Computers and Geotechnics 87 (2017) 178–187

Contents lists available at ScienceDirect

Computers and Geotechnics

journal homepage: www.elsevier.com/locate/compgeo

Research Paper

Performance of geosynthetic-reinforced granular piles in soft clays:

Model tests and numerical analysis
Murtaza Hasan ⇑, N.K. Samadhiya
Department of Civil Engineering, Indian Institute of Technology Roorkee, Roorkee 247667, Uttarakhand, India

a r t i c l e i n f o a b s t r a c t

Article history: The laboratory model tests and numerical analyses have been performed on reinforced granular piles
Received 31 July 2016 installed in very soft clay. The granular piles were reinforced with geosynthetic in the form of vertical
Received in revised form 9 December 2016 encasement, horizontal strips and combined vertical-horizontal reinforcement. The short term-
Accepted 21 February 2017
displacement control model tests were carried out either only a granular pile loaded or with an entire
area loaded. The laboratory results in the form of vertical load intensity-settlement behaviour were com-
pared with that obtained from FEM software, PLAXIS 3D. The results indicated significant improvement in
Keywords:
ultimate load intensity and stiffness of treated ground due to inclusion of geosynthetic.
Stone columns
Geotextile
Ó 2017 Elsevier Ltd. All rights reserved.
Geogrid
Plaxis 3D
Soft clays
Numerical analysis

1. Introduction carrying capacity of the granular piles due to encasement. The

Granular piles (also known as stone columns) have been com-
monly installed in grounds to improve the load bearing capacity,
to reduce settlements and to increase the rate of consolidation.
The applications of granular piles include the support of embank-
ments, liquid storage tanks, raft foundations and other low rise
structures [1–5]. The load carrying capacity of the granular piles
depends on the lateral confining pressure from the surrounding
soils. In very soft clays, granular piles do not achieve significant
load carrying capacity due to low lateral confinement. Therefore
granular piles of additional confinement are needed for the better
performance. In recent years, geosynthetic reinforced granular
piles have been successfully adopted in very soft soils throughout
the world. Vertical encased granular piles have several advantages
like increased stiffness of pile by mobilization of hoop stress in the
reinforced material, preventing the loss of stones into the sur-
rounding soft clay, preserving the drainage and frictional proper-
ties of the stone aggregates. Van Impe and Silence [6] were
probably the first who thought of encasing the granular piles by
geosynthetic. The laboratory and small scale model tests have con-
ducted on vertical encased granular piles by various researchers
[7–15]. Murugesan and Rajagopal [8] found improvement in load
⇑ Corresponding author.
E-mail addresses: murtazadce@gmail.com (M. Hasan), nksamfce@iitr.ac.in
(N.K. Samadhiya).
effect of encasement was found to decrease with increase in the
diameter of the granular pile. Numerical analyses have performed
by various researchers on geosynthetic encased granular piles
[7,16–19]. Many authors [20–22] also performed field scale load
tests on vertical encased granular piles. The reinforcement in the
form of horizontal strips, placed at regular spacing in granular piles
increase load carrying capacity of improved ground. Granular piles
reinforced with horizontal strips control bulging by mobilising fric-
tional resistance between strips and stone aggregates [11,23–25].
Sharma et al. [23] investigated the effect of the number of geogrid
layers and the spacing between them on the load carrying capacity
of granular piles.
The reinforcement effect in form of either encasement or hori-
zontal strips has been very well establish through various labora-
tory studies, field applications and numerical analyses. Generally,
only one type of reinforcement (encasement or horizontal strips)
has been dealt with. Enough literature is not available on assess-
ment of improvement in the load carrying capacity of the granular
piles reinforced with bothy encasement and horizontal strips.
Therefore in the present study, combined encased-horizontal
stripped granular piles have been installed in very soft clay to
investigate the qualitative and quantitative improvement in the
ultimate load intensity of granular piles, and stiffness of improved
ground. The laboratory model tests were conducted on unrein-
forced, vertical encased, with horizontal strips and combined
vertical-horizontal reinforced granular piles. Numerical analyses

http://dx.doi.org/10.1016/j.compgeo.2017.02.016
0266-352X/Ó 2017 Elsevier Ltd. All rights reserved.
 M. Hasan, N.K. Samadhiya / Computers and Geotechnics 87 (2017) 178–187
have also been carried out on granular piles by using finite element
software; PLAXIS 3D. The parameters included in this study are
shear strength of clay, reinforcement, encasement stiffness and
length of granular piles. The ultimate load capacity of granular
piles has been found to increase with increase in the depth of
encasement [8,12]. In field studies, various authors [20–22] also
provided geotextile encasement throughout the length of granular
piles. Therefore in the present study, encasement and horizontal
strips up to full length of granular piles have been adopted.
2. Experimental programme
The outline of experimental programme is given in Table 1. Unit
cell idealisation, by assuming piles in a triangular pattern, was
adopted to simplify the design of the apparatus needed to assess
the behaviour of an interior granular pile in a large group of piles.
Various researchers [3,8,26] have used unit cell concept in their
experimental investigations.
The laboratory model tests were carried out on 75 mm diameter
single floating and end bearing granular piles. The length of gran-
ular piles was kept as 5d (375 mm) in case of floating piles and
7d (525 mm) for end bearing piles, where d is the diameter of
the pile. The granular piles were installed in very soft clay bed in
a 200 mm diameter and 525 mm high cylindrical tank. Undrained
shear strength (cu) of the very soft clay was kept close to 5 kPa
throughout the experimental work. Geotextile and geogrid were
used for vertical encasement and horizontal strips respectively.
Circular strips of geogrid having diameter 10 mm less than the
granular piles were placed over the entire length of granular piles.
These strips were placed at three different centre to centre spacing
(S) of 25 mm (d/3), 50 mm (2d/3) and 70 mm (d). First geogrid
strip in each case was placed 25 mm below the loading plate.
2.1. Modelling considerations
The single granular pile behaviour with unit cell concept simu-
lates the field behaviour for an interior pile when large number of
piles is simultaneously loaded. In the present study, the dimen-
sions are reduced by an appropriate scale factor to simulate the
behaviour of granular piles installed in the field. Following param-
eters were considered to be scaled down: (i) ratio of granular pile
diameter to unit cell diameter, (ii) ratio of length to diameter of
granular pile, (iii) ratio of pile diameter to aggregate size and (iv)
tensile strength of geosynthetic materials. Granular pile spacing
broadly varies in the range of 2–3 times the diameter of the pile.
Since the equilateral triangular pattern gives the densest packing,
Table 1
Details of experimental programme.
Test description Reinforcing material type
Type of piles Reinforcement type
– Clay bed – –
Floating Unreinforced – –
End bearing
Floating Vertical encased Geotextile
End bearing
End bearing Horizontal strips Geogrid
End bearing Vertical encased and horizontal strips Geogrid and geotextile
Total tests
179
hence the granular piles have been assumed to be installed in
the same pattern as well as the in same spacing. The area replace-
ment ratio, Ar (ratio of the area of the granular pile to the total area
within the unit cell) and corresponding centre to centre spacing
were kept 14% and 2.5d respectively. In the field, the ratio of length
to diameter (l/d) of granular pile is varied from 4.5 to 20 whereas in
the present study the ratio l/d is kept from 5 to 7, where l is length
of pile. In practice, granular piles are constructed in typical diame-
ters (d) varying between 0.6 m in case of stiff clays to 1.1 m in very
soft clays and lengths ranging from 5 to 20 m [27]. Well graded
stones aggregates of size (k) 2–75 mm are used, so that the ratio
d/k lies typically in the range of 8 and 550 [28]. The pile diameter
used in the model test was 75 mm. The particle size of crushed
stones was kept between 2 mm and 6.3 mm, so that the ratio of
d/k in model test lies between 8 and 45. In practice, the tensile
strength of geosynthetic materials used in granular piles is kept
up to 400 kN/m and stiffness is varied from 1000 to 4000 kN/m
[12]. Tensile strength of geosynthetic materials were also reduced
as per scaling laws proposed by Iai [29], the relationship between
prototype-scale reinforcement stiffness (JP) and model-scale stiff-
ness (Jm) can be calculated as JP = Jmk2, where 1/k is the model scale.
In the present study, 1/k is equal to 1/10 therefore stiffness was
kept typically in the range of 10–40 kN/m. For laboratory model
tests, various researchers [9–12] used tensile strength of geosyn-
thetic material in the range of 1.5–20 kN/m. Geotextile and geogrid
having tensile strength of 4.4 kN/m and 7.96 kN/m respectively
were used in the present laboratory investigation. Reinforced gran-
ular piles are commonly used in soft clayey soils (undrained shear
strength less than 7 kPa) to control lateral confinement [11,12,19].
Undrained shear strength of clay in the present study has been
kept close to 5 kPa.
2.2. Materials properties
The dried clay, classified as CI as per IS: 1498:2000 [30] was
converted into fine powder by grinding and stored in air dried
room. The physical properties of clay are presented in Table 2.
The crushed granite aggregates in sizes ranging 2–6.3 mm were
used to construct granular piles. The maximum and minimum dry
unit weights of the aggregate are 15.04 kN/m3 and 13.41 kN/m3
respectively. The relative density of stone aggregates in granular
piles was considered 70% to ensure negligible bulging during con-
struction of pile. The dry unit weight and angle of internal friction
of stone aggregates at 70% relative density are 14.51 kN/m3 and 42°
respectively. The particle size distribution curves for both clay and
crushed stones are shown in Fig. 1
Strip spacing (mm) Loading type Tests
Pile alone Entire area
U U 2
U U 2
– U U 2
– U U 2
– U U 2
25 U – 1
50 U – 1
70 U – 1
25 U U 2
50 U U 2
70 U U 2
19
180 M. Hasan, N.K. Samadhiya / Computers and Geotechnics 87 (2017) 178–187

Table 2
Physical Properties of clay.

Properties Value
Specific gravity 2.73
Optimum moisture content (%) 17.56
Maximum dry unit weight (kN/m3) 17.22
Liquid limit, wl (%) 48
Plastic limit, wP (%) 18
Plasticity index (%) 30
Dry unit weight at 34% water content (kN/m3) 13.85
Undrained shear strength at 34% water content (kPa) 5
Classification CI

Fig. 2. Tensile stress strain behaviour of geosynthetic materials.

Fig. 1. Particle size distribution curves for clay and stone aggregates.

Fig. 3. Variation of undrained shear strength with water content.

Nonwoven geotextile and biaxial geogrid were used as encase-
ment and horizontal strips respectively. The properties of geosyn-
thetic materials used are listed in Table 3. The ultimate tensile
strength of materials was determined from standard wide-width
tension tests (ASTM D4595) [31]. The tensile stress-strain plots
for geosynthetic are shown in Fig. 2. Geotextile was patched by
special glue to form a hollow tube with an overlapping width
20 mm to be used as for encasement of the granular pile. Ultimate
seam strength of glued geotextile was also determined with spec-
imen having a horizontal seam at mid length. The ultimate
strength observed from tensile tests on glued geotextile was com-
parable to that of unpatched geotextile.
2.3. Preparation of clay bed
The required quantity of clay corresponding to dry unit weight
(13.85 kN/m3) was mixed thoroughly with water (34%) in a large
steel container to ensure uniformity throughout clay. Grease in
thin coat was applied on the inner surface of cylindrical tank to
create frictionless boundary between clay and inner wall. The tank
was filled in layers (30 mm height) and compacted properly. After
completing clay bed (525 mm in height), tank was covered by wet
jute bag and left for two days to gain uniformity. Vane shear tests
were conducted at the centre of clay bed to verify the shear
strength before installation of granular piles. All vane shear tests
had reasonable results, equivalent to 5 kPa. The same procedure
was adopted for all the experiments. Fresh clay was used in each
laboratory test for better control on moisture content.

In the present study, clay bed of 525 mm height and having

undrained shear strength close to 5 kPa was prepared for all exper- 2.4. Construction of granular pile
iments. The laboratory unconfined compression tests were con-
ducted to find undrained shear strength corresponding to Granular piles of diameter 75 mm was constructed by using
different moisture content. Perusal of Fig. 3 shows that water con- replacement method in all laboratory model tests. Generally, dis-
tent and dry unit weight corresponding to 5 kPa shear strength placement granular piles are popular in soft clays. The displace-
were found to be 34%, and 13.85 kN/m3 respectively. ment technique is difficult to implement in a small scale
laboratory model tests whereas the replacement technique is
known to produce the granular piles of excellent consistency. A
Table 3 thin seamless steel pipe (outer diameter equivalent to granular pile
Properties of geosynthetic materials. diameter) with the help of auger guide was inserted smoothly in
Properties Geotextile Geogrid the centre of clay bed up to the desired height of granular pile as
Type Nonwoven Biaxial
shown in Fig. 4a. Oil was applied on both outer and inner surface
Ultimate tensile strength (kN/m) 4.41 7.96 of steel pipe to allow easy insertion as well as withdrawal and to
Strain at peak load (%) 54.62 20.21 avoid sticking of clay to pipe. A 73 mm diameter helical steel auger
Axial stiffness (kN/m) 8.07 38.01 was used to scoop out clay within the steel pipe. To avoid suction
Ultimate seam strength (kN/m) 3.94 –
between pipe and auger, Approximately 20 mm height of clay was
Thickness (mm) 2 1.5
scooped in a single stroke. For encased granular pile, a 72 mm
 M. Hasan, N.K. Samadhiya / Computers and Geotechnics 87 (2017) 178–187
Fig. 4. Construction of granular pile (a) insertion of steel pipe, (b) excavated hole, (c) pile and clay bed.
diameter circular wooden rod was used for vertical placing of geo-
textile in excavated hole. Then crushed stone aggregates in prede-
termined quantity were charged into hole in 10 layers to maintain
dry unit weight (14.51 kN/m3). Each layer was light compacted
(2.5 kg circular steel tamper, 5 blows with 150 mm drop) to ensure
that no lateral bulging occurs during the construction of granular
pile. In case of granular pile with horizontal strips, the aggregates
in predetermined quantity between two geogrid layers were
poured and then compacted within hole before placing next circu-
lar geogrid. Fig. 4b–c shows clay bed with excavated hole and after
formation of granular pile. Fresh aggregates were used for each
laboratory test for better results.
2.5. Test set up and procedure
Fig. 5 shows a schematic view of test setup for laboratory model
tests. Vertical load was applied either over the entire cylindrical
tank area or only over the area of granular pile. The load was
Fig. 5. Schematic view of test setup.
181
applied at a constant displacement rate of 1.2 mm/min. A proving
ring was used to measure the applied load. Dial gauge was used to
measure the settlement of loading plate. In case of only granular
pile loading, the load was applied (up to 35 mm settlement)
through a loading plate of 15 mm thickness and diameter equal
to granular pile. However in case of entire area loading, the load
was applied (up to 30 mm settlement) through a steel plate of
diameter 10 mm less than the inside diameter of the test tank.
Loading period is kept short to ensure undrained loading condition
which simulates loading during construction. Vertical load inten-
sity settlement behaviour was studied for the assessment of per-
formance of the granular piles. After completion of each test,
aggregates were scoop out carefully, slurry of plaster of Paris was
poured into the hole thus obtained and allowed to set for 24 h. Sur-
rounding clay was removed carefully and deformed shape of gran-
ular pile was obtained and then painted in white colour for better
visibility of bulging.
3. Finite element analysis
FEM analysis was carried out by finite element software, PLAXIS
3D. The vertical load intensity settlement plots were compared
with these obtained by laboratory investigations. Plaxis numerical
model was validated from result of the laboratory model load test
on granular piles, reported by Ambily and Gandhi [26]. Ambily and
Gandhi [26] assumed unit cell concept with short term loading
condition, so numerical model is close to present study. Ambily
and Gandhi conducted the tests on 100 mm diameter stone col-
umns surrounded by soft clay for a height of 450 mm in 210 mm
diameter cylindrical tank. A model was generated in PLAXIS 3D
and analysed using Mohr-Coulomb failure criterion. The material
properties used for numerical model are given in Table 4. The mesh
generated and total displacements in the model are shown in
Fig. 6. The comparison of experimental and PLAXIS 3D results is
presented in Fig. 7. The results are in reasonably good agreement.
In the present study, the linear elastic perfectly plastic, Mohr
Coulomb model has been adopted for clay and stone aggregates.
This model has also adopted for similar studies by various
researchers [12,17,19,26,32]. The input parameters (E, cu, /, cd,
w) required for this model can be easily determined from the rel-
evant laboratory tests. In the present study, the cylindrical unit cell
of equivalent diameter was simulated as square unit cell due to
geometry limitation in PLAXIS. Ng and Tan [33] investigated that
the results in Plaxis 2d and 3d models are similar especially for
182 M. Hasan, N.K. Samadhiya / Computers and Geotechnics 87 (2017) 178–187

Table 4
Material properties for PLAXIS 3D.

Parameters Ambily and Gandhi [26] Present study

Clay Stone aggregates Clay Stone aggregates
Young’s modulus, E (kPa) 5500 55,000 420 42,500
Undrained shear strength, cu (kPa) 30 – As per Table 5 –
Angle of internal friction, / (°) 0 43 0 42
Dilation angle, w (°) – 10 – 10
Poisson’s ratio, l 0.42 0.3 0.48 0.3
Dry unit weight (kN/m3) 15.56 16.62 13.85 14.51
Bulk unit weight (kN/m3) 19.45 – 18.58 –

Fig. 6. Mesh generation and total displacements in model, Ambly and Gandhi [26].

Fig. 8. Vertical encased and horizontal stripped granular piles models.

loading plate. Geosynthetic was modelled as elastic material.

Deformation of the granular piles is mainly due to radial bulging,
and no significant shear is possible on the interface between gran-
ular pile and clay [4,26]. Therefore an interface element has not
been used in the present study. The material parameters used in
model were determined from relevant laboratory tests and are
given in Table. 4. Modulus of elasticity of soft clay was determined
by consolidation test corresponding to a pressure range of 100–
200 kPa as reported by Ambily and Gandhi [26]. The Poisson’s ratio
was taken as per typical values suggested by Bowles [34]. The gen-
erated meshes used for encased and horizontal stripped end bear-
ing granular pile models are presented in Fig. 8.
Fig. 7. PLAXIS validation through experimental result, Ambly and Gandhi [26].
4. Results and discussion

load-settlement behaviour and failure mechanism. Square unit cell
was also simulated by various researchers [7,12,26,32]. The bottom
boundary of the unit cell model is restricted to move in all the
three directions whereas the vertical boundaries can move only
in the vertical directions. Geogrid strips were created by using
function of polycurve with circular shape converting into surface
and then applying geogrid in PLAXIS model. The vertical load
was applied either on the granular pile or entire unit cell area in
the form of prescribed displacement assuming rigid behaviour of
4.1. Comparison of laboratory model tests and FEM when granular pile
alone is loaded
The ultimate load intensity of granular piles has been deter-
mined from laboratory model tests by applying load on only gran-
ular pile. The experimental and FEM analysis results, in terms of
vertical load intensity settlement relationship for clay bed, unrein-
forced, encased floating as well as end bearing granular pile are
presented in Fig. 9a-b.
 M. Hasan, N.K. Samadhiya / Computers and Geotechnics 87 (2017) 178–187 183

Fig. 9. Vertical load intensity settlement behaviour of granular piles; (a) unrein-
forced, (b) vertical encased.
As may be seen from the figure, the results of laboratory model
tests and numerical analysis are in close agreement. The results
have also been summarised in Table 5. The unreinforced granular
piles and reinforced floating granular piles exhibit clear ultimate
load, whereas reinforced end bearing granular piles do not show
failure, even at 35 mm settlement levels. There seems to be mar-
ginal increase in the vertical load intensity of end granular piles
after 30 mm settlement. Therefore for the comparison purpose,
Fig. 10. Variation of the ultimate load intensity with; (a) undrained shear strength,
(b) axial stiffness of geotextile.
load intensity corresponding to 30 mm settlement was assumed
to be ultimate load intensity. The increment in ultimate load inten-
sity of granular pile surrounding the very soft clay was expressed
in the term of ultimate load intensity ratio (a). The ultimate load
intensity ratio (a) is derived as the ratio of ultimate load intensity
of granular piles to that of clay bed. It was found in the range of

Table 5
Results of laboratory model tests and FEM analysis.

Test description Reinforcing material type Strip spacing (mm) Pile alone loaded Entire area loaded
Type of piles Reinforcement type cu (kPa Ultimate load cu (kPa Stiffness (kPa)
intensity (kPa)
Exp. FEM Exp. FEM
– Clay bed – – 5.12 33.85 33.64 5.18 48.09 46.32
Floating Unreinforced – – 5.01 96.65 94.87 5.11 75.21 76.18
End bearing – 5.09 114.58 109.08 5.16 96.41 90.38
Floating Vertical encased Geotextile – 5.21 174.16 178.06 5.34 81.12 85.01
End bearing – 5.41 221.97 217.90 5.29 117.95 110.87
End bearing Horizontal strips Geogrid 25 5.34 218.62 221.72 – – –
50 5.25 202.25 194.39 – – –
70 5.10 148.25 140.83 – – –
End bearing Vertical encased and horizontal strips Geogrid and geotextile 25 5.38 383.57 372.16 5.32 137.20 133.79
50 5.19 298.65 295.70 5.18 122.45 124.17
70 5.12 266.77 264.42 5.22 116.18 117.40
184 M. Hasan, N.K. Samadhiya / Computers and Geotechnics 87 (2017) 178–187

2.66–3.19 (exp.) and 2.79–3.25 (FEM) for unreinforced granular ging. The ultimate load intensity of end bearing granular piles
pile. While for vertical encased granular piles, a is varying in the increased almost linearly due to increase in the encasement
range of 4.91–6.26 (exp.) and 5.31–6.49 (FEM).The encased granu- stiffness.
lar pile offers higher resistant to surrounding clay by mobilising The experimental and FEM results of end bearing granular piles
hoop stress in geotextile and thereby improving the lateral reinforced with horizontal stripes in three different spacing are
confinement. presented in Fig. 11. The ultimate load intensity for 25 mm spaced
The influence of shear strength of the soft clay and stiffness of horizontal striped granular pile has been observed to be very close
encasement on the behaviour of encased floating as well as end to that of vertical encased end bearing granular pile. The ultimate
bearing granular piles was studied by numerical analysis. The load intensity of granular piles increased with the reduction in ver-
shear strength of clay bed was varied between 2 kPa and 15 kPa.
The axial stiffness of geotextile was kept as 7.32 kN/m. The varia-
tion of ultimate load intensity with shear strength is shown in
Fig. 10a. The ultimate load intensity for granular pile increased
with increase in the shear strength of clay, which may be attribu-
ted to higher lateral confining pressure. It may also be noted that
for the shear strength of clay more than 8 kPa, the ultimate load
intensity of floating and end bearing granular piles are more or less
equal. The encasement stiffness of reinforced granular pile was
varied in the range of 7–250 kN/m. The shear strength of clay
was kept as 5 kPa. The effect of stiffness on the ultimate load inten-
sity of granular pile is shown in Fig. 10b. For floating granular pile,
the ultimate load intensity has been found to increase up to 50 kN/
m stiffness of encasement, and then is constant for higher stiffness.
Due to higher stiffness of encasement in the floating granular piles,
it was observed that pile penetrated into soft clay rather than bul-

Fig. 13. Variation of a in term of spacing for only horizontal strips and combined
reinforced granular piles.

Fig. 11. Vertical load intensity settlement behaviour of end bearing granular piles
with horizontal stripped.

Fig. 12. Vertical load intensity settlement behaviour of combined end bearing Fig. 14. Vertical load intensity settlement behaviour of granular piles (entire area
reinforced granular piles. loaded); (a) unreinforced, (b) vertical encased.
 M. Hasan, N.K. Samadhiya / Computers and Geotechnics 87 (2017) 178–187
tical spacing of strips. It was found that a is varying in range of
4.40–6.49 (exp.) and 4.20–6.61 (FEM) for horizontal reinforced
granular pile. The reason for the improvement in ultimate load
intensity may be due to mobilising frictional resistance between
strips and stone aggregates.
Fig. 15. Vertical load intensity settlement behaviour of combined reinforced end
bearing granular piles (entire area loaded).
Fig. 16. Deformed shapes of some granular piles; (a) laboratory model tests, (b) FEM.
185
Laboratory model tests were also conducted to study the com-
bined effect of vertical encased-horizontal stripped end bearing
granular piles, installed in very soft clays. The results of combined
reinforced granular pile from experimental and FEM analysis in
term of vertical load intensity settlement relationship are pre-
sented in Fig. 12. The ultimate load intensity of granular piles fur-
ther improved due to mobilising hoop stresses in geotextile as well
as frictional resistance between geogrid strips and aggregates. The
variation of a in terms of spacing for only horizontal strips and
combined reinforced granular piles is presented in Fig. 13. The ulti-
mate load intensity ratios for combined reinforced granular piles
(a) were found much higher as compared to unreinforced, rein-
forced and horizontal reinforced granular piles. For combined rein-
forced granular piles, a was found in the range of 7.92–11.38 (exp.)
and 7.88–11.09 (FEM).
4.2. Comparison of laboratory model tests and FEM when entire area is
loaded
Some laboratory model tests and FEM analyses have also been
performed to evaluate the improvement in the stiffness of granular
piles treated ground. It represents actual field behaviour of an inte-
rior granular pile when large groups of piles are loaded simultane-
ously [26]. The experimental and FEM results in terms of vertical
load intensity settlement behaviour of untreated, treated ground
186 M. Hasan, N.K. Samadhiya / Computers and Geotechnics 87 (2017) 178–187
with unreinforced (UR) and encased floating as well as end bearing
granular pile are presented in Fig. 14a-b.
Fig. 15 presents results of experimental and FEM analysis in
terms of vertical load intensity settlement behaviour of combined
reinforced end bearing granular pile treated ground. It may be
noted that curves are almost linear up to settlement of 10–
12 mm and then the vertical intensity is more or less constant.
For comparison purpose, vertical load intensity corresponding to
25 mm settlement was selected for all tests and presented in
Table 5. This settlement (25 mm) is about 14% of loading plate.
The improvement in the stiffness of treated ground due to presence
of the granular pile has been evaluated in terms of stiffness
improvement factor (b). The stiffness improvement factor (b) is
expressed as the ratio of vertical load intensity of treated ground
to that of untreated ground at same settlement (25 mm). The stiff-
ness improvement factor (b) was found in the range of 1.56–2.0
(exp.) and 1.64–1.95 (FEM) for unreinforced granular pile, whereas
for encased granular piles, b was varying in the range of 1.68–2.45
(exp.) and 1.83–2.39 (FEM). For combined reinforced granular
piles, b has been found in the range of 2.41–2.85 (exp.) and
2.53–2.88 (FEM).
4.3. Deformation and failure modes
Deformed shapes of granular piles were obtained after comple-
tion of laboratory model tests and PLAXIS 3D studies. Fig. 16 shows
deformed shapes for floating and end bearing granular piles (unre-
inforced, encased and combined) when these were loaded alone.
The failure mainly attributed to bulging in unreinforced and
encased granular piles due to poor lateral confinement. The bul-
ging reduced in the case of encased granular pile due to additional
confinement provided by geotextile. The depth of maximum bul-
ging was observed in the range of 1–1.6 times the diameter of
the pile from the top. Whereas, maximum bulging was observed
close to top of pile in case of unreinforced granular piles compared
to encased granular piles. The total length of the granular pile
experiencing bulging was found to be in the range of 1.5–2.5 times
the diameter of the pile from the top. When entire cross sectional
area is loaded, unreinforced and reinforced end bearing granular
piles deformed with negligible bulging.
4.4. Comparison with existing theory and field studies
A chart developed for the comparison of present study with
existing theory and field studies [26,35,36]. The relationship
Fig. 17. Comparison of ultimate load intensity with existing theory and field
studies.
between shear strength of clay, internal friction angle of granular
material and ultimate load intensity of granular pile is shown in
Fig. 17. The existing theories predict the load capacity of a single
granular pile installed in infinite soil mass without consider the
effect of spacing. The results of present study are in close agree-
ment with existing theory.
5. Conclusions
In the present investigation, laboratory model tests as well as
numerical analyses were carried out on 75 mm diameter granular
piles (floating and end bearing). The effect of reinforcement, shear
strength of clay, encasement stiffness, and length of granular piles
was studied. The vertical load intensity-settlement plots of labora-
tory model tests were compared with that obtained from PLAXIS
3D. The following conclusions can be drawn from the study:
1. The ultimate load intensity of granular piles installed in very
soft clay has been found to improve due to inclusion of geosyn-
thetic materials. It shows further enhancement with the
increase in the length of granular piles.
2. Unreinforced end bearing granular piles can favourably be
replaced by reinforced floating granular piles.
3. The ultimate load intensity of encased granular piles improves
with increase in the shear strength of clay. For the shear
strength of clay higher than 8 kPa, ultimate load intensity of
encased floating and end bearing granular piles is found to be
more or less equal.
4. For floating granular piles, the ultimate load intensity is found
to increase up to 50 kN/m stiffness of geotextile, and then is
constant for higher stiffness. The ultimate load intensity of
end bearing granular piles increases linearly with the increase
in stiffness of geotextile.
5. The ultimate load intensity of end bearing granular pile with
horizontal strips increases with the reduction in vertical spac-
ing of geogrid strips. Granular pile reinforced with horizontal
strips @ 25 mm c/c spacing can be the substituted for encased
granular pile.
6. The ultimate load intensity ratio (a) for combined reinforced
end bearing granular piles (a) is found much higher and lie in
the range of 7.92–11.38 (experimental),7.88–11.09 (FEM). For
practical applications, the number of granular piles can be sig-
nificantly reduced as compared to unreinforced, encased and
horizontal reinforced granular piles supported structures.
7. The reinforcement provided in the granular piles controls bul-
ging of the piles. When granular piles are loaded alone, the
depth of maximum bulging in unreinforced and encased granu-
lar lies in the range of 1–1.6 times the diameter of the pile. The
total length of the granular pile experiencing bulging is found to
be in the range 1.5–2.5 times the diameter of the pile.
8. The stiffness improvement factor (b) of improved ground
increases for encased granular piles and further enhances when
the length of granular piles increases. It improves in the reduc-
tion of vertical spacing of strips for combined reinforced end
bearing granular piles.
6. Limitations and future scope
It may be noted that present laboratory and numerical study
has been carried out under short term loading condition with con-
stant area replacement ratio. Long term behaviour of granular piles
with different area replacement ratios is still required under in-situ
conditions. Fabrication of horizontal reinforcement layers in full
scale granular piles is deficient in practical applications. Hence a
further research to thoroughly study merits of horizontal rein-
 M. Hasan, N.K. Samadhiya / Computers and Geotechnics 87 (2017) 178–187
forcement is required. Well documented case histories of success-
ful utilization of encased granular piles are rather limited. There-
fore the well documented data sets of field performance are still
required.
Funding
This research did not receive any specific grant from funding
agencies in the public, commercial, or not-for-profit sectors.
References
[1] Greenwood DA. Mechanical improvement of soils below ground surfaces. In:
Proceedings, ground engineering conference, institution of civil engineers,
London; 1970. p 11–22.
[2] Castro J, Cimentada A, Costa A, Canizal J, Sagaseta C. Consolidation and
deformation around stone columns: Comparison of theoretical and laboratory
results. Comput Geotech 2013;49:326–37.
[3] Barksdale RD, Bachus RC. Design and construction of stone columns. Federal
highway administration RD, 83/026; 1983.
[4] Mitchell JK, Huber TR. Performance of a stone column foundation. J Geotech
Eng ASCE 1985;111(2):205–23.
[5] Han J, Ye SL. Simplified method for consolidation rate of stone column
reinforced foundation. J Geotech Geoenviron Eng ASCE 2001;127(7):597–603.
[6] Van Impe W, Silence P. Improving of the bearing capacity of weak hydraulic
fills by means of geotextiles. In: Proceedings of the 3rd international
conference on geotextiles, Vienna, Austria; 1986. p. 1411–6.
[7] Murugesan S, Rajagopal K. Geosynthetic-encased stone columns: numerical
evaluation. Geotext Geomembr 2006;24:349–58.
[8] Murugesan S, Rajagopal K. Model tests on geosynthetic encased stone
columns. GeosynthetInt J 2007;24(6):346–54.
[9] Gniel J, Bouazza A. Improvement of soft soils using geogrid encased granular
columns. Geotext Geomembr 2009;27(3):167–75.
[10] Murugesan M, Rajagopal K. Studies on the behaviour of single and group of
geosynthetic encased stone columns. J Geotech Geoenviron Eng 2010;136
(1):129–39.
[11] Ali K, Shahu JT, Sharma KG. Model tests on geosynthetic-reinforced stone
columns: a comparative study. Geosynth Int 2012;19(4):292–305.
[12] Ghazavi M, Afshar JN. Bearing capacity of geosynthetic encased stone columns.
Geotext Geomembr 2013;38:26–36.
[13] Dash SK, Bora MC. Influence of geosynthetic encasement on the performance
of stone columns floating in soft clay. Can Geotech J 2013;50(7):754–65.
[14] Zhang L, Zhao M. Deformation analysis of geotextile-encased stone columns.
ASCE Int J Geomech 2015;15(3):04014053.
[15] Mohapatra SR, Rajagopal K, Sharma J. Direct shear tests on geosynthetic-
encased granular columns. Geotext Geomembranes 2016;44:396–405.
187
[16] Yoo C. Performance of geosynthetic-encased stone columns in embankment
construction: numerical investigation. J Geotech Geoenviron Eng ASCE
2010;136(8):1148–60.
[17] Pulko B, Majes B, Logar J. Geosynthetic-encased stone columns: analytical
calculation model. Geotext Geomembranes 2011;29(1):29–39.
[18] Zhang Y, Chan D, Wang Y. Consolidation of composite foundation improved by
geosynthetic-encased stone columns. Geotext Geomembranes 2012;32:10–7.
[19] Chen JF, Li LY, Xue JF, Feng SZ. Failure mechanism of geosynthetic encased
stone columns in soft soils under embankment. Geotext Geomembr 2015;43
(5):424–31.
[20] Alexiew D, Brokemper D, Lothspeich S. Geotextile encased columns (GEC):
load capacity, geotextile selection and pre-design graphs. In: Contemporary
Issues in Foundation Engineering, Geotechnical Special Publication No. 131,
ASCE, Reston, VA, USA; 2005. p. 1–14.
[21] Yoo C, Lee D. Performance of geogrid-encased stone columns in soft ground:
full-scale load tests. Geosynth Int 2012;19(6):480–90.
[22] Almeida M, Hosseinpour I, Ricci M, Alexiew D. Behavior of geotextile-encased
granular columns supporting test embankment on soft deposit. J Geotech
Geoenviron Eng ASCE 2014:04014116. 1943-5606.0001256.
[23] Sharma RS, Kumar BRP, Nagendra G. Compressive load response of granular
piles reinforced with geogrids. Can Geotech J 2004;41(1):187–92.
[24] Ayadat T, Hanna AM, Hamitouche A. Soil improvement by internally reinforced
stone column. Ground Improv 2008;161(2):55–63.
[25] Hong YS, Wu CS. The performance of a sand column internally reinforced with
horizontal reinforcement layers. Geotex Geomembranes 2013;41:36–49.
[26] Ambily AP, Gandhi SR. Behavior of stone columns based on experimental and
FEM analysis. J Geotechgeoenviron Eng ASCE 2007;133(4):405–15.
[27] Ranjan G. Ground treated with granular piles and its response under load.
Indian Geotech J 1989;19(1):1–86.
[28] IS: 15284 (part 1). Indian standard code of practice for design and construction
for ground improvement-guidelines. New Delhi, India: Indian Standards
Institution; 2003.
[29] Iai S. Similitude for shaking table tests on soil-structure fluid models in 1g
gravitational field. Soils Found 1989;29(1):105–18.
[30] IS: 1498. Classification and identification of soils for general engineering
purposes. New Delhi, India: Indian Standards Institution; 2000.
[31] ASTM D4595. Standard test method for tensile properties of geotextiles by the
wide-width strip method. ASTM International, West Conshohocken, PA, USA;
1986.
[32] Mohanty P, Samanta M. Experimental and numerical studies on response of
the stone column in layered soil. Int J Geosynth Ground Eng 2015;1:27.
[33] Ng KS, Tan SA. Stress transfer mechanism in 2D and 3D unit cell models for
stone column improved ground. Int J Geosynth Ground Eng 2015;1:3.
[34] Bowles JE. Foundation analysis and design. Singapore: McGraw Hill
International Editions; 1997.
[35] Hughes JMO, Withers NJ. Reinforcing of soft cohesive soils with stone columns.
Ground Eng 1974;7(3):42–9.
[36] Hughes JMO, Withers NJ, Greenwood DA. A field trial of reinforcing effect of
stone column in soil. In: Proc., Ground Treatment by Deep Compaction,
Institution of Civil Engineers, London; 1976. p. 32–44.