Downloaded from ascelibrary.org by UFRGS - Universidade Federal do Rio Grande do Sul on 08/15/15. Copyright ASCE.

For personal use only; all rights reserved.

Parameters Controlling Tensile and Compressive Strength of

Artificially Cemented Sand
Nilo Cesar Consoli, Ph.D.1; Rodrigo Caberlon Cruz, D.Sc.2; Mrcio Felipe Floss3; and Lucas Festugato4
Abstract: The enhancement of local soils with cement for the construction of stabilized pavement bases, canal lining, and support layer
for shallow foundations shows great economical and environmental advantages, avoiding the use of borrow materials from elsewhere, as
well as the need of a spoil area. The present research aims to quantify the influence of the amount of cement, the porosity, and the
voids/cement ratio in the assessment of unconfined compressive strength qu and splitting tensile strength qt of an artificially cemented
sand, as well as in the evaluation of qt / qu relationship. A program of splitting tensile tests and unconfined compression tests considering
three distinct voids ratio and seven cement contents, varying from 1 to 12%, was carried out in the present study. The results show that
a power function adapts well qt and qu values with increasing cement content and with reducing porosity of the compacted mixture. The
voids/cement ratio is demonstrated to be an appropriate parameter to assess both qt and qu of the sand-cement mixture studied. Finally, the
qt / qu relationship is unique for the sand-cement studied, being independent of the voids/cement ratio.
DOI: 10.1061/ASCEGT.1943-5606.0000278
CE Database subject headings: Tensile strength; Compressive strength; Soil cement; Compacted soils.
Author keywords: Tensile strength; Compressive strength; Soil cement; Compacted soils.

Introduction
Portland cement is used worldwide in the improvement of local
soils, particularly as a soil-cement mixture of a compacted layer
over a low bearing capacity soil. In such cases, Consoli et al.
2003 has shown that the system failure mechanism usually start
up under tensile stresses at the base of the improved layer. Although it would seem more reasonable to use the tensile strength
as a direct measure of the soil-cement strength, there are no dosage methodologies based on rational criteria considering the effect of different variables e.g., amount of cement and porosity
on the soil-cement tensile strength.
The first rational dosage methodology for soil-cement was developed by Consoli et al. 2007 considering the voids/cement
ratio / Cv, defined by the porosity of the compacted mixture
divided by the volumetric cement content, as an appropriate parameter to evaluate qu of the soil-cement mixture. Nowadays,
even though it is recognized that compressive and tensile
strengths are intimately related on artificially cemented soils e.g.,
Clough et al. 1981; Consoli et al. 2001, it is still not clear

whether there is a straight proportionality between qu and qt and if

such relation is a function of porosity and cement content. The
unconfined compression test has therefore been used as the most
convenient means to investigate the effect of different variables
on the soil-cement strength and to carry out dosage methodologies. Questions that remain unanswered are: is it correct to carry
out dosage methodologies based on unconfined compression tests
for the cases where tensile stresses are the basic variables? Is
there a straight proportionality between qu and qt? Is the qt / qu
index for a given soil-cement relationship a function of voids/
cement ratio? This study aims at approaching some of these questions by quantifying the influence of the amount of cement and
the porosity on the tensile strength of an artificially cemented
sand, as well as to evaluate the use of voids/cement ratio to assess
its splitting tensile strength qt, unconfined compressive strength
qu and their ratio qt / qu. The main contributions of present
work are showing that the voids/cement ratio / Cv is an appropriate index parameter to evaluate not only unconfined compressive strength qu of soil-cement mixtures, but splitting tensile
strength qt as well, and that the qt / qu relationship appears to be
unique for a given soil-cement studied, being independent of the
voids/cement ratio.

Associate Professor, Dept. of Civil Engineering, Federal Univ. of Rio

Grande do Sul, Rio Grande do Sul 90035-190, Brazil corresponding
author. E-mail: consoli@ufrgs.br
2
Research Fellow, Federal Univ. of Rio Grande do Sul, Rio Grande do
Sul 90035-190, Brazil. E-mail: rccaberlon@hotmail.com
3
Ph.D. Student, Federal Univ. of Rio Grande do Sul, Rio Grande do
Sul 90035-190, Brazil. E-mail: mfloss@gmail.com
4
Ph.D. Student, Federal Univ. of Rio Grande do Sul, Rio Grande do
Sul 90035-190, Brazil. E-mail: lufega@yahoo.com.br
Note. This manuscript was submitted on March 13, 2009; approved
on November 1, 2009; published online on November 5, 2009. Discussion period open until October 1, 2010; separate discussions must be
submitted for individual papers. This technical note is part of the Journal
of Geotechnical and Geoenvironmental Engineering, Vol. 136, No. 5,
May 1, 2010. ASCE, ISSN 1090-0241/2010/5-759763/$25.00.

Experimental Program
The experimental program has been carried out in two parts. First,
the geotechnical properties of the soil and cement were characterized. Then, a number of splitting tensile and unconfined compression tests was carried out. In addition, measurements of matric
suction were also performed in most specimens to check a possible influence on the results.
Materials
The Osorio sand used in the testing was obtained from the region
of Porto Alegre, in Southern Brazil, being classified ASTM

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING ASCE / MAY 2010 / 759

J. Geotech. Geoenviron. Eng. 2010.136:759-763.

Downloaded from ascelibrary.org by UFRGS - Universidade Federal do Rio Grande do Sul on 08/15/15. Copyright ASCE. For personal use only; all rights reserved.

1993 as nonplastic uniform fine sand with rounded particle shape

and specific gravity of the solids 2.65. Mineralogical analysis
showed that sand particles are predominantly quartz. The grain
size is purely fine sand with a mean effective diameter D50 of
0.16 mm, being the uniformity and curvature coefficients of 1.9
and 1.2, respectively. The minimum and maximum void ratios are
0.6 and 0.9, respectively.
Portland cement of high initial strength Type III was used as
the cementing agent. Its fast gain of strength allowed the adoption
of seven days as the curing time. The specific gravity of the
cement grains is 3.15. Distilled water was used for these characterization tests and tap water for molding specimens for the tensile and compression tests.

capacities of 10 and 50 kN and resolutions of 0.005 and 0.023 kN,

respectively, were used for the unconfined compression tests.
After curing, the specimens were submerged in a water tank
for 24 h for saturation to minimize suction. The water temperature
was controlled and maintained at 23 3 C. Immediately before
the test, the specimens were removed from the tank and dried
superficially with an absorbent cloth. Then, the splitting tensile
test was carried out and the maximum load recorded. As acceptance criteria, it was stipulated that the individual strengths of
three specimens, molded with the same characteristics, should not
deviate by more than 10% from the mean strength.
Unconfined Compression Tests

Methods
Molding and Curing of Specimens
For the splitting tensile and unconfined compression tests, cylindrical specimens 50 mm in diameter and 100-mm high were used.
Once established a given voids ratio e, the target dry unit weight
d was calculated according to Eq. 1
e=

s
1
d

where s = solids unit weight. A target dry unit weight for a given
specimen was then established through the dry mass of soilcement divided by the total volume of the specimen. In order to
keep the dry unit weight of the specimens constant with increasing cement content, a small portion of the soil was replaced by
cement. As the specific gravity of the cement grains 3.15 is
greater than the specific gravity of the soil grains 2.65, for the
calculation of void ratio and porosity, a composite specific gravity
based on the soil and cement percentages in the specimens was
used.
After the soil, cement, and water were weighed, the soil and
cement were mixed until the mixture acquired a uniform consistency. The water was then added continuing the mixture process
until a homogeneous paste was created. The amount of cement for
each mixture was calculated based on the mass of dry soil and the
moisture content. The specimen was then statically compacted in
three layers inside a cylindrical split mold, which was lubricated,
so that each layer reached the specified dry unit weight. The top
of each layer was slightly scarified. After the molding process, the
specimen was immediately extracted from the split mold and its
weight, diameter, and height measured with accuracies of about
0.01 g and 0.1 mm, respectively. The samples were then placed
inside plastic bags to avoid significant variations of moisture content. They were cured for 6 days in a humid room at 23 2 C
and relative humidity of above 95%.
The samples were considered suitable for testing if they met
the following tolerances:
Dry unit weight d: degree of compaction between 99 and
101% the degree of compaction being defined as the value
obtained in the molding process divided by the target value of
d; and
Dimensions: diameter to within 0.5 mm and height of 1
mm.
Splitting Tensile Tests
Splitting tensile tests followed Brazilian standard NBR 7222
Brazilian Standard Association 1983. An automatic loading machine, with maximum capacity of 50 kN and proving rings with

Unconfined compression tests have been systematically used in

most experimental programs reported in the literature in order to
verify the effectiveness of the stabilization with cement or to access the importance of influencing factors on the strength of cemented soils. One of the reasons for this is the accumulated
experience with this kind of test for concrete. The tests usually
followed Brazilian standard NBR 5739 Brazilian Standard Association 1980, being simple and fast, while reliable and cheap.
The automatic loading machine and proving rings were the
same used for the splitting tensile tests. Curing of specimens and
acceptance criteria were exactly the same as for splitting tensile
tests.
Matric Suction Measurements
At their molding moisture contents, all specimens were in an
unsaturated state exhibiting a certain level of suction. Suction
measurements aimed to verify its magnitude and examine if there
was significant variation between specimens of different porosities and cement contents.
The matric suction, i.e., that arising from the capillary forces
inside the sample, was measured using the filter paper technique.
The filter paper used was Whatman No. 42. Its initial moisture
content in the air dried state is approximately 6%, which allows
measurements of suction from zero to 29 MPa. The calibration
equations for this filter paper are those presented by Chandler et
al. 1992.
Program of Splitting Tensile and Unconfined Compression
Tests
The splitting tensile and unconfined compression tests constituted
the main part of this research. The program was conceived in such
a way as to evaluate, separately, the influences of the cement
content, porosity, and voids/cement ratio on the mechanical
strength of the artificially cemented soil.
The molding points were chosen considering voids ratio of
0.64, 0.70, and 0.78 corresponding, respectively, to high, medium, and reduced relative densities, with the same moisture
content 10%. Each point was molded with seven different
cement percentages: 1, 2, 3, 5, 7, 9, and 12%. These percentages
were chosen following Brazilian and international experience
with soil-cement e.g., Mitchell 1981; Consoli et al. 2003, 2006,
2007, 2009; Thom et al. 2005. Because of the typical scatter of
data for both splitting tensile and unconfined compression tests, a
minimum of three specimens were tested for each point.

760 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING ASCE / MAY 2010

J. Geotech. Geoenviron. Eng. 2010.136:759-763.

Downloaded from ascelibrary.org by UFRGS - Universidade Federal do Rio Grande do Sul on 08/15/15. Copyright ASCE. For personal use only; all rights reserved.

Fig. 1. Variation of splitting tensile strength qt with cement content

Results
Effect of the Cement Content and Porosity on Tensile
and Compressive Strengths
Fig. 1 shows the raw data for the three studied voids ratio and
the fitted lines for the splitting tensile strength qt as a function
of the cement content C, the latter defined by Eq. 2 as
C=

Wc
Ws

where Wc = mass of cement and Ws = mass of dry soil. It can be

observed in Fig. 1 that the cement content has a great effect on the
tensile strength of this sand-cement mixture, where a small addition of cement is enough to generate a significant gain in strength.
The lines shown on the figure are best fit lines, demonstrating that
a power function adapts well the relation qt C.

Fig. 2. Variation of splitting tensile strength qt with porosity

Fig. 3. Variation of unconfined compressive strength qu with cement content

Fig. 2 shows how the porosity affects the splitting tensile

strength of the sand-cement mixture. The tensile strength increases with reducing porosity of the compacted mixture. The
beneficial effect of a decrease in porosity on the tensile strength
has been reported by several researchers e.g., Moore et al. 1970.
In particular, Chang and Woods 1992 have already shown,
through electron microscopy on different sands with various kinds
of cement, that the existence of a larger number of interparticle
contacts and, consequently, a greater possibility of the cement to
promote effective bonds at these contacts, explains the increase in
the rate of tensile strength gain with the reduction in the porosity.
The unconfined compression strength qu variation with the
amount of cement is shown in Fig. 3. With a similar pattern to the
splitting tensile tests, a power function also fits well to the relation qu C. Besides that, the soil-cement mixtures present clear
increase in the unconfined compressive strength gain rate with
decreasing voids ratio Fig. 4.
The process of submerging the specimens for 24 h before the
splitting tensile tests was found to be satisfactory to ensure a high
and repeatable degree of saturation. An average degree of saturation of 92% was obtained for specimens after submersion, irre-

Fig. 4. Variation of unconfined compressive strength qu with porosity

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING ASCE / MAY 2010 / 761

J. Geotech. Geoenviron. Eng. 2010.136:759-763.

Downloaded from ascelibrary.org by UFRGS - Universidade Federal do Rio Grande do Sul on 08/15/15. Copyright ASCE. For personal use only; all rights reserved.

Fig. 5. Variation of splitting tensile strength qt with voids/cement

ratio

spective of the initial porosity or cementitious material content.

The values of suction measured were low with values ranging
from about 14% of the tensile strength. These measurements
were made on the specimens after failure in the splitting tensile
tests and are therefore likely to overestimate the real value, because a slight drying of the sample may have occurred during the
few minutes from the start of the test until the measurement was
made. Given the small values of matric suction measured in these
specimens, the small effect arising from the unsaturated nature
were disregarded.
Effect of Voids/Cement Ratio on Tensile and Compressive
Strengths
Fig. 5 presents the splitting tensile strength as a function of the
voids/cement ratio / Cv expressed as porosity divided by
the volumetric cement content Cv, the latter expressed as a percentage of cement volume regarding total volume, defined by Eq.
3

Vv

Vtotal
Vv
=
=
Cv
Vc
Vc
Vtotal

Fig. 6. Variation of unconfined compressive strength qu with voids/

cement ratio

qukPa = 28,327

Cv

1.30

Fig. 7 summarizes all measurements from Figs. 5 and 6. By

examining this figure, as well as Eqs. 4 and 5, it can be seen
that they present rather similar trends. In order to check whether a
qt / qu relationship for the sand-cement mixture is a function of
porosity, cement content, or voids/cement ratio, Eq. 4 is divided
by Eq. 5 which yields the ratio

Cv
qt
=
qu

28,327
Cv

1.30

4,266

1.30

= 0.15

It can be seen in Eq. 6 that qt / qu is a scalar for the sandcement blend, being independent of porosity, cement content, or
voids/cement ratio. So, there is a straight proportionality between
tensile and compressive strengths, which is valid for the whole

where Vv = volume of voids water+ air of the specimen; Vc

= volume of cement of the specimen; and Vtotal = total volume of
the specimen.
A simple observation of this figure suggests that the voids/
cement ratio is useful in normalizing results. A good correlation
R2 = 0.97 can be observed between this ratio / Cv and the
splitting tensile strength qt of the sand-cement studied see Eq.
4
qtkPa = 4,266

Cv

1.30

Fig. 6 presents a good quality correlation R2 = 0.98 between

/ Cv and the unconfined compressive strength qu of the sandcement studied see Eq. 5

Fig. 7. Variation of both splitting tensile qt and unconfined compressive strengths qu with voids/cement ratio

762 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING ASCE / MAY 2010

J. Geotech. Geoenviron. Eng. 2010.136:759-763.

Downloaded from ascelibrary.org by UFRGS - Universidade Federal do Rio Grande do Sul on 08/15/15. Copyright ASCE. For personal use only; all rights reserved.

range of voids ratio and cement content studied in the present

research program. As a consequence, it is possible to conclude
that any rational dosage methodology considering the effect of
different variables can be centered on tensile or compression tests,
once they are intimately related through a scalar 0.15 for the
sand-cement studied in the present research. Finally, the results
presented in this paper suggest that the voids/cement ratio can be
an extremely useful index for practitioners from which an engineer can choose the amount of cement appropriate to provide a
mixture that meets the strength required by the project at the
optimum cost. The voids/cement ratio can also be useful in a field
control of soil-cement layers. Once a poor compaction has been
identified, it can be readily taken into account in the design being
adopted through the qt versus / Cv or qu versus / Cv curves,
with corrective measures accordingly such as the reinforcement
of the treated layer or the reduction in the load transmitted.

Conclusions
From the data presented in this technical note, the following conclusions can be drawn:
A power function adapts well to both qt C and qu C sandcement mixture relations;
The reduction in porosity of the compacted mixture increases
both the tensile and compressive strengths;
The voids/cement ratio / Cv has been shown to be an appropriate index parameter to evaluate both splitting tensile qt
and unconfined compressive qu strength of sand-cement mixtures. Both qt and qu reduce with increasing / Cv values; and
The qt / qu ratio is a scalar 0.15 for the sand-cement mixture
evaluated in the present study, being independent of voids/
cement ratio. As a consequence, dosage methodologies based
on rational criteria can concentrate either on tensile or compression tests, once they are interdependable.

Acknowledgments
The writers wish to express their gratitude to Brazilian Research
Council CNPq/MCT Projects Produtividade em Pesquisa Grant
No. 301869/2007-3, Edital Universal Grant No. 472851/2008-0,
PNPD Grant No. 558474/2008-0, and INCT, to Brazilian Electrical Energy Agency ANEEL Project P&D Grant No. 0089-036/
2006-CEEE-GT/9936455, and to PRODOC CAPES for their
financial support to the research group.

Notation
The following symbols are used in this technical note:
C cement content expressed in relation to mass
of dry soil;
Cv volumetric cement content expressed in
relation to the total specimen volume;

D50
e
qt
qu
R2
Wc
Ws
d
s

/ Cv

mean effective diameter;

voids ratio;
splitting tensile strength;
unconfined compressive strength;
coefficient of determination;
mass of cement;
mass of dry soil;
dry unit weight;
solids unit weight;
porosity;
voids/cement ratio; and
moisture content.

References
ASTM. 1993. Standard classification of soils for engineering purposes. D2487-93, West Conshohocken, Pa.
Brazilian Standard Association. 1980. Mortar and concreteTest
method for compressive strength of cylindrical specimens. NBR
5739, Rio de Janeiro, Brazil in Portuguese.
Brazilian Standard Association. 1983. Mortar and concreteTest
method for splitting tensile strength of cylindrical specimens. NBR
7222, Rio de Janeiro, Brazil in Portuguese.
Chandler, R. J., Crilly, M. S., and Montgomery-Smith, G. 1992. A
low-cost method of assessing clay desiccation for low-rise buildings.
Proc.-Inst. Civ. Eng., 922, 8289.
Chang, T. S., and Woods, R. D. 1992. Effect of particle contact bond
on shear modulus. J. Geotech. Engrg., 1188, 12161233.
Clough, G. W., Sitar, N., Bachus, R. C., and Rad, N. S. 1981. Cemented sands under static loading. J. Geotech. Engrg. Div., 1076,
799817.
Consoli, N. C., Foppa, D., Festugato, L., and Heineck, K. S. 2007. Key
parameters for strength control of artificially cemented soils. J. Geotech. Geoenviron. Eng., 1332, 197205.
Consoli, N. C., Prietto, P. D. M., Carraro, J. A. H., and Heineck, K. S.
2001. Behavior of compacted soil-fly ash-carbide lime-fly ash mixtures. J. Geotech. Geoenviron. Eng., 127, 774782.
Consoli, N. C., Rotta, G. V., and Prietto, P. D. M. 2006. Yieldingcompressibility-strength relationship for an artificially cemented soil
cured under stress. Geotechniqu, 561, 6972.
Consoli, N. C., Vendruscolo, M. A., and Prietto, P. D. M. 2003. Behavior of plate load tests on soil layers improved with cement and
fiber. J. Geotech. Geoenviron. Eng., 1291, 96101.
Consoli, N. C., Viana da Fonseca, A., Cruz, R. C., and Heineck, K. S.
2009. Fundamental parameters for the stiffness and strength control
of artificially cemented sand. J. Geotech. Geoenviron. Eng., 1359,
13471353.
Mitchell, J. K. 1981. Soil improvementState-of-the-art report.
Proc., 10th Int. Conf. on Soil Mechanics and Foundation Engineering, International Society of Soil Mechanics and Foundation Engineering, Stockholm, Sweden, 509565.
Moore, R. K., Kennedy, T. W. and Hudson, W. R. 1970. Factors affecting the tensile strength of cement-treated materials. Highway Research Record: Soil Stabilization: Multiple Aspects, Vol. 315,
Highway Research Board, Washington, D.C., 6480.
Thom, A., Donato, M., Consoli, N. C., and Graham, J. 2005. Circular
footings on a cemented layer above weak foundation soil. Can. Geotech. J., 426, 15691584.

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING ASCE / MAY 2010 / 763

J. Geotech. Geoenviron. Eng. 2010.136:759-763.