ACCURATE DETERMINATION OF THE CRITICAL STATE FRICTION

ANGLE FROM TRIAXIAL TESTS

Mohsen Ghafghazi, Department of Civil Engineering, University of British Columbia, Vancouver, BC, Canada
Dawn Shuttle, Department of Civil Engineering, University of British Columbia, Vancouver, BC, Canada

ABSTRACT

The idea that soils sheared to very high values of strain will eventually reach a constant void ratio, e c , and friction
angle, φcv , is well established in soil mechanics. This constant volume state, usually termed the critical state, is also
inherent in many, if not most, advanced constitutive soil models , which are sensitive to the estimated shear strength
parameters. Hence it is important that geotechnical engineers have access to good parameter estimates based on
good industry quality data. However, despite the importance of the critical state to understanding soil behaviour,
measurement of φcv remains problematic. This paper uses good industry quality triaxial data from two well-known
Erksak and Ticino sands to investigate the accuracy of four different methods of obtaining shear strength parameters.
Based on theoretical and practical limitations of these methods , recommendations are made on how to obtain the
most reliable shear strength parameters from a limited amount of data from drained triaxial tests on dense samples.

RÉSUMÉ

L'idée que les sols cisaillés à très hautes déformations atteindront un indice de vides constants, ec, et un angle de
friction φcv , est bien établie dans la mécanique de sol. Cet éta t de volume constant, habituellement nommé l’état
critique, est aussi inhérent à beaucoup, si pas dans la plupart des modèles constitutifs avancés qui sont sensibles
aux paramètres de résistance au cisaillement. Par conséquent il est important que les ingénieurs géotechniques
aient accès à bonnes estimations de paramètres basés en bonnes données de qualité d'industrie. Cependant,
malgré l’importance de l’état critique pour comprendre le comportement de sol, la mesure de φcv demeure
problématique. Cet article emploi des données de bonne qualité d’industrie d’essais triaxiaux de deux sables bien
connus, Erksak et Ticino pour étudier l’exactitude de quatre méthodes différentes d’obtenir les paramètres de
résistance au cisaillement. Basé sur des limitations théo riques et pratiques de ces méthodes, des
recommandations sont faites sur la façon d’obtenir les paramètres les plus fiables de résistance au cisaillement
d'une quantité limitée de données des essais triaxiaux drainés sur les échantillons denses.

1. INTRODUCTION This paper describes four different methods from the

The idea that soils sheared to very high values of strain
will eventually reach a constant void ratio, e c , and friction
angle, φcv , is well established in soil mechanics. This
constant volume state, usually termed the critical state,
is also inherent in many, if not most, advanced
constitutive soil models, including the well known critical
state soil models Cam Clay and Modified Cam Clay.
Geotechnical predictions are sensitive to the assumed
shear strength parameters. Hence it is important that
geotechnical engineers have access to good parameter
estimates. However, despite the importance of the
critical state to understanding soil behaviour,
measurement of φcv remains problematic. This is
particularly the case in engineering practice, where
typically only a limited number of soil tests are available.
literature to obtain the critical state friction ratio in triaxial
conditions. The accuracy of each of thes e methods is
discussed based on previous ly published data from the
well-known Ticino sand, taking into consideration the
limited number of tests typically available in practice.
Then two of the methods being more promising in
determining shear strength parameters are applied to
data from another well-known sand, Erksak, to
investigate capabilities of each of these methods in
determining the critical state friction ratio. Finally
recommendations are made on how to acquire the most
accurate parameters from a limited amount of data.
1.1 Definition of the Critical State
The critical state (also called steady state) was defined
by Roscoe et al. (1958) as the state at which a soil
"continues to deform at constant shear stress and
constant velocity". Poulos (1981) gives a more precise
definition as "The steady state of deformation for any
mass of particles is that state in which the mass is
continuously deforming at constant volume, constant
normal effective stress, constant shear stress and
constant velocity". Writing these definitions in a formal
form we have:
∃C (e, q, p′) &p′=0 ∋ ε&v ≡ 0 ∧ ε&&v ≡ 0∀ε q [1]
Where C( ) is the function defining the CSL, e is the void
ratio, q is the deviatoric stress, p' is the mean effective
stress and εv and εq are volumetric and deviatoric strains
respectively. The above equation includes two important
conditions: first, the volumetric strain rate must be zero;
second, the rate of change of this strain rate must also
be zero. Hence, both dilatancy and rate of change of
dilatancy must be zero during shearing at the critical
state. There are no strain rate terms in C( ) , making the
CSL identical to the steady state of Poulos (1981).
Constant mean stress is invoked in the equation to avoid
a less easily understood definition for the situation in
which mean stress is increased while the soil is
continuously sheared at the critical state (Jefferies
1993).
These definitions can be used to infer the existence of a
unique critical state line (not necessarily linear) in e –
log p' space. This line is the locus of end points of state
path (e – log p') for different tests sheared to high strain
values. Many authors have presented data on existence
of a unique critical state line (e.g. Castro 1969, Been and
Jefferies 1985, Been et al. 1991, Vaid and Sasitharan
1992, Garga and Sedano 2002). It has been shown that
all soil samples end on this line regardless of stress
path, sample preparation (also known as fabric), mean
effective stress and initial void ratio.
should be treated as a function of the intermediate
principal stress, represented by the lode angle (θ). In
doing this, triaxial compression conditions are taken as
the reference case in which soil properties are
determined. Thus, M tc becomes the soil property (where
subscript 'tc' denotes triaxial compression), and M(θ) is
evaluated in terms of this property (Jefferies and Shuttle
2002). For known stress conditions the friction angle is
directly related to stress ratio and the two parameters
can be applied interchangeably; for example, in triaxial
compression we have
6 sin φcv
M tc = [3]
3 − sin φ cv
This paper only considers triaxial compression
conditions.
1.2 Stress-Dilatancy Definitions
The stress-dilatancy plot is very useful in understanding
the behaviour of soils. The ratio between two stress
invariants (η = q/p') is plotted on the vertical axis. (Total)
dilatancy, which is plotted on the horizontal axis, is
defined as the ratio of rate of volume change over rate of
change in shear strain. In mathematical form:
ε&v
D= [4]
ε&q
To reduce the noise in processing raw data into the
dilatancy, a central-difference approach is employed
here. That is,

Equation 1 also requires the existence of a unique

ε v, j +1 − ε v , j −1
critical state locus in p'-q space. Since the models are Dj = [5]
cast in terms of stress invariants, and the relationship ε q , j +1 − ε q, j −1
between these invariants needs to be expressed for the
critical state, it is convenient to replace the critical state
friction angle with a parameter which is directly related to where j subscript means that the value of parameter at j
th

stress invariants . The convention is to introduce a critical data point is considered.

stress ratio, M, so that, at the critical state:
Stress -dilatancy is a concept regarding plastic strain
rates, and thus, equation 5 is not itself sufficient to
q = Mp ′ [2] reduce test data, as there is an elastic component of
strain rate at other than peak or critical strengths. The
plastic dilatancy is estimated as
Experimental data suggest that constant M will yield
unrealistic friction angles for different stress conditions
in a general 3-D stress space (e.g. Bishop 1966 and
Wanatowski and Chu 2006). It is suggested that M
 ε&vp (ε v, j +1 − ε v, j −1 ) − (p ′j +1 − p ′j −1 ) K
D jp = =
ε&qp (ε q , j +1 − ε q , j −1 ) − (q j +1 − q j −1 ) 3G
[6]
K being the elastic bulk modulus and G the elastic
shear modulus ( K =
2(1 + υ ) ).
G
3(1 − 2υ )
2. CRITICAL STATE FRICTION RATIO
DETERMINATION
The critical state of soils is usually achieved at very high
strains. The ring shear device can measure the critical
state directly, but this test also has its own
shortcomings. The major difficulty with ring shear device
is that the complete stress conditions are not known and
only a friction angle can be determined (i.e. p' is simply
unknown in ring shear). Some authors have used this
device to obtain shear strength at the critical state
(Negussey et al. 1988, Garga and Sedano 2002). The
more popular triaxial testing apparatus typically cannot
achieve the strains required to get to the critical state,
and at larger strains shear banding may occur, negating
the measured stresses. This has caused a tendency
among researchers to obtain critical state shear strength
properties using peak and pre-peak data obtained from
triaxial tests.
2.1 End of Test (ET) Method
A direct way of obtaining M tc from loose soil samples is
to plot the stress ratio η versus deviatoric strain ( εq) for a
drained test sheared up to about 20% of strain. The final
stress ratio of such a test is taken as Mtc for that test.
Figure 1 shows stress ratio versus deviatoric strain from
a drained triaxial test on a slightly dense sample of
Ticino sand. M tc parameter obtained from this method
will be denoted by (M tc) ET in the following; where ET
subscript stands for End of Test.
A major problem with this method is that we are again
dealing with data obtained from triaxial test in relatively
large strains where localization phenomenon can occur.
As shown in figure 1, at the end of the test neither the
stress ratio nor the volumetric strain has reached a
constant value. It is very unlikely that dense samples
reach the critical state at strains achievable in triaxial
apparatus; but for loose samples η and εv versus εq
plots almost level off at the end of the test indicating
being near the critical state conditions at that strain level.
2.2 Maximum Contraction (PT) Method
Maximum contraction during shearing of a soil sample is
identified as the point where the sample reaches its
minimum volume as illustrated in figure 1. At this point
only one of the conditions of critical state ε& v = 0
mentioned in equation 1 are satisfied.
Negussey et al. (1988) suggested obtaining the critical
state friction angle φcv from mobilized friction angle at
maximum contraction based on data including ring
shear tests on Ottawa sand, two tailings sand, granular
copper, lead shot and glass beads. M tc parameter
obtained from this method will be denoted by (M tc )MC in
the following; where PT subscript stands for Phase
Transformation.
2.3 Bishop Method (BM)
Bishop (1971) suggested a method of obtaining critical
state friction angle using results of several drained tests
on dense samples with varying densities. Each test is
then reduced to a value of peak measured dilatancy Dmin
at peak strength φ ′peak . For a sand that reaches the
critical state directly, i.e. without any dilation, the peak
stress ratio corresponds with the critical state of η = M tc
and Dmin=0. According to stress-dilatancy theory (Rowe
1962, Bolton 1986) peak dilatancy should occur at the
same point in the test as peak strength. Mtc is then
determined by extrapolation to zero dilatancy in order to
find the strength of the sample failing at zero rate of
volume change. It is very convenient to use stress ratio η
instead of φ′ to directly obtain the stress ratio at critical
state (M tc ). Thus, we plot ηmax instead of φ ′peak and the
corresponding dilation rate (peak dilatancy is Dmin ≤ 0
because of the compression-positive convention). Figure
2 shows the application of this method in ηmax - Dmin
space applied to Ticino sand.
As illustrated in figure 2, the intercept of this line is taken
as Mtc . The slope of this line (N-1) has been noticed by
some researchers as a material parameter (Nova 1982,
Jefferies 1993); N is called the volumetric coupling
parameter.
2.4 Stress-Dilatancy (SD) Method
By definition, the stress ratio at which no volume change
occurs (D=0) and remains there with continuing
shearing is the stress ratio at the critical state. Most
triaxial tests do not reach the critical state within the
strain limitations of the apparatus. Additionally, as
discussed earlier, localization may occur post peak
making the measured stress ratio unreliable. But it is
possible to infer M tc by extrapolating the post-peak
portion of a stress ratio versus dilatancy plot to zero
 dilatancy (the vertical axis where D=0 ). The post-peak
1.8
(hook) portion of the graph is usually linear, and this
1.6
1.4
(M tc )ET extrapolation to the critical state can be done by
assuming that a linear trend continues to the critical
1.2 (M tc )MC state. The line drawn through the stress-dilatancy plot in
η =(q/p')

1.0 Figure 3 illustrates how this method is applied to a

0.8 triaxial test on a dense sample.
0.6
0.4 Post peak data should be used with significant caution.
From an experimental point of view, shear banding and
0.2
localization may happen as soon as the sample leaves
0.0
peak strength towards critical state. In larger strains
0 5 10 15 20 25 (about 15% and more) other problems such as a higher
εq(%) effect of membrane penetration, tilting or bulging of the
sample can occur; hence the data should be looked at
very cautiously.
1
In this method the closer the test gets to the critical state,
0 without developing shear bands , or other sources of
error, the more accurate the extrapolation will be. In the
-1
case where the dilatancy path deviates from the linear
trend approaching the end of the test (see Figure 9), it is
εv(%)

-2
recommended that the initial part of the hook gets a
-3 higher weight in determining the location of the
extrapolation line, and the second part be discarded. In
-4 this work the slope of this line is identified as ( N ′ − 1 ).
-5

-6 3. OBTAINING THE CRITICAL STATE STRESS RATIO

OF TICINO SAND
Figure 1. Stress ratio and volumetric strain versus
Ticino sand is a medium to coarse predominantly quartz
deviatoric strain for Ticino sand (test 08-CID_D169) subangular sand. A summary of this sand's properties
and the test program is available in Been et al. (1987)
and Bellotti et al. (1996). Data presented here are from
1.8
test performed in Golder Associates Calgary laboratory
in 1987. All samples were dry pluviated and taken from
1.7 bulk samples known as Ticino 08 and 09 used for SBPT,
1
N −1 Bishop's Dilatometer and CPT chamber tests.
1.6
Line
3.1 Processing the Data
η max

1.5
TICINO-09 For Ticino sand Shuttle and Jefferies (1998)
suggest using
TICINO-08 1.4
TICINO-04
Ticino 08,09-M=1.33,N=0.4 1.3 0.48
M tc 6.5  p ′ 
Ticino 04-M=1.26,N=0.35 G=   [7]
1.2 e 1.3  p a 
-0.8 -0.6 -0.4 -0.2 0
Dmin
where p a is the atmospheric pressure. For th is test,
Figure 2. Application of Bishop's method to Ticino sand knowing that p'=300 kPa and e=0.686 at the beginning
of the test, we will have G=180MPa for the example
P
shown in figure 4 leads us to calculating D-D for the
whole test, which can be plotted against stress ratio as
shown in the figure.
 1.60
1.50
Dilatancy
N ′ −1
1.50 Plastic Dilatancy
1 D-Dp
1.00
1.40

η(q/p)
Figure 3
η (q/p)

(M tc )S D
1.30
0.50
Total Dilatancy
Plastic Dilatancy 1.20

D-Dp 0.00
1.10 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2
-0.4 -0.2 0.0 0.2 Dilatancy
Dilatancy Figure 4. Stress-dilatancy plot (η - D) for Ticino sand
Figure 3. Stress-dilatancy plot (η - D), around peak (test 09-CID_D169)

The MC method suggests that the hook finally falls on

Theoretically, the difference between total and plastic
P the maximum contraction point when the sample
dilatancy (D-D ) is equal to zero at the peak and at critical
reaches the critical state. None of the triaxial data
state. One important observation made in the figure 3 is
available demonstrate s such behaviour and as
that the difference between the total and plastic dilatancy
P illustrated in Figure 5, this method predicts a much lower
(D-D ) is very small between the two points where it
shear strength ((M tc ) PT =1.25) than the Stress Dilatancy
exactly equals zero. Hence, when the emphasis is on
(SD) and Bishop (BM) methods.
this part of the stress-dilatancy plot, it is possible to use
P
total dilatancy (D) instead of plastic dilatancy (D ). This is
(M tc ) SD = 1.345 is very close to (Mtc ) BM = 1.33 and the
very convenient because elastic modulus is not
difference is smaller than ± 0.03 resolution of M tc
accurately determined in all triaxial tests. Ideally, this
determination suggested by Jefferies and Been (2006).
should be done through application of bender elements.
In order to show how BM results compare to the results
Usually, it is also not known how the elastic modulus of
the sample varies during the test. Thus, total dilatancy of obtaining M tc from the stress-dilatancy plot, we need
P to plot the results in the same space as shown in Figure
(D) has been used instead of plastic dilatancy (D ) in the
rest of this work. 6.

It can be seen that the two tests for which the entire
Figure 5 shows a comparison between the Mtc
stress -dilatancy path is plotted in the figure, are in
parameters obtained using each of the four methods
complete coincidence with the line passing through the
described earlier; these values range from 1.25 for MC to
peak points. That is, the second part of stress-dilatancy
1.445 for ET with BM and SD yielding similar values of path (the hook) lies on this line. Consequently, Bishop
1.33 and 1.345 respectively.
m ethod yields same results the stress-dilatancy method
can yield. In other words, we have (M tc ) SD = (Mtc ) BM and
3.2 Discussion of Results for Ticino Sand
N=N'. In fact, N=N' is necessary for these two methods
to yield equal M tc values; otherwise (if N ≠ N ′) then the
The strength from the end of test method, (M tc ) ET =1.445
hook part of stress -dilatancy plot will leave Bishop's line
is likely overestimating Mtc for the dense exampl e used
after the peak and intersect D=0 axis at a different M tc .
here; the sample is not at the critical state at the end of
Since s tress-dilatancy plot for all of the ten tests show a
the test, as expected for the strains normally achieved in
relatively good agreement with Bishop's method, we can
triaxial tests. This is confirmed in Figure 5, where the
dilation rate is not zero at the end of the test. use Bishop's M tc=1.33 and N=0.4 for Ticino sand.
 4.1 Processing the Data
1.60
Dmin ,η max
It was shown that for Ticino sand ET and MC methods
cannot yield accurate M tc values due to their obvious
1.50 theoretical shortcomings. Hence, only BM and SD
(M tc )ET = 1.445 methods are applied to Erksak data and discussed
here.
1.40
(M tc )SD = 1.345 Dmin,ηmax pairs are plotted in figure 7 in order to locate
η (q/p)

Bishop's line. It seems quite reasonable to draw the

1.30 thinner line shown in the figure passing through six
(M tc )BM = 1.33 points with a very good resolution and ignoring the three
points which are off the line. Th is results in M tc=1.10 and
(M tc )MC = 1.25 N=0.065 ; the N value is lower than expected (usually
1.20
Total Dilatancy 0.2-0.4 ) and this leaves the chosen line with some
uncertainty. The best fit trend-line is also plotted in figure
7 accounting for all the nine tests. This results in
1.10 M tc =1.26 and N=0.41, which are more common values
-0.4 -0.2 0.0 0.2 for these two parameters. Still there is quite a scatter
Dilatancy 2
around this line (R =0.79) and it does not seem
Figure 5. Stress-dilatancy plot (η - D) for Ticino sand reasonable to merely rely on this line to obtain the critical
(test 09_CID169) state parameters. This example shows that Bishop's
method is very sensitive to selection of data and the
approach taken in drawing the line.

1.8 Trying to obtain an accurate critical state stress ratio

Dmin ,ηmax from of a limited number of tests, we now look at the
1.7
stress dilatancy plot for the entire test path. Figure 8
1.6 shows how the extrapolation is done for test G667 and
how the final part of the hook is ignored as described
1.5
earlier. The same process is applied to all nine tests,
ηmax

1.4 and M and N' parameters obtained are reported in Table

Peaks 1. The average values M tc =1.28 and N=0.40 are very
1.3
close to those obtained from the second approach
CID_D165 1.2 described above (best fit line) .
CID_D169 1.1 4.2 Discussion of Results for Erksak Sand
1
It is of interest to note that this set of nine triaxial tests on
-0.8 -0.6 -0.4 -0.2 0 0.2
dense samples would be considered a luxurious set of
Dmin
data in most engineering, and even academic, projects.
Figure 6. Stress-dilatancy plot (η - D) for Ticino sand; This makes it very important to be able to estimate the
including Dmin,η max critical state friction angle (or stress ratio) with
acceptable accuracy relying on a limited number of tests.
In this case, we have the virtue of having results of a
4. OBTAINING THE CRITICAL STATE STRESS RATIO research project (Vaid and Sasitharan 1992) on the
OF ERKSAK SAND same sand, which makes it possible to compare the two
proposed Bishop's lines as well as the average SD
Erksak sand is a uniformly graded, medium to fine grain values to a much more reliable (because of the large
predominantly quartz subrounded sand. A complete number of tests) reference. M tc =1.26 and N=0.37 are
description of this sand and the test program undertaken obtained from Figure 9; these are in good agreement
is available in Been et al. (1991). Nine of the tests with those obtained from fitting the best line passing
described by them are used here. These tests are through nine data points in BM and the average values
presented in table 1 and performed on wet pluviated obtained from SD method.
samples.
 1.8

1.6 1.7
Best fit line
considering all data
1.5
points 1.6

ηmax
1.4
1.5

ηmax
1.3
1.4
3 tests ignored in
identifying Bishop's line 1.2
1.3
1.1
1.2
1.0 -0.8 -0.6 -0.4 -0.2 0
-0.6 -0.4 -0.2 0 Dmin
Dmin
Figure 9. Stress-dilatancy (Dmin−ηmax) plot from triaxial
Figure 7. Stress-dilatancy (Dmin−ηmax) plot from triaxial data on Erksak 330/0.7 sand (data after Vaid and
data on Erksak 330/0.7 sand Sasitharan, 1992)

Table 1. Mtc and N' parameters for 9 wet pluviated

N ′ −1 1.6
triaxial tests on dense Erksak 330/0.7
1
1.5 p′0
Test name (kPa) OCR e M N'
η (q/p)

1.4 CID_G661 140 1.0 0.676 1.28 0.36

Bishop’s line
CID_G662 60 1.0 0.595 1.26 0.36
1.3 CID_G663 300 1.0 0.601 1.30 0.70
CID_G664 300 1.0 0.570 1.23 0.33
(M tc ) SD
1.2 CID_G665 130 1.0 0.610 1.18 0.30
CID_G666 60 1.0 0.637 1.24 0.20
1.1 CID_G667 130 1.0 0.527 1.38 0.41
-0.6 -0.4 -0.2 0 CID_G761 250 4.0 0.589 1.32 0.45
Dilatancy CID_G762 250 4.0 0.491 1.32 0.50
Figure 8. Stress-dilatancy plot (η - D) for Erksak 330/0.7 Average 1.28 0.40
sand (test CID_G667)

One important point about test G667 (and many other
tests) is that although the (Dmin−ηmax ) point neatly
coincides with Bishop’s line, the dilatancy path leaves
this line and ends up having a higher Mtc than that of
Bishop method; this also implies N ≠ N ′ . However, as
shown in Table 1, the average parameters are only
slightly higher than r esult of Bishop method.
Based on the above discussion we can choose
Mtc =1.27 and N=0.40 for this sand taking into account
both the Stress-Dilatancy method and Bishop's method.
5. CONCLUSION
Four methods of obtaining critical state friction angle are
briefly discussed and advantages and disadvantages of
each method are described based on previously
published data from drained triaxial tests on two well-
known sands.
Two of the methods (ET and MC) should not be
employed in determining the critical state stress ratio
from triaxial tests due to their apparent inconsistency
with definition of the critical state and the available data.
 A common geotechnical laboratory testing program
usually includes a limited number of acceptable drained
tests on dense samples. Using Bishop method of
obtaining the critical state stress ratio can result in
significant errors because it is probable that a wrong
trend line is picked up to fit a small number of data
points.
It is recommended that the whole stress-dilatancy path
of tests be observed. The stress-dilatancy method is
based on extrapolating the final part of Stress-Dilatancy
path to obtain the critical state stress ratio. These
obtained values should be compared to results of
Bishop's method to determine the final value of the
critical state stress ratio.
One important advantage of the Stress-Dilatancy method
is that it can yield the required parameters for every
single test, the average values can then be compared
with that of Bishop's method. This makes Stress -
Dilatancy method especially helpful when dealing with a
small number of acceptable tests.
In order to obtain the most accurate critical state shear
parameters it is necessary to apply engineering
judgement in two stages: when choosing acceptable
triaxial tests to be applied in either method; and in the
case of Stress-Dilatancy method, when choosing data
included in obtaining average parameter values. That is,
Mtc and N parameters which seem to be far away from
the average value should be excluded from the data.
6. ACKNOWLEDGEMENTS
The authors would like to thank Natural Sciences and
Engineering Research Council of Canada for funding the
research program leading to publication of this paper.
References
Been, K., and Jefferies, M. G. (1985). "A State Parameter
for Sands", Geotechnique, 35, 2, 99-112.
Been, K., Jefferies, M. G., Crooks J. H. A. and
Rothenberg, L. (1987). "The cone penetration test in
sands: Part II, general inference of state",
Geotechnique, 37, 285-299.
Been, K., Jefferies, M.G., and Hachey, J.E. (1991). "The
critical state of sands", Geotechnique, 41, 3, 365-381.
Bellotti, R., Jamiolkowski, M., Lo Presti, D. C. F. and
O'Neil, D. A. (1996). "Anisotropy of small strain
stiffness in Ticino sand", Geotechnique, 46 (1), 115 -
131.
Bishop, A. W. (1966). " Strength of soils as engineering
th
materials". 6 Rankine Lecture. Geotechnique 16,
89-130.
View publication stats
Bishop, A. W. (1971). " Shear strength parameters for
undisturbed and remoulded soil specimens". Proc.
Roscoe Memorial Symp., Cambridge, 3 -58.
Bolton, M. D. (1986). " Strength and dilatancy of Sands",
Geotechnique, 36, 1, 65-78.
Garga, V. K. and Sedano, J. A. I. (2002). "Steady State
Strength of Sands in a Constant Volume Ring Shear
Apparatus", Geotechnical Testing Journal, Vol. 25,
No. 4.
Jefferies, M. G. (1993). "Nor-sand: a simple critical state
model for sand", Geotechnicque, 43, 1, 91-103.
Jefferies, M. G. and Shuttle, D. A. (2002). "Dilatancy in
General Cambridge -Type Models", Geotechnique,
52, 625-638.
Jefferies, M.G. & Been, K. (2006); "Soil liquefaction: a
critical state approach". Taylor & Francis (Abingdon &
New York). ISBN 0-419-16170-8.
Klotz, E. U., Coop, M. R. (2002). "On the Identification of
Critical State Lines for Sands", Geotechnical Testing
Journal, Vol. 25, No. 3.
Negussey, D., Wijewickreme, W. K. D., and Vaid, Y. P.
(1988). "Constant Volume Friction Angle of Granular
Materials", Can. Geotech. J., Vol. 25, 50-55.
Poulos, S. J. (1981). "Th e Steady State of Deformation",
J. Geotech. Engng Am. Soc. Civ. Engrs, 17, G75, 553-
562.
Roscoe, K. H., Schofield, A. N. and Wroth, C. P. (1958).
"On the Yielding of Soils", Geotechnique, 8, 1, 22-53.
Rowe, P. W. (1962), "The Stress Dilatancy Relation for
Static Equilibrium of an assembly of particles in
Contact", Proc. Royal Society of London, A 269, 500-
527.
Shuttle, D. A. and Jefferies M. J. (1998). "Dimensionless
and Unbiased CPT Interpretation in Sand", Intl. J. for
Numerical and Analytical Methods in Geomechanics,
22, 351-391.
Vaid, Y.P., and Sasitharan, S. (1992). "The strength and
dilatancy of sand", Can. Geotech. J., 29, 522-526.