Louisiana State University LSU Digital Commons
LSU Historical Dissertations and Theses Graduate School
1990 In Situ and Laboratory Evaluation of Liquefaction Resistance of a Fine Sand. Behnam Mahmoodzadegan Louisiana State University and Agricultural & Mechanical College
Follow this and additional works at: https://digitalcommons.lsu.edu/gradschool_disstheses
Recommended Citation Mahmoodzadegan, Behnam, "In Situ and Laboratory Evaluation of Liquefaction Resistance of a Fine Sand." (1990). LSU Historical Dissertations and Theses. 5075. https://digitalcommons.lsu.edu/gradschool_disstheses/5075
This Dissertation is brought to you for free and open access by the Graduate School at LSU Digital Commons. It has been accepted for inclusion in LSU Historical Dissertations and Theses by an authorized administrator of LSU Digital Commons. For more information, please contact [email protected]. INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.
The quality of this reproduction isdependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand corner and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.
Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order.
University Microfilms International A Bell & Howell Information Company 3 0 0 North Z e e b Road, Ann Arbor, Ml 4 8 1 0 6 -1 3 4 6 USA 313/761-4700 800 521-0600
Order Number 9123218
In situ and laboratory evaluation of liquefaction resistance of a fine sand
Mahmoodzadegan, Behnam, Ph.D.
The Louisiana State University and Agricultural and Mechanical Col., 1990
UMI 300 N. Zeeb Rd. Ann Arbor, MI 48106
NOTE TO USERS
THE ORIGINAL DOCUMENT RECEIVED BY U.M.I. CONTAINED PAGES WITH SLANTED AND POOR PRINT. PAGES WERE FILMED AS RECEIVED.
THIS REPRODUCTION IS THE BEST AVAILABLE COPY.
IN-SITU AND LABORATORY EVALUATION OF LIQUEFACTION RESISTANCE OF A FINE SAND
A Dissertation
Submitted to the Graduate Faculty of the Louisiana State University and Agricultural and Mechanical College in partial fulfillment of the requirements for the degree of Doctor of Philosophy
in
The Department of Civil Engineering
by
Behnam Mahmoodzadegan B.S., State University of New York, 1983 M.S., State University of New York, 1984 December, 1990 To My Wife, Fariba
ii ACKNOWLEDGEMENTS
This research was carried out under the guidance of
Professor Ilan Juran, whose constant encouragement and technical assistance were of great value during the course of this study.
I would also like to thank Professors M. T. Tumay, A.
Arman, and J. N. Suhayda from the Civil Engineering
Department, Professor P. E. Connor from the Department of
Mathematics, and Professor S. Iyengar from the Computer
Science Department for serving on my dissertation committee.
In particular, Professor M. T . Tumay is thanked for his useful comments regarding the in-situ test results, and for his careful review of the draft.
My deep appreciation goes to Professor A. Arman, the
Director of the Louisiana Transportation Research Center
(LTRC) at the time this research started, for authorizing the use of the LTRC facilities for installation and maintenace of the triaxial loading system and its auxiliary units.
The research grant from the French Ministry of Research al1ocated for the in-situ phase of this study is great 1y appreciated. Specially, Professor F. Schlosser is thanked for initiating the joint research program between LSU and the
Ministry of Research.
A special thank goes to Mr. Bil 1 Tierney for his assistance in performing the in-situ tests. I wish to also extend my appreciation to my parents for their continued moral support, and specially to my wife without whose constant patience and encouragement this work would have never been completed.
iv TABLE OF CONTENTS
Page
ACKNOWLEGEMENTS ...... iii
LIST OF TABLES ...... vii
LIST FIGURES ...... viii
LIST OF SYMBOLS ...... xi
ABSTRACT ...... xii
Chapter
1...... INTRODUCTION ...... 1
2 LITERATURE REVIEW ...... 5
2.1 Basic Definitions of Soil Liquefaction ...... 6
2.2 Laboratory determination of liquefaction Potential ...... 7
2.3 In-situ Evaluation of Liquefaction Potential ...... 14
2.4 Use of Piezocone Penetration Tests ...... 18
3 IN-SITU EVALUATION OF THE LIQUEFACTION POTENTIAL OF SAND DEPOSITS: A COMPARATIVE STUDY ...... 21
3.1 Available CPT-based Liquefaction Evaluation Procedures ...... 22
3.2 Comparative Evaluation of the Available Procedures ...... 31
3.3 Potential Use of the Piezocone Data .... 32
4 STATIC AND CYCLIC TRIAXIAL LIQUEFACTION RESISTANCE OF THE HRD SAND ...... 59
4.1 Monotonic Triaxial Liquefaction Tests .. 60
4.2 Cyclic Triaxial Liquefaction Tests ...... 79
v 4.3 Relationship Between Monotonic and Cyclic Liquefaction ...... 92
5 DEVELOPMENT OF AN EQUIPMENT AND NEW TESTING PROCEDURES FOR EVALUATION OF STATIC AND CYCLIC LIQUEFACTION RESISTANCE OF SOILS 95
5.1 Hollow Cylinder Cell Test: An Overview.. 96
5.2 The Proposed Interpretation Procedure... 99
5.3 Verification of The Proposed Interpretation Procedure ...... 106
5.4 Evaluation of the Boundary Condition Effect ...... 120
5.5 Definition of the Cyclic Stress Ratio for Cyclic HCC Tests ...... 122
6 USE OF HOLLOW CYLINDER CELL EQUIPMENT FOR EVALUATION OF STATIC AND CYCLIC LIQUEFACTION RESISTANCE OF HRD SAND ...... 128
6.1 Monotonic HCC Liquefaction Tests ...... 129
6.2 Cyclic HCC Liquefaction Tests ...... 146
6.3 Relationship Between Static and Cyclic Liquefaction ...... 155
7 SUMMARY AND CONCLUSIONS ...... 160
REFERENCES ...... 175
VITA ...... 186
vi LIST OP TABLES
Table Page
1 The In-situ Testing Program ...... » . . 33
2 Liquefaction Potential of the Sand Layers at the Heber Road and Wildlife Sites ...... 36
3 The Parameters Used for Numerical Simulation of the Cavity Expansion Tests ...... 110
vii LIST OF FIGURES
Figure Page
1 Liquefaction Evaluation Procedure Based on Shear Wave Velocity (Bierschwale and Stokoe, 1985) ...... 21
2 Liquefaction Evaluation Procedure Based on SPT Results (Seed and Idriss, 1981) .... 25
3 Liquefaction Evaluation Procedure Based on SPT-CPT Correlation (Seed, 1983) ...... 27
4 Liquefaction Evaluation Procedure Based on Relative Density (Seed and Idriss, 1981) 28
5 Liquefaction Evaluation Procedure Based on Tip Resistance (Robertson and Campanella, 1985) ...... 30
6 Liquefaction Evaluation Procedure Based on Direct Observation (Robertson and Campanella, 1985) ...... 32
7.a Soil Stratification at the Heber Road Site (Youd and Bennett, 1983) ...... 34
7.b Soil Stratification at the Wildlife Site (Youd and Bennett, 1983) ...... 35
8 Use of Dual Piezocone Test for Identifying Loose Sand Seams ...... 40
9 Effect of Relative Density on the Deviatoric Stress During Monotonic Triaxial Tests .... 68
10 Effect of Relative Density on the Excess Pore Water Pressure During Monotonic Triaxial Tests ...... 69
11 The Steady-state Line for the HRD Sand Obtained From Monotonic Triaxial Tests .... 72
12 Effect of Relative Density on the Effective Stress Paths Obtained From Monotonic Triaxial Tests ...... 74
13 Effect of Confining Pressure on the Deviatoric Stress During Monotonic Triaxial Tests .... 76
viii 14 Effect of Confining Pressure on the Excess Pore Water Pressure During Monotonic Triaxial Tests ...... 77
15 Effect of Confining Pressure on the Effective Stress Paths From Monotonic Triaxial Tests ...... 80
16 Effect of Relative Density on Cyclic Liquefaction Resistance of HRD Sand From Triaxial Tests ...... 86
17 The Excess Pore Pressure Response for Dt =80% 87
18 The Axial Strain Response for Dr =80% ...... 88
19 The Effective Stress Path for Dt =80% ...... 89
20 Effect of Confining Pressure on Cyclic Liquefaction Resistance of HRD Sand From Triaxial Tests ...... 91
21 Effective Stress Paths for Monotonic and Cyclic Triaxial Tests on HRD Sand ...... 93
22 Schematic of the Hollow Cylinder Cell (BenSaid, 1986) ...... 97
23 Constitutive Soil Model (Juran and Beech, 1986) ...... 102
24 (a) Typical Results of An Undrained HCC Test (b) Derivation of Model Parameters by the Hyperbolic Transformation ...... 107
25 Numerical Simulations of the Triaxial Tests on Fontainbleau Sand ...... 108
26 The Cavity Expansion Test Results (BenSaid, 1986) ...... 112
27 Comparison of the Experimental Response Curves for the Triaxial Tests with Numerical Simulations for the HCC Tests ...... 115
28 Numerical Simulations of the Stress Ratio and the Volumetric Strain for the CET Tests ... 116
29 Effect of the Applied Stress Path on Soil Properties ...... 118
ix 30 Numerical Simulation of Effective Stress Paths for the Cavity Expansion Tests ..... 119
31 Effect of the Boundary Conditions and the Analytical Procedure on the Soil Response in Cavity Expansion Test ...... 121
32 Numerical Simulations of Triaxial CU, CD, and CET Tests Using the Proposed Algorithm .... 125
33 Definition of CSR for Cyclic HCC Tests .... 126
34 Reproducibility of the Monotonic Liquefaction Tests in the HCC Apparatus ...... 130
35 Effect of Cavity Volume Change Rate on the Soi1 Response in Undrained CET Tests ..... 135
36 Effect of Relative Density on the Cavity Pressure During Monotonic HCC Tests ...... 137
37 Effect of Relative Density on the Excess Pore Pressure During Monotonic HCC Tests ...... 138
38 The Steady-state Line Derived From the Results of Undrained Monotonic HCC Tests .. 142
39 Effect of Confining Pressure on Cavity Pressure in Monotonic HCC Tests ...... 144
40 Effect of Confining Pressure on the Excess Pore Pressure in Monotonic HCC Tests ..... 145
41 Effect of Relative Density on the Excess Pore Pressures in Cyclic HCC Tests ...... 149
42 Effect of the Applied Stress Path on the Cyclic Liquefaction Potential of the HRD Sand: the Dc effect ...... 152
43 Effect of Confining Pressure on the Excess Pore Pressures in Cyclic HCC Tests ...... 154
44 Effect of the Applied Stress Path on the Cyclic Liquefaction Potential of HRD Sand: the o'c effect ...... 156
45 Comparison of Effective Stress Paths in Monotonic and Cyclic HCC Tests ...... 158
x LIST OF SYMBOLS
A Deformed cross-sectional area of the triaxial specimen
A 0 Initial specimen cross-sectional area
a x Maximum ground acceleration
Dr Relative density
g Acceleration due to gravity
Liquefaction potential p' Average effective stress
q Deviatoric stress r Internal cavity diameter
R External cavity diameter
Au Excess pore pressure
O ’l Major effective principal stress
O's Minor effective principal stress
O'c Consolidation pressure
*^c y c Cyclic shear stress
*P Peak soil friction angle
4>cv Constant volume or critical state friction angle
€ Axial strain
Y Shear strain
V Dilation angle
U Dilatancy coefficient
Plastic dilatancy rate
P Eulerian spatial coordinate
xi ABSTRACT
Liquefaction is a phenomenon during which saturated
cohesionless soils lose a major portion of their shear
strength. Liquefaction has been identified as the source of
several major failures. In this study, a comprehensive
program to investigate the in-situ and laboratory liquefaction
potential of sandy soils was carried out.
The in-situ testing program included friction cone,
seismic cone, piezo-cone, and dual piezo-cone penetration
tests. Several in-situ liquefaction procedures were reviewed.
Four procedures were selected to evaluate the liquefaction
potential of a number of sandy sublayers in two well-
documented sites in the southern Imperial Valley, California.
The laboratory testing program included monotonic and
cyclic liquefaction tests using two different testing
equipment: triaxial and hollow cylinder cell. A new hollow cylinder cell equipment was designed, fabricated and
calibrated. This new equipment allows performing monotonic or cyclic cavity expansion tests.
Effect of the main testing variables on the laboratory liquefaction potential of a fine sand, using the two equipment, were examined. The effect of the applied stress path on the monotonic and cyclic liquefaction potential of the soil was illustrated. Also, relationships between cyclic and monotonic liquefaction were considered, using both devices. Chapter 1
INTRODUCTION
Soil liquefaction has been identified as the source of several major catastrophic failures. Liquefaction has been conventionally defined as the collapse process in loose, saturated non-cohesive soils where due to the buildup of excess pore water pressure the effective stresses, and, therefore, the shear strength of the soil are drastically reduced. Liquefaction failures have been reported to occur due to seismic as well as nonseismic loading sources.
Unti1 the late 1970's laboratory evaluation of the liquefaction potential of granular soiIs were mainly based on cyclic triaxial or simple shear test results. Several investigators have studied the effect of different testing variables on the liquefaction susceptibility of laboratory prepared specimens. The testing variables often considered include: confining pressure, relative density, cyclic stress ratio, loading wave frequency and form, overconsolidation ratio, and existance of initial shear stress. Liquefaction susceptibility of granular soils is largely influenced by the shear induced volume changes which in turn depends primarily upon the type of the stress path to which the soil is subjectd. The effect of stress path on the liquefaction potential of granular soiIs has not been adequately investigated. In fact, the extent of the research conducted on this topic so far has been very limited. Furthermore, new liquefaction evaluation procedures have been proposed which are entirley based on the results of monotonic tests. However, possible relationships between mechanisms leading to
1iquefaction under monotonic and cyclic loading have been the subject of very limited investigation.
During the past three decades several techniques for the in-situ evaluation of the liquefaction potential of soils have been proposed. The in-situ techniques used for this purpose include standard and cone penetration tests, geophysical techniques such as cross-hole and seismic cone penetration tests, electrical resistivity methods, and dilatometer test.
Throughout the 1980's, cone penetration test (CPT) and piezocone penetration test (PCPT) have progressively emerged as reliable tools for soil stratification. Various CPT and
PCPT based soi1 stratification charts have been proposed by different investigators. Although various CPT based
1iquefaction evaluation procedures have been proposed, a comparative analysis of their predictive capabilities has not yet been reported in the literature. Furthermore, use of piezocone data for prediction of liquefaction potential of soils has so far been limited to qualitative analyses.
In order to address the above mentioned research needs, a comprehensive research program involving field testing, laboratory studies, and numerical modeling of soil behavior was designed. The major goals of this research program have been to: (a) determine the effect of stress path and the main testing variables (specifically, relative density and
consolidation pressure) on the static as well as cyclic
liquefaction resistance of a sandy soil, (b) investigate possible relationships between monotonic and cyclic
liquefaction resistance of this sand under two distinct stress paths, (c) conduct a comparative study of several well-known in-situ liquefaction evaluation procedures. As the nature of this research program required, other issues were also to be addressed. In particular, the following research tasks were undertaken:
1. Several friction cone, piezo-cone and seismic cone tests were performed in selected wel1-documented sites in Imperial
Valley, California. The results of these tests were analysed to establish comparisons among four well-known liquefaction evaluation procedures. Also, during the course of the field testing program detailed soil stratifications were recorded at the sites under consideration to identify a suitable sandy soil for the proposed laboratory investigations. A fine sand was selected and several hundred pounds of this material was transferred to LSU geotechnical engineering laboratory.
Results of laboratory identification and index tests on the selected Heber Road Dune sand are presented in Appendix A.
2. Monotonic and cyclic triaxial liquefaction tests were performed on laboratory reconstituted specimens of this sand under different initial states of consolidation pressure and relative density. To conduct these tests, a closed loop servo- hydraulic triaxial system with 22 kip axial load capacity manufactured by Material Testing Systems, Inc. was used. The system had to be fully calibrated and a data acquisition package for complete testing automation was developed.
3. In order to perform the monotonic and cyclic cavity expansion tests, a hollow cylinder cell apparatus was designed, fabricated, calibrated, and integrated into the main testing equipment (MTS).
4. A numerical algorithm for interpretation of monotonic cavity expansion test results was developed using an elasto- plastic constitutive soil model. This interpretation procedure was used to: (a) investigate the effect of geometric boundary conditions (i.e. scale effects) on the hollow cylinder cell test results, and (b) determine values of cyclic stress ratios for the hollow cylinder cell tests which would be compatible with those used for cyclic triaxial tests. Chapter 2
LITERATURE REVIEW
Soil liquefaction has been identified as the major cause
of catastrophic failures in a variety of geotechnical
engineering applications. Liquefaction has been conventionally
defined as the collapse process in loose non-cohesive soils
subjected to seismic loading as the generated excess pore water pressure attains the total confining stress (Seed and
Lee, 1966). However, liquefaction failures (sometimes cal1ed
"flow slides") are reported to have also occured due to other causes including low-level vibrations from non-seismic sources, construction operations, and rapid monotonic loadings such as tidal fluctuations and rapid drawdown in hydraulic fill dams (Edgers and Karlsrud, 1982).
2.1 Basic Definitions of Soil liquefaction
Ever since its first usage by Hazen (1920), to describe the failure of the Calaveras Dam in California in 1918, the term 1iquefaction has been used in reference to a number of static or cyclic col lapse mechanisms for loose granular soi1 deposits:
Initial Liquefaction. Seed and Lee (1966) used this term to define a stage at which the measured excess pore water pressure during cyclic triaxial tests attained the value of the initial confining stress (100% pore pressure ratio). The parameter which is commonly used to define initial
5 6 liquefaction is cyclic stress ratio (CSR) defined as the ratio of the applied cyclic deviatoric stress to the effective confining pressure. Several investigators (Lee and Seed, 1967;
Finn et al. , 1971; Mulilis, 1975; Castro and Poulos, 1976, and others) have studied the factors which determine the required
CSR to yield initial 1iquefaction.
Cyclic Mobility. Several investigators (Casagrande, 1975;
Castro and Poulos, 1976; Poulos et al., 1985, and others) have used this term to define a stage at which subsequent to occurence of initial 1iquefaction the soil would stabilize either due to dilation or due to its residual strength and a total 1oss of shear strength would not occur. Other investigators (Seed, 1979) have used the term "cyclic initial
1iquefaction with 1imited strain potential" or "partial
1iquefaction" to refer to the same phenomenon.
Static Liquefaction. Several investigators (Poulos, 1981;
Castro et al. , 1982; Poulos et al., 1985; Sladen et al., 1985) have proposed the concept of static liquefaction to describe the phenomenon of steady-state deformation at which a mass of soil would continuously deform under constant normal and shear stresses, constant volume, and constant rate of shear strain.
Castro et al. (1982), and Poulos et al. (1985) reported that, for sand specimens consolidated under different pressures, the piot of void ratio versus effective minor principal stress at failure obtained from undrained triaxial tests yields a straight line called the steady-state line (SSL). The location and orientation of the SSL is assumed to only depend on the in-situ void ratio. It defines the limiting state of steady-state deformation. Once the steady state is reached, the soi1 shear strength reduces to a residual or critical state value, which is assumed to be solely a function of the in-situ void ratio.
The laboratory procedures currently used to define the cyclic initial liquefaction and static liquefaction potential of soils are discussed in the subsequent section.
2.2 Laboratory Determination of Liquefaction Potential
At present, the laboratory procedures for determination of the liquefaction potential of soils can be broadly classified into two major categories:
Cyclic Liquefaction Tests. Cyclic 1iquefaction potential of a laboratory prepared specimen is commonly defined as the number of cycles required to induce initial liquefaction (100% pore pressure ratio condition) under an applied stress component (stress-control led tests) or an applied strain amplitude (strain-controlled tests). The tests most commonly used for this purpose are cyclic triaxial (Lee and Seed, 1967;
Finn et al. , 1970; Ishihara, 1985), cyclic torsional shear
(Ishibashi and Sherif, 197 4; Robinet et al., 1983; Alarcon et al. , 1986), and simple shear (Peacock and Seed, 1968; Finn et al., 1971; Amer et al., 1986).
The most significant soil parameters and testing variables affecting the cyclic liquefaction potential of granular soils are (Seed, 1976):
1. relative density,
2. confining pressure,
3. specimen preparation technique and sample disturbance,
4. soil type (grain size distribution, grain geometry,
coarseness, percentage of fines),
5. stress and strain history,
6. loading characteristics (frequency, duration, form),
7. consolidation ratio (presence of initial shear stress),
8. period under sustained load.
A detailed discussion of the effect of each and every parameter in the above list has been reported elsewhere
(Townsend, 1978). In the following paragraphs, a summary of the effect of the more fundamental parameters on the cyclic liquefaction resistance of sandy soils are presented:
Specimen Preparation. Several authors have investigated the effect of specimen preparation method on the cyclic liquefaction resistance of soils (Mulilis, 1975; Ladd, 1976;
Canou, 1989; and others). The most comprehensive set of data on this topic was presented by Mulilis (1975). He reported results of cyclic triaxial liquefaction test performed on specimens of Monterey No. 0 sand prepared using different procedures. The preparation methods included various pluviation (raining), vibratory, tamping and roding techniques. The main finding of this study was that high frequency vibration and air-pluviation would yield the highest and lowest liquefaction resistance, respectively.
At present, a standard specimen preparation technique for
cyclic liquefaction tests does not exist. For the purpose of
this study, a specimen preparation technique recently proposed
by Canou (1989) was used. This technique, which was
successfully used to prepare very loose sand specimens (Dr
values as low as 10 to 20 %) will be discussed in a subsequent
section.
Confining Pressure. Several investigators (Townsend,
1978; Mulilis, 1975; Castro and Poulos, 1976; Seed, 1976, and
others) have shown that for a given relative density, the
cyclic liquefaction resistance of sandy soils increases with
increasing confining pressure.
Relative Density. Townsend (1978) has reported the
results of several cyclic liquefaction tests indicating that
for a given confining pressure, cyclic liquefaction resistance increases with increasing relative density.
Consolidation Ratio. Several investigators (Lee and Seed,
1966; Castro and Poulos, 197 6; Vaid and Chern, 1983; Mohamad and Dobry, 1986, and others) have studied the effect of consolidation ratio on the cyclic 1iquefaction resistance of sands. In a comprehensive study, Mohamad and Dobry (1986) showed that this effect depends on the soil dilatancy / contractancy tendency. More specifically, they showed that for dilative specimens, cyclic liquefaction resistance always increases with increasing initial shear stress, while for 10
contractive specimens, cyclic liquefaction resistance
could increase or decrease depending on the relative
magnitudes of the initial shear stress, cyclic deviator
stress, and undrained steady-state shear strength of the soil.
Applied Stress Path. Liquefaction potential of soils is
mostly influenced by the volume changes which occur during
in-situ loading. These volume changes are in turn control 1ed
by the type of stress path to which the in-situ soi 1 is
subjected. Therefore, a study of the stress path effect on the
(monotonic or cyclic) liquefaction resistance of the particular soil is essential in every field investigation
concerning liquefaction failures. In order to address this issue, several testing equipment have been used in practice, including cyclic triaxial test (Lee and Seed, 1967; Finn et al. , 1970; Ishihara, 1985), cyclic torsional shear test
(Ishibashi and Sherif, 1974; Robinet et al., 1983; Alarcon et al., 1986), and cyclic simple shear test (Peacock and Seed,
1968; Finn et al., 1971; Amer et al., 1986). Also, a limited number of strain-controlled cyclic cavity expansion test results have been reported by Schwab and Dormieux (1985).
The investigations cited above, however, do not address two critical issues: (1) comparison between the liqefaction resistance of the same soil, prepared to identical initial conditions, tested in different equipment, and (2) comparison between monotonic and cyclic liquefaction of the same soil, prepared to identical conditions, using the same equipment. 11
In this study, monotonic and cyclic triaxial tests as
well as monotonic and cyclic hollow cylinder cell (HCC) tests
were performed to address the research needs described above..
Static Liquefaction Tests. Recently, Poulos et al. (1985)
proposed a procedure for determination of the liquefaction
potential of sandy soils from monotonic triaxial undrained
tests. The method which is based on the concept of the
steady-state line (Castro, 1969; Poulos, 1981) involves five
steps:
1. determine in-situ void ratio for the given soil,
2. determine steady-state void ratio, or density, as a
function of effective stress using compacted specimens,
3. determine undrained steady-state strength for
undisturbed specimens,
4. correct measured undrained steady-state strengths to
in-situ void ratio,
5. calculate in-situ driving shear stress and the
liquefaction potential, or safety factor.
In order for the proposed procedure to be of practical
relevance, two fundamental issues must be adequately adressed.
Firstly, as indicated by Poulos et al. (1985), both the
location and the inclination of the steady-state line depend
on, and are sensitive to, the in-situ void ratio of the given sandy soil. With the present sampling techniques for granular soils, a precise determination of the in-situ void ratio is very difficult, if not unfeasible. Therefore, the effect of 12
sampling procedures on the static liquefaction potential of
soils must be taken into account. Secondly, the proposed method has so far been used to determine the liquefaction potential of soil specimens subjected to only one type of
stress path (i.e. that of undrained triaxial compression test). Thus, in order for this method to be of practical
relevance, the fundamental concept of the steady-state of deformation and an appropriate testing procedure to establish
the steady-state 1ine need to be thoroughly investigated for other stress paths.
Sladen et al. (1985) have proposed an alternative procedure to evaluate the static liquefaction potential of sandy soils. Their method is based on a number of key observations made during stress- and strain-controlled undrained triaxial tests on sandy soils: (a) specimens prepared at the same void ratio (relative density) failed at the same point on the steady-state 1ine (SSL) regardless of their consolidation pressures, (b) the peak points of the observed effective stress paths (in the q-p' plane) for specimens prepared at the same void ratio fell on a straight
1ine passing through their common failure point on the SSL.
Moreover, the authors have suggested that due to the inter-dependence of the three test parameters (i.e. q, p', and e) in determining the liquefaction potential of a given specimen, a three-dimensional "collapse surface" can be established. The proposed surface is similar in concept to the 13
"State Boundary Surface" proposed by Schofield and Wroth
(1968) in the sense that specimens with initial states
approaching or on the collapse surface would be susceptible to
liquefaction. At present, data supporting the concept of a
collapse surface have been established only from undrained
triaxial compression tests. A rational use of this concept in
predicting the liquefaction potential of soils in-situ
requires its verification for other stess paths.
The effect of a number of testing variables on the static
liquefaction potential of sands defined by the steady-state
line have been reported in the 1iterature. Castro and Poulos
(1976), Kramer and Seed (1988), and Canou (1989) have
investigated the effectf of relative density, confining
pressure, and consolidation ratio on the static liquefaction
resistance of various type of sands. These studies have
concluded that the static liquefaction resistance of initial 1y
1 oose specimens increases with increasing relative density and
consolidation pressure and decreases with increasing
consolidation ratio.
Loading Mode. Castro et al . (1982) and Sladen et al .
(1985) have investigated the effect of stress- versus
strain-controlled loading procedure on the steady-state line.
They have reported that the loading procedure has practically no effect on the obtained steady-state 1ine.
Limitations of Laboratory Procedures. Laboratory tests in general present a number of limitations: 14
1. The laboratory prepared samples do not adequately
represent the in-situ soil with regard to:
a. the initial in-situ states of stress and density
b. depositional history (aging, cementation, etc.)
c. in-situ structure and fabric
d. stress history (overconsolidation),
2. The state of stress/strain induced within the specimen
during laboratory tests is not uniform,
3. Presently, it is difficult to precisely simulate in the
laboratory the actual in-situ stress/strain path.
Limitations of the existing laboratory testing procedures have stimulated the development and growing use of in-situ testing for the evaluation of liquefaction potential of soils.
A critical review of some common in-situ liquefaction evaluation techniques is presented in the following section.
2•3 In-situ Evaluation of Soil Liquefaction Potential
Several in-situ techniques have been used to determine the liquefaction susceptibility of soils. The essential features of each test are described below:
Standard Penetration Test (SPT). The most commonly used method for determination of the liquefaction potential of sandy soils is based on SPT results (Seed and Idriss, 1971;
Iwasake, 1982; Seed, et al., 1983; Nixon, 1982, and others).
The liquefaction susceptibility of any given soil layer in the field is determined through the use of empirically developed charts. The charts correlate the anticipated in-situ cyclic 15
stress ratio, determined from the semi-empirical formula proposed by Seed and Idriss (1971), to either the SPT blow
count number or to in-situ relative density.
The main advantage of the SPT lies in the availability of a large data base of results which is of significant practical value. However, SPT suffers from several disadvantages:
1. the in-situ soil is subjected to a great amount of
disturbance,
2. the induced stress path is complex resulting in
difficulties in analytical interpretation of test
results,
3. the testing procedure in practice is not standardized
raising questions with regard to reliability and
reproducibility of results,
4. the testing procedure does not allow for determination of
the main parameters controlling the liquefaction
resistance; that is the generated excess pore pressures
or soil volume changes,
5. the test does not yield a continuous soil profile and,
therefore, it is practically unfeasible to detect thin
sandy seams in which liquefaction is often initiated.
Cone Penetration Test. Friction cone penetration test
(CPT) offers two significant advantages over SPT: (a) standardized testing procedure (ASTM-D3441, 1979) yielding reproducible results, (b) ability to produce continuous soil profile and thereby to detect thin seams of sandy soils in 16 which liquefaction may be initiated. However, the use of CPT
for liquefaction evaluation presents major concerns with
regard to: (a) soil disturbance during penetration, (b) difficulties in analytical interpretation due to complex induced stress path, (c) inability to measure generated excess pore pressures or soil volume changes.
A number of empirical CPT-based liquefaction evaluation charts have been proposed:
CPT-SPT Correlation. Several investigators (Douglas et al. , 1981; Seed, 1983, and others) have proposed liquefaction evaluation procedures based on converting the CPT tip resistance (qc) values into equivalent values of SPT blow count (N). This permits direct utilization of the existing large data base of SPT results in various soils.
Cyclic Stress Ratio-Relative Density Correlation. Seed and Idriss (1981) have proposed an empirical procedure for liquefaction evaluation based on acorrelations between the in-situ determined relative density (using SPT or CPT data) and the average in-situ cyclic stress ratio.
Cyclic Stress Ratio-Tip Resistance Correlation. Robertson and Campanella (1985) have proposed a liquefaction evaluation chart based on an empirical correlation between the in-situ induced average cyclic stress ratio and the modified cone bearing (qc). This correlation was established by using the cyclic stress ratio-relative density (Dr) chart developed by
Christian and Swiger (1975) along with the Dr-qc chart 17
s ' proposed by Baldi et al. (1982). They have further proposed an upper bound for this correlation to account for the effect of
cementation and aging.
Similar charts had been proposed by Zhou (1980) and Seed et al. (1983).
Shear Wave Velocity Measurement. Shear wave velocity measurements have been used in two ways to develop liquefaction evaluation charts. A number of investigators
(Seed et al. , 1983; Robertson et al., 1986; NCEE, 1985) have proposed empirical correlations between the shear wave velocity and the SPT blowcount number (N) allowing the direct utilization of SPT-based liquefaction potential charts. Other investigators (Bierschwale and Stokoe, 1985; NCEE, 1985; Juran et al. , 1989) have reported the use of in-situ shear wave velocity and ground surface acceleration measurements, along with direct site observations to evaluate liquefaction potential of in-situ soils.
In-situ shear wave velocity can be determined by a variety of techniques including seismic cone test, and cross-hole, or down-hole method.
Flat-plate Dilatometer Test. The dilatometer equipment has been developed by Marchetti (1980), and related testing procedure has been described in details by Marchetti and
Crapps (1981).
Empirical procedures have been proposed by Marchetti
(1982) and by Robertson and Campanella (1984) for evaluation 18
of the liquefaction potential of sandy soils from DMT results.
These procedures involve correlations between the in-situ
average cyclic stress ratio required to cause liquefaction and
the horizontal stress index, Kd .
The main advantage of DMT is the well-defined stress path
(i.e. cylindrical cavity expansion) to which the soi1 is
subjected during the test. This allows development of
analytical interpretation procedures based on the available
theories of cavity expansion. However, use of dilatometer for
in-situ 1iquefaction evaluation raises questions with regard
to: (a) soil disturbance during the placement phase, and (b)
inability to measure generated excess pore pressures during
the test.
2.4 Use of Piezo-cone Penetration Test (PCPT). Piezo-cone
(Wissa et al. , 1975; Torstensson, 1975) is an in-situ testing
equipment which allows simultaneous measurement of the tip
resistance, the shaft friction and the excess pore water pressure generated during the penetration. Piezo-cone
penetration test (PCPT) presents the unique capability of providing a continuous, accurate soil stratification (Baligh
et al . , 1980; Tumay et al. , 1981; Robertson et al . , 1985;
Juran and Tumay, 1989).
A number of PCPT-based soil classification charts and
procedures have been offered using the tip resistance and the
excess pore pressure measured along the shaft, immediately
above the tip (Tumay et al. , 1981; Senesset et al. , 1982; 19
Robertson et al. , 1985).
The increasing use of piezo-cone has stimulated an extensive research on both practical and theoretical aspects of cone penetration. Investigations on the practical aspects of cone penetration have been focused on the effect of factors such as cone geometry (Levadoux and Baligh, 1980; Tumay et al. , 1981), penetration rate (Campanella and Robertson, 1981;
Canou et al . , 1988, Juran et al . , 1989), location of the piezo-cell (Wroth, 1984; Campanella et al. , 1984; Juran and
Tumay, 1989), and compressibility of the piezo-cell (Rad and
Tumay, 1985; Juran et al. , 1989) on the magnitude of the excess pore pressures generated during penetration.
Theoretical analyses of the cone penetration problem have been mainly focused on the strain path (Baligh, 1985) and the flow fields (Tumay et al. , 1985) around a penetrating cone, generation and dissipation of excess pore water pressures during and after penetration (Levadoux, 1980; Chan, 1982;
Kiousis et al . , 1988), and large strain finite element simulation of cone penetration mechanism (Karim, 1985;
Kiousis, et al., 1988).
A rigorous quantitative analysis of the soil-cone- sleeve interaction is complex. However, differences in pore water pressures measured at the center of the cone and along the shaft can be qualitatively related to the induced strain paths in the soil. In the vicinity of the cone center, the soil distortions are coupled with high compressive strains 20
resulting in large positive excess pore pressures. Along the
shaft, however, the soil is subjected to large distortional
strains which can induce either positive or negative excess
pore water pressure depending on the tendency of the soil to
contract or dilate, respectively.
Tumay et al. (1981), Campanella et al. (1984), and Juran
and Tumay (1989) measured generated excess pore water
pressures both on the cone and along the friction sleeve in
loose and dense sand seams. Tumay et al. (1981) and Juran and
Tumay (1989) used a dual piezo-cone which provides excess pore
pressure measurements at the cone and along the shaft, in
addition to measurements of the tip resistance and sleeve
friction. The results reported by Juran and Tumay (1989)
clearly indicated the advantage of the dual piezo-cone in
detecting very thin sand seams in which liquefaction is often
initiated.
Despite its clear advantages over SPT and CPT, use of
PCPT for in-situ liquefaction evaluation raises a number of
questions with regard to: (a) difficulties involved in
analytical interpretation of test results due to complex
stress path, (b) soil remolding and disturbance during the
test. Furthermore the data base of PCPT results is limited at
the present time. Development of rational PCPT-based
liquefaction evaluation procedures necessitates establishing much broader testing programs in various soil types and
conditions. Chapter 3
IN-SITU EVALUATION OF THE LIQUEFACTION POTENTIAL OF SAND DEPOSITS: A COMPARATIVE ANALYSIS
Cone penetration test (CPT) has been increasingly used during the past two decades for soil stratification.
Developments and engineering use of piezo-cone (PCPT) and seismic cone have enhanced the cone capabilities to identify soil strata and to provide data pertaining to the engineering properties of the in-situ soil. Specifically, the dual piezo cone (DPCFT) recently developed at L.S.U. (Tumay, 1985) yields measurements of the excess pore water pressures at the tip and along the friction sleeve, in addition to the point resistance and shaft friction, which can be related to the volume change properties of the soil, and therefore to its liquefaction potential. The seismic cone yields the in-situ shear wave velocity which can be empirically related to the shear modulus and to its liquefaction potential.
Several interpretation methods and classification charts have been established to use cone penetration data for the in- situ evaluation of soil liquefaction potential. In order to conduct a comparative analysis of the existing CPT/PCPT based liquefaction evaluation procedures, and to assess the capabilities of the DPCPT and the seismic cone to predict soil liquefaction potential , an in-situ testing program including seismic and piezo-cone penetration tests was established. The tests were conducted in two well documented sites in Imperial
21 22
Valley, California.
Analysis of the test results using the available
interpretation methods for assessment of soil liquefaction potential is presented in this chapter. This analysis demonstrates that although these methods rely on different empirical assumptions and involve correlations with different testing variables, they yield consistent results. Results of this study are also compared with previously published data on the same sites.
3.1 Available CPT-based Liquefaction Evaluation Procedures
Several procedures have been proposed for the in-situ evaluation of the liquefaction potential of sandy soils using seismic or friction cone penetration tests. They can be mainl y classified into two categories: (a) the seismic cone approach, and (b) friction cone based approaches.
The Seismic Cone Approach. Several investigators
(Robertson et al., 1986; Bierschwale and Stokoe, 1984; NCEE,
1985; Juran et al., 1989) have reported evaluation of liquefaction susceptibility of sand deposits based on seismic cone test results. In this approach (Fig. 1), the likelihood of liquefaction occurance, at a certain depth, is determined from the estimated values of the maximum ground acceleration and shear wave velocity.
Friction Cone Based Approaches. Several procedures for evaluation of liquefaction potential of sandy soils based on the friction cone penetration (CPT) test results have been iue 1.Figure LiquefactionEvaluationProcedure on ShearBased Shear Wave Velocity of Liquefiable Sand Layer (fps) 0 5 5 0 0 3 0 0 6 0 5 3 0 0 4 0 0 5 0 5 4 mx t ie oprbe o ceeorp Sain (g) Station Accelerograph to Comparable Site at Omax ol a te R site HR the at Soils 9 O LIQUEFACTION NO aeVlct (BierschwaleStokoe,Velocityand 1985)Wave ol a te L site WL the at Soils
iquefactio kely LIQUEFACTION
• • I 2 23 24
proposed. These procedures can be classified into four main
categories:
CPT-SPT Correlation. Until the late 1970's, standard
penetration test (SPT) was the only in-situ testing technique
for evaluation of the liquefaction potential of soils, and,
therefore, there exists a large data base of SPT results for
a variety of soil types and characteristics. In the SPT-based
approach (Fig. 2), the in-situ liquefaction potential is usual 1y determined based on an empirical correlation between
the blowcount (N) value and the average in-situ cyclic stress ratio. The average cyclic stress ratio is defined as the ratio of the expected induced cyclic shear stress (xaV(8), generated by an earthquake of given duration and magnitude, to the vertical effective stress at the specific depth (o'wo). The use of SPT for the in-situ liquefaction evaluation, however, raises basic questions with regard to: (1) non-standard testing procedure and equipment design, (2) interpretation of the soil response due to the impact type of loading during the test, (3) difficulties involved in identifying 1oose sand seams in which 1iquefaction is often initiated, (4) a discontinous soil profile obtained due to the testing procedure. Cone penetration test (CPT), on the other hand, involves a standardized testing procedure, yields continuous soil profiles, and presently offers the only feasible in-situ testing technique to detect loose liquefiable sand seams (Juran and Tumay, 1989). Figure 2.on LiquefactionFigureSPTChartBasedEvaluation Cyclic Stress Ratio, T/cr^ for cr^s I ton per sq. ft. 0.5 0.2 0.4 0.3 1 20 3 40 4 30 0 2 10 0 Lmt st y hns Cd (1974) Code Chinese by set Limits ® Liquefaction • N liquefaction No o oiid eerto Rssac, j blows/ft. Nj - Resistance, Penetration Modified eut (Seed Idriss,and 1981)Results
ae Amatitlan Lake o o Niigata
The main difficulty associated with the use of CPT test for evaluation of the liquefaction potential of soils lies in the limited data base of CPT results presently available.
Therefore, several investigators (Douglas et al., 1981; Seed,
1983; Robertson et al. , 1985; and others) have proposed liquefaction evaluation procedures based on the semi-empirical or statistical correlations which allow conversion of CPT tip resistance (qc) values to equivalent SPT N values. Figure 3 shows the application of one such liquefaction prediction chart, proposed by Seed (1983), to assess the liquefaction sucseptibility of three sites during an earthquake with a magnitude of 7.5.
Cyclic Stress Ratio-Relative Density Correlation. The response of the in-situ sandy soils to a particular stress or strain path (such as that inducing liquefaction) is mainly governed by the in-situ effective overburden pressure, o'vo, and the relative density, Dr . Therefore, a number of authors
(Schmertmann, 1978; Baldi et al., 1982; Robertson et al.,
1985; and others) have proposed CPT based correlations between the tip resistance and the relative density of the in-situ soil. Some of these methods use directly the results of the
CPT tests to determine the in-situ relative density, while others make use of CPT-SPT correlations. Once established, such correlation can be used to correlate the in-situ relative density and the earthquake induced average cyclic stress ratio which would most probably induce liquefaction (Fig. 4). Cyclic Stress Ratio T crv' for crv' » I kg/cm 0 0 3 0 O O I Figure3. Liquefaction Evaluation Procedure on Based
0 2 0 Bsd n cN 4 to 5 1 4 qc/N on IBosed oiid oe eerto Rssac, c kg/cm qc) Resistance, Penetrotion Cone Modified D50 > 2 m O m 5 SPT-CPTCorrelation (Seed, 1983) lo Sonds Cleon 0 5 LiQue LiQue
foction
kg/cm 100 150 150 o Liquefoction No - Based on Based - r P correlationSPT ae on Based otr Zhou -otter — ------S 2 M : y— J — Chinese 71/2
— ) 0 8 9 1 criteria 0 0 2
27 28
0.5
0.4
0.3 ave. vo
0.2
O 20 40 60 80 100
Dr (%)
Figure 4. Liquefaction Evaluation Procedure Based on Relative Density (Seed and Idriss, 198) 29
Cyclic Stress Ratio - Tip Resistance Correlation. Several authors (Seed et al. , 1983; Robertson and Campanella, 1985, and others) have proposed procedures for evaluation of liquefaction potential of sandy soils based on the correlation of cyclic stress ration (CSR) induced in the field during an earthquake and the CPT tip resistance (Qt). One such
1iquefaction evaluation chart was proposed by (Fig. 5)
Robertson and Campanel1 a (1983) . This chart was modified in a later work (Robertson and Campanel la, 1985) to take into account the effectf of aging and cementation on soil liquefaction resistance.
Field studies at sites affected by the 1978 earthquake in
Tangshan, China led to the development of an empirical procedure for evaluation of liquefaction potential of clean sands (Zhou, 1980). In this procedure, a critical tip resistance value, Qcrit' seperating liquefiable from nonliquefiable deposits of clean sands upto 15 m below the ground surface is determined as a function of the depth to the layer under consideration, the location of the ground water table, and the earthquake intensity. Seed (1983) has shown that the procedure proposed by Zhou (1980) would yield almost identical results to those predicted using SPT-CPT correlation charts.
Correlations Based on Direct Observation. Robertson and
Campanella (1985) have proposed a liquefaction evaluation chart based on observation of soil response during several 30
I bar = 100 kPa = 1.02 kg/cm‘
• • i • 4/2 . o 1 tr ej 4 ® Soils at the HR site
o H Soils at the WL site h- / / < / / cr 5 6 i B 8 / CO D50< 0.15 mm / CO UJ cr 1 / f- co / / ^ D 5 0 > 0 .2 5 mm u
Z i ' u i > u
J ' I 1 ^ 1 1 J J 0 100 200
MODIFIED CONE BEARING , 0 C , bar
Figure 5. Liquefaction Evaluation Procedure Based on Tip Resistance (Robertson and Campanella, 1983) 31 earthquakes. This chart, shown in Figure 6, defines the envelope of most liquefiable soils, denoted as Zone A.
3.2 Comparative Evaluation of the Available Procedures. In order to assess the applicability of the various liquefaction evaluationtion procedures summarized above, the testing program outlined in Table 1 was planned. The tests were performed in the Heber Road and Wildlife sites, Imperial
Valley, California, and the tests results are presented in
Appendix A.
Figures 7.a and 7.b illustrate the soil profiles at the two sites as reported by Youd and Bennett (1983). The penetration test results indicated the presence of several sublayers of sandy deposit. In order to conduct the comparative study, five different sand layers were selected and analysed using the seismic cone based procedure (Fig. 1) and three of the four CPT-based procedures (Figs. 3, 4, 6) outlined above. Youd and Bennett (1983) have reported the use of the fourth CPT-based procedure to determine the liquefaction susceptibility of the same deposits. Table 2 summarizes the method predictions and indicates that the four procedures yield consistent evaluations of the liquefaction potential of the sand layers considered. Moreover, these results are in good agreement with those reported by Youd and
Bennett (1983).
3.3 Potential Use of Piezo-cone Data
In order to use the piezocone data for in-situ " 100 •"
Figure 6.Liquefaction Evaluation Procedure Bases on Direct CONE BEARING , bevto (Robertson 1985)andCampanella, Observation 4 5 4 3 2 1 FRICTION s d n a s ' SLY B /SILTY ITS / SILT S / AND SILT S, / / ,% R F , O I T A R SANDY ^ IT GLADYS SILTY AN / SIL / ol a te HR site the at Soils # / Soils at the WL site WL the at Soils / y e y a l c / T / LTS / D / CLAYS T PEAT 32 N O . O F POROUS PENETRATION DISSIPATION SITE LOCATION TEST TYPE TESTS ELEMENT RATE (cm/sec) TESTS
F C 4 N.A. 2.0 N.A.
Heber Road SPC 3 Plastic & Ceramic 2.0 & 17.0 10
D P C 4 Plastic &. Ceramic 2 . 0 & I 7 . 0 3 I
F C 3 N.A. 2.0 N.A.
Wildlife S P C I Plastic 2 . 0 -
D P C 3 Plastic &. Ceramic 2 . 0 & 17 . 0 2 7
Table 1- The In-situ Testing Program
U) Depth, in meters Channel Artificial il sand fill, Figure7.a. Soil Stratification at Heber Road Site overbank sand A3 sand overbank Point bar bar Point ad Aj sand (Youd and Bennett, 1983) meters OtherLevee and ly and Clay ae table Water 100 iorm pr qae centimeter square per in kilograms reisistance, Penetration 0 rssac, n lw pr foot per blows in resistance, 8
f Standard-penetration Standard-penetration Cone Resistance
250 aat E •is. CO Depth, in feet West 40 35 30 25 20 5 0 10 15 ie Ra 3 Road River hs ra iy- area This iue 22 figure shown shown B Clay and clayey clayey and Clay B n rfin Figure 7.b. Soil Stratification at Wildlife Site it n sand and Silt
silt (Youdand Bennett, 1983) M m meters Standard-penet rat Standard-penet ion lw pr foot per blows in res i stance Road 100
eerto resistance Penetration n iorm pr qae centimeter square per kilograms in oe resistance Cone
East •M 42 S? 6 10 0 3 1 14 a V Q. B t Not Very Likely Heber Road 7.2-7.8 160 18 70 8 0 0 I Not in Zone A (23 1 Not Very Likely 12 Not Very Likely I Not Likely Heber Road 9.2-10.5 240 22 85 150 I Not in Zone A (3) 1 Not Likely 12 Not Likely 1 Likely Heber Road 12.4-13.3 150 30 55 4 8 0 I In Zone A, Likely (4) 1 Likely 12 Likely I Very Likely Wildlife 2 - 3 .5 3 0 7 15 3 0 0 n In Zone A. Likely (5) i Borderline Case, Likely 12 Very Likely I Very Likely Wildlife 4-6.5 60 12 35 4 0 0 I In Zone A, Likely (6) 111 Borderline Case, Likely 12 Very Likely Table 2- Liquefaction Potential ofth e S a n d Layers At the Heber Road and W ildlife Sites 37 liquefaction evaluation, effect of the main testing variables (location of porous element, porous element compressibility, and penetration rate) on the results of penetration tests must be adequately investigated. A parametric study was performed during this study to evaluate the effect of each testing variable on the measured soil response in terms of tip resistance, shaft friction, and generated excess pore pressure(s). Location of Porous Element (s). Previous experimental and theoretical studies (Al Awkati 1975; Parez et al., 1976; Baligh et al. , 1980; Tumay et al. , 1981; Robertson and Campanel1 a, 1983; Tumay et al., 1985; Lunne et al., 1986; Kiousis et al. , 1988; Bruzzi and Battaglio, 1988; Zudiberg, 1988) have pointed out the individual merits of piezocone testing with pore pressures measured either on the shaft above the cone tip or at the midsection of the cone tip. It has also been suggested that piezometric measurement on the cone tip is best suited for investigations regarding soil classification, whereas the pore pressures measured along the shaft tend to reflect the stress history of the sediments penetrated. Theoretical studies by Al-Awkati (1975), Tumay et al. (1981), and Kiousis et al . (1988) have further indicated the likelihood of the presence of a significant unloading zone (i.e. tendency for seperation of the soil and shaft interface) immediately behind the cone tip extending approximately 2-3 times the radius of the shaft. The concept of the dual pore pressure piezocone penetration test (DPCPT) has thus evolved from the necessity of making reliable measurements of pore pressures generated during a OPT for proper soil stratification / classification and stress history (OCR) identification. The respective locations of the piezometric elements were initially envisaged to be on the face of the cone tip and at midsection of the friction sleeve about 3 diameters behind the cone. Observations on the wear of the cone tip and friction sieeve with respect to penetrometer use (Tumay et al . , 1981) have demonstrated these locations as being subjected to maximum soi1-penetrometer interaction. Due to practical reasons, however, the piezometric element on the shaft was finally emplace immediately behind the friction sieeve 17 cm above the cone tip. Several investigators (Baligh, et al. , 1980; Tumay and Yilmaz; 1981, Campanella et al. , 1985; Lunne, et al. , 1986; Juran and Tumay, 1989; etc.) have discussed the effect of the 1ocation of porous element on the measured excess pore water pressures during a piezocone penetration test. Most of these studies have been conducted in fine grained 1ow permeabi1ity soi 1 s. They have demonstrated that the location of the piezometric element has a major effect on the magnitude of the measured pore water pressure. The fundamental differences in strain paths at the cone tip and along the penetrometer shaft results in a significantly different pore water pressure response. At the cone tip, the soil is subjected to both 39 maximum compression and interface shear. The generated pore water pressures are generally dominated by the increase of normal stress which can be related to the increase in the point resistance. Along the penetrometer shaft and in particular immediately behind the friction sleeve, the soil is subjected mainly to shearing and the measured pore water pressures depend primarily on the tendency of the saturated soil to dilate or contract during shearing. The pore water pressures measured immediately above the cone base were found to be highly dependent upon the stress history (i.e. OCR), sensitivity, and stiffness to strength ratio of the soil. Therefore, several classification charts have been developed using the point resistance and the excess pore water pressures measured immediately above the cone base (Senneset et al. , 1982; Jones and Rust, 1982; Robertson et al., 1985). Figure 8 illustrates the results of a DPCPT penetration test at the Heber Road site. In addition to the pore pressure records uL and u2 , the tip resistance and sleeve friction measured with the DPCPT equipment, the pore pressure measured just above the tip, during a second sounding, is presented in this figure. As indicated, as the tip penetrates the loose contracting sand seam, all three measured excess pore pressures increase. However, the excess pore pressure measured at the tip shows a much larger increase as compared with the other two. This is due to a large increase in the normal DEPTH (m) 10.4 12.4 8.4 9.4 0 6 2 040 80 120 160 200 Figure 8. Useof DPCPT Test in Identifying Loose Sand Seams os Sn Seam Sand Loose c (qc bar) f- (bar) 0 yrsai Pressure Hydrostatic 1 15 10 5 U|au |,a aU 2 (bar) 20 o 41 (octahedral) stress around the tip. The tip resistance and the sleeve friction readings facilitate identification of the loose sand seam. In this research program, in order to evaluate the effect of the location (along the shaft) of the porous element on the measured excess pore water pressures, two types of piezocones were used: a dual piezocone with one porous element behind the friction sleeve, and a seismic piezocone with a single porous element immediately above the conical tip. Analysis of the test results indicated several significant points which are discussed below: Heber Road Site. A total of 7 piezocone tests were performed at the Heber Road site. Out of these, 4 tests (9.1, 16.1, 17.1, 19.1) were conducted using the dual piezocone, while the other 3 (8.1, 15.1, 18) were performed using the seismic cone. Out of these seven tests, two tests (no. 18 and 19.1) were conducted at location 442' while the other five were performed at location 700'. The test results are presented in Appendix A. In order to analyse the effect of the location of the porous element on the measured excess pore water pressures, location 700' was selected. In particular, the results of the dual piezocone test no. 17.1 (with ceramic piezo-element) and the seismic piezocone test no. 8.1 will be considered: 0 to 2.0 m deep: A prepunch of 2 meters was used in this test. 42 2 to 5.5 m deep: In this sandy soil, the excess pore pressure record at the tip (Pore 1) consistently recorded higher values than those behind the friction sleeve (Pore 2). Moreover, the magnitudes of these pore pressures are small (< 1.0 bar). These results, typical of a relatively thick layer of a sandy deposit, are consistent with those reported by Juran and Tumay (1989) at the Dunkerque site. An important observation to be pointed out is the very close agreement between the magnitude of the measured excess pore water pressure behind the friction sleeve (Pore 2 in test 17.1) and that immediately above the cone tip (Pore 1 in test 8.1). As discussed above, the difference in the pore pressures measured at these different locations has not yet been rigorously investigated. These results indicate that for this relatively thick sandy deposit the difference between the pore pressures measured at these two locations are practically negligible. The dissipation test no. 17.2 performed at the depth of 4.0 meters permits verification of the low magnitude and rapid dissipation of excess pore water pressures generated during penetration in this soil layer. It is important to point out that in a few locations (e.g. 2.1, 2.5, 5.3, 5.5, 6.2, 6.5, 8.8, 9.2, 9.6) the pore pressure record immediately above the cone tip shows negative values at several depths . This observation would tend to agree with the results of numerical simulations conducted by 43 Kiousis et al. (1988) suggesting the likelihood of formation of a separation zone above the cone, indicating discontinuities in the pore pressure profile. 5.5 to 7.2 m deep: In this clayey layer, the excess pore water pressure values measured at the tip (pore 1 record in test 17.1) are consistently higher than those recorded behind the friction sleeve (Pore 2, test 17.1). In particular Pore 1 record values vary between 4 to 8 bars, whereas those recorded by Pore 2 range only between 2 and 4 bars. A comparison between the pore pressure recorded (using the seismic cone) immediately above the tip with that recorded (using dual piezocone) behind the friction sleeve indicates once again the close agreement between these two measurements in a clayey soil. 7.2 to 7.8 m deep: The excess pore pressures at all three locations (Pore 1 and Pore 2 in test no. 17.1, and Pore 1 in test no. 8.1) reduce to the hydrostatic pressure as the cone penetrates this sandy deposit. An important observation to be made is that dissipation tests 17.4 (using the dual piezocone) and 8.4 (using the seismic piezocone) indicate practically identical dissipation response behind the friction sleeve and immediately above the cone tip. 7.8 to 9.2 m deep: As the cone penetrates through this clayey deposit, a systematic increase in the excess pore water pressures at all three locations (Pore 1 and Pore 2 in test 44 no. 17.1, and Pore 1 in test no. 8.1) is observed. It is important to point out that in this layer, the pore pressure values measured behind the friction sleeve are close to those recorded just above the tip. 9.2 to 10.5 m deep: In this sand layer, a major difference between the pore water pressure measurements behind the sleeve (Pore 1, test no. 17.1) and above the tip (Pore 1, test no. 8.1) can be noticed. The pore pressure behind the sleeve seems to be almost uniform at about 0.7 bar, whereas the pore pressure above the tip showed continuous fluctuations even becoming negative. These observations tend to suggest that immediately above the tip, soil is subjected to a complex interaction with the cone penetrometer. These observations seem to be consistent with the results of numerical simulations reported by Kiousis et al. (1988). Results of the dissipation test no. 17.6 and no. 8.5 indicate that (a) with the tip at the depth of 10.06 m, the excess pore pressure generated at the cone tip dissipates very fast (Tso % 20 seconds) whereas it remains almost at the hydrostatic 1evel behind the friction sieeve, and (b) immediately above the tip, the excess pore pressure is below the hydrostatic 1evel implying an unioading zone at that 1 ocation, which would be consistent with the soi 1 -penetrometer separation predicted by finite element simulations (Kiousis et al., 1988). 10.5 to 12.3 m deep: There are a number of important 45 observations which need to be pointed out with regard to the excess pore water pressure records obtained within this clayey layer: (1) From 10.5 to 11.6 meter, the pore pressure at tip is consistently higher than that recorded above the cone and behind the friction sleeve. In this clayey layer, this is anticipated. (2) At the depth of 11.6 m, there exists a loose sand seam inclusion. In the inclusion, the excess pore pressures in all three locations increase. The increase in the pore water pressure at the tip is substantial (about twice that in the surrounding clay layer). It is important to notice that the pore pressure response behind the sleeve is very similar to that at the tip (i.e. an increase of about two times), while the response immediately above the tip is not as sensitive. This observation tends to suggest that the dual piezocone device may be better suited for detection of loose liquefiable sand seams. Similar results were obtained by Juran and Tumay (1989) at the Dunkerque site. (3) Dissipation test no. 17.7 indicates that with the tip in the loose sand inclusion, the excess pore pressure at the tip dissipates immediately (T50 % 9 seconds). Wildlife Management Site. A total of 4 piezocone tests were performed at the Wildlife site, three dual piezocone tests (no. 14.1, 20.1, 22.1) and one seismic piezocone test (no. 21). The results of the seismic piezocone test indicate 46 that the porous element was not completely saturated. Therefore, the following discussion will be limited to a comparison of the excess pore pressure measurements at the tip and immediately behind the friction sleeve along the shaft. For this purpose, results of dual piezocone test 20.1 along with the relevant dissipation tests are discussed below: 0 to 2 m deep: A prepunch of 2 m was used in this test. 2 to 4 m deep: This is a silty layer with clayey interbeds. Throughout this layer the pore pressure measured at the tip (Pore 1 record) is consistently higher than that measured behind the friction sleeve (Pore 2). 4 to 6.8 m deep: This is a layer of loose sandy deposit with almost no interbeds of silty or clayey materials. Pore pressure records (Pore 1 and Pore 2) indicate almost hydrostatic pressures throughout this layer. This is mainly due to the high permeability of this deposit. The results of the dissipation test no. 20.2 conducted at the depth of 6.18 m confirms that in this highly permeable deposit, the excess pore water pressures generated both at the tip and along the shaft are very small. They also indicate a ground water table depth of about 1.2 m. 6.8 to 11.2 m deep: This is a layer of clayey soil with several silty seams. The excess pore water pressures measured at the tip are consistently (about 2 times) higher than those along the shaft. At the levels of silt seams (e.g. at depths 7.5, 7.8, 8.5, etc.) the magnitudes of both measurements drop 47 indicating penetration through a material of higher permeability. The results of dissipation test no. 20.5 at the depth of 9.98 m indicates: (a) the excess pore pressure at the tip is about twice that along the shaft (behind the friction sleeve) at the start of the dissipation test, and (b) both records show a slow dissipation rate with T50 * 600 seconds indicating the presence of a low permeability soil typical of a silty clay. 11.2 to 12.6 m deep: This is a layer of medium dense sand deposit. The excess pore water pressures measured at the tip are consistently higher than those measured behind the friction sleeve. At the depth of 11.6 m, the tip resistance reduces from 120 to 80 bars, the pore pressure at the tip reduces from 16 to 2 bars, while the pore pressure along the shaft practically remains at the hydrostatic level. The subsequent increase in the tip resistance from 80 to 115 bars results in a substantial increase in the tip pore water pressure but has practically no effect on the pore water pressure measured at the sleeve in the medium dense sand. The results of the dissipation test no. 20.7 conducted at the depth of 11.58 m, indicates that the high excess pore pressures (reaching 11 bars) generated as the tip penetrated the medium dense sand immediately dissipate (Tso % 8 seconds). Porous Element Compressibility. The effect of saturation and compressibility of the piezo-element on the measured 48 excess pore water pressures during piezocone penetration tests have been the topic of a number of investigations (Campanella and Robertson; 1981, Rad and Tumay; 1985, Gupta and Davidson; 1986 and others). These have been primarilly focused on evaluation and improvement of procedures used for: 1. deairation of the porous element 2. deairation of the cone 3. protection of the deaired cone-element assembly prior to beginning of the test The porous element compressibility is a fundamental parameter which directly influences the quality of excess pore water pressure penetration / dissipation records (Campanella and Robertson, 1981; Lunne et al . , 1986). In order to investigate the effect of porous element stiffness on the measured excess pore water pressure records, two different porous elements were used in this study, a ceramic (aerolith 10 with hydraulic conductivity of 10'2 cm/sec) and a plastic element (with practically the same hydraulic conductivity but 1ower rigidity). Heber Road Site. In order to evaluate the effect of porous element stiffness on the excess pore pressures generated during the piez-cone penetration tests, two dual piezocone tests were performed at location 700’, test no. 17.1 with the ceramic tip and test no. 9.1 with the plastic tip. An anlysis of the results of the excess pore pressures measured in these two tests is presented below: 49 0 to 2 m deep: A prepunch of 2 m was used in this test. 2 to 5.5 m deep: In this relatively loose sand, the excess pore water pressures measured using the ceramic tip (Pore 1, test 17.1) were consistently (upto 1.5 times) higher than those measured using the plastic element (Pore 1, test 9.1). 5.5 to 7.2 m deep: In this clayey layer, once again, the pore pressures measured using the ceramic tip were higher than those measured using the pi astic tip. As the two tips penetrate through the silty sand seam at the depth of 5.8 m, both records show a drop in the measured pore pressure. Moreover, the plastic tip recorded much smaller pore pressures for the depths of 6 to 6.5 m. 7.2 to 7.8 m deep: In this medium dense sand deposit, the pore pressures measured with the two different tips were almost the same. In fact, the two tests indicated a rather different soil profile for the depths of 7.5 to 7.8 m. The profile predicted using the ceramic tip is, however, more consistent with that obtained using the friction cone (test no. 6). 7.8 to 9.1 m deep: In this clayey deposit, the pore pressures measured using the ceramic element are consistently (upto 1.5 times) higher than those measured using the plastic i tip. 9.1 to 10.3 m deep: At the start of this sandy layer, the pore pressure obtained using the plastic tip is higher than 50 that measured using the ceramic tip. As the tips penetrated through a silty inclusion (9.4 to 9.9 m), both measurements indicated a drop in the pore pressures, with the drop in the case of the ceramic tip (from 8 to 3.5 bars) being about half of that measured with the plastic tip (from 12 to 2 bars). As the tips pass through the silt seam back into the sandy deposit, the pore pressure measured using the plastic tip reaches about 10 bars, while that measured with the ceramic tip only reaches 6 bars. 10.3 to 12.4 m deep: In this clayey layer, pore pressure measurements using both tips indicated high (between 8 to 16 bars) values with several fluctuations. The magnitudes of the pore pressures measured using the plastic tip were in general higher than those using the ceramic tip. Wildlife Management Site. The results of the piezocone test no. 14.1 (using the plastic tip) and no. 20.1 (using the ceramic tip) conducted at the WSW position are discussed below: 0 to 2 m deep: A prepunch of about 2 m was used in this test. 2 to 6.8 m deep: This is a loose to medium dense sand deposit interspersed with very loose sand seams. Throughout this layer, the excess pore water pressures measured using the ceramic element were higher than those measured using the plastic tip. At the depth of about 3.7 m the tips penetrated a very loose sand seam. The pore pressures measured were about 51 5 bars, using the ceramic tip and about 2.5 bars, using the plastic tip. 6.8 to 11.3 m deep: In this clayey soil deposit, the pore pressures measured using the ceramic tip were always higher than those using the plastic tip. The only exception was at the depth of 7 m where the excess pore pressure using the plastic tip reached about 8 bars, while that using the ceramic tip was about 6 bars. Penetration Rate. A number of investigators (Campanella and Robertson; 1981, Canou; 1989, Juran and Tumay; 1989) have studied the effect of penetration rate on the induced excess pore water pressures and the cone resistance during cone penetration. The standard penetration rate for piezocone testing is 2 cm/sec. The main issue affecting the choice of a specific penetration rate is the drainage conditions during penetration. For high permeability clean sands and coarser materials, the excess pore pressures dissipate immediately and the penetration occurs under essentially drained conditions. However, for low permeability silty or clayey soils, the penetration takes place under partially drained to almost undrained conditions. Heber Road Site. In order to investigate the effect of the penetration rate on the measured tip resistance and excess pore water pressures at this site, one dual piezocone test (no. 16.1) and one seismic piezocone penetration test (no. 52 15.1) were performed at 17 cm/s. in addition to other tests (no. 17.1, 9.1, 8.1) which were conducted at the normal rate (2 cm/sec). A discussion of the results of tests no. 16.1 (fast rate, dual piezocone), and no. 17.1 (normal rate, dual piezocone) is presented below: 0 to 2 m deep: A prepunch of about 2 m was used in all three tests. 2 to 5.5 m deep: In this medium dense sand layer, the point resistance values obtained at the two rates agreed fairly well. The excess pore water pressures obtained at the tip under the fast rate were generally higher and included many more peaks as compared with those obtained under the normal rate. For this sandy layer, this result could be expected as under the fast rate the time allowed for the excess pore pressures to dissipate is much shorter and, therefore, the pore pressure response curve would include more fluctuations. The excess pore pressures measured along the shaft are almost the same at the two rates. 5.5 to 7.2 m deep: In this clayey silt layer, the point resistance values obtained at the two rates agreed very closely. However, the penetration records of friction ratio and pore water pressures obtained under the two rates indicate significantly different soil layers. From 5.5 to 6 m, the two penetration records indicate a sandy silt layer with approximately equal tip pore pressures. Higher pore pressures 53 were recorded at the sleeve under the faster penetration rate, probably due to larger residual pore pressures generated at the tip during penetration. From 6 to 7 m, the penetration rate seems to have a significant effect on the friction ratios and the related pore pressures. The faster penetration rate yields significantly higher friction ratios indicating a much stiffer (clayey) layer. The observed differences in the friction ratio profiles can also be related to different pore pressure response. In the stiffer cl ay, lower pore pressures are recorded at both the tip and the sleeve with negative pore pressure at the sleeve. This basic difference is also illustrated by comparing the results of the dissipation tests no. 16.3 and 16.4 with those of tests no. 17.2, 17.3 and 17.4. The higher penetration rate results in an increase of the tip pore pressures during dissipation which would suggest an apparently partially saturated clayey soil. 7.2 to 7.8 m deep: In this medium dense sand layer, the penetration rate seems to have practically no effect on the measured tip resistance. It does, however, significantly affect the excess pore water pressures measured at the tip. Under the fast rate, an excess pore pressure of about 20 bars was reached, while under the normal rate the pore pressure reached only 14 bars. The tip pore water pressure recorded under the fast rate is much more sensitive to the soil stratification as detected by the change in the tip resistance. The excess pore water pressures measured at the 54 sleeve under the two penetration rate were almost equal. 7.8 to 9.1 m deep: In this sandy silt to clayey silt layer, the tip resistance values obtained under the two rates were almost identical. The excess pore water pressure records obtained at both the tip and the sleeve under the fast rate were consistently lower than those obtained at the normal rate. The lower pore water pressure recorded under the fast penetration rate can be again related to higher friction ratio indicating an apparently stiffer clay layer. It is also of interest to indicate that for locations where the friction ratio under the two penetration rates are approximately equal, the excess pore water pressures recorded at both the tip and the sleeve under the two rates are also about the same. However, the observed correlations among penetration rate, friction ratio, and pore water pressure in silty clay layers must be further investigated. 9.1 to 10.5 m deep: In this medium dense sand deposit, the difference in the penetration rate yields the following observations: 1. The point resistance records obtained at the fast rate indicate a more stratified sand layer with two distinct sublayers of lower relative density at about 9.2 and 9.6 m depths, respectively. The tip resistance is reduced from 180 to 100 bars in the first sublayer, and from 170 to 130 bars in the second. The tip resistance profile obtained at the normal 55 rate shows only one such sublayer with a net reduction in the relative density. 2. The pore water pressures recorded under the two rates are very sensitive to the change in the tip resistance. The pore pressures obtained under the fast penetration rate were about 2 to 3 times those obtaines at the normal rate. This observation once again points out the significance of the penetration rate in detecting sublayers of lower relative density within a relatively thick sand deposit. This aspect is of practical consequence as liquefaction is most likely to be initiated in these loose sublayers. 3. The pore pressure measured at the sleeve (us) under the fast rate was also higher than that measured under the normal rate. However, this difference in magnitude is primarily due to the fact that under the fast penetration rate the piezoelement at the sleeve records residual tip pore pressures, while under the normal penetration rate these residual pore pressures would entirely dissipate. Therefore, this observed difference in the measured pore pressures is not an indication of contractancy / dilatancy soil behavior and is difficult to analyse. 10.5 to 12.2 m deep: This is a silty soil layer with a loose sand seam at the depth of about 11.5 m. The penetration rate seems to have very slight effect on the tip resistance values, however, it substantially affects the sleeve friction values and therefore the soil classification. The results of 56 fast penetration indicates high friction ratios and tip excess pore water pressures close to the hydrostatic value, suggesting a stiff clayey layer. The normal penetration showed a substantial increase in the tip excess pore pressure to about 24 bars and friction ratio of about 2 which would suggest a loose silty sand seam. The dissipation test no. 17.7, conducted with the tip at 11.58 m, indicates the high permeability of this sand seam. Under the fast rate, the excess pore water pressure measured at the sleeve is larger than those at the tip, probably due to the time lag in the excess pore water pressure response in the silty soil (or the stiff clayey soil) which did not allow the tip pore pressures fully dissipate within the short tip-sleeve travel time of about 1 second. This time lag is clearly illustrated in the results of the dissipation test no. 17.8. Under the normal rate, the tip excess pore water pressures were higher than those measured at the sleeve. 12.2 to 13.2 m deep: The peak tip resistance values are practically not affected by the change in the penetration rate. The tip excess pore pressure records appear to indicate a higher sensitivity to the changes in the tip resistance under the normal penetration rate. The excess pore water pressures measured at the tip under the normal rate are higher than those measured under the fast rate. Under the normal rate, the tip excess pore water pressures were higher than those recorded at the sleeve. Under 57 the fast rate, due to the short time allowed for dissipation of residual tip excess pore pressures, the excess pore pressures measured at the sleeve are higher than those at the tip. Wildlife Management Site. A comparison between the results of the dual piezocone penetration tests 22.1 (at 17 cm/sec) and 20.1 (at 2 cm/sec) is presented below: 0 to 2 m deep: A prepunch of about 2 m was used in this test. 2 to 4 m deep: In this loose sandy deposit, the point resistance values measured under the two rates were practically the same. The excess pore pressures measured both at the tip and at the sleeve were practically identical. 4 to 6.5 m deep: In this medium dense sand layer, the tip resistance values measured under the fast rate were higher than (about 1.5 to 2 times) those measured under the normal rate. The fast rate tip resistance and tip pore pressure indicate a sublayer of higher relative density at about 5 m depth. As previously discussed, in general, an increase in the tip resistance yields an increase in the tip pore pressure. The pore pressures measured at the sleeve under both penetration rates remained very close to the hydrostatic pressure, which would be expected for medium dense sands. 6.5 to 11.3 m deep: In this silty clay deposit, the tip resistance values obtained at the two rates agreed very closely. The excess pore water pressures measured at the tip 58 were slightly higher in the case of the fast rate penetration (test no. 22.1). The excess pore water pressures measured along the shaft indicated significant differences in results for the two tests. The values obtained under the normal rate indicated large excess pore pressures in the order of 4 to 8 bars with some fluctuations, whereas those obtained under the fast rate were very small becoming negative at depths between 6.8 to 8.4 m, 8.8 to 9.2 m, 9.6 to 10.0 m, and 10.5 to 10.9 m. The effect of the penetration rate on the excess pore water pressures measured at the sleeve is not well understood. Two patterns can be consistently identified in both Heber Road and Wildlife Management sites. Firstly, the residual tip pore pressures cause the measured pore pressure at the sleeve to be higher than that under the normal rate due to the short dissipation time (about 1 second). Secondly, it seems that the high values of the friction ratios determined under the fast penetration rate in a certain layer could be correlated to the pore pressure measured at the sleeve. Comparing the sleeve pore pressure records obtained under the fast rate in the two sites, it appears that peak friction ratios are associated with low to negative excess pore pressures at the sleeve. In silty layers where the two penetration rates yield approximately the same friction ratio values, the recorded excess pore pressures at the sleeve were also close. These trends need to be further investigated. Chapter 4 STATIC AND CYCLIC TRIAXIAL LIQUEFACTION RESISTANCE OF THE HRD SAND As pointed out in chapter 2, liquefaction of sandy soils has been the topic of several studies dating back to the early 1960's. Presently, it seems that there exists two classes of thought with respect to liquefaction phenomena and the triggering mechanism. The first, originated by the work of Seed and Lee (1966), and further elaborated on through the works of Seed and his co-workers (Seed and Peacock, 1971; Seed et al., 1975; Seed, 1976; Seed, 1979), depicts liquefaction as a dynamic loading phenomenon which should be investigated through laboratory cyclic tests. The second school of thought, originated by the work of Castro (1969), and elaborated on through the works of Castro and his co-workers (Castro and Poulos, 1976; Poulos, 1981; Castro et al. , 1982; Poulos et al . , 1985), describes liquefaction as a static loading phenomenon which must be investigated through laboratory monotonic tests. However, as mentioned in chapter 1, so far there has been no or little research on: (a) possible relationships between mechanisms initiating static and cyclic liquefaction, and (b) the effect of stress path on liquefaction resistance of a soil. In order to address the above research needs, a comprehensive experimental program involving monotonic and cyclic triaxial liquefaction tests, and monotonic and cyclic 59 60 hollow cylinder cell liquefaction tests was established. The results of monotonic and cyclic triaxial tests are presented and discussed in this chapter. 4.1 Monotonic Triaxial Liquefaction Tests The purpose of these tests were to: (a) determine the effect of testing variables, specifically consolidation pressure and relative density, on the static liquefaction resistance of the HRD sand, (b) determine the steady-state (failure) line for this sand, (c) evaluate the collapse surface concept proposed by Sladen et al. (1985), and (d) provide, along with cyclic liquefaction test results, the necessary data base for studying mechanisms of monotonic versus cyclic liquefaction. Testing Equipment. To perform the strain-controlled monotonic liquefaction tests, a classical triaxial testing system manufactured by Wykeham Farrace, Inc. with maximum loading capacity of 5 tons, and a specimen size of 76mm x 38mm was used. The instrumentation involved a force transducer to measure the axial load, and therefore the major principal stress, a LVDT to determine the axial displacement, and therefore the axial strain, and two pressure transducers, one to measure the back-pressure during saturation stage, and another one to measure the excess pore water pressure generated during the test. The range of displacement rates which could be applied to the specimen for this triaxial system was from .1 to 50 mm/min. Data Acquisition and Control. The data acquisition system consisted of a 16 channel data logging unit, a A/D board and an a personal computer with a 286 microprocessor. During the test, analog data were continuously sent by the force transducer, the LVDT, and the pore pressure transducer to the acquistion system. The analog data were intercepted at the A/D interface, converted into digital readings and collected and stored by the computer using a pre-written software provided by the Wykeham Farrance, Inc. The use of this data acquisition system ensured complete test automation and thereby helped reducing the human errors involved in performing the test. Specimen Compaction Technique. As mentioned in chapter 2, a standard specimen compaction technique for laboratory, monotonic or cyclic, liquefaction tests does not presently exist. Castro et al. (1982) and Poulos et al. (1985) have reported that specimen preparation technique does not affect the monotonic liquefaction resistance of sands, while Mulilis (1975) and Townsend (1978) have shown that the preparation method can have an appreciable effect on the cyclic liquefaction resistance of sandy soils. For the purpose of this study, a specimen preparation method proposed by Canou (1989) was used. This technique, which is a modification to the procedure used by Castro (1969), involves mixing the dry soil with water prior to its placement into the mold. The premoistening of sand induces an artificial adhesion and reduces compaction due to placement of successive layers. A 62 mixing water content of 4% by dry weight, suggested by Canou (1989), was used throughout this study. Saturation Technique. Presently, back-pressure saturation seems to be the most widely used saturation technique in experimental investigations (Head, 1986). The sand type used in this study is very fine and even with the backpressure saturation technique, it required incrementally increasing the backpressure and the cell pressure to about 70 psi before a value of B-parameter of 0.95 or higher could be reached. In order to speed up the saturation process in case of fine sands, use of liquid C02 perculation (purging) is becoming a standard practice (Alarcon, 1986; Frost, 1989; Canou, 1989). In this study liquid C02 was perculated through the specimen for 10 to 15 minutes under a low pressure gradient of about 3 psi prior to incremental back-pressure saturation stage. This extra step not only reduced the overall time taken for saturating each specimen but also decreased the required maximum back-pressure from about 70 psi to about 50 psi. Specimen Preparation Procedure. The standard triaxial CU testing procedure (Head, 1986) was used throughout this part of the research. This procedure is briefly described below: 1. A preweighed amount of sand was mixed with distilled water at a proportion of 4% by dry weight. 2. A standard three-piece split mold, for 76mm x 38mm specimens, was assembled on the triaxial cell base. A latex membrane was placed inside the mold, two O-rings were placed 63 at the top and bottom ends of the mold, and a vacuum of about 6 psi was applied to the membrane through an inlet to control the compaction during the specimen placement stage. 3. The lower end of the membrane was fixed around the bottom end cap using the lower O-ring. After placing a porous stone on the bottom end cap, the sand was placed in layers inside the mold using a spatula. To achieve a higher specimen uniformity, a total of 8 to 9 layers were used. 4. After the specimen reached the desired height, with the vacuum on, the top cap and porous stone were gently put in place, and the membrane was fixed around the top cap using the upper O-ring. 5. A vacuum of about 3 psi was applied to the specimen through a pressure inlet at the base, the mold was gently disassembled, the specimen dimensions were carefully determined using a vernier, the triaxial cell was put in place and was screwed down into the base. 6. The cell was filled with de-airea water, a confining pressure of 5 psi was applied to the specimen, and the vacuum was disconnected from the base of the cell. 7. Liquid C02 was perculated through the specimen for 10 to 15 minutes under a low pressure gradient of about 2 psi. 6. Using the standard incremental back-pressure saturation technique (Head, 1986), the sample was saturated. The criterion used in determining the end of the saturation stage was to reach a minimum value of .98 for the B-coefficient. 64 7. After the sample was saturated, the cell pressure and the back-pressure were reduced and the sample was allowed to consolidate under an effective consolidation pressure of predetermined value. During the saturation, the axial specimen deformation was accurately monitored but generally no appreciable reduction in specimen height was recorded. 8. After the consolidation stage was completed, determined by a condition of zero excess pore pressure maintained over a time period of at least 2 hours, the drainage valve was closed, and the sample was sheared to failure. This was achieved by application of a constant vertical displacement rate to the specimen. Vertical Displacement Rate. Castro et al . (1982), Poulos et al . (1985), and Canou (1989), among others, have indicated that the displacement rate has no effect on the monotonic triaxial liquefaction resistance of sandy soils. For instance, Canou (1989) has shown that for a displacement range of .01 to 1.0 mm/min the monotonic liquefaction resistance is essentially unchanged. A displacement rate of .5 mm/min was adopted for all monotonic triaxial liquefaction tests performed in this study. Area Correction. It has been customary (Bishop and Henkel, 1962; Castro, 1969; Head, 1986) to correct the specimen cross-sectional area during a triaxial undrained test using the following relationship: A = Ac/(l-ea ) where: A = initial specimen cross-sectional area A0 = corrected cross-sectional area ea = axial strain The above correction was also suggested by other investigators (Castro, 1969) for monotonic and cyclic triaxial liquefaction tests. It must be pointed out that although this relationship was originally developed for materials with a Poisson ratio of 0.5, it has been extensively used for undrained tests on sands. In this study, the above relationship was used to correct the deviatoric stress during undrained tests. Correction for Membrane Rigidity. Bishop (1962), Castro (1969), among others, have proposed the following correction to the applied deviatoric stress to take into account the membrane rigidity: Aod = 4 M ea (l-ea) / D where: M = modulus of the membrane per unit width D = initial specimen diameter A value of M ranging from 0.21 to 0.32 kg/cm has been used in other investigations (Castro, 1969). For the tests performed in this study, an average value of 0.26 kg/cm was used for the modulus M. Correction for Membrane Penetration. A net inward pressure acts at the outer surface (perimeter) of a specimen during various stages of any static or cyclic undrained 66 triaxial tests. In case of granular non-cohesive soils, such as sands or silty sands, this net inward pressure could cause the membrane to penetrate into the specimen thereby changing the system compliance, which in turn affects the precision of the test results. Several investigators (Newland and Allely, 1959; Frydman et al.', 1973; Lade and Hernandez, 1977; Wu and Chang, 1982; Kramer, 1985, and others) have studied the effect of membrane penetration on results of undrained triaxial test, and proposed various correction schemes. Recently, Sladen et al. (1985) have presented a review of several of these correction procedures. They showed that the membrane penetration effect increases logarithmically with increasing mean grain size, D50, of a soil. Moreover, they showed that for granular soils with D50 less than 0.60 mm membrane penetration is practicaly negligible. Therefore, for the HRD sand with D50=0.13 mm no membrane penetration correction would be necessary. Effect of Testing Variables: A Parametric Study In order to evaluate the effect of consolidation pressure and relative density on the monotonic liquefaction resistance of the HRD sand, a testing program was carried out involving 9 tests. The results of these tests are presented and discussed in the following section. Effect of Relative Density. As shown in table 1 in chapter 3, the in-situ relative density and effective overburden pressure for the HRd sand were about 20% and 10 psi, respectively. In order to determine the effect of relative density on the monotonic liquefaction resistance of the HRD sand, 5 tests were performed on the specimens prepared under an initial consolidation pressure (o'c ) of 10 psi and relative density (Dr ) values of 10, 20, 30, 40, and 50%. For each test the stress deviator, (o1-o3), and generated excess pore water pressure measured during the test are plotted as a function of the axial strain, ea . The results of theses tests are shown in figures 9 and 10. The results of each test, plotted seperately, are presented in Appendix B. Interpretation of The Test Results. Figures 9 and 10 illustrate the effect of the relative density on the deviatoric stress and excess pore pressure response of the HRD sand specimens during monotonic triaxial tests. As shown in figure 9, for the sample with a relative density of 10% the deviatoric stress rapidly increases to a maximum of about 4.2 psi at a low axial strain of about 1% and thereafter decreases to a very small, almost zero, value. The corresponding excess pore pressure response curve for this test, shown in figure 10 indicates that at the peak deviatoric stress the excess pore pressure has not reached the consolidation of 10 psi, and in fact attains a value of about 8.50 psi. This observation confirms those previously reported by ohter investigstors (Canou, 1989; Kramer, 1985). Deviatoric Stress (psi) 20 25 10 15 5 0 0 Figure 9. Effect ofRelative Density onthe Deviatoric Stress During MonotonicTriaxial Tests 5 i s p 0 1 - xa Sri (%) Strain Axial 10 Dr=10% D=30 % D=20% ^5 % D^=50 D=40 % 15 20 68 Excess Pore Pressure (psi) Figure10. Effect of Relative Density onthe Excess Pore 10 6 B 2 0 4 0 WaterPressure During Monotonic Triaxial Tests 10 5 xa Sri (%) Strain Axial 15 20 69 Figure 9 also indicates that the maximum deviatoric stress reached during the test increases with an increase in the relative density of the specimen. However, the axial strain corresponding to the peak deviatoric stress does not significantly change, and in fact it only reaches a value of about 1.5%, for specimens with Dt values of 30 and 40%. As indicated in figure 9, the change in relative density also affects the residual deviatoric stress, that is the deviatoric stress value at large strain. For instance, an increase in relative density from 20 to 30% increases the residual deviatoric stress by about 1.70 psi. Perhaps the most significant observation to be made from the results shown in figures 9 and 10 is the change in material response from contracting (i.e. liquefiable) to dilating or nonliquefiable. As the specimen relative density increases to 30% and higher the post-peak softening, or decrease in deviatoric stress, gradually disappears. At the 40% relative density there is essentially no post-peak softening, and the specimen prepared at 50% shows a significant increase in the deviatoric stress starting at an axial strain of about 4%. The change in the material response from contracting to dilating can be further explained through the excess pore pressure response curves. The specimen prepared to a relative density of 10% represents the most liquefiable state among the five specimens tested. The excess pore pressure for this specimen reached the maximum value of 71 the consolidation pressure, or 10 psi, indicating a condition of zero effective stress. On the other hand, the excess pore pressure for the specimens prepared to relative densities of 20 and 30% are continuously increasing until a certain axial strain is reached and then remain constant. It is important to notice that the maximum value of excess pore pressure reached for the specimens prepared to 10, 20, and 30% relative densities decreased with an increase in the specimen relative density. This clearly indicates that an increase in the relative density increases the soil resistance to liquefaction. As the specimen relative density is further increased to 40 and 50%, the material response drastically changes, in fact the excess pore pressure decreases beyond a certain axial strain as the material response changes from liquefiable to nonliquefiable. Determination of the Steady-state Line. As discussed in chapter 2, the steady-state line approach is presently used almost exclusively to determine the triaxial monotonic liquefaction potential of soils. Figure 11 illustrates the steady-state line for the HRD sand. The open and filled circles represent, respectively, the initial and failure states in terms of the specimen void ratio, e, and the average effective stress, P ' = (a 'x+o'3)/2 . The monotonic liquefaction potential of soils has been defined to be a function of the horizontal distance from the initial state to the steady-state 1 ine: Void Ratio (e) .95 9 0. 0.90 1.05 .B5 B 0. 1.00 Figure11. The Steady-state Line for the SandHRD ObtainedFrom Triaxial Tests tay tt Line State Steady 1 ' (psi) P' CK K C • D =50 % 0 5 = D t Q 10 % 0 3 = D % 0 4 = D 108 72 73 Lp = (P'i-P’t )/P't ...... Canou (1989) Lp = (o'3i-a’3£)/o'3£ ...... Casagrande (1976) where P' and o'3 represent, respectively, the average and minor effective principal stresses, and subscripts i and f refer to the initial and failure states of a specimen, respectively. As can be seen in figure 11, the specimen with a relative density of 10% has the highest liquefaction potential among the five specimens tested. Effective Stress Paths. The concept of effective stress path (ESP) is used extensively in geotechnical engineering to study the stress states to which a soil is subjected during a particular loading event in field or in laboaratory. Figure 12 illustrates the effective stress paths for the five monotonic tests discussed earlier in this section. The coordinate axes in this figure are defined as the average stresses P ' = (c+o'3)/2 and q=(o\ - a '3 )/2, which are measured throughout the test. As illustrated in figure 12, the specimen relative density highly affects the obtained effective stress paths. The ESP for the highly liquefiable specimens with relative densities of 10 and 20% indicate that both P' and q decrease substantially toward failure, becoming almost equal to zero in case of the specimen at 10% relative density. However, as the relative density is increased the residual q value increases Figure 12. Effect of Relative Density on the Effective Stress (psi) PathsObtained From MonotonicTriaxial Tests O'c=10psi * (psi)P* 74 75 and the decrease in the value of P ' is not as large. In fact, for the specimens prepared to relative densities of 40 and 50%, the value of P ’ actually starts to increase as the failure is approached. The five effective stress paths, shown in figure 12, define a unique failure envelope marked by the solid straight line passing through the failure points. Effect of Consolidation Pressure. In order to determine the effect of consolidation pressure on the monotonic liquefaction resistance of the HRD sand, 4 tests were performed on the specimens prepared to a relative density (Dr ) of 20% and consolidation pressure (o'c ) values of 10, 20, 30, and 40 psi. For each test the stress deviator, (ox-o2 ), and generated excess pore water pressure measured during the test are plotted as a function of the axial strain, ea. The results of these tests are shown in figures 13 and 14. The results for each test, plotted separately, are presented in Appendix B. Interpretation of the Test Results. Figures 13 and 14 illustrate the effect of the consolidation pressure on the deviatoric stress and excess pore pressure response of the HRD sand specimens during monotonic triaxial tests. As shown in figure 13, for each specimen the deviatoric stress rapidly increases to a maximum value at a low axial strain of about 1% and thereafter decreases to a much lower value. The corresponding excess pore pressure response curves for these tests, shown in figure 14, indicate that for none of these tests the excess pore pressure at peak deviatoric stress Deviatoric Stress (psi) 20 25 15 10 Figure 13. Effect ofConfining Pressure on the Deviatoric 5 0 0 StressDuring Monotonic Triaxial Tests 5 xa Sri (%) Strain Axial D% =20 10 -40psi =30psi =20psi =10psi 15 20 76 Figure 14. Effect of Confining Pressure on the Excess Pore Excess Pore Pressure (psi) 20 25 0 WaterPressure During Monotonic Triaxial Tests 5 xa Sri (%) Strain Axial 0 2 = D 10 i s p 0 4 = c ' O i s p 0 1 = c ' C i s p 0 2 = c ' a 15 20 77 78 reaches a value equal equal to the corresponding consolidation pressure. This observation has been previously reported by ohter investigators (Canou, 1989; Kramer, 1985). Figure 13 also indicates that the maximum (peak) deviatoric stress reached during the test increases with an increase in the consolidation pressure of the specimen. However, the axial strain corresponding to the peak deviatoric stress does not significantly change. As indicated in this figure, the change in consolidation pressure also affects the residual deviatoric stress, that is the deviatoric stress value at large strain. For instance, an increase in o'c from 20 to 30 psi increases the residual deviatoric stress by about 2.50 psi. It is important to notice that under a low relative density of 20%, even application of a high confining pressure of 40 psi will not prevent the specimen from undergoing monotonic liquefaction. On the other hand, as discussed above, under a confining pressure of 10 psi, an increase in the relative density to about 50% will be sufficient to prevent monotonic liquefaction. These observations are of high practical relevance as far as the in-situ liquefaction susceptibility of the HRD sand is concerned. They seem to suggest that for the HRD sand, with in-situ overburden pressure and relative density of 50 psi and 20%, respectively, an in-situ densification method would be a much more efficient means of reducing the liquefaction risks. 79 Effective Stress Paths and Collapse Surface. The concept of effective stress path (ESP) is used extensively in geotechnical engineering to study the stress states to which a soil is subjected during a particular loading event in field or in laboratory. Figure 15 illustrates the effective stress paths for the four monotonic tests discussed earlier in this section. The coordinate axes in this figure are defined as the average effective stresses P' = (a\ +o’3)/2 and q=(c \ - a *3 )/2 , which are monitored throughout the test. As expected, under a constant relative density of 20%, the obtained effective stress paths shown in figure 15 are geometrically similar indicating that this soil response is independent of the applied confining pressure. As discussed in chapter 2, Sladen et al. (1985) have discussed the use of the collapse surface concept in determining monotonic liquefaction potential of sandy soils. In figure 15, the collapse surface is shown as the soild straight line connecting the points of maximum q on each ESP and passing through their commom intersection point, P, on the failure line. 4.2 Cyclic Triaxial Liquefaction Tests The purpose of these tests were to: (a) determine the effect of testing variables, specifically consolidation pressure and relative density, on the cyclic liquefaction resistance of the HRD sand, and (b) provide, along with static liquefaction test results, the necessary data base for Figure15. Effect of Confining Pressure onthe Effective (psi) 10 8 6 2 4 0 0 2 % =20 D Stress Paths From MonotonicTriaxial Tests 20 ' (psi) P' 30 050 40 81 studying mechanisms of monotonic versus cyclic liquefaction. Testing Equipment. To perform the cyclic liquefaction tests, a closed loop servo-hydraulic system manufactured by MTS, Inc. was used. The system included two basic parts: a main unit including the triaxial load frame, triaxial cell, and accessories, and a subsystem including the specially designed and constructed hollow cylinder cell along with the lateral pressure application system, developed for the purpose of the cavity expansion tests. The triaxial unit consisted of a hydraulically operated loading frame with a maximum axial load capacity of 22 kips, a plexiglass triaxial cell allowing for specimen sizes up to 5 inches in diameter (10 inches in height) and a maximum confining pressure of 300 psi, and the instrumentation including a dynamic rated force transducer, a LVDT, and a pore pressure transducer. The hollow cylinder cell subsystem will be discussed in a later chapter. Data Requisition and Control. The data acquisition hardware included a high performance 12 channel, 16 bit A/D board manufactured by Data Translation, Inc., a signal conditioning / amplifying interface, and an IBM PS/2 computer with a model 80286 (AT Compatible) Intel microprocessor equipped with a math co-processor to speed up the digital operations. The software used was Labtech Notebook Release 4.30 developed by the Cyber Research, Inc. This software 82 allows a real-time access to the experimental data which permits on-screen observation of the experimental results. Specimen Compaction, Saturation, and Preparation. The same procedures described in section 4.1, in the case of monotonic triaxial liquefaction tests, were used. Loading Frequency. Townsend (1978) has reported a review of several investigations indicating that for a loading frequency range of .1 to 1 Hz, the change in the liquefaction resistance is less than 10%. Therefore, for the purpose of this study a frequency of 1 Hz was used in all cyclic liquefaction tests. Loading Wave Form. Townsend (1978) has reported a review of several investigations indicating that the form of the loading wave (i.e. ramp, sinusoidal, rectangular, etc.) has virtually no effect on the cyclic shear strength of sands. Therefore, a sinusoidal loading wave, with a frequency of 1 Hz, was used in all cyclic liquefaction tests in this study. Applied Cyclic Stress Ratio. In order to establish a realistic correlation between laboratory and in-situ liquefaction resistance of a soil, it would be necessary to subject the laboratory specimens to the field loading conditions. However, as discussed in chapter 2, this is difficult to achieve because: (a) the actual earthquake load is a transient, non-periodic stress wave which cannot be easily reproduced in the laboratory, (b) the laboratory specimens are not true representative of the actual field 83 conditions. A significant source of uncertainty in trying to establish laboratory-field liquefaction resistance correlations, for a given soil, is the means by which the laboratory applied cyclic stress ratio is determined. In the field, as discussed in chapter 3, this value is estimated using the semi-empirical procedure proposed by Seed and Idriss (1970). Determination of the appropriate cyclic stress ratio for the 1aboratory specimens, however, is not a straight forward task. It is well established that laboratory tests on reconstituted specimens underestimate the liquefaction resistance. The in-situ liquefaction resistance is normaly much higher than that measured in laboratory due to cementation and aging effects (Seed, 1976; Rad and Clough, 1984). Therefore, a certain amount of engineering judgement is required to decide the value of the applied cyclic stress ratio for laboratory tests. At the Heber Road site, as discussed in chapter 3, the average cyclic stress ratio for an earthquake of magnitude 6.5 M was conservatively estimated to be about 0.50. However, other investigations at the Imperial Valley sites (Holzer et al . , 1989) have revealed that the actual in-situ cyclic stress ratio could be much lower, and that for the purpose of laboratory tests on reconstituted specimens cyclic stress ratios between 0.20 to 0.30 may be more appropriate. Thus, for the purpose of triaxial cyclic liquefaction tests performed in this study, a cyclic stress ratio of 0.20 was adopted. 84 Effect of Testing Variables: A Parametric Study In order to determine the effect of relative density and consolidation pressure on the cyclic liquefaction potential of the HRD sand a series of 9 tests were performed. The test ) results are presented and discussed in the following sections. Effect of Relative Density. In order to determine the effect of the relative density (Dr) on the cyclic liquefaction resistance of the HRD sand, a series of 5 tests were performed on the specimens prepared at relative densities of 20, 30, 40, 50, and 80% consolidated under a pressure of 10 psi. Interpretation of The Test Results. For each test, the excess pore pressure and the axial strain developed during the test were directly monitored using a LVDT and a pressure transducer, respectively. Furthermore, the data acquisition 16 channel board was programmed to calculate the major and minor effective principal stresses (o^ and o'3 ) throughout the test, so that the effective stress path (q-p' diagram) are obtained for each test. The values of were calculated by measuring the axial force, using the fatigue rated load cell, dividing that by the corrected specimen cross-sectional area, and subtracting from that the measured excess pore pressure. The o'3 values were simply determined by subtracting the excess pore pressure generated during the test from the total confining pressure, o3 , which was maintained constant during the test. 85 The test results for the specimens with relative densities of 20 to 50 are presented in Appendix C. These results clearly indicate that as the specimen relative density is increased, the rate at which the generated excess pore pressure and axial strain develop decrease. Figure 16 illustrate the number of stress cycles, at 1 Hz frequency, required to liquefy specimens of the HRD sand prepared under relative densities of 20, 30, 40, and 50%. It is clearly indicated that an increase in the relative density results in a denser specimen which is more resistant to liquefaction. In order to further investigate this observation, a test was performed on a specimen prepared to 80% relative density and 10 psi consolidation pressure. The test results shown in figures 17 through 19 indicate very significant changes in the soil response. The rate of generation of the excess pore pressure decreases after about 330 seconds and the value of the excess pore pressure attains a constant value of about 4.5 psi after about 410 seconds. This phenomenon, as discussed in chapter 2, is referred to as cyclic mobility. As the axial strain response indicates further stress cycles could induce limited axial strains but no excess pore pressure. Therefore, a complete liquefaction or flow will not take place. Effect of Consolidation Pressure. In order to determine the effect of the consolidation pressure (o'c) on the cyclic liquefaction resistance of the HRD sand, a series of 4 tests Figure 16. Effect of Relative Densityon Cyclic Liquefaction No. Cycles to Liquefaction 200 150 100 50 0 0 Resistance of SandHRD From Triaxial Tests 10 20 eaie est (%) Density Relative 30 40 50 60 70 86 Excess Pore Pressure (psi) 5 6 7 3 4 2 1 0 Figure 17. The Excess Pore Pressure Response for Dt=80 % 0 Dr=80% 0 8 = r D ; i s p 0 1 = ' a 100 (seconds) e m i T 200 300 400 0 0 5 87 Axial Strain (%) 12 15 IB 6 0 g 3 0 Figure18. The Axial Strain Response for Dt= 80 % Dr-80% 0 8 - r D ; i s p 0 1 = c ' G 100 (second) e m i T 200 0 0 3 400 0 0 5 88 CT'C=10 psi D = 8 0 % e b 10 12 P\ psi Figure 19. The Effective Stress Path for Dr=80 % 90 were performed on the specimens prepared at a relative density of 20% consolidated under effective pressures of 10, 20, 30, and 40 psi. Interpretation of the Test Results. For each test, the excess pore pressure and the axial strain developed during the test were directly monitored using a LVDT and a pressure transducer, respectively. Furthermore, the 16 channel data acquisition board was programmed to calculate the major and minor effective principal stresses and o'3 ) throughout the test, so that the effective stress path (q-p' diagram) are obtained for each test. The values of were calculated by measuring the axial force, using the fatigue rated load cell, dividing that by the corrected specimen cross-sectional area, and subtracting from that the measured excess pore pressure. The o'3 values were simply determined by subtracting the excess pore pressure generated during the test from the total confining pressure, o3 , which was maintained constant during the test. The test results are presented in Appendix C. These results clearly indicate that as the consolidation pressure is increased, the rate at which the generated excess pore pressure and axial strain develop decrease. Figure 20 illustrate the number of stress cycles, at 1 Hz frequency, required to liquefy specimens of the HRD sand prepared under consolidation pressures of 10, 20, 30, and 40 psi. It is clearly indicated that an increase in the Figure 20. Effect ofConfining Pressureon Cyclic Liquefacti No. Cycles to Liquefaction 200 300 250 150 100 50 0 0 Resistance of HRDSand From Triaxial Tests 10 ofnn Pesr (psi) Pressure Confining D=20% 20 30 40 50 92 consolidation pressure results in an increase in the specimen resistance to liquefaction. 4.3 Relationship Between Monotonic and Cyclic Liquefaction As pointed out earlier, the two existing schools of thought regarding liquefaction assume monotonic and cyclic liquefaction to be two fundamentally independent and unrelated phenomena. Initial liquefaction is believed to be induced when, under cyclic loading, the generated excess pore pressure becomes equal to the confining pressure. On the other hand, monotonic liquefaction is considered to occur when the driving shear stresses exceed the steady-state shear strength of the soil. In order to investigate possible relationships between monotonic and cyclic liquefaction, the effective stress paths obtained from various static and cyclic tests were examined. Figure 21 illustrates the effective stress paths obtained from static and cyclic triaxial tests on specimens with relative density of 20% and consolidation pressure of 30 psi. The hatched line shown in this figure, passing through the peak point of the static ESP, has been called the critical stress ratio (CSR) line (Ishihara et al., 1975; Vaid and Chern, 1983; Ishihara, 1985). According to theses authors, the CSR line represents the onset of the contractive response, in other terms limited initial liquefaction. The excess pore pressure starts to rise largely once the soil state reaches the CSR line. The soil would then undergo initial or limited 93 p* (psi) and ^ fnr Monotonic < stress Paths to\, hrd Sand ,, Elective Tests on the H* Figure 2 1 - “ Jolic Triaxrei 94 liquefaction which could ultimetely result in complete liquefaction or collapse. It has to be pointed out that the critical stress ratio line is different from the failure (steady-state) line which is represented by the linear portion of the monotonic effective stress path. Comparisons between the effective stress paths obtained for monotonic and cyclic triaxial tests under different confining pressures were established. These results are presented in Appendix C. The tests performed under confining pressures of 30 and 40 psi confirm the existence of a CSR line which is clearly distinct from the failure line. However, this observation is not completely verified by the results obtained from the other two tests, that is for consolidation pressures of 10 and 20 psi. One possible explanation could be that under a relative density of 20%, a low confining pressure, such as 10 to 20 psi, produces a specimen which is highly liquefiable. For a specimen with such initial conditions it would be extremely difficult to distinguish a transformation of state from limited liquefaction to collapse. Chapter 5 DEVELOPMENT OF AN EQUIPMENT AND NEW TESTING PROCEDURES FOR EVALUATION OF STATIC AND CYCL’C LIQUEFACTION RESISTANCE OF SANDS As discussed in chapter 1, one of the objectives of this study was to try to evaluate the effect of the laboratory applied stress path on static and cyclic liquefaction resistance of the HRD sand. In order to achieve this objective, a hollow cylinder cell equipment, which allows perfoming static and cyclic cavity expansion tests, was designed, fabricated and calibrated. Furthermore, a new interpretation procedure for static cavity expansion test results was developed and verified. This procedure was then used to defined the cyclic stress ratio for the cyclic cavity expansion tests. This chapter presents detailed descriptions of the apparatus, the new test interpretation procedure and its verification, and new definition of the cyclic stress ratio for cyclic hollow cylinder cell tests. 5.1 Hollow Cylinder Cell Test: An Overview Among the in-situ testing technique presently available, pressuremeter (PMT) and dilatometer (DMT) tests are used increasingly in determining the engineering properties of in- situ soils. One of the major advantages of these tests over other available in-situ techniques, such as SPT and CPT, is the well defined stress path to which the soil is subjected, 95 96 that of an expanding cavity. The interpretation procedures for PMT and DMT tests have been based on the theories of cylindrical or spherical cavity expansion. Therefore, in order to study the soil response to cavity expansion a number of investigators (Schwab and Dormieux, 1985; Juran and Beech, 1986; Juran and BenSaid, 1987) have developed hollow cylinder cell (HCC) equipment. As shown in figure 22, use of a HCC equipment allows application of three independent stresses, axial stress, lateral confining stress, and radial cavity pressure, to annular soil specimens. During the test, the cavity pressure, cavity volume change, and excess pore pressure or soi1 volume change are monitored. A major consideration in the design of the hoi low cylinder cel 1 equipment is usually minimizing the boundary condition effect. In order to achieve this objective, one needs to determine the effect of the internal to external diameter ratio on the observed soi1 response, specifically the obtained effective stresses. This in turn requires an appropriate test interpretation procedure and the use of a realistic soil model. The available interpretation procedures for cavity expansion, or PMT, tests on soils have involved use of numerical methods (Carter et al . , 197 9; Banerjee and Fathallah, 197 9; Nahra and Frank, 1985; Nahra, 1985; Baguelin et al., 1986; Lassoudiere and Zanier, 1986) and analytical solutions based on different soil models (Gibson and Anderson, 97 POROUS STONE / ) SECTION A-A' DRAIN ANNULAR SPECIMEN f------1 ^BACK PRESSURE } PORE WATER J PRESSURE TRANSDUCERS Figure 22. Schematic of the Hollow Cylinder Cell Apparatus (BenSaid, 1986) 98 1961; Ladanyi, 1963; Baguelin et al., 1972, 1978; Palmer, 1972; Vesic, 1972; Wroth and Windle, 1975; Prevost and Hoeg, 1975; Denby and Clough, 1980; Jewel 1 et al. , 1980; Fahey, 1986; Juran and Beech, 1986). Hughes et al. (1977), Wroth (1982), and Robertson and Hughes (1986) have proposed closed form solutions to determine the dilatancy properties and peak plane strain friction angle of sands. These solutions assume that sands are 1inearly elastic-perfectly plastic and the plastic flow is defined by a constant rate of dilation. In these analyses, the stress ratio-dilatancy relationship proposed by Rowe (1962) is used to define the dilatancy angle, and an independent determination of the constant volume friction angle (<|>cv ) is required. The main drawback of these analyses lies in the assumed soil model for sands. It has 1 ong been recognized that dense sands undergo initial contraction and strain hardening prior to the peak principal stress ratio fol1 owed by a post peak strain softening. At the peak principal stress ratio, the volume change is approximately proportional to the shear strain and the dilatancy rate attains its maximum. However, at larger strains the sand wi11 approach the critical state, or plastic flow under constant volume. Therefore, the dilatancy angle of the sand depends not only on its relative density but also on the strain path and on the shear strain level. The post-peak strain softening can significantly affect the response of the in-situ soil to cavity expansion. Hence, the 99 assumption of a rigid plastic behavior associated with a constant dilation rate made by Hughes et al. (1977), Worth (1982), and Robertson and Hughes (1986) appears to be too restrictive for a rational interpretation of the pressuremeter test. Moreover, the interpretation procedure proposed by these authors cannot be used to derive the volume change properties and shear strength characteristics of loose sands which undergo contraction during shearing, except under extremely low confining stresses. In addition, as a limit state of stress is assumed in this analysis, the low stress level engineering properties of sands, such as shear modulus, cannot be obtained. In order to address the need for a more adequate interpretation procedure, a new approach is proposed which permits the determination of the shear strength characteristics, dilatancy properties, and shear modulus of sands from the analysis of pressuremeter test results. A relatively simple soil model has been developed which takes into account the contracting / dilating behavior of the soil during cavity expansion. The soil is assumed to be an elasto- plastic material with a non-associated f1ow rule. 5.2 The Proposed Interpretation Procedure In this section an analytical procedure for the interpretation of cavity expansion tests (CET), or PMT tests, in sands is proposed. The procedure is based on a rheological model developed by Juran and Beech (1986) considering the soil 100 as a homogeneous, isotropic, and strain hardening elasto- plastic material with a non-associated flow rule. In this model elastic strains during shearing are neglected, and the following constitutive equations are considered: Yield function. A Mohr-Coulomb type yield criterion is considered: F(Oi j ,y) = t/s' - h(y) = 0 (1) where o i j is the stress tensor, t = ( a \ - o'3 )/2 is the deviatoric stress, s' = ( o ' l + o'3)/2 is the average effective stress, o ’x and o'3 are, respectively, the major and minor effective principal stresses, and h(y) is the strain hardening function relating the actual yield surface to the current strain state. The strain hardening and the post-peak strain softening are assumed to be isotropic with the deviatoric strain (y = ex - e3) being the hardening parameter. For 1oose contracting sands, a hyperbolic strain hardening function, of the type proposed by Kondner (1963), with two material constants is used: h(y) = y / (a + by) (2) For triaxial test: a = o0/G; b = 1/sin For dense dilating sands, it is assumed that the 101 hardening function h(y) is parabolic and can be written as: h(y) = cy(y - a) / (y + b)2 (3) where the constants a, b and c are determined from the following conditions (Figure 23): (1) the initial tangent modulus of the h(y) function is equal to G/o0, (2) at peak stress ratio, h(y) = sin 4>'/ and (3) at the critical state, h(y) = sin a = - 4(o0/G) [sin2 4>' • r2 )]/sin 4>cv b = 2(o0/G) [sin c = sin <|>cv with r = 1 + [1 - (sin <|>cv/sin Plastic potential function. The following linear stress- dilatancy relation (Nova, 1979) is considered: t| = sin v = devp = (l/uB ). [sin<|>cv - t/s’] (5) where: devp = plastic volumetric strain increment dyp = plastic deviatoric strain increment v = dilation angle and u, is a correction modulus defined as: 102 (in max OILATING CONTRACTING DILATING CONTRACTING ,min 6/P, d€, (a) dx OILATING CONTRACTING DILATING + « min CONTRACTING ( b ) Figure 23. Constitutive Soil Model (Juran and Beech, 1986) 103 rUi when t/s' < sin Anon-associated plastic potential function Q(t,s') = 0 can be derived, assuming coincidence of principal axes of stresses and plastic strain increments. Figures 23(a) through 23(c) show schematically the strain hardening function, h(y), the associated volumetric strain response, and the corresponding stress ratio-dilatancy rate function f|(t/s') for both contracting (loose) and dilating (dense) sands. The maximum plastic dilatancy rate * = max(d£v/dy) is approximately equal to the slope of the volumetric strain-shear strain curve at the peak of the h(y) function, whereas the dilatancy rate at the critical state is equal to zero. The proposed soil model requires five parameters: G/o0, 4>\ «f>cv / and u2 which, as illustrated in figures 23(a) and 23(c), can be determined from the analysis of conventional triaxial test results. It should be pointed out that these sand characteristics are dependent on the applied confining pressure and initial relative density. Therefore, average characteristics of a sand compacted to a given relative density are used to represent its behavior in each specified range of confining stresses. Using this soi 1 model, a procedure for the interpretation of pressuremeter tests in sands is proposed. This procedure is 104 derived from the solution to the cylindrical (plane strain) cavity expansion through the analysis of compatibility and radial equilibrium conditions. For a cylindrical cavity expansion, current strains are given by: er = - dx/dr ; ee = X/r ; e8 = 0 (6) where er, e8 , and e, are the radial, circumferential, and vertical principal strains, respectively, and X is the radial displacement. For the purpose of this analysis, the ratio of the volumetric strain (ev = et + e8 + es ) to the deviatoric strain (y = er - ee) is defined by the parameter N: N = sin f = ev/y (7) Hence, et/e8 = - (1 + sin f) / (1 - sin *) (8) and dy = - 2dee / (1 - sin v) (9) The parameter N can be related to the dilation angle (v) as: sin v = sin f + y • d (sin f)/3y (10) The compatibility condition implies: er = ee + r .(dsQ/dr) (11) Substituting relation (8) into equation (11) yields: 105 8e0/dr = - (1/r). (2e0 /1-sin f) (12) The radial equilibrium equation can be expressed as: 8oc/8p + (or-o0)/p = 0 (13) where p is the actual radius (Eulerian coordinate), and or and oe are, respectively, the radial and circumferential stresses. Combining equations (12) and (13), the shear curve can be incrementally derived as: t = (ot -oe )/2 = (8os/8e0 ). (e0/1-sin f). (p/r). (dr/dp) = (8oc/8e0 ). (e0/1-sin f). (l+e0/l-e0 ) (14) This expression is independent of any assumption regarding the constitutive equation of the soil. At the cavity facing the radial stress, or , is equal to the applied cavity pressure, oc . The average effective stress at the cavity facing is given by: s ’ = (p0 + Aoc ) - t - Au (15) where Au is the excess pore water pressure generated during the expansion and p0 is the initial cavity pressure corresponding to the at-rest stress state in the in-situ soil. It should be mentioned that the constitutive equations of the soil (i.e. the yield function, t/s' = h(y)/ and the stress ratio-dilatancy relationship, dev p/dyp = t|(y)) are intrinsic and are therefore independent of the applied boundary conditions. To simulate a cavity expansion test, the following 106 incremental procedure is used: 1. For a specified increment of displacement x at the cavity facing, the dilation angle (v) is calculated from equation (5), and the deviatoric strain increment (dy) is determined from equation (9). 2. The strain ratio (N) is determined using equation (7). 3. The deviatoric stress (t) and the average effective stress (s') are calculated from equations (14) and (15), respectively. Both the shear curve and the effective stress path at the cavity facing can therefore be incrementally obtained using this procedure. 5.3 Verification of the Proposed Interpretation Procedure In order to evaluate the proposed intererpretation procedure, it was used to simulate the results of cavity expansion tests (CET) on Fontainbleau sand reported by BenSaid (1986). Figure 24(a) shows typical results of an undrained expansion test including an expansion curve and measured excess pore water pressures within the specimen. The vertical displacements measured during the tests were reported to be very small illustrating a plane strain expansion. The mechanical properties of the Fontainbleau sand were obtained from the analysis of consolidated drained triaxial test results reported by Habib and Luong (1979). Figures 25(a) and 25(b) show the stress ratio-shear strain curves Figure24. (a) Typical Results of An Undrained HCCTest (crc - Po), Au (kPa) CL o -100 b° X i 200 400 300 00 2 4 0 5 3 (b) Derivationof Model Parameters bythe AIY (%) 0 V / V A E G N A H C E M U L O V CAVITY HyperbolicTransformation (Kondner, 1963) ET NO.TEST4 40 Au 80 (a) 120 30 160 40 200 107 .8 .6 *cv= 3 2 -5 \ <£> 8 3 7 ® < .4 in .2 G /o i = 6 0 0 3 DILATION RATE 17sd«v'dr Figure 25. Numerical Simulations 2 of the Triaxial Tests on Fontainbleau Sand a--*— * 0 0 5 10 15 20 2 5 (c) SHEAR STRAIN , y (%) 108 109 [ t/s' = h(Y)3 and the volumetric versus deviatoric strain curves [ev = f(y)] obtained under confining pressures of o0 = 100, 300 and 600 kPa. They illustrate the effect of confining pressure on the peak shear strength characteristics and volume change properties of this sand. Figure 25(c) shows that in spite of significant differences in volume changes measured in these three tests, the confining pressure has a rather minor effect on the stress ratio-dilatancy relationship T)(t/s) which can be approximately represented by a bilinear curve with a contractancy modulus pL, and a dilatancy modulus p2 • The experimental curves are compared with numerical simulations using the proposed soil model with the following mechanical properties: for 0q = 300 kPa: G/o0 = 60, Uj = 2, u2 =0.85 for o0 = 600 kPa: G/o0 = 60, «§>' = 37°, <|>cv = 32.5°, Pi = 1.7, p2 = 1.0 As shown in figure 25, the model predictions agree fairly well with the experimental results. For the sake of analysis, the fol1 owing average soi1 properties are considered: G/Oq = 60, <|>' = 38.5° , «f>cv = 32.5°, px = 1.8, p2 = 0.9 The hoi low cylinder cel Is with internal to external radius ratios, r0/R, of 1/5 and 1/10 were used in these tests in order to evaluate the effect of the boundary conditions on the constitutive equations of the sand. The test variables are indicated in Table 3. Elastic- Test Linear Selected Cavity Perfectly Triaxial Characteristics Regression Values Expansion Plastic R/r0 PQ (kPa) a b a b a b G/P0 G/°o 270 .006 .089 Oil .072 .006 .084 167 10 220 .013 -050 018 .110 .018 .110 56 60 170 .011 .077 012 .230 .011 .120 91 5 120 .023 .131 014 .330 .018 .170 56 Table 3. The Parameters Used for the Numerical Simulation of the Cavity Expansion Tests 110 Ill Numerical Simulation of CET Results. Figure 26 illustrates the results of the cavity expansion tests reported by BenSaid (1986). Derivation of the stress-strain curves using the proposed interpretation procedure requires incremental deter-mination of the slope of the expansion curve; d ( a c - p0)/3e0. The accuracy of this derivation process was found to be highly dependent upon the data selection and input procedures. In order to minimize the effect of the data input procedure on the derived soil properties, an extrapolation function of the type proposed by Kondner (1963) was considered: oc - p0 = e0 / (a+be8) (16) where: a = 0(oc - po )/0ee as e0-*O is the initial tangent of the expansion curve, and 1/b = (oc - p0 ) as ee-co is the limit cavity pressure, pL . As shown in figure 24(b), these two constants can be determined from a linear regression analysis of the experimental data, presented as: 1/(oc - p0) = a.(l/e0 ) + b (17) Alternatively, values of a and b can be determined as follows: (1) Determination of a - It is assumed that at the Figure 26. The CavityExpansion Test(BenSaid, Results 1986) CAVITY P R E S S U R E ( crc - Po)( kPa ) 1000 0 0 5 0 0 4 0 0 3 200 0 0 6 0 0 7 0 0 8 0 0 9 100 0 VOLUME CHANGE, AV/V0(%) 0 V / V A , E G N A H C E M U L O V Y T I V A C 10 0. kPa 909. 833. kPa 833. 588. kPa 588. .190 kPa 20 112 0 3 113 initial stage of expansion, soil response is linear elastic. Hence: a = p0/G, where G is the shear modulus. (2) Determination of b - It is assumed that at failure, the sand is at a limit state of stress defined by the Mohr- Coulomb failure criterion. Hence, ot/o8 = K. = (1 - sin «f»’)/ (1 + sin #') (18) Substituting equation (18) into the radial equilibrium equation (13), the expression for the limit cavity pressure is obtained as: PL = P0 -[(R/r0 )1_K0 - 1] = 1/b (19) where R is the external specimen radius and r0 is the internal cavity radius. It should be pointed out that equation (19) requires a pre-estimation of the peak friction angle «j>' and provides only a preliminary estimate of the b value. This equation is derived assuming a rigid plastic behavior and therefore the post-peak strain softening of the sand is neglected. This leads to an overestimation of the limit cavity pressure, PL. The two procedures described above were used to obtain the range of probable values for the a and b parameters. A curve fitting procedure was then used to select the most appropriate values for the purpose of analysis. The parameters obtained by each procedure and the selected values are summarized in Table 3. Figures 27(a) and 27(b) i1lustrate a comparison between 114 the stress-strain-volume change response curves of the sand derived from triaxial and cavity expansion tests conducted under confining pressures of o0 = 100 kPa and P0 = 120 kPa. Figures 27(c) and 27(d) show a similar comparison for the tests conducted under o0 = 300 kPa and P0 = 270 kPa. It should be pointed out that triaxial tests on dense sands generally yield a dilatancy modulus (u2 ) of about 1.0 (Baldi and Nova, 1981; Nova, 1982). The value of the contractancy modulus (pij ) is more difficult to determine. Therefore, in order to assess the effect of the latter parameter on the derived soil properties, the numerical simulations were conducted with l/jij values of 0 and 1. These simulations showed that the value has a relatively smal 1 effect on the values of the peak friction and dilation angles. Variation of l/ux from 0 to 1 results in an increase of of about 44 and an increase in v of about 3°. Furthermore, variation of u2 within this range has practical ly no effect on the value of the normalized shear modulus (G/P0). Based on these observations, the value of 1/uj = .50 obtained from the triaxial tests (Fig. 25) was retained for the purpose of analyzing the cavity expansion test results. Figures 28(a) and 28(b) illustrate the stress-strain- volume change response curves of the sand obtained from the analysis of the four cavity expansion tests described in Table 3. These results indicate that the sand properties derived using the proposed interpretation procedure (i.e., Figure27. Comparisonof theExperimental ResponseCurves VOLUMETRIC STRAIN. *J%) STRESS RATIO, t/s 6 0 0.2 0.3 7 0 0.4 0.5 0.0 0.6 -I 4 0 3 2 fc) fc) 1 2 3 0 10 0 30 20 10 0 (a) forthe HCCTests for the Triaxial Tests with Numerical Simulations l/Mi» 1-0 DEVIATOR 1C STRAIN, Figure28. Numerical Simulations of the Stress Ratioand VOLUMETRIC STRAIN e„ (%) STRESS RATIO t/s theVolumetric Strain for the CET Tests N I A R T S C I R O T A I V E D EITRC N I A R T S DEVIATORIC 0 0 30 20 10 /?✓ -inf ° 20 ul a Test No.a ul 2 --V = e-o- Ts No.Test 4 Y Test No> No> 1 Test (%) (%) 30 _L 116 117 maximum dilation angle, peak and residual friction angles, shear modulus) are not affected by the geometric boundary conditions, governed by the R/r0 ratio. The engineering properties of the sand derived from these expansion tests are compared in figure 29 with those obtained from the triaxial tests. These comparisons confirm that the applied stress path affects the soil response thereby yielding different properties: the plain strain cavity expansion tests result in higher peak friction angle (#'), and dilation angle (v), as compared to the triaxial compression tests. They also exhibit a more pronounced post peak strain softening. These observations are consistent with the results reported by Bishop (1971) and Marachi et al. (1981) comparing the stress- strain response of sand specimens during plane strain and triaxial tests. Figure 29 shows that as the applied confining pressure during the expansion is increased from 120 to 270 kPa, the peak friction angle is decreased from 48° to 44°, while the dilation angle is decreased from 13° to 10°. The triaxial tests show similar trends. It is also of interest to note that, as indicated in Table 3, the average normalized shear modulus (G/P0 = 68) value obtained from the cavity expansion tests under P0 of 220, 170, and 120 kPa corresponds fairly well to that (G/o0 = 60) determined from the triaxial tests. Figure 30 shows the effective stress paths at the cavity facing obtained from the interpretation of the four expansion Figure 29. Effect of the Applied Stress Path on Soil Properties DILATION ANGLE v (degrees) PEAK FRICTION ANGLE <£p (degrees) 0 4 0 3 5 4 35 0 5 10 15 0 5 Po, cro E R (kPa) U S S E R P G N I N I F N O C G N I N I F N O C Cavity Expansion • Tests Triaxialo Tests ® Cavity Expansion Tests Triaxial Testso 00 200 300 400 500 600 0 6 0 0 5 0 0 4 0 0 3 0 0 2 0 10 0 200 300 400 500 600 0 6 0 0 5 0 0 4 0 0 3 0 0 2 100 E R U S S E R P (a) Po, cro (kPa) cro Po, 118 Figure30. Numerical Simulation of Effective Stress Paths DEVIATORIC STRESS, t(kPa) 200 300 500 400 100 VRG EFCIE TES S (ka ) (kPa S' STRESS, EFFECTIVE AVERAGE a Ts No. Test 3 □ No. Test 2 ^ No. I Test o et No. Test 4 for the Cavity Expansion Tests 100 200 300 0 500 400 cv 119 120 tests. The peak shear strength envelope yields a 4>,p®Sk value of about 46*. At larger strains, the effective stress paths approach the critical state failure envelope ( 5.4 Evaluation of The Boundary Condition Effect As pointed out in section 5.1, above, in order to design the hoi low cylinder cel 1 the effect of geometric boundary conditions, more specifically the R/r0 ratio, on the obtained cavity expansion test results had to be determined. In order to achieve this objective, a parametric study involving both total and effective stress analysis procedures were performed (Juran and Mahmoodzadegan, 1990). For the total stress analysis, the procedure proposed by Baguelin et al. (1978) was adopted. This procedure involves assuming a unique shear curve, t = F(ee ), which can be determined as: t = (l+2g0) . g0 . d(oc-P0)/dg0 (20) where g0 ~ e0 = (-T3/3)y. Figure 31 illustrates the comparison between predictions of the total stress analysis procedure proposed by Baguelin et al. (1978) and the effective stress procedure developed in this study. This comparison indicates that the obtained shear curve by the total stress anlysis procedure is noticeably affected by the geometric boundary conditions, governed by the R/r0 ratio. On the other hand, the effective stress analysis interpretation procedure developed in this study provides a unique soil response to cavity expansion. 121 q(kPa) 50 a a a aaa 40 & R/r0 = 50 •••••• . • ••••• R/r0 = 10 30 o o oo °°00 R /r0 s 5 20 j Total Stress Analysis - Effective Stress Analysis Figure 31. Effect of the Boundary Conditions and the Analytical Procedure on the Soil Response in Cavity Expansion Test 122 5.5 Definition of The Cyclic Stress Ratio For CET Tests A major difficulty in evaluating the effect of the applied stress path on cyclic liquefaction resistance of soils is the appropriate definition of the applied cyclic stress ratio (CSR). As discussed in chapter 3, CSR has been defined as the ratio of the applied cyclic deviatoric stress to the consolidation pressure for cyclic triaxial tests. In an attempt to define an equivalent definition of CSR for cyclic simple shear tests, Peacock and Seed (1968) examined four alternative stress ratios: (a) the maximum ratio of the shear stress developed during cyclic loading to the normal stress during consolidation on any plane in the sample, (b) the maximum ratio of the change in shear stress on any plane during cyclic loading to the normal stress on that plane during consolidation, (c) the ratio of the maximum shear stress induced in a sample during cyclic loading to the mean principal stress on the sample during initial consolidation, and (d) the ratio of the maximum change in the shear stress on any plane during cyclic loading to the mean principal stress on the sample during consolidation. They further reported results of cyclic triaxial and simple shear tests, on specimens of Monterey sands under various initial conditions. These results indicated that for specimens prepared at a relative density of 50% and consolidations pressures of 1 to 8 kgf/cm2 , and subjected to the same value of CSR, the cyclic liquefaction resistance of the triaxial test specimens were 123 consistently higher than those of simple shear test. The ratio of the two cyclic strengths was reported to increase with increasing confining pressure and ranged from about 1.75 at a confining pressure of 1 kgf/cm2 to about 2.65 at 8 kgf/cm2 . Recently, for cyclic 1iquefaction tests in a cubical shear device, Clough et al. (1989) defined the cyclic stress ratio as the ratio of the maximum octahedral shear stress to the isotropic confining pressure, xoctj Bax/ooct . In order to analyze the effect of the applied stress path on static and cyclic response liquefaction potential of the HRD sand, static and cyclic liquefaction tests were performed in triaxial and hoi 1ow cylinder cel 1 equipment. The static liquefaction tests were strain-controlled and no difficulty in the test interpretation was encountered. However, in order to perform the cyclic hoi low cylinder cel 1 tests, an appropriate definition for the applied CSR was required. To adequately address this issue, the foil owing methodology was adopted: 1. An effective stress analysis interpretation procedure was developed which allows simulation of cavity expansion and triaxial tests on loose or dense specimens. 2. The proposed effective stress analysis procedure was verified in two steps: (a) results of triaxial CD and undrained cavity expansion tests on dense specimens of Fontainbleau sand were simulated (Fig. 27). These results indicate that the stress ratio, t/s', and volume change response of soils are essentially stress path independent and 124 can be uniquely determined using the propsed numerical procedure, (b) results of triaxial CD, CU, and undrained cavity expansion tests on specimens of a normally consolidated silty clay (Juran and Mahmoodzadegan, 1990) were simulated (Fig. 32). The results shown in figure 32 indicate that the proposed interpretation procedure can also be used to predict the response of loose soils to loading. The results shown in figure 32 are important not only because they illustrate the stress path effect on the soil response, but also because they establish the framework for definition of an appropriate CSR for cyclic cavity expansion test. 3. Figure 33 illstrates an idealized version of the results shown in figure 32. This schematic presentation is based on the principle of normalisation (Roscoe et al. , 1958; Schofield and Wroth, 1968), which essentially states that the effective stress paths for specimens with similar stress history and initial void rario would be geometrically similar, for CET tests. In order to define the CSR for cyclic cavity expansion tests, considering the results shown for the tests with P'0=20, one should recall that the applied cyclic stress ratio for the cyclic triaxial test is defined as qa/2P'B, where P'a and qa are, respectively, the mean effective and deviatoric stress at point A . 4. In order to define the CSR for the cyclic cavity expansion test under P'0=20, the coordinates of point B were calculated using the effective stress procedure described above. The 125 q / p Triaxial CD Tests Triaxial CU Tests Cavity Expansion Tests sin £ cv 0.5 (o) d « 0.5 (a) q(kPa) Effective Stress Paths — — Triaxial CD Test —— Triaxial CU Test ►----- Cavity Exponsion Test 100 (b ) p'(kPa) 100 3 0 0 lb) Figure 32. Numerical Simulations of Triaxial CU, CD, and CET Tests Using the Proposed Algorithm Figure 33.of CSRforFigure DefinitionCylinder CellHollow Tests cr (psi) alr Line Failure P* i Collapse Surface 126 127 cyclic radial cavity pressure to be applied, ortCTC, was then determined from equation 15, section 5.2. 5. Using the above procedure along with the results of triaxial monotonic liquefaction tests (Dr=20%) presented in chapter 4, the cyclic cavity pressures for cavity expansion tests were determined: o ' c (Psi) of l C jc (Psi) 10 4 20 7 30 11 40 16 The above values of or>cyc were used for the cyclic hollow cylinder liquefaction tests, to be discussed in the next chapter. Chapter 6 USE OF HOLLOW CYLINDER CELL EQUIPMENT FOR EVALUATION OF STATIC AND CYCLIC LIQUEFACTION RESISTANCE OF THE HRD SAND The steps involved in the development and calibration of the hollow cylinder cel1 (HCC) were discussed in chapter 5, along with a discussion of the test interpretation procedures. The developed HCC equipment was integrated into the MTS cyclic loading frame, along with a servo-controlled lateral pressure system, in order to perform monotonic and cyclic cavity expansion tests. This pressure system, fabricated by Material Testing Systems, Inc., allows application of monotonic or cyclic cavity pressure independently from the axial stress and cel1 (confining) pressure. Using this integrated equipment, monotonic and cyclic liquefaction tests were performed on specimens of HRD sand prepared under different initial conditions. The results of hoi 1ow cylinder cel1 monotonic and cyclic liquefaction tests are presented in this chapter. 6.1 Monotonic HCC Liquefaction Tests The purpose of these tests were to: (a) determine the effect of testing variables, specifically consolidation pressure and relative density, on the static liquefaction resistance of the HRD sand as determined in the HCC equipment, (b) determine whether the steady-state and/or collapse surface concepts can be applied to the stress path conditions imposed 128 129 by the HCC device, and (c) provide, along with the results of the HCC cyclic liquefaction tests, the necessary data base for studying mechanisms of monotonic versus cyclic liquefaction. Sensitivity Analysis. The HCC device developed for the purpose of this study allows performing monotonic and cyclic cavity expansion tests on annular (hollow) soil specimens. Use of this new equipment for evaluation of static and cyclic liquefaction potential of the HRD sand required an evaluation of the reproducibility of the test results. In order to meet this objective, a sensitivity analysis program involving several static tests on specimens prepared under different initial conditions were carried out. Figure 34 presents typical results of three monotonic HCC liquefaction tests performed on specimens of the HRD sand under an initial relative density of 20% and consolidation pressure of 10 psi. The test results obtained during the sensitivity analysis program, including but not limited to those reported in figure 34 indicated an adequate degree of reproducibi1ity. Data Acquisition and Control. The test variables measured during a HCC test were described in chapter 5. The data acquisition system consisted of a 12 channel data 1 ogging unit, an A/D board and an a personal computer with a 286 microprocessor. During the test, analog data were continuously sent from the instrumentation to the acquistion system. The analog data were intercepted at the A/D interface, converted Figure34. Reproducibilityof the MonotonicLiquefaction Cavity Pressure (psi) 20 10 25 15 0 5 Tests inthe Hollow Cylinder Cell Apparatus 5 aiy oue hne (%) Change Volume Cavity 0pi ; Dr-20% psi 10 20 2515 130 131 into digital readings and collected and stored by the computer thereby minimizing human errors involved in monitoring the test results. Specimen Compaction Technique. As mentioned in chapter 2, a standard specimen compaction technique for laboratory, monotonic or cyclic, liquefaction tests does not presently exist. The compaction technique proposed by Canou (1989), discussed in chapter 4, was used throughout this study. Saturation Technique. The liquid C02 perculation and back-pressure saturation techniques, discussed in chapter 4, were used throughout this testing program. Specimen Preparation Procedure. The specimen preparation procedure for hoi low cylinder cel 1 tests is quite involved and requires a great amount of practice. This procedure involves the following steps: 1. A preweighed amount of sand was mixed with distilled water at a proportion of 4% by weight. 2. A three-piece split mold, for 8” x 4" specimens, was assembled on the cel 1 base. A latex membrane 4" in diameter was placed inside the mold, two O-rings were placed at the top and bottom ends of the mold, and a vacuum of about 6 psi was applied to the membrane through an iniet to control the compaction during the specimen placement stage. A second membrane equal in diameter (0.4") with the internal cavity was placed around a 10" long rod. The lower end of the rod was machined down to the size of the pressure ini et for the 132 application of the cavity pressure. The lower end of the 0.4" membrane was fixed around the stem of a bolt with the help of an O-ring. The stem of the bolt was drilled to allow for the passage of cavity pressure. The lower, machined, end of the rod was piaced into the drilled stem of the bolt. 3. The lower end of the 4" membrane was fixed around the lower end cap using two O-rings. After placing a porous stone on the bottom end cap, the sand was placed in layers inside the mold using a spatula. A great amount of care had to be taken to place the soi1 in the area between the mold and the rod. To achieve a higher specimen uniformity, a minimum of 20 layers were used. 4. After the specimen reached the desired height, with the vacuum still being applied to the membrane, the top porous stone and the cap were gently put in place. The upper end of the 4" membrane was fixed around the outer perimeter of the cap using two O-rings. The upper end of the 0.4" membrane was then placed aroung the projected area in the center of the top cap, and the rod was pul led out of the cavity. In order to cover the top cap central hole, to seal the cavity, a circular assembly cap was designed which was screwed onto the top end cap after the rod was pul led out. 5. A vacuum of about 5 psi was applied to the specimen through a pressure iniet at the base, the three-piece mold was gently disassembled and the specimen dimensions were carefully determined using a vernier caliper. It should be pointed out 133 that despite the fact that vaccum was applied to the specimen while the mold and rod were being removed, due to several intermediate stages involved, specially placement of the top cap assembly, it was essentially impossible to achieve relative densities below 20% by this procedure. 6. The hydraulically operated pressure cel 1 was lowered onto the base, the cel1 was filled with de-aired water, confining and cavity pressures of 5 psi were applied, and the vacuum was disconnected from the base of the cel 1, 7. Liquid C02 was perculated through the specimen for 10 to 15 minutes under a low pressure gradient of about 2 psi. 6. Using the standard incremental back-pressure saturation technique (Head, 1986), the sample was saturated. Care was taken to increment the confining and the cavity pressures equally and at the same time. This was made possible through a special feature on the pressure panel. 7. After the sample was saturated, determined by reaching a 13- coefficient vaue of at least 0.98, the back-, confining, and cavity pressures were reduced and the sample was al1 owed to consolidate under an effective consolidation pressure of predetermined value. During the saturation and consolidation stages the axial specimen deformation was accurately monitored but generally no appreciable reduction in specimen height was recorded. 8. After the consolidation stage was completed, determined by a condition of zero excess pore pressure maintained over a 134 time period of at least 2 hours, the drainage valve was closed, and the sapmle was sheared to failure. This was achieved by application of a constant rate of volume change to the cavity. Cavity Volume Chance Rate. Monotonic cavity expansion liquefaction tests had not been reported prior to this study, and therefore, the effect of the cavity volume change rate on the soil response had to be evaluated. Figure 35 shows the results of a sensitivity analysis performed to address this need. These results indicate that a change in the rate of application of cavity volume from 0.2% to 1.0% per minute would not practically affect the soil response. Therefore, a rate of cavity volume change of 1.0% per minute was used in all monotonic tests. Correction for Membrane Penetration. As discussed in chapter 4, for the fine HRD sand with Dso=0.13 mm no membrane penetration correction was necessary. Effect of Testing Variables; A Parametric Study A testing program involving 9 monotonic HCC liquefaction tests on specimens of the HRD sand, prepared under different initial conditions, was carried out. The purpose of these tests were to: (1) evaluate the effect of consolidation pressure and relative density on the monotonic liquefaction resistance of the HRD sand, (2) establish the necessary data base of static HCC liquefaction tests for evaluation of the stress path effect. The second objective involved: Figure35. Effect of Cavity Volume Change Rate onthe Soil Cavity Pressure (psi) 10 20 15 25 0 5 0 Response in UndrainedCavity Expansion Tests 5 aiy oue hne (%) Change Volume Cavity .%pr minute per 0.5% .%pr minute per 0.2% .%pr minute per 1.0% 10 a'c-10 ;psi Dr=20% N 525 15 20 135 136 (a) comparison of the static and cyclic HCC liquefaction test results to study the liquefaction mechanism in each case, and (b) comparison of static HCC test results with those obtained from monotonic triaxial tests to evaluate the validity of the available interpretation procedures, specifically the steady- state line approach, for the stress path during a monotonic HCC test. Effect of Relative Density. As shown in table 1 in chapter 3, the in-situ relative density and effective overburden pressure for the HRD sand were about 20% and 10 psi, respectively. In order to determine the effect of relative density on the monotonic liquefaction resistance of the HRD sand, 5 tests were performed on the specimens prepared under an initial consolidation pressure (P0) of 10 psi and relative density (Dr) values of 20, 30, 40, 50, and 80%. The cavity pressure,(or-P0), and generated excess pore pressure measured during each test are shown in figures 36 and 37, as functions of the cavity volume change. The individual test results are presented in Appendix D. Interpretation of the Test Results. Figures 36 and 37 illustrate the effect of the relative density on the obtained soi1 response in terms of the cavity pressure and generated excess pore pressure. As shown in figure 36, for the specimen with a relative density of 20% the cavity pressure rapidly increases to a maximum of about 4.5 psi at a small volume change of about 3% and thereafter continuously decreases to a Figure 36. Effect of Relative Density on the Cavity Pressure Cavity Pressure (psi) 10 20 15 0 5 0 During Monotonic HollowCylinder Cell Tests c'c-10 psi c'c-10 5 aiy oue hne (%) Change Volume Cavity 10 15 20 D% “40 530 25 137 Excess Pore Pressure (psi) Figure37. Effect ofRelative Density onthe Excess Pore -2 10 0 2 4 6 8 Pressure During Monotonic Hollow Cylider Tests aiy oue hne (%) Change Volume Cavity D “20%D 25 138 139 very small value. The corresponding excess pore pressure response curve for this test, shown in figure 37, indicates that at the peak cavity pressure the excess pore pressure has not reached its maximum value, and in fact attains a value of about 8 psi. A similar observation was made in the case of triaxial static tests, discussed in chapter 4. Figure 36 indicates that the maximum cavity pressure reached during the test increases with an increase in the relative density of the specimen. However, the cavity volume change corresponding to the peak cavity pressure does not significantly change, and in fact it only reaches a value of about 4%, for specimens with Dc values of 30 and 40%. Also indicated in figure 36, is the effect of relative density on the residual cavity pressures, that is values of cavity pressures at large strain. For instance, an increase in relative density from 20% to 30% increases the residual cavity pressure by about 1.50 psi. Perhaps the most significant observation to be made from the results shown in figures 36 is the change in material response from contracting (i.e. liquefiable) to dilating or nonliquefiable. As the specimen relative density increases to 30% and higher, the post-peak softening, or decrease in cavity pressure, gradually disappears. For the specimen with a relative density of 40% there is only minor post-peak softening, and the specimen prepared at 50% relative density actually exhibits some post-peak hardening. 140 The change in the material response from contracting to dilating can be further explained through the excess pore pressure response curves shown in figure 37. The specimen prepared to a relative density of 20% represents the most liquefiable state among the five specimens tested. The excess pore pressure for this specimen reached the maximum value of the consolidation pressure, or 10 psi, indicating a condition of zero effective stress. As shown in figure 37, the excess pore pressures for the specimens prepared to relative densities of 20 and 30% are continuously increasing unti1 a certain axial strain is reached and then remain constant. It is important to notice that the maximum value of excess pore pressure reached during these tests decreased with an increase in the specimen relative density. This cl early indicates that an increase in the relative density decreases the soi 1 susceptibility to liquefaction. As the specimen relative density is further increased to 40% or higher, the material response drastically changes, in fact the excess pore pressure decreases beyond a certain value of the cavity volume change as the material response changes from 1iquefiable to nonliquefiable. Determination of The Steady-state Line. The steady-state line (SSL) approach has so far been used exclusively with the triaxial compression tests. The conventional definition of the SSL requires determination of either P' or o'3 at failure. However, during a HCC test only the cavity pressure,(ot-P0 ), 141 is directly measured, and, therefore, the choice of the stress parameter to be used for establishing the SSL is not obvious. Figure 38 illustrates a piot of the cavity pressure at failure versus void ratio for the specimens prepared to relative densities of 20, 30, 40, and 50% under a consol idation pressure of 10 psi. The specimens prepared at relative densities of 20 and 30%, as il lustrated in figures 36 and 37, did undergo monotonic 1iquefaction. However, the specimens prepared at relative densities of 30 and 40 % underwent dilation and did not liquefy. The failure points for specimens with relative densities of 20 and 30 % establish a straight 1ine. This 1ine could represent the SSL for the HCC tests. This definition is not, however, identical to that proposed by Castro (1969) for triaxial compression tests as the stress parameters for the two tests are different. The critical relative density at which the material response would change from contracting to dilating seems to be about 32 %, which is considerably lower than that observed in the case of the monotonic triaxial tests. As discussed earlier, due to the steps involved in specimen preparation, it is practical 1y impossible to prepare specimens with Dr<20% using the current procedure. On the other hand, only specimens with relative densities of 20 and 30% did undergo complete monotonic liquefaction. Therefore, it is difficult to directly compare the results shown in figure Figure38. The Steady-state Line DerivedFrom the Results of Void Ratio 0.80 .5 ■ 0.85 0.90 0.95 1.00 10’1 Undrained MonotonicHollow Cylinder Cell Tests aiy rsue t alr (psi) Failure at Pressure Cavity 1 10 0 5 = D • D =20 % 0 2 = D O 0 4 = D 102 142 143 10 with those obtained from the triaxial tests (chapter 4). Effect of Consolidation Pressure. In order to determine the effect of consolidation pressure on the monotonic liquefaction resistance of the HRD sand, 4 tests were performed on the specimens prepared to a relative density (Dr) of 20% and consolidation pressure (P0 ) values of 10, 20, 30, and 40 psi. The cavity pressure, (oc-P0 ), and generated excess pore water pressure measured during each test are plotted as functions of the cavity volume change in figures 39 and 40. The individual test results are presented in Appendix D. Interpretation of The Test Results. Figures 39 and 40 illustrate the effect of the consolidation pressure on the cavity pressure and generated excess pore pressure obtained during the four monotonic HCC tests. As indicated in figure 39, during all tests the cavity pressure rapidly increases to a maximum value at a cavity volume change of less than 4% and thereafter deccreases to a much lower value. The corresponding excess pore pressure response curves for these tests, shown in figure 40, indicate that the excess pore pressure at peak cavity pressure did not become close to the value of corresponding consolidation pressure for any of the four tests. A similar observation was made in case of the monotonic triaxial liquefaction tests discussed in chapter 4. Figure 39 also indicates that the maximum (peak) cavity pressure reached during a test increases with an increase in the consolidation pressure of the specimen. However, the Cavity Pressure (psi) Figure 39. Effect of Confining Pressure on Cavity Pressure 10 15 20 0 5 0 5 in Honotonic HollowCylinder Cell Tests aiy oue hne (%) Change Volume Cavity 10 15 20 i s p 0 1 = ' o i s p 0 2 = c ' O ' i s p 0 3 = O' i s p 0 4 - ' O 25 30 144 Figure 40. Effect of Confining Pressure on the Excess Pore Excess Pore Pressure (psi) 20 10 5 2 15 0 3 5 3 40 0 5 0 Pressure in Monotonic Hollow Cylinder Cell Test 5 aiy oue hne (%) Change Volume Cavity 10 15 20 i s p 0 2 - c ' Q i s p 0 3 = c ' G 5 2 0 3 145 146 cavity volume change corresponding to the peak cavity pressure does not significantly change. As indicated in this figure, the change in consolidation pressure also affects the residual cavity pressure, that is the value of the cavity pressure at large strain. For instance, an increase in P0 (o'c ) from 20 to 30 psi increases the residual cavity pressure by 2.30 psi. It is important to note that under a low relative density of 20%, even application of a high confining pressure of 40 psi will not prevent the specimen from undergoing monotonic liquefaction. On the other hand, as discussed above, under a confining pressure of 10 psi, an increase in the relative density to about 50% wi11 be sufficient to prevent monotonic liquefaction. These observations are of high practical relevance as far as the in-situ liquefaction susceptibility of the HRD sand is concerned. 6.2 Cyclic HCC Liquefaction Tests The.purpose of these tests were to: (a) determine the effect of testing variables, specifically consolidation pressure and relative density, on the cyclic liquefaction resistance of the HRD sand, and (b) provide, along with static liquefaction test results, the necessary data base for studying mechanisms of monotonic versus cyclic liquefaction, and (c) evaluate the effect of stress path on the cyclic liquefaction potential of the HRD sand as determined from HCC and triaxial tests. 147 Testing Equipment. To perform the cyclic HCC liquefaction tests, a servo-controlled lateral pressure system was designed and constructed by the MTS company. This system allowed for application of monotonic as well as cyclic confining and/or cavity pressures of upto 300 psi. The lateral confining pressure system was integrated into the main triaxial system to allow for the use of the same electronic control units. Data Acquisition and Control. The data acquisition hardware included a high performance 8 channel, 12 bit A/D board manufactured by Data Translation, Inc., a signal conditioning / amplifying interface, and an IBM PS/2 computer with a model 80286 (AT Compatible) Intel microprocessor equipped with a math co-processor to speed up the digital operations. The software used was Labtech Notebook Release 4.30 developed by the Cyber Research, Inc. This software al1ows a real-time access to the experimental data which permits on-screen observation of the experimental results. Specimen Compaction, Saturation, and Preparation. The same procedures described in section 6.1, above, were used. Loading Frequency and Wave Form. During all cyclic liquefaction tests in this study a sinusoidal wave with a frequency of 1 Hz was used. Applied Cyclic Cavity Pressure. Determination of the cyclic cavity pressure, o'CyC.r' was based on the interpretation procedure developed for static HCC tests. As discussed in section 5.5, this procedure yields a value for 148 °'cyc,r based on the analysis of the results of the monotonic test on a sample with similar initial conditions. Effect of Testing Variables: A Parametric Study In order to determine the effect of relative density and consolidation pressure on the cyclic liquefaction potential of the HRD sand a series of 8 tests were performed. The test results are presented and discussed in the following sections. Effect of Relative Density. In order to determine the effect of the relative density (Dr) on the cyclic liquefaction resistance of the HRD sand, a series of 4 tests were performed on the specimens prepared at relative densities of 20, 30, 40, and 50% and consolidation pressure of 10 psi. For each test, the cavity pressure and exceee pore pressure generated during the test were monitored. These results are presented in Appendix E. For the purpose of comparison, the generated excess pore water pressures for al1 four tests were shown in figure 41. Interpretation of The Test Results. The obtained cavity pressure response for the specimens with relative densities of 20 and 30%, indicate that after a certain number of loading cycles the soi 1 undergoes complete liquefaction with essentially zero residual shear strength. The pore pressure response for these two specimens, shown in figure 41, indicate that the excess pore pressures generated during the two tests did in fact reach 10 psi, the value of the consolidation pressure, thereby inducing a zero effective stress (initial Figure 41. Effect ofRelative Density onthe Excess Pore Excess Pore Pressure (psi) 10 0 2 4 6 B 0 Pressures inCyclic HollowCylinder Tests 0 3 5 aiy oue hne (%) Change Volume Cavity 10 % 0 3 = D 15 20 % 0 5 = D i s p 0 1 = c ' O 5 2 149 150 liquefaction) condition. As the specimen relative density is increased to 40%, the soil response significantly changes. The cavity pressure response indicates that a finite, non-zero residual cavity pressure of about 1.5 psi is reached at a cavity volume change of about 20%. More importantly, it is noticed in this figure that further loading cycles have practically no effect on the value of the residual cavity pressure. As indicated in figure 41, the rate of change of the excess pore pressure generated during this test starts to decrease at a cavity volume change of about 15%, reaching a practically constant value of 6 psi at about 20%. Further loading cycles does not seem to have any effect on the value of the excess pore pressure. The results obtained for the specimen with a relative density of 50% indicate that an increase in the relative density yields an increase in the residual cavity pressure. A residual cavity pressure of 5.5 psi is reached for this specimen which is considerably higher than the value obtained in the case of the specimen at 40% relative density. The residual cavity pressure obtained remains virtually constant upon application of additional stress cycles. Similarly, an increase in the relative density causes a decrease in the value of the residual excess pore water pressure. As indicated in figure 41, the residual excess pore pressure of 3 psi is significantly lower than that obtained for the specimen with 40% relative density. It is clear from theses results that an 151 increase in the initial relative density of the specimen increases its resistance to development of initial liquefaction condition. Figure 42 illustrates a comparison between the cyclic liquefaction potential of the HRD sand specimens determined from triaxial and hollow cylinder cell tests. It is important to note that, for specimens prepared under identical initial conditions, the liquefaction resistance is consistently higher in case of the hollow cylinder cell tests. The difference in the number of cycles required to induce liquefaction in the two tests increases with an increase in relative density. The above observations point out the influence of the applied stress path on the cyclic liquefaction of the HRD sand. They also imply that for the stress path induced during a cyclic cavity expansion, the initially contracting material could undergo densification which tends to increase its resistance to the buildup of excess pore pressure. Effect of Consolidation Pressure. In order to determine the effect of the consolidation pressure (o'c or P0) on the cyclic liquefaction resistance of the HRD sand, a series of 4 tests were performed on the specimens prepared at a relative density of 20% consolidated under effective pressures of 10, 20, 30, and 40 psi. For each test, the cavity pressure and excess pore pressure generated during the test were monitored. These results are presented in Appendix E. In order toillustrate the effect of the consolidation Figure 42. Effect of the Applied Stress Pathon the Cyclic No. Cycles to Liquefaction 200 100 150 Liquefaction Potentialof the HRD Sand*Dr effect ylcHC v CyclicHCC Cyclic Triaxial- Relative Density Density Relative 10 psi 10 [%) 152 153 pressure on the generated excess pore water pressures, the pore pressures for all four tests are shown in the figure 43. Interpretation of The Test Results. The obtained cavity pressure response for the specimens with consolidation pressures of 10 and 20 psi indicate that after a certain number of loading cycles the soil undergoes complete liquefaction with essentially zero residual shear strength. The pore pressure response curves for these two specimens, shown in figure 43, indicate that the excess pore pressures generated during both tests did in fact reach 10 psi, the value of consolidation pressure, thereby inducing a zero effective stress condition. As the initial confining (i.e. consolidation) pressure is increased to 40%, the soil response significantly changes. The cavity pressure response indicates that a finite, non-zero residual cavity pressure of about 2.2 psi is reached at a cavity volume change of about 15 %. It is also indicated that further 1oading cycles have practically no effect on the value of the residual cavity pressure. As indicated in figure 43, the rate of change of the excess pore pressure generated during this test starts to decrease at a cavity volume change of about 17%, reaching a practically constant value of 6 psi at about 20%. Further loading cycles seem to have no effect on the value of the excess pore pressure. The cavity pressure response curve obtained for the specimen with a relative density of 50% indicates that an Figure 43. Effect of Confining Pressure on theExcess Pore Excess Pore Pressure (psi) 20 10 15 5 2 0 3 5 3 40 0 5 0 Pressures in Cyclic Hollow Cylinder Cell Tests 5 % 0 2 = D aiy oue hne (%) Change Volume Cavity 10 psi 10 10 20psi 15 20 30psi 40psi 5 2 154 155 increase in the consolidation pressure yields an increase in the residual cavity pressure. A residual cavity pressure of 6.1 psi is reached for this specimen which is about 3 times that reached in the case of the specimen at 40% relative density. The residual cavity pressure obtained remains virtually constant upon application of additional stress cycles. As indicated in figure 43, however, an increase in the consolidation pressure results in an increase in the value of the residual excess pore pressure. This observation was also made in the case of cyclic triaxial tests discussed in chapter 4. These results clearly indicate that an increase in the consolidation pressure results in an increase in the specimen resistance to liquefaction. Figure 44 illustrates the effect of the applied stress path on the cyclic liquefaction potential of the HRD sand as determined from the triaxial and hollow cylinder cell tests. The cyclic liquefaction potential of the triaxial specimens are consistently higher than the corresponding hollow cylinder cell specimen. The difference in the number of cycles required to induce liquefaction increases with an increase in the confining pressure. 6.3 Relationship Between Static and Cyclic Liquefaction As pointed out earlier, monotonic and cyclic liquefaction potential of sandy soils have been studied independently in the past. One of the objectives of this study was to investigate possible relationships between monotonic and 156 300 250 Cyclic Triaxial T, 200 Cyclic HCC— cr _J 150 i—* o . 100 50 0 0 10 20 30 40 50 Confining Pressure (psi) Figure 44. Effect of the Applied Stress Path on the Cyclic Liquefaction Potential of HRD Sand: o'c effect 157 cyclic liquefaction potential of the HRD sand through a fundamental study of the liquefaction mechanism in either case. In order to meet this objective, monotonic and cyclic liquefaction tests were conducted in the HCC equipment developed at an initial stage of this study. The results of these tests, focused on studies of the effect of Dr and o'c, were discussed in sections 6.1 and 6.2, above. A useful means of studying the 1 iquefaction mechanism(s) during HCC tests is to compare the the cavity expansion response curves obtained during monotonic and cyclic tests. Figure 45 illustrates such a comparison for tests conducted on specimens with initial relative density of 20% and consolidation pressure of 20 psi. As indicated in this figure, at a certain cavity volume change during the monotonic test, the specimen undergoes steady-state flow, and the shear strength is reduced to a very low value. On the other hand, during the cyclic test, the specimen undergoes complete col lapse with essentially zero residual shear strength. It is important to note that the starting point of collapse on the cyclic cavity expansion response curve approximately corresponds to the point of the peak cavity pressure on the monotonic response curve. In other words, a straight line connecting the origin of the coordinate axes to the point of the peak monotonic cavity pressure would approximately pass through the point where collapse is initiated during cyclic loading. This observation is confirmed by the results obtained Figure 45. Comparison of Effective Stress Pathsin Monotonic Cavity Pressure (psi) 20 5 1 15 20 2 5 1 10 5 w and Cyclic Hollow Cylinder CellandCylinderCyclic TestsHollow aiy oue hne (%) Change Volume Cavity 158 159 for the specimens with consolidation pressures of 20, 30 and 40 psi, presented in Appendix E. It should be recalled that a similar observation was made in case of the triaxial tests, discussed in chapter 4, where it was pointed out that the point of initiation of collapse during cyclic loading was related to the point of maximum stress ratio, q/p', during monotonic tests. A similar observation is made when comparing the results of the monotonic and cyclic HCC tests conducted on specimens with a consolidation pressure of 10 psi and different initial relative densities. Chapter 7 SUMMARY AND CONCLUSIONS A comprehensive research program focusing on the in-situ and laboratory evaluation of the liquefaction potential of fine sands was carried out. The in-situ testing program, which involved cone penetration (CPT), piezocone penetration (PCPT), dual piezocone penetration (DPCPT) and seismic cone penetration (SCPT) tests, was carried out at two wel 1 - documented sites in the southern Imperial Valley, California. During the in-situ phase of the study, a fine, metastable sand layer was identified and large quantities of the soil was transferred to the Louisiana State University geotechnical laboratory. The laboratory testing program involved monotonic and cyclic triaxial tests, as well as monotonic and cyclic hollow cylinder tests. In addition a new hollow cylinder testing equipment was developed which required numerical modeling of soil response to the specific stress path of a cylindrical cavity expansion. A summary of the main findings in this study is presented in this chapter. For the purpose of better organization, the findings are categorized into four subsections: (a) In-situ Liquefaction Analysis, (b) Monotonic Liquefaction Potential, (c) Cyclic Liquefaction Potential, (d) New Hollow Cylinder Cell Equipment. 160 161 7.1 In-situ Liquefaction Analysis Several existing in-situ liquefaction evaluation procedures were reviewed, and the main advantages and dis advantages of each procedure was pointed out. Four of the most widely used liquefaction evaluation procedures were selected for a comparative study of their predictive capabilities. Two of the selected procedures are based on direct use of cone penetration test results, one procedure is based on the correlations between SPT and CPT results, and the fourth is based on the seismic cone test results. The four selected methods were used to evaluate the in-situ liquefaction potential of six sublayers of sandy deposits in the two sites. The results of this comparative study indicated that although these four procedures are based on different intrinsic assumptionsthe, they yield quite consistent predictions. As a part of this study, a large number of PCPT tests performed in the two sites to investigate the feasibility of using its results to evaluate the in-situ liquefaction potential. Also, several DPCPT tests were conducted. The results of PCPT and DPCPT tests indicated that the measured test variables can be related directly to the soil contracting or dilatancy tendency. In particular, a combination of the friction ratio, tip resistance and the excess pore pressure at the tip can be used to determine thin liquefiable soil layers which are very hard detect using any other in-situ technique. Moreover, the difference between the excess pore pressures 162 measured at the tip and along the shaft, using the dual piezocone, is indicative of the stress path dependent nature of the shear induced volume changes of the soil. Specifically, in liquefiable soil deposits pore pressures measured at the tip and along the shaft are both positive, with the value measured at the tip being smaller. On the other hand, in nonliquefiable soil layers the pore pressure measured along the shaft is either very small or negative due to the applied shear stress. The specific conclusions of the PCPT and PCPT tests are summarized below. Analysis of the effect of the piezocell location on the measured pore water pressures suggests: 1. In sandy deposits (e.g. Heber Road site location 442', 2 to 5.5 m, and Wildlife site location WSW, 4 to 6.8 m) the excess pore water pressure measured at the tip increases with the point resistance due to an increase in the normal stress. The excess pore pressures measured at the sleeve were essentially negligible. 2. The excess pore pressure measured on the cone tip is much more sensitive to the type of soil penetrated which can be correlated to the measured values of the tip resistance. 3. The excess pore pressures measured at the sleeve appears to be sensitive to the tendency of the sand to contract or dilate during shearing. In loose sand seams (e.g. Heber Road 700', at 11.6 m) positive excess pore water pressures are generated 163 both at the tip and at the sleeve (where the generated excess pore water pressures are much smaller. In medium dense sand layers which have a low dilatancy potential (e.g. Heber Road 700', 9 to 10 m) large positive excess pore pressures are measured at the tip, while the pore pressures at the sleeve are negligible. 4. In loose, potentially liquefiable sand inclusions (e.g. Heber Road 700' , at 11.6, 13.5 and 14.0 m) large positive excess pore pressures were measured at the tip, with the excess pore pressures measured at the sleeve being positive but smaller. 5. In the two sites investigated dense sand inclusions were not encountered. However, in such inclusions negative excess pore pressures would be expected at the sieeve. 6. For most cases in both clay and sand layers, the excess pore water pressures recorded just above the cone, (u8) in the seismic piezocone, and behind the sleeve, in the dual piezocone, were practical 1y the same. Therefore, for most practical cases, the excess pore pressures measured with the dual piezocone at the sleeve could be used in the classification charts based on u0 , such as that proposed by Robertson et al. (1985). 7. In some cases (e.g. test no. 17.2, at 2.5 m, 5.3 m, 8.8 m), the excess pore pressure measured just above the cone became negative, while that measured behind the sleeve was positive. This observation tends to agree with the results of numerical 164 simulations by Kiousis et al. (1988) suggesting the formation of an unloading zone at that location. 8. In light of the results obtained throughout this study, it appears that the location of the piezocell in the dual piezocone would be better suited for soil stratification and, specifically, for detection of loose liquefiable sandy layers. Analysis of the PCPT test results with different piezo cell materials indicated the significant effect of the compressibility of the porous element and demonstrated that the stiffer ceramic element is more sensitive and yields more reliable data. Analysis of the effect of the penetration rate on the measured tip resistance and the excess pore pressures at the two sites suggests: 1. In sand layers, the penetration rate most generally has practically no effect on neither the point resistance nor the friction ratio, and, therefore, would not affect the friction cone based soil classification. 2. The excess pore pressures measured at the tip are highly dependent upon the ratio of the penetration rate to the hydraulic conductivity of the soil (Juran and Tumay, 1989), therefore, the penetration rate has a significant effect on the measured excess pore pressures. In the loose to medium dense sand layers (e.g. Heber Road site, location 440', depths 2 to 5 m, 7 to 8 m and 9 to 10m) the excess pore pressures measured under the normal 165 penetration rate were upto 2 times those measured under the fast rate. The penetration rate also seem to affect the cone capability to establish site stratification and to detect loose sand seams (e.g. Heber Road site, 13.5 and 14 m) in which liquefaction can be initiated (Juran and Tumay, 1989). The normal penetration rate seems to yield a more detailed site stratigraphy as well as being able to more effectively detect loose sand seams. 3. The tip pore pressure is most sensitive to the variation in the soil stratigraphy, and can be correlated to variation in the tip resistance. Therefore, it can be used more effectively for detecting loose liquefiable sand seams. 4. During the fast penetration (i.e. 17 cm/s), the pore pressure measured by the piezocell at the sleeve (i.e. about 17 cm above the tip) includes a component due to the residual pore pressures generated at the tip. Therefore, the effect of the penetration rate on the excess pore pressures measured at the sleeve specifically in sand layers is difficult to analyse. Furthermore, the limited data base obtained during this study suggests that the fast rate penetration would not necessarily improve the piezocone capability to detemine the liquefaction potential of soils. 5. In clayey soils, it appears that fast penetration would indicate an apparent stiffness. This observation is supported not only by higher friction ratios measured under the fast 166 rate, but also by lower pore pressures both at the tip and at the sleeve, with the pore pressure at the sleeve sometimes becoming negative. 6. The presently available classification charts are developed on the basis of the standard penetration rate of 2 cm/s. Significant research is therefore required to develop methodologies to integrate the penetration rate into the liquefaction evaluation procedures. 7.2 Monotonic Liquefaction Tests The monotonic liquefaction tests performed in this study included both triaxial and hollow cylinder cell (HCC) tests on specimens of the Heber Road Dune (HRD) sand prepared under different initial relative densities and confining pressures. The results of monotonic triaxial liquefaction tests performed on specimens of the HRD sand were discussed in section 4.1. These results seem to suggest a number of significant points: 1. At any given relative density between 10 and 50%, for a consolidation pressure of 10 psi, the deviatoric stress rapidly increases to its peak value at a low axial strain of about 0.5 to to 1.5% and thereafter deccreases continuously to a residual, in some cases negligible, value. 2. The corresponding excess pore pressure response curves for these tests indicate that at the peak deviatoric stress the excess pore pressure does not reach the consolidation pressure. 167 3. The peak deviatoric stress reached during any test increases with an increase in the initial relative density of the specimen. However, the axial strain corresponding to the peak deviatoric stress does not significantly change. 4. An increase in the relative density results in an increase in the residual deviatoric stress. 5. As the specimen relative density increases, the material behavior changes from liquefiable, with a large post-peak strain softening, to nonliquefiable, with essentially no post peak softening. The experimentally established SSL for the sand suggests the limiting relative density for this change of state to be between 38 to 40%. 6. The initial specimen relative density highly affects the obtained effective stress paths, which collectively define a unique failure envelope. 7. For any consolidation pressure between 10 and 40 psi, at a relative density of 20%, the peak deviatoric stress increases appreciably with an increase in the consolidation pressure of the specimen, while the axial strain corresponding to the peak deviatoric stress does not significantly change. 8. An increase in the consolidation pressure also results in an increase in the residual deviatoric stress. 9. For the specimen initial conditions described above, an increase in either relative density or consolidation presure increases monotonic liquefaction resistance. The results of HCC monotonic liquefaction tests conducted 168 in this study were discussed in section 6.3. These results seem to be in agreement with some of the findings enumerated above for the stress path of the triaxial test, while there are also certain points of distinction: 1. At any given relative density between 10 and 50%, for a consolidation pressure of 10 psi, the cavity pressure rapidly increases to its peak value at a low cavity volume change of less than 4.5% and thereafter decreases continuously to a residual, in some cases negligible, value. 2. The corresponding excess pore pressure response curves indicate that at the peak cavity pressure the excess pore pressure does not reach the consolidation pressure. 3. The peak cavity pressure increases with an increase in the initial specimen relative density, while the corresponding value of the cavity volume change does not significantly change. 4. An increase in the relative density causes an increase in the residual cavity pressure. 5. As the specimen relative density increases, the material behavior changes from liquefiable to nonliquefiable. The experimentally established SSL suggests the limiting relative density to be between 32 to 34%. 6. The effective stress paths cannot be experimentally derived with the present instrumentation. A direct measurement of major and minor principal stresses during cavity expansion tests is necessary for this purpose. 169 7. For any consolidation pressure between 10 and 40 psi, at a relative density of 20%, the peak cavity volume change increases appreciably with an increase in the consolidation pressure of the specimen, while the corresponding cavity volume change does not significantly change. 8. An increase in the consolidation pressure also results in an increase in the residual cavity pressure. 7.3 Cyclic Liquefaction Tests Both cyclic triaxial and HCC tests were performed on specimens of the HRD sands with different initial conditions. The results of cyclic triaxial liquefaction tests performed on specimens of the HRD sand were discussed in section 4.2. These results seem to suggest a number of significant points: 1. For a given relative density between 20 and 50%, with a consolidation pressure of 10 psi, the number of stress cycles, at 1 Hz frequency, required to induce liquefaction increases with increasing relative density. 2. At a relative density of 80%, under consolidation pressure of 10 psi, the specimen did not liquefy although it did undergo cyclic mobility deformation. 3. The rate of generation of the excess pore pressure decreases with increasing relative density for the values of Dc between 20 and 50%, although all four specimens did liquefy when enough number of cycles were applied. For the specimen prepared at the relative density of 80%, a constant excess pore pressure was attained after a certain number of loading 170 cycles, and further cycles did not cause any increase in this constant excess pore pressure value. 4. For a given consolidation pressure between 10 and 40 psi, at a relative density of 20%, the rates of generation of the excess pore pressure and axial strain decrease, although after application of a large number of stress cycles all specimens did finally liquefy. The results of HCC cyclic liquefaction tests conducted in this study were discussed in section 6.3. These results seem to be in agreement with some of the findings enumerated above for the stress path of the triaxial test, while there are also certain dissimilarities: 1. Under a similar stress level, determined by the value of the applied CSR, the liquefaction resistance of the HCC specimen, is always higher than that for the corresponding triaxial specimen prepared under identical initial conditions. 2. Under a consolidation pressure of 10 psi, only the specimens at relative densities of 20 and 30% did actually liquefy. For the specimens at 40 and 50%, a constant residual shear strength was reached, which was not affected by additional stress cycles. This observation tends to suggest that under the cyclic stress path of a cavity expansion test the specimens prepared at relative densities of 40 and 50% underwent a change of state from their initially liquefiable condition, perhaps due to some compaction mechanism. It is important to recall that the specimens prepared to identical 171 initial conditions did liquefy under the stress path of cyclic triaxial test. 3. The rate of generation of the excess pore pressure did decrease with an increase in the initial specimen relative density. 4. Under a relative density of 20%, only the specimens with consolidation pressures of 10 and 20 psi did liquefy. The specimens with consolidation pressures of 30 and 40 psi only underwent cyclic mobi1ity deformation, developing excess pore pressures which reached a constant residual value after a certain number of loading cycles. Again, it is significant to point out that under the stress path of cyclic triaxial test, specimens with consolidation pressures of 30 and 40 psi did undergo complete 1iquefaction.A useful means of studying the liquefaction mechanism(s) during HCC tests is to compare the the cavity expansion response curves obtained during monotonic and cyclic tests. The results of both triaxial and hollow cylinder cell tests seem to suggest that the starting point of collapse during cyclic tests approximately corresponds to the point of the peak deviatoric stress (the peak cavity pressure in case of HCC tests) during monotonic test. In other words, a straight line connecting the origin of the coordinate axes to the point of the peak point on the monotonic stress respose curve (either the deviatoric stress or the cavity pressure) would approximately pass through the point where collapse is 172 initiated during cyclic loading. 7•4 New Hollow Cylinder Cell Equipment A hollow cylinder cell equipment, which allows perfoming static and cyclic cavity expansion tests, was designed, fabricated and calibrated. A major consideration in the design of the hollow cylinder cell equipment was to minimize the boundary condition effect. In order to achieve this objective, effect of the ratio of the internal to external radius, R/r0 , on the observed soil response was to be determined. This in turn required an appropriate test interpretation procedure and the use of an appropriate soil model. In order to address the need for a more adequate interpretation procedure, a new approach was proposed which permits the determination of the shear strength properties, dilatancy properties, and shear modulus of sands from the analysis of HCC test results. A relatively simple soil model was developed which takes into account the contracting / dilating behavior of the soil during cavity expansion. The soil is assumed to be an elasto-plastic material with a non associated flow rule. In order to evaluate the proposed intererpretation procedure, it was used to simulate the results of cavity expansion tests (CET) on Fontainbleau sand. Numerical simulation of the test results using the proposed effective stress analysis procedure indicated that the effective stress soil response are not affected by the geometric boundary 173 conditions, governed by the R/r0 ratio. However, comparison of the engineering properties of the sand derived from the expansion tests and those obtained from triaxial tests shows that the applied stress path affects the soil response. Also, the applied stress path affects soil properties: the plain strain cavity expansion tests result in higher peak friction angle ($'), and dilation angle (v), as compared to the triaxial compression tests. These findings are consistent with the results previously reported in the literature. In order to further validate the effective stress analysis procedure proposed in this study, its predictions were compared with those obtained using the conventional total stress analysis procedures. This comparative analysis indicated that the obtained shear curve by the total stress anlysis procedure is noticeably affected by the geometric boundary conditions, governed by the R/r0 ratio. On the other hand, the effective stress analysis interpretation procedure developed in this study provides a unique soil response to cavity expansion. Recommendations for Future Research. The work reported herein presents only a minor contribution to the State-of-the- Art in the evaluation of the 1iquefaction potential of sandy soil deposits. Certain areas of the present work could be extended to provide better understanding of the mechanisms involved in monotonic and cyclic liquefaction: 1. The design of the hollow cylinder cell equipment could be 174 modified to allow for preparation of specimens with Dr <20%. 2. Effect of several testing variables on the soil response to cyclic HCC cavity expansion could be investigated. These parameters include specimen preparation, loading mode, rate of application of cavity pressure or volume change, etc. 3. The effective stress analysis procedure developed herein could be extended to the strain softening soil response, and, therefore, to simulate monotonic liquefaction. This would be possible mainly due to the fact that, as shown in this study, the stress ratio, q/p', is independent of the stress path. 4. Self-boring pressuremeter test (SBPMT) presents primarily two advantages over the other in-situ tests with regard to the in-situ liquefaction analysis: capability to measure both volume changes and the excess pore pressures during the shearing, and minimized soil disturbance. Hoilow cylinder celIs provide adequate laboratory models of a pressuremeter test. Honotonic and/or cyclic HCC tests could be coupled with in-situ pressuremeter tests with measurements of the excess pore water pressure for laboratory-field correlations. The outcome of such studies could be a liquefaction evaluation procedure which combines our fundamental knowledge of the soil response in the field and in laboratory under both monotonic and cyclic loading. This study could further be enhanced through numerical simulations of the test results. The effective stress analysis developed in this study presents a simple but efficient tool for this purpose. REFERENCES Acar, Y. B. , Tumay, M. T. and Chan, A. (1982), "Interpretation of the Dissipation of Penetration Pore Pressures," Proc. , Int. Symp. Num. Models in Geomechanics, Zurich, Sept., pp. 353-358. Alarcon, A., Chameau, J. L ., and Leonards, G. A. (1986), "A New Apparatus for Investigation the Stress-Strain Characteristics of Sands," Geot. Testing J., ASTM, Vol. 9, No. 4, pp. 204-212. Amer, M. I., Aggour, M. S., adn Kovacs, W. D. (1986), "Testing Using a Large-Scale Cyclic Simple Shear Device," Geot. Testing J., ASTM, Vol. 9, No. 3, pp. 140-146. Anderson, W. F., Pyrah, I. C. and Haji Ali, F. (1987), "Rate Effects in Pressuremeter Tests in Clays," J. Geot. Eng. Div., ASCE, Vol. 113, No. 11, pp. 1344-1358. Arulmoli, K., Arulanandan, K., and Seed, H. B. (1985), "New Method for Evaluating Liquefaction Potential," J . Geot. Eng. Div., ASCE, Vol. Ill, No. GTl, pp. 95-114. Baguelin, F., Frank, R. A., and Nahra, R. (1986), "A Theoretical Study of Pore Pressure Generation and Dissipation around the Pressuremeter," The Pressuremeter and Its Marine Application, Second Int. Symp., ASTM STP 950, pp. 169-186. Baguelin, F., Jezequel, J. F., Lemee, E., and Le Mehaute, A. (1972), "Expansion of Cylindrical Probes in Cohesive Soils," J. Soil Mech. Found. Div., ASCE, 98(11), pp. 1129-1142. Baguelin, F. , Jezequel, J. F. and LeMehaute, A. (1973), "Study of Pore Pressures Occurring During Pressuremeter Test," Proc., 8th ICSMFE, Moscow, Vol. 1, pp. 19-24. Baguelin, F ., Jezequel, J. F . and Shields, D. H. (1978), "The Pressuremeter and Foundation Engineering," Trans. Tech. Publications, Clausthal1. Baguelin, F. J. and Jezequel, J. F. (1983), "The PLC Pressiopenetrometer," Proc., Conf. on Geotechnical Practice in Offshore Engineering, Univ. of Texas, Austin, April 27-29, S. G. Wright (Ed.), pp. 203-219. Baligh, M. M. , Vivatrat, V. and Ladd, C. C. (1980), "Cone Penetration in Soil Profiling," J. Geot. Eng. Dov., ASCE, 175 176 Vol. 106, No. 4, pp. 447-461. Baligh, M. H. (1985), "Strain Path Method," J. Geot. Eng. Div., ASCE, Vol. Ill, No. 9, pp. 1108-1136. Baner jee, P. K. , and Fathallah, R. C. (1979), "An Eulerian Formulation of the Finite Element Method for Predicting the Stresses and Pore Water Pressure Changes around a Driven Pile," Proc. Third Int. Conf. Num. Meth. Geomech., Aachen, April 2-6, 1979. Benoit, J . and Clough, G. W . (1986), "Self-boring Pressuremeter Tests in Soft Clay," J. Geot. Eng. Div., ASCE, Vol. 112, No. 1, pp. 60-78. Bierschwale, J. G. and Stokoe, K. H., II (1984), "Analytical Evaluation of Liquefaction Potential of Sands Subjected tot he 1981 Westmoreland Earthquake," Geot. Engr. Report GR-84-15, Civ. Eng. Dept., Univ. of Texas, Austin. Bjerrum, L. , Kringstad, S. and Kummeneje, O. (1961), "The Shear Strength of a Fine Sand," Proc. 5th ICSMFE, Paris, Vol. 1, pp. 29-37. Briaud, J. L . Lytton, R. L . , and Hung, J. T. (1983), "Obtaining Moduli from Cyclic Pressuremeter Tests," J. Geot. Eng. Div., ASCE, Vol. 109, No. 5, pp. 657-665. Campanella, R. G. and Robertson, P. K. (1981), "Applied Cone Research," Proc., ASCE Conf. on Cone Penetration Testing and Experience, St. Louis, pp. 343-362. Campanella, R. G ., et al. (1982), "Pore Pressure During Cone Penetration Testing," Proc., 2nd ESOPT, Amsterdam, pp. 507-512. Campanella, R. G. , et al. (1984), "Recent Developments in In- situ Testing of Soils," Soil Mechanics Series No. 84, Dept. of Civil Eng., Univ. of British Columbia, Vancouver, 7 p. Canou, J. and Tumay, M. T . (1986), "Field Evaluation of French Self-boring Pressuremeter PAF-7 6 in a Soft Deltaic Louisiana Clay," Proc., 2nd ISPMT, pp. 97-118. Canou, J., Hachem, M. E ., Kattan, A., and Juran, I. (1988), "Mini Piezo-cone (M-CPTU) Investigation Related to Sand Liquefaction Analysis," Int. Symp. on Penetration Testing, March 20-24, Orlando, Florida. Carter, J. P. , Randolph, M . F . , and Wroth, C . P. (1979), "Stress and Pore Pressure Changes in Clay During and 177 After the Expansion of a Cylindrical Cavity," Int. J. Num. and Anal. Meth. in Geomech., 3(4), pp. 305-322. Casagrande, A. (1975), "Liquefaction and Cyclic Deformation of Sands: A Critical Review," Proc. 5th Pan-American Conf. on Soil Mechanics and Foundation Engineering, Vol. 5, Buenos Aires, Argentina, pp. 79-133. Casagrande, A. (1976), "Liquefaction and Cyclic Deformation of Sands - A Critical Review," Harvard Soil Mechanics Series, No. 88, Cambridge, MA. Castro, G. (1969), "Liquefaction of Sands," Harvard Soil Mechanics Series, No. 81, Cambridge, MA. Castro, G. (1975), "Liquefaction and Cyclic Mobility of Saturated Sands," Journal of Geot. Eng. Div., ASCE, Vol. 101, No. GT6, pp. 551-569. Castro, G . and Poulos, S. J. (1976), "Factors Affecting Liquefaction and Cyclic Mobility," Proc. ASCE National Convention, Liquefaction Problems in Geotechnical Engineering, Sept. 27 - Oct. 1, pp. 105-138. Castro, G. , Poul os, S. J . , France, J . W. , and Enos, J. L . (1982) , "Liquefaction Induced by Cyclic Loading," Report to NSF, Geotechnical Engineers, Inc., Winchester, MA. Chan, A. S. K. (1982), "Analysis of Dissipation of Pore Pressures after Cone Penetration," M.S. Thesis, Louisiana State University, Aug., 132 p. de Ruiter, J. (1982), "Static Cone Penetration Test; State-of- the-Art Report," Proc., 2nd ESOPT, Amsterdam, pp. 389- 405. Dobry, R., Stokoe, K. H., Ladd, R. S., and Youd, T . L . (1981), "Liquefaction Susceptibility for S-wave Velocity," Proc. , In-situ Testing to Evaluate Liquefaction Susceptibility, ASCE National Convention, St. Louis, MO, Session No. 24. Douglas, B. J. and Olsen, R. S. (1981), "Soil Classification Using Electric Cone Penetrometer," Proc. ASCE Conf. Cone Penetration Testing and Experience, G. M. Norris and R. D. Holtz (Editors), St. Louis, pp. 209-228. Fahey, M. (1986), "Expansion of a Thick Cylinder of Sand: A Laboratory Simulation of the Pressuremeter Test," Geotechnique, 36(3), pp. 397-424. Finn, W. D. L ., Bransby, P. L . and Pickering, D. J . (1970), 178 "Effect of Strain History on Liquefaction of Sand," J. Soil Mech. and Found. Div., ASCE, Vol. 96, SM6, pp. 1917- 1933. Finn, W. D. L., Pickering, D. J. and Bransby, P. D. (1971), "Sand Liquefaction in Triaxial and Simple Shear Tests," J. Soil Mech. Found. Div., ASCE, Vol. 97, SM4, pp. 639- 659. Finn, W. D. L. , Lee, K. W. , and Martin, G. R. (1977), "An Effective Stress Model for Liquefaction," J. Geot. Eng. Div., ASCE, Vol. 103, No. GT6, pp. 517-533. Finn, W . D. L . and Vaid, Y . P. (1977), "Liquefaction Potential from Drained Constant Volume Cyclic Simple Shear Tests," Proc., 6th World Conf. on Earthquake Eng., Session 6, pp. 7-12. Finn, W. D. L ., Bhatia, S. K. and Pickering, D. J. (1982), "The Cyclic Simple Shear Test," in: Soil Mechanics-- Transient and Cyclic Loads, G. N. Pande and 0. C. Zienkiewicz (editors), John Wiley and Sons, pp. 583-607. Foray, P. (1985), "Essais aux Piezo-cone en Chambre de Calibration," Report de Recherche MRES-IMG, Grenoble. Ghionna, V., et al. (1981), "Performance of Self-boring Pressuremeter Tests in Cohesive Deposits," Report FHWA/RD-81/173, Dept. of Civil Engineering, MIT, Cambridge, MA. Gibson, R. E. , and Anderson, W . F. (1961), "Measurement of Soi1 Properties with the Pressuremeter," Civ. Eng. and Public Works Review, 3(6). Gudehus, G. , Darve, F . , and Vardoulakis, I. (Eds.) (1982), "Constitutive Relations for Soils," Proc., Int. Workshop, Sept. 6-8, Grenoble, France. Hazen, A. (1920), "Hydraulic Fil 1 Dams," Trans., ASCE, Vol. 83, pp. 1713-1745. Hughes, J . M. O. , Wroth, C . P. , and Windle, D. ( 1977) , "Pressuremeter Tests in Sand," Geotechnique, 27(4), pp. 455-477. Hujeux, J. C. , Aubry, D. , Lassoudiere, F. , and Meimon, Y . (1984), "A Double Memory Moden with Mu11ip1e Mechanisms for Cyclic Soi1 Behavior," Int. Symp. on Numerical Models in Geomechanics, Zurich, Vol. 1, pp. 3-13. Ishibashi, I. and Sherif, M. A. (1974), "Soil Liquefaction by 179 Torsional Simple Shear Device,” J. Geot. Div., ASCE, Vol. 100, No. GT8, pp. 871-888. Ishihara, K. and Koga, Y. (1981), "Case Studies of Liquefaction in the 1964 Niigata Earthquake," Soils and Foundations, Vol. 21, No. 3, pp. 35-53. Ishihara, K., Tatsuoka, F. and Yasuda, S. (1975), "Undrained Deformation and Liquefaction of Sands Under Cyclic Stresses," Soils and Foundations, Vol. 15, NO. 1, pp. 29- 44. Ishihara, K. , Yasuda, K. and Yokota, K. (1981), "Cyclic Strength of Undistrubed Mine Tailings," Int. Conf. on Recent Advances in Geotechnical Earthquake Engineering and Soil Dynamics, St. Louis, Vol. 1, pp. 53-58. Ishihara, K. (1985), "Stability of Natural Deposits During Earthquakes," Proc. 11th ICSMFE, Balkema Publishers, Netherlands. Iwasake, T. , et al. (1982), "Prediction of Liquefaction Potential in Japan Using N-values of SPT," Proc., 2nd European Symp. on Penetration Testing, Amsterdam, Vol. 1, pp. 79-84. Jennings, J . E. (1979), "The Failure of a Slimes Dam at Bafokeng," Civ. Eng. Soc. Afr., Vol. 6, pp. 135-140. Jewel 1, R . J., Fahey, M. , and Wroth, C. P. (1980), "Laboratory Studies of the Pressuremeter Test in Sand," Geotechnique, 30(4), pp. 507-531. Jezequel, J . F. and Le Mehaute, A. (1982), "Cyclic Tests with the Self-boring Pressuremeter," Proc., Symp. on Pressuremeter and its Marine Applications, Houston, pp. 209-221. Juran, I., Canou, J., BenSaid, A., Te Kamp, W ., and Tumay, M. T. (1983), "Application of Static Cone Penetrometer and Electric Piezo-cone in In-situ Soi 1 Investigations," Proc., Int. Symp. In-situ Testing of Soils and Rocks, Paris, Vol. 2, pp. 309-315. Juran, I. and Beech, J . 1986, "Effective Stress Analysis of Soil Response in a Pressuremeter Test," Proc., 2nd ISPMT, Col 1ege Station, pp. 150-168. Juran, I. and BenSaid, A. (1986), "Mesures in-situ des Pressions Interstitiel1es; Application a la Reconnaissance des Sols," CERMES-ENPC Report to the French Ministry of DAEI. 180 Juran, I., Mahmoodzadegan, B. and Tumay, M. T. (1989), "Geotechnical Investigation and Piezocone Testing in Well Documented Liquefiable Sites in California," Geotechnical Research Report, Department of Civil Engineering, Louisiana State University, Submitted to The French Ministry of Research, Dec. 1989. Juran, I. and BenSaid, M. A. (1987), "Cavity Expansion Tests in a'Hollow Cylinder Cel 1," Geot. Testing Journal, ASTM, Vol. 10, No. 4, pp. 203-212. Juran, I. and Mahmoodzadegan, B. (1989), "Interpretation Procedure for Pressuremeter Tests in Sand," ASCE J. Geot. Eng. Div., Vol. 115, No. 11, pp. 1617-1633. Juran, I. and Tumay, M. T. (1989), "Soil Stratification Using Dual Pore Pressure Piezocone Test," Proc., 68th Annual Meeting, Transportation Research Board, Jan. 22-26, Washington, D.C. Kiousis, P. D. (1985), "Large Strain THeory as Applied to Penetration Mechanism in Soils," Ph.D. Dissertation, Louisiana State University, 98 pp., Kishida, H. (1966), "Damage to Reinforced Concrete Buildings in Niigata City with Special Reference to Foundation Engineering," Soils and Foundations, Vol. 7, No. 1. Kleiner, D. E. (1976), "Design and Construction of an Embankment Dam to Impound Gypsum Wastes," Proc., 12th Int. Cong. Large Dams, ICLD, Mexico City, Mexico, pp. 235-249. Kramer, S. L . and Seed, H. B. (1988), "Initiation of Soil Liquefaction Under Static Loading Conditions," J. Geot. Eng. Div., ASCE, Vol. 114, No. GT4, pp. 412-430. Ladanyi, B. (1963), "Evaluation of Pressuremeter Tests in Granular Soils," Proc. Second Panamerican Conf. on Soil Mechanics and Foundation Engineering, pp. 3-20. Lee, K. L . and Seed, H . B. (1967), "Cyclic Stress Conditions Causing Liquefaction of Sands," J. Soil Mech. Found. Div., ASCE, Vol. 93, SMI, pp. 47-70. Lee, K. L . and Fitton, J. A. (1969) , in Vibration Effects of Earthquakes on Soils and Foundations, ASTM, STP 450. Lee, K. L . and Focht, J . A. , Jr. (1975), "Liquefaction Potential at Ekofisk Tank in North Sea,” J. Geot. Div., ASCE, Vol. 100, GT1, pp. 1-18. 181 Lee, M. K. W. and Finn, W. D. L. (1975), "DESRA-1, Program for the Dynamic Effective Stress Response Analysis of Soil Deposits Including Liquefaction Evaluation," Soil Mechanics Series, No. 36, Dept. of Civ. Eng., Univ. of British Columbia, Vancouver, B.C., Canada. Lee, K. L. , et al. (1975), "Properties of Soil in the San Fernando Hydraulic Fil1 Dams," J. Geot. Eng. Div., ASCE, Vol. 101, GT8, pp. 801-821. Lee, M. K. W . and Finn, W . D . L . (1978), "DESRA-21, Dynamic Effective Stress Analysis of Soil Deposits with Energy Transmitting Boundary Including Assessment of Liquefactin Potential," Soil Mechanics Series, No. 38, Dept. of Civ. Eng. , Univ. of British Columbia, Vancouver, B.C., Canada. Levadoux, J. N . (1980), "Pore Pressures in Clays due to Cone Penetration," Ph.D. Diss., MIT. Levadoux, J . N . and Baligh, M. M. (1980), "Pore Pressure During Cone Penetratino," Sea Grant Report No. 80-12, MIT, 310 p. Levadoux, J. L . and Baligh, M. M. (1986), "Consolidation After Undrained Piezocone Penetration; I: Prediction," J. Geot. Eng. Div., ASCE, Vol. 112, No. 7, pp. 707-726. Liou, C. P. , Streeter, V. L . , and Richart, F. E. (1977), "Numerical Model for Liquefaction," J. Geot. Eng. Div., ASCE, Vol. 103, GT6, pp. 589-606. Lucia, P. C. (1981), "Flow Failures of Tai1ing Dams, Waste Impoundments, and Embankments," Ph.D. Thesis, Univ. of Calif., Berkeley. Marchetti, S. (1982), "Detection of Liquefiable Sand Layers by Means of Quasi-Static Penetration Tests," Proc., 2nd ESOPT, Amsterdam, Vol. 2, pp. 689-695. Middlebrooks, T . A. (1942), "Fort Peck Slide," Trans. ASCE, Vol. 107, pp. 723-764. Mohamad, R. and Dobry, R. (1986), "Undrained Monotonic and Cyclic Triaxial Strength of Sand," J. Geot. Eng. Div., ASCE, Vol. 112, No. 10, pp. 941-958. Mulilis, J. P. (1975), "The Effects of Method of Sample Preparation on the Cyclic Stress-Strain Behavior of Sands," EERC Report 75-18, Dept. of Civil Eng., Univ. of Calif., Berkeley. Nasr, A. N. and Gangopadhyay, C. R. (1988), "Study of 182 Predicted by Pressuremeter Test," J. Geot. ENg. Div., ASCE, Vol. 114, No. 11, pp. 1209-1225. Nixon, I. K. (1982), "Standard Penetration Test; State-of-the- Art Report," Proc. 2nd ESOPT, Amsterdam, pp. 3-24. Ohsaki, Y. (1966), "Niigata Earthquakes, 1964 Building Damage and Soil Conditions," Soils and Foundations, Vol. 6, No. 2, pp. 14-37. Palmer, A. C. (1972), "Undrained Plane-strain Expansion of a Cylindrical Cavity in Clay: A Simple Interpretation of the Pressuremeter Test," Geotechnique, 22(3), pp. 451- 457. Peacock, N . H. and Seed, H . B. (1968), "Sand Liquefaction Under Cyclic Loading Simple Shear Conditions," J. Soil Mech. Found. Div., ASCE, SM3, pp. 689-708. Poulos, S. J. (1981), "The Steady State of Deformation," J. Geot. Eng. Div., ASCE, Vol. 107, GT5, pp. 553-562. Poulos, S. J., Castro, G. and France, J. W. (1985), "Liquefaction Evaluation Procedure," J. Geot. Eng. Div., ASCE, Vol. Ill, No. 6, pp. 772-791. Rad, N . S. and Tumay, M. T . (1985), "Pore Pressure Response of the Piezocone Penetrometer," Geotechnical Testing Journal, ASTM, Vol. 8, No. 3, pp. 125-131. Robertson, P. K. and Campanella, R. G. (1983), "Evaluation of Liquefaction Potential Using the Cone Penetration Test," Soil Mechanics Series No. 64, Dept. of Civ. Eng., Univ. of British Columbia, Vancouver. Robertson, P. K. and Campanella, R. G. (1985), "Liquefaction Potential of Sands Using the CPT," J. Geot. Eng. Div. , ASCE, Vol. Ill, No. 3, pp. 384-403. Robertson, P. K. , et al. (1985), "Using of Piezometer Cone Data," Research Report No. 92, Dept. of Civi1 Eng., Univ. of British Columbia, Vancouver. Robertson, P. K., and Hughes, J. M. 0. (1986), "Determination of Properties of Sand from Self-boring Pressuremeter Tests," The Pressuremeter and Its Marine Applications, Second Int. Symp., ASTM STP 950, pp. 283-302. Robinet, J. C ., Golcheh, Y., Deffayet, M., and Fau, D . (1983) , "Appareil de Cisaillement des Sols par Torsion d ’un Cylindre Creux," Groupe Francais de Rheologie, Paris. 183 Roscoe, K. H. and Burland (1968), "On the Generalized Stress- Strain Behavior of ’Wet' Clay," in: Engineering Plasticity, J. Heyman and F. A. Leckie (Editors), Cambridge, England, pp. 535-609. Rowe, P. W . (1962), "The Stress-dilatancy Relation for Static Equilibrium of an Assembly of Particles in Contact," Proc. Royal Soc., A. 269, pp. 500-527. Saada, A. S. and Bianchini, G. F. (Eds.) (1987), "Constitutive Equations for Granular Non-cohesive Soils," Proc. Int. Workshop, July 22-24, 1987, Case Western Reserve University, Cleveland, OH. Schmertmann, J. H. (1978), "Use the SPT to Measure Dynamic Soil Properties?--Yes, but ...!" in Dynamic Geotechnical Testing, ASTM SPT 654, pp. 341-355. Schmertmann, J . H. (1978), "Guidelines for Cone Penetration Test--Performance and Design," Federal Highway Administration Report No. FHWA - T S - 7 8 -209 , Washington, D.C.. 145 pp. Schnabel, P. B ., Lysmer, J ., and Seed, H. B. (1972), "SHAKE: A Computer Program for Earthquake Response Analysis of Horizontally Layered Sites," Report No. EERC/72-12, Univ. of California, Berkeley. Schwab, E. and Dormieux, L . (1985), "Liquefaction Due to Expansion of a Cylindrical Cavity," Proc. 11th ICSMFE, San Francisco, Vol. 2, pp. 1049-1054. Seed, H. B. and Lee, K. L . (1966), "Liquefaction of Saturated Sands During Cyclic Loading," J. of SMFD, ASCE, Vol. 92, SM6, pp. 105-134. Seed, H. B . , Lee, K. L. , and Idriss, I. M. (1968), "An Analysis of the Sheffield Dam Failure," Soil Mechanics and Bituminous Materials Res. Lab., Dept. of Civ. Eng., Univ. of Calif., Berkeley. Seed, H . B. and Peacock, W. H . (1971), "Test Procedures for Measuring Soi 1 Liquefaction Characteristics," J. Soil Mech. Found. Div., ASCE, SM8, pp. 1099-1119. Seed, H . B . and Idriss, I. M. (1971), "Simplified Procedure for Evaluating Soil Liquefaction Potential," J . Soi 1 Mech. Found. Div., ASCE, Vol. 97, SM9. Seed, H. B. (1979), "Soil Liquefaction and Cyclic Mobi1ity Evaluation for Level Ground during Earthquakes," J. Geot. ENg. Div., ASCE, Vol. 105, No. GT2, pp. 201-255. 184 Seed, H. B. (1983), "Earthquake Resistant Design of Earth Dams," Proc., Symp. on Seismic Design of Embankments and Caverns, ASCE, pp. 41-64. Seed, H. B ., Idriss, I. M. and Arango, I. (1983), "Evaluation of Liquefaction Potential Using Field Performance Data," J. Geot. Eng. Div., ASCE, Vol. 109, No. 3, pp. 458-482. Seed, H. B. (1983), "Evaluation of the Dynamic Characteristics of Sands by In-situ Testing Techniques," Proc., Int. Conf. on In-situ Testing and Investigation in Soil and Rock, May 18-20, Paris. Seed, H. B ., Tokimatsu, K., Harder, L. F., and Chung, R. M. (1985), "Influence of SPT Procedures in Soil Liquefaction Resistance Evaluation," J. Geot. Eng. Div., ASCE, Vol. Ill, GTl2, pp. 1425-1445. Seed, H. B. , Seed, R. B . , Schlosser, F. , Blondeau, F. , and Juran, I. (1988), "The Landslide at the Port of Nice on October 16, 1979," Report No. UCB/EERC-88/10, Earthquake Engineering Research Center, Dept. of Civ. Eng., Univ. of Calif., Berkeley. Senneset, K., Janbu, N . and Svano, G. (1982), "Strength and Deformation Parameters from Cone Penetration Test," Proc., 2nd ESOPT, Amsterdam, pp. 863-870. Si 1ver, M. L . , et al. (1976), "Cyclic Triaxial Strength of Standard Test Sand," J. Geot. Eng. Div., ASCE, Vol. 102, No. 5, pp. 511-523. Townsend, F. C . (1978), "A Review of Factors Affecting Cyclic Triaxial Tests," Dynamic Geotechnical Testing, ASTM STP 654, pp. 356-383. Tumay, M. T., Boggess, R. L . and Acar, Y . B. (1981), "Subsurface Investigations with Piezo-cone Penetrometer," Proc., ASCE Conf. on Cone Penetration Testing and Experience, St. Louis, pp. 325-342. Tumay, M. T. , Yilmaz, R., Acar, Y . B ., and Deseze, E. (1982), "Soil Exploration in Soft Clays with the Electric Cone Penetrometer," Proc., 2nd ESOPT, Vol. 1, pp. 915-921. Tumay, M. T. , et al . (1985), "Flow Field Around Cones in Steady Penetration," J. Geot. Eng. Div., ASCE, Vol. Ill, No. 2, pp. 193-203. Tumay, M. T . and Acar, Y. B. (1985), "Piezocone Penetration Testing in Soft Cohesive SoiIs," Strength Testing of Marine Sediments: Laboratory and In-Situ Measurements, 185 ASCE STP 883, R. C. Chaney and K. R. Demars, Eds., Philadelphia, pp. 72-82. Vaid, Y. P. , Byrne, P. M. , and Hughes, J. M. 0. (1981), "Dilation Angle and Liquefaction Potential," J. Geot. Eng. Div., ASCE, Vol. 107, GT7, pp. 1003-1008. Vaid, Y. P. and Chern, 3. C. (1983), "Effect of Static Shear on Resistance to Liquefaction," Soils and Foundations, Vol. 23, No. 1, pp. 47-60. Vaid, Y.P. and Chern, J.C. (1983), "Mechanism of Deformation During Cyclic Undrained Loading of Saturated Sands," Int. J. Soil Dynamics and Earthquake Engineering, Vol. 2, No. 3, pp. 171-177. Vesic, A. S. (1972), "Expansion of Cavities in Infinite Soil Mass," 3. Soil Mech. Found. Div., ASCE, 98(3), pp. 265- 290. Wang, M. S. (1972), "Liquefaction of Triaxial Sand Samples Under Different Frequencies of Cyclic Loading," M.S. Thesis, Univ. of Western Ontario, May 1972. Wang, W . (1979), "Some Findings in Soil Liquefaction," Water Conservancy and Hydroelectric Power Scientific Research Institute, Beijing, China, Aug. Wong, R. T . , Seed, H. B . , and Chan, C. K. (1975), J. Geot. Eng. Div., ASCE, Vol. 101, GT6, pp. 571-583. Wroth, C. P., and Windle, D. (1975), "Analysis of Pressuremeter Test Allowing for Volume Changes," Geotechnique, 25(3), pp. 598-604. Wroth, C. P. (1982), "British Experience with the Self-boring Pressuremeter," Proc. Symp. Pressuremeter and Its Marine Applications, Paris, pp. 143-164. Wroth, C . P. (1984), "The Interpretation of In-situ Soi 1 Tests," 24th Rankine Lecture, Geotechnique, Vol. 34, No. 4, pp. 449-489. Zhou, S. G. (1980), "Evaluation of the Liquefaction of Sand by Static Cone Penetration Test," Proc. , 7th World Conf. on Earthquake Engineering, Vol. 3, Istanbul, Turkey. VITA Behnam Mahmoodzadegan was born on September 30, I960, in Tehran, Iran. He received his Bachelor of Science Degree in Civil Engineering from the State University of New York, Buffalo in August 1983. He was awarded a Graduate Teaching Assistantship in September 1983. In August 1984 he received his Master of Science Degree in Civil Engineering from SUNY, Buffalo. He left for Tehran, Iran in September 1984, where he worked for two years for a major geotechnical engineering consulting company. He enrolled in the Doctorate Degree Program in the Civil Engineering Department at Louisiana State University, Baton Rouge in September 1986. He was awarded a Graduate Research Assistantship in January 1987. He has been a candidate for the Degree of Doctor of Philosophy in Civil Engineering since May 1989. 186 DOCTORAL EXAMINATION AND DISSERTATION REPORT Candidate: Behnam MAHMOODZADEGAN Major Field: Civil Engineering (Geotechnical) Title of Dissertation: In-Situ and Laboratory Evaluation of Liquefaction Resistance of a Fine Sand Approve*" _ £ X x V fi, ,i| r, ft A * 0 — ~7 Major Professor and Chairmary » vy? MehmetricllUttzL T. Tumay D e a n of the Graduate School EXAMINING COMMITTEE: Ilan Jnran (Co-Chair) .ra Arman P.E. Connor Sitharma Iyengar / M J.N. Suhayda Date of Examination: November 19 ,— L9-9-Q