Ocean Engineering 218 (2020) 108143

Contents lists available at ScienceDirect

Ocean Engineering
journal homepage: www.elsevier.com/locate/oceaneng

Long-term setup of a displacement pile in clay: An analytical framework

Weibing Gong a, Lin Li a, *, Sheng Zhang b, c, Jingpei Li a
a
Department of Geotechnical Engineering, Tongji University, Shanghai 200092, China
b
National Engineering Laboratory for High-Speed-Railway Construction, Central South University, Changsha 410083, China
c
School of Civil Engineering, Central South University, Changsha 410083, China

A R T I C L E I N F O A B S T R A C T

Keywords: This paper integrates the effects of pile installation, reconsolidation, pile loading and soil ageing into an
Setup analytical framework to predict the long-term setup of a displacement pile in clay. The increase of mean effective
Reconsolidation stress and the decrease of void ratio of surrounding clay during reconsolidation and soil ageing are determined to
Soil ageing
quantify the changes in undrained shear strength and shear modulus, which are the two key soil parameters
Displacement pile
Overconsolidation ratio
controlling load carrying behaviours of piles. The bearing capacity factors and load-transfer curves are developed
Secondary compression index by incorporating the developed soil parameters. The analytical framework is then proposed and validated by
predicting a well-documented pile field test. Good agreements between the analytical results and the measured
data demonstrate that the pile setup can be reasonably predicted by the proposed framework. Comprehensive
parametric studies are conducted to investigate the influences of overconsolidation ratio, secondary compression
index and pile loading on the pile setup. The results indicate that a displacement pile in clay with a smaller
overconsolidation ratio or a larger secondary compression index, or subjected to a larger load, exhibits a larger
long-term bearing capacity.

1. Introduction piles. The semilogarithmic empirical relationship proposed by Skov and

Denver (1988) is one of the milestones of these empirical equations.
An interesting phenomenon that the axial bearing capacity of a However, these empirical equations had limitations because they were
displacement pile in clay increases with the elapse of time after its derived based on the field tests in the specific geological conditions. On
installation has been recognized by many geotechnical researchers and the other hand, field tests are expensive and time-consuming, especially
practitioners, which is well denoted as pile setup. Pile setup is mainly when the pile setup is taken into consideration. Because such in­
caused by recovery of strength and stiffness of surrounding clay due to vestigations require the field tests lasting for several years or even de­
the reconsolidation (e.g., Randolph et al., 1979; Ng et al., 2012) and soil cades, which make the tests quite difficult to realize.
ageing (e.g., Tavenas and Audy, 1972; Schmertmann, 1991). Incorpo­ The finite element simulation, apart from field tests, has been served
rating setup into the design of displacement piles can undoubtedly help as another approach to explore the setup of a displacement pile in clay
save construction costs through reducing the number or dimension of (e.g., Titi and Wathugala, 1999; Whittle and Sutabutr, 1999; Abu-­
required piles. In addition, accurate prediction of pile setup is of great Farsakh et al., 2015), since a variety of complicated factors, such as pile
significance to estimate the load carrying behaviour of existing installation, reconsolidation as well as soil ageing, can be incorporated
displacement piles and to further utilize them to construct pile foun­ into the modeling process. However, some of the finite element simu­
dation via being mixed with new piles for rehabilitating or constructing lations utilized the soil properties obtained from laboratory tests on
buildings. Up to now, field tests have been widely conducted by re­ undisturbed soil samples, while significant changes in mechanical
searchers to explore the setup of a displacement pile in clay (e.g., Konrad properties of surrounding clay have been induced due to the great
and Roy, 1987; Bullock et al., 2005a,b; Ng et al., 2012; Chen et al., 2014, disturbance in the soil structure caused by pile installation. Therefore,
Haque et al., 2014). From the well-documented results of field tests, these simulations are unable to predict the actual time-dependent axial
many empirical equations were developed to predict the bearing capacity of displacement piles. In addition, the finite element
post-installation increase in the axial bearing capacity of displacement simulations are limited in their wide applicability, since one finite

* Corresponding author.
E-mail address: [email protected] (L. Li).

https://doi.org/10.1016/j.oceaneng.2020.108143
Received 23 April 2019; Received in revised form 22 June 2020; Accepted 20 September 2020
Available online 1 October 2020
0029-8018/© 2020 Elsevier Ltd. All rights reserved.
W. Gong et al. Ocean Engineering 218 (2020) 108143

element model can just serve for one geotechnical engineering project. If
different projects need to be simulated, the finite element model may
need to be rebuilt from scratch. As a result, it is difficult to build a
general framework through finite element simulations to investigate the
setup of displacement piles with different dimensions under different
geological conditions.
The objective of this paper is to present a general analytical frame­
work to predict the long-term setup of a displacement pile in clay. The
changes occurring in the two key soil parameters, undrained shear
strength and shear modulus, of surrounding clay at different phases,
including installation, reconsolidation, pile loading as well as soil
ageing, are quantified and incorporated into the proposed analytical
framework. The installation effect is modeled by an anisotropic cavity
expansion and the generated excess pore water pressure immediately
after pile installation is determined based on the solution of cavity
expansion. The radial consolidation theory with a variable coefficient of
consolidation is employed to model the reconsolidation of surrounding
clay. The soil ageing is represented by means of defining the quasi-
overconsolidation ratio. These effects are then integrated and incorpo­ Fig. 1. Reconsolidation and ageing of surrounding clay.
rated into the undrained shear strength and shear modulus, based on 2.2. Pile installation and dissipation of excess pore water pressure
which the bearing capacity factors and load transfer curves are devel­
oped to predict the pile setup. The proposed framework is validated by The surrounding soil is displaced predominantly outward during pile
predicting a well-documented pile field test, which was carried out for installation. The strain field generated by pile installation in the sur­
exploring the long-term loading carrying behaviour of a displacement rounding clay resembles the strain field determined from the cavity
pile. Parametric studies are conducted to investigate the effects of expansion theory, the rationality of which has been verified by Randolph
overconsolidation ratio (OCR), secondary compression index and et al. (1979) and Randolph (2003). Besides, the cavity expansion theory can
magnitude of pile loading on the long-term pile setup. provide a closed-form solution for predicting the radial stress and the excess
pore water pressure in the surrounding clay. Therefore, an anisotropic
2. Changes in mechanical properties of surrounding clay cavity expansion solution (Li et al., 2016), which employs the K0 -based
anisotropic modified Cam-clay model, is adopted to model the process of
2.1. Definitions of significant soil parameters pile installation. The notation K0 denotes the coefficient of earth pressure at
rest. The reason for adopting the anisotropic solution of cavity expansion is
The undrained shear strength of in-situ clay, su0 , under a conven­ that natural clays generally consolidate in the K0 condition and show the
tional triaxial compression test can be defined as (Wood, 1990) initial stress anisotropy and the initial stress-induced anisotropy and the
1
(
OCR
)Λ adopted anisotropic solution can well capture these two anisotropies.
(1) Following the solution of cavity expansion in K0 -consolidated clay pro­
′
su0 = Mp0
2 2
posed by Li et al. (2016), the excess pore water pressure induced by pile
where OCR is the overconsolidation ratio; p0 is the initial mean effective
′ installation can be given as
{ [√̅̅̅ ] }
stress; M is the slope of critical state line, equal to 6sinϕ / (3 − sinϕ ) ; [ ( )Λ ]
′ ′
K0 OCR 1 ′ 1 3G0 (rp )2
+ √̅̅̅p0 η*y + χ ln ′ *
′
u0 (r) = p0 − − 1
ϕ is the effective internal friction angle; and Λ represents the plastic
′
2K0 + 1 2 3 2 p0 ηy r
volumetric strain ratio, defined as Λ = 1 − κ/λ; and κ and λ represent the √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
2
slopes of swelling and compression lines in the e- lnp plane,
′ 4qf − 3χ 2

±
respectively. 6
The shear modulus of in-situ clay can be defined as (Sun et al., 2004) (3)
′ ′

G=
3(1 − 2v )υp
(2) where η*y is the relative stress ratio at the boundary of plastic zone
2(1 + v′ )κ developed around the pile, equal to
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
[ ]̅
where v is the effective Poisson’s ratio; and υ( = 1 + e) is the specific
′

(OCR − 1) M2 − |3(1 − K0 )/(2K0 + 1)|2

volume.
The pile setup mainly results from the recovery in strength and √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
√̅̅̅̅
stiffness of surrounding clay, which is caused by the effects of reconso­ ; the parameter χ is defined as pf [4M2 − 9|2(1 − K0 )/(2K0 + 1)|2 ]/ 3;
′

lidation and soil ageing. Although both effects simultaneously rp is the pile radius; and pf and qf are the mean effective stress and the
′

contribute to the pile setup, it is unnecessary to consider both the effects deviator stress at the critical state, respectively, which can be given as
together in the same phase. Because during reconsolidation, the dissi­
( )Λ
pation of excess pore water pressure dominates the recovery of strength OCR
(4)
′ ′
pf = p 0
and stiffness of surrounding clay while the effect of soil ageing is 2
negligible (Augustesen, 2006; Bullock et al., 2005). During soil ageing,
( )Λ
the situation is reversed. Therefore, it can be assumed that the soil OCR
(5)
′
qf = Mp0
ageing occurs primarily after reconsolidation (see Fig. 1) to simplify the 2
analytical analysis in this study. The similar assumption has also been
suggested by other researchers (e.g., Ladd et al., 1977; Mesri and Choi, After pile installation, the induced excess pore water pressure mainly
1985; Augustesen, 2006). dissipates in the radial direction, which is depicted in Fig. 2. This
dissipation process can be properly modeled by the classical Terzaghi’s
theory of consolidation. The governing equation can be written as

2
W. Gong et al. Ocean Engineering 218 (2020) 108143

critical state, pf ,.with the dissipation of excess pore water pressure,

′

which leads to the increase in undrained shear strength of surrounding

clay. Because the relaxation effect exists in the process of reconsolida­
tion (Randolph and Wroth, 1979a), the excess pore water pressure is
unable to fully transform into the increase of mean effective stress.
Hence, an effective stress transfer parameter, ξ, proposed by Randolph
and Wroth (1979a), is adopted here to consider the relaxation effect.
After taking the relaxation effect into consideration, the mean effective
stress of surrounding clay during reconsolidation can be calculated
based on the principle of effective stress as

(11)
′ ′
p (t) = pf + ξ[u0 − u(t)]

where
′ [ ]
p0 3OCR 5(1 + 2K0 ) u0
ξ= ln 1 + ′ (12)
u0 5(1 + 2K0 ) 3OCR p0

Due to the rearrangement of remolded soil particles, the surrounding

clay also regains its strength with time after pile installation. This phe­
nomenon is referred to as thixotropy (Abu-Farsakh et al., 2015) and
occurs no matter whether the surrounding clay is subjected to recon­
Fig. 2. Dissipation of installation-induced excess pore water pressure.
solidation. To consider the thixotropic effect, a time-dependent param­
(
1
)
∂e k 1 ∂
(
∂u
) eter, β(t), is introduced and defined as (Abu-Farsakh et al., 2015)
= r (6)
1 + e0 ∂t γw r ∂r ∂t β(t) = β(∞) − [β(∞) − β(0)]e− t/τ
(13)
where e0 is the initial void ratio; k is the permeability coefficient; and γw where β(∞) represents that the surrounding clay regains its strength
is the unit weight of water. from the disturbed state for a quite long time after pile installation and
The general solution of Eq. (6) can be obtained based on the variable its value can be taken as 1; β(0) is a reduction parameter denoting the
separation method as follows (Gong et al., 2017) stress state of surrounding clay immediately after pile installation, equal
( )
∑∞ (
J1 λn rp
) to 0.75 (Abu-Farsakh et al., 2015); and τ is a time constant, assumed as
(7)
2
u1j (r, t) = An J0 (λn r) − (
Y1 λn rp
)Y0 (λn r) e− λn Ch1j t the time for 90% dissipation of excess pore water pressure.
The surrounding clay suffers from the strong remolding effects dur­
n=1

where J0 and J1 are the first-kind Bessel functions of first order and ing pile installation, including the collapse of soil structure and the
second order, respectively; Y0 and Y1 are the second-kind Bessel func­ erasing of stress history. The remolding effects lead to the surrounding
tions of first order and second order, respectively; j represents the jth clay exhibiting the mechanical properties similar to the normally
small interval of total dissipation time; λn are the eigenvalues of Bessel consolidated clay after pile installation (Randolph and Wroth, 1981; Yao
et al., 2009). Thus, the overconsolidation ratio of surrounding clay can
function; and Ch is a variable coefficient of consolidation, which can be
be taken as unity. Correspondingly, the undrained shear strength of
expressed as
surrounding clay after pile installation can be expressed as
Ch0 ′
Ch = ′ α (p (t))
α
(8) 1
( )Λ
1
(14)
′
(p0 ) su = Mp
2 2
where Ch0 is the initial coefficient of consolidation for radial horizontal Through substituting Eqs. (11) and (13) into Eq. (14), the time-
drainage; and α is a dimensionless factor. The coefficient, An , in Eq. (7) dependent undrained shear strength of surrounding clay after pile
can be obtained based on the orthogonality of Bessel function as installation, which appropriately incorporates the effects of pile instal­
∫ ry
u0 (r)V0 (λn r)rdr lation, reconsolidation and thixotropy, can be determined as follows
(9)
r
An = p ∫ r 2 ( )Λ+1 [ ( )Λ ]
V (λn r)rdr 1 OCR
(15)
rp 0 ′
su (t) = M p0 + ξΔu(t) β(t)
2 2
where ry is the radius of plastic zone developed around the displacement
In the e- logp plane, the relationship between the void ratio and the
′
pile; and the expression of V0 (λn r) is
( ) mean effective stress can be written as
J1 λn rp ( ′)
V0 (λn r) = J0 (λn r) − ( )Y0 (λn r) (10) p
Y1 λn rp e = e0 − Cc log ′ (16)
p0
During pile installation, a critical state zone is developed around the
displacement pile, where the mean effective stress of surrounding clay where Cc is the compression index and represents the slope of
remains constant and always equals the mean effective stress at the compression curve in the e- logp plane.
′

critical state, pf . After pile installation, the mean effective stress of Substituting Eq. (11) into Eq. (2), along with the substitution of Eqs.
′

surrounding clay, p (t), increases from the mean effective stress at the
′
(4), (7) and (16), the time-dependent shear modulus of surrounding clay
after pile installation can be expressed as

3
W. Gong et al. Ocean Engineering 218 (2020) 108143

{ [( )Λ ]}[ ( )Λ ]
′ ′ ′
3(1 − 2v ) 1 + e0 − Cc log OCR
2
+ ξΔu(t) /p0 p0 OCR
2
+ ξΔu(t)
G(t) = (17)
2(1 + v′ )κ

To better describe the quasi-overconsolidated behaviour of clay, a

parameter called quasi-overconsolidation ratio, R, is defined as the ratio
of quasi-preconsolidation pressure to the initial mean effective stress,
pcq /p0 . According to the relationship in the e -lnp plane, the decrease of
′ ′ ′

2.3. Soil ageing

void ratio can be also determined as
As shown in Fig. 3, the loading of the displacement pile gives rise to pcq
′

the decrease in the void ratio of surrounding clay (from point A to point Δe = e − e0 = − λln ′ (19)
p (tre )
B). In addition, the void ratio of surrounding clay decreases with the
elapse of time due to the effects of soil ageing (from point B to point C), Substituting Eq. (19) into Eq. (18), the quasi-overconsolidation ratio
although the mean effective stress of surrounding clay remains constant. can be obtained as
Both the decreases in void ratio lead to the surrounding clay showing [ ( )]
Cc Cα t
more stable structure and exhibiting the quasi-overconsolidated R(t) = exp log(LR) + log , t > tre (20)
λ λ tre
behaviour. To account for this quasi-overconsolidated behaviour, the
secondary relations built based on the Bjerrum’s theory of time- Combining Eqs. (15) and (20) as well as combining (17) and Eq. (20),
dependent compression (Bjerrum, 1967) and developed by Garlanger the time-dependent undrained shear strength and shear modulus of
(1972) is adopted to describe the decrease of void ratio resulting from surrounding clay, which incorporates the effects of soil ageing, can be
the pile loading and the soil ageing. The expression can be specifically given as follows
written as ( )Λ+1 [ ( )Λ ]
1 OCR
( ) (21)
′

t su (t) = MR(t) p0 + ξΔu(tre ) β(t), t > tre

e(t) = e0 − Cc log(LR) − Cα log , t > tre (18) 2 2
tre

[ ( )Λ ][ ( )]
′ ′
3(1 − 2v )R(t) p0 OCR
2
+ ξΔu(tre ) 1 + e0 − Cc log(LR) − Cαe log ttr
G(t) = , t > tre (22)
2(1 + v′ )κ

where Cα is the secondary compression index; tre is the reference time,

which here should be denoted as the end time of the dissipation of excess
pore water pressure; and LR ​ (≥ 1) is defined as the loading parameter
and equal to σ z,p /σ z (tre ), where σ z,p is the vertical effective stress of
′ ′ ′

3. Time-dependent axial bearing capacity

surrounding clay when an axial load is applied to the pile head. The
loading parameter can well reflect the magnitude of pile loading, as a
Fig. 4 shows the stress state of a soil element adjacent to the pile shaft
larger loading parameter corresponds to a larger load applied to the pile
when the pile is subjected to an axial load. Randolph and Wroth (1981)
head.
indicates that the shearing state of a soil element adjacent to the pile
shaft can be equivalent to that in a simple shear test. Hence, the unit
shaft resistance of a displacement pile, fs , can be developed in terms of
the undrained shear strength under the plane strain condition, spu , as (Li
et al., 2017a)

(23)
′
fs = τyx = spu cosϕ

Note that the undrained shear strength defined by Eq. (1) is obtained
under conventional triaxial compression tests and is different from the
undrained shear strength under the plain strain condition, because of its
intrinsic three-dimensional mechanical properties. Therefore, the stress-
transformed method combined with the SMP criterion presented by
Matsuoka et al. (1999) is adopted to develop the relationship between
the undrained shear strength under the plane strain condition and the
undrained shear strength obtained from the three-dimensional
compression test. The corresponding relationship between them can
be derived as (Li et al., 2017a)

Fig. 3. Development of quasi-preconsolidation pressure.

4
W. Gong et al. Ocean Engineering 218 (2020) 108143

for a small part of the total bearing capacity (Bullock et al., 2005a, b). In
the framework of the total stress method, the unit pile end resistance can
be easily estimated as
qb = Nc su0 (31)

where Nc is the pile end bearing capacity factor and can be empirically
taken as 9 (Wardle et al., 1992).
Due to the effects of dissipation of excess pore water pressure and soil
ageing, the unit pile end resistance of the displacement pile also exhibits
the time-dependent effects. Therefore, it can be expressed as
qb (t) = Nc (t)su0 (32)
Substituting Eqs. (1) and (15) into Eq. (32), the time-dependent pile
end bearing capacity factor, which considers the effects of reconsoli­
dation and thixotropy, can be written as
′
p (t) Nc
Nc (t) = ′ β(t) (33)
p0 OCRΛ
The time-dependent pile end bearing capacity factor, considering the
Fig. 4. Stress states of a soil element adjacent to the pile shaft and in a simple
effects of soil ageing and thixotropy, can be obtained by substituting Eq.
shear test.
(21) into Eq. (32) as follows
3sinψ f
(24)
′
spu = ( )su
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ p (tre ) R(t)Nc
M 2 + sin2 ψ f Nc (t) = ′ β(t), t > tre (34)
p0 OCRΛ
For a displacement pile in clay, the total axial bearing capacity
where ψ f is the stress-transformed parameter, written as
consists of the shaft bearing capacity and the end bearing capacity. Thus,
√̅̅̅
2M the total axial bearing capacity of a displacement pile in clay, which
sinψ f = √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ (25) considers the effects of both reconsolidation and soil ageing, can be
9 + 3M
estimated by the following equation
Substituting Eqs. (1), (15) and (24) into Eq. (23) yields the unit time-
dependent shaft resistance as ∫h
′
Q(t) = Qs (t) + Qb (t) = 2πrp fs (t)dz + πrp2 qb (t) (35)
′
3sinψ f cosϕ p (t) su0
fs (t) = √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
( ) ′ Λ
β(t) (26) 0

M 2 + sin2 ψ f p0 OCR
4. Time-dependent load-displacement response
The unit shaft resistance of a displacement pile, fs (t), which in­
corporates the time-dependent effects, can also be easily evaluated The surrounding clay exhibits the time-dependent undrained shear
based on the total stress method as follows strength and shear modulus after pile installation because of the effects
of reconsolidation and soil ageing. The load-displacement response of an
fs (t) = α(t)su0 (27)
axially loaded displacement significantly relies on these two soil pa­
where α(t) is the time-dependent pile shaft bearing capacity factor. It is rameters (Li et al., 2017 a; b). Thus, a displacement pile in clay exhib­
worth mentioning that in the total stress method, the pile interface is iting evidently time-dependent load-displacement response. This
deemed to be completely rough and without slippage.
Comparing Eq. (26) with Eq. (27), the time-dependent pile shaft
bearing capacity factor, which considers the effects of reconsolidation
and thixotropy, can be given as
′
3sinψ f p (t) β(t)
α(t) = √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
( ) ′ Λ
(28)
M 2 + sin2 ψ f p0 OCR

Substituting Eqs. (1), (21) and (24) into Eq. (23) and comparing with
Eq. (27), the time-dependent pile shaft bearing capacity factors, incor­
porating the effects of soil ageing and thixotropy, can be expressed as
′
3sinψ f R(t)p (tre ) β(t)
α(t) = √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
( ) ′ , t > tre (29)
M 2 + sin2 ψ f p0 OCRΛ

Conducting the integral along the pile shaft, the shaft bearing ca­
pacity of a displacement pile in clay, which properly incorporates the
time-dependent effects, can be determined as
∫h
Qs (t) = 2πrp fs (t)dz (30)
0

The end bearing capacity of a displacement pile in clay only accounts Fig. 5. Hyperbolic nonlinear relationship between shaft shear stress and pile-
soil relative displacement.

5
W. Gong et al. Ocean Engineering 218 (2020) 108143

time-dependent response can be explored by incorporating the

time-dependent soil mechanical properties and the correspondingly
developed time-dependent pile bearing capacity factors into the model
parameters of load-transfer curves. Up to now, geotechnical researchers
have extensively investigated the relationship between the shaft shear
stressτs,z , and the pile-soil relative displacement, Ds,z . The results of both
field and laboratory tests indicate that there is a nonlinear relationship
between the shaft shear stress and the pile-soil relative displacement and
this nonlinear relationship can be properly described by an exponential
function or a hyperbolic function (Zhang and Zhang, 2012; Wang et al.,
2012). Therefore, a hyperbolic function (see Fig. 5) is adopted to
simulate the nonlinear behaviour of the pile-soil interface, the expres­
sion of which can be written as
Ds,z
τs,z = (36)
as + bs Ds,z

where as and bs represent the initial stiffness and the reciprocal of the
asymptote of the pile shaft load-transfer curve, respectively.
Since the time-dependent load-displacement response is one of the
focuses of this study, Eq. (36) can be rewritten as
Ds,z
τs,z (t) = (37)
as (t) + bs (t)Ds,z

The value of as (t)in the above equation can be determined as (Ran­

dolph and Wroth, 1979b)
( )
rp rm
as (t) = ln (38)
G(t) rp

where rm is the limiting influence radius and equal to 2.5L(1 − v),

beyond which the shaft shear stress induced by pile installation becomes
negligible.
The asymptote of the pile shaft load-transfer curve is equal to τsu (t),
and bs (t) is the reciprocal of the asymptote. Thus, the coefficient, bs (t),
can be determined as
1 1
bs (t) = = (39)
τsu (t) fs (t)
The nonlinear load-displacement relationship developed at the pile
base is also assumed to conform to a hyperbolic function. Therefore, the Fig. 6. Illustration of single pile notation.
relationship between the unit pile end resistance, qb , and the pile end
(1) To determine the eigenvalues of Bessel function, λn , in Eq. (7), it
settlement, Db , can be expressed as
first needs to determine the first-zero point and period of the
Db following equation
qb (t) = (40)
ab (t) + bb (t)Db ( ) ( ) ( ) ( )
J0 xry Y1 xrp − Y0 xry J1 xrp = 0 (43)
where ab (t)is the initial stiffness of the pile end load-transfer curve,
which can be derived as (Randolph and Wroth, 1979b) (2) Utilizing Eq. (43) gives the first 50 values of zero points (i.e. λn ).
π(1 − v)rp Note that although Eq. (43) owns infinite zero points because the
ab (t) = (41) Bessel functions are oscillating functions, the first 50 values can
4G(t)
give sufficient accuracy. In addition, equal roots may be obtained
and bb (t) is the ultimate pile end resistance, given by when calculating Eq. (43), however only one of these equal roots
should be remained while others should be deleted;
1 1
bb (t) = = (42) (3) The coefficient, An , is determined by substituting the eigenvalues
qbu (t) Nc (t)su0
obtained from Step (2) into Eq. (9);
(4) The excess pore water pressure at time t, i.e.u(t) , is determined
5. Computational scheme from Eq. (7);
(5) With the elapse of time, the excess pore water pressure dissipates.
The stress states of soil elements adjacent to the displacement pile are When u(t)/u0 ≤ 10%, the time constant, τ, in Eq. (13) is deter­
different along the pile shaft, which requires the displacement pile being
mined. When u(t) ≤ 10− 2 , it can be considered that the recon­
divided into very small segments to guarantee that the stress along the
solidation is completed and this corresponding moment is regard
pile segment can be assumed to be equal. Hence, as shown in Fig. 6, the
to the reference time, tre , in Eq. (20);
displacement pile is divided into nsmall segments with length l. A flow
(6) The mean effective stress of surrounding clay during reconsoli­
chart for predicting the time-dependent bearing capacity of a displace­
dation is determined from Eq. (11) and the quasi-
ment pile in clay is provided in Fig. 7 and described below.
overconsolidation ratio during soil ageing is determined from
Eq. (20);

6
W. Gong et al. Ocean Engineering 218 (2020) 108143

Fig. 7. Illustration of calculation process of time-dependent bearing capacity of a displacement pile in clay.

(7) The bearing capacity factors α(t) and Nc (t) during reconsolidation
Qtn (t) + 3Qb (t) l
and soil ageing are obtained based on Eqs. (28), (29), (33) and Ducn = Dbn + (45)
4 2Ep Ap
(34), respectively;
(8) The time-dependent end bearing capacity and the time-depend
shaft bearing capacity are summed to obtain the total time- (14) A very small tolerance (i.e. 10− 6 m) is specified between the
dependent axial bearing capacity of the displacement pile in updated midpoint displacement, Ducn , and the first assumed
clay from (35). ⃒ ⃒
midpoint displacement, Dcn . If ⃒Ducn − Dcn ⃒ > 10− 6 , the updated
midpoint displacement, Ducn , needs to be set as the new value of
Note that the first 6 steps of the procedure for exploring the time- ⃒
Dcn , and Steps (10)–(14) need to be repeated until ⃒Ducn −
dependent load-displacement response of a displacement pile in clay ⃒
are the same as those of investigating the time-dependent bearing ca­ Dcn ⃒ ≤ 10− 6 ;
pacity. Hence, the flow chart and the specific procedures for predicting (15) The displacement and the load at the top of pile segment nare
the time-dependent load-displacement response directly starts from Step calculated as
(7). The flow chart can be referred to Fig. 8 and the procedures are Dtn = Dbn + Ducn (46)
described in the following.
Qtn (t) = Qb (t) + πlτusn (t) (47)
(7) The shear modulus of the surrounding clay at a given time t,
G(t)is obtained from Eq. (17) by substituting the corresponding
mean effective stress into it when t < tre or from Eq. (22) by (16) Steps (10)–(15) are repeated from pile segment nto pile segment 1
substituting the corresponding quasi-overconsolidation ratio and to obtain the load-displacement response developed at the pile
the mean effective stress into it when t > tre . The unit shaft head;
resistance, fs (t), and the unit pile end resistance, qb (t), at a given (17) Steps (8)–(16) are repeated with an assumed pile end displace­
time t, are determined from Eqs. (27) and (32) with the substi­ ment from zero to a desired value to obtain the load-displacement
tution of corresponding pile bearing capacity factors, curve of a displacement pile in clay at a given time after pile
respectively; installation.
(8) A small pile end displacement, Dbn , is assumed;
(9) The pile end resistance, qb (t), is obtained by substituting Dbn into 6. Validation
Eq. (40), along with the substitution of Gb (t)and Nc (t)into Eqs.
(41) and (42). The pile end load, Qb (t) is equal to πrp2 qb (t); 6.1. Field test
(10) In the first trial, the displacement, Dcn , at the midpoint of pile
segment nis assumed to equal the pile end displacement, Dbn , i.e. The field test selected for validation was conducted by Konrad and
Dcn = Dbn ; Roy (1987) at the Saint-Alban test site, which is located about 80 km
(11) The shear stress along the shaft of pile segment n, τsn (t), is ob­ west of Quebec City at the northern fringe of the Saint-Laurent lowlands.
tained by substituting Dcn and the corresponding as (t)and The reason for selecting this field test instead of other available field
bs (t)into Eq. (37); tests is that the authors not only well rerecorded the test results of the
(12) The load at the top of pile segment n, Qtn (t), is calculated as setup of displacement piles, but also listed many soil parameters that are
vital for the proposed analytical analysis. The soil at the test site is the St
Qtn (t) = Qb (t) + πlτsn (t) (44) Alban clay, which is slightly overconsolidated and with an OCR in the
order of 2.1. The initial coefficient of radial consolidation and the sec­
(13) The updated displacement at the midpoint of pile segment n, Ducn , ondary compression index were not measured on St-Alban clay samples
is calculated as during field test. Nonetheless, the initial coefficient of radial consoli­
dation can be taken about 0.0008 m2/h to represent the destructed clay
in a slightly overconsolidated state (Leroueil et al., 1979). In addition,
according to the literature (Yin, 2007), the secondary compression index

7
W. Gong et al. Ocean Engineering 218 (2020) 108143

Fig. 8. Illustration of calculation process of time-dependent load-displacement response of a displacement pile in clay.

8
W. Gong et al. Ocean Engineering 218 (2020) 108143

Table 1 displacement pile in clay, the fifty-day variation of excess pore water
Summary of soil properties of St Alban clay. pressure, shaft bearing capacity and total axial bearing capacity are
Soil parameter Value predicted for comparison in Fig. 9. The dissipation of excess pore water
3 pressure chosen here was measured at the depth of 3.05 m by the pore
Effective unit weight, γ’ (kN/m ) 7.42
Effective Poisson’s ratio, v’ 0.3 pressure transducer equipped in the tested pile. From the figure, it can
Effective internal friction angle, ϕ’ (◦ ) 28 be found that the predicted dissipation of excess pore water pressure and
Slope of swelling line, κ 0.0287 shaft bearing capacity show good agreements with the measured values.
Initial void ratio, e0 1.917 The total axial bearing capacity is overestimated about 20% at 4 days
after pile installation. This may be because the actual thixotropy is
stronger than the thixotropy predicted by Eq. (13), which makes the
undrained shear strength at the early stage slightly overestimated.
However, it still shows a satisfactory agreement between the predicted
and field-measured results of total axial bearing capacity as time goes
on. In addition, it can be seen that the reconsolidation lasts for about 41
days since the completion of pile installation. After that, the effects of
soil ageing mainly contribute to the increase of bearing capacity of the
displacement pile.
Fig. 10 shows the complete predicted variation of total axial bearing
capacity and shaft bearing capacity of the displacement pile at the Saint-
Alban test site for 2 years. It can be found that both the predicted shaft
bearing capacity and total axial bearing capacity at 2 years agree very
well with the measured values. It is worth noting that the increase of pile
bearing capacity caused by the effect of soil ageing seems not very
obvious, compared with that resulting from the effect of reconsolidation.
This is because the rate of soil ageing is much slower than the rate of
reconsolidation. The increase of soil strength caused by soil aging is
much smaller than that resulted from reconsolidation in a short time
period. Correspondingly, the increase of bearing capacity arising from
Fig. 9. Comparison of total bearing capacity, shaft bearing capacity and soil ageing in 2 years is comparatively small. Nonetheless, when time for
dissipation of excess pore water pressure between predicted and measured re­ soil ageing is long enough, the pile setup caused by the effects of soil
sults for 50 days. ageing will be much more obvious. Therefore, it suggests that the in­
crease in pile bearing capacity due to the reconsolidation is important in
the economy for new structures, while the effect of soil ageing is more
significant for rehabilitating or building new structures utilizing existing
piles from previous structures.

6.3. Load-displacement response

The validation of the proposed framework is examined again by

predicting the time-dependent load-displacement response of Pile A
measured on the 4th day and 8th day after pile installation. It can be
seen from Fig. 11 that although a softer axial response is predicted for
the displacement pile on the 4th day after pile installation, both the load-
displacement curves still agree well with the measured results, espe­
cially the curve predicted on 8th day. Therefore, satisfactory agreements
are demonstrated between the predicted and measured load-

Fig. 10. Comparison of total axial bearing capacity and shaft bearing capacity
between predicted and measured results for 2 years.

of clay with an OCR larger than 2 can be empirically taken as 0.001. The
other soil parameters required in the analysis are tabulated in Table 1.
The test pile chosen for comparison is the Pile A in the field test. The pile
is a circular steel pipe with 8 mm thick wall, whose length and diameter
are 7.6 m and 0.22 m, respectively. The elastic modulus of the pile is
about 210 GPa. Its bearing capacity was measured carefully at time
intervals of 4 days, 8 days, 20 days, 33 days and 2 years after pile
installation. Further details related to this field test can refer to Konrad
and Roy (1987).

6.2. Bearing capacity

Fig. 11. Comparison of predicted and measured load-displacement responses
To investigate the effect of reconsolidation on the setup of a at different time.

9
W. Gong et al. Ocean Engineering 218 (2020) 108143

Fig. 12. Effects of different OCRs on pile end bearing capacity factors (a) Fig. 13. Effects of different OCRs on pile shaft bearing capacity factors (a)
mainly during reconsolidation; (b) during soil ageing. mainly during reconsolidation; (b) during soil ageing.

displacement responses. It is worth pointing out that the reason for the ′
(48)
′
K0 = OCRsinϕ (1 − sinϕ )
softer axial response may be the slight underestimation of the shear
modulus of the surrounding clay. It can also be interestingly found that To better illustrate the effects of stress history on pile bearing ca­
the stiffness of the displacement pile in clay increases with the elapse of pacity, the results are separately illustrated into Figs. 12 and 13 ac­
time. This observed time-dependent property should be taken into ac­ cording to the two different phases of pile setup, i.e. reconsolidation and
count when designing displacement piles in clay. soil ageing. From Figs. (12a) and Figs. (13a), it can be seen that during
reconsolidation the bearing capacity factors of a displacement pile in
7. Results and discussions

7.1. Effects of overconsolidation ratio

The stress history significantly affects the clay behaviours, like

compression and consolidation, which certainly exerts influences on the
bearing capacity of a displacement pile in clay. Compared with the
undrained shear strength that was adopted by some researchers (e.g.
Weltman and Healy, 1978; Kulhawy and Jackson, 1989), the OCR is
independent of other significant soil parameters, such as internal friction
angle, volumetric strain ratio, unit weight, etc. and thus it is more
rational to utilize the OCR to explore the effects of stress history on pile
bearing capacity. Here, three different OCRs (i.e. 1, 2 and 5), which
represent the normally consolidated clay, the slightly overconsolidated
clay, and the moderately overconsolidated clay, are adopted to inves­
tigate the effects of stress history on pile bearing capacity. The ratio of
pile length to pile diameter is 35 and the soil parameters adopted for
analysis is the same as those listed in Table 1. The values of coefficients
of earth pressure at rest of clays with different OCRs can be evaluated by
the following equation (Mayne and Kulhawy, 1982) Fig. 14. Effects of different secondary compression indexes on pile end bearing
capacity factors.

10
W. Gong et al. Ocean Engineering 218 (2020) 108143

Fig. 15. Effects of different secondary compression indexes on pile shaft Fig. 17. Effects of different magnitudes of pile loadings on pile shaft bearing
bearing capacity factors. capacity factors.

clay with larger OCRs increase much faster than those of a displacement pile in clay. Therefore, it is quite meaningful to explore the effects of
pile in clay with smaller OCRs. Correspondingly, the bearing capacity of different secondary compression indexes on pile bearing capacity fac­
displacement piles in clay with larger OCRs increases much faster than tors. Note that the secondary compression index only affects the pile
piles in clay with smaller ones during reconsolidation. The end time of bearing capacity during soil ageing. Thus, only variations of pile bearing
pile setup caused by reconsolidation increases with the increase of OCR. capacity factors during soil ageing are shown in Figs. 14 and 15. Since
Hence, when designing new displacement piles in clays with different the secondary compression index of the clay in the test site is equal to
OCRs, special attention should be paid to the time of applying loads to 0.001, another three values of secondary compression index with the
the pile head with the aim of incorporating the full pile setup caused by same order of magnitudes, which are 0.003, 0.005 and 0.007, respec­
the reconsolidation into the design. It should also be noted in Figs. (12b) tively, are adopted here. It can be found that the values of both the pile
and (13b) that the displacement piles in clay with smaller OCRs own end bearing capacity factor and the pile shaft bearing capacity factor of a
larger bearing capacity factors in the period of soil ageing, as the excess displacement pile in clay with larger secondary compression indexes are
pore water pressure totally dissipates and the clay has a stronger soil also larger than those of a displacement pile in clay with smaller sec­
aging effect. The larger bearing capacity factors demonstrate that the ondary compression indexes. In addition, both the pile bearing capacity
displacement piles in clay with smaller OCRs have larger long-term factors increase much faster when the surrounding clay processes the
bearing capacities. larger secondary compression index. This is because the larger un­
drained shear strength is developed for the clay with a larger secondary
compression index during soil ageing and the undrained shear strength
7.2. Effects of secondary compression indexes is the key soil parameter controlling the pile bearing capacity. The
phenomena indicate that a displacement pile in clay with a larger sec­
The rate of soil ageing is affected by the secondary compression ondary compression index will develop a larger long-term bearing
index. The larger compression indexes generally correspond to the capacity.
larger rates of soil ageing. Besides, it can be found from Eq. (20) that the
magnitude of quasi-overconsolidation ratio depends on the secondary
7.3. Effects of loadings
compression index and the quasi-overconsolidation ratio is a significant
parameter determining the long-term bearing capacity of a displacement
Figs. 16 and 17 illustrate the effects of different loads applied to the
pile head (reflected by loading parameter, LR) on pile bearing capacity
factors. The reason for this exploration is that the pile loading can lead to
the decrease of void ratio of surrounding clay and then can change the
mechanical properties of surrounding clay, which further affects the
bearing capacity of displacement piles in clay. Because it is assumed that
the time for pile loading is just at the end of dissipation of excess pore
water pressure, only the variation of pile bearing capacity factors during
soil ageing is shown in the figures. It can be observed that both the pile
end bearing capacity factor and the pile shaft bearing capacity factor
increase as the magnitude of the load applied to pile head increases,
which indicates that a larger long-term bearing capacity is developed for
the pile subjected to a larger load. This phenomenon can be attributed to
that a larger pile loading results in the larger decrease of void ratio and
then brings about stronger quasi-overconsolidated effects of surround­
ing clay. The stronger quasi-overconsolidated effects lead to the sur­
rounding clay developing larger quasi-overconsolidation ratios. As a
result, a larger undrained shear strength is developed for the sur­
rounding clay, which leads to the displacement pile processing larger
Fig. 16. Effects of different magnitudes of pile loadings on pile end bearing bearing capacity. Note that the statement that the pile subjected to a
capacity factors. larger load can result in a larger long-term bearing capacity may only

11
W. Gong et al. Ocean Engineering 218 (2020) 108143

hold true under the similar soil conditions as the field tests. the work reported in this paper.

8. Conclusions Acknowledgements

In this paper, an analytical framework has been presented to predict This study was financially supported by the National Natural Science
the setup of a displacement pile in clay, where the effects of pile Foundation of China (Grant No. 41772290).
installation, reconsolidation, pile loading and soil ageing were inte­
grated. The validity of the framework was examined by predicting the References
time-dependent bearing capacity and the time-dependent load-
displacement response of a displacement pile in a well-documented pile Abu-Farsakh, M., Rosti, F., Souri, A., 2015. Evaluating pile installation and subsequent
thixotropic and consolidation effects on setup by numerical simulation for full-scale
field test. The effects of overconsolidation ratio, secondary compression pile load tests. Can. Geotech. J. 52 (11), 1734–1746.
index and loading on pile setup were discussed in detail based on the Augustesen, A., 2006. The Effects of Time on Soil Behaviour and Pile Capacity. Thesis,
proposed framework. The proposed procedure can be applied to esti­ Aalborg University, Aalborg, Denmark.
Bjerrum, L., 1967. Engineering geology of Norwegian normally-consolidated marine
mate the setup of different types of piles, but for some types of piles it clays as related to settlements of buildings. Geotechnique 17 (2), 83–118.
will have some limitations. For example, for bored piles, only the soil Bullock, P.J., Schmertmann, J.H., McVay, M.C., Townsend, F.C., 2005a. Side shear setup.
ageing effects should be taken into consideration when exploring the I: test piles driven in Florida. J. Geotech. Geoenviron. Eng. 131 (3), 292–300.
Bullock, P.J., Schmertmann, J.H., McVay, M.C., Townsend, F.C., 2005b. Side shear setup.
setup and thus the proposed procedure only considering the effects of II: results from Florida test piles. J. Geotech. Geoenviron. Eng. 131 (3), 301–310.
soil ageing can be utilized. For driven piles, both the effects of recon­ Chen, Q., Haque, M.N., Abu-Farsakh, M., Fernandez, B.A., 2014. Field investigation of
solidation and the effects of soil ageing should be taken into account pile setup in mixed soil. Geotech. Test J. 37 (2), 268–281.
Garlanger, J.E., 1972. The consolidation of soils exhibiting creep under constant effective
when predicting the setup. Besides, the influences caused by the pile
stress. Geotechnique 22 (1), 71–78.
striking play an important role in driven pile setup and should be Gong, W., Li, J., Li, L., Zhang, S., 2017. Evolution of mechanical properties of soils
considered as well. Therefore, the procedure for predicting the load subsequent to a pile jacked in natural saturated clays. Ocean. Eng. 136, 209–217.
carrying behaviours of driven piles can be similar to the proposed one Haque, M.N., Chen, Q., Abu-Farsakh, M., Tsai, C., 2014. Effects of pile size on set-up
behavior of cohesive soils. In Geo-Congress 2014: geo-characterization and Modeling
for displacement piles, but the influences resulting from the pile striking for Sustainability, pp. 1743–1749.
should be integrated into the procedure for the sake of accuracy. The Konrad, J.M., Roy, M., 1987. Bearing capacity of friction piles in marine clay.
conclusions that can be drawn from this study are summarized in the Geotechnique 37 (2), 163–175.
Kulhawy, F.H., Jackson, C.S., 1989. Some observations on undrained side resistance of
following. drilled shafts. Current principles and practices, Foundation Engineering Congress,
ASCE, Reston, VA 1011–1025.
(1) Both the predicted pile setup and load-displacement response Ladd, C.C., 1977. Stress-deformation and Strength Characteristics, State of the Art
Report. Proc. 9th International Conference on Soil Mechanics and Foundation
show good agreements with the test results of a displacement pile Engineering, Tokyo, Japan, pp. 421–494.
in the field test. It well proves that the proposed analytical Leroueil, S., Tavenas, F., Brucy, F., La Rochelle, P., Roy, M., 1979. Behaviour of
framework can be effectively applied to evaluate the pile setup destructured natural clays. J. Geotech. Eng. Div. 105 (6), 759–778.
Li, L., Li, J., Sun, D.A., 2016. Anisotropically elasto-plastic solution to undrained
resulting from the effects of reconsolidation and soil ageing. cylindrical cavity expansion in K0-consolidated clay. Comput. Geotech. 73, 83–90.
(2) For a displacement pile in clay with a larger OCR, its bearing Li, L., Li, J., Sun, D.A., Gong, W., 2017a. Analysis of time-dependent bearing capacity of a
capacity after pile installation increases much faster than the driven pile in clayey soils by total stress method. Int. J. GeoMech. 17 (7), 10.1061/
(ASCE)GM. 1943-5622.0000860.
displacement pile in clay with a smaller OCR, due to the faster
Li, L., Li, J., Sun, D.A., Gong, W., 2017b. Semi-analytical approach for time-dependent
dissipation of installation-induced excess pore water pressure. load–settlement response of a jacked pile in clay strata. Can. Geotech. J. 54 (12),
However, its long-term bearing capacity is smaller than the 1682–1692.
displacement pile in clay with a smaller OCR. Matsuoka, H., Yao, Y., Sun, D., 1999. The Cam-clay models revised by the SMP criterion.
Soils Found. 39 (1), 81–95.
(3) The magnitude of secondary compression index affects the long- Mayne, P.W., Kulhawy, F.H., 1982. K0-OCR relationship in soil. J. Geotech. Eng. Div. 108
term bearing capacity of displacement piles in clay. Larger in­ (6), 851–872.
crease rate and magnitude of long-term bearing capacity are Mesri, G., Choi, Y.K., 1985. Settlement analysis of embankments on soft clays. Journal of
Geotechnical Engineering 111 (4), 441–464.
developed for a displacement pile in clay with a larger secondary Ng, K.W., Roling, M., AbdelSalam, S.S., Suleiman, M.T., Sritharan, S., 2012. Pile setup in
compression index. Besides, a larger load applied to the pile head cohesive soil. I: experimental investigation. J. Geotech. Geoenviron. Eng. 139 (2),
during pile loading also leads to the displacement pile processing 199–209.
Randolph, M.F., 2003. Science and empiricism in pile foundation design. Geotechnique
a larger long-term bearing capacity. 53 (10), 847–876.
(4) The parametric study focusing on the investigations of bearing Randolph, M.F., Carter, J.P., Wroth, C.P., 1979. Driven piles in clay—the effects of
capacity factors provides a feasible way for exploring the pile size installation and subsequent consolidation. Geotechnique 29 (4), 361–393.
Randolph, M.F., Wroth, C.P., 1979a. An analytical solution for the consolidation around
effect on setup just through incorporating the pile radius and the a driven pile. Int. J. Numer. Anal. Methods GeoMech. 3 (3), 217–229.
bearing capacity factors into Eq. (35). It can be easily found from Randolph, M.F., Wroth, C.P., 1979b. An analysis of the vertical deformation of pile
the equation that a pile with a larger size will experience more groups. Geotechnique 29 (4), 423–439.
Randolph, M.F., Wroth, C.P., 1981. Application of the failure state in undrained simple
setup, which is similar to the conclusion suggested by Haque et al.
shear to the shaft capacity of driven piles. Geotechnique 31 (1), 143–157.
(2014). Schmertmann, J.H., 1991. The mechanical aging of soils. Journal of Geotechnical
Engineering 117 (9), 1288–1330.
Skov, R., Denver, H., 1988. Time-dependence of bearing capacity of piles. Proc. In:
CRediT authorship contribution statement
Fellenius, B.H. (Ed.), Third International Conference on the Application of Stress-
Wave Theory to Piles, pp. 879–888. BiTech.
Weibing Gong: Software, Methodology, Data curation, Writing - Sun, D.A., Matsuoka, H., Yao, Y.P., Ishii, H., 2004. An anisotropic hardening elastoplastic
original draft. Lin Li: Conceptualization, Methodology, Software, model for clays and sands and its application to FE analysis. Comput. Geotech. 31
(1), 37–46.
Methodology, Writing - review & editing. Sheng Zhang: Conceptuali­ Tavenas, F., Audy, R., 1972. Limitations of the driving formulas for predicting the
zation, Supervision, Writing - review & editing. Jingpei Li: Conceptu­ bearing capacities of piles in sand. Can. Geotech. J. 9 (1), 47–62.
alization, Supervision, Funding acquisition, Writing - review & editing. Titi, H., Wathugala, G., 1999. Numerical procedure for predicting pile capacity-setup/
freeze. Transport. Res. Rec.: Journal of the Transportation Research Board 1663,
25–32.
Declaration of competing interest Whittle, A., Sutabutr, T., 1999. Prediction of pile setup in clay. Transport. Res. Rec.:
Journal of the Transportation Research Board 1663, 33–40.
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence

12
W. Gong et al. Ocean Engineering 218 (2020) 108143

Wang, Z., Xie, X., Wang, J., 2012. A new nonlinear method for vertical settlement Wood, D.M., 1990. Soil Behaviour and Critical State Soil Mechanics. Cambridge
prediction of a single pile and pile groups in layered soils. Comput. Geotech. 45, University Press, Cambridge, U.K.
118–126. Yao, Y.P., Hou, W., Zhou, A.N., 2009. UH model: three-dimensional unified hardening
Wardle, I.F., Price, G., Freeman, J., 1992. Effect of time and maintained load on the model for overconsolidated clays. Geotechnique 59 (5), 451–469.
ultimate capacity of pile in stiff clay. In: Proc., Piling: European Practice and Yin, Z.Z., 2007. Principles of Soil Mechanics. China Water and Power Press, Beijing,
Worldwide Trends. The Institution of Civil Engineers, London, pp. 92–99. China (in Chinese).
Weltman, A.J., Healy, P.R., 1978. Piling in boulder clay and other glacial tills. Report Zhang, Q.Q., Zhang, Z.M., 2012. Simplified calculation approach for settlement of single
PG5, construction industry research & information association, London. pile and pile groups. J. Comput. Civ. Eng. 26 (6), 750–758.

13