Engineering Science and Technology, an International Journal 23 (2020) 744–757

Contents lists available at ScienceDirect

Engineering Science and Technology,

an International Journal
journal homepage: www.elsevier.com/locate/jestch

Full Length Article

Experimental and numerical studies of laterally loaded piles located near

oil-contaminated sand slope
Reda A. Abdelhalim a,b,⇑, Mostafa El Sawwaf a, Ahmed M. Nasr a, Ahmed Farouk a
a
Structural Engineering Department, Faculty of Engineering, Tanta University, Tanta, Egypt
b
Department of Civil Engineering, Delta Higher Institute for Engineering and Technology, Mansoura, Egypt

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

Article history: Oil spillage occurrences are due to wars, accidents, drilling, storage, natural disasters, and rapid industri-
Received 3 April 2019 alization, which lead to huge amounts of waste discharge into adjacent areas. Soil contamination is also
Revised 11 January 2020 detrimental to associated structures and foundations. Therefore, in this research, experimental and
Accepted 4 March 2020
numerical studies on laterally loaded piles near an oil-contaminated sand slope were conducted to esti-
Available online 16 March 2020
mate the ultimate lateral capacity, lateral displacement, and maximum bending moment of a single pile
embedded in contaminated and uncontaminated sand slope. To simulate the contamination content as in
Keywords:
field conditions, preparation of oil-contaminated layers of sandy soil was performed by blending clean
Sand slope
Contaminated material
sand samples with heavy motor oil at proportions of 0–6% of the dried weight of soil. These studies were
Lateral loading performed by varying the contaminated sand layer thickness (LC), oil content (O.C), pile slenderness ratio
Model tests (LP/DP) and distance from the slope crest (X) to the pile head. The results demonstrated a considerable
Three-dimensional (3D) numerical analysis reduction in the friction angle (Ø) with increasing oil content. Therefore, the ultimate horizontal load
(Hu) of the pile also reduced. Moreover, the raise in the thickness of oil-contaminated sandy soil layer
(LC) has resulted in an increment in the lateral movement of the pile (y) and a reduction in the ultimate
horizontal load (Hu). When the proportion of the oil content increased to 2% over that in clean sand, the
maximum bending moment raised by about 11%. In addition, a good consistency between the numerical
analysis and experimental tests was observed.
Ó 2020 Karabuk University. Publishing services by Elsevier B.V. This is an open access article under the CC
BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction tance which the slope is free of influence was approximately 12DP
for a slope of 2 V to 3H, and 8 DP for a slope of 1 V to 2H. Otani et al.
Many tall constructions, communication towers, and irrigation [27] had performed a three-dimensional (3D) analysis using X-ray
structures like bridges are established adjacent to sand slopes to study the failure behavior of piles in sand due to lateral load.
and are founded by deep foundations especially piles. Lateral loads They demonstrated that the failure zone size reduces with increas-
are considered significant on these structures since they are trans- ing depth over the modeled pile and in a three dimensional view,
ferred through piles foundation. In the cases of installed piles in the this zone always takes a conical shape. Hajialilue-Bonab et al.
level ground, the analysis of lateral load has received an extensive [14] demonstrated that in the case of laterally loaded piles, a
interest over the past decades, [36,9,26]. Besides, several studies cone-shaped passive area is formed facing these modeled piles.
related to the ground slope effect on the pile lateral capacity have Chae [7] conducted a three-dimensional (3D) analysis on the
been investigated either through small scale tests in the laboratory, small-scale tests and prototype test of pile lateral load adjacent
[18] or by using software analysis [13,7]. Mezazigh and Levacher to sandy slope to validate the 3D elasto–plastic (finite element
[18] carried out laboratory tests to investigate the effect of sloping method) by comparisons with the experimental tests. Despite the
ground on the behavior of horizontal load of tube aluminum piles numerous studies on the capacity optimization of a lateral pile
driven in dry sandy soil. The results revealed that the effective dis- exposed to different loads through the improvements of length
and diameter of pile, a lot of investigations are yet required to be
carried out in order to modify the smoothness of pile’s material
⇑ Corresponding author.
via corrugations which can considerably improve the pile charac-
E-mail addresses: [email protected] (R.A. Abdelhalim), mosta-
[email protected] (M. El Sawwaf), [email protected] (A.M. Nasr),
teristics and increase the pile-soil adhesion [19,33].
[email protected] (A. Farouk). In addition to the above statement, a considerable amount of
Peer review under responsibility of Karabuk University. manufacturing waste flows out to surrounding areas due to fast

https://doi.org/10.1016/j.jestch.2020.03.001
2215-0986/Ó 2020 Karabuk University. Publishing services by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
 R.A. Abdelhalim et al. / Engineering Science and Technology, an International Journal 23 (2020) 744–757 745

Nomenclature

C cohesion LC thickness of contaminated sand layer

DP model pile diameter LP pile length
Dr relative density gLP dimensionless embedment length
D50 average grain size M bending moment of model pile
D10 effective grain size nh horizontal subgrade reaction modulus
Cu uniformity coefficient Rinter strength reduction factor of interface
Eref modulus of elasticity of soil y lateral displacement
EP modulus of elasticity of pile material m Poisson ratio
Hu (Clean) ultimate horizontal (lateral) load of single pile Ø friction angle of soil
embedded in clean sand slope Øi friction angle of interfaces
Hu (Oil) ultimate horizontal (lateral) load of single pile embed- W dilatancy angle
ded in contaminated sand slope y/DP normalized pile lateral displacement to diameter ratio
HR Horizontal (lateral) load of single pile LCR lateral capacity ratio
IP moment of inertia of pile cross-section Nc bearing capacity factor

industrialization, causing contamination in soil. Surface contami- In this investigation, the model piles dimensions were chosen
nation commonly presents in leaked pipelines, damaged petro- carefully to verify the geometric similarity. The tested model piles
leum wells and sub-surface storage tanks. Patel, [29] stated that were fabricated using roll-formed steel pipe with outer and inner
about 25% of crude oil production in the United States of America diameter of 21 mm and 17.4 mm, respectively. The lengths of the
was spilling and contaminating the soils surrounding a plant. Al- piles (LP) located inside the soil were 350 and 450 mm. The ratios
Sanad et al. [3] and Al-Duwaisan and Al-Naseem [1] reported that of pile length to pile diameter used were 16.6 and 21.4 to simulate
around 600 to 700 oil wells were damaged during the Gulf War rigid short piles and intermediate piles, respectively. For confirma-
(1991), causing a large amount of ground pollution. The variations tion purpose, the rigidity of the piles can be determined by calcu-
in the chemical and physical properties of oil-contaminated soil lating the dimensionless embedment length (gLP) [6]. The
are due to the difference in oil properties; such as kinematic vis- coefficient, g is calculated using Eq. (1) as follows:
cosity, pour point, and density; and mineral composition of soil rffiffiffiffiffiffiffiffiffi
nh
[16]. Nasr, [22] studied the influence of heavy oil-contaminated g¼ 5
ð1Þ
sandy soil on strip footing resting on slope ground using small- EP I P
scale tests and finite element analysis, the study aimed to calculate where IP = moment of inertia of the pile section; EP = Young’s mod-
the percentage of decline in the bearing capacity factor Nc due to ulus of the pile material; and nh is the modulus of horizontal sub-
soil contamination. grade reaction.
There are limited studies on the behavior of laterally loaded As reported by Terzaghi, [35], the extent of the nh coefficient
piles adjacent to oil-contaminated sand slope. Therefore, the cur- based on Eq. (1) is approximately 5500 kN/m3 for medium sand.
rent study aims to perform small-scale lab tests and numerical For practical purposes, piles exposed to lateral load may be consid-
analysis using software programming to determine the influence ered short and rigid if the dimensionless embedment length of the
of soil contamination on the performance of piles exposed to lat- pile is smaller than 2 (gLP  2), and supposedly long and flexible if
eral load and are located adjacent to a sandy slope. The major the dimensionless embedment length of the pile is higher than 4
parameters in this investigation involve various significant factors (gLP  4) [6,8]. Table 1 presents the relative flexibility of the test
affecting the contaminated soil including the contaminated sand model piles. A particular plate was prepared from a mild steel with
layer thickness (LC), the proportion of oil content (O.C) and dimen- a thickness of 2.0 mm for connecting the model piles to the loading
sions of pile (LP/DP). Furthermore, a behavioral comparison was cell.
made between the piles located near an uncontaminated (clean)
sandy slope and an oil-contaminated sandy slope. 2.2. Testing materials

2.2.1. Uncontaminated sand properties

2. Small-Scale model tests The sandy soil utilized in the experimental tests was collected
from domestic suppliers. Numerous small-scale experiments were
2.1. Experimental model tank and piles conducted on the sandy soil in accordance to the American Society
for Testing and Materials [4] to evaluate its geotechnical engineer-
Experimental model tests were performed in a mild steel tank, ing characteristics. The size distribution of particles was calculated
with inner dimensional length, width and height of 1000, 500 and using the dry sieve analysis and the outcomes are presented in
600 mm, respectively. These dimensions were selected to prevent Fig. 2. Based on the figure, the coefficient of uniformity,
the collapsing wedges on every side of the tested pile from stretch- Cu = 4.25, the average diameter, D50 = 0.6 mm and the effective
ing to the tank walls, which is as recommended by Prakash [31]. diameter, D10 = 0.16 mm. The physical characteristics of the sand
Mark was placed at every 50 mm distances on the interior faces used were a specific gravity of 2.653, maximum and minimum void
of the box tank to ensure more convenient and accurate setup ratios of 0.596 and 0.42, respectively and a proportion of fine sand
and configuration of the sandy soil layers inside the tank. The ver- of 12%. Also, in accordance to the Unified Soil Classification System
tical borders of the model box were supported by utilizing roll- (USCS), the sand used is categorized as poorly graded sand (SP).
formed steel angles at the top and the bottom of the borders.
Fig. 1 indicates a graphical view of the experimental apparatus 2.2.2. Properties of the contaminant
setup. To reduce friction between the tank walls and soil, the inner High viscosity engine oil that is obtainable from local stores was
tank walls were polished smooth. utilized to purposely pollute the sandy soil. This contaminant rep-
746 R.A. Abdelhalim et al. / Engineering Science and Technology, an International Journal 23 (2020) 744–757

Fig. 1. Graphical view of the laboratory apparatus setup (without scale).

Table 1
Relative flexibility of model test piles.

Soil type Dr (%) nh (kN/m3) gðm1 Þ LP (m) gLP Pile flexibility

Medium Sand 40 5500 5.53 0.35 1.93 Short pile

0.45 2.49 Intermediate pile

100

80
Finer Percentage (%)

60

40

20

0
0.01 0.1 1 10
Particle Diameter (mm)

Fig. 2. Particle-size distribution curve of the sandy soil used.

 R.A. Abdelhalim et al. / Engineering Science and Technology, an International Journal 23 (2020) 744–757 747

resents unused oil, and utilized to simulate the conditions Table 2

observed at contaminated locations [12,3]. The contaminant used Internal friction angle of sand at various proportions of oil content.

in this study is similar to that used by Nasr, [26], which has these Relative density Proportion of oil Internal friction Reduction
properties: (Dr) (%) content (%) angle (£Þ (%)
Specific gravity of 0.88, kinematic viscosity of 169.8  106 m2/ 40 0.0 37.0 –
s, lower pour point of 10 °C and density of 8.79 kN/m3. All exper- 2.0 34.4 7.00
iments were conducted at a laboratory temperature of about 4.0 32.5 12.16
6.0 31.3 15.4
27 ± 1.0 °C.

2.3. Preparation and characteristics of the contaminated sand Nasr and Krishna Rao, [21] had performed several triaxial com-
pression experiments to calculate the angle of friction (Ø). The
In this investigation, contaminated sand samples were prepared results revealed that the contaminated sandy soil shows no clear
by blending clean sand with unused engine oil at various propor- high strength value at all confining levels, contrasting from the
tions of oil content (O.C) of 2, 4 and 6% (according to dry sand clean sandy soil. This can be attributed by the covering of heavy
weight) to verify the similarity of oil percentages found at contam- oil on the sandy soil grains, which causes a decrease in the friction
inated sites. Due to the high viscosity of the engine oil, the samples strength when the sandy soil grains skid above one another. When
of oil-contaminated sand were placed into locked boxes for 72 h to the heavy motor oil and clean sand are mixed, the lubricating
attain balance for this mixing, and to permit any probable reac- impacts of the contaminant at sand grains surfaces have decreased
tions among uncontaminated sandy soil and the contaminant. the shear strength of this mixture.
Due to the relationship between resistance of sandy soil and the
angle of internal friction, many shear box experiments on the con- 2.4. Experimental procedure and test program
taminated and uncontaminated sandy soil were carried out at nor-
mal stress values of 14.3, 24.1, and 43.7 kN/m2 to investigate the In the study of soil–pile interaction due to horizontal load,
impact of the proportion of oil content (O.C) on the shear strength sandy soil preparation around the model pile is a significant factor
of the sand. The results in terms of shear stress versus horizontal which has a considerable action on the results [14]. Prior to sand
displacement for clean and oil-contaminated soil are given in slope preparation, the model pile was placed in a certain location
Fig. 3. In addition, it should be noted that the uncontaminated to simulate a non-displacement pile. During installation, this tech-
and contaminated specimens exhibited low cohesion values which nique does not reflect any displacement in the soil around the pile
could be due to the surface tension force of existing water in the to keep the target relative density of sand at limited value during
soil and the viscosity of the oil. Nevertheless, the friction angle of the tests. It can be deemed figuratively as auger or bored piles,
all specimens decreases as oil content increases. The values of which cause low displacement in the soil around the pile during
cohesion were 3.7, 5.6, 6.2, and 6.3 kPa at oil contents of 0, 2, 4, installation. This installation method is similar to that applied by
and 6% respectively. These findings are consistent with those Nasr [23]. Next, a model of sand slope was prepared in layers of
investigated by Khamehchiyan et al. [15], where low values of 50 mm thick. Based on the relative density of the sand sample,
cohesion for (SP) soil were also observed. Laboratory experiments the unit weight of the sand was determined and managed by con-
were performed for relative density, Dr of 40%. Table 2 demon- sidering the pre-estimated weight of the soil sample inside the
strates the different values of friction angle (Ø) for both the uncon- tank at each sandy layer (50 mm) individually. The unit weight
taminated and contaminated sandy soil. of the sandy samples was obtained through experimental tests
by gathering several samples in small cans (40  40  40 mm) with
fixed volume (64000 mm3). The average value of sample relative
50
OC = 0% density (Dr) utilized in this investigation was 40%, which gives a
O.C = 2% unit weight of 17.00 kN/m3 ± 0.3. Then, an applied load of 5 N
Shear stress (kN/m2)

40
OC = 4% was placed as the lateral load, which was affected by a high-
30 OC = 6% tension steel wire of 2 mm diameter hooked to the pile head
through an eye bolt. The other hand of the steel wire moves
20 through a polished flat roller with a diameter of 100 mm and con-
nected to the frame of the model tank. Finally, the horizontal
10 deflections were recorded using two high sensitivity dial gauges
accurate to 0.01 mm. A small-scale testing scheme was prepared
0
0 2 4 6 8 10 to investigate the behaviors of the short and intermediate laterally
Horizontal displacement (mm) loaded piles located near a heavy oil-contaminated sandy slope.
50 Table 3 demonstrates an outline of the parameters studied in the
laboratory model tests and their values. Furthermore, the sandy
Shear stress (kN/m2)

40 soil geometry, model pile geometry, and oil-contaminated sand

are presented in Fig. 4.
30

20 O.C = 0% 3. Numerical analysis

O.C = 2%
10 O.C = 4% 3.1. Finite-element modeling
O.C = 6%
0 In this study, the problem was examined numerically using a
0 2 4 6 8 10
software analysis of Plaxis 3D Foundation [30]. For the accuracy
Horizontal displacement (mm)
of the results, finite element (FE) analysis was performed to inves-
Fig. 3. Shear stress versus horizontal displacement curves at various oil contents: tigate the maximum bending moment and lateral capacity of the
(a) Normal stress = 24.1 kN/m2; (b) Normal stress = 43.7 kN/m2. pile group embedded in contaminated sand without slope [21].
748 R.A. Abdelhalim et al. / Engineering Science and Technology, an International Journal 23 (2020) 744–757

Table 3
Parameters of laboratory model tests.

Series Sand Constant parameters Variable parameters

I Clean Short pile (LP =Dp ¼ 16:6Þ; ðLP =Lv ¼ 1:75Þ X=DP ¼ ð0:0; 2:0; 4:0; 6:0; 8:0Þ
Intermediate pile (LP =Dp ¼ 21:4Þ; ðLP =Lv ¼ 2:25Þ X=DP ¼ ð0:0; 2:0; 4:0; 6:0; 8:0Þ
 
II Contaminated Lc =Lp ¼ 0:1 ;(LP =Dp ¼ 16:6Þ,ðLP =Lv ¼ 1:75Þ X=DP ¼ ð0:0; 2:0; 4:0; 6:0Þ(Oil content = 2%, 4%, 6%)
 
Lc =Lp ¼ 0:1 ;(LP =Dp ¼ 21:4Þ,ðLP =Lv ¼ 2:25Þ X=DP ¼ ð0:0; 2:0; 4:0; 6:0Þ(Oil content = 2%, 4%, 6%)
III (Oil content = 4%), (LP =Dp ¼ 16:6ÞðLP =Lv ¼ 1:75Þ X=DP ¼ ð0:0; 2:0; 4:0; 6:0Þ Lc =Lp ¼ ð0:1; 0:2; 0:3; 0:5Þ

Fig. 4. Geometric parameters investigated in the experimental test.

Comparisons were made between the model test and the FE results the numerical analysis of sand, friction angle was calculated accu-
to verify their accuracy and therefore new variables can be studied rately using a direct shear test as mentioned previously, and
solely from the finite element tests only. The results showed that Young’s modulus was chosen according to Bowles [5]. Table 4
the FE results have an appropriate consistency with the experi- shows a summary of the properties of the materials used for the
mental test results. Also, Nasr [25] had carried out numerical and MC- Soil model and steel pile.
experimental tests to examine the strip footing behavior on con-
taminated sand slope. The results showed a close agreement 4. Results and discussions
between both results and also proved that FE analysis by Plaxis
software can be used to supplement the experimental results. To provide significant discussions for the various parameters,
Therefore, in the current study, finite element analysis was per- the ultimate horizontal resistance of the pile (Hu) must be deter-
formed to supplement the experimental results and to evaluate mined beforehand, which was estimated using the tangent inter-
the bending moment along the pile depth. The pile dimensions section technique as stated by Rahman et al. [32] and Al-Mhaidib
simulated in the numerical analysis are similar to the dimensions [2]. In this technique, tangent lines are sketched from the begin-
used in the experimental model. The meshes shown in Fig. 5(a), ning and ending points of the horizontal load (HR)–lateral displace-
5(b) and 5(c) were selected according to their sensitivity and the ment (y) curve, and the point of crossing of these tangents is
computational period suitable for the model analysis. Fig. 5(d) pre- extended to the horizontal axis (x-axis) to obtain the ultimate hor-
sents the 3D view of the single short model pile utilized in all the izontal resistance of the pile (Hu). Fig. 6 shows an instance of such
numerical investigations. Both the finite element numerical and tangent intersection method. To compare the test data, the expres-
experimental models have the same boundary conditions suitable sion lateral capacity ratio (LCR) is utilized and described as
for making accurate comparisons. Thus, at the lower end of the
three-dimensional finite element mesh, complete fixities that HuðoilÞ
LCR ¼ ð2Þ
restricted the motions in both vertical and horizontal dimensions HuðcleanÞ
were utilized. For the left and right directions, roller supports were where HuðoilÞ and HuðcleanÞ are the ultimate horizontal loads for single
utilized to restrict the horizontal movements while the vertical
pile located near oil-contaminated sand slope and uncontaminated
movement was released without fixation.
(clean) sand slope, respectively.

3.1.1. Modeling parameters of the analysis 4.1. Behavior of laterally loaded single piles in clean sand slope
For the case of lateral load of pile under static load, the Mohr-
Coulomb (MC) model is considered as a suitable model that can To acquire a better comprehension on the behavior of a single
be used to simulate the soil behavior in general. Furthermore, pile in heavy oil-contaminated sand slope, a series of small-scale
many researchers have utilized the Mohr-Coulomb (MC) model lateral load tests was conducted on a single pile in clean sand slope
to simulate the drained attitude of granular soils. The model (Series I). In these small-scale tests, the relative density (Dr), verti-
includes five essential parameters that have been used in most pre- cal length of slope (LV), and the sand slope were kept constant at
vious researches. These parameters are internal friction angle (Ø), 40%, 200 mm and 2H: 1 V, respectively. Fig. 7 shows plots of the
cohesion (c), dilatancy angle (w), Young’s modulus (Eref) and Pois- lateral load–lateral displacement curves determined from experi-
son’s ratio (m). The dilatancy angle w was set as w = Ø – 30, accord- mental tests for both short and intermediate piles embedded in
ing to the Plaxis manual [30]. Due to the fact that the Young’s clean sand slope. Fig. 8 on the other hand shows plots of lateral
modulus and friction angle are the most effective parameters in load–lateral displacement curves from numerical analysis for short
 R.A. Abdelhalim et al. / Engineering Science and Technology, an International Journal 23 (2020) 744–757 749

(a) (b)

(c) (d)
Fig. 5. The modeling of finite element used to simulate a lateral load of single pile:(a) 3D mesh view; (b) soil components; (c) plan view of mesh; and (d) 3D view of model
pile.

Table 4 the intermediate pile was less than the lateral displacement in
Properties of materials utilized in the numerical analysis. the short pile, with a decrease of about 19–45% at the same applied
Parameters Uncontaminated Contaminated Pile lateral load.
Sand Sand
Material model Mohr–Coulumb Mohr–Coulumb Linear
elastic 4.2. The effect of pile location comparative to slope crest (X)
Type of material Drained Drained Non-
behavior porous The ratio X/DP expresses the nearness of a model pile to the
Young’s modulus, Eref: 15,000 (12500, 11000, 21  107
kN/m2 10000) *
slope crest. To investigate the influence of this ratio on the horizon-
Poisson ratio, m 0.3 0.3 0.2 tal load and horizontal displacement of the piles, laboratory and
Friction angle, Ø: degree 37 (34.4, 32.5, 31.3) * ___ finite element tests were conducted on both the contaminated
Dilatancy angle, w: 7 (4.4, 2.5, 1.3) * ___ and uncontaminated sand slopes with all the variables kept at con-
degree
stant and X/DP ratios of 0, 2, 4, 6, and 8. Fig. 9 presents the variation
Interface Reduction 0.75 0.7 ___
Factor, Rinter of the ultimate horizontal load of the pile versus the ratio X/DP for
both short and intermediate piles. As the model pile situates fur-
*At 2, 4 and 6% oil contents, respectively.
ther away from the slope crest, its ultimate lateral capacity
enhances. Fig. 10 provides plots of the measured bending moment
pile embedded in clean sand slope. Table 5 shows the values of against depth for short pile located in clean sand slope at constant
ultimate horizontal load (Hu) obtained from the tangent intersec- horizontal load, HR of 80 N. From Fig. 10, the computed bending
tion method for short and intermediate piles embedded in clean moment along the pile decreases as the pile is located further away
sand. The results demonstrated that the lateral displacement in from the slope crest, and reaches maximum at a depth of 0.135 m
750 R.A. Abdelhalim et al. / Engineering Science and Technology, an International Journal 23 (2020) 744–757

consideration three various oil contents, as presented in Series II.

In this series, the oil-contaminated sand layer thickness was kept
constant at LC/LP = 0.1. Fig. 11(a) and 11(b) show the lateral
load–lateral displacement curves at various oil contents. Based
on these curves, the values of ultimate horizontal load (Hu) are
determined and given in Table 6. Meanwhile, the bending moment
and lateral capacity ratio (LCR) were calculated and presented in
Fig. 12 and Fig. 13 respectively. Fig. 12 provides a plot of measured
bending moment versus depth for short pile located at both con-
taminated and uncontaminated sand slope at a constant lateral
load (HR = 80 N). For the same load level, the bending moment
increases with increasing percentage of contamination. The finite
element results indicated that the maximum increase in bending
moment was observed at the first percentage of contamination at
O.C = 2–4% for piles at any distance from slope crest (X). This is
due to the lowering of the angle of internal friction, which mini-
mizes the shear strength of the sandy soil and increases the bend-
ing moment on the pile. The maximum bending moment values of
the contaminated soil at 2%, 4%, and 6% oil contents are relatively
close to each other but however are relatively far from that at O.
C = 0% (uncontaminated soil). This is possibly due to that the con-
tact between the pile and the surrounding soil was modeled in the
finite element analysis using interface elements created along the
pile’s circumference, and according to Coulomb’s shear stress crite-
rion, the strength properties of interfaces (Øi) are obtained from
both the strength properties of each layer (Øsoil) and reduction fac-
tor of interfaces (Rinter) by applying the following equation.

tan Øi ¼ Rinter tan Øsoil ð3Þ

Fig. 6. Example of a tangent intersection technique.

where Øi is the friction angle of the interface. Thus, in case of a low

from the ground surface, at a ratio of 0.386 pile length (Mmax at normal stress, in addition to the constant value of Rinter for all con-
0.386 LP). taminated soils in the finite element analysis, the shear strength
characteristics for contaminated soil will be relatively close to each
4.3. The effect of the proportion of oil content (O.C) other. Therefore, it is preferable to calculate the value of Rinter
instead of assuming it. However, in this article, the value of Rinter
To examine the effect of the proportion of oil content, tests were was set at constant (0.7) for all contaminated layers, in accordance
performed on both the short and intermediate pile by taking into to Plaxis manual’s recommendations as it is out of the scope of this
Lateral Load, HR (N)
Lateral load , HR (N)
30 60 90 120 150 180 210 240
30 60 90 120 150 180 210 240
0
0 Intermediate Pile
Short Pile 2 LP/DP = 21.4
2
LP/DP = 16.6 Slope 2H: 1V
Slope 2H: 1V 4
4 Dr = 40%
Dr = 40%
Lateral displacement, y (mm)

6
6
Lateral displacement, y (mm)

8
8
10
10
12
12
14
14
16
16
18
18
20
20 at crest
22 at Crest
22 X = 2Dp
X = 2Dp
X = 4Dp 24
24 X = 4Dp
X = 6Dp 26
26 X = 6Dp
X = 8Dp 28 X = 8Dp
28
30
30

Fig. 7. Variation of horizontal load (HR) with horizontal displacement (y) for piles located at different distances from the slope crest, based on the experimental results.
 R.A. Abdelhalim et al. / Engineering Science and Technology, an International Journal 23 (2020) 744–757 751

Lateral Load, HR ( N ) 220

Slope 2H:1V

Ultimate Lateral Load, Hu (N)

Dr = 40%
30 60 90 120 150 180 210 240 270 Experimental results
180
0

2 Short Pile
LP/DP = 16.6 140
4 Slope 2H: 1V
Lateral displacement, y (mm)

6 Dr = 40%
100
8 Short Pile (Lp/Dp=16.6)
Intermediate Pile (Lp/Dp=21.4)
10
60
12 0 2 4 6 8 10
X/DP
14
16 Fig. 9. Variation of ultimate lateral load (Hu) versus the ratio X/DP for both short
and intermediate piles embedded in clean sand.
18

20
at Crest Bending Moment(10-3kN.m)
22 -0.15
at X = 2Dp
24 -0.1
at X = 4Dp
26 -0.05
at X = 6Dp
-150 -130 -110 -90 -70 -50 -30 -10 10 30 50
28
at X = 8Dp 0
30
0.05
Depth (m)

Fig. 8. Variation of horizontal load (HR) with horizontal displacement (y) for piles
0.1
located at different distance from slope crest, from the numerical analysis.
0.15

0.2
study to investigate the influence of surface roughness on the pile
lateral behavior. 0.25
Fig. 13 shows that the increment in the oil content percentage 0.3
at X = 2 Dp
significantly minimizes the lateral capacity ratio (LCR) until oil
at X = 4 Dp 0.35
content percentage approaches 4%; after this percentage, the low-
at X = 6 Dp
ering in the LCR is lightly decreasing. This is due to the increment
0.4
in the proportion of oil content, which raises the lubricating impact
among soil particles, and therefore decreases the shear strength of Fig. 10. Comparison of bending moment with depth for short pile (F.E results).
the sandy soil. Thus, the oil-contaminated sand with an extreme
proportion of oil content will be easily sheared while a shearing
load is being applied. However, in the results of small-scale and C) > 4% the reduction rate of internal friction angle (Ø) was
numerical tests, an extreme decrease in the internal friction angle minimal.
was observed at the first percentage of contamination. Based on
the results presented in Table 7, the percent of decreasing in the 4.4. The influence of thickness of contaminated sandy soil layer(LC)
ultimate lateral load for oil-contaminated sand in comparison with
clean sand are given in Tables 8, and 9. To examine the influence of this parameter, several tests were
The maximum bending moment values and the percent of carried out on a single short rigid pile with various parameters.
increasing due to contamination of the soil in comparison with In these experiments, the proportions of oil content (O.C) were
clean sand are given in Tables 10 and 11, respectively. retained at a proportion of 4%. The horizontal load (HR) – horizon-
Fig. 14 shows the variation of single piles resistance at various tal displacement (y) curves for various thickness of the contami-
proportions of oil content. The results showed that the rise in the nated layer is presented in Fig. 15.
oil content percentage significantly decreases the ultimate hori- The results indicated that the finite element analysis has the
zontal loads of piles. These conclusions are consistent with the same trend as the experimental analysis with increasing thick-
results of Shin and Das [34] which stated that the maximum inter- nesses of the contaminated sand layer. As observed from the
nal friction angle value (Ø) was extremely declined with increasing results presented in Fig. 15, the values of ultimate horizontal load
the oil content (O.C) from zero to 2%. Nonetheless, at oil content (O. (Hu) were determined and presented in Table 7. Furthermore,

Table 5
Test results for piles embedded in clean sand slope at different locations from slope.

Short pile (Experimental results) Intermediate pile (Experimental results)

X 0 2 4 6 8 0 2 4 6 8
Dp
Hu (N) 110 132 151 171 177 123 143 165 179 185
y
Dp
0.15 0.16 0.175 0.175 0.165 0.095 0.11 0.125 0.115 0.11
752 R.A. Abdelhalim et al. / Engineering Science and Technology, an International Journal 23 (2020) 744–757

Lateral Load, HR ( N ) Lateral Load, HR ( N ) Lateral Load, HR ( N )

30 60 90 120 150 180 210 240 270 30 60 90 120 150 180 210 240 270 30 60 90 120 150 180 210 240 270
0 0 0
Clean Sand Clean Sand Clean Sand
2 2 2
Oil Content = 2% Oil Content = 2% Oil Content=2%

Lateral displacement, y (mm)

4 4 Oil Content = 4% 4 Oil Content=4%
Oil Content = 4%
Lateral displacement, y (mm)

Lateral displacement, y (mm)

Oil Content = 6% 6 Oil Content=6%
6 Oil Content = 6% 6
8 8 8

10 10 10

12 12 12

14 14 14

16 16 16

18 18 18

20 20 20

22 22 22

24 24 24

26 Short Pile 26 Intermiate Pile 26 Short Pile

28 28 28

30 30 30

Fig. 11. Variation of horizontal load (HR) with lateral displacement (y) for piles located at different oil contents.

Table 6
Test results for short and intermediate piles embedded in heavy oil-contaminated sand with variable of percentage of contamination (O.C) at constant LC / LP (0.1).

O.C 2% 4% 6%
Pile type X 0 2 4 6 0 2 4 6 0 2 4 6
DP
Short pile Hu (N) 102 114 121 127 95 105 115 122 91 101 112 118
y
DP
0.16 0.15 0.165 0.16 0.165 0.175 0.187 0.18 0.17 0.185 0.20 0.20
Intermediate pile Hu (N) 113 127 144 154 110 124 138 148 108 121 134 143
y
DP
0.14 0.14 0.125 0.14 0.156 0.155 0.14 0.18 0.185 0.175 0.15 0.195

1.2
Bending Moment (e-3 kN.m) 1.2
Intermediate pile
Short pile
-0.15
Short Pile 1 1
Lp/Dp=16.6 -0.1
LCR = Hu(oil) / Hu(clean)

at X= 2Dp
LCR = Hu(oil) / Hu(clean)

-0.05 0.8 0.8

-150 -130 -110 -90 -70 -50 -30 -10 10 30 50

0
Depth (m)

0.6 0.6
0.05
at Crest at Crest

at X=2Dp at X=2Dp
0.1 0.4 0.4 at X=4Dp
at X=4Dp
at X=6Dp
at X=6Dp
0.15
0.2 0.2
0.2 0 2 4 6 8 0 2 4 6 8
Oil Content,% Oil Content,%
0.25 Fig. 13. Variation of lateral capacity ratio (LCR) with different oil contents for short
pile and intermediate pile (Experimental results).
Clean 0.3
variations of the ultimate lateral load (Hu) and the lateral capacity
O.C = 2%
ratio (LCR) versus LC/LP are shown in Fig. 16 and Fig. 17, respec-
O.C = 4% 0.35
tively. The results obviously showed that the lateral capacity ratio
O.C = 6% (LCR) significantly declines with an increment in the thickness of
0.4
the contaminated sand layer (LC). This is due to the passive cone-
Fig. 12. Comparison of bending moment with depth for short pile (F.E results). shaped zone, which supports the lateral capacity around the model
piles intersecting with this contaminated sand layer. The percent of
 R.A. Abdelhalim et al. / Engineering Science and Technology, an International Journal 23 (2020) 744–757 753

Table 7
Test results for short piles embedded in heavy oil-contaminated sand slope with variable of thickness of contaminated layer (LC) at constant O.C (4%).

LC 0.1 0.2 0.3 0.5

Lp

X 0 2 4 6 0 2 4 6 0 2 4 6 0 2 4 6
DP
Hu (N) 95 105 115 122 92 97 105 108 89 92 95 100 87 89 93 95
y
DP
0.16 0.17 0.18 0.18 0.185 0.18 0.19 0.2 0.2 0.15 0.2 0.225 0.2 0.18 0.215 0.23

Table 8
The percent of decreasing in the ultimate lateral load due to soil contamination in Lateral Load, HR (N)
short pile.
30 50 70 90 110 130 150 170 190
Short Pile O.C = 2% O.C = 4% O.C = 6% 0
at crest 7.3 13.6 17.3 Short pile
2
at X = 2 DP 13.6 19.7 23.5 LP/DP = 16.6
4 O.C = 4%
at X = 4 DP 19.8 23.8 25.8

Lateral displacement, y (mm)

Dr = 40%
at X = 6 DP 25.7 28.6 31.0
6 Slope 2H:1V

Table 9 8
The percent of decreasing in the ultimate horizontal load due to soil contamination in
intermediate pile.
10

Intermediate Pile O.C = 2% O.C = 4% O.C = 6%

12

at crest 8.1 10.6 12.2 14

at X = 2 DP 11.2 13.3 15.4
16
at X = 4 DP 12.7 16.3 18.8
at X = 6 DP 13.9 17.3 20.1 18

20 Clean Sand(FEM) Clean Sand(EXP)

Table 10
The values of maximum bending moment, (e-3 kN.m) due to soil contamination for 22 Lc/Lp = 0.1(FEM) Lc/Lp = 0.1(EXP)
short pile at different distance from slope crest (x). Lc/Lp = 0.2(FEM) Lc/Lp = 0.2(EXP)
24
Short Pile O.C = 0% O.C = 2% O.C = 4% O.C = 6% Lc/Lp = 0.3(FEM) Lc/Lp = 0.3(EXP)
26
at X = 2DP 117.6 130.29 131.73 136.51 Lc/Lp = 0.5(FEM) Lc/Lp = 0.5(EXP)
at X = 4 DP 113.7 125.65 126.98 128.12
at X = 6 DP 106.7 117.69 118.02 119.09
Fig. 15. Variation of horizontal load (HR) with lateral displacement (y) for short
piles located at X = 2DP with different LC/LP ratios.
Table 11
The percent of increasing in maximum bending moment due to soil contamination for decreasing in ultimate horizontal load due to varying the thickness
short pile. of contaminated layer (LC) is given in Table 12.
Short Pile O.C = 2% O.C = 4% O.C = 6% The results indicated that the maximum decreasing value in the
ultimate horizontal load of pile (Hu) was observed at LC/LP = (0.1–
at X = 2DP 10.79 12.02 16.08
at X = 4 DP 10.51 11.68 12.68 0.3), LC/DP = 1.75–5.25, and the rate of reduction in Hu becomes
at X = 6 DP 10.29 10.61 11.61 smaller when the thickness of the contaminated sand layer (LC)

Intermediate
160 ( 143 N)
Short
(132 N) Intermediate Intermediate Intermediate
140 Short ( 127 N) ( 124 N) ( 121 N)
Ultimate Lateral Load, Hu ( N )

Short
(115 N) (114 N) Short Short
120 Short (105N) (101 N)
(95 N) Short
Short
100 (90 N)
(86 N)

80

60

40

20

0
an M

an XP

(O 2% EM

), P
P

(O 4% EM

(O 6% EM

), P
P
X

2% X
EX

4% X
EX

6% X
EX
C , FE

C ,E
,E

= ), E

= ), E

= ), E
F

F
),
.C ),

.C ),

.C ),
an

(O 2%

(O 4%

(O 6%
le

le
le
C

=
=

=
.C

.C

.C
.C

.C

.C
(O

(O

(O

Fig. 14. Variation of the ultimate horizontal load (Hu) with different oil contents (O.Cs) at X = 2Dp.
754 R.A. Abdelhalim et al. / Engineering Science and Technology, an International Journal 23 (2020) 744–757

200 analyze the slope movement. Fig. 18 shows the soil cracks around
the pile which were observed during the lateral loading tests of
Ultimate Lateral Load, Hu (N)

short pile embedded in clean sand slope at different distances from

160 Oil Content = 4% the slope crest (X). Large gaps were observed behind both piles as
well as ground heaving in front of piles during the tests, which are
similar to the investigations made by Muthukkumaran [20]. The
120
deformed cracks formation during the tests indicated that the soil
movement along the line is slightly less than 45°measured from
80 the pile center in the direction perpendicular to the load. For the
at crest short and intermediate piles which are located at 8DP from the
at X= 2Dp slope crest, no visible cracks were observed on the slope through-
40 out the loading tests. This proves that the passive resistance
at X= 4Dp
extends in the direction of loading to distance (X) shorter than
at X= 6Dp 8DP in front of the pile.
0
0 0.1 0.2 0.3 0.4 0.5 0.6 For the laterally loaded piles in contaminated sand slope, there
are a few differences in the crack width due to the oil contamina-
L C / LP tion content in the soil. This is because contamination of soil with
high viscosity oil drastically reduces the friction among soil parti-
Fig. 16. Variation of thickness of contamination layer (LC/LP) with ultimate lateral
load (Hu). cles and encourages the sliding of particles under the lateral load
effect. Besides, the lateral load-lateral displacement curves (failure
behavior) shown in Fig. 11 are consistent with this observation.
Furthermore, due to the cracks resulting from soil contamination,
the ultimate horizontal load of the pile (Hu) declined as shown in
Tables 8 and 9. Despite that, cracks generally have the same trend
of failure. Fig. 19 displays the soil cracks observed during the lat-
eral loading tests for short pile embedded in oil-contaminated sand
slope at different distances from the slope crest (X).

4.6. Scale effects and limitations

Due to the high cost and time consuming factors in conducting

full-scale tests on full-dimensions piles, experimental tests using
small-scale models are more preferred as the commonly used
method. However, there are numerous significant factors that
causing the use of small-scale laboratory tests less effective. These
include the gradation of the sand particle relative to pile dimension
(which is known as the ‘particle size effect’), friction between tank
walls and model of the pile, methods of construction and boundary
Fig. 17. Variation of thickness of contamination layer (LC/LP) with lateral capacity
conditions. In this research, the ratio of model pile diameter to the
ratio (LCR).
average diameter of soil particles (DP/D50) was 35, which exceeds
the minimum value of 30. This indicates insignificant scale influ-
Table 12 ence on the lateral load for the pile foundation [28]. Besides, in
The percent of decreasing in ultimate lateral load (Hu) due to varying the thickness of order to avoid friction between the walls of tank and soil particles,
contaminated layer (LC). the internal side of the tank walls were polished to minimize any
LC = 0.1 LP LC = 0.2 LP LC = 0.3 LP LC = 0.5 LP friction effects between these walls and sand particles as much
LC = 1.75 DP LC = 3.5 DP LC = 5.25 DP LC = 10.5 DP as possible. Prakash [31] stated that the dimensions of a soil tank
X/DP = 0 14% 17% 19% 21% can be extended up to (8–12) DP and (3–4) DP, where DP = (pile
X/DP = 2 20% 27% 31% 33% diameter), in the direction and perpendicularly to the lateral load-
X/DP = 4 24% 31% 37% 39% ing, respectively. In addition, to decrease the effect of tank bound-
X/DP = 6 28% 36% 42% 44% aries, the depth of sand was kept beneath the pile bearing of
minimum of 6DP. Furthermore, the sides of the model experimen-
is approaching to 0.5LP (LC/DP = 10.5). This is due to the extension tal box were strengthened by installing steel angles to increase its
of the collapse region (the passive cone-shaped zone), which has rigidity and restrain any side movements of the tank walls. Finally,
the largest portion at the ground surface and the extension of this this investigation provided a useful fundamental idea for further
area declines with the increment in the soil depth along the pile researches involving full-scale tests and centrifugal tests which
length [27]. In addition, these findings are consistent with the will focus more on analyzing the actual behavior of lateral loads
results by Duncan et al. [10], whom revealed that the effective of piles located near an oil-contaminated sand slope.
depth of laterally loaded pile is at a distance of eight times the
diameter of the pile (8DP) measured from the ground surface.
5. Conclusions
4.5. Failure behavior of single pile located near the uncontaminated
and Oil-contaminated sand slope The research about the ultimate lateral load of single piles
located near oil-contaminated soil especially sand slopes is very
The results from the small-scale tests were analyzed using dig- limited. So, the ultimate lateral load of short and intermediate piles
ital images. The digital images taken during testing were used to located near heavy oil-contaminated sand slope was discussed
 R.A. Abdelhalim et al. / Engineering Science and Technology, an International Journal 23 (2020) 744–757 755

Cracks from side view Cracks from elevation

(a) Images at = 4DP

Cracks from side view Cracks from elevation

(a) Images at X = 8DP

Fig. 18. Images during lateral loading tests for short pile embedded in clean sand slope at different locations from slope crest (X).

using experimental and numerical analysis. The conclusions that 5) The ultimate horizontal load (Hu) decreases due to the incre-
can be drawn on the basis of this current study are: ment in the thickness of the contaminated sand layer (LC).
The peak decrease in the Hu and LCR was observed at LC/
1) For both uncontaminated (clean) and heavy oil- LP = 0.1–0.3, LC/ DP = (1.75–5.25), and the rate of depletion
contaminated sand slope; as the model pile is located away becomes relatively slight when the thickness of the contam-
from the slope crest, the ultimate lateral capacity of the pile inated sand layer (LC) is higher than 0.3 LP.
enhances until a value of X/DP = 6, after which the influence
of sand slope on the pile lateral capacity is negligible. 6. Recommendations from the current study
2) The increment in oil content percentage significantly mini-
mizes the lateral capacity ratio (LCR), until the oil content The recommendations that can be drawn from this study are
percentage approaches 4%; where beyond this percentage, established based on cases of research (e.g. slope gradient, pile
the lowering in the LCR is lightly decreasing. characteristics, and loading type). For the construction of piles in
3) For both uncontaminated (clean) and oil-contaminated sand a contaminated sandy slope, they should be installed at a distance
slope; the computed bending moment for the model pile (x) of not less than 6DP to mitigate the slope’s influence. Since the
decreases when the model pile is located further away from maximum reduction in Hu was observed at LC = 5.25 DP. Therefore,
the slope crest and reaches maximum at a depth of 0.135 m as a design guideline, it is essential to determine the actual LC on
from the ground surface, at a ratio of 0.386 pile length (Mmax site to avoid the effect of critical depth as well as increasing the lat-
at 0.386 LP). eral pile capacity, these can be achieved either by strengthening
4) The peak values of bending moment along the pile length the contaminated layer using a suitable stabilization method [24]
raised by about 11, 12, and 16% at oil contents of 2, 4, and or slope reinforcement technique [11], particularly the Honeycomb
6% respectively, compared to piles in uncontaminated cellular structures due to their light weight and high efficiency to
(clean) sand slope located at X = 2Dp from the slope crest. resist different loads [17]. Furthermore, the most effective reduc-
756 R.A. Abdelhalim et al. / Engineering Science and Technology, an International Journal 23 (2020) 744–757

For (O.C) = 2% For (O.C) = 4%

For (O.C) = 6%

Fig. 19. Images at X = 6DP during lateral loading tests of short pile embedded in oil-contaminated sand slope at different oil contents.

tion in Hu was observed at 4% contamination; therefore, prior to a [10] J.M. Duncan et al., Lateral load analysis of single piles and drilled shafts, J.
Geotechn. Eng. 120 (6) (1994) 1018–1033.
construction, the percentage of contamination should be well-
[11] M. El Sawwaf, Lateral behavior of vertical pile group embedded in stabilized
investigated beforehand. earth slope, J. Geotech. Geoenviron. Eng. 134 (7) (2008) 1015–1020.
[12] E. Evgin, B. Das, Mechanical Behavior Of An Oil Contaminated Sand,
Environmental Geotechnology, Usmen Acar, Balkema, Rotterdam, The
Declaration of Competing Interest Netherlands, 1992.
[13] M.A. Gabr, R.H. Borden, Lateral analysis of piers constructed on slopes, J.
The authors declare that they have no known competing finan- Geotechn. Eng. 116 (12) (1990) 1831–1850.
[14] M. Hajialilue-Bonab et al., Study of strain wedge parameters for laterally
cial interests or personal relationships that could have appeared loaded piles, Int. J. Geomech. 13 (2) (2013) 143–152.
to influence the work reported in this paper. [15] M. Khamehchiyan et al., Effects of crude oil contamination on geotechnical
properties of clayey and sandy soils, J. Eng. Geol. 89 (3–4) (2007)
220–229.
References [16] E. Khosravi et al., Geotechnical properties of gas oil-contaminated kaolinite,
Eng. Geol. 166 (2013) 11–16.
[1] D.B. Al-Duwaisan, A.A. Al-Naseem, Characterization of Oil Contaminated Soil [17] M.Z. Mahmoudabadi et al., Theoretical and experimental crushing analysis of
(Kuwait Oil Lakes), 2nd International Conference on Environmental Science metal square honeycombs under quasi-static loading, ASME 2010 10th
and Technology, IACSIT Press, Singapore, 2011. Biennial Conference on Engineering Systems Design and Analysis, American
[2] A.I.J. Al-Mhaidib, Effect of loading rate on pile groups in sand, International Society of Mechanical Engineers, 2010.
Conference on Geotechnical Engineering, 2004. [18] S. Mezazigh, D. Levacher, Laterally loaded piles in sand: slope effect on PY
[3] H.A. Al-Sanad et al., Geotechnical properties of oil-contaminated Kuwaiti sand, reaction curves, Can. Geotech. J. 35 (3) (1998) 433–441.
J. Geotechn. Eng. 121 (5) (1995) 407–412. [19] H. Mozafari et al., Numerical and experimental investigation of corrugated
[4] ASTM, American Society for Testing and Materials Specifications, ASTM tubes under lateral compression, Int. J. Crashworthiness 23 (4) (2018) 461–
International, West Conshohocken, PA, USA, 2010. 473.
[5] L. Bowles, Foundation Analysis and Design, McGraw-hill, 1996. [20] K.J.I.J. Muthukkumaran, o. G, Effect of slope and loading direction on laterally
[6] B.B. Broms, Lateral resistance of piles in cohesionless soils, J. Soil Mech. Found. loaded piles in cohesionless soil, Int. J. Geomech. 14 (1) (2014) 1–7.
Division 90 (3) (1964) 123–158. [21] A. Nasr, S. Krishna Rao, Behaviour of laterally loaded pile groups embedded in
[7] K. Chae et al., Lateral resistance of short single piles and pile groups located oil-contaminated sand, Géotechnique 66 (1) (2016) 58–70.
near slopes, Int. J. Geomech. 4 (2) (2004) 93–103. [22] A.M. Nasr, Experimental and theoretical studies for the behavior of strip
[8] Chari, T. R. and G. G. J. C. G. J. Meyerhof, 1983. Ultimate capacity of rigid single footing on oil-contaminated sand, J. Geotech. Eng. 135 (12) (2009)
piles under inclined loads in sand. 20(4), pp. 849-854. 1814–1822.
[9] D. Dewaikar et al., Lateral resistance of long piles in cohesive soils using py [23] A.M. Nasr, Experimental and theoretical studies of laterally loaded finned piles
curves, J. Struct. Eng.-ASCE 38 (3) (2011) 222–227. in sand, Can. Geotech. J. 51 (4) (2013) 381–393.
 R.A. Abdelhalim et al. / Engineering Science and Technology, an International Journal 23 (2020) 744–757 757

[24] A.M. Nasr, Utilisation of oil-contaminated sand stabilised with cement kiln [30] PLAXIS, PLAXIS 3D foundation user manual, version 1.6, Plaxis Inc., Delft, the
dust in the construction of rural roads, J Int. J. Pavement Eng. 15 (10) (2014) Netherlands, 2008.
889–905. [31] S. Prakash, Behavior of Pile Groups Subjected to Lateral Loads PhD thesis,
[25] A.M. Nasr, Behaviour of strip footing on oil-contaminated sand slope, J. Int. J. University of Illinois, Champaign, IL, USA, 1962.
Phys. Modell. Geotech. 16 (3) (2015) 134–151. [32] M.M. Rahman et al., Investigation of Lateral Load Resistance of Laterally
[26] A.M. Nasr, Behaviour of strip footing on oil-contaminated sand slope, Int. J. Loaded Pile In Sandy Soil, in: Proceedings of the 4th international conference
Phys. Modell. Geotechn. 16 (3) (2016) 134–151. on bored and auger piles, BAPIV, Ghent, Belgium, 2003, pp. 209–215.
[27] J. Otani et al., Investigation of failure patterns in sand due to laterally loaded [33] A. Sadighi et al., A novel axially half corrugated thin-walled tube for energy
pile using X-ray CT, J. Soils Found. 46 (4) (2006) 529–535. absorption under Axial loading, J. Thin-Walled Struct. 145 (2019) 106418.
[28] N.K. Ovesen, The scaling law relationship-panel discussion, Proc. 7th European [34] E.C. Shin, B.M. Das, Bearing capacity of unsaturated oil-contaminated sand, Int.
Conference on Soil Mechanics and Foundation Engineering, Brighton, UK, J. Offshore Polar Eng. 11 (03) (2001).
1979. [35] K. Terzaghi, Evalution of conefficients of subgrade reaction, Geotechnique 5 (4)
[29] A. Patel, Study of Geotechnical properties of black cotton soil contaminated by (1955) 297–326.
castor oil and stabilization of contaminated soil by sawdust, National [36] L. Zhang et al., Ultimate lateral resistance to piles in cohesionless soils, J.
Conference on Recent Trends in Engineering, 2011. Geotechn. Geoenviron. Eng. 131 (1) (2005) 78–83.