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