Clay Reinforcement Using Geogrid Embedded In Thin Layers

of Sand
M.R. Abdi1,*, S. A. Sadrnejad2 and M.A. Arjomand3
Received: October 2008 Accepted: September 2009

Abstract: Large size direct shear tests (i.e.300 x 300mm) were conducted to investigate the interaction between clay
Downloaded from ijce.iust.ac.ir at 20:52 IRDT on Friday March 22nd 2019

reinforced with geogrids embedded in thin layers of sand. Test results for the clay, sand, clay-sand, clay-geogrid, sand-
geogrid and clay-sand-geogrid are discussed. Thin layers of sand including 4, 6, 8, 10, 12 and 14mm were used to
increase the interaction between the clay and the geogrids. Effects of sand layer thickness, normal pressure and
transverse geogrid members were studied. All tests were conducted on saturated clay under unconsolidated-undrained
(UU) conditions. Test results indicate that provision of thin layers of high strength sand on both sides of the geogrid
is very effective in improving the strength and deformation behaviour of reinforced clay under UU loading conditions.
Using geogrids embedded in thin layers of sand not only can improve performance of clay backfills but also it can
provide drainage paths preventing pore water pressure generations. For the soil, geogrid and the normal pressures
used, an optimum sand layer thickness of 10mm was determined which proved to be independent of the magnitude of
the normal pressure used. Effect of sand layers combined with the geogrid reinforcement increased with increase in
normal pressures. The improvement was more pronounced at higher normal pressures. Total shear resistance provided
by the geogrids with transverse members removed was approximately 10% lower than shear resistance of geogrids
with transverse members.

Keywords: geogrid, reinforcement, clay, sand, interaction, direct shear test

1. Introduction higher post-construction creep potential are the

main concerns expressed about the use of
Soil reinforcement is a highly attractive cohesive soils in soil reinforcement [2]. These
alternative for embankment and retaining wall concerns may represent unrealistic restrictions in
projects because of the economic benefits it actual practice, where many highway
offers in relation to conventional retaining embankments are constructed of compacted
structures. The rapid acceptance of soil clays. One potential solution for reinforcing
reinforcement can be attributed to a number of marginal soils is the use of permeable
factors, including low cost, aesthetics, reliability, geosynthetics that function not only as
simple construction techniques, and the ability to reinforcement but also as lateral drains (Zornberg
adapt to different site conditions. However, these and Mitchell [3]). This would eliminate the need
economic benefits have often been limited by the for expensive backfill and reduce transportation
availability of good-quality granular material. and structural costs as well as improving
These materials have been the preferred backfill performance of compacted clay. Undoubtedly,
material due to their high strength and ability to substantial cost savings and new soil
prevent development of pore water pressure [1]. reinforcement applications would result if
Build up of pore water pressure, lower cohesive soil as well as industrial and mine
frictional strength and compactibility as well as wastes that would otherwise require disposal
could be used in reinforced soil construction.
* Corresponding Author: Email: [email protected] Interestingly, the first geotextile-reinforced
wall constructed in 1971 by the French Highway
1 Assistant Professor, Faculty of Civil Engineering, Administration in Rouen, used poorly draining
KNT University, Tehran, Iran.
cohesive soil as backfill material. The purpose of
2 Professor, Faculty of Civil Engineering, KNT
University, Tehran, Iran. this structure was to test its stability and to verify
3 PhD Student, Faculty of Civil Engineering, KNT the magnitude of deformations caused by soil-
University, Tehran, Iran. geotextile interaction (Puig et al. [4]). Although

224 International Journal of Civil Engineerng. Vol. 7, No. 4, December 2009

 marginal soils have been successfully reinforced, to 40%, while reinforcements with low
failures have also occurred mainly because pore transmissivity can decrease the undrained
water pressure generations were not correctly strength by a similar magnitude. The use of non-
addressed during design. woven geotextiles for reinforcing a near-
saturated silty clay was also evaluated by Ling
2. Literature review and Tatsuoka [8] using a plane strain device. The
reinforcement effect, in terms of strength and
Downloaded from ijce.iust.ac.ir at 20:52 IRDT on Friday March 22nd 2019

Reinforced soil derives its superior behaviour stiffness, was reported to be greater in drained
due to the stress transfer from the soil to the compared to undrained tests. At small strain
reinforcement at the interface. Adequate soil- levels, excess pore water pressures adversely
reinforcement interaction has to be ensured to affected the stress-strain response of the
enable such a mechanism to take place. In the reinforced soil samples tested under undrained
case of clay soil, the interfacial strength is low conditions. In the drained tests, tensile stresses
resulting in an early failure of the interface before were mobilized in the geotextile ensuring a
the full strength of reinforcement can be positive reinforcement effect.
mobilised. Thus the strength of reinforcement Shear failure at the interface may happen due
may be largely underutilized due to failure of the to the high shear stresses near the reinforcement
interface. A number of experimental studies using as seen in experimental observations by Jewell
triaxial tests have been conducted to develop an and Wroth [9], Milligan et al. [10] and Sridharan
understanding of the interaction between et al. [11. They have found that the shear stresses
cohesive soil and different reinforcement are highest around the reinforcement and
systems. Results of drained and undrained decrease rapidly away from the reinforcement.
compression tests on normally consolidated clay Hence, when poor quality backfill is used for
samples reinforced with several disks cut out of construction, it is advantageous to place thin
aluminum foil or porous plastics were presented layers of high-strength granular soil around the
by Ingold and Miller [5] and Ingold [6]. Results reinforcement to resist these high shear stresses
showed reductions in undrained axisymetric near the interface. This method of construction
compressive strength of more than 50% relative called "sandwich technique" will improve the
to unreinforced samples. The premature failure of stress transfer mechanism because of the
the specimen was attributed to pore-water interface properties. Sridharan et al. [11] also
pressures induced in the reinforced specimen reported significant improvement in the pullout
which greatly exceeded those measured in a capacity of geogrids embedded in weak soils
similar unreinforced specimen. Decreasing the because of sandwich layers. Based on laboratory
spacing between the horizontal layers of tests on model retaining walls employing
reinforcement resulted in an increase in both the sandwich layers, Sreekantiah and Unnikrishnan
drained shear strength and the secant modulus of [12] also reported improvement in the response
the reinforced sample. Based on the radiographic of retaining walls. Unnikrishnan et al. [13] by
investigation, the strength enhancement was conducting UU triaxial compression tests,
attributed, as in the case of sand reinforcement, to reported improvement in strength and
radial strain control arising from shear stress deformation behaviour of reinforced clay soils
mobilized on the soil-reinforcement interface. under static and cyclic loading.
Fabian and Fourie [7] presented the results of A promising approach for design of reinforced
undrained triaxial tests performed on silty clay marginal soils is to promote lateral drainage in
samples reinforced with various geosynthetics combination with soil reinforcement. This maybe
having different in-plane transmissivities, achieved by using geocomposites with in-plane
including woven geotextiles, nonwovens, and drainage capabilities or thin layers of granular
geogrids. Their results showed that soil in combination with the geo-synthetic
reinforcements with high transmissivity can reinforcements. This design approach may even
increase the undrained strength of the clay by up lead to the elimination of external drainage

225
 requirements. The potential use of permeable used to investigate the effect of confining pressure
inclusions to reinforce poorly draining soils is for reinforced and unreinforced samples for a given
well documented by Zornberg and Mitchell [3] sand layer thickness. Samples were tested under
and Tatsuoka and Yamauchi [14]. Although there unconsolidated-undrained (UU) condition to
is already strong experimental evidence that simulate the behaviour of clays subjected to quick
permeable inclusions can effectively reinforce loading immediately after construction. Clay soils
poorly draining backfills, but there is no general will have the least shear strength under this type of
Downloaded from ijce.iust.ac.ir at 20:52 IRDT on Friday March 22nd 2019

design methodology for reinforced soil structures loading and hence it has been used to examine the
built with cohesive soils. influence of sand and sand-geogrid layers on its
strength and performance. A constant horizontal
3. Scope of Research displacement rate of 1mm/min was used throughout
the tests as recommended by ASTM D:5321 and in
This paper investigates the effects of embedding order to be consistent with previous investigations.
reinforcements in thin layers of granular material The geogrid specimens were large enough to
within a clay soil (i.e. sandwich technique) using completely cover the apparatus plan area.
large direct shear tests. A large number of tests were Complementary direct shear tests were also
done by varying thickness of granular layer, conducted with the transverse members of the
magnitude of normal pressures and transverse geogrids removed in order to evaluate their
members of the geogrids removed. contribution to the overall shear resistance.

4. Testing program 5. Materials Used

5.1 Soils
Laboratory tests were conducted using 300 x 300
mm direct shear tests on samples of clay, sand, clay- Kaolinite was used as the clay soil and Firozkoh
sand and clay-sand-geogrids. A single horizontal sand used for casting was selected as the granular
layer of geogrid was used as reinforcement with material. The index properties of the clay and sand
variable sand layer thickness. Sand layer were determined according to the appropriate
thicknesses of 2, 3, 4, 5, 6 and 7 mm on either side ASTM standards and are summarized in Table1.
of the reinforcement were used to investigate its According to Unified Soil Classification System
effects on the peak shear stress. Normal stress (USCS) the clay was classified as CL (inorganic
combinations of 25, 50 and 75 kN/m2 were also clay of low plasticity) and the sand as SW (well

Table 1. Clay and sand characteristics

Soil Description / Property ASTM Standard Value
Liquid limit ASTM D4318 53%
Plastic limit ASTM D4318 33%
Plasticity index ASTM D4318 20%
Clay Optimum moisture content (Proctor compaction ) ASTM D698 23%
Maximum dry density (Proctor compaction) ASTM D698 1550(kg/m3)
Cohesion (at optimum moisture content) ASTM D3080 23.2(kN/m2)
Angle of friction (at optimum moisture content) ASTM D3080 10 
D10 ASTM D2826 0.4
D30 " 1.3
D60 " 2.5
Uniformity coefficient (cu) " 6.25
Sand Coefficient of curvature (cc) " 1.69
Optimum moisture content (Proctor compaction ) ASTM D698 4%
Maximum dry density (Proctor compaction) ASTM D698 1600(kg/m3)
Angle of friction (at optimum moisture content) ASTM D3080 33.7 

226 International Journal of Civil Engineerng. Vol. 7, No. 4, December 2009

 Table 2. Interfacial properties of the geogrid and its characteristics

Geogrid Interfacial 1 Interfacial 1 Interfacial C

with sand (Deg.) with clay (Deg.) with clay (kN/m2)
Miragrid 50/25-30 36.1 7.9 28.8

Description Symbol/Value
Raw material PET
Downloaded from ijce.iust.ac.ir at 20:52 IRDT on Friday March 22nd 2019

Coating PVC
Ultimate longitudinal tensile strength (Tult) 50 (kN/m)
Ultimate lateral tensile strength (Tult) 25 (kN/m)
Longitudinal strain at Tult 11%
Lateral strain at Tult 13%
Ratio of reinforcement shear relative to total shear area ( , ds ) 10%

graded sand). In the case of geogrid embedded in sand

layers, after placing and compacting clay in three
5.2 Geogrid equal layers, predetermined amounts of moist
sand were weighted, poured, spread and
Miragrid 50/25-30 geogrid, a uniaxial polymer compacted to fill the lower half of the shear box.
normally used for reinforced soil walls and steep The geogrid was then laid and the top half of the
slopes, was employed as reinforcement. Geogrid shear box positioned and tightly secured.
specimens 300x300mm covering the whole plan Subsequently the same amount of moist sand was
area of the shear box were used. The interface poured, spread and compacted and the remaining
shear strength properties obtained by conducting volume of the top shear box filled with three
modified direct shear tests together with the equal layers of moist clay. Test set up including
geogrid characteristics provided by the producer the position of the sand layer and the geogrid is
are listed in Table 2. schematically shown in Figure 1.
After completion of each test, samples were
6. Sample preparation taken for density and moisture content
determination. A maximum variation of 3% was
Samples were prepared by static compaction of observed which was considered acceptable. The
soil to a predetermined dry density and moisture procedures for specimen preparation and testing
content. Accurately measured quantities of dry were standardized to achieve repeatability in the
powdered clay and water corresponding to
maximum dry density (MDD) and optimum
moisture content (OMC) were thoroughly mixed
and kept in plastic containers for 24 hours for
uniform moisture distribution. Initially the lower
half of the shear box was filled with three equal
layers of clay and lightly tamped with the specially
adopted tamping device. Subsequently geogrid
specimen covering the whole surface of the
sample was horizontally laid and clamped to the
inner face of the shear box. Then the upper half of
the shear box was placed, secured and filled with
moist soil in the same manner. After imposing the
desired normal pressure and setting up gauges for
measuring vertical and horizontal displacements Fig.1. Cross – section of the shear box showing the
as well as the shear force, testing commenced. position of sand layer and the reinforcement

M.R. Abdi, S. A. Sadrnejad and M.A. Arjomand 227

 test results. All the initial tests were repeated resistance at particle contact points. Bergado et
until consistent results were obtained. al. [15] investigating the interaction between
cohesive-frictional soil and various grid
7. Results and discussions reinforcements concluded that owing to the
7.1. Reinforced and unreinforced clay influence of the apertures on the grid
reinforcements, the shear resistance between the
Results of direct shear tests conducted on grids and the soil in a direct shear test can be
Downloaded from ijce.iust.ac.ir at 20:52 IRDT on Friday March 22nd 2019

samples of reinforced and unreinforced clay are equal to or larger than the shear resistance
shown in Figure 2. Variations of shear stress between soil and soil. Touahamia et al. [16]
versus shear displacement for both the reinforced investigating the shear strength of reinforced-
and unreinforced clay show an increasing trend recycled material also reported that overall
by increase in normal pressures. Slope of the restraint provided by the geogrid is determined
curves is significant at the early ages of shearing by the particle size and particle grading.
and it reduces by further shear displacement. The shear failure at the interface may happen
Although reinforced clay samples consistently due to high shear stresses developed near the
showed slightly higher shear stresses compared reinforcement as observed experimentally by
to unreinforced samples subjected to the same Jewell and Wroth [9], Milligan et al. [10] and
normal pressure, but the increase was not Sridharan et al. [11].They have found that the
significant. This meant that failure occurred in shear stresses are the highest around the
the clay by way of full mobilization of cohesive reinforcement and decrease rapidly away from
strength. This behaviour is an indication that the the reinforcement. Hence, when poor quality
clay-geogrid interface resistance is low which backfill is used for construction, it is
results in premature failure of the interface before advantageous to place thin layers of high strength
the full strength of the reinforcement can be granular soil around the reinforcement to resist
mobilised. these high shear stresses near the interface. This
Thus, the strength of reinforcement may be will probably improve the stress transfer
largely underutilized due to the failure of the mechanism because of the better interface
interface. Another possible reason for such properties. Alfaro et al. [17] also showed the
behaviour can be the mesh size of the geogrid in mobilization of direct shear resistance to be away
comparison with the clay particle size. The from the interface into the soil and that mobilized
geogrid can restrain particle movement and shear strain in the direct shear test to be very
therefore increase the mobilised frictional uniform along the soil-geogrid interface.
50
50
Clay , Clay-Reinforcement Clay , Clay-Reinforcement

40
ın=75kPa 40
Shear Strength (kPa)

ın=50kPa
Shear Stress (kPa)

30
ın=25kPa
30

20

20
ƕ
10 Clay-Reinforcement , C=28.8 kPa ,Ɏ=7.9
ƕ
Clay , C=23.2 kPa ,Ɏ=10
Clay- Reinforcement Clay-Clay
10
0 0 10 20 30 40 50 60 70
0 5 10 15 20 Normal Stress (kPa)
Shear Displacement (mm)

Fig. 2. Shear stress versus shear displacement curves for Fig. 3. Failure envelopes for clay and clay – reinforcement
reinforced and unreinforced clay samples

228 International Journal of Civil Engineerng. Vol. 7, No. 4, December 2009

 Figure 3 shows the failure envelopes for the displacement-softening behaviour and finally
reinforced and the unreinforced clay samples. It reach a steady state. The display of hardening and
can be observed that geogrid resulted in slightly then softening behaviour is attributed to the
reducing the angle of friction of the clay and amount of particle rearrangements that can occur.
increasing its cohesion. The overall effect of At low normal pressure (i.e. 25kN/m2), the
geogrid inclusion has been to slightly increase the reinforced and the unreinforced samples show
shear strength of the clay. The linear envelopes are approximately the same maximum shear and
Downloaded from ijce.iust.ac.ir at 20:52 IRDT on Friday March 22nd 2019

an indication of the absence of particle ultimate strength. By increasing normal pressure

interlocking and subsequently no dilation was to 75kN/m2, the reinforced samples show higher
observed. As the shear resistance from the clay- maximum shear and ultimate strengths. These
geogrid test is close to that of the clay test, the changes are because the particles on the interface
results seem to agree with the suggestion of Jewell surface are less likely to be rearranged during
et al. [18]. They suggested that for a sandy gravel shearing if the shear stress is not large enough to
type of backfill and grid reinforcement, the direct overcome the internal friction. The stress-
sliding mechanism will be such that the rupture hardening behaviour, especially at high normal
zone is forced away from the interface into the stress, may result from the plowing of angular
soil. In this case, the direct shear resistance of the particles into geogrid material surface also
soil-reinforcement interface would be equal to the reported by Han [19]. No dilatancy as such was
full shear resistance of the soil. observed, which confirms that the stress induced
during shearing at the interface is not large
7.2. Reinforced and unreinforced sand enough to disturb the whole specimen. Under
these conditions, sliding dominates the shear
Figure 4 shows the results of direct shear tests resistance as shown by Dove [20]. Results also
conducted on reinforced and unreinforced sand. show that horizontal displacement at failure is
It can clearly be seen that the curves do not enhanced with an increase in normal pressure
display an obvious peak and by increasing the (i.e. confining pressure) in a way that it is 1-2 mm
normal pressures the shear strength of both the for 25kN/m2 and 4-5 mm at 75kN/m2. These
reinforced and unreinforced samples increase. changes have also been reported by Haeri et al.
The shear strength of all samples initially [21] from their investigation into the effect of
increased significantly with only a limited shear geotextile reinforcement on the mechanical
displacement. After reaching their maximum behaviour of sand.
shear strength samples exhibit post-peak The peak shear stress – normal stress
80
sand , sand-Reinforcement 70
70 Sand , Sand-Reinforcement
60
60 ın=75kPa
50
Shear Stress (kPa)

Shear Strength (kPa)

50

ın=50kPa 40
40

30
30
ın=25kPa 20
20
º
10 Sand , C=12.1 kPa , Ɏ=33.7
º
10 Sand-Reinforcement ,C=12 kPa , Ɏ=36.1
Sand-Reinforcement Sand-Sand 0
0 0 10 20 30 40 50 60 70
0 5 10 15 20 Normal Stress (kPa)
Shear Displacement(mm)

Fig. 4. Shear stress versus shear displacement curves for Fig. 5. Failure envelopes for sand and sand – reinforcement
reinforced and unreinforced sand samples

M.R. Abdi, S. A. Sadrnejad and M.A. Arjomand 229

 envelopes for the reinforced as well as the then continues gradually until it reaches a
unreinforced sand are shown in Figure 5. Failure maximum at a shear displacement varying
envelopes for the peak shear stresses are linear between 2 to 5mm depending on the sand layer
which indicates the absence or very little thickness. It can be seen that only the clay sample
dilatancy. The small adhesion intercepts obtained reinforced with sand layer thickness of 14mm
are attributed to the open apertures in the displays a distinctive maximum shear stress
geogrids which permitted soil-to-soil contact and which reduces with further shear displacement.
Downloaded from ijce.iust.ac.ir at 20:52 IRDT on Friday March 22nd 2019

adhesion to exist because of suction in the soil. Other samples do not display a distinctive peak
This cohesion is reflected as adhesion in the and their maximum shear stresses coincide with
geogrid-soil interface, unlike other geosynthetics. their ultimate shear stresses (i.e. steady state
Goodhue et al. [22] attributed the small adhesion condition) displaying a plastic behaviour.
observed to be caused by matric suction at soil- The points of maximum shear stress displayed
to-soil contact and machine friction. Suction by the samples seem to shift to the left by
causes an increase in adhesion at the interface increasing the sand layer thickness and are
between two porous materials but not at the reached at smaller shear displacements. For
interface between soil and geosynthetic. example, for the clay soil reinforced with 4mm
Athanasopoluos et al. [23] also reported the and 14mm sand layers, shear displacements
development of adhesion which they considered corresponding to the maximum shear stresses are
negligible for practical applications. 5mm and 3mm respectively. Examination of
specimens with sand layers after the tests
7.3. Clay reinforced with thin layers of sand revealed that sand had penetrated a little into the
clay and had established a good bond at the
Changes in shear stress versus shear interface. As mentioned earlier, the shear stress in
displacement for clay samples reinforced with the soil reduces as the distance from
thin sand layers of varying thickness are shown in reinforcement increases. This reduced shear
Figure 6. It can be observed that the inclusion of stress at some distance from the reinforcement
sand layers significantly improves the shear can be resisted easily by the sand-clay interface.
strength of the clay soil. The improvement The results are a clear indication of the
increases with increase in sand layer thickness. effectiveness of including thin sand layers for
The shear stresses increase substantially during improving clay soil performance. Inclusion of
the early parts of the tests (i.e. 1-2mm shear sand layers apart from improving the
displacement) with samples displaying a
hardening behaviour. The increase in shear stress Clay-Sand
60

70
Clay -Sand (ın=75kPa) 50
60
Shear Strength (kPa)

40
50
Shear Stress (kPa)

40
30

30 0 mm thick sand layer

`
4mm thick sand layer Clay , C=23.2 kPa , Ɏ=10º
20
6mm thicksand layer t=4mm , C=14.5 kPa , Ɏ=15.5º
20 t=6mm , C=15.9 kPa , Ɏ=16.8º
8mm thick sand layer
t=8mm , C=16.3 kPa , Ɏ=17.3º
10mm thick sand layer
10 t=10mm , C=18.5 kPa, Ɏ=19.9º
10 12mm thick sand layer
t=12mm , C=17.4 kPa ,Ɏ=25.8º
14mm thick sand layer
t=14mm , C=17.5 kPa ,Ɏ=28º
Sand Sand , C=12.1 kPa, Ɏ=33.7º
0
0
0 2 4 6 8 10 12 14 16
0 20 40 60
Shear Displacement (mm) Normal Stress (kPa)

Fig. 6. Shear stress versus shear displacement curves for Fig. 7. Failure envelopes for clay, sand and clay – sand
clay reinforced with thin layers of sand samples

230 International Journal of Civil Engineerng. Vol. 7, No. 4, December 2009

 performance of clay backfills, perhaps more be useful in understanding the progressive global
importantly, can help drain and prevent pore mobilization of direct-shear interaction resistance
water pressure build up. that will likely occur in actual cases.
Failure envelopes for the clay samples By increasing sand layer thickness around the
reinforced with different sand layer thickness are geogrid to 10, 12 and 14mm, gradually
shown in Figure 7. Results clearly show the distinctive maximum and ultimate shear stresses
significant improvement in clay shear strength can be observed. Unlike the clay samples which
Downloaded from ijce.iust.ac.ir at 20:52 IRDT on Friday March 22nd 2019

parameters by the inclusion of thin layers of sand. showed increase in shear stress with sand layer
Taking clay shear strength parameters as the thickness (i.e. Figure 6), inclusion of geogrid
base, it can be observed the cohesion of the increases the shear stress only up to 10mm sand
composites decrease and their angles of friction layer thickness. Further increasing the sand layer
increase. For example, for the clay reinforced thickness to 12 and 14mm resulted in lowering
with 14mm sand layer, cohesion changes from the maximum shear stress. The results clearly
23.2 to 17.5 kPa, a reduction of 24.6%, and the show that the full soil-reinforcement interface
angle of friction increases from 10 to 28 degree, capacity has been mobilised even with thin layers
showing an increase of 180%. of sand and further increase of sand layer
thickness does not lead to improved performance
7.4. Clay–thin layers of sand-geogrid of the composite. This means that there is an
optimum sand layer thickness for achieving the
The results of direct shear tests on samples of maximum shear stress. After the tests all the
clay-sand-geogrid with different sand layer geogrids remained intact which is mainly
thicknesses using the whole plan area of (300x attributed to the fact that geogrid's modulus and
300mm) the apparatus are shown on Figure 8. strength is much higher than the surrounding soil.
Results show that provision of sand layers around This is in accordance with observations of
the geogrid reinforcement significantly improves Unnikrishnan et al. [13] and Gray and Al-Refeai
the strength of the clay soil. The shear stresses [24]. Unnikrishnan et al. [13] have also stated,
initially increase sharply and become more the fact that none of the geogrids tested ruptured
gradual with further shear displacement. Clay during the tests indicates that failure occurred
samples reinforced with 4, 6 and 8mm sand mainly by pullout.
layers around the geogrids do not display peak Shear stress versus normal stress envelopes for
shear stresses and did not reach a steady state the clay-sand-geogrid composites are shown in
condition by the end of the tests. This behaviour Figure 9. It can be observed that embedding
indicated a progressive type of failure which can

70 Clay-Sand - Reinforcement (ın=75kPa)

Clay - Sand - Reinforcement
60

60
Shear Strength(kPa)

50
Shear Stress (kPa)

40
40

30 0 mm thick sand layer t=0 mm , C=28.8 kPa , Ɏ=7.9º

4mm thick sand layer t=4mm , C=20.8 kPa , Ɏ=23º
6mm thick sand layer t=6mm , C=23.5 kPa , Ɏ=24.1º
20 20
8mm thick sand layer
t=8mm , C=23.6 kPa , Ɏ=26.4º
10mm thick sand layer
t=10mm , C=24.5 kPa , Ɏ=28.4º
12mm thick sand layer
10 t=12mm , C=22.3 kPa , Ɏ=26.4º
14mm thick sand layer
sand - geogrid t=14mm , C=18.9 kPa , Ɏ=25.3º
sand- geogrid,C=12 kPa , Ɏ=36.1º
0
0
0 5 10 15 20
Shear Displacement (mm) 0 20 40 60
Normal Stress (kPa)

Fig. 8. Shear stress versus shear displacement curves for Fig. 9. Failure envelopes for clay – reinforcement, sand-
clay reinforced with thin layers of sand and reinforcement reinforcement and Clay – sand - reinforcement samples

M.R. Abdi, S. A. Sadrnejad and M.A. Arjomand 231

 geogrid in thin layers of sand increases the shear pressures. The change is more pronounced at
resistance of the composite. The combined use of higher normal pressures increasing at a faster rate.
sand and geogrid as compared to using only sand This behaviour is attributed to the greater
for reinforcing clay resulted in mobilizing the confining effects provided by the geogrid. Results
maximum shear resistance at a smaller sand layer clearly show that increasing sand layer thickness
thickness (i.e. 10mm instead of 14mm). up to 10mm increases maximum shear stress and
Compared to reinforced clay (i.e. C=28.8kPa, further increase in sand layer thickness causes a
Downloaded from ijce.iust.ac.ir at 20:52 IRDT on Friday March 22nd 2019

φ = 7.9o ), the shear strength parameters of the reduction in maximum shear stress. The optimum
clay-10mm sand-geogrid system changes to sand layer thickness (i.e. 10mm) seems to be
C=24.5kPa φ = 28.4o and which are very close to independent of the normal pressure used. These
the parameters produced by the clay-14mm sand results are in contrast to the observations reported
system (i.e. C=17.5kPa, and φ = 28o ). The by Unnikrishnan et al. [13]. They reported that the
combined effects of sand-geogrid in reinforcing relative increase in the additional confining stress
clay has resulted in reducing cohesion by 15% induced by the woven geotextile reinforcement is
and increasing the angle of friction by 259% higher at lower confining pressures. At higher
which is a substantial improvement. This is a normal pressures (i.e. 158kPa), beyond a sand
clear indication that for a particular soil, layer thickness of 8mm, the maximum shear stress
reinforcement, loading condition and normal did not increase appreciably whereas at lower
pressure, an optimum sand layer thickness exists normal pressures the increase continued up to
which mobilizes the maximum shear strength. 15mm. They also concluded that the optimum
The provision of thicker sand layers will not lead thickness of sand layer depends on the operative
to further improvement in the performance of the range of stresses in the soil.
system also reported by Unnikrishnan et al. [13]
investigating the behaviour of reinforced clay 8. Bond strength
under monotonic and cyclic loading.
The bond coefficient between the soil and
7.5. Effect of normal pressure reinforcement is defined as the ratio of the
resistance between soil and reinforcement to the
To study the effects of normal pressures, results resistance between soil and soil. For the
are presented as maximum shear stress versus soil/reinforcement direct – shear – interaction
thickness of sand layer in Figure 10. Results show mechanism, the resistance between soil and soil
that the combined effect of sand layers and geogrid is the direct shear resistance of the soil with the
reinforcement increases with increase in normal same shear area as that of the soil/reinforcement
interface. Bond coefficient is the parameter that
Clay - Sand-Reinforcement
ın=75kPa 1.5
65

ın=50kPa
Max. shear stress (kPa)

55 1.3
Bond Coefficient

45
1.1
ın=25kPa

35
0.9 Clay -Sand- Reinforcement
Clay - Reinforcement
25 Sand - Reinforcement
0 4 6 8 10 12 14
Thickness of sand layers(mm) 0.7
20 40 60 80
Fig. 10. Maximum shear stress versus thickness of sand Normal Stress (kPa)
layers for clay reinforced with thin layers of sand and
reinforcement Fig. 11. Bond coefficient versus normal stress

232 International Journal of Civil Engineerng. Vol. 7, No. 4, December 2009

 expresses the efficiency of the grid reinforcement geogrid, several tests were conducted with these
for providing shear resistance. Figure 11 shows members removed. All the samples had the same
the variation of the bond coefficient as function number of longitudinal members as the geogrid
of normal stress for the clay, clay-sand and clay- with transverse members. A comparison of the
sand-geogrid systems. The corresponding values total resistance of the reinforcement with and
vary between 1.0 and 1.4. It is indicated that the without transverse members was made and the
shear resistance between clay and the geogrid results are shown in Figures 12 and 13. The shape
Downloaded from ijce.iust.ac.ir at 20:52 IRDT on Friday March 22nd 2019

was higher than that of the clay soil. In a situation of the shear stress – shear displacement curves
where the location of the shear surface is for both sets of samples are similar. Results show
constrained to pass along the soil-reinforcement that removal of the transverse members slightly
interface, this is possibly due to the influence of reduces the shear resistance between the soil and
the apertures on the geogrid, which may provide the reinforcement. It was found that the direct
some amount of bearing resistance during shear. shear resistance of the geogrid without transverse
However, it is difficult to measure the bearing members was approximately 90% of the direct
effect of the apertures on the geogrid shear resistance of the geogrid with transverse
quantitatively. In real situations, the shear plane members. Pullout tests conducted by Bergado et
will pass through the plane with lowest al. [15] on Tensar geogrids and bamboo grids
resistance, so that the bond coefficient cannot with transverse members removed respectively
exceed unity. This would indicate that the showed pullout resistances equal to 90-100% and
soil/geogrid reinforcement can provide the same 80-90% of the grids with transverse members.
shear strength as the soil itself. The range of bond They reported that the total resistance of the grids
coefficients determined in the present study is with and without transverse members is very
slightly wider than the range of 1.0 to 1.2 close which they attributed to the small spacing
reported by Bergado et al. [15] whom between the longitudinal grid members and three
investigated the interaction between cohesive- dimensional effects.
frictional soil and various grid reinforcements.
This difference is attributed to the different grids 10. Conclusions
used.
A large number of 300x300mm direct shear
9. Effect of transverse members tests were carried out to investigate the behaviour
of reinforced clays with geogrids encapsulated in
To investigate the effects of passive resistance thin sand layers (i.e. sandwich technique). It was
provided by the transverse members of the observed that using thin layers of sand to cover

Clay- Sand - Reinforcement

clay -sand - Reinforcement

60
60 ın=75kPa
Shear Strength (kPa)
Shear Stress (kPa)

ın=50kPa
40 40

ın=25kPa

20
20

with transverse members , C=24.5kPa , Ɏ=28.4º

with transvers member without transeverse member

without transverse members , C=15.6kPa , Ɏ=30.6º
0
0 5 10 15 20 0
Shear Displacement(mm)
0 20 40 60
Normal Stress (kPa)
Fig. 12. Shear stress versus shear displacement curves for
clay – sand – reinforcement with and without transverse Fig. 13. Failure envelopes for clay – sand - reinforcement
members samples with and without transverse members

M.R. Abdi, S. A. Sadrnejad and M.A. Arjomand 233

 the geogrids significantly improves the response [3] Zornberg, J.G. and Mitchell, J.K.: 1994,
of clay soils through interfacial enhancement. Reinforced soil structures with poorly drained
The improvement is the result of more effective backfills. Part I: Reinforcement interactions and
interlocking of sand within the geogrid openings. functions,Geosynthetics International,Vol.1,No.
Owing to the influence of the apertures on the 2,103-148.
geogrid, the shear resistance between the geogrid
and the soil can be equal to or larger than the [4] Puig, J., Blivet, J.C. and Pasquet, P.: 1977,
Downloaded from ijce.iust.ac.ir at 20:52 IRDT on Friday March 22nd 2019

shear resistance of the soil itself. Remblami arme' avec un textile synthe'tique,
For the soil, geogrid and normal pressures Proceedings of the international Conference on
used, an optimum sand layer thickness was the use of Fabrics in Geotechnics, Paris, France,
determined which resulted in most improvement. 85-90 (In French).
The provision of thicker sand layers did not lead
to further improvement in the behaviour of the [5] Ingold, T.S. and Miller, K.S.: 1982, The
composite system. Combined effects of sand performance of impermeable and permeable
layers and geogrid reinforcement increased with reinforcement in clay subject to undrained
increase in normal pressures. The change was loading, Quarterly Journal Of Engineering
more pronounced at higher normal pressures. By Geology, Vol. 15, 201-208.
increasing sand layer thickness up to 10mm,
maximum shear stresses displayed by samples [6] Ingold, T.S.: 1983, A laboratory investigation of
increased and thicker sand layers caused grid reinforcements in clay, Geotechnical
reduction in maximum shear stresses. The Testing Journal, ASTM, Vol, 16, No. 3, 112-
optimum sand layer thickness (i.e. 10mm) 119.
seemed to be independent of the normal pressure
used. [7] Fabian, K.J., and Fourie, A.B.: 1986,
Shear resistance provided by the geogrids Performance of geotextile reinforced clay
without the transverse members was determined samples in undrained triaxial tests, Geotextiles
to be approximately 10% less than the shear and Geomembranes 4, 53-63.
resistance of geogrids with the transverse
members. [8] Ling, H.I., and Tatsuoka, F.: 1993, Laboratory
evaluation of nonwoven geotextiles for
11. Acknowledgement reinforcing on-site soil, Proc. of Geosynthetics
93, Vol. 2, Vancouver, Canada, 533-546.
The kind co-operation and technical support
provided by the Faculties of Civil Engineering at [9] Jewell, R.A. and Wroth,C.P.:1987, Direct shear
Khaje-Nasir-Toosi and Shahid Rajaei tests on reinforced sand, Geotechnique 37,53-
Universities is gratefully acknowledged. 68.

References [10] Milligan, G.W.E., Earl, R.F., and Bush, D.I.:

1990, Observation of photo-elastic pullout tests
[1] Elias, V., and Christopher, B.B.: 1996, on geotextile and grids, Proceeding of the
Mechanically stabilized earth walls and Fourth Inter. Conf. on Geotextiles,
reinforced soil slopes, design and construction Geomemebranes and Related Products, Hague,
guidelines, Federal Highway Administration, Vol. 2, 747-751.
FHWA-Sa-96-071.
[11] Sridharan, A., Murthy, S., Bindumadhava, B.R.,
[2] Mitchell, J.K.: 1981, Soil improvement: State of And Revansiddappa, K.: 1991, Technique for
the Art, Proc. of Tenth Inter. Conf. on Soil using fine-grained soil in reinforced earth, Jour.
Mechanics and Found. Eng., Stockholm, Of Geotechnical Eng. Division, ASCE 117,
Sweden, Vol. 4, 509-565. 1174-1190.

234 International Journal of Civil Engineerng. Vol. 7, No. 4, December 2009

 [12] Sreekantiah, H.R., and Unnikrishnan, N.: 1992, and geogrids, Proc. Symposium on Polymer
Behaviour of geotextile under pullout, Proc. of Grid Reinforcement in Civil Engineering, 18-
the Indian Geotechnical Conference, Calcutta, 30.
215-228.
[19] Han, J.: 1997, An experimental and analytical
[13] Unnikrishnan, N, Rajagopal, K., and study of fiber reinforced polymer piles in sand
Krishnaswamy, N.R.: 2002, Behavior of and pile-sand interactions, PhD Dissertation,
Downloaded from ijce.iust.ac.ir at 20:52 IRDT on Friday March 22nd 2019

reinforced clay under monotonic and cyclic Georgia Institute of Technology, Atlanta.
loading, Geotextile and Geomembrane, 20. ,
117-133. [20] Dove, J.E.: 1996, Particle – geomembrane
interface strength behaviour as influenced by
[14] Tatsuoka, F. and Yamauchi, H.: 1990, A surface topography, PhD Dissertation, Georgia
reinforcing method for steep clay slopes using a Institute of Technology, Atlanta.
non-woven geotextile, Geotextile and
Geomembranes, Vol.4, No.3-4. [21] Haeri, S.M., Noorzad, R. and Oskoorouchi,
A.M.: 2000, Effect of geotextile reinforcement
[15] Bergado, D.T., Chai, J.C., Abiera, H.O, Alfaro, on the mechanical behavior of sand, Geotextile
M.C., and Balasubramaniam, A.S.: 1993, and Geomembranes, Vol. 18, Issue 6, 385-402.
Interaction between cohesive-frictional soil and
various reinforcements, Geotextile and [22] Goodhue, M.J., Edil, T.B., and Benson, C.H.:
Geomembranes 12, 327-349. 2001, Interaction of foundry sands with
geosynthetics, Jour. of Geotechnical and
[16] Touahamia, M., Sivakumar, V. and McKelvey, Geoenvironmental Eng., April, 353-362.
D.: 2002, Shear strength of reinforced-recycled
material, Construction and Building Materials, [23] Athanasopoluos, G.A., Katsas, C.E., Ioannidis,
No.16, 331-339. A.A., and Pelekis, P.C.: 2002, Evaluation of
sand-geotextile interface friction angle by a
[17] Alfaro, M.C., Miura, N., and Bergado, D.T.: modified 300 300mm direct shear box,
1995, Soil-geogrid reinforcement interaction by Geosynthetics- 7th ICG- Delmas.
pullout and direct shear tests, Geotechnical
Testing Journal, Vol. 18, No. 2, 157-167. [24] Gray, D.H., and Al-Refeai, T.: 1986, Behavior
of fabric vs. fiber-reinforced sand, Journal of
[18] Jewell, R.A., Milligan, G.W.E., Sarsby, R.W., Geotechnical Foundations, ASCE, 94 (SM1),
and Dubois, d.: 1984, Interaction between soil 271-290.

M.R. Abdi, S. A. Sadrnejad and M.A. Arjomand 235