Geotextiles and Geomembranes 42 (2014) 599e610

Contents lists available at ScienceDirect

Geotextiles and Geomembranes

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

Behavior of strip footing on fiber-reinforced cemented sand adjacent

to sheet pile wall
Ahmed M. Nasr
Structural Engineering Department, Faculty of Engineering, Tanta University, Egypt

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

Article history: In urban areas, shallow foundations are often placed along the ground surface above a sheet pile wall. In
Received 3 June 2014 this research, the potential benefits of reinforcing the active zone behind a model sheet pile wall by using
Received in revised form polypropylene fiber and cement kiln dust have been investigated experimentally and numerically. Tests
16 September 2014
were conducted by varying parameters including fiber ratio (RF), cement kiln dust (CKD) ratio, thickness
Accepted 25 October 2014
of reinforced layer, footing location relative to the sheet pile wall and curing time of reinforced layer.
Available online 14 November 2014
Finite element computer code PLAXIS 2D foundation was used for numerical modeling. Close agreement
between the experimental and numerical results was observed (maximum difference 14%). Experimental
Keywords:
Geosynthetics
and numerical results clearly show that fiber insertion into the cemented soil causes an increase in
Sheet pile wall ultimate bearing capacity of footing and significant reduction in the lateral deflection of the sheet pile
Fiber-reinforced sand wall. At higher fiber ratios (RF
0.75%), the bearing capacity ratio (BCR) increased by about 42% and the
Cement kiln dust effect of CKD ratio on BCR is more pronounced. The addition of fibers changed the brittle behavior of
Strip footing cemented sand to a more ductile one. Critical values of reinforcing parameters for maximum reinforcing
Active zone effects are established.
© 2014 Elsevier Ltd. All rights reserved.

1. Introduction
Shallow foundations, such as the roadway or continuous wall
footings for buildings are often placed along the ground surface
above a sheet pile wall. These foundations may impose significant
lateral pressure on the sheet pile wall (Ghanbari and Taheri, 2012).
In earth-retaining problems, the subject of reinforcing the active
zone has attracted a great deal of attention (Santos and
PalmeiraBathurst, 2013). Nasr and Nazir (2013) studied the
behavior of a sheet pile wall embedded in non-cohesive soil rein-
forced by geosynthetics in the active zone adjacent to the strip
footing. The results indicate that the inclusion of geogrid rein-
forcement in the active zone leads to a reduction by about 48%e75%
in the lateral deflection. Nasr (2009) showed that the provision of
discrete vertical reinforcement behind the sheet pile wall has a
significant effect in increasing the stiffness of the soil and decreases
the lateral deflection of the sheet pile wall produced by the lateral
stress of the strip load. Georgiadis and Anagnostopoulos (1998)
used a model sheet pile wall embedded in sand to investigate the
effect of surcharge strip loads on wall behavior. From the results, it
was concluded that the bending moments along the sheet pile wall
E-mail address: amanasrg@hotmail.com.
increase with increasing surcharge load and decreasing strip load
distance to the excavation.
Several stabilization techniques have been used to stabilize and
improve the properties of the sandy soil, but improvement with the
addition of cement is the most generally and successfully used the
technique. Natural or artificial cementation of soil particles con-
tributes to settlement reduction and bearing capacity increase
(Asghari et al., 2003), which are the two key design considerations
in the field of geotechnical engineering. Shear strength behavior of
cemented sands has been widely studied (Schnaid et al., 2001;
Consoli et al., 2012). Ismail (2005) used cemented in situ soil for
retaining walls. According to these studies, cementation of sand
results in increased brittle behavior as peak shear strength in-
creases. However, the high cost of cement motivated the search for
a suitable alternative stabilizing material with low or no cost, such
as cement kiln dust (CKD) for the stabilization of sandy soil. The
high alkalinity and fine particle size of the CKDs, in addition to their
cementitious properties, make these materials suitable for several
applications: waste solidification (Conner et al., 1992; Mckay and
Emery, 1992) and mineral fillers in asphalt paving and mine
reclamation operations. According to Portland Cement Association
(PCA, 2007), more than half million metric tons of CKD were used
for soil stabilization and as a binder in soil stabilized base and sub-
base pavement applications.

http://dx.doi.org/10.1016/j.geotexmem.2014.10.004
0266-1144/© 2014 Elsevier Ltd. All rights reserved.
600 A.M. Nasr / Geotextiles and Geomembranes 42 (2014) 599e610

The behavior of fiber reinforced sand has been studied by a
number of researchers (Park and Tan, 2005; Chauhan et al., 2008;
Ahmad et al., 2010; Diambra et al., 2010; Lovisa et al., 2010;
Falorca and Pinto, 2011; Tang et al., 2007). These studies showed
that adding fiber to sandy soil results in greater peak shear strength
and more ductile behavior. In particular, when applying cemented
soils at a shallow depth, the degree of brittle failure may be more
pronounced due to a low confining stress. A number of studies have
also reported on the influence of both cement and fiber on the
mechanical behavior of sandy soils. Maher and Ho (1993) con-
ducted static and dynamic triaxial compression and extension tests
on cemented sand containing randomly distributed fibers and re-
ported more shear strength and energy absorption for soil con-
taining fiber. Consoli et al. (1998, 2004) conducted consolidated
drained triaxial tests on sand reinforced with fiber and cement.
They found that the addition of fiber increased peak and residual
shear strength and reduced residual dilation. The addition of fiber
to cemented soil produces bonding and friction between the soil
and the fibers. A fiber-reinforced cemented soil can sustain a load
even after failure of a cemented soil and thus, can effectively
improve the brittle behavior of the soil.
From the above literature, only limited information has been
reported on the use of fiber-reinforced cemented sand in active
zone behind the sheet pile wall. Therefore, the present work de-
scribes a study of the behavior of the strip footing resting on fiber-
reinforced cemented sand adjacent to the sheet pile wall.
2.1.2. Fiber
Polypropylene fiber is the most widely used inclusion in the
laboratory testing of soil reinforcement (Ple and Le, 2012; Tang
et al., 2007). Therefore, white monofilament polypropylene fibers
with circular cross section were used throughout this investigation
to reinforce the cemented sand. Short fiber was used in this present
study to be consistent with model tests. The fibers were 12 mm in
length and 0.023 mm in diameter (consequently aspect ratio of
521), with a specific gravity of 0.91, tensile strength of 350 MPa, and
elastic modulus of 6 GPa. Also, the fiber used is a hydrophobic and
chemically inert material which does not absorb or react with the
soil moisture.
2.1.3. Cement kiln dust (CKD)
The chemical composition of CKD used in this study was
determined by X-ray fluorescence and is given in Table 2. For
comparative purposes, the chemical composition of ordinary
Portland cement (from the same company) is also included in
Table 2. Relative to cement compounds, CKD contains the main
oxides in proportions that produce intrinsic cementitious proper-
ties as indicated by the calculated hydration modulus (Kamon and
Nontananandh, 1991). The beneficial properties of CKD and its cost
effectiveness compared with other types of stabilizers have led to
its use as a popular stabilization agent in recent times (Miller and
Azad, 2000). Kamon and Nontananandh (1991) have suggested
that in order that reactions take place, the hydration modulus must
be greater than 1.7.

Cao; %
2. Experimental investigation Hydration modulus ¼
1:7 (1)
Sio2 ; % þ AL2 o3 ; % þ Fe2 o3 ; %

2.1. Material used for the testing
2.1.1. Sand
The experiments were carried out on clean, oven dried,
commercially available sand. Mineralogical analysis showed that
sand particles are predominantly quartz. According to the Unified
Soil Classification System (ASTM, 2010), the soil is classified as
poorly graded sand with letter symbol (SP). The sand was placed in
a test tank at a unit weight of 15.57 ± 0.15 kN/m3, which corre-
sponds to relative density (Dr) of 50%. The friction angle of the sand
(Ø) at 50% relative density, as determined from standard Uncon-
solidated Undrained (UU) triaxial compression tests on a dry sand
sample, was found to be 36 . Other physical properties of the sand
are summarized in Table 1.
Table 1
Physical and mechanical properties of sand used in the model tests.
Property
Based on the values of each component in Table 2, the hydration
modulus (Eq. (1)) of CKD and Portland cement are found to be 2.30
and 2.17, respectively. This indicates that CKD can be considered a
soil stabilizer.
2.2. Test tank
A series of laboratory model tests was executed in a test tank
made of steel, having inside dimensions of 1.50 m long, 0.5 m wide,
and 0.9 m high. These tank dimensions were chosen to ensure that
the failure zone did not extend up to the walls (based on pre-
liminary results of Plaxis 2D and other investigators such as Nasr
and Nazir, 2013. To ensure the rigidity of the tank, the sides of
the tank were braced with vertical and horizontal stiffeners to
prevent any deflection during the loading process. Furthermore, to
achieve almost frictionless side faces, the inside fixed walls of the
tank were polished smooth. Then, the longitudinal sides of the tank
were coated with lubricating oil in order to reduce the effect of side
Value
wall friction. A schematic diagram of the model test configuration is

Specific gravity, Gs 2.64

Grain size analysis Table 2
Effective grain size, D10 (mm) 0.10 Chemical composition of CKD and ordinary Portland cement.
Average grain size, D50 (mm) 0.30
Uniformity coefficient, Cu 3.80 Chemical composition Analysis method Composition (% by weight)
Coefficient of curvature, Cc 1.05
CKD OPC
Coarse to medium sand, % 30.0
Fine sand, % 68.5 SiO2 XRF 15.42 21.20
Fines (<0.075 mm) 1.50 Al2O3 XRF 3.92 4.95
Classification (USCS) SP Fe2O3 XRF 2.95 2.82
Maximum dry unit weight, gd max (kPa) 17.10 CaO XRF 51.23 62.81
Minimum dry unit weight, gd min (kPa) 14.30 MgO XRF 2.73 4.0
Maximum void ratio, emax 0.811 SO3 XRF 4.36 2.63
Minimum void ratio, emin 0.515 Na2O3 XRF 0.35 0.2
Compaction study K2O XRF 1.91 0.3
Optimum moisture content, Wc(%) 9.50 Loss on ignition (LOI) ASTM C575 19.82 1.72
Maximum dry density, (kPa) 16.90 Specific gravity, Gs ASTM D 854 2.78 3.13
 A.M. Nasr / Geotextiles and Geomembranes 42 (2014) 599e610
Fig. 1. Schematic diagram of the model test configuration (not to scale).
shown in Fig. 1. A raining device consisting of an upper steel box
(raining box) was designed and constructed in the loading frame to
obtain the desired density of sand.
2.3. Model footing
The footing was 499 mm in length, 100 mm width, B, and
25 mm in thickness. The length of the footing was made almost
equal to the width of the tank in order to maintain the plane strain
conditions. All tests were performed with the footing resting on
the sand surface (Df ¼ 0.0) in the active zone. To overcome po-
tential error due to friction on the sides of the tank, the ends of the
footing box section were polished smooth to minimize the end
friction effect. In order to simulate a valid, real condition, a rough
base condition was achieved by attaching sandpaper to the bottom
of the footing.
2.4. Sheet pile wall
The sheet pile wall consisted of a flexible steel plate 750 mm
in length, 499 mm width, and 3.3 mm in thickness. The sheet pile
was embedded 500 mm into the sand bed. The base of the sheet
pile wall was tapered to facilitate the penetration into the soil.
The sheet pile wall was polished smooth to minimize the end
friction effects and friction with the soil bed. The flexural stiff-
ness of the model sheet pile wall (EI) about the ZeZ axis was
0.629 kN m2/m.
2.5. Loading system
A rigid loading frame was used in the experimental work. The
footing was loaded by a hand operated hydraulic jack of 50 kN
capacity. To push the footing slightly into the sand bed, load
applied through the jack was recorded with the help of a pre-
calibrated proving ring fitted to the piston of the jack. In order
to record the correct settlement of the footing after application of
the load, two dial gauges with a measurement of at least
0.01 mm were used. Also, to measure the lateral deflection along
the sheet pile wall, four LVDT'S were placed in the center of the
sheet pile wall at different elevations in the passive zone as
shown in Fig. 1.
601
2.6. Preparation of reinforced-sand layer
This section describes briefly the stages followed in preparation
of the reinforced e sand layer. The cement kiln dust ratio (RCKD) is
based on dry sand defined as follows:
RCKD ¼ ðWCKD =WS Þ  100ð%Þ (2)
where WCKD is the weight of CKD and WS is the weight of dry sand.
The different values adopted in the present study for RCKD are 3, 6, 9,
and 12%. The fiber ratio (RF) is based on the amount of the fiber used
in each fiber inclusion layer. It is defined as follows:
RF ¼ WF =ðWS þ WCKD Þ  100ð%Þ (3)
where WF is the weight of fiber. The reinforced-sand layer was
prepared at four fiber ratios, namely, 0.25, 0.5, 0.75, and 1.0%.
The fiber-reinforced cemented sand used in the experiments
was prepared by hand-mixing dry sand, CKD, polypropylene fibers
with optimum moisture content of 9.5% (as calculated from
compaction test). Proper care was taken to prepare homogenous
mixtures at each stage of mixing.
2.7. Sand deposit preparation
The inside walls of the tank were marked at every 100 mm.
Then, the sand was placed in 100 m layers of thickness using a
raining technique through the air. The height of fall to achieve the
desired unit weight was determined a priori by performing a series
of trials with different heights of the fall. In each trial, the unit
weights were calculated by collecting samples in small cups of
known volume placed at different locations in the test tank. In all
model tests, the average unit weight and relative density of the
sand were kept as 15.57 ± 0.15 kN/m3 and 50%, respectively.
2.8. Test procedures
A sand bed, initially 600 m deep (100 mm for each layer), was
prepared in the test tank using the sand raining technique as dis-
cussed previously. The position of the sheet pile wall was marked
vertically for reference. Then, the model sheet pile wall was driven
vertically into place. A supporting wooden box was made to guide
the sheet pile wall on one side of the wall (passive zone). The sand
602 A.M. Nasr / Geotextiles and Geomembranes 42 (2014) 599e610

Table 3
Details of parametric study.

Series Study on effect of Constant parameters Variable parameters

A Unreinforced active zone (no fiber) X/H ¼ 0.6, U/H ¼ 0.3, Time ¼ 0.0 RCKD (%) ¼ 3, 6, 9, 12
B Fiber ratio þ CKD ratio RCKD ¼ 3%, X/H ¼ 0.6, U/H ¼ 0.3, Time ¼ 0.0 RF (%) ¼ 0.25,0.5, 0.75, 1.0
RCKD ¼ 6%, X/H ¼ 0.6, U/H ¼ 0.3, Time ¼ 0.0 RF (%) ¼ 0.25,0.5, 0.75, 1.0
RCKD ¼ 9%, X/H ¼ 0.6, U/H ¼ 0.3, Time ¼ 0.0 RF (%) ¼ 0.25,0.5, 0.75, 1.0
RCKD ¼ 12%, X/H ¼ 0.6, U/H ¼ 0.3, Time ¼ 0.0 RF (%) ¼ 0.25,0.5, 0.75, 1.0
C Reinforced layer thickness RCKD ¼ 9%, X/H ¼ 0.6, Time ¼ 0.0, RF ¼ 0.50% U/H ¼ 0.1, 0.2, 0.3, 0.4, 0.5
RCKD ¼ 9%, X/H ¼ 0.6, Time ¼ 0.0, RF ¼ 0.75% U/H ¼ 0.1, 0.2, 0.3, 0.4, 0.5
RCKD ¼ 9%, X/H ¼ 0.6, Time ¼ 0.0, RF ¼ 1.00% U/H ¼ 0.1, 0.2, 0.3, 0.4, 0.5
D Footing location relative to SPW RCKD ¼ 9%, U/H ¼ 0.3, Time ¼ 0.0, RF ¼ 0.50% X/H ¼ 0.4, 0.6, 0.8, 1.0
RCKD ¼ 9%, U/H ¼ 0.3, Time ¼ 0.0, RF ¼ 0.75% X/H ¼ 0.4, 0.6, 0.8, 1.0
RCKD ¼ 9%, U/H ¼ 0.3, Time ¼ 0.0, RF ¼ 1.0% X/H ¼ 0.4, 0.6, 0.8, 1.0
E Curing time of reinforced layer RCKD ¼ 6%, U/H ¼ 0.3, RF ¼ 0.75%, X/H ¼ 0.6 Time ¼ 0.0, 7, 14, 21 days
RCKD ¼ 9%, U/H ¼ 0.3, RF ¼ 0.75%, X/H ¼ 0.6 Time ¼ 0.0, 7, 14, 21 days

raining process continued until the required height behind the represented by a non-dimensional parameter called the bearing
sheet pile wall (active zone) was achieved. The fiber-reinforced capacity ratio (BCR) and described as.
cemented sand was also prepared as explained previously. At a
maximum dry density (gd max ¼ 16.90 kN/m3, as calculated from
quðfiberÞ
BCR ¼ (4)
compaction test), the quantity of fiber-reinforced cemented sand quðnofiberÞ
for the required thickness layer was deposited in the active zone.
The reinforced layer was compacted to reach the required layer where qu (fiber) and qu (no fiber) ¼ ultimate bearing capacities of the
thickness by using a flat bottom hammer (0.1 m  0.1 m, weighting footing on fiber reinforced and unreinforced cemented active zone,
20 N). As shown in Fig. 1, after preparing the reinforced layer, the respectively.
surface was leveled, and the footing was placed in the desired Another factor to consider the settlement characteristics of the
position. The guiding wooden box remained in place (passive zone) system is the “deformability index” is proposed and defined to
until the model was fully prepared. Then, the supporting wooden describe the ductility of fiber-reinforced cemented sand layer as
box was removed. Two dial gauges were placed at equal distance follows:
from the edges of the hydraulic jack. Finally, after the wooden box
was removed, LVDT'S were placed to measure the lateral deflection SðfiberÞ
DI ¼ (5)
of the sheet pile wall (see Fig. 1). The footing was loaded statically SðnofiberÞ
using the hydraulic jack and the average vertical displacements
measured by gauges were taken. The lateral deflection of the sheet where S(fiber) and S(no-fiber) ¼ vertical surface settlement obtained at
pile wall at each load was recorded. the ultimate bearing capacity conditions for fiber reinforced and
unreinforced cemented sand in the active zone, respectively.

2.9. Test program

3.1. Effect of cement kiln dust inclusion on the behavior of strip
footing
A total of 47 tests were conducted to study the effect of fiber-
reinforced cemented sand in the active zone on the behavior of
Since the primary objective of the experimental investigation is
strip footing and sheet pile wall. First, the response of the model
to evaluate the effect of fiber-reinforced cemented sand in
footing constructed on the unreinforced active zone (without fiber
improving the load carrying capacity of the strip footing adjacent to
reinforcement) was studied at different CKD ratios to obtain the
necessary reference data for the purpose of comparing on (series A).
Then, four series of tests (from B to E) were performed to study the
effects of different parameters on the strip footing and the sheet
pile wall behavior. Each series was conducted to study the response
of one parameter while the other variables were remaining con-
stant. Table 3 shows the details of the parametric study.

3. Interpretation and discussion of results

The footing pressure-settlement curves were obtained from the

results of the experimental model strip footing tests on the active
zone of sheet pile wall. The ultimate bearing capacities for the
model footing were determined from the footing pressure-
settlement curves as the pronounced peaks and after which the
footing collapsed. In curves that did not exhibit a definite failure
point, the ultimate bearing capacity was taken as the point at which
the slope of the footing pressure-settlement curve first reaches zero
or reaches a steady minimum value (Vesic, 1973). In order to
evaluate the fiber reinforcement effects, the bearing capacity Fig. 2. Variation of footing stress q with settlement for different RCKD (unreinforced
improvement of footing on fiber-reinforced cemented sand is case).
 A.M. Nasr / Geotextiles and Geomembranes 42 (2014) 599e610 603

Fig. 3. Variation of q with settlement for physical model strip footing with different RF (reinforced case).

a sheet pile wall; it is convenient to present the results of the
footing on cemented sand without any reinforcement (unrein-
forced case). Initially the tests were conducted with strip footing
placed on the top of the unreinforced cemented sand layer pre-
pared at maximum dry density and optimum moisture content. The
tests were carried out by keeping the footing distance from the
sheet pile wall X/H ¼ 0.6 and thickness of cemented layer U/H ¼ 0.3
constants (series A). Typical footing stressesettlement curves for
four different CKD ratios are presented in Fig. 2. It can be observed
from Fig. 2 that the CKD ratio (RCKD) has a great effect on the ulti-
mate bearing capacity of the strip footing. The ultimate bearing
capacity increases with an increase in CKD ratio. The cemented
Table 4
Summary of the results for fiber-reinforced cemented sand in active zone tests
(Series B).
sand layer had 3%, 6%, 9%, and 12% (by weight of dry soil) of CKD
ratio, having ultimate bearing capacities, qu, of 150, 191, 239, and
265 kPa, respectively. It is probably for the following reason: the
additional of cement kiln dust binds sand particles together and
makes the matrix more compacted, and thus restricts the rear-
rangement of particles and increases the effective contact area.
Furthermore, the cement kiln dust provides tensile strength capa-
bility, which will increase cohesion values.
In all tests, general shear failures happened as expected in the
compacted cemented sand layer and load dropped after failure.
Vertical settlement values at failure (Sf) increased as CKD increased
until 9%. At 12% CKD ratio, the vertical settlement at failure is
9.25 mm, which is smaller than that at 6% and 9% CKD ratios. Thus,
at 12% CKD ratio, the cemented sand layer exhibited a more brittle
behavior.

RCKD RF, qu, Settlement Max. Lateral deflection BCR DI 3.2. Effect of fiber ratio on the behavior of strip footing
(%) (%) (kPa) at Peak, (mm) at peak, (mm)

3 0.25 167 9.73 6.53 1.11 1.12 In this series of tests, the effect of fiber inclusion on cemented
0.50 198 10.40 4.50 1.32 1.20 sand was investigated. The tests were carried out using four
0.75 234 10.83 3.80 1.56 1.25 different percentages of fiber ratios 0.25%, 0.5%, 0.75%, and 1.0%
1.00 259 11.28 3.53 1.73 1.29 (series B). Fig. 3 shows the behavior of the strip footing on the top of
6 0.25 219 12.75 5.20 1.15 1.25
0.50 285 14.28 3.83 1.49 1.40
fiber-reinforced cemented sand and adjacent to the sheet pile wall.
0.75 323 15.96 3.17 1.69 1.56 A summary of the test results is given in Table 4. It is readily
1.00 349 16.55 2.97 1.83 1.62 observed that fiber reinforcement improves the footing-settlement
9 0.25 295 12.50 3.96 1.23 1.14 behavior. In other words, at any CKD content, footing supported on
0.50 391 14.00 3.26 1.64 1.27
fiber-reinforced cemented sand provided considerable higher ul-
0.75 471 15.35 2.83 1.97 1.40
1.00 520 16.93 2.59 2.17 1.54 timate loads than the corresponding cemented sand only without
12 0.25 342 11.17 3.35 1.29 1.21 fiber reinforcement. This can be attributed to both the internal
0.50 463 12.19 2.75 1.75 1.31 friction angle and cohesion intercept increase when fiber was
0.75 575 14.79 2.56 2.17 1.60 added (Amir and Mahdi, 2013). The addition of fiber to cemented
1.00 655 16.34 2.47 2.47 1.77
sand produces bonding and friction between the soil and the fibers.
604 A.M. Nasr / Geotextiles and Geomembranes 42 (2014) 599e610
Fig. 4. Variation of BCR and DI with fiber ratio for different CKD ratios.
Amir and Mahdi (2013) stated that, at 1.0% polypropylene fibers, the
internal friction angle of cemented sand increased from 39 to 41.5
for 50% relative density. Furthermore, the failure of the fiber rein-
forced cemented sand occurred due to slippage between the sand
particles and the fibers. For cemented sand, the use of fiber seems
to produce a strong frictional interaction between the fibers and
the sand particles.
Upon comparison with unreinforced cemented sand (RF ¼ 0.0%),
it can be seen that the inclusion of fibers within the cemented sand
reduces the brittleness of the response. The vertical settlement at
failure increased with increasing in fiber ratio. It is clear that at
RCKD ¼ 3%, the footing stressesettlement curves at different fiber
contents did not exhibit a peak point (failure point). This result
indicates that the addition of fiber delay failure and mitigates the
brittle behavior. The above results are in good agreement with the
findings of Consoli et al. (1998, 2004). In case of fiber-reinforced
cemented sand, RCKD ¼ 9% and 12%, the footing stress increases
with an increase in vertical settlement until the peak value is
reached, followed by reduction of post-peak stress gradually. Un-
doubtedly, one of the main advantages of fiber reinforcement when
applied to soil is the improvement in material ductility. The above-
mentioned behavior of fiber-reinforced cemented sand explains
why the fiber in cemented stabilized sand shows higher
Fig. 5. Lateral deflection along the sheet pile wall at different fiber ratios.
reinforcement benefit than that in cemented sand without fiber
reinforcement.
The variations of bearing capacity ratio, BCR (Eq. (4)), and
deformability index, DI (Eq. (5)), with fiber ratio are shown in Fig. 4
for different CKD ratios. From Fig. 4-A, it is evident that at any CKD
ratio, the fiber ratio has a significant effect on BCR. The bearing
capacity ratio increases with increasing fiber ratio. It is clear that
the effect of the CKD ratio on BCR is minimized when the footing is
placed on reinforced sand with RF ¼ 0.25%. At RF ¼ 0.25%, the BCR
increased by about 16% when the CKD ratio increased from 3% to
12%. On the contrary, at RF ¼ 1.0%, the BCR increased by about 42%
and the effect of the CKD ratio on BCR is more pronounced.
Furthermore, Fig. 4-A shows that at any fiber ratio the BCR of a
footing on a fiber-reinforced cemented sand is considerably higher
than that of a footing on an unreinforced cemented sand where BCR
is greater than 1.0 (BCR > 1.0). Thus, reflecting the beneficial effect
of fiber reinforcement in improving the bearing capacity of footing
in active zone adjacent to the sheet pile wall.
Fig. 4-B illustrates the variation of deformability index DI (Eq.
(5)) with fiber ratio for different CKD ratios. The results clearly
indicate that the DI increases with increasing the fiber ratio. The
value of DI increases from 1.12 to 1.77, as the fiber ratio increases
from 0.25% to 1.0%. However, at low levels of the fiber ratio
(RF  0.5%), DI was weakly influenced by the cement kiln dust ratio.
The deformability index can be applied to evaluate the ductility or
deformability of cemented sand that is reinforced by various fibers.
This index may indicate the degree of fiber extensibility in
cemented sand and aid in the selection of suitable fibers for a given
cementation. In the case of RCKD ¼ 3%, the variation of DI with fiber
ratios was minimal.
The main observation of the results of the experimental work
was the deformation of the sheet pile wall. From Table 4, the var-
iations of the maximum lateral deflection of the sheet pile wall at
ultimate bearing capacity with fiber ratio at four different CKD ra-
tios are presented. The results indicate that the fiber ratio, RF, has a
major influence on the maximum lateral deflection of the sheet pile
wall. Increasing the fiber ratio, RF, decreases the maximum lateral
deflection at peak stress. At cement kiln dust ratio, RCKD ¼ 9%, the
maximum lateral deflection decreased by about 51% and 55% at
fiber ratio of 0.75% and 1.0%, respectively. It can be concluded that
by using fiber ratio, RF ¼ 0.75%, polypropylene fiber reinforcement
provided a beneficial reduction in the lateral behavior of the sheet
pile wall adjacent strip footing. It is clear that the decrease in the
maximum lateral deflection values was reduced significantly when
the RF ratio was increased to 0.75%. At the same time, increasing the
CKD ratio decreases the maximum lateral deflection of the sheet
pile wall at peak stress. From the LVDT readings along the free
 A.M. Nasr / Geotextiles and Geomembranes 42 (2014) 599e610 605

length of the sheet pile wall, Fig. 5 shows the lateral deflection
along the sheet pile wall at different fiber ratios. The CKD ratio was
kept constant at 9%. The figure demonstrates that increasing the
fiber ratio not only decreases the maximum lateral deflection at the
ground surface, but also decreases the lateral deflection along the
length of the sheet pile wall. At the end of a fiber-reinforced
cemented sand layer, the lateral deflection decreased by about
65% when the fiber ratio increased from 0.25% to 1.0%.

3.3. Effect of thickness of fiber-reinforced cemented sand layer

The variation of ultimate bearing capacity, qu, with the thickness

of the fiber-reinforced cemented sand layer, U/H, is shown in Fig. 6.
From the aforementioned figure, it is found that the value of qu
depends markedly on the thickness of the fiber-reinforced
cemented sand layer. It is clear that increasing the thickness of
the reinforced layer will increase the ultimate bearing capacity, qu.
For the three RF ratios used in this series of tests, most of the in-
crease in the qu takes place between U/H ratios of 0.1 and 0.4. After
that, for higher fiber ratios (RF
0.75%) and at U/H ¼ 0.5, the in-
crease in qu is not significant, and the effect of fiber reinforcement
on qu was diminished. This can be attributed to shearing resistance
and cohesion between sand particles (Amir and Mahdi, 2013)
caused by the fiber reinforcement in the upper cemented sand
layer, which intersects the collapse mechanism (shear zone resis-
tance) of the soil under the footing. At the same time, by increasing
U/H, the area of the fiber-reinforced cemented sand that has greater
shear strength increased under the footing. However, at RF ¼ 0.75%
and 1.0%, it is clear that using fiber-reinforced cemented sand layer
with thickness U/H ¼ 0.4 is more effective than that at U/H < 0.4 in
increasing the ultimate bearing capacity of the strip footing adja-
cent to the sheet pile wall. For example, at U/H ¼ 0.4 and fiber
ratio ¼ 0.75%, the ultimate bearing capacity increased by about
121% than that at U/H ¼ 0.1.
To investigate the effect of U/H on the lateral deflection along the
sheet pile wall, the LVDT readings were recorded, as shown in Fig. 7.
The increase in the thickness of the fiber-reinforced cemented sand
layer decreased the lateral deflection along the sheet pile wall.
Furthermore, the results in Fig. 7 clarify that the lateral deflection
values and the lateral deflection shape along the sheet pile wall
strongly depend on the thickness of the reinforced layer. As the
lateral deflection is the main cause of failure of sheet pile walls
(Georgiadis and Anagnostopoulos, 1998), using fiber reinforcement
can improve the overall behavior of a sheet pile wall. It was evident
that in the case of U/H ¼ 0.1, 0.2, and 0.3, the top of the sheet pile
Fig. 6. Variation of ultimate bearing capacity qU with U/H at different fiber ratios.
Fig. 7. Lateral deflection along the sheet pile wall at different U/H ratios.
was the most affected zone by lateral deflection. In contrast, when
the thickness of the fiber-reinforced layer increased to 0.4 and 0.5,
the top lateral deflection decreased significantly. It is also of interest
to note that the increase of the thickness of reinforcement layers to
U/H ¼ 0.5 tends to cause a shift of the maximum lateral deflection
from the top of the sheet pile to a deeper depth along the sheet pile
wall.
3.4. Effect of footing location relative to the sheet pile wall
A series of tests was performed to study the effect of strip
footing location behind and parallel with the cantilever sheet pile
wall. The variations of ultimate bearing capacity, qu, and maximum
lateral deflection at different footing locations are shown in Fig. 8A-
B, respectively. From Fig. 8-A, the results clearly indicate that at any
fiber ratio, the ultimate bearing capacity, qu, of the strip footing
adjacent to a sheet pile wall increases with increasing distance
between the footing and the sheet pile wall. This clearly indicates
the adverse effect of sheet pile wall on footing behavior. Further-
more, at any X/H ratio, the qu increase considerably due to reflecting
the beneficial effect of fiber reinforcement in improving the
behavior of the strip footing adjacent to the sheet pile wall. Also, for
fiber ratios of 0.75% and 1.0%, most of the increase in the qu takes
place between X/H ratios of 0.4 and 0.8. After that, if X/H is more
than 0.8 there is no appreciable increase in the ultimate bearing
capacity of the footing (about 4.8%).
Strip footing supported by a uniform backfill soil and subjected
to a central vertical loading will undergo lateral displacement only
when their lateral supports are not equal on both sides. When a
footing is nearby a wall, the lateral support provided by the wall
may be smaller or greater than that on the other side depending on
the footing location and wall stiffness. The variation of maximum
lateral deflection of sheet pile wall with different X/H ratios is
presented in Fig. 8-B. From Fig. 8-B, the maximum lateral deflection
decreases with increasing X/H for any fiber ratio. Markedly, for
fiber-reinforced active zone and at X/H ¼ 0.8, the maximum lateral
deflection decreased relative to the case of X/H ¼ 0.4 by about 52%,
54%, and 47% for fiber ratios of 0.5%, 0.75%, and 1.0%, respectively.
However, the rate of decrease of maximum lateral deflection
decreased with increasing X/H ratio to 1.0. At X/H ¼ 1.0, the effect of
fiber ratio in reducing the maximum lateral deflection of the sheet
606 A.M. Nasr / Geotextiles and Geomembranes 42 (2014) 599e610
Fig. 8. Variation of ultimate bearing capacity and maximum lateral deflection with X/H at different fiber ratios.
pile wall can be neglected. The above results agree with the findings
of Nasr and Nazir (2013).
3.5. Effect of curing time on the behavior of strip footing
Obviously, when CKD-mixed with sand, the cemented sand will
continue to improve with time (like concrete). Therefore, a series of
tests were conducted to study the effect of curing time on the
bearing capacity of the strip footing resting on fiber-reinforced
cemented sand layer. The fiber ratio was kept constant at 0.75%
and two different CKD ratios of 6% and 9% were used (series E).
Ultimate bearing capacity, qu, is plotted against curing time in Fig. 9.
The results clearly indicate that, for both CKD ratios, the ultimate
bearing capacity increases with increasing curing time. This is ex-
pected since fiber-reinforced cemented sand mixture requires time
to gain strength (Sulapha et al., 2008). Also, it can be seen that the
value of qu rapidly increased at the beginning of the curing time
(Tang et al., 2007). It is clear from the above figure that the greatest
rate of strength gain occurs within the first 14 days. For example, at
6% CKD ratio, the ultimate bearing capacity, qu, at 21 days is 1.46
times the qu after 7 days of curing, and 1.07 times after 14 days of
curing. At 9% CKD ratio, the qu at 21 days is 1.36 times the qu after 7
days of curing, and 1.1 times after 14 days of curing. This can be
attributed to higher fineness and the surface area of CKD. The same
results were obtained by Miller and Azad (2000). Therefore, sig-
nificant increases in qu occurred with increases in the CKD ratio in
reinforced layer at different curing times.
Fig. 9. Variation of ultimate bearing capacity qU with curing time for different CKD
ratios.
4. Finite-element analysis
A series of two-dimensional (2D) finite element analyses (FEAS)
on a model-scale strip footing and sheet pile wall system was
performed to supplement the results of the model laboratory tests.
The dimensions of strip footing and sheet pile wall modeled in the
FEAS are similar to the dimensions used in the experimental model.
Plane strain elasto-plastic FEAS were performed using the com-
mercial program PLAXIS 7.2 (Bringkgreve and Vermeer, 1998). Note
that the FE model was assumed to be constructed using the same
materials used in the laboratory tests. The dimensions of the mesh
size were chosen large enough to ensure that the boundaries will
have no effect on the calculated results (width of mesh ¼ 15B, and
height of mesh ¼ 8.5B). The analyzed strip footing, sheet pile ge-
ometry, and generated mesh are shown in Fig. 10.
4.1. Finite element modeling
A variety of soil models are built into the computer code chosen
for this study. However, it was decided to use the non-linear
MohreCoulomb criteria to model the sand for its simplicity, prac-
tical importance and the availability of the parameters needed. The
MohreCoulomb soil model parameters used in the FEAs were
derived from a series of laboratory tests as shown in the reduced
scale laboratory model. For pure sand, an internal friction angle (Ø)
of 36 and a dilatancy angle (j) of 6 were used, where j ¼ Ø  30 .
For fiber reinforced-cemented sand, the results of material tests
which used in the FE analysis are reported in Table 5. The analyses
were carried out for draining condition without pore water pres-
sure and the FE calculations for soil elements were based on 15-
node triangular elements. From the Plaxis manual, to avoid com-
plications during a numerical process, enter at least a small value
for cohesion (c > 0.2 kPa), but its effect on the resulting collapse
load values is negligible (less than 0.1%). Therefore, the cohesion in
all calculation process was taken 1.0 kPa. Table 5 summarizes the
parameters used in the analysis.
The linear elastic behavior of the sheet pile wall and the footing
were modeled as plate elements. Plates in the 2D finite element
model are composed of beam elements with three degrees of
freedom per node. When 15-node soil elements are employed then
each beam element is defined by 5-nodes. The beam elements are
based on Mindlin's plate theory. This theory allows for beam de-
flections due to shearing as well as bending. For the modeling of
elastic plates, an axial stiffness, EA, and a flexural rigidity, EI, must
be specified as material properties (Table 5).
The interaction between the sheet pile wall, footing, and the
surrounding soil was simulated by interface elements located at the
contact between any two elements. The interface elements, which
 A.M. Nasr / Geotextiles and Geomembranes 42 (2014) 599e610 607

Fig. 10. Mesh used in the finite-element analysis.

Table 5
Material parameters used in the finite element analysis.

Parameter Pure sand Fiber-R.C. sand Footing Sheet pile

Material model MohreCoulomb model MohreCoulomb model Linear elastic Linear elastic
Type of material behavior Drained Drained Non-porous Non-porous
ref
Primary loading stiffness, E50 (kPa) 30,000 35,000 21  107 21  107
Poisson's ratio, n 0.30 0.30 0.20 0.2
Cohesion, c (kPa ) 1.00 (45, 66, 92, 145, 203)* e e
Friction angle, Ø ( ) 36 (39, 41, 42.5, 44.6, 45)* e e
Dilatancy angle, j ( ) 6 (9, 11, 12.5, 14.6, 15) e e
Interface reduction factor, Rint 0.75 0.75 e e
Failure ratio, Rf 0.90 0.90 e e
Axial stiffness, EA (kN/m) e e 420,000 2.1  106
Flexural stiffness, EI (kPa/m) e e 14 0.629

( )*Results from triaxial tests at RCKD ¼ 9% and RF varied from 0.0 to 1.0%.
( )*Tests were conducted at max. dry density and optimum moisture content.

have zero physical thickness, were connected to soil elements by that the difference in the ultimate bearing capacity calculated from
five pairs of nodes. The stiffness matrix of interface elements was finite element method and that calculated from experimental tests
obtained using Newton-Cotes integration points. The interface was about 5% and 14% at fiber ratio 0.75% and 0.25%, respectively. It
parameters were simulated in PLAXIS by assigning a suitable value is also clear that by increasing the fiber ratio, the behavior of the
(as stated in program manual) for the strength reduction factor in strip footing at failure was more ductile. For practical purposes, the
the interface compared with the corresponding soil strengths.
The modeled boundary conditions were assumed to fix parts of
the model in the directions of X and Y. All the boundary conditions
assumed in the physical model were simulated in the FE model.
Therefore, the vertical geometry lines of the model were con-
strained horizontally (ux ¼ 0.0), and the horizontal bottom line was
constrained in both the horizontal and vertical directions
(ux ¼ uy ¼ 0.0). The load was applied uniformly on the footing
(beam element) by using the distributed load option.

4.2. Results of finite-element analysis

The results of the finite-element analyses are shown in

Figs. 11e14. In these series, all the tests were performed on the
footing placed at U/H ¼ 0.3 and a CKD ratio of 9%, while the fiber
ratio, RF, and footing distance from the sheet pile wall, X/H, were
varied. Fig. 11 shows the variations of footing stress with a footing
settlement for different fiber ratios. The figure clearly shows that by
increasing the fiber ratio, much has improved for both the initial
stiffness (initial slope of the load-settlement curve) and the ulti-
mate bearing capacity of the footing. The above results agree with Fig. 11. Variation of q with settlement for FE model sheet pile wall with different fiber
the experimental results in terms of the fiber ratio. It was evident ratios.
608 A.M. Nasr / Geotextiles and Geomembranes 42 (2014) 599e610
Fig. 12. Comparison of BCR and DI with various fiber ratios as obtained from the physical model and FEAs.
optimum value of fiber ratio, was RF ¼ 0.75%, whereas the increase
in the ultimate bearing capacity between RF ¼ 0.75 and RF ¼ 1.0 did
not exceed 10%.
The BCRS with different RF were calculated by using the FEAS
(model scale) and compared with the results obtained from model
experimental tests as shown in Fig. 12-A. The finite element results
provide a reasonable fit with the experimental data (BCR obtained
from the FEA was defined in the same way as defined in the model
tests). The BCR increases with the increase of the fiber ratio, RF, for
both experimental and numerical results. Although, the BCR ob-
tained from the FEA appears to be smaller than that for model scale
footing, the maximum difference in BCR between experimental
results and FEAS was about 14% at RF ¼ 1.0. Fig. 12-B shows the
variation of DI with various RF obtained from model tests and FEAS.
This figure demonstrated that the results obtained from the
experimental tests are in close agreement with those from the
finite-element method. It is clear also that at RF ¼ 1.0%, the
maximum difference in DI between experimental results and FEA
was 11.6%. However, the finite element analyses showed that the
rate of the increase in DI reduces with increasing the RF to 1.0.
Comparison between experimental and numerical results for
ultimate bearing capacity, qu, at different X/H ratios can be seen in
Fig. 13. Finite element results also indicate that the ultimate bearing
capacity of the strip footing increases with increasing distance
between footing and sheet pile wall, X/H. As can be seen, the results
are similar to experimental values. It is observed that the experi-
mental results curve is above the curve from the numerical
modeling. This can be attributed to the soil parameters used in the
Fig. 13. Comparing Plaxis and laboratory test results of ultimate bearing capacity at
various X/H ratios.
numerical analyses such as modulus of elasticity and the method
used to model the soil in the analysis. The maximum difference in
qu between experimental and numerical results was about 10% at X/
H ¼ 1.0. Based on the results of the FEAS, the distribution of
displacement vectors and contours of the deformation in the active
zone under the footing and behind the sheet pile wall is shown in
Fig. 14 at different X/H ratios. At X/H  0.6, it is clear from Fig. 14 that
the displacement vectors under the footing moved horizontally in
the direction of the sheet pile wall. The maximum lateral deflection
occurred at the top of the sheet pile and vanished at about 80% of
the sheet pile length. When the footings are not far behind the
sheet pile walls, the lateral supports under the footing are not equal
on both sides. Furthermore, the footings and the sheet pile walls
may interact resulting in an adverse effect on the stability of the
footings and the walls. When the footing is located farther away at
X/H
0.8 behind the wall, the displacement vectors shown in
Fig. 14 also skew toward the sheet pile wall, but the most of
displacement vectors do not reach the wall. At different X/H ratios,
the difference in the lateral movement of the sheet pile wall be-
comes more evident in the vector plots which decrease as X/H in-
crease. Furthermore, Fig. 14 shows that the shape of the contours of
the deformation along the sheet pile wall and under the footing is
changed relative to the footing location. At X/H
0.8, the de-
formations are concentrated under the footing and approximately
distributed equally on both sides. This helps to explain that, when
X/H
0.8, there is no appreciable increase in the ultimate bearing
capacity of the footing.
5. Scale effects and limitations
Physical models have been commonly used to study bearing
capacity and settlement behavior of foundations because of the
difficulties associated with loading full-scale prototypes to failure.
However, as in all reduced-scale physical model tests, scaling ef-
fects must be minimized to ensure that the behavior observed on a
reduced scale can be extrapolated to prototype behavior (Muir
Wood, 2004). For reduced-scale models, the particle size and
density of sand, construction techniques, boundary conditions,
reinforcement stiffness, and dilatancy at low stress are the impor-
tant factors to be considered.
Considering the model footing size and the scale factor effect,
relatively fine-grained sand (D50 ¼ 0.3 mm) was used for tests.
Habib (1974) stated that the particle size effect has a negligible
effect on bearing capacity at B/D50 ratio less than 200, and no effect
at B/D50 values greater than 200. Kusakabe (1995) summarized test
data and indicated that the particle size effect on the test results
becomes less marked for a D50/B ratio smaller than 0.01. In this
 A.M. Nasr / Geotextiles and Geomembranes 42 (2014) 599e610
Fig. 14. Displacement vectors and contours of the deformation at various X/H ratios.
study, the effect of the particle size should be smaller, as the D50/B
ratio used in the model was 0.003. Furthermore, according to Amy
and Alan (2007), the model-scale test must be performed on sand
that is looser than the sand underneath a full-scale test. Therefore,
all model tests were conducted on sand with relative density 50%.
To satisfy the plane strain conditions (Ko and Davidson, 1973),
first, the friction between soil and the front and back internal sur-
faces of the tank should be zero. Balachandran (1996) stated that
the peak friction angle between sand and treated surfaces de-
creases by about 65% relative to the case of untreated surfaces.
Therefore, the inside walls were polished smooth and coated with
lubricating oil in order to reduce friction as much as possible.
Second, deformation in longitudinal direction should be zero. This
implies that plane strain models should have rigid front and back
planes. Therefore, the deformations of the test tank in longitudinal
direction were reduced, whereas the vertical sides were strength-
ened by using steel angles in the middle and the top of the sides. In
all reduced-scale model tests, relatively wide tanks are used. The
length of the tank was about 15 times the footing width, and the
soil layer thickness under the footing was more than 8 times the
footing width. According to the tank dimensions, it may be seen
there is sufficient space for the development of active and passive
wedges as established from the series of FE analyses. Therefore, the
boundary conditions do not have any significant effect on the re-
sults (Bransby and Smith, 1975).
To simulate the fiber reinforcement in the model tests, proto-
type short fiber with small diameter was preferred for use. There-
fore, the results from the reduced-scale model may be not
influenced. The difference in the stress level between the model
tests and the field tests was the imperative factor invalidating the
use of small-scale models, which have been constructed in sand
and tested at 1-g, for the accurate determination of deformation
behavior. The stress level under the footing in a small-scale model
is much smaller than that under the footing in the field. This low
stress in granular soil corresponds to a greater angle of internal
friction than that at higher stress level. In consequence, the results
of reduced-scale models should be used with extreme caution.
Some components of the physical model verified in this study
are reduced to a certain scale and therefore, does not relate to
prototype footing-sheet pile wall systems faced in the field.
Nevertheless, the goals of this study are to expose the reflection of
using fiber reinforcement in the active zone behind the sheet pile
wall. The study indicated the benefits for the overall behavior of
strip footing and sheet pile wall that can be achieved when using
fiber reinforcement in the active zone. However, it is important to
make clear that the results observed herein are relevant for the soil,
CKD, and fiber type used in the present research and that further
studies are necessary to generalize such findings. Furthermore, full-
609
scale and centrifugal tests must be conducted to increase under-
standing of the real behavior and accurate design in the application
of fiber reinforced sand.
6. Conclusions
The interaction behavior between strip footing resting on fiber-
reinforced cemented sand in the active zone and the cantilever
sheet pile wall was investigated through physical model tests and
two-dimensional non-linear FE analyses. The results obtained from
model test studies and numerical analyses were compared. Based
on the investigations the following main conclusions can be drawn:
1 The addition of cement kiln dust, CKD, improves greatly the
ultimate bearing capacity of the strip footing resting on fiber-
reinforced and unreinforced sand in the active zone. The rate
of ultimate bearing capacity gain, increased with the increase in
the fiber ratio, RF, of the reinforced cemented sand for the whole
range of cement kiln dust studied.
2 In physical model results and FE analyses, the addition of poly-
propylene fibers to the cemented sand increases the deform-
ability index, DI, which represents the ductility. At the lowest
level of the fiber ratio (RF  0.5%), DI was weakly influenced by
the cement kiln dust ratio.
3 The fiber insertion in the cemented sand behind the sheet pile
wall has a significant effect in decreasing the lateral deflection
along the sheet pile wall. At cement kiln dust ratio, RCKD ¼ 9%,
the maximum lateral deflection decreased by about 51% at fiber
ratio of 0.75%.
4 The ultimate bearing capacity of the strip footing on fiber-
reinforced cemented sand in the active zone increases with an
increase in the thickness of reinforced sand layer. However, for
higher fiber ratios (RF
0.75%), at a thickness of fiber-reinforced
cemented sand layer greater than U/H ¼ 0.4, the increase in
ultimate bearing capacity is not significant.
5 Increasing the distance between the strip footing and the sheet
pile wall, X/H, leads to a significant increase in the ultimate
bearing capacity of the footing and a decrease in the maximum
lateral deflection of the sheet pile wall. This clearly indicates the
adverse effect of the interaction between footing and sheet pile
wall. However, the experimental and numerical results at a
higher fiber ratio (RF
0.75%) showed that if X/H is more than
0.8 there is no appreciable increase in the ultimate bearing ca-
pacity of the footing (about 4.8%). The above result allows a
guideline to be formulated for designers to inform them when
they need to consider wallefoundation interaction.
6 There is close agreement between the experimental and nu-
merical results in term of the fiber ratio. However, the numerical
610 A.M. Nasr / Geotextiles and Geomembranes 42 (2014) 599e610

results show that the BCR has a significant decrease (about 14%) Falorca, I.M.C.F.G., Pinto, M.I.M., 2011. Effect of short randomly distributed poly-
propylene microfibers on shear strength behavior of soils. J. Geosynth. Int. 18
compared with the experimental model results.
(1), 2e11.
Georgiadis, M., Anagnostopoulos, C., 1998. Lateral pressure on sheet pile walls due
It should be noted that the results presented in this study are to strip load. J. Geotech. Geoenviron. Eng. 124 (1), 95e98.
related to model-scale footings on fiber-reinforced sand and are Ghanbari, A., Taheri, M., 2012. An analytical method for calculating active earth
pressure in reinforced retaining walls subject to a line surcharge. J. Geotext.
limited to these conditions. Caution should be used when applying Geomembr. 34, 1e10.
model-scale footing test results in the behavior of full-scale foun- Habib, P.A., 1974. Scale effect for shallow footings on dense sand. J. Geotech. Eng.
dations. Therefore, further tests using full-scale or a large models 100 (GT1), 95e99.
Ismail, M.A., 2005. Performance of cement-stabilized retaining walls. Can. Geotech.
are recommended. Furthermore, the effect of some other parame- J. 39, 876e891.
ters such as scale effect, backfill soil type, rigidity of the sheet pile Kamon, M., Nontananandh, S., 1991. Combining industrial wastes with lime for soil
wall, type of reinforcement, loading conditions, etc. have not been stabilization. J. Geotech. Eng. 117 (1), 1e17.
Ko, H., Davidson, W., 1973. Bearing capacity of footings in plane strain. J. Soil Mech.
investigated. Found. Div. ASCE 99 (1), 1e23.
Kusakabe, O., 1995. Foundations. In: Taylor, R.N. (Ed.), Geotechnical Centrifuge
References Technology. Blackie Academic and Professional, London.
Lovisa, J., Shukla, S.K., Sivakugan, N., 2010. Shear strength of randomly distributed
moist fibre-reinforced sand. J. Geosynth. Int. 17 (2), 100e106.
Ahmad, F., Bateni, F., Azmi, M., 2010. Performance evaluation of silty sand reinforced
Maher, M.H., Ho, Y.C., 1993. Behavior of fiber-reinforced cemented sand under static
with fibers. J. Geotext. Geomembr. 28 (1), 93e99.
and cyclic loads. Geotech. Test. J. 16 (3), 330e338.
Amir, H., Mahdi, H., 2013. Effect of fiber reinforcement on triaxial shear behavior of
Mckay, M., Emery, J., 1992. Stabilization/solidification of contaminated soils and
cement treated sand. J. Geotext. Geomembr. 36, 1e9.
sludges using cementitious systems. In: Proceedings of the Cement Industry
Amy, B., Alan, J., 2007. Scale effects of shallow foundation bearing capacity on
Solutions to Waste Management, 7e9 October. Calgary, Alberta, Canada,
granular material. J. Geotech. Geoenviron. Eng. 133 (10), 1192e1202.
pp. 135e151.
Asghari, E., Toll, D.G., Haeri, S.M., 2003. Triaxial behavior of a cemented gravely
Miller, G.A., Azad, M., 2000. Field and Laboratory Evaluation of Cement Kiln Dust as
sand, Tehran alluvium. Geotech. Geol. Eng. 21, 1e28.
a Soil Stabilizer, vol. 1714. Transportation Research Board, Washington, DC.
ASTM., 2010. Standard Practice for Classification of Soils for Engineering Purposes
Muir Wood, D., 2004. Geotechnical Modeling. Spon Press, London.
(Unified Soil Classification System). ASTM standard D2487. American Society for
Nasr, A.M., 2009. Use of discrete vertical reinforcement in active zone to improve
Testing and Materials, West Conshohocken, Pa.
the lateral response of the sheet pile wall. Alex. Eng. J. 48 (6), 743e759.
Balachandran, S.B., 1996. Modeling of Geosynthetic Reinforced Soil Walls. Ph.D.
Nasr, A.M., Nazir, A.K., 2013. Effect of geosynthetic reinforcement in active zone on
thesis. University of Cambridge, Cambridge, England.
the behavior of sheet pile walls. Geotech. Test. J. 36 (3), 331e344.
Bransby, P.L., Smith, I.A., 1975. Side friction in model retaining wall experiments.
Park, T., Tan, S.A., 2005. Enhanced performance of reinforced soil walls by the in-
J. Geotech. Eng. GT7, 615e632.
clusion of short fiber. J. Geotext. Geomembr. 23 (4), 348e361.
Bringkgreve, R., Vermeer, P., 1998. PLAXISdFinite Element Code for Soil and Rock
PCA, 2007. Report on Sustainable Manufacturing. Portland Cement Association,
Analysis, Version 7. Plaxis B. V., Netherlands.
Skokie, IL.
Chauhan, M.S., Mittal, S., Mohanty, B., 2008. Performance evaluation of silty sand
Ple, O., Le, T.N., 2012. Effect of polypropylene fiber-reinforcement on the mechanical
subgrade reinforced with fly ash and fiber. J. Geotext. Geomembr. 26 (5),
behavior of silty clay. J. Geotext. Geomembr. 32 (3), 111e116.
429e435.
Santos, E.C., Palmeira, E.M., Bathurst, R., 2013. Behavior of a geogrid reinforced wall
Conner, J.R., Cotton, S., Lear, P.R., 1992. Chemical stabilization of contaminated soils
built with recycled construction and demolition waste backfill on a collapsible
and sludges using cement and cement by products. In: Proceedings of the
foundation. J. Geotext. Geomembr. 39, 9e19.
Cement Industry Solutions to Waste Management, 7e9 October. Calgary,
Schnaid, F., Prietto, P.D.M., Consoli, N.C., 2001. Characterization of cemented sand in
Alberta, Canada, pp. 73e97.
triaxial compression. J. Geotech. Geoenviron. Eng. 127 (10), 857e867.
Consoli, N.C., Cruz, R.C., da Fonsea, A.V., Coop, M.R., 2012. Influence of cement-voids
Sulapha, P., Jan, O., Janet, L., 2008. Influence of chemical and physical characteristics
ratio on stress-dilatancy behavior of artificially cemented sand. J. Geosynth. Int.
of cement kiln dusts (CKDs) on their hydration behavior and potential suit-
138 (1), 100e109.
ability for soil stabilization. Cem. Concr. Res. J. 38, 803e815.
Consoli, N.C., Montardo, J.P., Donato, M., Prietto, D.M., 2004. Effect of material
Tang, C., Shi, B., Gao, W., Chen, W., Cai, Y., 2007. Strength and mechanical behavior of
properties on the behavior of sand-cement-fiber composites. Ground Improv. J.
short polypropylene fiber reinforced and cement stabilized clayey soil.
8 (2), 77e90.
J. Geotext. Geomembr. 25 (3), 194e202.
Consoli, N.C., Prietto, P.D.M., Ulbrich, L.A., 1998. Influence of fiber and cement
Vesic, A.S., 1973. Analysis of ultimate loads of shallow foundations. J. Soil Mech.
addition on behavior of sandy soil. J. Geotech. Geoenviron. Eng. 124 (12),
Found. Div. 94 (SM3), 661e688.
1211e1214.
Diambra, A., Ibraim, E., Muir Wood, D., Russell, A.R., 2010. Fiber reinforced sands:
experiments and modeling. J. Geotext. Geomembr. 28 (3), 238e250.