CLAY-02989; No of Pages 9

Applied Clay Science xxx (2014) xxxxxx

Contents lists available at ScienceDirect

Applied Clay Science

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

Research paper

Chemo-mechanical modelling of lime treated soils

V. Robin a,b, O. Cuisinier a,, F. Masrouri a, A. A. Javadi b
a
b

LEMTA, UMR 7563 CNRS, Laboratoire d'Energtique et de Mcanique Thorique et Applique, Universit de Lorraine, France
Computational Geomechanics Group, Department of Engineering, University of Exeter, United Kingdom

a r t i c l e

i n f o

Article history:
Received 21 January 2014
Received in revised form 9 April 2014
Accepted 10 April 2014
Available online xxxx
Keywords:
Quicklime
Mechanical behaviour
Chemo-mechanical coupling
Constitutive modelling

a b s t r a c t
Lime treatment has been widely used to improve mechanical properties of soils. However, less has been done to
account for the effect of the treatment on constitutive relationships. In this study, a comprehensive programme of
isotropic consolidation tests and drained triaxial experiments were designed and carried out on saturated specimens of a silt treated with quicklime. The chemical composition in hydrates, portlandite, and carbonates was determined using thermogravimetric analysis and differential thermal analysis. The modications to the
mechanical parameters of the soil were evaluated in the framework of the Cam Clay elastoplastic model. The experimental results revealed that the addition of lime leads to the modication of the critical state. For concentrations in lime higher than 1%, the treated specimens displayed a different normal compression line compared to
the untreated state. Chemical analysis revealed the production of cementitious compounds for every concentration tested. A constitutive model was proposed to describe the observed behaviour of lime treated soils in the
framework of the Structured Cam Clay. The model accounts for the modications on the mechanical parameters
of the soil. A chemo-mechanical coupling was established between the yield stress and the mass concentration in
cementitious compounds.
2014 Elsevier B.V. All rights reserved.

1. Introduction
For economic and environmental considerations, engineering companies are highly encouraged to use on site materials to construct
earth structures like embankment, river levees, and earth dams. For
soils with low mechanical characteristics, lime treatment (CaO) is an efcient method to improve their properties and allow their use in geotechnical structures (e.g. Little, 1995). It is of the greatest interest to
account for articial treatments in geotechnical design.
When added to the soil, quicklime reacts with the free water to produce hydrated lime, known as portlandite (Ca(OH)2). Then, calcium cations Ca2+ and hydroxides anions HO are released in the soil, which
increase the pH of the porous medium. Cation exchange process combined with the presence of large amounts of calcium ions adsorbed on
the clay particles surface leads to a reduction of the size of the double
diffuse layer of the clay particles. This results in the lowering of the repelling forces between clay particles and thereby giving rise to the occulation of the clay particles. It is generally accepted that cation
exchange and occulation processes occur immediately after the addition of lime and result in a decrease of the soil plasticity (e.g. Eades
and Grim, 1966; Rogers and Glendinning, 1996). Moreover, the addition
of lime leads to a high pH environment (near 12.4), which enables the
Corresponding author at: LEMTA UMR 7563, Universit de Lorraine / CNRS 2 Rue du
Doyen Marcel Roubault TSA 7060554518 Vanduvre-les-Nancy Cedex France.
E-mail address: [email protected] (O. Cuisinier).

dissolution of both silica and alumina present in the soil (e.g. North
et al., 2008). These later react with the calcium. As a result of these
chemical reactions, known as pozzolanic reactions, cementitious compounds such as calcium silicate hydrates (CSH), calcium aluminate hydrates (CAH), and calcium aluminosilicate hydrates (CASH) can be
formed (e.g. Metelkov et al., 2012; Pomakhina et al., 2012). The formation of these compounds leads to a modication in the soil mechanical
properties (e.g. Little, 1995).
Many authors have studied the mechanical behaviour of materials treated with lime or cement (e.g. Consoli et al., 2011; Cuisinier
et al., 2008; Cuisinier et al., 2011; Malandraki and Toll, 2001;
Oliveira et al., 2013; Stoltz et al., 2012). In most of the cases, the addition of quicklime leads to a modication of the mechanical parameters such as the cohesion, the friction angle, and the yield stress.
Improvements obtained with lime treatment regarding tensile
strength and Young's modulus have been integrated in the design
of pavement (Thompson, 1965). Nevertheless, less has been done
to consider these improvements in the design of structures for global
stability or settlement analysis. This last issue could be associated
with the fact that there is no specic constitutive relationship accounting for the specic aspects of the mechanical behaviour of
lime-treated soils and for the coupling between physicochemical
processes and mechanical behaviour. Such relationship is required
to account for the treatment in the design of geotechnical structures,
in the short term and also to foretell the long term behaviour of
earthen structures built with lime-treated soils.

http://dx.doi.org/10.1016/j.clay.2014.04.015
0169-1317/ 2014 Elsevier B.V. All rights reserved.

Please cite this article as: Robin, V., et al., Chemo-mechanical modelling of lime treated soils, Appl. Clay Sci. (2014), http://dx.doi.org/10.1016/
j.clay.2014.04.015

V. Robin et al. / Applied Clay Science xxx (2014) xxxxxx

Lime treatment induces the formation of cementitious compounds

that bind soil particles. Some studies (Leroueil and Vaughan, 1990; Liu
and Carter, 2003) have shown that naturally structured soils and articially treated materials have common mechanical features, treatment
appearing to create a structure in the soil. In this paper, structure refers to the combination of the fabric and the bonding of the soil skeleton
(Burland, 1990). Fabric accounts for the arrangement of particles, which
depends on the state of compaction and their geometry.
Several constitutive models have been proposed for structured materials (Baudet and Stallebrass, 2004; Cotecchia and Chandler, 2000;
Gens and Nova, 1993; Lee et al., 2004; Nova et al., 2003; Vatsala et al.,
2001). Liu and Carter (2002) proposed a Cam Clay based constitutive
model for structured materials, adding only three extra parameters to
the original Modied Cam Clay (Roscoe and Burland, 1968). Since
then several enhancements (Horpibulsuk et al., 2010a; Suebsuk et al.,
2011, 2010) have been proposed. Most of these constitutive models
use the destructured state as reference to describe the mechanical behaviour of the structured soils. The destructured state accounts for the
intrinsic parameters of the soil, and especially the geometry of the particles. Articial cementation arising from the treatment involves chemical reactions that could alter the soil particles. This aspect should be
considered to adapt the existing framework to the case of lime stabilised
compacted soils.
A key aspect of lime stabilisation is the fact that the structure, and
thus the mechanical behaviour, relies on physicochemical processes
that need to be quantied. Based on the soil water transfer method
(Zhu et al., 2007), Chiu et al. (2009) used centrifuge method to determine the variation of soil water composition in dredged materials with
high water content. A different approach was used in this study to assess
the chemical composition of lime treated compacted soils with lower
water content. Chemical composition in hydrates and portlandite in cement and concrete was successfully determined using thermogravimetric analysis (TGA) and differential thermal analysis (DTA). Some studies
(Horpibulsuk et al., 2010b; Maubec, 2010) have shown the suitability of
this method to determine the chemical modications following a lime
treatment. The objective of the paper is to investigate the couplings between mechanical behaviour and physicochemical processes, and then
to introduce these aspects in a constitutive relationship.
To account for the treatment in the structure design, a chemomechanical coupling is required to predict the mechanical improvements resulting from the addition of a given amount of lime. This implies a good understanding of the mechanical behaviour of lime
treated soils and of the chemical modications arising from the treatment. To this end, the paper aims at shedding light on three aspects:

Table 1
Characteristics of the silt.
Value
Geotechnical properties
Liquid limit LL (%)
Plastic limit PL (%)
Plasticity index PI
Particle density s (Mgm3)
Methylene blue value (g/100 g of dry soil)
USCSa

31.2
7.8
23.4
2.66
2.1
ML

Physical chemistry properties

pH ()
Cation-exchange capacity (cmol+/100 g)
Carbonate mass concentration (%)

5.7
7.41
0.6

Initial conditions
Dry density (Mgm3)
Water content w (%)
Initial void ratio ei ()
a

1.68
20
0.6

Unied Soil Classication System.

specimen (0%) was used as reference to assess the mechanical and the
chemical modications arising from the treatment.
All the specimens used for the mechanical characterisation of the
soil, with or without treatment, were prepared at the same moisture
content and dry density. Indeed, the optimal moisture content and maximum dry density of the silt used in this study were not signicantly
modied by the lime-treatment (Table 2). Therefore, all the specimens
have been prepared at the same mean initial water content of 20% and
a dry density of 1.68 Mgm3. Thus, any modication of the mechanical
behaviour of the soil after treatment can be attributed mostly to structure modication, i.e. fabric and bonding, but not to density.
The soil was rst prepared at the target water content. After an
equilibration period of several days, quicklime was added to the soil,
both being mixed thoroughly until a homogenous mixture was obtained. Before the specimens were compacted, the quicklimesoil mixture
was sealed in hermetic plastic bags for 1 h before compaction. Then,
35 mm 70 mm triaxial specimens were statically compacted up to
the target dry density. After these steps, the actual water content of
the specimens was determined. It varied between 19.6% and 20.2%.
The specimens were then wrapped in plastic bags to prevent any exchange with the atmosphere, and a curing time of 28 days was
respected.

2.2. Drained triaxial experiments

a) How can the effects of a lime treatment on a material and its consequences on the mechanical properties be described?
b) How can a chemo-mechanical coupling be established between the
mechanical improvements and the chemical modications?
c) How can a lime treated soil be modelled?

2. Material and methods

2.1. Tested soil and specimens preparation
The material selected in this study is a silt from the east part of
France. The mineral composition is as follows: quartz (55%), kaolinite
(12%), feldspar (11%), illite (10%), montmorillonite (4%), chlorite (1%),
goethite (6.4%), and carbonates (0.6%). The lime xation point is near
1% and was determined according to the ASTM Standard D6276
(2006). Before proceeding to the moulding of the specimen, the material was rst heated at 60 C and then sieved to get particles smaller than
2 mm. The main characteristics of the soil are listed in Table 1.
Five concentrations in quicklime were considered in this study: 0,
0.5, 1, 2, and 5% expressed on a soil dry weight basis. The non-treated

To determine the shape of the yield function, three kinds of stress

path were performed in this study: isotropic consolidation, drained
paths ( 3 constant), and same stress ratio paths ( = p/q = 0.39).
Saturation was achieved by rst creating an upward water ow
through the specimen. This was achieved by applying a pressure gradient between the bottom and the top of the specimen. Deaerated water
was used to maximise the dissolution of air. Pore pressure was then increased in three stages (Table 3) while maintaining a constant effective
mean stress.

Table 2
Results of proctor compression test for different concentrations in lime.
Lime content
(%)

Optimal moisture content

(%)

Optimal dry density

(Mgm3)

0
1
2
3

17.9
18.8
20.3
20.9

1.76
1.72
1.68
1.64

Please cite this article as: Robin, V., et al., Chemo-mechanical modelling of lime treated soils, Appl. Clay Sci. (2014), http://dx.doi.org/10.1016/
j.clay.2014.04.015

V. Robin et al. / Applied Clay Science xxx (2014) xxxxxx

Table 3
Saturation stages performed.
Stage 1

Stage 2

Stage 3

Duration (days)

3 (kPa)
Pbottom (kPa)
Ptop (kPa)
03 (kPa)

50
30
20
25

100
80
70
25

200
180
170
25

Pbottom: pressure at the bottom of the specimen; Ptop: pressure at the top of the specimen.

After every triaxial experiment, saturation of the specimen was

checked by determining the degree of saturation with parafn according to the ASTM Standard D7263-09 (2009).
To ensure full drainage of the specimens during the experiments, lter paper strips were applied to the surface of the specimens. Low rates
of consolidation and axial deformation were chosen in order to generate
small excess of pore pressure. An external probe was used to measure
accurately the pore pressure at the bottom of the specimen during
shearing.
For isotropic consolidation, a rate of 3.47 Pas1 was chosen and
specimens were consolidated up to a total effective stress of 3320 kPa.
Validation tests were performed to verify that there was no excess
pore pressure in the specimen under this rate.
Triaxial compression tests were carried out with an axial displacement rate of 2.46 m min1, leading to an axial deformation of
about 5% per day. Validation tests have conrmed the suitability of
this rate. All the experiments were performed following the procedure
described in the ASTM Standard D7181-11 (2011). The initial conditions
of the specimens are given Table 4.

test is completed with a differential thermal analysis (DTA), which compares any temperature difference between the specimen and an inert
reference (MacKenzie, 1970). Each peak on the curves is correlated to
a chemical compound. Using the associated mass loss, the mass concentration of the species can be estimated (Mackenzie, 1972).
TGA/DTA analysis was rst used successfully to determine the chemical composition of cements and concretes in hydrated products,
portlandite, and carbonates. Some studies have shown that the decarboxylation processes of these three compounds were associated with
three temperature domains (Alarcon-Ruiz et al., 2005; Das et al., 1996;
Saikia et al., 2002). Recent studies have applied TGA/DTA analysis on articially treated soils to assess the mass concentration in portlandite
(unreacted lime), cementitious compounds, and calcium carbonates in
articially treated specimens (Horpibulsuk et al., 2010b; Maubec,
2010).
The procedure for the determination of the chemical composition is
illustrated in Fig. 1. TGA/DTA analysis provides three different curves:
the variation of the energy (TDA), the mass loss (TG), and the rst derivative of the weight loss (DTG). First, TDA and DTG results were used to
determine accurately the three temperature domains (Table 5).
Each peak on the DTG curve (Fig. 1) is associated to a variation of the
weight loss and corresponds to the decarboxylation of a chemical compound. The two temperatures delimitating the peak correspond to the
temperature domains, and the chemical compound is assumed to
have completely disappeared when the upper temperature limit was
reached. Therefore, the variation of weight (TG curve) between these
two temperatures gives the mass of the chemical compound. Analyses
were performed using a NETZSCH STA 409 PC/PG device. The accuracy
of the TGA weighing scale was 0.01 mg.

2.3. Monitoring of the physicochemical reactions

0.05

CaOs H2 0l CaOH2 aq 1140 kJ=kg of CaO:

Dissociation of portlandite induces release of hydroxide anions, thus

increasing the pH. Once a suitable pH is reached in the porous medium,
portlandite is consumed by pozzolanic reactions to produce hydrates.
Due to the high sensibility of lime with carbon dioxide, we also considered the evolution in carbonates in the specimens. Carbonates are
produced by reaction of quicklime and carbon dioxide according to
the following reaction:

0.00

0.05

0.10

0.15

Free Water
0.20

100

Therefore, to monitor the progress of the physicochemical processes

in the specimens, it is necessary to determine the amount of portlandite,
hydrates, and carbonates.
Thermogravimetric analysis (TGA) consists of measuring the mass
loss of a specimen as a function of the increasing temperature. This

98

Table 4
Characteristics of the specimens.

20.0
19.6
20.2
19.6
20.2

200

Portlandite

300

400

500

Carbonates
600

700

800

TGA

96

94

92

Lime content Water content d (Mgm3) eo ()

(%)
(%)
0
0.5
1
2
5

100

Hyd.

Temperature (C)

Weight loss (%)

CaOs CO2 g CaCO3 s :

Rateofweightloss(%/min)

DTG
Lime can react with soil particles to produce only three different
chemical compounds: portlandite, hydrates (cementitious compounds
CAH, CSH, and CASH), and carbonates. Portlandite is produced by the
hydration of quicklime introduced in the material, according to the following reaction:

1.67
1.67
1.67
1.67
1.66

0.01
0.01
0.01
0.01
0.01

0.60
0.60
0.60
0.60
0.60

Hydraulic permeability
k (m/s)
0.02
0.01
0.01
0.01
0.02

109108
109108
109108
108107
108107

Free Water
90

100

200

Hyd.
300

Portlandite
400

500

Carbonates
600

700

800

Temperature (C)
Fig. 1. TGA/DTA results for a lime treated specimen with the three temperatures domains
Hyd.: hydrates, DTG: differential thermo gravimetric, TGA: thermogravimetric analysis.

Please cite this article as: Robin, V., et al., Chemo-mechanical modelling of lime treated soils, Appl. Clay Sci. (2014), http://dx.doi.org/10.1016/
j.clay.2014.04.015

V. Robin et al. / Applied Clay Science xxx (2014) xxxxxx

1800

Table 5
Ranges of temperatures used for the determination of the TGA/DTA.
Chemical compound

20 C217 C
217 C350 C
350 C610 C
610 C800 C

Free and adsorbed water

Cementitious compounds
Portlandite (Ca(OH)2)
Carbonates (CaCO3)

Yield stress (kPa)

Range of temperature

1600
1400
1200
1000
800
600
400
200
0

3. Results

[CaO] (%)
The experiments performed in this paper aimed: 1) to determine the
effects of the quicklime on the intrinsic parameters of soils, 2) to assess
the shape of the yield function in order to choose the most suitable
model for lime-treated soils, and 3) to measure the concentrations in
portlandite, hydrates, and carbonates of lime-treated specimens.
3.1. Mechanical behaviour
3.1.1. Isotropic consolidation
To assess the effects of the treatment on the yield stress, isotropic
consolidation tests were performed on 0, 0.5, 1, 2, and 5% lime-treated
specimens. For concentrations greater than 1%, three isotropic tests
were carried out to ensure the highest representativeness. Only one result is reported in Fig. 2. The yield stress was determined using
Casagrande's method.
Yield occurred for an effective mean stress of 177 kPa for the untreated specimen (Fig. 2). When lime was added to the soil, the yield
stress was increased even for low lime content of 0.5%. The evolution
of the yield stress as a function of the lime content is given in Fig. 3.
For concentrations between 0.5% and 2%, the gradient of the curve
was signicantly increased. Above 2% the slope decreased, and as concentration increased, yield stress value seemed to approach 2000 kPa.
One can note that there exists a one-to-one correspondence between
the lime content and the resulting yield stress after 28 days of curing.
At yield the treated soil displayed higher specic volume compared
to the reference state (Fig. 2). This could be due to the structure. When
the effective mean stress reached the yield stress, the additional specic
volume began to decrease as the effective mean stress was increased
until a new normal compression line was reached, parallel to the one

1.65

1.60

Specific volume (-)

1.45

1.30

3.1.2. Shear behaviour

Drained triaxial experiments were performed to assess the shape of
the yield function and the gradient of the critical state line. Different
stress paths and pre-consolidation pressures were performed on 0, 1,
2, and 5% lime-treated specimens.
The shear behaviour of the non-treated specimens (Fig. 4a) was signicantly improved with 5% lime treatment (Fig. 4b). For an isotropic
pre-consolidation pressure of 245 kPa, the yield locus of the reference
state was reached for a deviatoric stress of 400 kPa. With 5% of lime,
this value was increased up to 1250 kPa.
As the conning pressure was increased, one can see that contraction behaviour was gradually observed. This feature was particularly
marked for the 5% lime-treated specimens (Fig. 4b). For a conning
pressure of 20 kPa, the specimens had signicant dilation behaviour.
As the conning pressure increased, the specimens started to exhibit a
contraction behaviour.

Yield loci and critical states were determined from the previous results and plotted in the (p,q) plane (Fig. 5). Results from the same stress
ratio paths ( = 0.39) were also included.
To use these results in a constitutive model, yield loci and critical
states were determined based on the critical state theory (Muir Wood,
1991). Yield was assumed to occur when plastic deformation appeared
(pq N 0), and critical state was reached when:

1.50

1.35

of the non-treated state. This decrease of the additional specic volume

corresponds to the loss of the structure. Only 0.5% lime treated specimen has shown to completely lose the additional specic volume at
high mean stresses. In this case, the normal compression line matched
the one of the non-treated state. For lime contents greater than 1%, specimens did not appear to converge to the non-treated state for mean effective stresses lower than 3320 kPa. At some point, a secondary normal
compression line, different from the non-treated state but still parallel,
seems to be reached. This feature is particularly noteworthy for the 1%
treated specimen. The 2% and 5% lime-treated specimens have shown
such signicant improvements in mechanical properties that this feature could not be seen because of the limits of the applied pressure in
the triaxial cell (3500 kPa).

3.2. Yield loci for lime-treated soils

1.55

1.40

Fig. 3. Evolution of the yield stress after 28 days of curing with the lime.

Isotropic - 0%
Isotropic - 0.5%
Isotropic - 1%
Isotropic - 2%
Isotropic - 5%
102

p
q
v

0
dq dq dq

103

Mean effective stress (kPa)

Fig. 2. Isotropic consolidation curves for ve concentrations in lime.

The elastic domain showed to be signicantly increased with the increase of lime content, without any signicant anisotropic behaviour,
and the critical state lines appeared to be modied with the treatment.
The gradient and the y-intercept of the critical state lines both increased
with the increase of the amount in lime. The results appear to be well
described by the Modied Cam Clay model.

Please cite this article as: Robin, V., et al., Chemo-mechanical modelling of lime treated soils, Appl. Clay Sci. (2014), http://dx.doi.org/10.1016/
j.clay.2014.04.015

V. Robin et al. / Applied Clay Science xxx (2014) xxxxxx

450
400

= 20kPa

= 100kPa

= 245kPa

3000

2500

300

q = 1 3 (kPa)

q = 1 3 (kPa)

350

250
200
150

2000

= 20kPa

= 100kPa

= 245kPa

= 900kPa

1500

1000

100
500
50
0

10

Axialstrain a = l/l0 (%)

1.0

0.5

10

10

p = V/V0 (%)

p = V/V0 (%)

0.0
0.5
1.0
1.5

1
0
1

2.0

2.5
3.0

Axialstrain a = l/l0 (%)

10

Axialstrain a = l/l0 (%)

Axialstrain a = l/l0 (%)

Fig. 4. Drained triaxial results on 0% and 5% lime treated specimens p: volumetric deformations, a: axial strain, q: deviatoric stress. a: shear behaviour for untreated specimens; b: shear
behaviour for 5% lime treated specimens.

3.3. Chemical analysis

2.0
YL-0%
CS-0%

YL-1%
CS-1%

YL-2%
CS-2%

YL-5%
CS-5%

Deviatoric stress q = 1 3 (MPa)

M 0%

1.5

M 2%

M 5%

M 1%

1.0

0.5

0.0
0.0

0.5

1.0

Mean effective stress

1.5

2.0

+ 23
(MPa)
= 1
3

Fig. 5. Yield loci and critical states of 0%, 1%, 2%, and 5% lime treated specimens YL: yield
loci, CS: critical state, M: gradient of critical state line.

The results of TGA/DTA tests to measure the chemical composition in

portlandite, hydrates, and carbonates on 0, 0.5, 1, 2, and 5% lime treated
specimens are plotted in Fig. 6. The y-axis values correspond to the variation of mass compared to the untreated specimen.
The chemical composition of the soil appears to be modied after addition of lime and a curing time of 28 days. Concentration in portlandite
is increased only for lime contents greater than 0.5%. The latter is used
by pozzolanic reactions to produce cementitious compounds. For a concentration of 0.5%, all the portlandite produced by the hydration of the
quicklime was used by pozzolanic reactions.
Concentration in hydrates was increased for all the lime contents. It
is worth to note that there is a one-to-one correspondence between the
lime content and the mass concentration in hydrates. The fact that neoportlandite is measured for high lime contents means that pozzolanic
reactions are unnished and concentration in hydrates is likely to
evolve if the curing time is extended.
Concentration in carbonates revealed to be increased only for
high lime contents. Probability of carbonation is higher at high
lime contents and this could explain the sudden formation of carbonates. It is more likely that these new carbonates are the result
of the reaction of the unhydrated lime with the atmospheric carbon
dioxide.

Please cite this article as: Robin, V., et al., Chemo-mechanical modelling of lime treated soils, Appl. Clay Sci. (2014), http://dx.doi.org/10.1016/
j.clay.2014.04.015

V. Robin et al. / Applied Clay Science xxx (2014) xxxxxx

0.40

Portlandite

0.35

(%)

0.30
0.25

0.20
0.15
0.10
0.05
0.00
0

[CaO] (%)
0.25

Hydrates

(%)

0.15

0.20

0.10
0.05
0.00
0

[CaO] (%)
0.9

Carbonates

0.8

(%)

0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
0

[CaO] (%)
Fig. 6. Inuence of a lime treatment on the variation of the mass concentration in hydrates
(cementitious compounds), portlandite, and carbonates for a curing time of 28 days.
mp: mass concentration in portlandite, mh: mass concentration in hydrates, mc:
mass concentration in carbonates.

cementation and the fabric. In the case of structured soils, Leroueil and
Vaughan (1990) showed that isotropic consolidation induced a progressive loss of structure at yield. After total loss of structure, the normal
compression line of the structured soil converges to that of the same
soil prepared in a remoulded state before compression.
In the case of a treated soil with quicklime, this assumption was only
valid for a soil treated with 0.5% of quicklime. Isotropic consolidation results revealed a full loss of the structure for 0.5% lime treated specimens,
which displayed the same virgin consolidation line as the non-treated
specimens at a mean stress value of 1340 kPa. However, for concentrations equal or greater than 1% it could be assumed that the position of
the virgin consolidation line following the destructuration is shifted.
This feature is usually controlled by the fabric, and especially by the geometry of the particles. Therefore, one may assume that the geometry of the
particles is modied by the treatment and is lime content-dependent.
This observation is corroborated by the drained triaxial test results,
which revealed a modication of the critical state line. The increase of
the y-intercept is a result of the cementation (Fig. 5). The slope of the critical state line, M, is a direct function of the angle of friction (Schoeld and
Wroth, 1968) and describes the relationship between the particles and
their geometry. A modication of M (Table 6) implies a modication of
the angle of friction, and therefore of the geometry of the particles
(Stocker, 1974; Wissa, 1965). Therefore, in the framework of the limetreated soils, our results show that these mechanical parameters are
modied.
Parameters like M and the normal compression line are considered
as intrinsic and invariable in the framework of naturally structured
soils (Liu and Carter, 2002). For these materials, reconstituted state is
used as reference to assess the mechanical improvements due to the
structure. In this case, the soil particles have already experienced cementation processes. The soil particles of the untreated state have not
been in contact with lime and hence no modication of their geometry
has occurred. From this point of view, the use of the mechanical parameters of the non-treated state as reference appears to be inappropriate
to assess the effects of the treatment. This is important to describe the
improvements, but not enough to fully describe the mechanical behaviour of treated soils. To do this, an intermediate state appears to be required, the destructured state of the lime treated soil. This state would
account for the effects of the treatment on the mechanical parameters.
These results have showed that mechanical parameters of a soil are
modied with the addition of quicklime. Therefore, it appears of the
greatest interest to compare those with the chemical modications arising from the treatment.

4.2. Correlation between physicochemical processes and yield stress

4. Discussion
In the rst part of the Discussion section, the inuence of lime on the
mechanical properties of soil is discussed. Moreover, a quantication of
the physicochemical processes induced by lime-treatment after 28 days
of curing is provided. The main objective of the second part of the Discussion section is to highlight the coupling between the mechanical behaviour and the amount of the different compounds formed during the
curing period and to propose an equation for the coupling. In the last
part, a framework to account for the effects of the treatment on the intrinsic parameters is proposed.
4.1. Inuence of lime on mechanical parameters
Lime has appeared to modify the mechanical behaviour of the soil
and parameters like the position of the virgin consolidation line and
the slope of the critical state line, M. Following the denition proposed
by Burland (1990), structure consists of the combination of the

It has appeared that lime treatment modied the nature of the soil
by altering parameters classically considered as intrinsic. In light of
these results and using the approach proposed by Chiu et al. (2009),
the yield stress was chosen as the mechanical coupling parameter.
Apart from controlling the size of the initial yield function in most of

Table 6
Inuence of lime on the mechanical properties.
Parameters

Values
0%

0.5%

1%

2%

5%

y (kPa)
v0

M
()
c (kPa)

177
1.602
0.075
0.005
1.092
27.5
11.4

300
1.609
0.088
0.005
N/A
N/A
N/A

731
1.612
0.085
0.005
1.136
28.5
47.9

1211
1.606
N/A
0.005
1.246
31.1
152.9

1691
1.626
N/A
0.005
1.464
36.0
207.6

c: cohesion, : angle of friction.

Please cite this article as: Robin, V., et al., Chemo-mechanical modelling of lime treated soils, Appl. Clay Sci. (2014), http://dx.doi.org/10.1016/
j.clay.2014.04.015

V. Robin et al. / Applied Clay Science xxx (2014) xxxxxx

(b) Above a critical lime content, yield stress does not increase anymore (Rotta et al., 2003). To account for this feature in the coupling, the interpolation function must converge to a nite value
for high concentrations in hydrates.

0.8

0.0

0.4

(%)
0.30

0.15

0.00

(%)

800

(kPa)

1600

YieldStress

To account for these two conditions, we chose a generalised logistic

function (also known as Richards's curve) of the form:

(%)

0.30



mh f mh py; min py; max py; min 

Portlandite

0.15

1
1 emh x

0.00

with:

0.1

(%)

0.2

Hydrates

0.0

(%)

0.8

Carbonates

0.4

(kPa)

0.1

0.2

0.0

1600

800

0.0

(%)

Fig. 7. Scatterplot matrix between the yield stress and the mass concentrations in cementitious compounds, portlandite, and carbonates.

the constitutive models, there is a one-to-one correspondence between

the lime content and the resulting yield stress. The mechanical results
and the chemical compositions are plotted in a scatterplot matrix
(Fig. 7) to assess the trends between the yield stress and the mass concentration in cementitious compounds, portlandite, and carbonates.
The trends between the yield stress and the portlandite, and carbonates and hydrates are given in Fig. 7. The results show a linear correlation between the yield stress and the portlandite. Hydrates also
appear to be linearly correlated to the yield stress. The production of hydrates was noted for every lime content and a one-to-one correspondence with the lime content. A second trend can be noticed between
the yield stress and the portlandite. Chemical analysis revealed that all
the portlandite has been consumed by pozzolanic reactions for low
lime-contents (b 0.5%). However, mechanical results showed that this
lime-content is enough to increase the yield stress. Therefore,
portlandite cannot be considered as a relevant parameter for a chemomechanical coupling.
Production of carbonates was only observed for concentrations
greater than 2%, but mechanical improvements can be measured for
small lime contents. Maubec (2010) has shown that the contribution
of the carbonates in the mechanical behaviour can be neglected. For
the same reasons as the portlandite, carbonates are not suitable for a
chemo-mechanical coupling. Lime can be easily carbonated in contact
with CO2 (Eq. (2)).
In light of these observations, mass concentration in hydrates is the
only parameter systematically modied by the addition of lime and presenting a one-to-one correspondence with the lime content. Since yield
stress has also a one-to-one correspondence with the lime content, a
coupling between these two parameters appears to be the most relevant choice.
In order to establish a coupling, a relationship was established between yield stress and the mass concentration in hydrates. To model accurately a lime-treated soil, the regression function must full the
following criteria:
(a) The point of lime xation describes the minimum lime-content
required to measure mechanical improvements. Thus, for low
concentrations in hydrates (b0.05%), regression function must
show a low gradient.

py,min The lower asymptote

py,max The upper asymptote
The growth rate
x mh value for which the rst derivative is maximum and
2
df
0.
2
d mh

Using a non-linear least square method, the 4 variables py,min,

py,max,, and x were calculated to determine the optimal set of
parameters (Table 7) based on the results of the isotropic tests.
For lime concentrations greater than 1%, three isotropic tests
were performed to reduce the error. The interpolation appears to
satisfactorily describe the results (Fig. 8) and full the conditions
(a) and (b).
The same procedure can be applied to link the mass concentration in hydrates with the lime content. If introduced in the Eq. (7),
this will result in a direct coupling between the amount in lime
introduced in the soil and the resulting yield stress. This correlation is allowed since there is a systematic one-to-one correspondence between all the variables, which proves that a robust
coupling between the mechanical behaviour and chemical composition exists.
4.3. Modication of the Structured Cam Clay model for application on lime
treated soils
Several studies (Leroueil and Vaughan, 1990; Liu and Carter,
2003) have shown that naturally structured soils exhibit similar behaviour to articially treated materials. Under isotropic consolidation, structured and treated soils display at yield a higher void
ratio compared to the remoulded/untreated state (Callisto and
Rampello, 2004; Coop and Atkinson, 1993). This additional void
ratio tends to disappear as the mean effective stress is increased to
nally match the behaviour of the remoulded/untreated state
(Burland et al., 1996). For drained experiments, Leroueil and
Vaughan (1990) noted that a structured material exhibits a rate of
maximum dilatation after the peak while this occurs before the
peak for a non-structured material. Based on these considerations
and the results of this study, one can assume that naturally structured soils and lime treated materials have a similar mechanical behaviour for both isotropic and shear stresses.
We propose in this paper a simple modication to the Structured
Cam Clay model (SCCM) (Liu and Carter, 2002) to account for the effects
of lime treatment on the intrinsic parameters. The SCCM was originally
designed for naturally structured soils, but the results prove that it can
be applied to lime treated specimen, provided that a modication is

Table 7
Optimal set of parameters for a logistic interpolation.
Parameter

py,min

py,max

R2

Value

100

2000

17.4

0.16

0.95

Please cite this article as: Robin, V., et al., Chemo-mechanical modelling of lime treated soils, Appl. Clay Sci. (2014), http://dx.doi.org/10.1016/
j.clay.2014.04.015

V. Robin et al. / Applied Clay Science xxx (2014) xxxxxx

2000

1.60

1.55
1500

Yield stress

(kPa)

Specific volume (-)

1000

500

1.45

1.40

Experimental Data
Interpolation
0
0.0

0.1

0.2

0.3

0.4

1.35
0.5

e e e e ei

!b

e* is the void ratio for the reconstituted soil,

ei is the additional void ratio at yield compared to the reconstituted
soil at the same stress state,
py,i is the yield stress of the structured soil, and
b is the destructuring index.
The formulation of the additional void ratio is chosen to converge to
zero for high levels of mean stress. In the case of a lime treatment, normal compression line appears to be vertically translated by a constant
void ratio, called ec. The modied equation of the additional void
ratio can be written as follows:
"

p0y;i
p0

!b #

103

Fig. 9. Isotropic compression for 0% and 1% lime treated specimens MSCC: Modied
Structured Cam Clay; ei: additional void ratio at yield; ec: vertical translation of the virgin compression line.

The suitability of this formulation for the 1% lime treated specimen

of this study is given in Fig. 9. Parameters used for the simulation are
given in Table 8.
5. Conclusion

where:

e ei ec

102

(%)

implemented to introduce the secondary normal compression line for

lime treated materials.
The SCCM models structured soils by introducing an additional void
ratio e to the Cam Clay equation of the void ratio as:
p0y;i
0

Isotropic - 0%
Isotropic - 1%
MSCC Liu and Carter
MSCC Robin et al.

Effective pressure (kPa)

Fig. 8. Logistic interpolation for a chemo-mechanical coupling mh: mass concentration

in hydrates.

1.50

p0

ec ec :

This formulation ensures the convergence toward the destructured

state (intermediate state), rather than the untreated state. This additional parameter ec can be measured from the same isotropic test
used for the determination of ei and py,i. In the case where ec = 0,
Eq. (5) reverts back to the formulation of the MSCC. For use in constitutive modelling Eq. (5) can be written in an incremental form:

b
dp0
0
0  0 b1
dv 0 ei ec  py;i  b  dp  p
:
p

This study has shown the potential of lime treatment to improve the
mechanical properties of the materials initially unsuitable for geotechnical structures. Thus, a chemo-mechanical coupling to account for the
treatment in the design is of the greatest interest.
This paper aimed rst to describe in details the effects of lime on the
mechanical behaviour of a soil and also on parameters usually considered as constant. The results have shown that there was an improvement of the mechanical parameters due to the chemical reactions
occurring after the addition of the lime. For lime contents greater than
1%, the specimens converged toward a normal compression line different from the untreated state but parallel to it. Therefore, in the framework of the lime treated soils, the mechanical parameters of treated
soils could be different from those of the untreated state.
The chemical composition of lime treated specimens has been successfully assessed using thermogravimetric and thermodifferential
analyses. The results have shown that concentrations in cementitious
compounds and portlandite increased with the lime content. They permitted to explain the mechanical results and especially the modication
of some parameters such as the angle of friction. The production of the
cementitious compounds involves several preliminary reactions
where soil particles experience dissolution due to the high pH. From
this point of view, the nature of soil is seriously altered, leading to an
evolution of the mechanical parameters. Chemical results revealed
that cementitious compounds were produced for every concentration
tested. A chemo-mechanical coupling was established using a logistic
Table 8
Parameters for the Structured Cam Clay model adapted for lime treated soils.

The rst part is the classic incremental form of the normal compression line from the Cam Clay model and describes the untreated state.
The second part adds the effects of the structure as an additional void
ratio.

MSCC parameters

b
ei
ec

Lime content
0%

0.5%

1%

2%

5%

0.0
0.0
0.0

1.8
0.027
0.0

7.2
0.066
0.046

0.129

0.159

MSCC: Modied Structured Cam Clay.

Please cite this article as: Robin, V., et al., Chemo-mechanical modelling of lime treated soils, Appl. Clay Sci. (2014), http://dx.doi.org/10.1016/
j.clay.2014.04.015

V. Robin et al. / Applied Clay Science xxx (2014) xxxxxx

function between the amount of cementitious compounds and the yield

stress. The proposed equation appears to describe accurately the observations and ensures a one-to-one correspondence between the mass
concentration in cementitious compounds and the yield stress.
A modication of the formulation of the Structured Cam Clay model
was proposed by introducing a new parameter ec describing the position of the second normal compression line. Simulations proved the
suitability of the new formulation for use in constitutive modelling.
The proposed coupling may be used to account for the treatment in
the design of geotechnical structures. However, one must be aware of
the durability of the treatment. The bonding structure resulting from
chemical reactions occurring at high pH environment and treated materials are often subjected to environmental conditions. For example, in
the event of a continuous rainfall, water ow may occur changing the
pH environment and altering the chemical equilibriums. This would result in the dissolution of the cementitious compounds and, therefore, in
the extinction of the bonding structure. It is thus important to assess the
durability for the treatment in the design of geotechnical structures.
References
Alarcon-Ruiz, L., Platret, G., Massieu, E., Ehrlacher, A., 2005. The use of thermal analysis in
assessing the effect of temperature on a cement paste. Cem. Concr. Res. 35, 609613.
http://dx.doi.org/10.1016/j.cemconres.2004.06.015.
ASTM Standard D6276, 2006. Standard Test Method for Using pH to Estimate the Soil
Lime Proportion Requirement for Soil Stabilization. pp. 14.
ASTM Standard D7181-11, 2011. Method for consolidated drained triaxial compression
test for soils. ASTM Int. 111. http://dx.doi.org/10.1520/D7181.
ASTM Standard D7263-09, 2009. Standard test methods for laboratory determination of density (unit weight) of soil specimens. ASTM Int. 17. http://dx.doi.org/10.1520/D7263-09.
Baudet, B., Stallebrass, S., 2004. A constitutive model for structured clays. Geotechnique
54, 269278.
Burland, J.B., 1990. On the compressibility and shear strength of natural clays.
Geotechnique 40, 329378.
Burland, J.B., Rampello, S., Georgiannou, V.N., Calabresi, G., 1996. A laboratory study of the
strength of four stiff clays. Geotechnique 46, 491514. http://dx.doi.org/10.1680/
geot.1996.46.3.491.
Callisto, L., Rampello, S., 2004. An interpretation of structural degradation for three natural clays. Can. Geotech. J. 41, 392407. http://dx.doi.org/10.1139/T03-099.
Chiu, C.F., Zhu, W., Zhang, C.L., 2009. Yielding and shear behaviour of cement-treated dredged
materials. Eng. Geol. 103, 112. http://dx.doi.org/10.1016/j.enggeo.2008.07.007.
Consoli, N.C., da Lopes, L.S., Prietto, P.D.M., Festugato, L., Cruz, R.C., 2011. Variables controlling stiffness and strength of lime-stabilized soils. J. Geotech. Geoenviron. Eng. 137,
628632. http://dx.doi.org/10.1061/(ASCE)GT.1943-5606.0000470.
Coop, M.R., Atkinson, J.H., 1993. The mechanics of cemented carbonate sands.
Geotechnique 43, 5367.
Cotecchia, F., Chandler, R.J., 2000. A general framework for the mechanical behaviour of
clays. Geotechnique 50, 431447.
Cuisinier, O., Masrouri, F., Pelletier, M., Villieras, F., Mosser-Ruck, R., 2008. Microstructure
of a compacted soil submitted to an alkaline PLUME. Appl. Clay Sci. 40, 159170.
http://dx.doi.org/10.1016/j.clay.2007.07.005.
Cuisinier, O., Auriol, J.-C., Le Borgne, T., Deneele, D., 2011. Microstructure and hydraulic
conductivity of a compacted lime-treated soil. Eng. Geol. 123, 187193. http://dx.
doi.org/10.1016/j.enggeo.2011.07.010.
Das, S.K., Mitra, A., Das Poddar, P.K., 1996. Thermal analysis of hydrated calcium aluminates. J. Therm. Anal. Calorim. 47, 765774.
Eades, J.L., Grim, R.E., 1966. A quick test to determine lime requirements for lime stabilization. Highw. Res. Rec. 139, 6172.
Gens, A., Nova, R., 1993. Conceptual bases for a constitutive model for bonded soils and
weak rocks. Geotech. Eng. Hard Soils-Soft Rocks 1, 485494.
Horpibulsuk, S., Liu, M.D., Liyanapathirana, D.S., Suebsuk, J., 2010a. Behaviour of cemented
clay simulated via the theoretical framework of the Structured Cam Clay model.
Comput. Geotech. 37, 19. http://dx.doi.org/10.1016/j.compgeo.2009.06.007.
Horpibulsuk, S., Rachan, R., Chinkulkijniwat, A., Raksachon, Y., Suddeepong, A., 2010b.
Analysis of strength development in cement-stabilized silty clay from microstructural

considerations. Constr. Build. Mater. 24, 20112021. http://dx.doi.org/10.1016/j.

conbuildmat.2010.03.011.
Lee, K., Chan, D., Lam, K., 2004. Constitutive model for cement treated clay in a critical
state frame work. Soils Found. 44, 6977.
Leroueil, S., Vaughan, P.R., 1990. The general and congruent effects of structure in natural
soils and weak rocks. Geotechnique 40, 467488.
Little, D.N., 1995. Stabilization of Pavement Subgrades and Base Courses with Lime. National Lime Association, Arlington.
Liu, M.D., Carter, J.P., 2002. A structured Cam Clay model. Can. Geotech. J. 39, 13131332.
http://dx.doi.org/10.1139/T02-069.
Liu, M.D., Carter, J.P., 2003. Volumetric deformation of natural clays. Int. J. Geomech. 3,
236252. http://dx.doi.org/10.1061/(ASCE)1532-3641(2003) 3:2(236).
MacKenzie, R.C., 1970. Differential thermal analysis. Fundamental Aspects, vol. 1. Academic Press London, London-New York.
Mackenzie, R.C., 1972. Differential thermal analysis. In: Academic, P. (Ed.), Applications,
vol. 2. Academic Press London, London-New York.
Malandraki, V., Toll, D.G., 2001. Triaxial tests on weakly bonded soil with changes in stress
path. J. Geotech. Geoenviron. Eng. 127, 282291.
Maubec, N., 2010. Approche multi-echelle du traitement des sols a la chaux. Etudes des
intractions avec les argiles. Universit de Nantes, France, Thse de Doctorat.
Metelkov, Z., Boh, J., Pikryl, R., Sedlov, I., 2012. Maturation of loess treated with
variable lime admixture: pore space textural evolution and related phase changes.
Appl. Clay Sci. 61, 3743. http://dx.doi.org/10.1016/j.clay.2012.03.008.
Muir Wood, D., 1991. Soil Behaviour and Critical State Soil Mechanics.
North, J.M., Becker, J.G., Seagren, E.A., Ramirez, M., Peot, C., 2008. Methods for quantifying
lime incorporation into dewatered sludge. I: Bench-scale evaluation. J. Environ. Eng.
134, 750761.
Nova, R., Castellanza, R., Tamagnini, C., 2003. A constitutive model for bonded geomaterials subject to mechanical and/or chemical degradation. Int. J. Numer. Anal. Methods
Geomech. 27, 705732.
Oliveira, P.J.V., Correia, A.A.S., Garcia, M.R., 2013. Effect of stress level and binder composition on secondary compression of an articially stabilized soil. J. Geotech.
Geoenviron. Eng. 139, 810820. http://dx.doi.org/10.1061/(ASCE)GT.1943-5606.
0000762.
Pomakhina, E., Deneele, D., Gaillot, A.-C., Paris, M., Ouvrard, G., 2012. 29Si solid state NMR
investigation of pozzolanic reaction occurring in lime-treated Ca-bentonite. Cem.
Concr. Res. 42, 626632. http://dx.doi.org/10.1016/j.cemconres.2012.01.008.
Rogers, C.D.F., Glendinning, S., 1996. The role of lime migration in lime pile stabilization of
slopes. Q. J. Eng. Geol. Hydrogeol. 29, 273284. http://dx.doi.org/10.1144/GSL.QJEGH.
1996.029.P4.02.
Roscoe, K.H., Burland, J.B., 1968. On the generalized stress-strain behaviour of wet clay.
Eng. Plast. 535609.
Rotta, G.V., Consoli, N.C., Prietto, P.D.M., Coop, M.R., Graham, J., 2003. Isotropic yielding in
an articially cemented soil cured under stress. Geotechnique 53, 493501. http://dx.
doi.org/10.1680/geot.2003.53.5.493.
Saikia, N.J., Sengupta, P., Gogoi, P.K., Borthakur, P.C., 2002. Hydration behaviour of limeco-calcined kaolin-petroleum efuent treatment plant sludge. Cem. Concr. Res. 32,
297302. http://dx.doi.org/10.1016/S0008-8846(01)00676-7.
Schoeld, A.N., Wroth, C.P., 1968. Critical State Soil Mechanics. McGraw-Hill.
Stocker, P.T., 1974. Diffusion and diffuse cementation in lime and cement stabilised clayey
soils studies of plasticity and aggregation. Aust. Road Res. 5, 5175.
Stoltz, G., Cuisinier, O., Masrouri, F., 2012. Multi-scale analysis of the swelling and shrinkage of a lime-treated expansive clayey soil. Appl. Clay Sci. 61, 4451. http://dx.doi.
org/10.1016/j.clay.2012.04.001.
Suebsuk, J., Horpibulsuk, S., Liu, M.D., 2010. Modied structured cam clay: a generalised
critical state model for destructured, naturally structured and articially structured
clays. Comput. Geotech. 37, 956968. http://dx.doi.org/10.1016/j.compgeo.2010.08.
002.
Suebsuk, J., Horpibulsuk, S., Liu, M.D., 2011. A critical state model for overconsolidated
structured clays. Comput. Geotech. 38, 648658. http://dx.doi.org/10.1016/j.
compgeo.2011.03.010.
Thompson, M.R., 1965. Shear Strength and Elastic Properties of LimeSoil Mixtures. Bureau of Public Roads /US, Washington.
Vatsala, A., Nova, R., Srinivasa Murthy, B.R., 2001. Elastoplastic model for cemented soils.
J. Geotech. Geoenviron. Eng. 127, 679687.
Wissa, A.E.Z., 1965. Effective StressStrength Behavior of Cemented Soils. (Thesis (Sc. D.))
Massachusetts Institute of Technology.
Zhu, W., Zhang, C.L., Chiu, A.C.F., 2007. Soilwater transfer mechanism for solidied
dredged materials. J. Geotech. Geoenviron. Eng. 133, 588598. http://dx.doi.org/10.
1061/(ASCE)1090-0241(2007) 133:5(588).

Please cite this article as: Robin, V., et al., Chemo-mechanical modelling of lime treated soils, Appl. Clay Sci. (2014), http://dx.doi.org/10.1016/
j.clay.2014.04.015