Proceedings of Indian Geotechnical Conference

December 22-24,2013, Roorkee

LATERAL LOAD CAPACITY OF PILES IN STRATIFIED SOIL DEPOSIT

SUBJECTED TO SCOUR

A. K. Yadav, Resident Engineer, Jayprakash Associates Pvt. Ltd., Noida, India, akyiitg30@gmail.com
A. Dey, Assistant Professor, IIT Guwahati, Assam-781039, India, arindam.dey@iitg.ac.in

ABSTRACT: Pile foundations of bridges are often subjected to detrimental changes in lateral load capacity in
event of extreme scouring. Based on the permissible deflection of the pile at its cut-off level, this article reports the
estimation of lateral load capacity of a single free-headed pile embedded in stratified deposit. Two softwares with
different working principle, BEF and OASYS ALP, have been utilized. Extreme scouring depths (in the tune of 5-7
m has been accounted) portray vivid degradation of the lateral load capacity of pile, guided by the thickness and
stiffness of the substratum.

INTRODUCTION
For piles supporting bridge structures, it is very difference theory to solve for the flexural response
important to estimate the ultimate lateral load of the beam in terms of deflection, bending
capacity of the pile in order to have a-priori idea moment, shear stress and contact stress profiles.
about its failure. Conventionally, such piles are The responses are provided as graphical response
estimated for their ultimate load based on specific envelopes as well as in tabulated format as nodal
deflection criterion. This article reports the information. Fig. 1 depicts an arbitrarily loaded
findings of a study to estimate the lateral load prismatic beam of unit width and length L resting
capacity of a single pile embedded in stratified soil on an elastic foundation with variable foundation
deposit in river bed, based on the permissible reaction.
deflection of 10% of the pile diameter at the cut-off
(in absence of scour) or at the maximum scour
level (in the presence of scour). Flexural response
of the pile (in terms of deflection, bending moment
and shear force and contact stress profiles) has
been illustrated using two different softwares
namely Beams on Elastic Foundations (BEF)
[linear elastic analysis] and OASYS ALP v19.1
[nonlinear P-Y analysis]. Analysis of Laterally
Loaded Piles (ALP) has been found to be more
efficient since it has the ability to consider the Fig. 1 Arbitrarily loaded prismatic beam resting on
effect of water table and nonlinear earth pressure Winkler foundation
generation in the soil surrounding the pile. Based
on the results, the article reports the lateral load The governing deflection response of the above
capacity of the piles embedded in stratified deposit beam is expressed as:
with or without the effect of scour. The effect of
4 y
scour on the lateral load capacity of the pile has EI w ky (1)
x 4
also been reported.
where, E is the modulus of elasticity and I is the
moment of inertia of the beam, y is the deflection
WORKING PRINCIPLE OF BEF
at any point of the beam, w is the magnitude of
BEF [1] utilizes the concept of beam resting on
applied load, and k is the modulus of subgrade
elastic Winkler foundation [2,3] and uses finite
reaction of the elastic foundation.

Page 1 of 9
A. K. Yadav & A. Dey

In order to solve the above problem numerically,
the beam is discretized into several nodes, and the
central difference scheme of the finite difference
theory is utilized to express the governing
deflection response at any node i as:
ki h4
i 2
y 4 yi 1 (6 ) yi 4 yi2 yi2 i
EI
where, Pi is the computed equivalent nodal load,
and h is the uniform spacing of two nodes.
analyzed for three different diameters (1000 mm,
1200 mm and 1500 mm). The pile is considered to
be made of M25 grade concrete, and is assumed to
be free-headed. Seven different soil-pile
configurations have been analyzed, although not
Ph 3
presented here for the sake of brevity [4].
(2)
EI

The above expression is further modified for

appropriate boundary conditions and changes in the
flexural rigidity of the beam and subgrade modulus
of the soil.

WORKING PRINCIPLE OF ALP

ALP models the pile as a series of elastic beam
elements and the soil as a series of non-interactive,
non-linear Winkler springs. The outcome conforms
to the prediction of contact pressures, horizontal
displacements, shear forces and bending moments
induced in a pile when subjected to lateral loads.
The load-deflection behaviour of soil is modelled Fig. 2 General soil-pile interaction model in ALP
either assuming an Elastic-Plastic behaviour, or by
specifying or generating load-deflection (P-Y) data
(depending on whether the soil is clayey or sandy).
The pile is discretized into several nodes, and two
separate stiffness matrices (one for the pile in
bending and one for the adjacent soil) relating
nodal forces to displacements are developed. The
pile stiffness is provide at each node, and remains
constant between successive nodes. The software
allows for incremental load application which aids
to monitor the progressive change in flexural
response. ALP is capable of considering the
presence of water table. Fig. 2 describes the
general schematic of a soil-pile interaction model
in ALP.

PROBLEM STATEMENT
Fig. 3 depicts a single pile embedded in a stratified
soil. BEF and ALP have been used to determine
the lateral load capacity for a specified deflection
at its cut-off level (maximum allowable deflection
criterion considered as 10% of the pile diameter),
the length of fixity of the pile and the
corresponding flexural response envelopes at the Fig. 3 Soil-pile configuration of the present study
lateral load capacity of pile. The pile had been

Page 2 of 9

BEF LINEAR-ELASTIC ANALYSIS
A linear elastic analysis has been carried out using
BEF, considering the laterally loaded pile as a
beam resting on elastic foundation (representing
the soil stratification with varying subgrade
modulus). A vertical load is applied at the tip of the
free-headed pile. The bottom of the pile is
considered to be restrained both from translation
and rotation. BEF does not account for water table
in the analysis.
Estimation of Model Parameters
For an analysis in BEF, the contributory
parameters are (a) Pile - the modulus of elasticity
of pile material, geometry configuration (length
and cross-section dimensions) and moment of
inertia, and (b) Soil - modulus of elasticity and
subgrade modulus of soil.
Depending on the grade of concrete, the elastic
modulus of concrete (Es) is estimated (using Indian
Standards [5]) as:
Ec 5000 fck 25 kN/mm2 (3)
2
where, fck is the grade of concrete (in kN/mm )
Based on standard literature [6], the modulus of the
clayey and sandy stratum is estimated as:
Es clay 350c 227.5 kg/cm2 (4) response of the pile. Since BEF operates on linear
Es sand 400 10.5 N 998.5 kg/cm2 (5) obtained as linear. It is observable that the lateral
where, c is the cohesion of clayey soil, and N is the
SPT blow count obtained in a sandy soil.
Considering a 1000 mm diameter of pile, the
moment of inertia is estimated to be 4.9x106 cm4.
Using Vesics expression [7], the modulus of
subgrade reaction (ks) (Pile diameter = 1000 mm)
is estimated to be 11.76 and 54.88 kN/m2/mm
respectively for silty-clayey and sandy strata as
shown in Fig. 3.
0.65Es E D4
ks 12 s
(6)
D 1 s2 Ec I p
where, D is the pile diameter, Es is modulus of
elasticity of soil, s is the Poissons ratio of soil
(0.35 for clayey soil and 0.25 for sandy soil), and Ip
is moment of inertia of the pile.
For piles of other diameters, the subgrade modulus
has been similarly calculated.
Lateral load capacity of piles in stratified soil deposit subjected to scour
Mesh Discretization
Finite difference method is subjected to various
types of numerical instabilities originating due to
either sparse meshing (missing response variations)
or even very dense meshing (progressive error
accumulation). Both the instabilities result in a
non-convergent solution. Hence, it is necessary to
carry out a sensitivity study and arrive at an
optimum number of nodes which results in a stable
and convergent solution which does not vary with
small changes in the number of nodes [8]. In BEF,
the sensitivity study was carried with 151, 677 and
1201 nodes and the changes in the flexural
responses were investigated. 151 nodes were
considered to be optimum and have been used in
further studies.
Load-Deformation Response and Lateral Load
Capacity of Pile
The lateral load capacity of the free-headed pile is
estimated based on the maximum permissible
deformation criterion of 10% of the pile diameter
at its cut-off level. For the present study, the load at
the pile head is continually increased till the
permissible deflection at the pile head or cut-off
level is reached. Fig. 4a depicts the load-deflection
elastic springs, the response curves have also been
load capacity of pile increases with the diameter of
the pile due to the enhanced flexural stiffness of
the pile.
The lateral load capacity of the pile, as estimated
above, is further used to determine the flexural
response and estimate the maximum bending
moment, shear force and contact stress generated at
the verge of failure. Fig. 4b depicts such a typical
flexural response as obtained from the software. In
actual practice, the pile needs to be designed based
on these maximum magnitudes. Moreover, the
length of fixity of the laterally-loaded pile has also
been determined from the maximum moment/zero
shear force criterion. The point of fixity is referred
to as that point above which the pile can be
modelled and analyzed a fixed cantilever beam.
Beneath this point, the pile is considered to be
fixed and does not contribute to rotation of the pile.
Table 1 enlists the response of the pile (for various
Page 3 of 9
A. K. Yadav & A. Dey

diameters) at its lateral load capacity. The table Effect of Scouring

reflects that increase in pile diameter results in
increment of the lateral load capacity as well as the
length of fixity of the same.
Fig. 4a Load-deformation and lateral load capacity
of pile using BEF
Scouring is the removal of sediment such as sand
and rocks from around bridge abutments or piers.
Scour, caused by swiftly moving water, can scoop
out scour holes, compromising the integrity of a
structure. For Indian rivers and river bed
conditions, guidelines are available related to the
estimation of scour depth in accordance to the
Laceys theorem [9, 10, and 11]. For Indian rivers
like Ganga and Brahmaputra, scour depth as high
as 18m have been reported [12]. Fig. 5 depicts the
flow structure and scouring around a bridge pier.
An attempt has been made in this study to
comprehend the effect of scouring on the lateral
load capacity of the pile. A scouring depth of 11 m
has been considered in the present study. The
methodology, as stated earlier for BEF, has been
used to estimate the lateral load capacity with
increasing scour depth by taking into account the
increase in the cut-off level due to scouring. The
lateral load capacity is estimated based on the
permissible deflection of 10% of pile diameter at
the pile-head. Fig. 6 reveals the degradation in the
lateral load capacity of pile due to increase in the
scour depth. It is clearly understandable that
scouring of softer soil layers result in faster
degradation of the lateral load capacity of the pile.

Fig. 4b Typical flexural response of pile as

obtained from BEF

Table 1 Response of the pile at lateral load

capacity as obtained from BEF
D ycut- Pu Mmax Qmax pmax Lfix
(mm off (kN) (kNm) (kN) (kPa (m)
) (mm )
) Fig. 5 Flow structure and scouring around a bridge
1000 10 222 817 147 115 5.3 pier
1200 12 366 1541 233 147 6.5
1500 15 675 3342 413 197 7.2 OASYS ALP NONLINEAR P-Y ANALYSIS
Note: ycutoff - Deflection at cut-off level, Pu - Unlike BEF, ALP is capable of considering the
Lateral load capacity, Mmax - Maximum moment, effect of water table. The software also facilitates
Qmax - Maximum shear force, pmax - Maximum to determine the flexural response through
contact stress, Lfix - Length of fixity from the automatic load increment, and hence compared to
bottom of the pile.

Page 4 of 9

Fig. 6 Lateral load capacity degradation with
increased scouring
BEF, utilizes lesser time to estimate the failure
load. ALP also considers the passive resistance of
the soil (not available in BEF) and indicates
whenever the passive resistance of the soil is
exceeded. The methodology taking into account of
the generated P-Y curves has been considered in
the present study. Fig. 7 depicts a typical sample
output as obtained from ALP.
Fig. 7 Typical flexural response envelopes as
obtained from ALP 19.1
Estimation of Model Parameters
Both elastic and plastic soil properties are required
to be provided in the ALP input model. Apart from
the modulus of subgrade reaction (identical as used
in the BEF software), depending on the soil type,
Lateral load capacity of piles in stratified soil deposit subjected to scour
the other parameters required as input are: (a) Unit
weight, percentage strain, and cohesion for clayey
soils and (b) unit weight, angle of internal friction
and coefficient of earth pressure at-rest for sandy
soil. The unit weight of soil for all the problems is
considered to be 18 kN/m3, and the strain is
considered to be of magnitude 5% for all soils. The
coefficient of earth pressure at rest is computed
using Jakys expression [13] as follows:
K0 1 sin (7)
The pile has been discretized into nodes with a
uniform spacing of 0.5 m, with required
refinements being made in the intersection with the
water table and soil strata interface.
Load-Deformation Response and Lateral Load
Capacity of Pile
The advanced feature of automatic load increment
available in the software OASYS ALP 19.1 has
been used to estimate the lateral load capacity of
the pile. For the present problem, the load acting at
the pile head has been set to a reasonable value
(e.g. 250 kN) and the number of increments to
reach the load has been provided (13 increments).
This enables to obtain the cumulative flexural
response of the pile with the load increments. Fig.
8a-8d shows a typical load increment procedure in
terms of the deflection response of the pile (1000
mm diameter).
Fig. 8a Deflection envelopes for load increments
(Pile diameter = 1000mm)
Page 5 of 9
A. K. Yadav & A. Dey
Fig. 8b Bending moment envelopes for load
increments (Pile diameter = 1000mm)
Fig. 8c Shear force envelopes for load increments
(Pile diameter = 1000mm)
Fig. 8d Contact pressure envelopes for load
increments (Pile diameter = 1000mm)
Fig. 9 Load-deformation curves and lateral load
capacity of piles using OASYS ALP 19.1
From the above observation, considering the
deflection criterion as 10% of the pile diameter at
cut-off level, the lateral load capacity of the pile
has been estimated. Fig 9 depicts the lateral load
capacity as estimated from the ALP software. It
can be observed from the plots that, unlike BEF,
the load-deflection curves reveal nonlinear
behavior for higher diameter piles.
Fig. 10 reveals the difference in the estimated
lateral load estimated obtained from BEF and ALP
considering the permissible deflection criterion as
10% of the pile diameter at the cut-off level of the
pile. It is noticeable that the lateral load capacity
estimated from ALP is lower than that obtained
using BEF.
Fig. 10 Comparison of lateral load capacity of piles
of varying diameters obtained from BEF and ALP
Page 6 of 9
 Lateral load capacity of piles in stratified soil deposit subjected to scour

This may be attributed to the nonlinear analysis of condition. The sandy layers show mostly bilinear
ALP which inadvertently considers the soil to be behavior and, as indicated in the figure, needs to be
remaining in a softer state at any stress level as subjected to large load values to reach their
compared to the linear elastic analysis, and hence ultimate state. Hence, the linear/nonlinear flexural
shows lower lateral load capacity as the soil offers behavior of the pile will be governed by the
more displacement as compared to the other thickness and location of the clayey or sandy
procedure. Moreover, the consideration of the stratum.
water table results in soil layers remaining in the
submerged condition, which further lowers the Table 2 Response of the pile at lateral load
lateral support of the soil due to the reduction of capacity as obtained from ALP
the effective unit weight of the soil. D ycut-off Pu Mmax Lfix
(mm) (mm) (kN) (kNm) (m)
Using the estimated lateral load capacity, the 1000 10 101 488 7
flexural response of the pile has been determined 1200 12 181 984 7
for various diameters and is represented in Figures 1500 15 388 2370 7.5
11a-11e. Similar to BEF, it is observed that with
the increase in the pile diameter, the maximum
responses (deflection, rotation, bending moment,
shear force, and contact stresses) also reveal
increment in their magnitudes. However, it is
worth noticing that the point of fixity, represented
by the position of maximum bending moment, does
not reveal significant change with the change in the
diameter of the pile. This is dissimilar in
observation to BEF. This may be attributed to the
fact that BEF considers solely the modulus of
subgrade reaction as the support system while
neglecting the passive resistance offered by the
soil. Moreover, the response of soil is generally
nonlinear under high load and especially when the
structures reach the verge of failure. The nonlinear Fig. 11a Deflection envelope of pile as obtained
P-Y curves provide much realistic behavior of the from ALP 19.1
soil under such condition. The deformation
behavior of the soil does not proportionally
increase with the increment in the load, and hence
shows relatively softer behavior in comparison to
the linear elastic medium. Moreover, the load-
deformation behavior of the alternatively placed
sandy and clayey stratum is significantly different.
All this factors might actually result in maintaining
the point of fixity of the pile to be same despite
change in pile diameter. The results are enumerated
in Table 2.

Figs. 12a-12b depicts the P-Y curves that are

generated at the top and bottom of each soil layer.
It is noticeable that the clayey layers (which are
more plastic than the sandy layers) show more non- Fig. 11b Rotation envelope of pile as obtained
linear behavior, and reaches the plastic limit from ALP 19.1

Page 7 of 9
A. K. Yadav & A. Dey

The properties of the soil remain the same as

discussed earlier. Table 3 enumerates the lateral
load capacity of the pile of various diameters under
the extreme scouring condition. It is observed that
the lateral load capacity increases while the length
of fixity is not significantly increased with the
increase in pile diameter.

Fig. 11c Bending Moment envelope of pile as

obtained from ALP 19.1

Fig. 12a Generated P-Y curves for the clayey soil

Fig. 11d Shear force envelope of pile as obtained

from ALP 19.1

Fig. 12b Generated P-Y curves for the sandy soil

Table 3 Response of the pile at lateral load

capacity under 5 m scouring as obtained from ALP
D ycut-off Pu Mmax Lfix
(mm) (mm) (kN) (kNm) (m)
1000 10 70 491 8.5
Fig. 11e Contact stress envelope of pile as obtained 1200 12 134 964 10
from ALP 19.1 1500 15 278 2323 10

Effect of Scouring CONCLUSIONS

The pile as depicted in Fig.3 has been analyzed for
a condition where the maximum scouring was
observed to be 5 m from the existing ground level
(EGL). The rest of the pile passes through 4.5 m of
silty-clay soil underlain by 15.5 m of sandy soil.
Based on the conducted study, the following
important conclusions can be stated:
Mesh refinement in BEF revealed that 151
nodes proves to be sufficient to obtain a stable
and convergent solution and can be considered

Page 8 of 9

as optimal number of nodes for the reported
problems.
BEF models the soil as a linear elastic
material, and hence, the load-deformation
behavior of the soil has also obtained to be
nonlinear. Modeling with ALP resulted in the
nonlinear portrayal of the load-deformation
behavior of the pile, since in this case, the soil
is modeled using nonlinear Winkler springs
and generated P-Y curves.
Owing to enhanced flexural rigidity, lateral
load capacity of a pile increases with the
increase in the diameter of the pile.
As observed from the results of both BEF and
ALP, increase in the diameter of the embedded
piles results in the increase in the maximums
of the flexural responses of the pile.
BEF results revealed that the length of fixity of
the pile increased with the increase in the pile
diameter. However, ALP results revealed that
the length of fixity is not significantly affected
by the change in the pile diameter. This is
attributed to the fact that the length of fixity is
not governed alone by the flexural rigidity of
the pile, but by the relative flexural rigidity of
the pile and the surrounding soil. This feature
is portrayed better by the software ALP and
hence, length of fixity remain primarily
invariant with the change in the pile diameter
Increase in the scouring significantly decreases
the lateral load capacity of the soil. The
degradation is more prominent when scouring
takes place in the softer soil stratum.
The presence of water table, thus reducing the
effective unit weight of the submerged soil,
also decreases the lateral load capacity of the
embedded pile.
P-Y curves generated reveal that the clay
layers show prominence in nonlinear behavior
and reaches the plastic limit under failure load
condition, while sandy layers, commonly
behaving as elastic, show at most bilinear
behavior and indicates that this kind of soil
require high loads to reach into their plastic
limit.
Lateral load capacity of piles in stratified soil deposit subjected to scour
REFERENCES
1. Jones, G. (1997), Analysis of Beams on Elastic
Foundations using Finite Difference Theory,
Thomas Telford, UK.
2. Hetenyi, M. (1946), Beams on Elastic
Foundations: Theory with Applications in the
field of Civil Engineering, University of
Michigan Press, Ann-Arbor.
3. Winkler, E. (1867), Die Lehre von der
Elastizaitat unde Festigkeit, H. Dominicus,
Prague.
4. Yadav A.K. (2013), Lateral Load Capacity of
Piles in stratified Soil Deposit Subjected to
Scour, BTP Report, Department of Civil
Engineering, Indian Institute of Technology,
Guwahati, India.
5. IS 456 (2000) Plain and Reinforced Concrete,
Bureau of Indian Standards, New Delhi, India.
6. Bowles, J.E. (1997), Foundation Analysis and
Design, Tata McGraw Hill, USA.
7. Vesic, A. B. (1961), Beams on elastic subgrade
and the Winklers hypothesis, Proceedings of
the 5th International Conference on Soil
Mechanics and Foundation Engineering, Paris,
France, Vol. 1, 845-850.
8. Dey, A. (2009), Nonlinearly Elastic and
Viscoelastic Unreinforced and Reinforced
Foundation Beds: Modeling and Model
Parameter Estimation, Ph.D. Thesis, IIT
Kanpur.
9. IRC 45 (1972), Recommendations for
estimating the resistance of soil below the
maximum scour level in the design of well
foundations, Indian Roads Congress, New
Delhi.
10. IRC 78 (2000), Standard specifications and
code of practice for road bridges. Section-VII,
Indian Roads Congress, New Delhi.
11. IRC: 5 (1998), Standard specifications and
code of practice for road bridges. Section 1,
Indian Roads Congress, Delhi.
12. Kothyari, U. (2007), Indian practice on
estimation of scour around bridge piers A
comment, Sadhana, Vol. 32, Part 3, 187-197.
13. Alpan, I. (1967), The empirical evaluation of
coefficient of K0 and K0R, Soils and
Foundations, Vol. 7, No. 1, 31-40.
Page 9 of 9