The Japanese Geotechnical Society

Soils and Foundations

Soils and Foundations 2014;54(3):488–501

http://d0038-0

nCorE-m

ceeleefh1(PrePeer

x.doi.org/1806/& 201

respondinail addre@nus.edusently witreview un

www.sciencedirect.comjournal homepage: www.elsevier.com/locate/sandf

Yielding of cement-treated marine clay

HuaWen Xiaoa,n, Fook Hou Leea, Kheng Ghee Chinb,1

aDepartment of Civil and Environmental Engineering, National University of Singapore, Singapore 117576, SingaporebDepartment of Civil Engineering, National University of Singapore, Singapore 117576, Singapore

Received 8 April 2013; received in revised form 7 November 2013; accepted 30 January 2014Available online 9 May 2014

Abstract

This paper presents the findings of an experimental study on the primary yielding and post-yield behavior of cement-treated Singapore marineclay. The study was conducted using unconfined compression tests and triaxial tests. The results show that all the primary yield loci for thecement-treated marine clay have a consistent shape regardless of the mix ratio, curing stress or curing period. Three relationships are proposed fordetermining the size of the primary yield locus. The first two involve the direct determination of the isotropic primary yield stress, whereas thethird makes use of the unconfined compressive strength. The first two relations are valid only for 7-day specimens. The third appears to haveslightly larger scatter, but it is also applicable over a wider range of curing period and curing stress. Post-yield, over-consolidated samples wereobtained by compressing specimens isotropically under effective stress levels higher than their isotropic primary yield stress and then allowingthem to swell back to a lower effective confining stress prior to shearing. The normalized yield loci of these pre-yielded samples show a“collapse” from steep arches to more-rounded ellipses, while the yield loci expand with isotropic pre-compression pressure.& 2014 The Japanese Geotechnical Society. Production and hosting by Elsevier B.V. All rights reserved.

Keywords: Cement-treated clay; Primary yielding; Yield locus; Unconfined compressive strength; Triaxial test; Tension cut-off

1. Introduction

Cement-admixtures are commonly used for improving softfine-grained soils. The strength and failure envelopes ofcement-treated soft clay, as well as the factors affecting them,have been studied extensively (e.g., Kawasaki et al., 1981;

0.1016/j.sandf.2014.04.0214 The Japanese Geotechnical Society. Production and hosting by

g author. Tel.: þ65 94390766; fax: þ65 67761002.sses: ceexh@nus.edu.sg (H.W. Xiao),.sg (F.H. Lee), kheng.chin@aecom.com (K.G. Chin).h AECOM).der responsibility of The Japanese Geotechnical Society.

Gallavresi, 1992; Yoshizawa et al., 1996; Yu et al., 1998;Miura et al., 2001; Horpibulsuk et al., 2003, 2006, 2011; Leeet al., 2005; Consoli et al., 2006; Namikawa and Koseki, 2006;Kasama et al., 2007; Kongsukprasert et al., 2007; Chiu et al.,2008; Ezaoui et al., 2010; Rabbi et al., 2011; Seng and Tanaka,2011). Other studies have also examined the isotropic com-pression and the drained and undrained shear behaviour ofcement-treated soils (e.g., Hirai et al., 1989; Matsuoka andSun, 1995; Uddin et al., 1997; Kasama et al., 2000, 2006;Rotta et al., 2003; Bergado et al., 2006; Namikawa and Mihira,2007; Taheri et al., 2012). More recently, the strengthand deformation characteristics of soils cemented withother stablization materials have been investigated (e.g.,Horpibulsuk et al., 2013; Yasuhara et al., 2012; Kamei et al.,2013; Vichan and Rachan, 2013). The physico-chemical andmicro-structural aspects of cement-treated Singapore marine

Elsevier B.V. All rights reserved.

Table 1Summary of specimen mix ratios and curing conditions in this study.

Mix ratio(S–C–W)

Soil–cement ratio(S:C)

Water–cement ratio(W:C)

Water content Cw

(%)Curing load p0cur(kPa)

Curing time t(days)

Tests in this study

10–1–11 10:1 11:1 100 0–350 7–28 UCTn, ICT†, CIU1;‡, CID2;§

20–3–23 20:3 23:3 100 0–250 7–210 UCT, ICT, CIU, CID, SP3;║

5–1–6 5:1 6:1 100 0–250 7–180 UCT, ICT, CIU, CID, SP10–3–13 10:3 13:3 100 0 7–28 UCT, ICT, CIU, CID, SP10–3–17.3 10:3 17.3:3 133 0–100 7–28 UCT, ICT10–3–19.5 10:3 19.5:3 150 0–100 7–28 UCT, ICT2–1–3 2:1 3:1 100 0 7–28 UCT, ICT, CIU, CID, SP2–1–4 2:1 4:1 133 0–250 7–180 UCT, ICT, CIU, CID, SP2–1–4.5 2:1 4.5:1 150 0–100 7–100 UCT, ICT2–1–5 2:1 5:1 167 0–200 7–28 UCT, ICT, CIU, CID2–1–5.5 2:1 5.5:1 183 0 7–28 UCT, ICT, CIU, CID1.3–1–3.06 1.3:1 3.06:1 133 0–100 7–28 UCT, ICT1.3–1–3.45 1.3:1 3.45:1 150 0–100 7–28 UCT, ICT1.3–1–3.5 1.3:1 3.5:1 152 0–100 7–28 UCT, ICT, CIU, CID1–1–2 1:1 2:1 100 0–100 7–28 UCT, ICT, CIU, CID1–1–2.66 1:1 2.66:1 133 0–100 7–28 UCT, ICT1–1–3 1:1 3:1 150 0–100 7–90 UCT, ICT10–1–7.9 10:1 7.9:1 72 0–100 7–28 UCT, ICT6–1–5 6:1 5:1 71 0–100 7–28 UCT, ICT4–1–3.6 4:1 3.6:1 72 0–100 7–28 UCT, ICT

nUnconfined compressive strength test.†Isotropic compression test.‡Isotropic consolidated undrained compression test.§Isotropic consolidated drained compression test.║Constant stress ratio test.

H.W. Xiao et al. / Soils and Foundations 54 (2014) 488–501 489

clay (e.g., Chew et al., 2004; Chin, 2006; Kamruzzaman et al.,2009) have also been studied. However, their results arequalitative and restricted to a few mix ratios, mostly withcement contents of about 20% or lower.

Although a few constitutive models for cement-treated clayhave been proposed, their yield surfaces were assumed to be thesame as those for natural soils (e.g., Lee et al., 2004; Suebsuket al., 2010; Horpibulsuk et al., 2010). This may be attributableto the dearth of data on yielding and post-yield behaviour ofcement-treated clay, particularly for admixtures with highcement contents. Theoretically, yielding is readily defined asthe onset of irrecoverable or plastic strain. However, this is oftendifficult to discern experimentally. Hence, yielding is usuallyidentified by a discontinuity in the stress–strain behavior(Vaughan, 1988; Maccarini, 1987; Bressani, 1990; Jardineet al., 1991; Jardine, 1992; Smith et al., 1992; Malandraki andToll, 1996; Leroueil and Vaughan, 1990; Bergado et al., 2006)or an abrupt decrease in stiffness (e.g., Mitchell, 1970; Wongand Mitchell, 1975; Callisto and Calabresi, 1998; Coop andAtkinson, 1993; Rotta et al., 2003; Jongpradist et al., 2011)under monotonic stress changes. This discontinuity or abruptdecrease is more significant for structured soil, and it is taken tosignify the onset of the loss of structure in the soil. This point isknown as primary yield. As the structure is due to cementationfor cement-treated soil, the primary yield indicates the beginningof the loss of bonding in the soil.

There are different ways to identify and to determineprimary yielding. For example, the primary yield is taken to

occur at the point at which the compression curve starts todeviate from the initial linear behaviour (Rotta et al., 2003;Coop and Atkinson, 1993). Cotecchia and Chandler (2000)defined an alternative ‘gross yield point’ as the point oftangency between the compression curve and a line drawnparallel to the intrinsic compression line (Burland, 1990), atwhich the ratio of the stress on the compressive curve to thaton the intrinsic compression line is maximum.Loss of structure (or bonding in cement-treated soil) after

primary yielding (Leroueil and Vaughan, 1990) is progressivewith additional straining. This is termed herein as post-yieldbehavior. For cement-treated soil, this post-yield behaviorinvolves a gradual loss of bonding with strain after primaryyielding. This paper presents the results of tests on the primaryyield and the post-yield behaviour of cement-treated Singaporemarine clay, based on triaxial tests conducted over a wide rangeof mix ratios and different curing conditions. The results showthat, regardless of mix ratios and conditions, the primary yieldloci have a generic shape. The size of the yield locus is shown tobe well-correlated to the unconfined compressive strength,which in turn can be correlated to the mix ratios and conditions.

2. Experiment investigation

2.1. Materials

Specimens were prepared from Singapore Upper marineclay and Ordinary Portland Cement (OPC). The marine clay

Fig. 1. Stress paths and fitted primary yield locus for cement-admixed marine clay specimen with mix proportion 10–3–13 (cement content 30%, water content100%, cured under atmospheric pressure for 7 days), CIU and CID represent consolidated undrained and drained tests, respectively, and the numeric denotes theeffective confining stress.

H.W. Xiao et al. / Soils and Foundations 54 (2014) 488–501490

was collected from a depth of 4 to 5 m below the seabed offSingapore's north eastern coast. Its main properties are asfollows: in-situ water content 70%, liquid limit 88%, plasticlimit, 38%, and sensitivity 1.5–6. Its properties and mineralogyare similar to those of the clay studied by Chew et al. (2004).Following Lee et al. (2005), the mix ratio is defined as themass of soil (S): mass of cement (C): mass of water (W) in themix, hereafter expressed in the form S–C–W. The watercontent is defined as the ratio of the mass of water in theresulting mix to the mass of dry soil and cement, while thecement content, Aw, is defined as the ratio of the mass ofcement solids to the mass of soil solids. The tested mix ratiosspan roughly over the range used in deep mixing and jet-grouting studies involving soft clay (Lee et al., 2005) (seeTable 1).

2.2. Specimen preparation, curing, and testing

The sample preparation procedure follows that used by Chinet al. (2004). The natural marine clay was first mixed with theprescribed amount of water to achieve a moisture content of100% and then remoulded. Cement slurry with the water–cement ratio needed to achieve the desired mix ratio was thenadded to the marine clay in a Hobart Mixer and mixed at arotational speed of 125 rpm for around 10 min.

The mixture was then placed into a 50 mm (diameter) by100 mm (height) cylindrical polyvinyl chloride (PVC) split-mould. No compaction was applied during the placement.Samples which were cured without loading were submerged indistilled water within their split-moulds. For samples whichwere cured under isotropic pressure, hereafter termed load-curing, the inner wall of the split-mould was pre-lined with arubber membrane, within which the mixture was placed. Theassembly was then placed on the platen of a triaxial pedestaland the split-mould was removed before applying isotropicpressure. During load-curing, drainage was permitted via bothends of the specimen. The elapsed time between the comple-tion of mixing and the application of curing stress was kept toless than 20 min, so as to minimize structure formation withinthe treated clay.All specimens were trimmed to 38 mm (diameter) by 76 mm

(height) for testing. The test procedures for the unconfinedcompression (UCT), isotropic consolidation (ICT), and con-solidated undrained/drained (CIU/CID) tests followed thoseprescribed in ISO/TS 17892 (2004). The strain-rates alsofollowed those in ISO/TS 17892 (2004). For UCT, therecommended ram velocity was 0.5 to 2 mm/min, correspond-ing to an axial strain rate of 0.66–2.63%/min. In this study, astrain rate of 1.32% (1 mm/min) was used for the UCT tests.For the CIU tests, pore pressure equilibration guidelines led to

Fig. 2. Stress paths for tests on cement-treated marine clay. Prefix ‘a’ denotesspecimens with 7 days of curing under atmospheric pressure. Prefix ‘b’ denotesspecimens with curing period longer than 7 days or under isotropic pressurep0cur during 7 days of curing. Prefix ‘c’ denotes specimens pre-compressed to apressure p00 higher than the isotropic primary yield stress p0py. Suffixes UCT,CIU, CID, ISO, and SP denote unconfined compression test, consolidatedundrained test, drained test, isotropic compression test, and constant stress ratiotest, respectively. Points labeled “YP” are yield point.

H.W. Xiao et al. / Soils and Foundations 54 (2014) 488–501 491

a strain rate of 0.0132%/min. Preliminary trials over a strainrate of 0.0132%/min to 0.197%/min showed virtually nodifference in specimen behaviour, thereby indicating that thestrain rate effect for the cement-treated marine clay is insig-nificant. For the CID tests, a slower rate was used to ensure anegligible change in pore pressure during shearing. From theresults of the preliminary tests, the strain rate used for CID wasfrom 0.0066%/min to 0.004%/min, the lower rate being usedfor tests with effective confining stress higher than 1000 kPa.Once again, trial tests over this range showed a negligibledifference, indicating that the strain rate effect is unlikely to besignificant.

3. Primary yielding and yield locus

3.1. Identification of yield point

Fig. 1 shows the results of isotropic compression, constantstress ratio, undrained, and drained triaxial tests for sampleswith a mix ratio of 10–3–13. Yielding is often identifiedexperimentally by an abrupt decrease in stiffness (e.g.,Mitchell, 1970; Wong and Mitchell, 1975; Callisto andCalabresi, 1998; Coop and Atkinson, 1993; Rotta et al.,2003). In this study, yielding is identified through volumetriccompression curves (Mitchell, 1970) and is taken to occur atthe point at which the compression curve starts to deviate fromthe initial linear behaviour; this is Rotta’s et al. (2003) primaryyield point (YP). The determination of the YP is shown inFig. 1, which is done by best fitting a straight line with theinitial part of the compression curve on the compression spaceby trial and error. Once the best fitting is done (the relativeerror between the fitting line and the experiment data is lessthan 0.01%), the first point to deviate from the straight line(linear behavior) is identified as the primary yield point.Numerically, the point with a relative error larger than0.01% is considered as the first point on the experiment curveto deviate from the straight line.

The triaxial specimens in Fig. 1 were consolidated to meaneffective stress levels less than the isotropic primary yieldstress (IPYS), that is, the primary yield stress under isotropiccompression; these specimens will hereafter be termed pre-yield specimens. As Fig. 1 shows, yielding in the drainedspecimens is much more readily identified from the compres-sion curves than from the stress–strain curves, which show amore gradual change in stiffness. All the drained specimenstested show post-yield hardening, accompanied by largevolumetric compression. Using the results from drained tests,self-consistent primary yield loci can be obtained based onRotta’s et al. (2003) primary yield definition. The stress pathsof the undrained tests conducted at low confining pressuresalso appear to converge towards the intersection of the yieldlocus and the tension cut-off line.

The behaviour of the undrained specimens depends upon theconfining pressure. Under low confining pressures, theundrained triaxial stress paths start off trending roughlyvertical, until they reach the tension cut-off line. At higherconsolidation pressures, the undrained stress paths veer

leftwards, indicating the generation of positive excess porepressure, until they reach the tension cut-off line. The peakstrength is often reached along the tension cut-off line. This issimilar to reported observations for cement-treated clays(e.g., Tatsuoka and Kobayashi, 1983; Uddin et al., 1997;Kamruzzaman et al., 2009; Åhnberg, 2007).The tension cut-off line corresponds to a state in which the

effective radial stress is zero. Once this state is reached, thepore pressure is equal to the confining stress and the rubbermembrane detaches from the specimen. Consequently, waterdrains into the intervening space and an undrained state cannotbe maintained. Moreover, the fact that the effective stress pathfollows that of the drained tests after reaching the tension cut-off suggests that the conditions of the test have changed fromundrained to drained. Yielding is not readily identified sincethe compression curve is a horizontal straight line and thestress–strain curve does not show any abrupt change ingradient. In summary, all the stress paths employed in thispaper were summarized in Fig. 2, which includes primaryyielding and post-yielding.

3.2. Primary yield locus

Fig. 3a–d show the primary yield loci from drained tests andthe stress paths from undrained tests, for different mix ratios,

Fig. 3. Undrained stress paths and primary yield loci for cement-admixed marine clay specimens for different mix ratios, curing periods, and curing stresses. (a)Water content 100%, after 7 days curing without loading. (b) Cement content 50%, after 7 days curing without loading. (c) Mix ratio 2–1–4 (cement content 50%and water content 133%) after 7 days curing under different curing stresses. (d) Mix ratio 5–1–6 after different curing periods without loading.

H.W. Xiao et al. / Soils and Foundations 54 (2014) 488–501492

curing stresses, and curing periods. All the yield loci from thedrained tests have qualitatively similar shapes. The undrainedstress paths trend along the tension cut-off line to approach,but not exceed, their respective yield loci. Most of theundrained samples reached peak strength just before theirstress paths reached their yield loci, indicating that the primaryyield locus intersects the tension cut-off line at or just abovethe peak points of the undrained stress paths.

Fig. 4 shows the yield loci from a large number of drainedtests, with mean effective and deviator stresses p0 and qnormalized by their respective IPYS, p0py. All of the yield locican be fitted by the relation

q

p0py

!2

¼ N2 p0

p0py

!� p0

p0py

!28<:

9=; ð1Þ

with N¼3.65. Hence, if the IPYS is known, the yield locuscan be determined. In fact, for remoulded cement-treatedmarine clay, Xiao and Lee (2010) reported a critical statefriction coefficient M of about 1.5 to 1.7, which, althoughmuch higher than that of untreated marine clay, is still muchless than 3.65.

3.3. Relation of IPYS to post-curing void ratio

The primary yield locus can be completely determined if aparameter which characterizes its size, such as the IPYS, canbe evaluated. Rotta et al. (2003) showed that the differencebetween IPYS p0py and curing stress p0cur, termed hereafter asthe incremental IPYS p0pyΔ, can be related to the post-curingvoid ratio ecur via a log-linear relationship. Post-curing void

Fig. 4. Primary yield loci normalized by isotropic primary yield stress forcement-admixed marine clay specimens with different mix ratios, curingstresses, and periods (fitted using Eq. (1) with N¼3.65).

Fig. 5. Effect of post curing void ratio and ratio of water content to the cementcontent on IPYS for 7-day cement-admixed marine clay specimens. (a)Relationship between incremental isotropic primary yield stress and post-curing void ratio. and (b) Relationship between A1 or B1 and ratio of watercontent to cement content.

H.W. Xiao et al. / Soils and Foundations 54 (2014) 488–501 493

ratio ecur is the volume ratio of void (water) to solid (unreactedsoil particles and cementitious products) after curing.

The use of the post-curing void ratio, rather than the voidratio at the point of mixing, enables avoidance of the issue ofwhether the cement is solid or pore fluid, because thecomponents of the cement which dissolve and react wouldhave already done so. The post-curing solid phase consists ofunreacted soil minerals and cementitious products. Since thespecimens were submerged in water during curing, it wasassumed that they were fully saturated and that the post-curingvoid ratio is the product of the water content and the specificgravity of the post-curing solid phase. The inferred post-curingvoid ratio was also cross-checked using the bulk density of thepost-curing soil specimen. The void ratios inferred from thewater content and the bulk density were found to be very close.The specific gravity is measured in a standard manner, using aspecific gravity bottle. Since the solid phase may consist ofmultiple products, e.g., unreacted soil solid and cementitiousproducts, the measured specific gravity is a representativevalue, rather than a true value of any of the products. This issimilar to natural clays which often consist of different soilminerals.

As Fig. 5 shows, for samples which were subjected to 7days of curing,

ecur ¼ A1 lnðp0pyΔ=paÞþB1 ð2Þ

where pa¼1 kPa is a reference incremental pressure, andB1¼ecur0 is the void ratio corresponding to the referenceincremental pressure pa.

Rotta et al. (2003) used the same water content in all theirsamples, hence their parameters A1 and B1 are functions ofcement content only. However, water content also affects thepost-curing void ratio. Miura et al. (2001) noted that forcement-stabilized clay at high water contents, the ratio of the

water content to the cement content significantly influences thestrength and deformation behaviour. In Fig. 5a, series 2–1–4and 2–1–5 have a cement content of 50%, but different watercontents. As can be seen, their post-curing void ratios lie ondifferent lines. As Fig. 5b shows, coefficients A1 and B1 inEq. (2) can be related to the ratio of the water content to thecement content via the following relationships:

A1 ¼ �0:566�0:461ðCw=AwÞ�0:8919 ð3Þ

B1 ¼ 9:1372ðCw=AwÞ0:2198 ð4ÞSubstituting Eqs. (3) and (4) into Eq. (2) leads to

p0pyΔ ¼ p0py�p0cur ¼ expecur�9:1372ðCw=AwÞ�0:2198

�0:566�0:461ðCw=AwÞ�0:8919

� �ð5Þ

Eqs. (3) and (4) differ significantly from Rotta’s et al. (2003)corresponding relations. This can be attributed to the fact thatRotta’s et al. (2003) study did not take into account the effectof the water content and that it was conducted on a lowplasticity soil comprising mainly sand and silt, whereas thetests herein are conducted on medium-high plasticitymarine clay.For specimens which were not load-cured, the post-curing

void ratio may also be dependent upon the cement content andthe water content. As Fig. 6a and b show, for the 7-dayspecimens cured without load, the IPYS increases approxi-mately linearly with the cement content, but the rate of increase

cement content Aw

0

2

4

6

8

10

12

p'py

/Pap

Cw=72%Cw=100%Cw=133.3%Cw=150%

2

2

21.6420.592

AB

==

2

2

10.520.374

AB

==

2

2

4.8750.113

AB

==

2

2

3.4460.0543

AB

== −

' / 2 2py ap w= +

0 0.2 0.4 0.6 0.8 1 1.2

0.6 0.8 1 1.2 1.4 1.6water content Cw

0

5

10

15

20

25

A2

-0.2

0

0.2

0.4

0.6

B2

( 2.504)2 9.865 wA C −=

2 wB C

p p A A B

0.820 1.189= − +

Fig. 6. Effect of cement content and water content on IPYS for 7-dayspecimens cured without loading. (a) Variation in IPYS with cement andwater contents. and (b) Relationship between A2 or B2 and water content.

q1 (kPa)

0

1000

2000

3000

q u +

1.79

p' c

ur o

r q p

k +1

.79

p'cu

r (kP

a) qu (curing stress)

qpk (curing stress)qu or qpk=q1

500 1000 1500 2000 2500

500 1000 1500 2000 2500

q1 (kPa)

0

500

1000

1500

2000

2500

q u o

r qpk

(kP

a)

qu (no curing stress)qu (curing stress)qpk (no curing stress)qpk (curing stress)qu or qpk=q1

Fig. 7. Correlation between peak undrained strength and unconfined compres-sive strength with q1 from Eq. (7) for cement-admixed marine clay specimenswith same mix ratios. (a) Measured peak undrained strength qpk andunconfined compressive strength qu plotted against q1 from Eq. (7). and (b)Measured peak undrained strength qpk and unconfined compressive strength quplus 1.79 times curing stress plotted against q1 from Eq. 7.

H.W. Xiao et al. / Soils and Foundations 54 (2014) 488–501494

decreases with the inverse power of the water content, while theintercept decreases linearly with the water content. This suggeststhe relation

p0pypap

¼ c1CdwAwþe1Cwþ f ð6Þ

where pap¼101.325 kPa (1 atm), c1¼9.865, d¼�2.504, ande1¼�0.820, f¼1.189.

3.4. Relation between IPYS, undrained peak strength, andunconfined compressive strength

Eqs. (5) and (6) apply only to the 7-day specimens. Thissection examines the possibility of using other parameterswhich are commonly measured in laboratory tests as sizeparameters of the yield loci. From Eq. (1), the point ofintersection of the yield locus with the tension cut-off(p01, q1) (Fig. 4) can be related to the IPYS via

ðp01; q1Þ ¼ ð0:597p0py; 1:79p0pyÞ ð7ÞFig. 7a shows the peak strength from consolidated

undrained tests, qpk, hereafter termed peak undrained strength,and unconfined compressive strength, qu, plotted against q1from Eq. (7), for specimens with the same mix ratios. Forspecimens that were not load-cured, the peak undrainedstrength is approximately equal to or falls just short of q1.This is consistent with the earlier observation that theundrained stress path trends along the tension cut-off, withthe peak stress often falling just short of the yield locus. Theunconfined compressive strength appears to lie even closer toq1, although the difference is small. Hence, for specimens thatwere not load-cured, the unconfined compressive strength andpeak undrained strength are good estimators of q1. In otherwords,

p0py � 0:559qu ¼ 0:559qpk ð8ÞFor specimens which were load-cured, the unconfined

compressive strength and the undrained peak strength fallbelow the value of q1, in many cases, quite substantially. Onthe other hand, as shown in Fig. 7b, if the unconfinedcompressive strength is increased by 1.79 times the curingstress, then the resulting quantity is a reasonably goodestimator of q1, that is,

q1 ¼ 1:79p0py � quþ1:79p0cur ð9aÞor

p0py�p0cur � 0:559qu ð9bÞHowever, this adjustment does not work for the peak

undrained strength which remains rather scattered.The close correlation between q1 and the unconfined

compressive strength for specimens without curing stressmotivates the possibility that unconfined compressive testspecimens might also have reached their peak strength at theintersection point between the tension cut-off and the primaryyield locus. This is a plausible notion since unconfinedcompression samples are exposed and any pore pressure will

500 1000 1500 2000 2500predicted qu (kPa)

0

500

1000

1500

2000

2500m

easu

red

q u (k

Pa)

7 days28 days1:1 line

Fig. 8. Comparison between measured and predicted qu by Eq. (10) (withq0¼13,000 kPa and 20,000 kPa for 7- and 28-day strengths, m¼0.28 andn¼2.93) for cement-admixed marine clay specimens with different curingperiods and zero curing stress.

40 80 120 160 200curing time t (days)

0

10000

20000

30000

40000

50000

q 0 (k

Pa)

q0 r =0.8194q0a r =0.9903standard deviation

Generalized Hyperbolic fit

1σ

1σ

Fig. 9. Parameter q0 versus curing time for cement-admixed marine clayspecimens after different curing periods with no curing stress. q0a denotes theaverage q0 for a given curing time. r denotes the coefficient of correlationbetween the data and the fitting. s denotes the standard deviation around theaverage value for each curing time.

H.W. Xiao et al. / Soils and Foundations 54 (2014) 488–501 495

dissipate towards the sides, if sufficient time is given, causingthe effective stress path to trend along the tension cut-off. Fortriaxial specimens that are isotropically consolidated withradial drainage to effective stress levels below the IPYS, themeasured t90 consolidation time was approximately 30 s, thisbeing consistent with the high pre-yield stiffness of thespecimens. The typical time to failure in an unconfinedcompression test is about 1.5 min. During the test, water wasobserved to be seeping out through the sides of the specimens.All these suggest that significant pore pressure dissipationoccurred during the unconfined compression tests. Hence, theeffective stress paths of unconfined compression samples canbe reasonably expected to trend along the tension cut-off.

When the cement content approaches zero, the unconfinedcompressive strength will become very small, and Eq. (9b)indicates that the IPYS will converge towards the curing stress.Physically, this is reasonable since both are equal to the pre-consolidation pressure. Hence, Eq. (9b) remains intuitivelysound when the cement content approaches zero.

3.5. Effect of mix proportion, curing stress, and curing periodon unconfined compressive strength

The use of unconfined compressive strength as an estimatorof the size of the yield locus opens up the possibility ofestimating the primary yield locus for a wide range of mixratios and curing conditions. Many correlations betweenunconfined compressive strength and mix ratios have alreadybeen proposed. For instance, Lee et al. (2005) proposed that,for cement-treated soil specimens which were not subjected toload curing,

qu ¼ q0expðmð1=AwÞÞ

ðW=CÞn ð10Þ

in which q0 is a constant and m and n are indices, Aw¼C/S.For the clay tested by Lee et al. (2005), q0¼4000 kPa and6000 kPa for 7-day and 28-day strengths, respectively, whilem¼0.62 and n¼3.

The unconfined compressive strengths obtained herein aregenerally higher than those of Lee et al. (2005), probablybecause of the different properties of the two clays, which,albeit marine clay, were taken from different parts of Singa-pore. The plasticity index of the clay used by Lee et al. (2005)was 59%, whereas that used herein has the lower plasticityindex of 50%. As Fig. 8 shows, the current data are bettermatched by q0¼13,000 kPa and 20,000 kPa for 7- and 28-daystrengths, respectively, with m¼0.28 and n¼2.93.Eq. (10) can be extended to take into account the effect of

different curing times. Fig. 9 shows the values of q0 inferredfrom measured qu using Eq. (10), against curing time. Therelatively large scatter arises because a small change inmeasured qu causes a large change in the inferred q0. None-theless, a general trend can be discerned. Furthermore, theaverage q0 corresponding to each curing period is well fitted bya “generalized” hyperbolic curve of the form

q0 ¼ q1 1� 1

1þ αtq1

� �r8<:

9=; ð11Þ

in which q1 is the long-term value for q0, α is the initial rate ofincrease in q0 with time, and r is a fitted index. Based on theaverage values for q0, initial fitting gives q1¼39,000 kPa,α¼2000 kPa/day, and r¼0.9.However, in some cases, the number of samples is very

small and the average q0 is susceptible to errors. To reducethese errors, a second stage of correlation is applied. Combin-ing Eqs. (10) and (11) gives

qu ¼ q1 1� 1

1þ αtq1

� �r8<:

9=; expðmð1=AwÞÞ

WC

� �n( )

ð12Þ

As shown in Fig. 10, a remarkably good fit between Eq. (12)and the measured unconfined compressive strength can beobtained by using q1¼40,000 kPa, α¼1300 kPa/day,r¼0.52, m¼0.3, and n¼2.92.

500 1000 1500 2000 2500measured qu (kPa)

0

500

1000

1500

2000

2500pr

edic

ted

q u (k

Pa)

y =1.0062x,R2 = 0.9093

Fig. 10. Comparison between measured and predicted qu using Eq. (12) forcement-admixed marine clay specimens with different curing periods and zerocuring stress.

50 100 150 200 250 300 350curing stress p'cur (kPa)

0

500

1000

1500

2000

2500

3000

3500

Δqu

(kP

a)

5−1−6, Δqu=1.9442p'cur2−1−4, Δqu=7.1087p'cur2−1−5, Δqu=9.492p'cur20−3−23, Δqu=2.1087p'cur10−1−11, Δqu=1.6393p'cur1−3−3.5, Δqu=17.33p'cur1−1−2.66, Δqu=32.18p'cur10−1−7.9, Δqu=23.796p'cur6−1−5, Δqu=2.9074p'cur4−1−3.6, Δqu=4.96p'cur10−3−19.5, Δqu=5.324p'cur2−1−4.5, Δqu=9.1p'cur

Fig. 11. Measured qu increment versus curing stress for cement-admixedmarine clay specimens with 7 days curing and different curing stress.

0 0.2 0.4 0.6 0.8 1cement soil ratio C/S

0

10

20

30

40

q u in

crem

ent p

er c

urin

g st

ress

Δq u

/p' c

ur

' 2/ 22( / ) 9( / )u curq p C S C SΔ = +

Fig. 12. qu increment per unit curing stress versus cement soil ratio forcement-admixed marine clay specimens with 7 days curing and different curingstress.

0 1000 2000 3000 4000measured qu (kPa)

0

1000

2000

3000

4000

20.9967 , 0.936y x R= =

pred

icte

d q

(kP

a)

u

Fig. 13. Measured qu and predicted qu using Eq. (14) for cement-admixedmarine clay specimens with 7 days curing and curing stress up to 250 kPa.

H.W. Xiao et al. / Soils and Foundations 54 (2014) 488–501496

Eq. (12) does not include the effect of curing stress. In thisstudy, specimens were only load-cured for 7 days. As Figs. 11and 12 show, the unconfined compressive strength qu increaseswith curing stress and the strength increment per unit curingstress, i.e., Δqu=p0cur, increases with the cement content, that isthe cement:soil ratio, following the relation

Δqu ¼ fC1ðAwÞ2þC2ðAwÞgp0cur ð13Þin which Δqu is the difference between qu and p0cur; C1=22 andC2=9 for the 7-day curing time. Combining Eqs. (12) and (13)leads to

qu ¼ q1 1� 1

1þ αtq1

� �r8<:

9=; expðmð1=AwÞÞ

WC

� �n( )

þfC1ðAwÞ2þC2ðAwÞgp0cur ð14Þ

As Fig. 13 shows, Eq. (14), with the fitted values of C1 andC2, agrees with the measured 7-day unconfined compressivestrength reasonably well for curing stress up to 250 kPa.Finally, combining Eqs. (9b) and (14) gives

p0py ¼ 22360 1� 1

1þð0:0325tÞ0:52�

expð0:3ð1=AwÞÞWC

� �2:92( )

þf12:30ðAwÞ2þ5:04ðAwÞþ1gp0cur ð15Þfor specimens without curing stress up to 270 days as well as7-day specimens with curing stress up to 350 kPa. As Fig. 14shows, the IPYS predicted by Eq. (15) has a slightly largererror range with respect to the measured results than the twoother two methods, but it also has the widest applicability.

3.6. Post-yield behaviour and evolution of yield locus

Post-yield behaviour was investigated by isotropically con-solidating the samples to an effective stress that is higher than

400 800 1200 1600measured p'py (kPa)

0

400

800

1200

1600

2000

pred

icte

d p'

py(k

Pa)

Equation 5Equation 6Equation 15

25%+

25%−

36%−

36%+

1:1

Fig. 14. Comparison between experimental and predicted IPYS for cement-admixed soil specimens.

H.W. Xiao et al. / Soils and Foundations 54 (2014) 488–501 497

the IPYS. These samples were then sheared at their isotropicpre-compression pressure p00 or else allowed to swell to a lowerpressure p0c before shearing. All post-yield specimens werecured for 7 days without loading. The states of these post-yieldsamples are characterized by an isotropic over-consolidationratio (IOCR), which is the ratio of the isotropic pre-compression pressure to the mean effective stress after swel-ling, that is, p00/p0c. In Fig. 15, these “over-consolidated”specimens are identified by two numbers which refer to theirisotropic pre-compression pressure and mean effective stressduring shearing.

As Fig. 15 illustrates, at lower IOCR, e.g., CID 1500-500,the post-yield samples exhibit a similar trend to the pre-yieldsamples in the drained tests, with large volumetric compres-sion occurring after an initially stiff segment in the compres-sion curve with significant post-yield hardening. At highIOCR, the drained samples, e.g., CID1500-50, show stiff aresponse up to the peak strength, which is reached shortly afteryielding. Post-peak strain softening is accompanied by dila-tion. Consistent yield loci can also be drawn through theisotropic pre-compression pressure and the yield pointsdeduced from drained tests, using the change in gradient inthe compression curve.

The undrained samples sheared under high IOCR presentstress paths that trend along the tension cut-off line until thepeak strength is reached; this is similar to the pre-yieldspecimens. At lower IOCR, the undrained specimens, e.g.,CIU1500-750 and CIU2000-1000, show stress paths that areroughly vertical at the initial stages, but subsequently turn totrend along the yield loci just before peak strength is reached.Normally consolidated specimens, viz. those with an IOCR of1, trend roughly along the yield loci from the start of shearing.This suggests that, for the undrained specimens with a lowIOCR, the yield points can be identified by the points atwhich the stress paths turn to trend along the yield loci.

This definition of yield stress would be similar to that adoptedby Airey (1993) for the undrained tests.As Fig. 15 shows, the yield loci increase in size with

isotropic pre-compression pressure. However, all normalizedyield loci (YL) change progressively from steep arches to morerounded ellipses, as seen in Fig. 16. Moreover, by decreasingthe value of N in Eq. (1) progressively, a succession of uprightelliptical yield loci can be generated; this gives a reasonable fitto the measured yield points. As N decreases, the “arch” shapeof the YL changes to a more rounded ellipse. An ellipticalyield locus with a larger value for N will be able to sustainlarger levels of deviator stress (for the same mean effectivestress) before yielding, compared to one with a smaller valuefor N. Hence, a decreasing value for N during loading impliesthat the soil skeleton is able to sustain progressively smallerlevels of deviator stress before yielding, given the sameeffective stress. This can be attributed to a loss in bonding,and the decrease in N is taken herein to be a measure of thisloss in bonding.The occurrence of irrecoverable strain in soil is generally

associated with the onset of large-scale particle slippage andre-arrangement (e.g., Jardine, 1992), which is also postulatedherein to be the main cause of the loss in bonding. Since N is ameasure of the bonding, it can be surmised that N can berelated to the plastic strain. In this study, the decrease in N wasonly investigated for isotropically pre-compressed specimens;none of the specimens were subjected to pre-shearing. For thisreason, the effect of plastic shear strain cannot be elucidated,although it is quite likely to be present. Hence, it is postulatedherein that the measured values of N can be related to plasticvolumetric strain εvp. As Fig. 17 shows, N can be reasonablyrelated to the plastic volumetric strain via

N ¼ 1:23ε�0:254vp ð16Þ

When the isotropic pre-compression pressure reaches 2500 kPa,the stress ratio corresponding to the crown of the yield locus fallsto between 1.5 and 1.9. This is similar to the range in values forremoulded cement-treated marine clay (Xiao and Lee, 2010), butis still significantly higher than the critical state friction coefficientof approximately 0.9 for the reconstituted marine clay. Hence,while some loss of structure might have occurred, the cemented-treated soil did not reach the same state as the untreated marineclay. This suggests that cement-treatment does not merely super-impose bonding on the marine clay particles; the cement actuallyreacts with the clay mineral and changes its mineralogy. Thus,untreated marine clay does not represent the intrinsic unstructuredstate for the cement-treated clay.

4. Conclusions

The findings suggest that, in spite of the wide range in mixratios and curing conditions, relatively simple bases for thebehaviour of cement-treated clays may still be established. Oneof these is the notion that the primary yield locus of cement-treated medium-high plasticity clay has a consistent shaperegardless of the mix ratio, the curing stress or the curingperiod. This allows the primary yield locus to be specified

Fig. 15. Typical post-yield stress–strain behavior of cement-admixed marine clay specimen (mix proportion 5–1–6, 20% cement content, 100% water content),cured under atmospheric pressure for 7 days, CIU, CID, and SP represents consolidated undrained, drained, and constant stress ratio tests, respectively, and numericdenotes pre-compression pressure and mean effective stress during shearing.

H.W. Xiao et al. / Soils and Foundations 54 (2014) 488–501498

using a single-size parameter. The foregoing discussionindicates that one method of evaluating the size parameter isthrough the unconfined compressive strength, a commonly andreadily measured strength parameter.

In the laboratory, the curing stress is relatively easy tocontrol since the specimen size is small and the drainagepath is short. In view of this, it is quite likely that the soilskeleton will be subjected to the full curing stress prior to thedevelopment of much of the cementation. In the field,however, excess pore pressure may take much longer todissipate. In such situations, the development of cementationmay further inhibit the compression of the soil skeleton beforea significant volume change can occur. This uncertainty hasalso been alluded to by Tatsuoka and Kobayashi (1983). Forthis reason, the curing stress in field scenarios may besignificantly lower than the geostatic stress. In such cases, itmay be conservative to assume that the curing stress is zero.With this assumption, the framework presented above wouldindicate that the size of the yield locus is directly related to theunconfined compressive strength.

The “collapse” of the normalized yield locus, whichpresumably reflects the loss of structure under isotropiccompression, may also have significant implications in con-stitutive modeling. This feature is still not widely replicated inconstitutive models for cement-treated clays. The relativelyhigh N-value of the post-yield specimens, even under com-pression pressure that far exceeds the primary yield stress, aswell as the high critical state friction coefficient of theremoulded cement-treated marine clay, suggests that untreatedmarine clay may not be a representative unstructured referencestate for the cement-treated clay.One implication of the evolution of the yield locus is the

significance of the tension cut-off. In post-yield specimens, therounded profile of the yield locus allows undrained stress pathsof specimens with low IOCR to reach it before the tension cut-off. The resulting stress–strain curve is more ductile, and post-yield strain hardening is evident. In contrast, for pre-yieldspecimens, the yield locus is much steeper and undrained stresspaths usually reach the tension cut-off before the yield locus.For such specimens, the unconfined compressive strength is a

Fig. 16. Evolution of yield locus on normalized stress space for cement-admixed marine clay specimens for different mix proportions, fitted by decreasing the valuefor N in Eq. (1). (a) Mix ratio 10–1–11. (b) Mix ratio 5–1–6. (c) Mix ratio 10–3–13. (d) Mix ratio 2–1–3.

Fig. 17. Relationship between stress ratio at the top of the yield locus (N valuein Eq. (1)) and plastic volumetric strain for cement-admixed marine clayspecimens for different mix ratios.

H.W. Xiao et al. / Soils and Foundations 54 (2014) 488–501 499

good estimate of the peak undrained shear strength. In the field,cement-treated soil masses are more likely to be in pre-yieldrather than post-yield states. Hence, the unconfined compressive

strength may be a good indicator of the peak undrained strengthof the cement-treated soil in the field.It is likely that the specific correlation values presented will

only be applicable to tropical marine clay with similarmineralogy and plasticity. However, the observed qualitativetrends are consistent with those reported previously (e.g.,Uddin et al., 1997; Kamruzzaman et al., 2009). This suggeststhat the overall framework probably encompasses widergenerality.

Acknowledgements

This research was supported by the National ResearchFoundation Singapore under its Competitive Research Pro-gramme (CRP Award no. NRF-CRP 6-2010-03) and the NUSResearch Scholarship.

H.W. Xiao et al. / Soils and Foundations 54 (2014) 488–501500

References

Åhnberg, H., 2007. On yield stresses and the influence of curing stresses onstress paths and strength measured in triaxial testing of stabilized soils.Can. Geotech. J. 44 (1), 54–66.

Airey, D.W., 1993. Triaxial testing of naturally cemented carbonate soil.J. Geotech. Eng., ASCE 119 (9), 1379–1398.

Bergado, D.T., Taechakumthorn, C., Lorenzo, G.A., Abuel-Naga, H.M., 2006.Stress and deformation behavior under anisotropic drained triaxial con-solidation of cement-treated soft Bangkok clay. Soils Found. 46 (5),629–637.

Bressani, L.A., 1990. Experimental Properties of Bonded Soils. Ph.D. Thesis.University of London.

Burland, J.B., 1990. On the compressibility and shear strength of natural clays.Geotechnique 29 (3), 329–378.

Callisto, L., Calabresi, G., 1998. Mechanical behaviour of a natural soft clay.Geotechnique 48 (4), 495–513.

Chew, S.H., Kamruzzaman, A.H.M., Lee, F.H., 2004. Physicochemical andengineering behavior of cement treated clays. J. Geotech. Geoenviron.Eng., ASCE 130 (7), 696–705.

Chin, K.G., Lee, F.H., Dasari, G.R. (2004): Effects of curing stress onmechanical properties of cement-treated soft marine clay. In: Proceedingsof the International Symposium on Engineering Practice and Performanceof Soft Deposits, Osaka, Japan, pp. 217–222.

Chin, K.G., 2006. Constitutive Behavior of Cement Treated Marine Clay.National University of Singapore, Singapore (Ph.D. Thesis).

Chiu, C.F., Zhu, W., Zhang, C.L., 2008. Yielding and shear behaviour ofcement-treated dredged materials. Eng. Geol. 103 (1–2), 1–12.

Consoli, N.C., Rotta, G.V., Prietto, P.D.M., 2006. Yielding-compressibility-strength relationship for an artificially cemented soil under stress. Geo-technique 56 (1), 69–72.

Coop, M.R., Atkinson, J.H., 1993. The mechanics of cemented carbonatesands. Geotechnique 43 (1), 53–67.

Cotecchia, F., Chandler, R.J., 2000. A general framework for the mechanicalbehaviour of clays. Geotechnique 50 (4), 431–447.

Ezaoui, A., Tatsuoka, F., Sano, Y., Iguchi, Y., Meada, Y., Sasaki, Y., Dutine,A., 2010. Ageing effect on the yielding characteristics of cement-mixedgranular materials. Soils Found. 50 (5), 705–724.

Gallavresi, F., 1992. Grouting improvement of foundation soils. In: Proceed-ings of the Grouting, Soil Improvement and Geosynthetics, ASCE, NewYork, vol. 1, pp. 1-38.

Hirai, H., Takahashi, M., Yamada, M., 1989. An elastic-plastic constitutivemodel for the behavior for improved sandy soils. Soils Found. 29 (2), 69–84.

Horpibulsuk, S., Miura, N., Nagaraj, T.S., 2003. Assessment of strengthdevelopment in cement-admixed high water content clays with Abrams lawas a basis. Geotechnique 53 (4), 439–444.

Horpibulsuk, S., Katkan, W., Sirilerdwattana, W., Rachan, R., 2006. Strengthdevelopment in cement stabilized low plasticity and coarse grained soils:laboratory and filed study. Soils Found. 46 (3), 351–366.

Horpibulsuk, S., Liu, M.D., Liyanapathirana, D.S., Suebsuk, J., 2010.Behaviour of cemented clay simulated via the theoretical framework ofstructured Cam Clay model. Comput. Geotech. 37 (1–2), 1–9.

Horpibulsuk, S., Rachan, R., Suddeepong, A., Chinkulkijniwat, A., 2011.Strength development in cement admixed Bangkok clay: laboratory andfield investigations. Soils Found. 51 (2), 239–251.

Horpibulsuk, S., Phetchuay, C., Chinkulkijniwat, A., Cholaphatsorn, A., 2013.Strength development in silty clay stabilized with calcium carbide residueand fly ash. Soils Found. 53 (4), 477–486.

Jardine, R.J., St. John, H.D., Hight, D.W., Potts, D.M. (1991): Some practicalapplications of a non-linear ground model. In: Proceedings of the 10thEuropean Conference on Soil Mechanics and Foundation Engineering,Florence, vol. 1, pp. 223–228.

Jardine, R.J., 1992. Some observations on the kinematic nature of soil stiffness.Soils Found. 32 (1), 11–24.

Jongpradist, P., Youwai, S., Manorat, P., Kongkitkul, W., Chucheepsakul, S.,2011. Influence of curing stress on one-dimensional yielding of cement-admixed Bangkok clay at high water content. Soils Found. 51 (2), 351–357.

ISO/TS17892 (2004): Geotechnical Investigation and Testing—LaboratoryTesting of Soil. International Organization for Standardization, first ed.,Part 7-9.

Kasama, K., Zen, K., Iwataki, K., 2007. High-strengthening of cement-treatedclay by mechanical dehydration. Soils Found. 47 (2), 171–184.

Kamei, T., Ahmed, A., Ugai, K., 2013. Durability of soft clay soil stabilizedwith recycled Bassanite and furnace cement mixtures. Soils Found. 53 (1),155–165.

Kamruzzaman, A.H.M., Chew, S.H., Lee, F.H., 2009. Structuration anddestructuration behavior of cement-treated Singapore marine clay.J. Geotech. Geoenviron. Eng., ASCE 135 (4), 573–589.

Kasama, K., Ochiai, H., Yasufuku, N., 2000. On the stress–strain behavior oflightly cemented clay based on an extended critical state concept. SoilsFound. 40 (5), 37–47.

Kasama, K., Zen, K., Iwataki, K., 2006. Undrained shear strength of cement-treated soil. Soils Found. 46 (2), 221–232.

Kawasaki, T., Niina, A., Saitoh, S., Suzuki, Y., Honjo, Y. (1981): Deep mixingmethod using cement hardening agent. In: Proceedings of the 10thInternational Conference on Soil Mechanics and Foundation Engineering,Stockholm, vol. 3, pp. 721–724.

Kongsukprasert, L., Tatsuoka, F., Takahashi, H., 2007. Effects of curing periodand stress conditions on the strength and deformation characteristics ofcement-mixed soil. Soils Found. 47 (3), 577–596.

Lee, F.H., Lee, Y., Chew, S.H., Yong, K.Y., 2005. Strength and modulus of marineclay-cement mixes. J. Geotech. Geoenviron. Eng., ASCE 131 (2), 178–185.

Lee, K., Chan, D., Lam, K., 2004. Constitutive model for cement treated clayin a critical state frame work. Soils Found. 44 (3), 69–77.

Leroueil, S., Vaughan, P.R., 1990. The general and congruent effects ofstructure in natural soils and weak rocks. Geotechnique 40 (3), 467–488.

Maccarini, M., 1987. Laboratory Studies of a Weakly Bonded Artificial Soil.University of London (Ph.D. Thesis).

Malandraki, V., Toll, D.G., 1996. The definition of yield for bonded materials.Geotech. Geol. Eng. 14, 67–82.

Matsuoka, H., Sun, D.A., 1995. Extension of spatially mobilized plane (SMP)to frictional and cohesive materials and its application to cemented sands.Soils Found. 35 (4), 63–72.

Mitchell, R.J., 1970. On the yielding and mechanical strength of Leda clays.Can. Geotech. J. 7 (3), 297–312.

Miura, N., Horpibulsuk, S., Nagagaj, T.S., 2001. Engineering behavior ofcement stabilized clay at high water content. Soils Found. 41 (5), 33–45.

Namikawa, T., Koseki, J., 2006. Experimental determination of softeningrelations for cement-treated sand. Soils Found. 46 (4), 491–504.

Namikawa, T., Mihira, S., 2007. Elasto-plastic model for cement-treated sand.Int. J. Numer. Anal. Methods Geomech. 31 (1), 71–107.

Rabbi, A.T.M.Z., Kuwano, J., Deng, J., Boon, T.W., 2011. Effects of curingstress and period on the mechanical properties of cement-mixed sand. SoilsFound. 51 (4), 705–724.

Rotta, G.V., Consoli, N.C., Prietto, P.D.M., Coop, M.R., Graham, J., 2003.Isotropic yielding in an artificially cemented soil cured under stress.Geotechnique 53 (5), 493–501.

Seng, S., Tanaka, H., 2011. Properties of cement-treated soils during initialcuring stages. Soils Found. 51 (5), 775–784.

Smith, P.R., Jardine, R.J., Hight, D.W., 1992. The yielding of Bothkennar clay.Geotechnique 42 (2), 257–274.

Suebsuk, J., Horpibulsuk, S., Liu, M.D., 2010. Modified structured Cam Clay:a generalized critical state model for destructured, naturally structured andartificially structured clays. Comput. Geotech. 37 (7–8), 956–968.

Taheri, A., Sasaki, Y., Tatsuoka, F., Watanabe, K., 2012. Strength anddeformation characteristics of cement-mixed gravelly soil in multiple-steptriaxial compression. Soils Found. 52 (1), 126–145.

Tatsuoka, F., Kobayashi, A. (1983): Triaxial strength characteristics of cement-treated soft clay. Proceedings of the Eighth European Conference on SoilMechanics and Foundation Engineering, vol. 8, No. 1, pp. 421–426.

Uddin, K., Balasubramaniam, A.S., Bergado, D.T., 1997. Engineeringbehavior of cement-treated Bangkok Soft Clay. Geotech. Eng 28 (1),89–119.

Vaughan, P.R. (1988): Characterising the mechanical properties of the in-situresidual soil. In: Proceedings of the Second International Conference on

H.W. Xiao et al. / Soils and Foundations 54 (2014) 488–501 501

Geomechanics in Tropical Soils, Singapore, Balkema, Rotterdam, vol. 2,pp. 469–487.

Vichan, S., Rachan, R., 2013. Chemical stabilization of soft Bangkok clayusing the blend of calcium carbide residue and biomass ash. Soils Found.53 (2), 272–281.

Wong, P.K.K., Mitchell, R.J., 1975. Yielding and plastic flow of sensitivecemented clay. Geotechnique 25 (4), 763–782.

Xiao, H.W., Lee, F.H. (2010): Stress–strain behavior of remoulded cement-treated Singapore marine clay. In: Proceedings of the InternationalSymposium on Geomechanics and Geotechnics: From Micro to Macro,Shanghai, China, vol. 1, pp. 177–182.

Yasuhara, H., Neupane, D., Hayashi, K., Okamura, M., 2012. Experiments andpredictions of physical properties of sand cemented by enzymatically-induced carbonate precipitation. Soils Found. 52 (3), 539–549.

Yoshizawa, H., Okumura, R., Hosoya, Y., Sumi, M., Yamada, T. (1996): JGSTechnical Committee Report: factors affecting the quality of treated soilduring execution of DMM. Proceedings of the Second InternationalConference on Ground Improvement Geosystems, Grouting and DeepMixing, Tokyo 2, pp. 931–937.

Yu, Y.Z., Pu, J.L., Ugai, K., 1998. A damage model for soil–cement mixture.Soils Found. 38 (3), 1–12.