Theoretical t -z Curves for Axially Loaded Piles

Abigail H. Bateman 1; Jamie J. Crispin, Ph.D. 2; Paul J. Vardanega, Ph.D., M.ASCE 3;

and George E. Mylonakis, Ph.D., P.E., M.ASCE 4

Abstract: Estimation of nonlinear pile settlement can be simplified using one-dimensional “t-z” curves that conveniently divide the soil into
multiple horizontal “slices.” This simplification reduces the continuum analysis to a two-point boundary-value problem of the Winkler type,
which can be tackled by standard numerical procedures. Theoretical “t-z” curves can be established using the “shearing-of-concentric-
cylinders” theory of Cooke and Randolph-Wroth, which involves two main elements: (1) a constitutive model cast in flexibility form,
Downloaded from ascelibrary.org by 39.40.1.215 on 03/09/23. Copyright ASCE. For personal use only; all rights reserved.

γ ¼ γðτ Þ; and (2) an attenuation function of shear stress with radial distance from the pile, τ ¼ τ ðrÞ. Soil settlement can then be determined
by integrating shear strains over the radial coordinate, which often leads to closed-form solutions. Despite the simplicity and physical appeal
of the method, only a few theoretical “t-z” curves are available in the literature. This paper introduces three novel attenuation functions for
shear stresses, inspired by continuum solutions, which are employed in conjunction with eight soil constitutive models leading to a set of
32 “t-z” curves. Illustrative examples of pile settlement calculation in two soil types are presented to demonstrate application of the method.
DOI: 10.1061/(ASCE)GT.1943-5606.0002753. This work is made available under the terms of the Creative Commons Attribution 4.0
International license, https://creativecommons.org/licenses/by/4.0/.
Author keywords: Piles; Settlement; Nonlinear analysis; “t-z” curves; Soil/structure interaction.

Introduction are easy to implement and available in different formats, including

finite-differences (Ensoft TZPile, Reese et al. 2014), finite-elements
Nonlinear settlement estimation of axially loaded piles requires (Oasys Pile, Oasys 2017), and alternative formulations (Scott 1981;
continuum analysis using constitutive models in multiple dimen- Motta 1994; Seo et al. 2009; Crispin et al. 2018; Psaroudakis et al.
sions, which might be difficult to implement in practice and are 2019). The stress-displacement response at the pile tip (i.e., “q-z”
unavailable for many real soils. The problem can be simplified curves) and group effects lie beyond the scope of this paper; there-
to a single dimension using the Winkler model of soil reaction fore, the Illustrative Examples presented in the following work
through the introduction of pertinent stress-displacement curves involve only end-bearing solitary piles.
(“t-z” curves) along the pile shaft (Seed and Reese 1957; Coyle Some empirical “t-z” curves are available in the literature
and Reese 1966) and a stress-displacement curve (“q-z” curve) [e.g., Coyle and Reese (1966); Coyle and Sulaiman (1967); Reese
at the pile tip (Frank and Zhao 1982; Chow 1986; Armaleh and
et al. (1969); Vijayvergiya (1977); Frank and Zhao (1982); Abchir
Desai 1987; API/ANSI 2011). The pile settlement can then be es-
et al. (2016)] but are limited to specific test configurations. In ad-
timated by discretizing the pile, which is modeled as an elastic-
dition, the American Petroleum Institute and American National
compressible rod, into a number of elements separated by nodes.
Standards Institute (API/ANSI) (2011) provides general recom-
A “t-z” curve for each node and a “q-z” curve at the base are se-
mendations to develop idealized “t-z” curves for axially loaded
lected. A numerical integration approach of a two-point nonlinear
boundary value problem is then applied to compute the settlement piles based on a database of field tests. Analogous (“p-y”) curves
at the pile head due to a given load [e.g., Poulos and Davis (1980); are available for laterally loaded piles [e.g., McClelland and Focht
Salgado (2008); Guo (2012); Poulos (2017)]. Relevant algorithms (1956); Matlock (1970); Lam and Martin (1986); Reese and Van
Impe (2011); Khalili-Tehrani et al. (2014)].
1 Instead, theoretical “t-z” curves have been derived from the con-
Ph.D. Student, Dept. of Civil Engineering, Univ. of Bristol, Bristol
BS8 1TR, UK. ORCID: https://orcid.org/0000-0003-3454-1756 sideration of a (plane-strain) horizontal soil “slice.” Early solutions
2
Formerly, Ph.D. Student, Dept. of Civil Engineering, Univ. of Bristol, (Cooke 1974; Randolph and Wroth 1978; Baguelin and Frank
Bristol BS8 1TR, UK. ORCID: https://orcid.org/0000-0003-3074-8493 1979) assumed linear-elastic soil and employed the concentric
3
Associate Professor in Civil Engineering, Dept. of Civil Engineering, cylinder model to describe the attenuation of shear stress with a
Univ. of Bristol, Bristol BS8 1TR, UK. ORCID: https://orcid.org/0000 radial distance from the pile. These idealizations have been used
-0001-7177-7851 to establish linearly elastic springs for axially loaded piles which
4
Professor, Dept. of Civil Infrastructure and Environmental Engineer-
ing, Khalifa Univ., Abu Dhabi, United Arab Emirates; University Chair
are widely used in practice [Scott (1981), Guo and Randolph
in Geotechnics and Soil-Structure Interaction, Dept. of Civil Engineering, (1998), Mylonakis and Gazetas (1998), Salgado (2008), Fleming
Univ. of Bristol, Bristol BS8 1TR, UK; Adjunct Professor, Dept. of Civil et al. (2009), Viggiani et al. (2012), Guo (2012), Crispin et al.
and Environmental Engineering, Univ. of California Los Angeles, CA (2018)]. Following the same approach, nonlinear “t-z” curves
90024 (corresponding author). ORCID: https://orcid.org/0000-0002-8455 were obtained by Vardanega et al. (2012), Kraft et al. (1981) and
-8946. Email: [email protected] Chang and Zhu (1998) using power law, hyperbolic, and modified
Note. This manuscript was submitted on April 28, 2021; approved on
hyperbolic constitutive models, respectively. These curves have
November 5, 2021; published online on May 6, 2022. Discussion period
open until October 6, 2022; separate discussions must be submitted for in- been employed in various studies by Chow (1986), Zhu and
dividual papers. This paper is part of the Journal of Geotechnical and Chang (2002), Randolph (1994), Guo (2012), and Voyagaki et al.
Geoenvironmental Engineering, © ASCE, ISSN 1090-0241. (2021).

© ASCE 04022052-1 J. Geotech. Geoenviron. Eng.

J. Geotech. Geoenviron. Eng., 2022, 148(7): 04022052

 This method is versatile because it can incorporate variable soil 4. The attenuation of shear stresses with radial distance from the
properties with radial distance from the pile, installation effects, pile, τ ¼ τ ðrÞ, can be established (often taken as proportional
interface action between soil and pile, variable soil properties with to 1=r by equilibrium considerations, as discussed in the
depth, negative skin friction, and dynamic phenomena (Akiyoshi following).
1982; El Naggar and Novak 1994; Michaelides et al. 1998; Crispin 5. Deformations are infinitesimal.
and Mylonakis 2021). In addition, the method provides information Note that the second assumption can be relaxed using an
about the stress and displacement fields around a pile that can be extended set of attenuation functions, τ ðrÞ, to be discussed
used for developing solutions for pile groups, such as those by in the next section. To obtain explicit solutions, two additional
Randolph and Wroth (1979) and Mylonakis and Gazetas (1998). assumptions (6 and 7) are needed:
Moreover, the formulation can accommodate both total and effec- 6. Soil stress-strain response in shear is cast in flexibility form, that
tive stresses (by using appropriate constitutive relations); hence, is, γ ¼ γðτ Þ, where τ and γ are the shear stress and shear strain
it can handle both fine- and coarse-grained soil materials under acting on a soil element, respectively.
drained and undrained conditions. Nevertheless, despite the elegance 7. The composite function γ ¼ γðτ ðrÞÞ is integrable over the radial
and practical appeal of this method, only a handful of theoretical coordinate, r.
“t-z” curves are reported in the literature. Using these assumptions, shear strain can be integrated over the
Downloaded from ascelibrary.org by 39.40.1.215 on 03/09/23. Copyright ASCE. For personal use only; all rights reserved.

This paper derives an extended set of “t-z” curves pertaining to radial distance to provide the soil settlement at the pile-soil inter-
different conditions based on the same horizontal soil slice model face due to an applied shear stress, τ 0 , at any given depth (Fig. 1)
as the aforementioned theoretical solutions. A “toolbox” of “t-z” Z ∞ Z ∞
curves is provided to allow a designer to select an appropriate curve u0 ¼ γðrÞdr ¼ γ½τ ðrÞdr ð1Þ
d=2 d=2
from a selected soil model and an easily fitted attenuation function.
The unloading/reloading case is not investigated in this work (yet, where u0 = soil settlement at the pile circumference; r = radial dis-
no relevant restrictions are imposed by the derived “t-z” curves). tance from the pile centerline; and d = pile diameter. Since volu-
Two illustrative calculation examples involving end-bearing piles metric strains do not explicitly appear in Eq. (1), the analysis
in natural Pisa and remolded kaolinite are provided to demonstrate highlights how immediate settlement develops in nearly incom-
the simplicity of calibrating the “t-z” curves for application to a pressible media, such as saturated clay. Regarding the upper inte-
design problem. In these examples, the soil constitutive models gration limit in Eq. (1), this paper shows that it is not always
and attenuation functions are fitted to available triaxial test data feasible to extend to infinity under certain conditions.
and numerical data, respectively. A preliminary version of some After dividing the total side friction per unit pile length (πdτ 0 )
of the work presented in this paper is available in Bateman (2019). by Eq. (1), the nonlinear secant Winkler modulus, kðτ 0 Þ, of the soil
slice due to the imposed shear stress τ 0 can be readily obtained as
πd πd
Horizontal Soil Slice Model kðτ 0 Þ ≡ τ 0 ¼ τ0 R ∞ ð2Þ
u0 d=2 γ½τ ðrÞdr
Fig. 1 indicates the soil strip considered in the horizontal soil slice
model, which is based on the following main assumptions. This stiffness parameter naturally carries units of stress and rep-
1. The soil around the pile is divided into a series of independent resents a nonlinear extension of the classical modulus of subgrade
horizontal “slices” of infinitesimal thickness, Fig. 1, subjected to reaction for axially loaded piles. It is noted that no assumptions
shearing on horizontal and vertical planes. have been made, or restrictions imposed, in Eqs. (1) and (2) regard-
2. Since the continuity of the soil medium with depth is not ing the spatial distribution of soil properties, including the low
considered, the slices provide resistance only to shearing, Fig. 1. strain shear modulus and strength, which might vary with the radial
Additional resistance due to pile-induced normal stresses acting distance from the pile [e.g., observed in field tests by Kalinski et al.
on the upper and lower faces of the slices is neglected. (2001) and O’Neill (2001)]. However, the explicit solutions ob-
3. Shear strains depend solely on vertical soil displacements tained later in this paper assume radially homogenous soil. Correc-
(i.e., γ ≅ ∂u=∂r). The effect of radial displacements on shear tions to the solutions provided to account for radial inhomogeneity
strains is small and can be neglected. are available [e.g., Kraft et al. (1981), Bateman and Crispin (2020)].
Eq. (1) indicates that two functions are required to obtain a theo-
retical “t-z” curve from the horizontal soil slice model describing,
first, the attenuation of shear stresses, τ ðrÞ, and, second, the soil
constitutive relationship (in flexibility form), γðτ Þ. Previous au-
thors have utilized the concentric cylinder model to obtain an at-
tenuation function, leading to a 1=r dependence (discussed later).
This paper goes on to introduce three novel attenuation functions
inspired by available continuum solutions [Mylonakis (2001a)],
resulting in a combination of generalized power law and exponen-
tial decay functions. These attenuation functions are employed in
conjunction with eight simplified soil constitutive models from the
literature, leading to 32 families of theoretical “t-z” curves, the vast
majority (28) of which are, to the authors’ knowledge, novel.

Fig. 1. Horizontal soil slice model. [Reprinted from Bateman and Attenuation Functions
Crispin (2020), under Creative Commons-BY-4.0 license (https://
To derive attenuation functions of shear stresses, the cylindrical pile
creativecommons.org/licenses/by/4.0/).]
depicted in Fig. 2 is considered. Taking the vertical equilibrium of

© ASCE 04022052-2 J. Geotech. Geoenviron. Eng.

J. Geotech. Geoenviron. Eng., 2022, 148(7): 04022052

 Fig. 2. Equilibrium of a soil element in cylindrical coordinates. [Reprinted from Bateman and Crispin (2020), under Creative Commons-BY-4.0
license (https://creativecommons.org/licenses/by/4.0/).]
Downloaded from ascelibrary.org by 39.40.1.215 on 03/09/23. Copyright ASCE. For personal use only; all rights reserved.

an infinitesimal soil element in cylindrical coordinates and focusing An approximate solution to Eq. (5) can be obtained using a
on pile-induced stresses yields [Randolph and Wroth (1978), technique analogous to the one employed for spread footings by
Mylonakis (2001b), Anoyatis and Mylonakis (2012)] Vlasov and Leontiev (1966), previously applied to this problem
by Vallabhan and Mustafa (1996), Lee and Xiao (1999), and
∂ðrτ Þ ∂σ Mylonakis (2000, 2001b) (similar approaches have been employed
þ r¼0 ð3Þ
∂r ∂z for laterally loaded piles, including Guo and Lee (2001), Basu
and Salgado (2008), Shadlou and Bhattacharya (2014), and
where σ and τ = vertical normal and shear stresses, respectively;
Bhattacharya (2019), among others). By separating the displace-
and z = depth below ground level.
ment function in Eq. (5), uðr; zÞ, into radial, uðrÞ, and vertical,
A simplified solution to this equation can be derived by assum-
χðzÞ, components [uðr; zÞ ¼ uðrÞχðzÞ], multiplying by a virtual
ing that variations in the vertical stress with a depth due to pile load-
displacement χðzÞ and integrating over the vertical coordinate, z,
ing are negligible; accordingly, setting ∂σ=∂z ¼ 0 and integrating
these authors obtained an alternative form of Eq. (5) that is now
over r yields the elementary solution [Cooke (1974), Randolph and
independent of depth [modified from Mylonakis (2000)]
Wroth (1978); Baguelin and Frank (1979)]
     2
d2 uðrÞ 1 duðrÞ 2q
τ ðrÞ ¼ τ 0
d
ð4Þ þ − uðrÞ ¼ 0 ð6Þ
2r dr2 r dr d

This result is commonly known as the concentric cylinder (or where q = compressibility constant that results from the integration
plane strain) model, in which the soil is treated as a series of con- with depth and is proportional to η. The specific form of q is chosen
centric cylinders with respective shear forces and no resistance by to enable it to be employed as a dimensionless fitting parameter to
means of vertical normal stresses. This model provides a simple simplify the problem.
attenuation function that has been extensively employed in the Contrary to the elementary concentric cylinder model, Eq. (6)
literature (Scott 1981; Fleming et al. 2009; Guo 2012; Viggiani duly accounts for the continuity of the soil in the vertical direction.
et al. 2012). Eq. (6) is of the Bessel type and admits the following solutions for
A difficulty associated with the use of this equation lies in the displacements and stresses (Mylonakis 2000; Olver et al. 2010)
singular nature of the associated displacement field. Indeed, inte- K 0 ð2qr=dÞ
grating shear stresses over the radial coordinate leads to a logarith- uðrÞ ¼ u0 ð7aÞ
K 0 ðqÞ
mic solution for displacements that diverges with increasing r. To
correct this problem, Randolph and Wroth (1978) suggested an
empirical radius, rm , beyond which the vertical soil settlement K 1 ð2qr=dÞ
τ ðrÞ ¼ τ 0 ð7bÞ
can be assumed negligible; rm is usually taken as proportional to K 1 ðqÞ
the pile length, L, and specific values have been suggested by
Randolph and Wroth (1978), Guo and Randolph (1998), Fleming where K 0 and K 1 = modified Bessel functions of the second kind
et al. (2009), and Guo (2012). and order 0 and 1, respectively. Interestingly, these solutions are
A more rigorous yet simple approach for tackling this problem identical to those of the related dynamic plane strain problem
is possible by first casting the governing equation [Eq. (3)] in dis- for axially loaded piles pioneered by Baranov (1967) and Novak
placement form through the introduction of the approximate stress- (1974), and tend to zero with increasing r (thus rendering empirical
displacement relations τ ≅ −Gs ∂u=∂r and σ ≅ −Ms ∂u=∂z [as, corrections such as rm unnecessary). A discussion of this remark-
for instance, done by Nogami and Novak (1976), Mylonakis able similarity is provided in Mylonakis (2001a).
(2001b), and Anoyatis et al. (2019)] Due to its complexity, this expression does not result in closed-
  form “t-z” curves as desired. Concentrating on the solution for
∂ ∂uðr; zÞ ∂ 2 uðr; zÞ stresses in Eq. (7b), this can be simplified considering the asymp-
r þ η2 r ¼0 ð5Þ
∂r ∂r ∂z2 totic expansions of the associated Bessel function in different
regimes. For instance, for small radial distances from the pile
where u = vertical soil settlement; η2 ¼ Ms =Gs is a dimensionless centerline, the asymptotic form of the modified Bessel function
compressibility coefficient; Gs = elastic shear modulus; Ms = is the power law function K 1 ðxÞ ∼ ð1=2ÞΓðvÞð2=xÞ1 , where x =
compressibility modulus of the soil material; and η = function independent variable and ΓðvÞ = Gamma function [Olver et al.
of the Poisson’s ratio of the soil, ν s . Suitable values of η (and, thus, (2010)]. Substituting this expression into Eq. (7b) yields Eq. (4),
M s ) are discussed in Appendix I and in Mylonakis (2001a, b). which implies that the first term in Eq. (3) governs the behavior

© ASCE 04022052-3 J. Geotech. Geoenviron. Eng.

J. Geotech. Geoenviron. Eng., 2022, 148(7): 04022052

 of the solution in the vicinity of the pile, as assumed in the classical in Appendix I) is detailed subsequently in the paper. The approxi-
concentric cylinder model, regardless of soil compressibility. mate range of the parameters from these examples, given in Table 2,
For large radial distances from the pile centerline, the pertinent are 0.1 < q < 0.3, 1.0 < m < 1.25, and 0.75 < n < 0.95.
asymptotic expression is K 1 ðxÞ ∼ ðπ=2xÞ1=2 e−x [Olver et al. (2010)], The attenuation functions in Eqs. (4), (8), (9), and (10) are in-
which yields the approximate solution (referred in the ensuing as dicated in Fig. 3 with the fitted parameters from illustrative
“power-exponential decay”) Example B. These are compared with finite element method (FEM)
 −1=2    results in PLAXIS 2D (Brinkgreve et al. 2019) for both linear and
2r 2r hyperbolic soil constitutive models (using the Mohr-Coulomb and
τ ðrÞ ¼ τ 0 exp −q −1 ð8Þ
d d HS Small soil models to match the two constitutive models used
here). An elastostatic FEM analysis using axisymmetric conditions
Eqs. (4) and (8) are based on sound theoretical arguments that was carried out to model an end-bearing pile for each of the Illus-
include some necessary simplifying assumptions. Therefore, to trative Examples, which are discussed in the corresponding section.
provide flexibility in the ensuing analyses, these can be generalized The pile (modeled using a linear-elastic material) has a fully fixed
in the form shown in Eqs. (9) and (10). By applying the additional rigid layer at the pile base, and horizontal fixities are provided at
fitting parameters (m and n), a better fit can be achieved the pile centerline and at a distance of 75d (Example A) or 50d
 −m
Downloaded from ascelibrary.org by 39.40.1.215 on 03/09/23. Copyright ASCE. For personal use only; all rights reserved.

2r (Example B). A mesh with 15-node triangular elements was gen-

τ ðrÞ ¼ τ 0 ð9Þ erated by PLAXIS (total number of elements: 6026 for Example A
d and 3245 for Example B). The resulting attenuation functions for
 −n    Example B are also plotted in Fig. 3. Further details on this example
2r 2r
τ ðrÞ ¼ τ 0 exp −q −1 ð10Þ are provided later in this paper.
d d Overall, the differences between the various curves are minor
and start to manifest themselves at medium radial distances
where expðxÞ = exponential function. The first [Eq. (9)] of these
(r > 5d), which agrees with the observations from the asymptotic
modified solutions is referred in the ensuing as “generalized con-
solutions in Eqs. (4) and (8). These results suggest that the attenu-
centric cylinder,” whereas the second [Eq. (10)] is referred to as
ation of shear stresses in the vicinity of the pile is governed by
“generalized power-exponential decay” model. The last model en-
equilibrium, not stress-strain behavior. Soil compressibility starts
compasses all others by using appropriate values for parameters n
becoming important at higher distances, where shear stresses
and q. In addition, these equations have previously been suggested
and strains are generally lower than the corresponding normal
as approximate attenuation functions for displacements and asso-
stresses and volumetric strains. These traits justify using some elas-
ciated time derivatives in relevant dynamic problems involving fre-
ticity arguments [used to derive Eqs. (8) and (10)] for tackling the
quency and damping (Mylonakis 1995; Gazetas et al. 1998; Su
stress distribution in the nonlinear problem at hand. Similar patterns
et al. 2019); here, they are used as semiempirical static solutions
are observed in other relevant problems, such as the Boussinesq and
for stresses.
Mindlin solutions [e.g., Davis and Selvadurai (1996)].
Eqs. (9) and (10) enable theoretical “t-z” curves to be analyti-
cally derived without reliance on the empirical radius rm , which is
employed in the concentric cylinder model [provided m is greater Soil Constitutive Models
than 1 in Eq. (9)]. Instead, these equations require fitting the param-
eter m [Eq. (9)] or the parameters q and n [Eq. (10)] to field test Eight soil constitutive models are considered in this paper. To ob-
data, numerical continuum solutions or more rigorous analytical tain an explicit solution using Eq. (1), the selected constitutive re-
models, similar to how rm has been fitted by previous authors lations must be cast in flexibility form, that is, γ ¼ γðτ Þ. Each soil
(Randolph and Wroth 1978; Guo 2012). An example of fitting constitutive model (illustrated in Fig. 4) is given as follows in both
these parameters to a more rigorous analytical solution (that is the (a) stiffness and (b) flexibility forms. A table summarizing the
unsuitable for direct use in developing “t-z” curves and discussed models is provided in Appendix III. Each model contains a number

(a) (b)

Fig. 3. Attenuation functions (parameters from Example B, Table 2): (a) linear; and (b) log scale.

© ASCE 04022052-4 J. Geotech. Geoenviron. Eng.

J. Geotech. Geoenviron. Eng., 2022, 148(7): 04022052

 The bilinear model provides a simple adaption by splitting the
linear model into two regions, the first having a stiffness G1 and the
second having a reduced stiffness G2 , which is applicable after a
predefined yield stress, τ 1 , is reached. Additional stiffness changes
could be introduced to more closely model nonlinear soil behavior.
However, doing so would be at the expense of an increasingly com-
plex function involving many fitting parameters. The bilinear
model is given by (for τ < τ max )
8
>
< γG 1 γ ≤ τ 1 =G1

τ¼ τ1 ð12aÞ
>
: G2 γ − G þ τ 1 γ ≥ τ 1 =G1
1

8
> τ
(a) >
<G τ ≤ τ1
Downloaded from ascelibrary.org by 39.40.1.215 on 03/09/23. Copyright ASCE. For personal use only; all rights reserved.

1
γ¼ τ − τ1 τ1 ð12bÞ
>
> þ τ ≥ τ1
:
G2 G1

Alternatively, a simple power law relationship can be employed,

in this case defined with the shear strain when 50% of the soil shear
strength is mobilized, γ 50 , and a soil nonlinearity exponent, b. This
model was employed by Vardanega et al. (2012) to generate simple
“t-z” curves [see also Williamson (2014), Vardanega (2015),
Crispin et al. (2019)]. The power law relationship has the advantage
that rm is not required due to an infinite initial stiffness (b < 1). The
power law model is given by (for τ < τ max )
 
τ max γ b
τ¼ ð13aÞ
2 γ 50
 1
(b) 2τ b
γ ¼ γ 50 ð13bÞ
τ max
However, since infinite initial stiffness is not physically realiz-
able, a linear model can be employed at low stress (τ ≤ τ i ), with an
initial stiffness Gi , resulting in the linear-power law soil constitu-
tive model (for τ < τ max )
8
>
< γGi   γ ≤ τ i =Gi
τ ¼ τ max γ b ð14aÞ
>
: 2 γ ≥ τ i =Gi
γ 50
8
> τ
>
> τ ≤ τi
<G i
γ¼  1 ð14bÞ
>
> 2τ b
> γ
: 50 τ τ ≥ τi
max
(c)
where
Fig. 4. Illustration of constitutive models considered [Eqs. (11)–(18)].   b
τ max 2Gi γ 50 b−1
τi ¼ ð14cÞ
2 τ max
of parameters that should be fitted to site-specific laboratory data
and/or in situ testing. An example of fitting these constitutive mod- A combination of the linear and power law models was sug-
els to the soil test data is subsequently discussed. gested by Ramberg and Osgood (1943), which is defined here with
The simplest constitutive relation considered is the linear stress- a reference shear strain, γ r , and two fitting constants, c1 and c2
strain model defined by a single shear modulus value, G, given in (Ishihara 1996). This model is available in only the flexibility form
stiffness and flexibility form by (for τ < τ max ) (for τ < τ max )
   
τ τ c2
τ ¼ γG ð11aÞ γ ¼ γr þ c1 ð15Þ
τ max τ max
τ The five constitutive relationships previously discussed do not
γ¼ ð11bÞ
G approach a limiting stress; therefore, a shear strength cap, τ max ,

© ASCE 04022052-5 J. Geotech. Geoenviron. Eng.

J. Geotech. Geoenviron. Eng., 2022, 148(7): 04022052

    
is introduced. For undrained conditions, this ultimate strength τ Rf τ c3 −1
should be taken as equal to the undrained shear strength, τ max ¼ γ¼ 1− ð17Þ
Gi τ max
cu , which, in turn, depends on the shear mode of the soil test
[e.g., Mayne (1985), Beesley and Vardanega (2020)]. Nevertheless, A similar asymptotic property can be obtained using an expo-
a generic parameter τ max is employed because the “t-z” curves are nential relationship resulting in the exponential model indicated
not linked to a particular test or limited to undrained conditions. with a similar fitting parameter, Rf , employed as before
In practice, the shear strength of the resultant “t-z” curves might    
be capped at a lower value to represent slip at the pile-soil interface. τ max R f Gi
τ¼ 1 − exp −γ ð18aÞ
To this end, empirical factors such as the familiar adhesion Rf τ max
parameter α (Tomlinson 1957; Skempton 1959; Meyerhof 1976;
 
Chakraborty et al. 2013) are available for undrained problems, τ max Rf τ
which can be used to limit the maximum shear stress. Note that γ¼− ln 1 − ð18bÞ
R f Gi τ max
in the interest of space, simple names such as “Bilinear model”
are adopted here over more accurate alternatives such as “capped Note that some of the formulations include an initial (low-strain)
Bilinear model” or “Bilinear-fully plastic model.” shear modulus, Gi , which is employed as a model parameter to
Downloaded from ascelibrary.org by 39.40.1.215 on 03/09/23. Copyright ASCE. For personal use only; all rights reserved.

Instead, the hyperbolic model, which approaches a single value enable a better fit of the constitutive relation to the available data
at large strain, can be employed. Different forms of this model are over the strain range of interest. Alternatively, when small strains
available in the literature [e.g., Kondner (1963)], and it was utilized are important, this modulus might be set equal to the maximum
by Kraft et al. (1981) to develop a “t-z” curve using the concentric shear modulus, Gmax , from high-quality experimental testing
cylinder model. The form of the hyperbolic model adopted by Kraft (e.g., seismic cone penetration testing).
et al. (1981) is employed with a fitting parameter, Rf , which acts as
a factor to τ max that alters the location of the asymptote. This en-
ables a better fit of this model over a desired stress range, although
this results in the model performing poorly at high stresses. In the
“t -z” Curves
classical model, Rf is expected to be less than 1 and requires a shear By substituting the four attenuation functions [Eqs. (4), (8), (9),
strength cap similar to the previous five models. However, in this and (10)] into each constitutive model [Eqs. (11)–(18)], theoretical
paper, it has been employed as a fitting parameter and allowed to “t-z” curves can be derived analytically using Eq. (1). As an exam-
vary outside this range. The hyperbolic model is given by ple, substituting the concentric cylinder model [Eq. (4)] into the
 −1 linear soil constitutive model Eq. (11b) results in a function, γðrÞ.
τ τ max
τ ¼ max þ1 ð16aÞ By inputting this directly into Eq. (1), infinite settlement would be
Rf Rf G i γ
predicted when integrating over an infinite distance; thus, the upper
 −1 integration limit is replaced with rm , and the following equation is
τ max τ max derived. This is valid until slip occurs at the pile-soil interface
γ¼ −1 ð16bÞ
Rf G i Rf τ (Randolph and Wroth 1978; Fleming et al. 2009)
Z  
In addition, a modified hyperbolic function is employed with an u0 τ 0 rm 1 τ0 2rm
¼ dr ¼ ln ð19Þ
exponent, c3 , to give more control over the stiffness degradation, as d 2G d=2 r 2G d
discussed by Fahey and Carter (1993). This model has previously
been utilized to develop a “t-z” curve in conjunction with the con- Theoretical “t-z” curves derived using the concentric cylinder
centric cylinder model by Chang and Zhu (1998). The modified model Eq. (4) are given as follows and are summarized in
hyperbolic model is given only in the flexibility form Appendix IV.

Bilinear [Eq. (12b)]

8  
>
> τ0 2rm
> ln τ0 ≤ τ1
u0 < 2G1 d
¼        ð20Þ
d > τ 0 ðG1 − G2 Þ
> G2 2rm τ 1 G1 τ0 τ1
>
: ln þ ln þ −1 τ0 ≥ τ1
2G1 G2 G1 − G2 d τ0 G1 − G2 τ1 τ0

Power Law, for b < 1 (Vardanega et al. 2012; Williamson 2014; Vardanega 2015; Crispin et al. 2019) [Eq. (13b)]
 1
u0 γ 50 b 2τ 0 b
¼ ð21Þ
d 2ð1 − bÞ τ max

Linear-Power Law [Eq. (14b)]

8  
>
> τ0 2rm
> ln τ0 ≤ τi
u0 < 2Gi d
¼  1   b−1    ð22aÞ
d >
> γ b 2τ 0 b 2ri b τ r
>
: 50
1− þ 0 ln m τ0 ≥ τi
2ð1 − bÞ τ max d 2Gi ri

© ASCE 04022052-6 J. Geotech. Geoenviron. Eng.

J. Geotech. Geoenviron. Eng., 2022, 148(7): 04022052

 where
 b
2ri 2τ 0 2γ 50 Gi 1−b
¼ ð22bÞ
d τ max τ max

Ramberg-Osgood [Eq. (15)]

       
u0 τ 0 γ r 2rm γr c1 τ 0 c 2 2rm 1−c2
¼ ln þ 1− ð23Þ
d 2τ max d 2ðc2 − 1Þ τ max d

Hyperbolic [Kraft et al. (1981)] [Eq. (16b)]

    
u0 τ 2rm Rf τ 0 Rf τ 0
¼ 0 ln − − ln 1 − ð24Þ
d 2Gi d τ max τ max
Downloaded from ascelibrary.org by 39.40.1.215 on 03/09/23. Copyright ASCE. For personal use only; all rights reserved.

Modified Hyperbolic [Chang and Zhu (1998)] [Eq. (17)]

         
u0 τ0 2rm c3 Rf τ 0 c3 Rf τ 0 c3
¼ ln − − ln 1 − ð25Þ
d 2Gi c3 d τ max τ max

Exponential [Eq. (18b)]

          
u0 τ Rf τ 0 2r Rf τ 0 d Rf τ 0 2rm Rf τ 0 Rf τ 0
¼ max ln 1 − − m ln 1 − þ ln − − ln 1 − ð26Þ
d 2Rf Gi τ max d τ max 2rm τ max d τ max τ max

Theoretical “t-z” curves derived using the generalized concentric cylinder model [Eq. (9)] are given in Eqs. (27)–(34) and are summarized
in Appendix IV. The rm value has been included in these solutions; however, when m > 1, the relevant terms vanish and the solutions can be
further simplified.
Linear
   
u0 τ0 2rm 1−m
¼ 1− ð27Þ
d 2Gðm − 1Þ d

Bilinear
8    
>
> τ0 2rm 1−m
>
< 1− τ0 ≤ τ1
u0 2G1 ðm − 1Þ d
¼      1−m    1−m  ð28Þ
d >
> τ 0 ðG1 − G2 Þ 1 G1 G2 2rm 1−m τ τ τ
>
: − − 0
m
þ 1− 0
m
τ0 ≥ τ1
2G1 G2 m − 1 G1 − G 2 G1 − G2 d τ1 τ0 τ1

Power Law, for b < m

 1
u0 γ 50 b 2τ 0 b
¼ ð29Þ
d 2ðm − bÞ τ max

Linear-Power Law
8    
>
> τ0 2rm 1−m
>
> 1− τ0 ≤ τi
u0 < 2Gi ðm − 1Þ d
¼   1  b − m   1−m    ð30aÞ
d >
> γ 50 b 2τ 0 b 2ri τ0 2ri 2rm 1−m
>
> 1− b þ − τ0 ≥ τi
:
2ðm − bÞ τ max d 2Gi ðm − 1Þ d d

where
   b 1
2ri 2τ 0 2γ 50 Gi 1−b m
¼ ð30bÞ
d τ max τ max

Ramberg-Osgood
         
u0 τ 0γr 2rm 1−m γr c1 τ 0 c 2 2rm 1−c2 m
¼ 1− þ 1− ð31Þ
d 2τ max ðm − 1Þ d 2ðc2 m − 1Þ τ max d

© ASCE 04022052-7 J. Geotech. Geoenviron. Eng.

J. Geotech. Geoenviron. Eng., 2022, 148(7): 04022052

 Hyperbolic
        
u0 τ max Rf τ 0 m1 τ 2rm m 1 τ 1
¼ B max ; ; 0 − B max ; ; 0 ð32Þ
d 2Rf mGi τ max Rf τ 0 d m Rf τ 0 m
where Bðx; a; bÞ = incomplete beta function (see Appendix II).
Modified Hyperbolic
          
u0 τ max Rf τ 0 m1 R f τ 0 c3 m − 1 Rf τ 0 d m c3 m − 1
¼ B ; ;0 − B ; ;0 ð33Þ
d 2Rf Gi c3 m τ max τ max c3 m τ max 2rm c3 m

Exponential
               
u0 τ max Rf τ 0 2rm Rf τ 0 d m Rf τ 0 m1 τ max 1 τ max 2rm m 1
¼ ln 1 − − ln 1 − þ B ; ;0 − B ; ;0 ð34Þ
d 2Rf Gi τ max d τ max 2rm τ max Rf τ 0 m Rf τ 0 d m

Theoretical “t-z” curves derived using the power-exponential decay [Eq. (8)] are given as follows and summarized in Appendix V.
Downloaded from ascelibrary.org by 39.40.1.215 on 03/09/23. Copyright ASCE. For personal use only; all rights reserved.

Linear
pffiffiffi
u 0 τ 0 eq π pffiffiffi
¼ pffiffiffi erfcð qÞ ð35Þ
d 2G q
where erfcðxÞ = complementary error function (see Appendix II).
Bilinear
8 pffiffiffi
>
> τ 0 eq π pffiffiffi
>
> p ffiffiffi erfcð qÞ τ0 ≤ τ1
u0 < 2G1 q
¼  pffiffiffi  rffiffiffiffiffiffiffiffiffiffi   ð36aÞ
d >
> τ 0 ðG1 − G2 Þ eq π G1 pffiffiffi 2qr1 τ 1 2r1
>
> pffiffiffi erfcð qÞ − erfc − −1 τ0 ≥ τ1
: 2G1 G2 q G1 − G2 d τ0 d

where
  2 
2r1 1 τ
¼ W 2qe2q 0 ð36bÞ
d 2q τ1
where WðxÞ denotes the Lambert W-Function (see Appendix II).
Power Law
 2b−1  1  
u0 γ 50 q b 2b 2τ 0 b 2b − 1 q
¼ eb Γ ; ð37Þ
d 2 q τ max 2b b
where Γðs; xÞ is the upper incomplete gamma function (see Appendix II).
Linear-Power Law
8 pffiffiffi
>
> τ eq π pffiffiffi
> 0 pffiffiffi erfcð qÞ
> τ0 ≤ τi
u0 < 2Gi q
¼  2b−1  1        ð38aÞ
d >
> γ 50 q b 2b 2τ 0 b 2b − 1 q 2b − 1 2ri q τ 0 eq 1 2ri
>
> e Γ ; − Γ ; þ pffiffiffi Γ ; q τ0 ≥ τi
: 2 b
q τ max 2b b 2b d b 2Gi q 2 d

where
     2b 
2ri 1 2τ 0 2 2γ 50 Gi 1−b
¼ W 2qe2q ð38bÞ
d 2q τ max τ max
Ramberg-Osgood
     
u0 τ γ eq 1 γ c2 −2 c1 τ 0 c 2 2 − c2
¼ 0 r pffiffiffi Γ ; q þ r ec2 q ðqc2 Þ 2 Γ ; qc2 ð39Þ
d 2τ max q 2 2 τ max 2

Hyperbolic
Z  rffiffiffiffiffi    −1
u0 τ ∞ τ max 2r 2r
¼ max exp q −1 −1 dr ð40Þ
d Rf dGi d=2 Rf τ 0 d d

Modified Hyperbolic
Z rffiffiffiffiffi     rffiffiffiffiffi   c3 −1
u0 τ ∞ d 2r Rf τ 0 d 2r
¼ 0 exp −q −1 1− exp −q −1 dr ð41Þ
d Gi d d=2 2r d τ max 2r d

© ASCE 04022052-8 J. Geotech. Geoenviron. Eng.

J. Geotech. Geoenviron. Eng., 2022, 148(7): 04022052

 Exponential
Z  rffiffiffiffiffi   
u0 τ ∞ Rf τ 0 d 2r
¼ − max ln 1 − exp −q −1 dr ð42Þ
d Rf G i d d=2 τ max 2r d

Theoretical “t-z” curves derived using the generalized power-exponential decay [Eq. (10)] are given as follows and summarized in
Appendix V.
Linear
u0 τ
¼ 0 eq qn−1 Γð1 − n; qÞ ð43Þ
d 2G
Bilinear
8
> τ 0 q n−1
>
> e q Γð1 − n; qÞ τ0 ≤ τ1
u0 < 2G 1
¼       ð44aÞ
>
> τ 0 ðG1 − G2 Þ q n−1 G1 2r1 τ 1 2r1
Downloaded from ascelibrary.org by 39.40.1.215 on 03/09/23. Copyright ASCE. For personal use only; all rights reserved.

d
>
: 2G G e q Γð1 − n; qÞ − Γ 1 − n; q − − 1 τ0 ≥ τ1
1 2 G1 − G2 d τ0 d

where
  1 
2r1 n q q τ n
¼ W en 0 ð44bÞ
d q n τ1

Power Law
 b−n  1  
u0 γ 50 q b b 2τ 0 b b − n q
¼ e b Γ ; ð45Þ
d 2 q τ max b b

Linear-Power Law
8
> τ 0 q n−1
>
> e q Γð1 − n; qÞ τ0 ≤ τi
u0 < 2G i
¼     1        ð46aÞ
> γ 50 q b b 2τ 0 b b−n q b − n 2ri q τ 2r
b−n
d >
> Γ ; −Γ ; þ 0 eq qn−1 Γ 1 − n; q i τ0 ≥ τi
: 2 eb q τ max b b b d b 2Gi d

    b 1 
2ri n q q 2τ 0 2γ 50 Gi 1−b n
¼ W e n ð46bÞ
d q n τ max τ max

Ramberg-Osgood
 c
u0 τ 0 γ r q n−1 γ r ec2 q c2 n−1 c1 τ 0
2
¼ e q Γð1 − n; qÞ þ ðqc2 Þ Γð1 − c2 n; qc2 Þ ð47Þ
d 2τ max 2 τ max

Hyperbolic
Z       −1
u0 τ ∞ τ max 2r n 2r
¼ max exp q −1 −1 dr ð48Þ
d Rf dGi d=2 Rf τ 0 d d

Modified Hyperbolic
Z  n      
u0 τ ∞ d Rf τ 0 d n −qð2r−1Þ c3 −1
¼ 0
2r
e−qð d −1Þ 1 − e d dr ð49Þ
d Gi d d=2 2r τ max 2r

Exponential (42), (48), (49), and (50)], for which explicit solutions were not
Z       found. The “t-z” curves [Eqs. (19)–(50)] are given normalized by
u0 τ max ∞ Rf τ 0 d n 2r the pile diameter, d, which indicates that the normalized settle-
¼− ln 1 − exp −q −1 dr
d Rf Gi d d=2 τ max 2r d ment, u0 =d, is independent of the pile diameter for a given pile
ð50Þ slenderness, L=d. The parameters dependent on the pile slender-
ness come from the attenuation functions (rm , m, q, n). Therefore,
The derived “t-z” curves are all presented in closed form with for a given pile slenderness, the settlement at the pile–soil
the exception of the hyperbolic, modified hyperbolic, and expo- interface is proportional to the pile diameter independent of the
nential constitutive models in Eqs. (8) and (10) [Eqs. (40), (41), constitutive model selected.

© ASCE 04022052-9 J. Geotech. Geoenviron. Eng.

J. Geotech. Geoenviron. Eng., 2022, 148(7): 04022052

 These “t-z” curves act as a toolbox, which a designer might can be found in Soga (1994). For both soils, Soga (1994) undertook
easily employ to calculate the pile settlement (at any depth). This an isotropic undrained triaxial compression test for a strain rate of
requires only the parameters for both the soil constitutive model 0.5% per min to obtain stress-strain results (τ –γ) for each material,
and the attenuation functions to be determined. First, the soil constit- plotted in Fig. 6. From this digitized data, cu was interpreted as the
utive parameters can be fit directly to a chosen stress range from ex- maximum shear stress reached [cu ¼ 45 kPa for Pisa clay and cu ¼
perimental test data (e.g., a standard triaxial test from a representative 29 kPa for kaolinite; see Figs. 6(a and b), respectively].
soil sample), similar to how input parameters for a standard numerical Parameters for each soil constitutive model [Eqs. (11)–(18)] are
analysis are determined. Second, the attenuation function can also be determined using a nonlinear least squares fit to the triaxial test data.
directly fitted to a numerical analysis or a more rigorous analytical In some cases, model parameters can be directly interpreted from the
solution accounting for material nonlinearity in the soil and the test results instead of being employed as a fitting parameter. In ad-
pile-soil interface [e.g., Akiyoshi (1982)], if available. dition to the undrained triaxial tests, Soga (1994) undertook cyclic
The “t-z” curves developed in this work can be used to establish torsional shear tests at 0.5 Hz on the same samples that allow a direct
the total head settlement. The pile is discretized, and a “t-z” curve is interpretation of Gmax [29 MPa for Pisa clay and 78 MPa for kao-
selected for each section (with fitted parameters as previously dis- linite; Soga (1994)]. Both combinations of fitted (i.e., Gi as a free
cussed). Various numerical integration techniques for two-point parameter) and interpreted (i.e., Gi ¼ Gmax ) stiffnesses were consid-
Downloaded from ascelibrary.org by 39.40.1.215 on 03/09/23. Copyright ASCE. For personal use only; all rights reserved.

boundary-value problems are then available to determine the total ered to ascertain which provides the best results for each model. It
head settlement, such as the Runge-Kutta methods. Alternatively, should be noted that to ensure a reasonable initial stiffness for the
numerical analysis software packages [e.g., Oasys Pile, Oasys modified hyperbolic model, Gi ¼ Gmax is selected.
(2017)] often enable user-defined “t-z” curves to be inputted. Two Moreover, a desired stress range of 0.2cu ≤ τ ≤ 0.8cu was se-
illustrative examples involving end-bearing single piles are pro- lected for both examples. Thus, only data points within this region
vided to demonstrate the simplicity of determining these parame- were considered during the fitting stage to optimize the results. This
ters in practice. stress range was chosen to better reflect the response at moderate
strains and to limit the effect of the scatter at the onset of plastic
behavior at the expense of accurately modeling the small strain
Illustrative Examples response [this approach was also employed in Vardanega and
Bolton (2011) when calibrating a power-law using collected soil
Two example scenarios, indicated in Fig. 5, are examined using two stress-strain data]. Although the two examples in this paper are
representative soils to determine “t-z” curves at the mid-depth of assessed within a specific stress range (0.2cu ≤ τ ≤ 0.8cu ), an
each example pile and demonstrate the application of the method appropriate stress range should be selected that is specific to each
described in this paper. These two examples are: (A) an end-bearing problem. This should be decided based on the expected loading
concrete pile embedded in Pisa clay, and (B) an end-bearing steel conditions of the pile, which might affect which soil constitutive
pile embedded in kaolinite. Note that Example B is treated as an model is chosen.
equivalent solid pile with the same Ep A, where A is the cross- Since multiple fittings for each soil constitutive model are con-
sectional area of the pile. The example uses end-bearing piles to sidered, a method to select the most appropriate fitted line is re-
only consider the response of the pile shaft (“t-z” curves) and re- quired. To assess each fitting, the measured shear stress from
move the effect of the base response (a “q-z” curve), which is out of the triaxial test data can be compared with that predicted by each
the scope of this paper. In practice, this can be repeated for multiple model and an error bound obtained based on a selected proportion
depths to obtain the total pile settlement. of the data (95% is used here). The fitting with the smallest error
To analyze these examples, representative soil samples are se- bound is selected (or an alternative method can be applied). Note
lected using soil test data from Soga (1994) (note that this is an that this method does not bias the result toward conservative pre-
example analysis and not intended to provide site-specific design dictions. Full details and results of this process are given in the
values or comparisons to load test data). Example A is represented Supplemental Materials. A summary of selected parameters and
by an undisturbed sample of a high plasticity Pisa clay taken from a error bounds for each constitutive model are given in Table 1
depth of 10 m (z=L ¼ 0.5). Example B is represented by a floccu- and plotted in Fig. 6.
lated kaolinite which was remolded and subsequently tested at a For Pisa clay in Example A, the smallest error bound is observed
confining pressure of 98 kPa to represent a sample at a depth of with the Ramberg-Osgood model. However, the error bounds of
approximately 5 m (z=L ¼ 0.5). Full details of the material tested both the power law model and the linear-power law model are only
slightly larger (2.6% instead of 2.3%). In this case, the power law
model has been chosen for Pisa clay because it provides similarly
accurate results as the other two models and only requires two
model parameters to be fitted.
For the kaolinite, the Ramberg-Osgood model also has the
smallest error bound. In addition, the second best-performing
model—the hyperbolic model—employs the fitting parameter,
Rf , which has the effect of scaling cu to provide a better fit within
the desired stress range at the expense of an accurate fit at high
stresses. This is evident from the kaolinite triaxial test data for which
the model does not perform well beyond 0.8cu . Although not the
particular region of interest, the Ramberg-Osgood model also pro-
vides a reasonable fit in this range and is, thus, selected for the
kaolinite.
If the small strain response is of interest in a problem, an
Fig. 5. Examples of end-bearing piles in: Pisa clay (Example A) and
experimentally determined Gmax should be employed (e.g., in the
kaolinite (Example B).
linear-power law model). If the stress is unlikely to exceed the

© ASCE 04022052-10 J. Geotech. Geoenviron. Eng.

J. Geotech. Geoenviron. Eng., 2022, 148(7): 04022052

 (a)
Downloaded from ascelibrary.org by 39.40.1.215 on 03/09/23. Copyright ASCE. For personal use only; all rights reserved.

(b)

Fig. 6. Chosen fitting (this work) of each constitutive model to undrained triaxial compression test data for (a) Pisa clay; and (b) kaolinite fitted over
the range 0.2cu ≤ τ ≤ 0.8cu (fitted parameters given in Table 1). N denotes number of data points considered in each range. (Data from Soga 1994.)

Table 1. Parameters from the selected fitting for each soil constitutive model and error bounds (%) on Predicted versus Measured shear stress plots (Fig. 6)
that 95% of data (within 0.2cu –0.8cu ) fall between. cu and Gmax are interpreted values from Soga (1994). Corresponding figure numbers in Supplemental
Materials are indicated in brackets next to soil constitutive model name
Soil type Soil constitutive model Parameters (τ max ¼ cu ) Error bounds (%)
Pisa Linear [Fig. S1(a)] G ¼ Gmax 244
Bilinear [Fig. S2(a)] G1 ¼ 96.5 MPa, G2 ¼ 1.1 MPa, τ 1 ¼ 12.6 kPa 28
Power law [Fig. S3(a)] γ 50 ¼ 0.0079, b ¼ 0.41 2.6
Linear-Power law [Fig. S4(a)] Gi ¼ Gmax , γ 50 ¼ 0.0079, b ¼ 0.41 2.6
Ramberg-Osgood [Fig. S5(a)] γ r ¼ 0.00053, c1 ¼ 5.7, c2 ¼ 2.5 2.3
Hyperbolic [Fig. S6(a)] Gi ¼ 7.6 MPa, Rf ¼ 1.12 37
Modified Hyperbolic [Fig. S7(a)] Gi ¼ Gmax , Rf ¼ 1, c3 ¼ 0.17 22
Exponential [Fig. S8(a)] Gi ¼ 5.8 MPa, Rf ¼ 1.39 63
Kaolinite Linear [Fig. S1(b)] G ¼ 6.4 MPa 114
Bilinear [Fig. S2(b)] G1 ¼ 12.2 MPa, G2 ¼ 0.4 MPa, τ 1 ¼ 15.4 kPa 18
Power law [Fig. S3(b)] γ 50 ¼ 0.0028, b ¼ 0.24 36
Linear-Power law [Fig. S4(b)] Gi ¼ Gmax , γ 50 ¼ 0.0028, b ¼ 0.24 36
Ramberg-Osgood [Fig. S5(b)] γ r ¼ 0.0021, c1 ¼ 1.8, c2 ¼ 6.8 2.6
Hyperbolic [Fig. S6(b)] Gi ¼ 20 MPa, Rf ¼ 1.26 6.8
Modified Hyperbolic [Fig. S7(b)] Gi ¼ Gmax , Rf ¼ 1.21, c3 ¼ 0.18 8.6
Exponential [Fig. S8(b)] Gi ¼ 14.2 MPa, Rf ¼ 1.40 17

small strain region, the linear model might be suitable (Leung analytical solution of Eq. (5) developed by Mylonakis (2001b)
et al. 2010). In addition, these simplified constitutive models for perfectly end-bearing piles in linear-elastic material. This con-
are likely to break down at higher stresses, particularly in softer tinuum solution allows for a closed-form, depth-dependent attenu-
clays, so more advanced soil models (with specific treatment of ation function to be derived in the form of trigonometric series
yielding and/or plastic behavior) may be more appropriate. (Appendix I), which is indicated to depend on depth z=L, pile slen-
The parameters for the attenuation functions were then deter- derness L=d, and pile-soil stiffness ratio Ep =Es . An Es value is se-
mined by fitting the simplified functions to a more rigorous lected based on the fitted Gi of the chosen soil constitutive model

© ASCE 04022052-11 J. Geotech. Geoenviron. Eng.

J. Geotech. Geoenviron. Eng., 2022, 148(7): 04022052

 (except for the power law model for which Gmax must be selected) is calculated (Appendix I). This discrepancy is obtained from the
because much of the soil is under relatively low strain. area under the “t-z” curve from the linear constitutive model
With this in mind, the parameters of the simplified attenuation [∫ τ0max τ 0 ðuÞdτ 0 , proportional to the energy stored by the Winkler
functions in Eqs. (4), (8), (9), and (10) can be fitted. Both the con- spring] obtained with each fitted function compared with the rig-
centric cylinder model [Eq. (4)] and the power-exponential decay orous solution. The fitted parameters and the calculated discrep-
[Eq. (8)] require only one fitting parameter, rm and q, respectively, ancy are given in Table 2 for both example piles and plotted
whereas two parameters are required for the generalized concentric alongside the rigorous attenuation function in Fig. 7.
cylinder model [Eq. (9)] (m, rm ), and the generalized power- Due to the two fitting parameters, the generalized power-
exponential decay model [Eq. (10)] ðq; nÞ. exponential decay model [Eq. (10)] is observed to fit best in
These parameters, with the exception of rm , are fitted to linearly most situations. However, the simpler concentric cylinder model
spaced data points along the more rigorous attenuation function [Eq. (4)] provides a better fit for Example A at mid-depth. The gen-
using the method of nonlinear least squares. rm is calculated using eralized concentric cylinder model [Eq. (9)] provides a very similar
the following expression (Randolph and Wroth 1978): result as m ≈ 1, reducing this equation to the concentric cylinder
rm ≈ 2Lð1 − vs Þ ð51Þ model. For Example B, the power-exponential decay model
[Eq. (8)] performs similarly to the generalized power-exponential
Downloaded from ascelibrary.org by 39.40.1.215 on 03/09/23. Copyright ASCE. For personal use only; all rights reserved.

where ν s = Poisson’s ratio of the soil, taken here as 0.5. decay [Eq. (10)], while requiring only one parameter. The results in
This expression is assumed to be an appropriate estimate for rm Fig. 7 indicate that the power-exponential decay model [Eq. (8)]
for an end-bearing pile; although further work should be completed might be a better fit for compressible piles (Example B, low
to verify its suitability in this method for both the concentric cyl- Ep =Es ), whereas the concentric cylinder model [Eq. (4)] might
inder and the generalized concentric cylinder model. To assess each be a better fit for stiff piles (Example A, high Ep =Es ).
simplified attenuation function, the discrepancy (as a percentage Taking the parameters from the fitting of each soil constitutive
difference) between the fitted functions and the rigorous solution model and the simplified attenuation functions (Tables 1 and 2),

Table 2. Parameters from the fitting of each simplified attenuation function and the discrepancy compared with continuum solution by Mylonakis (2001b)
Percentage
discrepancy from
Parameter rigorous solution
Example Attenuation function z=L 0.25 0.5 0.75 0.25 0.5 0.75
Example A Concentric cylinder model [Eq. (4)] 2rm =d 100 100 100 18.3 2.4 5.7
(Ep =Es ¼ 287, L=d ¼ 50) Generalized concentric cylinder model [Eq. (9)] m 1.04 1.00 0.98 8.2 2.5 0.9
2rm =d 100 100 100
Power-exponential decay [Eq. (9)] q 0.16 0.14 0.13 22.3 26.2 28.5
Generalized power-exponential model [Eq. (10)] q 0.02 0.01 0.01 2.2 4.6 5.8
n 0.92 0.91 0.91
Example B Concentric cylinder model [Eq. (4)] 2rm =d 20 20 20 36.2 22.3 16.5
(Ep =Es ¼ 96, L=d ¼ 10) Generalized concentric cylinder model [Eq. (9)] m 1.23 1.17 1.15 0.9 4.1 5.8
2rm =d 20 20 20
Power-exponential decay [Eq. (8)] q 0.27 0.22 0.21 4.9 3.7 3.3
Generalized power- exponential decay [Eq. (10)] q 0.15 0.12 0.11 0.2 1.7 2.4
n 0.78 0.76 0.77

(a) (b)

Fig. 7. Fitting of attenuation functions at depth z=L ¼ 0.5 of (a) Example A; and (b) Example B (fitted parameters given in Table 2).

© ASCE 04022052-12 J. Geotech. Geoenviron. Eng.

J. Geotech. Geoenviron. Eng., 2022, 148(7): 04022052

 (a) (b)
Downloaded from ascelibrary.org by 39.40.1.215 on 03/09/23. Copyright ASCE. For personal use only; all rights reserved.

(c) (d)

(e) (f)

(g) (h)

Fig. 8. “t-z” curves for each illustrative example (parameters from Tables 1 and 2): (a) linear; (b) bilinear; (c) power law; (d) linear-power law;
(e) Ramberg-Osgood; (f) exponential; (g) hyperbolic; (h) modified hyperbolic.

“t-z” curves can be generated using Eqs. (19)–(50). These are plot- those predicted using the “t-z” approach and calculating the error
ted in Fig. 8 for both example piles and have been compared with bound within which 95% of the data points lie. Considering the
numerical “t-z” curves obtained from PLAXIS 2D for the linear stress range 0.2cu ≤ τ ≤ 0.8cu (as employed for fitting the constit-
and hyperbolic constitutive models, as discussed for the similar utive models), the linear and hyperbolic constitutive models re-
comparison in Fig. 3. The error in the “t-z” approach is obtained sulted in error bounds of 21% and 15%, respectively, for
by comparing the measured settlement from the FEM results to Example A and 4.0% and 3.8%, respectively, for Example B. Note

© ASCE 04022052-13 J. Geotech. Geoenviron. Eng.

J. Geotech. Geoenviron. Eng., 2022, 148(7): 04022052

 that all of these error bounds are smaller than those obtained from 5. The results from the illustrative examples indicate that the
fitting the constitutive models [Eqs. (11)–(18)], indicating that the power-exponential decay model in Eq. (8) might be a better
additional error introduced by the “t-z” assumption is very small fit for compressible piles (low Ep =Es ), whereas the concentric
than compared with the assumptions in soil parameters for model- cylinder model in Eq. (4) might be a better fit for stiff piles
ing both a single element and the spatial distribution. (high Ep =Es ).
In addition, Fig. 8 indicates that a strong variation exists in 6. A strong variation exists in the settlement at 50% of the ultimate
the normalized settlement (u0 =d) at one-half of the ultimate load, load among the various curves indicated in Fig. 8, ranging be-
ranging approximately between 0.1% and 1%. Ultimate loads are tween approximately 0.1% and 1% of pile diameter d. Ultimate
attained at settlements between 0.25% to more than 2%. These shaft loads are attained at values varying from 0.25% to more
results are problem-specific and should not be viewed as design than 2% of d. These results are clearly problem specific and
recommendations. should not be viewed as universal.
The “t-z” curves derived in this paper can be used to produce
design-orientated stress-displacement curves informed by experi-
Summary and Conclusions mental results for specific pile configurations. It should be noted
that the range of parameters obtained from the illustrative example
Downloaded from ascelibrary.org by 39.40.1.215 on 03/09/23. Copyright ASCE. For personal use only; all rights reserved.

A theoretical method for deriving “t-z” curves, employed to pre- presented in this paper should not be viewed as generic design
dict nonlinear pile settlement, has been presented. These curves recommendations but are provided as an example of obtaining
should be used in conjunction with a pressure-displacement rela- parameters for the method shown. Additionally, the method could
tion at the pile tip (“q-z” curve) to model the axial response of be experimentally verified by comparing against “t-z” curves from
floating piles. The “t-z” curves were obtained by combining an in situ pile tests. However, for a specific site, the ideal “t-z” curve
attenuation function τ = τ ðrÞ and a soil constitutive model γ = depends on the ground conditions, the pile properties, and the in-
γðτ Þ and integrating over the radial coordinate. The curves can stallation method. A systematic comparison, including relevant
be employed through hand or spreadsheet calculations or in avail- statistics, against a database of field data [e.g., the DINGO data-
able commercial software such as Oasys Pile [Oasys (2017)] or base; Vardanega et al. (2021); Voyagaki et al. (2021)] is a formi-
Ensoft TZPile [Reese et al. (2014)]. Use of these functions ena- dable task and lies beyond the scope of this work.
bles simple estimations of foundation performance without reli- As a final remark, it is fair to mention that no attempt has been
ance on numerical continuum methods in multiple dimensions, made here to identify the most suitable element test for fitting a
which are complex, time consuming, and often unavailable for theoretical τ − γ curve to experimental data, especially the shear
real soils. In addition, the method provides a much better physical strength parameter τ max (or cu ). As pile loading induces rotations
insight into the mechanics of the problem. Nevertheless, despite of principal stresses in the surrounding soil, no elementary shear
the elegance and practical appeal of the method at hand, only a test can faithfully reproduce the pertinent stress state. Evidently,
handful of such “t-z” curves are available to date—hence, the mo- the fitting process requires judgment and should be conducted
tivation for this study. on a case-by-case basis.
To this end:
1. Three novel stress attenuation functions with radial distance r
[Eqs. (8), (9), and (10)] were developed, inspired by an available Appendix I. Continuum Solution for Attenuation of
continuum solution by the corresponding author [Mylona- Shear Stress
kis (2001a)].
2. The new attenuation functions, in addition to the well-known con- A more rigorous solution to Eq. (6) can be derived by introducing
centric cylinder model (Eq. 4), were employed in conjunction the separation of variables and applying the boundary conditions of
with eight soil constitutive models [Eqs. (11)–(18)] to derive zero displacement at large radial distances, zero vertical stress at the
32 theoretical “t-z” curves [Eqs. (19)–(50)], 28 of which are, surface, and zero displacement at depth z ¼ L, which corresponds
to the authors’ knowledge, novel, and 26 are presented in closed to a perfectly end-bearing pile. Thus, the vertical displacements in
form. the soil medium can be written as an infinite trigonometric series
3. In principle, the generalized power-exponential decay model [Mylonakis (2001b)]
[Eq. (10)] can describe all other attenuation functions by using X
∞
appropriate values for parameters n and q. However, the corre- uðr; zÞ ¼ Bm K 0 ðαm ηrÞ cosðαm zÞ ð52Þ
sponding solutions in Eqs. (43)–(50) cannot describe all “t-z” m¼0
curves derived in this paper; these do not employ the empirical where Bm = coefficient to be determined; K 0 = modified Bessel
radius rm required due to the singular nature of the power function of the second kind with order 0; and αm is obtained by
law models. Moreover, simpler solutions are obtained with the imposing the requirement of zero displacement at the base of the
power law attenuation functions, whereas three of the constit- pile, as follows:
utive models are not integrable using Eq. (10).
π
4. Two illustrative example problems involving end-bearing piles cosðαm LÞ ¼ 0; αm ¼ ð2m þ 1Þ; m ¼ 0; 1 : : : ð53Þ
have been provided to demonstrate the process of employing 2L
these “t-z” curves by selecting the relevant model parameters. An attenuation function is obtained by substituting Eq. (52) into
To this end, the attenuation functions were fitted to a more rig- the stress-strain relation τ ≅ −Gs ∂u=∂r
orous continuum solution for two pile configurations, and the
soil constitutive models were fitted to triaxial test data for ∂u X∞
τ ðrÞ ≅ −Gs ¼ −Gs αm ηBm K 1 ðαm ηrÞ cosðαm zÞ ð54Þ
two materials. The most appropriate attenuation function and ∂r m¼0
constitutive model for each example was selected, which en-
abled the most suitable “t-z” curves to be calculated. All 32 Considering vertical equilibrium of the pile, assuming both a
“t-z” curves were plotted using the fitted parameters for both linear-elastic pile material and perfect bonding at the pile/soil
examples (Fig. 8). interface yields the following boundary condition:

© ASCE 04022052-14 J. Geotech. Geoenviron. Eng.

J. Geotech. Geoenviron. Eng., 2022, 148(7): 04022052

   2
π 2 ∂ 2 uðd=2; zÞ η2 ¼ ð59Þ
−Ep d ¼ πdτ ðd=2; zÞ þ FðzÞ ð55Þ 1 − νs
4 ∂z2

where FðzÞ = distributed axial force in the pile, assumed to be uni- Appendix II. Special Functions
formly distributed within the pile cross-section. This is determined
by resolving the force at the pile head into equivalent body forces The lesser known mathematical functions used in this study are de-
using a cosine series tailed as follows (Olver et al. 2010). First, the upper incomplete
gamma function
X
∞
2P
FðzÞ ¼ cosðαm zÞ ð56Þ Z ∞
m¼0
L Γðs; xÞ ¼ e−t ts−1 dt ð60Þ
x
where P = the axial force at the pile head. Eqs. (52), (54), and (56)
can be substituted into Eq. (55) and the cosine transformation This equation simplifies to the ordinary gamma function when
applied to obtain a solution for Bm x = 0. In addition, if s ¼ 1=2, this can be simplified using the com-
     plementary error function, erfcðxÞ, which is related to the standard
2P d Ep 2 αm ηd αm ηd −1 error function, erfðxÞ as indicated
Downloaded from ascelibrary.org by 39.40.1.215 on 03/09/23. Copyright ASCE. For personal use only; all rights reserved.

Bm ¼ α K þ αm K 1
πLdηGs 4η Gs m 0 2 2    
1 pffiffiffi pffiffiffi pffiffiffi
ð57Þ Γ ; x ¼ πerfc x ¼ π½1 − erfðxÞ ð61Þ
2
Substituting this back into Eq. (54) yields an analytical attenu-
The incomplete beta function is defined as
ation function for perfectly end bearing piles Z x
τ ðrÞ τ ðr; zÞ Bðx; a; bÞ ¼ ta−1 ð1 − tÞb−1 dt ð62Þ
¼ rz 0
τ0 τ rz ðd=2; zÞ
K 1 ðαm ηrÞ , Variations of the functions previously discussed are available in
X
∞
K 1 ðαm ηd=2Þ cosðαm zÞ
X∞
cosðαm zÞ
¼ common spreadsheet software (although often in the form of stat-
d Ep K 0 ðαm ηd=2Þ d Ep K 0 ðαm ηd=2Þ
m¼0 αm ð4ηÞðGs Þ K 1 ðαm ηd=2Þ þ 1 m¼0 αm ð4ηÞðGs Þ K 1 ðαm ηd=2Þ þ 1 istical distribution functions).
Finally, the lambert W function is defined as the solution to the
ð58Þ
equation
To select a value of η, specific assumptions regarding the radial WeW ¼ x ð63Þ
and tangential soil stresses and displacements must be chosen,
which are discussed in Mylonakis (2001b). The assumption of zero For the application in this paper, x ≥ 0, the principle branch is
radial stresses and zero tangential strains in the soil medium employed. This function is available in various numerical comput-
requires ing packages.

Appendix III. Soil constitutive models considered (τ ≤ τ max ).

Constitutive model Eq. Number Number of parameters Stiffness form, τ Flexibility form, γ
τ
Linear Eq. (11) 2 γG
G
8 8 τ
>
< γG
1 γ ≤ τ 1 =G1 >
<G τ ≤ τ1
 1
Bilinear Eq. (12) 4 τ τ − τ1 τ1
>
: G2 γ− 1 þ τ1 γ ≥ τ 1 =G1 >
: þ τ ≥ τ1
G1 G2 G1
   1
τ max γ b 2τ b
Power Law Eq. (13) 3 γ 50
2 γ 50 τ max
8 8 τ
>
< γGi   γ ≤ τ i =Gi >
<G τ ≤ τi
i  1
Linear-Power Law Eq. (14) 4 τ max γ b 2τ b
>
: 2 γ ≥ τ i =Gi >γ
: τ ≥ τi
γ 50 50
τ max
   
τ τ c2
Ramberg-Osgood Eq. (15) 4 — γr þ c1
τ max τ max
 −1  −1
τ max τ max τ max τ max
Hyperbolic Eq. (16) 3 þ1 −1
Rf Rf Gi γ Rf Gi Rf τ
   
τ Rf τ c3 −1
Modified Hyperbolic Eq. (17) 4 — 1−
Gi τ max
     
τ max Rf Gi τ Rf τ
Exponential Eq. (18) 3 1 − exp −γ − max ln 1 −
Rf τ max R f Gi τ max
Note: [τ i ¼ ðτ max =2Þð2Gi γ 50 =τ max Þb=ðb−1Þ ] [Eq. (14c)].

© ASCE 04022052-15 J. Geotech. Geoenviron. Eng.

J. Geotech. Geoenviron. Eng., 2022, 148(7): 04022052

 © ASCE

Appendix IV. Normalized Displacement, u 0 =d, for Each Constitutive Model Obtained Using Eqs. (4) and (9) (τ ≤ τ max )

a
Soil constitutive Concentric cylinder model
[Eq. (4)] Generalized concentric
cylinder
model [Eq. (9)]b
model τ ¼ τ 0 ðd=2rÞ τ ¼ τ 0 ðd=2rÞm
   1−m 
τ0 2rm τ0 2rm
Linear [Eq. (11)] ln 1−
2G d 2Gðm − 1Þ d
Downloaded from ascelibrary.org by 39.40.1.215 on 03/09/23. Copyright ASCE. For personal use only; all rights reserved.

8   8    
>
> τ0 2rm >
> τ 0 2rm 1−m
>
< 2G ln τ0 ≤ τ1 >
< 2G ðm − 1Þ 1 − τ0 ≤ τ1
1 d 1 d
Bilinear             1−m    1−m 
>
> τ 0 ðG1 − G2 Þ G2 2rm τ 1 G1 τ τ >
> τ 0 ðG1 − G2 Þ 1 G1 G2 2rm 1−m τ τ τ
[Eq. (12)] >
: ln þ ln 0 þ 1 − 1 τ0 ≥ τ1 >
: − − 0
m
þ 1− 0
m
τ0 ≥ τ1
2G1 G2 G1 − G2 d τ0 G1 − G2 τ1 τ0 2G1 G2 m − 1 G1 − G2 G1 − G2 d τ1 τ0 τ1
 1  1
γ 50 b 2τ 0 b γ 50 b 2τ 0 b
Power Law where b < 1 where b < m
[(Eq. (13)] 2ð1 − bÞ τ max 2ðm − bÞ τ max
8   8    
>
> τ0 2rm >
> τ0 2rm 1−m
>
< 2G ln d τ0 ≤ τi >
< 2G ðm − 1Þ 1 − τ0 ≤ τi
i i d
Linear-Power  1   b−1     1   b−m   1−m  1−m 
>
> γ 50 b 2τ 0 b 2ri b τ r >
> γ 50 b 2τ 0 b 2ri b τ0 2ri 2rm
Law [Eq. (14)] >
: 1− þ 0 ln m τ0 ≥ τi >
: 1− þ − τ0 ≥ τi
2ð1 − bÞ τ max d 2Gi ri 2ðm − bÞ τ max d 2Gi ðm − 1Þ d d
 b   b
2r 2τ 0 2γ 50 Gi 1−b 2r 2τ 0 2γ 50 Gi 1−b 1=m
where i ¼ where i ¼
d τ max τ max d τ max τ max
                 
τ 0 γr 2rm γr c1 τ 0 c2 2rm 1−c2 τ 0 γr 2rm 1−m γr c1 τ 0 c2 2rm 1−c2 m
Ramberg- ln þ 1− 1− þ 1−
04022052-16

Osgood 2τ max d 2ðc2 − 1Þ τ max d 2τ max ðm − 1Þ d 2ðc2 m − 1Þ τ max d

[Eq. (15)]       1    m   
τ0 2rm Rf τ 0 Rf τ 0 τ max Rf τ 0 m τ 2rm 1 τ 1
Hyperbolic ln − − ln 1 − B max ; ; 0 − B max ; ; 0
[Eq. (16)] 2Gi d τ max τ max 2Rf mGi τ max Rf τ 0 d m Rf τ 0 m
  c  c                
τ0 2rm 3 Rf τ 0 3 Rf τ 0 c3 τ max Rf τ 0 m1 Rf τ 0 c3 m − 1 Rf τ 0 d m c3 m − 1
Modified ln − − ln 1 − B ; ;0 − B ; ;0
hyperbolic 2Gi c3 d τ max τ max 2Rf Gi c3 m τ max τ max c3 m τ max 2rm c3 m
[Eq. (17)]             
τ max Rf τ 0 2r Rf τ 0 d τ max Rf τ 0 2r Rf τ 0 d m
Exponential ln 1 − − m ln 1 − ln 1 − − m ln 1 −
2Rf Gi τ d τ 2rm 2Rf Gi τ d τ 2r
[Eq. (18)]   max   max     max  max  m m 
Rf τ 0 2rm Rf τ 0 Rf τ 0 Rf τ 0 m1 τ max 1 τ max 2rm 1
þ ln − − ln 1 − þ B ; ;0 − B ; ;0
τ max d τ max τ max τ max Rf τ 0 m Rf τ 0 d m
Note: Constitutive model parameters are given in Eqs. (11)–(18); rm = empirical radius [Randolph and Wroth (1978)]; and Bðx; a; bÞ = incomplete Beta function (see Appendix II).
a
“t-z” curves available in the literature with the concentric cylinder model [Linear: Randolph and Wroth (1978), Power Law: Vardanega et al. (2012); Williamson (2014); Vardanega (2015); Crispin et al. (2019);
Hyperbolic: Kraft et al. (1981); Modified Hyperbolic: Chang and Zhu (1998)].
b
For m > 1, all terms ð2rm =dÞ1−m in the generalized concentric cylinder model (right column) vanish when rm approaches infinity, which simplifies the solutions.
J. Geotech. Geoenviron. Eng.

J. Geotech. Geoenviron. Eng., 2022, 148(7): 04022052

 Appendix V. Normalized Displacement, u 0 =d , for Each Constitutive Model Obtained Using Eqs. (8) and (10)
(τ ≤ τ max )

Soil constitutive Power-exponential decay [Eq. (8)]

Generalized
power-exponential decay [Eq.
(10)]
model τ ¼ τ 0 ð2r=dÞ −1=2 expð−qð2r=d − 1ÞÞ τ ¼ τ 0 ð2r=dÞ−n expð−qð2r=d − 1ÞÞ
pffiffiffi τ 0 q n−1
τ 0 eq π pffiffiffi e q Γð1 − n; qÞ
Linear [Eq. (11)] pffiffiffi erfcð qÞ 2G
2G q
8 pffiffiffi   8
> τ 0 eq π pffiffiffi > τ 0 q n−1
> >
>
< 2G1 q
p ffiffiffi erfc q τ0 ≤ τ1 < 2G1 e q Γð1 − n; qÞ
> τ0 ≤ τ1
Bilinear  pffiffiffi     
>
> τ 0 ðG1 − G2 Þ eq π τ 1 2r1 >
> τ 0 ðG1 − G2 Þ q n−1 τ 2r1
[Eq. (12)] >
: pffiffiffi X 1 − −1 τ0 ≥ τ1 >
: 2G G e q X1 − 1 −1 τ ≥ τ1
2G1 G2 q τ0 d 1 2 τ0 d
  rffiffiffiffiffiffiffiffiffiffi  
G1 pffiffiffi 2qr1 G1 2r
where X 1 ¼ erfc q − erfc where X 1 ¼ Γð1 − n; qÞ − Γ 1 − n; q 1
G1 − G2 d G1 − G2 d
  2    1 
Downloaded from ascelibrary.org by 39.40.1.215 on 03/09/23. Copyright ASCE. For personal use only; all rights reserved.

2r1 1 τ 2r1 n q q τ n
¼ W 2qe2q 0 ¼ W en 0
d 2q τ1 d q n τ1
 2b−1  1    b−n  1  
γ 50 q b 2b 2τ 0 b 2b − 1 q γ 50 q b b 2τ 0 b b − n q
Power Law eb Γ ; eb Γ ;
[Eq. (13)] 2 q τ max 2b b 2 q τ max b b
8 pffiffiffi   8
> τ 0 eq π p ffiffiffi > τ 0 q n−1
> >
>
< 2Gi q
pffiffiffi erfc q τ0 ≤ τi < 2Gi e q Γð1 − n; qÞ
> τ ≤ τi
Linear-Power  2b−1  1    b−n  1  
>
> γ q b 2b 2τ 0 b τ eq 1 2r >
> γ 50 q b b 2τ 0 b τ 2r
Law [Eq. (14)] >
: 50 eb Γg þ 0 pffiffiffi Γ ; q i τ0 ≥ τi >
: 2 eb q Γg þ 0 eq qn−1 Γ 1 − n; q i τ ≥ τi
2 q τ max 2Gi q 2 d τ max 2Gi d
       
2b − 1 q 2b − 1 2ri q b−n q b − n 2ri q
where Γg ¼ Γ ; −Γ ; where Γg ¼ Γ ; −Γ ;
2b b 2b d b b b b d b
  2   2b      b 1 
2ri 1 2τ 0 2γ 50 Gi 1−b 2ri n q q 2τ 0 2γ 50 Gi 1−b n
¼ W 2qe2q ¼ W en
d 2q τ max τ max d q n τ max τ max
       
τ 0 γ r eq 1 γ r c2 q c2 −2 c1 τ 0 c2 2 − c2 τ 0 γ r q n−1 γ ec2 q c τ c2
Ramberg- pffiffiffi Γ ; q þ e ðqc2 Þ 2 Γ ; qc2 e q Γð1 − n; qÞ þ r ðqc2 Þc2 n−1 1 0 Γð1 − c2 n; qc2 Þ
Osgood 2τ max q 2 2 τ max 2 2τ max 2 τ max
[Eq. (15)]  rffiffiffiffiffi    −1       −1
τ max ∞ τ max 2r 2r τ max ∞ τ max 2r n 2r
Hyperbolic ∫ d=2 exp q −1 − 1 dr ∫ d=2 exp q −1 − 1 dr
Rf dGi Rf τ 0 d d Rf dGi Rf τ 0 d d
[Eq. (16)] qffiffiffiffi     n   
d
exp −q 2r −1 exp −q 2rd −1
d
τ0 ∞ 2r d τ0 ∞ 2r
Modified ∫  qffiffiffiffi   c dr ∫      c dr
Gi d d=2 Rf τ 0 2r
3 Gi d d=2 Rf τ 0 d n 2r
3
hyperbolic 1 − τ max d
2r exp −q d − 1 1 − τ max 2r exp −q d − 1
[Eq. (17)]
 rffiffiffiffiffi         
τ max ∞ Rf τ 0 d 2r τ max ∞ Rf τ 0 d n 2r
Exponential − ∫ d=2 ln 1 − exp −q −1 dr − ∫ d=2 ln 1 − exp −q −1 dr
R f Gi d τ max 2r d Rf Gi d τ max 2r d
[Eq. (18)]
Note: Constitutive model parameters are given in Eqs. (11)–(18); erfcðxÞ = complementary error function; Γðs; xÞ = upper incomplete gamma function; and
WðxÞ = Lambert W-Function (see Appendix II). X 1 and Γg are parameters defined for convenience.

Data Availability Statement Notation

All data and models generated or used during the study appear in The following symbols are used in this paper:
the published article. All of the code that supports the findings of Bm = integration constant;
this study is available from the corresponding author on reasonable b = soil nonlinearity exponent;
request. cu = soil undrained shear strength;
c1 , c2 , c3 = fitting constants from constitutive models;
d = pile diameter;
Acknowledgments Ep , Es = stiffness of pile and soil, respectively;
FðzÞ = distributed axial force in pile;
The first and second authors would like to thank the Engineering
G, Gi = linear shear modulus, initial (low strain) shear
and Physical Sciences Research Council for their support (Grant modulus in constitutive models;
Nos. EP/T517872/1 and EP/N509619/1, respectively). Partial fund- Gs , Gmax = soil shear modulus, measured maximum (low strain)
ing was received by EU/H2020 under grant agreement number soil shear modulus;
730900 (SERA) with George Mylonakis as Principal Investigator G1 , G2 = low-strain shear modulus and reduced shear modulus
for University of Bristol. in bilinear constitutive model;

© ASCE 04022052-17 J. Geotech. Geoenviron. Eng.

J. Geotech. Geoenviron. Eng., 2022, 148(7): 04022052

 K v = modified Bessel function of second kind with order v; Basu, D., and R. Salgado. 2008. “Analysis of laterally loaded piles with
kðτ 0 Þ = nonlinear secant Winkler modulus; rectangular cross sections embedded in layered soil.” Int. J. Numer.
Anal. Methods Geomech. 32 (7): 721–744. https://doi.org/10.1002
L = pile length;
/nag.639.
Ms = soil compressibility modulus; Bateman, A. H. 2019. Theoretical “t-z” curves for axially loaded piles.
m = stress attenuation fitting parameter; Undergraduate Research Rep. No. 1819RP027. Bristol, UK: Dept. of
n = stress attenuation fitting parameter; Civil Engineering, Univ. of Bristol.
N = number of data points; Bateman, A. H., and J. J. Crispin. 2020. “Theoretical ‘t-z’ curves for piles in
P = axial force acting on pile head; radially inhomogeneous soil.” DFI J. 14 (1): 1–9. https://doi.org/10
q = shear attenuation fitting parameter; .37308/DFIJnl.20190930.210.
Beesley, M. E. W., and P. J. Vardanega. 2020. “Parameter variability of
Rf = fitting constant to scale τ max in constitutive models;
undrained shear strength and strain using a database of reconstituted
r, ri , r1 = radial distance from pile centerline, radius at which τ i soil tests.” Can. Geotech. J. 57 (8): 1247–1255. https://doi.org/10
occurs, radius at which τ 1 occurs; .1139/cgj-2019-0424.
rm = empirical radius beyond which vertical settlement is Bhattacharya, S. 2019. Design of foundations for offshore wind turbines.
assumed negligible; Hoboken, NJ: Wiley.
Downloaded from ascelibrary.org by 39.40.1.215 on 03/09/23. Copyright ASCE. For personal use only; all rights reserved.

u, u0 = vertical displacement, vertical displacement at pile- Brinkgreve, R. B. J., L. Zampich, and N. Ragi Manoj. 2019. PLAXIS con-
soil interface, respectively; nect edition V20. Delft, Netherlands: Plaxis bv.
x = real variable; Chakraborty, T., R. Salgado, P. Basu, and M. Prezzi. 2013. “Shaft resistance
of drilled shafts in clay.” J. Geotech. Geoenviron. Eng. 139 (4):
z = depth below ground level;
548–563. https://doi.org/10.1061/(ASCE)GT.1943-5606.0000803.
α = empirical adhesion factor; Chang, M. F., and H. Zhu. 1998. “A t-z curve with consideration of modulus
αm = eigenvalue parameter; degradation for pile analyses.” In Proc., 13th Southeast Asian Geotech-
γ, γ r , γ 50 = soil shear strain, reference shear strain, strain when ½ nical Conf., 461–466. Reston, VA: ASCE.
the peak shear stress is mobilized, respectively; Chow, Y. K. 1986. “Analysis of vertically loaded pile groups.” Int. J. Nu-
η = compressibility coefficient; mer. Anal. Methods Geomech. 10 (1): 59–72. https://doi.org/10.1002
vs = Poisson’s ratio of soil; /nag.1610100105.
σ = vertical normal stress acting on soil slice; Cooke, R. W. 1974. “The settlement of friction pile foundations.” In Proc.,
Conf. on Tall Buildings, 7–19. Kuala Lumpur, Malaysia: Institution of
τ , τ 0 = shear stress, shear stress at pile-soil interface,
Engineers.
respectively; Coyle, H. M., and L. C. Reese. 1966. “Load transfer for axially loaded piles
τ i , τ 1 = shear stress at change in stiffness in linear-power law in clay.” J. Soil Mech. Found. Div. 92 (2): 1–26. https://doi.org/10.1061
and bilinear models, respectively; and /JSFEAQ.0000850.
τ max = shear strength of soil. Coyle, H. M., and I. H. Sulaiman. 1967. “Skin friction for steel piles in
sand.” J. Soil Mech. Found. Div. 93 (6): 261–278. https://doi.org/10
.1061/JSFEAQ.0001055.
Supplemental Materials Crispin, J. J., C. P. Leahy, and G. Mylonakis. 2018. “Winkler model for
axially loaded piles in inhomogeneous soil.” Géotech. Lett. 8 (4):
Figs. S1–S8 and Tables S1–S8 are available online in the ASCE 290–297. https://doi.org/10.1680/jgele.18.00062.
Library (www.ascelibrary.org). Crispin, J. J., and G. Mylonakis. 2021. “Simplified models for axial static and
dynamic analysis of pile foundations.” In Analysis of pile foundations
subject to static and dynamic loading, edited by A. Kaynia, 131–184.
References Boca Raton, FL: CRC Press.
Crispin, J. J., P. J. Vardanega, and G. Mylonakis. 2019. “Prediction of pile
Abchir, Z., R. Burlon, J. Frank, S. Habert, and S. Legrand. 2016. “t-z curves settlement using simplified models.” In Proc., XVII European Conf. on Soil
for piles from pressuremeter test results.” Géotechnique 66 (2): Mechanics and Geotechnical Engineering, edited by H. Sigursteinsson,
137–148. https://doi.org/10.1680/jgeot.15.P.097. S. Erlingsson, and B. Bessason. Reykjavik, Iceland: Icelandic Geotechnical
Akiyoshi, T. 1982. “Soil-pile interaction in vertical vibration induced Society. https://doi.org/10/gmfpn4.
through a frictional interface.” Earthquake Eng. Struct. Dyn. 10 (1): Davis, R. O., and A. P. S. Selvadurai. 1996. Elasticity and geomechanics.
135–148. https://doi.org/10.1002/eqe.4290100110. Cambridge, UK: Cambridge University Press.
Anoyatis, G., and G. Mylonakis. 2012. “Dynamic Winkler modulus for El Naggar, M. H., and M. Novak. 1994. “Non-linear model for dynamic
axially loaded piles.” Géotechnique 62 (6): 521–536. https://doi.org/10 axial pile response.” J. Geotech. Eng. 120 (2): 308–329. https://doi.org
.1680/geot.11.P.052. /10.1061/(ASCE)0733-9410(1994)120:2(308).
Anoyatis, G., G. Mylonakis, and A. Tsikas. 2019. “An analytical con- Fahey, M., and J. P. Carter. 1993. “A finite element study of the pressure-
tinuum model for axially loaded end-bearing piles in inhomogeneous
meter test in sand using a non-linear elastic-plastic model.” Can. Geo-
soil.” Int. J. Numer. Anal. Methods Geomech. 43 (6): 1162–1183.
tech. J. 30 (2): 348–362. https://doi.org/10.1139/t93-029.
https://doi.org/10.1002/nag.2886.
Fleming, K., A. Weltman, M. Randolph, and K. Elson. 2009. Piling engi-
API (American Petroleum Institute) and ANSI (American National
Standards Institute). 2011. Part 4—Geotechnical and foundation de- neering. 3rd ed. Abingdon, UK: Taylor and Francis.
sign considerations: ANSI/API recommended practice 2GEO. 1st ed. Frank, R., and S. R. Zhao. 1982. “Estimation par les paramètres pressio-
Washington, DC: API. métriques de l’enfoncement sous charge axiale de pieux forés dans des
Armaleh, S., and C. S. Desai. 1987. “Load-deformation response of axially sols fins.” [In French.] Bull. Liaison Lab. Ponts Chaussées 119 (May/
loaded piles.” J. Geotech. Eng. 113 (12): 1483–1500. https://doi.org/10 Jun): 17–24.
.1061/(ASCE)0733-9410(1987)113:12(1483). Gazetas, G., P. Hess, R. Zinn, G. Mylonakis, and A. Nikolaou. 1998.
Baguelin, F., and R. Frank. 1979. “Theoretical studies of piles using “Seismic response of a large pile group.” In Proc., 11th European Conf.
the finite element method.” In Numerical methods in offshore piling, on Earthquake Engineering, 10. Rotterdam, Netherlands: A.A. Balkema.
83–91. London: Institution of Civil Engineers. CD-ROM.
Baranov, V. A. 1967. On the calculation of excited vibrations of an em- Guo, W. D. 2012. Theory and practice of pile foundations. Boca Raton, FL:
bedded foundation. [In Russian.] Riga, Latvia: Polytechnic Institute. CRC Press.

© ASCE 04022052-18 J. Geotech. Geoenviron. Eng.

J. Geotech. Geoenviron. Eng., 2022, 148(7): 04022052

 Guo, W. D., and F. H. Lee. 2001. “Load transfer approach for laterally Novak, M. 1974. “Dynamic stiffness and damping of piles.” Can. Geotech.
loaded piles.” Int. J. Numer. Anal. Methods Geomech. 25 (11): J. 11 (4): 574–598. https://doi.org/10.1139/t74-059.
1101–1129. https://doi.org/10.1002/nag.169. Oasys. 2017. Pile oasys geo suite for windows. London: Oasys Ltd.
Guo, W. D., and M. F. Randolph. 1998. “Rationality of load transfer ap- Olver, F. W. J., D. W. Lozier, R. F. Boisvert, and C. W. Clark. 2010. NIST
proach for pile analysis.” Comput. Geotech. 23 (1–2): 85–112. https:// handbook of mathematical functions. New York: Cambridge University
doi.org/10.1016/S0266-352X(98)00010-X. Press.
Ishihara, K. 1996. Soil behaviour in earthquake geotechnics. New York: O’Neill, M. W. 2001. “Side resistance in piles and drilled shafts.” J. Geo-
Oxford Science Publications. tech. Geoenviron. Eng. 127 (1): 1–16. https://doi.org/10.1061/(ASCE)
Kalinski, M. E., A. Ata, K. H. Stokoe, and M. W. O’Neill. 2001. “Use of 1090-0241(2001)127:1(3).
SASW measurements to evaluate the effect of line slurry conditioning Poulos, H. G. 2017. Tall building foundation design. Boca Raton, FL: CRC
in drilled shafts.” J. Environ. Eng. Geophys. 6 (4): 147–156. https://doi Press.
.org/10.4133/JEEG6.4.147. Poulos, H. G., and E. H. Davis. 1980. Pile foundation analysis and design.
Khalili-Tehrani, P., E. R. Ahlberg, C. Rha, A. Lemnitzer, J. P. Stewart, New York: Wiley.
E. Taciroglu, and J. W. Wallace. 2014. “Nonlinear load-deflection Psaroudakis, E. G., G. Mylonakis, and N. Klimis. 2019. “Non-linear
behaviour of reinforced concrete drilled piles in stiff clay.” J. Geotech. analysis of axially loaded piles using ‘t-z’ and ‘q-z’ curves.” Geotech.
Geoenviron. Eng. 140 (3): 04013022. https://doi.org/10.1061/(ASCE) Geol. Eng. 37 (4): 2293–2302. https://doi.org/10.1007/s10706-019
Downloaded from ascelibrary.org by 39.40.1.215 on 03/09/23. Copyright ASCE. For personal use only; all rights reserved.

GT.1943-5606.0000957. -00823-2.
Kondner, R. L. 1963. “Hyperbolic stress strain response: Cohesive soils.” Ramberg, W., and W. R. Osgood. 1943. Description of stress-strain curves
J. Soil Mech. Found. Div.: Proc. 89 (1): 115–143. https://doi.org/10 by three parameters. Technical Note No. 902. Washington, DC:
.1061/JSFEAQ.0000479. National Advisory Committee for Aeronautics.
Kraft, L. M., R. P. Ray, and T. Kagawa. 1981. “Theoretical t-z curves.” Randolph, M. F. 1994. “Design methods for pile groups and piled rafts.”
J. Geotech. Eng. Div. 107 (11): 1543–1561. https://doi.org/10.1061 In Proc., 13th Int. Conf. of Soil Mechanics and Foundation Engi-
/AJGEB6.0001207. neering, 61–82. New Delhi, India: Oxford and IBH Publishing
Lam, I. P., and G. R. Martin. 1986. Seismic design of highway bridge Company.
foundations. Volume II: Design procedures and guidelines. Rep. No. Randolph, M. F., and C. P. Wroth. 1978. “Analysis of deformation of
FHWA/RD-86-102. Washington, DC: US DOT, Federal Highway vertically loaded piles.” J. Geotech. Eng. Div. 104 (12): 1465–1488.
Administration. https://doi.org/10.1061/AJGEB6.0000729.
Lee, K., and Z. R. Xiao. 1999. “A new analytical model for settlement Randolph, M. F., and C. P. Wroth. 1979. “An analysis of the vertical de-
analysis of a single pile in multi-layered soil.” Soils Found. 39 (5): formation of pile groups.” Géotechnique 29 (4): 423–439. https://doi
131–143. https://doi.org/10.3208/sandf.39.5_131. .org/10.1680/geot.1979.29.4.423.
Leung, Y. F., K. Soga, M. Lehane, and A. Klar. 2010. “Role of linear Reese, L. C., B. S. Hudson, and B. S. Vijayvergija. 1969. “An investigation
elasticity in pile group analysis and load test interpretation.” J. Geo- of the interaction between bored piles and soil.” In Vol. 2 of Proc., 7th
tech. Geoenviron. Eng. 136 (12): 1686–1694. https://doi.org/10.1061 Int. Conf., of Soil Mechanical and Foundation Engineering, 211–215.
/(ASCE)GT.1943-5606.0000392. Mexico City, Mexico: Sociedad Mexicana de Mecanica.
Matlock, H. 1970. “Correlations for design of laterally loaded piles in soft Reese, L. C., and W. F. Van Impe. 2011. Single piles and pile groups under
clay.” In Proc., 2nd Offshore Technology Conf., 22–24. Houston: Off- lateral loading. 2nd ed. London: CRC Press.
shore Technology Conference. https://doi.org/10.4043/1204-MS. Reese, L. C., S. T. Wang, and J. A. Arrellaga. 2014. “TZPILE 2014 User’s
Mayne, P. W. 1985. “Stress anisotropy effects on clay strength.” J. Geotech. Manual—Analysis of load versus settlement for an axially loaded deep
Eng. Div. 111 (3): 356–366. https://doi.org/10.1061/(ASCE)0733-9410 foundation. Austin, TX: Ensoft Inc.
(1985)111:3(356). Salgado, R. 2008. The engineering of foundations. New York: McGraw-Hill.
McClelland, B., and J. A. Focht. 1956. “Soil modulus for laterally loaded Scott, R. F. 1981. Foundation analysis. Englewood Cliffs, NJ: Prentice-Hall.
piles.” J. Soil Mech. Found. Div. 82 (4): 1081. https://doi.org/10.1061 Seed, H. B., and L. C. Reese. 1957. “The action of soft clay along friction
/JSFEAQ.0000023. piles.” Trans. Am. Soc. Civ. Eng. 122 (1): 731–754. https://doi.org/10
Meyerhof, G. G. 1976. “Bearing capacity and settlement of pile foundations.” .1061/TACEAT.0007501.
J. Geotech. Eng. Div. 102 (3): 195–228. https://doi.org/10.1061/AJGEB6 Seo, H., D. Basu, M. Prezzi, and R. Salgado. 2009. “Load-settlement
.0000243. response of rectangular and circular piles in multi-layered soil.” J. Geo-
Michaelides, O., G. Gazetas, G. Bouckovalas, and E. Chrysikou. 1998. tech. Geoenviron. Eng. 135 (3): 420–430. https://doi.org/10.1061/(ASCE)
“Approximate non-linear dynamic axial response of piles.” Géotechni- 1090-0241(2009)135:3(420).
que 48 (1): 33–53. https://doi.org/10.1680/geot.1998.48.1.33. Shadlou, M., and S. Bhattacharya. 2014. “Dynamic stiffness of pile in a
Motta, E. 1994. “Approximate elastic-plastic solution for axially loaded layered elastic continuum.” Géotechnique 64 (4): 303–319. https://doi
piles.” J. Geotech. Eng. 120 (9): 1616–1624. https://doi.org/10.1061 .org/10.1680/geot.13.P.107.
/(ASCE)0733-9410(1994)120:9(1616). Skempton, A. W. 1959. “Cast in-situ bored piles in London clay.”
Mylonakis, G. 1995. “Contributions to static and seismic analysis Géotechnique 9 (4): 153–173. https://doi.org/10.1680/geot.1959.9.4.153.
of piles and pile-supported bridge piers.” Ph.D. thesis, Dept. of Civil Soga, K. 1994. “Mechanical behaviour and constitutive modelling of
Engineering, State Univ. of New ’York. natural structured soils.” Ph.D. thesis, Univ. of California at Berkeley.
Mylonakis, G. 2000. “‘A new analytical model for settlement analysis of a Su, H., N. Psyrras, J. J. Crispin, D. Karamitros, and G. Mylonakis. 2019.
single pile in multi-layered soil’, Discussion to Paper by K.M. Lee and “Transmission of foundation vibrations.” In Proc., Society for Earth-
Z.R. Xiao.” Soils Found. 40 (4): 163–166. quake and Civil Engineering Dynamic Conf., Paper 22.7, 10. London:
Mylonakis, G. 2001a. “Elastodynamic model for large-diameter end- Society for Earthquake and Civil Engineering Dynamics.
bearing shafts.” Soils Found. 41 (3): 31–44. https://doi.org/10.3208 Tomlinson, M. J. 1957. “The adhesion of piles driven in clay soils.” In
/sandf.41.3_31. Proc., 4th Int. Conf., on Soil Mechanical Foundation Engineering,
Mylonakis, G. 2001b. “Winkler modulus for axially loaded piles.” Géotech- 66–71. London: Butterworths Scientific Publications.
nique 51 (5): 455–461. https://doi.org/10.1680/geot.2001.51.5.455. Vallabhan, C. V., and G. Mustafa. 1996. “A new model for the analysis of
Mylonakis, G., and G. Gazetas. 1998. “Settlement and additional internal settlement of drilled piers.” Int. J. Numer. Anal. Methods Geomech.
forces of grouped piles in layered soil.” Géotechnique 48 (1): 55–72. 20 (2): 143–152. https://doi.org/10.1002/(SICI)1096-9853(199602)
https://doi.org/10.1680/geot.1998.48.1.55. 20:2<143::AID-NAG812>3.0.CO;2-U.
Nogami, T., and M. Novak. 1976. “Soil-pile interaction in vertical vibra- Vardanega, P. J. 2015. “Sensitivity of simplified pile settlement calculations
tion.” Earthquake Eng. Struct. Dyn. 4 (3): 277–293. https://doi.org/10 to parameter variation in stiff clay.” In Proc., 16 ECSMGE, 3777–3782.
.1002/eqe.4290040308. London: ICE Publishing.

© ASCE 04022052-19 J. Geotech. Geoenviron. Eng.

J. Geotech. Geoenviron. Eng., 2022, 148(7): 04022052

 Vardanega, P. J., and M. D. Bolton. 2011. “Strength mobilization in clays Vlasov, V. Z., and N. N. Leontiev. 1966. Beams, plates and shells on elastic
and silts.” Can. Geotech. J. 48 (10): 1485–1503. https://doi.org/10.1139 foundations translated from Russian by the Israel program for scientific
/t11-052, [corrigendum 49(5): 631]. translations. NIST No. N67-14238. Jerusalem, Israel: Israel Program
Vardanega, P. J., E. Voyagaki, J. J. Crispin, C. E. L. Gilder, and K. Ntassiou. for Scientific Translations.
2021. The DINGO database: Summary report: April 2021, v1.1. Voyagaki, E., J. J. Crispin, C. E. L. Gilder, K. Ntassiou, N. O’Riordan,
Bristol, UK: Univ. of Bristol. https://doi.org/10.5523/bris.89r3npvewe P. Nowak, T. Sadek, D. Patel, G. Mylonakis, and P. J. Vardanega.
l2ea8ttb67ku4d. 2021. “The DINGO database of axial pile load tests for the UK:
Vardanega, P. J., M. G. Williamson, and M. D. Bolton. 2012. “Bored pile Settlement prediction in fine-grained soils.” Georisk: Assess. Manage.
design in stiff clay II: Mechanisms and uncertainty.” Proc. Inst. Civ. Risk Eng. Syst. Geohazards https://doi.org/10.1080/17499518.2021
Eng. Geotech. Eng. 165 (4): 233–246. https://doi.org/10.1680/geng .1971249.
.11.00063, [corrigendum 166(5): 518]. Williamson, M. G. 2014. “Tunnelling effects on bored piles in clay.” Ph.D.
Viggiani, C., A. Mandolini, and G. Russo. 2012. Piles and pile foundations. thesis, Dept. of Engineering, Univ. of Cambridge.
Boca Raton, FL: CRC Press. Zhu, H., and M. F. Chang. 2002. “Load transfer curves along bored piles
Vijayvergiya, V. N. 1977. “Load movement characteristics of piles.” In considering modulus degradation.” J. Geotech. Geoenviron. Eng.
Proc., Ports’77: The 4th Annual Symp. of the Waterway, Port, Coastal 128 (9): 764–774. https://doi.org/10.1061/(ASCE)1090-0241(2002)
and Ocean Division, 269–284. Reston, VA: ASCE. 128:9(764).
Downloaded from ascelibrary.org by 39.40.1.215 on 03/09/23. Copyright ASCE. For personal use only; all rights reserved.

© ASCE 04022052-20 J. Geotech. Geoenviron. Eng.

J. Geotech. Geoenviron. Eng., 2022, 148(7): 04022052