World Academy of Science, Engineering and Technology

International Journal of Civil and Environmental Engineering

Vol:11, No:9, 2017

Efficiency of Geocell Reinforcement for Using in

Expanded Polystyrene Embankments via Numerical
Analysis
S. N. Moghaddas Tafreshi, S. M. Amin Ghotbi

 2004, are two of the most complete ones [2], [3]. However,
Abstract—This paper presents a numerical study for there are several points of interest for deeper examination
investigating the effectiveness of geocell reinforcement in reducing which were also explicitly introduced by the above-mentioned
pressure and settlement over EPS geofoam blocks in road materials. “Slope stability issues”, “seismic behavior”, and
embankments. A 3-D FEM model of soil and geofoam was created in
International Science Index, Civil and Environmental Engineering Vol:11, No:9, 2017 waset.org/Publication/10007845

“effectiveness of using geocell or geogrid above geofoam

ABAQUS, and geocell was also modeled realistically using
membrane elements. The accuracy of the model was tested by blocks” are the examples of such topics. The first two topics
comparing its results with previous works. Sensitivity analyses had the chance to gain more attention by researchers, and
showed that reinforcing the soil cover with geocell has a significant several papers and reports have been published about them.
influence on the reduction of imposed stresses over geofoam and However, the use of geogrid or geotextile has attracted much
consequently decreasing its deformation. less attention so that there is nearly no direct research on them
from 2004 up to now. It is obvious that reinforcement methods
Keywords—EPS geofoam, road embankments, geocell,
(if used consciously) introduce substantial benefits relative to
reinforcement, lightweight fill.
common techniques (e.g. using a concrete slab).
I. INTRODUCTION Since the previously cited guidelines contain a conclusive
and comprehensive discussion involving main research
U SING expanded polystyrene (EPS) geofoam as a
geotechnical material has a 50-year-old background in
engineering. EPS weighs about 1% of soil and less than 10%
materials published prior to their publication, we eliminate the
repetition of them here. Yet, we just give a short review on a
research accomplished in 2000 for its general similarity to our
of other lightweight fill substitutes. The significant benefit of present and future studies.
using such material is the reduction of the imposed loads to Reference [4] investigated the behavior of EPS geofoam
the nearby structures or the underlying soil. It is not a common used as subgrade and fill material under flexible pavement.
practice to use EPS geofoam for geotechnical applications, but They performed their tests in a special apparatus simulating
an engineer can tackle certain challenges by its application. It the wheel loading more realistically by moving the wheel
is very easy to handle without using heavy machines, which along an oval-shaped test track. In their test setup, pavement
generally yields in higher construction speed. It is also easier sections included a wearing course, a gravel base layer, and a
to work with EPS, as it is much less affected by weather sand subbase, all of which were positioned over the EPS
conditions and can be shaped and cut anytime on the project blocks inside a test box. They examined several effective
site. Finally, it is available in several densities and engineering factors including repeated traffic load, EPS block size, and
properties, and retains its initial characteristics through the side restraints. They concluded that resilient deformation of
service life. Its durability is also similar to other construction EPS geofoam at the subgrade level is much higher than that
material [1]. for compacted sand. The resilient deformation manifests as a
deeper rut on the pavement surface of the EPS geofoam
II. BACKGROUND AND SCOPE OF STUDY subgrade test section than on the compacted sand subgrade test
One of the most challenging concerns regarding the use of section, even when both have the same pavement structure.
EPS as filling material for embankments is to find the most The rut depth could be reduced, however, by using an
effective and reliable pavement system for distributing stress appropriately designed pavement structure (e.g. increase the
over geofoam and controlling its settlement. For this purpose, pavement thickness or use stiffer pavement).
a few guidelines have been published on the applications of Reference [5] presented construction and long-term
geofoam in highway embankments among which, NCHRP monitoring results for some of embankments with EPS in the
Web Document 65 and NCHRP report 529, both published in I-15 reconstruction project in Salt Lake City. They also used a
finite difference program to analyze stress distributions,
displacements and strains in specific embankments containing
S. N. Moghaddas Tafreshi is Professor of Geotechnical Engineering, K.N.
Toosi University of Technology, Tehran, Iran, 15875-4416 (phone: +98-21-
EPS geofoam. Typical configuration of the studied
88779474; e-mail: nas_moghaddas@kntu.ac.ir). embankments consisted of a 0.6-m base course covered with
S. M. Amin Ghotbi is PhD Candidate, K. N. Toosi University of 0.36-m concrete pavement and no kind of soil reinforcement
Technology, Tehran, Iran, 15875-4416 (phone: +98-9124430873; e-mail:
aminghotbi@gmail.com).
were investigated.

International Scholarly and Scientific Research & Innovation 11(9) 2017 1217 scholar.waset.org/1307-6892/10007845
 World Academy of Science, Engineering and Technology
International Journal of Civil and Environmental Engineering
Vol:11, No:9, 2017

It is clear from the literature review that there is no direct and 15°, respectively. For geocell, a simple elastic model with
research on the described topic of using reinforcement of soil Young’s modulus of 200 MPa and Poisson’s ratio of 0.35 was
above geofoam as a substitute for other methods. A detailed derived from [8]. For EPS, Mohr-Coulomb model values were
study is beneficent to our knowledge about it and will be chosen from [9]. Table I shows the values of different
valuable for achieving an optimum design procedure. In this parameters for soil, geocell, and EPS.
study, we aim to utilize a robust numerical method to assess
the incorporation of geocell reinforcement with the soil cover
over EPS fill to reduce deformation over geofoam blocks and
increase pavement’s reliability.

III. NUMERICAL MODEL

Different numerical methods are available for analyzing
stress-strain in engineering applications. Finite Element
Method (FEM) is one of the most popular, easy, and accurate
methods for this purpose. This method is available in many
International Science Index, Civil and Environmental Engineering Vol:11, No:9, 2017 waset.org/Publication/10007845

commercial and non-commercial software packages such as

ABAQUS, ANSYS, PLAXIS, etc. ABAQUS has great
capabilities including robust mesh generation techniques,
loading patterns, and constitutive laws for using in
geotechnical applications. ABAQUS was considered to be a
proper choice for our studies according to the mentioned (a)
points.
In this study, material properties and model dimensions
were selected based on our future experimental program and
laboratory settings. Soil properties and reinforcement have a
meaningful relation with those in the research of [6]. Thus, we
compare our results with it and further references to ensure its
correctness.
A. Geometry, Loading and Boundary Conditions
The future physical model will consist of a 2×2.5 m box in
plan and 90 cm in height. The walls and bottom of the box are
constructed from rough and rigid concrete. Loads will be
applied to the soil through a circular rigid steel plate with the (b)
diameter of 30 cm, which represents a typical tire loading area
Fig. 1 (a) Geometry of the embankment system. (b) Geometry of
according to [6]. Soil and geofoam will be placed in the box
geocell in ABAQUS
according to test plans. In the numerical model, sides were
fixed in the horizontal direction and set free in the vertical TABLE I
direction. It was observed that the boundary fixity in the MATERIAL PROPERTIES
horizontal direction will not impose any certain inaccuracies Material Soil EPS 30 Geocell
to the results. The dimensions of the box are sufficient to Density (kg/m3) 2000 30 333
Basic
remove any interference with the soil stresses from plate Young’s Modulus (MPa) 30 7.8 200
Properties
loading. Geocell will be placed at a specific depth in order to Poisson’s Ratio 0.4 0.17 0.35
prevent damage from direct contact with the rigid plate during Angle of Friction 45 2 -
Cohesion (kPa) - 62 -
loading. A typical geometry of the model and the geocell Plastic
Dilation Angle 15 0.1 -
created in ABAQUS are shown in Fig. 1. Properties
Yield Stress (kPa) 18 - -
B. Material Models and Properties Flow Stress Ratio 0.78 - -
Material properties were kept in close consistency with C. FE Mesh Details
those of [7] in order to derive comparable results for
validation of the model. A linear Drucker-Prager model As large deformations were expected for achieving the final
available in ABAQUS was used to simulate plastic bearing pressure during the analysis, a full scale 3D model
deformations in the soil and geofoam. This model is capable was created and meshed as shown in Fig. 2. For the validation
of reproducing stress distribution more accurately compared to part (which does not include EPS blocks), a total number of
traditional Mohr-Coulomb model. Soil’s Young’s modulus 71744 hexahedral linear elements with reduced integration
and Poisson’s ratio were set to 30 MPa and 0.4, respectively. formulation (C3D8R) were used for soil. Elements around the
Its angle of friction and dilation angle was also selected 45° plate edge were sized around 4 mm to enable calculating
extensive deformation and stress gradient in the soil under

International Scholarly and Scientific Research & Innovation 11(9) 2017 1218 scholar.waset.org/1307-6892/10007845
 World Academy of Science, Engineering and Technology
International Journal of Civil and Environmental Engineering
Vol:11, No:9, 2017

extreme loadings. It is also worth mentioning that finer mesh kPa.

was also used for the soil near the ground surface. It is clear
that element’s size was increased wherever higher accuracies
were not needed.
Geocell was modeled with a total number of 14830 with
linear membrane elements (M3D4R). An embedded
interaction was used between soil and geocell and it yielded
very good results for our study. This interaction assumes that
soil and geocell fully move together and it saves
computational cost to a great extent while generating reliable
results.
D. Analysis Procedure
Geostatic stresses were established through soil medium via
a primary step applying soil body forces. On the next step, the
foundation pressure was applied to a circular rigid plate atop
International Science Index, Civil and Environmental Engineering Vol:11, No:9, 2017 waset.org/Publication/10007845

the soil. The loading was increase until large deformations Fig. 2 FEM Mesh
occurred. The interaction between the rigid plate and soil was
considered to be “Frictionless” for tangential behavior and 1000

Soil Pressure at Depth of 300 mm

“Hard Contact” for the normal one. Loads for the both steps Reinforced
were applied through a smooth loading pattern to assure 800

reduction of numerical errors originated from dynamic effects

accompanied with more suddenly applied patterns (e.g. linear 600
(kPa)

pattern).
Explicit solver was selected for analyzing the system and 400
obtaining the final results. The explicit method uses very small
time increments and could produce accurate results if used 200
properly. It was observed that the results obtained from
explicit method had a very little difference with the implicit or 0
0 200 400 600 800 1000
static ones, while providing a much faster analysis.
Applied Pressure (kPa)
E. Validation
Fig. 3 Variation of transferred pressure with applied pressure on the
According to Fig. 3 vertical pressure in soil at depth of 300 footing surface
mm increase with the applied pressure for both unreinforced
and geocell reinforced cases, and their values are very close to
Bearing Pressure (kPa)
each other until the applied pressure reaches to about 600 kPa. 0 200 400 600 800 1000 1200
This result is compatible with those obtained by [7] when the 0
underlying and cover material has the same stiffness. In fact, 5
the stiffness of the reinforcing material has to be considerably
Footing Settlement, s/D (%)

10
larger than the values used in this research in order to mobilize
15
the mattressing effect when both material has the same
20
stiffness. However, for larger deformations where a substantial
25
part of the soil under the foundation has reached its plastic
limit, the geocell exhibits the effectiveness in reducing the 30
Reinforced
vertical stress at this depth. For both reinforced and 35
Unreinforced
unreinforced graphs, the value of the stresses locates under the 40
dashed line indicating the amount of applied pressure. When 45
the soil is unreinforced, its pressure at the 30-cm depth draws
back to the applied pressure after a certain pressure reaches a Fig. 4 Variation of bearing pressure with footing settlement
certain value (600 kPa for this case), due to the loss in soil
strength after yielding. IV. RESULTS
Fig. 4 demonstrates the variations of bearing pressure with In this section, the influence of EPS density is discussed for
the footing settlement. Pressure was applied on the foundation a model section with a soil cover of thickness 30 cm. The
until it reached an ultimate state. According to this figure, detailed properties of various EPS densities are given in Table
geocell reinforcement increases the bearing pressure of the II.
foundation by at least 50% although there is no significant Bearing pressure of the footing versus its settlement for
difference in footing settlement for the pressures less than 400 different densities of EPS is illustrated in Fig. 5. The density

International Scholarly and Scientific Research & Innovation 11(9) 2017 1219 scholar.waset.org/1307-6892/10007845
 World Academy of Science, Engineering and Technology
International Journal of Civil and Environmental Engineering
Vol:11, No:9, 2017

of each EPS type is indicated by their value in the legend, and a wider area by its special reinforcing mechanisms. Such
the reinforcement status is specified with the letters “R” for behavior is beneficent when using EPS in road embankments,
reinforced and “U” for unreinforced in front of it. The words as it is critical to avoid concentrated stress on the geofoam.
“Top” and “Bottom” in the figures refer to top of geofoam fill Figs. 7 and 8 display the pressure distribution on the top and
and under geofoam fill, respectively. It is concluded that the bottom of geofoam in unreinforced and reinforced states for
bearing pressure of all cases are less than that of the basic EPS20 and EPS30, respectively. While the amount of vertical
model (with no geofoam) except the reinforced case of EPS pressure on the top and bottom of geofoam increases with
30. In fact, the combined use of EPS 30 with geocell has increase in geofoam density, the percent of reduction between
caused a slight growth in the bearing pressure of the the reinforced and unreinforced cases decreases. For example,
foundation. Consequently, EPS 30 was selected as the suitable the pressure on top of EPS block reduces from 213 kPa to 163
density for the use in road embankments. It should be noted kPa for EPS20 (24% reduction) and from 249 kPa to 205 kPa
that the authors did not have access to the mechanical for EPS 30 (18% reduction). Comparing percentage of
properties of higher densities of EPS at this time. It is variation in the pressure over geofoams with different
therefore obvious that if higher densities show significant densities, increasing the density and Young’s modulus of EPS
improvement in the performance and cost effectiveness results in a higher increase of pressure over geofoam for
International Science Index, Civil and Environmental Engineering Vol:11, No:9, 2017 waset.org/Publication/10007845

simultaneously, their usage would be justifiable over lower reinforced cases.

densities. Although increasing geofoam densities leads to the
reduction of pressure in the bottom of EPS blocks (over the
TABLE II foundation) in the unreinforced cases, it causes an increase for
EPS PROPERTIES
the reinforced ones. Comparing the effect of reinforcement for
Material Properties EPS 15 EPS 20 EPS 30
each density, there are considerable decrease in the pressure
Density (kg/m3) 15 20 30
Young’s Modulus (MPa) 2.4 4 7.8
under geofoam in the reinforced cases relative to unreinforced
Poisson’s Ratio 0.1 0.12 0.17 ones for EPS15 and EPS20. For EPS30, there is a slight
Friction Angle 1.5 2 2.5 increase in the vertical stress under geofoam for the reinforced
Cohesion (kPa) 33.75 38.75 62 case compared to unreinforced case which can be assumed
unimportant.
Bearing Pressure (kPa) To summarize, geocell reinforcement generally has a
significant influence on the reduction of vertical stresses
0 200 400 600 800 1000 applied over EPS blocks. The amount of reduction varies from
0
EPS30-R 30% for EPS15 to 18% for EPS30. For the bottom of
EPS30-U geofoam, this reduction is 42% and 27% for EPS15 and
Footing Settlement , s/D (%)

5 EPS20-R EPS30 and 14% increase for EPS30.

EPS20-U
EPS15-R
10 -1 -0,8 -0,6 -0,4 -0,2 0 0,2 0,4 0,6 0,8 1
EPS15-U
0
25
Vertical Pressure (kPa)

15 50
75
20 100
125
150 EPS15-R-Top
25
EPS15-R-Bottom
175
EPS15-U-Top
Fig. 5 Variation of bearing pressure with footing settlement for
200 EPS15-U-Bottom
different EPS densities
Fig. 6 Pressure distribution over and under geofoam for reinforced
Fig. 6 displays top and bottom pressure distribution profiles and unreinforced cases at 700 kPa
of EPS15 for both reinforced and unreinforced cases. All
pressures were measured when the foundation pressure
reached 700 kPa. In agreement with the observations for the
case study in the validation part, geocell reinforcement
exhibits a great capability in reduction of stresses over very
soft fill material like EPS15. The vertical pressure over and
under geofoam had reductions of about 30% and 44%
respectively for the reinforced case compared to unreinforced
case. When the reinforcement is present, the stress distribution
diagram has been flattened, and a more uniform stress profile
was obtained. The reason is that geocell spreads the load over

International Scholarly and Scientific Research & Innovation 11(9) 2017 1220 scholar.waset.org/1307-6892/10007845
 World Academy of Science, Engineering and Technology
International Journal of Civil and Environmental Engineering
Vol:11, No:9, 2017

-1 -0,8 -0,6 -0,4 -0,2-25 0 0,2 0,4 0,6 0,8 1

0
25
Vertical Pressure (kPa)

50
75
100
125
EPS20-R-Top
150
EPS20-R-Bottom
175
EPS20-U-Top
200
EPS20-U-Bottom
225

Fig. 7 Pressure distribution over and under geofoam for reinforced

and unreinforced cases at 700 kPa

-1 -0,8 -0,6 -0,4 -0,2 0 0,2 0,4 0,6 0,8 1

International Science Index, Civil and Environmental Engineering Vol:11, No:9, 2017 waset.org/Publication/10007845

0
25
50
Vertical Pressure (kPa)

75
100
125
150
175
EPS30-R-Top
200
225 EPS30-R-Bottom
250 EPS30-U-Top
275 EPS30-U-Bottom

Fig. 8 Pressure distribution over and under geofoam for reinforced

and unreinforced cases at 700 kPa

REFERENCES
[1] Stark, T. D., Bartlett, S. F., Arellano, D., "Expanded Polystyrene (EPS)
Geofoam Applications and Technical Data," The EPS Industry Alliance,
2012.
[2] Stark, T. D., Arellano, D., Horvath, J. S., and Leshchinsky, D., "NCHRP
Web Document 65 (Project 24-11): Geofoam Applications in the Design
and Construction of Highway Embankments," Transportation Research
Board, Washington, D.C., 2004.
[3] Stark, T. D., Arellano, D., Horvath, J. S., and Leshchinsky, D., "NCHRP
Report 529: Guideline and Recommended Standard for Geofoam
Applications in Highway Embankments," Transportation Research
Board, Washington, D.C., 2004.
[4] Zou, Y., Small, J. C. and Leo, C. J., "Behavior of EPS Geofoam as
Flexible Pavement Subgrade Material in Model Tests," Geosynthetics
International, vol. 7, no. 1, pp. 1-22, 2000.
[5] Newman, M. P., Bartlett, S. F. and Lawton, E. C., "Numerical Modeling
of Geofoam Embankments," J. Geotech. Geoenviron. Eng., vol. 136, pp.
290-298, 2010.
[6] Brito, L. A. T., Dawson, A. R. and Kolisoja, P. J., "Analytical
Evaluation of Unbound Granular Layers in Regard to Permanent
Deformation," in Bearing Capacity of Roads, Railways and Airfields.
8th International Conference, 2009.
[7] Leshchinsky, B. and Ling, H. I., "Numerical modeling of behavior of
railway ballasted structure with geocell confinement," Geotextiles and
Geomembranes, vol. 36, pp. 33-43, 2013.
[8] Moghaddas Tafreshi, S. N., Khalaj, O. and Dawson, A. R., "Pilot-scale
load tests of a combined multilayered geocell and rubber-reinforced
foundation," Geosynthetics International, vol. 20, no. 3, pp. 143-161,
2013.
[9] Lal B. R. R., Padade, A. H. and Mandal, J. N., "Numerical Simulation of
EPS Geofoam as Compressible Inclusions in Fly Ash Backfill Retaining
Walls," in Geo-Shanghai 2014, Shanghai, China, 2014.

International Scholarly and Scientific Research & Innovation 11(9) 2017 1221 scholar.waset.org/1307-6892/10007845