Studies on geosynthetic-reinforced road

pavement structures
K. Rajagopal*, S. Chandramouli, Anusha Parayil and K. Iniyan
Many of the pavement structures fail well before their design life owing to the poor quality of
construction materials, inadequate compaction, inadequate preparation of the subgrade, over-
loading, etc. Two options are available to improve the longevity of the pavement. The first option is by
increasing the thickness of different pavement layers and the other option is by increasing the rigidity
of the layers within the system so as to reduce the stresses transferred to the lower layers. Of these
two methods it has been widely observed that increasing the strength and rigidity of the pavement
layers is a more efficient method to lower the stresses on the pavement layers thereby increasing the
life of the pavement.
In the present research work, the improvement in the strength and stiffness of the subbase layer in a
flexible pavement system through the use of geosynthetic layers was investigated by conducting
field plate load tests and a series of laboratory plate load tests. The improvement in the strength of
the pavement is reflected by the increase in modulus of the section reinforced with geosynthetic
layers. This paper will describe the field and laboratory tests, interpretation of the data from these
tests, and the application of this data for design of flexible pavements and their economic analyses.
Keywords: Geosynthetics, Flexible pavements, Geogrids, Geocells, Geotextiles

This paper is part of a special issue on geosynthetics

Introduction The performance of the geocells as surface confinement

layers and as reinforcement layers has been reported by
The performance of highway pavements is governed by several researchers in the past. Bathurst and Rajagopal
the strength and stiffness of the pavement layers. The cost (1993) and Rajagopal et al. (1999) have reported the
and duration of construction are dependent on the strength and stiffness behavior of soils confined in single
availability of aggregate materials for construction. geocell and multiple geocell pockets. Madhavi Latha et al.
Scarcity of natural resources often delays the projects or (2008, 2009) have reported the benefit of using geocells as
escalates the costs due to large lead distances from the basal reinforcement layers for embankments constructed
borrow areas. Hence, it is essential to look at alternatives on soft foundation soils. It was reported that the factor of
to achieve improved quality of pavements using new safety of the slopes can be increased significantly because
materials and reduced usage of natural materials, Giroud of the interception of the slip surface by the geocell layer.
and Han (2004). This paper reports on the studies of the Unni (2010) and Chandramouli (2011) have reported the
performance of geosynthetic-reinforced flexible pave- construction of geocell-reinforced unpaved road pave-
ments. Different types of geosynthetics like planar ments and their performance on different types of
(geogrids and geotextiles) and three dimensional (geocells) subgrade layers. Iniyan (2012) has reported the use of
can be employed for strengthening the pavement bases. geogrids for construction of pavements and the improve-
The geocells are three-dimensional honeycomb geosyn- ment of the strength of the pavement sections. Based on
thetic products that provide all round confinement to the the higher modulus obtained with geosynthetic reinforce-
soils. The geocell-confined soil acts like a semi-rigid mat in ment layers, he has discussed that the pavement thickness
distributing the surface loads over a wide area of the can be reduced while maintaining the same level of design
foundation soil. parameters.
Geocells and geosynthetics are adopted in several road
and ground stabilization projects across the globe. Han
et al. (2008, 2010, 2011) have described the influence of the
Department of Civil Engineering, IIT Madras, Chennai 600 036, India infill material and the stiffness on the performance of
*Corresponding author, email [email protected] geocells in pavements. Unni (2010), Chandramouli (2011),

ß 2014 W. S. Maney & Son Ltd

Received 13 December 2013; accepted 10 January 2014 International Journal of
DOI 10.1179/1939787914Y.0000000042 Geotechnical Engineering 2014 VOL 8 NO 3 287
Rajagopal et al. Geosynthetic-reinforced road pavement structures

Iniyan (2012), and Parayil (2013) have reported the

performance of geosynthetic-reinforced flexible pavement
sections. They reported that the stresses below the
reinforced layers are two to three times lesser than the
surface stresses. Emersleben and Meyer (2010) conducted
test box analysis for 200 mm height geocell filled with sand
above very soft clay and observed that the stresses can be
reduced by 30 and 36% depending on the applied load.
The load carrying capacity could be improved up to 1?5
times due to the reinforcement of dry sand with geocells.
Shin et al. (2010) conducted field plate load tests on
reinforced and unreinforced subgrade soil and analyzed by
using finite element software. They gave the subgrade
improvement factor of 2. Bush et al. (1990) carried out
research on geocell-reinforced embankment and con-
cluded that the 1 m high geocell with local soil infill will
have 33% lesser settlements after 4 years when compared
to systems with horizontal layers of reinforcement.
Further, the cost savings of more than 31% were reported
for geocell-treated constructions. 1 Cross-section of the pavement section at Govind Dairy
The current paper investigates the performance of the Factory
reinforced flexible pavements under monotonic and
repeated loads. The granular subbase and sand materials
were obtained from a highway construction site near opening dimensions are approximately 2106250 mm. The
Chennai. All the index tests were performed to character- tensile strength of the geocell material in strip tension test
ize these materials. Field and laboratory plate load tests was found to be 0?25 kN (ASTM D638-2003) and the peel
were conducted on the flexible pavements. The pressure– strength of the weld is 0?2 kN from ASTM D6392-99
settlement data was used to back-calculate the elastic standard tensile strength tests. There was no change of
modulus of the geosynthetic-reinforced pavement layers. dimensions when pieces of the geocell were exposed to
100uC temperature in an oven for 1 h duration (ASTM
D1204).
Field studies on geosynthetic flexible The construction at the site proceeded by excavating the
pavements soil to the required depth. The hydrated lime was spread
on the soil and mixed by a tractor with a plow attachment.
Geocell-reinforced pavements
The lime was mixed in proportion of 4% by weight. This
The internal access roads at Govind Dairy Factory in percentage was decided based on prior experience with
Phaltan, Maharashtra required frequent repairs. The similar soils in India. The addition of 3% lime was found
foundation soil is typically black cotton soil, which to drastically reduce the plasticity index values by as much
undergoes severe swelling and shrinking. The properties as 50%. Hence, 4% lime mixing was recommended to
of this soil are given in Table 1. The roads are typical account for some loss during and after the construction.
unpaved roads with thick layers of water-bound Macadam After the compaction of the lime-treated soil and the
(WBM) and granular subbase (GSB) materials. Nearly granular subbase layers were completed, the geocell layer
200 m long stretch of this road was treated with 150 mm was spread on the road section and held in place by use of
thick geocell layer on an experimental basis to study the stakes driven into ground at 485 mm c/c spacing. The
performance improvement.
geocell pockets were filled with GSB material by a tipper
Based on the soil properties and the traffic data, the truck and spread using a dozer. Care was taken to make
following designed section of pavement as shown in Fig. 1 sure that the vehicles do not pass directly on unfilled
was used for reconstructing the road using geocell
geocell section. After the geocell pockets were filled with
reinforcement. The geocell pockets were filled with GSB
GSB material and 75 mm cover material was placed, the
materials. Water-bound Macadam layers were not used
entire section was compacted using normal 10-ton vibro
within this stretch of road where geocell was used as a
roller passes. The photographs in Figs. 2–5 illustrate the
reinforcement layer. The bottom most layer was treated
construction procedure adopted at the site.
with 4% lime (hydrated lime) in order to stabilize the
expansive foundation soil. Addition of 3% lime itself was
found to reduce the plasticity index substantially. Hence,
slightly higher percentage of 4% lime addition was Table 1 Properties of subgrade soil
recommended in order to account for any losses during California bearing ratio (CBR) 4%
installation and service life. Swell index 150%
The geocell is 150 mm high and made of a polymeric Liquid limit 60%
Plastic limit 25%
alloy. The thickness of the geocell walls is approximately
Shrinkage limit 8%
1?2 mm. The c/c weld distance is 330 mm and the pocket

288 International Journal of Geotechnical Engineering 2014 VOL 8 NO 3

 Rajagopal et al. Geosynthetic-reinforced road pavement structures

2 Mixing of the lime by a tractor

5 Compaction by a vibratory roller

The construction of the pavement took place in March

2010. The unreinforced pavement sections were also
constructed in the same manner without the geocell
reinforcement at the subbase level. This stretch of road
was provided with layers of WBM material, which is more
expensive compared to GSB material. In place of the
400 mm thick GSB layer, two layers of 200 mm thick
WBM layers were provided. Over this 150 mm thick layer
of GSB was provided. The thickness of the lime-treated
soil is the same as shown in Fig. 1.
The performance of the geocell-reinforced pavement
and the adjacent unreinforced sections were monitored for
their performance. The year 2010 was characterized by
unusually heavy rainfall in that region. The unreinforced
pavement had undergone severe rutting and had to be
reconstructed at least three times by dumping of aggregate
3 Geocell layer spread over the road section and recompaction during the period March–December
2010. The photographs of the unreinforced and the

4 Filling the geocell pockets with a dozer 6 Settlements observed in the unreinforced section

International Journal of Geotechnical Engineering 2014 VOL 8 NO 3 289

Rajagopal et al. Geosynthetic-reinforced road pavement structures

IS 1888–1988 in the geocell-treated area and the unrein-

forced areas. Two tests were performed in geocell-
reinforced area (R-1 and R-2) and two tests were
performed in unreinforced pavement area (UR-1 and
UR-2). One test was performed at subgrade level for
comparative purposes. All the tests were performed at
surface level after scraping the top 50 mm of pavement
material. The observed pressure–settlement responses are
shown in Fig. 8. The pressure–settlement responses of
both the tests performed in geocell-treated pavement were
very close to each other. The responses from the
unreinforced areas are also very similar. The unreinforced
pavement area was repaired several times by dumping
large size stones, which are in excess of 200–300 mm in
size. The test plate may have been located inadvertently
7 Uniform surface observed in the geocell section
over a large size stone in UR-2, which gave a stiffer
response than the tests performed in geocell-treated area.

Geogrid-reinforced pavements
reinforced pavement sections after 9 months of traffic Two sections of a highway under construction near
loading are shown in Figs. 6–7 for comparison purposes. Chennai were reinforced with two different types of
The unreinforced pavement section had undergone geogrids (flexible and stiff). Both geogrids are biaxial type
severe surface depressions as indicated by the arrows. having tensile strengths in the same range. The flexible
On the other hand, the geocell-reinforced road section had geogrid was a knitted polyester geogrid having tensile
maintained a uniform surface. The trucks had to negotiate strength of 100 kN m21 at a strain of 10%. The stiff
the unreinforced sections at a slow speed while they could geogrid was an extruded and welded polyester geogrid,
maintain their normal speed in the reinforced sections. which is much heavier and stiffer. The stiff geogrid had a
This difference in the performance clearly shows the tensile strength of nearly 130 kN m21 at a strain of 6%.
improvement in the performance of the flexible pavements The geogrid layers were placed within the subbase layer of
with geocell reinforcement. The performance of the the pavement at a depth of 200 mm below the surface.
geocell-treated section is very good even 3 years after its These two trial stretches were constructed next to each
installation in 2010. The client has decided to reconstruct other so that the subgrade soil is similar. The subgrade soil
the entire stretch of their internal roads using geocell at this site has a soaked CBR value of 8%. The pressure–
reinforcement. settlement data of test performed at subgrade level is also
In order to differentiate the strength of the pavement shown for comparison. Two tests were performed on top
sections, plate load tests were performed at the site as per of 200 mm thick granular subbase material without any

8 Pressure–settlement response from different field plate load tests

290 International Journal of Geotechnical Engineering 2014 VOL 8 NO 3

 Rajagopal et al. Geosynthetic-reinforced road pavement structures

9 Pressure–settlement response with two different types of geogrids

geogrid reinforcement and one plate load test was having plan dimensions of 120061200 mm and height of
performed within each of the two different types of 1200 mm. The steel tank was made of 8 mm thick steel
geogrids. The observed pressure–settlement data is shown plate and supported around its periphery by a frame made
in Fig. 9. It could be seen that the response with the of 75 mm L-angles. The diameter of the plate used for
geogrid layers is stronger compared to the unreinforced loading was 150 mm and the thickness was 30 mm. The
sections. The observed pressures with stiffer geogrid are load was applied to the plate through a hydraulic jack
higher than those with flexible geogrid owing to the higher fixed against a rigid self-reacting frame.
modulus of the stiff geogrid. All the tests were performed by constructing 500 mm
thick subgrade made up of dry sand placed at 60% relative
Discussion on field test results density. The sand is uniformly graded coarse sand with
It is interesting to note the following points from the field angular particles having the properties listed in Table 2.
test results (Figs. 8 and 9): The sand was placed at 60% relative density in all the tests.
1. The improvement of the response with geogrid It was closely monitored in all the tests by collecting soil
reinforcement layers is practically nil at low settle-
ment levels.
2. The marginal improvement with geogrid layers is
Table 3 Properties of the granular subbase (GSB) material
seen only at large settlement levels.
3. The improvement with geocell reinforcement is Material type GSB (granular subbase)
substantial even at low settlements.
4. The improvement at large settlements is substantial Dry density 2.19 g cc21
Optimum 5.31%
with geocell reinforcement.
moisture content
CBR value 23%
Laboratory studies in loose state
Particle size .4.75 mm 71.6%
The laboratory plate load tests were performed under distribution 4.75–2.36 mm 5.2%
monotonic and cyclic conditions to examine the benefit of 2.36–0.075 mm 18.5%
the reinforcement layers under repeated loadings. The ,0.075 mm 4.7%
laboratory tests were performed in a rigid steel tank Shear strength properties c50, w565u

Table 2 Properties of sand used in the laboratory tests

Specific Effective Coefficient of Coefficient of Maximum unit Minimum unit

gravity size (D10)/mm uniformity Cu curvature Cc weight/kN m23 weight/kN m23 Friction angle/u

2.66 0.55 2.9 1.6 16.1 14.5 46u

International Journal of Geotechnical Engineering 2014 VOL 8 NO 3 291

Rajagopal et al. Geosynthetic-reinforced road pavement structures

10 Schematic of the load testing in laboratory

samples using steel containers buried in the soil at different compaction to achieve 60% relative density of the soil. The
depths. relative density achieved was carefully monitored by
The base course was constructed using GSB material of placing steel cups of known volume and collecting soil
coarse particles having the properties listed in Table 3. samples after each layer of compaction.
The GSB was placed in the test tank in a relatively loose The geosynthetic reinforcement (geotextile, geogrid, or
state by hand packing. The laboratory CBR value for this geocell) was placed on the prepared subgrade. While
packing was nearly 23%. placing the geocell, each pocket was stretched to at least
The base layers were reinforced with geocell or geogrid 180 mm size and the corners were filled with aggregate to
layers. In all the tests, the sand and GSB material were keep it in position. Then the other pockets of the geocell
separated by a woven geotextile layer. The properties of were carefully filled with GSB by hand packing. In case of
different geosynthetic materials are listed tables 4 and 5. geocell an additional 10 mm of GSB was provided over the
Three different heights of geocell layers 50, 100, and weld to prevent direct load application on to the geocell.
150 mm were used in this testing program. The schematic representation of testing is shown in
The geogrid used is a polyester knitted type geogrid Fig. 10. The settlements were measured using electronic
having the properties listed in Table 5. The tensile linear voltage differential transducers (LVDTs) and the
properties of the geogrid layer were determined as per load was measured using an electronic load cell. The load
ASTM D 6647-11. The strain corresponding to the was applied in small increments. The load increment was
ultimate strength is 10%. The same geocell used for field maintained constant until the settlements under that
tests was also used in the laboratory tests. increment cease to increase. Then the next increment of
The total thickness of the subgrade of 500 mm was load was applied. Since the subgrade is sand, settlements
prepared in four layers of thickness 150, 150, 100, and have become constant within 1 min of load application.
100 mm. The quantity of the sand required in each The test was terminated when the base could not continue
layer was pre-weighed and placed in layers with light to hold any further load increment.

Table 4 Properties of geotextile layer used in the laboratory tests

Geotextile type Mass per unit area/g m22 Nominal thickness/mm Ultimate tensile strength/kN m21 Maximum Elongation/%

Woven 240 0.5 115 14

Table 5 Properties of geogrid used in the tests

Nominal thickness/mm Ultimate tensile strength/kN m21

Mass per unit area/g m22 MD CD MD CD

918 2.1 1.67 110 23

MD: machine direction; CD: cross-machine direction.

292 International Journal of Geotechnical Engineering 2014 VOL 8 NO 3

 Rajagopal et al. Geosynthetic-reinforced road pavement structures

11 Pressure–settlement data from two tests performed on 100 mm thick granular subbase (GSB) layer

The consistency of the test results is very important measured pressure–settlement data, the ultimate bearing
when results from different tests are to be compared. capacities for different test configurations are obtained as
Several trials were made initially until achieving consis- shown in Table 6. The ultimate pressures were obtained by
tent pressure–settlement data. Typical pressure–settlement drawing tangent lines to the initial and the final parts of
data from two different tests with the same configuration the pressure–settlement responses. The ultimate pressure is
is shown in Fig. 11. It could be seen that the two pressure– given by the intersection point of these two tangent lines as
settlement responses are almost identical. Hence, it could illustrated in Fig. 12.
be concluded that the results from this testing program After each test was completed, the GSB layer was
are repeatable. Typical pressure–settlement responses for carefully removed to examine the settlement bowl at the
150 mm thick GSB layer and different forms of geosyn- surface of the subgrade. After the subgrade was exposed,
thetic reinforcement are illustrated in Fig. 12. From the the radial extent of the settlement bowl and the settlements

12 Pressure–settlement response with different types of reinforcements for 150 mm thick granular subbase (GSB) base layer:
UR: unreinforced; GT: geotextile; GGzGT: geogridzgeotextile; GCzGT: geocellzgeotextile

International Journal of Geotechnical Engineering 2014 VOL 8 NO 3 293

Rajagopal et al. Geosynthetic-reinforced road pavement structures

It is seen that for geosynthetic-reinforced cases, the

diameter of the settlement bowl is much bigger leading to
lesser pressures transmitted to the subgrade soil. The
settlement bowl for the geocell reinforcement is found to
be biggest among all the reinforced cases.
The pavements are subjected to number of load
repetitions during their service life. The response of the
pavements under repeated loading may be much different
from that under static loading. Hence, several tests were
performed by subjecting the pavement sections to repeated
loading (cyclic loading). The loading was applied as one-
way cyclic loading. The load was increased to a pre-set
maximum value and reduced to a lower value (1 kN) and
increased once again. The same configurations used for
static load tests were also used for the cyclic plate load
tests.
The maximum load in the cyclic load tests was decided
based on the ultimate bearing pressure observed in the
13 Screen shot of the computer screen during cyclic load static load tests. The pavement section was prepared in the
tests same manner as in the monotonic plate load test. After
the section is set up, two dial gages were fixed on top of the
load plate. The cyclic loading was applied through a servo-
within the bowl were carefully measured. In general, it was controlled hydraulic actuator. The system is operated
noticed that the loads are distributed over a larger area of using a computer program that acquires the load and
the subgrade due to the provision of the geosynthetic deformation data. The cyclic loading was applied at a
layer. The tests were performed with a single layer of frequency of 0?7 Hz for 50 000 cycles. This loading
geotextile, combination of geotextile and geogrid, and a frequency was reported by Pokharel et al. (2010) and
geocell underlain by a geotextile, which acts as a separator Han et al. (2011) as that representative of the traffic
layer. The ratio between the diameters of the settlement loading. The pressures applied in these tests are the
bowl and the loading plate gives an idea of the pressures ultimate pressures observed for respective thickness of the
transmitted to the subgrade. These ratios for different test unreinforced sections viz. 610, 725, and 930 kPa for 50,
configurations are given in Table 7. 100, and 150 mm thick GSB layers, respectively.

Table 6 Ultimate bearing pressures for different test configurations

Ultimate bearing capacity/kPa

Base course thickness/mm Unreinforced Geotextile Geogridzgeotextile Geocellzgeotextile

50 610 630 920 1560

100 725 1320 1510 1980
150 930 1370 1530 2045

Table 7 Ratio of the diameter of settlement bowl and the loading plate

Thickness of base layer/mm Unreinforced Geotextile Geogridzgeotextile Geocellzgeotextile

50 2.15 2.2 2.25 2.43

100 2.33 2.53 2.59 2.77
150 2.56 2.62 2.74 2.87

Table 8 Maximum settlements at the end of 50 000 load cycles

Maximum settlement/mm

Thickness of base course/mm Applied pressure/kPa Unreinforced Geotextile Geogridzgeotextile Geocellzgeotextile

50 610 23.17 14.48 10.49 8.08

100 725 39.18 24.5 19.88 15.2
150 930 46.12 28.61 24.6 20.9

294 International Journal of Geotechnical Engineering 2014 VOL 8 NO 3

 Rajagopal et al. Geosynthetic-reinforced road pavement structures

14 Variation of settlement with number of load repetitions

A screen shot of the cyclic loading program is shown in Knowing the values of the applied load P, radius of the
Fig. 13. The green line shows the load calculated as per the loading plate a, and the settlement s, the modulus of
applied load and the red line shows the load measured by subgrade reaction can be calculated from the data as
the load cell in the actuator. It is seen that both are very
P
close to each other. M~ (1)
Typical data on the variation of settlement with p a2 s
different types of geosynthetic reinforcement layers for
50 mm thick GSB layer are shown in Table 8 and Fig. 14. The results showed that initially the modulus of the
From the above measured data, the modulus of subgrade is very high and as the settlement increased with
subgrade reaction can be estimated as the ratio of the number of cycles, the modulus value decreased and
applied pressure, the base area, and the settlement. reached a constant value toward the end of 20 000 cycles.

15 Variation of the modulus of subgrade reaction with number of load cycles

Table 9 Modulus of subgrade reaction obtained at the end of 50 000 cycles

Modulus of subgrade
reaction/kN m23 Thickness of base course/mm Unreinforced Geotextile Geogridzgeotextile Geocellzgeotextile

50 3011 4813 5802 7031

100 1941 3224 3866 4650
150 1861 3150 3868 4489

International Journal of Geotechnical Engineering 2014 VOL 8 NO 3 295

Rajagopal et al. Geosynthetic-reinforced road pavement structures

16 Variation of modulus improve factor with the number of cycles for geocell case

The typical variation of the subgrade modulus with modulus improvement is higher for geocell-reinforced
different number of cycles is shown in Fig. 15 for cases and its value is higher for larger heights of the base
100 mm thick GSB layer. layers.
The modulus values obtained at the end of 50 000 cycles The modulus improvement factor was found to vary
are shown in Table 9.The geosynthetic reinforcement with the number of cycles. The modulus improvement
improved the modulus of subgrade reaction of the test factor was also found to be high initially and as the number
section for all thicknesses of the base course layer. The of load cycles increased, the value was found to decrease as
reduction in modulus value with increased base layer illustrated in Fig. 16. This decrease is mainly attributed to
thickness is due to the compressions within the base layer the loose state of the base layer, which has undergone
as it was placed in loose state. compressions due to repeated load applications.
The modulus improvement factor is defined as the ratio
of elastic modulus of the reinforced base course to that of Back calculation of elastic modulus of the
unreinforced base course of the same thickness. The
modulus of subgrade reaction values at the end of 50 000
system
cycles are used to calculate the modulus improvement Elastic finite element analyses were performed to estimate
factors for illustration purposes. The improvement factors the equivalent elastic modulus of the unreinforced and
thus obtained are given in Table 10. It can be seen that the reinforced pavement systems. The finite element analyses

Table 10 Modulus improvement factors from cyclic plate load test data

Thickness of base course/mm Geotextile Geogridzgeotextile Geocellzgeotextile

Modulus improvement factor 50 1.6 1.93 2.34

100 1.67 1.99 2.40
150 1.70 2.08 2.41

Table 11 Modulus of elasticity of granular subbase (GSB) from finite element analyses

Modulus of elasticity/kPa

Thickness/mm Unreinforced Geotextile Geogridzgeotextile Geocellzgeotextile

50 25 500 44 100 52 000 70 500

100 29 100 50 500 61 100 84 500
150 33 000 58 000 74 000 98 000

Table 12 Modulus improvement factors from finite element results

Improvement factors

Thickness/mm Geotextile Geogridzgeotextile Geocellzgeotextile

50 1.73 2.04 2.76

100 1.74 2.10 2.90
150 1.76 2.24 2.97

296 International Journal of Geotechnical Engineering 2014 VOL 8 NO 3

 Rajagopal et al. Geosynthetic-reinforced road pavement structures

were performed by using axi-symmetric model and 15- improvement in the pavement performance as compared
node triangular elements. The rough, rigid footing was to the planar type products like geotextiles and geogrids.
simulated by applying uniform settlements at the nodes The modulus improvement factors obtained from both
corresponding to the footing and restraining their lateral monotonically applied tests and the cyclic load tests are
deformations. close to each other. Significant improvement is observed
The equivalent elastic modulus was determined by trial for all types of geosynthetic reinforcement systems. The
and error by matching the finite element calculated footing modulus improvement factor is seen to be higher for
pressure at 1?5 mm settlement (equal to 1% of plate monotonic loading as compared to the cyclic load tests.
diameter) with the measured pressures in the laboratory This could be due to the loose packing of the GSB layer in
tests. It is assumed that the response of the system at a laboratory tests leading to continuous compressions
small settlement equal to 1% of the footing diameter is within the GSB layer under cyclic loading.
within the elastic limit. The elastic modulus value of the
continuum was varied until the estimated pressure Ackowledgement
matches with the experimentally measured values. The
results of the monotonic plate load tests were used for The authors are grateful to the support from M/s PRS
these analyses. The elastic modulus values back calculated Mediterranean, Tel Aviv, Israel for their support in
for different cases are listed in Table 11. performing the field tests at a site near Pune and for the
The modulus improvement factor for the reinforced supply of the geocell materials free of cost.
cases is calculated as the ratio between the modulus of the
reinforced system and the corresponding modulus of the References
unreinforced cases. These values are reported in Table 12.
It is interesting to note that these improvement factors fall ASTM D 0638. 2003. Standard test method for tensile properties of
plastics, West Conshohocken, PA, USA, ASTM International.
within the same range as those estimated using the cyclic ASTM D1204. 2008. Linear dimensional changes of nonrigid thermo-
load test results. Hence, it may be possible to utilize plastic sheeting or film at elevated temperature, West
the results from static load tests for preliminary design Conshohocken, PA, USA, ASTM International.
purposes without incurring too much of an error. ASTM D 6392. 2008. Determining the integrity of nonreinforced
geomembrane seams produced using thermo-fusion methods, West
However, the designs will not be conservative as the
Conshohocken, PA, USA, ASTM International.
modulus improvement factors from static load tests are ASTM D6647. 2001. Test method for determining tensile properties of
about 15% higher than those from cyclic load tests. geogrids by the single or multi-rib tensile method, West
The modulus improvement factors are required in Conshohocken, PA, USA, ASTM International.
mechanistic-based design of flexible pavements in which Bathurst, R. J. and Rajagopal, K. 1993. Large scale triaxial compression
testing of geocell-reinforced granular soils, Geotech. Test. J., 16, 3,
the modulus values of each pavement layers are to be 296–303.
given as input values, e.g. CIRCLY program for design of Bush, D. I., Jenner, C. G. and Bassett, R. H. 1990. The design and
pavements. These modulus improvement factors can be construction of geocell foundation mattresses supporting embank-
used to represent the equivalent behavior of the geosyn- ments over soft ground, Geotext. Geomembr., 9, 83–98.
Chandramouli, S. 2011. Performance and economic evaluation of geocells
thetic-reinforced pavement sections. The use of higher in the road pavement structure, Thesis submitted in partial
modulus for the pavement layers results in lesser thickness fulfillment of the requirements of the MTech degree, Department
for the layers as the pressure transmitted to the subgrade of Civil Engineering, Indian Institute of Technology Madras,
reduces with increase in the modulus values. By using Chennai, India.
different modulus improvement factors in the CIRCLY CIRCLY. 2013. Software for design of flexible pavements, Richmond,
VIC, Australia, MINCAD Systems Pvt. Ltd.
program, Iniyan (2012) has studied the influence of geocell Emersleben, A. and Meyer, N. 2010. The influence of hoop stresses and
and geogrid layers on the thickness of the pavement layers. earth resistance on the reinforcement mechanism of single and
Reduced thickness of the pavement layers results in lesser multiple geocells, 9th International Conference on Geosynthetics,
total cost of the pavement and lesser construction times. Brazilian Chapter of International Geosynthetics Society, IGC-
2010, Guaruja, Brazil, May 23-27, 713–716.
This will also lead to lesser carbon footprint as reduced Giroud, J. P. and Han, J. 2004. Design method for geogrid-reinforced
quantities of natural aggregate materials are required for unpaved roads. I. Development of design method and II. Calibration
construction. of applications, J. Geotech. Geoenviron. Eng., 130, (8), 775–797.
Han, J., Pokharel, S. K., Parsons, R. L., Leshchinsky, D. and Halahmi,
I. 2010. Effect of infill material on the performance of geocell-
Conclusion reinforced bases, 9th Int. Conf. on ‘Geosynthetics’ (ICG 2010),
Brazil, May 23–27.
This paper has presented some results from field and Han, J., Pokharel, S. K., Parsons, R. L., Leshchinsky, D. and Halahmi, I.
laboratory tests on the performance of pavements with 2010. Effect of infill material on the performance of geocell-
different types of geosynthetic reinforcements. It is seen reinforced bases, 9th International Conference on Geosynthetics,
Brazilian Chapter of International Geosynthetics Society ICG 2010,
that both the strength and stiffness of the pavement system
Guaraja, Brazil, May 23-27, 1503–1506.
can be improved by the use of geosynthetics. The Han, J., Yang, X. M., Leshchinsky, D. and Parsons, R. L. 2008. Behavior
performance under repeated loads is also better with of geocell-reinforced sand under a vertical load, J. Transp. Res.
geosynthetic reinforcement layers. Board, 2045, 95–101.
The improvement in the overall performance is by Iniyan, K. 2012. Influence of geogrid reinforcement on the carbon
footprint of flexible pavements, Thesis submitted in partial
distributing the applied loads over a much wider area of fulfillment of the requirements of the MTech degree, Department
the subgrade thus reducing the stresses at the subgrade of Civil Engineering, Indian Institute of Technology Madras,
level. The geocell reinforcement gives much higher Chennai, India.

International Journal of Geotechnical Engineering 2014 VOL 8 NO 3 297

Rajagopal et al. Geosynthetic-reinforced road pavement structures

IS 1888. 1988. Method of load test on soils, Bureau of Indian Standards, single geocell-reinforced based under static loading, Geotext.
New Delhi. Geomembr., 28, 6, 570–578.
Madhavi Latha, G., Dash, S. K. and Rajagopal, K. 2008. Equivalent Rajagopal, K., Krishnaswamy, N. R. and Madhavi Latha, G. 1999.
continuum simulations of geocell reinforced sand beds supporting Behavior of sand confined with single and multiple geocells,
strip footings, Geotech. Geol. Eng., 6, (4), 387–398. Geotext. Geomembr., 17, 3, 171–184.
Madhavi Latha, G., Dash, S. K. and Rajagopal, K. 2009 Numerical Shin, E. C., Kim, S. H. and Oh, Y. I. 2010. Comparison of bearing
simulation of the behaviour of geocell reinforced sand in founda- capacity on geocell-reinforced subgrade, 9th International
tions, ASCE Int. J. Geomech., 9, 143–152. Conference on Geosynthetics, Brazilian Chapter of International
Parayil, A. 2013. Performance of geosynthetic reinforced flexible Geosynthetics Society, IGC-2010, Guaruja, Brazil, May 23-27,
pavements, Thesis submitted in partial fulfillment of the require- 1441–1444.
ments of the MTech degree, Department of Civil Engineering, Unni, A. 2010. Structural and economic evaluation of geocells in the road
Indian Institute of Technology Madras, Chennai, India. pavement structure, Thesis submitted in partial fulfillment of the
Pokharel, S. K., Han, J., Leshchinsky, D., Parsons, R. I. and Halahmi, requirements of the MTech degree, Department of Civil Engineering,
I. 2010. Investigation of factors influencing behavior of Indian Institute of Technology Madras, Chennai, India.

298 International Journal of Geotechnical Engineering 2014 VOL 8 NO 3