International Journal of Geosynthetics and Ground Engineering (2024) 10:34

https://doi.org/10.1007/s40891-024-00541-7

ORIGINAL PAPER

Behaviour of Geocell Reinforcement in a Multi‑Layered Flexible

Pavement Under Repeated Loading
Sayanti Banerjee1 · Bappaditya Manna1 · J. T. Shahu1

Received: 31 December 2023 / Accepted: 12 March 2024 / Published online: 3 May 2024
© The Author(s), under exclusive licence to Springer Nature Switzerland AG 2024

Abstract
This study aims to comprehensively analyse the behaviour of geocell-reinforced pavement and discern the key factors affect-
ing its performance through a combined approach of experimental and numerical analyses. It begins with full-scale model
testing of reinforced and unreinforced sections, followed by numerical analysis conducted using the three-dimensional finite
element program PLAXIS 3D. The numerical models were calibrated against the results obtained from the experimental
study. The numerical investigation primarily focuses on evaluating how different parameters, including base material, diam-
eter of wheel contact area, and subgrade conditions, impact the performance of geocell-reinforced pavement sections. The
incorporation of geocells in the base layer has shown a marked reduction in both permanent deformation and concentration
of subgrade pressure compared to unreinforced sections across all pavement sections. The rut depth reduction value is found
to be influenced by the subgrade strength and the diameter of the wheel contact area. However, the study highlights a more
significant decrease in rut depth and subgrade vertical stress in reinforced pavement sections constructed with a base material
of fly ash compared to the sections with bases made of wet mix macadam (WMM) and sand materials. The study emphasizes
fly ash as an optimal infill material choice, demonstrating minimal rut depth and lower pressure values at the subgrade level.

Keywords Fly ash · Wheel contact area · Geocell · Rut depth · Pressure distribution

Introduction and rigidity of the base course are critical factors that signifi-
cantly impact the longevity of the road. Nevertheless, it is
Roads play an essential role in the transportation infrastruc- important to note that the majority of unpaved roads tend to
ture by facilitating connectivity between various cities, sup- undergo rapid deterioration, resulting in a reduced lifespan
porting economic endeavours, and contributing to the overall compared to their intended design lifespan. The most com-
growth of a nation. Roads can be categorised into many clas- mon problem in unpaved roads is permanent deformation, or
sifications, such as highways, expressways, arterial roads, rutting, often associated with excessive lateral displacement
local roads, and rural roads. A significant proportion of the in unbound aggregate bases under repetitive traffic load. To
worldwide road infrastructure consists of rural and local mitigate pavement deformation caused by the substantial
roads, with the majority of these roads being unpaved and increase in traffic volume, researchers are actively inves-
having low traffic volume. Typically, unpaved roads consist tigating alternative approaches for road construction and
of granular bases that are laid on a properly prepared sub- rehabilitation works.
grade. Therefore, in low-volume unpaved roads, the strength Over the past decade, there has been a growing interest
in the application of geosynthetics to enhance the perfor-
* Sayanti Banerjee
mance of paved roadways and reduce pavement deforma-
[email protected] tion. The incorporation of geosynthetic reinforcement has
Bappaditya Manna
led to a notable enhancement in both the load-carrying
[email protected] capacity and the design lifespan of pavements [1–12]. A
J. T. Shahu
more recent and advanced reinforcement technique, known
[email protected] as Geocell, has been introduced [13]. Geocell is designed
to provide three-dimensional confinement to the aggregate
1
Department of Civil Engineering, Indian Institute layer, effectively mitigating lateral movement within that
of Technology Delhi, Hauz Khas, New Delhi 110016, India

Vol.:(0123456789)
34 Page 2 of 17 International Journal of Geosynthetics and Ground Engineering (2024) 10:34

layer. Numerous researchers [14–21] have explored the permanent deformation value at the pavement surface and
advantageous effects of employing geocell to enhance the the subgrade stress distribution profile are obtained from
load-bearing capacity of soil and reduce subgrade settle- finite element analyses. Also, the benefit of geocell has been
ment. According to Pokharel et al. [22] and George et al. quantified in terms of rut depth reduction ratio (RDRR).
[23], the geocell mattress acts like a large slab that transfers
the pressure from the footing to the soil layer below and
spreads it over a wider area. The inclusion of the geocell
mattress results in a 2.5–3.3 times increase in the stiffness Laboratory Model Test
of the geocell-reinforced section. Furthermore, the impact
of many aspects of geocell reinforcement, such as its dimen- Material Properties
sions (height, width, and pocket size), pattern of formation,
and placement position, on the performance of the founda- The pavement section in this study comprises three differ-
tion bed has been thoroughly examined in previous stud- ent layers, namely the subgrade, subbase, and base layer.
ies [24–26]. Suku et al. [27] examined the performance of The subgrade layer utilises locally available riverbed sand,
geocell-reinforced granular bases subjected to repeated load- constructed at a 60% relative density. The Cu (coefficient
ing in unpaved roadways. The study demonstrated that the of uniformity) and Cc (coefficient of curvature) were deter-
resilient modulus is enhanced as a result of the inclusion of mined to be 2.29 and 1.18, respectively. According to the
reinforcement. Unified Soil Classification System, the sand was categorised
Hegde [28] and Banerjee et al. [29] provided a thorough as SP (poorly graded sand) [30]. A direct shear test was
overview of the current and previous research on geocell, performed on sand with a relative density of 60% according
presenting significant findings. It is important to high- to the ASTM standards [31]. The measured friction angle
light that most previous studies on geocells have focused was determined to be 40.85°. The dilation angle, calculated
on conducting static tests on a small scale. Moreover, lim- by subtracting the residual friction angle from the peak fric-
ited research has been conducted on the performance of tion angle, was determined to be 4°. In addition, the selec-
geocell-reinforced pavement subjected to repeated loading tion of the aggregate layer for the base (Wet mix macadam)
conditions. However, these studies did not utilise factory and subbase layers of the pavement sections was based on
manufactured geocells, which are predominantly employed the requirements provided by MORTH specifications. The
in real-world applications. Furthermore, it was noted that optimum moisture content and maximum dry unit weight
conventional materials, such as aggregate, crushed rock, and of granular subbase (GSB) and wet mix macadam (WMM)
sand, were mostly utilized as infill material for construct- were determined by conducting modified Proctor compac-
ing the geocell-reinforced base layer. Prior research failed tion tests in accordance with IS 2720 Part 8 [32]. The GSB
to investigate the effects of different waste materials used material exhibited a maximum dry density of 21.65 kN/m3
as infill in the geocell-reinforced pavement under repeated at an optimum moisture content of 6%. Similarly, the WMM
loading conditions. Also, the diameter of the wheel con- material showed a maximum dry density of 22.85 kN/m3 at
tact area played a significant role in the failure profile of an optimum moisture content of 5.95%. The granular sub-
the pavement section, which was not addressed in the prior base (GSB) layer was constructed with a moisture content of
study. Furthermore, previous studies conducted a very lim- 6%, which corresponds to a CBR value of 85%. Additionally,
ited numerical investigation that considered the real cur- the Wet mix macadam (WMM) layer was constructed with
vature and dimensions of the geocell layer. Therefore, it is a moisture content of 5.95%, resulting in a CBR value of
essential to perform three-dimensional numerical modeling 98.6%. The friction angles for the base and subbase layers
that considers the actual curvature and size of the geocell are 48 and 43°, respectively, as adapted from the study of Pai
layer to comprehensively understand the effects of geocells et al. [33]. The same procedure and compaction energy were
on pressure redistribution and surface settlements under used to prepare both unreinforced and reinforced pavement
repetitive loading conditions. sections. Moreover, during the construction of each pave-
Therefore, the current study conducted extensive full- ment layer, a density check has been carried out. This study
scale model testing on both reinforced and unreinforced uses a reinforced layer composed of High-Density Polyeth-
pavement subjected to repeated loading conditions to ylene (HDPE) polymeric sheets, known as a geocell. When
ascertain the advantageous impact of geocell on pavement the geocell layer expanded to its maximum size, the pockets
sections. In addition, a comprehensive investigation was formed a honeycomb shape. These pockets are connected to
conducted in numerical simulations using the finite ele- neighbouring pockets through ultrasonically welded joints.
ment package PLAXIS 3D by varying the parameters of the The height and weld spacing of the geocell layer are 150 and
base material, the diameter of the wheel contact area, and 330 mm respectively. According to the stress–strain char-
the strength properties of the subgrade layer. Finally, the acteristics of the geocell material, the tensile strength was
International Journal of Geosynthetics and Ground Engineering (2024) 10:34 Page 3 of 17 34

determined to be 19.04 kN/m at an axial strain of 10%, using of the subbase, and each geocell pockets were filled with
the ASTM standard [34]. a certain quantity of aggregate. To prevent direct contact
between the loading plate and geocell walls, a 100 mm-thick
Preparation of Model Pavement Sections aggregate cover was provided over the base layer.

A sequence of full-scale model tests was conducted in a

tank with dimensions of 2 m in length, 2 m in width, and Model Testing Procedure
1.8 m in height. The tests involved a 300 mm-diameter and
25 mm-thick steel plate for applying loads to the surface of In order to study the effect of repeated vehicular load on
the pavement, utilizing a hydraulic actuator with a maximum both reinforced and unreinforced pavement, the experimen-
load capacity of 100 kN. The important pressure bulb, as tal study has been carried out on both the model pavements
stated by Selig and Mckee [35], extends from 2 times to 2.5 under repeated loading conditions. Repetitive loads were
times the loading plate diameter beneath the loading region applied through an actuator with a 300 mm loading plate,
and on each side, accordingly. The test beds in this study employing a haversine load pulse with a frequency of 1 Hz
comprised three layers: a 1000 mm-thick subgrade layer, a and a loading amplitude of 40 kN [36]. The test has been
150 mm-thick subbase layer, and a 250 mm-thick base layer. carried out for both reinforced and unreinforced pavements
The dimensions of the test bed exceeded those of the signifi- up to 1000 loading cycles.
cant pressure bulb zone. Figure 1 provides a schematic view
of the pavement section used for the laboratory model tests.
The subgrade layer, with the required thickness, was con-
structed using river sand at 60% relative density. Following Numerical Model Description
the subgrade, the subbase layer of the required thickness was
prepared using aggregate with a moisture content of 6%. The software PLAXIS 3D [37], which is based on the finite-
The overall thickness of the subbase layer was partitioned element (FE) method, was employed to conduct numerical
into three equal parts, each measuring 50 mm in thickness. modeling of both unreinforced and geocell-reinforced pave-
Each section of 50 mm was filled with a specific amount of ment sections. The dimensions of the FE model were chosen
soil mass and compacted using a hand compactor. For unre- to accurately reflect the dimensions and characteristics of
inforced sections, the construction of the base layer follows the test tank, soil geometry, and geocell. The subsequent
the same technique. Conversely, for reinforced sections, the sections provide detailed information on the numerical mod-
150 mm high geocell layer was evenly distributed on top elling process.

Fig. 1  Schematic view of the

model pavement section
34 Page 4 of 17 International Journal of Geosynthetics and Ground Engineering (2024) 10:34

unreinforced and reinforced pavement sections, respectively.

Also, the mesh geometry of a single geocell pocket is shown
in Fig. 3c.

Material Model and Properties

Modelling Details for Different Soil Layers

The various soil layers are discretized in the current numeri-

cal analysis using 10-noded tetrahedral elements, each ele-
ment has three translational degrees of freedom per node.
Firstly, in order to validate the numerical model with the
experimental test results, the material properties utilized
in the numerical study are chosen to accurately replicate
those used in the laboratory investigation. In the experiment
study, river sand and aggregates are used as subgrade, sub-
base (Granular subbase), and base (wet mix macadam) layer
Fig. 2  Mesh convergence study respectively. Furthermore, in the numerical analysis, to rep-
Boundary Conditions and Mesh Discretization resent the behaviour sandy subgrade layer, the elastic-per-
fectly plastic Mohr–Coulomb (MC) constitutive relationship
Boundary Conditions was used to represent the sandy subgrade layer. Bhowmik
et al. [38] examined two distinct soil models, namely the
Roller supports have been implemented at all lateral soil hardening soil (HS) model and the MC model, to simulate
boundaries in the model, indicating that deformations are the sandy soil. The study revealed that both models yielded
permitted in the plane while being restrained in the normal similar responses. To represent the behaviour of the sandy
direction. Additionally, a fixed support is applied at the bot- soil, the Mohr–Coulomb model requires six input param-
tom boundary, signifying that both horizontal and vertical eters, and the properties of all these parameters are obtained
movements are restricted at this boundary. from laboratory tests provided in Table 1. The analyses uti-
lised a cohesion value of 0.5 kPa to ensure numerical stabil-
Mesh Discretization and Sensitivity Analysis ity [38]. The value of Poisson’s ratio was assumed to be 0.35
based on the prior literature data for sand [39]. Also, the
Before conducting any FE analyses, the entire model must dilation angle (ψ) of sand is 4°. Furthermore, the granular
undergo discretization into a finite number of elements. subbase (GSB) layer was modelled using the MC model. The
There are five primary meshing options available, namely material properties are taken from the study of Pai et al. [33]
‘very coarse’, ‘coarse’, ‘medium’, ‘fine’, and ‘very fine’, and the values are given in Table 1. The wet mix macadam
based on their level of ‘mesh coarseness factor’. In order (WMM) layer was modelled using the MC model. The mate-
to ascertain the most suitable mesh density, a mesh sen- rial properties of the WMM layer were taken from the study
sitivity analysis was conducted to achieve an equilibrium of Pai et al. [33], as given in Table 1. The value of the dila-
between accuracy and computing time. Figure 2 displays tion angle (ψ) of both the aggregate layers was assumed to
the results of the mesh convergence analysis conducted for be 14.4° based on the prior literature data for aggregate [40].
the geocell-reinforced pavement. Staged mesh refinement Furthermore, for the parametric study, two different
was implemented for all five mesh densities, specifically materials are used for the subgrade layer, namely sand and
focusing on refining the mesh for the geocell wall, load- clayey soil. The elastic-perfectly plastic Mohr–Coulomb
ing plate, and the soil volume near the geocell and loading (MC) constitutive relationship was used to represent the
plate. Among the several mesh sizes, namely the very course sandy subgrade layer. The material properties of sand are
mesh (30,303 elements), coarse mesh (30,594 elements), the same as given in Table 1. Furthermore, the clayey sub-
medium mesh (46,725 elements), fine mesh (74,112 ele- grade soil was represented using the hardening soil (HS)
ments), and very fine mesh (133,572 elements), the medium model. Several researchers [41–43] used the HS model to
mesh provided the most favourable outcomes in terms of represent the behaviour of the clayey subgrade. The pri-
mary input parameters for the HS model include, E50 (the
ref
the balance between the output and the computation time.
reference secant stiffness at a 50% stress level), Eoed (the
ref
Consequently, the medium mesh was selected for subsequent
reference oedometer tangent stiffness), Eur (the unload-
ref
investigations. Figures 3a and b illustrate typical discretized
ing/reloading stiffness) and the power m. The values of
International Journal of Geosynthetics and Ground Engineering (2024) 10:34 Page 5 of 17 34

Fig. 3  Three-dimensional complete mesh geometry of finite element model, a Unreinforced pavement section; b Reinforced pavement section; c
Sigle geocell pocket

Table 1  Material properties Parameters Sand GSB [33] WMM [33]

of different pavement layers
used in numerical analysis for Material model Mohr–Coulomb Mohr–Coulomb Mohr–Coulomb
validation purpose
Poisson’s ratio (v) 0.35 0.35 0.35
Cohesion (c) (kPa) 0.5 0.5 0.5
Angle of internal friction (φ) (°) 40.85 43 48
Young’s modulus (E) (MPa) 10.35 67 80

all the input parameters of clay are taken from the study Modelling Details for Geocell Layer
of Liu et al. [43], as given in Table 2. Moreover, for the
parametric study, three different materials were used for Modelling the precise curvature of the geocell is chal-
the base layer, namely wet mix macadam (WMM), sand, lenging due to the honeycomb structure of the expanded
and fly ash. For all three types of base materials, the geocell pockets. Furthermore, it cannot be accurately
MC model was considered. The material properties of the described using basic curves or trigonometric functions.
WMM layer and sand are given in Table 1. Furthermore, Therefore, the decision was made to convert an image
the material properties of fly ash were taken from the of a quarter section of an enlarged geocell pocket into a
study of Sarkar and Dawson [44], as given in Table 2. digital format and utilise the coordinates to create a single
geocell pocket within an AutoCAD program. After the
34 Page 6 of 17 International Journal of Geosynthetics and Ground Engineering (2024) 10:34

Table 2  Material properties of Parameters Clay [43] Fly ash [44]

soil used in numerical analysis
Material model Hardening soil Mohr–Coulomb
Poisson’s ratio (v) 0.2 0.35
Cohesion (c) (kPa) 13.7 0.5
Angle of internal friction (φ) (°) 10.1 28.9
Young’s modulus (E) (MPa) – 70
3.8 –
( )
ref
E50
Secant stiffness at a 50% stress level (MPa)
5.7 –
( )
ref
Eoed
Tangent stiffness at oedometer loading (MPa)
30.4 –
( )
ref
E
Unloading/reloading stiffness ur (MPa)

Analysis

The different phases are taken into consideration to depict

the various stages of construction in the real field situation.
The analysis is done in four steps:

(a) Initial phase—Initial stresses of the subgrade soil layer

have been generated;
(b) Phase 1—In this phase, the subbase layer is activated
and corresponding stresses are generated;
(c) Phase 2—Activation of the base layer for the unrein-
Fig. 4  Photograph of a single geocell pocket, a Factory-manufactured forced case has been carried out. For reinforced cases,
geocell; b Numerically developed geocell geocell and infill materials with a cushion layer are also
activated;
(d) Phase 3- The loading condition is enabled in this phase
generation of a single pocket, additional pockets were to simulate the deformation of the pavement section.
produced and connected to the existing ones to form the
complete geocell layer using the AutoCAD software. Sub-
sequently, the produced geocell shape was imported into Validation
the PLAXIS 3D program. Figure 4 exhibits a photograph
of both an actual geocell pocket and a geocell pocket gen- The laboratory testing findings of both pavement sections
erated through numerical approaches. with or without reinforcement are utilized to validate the
Now, in order to represent the behaviour of the geo- precision of the numerical model before performing fur-
cell layer, it has been modelled using the linear-elastic ther numerical analysis using this model. The dimension
model. Numerous studies [26, 45–50] have reported that of the pavement domain and material properties utilized
geocell layers undergo strains within the elastic range in the numerical study are chosen to accurately replicate
under loading conditions. The input value for the mate- those used in the laboratory investigation. The material
rial properties of the geocell layer is the axial stiffness properties of the subgrade (sand) layer used in numeri-
(EA) of 190.4 kN/m. This value is determined by dividing cal analysis are given in Table 1. Additionally, the char-
the ultimate load per meter width of the reinforcement acteristics of the subbase (GSB) and base (WMM) are
by the corresponding axial strain. Further, the values of provided in the previous subsection. For the reinforced
EA1 and EA2 are considered identical on the assumption section, the geocell layer is placed at the base layer. To
that the behaviour of the geocell layer is isotropic. The replicate the behaviour of the geocell layer, the axial stiff-
Geogrid structural elements, which are an inherent com- ness (EA) of 190.4 kN/m is used in numerical analysis.
ponent of PLAXIS 3D, are employed for simulating the A haversine load pulse is applied at a frequency of 1 Hz
geocell layers. and with a loading amplitude of 40 kN. The numerical
analysis of up to 1000 cycles required 96 h of computing
time on a desktop computer that featured the latest Intel(R)
Xeon(R) W-2275 CPU (3.3 GHz) central processing unit
International Journal of Geosynthetics and Ground Engineering (2024) 10:34 Page 7 of 17 34

and 128 GB RAM. Therefore, the numerical analysis has Figs. 5 and 6, respectively. Additionally, the current study
been carried out for a total of 1000 cycles. The accuracy compares the deformation profile at the base layer for both
of the results has been validated upto 1000 loading cycles reinforced and unreinforced sections with the results of
through comparison with experimental data. The defor- experimental investigations for various loading cycles, as
mation profiles, obtained through both experimental and illustrated in Fig. 7. From Fig. 7, it can be concluded that
numerical analyses, for unreinforced and reinforced pave- the numerical results demonstrate satisfactory agreement
ment sections at the end of 1000 cycles are depicted in with the experimental findings, showing variations of less

Fig. 5  Deformation profile for unreinforced pavement sections at the end of 1000 cycles, a Experimental analysis; b Numerical analysis

Fig. 6  Deformation profile for

reinforced pavement sections
at the end of 1000 cycles,
a Experimental analysis; b
Numerical analysis

Fig. 7  Vertical deformation

profile at different loading
cycles and relative distances for
different pavement sections, a
Unreinforced case;b Reinforced
case
34 Page 8 of 17 International Journal of Geosynthetics and Ground Engineering (2024) 10:34

than 10%. As a result, the validated Finite Element (FE) permanent deformation (or rut depth) of the unreinforced
model is now utilized for the parametric study. layer (du) and the permanent deformation of the reinforced
layer (dr) to the permanent deformation of an unreinforced
layer at a specific repeated load cycle. The mathematical
Parametric Study representation is denoted as Eq. (1).
du − dr
To study the behaviour of geocell layer, a series of three- RDRR = × 100 (1)
dimensional FE analyses was carried out using the validated du
FE model. The primary aim of this parametric study is to
assess the influences of significant geometric parameters,
i.e., type of base material (B), strength of subgrade, diameter Effect of Different Base Material
of the wheel contact area (D) on the settlement response
and vertical stress distribution at the subgrade layer for The investigation involved varying the base material
both reinforced and unreinforced sections. Moreover, the (WMM, fly ash, and sand) while keeping the diameter of
effects of these parameters on the rut depth reduction ratio wheel contact area constant (D = 300 mm). The aim was to
(RDRR) are evaluated as well. The height and weld spacing evaluate the influence of the base material on the permanent
of the geocell used in this entire parametric study is kept deformation behaviour of both reinforced and unreinforced
constant, i.e. 150 and 330 mm respectively, which is used pavement. This study also intends to investigate the influ-
in the experimental study. The different cases considered in ence of different base materials on various subgrade condi-
the parametric study are presented in Table 3. The material tions. The impact of various base materials on the permanent
properties of subgrade and base materials are depicted in deformation value for different numbers of loading cycles
Tables 1 and 2. The analysis has been carried out for repeti- and pavement sections overlying different subgrade condi-
tive loading conditions with a frequency of 1 Hz and loading tions (namely, sandy and clayey subgrade) is illustrated in
amplitude of 40 kN. Figs. 8a–c.
By comparing Figs. 8a–c, it is evident that the base layer
built with sand material exhibits a larger permanent defor-
Results and Discussion mation value of approximately 35% than the deformation
value of the pavement sections constructed using wet mix
Vertical Deformation Profile at the Base Layer macadam (WMM) and fly ash. This is likely due to the infe-
rior mechanical properties of sand, such as lower modu-
This section explains the vertical settlement profile for the lus value. Additionally, Fig. 8c reveals that the base layer
pavement, both with and without reinforcement, along with constructed with sand material experienced rutting failure
the corresponding number of cycles. This study comprehen- (rutting greater than or equal to 20 mm, as per IRC 37 guide-
sively investigates the impact of subgrade condition, base line) by the end of 1000 cycles for the unreinforced section
material type, and diameter of wheel contact area (loading overlying clayey subgrade. Conversely, the base layer built
plate diameter) on the settlement response of pavement. The with WMM and fly ash (Figs. 8a and b) exhibited rut depths
enhancement in the efficiency of geocell-reinforced pave- of approximately 13 and 16 mm, respectively, at the end of
ment is quantified by one non-dimensional factor, i.e., the 1000 cycles for the unreinforced section overlying clayey
rut depth reduction ratio (RDRR). RDRR, as defined by subgrade. Furthermore, the base layer built with WMM, fly
Saride et al. [51], is the ratio of the difference between the ash, and sand exhibited rut depths of approximately 8, 10,

Table 3  The different case Analysis Reinforcement type Subgrade Variable parameters No of analysis
for numerical analysis of sequence material
unreinforced and reinforced
sections 1 Unreinforced Sand B = WMM, sand, fly ash 18
D = 150,300, 400
Clay B = WMM, sand, fly ash
D = 150,300, 400
2 Geocell reinforced Sand B = WMM, sand, fly ash 18
D = 150,300, 400
Clay B = WMM, sand, fly ash
D = 150,300, 400

Note: B = type of base material, D = diameter of wheel contact area (mm)

International Journal of Geosynthetics and Ground Engineering (2024) 10:34 Page 9 of 17 34

Fig. 8  Permanent deformation versus number of loading cycles for different pavement sections, a WMM; b Fly ash; c Sand

a pivotal role in enhancing the overall performance of

the pavement system by actively sharing the load and
providing anchorage to the infill material. Through its
interconnected cells, the geocell efficiently distributes
applied loads across a wider area, mitigating localized
stress concentrations and reducing the rut depth of the
pavement section. Additionally, the anchorage mechanism
of the geocell prevents lateral movement and displace-
ment of the infill material under load, ensuring stability
and integrity even in dynamic loading conditions. This
combined load-sharing and anchorage function signifi-
cantly improves the resistance of pavement sections to
rutting, cracking, and other forms of distress, leading to
enhanced durability, longevity, and overall performance
of the pavement structure. Furthermore, the displace-
Fig. 9  Deformation profile of geocell
ment contour of unreinforced and reinforced pavement is
shown in Fig. 10. The observation from Fig. 10a indicates
and 14 mm, respectively, at the end of 1000 cycles for the that under applied load, soil particles experience lateral
unreinforced section lying over a sand subgrade. This indi- movement on both sides of the plate, resulting in higher
cates that the influence of subgrade conditions is apparent lateral displacement in the unreinforced pavement sec-
across all pavement sections, regardless of the material used tion. However, in geocell-reinforced pavement sections,
in constructing its base layers. the confinement effect provided by the anchorage func-
Figures 8a–c demonstrate that the incorporation of tion of geocell significantly reduces the lateral movement
geocell in the base layer led to a reduction in permanent of soil particles. This is evident from the observation of
deformation values by approximately 90%, as compared less lateral displacement in the reinforced pavement sec-
to the permanent deformation values of the unreinforced tions compared to the unreinforced case. The geocell layer
section, across all pavement sections. Geocells improve effectively restrains the soil particles, preventing exces-
the shear resistance of the base layer material by increas- sive lateral spreading and contributing to improved stabil-
ing its internal friction and cohesion. The confinement ity and performance of the pavement structure. Moreover,
provided by the geocells prevents the individual particles the impact of various base materials on the Rut Depth
of the base layer material from lateral movement under Reduction Ratio (RDRR) value for different pavement
load, thereby resisting shear deformation and reducing sections is illustrated in Fig. 11. Figure 11b highlights
rutting. In order to better understand the reinforcement that the base layer constructed with fly ash exhibits a
mechanism of geocell, the deformation profile of geocell higher RDRR value of approximately 4% when compared
layer is given in Fig. 9. From Fig. 9, it is seen that the to pavement sections constructed with base materials of
displacement in the geocell is more prominent around WMM and sand. Conversely, Fig. 11c highlights that the
the loading region and gradually reduces towards the base layer constructed with sand exhibits a lower RDRR
ends. The geocell layer positioned at the base layer plays value of approximately 5% when compared to pavement
34 Page 10 of 17 International Journal of Geosynthetics and Ground Engineering (2024) 10:34

Fig. 10  Soil displacement con-

tour diagram; a Unreinforced
case; b Reinforced case

Fig. 11  Rut depth reduction

ratio versus number of loading
cycles, a WMM; b Fly ash; c
Sand

sections constructed with base materials of WMM and are prevalent choices for infill material, the findings pro-
fly ash. The study suggests that employing fly ash as an pose that fly ash can be a viable alternative in achieving
infill material in reinforced pavement results in a more similar or improved performance.
effective confinement effect, leading to a reduction in
permanent deformation values. While WMM and sand
International Journal of Geosynthetics and Ground Engineering (2024) 10:34 Page 11 of 17 34

Fig. 12  Permanent deformation

versus number of loading cycles
for different pavement sections,
a Sand; b Clay

Effect of Different Subgrade Material Furthermore, the beneficial effect of geocell reinforce-
ment is quantified in terms of rut depth reduction ratio
To evaluate the effect of the subgrade condition on the per- (RDRR) and the effect of subgrade condition on RDRR
manent deformation response of both reinforced and unre- value for different loading cycles is shown in Fig. 13.
inforced pavement, the subgrade condition was varied while Figure 13 demonstrates that the rut depth reduction value
maintaining the type of base material (i.e., sand) constant. decreases as the number of loading cycles increases. This
This study also aims to examine the impact of subgrade suggests that there is a greater amount of permanent defor-
conditions on different diameters of wheel contact areas. mation during the early loading cycles. However, as the
Increasing the modulus value of the subgrade layer can lead number of cycles increases, the permanent deformation
to a reduction in permanent deformation for both pavement values decrease for various subgrade conditions. Moreo-
sections, with and without the geocell layer, as depicted ver, it has been noted that the RDRR value is higher for the
in Fig. 12. For all pavement sections, the clayey subgrade clayey subgrade (as shown in Fig. 13b) compared to the
layer (as shown in Fig. 12b) exhibits approximately 50% RDRR values for the sandy subgrade (as shown in Fig. 13a).
higher deformation compared to the sandy subgrade layer Furthermore, Figs. 13a and b demonstrate that, in all pave-
(as shown in Fig. 12a). The research findings indicate that ment sections, the RDRR values increase when the size of
rut depth escalates with an increasing number of loading the wheel load area decreases, irrespective of the subgrade
cycles. Nonetheless, the incorporation of geocell leads to conditions. These findings suggest that the RDRR values
a substantial reduction in rut depths, approximately 89% exhibit an inverse correlation with the subgrade strength and
less compared to the unreinforced section. The geocell the diameter of the wheel contact area. It implies that as
with encapsulated soil layer functions like a mattress and the subgrade strength decreases, the rut depth in the unre-
reduces the overall deformation value of the reinforced sec- inforced section tends to increase. This is because a weaker
tion. Moreover, the impact of the diameter of the wheel load subgrade is more prone to deformation under traffic loads.
area is evident across all pavement sections, regardless of the On the other hand, the geocell-reinforced section experi-
subgrade conditions. ences a reduction in rut depth due to the confinement and

Fig. 13  Rut depth reduction

ratio versus number of loading
cycles, a Sand; b Clay
34 Page 12 of 17 International Journal of Geosynthetics and Ground Engineering (2024) 10:34

stabilization provided by the geocells. This reduction is more highlights the crucial role of the diameter of wheel contact
pronounced in weaker subgrade conditions. Similar find- area in distributing the load imposed by a vehicle onto the
ings in RDRR values at the base layer for geogrid-reinforced pavement. A larger diameter of wheel contact area contrib-
pavement were observed by Baadiga et al. [52] in their study. utes to distributing the load over a larger surface, thereby
reducing stress concentration and the deformation value of
Effect of the Diameter of Wheel Contact Area the pavement. Also, Fig. 14 depicts that the influence of sub-
grade conditions is apparent across all pavement sections,
In this study, the diameter of wheel contact area (loading regardless of the diameter of wheel contact area.
plate diameter) was varied at three different sizes (D = 150, Moreover, the beneficial effect of the geocell layer on
300, and 400 mm) while maintaining a consistent base mate- reducing permanent deformation is evident when compar-
rial type (WMM). The objective was to evaluate the impact ing the deformation values of sections without reinforce-
of the wheel contact area on the permanent deformation ment, regardless of the diameter of wheel contact area. The
response of both reinforced and unreinforced pavement. This advantageous effect of geocell reinforcement is quantified
study also aims to examine the impact of various diameters in terms of the rut depth reduction ratio (RDRR), and the
of wheel contact areas on different subgrade circumstances. results for different loading areas are presented in Fig. 15.
The impact of different diameters of wheel contact area on Figure 15 depicts that the pavement sections correspond-
the permanent deformation values for various numbers of ing to a 150 mm loading plate diameter exhibit a higher
loading cycles and pavement sections over different sub- RDRR value, approximately 2% more than the RDRR value
grade conditions (sandy and clayey subgrade) is depicted in of the same pavement sections corresponding to 300 and
Figs. 14a–c. Upon comparing these Figs. 14a–c, it becomes 400 mm loading plate diameter. When the loading plate is
evident that the unreinforced and reinforced sections corre- relatively small compared to the geocell pocket diameter,
sponding to a 150 mm loading plate diameter exhibit a larger the load is concentrated over a smaller area. This can lead
permanent deformation value, approximately 43% more than to higher stresses being applied to the soil within the geocell
the deformation value of the same pavement sections cor- pockets directly beneath the loading plate. As a result, the
responding to 300 and 400 mm loading plate diameter. This geocells in that area experience greater confinement, which

Fig. 14  Permanent deformation versus number of loading cycles for different pavement sections, a D = 150; b D = 300; c D = 400 mm

Fig. 15  Rut depth reduction

ratio versus number of loading
cycles, a D = 150; b D = 300; c
D = 400 mm
International Journal of Geosynthetics and Ground Engineering (2024) 10:34 Page 13 of 17 34

Fig. 16  Pressure distribution

profile at subgrade layer versus
number of loading cycles, a
Unreinforced case; b Reinforced
case

enhances their effectiveness in distributing the load and diameter of wheel contact area on the pressure distribution
reducing vertical deformation. Conversely, when the load- profile at the subgrade layer is elaborately discussed.
ing plate is larger relative to the geocell pocket diameter, the
load is distributed over a larger area. Due to the interference
to a greater number of geocell pockets for larger loading Pressure Distribution Profile at the Subgrade Layer
areas, lower stresses are being applied to the soil within each
geocell pocket, reducing the degree of confinement. As a For both reinforced and unreinforced pavement sections,
result, the geocells may be less effective in restraining lateral it is essential to obtain the maximum induced vertical
deformation and redistributing loads compared to the case stresses at the interface of the subgrade and subbase lay-
of a smaller loading plate. Similar results were observed ers. It is critical because the stresses that exceed the elastic
by Rajagopal et al. [53]. Also, it was observed that as the limit can cause significant deformation at the subgrade,
number of cycles increases the RDRR value also increases, which may result in shear failures. Therefore, in this sec-
regardless of the diameter of the wheel contact area. Con- tion, the pressure distribution profile at the subgrade layer
sequently, the pressure distribution value at the subgrade is obtained for the pavement, both with and without rein-
layer also increases with the number of cycles, as shown in forcement, along with the corresponding number of cycles.
Fig. 16. Figure 16 shows the pressure distribution profile Also, the study comprehensively investigates the impact of
for the different numbers of cycles for both the reinforced subgrade condition, base material type, and wheel contact
and unreinforced pavement with WMM base layer overlying area (loading plate diameter) on the subgrade pressure dis-
sandy subgrade. Furthermore, in the subsequent section, the tribution profile of pavement.
effect of different base materials, subgrade conditions, and

Fig. 17  Subgrade pressure distribution for different pavement sections at the end of 1000 loading cycles, a WMM; b Fly ash; c Sand
34 Page 14 of 17 International Journal of Geosynthetics and Ground Engineering (2024) 10:34

Effect of Different Base Material Effect of Different Subgrade Material

In this study, to evaluate the pressure distribution at the In this case, the vertical pressure distribution at the subgrade
subgrade layer, vertical pressure is obtained along the layer is determined for both reinforced and unreinforced
loading axis and away from the loading axis, specifically pavement sections lying over different subgrade conditions,
at 0, 300, and 600 mm. Figure 17 illustrates the verti- specifically clayey and sandy. Furthermore, this study has
cal pressure distribution at the subgrade layer for sev- been conducted for various diameters of wheel contact areas
eral pavement sections constructed using different base while maintaining a consistent base material (i.e., sand). Fig-
materials, namely WMM, fly ash, and sand. This study ure 18 depicts the impact of different subgrade conditions on
investigates various subgrade conditions, including sand the pressure distribution for various pavement sections after
and clay, while maintaining a constant diameter of wheel 1000 loading cycles. Figures 18a and b reveal that the pave-
contact area of 300 mm. Figures 17a–c show that the ment sections built over the clayey subgrade layer exhibit a
sand material used in the base layer leads to a higher higher pressure value at the subgrade layer, approximately
pressure at the subgrade, roughly 23%, compared to the 34% more than that observed for the pavement section over-
vertical pressure values of pavement sections built with laying the sandy subgrade. Also, the incorporation of geo-
wet mix macadam (WMM) and fly ash. Conversely, the cell leads to a substantial reduction in pressure distribution
study highlights that the base layer constructed with fly at the subgrade, approximately 45% less compared to the
ash exhibits a lower vertical pressure at the subgrade unreinforced section. Moreover, it has been noted that the
layer, roughly 15%, compared to the vertical pressure reduction in localized pressure value is higher for the clayey
values of pavement sections constructed with base mate- subgrade (as shown in Fig. 18b) compared to the pressure
rials of sand. Also, among the unreinforced section, it values for the sandy subgrade (as shown in Fig. 18a). Due
was observed that the base layer constructed with WMM to the flexural rigidity and confinement effect, the geocell
exhibits a lower vertical pressure at the subgrade layer, layer effectively distribute the vertical pressure over wider
compared to the vertical pressure values of pavement sec- area and reduce the stress concentration at a specific point
tions constructed with base materials of sand and fly ash. of pavement section.
Furthermore, the unreinforced sections exhibit a higher
vertical pressure at the subgrade, showing an approximate Effect of the Diameter of Wheel Contact Area
increase of 43% compared to the reinforced sections. This
outcome highlights the capability of geocell reinforce- The study was conducted to assess the influence of differ-
ment to mitigate stress concentration and uniformly dis- ent diameters of wheel contact areas (D = 150, 300, and
tribute vertical pressure in the subgrade layer. Addition- 400 mm) on the vertical pressure distribution at the sub-
ally, the inclusion of geocell results in a more significant grade layer. In this case, the type of subgrade layer is varied
reduction in subgrade stress observed in reinforced pave- (sandy and clayey subgrade) while keeping a consistent base
ment sections constructed with a base material of fly ash material type (fly ash). The impact of different wheel contact
compared to the sections with bases made of WMM and areas on vertical pressure distribution value for various pave-
sand materials. Based on the observed performance, fly ment sections over different subgrade conditions (sandy and
ash emerges as an optimized choice for an infill material, clayey subgrade) is depicted in Figs. 19a–c. Figures 19a–c
providing minimal pressure values at the subgrade. show that the unreinforced and reinforced sections cor-
responding to a 150 mm loading plate diameter exhibit a

Fig. 18  Subgrade pressure dis-

tribution for different pavement
sections at the end of 1000 load-
ing cycles, a Sand; b Clay
International Journal of Geosynthetics and Ground Engineering (2024) 10:34 Page 15 of 17 34

Fig. 19  Subgrade pressure distribution for different pavement sections at the end of 1000 loading cycles, a D = 150; b D = 300; c D = 400 mm

larger subgrade stress value, approximately 15% more than of 43% compared to the reinforced sections. This out-
the stress value of the same pavement sections correspond- come highlights the capability of geocell reinforcement
ing to 300 and 400 mm loading plate diameter. Furthermore, to mitigate stress concentration and uniformly distribute
Fig. 19 clearly illustrates that the impact of subgrade condi- vertical pressure in the subgrade layer.
tions is evident in all pavement sections, irrespective of the 4. The sand material used in the base layer leads to a higher
wheel contact area. Also, the incorporation of geocell leads rut depth and maximum pressure at the subgrade com-
to a substantial reduction in pressure distribution at the sub- pared to the same for the pavement sections built with
grade, approximately 43% less compared to the unreinforced base material of wet mix macadam (WMM) and fly ash.
section. 5. However, the study highlights a more significant
decrease in rut depth and subgrade vertical stress in rein-
forced pavement sections constructed with a base mate-
Conclusions rial of fly ash compared to the sections with bases made
of WMM and sand materials. The study emphasizes fly
This study aims to comprehensively analyse the behaviour ash as an optimal infill material choice, demonstrating
of geocell-reinforced pavement and discern the key factors minimal rut depth and lower pressure values at the sub-
affecting its performance through a combined approach of grade level.
experimental and numerical analyses. The numerical models
have been validated using the laboratory test results, and
Acknowledgements This research was funded by the "Performance
further parametric assessments were conducted. The present Study of Geocell Reinforced Road Pavement at Dholera Activation
study yields the following conclusions. Area" project sponsored by Dholera Industrial City Development
Ltd., A Government of Gujarat Undertaking, under Sanction No.
1. The incorporation of geocell in the base layer led to a FT/05/299/2021.
reduction in permanent deformation values by approxi- Authors’ Contributions SB: study conceptualization, methodology,
mately 90%, as compared to the permanent deformation numerical analysis, writing original draft, BM: study conceptualiza-
values of the unreinforced section, across all pavement tion, supervision, writing review and editing. JTS: supervision, writing
sections. review and editing.
2. The rut depth reduction ratio (RDRR) value is higher for Funding Dholera Industrial City Development Ltd., A Government of
the clayey subgrade compared to the RDRR values for Gujarat Undertaking, FT/05/299/2021, Bappaditya Manna
the sandy subgrade. Also, for all pavement sections, the
RDRR values increase when the diameter of the wheel Data Availability All data, models, and code generated or used during
the study appear in the submitted article.
contact area decreases, irrespective of the subgrade con-
ditions. These findings suggest that the RDRR values Declarations
exhibit an inverse correlation with the subgrade strength
and the diameter of the wheel contact area. Conflict of Interest The authors declare no conflict of interest.
3. The unreinforced sections exhibit a higher vertical pres-
sure at the subgrade, showing an approximate increase
34 Page 16 of 17 International Journal of Geosynthetics and Ground Engineering (2024) 10:34

References 20. Khan MA, Puppala AJ (2023) Sustainable pavement with geo-
cell reinforced reclaimed-asphalt-pavement (RAP) base layer.
J Clean Prod 387:135802
1. Haas R, Walls J, Carroll RG (1988) Geogrid reinforcement of
21. Gottumukkala B, Mehar B, Minchala D, Pulikanti SP, Kuna KK
granular base in flexible pavements. In: 67th Annual Meeting,
(2023) Laboratory and field evaluations of geocell reinforced
Transportation Research Board, Washington DC, USA
bases for locally available material in the Himalayan region. Int
2. Leng J, Gabr MA (2005) Numerical analysis of stress—defor-
J Geosynth Ground Eng 9:74
mation response in reinforced unpaved road sections. Geosynthe
22. Pokharel SK, Han J, Leshchinsky D, Parsons RL (2018) Experi-
Int 12:111–119. https://​doi.​org/​10.​1680/​gein.​2005.​12.2.​111
mental evaluation of geocell reinforced bases under repeated
3. Shahu JT (2007) Pullout response of inextensible sheet rein-
loading. Int J Pavement Res Technol 11(2):114–127
forcement subject to oblique end-force. J Geotech Geoenviron
23. George AM, Banerjee A, Puppala AJ, Saladhi M (2021) Per-
Eng, ASCE 133(11):1440–1448
formance evaluation of geocell reinforced reclaimed asphalt
4. Patra S, Shahu JT (2012) Pasternak model for oblique pullout of
pavement (RAP) bases in flexible pavements. Int J Pavement
inextensible reinforcement. J Geotech Geoenviron Eng, ASCE
Eng 22(2):181–191
138(12):1503–1513
24. Dash SK, Rajagopal K, Krishnaswamy NR (2007) Behaviour of
5. Shukla SK (2012) Handbook of geosynthetic engineering. ICE
geocell-reinforced sand beds under strip loading. Can Geotech
Publishing, London
J 44(7):905–916
6. Abu-Farsakh MY, Gu J, Voyiadjis GZ, Chen Q (2014) Mecha-
25. Latha GM, Somwanshi A (2009) Bearing capacity of square
nistic–empirical analysis of the results of finite element analysis
footings on geosynthetic reinforced sand. Geotext Geomembr
on flexible pavement with geogrid base reinforcement. Int J
27(4):281–294
Pavement Eng 15:786–798. https://​doi.​org/​10.​1080/​10298​436.​
26. Gedela R, Karpurapu R (2021) Laboratory and numerical stud-
2014.​893315
ies on the performance of geocell reinforced base layer overly-
7. Bagshaw SA, Herrington PR, Kathirgamanathan P, Cook SR
ing soft subgrade. Int J Geosynth Ground Eng 7:7. https://​doi.​
(2015) Geosynthetics in basecourse stabilisation. NZ Transport
org/​10.​1007/​s40891-​020-​00249-4
Agency research report 574, p 64, ISBN 978-0-478-44577-0
27. Suku L, Prabhu SS, Ramesh P, Babu GLS (2016) Behavior of
8. Bhowmik R, Shahu JT, Datta M (2018) Failure analysis of a
geocell-reinforced granular base under repeated loading. Transp
geomembrane lined reservoir embankment. Geotext Geomembr
Geotech 9(17–30):923. https://​doi.​org/​10.​1016/j.​trgeo.​2016.​06.​
46(1):52–65
002
9. Bhowmik R, Shahu JT, Datta M (2019) Experimental studies
28. Hegde A (2017) Geocell reinforced foundation beds-past find-
on inclined pullout behavior of geosynthetic sheet vis-à-vis
ings, present trends and future 859 prospects: a state-of-the-art
geogrid—effect of type of anchor and sand. Geotext Geomembr
review. Constr Build Mater 154:658–674. https://​doi.​org/​10.​
47(6):769–779
1016/j.​conbu​ildmat.​2017.​07.​230
10. Banerjee S, Srivastava MVK, Manna B, Shahu JT (2022) A
29. Banerjee S, Manna B, Shahu JT (2023) Geocell as a prom-
novel approach to the design of geogrid-reinforced flexible
ising reinforcement technique for road pavement: a state of
pavements. Int J Geosynth Ground Eng 8:29. https://​doi.​org/​
the art. Indian Geotech J:2277–3347. https://​doi.​org/​10.​1007/​
10.​1007/​s40891-​022-​00373-3
s40098-​023-​00818-0.
11. Chua BT, Hossam AN, Nepal KP (2022) Analysis and design of
30. ASTM D2487–17e1 (2017) Standard practice for classification
geogrid-reinforced unbound granular pavement over soft sub-
of soils for engineering purposes (Unified Soil Classification
grade for low volume roads. Aust Geomech 57(1):25–41
System). ASTM International, West Conshohocken
12. Deshmukh R, Patel S, Shahu JT (2022) Full-scale field perfor-
31. ASTM D3080–04 (2004) Standard test method for direct shear
mance of geocell reinforced-fly ash in the subbase course of
test of soils under consolidated drained conditions. ASTM Inter-
flexible pavement. Int J Geosynth Ground Eng 8:36. https://​doi.​
national, West Conshohocken
org/​10.​1007/​s40891-​022-​00383-1
32. BIS (Bureau of Indian Standards) (2006) Determination of
13. Bathurst RJ, Karpurapu R (1993) Large-scale triaxial testing of
water content- dry density relation using heavy compaction. IS
geocell-reinforced granular soils. Geotech Test J 16:296–303
2720 Part 8. New Delhi, India: BIS
14. Latha GM, Rajagopal K, Krishnaswamy NR (2006) Experimen-
33. Pai RR, Bakare MD, Patel S, Shahu JT (2021) Structural
tal and theoretical investigations on geocell supported embank-
evaluation of flexible pavement constructed with steel slag-
ments. Int J Geomech 6(1):30–35
fly ash-lime mix in the base layer. J Mater Civ Eng ASCE
15. Pokharel SK, Han J, Leshchinsky D, Parsons RL, Halahm I
33(6):04021097
(2010) Investigation of factors influencing behavior of sin-
34. ASTM D 4595 (2009) Standard test method for Tensile properties
gle geocell-reinforced bases under static loading. Geotext
of geotextiles by the wide-width strip method. ASTM Interna-
Geomembr 28(6):570–578
tional, West Conshohocken, PA, 2016. www.​astm.​org
16. Yang X, Han J, Pokharel SK, Manandhar C, Parsons RL,
35. Selig ET, McKee KE (1961) Static and dynamic behaviour of
Leshchinsky D, Halahmi I (2012) Accelerated pavement testing
small footings. J Soil Mech Found Division 87:29–47
of unpaved roads with geocell-reinforced sand bases. Geotext
36. IRC 37 (2018) Guidelines for the design of flexible pavements.
Geomembr 32:95–103
Indian Code of Practice, Indian Road Congress, New Delhi, India
17. Mamatha KH, Dinesh SV (2019) Performance evaluation of
37. Plaxis BV (2023) PLAXIS 3D-Reference Manual. Delft. Plaxis,
geocell-reinforced pavements. Int J Geotech Eng 13(3):277–286
The Netherlands
18. Ghothi Siabil SG, Moghaddas Tafreshi SN, Dawson AR (2020)
38. Bhowmik R, Shahu JT, Datta M (2020) Finite element mode-
Response of pavement foundations incorporating both geocells
ling of geogrids in trenches under inclined pull. Int J Geomech
and expanded polystyrene (EPS) geofoam. Geotext Geomembr
20:04020129
48(1):1–23. https://​doi.​org/​10.​1016/j.​geote​xmem.​2019.​103499
39. Bowles JE (1997) Foundation analysis and design. McGraw-Hill,
19. Sheikh IR, Mandhaniya P, Shah MY (2021) A parametric study
Singapore
on pavement with geocell reinforced rock quarry waste base on
40. Juneja G, Sharma RK (2023) Experimental and numerical analy-
dredged soil subgrade. Int J Geosynth Ground Eng 7:32. https://​
sis of geocell-reinforced base layer with different infill materials
doi.​org/​10.​1007/​s40891-​021-​00275-w
overlying clay. J Hazard Toxic Radioac Waste 27(4):04023024
International Journal of Geosynthetics and Ground Engineering (2024) 10:34 Page 17 of 17 34

41. Cudny M, Truty A (2020) Refinement of the hardening soil model 49. Han J, Yang X, Leshchinsky D, Parsons RL (2008) Behaviour of
within the small strain range. Acta Geotechnica 15(8):2031–2051. geocell reinforced sand under a vertical load. J Transp Res Board
https://​doi.​org/​10.​1007/​s11440-​020-​00945-5 20(45):95–101. https://​doi.​org/​10.​3141/​2045-​11
42. Kung GT-C, Ou C-Y, Juang CH (2009) Modeling small-strain 50. Dutta S, Mandal JN (2017) Numerical analyses on cellular mat-
behavior of Taipei clays for finite element analysis of braced exca- tress-reinforced fly ash beds overlying soft clay. Int J Geomech
vations. Comput Geotech 36(1–2):304–319 17(4):04016095. https://​doi.o​ rg/​10.​1061/​(ASCE)​GM.​1943-​5622.​
43. Liu C, Yan C, Zheng G, Liu T, Yang Y (2023) Field testing and 00007​72
numerical analysis of supporting performance of oblique piles 51. Saride S, Rayabharapu VK, Suraj V (2015) Evaluation of rutting
used in pit excavation. Int J Geomech 23(11):1532–3641 behaviour of geocell reinforced sand subgrades under repeated
44. Sarkar R, Dawson AR (2017) Economic assessment of use of pond loading. Indian Geotech J 45(4):378–388
ash in pavements. Int J Pavement Eng 18(7):578–594. https://​doi.​ 52. Baadiga R, Balunaini U, Saride S, Madhav MR (2021) Influence
org/​10.​1080/​10298​436.​2015.​10959​15 of geogrid properties on rutting and stress distribution in rein-
45. Leshchinsky B, Ling H (2013) Effects of geocell confinement on forced flexible pavements under repetitive wheel loading. J Mater
strength and deformation behaviour of gravel. J Geotech Geoen- Civ Eng 33(12):14
viron Eng 139(2):340–352. https://​doi.​org/​10.​1061/​(ASCE)​GT.​ 53. Rajagopal K, Krishanaswamy NR, Latha GM (1999) Behav-
1943-​5606.​00007​57 iour of sand confined with single and multiple geocells. Geotext
46. Hegde A, Sitharam TG (2015) 3-Dimensional numerical analy- Geomembr 17:171–184
sis of geocell reinforced soft clay beds by considering the actual
geometry of geocell pockets. Can Geotech J 52(9):1396–1407. Publisher's Note Springer Nature remains neutral with regard to
https://​doi.​org/​10.​1139/​cgj-​2014-​0387 jurisdictional claims in published maps and institutional affiliations.
47. Hegde A, Sitharam TG (2015) 3-Dimensional numerical mod-
elling of geocell reinforced sand beds. Geotext Geomembr Springer Nature or its licensor (e.g. a society or other partner) holds
43(2):171–181. https://d​ oi.o​ rg/1​ 0.1​ 016/j.g​ eotex​ mem.2​ 014.1​ 1.0​ 09 exclusive rights to this article under a publishing agreement with the
48. Yang X, Han J, Parsons LR, Leshchinsky D (2010) Three-dimen- author(s) or other rightsholder(s); author self-archiving of the accepted
sional numerical modelling of single geocell reinforced sand. manuscript version of this article is solely governed by the terms of
Front Archit Civ Eng J 4(2):233–240. https://​doi.​org/​10.​1007/​ such publishing agreement and applicable law.
s11709-​010-​0020-7