See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/340850259

Performance of a geosynthetic-encased column (GEC) in soft ground:

numerical and analytical studies

Article · January 2013

DOI: 10.1680/gein.13.00015]

CITATIONS READS

19 292

4 authors, including:

Iman Hosseinpour Mario Riccio

University of Guilan Federal University of Juiz de Fora
63 PUBLICATIONS 1,295 CITATIONS 54 PUBLICATIONS 874 CITATIONS

SEE PROFILE SEE PROFILE

Marcio Almeida
Federal University of Rio de Janeiro
298 PUBLICATIONS 2,559 CITATIONS

SEE PROFILE

All content following this page was uploaded by Marcio Almeida on 22 April 2020.

The user has requested enhancement of the downloaded file.

 Geosynthetics International, 2013, 20, No. 4

Performance of a geosynthetic-encased column (GEC)

in soft ground: numerical and analytical studies
M. S. S. Almeida1 , I. Hosseinpour2 and M. Riccio3
1
Professor (corresponding author), Department of Civil and Environmental Engineering, COPPE,
Federal University of Rio de Janeiro, Ilha de Fundao, RJ 21945-970, Brazil,
Telephone: +55 21 25627772, Telefax: +55 21 25627200, E-mail: [email protected].
2
Doctoral student, Department of Civil and Environmental Engineering, COPPE, Federal University of
Rio de Janeiro, Ilha de Fundao, RJ 21945-970, Brazil, Telephone: +55 21 80253856,
Telefax: +55 21 25627200, E-mail: [email protected].
3
Post-doctoral researcher, Department of Civil and Environmental Engineering, COPPE, Federal
University of Rio de Janeiro, Ilha de Fundao, RJ 21945-970, Brazil, Telephone: +55 21 81875422,
Telefax: +55 21 25627200, E-mail: [email protected].

Received 15 October 2012, revised 20 May 2013; accepted 7 June 2013

ABSTRACT: Stone columns with granular material are often used to support embankments over soft
soil. In extremely soft soils, encasing the granular column with geosynthetic material improves their
stiffness and strength. Based on the soil and column properties, settlements and stresses can be
determined by means of analytical solutions or numerical analyses, and this paper compares the
results of these two procedures by means of parametric analyses. Two-dimensional finite-element
analysis (FEA) was used to simulate the behavior of a geosynthetic-encased stone column in soft
clay loaded by an embankment. Parametric studies are then carried out by varying the thickness of
the soft soil layer and the tensile stiffness of the geosynthetic encasement for a given set of soft clay
and column parameters. The results of FEA have shown that the influence of the geosynthetic
encasement on settlement improvement increases in shallow soft soil layers. It was also observed that
the dependence of stress concentration on the thickness of the soft soil increases with geosynthetic
stiffness. The critical embankment height where ground settlement is negligible was also investigated.

KEYWORDS: Geosynthetics, Granular column, Numerical analyses, Settlements, Stresses, Tensile

forces.

REFERENCE: Almeida, M. S. S., Hosseinpour, I. & Riccio, M. (2013). Performance of a geosynthetic-

encased column (GEC) in soft ground: numerical and analytical studies. Geosynthetics International, 20,
No. 4, 252–262. [http://dx.doi.org/10.1680/gein.13.00015]

Soil reinforcement is one of the most useful techniques

1. INTRODUCTION for improving the strength properties of soft soils, and
Structures built on soft strata may experience problems geosynthetics are the main materials used for increasing
such as excessive settlement, large lateral deformation of the strength and stability of geotechnical structures (Gray
the soft layer beneath the structures, and global or local and Ohashi 1983; Reugger 1986; Hosseinpour et al.
instability. Ground improvement with an emphasis on the 2010). Encasing the granular columns with suitable geo-
granular column technique overcomes these problems by synthetic material is one of the best ways to improve the
reducing the total settlement under loading, and speeding performance of columns in very soft soil (Su , 15 kPa)
up the consolidation process. The existence of the columns foundations, or even to make their application possible at
creates a composite material, stiffer than the original soil, all in very soft soils. In addition to this, encasement
which attains its load capacity from the confinement prevents the lateral squeezing of stones into the surround-
provided by the surrounding soil. When granular columns ing soil, and vice versa (Murugesan and Rajagopal 2010).
are installed in extremely soft clay, insufficient lateral The concept of encasing individual granular columns with
confinement, especially in the upper portion of the geosynthetic materials, over the full or partial length, in
columns, may significantly reduce their capacity (Green- order to increase their load-carrying capacity has been
wood 1970; Barksdale and Bachus 1983; Murugesan and evaluated by numerous researchers (Raithel and Kempfert
Rajagopal 2007). 2000; Raithel et al. 2002; Alexiew et al. 2005; Brokemper
1072-6349 # 2013 Thomas Telford Ltd 252

Downloaded by [ Universidade Federal do Rio de Janeiro] on [29/11/17]. Copyright © ICE Publishing, all rights reserved.
 Performance of a geosynthetic-encased column (GEC) in soft ground: numerical and analytical studies 253

et al. 2006; Murugesan and Rajagopal 2006, 2007, 2010; (2000) method was used to allow a direct comparison for
Lee et al. 2007; Araujo et al. 2009; Gniel and Bouazza the long-term condition between FEM and the analytical
2009; Yoo and Kim 2009; Khabbazian et al. 2010; Yoo method.
2010; Pulko et al. 2011; Zhang et al. 2011; Ali et al. The Raithel and Kempfert (2000) analytical model was
2012; Hong 2012; Keykhosropur et al. 2012; Yoo and Lee proposed as the basis for the conventional calculation of
2012) and will be studied here by means of numerical and granular columns with geotextile casing. As shown in Figure
analytical methods. 1, there is an incremental horizontal stress in the column,
In addition to the various experimental and theoretical ˜h,c (where subscript h ¼ horizontal, c ¼ column), owing
studies, numerical modeling can be used as an effective to the additional vertical stress ˜v,c (where subscript
tool for investigating the effectiveness of geosynthetic v ¼ vertical) over the column head. By knowing the radius
encasement in stone column applications for soft ground rgeo and the change in displacement ˜rgeo of the geotextile
improvement. Malarvizhi and Ilamparuthi (2007) carried casing, and its tensile stiffness J, the horizontal stress h,geo
out numerical analyses to understand the behavior of an acting on the geotextile can be determined from
encased stone column in soft clay to produce the param- FR
eters that play a dominant role in load sharing and  h,geo ¼ (1)
rgeo
settlement reduction. Axisymmetric analyses and the unit
cell concept have been adopted to evaluate the effect of where the ring tensile force FR is given by
geosynthetic encasement on the behavior of stone columns ˜rgeo
(e.g. Murugesan and Rajagopal 2006; Yoo and Kim 2009; FR ¼ J (2)
rgeo
Khabbazian et al. 2010), sometimes complemented by
full-scale embankment loading (Yoo and Kim 2009). By use of the separate horizontal stresses, a differential
This paper presents a numerical evaluation by the horizontal stress can be defined, which represents the
finite-element method (FEM) considering an axisym- partial mobilization of the passive earth pressure in the
metric model and using the unit cell concept of a GEC surrounding soft soil, as
(geosynthetic-encased stone column) (EBGEO 2010).  h,diff ¼  h,c  ( h,s þ  h,geo ) (3)
Here, finite-element analysis (FEA) is performed for long-
term conditions (drained analysis) to assess the influence The stress difference (h,diff ) results in an expansion of
of embankment loading on the settlement, the vertical the column. More details of the method can be found in
stress on the column and soil, and the tensile force in the Raithel and Kempfert (2000).
geosynthetic material. First, the adopted model is verified The hypothetical problem modeled is shown schemati-
by the Raithel and Kempfert (2000) analytical method, cally in Figure 2a, which consists of a 2.0 m diameter
and then numerical modeling and analytical solution are GEC unit cell (rE ¼ 1.0 m) and geosynthetic tensile stiff-
compared by varying some of the parameters to investi- ness J. The column is arranged in a triangular grid pattern
gate the changes in the stresses and settlement in both at 1.90 m spacing (s), and the maximum applied vertical
methods. stress is 120 kPa, simulating a 6.0 m high embankment.
The parameters used in the analytical model are presented
in Table 1 (subscripts s for soft soil and c for column).
2. CALCULATION METHODS Most of these parameters were defined above; ª9 is the
submerged weight, Eoed,s is the oedometer (constrained)
2.1. Analytical solution modulus of the soft soil, pref is the reference mean stress,
Raithel and Kempfert (2000) presented a closed- and m is the power of the stress dependence of stiffness.
form analytical solution for calculating and designing a
geotextile-encased column foundation, thus allowing the 2.2. Numerical modeling and material parameters
stresses and deformations in both the encased column and In order to simulate the unit cell, an axisymmetric model
the soft soil to be obtained. The method adopted in was undertaken using the FEM. The PLAXIS 2D program
EBGEO (2010) is in essence the Raithel and Kempfert (Brinkgreve and Vermeer 2002) was used, thus allowing
(2000) method. The solution of the proposed equations is calculation of the tensile strains and hoop tensile forces
iterative, as the method assumes equal settlements for the acting on the geosynthetic at any depth, as well as
soft soil and columns. The Raithel and Kempfert (2000) settlements in the column and surrounding soil separately.
computations are for the long-term drained condition at By contrast, the analytical solution considers equal dis-
which maximum settlements and ring tension forces are placements for the soil and the column. Concerning the
attained. It is also assumed that the column reaches an constitutive models, the soft clay was simulated using the
active lateral pressure condition, so that the coefficient of HS model, so that the oedometer modulus Eoed,s is stress
active earth pressure (Kac ) applies, and also that the dependent. An elasto-plastic Mohr–Coulomb model was
geosynthetic encasement is linearly elastic. Castro and adopted for both the granular column and the embankment
Sagaseta (2011) also proposed an analytical solution, material. The geosynthetic was simulated as a linear
using a unit cell for an end-bearing, granular-encased elastic material with tensile stiffness J. Table 2 shows the
column and its surrounding soil, including the consolida- parameters used in the FEM. The material properties
tion process. The present study is restricted to the final chosen for the column and the embankment are typical of
drained condition: therefore the Raithel and Kempfert these materials.
Geosynthetics International, 2013, 20, No. 4

Downloaded by [ Universidade Federal do Rio de Janeiro] on [29/11/17]. Copyright © ICE Publishing, all rights reserved.
 254 Almeida, Hosseinpour and Riccio
Δσ

Horizontal geogrid layer

σv,c

σv,s Encasement

Soft soil Column

r

Granular column
Bearing layer

Soft soil

σh,s,tot ⫽ Column pattern

σh,c
Hs σh,s ⫹ σh,diff

Soft soil S1
Geosynthetic

Encased column, Ac
dc S2

S⬘
2rc Unit cell, AE
2rgeo
aE ⫽ Ac /AE
Aunitcell ⫽ S1(S22 ⫺ S12/4)1/2
2rE
S⬘ ⫽ (S22 ⫹ 2S12)1/2/3 ⫹ d/2

Figure 1. Axisymmetric unit cell

Table 1. Geometric and material parameters used in

analytical model
He ⫽ 6 m

Parameter Value
Embankment
Area replacement ratio, aE 0.122
Encased column radius, rc (m) 0.35
Geotextile stiffness, J (kN/m) 1000
y
Oedometer constrained modulus of soft soil, Eoed,s (kPa) 750
r Effective cohesion of soft soil, c9s (kPa) 7.0
x
Effective friction angle of column, 9c (degrees) 38.0
Effective friction angle of soft soil, 9s (degrees) 24.0
Geosynethetic, J
Submerged unit weight of soft soil, ª9s (kN/m3 ) 5.0
Submerged unit weight of encased column, ª9c (kN/m3 ) 8.0
Lateral earth pressure of soft soil, K0,s (dimensionless) 0.55
Reference mean stress/power of stress dependence of 100/1.0
Soft soil stiffness, pref /m
Hs

γ2, φs, cs

2.3. Model description

Column
γc, φc
In the numerical model adopted in this study, shown in
Figure 2, the axis of symmetry coincides with the center
r ⫽ 0.35 m of the column (Figure 2a). Figure 2b shows the boundary
rgeo c aE ⫽ 0.122
z conditions and the materials involved: column, soft soil
rE ⫽ 1.0 m
and geosynthetic material. As can be seen, the horizontal
(a) (b)
displacements were restrained in the axis of the column,
Figure 2. Finite-element axisymmetric simulation of and also laterally (external part of the unit cell). It is also
geosynthetic-encased column in unit cell concept: (a) scheme assumed that the column base rests on a rigid stratum:
of GEC adopted in numerical analyses; (b) boundary thus horizontal and vertical fixities were employed at the
condition and finite-element mesh bottom of the column. Simulation of the installation of the
Geosynthetics International, 2013, 20, No. 4

Downloaded by [ Universidade Federal do Rio de Janeiro] on [29/11/17]. Copyright © ICE Publishing, all rights reserved.
 Performance of a geosynthetic-encased column (GEC) in soft ground: numerical and analytical studies 255

Table 2. Constitutive models and material properties used in FEM

Property Granular column Soft clay Embankment

Mohr–Coulomb Hardening soil Mohr–Coulomb

Saturated unit weight, ªsat (kN/m3 ) 18.0 15.0 20.0

Drained Young’s modulus, E9 (kPa) 15500 725 43000
Effective Poisson’s ratio, 9 0.30 0.40 0.33
Effective cohesion, c9 (kPa) 0.0 7.0 5.0
Effective friction angle, 9 (degrees) 38.0 20.0 25.0
Angle of dilatancy, ł (degrees) 5.0 0.0 0.0
Lateral earth pressure coefficient, K0 0.38 0.55 0.58
Power of stress dependence of stiffness, m – 1.0 –
Reference mean stress, pref (kPa) – 100 –
Oedometer modulus, Eoed (kPa) – 750 –

encased column, which is usually by lateral displacement, Verification is carried out in terms of settlement, vertical
was not considered. However, encased columns are less stress on the soil and the encased column, and the
commonly installed by non-displacement methods (i.e. geosynthetic tensile force is plotted against the embank-
using open-ended tubes with removal of the soft clay ment height with a maximum value of 6 m. In Figure 3a
inside the tube): therefore the present numerical simulation the embankment height–settlement response obtained
is applied to this installation method. In addition, the from the FEM is compared with the AM. It can be seen
purpose of this paper is a direct comparison between the that the FEM appears to meet the general trend of the
FE and analytical methods, and the laiter does not consider analytical results well. The variations in vertical stresses
column installation effects. It is observed in Figure 2 that and tensile force in the geosynthetic obtained from the
in the numerical analyses the embankment loading was current model, illustrated in Figures 3b and 3c, respec-
simulated by actual embankment material placed (instanta- tively, compared even better than the settlements. These
neously) on the top of the soil/column surface. No curves suggest that the FEM adopted appears to be
geosynthetic reinforcement was considered at this inter- appropriate for a parametric study of the GEC.
face, also because this interface is not considered in the
analytical model, and one of the purposes of the present
study is to compare the FE and analytical methods.
4. PARAMETRIC STUDY
An axisymmetric finite element was used here to FEM and AM are compared here by means of parametric
simulate the cylindrical GEC, and triangular finite ele- studies to assess the influence of various parameters on
ments with 15 nodal points were employed for the soil the overall behavior of GECs. In all analyses, the 2.0 m
elements. Therefore each geosynthetic element was auto- diameter GEC unit cell (aE ¼ A/AE ¼ 0.352 /1.02 ¼ 0.122),
matically defined by five-node line elements, which acts in a soft clay layer with varying thickness, was loaded by
as a membrane, and the hoop stress can be determined in a 6.0 m high embankment, which appear to be representa-
the geosynthetic. The geosynthetic encasement was mod- tive values for practical applications. The parameters
eled by geogrid elements in PLAXIS by assigning only an chosen for comparison are the settlement reduction ratio
axial stiffness, and for the axisymmetric conditions (SRR), the vertical stresses acting on the top of the soft
adopted here; hoop stresses are generated through the soil and the column, the stress concentration factor (SCF),
length proportionally with the radial deformation of the the differential settlement (DS), and the hoop tensile force
geosynthetic. Therefore the membrane effect is simulated, in the geosynthetic. (All these parameters will be defined
with additional confinement provided by the geosynthetic. later, separately.) The analyses were performed by varying
the thickness of the soft soil, Hs , in the range 5–20 m, and
the tensile stiffness of the geosynthetic, J, in the range
3. MODEL VALIDATION 0–4000 kN/m. Table 3 shows the parameters considered
Finite-element modeling and the analytical method (AM) in the parametric analyses.
are compared here so that an overall validation can be
performed. In the AM, the embankment material is simply 4.1. Soft-soil thickness and settlement response of
simulated by surface loading, but preliminary FE analyses GEC
showed that the settlement–loading curve varies slightly Figure 4 represents the variation in settlement on the top
with the embankment parameters frictional angle, cohe- of the column plotted against embankment height (He ) for
sion and elastic modulus. Therefore the embankment various thicknesses of soft soil (Hs ) by FEM and AM. As
material parameters presented in Table 2 were suitably expected, the settlement increases with embankment
chosen to allow the comparison of FEM with AM. The height and soft-soil thickness. There is good agreement
validation analyses here are performed for the case where between settlement calculated by the FEM and AM in the
the thickness of the soft soil and the geosynthetic tensile elastic strain range (small values of He ), but the difference
stiffness are Hs ¼ 10 m and J ¼ 1000 kN/m, respectively. between the two increases with both He and Hs : This is
Geosynthetics International, 2013, 20, No. 4

Downloaded by [ Universidade Federal do Rio de Janeiro] on [29/11/17]. Copyright © ICE Publishing, all rights reserved.
 256 Almeida, Hosseinpour and Riccio
Embankment height, He (m)
0 1.0 2.0 3.0 4.0 5.0 6.0
Table 3. Case considered in parametric analyses.
0
Property Values

Soft-soil thickness, Hs (m) 5, 10, 15, 20

0.20 Geosynthetic stiffness, J (kN/m) 0, 500, 1000, 2000, 4000
Settlement (m)

0.40 Embankment height, He (m)

0 1.0 2.0 3.0 4.0 5.0 6.0
0
FEM
0.60
0.20
AM

Settlement on top of column (m)

0.40
0.80
0.60

0.80
1.00 FEM
(a) 1.00
AM
Embankment height, He (m) 1.20
Hs ⫽ 5 m
0 1.0 2.0 3.0 4.0 5.0 6.0 Hs ⫽ 10 m
0 1.40
Hs ⫽ 20 m

Soft soil 1.60

100
1.80

200 Figure 4. Settlement against embankment height for encased

Vertical stress (kPa)

column (J 1000 kN/m)

300

FEM because the AM does not take the soil plasticity into
400
account, whereas the FEM result shows plastic points
AM
developed along the column and soft soil, and therefore
500 the difference between the settlements calculated by the
two methods increase with increase in He and Hs (plastic
600 strain range).
Figure 5 shows the final settlement on the top of the
Encased column
700 encased column plotted against the geosynthetic stiffness,
(b) for various thicknesses of soft soil. For both the FEM and
Embankment height, He (m) AM, it can be seen that the final settlement reduces
0 1.0 2.0 3.0 4.0 5.0 6.0
0
3.0

10
Tensile force in geosynthetic (kN/m)

2.5
Final settlement on top of column (m)

FEM Hs ⫽ 5 m
20
AM Hs ⫽ 10 m
2.0
FEM Hs ⫽ 20 m
30
AM 1.5

40
1.0

50

0.5

60
(c)
0
0 1000 2000 3000 4000
Figure 3. Validation of FEM with Raithel and Kempfert Geosynthetic stiffness, J (kN/m)
(2000): (a) settlement against embankment height;
(b) vertical stress against embankment height; (c) tensile Figure 5. Final settlement on column’s top against
force in geosynthetic against embankment height geosynthetic stiffness
Geosynthetics International, 2013, 20, No. 4

Downloaded by [ Universidade Federal do Rio de Janeiro] on [29/11/17]. Copyright © ICE Publishing, all rights reserved.
 Performance of a geosynthetic-encased column (GEC) in soft ground: numerical and analytical studies 257
80
gradually with increases in the geosynthetic stiffness. For
the cases simulated here, the influence of the stiffness on
the settlement decreases for stiffness values higher than 70

Tensile force in geosynthetic (kN/m)

FEM
J ¼ 2000 kN/m, particularly for thinner soft layers.
AM
4.2. Soft-soil thickness and settlement reduction ratio 60

(SRR)
SRR is defined here by the difference (ren  run ) in the 50
settlement between the encased and un-encased columns
(J ¼ 0), divided by the settlement of the un-encased Hs ⫽ 5 m
40
column (run ) for the identical state of geometry and
Hs ⫽ 20 m
embankment loading (SRR ¼ (ren  run )=run ).
Figure 6 shows the variation in SRR plotted against the 30
geosynthetic stiffness for two values of soft-soil thickness
(Hs ¼ 5 and 20 m). It is quite clear that for a constant Hs
the SRR increases as the geosynthetic stiffness increases, 20
0 1000 2000 3000 4000
and also that for any geosynthetic stiffness the SRR
Geosynthetic stiffness, J (kN/m)
increases when the thickness of the soft soil decreases.
Also, the efficiency of the geosynthetic in settlement Figure 7. Tensile force in geosynthetic against geosynthetic
improvement increases as the thickness of the soft soil stiffness
decreases. The reason for this is the additional confining
pressure provided by the geosynthetic, a topic explored in
A⬘ B⬘
Figure 7, which presents the maximum tensile force in the
geosynthetic plotted against geosynthetic stiffness. For
both the FEM and AM, it can be seen that the tensile Embankment
force increases gradually with geosynthetic stiffness, but
for constant stiffness the tensile force is greater as the
thickness of the soft soil reduces. Also, the AM shows a
higher tensile force than the FEM, and these differences
increase with an increase in geosynthetic stiffness.
A
B
4.3. Embankment height and differential settlements
A typical deformed shape of a GEC stabilized bed
obtained by FEM is shown in Figure 8 for an encased Encased Soft soil
column
column with geosynthetic stiffness J ¼ 4000 kN/m and
soft soil thickness Hs ¼ 10 m. It is clear that the encased
column does not settle equally with the surrounding soil,
whereas the AM assumes the same settlement value for
the top of the encased column and soft soil. Geotextile

0.9

0.8
FEM
AM Figure 8. Deformed mesh for GEC modeled by FEM
0.7
Settlement reduction ratio

(Hs 10 m, He 6 m)

0.6
This figure also represents points A and B at the top
0.5 (center) of the column and at soft soil in the middle space
between columns, respectively. Figure 9 shows the varia-
Hs ⫽ 5 m tion in settlement plotted against the embankment height
0.4
Hs ⫽ 20 m for these two points. It ca n be seen that the settlement
difference between points A and B increases for embank-
0.3
ment heights (He ) greater than 2.0 m. This difference
increases steadily, and reaches 0.1 m settlement at
0.2
0 1000 2000 3000 4000
He ¼ 6.0 m; this is quite an important difference between
Geosynthetic stiffness, J (kN/m)
the analytical and numerical methods. In practical applica-
tion, the differential settlement (DS) at the top of the
Figure 6. Settlement reduction ratio (SRR) against embankment – that is, at the ground’s surface – is more
geosynthetic stiffness important than the settlement at the level of the column.
Geosynthetics International, 2013, 20, No. 4

Downloaded by [ Universidade Federal do Rio de Janeiro] on [29/11/17]. Copyright © ICE Publishing, all rights reserved.
 258 Almeida, Hosseinpour and Riccio
Embankment height, He (m)
0 1.0 2.0 3.0 4.0 5.0 6.0
spacing s between the columns for a triangular layout (see
0 Figure 1) is s ¼ 1.05 3 dE ¼ 2.1 m. Geometry gives
1=2
s9 ¼ (3s2 )  d c =2 ¼ 0:86 m, which, substituted in Equa-
0.10 tion 4 with d c ¼ 0.70 m, gives Hcrit ¼ 2.00 m, greater than
the 1.55 m indicated in Figure 10. This difference is
0.20 expected, as pile elements in piled embankments deform
less than encased columns, and thus for the same area
Settlement (m)

0.30
ratio the critical height of the piled embankments is
expected to be higher than the for embankments on
encased columns.
0.40

Soft soil 4.4. Vertical stress ratio and stress concentration

0.50 A
Encased column The parameter vertical stress ratio is defined here as the
magnitude of vertical stress on the encased column (v,c ) or
0.60 soft soil (v,s ) divided by the vertical stress applied by the
B embankment, ˜ ¼ 120 kPa in the present case (see nota-
0.70 tion in Figure 1). Figure 11 shows the variation in vertical
stress ratio against geosynthetic stiffness for both the
Figure 9. Settlement against embankment height for top of
soft soil (B) and encased column (A) by FEM (Hs 10 m,
granular column (un-encased and encased) and the soft soil
J 4000 kN/m) with Hs ¼ 15 m. It can be seen that encasing increases the
value of the vertical stress ratio in the column, and inversely
decreases it in the surrounding soil. Also, higher geosyn-
Therefore, here, a normalized DS is defined by the thetic stiffness increases the value of the stress transferred
difference in settlement at the embankment surface (points in the encased column by as much as almost seven times
A9 and B9) divided by the embankment height: that of the embankment vertical stress for a geosynthetic
DS ¼ (rA9  rB9 )=H e ). Figure 10 shows the variation of with J ¼ 4000 kN/m. The results also show that the influ-
DS at the top of the embankment against embankment ence of the geosynthetic on vertical stress gradually reduces
height. It can be seen that, for the present case, DS occurs for stiffnesses higher than J ¼ 2000 kN/m for the case
only when the embankment height is less than analyzed.
He  1.55 m. This has been called the critical height of The parameter stress concentration factor (SCF) is
the embankment, Hcrit (McGuire et al. 2011), in relation to defined here as the ratio of axial stress experienced by the
piled embankments. Based on physical and numerical column (v,c ) to the vertical stress sustained by the soil
modeling, as well as field tests of piled embankments, (v,s ) for the same applied embankment loading:
McGuire et al. (2011) proposed the equation SCF ¼  v,c = v,s : Figure 12 shows the variation in SCF
H crit ¼ 1:15s9 þ 1:14d c (4) against geosynthetic stiffness for different values of Hs : It
can be seen that, for a granular column without encase-
where s9 is the diagonal half-span between the columns, ment (J ¼ 0), the value of the SCF does not change with
and dc is the column diameter, equal to 0.70 m in the the variation of soft soil thickness, as it is almost equal to
present study. For a unit cell diameter dE ¼ 2.0 m, the
8
10
7
Normalised differential settlement (%)

8 6
Vertical stress ratio

5
FEM
6 FEM
4
Numerical trend
AM

4 3

Encased column
y ⫽ ⫺6.0904x ⫹ 9.385
2
Soft soil
2
1

0 0
0 0.3 0.6 0.9 1.2 1.5 1.8 0 1000 2000 3000 4000
Embankment height, He (m) Geosynthetic stiffness, J (kN/m)

Figure 10. Differential settlement against embankment height Figure 11. Vertical stress ratio against geosynthetic stiffness
by FEM (Hs 10 m, J 4000 kN/m) (Hs 15 m).
Geosynthetics International, 2013, 20, No. 4

Downloaded by [ Universidade Federal do Rio de Janeiro] on [29/11/17]. Copyright © ICE Publishing, all rights reserved.
 Performance of a geosynthetic-encased column (GEC) in soft ground: numerical and analytical studies 259
50
geosynthetics, and the SCF reaches quite high and some-
Hs ⫽ 5 m what unexpected values (SCF ¼ 46 for Hs ¼ 5.0 m and
Hs ⫽ 10 m J ¼ 4000 kN/m): thus these results should be confirmed
40
Hs ⫽ 15 m experimentally.
Hs ⫽ 20 m
Stress concentration factor

4.5. Distribution of tensile force along column depth

30
Figure 13 shows the distribution of tensile force in the
geosynthetic with depth. These graphs are presented for
granular columns encased by different geosynthetic stiff-
20
ness for both the FEM and AM. Variations in the tensile
force by the FEM follow the same pattern of the lateral
deformation undergone by the stone column, and the
10 magnitude of the tensile force varies from 36 kN/m to
72 kN/m for different stiffness values. Gniel and Bouazza
(2009) also reported that hoop tensile force in geosyn-
0 thetics follows a variable pattern with depth, whereas
0 1000 2000 3000 4000
the AM assumes a constant tensile force value along the
Geosynthetic stiffness, J (kN/m)
depth, as shown in Figure 14. It may be observed that the
Figure 12. Stress concentration factor (SCF) against tensile forces become less variable as the geosynthetic
geosynthetic stiffness by FEM modulus increases.
This variable pattern between geotextile tensile forces
and depth is investigated in Figure 14. Figure 14a presents
the tensile forces and shear stress distribution for a short
4 for all values of Hs , and this value is quite typical for encased column with Hs /dc ¼ 2.0 (dc ¼ 0.70 m is the col-
conventional granular columns (Mitchell and Huber umn diameter) and geosynthetic stiffness J ¼ 1000 kN/m
1985). loaded by a 6.0 m high embankment. The GEC shear
For an encased column, though, as the geosynthetic stresses in Figure 14a can be associated with an ordinary
stiffness varies, the SCF increases proportionally, and the triaxial specimen with height twice the diameter. In fact, the
magnitude of the increase depends on the geosynthetic pattern of shear stress distribution is in accordance with the
stiffness and also on the thickness of the soft soil. For well-known behavior of ‘dead’ and ‘active or dilating’ zones
example, for an encased column modeled in soft soil with in triaxial specimens, as shown in Figure 14b, a phenomen-
Hs ¼ 20 m, the stress concentration of the un-encased on discussed in detail by Rowe and Barden (1964). Accord-
granular column increases from 4 to 26 when it is encased ing to Figure 14a, the inclination of the ‘active zone’ in
by geosynthetic with J ¼ 4000 kN/m. The results in Figure relation to the horizontal is equal to Æ ¼ 63.48, which is
12 also show that the stress concentration factor increases almost equal to the slip surface predicted by 45 +
as the thickness of the soft soil decreases for identical c /2 ¼ 648, for c ¼ 38.08 (Tables 1 and 2).

Tensile force (kN/m) Tensile force (kN/m) Tensile force (kN/m) Tensile force (kN/m)
0 10 20 30 0 20 40 0 20 40 60 0 20 40 60 80
0 0 0 0

1 1 1 1

2 2 2 2

3 3 3 3

4 4 4 4
Depth (m)
Depth (m)

Depth (m)

Depth (m)

5 5 5 5

6 6 6 6

7 7 7 7

8 8 8 8

9 9 9 9

10 10 10 10
(a) (b) (c) (d)

Figure 13. Tensile force distribution against depth for different geosynthetic stiffness (solid line, FEM; dashed line, AM)
Geosynthetics International, 2013, 20, No. 4

Downloaded by [ Universidade Federal do Rio de Janeiro] on [29/11/17]. Copyright © ICE Publishing, all rights reserved.
 260 Almeida, Hosseinpour and Riccio

Dilating zone
dc Dead zone
Dead zone

L ⫽ 2D

ne
α

zo
Dilating

ing
He
zone

la t
Di
Dead zone
α

(a) (b) (c)

Figure 14. Tensile force in geosynthetic and shear zones in encased column: (a) Hs /dc 2.0; b) Hs /dc 7.14; (c) slip surface in
triaxial specimen (Rowe and Barden 1964)

From Figure 14a it can also be seen that the maximum 5. The parameter DS was defined to assess the
tensile force occurs at the height where the shear stress is differential settlement at the ground surface. The
at a maximum, and the minimum is associated with the result of the FEM showed that the DS at the ground
‘dead’ zone region. Figure 14c presents results for a surface decreases as the embankment height in-
longer column with Hs /dc ¼ 7.14, and similar patterns of creases, and becomes zero at a given embankment
dead zones, with smaller tensile stresses, and dilating height corresponding to full arching. This is the
zones, with larger tensile stresses, are observed. critical embankment height, which is a function of
the span between the columns and the column
diameters.
5. CONCLUSIONS 6. The value of the critical embankment obtained here
In this paper, a parametric study using both the FEM and is slightly smaller than that found for piled
AM for a granular column and a GEC loaded by an embankments with reinforced platforms.
embankment was investigated. The unit cell modeling 7. FEA was also conducted in the drained condition,
approach was used to compare results from the FEM and because PLAXIS does not take into account the
AM. Based on the findings obtained for settlements, permeability characteristics of the geogrid element,
vertical stresses in the soil and column and the hoop and therefore the significant role of the geosynthetic
tensile force in the geosynthetic, the following conclusions in speeding up the consolidation is neglected. Also,
may be presented. the influence of column installation was not consid-
ered in the current study, which would affect the
1. The result of a 2D unit cell of GEC showed good lateral deformation of the soil beneath the embank-
agreement in general with the AM. However, unlike ment. In comparison with the 3D analysis, the 2D
the AM, the FEM indicated that settlement is not the unit was not used for modeling of the full
same for soft soil and the encased column. embankment over a group of GECs at which the
2. Parametric analyses showed that the enhancement of plane-strain condition should be used. Because the
the geosynthetic encasement to reduce settlement plane-strain condition is used, the membrane effect
improves with decreasing soft-soil thickness. Further- of the geosynthetic casing will be neglected.
more, it was observed that the influence of the
geosynthetic on settlement gradually reduces for
NOTATION
geosynthetic stiffness higher than J ¼ 2000 kN/m for
the cases analyzed here. Basic SI units are shown in parentheses.
3. Tensile force obtained by the FEM compared well
with the AM. However, whereas the AM determines Ac cross-sectional area of encased column (m2 )
a constant value along the depth, tensile forces in the AE cross-sectional area of unit cell (m2 )
FEM vary with depth, showing dilating zones in the aE area replacement ratio (dimensionless)
GEC associated with maximum tensile forces. cs soft soil cohesion (Pa)
4. The results showed that for an un-encased column c9s effective soft soil cohesion (Pa)
(J ¼ 0) the SCF does not change with soft-soil DS differential settlement (dimensionless)
thickness, whereas for an encased column the stress dc diameter of encased column ¼ 2rc (m)
concentration factor increases with an increase in dE diameter of unit cell ¼ 2rE (m)
geosynthetic stiffness, and also decreases in the E9 drained Young’s modulus (Pa)
thickness of the soft soil. The values of SCF obtained Eoed,s oedometer (constrained) modulus of soft
here are quite high, and deserve further inspection. soil (Pa)
Geosynthetics International, 2013, 20, No. 4

Downloaded by [ Universidade Federal do Rio de Janeiro] on [29/11/17]. Copyright © ICE Publishing, all rights reserved.
 Performance of a geosynthetic-encased column (GEC) in soft ground: numerical and analytical studies 261

FR tensile force in geotextile (N/m) Research and Development, Report No. FHWA/RD-83/026, Federal
Hcrit critical height of embankment (m) Highway Administration, Washington, DC, USA.
Brinkgreve, R. B. J. & Vermeer, P. A. (2002). PLAXIS: Finite Element
He embankment height (m)
Code for Soil and Rock Analysis, Version 8, Balkema, Rotterdam,
Hs thickness of soft soil (m) The Netherlands.
J geotextile stiffness (N/m) Brokemper, D., Sobolewski, J., Alexiew, D. & Brok, C. (2006). Design
K0,s lateral earth pressure coefficient of soft soil and construction of geotextile encased columns supporting geogrid
(dimensionless) reinforced landscape embankments: Bastions Vijfwal Houten in the
m power of stress dependence of stiffness Netherlands. Proceedings of the 8th International Conference on
Geosynthetics, Millpress, Amsterdam, The Netherlands, pp. 1681–
(dimensionless) 1684.
pref reference mean stress (Pa) Castro, J. & Sagaseta, C. (2011). Deformation and consolidation around
rc encased column radius (m) encased stone columns. Geotextiles and Geomembranes, 29, No. 3,
rE unit cell radius (m) 268–276.
rgeo radius of geotextile encasement (m) EBGEO (2010). Recommendations for Design and Analysis of Earth
Structures using Geosynthetic Reinforcements, German Geotechni-
˜rgeo radial displacement of geotextile (m) cal Society (DGGT), Berlin, Germany.
SCF stress concentration factor (dimensionless) Gniel, J. & Bouazza, A. (2009). Improvement of soft soils using geogrid
SRR settlement reduction ratio (dimensionless) encased stone columns. Geotextiles and Geomembranes, 27, No. 3,
s axial distance between columns (m) 167–175.
Gray, D. H. & Ohashi, H. (1983). Mechanism of fiber reinforcement in
s9 diagonal halfispan between columns (m)
sand. Geotechnical Engineering, 109, No. 3, 335–353.
ª9 submerged unit weight (N/m3 ) Greenwood, D. A. (1970). Mechanical improvement of soils below
ªc unit weight of encased column (N/m3 ) ground surface. Proceedings of the Ground Engineering Con-
ª9c submerged unit weight of encased column ference, Institution of Civil Engineers, London, pp. 11–22.
(N/m3 ) Hong, Y.-S. (2012). Performance of encased granular columns consider-
ªs soft soil unit weight (N/m3 ) ing shear-induced volumetric dilation of the fill material.
Geosynthetics International, 19, No. 6, 438–452.
ª9s submerged unit weight of soft soil (N/m3 ) Hosseinpour, I., Mirmoradi, S. H., Barari, A. & Omidvar, M. (2010).
ªsat saturated unit weight (N/m3 ) Numerical evaluation of sample size effect on the stress–strain
9 effective Poisson’s ratio (dimensionless) behavior of geotextile-reinforced sand. Journal of Zhejiang
r settlement on top of embankment (m) University – Science A, 11, No. 8, 555–562.
ren settlement of encased column (m) Keykhosropur, L., Soroush, A. & Imam, R. (2012). 3D numerical
analyses of geosynthetic encased stone columns. Geotextiles and
run settlement of un-encased column (m) Geomembranes, 35, 61–68.
h,geo horizontal stress acting on geotextile (Pa) Khabbazian, M., Kaliakin, V. N. & Meehan, C. L. (2010). Numerical
h,diff horizontal stress difference (Pa) study of the effect of geosynthetic encasement on the behaviour of
v,c vertical stress on encased column (Pa) granular columns. Geosynthetics International, 17, No. 3, 132–143.
v,s vertical stress on surrounding soil (Pa) Lee, D. Y., Song, A. R., Kim, S. B. & Yoo, C. (2007). Short-and long-
term load carrying capacity of geogrid-encased stone column: A
˜h,c horizontal stress in column (Pa) numerical investigation. Journal of Geotechnical and Geoenviron-
˜h,s horizontal stress in soil (Pa) mental Engineering, ASCE, 136, No. 1, 129–139.
9 effective friction angle (degrees) Malarvizhi, S. N. & Ilamparuthi, K. (2007). Comparative study on the
ł angle of dilatancy (degrees) behavior of encased stone column and conventional stone column.
Æ Slope of shear zone in triaxial specimen (degree) Soils and Foundations, 47, No. 5, 873–885.
McGuire, M. P., Sloan, J., Collin J. & Filz, G. M. (2011) Critical height
of column-supported embankments from bench-scale and filled
ABBREVIATIONS scale tests. International Society for Soil Mechanics and
Geotechnical Engineering Technical Committee 211, International
Symposium on Ground Improvement, Brussels, Belgium.
AM analytical model Mitchell, J. K. & Huber, T. R. (1985). Performance of stone column
FEA finite-element analysis foundation. Geotechnical Engineering, 111, No. 2, 205–223.
FEM finite-element model Murugesan, S. & Rajagopal, K. (2006). Geosynthetic-encased stone
GEC geosynthetic-encased stone column column: Numerical evaluation. Geotextiles and Geomembranes, 24,
No. 6, 349–358.
Murugesan, S. & Rajagopal, K. (2007). Model tests on geosynthetic
REFERENCES encased stone columns. Geosynthetics International, 14, No. 6,
346–354.
Alexiew, D., Brokemper, D. & Lothspeich, S. (2005). Geotextile encased Murugesan, S. & Rajagopal, K. (2010). Studies on the behavior of single
columns (GEC): load capacity, geotextile selection and pre-design and group of geosynthetic encased stone column. Journal of
graphs. GeoFrontiers 2005: Proceedings of Geo-Frontiers Congress Geotechnical and Geoenvironmental Engineering, ASCE, 136, No.
2005, Austin, TX, USA, Geotechnical Special Publication 131, 1, 129–139.
Reston, VA, USA, pp. 497–510. Pulko, B., Majes, B. & Logar, J. (2011). Geosynthetic-encased stone
Ali, K., Shahu, J. T. & Sharma, K. G. (2012). Model tests on columns: analytical calculation model. Geotextiles and Geomem-
geosynthetic-reinforced stone columns: a comparative study. branes, 29, No. 11, 29–39.
Geosynthetics International, 19, No. 4, 292–305. Raithel, M. & Kempfert, H. G. (2000). Calculation models for dam
Araujo, G. L. S., Palmeira, E. M. & Cunha, R. P. (2009). Behaviour of foundations with geotextile coated sand column. Proceedings of
geosynthetic-encased granular columns in porous collapsible soil. GeoEngineering 2000, Melbourne, Australia, Technomic Publish-
Geosynthetics International, 16, No. 6, 433–451. ing, Lancaster, PA, USA, p. 347.
Barksdale, R. D. & Bachus, R. C. (1983). Design and Construction of Raithel, M., Kempfert, H. G. & Kirchner, A. (2002). Geotextile encased
Stone Columns, Office of Engineering and Highway Operations columns (GEC) for foundation of a dike on very soft soils.

Geosynthetics International, 2013, 20, No. 4

Downloaded by [ Universidade Federal do Rio de Janeiro] on [29/11/17]. Copyright © ICE Publishing, all rights reserved.
 262 Almeida, Hosseinpour and Riccio

Proceeding of the 7th International Conference on Geosynthetics, Geotechnical and Geoenvironmental Engineering, ASCE, 136, No.
Balkema, The Netherlands, pp. 1025–1028. 8, 129–139.
Rowe, P. W. & Barden, L. (1964). Importance of free ends in triaxial Yoo, C. & Kim, S. B. (2009). Numerical modeling of geosynthetic encased
testing. Journal of the Soil Mechanics and Foundation Division, stone columns. Geosynthetics International, 16, No. 3, 116–126.
ASCE, 90, No. SM1, 1–27. Yoo, C. & Lee, D. (2012). Performance of geogrid-encased stone columns
Reugger, R. (1986). Geotextile reinforced soil structures. Proceedings of in soft ground: full-scale load tests. Geosynthetics International, 19,
the 3rd International Conference on Geotextiles, Vienna, Austria, No. 6, 480–490.
pp. 453–458. Zhang, Y., Li, T. & Wang, Y. (2011). Theoretical elastic solution for
Yoo, C. (2010). Performance of geosynthetic-encased stone columns in foundations improved by geosynthetic-encased columns. Geosyn-
embankment construction: numerical investigation. Journal of thetics International, 18, No. 3, 12–20.

The Editor welcomes discussion on all papers published in Geosynthetics International. Please email your contribution to
[email protected] by 15 February 2014.

Geosynthetics International, 2013, 20, No. 4

Downloaded by [ Universidade Federal do Rio de Janeiro] on [29/11/17]. Copyright © ICE Publishing, all rights reserved.
View publication stats