See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/311730533

Effects of scour-hole dimensions and soil

stress history on the behavior of laterally
loaded piles in soft clay under scour conditions

Article in Computers and Geotechnics · April 2017

DOI: 10.1016/j.compgeo.2016.12.008

CITATIONS READS

0 29

3 authors, including:

Fayun Liang
Tongji University
49 PUBLICATIONS 220 CITATIONS

SEE PROFILE

All content following this page was uploaded by Fayun Liang on 02 January 2017.

The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the original document
and are linked to publications on ResearchGate, letting you access and read them immediately.
 Computers and Geotechnics 84 (2017) 198–209

Contents lists available at ScienceDirect

Computers and Geotechnics

journal homepage: www.elsevier.com/locate/compgeo

Research Paper

Effects of scour-hole dimensions and soil stress history on the behavior

of laterally loaded piles in soft clay under scour conditions
Hao Zhang a, Shengli Chen b, Fayun Liang a,⇑
a
Department of Geotechnical Engineering, Tongji University, Shanghai 200092, China
b
Department of Civil & Environmental Engineering, Louisiana State University, Baton Rouge, LA 70803, USA

a r t i c l e i n f o a b s t r a c t

Article history: Bridge pile foundations in the midst of water current are often subjected to scour, which induces the loss
Received 1 September 2016 of soil support around the piles and thus results in a significant decrease of foundation capacities or even
Received in revised form 9 November 2016 the failure of bridges. In the current practice, when the scour-affected behavior of the pile foundations is
Accepted 8 December 2016
analyzed, either the whole scour-hole geometry or the possible changes in the stress history of the
remaining soils is often ignored. In reality, however, scouring creates scour holes with certain dimensions
around the pile foundations and the remaining soils that are not scoured away undergo an unloading
Keywords:
process at the same time, which will increase the over-consolidation ratios of the remaining soils and
Laterally loaded piles
Scour-hole dimensions
accordingly changes of their mechanical properties. This paper examines the behavior of laterally loaded
Soil stress history piles in soft clay under scour conditions by using the p-y method. The conventional p-y curves have been
Scour modified appropriately to reasonably reflect the effects of three-dimensional scour-hole geometry as well
Soft clay as the stress history of the soils, with the aid of integration of Mindlin’s fundamental solutions. A field test
is used to serve as a reference case and to investigate the effects of stress history, scour depth, scour
width, and scour-hole slope angle on the responses of laterally loaded piles in soft clay. The results indi-
cate that neglecting the stress history effect can be unconservative for scour-affected pile foundations in
soft clay, whereas neglect of the scour-hole dimensions and geometry would lead to over-conservative
predictions/design of the laterally loaded piles under scour condition.
Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction flood/scour conditions. Bennett et al. [3] evaluated the behavior

Scour is a natural phenomenon caused by erosion or removal of
streambed or bank material from bridge pile foundations due to
flowing water, which can thus result in a reduced capacity of the
pile foundations in either lateral or vertical direction. It is well
known that scour is a major cause of the bridge failures [4], pro-
ducing significant financial losses and threats to public safety
[21,22,8]. Therefore, it is of great importance and demand to have
rational analysis methods for evaluating the scour susceptibility of
bridge pile foundations.
In contrast to the extensive research on the evaluation of the
scour behavior, only limited number of studies have been con-
ducted to investigate the scour effects on the responses of pile
foundations and their supporting bridges. Daniels et al. [5] investi-
gated the pushover failure of pile bent bridges under extreme
⇑ Corresponding author.
E-mail addresses: zhanghaotumu@163.com (H. Zhang), shenglichen@lsu.edu
(S. Chen), fyliang@tongji.edu.cn (F. Liang).
of a laterally loaded bridge pile group under a scour condition
using the equivalent group pile method. Lin et al. [11], Lin et al.
[12] examined the scour effect on the behavior of laterally loaded
piles considering the stress history of sands and soft clay, respec-
tively. Li et al. [9] conducted a parametric study using the three-
dimensional finite-difference software (FLAC3D) to investigate the
effects of scour depth, scour width, and slope angle on the behavior
of laterally loaded piles in marine clay. Liang et al. [10] presented
an analytical model with modified lateral subgrade modulus to
investigate the extreme scour effect on the buckling of bridge piles
in soft clay considering the stress history of the remaining soils. By
using the wedge failure model and accordingly the modifications of
the widely used p-y curves for piles, Lin et al. [13], Lin et al. [14]
developed a simplified method to analyze the effects of scour-
hole dimensions on laterally loaded piles in both sand and soft
clay.
The above-mentioned literatures have been largely limited to
ignoring either the effect of scour-hole geometry, or that of the
stress history. In reality, however, for the isolated pile foundations

http://dx.doi.org/10.1016/j.compgeo.2016.12.008
0266-352X/Ó 2016 Elsevier Ltd. All rights reserved.
 H. Zhang et al. / Computers and Geotechnics 84 (2017) 198–209 199

in the midst of water current, scour holes formed around the piles
usually involve certain shapes and sizes and meanwhile, the
remaining soils have undergone an unloading process associated
with new stress states. The latter of unloading produces the
changes of stress history of the remaining soils and hence
their resistances provided to the pile foundations as well. Note that
the soil stress history can be characterized by the over-
consolidation ratio (OCR), which is generally defined as the ratio
of the historically maximum stress the soil has experienced to its
current stress.
The objective of this paper is to present a simplified method for
the analysis of laterally loaded piles in soft clay under scour
conditions, which can account properly for the effects of three-
dimensional scour-hole dimensions and the soil stress history
simultaneously. In the proposed model, the stress changes of the
remaining soils at the corresponding pile location due to the
three-dimensional scour-hole formation are first determined based
on the Mindlin’s elastic solutions. Then, the calculated new soil
stresses are adopted to re-examine the undrained shear strength
and other properties of the remaining soils undergoing a stress history
due to scour. With these key soil parameters obtained, the widely
used p-y curves proposed by Matlock [15] are further modified to
reasonably reflect the effects of three-dimensional scour-hole
geometry as well as of the stress history of the soils. A field test
reported in Matlock [15] is subsequently used as a case-study
example to analyze the scour-affected behavior of laterally loaded
piles in soft clay. Finally, the effects of stress history, scour depth,
scour width, and scour-hole slope angle on the laterally loaded
piles are discussed based on the computed results from the
simplified method.

Fig. 1. Soil and pile profile for studies under scour conditions: (a) front view and;
2. Analysis method (b) plan view.

2.1. Definition of the problem

The pile is assumed to be embedded in a normally consolidated Similarly, the triangle distributed loads Psidewall
p that are acted
homogenous soft clay before the scour event, which has an parallel to the surrounding sidewalls of the scour hole can be
effective unit weight of c0 , Poisson’s ratio of m, and effective internal written as
friction angle of u0 . Furthermore, the scour hole, which is modeled
ð1  K 0 Þ 0
as a truncated circular cone with the dimensions defined by the Psidewall
p ¼ c zp sin 2a ð2Þ
2
scour depth, the scour width and the scour-hole slope angle, is
assumed to be formed symmetrically around the pile as shown In Eqs. (1) and (2), zp is the depth coordinate of loading action
in Fig. 1. In this figure, the center of the scour hole at the ground point (ranging from 0 to Sd) and K0 refers to the earth pressure
surface is set to be the origin of the axisymmetrical cylindrical coefficient at rest. In the absence of test data, the empirical formula
coordinate system (r, h, z). Sd is the scour depth at the pile, i.e., [16] can be used, as shown in Eq. (3)
the distance between the pre-scour and post-scour ground surface;
K 0 ¼ ð1  sin u0 ÞOCRsin u
0

scour width, Sw, refers to the bottom width of the scour hole
throughout this study for simplicity, which is measured from one
edge of the scour hole to the other; the slope formed inside the
scour hole is represented by the scour-hole slope angle, a. It should
be noted that the scour shapes and dimensions are assumed to be
known in this study for the demonstration purposes, although
their determinations were focused in other research [4,21,18,1].
2.2. Calculation for the scour-induced stress changes of soils
Consider the scour-hole geometry and its symmetry as shown
in Fig. 1. The circular uniform loads Pbottom
v , which can be calculated
by c0 Sd, are acted vertically on the scour-hole bottom before the
scour is underway. Simultaneously, based on Mohr’s stress circle,
the triangle distributed loads Psidewall
n that are acted normal to the
surrounding sidewalls of the scour hole can be expressed as
ð1 þ K 0 Þ 0 ð1  K 0 Þ 0
Psidewall
n ¼ c zp þ c zp cos 2a
2 2
ð3Þ
where OCR denotes the over-consolidation ratio and particularly,
for the normally consolidated homogenous soft clay before scour,
the earth pressure coefficient at rest is thus given by
K 0 ¼ 1  sin u0 ð4Þ
bottom
When the scour hole is formed, equivalent values of P v ,
Psidewall
n and P sidewall
p , which have been changed into opposite direc-
tion, are applied to the bottom and surrounding sidewalls to sim-
ulate the scour-induced unloading effects. Then, Mindlin [17]
Green’s function for vertical and horizontal loads in a semi-
infinite half-space can be thus used to calculate the scour-
induced soil unloading stress. To facilitate the use of the Green’s
function, the triangle distributed loads Psidewall and Psidewall need to
n p
be rewritten as the corresponding triangle distributed vertical
loads Psidewall
v and horizontal loads Psidewall
h , which are also acted
reversely on the surrounding sidewalls of the scour hole to simu-
ð1Þ
late the unloading effects. Hence
200 H. Zhang et al. / Computers and Geotechnics 84 (2017) 198–209

rulz ¼ rbottom þ rsidewall v þ rsidewallh

Psidewall
v ¼ Psidewall
n cos a þ Psidewall
p sin a ð5aÞ z z z ð11Þ
The double integrals in Eqs. (6), (8) and (10) can be calculated
by the mean of Gauss-Legendre numerical technology [6].
Psidewall
h ¼ Psidewall
n sin a  P sidewall
p cos a ð5bÞ

Accordingly, considering the loads are symmetrical about the z
axis (pile axis), the vertical unloading stress at the point of pile axis
(0, 0, z) due to the unit vertical loading P bottom
v rdrdh acting at the
point (r, h, Sd) of scour-hole bottom can be given as follows
2.3. Solution for soil stresses and properties after scour
After scour, according to the first step analysis, the present
stress rsc
z of the remaining soil at the point of pile axis (0, 0, z),

ZZ ( )
Pbottom
v ð1  2tÞðz  Sd Þ 3ðz  Sd Þ3 ð1  2tÞðz  Sd Þ ½3ð3  4tÞzðz þ Sd Þ2  3Sd ðz þ Sd Þð5z  Sd Þ 30zSd ðz þ Sd Þ3
rbottom ¼ þ  þ þ rdrdh
Cb 8pð1  tÞ
z
R31 R51 R32 R52 R72
ð6Þ

where Ub is the integral domain for the scour-hole bottom, the vari- which can consider the effect of three-dimensional scour-hole
ables R1 and R2 are geometry, is thus given by
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
R1 ¼ r 2 þ ðz  Sd Þ2 ; R2 ¼ r2 þ ðz þ Sd Þ2 ð7Þ rscz ¼ rint
z  rz
ul
ð12Þ

Then since the radius coordinate of loading action point of

scour-hole sidewall varies with its depth coordinate (i.e., where rint
z is the soil stress at the point of pile axis (0, 0, z) before
r ¼ Sw =2 þ ðSd  zp Þ= tan a), the vertical unloading stress at the scour, which can be calculated by c0 z.
point of pile axis (0, 0, z) due to the unit vertical loading In the second step, to further address the stress history effect on
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi the behavior of laterally loaded piles, the key soil parameters (i.e.,
2
Psidewall
v r 1 þ ðdr=dzp Þ dzp dh acting at the point (r, h, zp) of scour- effective unit weight, strain at 50% of the maximum stress, and
hole sidewall can be obtained as undrained shear strength) of the p-y curve for soft clay [15] after

ZZ ( )
sidewallv Psidewall
v ð1  2tÞðz  zp Þ 3ðz  zp Þ3 ð1  2tÞðz  zp Þ ½3ð3  4tÞzðz þ zp Þ2  3zp ðz þ zp Þð5z  zp Þ 30zzp ðz þ zp Þ3
r ¼ þ  þ þ
Cs 8pð1  tÞ
z
R33 R53 R34 R54 R74
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2
r 1 þ ðdr=dzp Þ dzp dh ð8Þ

scour, should be re-examined by using the present stress rsc z of

where Us is the integral domain for the scour-hole sidewall, the the remaining soil calculated by Eq. (12). Also, since the strain at
variables R3 and R4 are 50% of the maximum stress only depends on the soil undrained
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi shear strength and can be also selected based on the consistency
R3 ¼ r 2 þ ðz  zp Þ2 ; R4 ¼ r2 þ ðz þ zp Þ2 ð9Þ of clay soils [20], which does not change during scour, the key soil
parameters needed to be re-examined are limited to the effective
Similarly, due to the unit horizontal loading
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi unit weight and the undrained shear strength. In other words, once
2
Psidewall
h r 1 þ ðdr=dzp Þ dzp dh acting at the point (r, h, zp) of scour- the soil’s effective unit weight and the undrained shear strength
hole sidewall, the vertical unloading stress at the point of pile axis have been determined with the remaining soil stress given by Eq.
(0, 0, z) can be expressed as (12), the p-y curve proposed by Matlock [15] can be readily modi-
fied to consider the effects of scour-hole dimensions and stress his-

ZZ ( )
ð1  2tÞ 3ðz  zp Þ2 ð1  2tÞ ½3ð3  4tÞðz þ zp Þ  6z2p  6zp ð1  2tÞðz þ zp Þ 30zzp ðz þ zp Þ2
2
Psidewall r
rzsidewallh ¼ h
 þ þ þ 
Cs 8pð1  tÞ R33 R53 R34 R54 R74
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2
r 1 þ ðdr=dzp Þ dzp dh ð10Þ

tory simultaneously. The detailed modifications to the p-y curves

Finally, the scour-induced soil unloading stress rul
z at the point
are described as below.
of pile axis (0, 0, z) can be obtained by considering the unloading Before scour, the major effective stress r01 and the minor effective
effects for the bottom and sidewall of scour hole stress r03 at the point of pile axis (0, 0, z) are respectively shown in
 H. Zhang et al. / Computers and Geotechnics 84 (2017) 198–209 201

r01 ¼ rint
z ¼ cz
0
ð13aÞ Hence, by combining Eqs. (14), (16), (17) and (21), the
undrained shear strength after scour can be finally determined
r03 ¼ K 0 rint
z ¼ K0c z
0
ð13bÞ by Eq. (23)
!K1
where K0 refers to the earth pressure coefficient at rest before scour, ð3  2 sin u0 ÞOCR
C sc ¼ C int ð23Þ
which is previously given by Eq. (4). Thus the mean effective stress 1 þ 2ð1  sin u0 ÞOCRsin u
u u 0

is determined
p0int ¼ ðr01 þ 2r03 Þ=3 ¼ c0 zð3  2 sin u0 Þ=3 ð14Þ
2.4. Modified p-y curve and pile response considering scour effect
After scour, the major effective stress r01 and the minor effective
stress r03 at the point of pile axis (0, 0, z) are respectively computed In practice, the p-y curve proposed by Matlock [15] is widely
used in the p-y method to analyze the behavior of laterally loaded
r ¼r ¼r r
0
1
sc
z
int
z
ul
z ð15aÞ piles in soft clay, and is formulated as follows
 1=3
r03 ¼ K 0 rscz ¼ K 0 ðrint
z  rz Þ
ul
ð15bÞ p
¼ 0:5
y
ð24Þ
pult y50
where K0 is the earth pressure coefficient at rest after scour, which
can be determined by using Eq. (3). Accordingly, the mean effective where p is lateral soil resistance per unit length of a pile; y is lateral
stress is calculated as shown soil displacement; y50 is lateral displacement at half the maximum
soil stress, which can be determined by Eq. (25); pult is ultimate soil
sin u 0
p0sc ¼ ðr01 þ 2r03 Þ=3 ¼ rsc
z ½1 þ 2ð1  sin u ÞOCR
0
=3 ð16Þ resistance per length, which is equal to the smaller value of pult1 and
pult2 calculated in Eqs. (26) and (27); and when y is greater than
where the over-consolidation ratio, OCR, is defined as
8y50 , p remains a constant value of pult
OCR ¼ rint
z =rz ¼ c z=rz
sc 0 sc
ð17Þ
y50 ¼ 2:5e50 D ð25Þ
By using Eqs. (14), (16) and (17), the specific volume after scour, vsc,  
can be expressed by Eq. (18) based on the compression and swelling c0 J
pult1 ¼ 3 þ z þ z CuD ð26Þ
curve Cu D
 0 
pint
v sc ¼ v int þ j ln pult2 ¼ 9C u D ð27Þ
p0sc
! where D is pile diameter; e50 is the strain at one-half the maximum
c0 zð3  2 sin u0 Þ
¼ v int þ j ln ð18Þ stress, typically between 0.01 and 0.02; z is depth below the pre-
rz ½1 þ 2ð1  sin u0 ÞOCRsin u 
0
sc
scour ground surface; C u is undrained shear strength of the soft
clay; and J is a constant value, typically using 0.5.
where j is the swelling index from isotropic consolidation tests;
By substituting the soil effective unit weight after scour
and v int is the specific volume before scour, which can be deter-
(Eq. (20)) and undrained shear strength after scour (Eq. (23)) into
mined by using Eq. (19) if the soil is saturated
Eqs. (26) and (27), the equations for the ultimate resistance of soft
cw þ cw w
v int ¼ ð19Þ clay after scour can be rewritten as shown in Eqs. (28) and (29)
cw  c0 w !K1
int ð3  2 sin u0 ÞOCR
where cw is the unit weight of water; and w refers to the soil pult1 ¼ ½3D þ Jðz  Sd ÞC u
1 þ 2ð1  sin u0 ÞOCRsin u
0

moisture content. ( ," ! #)

ð3  2 sin u0 ÞOCR
Then, the soil effective unit weight after scour can be obtained þ v int c0 v int þ j ln ðz  Sd ÞD
½1 þ 2ð1  sin u0 ÞOCRsin u 
0
by
ð28Þ
c0sc ¼ v int c0 =v sc
," !#
ð3  2 sin u0 ÞOCR !K1
¼ v int c0 v int þ j ln ð20Þ ð3  2 sin u0 ÞOCR
0 ÞOCR sin u pult2 ¼ 9DC int
0
½1 þ 2ð1  sin u  ð29Þ
1 þ 2ð1  sin u0 ÞOCRsin u
u 0

In addition, as the normally consolidated soil after scour

That is to say, once the soil parameters before scour (e.g.,
becomes over-consolidated and the earth pressure coefficient at
effective unit weight of c0 , Poisson’s ratio of m, soil effective internal
rest increases during the unloading process (see Eq. (3)), the rela-
friction angle of u0 , moisture content w, and undrained shear
tionship of the undrained shear strengths before and after scour
strength Cu) and the scour-hole dimensions (e.g., scour depth Sd,
can be established in terms of the mean effective stress by using
scour width Sw, and scour-hole slope angle a) are known, the p-y
Eqs. (21) and (22), which were derived based on the modified
curve described in Eq. (24) can be modified by applying Eqs. (17),
Cam-clay model [19]
(19), (24), (28), and (29) to consider the effects of scour-hole
 0 K
ðC sc 0
u =psc Þ ðC u =p0 ÞOC pint dimensions and stress history simultaneously. Then, the modified
¼ ¼ ð21Þ p-y curve is exported into the commercial software LPILE 6.0 to
ðC int 0
u =pint Þ
ðC u =p0 ÞNC p0sc
determine the scour affected response of laterally loaded piles in
soft clay.
kj C ur
K¼ ¼1 ðapproximately 0:8Þ ð22Þ
k Cc
3. Validation of the simplified model and case study
where C int sc
u and C u are soil undrained shear strength before and after
scour, respectively; k is the compression index from isotropic con- A thorough review of the literature available in this field indi-
solidation tests; Cc and Cur are compression and swelling indexes cated that experimental data regarding scour effects on the lateral
obtained from the oedometer tests, respectively. pile response is relatively scarce. Therefore, the simplified model is
202 H. Zhang et al. / Computers and Geotechnics 84 (2017) 198–209

verified by comparison with Lin et al. [12], Lin et al. [14] solution 120
for scour-affected laterally loaded piles (only considering stress Considering stress history effect, Sd =5D
history effect or scour-hole dimensions respectively) based on an
100
in-situ pile lateral-load test in soft clay conducted near Lake
Austin, Texas [15], which is subsequently selected as a case-

Lateral load, Ft (kN)

study example to further investigate the effects of scour-hole 80
dimensions and stress history on the behavior of laterally loaded
piles in the remaining soil under scour condition.
60

3.1. Site conditions Lin et al. (2014a), ignoring scour-

40 hole geometry, Sw=∞
The soil is a fat clay (Unified Soil Classification System classifi- Present method, ignoring scour-
cation of CH) with its properties summarized in Table 1. Some of hole geometry, Sw=∞
20 Present method, considering scour-
these properties are estimated based on empirical relationships
for demonstration purposes. They can be more accurately deter- hole geometry, Sw=0, α=40
mined by laboratory or in situ tests. The compression index, Cc, is 0
0 50 100 150 200 250
estimated based on the moisture content [2], and the swelling
index, Cur, is taken as (1/5) Cc [7]. Note that usually k and j are Lateral pile-head deflection, yt (mm)
assumed approximately equal to Cc/2.3 and Cur/2.3, respectively.
Fig. 2. Pile head lateral deflection versus lateral load.
Because no value for the effective friction angle and Poisson’s ratio
were provided from Lake Austin test, a typical friction angle of
u0 ¼ 20 and Poisson’s ratio of t ¼ 0:3 [23] are assumed for the soft 70
clay in this case study. The soil specific volume before scour, v int , is
estimated to be 2.6 in Eq. (19). The water table is kept above the 60
ground surface. The test pile is a pipe section with its parameters
summarized in Table 2. 50
Lateral load, Ft (kN)

3.2. Verification 40

Fig. 2 shows the comparison of the pile head deflection versus 30

lateral load curves from Lin et al. [12] solution and computed by Considering scour-hole geometry,
the presented method with scour depth of 5D. Reasonable agree- Sd =6D, Sw=0, α=40
20
ment has been achieved by these two models on the deflection
Lin et al. (2016), ignoring stress history
at the pile head, although Lin et al. [12] solution seems to over pre-
10 Present method, considering stress history
dict the deflection as shown in Fig. 2. This is because Lin et al. [12]
solution ignored the scour-hole geometry (Sw = 1), while the scour
width of 0 and scour-hole slope angle of 40 are assumed herein to 0
0 20 40 60 80 100 120
consider the scour-hole dimensions by using the presented
Lateral pile-head deflection, yt (mm)
method. As a matter of fact, if we ignore the scour-hole geometry
in the present model (Sw = 1), the present model predicts larger Fig. 3. Lateral load-deflection curves of piles.
deflection at the pile head, which is much closer to the results from
Lin et al. [12], as shown in Fig. 2. This is expected as the soil
resistance from the soils below the base of the scour hole and away history effect of the remaining soil is ignored. In reality, however,
from the pile will be enhanced to some extent due to the still the over consolidation induced by the removal of the soil changes
presence of the soils above the scour level. the in situ soil state and thus the soil resistances provided to the
In addition, Fig. 3 compares the present model with that of Lin pile foundations.
et al. [14] on laterally loaded piles under scour condition (Sd = 6D,
Sw = 0, a ¼ 40 ). It could be observed that the comparison of the 3.3. Parametric analysis
scour affected response of laterally loaded piles reveals reasonable
agreement. Fig. 3 also shows that Lin et al. [14] solution leads to an For the purpose of the case study considered in this paper,
underestimation of the deflection at the pile head, as the stress extreme scour-hole dimensions (Sd = 5D or 10D, Sw = 1, a ¼ 0 ,

Table 1
Properties of soft clay.

Effective unit Water content, w (%) Effective friction Strain at half

weight, c0 (kN/m3) angel, u0 ( ) maximum stress, e50
10 44.5 20 0.012
Compression Index, Cc Swelling index, Cur Poisson’s Ratio, t Undrained shear strength, Cu (kPa)
0.40 0.08 0.3 32.3

Table 2
Pile parameters.

Length L (m) Outer diameter D (m) Thickness t (m) Moment of inertia Ip (m4) Elastic modulus Ep (kPa) Yielding moment My (kN m)
12.8 0.319 0.0127 1.44  104 2.18  108 231
 H. Zhang et al. / Computers and Geotechnics 84 (2017) 198–209 203

i.e., total removal of the soil layer to the scour depth without con- 40
sidering the effect of scour-hole dimensions) are first assumed
herein to compare the scour-affected response of laterally loaded
piles by considering and ignoring the stress history effect. Then, 30
to evaluate the scour depth effects, the scour depth is varied from
0, 2D, 4D, 6D, to 8D, whereas the scour width (Sw ¼ 0 or 1) and the

p (kN/m)
scour-hole slope angle (a ¼ 40 ) are kept unchanged. Additionally, 20
five scour widths (i.e., Sw = 0, 5D, 10D, 15D, and 1) are considered
to evaluate their effects on the response of laterally loaded piles, Unmodified
while the scour depth is constant at 4D and the scour-hole angle 10 Modified, Sd =5D
is 40 . Finally, four scour-hole slope angles of 60 , 40 , 20 , and
0 are investigated, when the scour depth is maintained constant Modified, Sd =10D

at 4D and the scour width is 0. 0

It is worth noting that the remaining soils undergo an unloading 0 50 100 150 200 250
process due to scour and thus high OCR can be obtained near the y (mm)
soil surface. To avoid a passive earth failure, however, a limited
Fig. 4. Effect of stress history of soft clay on the p-y curves at the soil depth of half
OCR of 27 can be determined for this study (u0 ¼ 20 ) by using pile diameter (0.5D) below the post-scour ground surface.
Eq. (30) [16], which is based on the Ko-OCR relationship in soils.
" #ð1= sin u0 Þ 120
1 þ sin u0
OCRlimit ¼ ð30Þ
ð1  sin u0 Þ
2
100

Therefore, the OCRlimit in Eq. (30) aids in determination of soil

Lateral load, Ft (kN)

depth below which the effective unit weight and undrained shear 80
strength derived from Eqs. (20) and (23) are valid.
The computed results obtained by the simplified method after 60 Measured
analyzing a free-head single pile in soft clay (soil properties from Model before scour
the Lake Austin test) with different scour-hole dimensions under 40 Unmodified, Sd =5D
lateral loading, considering or ignoring stress history, are compared Modified, Sd =5D
in the following discussion. Comparisons of the results include the 20 Unmodified, Sd =10D
properties of the remaining soils after scour, the p-y curves or its Modified, Sd =10D
ultimate soil resistance per length, pile-head deflection, and
0
maximum bending moment of the pile. In addition, discussion is 0 100 200 300 400 500
provided on the effects of stress history, scour depth, width, and Lateral pile-head deflection, yt (mm)
slope angle on the response of the laterally loaded single pile.
Fig. 5. Effect of stress history of soft clay on the relationship of lateral load versus
lateral pile-head deflection.
4. Results and discussion

4.1. Effect of stress history under the extreme scour condition (i.e., Sd = 5D or 10D, Sw = 1,
a ¼ 0 ) as presented in Tables 3 and 4. The results show that the
Based on the procedures discussed earlier for modifying the p-y undrained shear strength of the soft clay after scour decreases sig-
curves, the properties of the remaining soft clay are determined nificantly when compared with those before scour, while a minor

Table 3
Calculated soil properties considering effect of stress history (Sd = 5D, Sw = 1, a ¼ 0 ).

Soil depth from mud line before/after scour (m) c0 (kN/m3) c0sc (kN/m3) C int
u (kPa)
C sc
u (kPa)
OCR

2.30/0.70 10.00 9.88 32.3 26.8 3.3

3.39/1.80 10.00 9.93 32.3 29.2 1.9
3.70/2.10 10.00 9.94 32.3 29.5 1.8
4.30/2.71 10.00 9.95 32.3 30.0 1.6
5.69/4.10 10.00 9.96 32.3 30.6 1.4
7.25/5.66 10.00 9.97 32.3 31.0 1.3
9.47/7.88 10.00 9.98 32.3 31.4 1.2
15.0/13.41 10.00 9.99 32.3 31.7 1.1

Table 4
Calculated soil properties considering effect of stress history (Sd = 10D, Sw = 1, a ¼ 0 ).

Soil depth from mud line before/after scour (m) c0 (kN/m3) c0sc (kN/m3) C int
u (kPa)
C sc
u (kPa)
OCR

3.39/0.20 10.00 9.72 32.3 20.9 17.0

3.70/0.51 10.00 9.80 32.3 23.7 7.3
4.30/1.11 10.00 9.86 32.3 26.1 3.9
5.69/2.50 10.00 9.91 32.3 28.3 2.3
7.25/4.06 10.00 9.94 32.3 29.4 1.8
9.47/6.28 10.00 9.96 32.3 30.2 1.5
15.0/11.81 10.00 9.97 32.3 31.1 1.3
204 H. Zhang et al. / Computers and Geotechnics 84 (2017) 198–209

120 decrease in the effective unit weight are observed, which can be
negligible during scour.
Then, a comparison of the unmodified and modified p-y curves
100
after scour (Sd = 5D and 10D) at a soil depth of 0.5D below the post-
scour ground surface are presented in Fig. 4. In Fig. 4, when the
Lateral load, Ft (kN)

80
scour depths are 5D and 10D, the modified p-y curve shows
approximately 17–35% lower ultimate soil resistance than the
60 Measured unmodified curve at a given lateral displacement. These results
Model before scour demonstrate that the approach ignoring the stress history effect
40 Unmodified, Sd =5D and keeping the properties of the remaining soil unchanged over-
Modified, Sd =5D estimates the lateral soil resistance to the pile foundation in soft
20 Unmodified, Sd =10D clay affected by scour.
Modified, Sd =10D Furthermore, lateral pile-head deflections and maximum bend-
0 ing moments under various lateral loads are computed for the pre-
0 100 200 300 400 500 scour condition and compared with the field measurement in
Maximum bending moment, Mmax (kN.m) Figs. 5 and 6, respectively, to further verify the reasonableness of
the simplified model. Also in Figs. 5 and 6, lateral pile-head deflec-
Fig. 6. Effect of stress history of soft clay on the relationship of lateral load versus tions and maximum bending moments are calculated by using the
maximum bending moment.
unmodified and modified p-y curves for the post-scour case. It is

Table 5
Calculated soil properties considering effects of scour-hole dimensions and stress history (Sd = 2D, Sw = 0 or 1, a ¼ 40 ).

Soil depth from mud line before/after scour (m) ðc0sc Þ0 (kN/m3) ðc0sc Þ1 (kN/m3) ðC u Þ0 (kPa) ðC u Þ1 (kPa) OCR0 OCR1

1.07/0.43 9.98 9.91 31.4 28.0 1.2 2.5

2.30/1.66 10.00 9.97 32.2 30.7 1.0 1.4
3.39/2.75 10.00 9.98 32.3 31.2 1.0 1.2
3.70/3.06 10.00 9.98 32.3 31.3 1.0 1.2
4.30/3.66 10.00 9.98 32.3 31.5 1.0 1.2
5.69/5.05 10.00 9.99 32.3 31.7 1.0 1.1
7.25/6.61 10.00 9.99 32.3 31.8 1.0 1.1
9.47/8.83 10.00 9.99 32.3 31.9 1.0 1.1
15.0/14.36 10.00 10.00 32.3 32.1 1.0 1.0

Table 6
Calculated soil properties considering effects of scour-hole dimensions and stress history (Sd = 4D, Sw = 0 or 1, a ¼ 40 ).

Soil depth from mud line before/after scour (m) ðc0sc Þ0 (kN/m3) ðc0sc Þ1 (kN/m3) ðC u Þ0 (kPa) ðC u Þ1 (kPa) OCR0 OCR1

1.62/0.34 9.94 9.84 29.7 25.3 1.7 4.8

2.30/1.02 9.98 9.92 31.6 28.4 1.2 2.3
3.39/2.11 10.00 9.95 32.1 29.9 1.0 1.6
3.70/2.42 10.00 9.96 32.1 30.2 1.0 1.5
4.30/3.02 10.00 9.96 32.2 30.5 1.0 1.4
5.69/4.41 10.00 9.97 32.3 31.0 1.0 1.3
7.25/5.97 10.00 9.98 32.3 31.3 1.0 1.2
9.47/8.19 10.00 9.98 32.3 31.6 1.0 1.2
15.0/13.72 10.00 9.99 32.3 31.8 1.0 1.1

Table 7
Calculated soil properties considering effects of scour-hole dimensions and stress history (Sd = 6D, Sw = 0 or 1, a ¼ 40 ).

Soil depth from mud line before/after scour (m) ðc0sc Þ0 (kN/m3) ðc0sc Þ1 (kN/m3) ðC u Þ0 (kPa) ðC u Þ1 (kPa) OCR0 OCR1

2.30/0.39 9.93 9.82 29.0 24.5 2.0 5.9

3.39/1.48 9.98 9.91 31.5 28.3 1.2 2.3
3.70/1.79 9.99 9.92 31.7 28.8 1.1 2.1
4.30/2.39 9.99 9.94 31.9 29.4 1.1 1.8
5.69/3.78 10.00 9.96 32.1 30.3 1.0 1.5
7.25/5.34 10.00 9.97 32.2 30.8 1.0 1.4
9.47/7.56 10.00 9.98 32.3 31.2 1.0 1.3
15.0/13.09 10.00 9.99 32.3 31.6 1.0 1.1

Table 8
Calculated soil properties considering effects of scour-hole dimensions and stress history (Sd = 8D, Sw = 0 or 1, a ¼ 40 ).

Soil depth from mud line before/after scour (m) ðc0sc Þ0 (kN/m3) ðc0sc Þ1 (kN/m3) ðC u Þ0 (kPa) ðC u Þ1 (kPa) OCR0 OCR1

2.85/0.30 9.88 9.78 27.0 22.8 3.1 9.5

3.39/0.84 9.96 9.86 30.2 25.9 1.5 4.0
3.70/1.15 9.97 9.88 30.8 26.9 1.3 3.2
4.30/1.75 9.98 9.91 31.4 28.0 1.2 2.5
5.69/3.14 9.99 9.94 31.9 29.4 1.1 1.8
7.25/4.70 10.00 9.95 32.1 30.1 1.0 1.5
9.47/6.92 10.00 9.97 32.2 30.7 1.0 1.4
15.0/12.45 10.00 9.98 32.3 31.3 1.0 1.2
 H. Zhang et al. / Computers and Geotechnics 84 (2017) 198–209 205

40 Lateral pile-head deflection, yt (mm)

0 50 100 150 200 250
Ultimate soil resistance per length, pult

35 0

30 1
Modified , F t =40kN, Sw=0
2 Modified , F t =40kN, Sw=
25

D)
Modified , F t =80kN, Sw=0
(kN/m)

3
20

Scour depth, Sd (
Modified , F t =80kN, Sw=
4
15
5
Modified , Sw=0
10 6
5 Modified , Sw= 7

0 8
0 2 4 6 8 10
Scour depth, Sd ( D) 9

Fig. 10. Lateral pile-head deflection versus scour depths (Sw ¼ 0 or 1, a ¼ 40 ).
Fig. 7. Effect of scour depth of soft clay on the ultimate soil resistance per length at
different scour depths (z ¼ 0:5D).

Maximum bending moment, Mmax (kN.m)

120
0 50 100 150 200 250 300 350
0
100 Modified , F t =40kN, Sw=0
1
Modified , F t =40kN, Sw=
2
Lateral load, Ft (kN)

80 Modified , F t =80kN, Sw=0

D)

Measured 3 Modified , F t =80kN, Sw=

Scour depth, Sd (

60 Model before scour

Modified ,S d =2D, S w=0 4
Modified ,S d =2D, S w=
Modified ,S d =4D, S w=0 5
40
Modified,S d =4D, S w=
Modified ,S d =6D, S w=0 6
20 Modified ,S d =6D, S w=
Modified ,S d =8D, S w=0 7
Modified ,S d =8D, S w=
8
0
0 50 100 150 200 250 300 350
9
Lateral pile-head deflection, yt (mm)
Fig. 11. Maximum bending moment at various scour depths(Sw ¼ 0 or 1, a ¼ 40 ).
Fig. 8. Effect of scour depth of soft clay on the relationship of lateral load versus
lateral pile-head deflection (a ¼ 40 ).
pile has approximately 4–6% higher maximum bending moment
using the modified p-y curves than that using the unmodified p-y
120 curves. The results presented in Figs. 5 and 6 further indicate that
considering the effect of stress history of the remaining soil results
100 in decreased lateral support for the pile and thus a conservative
analysis of the laterally loaded piles under scour.
Lateral load, Ft (kN)

80
Measured 4.2. Effect of scour depth
60 Model before scour
Modified ,S d =2D, S w=0 To study this effect, the properties of the remaining soft clay for
Modified ,S d =2D, S w=
40 Modified ,S d =4D, S w=0 the scour-hole dimensions with five scour depths, two scour
Modified,S d =4D, S w= widths, and one scour-hole slope angle (i.e., Sd = 0, 2D, 4D, 6D, to
Modified ,S d =6D, S w=0 8D, Sw = 0 or 1, a ¼ 40 ) are first calculated and compared in
20 Modified ,S d =6D, S w=
Modified ,S d =8D, S w=0 Tables 5–8. The results indicate that considering the scour-hole
Modified ,S d =8D, S w= dimensions (Sw = 0) results in a higher undrained shear strength
0 and effective unit weight of the soft clay when compared with
0 100 200 300 400 500
Maximum bending moment, Mmax (kN.m)
the case in which scour-hole dimensions are neglected (Sw = 1).
Furthermore, the OCR increases with increasing scour depth, while
Fig. 9. Effect of scour depth of soft clay on the relationship of lateral load versus it decreases with soil depth and gradually approaches unity at
maximum bending moment (a ¼ 40 ). greater depths.
Fig. 7 then illustrates the calculated ultimate soil resistance per
pile length, pult at soil depth, z = 0.5D below the post-scour ground
shown that the calculated deflections and bending moments match surface using the simplified method. When the scour depth is
well with the field measurements. Under the scour conditions increased from 0 to 8D, the difference of the ultimate soil resis-
(Sd = 5D and 10D), considering the stress history effect results in tance per length, pult, by considering and ignoring the effects of
12–16% higher pile-head deflection when compared with the case the scour-hole dimensions becomes increasingly evident, i.e., the
in which stress history effect is neglected. Similarly, after scour, the percentage difference (difference of pult between Sw = 0 and
206 H. Zhang et al. / Computers and Geotechnics 84 (2017) 198–209

Table 9
Calculated soil properties considering effects of scour-hole dimensions and stress history (Sd = 4D, a ¼ 40 ).

Soil depth from mud line Soil properties Sw = 0 Sw = 5D Sw = 10D Sw = 15D Sw = 1

before/after scour (m)
1.62/0.34 c0 sc (kN/m3) 9.94 9.90 9.89 9.87 9.84
(Cu)sc (kPa) 29.7 27.7 27.1 26.3 25.3
OCR 1.7 2.6 3.0 3.7 4.8
2.30/1.02 c0 sc (kN/m3) 9.98 9.96 9.94 9.93 9.92
(Cu)sc (kPa) 31.6 30.2 29.6 29.1 28.4
OCR 1.2 1.5 1.7 1.9 2.3
3.39/2.11 c0 sc (kN/m3) 10.00 9.99 9.98 9.97 9.95
(Cu)sc (kPa) 32.1 31.6 31.1 30.7 29.9
OCR 1.0 1.1 1.3 1.4 1.6
3.70/2.42 c0 sc (kN/m3) 10.00 9.99 9.98 9.97 9.96
(Cu)sc (kPa) 32.1 31.8 31.3 31.0 30.2
OCR 1.0 1.1 1.2 1.3 1.5
4.30/3.02 c0 sc (kN/m3) 10.00 9.99 9.99 9.98 9.96
(Cu)sc (kPa) 32.2 32.0 31.7 31.4 30.5
OCR 1.0 1.1 1.1 1.2 1.4
5.69/4.41 c0 sc (kN/m3) 10.00 10.00 10.00 10.00 9.97
(Cu)sc (kPa) 32.3 32.1 32.0 31.8 31.0
OCR 1.0 1.0 1.1 1.1 1.3
7.25/5.97 c0 sc (kN/m3) 10.00 10.00 10.00 10.00 9.98
(Cu)sc (kPa) 32.3 32.2 32.1 32.0 31.3
OCR 1.0 1.0 1.0 1.1 1.2
9.47/8.19 c0 sc (kN/m3) 10.00 10.00 10.00 10.00 9.98
(Cu)sc (kPa) 32.3 32.3 32.2 32.2 31.6
OCR 1.0 1.0 1.0 1.0 1.2
15.0/13.72 c0 sc (kN/m3) 10.00 10.00 10.00 10.00 9.99
(Cu)sc (kPa) 32.3 32.3 32.3 32.3 31.8
OCR 1.0 1.0 1.0 1.0 1.1

Sw = 1 divided by pult at Sw = 1) increases with the scour depth:

160
11.9%, 17.1%, 18.0% and 18.1% at the above scour depths.
Lateral pile-head deflection, yt (mm)

By using modified p-y curves, Figs. 8 and 9 examine the lateral

pile-head deflections and maximum bending moments under var-
ious lateral loads for the aforementioned post-scour condition, 140
respectively. It is shown that as the scour depth increases, the dif-
ference in the lateral load-deflection results by considering a circu- Modified , F t =80kN, Sd =4D, α=40
lar cone scour hole (Sw = 0) and a complete removal of the soil layer 120
Modified, F t =100kN, Sd =4D, α=40
to the scour depth (Sw = 1) becomes increasingly substantial. For
example, at a lateral load of 80 kN, the pile-head deflection consid-
ering a total removal of the soil layer are 7.2%, 7.9%, 8.8%, and 9.0%
100
higher than those considering the scour-hole dimensions when the
scour depths are 2D, 4D, 6D, and 8D, respectively. Also under the
scour condition, considering the effect of scour-hole dimensions
80
0 5 10 15
36 Scour width, Sw ( D)
Ultimate soil resistance per length, pult

34 Fig. 13. Lateral pile-head deflection versus scour width (Sd = 4D and a = 40°).

32

30 results in (approximately 3% at a lateral load of 80 kN) lower max-

imum bending moments when compared with the case in which
(kN/m)

28 scour-hole dimensions are neglected.

Figs. 10 and 11 present the variation of the lateral pile-head
26 deflection and maximum bending moments at different scour
depths obtained by considering and ignoring the scour-hole
24
Modified , Sd =4D, α=40 dimensions, respectively. The results show that the lateral pile-
22 head deflection and maximum bending moments increase signifi-
cantly as scour depth increases, and particularly the lateral
20 pile-head deflection increases with scour depth in a nonlinear
0 5 10 15
manner (at an accelerating rate). When the scour depth increases
Scour width, Sw ( D)
from 0 to 8D, the lateral pile-head deflection and maximum
Fig. 12. Effect of scour width of soft clay on the ultimate soil resistance per length bending moment increase by approximately 500% and 200%,
(z ¼ 0:5D). respectively.
 H. Zhang et al. / Computers and Geotechnics 84 (2017) 198–209 207

280 40
Maximum bending moment, Mmax (kN.m)

Ultimate soil resistance per length, pult

38
260 36
34
240 Modified , F t =80kN, Sd =4D, α=40 32

(kN/m)
Modified, F t =100kN, Sd =4D, α=40 30
220 28
26
200 Modified , Sd =4D, Sw =0
24
22
180
0 5 10 15 20
0 10 20 30 40 50 60 70
Scour width, Sw ( D) Scour-hole slope angle, α (dgree)

Fig. 14. Maximum bending moment versus scour width (Sd = 4D and a = 40°). Fig. 15. Effect of scour-hole slope angle of soft clay on the ultimate soil resistance
per length (z ¼ 0:5D).

4.3. Effect of scour width
Table 9 presents the effect of the scour-hole width on the prop-
erties of the remaining soft clay at a scour depth of 4D and a scour
hole slope angle of 40 . The results also indicate that the undrained
shear strength and effective unit weight of the soft clay decrease as
the scour width increases from zero to infinite, while the OCR
increases with increasing scour width.
Fig. 12 shows the calculated ultimate soil resistance per pile
length, pult at soil depth, z = 0.5D below the post-scour ground sur-
face at different scour width. When the scour width is increased
from zero to infinite, the ultimate soil resistance per length, pult,
decreases first (up to 15%) and then becomes constant at Sw = 15D.
Additionally, lateral pile-head deflections and maximum bend-
ing moments under various scour widths at lateral loads of 80 kN
and 100 kN are computed for the post-scour condition in Figs. 13
and 14, respectively. It is shown that as the scour width increases,
the pile-head deflection and maximum bending moment increase
slightly, with a maximum increase of approximately 8% and 3%
respectively. Moreover, the lateral pile-head deflection and maxi-
mum bending moment approach a constant value when the scour
width exceeds 15D. When the scour width exceeds 15D, the effect
of the scour width is found to be the same as that complete
removal of the soil layer as the scour depth is 4D, in other words,
the effects of the scour-hole dimensions on the behavior of later-
ally loaded piles can be neglected.

Table 10
Calculated soil properties considering effects of scour-hole dimensions and stress history (Sd = 4D, Sw = 0).

Soil depth from mud line Soil properties a = 0° a = 20° a = 40° a = 60°
before/after scour (m)
1.62/0.34 c0 sc (kN/m3) 9.84 9.89 9.94 9.98
(Cu)sc (kPa) 25.3 27.2 29.7 31.4
OCR 4.8 2.9 1.7 1.2
2.30/1.02 c0 sc (kN/m3) 9.92 9.96 9.98 10.00
(Cu)sc (kPa) 28.4 30.3 31.6 32.1
OCR 2.3 1.5 1.2 1.0
3.39/2.11 c0 sc (kN/m3) 9.95 9.98 10.00 10.00
(Cu)sc (kPa) 29.9 31.5 32.1 32.3
OCR 1.6 1.2 1.0 1.0
3.70/2.42 c0 sc (kN/m3) 9.96 9.99 10.00 10.00
(Cu)sc (kPa) 30.2 31.7 32.1 32.3
OCR 1.5 1.1 1.0 1.0
4.30/3.02 c0 sc (kN/m3) 9.96 9.99 10.00 10.00
(Cu)sc (kPa) 30.5 31.9 32.2 32.3
OCR 1.4 1.1 1.0 1.0
5.69/4.41 c0 sc (kN/m3) 9.97 10.00 10.00 10.00
(Cu)sc (kPa) 31.0 32.1 32.3 32.3
OCR 1.3 1.0 1.0 1.0
7.25/5.97 c0 sc (kN/m3) 9.98 10.00 10.00 10.00
(Cu)sc (kPa) 31.3 32.2 32.3 32.3
OCR 1.2 1.0 1.0 1.0
9.47/8.19 c0 sc (kN/m3) 9.98 10.00 10.00 10.00
(Cu)sc (kPa) 31.6 32.3 32.3 32.3
OCR 1.2 1.0 1.0 1.0
15.0/13.72 c0 sc (kN/m3) 9.99 10.00 10.00 10.00
(Cu)sc (kPa) 31.8 32.3 32.3 32.3
OCR 1.1 1.0 1.0 1.0
208 H. Zhang et al. / Computers and Geotechnics 84 (2017) 198–209

4.4. Effect of scour-hole slope angle 280

Maximum bending moment, Mmax (kN.m)

Table 10 shows the effect of the scour-hole slope angle on the
260
properties of the remaining soft clay when the scour depth is 4D
and the scour width is 0. It is found that with the increase of the
scour scour-hole slope angle, the undrained shear strength and 240
effective unit weight of the soft clay gradually increase, while the Modified, F t =80kN, Sd =4D, Sw=0
OCR decreases with increasing scour-hole slope angle.
Fig. 15 presents the variations of the ultimate soil resistance per 220 Modified, F t =100kN, Sd =4D, Sw=0
pile length, pult at a soil depth of 0.5D below the post-scour ground
surface at various scour-hole slope angles. An increase of the slope
200
angle increases the soil resistance. For example (Sd = 4D, Sw = 0),
the ultimate soil resistance per length increases approximately
24% when the scour-hole slope angle is increased from 0 to 60°. 180
The results presented in Fig. 15 also indicate that a larger scour- 0 10 20 30 40 50 60 70
hole slope angle means more overburden soil remaining above Scour-hole slope angle, α (degree)
the post-scour ground surface, and thus more soil resistance is Fig. 17. Maximum bending moment versus scour-hole slope angle(Sd = 4D and
induced. Sw = 0).
Figs. 16 and 17 also show the lateral pile-head deflections and
maximum bending moments computed by the simplified method
with various scour-hole slope angles at lateral loads of 80 kN and Lateral pile-head deflection, yt (mm)
100 kN, respectively. These figures indicate that an increase of 0 20 40 60 80 100 120
0
the scour-hole slope angle results in a decrease of the pile-head
deflection and maximum bending moment by approximately 9% Lin et al. (2014a), only considering
1
and 4% respectively, when Sd = 4D and Sw = 0. stress history
Lin et al. (2016), only considering
2
D)

scour-hole dimensions
4.5. Remarks on the scour effects on the response of laterally loaded
Present method, considering stress
piles
Scour depth, Sd (

3
history and scour-hole dimensions
Comparisons of the results in the above discussion indicate that 4
scour depth is the most important factor among the three scour Sw=0, α=40 , Ft=40kN
hole dimensions (scour depth, scour width, scour-hole slope angle) 5
influencing the responses of laterally loaded piles. Also, the change
of the undrained shear strength of the remaining soil after scour is 6
found to have more significant effect on the behavior of laterally
loaded piles, when compared with the other soil properties 7
considered.
8
Thus, to some extent, the scour effect on the lateral behavior of
piles may be simply characterized by the effects of scour depth and Fig. 18. Lateral pile-head deflection versus scour depths (Sw = 0, a = 40°, Ft = 40 kN).
soil stress history (e.g., the change of the undrained shear strength
of the remaining soil after scouring) for design purpose. This is fur-
ther demonstrated through comparisons with the results from Lin scour-hole dimensions and stress history of soils simultaneously)
et al. [12], Lin et al. [14], which consider only the effect of stress are larger than that from the solution of Lin et al. [14], but slightly
history of soils and only the scour-hole dimensions, respectively, smaller than in Lin et al. [12]. In other words, neglecting the stress
as shown in Fig. 18 for various scour depths. It can be noted that history effect tends to be unconservative for scour-affected pile
the pile head deflection obtained from our analysis (considering foundations in soft clay, whereas neglect of the scour-hole dimen-
sions and geometry would lead to over-conservative predictions/
design of the laterally loaded piles under scour condition.
160
5. Conclusions
Lateral pile-head deflection, yt (mm)

140 A simplified method using the modified p-y curves has been
developed in this paper to investigate the scour effect on the
responses of laterally loaded piles in soft clay considering the
Modified , F t =80kN, Sd =4D, Sw=0 stress history and scour-hole dimensions simultaneously. The fol-
120
lowing conclusions can be drawn from the present work:
Modified, F t =100kN, Sd =4D, Sw=0
1. Scour tends to slightly reduce the effective unit weight of the
100 remaining soil after scour, which is negligible when the stress
history is considered. In contrast, scour is found to be able to
significantly increases the over-consolidation ratio of the
80 remaining soft clay and reduce the undrained shear strength,
0 10 20 30 40 50 60 70
which contributes to the great difference between the behavior
Scour-hole slope angle, α (degree)
of laterally loaded piles including and ignoring the stress his-
Fig. 16. Lateral pile-head deflection versus scour-hole slope angle(Sd = 4D and tory effect. When the scour depth is increased from 5 to 10
Sw = 0). times of the pile diameter, consideration of the stress history
 H. Zhang et al. / Computers and Geotechnics 84 (2017) 198–209
effect is found to result in a 17–35% lower ultimate soil resis-
tance, and therefore increase the lateral pile-head deflection
by 12–16% and the pile maximum bending moment by 4–6%,
respectively.
2. Considering the scour-hole dimensions has a further significant
effect on the change of the undrained shear strength of the
remaining soils after scour, while its impact on the change of
the effective unit weight can also be neglected. Moreover, when
the total removal of the soil layer down to the scour depth is
considered, the lateral resistance of the soil to the pile is much
less than the case that considers the scour-hole geometry, and
thus results in larger pile-head deflection and maximum bend-
ing moment in the pile. In particular, it is found that the scour
depth plays the most important role among the different scour-
hole dimensions (scour depth, scour width, scour-hole slope
angle) in influencing the responses of laterally loaded piles.
3. As expected, the ultimate resistance of the remaining soils
decreases as the scour width increases from zero to infinity.
For the specific case studied with a scour depth of 4D and
scour-hole slope angle of 40°, the pile-head deflection and the
maximum bending moment in the pile are found to increase
slightly by approximately 8% and 3% respectively, and approach
a constant value when the scour width is over 15D.
4. Similarly, increase in the scour-hole slope angle from 0 to 60°
corresponds to an increase of the undrained shear strength of
the remaining soil and thus the lateral resistance of the soils
to the pile. The pile-head deflection and the maximum bending
moment in the pile are slightly decreased by approximately 9%
and 4% respectively, when the scour depth is 4D and the scour
width is 0.
5. To some extent, the scour effect on the lateral behavior of piles
may be simply characterized by the effects of scour depth and
soil stress history for design purpose.
Acknowledgements
This work was supported by National Key R&D Program of
China (Grant No. 2016YFC0800204), and the National Natural
Science Foundation of China (Grant No. 41672266). Financial sup-
port from these organizations is gratefully acknowledged.
References
[1] Annandale GW. Scour technology mechanics and engineering practice. New
York: McGraw-Hill; 2006.
[2] Azzouz AS, Krizek RJ, Corotis RB. Regression analysis of soil compressibility.
Soils Found 1976;16(2):19–29.
View publication stats
209
[3] Bennett CR, Lin C, Han J, Parsons RL. Bridge pile group under scour conditions.
In: Proceedings of SEI 2009 Structures Congress. Austin, Texas.
[4] Briaud JL, Ting FCK, Chen HC, Gudavalli R, Perugu S, Wei G. SRICOS: prediction
of scour rate in cohesive soils at bridge piers. J Geotech Geoenviron Eng
1999;125(4):237–46.
[5] Daniels J, Hughes D, Ramey GE, Hughes ML. Effects of bridge pile bent
geometry and levels of scour and P loads on bent pushover loads in extreme
flood/scour events. Practice Period Struct Des Construct 2007;12(2):122–34.
[6] Davis PJ, Rabinowitz P. Methods of numerical integration. 2nd ed. New
York: Academic Press; 1984.
[7] Kulhawy FH, Mayne PW. Manual on estimating soil properties for foundation
design, Rep. No. EL-6800. Palo Alto, CA: Electric Power Research Institute;
1990.
[8] Lagasse PF, Clopper PE, Zevenbergen LW, Girard LW. Countermeasures to
protect bridge piers from scour. NCHRP Rep. 593. Washington,
DC: Transportation Research Board; 2007.
[9] Li F, Han J, Lin C. Effect of scour on the behavior of laterally loaded single piles
in marine clay. Mar Georesour Geotechnol 2013;31(3):271–89.
[10] Liang FY, Zhang H, Huang MS. Extreme scour effects on the buckling of bridge
piles considering the stress history of soft clay. Nat Hazards 2015;77
(2):1143–59.
[11] Lin C, Bennett C, Han J, Parsons RL. Scour effects on the response of
laterally loaded piles considering stress history of sand. Comput Geotech
2010;37(7–8):1008–14.
[12] Lin C, Han J, Bennett C, Parsons RL. Technical note: behavior of laterally loaded
piles under scour conditions considering the stress history of undrained soft
clay. J Geotech Geoenviron Eng 2014;140(6):06014005.
[13] Lin C, Han J, Bennett C, Parsons RL. Analysis of laterally loaded piles in sand
considering scour hole dimensions. J Geotech Geoenviron Eng 2014;140
(6):04014024.
[14] Lin C, Han J, Bennett C, Parsons RL. Analysis of laterally loaded piles in soft clay
considering scour-hole dimensions. Ocean Eng 2016;111:461–70.
[15] Matlock H. Correlation for design of laterally loaded piles in soft clay. In:
Proceedings of the 2nd annual offshore technology conference.
Houston: American Institute of Mining, Metallurgical, and Petroleum
Engineers; 1970. p. 277–594.
[16] Mayne PW, Kulhawy FH. Ko-OCR relationship in soils. J Geotech Engrg Div
1982;108(6):851–72.
[17] Mindlin RD. Forces at a point in the interior of a semi-infinite solid. J Appl Phys
1936;7(5):195–202.
[18] Mohamed TA, Noor MJMM, Ghazali AH, Huat BK. Validation of some bridge
pier scour formulae using field and laboratory data. Am J Environ Sci 2005;1
(2):119–25.
[19] Muir Wood D. Soil behavior and critical state soil mechanics. Cambridge,
UK: Cambridge University Press; 1990.
[20] Reese LC, Van Impe WF. Single piles and pile groups under lateral
loading. Leiden, Netherlands: Balkema; 2001.
[21] Richardson EV, Davis SR. Evaluating scour at bridges. 4th ed. Washington,
DC: Federal Highway Administration; 2001. FHWA NHI 01–001 (HEC 18).
[22] Wardhana K, Hadipriono FC. Analysis of recent bridge failures in the United
States. J Perform Constr Facil 2003;17(3):144–50.
[23] Wu JJ, Li Y, Cheng QG, Wen H, Liang X. A simplified method for the
determination of vertically loaded pile-soil interface parameters in layered
soil based on FLAC3D. Front Struct Civ Eng 2016;10(1):103–11.