You are on page 1of 9

Geomechanics and Geoengineering: An International Journal

Vol. 1, No. 1, March 2006, 63--71

Analytical solutions for consolidation aided by vertical drains


D. BASU, P. BASU and M. PREZZI*
School of Civil Engineering, Purdue University, West Lafayette, IN 47907, USA

(Received 3 November 2005; in final form 12 December 2005)

Soil disturbance caused during the installation of vertical drains reduces the in situ hydraulic conductivity of soft deposits in the immediate vicinity of
the drains, resulting in a slower rate of consolidation than would be expected in the absence of disturbance. Experimental investigations have revealed
the existence of two distinct zones, a smear zone and a transition zone, within the disturbed zone around the vertical drain. The degree of change in the
hydraulic conductivity in the smear and transition zones is difficult to assess without performing of laboratory tests. Based on the available literature,
four different profiles of hydraulic conductivity versus distance from the vertical drain were identified. Closed-form solutions for the rate of
consolidation for each of these four hydraulic conductivity profiles were developed. It is found that different variations of the hydraulic conductivity
profiles in the disturbed zone result in different rates of consolidation.

Keywords: Vertical drain; Analytical solution; Consolidation; Smear; Soil disturbance; Ground improvement

1. Introduction compressibility increases (Holtz and Holm 1973, Lo and Mesri


1994). The reduction in hydraulic conductivity delays the con-
Soft soil deposits are characterized by low shear strength, high solidation process.
compressibility, and low hydraulic conductivity. In order to
improve the strength and stiffness of clayey soils, vertical
drains, combined with preloading, are often used in practice 2. Disturbed zone characterization
(Johnson 1970, Holtz 1987, Bergado et al. 1993a). The installa-
tion of vertical drains reduces the water drainage path, speeding The degree of change in the soil hydraulic conductivity k in the
up the dissipation of the excess pore pressure generated during disturbed zone with distance from the PVD is not known with
preloading. The fact that the flow is predominantly in the certainty. According to most researchers (Barron 1948,
horizontal direction (except near the top surface or close to a Casagrande and Poulos 1969, Hansbo 1981, 1986, 1987,
highly permeable silt/sand seam) further helps the process 1997, Bergado et al. 1991, 1993a), the soil within the disturbed
because, owing to depositional anisotropy, the hydraulic con- zone, which is also known as the smear zone, is completely
ductivity is generally greater in the horizontal direction than in remoulded, resulting in a spatially uniform hydraulic conduc-
the vertical direction. tivity ks which is smaller than the in situ hydraulic conductivity
Since the early 1970s, prefabricated vertical drains (PVD’s) kc (figure 1, case A). The degree of disturbance, quantified as
have been successfully used in various sites throughout the the ratio ks/kc of the hydraulic conductivities of the smear zone
world (Bergado et al. 1993a, 1997, 2002, Lo and Mesri 1994). and the undisturbed zone, and the extent of the smear zone have
PVDs consist of a plastic core surrounded by a filter sleeve with been the subject of many investigations based on back analysis
typical cross-sectional dimensions of 100 mm  4 mm (Holtz of case histories, laboratory tests on samples collected from
1987). They are installed in rectangular or triangular patterns field, model experiments, studies of pile driving, practical con-
with centre--centre distance ranging from about 1.0 m to 3.0 m siderations, and experience. According to Bergado et al.
(Holtz 1987). (1993a,b), Hansbo (1986, 1997), and Hird and Moseley
Closed-ended mandrels are generally used for the installation (2000), the degree of disturbance ks/kc varies between 0.1 and
of PVDs. Soil adjacent to the mandrel is displaced and dragged 0.33. However, Casagrande and Poulos (1969) proposed a
down in order to produce a space for the PVD (Hird and value as low as 0.001, while Bergado et al. (1991) suggested
Moseley 2000). This creates a disturbed zone of soil around that ks/kc ranges from 0.5 to 0.66.
the drain around the PVD in which the strength and the hydrau- The dimensions and shape of the disturbed zone depend on
lic conductivity in the horizontal direction decrease, and the many factors, such as the size and shape of the mandrel, the rate
of mandrel penetration, the type of mandrel shoe, and the soil
*Corresponding author. Email: mprezzi@purdue.edu properties (Hird and Moseley 2000, Holtz et al. 1991). In the

Geomechanics and Geoengineering: An International Journal


ISSN 1748-6025 print=ISSN 1748-6033 online Ó 2006 Taylor & Francis
http:==www.tandf.co.uk=journals
DOI: 10.1080=17486020500527960
64 D. Basu et al.

rc Hansbo 1986, 1997, Bergado et al. 1991, 1993b, Mesri et al.


rtr 1994, Chai et al. 2001) have concluded that the equivalent
Sof t deposit rsm
rd smear zone radius (radius of the smear zone measured from
the centre of the drain) can be taken as approximately two to
four times the equivalent mandrel radius rm,eq.
Transition zone The above discussion is based on the assumption that the
Undisturbed hydraulic conductivity remains constant within the disturbed
zone zone (case A). However, recent experimental investigations
Vertical drain have shown that the assumption of a single value for the
hydraulic conductivity in the disturbed zone is not valid
Smear zone
(Onoue et al. 1991, Madhav et al. 1993, Indraratna and
Redana 1998, Sharma and Xiao 2000). Madhav et al. (1993)
Unit cell performed a field-scale study to investigate the variation of the
hydraulic conductivity profile in the disturbed zone. Soil sam-
Pervious boundary r ples were collected from soft ground in which PVD’s were
Impervious boundary installed and tested in the laboratory to obtain the hydraulic
conductivity profile. The results of Madhav et al. (1993) are
(a) reproduced in figure 2, where the degree of disturbance
k kc (expressed as the ratio k/kc) is plotted as a function of the
1 normalized distance from the drain (normalization is performed
Case A with respect to the equivalent mandrel radius rm,eq). Based on
β
these results, both Madhav et al. (1993) and Miura et al. (1993)
r
k kc suggested that the disturbed zone comprises of two distinct
1 zones: the smear zone and the transition zone. In the completely
Case B remoulded smear zone immediately surrounding the drain, the
β
soil has a constant hydraulic conductivity ks. In the transition
r
k kc zone, which surrounds the smear zone, the degree of distur-
1 bance gradually decreases as the distance from the drain
Case C βt increases. Madhav et al. (1993) further suggested that the
β hydraulic conductivity increases linearly (figure 1, case B)
r from a value equal to ks at the smear zone boundary (i.e. the
k kc boundary between the smear zone and the transition zone) to the
Case D 1 in situ value kc at the transition zone boundary (i.e. the bound-
β ary between the transition zone and the undisturbed zone).
r
k kc
Case E 1
βp 1
β
rp r
0.8
Case E
(b)
Figure 1. (a) Idealized domain: a unit cell with smear and transition zones. (b)
Variation of the hydraulic conductivity with distance from the centre of the
0.6
drain for different cases. Case B
k/kc

case of circular sand drains, the smear zone is probably also 0.4
Case C
circular. A variety of mandrels with different cross-sections
(circular, rectangular or diamond shaped) are used for PVD’s
0.2 σ ' = 118 kPa
(Holtz et al. 1991, Bo et al. 2003). The disturbed zone of a non-
circular mandrel is likely to have a non-circular cross-section. σ ' = 235 kPa
However, it is customary to convert the areas of non-circular 0
mandrels and the corresponding cross-sections of the smear 0 4 8 12 16 20
zones surrounding the PVD’s to equivalent circular areas.
Assuming that the disturbed zone has a single value of Normalized distance, r/rm,eq
hydraulic conductivity ks (figure 1, case A), a number of Figure 2. Normalized hydraulic conductivity profiles from field samples.
researchers (Holtz and Holm 1973, Jamiolkowski et al. 1983, (Reproduced from Madhav et al. 1993.)
Consolidation aided by vertical drains 65

To our knowledge, only three laboratory model studies 1


(Onoue et al. 1991, Indraratna and Redana 1998, Sharma and
Xiao 2000) have been performed to investigate the variation of Sharma and
hydraulic conductivity in the disturbed zone. The results of 0.8 Xiao (2000)
these studies are reproduced in figure 3. Onoue et al. (1991) Indraratna and Redana
(1998)
used a circular steel drain, which acted as a mandrel, in their
experiments. Consequently, in figure 3, r/rm,eq starts from 1 for 0.6

k/kc
the data of Onoue et al. (1991). Based on their study, Onoue Onoue et al. (1991)
et al. (1991) proposed a two-zone model for the disturbed zone. 0.4
However, unlike Madhav et al. (1993) and Miura et al. (1993),
Onoue et al. (1991) assumed a linear variation for the hydraulic
conductivity in the smear zone (figure 1, case C). This results in 0.2
a bilinear variation for the hydraulic conductivity; k increases at
one rate from ks at the drain boundary (i.e. the drain--soil inter-
face) to kt at the smear zone boundary, and at another rate from 0
kt at the smear zone boundary to kc at the transition zone 0 4 8 12 16 20
boundary. Normalized distance, r/rm,eq
Case C fits the PVD data of Sharma and Xiao (2000) well
(figure 3). However, no information regarding the variation of Onoue et al. (1991)
the hydraulic conductivity for the zone lying between r/rm,eq = 0
Onoue et al. (1991)
and r/rm,eq = 2 is available from their study (r/rm,eq for their
experiment starts from zero since they performed tests with Indraratna and Redana (1998)
PVDs of negligible thickness). Case C also fits the hydraulic Indraratna and Redana (1998)
conductivity data of Madhav et al. (1993) reasonably well Indraratna and Redana (1998)
(figure 2).
Indraratna and Redana (1998)
Holtz and Holm (1973) and Holtz et al. (1991) suggested that
the degree of disturbance decreases monotonically as the dis- Indraratna and Redana (1998)
tance from the drain increases, and therefore there is no distin- Indraratna and Redana (1998)
guishable smear zone (figure 1, case D). The data of Indraratna Sharma and Xiao (2000)
and Redana (1998) (figure 3) appear to follow the profile of
Sharma and Xiao (2000)
case D, although a paucity of data immediately adjacent to the
drain makes it difficult to ascertain the actual variation of the Sharma and Xiao (2000)
hydraulic conductivity in the smear zone. Figure 3. Normalized hydraulic conductivity profiles from laboratory model
A new case for the variation of hydraulic conductivity (figure studies.
1, case E) may be identified for the data of Sharma and Xiao
(2000) if the hydraulic conductivity is assumed to be constant assumed to lie at a distance of 2rm,eq to 5rm,eq from the centre
(with a value equal to the value at r/rm,eq = 2) in the zone between of PVD, and the transition zone boundary (beyond which the
r/rm,eq = 0 and r/rm,eq = 2. For case E, the hydraulic conductivity hydraulic conductivity does not vary with increasing distance
remains constant at ks within the smear zone and increases in the from the drain) can be assumed to vary between 6rm,eq and
transition zone following a bilinear curve with one slope from ks 15rm,eq. Jamiolkowski et al. (1983) suggested that the transition
at the smear zone boundary to kp at any intermediate point within zone radius can be up to 20rm,eq based on studies of pile driving
the transition zone (at r = rp) and a different slope from kp (at in clay. However, more laboratory and field studies are neces-
r = rp) to kc at the transition zone boundary. The hydraulic sary to determine the hydraulic conductivity profile and the
conductivity profile (figure 2) obtained by Madhav et al. (1993) corresponding dimensions of the smear and transition zones
can also be described by case E. that are most likely to occur in the field.
No definite conclusions regarding the variation of the
hydraulic conductivity in the disturbed zone can be drawn
from the experimental studies of PVD disturbance (Onoue 3. Theoretical studies on soil disturbance
et al. 1991, Madhav et al. 1993, Indraratna and Redana 1998,
Sharma and Xiao 2000) which take the transition zone into Theoretical studies on soil disturbance have generally been
account. These studies suggest that k/kc can be assumed to be restricted to case A. Analytical solutions for case A, assuming
about 0.2 in the immediate vicinity of the drain (r/rm,eq = 0) for a radial flow of water into the drain, were developed by Barron
cases B, C, D, and E, and for case C, k/kc can be assumed to vary (1948) and Hansbo (1981); their solutions can be used to
between 0.5 and 0.8 at the smear zone boundary. Based on the calculate the degree of consolidation as a function of time.
studies by Onoue et al. (1991), Madhav et al. (1993) and These formulations consider a vertical drain with a circular
Sharma and Xiao (2000), the smear zone boundary can be cross-section. The solution obtained by Barron (1948) is
66 D. Basu et al.

based on the Terzaghi--Rendulic theory of radial consolidation between the drain and the unit cell. This results in a radially
(Terzaghi 1925, Rendulic 1935, 1936), while that obtained by convergent horizontal flow of water into the drain. If a homo-
Hansbo (1981) is a simplified approach based on the continuity geneous deposit with no horizontal strain in the soil cylinder is
of flow and Darcy’s law. The Hansbo (1981) solution matches assumed, flow patterns are identical along any horizontal plane.
closely the rigorous solution obtained by Barron (1948) and is Consideration of only one such horizontal plane with axisym-
widely used in practice. Leo (2004) developed analytical solu- metric flow is sufficient to solve this problem. In addition, the
tions considering both radial and vertical flow. Numerical solu- flow of water is assumed to follow Darcy’s law. It is further
tions considering only smear (case A) also exist (Indraratna and assumed that the vertical strain within the unit cell is spatially
Redana 1997; Basu and Madhav 2000). uniform. This represents the case of ‘equal strain’ consolidation
Numerical studies of the variation of the hydraulic conduc- (Richart 1959).
tivity in the transition zone represented in cases B and C have For cases B, C, and E, the smear and transition zones are
also been reported (Madhav et al. 1993, Hawlader et al. 2002, assumed to have annular cross-sections with outer radii (as
Basu et al. 2005). Madhav et al. (1993) considered case B with measured from the centre of the drain) rsm and rtr, respectively
a simplified assumption that the hydraulic conductivity in the (rsm and rtr are the smear zone radius and the transition zone
transition zone is constant at a value equal to the average of ks radius, respectively). As can be seen in figure 1, rd , rsm , rtr , rc.
and kc, and used finite-difference analysis to study the PVD For case D, no smear zone is considered (figure 1). For all
response. Basu et al. (2005) also considered case B but used cases (B, C, D and E), the undisturbed zone lies between rtr 
finite-element analysis, taking into account the actual linear r  rc with r measured radially outward from the centre of the
variation of the hydraulic conductivity in the transition zone. drain.
Hawlader et al. (2002) considered case C and analysed the PVD
performance using an elasto-viscoplastic constitutive model.
However, most of these numerical studies are case specific 4.2 Average excess pore pressure
and cannot be directly used in design. Chai et al. (1997)
obtained an analytical solution for consolidation by PVD for 4.2.1 Case B. A radial coordinate system, where r represents
case D; however, their expressions are too complex for use in the radial distance from the centre of the drain, is used in the
routine design. analysis. In this case, the hydraulic conductivity ksm(r) within the
smear zone (i.e. for rd  r  rsm) is assumed to be a constant equal
to ks. In the transition zone (i.e. for rsm  r  rtr), the hydraulic
3.1 Scope of the present study conductivity ktr(r) increases linearly from ks at the smear zone
boundary (r = rsm) to kc at the transition zone boundary (r = rtr).
In this paper, we develop analytical solutions for consolidation
The hydraulic conductivity kc remains constant in the
by vertical drains, considering both the smear and the transition
undisturbed zone (i.e. for rtr  r  rc). The linear variation of
zones, which are easy to use. Solutions are obtained for cases B,
ktr(r) can be expressed mathematically as
C, D, and E using a methodology similar to that of Hansbo
(1981). r  rsm
In practice, a number of drains are installed in the ground, ktr ðrÞ ¼ ks þ ðkc  k s Þ for rsm  r  rtr : ð1aÞ
rtr  rsm
and each drain has a zone of influence. This zone of influence is
called a ‘unit cell’ because each cell behaves identically (for which can be rearranged as
homogeneous deposits), and water within one unit cell does not
flow into another unit cell. The analysis considers one such unit ktr ðrÞ ¼ A þ Br ð1bÞ
cell with a circular cross-section. The cross-sections of the
drain and the disturbed zone are assumed to be circular. where

ks rtr  kc rsm
A¼ ð2Þ
4. Analysis rtr  rsm

4.1 Definition of the problem and assumptions kc  ks


B¼ : ð3Þ
rtr  rsm
It is assumed that a drain with a circular cross-section of radius
rd is installed in a saturated soft soil deposit. The length of the The specific discharge vc in the undisturbed zone can be
drain spans the entire thickness of the soil deposit. An annular written as
cylinder of soil with inner and outer radii rd and rc (measured kc @uc
from the centre of the drain) is considered as the unit cell (figure vc ¼ for rtr  r  rc ð4aÞ
w @r
1) (rd and rc are the drain radius and the unit cell radius,
respectively). The effect of the flow of water in the vertical
direction within the unit cell is negligible (Leo 2004). Therefore where  w is the unit weight of water and uc is the excess pore
the only pervious boundary of the unit cell is the interface pressure at a distance r in the undisturbed zone. Similarly, the
Consolidation aided by vertical drains 67

specific discharges within the transition and smear zones can be 8 9  8 9
r2 r k 1 k @"v
written as þ c ln>
: tr s > ;
2
k c k s Aln >
: c> ; : ð7cÞ
A rsm kc B ks @t
ktr @utr
vtr ¼ for rsm  r  rtr ð4bÞ Let u be the average excess pore pressure throughout the unit
 w @r
ksm @usm cell. Then we can write the following equation:
vsm ¼ for rd  r  rsm ð4cÞ
 w @r rð rðtr
sm ðrc
 
The total volume of water entering a cylinder of arbitrary radius r  rc2  rd2 u ¼ 2rusm dr þ 2rutr dr þ 2ruc dr: ð8Þ
(r , rc) within the unit cell from the outer hollow cylinder (of rd rsm rtr
thickness rc -- r) must be equal to the change in volume of the
outer hollow cylinder. Using this concept, the pore pressure at any Substituting usm, utr, and uc from equations (7a), (7b), and (7c),
distance r within the unit cell can be related to the vertical strain ev respectively, in equation (8) and rearranging terms we obtain
(which is assumed to be uniform throughout the unit cell) as follows:
w rc2 @"v
  @"v u ¼  ð9Þ
2rvc ¼   r rc2 2
for rtr  r  rc ð5aÞ 2kc @t
@t
 2  @"v where
2rvtr ¼  rc  r 2 for rsm  r  rtr ð5bÞ
@t
  @"v  8 9 8 9 8 9 
2rvsm ¼  rc2  r 2 for rd  r  rsm ð5cÞ rc2 > rc > kc >rsm > kc ðrtr rsm Þ > ks rtr > 3
¼ 2 2 ln : ; þ ln : ; þ ln: ;
@t rc rd rtr ks rd ðks rtr kc rsm Þ kc rsm 4
  2 2 
where t is time. 1 kc  2  kc ðrtr rd Þ rtr rd
 2 2 rsm rd2 rtr2 þ
Replacing vc, vtr and vsm in equations (5a), (5b), and (5c) by rc rd ks ðks rtr kc rsm Þ
equations (4a), (4b), and (4c), respectively, we obtain 
1 kc  4  kc  3 3 
þ 2 2 2 rsm rd4 þ r r ðrtr rsm Þ
8 2 9 rc rc rd 4ks 3ðkc ks Þ tr sm
@uc w >> rc >
> @"v
¼ :  r ; for rtr  r  rc ð6aÞ kc ðks rtr kc rsm Þðrtr rsm Þ2
@r 2kc r @t  f5ðks rtr kc rsm Þþðrtr kc ks rsm Þg
8 2 9 2ðkc ks Þ3
@utr w >> r > @"v #
¼ : c  r> ; for rsm  r  rtr ð6bÞ 8 9
@r 2ktr r @t kc ðrtr rsm Þðks rtr kc rsm Þ3 >kc > rtr4
8 2 9  ln: ; :
@usm w > > r > @"v ðkc ks Þ4 ks 4
¼ : c  r>; for rd  r  rsm : ð6cÞ
@r 2ks r @t ð10Þ
Integrating equation (6c) and applying the boundary condi-
tion that the excess pore pressure is fully dissipated at the drain We now define the following dimensionless terms:
boundary (i.e. usm = 0 at r = rd), we obtain rc
n¼ ð11Þ
 8 9  rd
w 2 > r > 1 2 2
 @"v
usm ¼ rc ln: ;  r  rd : ð7aÞ
2ks rd 2 @t rsm
m¼ ð12Þ
rd
Integrating equation (6b) and using the continuity condition
utr = usm at r = rsm, we obtain
rtr
   8 9 q¼ ð13Þ
w rc2 ks r 1 AþBr > rd
utr ¼ ln  2 AþBr ks Aln>
: ;
2 A ðAþBrÞrsm B ks ks
 8 9  ¼ : ð14Þ
1 2 >rsm > 1  2  @"v kc
þ rc ln: ;  rsm rd2 : ð7aÞ
ks rd 2 @t
Equation (10) can then be rewritten in terms of these quantities
Similarly, integrating equation (6a) and using the continuity as
condition uc = utr at r = rtr, we obtain  8 9 8 9 
n2 >
> n>> 1 ðqmÞ >q> 3
  8 9  ¼ 2 ln: ;þ lnðmÞþ ln: ;
w 1 2 > r > 1  2 2  n 1 q  ðqmÞ m 4
uc ¼ r ln : ;  r rtr
2 kc c rtr 2  
 8 9   1 1 2  ðq  mÞðq2  m2 Þ
1 2 >rsm > 1 2 2
  2 2
m 1 q þ
þ r ln : ;  rsm rd n 1  ðq  mÞ
ks c rd 2
68 D. Basu et al.

 transition zone boundary (r = rtr). The expression for ktr(r) can


1 1 4 
be obtained from equation (1a) by replacing rsm by rd. As
þ 2 2 m 1
n ðn  1Þ 4 before, the hydraulic conductivity kc in the undisturbed zone
1  3  is a constant. The expression for m (associated with equation
þ q  m3 ðq  mÞ (9)) is derived following the same procedure as outlined for
3ð1  Þ
case B. After eliminating the terms which make a negligible
contribution, the following equation is obtained for m:
ðq  mÞðq  mÞ2
 f5ðq  mÞ þ ðq  mÞg 8 9
2ð1  Þ3  ¼ ln>
> n> ð q  1Þ 3
: > ;þ lnðqÞ  : ð19Þ
q ðq  1Þ 4
8 9 #
ðq  mÞðq  mÞ3 > 1 > q4
 ln: ;  : ð15aÞ
ð1  Þ4  4
4.2.4 Case E. In this case, the hydraulic conductivity ksm(r)
has a constant value ks within the smear zone (i.e. for rd  r 
Equation (15a) is too cumbersome for use in routine design.
rsm) and increases in the transition zone, following a bilinear
However, a number of terms on the right-hand side make a
curve with one slope between ks (at r = rsm) and kp (at r = rp, say)
negligible contribution to the value of m. If we neglect these
and another slope between kp (at r = rp) and kc (at r = rtr).
terms, equation (15a) simplifies to
Thereafter, the hydraulic conductivity in the undisturbed zone
8 9 8 9 remains constant at kc. This variation can be described
> n> 1 ðq  mÞ >q> 3
 ¼ ln>
: > ; þ lnðmÞ þ ln: ;  : ð15bÞ mathematically as follows:
q  ðq  mÞ m 4
8 9 8 9
>
> n>> 1 ðp  mÞ >
> p >>
The ratio n2/(n2 -- 1) is close to unity for the typical unit cell and  ¼ ln: ; þ lnðmÞ þ ln: ;
q  8 9 ðp  p mÞ p m
drain diameters used in practice, and is not included in equation ð q  pÞ >  p q> 3
(15b). þ  ln>
: > ; ð20Þ
p q  p p 4

4.2.2 Case C. In this case, the hydraulic conductivity where the dimensionless terms p and bp are defined as
ksm(r) in the smear zone varies from ks at the drain--soil rp
interface (r = rd) to kt at the smear zone boundary (r = rsm), p¼ ð21Þ
rd
and is given by
kp
r  rd p ¼ ð22Þ
ksm ðrÞ ¼ ks þ ðkt  ks Þ for rd <r<rsm : ð16Þ kc
rsm  rd
where rd , rsm , rp , rtr , rc and ks , kp , kc.
The expression for the hydraulic conductivity ktr(r) in the
transition zone can be obtained from equation (1a) by
replacing ks by kt. Again, the hydraulic conductivity kc in 4.3 Degree of consolidation
the undisturbed zone is a constant. Following steps similar to
those of case B, and eliminating the terms which make a If we assume that all the excess pore pressure due to preloading
negligible contribution to the value of m, we obtain the is developed instantly, we can write the following relationship:
following expression for m (associated with equation (9)): @"v @ 0 @u
¼ mv ¼ mv ð23Þ
8 9 8 9 @t @t @t
> n> ðm  1Þ >m>
 ¼ ln>
: > ;þ ln: ;
q ðm8 t Þ9 t where 0 is the average effective stress in the unit cell due to
ð q  m Þ >  t q> 3 preloading at the end of consolidation, u is the average excess
þ ln: ;  : ð17Þ
ðt q  mÞ m 4 pore pressure at the time of load application, and mv is the
coefficient of volume compressibility.
The dimensionless term bt is defined as follows: The coefficient of consolidation ch in the horizontal direction
and the time factor T are defined as follows:
kt
t ¼ : ð18Þ kc
kc ch ¼ ð24Þ
mv w

4.2.3 Case D. In this case the disturbed zone consists of the


ch t
transition zone of radius rtr, and the hydraulic conductivity T¼ ð25Þ
ktr(r) varies from ks at the drain boundary (r = rd) to kc at the 4rc2
Consolidation aided by vertical drains 69

Substituting equation (9) into equation (23) we obtain the linear


differential equation 100

n = 17.05
m = 2.69

Degree of consolidation, U (%)


d
u 2kc 80
þ u ¼ 0:
 ð26Þ q = 16.17
dt mv w rc2  β = 0.2
β t = 0.6
u ¼ u0 at t = 0,
Solving equation (26) using the initial condition  60

where u0 is the initial average excess pore pressure, and using
the dimensionless terms defined in equations (24) and (25), we 40
obtain the change in average excess pore pressure with time:
8T
20
 u0 e   :
u¼ ð27Þ
The degree of consolidation U at a particular time t (or time
0
factor T ) is the ratio of the excess pore pressure dissipated to the
0.001 0.01 0.1 1 10
excess pore pressure induced at that time. U can be expressed
mathematically as follows: Time f actor, T
Case A

u Case B
U ¼1 : ð28Þ

u0 Case C
Case D
Substituting equation (27) in equation (28) gives the following
(a)
expression for the degree of consolidation:

8T 100
U ¼ 1  e  : ð29Þ
n = 51.14
Degree of consolidation, U (%)

m = 5.11
80
q = 30.67
β = 0.2
5. Results β t = 0.6
60
5.1 Consolidation rates for different cases
40
In order to determine the influence of the various hydraulic
conductivity profiles described above on the consolidation rate,
the solution for case A given by Hansbo (1981) is reproduced 20
here so that a comparison can be made:
8n9 1 3 0
 ¼ ln: ; þ lnðmÞ  : ð30Þ 0.001 0.01 0.1 1 10
m  4
Time f actor, T
Hansbo (1981) suggested that, by using an equivalent radius
rd,eq, the analytical solutions can also be applied to PVD’s. The Case A
equivalent radius is calculated as follows: Case B
1 Case C
rd;eq ¼ ðbw þ bt Þ ð31Þ Case D

(b)
where bw and bt are the width and thickness, respectively, of the
PVD. Rectangular or hexagonal unit cells are obtained when Figure 4. Plots of degree of consolidation versus time factor for different
hydraulic conductivity profiles: (a) spacing of 1 m; (b) spacing of 3 m. The
PVD’s are installed in rectangular or triangular patterns (Holtz curve for case D is just to the right of the curve for case A. The curve for case A
et al. 1991). In order to use the analytical solutions, these shapes corresponds to the solution obtained by Hansbo (1981).
need to be replaced by equivalent circles which have the same
area as the rectangular or hexagonal unit cell. The equivalent
radius rc,eq of the unit cell for a rectangular installation pattern is
rffiffiffiffiffiffiffiffi where sx and sy are the spacings of the PVD’s in two mutually
sx sy perpendicular directions. For a triangular pattern, the equivalent
rc;eq ¼ ð32Þ
 radius is given by
70 D. Basu et al.

Table 1. Solution of examplea

Case rsm (mm) rtr (mm) rp (mm) m q p n b bt bp m

B 135.4 744.7 -- 4.09 22.50 -- 34.09 0.2 -- -- 11.00


C 304.7 880.1 -- 9.20 26.59 -- 34.09 0.2 0.75 -- 7.50
E 135.4 1015.5 466.9 4.09 30.68 14.11 34.09 0.2 -- 0.9 10.32
a
rd = 33.1 mm; rc = 1128.4 mm.

sffiffiffiffiffiffi
pffiffiffiffi
3 5.1.2 Example In order to understand the impact of the
rc;eq ¼ s ð33Þ various hydraulic conductivity profiles on the rate of
2
consolidation, a practical example is analysed for all the
hydraulic conductivity profiles of figure 2. It is assumed that
where s is the PVD spacing.
the PVDs were installed with a mandrel of cross-section 120
Figures 4(a) and 4(b) show plots of the degree of con-
mm  120 mm (rm,eq = 67.7 mm), the PVDs have a cross-
solidation U versus time factor T for PVDs installed in a
section of 100 mm  4 mm (rd,eq = 33.1 mm), and the clay at the
rectangular arrangement with centre--centre spacings of 1 m
site has ch = 10 m2/year.
(rc,eq = 564.2 mm) and 3 m (rc,eq = 1692.6 mm), respec-
For a hydraulic conductivity profile corresponding to case B,
tively. Four hydraulic conductivity profiles (cases A, B, C
the smear zone extends to 2rm,eq and the transition zone extends
and D) are considered. The PVDs are assumed to have a cross-
to 11rm,eq (figure 2). If the hydraulic conductivity profile cor-
section of 100 mm  4 mm (rd,eq = 33.1 mm). Mandrels with a
responds to case C, rsm and rtr are 4.5rm,eq and 13rm,eq, respec-
rectangular cross-section (a  b) are considered, with dimen-
tively. However, if the hydraulic conductivity profile
sions 125 mm  50 mm (Saye 2003) (rm,eq = 44.6 mm) for a
corresponds to case E, rsm, rp, and rtr are equal to 2rm,eq,
spacing of 1 m and 150 mm  150 mm (Bergado et al. 1993b)
7rm,eq, and 15rm,eq, respectively. The degree of disturbance b
(rm,eq = 84.6 mm) for a spacing of 3 m. The equivalent mandrel
near the drain can be taken as 0.2 for all the cases (figure 2). For
radii are obtained from equation (32) by replacing sx and sy
case C, bt = 0.75 and for case E, bp = 0.9 (figure 2). A square
by a and b, respectively. The extent of the disturbed zone is
arrangement of PVDs with a centre--centre spacing of 2 m
defined by rsm = 2rm,eq (except for case D) and rtr = 12rm,eq.
(rc,eq = 1128.4 mm) is chosen. The values of m calculated
The degree of disturbance b at the drain surface is taken as
for cases B, D, and E (table 1) are 11.00, 7.50, and 10.32,
0.2. For case C, bt = 0.6 is assumed.
respectively. The value of T for U = 90% is calculated
Figures 4(a) and 4(b) indicate that the hydraulic conductivity
from equation (29) as 3.17, 2.16, and 2.97 for cases B, C,
profile in the disturbed zone has a definite impact on the rate of
and E, respectively. For ch = 10 m2/year, the actual times
consolidation. In figure 4(a), the time factors T at U = 90%
required for 90% consolidation are 1.6 years, 1.1 years, and
corresponding to cases A, B, C, and D are 1.74, 2.54, 1.37, and
1.5 years for cases B, C, and E, respectively.
2.09, respectively. For ch = 1 m2/year, the corresponding actual
times are 2.2, 3.2, 1.7 and 2.7 years. With respect to case A
(Hansbo 1981), the increase in time (or time factor) required for
90% consolidation is 46% and 20% for cases B and D, respec- 6. Conclusions
tively; for case C, the time required for 90% consolidation
decreased by 21%. The time factors corresponding to U = 90% Installation of vertical drains disturbs the soil around the drain.
in figure 4(b) are 2.79, 3.59, 2.13, and 2.95 for cases A, B, The hydraulic conductivity of the disturbed soil is less than that of
C, and D, respectively. The increases in T for cases B and D the original soil, reducing the acceleration of the consolidation
compared with case A are 29% and 6%, respectively, while process caused by the presence of the drains to less than it would
for case C the decrease in T relative to case A is 24%. be in the absence of disturbance. A number of researchers have
It is clear from these results that a proper knowledge of the proposed various hydraulic conductivity profiles in the disturbed
hydraulic conductivity profile in the disturbed zone is needed zones. Five possible hydraulic conductivity profiles (cases A, B,
for accurate design. In addition, neglecting the transition zone C, D, and E) have been considered in this paper. An analytical
in design may lead to errors in the estimation of the consolida- solution for the rate of consolidation, corresponding to case A, is
tion rate. Knowledge of the degree of soil disturbance in the already available in the literature (Hansbo 1981). Analytical solu-
immediate vicinity of the drain is of utmost importance for tions for the remaining cases have been developed in this paper.
predicting drain performance. This is evident by comparing Our analyses showed that the transition zone has a definite
the curves for cases B, C, and D. For cases B and D, k/kc is impact in slowing down the consolidation process and therefore
approximately 0.2 in the vicinity of the drain. However, for case must be considered in design. Moreover, the rate of consolidation
C this ratio increases from 0.2 to 0.6 in the vicinity of the drain. can vary greatly depending on how the hydraulic conductivity
Consequently, the difference in response between cases C and varies within the transition zone. Hence, proper identification of
B or cases C and D is more than that observed when cases B and the hydraulic conductivity profile around a vertical drain is
D are compared. necessary for accurate prediction of the rate of consolidation.
Consolidation aided by vertical drains 71

The experimental data available in the literature concerning the Hansbo, S., Practical aspects of vertical drain design, in Proceedings of the 14th
International Conference on Soil Mechanics and Foundation Engineering,
variation of the hydraulic conductivity within the transition zone Hamburg, 1997, pp. 1749--1752.
was collected and analysed. Definite conclusions regarding the Hawlader, B.C., Imai, G. and Muhunthan, B., Numerical study of the factors
most likely hydraulic conductivity profile could not be reached affecting the consolidation of clay with vertical drains. Geotext
Geomembranes, 2002, 20, 213--239.
because of the limited amount of experimental data. Until more Hird, C.C. and Moseley, V.J., Model study of seepage in smear zones around
information regarding this issue becomes available, all possible vertical drains in layered soil. Geotechnique, 2000, 50(1), 89--97.
hydraulic conductivity profiles, as outlined in this paper, should Holtz, R.D., Preloading with prefabricated vertical strip drains. Geotext
Geomembranes, 1987, 6, 109--131.
be considered before final design decisions are made. Holtz, R.D. and Holm, B.G., Excavation and sampling around some sand drains
in Skå-Edeby, Sweden. Sartryck och Preliminara Rapporter, 1973, 51,
79--85.
Holtz, R.D., Jamiolkowski, M.B., Lancellotta, R., and Pedroni, R.,
Prefabricated Vertical Drains: Design and Performance, 1991
References (Butterworth Heinemann: Oxford).
Indraratna, B. and Redana, I.W., Plane-strain modeling of smear effects asso-
Barron, R.A., Consolidation of fine-grained soils by drain wells. Trans. ASCE, ciated with vertical drains. J. Geotech. Geoenviron. Eng., 1997, 123(5),
1948, 113, 718--742. Reprinted in A History of Progress, Vol. 1, pp. 324- 474--478.
348, 2003 (ASCE: Reston, VA). Indraratna, B. and Redana, I.W., Laboratory determination of smear zone due to
Basu, D. and Madhav, M.R., Effect of prefabricated vertical drain clogging on vertical drain installation. J. Geotech. Geoenviron. Eng., 1998, 124(2),
the rate of consolidation: a numerical study. Geosynth Int., 2000, 7(3), 180--184.
189--215. Jamiolkowski, M., Lancellotta, R. and Wolski, W., Precompression and speed-
Basu, D., Basu, P., and Prezzi, M. Study of consolidation by prefabricated ing up consolidation, in Proceedings of the 8th European Conference on
vertical drain. Internal Geotechnical Report 2005-01, 2005 (Purdue Soil Mechanics and Foundation Engineering, 1983, Vol. 3, pp. 1201--1226
University: West Lafayette, IN). (A.A. Balkema: Rotterdam).
Bergado, D.T., Asakami, H., Alfaro, M.C. and Balasubramaniam, A.S., Smear Johnson, S.J., Foundation precompression with vertical sand drains. J. Soil Mech.
effects on vertical drains on soft Bangkok clay. J. Geotech. Eng.--ASCE, Fdn. Div., 1970, 96(SM1), 145--175.
1991, 117(10), 1509--1530. Leo, C.J., Equal strain consolidation by vertical drains. J. Geotech. Geoenviron.
Bergado, D.T., Alfaro, M.C. and Balasubramaniam, A.S., Improvement of soft Eng., 2004, 130(3), 316--327.
Bangkok clay using vertical drains. Geotext Geomembranes, 1993a, 12, Lo, D.O.K., and Mesri, G., Settlement of test fills for Chek Lap Kok airport. in
615--663. Vertical and Horizontal Deformations of Foundations and Embankments,
Bergado, D.T., Mukherjee, K., Alfaro, M.C. and Balasubramaniam, A.S., edited by A.T. Yeung and G. Feaalio, pp. 1082--1099, 1994 (American
Prediction of vertical-band-drain performance by the finite-element Society of Civil Engineers: New York).
method. Geotext Geomembranes, 1993b, 12, 567--586. Madhav, M.R., Park, Y.-M. and Miura, N., Modelling and study of smear zones
Bergado, D.T., Balasubramaniam, A.S., Fannin, R.J., Anderson, L.R., and around band shaped drains. Soils Found, 1993, 33(4), 135--147.
Holtz, R.D., Full scale field test of prefabricated vertical drain (PVD) on Mesri, G., Lo, D.O.K., and Feng, T-W., Settlement of embankments on soft
soft Bangkok clay and subsiding environment. in Ground Improvement, clays. in Vertical and Horizontal Deformations of Foundations and
Ground Reinforcement, Ground Treatment: Developments 1987--1997, Embankments, edited by A.T. Yeung and G. Feaalio, pp. 8--56, 1994
edited by V.R. Schaefer, pp. 372--393, 1997 (American Society of Civil (American Society of Civil Engineers: New York).
Engineers: New York). Miura, N., Park, Y. and Madhav, M.R., Fundamental study on drainage perfor-
Bergado, D.T., Balasubramaniam, A.S., Fannin, R.J. and Holtz, R.D., mance of plastic-board drains. J. Geotech. Eng.--JSCE, 1993, 481(III-25),
Prefabricated vertical drains (PVDs) in soft Bangkok clay: a case study of 31--40. (in Japanese).
the new Bangkok International Airport project. Can. Geotech. J., 2002, 39, Onoue, A., Ting, N.-H., Germaine, J.T. and Whitman, R.V., Permeability of
304--315. disturbed zone around vertical drains, in Geotechnical Engineering
Bo, M.W., Chu, J., Low, B.K., and Choa, V., Soil Improvement: Prefabricated Congress, Proceedings of the Congress of the Geotechnical Engineering
Vertical Drain Techniques, 2003 (Thomson Learning: Stamford, CT). Division, 1991, pp. 879--890 (American Society of Civil Engineers: New
Casagrande, L. and Poulos, S., On the effectiveness of sand drains. Can. York).
Geotech. J., 1969, 6(3), 287--326. Rendulic, L., Der hydrodynamische Spannungsausgleich in zentral entwässer-
Chai, J.C., Miura, N. and Sakajo, S., A theoretical study on smear effect around ten Tonzylindern. Wasserwirtsch-Wassertech, 1935, 2, 250--253.
vertical drain, in Proceedings of the 14th International Conference on Soil Rendulic, L., Porenziffer und Porenwasserdruck in Tonen. Bauingenieur, 1936,
Mechanics and Foundation Engineering, Hamburg, 1997, pp. 1581--1584. 17, 559--564.
Chai, J.-C., Shen, S.-L., Miura, N. and Bergado, D.T., Simple method of Richart, F.E., Review of the theories for sand drains. Trans. ASCE, 1959, 124,
modeling PVD-improved subsoil. J. Geotech. Geoenviron. Eng., 2001, 709--736.
127(11), 965--972. Saye, S.R., Assessment of soil disturbance by the installation of displacement
Hansbo, S., Consolidation of fine-grained soils by prefabricated drains, in sand drains and prefabricated vertical drains. in Soil Behavior and Soft
Proceedings of the 10th International Conference on Soil Mechanics and Ground Construction, pp. 372--393, 2003 (American Society of Civil
Foundation Engineering, Stockholm, 1981, pp. 677--682. Engineers: New York).
Hansbo, S., Preconsolidation of soft compressible subsoil by the use of pre- Sharma, J.S. and Xiao, D., Characterization of a smear zone around vertical
fabricated vertical drains. Ann Trav Publics Belg, 1986, 6, 553--563. drains by large-scale laboratory tests. Can. Geotech. J., 2000, 37,
Hansbo, S., Design aspects of vertical drains and lime column installations, in 1265--1271.
Proceedings of the 9th Southeast Asian Geotechnology Conference, Terzaghi, K., Erdbaumechanik auf bodenphysikalischer Grundlage, 1925
Bangkok, 1987, pp. 1--12. (Deuticke: Vienna).

You might also like