You are on page 1of 7

Modified shear deformation method for analyzing the

load-displacement response of a single pile in


expansive soil upon infiltration
Yunlong Liu & Sai. K. Vanapalli
Department of Engineering University of Ottawa, Ottawa, Ontario, Canada

ABSTRACT
Safety and stability of various infrastructure constructed in regions with expansive soil deposits are significantly
influenced by heave or settlement problems due to changes associated with the soil natural water content. Piles are
widely accepted in conventional geotechnical engineering practice as most suitable foundations to carry the loads safely
alleviating heave and settlement problems. However, the mechanical behavior of piles are significantly influenced by
infiltration or evaporation of natural water content (and matric suction) in expansive soils. In this study, considering the
ground heave and variations of soil properties with respect to matric suction, the traditional shear deformation method
which is based on saturated soil mechanics principles, is modified for the analysis of the head load-displacement
response of a single pile in an expansive soil deposit upon infiltration. An example problem of single pile in a typical
expansive clay is analyzed to illustrate the modified method. The modified method is capable of simulating the pile head
load-displacement response upon infiltration using limited soil parameters that can be determined from conventional
laboratory tests. The proposed method facilitates a quick and reasonable evaluation of displacement of pile in expansive
soil upon infiltration that can be used in engineering practice applications.

RSUM
La scurit et la stabilit des diverses infrastructures construites dans des rgions o les gisements expansifs des sols
sont fortement influencs par des problmes d'asschement ou de rglement en raison des changements associs la
teneur en eau naturelle du sol. Les piles sont largement acceptes dans les pratiques d'ingnierie gotechnique
conventionnelles, car les fondations les plus appropries pour transporter les charges soulvent en toute scurit les
problmes d'tablissement et de rglement. Cependant, le comportement mcanique des piles est fortement influenc
par l'infiltration ou l'vaporation de la teneur en eau naturelle (et l'aspiration matricielle) dans les sols expansifs. Dans
cette tude, compte tenu de l'lvation du sol et des variations des proprits du sol par rapport l'aspiration matricielle,
la mthode traditionnelle de dformation par cisaillement base sur les principes de la mcanique des sols saturs est
modifie pour l'analyse de la rponse de dplacement de la charge de tte d'une seule pile dans Un dpt de sol
expansif lors de l'infiltration. Un exemple de problme de pile unique dans une argile expansive typique est analys pour
illustrer la mthode modifie. La mthode modifie est capable de simuler la rponse de charge-dplacement de la tte
de pile lors de l'infiltration en utilisant des paramtres de sols limits qui peuvent tre dtermins partir de tests de
laboratoire conventionnels. La mthode propose facilite une valuation rapide et raisonnable du dplacement de la pile
dans le sol expansif lors de l'infiltration qui peut tre utilis dans les applications de la pratique de l'ingnierie.

1 INTRODUCTION rational and reliable design and construction for


sustainable infrastructure.
Expansive soils are considered to be problematic soils In this study, considering the variation of ground
because of their sensitivity to volume changes (i.e. heave and soil properties with respect to matric suction,
swelling and shrinkage) associated with natural water the traditional shear deformation method originally
content changes. Seasonal ground heave or settlement developed by Cooke (1974) is modified for the analysis of
due to fluctuations of natural water content (and matric the head load-displacement response of a single pile in
suction) often cause distress to various infrastructures an expansive soil deposit upon infiltration. While the
constructed in regions with expansive soils deposits. The traditional shear deformation method is based on
economic losses caused by damages induced saturated soil mechanics principles; the modified method
by expansive soil is significant and is comparable to the employs the mechanics of unsaturated soils. The
economic losses by natural hazards (i.e. floods, proposed modified method is simple for use in practice
hurricanes and earthquakes) (Jaremski 2012). applications because it requires limited number of soil
Expansive soils are widespread in various countries of properties which can be measured from conventional soil
the five continents: Canada, United States and Argentina properties.
in America; Sudan and Algeria in Africa, China, India and
Israel in Asia; Spain and United Kingdom in Europe, and 2 BACKGROUND
Australia, to list a few (Chen 1988; Al-Rawas and
Qamaruddin 1998; Rao et al. 2001). These expansive Piles are the most suitable of the presently available
soils are a nightmare to the geotechnical and structural foundations for combating the problems associated with
engineers because of challenges they offer for providing expansive soils (Chen 1965; Al-Rawas and Goosen 2006;
Nelson et al. 2015). Shallow foundations in expansive soil Prior to water infiltration, positive friction is distributed
regions are usually constructed within the active zone. along the entire length of the pile. Applied load is typically
Upon infiltration, the shear strength of the expansive soil carried by shaft friction with some contribution arising
decreases due to a reduction in matric suction. Due to this from the pile end [as shown in Figure 1 (a)]. However, as
reason, there will be also a reduction in the bearing water infiltrates into the active zone, expansive soil swells.
capacity. Furthermore, it is likely that expansive soil Positive friction changes in the active zone due to the
volume increases within the active zone and may lift the mobilization of lateral swelling pressure and the reduction
shallow foundation. This may lead to safety and stability of the interface shear strength properties [as shown in
problems to the superstructure. In order to alleviate these Figure 1 (b)] (Liu and Vanapalli 2017). The positive friction
problems, significantly long piles with high strength and in piles supporting light structures can experience a
stiffness are used. Typically, the depth of pile extends significant increment that contributes to their uplift. For
below active zone (moisture content changes in this zone this reason, the increasing positive friction is generally
are not influenced by seasonal factors) in expansive soil referred to as uplift friction. Once the pile has an upward
with the pile end placed on rigid bedrock or at a lower movement, negative friction generates in the stable zone
stable soil stratum. Such a construction technique and the pile base bearing capacity decreases
facilitates in achieving a high bearing capacity to safely significantly. The net contribution that arises from
carry the loads from the superstructure and effectively negative shaft friction, end bearing capacity and
limit the non-uniform displacement that is likely due to the surcharge combine to balance the increased uplift shaft
changes in the natural water content of the expansive friction.
soils, particularly in the active zone. The design methods Several investigators made contributions to our
are based on saturated soil mechanics principles and are present understanding of the mechanical behavior of pile
similar to other conventional pile foundations placed in foundations in soils upon infiltration of water. These
sands or other clays. However, expansive soils are methods fall into two categories: (i) Theoretical methods
typically in a state of unsaturated condition and may not which include the traditional load transfer method
reach fully saturated condition during their design life proposed by Coyle and Reese (1966) and the shear
period of the pile foundations. Expansive soils as deformation method proposed by Cooke (1974); (ii)
discussed earlier are sensitive to changes in the natural Numerical simulation methods which include the
water content and pile-soil interaction of the physical- boundary element method (Poulos 1993) and finite
chemical-mechanical mechanical behavior is more element method (Nelson et al. 2012). Numerical
complex (Xiao et al. 2011). simulations have been found to be robust options for the
Nelson and Miller (1992) and Fredlund and Rahardjo pile load-displacement analysis with high reliability.
(1993) studies suggest that heave is predominant in the However, numerical simulation is rather complex and time
upper few meters due to water content (and associated consuming process.
matric suction) changes; however, heave the deep seated In this study, the traditional shear deformation method
expansive soil layers is rare. The main reason for heave proposed by Cooke (1974) is modified for the estimation
in the top layers is the desiccation cracks that arise in the of the load-displacement response of a single pile in
expansive soil due to climatic and environmental factors expansive soil upon infiltration, considering the influence
which greatly facilitate the infiltration of water (due to rain, of ground heave and variation of soil shear modulus of
flooding or snow melting). Decayed root system of plants properties. In the present study, the modified shear
in the top layer expansive soils also facilitate easy deformation method is employed for estimation of the
movement of water. The soil layer which is sensitive to load-displacement response of a single pile in expansive
climatic and environmental factors is referred to as zone soil upon infiltration.
of seasonal fluctuations or the active zone. Expansive soil
below the active zone has no cracks; due to this reason, 3 PROPOSED METHOD
water cannot easily percolate. In addition, the overburden
pressure contribution that arises from the active zone 3.1 Basic assumptions
layer significantly reduces limits the heave.
The seasonal volume changes within the active zone (i) There is no slip in the pile-soil interface;
of the expansive soil and soil-structure interaction (ii) Expansive soil around the pile is assumed to be
associated with water content (and suction changes) can homogeneous, isotropic and linear elastic;
significantly influence the mechanical behavior of pile (iii) The cross section is constant along the pile and
foundations. For piles buried in expansive soils, increased the pile is capable of withstanding both compressive and
lateral earth pressure due to the mobilization of lateral tensile stresses that are applied or arise during loading.
swelling pressure contributes to the development of uplift
friction along the pile shaft within the active zone. The 3.2 Theoretical analysis and derivation
combined action of the lateral earth pressure and ground
heave not only results in the redistribution of pile axial The soil around the pile can be assumed to be a series of
forces, but also can contribute to the pile uplift (Liu and concentric cylinders (as shown in Figure 2). The vertical
Vanapalli 2017). settlement of soil around the pile at a certain depth z can
Figure 1 illustrates the variation of pile shaft friction be given by Eq. 1 extending the traditional shear
and end bearing capacity for a single pile (supporting light deformation method proposed by Cooke (1974).
structures) in expansive soil prior to and after infiltration.
0 r0 r 0 r0 rm the radius of pile, r = the horizontal distance between the
w( z, r )
Gs r0 r

Gs
ln( )
r
[1] calculated point and the pile axis, rm = the maximum
influencing radius of pile on the soil, generally rm can be
where w = the vertical settlement of soil around the estimated using Eq. 2 (Xiao et al. 2011).
pile, Gs = the shear modulus of soil around the pile, r0 =

Surcharge
Surcharge

Ground heave

Positive Changes in friction


Active Zone

Active Zone
Friction direction

Positive Increasing lateral


Friction swelling pressure
Changes in
interface Shear
strength properties
Stable Zone

Stable Zone Negative


Friction Changes in
interface shear
strength
End Bearing Capacity End Bearing Capacity

(a) Prior to infiltration (b) After infiltration

Figure 1. Distribution of shaft friction along a pile prior to and after infiltration

The pile displacement can be estimated using Eq. 6,


which is obtained by solving Eq. 5. In Eq. 6, A and B are
rm 2.5L(1 s ) [2] the unknown integration constants in the general solution
which can be determined by applying boundary conditions
and solving for them.
where L is the length of the pile, s = Poisson's ratio of
the soil around the pile.
wpz Ae r1z Be r2 z [6]
The relationship between axial force of pile and the
k ; k
shaft friction around the pile is given as Eq. 3. The r1 r2
relationship between the axial force of pile and the pile Ap E p Ap E p
displacement is given as Eq. 4.
However, Eq. 6 is only suitable for piles in non-
Pz
2r0 0 [3] swelling soils. For pile expansive soils, possible ground
z heave in the active zone should be taken into account as
well. The heave prediction equation in the elastic range
wpz Pz presented by Adem and Vanapalli (2016) (Eq. 7) can be
[4] extended for the calculation of heave in the active zone of
z E p Ap the expansive soil. For soil layers at different depth, the
matric suction variations can be different. Also, the heave
Eq. 5 can be acquired by combining Eq. 1, Eq. 3 and is accumulative from bottom to the top of active zone. As
Eq. 4. a consequence, the heave calculated using Eq. 7 for
2 w pz k [5] different soil layers is different. In this study, within a
wpz 0 certain soil layer, the development of the heave is
z 2 Ap E p
simplified as a linear distribution (using Eq. 8) as shown in
Figure 3.
rm
k 2Gs / ln( )
r0 (1 )(1 2 )
h h [ ](ua uw ) [7]
Ea (1 )
where h = heave of soil, h = thickness of the soils and suggested that 1 = 2 is typically and 2 varies
calculated soil layer, Ea = average elastic modulus over from 0.05 to 0.15 for expansive soils. In this study, an
the matric suction variation range, (ua - uw) = variation in average value, 2 equals to 0.1 is used; Pa = atmospheric
matric suction. pressure.

h Hz a [8] The pile displacement can be estimated using Eq. 10,


which takes account of influence of the ground heave.
Vanapalli and Oh (2010) proposed a semi-empirical The pile axial force can be calculated using Eq. 11, which
model for estimating modulus of elasticity of coarse and is obtaining by manipulating Eq. 10 and Eq. 4.
fine-grained unsaturated soils. Adem and Vanapalli Furthermore, Eq. 12 can be obtained by expressing Eq.
(2014) extended this approach for several expansive 10 and Eq. 11 in a matrix form. Eq. 12 can also be
soils. In this paper, the simple semi-empirical Eq. 9 represented in a simplified form as shown in Eq. 13.
developed by Vanapalli and Oh (2010) and Adem and
Vanapalli (2014) is used to estimate the modulus of wpz ( Hz a) Ae r1z Be r2 z [10]
elasticity of unsaturated expansive soils, Eu. This equation
suggests the E decreases with decreasing matric suction
(i.e. during the infiltration process). In the calculation, Ea wpz
Pz Ap E p Ap E p [ Ar1e r1 z Br2e r2 z H ]
equals the average value of various E calculated using z [11]
Eq. 9 over the matric suction variation range. The
influence of mechanical stress (confinement) is neglected Ap E p [ Ar1e r1 z Br1e r1 z H ]
in this method. Such an assumption is conservative and
can be extended in practice for pavements and lightly

w pz e r1z e r1z A Hz a

loaded residential structures, where the soil matric suction r1z
Pz Ap E p r1e
r1z
changes have a predominant influence on the behavior of Ap E p r1e B Ap E p H
unsaturated expansive soils (Adem and Vanapalli, 2014).
[12]
(ua u w ) 1
Eu Esat [1 2 S ] [9] wpz A Hz a
Pa 100 T ( z ) [13]
where 2 and 1 = fitting parameters, Adem and
Pz B Ap E p H
Vanapalli (2016) calculated Eu for five different expansive

Pile head load

Pz1 wpz1
L1 1
Positive Active
Friction Zone Pz2 wpz2
L2 2

Pz3 wpz3
L3 3

Negative Pz(n-1) Wpz(n-1)


Stable 5
Friction
Zone
Pzn wpzn

Figure 2. Analytical model of pile and soil around the pile


L1 h1 H3
a1

L2 h2 H2
Active zone
a2 1

L3 h3 H3

z a3 1

Depth Calculation of ground heave

Figure 3. Simplification of the ground heave calculation model

The shear modulus of the soil, Gs similar to elastic w pz1 1


w pz2 1
H1 z2 a1
modulus is significantly influenced by variations in matric [T( z1 ) ][T( z2 ) ] [T( z1 ) ][T( z2 ) ]
suction due to water content changes. In this study, Gs is Pz1 Pz2 A p E p H 1
deduced from elastic modulus and expressed as Eq. 14. H1 z1 a1
Poisson's ratio of expansive soils, is usually assumed,
estimated or measured from triaxial tests using the Ap E p H 1
information of linear and lateral strain. [18]

Eu w pz2 wz3 1
H 2 z3 a2
Gu [14]

1
[T( z2 ) ][T( z3 ) ] [T( z2 ) ][T( z3 ) ]
2(1 s )
Pz2 Pz3 A p E p H 2
where Gu = shear modulus of unsaturated soil.
H 2 z 2 a2

Water content changes associated with infiltration A p E p H 2
contribute to a certain matric suction profile along the pile [19]
depth. As discussed earlier, soil properties are influenced
by the matric suction. As a consequence, soil properties
Eq. 20 and Eq. 21 can be deduced by setting
are different with respect to the soil depth. In this study,
both the pile and soil are divided into several segments. [T( z1 ) ][T( z2 ) ]1 [T1 ] and [T( z2 ) ][T( z3 ) ]1 [T2 ] ,.
For a typical pile segment 1 as shown in Figure 2, the pile
axial force and displacement equation at the top and
w pz2 w pz3 H 2 z 3 a2 H 2 z 2 a2
bottom of the segment are given as Eq. 15 and Eq. 16, [T2 ] [T2 ]
respectively. Pz2 Pz3 Ap E p H 2 Ap E p H 2
[20]
w pz1 A H1 z1 a1
[T( z1 ) ] [15]
Pz1 B Ap E p H 1 wpz1 wpz3 H 2 z 3 a2
[T1 ][T2 ] [T1 ][T2 ]
Pz1 Pz 3 Ap E p H 2
w pz2 A H 2 z 2 a1
[T( z2 ) ] [16] H 2 z 2 a2 H1 z2 a1 [21]
Pz 2 B Ap E p H 2 [T1 ] [T1 ]
Ap E p H 2 Ap E p H1
For the same pile segment, the unknown integration H1 z1 a1
constants in the general solution A and B have the same
value (as shown in Eq. 17). Eq. 18 and Eq. 19 can be Ap E p H1
deduced by eliminating A and B for the same pile
segment. Definition of these parameters are available in Similarly, if the pile is divided into "n" segments, a
Figure 2. general Eq. 22 can be given. It should be noted that Eq.
22 totally has (2n - 1) items.
A A As discussed earlier, piles as foundation usually
[17] penetrate into the active zone and the pile end rests on
B z1 z B z1 z L the bedrock or a soil layer that has a relatively high
stiffness. For this reason, the pile base is no longer
influenced by seasonal moisture water content changes. b k sb wb
In other words, the pile below the active zone has a stable [23]
mechanical behavior. However, possible movement of k 4 Gsb
pile base caused by volume expansion or shrinkage of the sb r0 (1 b )
expansive soil in the active zone can lead to changes in where b = unit end resistance, wb = pile-end
end bearing capacity. The pile base settlement can be settlement, ksb = compressive rigidity of soil under pile
estimated using Eq. 23 (Randolph and Wroth in 1978). base, Gsb = shear modulus of soil under pile base, b =
Based on Eq. 22 and Eq. 23, assuming a series of pile Poisson's ratio of soil under pile base, r0 = pile radius.
base settlement, corresponding pile base load can be
calculated using Eq. 23. 4 EXAMPLE PROBLEM

The proposed algorithm in this paper is illustrated with an


wpz1 w pzn example problem in this section. Detailed geometry of the
[T1 ] [Tn 1 ] example problem is shown in Figure 4. It is assumed that
P
z1 Pz n a single pile is constructed in a typical expansive soil.
H n 1 zn an 1 During wet season, water infiltrates into the active zone
[T1 ] [Tn 1 ] along the cracks and contributes to a matric suction
Ap E p H n 1 reduction from 200kPa to zero value. The pile is divided
into two segments. The pile head load-displacement
H n 1 zn 1 an 1
[T1 ] [Tn 2 ] [22] response is analyzed using the modified shear
Ap E p zn 1 deformation method proposed in this study. The elastic
modulus of the pile is set as 500MPa. The saturated
H n 2 zn 1 an 2 elastic modulus and Poisson's ratio for a typical expansive
[T1 ] [Tn 2 ]
Ap E p zn 2
soil clay are set as 5MPa and 0.3, respectively (Ranjan
and Rao 2000).
H n 2 zn 1 an 2
[T1 ] [Tn 3 ]
Ap E p H n 2
H1 z2 a1 H1 z1 a1
[T1 ]
Ap E p H1 Ap E p H1

Ground Pz1 wpz1 Matric


surface suction

Two
4m 200kPa
segments

Active
Pz2 wpz2 zone
Stable
0.5m zone

Pz3 wpz3 z

Figure 4. Simplification of the ground heave calculation model

pile head settlement increases with increasing pile head


The heave in the active zone for the soil properties in load. When the pile head load is zero, the pile head
the example problem are calculated using Eq. 7. The displacement has a negative value. This means that the
shear modulus, Gs in the active zone is calculated using pile moves in the direction of the applied load due to
Eq. 14, taking account of the influence of matric suction contribution of the uplift friction that generates in the
on the soil properties. Figure 5 illustrates the pile head active zone. Since the plasticity of the soil and the shear
load displacement response calculated using Eq. 20. The strength softening are not considered, the modified shear
deformation method is only suitable for the pile head load Adem, H.H. and Vanapalli, S.K. 2014. A simple
displacement analysis under small load (which is less method for prediction of the modulus of elasticity of
than the bearing capacity). In other words, the maximum unsaturated expansive soils, UNAST2014 Conference,
bearing capacity cannot be estimated using proposed Unsaturated Soils: Research and Application, Sydney,
method since the interface shear strength softening is not Australia.
taken into account. Adem, H.H. and Vanapalli, S.K. 2016. Heave
prediction in a natural unsaturated expansive soil deposit
Pile had load (kN) under a lightly loaded structure, Geotechnical and
0 200 400 600 800 1000
Geological Engineering, 34(4): 1181-1192.
-0.02 Chen, F.H. 1988. Foundations on expansive soils,
Elsevier, New York, USA.
Pile head displacement (m)

Chen, F.H. 1965. The use of piers to prevent the uplift


0.00 of lightly loaded structures founded on expansive soils.
Proceeding of International Conference on Expansive
Soils, Texas A&M Press, College Station, Texas, USA.
0.02 Coyle, H.M. and Reese, L.C. 1966. Load transfer for
axially loaded piles in clay, Journal of the Soil Mechanics
and Foundation Engineering, 92(2): 122-126.
0.04
Cooke, R.W. 1974. The settlement of friction pile
foundations, Proceeding of Conference on Tall Buildings,
0.06 Kuala Lumpur, Malaysia.
Fredlund, D.G. and Rahardjo, H. 1993. Soil mechanics
Figure 5. Estimation of the pile head load settlement for unsaturated soils, Wiley, New York.
relationship Jaremski, J. 2012. Application of the joining pile for
foundations on expansive clays and rocks, American
5 CONCLUSIONS International Journal of Contemporary Research, 2(10):
53-58.
Pile foundations have been widely used in engineering Liu, Y. and Vanapalli, S.K. 2017. Influence of Lateral
practice to combat various problems induced by Swelling Pressure on the Geotechnical Infrastructure in
expansive soils upon infiltration. Currently, pile Expansive Soils. Journal of Geotechnical and
foundations in expansive soil are designed extending the Geoenvironmental Engineering, ASCE, 04017006.
principles of saturated soil mechanics. However, as water Nelson, J.D. and Miller, D.J. 1992. Expansive soils:
infiltrates into the active zone, which is typically in a state problems and practice in foundation and pavement
of unsaturated condition, uplift friction generates along the engineering, New York, John Wiley & Sons: 117-144,
pile. Due to this reason, it is likely that there will be an USA.
uplift of the pile. The pile uplift can significantly influence Nelson, J., Thompson, E., Schaut, R., Chao, K.,
the safety and stability of the supported superstructure. Overton, D. and Dunham-Friel, J. 2012. Design
For this reason, it will be appropriate to use unsaturated Procedure and Considerations for Piers in Expansive
soil mechanics for rational analysis of pile foundations in Soils, Journal of Geotechnical and Geoenvironmental
expansive soils. In the present study, the traditional shear Engineering, ASCE, 138(8): 945956.
deformation method originally developed by Cooke (1974) Nelson, J.D., Chao, K.C., Overton, D.D. and Nelson,
is modified for the estimation of the pile head load- E.J. 2015. Foundation Engineering for Expansive Soils,
settlement response in expansive soil using the John Wiley & Sons, Hoboken, New Jersey, USA.
mechanics of unsaturated soils. The application of the Poulos, H. 1993. Piled Rafts in Swelling or
proposed method is illustrated using an example problem, Consolidating Soils. Journal of Geotechnical Engineering,
which illustrates the uplift phenomenon for a single pile. 119(2): 374380.
The example problem shows the suggested method is Ranjan, G. and Rao, A.S.R. 2000. Basic and Applied
simple and can be used in engineering practice Soil Mechanics (2nd Ed), New Age International (P) Ltd,
applications. The proposed method however is only New Delhi, India,
suitable for piles under a head load which is less than its Rao, S. M., Reddy, B. V. V., and Muttharam, M. 2001.
bearing capacity. More future studies are required to The impact of cyclic wetting and drying on the swelling
extend the use of proposed method by considering behavior of stabilized expansive soils, Engineering
plasticity of soil and the interface shear strength softening. Geology, 60:223233.
Vanapalli S.K. and Oh W.T. 2010. A method for
6 REFERENCES predicting the modulus of elasticity of unsaturated soils
using the soil-water characteristic curve, International
Al-Rawas, A.A. and Qamaruddin, M. 1998. Journal of Geotechnical Engineering, 4(4): 425-433.
Construction problems of engineering structures founded Xiao, H., Zhang, C., Wang, Y., and Fan, Z. 2011. Pile-
on expansive soils and rocks in northern Oman. Building Soil Interaction in Expansive Soil Foundation: Analytical
and Environment, 33: 159171. Solution and Numerical Simulation, International Journal
Al-Rawas, A.A. and Goosen, M.F.A. 2006. Expansive of Geomechanics, ASCE, 11(3): 159166.
soils: Recent advances in characterization and treatment,
Taylor & Francis, London, England.

You might also like