Jiang N.-J. et al. (2014) Géotechnique Letters 4, 27–32, http://dx.doi.org/10.1680/geolett.13.

00079

Experimental investigation of the compressibility behaviour of

cement-solidified/stabilised zinc-contaminated kaolin clay
N.-J. JIANG*, Y.-J. DU{, S.-Y. LIU{ and J.-J. ZHU{

Heavy metal contamination has become a major environmental pollution issue in China, especially
under the process of rapid urbanisation. Zinc contaminants are reported to be especially harmful for
local ecosystems and biodiversity. In this study, a cement-based solidification/stabilisation technique
was adopted to remediate zinc-contaminated kaolin clay and a series of one-dimensional oedometer
tests was performed to investigate the compressibility of cement-stabilised zinc-contaminated soils.
Compression curves were analysed using a simulation of the logistic function, and the yield stress
(s9vp), swelling index (Cs) at the pre-yield state and compression index (Cc) at the post-yield state were
derived on the basis of the mathematical simulation. The effects of zinc concentration on s9vp, Cs and
Cc are discussed. It is found that relatively high zinc concentrations (i.e. .0?1%) significantly
influence the compression parameters and hence deteriorate the resistance to compression of the
stabilised soils. The reduction in resistance to compression at high zinc concentrations is intrinsically
attributed to the destruction of cementation structure/bonding in the stabilised soils. Physico-
chemically, this is attributed to the retardation effect of zinc contaminants on cement hydration
processes.

KEYWORDS: compressibility; contaminated material; pollution migration/control; remediation; soil stabilisation

ICE Publishing: all rights reserved

INTRODUCTION of construction on remediated lands. Most previous studies

Heavy metal contamination has become a major environ- have focused on the strength properties of S/S-treated soils
mental pollution issue in China, especially with the (Boardman et al., 2004; Voglar & Lestan, 2010; Du et al.,
tendency of rapid urbanisation (Du et al., 2012a). Major 2012b, 2013), and research on the compressibility of S/S-
heavy metals in soils that present potential risk to human treated heavy metal contaminated soils is limited, although
health and local ecosystem include, but are not limited to, the compression of chemically treated natural soils using
zinc (Zn), lead (Pb), chromium (Cr), cadmium (Cd), nickel cement or lime has been studied extensively (Tremblay
(Ni) and copper (Cu). Among these, zinc is reported to be et al., 2001; Rotta et al., 2003; Horpibulsuk et al., 2004;
especially phytotoxic for local microbial communities. It Rios et al., 2012). Indeed, based on the authors’ pre-
has been reported that zinc contaminants could decrease vious studies on the strength of stabilised heavy metal-
the ability of a local microbial community to resist repeated contaminated soils, high contaminant concentrations lead
contamination (Klimek, 2012). In addition, zinc contam- to serious destruction of the cement–soil matrix and thus
ination decreases the bacterial diversity of agricultural soils result in a loss of resistance to shear stress or compression
(Moffett et al., 2003). In soils, zinc contaminants are often pressure (Du et al., 2012b, 2013).
absorbed to various soil components such as clay minerals, The objective of this study was to investigate the
organic matters and iron and manganese (hydr)oxides compressibility of cement-stabilised zinc-contaminated soil.
(Voegelin et al., 2005). A series of one-dimensional oedometer tests was carried
The solidification/stabilisation (S/S) technique is widely out. Specifically, the impact of zinc concentration on yield
used for remediating lands contaminated with heavy stress (s’vp ), swelling index at the pre-yield state (Cs) and
metals. With S/S treatment, the leaching of heavy metals compression index at the post-yield state (Cc) are investi-
from contaminated soils can be controlled below regulatory gated. This article also offers a discussion that interprets
limits (Voglar & Lestan, 2010; Du et al., 2012a) and, the effect of zinc concentration on soil compressibility from
compared with untreated soils, the soil strength can be the perspective of fundamentally physiochemical changes.
increased to some extent (Boardman et al., 2004; Du et al.,
2012b). As a large proportion of remediated lands are
reused as bearing strata supporting civil infrastructures MATERIALS AND METHODS
(e.g. buildings and roadways), the mechanical properties of The base soil used in this study was kaolin clay; its basic
treated soils are of increasing concern from the perspective properties are shown in Table 1. Kaolin was selected due to
its uniform composition as well as its relatively well-defined
physical and chemical properties as these allow a better
Manuscript received 4 November 2013; first decision 5
December 2013; accepted 24 December 2013. fundamental understanding of the key factors controlling
Published online at www.geotechniqueletters.com on 7 compressibility. Based on ASTM D2487, this kaolin clay is
February 2014. classified as low-plasticity clay (CL). X-ray diffraction
*Institute of Geotechnical Engineering, Southeast University, analysis revealed that it is composed of kaolinite (96%) and
Nanjing, China (currently at Department of Engineering, quartz (4%). Locally produced cement (equivalent to type I
University of Cambridge, Cambridge, UK) Portland cement) was used as the sole binder and zinc
{Institute of Geotechnical Engineering, Southeast University, nitrate hydrate (Zn(NO3)2?6H2O) (chemical analytical
Nanjing, China agent) was used as the contaminant source. Five levels of

27
28 Jiang, Du, Liu and Zhu
zinc concentration (0?02, 0?1, 0?2, 0?5 and 1?0%, dry soil asymmetrical characteristic that could well distinguish
weight basis) were selected. A cement content of 8% (dry primary compression from secondary compression, it was
soil weight basis) was adopted and the initial water content found that the logistic function fits better with the
was set at about 60% (dry soil weight basis). Cement- experimental data presented in this study. Since secondary
treated clean kaolin clays with the same binder dosage and compression is not considered in this study, the symme-
initial water content were prepared as control samples. trical logistic function is accepted to just fit the primary
To prepare stabilised soil samples, a predetermined mass compression part.
of Zn(NO3)2?6H2O was thoroughly dissolved in de-aired Gregory et al. (2006) and Keller et al. (2011) indicated
distilled water and then the solution was mixed with air- that the inflection point of a sigmoidal function (i.e. m)
dried kaolin clay to achieve a slurry. A predetermined should be within the range of the measured data (i.e. lower
quantity of cement powder was then poured into the initial than the logarithm of maximum vertical effective stress). In
slurry and the binary mixture was agitated thoroughly for this study, the maximum vertical effective stress was
10 min to achieve uniformity. The homogeneous paste was 1600 kPa, therefore the maximum value of the fitted m
then transferred into oedometer rings (diameter 61?8 mm, should be less than 3?204. It was found that the value of the
height 20 mm) with two glass sheets enclosing both open fitted m for all cases in this study satisfied this criterion,
ends. Entrapped air bubbles were carefully eliminated except for the case of sample 28-8-0. Hence, the compres-
during this procedure. The preparation procedure was sion parameters (s’vp , Cs and Cc) were estimated from the
completed within 45 min to avoid initial hardening. Soil fitted curves using equation (1). Regarding sample 28-8-0, a
samples together with rings and glass sheets were then separate linear regression of points falling in the post-yield
wrapped in sealed plastic bags and cured in a controlled region was performed to obtain Cc. Then, Cc could be
environment (20uC and 95% relative humidity) for periods estimated as the modulus of slope at the inflection point of
of 7 and 28 d. After curing, the samples were soaked under the sigmoidal curve. Mathematically, this is expressed by
water and were then subjected to vacuum for 10 h to (Gregory et al., 2006)
facilitate saturation prior to the oedometer tests. This
vacuum time is generally regarded as sufficient to eliminate Cc ~bc=4 (2)
the impact of suction on compression. The initial void ratio Cs is determined using the elastic part of the compression
(e0) of the samples was determined by measuring the water curve and as the slope of the loading path up to the smallest
content, density and specific gravity of an identical sample vertical loading (Keller et al., 2011), which is 12?5 kPa in
immediately after vacuum treatment. The incremental this study. This can be expressed by
loads applied were 12?5, 25, 50, 100, 200, 400, 800 and
1600 kPa, and each increment was kept constant for a e0 {e12:5
Cs ~ (3)
duration of 24 h. log 12:5
In the following text, figures and tables, the notation i-j-k
represents a sample spiked with k% zinc contaminants, where e0 is the void ratio before loading and e12?5 is the void
stabilised with j % cement and cured for i days (e.g. 28-8-0?2 ratio after the first incremental loading.
represents 28 d curing time, 8% cement and 0?2% zinc). Unlike Cc and Cs, the magnitude of s’vp is dependent on
the methods used for estimation. In this study, both the
Casagrande (C) method and the maximum curvature (MC)
DETERMINATION OF COMPRESSION PARAMETERS method were adopted. The C method defines s’vp as the stress
Data from oedometer tests are often interpreted as at the intersect of the linear virgin compression curve and the
compression curves in a semi-logarithmic diagram (e– bisect of a horizontal and tangent drawn at the point of maxi-
log s’v ). Three important compression parameters can be mum curvature (Casagrande, 1936). The MC method simply
obtained from the compression curves, namely s’vp , Cs and uses the point of maximum curvature (Arvidsson & Keller,
Cc. In this study, the compression curves for all samples 2004). It should be noted that s’vp determined by the MC
were fitted by a logistic function, as attempted by Gregory method is always smaller than that by the C method (Gregory
et al. (2006), using a non-linear least-squares fitting method. et al., 2006). Detailed mathematical calculations for the C
The logistic function is expressed by (Gregory et al., 2006) method and the MC method are shown in Appendix 1 and
Appendix 2.
e~azc=(1zeb(log s’v {m) ) (1)
where e is the void ratio, s’v is the vertical effective stress TEST RESULTS
and a, b, c and m are fitting parameters. Although Gregory Figure 1 shows the e–s’v relations in a semi-logarithmic
et al. (2006) recommended the use of the Gompertz scale. The compression data were then fitted using the
function to fit soil compression behaviour due to its logistic function and the fitted parameters are summarised
in Table 2. Values of s’vp were estimated by both the C and
MC methods based on the fitted sigmoidal curves. In parti-
Table 1. Properties of kaolin clay used in this study
cular, s’vp for sample 28-8-0 was estimated by qualitative
Property Value comparison of the fitted curve with those of samples 28-8-
0?02 and 28-8-0?1 and the magnitude was between 900 and
Specific gravity Gs 2?65 1000 kPa.
Plastic limit wP: % 23 The relationship between s’vp and the initial zinc
Liquid limit wL: % 34 concentration is shown in Fig. 2. It is evident that, with
Grain size distributiona an increase in zinc concentration from 0 to 1?0%, s’vp
Clay (,0?002 mm): % 19?3
Silt (0?002–0?060 mm): % 76?0
decreased steadily from more than 600 kPa to less than
Sand (0?06–2?00 mm): % 4?7 75 kPa and from more than 900 kPa to less than 200 kPa
pHb 8?7 for curing periods of 7 d and 28 d, respectively. In
addition, the magnitudes of s’vp obtained using the C and
a MC methods are quite close to each other and they show
Based on BSI (1999)
b
Based on ASTM D4972 (2007) exactly the same variation patterns with respect to zinc
 Compressibility behaviour of cement-solidified/stabilised zinc-contaminated kaolin clay 29
2
The squares of the correlation coefficient (R ) are 0?86 and
1.6 0?88, respectively, indicating that the derived parabolic
formulations can be used as a useful engineering tool to
1.4 characterise the relationship between s’vp and qu. It was
also found that the fitted parabolic formulations in this
1.2 study lie well above the linear relation reported by
Horpibulsuk et al. (2004).
1.0 Figure 4 shows variations of Cs and Cc with zinc
7-8-0
e

7-8-0.02
concentration. The magnitudes of Cs for all samples with
0.8 7-8-0.1 a zinc concentration # 0?2% are almost identical regardless
7-8-0.2 of the zinc concentration and curing time. This indicates
7-8-0.5 that, when the zinc concentration is # 0?2%, the pre-yield
0.6
7-8-1
Fitting curve
stiffness is marginally affected by the addition of zinc
0.4 contaminants. In contrast, Cs increased from less than
10 100 1000 0?02 to 0?08 (at 28 d) and 0?20 (at 7 d) when the zinc
(a) concentration increased from 0?2 to 1?0%. This confirms
that the pre-yield stiffness seriously deteriorates at rela-
1.6 tively high zinc concentrations.
With regard to the post-yield stiffness, it is apparent that,
1.4 when the zinc concentration is # 0?1%, Cc remained at
around 0?6 and 0?7 at 7 and 28 d of curing, respectively (see
1.2 Fig. 4). As summarised by Mitchell & Soga (2005), most
intact natural clays have values of Cc less than 0?5.
1.0 Therefore, it appears that a cement-stabilised kaolin clay
e

28-8-0 with a low zinc concentration has greater resistance to

28-8-0.02 compression than most intact natural clays. However,
0.8
28-8-0.1
28-8-0.2 when the zinc concentration exceeded 0?1%, a dramatic
0.6 28-8-0.5 reduction in Cc was observed irrespective of the curing
28-8-1 time. Cc decreased to around 0?25 at 1?0% zinc concentra-
Fitting curve
0.4 tion, which is similar to the value found in untreated
10 100 1000 weakly structured natural clays or remoulded natural clays
σv′ : kPa as reported by Mitchell & Soga (2005). It has been reported
(b) that, for cemented sand (Cuccovillo & Coop, 1999) and
structured clay (Leroueil et al., 1983), Cc values of intact
Fig. 1. The e–log s9v relations and fitted logistic curves for samples were significantly larger than those of reconsti-
cement-solidified/stabilised zinc-contaminated kaolin clay after tuted specimens, which is attributed to the structural
(a) 7 d curing and (b) 28 d curing breakdown associated with greater compression pressure
than the yield stress. Leroueil et al. (1983) also indicated
concentration. Basically, the variations in s’vp with zinc that the Cc of structured clays would increase with
concentration are quite consistent with those of uncon- increasing sensitivity. As sensitivity is a widely used
fined compressive strength (qu) with zinc concentration indicator of soil structure, it is rational to conclude that
reported by Du et al. (2012b). Hence, s’vp is correlated Cc is positively related to the degree of cementation
with qu reported by Du et al. (2012b) with the same binder structure/bonding and hence the resistance to compression
dosage, curing time, zinc concentration and soil type (i.e. of the treated soils.
kaolin soil), as shown in Fig. 3. s’vp has a good parabolic The relationship between Cc and the liquid limit (wL) of
relationship with qu, as expressed by the stabilised kaolin clays is depicted in Fig. 5. An earlier
study that investigated the relationship between wL and
:
s’vp ~101:6q0u 44 ðCÞ (4) zinc concentration in stabilised kaolin clay (Du et al.,
2012b) is summarised in Table 3. A linear regression of Cc
: and wL was conducted and the following equation was
s’vp ~86:9qu0 46 ðMCÞ (5) obtained (R2 5 0?93)

Table 2. Fitted parameters of the logistic function for compression curves

Sample ID a b c m R2

7-8-0 0?754 3?828 0?678 3?012 0?998

7-8-0?02 0?647 3?098 0?763 3?093 0?998
7-8-0?1 0?780 3?465 0?686 2?938 0?999
7-8-0?2 0?685 2?156 0?799 2?763 0?997
7-8-0?5 0?511 1?316 0?953 2?586 0?997
7-8-1?0 0?126 0?880 1?187 2?957 0?999
28-8-0a 246?073 2?606 47?484 5?107 0?995
28-8-0?02 0?620 3?586 0?793 3?171 0?999
28-8-0?1 0?472 2?989 0?979 3?145 0?998
28-8-0?2 0?716 3?228 0?732 3?039 0?999
28-8-0?5 0?517 1?149 0?962 3?133 0?999
28-8-1?0 0?593 1?342 0?738 2?938 0?999
a
This case does not satisfy the requirement for fitted m
30 Jiang, Du, Liu and Zhu

1200 0.8
7 d (MC method)
0.7
28 d (MC method)
1000 900–
7 d (C method) 0.6
1000 kPa
28 d (C method) 0.5
800 ′ at Zn 0%
7 d (Cc)

Cc (Cs)
σ vp
0.4 28 d (Cc)
′ : kPa

656 kPa 7 d (Cs)

630 kPa
600 0.3 28 d (Cs)
σvp

0.2
400
0.1

200 0
0 0.01 0.1 1
Zn concentration: %
0
0.01 0.1 1
Zn concentration: % Fig. 4. Effect of zinc concentration on the swelling index (Cs) at
pre-yield and compression index (Cc) at post-yield of cement-
solidified/stabilised kaolin clay
Fig. 2. Effect of zinc concentration on yield stress (s9vp) of
cement-solidified/stabilised kaolin clay
DISCUSSION
This study revealed that the compressibility of cement-
Cc ~0:019(wL {19:3) (6) stabilised kaolin clays varies with zinc concentration. This
The Cc–wL relationship obtained from this study was observation is attributed to the negative influence of zinc
compared with that reported by Terzaghi & Peck (1948) for contamination on the cementation structure/bonding in the
soils. Du et al. (2012b) reported that, with an increase in
normally consolidated moderately sensitive intact clays
zinc concentration, the volume and mean diameters of inter-
(equation (7)), Bowles (1979) for Brazilian clays (equation
aggregate pores of cement-stabilised kaolin clays increased
(8)) and Nagaraj & Murthy (1985) for several natural clays
noticeably, whereas the magnitude of the intra-aggregate
(equation (9)). These correlations are also plotted in Fig. 5.
pores was nearly unchanged. Apparently, an increase in the
Cc ~0:009(wL {10) (7) volume and diameters of inter-aggregate pores alters the
microstructure of the stabilised soil and deteriorates
the initially developed strong cementation bonding. In
Cc ~0:0046(wL {9) (8) addition, Du et al. (2012b) indicated that the quantities of
major hydration products (e.g. C-S-H, portlandite and
Cc ~0:2343(wL =100)Gs (9) ettringite) decreased with an increase in zinc concentration,
and ultimately disappeared. Since these hydration products
where Gs is the specific gravity of the soil. The linear fitting are believed to be responsible for the strong cementation
curve in this study is well above those reported in the bonding and structure of stabilised soils, their absence
literature. In other words, the soils tested in this study definitely weakens the cementation bonding between parti-
exhibited a higher Cc relative to the natural intact clays at a cles. A new product, CaZn2(OH)6?2H2O, was observed to be
given wL. As interpreted earlier, a higher degree of formed in cement-stabilised kaolin when the zinc concentra-
structure/bonding in natural intact soils corresponds to a tion was higher than 0?2% (Du et al., 2012b). This product
larger Cc; therefore, the stabilised kaolin clays would have surrounds the unhydrated cement clinkers, preventing
stronger structure/bonding relative to those natural clays further hydration reactions and cementation bonding within
reported by earlier researchers. the soil matrix.

1000 0.9
′ = 101.6qu0.44
σvp 7d
0.8 28 d
R2 = 0.86
Fitting cuvre in this study
800 Terzaghi & Peck (1948)
0.7
Bowles (1979)
Nagaraj & Murthy (1985)
0.6
600
σ ′vp: kPa

0.5
Cc

400 ′ = 86.9qu0.46
σvp 0.4
R2 = 0.87 0.3
200 MC method
C method 0.2
Fitting curve (MC)
0 Fitting curve (C) 0.1

0
0 50 100 150 200 250 30 35 40 45 50 55 60 65
qu: kPa WL

Fig. 3. Relationship between qu and s9vp for cement-solidified/ Fig. 5. Relationship between compression index (Cc) at the
stabilised kaolin clay (zinc concentrations 0, 0?02, 0?1, 0?2, 0?5 post-yield state and the liquid limit (wL) for cement-solidified/
and 1?0%; curing times 7 and 28 d) stabilised kaolin clay
 Compressibility behaviour of cement-solidified/stabilised zinc-contaminated kaolin clay 31

Table 3. Variations in liquid limit with zinc concentration in cement-solidified/stabilised soil (modified from Du et al. (2012b))

Liquid limit, wL: %

Zinc concentration: %a

Curing time: d 0 0?02 0?1 0?2 0?5 1?0 Clean kaolin

7 53 54 49 39 35 36 34
28 60 55 53 49 34 34 34
a
Dry soil weight basis

Further work regarding the compressibility of stabilised lnf½(bc{4)(bc{12)1=2 =4{(bc=4)z2g

heavy metal contaminated field soils is underway, with log s’vMC ~m{
b
more focus on the impact of heavy metal speciation on
compressibility.
eMC ~azc=(1zeb(log s’vMC {m) )

CONCLUSIONS
The following conclusions can be drawn from this study. APPENDIX 2: CASAGRANDE METHOD
N The yield stress (s’vp ) decreases gradually with increas- The horizontal line at the point of maximum curvature is
ing zinc concentration in cement-stabilised kaolin clay e~eMC
after both 7 and 28 d of curing. s’vp has a good
parabolic relationship with the unconfined compressive The tangent line at the point of maximum curvature is
strength (qu).
N For zinc concentrations # 0?2%, the swelling index (Cs) de
e{eMC ~ |(log s’v {log s’vMC )
at the pre-yield state is marginally affected by zinc d(log s’v ) s’vMC
contaminants; however, it increases sharply at higher
zinc concentrations. The bisect of the horizontal and tangent drawn at the point
N The compression index (Cc) at the post-yield state is of maximum curvature is
stable at zinc concentrations # 0?1%, but exhibits a de
significant drop beyond this concentration. Cc is linearly e{eMC { |(log s’v {log s’vMC )
correlated with the liquid limit (wL) and stabilised kaolin d(log s’v ) s’vMC
e{eMC ~ #2 11=2
clays obtain a higher Cc relative to intact natural clays.
0 "
N The impact of zinc contamination on compressibility is @1z de A
attributed to the negative influence of zinc on cement d(log s’v ) s’vMC
hydration and thus the phases of hydration products
that alter the pore profiles in the soil matrix, leading to a The linear virgin compression curve is then
deteriorated cementation structure/bonding.
bc c
e~{ log s’v zaz (2zbm)
4 4
APPENDIX 1: MAXIMUM CURVATURE METHOD
The logistic function is

e~azc=(1zeb(log s’v {m) ) ACKNOWLEDGEMENTS

This work was funded by the National Natural Science
The first derivative of the logistic function is Foundation (no. 51278100) of China, the Natural Science
Foundation (no. BK2010060 and no. BK2012022) of Jiangsu
de {bc eb(log s’v {m) Province of China and the National High Technology
~
d(log s’v ) (1zeb(log s’v {m) )2 Research and Development Program of China (grant
no. 2013AA06A206). The authors thank the undergraduate
and the second derivative of the logistic function is students for their work in the relevant laboratory tests.

d2 e ({b2 c eb(log s’v {m) )(1{eb(log s’v {m) ) REFERENCES

2
~
d(log s’v ) (1zeb(log s’v {m) )3 Arvidsson, J. & Keller, T. (2004). Soil precompression stress: I. A
survey of Swedish arable soils. Soil Till. Res. 77, No. 1, 85–95.
ASTM (2007). ASTM D4972: Standard test method for pH of
The curvature function is soils. West Conshohoken, PA: ASTM International.
Boardman, D., Glendinning, S. & Rogers, C. (2004). The
d2 e=d(log s’v )2 influences of iron (III) and lead (II) contaminants on lime-
k~
½1z(de=d(log s’v ))2 3=2 stabilised clay. Géotechnique 54, No. 7, 467–486.
Bowles, J. W. (1979). Physical and geotechnical properties of soils.
and maximum curvature is New York: McGraw-Hill.
BSI (1999). BS 5930:1999: Code of practice for site investigations.
dk London: BSI.
~0 Casagrande, A. (1936). The determination of pre-consolidation load
d(log s’v )
and its practical significance. Proc. 1st Int. Conf. on Soil Mechanics
The point of maximum curvature is given by and Foundation Engineering, Cambridge, MA, USA 3, 60–64.
32 Jiang, Du, Liu and Zhu
Cuccovillo, T. & Coop, M. R. (1999). On the mechanics of Mitchell, J. K. & Soga, K. (2005). Fundamentals of soil behavior,
structured sands. Géotechnique 49, No. 6, 741–760. 3rd edn. New York: Wiley.
Du, Y.-J., Jiang, N.-J., Shen, S.-L. & Jin, F. (2012a). Experimental Moffett, B. F., Nicholson, F. A., Uwakwe, N. C., Chambers, B. J.,
investigation of influence of acid rain on leaching and Harris, J. A. & Hill, T. C. J. (2003). Zinc contamination
hydraulic characteristics of cement-based solidified/stabilized decreases the bacterial diversity of agricultural soil. FEMS
lead contaminated clay. J. Hazard. Mater. 225–226, No. 12, Microbiol. Ecol. 43, No. 1, 13–19.
195–201. Nagaraj, T. & Murthy, B. S. (1985). Prediction of the
Du, Y.-J., Jiang, N.-J., Wang, L. & Wei, M.-L. (2012b). Strength preconsolidation pressure and recompression index of soils.
and microstructure characteristics of cement-based solidified/ Geotech. Test. J. 8, No. 4, 199–202.
stabilized zinc-contaminated kaolin. Chinese J. Geotech. Engng Rios, S., Viana da Fonseca, A. & Baudet, B. (2012). Effect of
34, No. 11, 2114–2120. the porosity/cement ratio on the compression of cemented
Du, Y.-J., Wei, M.-L., Jin, F. & Liu, Z.-B. (2013). Stress-strain soil. J. Geotech. Geoenviron. Engng 138, No. 11, 1422–
relation and strength characteristics of cement treated zinc- 1426.
contaminated clay. Engng Geol. 167, No. 1, 20–26. Rotta, G., Consoli, N., Prietto, P., Coop, M. & Graham, J. (2003).
Gregory, A. S., Whalley, W. R., Watts, C. W., Bird, N. R. A., Isotropic yielding in an artificially cemented soil cured under
Hallett, P. D. & Whitmore, A. P. (2006). Calculation of the stress. Géotechnique 53, No. 5, 493–501.
compression index and precompression stress from soil Terzaghi, K. & Peck, R. B. (1948). Soil mechanics in engineering
compression test data. Soil Till. Res. 89, No. 1, 45–57. practice. New York: Wiley.
Horpibulsuk, S., Bergado, D. & Lorenzo, G. (2004). Compressibility Tremblay, H., Leroueil, S. & Locat, J. (2001). Mechanical
of cement-admixed clays at high water content. Géotechnique 54, improvement and vertical yield stress prediction of clayey soils
No. 2, 151–154. from eastern Canada treated with lime or cement. Can.
Keller, T., Lamandé, M., Schjønning, P. & Dexter, A. R. (2011). Geotech. J. 38, No. 3, 567–579.
Analysis of soil compression curves from uniaxial confined Voegelin, A., Pfister, S., Scheinost, A. C., Marcus, M. A. &
compression tests. Geoderma 163, No. 1–2, 13–23. Kretzschmar, R. (2005). Changes in zinc speciation in field soil
Klimek, B. (2012). Effect of long-term zinc pollution on soil after contamination with zinc oxide. Environ. Sci. Technol. 39,
microbial community resistance to repeated contamination. No. 17, 6616–6623.
Bull. Environ. Contam. Toxicol. 88, No. 4, 617–622. Voglar, G. E. & Lestan, D. (2010). Solidification/stabilisation of
Leroueil, S., Tavenas, F. & Le Bihan, J.-P. (1983). Propriétés metals contaminated industrial soil from former Zn smelter in
caractéristiques des argiles de l’est du Canada. Can. Geotech. Celje, Slovenia, using cement as a hydraulic binder. J. Hazard.
J. 20, No. 4, 681–705. Mater. 178, No. 1–3, 926–933.

WHAT DO YOU THINK?

To discuss this paper, please email up to 500 words to
the editor at journals@ice.org.uk. Your contribution will
be forwarded to the author(s) for a reply and, if
considered appropriate by the editorial panel, will be
published as a discussion.