Construction and Building Materials 71 (2014) 354362

Contents lists available at ScienceDirect

Construction and Building Materials

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

Analysis of strength development in soft clay stabilized

with cement-based stabilizer
Ma Cong, Chen Longzhu, Chen Bing
Department of Civil Engineering, Shanghai Jiaotong University, Shanghai 200240, PR China

h i g h l i g h t s
 The mechanical properties of soft clay stabilized with cement-based stabilizer were investigated.
 The predicted models for compressive strength and secant modulus were analyzed.
 The effect of the supplementary cementing materials was conrmed by SEM imaging.

a r t i c l e

i n f o

Article history:
Received 10 June 2014
Received in revised form 18 August 2014
Accepted 27 August 2014
Available online 18 September 2014
Keywords:
Clay stabilizer
Development and analysis
Unconned compressive strength
Secant modulus
Pozzolanic effect

a b s t r a c t
Sodium silicate and promoters have been widely used as supplementary cementing materials substituting ordinary Portland cement to improve the properties of stabilized soils. In this paper, the developments of mechanical properties of specimens stabilized with cement-based stabilizer, which consists
of cement, sodium silicate and composite promoter, is investigated by unconned compressive strength
(UCS) test and scanning electron microscope (SEM). The failure strain and secant modulus are also
obtained in UCS tests. The test results indicate that the supplementary cementing materials perform
effectively pozzolanic reactions and improve the mechanical properties of cement stabilized clay. The
contribution of pozzolanic effect is regarded akin as an addition of cement and the total cement content
is the summation of cement content and equivalent cement content of the supplementary cementing
materials. According to this premise, the claywater/cement ratio hypothesis for stabilized clay is proposed for analyzing and assessing the development of the mechanical properties. The phenomenological
models for predicting the compressive strength and secant modulus of specimens stabilized with the
selected cement-based stabilizer are developed and veried. The comparison of predicted results and laboratory results indicates that the deviation is mostly within 10%. The microstructural analysis observes
the changes of cementitious products in stabilized clays and conrms the pozzolanic effect of the supplementary cementing materials. On account of the efciently pozzolanic effect, the addition of sodium silicate (<2%) and composite promoter (<4%) can be equivalent to several times cement content. Hence, the
economic and environmental mix design is developed with the addition of the supplementary cementing
materials.
2014 Elsevier Ltd. All rights reserved.

1. Introduction
Clayey soils occur in most regions of South and East China [1,2].
Soft clays are often encountered on many civil engineering project
sites, which lack sufcient strength to support the loading either
during construction or throughout the service life. To improve
the strength and stiffness of those less competent soils, soil stabilization with cementitious materials has been widely practiced
[35].
Corresponding author. Tel.: +86 13917109022.
E-mail address: hntchen@sjtu.edu.cn (C. Bing).
http://dx.doi.org/10.1016/j.conbuildmat.2014.08.087
0950-0618/ 2014 Elsevier Ltd. All rights reserved.

Soil stabilization is a technique introduced many years ago with

the main purpose to render the soils capable meeting the requirements of the specic engineering projects [6]. One of the effective
soft ground improvement techniques is in situ deep mixing [710].
This method has been developed during last two decades primarily
to pile columnar inclusions into the soft ground to composite
ground. It has been applied popularly and successfully in Southeast
Asia [11,12]. The commonly used stabilizers include ordinary Portland cement (OPC) and lime, with their stabilization mechanisms
being relatively well understood [1315]. However, the quicklime
reacts with water rapidly and it increases the difculty of deep
mixing. In China, OPC is the most common stabilizer since it is

355

M. Cong et al. / Construction and Building Materials 71 (2014) 354362

readily available at reasonable cost [16,17]. Nevertheless, a major

issue with OPC is that its production processes are energy intensive
and emit a large quantity of CO2 [18,19].
To improve the environmental acceptability and reduce the
construction cost of the deep mixing method, the replacement of
the cement by supplementary cementing materials such as y
ash and sodium silicate is one of the best alternative ways [20].
Application of sodium silicate for geotechnical works has been
reported by many researchers [2123]. Used as a component of soil
stabilizer, sodium silicate has unique advantages: (i) reliable and
proven performance, (ii) safety and convenient for construction,
and (iii) environmental acceptability and compatibility [22]. In
China, liquid sodium silicate with high silica modulus is applied
in the improvement of tunnel and subway foundation.
The fundamental mechanical properties of cement stabilized
soft clays have been experimentally investigated by many
researchers [2427]. These investigations mainly focus on the
inuence of water content and cement content on the mechanical
properties. Compressive strength development of cement stabilized y ashsoil mixtures and several models have been studied.
Claywater/cement ratio, which is the ratio of clay water content
to cement content (both reckoned in percentage) has been found
to investigate the combination effect from both water content
and cement content [28]. While the clay water content reects
the micro-fabric of soft clay, the cement content inuences the
level of bonding of that fabric. Horpibulsuk et al. [29] have introduced a mathematical model for predicting laboratory strength
development in cement stabilized clays at various water contents,
cement contents, and curing ages.
The strength of stabilized clay is governed by the growth of
cementitious products which are controlled by the hydration of
cement and the pozzolanic reactions. The role of sodium silicate
on the mechanical properties of the cement stabilized clay has
been investigated both from micro- and macro observations
[22,30]. Sodium silicate reacts with Ca2+ generated by the hydration of cement, forming insoluble calcium silicate which polymerizes further to form a gel that binds soil or sediment particles
together and lls voids [31]. However, the consumption of Ca2+
leads to the lower Ca(OH)2 which reduces the pozzolanic effect.
In this study, two clay stabilizers, OPC and CSCN consisting of
OPC, sodium silicate and composite promoter were used. To investigate the performance of clay stabilizers, the following properties
of the stabilized clays were determined: unconned compressive
strength (qu), failure strain (ef), and secant modulus (Es). The methods for assessing the strength development of stabilized clay are
veried on the basis of the claywater/cement content ratio
hypothesis. The pozzolanic effect of sodium silicate and composite
promoter is estimated by calculating the equivalent cement content. The mathematical models for predicting the compressive
strength and secant modulus of CSCN stabilized clay are developed.
Furthermore, the microstructures of clays stabilized with OPC and
selected CSCN stabilized clays were investigated with the aid of
SEM imaging, respectively.

Table 1
Chemical composition of clay and OPC.
Oxide

Silicon dioxide (SiO2)

Calcium oxide (CaO)
Aluminum oxide (Al2O3)
Ferric oxide (Fe2O3)
Magnesium oxide (MgO)
Sodium oxide (Na2O)
Potassium oxide (K2O)
Sulfur trioxide (SO3)
Loss on ignition (LOI)

Chemical composition (%)

Clay

OPC

57.02
3.63
16.42
6.79
3.68
0.81
3.59
0.05
6.43

21.60
64.44
4.13
4.57
1.06
0.11
0.56
1.74
0.76

The binding agents consist of Portland cement and the supplementary cementing materials. ASTM Type I ordinary Portland cement (hereinafter referred to as
cement, OPC) was used for all stabilized clay mixtures in this study. The chemical
composition of OPC is also shown in Table 1. The supplementary cementing materials are sodium silicate and composite promoter. Sodium silicate (SS), a syrupy
liquid, consists of SiO2 (29.48%), Na2O (9.52%) and the silica modulus (molar ratio
of SiO2 and Na2O) is 3.2. The density and pH are 1.43 g/cm3 and 11.98, respectively.
Composite promoter (CN) consists of sodium hydroxide and calcium chloride at the
mass ratio of 1:1. And sodium hydroxide and calcium chloride used in this study are
both chemically pure.
2.2. Casting, curing, and testing of specimens
In order to investigate the mechanical properties of OPC and CSCN stabilized
clays, different contents of clay stabilizers were admixed with specic amounts of
oven-dry clay. For preparing the samples, the water was added through a two-step
process. Firstly, the predetermined quantity of water was initially mixed with the
oven-dry clay, and the mass ratio of water and clay was 0.7. Then the rest of the
water was blended with the binders at the mass ratio of 0.5, and the binders consisting of the desired amount of promoter, OPC and sodium silicate were added
by the weight of dry clay. The soluble promoter was added into the mixtures in
the form of solution, and the sequential mixing with CaCl2 solution followed by
NaOH solution was selected. Initial mixing was carried out in a laboratory mixer
for at least 10 min and the mix was subsequently transferred to a cubic mold
(70.7 mm in length). For squeezing the air and achieve a homogeneous mixture,
the mold was put onto vibrating table to vibrate for at least 2 min. Then, the mold
was sealed and stored in the curing room (20 2 C, 98 2% RH) for 7 days. Afterwards, the stabilized samples were demolded and put into airtight vinyl bags in
the curing room until the testing ages.
The unconned compressive strength (UCS) test was performed on the samples
after 7, 28, 60 and 90 days of curing. The rate of vertical displacement was xed at
0.5 mm/min. At least three replicates of each sample set were prepared and tested
under the same conditions to assure reproducibility. In most cases, the results
under the same testing condition were reproducible with low mean standard deviations, SD (SD/
x < 10%, where 
x = the mean strength value). The failure strain (ef) of
the stabilized specimens was determined in UCS tests. The secant modulus (Es) is
dened as the ratio of one half of the compressive strength to the axial strain corresponding to this stress.
To conrm the pozzolanic reactions of the supplementary cementing materials
and examine the micromorphological change of the stabilized specimens, SEM
imaging was performed on selected samples. SEM samples were prepared by following a procedure suggested in previous study [35]. A 1  1  1 cm cubic specimen was trimmed off and then air dried in a desiccator at ambient temperature.
The dried specimen was broken into two parts and the debris on the surface was
removed with an adhesive tape. The specimen was mounted on an alumina stud
with conductive tapes, and then sputter coated with goldpalladium alloy.

3. Claywater/cement ratio and pozzolanic effect hypothesis

2. Experiment details
2.1. Materials
The soil sample is a kind of soft clay collected from the Shanghai Jiao Tong University campus in Shanghai, China, at a depth of 6 m. The soil contains high ne particle content, similar to many marine soft clayey soils. Particle size analysis was
performed on the soil by following the standard methods [32,33], about 100%
and 80% of the soil are ner than 2 mm and 0.075 mm, so clay and ne sand are
the major components in this soil. Its specic gravity is 2.699. The liquid and plastic
limits are approximately 42.36% and 24.27%, respectively. According to the Unied
Soil Classication System [34], this soil is a CL. The natural water content and pH
value were approximately 41.38% and 7.138%. The chemical composition of the clay
is shown in Table 1.

According to ASTM C 595, a pozzolan is dened as a siliceous

or siliceous and aluminous material, which in itself possesses little
or no cementitious value but will, in nely divided form and in the
presence of moisture, chemically react with calcium hydroxide to
form compounds possessing cementitious properties (pozzolanic
activity). Based on this concept, sodium silicate can be regarded
as a pozzolan. For composite promoter consisting of CaCl2 and
NaOH, the permeation of CaCl2 and NaOH solutions is expected
to facilitate the precipitation of Ca(OH)2 according to the reaction:

CaCl2 NaOH ! CaOH2 2NaCl

356

M. Cong et al. / Construction and Building Materials 71 (2014) 354362

Table 2
Mechanical properties of OPC stabilized clay at different curing ages.
OPC (%)

wc (%)

wc/C

10
12
15
20
25
30
40
50
60
70
80

75.0
76.0
77.5
80.0
82.5
85.0
90.0
95.0
100.0
105.0
110.0

7.50
6.33
5.17
4.00
3.30
2.83
2.25
1.90
1.67
1.50
1.38

Compressive strength (kPa)

Failure strain (%)

Secant modulus (MPa)

7 days

28 days

60 days

7 days

28 days

60 days

7 days

28 days

60 days

34.4
51.4
120.0
198.1
286.6
385.7
706.6
1473.9
1556.8
2343.7
2663.4

79.7
99.8
232.7
432.9
565.3
683.9
1147.4
2376.3
3277.7
3687.3
4806.7

106.8
142.7
335.9
538.6
736.8
845.2
1577.0
2585.6
4253.7
5043.9
6242.2

3.86
1.94
2.64
1.58
1.51
1.40
1.38
1.63
1.77
1.44
1.61

2.85
1.66
1.91
1.09
1.38
1.28
1.54
1.82
1.72
1.95
2.14

1.16
1.46
1.63
1.68
1.45
1.34
1.46
1.76
1.89
2.09
2.22

1.4
6.1
16.1
47.3
46.4
52.3
67.8
137.5
171.6
196.3
228.3

4.4
16.0
30.6
57.8
78.9
71.9
113.8
270.0
277.4
309.4
377.5

32.3
33.1
40.8
69.7
82.5
84.6
154.2
325.8
306.6
349.8
421.6

Table 3
Mechanical properties of CSCN stabilized clay at different curing ages.
Mix No.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16

OPC (%)

10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10

SS (%)

1
1
1
1
2
2
2
2
3
3
3
3
4
4
4
4

CN (%)

1
2
4
6
1
2
4
6
1
2
4
6
1
2
4
6

wc (%)

76.0
76.5
77.5
78.5
76.5
77.0
78.0
79.0
77.0
77.5
78.5
79.5
77.5
78.0
79.0
80.0

Compressive strength (kPa)

Failure strain (%)

7 days

28 days

60 days

7 days

28 days

60 days

7 days

28 days

60 days

192.6
213.8
312.7
299.1
277.3
312.3
489.2
347.5
217.3
359.3
325.7
345.1
5.3
25.6
162.7
138.8

348.7
436.6
566.3
627.9
426.9
546.7
739.4
717.3
301.7
456.4
441.5
474.5
40.7
105.3
301.4
266.9

497.6
548.0
756.5
697.8
529.9
627.8
917.6
817.7
386.5
535.0
568.6
625.8

2.00
1.99
1.07
0.64
1.32
1.17
1.05
0.76
0.99
1.74
1.53
2.43
3.76
3.44
2.37
1.66

1.28
1.15
0.92
0.83
1.38
1.46
0.91
0.76
1.97
1.12
2.02
1.27
2.71
2.73
1.30
1.21

1.13
1.34
1.41
1.73
1.45
1.12
0.85
0.93
1.57
1.44
1.81
1.63

24.2
30.3
55.23
137.8
50.3
58.5
86.1
173.3
38.6
63.2
65.4
48.9
0.1
0.5
31.0
18.3

64.4
78.4
99.6
88.5
75.0
91.3
106.1
113.3
26.9
97.1
40.0
93.9
0.8
12.3
83.8
71.3

81.4
92.2
112.9
53.8
88.1
100.6
121.5
98.4
63.4
79.5
56.4
117.8

For the reason of the very high solubility and diffusion rates of CaCl2
and NaOH in water in contrast to the solubility and diffusion rates
of Ca(OH)2, CaCl2 and NaOH solutions are sequentially permeated in
microporosity or very small pores of stabilized clays. The formation
of Ca(OH)2 in microporosity or very small pores developed much
stronger soilCa(OH)2 pozzolanic reactions than those caused by
directly adding Ca(OH)2 [36]. The possible mechanisms may be that
render CaCl2 and NaOH solutions are sequentially permeated stabilized clays. Hence, the pozzolanic effect of the supplementary
cementing materials can be regarded akin as an addition of cement
[37]. By considering that the content of sodium silicate and composite promoter can be equivalent to cement content, the equivalent
cement content (Ce) is calculated by the following equation:

C e k1  C1 k2  C2

where k1 and k2 are pozzolanic factors, C1 and C2 are sodium silicate and composite promoter content, respectively.
The total cement content (C) for the stabilized specimens is thus
the summation of input of cement (Ci) and equivalent cement content (Ce). The claywater/cement ratio hypothesis for the stabilized
clays developed as follows For given set of cement stabilized clay
specimens, the strength development depends only on the clay
water/cement ratio, wc/C [38]. The Ce is mainly dependent upon
the pozzolanic effect governing by the content of supplementary
cementing materials and alters with curing age.
To predict the mechanical properties of CSCN stabilized clay
specimens, the phenomenological models of OPC stabilized clay
and values of the pozzolanic factors are needed. The phenomenological models are the mathematical formulas which can

Secant modulus (MPa)

quantitatively describe the phenomena presented in experiments,

and the equations could be obtained from the tted curves with
a high coefcient of determination (R2 > 0.95) on the basis of
experimental results shown in Table 2. According to the experimental results of CSCN stabilized clay specimens, the C and Ce
are achieved. Then, the pozzolanic factors can be drawn by multiple linear regression analysis. Compare the predicted values and
actual results of selected CSCN stabilized clay specimens, the
precision of the models is veried.
4. Test results
Tables 2 and 3 show the experimental results of mechanical
properties of OPC and CSCN stabilized clay specimens at different
curing ages, respectively. The compressive strength and secant
modulus of OPC stabilized clays increase with increasing the OPC
content and decreasing the wc/C. After 28 days of curing, the compressive strength at 60 days increases signicantly with the aid of
pozzolanic effect. The secant modulus increases with curing age,
which is approximately consistent with the tendency of strength
development. Fig. 1 shows the stressstrain of specimens stabilized with different contents of OPC at 28 days of curing. It is noted
that the failure strain generally decreases with increasing the OPC
content, but the change in failure strain becomes irregular when
the OPC content is greater than 40% (wc/C < 2.25).
As shown in Table 3, the compressive strength at different curing ages is extremely sensitive to the contents of sodium silicate
and composite promoter, which implies that the supplementary
cementing materials perform effectively pozzolanic reactions. For

357

M. Cong et al. / Construction and Building Materials 71 (2014) 354362

4000

800
15% OPC wc=5.17

3500

30% OPC w c=2.83

600

50% OPC wc=1.90

Axial stress (kPa)

Axial stress (kPa)

3000

70% OPC wc=1.50

2500

8% SCSCN
10% SCSCN
12% SCSCN
14% SCSCN
16% SCSCN

700

20% OPC wc=4.00

2000
1500

500
400
300

1000

200

500

100

0
0

Axial strain (%)

Axial strain (%)

all mixtures stabilized with CSCN, the compressive strength rstly

increases with the content of supplementary cementing materials,
and then achieves the peak values when the contents of sodium
silicate and composite promoter are 2% and 4%, respectively. That
is to say, the compressive strength of CSCN stabilized clay
decreases as the contents of sodium silicate and composite
promoter exceeding 2% and 4%. The secant modulus increases with
increasing the content of composite promoter when the contents
of sodium silicate are 1% and 2% and generally increases with curing age. Nevertheless, the secant modulus of No. 4 and 8 specimens
decreases with curing ages. Therefore, the analysis of pozzolanic
factors (k1, k2) and equivalent cement content (Ce) is based on
the results of No. 13 and No. 57 specimens.
To verify the precision of the pozzolanic factors, experimental
tests on the specimens stabilized with SCSCN which consists of
OPC, sodium silicate and composite promoter at the mass ratio of
5:1:2 were performed and the results are shown in Table 4. The
compressive strength and secant modulus at different curing ages
increase with the increasing content of SCSCN, which is similar
with that presented in OPC stabilized specimens. Fig. 2 shows
the stressstrain of specimens stabilized with different contents
of SCSCN at 28 days of curing. The similar stressstrain behavior
of specimens stabilized with OPC and SCSCN is gured out. As
shown in Fig. 2, an approximate tendency in failure strain of SCSCN
stabilized specimens with that of OPC stabilized specimens is
obtained.

5. Analysis and discussion

5.1. Unconned compressive strength
The wc/C can be applied to analyze the compressive strength
development of OPC stabilized clay with different OPC content at

Fig. 2. Stressstrain relationship of SCSCN stabilized specimens at 28 days of

curing.

Unconfined compressive strength (kPa)

Fig. 1. Stressstrain relationship of OPC stabilized specimens at 28 days of curing.

7 days
28 days
60 days

6000
5000
4000
3000
2000
1000
0
1

wc /C
Fig. 3. Fitted curves of strength development in the OPC stabilized clay using the
claywater/cement ratio hypothesis.

a given curing age. The relationship between wc/C and compressive

strength (qu) can be expressed as the following equation [37]:

qu

wc =CB

where qu is the compressive strength of OPC stabilized clay at a

given curing age, wc is water content in the mixture expressed in
percentage, C is the total cement content, A and B are empirical
constants.
From the test results, the phenomenological models of OPC
stabilized clay on the basis of the generalized two-dimensional
quwc/C plot is presented in Fig. 3. According to the tted curves,

Table 4
Mechanical properties of SCSCN stabilized clay at different curing ages.
Mix No.

1
2
3
4
5

OPC (%)

5
6.25
7.5
8.75
10

SS (%)

1
1.25
1.5
1.75
2

CN (%)

2
2.5
3
3.5
4

wc (%)

74.0
75.0
76.0
77.0
78.0

Compressive strength (kPa)

Failure strain (%)

7 days

28 days

60 days

7 days

28 days

60 days

Secant modulus (MPa)

7 days

28 days

60 days

86.3
124.4
203.1
328.6
489.2

164.5
247.7
416.9
583.4
739.4

221.4
327.6
532.7
716.2
917.6

2.29
1.43
1.26
1.30
1.05

1.84
1.55
1.46
1.01
0.91

1.50
1.17
1.11
1.37
0.85

10.5
36.4
49.0
62.7
86.1

36.9
43.5
68.1
94.0
106.1

41.5
61.8
78.6
120.0
121.5

358

M. Cong et al. / Construction and Building Materials 71 (2014) 354362

Table 5
The C, Ce and pozzolanic factors for the compressive strength of CSCN stabilized clay at different curing ages.
Mix no.

Ci

C1

C2

1
2
3
5
6
7

10
10
10
10
10
10

1
1
1
2
2
2

1
2
4
1
2
4

7 days

28 days

Ce

19.26
20.21
23.84
22.43
23.67
28.70

9.26
10.21
13.84
12.43
13.67
18.70

Factors

Ce

k1 = 5.13
k2 = 2.19

19.91
21.90
24.57
21.71
24.07
27.45

9.91
11.90
14.57
11.71
14.07
17.45

the mathematical expressions for the phenomenological models at

different curing ages are:

5958

qu7days
qu28days
qu60days

R2 0:997

2:5

wc =C
10601

wc =C2:5
13732

wc =C2:5

4a

R 0:998

4b

R2 0:996

4c

It is observed that the variation of parameter A is signicant and is

dependent on curing ages. The variation of parameter B is approximately constant and irrespective of curing ages. This observation is
consistent with the result presented in Ref. [39].
To observe the total cement content (C) of CSCN stabilized clay
specimens, the compressive strength (qu1) presented in Table 3 is
needed. The solution formula is deduced from Eq. (3), proposes
the following function:

wc
C 0  1
B

60 days

A
qu1

where qu1 is obtained from the experimental tests.

As mentioned before, the pozzolanic effect of the supplementary cementing materials varies with curing age. Hence, the equivalent cement content (Ce) of sodium silicate and composite
promoter also varies with curing ages, which indicates that the calculation of C depends on curing ages. By combination of Eqs. (4)
and (5), the calculational results of the C, Ce and pozzolanic factors
are shown in Table 5. It is notable that k1 decreases and k2
increases with curing ages, this implies that the long-term
pozzolanic effect of composite promoter is stronger than that of

Ce

Factors

k1 = 4.89
k2 = 2.35

20.16
21.09
24.31
20.81
22.41
26.43

10.16
11.09
14.31
10.81
12.41
16.43

k1 = 4.31
k2 = 2.39

sodium silicate. The C of CSCN stabilized clay specimens at different curing ages can be determined as the following equations:

C 0 7 days C i 5:13C1 2:19C2 R2 0:993

2

6a

C 28 days C i 4:89C1 2:35C2 R 0:978

6b

C 0 60 days C i 4:31C1 2:39C2 R2 0:969

6c

That is to say, to achieve the equivalent strength of specimens stabilized with 20% OPC, only about 8.5% OPC, 1.7% sodium silicate and
3.4% composite promoter are needed when the mass ratio of
sodium silicate and composite promoter is 1:2.
Based on the above analysis, the theoretical prediction models
for the compressive strength of CSCN stabilized clay specimens
are drawn and its mathematical expressions are expressed by combination of Eqs. (4) and (6):

5958

q0u7days 

wc
C i 5:13C12:19C2

7a

2:5

10601

q0u28days 

2:5

7b

2:5

7c

wc
C i 4:89C12:35C2

13732

q0u60days 

wc
C i 4:31C12:39C2

where q0u is the predicted compressive strength of CSCN stabilized

clay.
Figs. 46 shows the comparison of the actual values (qu) of
unconned compressive strength of SCSCN stabilized clay speci

mens at different curing ages with those predicted q0u by Eqs.
(7). As shown in Figs. 46, the actual values are mainly below
the predicted values, and the actual values mostly differ within
10% from the predicted values.

1100

600

1000
actual
predicted = actual
10% from predicted

actual
predicted = actual
10% from predicted

900

Actual qu(28days) (kPa)

500

Actual qu(7days) (kPa)

Factors

400

300

200

800
700
600
500
400
300
200

100

100
0
0

100

200

300

400

500

600

Predicted q'u(7days) (kPa)

Fig. 4. Comparison of predicted and actual values of 7-day compressive strength of
SCSCN stabilized clay specimens.

100

200

300

400

500

600

700

800

900 1000 1100

Predicted q'u(28days) (kPa)

Fig. 5. Comparison of predicted and actual values of 28-day compressive strength
of SCSCN stabilized clay specimens.

359

M. Cong et al. / Construction and Building Materials 71 (2014) 354362

3.0

1200
1100

actual
predicted = actual
10% from predicted

Actual qu(60days) (kPa)

1000
900

7 days
28 days
60 days

2.5

800

2.0

700
600

1.5

500
400

1.0

300
200

0.5

100
0
0

100 200 300 400 500 600 700 800 900 1000 1100 1200

0.0
2.4

Predicted q'u(60days) (kPa)

2.8

3.2

3.6

4.0

4.4

4.8

wc/C'

Fig. 6. Comparison of predicted and actual values of 60-day compressive strength

of SCSCN stabilized clay specimens.

Fig. 8. Failure strain of CSCN stabilized specimens at different curing ages.

to decrease as decreasing the wc/C or wc/C, which is consistent

with earlier ndings by Horpibulsuk [34].

5.2. Failure strain

The tendency of failure strain for OPC or geopolymer stabilized
soil has been reported in Refs. [40,41]. It is difcult to propose a
regression model for OPC stabilized soil because of the distractiveness of failure strain of specimens stabilized with different contents of OPC. Although a linear model presented in Ref. [31] is
applied to predict the values of failure strain geopolymer stabilized
soil, it has a low coefcient of determination (R2 < 0.8).
Figs. 7 and 8 show the failure strain of specimens stabilized
with OPC and CSCN at different curing ages, respectively. As shown
in Fig. 7, the failure strain itself is small, varying from 1% to 4% and
the tendency of the failure strain is indeterminate. The failure
strain generally tends to decrease with curing age when the wc/C
is higher than 5 and the opposite tendency is achieved as the
wc/C less than 5. The C presented in Fig. 8 is obtained from
Eq. (3). In general, the failure strain of CSCN stabilized clay varies
from 0.5% to 2.5%, which is smaller than that of OPC stabilized clay.
In comparison to the tendency of failure strain of OPC stabilized
clay as the wc/C ranging from 3 to 5, that of CSCN stabilized clay
is much more irregular. Hence, it is almost impossible to develop
the phenomenological models for the prediction of failure strain
of the specimens stabilized OPC-based cementing materials. Still,
the failure strain of stabilized clay at different curing ages tends

5.3. Secant modulus

Due to the similar development of compressive strength and
secant modulus, the wc/C can also be used to analyze the secant
modulus development of OPC stabilized clay. Based on the analysis
of strength development in OPC stabilized clay, the following equation to express the relationship between wc/C and Es is thus
proposed.

Es

wc =CN

where Es is the secant modulus of OPC stabilized clay at a given curing age, wc is water content in the mixture expressed in percentage,
C is the total cement content, A and B are empirical constants.
Fig. 9 shows the phenomenological models of secant modulus
development in the OPC stabilized clay by tting the experimental
results on the basis of the generalized two-dimensional Eswc/C
plot. According to the tted curves, the general relationship
between wc/C and Es at different curing ages can be advanced as:

480
4.0

420

3.5
3.0

Secant modulus (MPa)

7 days
28 days
60 days

2.5
2.0
1.5

7 days
28 days
60 days

360
300
240
180
120
60

1.0
0
0.5

wc/C
Fig. 7. Failure strain of OPC stabilized specimens at different curing ages.

wc/C
Fig. 9. Fitted curves of secant modulus development in the OPC stabilized clay
using the claywater/cement ratio hypothesis.

360

M. Cong et al. / Construction and Building Materials 71 (2014) 354362

Table 6
The C, Ce and pozzolanic factors for the secant modulus of CSCN stabilized clay at different curing ages.
Mix no.

Ci

C1

C2

1
2
3
5
6
7

10
10
10
10
10
10

1
1
1
2
2
2

1
2
4
1
2
4

417

Es7days
Es28days
Es60days

7 days

28 days

Ce

16.43
18.68
26.14
24.53
26.79
33.40

6.43
8.68
16.14
14.53
16.79
13.40

Factors

Ce

k1 = 5.70
k2 = 2.63

21.22
23.74
27.35
23.18
25.84
28.48

11.22
13.74
17.35
13.18
15.84
18.48

R2 0:980

wc =C1:86
691

wc =C1:86
790

wc =C1:86

R2 0:964

Ce

Factors

k1 = 5.33
k2 = 2.74

22.39
24.10
27.23
23.52
25.42
28.51

12.39
14.10
17.23
13.52
15.42
18.51

k1 = 5.48
k2 = 2.69

expressions of the theoretical prediction models for the secant

modulus CSCN stabilized clay are given as the following:

9b

E0s7days 

417

wc
C i 5:70C12:63C2

R2 0:955

9c

wc
C 0  1
N

12a

1:86

691

E0s28days 

1:86

12b

1:86

12c

wc
C i 5:33C12:74C2

790

E0s60days 

wc
C i 5:48C12:69C2

where E0s is the predicted secant modulus of CSCN stabilized clay.

10

M
Es1

140

C 0 7 days C i 5:70C1 2:63C2 R2 0:995

11a

C 0 28 days C i 5:33C1 2:74C2 R2 0:988

11b

C 60 days C i 5:48C1 2:69C2 R 0:984

100

60
40
20
20

40

60

80

100

120

140

Predicted E's(28days) (MPa)

Fig. 11. Comparison of predicted and actual values of 28-day secant modulus of
SCSCN stabilized clay specimens.

180

actual
predicted = actual
10% from predicted

80

80

11c

The pozzolanic factors vary with curing age. The 60-day pozzolanic
factors are higher than 28-day pozzolanic factors, which is different
from the tendency in the pozzolanic factors for compressive
strength. By combination of Eqs. (9) and (11), several mathematical

90

actual
predicted = actual
10% from predicted

120

Actual Es(28days) (MPa)

where Es1 is the secant modulus of CSCN stabilized clay obtained

from the experimental tests.
The total cement content (C), OPC content (Ci) and equivalent
cement content of the supplementary cementing materials (Ce)
can be obtained by combination of Eqs. (9) and (10). Table 6 shows
the values of C, Ce and pozzolanic factors. Based on the testing
results, their generalized interrelationship can be presented as:

actual
predicted = actual
10% from predicted

160

Actual Es(60days) (MPa)

70
60
50
40
30
20

140
120
100
80
60

10
0

Factors

9a

Similar with the phenomenological models for the strength development in OPC stabilized clay, the increase of parameter M with
curing age is signicant and parameter N keeps constant.
The total cement content (C) for secant modulus of CSCN stabilized clay can be calculated based on the following expression:

Actual Es(7days) (MPa)

60 days

40
0

10

20

30

40

50

60

70

80

90

Predicted E's(7days) (MPa)

Fig. 10. Comparison of predicted and actual values of 7-day secant modulus of
SCSCN stabilized clay specimens.

40

60

80

100

120

140

160

Predicted E's(60days) (MPa)

Fig. 12. Comparison of predicted and actual values of 60-day secant modulus of
SCSCN stabilized clay specimens.

M. Cong et al. / Construction and Building Materials 71 (2014) 354362

Figs. 1012 shows the comparison of the actual values (Es) of

secant modulus of SCSCN stabilized clay specimens at different


curing ages with those predicted E0s by Eqs. (7). As shown in
Fig. 10, the deviation between actual and predicted values of specimens stabilized with 8% SCSCN is much larger than the others.
And as shown in Figs. 10 and 12, the actual values at different curing ages are occasionally greater than the predicted values, which
is different with the tendency in compressive strength. Generally,
the actual values mostly differ within 10% from the predicted
values.
5.4. SEM characterization
Fig. 13 shows the surface morphology of untreated clay, and
specimens stabilized with different contents of OPC and SCSCN at
28 days of curing, respectively. Compared with untreated clay

361

(Fig. 13(a)), there are some brous CSH gels appeared in 10%
OPC stabilized clay (Fig. 13(b)). Besides, more brous CSH gels
are formed as the OPC content increasing to 20% (Fig. 13(c)). However, as shown in Fig. 13(d)(f), the SCSCN stabilized clay does not
have many brous CSH gels. There exist much anomalously
agglomeration and thin laminated CSH gels in the specimen of
SCSCN stabilized clay, which is much higher than that of the OPC
stabilized clay. Although the microstructure of specimens stabilized with 12% and 16% SCSCN is still rough, the discrete particles
were bonded signicantly closely and the gaps among the particles
were more tightly lled. This implies that the supplementary
cementing materials can react with OPC and active clay minerals
and generates higher amount of hydration and pozzolanic products
which bond among the clay particles. The higher degree of bonding
and more compact microstructure are believed to result in higher
compressive strength and secant modulus.

Fig. 13. SEM photos of (a) untreated clay, (b) 10% OPC, (c) 20% OPC, (d) 8% SCSCN, (e) 12% SCSCN and (f) 16% SCSCN stabilized specimens for 28 days.

362

M. Cong et al. / Construction and Building Materials 71 (2014) 354362

6. Conclusions
1. The addition of the supplementary cementing materials
(i.e. sodium silicate and composite promoter) can signicantly improve the compressive strength and secant
modulus of OPC stabilized clay. The optimal improvement
is found at the sodium silicate content of 2% and the
composite promoter content of 4%. Hence, the analysis for
compressive strength and secant modulus is based on the
test results of specimens stabilized with less than 2% of
sodium silicate and 4% of composite promoter.
2. Based on claywater/cement ratio and pozzolanic effect
hypothesis, wc/C is used as the prime parameter to achieve
the phenomenological models for predicting the mechanical properties of specimens stabilized CSCN and the supplementary cementing materials can be regarded akin as the
corresponding addition of cement. Therefore, the summation of OPC content and equivalent cement content of the
supplementary cementing materials is the total cement
content.
3. The interrelationship between mechanical properties (i.e.
compressive strength and secant modulus) and wc/C at
different curing ages is proposed. The models are useful
in predicting the compressive strength and secant modulus
wherein the content of OPC and the supplementary
cementing materials varies. And the deviation between
predicted values and laboratory results is mostly within
10%. Its notable that it is difcult to identify an accurate
relationship between failure strain and claywater/cement
ratio.
4. With the aid of SEM imaging, brous and thin laminated
CSH gels are formed in the specimens stabilized with
OPC and SCSCN, respectively. A more compact microstructure is formed in the specimens stabilized with SCSCN,
which conrmed that the supplementary cementing
materials perform effectively pozzolanic reactions. The
addition of the supplementary cementing materials is
recommended for the economic and environmental mix
design.

Acknowledgement
This research work was nancially supported by the National
Natural Science Foundation of China (Nos. 50708059 and
51379122) and National Key Technology R&D Program (No.
2014BAL03B03).
References
[1] Gao GR. Formation and development of the structure of collapsing loess in
China. Eng Geol 1988;25(2):23545.
[2] Gao GR. The distribution and geotechnical properties of loess soils, lateritic
soils and clayey soils in China. Eng Geol 1996;42(1):95104.
[3] Terashi M, Tanaka H, Mitsumoto T, Niidome Y, Honma S. Fundamental
properties of lime and cement treated soils (2nd report). Rep. Port Harbour
Res. Inst. 1980;19(1):3362 [in Japanese].
[4] Miller GA, Azad S. Inuence of soil type on stabilization with cement kiln dust.
Constr Build Mater 2000;14(2):8997.
[5] Degirmenci N, Okucu A, Turabi A. Application of phosphogypsum in soil
stabilization. Build Environ 2007;42(9):33938.
[6] Rogers CDF, Glendinning S, Roff TEJ. Lime modication of clay soils for
construction expediency. Geotech Eng 1997;125(4):2429.
[7] Kawasaki T, Niina A, Saitoh S, Suzuki Y, Honjo Y. Deep mixing method using
cement hardening agent. In: Proceedings of the 10th international conference
on soil mechanics and foundation engineering. Stockholm; 1981. p. 7214.
[8] Bruce DA, Bruce MEC, Dimillio AF. Dry mix methods: a brief overview of
international practice. In: Proceedings of international conference on dry mix
methods for deep soil stabilization. Rotterdam; 1999. p. 1525.

[9] Saitoh S, Suzuki Y, Shirai K. Hardening of soil improved by deep mixing

method. In: Proceedings of 11th international conference on soil mechanics
and foundation engineering, vol. 3. San Francisco; 1985. p. 17458.
[10] Kamon M, Bergado DT. Ground improvement techniques. In: Proceeding of 9th
Asian regional conference on soil mechanics and foundation engineering.
Bangkok; 1991. p. 52646.
[11] Porbaha A. State of the art in deep mixing technology: Part I. Basic concepts
and overview. In: Proceedings of the ICE-ground improvement; 1998. p. 81
92.
[12] Ye GB, Ye SL. Field study of improved soft soil by cementsoil mixed piles. J
Tongji Univ 1995;23(3):2705 (in Chinese).
[13] Prusinski JR, Bhattacharja S. Effectiveness of Portland cement and lime in
stabilizing clay soils. Transp Res Rec: J Transp Res Board
1999;1652(1):21527.
[14] Niazi Y, Jalili M. Effect of Portland cement and lime additives on properties of
cold in-place recycled mixtures with asphalt emulsion. Constr Build Mater
2009;23(3):133843.
[15] Rajasekaran G, Narasimha Rao S. Lime stabilization technique for the
improvement of marine clay. Soils Found 1997;37(2):97104.
[16] Duan JW, Gong XN, Zheng GX. Load transfer behavior of cement treated soil
column. Chin J Geotech Eng 1994;16(4):18 [in Chinese].
[17] Huang HS, Yan TN, Lan K. Laboratory experiment of the anticorrosion of
cement stabilized soft soil in deep mixing pile. Geol Sci Technol Inform
2005;24(7):858 [in Chinese].
[18] Gartner E. Industrially interesting approaches to low-CO2 cements. Cem
Concr Res 2004;34(9):148998.
[19] Meyer C. The greening of the concrete industry. Cem Concr Compos
2009;31(8):6015.
[20] Shen W, Zhou M, Zhao Q. Study on limey ashphosphogypsum binder.
Constr Build Mater 2007;21(7):14805.
[21] Ruff CG, Davidson DT. Lime and sodium silicate stabilization of
montmorillonite clay soil. Highway Res Board Bull 1961;304:7692.
[22] Kazemian S, Prasad A, Huat BB, Bazaz JB, Abdul Aziz FN, Mohammad Ali TA.
Inuence of cementsodium silicate grout admixed with calcium chloride and
kaolinite on sapric peat. J Civ Eng Manage 2011;17(3):30918.
[23] Yang KH, Song JK, Ashour AF, Lee ET. Properties of cementless mortars
activated by sodium silicate. Constr Build Mater 2008;22(9):19819.
[24] Terashi M, Tanaka H, Okumura T. Engineering properties of lime treated
marine soils and DMM. In: Proceedings of 6th Asian regional conference on
soil mechanics and foundation engineering; 1979. p. 1914.
[25] Horpibulsuk S, Miura N, Bergado DT. Undrained shear behavior of cement
admixed clay at high water content. J Geotech Geoenviron Eng, ASCE
2004;130(10):1096105.
[26] Horpibulsuk S, Rachan R, Suddeepong A, Chinkulkijniwat A. Strength
development in cement admixed Bangkok clay: laboratory and eld
investigations. Soils Found 2011;51(2):23951.
[27] Horpibulsuk S, Phojan W, Suddeepong A, Chinkulkijniwat A, Liu MD. Strength
development in blended cement admixed saline clay. Appl Clay Sci
2012;55:4452.
[28] Horpibulsuk S, Miura N, Nagaraj TS. Claywater/cement ratio identity for
cement admixed soft clays. J Geotech Geoenviron Eng, ASCE
2005;131(2):18792.
[29] Horpibulsuk S, Miura N, Nagaraj TS. Assessment of strength development in
cement-admixed high water content clays with Abrams law as a basis.
Geotechnique 2003;53(4):43944.
[30] Kazemian S, Prasad A, Huat BBK, Bazaz JB, Mohammed TA, Aziz FA. Effect of
aggressive pH media on peat treated by cement and sodium silicate grout. J
Cen South Univ Technol 2011;18(3):8407.
[31] Yonekura R, Kaga M. Current chemical grout engineering in Japan. In:
Proceeding of grouting, soil improvement and geosynthetics. ASCE; 1992. p.
72536.
[32] GB/T 50123. Standard for soil test method. Ministry of construction of the
Peoples Republc of China; 1999 [in Chinese].
[33] JTG E40. Test methods of soils for highway engineering. Ministry of
communications of the Peoples Republc of China; 2007 [in Chinese].
[34] ASTM2487. Standard practice for classication of soils for engineering
purposes (unied soil classication system). ASTM International; 2011.
[35] Barden L, Sides G. Sample disturbance in the investigation of clay structure.
Geotechnique 1971;21(3):21122.
[36] Thyagaraj T, Rao SM, Sai Suresh P, Salini U. Laboratory studies on stabilization
of an expansive soil by lime precipitation technique. J Mater Civ Eng
2012;24(8):106775.
[37] Papadakis VG, Tsimas S. Supplementary cementing materials in concrete: Part
I: Efciency and design. Cem Concr Res 2002;32(10):152532.
[38] Horpibulsuk S, Rachan R, Chinkulkijniwat A, Suddeepong A. Analysis of
strength development in cement-stabilized silty clay from microstructural
considerations. Constr Build Mater 2010;24(10):201121.
[39] Horpibulsuk S, Rachan R, Suddeepong A. Assessment of strength development
in blended cement admixed Bangkok clay. Constr Build Mater
2011;25(4):152131.
[40] Kaniraj SR, Havanagi VG. Compressive strength of cement stabilized y ash
soil mixtures. Cem Concr Res 1999;29(5):6737.
[41] Zhang M, Guo H, El-Korchi T, Zhang G, Tao M. Experimental feasibility study of
geopolymer as the next-generation soil stabilizer. Constr Build Mater
2013;47:146878.