You are on page 1of 11

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/267393769

Effect of Intermediate Principal Stress on the Mechanical Behavior of Angular


Sand

Chapter  in  Geotechnical Special Publication · May 2014


DOI: 10.1061/9780784413388.042

CITATIONS READS
2 1,058

2 authors:

Ramesh kannan Kandasami Tejas Gorur Murthy


University of Cambridge Indian Institute of Science
15 PUBLICATIONS   62 CITATIONS    78 PUBLICATIONS   644 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Ecological, engineering and biochemical aspects of termite mound construction. View project

A study of bonded geomaterials View project

All content following this page was uploaded by Tejas Gorur Murthy on 24 March 2015.

The user has requested enhancement of the downloaded file.


Soil Behavior and Geomechanics GSP 236 © ASCE 2014 406

Effect of Intermediate Principal Stress on the Mechanical Behavior of


Angular Sand

Ramesh K Kandasami1 and Tejas G Murthy2


1
Research student, Indian Institute of Science, India
2
Assistant professor, Indian Institute of Science, India

ABSTRACT: Sand is a naturally occurring cohesionless, granular material, with


varying morphology. It has been well studied as a model frictional material; and the
strength of the sand ensemble is derived from the inter-granular friction. Facets of
sand behaviour such as the inherent anisotropy, effect of intermediate principal stress
are extremely important to model and predict the constitutivity. A complex stress
field is required in order to investigate the effect of intermediate principal stress on
the failure behaviour of angular sand and such stress fields can be replicated in the
laboratory using a hollow cylinder torsional testing apparatus (HCT). A slew of
experiments are carried out at a particular density under drained conditions by
varying intermediate principal stress ratio ‘b’. The experimental observations are
studied in the framework of classical critical state soil mechanics and the results are
analyzed using plasticity theory. Effect of intermediate principal stress ratio on the
non-coaxial and failure behaviour is highlighted.

INTRODUCTION

Prediction of the mechanical behaviour of sand is vital for applications in


construction, infrastructure and manufacturing industries. The mechanical behaviour
of sand is very complex due to an array of interesting properties that sand exhibits,
such as mean stress dependence, dilatancy, shear banding and localization etc.
Applying the basic tenets of continuum mechanics and performing elemental tests to
understand the mechanical response of sand under generalized loading conditions has
been a cornerstone in development of predictive tools such as constitutive models.
The triaxial stress condition has remained ubiquitous in soil mechanics research and
practice, and has been extensively used as a standard test of understanding soil
behaviour. In the conventional triaxial test condition, the intermediate principal stress
made equal to either the major or minor principal stress, thus allowing a cylindrical
sand sample surrounded by a fluid to simulate this stress condition. An important
aspect which is generally not considered in experiments is the effect of intermediate
Soil Behavior and Geomechanics GSP 236 © ASCE 2014 407

principal stress on the behaviour of sand. This nescience is due to the limited
availability of experimental apparatus which are capable of controlling the
intermediate principal stress. Unlike traditional triaxial and simple shear apparatus
where independent control of all the three principal stresses is impossible, true triaxial
apparatus and hollow cylinder apparatus allow an independent control on the
magnitude of the three principal stresses. In addition to this, the hollow cylinder
torsional apparatus has a distinct advantage of controlling direction of the principal
stresses (α).
It is well documented that the intermediate principal stress plays a significant role
in controlling the mechanical response of sands. According to Bishop (1966), the
influence of intermediate principal stress can be described by a dimensionless
parameter ‘b’- called intermediate principal stress ratio ‘b’, where b = (σ2-σ3)/(σ1-σ3)
& varies from 0 to 1. Also ‘b’ is defined as a specific plane in the principal stress
space. b = 0 represents a compression plane while b = 1 represents an extension plane
(Atkinson and Bransby, 1982).
In 1957, Kirkpatrick was the first to investigate the failure conditions of sand using
a hollow cylinder torsional apparatus. Lade and Duncan, (1973) performed a series of
true triaxial tests in order to study the effect of intermediate principal stress on sands.
The failure points were plotted on the octahedral plane and concluded that the shape
of the yield surface is a triangle with curved edges. Symes et al. (1988) used hollow
cylinder apparatus and studied the effect of intermediate principal stress ratio and
principal stress rotation separately in a three dimensional space under drained
conditions. Only limited regions of the stress space were explored by change in ‘b’.
Dakuolas and Sun (1992) performed hollow cylinder experiments on a clean Ottawa
sand, under a slew of loading conditions like compression, extension, simple shear
and combined loading at different intermediate principal stress ratio. The failure
points obtained from the results were plotted on the octahedral plane these
experimental results were used to benchmark Lade’s single hardening constitutive
model. Also, Yoshimine et al. (1998), used Toyoyura sand and performed a series of
tests by varying ‘b’ and ‘α’ under undrained conditions. From their tests it was
observed that as ‘b’ increased the sand showed increased contractive behaviour.
Sayao and Vaid (1996), from a series of stress path tests conducted on Ottawa sand
showed that the non-dimensional parameter ‘b’ has an effect on the stress strain
behaviour of sand. In a set of tests performed by these authors at different values of
‘b’, while keeping a constant α = 45°, the stress ratio ‘R’ (R = σ1/σ3) was found to be
maximum at b = 0.3 indicating that the specimen was stiffest at b=0.3. The friction
angle at failure was also seen to increase as ‘b’ increased upto 0.3 and then decreases,
at variance to results presented by Bishop (1966), Lade and Duncan (1973) etc.
This paper reports on a series of hollow cylinder tests conducted in order to
understand mechanical behaviour of angular sand under drained conditions. Tests
were performed at different intermediate principal stress ratios in order to understand
the effect of intermediate principal stress on the strength at critical state, non-
coaxiality between the stress and strain directions and the shape of the failure surface
on a deviatoric/octahedral plane.
Soil Behavior and Geomechanics GSP 236 © ASCE 2014 408

EXPERIMENTAL

A series of drained tests on a HCT apparatus was conducted in this research


programme in order to investigate the effect of intermediate principal stress ratio. A
fully automated hollow cylinder torsion testing system built by GDS instruments
(GDS Instruments Co., UK) was used here. This apparatus provides an independent
control of the axial load (W), torque (Mt), internal pressure (Pi) and external pressure
(Po) on a hollow cylinder, which in-turn is used to control the magnitude and
direction of the three principal stresses. The average axial, radial, tangential and shear
stresses on an element is obtained by solving the balance equations (Hight et al,
1983). The average axial displacements, and rotation angles were measured using
high precision digital encoders, while the change in the volume of the specimen is
recorded using the digital pressure volume controllers (DPVC). In a hollow cylinder
torsional apparatus, the intermediate principal stress is always equal to the radial
stress (σ2 = σr). For a special case when Po = Pi, the intermediate principal stress ratio
b = sin2α (the intermediate principal stress ratio is related to the principal stress
inclination) can be used.

Materials and specimen preparation

Cauvery sand, which is a quartzitic sand is used in this study. It has a specific
gravity of about 2.65. The sand particles are angular with a mean grain size of about
0.45 mm. The angular shape of the particles is very apparent in the figure 1 captured
using a scanning electron microscope. The sand is poorly graded with uniformity co-
efficient of 3.71 and co-efficient of curvature of about 0.79. Figure 2 shows the grain
size distribution of this sand, the maximum and minimum void ratio of this angular
sand is 0.97 and 0.53 (ASTM D4254 & D4253) respectively. The chemical
composition of the sand is analyzed using both X-Ray diffraction and ICPOES
(Induction coupled plasma optical emission spectroscopy) studies. This quartzitic
sand consists of SiO2 – 92.1 %, Al2O3 – 1.6 %, Fe2O3 – 1.4 %, Na2O – 0.5 %, CaO –
0.3 %, K2O – 0.1 %, MgO – 0.1 % (with some loss due to ignition).
Specimens were prepared such that the density and fabric is maintained uniform
throughout the specimen. Pluviation technique (Vaid and Negussey, 1984, Cresswell
et al., 1999) was used to attain a particular packing/density. A zero drop height was
used in raining the sand into a split mold filled with water. The specimens were
prepared at a relative density of about 38% (e = 0.80), height about 200 mm, external
diameter of 100 mm and internal diameter of 60 mm. The dimensions of the specimen
were fixed based on the criteria given by Sayao and Vaid (1991) so that the non-
uniformities were kept as small as possible. Also Hight et al, 1983 quantified the level
of stress non-uniformities across the specimen and provided a range of stress ratio
(ratio of external pressure Po and internal pressure Pi) within which the stress non-
uniformities were minimum (0.9 < Po/Pi < 1.2), hence, the external and internal
pressures were maintained within this range.
Soil Behavior and Geomechanics GSP 236 © ASCE 2014 409

Fig. 1. Scanning Electron Micrograph which shows the angular shape of the
clean Cauvery sand.

Fig. 2. Grain size distribution of the poorly graded angular sand, with a mean
grain size of about 0.45mm.

Tests

A negative pressure of 50 kPa was applied to the specimen during the setting up
stage followed by the application of the back pressure of about 300 kPa. With an
effective pressure of 50 kPa, the specimen was saturated until the B-value reached a
value of 0.97 or more. Further the specimen was isotropically consolidated to an
Soil Behavior and Geomechanics GSP 236 © ASCE 2014 410

effective stress of about 300 kPa in several stages before shearing the specimen. A
series of stress-controlled tests at different values of ‘b’ was performed in this testing
programme keeping the major principal stress direction to be vertical. The non-
dimensional parameter ‘b’ depicts a particular plane in the principal stress space.
Figure 3 shows the extreme conditions of b = 0 - the triaxial compression plane (σ1 >
σ2 = σ3) and b = 1 - the triaxial extension plane (σ1 = σ2 > σ3). Varying ‘b’ is useful in
understanding the sand response with changing intermediate principal stresses.

Fig. 3. Stress space where different planes are explored by varying the
intermediate principal stress ratio (b = 0 for σ2 = σ3 plane, b = 1 for σ1 = σ2 plane)

RESULTS

From the experiments performed, the behaviour of sand under various loading
conditions is analyzed. Some of the typical results of stresses and strains obtained
from a HCT test are initially discussed followed by the study on the effect of
intermediate principal stress ratio on the behaviour of sand. From the tests conducted
by varying intermediate principal stress ratio, ancillary studies are carried out to
understand the non-coaxiality and failure (stress space) behaviour of sand.

Typical stress strain plots

The four components of stress tensor (axial, radial, tangential and shear stresses)
evolve through the test. The variation of these stress components for the case of α =
0° and b=0 are such that two of the three normal stresses (radial and tangential stress)
are decreasing while the axial stress is increasing with axial strain. The shear stress is
zero and this condition α = 0° and b = 0 is similar to triaxial compression test. As the
condition b = 0 is imposed on the test, the stress components are controlled in such a
way that two of the three principal stress are equal and decreasing while the third
principal stress is increasing. All the tests in this study were performed at a constant
mean effective normal stress and hence figure 4 shows σ2 = σ3 decreasing while σ1
was increasing. The variation of strain components are such that the radial and
tangential strains are decreasing while the axial strain is increasing. It is also found
that as ‘b’ increases from 0 to 1, the radial stress which is decreasing on the
Soil Behavior and Geomechanics GSP 236 © ASCE 2014 411

compression plane (b = 0), increases and becomes equal to major principal stress on
the extension plane (b = 1). Thus on the extension plane, the major and intermediate
principal stresses are equal and increasing while the minor principal stress is
decreasing. Failure or critical state here in this study under all conditions is defined as
the peak value of the deviatoric stress on a p’ vs. q plot.

Fig. 4. the radial stress becomes the intermediate principal stress and the
tangential stress becomes the minor principal stress where both the stresses were
decreasing while the axial stress becomes the major principal stress and is
increasing with the axial strain (at b = 0 and α = 0°).

Effect of ‘b’ on deviatoric/shear stress

The tests performed by varying ‘b’ from 0 to 1 indicate the transition from triaxial
compression to extension. Figure 5 shows the effect of intermediate principal stress
ratio on the deviatoric stress. The value of deviatoric stress when the sample reaches a
critical state is almost equal when ‘b’ values ranges from 0 to 0.6. When the value
goes beyond b = 0.6, the critical deviatoric stress decreases showing a minimum
value at b = 1. This behaviour is attributed to the increasing contractivity of sand as
‘b’ increases from 0 to 1. Angular particles have a propensity for orienting themselves
in a plane normal to the direction of deposition under gravity (Oda and Iwashita,
1999); hence an inherent anisotropy due to this fabric (i.e. orientation of the angular
particles) can be ascribed to these specimens. The friction angle at critical state is also
almost constant until ‘b’ reaches 0.6 beyond which it decreases (since p’ is constant,
‘q’ defines the critical state friction angle). Lade (1973), performed experiments on a
Soil Behavior and Geomechanics GSP 236 © ASCE 2014 412

sub-rounded sand and observed that the friction angle increases with increase in ’b’
upto 0.5 after which it decreases. The variation of friction angle with ‘b’ for rounded
and angular particles are slightly different because of the dissimilarity in the
depositional fabric. It is also identified that the octahedral shear strain at critical state
or failure increases as ‘b’ increases and reaches maximum values at b = 0.4 to 0.6 and
further decreases. This behaviour signifies that at around b = 0.5, the maximum
straining occurs as the test is akin to simple shear.

Fig. 5. The effect of intermediate principal stress ratio on the deviatoric stress
was shown and the deviatoric stress at critical state/failure is found to be 22%
lesser on the extension plane compared to the compression plane.

Non-coaxiality

A material behaviour is said to be non-coaxial when the principal stress direction


(α) does not coincide with the direction of the principal strain increment (αdε) during
the plastic deformation. The earliest experimental studies on the sand to understand
the non-coaxiality was reported by Roscoe et al in 1967 by testing sand on a simple
shear apparatus. Cai et al. (2013) performed a series of experiments on Portaway sand
by varying the principal stress inclination (direction of major principal stress with
respect to vertical) to study its non-coaxial behaviour. However studies on the non-
coaxial behaviour of angular sand due to the variation of ‘b’ are few and far between.
From the series of tests at various ‘b’ values (between b = 0 to b = 1 at 0.2
intervals) keeping principal stress inclination ‘α’ at 0°, a study on the evolving
magnitudes of inclinations of principal stresses and principal strains increment are
calculated. With increase in the value of ‘b’, the deviation between the stress and
Soil Behavior and Geomechanics GSP 236 © ASCE 2014 413

strain increment directions at failure/critical state is also increased. The principal


strain increment at failure is found to be maximum on the plane b = 1 (αdε = 1.24°)
and minimum on the plane b = 0 (αdε = 0.001°) as shown in figure 6. This deviation
from the coaxial behaviour is ascribed to the anisotropic nature of sand. This initial
scatter in the principal strain increment direction is conjectured to be due to the
inaccuracies in measurement of strains at low deviatoric stress levels (Cai et al.,
2013).

Fig. 6. The non-coaxial behaviour of sand at different intermediate principal


stress ratios with principal stress direction at 0°.

Failure behaviour

In the current study, the term failure/yield is adopted when the material reaches the
peak value of deviatoric stress in p’ vs. q space. Figure 7 is a two dimensional
representation of the three dimensional stress space where the yield points are plotted.
This two dimensional representation of the yield points is fitted by rotating the
intermediate principal stress (σ2) so as to coincide with the hydrostatic axis (σ1 = σ2 =
σ3), while the other two principal stresses lie on the octahedral plane (Rao and Nott,
2005). Two variables a1/S and a3/S can be defined as follows (equation 1 to 3),

a1 ((2* 1 ) 2  3 )/ 6 (1)

a3 ( 2  3 )/ 2 (2)

S ( 1  2  3 ) (3)
Soil Behavior and Geomechanics GSP 236 © ASCE 2014 414

Figure 7 shows the experimentally obtained failure points plotted along with the
benchmarked Mohr coulomb and Lade’s failure models with friction angle φ = 34°
and dimensionless constant η = 54 respectively. Using six fold symmetry the entire
yield loci can be constructed using these experimental results (Davis and Selvadurai,
2002). The experimentally obtained yield loci is a curved triangle which is akin to
Lade’s failure surface (Lade and Duncan, 1975).

Fig. 7. The failure loci obtained by varying ‘b’ and plotted along with the
benchmarked Mohr Coulomb and Lade failure models.

CONCLUSIONS

This paper reports on a series of drained stress controlled HCT tests on an


angular sand performed to investigate the influence of intermediate principal stress.
Particle shape (angularity) plays a very important role in depicting the anisotropy
nature on the failure behaviour of sand. The shear stress at critical state decreases
with increasing ‘b’ signaling increased contractility. The intermediate principal stress
also has an influence on the non-coaxial behaviour of sand, i.e. with increase in ‘b’
the specimens showed a marginal non-coaxiality (the deviation between ‘α’ and ‘αdε’
is just about 1°). The failure points that are plotted on the stress space show a shape
(curved triangle) akin to a Lade failure criterion. These suites of experimental results
at the continuum scale clearly delineate the effects of the intermediate principal
stress. These tests can be used to benchmark constitutive models, in addition to
validating DEM simulations.
Soil Behavior and Geomechanics GSP 236 © ASCE 2014 415

REFERENCES
Atkinson, J.H. and Bransby, P.L. (1982), “The mechanics of soils – An introduction
to critical state soil mechanics.” McGraw-Hill.
Bishop, A.W. (1966), “The strength of soils as engineering materials.” Geotechnique,
16 (2): 91-130.
Cai, Y., Yu, H., Wanatowski, D. and Li, X. (2013), “Noncoaxial behavior of sand
under various stress paths.” J. Geotechnical and Geoenvironmental Engineering
ASCE, 139(8): 1381–1395.
Cresswell, A., Barton, M.E. and Brown, R. (1999), “Determining the maximum
density of sands by pluviation.” Geotechnical Testing Journal, 22 (4): 324-328.
Dakoulas, P. and Sun, Y. (1992), “Fine Ottawa sand – Experimental behaviour and
theoretical predictions.” J. Geotechnical Engineering, ASCE, 118 (12): 1906-1923.
Davis, R.O. and Selvadurai, A.P.S. (2002), “Plasticity and geomechanics.”
Cambridge University Press.
Hight, D.W., Gens, A. and Symes, M.J. (1983), “The development of a new hollow
cylinder apparatus for investigating the effects of principal stress rotation in soils.”
Geotechnique, 33 (4): 355-383.
Kirkpatrick, W.M. (1957), “The condition of failure of sands.” Proceedings, Fourth
International Conference on Soil Mechanics and Foundation Engineering,
London, (1): 172-178.
Lade, P.V. and Duncan, J.M. (1973), “Cubical triaxial tests on cohesionless soil.”
Journal of Soil Mechanics and Foundation Division, ASCE, (99): SM10, 793-811.
Lade, P.V. and Duncan, J.M. (1975), “Elastoplastic stress-strain theory for
cohesionless soil.” J. Geotechnical Engineering Division ASCE, 101 (10): 1037-
1053.
Oda, M. and Iwashita, K. (1999), “Mechanics of granular materials – An
introduction.” A. A. Balkema Publishers.
Rao, K.K. and Nott, P.R. (2008), “An introduction to granular flow.” Cambridge
university press.
Roscoe, K. H., Bassett, R. H. and Cole, E. R. (1967), “Principal axes observed during
simple shear of a sand.” Proceedings of Geotechnical Conf., Oslo, (1): 231-237.
Saada, A.S., Liang, L., Figueroa, J.L. and Cope, C.T. (1999), “Bifurcation and shear
band propagation in sands.” Geotechnique, 49 (3): 367-385.
Sayao, A. and Vaid, Y.P. (1991), “A critical assessment of stress non uniformities in
hollow cylinder test specimens.” Soils and Foundations, 31 (1): 60-72.
Sayao, A. and Vaid, Y.P. (1996), “Effect of intermediate principal stress on the
deformation response of sand.” Canadian Geotechnical Journal, (33): 822-828.
Symes, M.J., Gens, A. and Hight, D.W. (1988), “Drained principal stress rotation in
saturated sand.” Geotechnique, 38 (1): 59-81.
Vaid, Y.P. and Negussey, D. (1984), “Relative density of pluviated sand samples.”
Soils and Foundations, 24 (2): 101-105.
Yoshimine, M., Ishihara, K. and Vargas, W. (1998), “Effects of principal stress
direction and intermediate principal stress on undrained shear behaviour of sand.”
Soils and Foundations, 38 (3): 179-188.
Yu, H.S. (2008), “Non co-axial theories of plasticity for granular materials.”
IACMAG, 363-378.

View publication stats

You might also like