Professional Documents
Culture Documents
(Asce) 1532-3641 (2008) 8 2
(Asce) 1532-3641 (2008) 8 2
Abstract: On the basis of critical state soil mechanics, this study derives theoretical formulas for predicting the undrained shear strength
of K0 consolidated soft soils in triaxial compression and extension. Although the modified Cam-clay model is often utilized to predict the
undrained shear strength of soft clays, it is applicable mainly to isotropically consolidated soils. Because of the anisotropy under K0
consolidation, an inclined elliptical yield surface is chosen, which is different from those methods based on the original Cam-clay model.
Downloaded from ascelibrary.org by Monash University on 09/26/13. Copyright ASCE. For personal use only; all rights reserved.
The inclined elliptical yield surface is testified to be appropriate to the K0 consolidated soft soil and results in a better prediction of
undrained strength, especially for the triaxial extension test. It is concluded that the analytical solutions obtained in this paper are in good
agreement with the available test results and back-analysis of slope failures. On the basis of the investigation of soil properties, a simple
formula is proposed for calculating the mean undrained shear strength along the failure surface.
DOI: 10.1061/共ASCE兲1532-3641共2008兲8:2共105兲
CE Database subject headings: Soil mechanics; Soft soils; Shear strength; Soil consolidation.
Introduction of clay, they found that the mean values of strength gain ratio
su / ⬘vc, for normally consolidated and overconsolidated clays
The investigation of undrained shear strength of soft soils began 共OCR= 2兲 were 0.23 and 0.21, respectively. On the basis of back
in the 1940s when Casagrande proposed the QRS methodology. analyses on 15 embankment failures 共for clays I p ⬍ 60%兲, Larsson
Skempton 共1954, 1957兲 presented experimental data for a number 共1980兲 concluded that the value of su / ⬘vc is 0.23 with a standard
of normally consolidated soils and deduced the relationship be- deviation of 0.04.
tween vane shear strength gain ratio su / ⬘v0 共⬘v0 means the effec- It must be pointed out that these findings mentioned above are
tive overburden pressure兲 and plasticity index. With gathering mostly based on investigation and back analysis of a large number
more experimental data, the data show a greater variety, and the of embankment failures. In other words, they are the overall val-
character of the dependence of su / ⬘v0 on plasticity index becomes ues that are mobilized along the failure surface. The measured
less clear. Bjerrum 共1972, 1973兲 analyzed a large number of field undrained shear strengths 共such as field vane shear strength or
data and proposed a nonlinear relationship between strength gain triaxial compression strength兲 are usually corrected by a coeffi-
ratio su / ⬘vc 共⬘vc means vertical preconsolidation pressure兲 and cient for application in slope stability analysis. Many researchers
plasticity index. Bjerrum 共1972兲 back-analyzed many case histo- emphasized that for the stability analysis of a soft soil slope 共Fig.
ries of embankment failure on soft ground and introduced a cor- 1兲, it is unreasonable to calculate the safety factor purely from the
rection factor. Bjerrum 共1973兲 emphasized that this correction shear strength measured by triaxial compression tests. On the
factor may arise partly from the effects of anisotropy of natural sliding surface, there are three different stress paths, including
soils and partly from the dependence of strength on rate of shear. triaxial compression 共TXC兲, direct simple shear 共DSS兲, and tri-
Mesri 共1975兲 analyzed the field vane strength of a number of late axial extension 共TXE兲, which correspond to three kinds of un-
glacial clays which involved both normally and overconsolidated drained shear strength 共sutxc, sudss, and sutxe兲. Ladd and Foott
soil samples. After introducing Bjerrum’s correction factor, Mesri 共1974兲 and Ladd 共1991兲 emphasized that the K0 consolidation
showed that the strength gain ratio su / ⬘vc, is roughly a constant condition should be considered for undrained shear strength
value of 0.22. Ladd and Foott 共1974兲 proposed the stress history analysis of soil.
and normalized soils engineering properties 共SHANSEP兲 method. Much effort has been made for modeling the anisotropy of
After carrying out the CK0U direct simple shear test for five kinds undrained shear strength since the 1940s. The anisotropy of the
undrained shear strength is considered to have two components:
1 the inherent anisotropy and the stress-induced anisotropy. The
Professor, Institute of Geotechnical Engineering, Zhejiang Univ.,
Zheda Rd. 38, Hangzhou 310027, China 共corresponding author兲. E-mail: inherent anisotropy results from the anisotropy of the soil skel-
wizzju@163.com eton structure such as the preferred sedimentation orientation of
2 soil particles, and the stress-induced anisotropy rises from the
Designer, Institute of Geotechnical Engineering, Zhejiang Univ.,
Zheda Rd. 38, Hangzhou 310027, China. anisotropic consolidation stress state. Many researchers have pro-
3
Designer, Institute of Geotechnical Engineering, Zhejiang Univ., posed the relationship between undrained shear strength and di-
Zheda Rd. 38, Hangzhou 310027, China. rection of shear 共Hansen and Gibson 1949; Nakase and Kamei
Note. Discussion open until September 1, 2008. Separate discussions 1983兲, but most of which are empirical. Prevost 共1979兲 first for-
must be submitted for individual papers. To extend the closing date by
mulated the undrained shear strength on the basis of a constitutive
one month, a written request must be filed with the ASCE Managing
Editor. The manuscript for this paper was submitted for review and pos- equation. Ohta and Nishihara 共1985兲 made use of the original cam
sible publication on April 3, 2006; approved on July 30, 2007. This paper clay model and derived the theoretical formula for undrained
is part of the International Journal of Geomechanics, Vol. 8, No. 2, shear strength which involved K0 consolidation-induced aniso-
April 1, 2008. ©ASCE, ISSN 1532-3641/2008/2-105–113/$25.00. tropy. Zdravkovic et al. 共2002兲 examined the stability of Saint-
p⬘c
p⬘ M2 v = v + ln p⬘ = N − 共 − 兲ln 共3兲
= 2 共1兲 p⬘
p⬘c M + 2
Substitution of Eq. 共1兲 into Eq. 共3兲 yields
where p⬘⫽effective mean stress corresponding to the current
state, p⬘c ⫽effective mean stress for normally consolidated soils M 2 + 2
indicating the reference size of the yield surface; and = q / p⬘ in v = N − 共 − 兲ln 共4兲
M2
which q⫽deviator stress.
According to Fig. 2共a兲, the specific volume at point II can be In accordance with the MCC model, we have the dilatancy of
obtained by isotropic consolidation soil
共14兲 where
path BC is also an undrained procedure, where we have ⌬v = 0,
which means
d*
dp⬘
3 sij
=− *
2 p⬘
冉
− ␣0
sijm sij
pm⬘ p⬘2
冊
pb⬘ M 2 − 共␣0K0nc兲2 + *b 2 Introduction of it into Eq. 共23兲 leads to
ln + 共 − 兲ln 2 =0 共15兲
冉 冊
p⬘ M − 共␣0K0nc兲2 + *2
sij sijm sij M 2 − 共␣0K0nc兲2 + *2
Since point B is on the initial yield surface given by Eq. 共8兲, the − ␣0 = 共24兲
p⬘ pm⬘ p⬘2 3
stress state at B must satisfy
For isotropic consolidation sijm = 0, Eq. 共24兲 becomes
p⬘b M 2 − 共␣0K0nc兲2
= 共16兲 = ±M 共25兲
M 2 − 共␣0K0nc兲2 + *b 2
Downloaded from ascelibrary.org by Monash University on 09/26/13. Copyright ASCE. For personal use only; all rights reserved.
p⬘c
This is the failure condition for isotropic consolidation soil which
Eliminating *b from Eqs. 共15兲 and 共16兲 yields is commonly used in critical state soil mechanics.
ln
p⬘b
p⬘
+ 共 − 兲ln
p⬘c
· 2冋
M 2 − 共␣0K0nc兲2
p⬘b M − 共␣0K0nc兲2 + *2
=0 册 共17兲
Undrained Shear Strength under Triaxial Stress States
Because p⬘b = pi⬘ = p⬘0 and the stress state at A also satisfies Eq. 共8兲, The triaxial stress condition is given as follows
Eq. 共17兲 becomes r⬘ = ⬘ , zr = z = r = 0
共26兲
p⬘0 ⬘ = ⬘0 = K0z0
r0 ⬘, zr0 = z0 = r0 = 0
ln + 共 − 兲
p⬘ * can be rewritten by substituting Eq. 共26兲 into its expression as
⫻ln n · 冋 M 2
+ 2K0nc
2
M −
− 2␣02K0nc
共␣0K0nc兲2
·
M − 共␣0K0nc兲
2 2
M − 共␣0K0nc兲2 + *2
2 =0 册 * = ± 冉 z⬘ − r⬘
p⬘
− ␣0
⬘ − rm
zm
pm⬘
⬘
冊
= ± 共 − ␣0K0nc兲 共27兲
共18兲
in which the plus sign represents the triaxial compression test
in which n⫽overconsolidation ratio in terms of the effective mean 共z⬘ ⬎ r⬘兲 and the minus sign represents the triaxial extension test
principle stress, i.e., n = pm⬘ / p⬘0 = pm⬘ / p⬘b = pm⬘ / pI⬘ 共see Fig. 4兲. Then 共z⬘ ⬍ r⬘兲, where = q / p⬘ = 共z⬘ − r⬘兲 / p⬘ and K0nc = qm / pm⬘ = 共zm
⬘
Eq. 共18兲 can be rewritten as ⬘ 兲 / pm⬘ = 3共1 − K0兲 / 共1 + 2K0兲.
− rm
冉冊冋 册
共−兲
The failure condition under triaxial stress condition can be
p⬘0 M 2 + 2K0nc − 2␣02K0nc M 2 − 共␣0K0nc兲2 obtained by substituting Eqs. 共26兲 and 共27兲 into Eq. 共24兲 as
· ·
p⬘ M 2 − 共␣0K0nc兲2 M 2 − 共␣0K0nc兲2 + *2 follows
· n共−兲 = 1 共19兲 = ±M 共28兲
when n = 1 in Eq. 共18兲, the undrained stress path for normally which is the same as the failure condition for isotropic consoli-
consolidated soil can be obtained by dation soil, as shown in Eq. 共25兲. This coincidence of isotropic
冉冊冋 册
共−兲 and anisotropic consolidation soil indicates that the critical state
p⬘0 M 2 + 2K0nc − 2␣02K0nc M 2 − 共␣0K0nc兲2 line in the p – q plane is independent of consolidation history and
· · 2 =1
p⬘ M − 共␣0K0nc兲
2 2
M − 共␣0K0nc兲2 + *2 stress path. The failure condition Eq. 共28兲 can also be rewritten as
共20兲 ±q
p⬘ = 共29兲
M
The intersection between the critical state line and undrained
Formula for Predicting Undrained Shear Strength
stress path is the failure point. Hence eliminating p⬘ by substitut-
ing Eqs. 共27兲 and 共29兲 into Eq. 共19兲 we have
冋 册
Failure Condition ⌳
q M 2 + 2K0nc − 2␣02K0nc M ± ␣0K0nc
The failure condition of the volumetric hardening plastic model is = ±M · · n⌳
p0⬘ M 2 − 共␣0K0nc兲2 2M
given by
共30兲
f f
d⑀vp = ␦ij = =0 共21兲 where
ij⬘ p⬘
−
where ⫽scalar plasticity multiplier and ⬎ 0; and f⫽yield ⌳= 共31兲
function given by
The undrained shear strength is defined by
p⬘ M 2 − 共␣0K0nc兲2
f= − 2 共22兲 兩q兩
p⬘c M − 共␣0K0nc兲2 + *2 su = 共32兲
2
The failure condition could be obtained by substituting Eq.
共22兲 into Eq. 共21兲 as Substituting Eq. 共32兲 into Eq. 共30兲, we obtain
=
p⬘0 2
冋
su M M 2 + 2K0nc − 2␣02K0nc M ± ␣0K0nc
M 2 − 共␣0K0nc兲2
·
2M
册 ⌳
· n⌳ 共33兲
Substituting Eq. 共26兲 into Eq. 共8兲 we have the yield function
p⬘0 =
1 + 2K0
3
· OCR · ⬘v0 共37兲 冉 冊
su
⬘v0 NCI
=
M
2⌳+1
共41兲
冉 冊 冉 冊
which is identical to the equation of the MCC model proposed by
su su ¯
⌳ Wood 共1990兲.
= OCR · 共38兲
⬘v0 OCA ⬘m
v NCA
冉 冊
parameters of the soils are listed in Table 1. There is one normally
su 1 + 2K0 M
= · consolidated sample and a few overconsolidated samples in each
⬘v0 3 2 group. In Fig. 6, the abscissa is the OCR and the ordinate is the
冋 册
NCA
M 2 + 2K0nc − 2␣02K0nc M ± ␣0K0nc ⌳ undrained strength gain ratio. The upper theoretical lines are ob-
⫻ · tained from Eqs. 共38兲 and 共39兲, while the lower ones are obtained
M 2 − 共␣0K0nc兲2 2M from the theory of Ohta and Nishihara 共1985兲. In triaxial com-
共39兲 pression it is seen that the results predicted by the formula pro-
posed in this paper and Ohta and Nishihara 共1985兲 are both close
Eqs. 共38兲 and 共39兲 give the formula for calculating the undrained
to the test results.
compression and extension shear strength under K0 consolidation
Fig. 7 shows the results of K0 consolidated triaxial extension
conditions. For isotropic consolidation soil, n = OCR; = ¯;
tests on three Japanese natural clays 共Shogaki and Nochikawa
K0nc = 0; and K0 = 1. Thus the undrained shear strength of isotro-
2004兲, and the relevant parameters are listed in Table 1 共Number
pic overconsolidated soil can be obtained by
13兲. As the parameter ⌳ is not reported, the authors adopt
冉 冊 su
⬘v0 OCI
= n⌳ · 冉 冊
su
v⬘m NCI
共40兲
Karube’s empirical equation: ⌳ = M / 1.75 as suggested by Ohta
and Nishihara 共1985兲. It is seen from Fig. 7 that the deviation
between the experimental values and predictions obtained from
where OCI indicates the isotropic overconsolidation, while NCI equations presented in this paper is much less than those obtained
denotes the isotropic normal consolidation. Thus the undrained from Ohta and Nishihara 共1985兲.
shear strength of isotropic normally consolidated soil can be ex- Fig. 8 shows a comparison of experimental and calculated
pressed as undrained shear strength of the normally K0 consolidated triaxial
tests reported in past publications, and the reference and soil pa-
rameters are listed in Table 1. Karube’s empirical equation is
employed in the case where the values of ⌳ are not fully reported
共except for Numbers 1 and 5兲 for calculating shear strength.
When the values of ⌳ ¯ are not reported, they are assumed to be
equal to ⌳ as Ohta and Nishihara 共1985兲 found. The abscissa is
the measured value and the ordinate is the theoretical value. It can
be seen from Fig. 8 that for triaxial compression the predictions
obtained in this study are slightly larger than the measured values,
while the predictions obtained by Ohta and Nishihara 共1985兲 are
slightly smaller; for triaxial extension there is a good agreement
between the predictions obtained in this study and the measured
values, whereas the calculated values of Ohta and Nishihara
共1985兲 are apparently less than the measured values.
It is found from Eq. 共39兲 that the undrained shear strength gain
ratio of the normally K0 consolidated clay is a function of K0nc,
Fig. 7. Relationship between su / ⬘v0 共CK0UE兲 and OCR 共experimen- K0, M, and ⌳. The initial stress ratio K0nc is related to K0 as
tal data from Ohta and Nishihara 1985兲: 共a兲 Mito clay; 共b兲 Kimhae K0nc = 3共1 − K0兲 / 共1 + 2K0兲. The earth pressure coefficient at rest
clay; and 共c兲 Kahokugata clay K0 could be related to the critical state frictional angle ⬘ as K0
= 1 − sin ⬘. According to critical state soil mechanics, the param-
Conclusion
On the basis of critical state soil mechanics and the fact that the
initial yield surface is a rotated and distorted ellipse on the p⬘ – q
plane, this paper takes K0 consolidation-induced anisotropy into
account and derives a theoretical formula for predicting the un-
Fig. 9. Plot of su / ⬘v0 versus ⬘, ⌳ of normally consolidated clays drained shear strength under the stress path related to triaxial
Downloaded from ascelibrary.org by Monash University on 09/26/13. Copyright ASCE. For personal use only; all rights reserved.
where ␣⫽correction factor which takes into account the effect of Acknowledgments
the rate of shear. Many embankment failures investigated by Bjer-
rum 共1972, 1973兲, Mesri 共1989兲, and Larsson 共1980兲 were located This research is sponsored by the Chinese National Science Foun-
dation 共Grant No. 50479045兲.
Mayne, P. W. 共1980兲. “Cam-clay predictions of undrained strength.” J. Skempton, A. W. 共1957兲. “Discussion: The planning and design of the
Geotech. Engrg. Div., 106共11兲, 1219–1242. new Hong Kong airport.” Proc., ICE, London, 7, 305–307.
Mesri, G. 共1975兲. “Discussion of ‘New design procedure for stability of Wheeler, S. J., et al. 共2003兲. “An anisotropic elastoplastic model for soft
soft clays’ by Ladd, C. C, and Foott, R.” J. Geotech. Engrg. Div., clays.” Can. Geotech. J., 40, 403–418.
101共4兲, 409–412. Wood, D. M. 共1990兲. Soil behavior and critical state soil mechanics,
Mesri, G. 共1989兲. “Reevaluation of su共mob兲 = 0.22⬘p using laboratory shear Cambridge University Press, New York.
test.” Can. Geotech. J., 26, 162–164. Zdravkovic, L., Potts, D. M., and Hight, D. M. 共2002兲. “The effect of
Mitachi, T., and Kitago, S. 共1976兲. “Changed undrained shear strength strength anisotropy on the behaviour of embankments ground.” Geo-
characteristics of saturated remoulded clay due to swelling.” Soils technique, 52共6兲, 447–457.