Journal of Rock Mechanics and Geotechnical Engineering 9 (2017) 159e169

Contents lists available at ScienceDirect

Journal of Rock Mechanics and

Geotechnical Engineering
journal homepage: www.rockgeotech.org

Full Length Article

Performance evaluation of cement-stabilized pond ash-rice husk ash-

clay mixture as a highway construction material
Deepak Gupta*, Arvind Kumar
Department of Civil Engineering, National Institute of Technology (NIT), Jalandhar, India

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

Article history: This paper reports the results of an investigation carried out on clay soil stabilized with pond ash (PA),
Received 25 January 2016 rice husk ash (RHA) and cement. Modified Proctor compaction tests were performed in order to
Received in revised form investigate the compaction behavior of clay, and California bearing ratio (CBR) tests were performed to
20 May 2016
determine the strength characteristics of clay. For evaluation purpose, the specimens containing different
Accepted 22 May 2016
Available online 8 December 2016
amounts of admixtures were prepared. Clay was replaced with PA and RHA at a dosage of 30%e45% and
5%e20%, respectively. The influence of stabilizer types and dosages on mechanical properties of clay was
evaluated. In order to study the surface morphology and crystallization characteristics of the soil sam-
Keywords:
California bearing ratio (CBR)
ples, scanning electron microscopy (SEM) and X-ray diffraction (XRD) analyses were carried out,
Pond ash (PA) respectively. The results obtained indicated a decrease in the maximum dry density (MDD) and a
Rice husk ash (RHA) simultaneous increase in the optimum moisture content (OMC) with the addition of PA and RHA.
Optimum moisture content (OMC) Multiple linear regression analysis (MLRA) showed that the predicted values of CBR tests are in good
Maximum dry density (MDD) agreement with the experimental values. Developed stabilized soil mixtures showed satisfactory
strength and can be used for construction of embankments and stabilization of sub-grade soil. The use of
locally available soils, PA, RHA, and cement in the production of stabilized soils for such applications can
provide sustainability for the local construction industry.
Ó 2017 Institute of Rock and Soil Mechanics, Chinese Academy of Sciences. Production and hosting by
Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/
licenses/by-nc-nd/4.0/).

1. Introduction A huge quantity of FA and PA as waste products is generated

With rapid industrialization growth, the industrial activities are
producing a large quantity of waste materials. Such wastes are
hazardous for health and environment. To minimize the impact of
such waste materials, proper management is required. Utilization
of the waste materials in different purposes is an effective way of
management. The waste materials such as fly ash (FA), pond ash
(PA), scrap tire, and rice husk ash (RHA) are generated in bulk and
their utilization for general purposes becomes difficult and chal-
lenging. Such wastes can be used in large quantities for civil engi-
neering construction. Utilization of such waste materials in the
form of admixtures for the improvement of soil characteristics is
one of the key ground improvement techniques. This methodology
can be used for the improvement of problematic soils like soft clay
or swelling soil, which are unable to sustain load from the structure.
* Corresponding author.
E-mail address: deepakg757@gmail.com (D. Gupta).
Peer review under responsibility of Institute of Rock and Soil Mechanics,
Chinese Academy of Sciences.
from power plants at global level. In India, about 200 million m2 of
land area is covered up with million tonnes of PA deposits. The
production of leachate compounds from the ash ponds leads to the
contamination of both the groundwater and surface water bodies
as well as soil, because of the presence of toxic elements and heavy
metallic substances within them.
The engineering properties of PA can be improved using various
techniques among which the incorporation of lime/cement in the
PA by mechanical mixing has been found to be the most reliable
approach. Kumar et al. (1999) presented the results of laboratory
investigation carried on silty sand and PA reinforced with randomly
distributed polyester fibers. They showed that incorporation of fi-
bers results in the improvement of various characteristics of soil
(e.g. peak compressive strength, California bearing ratio (CBR),
ductility, and peak friction angle). Sarkar et al. (2012) reported that
the CBR of PA improves with addition of cement. Bera et al. (2007)
reported that the increase in compaction leads to an increase in
maximum dry density (MDD) and a decrease in optimum moisture
content (OMC). Kumar and Gupta (2016) concluded that the
cement-stabilized and fiber-reinforced clay mixed with optimum

http://dx.doi.org/10.1016/j.jrmge.2016.05.010
1674-7755 Ó 2017 Institute of Rock and Soil Mechanics, Chinese Academy of Sciences. Production and hosting by Elsevier B.V. This is an open access article under the CC BY-
NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
160 D. Gupta, A. Kumar / Journal of Rock Mechanics and Geotechnical Engineering 9 (2017) 159e169
percentage of RHA and PA can be successfully used as fill material
for geotechnical applications. Ghosh (2010) studied the suitability
of stabilized PA for road base and sub-base construction purposes
by performing a series of laboratory tests. The test results revealed
that the utilization of Class F PA either individually or in combi-
nation with different dosages of lime (i.e. 4%, 6% and 10%) and
phosphogypsum (0.5% and 1%) is suitable for construction of sub-
base and road base applications. Roy and Chattopadhyay (2008)
studied the effectiveness of RHA and PA in the improvement of
subgrade quality for the purpose of road construction. It was shown
that with the addition of PA or RHA, the compaction characteristics
of alluvial soil are influenced significantly. Chand and Subbarao
(2007) studied the suitability of lime-stabilized PA for base and
sub-base courses of pavement.
Rice husks are the shells produced during de-husking opera-
tion of paddy, which varies from 20% (Mehta, 1986) to 23% (Della
et al., 2002) by the weight of paddy. The husk is a waste material
and is disposed off either by dumping or burning in the boiler for
processing paddy. The burning process of rice husk produces ash
which is about 20% of its weight (Mehta, 1986). This ash known as
RHA contains silica as a major constituent whose quality (percent
of amorphous and unburnt carbon) is influenced by the type of
burning process (Nair et al., 2006). The RHA is pozzolanic in na-
ture because of its high amorphous silica content (Mehta, 1986). It
has been reported that in India, the annual paddy production is
about 100 million tonnes which generates more than 4 million
tonnes of RHA (Ramakrishna and Kumar, 2008). However, RHA
cannot be used individually for the soil stabilization due to lack of
cementitious properties in it (Ali et al., 1992). Consequently, it is
used along with a binder (e.g. cement, lime, lime sludge, and
calcium chloride) for the stabilization of soil (Ali et al., 1992;
Muntohar and Hantoro, 2000; Basha et al., 2005; Sharma et al.,
2008; Brooks, 2009).
The utilization of RHA and PA for soil stabilization can help in
achieving a low-cost construction and also provide an environ-
mental friendly ways of their disposal. Their use would also reduce
the consumption of cement, thereby saving the energy and
reducing the emissions of greenhouse gases.
It has been shown that PA and RHA have sufficient capability in
the improvement of strength behavior of soil. In each case, there is
an optimum dosage of PA or RHA to be used beyond which the
improvement is insignificant. Therefore, it is expected that for
further improvement in strength, combination of different ad-
mixtures can be significant. In the current investigation, the soil
behavior has been studied in detail using different combinations of
PA, RHA and cement in it. This study aims to understand the
compaction behavior of soil which is very important for structures
(such as pavement and embankment). CBR test is considered for
strength behavior.
The geotechnical characteristics of clay mixed with PA, RHA and
cement were investigated. Cement was added to soil mix at dosage
of 0%e4%, whereas PA and RHA were added to the clay at 30%e45%
and 5%e20% by dry weight of sample, respectively. Test specimens
were subjected to compaction tests and CBR tests. Scanning elec-
tron microscopy (SEM) and X-ray diffraction (XRD) tests were also
performed. Specimens were soaked for 3 d after which they were
tested for CBR tests.
2. Experimental methods
2.1. Experimental materials
The different materials (i.e. soil, PA, RHA and cement) used in
the study have same properties as reported in previous research
(Kumar and Gupta, 2016).
2.1.1. Soil
The soil used in this investigation is kaolin clay. According to
unified soil classification system (USCS), the soil is classified as
clay with low plasticity. The index properties of soil are listed in
Table 1.
2.1.2. Pond ash
Light grey colored Class F PA was used in the study. The PA is of
nonplastic type possessing a specific gravity of 2.1. The ash may be
categorized as Class F type as per ASTM C618-12 (2012). The
compositions and properties of PA are listed in Table 2.
2.1.3. Rice husk ash
RHA was collected from local rice mill. The RHA is of non-plastic
type possessing a specific gravity of 1.95. Its properties also varied
depending on its burning temperature. The physicochemical
properties of RHA are listed in Table 3.
2.1.4. Ordinary Portland cement
Ordinary Portland cement (OPC) of Grade-43, having initial
and final setting times of 30 min and 600 min, respectively, was
used in the study. Physical properties of cement are listed in
Table 4.
2.2. Experimental programs
A comprehensive series of laboratory tests was performed on
the clayey soil specimens stabilized with different dosages and
combinations of the stabilizers, i.e. RHA, PA and cement. The tests
performed include Proctor compaction test and CBR test. Table 5
lists a summary of mixed various stabilizer combinations with
soil. The percentages used of RHA were 0%, 5%, 10%, 15% and 20%
and of PA were 30%, 35%, 40% and 45% of total mass of the mixture
with different percentages of cement as 0%, 2% and 4%. Such con-
tents are optimum contents for RHA and PA which are taken from
the literature. From the literature, it is found that the optimum
content in the case of RHA is 10%e15% (Basha et al., 2005) and in the
case of PA, it is 30%e40% (Bera et al., 2007). An amount of soil was
mixed with PA, RHA and cement to yield stabilized soil specimens.
All the specimens were prepared to the MDD and OMC, and tested
after 3 d of soaking.
2.3. Testing procedures
2.3.1. California bearing ratio test
CBR test on stabilized and unstabilized soil specimens was
conducted in accordance with ASTM D1883-99 (2000). The pro-
cedure adopted for the tests was the same as that used in previous
investigation (Kumar and Gupta, 2016).
2.3.2. Analysis of surface morphology and elemental composition
SEM was used to examine the surface morphology of the sam-
ples and energy dispersive spectroscopy (EDS) was used to detect
the presence and composition of different elements in them. A
small portion of samples was kept in an oven at 105
C for 24 h for
drying. The specimen was mounted on a specimen’s holder. A thin
conducting layer of silver about 200 Å in thickness was coated on
Table 1
Properties of soil.
Specific Liquid Plastic Plasticity OMC (%) MDD (kN/m3)
gravity limit (%) limit (%) index
2.7 43 19 24 18.3 18.6
 D. Gupta, A. Kumar / Journal of Rock Mechanics and Geotechnical Engineering 9 (2017) 159e169 161

Table 2
Properties of pond ash.

Composition (%) Color Shape Specific Plasticity OMC (%) MDD (kN/m3)
gravity, G index
Silt þ clay Fine sand Medium sand Coarse sand
(<75 mm) (75e425 mm) (425 mme2 mm) (2e4.75 mm)

41.6 44.6 12.2 1.6 Light gray Rounded/sub-rounded 2.1 Nonplastic 26 12.35

Table 3
Physicochemical properties of rice husk ash.

Chemical composition (%) Physical properties

Calcium oxide Silica (SiO2) Alumina Iron oxide Magnesia (MgO) Sodium oxide Potassium oxide Specific gravity OMC (%) MDD (kN/m3)
(CaO) (Al2O3) (Fe2O3) (Na2O) (K2O)

2.4 91.3 1.4 0.6 2.1 0.3 1.9 1.98 60 8.79

Table 4
Physical properties of cement.

Fineness Specific gravity, G Standard consistency (%) Initial setting time (min) Final setting time (min) Soundness (cement expansion) (mm)

3 3.15 38 30 600 3

Table 5
Stabilizer combination scheme for soil.
dust over them. RHA particles were rather angular and irregular in
shape (Fig. 1b) and had rough, gritty surface textures. It shows very
Stabilizer Combination Designation fine particle size to the order of a millimeter or less and that there
Single 0%, 5%, 10%, 15% and 0RHA, 5RHA, 10RHA, 15RHA, 20RHA are pores of varying sizes within the particles.
20% RHA Fig. 2a and b shows the surface morphology of soil-PA speci-
0%, 10%, 20%, 30%, 40% and 0PA, 10PA, 20PA, 30PA, 40PA, 50PA
mens after 7 d of curing. As indicated, the presence of plate-
50% PA
Mixed 0% RHA þ 0% PA 0RHA0PA shaped kaolin clay particles and spherical PA particles is
5% RHA þ 45% PA 5RHA45PA evident. Fig. 2c and d shows the EDS spectrum and the elemental
10% RHA þ 30% PA 10RHA30PA composition of soil-PA mixture, respectively. Fig. 3a and b shows
15% RHA þ 35% PA 15RHA35PA the SEM images of soil-RHA mixture after 7 d of curing, indicating
20% RHA þ 30% PA 20RHA30PA
the similar palatines (plate shaped) of kaolin clay particles as
observed in Fig. 3, with additional particles of RHA having irreg-
ular surface morphology. The EDS pattern of soil-RHA mixture is
the specimen surface with the aid of sputter coater Poloran E 5100. shown in Fig. 3c and its elemental composition is represented in
JEOL’s JSM 6510 series scanning electron microscope was used to Fig. 3d. Fig. 4a and b shows the SEM observations of soil-PA-RHA
examine the morphology of the mixes. mixture which reveals three different types of particles corre-
sponding to platy kaolin clay, spherical PA and irregular RHA.
2.3.3. XRD analysis for hydration behavior Fig. 4c and d respectively presents the EDS spectrum and
XRD analysis was carried out to determine the different phases elemental composition of soil-PA-RHA mixture. The effect of
presented in samples. A PANalytical X’PERT PRO diffract meter addition of cement as the stabilizing agent is shown in Fig. 5.
was used. The samples were scanned with the measuring angle of Fig. 5a and b shows the formation of CeSeH (calciumesilicatee
2q from 10
to 70
. The database of 2000 JCPDS-International hydrate) gel in a significant amount. The presence of some elon-
Center for Diffraction Data was used to identify the mineralog- gated crystalline structures can also be observed. The formation of
ical phases. CeSeH gel has been ascribed to the pozzolanic activity taking
place between cement as stabilizer and clay. Fig. 5c and d repre-
3. Results and discussion sents the EDS pattern and elemental composition of soil-PA-RHA-
cement mixture, respectively.
The influence of RHA, PA and cement on the geotechnical
characteristics of clayey soil was investigated by conducting 3.2. Results of XRD analysis
modified Proctor compaction tests and CBR tests. In addition, SEM
and XRD tests were carried out on certain samples in order to study The comparison of XRD pattern of soil-PA mixture and soil-PA-
the surface morphological characteristics and hydraulic com- cement mixture after 7 d of curing is shown in Fig. 6. The results
pounds formed. The detailed discussion and analysis of the results obtained have shown that stabilization of kaolin clay with cement
obtained are given in the following sections. leads to a noticeable reduction in the compositions of kaolinite
(represented as K) and quartz (represented as Q). This has been
3.1. Surface morphological characteristics evidenced by the disappearance of peaks for the particular crys-
talline phase in the diffractogram. The peak disappearance has been
Fig. 1a and b shows the SEM observations of PA and RHA, accompanied by a simultaneous appearance of new peaks in the
respectively. In terms of particle shape and surface features, PA was XRD patterns. These new peaks have appeared due to the formation
quite different from RHA. PA particles were spherical in shape as of crystalline phases (such as calcite, i.e. CaCO3) after the pozzolanic
evidenced by Fig. 1a. The particles had shiny surfaces without any reaction of cement with kaolin clay.
162 D. Gupta, A. Kumar / Journal of Rock Mechanics and Geotechnical Engineering 9 (2017) 159e169

Fig. 1. SEM images of (a) pond ash and (b) rice husk ash.

Fig. 2. SEM images of soil-PA mixture at magnifications of (a) 3000 and (b) 4500 times. (c) EDS spectrum of soil-PA mixture and (d) its elemental composition after 7 d of curing.

Fig. 7 shows the comparative XRD spectra of soil-RHA mixture
and soil-RHA-cement mixture, after 7 d of curing. The stabilizing
activities of both RHA and cement are indicated by the disappear-
ance of kaolinite and quartz in a similar manner as observed in the
case of PA. However, the pozzolanic action is rapid in the case of
RHA and cement as stabilizers as observed from high intensities of
pozzolanic product CaCO3 and reduced intensities of quartz peaks.
The XRD patterns obtained for the untreated and treated clay
samples are shown in Fig. 8. From the results obtained, it has
been noticed that the presence of kaolinite as the important
crystalline phase is evident from the diffraction peaks at posi-
tions 2q ¼ 24.8
, 26.6
and 60.5
(d ¼ 3.58 Å, 3.34 Å and 1.52 Å,
respectively). Further, the presence of quartz has been confirmed
from the peaks obtained at 2q ¼ 20.8
and 50.1
(d ¼ 4.25 Å and
1.8 Å, respectively) (Basha et al., 2005). However, when the clay
is subjected to stabilization process, the XRD pattern shows
noticeable changes in terms of peak intensities corresponding to
different crystalline phases. It has been observed that the
diffraction peaks corresponding to kaolinite and quartz have
shown a reduction in the intensities. This variation in peak in-
tensities has been possibly ascribed to the occurrence of pozzo-
lanic reactions which take place between the crystalline phases
of clay and the reactive pozzolanic material. The peak intensities
have shown variation with the variation in the content of stabi-
lizing material used in clay. This behavior has been clearly
depicted from the reduction in peak intensities of quartz with the
increasing content of stabilizing material. Apart from the
reduction or disappearance of quartz peaks, a simultaneous
appearance of calcite has been observed in the XRD patterns of
treated samples at peak positions of 2q ¼ 27.1
and 39.4

(d ¼ 3.3 Å and 2.28 Å, respectively). The formation of calcite is an
evidence of the fact that pozzolanic or cementing material has
 D. Gupta, A. Kumar / Journal of Rock Mechanics and Geotechnical Engineering 9 (2017) 159e169 163

Fig. 3. SEM images of soil-RHA mixture after 7 d of curing at magnifications of (a) 3000 and (b) 2000 times. (c) EDS spectrum of soil-RHA mixture and (d) its elemental composition
after 7 d of curing.

Fig. 4. SEM images of soil-PA-RHA mixture after 7 d of curing at magnifications of (a) 3000 and (b) 2000 times. (c) EDS spectrum of soil-PA-RHA mixture and (d) its elemental
composition after 7 d of curing.
164 D. Gupta, A. Kumar / Journal of Rock Mechanics and Geotechnical Engineering 9 (2017) 159e169

Fig. 5. SEM images of soil-PA-RHA-cement mixture after 7 d of curing at magnifications of (a) 3000 and (b) 2000 times. (c) EDS spectrum of soil-PA-RHA-cement mixture and (d) its
elemental composition after 7 d of curing.

taken part in chemical reactions with the clay minerals. This is
strongly noticeable in the case of treated clay containing
maximum amount of stabilizing material, which has shown a
significant disappearance of quartz and kaolinite, along with the
appearance of calcite peaks.
3.3. Compaction test
The compaction tests were performed as per ASTM D1557-09
(2009) specifications outlined for modified Proctor compaction
tests. Modified Proctor compaction tests were carried out on the
soil-RHA-cement mixture, PA-cement mixture and soil-PA-RHA-
Soil + RHA + Cement
C
Q C
Intensity (count)
cement mixtures proportions. The compaction tests were per-
formed for various combinations of soil-RHA-PA-cement mix-
tures as mentioned in Table 5. Figs. 9 and 10 present the variation
of MDD and OMC with different contents of RHA and PA. The
MDD and OMC have been found to increase with an increase in
RHA dosage from 0% to 20% and PA dosage from 0% to 50%.
Similar trends for lime, RHA and FA stabilized clay mixtures have
been reported (Rahman, 1987; Ali, 1992; Kaniraj and Havanagi,
1999; Al-Rawas et al., 2005; Kolias et al., 2005; Mir and Srid-
haran, 2013).
It can be observed from Fig. 11 that OMC increases with the
increase in combined dosage of RHA and PA in mix mode stabili-
zation. The increase in OMC has been attributed to the increased
rate of pozzolanic reaction between RHA/PA and soil components
(Hossain, 2011). A higher increase in OMC in the case of cement-
stabilized soil specimens has been ascribed to the extra water
needed for higher fineness and subsequent enhanced hydration.
Again from Fig. 11, with the combined dosage of PA and RHA
in mix mode stabilization, there is a further reduction in the
Q MDD and enhancement in OMC. Relatively lower specific gravity

K K K Q
K values of RHA and PA might be the causes for the reduction in
MDD values (Ali et al., 1992; Jha and Gill, 2006; Alhassan, 2008).
10 20 30 40 50 60 70 It is also observed that with increase in cement dosage in the soil
samples, there is further decrease in MMD values. The fall in
Soil + RHA density is due to the quick reaction of cement with the soil which
Q Q alters the base exchange aggregation and flocculation, thereby
C Q resulting in increased void ratio of the mixture. This subse-
K K C
K K quently leads to a decrease in the density of the overall mixture.
Since PA and RHA are replacing the clay particles because of
which overall weight of mixture decreases and mixture becomes
10 20 30 40 50 60 70
a light weight mixture. Such types of light weight mixtures
2 Theta (degree)
are preferable in the construction of retaining wall types of
Fig. 6. XRD patterns of clay treated with pond ash (Gupta and Kumar, 2016). structures.
 D. Gupta, A. Kumar / Journal of Rock Mechanics and Geotechnical Engineering 9 (2017) 159e169 165

cement-stabilized clay. In both figures, CBR value increases with

increase in RHA content up to 10% and this after the increase in
Soil + RHA + Cement RHA from 10% to 15% does not improve the load carrying capacity
significantly. It shows that RHA content around 10% is optimum
C for the improvement of RHA-clay mixture. After 3 d of curing, the
Q C Q
Intensity (count)

K K CBR value of unstabilized clay is found to be 2.1%. With the

K K Q
addition of 10% RHA and 4% cement to clay, the soaked CBR value
increases from 3.2% to 16.7% with the addition of 10% RHA and 4%
10 20 30 40 50 60 70 cement to clay.
Soil + RHA Figs. 14 and 15 represent the unsoaked and soaked CBR values
of unstabilized and PA/cement-stabilized clay. With the addition
Q Q of PA to clay, the CBR value increases. Clay has mainly cohesion
C Q
K K C component against the deformation, but the addition of PA fric-
K K
tional component also contributes to the resistance against
deformation. Due to this, load-carrying capacity increases with
10 20 30 40 50 60 70 increase in PA content. Figs. 14 and 15 also show that, increase in
2 Theta (degree) PA content from 40% to 50% does not improve the load-carrying
capacity significantly. It shows that PA content around 40% is the
Fig. 7. XRD patterns of clay treated with rice husk ash.
optimum for the improvement in strength of PAeclay mixture.
The CBR values of clay with 4% cement in the case of soaked and
3.4. California bearing ratio tests unsoaked conditions are found to be 4.5% and 3.1%, respectively.
With further addition of PA to this mixture, the CBR values in the
The CBR value of subgrade is an important factor in designing case of soaked and unsoaked conditions are found to be 18.3% and
composition of pavement thickness. Its value is commonly used to 13.7%, respectively.
evaluate the quality of pavement materials. Figs. 12 and 13 show For the soaked condition, as can be observed from Figs. 13 and
the unsoaked and soaked CBR values of unstabilized and RHA/ 15, upon addition of RHA/PA and cement to clay, the CBR value of

Fig. 8. XRD patterns of untreated clay (A), treated clay with PA, RHA and 0% cement (B), treated clay with PA, RHA and 2% cement (C), and treated clay with PA, RHA and 4% cement
(D) (Kumar and Gupta, 2016).
166 D. Gupta, A. Kumar / Journal of Rock Mechanics and Geotechnical Engineering 9 (2017) 159e169

20 24

19 22
MDD (kN/m3) or OMC (%)

MDD (kN/m3) or OMC (%)

18 20

17 18

16 16

15 14

14 12
0 5 10 15 20

Content of rice husk ash (%)

Cement=0%: MDD Cement=2%: MDD

Cement=0%: MDD Cement=2%: MDD
Cement=4%: MDD Cement=0%: OMC Cement=4%: MDD Cement=0%: OMC
Cement=2%: OMC Cement=4%: OMC Cement=2%: OMC Cement=4%: OMC

Fig. 9. Maximum dry density and optimum moisture content versus rice husk con- Fig. 11. Maximum dry density and optimum moisture content versus different mixture
tents for different percentages of cement. combinations for different percentages of cement.

21
16
cement=0%
20 14
cement=2%
MDD (kN/m3) or OMC (%)

12
19 cement=4%
10
CBR (%)

18 8

6
17
4
16
2

15 0
0 5 10 15 20
0 10 20 30 40 50
Content of rice husk ash (%)
Content of pond ash (%)
MDD 0% MDD 2% MDD 4% Fig. 12. Variations of unsoaked CBR with content of rice husk ash.

OMC 0% OMC 2% OMC 4%

Fig. 10. Maximum dry density and optimum moisture content versus pond ash con- 18
tents for different percentages of cement. cement=0%
16
cement=2%
14
the mixture increases. The soaked CBR of clay (alone) is less than cement=4%
12
that of unsoaked condition, since under soaked conditions, the
CBR (%)

surface tension forces (which were offering additional resistance to 10

penetration under the unsoaked condition) are destroyed (Pandian 8

and Krishna, 2003). The resulting improvement in soaked CBR
6
value with the addition of RHA/cement and PA/cement can be
attributed to the pozzolanic reaction between the clay and pozzo- 4
lanic material forming additional cementitous material that bounds 2
particles together and enhances the strength of the soil. The PA/
0
cement-stabilized clay has much higher CBR than RHA/cement- 0 5 10 15 20
stabilized clay.
Content of rice husk ash (%)
Figs. 16 and 17 show the unsoaked and soaked CBR values of
unstabilized and stabilized clay with the combination of PA, RHA Fig. 13. Variations of soaked CBR with content of rice husk ash.
 D. Gupta, A. Kumar / Journal of Rock Mechanics and Geotechnical Engineering 9 (2017) 159e169 167

16 30

cement=0% Cement=0%
14
25
Cement=2%
cement=2%
12
Cement =4%
cement=4% 20
10

CBR(%)
CBR (%)

15
8

6 10

4
5
2
0
0
0 10 20 30 40 50
Content of pond ash (%)
Fig. 17. Variations of soaked CBR with combination of rice husk ash and pond ash.
Fig. 14. Variations of unsoaked CBR with content of pond ash.

and cement. It can be observed that the combined effect of PA, RHA
20
and cement on load-carrying capacity is much greater than the PA/
Cement =0% cement and RHA/cement, when added to clay. Clay, PA and RHA
18
used in this study have sizes comparable to sizes of silt, sand and
16 Cement =2% fine particles, respectively. Mixture of these three materials has
better gradation and thus contributes to greater frictional resis-
14 Cement =4%
tance. Due to this reason, the combination has greater potential to
12 improve the strength of clay. Effects of different combinations of PA
CBR (%)

and RHA on CBR value of mixture are presented in Figs. 16 and 17. It
10
is observed from Figs. 16 and 17 that the CBR value increases with
8 the increase in contents of RHA and PA up to 10% and 40%,
respectively, and thereafter the CBR value decreases. It shows that
6 RHA content around 10% and PA content around 40% are the opti-
mum for the improvement of strength of the mixture.
4

2 4. Statistical analysis

0 The purpose of multiple linear regression analysis (MLRA) is

0 10 20 30 40 50
to discover two or more independent variables which explain the
Content of pond ash (%)
variations of a dependent variable. To perform the analysis, sta-
Fig. 15. Variations of soaked CBR with content of pond ash. tistical software SPSS 12.0 was used since it offers a choice of
regression methods. In this method, the independent variables
affecting the dependent variable are first determined. The
normal regression provides the alternative method for deter-
25
mining, in which independent variables are significant by iden-
Cement= 0% tifying good subset models that result in considerably less time
20 Cement =2% for computing.
A MLRA model can be formulated in the following form:
Cement= 4%

15
Y ¼ b0 þ b1 X1 þ . þ bn Xn þ 3 (1)
CBR (%)

where Y is the dependent variable, Xi (i ¼ 1, 2, ., n) is the inde-

10 pendent variable, bi is the predicted parameter, and 3 is the error.
In MLRA, the soaked and unsoaked CBR values have been
considered as the independent variable while remaining soil
5 properties have been taken as the dependent variables. MLRA can
be carried out using standard statistical software like data analysis
tool bar of Microsoft Excel in order to derive the relationship sta-
0 tistically. The soaked and unsoaked CBR values can be expressed as

CBRunsoaked ¼ 2:25 þ 0:162CPA þ 0:289CRHA

 
Fig. 16. Variations of unsoaked CBR with combination of rice husk ash and pond ash.
þ 0:625Ccement R2 ¼ 0:88 (2)
168 D. Gupta, A. Kumar / Journal of Rock Mechanics and Geotechnical Engineering 9 (2017) 159e169

25 combinations, and curing period. RHA-stabilized soils

showed higher strength as compared to their PA counter-
parts. Stabilization by RHA can be attributed to the fact
20
that RHA satisfies the requirement for self-cementing
Predicted CBR (%)

characteristics that allows RHA to react with soil in a

15 manner similar to Portland cement. Addition of PA pro-
duces better mechanical properties.
(4) CBR value of the single mode stabilizer increases with in-
10
crease in RHA/PA content up to a certain limit of RHA ¼ 10%
and PA ¼ 40% known as optimum contents, after which
5 further improvement in the CBR is not significant.
(5) CBR value of the combination of different stabilizers is
greater as compared to the single stabilizer. It is due to the
0
improvement in the gradation of the mixture with the
0 5 10 15 20 25
addition of RHA and PA.
Observed CBR (%)
(6) MLRA was performed to find the predicted CBR, which is in
(a) close agreement with the experimental values.
30 (7) Test results showed that the best combination for maximum
gain in CBR value is 4% cement, 10% RHA and 40% PA mixed in
25 clay.

20 Conflict of interest
Predicted CBR (%)

15 We wish to confirm that there are no known conflicts of interest

10
5
0
0 5 10 15 20
-5 Observed CBR (%)
(b)
Fig. 18. Cross-correlations of predicted and observed CBR values for multiple regres-
sion models of (a) unsoaked and (b) soaked conditions.
CBRsoaked ¼ 0:825 þ 0:054CPA þ 0:121CRHA
 
þ 1:674Ccement R2 ¼ 0:84
whereCPA , CRHA and Ccement are the contents of PA, RHA and
cement, respectively. The coefficient of determination, R2, is found
to be 0.88 and 0.84 for Eqs. (2) and (3), respectively. Fig. 18a and b
shows the linear scatter diagrams using Eqs. (2) and (3), respec-
tively. It may be observed that regression plot of predicted CBR
value against the experimental CBR value of testing data points lies
well within the 99% confidence interval.
5. Conclusions
In this study, different tests like SEM, XRD tests, modified
Proctor compaction tests and CBR tests were done to evaluate the
behavior of soil mixed with RHA, PA and cement. The major con-
clusions drawn are presented below:
(1) An increase in RHA or PA dosage causes a reduction in MDD
and enhancement in OMC.
(2) It is found that RHA and PA both have pozzolanic behavior as
confirmed from XRD results. Therefore, both can be used as
partial replacement materials for the cement.
(3) The potential benefit of stabilization is found to be
dependent on the type/amount of stabilizers, stabilizer
associated with this publication and there has been no significant
financial support for this work that could have influenced its
outcome.
References
25 30 Alhassan M. Potentials of rice husk ash for soil stabilization. Assumption University
Journal of Technology 2008;11(4):246e50.
Ali FH. Stabilization of a residual soil. Soils and Foundations 1992;32(4):178e85.
Ali FH, Adnan A, Choy CK. Geotechnical properties of a chemically stabilized soil
from Malaysia with rice husk ash as an additive. Geotechnical and Geological
Engineering 1992;10(2):117e34.
Al-Rawas AA, Hugo AW, Al-Sami H. Effect of lime, cement and artificial pozzolan on
the swelling potential of an expansive soil from Oman. Building and Environ-
ment 2005;40:267e81.
ASTM D1557-09. Test method for laboratory compaction characteristics of soil using
modified effort. West Conshohocken, PA, USA: ASTM International; 2009.
ASTM D1883-99. Standard test method for California bearing ratio of laboratory-
compacted soils. West Conshohocken, PA, USA: ASTM International; 2000.
(3) ASTM C618-12. Standard specification for coal fly ash and raw or calcined natural
pozzolana for use in concrete. West Conshohocken, PA, USA: ASTM Interna-
tional; 2012.
Basha EA, Hashim R, Mahmud HB, Muntohar AS. Stabilization of residual soil with
rice husk ash and cement. Construction and Building Materials 2005;19(6):
448e53.
Bera AK, Ghosh A, Ghosh A. Compaction characteristics of pond ash. Journal of
Materials in Civil Engineering 2007;19(4):349e57.
Brooks RM. Soil stabilization with fly ash and rice husk ash. International Journal of
Research and Reviews in Applied Sciences 2009;3(1):209e17.
Chand SK, Subbarao C. Strength and slake durability of lime stabilized pond ash.
Journal of Materials in Civil Engineering 2007;19(7):601e8.
Della VP, Kuhn I, Hotza D. Rice husk ash as an alternate source for active silica
production. Materials Letters 2002;57(4):818e21.
Ghosh A. Compaction characteristics and bearing ratio of pond ash stabilized with
lime and phosphogypsum. Journal of Materials in Civil Engineering 2010;22(4):
343e51.
Gupta D, Kumar A. Strength characterization of cement stabilized and fiber rein-
forced clay-pond ash mixes. International Journal of Geosynthetics and Ground
Engineering 2016;2(4):11e32.
Hossain KMA. Stabilized soils incorporating combinations of rice husk ash and
cement kiln dust. Journal of Materials in Civil Engineering 2011;23(9):1320e7.
Jha JN, Gill KS. Effect of rice husk ash on lime stabilization. Journal of the Institution
of Engineers (India) 2006;87:33e9.
Kaniraj SR, Havanagi VG. Geotechnical characteristics of fly ash-soil mixtures.
Geotechnical and Geological Engineering 1999;30(2):129e47.
Kolias S, Kasselouri RV, Karahalios A. Stabilisation of clayey soils with high calcium
fly ash and cement. Cement and Concrete Composites 2005;27(2):301e13.
Kumar A, Gupta D. Behavior of cement-stabilized fiber-reinforced pond ash, rice
husk ash-soil mixtures. Geotextiles and Geomembranes 2016;44(3):466e74.
Kumar R, Kanaujia VK, Chandra D. Engineering behavior of fibre-reinforced pond
ash and silty sand. Geosynthetics International Journal 1999;6(6):509e18.
 D. Gupta, A. Kumar / Journal of Rock Mechanics and Geotechnical Engineering 9 (2017) 159e169 169

Mehta PK. Concrete structures properties and materials. Englewood Cliffs, N.J.: Arvind Kumar obtained his Ph.D. from University of
Prentice Hall; 1986. Roorkee (now IIT Roorkee), India. He joined NIT Jalandhar
Mir BA, Sridharan A. Physical and compaction behavior of clay soilefly ash mix- in 1991 and became professor in 2007. He is active
tures. Geotechnical and Geological Engineering 2013;31(4):1059e72. consultant. He has published over 88 research papers in
Muntohar AS, Hantoro G. Influence of rice husk ash and lime on engineering national and international journals and conferences. He is
properties of a clayey subgrade. Electronic Journal of Geotechnical Engineering member of Indian Geotechnical Society, International So-
2000;5:1e13. ciety of Soil Mechanics and Geotechnical Engineering, In-
Nair DG, Jagadish KS, Fraaim A. Reactive pozzolanas from rice husk ash: an alter- dian Roads Congress, Institution of Engineers, and Indian
native to cement for rural housing. Cement and Concrete Research 2006;36(6): Society of Technical Education.
1062e71.
Pandian NS, Krishna KC. The pozzolanic effect of fly ash on the California bearing
ratio of black cotton soil. Journal of Testing and Evaluation 2003;31(6):1e7.
Rahman MA. Effect of cement-rice husk ash mixtures on geotechnical properties of
lateritic soils. Soils and Foundations 1987;27(2):61e5.
Ramakrishna AN, Kumar P. Stabilization of black cotton soil using rice husk ash and
cement. In: National Conference on Civil Engineering: Meeting the Challenges Deepak Gupta obtained a B.Tech degree from NIT Srinagar,
of Tomorrow. Ludhiana: GND Engineering College; 2008. p. 215e20. India. He obtained his M.Tech degree from Dr. B.R.
Roy TK, Chattopadhyay BC. A study on the effect of cement on alluvial soil Ambedkar National Institute of Technology, Jalandhar, In-
strengthened with pond and rice husk ash for construction of road subgrade. dia and now submitted his Ph.D. thesis at the same insti-
In: Proceedings of the 12th International Conference of International Associa- tute. He is member of Institution of Civil Engineers
tion for Computer Methods and Advances in Geomechanics (IACMAG), Goa, Ludhiana.
India; 2008.
Sarkar R, Abbas SM, Shahu JT. A comparative study of geotechnical behavior of lime
stabilized pond ashes from Delhi region. International Journal of GEOMATE
2012;1(3):273e9.
Sharma RS, Phanikumar BR, Varaprasada Rao B. Engineering behavior of remolded
expansive clay blended with lime, calcium chloride and rice husk ash. Journal of
Materials in Civil Engineering 2008;20(8):509e15.