Geotextiles and Geomembranes 27 (2009) 19–30

Contents lists available at ScienceDirect

Geotextiles and Geomembranes

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

Large scale direct shear tests of soil/PET-yarn geogrid interfaces

Chia-Nan Liu*, Yu-Hsien Ho, Jian-Wen Huang
University Road, Puli, Nantou, Civil Engineering Department, National Chi-Nan University, Nantou 545, Taiwan

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

Article history: The interface shear strength of soil against geosynthetic is of great interest among the researchers in
Received 2 November 2007 geosynthetic properties. This study conducts a series of large scale direct shear tests to investigate the
Received in revised form 24 February 2008 interface shear strength of different soils (sand, gravel, and laterite) against PET-yarn geogrids of various
Accepted 9 March 2008
tensile strengths, percent open area, and aperture patterns. First, the appropriateness of different set-ups
Available online 11 June 2008
of a lower shearing box is examined in this study. It reveals that a lower box which is filled with the test
soil and is of the same size as the upper box is more suitable for testing the soil/geogrid interface. The
Keywords:
test results show that the soil/PET-yarn geotextile interface has significantly lower shear strength than
Direct shear
Interface
soil strength. The ratio of shear strength soil/PET-yarn geotextile interface to internal shear strength of
PET-yarn geogrid soil is about 0.7–0.8 for Ottawa sand and for laterite, and it is about 0.85–0.95 for gravel. On the other
Shear strength hand, the soil/geogrid interface has higher shear strength. The ratio of shear strength soil/PET-yarn
Soil geogrid interface to internal shear strength of soil is about 0.9–1.05. It is found that the shear strength
ratio of soil/PET-yarn geogrid interface is positively correlated to the transverse tensile strength of the
PET-yarn geogrid. However, it is negatively correlated with the aperture length and percent open area of
the PET-yarn geogrid. The interface shear test results of PET-yarn geogrid against different soils are
compared with the test results predicted by a classical model for analyzing the applicability of the
classical model. Further, a simple model is proposed herein to estimate the bearing resistance provided
by the transverse ribs of geogrid. It shows this component to be about 0–15% when PET-yarn geogrid is
against Ottawa sand or laterite, while it is smaller when the PET-yarn geogrid is against gravel.
Ó 2008 Elsevier Ltd. All rights reserved.

1. Introduction include Jarret and Bathurst (1985), Cancelli et al. (1992), Bauer and
Zhao (1993), Cazzuffi et al. (1993), Bakeer et al. (1998), and Abu-
The understanding of soil/geosynthetic interface shear strength Farsakh and Coronel (2006). Research involving the interface of soil
is essential to the design and stability analysis of geosynthetic and geomembrane under different conditions such as degree of
reinforced soil structures. For example, an interface of a stronger saturation and bentonite contamination can be found in Fishman
shearing resistance in a geosynthetic lined slope can reduce the and Pal (1994), Hsieh and Hsieh (2003), Fleming et al. (2006),
tensile forces mobilized in the geosynthetics, as well as increase the Zabielska-Adamska (2006), Wu et al. (2008) and Vukelic et al.
slope inclination (Liu and Gilbert, 2003; Palmeira and Viana, 2003). (2008). The interface shear strength of soil to other types of
The shear strength of soil/geosynthetic interface is also essential to geosynthetics such as tire shreds, rubber chips, or geofoam is also
numerical simulation of the behavior of strip footing on geogrid investigated through direct shear tests (Bernal et al., 1997; Ho,
reinforced soft foundation (El Sawwaf, 2007). Though the shear 2001; Xenaki and Athanasopoulos, 2001).
strength of soil/geosynthetic interface has been investigated by The interface shear resistance of soil against geomembrane or
conducting other tests, such as tilt table tests (Wu et al., 2008), geotextiles comes from the friction between soil and geosynthetic,
direct shear test is still the most common testing method. For ex- because soil particles are not trapped into the small openings of
ample, direct shear tests involving interfaces between soil and geosynthetics. However, the direct shear resistance for the geogrid
geotextile have been performed by Richards and Scott (1985), Lee is more complex. The contact between ribs and soil provides
and Manjunath (2000), Mahmood et al. (2000), and Bergado et al. interface shear resistance. At the same time, the interlocking of soil
(2006). Studies involving interfaces between soil and geogrid particles above and below the geogrid is allowed within geogrid
apertures. Therefore, the shear resistance for the soil/geogrid
interface includes at least (i) shear resistance between soil and the
* Corresponding author. Tel.: þ886 49 2910960; fax: þ886 49 2918679. surface of geogrids ribs, and (ii) internal shear resistance of the soil
E-mail address: [email protected] (C.-N. Liu). in the openings of geogrid. The bearing resistance between the

0266-1144/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.geotexmem.2008.03.002
20 C.-N. Liu et al. / Geotextiles and Geomembranes 27 (2009) 19–30

minimum dimensions of 300 mm by 300 mm, should be used in

Notations the direct shear test of the geosynthetic/soil interface. The similar
regulation is denoted in Germany (DIN EN ISO 12957-1). This study
A shear area; uses a large scale direct shear device. The length, width, and
fds coefficient of resistance to direct sliding; thickness of the upper shear box is 450 mm  450 mm  130 mm.
RB resistance contributed by apertures; The movement of the lower shear box in the horizontal direction is
Rsoil/geogrid shear resistance of soil against the geogrid interface; controlled by a set of gears which are mobilized by an electric
a interface shear strength coefficient; motor. The vertical loading applied by a hydraulic jack is transferred
ads fraction of grid surface area that resists direct shear through the rigid reaction frame and adds on a rigid load plate
with soil; which is placed on top of the soils in the upper shear box. The
adsb fraction of transverse ribs surface area; variation of a normal load during the test is less than 2%, satisfying
b bearing resistance contribution ratio; the requirement regulated by ASTM D5321(ASTM 2002).
d angle of skin friction for soil on plane reinforcement A rigid plate is conventionally used as the loading plate in direct
surfaces; shear tests (for example, Bakeer et al., 1998; Lee and Manjunath,
4ds angle of friction for soil in direct shear; 2000). Hsieh and Hsieh (2003) found that in order to prevent
r percent open area; a concave pressure distribution at the shearing plane, a minimum
sn normal stress; of 50 mm thickness of soil in the upper shear box is recommended
ssoil internal shear strength of soil; so as to ensure the reproduction of normal pressures at the
ssoil/geogrid shear strengths of soil against the geogrid interface; shearing plane when a rigid load plate is used for applying
ssoil/geosynthetic shear strengths of soil against the geosynthetic the normal load. This device fulfills this requirement, because the
interface. thickness of soil in upper boxes is 130 mm. The system is capable of
applying a vertical force and shear force of up to 100 kN. Fig. 1
shows a frontal view of a large scale direct shear device used in this
study. The vertical force applied on the rigid plate and its vertical
geogrid and soils has been proposed by researchers through pullout displacement is measured during the tests. The horizontal move-
tests (Lopez, 2002; Alfaro and Pathak, 2005). Because a relative ment of the lower shear box and the shear force exerted during
movement between soil and geosynthetic is induced during the shearing testing are also recorded. These data are collected by using
pullout mode, the transverse ribs have been recognized as pro- two load cells and two linear variable displacement transformers
viding bearing resistance for the pullout mechanism (Jewell, 1990, (LVDT). The capacity for both load cells is 100 kN. The capacities for
1996; Bergado et al., 1993; Palmeira, 2004). On the other hand, the vertical and horizontal LVDT are 50 mm and 100 mm, respectively.
effect of transverse ribs of geogrid on direct shear resistance is not
clear. For example, Jewell et al. (1985) and Lopez (2002) state that 3. Test material
the contribution of passive earth pressure induced by the bearing
members of the geogrid is almost negligible for the direct shear The tests herein use various soils, including Ottawa sand, coarse
interaction mechanism. However, Bergado et al. (1993) state that gravel, finer gravel, and laterite. Table 1 lists the physical charac-
the apertures of the geogrid might be able to provide bearing re- teristics of each soil, while Fig. 2 shows their grain size distribution
sistance after he conducted a direct shear test, finding that the curves. The Ottawa sand, which is classified as SP according to the
shear strength of the HDPE geogrid/soil interface is greater than Unified Soil Classification System (USCS), is prepared into denser
that of soil. It is of interest to study the contribution of the trans- and looser conditions in its air-dried water content. The total unit
verse ribs to the shear strength of the soil/geogrid interface. weights for the denser and looser soil samples are, respectively,
The purpose of this study is to obtain further information of the 17.4 kN/m3 and 16 kN/m3, corresponding to relative densities (Dr)
soil/PET-yarn geogrid interface shear strength by conducting a se- of 80% and 40%. Two gravels, both classified as GP, have very uni-
ries of large scale direct shear tests. The details of these tests are form grain size distribution and an angular surface. The average
introduced thereafter. The characteristics of overall soil/PET-yarn diameters for the two gravels, which are denoted as coarser gravel
geogrid interface shear resistance are reported. The theoretical and finer gravel, are 9.5 mm and 4.75 mm, respectively. The
expression proposed by Jewell et al. (1985) for direct sliding re-
sistance between soil and geogrids is applied to predict the in-
terface shear strength. The predicted results are compared with the
test results to discuss the applicability of this theoretical model.
Finally, a model is proposed to evaluate the relative contribution of
different components including shear resistance between soil and
the surface of geogrids ribs, internal shear resistance of the soil in
the openings of geogrid, and bearing resistance provided by
transverse ribs are evaluated.

2. Test device

The size of the shearing device can influence the direct shear
test results. Generally, the boundary effect and device friction are
more significant for a smaller shear box. Ingold (1982) conducts
laboratory direct shear tests of the soil/geotextile interface by using
different sizes of shear boxes. He concludes that the friction angle
obtained from a 60 mm  60 mm shear area was 2–3 higher than
that obtained from a 300 mm  300 mm shear area. The dimension
of the shear box, as regulated by ASTM D5321(ASTM 2002), with Fig. 1. Large scale direct shear apparatus.
 C.-N. Liu et al. / Geotextiles and Geomembranes 27 (2009) 19–30 21

Table 1 weight within the shear box. The soil is compacted in three layers.
Soil characteristics The compaction of dense sand and clay is conducted by using an
Property Sand Fine gravel Coarse gravel Laterite electric vibrator and a standard Proctor hammer, respectively. The
Specific gravity 2.64 2.70 2.71 2.54 compaction of gravel and loose sand is conducted by using a man-
D10 (mm) 0.25 2.6 5.3 0.01 ual plastic hammer to hit the steel plate, which was placed on top of
D30 (mm) 0.30 2.9 6.2 0.05 the soil until reaching the target unit weight. The geosynthetic
D50 (mm) 0.36 3.2 7.4 0.07
specimen is positioned on top of the lower shear box and is
D60 (mm) 0.38 3.6 7.9 0.095
Coefficient of uniformity 1.52 1.38 1.49 9.50 clamped on the front edge of the lower shear box by an anchor
Coefficient of curvature 0.95 0.90 0.92 2.63 force of seven aligned bolts and a steel clamping block. These tests
Liquid limit – – – 53 are conducted using normal stresses of 42, 92, and 187 kPa. Normal
Plastic limit – – – 29 loading is applied on the specimen and the vertical deformation of
USCS symbol SP GP GP CH
the test specimen is then monitored. Shear loading is not applied
until the vertical deformation reaches its equilibrium. According to
maximum gravel sizes for coarser gravel and for finer gravel are ASTM D5321, a shear rate of 1 mm/min is used in this test program.
20 mm and 10 mm, respectively. This satisfies the general re- The test stops when the shear displacement reaches about 65 mm,
quirement that the ratio of the minimum size of the shear box to i.e., about 15% shear strain. The maximum shear strength during the
the maximum size of the soil particle is greater than 6. The air-dried shear process is recorded as the peak shear strength. The direct
gravels are prepared to a unit weight of 15.2 kN/m3. Research on the shear tests for soil–soil interfaces and soil/PET-yarn geotextile were
interface of geosynthetic and cohesive material is becoming quite conducted under the same normal loading and same testing pro-
popular (for example, Fourie and Fabian, 1987; Abu-Farsakh and cedures for the sake of comparison.
Coronel, 2006). The rationale behind the application of cohesive
material is of economical concern. For example, abundant laterite is 5. Selection of lower shear box for a soil/geogrid test
available in Taiwan. Therefore, the laterite which is classified as CH
is also selected as the test material. The laterite is compacted to Richards and Scott (1985) conclude that the test using a large
a 90% degree of compaction at the optimum water content (24.8%). solid block as the lower shear box is the best at replicating the
The total unit weight of the clay specimen is 17.1 kN/m3. testing results of the soil/geotextile interface. Jewell (1996) states
This study uses one geotextile and six geogrids, and they are that geotextile and geomembrane can be tested with a solid block
denoted as GT, GG3, GG4, GG6, GG7, GG8, and GG9, respectively. or soil in the lower part of the shear box while the geogrid must be
The geogrid and geotextile are woven from polyester yarns and tested by a device in which both parts of the shear test device have
coated with PVC, while GT, GG3, and GG4 are commercially avail- to be filled with soil. The set-up of the direct shearing device is not
able products of the same manufacture and the PVC coating of strictly regulated by testing standards. For example, only the
geotextile is specifically customized by the authors. For the purpose minimum size of shearing box is stated explicitly in ASTM D5321
of increasing the variation of geometric properties of the geogrid, for direct shear test of the soil–geosynthetic interface.
a fraction of the longitudinal and transverse ribs of GG3 and GG4 is Specifically, few research studies focus on the appropriate set-
removed. Here, GG6 and GG8 denote the geogrids in which half of up of direct shear testing device for the soil/geogrid interface. For
the transverse ribs of the GG4 and GG3 are removed, whereas GG7 example, the difference in the measured shear strength between
denotes the geogrid in which half of the transverse and longitudinal using different sizes of a lower shear box has not been discussed.
ribs of GG4 are removed, and GG9 indicates in which all the This study tests three different set-ups of a lower shearing box. The
transverse ribs of GG4 are removed. Table 2 lists the physical effect of different set-ups on the test results is evaluated by com-
characteristics of these geosynthetics. Fig. 3 presents a clear illus- paring the direct shear test results using different set-ups of a lower
tration of these geogrids. shear box. These set-ups include: a box with the same size
(450 mm  450 mm) that is filled with the desired soil, a box with
4. Test procedures a larger size (455 mm  584 mm) filled with the desired soil, and
a larger lower shear box filled with a solid block. These set-ups are
The soil used for the large scale direct shear testing program is symbolized as lower boxes (a), (b), and (c), respectively, (Fig. 4).
prepared by compacting the initially prepared soil to the target unit A comparison of the interface shear strength of soil against
geosynthetic is quantitatively represented by an interface shear
100 strength coefficient (a), which is the ratio of soil/geosynthetic shear
clay strength (ssoil/geosynthetic) to internal shear strength of soil (ssoil)
90
sand under the same normal stress, or in other words:
80  .
finer gravel
a ¼ ssoil=geosynthetic ðssoil Þ (1)
)

70 coarser gravel
Percent passing (

60 Fig. 5 shows the interface shear strength coefficients of soil/GG3

interfaces for different set-ups of a lower shear box. The term
50
a presented in this figure is the average value of the three a values
40 for three normal stresses (sn ¼ 42, 92, and 187 kPa). It is observed
30
that, among these set-ups, lower shear box (a) produces the
greatest interface shear strength. However, the use of a larger shear
20 box produces lower shear strength. When the lower shear box is
10 larger than the upper shear box, some soils in it are not directly
loaded by the normal loading applied through the smaller, upper
0 shear box, as shown in Fig. 4b. This soil has smaller shear strength
100 10 1 0.1 0.01 0.001 0.0001
because the normal loading for these soils are smaller. When the
Particle size (mm)
shearing is proceeding, the front edge is shearing a weaker, less
Fig. 2. Grain size distribution curves of soils. confined shear surface. Therefore, the measured shear strength is
22 C.-N. Liu et al. / Geotextiles and Geomembranes 27 (2009) 19–30

Table 2
Geosynthetic characteristics

Material Thickness of transverse Aperture Aperture Longitudinal Transverse rib Percent open Transverse ribs Total area Tulta (kN/m)
ribs (mm) length (mm) width (mm) rib width (mm) width (mm) area, r (%) surface area, adsb (%) of bearing ribs (mm2)
GT 0.96 – – – – – – – 100  50
GG3 2.2 26.6 18.5 10.1 4.6 55 14.7 9158 200  40
GG4 1.98 7 7 2.5 3.2 49 31.4 28662 100  100
GG6 1.98 17.2 7 2.5 3.2 61 15.7 14331 100  50
GG7 1.98 17.2 16.5 2.5 3.2 72 15.7 16531 50  50
GG8 2.2 57.8 18.5 10.1 4.6 60 7.4 4884 200  20
GG9 1.98 436 7 2.5 3.2 73 0.73 1303 100  0
a
Ultimate tensile strength, longitudinal  transverse.

lower. This explanation is confirmed by the observation that some mobilized within the grid opening is the friction between soil and
soil particles in the lower shear box struck through the geogrid the steel plate. Since the steel plate is smooth and there is no
apertures then piled at the front edge of the shear box (Fig. 6). particle interlocking, the friction between soil and the steel plate is
These soil particles were not confined by the upper shear box. The smaller than the friction between soil and soil particles. The test
placement of a steel plate on the lower shear box (lower shear box results reveal that when conducting the direct shear test of soil
(c)) also produces a lower shear box, because the resistance against the geogrid interface, the sizes of the upper and lower shear

Fig. 3. Geosynthetic specimens.

 C.-N. Liu et al. / Geotextiles and Geomembranes 27 (2009) 19–30 23

a upper shear box

geogrid

load cell soil

soil steel
clamp block

lower shear box

b not loaded by
upper shear box

soil

soil

Fig. 6. Soil particles are extruded out of the shear box and heaved up at the front edge
c of the lower, larger shear box.
soil

deformation versus shear displacement are shown in Fig. 8. The

steel block shear stress–shear displacement results from soil internal tests and
soil/GT are also included for direct comparison.
The results show that there is no well-defined peak shear
strength is observed for the soil/PET-yarn geogrid interface. In
Fig. 4. Different set-ups of lower shear box. general, a yield shear stress is reached at a small shear displace-
ment (usually less than 20 mm). The shear stress of the soil/PET-
yarn geogrid interface continues to increase, approaching values
boxes should be the same and they must be filled with the desired similar to the shear strength of the soil shear strength for com-
soil. paratively large shear displacement values (about 70 mm in this
study). Fig. 7 shows that the shear stress at any shear displacement
6. Large scale direct shear test results value obtained from direct shear tests on soil/geogrid is consis-
tently between the internal shear strength of soil and the shear
The results of the large-size direct shear tests are presented in stress obtained from direct shear tests on soil/GT interface. The
this section in terms of shear behavior, peak shear resistance for shear strength of soil/PET-yarn geogrid is bound between shear
different soil/PET-yarn geogrid interfaces. These typical direct shear strength of soil and interface shear strength of soil/PET-yarn
test results are discussed in this section. geotextile.
The results of vertical deformation versus shear displacement
6.1. Shear behavior of the soil/geogrid interface are shown in Fig. 8. At the initial stage, as the shear displacement is
small, the geogrid reinforced soil undergoes a vertical contraction.
To discuss the stress–strain behavior of the soil/PET-yarn geo- Following the contraction, the vertical deformation behavior then
grid interface, the results of direct shear tests on GG3 under a nor- depends on the soil type. The contraction would continue for soils
mal loading of 92 kPa are presented in Fig. 7. The results of vertical of high compressibility, such as loose sand and laterite. However,
for dense sand and for gravel, the test specimen exhibits dilatancy
with the increase of shear displacement. For the soils of expansion
1.1 behavior, the PET-yarn geogrid results in a decrease of dilatancy of
soil specimen.
1.0

6.2. Peak shear strength of soils and soil/gt interfaces

0.9
α value

Fig. 9 shows the peak shear strength for soils and the soil/geo-
0.8
textile interface. The results show an apparent cohesion in these
peak shear strengths, even for granular material (sands and
gravels). Possible explanations are in the non-linear relationship of
0.7 shear stress and in the machine friction under this stress level. The
lower box (a) lower box (b) lower box (c) interface shear strength of soil against geotextile is significantly
0.6 lower than the internal shear strength of corresponding soils.
dense sand loose sand coarse gravel clay
Fig. 10 shows the interface shear strength coefficients of soil/geo-
Fig. 5. Interface shear strength coefficient (a) of soil/GG3 interfaces for different set- textile interfaces. All the interface shear strength coefficients of
ups of lower shear box. soil/GT for different soils are less than 1. This is comparable to the
24 C.-N. Liu et al. / Geotextiles and Geomembranes 27 (2009) 19–30

90 70

80 a 60
b
70

Shear stress (kPa)

Shear stress (kPa)

50
60
50 40

40 30
30 dense sand internal loose sand internal
20
20 dense sand/GT interface loose sand/GT interface
10
10 dense sand/GG3 interface loose sand/GG3 interface
0 0
0 20 40 60 80 0 20 40 60 80
Shear displacement (mm) Shear displacement (mm)

140 120
c d
120 100

Shear stress (kPa)

Shear stress (kPa)

100
80
80
60
60
coarse gravel internal fine gravel internal
40
40
coarse gravel/GT interface fine gravel/GT interface
20 20
coarse gravel/GG3 interface fine gravel/GG3 interface

0 0
0 20 40 60 80 0 20 40 60 80
Shear displacement (mm) Shear displacement (mm)

120
e
100
Shear stress (kPa)

80

60

40 laterite internal

laterite/GT interface
20
laterite/GG3 interface
0
0 20 40 60 80
Shear displacement (mm)

Fig. 7. Stress–strain behaviors of different soil/soil, soil/GT, and soil/GG3 interfaces under a normal loading of 92 kPa, (a) dense sand, (b) loose sand, (c) coarse gravel, (d) fine gravel,
(e) laterite.

findings of many researchers (Cazzuffi et al., 1993; Bakeer et al., interface shear strength between soil and geosynthetic. However,
1998; Lee and Manjunath, 2000) in that the friction of the interface the angular-shaped gravels at the opposite sides of the 0.96-mm
between soil and geosyntehtics is smaller than the friction of the thick geotextile might be able to interlock to some extent, though
soil. This means that the shear strength of soil and the geotextile they could not penetrate through the geotextile. This interlocking
interface is less than the shear strength of the soil, indicating that effect is more significant under a higher stress level, as shown in
the soil/geotextile interface is the potential sliding surface when Fig. 10. It is suspected that the measured shear strength of gravel–
the direct shear mode is of concern. geotextile interface is composed of the friction between gravel and
In this study the average a values for PET-yarn geotextile against geotextile surface and some interlocking resistance between gravel
dense Ottawa sand, loose Ottawa sand, and laterite are similar: they particles. Therefore, the interface shear strength coefficients for the
are about 0.75. It is noticeable that the a values for PET-yarn geo- gravel–geotextile interface are significantly larger than that for
textile against coarse and fine gravels are much higher: about 0.9. sand or laterite against geotextile.
One possible explanation for this significant increase in the in-
terface shear strength coefficient for gravel is attributed to the 6.3. Peak shear strength of the soil/geogrid interface
interlocking of gravels. Since there is no aperture large enough for
direct contact between soils at the opposite sides of geotextile, the Fig. 10 shows the comparison of interface shear strength co-
shear resistance obtained from a direct shear test represents the efficients of soil against GG3 and soil against GT. It is observed that
 C.-N. Liu et al. / Geotextiles and Geomembranes 27 (2009) 19–30 25

0.4 0
a dense sand internal b

Vertical displacement (mm)

loose sand internal
Vertical displacement (mm)

0.3 -0.5
dense sand/GT interface loose sand/GT interface
dense sand/GG3 interface loose sand/GG3 interface
0.2 -1

0.1 -1.5

0 -2

-0.1 -2.5

-0.2 -3
0 20 40 60 80 0 20 40 60 80
Shear displacement (mm) Shear displacement (mm)

2.5 1
c d
Vertical displacement (mm)

Vertical displacement (mm)

2
0.5
1.5
1 0
0.5
0 -0.5

-0.5 coarse gravel internal fine gravel internal

coarse gravel/GT interface -1 fine gravel/GT interface
-1 coarse gravel/GG3 interface fine gravel/GG3 interface
-1.5 -1.5
0 20 40 60 80 0 20 40 60 80
Shear displacement (mm) Shear displacement (mm)

0
e laterite internal
Vertical displacement (mm)

-1 laterite/GT interface
laterite/GG3 interface
-2

-3

-4

-5

-6
0 20 40 60 80
Shear displacement (mm)

Fig. 8. Vertical displacement–shear displacement behaviors of different soil/soil, soil/GT, and soil/GG3 interfaces under a normal loading of 92 kPa, (a) dense sand, (b) loose sand, (c)
coarse gravel, (d) fine gravel, (e) laterite.

for each soil type and under different normal stress, the shear scale direct shear test results (Cancelli et al., 1992; Cazzuffi et al.,
strength of soil–GG3 interface is always higher than that of soil– 1993; Abu-Farsakh and Coronel, 2006) of soil/geogrid interfaces.
geotextile interface. These interface shear strength coefficients The interface shear strength coefficient obtained in this research is
range from 0.90 to 1.04. Compared to geosynthetic, GG3 is effective comparable to the results of other research studies, where the
in increasing the interface shear strength when it is placed in soil. a value for the interface of granular material (sand or gravel)
Alfaro et al. (1995) and Tatlisoz et al. (1998) state that the direct against an HDPE, PP, or PET geogrid ranges from 0.90 to 1.12. This
shear resistance of the soil/geogrid interface is composed of the soil value does not vary much with polymer type. For clay, the interface
to geosynthetic shear resistance and the soil to soil shear resistance shear strength coefficient estimated in this research is also within
within geogrid openings. As presented in Fig. 9, the internal soil the range estimated by Abu-Farsakh and Coronel (2006). It is noted
shear strength is higher than the interface shear strength of soil that the test results, especially for that of laterite/PET-yarn geogrid
against geotextile, and therefore this increase in shear strength is interfaces, might be affected by variability in the the material
mainly attributed to the interlocking of soil particles that penetrate properties and in the initial state of compaction and water content.
through geogrid apertures. In general, the interface shear strength Fig. 11 shows that the interface shear strength coefficients (a) of
coefficient for the soil/PET-yarn geogrid interface is somewhat sand/PET-yarn geogrid and laterite/PET-yarn geogrid are about 20–
higher in lower stress levels. 30% more than the a values of sand/GT and laterite/GT. However, for
Fig. 11 shows the average a values of different soil/PET-yarn gravel/PET-yarn geogrid interface, it is about only 10% more than
geogrid interfaces, where the average a value ranges from 0.89 to that of gravel/GT. It indicates that the interlocking of the soils in
1.01. It does not significantly vary with different soil/PET-yarn PET-yarn geogrid aperture is less significant for gravel particles. The
geogrid interfaces. Table 3 lists the test results and the other large explanation for this is that the particle size of gravel, especially of
26 C.-N. Liu et al. / Geotextiles and Geomembranes 27 (2009) 19–30

250 1.1
dense sand dense sand/GT GG3 GG4 GG6 GG7 GG8 GG9 GT
200 loose sand loose sand/GT
peak shear stress (kPa)

coarse gravel coarse gravel/GT 1.0

fine gravel fine gravel/GT
150
laterite laterite/GT

α value
0.9
100

50
0.8

0
0 50 100 150 200
normal stress (kPa) 0.7
dense sand loose sand coarse gravel fine gravel laterite

Fig. 9. Peak shear stresses for soils (solid symbols), and soil/geotextile interfaces (bold soil type
symbols).
Fig. 11. Average interface shear strength coefficient (a) for different interface of soil
against the geogrid.
the coarse gravel, is too large to allow these particles interlock ef-
fectively in geogrid aperture. Therefore, the shear strength of
this interface is not reliable, because its correlation coefficient (R) is
gravel/PET-yarn geogrid interface, especially for the geogrid which
only 0.07.
has small apertures (such as GG4 and GG6 in this study), is not
much more than that of gravel–geotextile interface.
Fig. 12 plots the average a value of soil/PET-yarn geogrid in- 7. Comparison with theoretical model by Jewell et al. (1985)
terface corresponding to the properties (nominal tensile strength,
aperture length, and percent open area) of the six geogrid speci- Jewell et al. (1985) pioneered the study of interaction between
mens. They are categorized according to the soil type. The best-fit soils and geogrids. In their classical work, the direct sliding re-
line for each category is also plotted to show the general trend. The sistance of soil/geogrid interface was proposed in terms of contri-
trend lines in Fig. 12a indicate that the PET-yarn geogrid of greater butions from shear between soil and the plane surface areas of the
longitudinal tensile strength has smaller interface shear strength grid, and between the soil shearing over itself in the grid apertures.
when it is against sand, while it has the opposite behavior when it This model was expressed as follows:
is against gravel or laterite. The confidence for this finding is low,
because the correlation between the longitudinal tensile strength fds tan fds ¼ ads tan d þ ð1  ads Þtan fds (2)
and a value is small. The correlation coefficient (R) for these trends where fds is the coefficient of resistance to direct sliding, 4ds is the
ranges between 0.29 and 0.68. A further study, by increasing the angle of friction for soil in direct shear, d is the angle of skin friction
variations in longitudinal tensile strength, is necessary to better for soil on plane reinforcement surfaces, ads is the fraction of grid
clarify the above-mentioned relationship. Fig. 12b and c show that surface area that resists direct shear with soil. It is noted that the
the effects of transverse tensile strength and aperture length of parameter, fds in Eq. (2) is equal to the parameter, a in Eq. (1). They
PET-yarn geogrid on the soil/geogrid interface shear strength are both denote the ratio of shear strength of soil/geogrid interface to
positive and negative, respectively. This finding holds for all soils internal shear strength of soil. The parameter fds can be expressed
(sand, gravel, and laterite). Fig. 12d shows that the PET-yarn geogrid by rearranging Eq. (2) as follows:
with a larger percent open area has lower interface shear strength
 
when it is against most soils. The tendency is less significant for tan d
dense sand-PET-yarn geogrid interface. However, the trend line for fds ¼ 1  ads 1  (3)
tan fds
that is, it can be calculated by the information of ads and the direct
shear test results of soil (4ds) and soil/plane reinforcement interface
1.1 (d). The value of ads, which represents the contact fraction of grid

1.0 Table 3
Interface shear strength coefficients of soil/geogrid estimated by conducting large
scale direct shear tests
0.9
Soil type a (polymer type) Reference
0.8 Sand 1.04 (HDPE) Cancelli et al. (1992)
α value

1.04–1.12 (PP)
1.08 (HDPE)
0.7
0.95–1.04 (HDPE) Cazzuffi et al. (1993)
dense sand/GG3 dense sand/GT
0.90–1.05 (PET) Abu-Farsakh and Coronel (2006)
0.6 loose sand/GG3 loose sand/GT 0.94 (PP)
coarse gravel/GG3 coarse gravel/GT 0.93–1.01 (PET) This research
0.5 fine gravel/GG3 fine gravel/GT Gravel 1.08–1.12 (HDPE) Cancelli et al. (1992)
1.04–1.12 (PP)
laterite/GG3 laterite/GT
0.83–0.9 (HDPE) Cazzuffi et al. (1993)
0.4 0.89–1.01 (PET) This research
0 50 100 150 200
Clay 1.0–1.2 (HDPE) Bergado et al. (1993)
normal stress (kPa) 0.82–1.15 (PET) Abu-Farsakh and Coronel (2006)
0.94–0.96 (PP)
Fig. 10. Interface shear strength coefficients (a) for different soils against geogrid (soil/
0.92–0.99 (PET) This research
GG3) and against geotextile (soil/GT).
 C.-N. Liu et al. / Geotextiles and Geomembranes 27 (2009) 19–30 27

1.05 1.05
a b
1.00 1.00

α value
α value
0.95 0.95

0.90 0.90

0.85 0.85
0 50 100 150 200 250 0 20 40 60 80 100 120
longitudinal tensile strength (kN/m) transverse tensile strength (kN/m)

1.05 1.05
c d
1.00 1.00
α value

α value
0.95 0.95

0.90 0.90

0.85 0.85
0 20 40 60 80 45 55 65 75
aperture length (mm) percent open area ( )
dense sand coarse gravel loose sand fine gravel laterite

Fig. 12. Effect of the geogrid properties on average interface shear strength coefficient (a).

surface area and soil, depends on the relative size of soil particles to between 1 and 100. The overestimation is attributed to the small
grid apertures. For fine soils, ads is equal to the ratio of the surface value of ads because in Eqs. (3) and (4), a small value of ads would
area of geogrid ribs to the total area of geogrid, that is, it is equal to result in a large value of fds.
(1-percent open area). For a coarse soil, the contact between geo- It is noted that Jewell et al. (1985) claimed that the bearing
grid and soil is limited to the top surface of transverse ribs because stresses provided by transverse ribs can be ignored because (i) the
rupture surface is above the top of grid. The ads thus reduces to the soil in the upper and lower halves of the grid apertures would shear
fraction of transverse ribs surface area. If the soil particles are in opposite directions so that bearing stresses could be balanced
coarser (similar size to grid aperture) the rupture zone would be across a transverse rib, and (ii) for granular soils, the rupture surface
fully located into the soil, then the ads reduces to 0. For soils with is above the top of grid thus the soil contained in the grid apertures
particles too large to penetrate the grid apertures, shear resistance would not displace relative to the transverse ribs. However, Bergado
might be provided only by soil particles in contact with the plane et al. (1993) state that the geogrid apertures might be able to provide
grid surface areas, the ads is equal to 1. bearing resistance under direct shear mode. The assessment of the
The model proposed by Jewell et al. (1985) is adopted as a basis bearing stresses provided by transverse ribs would be of interest.
for a rational analysis of the test data. First, the theoretical fds values Fig. 14 shows the discrepancy between model-predicted fds and test
for different soil/PET-yarn geogrid interface are calculated by using results of fds corresponding to the total area of bearing ribs for dif-
Eq. (4), which is modified from Eq. (3): ferent geogrids. The total area of bearing ribs (as listed in Table 2) is
  calculated as the product of total number of apertures in the shear
ssoil=plane reinforcement surface, thickness of transverse rib, and aperture width. A larger
fds ¼ 1  ads 1  (4)
ssoil value of total area of bearing ribs indicates a larger contribution of
bearing resistance (if this mechanism exists) would be potentially
the peak shear strengths of soil internal tests (ssoil), and peak shear
provided by the transverse ribs. Fig. 14 shows that the deficit of
strengths of soil against plane reinforcement (ssoil/plane reinforcement)
model-predicted fds increases with an increase of total area of
are obtained directly from test results. The ads value refers to rel-
bearing ribs. That is, this model which ignores the bearing stresses
ative size of soil particles to grid apertures and the geogrid char-
provided by transverse ribs tends to underestimate shear strength of
acteristics (Table 2). Fig. 13a shows the theoretical fds values versus
soil/geogrid interface more significantly for the geogrid with a larger
geogrid aperture width/D50. Its general trend compares very well
contribution of bearing resistance.
with that presented in Jewell et al. (1985). It is also noted that the
ratio value, 1, is the upper bound of fds evaluated from Eq. (4).
Fig. 13b shows the test results of fds values for different soil/PET- 8. Bearing resistance provided by transverse ribs
yarn geogrid interface. These fds values, though slightly reveal
a trend similar to that presented in Jewell et al. (1985), show a more The test results do not support the theoretical model that the
scattering distribution. Fig. 13c plots the discrepancy between resistance of the soil/geogrid interface for a direct shear mode is
model-predicted fds and test results of fds. The comparison shows attributed entirely to (i) shear resistance between soil and the
that the model-predicted fds values are smaller than the test results surface of geogrids ribs, and (ii) internal shear resistance of the soil
when a ratio of geogrid aperture width/D50 is greater than 100 or in the openings of geogrid. The other mechanism of soil re-
smaller than 1. However, the model tends to overestimate the fds inforcement interaction, bearing resistance provided by transverse
values for the interface with geogrid aperture width/D50 ranging ribs, is considered. It is of interest to study the contribution of the
28 C.-N. Liu et al. / Geotextiles and Geomembranes 27 (2009) 19–30

1.02 1.02
a b
1.00 1.00

0.98 0.98

0.96 0.96
GG3
fds

fds
0.94 GG4 0.94
GG3
GG6
0.92 0.92 GG4
GG7
GG6
0.90 GG8 0.90
GG7
GG9
0.88 0.88 GG8

αds = 1 αds = 0 αds = αds-b αds = 1-ρ GG9

0.86 0.86
0 1 10 100 1000 0 1 10 100 1000
aperture width/soil D50 aperture width/soil D50

0.10
0.08
c
(model value - test value)

0.06
0.04
0.02
Δfds

0.00
GG3
-0.02 GG4
-0.04 GG6

-0.06 GG7
GG8
-0.08
GG9
-0.10
0 1 10 100 1000
aperture width/soil D50

Fig. 13. (a) Coefficients of resistance to direct sliding for different soils evaluated from Eq. (4), (b) coefficients of resistance to direct sliding for different soils evaluated from test
results, and (c) difference between tested and predicted coefficients of resistance to direct sliding.

transverse ribs to the shear strength of the soil/geogrid interface. geogrids ribs, and (ii) internal shear resistance of the soil in the
This study proposes a simple equation to model the overall shear openings of geogrid. The last term of Eq. (5) represents (iii) bearing
resistance (Rsoil/geogrid) of the soil/geogrid interface under direct resistance provided by transverse ribs. The bearing resistance
shear mode: contribution ratio (b) is defined as the ratio of bearing resistance to
overall shear resistance of soil–geogrid interface. It can be calcu-
Rsoil=geogrid ¼ Að1  rÞssoil=geosynthetic þ Arssoil þ RB (5) lated by rearranging Eq. (5) as:
where A is the area of geogrid, r is the percent open area, and RB is h i.
the resistance contributed by transverse ribs. The first term and
b ¼ ssoil=geogrid ð1 rÞssoil=geosynthetic  rssoil ssoil=geogrid (6)
second term on the right-hand side of Eq. (5) correspondingly
Table 4 demonstrates the calculation of the bearing resistance
represent (i) shear resistance between soil and the surface of
contribution ratio of the loose sand/PET-yarn geogrid interface. The
information about the percent open area and the geosynthetic area
ratio can be obtained from Table 2, while the shear strength of loose
0.08 sand (ssoil) can be obtained from direct shear test. In addition, the
shear resistance between loose sand and the surface of the geo-
GG9 synthetic (ssand/geosynthetic) was defined in this study using the re-
0.06
Δfds (model value - test value)

sults from direct shear tests conducted on loose sand–geotextile

interfaces.
0.04 Fig. 15 shows the average bearing resistance contribution ratio
for different soil/PET-yarn geogrid interfaces. It is observed that the
0.02 GG8 bearing resistance contribution ratio also varies with soil that the
GG3 GG6 geogrid is against. The bearing resistance contribution ratio is
0.00
GG7 higher when PET-yarn geogrid is against sand or laterite soil.
However, the transverse ribs contribute less resistance when the
GG4 PET-yarn geogrid is against gravel. Also observed from Fig. 15, the
-0.02
bearing resistance contribution ratio is dependent on the aperture
pattern of PET-yarn geogrid. In general, compared to the modified
-0.04 geogrids (GG6, GG7, GG8, and GG9) those part ribs are removed, the
0 5000 10000 15000 20000 25000 30000
commercially available products (GG3 and GG4) have a larger
total area of bearing ribs
bearing resistance contribution ratio. It is noted that GG9 has the
Fig. 14. Difference between tested and predicted coefficients of resistance to direct smallest bearing resistance contribution ratio among these geogrid
sliding for geogrids of different total area of bearing ribs. specimens. Because all the transverse ribs of GG9 are removed, it
 C.-N. Liu et al. / Geotextiles and Geomembranes 27 (2009) 19–30 29

Table 4
Shear test results of loose sand/geogrid interfaces

Interface Shear strength r (%) sn ¼ 42 kPa sn ¼ 92 kPa sn ¼ 187 kPa mb sb

smax (kPa) b smax (kPa) b smax (kPa) b
Loose sand ssoil 35.82 – 65.26 – 126.78 – – –
Loose sand/GT ssoil/geosynthetic 25.59 – 51.66 – 100.47 – – –
Loose sand/GG3 ssoil/GG3 55 32.7 0.045 61.06 0.031 119.7 0.040 0.039 0.007
Loose sand/GG4 ssoil/GG4 49 33.9 0.097 65.89 0.115 125.54 0.097 0.103 0.010
Loose sand/GG6 ssoil/GG6 61 32.89 0.032 63.11 0.050 121.93 0.044 0.042 0.009
Loose sand/GG7 ssoil/GG7 72 34.09 0.033 62.94 0.024 122.87 0.028 0.028 0.005
Loose sand/GG8 ssoil/GG8 60 32.04 0.010 60.28 0.008 122.79 0.053 0.024 0.026
Loose sand/GG9 ssoil/GG9 73 33.59 0.016 60.98 0.010 115.89 0.033 0.009 0.024

verifies that no transverse ribs implies no bearing resistance because the soil interlocking is not allowed by using this set-up. The
contribution. test results reveal that a lower shear box that is filled with the test
It is noted that some calculated bearing resistance contribution soil and has the same size of the upper shear boxes is more
ratios are negative. The unreasonable results might be attributed to appropriate.
the overestimation of internal shear strength of soil (ssoil) in Eq. (6). The interface shear strength of soil/GT is smaller than soil shear
As the soil/PET-yarn geogrid interface is sheared, the interlocking of strength. It indicates that geotextile placed within the soil usually
soil particle in the geogrid openings is less effective than that as the acts as a weak interface in terms of direct sliding. The direct shear
soil is sheared. Therefore, the internal shear strength estimated resistance of the soil/geogrid interface is higher than the soil/GT.
from direct shear test of soil specimen would be greater than the The ratio of soil/PET-yarn geogrid shear strength to soil shear
internal shear resistance of the soil in the openings of geogrid. This strength ranges between 0.89 and 1.01. It does not significantly vary
overestimation phenomenon is more significant for gravel/geogrid with different types of soils. The characteristics of shear strength of
interface, as illustrated in the previous section. This explains the soil/PET-yarn geogrid interface are found to be in relation with
bearing resistance contribution ratio of the gravel/PET-yarn geogrid some geogrid properties. The geogrid that has larger transverse
interface being small. The small b value may also be attributed to tensile strength, smaller aperture length, or a smaller percent open
the overestimation of interface shear strength of soil/geosynthetic area will have larger interface shear strength.
(ssoil/geosynthetic) in Eq. (6). Specifically, the overestimation is sig- The examination of the test results shows that the shear re-
nificant for angular-shaped gravels, as described previously. These sistance of soil/PET-yarn geogrid interface under direct shear mode
overestimations of ssoil and ssoil/geosynthetic in Eq. (6) would yield an composed of the following components: (i) shear resistance be-
underestimation of b value. Therefore, the shear strength between tween soil and the surface of geogrids ribs, (ii) internal shear re-
soil against geogrid ribs (ssand/geosynthetic in Eq. (6)) is overestimated sistance of the soil in the openings of geogrid, and (iii) bearing
when the direct shear test results of sand–geotextile interfaces is resistance provided by transverse ribs. The bearing resistance is
adopted. induced when a relative movement occurs between transverse ribs
and soil particles in apertures during processing of direct shearing.
A simple model is proposed to estimate the bearing resistance
9. Conclusions
contribution ratio. In general, the transverse ribs contribute about
0–10% resistance to the overall interface shear strength of the soil/
This research has conducted a series of large scale direct shear
PET-yarn geogrid interface. This ratio is larger when the geogrid is
tests on soil/PET-yarn geogrid interfaces, where the interfaces of
against sand or against laterite, while it is smaller when it is against
four soils against six geogrids are investigated. The test results re-
gravel.
veal that the set-up of a lower shearing box is important for direct
shear tests of the soil/geogrid interface. A lower box that is larger
than the loading area is not appropriate because the soil outside the References
loading area is not confined well enough to provide lateral re-
Abu-Farsakh, M.Y., Coronel, J., 2006. Characterization of Cohesive Soil–Geosynthetic
sistance. The use of a rigid plane as a shear box is not appropriate Interaction from Large Direct Shear Test. 85th Transportation Research Board
Annual Meeting, Washington, D.C.
Alfaro, M.C., Miura, N., Bergado, D.T., 1995. Soil geogrid reinforcement interaction by
0.2 pullout and direct shear tests. Geotechnical Testing Journal 18 (2), 157–167.
Alfaro, M.C., Pathak, Y.P., 2005. Dilatant stresses at the interface of granular fills and
GG3 GG4 GG6 GG7 GG8 GG9
geogrid strip reinforcements. Geosynthetics International 12 (5), 239–252.
ASTM D5321, 2002. Standard test method for determining the coefficient of soil and
geosynthetic or geosynthetic and geosynthetic friction by the direct shear
method. ASTM Designation: D5321-02. ASTM, USA.
0.1 Bakeer, R.M., Sayed, M., Cates, P., Subramanian, R., 1998. Pullout and shear test on
geogrid reinforced lightweight aggregate. Geotextiles and Geomembranes 16
β value

(2), 119–133.
Bauer, G.E., Zhao, Y., 1993. Evaluation of shear strength and dilatancy behavior of
reinforced soil from direct shear tests. ASTM Special Technical Publication 1190,
138–157.
0.0 Bergado, D.T., Chai, J.C., Abiera, H.O., Alfaro, M.C., Balasubramaniam, A.S., 1993. In-
teraction between cohesive-frictional soil and various grid reinforcements.
Geotextiles and Geomembranes 12 (4), 327–349.
Bergado, D.T., Ramana, G.V., Sia, H.I., Varun, 2006. Evaluation of interface shear
strength of composite liner system and stability analysis for a landfill lining
system in Thailand. Geotextiles and Geomembranes 24, 371–393.
-0.1 Bernal, A., Salgado, R., Swan Jr., R.H., Lovell, C.W., 1997. Interaction between tire
dense sand loose sand coarse gravel fine gravel laterite
shreds, rubber-sand and geosynthetics. Geosynthetics International 4 (6), 623–
soil type 643.
Cancelli, A., Rimoldi, P., Togni, S., 1992. Frictional characteristics of geogrids by
Fig. 15. Average bearing resistance contribution ratio (b) for different interface of soil means of direct shear and pull-out tests. In: Proceedings of the International
against the geogrid. Symposium on Earth Reinforcement Practice, Kyushu, vol. 1, pp. 29–34.
30 C.-N. Liu et al. / Geotextiles and Geomembranes 27 (2009) 19–30

Cazzuffi, D., Picarelli, L., Ricciuti, A., Rimoldi, P., 1993. Laboratory investigations on Liu, C.N., Gilbert, R.B., 2003. Simplified method for estimating geosynthetic loads
the shear strength of geogrid reinforced soils. ASTM Special Technical Publi- in landfill liner side slopes during filling. Geosynthetics International 10 (1),
cation 1190, 119–137. 24–33.
El Sawwaf, M., 2007. Behavior of strip footing on geogrid-reinforced sand over a soft Lopez, M.L., 2002. Soil geosynthetic interaction. In: Shukla, S.K. (Ed.), Geosynthetics
clay slope. Geotextiles and Geomembranes 25, 50–60. and Their Applications. Thomas Telford Publishing, Thomas Telford Ltd, London.
Fishman, K.L., Pal, S., 1994. Further study of geomembrane/cohesive soil interface Mahmood, A., Zakaria, N., Ahmad, F., 2000. Studies on geotextile/soil interface shear
shear behavior. Geotextiles and Geomembranes 13 (9), 571–590. behavior. Electronic Journal of Geotechnical Engineering 5.
Fleming, I.R., Sharma, J.S., Jogi, M.B., 2006. Shear strength of geomembrane–soil Palmeira, E.M., 2004. Bearing force mobilisation in pull-out tests on geogrids.
interface under unsaturated conditions. Geotextiles and Geomembranes 24, Geotextiles and Geomembrances 22 (6), 481–509.
274–284. Palmeira, E.M., Viana, H.N.L., 2003. Effectiveness of geogrids as inclusions in cover
Fourie, A.B., Fabian, K.J., 1987. Laboratory determination of clay-geotextile in- soils of slopes of waste disposal areas. Geotextiles and Geomembranes 21 (5),
teraction. Geotextiles and Geomembranes 6 (4), 275–294. 317–337.
Ho, Y.H., 2001. Engineering Properties of Tire Shreds and Cohesive Soil Mix. Master Richards, E.A., Scott, J.D., 1985. Soil Geotextile Frictional Properties. Second Cana-
thesis, National Pingtong University of Technology (in Chinese). dian Symposium on Geotextiles and Geomenbranes, Edmonton. 13–24.
Hsieh, C., Hsieh, M.W., 2003. Load plate rigidity and scale effects on the frictional Tatlisoz, N., Edil, T.B., Benson, C.H., 1998. Interaction between reinforcing
behavior of sand/geomembrane interfaces. Geotextiles and Geomembranes 21 geosynthetics and soil–tire chip mixtures. Journal of Geotechnical and
(1), 25–47. Geoenvironmental Engineering 124 (11), 1109–1119.
Ingold, T.S., 1982. Some observations on the laboratory measurement of soil– Vukelic, A., Szavits-Nossan, A., Kvasnicka, P., 2008. The influence of bentonite ex-
geotextile bond. Geotechnical Testing Journal 5 (3), 57–67. trusion on shear strength of GCL/geomembrane interface. Geotextiles and Ge-
Jarret, P.M., Bathurst, R.J., 1985. Frictional Development at a Gravel Geosynthetic omembranes 26, 82–90.
Peat Interface. Proceedings of the Second Canadian Symposium of Geotextiles Wu, W., Wick, H., Ferstl, F., Aschauer, F., 2008. A tilt table device for testing
and Geomembranes, Edmonton. 1–6. geosynthetic interfaces in centrifuge. Geotextiles and Geomembranes 26 (1),
Jewell, R.A., 1990. Reinforcement bond capacity. Geotechnique 40 (3), 513–518. 31–38.
Jewell, R.A., 1996. Soil Reinforcement with Geotextiles. Thomas Telford, London. Xenaki, V.C., Athanasopoulos, G.A., 2001. Experimental investigation of the in-
Jewell, R.A., Milligan, G.W.E., Sarsby, R.W., Dubois, D.,1985. Interaction between Soil and teraction mechanism at the EPS geofoam–sand interface by direct shear testing.
Geogrids. Proc., Conference on Polymer Grid Reinforcement, London, pp. 18–29. Geosynthetics International 8 (6), 471–499.
Lee, K.M., Manjunath, V.R., 2000. Soil–geotextile interface friction by direct shear Zabielska-Adamska, K., 2006. Shear strength parameters of compacted fly ash–
test. Canadian Geotechnical Journal 37 (1), 238–252. HDPE geomembrane interfaces. Geotextiles and Geomembranes 24, 91–102.