASHTO (1997a, 1997b) for wall systems with
multiple levels of support in cohesionless soils (A3.11.5.6). A live load surcharge of 12 kPa
(equivalent to a uniform soil surcharge of about 0.63 m) is used for design. The calculations for
development of the lateral earth pressure and lateral pressure due to vehicular live load surcharge are
shown below.
Unfactored Apparent Earth Pressure (EH):
Pa = 0.65 ka s H = 0.65 tan2(45E - f/2) s H
Pa = 0.65 tan2 (45E - 36E/2) (18.865 kN/m3) (8 m)
Pa = 25.5 kPa
(A3.11.5.6-1)
(A3.11.6.1-1)
The apparent earth pressure and surcharge pressure diagrams are then distributed to the various
members of the system assuming tributary areas extending equidistant between the support levels as
shown below.
The unfactored horizontal force to be supported at Level I is the sum of the apparent earth and
vehicular surcharge lateral pressures from the ground surface down to a depth of 3.5 m (zI = 3.5 m)
as shown below:
Unfactored horizontal load from apparent earth pressure diagram:
P1 = Pa zI = (25.5 kPa)(3.5 m) = 89.25 kN/m
Unfactored horizontal load from vehicular surcharge lateral pressure:
P2 = p zI = (3.1 kPa)(3.5 m) = 10.85 kN/m
Total unfactored horizontal load:
Pa = P1 + P2 = 89.25 kN/m + 10.85 kN/m = 100.1 kN/m
The load factors from Chapter 4 (Tables 4-10 and 4-11) for horizontal earth pressure, EH, and live
load surcharge, LS, are:
EH = 1.50
LS = 1.75
Therefore, the total factored horizontal load is:
P = EH P1+ EH P2 =1.50 (89.25 kN/m) + 1.75 (10.85 kN/m) =152.9 kN/m
The unfactored horizontal force to be supported at Level II is the sum of the apparent earth and
vehicular surcharge lateral pressures from a depth of 3.5 m to a depth of 6.5 m (zII = 3.0 m) as shown
below:
Unfactored horizontal load from apparent earth pressure diagram:
P1 = Pa zII = (25.5 kPa)(3.0 m) = 76.50 kN/m
Unfactored horizontal load from vehicular surcharge lateral pressure:
P2 = p zII = (3.1 kPa)(3.0 m) = 9.30 kN/m
Total unfactored horizontal load:
Pa = P1 + P2 = 76.50 kN/m + 9.30 kN/m = 85.8 kN/m
Applying the load factors for EH and LS, the total factored horizontal load is:
P = EH P1+ LS P2 = 1.50 (76.6 kN/m) + 1.75 (9.30 kN/m) =131.0 kN/m
The unfactored horizontal force to be supported at the base of the wall for design of the soldier pile
B-51
embedment is the sum of the apparent earth and vehicular surcharge lateral pressures from a depth of
6.5 m to the base of the wall (zIII = 1.5 m) as shown below:
Unfactored horizontal load from apparent earth pressure diagram:
P1 = Pa zIII = (25.5 kPa)(1.5 m) = 38.25 kN/m
Unfactored horizontal load from vehicular surcharge lateral pressure:
P2 = p zIII = (3.1 kPa)(1.5 m) = 4.65 kN/m
Total unfactored horizontal load:
Pa = P1 + P2 = 38.25 kN/m + 4.65 kN/m = 42.9 kN/m
Applying the load factors for EH and LS, the total factored horizontal load at the wall base is:
PI = EH P1 + LS P2 = 1.50 (38.25 kN/m) + 1.75 (4.65 kN/m) = 65.5 kN/m
These calculations are summarized in the table below.
10.85
Unfactored
Horizontal
Load
(kN/m)
100.1
Factored
Horizontal
Load
(kN/m)
152.9
76.50
9.30
85.8
131.0
38.25
4.65
42.9
65.5
Unfactored
EH
(kN/m)
Unfactored
LS
(kN/m)
Level I Anchor
89.25
Level II Anchor
Excavation Base
Load
In addition to anchor design and soldier pile embedment, it is necessary to consider load distribution
and load factoring for structural design of the soldier piles and timber lagging. The distribution and
factoring of the apparent earth pressure and vehicular surcharge lateral pressures to the anchors
provides the necessary loading information for analysis of the vertical stability of each pile.
Step 2: Estimate Lateral Wall Deflection and Settlement Profile Behind Wall
Figure 13-3 shows the typical settlement profiles behind a wall of this type assuming good
construction practices are followed. For this problem, it is evident from Curve I in Figure 13-3 that
the maximum settlement at the wall will be about 0.3 percent of the excavation depth or 0.024 m.
The lateral wall deformations should be similar in magnitude to this settlement. It is assumed that
this level of deformation is acceptable for this problem.
Figure 13-3 (AC3.11.5.6-1)
B-52
An analysis of the global stability of the excavation is not presented herein, but would be checked
using a limit equilibrium analysis (A11.8.4.3) and is assumed to be adequate for this example.
Step 4: Preliminary Geotechnical Design of Anchors
As with ASD, design of an anchored wall includes preliminary estimation of anchor capacity to
permit selection of a reasonable anchor spacing and design of facing and other structural
components of the wall. For this problem, you can assume that the anchors will develop their
capacity within the medium dense sand deposit. In accordance with AASHTO (1997a) and FHWA
(Cheney, 1988) recommendations presumptive ultimate anchor capacity, Qa, as provided in Table
13-7 for estimating the length of the bond zone in sand for each anchor. The bond zone lengths are
estimated as shown in the following calculations. The adequacy of anchor capacity is confirmed
during construction in the same manner as for ASD, by means of proof and performance tests. For
the purposes of this example, the structural capacity of the anchor strands is assumed to be adequate.
B-53
Dense
30-50
190
Anchor Level I
Using the anchor inclination, i, of 35E from the horizontal, unfactored and factored horizontal loads
for Anchor Level I from Step 1, the total unfactored and factored anchor loads for a spacing, s, of 2.4
meters along the length of the wall between anchors are:
B-54
Level I
LRFD
Bond Zone Length
(m)
4.8
Level II
4.1
Anchor Level
B-55
Figure 13-5
Lateral Earth Pressure Diagram
for Soldier Pile Embedment into Cohesionless Soil
From Step 1, the load factor for active earth pressure, EH, is 1.50 and for live load surcharge, LS, is
1.75. For a 310 mm HP soldier pile (b = 0.31 m), the factored active pressure from Figure 13-5 is:
Pa2 = EH 0.5 ka s D2 b + (EH ka s H + LS p) D b
Pa2 = 1.5(0.5)(0.26)(18.865 kN/m3)D2(0.31 m)
+ [1.5(0.26)(18.865 kN/m3)(8 m) + 1.75(3.1 kPa)]D(0.31 m)
Pa2 = 1.14 D2 + 19.9 D
and the factored passive pressure from Figure 13-5 is:
Pp = (1.5 kp s D2 b)
Pp = 0.60 (1.5)(4.18)(18.865 kN/m3)(D2)(0.31 m)
Pp = 22.0 D2
Each soldier pile must resist the factored lateral load, Q, computed as the unit factored horizontal
load at the excavation base from Step 1 times the pile spacing, s, of 2.4 m so that:
Q = (65.5 kN/m)(2.4 m) = 157.2 kN
Equating the net factored passive resistance of the embedded vertical element to the load at the
B-56
excavation base:
Q = Pp - Pa2
157.2 kN = 22.0 D2 - 1.14 D2 - 19.9 D
or
0 = D2 - 0.95 D - 7.54
from which the required depth of embedment is:
D = 3.3 m
Step 6: Vertical Stability of Wall
The soldier piles must support the vertical component of anchor load by side and base resistance of
the soldier pile within the depth of embedment.
Tip Resistance
The nominal unit tip resistance, qp, of a pile in cohesionless soil is:
38 N corr D b
< ql
D
(A10.7.3.4.2a-1)
1920
N
N corr = 0.77 log10
v
(A10.7.3.4.2b-1)
qp =
for which:
and:
Ncorr =
N=
D=
qR =
N v =
Representative SPT blow count near the pile tip corrected for Nv (blows/0.3 m)
Measured SPT blow count (blows/0.3 m)
Pile width or diameter (m) Db = Depth of penetration in bearing stratum (m)
Limiting tip resistance taken as 400 Ncorr for sands and 300 Ncorr for silt (kPa)
Effective overburden pressure at pile tip (kPa)
For a pile tip at 3.3 m below the excavation base (or 3.3 m + 8.0 = 11.3 m) below the original ground
surface:
1920
20 = 14.7
Ncorr = 0.77 log10
2
(18.865 kN/ m )(11.3 m)
and
B-57
qp =
(38)(14.7)(3.3 m)
= 5946 kPa < ql = (400)(14.7) = 5880 kPa
0.31 m
qs = 0.96 N
where:
qs =
N =
QR = Qn = qp Qp + qs Qs (Eq. 9-18)
The resistance factor for a single pile for capacity estimation based directly on SPT results is
obtained from Table 9-5 in Chapter 9 as:
qp = qs = = 0.45
from which:
QR = 0.45 (547 kN) + 0.45 (77 kN)
QR = 281 kN
B-58
Factored Load
The total factored vertical load on an individual soldier pile is calculated as follows:
(A6.9.2.2-2)
where:
Pu =
Pr =
Mux =
Mrx =
Muy =
Mry =
The AASHTO LRFD Specification (1997a) provides simplified design equations for flexural
analysis of the vertical supporting elements dependent upon the average factored lateral pressure,
B-59
distance between vertical elements and height of the section of vertical element being considered.
Because the load and resistance factors in LRFD are calibrated for this approach, this example
problem uses these simplified design equations. Using these simplified equations, the maximum
moment, Mmax, will occur below the bottommost row of anchors as:
Mmax = 0.333 p L x2
(AC11.8.5.2-3)
where:
p=
L=
x=
In the AASHTO LRFD Specification (1997a), the average lateral pressure is taken as the average
factored lateral pressure and the maximum flexural moment is the maximum factored flexural
moment. The factored maximum moment below the Level II anchors is:
Mmax = 0.333 p L x2
(AC11.8.5.2-3)
where:
p = (1.5)(25.5 kPa) + (1.75)(3.1 kPa) = 43.7 kPa
Mmax = 0.333 (43.7 kPa)(2.4 m)(3 m)2 = 314.3 kN-m
From Section 6 of the AASHTO LRFD Specification (Steel Structures), structural design of the
soldier piles must also include consideration of the interaction effects from combined axial load and
flexure. From the calculations in Step 6, the factored axial load is 477.0 kN.
From the AASHTO LRFD Specification (1997a), Article 6.5.4.2, the resistance factors for flexure
and compression in LRFD are:
f = 1.00 (flexure)
c = 0.90 (compression)
First trying an HP 310 94 pile, the nominal moment capacity, Mn, of the HP soldier piles is
determined from the AASHTO LRFD Specification (1997a) Articles 6.12.2.2.1, 6.10.6.2 and
6.10.5.2.1-1 and Metric Properties of Structural Shapes (AISC, 1992) as:
Mn = M p
Mn = Zx Fy
Mn = (14.5 10-4 m3)(250 000 kPa)
Mn = 362.5 kN-m
where:
Mp = Plastic Moment resistance kN-m
B-60
Zx =
Fy =
Therefore:
Mr = f Mn = 1.00 (362.5 kN - m) = 362.5 kN - m
From AASHTO LRFD (1997a) Article 6.9.4.1, the nominal axial capacity, Pn, of the soldier pile is:
Pn = 0.66 Fy As
which is applicable for # 2.25, where is the normalized column slenderness factor defined as:
2
K l Fy
=
= 0.045
rs E
for:
As =
Fy =
E=
K=
R=
rs =
Through substitution back into the equation for the nominal axial capacity we find that:
Pn = 0.660.045 Fy As = 0.98 Fy As = 0.98 (250 000 kPa)(11.9 10-3 m2) = 2916 kN
Then the factored axial capacity, Pr is:
Pr = c Pn = 0.90 Pn = 0.90 (2916 kN) = 2624 kN
At this point, we can evaluate the axial resistance component of the interaction equation as:
P u 477.0 kN
=
= 0.18 < 0.2
P r 2624 kN
For this problem, it is reasonable to expect that there will be no significant bending in the plane of
the wall and therefore we can assume that Muy is zero. The factored flexural resistance about the x
axis is:
Mrx = Mn = (1.00) 362.5 kN-m = 362.5 kN-m
Therefore, the moment resistance component of the interaction equation is:
B-61
M ux 314.3 kN- m
=
= 0.87
M rx 362.5 kN- m
Substituting into the interaction equation, we find the following:
8
P u 8 M ux M uy
+
+
= 0.18 + (0.87 ) = 0.95 1
9
P r 9 M rx M ry
(A6.9.2.2-2)
Therefore, this soldier pile provides a section which is adequate for the design loading conditions for
LRFD. In addition to this combined stress check, the slenderness of the soldier beam should be
checked to assure that buckling will not control design. For this example, slenderness is not a
controlling factor.
Step 8: Timber Lagging Design
As for the structural design of the soldier piles, in this problem you can use the simplified design
equations to select a timber lagging section. The maximum factored moment is:
Mmax = 0.083 p L2
(AC11.8.5.3-2)
where:
p = Average factored lateral pressure = (1.50)(25.5 kPa) + (1.75)(3.1 kPa) = 43.7 kPa
L = Lagging span =2.4 m - 0.306 m = 2.094 m (Note: Pile flange width = 306 mm)
Substituting into Eq. AC11.8.5.3-2, the maximum factored moment is:
Mmax = 0.083 (43.7 kPa)(2.094 m)2 = (15.9 kN-m/m)
The factored flexural capacity, Mr, of the lagging is determined by multiplying the resistance factor
of 0.85 for flexure of wood from Article 8.5.2.2 of the AASHTO LRFD Specification (1997a) by the
nominal flexural resistance, Mn, as follows:
Mr = Mn = 0.85 Mn
(A8.6.1-1)
where the nominal flexural resistance of wood, Mn, from Article 8.6.2 of the AASHTO LRFD
Specification (1997a) is:
Mn = Fb S Cs (A8.6.2-1)
and:
Fb =
S=
B-62
(Table AC8.4.1.1.4-1)
Cs =
Evaluation of the size effect factor and the specified resistance in flexure both require calculations
which are not presented herein, although the results shown are consistent with the material and
geometry specified. For this problem, use Select Structural Southern Pine 102 mm thick and 254
mm wide for the timber lagging. For this material, Fb = 29 000 kPa, S = 4.4 10-4 m3, Cs = 1.0 and
the nominal flexural resistance of the lagging is:.
Mn = (29 000 kPa)(4.4 10-4 m3)(1.0) = 12.8 kN-m
The factored flexural resistance, Mr, is then:
Mr = 0.85 (12.8 kN-m) = 10.8 kN-m
The maximum factored moment on a 254 mm wide lagging section is:
Mmax = (15.9 kN-m/m) (0.254 m) = 4.0 kN-m
Therefore, the lagging is acceptable.
As indicated in Section 13.3.6.2, the thickness of timber lagging for temporary support is generally
selected based on empirical relationships with excavation depth (Goldberg, et al., 1975). For an
excavation depth of 8 m, a lagging thickness of 76 mm to 102 mm is prescribed for soldier piles
spaced at 2.5 m in medium dense sand. It is apparent from the above calculation that thinner lagging
would be acceptable.
Summary
y
This example problem illustrates the design of an anchored soldier pile and
lagging retaining wall by LRFD
The wall consists of driven steel H-piles supported by two rows of anchors
and spanned using timber lagging
B-63
REFERENCES
Agarwal, S. L., 1964, "Bearing Capacity of Piles in Cohesionless Soils: Discussion," Journal of
the Soil Mechanics and Foundation Division, ASCE, Vol. 90, No. SM1, pp. 130-131.
AASHTO, 1969, Standard Specifications for Highway Bridges, American Association of State
Highway Officials, Washington, D.C., 10th Edition.
AASHTO 1991, Standard Specifications for Highway Bridges, Division I - Design, American
Association of State Highway and Transportation Officials, Washington, D.C., 470p.
AASHTO, 1993, Guide to Metric Conversion, American Association of State Highway and
Transportation Officials, American Association of State Highway and Transportation Officials,
Washington, D.C., 106 p.
AASHTO, 1993, Standard Specifications for Transportation Materials and Methods of Sampling
and Testing. Part II Tests, American Association of State Highway and Transportation Officials
American Association for Transportation and Highway Officials, Washington, D.C.
AASHTO, 1995, Standard Specifications for Highway Bridges, Division I, American
Association for Transportation and Highway and Transportation Officials, Washington, DC, 15th
ed. (with Interim Specifications)
AASHTO, 1996, LRFD Highway Bridge Design Specifications, SI Units, American Association
of State Highway and Transportation Officials, Washington, D.C., First Edition (1994 with 1996
Interims)
AASHTO, 1997a, LRFD Highway Bridge Design Specifications, SI Units, American
Association of State Highway and Transportation Officials, Washington, D.C., First Edition
(1994 with 1996 and 1997 Interims)
AASHTO, 1997b, Standard Specifications for Highway Bridges, American Association of State
Highway and Transportation Officials, Washington, D.C., 16th Edition (1996 with 1997
Interims).
ACI, 1995, Building Code Requirements for Reinforced Concrete, American Concrete Institute,
Detroit.
AISC, 1994, Load and Resistance Factor Design Specification for Structural Steel Buildings,
American Institute of Steel Construction, Chicago, IL, 2nd Edition.
AISI, 1967, Handbook of Steel Drainage & Highway Construction Products, American Iron and
Steel Institute, New York, 1st Edition.
R-1
AISI, 1993, Handbook of Steel Drainage & Highway Construction Products, 4th Edition,
American Iron and Steel Institute, Washington, D.C.
ASCE, 1969, Design and Construction of Sanitary and Storm Sewers, American Society of Civil
Engineers, Manuals and Reports on Engineering Practice No. 37, New York.
ASCE, 1993, Standard Practice for Direct Design of Buried Precast Concrete Pipe Using
Standard Installations (SIDD), American Society of Civil Engineers New York, NY.
ASTM, 1993, Standard Practice for the Use of the International System of Units (SI), (The
Modernized Metric System), ASTM-E-380-93, American Society of Testing and Materials,
Philadelphia, PA, 35p.
ASTM, 1997, Annual Book of ASTM Standards, Volume 04.08, Soil and Rock (I): D 420 - D
4914, Philadelphia, PA, 1,000 p.
Azzouz, A.S., Baligh, M.M., Ladd, C.C. (1983), Corrected Field Vane Strength for
Embankment Design, Journal of Geotechnical Engineering, ASCE, Volume 109, No. 5, pp.
730-734.
Barker, R.M., J.M. Duncan, K.B. Rojiani, P.S.K. Ooi, C.K. Tan, and S.G. Kim 1991a, "Manuals
for the Design of Bridge Foundations," NCHRP Report 343, Transportation Research Board,
National Research Council, Washington, DC.
Barker, R.M., J.M. Duncan, K.B. Rojiani, P.S.K. Ooi, C.K. Tan, and S.G. Kim, 1991b, "Load
Factor Design Criteria for Highway Structure Foundations," Final Report, NCHRP Project 24-4,
Virginia Polytechnic Institute and State University, Blacksburg, VA.
Benjamin, J.R. and C.A. Cornell, 1970, Probability, Statistics, and Decision for Civil Engineers,
McGraw-Hill, New York, NY.
Blanchet, R., F. Tavenas, and R. Garneau, 1980, "Behavior of Friction Piles in Soft Sensitive
Clays", Canadian Journal of Geotechnical Engineering, Vol. 17, No. 2, pp. 203-224.
Bolton, M.D., 1986, "The Strength and Dilatancy of Sands", Geotechnique, Vol. 36, No. 1, pp.
65-78.
Borden, R.H. and M.A. Gabr, 1987, Analysis of Compact Pole-Type Footings; LTBASE:
Computer Program For Laterally Loaded Pier Analysis Including Base and Slope Effects,
Department of Civil Engineering, North Carolina State University at Raleigh, report prepared for
the North Carolina Department of Transportation and Federal Highway Administration, 215 p.
Bowles, J. E. (1968). Foundation Analysis and Design, McGraw-Hill, New York.
R-2
Briaud, J.-L., R. Viswanathan, R.K. Bush, A. Yeung, and J. Sangseom, (1994b), "Downdrag on
Uncoated and Bitumen-Coated Piles", Draft Submittal, prepared for NCHRP.
Briaud, J.L. and L.M. Tucker, 1994, Design and Construction Manual for Downdrag on
Uncoated and Bitumen-Coated Piles, Draft Report prepared for National Cooperative Highway
Research Program, Texas Transportation Institute, Texas A&M University, College Station, TX,
222p.
Briaud, J.L. and L. Tucker, 1984, "Coefficient of Variation of In Situ Tests in Sand,"
Probabilistic Characterization of Soil Properties - Bridge Between Theory and Practice, ASCE,
pp. 119-139.
Broms, B. B., 1963, "Bearing Capacity of Piles in Cohesionless Soils: Discussion", Journal of
the Soil Mechanics and Foundation Division, ASCE, Vol. 89, No. SM6, pp. 125-126.
Buckle, I., 1995, Chapter 7, Seismic Design and Retrofit of Bridges, Bridge Engineering, Volume
1, Participant Notebook, NHI Course No. 13064, Federal Highway Administration, Washington,
DC, 132p.
Bustamante, M. and L. Gianeselli, 1982, "Pile Bearing Capacity Prediction by Means of Static
Penetrometer CPT," Proceedings of the 2nd European Symposium on Penetration Testing,
Amsterdam, The Netherlands, pp. 493-500.
Canadian Geotechnical Society, 1992, Canadian Foundation Engineering Manual, 3rd Edition,
Bi-Tech Publishers, Ltd., Richmond, British Columbia, Canada, 512 p.
Canadian Geotechnical Society, 1985, Canadian Foundation Engineering Manual, Bitech
Publishers, Ltd., Second Edition, .
Carter, M. and S.P. Bentley, 1991, Correlations of Soil Properties, Pentech Press, London, U.K.,
130p.
Carter, J.P. and F.H. Kulhawy, 1987, Analysis and Design of Foundations Socketed into Rock,
Research Report 1493-4, Geotechnical Engineering Group, Cornell University, Ithaca, New
York.
Castilla, F., P. Martin, and J. Link, 1984, Fixity of Members Embedded in Concrete, Army
Construction Engineering Research Laboratory, CERLTRM-339.
Chen, Y.J. and F.H. Kulhawy, 1993, "Undrained Strength Interrelationships Among CIUC, UU,
and UC Tests," Journal of Geotechnical Engineering, Vol. 119, No. 11, pp. 1732-1750.
Cheney, R.S., 1988, Permanent Ground Anchors, Report No. FHWA-DP-68-1R Demonstration
Project, Federal Highway Administration, Washington, DC, 132 p.
R-3
Cheney, R.S. and R.D. Chassie, 1993, Soils and Foundations Workshop, Publication No.
FHWA-HI-88-009 (Revised); Federal Highway Administration, U.S. Department of
Transportation, Washington, D.C., 382p.
Christopher, B.R., S.A. Gill, J-P Giroud, I. Juran, J.K. Mitchell, F. Schlosser and J. Dunnicliff,
1990, Reinforced Structures, Volume I, Design and Construction Guidelines, FHWA-RD-89-043,
Office of Engineering and Highway Operations, Federal Highway Administration, McLean, VA,
301p.
Churner, R.M.,1970, "Effect of Particle Characteristics on Soil Strength", Journal of the Soil
Mechanics and Foundation Division, ASCE, Vol. 96, No. SM4, pp. 1221-1234.
Clough, G.W. and T.D. O'Rourke, 1990, "Construction Induced Movements of Insitu Walls,"
Design and Performance of Earth Retaining Structures, Geotechnical Special Publication No.
25, P.C. Lambe and L.A. Hansen, eds., American Society of Civil Engineers, New York, NY,
pp. 439-470.
Commonwealth of Pennsylvania, Pennsylvania Department of Transportation (PADOT), Bureau
of Design, 1995, Metric Training for Highway Publications, Second Edition, Harrisburg, PA,
212 p.
Danish Geotechnical Institute, 1985, Code of Practice for Foundation Engineering, DS 415,
Bulletin No. 36, Third Edition.
D'Appolonia, D.J., E. D'Appolonia and R.F. Brissette, 1970, Discussion of "Settlement of Spread
Footings in Sands", Journal of the Soil Mechanics and Foundations Division, ASCE, Vol. 96,
No. SM2, pp.754-761.
Davisson, M.T., F.S. Manuel, and R.M. Armstrong, 1983, Allowable Stresses In Piles, Report
No. FHWA/RD-83/059, Federal Highway Administration , 191 p.
Demsky, E., 1996, Lecture presented at Virginia Polytechnic Institute and State University,
Blacksburg, Va. on June 14, 1996.
Elias, V. And B.R. Christopher, 1996, Mechanically Stabilized Earth Walls and Reinforced Soil
Slopes, Design & Construction Guidelines, Office of Technology Applications, Federal Highway
Administration, Washington, D.C., 367 p.
Fang, H-Y, 1991, Foundation Engineering Handbook, Second Edition, H-Y Fang, ed., Van
Nostrand Reinhold, New York, NY, 923 p.
Federal Highway Administration (FHWA), 1988, Checklist Guidelines for Review of
Geotechnical Reports and Preliminary Plans and Specifications, Report No. FHWA-ED-88-053,
U.S. Department of Transportation, 33 p.
R-4
Federal Highway Administration (FHWA), 1991, Evaluating Scour at Bridges, Report No.
FHWA-1P-90-017, Hydraulic Engineering Circular 18, U.S. Department of Transportation,
McLean, VA.
Federal Highway Administration (FHWA), 1995, FHWA Geotechnical Metrication Guidelines,
FHWA, Office of Technology Applications, Washington, DC, 57p.
Federal Register, 1992, Metric Transition Policy, Vol. 57, No. 113, June 11, 1992, Washington,
DC, pp. 24843-24846.
Gifford, D.G., J.R. Kraemer, J.R. Wheeler, and A.F. McKown, 1987, Spread Footings for
Highway Bridges, Report No. FHWA/RD-86/185, Federal Highway Administration, McLean,
VA, 229p.
Glascock, B.C., 1980, Recommended Design Requirements for Elastic Buckling of Buried
Flexible Pipe, AWWA C950 Task Group Report on Buckling, American Waterworks
Association.
Goble. G.G. and F. Rausche, 1986, Wave Equation Analysis of Pile Driving - WEAP86 Program,
Volumes I-IV, U.S. Department of Transportation, Federal Highway Administration,
Implementation Division, McLean, VA.
Goble, Rausche, Likens and Associates, Inc. (GRL), 1996, Design and Construction of Driven
Pile Foundations, Volume I, NHI Course Nos. 13221 and 13222, US Department of
Transportation, Federal Highway Administration, Washington, DC.
Goldberg, D.T., W.E. Jaworski and M.D. Gordon, 1976, Lateral Support Systems and
Underpinning, 4 Volumes, Report Nos. FHWA-RD-75-128, FHWA-RD-75-129, FHWA-RD-75130, FHWA-RD-75-131, Federal Highway Administration, Washington, DC.
Grasso, P., Xu, S., and Mahtab, A., 1992, "Problems and Promises of Index Testing of Rocks",
Proceedings of the 33rd U.S. Symposium on Rock Mechanics, J.R. Tillerson and W.R.
Wawersik, eds., Balkema, Amsterdam, the Netherlands, pp. 879-888.
Hansell, W. C. and I. M. Viest, 1971, "Load Factor Design for Steel Highway Bridges," AISC
Engineering Journal, Vol. 8, No. 4, pp. 113-123.
Heger, F.J., 1963, Structural Behavior of Circular Reinforced Concrete Pipe - Development of
Theory, ACI Journal, November, American Concrete Institute, Detroit.
Heger, F.J., 1982, Structural Design Method for Precast Reinforced Concrete Pipe, TRB Record
878, Transportation Research Board, Washington, D.C.
R-5
Heger, F.J. and T.J. McGrath, 1982, Shear Strength of Pipe, Box-Sections and Other One-Way
Flexural Members, ACI Journal, November-December, American Concrete Institute, Detroit.
Holtz, R.D. and W.D. Kovacs, 1981, An Introduction to Geotechnical Engineering,
Prentice-Hall, Inc., Englewood Cliffs, NJ, 733 p.
Horvath, R.G., and T.C. Kenney, 1979, "Shaft Resistance of Rock-Socketed Drilled Piers,"
Proceedings, Symposium on Deep Foundations, ASCE, Atlanta, Georgia, pp. 182-214.
Hunt, R.E., 1986, Geotechnical Engineering Analysis and Practices, McGraw-Hill Book
Company, New York, NY, 729 p.
Koerner, R.M.,1970, "Effect of Particle Characteristics on Soil Strength", Journal of the Soil
Mechanics and Foundation Division, ASCE, Vol. 96, No. SM4, pp. 1221-1234.
Kulhawy, F.H., 1978, "Geomechanical Model for Rock Foundation Settlement", Journal of the
Geotechnical Engineering Division, ASCE, Vol. 104, No. GT2, pp. 211-227.
Kulhawy, F.H., Trautmann, C.H., Beech, J.F., O'Rourke, T.D., and McGuire, W., 1983,
Transmission Line Structure Foundations for Uplift-Compression Loading, EL-2870 Final
Report, Report prepared by Cornell University for the Electric Power Research Institute, Palo
Alto, CA, 370 p.
Kulhawy, F.H. and P.W. Mayne, 1990, Manual on Estimating Soil Properties for Foundation
Design, EL-6800 Final Report, Report prepared by Cornell University for the Electric Power
Research Institute, Palo Alto, CA, 296 p.
Kulhawy, F.H., Roth, M.J.S. and Grigoriu, M.D., 1991, "Some Statistical Evaluations of
Geotechnical Properties", Proceedings of ICASP6, Sixth International Conference on
Applications of Statistics and Probability in Civil Engineering, Volume 2, Mexico City, pp. 705712.
Leonards, G.A., and Stetkar, R.E., 1978, Performance of Buried Flexible Conduits, Purdue
University Joint Highway Research Project No. C-36-62F, Federal Highway Administration No.
JHRP-78-24.
Lind, N. C., 1971, "Consistent Partial Safety Factors," Journal of the Structural Engineering
Division, ASCE, Vol. 97, No. ST6, pp. 1651-1669.
Low, N. I., 1964, "Bearing Capacity of Piles in Cohesionless Soils: Discussion," Journal of the
Soil Mechanics and Foundation Division, ASCE, Vol. 90, No. SM1, pp. 129-130.
Marston, A., 1930, The Theory of External Loads on Closed Conduits in the Light of the Latest
Experiments, Bulletin 96., Iowa Engineering Experiment Station, Ames, Iowa.
R-6
Marston, A., and A. O. Anderson, 1913, The Theory of Loads of Pipes in Ditches and Tests of
Cement and Clay Drain Tile and Sewer Pipe, Bul. 31, Iowa Engineering Experiment Station,
Ames, Iowa.
McGrath, T.J., Tigue, D.B., and Heger, F.J., 1988, PIPECAR and BOXCAR Micro Computer
Programs for the Design of Reinforced Concrete Pipe and Box Sections, TRB Record 1191,
Transportation Research Board, Washington, D.C.
Means, 1996, Heavy Construction Cost Data, 10th Annual Edition, R.S. Means Company, Inc.
Meyerhof, G.G., 1956, "Penetration Tests and Bearing Capacity of Cohesionless Soils", Journal
of the Soil Mechanics and Foundation Division, ASCE, Vol. 82, No. SM1, pp. 1-19.
Meyerhof, G.G., and Baikie, L.D., 1963, Strength of Steel Culvert Sheets Bearing Against
Compacted Sand Backfills, with Discussion by R.K. Watkins, Highway Research Record No. 30,
pp. 1-19.
Meyerhof, G.G., 1970, "Safety Factors in Soil Mechanics," Canadian Geotechnical Journal,
Vol. 7, No. 4, pp. 349-355.
Meyerhof, G.G., 1976, "Bearing Capacity and Settlement of Pile Foundations", Journal of
Geotechnical Engineering, ASCE, Vol. 102, No. GT3, pp. 197-227.
Mitchell, J.K., and Gardner, W.S., 1975, "In-Situ Measurement of Volume Change
Characteristics", Proceedings of the ASCE Specialty Conference on In-Situ Measurement of Soil
Properties, Vol. 2, Raleigh, NC, pp. 279-345,
Mitchell, J.K. and W.C.B. Villet, 1987, Reinforcement of Earth Slopes and Embankments,
NCHRP Report 290, Transportation Research Board, National Academy of Sciences,
Washington, D.C., 323 p.
Moulton, L.K., 1986, Tolerable Movement Criteria for Highway Bridges, Report No. FHWATS-85-288, Federal Highway Administration, U.S. Department of Transportation, Washington,
DC, 93 p.
National Cooperative Highway Research Program (NCHRP), 1990, Truck Weight Limits: Issues
and Options, NCHRP Special Report 225, Transportation Research Board, Washington, D.C.,
307 p.
NAVFAC, 1986a, Soil Mechanics, Design Manual 7.01, Naval Facilities Engineering Command,
U.S. Government Printing Office, Washington, D.C., 348 p.
NAVFAC, 1986b, Foundations & Earth Structures, Design Manual 7.02, Naval Facilities
Engineering Command, U.S. Government Printing Office, Washington, D.C., 244 p.
R-7
R-8
Reese, L.C., S.T. Wang, K. Awoshika, and P.H.F. Lam, 1994, Documentation of Computer
Program Group, Version 3.0, Analysis of a Group of Piles Subjected to Axial and Lateral
Loading, Ensoft, Inc., 370 p.
Reese, L.C. and J.D. Allen, 1977, Drilled Shaft Design and Construction Guidelines Manual,
Vol. II - Structural Analysis and Design for Lateral Loading, Implementation Package 77-21,
U.S. Department of Transportation, Offices of Research and Development, Washington D.C.,
228 p.
Reese, L.C. and S.J. Wright, 1977, Drilled Shaft Design and Construction Guidelines Manual,
Vol. I - Construction of Drilled Shafts and Design for Axial Loading, Implementation Package
77-21; U.S. DOT, Offices of Research and Development, Washington, D.C., 140 p.
Reese, L.C., 1984, Handbook on Design of Piles and Drilled Shafts Under Lateral Load,
Publication No. FHWA-IP-84-11, Federal Highway Administration, U.S. Department of
Transportation, Office of Implementation, McLean, Virginia, 386p.
Riaund, J.L and J. Miran, 1992, The Cone Penetrometer Test, Report No. FHWA-SA-91-043,
Federal Highway Administration, Office of Technology Applications, Washington, D.C., 161 p.
Rosenblueth, E. and L. Esteva, 1972, "Reliability Basis for Some Mexican Codes," ACI
Publication SP-31, American Concrete Institute, Detroit, MI.
Schlick, W.J., 1932, Loads on Pipe in Wide Ditches, Bul. 108, Iowa Engineering Experiment
Station, Ames, IA.
Schlick, W.J., 1920, Supporting Strength of Drain Tile and Sewer Pipe Under Different PipeLaying Conditions, Bul. 57, Iowa Engineering Experiment Station, Ames, IA.
Seed, H. B. and DeAlba, P., 1986, Use of SPT and CPT Tests for Evaluating the Liquefaction
Resistance in Sands, ASCE, Proc. In Situ Testing '86, Blacksburg, VA, pp. 281-302.
Sehn, A.L., 1990, Experimental Study of Earth Pressures on Retaining Structures, Dissertation
submitted to the faculty of the Virginia Polytechnic Institute and State University in partial
fulfillment of the requirements for the degree of Doctor of Philosophy in Civil Engineering, 347
p.
Shahawy, M.A., and Issa, M., 1992, "Effect of Pile Embedment on the Development Length of
Prestressing Strands," PCI Journal, Vol. 37, No. 6, pp. 44-59.
Sherif, M.A., I. Ishibashi and C.D. Lee, 1982, 'Earth Pressures Against Rigid Retaining Walls, "
Journal of the Geotechnical Engineering Division, ASCE, Vol. 108, No. GT5, pp. 697-695.
R-9
Spangler, M.G., 1960, Soil Engineering, 2nd ed., International Textbook Company, Scranton,
PA, 483 p.
Spangler, M.G., 1950, Field Measurements of the Settlement Ratios of Various Highway
Culverts, Bul. 170, Iowa Engineering Experiment Station, Ames, Iowa.
Spangler, M.G., 1933, The Supporting Strength of Rigid Pipe Culverts, Bul. 112, Iowa
Engineering Experiment Station, Ames, IA.
Tang, W. H., D. L. Woodford and J. H. Pelletier, 1990, Performance Reliability of Offshore Pile,
22nd Annual Offshore Technology Conference, Houston, Texas, Paper No. OTC 6379.
Teng, W., Mesri, G., and Halim, I., 1992, "Uncertainty of Mobilized Undrained Shear Strength",
Soils and Foundations, Vol. 32, No. 4, pp. 107-116.
Terzaghi, K., 1934, "Retaining Wall Design for Fifteen-Mile Falls Dam," Engineering News
Record, May, pp. 632-636.
Terzaghi K. and R.B. Peck, 1967, Soil Mechanics in Engineering Practice, John Wiley & Sons,
Inc, New York, NY, 729 p.
Thurman, A.G., 1964 "Bearing Capacity of Piles in Cohesionless Soils: Discussion," Journal of
the Soil Mechanics and Foundation Division, ASCE, Vol. 90, No. SM1, pp. 127-129.
Timoshenko, S.P., and Gere, J.M., 1961, Theory of Elastic Stability, McGraw-Hill, New York,
2nd Edition.
Tomlinson, M.J., 1986, Foundation Design and Construction, 5th Ed., Longman Scientific and
Technical, London, England, 842 p.
Tomlinson, M.J., 1987, Pile Design and Construction Practice, Viewpoint Publication, 415 p.
Touma, F.T. and L.C. Reese, 1974, "Behavior of Bored Piles in Sand," Journal of the
Geotechnical Engineering Division, ASCE, Vol. 100, No. GT 7, pp. 749-761.
U.S. Army Corps of Engineers, 1993, Bearing Capacity of Soils, Engineer Manual EM 1110-11905, ASCE Press, New York.
Vanikar, S. N.,1986, "Manual on Design and Construction of Driven Pile Foundations," Report
No. FHWA-DP-66-1, U.S. Department of Transportation, Federal Highway Administration,
Washington, DC.
Vanmarcke, E. H., 1977, "Probabilistic Modeling of Soil Profiles," Journal of Geotechnical
Engineering, ASCE, Vol. 103, No. GT11, pp. 1227-1246.
R-10
Vesic, A.S., 1970, "Tests on Instrumented Piles, Ogeechee River Site", Journal of the Soil
Mechanics and Foundation Division, ASCE, Vol. 96, No. SM2, pp. 561-584
Vesic, A.S., 1975, Chapter 3, Bearing Capacity of Shallow Foundations, Foundation
Engineering Handbook, First Ed., H.F. Winterkorn and H.T. Fang, eds, Van Nostrand Reinhold
Company, New York, NY, pp. 121-147.
Vesic, A.S., T.L. Copas and H.A. Pennock, 1977, Design of Pile Foundations, National
Cooperative Highway Research Program, Synthesis of Practice 42, Transportation Research
Board, National Research Council, Washington, DC, 68p.
Wang, S-H and L.C. Reese, 1993, COM624P - Laterally Loaded Pile Analysis Program for the
Microcomputer, Version 2.0, Report No. FHWA-SA-91-048, Federal Highway Administration,
McLean, Virginia, 504 p.
Watkins, R.K., and Moser, A.P., 1969, The Structural Performance of Buried Corrugated Steel
Pipes, Research Report, Engineering Experimentation Station, Utah State University, American
Iron and Steel Institute.
Watkins, R.K., and Spangler, M.G., 1958, Some Characteristics of the Modulus of Passive
Resistance of Soil: A Study in Similitude, Proc. Highway Research Board, Vol. 37, p. 576.
Weatherby, D.E., 1982, Tiebacks, FHWA RD-82-047, Federal Highway Administration,
Washington, DC, 249p.
Westergaard, H.M., 1938, A Problem of Elasticity Suggested by a Problem in Soil Mechanics:
A Soft Material Reinforced by Numerous Strong Horizontal Sheets, Mechanics of Solids, by
Timoshenko, S., Macmillan, New York, NY.
Whyte, I.L., 1995, "The Financial Benefit From a Site Investigation Strategy", Ground
Engineering, Vol. 28, No. 8, pp. 33-36.
Wu, T.H., W.H. Tang, D.A. Sangrey and G.B. Baecher, 1989, "Reliability of Offshore
Foundations - State of the Art," Journal of Geotechnical Engineering, ASCE, Vol. 115, No. 2,
pp. 157-178.
R-11
R-12