Marine Structures 29 (2012) 198–210

Contents lists available at SciVerse ScienceDirect

Marine Structures
journal homepage: www.elsevier.com/locate/
marstruc

Full scale lateral behaviour of monopiles in granular marine

soils
A.S. Hokmabadi a, A. Fakher b, *, B. Fatahi a
a
Centre for Built Infrastructure Research (CBIR), School of Civil and Environmental Engineering,
University of Technology Sydney (UTS), P.O. Box 123, Sydney, Australia
b
School of Civil Engineering, College of Engineering, University of Tehran, P.O. Box 11155-4563,
Tehran 11365-4563, Iran

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

Article history: Monopiles are used in piers as mooring or berthing dolphins. This
Received 26 April 2010 article reports the results of full-scale lateral loading tests on
Received in revised form 16 May 2012 monopiles constructed as dolphins in the Pars Special Economic
Accepted 17 June 2012
Energy Zone in southern Iran. The length and diameter of the
monopiles were approximately 40 m and 2 m, respectively. Lateral
Keywords:
loading tests of such large monopiles are not commonly under-
Monopile
taken, thus there is limited data available. This research developed
Full-scale in-situ tests
Lateral behaviour a lateral analysis of piles computer code to examine analytical
p–y curves methods for pile analysis. Appropriate models were introduced
Strain wedge model resulting in accurate predictions in the analysis of lateral loaded
piles. The results showed that traditional p–y curves and strain
wedge models calculate larger pile head deflection in comparison
to the field test data and therefore local calibration is essential.
Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction

In general, there are two types of berth structures; quay and jetty. A quay (or wharf) is a landing
place parallel to a navigable waterway that provides access to ships and boats (Fig. 1a). Due to its high
lateral resistance, the fenders must be well-designed to absorb the berthing energy of a ship. A jetty (or
pier) extends out into the water from the shore. It is used vertically with the shoreline serving as

* Corresponding author. Tel.: þ98 21 66498981; fax: þ98 21 66461024.

E-mail addresses: aslan.sadeghihokmabadi@student.uts.edu.au (A.S. Hokmabadi), afakher@ut.ac.ir (A. Fakher), Behzad.
fatahi@uts.edu.au (B. Fatahi).

0951-8339/$ – see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.marstruc.2012.06.001
 A.S. Hokmabadi et al. / Marine Structures 29 (2012) 198–210 199

Fig. 1. (a) Schematic of a quay; (b) schematic of a jetty with two central berthing and four mooring dolphins at the sides; (c)
construction of a jetty with monopiles in Asalouyeh.

a landing place and where loading equipment allows the use of a lighter structure. Ships can berth
directly at the structure, but usually require separate structures, such as dolphins, to absorb the high
energy of the ship (Fig. 1b) [1].
In some cases, dolphins consist of a number of piles. This induces low lateral deformation and,
therefore, a reduced ability to absorb energy. A monopile is comprised of a single large-diameter pile
embedded in the soil and behaves as a console. The ability of monopiles to absorb a high amount of
energy, their low cost, and simple construction method has made them common alternatives for
offshore structures such as wind turbines and mooring or berthing dolphins.
In the analysis of monopiles, lateral behaviour is important and the interaction between the pile and
soil should be modelled accurately. One popular approach to determine the lateral behaviour of a pile is
the load-transfer method in which the pile is modelled as an elastic member and the soil is modelled as
a series of linear or nonlinear springs (p–y curves). Traditional p–y curves do not account for some
properties such as pile bending stiffness, cross-sectional shape, head restraint, and installation method
[2]. The strain wedge model (SWM) considers some of these limitations. However, both the traditional
p–y curve and SWM are semi-empirical methods. The main drawback to these approaches is that they
are based on empirical parameters (i.e., the modulus of subgrade reaction) which can only be
computed from the results of pile load tests [3]. In addition, the capability of these methods in pre-
dicting the behaviour of large diameter piles is questionable. Lensy et al. [4] studied the behaviour of
large diameter monopile foundation for typical loading conditions and denoted the shortcomings of
both p–y methods and SWM in predicting the behaviour of large diameter monopiles due to their
greater critical length and linear assumption of soil stiffness distribution in these methods, empha-
sising on the necessity of calibration for these methods.
The aim of this study is to assess the accuracy of these methods using the results of full-scale tests
and calibrate them for the laterally loaded monopiles at the Pars Special Economic Energy Zone Area
(Asalouyeh) in Iran. The details and results of the full-scale tests are presented and the characteristics
of the newly-developed computer code (LAP) have been described. The monopiles were then analysed
using LAP and the results are compared with the test data. The models have been calibrated, modifi-
cation factors have been recommended and practical conclusions have been drawn.
200 A.S. Hokmabadi et al. / Marine Structures 29 (2012) 198–210

2. Analysis of lateral behaviour of piles

The behaviour of piles under lateral loading is complex because of the interaction between the pile
and surrounding soil. A number of researchers have investigated laterally loaded pile behaviour,
providing a number of different approaches. These methods can be classified as continuum-based or
load-transfer (subgrade reaction) approaches.
In the continuum-based approach, the soil is modelled as a continuum media, requiring several soil
properties as the input for analysis [5]. The complexity and unavailability of soil properties of this first
approach make it less attractive. The load-transfer approach is more commonly used and has been
selected for this study.
The load-transfer method models the pile as an elastic member and the soil as a series of nonlinear
springs (p–y curves). The nonlinear soil springs describe the local variation of lateral soil–pile inter-
acting resistance with lateral displacement. Traditional p–y models were initially developed by Mat-
lock [6] and Reese et al. [7]. Later, a number of p–y curves were developed for sand and clay as tabulated
in Table 1. Fig. 2 shows p–y curve for sands developed by Reese et al. [7]. The curve has three segments
including two straight lines connected by a parabola, where A0 and B0 are the constants, k is the initial
stiffness, and the other parameters are defined in Table 1.
SWM approach [14] has been developed to predict the response of flexible piles under the lateral
loading. The modified version of this method can consider seismic loads as well [15]. In this model, soil
resistance against the lateral loading is determined by 3D passive wedge of soil developed in front of
the pile as shown in Fig. 3a, in which bm ¼ 45 þ 4m/2 is the base angle, h is the passive wedge depth,
4m is the mobilized friction angle, Dsh is the horizontal stress change at the wedge face, and s is the side
shear stress. SWM allows the assessment of the nonlinear p–y curve response of a laterally loaded pile
based on the envisioned relationship between 3D response of a flexible pile in the soil to the one-

Table 1
Summary of p–y curves for sand with modifications [8].

p–y curve function Reference Remarks

p ¼ Cy1=n Reese et al. [7] Developed function from results
Pm P ðP  Pm Þ of full-scale tests performed at
where C ¼
.  ; n ¼ m ; and m ¼ u
mym ðyu  ym Þ Mustang Island (function consists
1 n of 3 segments having 2 straight
ym
lines connected by a parabola).
 
kz Murchison and O’Neill [9] Developed function from back
p ¼ A  pu  tanh y
A  pu analyses of full-scale instrumented
where Pu ¼ minððC1 z þ C2 DÞgz; C3 DgzÞ pile load test on sand.
C1 ; C2 ; and C3 are constants; A is the empirical function;
z is the depth; and D is the pile diameter
y Kondner [10] Developed hyperbolic function
p ¼
1 y from results of stress–strain
þ
kini pu relationship of soil in triaxial
where pu is the ultimate subgrade reaction 9 compression tests.
s00 D Scott [11] Developed bilinear function from
Pk ¼
0:5
1 1 1 results of centrifuge tests.
þ
p sin2 4 3  4D
where s00 ¼ ðs01 þ s02 þ s03 Þ=3


y p Det Norske Veritas [12] Developed function from combined
p ¼ ; where f ¼ 1= 1  d
1 y k h yd hyperbolic and linear from results
þ
kh apd of full-scale and model tests.

n  m
z y Wesselink et al. [13] Developed function from the results
p ¼ Rd
z0 D of full-scale tests in calcareous sand
where z0 ¼ 1 m; R ¼ 850 kPa; of Bass Strait.
n and m are the emperical factors
P Ai DεEi Ashour et al. [2] Based on SWM related 3D response
ðEs Þi ¼ i ¼
yi dðh  xi Þ of pile in soil to its 1D beam on elastic f
oundation parameters.
 A.S. Hokmabadi et al. / Marine Structures 29 (2012) 198–210 201

Fig. 2. p–y curve for sand as suggested by Reese et al. [7].

dimensional response of a beam on elastic foundation [2] as depicted in Fig. 3b, where x0 is the depth of
zero deflection point, Y0 is the pile-head deflection, P0 is the applied horizontal load at pile head, Pn is
the soil-pile reaction or soil resistance for layer n, and Es is the modulus of subgrade reaction derived
from SWM. It should be noted that by considering soil continuity, this model calculates its own set of
non-unique p–y curves which are functions of both soil and pile properties.
In this study, p–y curves suggested by Murchison and O’Neill [9] and Reese et al. [7] for sands
together with SWM have been adopted.

3. Full-scale tests

Research conducted on the behaviour of laterally loaded piles can be divided into full-scale and
small-scale (or model) testing. Full-scale tests are generally believed to provide the most accurate
results, but they are rare because they are expensive and difficult to run. This makes the results of full-
scale tests valuable. In the present research, a number of full-scale tests were performed on monopiles.

3.1. Test location

Asalouyeh is located in southern Iran on the Persian Gulf. It is 300 km east of the city of Bushehr on
the coast of Iran. Pars Petrochemical Port in Asalouyeh has 15 berths. At piers 5 and 15, monopiles are
used as berthing and anchoring dolphins (Fig. 4).

Fig. 3. Configuration of strain wedge model [2]: (a) basic strain wedge in uniform soils; (b) soil-pile interaction in the multisublayer SWM.
202 A.S. Hokmabadi et al. / Marine Structures 29 (2012) 198–210

Fig. 4. Pars petrochemical port in Asalouyeh; (a) berthing dolphin; (b) port layout and berth locations.

Four monopiles were tested. Monopiles No. 1 and 2 are the inner and outer piles of Berth 15 at
a water depth of 14 m. Monopiles No. 3 and 4 are the inner and outer piles of Berth 5 at a water depth of
26 m. The final elevation of the monopile heads after installation was 5 m above the mean sea level.
These monopiles have a cylindrical shape and were made from three types of steel. The thickness and
types of steel used are variable in depth and are shown in Tables 2 and 3. Details of the monopiles are
shown in Fig. 5.
The soil parameters in the field were obtained for each layer using borings. Because of the high soil
stiffness, it was not possible to perform in-situ tests such as the standard penetration test. The
geotechnical properties of the soil are shown in Table 4. These parameters were obtained by describing
the disturbed samples and laboratory tests. The ground consists of 8 m of sand (SP), followed by 13 m of
gravely sand (GP) and 8 m of sand stone. Friction angle of top sand later and gravely sand layer are 38
and 40 , respectively. Table 4 presents the dry density (gd), wet density (gt), estimated value of
standard penetration test (Nspt), effective cohesion (C0 ), internal friction angle (40 ), Poisson’s ratio (y),
and the estimated value of Young modulus derived from the empirical relationships (Es).
Although the soil data obtained from the site are limited, they are sufficient to calculate the p–y
curves used in this study. As previously mentioned, the main advantage of the load-transfer
approaches (p–y curves) over the continuum-based methods is that less soil properties as the input
for the analysis are required due to their semi-empirical basis. For instance, in order to calculate the p–y
curves for sands as suggested by Reese et al. [7], Murchison and O’Neill [9], or SWM [2], soil density,
friction angle, and layer thickness are the main required soil parameters.

3.2. Test method

A heavy duty tension system was designed and constructed that uses a hydraulic jack to provide
force and a cable to transfer tension force from one monopile to another. The testing followed ASTM

Table 2
Characteristics of materials used in monopiles.

Type of steel Yielding stress (kN/m2) Ultimate stress (kN/m2) E (kN/m2) Poison ratio Density (kN/m3)
ST52 360,000 520,000 2.1 e þ 08 0.3 78.5
ST60 420,000 600,000 2.1 e þ 08 0.3 78.5
ST70 490,000 700,000 2.1 e þ 08 0.3 78.5
 A.S. Hokmabadi et al. / Marine Structures 29 (2012) 198–210 203

Table 3
Details of monopile sections.

Monopile no. 1 Monopile no. 2

Section Type of steel Outer diameter (m) Thickness (mm) Section Type of steel Outer diameter (m) Thickness (mm)
1 ST52 1.778 25.40 1 ST52 1.905 25.40
2 ST60 1.778 25.40 2 ST52 1.905 28.58
3 ST70 1.778 28.58 3 ST60 1.905 28.58
4 ST70 1.778 31.75 4 ST70 1.905 34.93
5 ST70 1.778 34.93 5 ST70 1.905 41.28
6 ST70 1.905 44.45

Monopile No. 3 Monopile No. 4

Section Type of steel Outer diameter (m) Thickness (mm) Section Type of steel Outer diameter (m) Thickness (mm)
1 ST52 1.778 25.40 1 ST52 1.905 25.40
2 ST60 1.778 25.40 2 ST60 1.905 25.40
3 ST70 1.778 28.58 3 ST70 1.905 28.58
4 ST70 1.778 31.75 4 ST70 1.905 31.75
5 ST70 1.778 34.93 5 ST70 1.905 34.93
6 ST60 1.778 34.93 6 ST70 1.905 44.45
7 ST70 1.905 41.28

D3966-81, item 24 (ASTM 1995) [16]. The tension system sat on one monopile and pulled the other one
(Fig. 6). Bolts placed in the head of the monopiles for a quick release system used for the temporary
installation of the tension system on one monopile and a pulley on the other. Cables were installed
between the tension system on one monopile and support on the other with a 1422 mm (56 in)

Fig. 5. Details of monopiles in Asalouyeh.

204 A.S. Hokmabadi et al. / Marine Structures 29 (2012) 198–210

Table 4
Geotechnical properties of soil in the field [17].

Layer Depth (m) Classification gd (kN/m3) gt (kN/m3) NSPT C0 (kN/m2) 40 (degree) y Es (MPa)
Sand 0.0–8.0 SP 1.7 2 >50 0 38 0.37 60
Sand and gravel 8.0–21.0 GP 1.95 2.1 >50 0 40 0.25 120
Sand stone 21.0–30.0 – 1.8 2.1 _ _ _ _ _

diameter pipe between them to support the weight of the cables and avoid any initial force induced by
them. Fig. 7 shows the allowance for axial displacement during the test. Analysis shows that the
maximum friction between the cable and the pipe was less than 3% of the applied load and may be
disregarded [17]. In addition, since the spacing between the piles (21.5 m) is more than eight times the
diameter of the piles, no pile group effect is expected [18].

3.3. Test results

Four monopiles were tested under the lateral static loading. Monopiles No. 1 and 2 were loaded in
five steps. Monopiles No. 3 and 4 were loaded in three steps to accommodate the displacement
limitation of the jacking system. At each step, the displacement of each pile was measured using Total
Station. The loading steps increased and, for each step, the load was maintained for 15 min for small
loads and 30 min for large loads. Table 5 shows the results.

4. Lap computer code

A computer code was developed to analyse the monopiles. The Lateral Analysis of Piles (LAP)
computer code was written in FORTRAN programming language to solve the governing equations for
a beam on an elastic foundation (Equation (1)) by Hetenyi [19],


2 

d4 y d y
EI þ Px þ Es y ¼ 0 (1)
dx4 dx2
where EI ¼ bending stiffness of the pile, Px ¼ axial load on the pile, y ¼ lateral deflection of the pile at
point x along the length of the pile, and Es ¼ soil subgrade reaction (spring stiffness). LAP uses the finite

Fig. 6. Plan and sections of test arrangement.

 A.S. Hokmabadi et al. / Marine Structures 29 (2012) 198–210 205

Fig. 7. Test arrangement and the pipe installed between two test monopoles.

difference method proposed by Matlock and Reese [20] to solve Equation (1). It considers four sets of
boundary conditions at the top of the pile, such as free-head or fixed-head pile. Furthermore, LAP uses
different spring stiffnesses as shown in Table 1.
To model the soil reaction against the lateral loads at any depth, the following models were
considered:

 Non-linear springs based on p–y curves suggested by Reese et al. [7].

 Non-linear springs based on p–y curves suggested by Murchison and O’Neill [9].
 SWM developed by Ashour et al. [2].

In addition, LAP can assess pile group behaviour under the lateral and dynamic lateral loading such
as earthquake loads [17]. In this research, LAP has used as a means to analyse the monopiles at Asa-
louyeh and to calibrate lateral spring models.
To examine the validity of the LAP results, a number of piles were analysed using LAP and COM624
[21], the conventional program to analyse single piles under lateral loads. The SAP2000 software
(CSI2003) [22] was used to examine the validity of the LAP as well. In the SAP2000 software the Link
element, as defined in the program manual, was used to model the desired p–y curves and the piles
were reanalysed. Fig. 8 presents the comparison of results for Monopile No.1 using Reese et al. [7]
curves for illustration. As shown, the results of LAP were in a good agreement with other programs.

Table 5
Results of lateral loading tests.

Load at Displacement Displacement of Load at Displacement Displacement of

monopile of monopile monopile no. monopile of monopile monopile
top (ton) no. 1 (mm) 2 (mm) top (ton) no. 3 (mm) no. 4 (mm)
30 78 60 31.4 318 247
60 204 140 50.3 555 418
70 230 155 62.8 786 585
80 280 190
90 320 220
206 A.S. Hokmabadi et al. / Marine Structures 29 (2012) 198–210

100

90

80

70
Force (ton)

60

50

40

30
LAP
20
COM624
10 SAP2000

0
0 10 20 30 40 50
Displacement (cm)

Fig. 8. Lateral analysis of monopile no. 1 compared to LAP and other programs.

100
100 Monopile No. 2
90 a Monopile No. 1 90 b
80
80
70
70
Force (ton)
Force (ton)

60
60
50 50

40 Test Data 40 Test Data

30 Reese curves [6] 30 Reese curves [6]

20 Murchinson curves [8] 20 Murchinson curves [8]

10 10 SWM [2]
SWM [2]
0 0
0 5 10 15 20 25 30 35 40 45 0 5 10 15 20 25 30
Displacement (cm) Displacement (cm)

70 70

c Monopile No. 3 d Monopile No. 4

60 60

50 50
Force (ton)
Force (ton)

40 40

30 30
Test Data Test Data

20 Reese curves [6] 20 Reese curves [6]

Murchinson curves [8] Murchinson curves [8]

10 10
SWM [2] SWM [2]
0 0
0 20 40 60 80 100 0 10 20 30 40 50 60 70
Displacement (cm) Displacement (cm)

Fig. 9. Pile head horizontal displacement versus lateral load for test data and LAP analysis for: (a) monopile no. 1, (b) monopile no. 2,
(c) monopile no. 3, (d) monopile no. 4.
 A.S. Hokmabadi et al. / Marine Structures 29 (2012) 198–210 207

5. Comparision and discussion

The tested monopiles were analysed by LAP using three different methods. In the first method, the
p–y curves suggested by Reese et al. [7] were adopted. In the second method, the p–y curves suggested
by Murchison and O’Neill [9] were used, and finally, SWM was used as the third method to analyse the
monopiles under lateral loads. Fig. 9 presents the results of these analyses for monopile head
displacement versus lateral load at the head of the monopiles.
As shown in Fig. 9, all three methods calculate larger pile head deflection in comparison to the field
test data. Modification factors (F) have been defined to calibrate the p–y curves in the first and second
methods. F is multiplied by parameter p of the p–y curve to equalize the results with the data. The F
values are presented in Table 6, and the modified monopile head displacement versus applied lateral
load is shown in Fig. 10. Considering the large free length of monopiles above the mud-line (sea floor),
the pile head deformations are dominated by this free length, which is why large calibration factors (F)
only have minor influence on the deformations. In the third method, SWM receives the force at the pile
head as the input and give the pile head displacement as the output [2]. This method calculates p–y
curves during computation for each case [23]. In other words, SWM does not define certain p–y curves,
as in the first and second methods, so F could not be calculated for the third method. It should be noted
that the linear behaviour for the monopile structure is assumed and the monopiles behave as long
piles. The type of the pile (long, intermediate or short) is defined based on soil and pile properties [24].
The real behaviour of pile head displacement is non-linear. The p–y models are different than real
behaviour in the flexure of the pile head-displacement curves. The reduction in lateral stiffness of in-

Table 6
Results of pile analysis under lateral load.

Monopile no. 1 Reese curves Murchinson curves SWM

Lateral load Test data Head disp. Head disp. with Head disp. Head disp. with Head disp. Lateral load
(ton) (cm) (cm) F ¼ 4.2 (cm) (cm) F ¼ 2 (cm) (cm) (ton)
30 7.8 11.51 10.176 11.22 10.23 9.93 27.51
60 20.4 25.08 20.799 23.25 20.84 25.26 63.55
70 23 30.37 24.467 27.52 24.5 28.3 70.21
80 28 35.49 28.216 31.95 28.24 31.02 75.96
90 32 40.7 31.977 36.54 32.06 34.84 84.27

Monopile no. 2 Reese curves Murchinson curves SWM

Lateral load Test data Head disp. Head disp. with Head disp. Head disp. with Head disp. Lateral load
(ton) (cm) (cm) F ¼ 3.5 (cm) (cm) F ¼ 2.5 (cm) (cm) (ton)
30 6 7.92 6.68 7.8 6.9 8.06 34.98
60 14 16.89 14.24 15.97 13.94 16.07 59.63
70 15.5 20.19 16.69 18.81 16.33 19.35 70.27
80 19 23.57 19.17 21.73 18.75 23.03 82.18
90 22 27.04 21.73 24.72 21.2 25.78 90.61

Monopile no. 3 Reese curves Murchinson curves SWM

Lateral load Test data Head disp. Head disp. with F ¼ 2 Head disp. Head disp. with F ¼ 1 Head disp. Lateral load
(ton) (cm) (cm) (cm) (cm) (cm) (cm) (ton)
31.4 31.8 39.73 37.94 38.39 38.39 38.29 30.96
50.3 55.5 67.2 61.88 63.08 63.08 65.59 50.8
62.8 78.6 85.87 78.03 80.03 80.03 80.5 61.47

Monopile no. 4 Reese curves Murchinson curves SWM

Lateral load Test data Head disp. Head disp. with Head disp. Head disp. with Head disp. Lateral load
(ton) (cm) (cm) F ¼ 1.5 (cm) (cm) F ¼ 1 (cm) (cm) (ton)
31.4 24.7 28.26 28.04 27.56 27.56 27.04 32.28
50.3 41.8 47.24 45.66 44.94 44.94 47.02 53.79
62.8 58.5 60.93 57.63 56.89 56.89 54.51 61.73
208 A.S. Hokmabadi et al. / Marine Structures 29 (2012) 198–210

100 100

90 a Monopile No. 1 90 b Monopile No. 2

80 80

70 70

Force (ton)
Force (ton)

60 60

50 50

40 40

30 30
Test Data Test Data
20 20
Modified Reese curves (F=4.2) [6] Modified Reese curves (F=3.5) [6]

10 10
Modified Murchinson curves (F=2) [8] Modified Murchinson curves (F=2.5) [8]

0 0
0 5 10 15 20 25 30 35 0 5 10 15 20 25
Displacement (cm) Displacement (cm)

70 70

c Monopile No. 3 d Monopile No. 4

60 60

50 Force (ton)
50
Force (ton)

40 40

30 30

20 20
Test Data Test Data

Modified Reese curves (F= 2) [6] Modified Reese curves (F=1.5) [6]
10 10
Modified Murchinson curves (F=1) [8] Modified Murchinson curves (F=1) [8]

0 0
0 10 20 30 40 50 60 70 80 90 0 10 20 30 40 50 60 70
Displacement (cm) Displacement (cm)

Fig. 10. Results when modification factor (F) is used: (a) monopile no. 1, (b) monopile no. 2, (c) monopile no. 3, (d) monopile no. 4.

situ soil occurs in smaller amount of displacement, whereas the p–y curve declines in larger
displacements and has an approximately linear behaviour under the test loads. This difference is due to
the development of a plastic region near the soil surface which yields under test loads. However, the
existing p–y models do not capture this behaviour. Consequently, for the modification of p–y curves,
the ultimate resistance of the non-linear springs should be reduced and their primary stiffness should
be increased. It should be mentioned that the overall behaviour of the pile-spring system is very
sensitive to the near surface spring stiffness, which should be selected carefully.

6. Parametric studies

As noted, a monopile may have a large diameter and long free length. In this research, a number of
p–y curves used in LAP were calibrated using data from full-scale in-situ tests. Based on these curves,
a parametric study was conducted using the parameters of soil strength (internal friction angle), free
length of monopile, embedded depth of monopile, and their effects on monopile head displacement
and depth of fixity as obtained in the equivalent cantilever beam method.
The results of the analysis of Monopile No. 2 with p–y curves suggested by Reese et al. [7] for a pile
head load of 800 kN are shown in Fig. 11.
As shown in Fig. 11a, by increasing the internal friction angle of the soil, the stiffness of springs
representing the soil media increases and consequently the head displacement and depth of fixity
decreases. An increase in the embedded depth of the monopile contributes to reduction of the pile
head displacement and depth of fixity (Fig. 11b). As mentioned earlier, one of the significant parameters
influencing the behaviour of monopiles and specifically pile head displacement prediction is the free
length of monopiles, and according to Fig. 11c, the pile head displacement increases significantly by
 A.S. Hokmabadi et al. / Marine Structures 29 (2012) 198–210 209

30
60
a b
25 50
Head Displacement (cm)

Head Disp.

Head Displacement (cm)

20
Depth of fixity (m)

Depth of fixity (m)

40
Depth of fixity
Head Disp.
15 30
Depth of fixity

10 20

5 10

0 0
33 34 35 36 37 38 39 40 41 42 43 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Internal friction angles (degree) Embedded depth (m)

50

45
c
40
Head Disp.
Head Displacement (cm)

35
Depth of fixity (m)

Depth of fixity
30

25

20

15

10

0
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Free length (m)

Fig. 11. Parametric study: (a) effect of internal friction angle of soil; (b) effect of free length of monopoles; (c) effect of embedded
depth of monopiles.

increasing the free length of the pile. Therefore, the effect of pile free length should be considered in
calibrating the load-transfer methods.

7. Conclusion

This article reports the details and results of full-scale lateral loading tests on monopiles con-
structed as dolphins in southern Iran. In addition, LAP computer code was developed to analyse piles
under lateral loading. This computer code has the ability to consider different boundary conditions and
types of p–y curves. In accordance with the results of full-scale in-situ tests, the accuracy of p–y curves
suggested by Reese et al. [7] and Murchison and O’Neill [9], and SWM [2] were investigated and
calibrated for local conditions by defining modification factors.
In granular marine soils, the traditional p–y curves and SWM result in larger pile head displacement
in comparison to the field measurements. This means that in-situ piles withstood larger amounts of
force for the specified displacements. Thus, application of existing p–y curves without calibration leads
to overestimation in the design. It is noted that, the predictions by SWM are more accurate than the
Reese curves [7], and Murchison and O’Neill [9] curves lead to the most accurate values. The shape of
the pile lateral displacement under real conditions declines sharper than the prediction as the
analytical models do not chapter soil plasticity near ground depth. Therefore, to modify p–y curves, the
ultimate resistance of the non-linear springs should be reduced and the primary stiffness should be
increased. In general, it is recommended to the practising engineers to calibrate the above mentioned
methods before applying to large diameter monopile foundations.
210 A.S. Hokmabadi et al. / Marine Structures 29 (2012) 198–210

References

[1] Quinn AD. Design and construction of ports and marine structures. USA: McGraw-Hill Inc.; 1972.
[2] Ashour M, Norris G, Pilling P. Lateral loading of a pile in layered soil using the strain wedge model. J Geotech and Geo-
envron Eng 1998;124(4):303–15.
[3] Basile F. Analysis and design of pile groups. In: Bull JW, editor. Numerical analysis and modeling in geomechanics. London:
Spon Press; 2003. p. 278–315.
[4] Lesny K, Paikowsky SG, Gurbuz A. Scale effects in lateral load response of large diameter monopiles. Geo-Denver 2007:
new peaks in geotechnics. Colorado: ASCE; 2007. 40.
[5] Fleming WGK, Weltman AJ, Randolph MF, Elson WK. Piling engineering. 2nd ed. Glasgow, UK: Blackie Academic &
Professional; 1992.
[6] Matlock H. Correlations for design of laterally loaded piles in soft clay. In: Proc. 2nd Ann. offshore Tech. Conf. Houston,
Texas: 1970. p. 577–94.
[7] Reese LC, Cox WR, Koop FD. Analysis of laterally loaded piles in sand. In: Proc. 6th Ann. offshore Tech. Conf. Austin, Texas:
1974. p. 473–85.
[8] Kim BT, Kim NK, Lee WJ, Kim YS. Experimental load-transfer curves of laterally loaded piles in Nak-Dong river sand. J
Geotech and Geoenvron. Eng 2004;130(2):416–25.
[9] Murchison JM, O’Neill MW. Evaluation of p–y relationships in cohesionless soil. In: Analysis and design of pile foundations.
New York: ASCE; 1984. p. 174–91.
[10] Kondner RL. Hyperbolic stress–strain response: cohesive soils. J Soil Mech and Fund Div 1963;89(1):115–44. ASCE.
[11] Scott RF. Analysis of centrifuge pile tests: simulation of pile driving. Res. Rep., OSAPR Project 13. Washington, D.C:
American Petroleum Institute; 1980.
[12] Det Norske Veritas. Rules for the design, construction and inspection of offshore structures, appendix F: foundations.
Hovik, Norway: Det NorkesVeritas; 1980.
[13] Wesselink BD, Murff JD, Randolph MF, Nunez IL, Hyden AM. Analysis of centrifuge model test data from laterally loaded
piles in calcareous sand. In: Engineering for calcareous sediments, vol. 1. Rotterdam, The Netherlands: Balkema; 1988. p.
261–70.
[14] NorrisGM. Theoretically based BEF laterally loaded pile analysis. In: Proc. third Int. Conf. on Numer. Meth. in offshore
piling. Nantes, France: 1986. p. 361–86.
[15] Hokmabadi AS, Fakher A, Fatahi B. Seismic strain wedge model for analysis of single piles under lateral seismic loading.
Aust Geomechanics 2001;46(1):31–41.
[16] American Society for Testing and Materials. Standard method of testing piles under lateral loads. In: Annual book of ASTM
standards, D3966–90. ASTM; 1995.
[17] Hokmabadi AS. Development of a computer program for the analysis of single piles and pile groups under lateral loads.
Post-graduate research thesis: University of Tehran. Iran: 2009.
[18] Reese LC, Vanimpe WF. Single piles and pile groups under lateral loading. Rotterdam, Netherlands: A.A. Balkema; 2001.
[19] Hetenyi M. Beams on elastic foundations. Ann Arbor: The University of Michigan Press; 1946.
[20] Matlock H, Reese LC. Generalized solution for laterally loaded piles. J Soil Mech and Fund Div 1961;86(5):673–94. ASCE.
[21] Reese LC, Sullivan WR. Documentation of computer program COM624, parts 1 and 2: analysis of stresses and deflections
for laterally loaded piles including generation of p–y curves. Geotech. Eng. Ctr., Bureau of Eng. Res.. Austin, Texas:
University of Texas; 1980
[22] Computer and Structures, Inc.. SAP2000 version 10.0.1 user’s Guide, manual. Computer and Structures, Inc.; 2003.
[23] Hokmabadi AS, Seyfi H, Fakher A. Analysis of single piles under lateral loading using the strain wedge model. In: presented
at the 8th Int. Cong. of Civil Eng.: Shiraz University. Shiraz, Iran: 2009 [in Farsi].
[24] Ashour M, Norris G. Report and user manual on strain wedge model computer program for piles and large diameter shafts
with LRFD procedure. Rep. CCEER-07–07. USA: The State of Nevada Department of Transportation (NDOT); 2007.