I EPRI Topics:

Soils
Testing
EPRI EL-6800
Project 1493-6
Final Report

I Electric Power
Research Institute
Foundations
Transmission towers
Transmission lines
August 1990

Design
I
I
Manual on Estimating Soil
Properties for Foundation Design

I
I
I
I
I
I
I
I
I Prepared by
Cornell University
Ithaca, New York

I "·

I
f
 REPORT S U M M A R yl
SUBJECTS Overhead structures and foundations I Overhead transmission
- - - - - - -
J
TOPICS Soils Transmission towers
Testing Transmission lines

AUDIENCE
Foundations Design
--------------------------------------------------------
Transmission managers and engineers
1
I
I
Manual on Estimating Soil Properties for •
Foundation Design 1
This manual provides foundation engineers with a comprehensive
reference on estimating engineering soil parameters from field or 'J
laboratory test data. Empirical correlations are used extensively to
evaluate soil parameters. The manual describes the most impor-
tant of these correlations completely and systematically with an
I
emphasis on the correlations of relatively common tests, including
those that are seeing increased usage in practice. I
BACKGROUND The analysis of all geotechnical problems, such as transmission structure
foundation design, requires the adoption of a soil behavioral model that
I
must include all relevant soil properties. These soil properties are not
known in advance and require the design engineer to either measure or
estimate properties using correlations. However, the source, extent, and
I
limitations of correlations are most often obscured in the presentation of
the relationships. When plotted, most correlations are presented as a sim-
ple line, but in reality they may be based on a veritable shotgun blast of 1
data points.

OBJECTIVE 1
To present a readily usable, comprehensive set of correlations for estimat-
ing soil properties with each correlation presented in the context of its

with new data when possible.

--------------------------------------------------------
I
historical evolution and statistical variability; to update existing correlations

APPROACH I
The researchers established a context for basic soil characterization, in-
eluding simple soil descriptions, classification, unit weight, relative density,
and consistency. Next, they developed correlations for in situ state of stress,
strength, elastic behavior, time-dependent deformability, and permeability-~
both for common tests and for newer tests coming into increasing use.

RESULTS This work is a collection of correlations that organize a huge body of dis-
persed knowledge into a coherent framework. Comprehensive correlations
are given for basic soil characterizations, in situ stress state tests, strength ·
I
tests, tests of elastic and time-dependent deformability, permeability tests,
and liquefaction resistance tests. Each correlation is constructed from its
beginnings in the literature. Some correlations are original amalgams of ,
I
I
EPAI EL-6800s Electric Power Research Institute
I several different presentations, and several correlations are consider-
ably enhanced by the addition of new data. r;:urther, many new correla-
tions were developed when sufficient data were available. All of the
I presentations give the foundation designer an immediate feel for the
variability of each relationship.

I EPRI PERSPECTIVE This manual is intended to make the job of the transmission structure
foundation designer easier. A second application is to aid in the devel-
opment of local soil property correlations specific to particular utility

I service areas. This use of the soil properties manual will tie in directly
with the use of the TLWorkstation™ foundation task modules, CUFAD
and MFAD (EPRI report EL-6420, volumes 16 and 17), and the recently

I released CUFAD+ EPRIGEMS module (report EL-6583-CCML). Finally,

the manual can serve to alert the design engineer, who previously had
only standard penetration test data on which to base soil characteriza-

t tions, that several other in situ tests are vastly superior predictors of
soil properties. The engineer is thus presented with the data to make a
cost-benefits analysis of the worth of better data on which to base de-

I sign. For other EPRI work on soil properties and foundation design see
EPRI reports EL-2870 and EL-6420, volume 2.

PROJECT RP1493-6

I EPRI Project Manager: Vito Longo

Electrical Systems Division
Contractor: Cornell University

I
f
I For further information on EPRI research programs, call
EPRI Technical Information Specialists (415) 855-2411.
I
I
I'
I
I
I
I
I
 I
I
Manual on Estimating Soil Properties for Foundation
Design I
i
EL-6800
Research Project 1493-6

Final Report, August 1990

I
·~

Prepared by
I
CORNELL UNIVERSITY
Geotechnical Engineering Group
Hollister Hall
I
Ithaca, New York 14853-3501

Authors
I
F H. Kulhawy
P W. Mayne

Principal Investigator
1
F. H. Kulhawy
I
I
I
I
Prepared for
I
Electric Power Research Institute
3412 Hillview Avenue
Palo Alto, California 94304
I
EPRI Project Manager
V. J. Longo
I
Overhead Transmission Lines Program
Electrical Systems Division I
I
I
I
I
I ABSTRACT

I This manual focuses on the needs of engineers involved in the geotechnical design
of foundations for transmission line structures. It also will serve as a useful

I reference for other geotechnical problems. In all foundation design, it is neces-

sary to know the pertinent parameters controlling the soil behavior. When it is

I
not feasible to measure the necessary soil parameters directly, estimates will have
to be made from other available data, such as the results of laboratory index tests
and in-situ tests. Numerous correlations between these types of tests and the

I necessary soil parameters exist in the literature, but they have not been synthe-
sized previously into readily usable form in a collective work. This manual summa-

I rizes the most pertinent of these available correlations for estimating soil param-
eters. In many cases, the existing correlations have been updated with new data,
and new correlations have been developed where sufficient data have been avail-

I able. For each soil parameter, representative correlations commonly are presented
in chronological order to illustrate the evolutionary development of the particular

I correlation. The emphasis is on relatively common laboratory and in-situ tests and
correlations, including those tests that are seeing increased use in practice.

I
I
t
I
I
I
I iii

I
 I
I
I
ACKNOWLEDGMENTS I
The authors appreciate the assistance of numerous people during the course of this
study and want to acknowledge their contributions. C. H. Trautmann and H. E.
I
Stewart of Cornell University provided many useful comments and suggestions as this
report was compiled, and they provided detailed review comments on both the first I
and final drafts of this report. K. J. Stewart prepared the text, and A. Avcisoy
drafted the figures.
I
Several colleagues graciously responded to a request for review and evaluation of
the first draft of this report. These included: J. I. Adams, Consultant; R. G.
I
Campanella, University of British Columbia; A. M. DiGioia, Jr. and L. F. Rajas-
Gonzalez, GAl Consultants; M. D. Grigoriu and S. Vidic, Cornell University; E. B.
Lawless, III, Potomac Electric Power Company; V. J. Longo, EPRI; J. K. Mitchell,
I
University of California at Berkeley; H. S. Radhakrishna, Ontario Hydro; T. E.
Rodgers, Jr., Virginia Power; J. W. Rustvold, Bonneville Power Administration; J.
H. Schmertmann, Schmertmann and Crapps; J. J. Wolf, Western Area Power Administra-
1
tion; and C. P. Wroth, University of Oxford. The detailed reviewers' comments were
very helpful in preparing the final text of this report.
I
I
I
t
I
I
I
v
I
I
I
I

CONTENTS

Section

1 INTRODUCTION AND BACKGROUND 1-1

Soil Correlations l-1
Soil and Test Variability 1-2
Soil Modeling 1-3
Scope of Manual 1-5
References l-7

2 BASIC SOIL CHARACTERIZATION 2-1

Simple Descriptions 2-l
Particle Size and Distribution 2-1
Index Parameters for Cohesive Soils 2-2
Index Parameters for Cohesionless Soils 2-3
Characterization by Simple Field Tests 2-6
Color and Odor 2-6
Classification 2-8
General Classification and Identification Systems 2-8
Cone Penetration Test (CPT) Classifications 2-8
Piezocone Penetration Test (CPTU) Classifications 2-14
Dilatometer Test (DMT) Classifications 2-16
Unit Weight 2-17
Relative Density of Cohesionless Soils from In-Situ Test
Correlations 2-17
Standard Penetration Test (SPT) Correlations 2-17
Cone Penetration Test (CPT) Correlations 2-26
Dilatometer Test (DMT) Correlations 2-34
Consistency of Cohesive Soils from In-Situ Test Correlations 2-35
Standard Penetration Test (SPT) Correlations 2-35
Cone Penetration Test (CPT) Correlations 2-37
Relationship Between SPT N and CPT qc Values 2-36
References 2-39

vii
 I
Section Page I
3 IN-SITU STRESS STATE 3-1
Basic Definitions
Reconstruction of Stress History
3-1
3-2
I
Effective Preconsolidation Stress in Cohesive Soils
Correlations with Index Parameters
3-6
3-7 I
Comments on Field Test Correlations 3-8
Correlations with VST Strength
Correlations with SPT N Value
3-8
3-8
1
Correlations with CPT qc Value
Correlations with CPTU Results
Correlations with PMT Results
3-10
3-10
3-13
I
Correlations with DMT Results
Effective Preconsolidation Stress in Cohesionless Soils
3-13
3-14
I
Overconsolidation Ratio for Cohesive Soils
Correlations with Index Parameters
3-14
3-15 I
Correlations with Laboratory Strength 3-16
Correlations with VST Strength
Correlations with SPT N Value
3-17
3-17
I
Correlations with CPT and CPTU Results
Correlations with DMT Results
Overconsolidation Ratio in Cohesionless Soils
3-18
3-18
3-19
I
Effective Horizontal Stress in Cohesive Soils
Correlations with Index Parameters
3-20
3-21
~
Direct Correlations with SBPMT and DMT Results
Indirect Correlations with SPT, CPT, CPTU, and DMT Results
3-24
3-26 't
Effective Horizontal Stress in Cohesionless Soils 3-29
Direct Correlations with SBPMT and DMT Results
Indirect Correlations with SPT and CPT Results
3-29
3-30
I
Combined DMT/CPT Approach for K0 of Sands
Empirical Approach
References
3-32
3-33
3-35
I
4 STRENGTH
Basic Definitions
4-1
4-1
i
Effective Stress Analysis
Total Stress Analysis
4-1
4-4
I
Relevance of Laboratory Strength Tests to Field Conditions 4-5
Effective Stress Friction Angle of Cohesionless Soils - General
Evaluation Basis 4-7
I
viii i
I
 Section

Typical Values 4-7

Correlations with Index Parameters 4-7
Influence of Strength Envelope Curvature 4-8
Influence of Test Boundary Conditions 4-12
Effective Stress Friction Angle of Cohesionless Soils
Correlated with In-Situ Tests 4-13
Correlations with SPT N Value 4-13
Correlations with CPT qc Value 4-14
Correlations with PMT Results 4-18
Correlations with DMT Results 4-20
Effective Stress Friction Angle of Cohesive Soils 4-21
Correlations with Critical Void Ratio Friction Angle 4-21
Influence of Test Boundary Conditions 4-22
Correlations with Residual Friction Angle 4-24
Undrained Shear Strength of Cohesive Soils - General Evaluation
Basis 4-27
Correlations with Index Parameters for Undisturbed Clays 4-28
Correlations with Index Parameters for Remolded Clays 4-30
General Behavior Under Triaxial Compression Loading 4-32
Influence of Overconsolidation 4-37
Influence of Test Boundary Conditions 4-39
Influence of Strain Rate During Testing 4-48
Summary of Factors Influencing the Undrained Strength
Ratio 4-48
Undrained Shear Strength of Cohesive Soils Correlated with
In-Situ Tests 4-49
Correlations with VST Results 4-50
Correlations with SPT N Value 4-53
Correlations with CPT qc Value 4-55
Correlations with CPTU Results 4-58
Correlations with PMT Results 4-59
Correlations with DMT Results 4-59
References 4-61

5 ELASTIC DEFORMABILITY 5-1

Basic Definitions 5-1
Poisson's Ratio 5-3
Undrained Modulus of Cohesive Soils 5-4
Typical Valt<es 5-6
Correlations with su 5-6

ix
 I
Section Page I
Correlations with SPT, CPT, and PMT Results 5-10
Back-Figured from Full-Scale Load Tests
Estimation from Dynamic Measurements
5-11
5-12
I !
Modulus for Cohesionless Soils
Typical Values
5-13
5-14 I
Correlations with Strength 5-15
Correlations with SPT N Value
Young's Modulus
5-17
5-17
I
Pressuremeter Modulus
Dilatometer Modulus
Correlations with CPT qc Value
5-18
5-18
5-19
I
Back-Figured from Full-Scale Load Tests
Estimation from Dynamic Measurements
5-20
5-20
~
Subgrade Reaction
References
5-23
5-24
·~
6 TIME-DEPENDENT DEFORMABILITY
Basic Definitions
Compression and Unload-Reload Indices for Cohesive Soils
6-1
6-1
6-3
a
Typical Values
Correlations with CPT qc Value
6-3
6-7 I
Constrained Modulus for Cohesive Soils
Typical Values
Correlations with SPT N Value
6-7
6-7
6-7
t
Correlations with CPT Results
Correlations with DMT Results
6-9
6-9
I
6-10
I
Compression Index for Cohesionless Soils
Constrained Modulus for Cohesionless Soils 6-11
Typical Values 6-11
Correlations with SPT, CPT, and DMT Results
Coefficient of Consolidation
6-13
6-13 I
I
Typical Values 6-13
Correlations with CPT and DMT Results 6-13
Coefficient of Secondary Compression 6-15

7
References

PERMEABILITY
6-17

7-1
t
Typical Values
References
7-1
7-4 I
X
I
I
I Section Page

t\ 8 LIQUEFACTION RESISTANCE
Cyclic Stress Ratio
Correlations with SPT, CPT, and DMT Results
8-1
8-1
8-1

Im
•·
Appendix
References 8-4

•t
t'
A STANDARD PENETRATION TEST
Procedure
Advantages and Disadvantages
Sources of Error, Reliability, and Cost
References
A-1
A-1
A-3
A-5
A-7

~
B CONE PENETRATION TEST B-1
Procedure B-2

~
Advantages and Disadvantages B-7
Sources of Error, Reliability, and Cost B-8
References B-10

c PRESSUREMETER TEST C-1

Procedure C-1
Advantages and Disadvantages C-5
Sources of Error, Reliability, and Cost C-5
References C-7

D DILATOMETER TEST D-1

Procedure D-1
Advantages and Disadvantages D-3
Sources of Error, Reliability, and Cost D-4
References D-5

E VANE SHEAR TEST E-1

Procedure E-1
Advantages and Disadvantages E-3
Sources of Error, Reliability, and Cost E-4
References E-5

F COMPARISON OF IN-SITU TEST METHODS F-1

References F-6

G CRITICAL STATE SOIL MECHANICS (CSSM) CONCEPT G-1

References G-3

xi
Appendix
Page
H CPT CALIBRATION CHAMBER DATA FOR SANDS
H-1
Data Summary
H-1
Chamber Boundary Influence
H-4
References
H-7
I UNIT CONVERSIONS
I-1
J SUMMARY CORRELATION TABLES
J-1 ~I
t

I
I
I
I
I
t
I
I
I
I
I
I
xii
I
I
I
It
i

Il
ILLUSTRATIONS

Figure

1-1 Common Laboratory Strength Tests and Field Tests 1-3

1-2 Importance of Proper Data Presentation 1-6

2-1 Liquidity Index Variations 2-3

2-2 Particle Roundness Definitions 2-5

2-3 Generalized Curves for Estimating emax and emin 2-5

2-4 Early Soil Classification by CPT 2-12

2-5 Soil Classification by Mechanical Friction CPT 2-13

2-6 Soil Classification by Fugro Electric Friction CPT 2-13

2-7 Simplified Soil Classification by Fugro Electric Friction CPT 2-13

2-8 Most Recent Soil Classification by Fugro Electric Friction CPT 2-13

2-9 Soil Classification Based on qT and Bq 2-15

2-10 Soil Classification Based on CPTU Data 2-15

2-11 Unequal End Areas of Electric Friction Cone 2-15

2-12 Determination of Soil Description and Unit Weight by DMT 2-16

2-13 Effect of Overburden Stress and Dr on SPT N Value 2-19

2-14 Relative Density-N-Stress Relationship 2-19

2-15 Relative Density-N-Stress Relationship for Several Sands 2-20

2-16 Donut and Safety Hammer Comparisons 2-21

2-17 Energy Ratio Variations 2-22

2-18 Comparison of SPT Overburden Corrections 2-23

2-19 Comparison of Recommended CN Factors and Available Data from OC

Sands 2-24

2-20 Particle Size Effect on Blow Count for Sands 2-24

xiii
Figure

2-21 Aging Effect on Blow Count for Sands 2-25

2-22 Relative Density from CPT for Uncemented and Unaged Quartz Sands 2-28

2-23 Relative Density from CPT, Including Soil Compressibility 2-28

2-24 Correlation Between Dr and Dimensionless qc (Uncorrected for

Boundary Effects) 2-29

2-25 Calibration Chamber Data on NC Sands 2-31

2-26 Calibration Chamber Data on OC Sands 2-32

1

2-27 Summary of Calibration Chamber Studies

Correlation Between DMT Horizontal Stress Index and Relative

2-33 I
2-28

2-29
Density for Normally Consolidated, Uncemented Sand

Variation of qc/N with Grain Size for Electric and Mechanical

2-34
J
I
Friction Cones 2-37

2-30 Recommended Variation of qc/N with Grain Size for Fugro Electric
Friction Cones 2-37

2-31 Variation of qc/N with Fines Content 2-38

2-32 Recommended Variation of qc/N with Fines Content 2-38

3-1 Stresses in Soil 3-l

3-2 Stress Paths for Simple Stress Histories 3-3

3-3 Horizontal Stress Coefficient for NC Soils from Laboratory Tests 3-4

3-4 Unload Coefficient for OC Soils 3-5

3-5 Reload Coefficient for OC Soils 3-5

3-6 Generalized ap - Liquidity Index - Sensitivity Relationships 3-7

3-7 ap Correlated with VST 5u 3-9

3-8 Field Vane Coefficient versus PI 3-9

3-9 ap Correlated with SPT N 3-10

3-10

3-11
ap Correlated with CPT ~

ap Correlated with CPTU <IT

3-ll

3-ll
I
3-12

3-13
ap Correlated with CPTU L'.Ut

ap Correlated with CPTU L'ltlbt

3-12

3-13
I
3-14 ap Correlated with SBPMT PL 3-14
I
xiv I
I
Figure
3-15
3-15 op Correlated with SBPMT Su and Ir
3-16
3-16 op Correlated with DMT Po

OCR Correlated with VST su 3-17

3-17

OCR Correlated with SPT N 3-18

3-18
3-19 OCR Correlated with CPTU qT 3-19

3-20 OCR Correlated with DMT Ko 3-20

3-21 OCR - Ko Relationships for Clays of Varied Geologic Origin 3-20

3-22 Konc Correlated with Atterberg Limits 3-21

3-23 K0 Correlated with PI and OCR 3-22

3-24 Apparent Lack of Trend Between Konc and PI for 135 Clay Soils 3-22

3-25 K0 Correlated with OCR 3-23

3-26 K0 Correlated with Undrained Strength Ratio 3-24

3-27 K0 from SBPMT Correlated with OCR 3-25

3-26
3-28 Ko Correlated with Ko
3-27
3-29 Ko from SBPMT Correlated with Ko
3-28
3-30 Ko Correlated with SPT N
3-28
3-31 Ko Correlated with CPTU qT
3-29
3-32 Ko Correlated with CPTU flu
3-33 Comparison of K0 Values for London Clay at Brent Cross 3-30

3-34 K0 Correlated with Ko in Sands 3-31

3-35 Cone Factor versus K0 as a Function of ~tc and D/B 3-31

3-36 Estimation of K0 in Coastal Plain Sand from CPT 3-32

3-37 Simplified qc - K0 - ~tc Relationships 3-33

3-38 K0 Correlated with qc and Ko 3-33

3-39 Tentative Correlation Between oho• ~. and Dr for NC and OC Sands

Tested in Calibration Chambers 3-34

3-40 Comparison of K0 Values at Stockholm Site 3-35

4-1 General Coulomb-Mohr Failure 4-2

4-2 Effective Stress Coulomb-Mohr Failure 4-2

XV
Figure

4-3 Strength Envelopes for a Range of Soil Types 4-3

4-4 Friction Angle Definitions 4-3

4-5 Total Stress Coulomb-Mohr Failure 4-4

4-6 Relevance of Laboratory Strength Tests to Field Conditions 4-6

4-7 ~tc versus Relative Density 4-8

4-8 ~tc versus Relative Density and Unit Weight 4-9

4-9 Nonlinear Failure Envelope Representation 4-9

4-10 Dilatancy Angle Relationships 4-11

4-11 Influence of Intermediate Principal Stress on Friction Angle 4-13

4-12 N versus ~tc 4-15

4-13 N versus ~tc and Overburden Stress 4-16

4-14 qc versus ¢tc and Vertical Stress for NC, Uncemented, Quartz Sands 4-17

4-15 ~tc from CPT Data 4-17

4-16 Simplified qc - K0 - ~tc Relationships 4-18

4-17 Trend of ¢tc with Normalized qc 4-19

4-18 PMT Data Representations 4-19

4-19 Friction Angle Evaluation from PMT Results 4-20

4-20 ~cv for NC Clays versus PI 4-22

4-21 ~tc Variation as a Function of Consolidation Stress for NC Clays 4-23

4-22 ~psc versus ~tc for NC Clays 4-23

4-23 ~te versus ~tc for NC Clays 4-24 I

4-24

4-25
~r from Ring Shear Tests and Field Studies

~r for Amuay Soils

4-26

4-26 I
4-26

4-27
su(VST)/ov0 versus PI for NC Clays

su/Ovo for NC Clay versus Liquidity Index

4-28

4-29
I
4-28 General Relationship Between Sensitivity, Liquidity Index,
and Effective Stress 4-31 I
4-29 Remolded Undrained Shear Strength versus LI 4-31

4-30 Undisturbed Undrained Shear Strength versus LI 4-32 I

xvi
I
I
I
I Figure

I 4-31 Comparison of Undrained Strength Ratio for NC Clays After

Anisotropic and Isotropic Consolidation 4-34

4-32 Undrained Strength Ratio versus ~tc for NC Clays 4-36

4-33 sufov0 versus OCR 4-38

4-34 Normalized sufovo versus OCR 4-38

4-35 Observed Trends Between Af and OCR 4-39

4-36 Relationship Between Af and PI for NC Clays 4-40

4-37 Undrained Strength Ratio as a Function of Test Type 4-41

4-38 Comparison of Undrained Strength Ratios from PSG and CK 0 UC Tests 4-42

4-39 Comparison of Undrained Strength Ratios from PSE and CK 0 UE Tests 4-42

4-40 Effect of Loading Direction on su 4-43

4-41 Undrained Strength Ratio from DSS versus ~tc 4-44

4-42 Comparison of Undrained Strength Ratios from DSS and CK 0 UC Tests 4-44

4-43 Comparison of Undrained Strength Ratios from DSS, CK 0 UC and CK 0 UE,

and PSG and PSE Tests 4-46

4-44 Comparison of Undrained Strength Ratios from CK 0 UE and CIUC Tests 4-47

4-45 Undrained Strength Ratios in Extension and Compression versus

Plasticity Index 4-47

4-46 Strain Rate Influence on Su 4-49

4-47 Normalized Undrained Strength Ratios for Major Laboratory Shear

Test Types 4-51

4-48 Field VST Correction Factor 4-52

4-49 Vane Undrained Strength Ratio versus Plasticity Index for NC, Young
and Aged Clays 4-52

4-50 Selected Relationships Between N and su 4-54

4-51 Apparent Decrease of N with Increasing Sensitivity 4-55

4-52 Relationship Between su and SPT N Value 4-56

4-53 Reported Range of Nk Factors from CPT Data 4-57

4-54 Effect of Pore Water Stress on Cone Tip Resistance 4-57

4-55 su as a Function of ~u in the CPTU 4-59

4-56 PMT Results in Bartoon Clay 4-60

xvii
Figure

4-57 su as a Function of KD from the DMT 4-60

5-l Modulus Definitions 5-2

5-2 Drained Poisson's Ratio Parameters for Granular Soils 5-5

5-3 Drained Poisson's Ratio versus PI for Several LOG Soils 5-5

5-4 Drained Poisson's Ratio versus OCR and Stress Level for
Sydney Kaolin 5-5

5-5 Normalized Undrained Modulus versus Stress Level and OCR 5-8

5-6 Generalized Undrained Modulus Ratio versus OCR and PI 5-8 I

5-7 Cam Clay Prediction of Undrained Initial Tangent Modulus Ratio 5-9

5-8 PMT Modulus of Clay versus N Value 5-10 I

5-9 Undrained Modulus for Deep Foundations in Compression
I
5-11

5-10 Undrained Modulus for (a) Drilled Shafts in Compression and Uplift
and (b) Spread Foundations in Uplift 5-11

5-11 Shear Modulus versus Shear Strain for Sands 5-13 I

5-12 Dynamic Shear Modulus versus N for Cohesive Soils 5-14

5-13 Comparative Plot of Drained Modulus Correlations for Sand 5-17 I

5-14 PMT Modulus of Sand versus N Value 5-18

5-15

5-16

5-17
Trend Between Dilatometer Modulus and N in Piedmont Sandy Silts

Variation of o with Dr for Sands in Calibration Chambers

CPT o Correlation for Ticino Sand

5-19

5-21

5-22
'I
5-18 Normalized Drained Modulus for (a) Drilled Shafts in Uplift and
(b) Spread Foundations in Uplift 5-22 I
5-19 Variation of Shear Modulus of Dry Sands with Void Ratio and
5-23

'I
Confining Stress

6-1 Consolidation Behavior 6-2

6-2 Time-Settlement Behavior 6-3

6-3 Representative Cc Relationships for Cohesive Soils 6-4

6-4 Compression and Unload-Reload Indices versus PI 6-5

6-6
I
6-5 Sensitivity-Compression Index Relationships

6-6 Compression Ratio versus Water Content 6-6 I

xviii t
I
I
I Figure

6-7 General Relationship Between Modulus Number and Porosity for NC

Soils 6-8

6-8 Modulus Number for NC Clay 6-8

6-9 SPT Constrained Modulus Coefficient f versus PI 6-9

6-10 Constrained Modulus versus qT from CPTU for Clays 6-10

6-11 Constrained Modulus from DMT Parameters 6-10

6-12 Effect of Grain Size on Sand Compressibility 6-12

6-13 Effect of Dr on Sand Compressibility 6-12

6-14 Modulus Number for NC Silts and Sands 6-13

6-15 Cv versus wL 6-14

6-16 Pore Water Stress Decay versus Piezocone Time Factor 6-15

6-17 Pore Water Stress Decay versus Dilatometer Time Factor 6-15

6-18 Coefficient of Secondary Compression versus Water Content for NC

Clays 6-16

7-1 Coefficient of Permeability versus Particle Size 7-2

7-2 Coefficient of Permeability versus Particle Size and Relative

Density 7-3

7-3 Vertical Coefficient of Permeability for Clay 7-3

7-4 Permeability Anisotropy for Various Natural Clays 7-4

8-1 Liquefaction Resistance Correlated with Modified SPT N Value 8-2

8-2 Liquefaction Resistance Correlated Indirectly with CPT Results 8-3

8-3 Liquefaction Resistance Correlated Directly with CPT Results 8-3

8-4 Liquefaction Resistance Correlated with DMT KD 8-4

A-1 Standard Split-Spoon Sampler A-1

A-2 SPT Safety Hammer A-2

A-3 SPT Donut Hammer A-3

A-4 Equipment Used to Perform the SPT A-4

B-1 Mechanical Cone Penetrometers B-1

B-2 Typical Designs of Electric Cone Penetrometers B-2

B-3 Electric CPT Data Acquisition System B-3

.-
xix
 I :.:?,

Figure Page
I~_,~·

B-4

B-5
Comparison of Begemann Mechanical and Fugro Electric Cones

Correlation of qc Between Electric and Mechanical Cones

B-4

B-5
j
, .•.
B-6 Common Piezocone Geometries B-6

B-7 Measured Pore Water Stresses in CPTU Tests B-7 I

C-1

C-2
Menard Pressuremeter Equipment

Typical Pressuremeter Test Curve

C-2

C-3
I
C-3

C-4
Self-Boring Pressuremeters

Examples of "Lift Off" Pressure

C-4

C-4
I
D-1

E-1
Dilatometer Test Equipment

Vane Geometries and Sizes

D-1

E-1
I
E-2

F-1
Common Vane Borers

Qualitative Relationship Between Relative Test Cost and Accuracy

E-2

F-6
I
G-1

G-2
Typical Soil Stress-Strain Behavior

CSSM Notation
G-1

G-2
I
H-1 Calibration Chamber Data for Various Sands H-5 ·J
H-2 CPT Calibration Chamber Correction Factor H-6

I
I
I
;I
l
il
!

:I
,.

!I
XX
~a

~
I
I
I
I TABLES

I Table

1-1 Categories of Analytical Methods for Soil Modeling 1-4

2-1 Soil Particle Size Identification 2-2

2-2 Relative Density of Cohesionless Soils 2-4

I 2-3 Approximate Plasticity and Dry Strength of Soil by Simple Tests 2-6

2-4 Approximate Cohesive Soil Strength by Simple Tests 2-7

-• 2-5

2-6

2-7
Approximate Cohesionless Soil Relative Density by Simple Tests

Unified Soil Classification System

Burmister Soil Identification System

2-7

2-9

2-10

2-8 Typical Soil Unit Weights 2-18

I'
2-9 Relative Density of Sand versus N 2-19
'
2-10 SPT Correction Factors for Field Procedures 2-22

2-11 SPT Correction Factors for Sand Variables 2-26

2-12 Relative Density of Sand versus qc 2-27

2-13 Consistency of Clay versus N 2-35

2-14 Consistency Index of Clay versus N and qc 2-36

4-1 Representative Values of ~tc 4-8

4-2 Relative Values of Effective Stress Friction Angles for

Cohesionless Soils 4-14

4-3 N versas ¢tc Relationships 4-15

4-4 qc versus ~tc 4-16

4-5 Relative Values of Effective Stress Friction Angles for Normally

Consolidated Cohesive Soils 4-25

4-6 Classification of Sensitivity 4-30

4-7 Typical Ranges in Af for Isotropically Consolidated Clays 4-33

xxi
Table

4-8 Evaluation of Modified Cam Clay Exponent A 4-48

4-9 Correction Factors for Su Compared with su from CIUC Test Results 4-50

4-10 Approximate Su versus N Relationship 4-54

5-l Typical Ranges of Drained Poisson's Ratio 5-4

5-2 Typical Ranges of Undrained Modulus for Clay 5-6

5-3 Typical Undrained Hyperbolic Modulus Parameters 5-7

5-4 Exponent M for Shear Modulus 5-13

5-5 Typical Ranges of Drained Modulus for Sand 5-15

5-6

5-7
Typical Drained Hyperbolic Modulus Parameters

Values of Rigidity Index for Selected Cohesionless Soils

5-16

5-16
I
6-l

6-2
Degree of Compressibility

Compressibility Data for Six Sands in Calibration Chamber Tests

6-4

6-ll
I
6-3

7-l
Compilation of Cae/Cc for Natural Soils

Coefficient of Permeability
6-17

7-l
I
A-1

B-1
Major Sources of Error in The Standard Penetration Test

Major Sources of Error in The Cone Penetration Test

A-6

B-9
I
C-1 Major PMT and SBPMT Variables

Major DMT Variables

C-6

D-4
I
D-1

E-1 Major Sources of Error in the Vane Shear Test E-4 I

F-1 Assessment of In-Situ Tests F-1

F-2 Usefulness of In-Situ Tests in Common Soil Conditions F-2 I

F-3 Historical Use, Mobilization and Access Requirements, and
Costs of In-Situ Tests

Estimates of In-Situ Test Variability

F-3

F-5
I
F-4

H-1 Calibration Chamber Data Base for Sands H-2

H-6
I
H-2 Boundary Conditions in Flexible-Wall Calibration Chambers

J-1 Correlations for Cohesive Soils J-2 I

J-2 Correlations for Cohesionless Soils J-4

I
xxii I
I
I

SYMBOLS

ENGLISH LETTERS - UPPER CASE

A - dilatometer test reading

Af - pore water stress parameter at failure

As - surface area of cone sleeve

Asl - cone area over which Ubt acts

As2 - cone area over which us acts

ASTM - American Society for Testing and Materials

B - foundation diameter or width; cone diameter; dilatometer test reading

Be - calibration chamber diameter

- piezocone parameter

I experimental constant in modified Cam clay model

I (
- N correction for aging

- N correction for borehole diameter

CER - N correction for energy ratio

CN - N correction for overburden stress

CoeR - N correction for overconsolidation

Cp - N correction for particle size

CR - N correction for rod length

cs - N correction for sampling method

Cc - compression index

Cq - qc correction for overburden stress

Cs - swelling index

Cu - uniformity coefficient D6Q/DlO

Cur - unload-reload index

xxiii
 I
Ca coefficient of secondary compression
I
Cere - Ca in terms of void ratio

Cere - Ca in terms of vertical strain I

I
CAUC consolidated anisotropic undrained triaxial compression

CF - clay fraction

CI - consistency index

CIDC consolidated isotropic drained triaxial compression I

CIUC

ClUE
consolidated isotropic undrained triaxial compression

- consolidated isotropic undrained triaxial extension

CK 0 UC - K 0 consolidated undrained triaxial compression

CK 0 UE - K 0 consolidated undrained triaxial extension

•I
COV

CPT
- coefficient of variation (standard deviation/mean)

- cone penetration test

I
GPTU

GR
- piezocone test

- compression ratio
I
GSL

GSSM
- critical state line

- critical state soil mechanics

I
D - depth; vane diameter; cone diameter I
Ds particle size at 5 percent finer

DlQ particle size at 10 percent finer I

Dso particle size at 50 percent finer

D6Q particle size at 60 percent finer

Dmax - maximum particle size

Dmin - minimum particle size

I
Dr

DMT
- relative density

- dilatometer test
= (emax - e)/(emax - emin)
I
DS - direct shear

DSS - direct simple shear

E - Young's modulus
I
xxiv I
I II
I
Eo - dilatometer modulus

EPMT - pressuremeter modulus

Ed - drained modulus

Eds - drained secant modulus

Ef - Young's modulus of foundation

Ei - initial tangent modulus

Es - secant modulus

Et - tangent modulus

Eu - undrained modulus

Eui - undrained initial tangent modulus

Eus - undrained secant modulus

Eut - undrained tangent modulus

ECPT - electric CPT

ER - energy ratio

F percent passing No. 200 sieve

Fv - Poisson's ratio parameter

FR - cone friction ratio - £sfqc (alternate for Rf)

G - shear modulus E/2(1 + v)

Gi - initial tangent G

Gmax - dynamic shear modulus

Gru - reload-unload G

Gs - specific gravity of solids; secant G

Gur - unload-reload G

Gv - Poisson's ratio parameter

H - vane height; height of drainage path

HOC - heavily overconsolidated

Io - density index; DMT material index

IRD - relative dilatancy index

If - moment of inertia of foundation

XXV
 I - rigidity index

Irr - reduced rigidity index

K - coefficient of horizontal soil stress = ah/Gv; vane constant

KA - coefficient of minimum active soil stress

Ko - DMT horizontal stress index

I
Kp coefficient of maximum passive soil stress

- in-situ or at-rest K = aho/Gvo J

Konc - normally consolidated K0

- K0 during primary unloading

I
- calibration chamber correction factor

subgrade reaction modulus; ratio of su in extension to compression

LI

LOG
- liquidity index = (wn - wp)/(wL - wp)

- lightly overconsolidated
I
M - critical state failure parameter; constrained modulus; earthquake
magnitude; exponent in Gmax relationship I
- drained constrained modulus

- drained secant constrained modulus

I
MCPT
- drained tangent constrained modulus

- mechanical CPT
I
N

N6Q
- standard penetration test value

- corrected N for field procedures

I
- cone bearing factor

- PMT bearing factor

I
- N60 corrected to reference stress of one atmosphere I
- piezocone factor

NC - normally consolidated I
oc - overconsolidated

OCR - overconsolidation ratio = ap/Uvo I

- isotropic OCR Pmax!Po
I
=

OCRlimit - limiting OCR at which passive failure occurs

xxvi I
I
I

I'
OCRmax - maximum OCR
j PI - plasticity index wL - wp

PMT - pressuremeter test

PSG plane strain compression

PSE - plane strain extension

Q - soil mineralogy and compressibility coefficient for strength dilatancy

Qc - soil mineralogy and compressibility coefficient for cone

Qco - dimensionless cone tip resistance= (qc/Pa)/(av0 /pa)O.S

QocR - qc correction for overconsolidation (equal to CoeR)

R - particle roundness; fitting coefficient; radius

Re equivalent radius

Rf - failure ratio; cone friction ratio = fsfqc

S - dynamic stiffness coefficient; particle sphericity

St - sensitivity

SBPMT - self-boring pressuremeter test

S.D. - standard deviation

SL - stress level (fraction of strength mobilized)

SPT - standard penetration test

T - time factor; torque

TC triaxial compression

TE - triaxial extension

U - unconfined compression test

UU unconsolidated-undrained triaxial compression test

V - volume

VST - vane shear test

ENGLISH LETTERS - LOWER CASE

a - cone area ratio; modified Cam clay parameter for anisotropic compression

- su correction for overconsolidation

- su correction for strain rate

xxvii
 aTEST - su correction for test mode

amax - maximum horizontal acceleration at ground surface

b - intermediate effective principal stress factor

c - cohesion intercept

effective stress cohesion intercept

- alternative form of su

- coefficient of consolidation
I
- horizontal cv

d - internal cone diameter; modified Cam clay parameter for plane strain
compression

e - void ratio (usually in-situ) I

emax - maximum void ratio

emin

e0

f
- minimum void ratio

initial void ratio

- unit side resistance; SPT modulus coefficient

'I
fs

fsn
- cone side resistance

- normalized cone side resistance

I
ft

g
- corrected cone side resistance

- gravitational acceleration
I
k

kh
- coefficient of permeability

- horizontal k; alternative form for ks

I
- subgrade reaction modulus
I
- vertical k

m - modulus number; OCR exponent I

- reload coefficient

- coefficient of volumetric compressibility I

n -hyperbolic modulus exponent; cone exponent; porosity- ej(l +e); number of

nh
measurements or data points

- coefficient of subgrade reaction

I
p - alternative form of avo; applied stress
I
xxviii I
I
I

~
p effective mean normal stress - (al + a2 + a3)/3; applied stress

Pl - DMT expansion stress

P2 - dilatometer C reading at a particular time

PL - PMT limit stress

Pa - atmospheric pressure or stress (See Appendix I for numerical values.)

Pc alternative form of up

Pe - PMT expansion stress

Pf - PMT yield stress

Pf p at failure

P~x - maximum p

Po - DMT contact stress; PMT total horizontal stress

Po initial effective p

q - shear stress

I qD

qT
- DMT tip resistance

corrected qc

qc - cone tip resistance

qn - qc standardized to reference stress of one atmosphere

qu - unconfined compression strength - 2 su

r critical state spacing ratio; sample correlation coefficient

r2 - coefficient of determination

s - slope of PMT plot of Pe vs. €c

su - undrained shear strength

sur - remolded su

Su(VST) - Su from VST

t - time

Ubt - pore water stress behind cone tip

urn - measured pore water stress during cone penetration

u0 - hydrostatic stress

Us - pore water stress behind cone sleeve

xxix
 - pore water stress on cone tip/face

- in-situ, natural water content

I
- liquid limit I
i
Wp - plastic limit

z - depth I
- DMT gage pressure deviation

GREEK LETTERS - UPPER CASE

I
~V - volume change

~u - excess pore water stress

~Ubt - excess pore water stress measured behind the cone tip I
~u 0 - ~u at time zero

~ut - excess pore water stress measured on the cone tip or face

~a1,3 - major, minor principal stress increment

~¢1-5 - corrections to ~cv I

~~r - change in residual friction angle

A - critical state parameter I

GREEK LETTERS - LOWER CASE I
- CPT parameter relating modulus to qc; K0 unload coefficient

0 VST - empirical vane factor I

- DMT K0 coefficient

f3o - DMT OCR coefficient I

- unit weight; shear strain

effective unit weight I

-I
'Yd - dry -y

'Ysat - saturated -y

-y total - total -y

'Yw - "'f of water

- stress rotation angle; displacement

I
XXX
I
I
 - strain

strain rate

ea - axial strain

ec - cavity strain

er - radial strain

ev - vertical or volumetric strain

K - modulus number; critical state parameter for isotropic swelling index

A - critical state parameter for isotropic compression index

p - micron (lo-6 meters); VST correction factor

v - Poisson's ratio

vd drained v

vdi - initial tangent drained v

vu - undrained v - 0.5

p - density

Pd - in-situ dry density

p dnax - maximum Pd

p dnin - minimum Pd

a - total normal stress

a effective normal stress

0 1,2 ,3 - major, intermediate, minor principal a

al,2,3 - major, intermediate, minor principal a

0 a,b,c - confining stress a, b, and c

a3c minor principal effective confining stress

am - initial horizontal a
am initial horizontal a
ai isotropic overburden a
ao .
- current vertical a· mean principal a
ap - maximum vertical a·. preconsolidation stress

qv - vertical a

xxxi
 vertical o
- mean ov

- mean ov

alternative form of op

vertical (or overburden) u

vertical (or overburden) o

- shear strength; shear stress
I
Tav

Th
- average cyclic stress

- shear stress in DSS

I
ifJ - friction angle I
~ - effective stress friction angle

~CV' - fully softened, constant volume, or critical state ~ I

~ds - direct shear ~

~p - peak ~ I
~psc plane strain compression ~

~r - residual ifJ
I
- relative friction angle- (~tc - 25.)/(45• - 25°)
ifJrel

~secant - secant ~
I
triaxial compression ~

- triaxial extension ifJ

I
- dilation angle
I
I
I
I
I
xxxii I
I
I

Section 1

INTRODUCTION AND BACKGROUND

This manual has been prepared to assist foundation engineers in the selection of
soil parameters, primarily for the geotechnical foundation design of transmission
line structures. It also will serve as a useful reference for other geotechnical
problems. Soil is a complex engineering material, and its properties are not
unique or constant. Instead, they vary with many environmental factors (e.g.,
time, stress history, water table fluctuation, etc.), as discussed in most geotech-
nical reference books.

Because of the complexity of soil behavior, empirical correlations are used exten-
sively in evaluating soil parameters. In this manual, an attempt has been made to
summarize the most pertinent of these empirical laboratory and in-situ test corre-
lations in an organized manner. The emphasis is on relatively common tests and
several newer tests that are seeing increased use in practice.

Within this section, the necessary background is presented to understand and appre-
ciate the nature of soil correlations and modeling, and the scope of this manual is
outlined.

I SOIL CORRELATIONS

The analysis of all geotechnical problems requires the adoption of a soil beha-
vioral model, complete with all relevant soil properties. These soil properties
are not known beforehand, and therefore the design engineer must either measure the
properties under controlled conditions in the laboratory or field or estimate the
I properties from other test data. These estimates are made most often from labora-
tory index te5ts and in-situ test results, which are correlated to the soil proper-

I ties either by calibration studies or by back-calculation from full-scale load test

data obtained in the field.

Comprehensive characterization of the soil at a particular site would require an

elaborate and costly testing program, well beyond the scope of most project budg-
ets. Instead, the design engineer must rely upon more limited soil information,

1-l
 I
and that is when correlations become most useful. However, caution must always be
exercised when using broad, generalized correlations of index parameters or in-situ
I
I
test results with soil properties. The source, extent, and limitations of each
correlation should be examined carefully before use to ensure that extrapolation is
not being done beyond the original boundary conditions. "Local" calibrations,
where available, are to be preferred over the broad, generalized correlations.
I
In addition, many of the common correlations in the literature have been developed
from test data on relatively insensitive clays of low to moderate plasticity and on I
unaged quartz sands reconstituted in the laboratory. Extrapolation of these corre-
lations to "special" soils, such as very soft clays, organic clays, sensitive
clays, fissured clays, cemented soils, calcareous sands, micaceous sands, collap-
I
sible soils, and frozen soils, should be done with particular care because the cor-
relations do not apply strictly to these soil deposits. Careful examination of the I
soil samples and reference to available geologic and soil survey maps should be
made to detect the possible presence of these soils. The same special care should
be exercised in remote areas and where no prior experience has been gained. If any
I
"special" soils are present, or if no experience has been documented in a given
area, a qualified geotechnical expert should be consulted for guidance. I
SOIL AND TEST VARIABILITY

Soil is a complex engineering material which has been formed by a combination of

I
various geologic, environmental, and chemical processes.
are continuing and may be modifying the soil in-situ.
Many of these processes
Because of these natural I
processes, all soil properties in-situ will vary vertically and horizontally. Even
under the most controlled laboratory test conditions, soil properties will exhibit
variability. This variability becomes more pronounced in the field where the natu-
I
ral geologic environment is introduced.
additional uncertainty is introduced.
When empirical correlations are used,
These levels of uncertainty must be con- I
sidered when assessing the reliability of a particular foundation design.

Variability also may be introduced by the type of laboratory or in-situ test used.
I
Each available test will provide a different test result because of differing boun-
dary conditions and loading mechanisms. Figure 1-1 illustrates these variables for I
some of the common laboratory strength tests and field tests. For the laboratory
strength tests, corrections are necessary to interrelate the particular test
results because of the different boundary conditions. For the field tests, differ-
I
ent in-situ responses are being measured in the different tests, as described in
Appendices A through E. Each test has its own variability, and the relative merits I
l-2
I
I
 Laboratory Strength Tests

~~®
TC TE OS DSS PSC PSE

Field Tests

SPT CPT PMT DMT VST

SYMBOLS: TC triaxial compression SPT - standard penetration test

TE - triaxial extension CPT - cone penetration test
DS - direct shear PMT - pressuremeter test
DSS - direct simple shear DMT - dilatometer test
PSG - plane strain compression VST - vane shear test
PSE - plane strain extension

Figure 1-1. Common Laboratory Strength Tests and Field Tests

of each test should be considered within the overall project context. Appendix F
provides a general comparison of these field test methods.

SOIL MODELING

Wroth and Houlsby (l) have stated succinctly that correlations ideally should be
(a) based on a physical appreciation of why the properties can be expected to be
related, (b) set against a background of theory, and (c) expressed in terms of
dimensionless variables to allow scaling. These thoughts should always be kept in
mind when using any type of correlation.

It also must be remembered how complex soil behavior really is. Ladd, et al. (£)
described this complexity as follows.

"A generalized model of the stress-strain behavior of soils

should ideally account for nonlinearity, yielding, variable
dilatancy (volume changes caused by shear stress), and ani-
sotropy (both inherent and stress system induced), plus the
behavioral dependence on stress path, stress system (orien-
tation of 01 and relative magnitude of 02), and stress his-
tory (both initial and changes due to consolidation)".

1-3
 1

I
Table 1-1 summarizes the major categories of analytical models that currently are
available for representing the behavior of soils. These models range from rather
I
complex (I) to advanced (II) to simple (III) descriptions of soil. Constitutive
models for soil behavior require input in the form of soil properties and in-situ I
parameters. In most cases for transmission line structure foundations, Category
III models may be most appropriate at the present time.
I
Jamiolkowski, et al. (1) also discuss the available laboratory and field tests in
use for characterizing soil. Their discussion focuses on a wide range of soil
behavior issues and might suggest that soil modeling is a most difficult task.
However, new efforts in research and development have resulted in considerable
progress in understanding soil behavior. The calibration and modification of soil
models have been made possible by the back-analysis of performance data from full-
scale field structures, such as deep foundations, embankments, tunnels, offshore I
platforms, and high-rise buildings. As additional field performance data become
available, newer and more reliable correlations undoubtedly will be developed.
This progress in research ideally will allow foundation design to evolve from Cate-
I
gory III in Table 1-1 to Categories II and then I, at which time all of the neces-
sary soil behavior issues will be addressed. I
I
Table 1-1

CATEGORIES OF ANALYTICAL METHODS FOR SOIL MODELING

I
I
Category Main Features of Models Determination of Soil Parameters

I Very advanced modeLs using non- Only from sophisticated laboratory

I
linear elastic-plastic time-depen- tests, with the exception of vari-
dent laws which possibly incorpo-
rate anisotropic behavior
ables which must be obtained from
in-situ tests I
II Advanced models using constitu- Laboratory tests which are only a
tive incremental elastic-plastic little more sophisticated than con-
laws and nonlinear elastic rela- ventional tests; in-situ tests also
tionships appropriate

lli Simple continuum, such as isotro-

pic elastic continuum, including
Conventional laboratory and in-situ
tests
I
layering and empirical models

Source: Adapted from Jamiolkowski, et al. (1), p. 58. I

l-4 I
I
I
At the present time, there is one modeling concept of soil behavior which is of
some practical use for estimating soil properties. This concept is known as Criti-
cal State Soil Mechanics (CSSM) and is described in Appendix G. With this concept,
a general predictor for soil behavior has emerged. Strictly speaking, CSSM is
applicable only to remolded, insensitive soils without aging, cementing, and other
environmental influences. However, the resulting model predicts well the behavior
of normally consolidated, insensitive soils, also without aging, cementing, and
other environmental influences. In other soils, the model effectively provides a
lower bound on the predicted property, such as the undrained shear strength. For
these reasons, property prediction by CSSM has been included in this manual as a
valuable reference on probable lower bound behavior of natural soils.

SCOPE OF MANUAL

In the following sections, commonly used correlations have been compiled that are
helpful for estimating soil properties. Within a particular topic, these correla-
tions are selected and presented in an approximate evolutionary order to represent
the development of the relationship as newer research findings became available.
In certain instances, it was necessary to develop new correlations to supplement
existing ones. Where new correlations have been developed, the complete data set
and regression analysis results are presented to provide a measure of the validity
of the relationship. The regression equation is presented first, normally using an
assumed intercept of zero for simplicity. The number of data points in the corre-
lation is denoted by n, and the standard deviation (S.D.) is given to allow assess-
ment of the dispersion around the regression line. Also given is the coefficient
of determination, r2, which is the ratio of the explained variation to the total
variation. For r2- l, a perfect correlation exists; for r2- 0, no correlation
exists; and for r2- 0.75, 75 percent of the observed variation in y may be attrib-
uted to x. In almost all cases presented, the value of r2 for a zero intercept was
only l or 2 percent less than the r2 for a regression line with an intercept. The
sample correlation coefficient, r, is the statistic for testing the significance of
a simple two-variable linear relationship (i.e., how well the data fit a linear
relationship). For r - 0, no linearity exists while, for r = ± l, direct linearity
exists.

By presenting the complete data set, the regression equation, and some pertinent
statistics (n, S.D., r2), the user will be able to assess the quality of the rela-
tionship and use the results accordingly. This format also will allow direct
incorporation of the results into evolving reliability-based design procedures.
Moroney (~) states rather directly in Figure l-2 the importance of presenting the

1-5
 IT IS DISHONEST TO PRESENT
THIS FOR THIS

y + y
.· : . .•
•
+•• • ••

••

X X

Figure 1-2. Importance of Proper Data Presentation

Source: Moroney (~). p. 29.

'
data properly.
I
Since this manual is directed toward the practicing engineer, its focus has been
limited to the more common tests available on a commercial basis and to those tests I
that are seeing increased use in practice. Included are the common laboratory
index and performance tests and the field standard penetration test (SPT), cone
penetration test (CPT), pressuremeter test (PMT), and vane shear test (VST). The
I
newer tests included are the dilatometer test (DMT), piezocone or cone penetration
test with pore water stress measurement (CPTU), and the self-boring pressuremeter
test (SBPMT). Intentionally not included are the wide variety of simple hand
I
devices which are intended primarily for field inspection purposes, such as the
pocket penetrometer, torvane, geostick, dynamic cone, etc. These are not design or
I
performance devices and should not be used as such. Also not included are scaled
tests such as the plate load test or centrifuge test, which may be used to model
full-scale foundation performance on a smaller scale.
I
Section 2 addresses basic soil characterization to define the soil material, while
I
Section 3 focuses on evaluating the in-situ soil stresses. The evaluation of soil
strength is covered in Section 4, while Sections 5 and 6 address elastic and time- I
dependent soil deformability, respectively. Section 7 covers soil permeability,
while Section 8 briefly addresses the special topic of liquefaction resistance.
I
Appendices A through F provide information on the various in-situ tests used in the
correlations, primarily for those readers who are not familiar with the tests. I
l-6
I
I
I
I This information was extracted largely from EPRI Reports EL-2870 (2) and EL-5507,
Vol. 2 (£). These reports should be consulted for further details on the tests.
Appendix G gives a brief summary of the Critical State Soil Mechanics concept, and
I Appendix H summarizes available CPT calibration chamber data used to develop a num-
ber of correlations in this manual.

I Within this manual, an effort has been made to present the relationships in dimen-
sionless form for ease in scaling to whatever units are desired by the user.
I Therefore, stresses have been made dimensionless by the atmospheric pressure or
stress, Pa• which is equal to 1.058 tsf, 14.7 psi, 101.3 kN/m2, etc. A simple,

I approximate conversion for preliminary work is that 1 atm ~ 1 tsf ~ 1 kg/cm2 ~ 100
kN/m2. These approximate conversions have been used liberally with previously
published work where the 1 or 2 percent variation would not be significant. All
I unit weights have been made dimensionless by the unit weight of fresh water,
which is equal to 62.4 pcf or 9.80 kN/m3.
~w.

Where lengths are included, dual units

I are given. A detailed unit conversions guide is given as Appendix I.

Lastly, Appendix J presents summary tables to assist the user in locating specific

I recommended correlations in this manual. These tables are not intended to be a

substitute for the text, which puts the correlations in proper perspective.

I Instead, they are intended to be a quick reference guide for the experienced user.

REFERENCES

I 1. Wroth, C. P. and Houlsby, G. T., "Soil Mechanics - Property Characterization

and Analysis Procedures", Proceedings, 11th International Conference on Soil
Mechanics and Foundation Engineering, Vol. 1, San Francisco, 1985, pp. 1-55.
I 2. Ladd, C. C., Foott, R., Ishihara, K., Schlosser, F., and Poulos, H. G.,
"Stress-Deformation and Strength Characteristics", Proceedings, 9th Interna-

I tional Conference on Soil Mechanics and Foundation Engineering, Vol. 2, Tokyo,

1977, pp. 421-494.

3. Jamiolkowski, M., Ladd, C. C., Germaine, J. T., and Lancellotta, R., "New
Developments in Field and Laboratory Testing of Soils", Proceedings, 11th
International Conference on Soil Mechanics and Foundation Engineering, Vol. 1,
San Francisco, 1985, pp. 57-154.

4. Moroney, M. J., Facts From Figures, 3rd Ed., Penguin Books, Baltimore, 1956,
472 p.

5. Kulhawy, F. H., Trautmann, C. H., Beech, J. F., O'Rourke, T. D., McGuire, W.,
~od, W. A., and Capano, C., "Transmission Line Structure Foundations for
Uplift-Compression Loading", Report El-2870, Electric Power Research Insti-
tute, Palo Alto, 1983, 412 p.

1-7
'6. Orchant, C. J., Kulhawy, F. H., and Trautmann, C. H., "Reliability-Based Foun-
dation Design for Transmission Line Structures: Critical Evaluation of In-Situ
Test Methods", Report EL-5507, Vol. 2, Electric Power Research Institute, Palo
Alto, 1988, 214 p.

i?J.

I
I
I
I
I
I
I
I
I
I
I
I
1-8 I
.I
(

Section 2

BASIC SOIL CHARACTERIZATION

One of the first steps in any geotechnical design problem is to develop an under-
standing and knowledge of the soil materials at the site. Soil is a complex engi-
neering material, and therefore it is important to know its basic characteristics
as thoroughly as possible before attempting to define its engineering design prop-
erties. In this section, procedures are presented to describe and classify soil,
to estimate its unit weight, and to estimate its physical characteristics. General
descriptions, simple index tests, and correlations with in-situ test results are
used where available.

SIMPLE DESCRIPTIONS

Simple descriptions for soil are useful because they help to establish the nature
and/or physical characteristics of the soil material in the laboratory or in-situ.
In terms of basic behavior, soils often are described simply as either cohesionless
or cohesive. Cohesionless soils include coarser-grained granular materials, such
as sands, gravels, and non-plastic silts. Cohesive soils include finer-grained
plastic materials, such as clays and plastic silts.

Particle Size and Distribution

The particle size and distribution are necessary to describe the basic nature of
soil. For coarse-grained soils, the size and distribution are determined using
nested sieves, as described in ASTM D422 (!) and D2217 (l). Identification by par-
ticle size is given in Table 2-1. For fine-grained soils, the size and distribu-
tion are determined by a hydrometer test(!). Clay-size particles generally are
defined as those being less than 2 microns (0.002 mrn).

From the particle size analyses, several parameters are defined which are of use in
later sections of this manual. These parameters are: D6o - particle size at which
60 percent of the sample is finer (by weight), Dso =mean grain size- particle
size at which 50 percent of the sample is finer, D10 effective grain size= par-
ticle size at which 10 percent of the sample is finer, and Cu = D6o/D10 =uniform-
ity coefficient. Soils with a high value of Cu are well-graded and contain a wide

2-1
 Table 2-1

SOIL PARTICLE SIZE IDENTIFICATION

Size Limits
Broad Group Name ASTM Sieve Number mm

Coarse-Grained Boulder > 12 in > 305

Cobble 12 in to 3 in 305 to 76

Coarse gravel 3 in to 3/4 in 76 to 19

Fine gravel 3/4 in to No. 4 sieve 19 to 4.75

Coarse sand No. 4 to No. 10 sieve 4.75 to 2.0 I

Medium sand No. 10 to No. 40 sieve 2.0 to 0.42

Fine sand No. 40 to No. 200 sieve 0.42 to 0.075 I

< No. 200 sieve < 0.075
I
Fine-Grained Silt and/or clay

Note: Particles finer than fine sand can not be discerned with the naked eye at
a distance of 8 in. (203 mm).

I
range of particle sizes, while soils with a low value of Cu are uniformly graded
and contain particles of similar sizes. I
Index Parameters for Cohesive Soils I
The relative consistency of cohesive soils is described by several useful index
parameters which are expressed as water contents at particular soil states.
consistency states are known as Atterberg limits, determined by ASTM D4318 (1).
These
I
The most common index parameters are: Wn = in-situ natural water content, wL =

liquid limit, wp = plastic limit, PI = wL - wp - plasticity index, and LI - (wn -

wp)/(wL - wp) = liquidity index. Soils with a liquid limit (wL) greater than 50
percent are termed "highly plastic". A plasticity index (PI) greater than 25 to 30
may mean troublesome soils with low strength, high compressibility, high shrink-
swell potential, etc. The liquidity index (LI) is an excellent indicator of geo-
logic history and relative soil properties, as shown schematically in Figure 2-1.
.,
I
2-2 I
.I
 wl
I
rn Wp
I . Decreasing water content
Sensitive _ _!-J_f;_ LOC HOC
Increasing OCR, K0
---- Increasing strength, modulus
Decreasing compressibility
>I 0 <0
Decreasing LI

NC normally consolidated
LOG lightly overconsolidated
HOC heavily overconsolidated
OCR overconsolidation ratio = apfavo
up maximum vertical effective stress in soil during its geologic history
avo vertical effective stress in-situ
aho horizontal effective stress in-situ
in-situ coefficient of horizontal soil stress

Figure 2-1. Liquidity Index Variations

Index Parameters for Cohesionless Soils

Cohesionless soils also can be represented by simple index parameters, generally

expressed in terms of either "unit weight" or "density". Unit weight (-y) is
defined as the soil weight per unit volume and is given by the units kN/m3 or
lb-forcejft3. Density (p) is defined as the soil mass per unit volume, with units
of kg/m3 or lb-mass/ft3. Although density actually is the preferred term in modern
SI usage, conventional engineering practice has favored unit weight, which will be
used in this manual. The ratio (-y/p) is the gravitational acceleration (g), which
is equal to 9.807 m/sec2 or 32.17 ftjsec2.

For cohesionless soils, the relative density (Dr) expresses the degree of compact-
ness with respect to both the loosest and densest states achieved by standard labo-
ratory procedures [ASTM D4253 (~)and D4254 (2)]. Most commonly, the relative den-
sity is expressed in terms of void ratio:

emax - e
(2-1)

in which e =in-situ void ratio, emax =maximum void ratio (loosest), and emin
minimum void ratio (densest). Alternatively, Dr can be expressed as:

2-3
 I
Pdmax(Pd - Pdmi,n)
(2-2)
I
Dr - Pd(Pdmax - Pdmin)

in which Pd- in-situ dry density, Pdmax- maximum dry density, and Pdmin- minimum
I
dry density. In this equation, unit weight can be used alternatively in place of
density. In some instances, the degree of relative compactness is described in
terms of the density index (ID):
I
Pd - Pdmin
I
ID - (2-3)
Pdmax - Pdmin
I
Relative density is a useful parameter for describing the relative behavior of
cohesionless soils. Standard terminology is given in Table 2-2.
to be used more commonly in the U.S.
Column (a) tends
Increasing Dr generally means increasing
I
strength and decreasing compressibility. If Dr is negative, a collapsible soil
structure may be present, such as can occur with honeycombed soils and very loose
cemented or calcareous sands with e > emax· The applicability of Dr is limited to
cohesionless soils having less than 15 percent fines. In practice, it has been
misapplied occasionally to soils having greater than 15 percent fines, with ques-
.I
tionable results. Since it is very difficult to obtain truly undisturbed samples
of clean sands, the direct measurement of Dr also is difficult. In addition, the I
Table 2-2
I
RELATIVE DENSITY OF COHESIONLESS SOILS
I
Relative Density Dr (%) I
(a) (b)

Very loose 0 to 15 0 to 20
I
Loose 15 to 35 20 to 40

Medium 35 to 65 40 to 60

Dense 65 to 85 60 to 80
I
•I
Very dense 85 to 100 80 to 100

a - Source: Lambe and Whitman (~), p. 31.

b Source: Meyerhof (?_) , p. 17.

2-4

I
 in-situ void ratio (e) is compared to emax and emin• both of which are subject to
considerable error in their determination in the laboratory. For these reasons, Dr
should be considered only as an index parameter.

For a variety of natural and artificially-prepared mixtures of sands, emax and emin
depend primarily on the particle roundness (R) and the uniformity coefficient
(Cu)· The roundness is defined as the ratio of the minimum radius of the particle
edges to the inscribed radius of the entire particle. Although R is difficult to
measure, it can be estimated from the apparent angularity of the grains, as shown
in Figure 2-2. Combined with a particle size analysis, the emax and emin values
can be estimated from Figure 2-3. This figure is valid for clean sands with normal
to moderately-skewed particle size distributions.

R<0.17 R:::::0.20 R:::::0.30 R:::::0.35 R:::::0.50 R:::::0.70

Very Angular Angular Subangular Subrounded Rounded Well-Rounded

Figure 2-2. Particle Roundness Definitions

Source: Adapted from Youd (~).

1.6

1.4
X
0
E 1.2 1.0
(1)

6 c
:;::
0
·e 0.8
0:: 1.0 (1)

-o ~
~ 0.8
0
0:: 0.6
E -o
~
::;)

E
·;;: 0.6 0.4
0 E
2 ::;)

E
0.4 ·c: 0.2
(a) 2 (b)
0.2 0
I 2 3 4 6 10 15 I 2 3 4 6 10 15
Uniformity Coefficient, Cu = 0 60 /0 10 Uniformity
Figure 2-3. Generalized Curves for Estimating emax and emin

Source: Youd (~), p. 108.

2-5
 I
J Characterization by Simple Field Tests
I
For preliminary reconnaissance studies and quality control during construction,
simple manual field tests are useful in describing the characteristics of in-place
soils. For cohesive soils, Table 2-3 provides guide 1 ;.nes for approximate plasti-
I
city characteristics. Similarly, manual tests can provide a crude index of the
unconfined compressive strength (qu) or undrained shear strength (su) of cohesive I
soils, as indicated in Table 2-4. A pocket penetrometer (for qu) or torvane (for
su) also can be used to provide these approximate values, even though these meas-
urements are crude.
I
Simple field tests similarly are available for evaluating the characteristics of
cohesionless soils. Table 2-5 provides rough guidelines for this purpose by use of
I
a reinforcing bar.
I
Color and Odor

Color also may be a useful indicator of some soil characteristics. For example, I
yellow and red hues often represent iron oxides in deeply weathered soil profiles.
Dark greens and browns often indicate organic soils, particularly when coupled with
the distinctive odor of decaying organic matter. Odor sometimes is an indicator of
I
contaminants as well. Color also can assist in differentiating topsoil and the
depth and extent of weathering. For these reasons, color and odor (if any) should I
always be considered an integral part of any soil description.

I
Table 2-3

APPROXIMATE PLASTICITY AND DRY STRENGTH OF SOIL BY SIMPLE TESTS

I
I
Plasticity PI (%) Dry Strength Field Test on Air-Dried Sample
I
Falls apart easily
Nonplastic

Slightly plastic
0 to 3

3 to 15
Very low

Slight Easily crushed with fingers 'I

Medium plastic 15 to 30 Medium Difficult to crush

Highly plastic > 30 High Impossible to crush with fingers I

Source: Sowers <2). p. 83.
I
2-6 I
J
I
Table 2-4

APPROXIMATE COHESIVE SOIL STRENGTH BY SIMPLE TESTS

I
qu

Strength (ksf) (kN/m2) Field Test

I Very soft 0 to 1/2 0 to 25 Squeezes between fingers when fist is closed

Soft 1/2 to 1 25 to 50 Easily molded by fingers

I Firm 1 to 2 50 to 100 Molded by strong pressure of fingers

I Stiff

Very stiff
2 to 3

3 to 4
100 to 150

150 to 200
Dented by strong pressure of fingers

Dented only slightly by finger pressure

I Hard

Note: qu
> 4 > 200

unconfined compressive strength

Dented only slightly by pencil point

su undrained shear strength

I Source: Sowers (~), p. 80.

I
Table 2-5

I APPROXIMATE COHESIONLESS SOIL RELATIVE DENSITY BY SIMPLE TESTS

I Density Dr (%) Field Test

Loose 0 to 50 Easily penetrated with 0.5 in. (12 mm) reinforcing rod
pushed by hand

Firm 50 to 70 Easily penetrated with 0.5 in. (12 mm) reinforcing rod
driven with 5 lb (2.3 kg) hammer

Dense 70 to 90 Penetrated a foot with 0.5 in. (12 mm) reinforcing rod
driven with 5 lb (2.3 kg) hammer

Very dense 90 to 100 Penetrated only a few inches with 0.5 in. (12 mm)
reinforcing rod driven with 5 1b (2.3 kg) hammer

Note: generally refers to shallow depths in uncemented quartz and feldspar sands

Source: Sowers (~), p. 81.

2-7
 I
CLASSIFICATION I
General Classification and Identification Systems

Classification systems are useful for grouping together soils of similar particle
size and plasticity characteristics. By this grouping into pre-established cate-
I
gories, consistent terminology can be employed to represent a soil fitting within
the bounds of a particular category. The most widely used of these systems is the
I
Unified Soil Classification System [ASTM D2487 (10) and D2488 (11)], given in Table
2-6.
needed.
To use this system properly, both particle size and Atterberg limits data are
With the particle size and Atterberg limits data, the soil is classified
I
using the pre-established group symbols in Table 2-6. Plastic soils utilize the
plasticity chart shown as well. Note that if any soils plot above the "U" line in
the plasticit~ chart, the data should be questioned and verified. Further details
are given in the ASTM Standards.

Other well-known special purpose classification systems have been developed by the
U. S. Department of Agriculture (USDA) for agricultural purposes, the Federal Avia-
I
tion Administration (FAA) for airport pavements, and the American Association of
State Highway and Transportation Officials (AASHTO) for highway pavements.
systems normally are not used in foundation engineering.
These I
As an alternative, Burmister (12, 13) developed a soil identification system for
I
both field and laboratory use. As compared with the classification systems which
use pre-established soil group categories, Burmister's approach uses rapid and sim-
ple visual-manual procedures to approximate the particle size and gradation and
I
overall plasticity index.
are given in Table 2-7.
Essential features of the resulting soil identification
With this system, approximate percentages of the principal
I
and minor components are estimated using the notation in Table 2-7a. Particle size
and gradation terms are defined in Tables 2-7b and c. For the fines (percent< No.
200 sieve), the overall plasticity is estimated and then described using the nota-
I
tion in Table 2-7d. Example identifications also are given with this table. Once
I
the straightforward visual-manual procedures are mastered, some 15 to 30 samples
per hour can be identified in terms of their approximate particle size distribution
and plasticity index. ,.
I
Cone Penetration Test (CPT) Classifications
'I
~

The CPT has been used widely for many years as a site investigation device.
:,
t,
Although no soil sample is recovered, the cone tip resistance (qc), cone side
resistance (fs), and friction ratio (Rf- FR- fs/qc) have been employed to

g
:•.

2-8
 Table 2-6

UNIFIED SOIL CLASSIFICATION SYSTEM

Soil Classification
Criteria lor Assigning Group Symbols and Group Names Using Laboratory TestsA Group
Group Name 8
Symbol
Coarse-Grained Soils Gravels Clean Gravels Cu 2: 4 and 1 :s Cc :s 3' GW Well-graded gravel'
More than 50 10 retained on No. More than 50 10 of coarse Less than 5 10 fines c
Cu < 4 and/or 1 > Cc > 3' GP Poorty graded graveiF
200 sieve fraction retained on No. 4
sieve Gravels wi1h Fines More Fines classify as Ml or MH GM Silty graveiF.O,H.
than 12 10 finesC
Fones classify as CL or CH GC Clayey graveiF.O.H

Sands Clean Sands Cu 2: 6 and 1 :5 Cc :5 3' SW Well-graded sand

50 10 or more of ooarse Less than 5 10 fines 0
traction passes No. 4 sieve Cu<6and/or1 >Cc>3' SP Poorty graded sand'
Sands wi1h Fines Fines dassify as ML or MH SM Silly sandO.H.J
More than 12 10 fines 0
Fines dassify as CL or CH sc Clayay sand 0 .H·'
Fine-Grained Soils Sins and Clays inorganic PI > 7 and plots on or above ·A· line J CL lean ctay""·LM
50 10 or more passes the No. liquid limit less than 50
PI< 4 or plots below "A" ilneJ Ml SinK.LM
200 sieve
organic liquid limit - oven dried < O. Qr~nic claf:·L.AI.H
75 OL
liquid 1Jm1t - not dned Organic sirtK·L.M.o
S~ts and Clays inorganic PI plots on or above "A" hne CH Fat clayK.LM
Liquid limit 50 or more
PI plots below "A" line MH Elastic siftK.L."'

orgarnc liquid lim1t - oven dried < O. OH 0r:2anic da'l:.L.M.P

75
liquid km1t - not dried Organic sift".L.AI,Q
Highly organic soils Primarily organic maner. dal1< "" color. and organic odor PT Peat
Based on the material passing the 3-<n. (75-mm) "' II soil contains 2: 30 plus No. 200, pre-
~
A '!;
sieve.
E Cu = Ooof010 dominantly gravel, add "gravelly" to group name.
0 10 x 0 60
8 II field sample contained cobbles or boulders. or "PI 2: 4 and plots on or above "A" line.
'II soil contains 2: 1510 sand. add "wilh sand" to 0
both. add "with cobbles or boulders. or both" to group name. PI< 4 or plots below "A" line.
group name. 0
If fines classify as CL-ML. use dual symbol GC- P PI plots on or above ·A· line.
0
c Gravels with 5 to 12 10 fines require dual GM. or SC-SM. PI plots below ·A" line.
symbols: " II fines are organic. add "with organic fines· to
GW-GM well.graded gravel with sin group name.
GW-GC well.graded gravel with day 1
1f soil contains 2: 15 '!; gravel. add "with gravel"
GP-GM poorly graded gravel with sin to group name.
GP-GC poony graded gravel w~h day J If Anerberg limits plot in hatched area. soil is a
0 Sands with 5 to 12 '!; fines require dual CL-ML. silty clay.
symbols: K II soil contains 15 to 2910 plus No. 200. add
SW..SM well-graded sand wi1h silt "with sand" or "wilh gravel." whichever is pre-
sw-sc well.graded sand with day dominant.
SP-SM poony graded sand wrth silt <If soil contains 2: 30 '!; plus No. 200. pre-
SP..SC poony graded sand with day dominantly sand, add "sandy" to group name.

60

.....
a..
X
Q) 40
"0
.....c:
->-
( .)

(;; MH or OH
0
a..

16 20 40 50 60 80 100
Liquid Limit, w L

Source: American Society for Testing and Materials (10), pp. 289, 292.

2-9
 I
Table 2-7

BURMISTER SOIL IDENTIFICATION SYSTEM

I
I
(a) Terms Describing Composition of Cohesionless Soils

Identification ProEortion
I
Component Written Symbol Written Symbol % by Weight

Principal GRAVEL G > 50

I
SAND
SILT
s
~
> 50
> 50 I
Minor Gravel G and a 35 to 50
Sand s some s 20 to 35 I
Silt ~ little 1 10 to 20
trace t 1 to 10
I
I
(b) Terms Describing Gradation of Cohesionless Soils
I
Designation
Written Symbol Defining Proportions
I
coarse medium to fine emf all fractions > 10% I
coarse to medium em < 10% fine
medium to fine mf < 10% coarse I
coarse c < 10% medium and fine
medium m < 10% coarse and fine
fine f < 10% coarse and medium

NOTE: For proportions in (a) and (b), use+ for upper limit and - for lower limit.

2-10
 Table 2-7 (cont'd)

BURMISTER SOIL IDENTIFICATION SYSTEM

(c) Particle Size Definitions

Soil Fraction Sieve Number and Size

Gravel coarse 3 in to 1 in (76 mm to 25 mm)

medium 1 in to 3/8 in (25 mm to 9.5 mm)
fine 3/8 in to No. 10 (9.5 mm to 2.0 mm)

Sand coarse No. 10 to No. 30 (2.0 mm to 0.6 mm)

medium No. 30 to No. 60 (0.6 mm to 0.25 mm)
fine No. 60 to No. 200 (0.25 mm to 0.075 mm)

Silt <No. 200 (< 0.075 mm)

(d) Terms Describing Cohesive Soils Based on Overall Plasticity

Overall Plasticity Principal Component Minor Component

Plasticity
Written Symbol Index Written Symbol Written Symbol

Non-plastic 0 SILT ~ Silt ~

Slight Sl 1 to 5 Clayey SILT Cy~ Clayey Silt Cy~

Low L 5 to 10 SILT & CLAY ~ & c Silt & Clay ~ & c

Medium M 10 to 20 CLAY & SILT c & ~ Clay & Silt c & ~

High H 20 to 40 Silty CLAY ~yc Silty Clay ~yc

Very High VH > 40 CLAY c Clay c

EXAMPLES: Full - coarse+ medium to fine- SAND, some- medium fine Gravel,
trace+ Silt
Abbreviated - c+mf- SAND, s-·mf Gravel, t+·silt
Shorthand - c+mf- S, s-·mfG, t+·~
Full - CLAY & SILT, little+ coarse- medium to fine+ Sand, Medium
Plasticity
Abbreviated - CLAY & SILT, l+·c-mf+ S, M·Pl
Shorthand - C & ~. l+·c-mf+S, M·Pl
NOTE: Principal component (> SO%) always listed first. If no principal component,
list sand first.

Source: Burmister (12, 13).

L ________ Z-11 - -
 classify the soil in-situ. Since soil classification by the CPT is an empirical
approach, it has been an evolutionary process which has required periodic updates
as new and larger data bases have been collected and evaluated. Two representative
examples of the earlier interpretations of CPT data are shown in Figure 2-4. Fur-
ther research led to empirical classification charts for the mechanical Begemann
friction-cone, as shown in Figure 2-5. Similar developments led to classification
charts for electric friction cones, as shown in Figure 2-6 in original form and in
Figure 2-7 in simplified form.

Recently, it has been realized that the correlations should be made dimensionless
by appropriate scaling factors (Wroth, 18). Numerous field studies have shown that
the cone side resistance increases proportionally with confining stress. For the
tip resistance, the proportionality varies with soil type (e.g., Jamiolkowski, et
al., 19). Therefore, at the present time, the most rational approach to soil clas-
sification by the CPT is by using dimensionless parameters, as given in Figure 2-8.
I
Soil classification using Figure 2-8 requires an iterative approach, since qc is
divided by a power function of the vertical effective stress, (ov 0 )n, and the expo-
nent (n) depends upon the soil type. This exponent (n) increases from about 0.5
I
for sands to approximately l for clays.
obtained from Figure 2-7.
An initial estimate of soil type may be
A first estimate of n for the iterative solution then I
I
400 Mechanical Friction 400 Electric Friction

0.0
......
u
r::T
Cone
1.3
Cone 1.1

~
I
300 300 1.4

I
Q)
1.6 0:
u u..
c::
.E 1.8 .9
"'
·;n
Q) 200 200 2.2 0:
0
0:
c.
i=
2.5 c::
2.9 .9
u
3.3 ·;::: I
,.·I
~ 100 100 4.0 u..
0 5.0
u 8.1
10.0
0~~~~==~~~~
0 3 4

Cone Side Resistance, fsiPa Cone Side Resistance, fsiPa

t
a. Mechanical Friction Cone b. Electric Friction Cone

Figure 2-4.

Source:
Early Soil Classification by CPT

Laboratorium voor Grondmechanica (14), p. 29.

II
•
f

2-12 I
I
1000
North
400 l2 Florida
(..)
0 Soils
0.0 0::: 0.0
........ w
0 :::E ........
CT :::i Dense or SILT-SAND CT
0

(/)
cemented MIXTURES
Q.l 100 0 Q.l
u
c z CLAYEY 0

Vi
0

IJ)
Q.l
0::
40
4
(/)

:J
..J
w
:X:
SAND
SANDS
AND
SILTS - 0

"iii
Q.l
c
1 /)

(/) 0:::
0. _v!!..ry stiff
>-
0::: Loose
0.
~

u
Q.l
c
0
10
8
w
>
-
Stif;--;,... ...

Medium- ~GANIC
/
i=
(1)
c
0
6 - - u
Soft-- I..L CLAYS 8
4 MIXED
SOILS
Very soft
2
0 2 4 6 2 4 6
Friction Ratio, fs/qc (%) Friction Ratio, fslqc (%)

Figure 2-5. Soil Classification by Figure 2-6. Soil Classification by Fugro

Mechanical Friction CPT Electric Friction CPT

Source: Schmertmann (15), p. 7. Source: Douglas and Olsen (16), p. 222.

400
c:
Sands/
-;:p
...... Grovel
0 a sand
/Silty •t?
100- / sands/
/ /sandy ~ 0
Q.l / / silts /clayey .!!!!
u 40./ /and silts/silts and I
c <V
0 / ;silty cloys/ 0
c:
Vi / / I 0
"iii u;
Q.l . // / /Cloys ·;;;
0:: 10,.......- / / - <V
0::
0. 8 / / c.
~ 6 / / F
Q.l 4 / / Peat-- Q)
c / / c:
0
0 / (..)
u 2 ./

IL---~--~----~--~·----~--~ 0.1
0 2 4 6
Friction Ratio, fslqc (%) Cone Side Resistance, fsn =fs /G-vo

Figure 2-7. Simplified Soil Figure 2-8. Most Recent Soil

Classification by Fugro Classification by Fugro
Electric Friction CPT Electric Friction CPT

Source: Robertson and Campanella (17), Source: Olsen and Farr (20), p. 858.
p. 721.

2-13
can be made from Figure 2-8.

As described in Appendix B, different results commonly are obtained using different

cones. Therefore, adjustments to the following figures may be warranted as a func-
tion of cone type and shape, as given in Appendix B.

Piezocone Penetration Test (CPTU) Classifications

With the recent development of the piezocone, which measures the total penetration
pore water stress (urn) in addition to qc and fs, the ability of the cone penetrome-
ter to delineate soil stratigraphy and provide an accurate classification of soil
type is enhanced greatly. In loose, contractive sands, the value of Um closely
follows the hydrostatic stress (u 0 ). In dense, dilatant sands, Urn may be less than
u0 . In clays, cone penetration generates excess pore water stresses which are
recorded by the pore water transducer. Two of the recent soil classification sys-
I
tems based on CPTU measurements are given in Figures 2-9 and 2-10. Other classifi-
cation charts are given by Robertson, et al. (23).
the parameter Bq is used, which is defined as:
In the first of these figures, I
I
(2-4)

in which Um =measured total pore water stress (usually behind the tip), u 0 =
I
hydrostatic pore water stress, qT = corrected cone tip resistance, and avo = total
overburden stress.
I
One important finding which has evolved from the development of piezocones is that
the cone tip and side resistances must be corrected for pore water stress effects
I
acting on unequal areas of the cone geometry.
given by:
The corrected tip resistance is
I
(2-5) I
in which qc =measured cone tip resistance, a= net area ratio for the particular
cone (See Figure 2-11.), and Ubt =pore water stress behind the tip. Similarly,
I
g
the correction for cone side resistance is given by:

(2-6)
I
in which Us pore water stress behind the sleeve, As surface area of the sleeve,
I
2-14 I
 0.0 140 Q.
0
...... ......
1-
CT 0
::>
cU 120 I
(,)
c::: E
J2 .3
Ill
·;;; .,;
Q) Ill
0:: ~
Q.
(i)
i=
~
...
Q)
c::: 0
0 3:
(..)

-o
...
Q)

Q) &
0Q)
... Q)
c:::
0
-I
0
(..)
u

0.4 0.8 1.2

B = (um-u 0 )
q (qT-crvo)

Figure 2-9. Soil Classification Based Figure 2-10. Soil Classification Based
on qT and Bq on CPTU Data

Source: Senneset and Janbu (21), p. 48. Source: Jones and Rust (22), p. 612.

As= Surface area

of cone sleeve
2
(typically 15,000 mm )

Cone sleeve measuring

side resistance

2
a= d2Jo
qr=qc+ubt( 1-a)

Figure 2-11. Unequal End Areas of Electric Friction Cone

2-15
fs -measured cone side resistance, and Asl and As2 are the net internal areas of
the sleeve, as given in Figure 2-11.

Dilatometer Test (DMT) Classifications

The flat dilatometer test (DMT) also is capable of providing an estimate of the
soil type and consistency. The original development of the DMT (Marchetti, 24)
included a classification based on the material index, In, defined as:

Pl - Po
In---- (2-7)
Po - Uo

I
in which p 0 - contact stress, Pl = stress to expand membrane 1 rom into soil, and u 0
= ambient equilibrium pore water stress (often assumed to be hydrostatic, although
not necessarily so). A more recent interpretation is shown in Figure 2-12, which
is based on In and the dilatometer modulus, En, defined as:
I
I
I
0

w
0.
........
0
Notes:
a - Number in parenthesis
I
is normalized unit
weight ( ylyw )
b- If PI>50, (y/yw) is
I
overestimated by
about 0.1
I

0.35 0.6 0.9 1.2 1.8 3.3

VERY SOFT
CLAY/PEAT
5L-~~~~~~uu~~~~~~~~~
0.1 0.2 0.5 10
Material Index, 10

Figure 2-12. Determination of Soil Description and Unit Weight by DMT '

Source: Schmertmann (25), p. 98.

2-16
'
I

I
 Eo= 34.7(pl- p 0 ) (2-8)

This correlation also provides an estimate of the soil unit weight.

UNIT WEIGHT

As previously defined, the soil unit weight (1) is determined as the weight of soil
per unit volume. The relationship between dry (1d) and total (1total) unit weight
is:

1total = (1 + wn)1d (2-9)

in which Wn- natural water content (as a decimal). Table 2-8 presents typical
soil unit weights.

RELATIVE DENSITY OF COHESIONLESS SOILS FROM IN-SITU TEST CORRELATIONS

The standard penetration test (SPT) N value and the CPT cone tip resistance (qc)
have been used extensively to estimate the relative density of cohesionless soils
in-situ. Although they are used commonly in practice, different approaches have
been adopted by different authors. Some of these differences in methodology result
from improvements in the understanding of penetration tests and the relevant fac-
tors affecting the test values. Also, the estimation of the relative density using
the SPT and CPT results is an evolutionary process during which newer and larger
data bases are compiled to allow for more statistically significant trends to be
established. Furthermore, some earlier studies were based on penetration tests
conducted in one type of soil. Testing of more soils of differing geologic ori-
gins, stress histories, and mineralogies allows for refinements and adjustments to
existing correlations.

Standard Penetration Test (SPT) Correlations

Early work on this subject simply correlated the SPT N value directly with relative
density, as shown in Table 2-9. Later laboratory research demonstrated that the
SPT N value also was influenced significantly by the overburden stress. Figure
2-13 shows these results, which were based on calibration chamber tests. For prac-
tical use in estimating Dr from Nand avo• these results were presented in alterna-
tive forms such as that shown in Figure 2-14.

Additional research showed that these relationships are even more complex and
dependent upon other factors, including vertical stress, stress history, and sand

2-17
 Table 2-8

TYPICAL SOIL UNIT WEIGHTS

Approximate Uniformity Normalized Unit Weight

Particle Size (mm) Coefficient Void Ratio Dry, 'Ydryhw Saturated, 'Ysati'Yw
Soil Type Dmax Dmin DlQ D6o/D1Q emax em in Min. Max. Min. Max.

Uniform granular soil

Equal spheres (theoretical) 1.0 0. 92 0.35
Standard Ottawa sand 0.84 0.59 0.67 1.1 0. 80 0.50 1.47 1. 76 1.49 2.10
Clean, uniform sand 1.2 to 2.0 1.00 0.40 1. 33 1. 89 1. 35 2.18
Uniform, inorganic silt 0.05 0.005 0.012 1.2 to 2.0 1.10 0.40 1.28 1. 89 1.30 2.18

Well-graded granular soil

Silty sand 2.0 0.005 0.02 5 to 10 0.90 0.30 1. 39 2.04 1.41 2.28
N
I Clean, fine to coarse sand 2.0 0.05 0.09 4 to 6 0. 95 0.20 1. 36 2.21 1. 38 2.37
..... Micaceous sand 1. 20 0.40 1. 22 1.92 1. 23 2.21
r:tJ
Silty sand and gravel 100 0.005 0.02 15 to 300 0. 85 0.14 1. 43 2.34 1.44 2.48

Silty or sandy clay 2.0 0.001 0.003 10 to 30 1. 80 0. 25 0.96 2.16 1.60 2.36
Gnp-graded silty clay w. gravel or larger 250 0.001 1.00 0.20 1. 35 2.24 1.84 2.42
\Jell-grnded gravel, sand, silt, and clay 250 0.001 0.002 25 to 1000 0. 70 0.13 1. 60 2.37 2.00 2.50

Clay (30 to 50% < 2~ size) 0.05 0. 5~ 0.001 2.40 0. 50 0.80 1. 79 1. 51 2.13
Colloidal clay (over 50% < 2~ size) 0.01 lOA 12.00 0.60 0.21 1. 70 1.14 2.05

Organic silt 3.00 0.55 0.64 1. 76 1. 39 2.10

Organic clay (30 to 50% < 2~ size) 4.40 0. 70 0.48 1.60 1. 30 2.00

Note: -rw- 62.4 lb/ft 3 - 1 gm/cm3 - 0.983 t/m3 - 9. 80 kN/m 3 (at STP conditions) .

Source: Hough (26), pp. 34, 35.

 Table 2-9

RELATIVE DENSITY OF SAND VERSUS N

N Value Relative
(blows/ft or 30S mm) Density Dr (%)

0 to 4 very loose 0 to lS

4 to 10 loose lS to 3S

10 to 30 medium 3S to 6S

30 to so dense 6S to 8S

> so very dense 8S to 100

Source: Terzaghi and Peck (27)' p. 341 and Lambe and Whitman (§_), p. 31.

E 100 SPT N (blows/ft or 305mm}

E
I()
0
r<l
80
.... 0
0 a.
-
.....
........ 60
'-
•t?
0

VI
~
0
o\olpo= If)
If)
-
Q)
..0
.!:::
Q)
40 (f)

.2 0
2
~
z 20 -....
u

Q)
>
1-
0...
(f) 0 3
0 20 40 60 80 100
Relative Density, Dr(%}

Figure 2-13. Effect of Overburden Stress Figure 2-14. Relative Density-N-Stress

and Dr on SPT N Value Relationship

Source: Gibbs and Holtz (28), p. 37. Source: Holtz and Gibbs (29), p. 441.

type (primarily compressibility influences), as a minimum. Figure 2-lS illustrates

some of these complexities. The studies presented in Figure 2-lS led to a corre-
lation for estimating Dr from SPT N values that includes the effect of overburden

2-19
 - E
E
100
-·-Platte river sand (A)
- - Standard concrete sand ( B l
l{) IIIII I I 0 psi,Reid Bedford model sand ( C l and Ottowa sand ( D l
0 :::-:·:::; 40psi,Reid Bedford model sand (C) and Ottowa sand (Ol
r0 \\\.\.' 80psi,Reid Bedford model sand (C)
.... - Gibbs and Holtz ( 1957 l for air-dry coarse (E) and fine sand

-
0
- - Gibbs and Holtz ( 1957) for submerged sands (E)
......
......... - - - Bozoroo (1967)
(f)
3: (I psi =6.89 kNim 2 l
0
:0
Q)
:::J
0
>
z
1-
0...
(f)
0
10 20 30 40 50 60 70 80 90 100

Figure 2-15.

Source:
Relative Density, Dr (%)

Relative Density-N-Stress Relationships for Several Sands

Marcuson and Bieganousky (30), p. 1301.

a
I
l

I
stress (av 0
as given below:
), particle size distribution (CU), and stress history (OCR op;-ov0 ),

I
Dr(%)- 12.2 + 0.75[222N + 2311- 711 OCR- 779(avo/Pa) - 50 CU2]0.5 (2-10)
I
Regression analyses of the data gave r2
equal to 1 or 3.
= 0.77. The data all were unaged with OCR
I
An important factor affecting the SPT N value is the energy efficiency of the drop
hammer onto the drill rods. The theoretical free-fall energy for the SPT is 140 lb
I
(0.623 kN) times 30 in (0.76 m) or 4200 in-lb (0.475 kN-m).
energy ratio (ER) is about 55 to 60 percent in the U.S.A., although this value can
Typically, the average
I
vary from 30 to 90 percent for particular drillers and SPT equipment in practice.

Skempton (31) reviewed SPT calibration data from Japan, China, the U.K., and the
U.S.A. and suggested correction factors based on standard practice in these coun-
tries. Some of the variables affecting the energy efficiency include the type of
hammer, age of the rope, borehole size, and use of liners in the split spoon sam-
pler. For example, the donut hammer is less efficient than the safety hammer, as
shown by the energy ratio examples in Figure 2-16.

2-20
Correcting the hammers to a

'
I
J
 SPT N Value (blows/ft or 305 mm)
0 20 40 60 0 20 40 60

8
34•
55-'--•45
(Measured)

/ .{oDonut
60,......._
/ 40
hammer
20

25
6

8
·-.
•' (Corrected)

J Hammer
':{corrected
20

25
56""-....•41 ' to ER=55
E ••
-
..c
Cl.
Q)
10
63•/
63•/
--•41
'•39
30

35
10 •.
~
' percent 30

35
---
0 /_,rSafety
\ '.
12
'-
14 Energy
63•~ammer

64•--•56
ratio (ERl-- 9
_.....-•47

..(s
I
40 12

45 14 --.
./

•
\
40

45

Figure 2-16. Donut and Safety Hammer Comparisons

Source: Robertson, et al. (32), p. 1454.

constant energy ratio eliminates the differences. The energy efficiency also
depends upon the size of cathead and number of turns of the rope, as indicated in
Figure 2-17. Standard U.S. practice is two turns of rope on a large cathead.

The SPT N value, corrected for field procedures, is given below:

(2-11)

in which N6o - N value corrected for field procedures to an average energy ratio of
60 percent, N =measured SPT N value, and CER• CB, Cs, and CR are correction fac-
tors for energy ratio, borehole diameter, sampling method, and rod length, respec-
tively, as given in Table 2-10.

Since the SPT N value also varies with stress level, overburden stress correction
factors are used to provide a consistent point of reference. This correction takes
the form:

(2-12)

in which (Nl)60 - N60 value corrected to a reference stress of one atmosphere and
CN = correction factor for overburden stress.

2-21
 I
1.0
[ ~-~pe th<O•n off ootheod
1.0
0
I
2
0.8
lMonuol ~0
----- _,-small
-- ootheod
1.2 "
a::
w
z...... I
•eleo•e o ~0 z
.2
0
a:
0.6 Trip hammer
~
o
1.6
-
.2
0
I
>- 0
....
C'
Lorge cotheod o a:
Q)
c::: 0.4r- 2.5
w

0.2
I
0~--~'------~------~------~----~
0 2 3
I
Nominal Number of Turns of Rope on Cotheod

Figure 2-17. Energy Ratio Variations

Source: Modified after Skempton (31), p. 427.

Table 2-10

SPT CORRECTION FACTORS FOR FIELD PROCEDURES

I
Factor Equipment Variables
Correction
Term Value tl
Energy ratio Safety hammer
Donut hammer
CER 0.9
0.75
:.
Borehole diameter 65 to 115 mm (2.5 to 4.5 in)
150 mm (6 in)
200 mm (8 in)
1.0
1.05
1.15
,I
Sampling method Standard sampler
Sampler without liner
1.0
1.2 I
> 10 m (> 30 ft) 1.0
I
Rod length
6 to 10 m (20 to 30 ft) 0.95
4 to 6 m (13 to 20 ft) 0.85
3 to 4 m (10 to 13 ft) 0.75

Source: Based on Skempton (31). I

I
2-22 I;I
Perhaps the simplest expression for GN is given below (Liao and Whitman, 33):

A comparison of different CN recommendations is given in Figure 2-18. Basically,

all methods give similar corrections for avo > 0.5 Pa within the range of expected
accuracy for the SPT. The correction factors proposed by Skempton are based large-
ly on laboratory test data, while the others have been derived from field data.

Although Equation 2-13 is simple, high values of CN develop at very low values of
Ovo· Alternatively, Skempton (31) suggested the following for fine sands:

(2-14)

This equation gives a maximum CN of 2 at the ground surface. Figure 2-19 shows
that both equations are adequate for avo > 0.5 Pa and also appear applicable for
use in overconsolidated sands.

Once the SPT N value has been corrected for field procedures and overburden effects
to give (N1)60• it can be used to evaluate the relative density as a function of
the soil characteristics. Figure 2-20 shows (Nl)6o/Dr2 as a function of the soil
particle size (D5o). The laboratory data in this figure were obtained from studies

SPT Overburden Correction Factor, CN

2.0 2.5

0.5

1.0
········· Peck, et al., 1974 (34)
IJ) - - - Liaa and Whitman,
IJ) 1986 (33)
....
Q) 1.5 - - Skempton,l986 (31):
(i) Coarse sands -
- - Skempton, 1986!31):
Fine sands -

-
0
u 2.0
- · - Skemptan, 1986 (31):
.... O.C. sands -
~ - .. -Seed, 1979 (35):
Dr=50% -
2.5
------Seed, 1979 (35):
Dr=70% -

Figure 2-18. Comparison of SPT Overburden Corrections

2-23
 I
SPT Overburden Correction Factor, CN I
-.
0 0.5 1.0 1.5 2.0 2.5
o~-----r------T-----~-----~~----_,

~ I
0
0.
........
lb>
0
2
I
vi
Vl
Q)
;:
(j)
I
I
0

-u
.....
~
Model Sand
I
Figure 2-19. Comparison of Recommended CN Factors and Available Data from OC Sands
I
I
Niigata 7 _P!:_C~ and _§aza..!:_a~ _ 7
'il
80
(NC, aged)\ 'il
· (field, aged, coarse) ·
I
N
...
_.. _..
~
--
'il-

0
........
60 Laboratory
(unaged) ll-"~•-
0
6 I
{~0
0

-z <D (NI )60

---= 60 + 25 log o50
40
0 D 2
r
I
WES data
20
0
1::!., A
0, •
Skemptan interpretation
Re-evaluation using Skempton method for overburden effect
Re-evaluation using Liao 8o Whitman method for overburden effect
I
0.5
0 50 (mm)
2 5
I
Figure 2-20. Particle Size Effect on Blow Count for Sands
I
at the Waterways Experiment Station (WES) on three sands (30, 36, 37).
data were for unaged, normally consolidated (NC) sands (OCR= 1), although a small
Most of the
I
series of tests was conducted on overconsolidated sands with OCR= 3.
interpretation (31) of these data is shown, but it is believed that the averaged
Skempton's
I
2-24 I
curves and smoothed data he used led to an underestimation of (Nl)6o/Dr2· Re-eval-
uation of the original data (36, 37) leads to the higher values shown, using either
Skempton's linearized overburden effect (31) or Liao and Whitman's (33) nonlinear
overburden effect. These results can be approximated as follows:

which is applicable for NC, unaged sands. The OC data give higher values than
Equation 2-15, and aged sands also give higher values. The data from Niigata,
Japan were tabulated by Skempton (31), but they were re-evaluated individually.
The Peck and Bazaraa (38) curve represents coarse sands (no exact particle size
given) from field test evaluations. These data represent aged sands that likely
were overconsolidated.

Figure 2-21 illustrates the data as a function of age of the deposits. The WES
laboratory data are plotted at an age of several days. The Niigata, Ogishima, and
Kawagishi data summarized by Skempton (31) represent NC recent fills that were
assigned approximate ages of 30 to 40 years. The time is not known for the OC,
aged, Peck and Bazaraa data, so it is estimated at 100 to 10,000 years. The other
four sites (A, B, C, D) are given by Barton, et al. (39). They represent OC, aged,
fine and fine to medium sands of four geologic periods, as noted.

zoo~~---.--.---.-~---.--.---.-~---.--,H~o--r--.
A, 8, C, 0 ore OC, aged, Grantham
fine and fine to medium sands (Jurassic)

c
.......
Folkestone
(Cretaceous)
A
N .._
Norwich ~
0 (Lower Pleistocene) Bogshot

__ --
........
0 Peck and Bozoroo- (Eocene)
tD coarse d. field, aged, coarse - - -
(Dso=4.75mm) me •um?--------? ----
z

I
-INiigOto:;;c::g; lging Effect
-Lb t
o oro ory- fine
CA=I.2+0.051og(t/100)
NC, unoged fine• Ogishimo-NC, aged
• Kowogishi-NC, aged
fine fine
(0 50 =0.074mml

OL-~~--~~--~~--~~--~~--~~~
2 4 6 8 10
I0- 2 10 10 10 10 10

Age or Time Since Deposition, t (years)

Figure 2-21. Aging Effect on Blow Count for Sands

2-25
Based on Figures 2-20 and 2-21, it is clear that particle size, aging, and overcon-
solidation significantly influence the (Nl)6o/Dr2 ratio. These effects can be
quantified as follows:

(2-16)

in which Cp, CA, and CoeR are the correction factors given in Table 2-11. Cp is
based on Figure 2-20~ CA is based on a conservative interpretation of the impre-
cise data in Figure 2-21. CoeR is based on direct evaluation of the WES data and
interpretation of the Niigata data. It also is consistent with the studies pre-
sented by Tokimatsu (40).

Finally, the complete expression for relative density (Dr) in terms of SPT N value,
I
including all corrections and modifying terms, is:

CER CB Cs CR CN N
I
Dr 2
Cp CA CoeR
(with Dr in decimal form) (2-17)
I
in which N =measured N value and the corrections are as follows: energy ratio
(CER), borehole diameter (CB), sampling method (Cs), rod length (CR), overburden
stress (CN), particle size (Cp), aging (CA), and overconsolidation (CoeR).
I
Cone Penetration Test (CPT) Correlations
I
Early work on this subject was similar to the SPT, and therefore the CPT qc value
I
Table 2-11

SPT CORRECTION FACTORS FOR SAND VARIABLES

I
I
Correction
Effect Parameter Term Value
I
Particle size D5o of sand Cp 60 + 25 log D5o
(D5o in mm) I
Aging Time (t) CA 1.2 + 0.05 log (t/100)

Overconsolidation CoeR OCRO.l8 I

2-26 I
simply was correlated directly to relative density, as shown in Table 2-12. As
with theN values, recent research has shown that the relationships are more com-
plex. Figure 2-22 shows the generalized relationship for Ticino sand, which is of
medium compressibility. The vertical effective stress can be used with this figure
if the sand is unaged and normally consolidated. The horizontal effective stress
should be used if the sand is aged or overconsolidated.

Figure 2-23 illustrates that the generalized CPT correlations vary for soils of
different compressibilities. Curve 3 corresponds to data on Monterey sand, which
is of low compressibility. Monterey sand is characterized by subrounded to sub-
angular grains, which are composed mainly of quartz and some feldspar, with zero
percent fines. Curve 2 is for Ticino sand, a granular soil of moderate compressi-
bility with subangular grains composed of quartz and 5 percent mica, with less than
1 percent fines. Curve 3 is for the high compressibility Hilton Mines sand, con-
sisting of angular iron mine tailings of quartz, feldspar, and mica composition,
with 3 percent fines.

To compare cone tip resistances obtained at different depths, it is necessary to

reference the values to a standardized reference stress level, usually taken as
Bvo/Pa = 1 atmosphere. The standardized cone tip resistance (~) then becomes:

Table 2-12

RELATIVE DENSITY OF SAND VERSUS qc

Cone Tip Relative

Resistance, qc/Pa Density Dr (%)

< 20 Very loose < 20

20 to 40 Loose 20 to 40

40 to 120 Medium 40 to 60

120 to 200 Dense 60 to 80

> 200 Very dense > 80

Source: Meyerhof (]_)' p. 17.

2-27
 Cone Tip Resistance, qc/p0
0 400
0 I{)
0
¢
0
......a. 0
0 "
·~
~
0.5 .....
ch .E
"'
Q) 0
;:
(/) ......a.0
2
Q)
.?! 1.0 •b"
u
-
w
0
Q)
.,vi
....
Q)

<i)
Q)
3
c0 1.5 .?!
N
·;::
u
~
0
:X: w 4
0
2.0 ~
03
> Dr(%)=40 50 60 70 80 90
5
Figure 2-22. Relative Density from CPT for Uncemented and Unaged Quartz Sands

Source: Robertson and Campanella (17), p. 723.

Cone Tip Resistance, qc /p0

0 100 200 300 400 500

0
Hilton mines sand
0
0..
...... Ticino sand
0
Monterey sand
•t?
ch
"'

-~
~
(j) 2
Compressibility
Q)
.?!
u low
I
2 3
w
0
:E
~. 4

Figure 2-23. Relative Density from CPT, Including Soil Compressibility

Source: Robertson and Campanella (17), p. 722.

in which qc -measured cone tip resistance, and Cq = overburden stress correction

factor. For all practical purposes, Cq is nearly identical to CN, proposed for the
SPT and given as:

Z-28
I
 (2-19)

Much research on the CPT has been conducted in calibration chambers, which are
described briefly in Appendix H. These studies allow the use of controlled sand
properties and in-situ stresses, which is not possible in the field. One summary
of Dr data from calibration chamber tests on five different normally consolidated
sands is shown in Figure 2-24. This figure illustrates the range in actual data
taken under controlled laboratory conditions after uniform soil placement. The
generalized figures shown earlier in this section do not show the data range and
perhaps suggest a high confidence level. Figure 2-24 shows what the actual ranges
are for only five sands under controlled laboratory conditions; field cases are
likely to exhibit more variability.

Calibration chamber data are useful, but the tests are performed with flexible
walls of limited dimensions. Therefore, the boundary effects result in lower qc
values than obtained for "field conditions", corresponding to an infinite half-
space. To correlate the field and chamber qc values, Jamiolkowski, et al. (19)
recommended dividing the field value of qc by Kq, as given below:

0~ 80
,_
0

(/)
c:
Q)
0
Q)
Sand
-
>
0
Q)
• Ticino
~:.
Ottawa
0::
o Edgar
• Hokksund
v Hilton mines
20
10 50 100 500 1000

Cone Tip Resistance,

Figure 2-24. Correlation Between Dr and Dimensionless qc (Uncorrected for Boundary

Effects)

Source: Jamiolkowski, et al. (19), p. 120.

2-29
 Kq = 1 + (Dr - 30)/300 (2-20)

before entering Figure 2-24. The equivalent chamber values then can be used to
evaluate Dr. This process requires iteration, because the value of Dr is not known.

As an alternative approach, the results of 24 sets of calibration chamber tests on

sands were compiled, in which the values of qc were corrected for the effects of
boundary conditions. These sands were predominantly fine and medium sands.
mary of this compilation is given in Appendix H.
A sum-
For the majority of chambers with
I
flexible walls, the boundary correction required an increase in qc to reflect
"field" values. I
The results of this study are given in Figures 2-25 and 2-26 for unaged, uncemented
sands. In all cases, a linear relationship was obtained for the square of the rel-
I
ative density (Dr2) versus the dimensionless cone tip resistance, given as shown by
QeD· Figure 2-25 shows the normally consolidated sands, separated into low, I
medium, and high compressibility. Low compressibility (Figure 2-25a) generally
corresponds to quartz sands with little, if any, fines. Medium compressibility
(Figure 2-25b) suggests quartz with some feldspar, with perhaps several percent
I
fines. High compressibility (Figure 2-25c) indicates more fines, mica, and other
compressible minerals. Most natural sands likely will be more toward the medium to I
high range of compressibility. As shown in these figures, the correlation is good
below the limit of possible particle crushing. This limit was established by sta-
tistical analysis of the data, optimizing the r2 value as a function of different
I
limiting QeD values from 250 up to the entire data set. The limiting QeD values
shown provide the maximum r2 for the data and define the boundary of possible par- I
ticle crushing. Data points beyond the limiting QeD values are not included in the
statistics. I
Figure 2-26 shows comparable calibration chamber data on overconsolidated sands,
separated into low(< 3), medium (3 to 8), and high (8 to 15) OCR ranges. These I
data also were optimized using r2 for different QeD limiting values, resulting in
the regression lines and possible particle crushing limits shown. I
A summary of these relationships is given in Figure 2-27 for all of the corrected
calibration chamber data. This figure clearly shows the influence of compressibil- I
ity and OCR on the relationship between Dr2 and the dimensionless cone tip resis-
tance. These relationships can be quantified approximately as follows:
I[I
~
I
2-30

l I
I
I
I
I
I
0.4

I
0.2 (a) Low

I Compressibility
NC Sands

I 600

I
I 1.0 .------..--1--~,.----.L,.,--'--.,--1--,.--,---,

I 0.8
- }Po~
0
:

- ~ 0
oru
1'1.. r-- Possible particle -
I 13'_ crushmg

0~~~
0.6 - -

~ 2 Oco Oco -
I 0.4: ~~'o
Dr = OF = 292

(n=145, r 2 =0.885,
-
S.D. =0.10)
-
I 0.2~ ® (b) Medium -
~~
Compressibility
-
I 1/00 1 I
0~---J----~----~----~----~--~
0 100 200 300
I
NC Sands
I
400 500
I
600

I
I Figure 2-25. Calibration Chamber Data on NC Sands

I 2-31

I
 r.o ,...----r---<$>-m,...,;..----.--o
I o0
ED ED O
o I oO 8
s5?)g:>~ 8 o og o
ED ~ 9()00 8
ED @BIB 0
I--

l
Possible particle crushing

ED z=Oco =Oco
0
0 O 'OF 280
em 00 (n=s:. r 2=0.769,
S.0.-0.14)

/
EBB (c) High

I
0
Compressibility
NC Sands

100 200 600

I
Figure 2-25. Calibration Chamber Data on NC Sands (continued)
I
I

I
'
I
II
'tl•
~
Figure 2-26. Calibration Chamber Data on OC Sands

2-32 I
 Cose OF n r2 S.D.
NC-high comp. 280 59 0.796 0.14
NC-med. comp. 292 145 0.885 0.10
NC-Iow. comp. 332 190 0.711 0.14
NC-overoge 305 0.776 0.13
Low OCR (<3l 390 0.711 0.14
Med. OCR (3-8) 403 0.849 0.10
0.859 0.12

Figure 2-27. Summary of Calibration Chamber Studies

Qco
Dr 2 (2-21a)
305 Qc QocR

Cq(qc/Pa)
Dr 2 (2-2lb)
305 Qc QocR

l (qc/Pa)
Dr 2 (2-2lc)
305 Qc OCR0.18 (ovo/Pa )0. 5

in which Qc -compressibility factor (0.91 for high, 1.0 for medium, and 1.09 for
low) and QocR =overconsolidation factor (- OCR0.18), comparable to CoeR for the
standard penetration test. The QocR factor was evaluated using the mean OCR values
for the low, medium, and high OCR data equal to 2.3, 5.1, and 10.1, respectively.
The majority of natural sands are likely to be of medium to high compressibility
and low to medium OCR.

It should be noted that Equation 2-17 for the SPT is similar in form to Equation
2-21 for the CPT, although some differences are evident. Perhaps the most

2-33
 important difference is that the SPT relationship includes aging, while the CPT
relationship is only for unaged sands. If the same f~nctional relationship for
aging holds for both the SPT and CPT, then CA (as given in Table 2-11) would be
introduced into the denominator of Equation 2-21. This addition is speculation at
this time. However, the CA changes qualitatively explain the effects of aging in a
reasonable manner.

Dilatometer Test (DMT) Correlations

The DMT is a relatively new test for which broad correlations have not yet been
developed for relative density (Dr). However, it has been used to estimate Dr in
normally consolidated, uncemented sands. This correlation is shown in Figure 2-28
for Dr as a function of the DMT horizontal stress index (KD), described in Appendix
D and defined as:

(2-22)

in which p 0 - initial contact stress, u 0 -hydrostatic stress, and avo - effective

vertical stress. This correlation is based on few data and should be considered
preliminary at this time.

10

0
f-
••
!::;. 0
Chamber tests
Field sites
-
:::.:::: 8-
><
Q)
- K0 ::::0.40
"0
c

I
...... 6

•I
(/)
(/)
Q)
'-
(/)
4 l
- 0
c
0 /
N
/.
-·--
'-
0 2
I ._!::;. 0

0 I I I I
0 20 40 60 80 100

Relative Density, Dr (%)

Figure 2-28. Correlation Between DMT Horizontal Stress Index and Relative Density
for Normally Consolidated, Uncemented Sand

Source: Robertson and Campanella (41), p. 39.

2-34
CONSISTENCY OF COHESIVE SOILS FROM IN-SITU TEST CORRELATIONS

The standard penetration test (SPT) N value and the cone penetration test (CPT) qc
value also have been used to estimate the consistency of cohesive soils in-situ.
However, little published work has been presented on these correlations, and there-
fore all should be considered approximate at best.

Standard Penetration Test (SPT) Correlations

The consistency of cohesive soils has been correlated with the N value, as shown in
Table 2-13. In general, these values are to be considered only approximate guide-
lines, since clay sensitivity can greatly affect theN value (Schmertmann, 42).

Although the correlations with N value in clay commonly are considered to be less
reliable than those in sand, increasing N values do, in general, reflect increasing
stiffness and therefore decreasing liquidity index. To express this general corre-
lation, the consistency index (CI) has been defined as follows:

CI 1 - LI (2-23)

which effectively is a mirror image of the liquidity index. Table 2-14 is

Table 2-13

CONSISTENCY OF CLAY VERSUS N

N Value
(blows/ft or 305 mm) Consistency

0 to 2 Very soft

2 to 4 Soft

4 to 8 Medium

8 to 15 Stiff

15 to 30 Very stiff

> 30 Hard

Source: Terzaghi and Peck (27), p. 347.

2-35
 I
Table 2-14 I
CONSISTENCY INDEX OF CLAY VERSUS N and qc
I
N Value
(blows/ft or 305 mm)
Cone Tip
Resistance, qc/Pa Consistency Consistency Index I
< 2 < 5 Very soft < 0.5 I
2 to 8 5 to 15 Soft to medium 0.5 to 0.75

8 to 15 15 to 30 Stiff 0.75 to 1.0 I

15. to 30 30 to 60 Very stiff 1.0 to 1.5

> 30 > 60 Hard > 1.5 I

Source: Szechy and Varga (43), p. 105.
'I
representative of the CI correlations.

Cone Penetration Test (CPT) Correlations

The consistency of cohesive soils also has been related to the cone tip resis-
tance. Again, as with the N values, the correlations in clay are less reliable.
typical correlation is given also in Table 2-14.
A

II
RELATIONSHIP BETWEEN SPT N AND CPT qc VALUES

Because of the numerous relationships developed for either SPT or CPT data, it is
II
~
advantageous to have a procedure to interrelate N and qc. Both are penetration
resistances (although the SPT is dynamic and the CPT is quasi-static), and they are
ll
the most common forms of in-situ testing used worldwide today.

A number of investigators have proposed single numerical values of qc/N. However,

r•
recent studies have shown that qc/N generally correlates with grain size, as shown
in Figure 2-29. Unfortunately, most of these data do not include N or qc value
I
corrections as noted previously.

Newer data (44 - 50) have been combined with the previous results in Figure 2-29 to
I
result in Figure 2-30. This new relationship confirms the general trend of the
data, and it extends the relationship to mean grain sizes up to 10 mm. The new
I
2-36
I z
8
Data from 18 sites

I .......

0..
0
6

..... •••
.
• •/' •
.....
...-• .. .
.......
u
0"
4 ••••••••
I 2
•~·
••
~·

I 0
0.001 0.005 0.01 0.05 0.1

Mean Particle Size, 0 50 (mm)

0.5

I Figure 2-29. Variation of qc/N with Grain Size for Electric and Mechanical
Friction Cones

I Source: Robertson and Campanella (17), p. 730.

I 0 Robertson and Campanella, 1983 (Lil

... ...
10
0 Zervogiannis and Kalteziotis, 1988 ( 4 5)
X Chin, et al., 1988 ( 46) ...
I 0 Jamiolkowski, et al., 1985 (44)
...
Andrus and Youd, 1987 (50)
0

•Kasim, et al., 1986 (47)

••
8

I z
.......

0..
0
6
Seed and deAiba, 1986 (48)
Muromachi, 1981 (49)

0
.......
u
0"
4
0
00
2 (qc/Pol/N=5.44 Dso0.26
(n=l97, r 2 =0.702, 5.0.=1.03)

0.005 0.01 0.05 0.1 0.5 5 10

Mean Particle Size, 0 50 (mm)

Figure 2-30. Recommended Variation of qc/N with Grain Size for Fugro Electric
Friction Cones

recommended relationship is given by the solid line and regression equation on the
figure.

2-37
In .other studies, the ratio of qc/N has been correlated to the percentage of fines
]
(clay and silt sizes). For example, Jamiolkowski, et al. (44) indicate the trend
presented in Figure 2-31 for Italian soils.
available data were summarized (46, 47, 49) to substantiate a general trend between
In addition to these data, other
I
the qc/N ratio and fines content, as shown in Figure 2-32. Use of Figures 2-30 and
2-32 will provide the best estimate relationship between qc and N, with the ratio J
decreasing with increasing fines content.

I
z
6

••
149
577
244 119
• • 46 40 28
f No. experimental
points
I
4
•• •
- .......

Cl.
.......
0
23
• 69
• 139 I
<..>
0'
2 •

o~~--~--~--~--~--~~~~---L~
0 20 40 60 80 I 00
I
Figure 2-31.
Percent Passing No. 200 Sieve
Variation of qc/N with Fl.nes Content
I
Source: Jamiolkowski, et al. (44), p. 1895.
I
IOr----r----T---~----~---.----.----,----,----,----,

f- + Jomiolkowski, et ol., 1985 (44)

I
6 Kosim, et ol., 1986 (47)
8f- 0
0
Muromochi, 1981 ( 49)
Chin, et ol., 1988 (46)
-
I
f-&
z
.......
0
6
60 c&
-
I
-
Cl.
-~~ 6.D Ef
I
....... 0 0
4 =c;-o6~-+-6~ ....
<..>
0' -
%?
cP -.~~ .(}) 0
~ rn. ooo~oo_:>+o o o «;
2~ ~ t...r
o _g fS)
0
o
0 Q:]
1:rt1 --e-0 Q.• CfJ Ji
016 I
(qc/p0 ) F 2 0
~-

0
- N - = 4.25-

20 40 60
1.
4 3
0~---~•---~~---~•---~•--~•----~'----~~----~~----~~----~
80 I 00
(n=I08, r =0.414, 5.0.=0.89)
I
Figure 2-32.
Percent Passing No. 200 Sieve, F
Recommended Variation of qc/N with Fines Content
II I
!

2-38 I
l I
 REFERENCES

1. American Society for Testing and Materials, "Standard Method for Particle-Size
Analysis of Soils [D422-63(1972)]", Annual Book of Standards, Vol. 4.08, ASTM,
Philadelphia, 1989, pp. 86-92.

2. American Society for Testing and Materials, "Standard Practice for Wet Prepa-
ration of Soil Samples for Particle-Size Analysis and Determination of Soil
Constants (D2217-85)", Annual Book of Standards, Vol. 4.08, ASTM, Philadel-
phia, 1989, pp. 270-272.

3. American Society for Testing and Materials, "Standard Test Method for Liquid
Limit, Plastic Limit, and Plasticity Index of Soils (D4318-84)", Annual Book
of Standards, Vol. 4.08, ASTM, Philadelphia, 1989, pp. 579-589.

4. American Society for Testing and Materials, "Standard Test Methods for Maximum
Index Density of Soils Using a Vibratory Table (D4253-83)", Annual Book of
Standards, Vol. 4.08, ASTM, Philadelphia, 1989, pp. 560-571.

5. American Society for Testing and Materials, "Standard Test Methods for Minimum
Index Density of Soils and Calculation of Relative Density (D4254-83)", Annual
Book of Standards, Vol. 4.08, ASTM, Philadelphia, 1989, pp. 572-578.

6. Lambe, T. W. and Whitman, R. V., Soil Mechanics, John Wiley and Sons, New
York, 1969, 553 p.

7. Meyerhof, G. G., "Penetration Tests and Bearing Capacity of Gohesionless

Soils", Journal of the Soil Mechanics and Foundations Division, ASGE, Vol. 82,
No. SMl, Jan. 1956, pp. 1-19.

8. Youd, T. L., "Factors Controlling Maximum and Minimum Densities of Sands",

Evaluation of Relative Density and Its Role in Geotechnical Projects Involving
Gohesionless Soils (STP 523), ASTM, Philadelphia, 1973, pp. 98-122.

9. Sowers, G. F., Introductory Soil Mechanics and Foundations: Geotechnical Engi-

neering, 4th Ed., Macmillan Publishing Go., New York, 1979, 621 p.

10. American Society for Testing and Materials, "Standard Test Method for Classi-
fication of Soils for Engineering Purposes (D2487-85)", Annual Book of
Standards, Vol. 4.08, ASTM, Philadelphia, 1989, pp. 288-297.

11. American Society for Testing and Materials, "Standard Practice for Description
and Identification of Soils (Visual-Manual Procedure) (02488-84)", Annual Book
of Standards, Vol. 4.08, ASTM, Philadelphia, 1989, pp. 298-307.

12. Burmister, D. M., "Physical, Stress-Strain, and Strength Responses of Granular

Soils", Symposium on Field Testing of Soils (STP 322), ASTM, Philadelphia,
1962, pp. 67-97.

13. Burmister, D. M., "Suggested Methods of Test for Identification of Soils",

Special Procedures for Testing Soil and Rock for Engineering Purposes (STP
479), ASTM, Philadelphia, 1970, pp. 311-332.

14. Laboratorium voor Grondmechanica, "Cone Penetration Testing", Giviele and

Bouwkundige Techniek, No. 3, May 1982, pp. 16-36.

2-39
15. Schmertmann, J. H., "Guidelines for Cone Penetration Test Performance and
Design", Report FHWA-TS-78-209, U.S. Department of Transportation, Washington,
1978, 145 p.

16. Douglas, B. J. and Olsen, R. S., "Soil Classification Using Electric Cone
Penetrometer", Cone Penetration Testing and Experience, Eds. G. M. Norris and
R. D. Holtz, ASCE, New York, 1981, pp. 209-227.

17. Robertson, P. K. and Campanella, R. G., "Interpretation of Cone Penetration

Tests. Part I: Sand", Canadian Geotechnical Journal, Vol. 20, No. 4, Nov.
1983, pp. 718-733.

18. Wroth, C. P., "Penetration Testing - A More Rigorous Approach to Interpreta-

tion", Proceedings, 1st International Symposium on Penetration Testing
(ISOPT-1), Vol. 1, Orlando, 1988, pp. 303-311.

19. Jamiolkowski, M., Ladd, C. C., Germaine, J. T., and Lancellotta, R., "New
Developments in Field and Laboratory Testing of Soils", Proceedings, 11th
International Conference on Soil Mechanics and Foundation Engineering, Vol. 1,
San Francisco, 1985, pp. 57-153.

20. Olsen, R. S. and Farr, J. V., "Site Characterization Using the Cone Penetro-
meter Test", Use of In-Situ Tests in Geotechnical Engineering (GSP No. 6), Ed.
S. P. Clemence, ASCE, New York, 1986, pp. 854-868.

21. Senneset, K. and Janbu, N., "Shear Strength Parameters Obtained from Static
CPT", Strength Testing of Marine Sediments (STP 883), Eds. R. C. Chaney and K.
R. Demars, ASTM, Philadelphia, 1985, pp. 41-54.

22. Jones, G. and Rust, E., "Piezometer Penetration Testing", Proceedings, 2nd
European Symposium on Penetration Testing, Vol. 2, Amsterdam, 1982, pp. 607-
613.

23. Robertson, P. K., Campanella, R. G., Gillespie, D., and Grieg, J., "Use of
Piezometer Cone Data", Use of In-Situ Tests in Geotechnical Engineering (GSP
No. 6, Ed. S. P. Clemence, ASCE, New York, 1986, pp. 1263-1280.

24. Marchetti, S., "In-Situ Tests by Flat Dilatometer", Journal of the Geotechni-
cal Engineering Division, ASCE, Vol. 106, No. GT3, Mar. 1980, pp. 299-321.

25. Schmertmann, J., "Suggested Method for Performing the Flat Dilatometer Test",
Geotechnical Testing Journal, ASTM, Vol. 9, No. 2, June 1986, pp. 93-101.

26. Hough, B. K., Basic Soils Engineering, 2nd Ed., Ronald Press, New York, 1969,
634 p.

27. Terzaghi, K. and Peck, R. B., Soil Mechanics in Engineering Practice, 2nd Ed.,
John Wiley and Sons, New York, 1967, 729 p.

28. Gibbs, H. J. and Holtz, W. G., "Research on Determining the Density of Sands
by Spoon Penetration Testing", Proceedings, 4th International Conference on
Soil Mechanics and Foundation Engineering, Vol. 1, Lqndon, 1957, pp. 35-39.

29. Holtz, W. G. and Gibbs, H. J., Discussion of "SPT and Relative Density in
Coarse Sand", Journal of the Geotechnical Engineering Division, ASCE, Vol.
105, No. GT3, Mar. 1979, pp. 439-441.

2-40

I
I
30. Marcuson, W. F., III and Bieganousky, W. A., "SPT and Relative Density in
Coarse Sands", Journal of the Geotechnical Engineering Division, ASCE, Vol.
103, No. GTll, Nov. 1977, pp. 1295-1309.

31. Skempton, A. W., "Standard Penetration Test Procedures and the Effects in
Sands of Overburden Pressure, Relative Density, Particle Size, Aging, and
Overconsolidation", Geotechnigue, Vol. 36, No. 3, Sept. 1986, pp. 425-447.

32. Robertson, P. K., Campanella, R. G., and Wightman, A., "SPT-CPT Correlations",
Journal of the Geotechnical Engineering Division, ASCE, Vol. 109, No. 11, Nov.
1983, pp. 1449-1459.

33. Liao, S. S. and Whitman, R. V., "Overburden Correction Factors for SPT in
Sand", Journal of Geotechnical Engineering, ASCE, Vol. 112, No. 3, Mar. 1986,
pp. 373-377.

34. Peck, R. B., Hansen, W. E., and Thornburn, T. H., Foundation Engineering, 2nd
Ed., John Wiley and Sons, New York, 1974, 514 p.

35. Seed, H. B., "Soil Liquefaction and Cyclic Mobility Evaluation for Level
Ground During Earthquakes", Journal of the Geotechnical Engineering Division,
ASCE, Vol. 105, No. 2, Feb. 1979, pp. 201-255.

36. Bieganousky, W. A. and Marcuson, W. F., III, "Liquefaction Potential of Dams

and Foundations: Laboratory Standard Penetration Tests on Reid Bedford Model
and Ottawa Sands", Research Report S-76-2 (Report 1), U.S. Army Engineer
Waterways Experiment Station, Vicksburg, 1976, 156 p.

37. Bieganousky, W. A. and Marcuson, W. F., III, "Liquefaction Potential of Dams

and Foundations: Laboratory Standard Penetration Tests on Platte River Sand
and Standard Concrete Sand", Research Report S-76-2 (Report 2), U.S. Army
Engineer Waterways Experiment Station, Vicksburg, 1977, 87 p.

38. Peck, R. B. and Bazaraa, A. R. S., Discussion of "Settlement of Spread Foot-

ings on Sand", Journal of the Soil Mechanics and Foundations Division, ASCE,
Vol. 95, No. SM3, May 1969, pp. 905-909.

39. Barton, M. E., Cooper, M. R., and Palmer, S. N., "Diagenetic Alteration and
Micro-Structural Characteristics of Sands", Penetration Testing in the U.K.,
Thomas Telford, London, 1988, pp. 57-60.

40. Tokimatsu, K., "Penetration Tests for Dynamic Problems", Proceedings, 1st
International Symposium on Penetration Testing (ISOPT-1), Vol. 1, Orlando,
1988, pp. 117-136.

41. Robertson, P. K. and Campanella, R. G., "Estimating Liquefaction Potential of

Sands Using the Flat Plate Dilatometer", Geotechnical Testing Journal, ASTM,
Vol. 9, No. 1, Mar. 1986, pp. 38-40.

42. Schmertmann, J. H., "Measurement of In-Situ Shear Strength", Proceed:i,ngs, ASCE

Conference on In-Situ Measurement of Soil Properties, Vol. 2, Raleigh, 1975,
pp. 57-138. (closure: pp. 175-179).

43. Szechy, K. and Varga, L., Foundation Engineering - Soil Exploration and Spread
Foundations, Akademiai Kiado, Budapest, 1978, 508 p.

2-41
44. Jamiolkowski, M., Baldi, G., Bellotti, R., Ghionna, V., and Pasqualini, E.,
"Penetration Resistance and Liquefaction of Sands", Proceedings, 11th Inter-
national Conference on Soil Mechanics and Foundation Engineering, Vol. 4, San
Francisco, 1985, pp. 1891-1896.

45. Zervogiannis, C. S. and Kalteziotis, N. A., "Experiences and Relationships

from Penetration Testing in Greece", Proceedings, 1st International Symposium
on Penetration Testing (ISOPT-1), Vol. 2, Orlando, 1988, pp. 1063-1071.

46. Chin, C. T., Duann, S. W., and Kao, T. C., "SPT-CPT Correlations for Granular
Soils", Proceedings, 1st International Symposium on Penetration Testing
(ISOPT-1), Vol. 1, Orlando, 1988, pp. 335-339.

47. Kasim, A. G., Chu, M. Y., and Jensen, C. N., "Field Correlation of Cone and
Standard Penetration Tests", Journal of Geotechnical Engineering, ASCE, Vol.
112, No. 3, Mar. 1986, pp. 368-372.

48. Seed, H. B. and de Alba, P., "Use of SPT and CPT Tests for Evaluating the
Liquefaction Resistance of Sands", Use of In-Situ Tests in Geotechnical Engi-
neering (GSP 6), Ed. S. P. Clemence, ASCE, New York, 1986, pp. 281-302.

49. Muromachi, T., "Cone Penetration Testing in Japan", Cone Penetration Testing
and Experience, Eds. G. M. Norris and R. D. Holtz, ASCE, New York, 1981, pp.
49-75.

50. Andrus, R. D. and Youd, T. L., "Subsurface Investigation of a Liquefaction-

Induced Lateral Spread, Thousand Springs Valley, Idaho", Miscellaneous Paper
GL-87-8, U.S. Army Engineer Waterways Experiment Station, Vicksburg, 1987, 131
p.

I
I
I
'

I
tj
I~
2-42
I
'I
 Section 3

IN-SITU STRESS STATE

In most geotechnical engineering problems, a knowledge of the in-situ state of

stress is necessary for two reasons. First, these stresses represent the original
conditions onto which any engineered construction imposes stress increments. These
initial through final stress conditions are used to evaluate the overall engineer-
ing performance of the constructed facility. Second, nearly all engineering prop-
erties of soil are a function of the soil stresses, either directly or indirectly.
Therefore, the stresses are needed to evaluate the soil properties.

In this section, procedures are presented to evaluate the in-situ stresses in both
cohesive and cohesionless soils. Vertical stresses are covered first, followed by
horizontal stresses. In each case, correlations are presented with soil index
parameters and in-situ test results, where available.

BASIC DEFINITIONS

The in-situ state of stress in soil is defined in terms of the current values of
effective vertical stress (uv0 ) and effective horizontal stress (~ 0 ). For hori-
zontal, level ground, the in-situ stress state is shown in Figure 3-l.

The current vertical stress is determined in a straightforward manner, being equal

to the effective overburden stress in which uv 0 - ~z. However, the horizontal
stress is more difficult to evaluate. The stress ratio is K0 , the at-rest coeffi-
cient of horizontal soil stress, which is defined as ~o/Ovo· As a lower bound, Ko
could equal KA, the coefficient of minimum active soil stress. The upper bound for
K0 is Kp, the coefficient of maximum passive soil stress. For horizontal, level

y =Effective unit
weight

Figure 3-l. Stresses in Soil

3-l
 i
ground and an effective stress cohesion (c) - 0, these limit states are given by
Rankine theory as below:
I
1 + sin ~psc
(3-1)
I
1 - sin ~psc

in which ~psc - effective stress friction angle for plane strain compression condi-
I
tions. Using these limits for a cohesionless soil with ~psc - 40°, for example, K0
could range from 0.2 to 4.6.
I
Many factors affect the in-situ state of stress in soil, including overconsolida-
tion, aging, chemical bonding, etc. Overconsolidation is probably most influen-
tial for the majority of soils, because it is caused by glaciation, erosion, desic-
cation, excavation, ground water fluctuations, and possibly other factors. In this
•I
regard, the effective vertical preconsolidation stress (denoted op, Ovmax• or Pc)
is an important measure of the soil stress history. This maximum past stress
affects the compressibility, strength, consistency, and overall state of stress.
I
It is often convenient to represent the stress history in terms of a dimensionless
parameter defined as the overconsolidation ratio (OCR):
lf
(3-2)

The magnitude of op and OCR can be evaluated directly from the results of one-di-
mensional consolidation tests conducted on undisturbed cohesive soil samples. Cor-
relations with other tests and soil types are presented in this section.

The magnitude of K0 may be measured directly either in the laboratory using special
testing equipment, or in the field using devices such as the pressuremeter or total
stress cells. However, these direct methods may be subject to unavoidable distur-
bance effects during sampling and in-situ testing. Alternatively, several empiri-
cal approaches can be used to evaluate the in-situ value of K0 , including: (1)
reconstruction of stress history, (2) correlations with soil index parameters, and
(3) correlations with in-situ test results. All three approaches are described in
this section.

RECONSTRUCTION OF STRESS HISTORY

Reconstruction of the soil stress history involves tracing the stress paths of the
soil as in Figure 3-2, from virgin loading, to primary unloading, to primary

3-2
 Primary unloading

o-ho
Virgin loading

Figure 3-2. Stress Paths for Simple Stress Histories

reloading, and then cyclical load-unload looping from water table fluctuations,
etc. (above pointE in figure). Virgin loading represents normally consolidated
(NC) soils with OCR- 1. All other stress paths represent overconsolidated (OC)
soils with OCR > 1.

Based on a study of 171 different laboratory-tested soils, Mayne and Kulhawy (!)
showed that a general equation can be used to model stress paths OB-BD-DE, as given
below:

Ko- Konc [ OCR + mr(l - OCR )] (3-3)

1-a OCRmax
OCRmax

in which Konc = K0 during virgin (normally consolidated) loading, a - at-rest

unload coefficient, mr = reload coefficient, OCR current overconsolidation ratio,
and OCRmax- maximum past OCR (e.g., point D for a soil currently at point E).

For virgin loading, the simplified Jaky equation (l) provides reasonable estimates
for Konc• as given below:

Konc - 1 - sin ~tc (3-4)

in which ~tc effective stress friction angle for triaxial compression. Figure
3-3 shows this equation to be a reasonable estimate for a wide range of soils. In
this figure, Konc was determined from oedometer or triaxial tests.

During rebound or unloading, the general relationship for K0 is often expressed as:

(3-5)

3-3
 u
c
0
~ Regression
In K0 nc= 0.97 (I- sin cj;tcl
·o (n=l24, r 2 =0.707, S.D.=0.06)
(/)

l)
z
....
.....0
c<I>
u
.....
.....<I>
0
l)

In
In
....<I>
({)

c0
0
• Cloy
o Silt
o Sand
I
N
....
I
0
0
0 0.8 1.0 I
sin 't'
::i.. tc

Figure 3-3. Horizontal Stress Coefficient for NC Soils from Laboratory Tests I
Source: Mayne and Kulhawy (l), p. 862.
I
As suggested by Schmidt (1), the exponent a may be expressed as a function of ~tc:

a - sin ~tc
(3-6)
•l
Alternatively, the exponent may be expressed as:

I
a - 1 - Konc

which also appears reasonable, as shown in Figure 3-4.

(3-7)

For reloading, the stress

,, "

path from D to E in Figure 3-2 may be approximated as a straight line with slope mr
= 8uh f8uvo·
0
Review of laboratory data from 35 soils (Figure 3-5) indicates that
the reload coefficient can be estimated adequately from:
I
mr = 0.75(1 - sin ~tc)
(3-8) I
Linear regressions on these data for mr give 0.76 Konc (r2 = 0.583 and S.D. = 0.06)
and 0.77(1- sin ~tc) with r2 = 0.534 and S.D.= 0.06.
I
3-4
I
f

~~ I
 a= ( 1- Konc)::!:: 0.1

0'
c
-o
0 0::
0 u
c 0
-
:::l

~
0
0'
0

10
0'
.9.
10
II
0 .o
c
•
0
Cloy
Silt
0 Sand
• Grovel
0
0 0.2 0.4 0.6 0.8 1.0

Horizonta I Stress Coefficient, Konc

Figure 3-4. Unload Coefficient for OC Soils

Source: Mayne and Kulhawy (l), p. 864.

19 Cloys I 5 Cloys
.... 0.6 16 Sands 0.6 13 Sands
E
-c
Q)

-(.)
Q)
0
0.4 0.4

u
-o 0.2 0.2
0
0
Q)
0::
o~~--~--~--~--~~--~--~
0 0.2 0.4 0.6 0.8 0.2 0.4 0.6 0.8
Horizontal Stress Coefficient, Konc sin ¢tc
Figure 3-5. Reload Coefficient for OC Soils

Source: Mayne (~), p. 269.

Combining the above relationships gives:

3-5
 f

_ _ _. . .;O:::..:C::.:.R:____ _ _ + 3 (l _ OCR ) J
K0 - (1 - sin ~tc) (3-9)
4 OCRmax
OCR (1 - sin "4>tc)
max

in which OCRmax is the OCR at point D in Figure 3-2. For primary unloading, OCR-
OCRmax and therefore:

(3-10)

For virgin loading, OCR - 1 and therefore:

Ko - 1 - sin ~tc - Konc (3-11)

Most natural soils have undergone a stress history of loading-unloading-reloading,

and therefore K0 is likely to be within points C and E in Figure 3-2. Therefore,
K0 at point E is an appropriate lower bound for the in-situ K0 . All that is needed
is ~tc• OCR, and OCRmax• which can be evaluated by direct laboratory measurements,
geologic generalization of the soil stress history, or experimental test programs
in-situ to establish the values. It should be noted that, if an NC assumption is
made (Equation 3-11), it will underestimate K0 in the majority of soil deposits.

One last point to mention regarding Figure 3-2 is that the soil can reach passive
failure during primary unloading if the vertical effective stress is reduced suffi-
ciently. This limit state can be developed from Equations 3-1 and 3-10 and is
given by:

OCRlimit- [(1 +sin ~psc)/(1- sin ~psc)(l- sin ~tc)J(l/sin ~tc) (3-12) I
As shown in Section 4, ~psc- 1.1 ~tc·
ure occurs, and the stresses change.
If this limit state is reached, soil fail-
It is uncertain what this stress state may
I
be, although K0 might approach 1.

EFFECTIVE PRECONSOLIDATION STRESS IN COHESIVE SOILS

I
Cohesive soils consolidate and stiffen during overconsolidation and effectively
retain a "memory" of the largest preconsolidation stress (up) to which they have
been subjected (e.g., point Bin Figure 3-2). This process was illustrated quali-
tatively in Figure 2-1 as a function of the water content and Atterberg limits.
Therefore, these index parameters represent a starting point for estimating ap·
Correlations with in-situ test results follow the index parameter correlations.

3-6
Details on the in-situ test strength parameters are given in Section 4.

Correlations with Index Parameters

The effective preconsolidation stress (ap) has been correlated with the liquidity
index by several authors. A recent analysis of laboratory consolidation test data
by Stas and Kulhawy (2) suggested the following:

ap/Pa- 10(1.11 - 1.62 LI) (3-13)

in which Pa = atmospheric stress in the desired stress units and LI - liquidity

index. This equation is based on 150 data points for clays with a sensitivity
between 1 and 10. This relationship has a standard deviation of 0.33 and r2 equal
to 0.740.

Other generalized relationships are shown in Figure 3-6, which gives the precon-
solidation stress as a function of liquidity index (LI) and sensitivity (St).

For comparison purposes to evaluate the soil stress history, the effective vertical
stress (av 0 ) is needed. This stress can be evaluated directly as in Figure 3-1, or
it can be estimated from the liquidity index. Based on the modified Cam clay model
and empirical observations, Wood (l) developed the following approximation for ov 0 :

avo/Paz 0.063•102(1-LI) (3-14)

.......
..J
x
Q) 0.8
-o
c:
....... 0.6
>-
:.0
:::l
cr 0.2
:.J

Preconsolidation Stress, O:P/p 0

Figure 3-6. Generalized ap - Liquidity Index - Sensitivity Relationships

Source: NAVFAC (~), p. 7.1-142.

3-7
 t
Although this equation strictly applies only to insensitive soils at the critical
state, it is a useful approximation for uncemented, low sensitivity soils.
I '

Comments on Field Test Correlations I

It should be noted that the following figures correlating up with field test meas-
urements are presented all in similar form, on log-log plots because of the range
in the parameters involved. These figures were developed from the sources noted,
I
and the symbols used correspond to the clay types referenced in the source papers.
For each figure, the number of intact and fissured clays is noted, the fissured
I
clays are located separately because their behavior is different, and a linear
regression equation is presented for the intact clays only.
done assuming a linear, arithmetic relationship through the origin.
The regression was
The statistics
I
given with each regression include the number of data points (n), coefficient of
determination (r2), and the standard deviation of up (S.D.) for a given field test
I
measurement. The given relationships should be used only as predictors for ap·

Correlations with VST Strength

l
The field vane shear test (VST) has been used for many years as an estimator of
up· In 1957, Hansbo (~)developed the following equation for Swedish clays:

(3-15)
'I
in which OVST is an empirical factor approximately equal to 222/wL, with wL -
liquid limit (in percent).
I
A more recent compilation of worldwide clays, shown in Figure 3-7, indicated the
I
general nature of Equation 3-15. This study further showed that OVST could be
correlated weakly with the plasticity index (PI), as shown in Figure 3-8. I
Correlations with SPT N Value

The standard penetration test (SPT) N value may be used to provide a first-order
estimate of up for cohesive soils.
clays.
Figure 3-9 shows the available data for 51
The regression shows a fair correlation with a relatively large standard
•I
deviation.

It should be noted that the reported N values have not been corrected for the fac-
a
tors which significantly affect the SPT N value. Until the N values are corrected
to a consistent standard, the SPT is likely to be of limited use in evaluating ap· I
3-8 I
I
I
I 50 / 0 I
~I /
~--"'
I a.
0

........
20
O:p = 3.54 su (VST)
(n = 205, r 2 =0.832,
fissured

~ +.,
I be.
S.D.=0.95p0 ) ~~~f/ij

'f
en
Vl
Q)
.... .f~

d~
(f)

c:
-
0:1
0 lie
[J
0
-o ~

~rn.P'&
f 0

,,
Vl
c: a
0 I'll
11. Qt})
u
Q)
.... 0 ~ <JII
Q_ a ~"~ 96 intact cloys
0.2 a

0.1
I'll~

0.01 0.02
/
0.05 0.1 0.2 0.5
I fissured cloy

2 5

' Field Vane Strength,

Figure 3-7.
Su

ap Correlated with VST su

{VST) /p 0

'
I~
Source: Based on Mayne and Mitchell(~), p. 154, and others (10).

I
1\ 1-
(f)
>

I
,,
'

"
I 0
0
I
20
I
40
I
60
I

80
I

100

Plasticity Index, PI (%)

'I, ;{_
~J
Figure 3-8.

Source:
Field Vane Coefficient versus PI

Mayne and Mitchell <2)' p. 154.

I 3-9

~~ <,;_
 I
a.
''be.
0
20

10
i ~~},

I
IJ)
IJ)
5
-
Q)
,_
( /)

I
I
0.2
49 intact cloys
2 fissured cloys
,,
SPT N Value (blows/ft or 305mm)
Figure 3-9. op Correlated with SPT N

Source: Based on Mayne and Kemper (11), p. 144, and others (12).

Correlations with CPT qc Value

The cone penetration test (CPT) tip resistance, qc, has been used effectively to
profile the preconsolidation stress in clays. Figure 3-10 presents the available
data from 49 clays. This correlation is somewhat better than with the N value, and
the standard deviation is smaller. This correlation also shows more clearly that
the fissured clays behave differently from the intact clays. However, it is impor-
tant to note that the data in Figure 3-10 are not corrected for pore water stress
effects.

Correlations with CPTU Results ~,

The piezocone (CPTU) provides additional data during penetration and generally is
considered to be a more sensitive type of cone penetration test. Tavenas and
Leroueil (20) demonstrated that the preconsolidation stress (op) was well-corre-
lated with the net corrected cone tip resistance (qT - av 0 ) for eleven Canadian
I
clays. A larger sample of piezocone data is shown in Figure 3-11.
in this case gave an even higher r2 with lower standard deviation.
The regression
II
In addition to measurements of cone tip resistance, piezocones provide the
l•
t

3-10 i
 50

0.
0 B"p=0.29 Qc
....... (n= 113, r 2 =0.858,
be.
S.D.= 2.31 p0 )
.n
(/)

- ....
Q}

( f)

c:
-0
0

"'0
0
(/)
c:
0
u
Q}
....
a._ 39 intact clays
0.2 10 fissured clays
0.1
0.5 2 5 10 20 50 100 200

Cone Tip Resistance, qc/pa

Figure 3-10. up Correlated with CPT qc

Source: Based on Mayne (13), p. 786, and others (14 - 19).

(/)
en
Q}
....
(f)

c:
0
0
"'0
0en
c:
0
u
....
Q}

a._

2 5 10 20 50 100

Figure 3-11. ap Correlated with CPTU qT

Source: Based on Mayne and Holtz (21), p. 25, and others (14, 15, 17 - 19).

3-11
magnitude of pore water stress (8u) caused by penetration. A relationship between
op and Aut from CPTU tests with tip or face pore water stress measurements is shown
in Figure 3-12. For pore water stress measurements behind the tip, the relation-
ship is given in Figure 3-13. The results are similar for the intact clays. How-
ever, for piezocones in heavily overconsolidated fissured clays, pore water
stresses measured behind the tip are near zero and sometimes are even negative. On
the cone tip, positive pore water stresses are observed for all clays at all OCR
values, regardless of whether fissuring is present.

From cavity expansion theory, the general relationship between op and the excess
pore water stress measured at the tip during piezocone penetration can be given by
the following (23):

(3-16)

in which M- critical state parameter (Appendix G) and Ir =rigidity index (G/su)·

For measurements behind the tip, the coefficient 3 becomes equal to 4. This equa-
tion gives values consistent with those in Figures 3-12 and 3-13 for the intact

0
0.
........
'be.
O:p=0.47 6ut
(n= 77, r 2 =0.838,
S.D.= 1.48 p0 l

c
g

T;p/loc~•
0
-o
0
VI
c
0
(.)
Q)
....
a..
37 intact cloys
5 fissured cloys

Excess Pore Water Stress from Piezocone, ~u,lp 0

Figure 3-12. op Correlated with CPTU 8ut

Source: Based on Mayne and Holtz (21), p. 23, and others (14, 18, 22).

3-12
 (/)
(/)
0)
.:::
(/)

c:
g
0
:52
0
(/)
c:
0
u
0)
....
a..

0.2 17 intact cloys

5 fissured cloys

Excess Pore Water Stress from Piezocone, 6ub 1/p0

Figure 3-13. op Correlated with CPTU 6Ubt

Source: Based on Mayne and Holtz (21), p. 24, and others (14, 15, 17, 19, 22).

clays.

Correlations with PMT Results

Several correlations have been attempted with the pressuremeter test (PMT) to esti-
mate the value of op. Early work with the Menard pressuremeter indicated that the
PMT creep pressure was approximately equal to op for Chicago area lake clays (24).
Later work showed that the limit stress from the self-boring pressuremeter test
(SBPMT) could be correlated with op, as shown in Figure 3-14. Other studies have
shown the correlations given in Figure 3-15, including the undrained shear strength
(su) and the rigidity index (Ir).

Correlations with DMT Results

The initial contact stress (p 0 ) from the dilatometer test (DMT) is a measure of the
induced total pore water stress caused by insertion of the DMT blade. Analogous to
the previous relationship between op and 6u for piezocone tests, a similar rela-
tionship applies for the DMT between op and (p 0 - Uo), as shown in Figure 3-16.

3-13
 0
a.
........
•tf"

c
g
0
"0
0
V>
c

'I
0
u
...
Q)

CL
0.5

0.2
• 35 intact cloys

• 6 fissured cloys

0.1
0.5
~~U---~~~~~~~----~~~~LL~
2 5 I0

SBPMT Limit Stress, pL/p0

20 50 I 00
I
Figure 3-14. op Correlated with SBPMT PL

Source: Data from Mayne and Kulhawy (25), and others (12, 16, 19, 26).

EFFECTIVE PRECONSOLIDATION STRESS IN COHESIONLESS SOILS

Cohesionless soils also consolidate and stiffen during overconsolidation and retain
a "memory" of the preconsolidation stress. However, cohesionless soils are diffi-
cult to sample and test in the laboratory in the undisturbed state, and therefore
little correlation information is available to estimate the preconsolidation stress
in these soils. More work has focused on evaluating OCR and K0 directly, as
described later.

OVERCONSOLIDATION RATIO FOR COHESIVE SOILS I

•
In lieu of describing soil stress history by the preconsolidation stress (op), the
in-situ overconsolidation ratio (OCR) may be estimated directly using normalized
! .
parameters developed from laboratory or field test measurements. These correla-
tions strictly apply only to insensitive clays. Furthermore, the same comments
made previously on field test correlations with respect to op also apply to the OCR

3-14

I
~ ..
I (/)
(/)
-
I
Q)
.!::
(f)

c:
:20
J :2
0(/)
c:
0

I u
Q)
~

a..

I 38 intact cloys
6 fissured cloys

I
I Figure 3-15. ap Correlated with SBPMT su and Ir

Source: Based on Mayne and Bachus (23), p. 293, and others (12, 16, 19, 26).

I
correlations.
I
Correlations with Index Parameters

I Equation 3-13 can be re-cast in terms of OCR as follows:

I OCR= (Pa/avo) lO(l.ll - 1.62 LI)

As noted previously, this relationship is based on statistical analysis of labora-

(3-17)

I tory consolidation test data on clays with sensitivity from 1 to 10.

I Based on the modified Cam clay model and empirical observations, Wood (l) developed
Equation 3-14 to correlate avo with LI. He also developed the following:

I log OCR z [2 - 2 LI - log (15.87 ovo/Pa)J/A (3-18)

I 3-15

I
 c:
.Q
0
:2
0V'l
c:
0
u
...
Q)

0..
24 intact cloys
7 fissured cloys
0.1~~~-W~~--~~~~~U---~~~~~~
0.2 0.5 2 5 10 20 50 100

Effective OMT Contact Stress, (p -u )/p

o o a
I
Figure 3-16. ap Correlated with DMT p 0

Source: Based on Mayne (27), p. 148, and others (26).

in which A= critical state parameter (Appendix G). Using a typical value of A-

0.8, and combining Equations 3-14 and 3-18, results in the following:

OCR= 10 [1 - 2.5 LI - 1.25 log (avo/Pa)] (3-19)

I
Although this equation strictly applies only to insensitive soils at the critical
state, it is a useful approximation for uncemented, low sensitivity soils, as noted
previously.
I
Correlations with Laboratory Strength

Laboratory undrained shear strength (su) data may be used to estimate the in-situ
OCR of clays. Using empirical observations from isotropically and anisotropically
consolidated triaxial compression tests, Mayne (28) observed the following for OCR:
'
I
(3-20)
I
(3-21) I
3-16 I
These results are consistent with the modified Cam clay model, which would predict
the following:

OCR= 2((suf0vo)/0.5 M]-1/A (3-22)

Correlations with VST Strength

The undrained strength from the field vane shear test (VST) may be related to the
in-situ OCR according to:

OCR = OVST (sufOvo)vsT (3-23)

in which avsT has been shown in Figure 3-8 to be related weakly to plasticity index
(PI). Figure 3-17 shows a direct relationship between OCR and sufovo for 96 clays.

Correlations with SPT N Value

Attempts have been made to correlate the SPT N value with OCR. Figure 3-18 is typ-
ical of these correlations, using uncorrected N values. This relationship is only
a first-order estimator.

50

a::
u
0

-0
0
a::
c

-
0
0
"'0
0
Vl
c
0
u
,_
(1)
>
2
0 96 intact clays
~
I
0.1 0.2 0.5 2 10

Normalized Field Vane Strength, (s 0 /~ 0 ) vsT

Figure 3-17. OCR Correlated with VST su

Source: Based on Mayne and Mitchell (2), p. 152.

3-17
 a:
u
0

g
0 OCR=0.58 Np0 /0:vo
a:
(n=ll2, r2=Q.661,
c:
.9 S.D.:: 3.82)
0
"0
:g 5
ell
c:
0
...
u
Q)
:>
0
48 intact cloys
3 fissured cloys

2 5 10 20 50 100
SPT N Value, N p0 /0:vo (blows/ft or 305 mm)

Figure 3-18. OCR Correlated with SPT N

Source: Based on Mayne and Kemper (11), p. 143, and others (12).

Correlations with CPT and CPTU Results

A number of authors (e.g., 13, 29) have demonstrated that OCR correlates with the
CPT qc value through the normalized cone tip resistance, (qc - av 0 )/ovo· However,
qc also should be corrected for pore water stresses acting on unequal areas of the
cone. Figure 3-19 shows the variation of OCR with the corrected cone tip resis-
tance, qT, as obtained from piezocones.

Other piezocone studies (31) suggested a general trend with Bq (Equation 2-4) and
OCR that was strongly dependent on the rigidity index. However, Bq is so site-
dependent that the relationship was of little predictive use. More recent work
(32) considered a combined critical state/cavity expansion model to correlate OCR
with piezocone results. However, at the present time, the relationship given in
Figure 3-19 probably is most appropriate to use.

Correlations with DMT Results

In the initial introduction of the dilatometer test, Marchetti (33) proposed the
correlation in Figure 3-20 between OCR and the DMT parameter Ko, given by:

3-18

I
 8

7
0::
(.)
0
6
.2
0
0::
5
c:
.2
0
"0 4
0
en
c:
0
u
..... 3
Q)
>
0
2
52 intact clays

I
0 15 20 25

Corrected Cone Tip Resistance, (qT- crv 0 ) I a-vo

Figure 3-19. OCR Correlated with CPTU qT

Source: Data from Mayne (30), and others (18, 19, 22).

OCR= (0.5 KD)l.56 (3-24)

in which KD =horizontal stress index= (p 0 - u 0 )/ovo• p 0 =initial contact stress,

u0 - hydrostatic pore water stress, and ovo =effective vertical stress. Subse-
quent research with the DMT in other countries suggests a more general expression:

(3-25)

in which the parameter Po depends upon the degree of fissuring, sensitivity, and
geologic origin, as shown in Figure 3-21.

OVERCONSOLIDATION RATIO IN COHESIONLESS SOILS

It is difficult to estimate the in-situ OCR of natural sand deposits. The best
approach is through a detailed geologic study to evaluate the stress history of the
formation. Indirectly, oedometer tests on interbedded clay strata or seams may
give clues to the in-situ OCR of the surrounding sands. With the DMT, a value of
OCR in sands can be back-calculated from the estimated K0 as (Bullock, 36):

3-19
 100

Fissured cloys
50 50 (~ 0 =0.75) ~
0::
I
0::
I

;•
• u
0

•"•
u .Q 20
20 0
0 Insensitive cloys
0::
(/30 =0.50)
i
...I ••
c
[original Marchetti]
0
0:: tV -
.Q

"'
0
i>" ."Q
c ¢::<:)

-
.2
0
~
....._Q
'-. . ; .§'
.:,.'b
~
01/)
c
0
u
5

~~
0
1/)
we.? ....
Q)
c .§ •• >
0
u
0 2
.... 2
Q)

0
>

I
I
'I
.M.-e
2 5 10 20 2 5 10 20

Horizontal Stress Index, K 0 Horizontal Stress Index, K 0

Figure 3-20. OCR Correlated with DMT Figure 3-21. OCR - KD Relationships for
KD Clays of Varied Geologic
Origin
Source: Marchetti (33), p. 315.
Source: Based on Marchetti (33), Powell
and Uglow (34), and Lacasse and
Lunne (35).

OCR [Ko/(1 - sin ~tc)J(l.25/sin ~tc) (3-26)

which is a form of Equation 3-10 that has been rearranged and modified to fit the
results of laboratory calibration chamber tests on sands.

EFFECTIVE HORIZONTAL STRESS IN COHESIVE SOILS

As noted previously, soils retain a "memory" of preconsolidation. With vertical

stresses, this memory is reflected by the preconsolidation stress (op) which, in OC
soils, is greater than the effective overburden stress (ov0 ). In the horizontal
direction, the process is somewhat different, because the soil can not unload as
freely as it can in the vertical direction. The result is that the retained memory
of the maximum horizontal effective stress is less clear. If the soil is young and
has experienced only a relatively simple stress history, then the procedures
described earlier under "Reconstruction of Stress History" can be used to evaluate
the horizontal effective stress in terms of K0 , defined as oho/Ovo· For older

3-20
soils or soils with more complex stress history, the reconstruction process can be
more difficult. By default in these cases, it may be necessary to assume only pri-
mary unloading, as shown in Figure 3-2. This assumption will result in an upper
bound on K0 , which must be used with some considered engineering judgment, taking
into account the loading level and differences between the virgin loading and pri-
mary unloading values of K0 .

Alternatively, K0 may be estimated from index parameters or correlations with

in-situ measurements. Ideally, these approaches reflect the soil in-situ and
therefore should be good indicators of the current K0 . However, all correlations
contain uncertainties and must be considered within the context of the stress
history of the soil. The predicted K0 should be consistent with this information.

Correlations with Index Parameters

A number of studies have attempted to correlate K0 with the Atterberg limits.

Figure 3-22 shows one of these relationships for NC clay. As shown, organic clays
should be excluded from the general trend. For OC soils, an early study demonstra-
ted the behavior shown in Figure 3-23, with the overconsolidation ratio (OCR) domi-
nating the resulting K0 value. These two figures suggest a high degree of correla-
tion with the Atterberg limits. However, more comprehensive data compilations show
the lack of correlation given in Figure 3-24, which has an r2 equal to 0.147.

One simple alternative estimator is to assume overconsolidation by simple unload-

ing, which was described previously as:

1.0 1.0

......
... .../
Konc 0.5 •
.
~·
/·
#
0
0.5 '-
.:./...·""-·..
....
• e" •
/

0 0 -
• •• • • 0 co 0
/. 0

• Inorganic cloys
o Organic cloys
0 0
0 40 80 0 40 80 120 160

Plasticity Index, PI (%) Liquid Limit, WL (%)

Figure 3-22. Konc Correlated with Atterberg Limits

Source: Larsson (37), p. 21.

3-21
 5 cloys and I sand

20 40 60 80
Plasticity Index, PI (%)
Figure 3-23. K0 Correlated with PI and OCR

Source: Brooker and Ireland (38), p. 14.

0.9 I I
I I

0
'-
cf>tc < 29o
r2
0.8 -
=0.147
... zoo < cf>tc < 30°
_ < cf> tc < 40°-
"i1 30°

0.7
0
• c/>tc > 40o
0
...1JtJ¢I 1:)0 ...

•
~ ...
"i1 • A ... o
C)
...
-
f. "' ..."'.
Konc 0.6 - "i1 "il ...
... "i1... ...

... 'V 'ilil ..

....
~
0.5
~Jf~·... ~"il -
~~."il~ \l/"il
0.4 I-
"i1
.... 135 cloys
-
"i1
I I I I I I I
0.3
0 20 40 60 80
Plasticity Index, PI (%)

Figure 3-24. Apparent Lack of Trend Between Konc and PI for 135 Clay Soils

Ko - (1 - sin ~tc) OCRsin ~tc (3-10)

Figure 3-25 illustrates this approach for 48 clay soils. Also, the following
approximation was given earlier:

3-22
 48 cloys

2.0

AEB

1.5

0.5

Overconsolidation Ratio, OCR

Figure 3-25. K0 Correlated with OCR

Source: Data from Mayne and Kulhawy (l).

OCR= (pa/uvo) 10(1.11 - 1.62 LI) (3-17)

By combining these two equations, K0 can be estimated simply from a knowledge of

~tc• Uvo• and LI.

One further simplification is to note that ~tc = 30° is a reasonable fit of the
data in Figure 3-25. Using this value, Equation 3-10 reduces to:

K z 0.5 OCR0.5 (3-27)

0

Then, combining this result with Equation 3-17 yields:

3-23
 (3-28)

which is a simple, first-order estimator requiring only avo and LI.

If information is available for the undrained shear strength (su), then the corre-
lation shown in Figure 3-26 can provide an estimate for K0 .

Direct Correlations with SBPMT and DMT Results

The self-boring pressuremeter test (SBPMT) has shown promise as one of the few
devices capable of providing a direct measurement of the in-situ horizontal
stress. There is no need for correlations because the stress is measured directly,
taking into account equipment calibrations. Figure 3-27 shows results summarized
for 56 clays in the literature, in which both K0 and OCR values were given.
be seen, the trends are consistent with those shown previously (Figure 3-25) for
As can I
laboratory data. It should be noted that the fissured and intact clays behave sim-
ilarly when tested with the SBPMT because this test involves an expanding device
which compresses the soil and fissures to mimic an intact soil.
I

I
I
I

13 cloys
I
CK 0 UC tests

(su /O:vo>oc
(suI c:Tvo>Nc

Figure 3-26. K0 Correlated with Undrained Strength Ratio

Source: Modified after Mayne (39).

3-24
 I-
~
a..
(I)
(/)

E
0
'-

2 5 10 20 50 100

OCR from Oedometer

Figure 3-27. K0 from SBPMT Correlated with OCR

Source: Based on Mayne and Kulhawy (25 and 40).

The original intent of the dilatometer test (DMT) was to model the soil modulus for
the laterally loaded pile problem, which requires an assessment of the horizontal
stress. However, all in-situ testing devices cause some disturbance upon insertion
into the ground. Therefore, Marchetti (33) found it necessary to develop a corre-
lation between a best estimate K0 and the DMT horizontal stress index (Kn). as
shown in Figure 3-28. The original Marchetti equation was based primarily upon
data from insensitive Italian clays and uncemented normally consolidated sands and
was given as:

K0 - (Kn/1.5)0.47 - 0.6 (3-29)

Powell and Uglow (34) tested heavily overconsolidated and fissured clays from the
United Kingdom with the DMT and found that, although the in-situ K0 trended with
Ko, the relationship was offset from the original one established for Italian
clays. Similarly, Lacasse and Lunne (35) used the DMT at several Norwegian sites

3-25 .··--
 Sensitive cloys
(,Bk=2.0)

Horizontal Stress Index, K 0

Figure 3-28. K0 Correlated with KD

Source: Based on Marchetti (33), Powell and Uglow (34), and Lacasse and Lunne (35).

and suggested further modifications to the original Marchetti correlation. Both

data sets also are shown in Figure 3-28. Considering these other data, a general
equation for K0 is:

(3-30)

in which Pk depends upon soil type and geologic origin.

Where possible, local calibration of the DMT should be made relative to K0 measure-
ments obtained with SBPMT or push-in spade cells. For preliminary estimating pur-
poses, the values of Pk in Figure 3-28 may be used.

Figure 3-29 shows a direct comparison of K0 from the SBPMT with KD from the DMT.
As can be seen, the SBPMT K0 for stiffer clays is higher than the original K0 pre-
diction by Marchetti (33).

Indirect Correlations with SPT, CPT, CPTU, and DMT Results

The standard penetration test (SPT), cone penetration test (CPT), and piezocone
test (CPTU) all are measurements of vertical penetration, and therefore they do not
address K0 directly. However, vertical penetration is coupled with the horizontal

3-26

I
 41-
I-
:!
a_
CD
(/)
3 -
E
-
-
0
.... 2

7 intact cloys -
5 fissured cloys
I I
10 15 20
Horizontal Stress Index, K 0

Figure 3-29. K0 from SBPMT Correlated with Ko

Source: Data from Mayne and Kulhawy (25 and 40).

stresses because they control the vertical "stiffness" of the soil and the shearing
resistance of the advancing in-situ device. Alternatively, the DMT provides meas-
urements of horizontal total stress. These measurements are taken immediately
after penetration of the blade into the clay and, as such, reflect large increases
in total horizontal and pore water stresses over the geostatic state of stress.
Consequently, the SPT, CPT, CPTU, and DMT provide indirect measurements of K0 .

Figure 3-30 shows the trend of K0 obtained from laboratory tests and DMT, PMT, and
SBPMT measurements with the normalized SPT N value. From regression analyses of
these data, K0 can be given by the following:

(3-31)

Figure 3-31 shows the trend of K0 from SBPMT measurements with the normalized cone
tip resistance. From these data, K0 can be given by the following:

(3-32)

K0 also can be estimated from the piezocone pore water stress, as shown in Figure
3-32. These data show that K0 can be given by:

(3-33)

3-27
 - -
4

1-

3-

-
2- -
- -
-
13 intact cloys
5 fissured cloys
I I I I

40 50 60

Figure 3-30. K0 Correlated with SPT N

Source: Kulhawy, et al. (41), p. 129.

I-
~
a..
(I)
3- -
(/)

-
0
.... -

-
12 intact cloys
5 fissured cloys

I I
0
0 10 20 30 40 50 60

{ qT_-a-vo) f rom CPT

CTVO

Figure 3-31. K0 Correlated with CPTU qT

Source: Kulhawy, et al. (41), p. 128, and others (40).

3-28
 5
•
f- 4
~
a..
(J)
(/)
3
E

-
0
....
2
0 .6u 1 at tip
~

12 intact clays
5 fissured clays
o~~~~~~--~~J-~-L~~--~~J_~~~~~
0 5 10 15 20

( ~Ut ) from CPTU

o-vo

Figure 3-32. K0 Correlated with CPTU ~u

Source: Data from Mayne and Kulhawy (15 and 40).

An example of K0 profiling by several in-situ tests in London clay is presented in

Figure 3-33. Measured values from the SBPMT and estimated values using the origi-
nal Marchetti (33) DMT correlation are given, along with correlations developed
from the SPT, CPT, and the liquidity index. Although there is obvious scatter, all
of the results are consistent with each other.

EFFECTIVE HORIZONTAL STRESS IN COHESIONLESS SOILS

Cohesionless soils also retain a "memory" of preconsolidation. However, as noted

previously, the stress history in cohesionless soils is more difficult to determine
because of sampling problems. Therefore, the focus has been almost exclusively on
in-situ tests.

Direct Correlations with SBPMT and DMT Results

The self-boring pressuremeter test (SBPMT) has shown promise as one of the few
devices capable of providing a direct measurement of the in-situ horizontal
stress. As such, there is no need for correlations because the stress ideally is
measured directly. However, the SBPMT has not been used widely in cohesionless
soils because of the relatively high cost, low productivity, and difficulties in
advancing the device in the field.

3-29
 Ko
0 2 3 4
or-----.-----.-----,-----r-----.-----r-----.-----,
London Cloy at
Brent Cross, Hendon
(Powell and Uglow, 1986) •
• SBPMT results
....
c .A DMT results
~

..c
0. •
0
Q)

•
0.073 N p 0

Figure 3-33. Comparison of K0 Values for London Clay at Brent Cross

Source: Kulhawy, et al. (41), p. 129.

One intent of the dilatometer test (DMT) was to provide a measurement of the hori-
zontal soil stress, as noted previously. Unfortunately, all in-situ testing
devices cause some disturbance upon insertion into the ground. Therefore, Mar-
chetti (33) found it necessary to develop a correlation between a best estimate K0
and the DMT horizontal stress index (Ko), as shown in Figures 3-28 and 3-29. How-
ever, Schmertmann (42) showed by calibration chamber tests that the original rela-
tionship should also be dependent upon the effective stress friction angle (~tc),
as given in Figure 3-34. Other correlations with CPT results are given below.

Indirect Correlations with SPT and CPT Results

No correlations have been developed to date between K0 and the standard penetration
test (SPT) N value. However, it was shown in Section 2 that the N value could be
correlated with the cone penetration test (CPT) Qc value. Therefore, the Qc corre-
lations below could be used approximately with N values converted to "equivalent"
Qc values.

For the CPT Qc value, Durgunoglu and Mitchell (44) developed a theory to relate the
cone factor, K0 , ~tc• and depth (D) to diameter (B) ratio. This theory has been
used to develop Figure 3-35, from which an estimate of K0 can be made. This figure
must be used cautiously because small changes in the cone factor or ¢tc can result

3-30
 Marchetti
(1980)

-
¢~~o_...........-
~ y~I
// I
y £
35o/ / I Schmertmonn
(1983)
j40°

2 10 20
Horizontal stress Inde X, K 0

Figure 3-34. Ko Correlated with Ko in Sands

Source: Marchetti (43)

- ' p. 2668.

"D/8=50
0 0/8 = 100
0 0/8 = 200

OJ
c:
0
u

I I

'
2 4 5

I Figure 3-35. Cone Factor versus Ko as a Function of '4>tc and D/B

I 3-31

I
in large K0 changes. However, careful us~ of this figure with a good knowledge of
the soil stress history can result in reasonable K0 predictions. An example using
this approach is given in Figure 3-36.

Marchetti (43) also used this theory and developed a more simplified relationship,
as shown in Figure 3-37. In this figure, B- 35.7 mm for a standard cone was
introduced. Note that these curves also are quite flat, and that small changes in
the input parameters can give large K0 changes.

Combined DMT/CPT Approach for K0 of Sands

In a novel approach, the combined results of DMT and CPT calibration chamber tests
on laboratory-prepared sand (Figure 3-38) indicated a best fit expression for K0 in
terms of both the horizontal stress index (Kn) and normalized cone tip resistance
(qcfov 0 ), as given below:

0.359 + 0.071 Kn - 0.00093 (qcfav 0 ) (3-34)

This equation was modified to account for field CPT and DMT measurements obtained
in a natural sand deposit. The differences between the laboratory and field rela-
I
tionships may be a result of aging effects. This phenomenon of aging is quite
important, but it is not very well understood at present, as noted in Section 2. I
I
I
• I
E
_c
a.
Q)
(using Fig. 3-35
with ~tc =35°)
I
0

15 I
I
Figure 3-36. Estimation of K0 in Coastal Plain Sand from CPT

Source: Kulhawy, et al. (41), p. 130. I

3-32
I
I
 10 20 50 100 200 500 1000 2000
Cone Tip Resistance, qc/O:vo

Figure 3-37. Simplified qc - K0 - ~tc Relationships

Source: Marchetti (43), p. 2668.

Horizontal Stress Index, K

0
Figure 3-38. K0 Correlated with qc and Kn

Source: Marchetti (43), p. 2672.

Empirical Approach

Data from CPT studies using electric cones in calibration chamber tests (e.g.,
Appendix H) indicate that the initial effective horizontal stress (ah 0 ) is more
influential on the magnitude of qc than the vertical stress. Furthermore, the
relationship between aho and qc appears to be independent of OCR. The advantages

3-33
of using laboratory chamber tests include known stress state, stress history, and
in-place density prior to penetration.

A tentative evaluation of the calibration chamber data is shown in Figure 3-39,

indicating a general trend between ~0 , ~. and Dr. The value of ~0 is the
imposed effective horizontal stress prior to cone penetration. With this figure,
measured values of qc and Dr are used to obtain ~0 , as given below:

(3-35)
Pa 35 exp (Drf20)

Once oho is known, KQ can be computed from ~o/Gvo·

Application of this empirical approach for estimating in-situ K0 from CPT data in
an overconsolidated sand near Stockholm is shown in Figure 3-40. The stress his-
tory of this sand has been documented well in the literature, and Equation 3-34 was

0.0

'•tf= 0

rn
rn

-,_
Q)

( /)

0
c0
N
"-
0
I
Q)

-
.2:
u

-w
Q)

Cone Tip Resistance, qc I P0

Figure 3-39. Tentative Correlation Between oho, qc, and Dr for NC and OC Sands
Tested in Calibration Chambers

Source: Kulhawy, et al. (41), p. 132.

3-34
I
I
 •• ....... -
-
E
~

..s::::
2
r;~r.-
•:/. I
/

. \
... Kp limit for
cj;tc=40o
CPT estimate

~-
a. 4
Stockholm sand
Q)
••
~F
0

.a.7 •
6
;;;./Calculated from
A stress history, A SPT estimate
A / K 0 =0.35 OCR 0 ·65 e PMT measurement

Figure 3-40. Comparison of K0 Values at Stockholm Site

Source: Kulhawy, et al. (41), p. 132.

used to evaluate the in-situ K0 . As can be seen, the agreements are quite good.

I The CPT approach may be extended to SPT results through an approximate correlation

I between the cone tip resistance and N value.

lated to mean particle size (expressed as Dso). as shown in Figure 2-30.
The ratio of qc to N has been corre-
For the
Stockholm site, the value of D50 averages about 0.9 + 0.1 mm, suggesting a qc/N

I ratio of about 6.5. This conversion has been used to estimate a profile of K0 from
SPT data using the CPT empirical procedure. Figure 3-40 shows reasonable agreement

I between the profiles of K0 estimated from CPT and SPT resistances and values deter-
mined from the known stress history and PMT data.

I REFERENCES

l. Mayne, P. W. and Kulhawy, F. H., "K0 - OCR Relationships in Soil", Journal of

I the Geotechnical Engineering Division, ASCE, Vol. 108, No. GT6, June 1982, pp.
851-872.

2. Jaky, J., "The Coefficient of Earth Pressure at Rest", Journal of the Society
I 3.
of Hungarian Architects and Engineers, Budapest, Oct. 1944, pp. 355-358.

Schmidt, B., Discussion of "Earth Pressures at Rest Related to Stress His-

tory", Canadian Geotechnical Journal, Vol. 3, No. 4, Nov. 1966, pp. 239-242.
I 4. Mayne, P. W., Discussion of "C0 /Cc Concept and K0 During Secondary Compres-
sion", Journal of Geotechnical Engineering, ASCE, Vol. 115, No. 2, Feb. 1988,

I pp. 267-270.

I 3-35

L
 5. Stas, C. V. and Kulhawy, F. H., "Critical Evaluation of Design Methods for
Foundations Under Axial Uplift and Compression Loading", Report EL-3771, Elec-
tric Power Research Institute, Palo Alto, Nov. 1984, 198 p.

6. NAVFAC, Soil Mechanics (DM 7.1), Naval Facilities Engineering Command, Alexan-
dria, 1982, 355 p.

7. Wood, D. M., "Index Properties and Critical State Soil Mechanics", Proceed-
ings, Symposium on Recent Developments in Laboratory and Field Tests and
Analysis of Geotechnical Problems, Bangkok, Dec. 1983, pp. 301-309.

8. Hansbo, S., "A New Approach to the Determination of Shear Strength of Clay by
the Fall Cone Test", Report 14, Swedish Geotechnical Institute, Stockholm,
1957.

9. Mayne, P. W. and Mitchell, J. K., "Profiling of Overconsolidation Ratio in

Clays by Field Vane", Canadian Geotechnical Journal, Vol. 25, No. l, Feb.
1988, pp. 150-157.

10. Marsland, A., "Design Parameters for Stiff Clays", Proceedings, European Con-
ference on Soil Mechanics and Foundation Engineering, Vol. 5, Brighton, 1979,
pp. 159-162.

11. Mayne, P. W. and Kemper, J. B. , "Profiling OCR in Stiff Clays by CPT and SPT",
Geotechnical Testing Journal, ASTM, Vol. 11, No. 2, June 1988, pp. 139-147.

12. Garrett, C. and Barnes, S. J., "The Design and Performance of the Dunton Green
Retaining Wall", Geotechnigue, Vol. 34, No. 4, Dec. 1984, pp. 533-548.

13. Mayne, P. W., "CPT Indexing of In-Situ OCR in Clays", Use of In-Situ Tests in
Geotechnical Engineering (GSP 6), Ed. S. P. Clemence, ASCE, New York, 1986,
pp. 780-789.

14. Campanella, R. G. and Robertson, P. K., "Current Status of the Piezocone

Test", Proceedings, 1st International Symposium on Penetration Testing (ISOPT-
1), Vol. 1, Orlando, 1988, pp. 93-116.

15. Lunne, T., Eidsmoen, T., Powell, J., and Quarterman, R. , "Piezocone Testing in
Overconsolidated Clays", Proceedings, 39th Canadian Geotechnical Conference,
Ottawa, 1986, pp. 209-218.

16. Mahar, L. and O'Neill, M. W., "Geotechnical Characteristics of Desiccated

17.
Clay", Journal of Geotechnical Engineering, ASCE, Vol. 109, No. 1, Jan. 1983,
pp. 56-71.

Mayne, P. W. and Frost, D. D., "Geotechnical Report, White House Communica-

tions Agency, Anacostia, Washington, D.C.", Report W6-5523, Law Engineering,
McLean, 1986, 95 p.
'
18. Powell, J. J. M., Quarterman, R., and Lunne, T., "Interpretation and Use of
the Piezocone Test in U.K. Clays", Penetration Testing in the U.K., Thomas
Telford, London, 1988, pp. 47-52.

19. Rad, N. S. and Lunne, T., "Direct Correlation Between Piezocone Test Results
and Undrained Shear Strength of Clay", Proceedings, 1st International Sympo-
sium on Penetration Testing (ISOPT-1), Vol. 2, Orlando, 1988, pp. 911-917.

3-36 '
I
I
 20. Tavenas, F. and Leroueil, S., "State-of-the-Art on Laboratory and In-Situ
Stress-Strain-Time Behavior of Soft Clays", Proceedings, International Sympo-
sium on Geotechnical Engineering of Soft Soils, Mexico City, 1987, pp. 1-46.

21. Mayne, P. W. and Holtz, R. D., "Profiling Stress History from Piezocone Sound-
ings", Soils and Foundations, Vol. 28, No. 1, Mar. 1988, pp. 16-28.

22. LaRochelle, P., Zebdi, M., Leroueil, S., Tavenas, F., and Virely, D., "Piezo-
cone Tests in Sensitive Clays of Eastern Canada", Proceedings, 1st Interna-
tional Symposium on Penetration Testing (ISOPT-1), Vol. 2, Orlando, 1988, pp.
831-841. .

23. Mayne, P. W. and Bachus, R. C., "Penetration Pore Pressures in Clay from CPTU,
DMT, and SBP", Proceedings, 12th International Conference on Soil Mechanics
and Foundation Engineering, Vol. 1, Rio de Janeiro, 1989, pp. 291-294.

24. Lukas, R. G. and de Bussy, B., "Pressuremeter and Laboratory Test Correlations
for Clays", Journal of the Geotechnical Engineering Division, ASCE, Vol. 102,
No. GT9, Sept. 1976, pp. 945-962.

25. Mayne, P. W. and Kulhawy, F. H., Discussion of "Independence of Geostatic

Stress from Overconsolidation in Some Beaufort Sea Clays", Canadian Geotechni-
cal Journal, Vol. 25, No. 3, Aug. 1988, pp. 617-621.

26. Powell, J. J. M. and Uglow, I. M., "Marchetti Dilatometer Testing in U.K.

Soils", Proceedings, 1st International Symposium on Penetration Testing
(ISOPT-1), Vol. 1, Orlando, 1988, pp. 555-562.

27. Mayne, P. W., "Determining Preconsolidation Stress and Penetration Pore Pres-
sures from DMT Contact Pressures", Geotechnical Testing Journal, ASTM, Vol.
10, No. 3, Sept. 1987, pp. 146-150.

28. Mayne, P. W., "Determining OCR in Clays from Laboratory Strength", Journal of
Geotechnical Engineering, ASCE, Vol. 114, No. 1, Jan. 1988, pp. 76-92.

29. Wroth, C. P., "Penetration Testing- A Rigorous Approach to Interpretation",

Proceedings, 1st International Symposium on Penetration Testing (ISOPT-1),
Vol. 1, Orlando, 1988, pp. 303-311.

30. Mayne, P. W., "Cavity Expansion/Critical-State Theory for Piezocone Penetra-

tion in Clays", submitted for review in 1990.

31. Robertson, P. K., Campanella, R. G., Gillespie, D., and Greigs, J., "Use of
Piezometer Cone Data", Use of In-Situ Tests in Geotechnical Engineering (GSP
~. Ed. S. P. Clemence, ASCE, New York, 1986, pp. 1263-1280.

32. Mayne, P. W. and Bachus, R. C., "Profiling OCR in Clays by Piezocone Sound-
ings", Proceedings, 1st International Symposium on Penetration Testing (ISOPT-
1), Vol. 2, Orlando, 1988, pp. 857-864.

33. Marchetti, S., "In-Situ Tests by Flat Dilatometer", Journal of the Geotechni-
cal Engineering Division, ASCE, Vol. 106, No. GT3, Mar. 1980, pp. 299-321.

34. Powell, J. J. M. and Uglow, I. M., "Dilatometer Testing in Stiff Overconsoli-

dated Clays", Proceedings, 39th Canadian Geotechnical Conference, Ottawa,
1986, pp. 317-326.

1 3-37

I
35. Lacasse, S. and Lunne, T., "Calibration of Dilatometer Correlations", Proceed-
ings, 1st International Symposium on Penetration Testing (ISOPT-1), Vol. 1,
Orlando, 1988, pp. 539-548.

36. Bullock, P., "The Dilatometer: Current Test Procedures and Data Interpreta-
tion", Civil Engineering Report, University of Florida, Gainesville, 1983, 308
p.

37. Larsson, R., "Basic Behavior of Scandinavian Soft Clays", Report 4, Swedish
Geotechnical Institute, Linkoping, 1977, 125 p.

38. Brooker, E. W. and Ireland, H. 0., "Earth Pressures at Rest Related to Stress
History", Canadian Geotechnical Journal, Vol. 2, No. 1, Feb. 1965, pp. 1-15.

39. Mayne, P. W., "K0 -cuf0vo Trends for Overconsolidated Clays", Journal of Geo-
technical Engineering, ASCE, Vol. 110, No. 10, Oct. 1984, pp. 1511-1516.

40. Mayne, P. W. and Kulhawy, F. H., "Direct and Indirect Determinations of In-
Situ K0 in Clays", Research Record xxxx, Transportation Research Board, Wash-
ington, 1990, in press.

41. Kulhawy, F. H., Jackson, C. S., and Mayne, P. W., "First-Order Estimation of
K0 in Sands and Clays", Foundation Engineering: Current Principles and Prac-
tices, Ed. F. H. Kulhawy, ASCE, New York, 1989, pp. 121-134.

42. Schmertmann, J. H., DMT Digest No. 1, GPE Inc., Gainesville, 1983, 3 p.

43. Marchetti, S., "On the Field Determination of K0 in Sand", Proceedings, 11th
International Conference on Soil Mechanics and Foundation Engineering, Vol. 5,
San Francisco, 1985, pp. 2667-2672.

44. Durgunoglu, H. T. and Mitchell, J. K., "Static Penetration Resistance of

Soils", Proceedings, ASCE Specialty Conference on In-Situ Measurement of Soil
Properties, Vol. 1, Raleigh, 1975, pp. 151-189.

I
I
3-38

•I
 Section 4

STRENGTH

A knowledge of the strength of soils is necessary for most geotechnical analyses.

From a foundation engineering standpoint, the strength is necessary primarily to
evaluate the capacity. However, soil strength varies with many parameters, and
therefore it is not uniquely defined. In this section, basic definitions are pre-
sented first to establish the general background, notation, and relevance of the
str~ngth tests to field conditions. Then methods for estimating the effective
stress friction angle are presented, first for cohesionless soils and second for
cohesive soils. For each soil type, typical values, influencing factors, and
in-situ test correlations are presented. Finally, methods for estimating the
undrained shear strength are presented, including typical values, influencing
factors, and in-situ test correlations.

BASIC DEFINITIONS

The strength of soils commonly is expressed by the Coulomb-Mohr failure criterion,

as illustrated in Figure 4-1. For this criterion, failure is given by:

T = c + a tan ~ (4-1)

in which r =shear stress at failure (i.e., shear strength), c cohesion inter-

cept, a = normal stress, and ~ = friction angle.

Effective Stress Analysis

Although Equation 4-1 is the general form of the criterion, it is rarely appro-
priate to use the complete equation. Instead, the criterion is used in two alter-
native forms. First, when effective stress analyses of cohesionless or cohesive
soils are conducted, Equation 4-1 is expressed as:

r = u tan ~ (4-2)

in which u = effective normal stress and ~ effective stress friction angle, as

shown in Figure 4-2.

4-1
 ...
en
en
- en
en
....
(1) (1)
.!::::
(j) (/)
.... ....
0 0
(1) (1)
.c. .c.
(/) (/)

(j3 a-,
Normal Stress, a- Effective Normal Stress, a-
Figure 4-1. General Coulomb-Mohr Figure 4-2. Effective Stress Coulomb-
Failure Mohr Failure

No effective stress cohesion intercept (c) is shown because it occurs only in spe-
cial cases, such as with truly cemented soils, partially saturated soils, and heav-
ily overconsolidated clays, in which c is interpreted as gradually decaying with
time on an engineering time-scale. For these special cases, c could be included in
Equation 4-2. However, it is prudent to seek expert geotechnical advice before
considering use of c for design.

Many times, effective stress laboratory test data are interpreted incorrectly to
show a moderately high c and an unrealistically low ¢ because the true failure
envelope curvature is not being addressed. Figure 4-3 shows actual curved fail-
ure envelopes, with c = 0, for a wide range of soils from clay to rockfill. Linear
interpretation of any of these data over a limited stress range would suggest a c
and ~. but these values would not be the true soil strength parameters.

The friction angle of soils also varies with many other factors, as will be
described throughout this section. For a given soil at a constant normal effective
stress (u), the friction angle varies with density state and strain, as shown in
Figure 4-4. Expressing ~ in terms of the effective major and minor principal I
stresses (ul and u3, respectively) gives:

(UJ!U 3) f - 1 (u 1 - u 3) f
sin ~ (4-3)
{UJ/U3) f + l (al + a3) f

~
in which the subscript f represents failure conditions.

Different peak friction angles <¢p) develop as a function of soil density state.
At one limit is the very dense cohesionless soil or the heavily overconsolidated
I
4-2
I
I'
 ~c?.----·
60 Silty sandy gravel
0
a.
.....,__

Connonsville Hom river

""' sand o
40 ~?-
II)
II)

,_
\_Avonmo~t-~
Q)
+-
(f)
... ----··clay
,_
0 .. Chofa.hoochee river sand
Q)
..c 20 \Oo __
(f) London cloy · · · · · ·
..... -· · · undisturbed
Londoii·-~ioy
consolidated from slurry

20 40 60 80 100
Effective Normal Stress, <T/p0
Figure 4-3. Strength Envelopes for a Range of Soil Types

Source: Bishop (l), p. 104.

(very dense,
_Q OC}
0
0:::
en
en
Q)
.._
(/)

0
c..
(.)
c
·.::
a. ~Contractive
Q)
> (very loose, NC}
(.)

-
~
w
Strain, E

Figure 4-4. Friction Angle Definitions

cohesive soil which exhibits strongly dilative behavior during shear. For these
soils, the peak friction angle is high, and it develops at very small strains,
typically on the order of a few percent. At the other limit is the very loose
cohesionless soil or the normally consolidated, insensitive, uncemented, cohesive
soil, which exhibits contractive behavior during shear. For these soils, the peak
I friction angle is lower, and it develops at larger strains, typically upwards of 10
to 20 percent. The difference between these limits occurs because of the volume

I change behavior during shear (dilative to contractive). Different behavior is

I 4-3

L
noted for sensitive, cemented, and other structured cohesive soils, which normally
peak at small strains, much like the intermediate curve in Figure 4-4.

As a dilative soil is strained past its peak, it strain-softens to a limiting state

known as the fully-softened or critical void ratio state (~cv). The contractive
soil strain-hardens to reach the critical void ratio state, which also corresponds
to its peak friction angle. The critical state (~cv) typically occurs at strains
upwards of 10 to 20 percent. Therefore, regardless of the initial density state,
~cv is unique for a given soil at a constant normal effective stress.

With subsequent large straining in cohesive soils, typically in excess of 100 per-
cent, ~cv is gradually reduced to an ultimate limit known as the residual state
(~r)· The resulting ~r is commonly several degrees lower than ~cv· For cohesion-
less soils, ~r is essentially equal to ~cv· The residual state would be considered
in foundation engineering only for very large strain problems, such as siting in
soils containing pre-existing shear failures. Common examples would be landslide
debris or slopes in stiff-fissured clays.

Total Stress Analysis

i
The second use of Equation 4-1 is defined as the total stress (or ~ 0) analysis
of cohesive soils, given by:
II
!

r = c = Cu (4-4)
I
in which all four terms can be used interchangeably to represent the undrained
shear strength of the soil. This relationship is shown in Figure 4-5. Also in ,I
this figure, qu is defined as the unconfined compressive strength = 2 Su·

In many older references, the term "cohesion" was used to designate Su. In recent I
....
I
en
C/)
,_
Q)

<f)
.....
¢=0
I
0

I
Q)
.c
<f)
0 qu o-3 o-1

Total Normal Stress, o-

Figure 4-5. Total Stress Coulomb-Mohr Failure I
4-4
I
I
 references, Su is referred to as the undrained shear strength or undrained shearing
resistance. The older definition has led to much confusion and misinterpretation
with the effective stress cohesion intercept (c).

Total stress analysis normally is adopted for simplicity. In reality, the failure
of all soils (sands, silts, and clays) occurs on the effective stress envelope
shown in Figure 4-2. In low permeability soils such as clays, loading generates
changes in pore water stresses (bu). These pore water stresses change the effec-
tive stresses, which in turn influence the state of stress relative to the effec-
tive stress envelope. Since the total stress loading path and the magnitude of the
changes in pore water stresses may not be known with confidence, a total stress
analysis provides a simple analysis alternative. However, it must be remembered
that Su includes ~ and bu, and it varies with stress level in-situ. Therefore, Su
must be determined carefully to represent the in-situ conditions at a particular
depth, as described in detail later in this section.

Relevance of Laboratory Strength Tests to Field Conditions

The strength of soils can be measured by a number of different laboratory strength

tests, as noted previously in Figure 1-1. Each of these tests will give different
results because each subjects the soil to different boundary conditions and loading
stress paths.

In the field, different elements of soil also will be subjected to different boun-
dary conditions and loading stress paths. Figure 4-6 shows a number of common
field loading cases and the test types pertinent for each case. For an embankment
loading, the bearing capacity is represented most correctly by a combination of
compression (PSG or TC), direct simple shear (DSS), and extension (PSE or TE) tests
along the potential shear surface noted. For ease in computation, an average of
these three test types normally is used. With a loaded wall, the direct simple
shear and extension test types are averaged. With a vertical cut, the compression
test is most relevant.

When addressing foundations, different strengths are appropriate for different

field loading and behavior modes. Thes~ modes are described in detail by Kulhawy,
et al. (£). For a drilled shaft in compression, the tip resistance can be evalu-
ated from an average of the triaxial compression, direct simple shear, and triaxial
extension tests. The side resistance is modeled by the direct simple shear test up
to first yield or slippage along the interface, after which direct shear is more
appropriate. The results of these two tests are similar, so they commonly are used

4-5
 a) Embankment b) Loaded Wall

~ .
DSS
or TE

c) Vertical Cut
Compression Direct Simple Extension
Test (PSC
or TC)
Shear Test
(DSS)
Test (PSE
or TE) __Wesc o• TC

d) Dri lied Shaft

Comp. Uplift ~oment

tDSS/OS
TEJj

e) Spread Foundation
Uplift Uplift

~ e~ng
c

TC~E
DSS - -
TC/DSS/TE

DSS

Note: Plane strain tests (PSC/PSE) used for long features

Triaxial tests (TC/TE) used for near symmetrical features
Direct shear (OS) normally substituted for DDS to evaluate f,
Figure 4-6. Relevance of Laboratory Strength Tests to Field Conditions
I
where they are best-suited, the DS being used for sands with the DSS being used for
clays. For a shaft in uplift, the side resistance is the same as in compression.
I
For lateral or moment loading, triaxial extension is more appropriate.

For spread foundations in compression, the same bearing capacity approach is used.
In uplift, the behavior can range from the normal situation of a vertical shear
surface to a vertical shear with cone breakout to a punching limit controlled by
bearing capacity. As noted in the figure, the shear case is given by the DSS. The
i'
i'
cone case is an average of TG and DSS. The punching is evaluated using an average
'
4-6
 of TC, DSS, and TE.

The various tests pertinent for a particular field condition are likely to be an
excessive requirement for common and routine design cases. Therefore, it is more
convenient to establish a standard "test of reference" which would be appropriate
for many design cases, and which would be simple and expedient from a commercial
testing standpoint. The recommended test (e.g., Wroth, l) is the isotropically
consolidated, triaxial compression test for undrained loading (CIUC) and for
drained loading (CIDC). Using the results of this test as a standard reference,
the results of all other tests can be compared simply and conveniently.

It should be noted that most soils in-situ actually will be consolidated aniso-
tropically. This difference in consolidation stresses has no appreciable influence
on the soil friction angle (~). However, it does influence the evaluation of the
undrained shear strength, as will be shown later.

EFFECTIVE STRESS FRICTION ANGLE OF COHESIONLESS SOILS - GENERAL EVALUATION BASIS

Correlations for estimating the effective stress friction angle for cohesionless
soils have been presented by numerous authors. Representative relationships are
given below.

Typical Values

Early work on this topic suggested simplified tabulated values for the effective
stress friction angle, such as those given in Table 4-1. Although never stated
explicitly, it is probable that these values refer to peak values measured in
triaxial compression tests (~tc)· Tabulated values such as these only establish
the general order of magnitude for ~tc· They should not be used for design.

Correlations with Index Parameters

Subsequent approaches have correlated the value of ~tc with one or more soil index
parameters, such as soil type, relative density, and unit weight or void ratio.
Figures 4-7 and 4-8 show two common relationships for estimating ~tc from soil
index parameters. Figure 4-7 refers specifically to ~tc from triaxial compression
tests on soils composed of hard minerals, at stress levels typical of those used in
footing design. Figure 4-8 is a more general relationship based on the groups in
the Unified Soil Classification System and presumably also refers to ~tc· Although
these figures address more of the variables, they still are simplifications of
actual behavior and tend to be somewhat conservative.

4-7
 Table 4-1

REPRESENTATIVE VALUES OF ~tc

~tc (degrees)
Soil Material Loose Dense

Sand, round grains, uniform 27.5 34

Sand, angular grains, well-graded 33 45

Sandy gravels 35 50

Silty sand 27 to 33 30 to 34

Inorganic silt 27 to 30 30 to 35

Source: Terzaghi and Peck(~), p. 107.

~
1-&
Q)

0"
c
<(
c

-
.Q
u
..._
35°

lJ..

0 20 40 60 80 100

Figure 4-7.
Relative Density, Dr (%)
~tc versus Relative Density
I
Source: Schmertmann (~), p. 41.
II
Influence of Strength Envelope Curvature

Table 4-1 and Figures 4-7 and 4-8 imply that the soil failure envelope is linear,
although data such as that in Figure 4-3 show that the failure envelopes normally
are nonlinear. This nonlinearity is well-established in the literature (e.g., l,
z, ~) and is attributed to soil dilatancy. This dilatancy increases with increas-

'
ing relative density and decreases with increasing stress level.

4-8
I
J
"
 45°
--
1-B- -
0

<1)

0'
c
<!
c
0
.....
u
"- For cohesionless
lL soils without
plastic fines
25°
1.2 1.4 1.6 1.8 2.0 2.2 2.4
Dry Unit Weight, yd /yw

1.2 1.0 0.8 0.6 0.4 0.2

Void Ratio, e for Gs=2.68
Figure 4-8. ~tc versus Relative Density and Unit Weight

Source: NAVFAC (~), p. 7.1-149.

The most convenient way to include the strength envelope curvature is to use secant
peak friction angles which vary with stress level, as illustrated in Figure 4-9.
By taking successive secants through the origin at varying normal stresses, the
values of ~secant with normal stress can be obtained. Loose soils approximate ~cv
and exhibit an essentially linear envelope.

It should be noted at this point that the soil behavior illustrated in Figures 4-3,
4-4, and 4-9 is general and that the same patterns will develop regardless of the
laboratory test type. From this point forward, it will be presumed that the fric-
tion angle given represents a peak value obtained as a secant to the failure

c
0
u
"'
"'
1-$-

--\_Loose - - - - -

Effective Normal Stress, a-

-
Figure 4-9. Nonlinear Failure Envelope Representation

4-9
 envelope. For clarity, the subscripts to be used will refer only to the test type,
such as ~tc for peak secant friction angle in triaxial compression. No test desig-
nation is needed for the critical void ratio friction angle (~cv) because this
value is unique and independent of test type (e.g.,~.~). The same is true for
the residual friction angle (~r).

Recent work by Bolton (~) has unified much prior research in a convenient way, uti-
lizing critical state concepts and a data base primarily of clean sands. This work
demonstrated that the dilatancy component of the friction angle can be estimated as
follows:

(¢psc ~cv) = 5 IRD for plane strain compression (4-5)

3 IRD for triaxial compression (4-6)

in which IRD is a relative dilatancy index, given by:

IRD = Dr[Q - ln(lOO Pf/Pa)J - R (IRD > 0) (4-7)

In this equation, Dr= relative density, Q = soil mineralogy and compressibility

coefficient (10 for quartz and feldspar, 8 for limestone, 7 for anthracite, 5.5 for
chalk), Pf =mean principal effective stress at failure [(ol + o2 + o3)f/3), Pa

~
atmospheric stress in the same units as Pf, and R = fitting coefficient (equal to 1
for the evaluated test conditions and data). Figure 4-10 illustrates this rela-
tionship for eight different quartz and feldspar sands. The equation noted on the
figure would be typical of triaxial compression tests on silica-type sands. The Ii

•I
relative dilatancy index (IRo) should be limited to 4 unless detailed laboratory
test data indicate otherwise.

Equation 4-7 unfortunately relates to the mean principal effective stress at fail-
ure, a parameter which includes the initial stress state, stress path to failure,
test conditions, and foundation type. For preliminary estimating purposes, Pf can
be assumed to approximate two times avo• which should lead to a computed (~ - ~cv)
within 1 to 2 degrees of the actual value for most cases. For final design, the
I
I
value of Pf corresponding to the specific foundation conditions should be used.

To estimate the value of ¢cv• Koerner's work (10) on single mineral soils can be
I
'I
considered, which led to the following:

4-10

I
 16°
>
c.>
•-& ''
~ 12°
•-&
c-
Q)
o Dr~o.a
(6 sands)
c
0
0. • Dr::::0.5
E
0 (6sonds)
u
>-
u
c
.E
.!2
0
0
0.1 10 100 1000
Mean Principal Effective Stress at Failure, pf /p0

Figure 4-10. Dilatancy Angle Relationships

Source: Bolton(~). p. 73.

tPcv (4-8)

in which: b.¢>1 correction for particle shape

b.¢>1 -6· for high sphericity and subrounded shape
b.¢>1 +20 for low sphericity and angular shape
b.¢>2 correction for particle size (effective size, d1o)
M2 -n· for dlQ > 2.0 nun (gravel)
M2= -9· for 2.0 > d1o > 0.6 (coarse sand)
b.¢>2 -4· for 0.6 > dlQ > 0.2 (medium sand)
M2 = 0 for 0.2 > d1o > 0.06 (fine sand)
b.t/>3 correction for gradation (uniformity coefficient, Cu)
b.¢>3= -2· for Cu > 2.0 (well-graded)
b.¢>3 -10 for Cu = 2.0 (medium graded)
b.¢>3 = 0 for Cu < 2.0 (poorly graded)
b.t/>4 = correction for relative density (Dr)
b.¢>4 = -1· for 0 < Dr < 0.5 (loose)
b.t/>4 = 0 for 0.5 <Dr< 0.75 (intermediate)
b.t/>4 +40 for 0.75 <Dr< 1.00 (dense)
A¢5 correction for type of mineral
M5 = o for quartz
b.t/>5 = +40 for feldspar, calcite, chlorite
A¢5 = +60 for muscovite mica

4-ll
Current understanding (e.g., ~) is that ~cv is essentially independent of relative
density, and therefore the relative density correctio~ (6~4) should be set equal to
zero. Relative density primarily influences the dilatancy component. Equation 4-8
also must be kept within the context of Bolton's work (~) on natural soils, which
showed that ~cv z 33o for representative quartz sands and ~cv z 40° for representa-
tive feldspar sands. However, most natural deposits of sand include silt. There-
fore, Bolton concluded that ~cv for most natural sand deposits rarely will be much
above 30° to 33°, and may be as low as 27o when the silt content is high.

Influence of Test Boundary Conditions

For simplicity, most analyses assume that the peak, secant, effective stress fric-
tion angle is independent of direction of loading, and therefore the intermediate
effective principal stress (o2) is disregarded. However, this influence can be
important in some loading cases. To evaluate this effect, the intermediate effec-
tive principal stress factor (b) can be defined as:

(4-9)

Normalized test data on five sands are shown in Figure 4-11 to illustrate the
importance of b. The mean and range are shown for both the loose and dense sands.
For plane strain compression (b = 0.3 to 0.4), the increase ranges from 7 to 18
percent with an average on the order of 12 percent. For triaxial extension (b =
1), the increase ranges from 0 to 23 percent, again with an average on the order of
12 percent. A similar increase should be expected when comparing plane strain

'
extension to compression.
'
Other studies (e.g., 2) have shown that the plane strain compression (PSG) and
direct shear (DS) tests can be interrelated as follows:
I
tan ~ds - tan ~psc cos ~cv (4-10) I
For typical ranges of ~cv• the PSG values from this equation will be some 2 to 7
degrees higher than the direct shear values, corresponding to increases from 4 to
19 percent. •I
Comparison of the direct shear values from Equation 4-10 with the triaxial compres-
sion values from Equations 4-5 and 4-6 indicates that the triaxial compression val-
ues may be larger or smaller than the direct shear values, depending on the values
I
4-12
I
I
 Triaxial Triaxial
compression extension

-0
II
.0
u
li:)- ·-e: 1.2

-
.2
a::
0
I. I
Q)
c;.
c
<{
Loose sands
c 1.0
-....
0
u
LL

(7"2- (7"3
b = ---:::-"-----:_;.,.
cr,- cr3
Figure 4-11. Influence of Intermediate Principal Stress on Friction Angle

Source: Data from Ladd, et al. (11), p. 431.

of ¢cv• relative density, and stress level.

Table 4-2 summarizes the relationships for friction angle as a function of test
type. As can be seen from this table and Figure 4-6, use of the triaxial compres-
sion friction angle <¢tc) alone will almost always be a conservative assumption.

EFFECTIVE STRESS FRICTION ANGLE OF COHESIONLESS SOILS CORRELATED WITH IN-SITU TESTS

At the present time, correlations of the effective stress friction angle have been
made with the standard penetration test (SPT), cone penetration test (CPT), pres-
suremeter test (PMT), and dilatometer test (DMT). The CPT correlations are perhaps
the best-developed, followed by the SPT. The PMT correlations are newer and less
developed, while the DMT correlations are of limited use at this time. In all
cases, it is presumed that the correlations use the triaxial compression friction
angle (~tc) corresponding to the appropriate stress andjor relative density condi-
tions.

Correlations with SPT N Value

Correlations of the effective stress friction angle with the SPT N value have been
made for many years. Early work on this subject attempted to relate N to ¢tc

4-13
 Table 4-2

RELATIVE VALUES OF EFFECTIVE STRESS FRICTION ANGLES FOR COHESIONLESS SOILS

Test Type Friction Angle (degrees)

Triaxial compression (TC) l.O;j,tc

Triaxial extension (TE) 1.12 :j, tc

PLane strain compression (PSG) 1.12 tP tc

PLane strain extension (PSE) 1.12 (for PSC/TC) x 1.12 (for TE/TC)
= 1. 25 :j, tc

Direct shear (DS) tan-1 [tan ¢psc cos ¢cvl

or tan-1 [tan (1.12 tfttc) cos tPcvl

directly, as shown in Table 4-3. The Peck, et al. (12) approach appears to be more
common, perhaps because it is more conservative. These values also are shown in
Figure 4-12.

As discussed in Section 2, theN value actually depends upon stress level. Figure
4-13 is representative of the correlations between Nand 4>tc as a function of
stress level. This correlation can be approximated as follows:

(4-11)

These results tend to be somewhat conservative and should not be used at very shal- I
low depths, less than 1 to 2 rn (3.3 to 6.6 ft). Improved correlations with the
other variables described in Section 2 have not been developed to date.

'
~
Correlations with CPT q c Value

Similarly, correlations of tPtc with cone tip resistance, qc, have been developed.
Early work attempted to correlate qc to 4>tc directly, as shown in Table 4-4.

As described in Section 2, qc is affected by the vertical stress. Therefore, 4>tc I

should be correlated to both qc and avo, as shown in Figure 4-14. This correlation
I
4-14
I
 Table 4-3

N VERSUS ~ te RElATIONSHIPS

.1\pproxima te <Pte (degrees)

N Value Relative
(blows/ft or 305 mm) Density (a) (b)

0 to 4 very loose < 28 < 30

4 to 10 loose 28 to 30 30 to 35

I 10 to 30 medium 30 to 36 35 to 40

30 to so dense 36 to 41 40 to 45

I > ~ very dense > 41 > 45

a - Source: Peck, Hanson, and Thornburn ( 12) ' p. 310.

I b - Source: Meyerhof (13), p. 17.

<b
~
I .§
~
~<b
o<b
,.::....
~<b

I
E
E
O.>l{)
2o
-j;r0 20
.....
z
I
0

t-:= 40
a..--..
(/)V>
~
0

I 60
..0

28° 32° 36° 40° 44°

Friction Angle, <f>tc

I Source:
Figure 4-12. N versus <Pte

Peck, Hanson, and Thornburn (12), p. 310.

I
can be approximated as follows:
I ~te ""ta.n-1 (0.1 + 0.38 log (qc/Ov0 )] (4-12)

I Adjustments to this figure and equation for soils of different compressibility and
stress history should be made as described in Section 2.
I
I 4-15

L
 SPT N Value, Blows/ft or 305mm
0 20 40 60
0

0
0.
.......
0
It)
Cl)
Cl)
Q)
.:::
(/)

-~8 2

Figure 4-13. N versus ~tc and Overburden Stress

Source: Schmertmann (14), p. 63.

Table 4-4

qc VERSUS ¢> tc

Normalized Cone Tip Relative Approximate i>tc

Resistance, qdPa Density (degrees)

<20 very loose < 30

20 to 40 loose 30 to 35

40 to 120 medium 35 to 40

120 to 200 dense 40 to 45

> 200 very dense > 45

Source: Meyerhof (13)' p. 17.

Villet and Mitchell (16) presented a more general approach to evaluating if>tc from
CPT data which includes qc, stress level, shape effects, and soil stress history.
'
Their results are shown in Figure 4-15 and are suitable for low compressibility
sands. I
4-16 I
I
 Cone Tip Resistance, Q/P0
200 400

0
c..
.......
0
•t:'
.n
"'....
Q)

Vi
Q)
.2! 2
u
~
w
0
~
.... 3
Q)
>

Figure 4-14. qc versus ~tc and Vertical Stress for NC, Uncemented, Quartz Sands

Source: Robertson and Campanella (15), p. 726.

Cone to sand Cone to sand

o/cf>tc=0.5

"

r:::r
>-..
z
..:
.£
(.)
0
LL.
Q)
c
0
u
3 (b)
10
30°
-
Friction Angle, cf>tc Friction Angle, cptc

Figure 4-15. ~tc from CPT Data

Source: Villet and Mitchell (16), p. 193.

4-17
Using the standard cone diameter (B) of 35.7 mm, Marchetti (17) reworked the data
in Figure 4-15 to result in the more simplified Figure 4-16. Consistent with the
development in Section 2 which related relative density to the normalized cone tip
resistance, a similar correlation has been developed from 20 data sets obtained in
calibration chambers and is shown in Figure 4-17. Mineralogy, particle shape, com-
pressibility, and percent fines largely account for the observed range of ~tc at
any normalized qc value.

Correlations with PMT Results

The results obtained from pressuremeter tests also can be correlated with the
effective stress friction angle, using procedures developed by either Schmertmann
(14) or Hughes, et al. (18). The Hughes, et al. approach is presented below.

In a pressuremeter test, the basic data obtained are the expansion stress (pe) and
the volume changes (~V) in the pressuremeter of known volume (V). The resulting
data can be plotted as shown in Figure 4-18a, using the cavity strain (Ec) which is
defined as the change in membrane radius divided by the initial radius and is given
by:

(4-13)

in which EV = 6VfV volumetric strain. These data then are re-plotted as in

0.5

0.2

10 20 50 100 200 1000 2000

Cone Tip Resistance, qc/crvo

Figure 4-16. Simplified qc - K0 - ~tc Relationships

Source: Marchetti (17), p. 2668.

4-18
I
I
 - 05 ~
,~...t C =17.6+11.0 log[(q C /p0 )/(0'YO /p)
'f' 0
· ] ~~
(n=633,
2
r =0.640, S.D. =2.8°) ~0 ~
~ 0
® 0

0 •

Solid Symbols- OC Sands

Other Symbols- NC Sands
(as given in Appendix H)

10 20 50 100 200 500 1000

0 5
Cone Tip Resistance, ( qc/Pa) I ( crv 0 /pa ) ·

Figure 4-17. Trend of ¢> tc with Normalized qc

0 ~7
0.. .......
(a l Typical Plot 5 (b) Normalized Plot
""'<u6 -0
::;1
0.. I
.,; Q)4
VI
0..
Q)
...
(/)
4 .,; 3
VI
Q)
....
c (/) 2
0
·;;; c

I
c 0
·;;;
0
0.. c
>< 0
I
w 0 0..
0 4 8 12 16 >< 0.4 0.6 2 4 6 10 16
w

I Cavity Strain, Ec (%)

Figure 4-18. PMT Data Representations

Cavity Strain, Ec (%)

I Source: Mair and Wood (19), p. 76.

I Figure 4-18b, subtracting the initial pore water stress at the pressuremeter
level. The resulting log-log plot is essentially linear with a slope, s.

I By considering cylindrical cavity expansion theory, s can be given by:

I 4-19

I
.....
 s - sin ~cv (1 + sin ~)/(1 + sin ~cv) (4-14)

in which ~cv - critical void ratio friction angle and ~ = dilation angle (~tc
~cv• as described previously). Equation 4-14 can be rearranged to give:

sin ~ = s (1 + sin ~cv)/sin ~cv (4-15)

Therefore, by re-plotting the PMT data to give s and estimating ~cv as described
previously, the friction angle (~tc) can be obtained. Figure 4-19 provides a
graphical procedure to evaluate ~tc• using Bolton's (~) approximation that:

~cv + 0.8 ~ (4-16)

~tc

Correlations with DMT Results

Recently, a correlation also has been presented between the effective stress fric-
tion angle and the thrust pressure (tip resistance) on the dilatometer during pene-
tration. Using the Durgunoglu and Mitchell (20) theory, Schmertmann (21) showed
that the dilatometer tip resistance (qo). obtained from thrust measurements during
penetration of the blade, could be related to the cone tip resistance (qc) and the
effective stress friction angle (~psc) under plane strain compression. This

u
~I
•-& I

I
c
g
u I
I
s
o log (pe-u 0 )
= o log €c
I
Figure 4-19. Friction Angle Evaluation from PMT Results I
Source: Mair and Wood (19), p. 78.

4-20 I
I
 I
I relationship is given below:

I ~psc - 25(2.3 - qofqc) (4-17)

To evaluate ~psc from the DMT results, an iterative process is necessary. An ini-

I tial estimate is made of ~psc for triaxial compression conditions, from which an
equivalent qc is determined from Figure 4-15 or 4-16. Using this qc and the qD
measurement, ~psc is computed from Equation 4-17.
I This plane strain ¢psc then is
converted to an equivalent triaxial ¢tc using the relationships shown in Figure
4-11 or Table 4-2. The final ~tc is compared with the initial assumption. If they

I agree, then ¢psc is correct. Otherwise, iteration must be done until the initial
estimate and final value converge. At the present time, the DMT versus ¢psc corre-

I lation should be considered only as a first order approximation until sufficient

field confirmations become available.

I EFFECTIVE STRESS FRICTION ANGLE OF COHESIVE SOILS

Correlations for estimating the effective stress friction angle for cohesive soils

I have focused on only two areas: (1) the friction angle for normally consolidated
(NC) and remolded clays, which will approximate ¢cv• and (2) the residual friction
angle (~r). No generally accepted procedure has been presented for estimating the
I peak friction angle of overconsolidated (OC) clays as a function of overconsolida-
tion ratio (OCR) and other controlling factors, although the behavior should be

I qualitatively similar to that for cohesionless soils. Similarly, no generally

accepted correlations have been presented with in-situ test results.

I Correlations with Critical Void Ratio Friction Angle

As described at the beginning of this section, the peak friction angle for insensi-

I tive, uncemented NC cohesive soils basically is equal to the critical void ratio
friction angle <¢cv). For sensitive, cemented, or other structured NC cohesive

I soils, ~cv will represent a lower bound for the peak friction angle. For OC soils,
remolding will destroy the stress history and therefore result in "newly-created NC
soil", with the friction angle being given by ¢cv· Other complex factors such as

I leaching, sensitivity, stress state, etc. influence this simple explanation to some
degree. However, first-order correlations can be made using this simple approach.

I Many authors have shown that ¢cv can be correlated with simple index parameters
such as the plasticity index. One such relationship is presented in Figure 4-20,

I which shows that ¢cv decreases with increasing plasticity index and increasing clay

I 4-21

L
 1.0
o Undisturbed soil
0.8 "' Remolded soil
cf>cv
40°
>
(.) 0.6
•-e-
c:
·v; 0.4
20°

0.2 10°

0
6 8 10 15 20 30 40 50 60 80 100 150 200
Plasticity Index, PI (%)
Figure 4-20. ~cv for NC Clays versus PI

Source: Mitchell (22), p. 284.

mineral activity (kaolinite~ illite~ montmorillonite). This general trend has

been corroborated by others (e.g., 11, 23). However, it should be noted that the
error band with this correlation is fairly large.

Influence of Test Boundary Conditions

Laboratory testing conditions can influence the friction angle of NC clays. The

•I
data in Figure 4-20 were obtained largely from isotropically consolidated,
undrained triaxial compression (CIUC) tests with pore water stress measurements.
In-situ, the initial stresses would correspond to anisotropic consolidation (CAUC),
most commonly restricted to K0 consolidation (CK 0 UC). Fortunately, comparative
studies such as that shown in Figure 4-21 have demonstrated that ~tc essentially is
the same, regardless of initial consolidation state. Although the regression shows
a small variation from equality, this variation is small and can be ignored. I
However, the same can not be said for other testing conditions.
compression, Wroth (1) suggested analytically that ~ps<~
For plane strain
would be approximately 9/8
I
times ~tc· Figure 4-22 illustrates that this relationship is satisfactory, al-
though the regression gives a slightly lower value.
for sands.
This value is similar to that
Figure 4-23 compares the friction angles for NC clays in extension and
I
compression. As can be seen, ~te always is equal to or greater than ~tc and, on
the average, ~tel~tc = 1.22. This average is the same for both anisotropic and
I
isotropic test conditions, even though their statistics differ a small amount.
Additional limited data (26) show that the pattern should be similar for plane
strain conditions as well.
I
4-22 I
I
 I
I
I 40°

u
I ::::>
<l
u
"0
30°
c

I
0

u
::::>
0 20°
:::c:::
u
I I-s:
u

I 4 7 intact clays

I ~tc (CIUC)
Figure 4-21. ~tc Variation as a Function of Consolidation Stress for NC Clays
I Source: Data from Mayne (24) and Nakase and Kamei (25).

I
I
I
I
I
I 12 intact cloys

I
Figure 4-22. ~psc versus ~tc for NC Clays

I Source: Data from Mayne and Holtz (26).

I 4-23

L
 e CK 0 U/CAU (n=37, r 2=0.730, $.0.=5.6°) 0

0 CIU (n=33, r 2 =0.843, 5.0.=5.0°) #

0
0 10° 20° 30° 40° 50°

f>tc ·.I
i

Figure 4-23. ~te versus ~tc for NC Clays

Source: Data from Mayne and Holtz (26) and Nakase and Kamei (25). I
Table 4-5 summarizes the relative values of the friction angle for the different I
testing conditions. Although no detailed comparisons have been presented for the
direct shear test, the results should exhibit patterns similar to those presented
earlier for cohesionless soils. Therefore, the same relationship is proposed for
I
cohesive soils. It should be noted that use of ~tc alone will almost always be a
conservative assumption. :I
Correlations with Residual Friction Angle
I
As described earlier in this section, the residual friction angle (~r) develops
when a cohesive soil undergoes very large strains, and the soil structure is
totally remolded and re-oriented into a minimum strength orientation. Currently,
I
4-24 I
 Table 4-5

RELATIVE VALUES OF EFFECTIVE STRESS FRICTION ANGLE

FOR NORMALLY CONSOLIDATED COHESIVE SOILS

Test Type Friction Angle (degrees)

Triaxial compression! (TC) 1.0 ~tc

Triaxial extension (TE) 1.22 ~tc

Plane strain compression (PSG) 1.10 ¢tc

Plane strain extension (PSE) 1.10 (for PSC/TC) x 1.22 (for TE/TC)
= 1.34 ~tc

Direct shear2 (DS) tan-1 [tan ¢psc cos ¢cvl

or tan-1 [tan(l.lO ~tc) cos ~cvl

1 - CIUC, CK 0 UC, or CAUC

2 - Speculative, based on results from sand

it is understood that the strains necessary to accomplish this remolding may exceed
100 percent. Earlier studies of this subject may not have subjected the soil to
the necessary strains, and therefore residual angles quoted in earlier sources may
be somewhat on the high side.

Extensive research (e.g., 27, 28) has shown that the clay fraction (percent finer
than two microns) and mineralogy perhaps are most important in evaluating ~r· If
the soil clay fraction is less than about 15 percent, the soil behaves much like a
cohesionless soil, with ~r typically greater than 25° and not much different from
~cv· If the clay fraction is greater than 50 percent, ~r is appreciably lower than
~cv and is governed entirely by sliding of the clay minerals. For the most common
clay minerals, ~r ranges approximately from 15° for kaolinite, to 10• for illite,
and then to 5• for montmorillonite. Soils with clay fractions between 15 and 50
percent exhibit transitional behavior, as shown in Figure 4-24.

The value of ~r also is stress-dependent because of curvature of the failure enve-

lope (22, 27, 29). Values given in Figure 4-24 are appropriate for an effective
normal stress equal to about one atmosphere. Figure 4-25a illustrates the typical
changes in ~r which occur with changes in effective normal stress and plasticity

I 4-25

I
 . ---
300 ·~ _
"\_ "-
_!2
" ' y R a n g e for 8 f1eld sites
with PI/CF =0.5 to 0.9
c: Sands \. \.

-0

.~
0
\.
'
'
tJ: '~ ......_"'o-- CO-- 0 +--Kaolin
0
::l
'0- 0 OCJ:)- - -8
"0 10° ----0---
Ill
<l>
0::
Values of ¢r at O:v /p0 ::::: I Bentonite ---•

60
Clay Fraction, CF (%}
Figure 4-24. ¢r from Ring Shear Tests and Field Studies

Source: Skempton (28), p. 14 .

.-;:;: c:Tip 0 =0.2

Q)
0.5
0'
c: I
<! zoo 2
c:
.2
u
....
lL

0
::>
-o
If) (a)
Q)
0::

a)
c;.
c:
<{
60
• I
Plasticity Index, PI(%)

I I I
•I
c:
.Q
u
lt zo \. -
I
gl-e:
~<J
Q)
0 ~. -
-

I
0:: -
.5
Q)
0'
c:
0
-zo
(b)
"---------· - i

I
.c I I I I
u 0.5 1.0 1.5 2.0 2.5

Effective Normal Stress, c:i I p 0

Figure 4-25.

Source:
~r for Amuay Soils

Based on Lambe (29), p. 144.

I
4-26 I
J
 index for the soils at the Amuay landslide sites. These curves essentially are
parallel, indicating that the change in ~r as a function of stress change is inde-
pendent of the plasticity index. Re-plotting these changes in friction angle (6~r)
results Figure 4-2Sb. Other data (e.g., 27) are consistent with these 6~r values.

The final values of ¢r therefore should be evaluated from Figure 4-24, modified for
effective normal stress level as given in Figure 4-2Sb.

UNDRAINED SHEAR STRENGTH OF COHESIVE SOILS - GENERAL EVALUATION BASIS

The undrained shear strength (su) may very well be the most widely used parameter
for describing the consistency of cohesive soils. However, su is not a fundamental
material property. Instead, it is a measured response of soil during undrained
loading which assumes zero volume change. As such, Su is affected by the mode of
testing, boundary conditions, rate of loading, confining stress level, initial
stress state, and other variables. Consequently, although not fully appreciated by
many users, Su is and should be different for different test types (See Figure l-1
for test types.).

As described earlier in this section, it is appropriate to use a standard "test of

reference", which is the isotropically consolidated, triaxial compression test for
undrained loading (CIUC). With the CIUC test as a standard reference, the results
of all other tests can be compared simply and conveniently. It should be noted
that simpler forms of triaxial test are available, such as the unconsolidated,
undrained (UU) triaxial and unconfined (U) compression tests. With the UU test, a
total confining stress is applied, but no soil consolidation is allowed under this
confining stress. With the U test, the soil is unconfined with a zero confining
stress.

Many detailed studies (e.g., 11, 23) have shown that the UU and U tests often are
in gross error because of sampling disturbance effects and omission of a reconsoli-
dation phase. Based on studies such as these, the CIUC test also is considered to
be the minimum quality laboratory test for evaluating the undrained shear strength

I of cohesive soils. Other simple tests such as the torvane and pocket penetrometer
have an error potential that is comparable to that of the UU and U tests. There-
fore, these tests should only be considered general indicators of relative beha-

I vior. They should never be used directly for design.

I Since su is stress-dependent, its value commonly is normalized by the vertical

effective overburden stress (av 0 ) at the depth where 5u is measured. This

I 4-27

I
undrained strength ratio, sufavo• has been expressed in many alternate forms in the
literature, including Su/ao, eu/Uo, eu;-ov. cjp, etc. All are equal to Su;-OVo·
which will be used in the remainder of this section.

Correlations with Index Parameters for Undisturbed Clays

Early work by Skempton (30) suggested the general correlation in Figure 4-26 for su
determined from the field vane shear test (VST) as a function of the plasticity
index. All of the data are for normally consolidated (NC) clays. A linear fit of
these data results in:

su(VST)/avo = 0.11 + 0.0037 PI (4-18)

In general, this relationship has been corroborated by others (e.g., 31), but there
usually is more spread in the data than that shown in Figure 4-26. Recent work by
Chandler (32) suggests that this approximation may also be valid for OC clays,
using the modification below with the preconsolidation stress (ap):

su(VST)/ap z 0.11 + 0.0037 PI (4-19)

He notes that the accuracy of this method will be on the order of ± 25 percent, but
he cautions against its use in fissured, organic, sensitive, or other unusual clays.

However, in a surprisingly large number of case histories, direct use of su from

the field VST in stability analyses of numerous embankments, excavations, and foot-
ings in clay has led to failures. Back-analysis of these failures has led to

0
0.6 s
u
{VST)

O"vo
=0.11+0.0037 Pil •
•t?
........
1-
(j)
>
0.4
........
·-. ~·
~·
.....
(/)
:::J 0.2

0
0
-• 20 40 60 80 100 120
Plasticity Index, PI (%)
Figure 4-26.

Source:
su(VST)/av0 versus PI for NC Clays

Skempton (30), p. 306.

I
4-28 I
empirical correction factors for the field VST. These factors will be described
later in the section on Su correlations with the VST.

Subsequent studies (e.g., 33) showed that sensitive clays with high liquidity index
did not fit the trend in Figure 4-26 very well. For these sensitive clays, the
undrained strength ratio could be correlated better with the liquidity index, as
shown in Figure 4-27. These data were obtained from triaxial compression tests on
NC clays.

The undrained strength ratio for triaxial compression also can be determined from
Critical State Soil Mechanics (CSSM) using the modified Cam clay model (e.g., 34).
For NC clay, this relationship is given by:

0.129 + 0.00435 PI (4-20)

in which oi = effective overburden stress after isotropic consolidation.

Other useful approximations include the following for low OCR clays with low to
moderate PI (Jamiolkowski, et al., 35):

su/Op = 0.23 + 0.04 (4-21)

in which up = preconsolidation stress. Alternatively, Mesri (36) suggested the

following:

>- 0.4
0
<..)

<..)
z 0.3
'-
0
._
.....
\.a •

. .-------·• •• .• • • ••·-
,,.......-'!'
u
:::>
.....
~ 4i
0 0.1
tb>
........
:=I
(/)
0
0 2 3 4

Liquidity Index, LI
Figure 4-27. su/Uvo for NC Clay versus Liquidity Index

Source: Bjerrum and Simons (33), p. 722.

4-29
 I
Su/O'p - 0.22 (4-22)
I
In both cases, the su corresponds approximately to direct simple shear (DSS) condi-
tions. I
Correlations with Index Parameters for Remolded Clays

The sensitivity (St) is defined as Su in the undisturbed state divided by su when

I
remolded (both tested normally in unconfined compression at the same natural water
content), and therefore it is a measure of strength loss upon disturbance. Table
I
4-6 gives the typical terminology used to describe sensitivity, while Figure 4-28
illustrates a generalized relationship for sensitivity as a function of liquidity
index and effective stress. The undrained remolded strength represents the lower
I
bound on su and, when St approaches one, Su z sur·
I
Figure 4-29 indicates that sur correlates reasonably well with the liquidity
index. Data on undisturbed natural clays of low sensitivity are presented in
Figure 4-30 and indicate good agreement with Figure 4-29, suggesting that sur is a
I
fair predictor of su for many clays of low sensitivity.
I
The undrained shear strength for triaxial compression also can be predicted from
the modified Cam clay model as follows (Wroth and Wood, 38):

ln S - (1 - LI) ln R (4-23)

Table 4-6

CLASSIFICATION OF SENSITIVITY

Clay Description St Clay Description St

Insensitive z 1 Slightly quick 8 to 16

Slightly sensitive 1 to 2 Medium quick 16 to 32

Medium sensitive 2 to 4 Very quick 32 to 64

Very sensitive 4 to 8 Extra quick > 64

Source: Mitchell (22), p. 208.

4-30
 .......
-'
.
><
<l)
"0
c:
.......
+-
>-
"0 0.5
::l
g
-'

10 100

Effective Vertical Stress, O:v 0 1P0

Figure 4-28. General Relationship Between Sensitivity, Liquidity Index, and

Effective Stress

Source: Mitchell (22), p. 229 .

.
><
Q)
"0
c:
.......

0.0001

Remolded Undrained Shear Strength, su/Po

Figure 4-29. Remolded Undrained Shear Strength versus LI

Source: Mitchell (22), p. 228.

4-31
 .......
..J
x
Q)
"0
c:
.......

-:;:...

:'2
:::l
0.4

0'"
:.J
0

-0.4~--~--~~~~ww~--~--~~~~L---~~~
0.05 0.1 0.5 50
Undrained Shear Strength, Su /p 0

Figure 4-30. Undisturbed Undrained Shear Strength versus LI

Source: Wood (37), p. 7.

in which S
(su at wL) z 0.017 Pa• yielding:

su/Pa- 1.7 e-4.6 LI (4-24)

This equation is plotted as the straight line in Figure 4-30 and shows good agree-
ment with the data in the range 0.1 < sufpa < 3.

General Behavior Under Triaxial Compression Loading

The undrained strength ratio in triaxial compression can be expressed in terms of

more fundamental soil parameters by analysis of the Coulomb-Mohr failure envelope
geometry. For K0 consolidation, the undrained strength ratio is given as:

K0 )] sin ~tc
(su/0 vo)CAUC - ------------------------ (4-25)
1 + (2Af - 1) sin ~tc

in which K0 = coefficient of horizontal soil stress and Af Skempton's pore water

stress parameter, defined as:

(4-26)

4-32
for saturated soil with 6u = excess pore water stress developed during loading,
bal - major principal stress increment, and ba3 - minor principal stress incre-
ment. Typical ranges in Af are shown in Table 4-7. For isotropic consolidation
(K 0 = 1), Equation 4-25 reduces to:

sin if>tc
(sufavo)CIUC = ------------------- (4-27)
1 + (2Af - 1) sin ~tc

In both cases, if>tc is used since it was shown earlier that the consolidation state
does not influence the friction angle.

The undrained strength ratio in triaxial compression also can be predicted from the
modified Cam clay model (e.g., l). For isotropic consolidation, this ratio is:

0.5 M (0.5)/\ (4-28)

with M and A defined in Appendix G. For anisotropic consolidation, this ratio is:

A
sin if>tc 2
(a + l) (4-29)
2a 2

in which

Table 4-7

TYPICAL RANGES IN Af FOR ISOTROPICALLY CONSOLIDATED CLAYS

Clay Type Af

High sensitivity 0.75 to 1.5

Normally consolidated (NC) 0.5 to 1.0

Lightly overconsolidated (LOG) 0 to 0.5

Heavily overconsolidated (HOC) -0.5 to 0

Source: Skempton (39), p. 146.

4-33
 3 - sin <Pte
a - ---------------- (4-30)
2(3 - 2 sin ~tc)

For isotropic consolidation, this model also predicts Af in NC clays, as follows:

Af - [2A + (M/3) - 1]/M (4-31)

Typical values of A range between 0.7 and 0.8, with 0.8 being used most often.

To examine the applicability of these relationships for predicting the undrained

strength ratio of NC clays, a data base of 48 different clays (24, 25, 40) was used
for comparison. Figure 4-31 shows the direct comparisons between the undrained
strength ratios for isotropic consolidation and for K0 or anisotropic consolida-
tion. It should be noted that the data base consisted of tests that were: (1)
accurately consolidated using K0 testing procedures, (2) consolidated to estimated
K0 stress values, or (3) consolidated to some general anisotropic stress which may
or may not be equal to K0 . Linear regression of these data showed the following:

Su )
( --- =0.15+0.49 ( ---
Su )
O'"vo CAUC O"vo CIUC

u
:::::>
<l
u
....
0
u
:::::>
0
:lie

-It?u
0 0.2
.......
:;)
(/) ---Coulomb- Mohr w. A f noted
- · - Modified Com cloy w. A noted
0.1

48 intact cloys

0.1 0.2 0.3 0.4 0.5 0.6 0.7

(suI ii-vo) CIUC

Figure 4-31. Comparison of Undrained Strength Ratio for NC Clays After Anisotropic
and Isotropic Consolidation

Source: Data from Mayne (24) and others (25, 40).

4-34
I
J
 (sufovo)CAUC = 0.15 + 0.49(Suf0vo)CIUC (4-32)

with the statistics shown on the figure. A linear regression through the origin
was inappropriate for these data. Also shown on this figure are the predictions
from the Coulomb-Mohr failure envelope geometry (Equations 4-25 and 4-27), using
typical limiting values for Af and K0 = (1 - sin ~tc) from Section 3, and from the
modified Cam clay model (Equations 4-28 and 4-29), using typical values for A. As
can be seen, the Coulomb-Mohr predictions bound the data well and modified Cam clay
provides an accurate prediction, although slightly conservative.

Figure 4-32 provides a more detailed comparison of these data, with each undrained
strength ratio plotted versus ~tc· For isotropic consolidation (Figure 4-32a),
linear regression on these data gave the following:

(sufov 0 )ciUC = 0.0120 ~tc (4-33)

with the statistics shown on the figure. The regression line and the modified Cam
clay prediction agree well. The Coulomb-Mohr prediction bounds much of the data,
but tends to be somewhat on the high side. It should be noted that the A - 0.8
line from the modified Cam clay model is identical to the Coulomb-Mohr model using
Af predicted by the modified Cam clay model (Equation 4-31). These data further
show that the following provides a reasonable lower bound for the data:

for lower bound (4-34)

This observation is in general agreement with Wroth's suggestion (l) that:

(4-35)

The remolded and critical void ratio values are consistent with lower bounds on
natural soils.

For anisotropic consolidation (Figure 4-32b), linear regression on the data gave
the following:

(4-36)

with the statistics shown on the figure. The Coulomb-Mohr predictions tend to be
on the high side, while the modified Cam clay predictions tend to be on the low

4-35

L
 0.2

---Coulomb-Mohr w. A 1 noted
-·-Modified Com clay w. A noted
0.1
81 intact cloys
I fissured cloy

( b) Anisotropic Consolidation

Su ) -
( --- =0.0117 ¢,
fJvo CAUC c
2
(n=75, r =0.578,
u
:::1 S.D. =0.044)
<(
u
0
•t?'
.......
;:;,
(/')
0.2

---Coulomb-Mohr w. A 1 noted
- · - Modified Cam cloy w. A noted
0.1
71 intact clays
I fissured cloy

-
cl>tc
Figure 4-32. Undrained Strength Ratio versus ~tc for NC Clays 'I
1

Source: Data from Mayne (41) and others (25, 26, 40).

4-36
I
J
 side. The ~tc/100 relationship still provides a lower bound estimate, although not
as reliably as for isotropic consolidation.

Figures 4-32a and b suggest that the su/avo ratios for isotropic and anisotropic
consolidation are nearly the same, although it is clear from Figure 4-31 that the
anisotropic value is less than the isotropic and the difference increases with
increasing su/avo· Part of the reason for this apparent anomaly is that the data
bases in these three figures are not the same. Also, it is clear that the data in
Figure 4-32a exhibit a near-linear trend, while the data in Figure 4-32b exhibit a
pronounced curved trend. For these reasons, it is believed that the regression
lines given in Figures 4-31 and 4-32a should be more reliable and be used to inter-
relate (su/av 0 )cAUC, <Suray 0 )ciUC, and '4>tc for "4>tc > 20•. As a preferred alterna-
tive, the modified Cam clay model (Equations 4-28 and 4-29) can be used directly.
It agrees well with the regression line in Figure 4-32 for predicting (su/av 0 )crUC·
and it gives a slightly conservative value of (su/av 0 )cAUC• typically 0.01 to 0.02
less than the regression, and a better fit for low su/avo values.

Influence of Overconsolidation

The undrained strength ratio increases with increasing overconsolidation, as meas-

ured by the overconsolidation ratio (OCR= apray0 ). Figure 4-33 shows typical
experimental data illustrating this OCR effect, as measured in direct simple shear
(DSS) tests. The concept of SHANSEP (~tress ~istory and ~ormalized ~oil ~ngineer­

ing Rarameters) addresses this phenomenon and uses this behavior to correct labora-
tory test results for sample disturbance effects (e.g., 11). For example, Figure
4-34 shows these same data in normalized form, indicating a rather narrow band.
Based on data of this type, the following general equation was suggested (e.g., 11):

(4-37)

with m- 0.8. However, a better fit occurs when m = 0.85 to 0.75 with increasing
OCR. This experimental observation also is the basis for the approximation made by
Jamiolkowski, et al. (35) for low to moderate PI soils, as given below:

su/av0 - (0.23 + 0.04) OCR0.8 (4-38)

This equation basically is a revised form of Equation 4-21, corresponding approxi-

mately to DSS conditions.

4-37

L
 : 0il wl(%) Moine Organic
Cloy 1
1.6 65
2 65 6
3 95
4 71
5 41
1.2 6C 65 Soils CD to@
68 35

u u

0.8 , Conn. Volley 6 --•b --•t:

0
0
z
0

I
/ ' Varved Cloy ..............
/

--
/ :::> :::>
/ en en
/
/
/
/

I
/

/
.., .., /

0
I 2 10 20 I
OCR OCR
Figure 4-33. suf0 vo versus OCR Figure 4-34. Normalized 5u;-0y0 versus
OCR
Source: Ladd, et al. (11)' p. 26.
Source: Ladd, et al. (11), p. 26.
I
It should be noted that the general form of Equation 4-37 will hold regardless of
strength test type (e.g., 1. 35, 42). However, (sufavo)Nc will vary significantly
I
with test type and m will vary to a limited degree, as described subsequently.
I
This general behavior also is predicted by the modified Cam clay model, as follows
(e. g., 34): I
(4-39)
I
with A typically about 0.8. Fundamentally, this equation applies to CIUC test con-
ditions (Equation 4-28).

The modified Cam clay model also can be used for predicting Af, as follows:

[(2/0CR)A + (M/3) - 1]/M

(4-40)
I
4-38
i
Jl
I
I
II
This equation is in general agreement with available data, as shown by the observed
trends in Figure 4-35, except at higher OCR values where Equation 4-40 tends to

I i
predict a larger negative value than observed. Figure 4-35 also shows that the
initial consolidation state plays a significant role in evaluating in Af.
age, the difference between isotropic and K0 consolidation data based on Figure
On aver-

I 4-35 can be estimated as:

I
Af(CIUC)
~ 2.3 - 3 log OCR (for OCR > 2) (4-41)
Af(CK 0 UC)

I It also should be noted that Af will differ in extension and compression, as shown
in Figure 4-36. This general pattern has been observed by others (e.g., 35),

a although the indicated variation with PI may not be truly general for other soils.

Influence of Test Boundary Conditions

'I The influence of test boundary conditions has a pronounced effect on the undrained
strength ratio (sufov 0
tion angle (~tc)·
), much larger than the effect on the effective stress fric-
As shown previously for triaxial compression (Equations 4-28 and
4-29), the anisotropic or K0 consolidation normally gives a lower sufov0 than

I 1.2

--If)

I f)

' ~
·u;
c
0
If)

I'
Q)
.....
0.
E
0
u
0.4
0
Range of data (CK 0 UC)
X
0
..... 0.2

J 1-
E
0
.........
t <! -0.2

-0.4
I 2 5 10 20 50 100
OCR
Figure 4-35. Observed Trends Between Af and OCR

Source: Based on Mayne and Stewart (40), pp. 1341, 1342.

4-39

l
 ..
CIUC CK 0 UC ClUE CKJJE
Kawasaki cloy-
.c:, mixture series 0 0 .c:, 'V
Natural marine
cloys e • A

<t -
-..:
Q)
Q)

E
....0
0
a..
rn
c:
0
0..
E
Q)
x
(f)

for 12! NC Japanese cloys

I!
0
0 20 40 60
Plasticity Index, PI {%)
Figure 4-36. Relationship Between Af and PI for NC Clays

Source: Nakase and Kamei (25), p. 37.

isotropic consolidation, as shown in Figure 4-37. The validity of these relation-

ships was shown in Figure 4-31.

The second test boundary condition of importance is the intermediate principal

stress, which most commonly is evaluated by comparing plane strain and triaxial
tests. From the modified Cam clay model (e.g., l), the undrained strength ratio of
NC clay in plane strain compression can be expressed as:
I
I
sin ~psc d2 + 1 A
(4-42)

:
(sufovo )psc = 2d ( 2 )

in which

d - 1/(2 - sin ~psc) (4-43)

I
4-40 I
I
I 0.6

Test Equation

I CIUC
PSC
CKOUC
4-28
4-42
4-29

t oss
PSE
CK 0 UE
4-47
4-52
4-53

I 'b
'-
0
>
A =0.8

:::l

I
rn

a
0.2

CKOUE
0.1

t
cf>tc

I Fig~re 4-37. Undrained Strength Ratio as a Function of Test Type

•I and ~psc ~ 1.1 ~tc• as shown previously.

equations with the limited data available.
conservative.
triaxial extension tests.
Figure 4-38 shows a comparison of these
The agreement is very good and slightly
For comparison, Figure 4-39 shows the results for plane strain and
The plane strain extension results also are larger than
the triaxial, and the differences in extension are larger than in compression.
Equation 4-42 also is plotted on Figure 4-37 for reference with other test types.

The next important test boundary condition is the loading direction or stress rota-
tion. For natural clays, strength anisotropy can develop from both stress aniso-
tropy (K0 stresses) and structural anisotropy (layering, fabric, sensitivity,
etc.). The complete range of loading angles and stress rotation effects can be
investigated only in sophisticated, hollow cylinder tests. However, since these
tests are rather expensive and difficult to perform, it is more common to use sim-
pler tests with more limited loading directions (S). Most commonly, triaxial com-
pression (S = o·), direct simple shear (S ~ 45°), and triaxial extension (S- 90•)
tests are used, as illustrated in Figure 4-40. This figure shows the general

4-41
 0.4

0.3

'
0 ..2 Theory
(Equations 4-29, 4-42)

0.1
10 intact cloys

0.4 0.5

Figure 4-38. Comparison of Undrained Strength Ratios from PSG and CK 0 UC Tests

Source: Data from Ladd, et al. (11) and Mayne and Holtz (26).

Su )
( -- =129 ( -Su
-)
crvo PSE Uvo CK 0 UE
(n=6, r 2 =0.639,
5.0.=0.030) ~

t
w
Vl

-
lb>
Q.

....._ I>
:::>
VI 0.2

I
•I
0.1

6 intact cloys

0.1 0.2 0.3 0.4 0.5

Figure 4-39.

Source:
Comparison of Undrained Strength Ratios from PSE and CK 0 UE Tests

Data from Ladd, et al. (11) and Mayne and Holtz (26).
I
4-42 I
I
 CK 0 UC oss CK 0 UE

~
(J

_L,
§? ~· 6.a-,
1.0

0
0
II
00
0
- 0
•b'"
........
on
::J -•b> 0
0.5
0 Voterlond cloy
........ ~ Boston blue cloy
::>
on b. Drommen plastic cloy

0
• Drommen leon cloy

oo 30° 60° 90°

Major Principal Stress Loading Direction, o

Figure 4-40. Effect of Loading Direction on su

Source: Hansen and Clough (43), p. 258.

observed pattern, with the DSS results intermediate between the triaxial compres-
sion and extension results.

At the present time, there is no general agreement on methods of interpreting the

DSS test results in terms of effective stresses. Wroth (1) discusses many of the
pertinent issues involved and presents three possible equations for interpretation,
as given below:

(su/avo)oss-1 = 0.5 sin ~tc (4-44)

(su/avo)oss-2 tan </>psc (1 - sin ~sc )/(1 + sin ~sc) (4-45)

- - 2 (4-46)
sin </>psc/(1 + sin </>psc)

with ~psc z 1.1 ~tc as shown previously. Figure 4-41 compares these equations with
available data on 41 clays. As can be seen, the DSS-1 interpretation typically is
high, especially at high ~tc values. The DSS-2 interpretation is consistently very
low, while the DSS-3 interpretation appears to be adequate.

Figure 4-42 shows the DSS data plotted against triaxial compression data, showing

4-43
 Wroth, 1984 / /
/
DSS-1 0.5 sin <f>tc //
_
-
1-sin c/>psc _ • ~/
/
Dss 2 . - toncf>psc ~ • ....-::
(f)
I + Sin !psc A g e.f'/ ~
/ '7 f{i~·
(f)
Cl DSS-3 sin c/>psc (

(l+sin ~ps~-~~4..~....
•• • •
.=·····,[)IV
"- rn.....-~
4~
* ......... ········
./·~· , DSS-1
0.1
./ /....-:
"
/hh & DSS vs. CK 0 UC 41 cloys
9

Figure 4-41. Undrained Strength Ratio from DSS versus ~tc

Source: Data from Mayne (44) and others (26, 35, 41).

Figure 4-42. Comparison of Undrained Strength Ratios from DSS and CK0 UC Tests

Source: Data from Mayne (44) and others (26, 35, 41).

two of the DSS interpretation methods. The DSS-1 method is consistently high,
while the DSS-3 method exhibits a high degree of curvature which is not evident in

4-44 I
J
I
I the data. Because of these problems with the theoretical models in describing the
data, it is more prudent to rely on the regression line for the data, given by:

I (4-47)

I This regression line is plotted on Figure 4-41 and provides as good a predictor as
the DSS-3 interpretation method. For these reasons, Equation 4-47 will be the

I recommended method for evaluating the DSS undrained strength ratio.

also is plotted in Figure 4-37 for comparison with the other test types.
This equation

I Lastly, it is necessary to address the behavior in extension and compression. Pre-

vost (45) developed simple relationships between the different tests, as given

I below:

(4-48)

'
I These relationships generally are consistent with previous experimental observa-
(4-49)

I
J
tions (e.g., 46) that the DSS strength is roughly equal to the average of the tri-
axial compression and extension strengths. Available data for the DSS and triaxial
tests are shown in Figure 4-43 and indicate general agreement. However, the

'
I
regressions for both the triaxial and plane strain data are lower than Prevost's
model (45). Therefore, to be consistent with the data, the relationships given by
Equations 4-48 and 4-49 should be changed as follows:

(4-50)

I (4-51)

I Equations 4-50 and 4-51 then can be rearranged to yield the extension strengths
directly, as below:

(4-52)

(4-53a)

(using Equation 4-47] (4-53b)

4-45
 (/)
(/)
0
---•t?'
0
DSS=0.40 (PSC +PSE)
....... (n=6, r 2 =0.442,
:::>
(/) S.D. =0.026)

0.1

0.1 0.2 0.3 0.4 0.5 0.6 0.7

Figure 4-43. Comparison of Undrained Strength Ratios from DSS, CK 0 UC and CK 0 UE, and
PSG and PSE Tests
'
Source: Data from Mayne (44) and others (26, 35).

These relationships then can be added to those on Figure 4-37 to provide a general
comparison of the different test results. Available data comparing the CK 0 UE and
CIUC results with Equation 4-53 are shown in Figure 4-44. As can be seen, the
theory underestimates the triaxial extension strength by a modest amount and there-
fore is somewhat conservative.

As an alternative approach for evaluating the extension strength, it has been sug-
gested by Ladd, et al. (11) that the ratio of undrained strengths in extension to
compression generally increases with increasing plasticity index, as shown in Fig-
ure 4-45. As can be seen, this is a fair alternative and could be used as a check
on the analytical prediction from Equations 4-52 and 4-53.

The available data bases also provide an opportunity to evaluate the exponent A in
the modified Cam clay model. Table 4-8 summarizes these data, showing that A
ranges from 0.72 for compression tests, to 0.78 for shear tests, to 0.82 for exten-
sion tests.
cent.
Overall, A is given by 0.75 with a coefficient of variation of 13 per-
This value is close to the common assumption that A z 0.8.
I i

4-46
•I l

I
 0.4-

w
::>0 0.3-
::.::
u

0.1 r- -

23 intact cloys
I I I l I
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

Figure 4-44. Comparison of Undrained Strength Ratios from CK 0 UE and CIUC Tests

Source: Data from Mayne and others (26, 41).

"'
y:

c:
c: ·;;;
0
0
·;;; rn
c: ,_
OJ
OJ a.
"';< E
OJ 0 0.4
u
c:
c:
:;J
1/) .. x
"*
:;J
1/) @ ~ Triaxial: 64 cloys
0.2@- 2 sands
Plane strain: 8 cloys
(denoted by p}

Plasticity Index, PI(%)

Figure 4-45. Undrained Strength Ratios in Extension and Compression versus
Plasticity Index

Source: Data from Mayne and Holtz (26) and others (25, 35, 41, 47).

4-47
 Table 4-8

EVALUATION OF MODIFIED CAM CLAY EXPONENT A

Reference A
Basis Test Type n mean cov (%)

Test CIUC 50 0.709 15.5

Evaluation PSG 2 0.730 2.0
CK 0 UC 34 0.738 12.7
DSS 30 0. 776 11.7
PSE 3 0.843 6.8
CK0 UE 19 0.818 13.8

Design Compression 86 0.72 14.1

Recommendation Simple shear 30 0.78 11.7
Extension 22 0.82 12.8

Overall Mean All Types 138 0.75 13.4

Source: Data from Mayne and others (26, 35, 41, 44).

Influence of Strain Rate During Testing

Soils exhibit a change in strength as a function of strain rate during loading. In

general, for triaxial compression tests, each log cycle increase in strain rate is
accompanied by a 10 percent increase in su (46). This observation is confirmed by
the data shown in Figure 4-46 for 26 clays tested in triaxial compression.
et al. (48) have shown that these trends also are observed for DSS and CK0 UE
tests. A testing rate of 1 percent/hour is considered as the standard reference
Graham,
''
I'
rate. For other than the standard rate, the following should be used:

su!<su for € - 1%/hr) ~ 1.0 + 0.1 log € (4-54)

I.
For most conventional loading cases, the standard rate would be appropriate for
design.

Summary of Factors Influencing the Undrained Strength Ratio

As described previously, many factors influence the measured value of Su. Using
the CIUC test as a standard reference, the value of su/ovo can be determined as
follows:
I
4-48 I
I
 I. 5.-----.,.~----..-l---~.-----.,.,---·<l---,--l--..,.l------.

Su <J ~
----=;...,-- = 1.00 + 0. 10 log i. <l ..,
su at 1% ~""ee

(,~209, < 2 ~0.802, S.D~~( '0 '0

it('. -

-g~~~~ St~eo,VIoMey
1.0- "

0 ;:,cloy'., Ledo (55)

~ ~ ® e Grande Bole~ne (50) g Drommen (56)

-
'-
0 ® Olga (50) \1 Fukokuso (57)
IB Broodbock (50) .6. Honey (58)
lSI Belfast (48) 0 Mexico City (59)_
0.5- (60)
r&l Lyndhurst (48) o Kawasaki M-30
Y Mostemyr ( 48) ..t. Kawasaki M-15 (60)
<>Winnipeg (48) W Kawasaki M-10 (60)
A. Bangkok ( ~) e Boston Blue (§l)
+ Bangpli (52) 0 Weald (62)
~ Rangsit (52) <J Grundite (63)
6l Vicksburg (53) [> Sodium Illite (64)
rn Atchofaloya (54 l () Khor -AI-Zubair (54) -

0~----~'----~'~----L-----~'------~~----~'~---~'-~ 5
ro- 3 ro- 2 ro-' 1o' ro 2 ro 3 ro 4 10
.
Strain Rate, E (%/hr)
,
Figure 4-46. Strain Rate Influence on su

(4-55)

in which the a coefficients are given in Table 4-9 and Figure 4-47, and (sufav 0 )cruc
is given as:

(sufavo)CIUC = 0.5 M (0.5)A (4-28)

Table 4-9 also gives a simple linear approximation for aTEST which may be useful
for first-order estimations.

UNDRAINED SHEAR STRENGTH OF COHESIVE SOILS CORRELATED WITH IN-SITU TESTS

In-situ tests can provide either a measurement or estimation of Su in clay depos-

its. At the present time, direct determinations of su are obtained from·the field
vane shear test (VST). The values of su from the standard penetration test (SPT),
cone penetration test (CPT), piezocone penetration test (CPTU), pressuremeter test
(PMT), and dilatometer test (DMT) currently are obtained from analytical models,

4-49
 Table 4-9

CORRECTION FACTORS FOR su COMPARED WITII su FROM CIUC TEST RESULTS

Test
Linear
Approximation
Within
'
I
Influence Term Type Value 20o < ~tc < 40o
I
Test Mode aTEST CIUC 1.0 1.0
PSG
CK 0 UC
DSS
Ratio
Ratio
Ratio
of
of
of
Eq.
Eq.
Eq.
4-42/Eq.
4-29/Eq.
4-47/Eq.
4-28a
4-28
4-28
1.22
1.13
0.77
-
-
-
0.0112
0.0094
0.0064
~tc
~tc
~tc
I
PSE Ratio of Eq. 4-52/Eq. 4-28 0.71 - 0.0052 ~tc
CK 0 UE Ratio of Eq. 4-53/Eq. 4-28 0.56 - 0.0046 ~tc

Strain Rate aRATE All 1 + 0.1 log ib

Overconsolidation aocR All OCRA

A 0. 72 for compressionc
A 0.78 for simple shearc
A 0.82 for extensionc

a- See plots given in Figure 4-47.

-I
b - Normal reference rate is 1 percent per hour.
c - See Table 4-8 for additional details.

empirical correlations, or calibration with a known reference strength. As noted

previously, each in-situ test provides a different su particular to the boundary
conditions imposed, rate of loading, direction of loading, etc.
I
I
Correlations with VST Results

The vane shear test (VST) is one of the oldest in-situ tests for the evaluation of
11
I
su in clays. The value of Su is determined from the torque required to rotate a
four-bladed vane in the clay. Both the peak and remolded Su can be determined, and

II
therefore the sensitivity (St) of the clay can be computed. Details of the VST are

,,
given in Appendix E.

The value of su determined from the VST should not be used directly in analysis,
because it needs to be corrected for the strain rate during testing and the soil
anisotropy. Bjerrum (65) reviewed a number of failure case histories from embank-
ments, excavations, and foundations which had been evaluated using Su from the VST 1?1
and developed a correction factor(~) that is to be applied to su(VST). This

4-50 I
J
I
I
I
0.8 ..
... ... ...
u ... ...
... ...
I lb
0

.......
::I
lb>
::>
.....
u
0

.......
0.6
----
... ... ...
...
....
-:.:.:.::·.:.::.:_......
---·:.:.:..:.·:..:..:..; ... ,
_rOSS
en

'
::I
IJl PSE-:;.:.:..:.:·.:.:..:.. ........._ ..
0.4

j
0.2

I ¢psc = 1.1 ¢tc

A = 0. 7 2 for compression

I cl>tc
Figure 4-47. Normalized Undrained Strength Ratios for Major Laboratory Shear Test

'• Types

factor apparently is correlated with the plasticity index. A recent update of this
''
i

I correlation is given in Figure 4-48.

In addition to the~ correction for strain rate and anisotropy, Bjerrum noted that
there were differences in the apparent preconsolidation stress caused by aging
which influenced the computed sufavo ratio. To account for this aging, he recom-
mended that the data be presented in terms of the preconsolidation stress, as shown
I below:

I su(VST)/ap = (Su(VST);-av0 ] OCR (4-56)

Figure 4-49 shows a typical plot of this type, which includes the recommended cor-
I relations of Bjerrum (66) and Skempton (30), given earlier as Equation 4-18.
can be seen, the Bjerrum correlation fits the data for inorganic clays somewhat
As

I better. It should be remembereci that Chandler (32) cautions against use of these

I 4-51

I-
 I
:::t.
1.4
•
su (field)= fL Su (VST)
I
..:
-0
<->
1.2 •
•• I
" .. I
0
LL
• le-i
1.0

.N
c:

-0
"~ ~ I
..
<-> 1 Bjerrum ( 1972)
,_
(1)
,_ 0.8 • I
.:
u
0
• • •......__
I-
(/)
>
0.6
~
••
.........
• I
0.4
0 20 40 60 80 100 120 I
Plasticity Index, PI {%)
Figure 4-48. Field VST Correction Factor I
Source: Ladd, et al. (11), p. 469.
i
0
Skempton 09571 I
0 0

I
0
a. 0 0 ....... -
'b
........ .... 8 (].... ......... -
~
0
• • _______ ................
•I
0 _....... - .J0
I-
(/) .......ct .................
> 0
0

0.2
Inorganic cloys
Eastern Canada e
Others
Organic cloys
A
0 I
Figure 4-49.

Source:
Plasticity Index, PI (%)
Vane Undrained Strength Ratio versus Plasticity Index for NC, Young
and Aged Clays

Tavenas and Leroueil (23), p. 21.

I
~
~
f
~
'I
types of relationships in fissured, organic, sensitive, or other unusual clays.
I
"

Equation 4-56 includes another important difference in evaluating the undrained I I

1
~-

4-52 I
LI
I
I strength of OC clays using the VST. Earlier it was shown that, for laboratory
tests, the strength increased with increasing OCRA. Typical A values ranged from
about 0.7 to 0.8. However, with the VST in the field, A basically is equal to
I unity.
32, 67).
This point has been demonstrated effectively by several authors (e.g., 23,

I Subsequent examination by Mesri (68) of Bjerrum's correction factor (Figure 4-48)

and su(VST)/ap relationship (Figure 4-49) suggested the following:
I
-I
(4-57)

in which su(field) represents the average mobilized undrained strength in the field
for stability problems such as embankments on soft clay and foundation bearing
capacity. This relationship has been corroborated in independent studies by Trak,
et al. (69) and Larsson (70). Recent studies by Mesri (36) have reconfirmed this

I relationship and have noted further that:

(4-58)

I These last two equations link the direct field and laboratory shear tests and
provide a general basis for evaluating the actual field value of su for design. As

'I noted previously, caution is warranted in unusual clays.

Correlations with SPT N Value

Correlations have been attempted for estimating su from SPT N values, even though

I it is known that these correlations are weak. The most common of these is shown in
Table 4-10, which was developed primarily using unconfined compression tests. From

I the results of this table, su can be approximated as follows:

(4-59)
sufpa"' 0.06 N

I Many other relationships have been proposed as well, and several of these are shown

I in Figure 4-50. It is clear that these relationships represent a wide variety of

interpretations of soil types and testing conditions and that a universal relation-
ship between su and N is unlikely. Several other serious problems exist with

I Figure 4-50.
energy level.
First, the SPT N values have not all been standardized to the same
Second, there is no indication of the reference strength used to

I determine su. The mixing of different undrained strength data is inconsistent, and
it increases the scatter in the reported trends. Third, the sensitivity of the

I 4-53

I
 Table 4-10

APPROXIMATE su VERSUS N RELATIONSHIP

N Value
(blowsjft or 305 mm) Consistency Approximate sufpa

0 to 2 very soft < 1/8

2 to 4 soft 1/8 to 1/4

4 to 8 medium 1/4 to 1/2

8 to 15 stiff 1/2 to 1

15 to 30 very stiff 1 to 2

> 30 hard > 2

Source: Terzaghi and Peck(~), p. 347.

Q)
::J
0
>
z
1-
0..
(/)

Figure 4-50. Selected Relationships Between N and Su

Source: Djoenaidi (71), p. 5-93.

4-54
clay can affect the N value greatly, as shown in Figure 4-51. Apparently, the
penetration process causes temporary excess pore water stresses which reduce the
effective stresses in the vicinity of the sampler, thereby resulting in an appar-
ently lower N value.

However, for clays within a given geology, a reasonable correlation might be expec-
ted between su and N. Figure 4-52 indicates this behavior over a wide range of N
values where the same drilling equipment, SPT procedure, and consistent reference
strength (UU triaxial) were employed. For these data, the reported regression is
given by:

0.29 N0.72 (4-60)

This equation tends to predict su/Pa on the high side of the relationships shown in
Figure 4-50.

Correlations with CPT qc Value

The theoretical relationship for the cone tip resistance in clay is given by:

(4-61)

in which qc =cone tip resistance, avo= total overburden stress, and Nk =cone
bearing factor. The application of classical plasticity theory to this bearing
capacity problem suggests Nk on the order of 9 for a general shear model. Cavity

1.0

--II

(j)

0
de Mello
-z
z
.......
0.5
-o
...
0.>
::l
Vl
0 Schmertmonn
0.>
~
0
0 2 4 6 8 10

Sensitivity, St
Figure .4-51. Apparent Decrease of N with Increasing Sensitivity

Source: Schmertmann (14), p. 66.

4-55
 25 cloy sites in Japan
(PI:::: 10-95, OCR=I-3)

5 10 50 100
SPT N Value
Figure 4-52. Relationship Between su and SPT N Value

Source: Hara, et al. (72), p. 9.

expansion theories give Nk increasing in the range of 7 to 13 for increasing values

of rigidity index (Ir = G/su, with G- shear modulus). Steady penetration theory
provides a narrow range for Nk between 14 and 18 for a wide range of Ir.

With the various uncertainties in choosing appropriate theoretical models, it is

not surprising that Nk usually is determined empirically by calibrating CPT data
with a known measured value of su.
CPT data is presented in Figure 4-53.
The range of values of Nk back-calculated from
This wide range of Nk values must be scru-
I
tinized for several reasons: (1) inconsistent reference strengths, (2) mixing of
different type cones (electric and mechanical), and (3) need for correction of qc
I
for pore water stress effects (Appendix B). These factors can change Nk dramati-
cally. I
I
l

The importance of correcting qc for pore water stress effects has been discussed
previously and is illustrated by Figure 4-54 for two piezocones with different area I
ratios. The corrected cone tip resistance (qT) can be obtained only by use of
piezocones with porous elements located behind the tip. Consequently, the large
scatter observed in empirical determinations of Nk may result, in part, from use of
I
4-56 I
I
 3.0r----.----••----~•.----.-l---.----.----.l----.-----.-,---,

2.5t-

2.01-

0
0.
1.5
'en
~

1.0

I I I
30 40 50

Figure 4-53. Reported Range of Nk Factors from CPT Data

Source: Djoenaidi (71), p. 5-83.

Measured Cone Tip Resistance, Corrected Cone Tip Resistance,

qc/Pa Qr1Pa
0 2 4 6 0 2 4 6
0 I 0 0 0

'
/• o:I q cB 0
··.. (o=0.86)
10
Q~
B~·o...,. 10
4 •I ·o QT wq q A

E • ~ ~~:. T
"q
- qAJ! 20
-
20 .,_
0 •

-
J::.
0.
Q) 8 c
(o=0.38) '1· 0:
·o
.·
·o
8 Q
•.• I

/at\
·.. I
•
0
•I
o.
30 oe
'\
30
• Q.~,
a.
12
T
Emmerstod •
·o.
··...:o 4ol2
\· ..
~,_~40
Quick Cloy ). o· o· •
I I I

Figure 4-54. Effect of Pore Water Stress on Cone Tip Resistance

Source: Aas, et al. (67), p. 19.

I 4-57
 an uncorrected qc.

The value of Nk ideally should be determined experimentally by comparison with a

consistent reference strength. Often, the field VST is used as the reference. In
this regard, it is important to recall that the VST requires a correction for su in
itself. Early correlations (e.g., Battaglio, et al., 73) for Nk using uncorrected
VST data suggested a trend for Nk in terms of the plasticity index (PI). However,
upon later re-analysis of the same data using the corrected VST strength
[~ su(VST)], Nk apparently was independent of PI.

Subsequent studies by Keaveny and Mitchell (74) and Konrad and Law (75) have demon-
strated that Vesic's cavity expansion theory (76) provides a reasonable estimate
for Nk, as given below:

2.57 + 1.33 (ln Ir + 1) (4-62)

Keaveny and Mitchell suggest using CK 0 UC triaxial compression tests to evaluate Ir,
while Konrad and Law recommend using the self-boring pressuremeter test.

Recent theoretical developments (Houlsby and Teh, 77) suggest that more refined
procedures for determining su from the CPT may be appropriate. However, these
models currently require a number of parameters that are difficult to determine.
Further testing in the future may allow convenient determination of these parame-
ters and a better estimation of Su·

Correlations with CPTU Results

The piezocone penetration test (CPTU) permits determination of Su from the cor-
rected cone tip resistance (qT), as described previously, and also allows for a
separate estimate of su from the pore water stress measurement. Research on this
subject (e.g., Robertson, et al., 78) has suggested the following:

(4-63)

in which ~u = measured excess pore water stress (urn - u 0 ) and N~u = pore water
stress ratio, which may be estimated from Af and either the PI or rigidity index,
as shown in Figure 4-55. Alternative recommendations by Konrad and Law (75) sug-
gest a more complex relationship, including a number of parameters which are some-
what difficult to evaluate.

4-58 I
I
 >- ....... 15 500
.-=:

l
(L
:;::u >< 200
(/)
o-o
Q)
100 (91~
-a.. ......
c
75
50 x
Q)
20 -o
c
10
->-
-o
~ 01
i:i:

10
u 8 6 4 2 0

Pore Water Stress Ratio, N

6u
=t.u
su

Figure 4-55. su as a Function of ~u in the CPTU

Source: Robertson, et al. (78), p. 1273.

Correlations with PMT Results

The pressuremeter test (PMT) ideally provides a measurement of Su at the PMT limit
stress. Based on cavity expansion thecry (Baguelin, et al., 79), su can be eval-
uated from:

(4-64)

in which PL ~ PMT limit stress, p 0 = PMT total horizontal stress, Np ~ l + ln(EPMTI

3su), and EpMT = PMT modulus. Values of Np may range from 2 to 20 (Mair and Wood,
19), but typical values usually range from 5 to 12, with an average of 8.5. Dif-
ficulties in choosing the correct value of Np are compounded by possible measure-
ment errors in both PL and Po·

An alternative and more direct method to obtain su is shown in Figure 4-56. By

re-plotting the basic data as shown in Figure 4-56b, a straight line develops. The
slope of this line is su. Wroth (l) notes that su from the PMT should be close to
the value obtained from plane strain compression (PSG) tests.

Correlations with DMT Results

The dilatometer test (DMT) horizontal stress index, Ko = (p 0 - u 0 )/av0 , has been
correlated with su, as shown in Figure 4-57. Based on these data for Italian
clays, the following correlation was suggested:

4-59
 (a ) Basic Data F1 ( b ) Processed Results
I
- a.
0

......... I
(
I
19

a. 18
I
V'l
I
V'l I
17
-....Cl>
( /)

'"0
Unloading\;/
I

16
.!!! /
a. /
a. /
<! / 15 • D \ _ Initial loading
/
/"-

0
/

4 8 12 6 7 8 9 10 12 14 16 18 20
-2
I
Strain, e (%) Volumetric Strain, t:NIV= 1-(l+e)
Figure 4-56. PMT Results in Bartoon Clay I
Source: Wroth (i), p. 462.

I
5
For 1
0
~ 1.2
I

./
..
I

e/
/
./
I
.
2
7.&
It:
0
) .,~ I
·v/·
......... - ~ -
::> I

I
V'l

0.5 1-

0.2
I
.
/
2
/~
-t
_.{

I
5 10
I
.•• u
uu
VST

20
-

I
Figure 4-57.
Horizontal Stress Index, K0
su as a Function of Ko from the DMT
I
Source: Marchetti (80), p. 317.
I
(sufUvo)PMT = 0.22 (0.5 Ko)l.25 (4-65)

This equation originally was based on clays with a material index, lo, less than or
equal to 1.2, Current recommendations (Schmertmann, 81) are to limit this rela-
tionship to clays with lo ~ 0.6. The strength data initially were obtained from
unconfined compression tests (U), unconsolidated-undrained triaxial compression

4-60
 tests (UU), and field vane shear tests (VST). Subsequent work by Lacasse and Lunne
(82) suggests that the 0.22 coefficient should vary with test type as follows:
0.14 for direct simple shear, 0.20 for triaxial compression, and 0.17 to 0.21 for
field VST. Other data by Powell and Uglow (83) indicate different factors for
fissured clays and glacial tills if the reference su is determined from plate load
tests or the self-boring pressuremeter test.

REFERENCES

1. Bishop, A. W., "The Strength of Soils as Engineering Materials", Geotechnigue,

Vol. 16, No. 2, June 1966, pp. 89-130.

2. Kulhawy, F. H., Trautmann, C. H., Beech, J. F., O'Rourke, T. D., McGuire, W.,
Wood, W. A., and Capano, C., "Transmission Line Structure Foundations for Up-
lift-Compression Loading", Report EL-2870, Electric Power Research Institute,
Palo Alto, 1983, 412 p.

3. Wroth, C. P., "Interpretation of In-Situ Soil Tests", Geotechnigue, Vol. 34,

No. 4, Dec. 1984, pp. 449-489.

4. Terzaghi, K. and Peck, R. B., Soil Mechanics in Engineering Practice, 2nd Ed.,
John Wiley and Sons, New York, 1967, 729 p.

5. Schmertmann, J. H., "Guidelines for Cone Penetration Test Performance and

Design", Report FHWA-TS-78-209, U.S. Department of Transportation, Washington,
1978, 145 p.

6. NAVFAC, Soil Mechanics (DM 7.1), Naval Facilities Engineering Command, Alexan-
dria, 1982, 355 p.

7. Baligh, M. M., "Cavity Expansion in Sands with Curved Envelopes", Journal of

the Geotechnical Engineering Division, ASCE, Vol. 102, No. GTll, Nov. 1976,
pp. 1131-1146.

8. Bolton, M. D., "The Strength and Dilatancy of Sands", Geotechnigue, Vol. 36,
No. 1, Mar. 1986, pp. 65-78.

9. Rowe, P. W., "The Relation Between the Shear Strength of Sands in Triaxial
Compression, Plane Strain and Direct Shear", Geotechnigue, Vol. 19, No. 1,
Mar. 1969, pp. 75-86.

10. Koerner, R. M., "Effect of Particle Characteristics on Soil Strength", Journal

of the Soil Mechanics and Foundations Division, ASCE, Vol. 96, No. SM4, July
1970, pp. 1221-1234.

11. Ladd, C. C., Foote, R., Ishihara, K., Schlosser, F., and Poulos, H. G.,
"Stress-Deformation and Strength Characteristics", Proceedings, 9th Inter-
national Conference on Soil Mechanics and Foundation Engineering, Vol. 2,
Tokyo, 1977, pp. 421-494.

12. Peck, R. B., Hanson, W. E., and Thornburn, T. H., Foundation Engineering, 2nd
Ed., John Wiley and Sons, New York, 1974, 514 p.

4-61
 13. Meyerhof, G. G., "Penetration Tests and Bearing Capacity of Cohesionless
Soils", Journal of the Soil Mechanics and Foundations Division, ASCE, Vol. 82,
No. SMl, Jan. 1956, pp. 1-19.

14. Schrnertmann, J. H., "Measurement of In-Situ Shear Strength", Proceedings, ASCE

Specialty Conference on In-Situ Measurement of Soil Properties, Vol. 2,
Raleigh, 1975, pp. 57-138. (closure: pp. 175-179).

15. Robertson, P. K. and Campanella, R. G., "Interpretation of Cone Penetration

Tests. Part I: Sand", Canadian Geotechnical Journal, Vol. 20, No. 4, Nov.
1983, pp. 718-733.

16. Villet, W. C. B. and Mitchell, J. K., "Cone Resistance, Relative Density, and
Friction Angle", Cone Penetration Testing and Experience, Eds. G. M. Norris
and R. D. Holtz, ASCE, New York, 1981, pp. 178-208.

17. Marchetti, S., "On the Field Determination of K0 in Sand", Proceedings, 11th
International Conference on Soil Mechanics and Foundation Engineering, Vol. 5,
San Francisco, 1985, pp. 2667-2672.

18. Hughes, J. M. 0., Wroth, C. P., and Windle, D., "Pressuremeter Tests in
Sands", Geotechnigue, Vol. 27, No. 4, Dec. 1977, pp. 455-477.

19. Mair, R. J. and Wood, D. M., Pressuremeter Testing, Butterworths, London,

1987, 160 p.

20. Durgunoglu, H. T. and Mitchell, J. K., "Static Penetration Resistance of

Soils", Proceedings, ASCE Specialty Conference on In-Situ Measurement of Soil
Properties, Vol. 1, Raleigh, 1975, pp. 151-189.

21. Schmertmann, J. H., "A Method for Determining the Friction Angle in Sands from
the Marchetti DMT", Proceedings, 2nd European Symposium on Penetration Test-
ing, Vol. 2, Amsterdam, 1982, pp. 853-861.

22. Mitchell, J. K., Fundamentals of Soil Behavior, John Wiley and Sons, New York,
1976, 422 p.

23. Tavenas, F. and Leroueil, S., "State-of-the-Art on Laboratory and In-Situ

Stress-Strain-Time Behavior of Soft Clays", Proceedings, International Sympo-
sium on Geotechnical Engineering of Soft Soils, Mexico City, 1987, pp. 1-46.

24. Mayne, P. W., "Stress Anisotropy Effects on Clay Strength", Journal of Geo-
technical Engineering, ASCE, Vol. 111, No. 3, Mar. 1985, pp. 356-366.

25. Nakase, A. and Kamei, T., "Undrained Shear Strength of Remolded Marine Clays",
Soils and Foundations, Vol. 28, No. 1, Mar. 1988, pp. 29-40.

26. Mayne, P. W. and Holtz, R. D., "Effect of Principal Stress Rotation on Clay
Strength", Proceedings, 11th International Conference on Soil Mechanics and
Foundation Engineering, Vol. 2, San Francisco, 1985, pp. 579-582.

27. Lupini, J. F., Skinner, A. E., and Vaughan, P. R., "The Drained Residual
Strength of Cohesive Soils", Geotechnigue, Vol. 31, No. 2, June 1981, pp. 181-
213.

28. Skempton, A. W., "Residual Strength of Clays in Landslides, Folded Strata, and
the Laboratory", Geotechnigue, Vol. 35, No. 1, Mar. 1985, pp. 3-18.

4-62
I
29. Lambe, T. W., "Amuay Landslides", Proceedings, 11th International Conference
on Soil Mechanics and Foundation Engineering, Jubilee Volume, San Francisco,
1985, pp. 137-159.

30. Skempton, A. W., Discussion of "Planning and Design of New Hong Kong Airport",
Proceedings, Institution of Civil Engineers, Vol. 7, June 1957, pp. 305-307.

31. Karlsson, R. and Viberg, L., "Ratio c/p in Relation to Liquid Limit and Plas-
ticity Index, with Special Reference to Swedish Clays", Proceedings, Confer-
ence on Shear Strength Properties of Natural Soils and Rocks, Vol. 1, Oslo,
1967, pp. 43-47.

32. Chandler, R. J., "The In-Situ Measurement of the Undrained Shear Strength of
Clays Using the Field Vane", Vane Shear Strength Testing in Soils: Field and
Laboratory Studies (STP 1014), ASTM, Philadelphia, 1988, pp. 13-44.

I 33. Bjerrum, L. and Simons, N. E., "Comparison of Shear Strength Characteristics

of Normally Consolidated Clays", ASCE Research Conference on Shear Strength of
Cohesive Soils, Boulder, 1960, pp. 711-726.

34. Wroth, C. P. and Houlsby, G. T., "Soil Mechanics - Property Characterization

and Analysis Procedures", Proceedings, 11th International Conference on Soil
Mechanics and Foundation Engineering, Vol. 1, San Francisco, 1985, pp. 1-55.

I 35. Jamiolkowski, M., Ladd, C. C., Germaine, J. T., and Lancellotta, R., "New
Developments in Field and Laboratory Testing of Soils", Proceedings, 11th
International Conference on Soil Mechanics and Foundation Engineering, Vol. 1,
San Francisco, 1985, pp. 57-153.

36. Mesri, G., "A Re-evaluation of Su(mob) ~ 0.22 ap Using Laboratory Shear
Tests", Canadian Geotechnical Journal, Vol. 26, No. 1, Feb. 1989, pp. 162-164.
I 37. Wood, D. M., "Index Properties and Critical State Soil Mechanics", Proceed-
ings, Symposium on Recent Developments in Laboratory and Field Tests and
Analysis of Geotechnical Problems, Bangkok, Dec. 1983, pp. 301-309.

38. Wroth, C. P. and Wood, D. M., "The Correlation of Index Properties with Some
Basic Engineering Properties of Soils", Canadian Geotechnical Journal, Vol.
15, No. 2, May 1978, pp. 137-145.

39. Skempton, A. W., "The Pore Pressure Coefficients A and B", Geotechnigue, Vol.
4, No. 4, Dec. 1954, pp. 143-147.

40. Mayne, P. W. and Stewart, H. E., "Pore Pressure Behavior of K0 -Consolidated

Clays", Journal of Geotechnical Engineering, ASCE, Vol. 114, No. 11, Nov.
1988, pp. 1340-1346.

41. Mayne, P. W., "Determining OCR in Clays from Laboratory Strength", Journal of
Geotechnical Engineering, ASCE, Vol. 114, No. 1, Jan. 1988, pp. 76-92.

42. Koutsoftas, D. C. and Ladd, C. C., "Design Strengths for an Offshore Clay",
Journal of Geotechnical Engineering, ASCE, Vol. 111, No. 3, Mar. 1985, pp.
337-355.

43. Hansen, L. A. and Clough, G. W., "Characterization of the Undrained Anisotropy

of Clays", Application of Plasticity and Generalized Stress-Strain in Geotech-
nical Engineering, Ed. by R. N. Yong and E. T. Selig, ASCE, New York, 1980,
pp. 253-276.

4-63
 44. Mayne, P. W., "A Review of Undrained Strength in Direct Simple Shear", Soils
and Foundations, Vol. 25, No. 3, Sept. 1985, pp. 64-72.

45. Prevost, J. H., "Undrained Shear Tests on Clays", Journal of the Geotechnical
Engineering Division, ASCE, Vol. 105, No. GTl, Jan. 1979, pp. 49-64.

46. Ladd, C. C. and Foott, R., "New Design Procedure for Stability of Soft Clays",
Journal of the Geotechnical Engineering Division, ASCE, Vol. 100, No. GT7,
July 1974, pp. 763-786.

47. Hight, D. W., Shibuya, S., and Symes, M. J., Discussion of "The Engineering
Application of Direct and Simple Shear Testing", Geotechnique, Vol. 38, No. 1,
Mar. 1988, pp. 139-140.

48. Graham, J., Crooks; J. H. A., and Bell, A. L., "Time Effects on the Stress-
Strain Behavior of Natural Soft Clays", Geotechnique, Vol. 33, No. 3, Sept.
1983, pp. 327-340.
I
49. Vaid, Y. P., Robertson, P. K., and Campanella, R. G., "Strain Rate Behavior of
Saint Jean Vianney Clay", Canadian Geotechnical Journal, Vol. 16, No. 1, Feb.
1979, pp. 34-42.
I
50. Lefebvre, G. and LeBoeuf, D., "Rate Effects and Cyclic Loading of Sensitive
Clays", Journal of Geotechnical Engineering, ASCE, Vol. 113, No. 5, May 1987,
pp. 476-489.
I
51. Eide, 0. and Holmberg, S., "Test Fills to Failure on Soft Bangkok Clay", Pro-
ceedings, ASCE Specialty Conference on Performance of Earth and Earth-Sup-
I
ported Structures, Vol. 1, West Lafayette, 1972, pp. 159-180.

52. Brand, E. W., Discussion of "Time Effects on the Stress-Strain Behavior of

Natural Soft Clays", Geotechnique, Vol. 34, No. 3, Sept. 1984, pp. 435-438.
I
53. Richardson, A.M. and Whitman, R. V., "Effect of Strain Rate Upon Undrained
Shear Resistance of a Saturated Remolded Fat Clay", Geotechnique, Vol. 13, No.
4, Dec. 1963, pp. 310-324.
I
54. Lacasse, S., "Effect of Load Duration on Undrained Behavior of Clay and Sand",
Report 40007-1, Norwegian Geotechnical Institute, Oslo, Apr. 1979, 118 p. I
55. Crawford, C. B., "The Influence of Rate of Strain on Effective Stresses in
Sensitive Clay", Papers on Soils (STP 254), ASTM, Philadelphia, 1959, pp. 36-
48.
I
56. Berre, T. and Bjerrum, L., "Shear Strength of Normally Consolidated Clays",
Proceedings, 8th International Conference on Soil Mechanics and Foundation
Engineering, Vol. 1.1, Moscow, 1973, pp. 39-49.
I
57. Akai, K., Adachi, T., and Ando, N., "Existence of a Unique Stress-Strain-Time
Relation of Clays", Soils and Foundations, Vol. 15, No. 1, Mar. 1975, pp. 1-16. I
58. Vaid, Y. P. and Campanella, R. G., "Time-Dependent Behavior of Undisturbed
Clay", Journal of the Geotechnical Engineering Division, ASCE, Vol. 103, No.
GT7, July 1977, pp. 693-709. I
59. Alberro, J. and Santoyo, E., "Long-Term Behavior of Mexico City Clay", Pro-
ceedings, 8th International Conference on Soil Mechanics and Foundation Engi-
neering, Vol. 1.1, Moscow, 1973, pp. 1-9.
I
4-64 I
 60. Nakase, A. and Kamei, T., "Influence of Strain Rate on Undrained Shear Charac-
teristics of K0 -Consolidated Cohesive Soils", Soils and Foundations, Vol. 26,
No. l, Mar. 1986, pp. 85-95.

61. Taylor, D. W., Fundamentals of Soil Mechanics, John Wiley and Sons, New York,
1948, 700 p.

62. Bishop, A. W. and Henkel, D. J., The Measurement of Soil Properties in the
Triaxial Test, Edward Arnold, London, 1962, pp. 174-179.

63. Perloff, W. H. and Osterberg, J. 0., "The Effect of Strain Rate on the Un-
drained Shear Strength of Cohesive Soils", Proceedings, 2nd Pan-American Con-
ference on Soil Mechanics and Foundation Engineering, Vol. l, Rio de Janeiro,
1963, pp. 103-128.

64. Olson, R. E., "Shear Strength Properties of a Sodium Illite", Journal of the
Soil Mechanics and Foundations Division, ASCE, Vol. 89, No. 1, Jan. 1963, pp.
183-208.

65. Bjerrum, L., "Embankments on Soft Ground", Proceedings, ASCE Specialty Confer-
ence on Performance of Earth and Earth-Supported Structures, Vol. 2, Lafay-
ette, 1972, pp. 1-54.

66. Bjerrum, L., "Problems of Soft Mechanics and Construction on Soft Clays", Pro-
ceedings, 8th International Conference on Soil Mechanics and Foundation Engi-
neering, Vol. 3, Moscow, 1973, pp. 111-159.

67. Aas, G., Lacasse, S., Lunne, T., and H~eg, K., "Use of In-Situ Tests for Foun-
dation Design on Clay", Use of In-Situ Tests in Geotechnical Engineering (GSP
&1. Ed. S. P. Clemence, ASCE, New York, 1986, pp. 1-30.
68. Mesri, G., Discussion of "New Design Procedure for Stability of Soft Clays",
Journal of Geotechnical Engineering, ASCE, Vol. 101, No. GT4, Apr. 1975, pp.
409-412.

69. Trak, B., LaRochelle, P., Tavenas, F., Leroueil, S., and Roy, M., "A New Ap-
proach to the Stability Analysis of Embankments on Sensitive Clays", Canadian
Geotechnical Journal, Vol. 17, No. 4, Nov. 1980, pp. 526-544.

70. Larsson, R., "Undrained Shear Strength in Stability Calculation of Embankments

and Foundations on Soft Clays", Canadian Geotechnical Journal, Vol. 17, No. 4,
Nov. 1980, pp. 591-602.

71. Djoenaidi, W. J., "A Compendium of Soil Properties and Correlations", M. Eng.
Sc. Thesis, University of Sydney, Sydney, 1985, 836 p.

72. Hara, A., Ohta, T., Niwa, M., Tanaka, S., and Banno, T., "Shear Modulus and
Shear Strength of Cohesive Soils", Soils and Foundations, Vol. 14, No. 3,
Sept. 1974, pp. 1-12.

73. Battaglio, M., Bruzzi, D., Jamiolkowski, M., and Lancellotta, R., "Interpreta-
tion of CPT and CPTU", Proceedings, Field Instrumentation and In-Situ Measure-
ments, Singapore, Nov. 1986, pp. 129-143.

74. Keaveny, J. M. and Mitchell, J. K., "Strength of Fine-Grained Soils Using the
Piezocone", Use of In-Situ Tests in Geotechnical Engineering (GSP-6), Ed. S.
P. Clemence, ASCE, New York, 1986, pp. 668-685.

4-65
75. Konrad, J. M. and Law, K. T., "Undrained Shear Strength from Piezocone Tests",
Canadian Geotechnical Journal, Vol. 24, No. 3, Aug. 1987, pp. 392-405.

76. Vesic, A. S., "Design of Pile Foundations", Synthesis of Highway Practice 42,
Transportation Research Board, Washington, 1977, 68 p.

77. Houlsby, G. T. and Teh, C. I., "Analysis of the Piezocone in Clay", Proceed-
ings, 1st International Symposium on Penetration Testing (ISOPT-1), Vol. 2,
Orlando, 1988, pp. 777-783.

78. Robertson, P. K., Campanella, R. G., Gillespie, D., and Greig, J., "Use of
Piezometer Cone Data", Use of In-Situ Tests in Geotechnical Engineering (GSP
&1. Ed. S. P. Clemence, ASCE, New York, 1986, pp. 1263-1280.
79.

80.
Baguelin, F., Jezequel, J. F., and Shields, D. H., The Pressuremeter and Foun-
dation Engineering, Trans Tech Publications, Clausthal, 1978, 617 p.

Marchetti, S., "In-Situ Tests by Flat Dilatometer", Journal of the Geotechni-

I
81.
cal Engineering Division, ASCE, Vol. 106, No. GT3, Mar. 1980, pp. 299-321.

Schmertmann, J. H., "Personal Communication", 12 Dec. 88 letter to F. H.

I
Kulhawy.

82. Lacasse, S. and Lunne, T., "Calibration of Dilatometer Correlations", Proceed-

ings, 1st International Symposium on Penetration Testing (ISOPT-1), Vol. 1,
I
Orlando, 1988, pp. 539-548.

83. Powell, J. J. M. and Uglow, R. S. T., "Dilatometer Testing in Stiff Overcon-

solidated Clays", Proceedings, 39th Canadian Geotechnical Conference, Ottawa,
I
1986, pp. 317-326.
I
I
I
I
I
I
I
I
4-66
I
I
I Section 5

ELASTIC DEFORMABILITY
I
I A knowledge of the so-called elastic behavior of soils is necessary for evaluating
the initial, time-independent, movement of foundations under static loads. These
deformation properties vary with many parameters and therefore are not defined

I uniquely. In this section, basic definitions are presented first to establish the
general background and notation. Methods for estimating Poisson's ratio are pre-

I sented next, followed by methods for estimating the soil modulus.

and cohesionless soils are included.
Both cohesive
Where available, typical values, influencing
factors, and in-situ test correlations are given. For the soil moduli, correla-

I tions with dynamic measurements also are given, even though the focus is on static
soil properties. The section is concluded with a brief discussion of the concept

I of subgrade reaction and evaluation of pertinent parameters for this concept.

BASIC DEFINITIONS

I The deformation properties of elastic materials are described most often by Young's
modulus (E) and Poisson's ratio (v). Although these parameters strictly are

I defined only for elastic materials under uniaxial loading, they are used commonly
in a "generic" sense with inelastic materials such as soils. These properties are
obtained most often from the results of triaxial compression tests. The modulus is
I the ratio of stress to strain and is obtained from the slope of deviator stress-
axial strain curves, as shown in Figure 5-l and given below:

I (5-l)

II in which (al - a3)

strain.
= deviator stress or principal stress difference and €a =

For any particular stress-strain curve, the modulus can be defined as the
axial

initial tangent modulus (Ei), the tangent modulus (Et) at a specified stress level,

'• or the secant modulus (Es) at a specified stress level.

curve.
pendent.
These moduli also will
vary with the confining stress (aa, ab, or ac in Figure 5-l) for each stress-strain
Therefore, soil moduli are described as being both nonlinear and stress-de-
In sophisticated numerical models, the actual stress path can be fol-

-I lowed, and the modulus can be evaluated for each stress state along the stress

5-l

I
 I
..,
b
b I
I
.....
--
0
0
>
Q)
I
0

Axial Strain, ~ 0
I
Figure 5-l. Modulus Definitions
I
path. In simpler, closed-form solutions, an effort must be made to estimate the
overall average modulus from the initial to the final stress states.
I
Poisson's ratio (v) is defined in an analogous form for triaxial tests in which
I
both axial and volumetric strains are measured. From these data, the axial and
radial strains can be obtained. Poisson's ratio is the ratio of the radial strain
(tr) to the axial strain (ta), as given below:
I
I
As with the modulus, Poisson's ratio is both nonlinear and stress-dependent.
ever, the range of v is relatively small compared with the range of E, and there-
How- I
fore less effort usually is made in evaluating v precisely.
I
For elastic materials, Young's modulus and Poisson's ratio are interrelated
uniquely with the shear modulus (G) as follows: I
G - E/2(1 + v) (5-3)
I
The shear modulus also is defined as the slope of the shear stress (r)-shear strain
(7) curve, which resembles that in Figure 5-l, and is given below: I
G - arja7 (5-4)
I
As with E and v, G is nonlinear and stress-dependent.
I
5-2
I
-I
I
I Another useful elastic parameter is the constrained modulus (M). This modulus is
defined for one-dimensional compression, where the lateral strains are zero, as
follows:
I
(5-5)

I in which av- vertical stress, €v- vertical strain, and mv- coefficient of volu-

•I
I
!
metric compressibility.

E(l -
M = (l + v)(l - 2v)
v)
From elastic theory, M is related to E and v as follows:

(5 -6)

-
The constrained modulus also is nonlinear and stress-dependent.

I POISSON'S RATIO

- Relatively little information is available in the literature for correlation stud-

ies with Poisson's ratio (v). However, this parameter does not vary greatly.
isotropic elastic materials, the entire range of v is from 0 to 0.5.
For
For dilatant

'

' soils that are inelastic, v may exceed 0.5.

the behavior is no longer elastic in this case.
However, it should be remembered that

' For undrained (¢ = 0) loading of saturated cohesive soil, no volume change occurs.
Therefore, the undrained Poisson's ratio (vu) is equal to 0.5 by definition.

For drained loading, volume changes occur, and the drained Poisson's ratio (vd)
varies with soil type and consistency. Typical values are given in Table 5-l,
which are representative of secant values at common design stress levels.

I For convenience in computer code implementation, Trautmann and Kulhawy (1) approxi-

•I
mated vd as follows:

' vd z 0.1 + 0.3 ¢rel (5-7)

i
with
I

(0 < ¢rel < l) (5-8)

'' in which ¢rel =

the soil density state.

relative friction angle that is convenient to use for approximating

5-3

:
 I
Table 5-l I
TYPICAL RANGES OF DRAINED POISSON'S RATIO

I
Soil Drained Poisson's Ratio, vd

Clay 0.2 to 0.4

I
Dense sand 0.3 to 0.4
I
Loose sand 0.1 to 0.3

I
An alternative approach is to use a hyperbolic model for the initial tangent
drained Poisson's ratio, as described by Kulhawy, et al. (1). This value is
I
expressed as:
I
(5-9)

in which a3c =minor principal effective confining stress, and~ and Fv are hyper-
I
bolic parameters given in Figure 5-2.
I
For cohesive soils, the drained Poisson's ratio also has been related to plasticity
index for several lightly overconsolidated (LOG) soils, as shown in Figure 5-3. I
However, vd also is nonlinear and stress-dependent, as shown in Figure 5-4 for one
clay as a function of stress level (amount of the failure stress mobilized) and
OCR. As can be seen in these two figures, the variation of vd is not great.
I
UNDRAINED MODULUS OF COHESIVE SOILS I
Cohesive soils exhibit time-dependent response to loading. For initial quick load-
ing conditions, the response is undrained. With time, the excess pore water
stresses developed during undrained loading will dissipate, leading to consolida-
I
tion and other long-term phenomena. These time-dependent phenomena and associated
soil properties are described in Section 6.
I
For undrained loading, the modulus of cohesive soils can be described by either the
undrained Young's modulus (Eu) or the shear modulus (G). The shear modulus actu-
I
ally describes the soil "skeleton" response, so it is independent of drainage con-
ditions, all other factors being equal. For undrained loading, Eu is equal to 3G
I
5-4 I
I
I 1.2 Sands
I
0
I
Dr= Relative
Density
\7 Cobble
0 GW
(> GP
!::.
0 SP
sw

10
1.0 1.0
I G11 0.8
0 9- /
06
0
0.8 Sands
/
Dr 0.6 o oo Fit 0.6 0 /0

I 0.4
0.2
0
/
tfi
/
Dr 0.4
0
CO/
/0
o·0 o
/

0.2
0 -t::6J---
I o~~~~--~~~--~~~~~
0 0.2 0.4 0.6 0.8 1.0
0
0 0.2 0.4 0.6 0.8 1.0
0.8.-.--.-.--.-.--.-.--r-~~~

I 0.6

G11 0.4
'<>
•JO.
w,
0
[J
Gravel and Rockfill

6_
0.6
0.4 Gravel and Rockfill
<o-fb OJ F
I 0.2
OL-~~~--~-L~--L-~-L--~
'-[]'- - < > - - \ 7 - -\7
\7
II 0.2
0
-o8-<trEtm
0
-81- o- - \ 7 -
\7
\7

0 0.2 0.4 0.6 0.8 1.0 0 0.2 0.4 0.6 0.8 1.0

I Figure 5-2.
Initial Void Ratio, ei
Drained Poisson's Ratio Parameters for Granular Soils

'I Source: Kulhawy (}), p. 76.

2/3

I 0
0
I /3

::::0
~ 0.6 I I I I I cr Stress level

I 0
(/)

.--· ------·
c:

_...--.-·
0
0 (/)
(/)
cr 0.4 ·c;
(/) a.. 0.2
I c:
0
(/)
(/)
·c;
a..
- -o
Q)
c:
·c;
0.2 0 0.1

I -o

0
Q)
c:
·c;
,_
-
o~~--~~--~--~~--~~
o~~--~--~·~·---·~-~·~·--~
0 20 40 60 80 I 3 5 9

I Plasticity Index, PI (%) OCR

Figure 5-3. Drained Poisson's Ratio Figure 5-4. Drained Poisson's Ratio

I versus PI for Several LOG

Soils
versus OCR and Stress Level
for Sydney Kaolin

Source: Wroth (~), p. 187. Source: Poulos (2), p. 104.

I
I 5-5

I
 I
from Equation 5-3 since vu- 0.5.
I
It should be noted that the factors affecting su (discussed in Section 4) also will
affect Eu. Therefore, the value of Eu will be dependent on test type and test spe- I
cifics.

Typical Values
I
A number of authors have given typical ranges for the undrained modulus, and these
ranges are summarized in Table 5-2. These values generally would be representative
I
of secant moduli at common design stress levels.

As an alternative, Kulhawy, et al. (£) suggested use of a hyperbolic model to esti-

I
mate the undrained tangent modulus (Eut), as given below:
I
(5-10)

in which Eui = undrained initial tangent modulus, SL = stress level (fraction of

I
strength mobilized), ac =isotropic confining stress, a1 =total major principal
stress, a3 = total minor principal stress, and K, n, and Rf = modulus parameters I
given in Table 5-3. For CIUC or CAUC test conditions, ac would equal the minor
principal effective confining stress (o3c)· For UU test conditions, ac would equal I
Correlations with su
I
More commonly, the undrained modulus (Eu) is normalized directly by the undrained
I
Table 5-2 I
TYPICAL RANGES OF UNDRAINED MODULUS FOR CLAY

I
Normalized Undrained
Consistency Modulus, Eu/Pa
I
soft

medium
15 to 40

40 to 80
I
stiff 80 to 200
I
5-6
I
J
I
I Table 5-3

TYPICAL UNDRAINED HYPERBOLIC MODULUS PARAMETERS

I Unified Soil
Classification n Rf

I CL 100 to 200 l 0.9

I Source:
CH 100 to 300 l

Kulhawy, et al. (§)' p. 10-19.

0.9

I
I shear strength (su) from the same test to give Eu/Su·
independent of test type.
This ratio is assumed to be
Also common is the rigidity index (Ir), which is defined
as the ratio of the shear modulus (G) to the strength. For undrained (¢ = 0) load-

I ing, Ir is given as:

I (5-ll)

For undrained loading, Eu is equal to 3G and therefore:

I (5-12)

I Figure 5-5 illustrates typical test results obtained for a number of cohesive
soils. (The numbered soils were defined in Figure 4-33.) These data were obtained

I from direct simple shear tests and illustrate the range of the secant undrained
modulus ratio (Eus/su) as a function of stress level (given as shear stress ratio)

I and OCR. Based on data such as these, Duncan and Buchignani

broad generalization shown in Figure 5-6.
(~) suggested the

I Alternatively, the modified Cam clay model can be used to provide an estimate of
the undrained modulus ratio. Wroth, et al. (J) suggested the following:

I (Eu/su)oc (G/su)oc
[l + C ln OCR] OCR-A (5-13)
(Eu/su)Nc (G/su)Nc

I in which C is an experimentally determined constant that is likely to be between 0

I and 2. A value of C = l appears to be representative of the trends observed in

I 5-7

I
 I
2000
r~oy_
~........
<", (wL-35
Sen:il;,~
', ,._PI=I5)
i'o,
1000 500
I
~'-Q"-
I
'-., 5_ At Th/su=l/3 At Th/su =2/3

(/)
:l

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

' '
'<}· ..........
''··..s ··., . . . . .
'· '
....

·,•, ..., ......" ' 5

"\
\
5
- "'\
\
\
w
en
:l
'
' '-
...... ', 2
......

''
·, I
..... 4
'·
Q.....
' ' ....
(;;-'-'
\
\
\ I
' ' '-3 '·~. ' ' \
200
3
--- ...... .' ' \
0... \.
I
20 0
Clays
0
-
............... -co-...... ::--..:

0.2 0.4
Shear Stress Ratio, Th /su
0.6 0.8 I 2
OCR
5 10 I 2
OCR
5 10
I
Figure 5-5. Normalized Undrained Modulus versus Stress Level and OCR

Source: Ladd, et al. (2), pp. 436, 441.

I
I
I
I
:::>
(/)
........
II)
:l
w

I
I
Overconsolidation Ratio, OCR

Figure 5-6. Generalized Undrained Modulus Ratio versus OCR and PI I

Source: Duncan and Buchignani (~), p. 26.
I
laboratory test data.
I
The original Cam clay model can be used to provide an estimate of the undrained
initial tangent modulus ratio for normally consolidated clay (Eui/su)NC·
relationships given by Mayne and Swanson (10), the initial tangent modulus can be
Using
I
evaluated by differentiation as the strain approaches zero.
be normalized by su as given by the Cam clay model, resulting in the following:
This modulus then can
I
5-8
I
J
I
2M(l + e 0 ) ln 10
I (Eui/su)Nc = Cc A(l - A) exp(-A) (5-14)

I in which Cc virgin compression index (See Section 6.), e 0 =initial void ratio,
and M and A are defined in Appendix G. This equation corresponds to CIUC triaxial
compression conditions. Using a typical value of A= 0.8, Equation 5-14 simplifies
I as follows:

I (5-15)

Combining Equations 5-15 and 5-13 (with A 0. 8 and C 1) gives:

I
64M(l + e 0 )[1 + ln OCR)

I Cc OCR0.8
(5-16)

I For e 0 = 1, Figure 5-7 shows the relationship for Eui/su in terms of OCR, Cc, and
~tc· This figure is similar in form to Figure 5-6, but it is based on more funda-
mental soil properties. The tangent modulus at a particular stress level then can

I be computed from Equation 5-16, using the stress level (SL) reduction as given in

I Com Cloy
( CIUC triaxial)
I (e 0 = I)

I
I w
::>

I
Cc
I }o.l
}0.2

I 0
I 2 5 10 20
} 0.5
50 100
OCR

I Figure 5-7. Cam Clay Prediction of Undrained Initial Tangent Modulus Ratio

I
5-9

I
 I
Equation 5-10. Furthermore, the limit value of the secant modulus ratio approach-
ing zero stress level would be Eui/su. This value then can be used to estimate I
Eus/su at a particular stress level using the experimental relationships shown in
Figure 5-5.
I
Correlations with SPT, CPT, and PMT Results

Apparently, few studies have attempted to relate the undrained modulus (Eu) to the
I
SPT N value or the CPT cone tip resistance in cohesive soils. Ironicaliy, many
efforts have instead attempted to correlate the constrained modulus (M
under drained conditions to the N value and qc, although these penetration resis-
= 1/mv) I
tances occur most likely under undrained conditions.
discussed in Section 6.
These relationships will be
I
The pressuremeter test (PMT) provides a measurement of the horizontal modulus in
soils. In clays, it is assumed commonly that EPMT = Eu. For practical use,
I
attempts have been made to correlate EPMT with the SPT N value, as shown in Figure
5-8. Based on these data, it is clear that more than an order of magnitude varia-
I
tion is possible when using N values as the sole predictor.
I
500 I
200 I
100

0.0
......... 50
I
t-

w
:;!
a..
20
I
10
Tokyo
Alluvial
0
Cloy
Oiluviol
•
I
Nagoya 0 •
5 0

0
0 8 Osaka
Sokoide
A
<>
"'
• I
2 5 20 50 100 200
SPT N Value
I
Figure 5-8. PMT Modulus of Clay versus N Value

Source: Ohya, et al. (11), p. 129. I

5-10
I
I
I
I Back-Figured from Full-Scale Load Tests

Perhaps more useful than the in-situ test results are moduli back-figured from

I analysis of full-scale field load tests.

on limited data for driven piles and drilled shafts.
Figure 5-9 shows an interpretation based

I
l
Figure 5-lOa includes more data for drilled shafts as a function of depth (D) to

•I ll
• \ ~verage for

r-•-o
• ~rilled shafts

"r/-~--1>--­
"'
0
a.
"'
j' #e
I
........ !Average for
0
w"'
::> " I / driven piles
I /ll
ll 1 0

1 0/ "

~
100 Driven piles
ll/
~~ o Drilled shafts
50 • Drilled shafts in
London cloy
0

'•
0 2

!
Figure S-9. Undrained Modulus for Deep Foundations in Compression

Source: Poulos and Davis (12), p. 103.

Eus /su
0 1000 2000 3000 0 500 1000
Qt I • •~Mean
I I
0
I I
I .._ / !--Lower range
o•!ii.lf:>ll.o.t. ~
I• 1• I
I 'Jhrf • - • (!) 0.5 1-
I -
(!)
........
0
lOr-I" l
0~
0 /
y o
/
-
........
0 I
\Mean
?
l 10 Lower range 0 I 0/
0 o".

-
0 /
~Upper
0
1.0 -
10
0 -
I o oI
lo range /
0
a:: I• ~ a:: o I
..c.
0.
Q)
20 I•
I
0

0
I
Compression
o Streight _
c. Belled
-
..c.
a.
Q)
1.5 r-
Jg,I /c /\
0 • I Uplift 0 f Uppe• range -
I• • I
I
•I •lo •
Streight
I•
Belled I
0
(a) ( b)
30 • I, I 2.0 I I

Figure 5-10. Undrained Modulus for (a) Drilled Shafts in Compression and Uplift and
(b) Spread Foundations in Uplift

Source: Callanan and Kulhawy (13), pp. 3-28, 3-33.

5-ll
 I
diameter (B) ratio.
greater than 200.
Of particular interest to note is that Eus/su is normally
Figure S-lOb shows limited data for spread foundations with I
cohesive soil backfill. In this iigure, avm- mean vertical total stress over the
foundation depth.
able.
Although the data are limited, the range appears to be reason- I
Lastly, from analyses of the axial deformation of piles at working load levels, I
Randolph (14) suggested the following range for rigidity index (Ir = Gfsu):

150 < Ir < 200 (5-17)

I
For lateral loads, the range was suggested to be: I
75 < Ir < 100 (5-18) I
These generalized ranges are intended to be representative of common, simple design
situations. I
Estimation from Dynamic Measurements
I
Another method for estimating the modulus is based on shear wave velocity measure-
ments from the resonant column test. Hardin and Drnevich (15) developed the fol-
lowing equation to evaluate Gmax at low-amplitude (dynamic) shear strains:
I
321 (2.97 - e)2 ocRM (- I )0.5
1 + e ao Pa (5-19)
I
in which e =void ratio (not to exceed 2), M =exponent given in Table S-4, and a0 I
= mean principal effective stress.

However, it must be remembered that Gmax at small dynamic strains is much larger
I
than G at large static strains, as shown in Figure 5-11. From this figure, it is
clear that G for static loading is on the order of 5 to 10 percent of Gmax for I
dynamic loading. This general pattern holds for all soil types.

Wroth, et al. (2) reviewed a number of relationships for Gmax at dynamic strains
I
versus N, as shown in Figure 5-12.
scatter is present in the data.
From this figure, it is clear that considerable
From these data, they suggested the following: I
Gmax /p a = 120 N0.77 (5-20) I
5-12
I
I
I
I Table 5-4

EXPONENT M FOR SHEAR MODULUS

I
Plasticity Index, PI Exponent, M

I 0 0

I 40
20 0.18

0.30

I 60

80
0.41

0.48

I Source:
> 100 0.50

Hardin and Drnevich (15), p. 672.

I Resonant column Static plate

I
test bearing test

I <.9
><
0
E
0.5
........

I <.9

I Shear Strain, y (%)

I Figure 5-11. Shear Modulus versus Shear Strain for Sands

Source: Seed and Idriss (16).

I
with limits of the data being 60 N0.71 < GmaxiPa < 300 N0.8. The static shear mod-

I ulus then would be some 5 to 10 percent of the computed Gmax value.

I MODULUS FOR COHESIONLESS SOILS

Cohesionless soils such as sands do not exhibit significant time-dependency to

loading caused by excess pore water stress dissipation, and therefore the modulus
I under undrained loading conditions exists only briefly. Almost always, the modulus

I 5-13

I
 I
I
2000
I
a.
0

-.......
><
0
I
E

I
<.9

Suggested limits
by Wroth, et ol., 1979
I
50~--~-----L--~----~-----L--~
I 2 5 10 20
SPT N Value (blows/ft or 305 mm)
50 100
I
Figure 5-12. Dynamic Shear Modulus versus N for Cohesive Soils

Source: Wroth, et al. (2), p. 96. I

is considered for drained conditions. However, for finer-grained silts, some sig-
I
nificant time-dependency may develop which will have to be considered on a case-by-
case basis. I
For drained loading, the modulus can be described by the drained elastic modulus
(Ed), the shear modulus (G), or the drained constrained modulus (Md). E and G com-
I
monly are evaluated in triaxial compression, while M is evaluated in one-dimen-
sional compression. All of these are interrelated through Poisson's ratio, as I
noted previously in Equations 5-3 and 5-6. Unless otherwise stated, the moduli
will be secant values given by Eds and Mds·
I
Typical Values

A number of authors have given typical ranges for the modulus of cohesionless
I
soils. Table 5-5 is representative of these ranges for sands in general and for
driven piles in particular. These values generally would be representative of I
secant moduli within common design stress levels.

Alternatively, Duncan and Chang (18) suggested a hyperbolic model to estimate the
I
drained tangent modulus, starting from an initial isotropic stress, as follows:
I
5-14
I
I
I
I Table 5-5

TYPICAL RANGES OF DRAINED MODULUS FOR SAND

I
Normalized Elastic Modulus, Ed/Pa

I Consistency Typical Driven Pilesa

I loose

medium
100 to 200

200 to 500
275 to 550

550 to 700

I a
dense

- Source:
500 to 1000

Poulos (17), p. 207.

700 to 1100

I
(5-21)
I in which a1 and a3 = effective major and minor principal stresses, respectively,

I ¢tc =effective stress friction angle in triaxial compression, and

modulus parameters given in Table 5-6.
K, n, and Rf =
For convenience in computer code implemen-
tation, Trautmann and Kulhawy (l) approximated K as follows:
I K ~ 300 + 900 ¢rel (5-22)

I with ¢rel defined in Equation 5-8.

I Correlations with Strength

The shear modulus commonly is correlated to the effective soil strength through the

I rigidity index (Ir), as defined below for drained loading:

I Gj(a tan "4>tc) (5- 23)

Selected values for Ir are given in Teble 5-7. Of particular interest to note is

I that Ir increases with increasing relative

normal stress.
densi~y and decreases with increasing
It also is lower with more compressible soil minerals.

I When using the rigidity index (Ir) for drained loading, volume changes normally
have to be considered. Therefore, Ir must be corrected for the volumetric strains

I (€v) to yield a reduced rigidity index (Irr), as given below by Vesic (20):

I 5-15

I
 I
Table 5-6 I
TYPICAL DRAINED HYPERBOLIC MODULUS PARAMETERS
I
Unified Soil
Classification n Rf I
GW

GP
300 to 1200

500 to 1800
1/3

1/3
0.7

0.8
I
sw
SP
300 to 1200

300 to 1200
1/2

1/2
0.7

0.8
I
ML

Source:
300 to 1200 2/3

Kulhawy, et al. (§_), p. 10-19.

0.8
I
I
Table 5-7

VALUES OF RIGIDITY INDEX FOR SELECTED COHESIONLESS SOILS I

Soil
Relative Density
Dr (%)
Normalized Mean
Normal Stress, 0 o1Pa
Rigidity
Index, Ir
I
Chattahoochee
sand
80 0.1 200 I
80 1 118
80
80
10
100
52
12
I
20
20
0.1
1
140
85 I
Ottawa sand 82
21
0.05
0.05
265
89 I
Piedmont silt 0.70 10 to 30

Source: Vesic (20), p. 68.

I
I
(5-24)

Vesic (20) noted that Ev would be zero for dense $Oils and range from 0 to 0.05 for
I
5-16 I
I
I
loose soils in the stress range from 1 to 10 atmospheres. For convenience in com-
puter code implementation, Trautmann and Kulhawy (!) approximated €v as follows:

I (5-25)

I I
in which ov- vertical effective stress (up to 10 atmospheres), and
in Equation 5-8.
~el is defined

I Correlations with SPT N Value

Young's Modulus. Early correlations in the literature related Eds of sands direct-

I ly to the standard penetration test (SPT) N value.

are shown in Figure 5-13.
Several of these correlations
Others within the same ranges are given by Mitchell and

I Gardner (23). Later correlations attempted to relate the constrained modulus (M)
and Nasa function of overburden stress (e.g., 24).

I However, all attempts to date which correlate a modulus with N show considerable
scatter. This lack of correlation is to be expected because the SPT N value varies

-I )
with many factors, as described in Section 2, and these factors have yet to be
incorporated in these correlations.
following may be used:

(sands with fines)

Therefore, as a first order estimator, the

(5-26a)

' (clean NC sands) (5-26b)

DENSE V.DENSE
1200 LOO~E . M~DIUM.
0 I - 0 Appolomo, 0 Appolonto,
0.
and Brisette, 1970 ( £!.l
' "'
"U 2- Schmertmonn, 1970 (22l
w
800 Table 5-5

10 20 30 40 50 60

N (blows/ft or 305 mm)

Figure 5-13. Comparative Plot of Drained Modulus Correlations for Sand

Source: Callanan and Kulhawy (13), p. 3-16.

5-17
 I
(clean OC sands) (5-26c) I
in which N6o is the N value corrected for field procedures to an average energy
ratio of 60 percent. Equation 2-11 gives the appropriate correction factors.
I
Pressuremeter Modulus. The pressuremeter test (PMT) provides a direct measurement
of the horizontal modulus of cohesionless soils. This modulus (EPMT) often is pre-
I
sumed to be roughly equivalent to Young's modulus (E).
value and EPMT have been developed, as shown in Figure 5-14.
Correlations between theN
The scatter shown is
I
typical of other N correlations because of the reasons noted above.

Dilatometer Modulus. The dilatometer test (DMT) also provides a direct modulus
I
measurement for cohesionless soils.
Young's modulus as follows:
The dilatometer modulus (Eo) is related to
I
Eo - E/(1 - v2) (5-27) I
No general correlations of ED with N have been presented at this time.
the DMT and other in-situ tests can be used effectively to develop convenient
However,
I
I
200
E
PMT =g_QB N0.66
Po
I
(n= 370, r 2=0.482,
100 S.D.=61.3p0 ) I
0.0
........

w
1-
:::E
a.
50

D
D
I
20

10 Sand
I
I
Alluvial Diluviol
0 •
5 D
c:. 4
•
¢ +
0.5 2 5 10
SPT N Value
20 50 100 200
I
Figure 5-14.

Source:
PMT Modulus of Sand versus N Value

Ohya, et al. (11), p. 129.

I
5-18 I
I
I correlations within a specific geologic setting. For example, Mayne and Frost (25)
developed the relationship shown in Figure 5-15, which correlates the SPT N value
with both the Eo and the secant modulus (Eds) back-calculated from eight case his-
I tories of field performance data on building foundations. All of these data were
obtained in sandy silts of the Piedmont geologic province, in and around the Wash-

I ington, -D.C. area. Local correlations of this type normally are much more accurate
than generalized global correlations.

I Correlations with CPT qc Value

I Modulus values for cohesionless soils have been correlated with the cone penetra-
tion test (CPT) qc value. Initial correlation studies attempted to link Eds with
qc directly, using the general form below:

I (5-28)

I in which a = empirical parameter and Eds and qc are in the same units. Webb, et
al. (26) have shown that existing relationships suggest a values ranging from 1.5

I to 2.5 and intercepts for EdsiPa ranging from 0 to 80.

viously, the qc relationships actually are nonlinear and stress-dependent.
However, as noted pre-

I The majority of studies actually have focused on the tangent constrained modulus
(Mctt) instead of Young's modulus (E), primarily because Mdt corresponds to

I
I ~
'-
w
0
0

E 0 /p0 o 0.22 N°·

82
~

~¥it~·
600

I
I ....
Q)
(i;
400
f
• ~/

/.
•
E
.2
200
............ +• E 0 from 9 sites

I 0
0 • /~ •

0 ··~--~--~--~--~--~--4---~--~--~--~~
Eback- calculated
from case histories

0 10 20 30 40 50 60
I SPT N Value (blows /ft or 305 mm)
Figure 5-15. Trend Between Dilatometer Modulus and N in Piedmont Sandy Silts

I Source: Mayne and Frost (25), p. 22.

I 5-19

I
 I
one-dimensional compression and is easier to determine.
take the form:
The correlations typically
I
(5-29) I
in which a - empirical parameter and Mdt and qc are in the same units. Values of a
quoted in the literature typically range from 3 to 8 for normally consolidated (NC)
I
sands. However, Figure 5-16a shows further ranges in a and a definite trend with
relative density. These data were obtained from the calibration chamber studies I
reported in Appendix H.

For overconsolidated (OC) sands, a is much higher. Values quoted in the literature
I
typically range from 7 to 25 or more. However, Figure 5-16b shows further ranges
in a and a definite trend with relative density. These data also were from the I
calibration chamber studies.

For one sand tested extensively in a calibration chamber, the effects of relative
I
density, overconsolidation, and stress level adopt consistent patterns, as shown in
Figure 5-17. These patterns can be used as guidelines for other sands. I
Figures 5-16 and 5-17 show that the modulus is a function of relative density.
Section 2, it was shown that the relative density is a function of the cone tip
In
I
resistance normalized by (ov 0 )0.5. Therefore, the modulus should have the same
proportionality with the effective vertical stress, as shown by Janbu (28). This I
issue will be discussed further in Section 6.

Back-Figured from Full-Scale Load Tests

I
Perhaps more useful than the in-situ test results are moduli back-figured from
analysis of full-scale field load tests. Figure 5-18a shows secant modulus values
I
from analyses of drilled shafts in uplift, where the modulus was normalized by the
mean unit side resistance (f). Comparable data are shown in Figure 5-18b for I
spread foundations in uplift, except that the modulus was normalized by the mean
vertical effective stress over the foundation depth.

and Edsfovm is 200.

A lower bound on both Eds/f
I
Estimation from Dynamic Measurements I
Another method of estimating the modulus is based on shear wave velocity measure-
ments from the resonant column test. These measurements have been used to evaluate I
5-20
I
I
I
I 20 (a) 9 NC sands

I
u
CT

I -.....

2
II
"0 5

I
I 60 80 100
Relative Density, Dr (%)
I
50 (b) 4 OC sands
I 0 0

I
I
I
I
I Relative Density, Dr (%)
Figure 5-16. Variation of a with Dr for Sands in Calibration Chambers

I
I Gmax at low-amplitude (dynamic) shear strains. As shown with Figure 5-11, this
dynamic modulus represents an upper bound, and therefore it is denoted Gmax· For
static loading with relatively large strains, G is on the order of 5 to 10 percent

I of Gmax·

I Early laboratory studies on rounded and angular sands gave the relationships shown
in Figure 5-19. More recent studies by Hardin (30) suggested the following:

I 5-21

I
 I
I
I
I
I
II

I
I
I
Overconsolidation Ratio, OCR

Figure 5-17. CPT a Correlation for Ticino Sand

I
Source: Jamiolkowski, et al. (27), p. 277.
I
0
Ed/ Unit Side Resistance, f
2500 5000
Ed/Meon Vertical Effective Stress, <fvm
0 1000 2000
I
0 1 1 0 I
l ...
I.
I
•• • •
•
•
•
0
o
•
Tamped
Unknown placement
Untamped
I
o..!I • • -
co
0--
1
I
I
co
--
0 I
.2
.2
;:g i--
I

..c=
c;_
11>
20
I •
I
Lower range
-
0
0::
..c=
c;_
11>
I
o I • Straight D

30~
• Belled
- I
I'· (a)

I
I I

Figure 5-18. Normalized Drained Modulus for (a) Drilled Shafts in Uplift and (b)
Spread Foundations in Uplift

Source: Callanan and Kulhawy (13), pp. 3-30, 3-36. I

5-22
I
I
-
I
I a.
'-
0

>(
2000
0

I <.9
.n
E

:::;1
1000

:::;1
500

I -o
~
0

'-
0
200
(]) --Round groins

I .r:.
c.n 100

0.05 0.1 0.2

- - - Angular groins

2 5
I Confining Stress, o:3 /p0
Figure 5-19. Variation of Shear Modulus of Dry Sands with Void Ratio and Confining
Stress

I Source: Richart, Hall, and Woods (29), p. 385.

I
GmaxiPa = (5-30)

I 2(1 + v)(0.3 + 0.7e2)

in which S = stiffness coefficient, M = exponent, a0 = mean principal effective

I stress, and e =void ratio. Hardin (30) notes that OGRM often is taken as 1 for
convenience and that S for clean sands is in the range of 1200 to 1500.

I SUBGRADE REACTION

I In contrast with elastic theories that use Young's modulus (E), an alternative
method for analyzing load-displacement response is the concept of subgrade reac-
tion. This concept is used often for evaluating the behavior of footings, mat/raft

I foundations, and laterally loaded deep foundations.

there is a basic parameter which is analogous to a spring constant.
In subgrade reaction models,
This parameter

I is defined as the modulus of subgrade reaction (ks), given by:

ks - p/6 (5-31)

I in which p = applied stress and 6 - displacement under p. By this definition, ks

I is in units of force per length cubed.

stress level.
As with Young's modulus, ks varies with
However, unlike Young's modulus, ks also varies with foundation
width (e.g., Horvath, 31).

I
I 5-23

L
 I
To account for this width dependence, another subgrade reaction modulus (K5 ) was I
introduced as below:

(5-32)
I
in which K5 has units of force per length squared and B - foundation width. For I
deep foundations where k 5 varies with depth, z (and k 5 sometimes is known as kh),
an alternative coefficient of subgrade reaction (nh) sometimes is used, as given by:
I
(5-33)
I
Perhaps the most logical procedure to evaluate k 5 is to present it in terms of
Young's modulus (E) and Poisson's ratio (v) of the soil.
k 5 in this manner and determined the following:
Vesic (32) reinterpreted
I
(5-34)
I
I
in which Ef- foundation Young's modulus, If- foundation moment of inertia, and
Ef I f - foundation stiffness.
foundation material and geometry.
Ef and If normally are constants depending on the
Procedures for evaluating E and v were presented
I
earlier in this section.
I
REFERENCES

1. Trautmann, G. H. and Kulhawy, F. H., "GUFAD - A Computer Program for ~ompres­

sion and ~plift foundation ~nalysis and Qesign", Report EL-4540-CGM, Vol. 16,
I
Electric Power Research Institute, Palo Alto, Oct. 1987, 148 p.

2. Kulhawy, F. H., Duncan, J. M., and Seed, H. B., "Finite Element Analysis of
Stresses and Movements in Embankments During Construction", Contract Report
I
S-69-8, U.S. Army Engineer Waterways Experiment Station, Vicksburg, Nov. 1969,

3.
169 p.

Kulhawy, F. H., "Finite Element Analysis of the Behavior of Embankments",

I
Ph.D. Dissertation, University of California, Berkeley, 1969, pp. 40-80.

4. Wroth, G. P., "In-Situ Measurement of Initial Stresses and Deformation Charac-

I
teristics", Proceedings, ASCE Specialty Conference on In-Situ Measurement of

5.
Soil Properties, Vol. 2, Raleigh, 1975, pp. 180-230.

Poulos, H. G., "Normalized Deformation Parameters for Kaolin", Research Report

I
R336, School of Civil Engineering, University of Sydney, 1978, 18 p.

I
5-24 I
I
I
I 6. Kulhawy, F. H., Trautmann, C. H., Beech, J. F., O'Rourke, T. D., McGuire, W.,
Wood, W. A., and Capano, C., "Transmission Line Structure Foundations for
Uplift-Compression Loading", Report EL-2870, Electric Power Research Insti-
tute, Palo Alto, Feb. 1983, 412 p.

I 7. Ladd, C. C., Foott, R., Ishihara, K., Schlosser, F., and Poulos, H. G.,
"Stress-Deformation and Strength Characteristics", Proceedings, 9th Interna-
tional Conference on Soil Mechanics and Foundation Engineering, Vol. 2, Tokyo,

I 8.
1977, pp. 421-494.

Duncan, J. M. and Buchignani, A. L., An Engineering Manual for Settlement

Studies, Department of Civil Engineering, University of California, Berkeley,
I 9.
June 1976, 94 p.

Wroth, C. P., Randolph, M. F., Houlsby, G. T., and Fahey, M., "A Review of the

I Engineering Properties of Soils with Particular Reference to the Shear Modu-

lus", CUED/D- SOILS TR 75, University of Cambridge, 1979, 79 p.

10. Mayne, P. W. and Swanson, P. G., "The Critical State Pore Pressure Parameter

I from Consolidated-Undrained Shear Tests", Laboratory Shear Strength of Soil

(STP 740), Ed. R. N. Yong and F. C. Townsend, ASTM, Philadelphia, 1981, pp.
410-430.

I 11. Ohya, S., Imai, T., and Matsubara, M., "Relationships Between N Value by SPT
and LLT Pressuremeter Results", Proceedings, 2nd European Symposium on Pene-
tration Testing, Vol. 1, Amsterdam, 1982, pp. 125-130.

I 12. Poulos, H. G. and Davis, E. H., Pile Foundation Analysis and Design, John
Wiley and Sons, New York, 1980, 397 p.

13. Callanan, J. F. and Kulhawy, F. H., "Evaluation of Procedures for Predicting

I Foundation Uplift Movements", Report EL-4107, Electric Power Research Insti-
tute, Palo Alto, Aug. 1985, 124 p.

I 14. Randolph, M. F., "PIGLET- A Computer Program for the Analysis and Design of
Pile Groups Under General Loading Conditions", Engineering Department Report,
University of Cambridge, July 1983, 69 p.

I 15. Hardin, B. 0. and Drnevich, V. P., "Shear Modulus and Damping in Soils: Design
Equations and Curves", Journal of the Soil Mechanics and Foundations Division,
ASCE, Vol. 98, No. SM7, July 1972, pp. 667-692.

I 16. Seed, H. B. and Idriss, I. M., "Soil Moduli and Damping Factors for Dynamic
Response Analysis", Report EERC 70-10, University of California, Earthquake
Engineering Research Center, Berkeley, Dec. 1970.

I 17. Poulos, H. G., "Settlement of Isolated Foundations", in Soil Mechanics-

Recent Developments, Eds. S. Valliappan, S. Hain, and I. K. Lee, William H.
Sellen Pty., Zetland, 1975, pp. 181-212.

I 18. Duncan, J. M. and Chang, C.-Y., "Nonlinear Analysis of Stress and Strain in
Soils", Journal of the Soil Mechanics and Foundations Division, ASCE, Vol. 96,
No. SMS, Sept. 1970, pp. 1629-1653.

I 19. Vesic, A. S., "Design of Pile Foundations", Synthesis of Highway Practice 42,
Transportation Research Board, Washington, 1977, 68 p.

I
I 5-25

L
 I
20. Vesic, A. S., "Bearing Capacity of Shallow Foundations", Chapter 3 in Founda-
tion Engineering Handbook, Eds. H. Winterkorn and H. Y. Fang, Van Nostrand
I
Reinhold Company, New York, 1975, pp. 121-147.

21. D'Appolonia, D. J., D'Appolonia, E., and Brisette, R. F., Discussion of "Set-
tlement of Spread Footings in Sands", Journal of the Soil Mechanics and Foun-
I
dations Division, ASCE, Vol. 96, No. SM2, Mar. 1970, pp. 754-761.

22. Schmertmann, J. H., "Static Cone to Compute Static Settlement Over Sand", Jour-
nal of the Soil Mechanics and Foundations Division, ASCE, Vol. 96, No. SM3, May
I
1970, pp. 1011-1043.

23. Mitchell, J. K. and Gardner, W. S., "In-Situ Measurement of Volume Change I

Characteristics", Proceedings, ASCE Specialty Conference on In-Situ Measure-

24.
ment of Soil Properties, Vol. 2, Raleigh, 1975, pp. 279-345.

Schultz, E. and Melzer, K. J., "The Determination of the Density and the Modu-
I
lus of Compressibility of Non-Cohesive Soils by Soundings", Proceedings, 6th
International Conference on Soil Mechanics and Foundation Engineering, Vol. 1,
Montreal, 1965, pp. 354-358. I
25. Mayne, P. W. and Frost, D. D., "Dilatometer Experience in Washington, D.C. and
Vicinity", Research Record 1169, Transportation Research Board, Washington,
1989, pp. 16-23. I
26. Webb, D. L., Mival, K. N., and Allinson, A. J., "A Comparison of the Methods
of Determining Settlements in Estuarine Sands from Dutch Cone Penetration
Tests", Proceedings, 2nd European Symposium on Penetration Testing, Vol. 2,
Amsterdam, 1982, pp. 945-950.
I
27. Jamiolkowski, M., Ghionna, V., Lancellotta, R., and Pasqualini, E., "New Cor-
relations of Penetration Tests for Design Practice", Proceedings, 1st Interna-
I
tional Symposium on Penetration Testing (ISOPT-1), Vol. 1, Orlando, 1988, pp.

28.
263-296.

Janbu, N., "Soil Models in Offshore Engineering", Geotechnigue, Vol. 35, No.
I
3, Sept. 1985, pp. 241-281.

29. Richart, F. E., Jr., Hall, J. R., Jr., and Woods, R. D., Vibrations of Soils
and Foundations, Prentice-Hall, Englewood Cliffs, 1970, 414 p.
I
30. Hardin, B. 0., "Nature of Stress-Strain Behavior for Soils", Proceedings, ASCE
Specialty Conference on Earthquake Engineering and Soil Dynamics, Vol. 1, Pasa-
dena, 1978, pp. 3-90.
I
31. Horvath, J. S., "Modulus of Subgrade Reaction: New Perspective", Journal of
Geotechnical Engineering, ASCE, Vol. 109, No. 12, Dec. 1983, pp. 1591-1596.
I
32. Vesic, A. S., "Beams on Elastic Subgrade and the Winkler Hypothesis", Proceed-
ings, 5th International Conference on Soil Mechanics and Foundation Engineer-
I
ing, Vol. 1, Paris, 1961, pp. 845-850.

I
I
5-26 I
I
I
I
I Section 6

I TIME-DEPENDENT DEFORMABILITY

I The parameters that define the time-dependent deformability of soils are important
for evaluating the settlement of foundations. In this section, basic definitions
are covered first to describe the pertinent terms. Correlations then are presented

I to evaluate the consolidation settlement of both cohesive and cohesionless soils.

Where available, in-situ test correlations are included. The parameters defining

I the time-dependency of consolidation settlement are covered next, again including

in-situ test correlations where available. The final topic addresses the parame-
ters that control the long-term settlement caused by secondary compression.

I BASIC DEFINITIONS

I The time-dependent deformability covered in this section refers to the processes of

hydrodynamic consolidation and secondary compression. Without addressing theoreti-

I cal issues in detail, the basic terms are described below.

The effective preconsolidation stress (ap or Ovmax) is the maximum vertical (over-

I burden) stress experienced by the soil during its geologic history, as shown in
Figure 6-1. Most natural soils are preconsolidated to some degree, either by ero-

I sion, desiccation, past glacial activity, aging, or other factors.

preconsolidation stress to current effective overburden stress is defined as the
The ratio of

overconsolidation ratio (OCR- ap;-av 0 ) and is a convenient term for describing the

I stress state. Methods for estimating ap and OCR have been presented in Section 3.

I The compression index (Cc) is defined as the slope of the void ratio (e) versus log
vertical effective stress (av) curve for virgin loading. This slope corresponds to
the normally consolidated (NC) state with OCR = 1. An alternative form is to plot

I the vertical strain (€v) versus log av.

is defined as the compression ratio [CR =
The virgin compression slope in this case
Cc/(1 + e 0 )].

I If the soil is unloaded vertically, it will rebound or swell along an unloading

line, as shown in Figure 6-1. Subsequent reloading follows a similar path, also

I shown in Figure 6-1. The differences between unloading and reloading normally are

I 6-1

l
 I
I
Q)

0
0
0:: cycle
I
"'0

~ Virgin
compression
I
Log crv ( veriical effective stress) I
Figure 6-1. Consolidation Behavior
I
small and are neglected in practice.
used.
Therefore, an average value (Cur) often is
Soils existing on the Cur line represent overconsolidated (OC) states.
I
The coefficient of consolidation (cv) expresses the rate of primary settlement with
time and is found by interpreting laboratory curves of settlement with time. From
I
these data, the value of Cv is computed as: II
(6-1)

in which T = time factor, H =height of drainage path, and t = time. The coeffi-
I
cient of consolidation includes the permeability (k) and constrained modulus (M =
1/rnv) as follows:
I
k Mhw (6-2) I
in which ""Yw unit weight of water.
I
The time factor (T) depends upon the drainage boundaries, geometry, and percent
dissipation of excess pore water stresses.
factor for 50 percent consolidation (Tso) is 0.197.
For one-dimensional loading, the time
For 90 percent consolidation,
I
the time factor (Tgo) is 0.848.
I
Secondary compression follows primary consolidation and is defined by the coeffi-
cient of secondary compression (Ca)· If expressed in terms of vertical strain, Ca€
is defined over one log cycle of time, as shown in Figure 6-2. If the coefficient
I
of secondary compression is expressed by change of void ratio with logarithm of
time, then:
I
6-2 I
I
I
I Log Time

caE =~
I c
0
.....
(/)
a log t

I
Figure 6-2. Time-Settlement Behavior

I
I (6-3)

in which e 0 initial void ratio.

I An alternate procedure for representing consolidation data is to use vertical

I strain versus vertical stress curves on arithmetic scales. In this way, the
stress-strain curve provides a constrained modulus (M) which can be related to the
more familiar compression index (Cc) as follows:

I (1 + e) ln 10 av 2.3 (1 + e) ov
M (6-4)
I Cc Cc

For overconsolidated soils, Cc should be replaced by Cur in Equation 6-4.

I COMPRESSION AND UNLOAD-RELOAD INDICES FOR COHESIVE SOILS

I The compression and unload-reload indices have been examined in detail by many
authors, and a variety of correlations have been proposed. Representative corre-
lations are presented below.
I
Typical Values

I The degree of compressibility of clay, expressed in terms of the compression index

(Cc), commonly is described as in Table 6-1. Over 70 different correlations have

I been published for correlating Cc to the index properties of clays, and Figure 6-3
illustrates the ranges involved. Apparently, the correlations between Cc and Wn
are more consistent than those cited between Cc and WL or e 0 .

I Although there is considerable scatter, the Terzaghi and Peck (£) relationship for

I NC natural clay is still popular. This relationship is given by:

I 6-3

I
 I
Table 6-1 I
DEGREE OF COMPRESSIBILITY
I
Compressibility
I
slight or low < 0.2
moderate or intermediate 0.2 to 0.4 I
high > 0.4
I
Numbers represent
references given
I
0.8 by Djoenoidi
(.)
u

-o
><
Q)

c
-
0.6
I
.......

.2
c
en
en
~
I
0..
E
u
0

I
120
Liquid Limit. wl (%)
0.4 0.8 1.2
Initial Void Ratio, e0 Natural Water Content. Wn (%)
I
Figure 6-3. Representative Cc Relationships for Cohesive Soils I
Source: Djoenaidi (!), p. 6-67.

I
Cc ~ 0.009 (wL - 10) (6-5)
I
Based on the modifed Cam clay model, Wroth and Wood (l) showed that Cc also can be
estimated as follows:
I
Cc z 0.5 Gs (PI/100) (6-6)
I
in which Gs- specific gravity of solids. Using a typical Gs = 2.7 for clays gives:

Cc "' PI/74 (6-7)

I
6-4 I
I
I
I Furthermore, the modified Cam clay model utilizes the A parameter, which is rela-
tively constant for natural clays at approximately 0.8. Re-writing Equation G-2

I
(in Appendix G) for A, the unload-reload index can be calculated as:

(6-8)

I Using the typical value of A= 0.8 with Gs- 2.7 gives:

I Cur "' PI/370 (6-9)

I Figure 6-4 indicates general agreement between the measured values of Cc and Cur
and those predicted using the modified Cam clay model. Furthermore, the regression
lines are within several percent of the model values. These data confirm that the
I average Cur is approximately 20 percent of the average Cc.

I The sensitivity of the clay (St) also affects Cc, particularly for marine depos-
its. Figure 6-5 shows the dramatic influence of sensitivity on Cc.

I Regressions: Cc=PI/73 (n=l63, r =0663,

2
5.0.=0.160)

I
2
.... Cur= PI/385 (n= 117, r =0448, S.D. =0.051)
:::> 1.4~----~----~-----r~---r-----r----~----------~.-----.-----,
u
X
Q)
...
-
"0 Mayne, 1980
••
I "0
c:

0
.Q
1.2 Nokose, et ol., 1988
Been, et ol., 1987
Wesley, 1988 .t. •
Q)
1.0 \1 Lambe a Whitmon,1969 •
• •
I 0::
"0
0
I
e
<J
NA
Olsen, et ol., 1986
Imoi, et ol., 1984 ~~ .t. ~ • .~ • •
..
.Q
c: ~ NA Morin a Dowe, 1987 ~
0.8 (Refs.~ -!!l .. :
I :::>
"0 ~ ..... •...
.7.. .
c: ........... IT
0
~,-...· ......
. ....
uo 0.6 ' , ... Modified com cloy
... fiJ.
I X
Q) ... ~~ 1.....: • .~~-• , 0

.
"0
c: 0.4 ~~/-,..
......
~It ~•oo
I c:
.2
rn
·~-.=·- ~ 8
rn
....
Q)
o.2 • /-.ev o• tl~ -~o--nz""ll1"
c9
0~~~-~~~--~1---LI--~---?~--~~l
Cl..

I E
u
0
0 10 20 30 40 50 60 70 80 90 100
Plasticity Index, PI (%)
I Figure 6-4. Compression and Unload-Reload Indices versus PI

I 6-5

I
 I
4 I
u
<U
X
0

~
·b:log 5-v
I
I
'"0
c
......
c
::>
VI

I
VI
<U
.....
c.
E I
0
u

0
I
0 2 3 4

In-Situ Void Ratio, e 0 I

Figure 6-5. Sensitivity-Compression Index Relationships

Source: Leroueil, et al. (12), p. 696. I

An alternative to the compression index (Cc) is the compression ratio (CR), defined
I
as Cc/(1 + e 0 ), in which e 0 =initial void ratio. Normalizing Cc in this manner
tends to reduce the data scatter,
in CR reported by Lambe and
(e.g.,~).

Whitman(~).
Figure 6-6 shows the typical ranges I
I
u
.,..-...

01 +a? I
II

u
0:: I
-0
0

0:: I
c
0
VI
VI
<U
'-
c.
I
E
0
u 20
Natural Water Content, wn (%)
I
Figure 6-6.

Source:
Compression Ratio versus Water Content

Lambe and Whitman(~), p. 321.

I
6-6 I
I
I
I Correlations with CPT qc Value

Attempts also have been made to corrrelate Cc with the cone tip resistance, as
I described by Sanglerat (13). However, these correlations have not proved to be
useful to date. For example, they show that for qc/Pa > 20, Cc is likely to be

I between 0.05 and 0.2. For qc/Pa < 10, Cc could be nearly any value above 0.1.

CONSTRAINED MODULUS FOR COHESIVE SOILS

I Typical Values

I As described previously, the constrained modulus (M) is an alternative to Cc.

Early work on this subject by Janbu (14) demonstrated that the drained secant con-
strained modulus (Mds) is a function of the vertical effective stress (ov) and a

I modulus number (m). For NC clays, Mds is given by:

I (6-10)

For NC silts and sands, M is given by:

I (6-11)

I Figure 6-7 shows the general trend in mas a function of porosity for a variety of
NC soils and rocks.

I Since the constrained modulus is defined as Bo/8€ for one-dimensional compression,

I it can be shown simply that:

1 + e0

I (
Cc
) ln 10 m~ (6-12)

I Therefore, the modulus number for clays is simply 2.3/CR, where CR

ratio.
= compression
Figure 6-8 shows that the trend for m with water content for NC clays is
consistent with the previous correlation for CR and water content (Figure 6-6).

I For OC clays, the modulus number is 5 to 10 times that for the NC range.

I Correlations with SPT N Value

The constrained modulus from oedometer tests on clay also has been correlated by
Stroud (16) with N values obtained from the standard penetration test (SPT). This
I relationship is given by:

I 6-7

I
 I
I
o Rock
A Moraine
o Sand
o Silt
=== I
E • Cloy
I
11/11.111.
....
C1)
.D
E
::>
z
I
10
I
1~-L~~~~--~~--~~~__J
0 50 100
I
Porosity, n (%)
I
I
I
I
I
I
Natural Water Content, wn (%)
I
Figure 6-8.

Source:
Modulus Number for NC Clay

Janbu (15), p. 261.

I
I
(6-13)
Mds/Pa - f N

in which the empirical coefficient, f, has been related to PI, as shown in Figure
I
6-8 I
I
I
I 6-9. This correlation is not very strong and should be used with caution.

I Correlations with CPT Results

Numerous correlations have been suggested to relate the cone penetration test (CPT)

I
qc value to the constrained modulus of cohesive soils. All generally take the form
below:

I
in which a empirical coefficient. Compilations of a (e.g., 17) have shown sug-
I
=

gested values ranging from 0.4 to 8, with the majority of values between 1 and 3.
However, most of these values have been obtained using a variety of mechanical and

I electric cones of different geometries and test procedures.

Figure 6-10 shows the variation of Mds with high quality cone tip resistance data
I from 12 sites tested by piezocone. This figure provides a more useful estimator
forM in clays.

I Correlations with DMT Results

I The dilatometer test (DMT) provides an estimate of Mds through an empirical rela-
tionship between the dilatometer parameters Eo and Ko, as shown in Figure 6-11.
The effect of the dilatometer parameter Io on this relationship is given in explic-

I it equations by Marchetti (19).

I
I 8 I I I I I. I

~·
1-
61-

I f
1-
4:... .-1!-.·~-·-
• I -

I 21-
1- 9 British cloys
0 I I I I I _L

I 0 10 20 30 40
Plasticity In de)(, PI (%)
50 60 70

I Figure 6-9. SPT Constrained Modulus Coefficient f versus PI

Source: Stroud (16), p. 373.

I 6-9

I
 I
I
0 Troll e Onsoy

I
0
0. 6 Glove A Porto Tolle
........ 0 Potomac • Montalto
z "' "il Drommen "f' Toronto
"0
600
[>Yorktown + Hogo
VI
:::J
:::J
"0
0
<I Modingley
0 Brent Cross
(sensitive)
I
z
"0
(1)
c
·c;
I
-.....
~ 200
u
0
I
20 30 40 50 60 I
. qT-a-vo
Net Cone Tip Res1stance, - - -

Figure 6-10.
Po
Constrained Modulus versus qT from CPTU for Clays
I
Source: Database from Mayne, et al. (18).
I
VI
:::J
:::J
"0 "0
0 0
VI
:::J
:::J
4 I
"il Sand (Chamber
I I I I I
:z z tests)
3>--- 6 Sand (In-situ)
"il
-
"0
OJ
c Q)
.....
OJ
• Cloy • ~
I
--
0
..... E 2-
·~ -
VI
c
0
u 0
E
0
·h "il

- I
~,
II

0
1-
6
• -;
I
I
I
zw 0
I 2
I
3
I
5
I
10 20
.P_::o:._-_u.::::.o
K o---

Figure 6-11.
CTvo

Constrained Modulus from DMT Parameters

I
Source: Marchetti (19), p. 317.
I
COMPRESSION INDEX FOR COHESIONLESS SOILS I
For the predominant quartz-type cohesionless soils found throughout the world, the
compressibility characteristics are much less than for cohesive soils.
to this observation could include micaceous sands and the calcareous sands
Exceptions
I
6-10 I
I
I
I associated with coralline deposits, which show significant compressibility compared
with the more prevalent silica sands. The compression index of cohesionless soils
av
! is somewhat stress-dependent, indicating that e-log
most appropriate means of presenting one-dimensional compression data.
plots are perhaps not the
Typical
values for the compression index and unload-reload index of six different sands are

I given in Table 6-2.

I The effect of grain size distribution on sand compressibility is illustrated in

Figure 6-12 at a reference relative density of 40 percent. The effect of relative
density on sand compressibility is given in Figure 6-13. In both of these figures,

I the notation used is defined in Table 2-7.

CONSTRAINED MODULUS FOR COHESIONLESS SOILS

I Typical Values

I The stress-dependency effect on sand compressibility may be taken into account more
directly by using the constrained modulus (Mds):

I Table 6-2

I COMPRESSIBILITY DATA FOR SIX SANDS IN CALIBRATION CHAMBER TESTS

I Sand eo 0 v1Pa 1 to 3
Cc
0 v/Pa 20 to 30 Cur

I Monterey 0 0.854
0.782
0.021
0.018
0.085
0.090
0.006
0.007

I Ticino 0.917
0.827
0.025
0.026
0.130
0.085
0.007
0.006

Hokksund 0.870 0.024 0.095 0.005

I 0.790 0.018 0.056 0.005

•I
Ottawa 0.760 0.025 0.030 0.007
0.560 0.005 0.100 0.003

Reid-Bedford 0.900 0.013 0.090 0.005

0.650 0.005 0.019 0.003

Hilton Mines 0.950 0.038 0.210 0.009

0.732 0.022 0.100 0.006

Note: Details on these sands are given in Appendix H.

Source: Been, et al. <.§)' p. 295.

I 6-ll

I....
 I
-..-

ii-v I Po
I
0.2 0.5 I 2 5 10

c
...>
- 0.01
~~~~~3;;:::=-r-ri-r~sZu:rm::is:te~r~
ldentificohons
I
0,_ emS, s•tG
(/)

0
emS
ciS
cmfS
fS +
I
-(.)

,_
fS, I cmf8
ts,t• cmfS
~ fS, s• cS
I
I
Figure 6-12. Effect of Grain Size on Sand Compressibility

Source: Burmister (20), p. 82.

0.2 0.5
c;v I Po
2 5 10
I
0.01
Or(%l
100
70
I
40

...> coarse, medium to fine SAND

I
c (cmfS)

I
0
,_
0

(/)
0.5 2 5
0

I
Or(%)
-,_
(.)

100
Q) 0.01
> 70

I
40

medium to fine SAND

(mfS)
0

Figure 6-13. Effect of Dr on Sand Compressibility

I
Source: Burmister (20), p. 83.
I
(6-15)
j
in which the modulus number (m) has been correlated with porosity, as shown pre-
viously in Figure 6-7. More specific general relationships are shown in Figure
6-14 for silts and sands.

6-12 I
I
I
500r---~----,---~----~

I E
..:

I
OJ
..0
E
::::>
z

I
• 40
Porosity, n (%)
50

Figure 6-14. Modulus Number for NC Silts and Sands

Source: Janbu (15), p. 261.

Correlations with SPT, CPT, and DMT Results

In Section 5, correlations are presented which relate the constrained modulus of

cohesionless soils to the SPT N value and CPT qc value. DMT correlations were
given in Figure 6-11.

COEFFICIENT OF CONSOLIDATION

I Typical Values

The field value of the coefficient of consolidation (cv) is a difficult parameter

I to estimate because common field situations include sand seams and lenses, varves,
etc., which make laboratory-predicted values of cv different from in-situ values.
However, Figure 6-15 provides a first-order estimate for cv of clays using the

I liquid limit.

a Correlations with CPT and DMT Results

Several of the recent in-situ tests, particularly the piezocone and dilatometer,
have been utilized to give field estimates of horizontal permeability (kh) and hor-
zontal coefficient of consolidation (cvh) in clays. The basic equation for the
horizontal coefficient of consolidation is:

I
I 6-13

I
 0(1)
(/)
I
~
-E
u
c:
>
J
'
.Q

'
0
-o
0(/)
c:

I
0
u
-c
0

•I
(1)

-
u

( 1)
0 40 80 120 160
u
Liquid Limit, wL (%)

Figure 6-15. cv versus WL

Source: NAVFAC (21), p. 7.1-144.

I
(6-16)
i
in which T = time factor, R =equivalent cavity (piezocone) radius, and t = time to
achieve desired degree of excess pore water stress dissipation. The approach is I
based on cavity expansion theory, and therefore it depends on the rigidity index of
the soil (Ir = Gfsu, in which G =shear modulus and su =undrained shear strength). I
Figure 6-16 gives the piezocone time factors. Most commonly, the dissipation test
is conducted for a period of time (t) which will allow 50 percent dissipation of
I
the original insertion excess pore water stress (~u). The time factor correspond-
ing to this dissipation time then is introduced into Equation 6-16 to compute the
coefficient of consolidation. Cylindrical theory would be used for a pore water
I
sensor behind the tip, while spherical theory would be used for a sensor at the tip.
I
I
Similar developments by Robertson, et al. (23) have led to an empirical method for
determining cvh from the dilatometer C readings, using the following:

(6-17)
I
in which Re = equivalent radius for the 14 mm by 95 mm dilatometer blade (i.e.,
Re z 20.6 mm) and the time factor (T) is given in Figure 6-17. In this figure, P2
is the dilatometer C reading at a particular time. The test procedure for the DMT

6-14 I
I
I
.I 0
:::>
<1
-
.......

-0
:::>
0.5

<1

10 100
Time Factor, T
Figure 6-16. Pore Water Stress Decay versus Piezocone Time Factor

I Source: Jamiolkowski, et al. (22), p. 128.

I ~
:::>
0
1.0 .------.-----.------.------, 1.0
p2( t }-uo

p2(0}-uo

I -0
:::>
<1
0.5

500
0.5

I 0
0.01 0.1
Time Factor, T
10

I Figure 6-17. Pore Water Stress Decay versus Dilatometer Time Factor

Source: Robertson, et al. (23), p. 574.

I dissipation readings is similar to that for the piezocone.

I COEFFICIENT OF SECONDARY COMPRESSION

I The coefficient of secondary compression (Ca) defines the rate of settlement with
time after primary consolidation is complete. This coefficient may be expressed
either in units of strain (Ca€) or void ratio (Cae) per log cycle of time, as shown

I in the following:

I 8€/8 log t (6-18)

8e/8 log t (6-19)

I For a wide variety of clays, Ca€ has been correlated to the natural water content,
as shown in Figure 6-18. Based on this figure, the following was suggested for NC
clay:

6-15

I
 I
.,"'
I
u
c
.,
·u;
...
0

Q)
I
0.

u
E
0

...>-
0
0.1
I
"0

I
c=
0
u
Q)
(/)

-c
0
0.01

Q)

;g
Ci>
I
0

I
u 0.001
10 100 1000

Natural Water Content, w0 (%)

Figure 6-18. Coefficient of Secondary Compression versus Water Content for NC Clays I
Source: Mesri (24), p. 125.

I
Examination of available data indicates that 0.0005 < Ca! < 0.001 for most OC clays.
(6-20)
1
For NC clays, the ratio of the coefficient of secondary compression to the compres-
I
sion index (Cae/Cc - Ca!/CR) is relatively constant for a given soil.
"\
lists Cae/Cc for a variety of clays. On the average, the value of CaefCc is 0.04 +
Table 6-3
I
0.01 for the inorganic clays and silts. For the organic clays and silts, the value
averages 0.05 + 0.01. For the peats, the value averages 0.075 + 0.01.
stant also is applicable for inorganic OC clays which have Cae/Cur equal to 0.04
This con-
±
I
0.01.
1
I
I

6-16

I
I
I
\
Table 6-3

COMPILATION OF Cae/Cc FOR NATURAL SOILS

I Grouping Soil Type

I
I Inorganic Clays Whangamarino clay 0.03 to 0.04
and Silts

'I Leda clay

Soft blue clay
Portland sensitive clay
0.025 to 0.06
0.026
0.025 to 0.055

I San Francisco bay mud

New Liskeard varved clay
0.04
0.03
to 0.06
to 0.06
Silty clay C 0.032
Nearshore clays and silts 0.055 to 0.075
Mexico City clay 0.03 to 0.035
Hudson River silt 0.03 to 0.06

Organic Clays Norfolk organic silt 0.05

and Silts
Calcareous organic silt 0.035 to 0.06
Post-glacial organic clay 0.05 to 0.07
Organic clays and silts 0.04 to 0.06
New Haven organic clay silt 0.04 to 0.075

Peats Amorphous and fibrous peat 0.035 to 0.083

Canadian muskeg 0.09 to 0.10
Peat 0.075 to 0.085

l Peat
Fibrous peat
0.05
0.06
to 0.08
to 0.085

I Source: Mesri and Godlewski (25), p. 421.

I REFERENCES

1. Djoenaidi, W. J., "A Compendium of Soil Properties and Correlations", M. Eng.

I 2.
Sc. Thesis, University of Sydney, 1985, 836 p.

Terzaghi, K. and Peck, R. B., Soil Mechanics in Engineering Practice, 2nd Ed.,
John Wiley and Sons, New York, 1967, 729 p.
I 3. Wroth, C. P. and Wood, D. M., "The Correlation of Index Properties with Some
Basic Engineering Properties of Soils", Canadian Geotechnical Journal, Vol.

I 15, No. 2, May 1978, pp. 137-145.

I 6-17

I
 I
4. Mayne, P. W., "Cam-Clay Predictions of Undrained Strength", Journal of the
Geotechnical Engineering Division, ASCE, Vol. 106, No. GTll, Nov. 1980, pp.
I
1219-1242.

5. Nakase, A. , Kamei, T. , and Kusakabe, 0. , "Constitutive Parameters Estimated by

Plasticity Index", Journal of Geotechnical Engineering, ASCE, Vol. 114, No. 7,
I
July 1988, pp. 844-858.

6. Been, K., Jeffries, M. G., Crooks, J. H. A., and Rothenburg, L., "The Cone
Penetration Test in Sands: Part II, General Inference of State", Geotechnigue,
I
7.
Vol. 37, No. 3, Sept. 1987, pp. 285-299.

Wesley, L. D., "Compression Index: A Misleading Parameter?", Journal of Geo-

technical Engineering, ASCE, Vol. 114, No. 6, June 1988, pp. 718-723.
I
8. Lambe, T. W. and Whitman, R. V., Soil Mechanics, John Wiley and Sons, New
York, 1969, 553 p.
I
I
,,
9. Olsen, H. W., Rice, T. L., Mayne, P. W., and Singh, R. D., "Piston Core Prop-
erties and Disturbance Effects", Journal of Geotechnical Engineering, ASCE,
Vol. 112, No. 2, June 1986, pp. 608-625.

10. Imai, G., Yano, K., and Aok, S., "Applicability of Hydraulic Consolidation
Test for Very Soft Clayey Soils", Soils and Foundations, Vol. 24, No. 2, June
1984, pp. 29-42.

11. Morin, P. and Dawe, C. R., "Geotechnical Properties of Two Deep-Sea Marine
Soils from The Labrador Sea", Canadian Geotechnical Journal, Vol. 24, No. 4,
il
Nov. 1987, pp. 536-548.

12. Leroueil, S., Tavenas, F., and LeBihan, J. P., "Proprietes Caracteristiques
des Argiles de l'esr du Canada", Canadian Geotechnical Journal, Vol. 20, No.
4, Nov. 1983, pp. 681-705.
'I
13.

14.
Sanglerat, G., The Penetrometer and Soil Exploration, Elsevier, Amsterdam,
1972, 464 p.

Janbu, N., "Soil Compressibility as Determined by Oedometer and Triaxial

Tests", Proceedings, 3rd European Conference on Soil Mechanics and Foundation
Engineering, Vol. 1, Wiesbaden, 1963, pp. 19-25.
'J
15. Janbu, N., "Soil Models in Offshore Engineering", Geotechnigue, Vol. 35, No.
3, Sept. 1985, pp. 241-281. I
16. Stroud, M. A., "The SPT in Insensitive Clays and Soft Rocks", Proceedings,
European Symposium on Penetration Testing, Vol. 2.2, Stockholm, 1974, pp. 367-
375.
I
17. Mitchell, J. K. and Gardner, W. S., "In-Situ Measurement of Volume Change
Characteristics", Proceedings, ASCE Specialty Conference on In-Situ Measure-
ment of Soil Properties, Vol. 2, Raleigh, 1975, pp. 279-345.
I
18. Mayne, P. W. , Kulhawy, F. H. , and Kay, J. N. , "Observations on the Development
of Pore Water Stresses During Cone Penetration in Clays", Canadian Geotechni-
cal Journal, Vol. 27, No. 3, Aug. 1990.
I
19. Marchetti, S., "In-Situ Tests by Flat Dilatometer", Journal of the Geotechni-
cal Engineering Division, ASCE, Vol. 106, No. GT3, Mar. 1980, pp. 299-321.
I
6-18 I
I
I
II 20. Burmister, D. M., "Physical, Stress-Strain, and Strength Responses of Granular
Soils", Symposium on Field Testing of Soils (STP 322), ASTM, Philadelphia,
1962, pp. 67-97.

21. NAVFAC, Soil Mechanics (DM 7.1), Naval Facilities Engineering Command, Alexan-

'
dria, 1982, 355 p.

22. Jamiolkowski, M., Ladd, C. C., Germaine, J. T., and Lancellotta, R., "New
Developments in Field and Laboratory Testing of Soils", Proceedings, 11th
International Conference on Soil Mechanics and Foundation Engineering, Vol. 1,
San Francisco, 1985, pp. 57-154.

23. Robertson, P. K., Campanella, R. G., Gillespie, D., and By, T., "Excess Pore
Pressures and the Flat Dilatometer Test", Proceedings, 1st International Sym-

•'\ 24.

25.
posium on Penetration Testing (ISOPT-1), Vol. l, Orlando, 1988, pp. 567-576.

Mesri, G., "Coefficient of Secondary Compression", Journal of the Soil Mechan-

ics and Foundations Division, ASCE, Vol. 99, No. SMl, Jan. 1973, pp. 123-137.

Mesri, G. and Godlewski, P.M., "Time and Stress- Compressibility Interrela-

tionship", Journal of the Geotechnical Engineering Division, ASCE, Vol. 103,

•I
No. GT5, May 1977, pp. 417-430 .

I
I
I
I
,,
'I
I
I 6-19

I
 I
I
Section 7

PERMEABILITY

The coefficient of permeability (k) of soil, also known as the hydraulic conductiv-
ity, describes the rate of water flow through soil. This soil property often is
difficult to evaluate with certainty, because it varies over many orders of magni-
tude and in-situ soil conditions are highly variable. In addition to controlling
the amount and rate of ground water inflow into foundation excavations, the coeffi-
cient of permeability also governs the rate of primary consolidation and equaliza-
tion of pore water stresses. I
TYPICAL VALUES

The value of the coefficient of permeability can vary over a wide range, as shown
I
in Table 7-1.
particle size.
From this table, it is clear that k is highly dependent upon soil
To obtain a first-order estimate of k in sands, Figure 7-1 suggests
I
I
Table 7-1

COEFFICIENT OF PERMEABILITY I
Coefficient of Relative
Soil Permeability, k (mjsec) Permeability

gravel > 10-3 high

sandy gravel,
clean sand,
fine sand medium

sand,
dirty sand,
silty sand lo-s to 10-7 low

silt, silty clay 10-7 to lo-9

< lo-9
very low
I
clay practically

Source: Based on Terzaghi and Peck (l).

impermeable
I
7-1 I
I
I
I 10 5 lli0- 2

I E
c:
5
2

I )( 10- 2

i
2
'--
-
..X:

--::>:
I .0
0
Q)

E
'-
Q)
2
(.)
Q)
(/)

'E
I 0...

-c
0
0.2
Lob test values
(Cu=2 to 12, D10 1D5<1.4) 5
Ill 10- 3

Q) 0.1

I -
(.)

Q)
0
u
orCu>l2

I
I 0. 0 I ..__._...._.__,__1.-J._._I..-W'-'-----'--'---'-.J......L...w..J...J 5 )( I 0- 5
0.1 0.2 0.5 I
Effective Particle Size, 0 10 (mm)
2 5 10

I Figure 7-1. Coefficient of Permeability versus Particle Size

Source: NAVFAC (;), p. 7.1-139.

I
I an approach in terms of void ratio (e) and effective particle size (expressed as
Dlo). The effect of particle size distribution and relative density on k is shown
for several sands in Figure 7-2. The notation used is given in Table 2-7.

I The in-situ vertical permeability (kv) of clay may be estimated from the void

I ratio, plasticity index (PI), and clay fraction (CF), as shown in Figure 7-3.
geotechnical problems, drainage can occur horizontally as well as vertically.
ratio of horizontal to vertical permeability (khfkv) generally is less than 1.5 for
In
The

t marine clays and other massive deposits. However, in varved clays and stratified
fluvial deposits, kh/kv easily can exceed 10, as shown in Figure 7-4. Values of

I
kh/kv over 100 are possible.

I
I 7-2

i
 I
I
,I
u
Q)
(/l
.......
E
cS

i
~
:0
0
Q)
I
E
,_
Q)
Q.. --~--
cfS, s•tG,
t+s
,J
-c
fS, acS
cS
0

Q)

-
u

Q) cfS, scfS
I
0 I 0- 8 .__.___,_ _._-':_...___,_~.___.____,_--:-:' scf G
u
0

Relative Density, Dr (%)

100
I
Figure 7-2. Coefficient of Permeability versus Particle Size and Relative Density

Source: Burmister (l), p. 78.

I
I
I
Q)

0
0
2.0 I
0:::
"0

~
1.2 I
I
kv (m/sec) I
Figure 7-3. Vertical Coefficient of Permeability for Clay

Source: Tavenas, et al. (~), p. 658.

7-3
 • Champlain cloys

I o Other Canadian cloys

* Other cloys
• Triaxial tests

I u
Q)
IF)

'-
E
>
..X

kh {m/sec)

I Figure 7-4. Permeability Anisotropy for Various Natural Clays

Source: Tavenas and Leroueil (~), p. 34 .

•I REFERENCES

l. Terzaghi, K. and Peck, R. B., Soil Mechanics in Engineering Practice, 2nd Ed.,
John Wiley and Sons, New York, 1967, 729 p.

I 2. NAVFAC, Soil Mechanics (DM 7.1), Naval Facilities Engineering Command, Alexan-
dria, 1982, 355 p.

I 3. Burmister, D. M., "Physical, Stress-Strain, and Strength Responses of Granular

Soils", Symposium on Field Testing of Soils (STP 322), ASTM, Philadelphia,
1962, pp. 67-97.

I 4. Tavenas, F., Jean, P., LeBlond, P., and Leroueil, S., "The Permeability of
Natural Soft Clays. Part II, Permeability Characteristics", Canadian Geotech-
nical Journal, Vol. 20, No. 4, Nov. 1983, pp. 645-660.

5. Tavenas, F. and Leroueil, S., "State-of-the-Art on Laboratory and In-Situ

't Stress-Strain-Time Behavior of Soft Clays", Proceedings, International Sympo-

sium on Geotechnical Engineering of Soft Soils, Mexico City, 1987, pp. l-46.

I
I'
•I 7-4
 I
I
I
Section 8

LIQUEFACTION RESISTANCE
j
For foundations in seismic regions, it is important to assess the potential for I
liquefaction occurring in cohesionless soils. If the cyclic stresses become too
large and last for a long enough period of time, looser sands below the water table
can liquefy and lose essentially all of their supporting capacity. Although lique-
I
faction analysis is complex, simplified guidelines have been developed for three
common in-situ tests, as described in this section.

CYCLIC STRESS RATIO

-I
•I
In all of the in-situ test evaluations, the loading is described by the average
cyclic stress ratio at depth, given by rav/avo• in which rav- average cyclic
stress and avo =effective overburden stress. This ratio can be evaluated experi-
mentally using cyclic triaxial compression or direct simple shear tests or by using
shaking table tests. Seed (l) discusses these tests and their interrelationships.

Alternatively, the average cyclic stress ratio can be estimated from the following
(Tokimatsu and Yoshimi, ~):
I
ravfav0 = 0.1 (M-1) Bmax (ay 0 ;-ay0 ) (1 - 0.015z) (8-1) I
in which M- earthquake magnitude (7.5 is used commonly), amax- maximum horizontal
acceleration at ground surface (as a fraction of g, the acceleration from gravity), I
uvo = total overburden stress, avo effective overburden stress, and z - depth in
meters (for z <25m).

CORRELATIONS WITH SPT, CPT, AND DMT RESULTS

Extensive work has been done on evaluating the liquefaction potential of loose
't
sands using the standard penetration test (SPT) N value and the cone penetration
test (CPT) qc value. A recent summary of this work by Seed and de Alba (l) is I
given in Figure 8-1. In this figure, the N value has been corrected for the over-
burden stress and a constant energy ratio of 60 percent, as described in Section j

-I
2. Data for this figure were developed from Pan-American, Japanese, and Chinese

8-1
I
I 0.6~--~---,----~--~--~
(a) Clean Sands
0.6~--,.---.,-----r---.---~
(b) Silty Sands
Fmes=35°/o 15°/o ~5%

I lb
'-
...0
0

>
0.5
Fines ~5%

I
I
I
I
0.5 I
1
r I
: I
I
I
I
I
I
1

0.4 I 0.4 L .tque f act10n

. I I I
1

I -0
0
0:: 0.3
Liquefaction
I
I
I
I

0.3
I I
I I
I I
1 1
1 I
I
I
I

(/')

_/,f'
(/')

,_
Q)

if)
0.2 '~
/ No Hquofoolloo

I
u
u
>.
u
No liquefaction

M =7.5 earthquakes
0.1 /
M = 7. 5 earthquakes

I 40 50 0 10 50

•I Figure 8-1. Liquefaction Resistance Correlated with Modified SPT N Value

Source: Seed and de Alba (i). pp. 290, 291.

sources. As can be seen, silty sands exhibit a greater resistance to liquefaction

I at a given N value.

I By cross-correlating SPT and CPT data, Seed and de Alba (i) developed a comparable
relationship for liquefaction resistance in terms of the CPT qc value.
tionship is shown in Figure 8-2 and uses qn, the cone tip resistance corrected for
This rela-

the overburden stress which, from Equations 2-18 and 2-19, is given by:

(8-2)

Recently, a more direct relationship has been proposed by Shibata and Teparaksa

I (~). This relationship was developed directly from CPT data obtained at earthquake
sites in Japan, China, and the U.S. As shown in Figure 8-3, this new approach pro-
vides further refinement over prior recommendations.

'
I
A correlation also has been developed by Robertson and Campanella
liquefaction resistance in terms of the dilatometer test (DMT) horizontal stress
index (Ko). as shown in Figure 8-4. This correlation is new and based on limited
(~) to evaluate

I data for normally consolidated, unaged, uncemented sands.

generalizations are likely in the future.
Further refinements and

I 8-2

I
 I
I
0.7

0.6 %fines ~35 :::15:::10

M =7.5 earthquakes

~
~5
I
0 o50!mm} 0.2 0.25 0.25 0.4 OB

I
0.1
•ti

,,
qc/N60 3.5 4.1 4.4 4.4 4.8 5.3
'._.o> 0.5
Liquefaction

-0
0
0:::
0.4

(/)
0.3

I
(/)
Q)
....
(/)
u 0.2
-u
>.
u
0.1 I
o~~~~--~~~~--~~~~~
0 40 80 120
Modified Cone Tip Resistance, qn/Pa
160 200 240
I
Figure 8-2. Liquefaction Resistance Correlated Indirectly with CPT Results

Source: Seed and de Alba (2), p. 295.

I
I
•t:
0

'.._o>
0.5

0.1 0.15 ~0.25mm

No liquefaction
•I
-0

I
0
0:::
(/) ..-..>---#--0 50 = 0.25 mm
(/)

-....
Q) ( :S 5% fines)
( f)

u
-u
>-
0.1
--Shibata and Teparaksa, 1988 ( ~)
--- Robertson and Campanella,
1985 (5)
I
u -·-Seed o-;;d DeAlba, 1986 (~)
0
0 200 I
Modified Cone Tip Resistance, qn/Pa

Figure 8-3. Liquefaction Resistance Correlated Directly with CPT Results I

I
Source: Shibata and Teparaksa (~), p. 57.

8-3 I
I
I
I 0
0.5 I
/~
•t:
....... /

I >
...0
0.4
/
/
/
/

~ 0.3 /

I
/
0 L iquefoction /
0::: /
/
r.n

/0
/
r.n 0.2
....
Q)

I (/)
u
u
>.
0.1 liquefaction

0 I

I 0
0 5
Horizontal Stress Index, K 0
10

Figure 8-4. Liquefaction Resistance Correlated with DMT KD

Source: Robertson and Campanella (~), p. 39.

I
,,
REFERENCES
I l. Seed, H. B., "Soil Liquefaction and Cyclic Mobility Evaluation for Level
Ground During Earthquakes", Journal of the Geotechnical Engineering Division,
ASCE, Vol. 105, No. GT2, Feb. 1979, pp. 201-256.

'I 2.

3.
Tokimatsu, K. and Yoshimi, Y., "Empirical Correlation of Soil Liquefaction
Based on SPT N Value and Fines Content", Soils and Foundations, Vol. 23, No.
4, Dec. 1983, pp. 56-74.

Seed, H. B. and de Alba, P., "Use of SPT and CPT Tests for Evaluating the
Liquefaction Resistance of Sands", Use of In-Situ Tests in Geotechnical Engi-

I 4.
neering (GSP 6), Ed. S. P. Clemence, ASCE, New York, 1986, pp. 281-302.

Shibata, T. and Teparaksa, W., "Evaluation of Liquefaction Potential of Soils

I
Using Cone Penetration Tests", Soils and Foundations, Vol. 28, No. 2, June
1988, pp. 49-60.

5. Robertson, P. K. and Campanella, R. G., "Liquefaction Potential of Sands Using

I 6.
the CPT", Journal of Geotechnical Engineering, ASCE, Vol. 111, No. 3, Mar.
1985, pp. 384-403.

Robertson, P. K. and Campanella, R. G., "Estimating Liquefaction Potential of

t Sands Using the Flat Plate Dilatometer", Geotechnical Testing Journal, ASTM,
Vol. 9, No. 1, Mar. 1986, pp. 38-40.

I
I
t 8-4

I
 I
I
I
-I
Appendix A

STANDARD PENETRATION TEST

The standard penetration test (SPT) is performed during a test boring to obtain an
approximate measure of the soil resistance to dynamic penetration and a disturbed
sample of the soil. Although the test can be performed in a wide variety of soils,
the most consistent results are found in sandy soils where large gravel particles
are absent. Almost all U.S. soil drilling rigs are equipped to perform the SPT.
In fact, the SPT is the most common in-situ geotechnical test in the world (l).

PROCEDURE

The detailed procedure for the SPT is described in ASTM Dl586 (l), and a complete
I'r

theoretical analysis of the statics and dynamics of the SPT is given by Schmertmann
<.~. ~). I

'
To perform the test, the drilling crew, after advancing the test boring to the
desired depth, first removes the string of drill rods slowly and cleans out the
hole to the desired depth of testing. During this procedure, the head of water in
the hole is maintained at or above the ground water level to avoid an inflow of I
water into the hole that can disturb the soil and cause erroneously low (conserva-
tive) test results. After the drilling tools are removed, a standard 51 mm (2 in)
O.D. split spoon sampler, as shown in Figure A-1, is attached to the drill rods and
I
lowered carefully to the bottom of the hole. With the sampler resting at the bot-
tom of the hole, a 63.6 kg (140 lb) weight is allowed to fall freely 762 mm (30 in) I
0.1 in.
(2.5m ml
L ,.,. ,, 1
f 0 pens hoe f Tube I
'
\
16° to 23°
-
..z_- r-.1.375 in. ---
(34.9mml
•I
1.5in. _ _ _ '7
(38.1mml ~

I to 2 in.
(25 to 50mml
18 to 30 in.
(457 to 762 mml I
Source:
Figure A-1. Standard Split-Spoon Sampler

American Society for Testing and Materials (l), p. 223.

I
A-1

'l
I
I onto a collar that is attached to the top of the drill string until 460 mrn (18 in)
of penetration has been achieved (or 100 blows have been applied).

The two most common hammers in North American practice are the safety and donut
hammers. The safety hammer illustrated in Figure A-2 is a long weight which slides
over the drill rods and impacts against an internal anvil. The donut hammer illus-
trated in Figure A-3 is a short, wide weight centered on a guide pipe which strikes
an external anvil above the drill rods. Alternatively, but now uncommon in U.S.
practice, a 63.6 kg (140 lb) pin-guided weight is allowed to drop freely on the top
of the drill string. The overall equipment and setup for the SPT are shown in

I Figure A-4.

I The number of blows (or drops of the weight) is recorded for each of three 152 mm
(6 in) intervals; the first generally is considered a seating drive, and the number
of blows for the final 305 mm (12 in) is reported as the standard penetration
resistance or N value. After the sampler has been brought back to the surface, the

'I 47.75in.

I
(1.213m)

43.125in.

1l (1.095m) Sleeve

I
I l Guide cop

I
I Sin.
(127mm)

I Figure A-2.

Source:
SPT Safety Hammer

Kovacs, et al. (2), p. 11.

I A-2

I
 I
3.375in.
(85.7mml
I
I· "'I

Pull cop

I Eye bolts for

cot head rope
I
I
I

48 {to 42)in.
{1.219 to 1.067ml
8.75 {to 10. 75) in.
{222 to 273 mml
I 10 {to 6.75lin.
1254 to 171 mml
I
I I
I I
12.75 in.l
,.. ..,
l69.9mml t--+-Oonut hammer
I
I
I
Guide slip pipe
or drive pipe
I {to 1.25lin.
5in.{ 127mml
diameter drive
ring {optionoll
I
L---.J
I

]
{25.4 to 31.8mml

4in.U02mml
I
5in. {127mml

'I
diameter drive_/L---.----r--'
head or anvil
~g~----Coupling to drill rod

Figure A-3. SPT Donut Hammer

Source: Kovacs, et al. (2), p. 10. I

samples are removed and classified, before being placed into jars, labeled, and
sealed with wax for transport.
i
ADVANTAGES AND DISADVANTAGES
I
The advantages of the SPT are that it is relatively quick and simple to perform,
and it is widely available. It is relatively inexpensive and provides, wi~h one
t
procedure, both a sample and a soil test result. The test also provides a useful
index of the relative strength and compressibility of the soil in the immediate
vicinity of the test. In addition, the test is able to penetrate relatively diffi-
I
cult materials such as dense layers, gravels, and fills.
I
A-3
I
I
 I
I
~ Crown sheave( s)
or pulley(s)

I lin.(25mm)
manila rope

I
shown

Slip or
pipe

I Drill rod
Anvil

I ~
l''t

t' ::--Borehole
t~,.

I ''r'
lit ttl
wbl! _l
~"
·--,
1

18in. (457mm)

I Figure A-4. Equipment Used to Perform the SPT

Source: Kovacs, et al. (£), p. 3.

I
t The disadvantage of the SPT is that it has many sources of error, both random and
systematic (l- 10). The accuracy of the test is in large part dependent on the
details of the procedure followed and the equipment used by the drilling crew, so

I that the care and knowledge of the drillers forms a critical factor in the test
accuracy.

I The SPT should not be relied on in soils containing coarse gravel, cobbles, or
boulders, because the sampler can become obstructed, giving erroneously high and

'•
unconservative N values. The test also should not be relied on for cohesionless
silts, because dynamic effects at the sampler tip can lead to erroneous strength
and compressibility determinations. In addition, the test has little meaning in
soft and sensitive clays. In such soils, the SPT yields results inconsistent with
actual in-situ conditions.

I If the head of water in the hole is not maintained at or above the ground water

I'
A-4
 I
level, piping can occur at the bottom of the hole which can loosen the soil and
I
invalidate the test results. This problem can be minimized by returning water to
the hole as the drilling tools are removed prior to conducting the SPT. I
Studies by Kovacs (11) showed that the SPT is highly dependent on the method of
winding the hammer rope around the cathead on the drill rig. While seemingly a I
minor detail, these studies showed that when two turns of rope are used, as is com-
mon practice in the U.S., N values are about 40 percent higher than when a free-
fall trip monkey or one turn was used. This example illustrates the level of
I
uncertainty involved.
I
In addition, many older correlations of N values with engineering properties were
based on pin-guided weights, which are no longer used for the SPT. The rod-guided
hammers in present use can lead to slightly higher (unconservative) N values.
I
SOURCES OF ERROR, RELIABILITY, AND COST I
The SPT has numerous sources of error that limit its use in foundation design.
list of many of the important sources of error and their probable effects on the
A
I
SPT results is given in Table A-1. Factors that tend to increase the N values err
on the unconservative side by overestimating soil strength and/or stiffness.
ever, most correlations of the SPT with engineering properties tend to be somewhat
How-
I
conservative. Other important issues influencing the N value are discussed in
detail by Schmertmann (10). I
In addition to these sources of error, a number of soil mechanics factors affect
the test results and the correlations of N value with engineering properties.
t
These factors include particle size, shape, and mineralogy; soil sensitivity, per-
meability, and degree of saturation; time lapse between drilling and testing; spac-
ing of samples; depth of sampler penetration; relative depth of the boring; and
I
size of the vent area of the sampler.
I
The reliability of the SPT is best where it is used as an index test to determine
the approximate strength and compressibility of sandy soil strata for preliminary f
•I
design purposes. For example, a soil with anN value of 50 is unlikely to exhibit
any major problems with respect to strength or compressibility for spread footings;
on the other hand, a soil with an N value of 2 or 3 can be expected to pose signi-
ficant difficulties.

Although it is difficult to quantify the costs of SPT in remote areas, one approach

A-5
'I
I
I
Table A-1

I MAJOR SOURCES OF ERROR IN THE STANDARD PENETRATION TEST

I Influence on
Cause Effect N Val~e
I Inadequate cleaning of hole SPT is not made in original in-situ Increases

I soil, and therefore soil may become

trapped in sampler and be compressed
as sampler is driven, reducing
recovery

I Failure to maintain adequate

head of water in the borehole
Bottom of borehole may become quick Decreases

I Careless measurement of ham-

mer drop
Hammer energy varies (generally,
variations cluster on the low side)
Increases

Hammer weight inaccurate Hammer energy varies (driller sup- Increases or

I plies weight; variations of 5 to 7
percent are common)
decreases

I Hammer strikes drill rod col-

lar eccentrically
Hammer energy reduced Increases

Increases
Lack of hammer free fall Hammer energy reduced

I because of ungreased sheaves,

new stiff rope on weight,
more than two turns on cat-
head, incomplete release of

I rope during each drop

Sampler driven above bottom Sampler driven in disturbed, arti- Increases

I of casing ficially densified soil greatly

Careless blow count Inaccurate results Increases or

decreases

I Use of non-standard sampler Correlations with standard sampler

invalid
Increases or
decreases

I Coarse gravel or cobbles in

soil
Sampler becomes clogged or impeded Increases

•
Use of bent drill rods Inhibited transfer of energy of sam- Increases
pler

Source: Kulhawy, et al. (12), p. 5-24.

I A-6

I
 is to determine the daily drill rig charge and divide by the number of tests
obtainable in one day. All-terrain vehicles in 1990 cost about $1000 to $1500 per
I
day and, during a typical day, 10 to 20 tests might be obtained. Therefore, the
unit charge could be approximated as $50 to $150 per test, including drilling 1.0
I
to 1.5 m (3 to 5 ft) between tests. These figures are intended only as a relative
measure of the cost of performing the SPT for comparison with other field explora-
tion techniques.
I
REFERENCES
I
1. DeCourt, L., Muromachi, T., Nixon, I. K., Schmertmann, J. H., Thorburn, S.,
and Zolkov, E., "Standard Penetration Test: International Reference Test Pro-
cedures", Proceedings, 1st International Symposium on Penetration Testing
(ISOPT-1), Vol. 1, Orlando, 1988, pp. 3-26.
I
2. American Society for Testing and Materials, "Standard Method for Penetration
Test and Split-Barrel Sampling of Soils (Dl586-84)", Annual Book of Standards,
I
Vol. 4.08, ASTM, Philadelphia, 1989, pp. 221-225.

3. Schmertmann, J. H., "Statics of SPT", Journal of the Geotechnical Engineering

Division, ASCE, Vol. 105, No. GTS, May 1979, pp. 665-670.
I
4. Schmertmann, J. H. and Palacios, A., "Energy Dynamics of SPT", Journal of the
Geotechnical Engineering Division, ASCE, Vol. 105, No. GT8, Aug. 1979, pp.
909-926.
I
5. Kovacs, W. D., Salomone, L. A., and Yokel, F. Y., "Comparison of Energy Meas-
urements in the Standard Penetration Test Using the Cathead and Rope Method",
I
Report NUREG/CR-3545, Nuclear Regulatory Commission, Washington, 1983, 69 p.

6. Kovacs, W. D., Salomone, L. A., and Yokel, F. Y., "Energy Measurements in the
Standard Penetration Test", Building Science Series 135, National Bureau of
I
Standards, Washington, 1981, 73 p.

7. Fletcher, G. F. A., "Standard Penetration Test: Its Uses and Abuses", Journal
of the Soil Mechanics and Foundations Division, ASCE, Vol. 91, No. SM4, July
I
8.
1965, pp. 67-75.

Ireland, H. 0., Moretto, 0., and Vargas, M., "The Dynamic Penetration Test: A
Standard That Is Not Standardized", Geotechnigue, Vol. 20, No. 2, June 1970,
I
9.
pp. 185-192.

Orchant, C. J., Kulhawy, F. H., and Trautmann, C. H., "Reliability-Based Foun-

I
dation Design for Transmission Line Structures: Critical Evaluation of In-Situ
Test Methods", Report EL-5507, Vol. 2, Electric Power Research Institute, Palo
Alto, 1988, 214 p. I
•
10. Schmertmann, J. H., "Use the SPT to Measure Soil Properties? -Yes, But .. !",
Dynamic Geotechnical Testing (STP 654), ASTM, Philadelphia, 1978, pp. 341-355.

11. Kovacs, W. D., "What Constitutes a Turn?", Geotechnical Testing Journal, ASTM,
Vol. 3, No. 3, Sept. 1980, pp. 127-130.

A-7 I
I
I
I 12. Kulhawy, F. H., Trantmann, C. H., Beech, J. F., O'Rourke, T. D., McGuire, W.,
Wood, W. A., and Capano, C., "Transmission Line Structure Foundations for
Uplift-Compression Loading", Report EL-2870, Electric Power Research Insti-

I tute, Palo Alto, 1983, 412 p.

I
I
I
I
I
I
I
I
I
I
I
I
t
I
I A-8

I
 I
I
Appendix B

CONE PENETRATION TEST I

The cone penetration test (CPT), once known as the Dutch cone test, is a versatile I
sounding procedure that can be used to classify the materials in a soil profile and
to estimate their engineering properties. The CPT is becoming perhaps the most
popular and versatile in-situ test in the world (l). In the CPT, a conical pene-
I
trometer tip is pushed slowly into the ground and monitored. The earlier versions
of the CPT still are used widely and are known as mechanical friction cone (Bege-
mann) penetrometers (Figure B-1). Some of these penetrometers lack a friction
I
sleeve and measure only tip resistance, such as the Delft mantle cone (Figure
B-1). These devices provide less information about the soil conditions. Modern
I
I
devices, such as those shown in Figure B-2, contain electrical transducers to meas-
ure both tip and side resistances as the instrument is advanced; such devices are
known as electric friction cone penetrometers. In the U.S., the electric cone in

35.7mm
I
n
35.7mm
I
M
a) Delft Mantle
Cone
52.5mm
b) Begemann
Friction Cone
47mm 30mm
I
92mm
23mm
133.5mm

35.7mm
I
99mm
69mm I
35mm lllmm 32.5mm I
35mm
I
Figure B-1. Mechanical Cone Penetrometers
I
Source: American Society for Testing and Materials (l), pp. 415, 416.

B-l I
I
I
I seal

I
I
Wot" ~I""'
I : :_
CSlood oell ~ -==d oell
c:~
I Figure B-2.
+

Typical Designs of Electric Cone Penetrometers

Source: Schaap and Zuidberg (2), p. 842.

I most common use is the Fugro cylindrical cone. Unless otherwise noted, this cone
is assumed. More recently, piezocone penetrometers (CPTU) have been developed

I which measure the pore water stresses during penetration, as well as the cone tip
resistance and sleeve side resistance. Furthermore, several new cone devices have

I been introduced to measure additional parameters, including the seismic cone (for P
and S waves), acoustic cone, pressuremeter cone (with full-displacement PMT),
vibrating cone (for liquefaction assessment), lateral stress cone (for pile analy

I sis and K0 evaluation), logging cone (for nuclear density readings), and cone pene-
trometers for environmental work, including water sampling capabilities.

I PROCEDURE

The detailed procedure for the CPT is described in ASTM D3441 (~). To perform the

I test, an electric cone penetrometer tip is attached to a string of steel rods and
is pushed vertically into the ground at a constant rate of approximately 20 mm (0.8

I in)jsec. Wires from the transducers are threaded through the center of the rods,
and the tip and side resistances are recorded continuously on a strip chart record-
er (Figure B-3) until the desired depth is reached. A similar procedure is used

I for electric piezocone soundings, except that special measures are required for
ensuring saturation of the porous stone element.

i The procedure is modified slightly when a mechanical penetrometer tip is used. In

this case, the penetrometer tip is connected to an inner set of rods and is first
I advanced about 40 mm (1.6 in), giving the tip resistance. With further thrusting,
the tip engages the side friction sleeve and, as the inner rods advance, the rod

I B-2

I
 I
Lb I
-~V~AV~'
·~ ··~v. Analog
A/0
-
signal
conditioner converter
I
I I
~~
Strip chart
Computer
Processor
I
recorder
- I
t'"'"'J Magnetic
- ....._
Printer
taoe
J
Figure B-3. Electric CPT Data Acquisition System

Source: DeRuiter(~), p. 393. I

force equals the sum of the tip and side resistances. The tip resistance is sub-
I
tracted to give the side resistance. Finally, the outer rods are pushed to col-
lapse the entire device, and the process is repeated at approximate 200 rnrn (8 in) I
intervals. This mechanical process has several important sources of error not
characteristic of the electrical process and, where available, the electric pene-
trometer is recommended.
I
With standard mechanical and electric cones, the two most useful parameters meas- I
ured by the test are the tip resistance, qc, and the side resistance, fs· Piezo-
cones also provide readings of the maximum pore water stress, Urn.
I
The CPT can be used where a sample is not needed and soil conditions do not prevent
its penetration. In general, the CPT is less suitable in soils containing gravelly I
soils, cobbles or boulders, or cemented seams.

Cone penetrometers have been in general use since the 1930s in Europe, but only
I
within the past two decades have they gained wide usage in the U.S.
trometers can be employed in a variety of soils and, although they do not provide a
Cone pene-
I
sample, they have a number of advantages over the standard penetration test. The
CPT, especially when performed with an electrical tip, provides a continuous log of
soil conditions, while the SPT usually shows conditions only at discrete locations
I
in the soil profile, typically at 1 to 1.5 rn (3 to 5 ft) intervals.
CPT measures at least two parameters, it ideally gives more information about
Because the
I
B-3 I
I
 I
I in-situ soil consistency than the SPT. Furthermore, when comparisons of cone
soundings using both electric and mechanical cone tips are made, the profiles give

I similar trends, as shown in Figure B-4. However, the electric cone provides more
tip detail and shows less scatter in the side resistance profile, indicating that
soil boundaries can be located more accurately with an electric penetrometer tip.

I The mechanical and electric cones do not give the same results, largely because of

I the different geometry of the cones. The Delft and Begemann cones shown in Figure
B-1, as well as the Gouda cone (similar to the Delft), all have a reduction india-
meter beyond the cone tip. In contrast, the Fugro electric cone has the same
sleeve and tip diameter. Approximate correlations between these mechanical and
electric cones have been suggested (e.g.,£- 10). These studies generally have
shown that qc for electric cones is greater than qc for mechanical cones in sands,
while the reverse is true in clays and silts. To quantify these studies further,
data were summarized from 14 sands and 10 clays and silts tested by both Fugro

I electric cones and several mechanical cones.

and indicate a good correlation.
The results are shown in Figure B-5

I For side resistance (fs), the mechanical cones apparently give higher readings than
the electric cones in all soils. In sands, the ratio is about 2 (e.g.,~. l). In

I marine clays, the ratio varies from 2.5 to 3.5 (e.g., 18).

I Side Resistance, Tip Side Resistance, Tip

fs/Po
fs/Po
----+-----
I 2.5

I -~.:;."
~=--
----
-=--;...--
5

-~~-
-~

fa...,.
..,.
~

_, -~
~

.. -
7..:--15
-,__ _
-<-
--:.==-...--
E

...,..--
..c ..c
a.cu a.cu
20 0 0

I
Figure B-4. Comparison of Begemann Mechanical and Fugro Electric Cones

Source: DeRiuter (~). p. 466.

I B-4

I
 I
I
I
-
0..
0

.......
Q)
c:
I
0

-
u
u

u
I
Q)

~
0"
0

Mechanical
Sand Cloy/Silt
I
Cone Type
Delft
Begemonn
0
6
•
.&
I
Gouda 0
•
1~---L--~~~~~L----L--~~~~UW-----L--~~~~_u
I 2 5 10 20 50 100 200 500 1000
I
qc (mechanical cone) /p 0

Sand Cloy/Silt
I
0
0
Heijnen (~)
Joustro (!.! )
•
•
De Beer ( ~)
Dobie (.!_§) I
0 Kok (12)
• Schmertmonn (§_)
0
0
6
Schmertmonn ( §)
Amor C.2>
De Ruiter (_§)
•
.&
.&
Amor (_~)
Boligh, et of. ( 10)
Mayne, et ol. (!..?)
I
6
6
6
Jones
Rol
Smits
a
(§)
<z>
Rust ( 14) .& Mayne a Kemper (!§)
I
Figure B-5. Correlation of qc Between Electric and Mechanical Cones
I
CPT results also may vary as a function of electric cone type. A recent study by I
Lunne, et al. (19) compared the results of 14 different types of commercially-
available electric cones in the same sand. The variations in qc were relatively
small, but values of fs varied dramatically, in some cases by a factor of 3. These
I
results undoubtedly would influence all interpretations made from the test results,
so it is prudent to conduct verification and local calibration tests with specific
CPT equipment.
I
The introduction of the piezocone (CPTU) and the resulting comparative studies of
I
B-5
I
I
I
I the CPT and CPTU have shown that all cones require a correction for pore water
stresses acting on any unequal areas of the cone. Correction of the tip resistance

I is most important in soft clays where the values of qc and urn are of comparable
magnitude. Studies by Lunne, et al. (19) using 14 different cones at the Onsoy
site in Norway showed a wide range in the uncorrected cone tip resistance (qc), but

I a relatively narrow range in the corrected cone tip resistance (qT).

fs also must be corrected, but this correction requires additional pore water
The value of

I stress measurements behind the cone sleeve.

yet practical for commercial CPTU testing.
These additional measurements are not

One further complication with the piezocone is that its design has not yet been
standardized (e.g., 20). Most commercially-available piezocones place the porous

I element either on the cone tip face or just behind the cone tip, as shown in Figure
B-6. Technically, the measurement of pore water stresses behind the tip (ubt) is
required to correct the cone tip resistance (qc) for pore water stresses acting on

I unequal areas of the cone. On the other hand, pore water stress measurements on
the cone tip or face provide the maximum reading, which may be best for delineation

I of stratigraphy.

Many electric cone penetrometers in commercial use do not have the ability to meas-

I ure pore water stresses during penetration. Therefore, it is of interest to exam-

ine means of empirically correcting the measured cone tip resistance (qc) to obtain

I the corrected cone tip resistance (qT), as follows:

transducer

I
I Strain gouges stone
for tip load cell

I
I
I Strain gouges
for tip load cell

Figure B-6. Common Piezocone Geometries

I Source: Campanella and Robertson (21), p. 7.

I B-6

I
 I
I
in which a = net area ratio defined in Figure 2-ll.
cal values of "a" for commercial cones.
Lunne, et al.
The actual value of "a" should be deter-
(19) give typi-
I
mined by site calibration.

Piezocone data from numerous soil sites are summarized in Figure B-7 to illustrate
I
the variation in ubt as a function of soil type and structure.
analyses, Ubt = 0.53 qT for intact clays and Ubt = 0.58 qT for the highly sensitive
From regression
I
Leda clays. Silts and micro-fissured clays show values of Ubt that are only a
small fraction of qT. For fissured clays, Ubt is about zero. These trends will be
useful for estimating the corrected cone tip resistance on a preliminary basis.

ADVANTAGES AND DISADVANTAGES

I
The CPT has a number of advantages over other routine forms of in-situ testing.
Current trends indicate that usage will continue to increase, as more engineers I
become familiar with the types of information that it provides, and as more dril-
ling firms acquire the equipment to perform the test.
I
a.
0
30
Intact: I
Ubt=0.534 QT
' :0
::l

c£
i=
Ledo
ub1=0.579qT
n=l66
r 2 =0.897
5.0.=0.148 / /
/
/--) / I
"0 20 behind \ & _, /
n=70 /"&/
I
c: Q
:c
Q)
tip, ubi r2=0.912 ¢ 121 &/
ro S.D. =0.091 t / t·{''
1/) ~/_,.-"'
1/)
<>?.'fl. ,§> .dl
....
I
Q)

(j) m /~O lb.

.... m~
Q) 0 , lb.
0
" (J~ It :O:.j'w '"-rmfl~ I
3
....
Q)
•• :..
0

(:;:·~·;::~..........IIIDliJ
a..
"0
....
Q)

::l
1/)
0
Q)
0
Silts
I
·..;7:..... 8 1SJ0al~~~0~ ISl
=~=...~.= ISl
5?
............
Fissured
I
~

0 10 20
Corrected Cone Tip Resistance, qT /p0
I
Figure B-7. Measured Pore Water Stresses in CPTU Tests

Source: Mayne, et al. (22).

I
B-7 I
I
I
I The CPT is one of the faster and less expensive forms of in-situ testing in rela-
tively soft or loose soils. It provides a rapid method of identifying potential

I
problem soils, such as peat or soft clay strata, so that more sophisticated sam-
pling and testing procedures can be used as efficiently as possible. Typical pene-
tration rates during testing average about 1.2 m (4 ft) per minute and, except for

I problems caused by cemented layers or gravel, penetration is interrupted only to

add additional rods as the test advances. Data are recorded concurrently with the
test and, when the instrument is calibrated, the test personnel have a relatively

I minor influence on the results, compared to the SPT. The test can be performed in
a wide range of soils, although very hard soils or gravel can not be penetrated at

I the present time. Except for special, high-capacity cone trucks, most standard
cone equipment can penetrate soils with SPT N values up to 50 or thereabouts.

I A significant advantage of the electric cone penetration test is that it provides a

continuous record of soil conditions. Stratigraphy and soil identification are

I inferred from empirical classification charts developed for the mechanical and
electric cones. The new piezocone equipment offers the most accurate means of pro-
filing soil strata today (20). Subsurface conditions therefore may be inferred
I without retrieval of soil samples. In general, however, samples should be obtained
whenever feasible to confirm the interpretation of soil types made with the CPT.

I As with the SPT, the empirical correlations vary with soil type.

The CPT also has several disadvantages. First, no sample is obtained and the pene-
I trometer can not penetrate very dense soils or soils containing cobbles or boul-
ders. Excessive force in these materials can damage the penetrometer tip. These

I problems, however, also are faced by most other forms of in-situ testing.
many drilling contractors do not have the test equipment at the present time.
Second,

Third, the penetrometer may drift from vertical at depths below about 50 ft (15
I m). Many new electric penetrometer tips include an inclinometer to monitor verti-
cality, so that if the instrument does wander, the operator can determine imme-

I diately if the test should be repeated.

I
SOURCES OF ERROR, RELIABILITY, AND COST

Errors in the CPT have been described by several authors (~, 23, 24), and Table B-1
lists many of the sources of error in the standard mechanical and electric CPT.

I For the more sophisticated cone penetrometers such as the piezocone, specialized
personnel, electronics, and computer hardware are required, and therefore numerous

I other factors may affect the measurements.

I B-8

I
 I
Table B-1 I
MAJOR SOURCES OF ERROR IN THE CONE PENETRATION TEST
I
Cause Effect Influence on Results
I
Gravel or cobbles in soil Impedes penetration of pene-
trometer tip (can break tip or
rods)
Increases qc greatly
I
Causes penetrometer to wander
off vertical
Increases or
decreases qc and fs I
Worn penetrometer tip Tip may become dull and/or Increases or
surface roughness may become
greater or lesser than stan-
dard
decreases qc and fs
slightly I
Soil clogging end of
friction sleeve (mechan-
ical tips only)
Adds an erroneous end bearing
component to fs
Increases fs up to
about 80 percent I
Rusted or clogged inner
rods (mechanical tips
Impedes free travel of inner
rods because of friction
Increases qc and fs I
only) against outer rods

Hard soils (mechanical

tips only)
Causes elastic compression of
inner rods, giving false indi-
Measurement of qc
and f s may not be
I
cation that penetration has possible

Leaky water seal (elec-

occurred

Electrical transducers may Increases or

I
trical tips only) become corroded decreases qc and fs

Improper calibration Inaccurate measurements Increases or

decreases qc and fs
I
(electrical tips only)

Source: Kulhawy, et al. (23), p. 5-30. I

The reliability is stated effectively in ASTM D344l (V, "Because of the many
I
variables involved and the lack of a superior standard, engineers have no direct
data to determine the accuracy of this method. Judging from its observed repro- I
ducibility in approximately uniform soil deposits, plus the qc and fs measurement
effects of special equipment and operator care, persons familiar with this method
estimate its precision as follows: (l) mechanical tips - standard deviation of 10
I
percent in qc and 20 percent in fs, (2) electric tips - standard deviation of 5
percent in qc and 10 percent in fs". I
B-9
I
I
I
I The CPT may be conducted using either standard drilling rigs or special cone trucks
which maximize the effort by pushing through the system center of gravity. The

I former generally weigh about 10 tons (89 kN) and can achieve qcfpa up to 250, while
the special 20 ton (178 kN) cone trucks can reach qcfpa values of 600 or more.
Mobilization costs for the latter are higher. However, unit costs for both run

I about $4 to $6/ft ($13 to $20/m) for CPT profiling.

sion costs about $5 to $9/ft ($16 to $30/m).
The more specialized CPTU ver-

I REFERENCES

l. De Beer, E. E., Goelen, E., Heynen, W. J., and Joustra, K., "Cone Penetration
I Test: International Reference Test Procedures", Proceedings, lst International
Symposium on Penetration Testing (ISOPT-1), Vol. l, Orlando, 1988, pp. 27-52.

I 2. American Society for Testing and Materials, "Standard Test Method for Deep,
Quasi-Static, Cone and Friction-Cone Penetration Tests of Soil (D344l-86)",
Annual Book of Standards, Vol. 4.08, ASTM, Philadelphia, 1989, pp. 414-419.

I 3. Schaap, L. H. J. and Zuidberg, H. M., "Mechanical and Electrical Aspects of

the Electric Cone Penetrometer Tip", Proceedings, 2nd European Symposium on
Penetration Testing, Vol. 2, Amsterdam, 1982, pp. 841-851.

I 4. De Ruiter, J., "The Static Cone Penetration Test: State-of-the-Art Report",

Proceedings, 2nd European Symposium on Penetration Testing, Vol. 2, Amsterdam,
1982, pp. 389-405.

I 5. De Ruiter, J., "Electric Penetrometer for Site Investigations", Journal of the

Soil Mechanics and Foundations Division, ASCE, Vol. 97, No. SM2, Feb. 1971,
pp. 457-472.

I 6. Schmertmann, J. H., "Guidelines for Cone Penetration Test Performance and

Design", Report FHWA-TS-78-209, Federal Highway Administration, Washington,
1978' 145 p.

I 7. Smits, F. P., "Cone Penetration Tests in Dry Sand", Proceedings, 2nd European
Symposium on Penetration Testing, Vol. 2, Amsterdam, 1982, pp. 887-881.

I 8. Rol, A. H., "Comparative Study on Cone Resistances Measured with Three Types
of CPT Tips", Proceedings, 2nd European Symposium on Penetration Testing, Vol.
2, Amsterdam, 1982, pp. 813-819.

I 9. Amar, S., "The Use of the Static Penetrometer in the Laboratoires des Fonts et
Chaussees", Proceedings, European Symposium on Penetration Testing, Vol. 2.2,
Stockholm, 1974, pp. 7-12.

I 10. Baligh, M. M., Vivatrat, V., and Ladd, C. C., "Exploration and Evaluation of
Engineering Properties for Foundation Design of Offshore Structures", Report
R78-40, Dept. of Civil Engineering, M.I.T., Cambridge, Dec. 1978, 268 p.

I 11. Joustra, K., "Comparative Measurements on the Influence of the Cone Shape on
Results of Soundings", Proceedings, European Symposium on Penetration Testing,
Vol. 2.2, Stockholm, 1974, pp. 199-200.

I
I B-10

I
 I
12. Kok, L., "The Effect of the Penetration Speed and Cone Shape on CPT Results",
Proceedings, European Symposium on Penetration Testing, Vol. 2.2, Stockholm, I
1974, pp. 215-216.

13. Heij nen, W. J. , "The Dutch Cone Test: Study of the Shape of the Electrical
Cone", Proceedings, 8th International Conference on Soil Mechanics and Founda-
I
tion Engineering, Vol. 1.1, Moscow, 1973, pp. 79-83.

14. Jones, G. and Rust, E., "Piezometer Penetration Testing", Proceedings, 2nd
European Symposium on Penetration Testing, Vol. 2, Amsterdam, 1982, pp. 607-
I
613.

15. De Beer, E. E., "Scale Effects in Results of Penetration Tests Performed in

Stiff Clays", Proceedings, European Symposium on Penetration Testing, Vol.
I
2.2, Stockholm, 1974, pp. 105-115.

16. Mayne, P. W. and Kemper, J. B., "Profiling OCR in Stiff Clays by CPT and SPT",
Geotechnical Testing Journal, ASTM, Vol. 11, No. 2, June 1988, pp. 139-147.
I
17. Mayne, P. W., Frost, D. D., and Swanson, P. G., "Geotechnical Report, CEBAF
Project, Newport News, VA", Law E 1gineering NK5-1182 to Daniel, Mann, Johnson,
I
and Mendenhall, Washington, Dec. 1986, 580 p.

18. Dobie, M. J. D., "A Study of Cone Penetration Tests in the Singapore Marine
Clay", Proceedings, 1st International Symposium on Penetration Testing
I
(ISOPT-1), Vol. 2, Orlando, 1988, pp. 737-744.

19. Lunne, T., Eidsmoen, D., Gillespie, D., andHowland, J., "LaboratoryandField
Evaluation of Cone Penetrometers", Use of In-Situ Tests in Geotechnical Engi-
I
neering (GSP 6), Ed. S. P. Clemence, ASCE, New York, 1986, pp. 714-729.

20. Campanella, R. G. and Robertson, P. K., "Current Status of the Piezocone

Test", Proceedings, 1st International Symposium on Penetration Testing (ISOPT-
I
1), Vol. 1, Orlando, 1988, pp. 93-116.

21. Campanella, R. G. and Robertson, P. K., "State-of-the-Art in In-Situ Testing

of Soils; Developments Since 1978", Proceedings, Engineering Foundation Con-
I
ference on Updating Subsurface Sampling of Soils and Rocks and Their In-Situ
Testing, Santa Barbara, 1982, 23 p. I
22. Mayne, P. W. , Kulhawy, F. H. , and Kay, J. N. , "Observations on the Development
of Pore Water Stresses During Cone Penetration in Clays", Canadian Geotechni-
cal Journal, Vol. 27, No. 3, Aug. 1990. I
23. Kulhawy, F. H., Trautmann, C. H., Beech, J. F., O'Rourke, T. D., McGuire, W.,
Wood, W. A., and Capano, C., "Transmission Line Structure Foundations for
Uplift-Compression Loading", Report EL-2870, Electric Power Research Insti-
tute, Palo Alto, 1983, 412 p.
I
24. Orchant, C. J., Kulhawy, F. H., andTrautmann, C. H., "Reliability-BasedFoun-
dation Design for Transmission Line Structures: Critical Evaluation of In-Situ
Test Methods", Report EL-5507, Vol. 2, Electric Power Research Institute, Palo
I
Alto, 1988, 214 p.
I
I
B-11 I
I
I
I
I Appendix C

I PRESSUREMETER TEST

I The pressuremeter test (PMT) can be used in soil to determine the in-situ stress,
deformability, and strength. A cylindrical probe is advanced to the test depth by
one of several means and is then expanded incrementally with either liquid or gas
I pressure. During expansion, the diameter or volume of the expanding probe is meas-
ured accurately to yield a volume versus pressure curve, from which the in-situ

I stress, stress-strain behavior, and strength properties can be estimated.

ginal Menard-type PMT is performed in a prebored hole.
The ori-
A more sophisticated device
is the self-boring pressuremeter (SBPMT), which minimizes stress relaxation and
I soil disturbance during insertion. More recently, the push-in pressuremeter and
full-displacement pressuremeter have been introduced, primarily for offshore work,

I and these may be operated more quickly without need for a prebored hole.

PROCEDURE
I A standard procedure for the prebored PMT has been developed recently in the U.S.
as ASTM D4719 (l). Specific details on the traditional test equipment and inter-

I pretation are given by Baguelin, et al. (l). After calibration, the pressuremeter
probe is installed at the test location by lowering it down a borehole, jacking it

I into the ground, or by self-boring. The latter technique is useful in soft and
medium clays, but specialized equipment is required, and shells and gravel parti-
cles can obstruct proper functioning of the probe. Figure C-1 illustrates a PMT

I installation.

I The test is carried out by applying pressure in about ten equal steps. The pres-
sure is maintained constant for each step for the same period of time, such as 60
seconds. The volumetric expansion of the probe is measured at 15, 30, and 60

I seconds after each pressure step to determine a creep curve. The test ends when
the probe has been expanded to twice its deflated volume or when the pressure limit

I of the device has been reached. Once the test has been completed, the probe is
deflated, and the device is either advanced to a new depth or returned to the sur-
face.
I
I C-1

I
 I
Control
I
unit

I
/
I
I
Borehole
I
I
Probe
I
Figure C-1. Menard Pressuremeter Equipment
I
Source: Baguelin, et al. (£), p. 47.
I
PMT results are presented generally as a plot of pressure versus volume, as shown I
in Figure C-2. Three characteristic pressures are determined from this curve:

• p 0 , representing the pressure at which recompression of disturbed

soil is complete and expansion into undisturbed material is ini-
I
tiated (It often is assumed that p 0 - aho, the in-situ total hori-
zontal stress.)
I
• pf, an inflation point, known as the creep or yield pressure, where
the soil behavior changes from pseudo-elastic to plastic and shear-
ing is initiated
I
• PL• the limit pressure, representing the pressure to which the curve
becomes asymptotic
I
The limit pressure is never measured directly. Instead, it is determined by extra-
polation as the pressure at which the probe has expanded to twice its original vol-
ume. A review of the available methods for interpreting p 0 , pf, and PL is given by
I
Ladd, et al. (2). These three characteristic pressures are used to estimate anum-
ber of engineering soil properties and for direct semi-empirical correlations to I
C-2
I
I
I
I
I .._
Q)

:J

I V)
V)
.._
Q)

a_

I
I Volume

Figure C-2. Typical Pressuremeter Test Curve

I
I foundation capacity and settlement.
test and not a logging tool.
The PMT is considered a specific soil property
Therefore, the soil must be characterized in advance
of the test for the PMT results to be used efficiently and economically.

I The self-boring pressuremeter test (SBPMT) is a relatively recent development, and

I as such it is just reaching maturity in terms of the equipment and procedures

employed. ·Two basic types of self-boring probes are currently in use: a French
version, known as the PAFSOR, and an English device called the Camkometer. Both

I are shown in Figure C-3.

I Among the most attractive features of the SBPMT is its ability to provide reason-
able estimates of the in-situ horizontal stress. A graphical procedure is used to
estimate aho from Camkometer data. An enlarged plot of the initial portion of the

I expansion curve for each displacement transducer is analyzed. Then aho equals the
"lift off" pressure or the pressure at which volumetric expansion of the membrane

I is first recognized. An example of this procedure is shown in Figure C-4.

noted in this figure, the three feeler arms may display substantially different
As

"lift off" pressures. This phenomenon has been attributed to one or more of the

I following factors (e.g.,~): soil stiffness, relative stiffnesses of each feeler

arm, noncircular shape of SBPMT hole, mechanical compliance of the instrument,

I deviation of the probe from the vertical, non-uniform shear stress at the probe
soil interface, and anisotropy of the in-situ horizontal stress.

I In terms of deformation parameters, the pressure-strain curve obtained from the

I C-3

I
 I
\ ;n jection of water for
'l'embrone expansion
Slurried
water
I
and soil---._ +:---Flushing water
/Coble and gas
pressure tube I
Reinforced
rubber
membrane Rubber membrane
I
Feeler
Pore water pressure cell
I
Clomp J
I
a) PAFSOR

Figure C-3.
b) Camkometer

Self-Boring Pressuremeters
I
Source: Jamiolkowski, et al. (~). pp. 98, 99.
I
I
Q)
I
=:;

I
en
en
Q)
.....
a... 0
:2 -u
0
-u

I
0 0::
0::
o)Soft b) Hard
Cloy E Cloy
4

Radial Displacement Radial Displacement

I
Figure C-4.

Source:
Examples of "Lift Off" Pressure

Jamiolkowski, et al. (~). p. 100.

I
I
SBPMT can provide estimates of the initial tangent shear modulus (Gi), secant shear
modulus (Gs), unload-reload shear modulus (Gur), and reload-unload shear modulus
(Gru).
I
C-4 I
I
I
I ADVANTAGES AND DISADVANTAGES

The main advantage of the pressuremeter test is that it is one of the few in-situ

I measurement techniques that can assess directly the state of horizontal stress in
soil. This capability is a significant advantage for the design of deep founda-

I
tions because the capacity of these foundations is directly related to the in-situ
stress. In addition, the PMT is capable of yielding data on soil modulus and
shearing resistance when performed carefully in appropriate materials.

I The PMT also has a number of disadvantages. It generally is performed in soil

I
deposits that have been identified previously using other forms of in-situ testing
or sampling. Therefore, like the vane shear test, prior exploration is required
for proper interpretation of the test results.

I From a soil mechanics point of view, the test has several limitations. The drain-
age conditions in soils of intermediate permeability are generally unknown during
I the test, which can seriously impair test interpretation. Pressuremeters of the
self-boring variety can, in some cases, provide the most accurate data because they

I cause minimal soil disturbance, but they are most reliable in relatively soft,
fine-grained soils that do not contain shells, gravel particles, or cohesionless

I
sands. Recent improvements in self-boring techniques have extended the range of
soils that can be penetrated, but gravel particles remain an important limitation
for self-boring pressuremeters. Test accuracy is still subject to drilling proce-

I dures, insertion techniques, and the human element in both performance and inter-
pretation, which includes instrument calibration, the theory used for interpreta-
tion, and prior knowledge of soil stratification. Strain-rate effects are impor-
I tant, and semi-empirical correlations with documented case histories still are
required to use the test results in design. Also, long test times may be required

I for testing some relatively impermeable cohesive soils.

I
SOURCES OF ERROR, RELIABILITY, AND COST

The PMT has a number of potential sources of error, largely because of the complex
nature of the test equipment and procedure(~,~-£). Equipment calibration,

I leakage, borehole preparation, probe insertion, prior knowledge of soil stratifi-

cation, and test interpretation are all important considerations, and trained

I personnel must perform the test. In addition, the strength and modulus values
obtained from the PMT are not strictly comparable to those derived from other forms
of in-situ testing, so the values can not be used indiscriminately in classical

I design methods without leading to erroneous results in some cases. A list of the

I C-5

I
 I
major variables affecting the PMT and SBPMT is given in Table C-1. I
The reliability of the PMT is greatest in homogeneous, finer-grained soil.
skilled operators and good equipment and procedural controls, the test is highly
With
I
reproducible in these soil types.

Pressuremeter tests are higher in cost compared with the SPT and VST. All three
I
require the same type of test boring, but the PMT requires a skilled operator in
addition to the drilling crew. Taking into account drilling costs, the operator,
I
and a productivity of 5 to 8 tests per shift, the cost per test in 1990 is in the
I
Table C-1

MAJOR PMT AND SBPMT VARIABLES I

I
Variable Relative Effect on Test Results

Gage error Minor

I
Expansion of tubing

Frictional losses in tubing

Minor to moderate

Minor
I
Probe dimensions

Probe design (PMT)

Minor to moderate

Minor
I
Membrane aging Minor I
Size of cutting shoe (SBPMT) Moderate to significant

Cutter position (SBPMT) Minor I

Shape of probe (SBPMT) Minor to moderate

Drilling equipment (SBPMT) Minor to moderate I

Compliance of electrical sensors (SBPMT) Minor

Method of drilling and borehole preparation (PMT) Significant I

Minor to moderate
Rate of probe inflation

Relaxation time (SBPMT) Moderate to significant

I
Rate of probe advance (SBPMT)

Source: Orchant, et al. (~). pp. 4-49, 4-51.

Moderate to significant
I
C-6 I
I
I
I range of $150 to $250 for the standard PMT and $300 to $600 for the SBPMT.

REFERENCES
I 1. American Society for Testing and Materials, "Test Method for Pressuremeter
Testing in Soils (D4719-87)", Annual Book of Standards, Vol. 4.08, ASTM,

I 2.
Philadelphia, 1989, pp. 881-888.

Baguelin, F, Jezequel, J. F., and Shields, D. H., The Pressuremeter and Foun-
dation Engineering, Trans Tech Publications, Clausthal, 1978, 617 p.

I 3. Ladd, C. C., Germaine, J. T., Baligh, M. M., and Lacasse, S.M., "Evaluation
of Self-Boring Pressuremeter Tests in Boston Blue Clay", Report FHWA/RD-80/
052, Federal Highway Administration, Washington, 1980, 224 p.

I 4. Jamiolkowski, M., Ladd, C. C., Germaine, J. T., and Lancellotta, R., "New
Developments in Field and Laboratory Testing of Soils", Proceedings, 11th
International Conference on Soil Mechanics and Foundation Engineering, Vol. l,
I 5.
San Francisco, 1985, pp. 57-153.

Kulhawy, F. H., Trautmann, C. H., Beech, J. F., O'Rourke, T. D., McGuire, W.,

I
Wood, W. A., and Capano, C., "Transmission Line Structure Foundations for
Uplift-Compression Loading", Report EL-2870, Electric Power Research Insti-
tute, Palo Alto, 1983, 412 p.

I 6. Orchant, C. J., Kulhawy, F. H., and Trautmann, C. H., "Reliability-Based Foun-

dation Design for Transmission Line Structures: Critical Evaluation of In-Situ
Test Methods", Report EL-5507, Vol. 2, Electric Power Research Institute, Palo
Alto, 1988, 214 p.

I
I
I
I
I
I
I
I
I C-7

I
 I
I
Appendix D
I
DILATOMETER TEST
I
The flat-plate dilatometer test (DMT) became commercially availabie in the U.S. in
1979. Although the use of this test in routine site investigation practice has
I
been relatively recent, a number of factors associated with the DMT, including its
relative ease of operation and durability, suggest that its use will increase. I
PROCEDURE

The equipment required to perform DMT tests is shown in Figure D-1. The dilatome-
I
ter itself is a flat blade or plate, 14 mm (0.55 in) thick, 95 mm (3.74 in) wide,
and 220 mm (8.66 in) long. A flexible stainless steel membrane, 60 mm (2.36 in) in I
diameter, is located on the center and flush with one side of the blade. A combi-
nation gas and electrical line extends from a surface control box through the push
rods and into the blade. A special hydraulic system has been developed for off-
I
shore use.
I
I
I
I
I
60mm (2.36in.)
diameter membrane
95mm
I
0 ] (3.74 in.)

I
Figure D-1. Dilatometer Test Equipment

Source: Schmertmann (l), p. 95. I

D-1 I
I
-
 I
I Although the test is not yet standardized, a recommended ASTM procedure has been
developed by Schmertmann (1). The test is performed by pushing the blade to the

I desired test depth at a typical rate of penetration of 20 mm (0.8 in)jsec.

depths may be taken as frequently as 200 mm (8 in), although more typically in the
Test

U.S., the intervals are 300 mm (l ft). The blade can be pushed with a CPT hydrau-

I lic jacking rig, the hydraulics of a rotary drilling rig, or a hammer and rod sys-
tem as used ~n the SPT. Upon achieving the desired test depth, the operator uses a

I control valve at the surface to inflate the membrane with high pressure nitrogen
gas. Typically two readings are recorded, prompted by audio and visual signals at
the control box. The first, called the A reading, represents the pressure at which

I the membrane "lifts off" its sensing disc, which ideally represents initial contact
with the soil. The second, called the B reading, is made after l mm (0.04 in)
The operator vents the pressure after obtaining the B
I deflection has occurred.
reading. Recently, a third reading, designated as the closing pressure or C read-
ing, has been proposed as a measure of the total pore water stress. The C reading

I is similar to the A reading, except that it is obtained during deflation.

these measurements, the blade is pushed to the next test depth, at which the test
After

Each cycle typically takes l to 2 minutes to complete.

I
cycle is repeated. After
each complete profile, the membrane response of the dilatometer should be cali-
brated.

I Recent devices include load cells capable of measuring static thrust (1) and piezo-
metric elements to monitor the pore water stresses generated during penetration
I (£). The inclusion of these electronic sensors enhances the information obtained
from the DMT, but they also increase the complexity of the test substantially.

I The A and B readings obtained during the test must be corrected by calibration of
the measuring gage and the membrane response. During calibration of the device,
I two readings, ~and ~B, are made. ~ is the vacuum pressure required to keep the
membrane in contact with its seating, because after a number of expansions the mem-

I brane develops an outward curvature.

the membrane 1.1 mm (0.043 in) in air.
~B is the air pressure required to deflect
The corrected in-situ data for the contact
stress (p 0 ) and the expansion stress (pl) are expressed as:
I (D-1)
Po l.OS(A + ~ - Zm) - 0.05 Pl

I Pl B - ~B (D-2)

I with Zm = gage pressure deviation from zero when vented to atmospheric pressure.

I D-2

I
 I
Using these values, three index parameters are defined as follows (1): I
(D-3)
I
(D-4)

Eo= 34.7(pl- p 0 ) (D-5)

I
in which Io = material index, u 0 = assumed hydrostatic pore water stress, Ko = hor-
I
izontal stress index, ovo = in-situ effective vertical stress, and Eo= dilatometer
modulus. In the original work for this test (l), Equations D-1 and D-2 were some-
what different, and the coefficient in Equation D-5 was equal to 38.2. When this
I
test was introduced, correlations of these index parameters to a variety of soil
properties were proposed. Most were based on limited field data and are empirical,
I
although some of the more recent relationships have a more theoretical basis.

ADVANTAGES AND DISADVANTAGES

I
The DMT offers a number of advantages.
form.
First, the test is simple and rapid to per-
The equipment is rugged, and the test can be used in a wide variety of
I
soils. Also, the blade-like shape reduces the shear and volumetric strain associ-
ated with other penetration tests. As indicated above, the DMT tentatively has I
been correlated to a number of soil properties. Specifically, the test may provide
reasonable estimates of the horizontal stress and the overconsolidation ratio,
which are traditionally difficult properties to measure. The proposed empirical
I
correlations, although requiring a substantial database for verification, relate
the test results to basic geotechnical engineering parameters. The test data can I
be reduced quickly in the field, which allows evaluation of anomalous results. In
addition, these test results and inferred soil properties can be plotted in a
nearly continuous profile to illustrate the variations with depth. Also, the test
I
equipment is relatively inexpensive and, because the test is rapid, numerous data
points can be obtained quickly. I
The DMT also possesses several notable disadvantages.
is a recent test which has had limited field exposure.
First and most important, it
Therefore, the general
I
validity of the soil property correlations is uncertain.
do not possess the equipment required to perform the DMT.
Second, most contractors
Third, as with any pene- I
tration test, the DMT has limited use in very dense or cemented soils and in soils
containing appreciable gravel or coarser fragments. In the case of gravelly
deposits, the blade may deviate from vertical penetration, causing difficulty in
I
D-3 I
I
I
I interpreting the horizontal stress parameters and, in some cases, the blade may be
bent or the inflatable membrane may be torn. Fourth, the test requires the addi-
tional measurement of thrust to evaluate the strength and stress history of cohe-
I sionless soils. These thrust measurements, as well as other electronic sensors
such as pore water stress elements which facilitate interpretation of the DMT,

I detract from the simplicity, ruggedness, and low cost of the test. Finally, this
test suffers from the common limitation that it does not obtain soil samples.

I SOURCES OF ERROR, RELIABILITY, AND COST

The DMT has a number of potential sources of error, as noted in Table D-1. Perhaps

I most important is that the test is quite new, and experience with the test is lim-
ited. Its real potential as a field production tool has yet to be assessed, and

I correlations with the DMT parameters have been limited to date.

The reliability of the test is difficult to determine precisely at the present time

I
Table D-1
I MAJOR DMT VARIABLES

I
Variable Relative Effect on DMT Results

I Leaking seals Minor to significant

I Deformed membrane

Bent or deformed push rods

Moderate to significant

Minor to moderate

I Damaged blade

Poor electrical ground

Minor to significant

Significant

I Inclination of push rods Minor to moderate

Minor to moderate
Rate of testing

I Method of driving Minor to significant

Rod friction Minor (except for thrust measurement)

I Calibration error Minor to moderate

Waiting time after insertion Significant in silts

I Source: Modified after Orchant, et al. (~), p. 4-42.

I D-4

I
 I
because of a shortage of detailed test data. However, the limited data to date are
encouraging and suggest good reproducibility and relatively high reliability.
I
The costs associated with dilatometer testing are comparable, to slightly more I
expensive, than those described for the CPT. Typical DMT costs have been reported
to be about $12 to $15/ft ($40 to $50/m) with a CPT truck and about $15 to $20/ft
($50 to $65/m) with a standard drill rig. Because of its simple and expedient
I
operation, the DMT is becoming popular and available for production testing.
specialty in-situ testing firms with large cone trucks also offer DMT services.
Many
I
REFERENCES
I
1. Schmertmann, J. H., "Suggested Method for Performing the Flat Dilatometer
Test", Geotechnical Testing Journal, ASTM, Vol. 9, No. 2, June 1986, pp. 93-
101. I
2. Lutenegger, A. J., "Current Status of the Marchetti Dilatometer Test", Pro-
ceedings, lst International Symposium on Penetration Testing (ISOPT-1), Vol.
1, Orlando, 1988, pp. 137-156. I
3. Marchetti, S., "In-Situ Tests by Flat Dilatometer", Journal of the Geotechni-

4.
cal Engineering Division, ASGE, Vol. 106, No. GT3, Mar. 1980, pp. 299-321.

Orchant, G. J., Kulhawy, F. H., and Trautmann, C. H., "Reliability-Based Foun-

I
dation Design for Transmission Line Structures: Critical Evaluation of In-Situ
Test Methods", Report EL-5507, Vol. 2, Electric Power Research Institute, Palo
Alto, 1988, 214 p.
I
I
I
I
I
I
I
I
D-5 I
I
I
I
I Appendix E

I VANE SHEAR TEST

I The vane shear test (VST) is a moderately rapid and economical in-situ method for
determining the peak and remolded undrained shear strength of soft to medium stiff
clays. The test involves pushing a four-bladed vane into a clay stratum and slowly

I rotating it while measuring the resisting torque.

I PROCEDURE

The procedure for the VST is described in ASTM 02573 (l). Important related issues
(~. 1. ~).

I
are given elsewhere The test generally is used to determine the shear
strength of a cohesive soil once its location has been established. In the test, a
shear vane similar to those shown in Figure E-l is pushed into undisturbed soil and

I is rotated from the surface at a standard rate of 0.1 degrees per second. The peak

I
I Casing Diameter, 0
in. mm

I ~H=20
AX
BX
NX
1.5
2.0
2.5
38.1
50.8
63.5

j_
4in.(I02mm) 3.625 92.1·

I
~o--l

I
~;Sin. ~.0625in. {1.6mm) for AX and BX casing

I 2.7mml
~125in.
1 {3.2mm) for NX and 4in. (102mm) casing

I Rectangular Vane Tapered Vane

Figure E-1. Vane Geometries and Sizes

I Source: American Society for Testing and Materials (l), p. 304.

I E-1

I
 I
torque which develops is related to the peak shear strength on a cylindrical fail-
ure surface by a constant, which is a function of the shape and dimensions of the I
vane. Details are given in ASTM D2573 (!). The VST may be conducted either at the
bottom of a prebored hole or, in soft clays, by merely pushing the vane rods to the
desired test depth. The latter method requires a correction for rod friction.
I
After the peak torque has been determined, the vane is rotated quickly about ten I
times to remold the soil. The torque then is measured again to determine the
remolded shear strength.
the peak to remolded strength.
The sensitivity (St) may be calculated as the ratio of
Numerous tests can be performed sequentially in the
I
same deposit, but individual tests should be separated vertically by at least 0.75
m (30 in). I
Another method of testing uses vane borers, as shown in Figure E-2.
device, the rods are surrounded by a sleeve to minimize friction losses, and the
With the SGI
I
vane is covered by a protective shoe during penetration. At the desired test
I
I
I
I
950mm
[37.4 in.]

I
500mm
[19.7in.J
I
A-A
w 130 (IIO)mm
[5.12 (4.33lin.]

Vane_/
I
1-----..J
65(55)mm
[2.56 (2.17) in.]
I
a) Swedish Geotechnical b) Nilcon
Institude I
Figure E-2. Common Vane Borers

Source: Walker (~). p. 68. I

E-2
I
I
depth, the vane is advanced into the soil beneath the protective shoe. The other
device is the Nilcon vane borer, which does not have either a protective sleeve or
shoe. However, the vane is followed by a slip coupling during penetration, which
provides for rod friction calibration before each test.

The maximum measured torque (T) in the VST is used to calculate the undrained shear
strength (su) as follows (l):

T/K (E-1)

in which T = torque in N-m or lb-ft and K constant depending on the dimensions

and shape of the vane (m3 or ft3), where:

K ~(D2H/2) [1 + (D/3H)] for D and H in meters (E-2)

K (~/1728) (D2H/2) [1 + (D/3H)) for D and H in inches (E-3)

A number of assumptions are made in calculating the undrained shear strength from
these torque measurements (l), including:

• The soil is completely undrained, i.e., no consolidation takes place

during insertion of the vane or during the test.

• No disturbance is caused by the boring operation or installation of

the vane.

• The remolded zone around the vane is very small.

• There is no progressive failure so that the maximum applied torque

overcomes the fully-mobilized shear strength along the cylindrical
surface.

• Isotropic strength conditions exist in the soil mass.

ADVANTAGES AND DISADVANTAGES

The VST has many advantages when used in soil deposits for which it is intended.
The test is moderately rapid and economical, and it is reproducible in homogeneous
deposits. The scatter in test results is on the same order as that for the con-
fined and unconfined compression tests with which it is compared. The test has had
extensive usage during the past few decades, and a large body of literature is
available for use in correlations with other test and design methods. The effect
of the vane size is minor in most types of soil and, by using two vanes with dif-
ferent length to diameter ratios in the same stratum, the soil strength anisotropy

E-3
 I
can be inferred. Additionally, the test is an inexpensive way to determine the
properties of sensitive clays, which are characteristically difficult to obtain in
I
the laboratory without extreme care.
I
The VST has a number of important limitations that influence its usefulness. The
test is most easily interpreted for soft and medium stiff clays which have been
previously identified by some other test or sampling procedure. Also, it is useful
I
mainly for analyses requiring the undrained shear strength. II
SOURCES OF ERROR, RELIABILITY, AND COST

The VST may be in error because of excessive rod friction, poor torque calibra- I
tions, non-standard rotation rates, and other factors (~, 2, £). A list of the
major sources of error with the VST is given in Table E-1. I
In addition to these test uncertainties, the theoretical nature of the failure
I
Table E-1 I
MAJOR SOURCES OF ERROR IN THE VANE SHEAR TEST

I
Cause Effect
Influence on Strength
Measurement I
Friction between torque rods
and soil or casing
Measured torque includes spu-
rious component of resistance
Increases
I
I
Poorly calibrated torque Inaccurate torque Increases or
measurement decreases

Vane rotated too quickly Soil sheared too rapidly Increases

Test performed in disturbed

soil
Soil structure broken down Decreases I
Damaged vane Disturbed soil excessively Decreases
peak strength I
Unknown sand/silt/shell Drainage during test Increases
lenses

Measured torque includes spu- Increases

I
Isolated gravel/cemented
nodules

Source:
rious component of resistance

Adapted from Kulhawy, et al. (2), p. 5-34.

I
E-4 I
I
 mechanism is not fully understood. Therefore, the correlation between field and
laboratory measurements of the same soil contains a significant element of uncer-
tainty. On the basis of published studies, the random variations between tests
made in the same soil are much smaller than the uncertainties associated with the
test procedure.

Vane shear tests are comparable in cost to the SPT, taking into account that both
require a test boring. During an average shift, approximately 10 to 15 tests can
be performed. Based on 1990 drilling costs, this indicates that the average cost
of a VST is about $70 to $150. However, it should be noted that the VST can be
alternated with the SPT in a single test boring to optimize the return of informa-
tion from a single borehole.

REFERENCES

l. American Society for Testing and Materials, "Standard Test Method for Field
Vane Shear Test in Cohesive Soil [D2573-72(1978)]", Annual Book of Standards,
Vol. 4.08, ASTM, Philadelphia, 1989, pp. 308-310.

2. Walker, B. F., "Vane Shear Strength Testing", In-Situ Testing for Geotechnical
Investigations, Ed. M. C. Ervin, A. A. Balkema, Rotterdam, 1983, pp. 65-72.

3. Flaate, K., "Factors Influencing the Results of Vane Tests", Canadian Geotech-
nical Journal, Vol. 3, No. l, Feb. 1966, pp. 18-31.

4. Richards, A. F., "Overview", Vane Shear Strength Testing in Soils: Field and
Laboratory Studies (STP 1014), ASTM, Philadelphia, 1988, pp. 1-9.

5. Kulhawy, F. H., Trautmann, C. H., Beech, J. F., O'Rourke, T. D., McGuire, W.,
Wood, W. A., and Capano, C., "Transmission Line Structure Foundations for
Uplift-Compression Loading", Report EL-2870, Electric Power Research Insti-
tute, Palo Alto, 1983, 412 p.

6. Orchant, C. J., Kulhawy, F. H., and Trautmann, C. H., "Reliability-Based Foun-

dation Design for Transmission Line Structures: Critical Evaluation of In-Situ
Test Methods", Report EL-5507, Vol. 2, Electric Power Research Institute, Palo
Alto, 1988, 214 p.

I
I
I
I
I E-5

I
 I
I
Appendix F
I
COMPARISON OF IN-SITU TEST METHODS
I
Mitchell (l) has reviewed the various types of in-situ test procedures and classi-
fied each according to a variety of parameters. A modified version of his summary
I
is shown in Table F-1 and can be used qualitatively in designing a field explora-
tion program, once a preliminary study has been completed to determine the general I
types of geologic materials likely to be encountered along the route.

In addition to the common in-situ tests described in Appendices A through E, a num-

I
ber of other tests exist which serve special purposes or have not gained wide usage
to date. These other tests include the Iowa borehole shear device, the Gl6etzl I
Table F-1
I
ASSESSMENT OF IN-SITU TESTS
I
Standard Cone Vane Pressure- Flat
I
Comparison Penetration Penetration Shear meter Dilatometer
Basis Test Test Test Test Test
I
Simplicity of Simple, Complex, Simple, Complex, Simple,
apparatus

Ease of testing
rugged

Easy
rugged

Easy
rugged

Easy
delicate

Complex
rugged

Easy
I
Continuous profile
or point values
Point Continuous Point Point Semi-con-
tinuous I
Basis for inter- Empirical Empirical, Theory Empirical, Semi-empiri-
pretation theory theory cal, theory
I
Suitable soils Most types Most types Softer Most types Most types

Suitability in Routine Routine

clays

Routine Limited Routine

I
practice

Source: Modified from Mitchell (l), pp. 121, 123. I

F-1 I
I
I
I total pressure cell, seismic cone, K0 -stepped blade, acoustic cone, large diameter
penetration test (LPT), Becker probes, and screw-plate tests, among others. Some

I of these tests may become common in the future.

When comparing test methods, it is very important to consider the cost-effective-

I ness of the information obtained. Handy (~)

qualitative relationships for both field and laboratory tests.

considered this point and developed
Table F-1 summa-

I
rizes the general applicability of the five major types of in-situ tests covered in
this manual. The usefulness of the various in-situ test methods in different soils
is summarized in Table F-2.

I The degree of historical use and the general familiarity of an in-situ technique

I commonly are important considerations in assessing their applicability to a given

project, because there is an added element of risk involved in using techniques

I Table F-2

I USEFULNESS OF IN-SITU TESTS IN COMMON SOIL CONDITIONS

I gravel sand
Soil T e
silt clay
loose dense soft stiff
I
Test

SPT 2 to 3 l l 2 3 3

I MCPT 2 to 3 1 2 1 1 2

ECPT 3 1 2 1 1 2

I CPTU 3 1 2 1 1 2

VST 4 4 4 3 l 2

I IMT 3 1 2 1 1 2

FMT 2 2 1 1 1 1
I SBPMT 3 2 2 1 1 1

I Notes: 1
2
3
Highly applicable
Moderately applicable
Limited applicability
MCPT
ECPT
CPTU
-
-
-
Mechanical CPT
Electric CPT
Piezometric CPT
4 Not applicable SBPMT - Self-boring PMT

I Source: Orchant, et al. (1), p. 2-61.

I F-2

I
 I
which have not been field-proven. I
Also, access to testing locations for some projects may be difficult. Therefore,
logistical concerns of equipment mobilization and access requirements must be con-
I
sidered in the overall applicability of in-situ techniques for transmission line
site characterization. These considerations must be evaluated on a project by pro-
ject basis because requirements will vary.
I
Test costs are related to the above logistical concerns. Since limited allocations
I
are available for most geotechnical projects, test economics may govern their
application for a given project. Table F-3 summarizes the historical use, mobili-
zation and access requirements, and relative costs of the tests reviewed.
I
Selection of the most suitable test for a specific project is governed by the type
I
of information required for the applied design method. In some cases, specific
soil property estimates are required; in others, empirical design models based on
in-situ test results are employed. These factors must be addressed as well.
I
I
Table F-3

HISTORICAL USE, MOBILIZATION AND ACCESS REQUIREMENTS, AND COSTS OF IN-SITU TESTS
I
I
Historical Unit
Test Use Availability Access Cost
I
SPT Substantial Excellent Truck, trailer Medium

MCPT Substantial Good Limited portability - truck, trailer Low I

ECPT

CPTU
Moderate

Limited
Good

Poor
Limited portability - truck, trailer

Limited portability - truck, trailer

Low

Medium
I
VST

DMT
Substantial

Limited
Excellent

Fair
Limited portability - truck, trailer

Limited portability - truck, trailer

Medium

Low
I
PMT

SBPMT
Moderate

Limited
Good

Poor
Limited portability

Limited portability
-
-
truck, trailer

truck, trailer
Medium

High
I
Source: Orchant, et al. (1). p. 2-62.
I
F-3 I
I
 I
I In addition to the test conditions summarized in the above tables, the sources of
error and the magnitude of uncertainty associated with particular tests will influ-

I ence their applicability. These factors are considered in a qualitative sense in

assessing the ability of the test to obtain specific soil property data. However,
to allow direct comparisons between tests and assess their potential to provide

I reliable design input, quantitative information regarding the variability of the

test results is required.

I The variability of the various in-situ testing methods has been evaluated by
Orchant, et al. (l), and the expected coefficient of variation (COV =ratio of

I standard deviation to mean value) for each test is summarized in Table F-4.
analysis is based on a statistical review of data from numerous sites tested by
The

I
each apparatus. In terms of reliability, the electric cone and dilatometer appear
to be less variable than the vane shear test and pressuremeter. The mechanical
cone and standard penetration test are the most variable test methods.

I Finally, the relative accuracy of the device must be weighed against its relative

I
cost. A qualitative relationship between relative cost and accuracy for the vari-
ous field test methods is given in Figure F-l.

I
I
I
I
I
I
I
I
I F-4

I
 I
Table F-4 I
ESTIMATES OF IN-SITU TEST VARIABILITY

I
Test
rova (%)
Equipment
cov (%)
Procedure
cov (%)
Random
covb
Total
(%) cove
Range
(%)
I
Standard
Penetration
Test (SPT)
12 to 15 15 to 45
I
Mechanical Cone
Penetration
5 15 to 25
I
Test (MCPT)

Electrical Cone
Penetration
3 5 5 to 15 I
Test (ECPT)

Vane Shear Test

(VST)
5 8 10 14 10 to 20 I
Di1atometer
Test (DMT)
5 5 8 11 5 to 15
I
10 16 10 to 20h
Pressuremeter 12
I
5
Test (PMT)

Self-Boring 8 15 8 19 15 to 25h
Pressuremeter
Test (SBPMT) I
Notes:
a - COV = standard deviation/mean
b- COV(Total) = (COV(Equipment)2 + COV(Procedure)2 + COV(Random)2]~
c - Because of limited data and the judgment involved in estimating COV values,
I
ranges represent probable magnitudes of field test measurement error
d - Best case scenario for SPT test conditions
e - Worst case scenario for SPT test conditions
f - Tip resistance CPT measurements
I
g - Side resistance CPT measurements
h - It is likely that results may differ for p 0 , p f• and PL• but the data are
insufficient to clarify this issue
I
Source: Orchant, et al. (l), p. 4-63.
I
I
I
F-5 I
j
I
 I
I SPT Standard Penetration Test

I
DMT Oilatometer Test
MCPT Mechanical Cone Penetration Test
ECPT Electrical Cone Penetration Test

I -( /)
Q)
f-
,._ Self- Boring Pressuremeter

I Q)
0.

(/)
I
0 Menard Pressuremeter

I u
Q)
/
Borehole Shear
-
>
0
SPT /
Vane Shear

I
Q)
n:: I
/_..- ECPT
MCPT-OMT

Drive Cone
/
I /
Geostick
,
/
Hand Penetrometer

Heel Print

I L""
Relative Test Accuracy

I Figure F-1. Qualitative Relationship Between Relative Test Cost and Accuracy

Source: Handy (1), p. 242.

I
I
REFERENCES

1. Mitchell, J. K., "In-Situ Techniques for Site Characterization", Proceedings,

NSF Specialty Workshop on Site Characterization and Exploration, Ed. C. H.

I 2.
Dowding, Northwestern University, Evanston, June 1978, pp. 107-129.

Handy, R. L., "Realism in Site Exploration: Past, Present, Future, and Then
Some -All Inclusive", Proceedings, Symposium on Site Exploration in Soft

I Ground Using In-Situ Techniques, Report FHWA-TS-80-202, Federal Highway Admin-

istration, Washington, 1980, pp. 239-248.

3. Orchant, G. J., Kulhawy, F. H.,· and Trautmann, G. H., "Reliability-Based Foun-

I dation Design for Transmission Line Structures: Critical Evaluation of In-Situ
Test Methods", Report EL-5507, Vol. 2, Electric Power Research Institute, Palo
Alto, 1988, 214 p.

I
I
I F-6

I
 I
I
Appendix G
I
CRITICAL STATE SOIL MECHANICS (CSSM) CONCEPT
I
The concept of critical state soil mechanics (CSSM) evolved at the University of
Cambridge (e.g., !) and has been instrumental in improving our understanding of
I
soil behavior (e.g.,~. l). Basically (and very simplistically), the concept
states that there is uniqueness of soil behavior at the critical state in void
ratio (e) - shear stress (q) - effective mean normal stress (p) space. The details
I
are well beyond the scope of this manual; however, the general soil behavior is
illustrated in Figure G-1.
I
For the stiff soil, a peak strength is achieved which is followed by strain soften-
ing to a state of constant volume (i.e., constant void ratio) deformation. For the
I
soft soil, a peak strength is achieved at the state of constant volume deforma-
tion. This state is known as the critical state and represents the limit strength
I
of soil. Different critical states exist for different confining stresses (or,
more precisely, effective mean normal stresses) to define a unique e-p-q envelope
in void ratio-stress space.
I
With this concept, a number of theoretical/experimental soil models were developed, I
known as Cam clay, Granta gravel, and modified Cam clay (e.g.,!). From these
models, a general predictive tool for soil behavior emerged. Strictly speaking,
this tool is applicable only to remolded, insensitive soils without aging,
I
I
I
....
0
Q)
.r;
I
(f)
For constant
confining stress I
Axial Strain

Figure G-1. Typical Soil Stress-Strain Behavior

I
G-l I
I
I
I cementing, and other environmental influences. However, research has shown that
the modified Cam clay model predicts well the behavior of normally consolidated,

I insensitive soils, also without aging, cementing, and other environmental influ-
ences. In other soils, the model effectively provides a lower bound.

I The following is some of the notation used with the modified Cam clay model:

6 sin ~tc
I M =
3 - sin ¢> tc
(G-1)

I 1\ ="- A
- -
K, (typically ~ 0.8) (G-2)

I Pmax!Po (G-3)

I r = spacing ratio, defined in Figure G-2 (typically r = 2 for modified

Cam clay)

I in which tf>tc = effective stress friction angle in triaxial compression, A = iso-

I tropic compression index, K-

Cs = swelling (or unload-reload) index,

= isotropic swelling index, Cc = compression index,
p = effective mean normal stress
Ca1 + o2 + o3)/3, o1, o2, and o3 =effective maximum, intermediate, and minor prin-

I cipal stresses, Pmax =maximum p to which soil has been subjected, Po= current p,
and OCRi = isotropic overconsolidation ratio.

I In its most basic form, CSSM assumes that all stress paths terminate on a line
(termed the critical state line or CSL) which is parallel to the virgin compression

I
\
I Critical \._).
state ~ \.
line \.

I e

I
In p
I Figure G-2. CSSM Notation

I G-2

I
 I
line. It is able to account for undrained and drained behavior and normally con- I
solidated as well as overconsolidated states of stress (See Figure G-2.) The
advantage of the Cam clay models is their simplicity and ability to relate effec-
tive stress analysis with total stress analysis. In its most basic form, only
I
three soil parameters (~tc• Cc, Cs) are required to represent a variety of common
stress paths and boundary conditions. I
REFERENCES

1. Schofield, A. N. and Wroth, C. P., Critical State Soil Mechanics, McGraw-Hill

I
2.
Book Company, London, 1968, 310 p.

Wroth, C. P., "Interpretation of In-Situ Soil Tests", Geotechnique, Vol. 34,

I
No. 4, Dec. 1984, pp. 449-489.

3. Wroth, C. P. and Houlsby, G. T., "Soil Mechanics - Property Characterization

and Analysis Procedures", Proceedings, 11th International Conference on Soil
I
Mechanics and Foundation Engineering, Vol. 1, San Francisco, 1985, pp. l-55.

I
I
I
I
I
I
I
I
I
I
G-3 I
I
I
I
I Appendix H

CPT CALIBRATION CHAMBER DATA FOR SANDS

I
I It is very difficult to obtain undisturbed samples of clean sands for laboratory
testing. New methods of sampling by freezing techniques are available now, but
they are difficult to use and expensive. Therefore, most CPT correlations for

I sands have been developed from data obtained in laboratory calibration chambers
which allow control of the sand uniformity, density, initial stress state, and

I stress history. Triaxial compression tests on identically prepared samples allow

determination of the friction angle and modulus for comparison. In this appendix,
the calibration chamber data used within this manual are described for reference.

I DATA SUMMARY

I In this manual, CPT correlations with relative density (Dr), effective stress fric-
tion angle (¢), constrained modulus (M), and in-situ or at-rest horizontal soil
stress coefficient (K 0 ) have been developed from 24 different sets of calibration
I chamber data on sands.
Table H-1.
A listing of these sands and their properties is given in
The symbol column refers to that used on the correlation plots.

I All of the calibration chamber tests were conducted on reconstituted sands which
were unaged. The majority were clean quartz sands. The percent fines (percent
I less than No. 200 sieve) ranged from 0 to 6 percent, although most of the sands had
less than 1 percent fines. The particle size at 50 percent finer (Dso) ranged from

I 0.16 to 1.0 mm, with an average of 0.38 mm. The particle size at 10 percent finer
(Dlo) ranged from 0.10 to 0.70 mm, with an average of 0.25 mm. All of the sands
were uniformly graded, with a range of uniformity coefficient (Cu) from 1.10 to
I 2.60 and an average of 1.79.
an average of 2.68.
The specific gravity ranged from 2.65 to 3.02, with
The maximum void ratio ranged from 0.73 to 1.05, while the

I minimum void ratio ranged from 0.40 to 0.65.

For testing, the sands were prepared over a range of relative density (Dr) and
I overconsolidation ratio (OCR). Dr varied from 8 to 100 percent, while the OCR
ranged from 1 (normally consolidated) to about 14 (heavily overconsolidated). In

I general, the sands were consolidated under K0 conditions prior to testing.

I H-1

I
 I
Table

CALIBRATION CHAMBER
I
I
No. Symbol' Sand (Reference)

l 4> Earlston (!) 0.33 0.16 2.60 2.65 0.727 0.404

I
2 <l ~ Edgar <!, ~) 0.45 0.29 1.79 NA 0.919 0.543

4
3 $ Erksak (~)

Frankston (~)
0.35

0.31
0.18

0.18
2.20

2.05
2.65

NA
0.963

0.792
0.525

0.462
I
5 EB Hilton Mines (2) 0.20 0.15 2.00 3.02 1.050 0.620
I
oe
7
6

<>+
Hokksund

Hokksund
(~.

(~)
2) 0.44

0.39
0.27

0.21
2.20

2.20
2.70

2.70
0.906

0.878
0.539

0.535
I
8
9
~
8
Hostun Fine
Lanchester 25/52 (10)
(~) 0.35
0.40
0.18
0.30
2.22
1.40
NA
NA
1.000
0.818
0.650
0.563 I
10 ~ Leighton Buzzard (11) 0.37 0.26 1.50 NA 0.815 0.489
11
12
<D
121
Leighton Buzzard (12)
Lone Star 2 (13)
0.85
1.00
0.70
0.60
1.30
2.00
NA
2.66
0.790
0.766
0.490
0.482 I
13 Lone Star 30 (13) 0.39 0.22 1.86 2.66 0.824 0.537

14 Lone Star 60 (13) 0.30 0.18 1.48 2.66 0.908 0.566

I
15 Monterey 0 (14) 0.37 0.25 1.60 2.65 0.820 0.540
I
16 m Monterey 0/30 (15) 0.45 0.35 1.37 2.65 0.803 0.563

17
18
Oostershelde (16)
Ottawa (17)
0.17
0.28
0.10
0.26
1.80
1.10
NA
NA
0.887
0.868
0.562
0.545
I
19
20
e
®
Ottawa 90 <2)
Reid-Bedford (2)
0.22
0.24
0.13
0.15
1. 85
1. 70
2.65
2.66
0.789
0.871
0.486
0.549
I
21
22
S. Oakleigh Fine (!)
S. Oakleigh Medium (!)
0.17
0.32
0.12
0.17
1.60
2.20
2.65
NA
0.932
0.754
0.570
0.412
I
23 Ticino (~) 0.50 0.41 1. 58 2.67 0.915 0.568

I
24 Toyoura (18) 0.16 0.13 1.46 2.64 0.977 0.605

Symbols: Dso - particle size at 50% finer

I
D10 - particle size at 10% finer
Cu
Gs
-
-
uniformity coefficient
specific gravity of solids I
H-2
I
I
I
I H-1

DATABASE FOR SANDS

I Chamber Diam- Cone Diam- Dr in Tests OCR in

Angularity Mineralogy eter (mm) eter (mm) (%) Tests

I sub rounded quartz 760 50.0 20,45,65,73 1

subangular quartz 1220 35.7 56,95 1 to 10

I sub rounded

subangular
quartz, 6% fines,
trace chert
quartz
1400

1200
35.7

35.7
69 to 99

54 to 100
1

1 to 7.7
to rounded

I angular
(S- 0.72,
R - 0.23)
feldspar, quartz,
mica, muscovite,
iron, 3% fines
1220 35.7 30 to 84 1

I angular 45% feldspar, 35% 762,1220 25.2,35.7 8 to 100 1,8

quartz, 10% mica
subangular 35% quartz, 1200 20,25.4, 31,82,96 1,7.3,
to angular 10% mica 35.7 14.5

I subangular
subangular
sub rounded
quartz
95% quartz
quartz
180
254
1200
11.3
9.5
35.7
15 to 95
0 to 100
40 to 97
1
1
1

I subrounded
sub rounded
to subangular
quartz
quartz with
feldspar
900
760
35.7
35.7
20 to 90
22 to 66
1
1

sub rounded quartz with 760 35.7 20 to 84 1

I to subangular
sub rounded
to subangular
feldspar
quartz with
feldspar
760 35.7 17 to 79 1,1. 5.
3.6,5.9

I
subrounded quartz, trace 760 35.7 27 to 72 1
(S - 0.80, feldspar
R - 0.35)
subrounded quartz with 1500 23.2,35.7 24,64 1

I
to subangular feldspar
rounded quartz 1900 35.7 30 to 87 1
well- quartz 71.1 12.7 57 1,2,4
rounded

I rounded
subangular
(S- 0.76,
quartz, 0.2% fines
quartz,
some feldspar,
1220
1220
35.7
35.7
20 to 83
24 to 81
1
1

R- 0.29) trace calcite

I subangular
subangular
quartz
quartz
760
760
35.7,50.0
35.7,50.0
28 to 86
44 to 89
1
1,2,4,8
subangular 30% quartz, 1200 20,25.4, 16 to 98 1 to 14.7

I to angular
(S- 0.79,
R - 0.38)
5% mica 35.7

subangular high feldspar 790 35.7 33 to 86 1

I content

emax - maximum void ratio

emin minimum void ratio

I s
R
particle sphericity- (6 x pacticle volume/~)l/3;particle length
- particle roundness (See Figure 2-2.)

I H-3

I
 I
Most of the sands were placed in a dry state by air-pluviation (raining).
tions include Erksak sand (l), which was tamped moist, and Lanchester sand (10),
Excep-
I
which was prepared by raining, tamping, and vibrating methods.
of the CPT tests were performed on dry sand.
In addition, most
The exceptions include Erksak sand I
(l), which was saturated using back pressure, and Monterey 0 sand (14) and Ticino
sand(~), which were both dry and saturated. Furthermore, Jamiolkowski, et al.
(19) state that the Edger, Ottawa, Reid-Bedford, and Hilton Mines sands were tested
I
both "drained" and "submerged".
I
All tests used electric cones with a 60° cone angle. The cone diameters ranged
from 9.5 to 50.0 mm, although 85 percent of the data were obtained with the stan-
dard 35.7 mm diameter cone. All of the cones were of the standard Fugro cylindri-
I
cal shape, except for that of Villet and Mitchell (13), which had a reduced diame-
ter behind the cone. I
CHAMBER BOUNDARY INFLUENCE
I
Most of the available data were obtained using flexible-wall calibration chambers,
which allow yielding during cone penetration. This yielding gives measured cone
tip resistance (qc) values which are less than they would be in an infinite medium,
I
and therefore the qc values need to be corrected for these boundary effects. No
generally accepted approach has been developed yet for making these corrections. I
However, research has shown (e.g., 20) that qc increases with increasing ratio of
chamber to cone diameter (BcfB). In addition, the increase is more pronounced as
the relative density increases (e.g., l). The correction factor used herein was
I
derived from six available data sets from Table H-1 where the BcfB ratio was varied
to allow evaluation of the boundary effects. These data are summarized in Figure I
H-1. Based upon examination of these data and the trends noted above, the follow-
ing correction factor was developed:
I
-0.005 D
qc (corrected) qc (measured) [ (BcfB - 1)/70] r (H-1)
I
in which Dr= relative density in percent. This equation assumes that there are no
boundary effects when BcfB equals or exceeds 70. A plot of this equation is given
in Figure H-2, which shows increasing corrections needed for smaller BcfB ratios
I
and higher relative densities.
I
Four different types of boundary conditions may be applied in flexible-wall cali-
bration chambers (e.g., 21), as shown in Table H-2. Most of the tests summarized I
H-4 I
I
I
I Hokksund Hokksund
(Parkin, et al., 1980) (Parkin, et ol.,

I 200
all OCR=I
200
1980)
92
92.5

- Dr(%)= 92.5~ Dr(%)= 8 1 /

I 100
g5~
100
81 ~
62..............-
58.5
8-
oiiOCR=8

60.5 (a) 32
(b)

I 1
40
·1--"36.5
o~--~--~----~--~--~--~~--J
0 20 40
I I
60

I c.
0

........
- - South Oakleigh Medium
(Veismonis, 1974)
- - South Ookleigh Fine
(Veismonis, 1974) /
0
- Hokksund - Monterey # 0130 .

I
cr 300 300
(Baldi, et al., (Sweeney, 1987)
Q)
u
c
1986) ~II Dr(%) = 61.5
all OCR= I
0
-:n 200 Dr<%>= 9 6 . 3 V

I (/)
Q)
0::
f~~::
/
/
95.6
"
all OCR=I 75 --~all
64~
24-
c. I 00 63~,-- OCR=I

I 1-
Q)
c
0
69" _.
48' .....
( c)
53 ./"'
36/
24-----
(d)
u 0~--~--~----~--~---L--~--~

I 0 20

300.---.-I--.I---,I.---.---.-I--.I---.
40 60

300 . . I I '0CR=II4.6
~81
I 200r-
Ticino
(Baldi, et al., 1986)

Dr(%) = 93.8 =----------

_ -
011 OCR -I
WO
T1crno
-(Baldi, et al.,l986~6.7
Dr(%)= 94.7~35
~A -
95.5

I I OOr-
77.4----
- I 001-
1- 950

............... 2.8_
-

I - 56.5 - - - - - - - ( e )- 1-
50A
(f )
_

OL---L---L---~~--~~--~~--~~--~ OL---~~--~~--~~~--L-'--~'--~'--~
0 20 40 60 0 20 40 60

I Chamber Diameter I Cone Diameter

Figure H-1. Calibration Chamber Data for Various Sands

I
I in Table H-1 used Type A or C conditions, which more closely simulate field condi-
tions. The proposed correction factor applies to these cases.

I Only two of the sands tested used either Type B or D boundary conditions. The

I H-5

I
 I
1.0

0.9
I
u Dr(%)

I
C"

-....
"0
Q) 0.8
u
,_
Q)

u
0

.......
u
0.7
I
C" 0.6
"0
....
Q)
:;)
(/)
0
0.5
I
Q)
:E
0.4
I
0.3
0 20 40
Chamber Diameter /Cone Diameter
60
I
Figure H-2. CPT Calibration Chamber Correction Factor
I
Table H-2 I
BOUNDARY CONDITIONS IN FLEXIBLE-WALL CALIBRATION CHAMBERS
I
1)rpe Vertical Horizontal I
A

B
Stress constant

Change in strain is zero

Stress constant

Change in strain is zero

I
c
D
Stress constant

Change in strain is zero

Change in strain is zero

Stress constant
I
I
first was Toyoura sand (18), where Type B conditions were imposed.
the three Lone Star sands (13), where the chamber used was of a different design
The second was I
than most and Type D conditions were imposed. For these sands, no correction
factor was introduced because the data are insufficient to develop this factor. I
I
H-6 I
I
I
I REFERENCES

1. Veismanis, A., "Laboratory Investigation of Electrical Friction Cone Penetro-

meters in Sand", Proceedings, European Symposium on Penetration Testing, Vol.

I 2.
2.2, Stockholm, 1974, pp. 407-419.

Holden, J. C., "The Determination of Deformation and Shear Strength Parameters

for Sands Using the Electrical Friction Cone Penetrometer", Report 110, Norwe-
I 3.
gian Geotechnical Institute, Oslo, 1976, pp. 55-60.

Been, K., Lingnau, B. E., Crooks, J. H. A., and Leach, B., "Cone Penetration

I Test Calibration for Erksak Sand", Canadian Geotechnical Journal, Vol. 24, No.
4, Nov. 1987, pp. 601-610.

4. Chapman, G. A. and Donald, I. B., "Interpretation of Static Penetration Tests

I in Sand", Proceedings, lOth International Conference on Soil Mechanics and

Foundation Engineering, Vol. 2, Stockholm, 1981, pp. 455-458.

5. Schmertmann, J. H., "Study of Feasibility of Using Wissa-Type Piezometer Probe

I to Identify Liquefaction Potential of Saturated Fine Sands", Report S-78-2,

U. S. Army Engineer Waterways Experiment Station, Vicksburg, 1978, 60 p.

6. Parkin, A. K., Holden, J., Aamot, K., Last, N., and Lunne, T., "Laboratory In-
I vestigation of CPT's in Sand", Report 52108-9, Norwegian Geotechnical Insti-
tute, Oslo, 1980, 45 p.

I 7. Lunne, T. and Christoffersen, H. P., "Interpretation of Cone Penetrometer Data

for Offshore Sands", Report 156, Norwegian Geotechnical Institute, Oslo, 1985,
pp. 1-11.

I 8. Baldi, G., Bellotti, R., Ghionna, V., Jamiolkowski, M., and Pasqualini, E.,
"Interpretation of CPT's and CPTU's: Drained Penetration of Sands", Proceed-
ings, 4th International Geotechnical Seminar: Field Instrumentation and In-
Situ Measurements, Nanyang Technological Institute, Singapore, 1986, pp. 143-

I 9.
156.

Canou, J., El Hachem, M., Kattan, A., and Juran, I., "Mini Piezocone Investi-
gation Related to Sand Liquefaction Analysis", Proceedings, 1st International
I 10.
Symposium on Penetration Testing (ISOPT-1), Vol. 2, Orlando, 1988, pp. 699-706.

Thomas, D., "Deep Sounding Test Results and the Settlement of Spread Footings

I on Normally Consolidated Sands", Geotechnigue, Vol. 20, No. 4, Dec. 1968, pp.
472-488.

11. Chong, M. K., "Density Changes of Sand on Cone Penetration Resistances", Pro-

I ceedings, 1st International Symposium on Penetration Testing (ISOPT-1), Vol.

2, Orlando, 1988, pp. 707-714.

12. Houlsby, G. T. and Hitchman, R., "Calibration Chamber Tests of a Cone Penetro-

I 13.
meter in Sand", Geotechnigue, Vol. 38, No. 1, Mar. 1988, pp. 39-44.

Villet, W. C. B. and Mitchell, J. K., "Cone Resistance, Relative Density, and

Friction Angle", Cone Penetration Testing and Experience, Ed. G. M. Norris and
I R. D. Holtz, ASCE, New York, 1981, pp. 178-207.

I
I H-7

I
 I
14. Huntsman, S. R., Mitchell, J. K., Klejbuk, L. W., and Shinde, S. B., "Lateral
Stress Measurement During Cone Penetration", Use of In-Situ Tests in Geotech-
I
nical Engineering (GSP 6), Ed. S. C. Clemence, ASCE, New York, 1986, pp. 617-

15.
634.

Sweeney, B. P., "Liquefaction Evaluation Using a Miniature Cone Penetrometer

I
and Large Scale Calibration Chamber", Ph.D. Thesis, Stanford University, Mar.

16.
1987' 281 p.

Greeuw, G., Smits, F. P., and van Driel, P., "Cone Penetration Tests in Dry
I
Oosterschelde Sand and the Relation with a Cavity Expansion Model", Proceed-
ings, 1st International Symposium on Penetration Testing (ISOPT-1), Vol. 2,
Orlando, 1988, pp. 771-776. I
17. Lambrechts, J. R. and Leonards, G. A., "Effects of Stress History on Deforma-
tion of Sand", Journal of the Geotechnical Engineering Division, ASCE, Vol.
104, No. GTll, Nov. 1978, pp. 1371-1388.
I
18. Iwasaki, K., Tanizawa, F., Zhou, S., and Taksuoka, F., "Cone Resistance and
Liquefaction Strength of Sand", Proceedings, 1st International Symposium on
Penetration Testing (ISOPT-1), Vol. 2, Orlando, 1988, pp. 785-791.
I
19. Jamiolkowski, M., Baldi, G., Bellotti, R., Ghionna, V., and Pasqualini, E.,
"Penetration Resistance and Liquefaction of Sands", Proceedings, llth Inter-
national Conference on Soil Mechanics and Foundation Engineering, Vol. 4, San
I
Francisco, 1985, pp. 1891-1896.

20. Parkin, A. K. and Lunne, T., "Boundary Effects in the Laboratory Calibration
of a Cone Penetrometer for Sand", Proceedings, 2nd European Symposium on Pene-
I
21.
tration Testing, Vol. 2, Amsterdam, 1982, pp. 761-768.

Bellotti, R., Bizzi, G., Ghionna, V., Jamiolkowski, M., Marchetti, S., and
I
Pasqualini, E., "Preliminary Calibration Tests of Electrical Cone and Flat
Dilatometer in Sand", Proceedings, 8th European Conference on Soil Mechanics
and Foundation Engineering, Vol. 2, Brighton, 1979, pp. 195-200. I
I
I
I
I
I
I
H-8 I
I
II
~
Ii
Appendix I

UNIT CONVERSIONS

~
I Parameter Measure Conversions

I length foot (ft)

inch (in)
0.3048 meters (m)
25.4 millimeters (rnrn)

I mass

force
pound (lb)

ton (t)
0.4526 kilograms (kg)

2000 pounds (lb)

2 kips (k)

I stress atmosphere (atm)

8.896 kiloNewtons (kN)

1.058 tons/square foot (tsf)

2.116 kips/square foot (ksf)

I 1.033 kilograms/square centimeter (ksc)

101.3 kiloNewtonsjsquare meter (kN/m2)
101.3 kiloPascals (kPa)
0.1013 MegaNewtonsjsquare meter (MN/m2)
I 14.70 pounds/square inch (psi)
l. 013 bars

I unit weight poundjcubic foot (pcf)

(actually pound-force)
0.157 kiloNewtonsjcubic meter (kN/m3)

density pound/cubic foot (pcf) 16.02 kilograms/cubic meter (kg/m3)

I Note:
(actually pound-mass)

l atm (pa) ~ l tsf ~ 2 ksf ~ l ksc ~ 100 kN/m2 ~ 100 kPa ~ 0.1 MN/m2

I ~ 14.7 psi~ l bar

unit weight of fresh water (~w) = 62.4 pcf = 9.80 kN/m3

I
I
I
I
I I-1

I
 I
I
Appendix J
I
SUMMARY CORRELATION TABLES
I
Within this manual, numerous correlations have been presented that allow the user
to estimate a desired soil property from the results of laboratory index tests or
I
in-situ field tests, or from other simple procedures. To assist the user in loca-
ting specific recommended correlations, Tables J-1 and J-2 have been prepared for I
cohesive and cohesionless soils, respectively. In each table, the broad property
category is noted in the first column, followed by the specific soil property to be
estimated in Column 2. Column 3 gives the laboratory or other test methods used to
I
develop the laboratory or theoretical correlations noted in Column 4. The remain-
ing columns identify the correlations available for the common in-situ field tests. I
These tables are not intended to be a substitute for the text, which puts the cor-
relations in proper perspective. Instead, they are intended to be a quick refer-
I
ence guide for the experienced user.
I
I
I
I
I
I
I
I
J-1 I
I
 I
I Table

CORRELATIONS FOR

I
I Property Lab/Field Lab/Theory
Category Soil Property Test Method Correlation
I Basic Simple description Atterberg limits, pp. 2-l to 2-7
Characterization gradation,

I (Sec. l, 2)
Soil classification
simple field tests
Atterberg limits, pp. 2-9,10
gradation,

I Unit weight, -y
Consistency
visual - manual
gradation Table 2-8

I In-Situ Stress
(Sec. 3)
Preconsolidation
stress, op
Atterberg limits Eq. 3-13
Fig. 3-6
Overconsolidation Atterberg limits, Eq. 3-17 to 22

I ratio, OCR
Coef. of horizontal
soil stress, K 0
triaxial shear
Atterberg limits,
oedometer, triaxial
Eq. 3-10
Fig. 3-22,25,26

I Strength
(Sec. 4)
Effective stress
friction angle, ~
Atterberg limits Fig. 4-20,24,25
Table 4-5

I Undrained shear
strength, su
Atterberg limits,
triaxial, etc.
Fig. 4-21,22,24,55
Fig. 4-26 to 29
Table 4-9

I Elastic Poisson's ratio, v vu- 0.5

Deformabili ty Young's modulus, E Atterberg limits, Eq. 5-16,19,20
(Sec. 5) field load tests, Fig. 5-6,7,9,10
resonant column

I Time-Dependent
Subgrade modulus, ks

Compression indices, Atterberg limits

Eq. 5-34

Eq. 6-7,9
Deformabil i ty Cc + Cur Fig. 6-4,5,6

I (Sec. 6) Constrained modulus,

M
oedometer Eq. 6-12
Fig. 6-7,8
Coef. of consol- Atterberg limits Fig. 6-15

I idation, cv
Coef. of secondary
compression, Ca
water content Fig. 6-18
Table 6-3

I Permeability
(Sec. 7)
Coef. of permea-
bility, k
constant head Fig. 7-3,4
Table 7-1

a - SPT - standard penetration test PMT - pressuremeter test

I CPT - cone penetration test
CPTU - piezocone test
DMT - dilatometer test
VST - vane shear test

I J-2

I
 I
J-1 I
COHESIVE SOILS
I
Field Test Correlationa
I
sPTb CPTb CPTU PMT DMT VST
(App. A, F) (App. B, F) (App. B, F) (App. C, F) (App. D, F) (App. E, F)
I
Fig. 2-8 Fig. 2-9,10 Fig. 2-12
I
I
Table 2-13,14 Table 2-14

Fig. 3-9 Fig. 3-10 Fig. 3-11,

12,13
Fig. 3-14,15 Fig. 3-16 Eq. 3-15
Fig. 3-7
I
Fig. 3-18

Fig. 3-30 Fig. 3-31

Fig. 3-19

Fig. 3-32 Fig. 3-27

Eq. 3-25
Fig. 3-21
Eq. 3-30
Fig. 3-17
I
(direct Fig. 3-28,29
measurement)
I
Fig. 4-50
Table 4-10
Eq. 4-61,62
Fig. 4-53
Eq. 4-63
Fig. 4-55
Eq. 4-64 Eq. 4-65 Eq. 4-57,58
Fig. 4-49 I
Fig. 5-8
I
I
Fig. 6-9 Fig. 6-10 Fig. 6-11 I
Fig. 6-16 Fig. 6-17
I
I
b - See interrelationship of CPT qc and
SPT N values in Figures 2-29 to 2-32. I
J-3 I
I
I
I Table

CORRELATIONS FOR

I
I Property Lab/Field Lab/Theory
Category Soil Property Test Method Correlation
I Basic Simple description gradation, pp . 2 -1 to 2- 7
Characterization simple field tests
I (Sec. l, 2)
Soil classification gradation,
visual - manual
p. 2-9,10

Unit weight, -y gradation Table 2-8

I In-Situ Stress
Relative density, Dr

Coef. of horizontal
relative density

triaxial

I (Sec. 3)

Strength
(Sec. 4)
soil stress, K0

Effective stress
friction angle, ~
triaxial, etc. Eq. 4-5,6,8
Table 4-2

I Deformabili ty
(Sec. 5, 6)
Poisson's ratio, v
Young's modulus, E
triaxial, etc.
triaxial, etc. ,
Eq. 5-7,9
Eq. 5-21,23,30
field load tests, Fig. 5-18,19

I Constrained modulus, M
resonant column
oedometer Fig. 6-14
Compression index, Cc gradation, Fig. 6-12,13

I Subgrade modulus, ks
oedometer
Eq. 5-34

Permeability Coef. of permea- falling head Fig. 7-1,2

I (Sec. 7)

Liquefaction
bility, k

Cyclic stress cyclic triaxial,

Table 7-1

Resistance ratio, rav/Ov 0 etc.

I (Sec. 8)

a - SPT - standard penetration test PMT - pressuremeter test

CPT - cone penetration test DMT - dilatometer test

I CPTU - piezocone test VST - vane shear test

I
I
I
I J-4

I
 I
J-2 I
COHESIONLESS SOILS
I
Field Test Correlationa
I
SPTb CPTb CPTU PMT DMT
(App. A, F) (App. B, F) (App. B, F) (App. C, F) (App. D, F)
I
Fig. 2-8 Fig. 2-9,10 Fig. 2-12 I
Eq. 2-17 Eq. 2-21 Eq. 2-21 Fig. 2-28 I
Fig. 3-35,37,39 direct Fig. 3-34

Fig. 4-13 Fig. 4-15,16,17

measurement

Fig. 4-19 Eq. 4-17

I
Eq. 5-26 Fig. 5-14 Eq. 5-27
I
Fig. 5-13

Fig. 5-16
I
I
Fig. 8-1 Fig. 8-2,3 Fig. 8-4
I
b - See interrelationship of CPT qc and
I
SPT N values in Figures 2-29 to 2-32.
I
I
I
I
J-5 I
I