th

Proceedings of the 19 International Conference on Soil Mechanics and Geotechnical Engineering, Seoul 2017

Influence of geotextile encasement in triaxial tests on gravel

Influence de l'enrobage de géotextile sur des essais triaxiaux sur gravier

Marina Miranda
School of Engineering and The Built Environment ,Napier University, U.K, [email protected]

Almudena da Costa, Jorge Castro, César Sagaseta & Jorge Cañizal

Department of Ground Engineering and Materials Science, University of Cantabria, Spain

ABSTRACT: Stone columns installed in extremely soft soils may significantly reduce the effectiveness of this treatment due to the
insufficient lateral confinement provided by the soft soil. The encasement of columns with geotextiles is commonly used in these
situations with satisfactory results thanks to the extra confinement provided by the geotextile to the column. Drained triaxial tests on
encased and non-encased samples of gravel have been performed to study the influence of the encasement on the behavior of stone
columns. Two different densities of the gravel and two different geotextiles were used. The study is focused on the increase in
strength of encased samples compared with non-encased ones, the extra confining pressure provided by the geotextiles and the
mobilized friction angle of the gravel. In this paper, the results of some of these laboratory tests are compared with numerical
simulation. All of the results show the improvement achieved when the gravel is encased with the geotextiles.

RÉSUMÉ : Les colonnes de pierre installées dans des sols extrêmement doux peuvent réduire significativement l'efficacité de ce
traitement en raison du confinement latéral insuffisant fourni par le sol mou. L'encasement des colonnes avec des géotextiles est
couramment utilisé dans ces situations avec des résultats satisfaisants grâce au confinement supplémentaire fourni par le géotextile à la
colonne. Des essais triaxiaux drainés sur des échantillons de gravier encastrés et non encastrés ont été réalisés pour étudier l'influence de
l'encasement sur le comportement des colonnes de pierre. Deux densités différentes du gravier et de deux géotextiles différents ont été
utilisées. L'étude porte sur l'augmentation de la résistance des échantillons encastrés par rapport aux non encapsulés, la pression de
confinement supplémentaire fournie par les géotextiles et l'angle de friction mobilisé du gravier. Dans cet article, les résultats de certains
de ces tests de laboratoire sont comparés à la simulation numérique. Tous les résultats montrent l'amélioration obtenue lorsque le gravier
est enveloppé avec les géotextiles.
KEYWORDS: gravel column, encased column, triaxial compression test, numerical analysis.
tests on encased samples, such as the work of Rajagopal et al.
1 INTRODUCTION (1999), who tested samples of granular soil encased in single
and multiple geocells using different types of geotextiles, Wu
Stone columns with granular material are often used to improve and Hong (2009), who carried out triaxial compression tests on
bearing capacity, to accelerate the speed of consolidation and to reinforced and non-reinforced columns mainly to assess the
reduce settlements on soft soil strata. Insufficient lateral support influence of the encasement on the radial strains of the sample
in extremely soft soils (su<15 kPa) results in a significant and on the deviator stress, or Najjar et al. (2010) on normally
reduction in the effectiveness of this treatment with stone consolidated kaolin samples reinforced with single sand
columns. This lack of lateral confinement mainly occurs at columns.
shallow depths, causing bulging failure in the upper portion of In addition to the experimental studies, several numerical
the columns (e.g., Huges and Withers, 1974; Madhav and analyses have been carried out to study various factors that
Miura, 1994). In these cases, an improvement in stone column influence the behavior of the encased columns, such as the
behavior can be further enhanced by encapsulating the column stiffness of the encasement (e.g., Almeida et al., 2013;
with a flexible sleeve (geotextile or geogrid), which can be a Chungsik, 2010; Khabbazian et al., 2010; Murugesan and
continuous sleeve or can be formed with a longitudinal union. Rajagopal, 2006), the stiffness parameters of the compacted
The behavior of encapsulated stone columns has been stone (Lo et al., 2010), the encasement length (Keykhosropur et
studied by numerous research initiatives through the al., 2012), the shear-induced volumetric dilation of the fill
development of experimental tests, theoretical and numerical material (Hong, 2012), the behavior under no monotonic loads
analyses and field applications. (e.g., Prisco et al., 2006), or the influence of the finite element
An important part of the experimental studies has been modeling approach (e.g., Yoo and Kim, 2009).
performed by small-scale laboratory tests, focusing on the
analysis of load-settlement behavior (e.g., Black et al., 2007;
Ghazavi and Afshar, 2013; Gniel and Bouazza, 2009; 2 EXPERIMENTAL PROGRAM
Malarvizhi and Ilamparuthi, 2007; Murugesan and Rajagopal,
2007, 2010). For these experimental studies, the sleeves were Three series of triaxial tests were performed, the first one with
mainly fabricated with geotextiles with a longitudinal union, samples of only gravel and the second and third ones with
which was commonly made by an overlap of the fabric that was gravel encased with two different geotextiles. Each of the series
sewn (e.g., Murugesan and Rajagopal, 2007, 2010), by a glued was carried out with samples with two different relative
overlap of the fabric (e.g., Gniel and Bouazza, 2009), or by densities of the gravel, Dr=50% and 80%.
overlapping the encasement by a nominal amount and relying
on the interlock between the aggregate and the section of 2 .1 Test materials
overlap (e.g., Gniel and Bouazza, 2010). In any case, this union Details about the properties of the materials employed in the
creates a weak point that reduces the strength of the geotextile. tests can be found in Miranda and Da Costa (2016). Below is a
Other experimental analyses are based on triaxial compression summary of these properties.
 th
Proceedings of the 19 International Conference on Soil Mechanics and Geotechnical Engineering, Seoul 2017

Uniformly graded limestone gravel with particle sizes 2 .3 Test procedure

between 4 and 5 mm was employed for the laboratory tests.
This material is the same as that employed in Cimentada et al. Drained triaxial compression tests were conducted. The test
(2011) and Miranda et al. (2015) for the study of the behavior procedure consisted of the following stages. First, the sample
of soft soils improved with non-encased stone columns. was saturated, and the desired confining pressure was applied
Two different geotextiles were used for column with opened drainage until consolidation of the sample
reinforcement, each made using a different flat fabric. Both occurred. Four different confining pressures, pc, of 25, 50, 150
geotextiles, along with the properties of both fabrics, were and 300 kPa were chosen for the tests. Afterwards, the sample
provided by Huesker Synthetic Gmbh. In both cases, the sleeve was axially loaded under a constant vertical strain rate of 0.002
was prepared by cutting the fabric and preparing it in a mm/s, keeping the drainage open.
cylindrical shape with a longitudinal union. It is important to
note that, in real treatments, continuous sleeves without 3 NUMERICAL ANALYSIS
longitudinal unions are constructed with these fabrics such that
this weak point does not exist. Numerical analyses using the finite element software Plaxis 2D
The comparison of the numerical simulations with the 2012 (Brinkgreve, 2012) were performed to simulate some of
laboratory tests is presented only for one of the geotextiles. Fig1 the triaxial tests performed in the laboratory. In particular,
shows the load-strain behavior of both the fabric and triaxial tests on samples encased with geotextile1 and with a
geotextile1 (fabric+joint). Selected values from the curve of this relative density of gravel of Dr=50% were chosen for the
geotextile are summarized in Table1, along with the numerical simulation.
corresponding secant modulus. The Hardening soil model (Schanz et al., 1999) was used to
More details about the properties of these geotextiles can be simulate the behavior of the gravel due to the stress dependency
found in Miranda (2014). of the stiffness described in the experimental results. This
model is an extension of the hyperbolic model developed by
180 Duncan and Chang (1970), which is based on the theory of
Geotextile1 plasticity and includes soil dilatancy and introduces a yield cap.
160 The strength of the soil in this model is given by the peak
friction angle () and the corresponding dilatancy (). With
140 these to values, the critical and mobilized friction angles, and
fabric the mobilized dilatancy are obtained by the following
Tensile force (kN/m)

120
equations:
100
sin   sin
fabric+ joint sin cr  (1)
80 1  sin  sin

60
 1'   3'
sin  m  (2)
40  1'   3'  2c cot 

20
sin  m  3 4 sin    m  0
0  sin  m  sin cr  (3)
0 2 4 6 8 10 12 14 sin  m  3 4 sin  and   0  sin m  max ,0 
Tensile strain (%)  1  sin  m sin cr 
Figure 1. Tensile force versus strain of the geotextiles.
Different stiffness and friction and dilatancy angles were
employed for each of the four confining pressures used in the
laboratory tests. These values were estimated from the
Table 1. Results of tensile tests. Geotextile1 (fabric+joint) experimental results (Fig. 2¡Error! No se encuentra el origen
Strain [%] Tension [kN/m] Secant modulus Jg [kN/m] de la referencia.) and are summarized in Table 2.
2 13 650 TABLE 2. Gravel parameters for the numerical analyses
5 31 620 Chamber pressure, Friction Dilatancy E50ref Eoedref
pc [kPa] angle* [º] [º] [MPa] [MPa]
8 50 625 25 54 14 9.5 8.2
Maximum 12.3 77 626 50 47 12 10.0 10.0
150 43 5 13.5 12.0
300 40 0 16.5 13.0
2 .2 Specimen preparation
(*) peak values
The triaxial compression tests were performed on 200-mm-high
x 100-mm-in-diameter specimens of only gravel and gravel
The unloading-reloading stiffness was chosen as three times
encased with a geotextile. Two different dry unit weights of the
the secant stiffness. The values of the Poisson's ratio for each
gravel were used in the research, 14.5 kN/m3 and 15.8 kN/m3,
confining pressure have been obtained from the initial slope of
which correspond to relative densities of Dr=50% and Dr=80%.
the volume strain – axial strain curves (Fig. 2). A Poisson's ratio
The specimen preparation is described in Miranda and Da Costa
of 0.25 was obtained for the chamber pressures of 25 and 50
(2016).
kPa and 0.15 for chamber pressures of 150 and 300 kPa. The
best numerical analysis fit with the experimental results was
obtained for a power for stress dependency of the stiffness of
m=0.2.
 th
Proceedings of the 19 International Conference on Soil Mechanics and Geotechnical Engineering, Seoul 2017

Grid elements were used to represent the geotextile. The circumferential strain that lead to a continuous increase of axial
stiffness is the only parameter required for this material, and a stress, which is supported by the high stiffness of the
value of 620 kN/m was adopted for the calculations. geotextiles, resulting in higher axial stress increments for the
Two calculation phases were defined in the numerical model encased samples compared with the non-encased ones (only
in Plaxis to simulate the triaxial tests with the encased samples. gravel). The noticeable influence of the confining pressure in
The first one was a consolidation calculation in which the the results of the non-encased samples is not so noticeable in
chamber pressure is applied to the sample. During the second the results of the encased ones.
one, the axial stress is increased until a certain value that, for The strain hardening behavior observed in the stress-strain
each chamber pressure, was chosen as the maximum value curves of encased samples is attributed to the mobilization of
reached in the laboratory tests. the geotextile strength. During the test, increasing radial and
circumferential strains are developed. As a consequence, in
encased samples, the geotextile develops an increasing
4 RESULTS circumferential tensile force per unit length, the value of which
depends on the stiffness of the geotextile. This increase in the
Results from the laboratory tests can be found in Miranda and circumferential tensile force of the geotextile results in an extra
Da Costa (2016). The results presented in this section are confining pressure provided by the encapsulating geotextile to
focused on the comparision between the laboratory and the the gravel, which is added to the chamber pressure applied for
numercial results. the test.
As was expected, the numerical results do not show the
4 .1 Gravel specimens described initial behavior obtained in the laboratory tests where
The numerical simulation of the consolidated-drained triaxial the influence of the geotextile is not realized until a certain axial
compression tests on non-encased samples was performed using strain is reached. The reason is that, in the numerical
the Soil Test facility in Plaxis. The numerical results are given calculation, the geotextile develops radial strain from the
in 2, and they show a good fit with the experimental results. beginning of the second calculation phase, whereas in the
laboratory tests, some axial strain is needed to achieve a
1400 complete response of the geotextile. This is slightly noticeable
300 kPa Dashed lines: Numerical fitting Dr=50% in the axial stress increments, but it is clearly shown by the
Axial stress increment, 1 (kPa)

1200 150 kPa Gravel

50 kPa volumetric strain with a different initial response. However, if
25 kPa
1000 the numerical volumetric curves are horizontally translated to a
800 certain axial strain, a good fit with the laboratory results is
obtained.
600
Fig. 4 shows the ratio between the axial stress increments in
400 encased and non-encased samples obtained from the numerical
200
analysis and its comparison with the experimental results. The
agreement of numerical and experimental results is quite
0
0 2 4 6 8 10 12 14 16 18 20 22 satisfactory.
Axial strain (%)
3000
300 kPa GTEX1 Dr=50%
Axial stress increment, 1 (kPa)

12
150 kPa
10 2500 50 kPa
8 25 kPa
2000
Volumetric strain (%)

6
4 Dashed lines: Numerical fitting
2 1500
0
-2 0 2 4 6 8 10 12 14 16 18 20 22 1000
-4
-6 500
-8
-10 0
0 2 4 6 8 10 12 14 16 18 20 22
-12
Axial strain (%) Axial strain (%)

Figure 2. Numerical fitting of the drained triaxial tests of the gravel

4

4 .2 Encased gravel specimens 2

Volumetric strain (%)

0
The results of the axial stress increment and the volumetric 0 2 4 6 8 10 12 14 16 18 20 22
-2
strain against the axial strain are presented in Fig. 3, along with
those from the numerical analyses. -4

Comparing the results for non-encased and encased samples -6

the following conclusions are obtained. For low axial strains, -8
the behavior of encased and non-encased samples is similar, -10
with no influence of the geotextile. This is attributed to the
-12
volume change of the sample during the defrosting process and Axial strain (%)
isotropic consolidation. This volume change reduces the Figure 3. Numerical results of the drained triaxial tests in samples
diameter of the sample such that it is slightly smaller than the encased with geotextile1
diameter of the encasement; therefore, some axial strain is
needed so that the sample is in contact with the encasement.
This is not constant along the sample, so the process takes place
progressively. Once the contact is reached, at different axial
strains depending on the gravel density and the chamber
pressure, the influence of the encasement becomes noticeable.
The geotextile develops increasing tensile force and increasing
 th
Proceedings of the 19 International Conference on Soil Mechanics and Geotechnical Engineering, Seoul 2017

14 Hong Y.S. 2012. Performance of encased granular columns considering

13 300 kPa Dr=50% shear-induced volumetric dilation of the fill material. Geosynthetics
150 kPa GTEX1
12 International 19 (6), 438 –452.
1 (reinforced) / 1 (unreinforced)

50 kPa
11 25 kPa Huges J.M.O. and Withers N.J. 1974. Reinforcing of cohesive soils
10 with stone columns. Ground Engineering 7 (3), 42-49.
9 Dashed lines: Numerical fitting Khabbazian M., Kaliakin V.N. and Meehan C.L. 2010. Numerical study
8 of the effect of geosynthetic encasement on the behaviour of
7 granular columns. Geosynthetics International 17 (3), 132 –143.
6 Keykhosropur L., Soroush A. and Imam R. 2012. 3D numerical
5 analyses of geosynthetic encased stone columns. Geotextiles and
4 Geomembranes 35, 61-68.
3 Lo S.R., Zhang R. and Mak J., 2010. Geosynthetic-encased stone
2 columns in soft clay: A numerical study. Geotextiles and
1 Geomembranes 28 (3), 292-302.
0 Madhav M.R. and Miura N. 1994. Soil improvement. Panel report on
0 2 4 6 8 10 12 14 16 18 20 22 stone columns. In: Proceedings of the 13th International
Axial strain (%)
Conference on Soil Mechanics and Foundation Engineering, New
Figure 4. Numerical result for the ratio of axial stress increment Delhi, India. 5, 163-164.
(reinforced/unreinforced samples) Malarvizhi S.N. and Ilamparuthi K. 2007. Comparative study on the
behavior of encased stone column and conventional stone column.
Soils and Foundations 47 (5), 873-885.
5 CONCLUSION Miranda M. 2014. Influencia de la densidad y del confinamiento con
geotextil en columnas de grava. PhD dissertation. Spain. University
A series of consolidated-drained triaxial compression tests were of Cantabria.
performed to study the effectiveness of encapsulation on the Miranda M. and Da Costa A. 2016. Laboratory analyses of encased
strength of gravel columns. stone columns. Geotextiles and Geomembranes 44 (3), 269-277.
The improvement in the strength of encased samples was Miranda M., Da Costa A., Castro J. and Sagaseta C. 2015. Influence of
evaluated by the ratio of the axial stress increment in the gravel density in the behaviour of soft soils improved with stone
columns. Canadian Geotechnical Journal, 52 (12), 1968-1980.
encased and the non-encased samples. This improvement is Murugesan S. and Rajagopal K. 2006. Geosynthetic-encased stone
more significant for low confining pressures with a value of 9.5 columns: numerical evaluation. Geotextiles and Geomembranes 24
for pc=25 kPa at an axial strain of 17% and 2.5 for pc =300 kPa. (6), 349-358.
Numerical analyses were conducted for comparison with the Murugesan S. and Rajagopal K. 2007. Model tests on geosynthetic-
experimental results. This comparison was performed for encased stone columns. Geosynthetics International 14 (6), 346-
samples encased with geotextile1 and with Dr=50%. The 354.
numerical simulation matches the laboratory test results quite Murugesan S. and Rajagopal K. 2010. Studies on the behavior of single
well except for the initial behavior where the influence of the and group of geosynthetic encased stone columns. . J. Geotechnical
geotextile is not realized in the laboratory tests until a certain and Geoenvironmental Engineering 136 (1), 129-139.
Najjar S., Sadek,S. and Maakaroun T. 2010. Effect of sand columns on
axial strain has been developed. the undrained load response of soft clays. J. Geotech. Geoenviron.
Eng. ASCE, 136(9), 1263–1277.
Prisco C., Galli A., Cantarelli E.and Bongiorno D. 2006. Geo-
6 ACKNOWLEDGEMENTS reinforced sand columns: small scale experimental tests and
theoretical modeling. Proc. of 8th International conference of
The authors would like to thank HUESKER Synthetic GmbH Geosynthetics. Yokahama. 1685-1688.
Company for providing the geotextiles needed for the Rajagopal K., Krishnaswamy N.R. and Madhavi Latha G., 1999.
encasement and their properties. Behavior of sand confined with single and multiple geocells.
Geotextile and Geomembranes 17 (3), 171-184.
Schanz T., Vermeer P. and Bonnier P. 1999. The hardening soil model:
7 REFERENCES formulation and verification. Proc. Plaxis Symp. Beyond 2000 in
computational geotechnics Amsterdam, Balkema.
Almeida M. S. S., Hosseinpour I. and Riccio M. 2013. Performance of a Yoo, C., Kim, S.B., 2009. Numerical modeling of geosynthetic-encased
geosynthetic-encased column (GEC) in soft ground: numerical and stone column-reinforced ground. Geosynthetics International 16
analytical studies. Geosynthetics International 20 (4), 252-262. (3), 116 –126.
Black J. A., Sivakumar M. R., Madhav M. R. and Hamii, G. A. 2007. Wu, C.S., Hong, Y.S., 2009. Laboratory tests on geosynthetic-
Reinforced stone columns in weak deposits: laboratory model encapsulated sand columns. Geotextile and Geomembranes 27 (2),
study. J. Geotechnical and Geoenvironmental Engineering 133 (9), 107-120.
1154-1161.
Brinkgreve R.B.J. 2012. Plaxis finite element code for soil and rock
analysis, 2D. Rotterdam: Balkema.
Chungsik A.M. 2010. Performance of geosynthetic-encased stone
columns in embankment construction: Numerical investigation. J.
Geotech. Geoenviron Eng. ASCE. 136, 1148-1160.
Cimentada A., da Costa A., Cañizal J. and Sagaseta C. 2011.
Laboratory study on radial consolidation and deformation in clay
reinforced with stone columns. Canadian Geotechnical Journal
48(1), 36-52.
Duncan J.M. and Chang C.Y. 1970. Nonlinear analysis of stress and
strain in soils. J. Soil Mech. Found. Div. ASCE 96 (5), 1629-1653.
Ghazavi M. and Afshar J.N. 2013. Bearing capacity of geosynthetic
encased stone columns. Geotextiles and Geomembranes 38, 26-36.
Gniel J., Bouazza, A., 2009. Improvement of soft soils using geogrid
encased stone columns. Geotextiles and Geomembranes 27 (3),
167–175.
Gniel, J. and Bouazza A. 2010. Construction of geogrid encased stone
columns: A new proposal based on laboratory testing. Geotextiles
and Geomembranes 28 (1), 108-118.