Phenomenological Issues Related To Strain Localization in Sensitive Clays
Phenomenological Issues Related To Strain Localization in Sensitive Clays
Phenomenological Issues Related To Strain Localization in Sensitive Clays
DOI 10.1007/s10706-005-5818-z
Abstract. A negative second order work, strain softening, is often noticed in contractant
material like sensitive clays. Failure in such clays will lead to the formation of localized
deformation zone of intense inelastic strain, known as shear band. Conditions, emergence and
inclination of shear band has been very well demonstrated in past decades in different man-
ners, however a definite thickness of shear band is still an open question due to several reasons.
Mesh dependency, loss of ellipticity is another challenge associated with finite element analyses
for strain softening clays. This paper covers a comprehensive review of classical theories of
strain localization and associated limitation. Mesh dependency, ill-possed boundary value
problem is addressed using finite element simulation examples and experimental results.
Key words. finite element analysis, localization theory, quick clay, shear band, strain softening.
1. Introduction
Some soils display a decrease in shear resistance with increasing strain from a peak
value to an ultimate or residual value. This behaviour, recognized as strain softening
or strain weakening, is very pronounced in soft and sensitive clays. Strain-softening
behaviour may also be seen in stiff over consolidated clays (Bjerrum, 1967,
Skempton, 1964). The general understanding is that stiff over consolidated soils tend
to dilate and loose contracts, when sheared under drained conditions.
Strain localization theory was firstly proposed by Hadamard (1903), further, Hill’s
stability criteria (1962) and pioneering contribution from Thomas (1961), Mandel
(1963), Rudnicki and Rice (1975), Rice (1976) and Vardoulakis (1980) helped
in order to understand the localization theory. Several efforts have been made to
describe the orientation and thickness of shear band in different geo-materials.
Muhlhaus and Vardoulakis (1987) suggested criteria to measure the thickness of
shear band in granular materials, which is 8 to 10 times of mean grain size; however a
definite thickness of shear band in clay is still an open issue. On the other hand,
w
Vikas Thakur, Geotechnical Division, Department of Civil and Transport Engineering, Norwegian
University of Science and Technology (NTNU), Hogskoleringen 7a, Gloshaugen, Trondheim, Norway,
NO 7491. (tel.: 0047 7359 6871; fax: 0047 7359 4609; e-mail: vikas.thakur@ntnu.no)
1730 VIKAS THAKUR ET AL.
2. Strain Localization
2.1. STABILITY CRITERION
There is no guarantee that an equilibrium state of a non linear system is stable. An
equilibrium state is considered to be stable if the response of a vanishingly small
disturbance also remains vanishingly small (Hill, 1959; Koiter, 1960; Jostad, 1993).
Sensitive clays mainly behave as softening material and hence this paper will only
discuss about localization phenomenon under strain softening condition. Further,
strain softening is possible either due to structural i.e. geometric instability or
material instability. In common geotechnical engineering practice, structural i.e.
geometric instability like buckling or snap through is ignored. However, fissuring,
and cracking is always taken in account as material softening. On deriving constit-
utive model from experimental data, material softening needs to be isolated from
geometrical softening otherwise such models would obtain an overdose of material
softening (Vermeer and Brinkgreve, 1994).
Figure 1(a) shows the three well know different equilibrium condition; stable,
unstable and neutral. It must be noted that, a same system can pass through all these
equilibrium condition depending upon stress state and amount of energy dissipated
i.e. state of second order work, refer Figure 1(b).
Figure 1. (a) Equilibrium conditions. (b) Stress (r) –strain (e) response in softening material.
1732 VIKAS THAKUR ET AL.
Incremental internal energy based Hill’s stability criteria (1959) is widely recog-
nized and accepted to analyses the body undergoing through deformation (Jostad,
1993).
Z
1
Incremental internal energy dWint ¼ drij deji dV
2
v0
Neutral if dWint ¼ 0 for at least one dx and remain stable for the rest ð3Þ
where, dx is incremental displacement
Here it is important to remember that Hill’s stability criterion is necessary but not
a guaranteed condition for stability. However, to have a guaranteed stability crite-
rion, material must follow the constitutive condition.
drij deij ¼ Dijkl dekl deij 0
where, Dijkl is tangential stiffness matrix.
In sensitive clays follows non associated flow rule and thus tangential stiffness
matrix is very much influenced by angle of dilatancy (w) and thus rate of softening is
also function of angle of dilatancy. To demonstrate the affect of w, a 2 dimensional
simple box subjected to pure shear is considered and simulated in finite element code,
PLAXIS, as shown in Figure 2. Mohr–Coulomb model, elastic perfectly plastic
model with negative dilatancy is considered. Sensitive clays are less permeable and
hence analysis is made under undrained condition as pure frictional materal. A shear
modulus (G) of 5000 kPa, poison’s ratio (v) 0.25 and friction angle of 30 degree is
considered.
A finite element model is simulated to validate analytical equation (4). A higher pore
pressure modulus represents a higher dilating behaviour i.e. contraction or dilation.
It is found that proposed equation and finite element simulation is having quite
significant match as shown in Figure 4.
3. Localization Theory
Strain localization theory can be well demonstrated by widely recognized bifurcation
analysis.
displacement rate can be obtained. However, this classical theory has three impor-
tant limitation: first, theory analyses only the emergence of shear band but not its
development, second there is no guarantee that shear band will actually emerge and
last that thickness of shear band is left undetermined due to lack of internal length
characteristics.
Shear band formation may be treated as a bifurcation problem within the
framework of continuum mechanics. An infinite body subjected to uniform stress
field may emerge a shear band under the localized deformation condition mentioned
by Thomas (1961), Hill (1962), Mandel (1964), Rudniki and Rice (1975), Vardou-
lakis (1980) and Bardet (1990).
To describe kinematics of deformation, let us consider a body with a shear band of
small thickness, tsb. Shear band is embedded between two singular surface Xþ and
X with a unit normal n. The infinite body is subjected to uniform Cauchy stress r.
@vi 1 1
Lij ¼ Uij ¼ ðLij þ Lji Þ Wij ¼ ðLij Lji Þ ð5Þ
@xj 2 2
where vi, velocity; Lij, velocity gradient and Uij, Wij are deformation and spin tensor,
respectively. As long as body is subjected to uniform deformation, Lij will remain
same throughout body. But on set of localization, velocity gradient inside band, Lijb,
will no longer equal to Lij:
Lbij ¼ Lij þ DLij Ubij ¼ Uij þ DUij Wbij ¼ Wij þ DWij ð6Þ
PHENOMENOLOGICAL ISSUES RELATED TO STRAIN LOCALIZATION 1735
where
1 1
DLij ¼ xi nj DUij ¼ ðni xj þ nj xi Þ DWij ¼ ðxi nj xj ni Þ ð7Þ
2 2
xj is the component of a function x which depends on the distance across the band and
vanishes outside the band. Assume, X planer boundary is the datum line shear band.
The deformation variation across the band can be written as:
8
>
< uþ
d~ i nj xj > tsb : ðregion Xþ Þ
ui ¼
d~ 1
tsb uþ
nj xj ðd~ u
i d~ i Þ 0 nj xj tsb ð8Þ
>
:
u
d~ i nj xj 0 : ðregion X Þ
1736 VIKAS THAKUR ET AL.
The static incremental equilibrium condition, stress traction outside and inside band
must be equal:
dtbi þ dti ¼ 0 ð9Þ
However, it must be noted that equation 9 gives an equilibrium condition in the
direction of unit normal n, but stresses along the shear band is not necessarily be
equal and that may cause localization. Solution from Rudniki and Rice (1975),
Bardet (1990) and Jostad (1993) proposed a constitutive law to fulfil for bifurcation
condition. The possibility for the existence of a shear band to be when:
drij ¼ Dijkl dekl
ð10Þ
detðnj Dijkl nl Þ ¼ det H ¼ 0
The expression H is denoted as acoustics tensor, whose vanishing determinant gives
a necessary (but not sufficient) condition for emergence of shear band.
Classical continuum mechanics is not able to handle bifurcation problem related
to strain softening. This is one of the reasons why finite element methods are not very
much suitable to do complete analysis of shear band. Mesh dependent results can be
obtained and always thickness of shear band depends on element size.
W ¼ Bð~s dÞ
where,
0 1
Z
B1 C
~s ¼ @ sdcA ð13Þ
c
c
1738 VIKAS THAKUR ET AL.
The average shear stress ~s, is inversely proportional to shear strain, c, which means
higher shearing will reduce the total work done. In other words, two different stages
of shearing, from 1 and 2, refer Figure 7, will reduce the work done will be obtain or
higher energy dissipation is expected. Since a continuous shearing (tends to infinity)
will lead to more and more reduction in work done and eventually a strong
discontinuity will emerge in a form of slip line.
Mohr–Coulomb solution use hypothesis that the rupture planes are parallel to the
planes subjected to the principal stress combination; however the Roscoe solution is
application to kinematically admissible body. The later one has been more widely
accepted by the researcher due to the fact that shear band goes under a continuous
deformation and also a kinematics equilibrium condition has to satisfy to obtain
bifurcation condition. However, none of these equations guaranteed that shear band
will actually follow the same angle while propagating. These solutions are only valid
at the point of emergence of shear band and critical propagation angle will change as
a function of time and distance.
experience the plastic strain and rest of the body will be elastically unloaded.
Figure 9 represents a separate elastic and plastic strain in element.
Considering a plastic flow in local coordinates (n, t) a cumulative strain within
shear band can be specified as equation (18).
2 3 2 e 3 2 p 3 2 @ut 3
det det des @t
4 den 5 ¼ 4 deen 5 þ 4 dep 5 ¼ 6 4 @un 7
5 ð18Þ
n @n
dcnt dcent dcpnt @un
þ @ut
@t @n
dee ¼ C1 dr
8 e 9 2 3K þ 4G 3K þ 2G
38 9
< den = 1 12K þ 4G 12K þ 4G 0 < drn =
deet ¼ 4 3K þ 4G
0 5 drt ð19Þ
: e ; G 12K þ 4G : ;
dcnt SYM 1 ds
And assuming an associated flow rule incremental plastic strain can be written as;
1 @fT @f
dep ¼ dr
A @r @r
Here A is softening modulus;
2 2 3
8 p9 @f @f @f 8 9
< n = 1 6 @rn
de 0 @rn @s 7< drn =
dep ¼ 6 0 0 0 7 drt ð20Þ
: tp ; A 4 @f @f 2 5: ;
dc 0 @f ds
@rn @s @s
de ¼ D1 dr
1 1 @f @f T
D¼ C þ ð21Þ
A @r @r
The tangential stiffness matrix [D] or [D]G referred to the local co-ordinates (n, t),
transforming this matrix to the global co-ordinate;
½DG ¼ ½TT ½DL ½T ð24Þ
where
2 3
cos2 h0 sin2 h0 sin h0 cos h0
½T ¼ 4 sin2 h0 cos2 h0 sin h0 cos h0 5 ð25Þ
2sin h0 cos h0 2sin h0 cos h0 cos2 h0 sin2 h0
h0 need not be similar to h, measured with respect to a local coordinates.
Component D33, represents the tangential shear modulus and a shear band may
develop when it is equal to zero. Jostad and Andresen (2004) reported that the shear
band formation may be prevented due to change in rn . However, D33 close to zero
would reduce the rn increase at the front of propagating shear band.
Figure 10. Strain localization in clay rock after Viggiani et al. (2004).
higher water content than its liquid limit, as shown in Figure 12. In past, enormous
amount of research has been done to understand the behaviour of Norwegian quick
clay. Some of the classical work done by Bjerrum (1954, 1955, 1967), Bjerrum and
Landva (1966) can be referred.
A Consolidated undrained (CU) test is performed to unveil the behaviour of
Norwegian quick clay. A standard 54 mm diameter and 10 mm high sample was
chosen and rate of shear was 0.05 mm/min, From Figure 13, a clear strain softening
response and brittle failure can be noted i.e., a rapid decrease in shear resistance after
touching yield surface.
In order to investigate localization authors (2005) have presented a set of
unconfined compression test at different strain rates. Instrumental set up is depicted
in Figure 14.
A simple methodology was adopted to scan shear localization using a video
extensometer. A series of small dots were imprint, see Figure 15, on specimen surface
Figure 11. Sketch of normal marine clay (left) and quick clay (right) on a micro structural level.
PHENOMENOLOGICAL ISSUES RELATED TO STRAIN LOCALIZATION 1743
and traced throughout the test duration. Video extensometer helped in gathering sets
of deformation tensor for each dots.
Onset of localization, a microscopic observation is performed, which concludes
that shear band thickness is far below than 1 mm, in Norwegian quick clay as shown
in Figure 16. A series of test were performed and most of the sample had localization
in a form of strong discontinuity i.e. slip surface. Also the orientation of shear band
was approximately 37 degrees.
Figure 14. Instrumental set up 1: Video extensometer; 2: sample; 3: Uniaxial device; 4: sample with
markers (dots); 5: to data logger.
Figure 17 depicts the growth of incremental shear strain during the test. At the
beginning of test, when specimen behaves as an elastic body has uniform shear
strain. However, on set of localization, a distinct shear zone emerges which contains
all plastic shear strain and rest of the body undergone for elastic unloading.
PHENOMENOLOGICAL ISSUES RELATED TO STRAIN LOCALIZATION 1745
Figure 16. Microscopic view of shear band in failure plane, after Grimstad et al. (2005).
Figure 17. Incremental shear strain distribution plots at different strain level.
6. Conclusions
In this paper, classical stability criteria and localization theory is highlight. A sum-
mery of shear band analysis using bifurcation theory can be concluded as;
(a) The shear band criterion is necessary but not a guaranteed that shear band will
actually emerge.
(b) Orientation of shear band is not unique.
(c) Thickness of shear band left undetermined.
(d) Coulomb and Roscoe solution are lower and upper bound for shear band
respectively.
1746 VIKAS THAKUR ET AL.
Acknowledgement
International Centre for Geohazards (ICG), Norway is gratefully acknowledged for their
support and supervision. The Geotechnical division, Norwegian University of Science
and Technology, NTNU, is acknowledging with utmost respect for financial support.
References
Andresen, L. and Jostad, H.P. (2002) A constitutive model for anisotropic and strain-softening
clay, In Proceedings of the Numerical Models in Geomechanics – NUMOG VIII, Rome, Italy,
pp. 581–585.
Arthur, J.F.R., Dunstan, T., Assadi, Q.A.J. and Assadi, A. (1977) Plastic deformation and
failure in granular material, Geotechnique, pp. 53–74.
Bardet, J.P. (1990) A comprehensive review of strain localization in elastoplastic soils,
Computers and Geotechnics, 10, 163–188.
Benallal, A., Billardon, R. and Geymonat, G. (1989) Some methamatical aspects of the damage
softening rate problem Cracking and Damage, Elsevier, Amsterdam, pp. 247–258.
Bjerrum, L. and Landva, A. (1966) Direct simple shear test on a Norwegian quick clay,
Geotechnique, 16(1), 1–20.
Bjerrum, L. (1954) Geotechnlcal properties of Norwegian marine clays, Geotechnique, 3(21),
49–69.
Bjerrum, L. (1955) Stability of natural slopes in quick clay, Geotechnique, 5(1), 101–119.
Bjerrum, L. (1967) Seventh Rankine lecture: Engineering geology of Norwegian normally
consolidated marine clays as related to settlement of buildings, Geotechnique, 17(2), 81–118.
de Brost, R., Sluys, L.J., Mulhaus, H.B. and Pamin, J. (1993) Fundamental issues in finite
element analyses of localization of deformation, Engineering and Computer, 10, 99–121.
PHENOMENOLOGICAL ISSUES RELATED TO STRAIN LOCALIZATION 1747
Desrus, J. (1984) La localization de la dé formation dans les milieuz granularis, Théses de
doctorat d’état, Grenoble.
Grimstad, G., Thakur, V. and Nordal, S. (2005) Experimental observation on formation and
propagation of shear zone in norwegian quick clay, 11th International conference and field
trip on landslides, Trondheim, Norway.
Hadamard, J. (1903) Lecons sur la propagation des ondes et les e´quations de I’ hydrodynamique,
Librarie scientifique, A Hermann, Paris.
Hill, J. (1962) Accelaration waves in solids, Journal of Mech. Physics of Solids, 10, 1–16.
Høeg, K. (1972) Finite element analysis of strain softening clay, Journal of Soil Mech. and
Foundation Engineering, ASCE, 98(sm 1), 43–58.
Jostad, H.P. and Andresen, L. (2004) Modelling of shear band propagation in clays using
interface elements with finite thickness NUMOG, Canada.
Jostad, H.P. (1993) Bifurcation analysis of friction materials, Trondheim, Norway. Ph D thesis,
NTH, Norges TekniskE Høgskole.
Koiter, T.W. (1960) General Theorems for Elastic-Plastic Solids, In: Sneddon J.N. and Hill R.
(eds.). Progress in Solid Mechanics 1, North-Holland, Amsterdam, pp. 165–221.
Mandel, J. (1964) Condition de stabilite’ et postulate de Drucker, IUTAM Symposium on
Rheology and Soil Mechanics, Grenoble, pp. 58–68.
Morgenstern, N.R. and Tschalenko, J.S. (1967) Microscopic structures in kaolin subjected to
direct shear, Geotechnique, 17, 309–328.
Muhlhaus, H.B. and Vardoulakis, I. (1987) The thickness of shear band in granular materials,
Geotechnique, 37, 271–283.
Pietrusczak, and Mroz, (1981) Finite element analysis of deformation of strain softening
materials, International Nour. Num. Methods in Engineering, 17, 327–334.
Rice, J.R. (1976) The localization of plastic deformation, In Proceedings 14th Int. Congress on
Theoretical and Applied Mechanics, Delft, (edited by W.T. Koiter), Vol 1, pp. 207–220.
Roscoe, K.H (1970) The influence of strain in soil mechanics, Geotechnique, 20(2), 129–170.
Rudnicki, J.W. and Rice, J.R. (1975) Conditions for the localization of the deformation in
pressure-sensitive dilatant material, Journal of Mech. Physics of Solids, 23, 371–394.
Thakur, V. Jostad, H.P., Nordal, S. and Andresen, L. (2005) Kinematics of shear zone
deformation in Norwegian quick clay, 7th IWBIDG, Chania, Greece, June 13–16.
Thakur, V. and Jostad, H.P. (2005) Creep coupled strain softening model for sensitive clays,
In 2nd International Workshop for Young Doctors in Geomechanics, WHYDOC, Paris,
France, November 23–25.
Thakur, V., Nordal, S., Jostad, H.P. and Andresen, L. (2005) Generation dissipation of pore
water during shear banding in sensitive clay, 11th Conference of IACMAG, Turin, Italy,
Volume 4.
Thakur, V. and Nordal, S. (2005) Rate dependent elasto plastic deformation of shear band in
sensitive clays, Electronic Journal of Geotechnical Engineering, edition September, bundle F.
Thomas, T. (1961) Plastic Flow and Fracture in Solids Academic press, New york.
Vardoulakis, I. (1980) Shear band inclination and shear modulus of sand in biaxial tests,
International Journal of Num. Analy. Methods in Geomechanics, 4, 103–119.
Vardoulakis, I. (2002) Dynamic thermo-poro-mechanical analysis of catastrophic landslides,
Ge´otechnique, 52(3), 157–171.
Vardoulakis, I. and Aifantis, E.C. (1991) A gradient flow theory of plasticity for granular
materials, Acta Mech., 87, 197–217.
Vermeer, P.A. and Brinkgreve, R.B.J. (1994) A new effective non-local strain measure for
softening plasticity, In Proc. 3rd International Workshop on Localization and Bifurcation
Theory for Soils and Rocks, Grenoble, France, pp. 89–100.