HOSTED BY Available online at www.sciencedirect.

com

ScienceDirect
Soils and Foundations 57 (2017) 60–79
www.elsevier.com/locate/sandf

Physicochemical and consolidation properties of compacted lateritic

soil treated with cement
Emmanuel Mengue a,b,⇑, Hussein Mroueh a, Laurent Lancelot a, Robert Medjo Eko b
a
Laboratory of Civil Engineering and geo-Environment (LGCgE) - Polytech’Lille, University Lille 1 Sciences and Technologies,
Villeneuve d’Ascq 59655, France
b
Laboratory of Engineering Geology, Department of Earth Sciences, Faculty of Science, University of Yaoundé I, PO Box 812, Yaoundé, Cameroon

Received 14 February 2016; received in revised form 7 October 2016; accepted 25 October 2016
Available online 9 February 2017

Abstract

The consolidation of a fine-grained lateritic soil, treated with compound Portland cement (CEMII/BM 32.5 N) up to 9% by weight of
the dry soil and prepared at three different molding water contents (xDRY, OMC, and xWET), was investigated by means of a one-
dimensional consolidation test. The physicochemical and microstructural properties of the compacted lateritic soil-cement mixture were
investigated using Raman spectroscopy, polarized light microscopy (PLM), scanning electron microscopy (SEM), and pH measurement.
The results show that cement admixtures resulted in the formation of tobermorite, afwillite, ettringite, portlandite, and calcite. However,
tobermorite and afwilite, which are calcium silicate hydrates (CSH) whose mechanisms of formation are the pozzolanic and alkali silica
reactions, appear from 6% added cement. The fixing point of the pH (12.4) is also obtained from 6% added cement. It is the threshold
value at which the material begins to develop an adequate mechanical performance. In general, as the content of cement in the soil is
increased, the yield stress increases from 1 to 3 times in comparison to untreated soil. For effective vertical stresses smaller than the
cement-induced yield stress, the primary consolidation process for specimens treated with cement is 2–7 times faster than that for spec-
imens not treated with cement, while for effective vertical stresses higher than the cement-induced yield stress, the primary consolidation
process for specimens treated with cement is about 0.5–1.5 times faster than that for specimens not treated with cement. Permeability and
secondary compression are reduced 1–9 times and 2–11 times that of the untreated samples, respectively. These changes are attributed to
the creation of chemical bonds and aggregation that accompany the addition of cement. The results also show that it would be desirable
for soil samples to be prepared at the dry side of optimum (xDRY) when the optimum moisture content (OMC) is not reached at the site.
These results indicate that significant and desirable changes in soil behavior can be achieved when the soil is admixed with CEM II/BM
32.5 N cement, thus providing the possibility of using the tested lateritic soil in road construction.
Ó 2017 Production and hosting by Elsevier B.V. on behalf of The Japanese Geotechnical Society. This is an open access article under the CC BY-
NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Keywords: Lateritic soil; Cement mixes; Microstructure; Consolidation behavior

1. Introduction used for highway geotechnic, primarily in roadway

structures (Mesbah et al., 1999). With the evolution of
Until recently, in many parts of the world, especially in knowledge, needs, and the recent consideration of environ-
tropical and subtropical regions, gravelly soils have been mental constraints, fine-grained lateritic soils are increas-
ingly used in road construction to reduce costs. From a
geotechnical perspective, fine-grained lateritic soils (silt or
Peer review under responsibility of The Japanese Geotechnical Society.
⇑ Corresponding author at: Laboratory of Civil Engineering and geo- clays) are generally associated with high compressibility,
Environment (LGCgE) - Polytech’Lille, University Lille 1 Sciences and high rates of creep, and sometimes lower permeability
Technologies, Villeneuve d’Ascq 59655, France. Fax: +33 7 60 93 89 87. characteristics, which increase the risk of inadmissible
E-mail address: [email protected] (E. Mengue).

http://dx.doi.org/10.1016/j.sandf.2017.01.005
0038-0806/Ó 2017 Production and hosting by Elsevier B.V. on behalf of The Japanese Geotechnical Society.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
 E. Mengue et al. / Soils and Foundations 57 (2017) 60–79 61

settlements and/or foundation failure. As a result, con- The potential use of this amended material for road con-
struction on these soils poses significant challenges to struction is discussed.
geotechnical engineers. The viability of civil engineering
works consisting of lateritic soils is proven to be dependent 2. Materials and test methods
on the chemical and mineralogical compositions as well as
the mechanical characteristics of these raw materials 2.1. Materials
(Idrissa, 1994).
Several methods are generally used to modify and The soil samples used for this study were obtained from
improve the geotechnical properties of these problematic a borrow pit at Zoétele, in the Southern part of Cameroon
soils. The methods include densifying treatments (such as (located between latitudes 11° 450 and 11° 550 N and longi-
compaction or preloading), pore water pressure reduction tudes 3° 050 and 3° 100 E) using the method of disturbed
techniques (such as dewatering or electro-osmosis), the sampling (Fig. 1). The first 0.60 m of soil was removed to
bonding of soil particles (by ground freezing, grouting, get rid of the organic soil layer. Laterite was then taken
and chemical stabilization), and the use of reinforcing ele- from the A horizon whose thickness was approximately
ments such as geotextiles and stone columns (Ismaiel, 2006; 2.5 m. The soil samples were sealed in bags and then trans-
Bobet et al., 2011). In almost all tropical and subtropical ported to the laboratory. Previous geological and pedolog-
areas where residual lateritic soils abound, soil-cement mix- ical studies show that the samples taken belong to the
ing (chemical stabilization) is the method widely used to group of ferralitic tropical soils derived from acid igneous
improve the engineering properties of poor laterites and metamorphic rocks (Tchameni, 1997). The mineralog-
(Kazemian and Huat, 2009; Oyediran and Kalejaiye, ical and chemical compositions of the concerned lateritic
2011). Soil-cement mixing (cement stabilization) has also soil are summarized in Table 1. As can be seen in this table,
been used with success for many geotechnical engineering the raw sample is mainly composed of quartz, kaolinite,
applications in sub-Saharan Africa, such as pavement goethite, hematite, gibbsite, and traces of feldspar (oligo-
structures, roadways (the construction of highway clase). These mineral phases are those commonly present
embankments), building foundations, channel and reser- in laterites (Millogo et al., 2008). Nevertheless, the absence
voir linings, irrigation systems, water lines, and sewer lines, of swelling clay species, such as smectites, is noted. The
to avoid damage due to the settlement of soft soil or to the sample is composed almost exclusively of silica
swelling action (heave) of expansive soils (Bagarre, 1990; (SiO2 = 36.47%), iron (Fe2O3 = 29.13%), and alumina
Joel and Agbede, 2010). (Al2O3 = 19.34%). The other elements (Ti, Mn, Na, Ca,
Although today there are high quality data on the engi- K, and P) are negligible, while alkali and alkaline earth
neering properties of treated lateritic soils, little attention is cations are totally absent. The S/R ratio, defined according
paid to the investigation of their microstructures or to the to the SiO2 content, the Al2O3 content, and the Fe2O3 con-
one-dimensional consolidation behavior (primary consoli- tent, was used to determine the type of lateritic soil. This
dation and secondary consolidation) of treated fine- ratio is defined by the following relationship (Autret, 1980):
grained lateritic soils (Millogo et al., 2008; Li et al.,
S SiO2
2012). In addition, in almost all of the above-mentioned ¼ with RS O3 ¼ Fe2 O3 þ Al2 O3 ð1Þ
R RS O 3
works, the authors limited themselves to preparing their
samples only at the optimum moisture content. However, in which S/R is the Sesquioxyde Ratio. The Sesquioxide
it is difficult to reach the optimum moisture content on site; Ratio (S/R = 0.69 < 1.33) shows that the raw sample
this is why it is desirable to vary the molding water content would be a real laterite (Autret, 1980; Bagarre, 1990).
in order to approach the best condition of the reality at the The relatively high iron oxide content (29.13%) partly
site (CEBTP, 1984). Compound Portland cements (CEM explains the red color of the raw sample.
II) were only rarely used as stabilizers in these previous The particle size distribution of the lateritic soil used in
studies. this study (Fig. 2) shows that the sample consists of
This paper presents a series of experimental studies on 18.8 wt.% of gravel, 32.4 wt.% of sand, 14 wt.% of silt,
the feasibility of using local lateritic soil treated with and 34 wt.% of clay. The values of some geotechnical prop-
cement in road construction. In this study, a compound erties (determined according to the French standards) of
Portland cement (CEM II/BM 32.5 N) produced in the sample are gathered in Table 2. It is a reddish lateritic
Northern France was added at various weight propor- soil, classified as a fine-grained soil (A-7-5) in the
tions, to a fine-grained lateritic soil prepared at different AASHTO (American Association of State Highway and
molding water contents. Geotechnical and microscopic Transportation Officials) soil classification system
testing of the soil samples were performed to assess the (AASHTO 1986), inorganic silts (MH) in the USCS (Uni-
characteristics of the materials before and after the cement fied Soil Classification System) soil classification (ASTM
addition. The tests included Raman spectroscopy, polar- 1992), and sandy clay or silt (A2) in the GTR (French
ized light microscopy (PLM), scanning electron micro- ‘‘Guide des Terrassements Routiers”) soil classification
scopy (SEM), pH measurement, compaction tests, (GTR 1992). According to CEBTP (1984) and Bagarre
workability tests, and one-dimensional consolidation. (1990), the geotechnical properties of the studied sample
62 E. Mengue et al. / Soils and Foundations 57 (2017) 60–79

Fig. 1. Origin of fine-grained lateritic soil.

Table 1
Chemical and mineralogical characteristics of lateritic soil.
Sample Values Standard methods
Chemical characteristics SiO2 33.47 DIN 51418-1
Fe2O3 29.13
Al2O2 19.34
TiO2 0.19
K2O 1.84
Na2O 1.91
CaO 1.41
MgO 0.84
P2O5 0.43
Sesquioxyde ratio S/R = SiO2/RsO3 0.69 Autret (1980)
Mineralogical characteristics Major Quartz Raman spectroscopy
Kaolinite
Secondary Hématite
Goethite
Gibbsite
Illite
Feldspath

are poor and unsuitable for base and subbase courses in Africa and Cameroon in particular. On the other hand,
road construction. compound Portland cement is increasingly being used in
The selected cement in this study is a compound Port- the treatment of soil instead of Portland cement (CEM
land cement (CEM II/BM 32.5 N) whose main constituents I). Its chemical composition and some of its specifications
are Portland clinker, limestone, granulated blast furnace are presented in Table 3.
slag, and siliceous fly ash. The clinker content is between
65% and 79%. It is produced in Northern France. The 2.2. Test methods
choice of this cement was motivated by the fact that on
the one hand, it has substantially the same characteristics Before obtaining the soil-cement mixture used in this
as the CPJ 35 cement commonly used in sub-Saharan study, the lateritic soil was first oven-dried at 50 °C for
 E. Mengue et al. / Soils and Foundations 57 (2017) 60–79 63

Stone Gravel Sand Silt Clay

100

90
(2 ; 81.2)
80

Percentage passing (%)

70

60

50
(0.063 ; 48.8)
40 (0.002 ; 34)

30

20

10

0
10 1 0.1 0.01 0.001
Particle size (mm)

Fig. 2. Grain size distribution of tested soil.

Table 2
Engineering properties of lateritic soil.
Properties Values Standard method
Particles size distribution
Gravel content >2 mm (wt.%) 18.80
Skeleton <2 mm (wt.%) 81.20 NF P94-056 and NF P 94-057
Mortar <0.425 mm (wt.%) 55
Fine particles <80 lm (wt.%) 51.10
Clays <2 lm (wt.%) 34
Atterberg limits
Liquid limit LL (%) 53 NF P94-051
Plasticity limit PL (%) 30
Plasticity index PI (%) 23
Blue methylene value MBV (g/100 g) 2.58 NF P 94-068
Natural water content x (%) 15.50 NF P 11-300
Specific gravity Gs 2.65 NF P 94-054
Compaction characteristics
Optimum moisture content OMC (%) 17.80 NF P 94-093
Maximum dry density MDD (g/cm3) 1.79
California bearing ratio CBR (%) 18 NF P94-078
Soil solution pH <2 lm (%) 5.29 NF ISO 10390
Soil classification
American Association AASHTOa A-7-5 AASHTO M320
Unified Soil USCSb MH USCS ASTM
French guide GTRc A2 NF P11 300
a
AASHTO: American Association of State Highway and Transportation Officials.
b
USCS: Unified Soil Classification System.
c
GTR: French guide road earthworks.

24 h to significantly reduce the amount of water contained tor mold using a modified mechanical standard Proctor
in the soil. In the mixing phase, soil and water were mixed hammer. After modified Proctor compaction, specimens
at three water contents (xDRY = OMC  2%, OMC, and were extracted immediately using a core sampler (52 mm
xWET = OMC + 2%), and stored in hermetic plastic bags in diameter and 210 mm in height). For the untreated soil,
for 24 h for homogenization. Thereafter, the cement (3%, cores were immediately transferred into the oedometer
6%, and 9% by dry weight of soil) was added to the soil- rings (50 mm in diameter and 20 mm in height) and tested.
water mixture and full mixing was carried out until a For the treated soil, extracted samples were waxed and
homogenous wet soil-cement mixture without lumps was stored in plastic bags at room temperature (20 °C). After
obtained. After mixing, the soil was compacted into a Proc- the corresponding curing times (7, 28, and 60 days), the
64 E. Mengue et al. / Soils and Foundations 57 (2017) 60–79

Table 3 and a Scanning Electron Microscope (SEM) equipped with

Chemical composition of cement. a micro-analyzer system by Energy Dispersive Spec-
Chemical composition Values (%) Specification (%) EN197-1 troscopy (EDS). PLM analyses were carried out with an
CaO 56.2 – OPTIKA B-165PolB binocular PLM. This PLM is
SiO2 19.8 – equipped with two Polaroid filters (polarizer and analyzer)
Al2O3 7.1 – and a rotating (platinum) slide. The type of polarized light
Fe2O3 3.6 –
MgO 1.2 –
used here is light by transmission. The SEM observations
Na2O 0.38 – were done on pieces taken from the above-mentioned cylin-
K2O 0.69 – drical samples using an EVO MA 10 Zeiss scanning elec-
SO3 2.9 63,5 tron microscope. This microscope was equipped with a
Cl 0.06 60,10 micro-analyzer system by EDS. It allows for the chemical
Loss on ignition 6.5 –
Insoluble residue 5.4 –
analysis (punctual or distribution maps) of various nonde-
structive samples. The PLM examinations were done on
specimens were placed in oedometer rings and subjected to the same thin sections as those used during Raman spec-
consolidation tests. A part of the compacted and cured troscopy, while the SEM examinations were done on the
specimens was used for the analysis of the microstructure. samples rendered conductive by coating them with a thin
layer of conductive material.
Thin sections were previously prepared according to the
2.2.1. Mineralogical and pH tests dehydration standard method for samples ‘‘by the water/
The mineralogical composition of the sample before and acetone exchange‘‘ (Hanrion and Delvigne, 1976). At the
after treatment was determined by means of a LabRAM end of this process, a test with petroleum distillate was con-
HR 800 (HORIBA JOBIN YVON) Raman Spectrometer, ducted to ensure that there was more water. The samples
controlled by LabSpec software. The device was coupled to were then impregnated with crystic polyester resin and
an XYZ motorized stage to allow for the performance of stored to cold-cure them in a ventilated cabinet for up to
Raman mappings according to these three axes. Excited one month. When the resin has set, the samples were
by monochromatic light (an Argon ion laser at 514 nm) moved to an oven to be hard-cured. The cured block was
at a power of 20 mW, the specimen emitted scattered light then sliced and bonded to a glass slide. The excess sample
showing different frequencies than the laser light in the was cut away and the remaining sample was lapped leaving
spectrum (Raman bands). The differences in frequencies a 30-lm thick sample attached to the glass slide.
(Raman-shift) contain vibrational information on the The SEM examinations were done on the samples
molecules in the specimen, and that of its composition deposited on a carbon layer and sputter coated with a
and structure. The readings were done with an objective 4-nm-thick gold palladium layer to make them conductive.
to a magnification of 50. The obtained Raman spectrums The material was metallized using a device called a ‘‘sputter
provided qualitative and quantitative information, such as coater‘‘.
peak position, intensity of peak, and shifts in the position
of the peak and the width the structural disorder. The mea- 2.2.3. Physical and compaction tests
surements were carried out in the range in wave number of The grain size distribution was realized by dry sieving as
100–1200 cm1. The analysis was made on polished thin well as by sedimentation, according to NF P 94-056 (1996)
sections (4.5  3 cm), prepared before and after the treat- and NF P 94-057 (1992) standards, respectively. The for-
ment of the samples. These thin sections were obtained mer method was used for the classification of coarse parti-
after the samples had been dehydrated and then impreg- cles (>80 lm in size), while the second one was applied for
nated with resin under a vacuum. After hardening, the tiny grains (<80 lm). Atterberg limits, the methylene blue
samples were sliced to a thickness of 30 lm and then values, were realized in accordance with NF P 94-051
polished. (1993) and NF P 94-068 (1998) standards.
The pH before and after the treatment was measured Tests involving the moisture-density relationship were
using the pH meter according to the ratio 2/5 (NF ISO carried out on both the untreated and the treated soil
10390, 2005). A quantity of 2 g of soil sample was sus- according to the NF P94-093 (1999) standard. The com-
pended in a quantity of 5 ml of deionized water. The mix- pactive effort utilized throughout the tests was the Modi-
ture was stirred with a glass rod. The mixture was then fied Proctor. This compactive effort consisted of the
allowed to stand for 30 min, taking care to shake it 5 or energy derived from a 4535 ± 5 g rammer falling through
6 times during this period. Then, the pH was measured 457 ± 2 mm onto five layers, each receiving 25 blows.
with the pH meter. Dynamic compaction at the Modified Proctor energy is
the conventional energy level commonly used in the region.
2.2.2. Polarised light microscopy and scanning electron The workability period of a cement-treated material is
microscope tests the time during which the material can be put in place
Microstructures of untreated and treated specimens before that binder setting makes compaction impossible.
were examined using Polarised Light Microscopy (PLM) The workability period (WP) is defined as the time
 E. Mengue et al. / Soils and Foundations 57 (2017) 60–79 65

corresponding to a relative decrease in dry density of 2% the top and bottom of the sample. For the primary consol-
(Eq. (2)). idation phase, the compression readings were recorded
cOPM  ct between 15 s and 24 h during the loading stage for each
¼ 2% ð2Þ incremental load. For the secondary compression phase,
cOPM
the compression readings for each stage of loading were
where cOPM is the Proctor density obtained after the imme- three days.
diate compaction and ct is the Proctor density, measured The saturation of both untreated and cement-treated
after time. soil specimens was conducted according to the principle
The workability period (WP) was determined by the and methodology developed by Magnan et al. (1985). A
workability Proctor test developed in 1972 at the LR labo- back pressure of 100 kPa and a holding pressure of the pis-
ratory of Bordeaux (Messou, 1980). The test was con- ton in contact with the sample of 120 kPa were always used
ducted to differentiate the lateritic soil compaction for saturation. The filling of the interstitial water in the
treated with different doses of cement (3%, 6%, and 9% oedometer cell comprising the ring + sample was accompa-
cement), during increasing periods of time (0.5, 1, 2, 3, nied by the application on the piston of the oedometer,
and 4 h), always adopting the same compaction energy, with a slightly higher pressure than that of the back pres-
for the endogenous curing of the specimens at an ambient sure (The values of the pressure on the piston fell in the
room temperature (20 °C). The results are shown in a range of 20–120 kPa). The applied back pressure must
graph ((cOMC  ct)/cOMC) for the function of time (t). ensure the almost perfect saturation of the specimen. The
values of the back pressure fell in the range of 10–
2.2.4. One-dimensional consolidation test 100 kPa. The saturation device (Fig. 3) consisted of a plex-
The incremental loading oedometer test method was iglass cylinder filled with demineralized water, a Bishop
chosen in this study because of its simplicity and wide- type of graduated buret interposed on drainage circuits
spread use for the study of soil compressibility. The test connected to the against-pressure, and a pressure-
was performed on a specimen, 50 mm in diameter and applying system (called the controller of pressure and vol-
20 mm in height, placed in a rigid cylindrical stainless steel ume). The all being connected to the oedometric cell, the
ring in accordance with NF P94-051 (1997). The device value of the necessary back pressure, was determined as
applies a vertical axial force on the specimen, keeping it follows:
saturated in an oedometric cell for the entire duration of The back pressure was gradually increased by noting the
the test. The load was applied through calibrated weights volume of the deaerated water entering the soil specimen.
on a lever system. The steps of applied vertical stress during When the volume of water entering the sample became neg-
consolidation were 25, 50, 150, 250, 500, and 750 kPa (pre- ligible, while increasing the back pressure, the saturation
cision = ±0.1). Specimen drainage was permitted through was considered to be satisfactory and this same value for

Fig. 3. Control device for saturation of specimen (Magnan et al., 1985).

66 E. Mengue et al. / Soils and Foundations 57 (2017) 60–79

the back pressure was maintained throughout the oedome- soil-cement material. Calcium hydroxide (Ca(OH)2) and
ter test. Once the saturation of the specimen was complete, calcium carbonate (CaCO3) provide the soil-cement mate-
a few hours (2 h) were left to pass before starting the test in rial. Interesting basic properties, calcium silicate hydrates
order to allow the interstitial gases to dissolve more com- (CSH), the major constituents of Portland cements, are
pletely in the pore water. responsible for the main properties of cohesion and the sus-
The methods of Casagrande, Taylor, Mesri and God- tainability of the soil-cement material.
lewski, and Buisman (Holtz and Kovacs, 1981; Mesri The Raman spectra show an increase in ettringite, Ca
et al., 1997; Al-Shamrani 1998) were used to determine (OH)2, CaCO3, and CSH with an increasing cement con-
the parameters of consolidation and permeability, such as tent. The obtained CSH (tobermorite and afwillite) reveal
the yield stress (r0y ), the primary compression index (Cc), two formation mechanisms, namely, the pozzolanic reac-
the coefficient of primary consolidation (Cv), the coefficient tion for tobermorite and alkali silica reaction for afwillite.
of permeability (kv), the secondary compression index (Ca), The CSH (tobermorite and afwillite) appear from 6%
Ca/Cc ratio, and the secondary settlement (Ss). added cement and continue to grow up to 9% added
cement. This proves that the hydration, pozzolanic, and
alkali silica reactions increase as the cement content and
3. Results and discussions curing time increase.
The microstructure seen at low zoom for both the
3.1. Mineralogical and microstructural characterization untreated and the cement-treated lateritic soil is shown in
Figs. 5 and 6. Fig. 5 shows the thin sections of the
The Raman spectra of the untreated and cement-treated untreated and the cement-treated lateritic soil, observed
lateritic soil in the 100–1200 cm1 region are shown in in PLM under transmission light, while Fig. 6 shows the
Fig. 4. Cement additions gave way to the formation of sil- SEM micrographs of the untreated and the cement-
icate (ettringite) observed at band 1155 cm1, calcium car- treated lateritic soil.
bonate (calcite) observed in bands 544 and 1087 cm1, As shown in Figs. 5a and 6a, the micrograph of the
calcium hydroxide (portlandite) observed at band untreated lateritic soil exhibits a fairly open type of
261 cm1, and calcium silicate hydrate (afwillite and tober- microstructure, with numerous quartz grains relatively well
morite) observed in bands 393 and 706 cm1, respectively. calibrated and with fine particles (kaolinite) assembled in a
These new minerals formed as a result of the addition of dispersed arrangement, leaving many voids 50–200 lm in
cement in the lateritic soil and play an important role in dimension.
improving the mechanical properties of the latter. The for- An examination of the micrograph of the treated sample
mation of ettringite crystals (Ca6Al2(SO4)3(OH)l226H2O) (Figs. 5b and c, and 6b and c) allowed for the observation
is at the origin of the beginning of the stiffening of the of the changes in the microstructure of the matrix by the

Major Phases:
Qz - Quartz Ka - Kaolinite
Ka Qz Ge - Goethite Hm - Hematite
Gb - Gibbsite Fd - Feldspar
Pr - Portlandite Af - Afwillite
Ka Cl - calcite Tb - Tobermorite
Il - illite Et - Ettringite
Ge Ge Ka
Ka Hm Hm ka
Qz Pr Hm Af Cl Il Fd
Ka Tb
Ka Qz Ka Hm
Raman Intensity

Gb Gb Cl Et Ge
(A)

Qz
(B)

(C)

(D)

100 200 300 400 500 600 700 800 900 1000 1100 1200
Wavenumber (cm-1)
(A) 9% cement, (B) 6% cement, (C) 3% cement, and (D) untreated soil
Fig. 4. Raman spectra of untreated and cement-treated lateritic soil in 100–1200 cml region.
 E. Mengue et al. / Soils and Foundations 57 (2017) 60–79 67

(a) (a) Large openings

Kaolinite

Large openings

Less compact
structure

Quartz grains

(a). Untreated lateritic soil (a). Untreated lateritic soil

(b) (b)

Large openings
Structure very few open

Beginning of particle
accumulation
Quartz grains

(b). 3% cement-treated lateritic soil (b). 3% cement-treated lateritic soil

(c) (c)
Kaolinite

Tighter
structure
Compact structure

Quartz grains

Small pores

(c). 6% cement-treated lateritic soil (c). 9% cement-treated lateritic soil

Fig. 5. Thin sections Raman of untreated and cement-treated lateritic soil Fig. 6. Micrograph of untreated and cement-treated lateritic soil (low
(low zoom). zoom).
68 E. Mengue et al. / Soils and Foundations 57 (2017) 60–79

(a) (b)

Fine reticulated
Macropores laterite-cement structure

Agglomerate

(a). Untreated lateritic soil (b). 3% cement-treated lateritic soil

(c) (d)

High reticulation of
Micropores
laterite-cement structure
with opening

Floculated laterite-ciment
cluster

(c). 6% cement-treated lateritic soil (d). 9% cement-treated lateritic soil

(e)
Platy laterite-cement
cluster

(e). 9% cement-treated lateritic soil

Fig. 7. Micrograph of untreated and cement-treated lateritic soil (higher zoom).

presence of a more compact structure. As can be seen in the cement content increases to 6% and 9%
Figs. 5b and 6b, low cement content (3%) results in an open (Figs. 5c and 6c), the soil structure is tighter than that of
structure, with the beginning of particle accumulation. As the untreated sample and the agglomeration of particles
 E. Mengue et al. / Soils and Foundations 57 (2017) 60–79 69

becomes more important. Pores smaller than 50 lm in The increase in the degree of reticulation can be attributed
diameter can be observed. to the increase in the amount of CSH (tobermorite and
With a higher zoom, the tightly-knit organization afwillite), which is reticular in structure. This is consistent
between the fine particles (kaolinite) and the quartz grains with the results of the Raman analyses discussed earlier.
is clearly visible (Fig 7a, untreated soil). The fine particles Referring to some authors (Locat et al., 1990; Chew
are in fact aggregated and form a film that coats the et al., 2004; Kamruzzaman et al., 2009), the flocculated
larger-size quartz particles (measuring in the tens of structure can be attributed to the cation exchange process,
micrometers). This association thus constitutes an agglom- which results in the Ca2+ ions replacing the Na+ or K+
erate. This is in agreement with findings made by Lemaire cations.
et al. (2013). Moreover, the smaller pores are visible at the To confirm the presence of minerals found and the
surface of these agglomerates and make up the porosity, findings made below on the untreated and cement-
qualified as ‘‘intra-agglomerates”. treated lateritic soil, the microstructure was also analyzed
With a higher zoom (Fig. 7b–e), the images of cement- by using back-scattering SEM coupled EDS. The back-
treated lateritic soil always show a gel coating the agglom- scattering SEM coupled EDS performed on untreated
erates. As can be seen, at a 3% cement content (Fig. 7b), lateritic soil showed a heterogeneous structure consisting
some signs of reticulation appear. As the cement content of isolated and irregular particles of quartz (Fig. 8b)
increases to 6% and 9% (Figs. 7c and d), the flocculated coated or not coated with fine particles of kaolinite
nature of the structure becomes more evident, with treated (Fig. 8a). As can be seen on the EDS spectrums of
laterite particle clusters interspersed by large openings. At untreated lateritic soil, high intensities of aluminum
the same time, the platiness of the structure becomes less (Al), silicon (Si), followed by a lower level of iron (Fe)
evident and the degree of reticulation appears to increase. are evident (Fig. 8a; Zones A and B). The presence and

(a) AB

(b) CD

Fig. 8. SEM micrographs and EDS analyses relevant to untreated samples.

70 E. Mengue et al. / Soils and Foundations 57 (2017) 60–79

(a) AB

(a). 3% cement treated lateritic soil

(b) CD

s
D

(b). 6% cement treated lateritic soil

(c) EF

F s

(c). 9% cement treated lateritic soil

Fig. 9. SEM micrographs and EDS analyses relevant to treated samples.

intensities of these elements are consistent with the chem- quartz (Fig. 8b; Zones C and D). However, in lower
ical composition of kaolinite (Abdulhussein Saeed et al., cement-content samples (3% cement), there is the appear-
2014). On the other hand, the increase in the Si peak ance of small particles, which consist mainly of calcium
and the decrease in the Al peak indicate that this is and sulphur (Fig. 9a; Zones A and B). These might
 E. Mengue et al. / Soils and Foundations 57 (2017) 60–79 71

correspond to ettringite which could have developed due ments. These results show that CEM II/BM 32.5 N cement
to the aqueous reaction between gypsum and tricalcium without any further additions may be used in the treatment
aluminate (C3A (3CaOAl2O3)) (Millogo et al., 2008). of soil and produces substantially the same performance as
The increase in the cement content (6% and 9%) led to CEMI cement or lime commonly used in the existing liter-
the occurrence of sulphur-calcium rich zones comprised ature (Feng et al., 2001; Santamarina et al., 2001; Chew
of portlandite and/or calcite, ettringite, tobermorite, and et al., 2004).
afwillite (Fig. 9b and c; Zones C-F). Referring to the find- The performances obtained could also be explained by
ings of some authors (Yu et al., 1999; Mitchell et al., 1998; the high calcium content of the cement shown in Table 3,
Millogo et al., 2008), the latter CSH could have originated indicating that its soil stabilization potential is high. This
from the hydration of the components of cement (C3S and is especially true for lateritic soils under acidic pH condi-
C2S (2CaOSiO2)), but its partial formation from the poz- tions (Table 2). The formation of bonds through soil pH
zolanic reaction involving clay minerals and portlandite is elevation is expected initially, followed by cation exchange.
plausible. The previous researches (Wang and Siu, 2006) have shown
The pH of the sample after the cement treatment was that the higher pH converts the positive charge to negative,
determined to find the fixing point. It is the threshold value leading to changes in fabric of the cement-treated laterite
at which the material begins to develop an adequate (i.e., the formation of a flocculated laterite-cement matrix).
mechanical performance. In the literature, Eades and The consequence of this is that a closely packed structure is
Grim (1966), cited by Feng et al. (2001), set this value to obtained. The conclusions are consistent with the results
12.4. obtained in this study (Fig. 7) where one can observe a floc-
The obtained pH values of untreated and cement- culated structure from 6% added cement.
treated lateritic soil at different dosages are shown in
Fig. 10. The figure shows a significant increase in pH
between 0% and 6% added cement. Between 6% and 9% 3.2. Compaction characteristics and period of workability
cement, the pH continues to rise, but very little. It may also
be noted that the value of 12.4 is reached from 6% cement; Fig. 11 presents the dynamic compaction curves of soil
the required value for obtaining a favorable environment and soil-cement mixtures with 3%, 6%, and 9% added
for the development of cement hydrates, and thus, the cement. Soil-cement mixtures showed higher values of
development of an adequate mechanical performance. maximum dry density than the untreated soil, while the
Obtaining the fixing point from 6% added cement corrob- optimum molding water content varies little.
orates the above-obtained results by SEM and Raman For instance, the maximum dry density MDD increased
spectroscopy of the treated soil, where a significant and from 1.79 g/cm3 for natural (untreated) soil up to 1.91 g/
continuous training of CSH to 6% and 9% of added cement cm3 for the 9% cement-treated sample. Meanwhile, the
was observed (Figs. 4–8). The fixing point being reached optimum moisture content, OMC, decreased from
from 6% cement is a possible reason for the treatment from 17.80% (untreated) to 17.50% for 6% cement-treated soil
6% cement being more efficient and allowing the use of a and then increased to 18% for 9% cement. A similar trend
laterite soil-cement mixture in the layers of road pave- has been reported for calcareous soils of Saudi Arabia (sab-
kha soils), and some lateritic soils from southwestern Nige-
ria, stabilized with CEM I cement (Al-Amoudi, 2002;
14 Oyediran and Kalejaiye, 2009).
The increase in maximum dry density with the cement
13 addition would be attributed to the fact that the cement in
12.4
12 12.34
19 1.92
11
18.8 OMC MDD
1.9
10 18.6
pH

18.4 1.88
MDD (g/cm3)

9
OMC (%)

18.2 1.86
8 18
17.8 1.84
7
17.6 1.82
6 17.4
1.8
17.2
5
17 1.78
0 2 4 6 8 10 -1 0 1 2 3 4 5 6 7 8 9 10
Cement content (%) Cement content (%)

Fig. 10. Variation in pH with cement content; error bars show standard Fig. 11. Variation in Optimum Moisture Content (OMC) and Maximum
deviations. Dry Density (MDD) against cement addition.
72 E. Mengue et al. / Soils and Foundations 57 (2017) 60–79

Table 4 than that which occurs during the first 7 days. However, on
Obtained values of workability period. the dry side of the OMC (Fig. 12a), during the first 7 curing
% Cement xDRY (mn) OMC (mn) xWET (mn) days, the void ratio generally increases between 3% and 6%
3 145–162 159–177 138–163 cement treatment before decreasing, compared with the
6 150–163 175–186 148–173 untreated laterite. This phenomenon may be related to
9 164–173 187–193 161–174 the initial particle size distribution of the soil-cement mix-
ture and the pozzolanic reaction observed during testing
SEM.
the presence of water tends to lubricate the soil particles,
The most notable effect of the treatment on the compres-
thereby resulting in a denser packing during the com-
sion behavior observed in Figs. 12 and 13 is the increase in
paction process. Moreover, the fine cement particles tend
the yield stress (r’y indicated with arrows on the plots) of
to occupy the voids between the soil particles, resulting in
the treated lateritic soil, for all the tested water contents
a denser soil matrix. The slight decrease in OMC recorded
and curing times, compared with the untreated laterite.
in the cement-treated soil may be due to self-desiccation
For example, in the case of the OMC (Fig. 13b), the yield
of the water used. When no water movement to or from
stress of 120 kPa for the untreated lateritic soil, after treat-
the cement paste is permitted, the water is used up in
ment with 3%, 6%, and 9% cement by weight of dry soil,
the hydration reaction until too little is left to saturate
the yield stress (r0y ) increases to 125 kPa, 225 kPa, and
the solid surfaces; and hence, the relative humidity within
250 kPa, respectively, for 7 curing days and 200 kPa,
the paste decreases. The above-described process might
235 kPa, and 350 kPa for 60 curing days. In general, the
have affected the reaction mechanism of the stabilized
yield stress of the cement-treated soil increases from 1 to
lateritic soil.
3 times compared to the untreated soil. This increase in
Given the required specifications for the maximum dry
the yield stress is due to the effect of structuration which
density of soils in road construction in tropical countries
creates a cementation bond for the treated clay particles
(CEBTP 1984), the obtained values for the 6% and 9%
(Kamruzzaman et al., 2009).
cement-treated soils allow their use as subbase (MDD
The obtained results also show that, for all cement con-
should be at least 1.9 g/cm3). These results indicate that
tents (3%, 6%, and 9%) and curing times (7 days, 28 days,
the use of CEM II/BM 32.5 N cement as a stabilizing agent
and 60 days) tested, the yield stress of the specimens pre-
makes it possible to employ the tested lateritic soil in road
pared from the dry side of optimum (xDRY) is 1–1.5 times
construction.
stronger than that of the specimens prepared from the wet
The obtained values for the period of workability (WP)
side of optimum (xWET). It may be noted that the sole use
of the cement-treated lateritic soil are gathered in Table 4.
of compound Portland cement (CEM II/BM 32.5 N) as a
We can see that the periods of workability are between
stabilizing agent for this type of lateritic soil, reduces the
145–162 mn and 187–193 mn for the three cement contents
void ratio and increases the yield stress; and therefore,
and for all molding water contents tested. However, the
allows the soil to withstand increased pressure without set-
lowest WP values are obtained for the treated samples with
tling at higher cement contents, especially when the mate-
3% cement and prepared on the dry side of optimum
rial is prepared from the dry side of optimum (xDRY)
(xDRY), indicating that the treated soil at this cement con-
and the optimum moisture content (OMC). These results
tent would be more sensitive to the setting of the binder
show that the tested material is a little more resistant on
(cement) compared to the other samples treated at 6%
the side dry than on the wet side and that it would be desir-
and 9% of cement and prepared both on the wet side of
able for the soil sample to be prepared at xDRY when the
optimum (xWET) and at the optimum moisture content
OMC is not reached at the site.
(OMC).

3.3. Primary consolidation 3.3.2. Coefficient of consolidation (Cv) and coefficient of

permeability (kv)
3.3.1. Void ratio and yield stress The coefficient of consolidation Cv (in m2/s) is the
The one-dimensional compression curves for both the parameter used to describe the rate at which saturated clay
untreated and the cement-treated lateritic soil specimens, or other soil undergoes consolidation, or compaction,
given in terms of void ratio and the log of effective vertical when subjected to an increase in pressure.
stress (e–log r0v relationship), are shown in Fig. 12. The In this investigation, for each load increment, the coeffi-
water content, cement content, and curing time signifi- cient of consolidation was calculated using Taylor’s proce-
cantly affect the fabric (pore space among the clay dure (graphical method) which consists of determining t90
particles). by plotting the change in height against the square root
A decrease in void ratio occurs for cement contents of of time, and to express Cv, by the following equation:
3–9% compared to the untreated laterite. This decrease
occurs essentially during the first 7 days of curing. Between 0:848H 2
Cv ¼ ð3Þ
7 and 60 days, the decrease in void ratio is less substantial t90
 E. Mengue et al. / Soils and Foundations 57 (2017) 60–79 73

0.65
(a) Dry side
0.63

0.61 Untreated soil

3% cem - 7 days
Void ratio, e 0.59 3% cem - 28 days
3% cem - 60 days
0.57 6% cem - 7 days
6% cem - 28 days
0.55 6% cem - 60 days
9% cem - 7 days
0.53 9% cem - 28 days
9% cem - 60 days
0.51

0.49
15 150 150 0
Effective vertical stress, 'v (kPa)
0.65
Yield stress ( 'y) (b) Optimum moisture
0.63

0.61 Untreated soil

3% cem - 7 days
3% cem - 28 days
Void ratio, e

0.59
3% cem - 60 days
0.57 6% cem - 7 days
6% cem - 28 days
0.55 6% cem - 60 days
9% cem - 7 days
0.53 9% cem - 28 days
9% cem - 60 days
0.51

0.49
15 150 1500

Effective vertical stress, 'v (kPa)

0.77
(c) Wet side
0.73
Untreated soil
0.69 3% cem - 7 days
3% cem - 28 days
Void ratio, e

0.65 3% cem - 60 days

6% cem - 7 days
0.61 6% cem - 28 days
6% cem - 60 days
0.57 9% cem - 7 days
9% cem - 28 days
0.53 9% cem - 60 days

0.49
15 150 1500

Effective vertical stress, v (kPa)

Fig. 12. Compressibility curves (e–log r0v ).

where H is the average height of the specimen during the specimens with cement are 2–7 times that for the speci-
increment/2 (for two-way drainage) and t90 is the time mens without cement. On the other hand, for effective
for 90% primary consolidation. vertical stresses higher than the cement-induced yield
Fig. 14 shows the variation in the coefficient of consol- stress, the Cv values for the specimens treated with
idation (Cv) with the effective vertical stress. It is cement are about 0.5–1.5 times that for the specimens
observed that, for effective vertical stresses smaller than without cement. This means that, the use of cement to
the cement induced yield stress, the Cv values for the increase the coefficient of consolidation (Cv) is effective
74 E. Mengue et al. / Soils and Foundations 57 (2017) 60–79

400 400
7 curing days Molding water: DRY
350 350
DRY 7 days

(kPa)
Yield stress, 'y (kPa)

300 OMC 300 28 days

WET 60 days

y
250 250

200 200

150 150

100 100

50 50
0 2 4 6 8 10 0 2 4 6 8 10
Cement content (%) Cement content (%)

400 400

28 curing days Molding water: OMC

350 350
7 days
Yield stress, 'y (kPa)
DRY
Yield stress, 'y (kPa)

300 300 28 days

OMC
WET 60 days
250 250

200 200

150 150

100 100

50 50
0 2 4 6 8 10 0 2 4 6 8 10
Cement content (%) Cement content (%)
400 400

60 curing days Molding water: WET

350 350
(kPa)

DRY 7 days
Yield stress, 'y (kPa)

300 OMC 300

28 days
WET 60 days
y

250 250

200 200

150 150

100 100

50 50
0 2 4 6 8 10 0 2 4 6 8 10
Cement content (%) Cement content (%)

Fig. 13. Variation in yield stress; error bars show standard deviations.

only for effective vertical stresses below the cement- It should also be noted that the Cv of the untreated sam-
induced yield stress. This is in agreement with other ples generally remains low, reflecting the slow process of
works (Feng et al., 2001). primary consolidation. However, the Cv of the treated
 E. Mengue et al. / Soils and Foundations 57 (2017) 60–79 75

18 compacted from the wet side of optimum (xWET). For

Molding water: DRY instance, for a specimen treated at 6% cement and sub-
16
'y=115MPa
Coefficient of consolidation,

jected to an effective vertical stress below the yield stress

14
(50 kPa for example), the coefficient of consolidation (Cv)
of the specimens prepared from the dry side of optimum
Cv (10-8m2/s)

12 0% Cement
3% Cement
10 6% Cement
(xDRY) and the optimum moisture content (OMC) is
9% Cement 1.1–1.3 times higher than that of the specimens prepared
8 from the wet side of optimum (xWET). This is consistent
6 with the results of the yield stress discussed earlier which
show that it would be desirable for the soil sample to be
4
prepared at xDRY when the OMC is not reached at the site.
2 Compared to the results obtained by Feng et al. (2001), in
0 which the soil sample in their investigation was treated with
0 200 400 600 800 CEM I cement, the results obtained in this study with CEM
Effective vertical stress, 'v (kPa) II/BM 32.5 N cement show that the consolidation process
8 is twice as fast as that obtained by Feng et al. (2001). This
Molding water: OMC is important for project managers during the construction
7 'y=120MPa of road embankments and other geotechnical structures.
6 The permeability of soil is its capability to permit the
Coefficient of consolidation,

flow of fluids through its pore spaces. However, the use

5 of the material in road construction is dependent on the
Cv (10-8 m2/s)

4 degree of permeability, which determines whether the pro-

posed road foundation will be suitable for use and also
3
0% Cement
enables the design for the drainage of pavement structures.
2 3% Cement The coefficient of permeability (kv) for both the untreated
6% Cement and the cement-lateritic soil samples at stress levels of 50,
9% Cement
1 100, 150, 250, 500, and 750 kPa was determined on the
specimens kept saturated in the oedometric cell. This kv
0
0 200 400 600 800 was estimated from the coefficient of consolidation by
Effective vertical stress, 'v (kPa) using the following expression:
C v  cx Dr0v
8 kv ¼ with Eoed ¼ ð4Þ
Molding water: WET
Eoed DH =H 0
7 'y=95 MPa
where cx is unit weight of the pore fluid (kN/m3), Eoed is the
Coefficient of consolidation,

6
oedometer modulus (m2/kN), Dr0v is the variation in effec-
5 tive stress, DH is the variation in height of the specimen,
Cv (10-8m2/s)

and H0 is the initial height of the specimen.

4
Fig. 15 shows the variation in the coefficient of perme-
3 ability (kv) with cement contents to a stress level of
0% cement 750 kPa. It is observed that there is generally a decrease
2 3% Cement
6% Cement in the coefficient of permeability with an increase in the
1 9% Cement cement content. According to the results obtained by the
Raman spectrometer and SEM analyses, this decrease is
0
0 200 400 600 800
the result of the formation of cementitious compounds by
calcium from lime and the readily available silica and/or
Effective vertical stress, 'v (kPa)
alumina from both the soil and the cement, which fills
Fig. 14. Variation in Cv with effective vertical stress; error bars show the soil voids thereby obstructing the flow of water.
standard deviations. The permeability of all the cement-treated specimens
decreased to a minimum at a 6% cement content. The val-
ues of the coefficient of permeability at 9% cement were
laterite reaches its maximum increase for 6% cement. This higher than the values measured for the treated soil at
could be explained by the fact that the attachment point is 6% cement. A probable explanation for the decrease in kv
reached from the 6% added cement (Fig. 10). More impor- to a minimum at 6% cement content is that the hydration
tantly, the largest values of Cv for the treated samples are is accompanied by a continued increase in pH beyond the
obtained from the dry side of the optimum (xDRY) and point of attachment at 6% cement (Fig. 10) and Ca2+ ion
the optimum moisture content (OMC), reflecting a fast concentration on the fine particle surface (kaolinite). The
process of primary consolidation compared to samples increased concentration of Ca2+ ion causes shrinkage of
76 E. Mengue et al. / Soils and Foundations 57 (2017) 60–79

2.5 2.5

7 curing days Molding water: DRY

2 2
Coefficient of permeability,

Coefficient of permeability,
7 days
DRY 28 days
OMC
kv (10-10m/s)

kv (10-11m/s
1.5 1.5 60 days
WET

1 1

0.5 0.5

0 0
0 2 4 6 8 10 0 2 4 6 8 10
Cement content (%) Cement content (%)

2.5 2.5
Molding water: OMC
28 curing days
Coefficient of permeability,
Coefficient of permeability,

2 2
kv (10-11 m/s)
kv (10-11m/s)

1.5 DRY 1.5 7 days

OMC 28 days
WET 60 days
1 1

0.5 0.5

0 0
0 2 4 6 8 10 0 2 4 6 8 10
Cement content (%) Cement content (%)
2.5 2.5

60 curing days Molding water: WET

Coefficient of permeability,

2 2
Coefficient of permeability,

7 days
kv (10-11m/s)

kv (10-11m/s)

1.5 DRY 1.5 28 days

60 days
OMC
WET
1 1

0.5 0.5

0 0
0 2 4 6 8 10 0 2 4 6 8 10
Cement content (%) Cement content (%)

Fig. 15. Variation in kv with cement content for r0v = 750 kPa; error bars show standard deviations.
 E. Mengue et al. / Soils and Foundations 57 (2017) 60–79 77

the diffused double layer leading to flocculation, and hence, the parameter indicates a nearly constant value for a given
a slight increase in permeability after 6% cement has been load increment. Hence, the secondary settlement (Ss) can
added (Osinubi, 1998; De Brito Galvão et al., 2004). simply be determined by using the following equation:
For all the cured specimens, the coefficients of perme-
ability decreased with the curing age. In general, the perme- Ca tf
Ss ¼ H p log ð6Þ
ability of specimens with cement decreases to 1–9 times 1 þ ep tp
that of the untreated specimen. Furthermore, the decrease
in the coefficient of permeability with time is probably due
to the precipitation of calcium carbonate in the voids as the where ep is the void ratio at the start of the linear portion of
ionized calcium reacts with the dissolved carbon dioxide in the void ratio versus logarithm of time curve, Hp is the
the water. This study shows that a percentage of cement, thickness of the compressible soil layer at time tp (m), tp
from 3 to 6%, can be effectively used to reduce the perme- is the time corresponding to the completion of the primary
ability of A-7-5 lateritic soil. In view of the obtained consolidation, and tf is the time at which the secondary
results, and especially those obtained for permeability, it compression settlement is computed.
can be inferred that a 6% content of cement is sufficient The obtained values for Ca (Table 5) show that Ca
to obtain an adequate mechanical performance. increases as the effective vertical stress is increased, mean-
ing that Ca is strongly dependent on the effective stress.
3.4. Secondary compression Similar results were obtained by other authors when
cement was used to treat fine-grained soils (Mesri et al.,
3.4.1. Secondary compression index 1997).
The key parameter for the quantitative analysis of the It is also observed that Ca decreases as the cement con-
secondary compression is the secondary compression index tent is increased, including maturation. The Ca of the spec-
(Ca). It is defined as the variation in void ratio (De) as a imens with cement decreases to 2–11 times that of the
function of the change in time (t) plotted in the logarithmic untreated specimen. This is significant because it implies
scale. Accordingly, the coefficient of secondary compres- that the long-term and continuing settlement of lateritic
sion (Ca) is defined as soils, due to a sustained load from the infrastructure ele-
ments in the field, may be effectively stabilized with cement
De
Ca ¼ ð5Þ treatment at appropriate dosages. A decrease in Ca was
D log t obtained in other investigations when cement was used to
where De is the variation in void ratio and t the time in treat soft grounds and dredged sediments (Al-Shamrani,
minutes (mn) 1998; Bobet et al., 2011).
The secondary compression index (Ca) is by far the most If the effective vertical stress of the embankment load on
useful and commonly used parameter for describing sec- a layer of lateritic soil, Hp = 9.5 m in thickness, is increased
ondary compression (Al-Shamrani, 1998). The importance to r0v = 150 kPa, at which Ca/(1 + ep) = 0.0064, the end of
of this parameter stems from the fact that for some soils, primary consolidation is reached in tp = 25 days. Then,

Table 5
Average values of secondary compression index.
Ca with curing time Stess level r0v (kPa) Cement content (%)
0 3 6 9
Ca at 7 days (103) 50 1.1 0.2 – –
100 1 0.6 – –
150 1.1 0.6 0.8 0.4
250 1.8 0.6 0.6 0.5
500 1.8 1.3 0.7 0.4
750 1.9 1 0.8 0.7
Ca at 28 days (103) 50 1.1 0.2 – –
100 1 0.4 0.2 –
150 1.1 0.5 0.5 –
250 1.8 0.6 0.6 0.2
500 1.8 0.7 0.7 0.6
750 1.9 0.8 – 0.4
Ca at 60 days (103) 50 1.1 0.2 – –
100 1 0.4 0.2 –
150 1.1 0.4 0.6 0.1
250 1.8 0.6 0.6 0.2
500 1.8 0.6 0.6 0.6
750 1.9 0.5 0.5 0.4
78 E. Mengue et al. / Soils and Foundations 57 (2017) 60–79

0.04 (CSH) which were derived principally from cement hydra-

Inorganic clays and silts tion. The experimental results show that a 6% content of
0.035 cement is sufficient for obtaining an adequate mechanical
7 curing days performance. On the other hand, based on the measured
28 curing days
physical magnitudes of the consolidation properties, it
0.03 Shale and mudstone area 60 curing days
would be desirable to compact the soil sample on the dry
side of optimum when the optimum moisture content is
C /Cc

0.025 not reached at the site. However, the use of cement to

increase the coefficient of consolidation is more effective
0.02 only for effective vertical stresses below the cement-
induced yield stress. Cement treatment can alter the engi-
neering behavior (analyzed by the Ca/Cc ratio concept) of
0.015
Granular materials lateritic soils to a more desirable performance close to that
of a granular material.
0.01
0 2 4 6 8 10
Cement content (%)
Acknowledgements
Fig. 16. Variation in Ca/Cc with cement content; error bars show standard
deviations. The authors would like to thank the MUNDUS ACP 2
project for funding this research. Appreciation is also
extended to the location of the host laboratory, the
during a design life of 30 years, 15 cm of secondary settle-
LGCGE, where almost all of these experiments were
ment is expected to follow primary consolidation.
conducted.
3.4.2. Ca/Cc ratio
For a given soil, Ca/Cc is generally observed to remain References
constant, reflecting the similar nature of the mechanisms
that contribute to primary and secondary consolidation. AASHTO, 1986. Standard for Transportation Materials and Methods of
Moreover, for a given soil type, Ca/Cc is known to vary Sampling and Testing, 14th ed. AASHTO, Washington, D.C.
Abdulhussein Saeed, K., Anuar Kassim, K., Nur, H., 2014. Physico-
within a fairly limited range (Bobet et al., 2011). The value chemical characterization of cement treated kaolin clay. Gradevinar 66
for Ca/Cc is 0.036 for the specimens without cement and (06), 513–521.
between 0.023 and 0.012 for the specimens treated with Al-Amoudi, O.S.B., 2002. Characterization and chemical stabilization of
cement up to 9% by weight of dry soil, and hardened up Al-Qurayyah sabkha soil. J. Mater. Civ. Eng. 14 (6), 478–484.
to 60 days (Fig. 16). These data indicate that adding Al-Shamrani, M.A., 1998. Application of the C a/C c concept to
secondary compression of sabkha soils. Can. Geotech. J. 35 (1), 15–26.
cement tends to decrease the value of Ca/Cc. The decrease ASTM, 1992. Annual book of ASTM standards, West Conshohocken.
in Ca/Cc reflects the modifications in the lateritic soil engi- Autret, P., 1980. Contribution à l’étude des graveleux latéritiques traités
neering behavior, which shifts from that of inorganic clays au ciment Thèse de doctorat. ENPC.
and silts (between 0.035 and 0.045), to shale and mudstone Bagarre, E., 1990. Utilisation des graveleux latéritiques en technique
(between 0.025 and 0.035), and finally to a granular soil routière. Rev. ISTED-CEBTP.
Bobet, A., Hwang, J., Johnston, C.T., Santagata, M., 2011. One-
(from 0.025). Similar results were obtained by other dimensional consolidation behavior of cement-treated organic soil.
authors (Terzaghi et al., 1996). Can. Geotech. J. 48 (7), 1100–1115.
The same trend as that for the decrease in Ca/Cc was CEBTP, 1984. Guide pratique de dimensionnement des chaussées pour les
observed in other works when cement had been used to pays tropicaux.
treat the soil. The authors concluded that these results were Chew, S.H., Kamruzzaman, A.H.M., Lee, F.H., 2004. Physicochemical
and engineering behavior of cement treated clays. J. Geotech.
encouraging, since the compression behavior of the soils Geoenviron. Eng. 130 (7), 696–706.
appeared to have been fundamentally changed (by cement De Brito Galvão, T.C., Elsharief, A., Simões, G.F., 2004. Effects of lime
stabilization) to that of an approximate granular soil, on permeability and compressibility of two tropical residual soils. J.
which is considered to be an excellent foundation material Environ. Eng. 130 (8), 881–885.
by geotechnical and pavement engineers (Hwang, 2006). Eades, J.L., Grim, R.E., 1966. A quick test to determine lime requirements
for lime stabilization. Highway research record.
Feng, T.W., Lee, J.Y., Lee, Y.J., 2001. Consolidation behavior of a soft
4. Conclusion mud treated with small cement content. Eng. Geol. 59 (3), 327–335.
GTR, 1992. Guide technique pour la réalisation des remblais et couches de
CEM II/BM 32.5 N cement additions to fine-grained forme. SETRA/LCPC.
lateritic soil led to the formation of a homogeneous Hanrion, C., Delvigne, J., 1976. Techniques utilisées pour la préparation
des lames minces pétrographiques.
microstructure and improved the consolidation character- Holtz, R.D., Kovacs, W.D., 1981. Introduction à la géotechnique.
istics. These positive impacts are mainly related to the fact Editions de l’Ecole Polytechnique de Montréal.
that the addition of cement gives way to the formation of Hwang, J., 2006. Effects of cement treatment on the one-dimensional
ettringite, calcite, portlandite, afwillite, and tobermorite consolidation behavior of a highly organic soil. ProQuest.
 E. Mengue et al. / Soils and Foundations 57 (2017) 60–79 79

Idrissa, T., 1994. Contribution à l’étude des graveleux latéritiques dans les Millogo, Y., Hajjaji, M., Ouedraogo, R., Gomina, M., 2008. Cement-
pays du Sahel: cas des routes non revêtues. Thèse de M.Sc.A (génie lateritic gravels mixtures: microstructure and strength characteristics.
civil). Constr. Build. Mater. 22 (10), 2078–2086.
Ismaiel, H.A.H., 2006. Treatment and improvement of the geotechnical Mitchell, D.R.G., Hinczak, I., Day, R.A., 1998. Interaction of silica fume
properties of different soft fine-grained soils using chemical stabiliza- with calcium hydroxide solutions and hydrated cement pastes. Cem.
tion. Shaker, 2006. Concr. Res. 28 (11), 1571–1584.
Joel, M., Agbede, I.O., 2010. Mechanical-cement stabilization of laterite NF ISO 10390, 2005. Qualité du sol – Détermination du pH; Mai 2005.
for use as flexible pavement material. J. Mater. Civ. Eng. 23 (2), 146– NF P 94-051, 1993. Détermination des limites d’Atterberg. Limite de
152. liquidité à la coupelle. Limite de plasticité au rouleau. Mars 1993.
Kamruzzaman, A.H., Chew, S.H., Lee, F.H., 2009. Structuration and NF P 94-056, 1996. Analyse granulométrique. Méthode par tamisage à sec
destructuration behavior of cement-treated Singapore marine clay. J. après lavage. Mars 1996.
Geotech. Geoenviron. Eng. 135 (4), 573–589. NF P 94-057, 1992. Analyse granulométrie des sols. Méthode par
Kazemian, S., Huat, B.B., 2009. Compressibility characteristics of fibrous sédimentation. Mai 1992.
tropical peat reinforced with cement column. Electron. J. Geotech. NF P 94-068, 1998. Détermination de la valeur de bleu de méthylène d’un
Eng. 14, 1–13. sol par l’essai à la tâche, Aout 1998.
Lemaire, K., Deneele, D., Bonnet, S., Legret, M., 2013. Effects of lime and NF P94-051, 1997. Essai oedométrique; Décembre 1997.
cement treatment on the physicochemical, microstructural and NF P94-093, 1999. Détermination des références de compactage d’un
mechanical characteristics of a plastic silt. Eng. Geol. 166, 255–261. matériau; Octobre 1999.
Li, Y., Tan, Y.Z., Jiang, J., Xia, C.F., 2012. Experimental study on the Osinubi, K.J., 1998. Permeability of lime-treated lateritic soil. J. Trans-
compressibility of cement improved laterite soil. Appl. Mech. Mater. port. Eng. 124 (5), 465–469.
170–173, 371–374. Oyediran, I.A., Kalejaiye, M., 2011. Effect of increasing cement content on
Locat, J., Berube, M.A., Choquette, M., 1990. Laboratory investigations strength and compaction parameters of some lateritic soils from
on the lime stabilization of sensitive clays: shear strength development. southwestern Nigeria. Electron. J. Geotech. Eng. 16, 1501–1514.
Can. Geotech. J. 27 (3), 294–304. Santamarina, J.C., Klein, K., Fam, M., 2001. Soils and Waves. Wiley,
Magnan, J.P., Mieussens, C., Soyez, B., Vautrain, J., 1985. Essais New York.
œdométriques. Laboratoire central des Ponts et Chaussées. Paris, Tchameni, R., 1997. Géochimie et géochronologie des formations de
Méthode d’essai LPC, 13, 83 p. l’Archéen et du Paléoprotérozoı̈que du Sud-Cameroun (Groupe du
Mesbah, A., Morel, J.C., Olivier, M., 1999. Comportement des sols fins Ntem, Craton du Congo). Thèse de doctorat, Université de Yaoundé I.
argileux pendant un essai de compactage statique: détermination des Terzaghi, K., Peck, R.B., Mesri, G., 1996. Soil Mechanics in Engineering
paramètres pertinents. Mater. Struct. 32 (9), 687–694. Practice, third ed. John Wiley and Sons Inc.
Mesri, G., Stark, T.D., Ajlouni, M.A., Chen, C.S., 1997. Secondary Wang, Y.H., Siu, W.K., 2006. Structure characteristics and mechanical
compression of peat with or without surcharging. J. Geotech. properties of kaolinite soils. I. Surface charges and structural charac-
Geoenviron. Eng. 123 (5), 411–421. terizations. Can. Geotech. J. 43 (6), 587–600.
Messou, M., 1980. Comportement mécanique d’une couche de base en Yu, P., Kirkpatrick, R.J., Poe, B., McMillan, P.F., Cong, X., 1999.
graveleux latéritiques améliorés au ciment: cas des routes en Côte Structure of calcium silicate hydrate (C-S-H): near-, mid-, and far-
d’Ivoire” Thèse de doctorat. ENPC. infrared spectroscopy. J. Am. Ceram. Soc. 82 (3), 742–748.