GeoCongress 2012 © ASCE 2012 920

Accelerated Testing of Cement Treated Singapore Marine Clay Cured under

Elevated Temperature
Downloaded from ascelibrary.org by Sardar Vallabhbhai Natl Inst Of Tech (svnit) on 07/27/16. Copyright ASCE. For personal use only; all rights reserved.

Y. T. Lu1, T. S. Tan1 and K. K. Phoon1

1
Centre for Soft Ground Engineering, Department of Civil and Environmental
Engineering, National University of Singapore, P.O. Box 119260, Singapore; PH (65)
65162180; FAX (65) 68742248; email: luyitan@nus.edu.sg

ABSTRACT

Cement stabilized Singapore marine clay is the material associated with the
Stabilized Dredged Fill (SDF) technology practiced in Singapore. A novel
accelerated curing/testing procedure is proposed in this study making use of a
constant temperature hot water bath. The accelerated testing technique enables the
later-age strength of the SDF material to be predicted at a very early curing age
(~30hrs). The proposed accelerated curing procedure involves curing the cement
mixed clay specimens in a hot water bath at 60ºC for 24hrs followed by 6hrs of
cooling. Unconfined compression tests and bender element tests are conducted on
specimens cured under both the elevated temperature curing condition (30hrs) and
the normal room temperature curing condition (7 days and 28 days). Test results
show that both the unconfined compressive strength (UCS) and the small-strain shear
stiffness (Gmax) are controlled by the water/cement ratio under both curing conditions.
Strength-to-strength, stiffness-to-stiffness and strength-to-stiffness correlations are
established between the two conditions. The effect of soil type on theses correlations
is studied as well.

INTRODUCTION

Background

Huge amounts of unwanted soils are generated from urban construction and routine
dredging works. In land-scarce Singapore, it is an attractive proposition to reuse the
unwanted soils as reclamation fill, which solves not only the problem of the shortage
of fill materials but also the disposal of unwanted soils, as dumping these soils in
highly built-up Singapore is no longer viable.
Stabilized Dredged Fill (SDF) technology is a realization of this proposition.
In SDF technology, the dredged clay at high water content is mixed with small
amount of stabilizing agent, usually cement. The highly workable mixture is then cast
at the designated costal location through an air-compressed pumping system. SDF
technology has become more popular in recent years, notable projects include the
Ishinomaki reclamation project (Porbaha et al. 1999; Sakamoto 1998), Central Japan
International Airport man-made island project (Kitazume and Satoh 2003), and more
recently, parts of the reclaimed extension of Haneda Airport (Morohoshi et al. 2010).

GeoCongress 2012
 GeoCongress 2012 © ASCE 2012 921

The current SDF design criteria focus on the overall stability of the
reclamation bund, which is controlled by the strength of the SDF mixture after
curing. The 7-day and 28-day unconfined compression tests constitute the main
component of the SDF quality control (QC) scheme, as the test results are indicative
of the mixture final strength. Such a QC scheme is adopted in a land reclamation
project in Singapore using SDF technology. A similar QC scheme has been reported
Downloaded from ascelibrary.org by Sardar Vallabhbhai Natl Inst Of Tech (svnit) on 07/27/16. Copyright ASCE. For personal use only; all rights reserved.

by Kitazume and Satoh (2005). However, it would be too late for remedial measures
to be implemented if defective materials were detected after 7 days at earliest. Even
removal of the defective materials can be technically difficult if the cement dosage is
excessive. Moreover, considering the typical scale of a large reclamation project and
the associated high production rate, which can be as high as 1000 m3/hr (e.g.:
Kitazume and Satoh 2005), failure in detecting young defective materials may incur
serious economical losses.

Accelerated Testing Technique

The early strength of a cementitious material increases when cured under elevated
temperature. Accelerated testing technique using elevated curing temperature has
been employed as a part of the field quality control since as early as 1930s (Patch
1933), and received wider recognition in concrete industry since late 1970s, as was
marked by the publication of ACI SP-56 (1978). Accelerated testing of concrete
materials using elevated curing temperature has also been included in international
standards (e.g.: ASTM 2003; BSI 1983). Inspired by the accelerated testing technique
used for concrete materials, an accelerated testing procedure for SDF mixtures is
proposed based on a hot water bath to cure the SDF mixtures under elevated
temperatures (60 ºC). Unconfined compression (UC) tests and bender element (BE)
tests are conducted on SDF mixtures under both the elevated curing temperature and
the normal air-conditioned room temperature (23±2 ºC) conditions. Correlations of
engineering properties are attempted between the later age (7-day or 28-day in this
study) under normal curing condition and the early age under elevated curing
condition.

MATERIALS

The clay samples used in this study belong to two sub-types of the Singapore Upper
Marine Clay (UMC). The basic properties of the two UMC samples are slightly
different, as summarized in Table 1. Portland Blast Furnace Cement (PBFC) with
65% of slag content is used as the stabilizing agent for SDF production. The same
type of cement is being used in a land reclamation project in Singapore, in which the
SDF technology is practiced.

Table 1. Basic properties of the two different types of Singapore UMC

Particle Fraction (%)
Soil Type LL PL Gs Clay (<0.002mm) Silt Sand (>0.06mm)
UMC Type-A 85 36 2.69 52 46 2
UMC Type-B 92 46 2.67 54 45 1

GeoCongress 2012
 GeoCongress 2012 © ASCE 2012 922

METHODS

Sample Mixture Preparation

Samples are pretreated and water contents are determined one day before mixing with
cement. Pretreatment is essentially a mixing process in which the sample is
Downloaded from ascelibrary.org by Sardar Vallabhbhai Natl Inst Of Tech (svnit) on 07/27/16. Copyright ASCE. For personal use only; all rights reserved.

homogenized into slurry form by mechanical mixing. A small quantity of seawater is

usually added into the sample to facilitate the pretreatment using the Hobart mixer.
Seawater is used to remold the soil samples in this study, as the samples are dredged
from seabed in actual reclamation site.
Prior to mixing the soil sample with PBFC, the required additional seawater is
added into the pretreated sample to achieve a target water content based on the initial
water content of the pretreated sample. Mixing starts right after the desired amount of
PBFC is poured into the mixing bowl. The mixing process takes 10 mins in total, in
which one minute is allowed for scraping the materials attached to the side wall and
bottom of the mixer bowl using spatula so as to avoid any “dead corners” of the
machine mixing. The Hobart mixer operates at a constant rate of 180 rpm. The total
unit weight and the water content of the mixture are taken upon completion of mixing
to monitor the consistency of the mixture.

Test Specimen Preparation

The mixture is placed into plastic molds of 50 mm in diameter and 100 mm in height
immediately after mixing. To minimize the air voids trapped inside the specimen, the
mixture is compacted following a standardized working procedure: the mixture is
placed into the mold in three layers (40% full, 80% full and 100% full, respectively),
and the mixture is compacted after each layer has been placed. The manual
compaction is performed by slowly tamping the mold along with the mixture on the
ground by fixed number of blows (60 blows per layer). Similar specimen making
procedure has been reported by Tan et al. (2002).

Proposed Elevated Temperature Curing Procedure for Accelerated Testing

After the specimens have been prepared, the ones meant for accelerated curing are
fully immersed in seawater in the constant temperature water bath. The seawater
temperature is maintained at 60 ºC from the point of immersion, and the constant
temperature is maintained throughout the entire curing period (24 hrs) as well. The
rest of the specimens arising from the same batch of mixture are to be cured under
normal room temperature condition (room temperature = 23±2 ºC)
“Thermal shock” arises due to the internal stresses induced by nonlinear
temperature distribution within a solid body, which can be caused by a sudden change
of ambient temperature (Kristensen and Hansen 1994). One typical example of
“thermal shock” is that incandescent bulbs may be exploded by splashing cold water
onto its hot glass surface. In order to protect the specimens against thermal shock, a
cooling regime is proposed in this study. After the 24-hr curing period in the 60 ºC
water bath, the specimens are immediately put into a plastic box (L 25cm by W 18cm

GeoCongress 2012
 GeoCongress 2012 © ASCE 2012 923

by H 14cm) containing seawater preheated to 60 ºC as well. The box with specimens

immersed in seawater inside with the cover closed is then allowed to cool down
naturally in air. The seawater time-temperature history is observed to vary little from
test to test. Figure 1 shows the time-temperature history of the cooling regime from
five independent measurements of the water temperature in the cooling box using a
thermometer. Based on these results, it is judged that specimens cooled for 6 hrs in
Downloaded from ascelibrary.org by Sardar Vallabhbhai Natl Inst Of Tech (svnit) on 07/27/16. Copyright ASCE. For personal use only; all rights reserved.

the cooling box prior to testing would not be subjected to significant thermal shock.

Figure 1. Temperature-time history of the accelerated curing stages with cooling

regimes obtained from five independent measurements

Testing Programme

Specimens cured under either normal or elevated temperature condition are tested in
unconfined compression (UC). The bender element (BE) test is conducted before the
UC test due to its non-destructive nature. The detailed testing programme is
summarized in Table 2. Mixes are designed based on the commonly encountered
range of the SDF constituent proportions in the project practicing SDF technology in
Singapore. For each designed mix, the specimens are tested in triplicate for both the
UC and BE tests under either curing condition.

Bender Element Test

Bender element test is a simple, fast and non-destructive testing method, which
features make it a promising tool for the in-situ quality control of SDF. The small-
strain shear modulus (Gmax) can be obtained from the bender element test using the
following equation:
Gmax = ρ·vs2
where ρ is total density of the specimen, and vs is the measured shear wave velocity.
Basic testing and interpretation procedures can be found in Yamashita et al.
(2009). In this study, the GDS bender element system is used. The test specifications
are summarized in Table 3. Excitation frequencies of 1kHz, 2kHz, 5kHz and 10 kHz
were tested in a number of trial tests, and vs obtained under the different frequencies
were compared. It was found that signals under 2kHz showed the best possible
quality over the entire range of vs. Furthermore, there was but marginal difference in

GeoCongress 2012
 GeoCongress 2012 © ASCE 2012 924

the magnitude of vs between 2kHz and frequencies larger than 2kHz, and the “bump”
pattern of wave shape distortion implying the near field effect was rarely observed in
the amplitude-time plots under 2kHz. It is understood that there is much debate
regarding the “correct” wave path length to wave length ratio to avoid any near field
effect, however it is beyond the scope of this paper to discuss this issue in detail.
Shear wave signals are interpreted using the cross-correlation method proposed by
Downloaded from ascelibrary.org by Sardar Vallabhbhai Natl Inst Of Tech (svnit) on 07/27/16. Copyright ASCE. For personal use only; all rights reserved.

Viggiani and Atkinson (1995). The associated code is written in MATLAB, which
enables computerized automatic signal interpretation in an effort to minimize human
errors. Four shear wave velocity measurements are taken for each specimen. Very
consistent velocity measurements are obtained, showing error margin <±1% and
therefore the average values are used for analyses.

Table 2. Summary of experimental programme

w/s w/c s/c UCS correlations attempted Gmax correlations attempted
30-hr/7-day 30-hr/28-day 30-hr/7-day
150 7.50 5.0 A A
180 9.00 5.0 A A
135 7.43 5.5 AB A B
165 9.08 5.5 AB B
185 10.18 5.5 AB A B
142 8.50 6.0 A A
175 10.50 6.0 A A
135 9.45 7.0 AB A B
155 10.85 7.0 B A B
165 11.55 7.0 AB A B
185 12.95 7.0 AB AB B
140 11.20 8.0 A A
155 12.40 8.0 A A
170 13.60 8.0 A A
135 11.48 8.5 AB A B
155 13.18 8.5 AB A B
150 13.50 9.0 A A
139 12.50 9.0 A A
135 13.50 10.0 AB A B
140 14.00 10.0 AB A B
Note: (1) 30-hr refers to the accelerated curing condition and 7-day refers to the normal curing
condition. (2) w = mass of water, s = mass of soil solid and c = mass of cement. (3) A refers to treated
UMC type-A and B refers to treated UMC type-B

Table 3. Bender element test specifications

Waveform Sinusoidal (single cycle)
Input Voltage ±14V
Frequency 2kHz in this study
Sampling Frequency 500kHz

RESULTS AND DISCUSSIONS

Effect of Water/Cement Ratio on Strength

GeoCongress 2012
 GeoCongress 2012 © ASCE 2012 925

The 28-day UCS, 7-day UCS and accelerated UCS attained under elevated curing
temperature are plotted against the mixture water/cement ratio in Figure 2, Figure 3
and Figure 4 respectively. It is evident that the strength of the mixture decreases with
increasing water/cement ratio. The intentionally varied soil/cement ratio appears to
have relatively minor effect on the strength, as long as the water/cement ratio remains
unchanged. This observation suggests that the water/cement ratio is the primary
Downloaded from ascelibrary.org by Sardar Vallabhbhai Natl Inst Of Tech (svnit) on 07/27/16. Copyright ASCE. For personal use only; all rights reserved.

strength influencing factor for the range of the constituent proportions considered.
This finding is in line with those studies conducted on similar materials, i.e. cement
treated soft marine sediments at very high water contents (e.g.: Boutouil and
Levacher 2005; Horpibulsuk et al. 2003; Liu et al. 2008; Miura et al. 2001; Rahman
and Taiyab 2008). Furthermore, as the strength is controlled by the water/cement
ratio regardless of the curing condition (either room temperature curing or elevated
temperature curing), it may be inferred that the strength development mechanisms
between the two curing conditions are comparable.

Figure 2. 28-UCS versus Water/cement ratio

Figure 3. 7-day UCS versus Figure 4. Accelerated UCS versus

Water/cement ratio Water/cement ratio

GeoCongress 2012
 GeoCongress 2012 © ASCE 2012 926

Correlations between Strengths under Normal and Elevated Temperature

Curing Conditions

The 28-day UCS or 7-day UCS under normal curing condition is found to be strongly
correlated to the early (30hr) accelerated UCS, as presented in Figure 5. Using the
correlations shown in Figure 5, the later-age strength can be predicted at a very early
Downloaded from ascelibrary.org by Sardar Vallabhbhai Natl Inst Of Tech (svnit) on 07/27/16. Copyright ASCE. For personal use only; all rights reserved.

curing age (30hr in this study) based on the accelerated UCS. Such early prediction is
a significant improvement to the current SDF quality control, which requires a
waiting time of at least 7 days for the SDF mixture to gain enough strength for proper
strength testing.
It is worth noting that the correlation between the 7-day UCS and the
accelerated UCS seems to be little affected by the two different UMC samples used,
although the treated UMC type-B generally produces higher strength than treated
UMC type-A with the same mix design under either curing condition, which can be
observed in Figure 3 and Figure 4 respectively. Alternately, the effect of soil type can
be accounted for using the normalization approach proposed by Tan et al. (2002). As
shown in Figure 6, the UCS obtained under either curing condition can be normalized
by the 28-day UCS of the mixture with water/cement ratio = 12.95 and soil/cement
ratio = 7. Again from Figure 6, it can be inferred that the strength development
mechanisms underlain both curing conditions are comparable, as the correlations
between the normalized strengths are consistent with each other.

Figure 5. Later-age strength (7- Figure 6. Relationships between

day or 28-day) versus Accelerated normalized strengths under normal room
Strength temperature curing and elevated
temperature curing conditions

Effect of Water/Cement Ratio on Small-Strain Shear Stiffness

Similar to that discussed for UCS, the small-strain shear modulus Gmax obtained from
the bender element test also appears to be controlled primarily by the mixture

GeoCongress 2012
 GeoCongress 2012 © ASCE 2012 927

water/cement ratio under either curing condition, as shown in Figure 7 and Figure 8
respectively. It can also be observed that the treated UMC type-B shows higher Gmax
than the treated UMC type-A at the same water/cement ratio.
Downloaded from ascelibrary.org by Sardar Vallabhbhai Natl Inst Of Tech (svnit) on 07/27/16. Copyright ASCE. For personal use only; all rights reserved.

Figure 7. 7-day Gmax versus Figure 8. Accelerated Gmax versus

Water/cement ratio Water/cement ratio

Figure 9. 7-day Gmax versus Figure 10. 7-day UCS versus

Accelerated Gmax Accelerated Gmax

Stiffness-to-Stiffness and Strength-to-Stiffness Correlations between Normal

Curing and Elevated Temperature Curing Conditions

The stiffness-to-stiffness correlation can be established between the normal curing

and elevated temperature curing conditions, as shown in Figure 9. This correlation is
similar to the strength-to-strength correlation presented in Figure 5, as both the
strength and small-strain shear stiffness are controlled primarily by the water/cement

GeoCongress 2012
 GeoCongress 2012 © ASCE 2012 928

ratio. Due to the same reason, the normal 7-day UCS can be predicted using the
early-age (30hr) accelerated Gmax attained under elevated temperature curing
condition based on the correlation presented in Figure 10.

CONCLUSIONS
Downloaded from ascelibrary.org by Sardar Vallabhbhai Natl Inst Of Tech (svnit) on 07/27/16. Copyright ASCE. For personal use only; all rights reserved.

The following conclusions can be made from this study:

(1) The strength of the SDF mixture is primarily controlled by its water/cement
ratio under either the normal room temperature or the elevated temperature curing
condition. This observation suggests that the strength development mechanisms under
the both conditions are similar to each other.
(2) The use of elevated curing temperature accelerates the SDF early strength
gain and enables early predictions based on the correlations between the later-age
strengths and the early accelerated strengths, which may significantly improve the
current SDF quality control.
(3) Two types of UMC were used in this study. The strengths of the treated UMC,
when normalized, show consistent correlations under the both curing conditions. This
observation suggests that the strength development mechanisms under the both curing
conditions are similar to each other.
(4) Similar to strength, the small-strain shear stiffness is also found be primarily
controlled by the SDF water/cement ratio.
(5) The bender element test demonstrated its effectiveness as an alternative
approach to the in-situ SDF quality control by correlating the later-age strength with
the early accelerated small-strain shear stiffness.

ACKNOWLEDGEMENT

The work presented in this paper is part of the research project supported by the
Building and Construction Authority of Singapore under the “MND Research Fund
for the Built Environment”. The financial support is greatly appreciated. The authors
would also like to acknowledge the invaluable assistance offered by the Housing and
Development Board of Singapore and Surbana Corporation Pte Ltd.

REFERENCES

ACI. (1978). SP-56: Acceleraed Strength Testing, American Concrete Institute.

ASTM. (2003). "C684-99 Standard test method for making, acclerated curing and
testing concrete compression test specimens."
Boutouil, M., and Levacher, D. (2005). "Effect of high initial water content on
cement-based sludge solidification." Ground Improvement, 9(4), 169-174.
BSI. (1983). "1881-112: Testing concrete - Methods of accelerated curing of test
cubes." British Standards Institution.
Horpibulsuk, S., Miura, N., and Nagarj, T. S. (2003). "Assessment of strength
development in cement-admixed high water content clays with Abram's law
as basis." Geotechnique, 53, No.4, 439-444.

GeoCongress 2012
 GeoCongress 2012 © ASCE 2012 929

Kitazume, M., and Satoh, T. (2003). "Development of a pneumatic flow mixing

method and its application to Central Japan International Airport
construction." Ground Improvement, 7(3), 139-148.
Kitazume, M., and Satoh, T. (2005). "Quality control in Central Japan International
Airport construction." Ground Improvement, 9(2), 59-66.
Kristensen, L., and Hansen, T. C. (1994). "Cracks in concrete core due to fire or
Downloaded from ascelibrary.org by Sardar Vallabhbhai Natl Inst Of Tech (svnit) on 07/27/16. Copyright ASCE. For personal use only; all rights reserved.

thermal heating shock." ACI Materials Journal, 91(5), 453-458.

Liu, S. Y., Du, Y. J., Han, L. H., and Gu, M. F. (2008). "Experimental Study on the
electrical resisitivity of soil cement admixtures." Environmental Geology, 54
(6), 1227-1233.
Miura, N., Horpibulsuk, S., and Nagaraj, T. S. (2001). "Engineering behavior of
cement stabilized clay at high water content." Soils and Foundations, 41(5),
33-45.
Morohoshi, K., Yoshinaga, K., Miyata, M., Sasaki, I., Saitoh, H., Tokoro, M.,
Fukuda, N., Fujii, T., Yamada, K., and Ishikawa, M. (2010). "Design and
long-term monitoring of tokyo international airport extension project
constructed on super-soft ground." Geotechnical and Geological Engineering,
28(3), 223-232.
Patch, O. G. (1933). "An 8-Hour Accelerated Strength Test for Field Concrete
Control." American Concrete Institute Journal, 29, 318-324.
Porbaha, A., Hanzawa, H., and Shima, H. (1999). "Technology of air-transported
stabilized dredged fill. Part I: Pilot Study." Ground Improvement, 3, 49-58.
Rahman, M., and Taiyab, M. (2008). "A short-cut method for predicting strength of
cement treated soft Bangladesh clays and the alteration of other engineering
parameters." Geotechnical Engineering, 39(3), 159-167.
Sakamoto, A. (1998). "Cement and soft mud mixing technique using compressed air-
mixture pipeline: Efficient solidification at a disposal site." Terra et Aqua(73),
11-22.
Tan, T. S., Goh, T. L., and Yong, K. Y. (2002). "Properties of Singapore marine clays
improved by cement mixing." ASTM Geotechnical Testing Journal, 25(4),
422-433.
Viggiani, G., and Atkinson, J. H. (1995). "Interpretation of bender element tests."
Geotechnique, 45(1), 149-154.
Yamashita, S., Kawaguchi, T., Nakata, Y., Mikamt, T., Fujiwara, T., and Shibuya, S.
(2009). "Interpretation of international parallel test on the measurement of G
max using bender elements." Soils and Foundations, 49(4), 631-650.

GeoCongress 2012