EXPERIMENTAL STUDY ON STRENGTH OF CEMENT STABILIZED CLAY

WOO-SIK KIM , NGUYEN MINH TAM & DU-HWOE JUNG

Corresponding Author: Du-Hwoe Jung

Assoc. Prof., Dept. of Civil Engineering
Pukyong National University
Korea.

ABSTRACT
This paper describes the effect of factors on the strength characteristics of cement treated clay
from laboratory tests performed on cement mixed clay specimens. It is considered that several
factors such as soil type, sample preparing method, quantity of binder, curing time, etc. can have
an effect on strength characteristics of cement stabilized clay. A series of unconfined
compression tests have been performed on samples prepared with different conditions. The
results indicated that soil type, mixing method, curing time, dry weight ratio of cement to clay
(A w), and water-clay to cement (wc/c) ratio were main factors which can have an influence on
unconfined compressive strength, modulus of elasticity, and failure strain of cement stabilized
clay. Unconfined compressive strength of soil-cement samples prepared from dry mixing method
was higher than those prepared from wet mixing method.

KEYWORDS: Cement Stabilized Clay, Deep Mixing Method, soft clay, unconfined compressive
strength, modulus of elasticity, failure strain.

INTRODUCTION A comparative study, carried out by Shiells

The behavior of soft clay can be improved
with soil-cement columns created by a deep
mixing method (DMM). Soil-cement
columns are used primarily to reduce
settlement and improve stability. The deep
mixing method encompasses a group of
technologies that provide in situ soil
treatment. The method in which dry
powdered cement is used as a stabilizing
agent is generally known as the dry method
of deep mixing, whereas the use of
stabilizing agent in a slurry form is referred
to as the wet method of deep mixing. The
dry mixing method is usually used to
improve soft clay that has high natural water
content because of sufficient groundwater to
hydrate the cement (Esrig 1999) 2001). The
wet mixing method is generally
recommended for dry or arid environments
or sites with deep ground water tables.
Although the dry mixing technology is
predominant in Scandinavian countries and
the wet mixing method has been frequently
used in Japan, there is a very limited
conclusive research available on the choices
of one of these two methods, which one can
result in a better improvement.
JOURNAL OF ENGINEERING, DESIGN ANDTECHNOLOGY
VOLUME 3 NUMBER 2 2005 PP116-126
2005 CAPE PENINSULA UNIVERSITY OF TECHNOLOGY
et al. (2003), on the strength of soil cement
admixture of these two mixing methods,
stated that the wet mix method generally
used a dosage rate higher than the dry
mixing method. The dosage rate ranged
from 180 to 400 kg/m3 for the wet mixing
method and it ranged from 90 to 180 kg/m3
for the dry mixing method. However, Lin
and Wong (1999) inversely reported that the
unconfined compressive strength of
improved soil using cement slurry was
lower than that using cement powder. This
may be attributed to the difference in water
cement ratio of stabilized soils prepared
from two mixing methods. Specimens
prepared from the wet mixing method had
higher water content and further increased
the water-cement ratio, and thereby result ing
in lower strength. These results from Shiells
et al. (2003) and Lin and Wong (1999)
should not be compared each other because
sample preparation methods are different in
these two investigations.
The main purpose of this study is to
investigate the strength characteristics of
cement stabilized clay samples prepared
 EXPERIMENTAL STUDY ON STRENGTH OF CEMENT STABILIZED CLAY
W OO-SIK KIM, NGUYEN M INH TAM & DU-HWOE JUNG

with different mixing methods and mixing sample preparation in both dry and wet
conditions. Additionally, the effects of other mixing methods. Although some differences
important factors including curing time, dry exist in Scandinavian and Japanese methods,
weight ratio of cement to clay (A w), water- a sample preparation method proposed by
clay to cement (wc/c) ratio, on the strength Japanese Coastal Development Institute of
charcateristics of cement treated clay were Technology (CDIT, 2002) was employed in
also evaluated to provide a better the study.
understanding of the behavior of cement
stabilized clay. In order to investigate several factors
affecting strength characteristics of cement
PROPERTIES OF NATURAL SOFT CLAY stabilized clay, samples were prepared with
Two different types of clays were obtained different conditions as indicated in Table 2.
from Nak-dong river basin. Their physical
and mechanical properties were determined TEST R ESULTS AND DISCUSSION
and compared in Table 1. STRESS-STRAIN CURVES
Stress-strain curves obtained from
Table 1: Properties of natural clays unconfined compression tests on soil-
Properties Clay 1 Clay 2 cement samples cured for 28 days are
Unit weight (kN/m3 ) 17.8 18.2 presented in Figure 1 for various dry weight
Liquid Limit LL(%) 48.0 42.4
ratio of cement to clay (Aw). Samples with
Plastic Limit PL (%) 24.7 27.6
Water Content w(%) 52.7 45.2 higher cement content showed more brittle
Plasticity Index (Ip ) 23.3 14.8 failures. Samples with Aw of 15% and 20%
Liquid Index (IL ) 1.2 1.2 showed brittle failures, where the peak
Specific Gravity 2.61 2.67 strengths achieved at lower strains and their
Grain size distribution
residual strengths were much lower than
Sand 5% 8%
Sieve (% < #200) 95% 92% their peak strengths. On the other hand,
q u (kPa) 41 65 samples with lower Aw showed decreased
Axial failure strain f (%) 9.5 6.5 peak strengths and larger failure strains. It is
E50 (Kpa) 779 1239 also interesting to realize from Figure 1 that
the failure strain tends to decrease with
SAMPLE PREPARATION AND TEST increasing unconfined compressive strength.
CONDITIONS
Normal Portland cement was used for

Table 2: Sample preparation methods and test conditions employed in the study
Initial Additional Modified initial Cement mixed clay paste
Number water water to water content of water-clay to Initial
Test Soil Mixing
of content of cement ratio of natural soil after Dry weight ratio cement content Water
number group Method of cement toclay
samples natural clay cement slurry mixing ratio content
(%) (%) (%) (%) - (%)
Wet 10 45.22 100 55.22 10 5.52 50.20
1 Clay 1
Dry 10 45.22 100 55.22 10 5.52 50.20
6 52.71 100 57.71 5 11.54 54.96
6 52.71 100 60.71 8 7.59 56.21
Wet
2 Clay 2 6 52.71 100 62.71 10 6.27 57.01
method
6 52.71 100 67.71 15 4.51 58.88
6 52.71 100 72.71 20 3.64 60.59
6 52.71 50 57.71 10 5.77 52.46
6 52.71 100 62.71 10 6.27 57.01
Wet
3 Clay 2 6 52.71 150 67.71 10 6.77 61.55
method
6 52.71 200 72.71 10 7.27 66.10
6 52.71 300 82.71 10 8.27 75.19

117 JOURNAL OF ENGINEERING, DESIGN ANDTECHNOLOGY

VOLUME 3 NUMBER 2 2005 ISSN 1726-0531
 EXPERIMENTAL STUDY ON STRENGTH OF CEMENT STABILIZED CLAY
W OO-SIK KIM, NGUYEN M INH TAM & DU-HWOE JUNG

2000

Aw = 20%
Aw = 15%
1600 Aw = 10%
Aw = 8%

Deviator stress (Kpa)

Aw = 5%
1200

800

400

0
0 1 2 3 4

Axial Strain (%)

Figure 1: Stress-strain curve s of unconfined compression test on soil-cement samples tested

after 28 days of curing

EFFECT OF MIXING M ETHOD AND CURING specimens prepared from cement powder
TIME than those prepared from cement slurry. 3)
To investigate the effect of mixing methods The strength of soil cement specimens is
on strength characteristics of soil-cement gained as results of physical and chemical
samples, 10 soil-cement samples were reactions between stabilizing agent, clay
prepared with clay 1 by each mixing method minerals, and water including hydration,
as indicated in Table 2. As shown in Figure pozzolanic reaction, ion exchange,
2, the unconfined compressive strength of flocculation, carbonation etc. that were
soil-cement samples mixed by dry mixing dependent of sequence of soil-cement
method was much higher than that by wet mixing, or of mixing method.
mixing method. Furthermore, the rate of
strength gain was appeared to be much Unconfined compressive strength of both
faster for dry mixed samples compared to soil-cement samples tended to develop
wet mixed samples. These differences in rapidly in an early curing stage and the
two mixing methods may be explained by development of strength tended to slow
the reasons hereafter: 1) Hampton and Edil down afterwards. Strength gaining of
(1998) reported that the compaction of wet cement stabilized soil is mainly caused by a
mixed samples was much less uniform than combination of cement hydration and
the dry mixed specimens with the same pozzolanic reaction that are time dependent.
other experiment conditions. 2) The During the pozzo lanic reactions, cementing
specimens prepared from cement powder phases are produced and calcium and
shows a rapid increase in pH value (due to hydroxide ions in solution are constinuously
the hydration of cement) compared to consumed and more must be dissolved to
specimens prepared from slurry cement maintain solution equilibrium. Strengh
(Fam and Santamarina, 1996). This result increase with time reflects the cumulative
associates with the reports of Massaki effects of a cementation process (Locat et
(1996) and Gotoh (1996) that the strength of al., 1990). In the early stage of the curing
soil cement samples increase with increasing period where the pozzolanic reactions are
pH value. Thus, a stiffer structures matrix of mechanically felt as a rapidly increase in
stabilized soil would be formed. strength. During this period, more and more
Combination from these conclusions results efficiency particle or aggregate bonds are
in the higher strength gain of soil cement created by the neoformed minerals. After

JOURNAL OF ENGINEERING, DESIGN ANDTECHNOLOGY 118

VOLUME 3 NUMBER 2 2005 ISSN 1726-0531
 EXPERIMENTAL STUDY ON STRENGTH OF CEMENT STABILIZED CLAY
W OO-SIK KIM, NGUYEN M INH TAM & DU-HWOE JUNG

exceeding this period, a reduction in the rate For soils stabilized by the wet mixing
of increase of shear strength results from at method, the 28-day unconfined compressive
least three following reasons: i) completion strength, which is typically used for design
of pozzolanic reactions due to exhaustion of value, are generally about 1.5 times the 7-
cement, ii) reactions are still going on but day values, similar to results indicated by
with greater difficulty for solutes to diffuse Yang et al. (1998). The 56-day strengths is
within the soil cement matrix, or iii) reaction approximate 1.2 times the 28-day values or
products are still produced in abundance but 1.8 times the 7-days values of strength.
their effects on strength are not as Similarly, for soils stabilized by the dry
pronounced as in initial stage, since the soil method, the 28-day unconfined compressive
has already a new, more rigid structure. strengths are about 1.7 times the 7-day
values and about 0.8 times the 56 days value
Especially, it can be found from Figure 2b of strength. These test values is into the
that the unconfined compressive strength of range of qu with the change of curing time
soil-cement sample increases almost linearly indicated by Bruce (2000) that the 28-day
with a logarithmic curing time, which is strengths were about 1 to 2 times the 7-day
consistent with data published in literatures values for wet mixing method, whereas this
(Kitazume et al. 2003, and Hayashi et al. difference were about 1.5 to 2.4 times for
2003). Based on the predicted strength gains dry mixing method. The 60-day strengths
in the soil cement specimens, designers may can be 1.2 to 2.5 times the 28-day values for
estimate a time schedule for construction. wet method.

1000 1000
Dry mixing method
Unconfined compressive strength qu (Kpa)

Unconfined compressive strength qu (Kpa)

R2 = 0.9848 Dry mixing method

800 800
qu = 217 ln(t)+18; R2 = 0.9848

600 600

400 400

200 Wet mixing method 200

R2 = 0.9412 Wet mixing method
qu = 86 ln(t)+60.67; R2 = 0.9412
0 0
0 10 20 30 40 50 60 1 10 100
Curing time (days) Curing time (days)

(a) In arithmetic scale (b) In logarithmic scale

Figure 2: Relationship between qu of soil-cement samples and curing time for dry and wet
mixing methods
50000
5

Wet Mixing Method

40000 2
E50 = 64.523 x qu; R = 0.9404 4
Dry mixing method: R = 0.9295
2
Axial failure strain (%)

30000 3
E 50 (kPa)

20000 2

Dry mixing method

10000 (a) 2
E50 = 52.319 x q u ; R = 0.9163 1 (b)
2
Wet mixing method: R = 0.8176

0 0
0 200 400 600 800 1000 0 200 400 600 800 1000
Unconfined compressive strength q u (kPa) Unconfined compressive strength q u (kPa)

Figure 3: Comparison of test results between wet and dry mixing methods

119 JOURNAL OF ENGINEERING, DESIGN ANDTECHNOLOGY

VOLUME 3 NUMBER 2 2005 ISSN 1726-0531
 EXPERIMENTAL STUDY ON STRENGTH OF CEMENT STABILIZED CLAY
W OO-SIK KIM, NGUYEN M INH TAM & DU-HWOE JUNG

As shown in Figure 2 also, the unconfined
compressive strength of samples after 3 days
of curing for both dry and wet mixing
method is about 160 and 259 kPa,
respectively, that is much higher than the
unconfined compressive strength of natural
soft clay sample of 41 kPa. This implies that
the recovery of strength of cement mixed
clay after mixing is progressed very fast
which was also consistent with results
studied by Shen et al. (2003). This strength
recovery is attributed by a combination of
cementation (pozzolanic reaction),
exrothemic hydration process.
Figure 3a shows a relationship between the
modulus of elasticity and the unconfined
compressive strength of soil-cement samples
and Figure 3b shows a relationship between
the axial failure strain and the unconfined
compressive strength of soil-cement
samples. The modulus of elasticity tended to
increase with increasing unconfined
compressive strength, whereas the failure
strain tended to decrease with increasing
unconfined compressive strength. This
implies that soil-cement samples with higher
strength tend to show more brittle failures.
In addition, soil-cement samples using the
wet method provided slightly higher
modulus of elasticity whereas soil-cement
samples using the dry method provided
greater failure strain. This implies that soil-
cement samples using the wet method tend
to show more brittle failures compared to
those using the dry method.
EFFECT OF DRY WEIGHT RATIO OF CEMENT TO
CLAY, AW (C EMENT CONTENT)
Bergado et al. (1996) proposed that the
relationship between unconfined
compressive strength and cement content
can be divided into 3 zones: Inactive Zone,
Active Zone, and Inert Zone as shown in
Figure 4a. The shapes of the relationship
curve between qu and Aw presented by
Tatsuoka and Kobayashi (1983), and
Bouazza et al. (2004) were similar to the
curve shown in Figure 4a. However, Lin
(2000) proposed a different relationship
between unconfined compressive strength
and cement content of soil-cement as shown
in Figure 4b. This aspect was also
investigated in the study to validate these
two conflicting investigation results.
The influence of cement content on
unconfined compressive strength was
investigated by performing unconfined
compression tests on soil-cement samples
prepared with clay 2. The ratio of additional
water to cement was fixed at 100%. Soil-
cement samples were mixed with Aw of 5%,
8%, 10%, 15%, and 20% by wet method.
All these samples were tested after 7, 14 and
28 days of curing and test result was plotted
in Figure 5. As expected, unconfined
compressive strength showed a tendency to
increase with increasing cement content.
The unconfined compressive strength
increased rapidly increased up to the cement
content of 20% and then tended to slow
down. The shape of the relationship was
appeared to be similar to that proposed by
Bergado et al. (1996).

A
a) After Bergado et al. (1996) (b) After Lin (2000)
Figure 4: Comparison of the relationship between qu and Aw of cement stabilized clay

JOURNAL OF ENGINEERING, DESIGN ANDTECHNOLOGY 120

VOLUME 3 NUMBER 2 2005 ISSN 1726-0531
 EXPERIMENTAL STUDY ON STRENGTH OF CEMENT STABILIZED CLAY
W OO-SIK KIM, NGUYEN M INH TAM & DU-HWOE JUNG

2000

28 days curing: R = 0.9818

2

Unconfined compressive strength qu (Kpa)

1500
14 days curing: R2 = 0.9945

7 days curing: R = 0.9311

2

1000

500

0
0 5 10 15 20
Dry weight ratio of cement to clay, Aw (%)
Figure 5: Variation of qu with Aw

A conclusion about the more important

effect between the cement content Aw and Figure 6 depicts the role of wc/c ratio on
curing time can be also given from Figure 5 unconfined compressive strength
which the change of unconfined development of soil cement samples tested
compressive strength is higher due to after 28 days of curing. The relationship of
variation of cement content than due to the wc/c ratio and q u appears to be nonlinear
change of curing time. Consequently, the over the range of wc/c ratio investigated.
cement content has more influence on Unconfined compressive strength of soil-
unconfined compressive strength than cement samples tended to decrease with
curing time. increasing wc/c ratio.

EFFECT OF CLAY WATER/CEMENT RATIO The best trend line produced by fitting the
(WC/C) curve to the data is displayed under
Miura et al. (2001) suggested that a prime exponential Equation (1) in which the
factor governing the engineering parameters unconfined compressive strength decrease
of cement-stabilized soil is the clay- with increasing wc/c ratio.
water/cement ratio, wc/c. This is defined as a
the ratio of initial water content of the soil q u = ( wc / c ) (1)
b
(%) to the cement content. The wc/c is a
paramount parameter combining the effects where a and b are constant obtained by
of water content and cement content, which least-squares regression depending on the
is more advantageous than the parameter characteristic of clay, cement and curing
water content. The importance of this ratio time (Nagaraj, 2002). In genaral, R-square
is emphasized by Najaraj (2002) that the values in excess of 0.9 are obtained with this
natural clays, which have different relathionship. The relation shown in
mineralogy, exhibit different strength Equation (1) merits an examination for
characteristics after mixing with cement. engineering evaluation. It should be possible
However, the strength development of a to estimate how far increasing the cement
particular clay-cement mixture after a content can reduce the cutting time. On the
certain curing time is only dependent upon other hand, it should be possible to estimate
the wc/c, which is structural parameter how much cement content is needed for
reflecting both the effects of microfabric and obtaining a required level of strength
their cementation. The denominator Aw corresponding to the curing time and clay
would reflect the level of bonding that could water content.
be induced.

121 JOURNAL OF ENGINEERING, DESIGN ANDTECHNOLOGY

VOLUME 3 NUMBER 2 2005 ISSN 1726-0531
 EXPERIMENTAL STUDY ON STRENGTH OF CEMENT STABILIZED CLAY
W OO-SIK KIM, NGUYEN M INH TAM & DU-HWOE JUNG

4000

Unconfined compressive strength qu (Kpa)

(wc/c)
qu(28 days) = 5588.7/1.364 ;
2
R = 0.9456
3000

Nagaraj's trend line

Miura's trend line

2000

1000

Jacobson's trend line

0
2 4 6 8 10 12
Water-clay to cement content (wc/c) ratio
Figure 6: Comparison of test data and published data

1000
Unconfined compressive strength qu (Kpa)

Clay 2
qu = 212.22 ln(t)+118.76;
800
2
R = 0.9014

600

400

200
Clay 1
2
qu = 86 ln(t)+60.67; R = 0.9412
0
1 10 100

Curing time (%)

Figure 7. Influence of natural clay type on qu of soil cement mixture cured at 7, 14 and 28 days

A comparison of test results obtained from and 2 under the same conditions of curing
the study with data published in literatures is time, wet mixing method, dry weight ratio
presented in Figure 6. The trend line of cement to clay, and additional water to
obtained from unconfined compressive tests cement ratio. The index properties of two
on soil-cement samples cured for 28 days is clay type were summarized in Table 1. As
fairly consistent with the results reported by shown in Figure 7, unconfined compression
Miura et al. (2001) and Jacobson (2002). testing results of soil cement samples
However, the trend line proposed by prepared from two type of natural clay were
Nagaraj (2002) was appeared to be far off compared each other.
from others. This can be attributed to
differences in the characteristics of natural Each clay has its own mineralogy with
clay such as water content, mineralogy, different physical and chemical properties,
Atterberg limits, organic content, pH value and each of these properties may affect the
etc., and cement type. strength improvement (Gotoh, 1996). To
derive a strength relation that incorporates
EFFECT OF NATURAL CLAY TYPE all relevant factors, especially at a
In this section, the influence of natural clay fundamental level, is ideal and desirable, but
type on unconfined compressive strength of is extremely complicated. Comparison of
soil cement mixture is performed on clays 1 two tests indicated that the unconfined
JOURNAL OF ENGINEERING, DESIGN ANDTECHNOLOGY 122
VOLUME 3 NUMBER 2 2005 ISSN 1726-0531
 EXPERIMENTAL STUDY ON STRENGTH OF CEMENT STABILIZED CLAY
W OO-SIK KIM, NGUYEN M INH TAM & DU-HWOE JUNG
compressive strength of soil cement samples
made from natural clay 2 is greater than that
made from natural clay 1. This can is
explained by several reasons hereafter.
i. Due to difference in water content of
two natural clays, the water content of
natural clay 1 is higher of clay 2.
Resulting in the higher unconfined
compressive strength of soil cement
specimens prepared natural clay 2 than
prepared from natural clay 1.
ii. This reason can be attributed to the
grain size distribution or clay
mineralogy. The effect of cement
gradually increases with decreasing
clay content or increasing sand
proportion in natural clay. As
indicated in Table 1, the sand
proportion of clay 2 is 8%, while this
of clay 1 is only 5%. Thereby, the soil
cement specimens prepared from clay
2 produced higher unconfined
compressive strength than those
prepared from clay 1 in conformity
with study of Taki and Yang (1991).
iii. Bergado et al (1996) also reported that
the effect of cement decrease with
increasing plasticity index of natural
soil. The plasticity index of natural
soil clay 1 is 23.3, while this value for
natural soil clay 2 is only 14.76.
iv. The organic content and pH values are
considered to be the most influential
soil parameters to the strength of soil-
cement (Massaki 1996, Gotoh 1996).
The strength decreases with increasing
organic content or decreasing pH
value.
v. Another approach is also due to the
difference in chlorite, carbonates, and
sulfates being as reaction inhibitors
(Locat et al. 1990). Or general
speaking is that the ground chemistry
of each site is different will result in
different implications.
AXIAL FAILURE STRAIN OF SOIL CEMENT
SAMPLES
Determining the axial train of soil cement
mixture at failure state is significant part to
evaluate the interaction between deep
123 JOURNAL OF ENGINEERING, DESIGN ANDTECHNOLOGY
VOLUME 3 NUMBER 2 2005 ISSN 1726-0531
mixing cement columns and unstabilized
soil around the columns. Basing on this
interaction, the stability analysis of
structures founded on soft soil improved by
soil cement columns system can be
evaluated. The failure strain has also a large
influence on the reduction of the bearing
capacity and on the increase of lateral
displacement of the columns. In this section,
the axial failure strain is, therefore,
investigated.
The relationship between unconfined
compressive strength and failure strain of
soil cement samples obtained from test
results on unconfined compression tests in
laboratory is shown in Figure 8. It can seen
from the figure that the failure strain
decreases in general with increasing
unconfined compression strengt and that
failure occurs at relatively small axial
strains, at 1% to 3.5% with a unconfined
compressive strength largely varies about
from 100 to 2000 kPa. Similar and
consistent results have been reported e.g. by
Balasubramanian and Buensuceso (1989),
Erdil et al. (1991). Besides, the scatter in the
relationship between unconfined
compressive strength and axial failure strain
is fairly large as axial failure strain is high.
This scatter is partly due to the variability in
the soil from which the soil cement is
derived, different constituent mixing
proportion of soil cement mixture, type of
mixing method, and also degree of mixing,
or uniformity of soil cement that is
produced. That is, the higher the uniformity
of soil cement, the smaller is the scatter.
Compared to the failure strain of natural
clay, It can be seen that the axial strain at the
peak shear strength of soil cement samples
performed on unconfined compression test
was small, about 1% to 3.5%, while this
value of natural clays ware 9.5 for soil clay
1 and 6.5 for soil clay 2. Therefore, it is
important to consider the small failure strain
of the cement columns which will reduce the
interaction of the cement columns with the
unstabilized soil between the columns.
 EXPERIMENTAL STUDY ON STRENGTH OF CEMENT STABILIZED CLAY
W OO-SIK KIM, NGUYEN M INH TAM & DU-HWOE JUNG

5
Clay 1
Wet mixing method
Dry mixing method
4 Clay 2
Different dry weight ratio of cement to clay

Axial failure strain 0f (%)

Different ratio of additional water to cement

1
f (%) = 21.223xqu(-0.3624)
R2 = 0.7212
0
0 500 1000 1500 2000
Unconfined compressive strength qu (kPa)

Figure 8: Relationship between qu and axial failure strain of soil cement samples.

CONCLUSION cement content of 20% and then

Based on the results obtained from tended to slow down. The shape of the
unconfined compression tests carried out on relationship was appeared to be
soil-cement samples prepared with different similar to that proposed by Bergado et
conditions, the following can be appropriate; al. (1996).
1. Soil-cement samples with higher 5. Unconfined compressive strength of
cement content showed more brittle both soil-cement samples prepared
failures. with dry and wet mixing methods
2. The unconfined compressive strength tended to develop rapidly in an early
of soil-cement samples prepared from curing stage and the development of
dry mixing method was greater than strength tended to slow down
those prepared from wet mixing afterwards.
method. However, soil-cement 6. The cement content has more
samples prepared from dry mixing influence on unconfined compressive
method tended to show more brittle strength than curing time
failures compared to those prepared 7. The trend line showing an effect of
from wet mixing method. water-clay to cement content ratio
3. The modulus of elasticity tended to obtained from unconfined
increase with increasing unconfined compressive tests on soil-cement
compressive strength, whereas the samples cured for 28 days was fairly
failure strain tended to decrease with consistent with the results reported by
increasing unconfined compressive Miura et al. (2001) and Jacobson
strength. However, soil-cement (2002).
samples using the wet method 8. The axial failure strain of soil cement
provided slightly higher modulus of samples is much smaller than that of
elasticity whereas soil-cement samples natural clays and thus the interaction
using the dry method provided greater of the brittle cement columns with the
failure strain. unstabilized soil between the columns
4. Unconfined compressive strength will reduce. Result that the
showed a tendency to increase with applicability of a composite shear
increasing cement content. The strength analysis to ground stabilized
unconfined compressive strength with soil cement columns in present
increased rapidly increased up to the design practice is uncertain longer.
JOURNAL OF ENGINEERING, DESIGN ANDTECHNOLOGY 124
VOLUME 3 NUMBER 2 2005 ISSN 1726-0531
 EXPERIMENTAL STUDY ON STRENGTH OF CEMENT STABILIZED CLAY
W OO-SIK KIM, NGUYEN M INH TAM & DU-HWOE JUNG
REFERENCES
Balasubramaniam, A. S. and Buensuceso. B.
R. (1989). The Overconsolidated
Behaviour of Lime Treated Soft Clay
Proc 12th Int. Conf. Soil Mech. And
found. Engng., Jio de Janeiro, Brazil,
Vol. 2, 1335-1338.
Bergado, D. T., Anderson, L. R, Miura, N.
and Balasubramaniam, A. S. (1996).
Soft Ground Improvement in Lowland
and Other Environments, ASCE press,
234-304.
Bouazza, A., Kwan, P. S., and Chapman, G.
(2004) Strength Properties of Cement
Treated Coode Island Silt by the Soil
Mixing Method, Proc. Geotechnical
Engineering for Transportation
Projects, ASCE press, 1421-1428.
Bruce, D. A. (2000). "An Introduction to the
Deep Mixing Methods as used in
geotechnical applications - Volume
III: The Verification and Properties of
Treated Ground." DRAFT Federal
Highways Administration, FHWA-RD-
99-167.
CDIT (Coastal Development Institute of
Technology). (2002). The Deep
Mixing Method: Principle, Design and
Construction. A.A. Balkema, The
Netherlands
Erdil, R., Tuncer, Adnan, A., and Basma.
(1991). Strength and Stress-Strain
Characteristic of a Lime Treated
cohesive Soil. Tranportation
Research Record, No. 1295, Soil
Stabilization, Washington D.C.
Esrig, M. I. and MacKenna, P. E. (2001).
Lime Cement Column Ground
Stabilization for I-15 in Salt Lake
City. Practice Periodical on
Structural Design and Construction,
ASCE, Vol. 4, No. 3, 104-115.
Fam, M. A. and Santamaria, J. C. (1996).
Study of Clay-Cement Sluries with
Mechanical and Electromagnetic
Waves. Journal of Geotechnical
Engineering, ASCE, 365-373.
Gotoh, M. (1996). Study on Soil Properties
Affecting the Strength of Cement
Treated Soils. Proc. of IS-Tokyo 96,
2nd Int. Conf. on Ground Improvement
125 JOURNAL OF ENGINEERING, DESIGN ANDTECHNOLOGY
VOLUME 3 NUMBER 2 2005 ISSN 1726-0531
Geosystems, Tokyo, Vol. 1, 399-404.
Hampton, M. B., and Edil, T. B. (1998).
Strength Gain of Organic Ground
With Cement Type Binders. Soil
Improvement for Big Digs,
Geotechnical Special Publication,
ASCE, No.81, Boston, MA, October,
135-148.
Hayashi, H., Nishikawa, J., Ohishi, K., and
Terashi, M. (2003). Field
Observation of Long-Term Strength of
Cement Treated Soil. Grouting and
Ground Treatment, Proc. of the 3rd
In. Conf., ASCE Special Publication
No. 120, New Orleans, 598-609.
Jacobson, J. R. (2002). Factors Affecting
Strength Gain in Lime-Cement
Columns and Development of a
Laboratory Testing Procedure.
Master Thesis, Virginia Tech. Institute
Kitazume, M., Takeshi, N., Terashi, M., and
Ohishi, K. (2003) Laboratory Tests
on Long-Term strength of Cement
Treated Soil. Grouting and Ground
Treatment; Proceedings of the 3rd
International Conference, ASCE
Geotechnical Special Publication No.
120New Orleans, 586-597.
Lin, K. Q. (2000). Behaviour of DCM
Columns under Highway
Embankment at Bridge Approaches.
Thesis for the Degree of Doctor of
Philosophy, Nanyang Technical
University.
Lin, K. Q. and Wong, I. H. (1999). Use of
deep cement mixing to reduce
settlements at bridge abutments.
ASCE 125, No. 4, 309-320.
Locat, J., Berube M.A., Choquette M.
(1990). Engineering Behavior of
Cement Treated Bangkok Soft Clay.
Canadian Geotechnical Journal, Vol.
27, No. 3, January, 294-304
Massaki, G. (1996). Study on Soil
Properties Affecting the Strength of
Cement Treated Soils. Grouting and
Deep Mixing, Proc. of IS-Tokyo 96,
2nd Int. Conf. on Ground Improvement
Geosystems,14-17 May 1996, Tokyo,
399-404.
Miura, N., Horpibulsuk, S., and Nagaraj, T.
 EXPERIMENTAL STUDY ON STRENGTH OF CEMENT STABILIZED CLAY
W OO-SIK KIM, NGUYEN M INH TAM & DU-HWOE JUNG

S. (2001). Engineering Behavior of M. (2003). Deep Mixing: An Owner's

Cement Stabilized Clay at High Water
Content. Soils and Found.. Japanese
Geotechnical Society. Vol. 41, No. 5,
33 46.
Nagaraj, T. S. (2002). One Day Lecture on
Ground Improvement Techniques.,
Technical committee of Thick Deposit
(ATC-7), Dong A University, Busan,
Korea, 59-95.
Shen, S. L., Miura, N., Han, J., and Koga, H.
(2003). Elevation of Property
Changes in Surrounding Clays due to
Installation of Deep Mixing
Columns. Grouting and Ground
Treatment; Proc. of the 3rd Int. Conf.,
ASCE GSP, No. 120, New Orleans,
634-645.
Shiells, D. P., Pelnik III, T. W., and Filz, G.
Perspective. Grouting and Ground
Treatment, Proc. of the 3rd Int. Conf.,
ASCE GSP, No. 120, New Orleans,
489-500.
Taki, O. and Yang, D.S. (1991). Soil-
cement mixed wall technique, ASCE
special conference, Denver, CO.
Tatsuoka, F. and Kobayashi, A. (1983).
Triaxial strength characteristics of
cement-treated soft clay. Proc. of the
8th European Con. of SMFE, 421-26.
Yang, D. S., Yagihashi, J. N., and
Yoshizawa, S. S. (1998). Dry Jet
Mixing For Stabilization of Very Soft
Soils and Organics Soil., Soil
Improvement for Big Digs, ASCE
GSP, No.81, Boston, MA, October,
96-110

JOURNAL OF ENGINEERING, DESIGN ANDTECHNOLOGY 126

VOLUME 3 NUMBER 2 2005 ISSN 1726-0531