\

I
Stress path and steady state
\

Y. P. VAID
Department of Civil Engineering, The University of British Columbia, Vancouver, B.C., Canada V6T I W5
E. K. F. CHUNG
Binnie & Partners (Hong Kong), Hopewell Centre, 183 Queen's Road East, Hong Kong
AND
R. H. Kuerbis
Department of Civil Engineering, The University of British Columbia, Vancouver, B.C., Canada V6T I W5
Received March 3 1, 1989
Accepted September 21, 1989
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by YORK UNIV on 09/30/13

The effect of stress path on the steady state line of a liquefiable sand is investigated. Results from undrained triaxial
compression and extension tests on water-deposited sands show that the steady state line of a given sand, though unique
in the effective stress space, is not so in the void ratio - effective stress space. The sand is contractive over a much
larger range of void ratios in extension than in compression. While a single steady state line emerges for compression
loading, extension loading yields several lines, each characteristic to a given deposition void ratio. All these extension
lines lie to the left of the compression line in void ratio - effective stress space. Thus at a given void ratio, steady state
strength is smaller in extension than in compression, the difference increasing as the sand becomes looser. The implications
of the results are discussed in relation to practical design.
Key words: sand, liquefaction, steady state, stress path.

L'effet du cheminement des contraintes sur la ligne d'etat permanent d'un sable liquefiable est!iudie. Des resultats
d'essais triaxiaux non draines en compression et en extension sur des sables dCposCs dans l'eau montrent que la ligne
d'itat permanent d'un sable donne, quoiqu'unique dans l'espace des contraintes effectives, ne l'est pas dans l'espace
indice de vide - contrainte effective. En extension, le sable est contractant sur une plage beaucoup plus etendue d'indice
For personal use only.

de vide qu'en compression. Aiors qu'une seule ligne d'ktat permanent ressort du chargement en compression, le charge-
ment en extension fournit plusieurs lignes, chacune Ctant caracteristique d'une dCposition a un indice de vide donne.
Toutes ces lignes en extension se situent a gauche de la ligne de compression dans l'espace indice de vide - contrainte
effective. Ainsi, a un indice de vide donne, la resistance a 1'Ctat permanent est plus faible en extension qu'en compression,
la difference augmentant a mesure que le sable devient plus faible. Les implications des rksultats sont discutees en rapport
avec le calcul en pratique.
Mots clPs :sable, liquefaction, etat permanent, cheminement de contrainte.
[Traduit par la revue]

Can. Georech. J. 27, 1-7 (1990)

Introduction question of possible stress path dependency of steady state

Steady state concepts are being increasingly used to line is therefore of utmost practical importance.
analyze the stability of granular soil structures subject to Stress path dependence of undrained response of sands
monotonic or cyclic undrained loading. The key assumption has been reported by Bishop (1971), Miura and Toki (1982),
in such analyses is that the sand has a unique steady state Hanzawa (1980), Chang et al. (1982), Vaid and Chung
line in void ratio - effective confining stress space. This line (1989), Kuerbis et a/. (1988), and Kuerbis and Vaid (1988),
is determined by performing undrained triaxial compression using triaxial compression and extension tests. Compression
tests and it is tacitly assumed applicable regardless of stress tests correspond to a = 0 and extension tests to a = 90'.
path. The term stress path embodies the changes in both Further evidence of undrained behaviour undergoing a sys-
magnitude and direction of principal stresses on a soil tematic weakening (increasing contractiveness) with increase
element. in a is presented by Symes et a/. (1985) and Shibuya and
That soil behaviour is stress path dependent is widely Hight (1987). Given that the stress path dependence of
recognized. It is particularly true for fluvial and hydraulic undrained strength of clays has been routinely considered
fill sands, which assume inherently anisotropic character due in practice since the pioneering work of Bjerrum (1972), it
to their pluvial depositional mode. Consequently, the is surprising that similar concerns have not been expressed
assumption that the triaxial compression steady state line for sand. This variation in undrained strength with principal
for these materials is unique may not be true. Along a poten- stress direction a occurs even though the effective strength
tial failure surface through an embankment, different soil parameters are essentially independent of a. Thus, just as
elements follow different stress paths to failure that range for clays the critical state line is unique in stress space (4'
approximately from axial compression at the crest to axial essentially independent of stress path) but not in void ratio
extension at the toe. Owing to this variation a in the direc- stress space, for sands similar behaviour at steady state is
tion of major principal stress at failure in relation to the likely to be manifested.
vertical deposition direction, the steady state strength for The objective of this paper is to investigate the effect of
a given void ratio could vary along the failure surface. The stress path on the steady state line of sand. The triaxial test

Printed in Canada / IrnprimC au Canada

 2 CAN. GEOTECH. J. VOL. 27, 1990
,
(01 e ( n ) Phore Tronsformotion

-bm Limited Liquefoction

1
b-

Liquefoction
I
Axial Strain

/
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by YORK UNIV on 09/30/13

- --)---'-
EXTENSION COMPRESSION

- I I I I

-
-
-
-
I k9f/stnZ- 98 kPa \
-
-
t 1 1 I
For personal use only.

FIG. 2. Typical undrained loading response o f loose Ottawa

sand.
I
Undrained compression and extension tests were carried
a; out on sand specimens over a range of depositional void
FIG. 1. Types o f contractive deformation. ratios ei,confining pressures a;,, and anisotropic consolida-
tion stress ratios Kc = a;,/a;I,, where a:, and ah, refer to
is used for this purpose. Since this test can simulate only vertical (soil deposition direction) and radial principal con-
two undrained stress paths (compression and extension), as solidation stresses respectively. Only those void ratio and
such this study is limited to investigating possible differences stress conditions were considered that resulted in contrac-
between triaxial compression (a = 0) and extension ( a = 90") tive response in compression or extension.
steady state lines. Where differences do exist, their important
implications in the analysis of the stability of hydraulic fill
structures are discussed. Steady state and phase transformation state
Steady state (SS) for a contractive sand corresponds to
a state of residual undrained strength (following a peak) that
Experimentation is sustained with continued deformation (Fig. la). The term
Ottawa sand, ASTM designation C-109, was used in this liquefaction is used for this type of response (Castro 1969).
study. Ottawa sand is a medium uniform quartz sand The state of transient minimum strength associated with con-
(DS0 = 0.40 mm, C, = 1.5) with rounded particles. tractive response of the limited liquefaction type corresponds
Reference maximum and minimum void ratios determined to thr: phase transformation (PT) state (Ishihara et al. 1975).
by the ASTM test method D2049 are 0.82 and 0.50 This state signifies the start of dilation or decrease in pore
respectively. pressure and increase in deviator stress with further
Triaxial specimens, 63 mm in diameter x 126 mm high, undrained shearing.
were prepared by water pluviation. Water pluviation is con- Comprehensive studies by Vaid and Chern (1985) show
sidered to produce sand fabric similar to that of natural that contractive response of both types-liquefaction and
fluvial sands (Oda et al. 1978). Laboratory studies on limited liquefaction-can be treated within the same frame-
pluviated sands should therefore duplicate the characteristic work. The steady state line in void ratio - effective stress
behaviour of natural sands and hydraulic fills. Pluviation space associated with liquefaction-type contractive response
yielded the loosest state of deposition. A range of deposi- also encompasses response of the limited liquefaction type
tional void ratios were simulated by densification following (Figs. lb, lc). This line will herein be referred to as the SS
pluviation. Details of specimen preparation by water pluvia- and P T line. The single line in stress space describing stress
tion have been described by Vaid and Negussey (1988). conditions at SS and P T states (Fig. 1b) implies that friction
A back pressure of 200 kPa was used in all tests and satura- angle $pT mobilized at SS and P T states for a sand is
tion was insured by insisting on a B value of at least 0.99. unique. Work of Castro (1969) and later of others
 VAID ET AL.
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by YORK UNIV on 09/30/13

FIG. 3. Stress conditions at triggering of liquefaction, steady state, and phase transformation.
For personal use only.

(Mohamad and Dobry 1986; Sladen et al. 1985) also treated pore pressure in the two 1oadifig"modesare primarily respon-
liquefaction and limited liquefaction within the same frame- sible for the observed contrasts in the undrained stress-strain
work, when considering steady state line in void ratio -stress response.
space. The state of transient minimum strength associated Only deposition void ratios between about 0.686 and the
with P T state in response of the limited liquefaction type, loosest state of deposition resulted in contractive compres-
however, cannot be regarded as a steady state of sion response. The response was always of the limited
deformation. liquefaction type, regardless of ah, and Kc levels. The
extension response, in contrast, was of the liquefaction type
at the loosest void ratio ei = 0.715, and gradually changed
Test results to limited liquefaction type with decreasing void ratio. The
The test data presented cover a range of depositional void sand was contractive in extension over a much larger range
ratios ei from 0.715 to 0.657, initial confining pressures ahc of void ratios (ei = 0.628 and higher), when compared
from 200 to 2500 kPa, and Kc values from 1 to 2. Void with that under compression. Thus, in the range of deposi-
ratio ei = 0.715 was the loosest that could be obtained tion void ratios 0.686-0.628, compression response was
after water pluviation and imposing a confinement of about dilative but extension response contractive. Over the range
20 kPa during sample preparation. Sample deformation dur- of overlapping density and stress states (e,, afIc, and Kc) for
ing extension loading was very uniform well after the realiza- which both loading modes resulted in contractive response,
tion of SS or P T states. Necking appeared only at very large liquefaction was always triggered at a lower strain and
strains in excess of about 10% axial strain. deviator stress level in extension than in compression.
Typical undrained compression and extension behaviour
of two identical loose samples is shown in Fig. 2. The initial Triggering of liquefaction
state of the specimens prior to undrained loading was Stress conditions at the onset of liquefaction or limited
e, = 0.705, ahc = 350 kPa, and Kc = 1. Both loading liquefaction (peak of deviator stress) are shown in Fig. 3.
modes show a contractive response. Extension behaviour, Several data points shown in both compression and exten-
however, is much weaker, both in peak deviator stress and sion loading modes represent results from more than one
P T or SS deviator stress, than compression behaviour. test at a given void ratio. The results show that the data
Compression and extension tests were carried out by points in compression lie on a straight line passing through
respectively increasing and decreasing axial stress a, while the origin. This implies that liquefaction or limited liquefac-
the cell pressure a~ was held constant. Yet, despite tion in compression is triggered at a unique value of effec-
decreases in mean normal stresses in extension as opposed tive stress ratio (critical stress ratio, CSR) regardless of the
to increases in compression, pore pressures Au/afI, are initial state (e,, ah,, and Kc) of the sand. The CSR line in
larger in extension than in compression. This reflects radical Fig. 2 is identical to that already reported (Vaid and Chern
differences in shear-induced compressibility in the two 1983, 1985) for Ottawa sand.
loading modes. Drained tests carried out by Negussey (1984) Unlike compression, in extension liquefaction or limited
on the same sand at constant mean normal stress confirm liquefaction does not trigger at a unique value of CSR that
such a conclusion. Clearly, the differences in shear-induced is independent of ec, Kc, and ah, (Fig. 3). Instead, a unique
 CAN. GEOTECH. J. VOL. 27, 1990
,
\
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by YORK UNIV on 09/30/13

FIG. 4. Steady state and phase transformation lines of Ottawa sand in extension and compression loading.
CSR that is independent of the consolidation stress state a maximum and starts decreasing w i ~ hfurther straining. In
(ahc and Kc) seems to be associated with each deposition an attempt to seek a broader ~igni~ffcanceof of a sand,
void ratio ei. Two CSR lines are shown together with the Negussey et al. (1988) have ghown that 4,, equals
data points at two different deposition void ratios (ei = constant-volume friction angle,$,, for the same sand.
0.715 and 0.681). The value of CSR increases with increase
in ei, and for Ottawa sand this corresponds to a mobilized SS and PT state line in void ratio - stress space
friction angle 4' between 13 and 18". The CSR is therefore Relationships between void-ratio ec and effective confin-
For personal use only.

snbstantially smaller in extension than that in compression ing stress u; at SS or P T state for triaxial compression and
extension loadings are shown in Fig. 4. Compression
(4hobilized = 23").
Additional evidence in support of a unique CSR in response may be seen to yield a unique SS and P T line. This
undrained compression may be found in studies on other is the characteristic compression type of response reported
sands (Bishop 1971; Vaid and Chern 1985; Mohamad and for sands (e.g., Castro et al. 1985). In extension, however,
Dobry 1986). Liquefaction studies using moist tamped spec- the ec-a; relationship is not unique. A different SS and P T
imens have, in general, not revealed such a characteristic line is seen to emerge for each deposition void ratio e,. Four
of liquefiable sands. Sample inhomogeneity (Castro et al. such lines are shown in Fig. 4. A general consistency of data
1982) and possible differences in fabric at each tamped den- and lack of scatter are particularly noteworthy in all SS and
sity could be cited as possible reasons for this non- P T lines in Fig. 4. Compression steady state lines reported
uniqueness. The peak strength of sand that is contractive in the literature for moist tamped sand often suffer from
occurs at very small strain (less than 0.5% axial strain for considerable scatter (Castro et al. 1982). As explained
Ottawa sand). Consequently, fabric differences that affect before, this may be a consequence of sample nonuniformity
profoundly the nature of induced pore pressure at small together with fabric variation at each moist tamped density.
strain level can be expected to cause large changes in the Because of the relative flatness of SS and P T lines for
mobilized effective stress ratio at peak strength. rounded sands (such as Ottawa sand), small changes in void
ratio can influence a; at P T or SS by a large amount. The
SS and PT state line in stress space water pluviation technique used for sample reconstitution
Stress conditions at SS or P T states for contractive in this study results in very uniform specimens that can be
response are also shown in Fig. 3. The data points may be replicated easily (Vaid and Negussey 1988). The technique
seen to lie on two straight lines passing through the origin simulates closely the fabric of water-deposited sands and
of stress space for each of the two loading modes. Further- hydraulic fills and thus enables confident assessment of the
more, the slopes of the two lines are essentially identical. element properties of these materials under undrained
This implies that the friction angle r$, mobilized at SS or loading.
P T state following contractive response is unique for Ottawa In Fig. 4 all SS and P T lines in extension lie to the left
sand. It is independent of the deposition void ratio, con- of the unique line in compression. This implies that at a given
solidated state (e,, ahc and Kc) prior to undrained loading void ratio e, the SS or P T state strength SpTin extension
as well as the mode of loading-compression or extension. is smaller than that under compression. SpTis given by
Studies of the undrained behaviour of other sands also
show that 4pTis unique for a given sand (Bishop 1966,
1971; Vaid and Chern 1983, 1985; Ishihara et al. 1975;
Kuerbis et al. 1988). In fact, the uniqueness of +pT for a As an example, at loose state (e, = 0.705) strength in
given sand encompasses both contractive and dilative extension is only 10% of that under compression. The dif-
responses. For the dilative undrained response, the P T state ference in undrained strengths reduces with decrease in void
corresponds to the instant at which the pore pressure attains ratio. For deposition void ratios smaller than 0.686 there
 VAID
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by YORK UNIV on 09/30/13

AXIAL STRAIN Ea (%)

FIG. 5. Steady state and phase transformation lines of Brenda FIG.6. Deformation of Ottawa sand under virgin isotropic
sand in extension and compression loading (after Chern 1985). consolidation.
<
will be no concern for liquefaction under compression corrected for membrane penetration effects using procedures
loading but the sand would still be susceptible to liquefac- suggested by Vaid and Negussqy (1984).
tion in extension at deposition void ratios down to about The difference in the observed compression versus exten-
0.628. sion response of Ottawa sand is a reflection of this inherent
anisotropy. Compression mode corresponds to or = 0 (the
For personal use only.

Similar dramatic differences in SS and P T state lines in

compression and extension for an angular tailings sand are angle between al and-deposition direction) in contrast to
shown in Fig. 5 (Chern 1985). This sand is about 60% or = 90" for the extension mode. There is a close associa-
feldspar and 30% quartz and has maximum and minimum tion of the degree of inherent anisotropy and the magnitude
void ratios of 1.060 and 0.688 respectively. While the com- of difference in stress-strain response between the two
pression SS and P T state line is unique for the sand, the modes of undrained behaviour. Greater deformability in the
extension line (shown for only one deposition void ratio) horizontal direction is mainly responsible for a more pro-
lies substantially to the left of the compression line. At the nounced contractive response in extension than compression,
loosest void ratio shown, extension PT or SS strength is only or contractive behaviour in extension but dilative in com-
about 15% of the compression value. Further, radical dif- pression. Similar differences in triaxial compression and
ferences between compression and extension response of extension undrained response have been reported by others
several gradations of this sand and silty sands reconstituted (Bishop 1971; Miura and Toki 1982; Hanzawa 1980; Chang
by a process that simulates fluvial and hydraulic fill place- et a[. 1982; Vaid and Chung 1989). Further evidence of
ment are reported by Kuerbis et al. (1988). Although '
undrained behaviour undergoing systematic weakening
deposited in the loosest state, all sands showed dilative (increasing contractiveness and reduction in P T or SS
response in compression, while the extension response was strength) with increase in a is provided by Symes et al. (1985)
strongly contractive. and Shibuya and Hight (1987).
Although the undrained strengths at PT or SS in com-
Discussion of results pression and extension are different, the friction angles are
essentially identical (Fig. 3). Therefore, the P T or SS line
Water-deposited sands are inherently anisotropic in their for contractive response, though unique in stress space, is
deformation properties. Their response to loading would not so in void ratio - stress space. The differences in
therefore depend on the orientation of the principal stresses undrained PT or SS strengths are therefore due solely to
in relation to their deposition direction. Considerable larger pore pressures induced in extension on account of
evidence of this anisotropic behaviour has been demon- greater horizontal compressibility.
strated under drained loading (e.g., Arthur and Menzies
1972, Oda 1981). It is then expected that the undrained
response will also be anisotropic. Practical implications
The nature of inherent anisotropy in Ottawa sand tested Realistic characterization of the response of fluvial and
in this study can be assessed from strain response during hydraulic fill sands as to their susceptibility io liquefaction
virgin isotropic consolidation (Fig. 6). Equal axial and radial must consider
strain would imply isotropy. All deposition densities (ei's) (1) simulation of the deposition process of the deposit to
show larger radial strains than axial. This anisotropic com- be modelled in reconstituting laboratory test specimens, if
pressibility may be seen to be more accentuated at the looser it is not possible to obtain undisturbed samples;
void ratio. Radial strains were computed from measured (2) variation in undrained response with inclination or of a,
axial and volumetric strains. Volumetric strains were suitably with the deposition direction.
 6 CAN. GEOTECH. J. VOL. 27, 1990
\
' I ,
Fabric simulation is only 10% of that in compressjon, y d deposition void
The steady state line for a sand is normally determined ratios less than about 0.686 do not even show contractive
in the laboratory by compression tests on samples behaviour in compression, yet contractive response in exten-
reconstituted by the moist tamping technique. Moist tamped sion persists at void ratios down to about 0.628.
sand samples are contractive in compression over a range
of void ratios (Castro et al. 1985). These include rounded
as well as angular sands with a range of gradations. Many Conclusions
water-deposited sands, on the other hand, may be dilative Stress path effects on undrained behaviour of a water-
even in the loosest deposition state (Kuerbis et al. 1988). deposited sand were studied using triaxial compression and
Moist tamping must therefore be imparting to the sand a extension loading modes. A range of initial deposition den-
fabric that promotes contractive behaviour. sity states, confining pressures up to 2.45 MPa, and both
Untamped moist sand is in a bulked state with a density isotropic and anisotropic consolidation states were con-
looser than the ASTM minimum. Tamping is used to achieve sidered. Based on the test results, the following conclusions
any desired density. During tamping there is a resistance to can be drawn.
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by YORK UNIV on 09/30/13

particle packing on account of capillary tensions. This results (1) The range of void ratio over which contractive
in the development of a fabric clearly different from that response manifests in much higher in extension than in the
of the water-deposited sand, which is placed under gravita- compression mode. For a certain range of void ratio, the
tional forces. Since moist dumped sands are particularly undrained response could be dilative in compression but con-
prone to liquefaction due to their "metastable honeycomb" tractive in extension. Inherent anisotropy of water-deposited
structure (Casagrande 1976), some remnants of this sand is considered to be the main cause of the difference
characteristic is very likely to persist even after tamping to in the two modes of loading.
densities higher than the ASTM minimum. In some labora- (2) Liquefaction or limited liqvefaction (contractive
tory studies, loose moist tamped specimens of fine-grained response) was triggered in compression mode at a unique
soils have been shown to experience very large strains dur- value of effective stress ratio that is independent of the
ing the saturation process (Marcuson and Gilbert 1972; deposition density as well as t h e consolidation stress state
Chang et al. 1982; Sladen et al. 1985). This collapsing prior to undrained loading. In extension, however, such a
characteristic on mere removal of capillary tensions suggests unique value of effective stress ratio was associated with each
For personal use only.

a metastable fabric that is likely to promote contractive deposition void ratio. The stiess ratio increased somewhat
deformation during shear. with decrease in deposition void ratio, but it was always
Modelling loose water-deposited sand by moist tamping smaller than that mobilized under compression loading.
therefore might unjustifiably condemn it as being potentially (3) The friction angle mobilized at steady state (SS) or
liquefiable in conpression. In water-deposited loose sands phase transformation (PT) state was unique for the sand.
that are contractive in compression (such as the Ottawa sand It was independent of both the initial state (void ratio, con-
reported herein), moist tamping methods may tend to pro- fining pressure, and shear stress) and the loading mode
mote more pronounced contractive behaviour at a given (compression or extension).
density. (4) While the SS and P T line was independent of stress
path in stress space, it was not so in the void ratio - stress
Anisotropy of undrained response space. In compression a unique SS and P T line was found
At the instant of undrained failure along a curved sur- in ec-a; space. In extension, however, each deposition void
face through a hydraulic fill embankment, the orientation ratio el gave rise to a different SS and P T line. ?'he position
a of al to the vertical could vary from a = 0 (compression of these lines was always to the left of the unique line in
loading) at the crest to a = 90" (extension loading) at the compression, with the amount of shift increasing with
toe. Compression loading alone cannot therefore represent increase in el. Thus at a given void ratio e,, PT or SS
the undrained steady state strength along the entire poten- strength was smaller in extension than in compression, the
tial failure surface. If the failure surface analyzed is relatively difference increasing as the sand became looser.
flat, as was the case in the hydraulic fill failure of the lower (5) Laboratory assessment of liquefaction susceptibility
San Fernando dam (Castro et al. 1985), the undrained steady of fluvial and hydraulic fill sands must ensure sample
state strength would correspond to loading with a = 45 reconstitution by a process that simulates these water deposi-
+ 4,/2, where 4, is the friction angle at steady state. With tion processes. Moist tamping techniques of reconstitution
6, = 37", a = 63", which is closer to an extension rather simulate neither the fabric nor the range of void ratios
than a compression mode of loading. A reduction factor accessible to water-deposited sands and silty sands. In addi-
of 20 to the measured compression steady state strength to tion, the method of stability analysis must recognize and
obtain the strength operating at the instant of failure (as take into account the fact that at a given void ratio, steady
reported by Castro et al. 1985) would then not be necessary state strength varies with the inclination of al to the fill
if suitable a-dependent values of steady state strength are deposition direction.
used. Densification of undisturbed sand samples taken from
a field deposit could be blamed for an overestimate of the
in situ steady state strength. But, its effects are probably
overstated, since the a dependence of steady state strength Acknowledgements
was neglected. An estimate of the magnitude of difference This research was supported by a grant from the Natural
in undrained steady state strengths between compression and Sciences and Engineering Research Council of Canada. Kelly
extension loading modes may be obtained from the results Lamb typed the manuscript and Richard Brun drafted the
in Fig. 4. Loose Ottawa sand has extension strength that figures.
 VAID ET AL 7
\

ARTHUR,J.R.F., and MENZIES,B. 1972. Inherent anisotropy in MIURA,S., and TOKI,S. 1982. A sample pieparation method and
a sand. Giotechnique, 22: 115-128. its effect on static and cyclic deformation-stiength properties
BISHOP,A.W. 1966. Strength of soils as engineering materials. of sand. Soils and Foundations, 22(1): 61-77.
6th Rankine Lecture. Giotechnique, 16: 89-130. MOHAMAD, R., and DOBRY,R. 1956. Undrained monotonic and
1971. Shear strength parameters for undisturbed and cyclic triaxial strength of sand. ASCE Journal of Geotechnical
remolded soil specimens. Roscoe Memorial Symposium, Engineering, 112: 941-958.
Cambridge University, pp. 3-58. NEGUSSEY, D. 1984. An experimental study of the small strain
BJERRUM, L. 1972. Embankment on soft ground. ASCE Specialty response of sand. Ph.D. thesis, The University of British
Conference on Earth and Earth Supported Structures, Purdue Columbia, Vancouver, B.C.
University, vol. 1, pp. 1-54. NEGUSSEY, D., WIJEWICKREME, W.K.D., and VAID,Y.P. 1988.
CASAGRANDE, A. 1976. Liquefaction and cyclic deformation on Constant-volume friction angle of granular materials. Canadian
sands-a critical review. Harvard Soil Mechanics Series No. 88, Geotechnical Journal, 25: 50-55.
Harvard University, Cambridge, MA. ODA,M. 1981. Anisotropic strength of cohesionless sand. ASCE
CASTRO,G. 1969. Liquefaction of sands. Ph.D. thesis, Harvard Journal of the Geotechnical, 108: 1219-123 1.
University, Cambridge, MA. ODA,M., KOISHIKAWA, I., and HIGUCHI, T. 1978. Experimental
study of anisotropic shear strength of sand by plane strain test.
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by YORK UNIV on 09/30/13

CASTRO,G., POULOS,S.J., FRANCE,J.W., and ENOS,J.L. 1982.

Liquefaction induced by cyclic loading. Report to National Soils and Foundations, 18(1): 25-38.
Science Foundation, Washington, DC. SHIBUYA, S., and HIGHT,D.W. 1987. A bounding surface for
CASTRO,G., POULOS, S.J., and LEATHERS,F.D. 1985. granular materials. Soils and Foundations, 27(4): 123-136.
Re-examination of sIide of lower San Fernando Dam. ASCE SLADEN,J.A., D'HOLLANDER, R.D., and KRAHN,J. 1985. The
Journal of Geotechnical Engineering, 111: 1093-1 107. liquefaction of sands, a collapse surface approach. Canadian
CHANG,N.Y., HSIEH,N.P., SAMUELSON, D.L., and HORITA,M. Geotechnical Journal, 22: 564-578.
1982. Static and cyclic behaviour of Monterey no. '0' sand. Pro- SYMES,M.J., SHIBUYA, S., HIGHT,D.W., and GENS,A. 1985.
ceedings, 3rd Microzonation Conference, Seattle, pp. 929-944. Liquefaction with cyclic principal stress rotation. 11th Interna-
CHERN,J.C. 1985. Undrained response of saturated sands with tional Conference on Soil Mechanics ancl Foundation Engineer-
emphasis on liquefaction and cyclic mobility. Ph.D. thesis, The ing, San Francisco, vol. 4.
University of British Columbia, Vancouver, B.C. VAID,Y.P., and CHERN,J.C. 1983. Meihanism of deformation
HANZAWA, H. 1980. Undrained strength and stability of quick during cyclic undrained loading of saturated sands. Soil
sand. Soils and Foundations, 20(2): 17-29. Dynamics and Earthquake Engineering, 2, No. 3, pp. 171-177.
ISHIHARA, K., TATSUOKA, F., and YASUDA,S. 1975. Undrained -1985. Cyclic and monotonic undrained response of
For personal use only.

deformation and liquefaction of sand under cyclic stresses. Soils saturated sands. Advances in the Art of Testing Soils Under
1 and Foundations, 15(1): 29-44. Cyclic Conditions, ASCE Convention, Detroit, pp. 120-147.
/ KUERBIS,R.H., and VAID, Y.P. 1988. Sand sample prep- VAID,Y.P., and CHUNG,E.F.K. 1989. Preshearing and undrained
I aration-the slurry deposition method. Soils and Foundations, response of sand. Soil Mechanics Series No. 128, University of
1 28(4): 107-118. British Columbia, Vancouver, B.C.
j KUERBIS,R.H., NEGUSSEY, D., and VAID,Y.P. 1988. Effect of VAID,Y.P., and NEGUSSEY, D. 1984. A critical assessment of
I gradation and fines content on the undrained response of sand. membrane penetration in the triaxial test. Geotechnical Testing
!I ASCE Conference on Hydraulic Fill Structures, Geotechnical Journal, 7: 70-76.
Special Publication 21, pp. 330-345. -1988. Preparation of reconstituted sand specimens.
I MARCUSON, W.F., 111, and GILBERT,P.A. 1972. Earthquake American Society for Testing and Materials, Special Technical
liquefaction potential at Patoka Dam, Indiana. United States Publication, 997, pp. 405-420.
Army Engineer Waterways Experimental Station, Miscellaneous
Paper S-72-42.