You are on page 1of 15

947

ARTICLE
Analytical solution for radial consolidation considering soil
structure characteristics
Cholachat Rujikiatkamjorn and Buddhima Indraratna

Abstract: A system of surcharge load combined with vertical drains to speed up consolidation of soft soil by reducing the
drainage path is one of the most efficient and economical ground improvement techniques. In the field, conventional theories
including smear zone have been commonly employed to predict the radial consolidation behaviour induced by vertical drains
in soft clay. One of the key parameters in conventional analysis is the use of mean coefficient of volume compressibility and soil
permeability, which are often assumed to be constant. The effect of drain installation on the soil compressibility of the in situ
clay structure is often ignored. Laboratory testing has shown that the soil compressibility and permeability can vary nonlinearly
over a considerable range of applied surcharge pressure, and both these properties can be affected during the drain installation.
This study presents a mathematical model of radial consolidation via vertical drains incorporating the variations of soil
compressibility and permeability as well as highlighting the effects of drain installation on those parameters. The main
differences between the proposed and conventional models are elucidated, in terms of stress history and preloading (surcharge)
pressure. The effects of preconsolidation pressure and the magnitude of applied preloading are examined through the dissipa-
tion of average excess pore pressure and associated settlement. Supported by experimental observations, the proposed theory is
validated with field data of a selected case study in the town of Ballina, New South Wales, Australia.

Key words: prefabricated vertical drains, mandrel installation, smear zone.

Résumé : L’une des techniques les plus efficaces et économiques pour améliorer le sol est un système de surcharge combiné avec
des drains verticaux, qui permet d’accélérer la consolidation d’un sol mou en réduisant le parcours de drainage. Sur le terrain,
les théories conventionnelles, incluant la zone souillée, ont été couramment utilisées pour prédire le comportement en consol-
idation radiale induit par les drains verticaux dans l’argile molle. Les paramètres clés dans l’analyse conventionnelle sont les
coefficients moyens de compressibilité volumique et de perméabilité du sol, qui sont souvent supposés constants. L’effet de
l’installation du drain sur la compressibilité du sol et de sa structure d’argile in-situ est souvent ignoré. Des essais en laboratoire
ont démontré que la compressibilité et la perméabilité du sol peut varier de façon non linéaire sur une gamme considérable de
pressions de surcharge appliquée, et ces deux propriétés peuvent être affectées lors de l’installation du drain. Cette étude
présente un modèle mathématique de la consolidation radiale par des drains verticaux, qui inclus les variations de la compress-
ibilité et de la perméabilité du sol, ainsi que les effets de l’installation sur ces paramètres. Les principales différences entre le
modèle proposé et les modèles conventionnels sont présentées en termes d’historique des contraintes et de pression de
pré-chargement (surcharge). Les effets de la pression de pré-consolidation et la magnitude de la pré-charge appliquée sont
examinés par la dissipation de la pression interstitielle moyenne et par le tassement associé. La théorie proposée est supportée
par des observations expérimentales et validée avec des données de terrain provenant d’un site d’étude dans la ville de Ballina,
New South Wales, en Australie. [Traduit par la Rédaction]

Mots-clés : drains verticaux pré-fabriqués, installation de mandrin, zone souillée.

Introduction pression. Studies of the clay structure have shown that geotech-
nical properties including permeability and compressibility of
Much of the coastal regions of Australia and Southeast Asia
such clays are different in the horizontal and vertical directions
contain very soft clays (estuarine or marine), which have unfa-
(Gibson and Lo 1961; Randolph and Wroth 1979; Yin et al. 2010). In
vourable geotechnical properties such as low bearing capacity and
addition to overburden pressure, there are other factors that have
high compressibility. In the absence of appropriate ground im- been found to influence the anisotropy amongst which are the
provement, excessive settlement and lateral movement can ad- nature of deposition, overconsolidation, stress history, and ce-
versely affect the stability of surface infrastructure including mentation bonds.
highway and rail embankments. The constraints of restricted The traditional preloading method by a surcharge fill embank-
space, environmental and safety issues, and the desired longevity ment (facilitated by prefabricated vertical drains, PVDs) is gener-
of earth structures have continued to demand innovation in the ally an economical solution (Hansbo 1981; Indraratna and Redana
design and construction of essential infrastructure built on soft 2000; Rujikiatkamjorn and Indraratna 2010). However, in thick
clays. Almost all natural (structured) soil deposits are intrinsically soft soil sites, consolidation time can be overly lengthy due to the
anisotropic (Bishop 1966), and this is attributed to the process of very low soil permeability and the lack of efficient drainage. The
sedimentation followed by predominantly one-dimensional com- installation of PVDs is known to aid the rapid dissipation of pore-

Received 15 July 2014. Accepted 10 November 2014.


C. Rujikiatkamjorn and B. Indraratna. Centre for Geomechanics and Railway Engineering, School of Civil Engineering, Faculty of Engineering and
Information Sciences, University of Wollongong, Wollongong City, NSW 2522, Australia.
Corresponding author: Buddhima Indraratna (e-mail: indra@uow.edu.au).

Can. Geotech. J. 52: 947–960 (2015) dx.doi.org/10.1139/cgj-2014-0277 Published at www.nrcresearchpress.com/cgj on 12 November 2014.
948 Can. Geotech. J. Vol. 52, 2015

water pressure (Kjellman 1952; Bergado et al. 2002; Bo et al. 2003). Fig. 1. Concept to evaluate compression behaviour of soils
Indraratna and Redana (1998) and Chai and Miura (1999) showed (modified from Rujikiatkamjorn et al. 2013).
that drain installation causes a distinctly smeared zone where the
structure of the clay layer is altered due to mandrel driving, af-
fecting its lateral permeability and compressibility. Indraratna
and Redana (1998) demonstrated experimentally that the size of
the smear zone is a parabolic function of the diminished horizon-
tal coefficient of permeability to the vertical permeability ratio.
The smear effect and disturbance of soils due to driven PVDs have
also been investigated by other researchers (Onoue et al. 1991;
Indraratna and Redana 1998; Hird and Moseley 2000; Sharma and
Xiao 2000; Sathananthan and Indraratna 2006; Basu and Prezzi
2007). It is noteworthy that all studies mentioned above were
carried out using reconstituted soil samples where the effect
of soil structure could not be properly captured. Recently,
Rujikiatkamjorn et al. (2013) proposed a conceptual model that
captured the effects of soil structure via the performance of un-
disturbed large-scale testing, and suggested a three-zone model
around the vertical drain considering the variations of soil com-
pressibility and permeability.
The theories on vertical drains have been well developed to
study the behaviour of the vertical drain system. Barron (1948)
introduced the analytical solution for radial consolidation of soil behaviour of a structured (in situ) soil would be different from the
without smear effects. Subsequently, Hansbo (1981) incorporated same soil in a remoulded condition: (i) existing structure provides
the smear effect and well resistance into Barron’s formulation. As a material that is initially stiff at relatively low stress levels,
Hansbo’s theory is applicable to a small strain condition, the co- (ii) structured soil possesses a higher void ratio than the reconsti-
efficient of volume compressibility and associated soil permeabil- tuted soil, and (iii) during yielding, a structured soil generally has
ity, determined at the average effective stress, is assumed to be a higher compressibility than a reconstituted soil, but its com-
unchanged between the beginning and end of preloading. For a pressibility decreases and becomes that of reconstituted soil dur-
broader range of stress, the laboratory consolidation test has ing the destructuring process.
shown that the soil compressibility and permeability relation- A material idealization of the vertical compression of struc-
ships cannot be assumed to be constant on a linear scale and the tured and reconstituted soils during virgin compression is shown
stress state of the soil in relation to in situ stress and preconsoli- in Fig. 1 with the variation of the void ratio due to destructuration
dation pressure as well as the applied preloading pressure are also during consolidation (Liu and Carter 1999, 2000; Rujikiatkamjorn
vital parameters to accurately determine the time-dependent set- et al. 2013). The complete mathematical expressions for the recon-
tlement (Burland 1990; Indraratna and Balasubramaniam 1993). In stituted and structured soil curves in Fig. 1 are explained below.
the recent past, researchers and practitioners have pointed out
Salient characteristics of the conceptual model
field difficulties associated with PVD applications for in situ soil
(Chu et al. 2000; Bo et al. 2003). Indraratna et al. (2005) introduced 1. Under one-dimensional conditions, the compression line of
a preliminary analytical and numerical model capturing the vac- reconstituted soil can be expressed as
uum consolidation mechanisms, but the soil structure effect was
not captured in their proposed model. In any case, none of these (1) e∗ ⫽ eic

⫺ cc∗ log␴v
past approaches could correctly capture the role of soil structure
in relation to its permeability and compressibility in view of PVD
installation. where e* is the void ratio of reconstituted soil during vertical

In this paper, the analytical solution for radial consolidation compression, eic is the void ratio of reconstituted soil when
incorporating the effect of soil disturbance on soil compressibility ␴v = 1 kPa during virgin vertical compression, cc∗ is the gradient
and soil permeability is proposed. The effects of in situ soil struc- of the compression line of the reconstituted soil, and ␴v is the
ture and the magnitude of preloading are examined through the current vertical effective stress.
dissipation of excess pore pressure and associated settlement. 2. The compression curve of the undisturbed soil with the void
Subsequently, the predictions of settlements and associated ex- ratio (ei,u) at vertical effective stress at the initial yield point
cess pore pressures based on the authors’ model are presented in 共␴y i,u兲 can be determined using one-dimensional consolidation
comparison with field observations in the town of Ballina, New test given by

冉 冊
South Wales, Australia.
␴y i,u b

Compression behaviour capturing changes in soil (2) e ⫽ eic ⫹ ⌬ei,u ⫺ cc∗ log␴v for ␴v ≥ ␴y i,u
structure ␴v

The structure of in situ soils is different from the same material ␴v
in a reconstituted state, resulting in different behaviour (e.g., (3) e ⫽ ei,u ⫺ cs∗ log for ␴v ⬍ ␴y i,u
Burland 1990; Leroueil and Vaughan 1990; Cuccovillo and Coop ␴y i,u
1999). There have been studies in developing constitutive models

that consider the influence of soil structure, such as those pro- (4) ⌬ei,u ⫽ ei,u ⫺ eic ⫹ cc∗ log␴y i,u
posed by Gens and Nova (1993), Whittle (1993), Rouainia and Muir
Wood (2000), Liu and Carter (2002), and Wheeler et al. (2003). The
structure of the soil can normally be assessed using the variation In the above, b is a parameter representing the rate of destruc-
in void ratio with vertical effective stress obtained from oedom- turing and cs∗ is the gradient of the compression line when
eter compression tests. Liu and Carter (1999) proposed how the ␴v ⬍ ␴y i,u.

Published by NRC Research Press


Rujikiatkamjorn and Indraratna 949

3. For compression curves of a partially disturbed soil, it is as- Fig. 2. (a) Unit cell with vertical drain; (b) permeability variation;
sumed that the yield stress is located on the extrapolated line (c) initial void ratio variation, and (d) final void ratio variation.
LM (Fig. 1), which is perpendicular to the compression line of
the reconstituted soil and through the vertical effective stress
at the initial yield point 共␴y i,u兲 of the in situ void ratio of the
undisturbed soil (ei,u) (Nagaraj et al. 1994). The initial yield
stress of the partially disturbed soil (␴y i,D), and its respective
void ratio (ei,D) are expected to be located on this line. Then the
initial yield stress of the partially disturbed soil (␴y i,D) on line
LM can now be readily calculated using the following equation

␴y i,D ⫽ 兵10[cc (ei,D⫺ei,u)]其␴y i,u



(5)

4. The compression curves of a partially disturbed soil with void


ratio (ei,D) at a vertical effective stress at the initial yield point
(␴y i,D) is given by

冋 册
␴y i,D b

(6) e ⫽ eic ⫹ ⌬ei,D ⫺ cc∗ log␴v for ␴v ≥ ␴y i,D
␴v

(7) e ⫽ ei,D ⫺ cs∗ log 冋 册 ␴v


␴y i,D
for ␴v ⬍ ␴y i,D


(8) ⌬ei,D ⫽ ei,D ⫺ eic ⫹ cc∗ log␴y i,D

Analytical solutions for radial consolidation


considering soil-structure characteristics
Laboratory testing has shown that during the consolidation pro-
cess, the variation of soil volume compressibility and soil perme-
ability can be quantified, both of which impart a direct influence
on the shape of the void ratio–effective stress (e–log␴ ) and void
ratio–permeability (e–logkh) relationships (Lekha et al. 2003). In
the field, the nature of subsoil stress history gives different
consolidation responses (Seah and Juirnarongit 2003). There-
fore, to predict the time-dependent consolidation subjected to
radial drainage more accurately, it is necessary to incorporate the
relationships of e–log␴  and e–logkh as well as the change in soil
structure characteristics due to the installation of PVDs, and then
determine a new solution for the radial consolidation. It is as-
sumed that both vertical consolidation and well resistance are
negligible due to the long drain length and high discharge capac- The set of equations satisfying the above conditions are
ity of the PVDs.
Based on Walker and Indraratna (2007), the flow rate in the unit
cell by Darcy’s law with respect to the rate of volume change of
(12a) ks ⫽ 冉
k0,u A
␬ rw
r⫹B 冊
soil mass can be expressed as (Fig. 2)

where
共 兲
kh(r) ⭸u ⭸␧ 2
(9)
␥w ⭸r
2r dz ⫽ 共r ⫺ r 2兲 dz
⭸t e
␬⫺1
(12b) A⫽
s⫺1
where kh(r) is the permeability at radial distance r for an element
thickness of dz, ␥w is the unit weight of water, u is excess porewa- s⫺␬
(12c) B⫽
ter pressure at radial distance r, ␧ is the volumetric strain, re is the s⫺1
influence zone radius, and t is time.
k0,u
As shown in Fig. 2b, the linear distributions of permeability in (12d) ␬⫽
both smear and undisturbed zones proposed by Walker and k0(rw)
Indraratna (2007) can be expressed as
rs
(12e) s⫽
(10) ks(rw) ⫽ k0(rw) rw

and Combining eq. (9) with eqs. (10)–(12a), the pore-water pressure in
the smear and undisturbed zones, respectively, can be deter-
(11) ks(rs) ⫽ k0,u mined from

Published by NRC Research Press


950 Can. Geotech. J. Vol. 52, 2015

(13) us ⫽
␥wre2 ⭸␧ 1
2kh ⭸t

␬ ln
r
B rw
B
冉 冊 冉
1
⫹ 2 2 ⫺ ln B ⫹
An B
Ar
rw 冊冉 冊
Fig. 3. (a) Compression during preloading and (b) semi-log
permeability – void ratio relationship.


1⫺

An2
r
rw
; rw ⬍ r ⬍ rs 册
and

(14) u⫽
␥wre2 ⭸␧
2kh ⭸r 再冉 冊
␬ ln
r
s rw
1 r2
⫹ 2 s2 ⫺ 2
2n rw
冉 冊
⫹␬ 冋B1 ln(s) ⫺ sAn⫺ 1 ⫹ 冉AnB
2 2

1
B 冊 册冎;
ln(␬) rs ⬍ r ⬍ re

re
where n ⫽ .
rw
Let ūt be the average excess pore pressure of the smear and
undisturbed zones, at depth z, for vertical drain length l and for a
given time t; hence,

冤冕冕 冕冕 冥
l re l rs
1
(15) ūt ⫽ u2␲r drdz ⫹ us2␲r drdz
␲共re2 ⫺ rw
2
兲l 0 rs 0 rw

Substituting eqs. (13) and (14) into eq. (15) yields the following
expression for the average excess pore pressure, ūt, at any time,
knowing that de = 2re and ⭸␧ ⫽ ⭸ē/共1 ⫹ e0兲; thus,

⭸ē de2␮
(16) ūt ⫽ ␥
⭸t 8kh(1 ⫹ e0) w

Ignoring higher-order terms, the ␮ parameter from eq. (16) is


given by

(17) ␮ ⫽ ln 共 ns兲 ⫺ 43 ⫹ ␬s(s⫺⫺␬1) ln共 ␬s 兲 (20) ei(rw) ⫽ ei,u ⫹ ck log 冋 册 ki(rw)


ki

where rs is the smear zone radius, rw is the equivalent radius of (21) ei(rs) ⫽ ei,u
drain, and ks(r) is permeability in the smear zone at radial distance r.
The average void ratio – effective stress according to void ratio
The assumed linear distribution (Figs. 2c and 2d) that satisfies
distribution along a radius (Figs. 2c and 2d), and the void ratio–
the above conditions is
permeability relations along line LM shown in Figs. 1 and 3 can

冋 册
now be expressed by ei,u C
e(r) ⫽ r⫹D

冉冊
(22a)
f i rw
៮

(18a) ē ⫽ ē0 ⫺ c̄s log
v
; ៮ ⬍ ␴
␴ ៮ 
v yi fi ⫺ 1
␴v0 (22b) C⫽
s⫺1

(18b) ē ⫽ ē0 ⫺ c̄s log 冉冊៮


␴ 
yi

␴v0
⫺ c̄c log 冉冊
៮



v


yi
; ៮ ≥ ␴
␴v
៮ 
yi
(22c) D⫽
s ⫺ fi
s⫺1
ei,u

冉 冊
kh (22d) fi ⫽
ei(rw)
(19) ē ⫽ ē0 ⫹ ck log
kh0
rs
(22e) s⫽
rw
where ē is the average void ratio at a given depth, c̄c is the average
compressibility index for a given depth, ck is the permeability
index, and ē0 is the average initial void ratio. where ei,u is the in situ void ratio at yield stress of the undisturbed
The initial average void ratio at yield stress according to Fig. 2c soil, rs is the smear zone radius, and rw is equivalent radius of the
can be determined as drain.

Published by NRC Research Press


Rujikiatkamjorn and Indraratna 951

By integrating eq. (22a), the initial average void ratio can be ៮ ⫽ ␴


where ti is the time when ␴ ៮ 
v vi and ūt is the average excess pore
expressed as pressure at time t. The average excess pore pressure ratio is the
ratio between average excess pore pressure from the entire unit
(23) ēi(re ⫺ rw) ⫽ 冕冉 冋
rs

rw
ei,u C
f i rw 册冊
r ⫹ D dr ⫹ 冕rs
re
ei,u dr
cell and the applied preloading pressure. This is a useful expres-
sion to be used to calculate the pore pressure based degree of
consolidation (Indraratna et al. 2005).
⭸␨
By differentiation, the term can then be derived as (see
The exact solution of eq. (23) is given by (detailed derivations ⭸Th∗
can be found in Appendix A) Appendix B)

再冋 册 冎
⭸␨ 8
ei,u 1 (f i ⫹ 1)(s ⫺ 1) (29) ⫽⫺ ␨
(24) ēi ⫽ ⫹ (n ⫺ s) ⭸Th∗ ␮
(n ⫺ 1) f i 2

៮  where Th∗ is the modified time factor, defined by


Based on ēi, the average preconsolidation pressure (␴ vi) can be

冋冉 冊 册
calculated using eq. (5). 1⫺(c̄s/ck)

␴ 
In similar manner to eq. (24), the average initial void ratio (ē0) at ∗ yi
(30a) Th0 ⫽ Pav,0Th0 ⫽ 0.5 ⫹ 1 Th0 ;
stress (␴v0) and the average final void ratio (ēf) at final effective ␴v0
stress (␴v0 ⫹ ⌬␴ ), where ⌬␴  is the applied preloading pressure,
៮ ⬍ ␴
␴ ៮ 
are given by v yi and t ⬍ ti

再冋 册 冎 冋冉 冊 册
1⫺(c̄s/ck)
e0,u 1 (f0 ⫹ 1)(s ⫺ 1) ៮
␴ 
(25a) ē0 ⫽ ⫹ (n ⫺ s) (30b) Pav,0 ⫽ 0.5
yi
⫹1
(n ⫺ 1) f0 2
␴v0

冋冉 ␴v0
冊 册
1⫺(c̄c/ck)
and ∗ ⌬␴ 
(30c) Thi ⫽ Pav,iThi ⫽ 0.5 ⫹ ⫹ 1 Thi ;

␴  ៮
␴ 
re yi yi
(25b) n⫽ ៮ ≥ ␴
៮ 
rw ␴v yi and t ≥ ti

再冋
1 (f f ⫹ 1)(s ⫺ 1)
册 冎 冋冉 冊 册
ef,u ␴v0 1⫺(c̄c/ck)
(26) ēf ⫽ ⫹ (n ⫺ s) ⌬␴ 
(n ⫺ 1) f f 2 (30d) Pav,i ⫽ 0.5 ⫹ ⫹1

␴  ៮
␴ 
yi yi

where
Integrating eq. (30a) subject to the boundary condition that
ef,u e0,u ūt ⫽ ⌬␴ at Th0 = 0 and then integrating eq. (30c) subject to the
ff ⫽ and f0 ⫽ ៮
ef(rw) e0(rw) boundary condition that ūt ⫽ ␴v0 ⫹ ⌬␴  ⫺ ␴ 
vi at Thi = 0 and t ⫽ ti
leads to the following expressions:
Based on the final stress state (␴v0 ⫹ ⌬␴ ), the final void ratio in
the undisturbed zone (ef,u) and the final void ratio at the drain
interface (ef(rw)) can be determined using eqs. (2) to (4) and (6) to
(31a) ␨ ⫽ exp ⫺ 冉 ␮

8Th0
冊 ៮ ⬍ ␴
for ␴v
៮ 
yi and t ⬍ ti

冉 冊
(8), respectively.
When ēi and ēf are determined using eqs. (24) and (26), respec-
tively, the average compressibility index c̄c can be determined (31b) ␨⫽
⌬␴

␴v0 ⫹ ⌬␴  ⫺ ␴ 
yi

exp ⫺ 冉 冊 ␮

8Thi

from
៮ ≥ ␴
for ␴ ៮ 
v yi and t ≥ ti
ēi ⫺ ēf

冉 冊
(27) c̄c ⫽
␴v0 ⫹ ⌬␴  Substituting Th∗ in eqs. (31a) and (31b), the expression for excess
log  pore pressure ratio becomes
␴yi

再 冋冉 冊 册冎
1⫺(c̄s/ck)

␴ 
Th0
yi
Assuming that c̄s ⫽ cs, and substituting excess pore pressure (32a) ␨ ⫽ exp ⫺4 ⫹1
ratio ␨ ⫽ ūt/⌬␴  in eq. (16), subsequent algebraic manipulation ␴v0 ␮
gives ៮ ⬍ ␴
for ␴ ៮ 
v yi and t ⬍ ti

(28a) ␨⫽
⭸ē ⭸(␴ ⫺ ūt) kh0
⭸ ␴v ⭸t
de2␮

kh 8k (1 ⫹ ē )⌬␴  w
h0 0
(32b) ␨⫽ 冉 
␴v0
⌬␴ 

⫹ ⌬␴  ⫺ ␴ 
yi
冊 冉 冋冉
exp ⫺4



␴v0


⌬␴ 

␴  冊
1⫺(c̄c/ck)
⫹1
册冊
Thi

for ␴៮ ⬍ ␴
៮  yi yi
v yi and t ⬍ ti ៮ ≥ ␴
៮ 
for ␴v yi and t ≥ ti

⭸ē ⭸(␴ ⫺ ūt) khi de2␮


(28b) ␨⫽ ␥
⭸ ␴v ⭸t kh 8k (1 ⫹ ē )⌬␴  w When the effective stress equals the preconsolidation pressure
hi i ៮ ⫽ ␴
共␴ ៮ 
៮ ≥ ␴
៮  v yi兲, the corresponding time factor ti can then be determined
for ␴v yi and t ≥ ti using eqs. (32a) and (32b), thus,

Published by NRC Research Press


952 Can. Geotech. J. Vol. 52, 2015

de2␮
冉 冊
Table 1. Soil parameters used in the parametric analysis for normally
⌬␴ 

冋冉 冊 册
(33) ti ⫽ 1⫺(c̄s/ck)
ln consolidated clay.

␴  ៮
␴v0 ⫹ ⌬␴  ⫺ ␴ 
yi yi Walker and
4cho ⫹1
␴v0 Proposed model Indraratna (2007)
Soil
If the embankment construction from surcharge fill is assumed parameter Case A Case B Case C Case A Case B Case C
to be a ramp loading (i.e., the embankment load (⌬␴ ) increases ē0 1.950 1.950 1.950 2.112 2.112 2.112
linearly with time from ␴v0 up to a maximum value (␴v0 ⫹ ēf 1.734 1.630 1.553 1.872 1.757 1.673
⌬␴ ) at time t0 and keeps constant thereafter), then the excess ៮
␴v0

(kPa) 28.0 28.0 28.0 28.0 28.0 28.0
pore pressure ratio for ramp loading (␨L) can be expressed by ៮ (kPa)
␴ f
53.2 73.7 94.3 53.2 73.7 94.3
c̄c 0.78 0.76 0.75 0.90 0.87 0.85
⌬␴t ␨ ck 0.787 0.787 0.787 0.787 0.787 0.787
(34a) ␨L ⫽ ; 0 ⬍ t ⬍ t0
⌬␴  f0 1.38 1.38 1.38 N/A N/A N/A
ff 1.36 1.35 1.35 N/A N/A N/A
(34b) ␨L ⫽ ␨; t ⬎ t0 k0,u × 10−10 (m/s) 6.88 6.88 6.88 6.88 6.88 6.88
k0(rw) × 10−10 (m/s) 2.96 2.96 2.96 2.96 2.96 2.96
kh0 × 10−10 (m/s) 5.21 5.21 5.21 8.62 8.62 8.62
where ⌬␴t is the embankment load during a ramp loading at a Note: N/A, not applicable.
given time t.
When the value of c̄c/ck for a normally consolidated soil ap- Table 2. Soil parameters used in the parametric analysis for lightly
proaches unity, the authors’ solution converges to that of Walker overconsolidated clay.
and Indraratna (2007); hence,
Walker and

冉 冊
Proposed model Indraratna (2007)
8Th
(35) ␨ ⫽ exp ⫺ Soil

parameter Case A Case B Case C Case A Case B Case C

The average degree of consolidation based on the excess pore ē0 2.018 1.950 1.950 2.157 2.157 2.157
pressure can be obtained as follows: ēi 1.981 1.981 1.981 2.112 2.112 2.112
ēf 1.734 1.630 1.553 1.872 1.757 1.673

␴  13.96 13.96 13.96 13.96 13.96 13.96
(36) Up ⫽ 1 ⫺ ␨ v0 (kPa)

␴  24.7 24.7 24.7 28.0 28.0 28.0
yi (kPa)
៮
␴ (kPa) 53.2 73.7 94.3 53.2 73.7 94.3
The average degree of consolidation (Us) based on settlement f
(vertical strain) can then be given by the ratio ␳/␳∞, where ␳ and ␳∞ c̄s 0.149 0.149 0.149 0.149 0.149 0.149
are the settlement and ultimate settlement, respectively. c̄c 0.74 0.74 0.73 0.86 0.84 0.83
The associated settlement (␳) can then be evaluated by the fol- ck 0.787 0.787 0.787 0.787 0.787 0.787
lowing equations: f0 1.30 1.30 1.30 N/A N/A N/A
fi 1.29 1.29 1.29 N/A N/A N/A

冉冊
ff 1.36 1.35 1.35 N/A N/A N/A
Hc̄s ៮
␴ k0,u × 10−10 (m/s) 6.88 6.88 6.88 6.88 6.88 6.88
(37a) ␳⫽ log 
v ៮ ⬍ ␴
for ␴ ៮ 
(1 ⫹ ē0) v yi k0(rw) × 10−10 (m/s) 2.96 2.96 2.96 2.96 2.96 2.96
␴v0 kh0 × 10−10 (m/s) 6.36 6.36 6.36 9.55 9.55 9.55

冋 冉冊 冉 冊册
khi × 10−10 (m/s) 5.71 5.71 5.71 8.37 8.37 8.37

␴  ៮

(37b) ␳⫽
H
c̄s log 
yi
⫹ c̄c log
v ៮ ≥ ␴
for ␴ ៮ 
(1 ⫹ ē0) ␴v0 ៮ 
v yi
␴yi proposed model. Cases A, B, and C have the same initial effective
stress of 28 kPa. To study the effect of the preloading pressure
ratio, the final effective stresses after consolidation for cases A, B,
where ␳ is settlement at a given time, c̄c is the average compres- and C were 53.2, 73.7, and 94.3 kPa, respectively, as shown by
sion index, c̄s is the average recompression index, and H is the points A, B, and C in Fig. 4b. The associated preloading pressure
compressible soil thickness. ratios (⌬␴ /␴v0) are 0.9, 1.6, and 2.4 for cases A, B, and C, respec-
៮    tively. The axial strain made using Walker and Indraratna’s (2007)
Effects of preconsolidation pressure (␴ yi) and ⌬␴ /␴v0 ratio solution are greater than those from the current model (Fig. 4b),
Before using the new model to simulate the radial consolida- and this is because the average compressibility influenced by
tion behaviour of soft soil via a case study, a sensitivity analysis drain installation is considered in the latter. There is no doubt
was carried out to examine the influence of the parameters in- that the increase in applied preconsolidation pressure has re-
cluding preconsolidation pressure 共␴ ៮
yi兲 and the preloading pres- sulted in an increase of the final settlement. Figure 4c presents
sure ratio, which is the ratio between the applied surcharge and the predicted average excess pore pressure ratio. As expected,
the initial in situ stress (⌬␴ /␴v0) in accordance with the soil prop- excess pore pressure dissipation rates calculated by Walker and
erties tabulated in Tables 1 and 2. For normally consolidated soil, Indraratna (2007) are higher than those of the proposed model,
Fig. 4 illustrates the predicted values using the authors’ model which is attributed to the change in soil permeability during con-
(using eqs. (30a–33) and (35–37b)), in comparison with Walker and solidation.
Indraratna’s (2007) solution at different preloading pressure ra- Similar to cases A, B, and C, a further three cases (D, E, and F)
tios (⌬p/␴v0r). For Walker and Indraratna’s (2007) prediction, the were examined where the initial stress was reduced to 13.96 kPa
compressibility curve of undisturbed soil was employed, whereas for lightly overconsolidated soil. The preloading pressure ratios
the average compressibility curve capturing soil disturbance due (⌬␴ /␴v0) are 2.8, 4.3, and 5.8 for cases D, E, and F, respectively.
to PVD installation calculated by eqs. (25a)–(27) was adopted in the Figure 5a presents the compression curves of an undisturbed sam-

Published by NRC Research Press


Rujikiatkamjorn and Indraratna 953

Fig. 4. Comparison between predicted results of the proposed model and Walker and Indraratna’s (2007) solution at the embankment
centreline: (a) relationship between specific volume and effective stress, (b) total vertical strain with time, and (c) average excess pore pressure
with time at 4.5 m depth.

ple and a sample obtained at the PVD vicinity, and the average from the elastic to elastoplastic range. When the effective stress
compression curve based on soil parameters given in Table 2 us- exceeds the yield point, the excess pore pressure dissipation in
ing eqs. (20)–(27). Points D, E, and F show the location of the final case D (the lowest preloading ratio) is faster than others within
effective stress after consolidation. Figure 5b shows the compari- 200 days. Afterwards, it becomes slower than in cases E and F.
son of the time-dependent axial strain curves, as modelled using Therefore, to correctly evaluate the vertical drain performance,
eqs. (35)–(37), whereas Fig. 5c illustrates the average excess pore the permeability, compressibility, stress history, preconsolida-
pressure ratio with time using eqs. (30a)–(33). It is evident that the tion pressure, and magnitude of preloading pressure should all be
increase in ⌬␴ /␴v0 increases the ultimate settlement (Fig. 5b). It is considered and integrated rationally in the mathematical formu-
of interest to note that the soil compressibility and yield stress lation. More significantly, the role of the ⌬␴ /␴v0 ratio is found to
decrease when the effect of PVD installation is considered. The be important as demonstrated here.
former reduces overall settlement whereas the latter increases
the overall settlement. As a result, both models provide similar Application of the model to a selected case study
ultimate axial strain values. The dissipation of excess pore pres-
Site descriptions
sure in the predominantly elastic range (effective stress less than
៮   
The Pacific Highway was constructed to cater to the traffic de-
␴yi) is almost the same for all ⌬␴ /␴v0 values (i.e., initial part of the mand between Sydney and Brisbane. The bypass route via the
curve, Fig. 5c). Attributed to the slope change from c̄s to c̄c at ␴៮ 
yi, a town of Ballina was built on some highly compressible and satu-
“kink” in the excess pore pressure curve is observed in Fig. 5c. The rated estuarine and alluvial clay (thickness up to 30 m in some
exact location of this bifurcation depends on the value of the locations). The application of surcharge loading in conjunction
⌬␴ /␴v0 ratio. For instance, in case D the slope change is clearly with PVDs was adopted to reduce consolidation time and con-
visible in Fig. 5c. However, in reality this “kink” is not expected to struction cost. For the purpose of performance monitoring, a trial
be pronounced because the soil compressibility changes gradually embankment was built at the southern approach to Emigrant

Published by NRC Research Press


954 Can. Geotech. J. Vol. 52, 2015

Fig. 5. Settlements and excess pore pressure responses at the embankment centreline: (a) relationship between void ratio and effective stress,
(b) total vertical strain with time at the surface, and (c) average excess pore pressure ratio with time at 4.5 m depth.

Fig. 6. Instrumentation layout for the test embankments at Ballina Bypass (modified after Indraratna et al. 2012).

Published by NRC Research Press


Rujikiatkamjorn and Indraratna 955

Fig. 7. Plan view of sampling locations to characterize smear zone Fig. 8. Variations of (a) void ratio; (b) permeability, and
(modified after Indraratna et al. 2014). (c) normalised permeability along the distance away from vertical
drain (modified after Indraratna et al. 2014).

Creek, north of Ballina. At this site, there is a relatively uniform


soft silty clay layer (6–30 m) followed by firm silty clay to a depth
of 55 m. The groundwater level is almost at the ground surface.
The water content of the soft and medium silty clay layers varied
from 80% to 120% and was characterized by very low undrained
shear strength. The field vane shear tests indicated that the shear
strength was in the range of 5–40 kPa.
Indraratna et al. (2012) reported that PVDs were installed in a
square pattern at a spacing of 1.0 m. Figure 6 shows the locations
of field instrumentation, which consisted of surface settlement
plates, inclinometers, and piezometers. The piezometers were
placed at 1.3, 4.5, and 8 m below the ground level away from the
PVDs by 0.5 m. A total of eight inclinometers were installed at the within 1 h from depths between 2.5 and 2.95 m below the 600 mm
boundaries of the embankment. The embankment was split into thick working platform using Shelby tube samplers (50 mm diam-
two sections: section A without vacuum pressure and section B eter and 450 mm long). Ten samples were collected and their
with vacuum pressure. Within the scope of this study, only the locations are presented in Fig. 7. The tubes were then wrapped in
field performance of the nonvacuum area is discussed and ana- a shock-absorbing bubble wrap sheet and transported and stored
lyzed, where the surface settlement at settlement plate SP1 and in a room under low temperature (10 °C) and with 95% humidity,
the excess pore pressure measured by piezometer P1 at 4.5 m to prevent moisture loss before being tested in the laboratory.
depth at 0.5 m from each drain are used for comparison. The To obtain the horizontal permeability, specimens were ex-
time–settlement curves at SP1 and SP2 are very similar. Therefore tracted in the horizontal direction from the Shelby tube samples.
only the curve at SP1 is plotted in this paper. The time–settlement The Casagrande log-time method (graphical method) was em-
curve at SP2 is similar to that at SP1. The wick drains were 100 mm ployed to determine the horizontal coefficient of consolidation
wide, 3 mm thick, and the cross section of the mandrel was (ch) using time at a 50% degree of consolidation from samples
120 mm × 60 mm. Drains were installed in a square pattern 1 m extracted horizontally, and Terzaghi’s one-dimensional theory
apart, to a depth of 15 m, at a penetration rate of approximately was used to calculate the corresponding horizontal permeability
1.5 m/s. To characterize the smear zone, soil samples were col- (kh) in relation to the coefficient of compressibility (mv) and the
lected between the drains as reported by Indraratna et al. (2014). horizontal coefficient of consolidation (ch). As shown in Fig. 8, the
After installation, samples of undisturbed soil were recovered lateral permeability was almost constant beyond 400 mm away

Published by NRC Research Press


956 Can. Geotech. J. Vol. 52, 2015

Fig. 9. (a) Compression curves and (b) horizontal permeability variation with void ratio.

from the drain, but decreased towards the drain. To characterize The following three cases were examined:
the smear zone, normalized permeability defined as the ratio be-
tween the horizontal permeability (kh) and the horizontal perme- Case G: Authors’ proposed model capturing the smear zone
ability of the undisturbed zone (kh(undisturbed)) is plotted in Fig. 8c. characteristics including both compressibility and permeability
This shows that the value of the normalized permeability ratio parameters;
decreased swiftly with the radial distance close to the drain Case H: Walker and Indraratna’s (2007) model capturing the
boundary (highly disturbed zone), whereas the change in normal- linear variation of permeability in the smear zone, but ignoring
ized permeability became minimal further away from the drain. the change in soil compressibility caused by PVD installation; and
Irrespective of the applied pressure, all the curves are confined Case I: Idealized analysis where the effects of soil disturbance
within a relatively narrow band, clearly defining the extent of the were ignored.
smear zone. This data reveals that the normalized lateral perme-
Figures 10b and 10c present the variation of the predicted and
ability ratio within the smear zone varies from 0.2 to 1 (an average
measured excess pore pressures and the associated settlements,
of 0.6), and further away from the drain the kh/kh(undisturbed) ratio
respectively. The proposed model (case G) can capture the rate and
approaches unity. In general, the variations in the permeability
magnitude of consolidation better than cases H and I. Case I
indicate that the smear zone was about 6.3 times greater than the
yielded a different ultimate settlement and consolidation rate
equivalent dimension of the mandrel, a result that was higher
than cases H and I. As expected, the overall consolidation rate
than those obtained in previous laboratory studies (Indraratna
and Redana 1998; Sathananthan and Indraratna 2006). It is possi- based on case I was the highest followed by cases H and G. This is
ble that increased PVD length may have contributed to a much because the effects of soil disturbance on soil compressibility and
larger smear zone, i.e., at least six times the equivalent mandrel permeability were ignored in the calculations of cases H and I,
width compared to past studies with shorter PVDs, but further respectively. The measurement of the excess pore pressure was
studies are needed to establish a more convincing relationship plotted for 300 days due to the malfunctioning of the piezometer
between the PVD length and the extent of smear. Figures 9a and P1 at a depth of 4.5 m and at a distance of 0.5 m from each drain
9b present the compression curves for vertical samples and the (Fig. 6). It can be seen that the excess pore-water pressure predic-
horizontal permeability variation with void ratio, respectively. tion for case G (i.e., proposed model) underpredicts the field data,
They show that the soil becomes increasingly disturbed towards but the predicted results plot closer to reality than cases H and I.
the drain as the adjacent soil experiences severe remoulding due This can be attributed to the piezometric measurement at a par-
to drain installation. From e–log␴  and e–logkh plots, the slope of ticular point (location) in contrast to the authors’ solution, which
the e–log␴  line (Cc) of an undisturbed sample and the slope of the is an averaged value. This may also be attributed to partial clog-
e–logkh line (Ck) are found to be 1.06 and 0.79, respectively. ging (silt intrusion) of the PVDs after 250 days. In fact, random
exhuming of a few PVDs indicated partial clogging of the filter
Comparison between predicted and field data due to silt intrusion.
Settlement and associated pore pressure recorded by the settle-
ment plate and piezometer are shown in Fig. 10 with an embank- Conclusion
ment construction schedule. Soil parameters used in this section An analytical model for radial consolidation of soft clay incor-
are listed in Table 3. The embankments were initially raised to a porating the effects of soil disturbance has been proposed. The
height of 4.5 m in 120 days. The ramp embankment load was effects of soil disturbance during vertical drain installation on the
applied according to a single-stage construction. The unit weight permeability and compressibility were studied using samples ob-
of fill was assumed to be 17.5 kN/m3. Settlement and associated tained from a soft clay site in Ballina, NSW. From the results of
excess pore pressure predictions were carried out at the embank- this study, the following conclusions can be drawn:
ment centreline using the proposed analytical model (e.g., eqs. (30)–
(37)). At the centreline (zero lateral displacements), the analysis 1. The characteristics of the smear zone could be explained on
follows the basic one-dimensional consolidation theory, and the the basis of the variations in soil permeability and compress-
use of an EXCEL spreadsheet formulation was adopted. ibility. The radius of the smear zone was found to be six times

Published by NRC Research Press


Rujikiatkamjorn and Indraratna 957

Fig. 10. (a) Actual loading of the embankment; (b) predicted average Table 3. Soil parameters used in a case study analysis.
excess pore pressure and actual excess pore pressures at 4.5 m depth Case A: Case B: Case C: Ideal drain
and a distance of 0.5 m away from each drain, and (c) predicted and Soil Proposed Walker and with no soil
actual surface settlements in SP1 (measured data from Indraratna parameter model Indraratna (2007) disturbance
et al. 2012).
ē0 3.191 3.270 3.270
ēi 3.110 3.184 3.184
ēf 2.448 2.516 2.516

␴  19.70 19.70 19.70
v0 (kPa)

␴ 
(kPa) 31.20 33.50 33.50
yi
៮ (kPa)
␴ 99.7 99.7 99.7
f
c̄s 0.406 0.373 0.373
c̄c 1.31 1.41 1.41
ck 0.79 0.79 0.79
f0 1.30 N/A N/A
fi 1.29 N/A N/A
ff 1.35 N/A N/A
k0,u × 10−10 (m/s) 10 10 10
k0(rw) × 10−10 (m/s) 4.29 4.29 —
kh0 × 10−10 (m/s) 8.12 10 10
khi × 10−10 (m/s) 6.41 7.78 7.78
␮ 3.269 3.269 2.754
n 33.24 33.24 33.24
s 7.77 7.77 —

sure, stress history, and smear zone characteristics are signif-


icant factors influencing the consolidation of PVD stabilized
soft clays.
3. This study demonstrates that to make more realistic predic-
tions and improve current design practices, the inevitable
changes in soil permeability and compressibility within the
smear zone caused by the mandrel should be properly
captured.

Acknowledgements
The first author thankfully acknowledges the financial support
received from the Early Career Researcher Award through the
ARC Centre of Excellence for Geotechnical Science and Engineer-
ing. The support of Ph.D. student Darshana Perera during manu-
script preparation is very much appreciated.

References
Barron, R.A. 1948. Consolidation of fine-grained soils by drain wells. Transac-
tions ASCE, 113(1): 718–724.
Basu, D., and Prezzi, M. 2007. Effect of the smear and transition zones around
prefabricated vertical drains installed in a triangular pattern on the rate of
soil consolidation. International Journal of Geomechanics, 7: 34–43. doi:10.
1061/(ASCE)1532-3641(2007)7:1(34).
Bergado, D.T., Balasubramaniam, A.S., Fannin, R.J., and Holtz, R.D. 2002. Prefab-
ricated vertical drains (PVDs) in soft Bangkok clay: a case study of the new
Bangkok International Airport project. Canadian Geotechnical Journal,
39(2): 304–315. doi:10.1139/t01-100.
Bishop, A.W. 1966. The strength of soils as engineering materials. Géotechnique,
16: 91–130. doi:10.1680/geot.1966.16.2.91.
the equivalent mandrel radius, which was larger than that Bo, M.W., Chu, J., Low, B.K., and Choa, V. 2003. Soil improvement; prefabricated
observed in the past using reconstituted laboratory speci- vertical drain techniques. Thomson Learning, Singapore.
mens. The compression curves showed that the soil became Burland, J.B. 1990. On the compressibility and shear strength of natural clays.
Géotechnique, 40: 329–378. doi:10.1680/geot.1990.40.3.329.
increasingly disturbed towards the drain as the soil adjacent
Chai, J.-C., and Miura, N. 1999. Investigation of factors affecting vertical drain
to the drain experienced severe remoulding due to the man- behavior. Journal of Geotechnical and Geoenvironmental Engineering,
drel movement. 125(3): 216–226. doi:10.1061/(ASCE)1090-0241(1999)125:3(216).
2. The prediction using the proposed analytical model provided Chu, J., Yan, S.W., and Yang, H. 2000. Soil improvement by the vacuum preload-
a better agreement with the field measurement when applied ing method for an oil storage station. Géotechnique, 50(6): 625–632. doi:10.
1680/geot.2000.50.6.625.
to a selected case study in Ballina. The proposed solution gives Cuccovillo, T., and Coop, M.R. 1999. On the mechanics of structured sands.
better accuracy in prediction of settlement and excess pore Géotechnique, 49: 741–760. doi:10.1680/geot.1999.49.6.741.
pressure compared to the predictions using the Walker and Gens, A., and Nova, R. 1993. Conceptual bases for a constitutive model for
Indraratna (2007) model, which overestimated the settlement bonded soils and weak rocks. In Geotechnical engineering of hard soils–soft
rocks. Edited by A. Anagnostopoulos, F. Schlosser, N. Kalteziotis, and R. Frank.
as the change in compressibility of the smear zone was ig- A.A. Balkema, Rotterdam, the Netherlands. pp. 485–494.
nored. The findings of this study prove that the evaluation of Gibson, R.E., and Lo, K.Y. 1961. A theory of consolidation for soils exhibiting
soil compressibility in relation to the applied preloading pres- secondary compression. NGI.

Published by NRC Research Press


958 Can. Geotech. J. Vol. 52, 2015

Hansbo, S. 1981. Consolidation of fine-grained soils by prefabricated drains. In Yin, Z.-Y., Chang, C.S., Karstunen, M., and Hicher, P.-Y. 2010. An anisotropic
Proceedings of the 10th International Conference on Soil Mechanics and elastic viscoplastic model for soft clays. International Journal of Solids and
Foundation Engineering, Stockholm. pp. 677–682. Structures, 47(5): 665–677. doi:10.1016/j.ijsolstr.2009.11.004.
Hird, C.C., and Moseley, V.J. 2000. Model study of seepage in smear zones around
vertical drains in layered soil. Géotechnique, 50(1): 89–97. doi:10.1680/geot.
2000.50.1.89. List of symbols
Indraratna, B., and Balasubramaniam, A. 1993. Closure to “Performance of test
embankment constructed to failure on soft marine clay” by B. Indraratna, b parameter representing the rate of destructuring
A.S. Balasubramaniam, and S. Balachandran (January, 1992, Vol. 118, No. 1). C constant in eq. (22a)
Journal of Geotechnical Engineering, 119(8): 1326–1329. doi:10.1061/(ASCE) Cc, Ck slope of the e–log␴  and e–logkh lines, respectively
0733-9410(1993)119:8(1326). cc coefficient of consolidation for horizontal drainage
Indraratna, B., and Redana, I.W. 1998. Laboratory determination of smear zone cc∗ gradient of compression line of reconstituted soil
due to vertical drain installation. Journal of Geotechnical Engineering,
c̄c average compression index
124(2): 180–184. doi:10.1061/(ASCE)1090-0241(1998)124:2(180).
Indraratna, B., and Redana, I.W. 2000. Numerical modeling of vertical drains
ch horizontal coefficient of consolidation
with smear and well resistance installed in soft clay. Canadian Geotechnical ch0 initial coefficient of radial consolidation (m2/s)
Journal, 37(1): 132–145. doi:10.1139/t99-115. chi coefficient of radial consolidation at effective preconsolida-
Indraratna, B., Sathananthan, I., Rujikiatkamjorn, C., and Balasubramaniam, A.S. 2005. tion pressure (m2/s)
Analytical and numerical modeling of soft soil stabilized by prefabricated ck permeability index
vertical drains incorporating vacuum preloading. International Journal of cs∗ gradient of compression line in recompression region
Geomechanics, 5(2): 114–124. doi:10.1061/(ASCE)1532-3641(2005)5:2(114). c̄s average recompression index
Indraratna, B., Rujikiatkamjorn, C., Kelly, R., and Buys, H. 2012. Soft soil foun- D constant in eq. (22a)
dation improved by vacuum and surcharge loading. In Proceedings of the
de diameter of influenced zone (m)
Institution of Civil Engineers: Ground Improvement, 165(2): 87–96. doi:10.
1680/grim.10.00032. e void ratio
Indraratna, B., Perera, D., Rujikiatkamjorn, C., and Kelly, R. 2014. Soil distur- e* void ratio of reconstituted soil
bance analysis due to vertical drain installation. Geotechnical Engineering, ē average void ratio at a given depth
ICE. [Posted online ahead of print 7 November 2014.] doi:10.1680/geng.14. e0 initial void ratio
00052. ē0 average initial void ratio
Kjellman, W. 1952. Consolidation of clayey soils by atmospheric pressure. In e0,u initial void ratio of soil in undisturbed zone
Proceedings of a Conference on Soil Stabilization, Massachusetts Institute of e0共rw兲 initial void ratio of soil in disturbed zone next to the drain
Technology, Boston, Mass. pp. 258–263. ēf average final void ratio
Lekha, K.R., Krishnaswamy, N.R., and Basak, P. 2003. Consolidation of clays for
ef,u final void ratio of soil in undisturbed zone
variable permeability and compressibility. Journal of Geotechnical and Geo-
environmental Engineering, 129(11): 1001–1009. doi:10.1061/(ASCE)1090-0241 ef(rw) final void ratio of soil in disturbed zone close to the drain
(2003)129:11(1001). ēi average void ratio at yield stress

Leroueil, S., and Vaughan, P.R. 1990. The general and congruent effects of struc- eic void ratio of reconstituted soil when ␴v = 1 kPa
ture in natural soils and weak rocks. Géotechnique, 40: 467–488. doi:10.1680/ ei,D initial void ratio at yield stress of partially disturbed soil
geot.1990.40.3.467. ei(rw) void ratio at initial yield stress of soil in disturbed zone next
Liu, M.D., and Carter, J.P. 1999. Virgin compression of structured soils. Géotech- to the drain
nique, 49: 43–57. doi:10.1680/geot.1999.49.1.43. ei,u in situ void ratio at yield stress of undisturbed soil
Liu, M.D., and Carter, J.P. 2000. Modelling the destructuring of soils during ēu average void ratio for undisturbed soil
virgin compression. Géotechnique, 50: 479–483. doi:10.1680/geot.2000.50.4.
f0 constant in eq. (25a)
479.
Liu, M.D., and Carter, J.P. 2002. A structured Cam Clay model. Canadian Geotech-
ff constant in eq. (26)
nical Journal, 39(6): 1313–1332. doi:10.1139/t02-069. fi constant in eq. (22a)
Nagaraj, T.S., Pandian, N.S., and Narashima Raju, P.S.R. 1994. Stress-state- H thickness of clay (m)
permeability relations for overconsolidated clays. Géotechnique, 44(2): 349– k0共rw兲 initial horizontal permeability of soil in disturbed zone next
352. doi:10.1680/geot.1994.44.2.349. to the drain
Onoue, A., Ting, N.H., Germaine, J.T., and Whitman, R.V. 1991. Permeability of k0,u initial horizontal permeability of soil in undisturbed zone
disturbed zone around vertical drains. In Proceedings, ASCE Geotechnical (m/s)
Engineering Congress, Colorado. pp. 879–890. kh horizontal permeability
Randolph, M.F., and Wroth, C.P. 1979. An analytical solution for the consolida- kh0 average initial horizontal permeability of soil (m/s)
tion around a driven pile. International Journal for Numerical and Analytical khi average horizontal permeability at yield stress of soil
Methods in Geomechanics, 3(3): 217–229. doi:10.1002/nag.1610030302.
kh(r) horizontal permeability along the radius (m/s)
Rouainia, M., and Muir Wood, D. 2000. A kinematic hardening constitutive
model for natural clays with loss of structure. Géotechnique, 50: 153–164. ki average horizontal permeability at yield stress of soil (m/s)
doi:10.1680/geot.2000.50.2.153. ki(rw) horizontal permeability at yield stress of soil in disturbed
Rujikiatkamjorn, C., and Indraratna, B. 2010. Radial consolidation modelling zone next to the drain (m/s)
incorporating the effect of a smear zone for a multilayer soil with downdrag ks horizontal permeability of soil
caused by mandrel action. Canadian Geotechnical Journal, 47(9): 1024–1035. ks(r) horizontal permeability of soil in disturbed zone (m/s)
doi:10.1139/T09-149. ks(rw) horizontal permeability of soil at drain radius (m/s)
Rujikiatkamjorn, C., Ardana, M.D.W., Indraratna, B., and Leroueil, S. 2013. Con- mv coefficient of volume compressibility (m2/kN)
ceptual model describing smear zone caused by mandrel action. Géotech- mv0 coefficient of volume compressibility at ␴v ⫽ ␴ ៮ (m2/kN)
nique, 63(16): 1377–1388. doi:10.1680/geot.12.P.138. 0
mvi coefficient of volume compressibility at ␴v ⫽ ␴ ៮ 2
Sathananthan, I., and Indraratna, B. 2006. Laboratory evaluation of smear zone yi (m /kN)
and correlation between permeability and moisture content. Journal of r distance from centre of the unit cell (m)
Geotechnical and Geoenvironmental Engineering, 132(7): 942–945. doi:10. re radius of influenced zone (m)
1061/(ASCE)1090-0241(2006)132:7(942). rs radius of smear zone (m)
Seah, T.H., and Juirnarongit, T. 2003. Constant rate of strain consolidation with
rw radius of drain well (m)
radial drainage. Geotechnical Testing Journal, 26(4): 1–12. doi:10.1520/
GTJ11251J. Th dimensionless time factor
Sharma, J.S., and Xiao, D. 2000. Characterization of a smear zone around vertical Th∗ modified dimensionless time factor
drains by large-scale laboratory tests. Canadian Geotechnical Journal, 37(6): Th0 time factor defined by eq. (B1b)
1265–1271. doi:10.1139/t00-050. Thi time factor defined by eq. (B2b)
Walker, R., and Indraratna, B. 2007. Vertical drain consolidation with overlap- t time (s, days)
ping smear zones. Géotechnique, 57(5): 463–467. doi:10.1680/geot.2007.57.5. ti time when ␴ ៮ ⫽ ␴៮
463. v vi (s, days)
Wheeler, S.J., Näätänen, A., Karstunen, M., and Lojander, M. 2003. An anisotro-
Up average degree of consolidation based on excess pore pres-
pic elastoplastic model for soft clays. Canadian Geotechnical Journal, 40(2): sure
403–418. doi:10.1139/t02-119. Us average degree of consolidation based on settlement (vertical
Whittle, A.J. 1993. Evaluation of a constitutive model for overconsolidated clays. strain)
Géotechnique, 43: 289–313. doi:10.1680/geot.1993.43.2.289. u excess pore-water pressure (kPa)

Published by NRC Research Press


Rujikiatkamjorn and Indraratna 959

ūt average excess pore pressure at time t (kPa) chi(t ⫺ ti)


z depth (m) (B2b) Thi ⫽
␥w unit weight of water (kN/m3) de2
␧ volumetric strain
␬ constant defined by eq. (12d) khi(1 ⫹ ēi)
(B2c) chi ⫽ ;
␳ settlement at a given time (m) ␥w⭸ē/⭸ ␴v t⫽ti
␳∞ ultimate settlement (m)
␴ stress
⌬␴  applied preloading pressure (kPa) and
៮
␴ final average vertical effective stress at the end of consolidation
f
␴v current vertical effective stress ⭸ē/⭸ ␴v
៮ average vertical effective stress (kPa) (B2d) mv ⫽
␴v
(1 ⫹ ēi)

␴v0 initial vertical effective stress (kPa)
៮ ⭸ē/⭸ ␴v t⫽ti
␴v0 initial average vertical effective stress (B2e) mvi ⫽
៮  (1 ⫹ ēi)
␴ viaverage preconsolidation pressure

␴ 
yi average yield stress (kPa)
␴y i,D initial yield stress of the partially disturbed soil (kPa) Differentiating eq. (18) with respect to the effective stress (␴v)
␴y i,u
gives
initial yield stress of structured soil (kPa)
␨ average excess pore pressure ratio (kPa)
␨L average excess pore pressure for ramp loading (kPa) mv0 ⌬␴  ␨⌬␴  ៮ ⬍ ␴
៮ 
(B3a) ⫽1⫹  ⫺  for ␴v yi
mv ␴v0 ␴v0
Appendix A. Complete details of eq. (24) derivation
Integrating eq. (23) gives
mvi ␴v0 ⌬␴  ⌬␨␴  ៮ ≥ ␴
៮ 
⫽ ⫹ ⫺ for ␴
冉 冊
(B3b) v yi
ei,u Cr2 rs mv ៮  ៮  ៮ 
(A1) ēi(re ⫺ rw) ⫽ ⫹ Dr ⫹ (ei,ur)rrse ␴ yi ␴yi ␴yi
f i 2rw rw

Combining eqs. (18) and (19) gives


rs re

冉冊
Substituting s ⫽ and n ⫽ gives
⫺c̄s/ck

冉 冊
rw rw ៮
kh ␴v mv0 ⫺c̄s/ck
៮ ⬍ ␴
៮ 

再冋 冎
⫽ ⫽ for ␴

(B4a)
1 (f i ⫹ 1)(s ⫺ 1)
v yi
(A2) ēi ⫽
ei,u
⫹ (n ⫺ s)
kh0 ␴v0 mv
(n ⫺ 1) f i 2

冉冊
⫺c̄c/ck

冉 冊
Appendix B. Complete details of eq. (29) derivation kh ៮
␴ mvi ⫺c̄c/ck
(B4b) ⫽
v
⫽ ៮ ≥ ␴
for ␴ ៮ 

If the preloading pressure (⌬␴ ) is assumed to be an instanta- khi 
␴yi mv v yi

neous loading and total stress (␴) is constant throughout consoli-


dation process (i.e., ⭸␴/⭸t = 0), simplifying eq. (28) gives
Therefore
⭸␨ 8 mv0 kh ៮ ⬍ ␴

冉 冊

(B1a) ⫽⫺ ␨ for ␴
⭸Th0 ␮ mv kh0 v yi kh ⌬␴  ␨⌬␴  ⫺c̄s/ck
៮ ⬍ ␴
៮ 
(B5a) ⫽ 1⫹  ⫺  for ␴v yi
kh0 ␴v0 ␴v0

冉 冊
where
kh ␴v0 ⌬␴  ␨⌬␴ 
⫺c̄c/ck
៮ ≥ ␴
៮ 
(B5b) ⫽ ⫹ ⫺ for ␴v yi
ch0t khi ៮  ៮  ៮ 
(B1b) Th0 ⫽ ␴yi ␴yi ␴yi
de2

kh0(1 ⫹ ē0) Substituting eqs. (B4a), (B4b), (B5a), and (B5b) into eq. (B1) yields
(B1c) ch0 ⫽
␥w⭸ē/⭸ ␴v t⫽0 ⭸␨ 8
(B6a) ⫽ ⫺ P␨
⭸Th ␮
and
where
⭸ē/⭸ ␴v
mv ⫽
冉 冊
(B1d)
(1 ⫹ ē0) ⌬␴  ␨⌬␴  1⫺(c̄s/ck)
៮ ⬍ ␴
៮ 
(B6b) P⫽ 1⫹ ⫺ for ␴v yi
␴v0 ␴v0
⭸ē/⭸ ␴v t⫽0

冉 冊
(B1e) mv0 ⫽
(1 ⫹ ē0) ␴v0 1⫺(c̄c/ck)
⌬␴  ␨⌬␴  ៮ ≥ ␴
៮ 
(B6c) P⫽ ⫹ ⫺ for ␴v yi

␴  ៮
␴  ៮
␴ 
⭸␨ 8 mvi kh ៮
៮ ≥ ␴ 
(B2a) ⫽⫺ ␨ for ␴ yi yi yi
⭸Thi ␮ mv khi v yi

It can be seen that eq. (B6) is a nonlinear partial differential


where equation for radial consolidation under instantaneous loading,

Published by NRC Research Press


960 Can. Geotech. J. Vol. 52, 2015

incorporating the e–log␴v and e–logkh relations. The nonlinear


冋冉 冊 册
1⫺(c̄s/ck)

␴ 
differential eq. (B6a) with variable ␨ does not have a general solu- yi ៮ ⬍ ␴
៮ 

冉冊
(B7a) P ⫽ Pav,0 ⫽ 0.5 ⫹1 for ␴v yi
៮ 
1⫺共c̄s/ck兲 ␴v0
␴ yi ៮
៮ ⬍ ␴ 
tion and P varies from to 1 for ␴ yi, and from

冉 冊 冋冉 冊 册
v
␴v0
␴v0 ⌬␴ 
1⫺共c̄c/ck兲 ␴v0 ⌬␴ 
1⫺(c̄c/ck)
៮ ≥ ␴

⫹ ៮ ≥ ␴
to 1 for ␴ ៮  (B7b) P ⫽ Pav,i ⫽ 0.5 ⫹ ⫹1 for ␴ 
v yi. Hence, it can be assumed to v yi
៮  ៮  ៮
␴  ៮
␴ 
␴ yi ␴ yi yi yi
have an average value given by

Published by NRC Research Press


Copyright of Canadian Geotechnical Journal is the property of Canadian Science Publishing
and its content may not be copied or emailed to multiple sites or posted to a listserv without
the copyright holder's express written permission. However, users may print, download, or
email articles for individual use.

You might also like