Arabian Journal of Geosciences (2018) 11: 735

https://doi.org/10.1007/s12517-018-4045-x

ORIGINAL PAPER

The effect of silica fume and lime on geotechnical properties of a clay

soil showing both swelling and dispersive features
Murat Türköz 1 & Hasan Savaş 1 & Gorkem Tasci 1

Received: 28 February 2018 / Accepted: 8 November 2018 / Published online: 1 December 2018
# Saudi Society for Geosciences 2018

Abstract
Improvement of swell and dispersive characteristics of problematic soils which are present at various places in the world is an
important topic of research in engineering applications. In recent years, industrial side products are often used to improve
engineering features of problematic clay soils and both environmentalist and economic benefits are obtained. In the present
study, it is aimed to improve the high plasticity clay soil, which is determined to have both high swelling potential and dispersive
features, by using silica fume (SF) as a by-product material and lime (L). The minimum lime quantity that is required for the
pozzolanic reaction is fixed as 3% and experiments are made with SF additive mixtures of increasing percentages (0, 1, 3, 5, 10,
15, and 20%). Within this context, the swell percentage, swell pressure, crumb, pinhole, and unconsolidated–undrained (UU)
triaxial compression tests with different curing periods were conducted on the soil samples that are prepared by compressing at
compaction characteristics that are specified at each additive level at standard proctor energy. In addition to these tests, in order to
examine the changes caused by additives on soil structure, scanning electron microscopy (SEM) analyses were performed. As a
result of this study, it is found out that swelling and dispersive features of the clay soil improved and that curing period of first
7 days was more effective on the strength improvement and optimum stabilization was achieved at an addition of 3% lime
combined with 10% silica fume.

Keywords Dispersive clay . Swelling clay . Lime . Silica fume . Stabilization

Introduction expansive soils are thought to be the main cause of problems in

light structures (Nelson and Miller 1992).
Improvement of undesired features of clay soils such as Although not life-threatening or cataclysmic as compared to
dispersibility and swelling has an important place in engineering other natural events, expansive soils are certainly a natural haz-
applications. Dispersive and expansive soils are considered ard. In fact, expansive soil damage exceeds the average annual
problematic, and these soil properties cause serious problems damage caused by floods, hurricanes, earthquakes, and torna-
for many engineering structures (Sherard et al. 1976; Chen dos combined (Buhler and Cerato 2007). Expansive soils are
1988; Basma et al. 1995; Yong and Warkentin 1996; Abdullah frequently overlooked as a major problem because they take
et al. 1999; Goodarzi and Salimi 2015; Qi and Vanapalli 2015; years to cause extensive damage. However, in reality, expansive
Hassanlourad et al. 2017). Dispersive soils are thought to be the soils are a widespread and costly natural hazard. This damage to
cause of internal erosion in earthen structures (NRC 1983), and engineering structures results in significant financial losses. The
annual cost of damage to facilities and infrastructures caused by
these types of soils is estimated at £150 million in UK, $1000
* Murat Türköz
[email protected]
million in the United States, and several billions of dollars
worldwide (Qi and Vanapalli 2015). The annual cost of damage
Hasan Savaş to the engineering structures has also been reported in other
[email protected] studies (Chen 1988; Basma et al. 1995; Abdullah et al. 1999;
Gorkem Tasci Dua et al. 1999; Al-Rawas et al. 2002; Shi et al. 2002; Huang
[email protected] and Wu 2007; Phanikumar 2009; Turkoz and Tosun 2011;
Djellali et al. 2017).
1
Civil Engineering Department, Eskisehir Osmangazi University,
26480 Eskisehir, Turkey
735 Page 2 of 14 Arab J Geosci (2018) 11: 735

Dispersive clays are those clays that are confronted with There are many studies carried out for the purpose of
erosion due to their individual colloidal clay particles flowing studying the impacts of silica fume additive on the
slowly and creating suspension even in still water. This ero- problematic clay soil properties (Abd El-Aziz et al.
sion process is related to the physical–chemical situation of 2004; Al-Zairjawi 2009; Fattah et al. 2015a, 2015b;
clay fraction of a soil. This situation causes for individual clay Bharadwaj and Trivedi 2016; Goodarzi et al. 2016). In
particles to push each other in free water existence. Chemistry these studies, not only the swelling and dispersive prop-
of water is very important at this stage (Sherard et al. 1976; erties of the soil but also their physical and strength
Yong and Warkentin 1996; Penner and Lagaly 2001). parameters have been examined. In the study conducted
While dispersive clays are widely present in nature until by Abd El-Aziz et al. (2004), lime and silica fume
recent years, they were not defined as problematic soils. together have been used as stabilizers. It was found that
Earlier, it was considered that dispersive soils were only related the engineering properties of soil have been improved
to arid or semi-arid lands and alkali soils. In recent times, it is by adding lime (L) in the range of 5–9% combined with
found out that similar problems caused by dispersive clays have a 10% silica fume (SF). The plasticity index decreased
existed in moist climates in various places in the world. In the from 40% to 19% when subjected to an L-SF blend of
water projects which are present in various places in Australia, 11–15%. At L-SF 5–10%, the angle of internal friction
South Africa, India, Iran, Thailand, Middle East, and South concerning shear strength parameters would increase
America, big problems are experienced with dispersive clays. from 5.80° to 24.75°.
Problems associated with dispersive soils are reported from In the study carried out by Al-Zairjawi (2009), the impact
many parts of the world. A number of earthen dams, hydraulic of a cement and silica fume additive on the compaction char-
structures, and roadway embankments have suffered damage or acteristics and strength of a high plasticity clayey soil was
collapsed due to erosion problems (NRC 1983; Indraratna et al. examined. Test results showed that 8% cement + 6% silica
1991; Umesha et al. 2009; Abbasi and Nazifi 2013; Goodarzi fume decreased the maximum dry density from 1.64 to
and Salimi 2015; Hassanlourad et al. 2017). 1.55 g/cm3 and increased the optimum water content from
Studies are being made for improving swelling and disper- 19 to 23%. In addition, it has been observed that the curing
sive soils which are observed in various countries in the time has a positive effect on the unconfined compressive
world, by using additives such as cement, lime, gypsum, strength of the soil.
natural pozzolana, industrial wastes, magnesium chloride so- Goodarzi et al. (2015) also studied the effect of mi-
lution, and fly ash with different contents (Ouhadi and cro silica–lime admixture on geotechnical properties on
Goodarzi 2006; Murty and Praveen 2008; Yilmaz and the expansive soils and concluded that highly expansive
Civelekoglu 2009; Harichane et al. 2010; Vakili et al. 2013; clayey soils can be stabilized satisfactorily by a large
Turkoz et al. 2014; Latifi et al. 2016; Hassanlourad et al. amount of lime (at least 10%) and following adequate
2017; Moravej et al. 2018). curing due to development of the pozzolanic reaction.
In recent years, in order to improve engineering features of Also, with a further increase in the lime content, the
problematic clay soils, industrial by-products are used fre- pozzolanic activity cannot continuously occur and the
quently. Usage of additives such as volatile ash, rice husk, excessive addition of lime alone causes a reduction in
and silica fume for stabilization purposes provides both envi- the mechanical capacity. Moreover, it is found that the
ronmentalist and economic benefits. On the other hand, it is addition of silica fume (SF) alone, even up to 30%, has
observed that usage of these additives together with lime less effect on the engineering parameters of Na-smectite.
turned out to be more effective rather than when used alone. On the other hand, adding lime–silica fume mixture is
Lime improves the efficiency of stabilization considerably. very effective and improves the soil behavior with a
Some researchers (El-Aziz et al. 2004; Fattah et al. 2015a; lower percentage of lime.
Savas 2016; Alrubaye et al. 2017) have stated that after usage In the investigation performed by Bharadwaj and
of additives together, swell and dispersive features of Trivedi (2016), a series of laboratory experiments have
problematic soils have improved and that their shear strength been conducted on samples with 0, 5, 10, and 15% of
parameters and workability have significantly improved. For silica fume by weight of dry soil. The test results showed
example, in the study conducted by Savas (2016) on the sta- a significant change in consistency limits of samples con-
bilization of dispersive soils, it has been shown that with the taining silica fume. Also, the differential free swell de-
addition of 3% of natural pozzolanic additive to 2% lime con- creased from 48.46 to 9% showing an appreciable de-
tent, both swelling potential and compressibility of soils were crease in swelling behavior. It is seen that swelling poten-
reduced, and similarly Alrubaye et al. (2017) have shown that tial decreased with an increase in percent silica fume treat-
shear strength parameters of soils gave bigger values with ment. Amina and Rani (2017) used hydrated lime (Ca
soil + lime + silica fume mixture series when compared with (OH)2) and silica fume as a by-product material for stabi-
lime + soil and silica fume + soil mixture series. lizing the dispersive soil. They concluded that the addition
Arab J Geosci (2018) 11: 735 Page 3 of 14 735

of suitable additives such as lime and lime + silica fume 422-63), consistency limits (ASTM D 4318-00), and spe-
caused a significant decrease in the dispersive characteris- cific gravity (ASTM D 854-00) tests were performed to
tics of the soil. From the modified free swell test, crumb characterize the soil sample. Based on the identification test
test, and double hydrometer test, it was observed that 5% results, the sample was classified as CH (high plasticity
lime and 2% lime + 12% silica fume decreased the disper- clay) according to the Unified Soil Classification System
sive characteristics of the soil. (USCS) (ASTM D 2487-00). The grain-size distribution,
Silica fume (SF) is a by-product material produced in physical properties, and some chemical compositions of
large amounts throughout the world from the manufac- the soil sample are presented in Fig. 1 and Table 1, respec-
ture of silicon or ferrosilicon on alloys. The proper dis- tively. The engineering properties of the clay sample were
posal of SF, as an industrial waste, is one of the major measured according to American Society for Testing and
issues for environmentalists since leaving it directly in Materials methods (ASTM 1994). X-ray diffraction (XRD)
the environment may cause severe health problems analysis was carried out on the clay sample in Anadolu
(Zhang et al. 2016). University Material Science Laboratory (Fig. 2).
Silica fume which is obtained as a by-product in the pro- According to the result of XRD analysis, it could be said
duction of silicon and ferrosilicon alloys has got very small that illite is the dominant clay mineral.
spherical particles and as it contains 90% of amorphous silica, Chen (1988) classified a plasticity index over 35 as a very
it is a substance having high pozzolanic value. Silica fume is high swell potential, 29–35 as high, 10–35 as moderate, and
obtained by degrading quartz at 2000 °C in an open electrical 0–15 as low. On the basis of the physical properties, the swell
arc furnace. SiO, which turns into gas under high temperature, potential of the sample can be classified as very high.
are accumulated as SiO2 grains as they are oxidized during the It is seen from Table 1 that the values of Na (%), ESP, and
cooling process. SiO2, which does not crystallize due to sud- SAR, which are shown as the reason for dispersive soil be-
den cooling process, is formed as very small grains in an havior, are significantly high. According to Knodel’s (1991)
amorphous structure. classification based on the exchangeable sodium percentage
Total annual silica fume production in the world is (ESP), he has stated that soils with an ESP value higher than
around 1 million tons. Moreover, 130,000 tons of this 10 can be classified as dispersive soils. For this reason, this
amount are being produced in the USA and 120,000 tons soil sample also has dispersive characteristics.
are being produced in Norway. In North America, there are
13 plants where silicon, silica fume, and ferrosilicon are Additives
being produced. Silica fume can also be obtained from
other production ways which release silicon such as In this study, two additives which are hydrated lime and
f e r r o c h r om i u m ( F e C r ) , f e r ro m a n g a n e s e ( F e M n ) , silica fume have been used. Silica fume is obtained from
ferromagnesium (FeMg), and calcium silicon. Silica fume Antalya Elektrometalurji A.Ş. Silica fume is collected
has very small grains and as it has nearly 93% of SiO2 in its with dust catchers having special filters in ferrosilicium
structure, it is a very powerful pozzolanic substance (FeSi) and silicoferrochromium (SiFeCr) furnaces at
(Papworth 1997). Silica fume (SF), which has low unit Antalya Elektrometalurji A.Ş. factory as a by-product in
weight, low compactibility, and high pozzolanic activity, Turkey. Lime obtained from standard manufacturers is
is widely used for the improvement of problematic soils used for the treatment of the studied soil. The chemical
(Goodarzi et al. 2015). properties of the additives used in this study are shown
Each feature in the literature associated with the improve- in Table 2. Table 2 shows a significantly high content of
ment of the dispersive and swell properties of clay soil with amorphous SiO2 in SF, with small amounts of alumina,
additives was separately evaluated. In this study, the effects of iron, calcium, and alkali oxides. Lime contains a lower
lime additive mixed with different proportions of silica fume SiO2 and a higher CaO when compared to SF.
on swelling potential, dispersive, and strength characteristics Percentage of lime that is added to a soil can vary
of a natural clay soil were investigated. depending on whether it is used to improve stabilization
of soil or its workability. While small percentages are
sufficient to improve workability, an adequate quantity
Materials of lime percentage is needed to improve stabilization.
For determining the minimum percentage of lime required
Soil for stabilization of soil, pH experiment which is recom-
mended by Eades and Grim (1966) has been conducted.
The soil sample considered in this study was taken from In this experiment, when pH value of soil and lime mix-
the Afyon province located in the Central Anatolia Region ture reaches up to 12.4, the percentage of lime required
of Turkey. Sieve analysis, hydrometer analysis (ASTM D for the reaction is determined. When lime with a ratio of
735 Page 4 of 14 Arab J Geosci (2018) 11: 735

Fig. 1 Grain-size distribution of

the used soil sample

3% is added to the soil, pH has reached to 12.4 and in Methods

increased percentages of lime, no significant change was
observed in pH value (Fig. 3). Sample preparation

The soil sample used in the study has firstly been dried for
Table 1 Physical and some chemical characteristics of the considered 24 h at 105 °C in an oven, and after that, it has been ground
soil
and passed through a No. 4 sieve to obtain a uniform distri-
Property Value bution. Different amounts of the silica fume (0, 1, 3, 5, 10, 15,
and 20% by dry weight of the soil) and lime (3% by dry
Physical Grain size weight of the soil) were added and mixed into the prepared
Gravel (%) – soil samples. Soil–additive mixtures were prepared for each
Sand (%) 13 soil sample by mixing in an optimum amount of water, which
Silt (%) 54 was determined at the standard Proctor energy level (ASTM D
Clay (%) 33 698). As a result, compaction characteristics needed to prepare
Atterberg limits the samples for use in swell percentage, swell pressure, crumb,
Liquid limit, LL (%) 69 pinhole, and strength tests for each additive level were
Plastic limit, PL (%) 32 determined.
Plasticity index, PI (%) 37
Specific gravity, Gs 2.73
Classification (USCS) CH Experimental study
Activity, A 1.12
Maximum dry density (Mg/m3) 1.448 Swell potential referred to the swell percentage and swell
Optimum water content (%) 23.20 pressure of clays is best determined through direct measure-
Chemical pH 8.55 ments (Nelson and Miller 1992). The swell percentage and
Total dissolved salt, TDS (mg/L) 143.19 swell pressure tests were performed using direct methods.
Sodium adsorption ratio, SAR 53.36 Swell percentage test determines the amount of vertical heave
Sodium percentage, Na (%) 92.30 that will occur when the soil is wetted. This test is performed
Exchangeable sodium percentage, ESP (%) 33.85 on specimens prepared by compressing in the desired density
Conductivity (mmhos/cm) 20.60 and water content. The swell percentage is defined as the ratio
between the starting length of the sample and the final
Arab J Geosci (2018) 11: 735 Page 5 of 14 735

Fig. 2 XRD pattern of the soil

sample

deformation of the sample after being soaked in water under To determine the dispersive characteristics of the samples,
7 kPa pressure for 24 h or until swelling is complete (ASTM D pinhole and crumb tests were performed following the stan-
4829). Modified potential volume change (PVC) meter equip- dard procedures of the United States Bureau of Reclamation
ment that is developed by Turkoz et al. (2014) was used to (USBR) (USBR 5400 1989; USBR 5410 1989). The crumb
determine the swell pressure. There is no standard procedure test yields good qualitative results and is used to determine the
for the PVC meter test, so the method recommended by potential erodibility of clay soils.
Lambe (1960) was used. The PVC meter test involves deter- Unconsolidated–undrained (UU) test method was used to
mining the pressure arising from the inhibited swell deforma- evaluate the shear strength parameters of the samples with
tion that develops after saturating the compacted soil sample additives (ASTM D 2850). The samples were prepared in
with water. Immediately after, the samples placed in equip- stainless steel tubes so that the ratio of their height to their
ment were soaked in water, and the swell percentage and swell diameter was 2 (76 mm height and 38 mm diameter) and
pressure were measured at a series of time intervals (0, 0.17, compressed to the desired compaction characteristics of each
0.33, 0.50, 0.67, 0.83, 1, 2, 4, 8, 15, 30, 60, 120, 240, 360, … additive level. The samples were removed from the tubes,
min) using digital deformation meters connected to the data placed in plastic bags, and cured for 7 and 28 days in vacuum
logger. Swell percentage and swell pressure tests were per- desiccators. These tests were performed under confining
formed on the soil specimens prepared by compression in a
ring with a height of 2 cm and a diameter of 7 cm.

Table 2 Chemical compositions of silica fume and hydrated lime

Property Silica fume Hydrated lime

(%) (%)

SiO2 85.0–95.0 6.00

Al2O3 1.0–3.0 1.70
Fe2O3 0.5–1.0 0.70
CaO 0.8–1.2 86.90
MgO 1.0–2.0 0.70
Na2O 0.1–0.3 0.06
K2O – 0.18
Loss of ignition 0.5–1.0 6 (max)
Fig. 3 The effect of lime content on pH value of soil sample
735 Page 6 of 14 Arab J Geosci (2018) 11: 735

pressures of 100, 200, and 300 kPa, prior to curing and then has stated that LL value of clay soils with high plasticity
after curing for 7 and 28 days. decreased with increasing quantity of lime.
Finally, in addition to these experiments, in order to inves- Figure 5 shows the changing caused by additive content on
tigate the impact of additives on soil structure, SEM analyses the soil class. Soil class has gradually changed in the arrow
are conducted. In the analysis, ZEISS branded SUPRA 50 VP direction depending on the increase in the additive content. As
model Scanning Electron Microscope (SEM) has been used. mentioned by Abd El-Aziz et al. (2004), SF coats and binds all
clay particles which possess little cementitious value and large
particles which called the pozzolanic reaction between SF and
aluminous material. According to the soil classification sys-
Results and discussions tem, the soil class has changed as MH (high plasticity silt).
Similar results have been seen in the studies conducted by
Atterberg limits Kalkan (2011) and Harichane et al. (2018).

Atterberg limits tests were performed on each soil–additive Compaction test

mixture (ASTM D 4318). Figure 4 shows the effect of addi-
tive contents on Atterberg limits. It can be noticed that there is In order to determine the compaction characteristics used
a decrease in liquid limit and plasticity index with the addition in the preparation of the soil samples with additives,
of silica fume. Liquid limit (LL) value has decreased from 69 experiments were made at Standard Proctor energy level.
to 67% with 3% lime additive and it has reduced to 60% with Compaction curves which were obtained as a result of
3% L + 10% SF additives. No significant reduction in LL was compaction experiment that was conducted depending on
observed with increasing lime + silica fume additive percent- additive content are shown in Fig. 6 and changes in
ages. Plastic Limit (PL) value has increased from 32 to 34% optimum water content (OWC) and maximum dry densi-
with 3% lime additive and it has risen to 37% with 3% L + ty (MDD) values are shown in Table 3. As can be seen
10% SF additive levels, while PL has not changed with in- in Fig. 6 and Table 3, the MDD value decreases while
creased percentages of silica fume. Plasticity index (PI) value, the OWC value increases with increasing percentage of
which is defined as the difference between LL and PL water additives. A remarkable reduction in MDD value was
content, has shown an important reduction by decreasing from observed at 3% L + 10% SF additive level. MDD value
37 to 23% especially with the additive level of 3% L + 10% of 1.448 Mg/m 3 without additive has reduced to
SF. The decrease in plasticity index indicated an improvement 1.376 Mg/m3 at 3% L + 10% SF additive level and it
of the soil characteristics. In the study conducted by Harichane has not changed significantly with increasing percentage
et al. (2018) for the same soil class by only using the lime of additives. The decrease in MDD value was caused by
additive, it was seen that with 10% lime additive, the liquid additional gap volume that has developed due to low
limit has decreased from 84.8 to 76.4%. Attoh-Okine (1995) specific gravity (Gs) of samples with additives and grain
dimensions. The 3% L additive was effective in the in-
crease in OWC, whereas this increase was limited to
increasing SF additive percentages. Due to flocculation
and agglomeration, lime and silica fume additives around
clay grain surfaces reduce the specific surface area and
they require more water for pozzolanic reaction.

Swell potential test

Changes in swell percentage–time and swell pressure–time

relations with the additive content are shown in Figs. 7 and
8, respectively. From Figs. 7 and 8, it is seen that both the
swell percentage and the swell pressure values reached a final
value at a longer time in the case of no addition of the additive.
These variables reach their final values sooner together with
the addition of additives. It can be stated that this kind of
behavior could provide a big advantage with regard to appli-
cation. Especially in fillings which are constructed by
compressing in layers, providing stabilization in a short time
Fig. 4 Variation of Atterberg limits with different additive contents can be evaluated as superiority.
Arab J Geosci (2018) 11: 735 Page 7 of 14 735

Fig. 5 The effect of additive

contents on the sample
classification

Table 4 presents the changes in swelling potential as a Pinhole and crumb tests
whole. From Table 4, it is observed that both swell percentage
and swell pressure decreased depending on increasing addi- Crumb and pinhole tests were conducted separately for
tive content. At 3% L additive level, while there is a reduction each additive level with the aim to analyze the change in
of 32% in swell percentage and a reduction of 29.2% in swell dispersibility characteristic of the soil sample. Crumb test
pressure, this reduction has become more apparent in the silica results are collectively given in Fig. 9. The colloidal
fume additive percentages combined with lime. Especially, at clouding, which is an indicator of the dispersive nature of
3% L + 10% SF additive level, there has been a reduction of the soil sample without additive, was densely observed, and
95% in swell percentage and a reduction of 94% in swell the crumb test class was determined as K3. As can be seen
pressure. No significant changes have occurred in SF additive from Fig. 9, silica fume and lime additive reduced
percentages higher than 10%. The reason for this situation has clouding. Colloidal clouding decreased and crumb class
been evaluated such that silica fume is a pozzolanic substance was determined as K2 at 3% lime additive. At 3% lime
and that there is no sufficient amount of lime to activate a additive and the addition of increased amounts of silica
binding feature of silica with increasing silica fume fume, colloidal clouding was almost completely eliminated
percentages. and crumb test class was defined as K1.

Fig. 6 Compaction curves of

samples with different additive
contents
735 Page 8 of 14 Arab J Geosci (2018) 11: 735

Table 3 Effects of the additive content on the compaction dispersive properties of the soil sample are given in Table 5.
characteristics
In Table 5, it can be seen that dispersive features of the sample
Additive content OWC (%)a MMD (Mg/m3)b Gs have improved depending on increasing amounts of additives.
By using silica fume together with lime, dispersibility of sam-
0% 23.2 1.448 2.73 ple has considerably decreased.
3% L 27.8 1.418 2.68
3% L + 1% SF 29.3 1.403 2.67
3% L + 3% SF 29.5 1.400 2.67 Unconsolidated–undrained test
3% L + 5% SF 30.0 1.395 2.67
3% L + 10% SF 30.1 1.376 2.63 The unconsolidated–undrained (UU) triaxial compression test
3%L + 15%SF 30.6 1.370 2.63 was used to evaluate the shear strength parameters of the sam-
3% L + 20% SF 30.7 1.368 2.63 ple with and without the additive. These tests were performed
a
under confining pressures of 100, 200, and 300 kPa prior to
Optimum water content
b
curing and then after curing for 7 and 28 days. The effects of
Maximum dry density
the additive content and the curing time on the shear strength
parameters of the sample are presented in Table 6. It is ob-
However, it was understood that the more than 10% of the served that internal friction angle (Ø) and especially cohesion
silica fume was not able to bind with the soil and lime because (c) values increased depending on additive ratios and curing
silica fume transmitted its color to water at the rates of 3% L + periods. This situation reveals the positive impact of increased
15% SF and 3% L + 20% SF. When this situation is evaluated additive amounts and curing period on strength. In the study
with respect to crumb tests, it can be seen that 3% L + 10% SF conducted by Turkoz and Vural (2013), it is mentioned that
additive ratios are sufficient to reduce dispersion. curing period was more effective on increase of strength es-
Figure 10 shows the time–flow rate graphic obtained from pecially in clays with high plasticity. The impact of curing
pinhole test that is conducted depending on the percentage of period on cohesion and internal friction angle can be more
the additive. According to the pinhole test results, it can be clearly seen in Figs. 11 and 12, respectively. In clay soils,
observed that the soil sample shows dispersive (D2) proper- the cohesion has a strong influence on the shear strength,
ties. With increasing amounts of additives, the dispersibility and increasing cohesion is assumed to correspond to increas-
properties have changed. It exhibited intermediate-dispersive ing shear strength (Turkoz et al. 2014).
soil (ND3) behaviors at the additive percentages of 3% L and It is seen that there is a gradual increase in cohesion value
3% L + 1% SF and it was classified as non-dispersive soil with the increasing additive percentages (Fig. 11). While the
(ND2) at the additive percentages of 3% L + 3% SF and 3% increase starting at 3% L additive does not change much at
L + 5% SF. Finally, it demonstrated features of a non- additive percentages up to 3% L + 5% SF additive level, it has
dispersive soil (ND1) at the 3% L + 10% SF additive level a new increasing tendency at 3% L + 10% SF level and it has
and increased SF additive percentages. The changes of not changed much at additive percentages within the interval

Fig. 7 Swell percentage vs. time

plots for the soil samples mixed
with different additive contents
Arab J Geosci (2018) 11: 735 Page 9 of 14 735

Fig. 8 Swell pressure vs. time

plots for the soil samples mixed
with different additive contents

up to 3% L + 20% SF level. The cohesion value of 81.7 kPa in

the state without additives reached 169.2, 253.9, and 280 kPa
Table 4 Swell potential test results at various additive contents in 3% L, 3% L + 10% SF, and 3% L + 20% SF additives
Additive content Swell percentage (%) Swell pressure (kPa)
levels, respectively, after 7 days of curing. For the same addi-
tive levels, after a curing period of 28 days, cohesion has
0% 12.2 96.1 reached to the values of 202.2, 298, and 325 kPa. As can be
3% L 8.3 68.0 seen from these values, a significant increase in cohesion val-
3% L + 1% SF 5.7 49.7 ue has occurred at the level of 3% L + 10% SF after a curing
3% L + 3% SF 3.8 30.6 period of 7 days. As it was stated in the study conducted by
3% L + 5% SF 1.8 18.1 Latifi et al. (2016), curing on first 7 days had significant im-
3% L + 10% SF 0.6 5.9 portance on the increase of strength for clays with high plas-
3% L + 15% SF 0.4 3.1 ticity. This result is due to the pozzolanic reaction between the
3% L + 20% SF 0.2 2.4 lime and silica fume, which is more effective with soil parti-
cles (Alrubaye et al. 2017).

Fig. 9 Crumb test results of the

soil samples mixed with different
additive contents
735 Page 10 of 14 Arab J Geosci (2018) 11: 735

Fig. 10 Pinhole test results of the soil samples mixed with different additive contents

As can be seen in Fig. 12, the value of the internal friction produced the optimal results in swell, dispersibility, and
angle varied in a narrow range depending on the increasing strength tests.
percentage of additive and curing time. There was no signif- Figure 13 shows the images obtained from the analyses of
icant change in the value of the internal friction angle (22.1° to natural clay (Fig. 13a), of the clay sample with an additive
24.2°) especially without curing. This value was partially in- content of 3% lime (Fig. 13b), and of the clay sample with an
creased as a result of the pozzolanic reaction that was accom- additive content of 3% lime + 10% silica fume (Fig. 13c) tak-
panied by the increased cure time. en at the same magnification. The natural sample shows a
more disperse and dense structure (Fig. 13a). In case of 3%
L additive content, it exhibits a more aggregated structure
Scanning electron microscopy (SEM) analysis (Fig. 13b). The 3% L + 10% SF additive content caused the
particles to reorganize and the structural integrity to increase
To determine the microstructural changes in the soil, SEM (Fig. 13c). The resultant agglomeration reduces the interac-
analyses were performed on the natural samples, an additive tions between the surface areas and water, which in turn
content of 3% lime, and 3% lime + 10% silica fume. The 3% changes the engineering properties. This finding can be ex-
lime + 10% silica fume additive content was chosen because it plained by the increased silica concentration caused by the

Table 5 Dispersibility test results at various additive contents

Test Additive content

0% 3% L 3% L + 1% SF 3%L + 3% SF 3% L + 5% SF 3% L + 10% SF 3% L + 15% SF 3% L + 20% SF

Crumb test class K3 K2 K1 K1 K1 K1 K1 K1

Pinhole test class D2 ND3 ND3 ND2 ND2 ND1 ND1 ND1

D1 and D2, dispersive

ND3 and ND4, intermediate soil
ND1 and ND2, non-dispersive soil
K3 and K4, dispersive
K2, intermediate soil
K1, non-dispersive soil
Arab J Geosci (2018) 11: 735 Page 11 of 14 735

Table 6 Shear strength

parameters at various additive Additive content Without curing 7-day curing 28-day curing
contents and curing time
c (kN/m2) Ø (°) c (kN/m2) Ø (°) c (kN/m2) Ø (°)

0% 81.7 22.1
3% L 101.4 23.1 169.2 23.6 202.2 24.0
3% L + 1% SF 110.0 23.2 183.4 24.0 209.4 24.3
3% L + 3% SF 115.0 23.7 191.4 24.2 214.8 25.8
3% L + 5% SF 124.0 24.5 215.0 24.4 230.0 26.5
3% L + 10% SF 128.2 23.9 253.9 24.6 298.0 27.3
3% L + 15% SF 132.0 24.2 264.4 24.8 316.0 27.6
3% L + 20% SF 140.0 23.0 280.0 25.0 325.0 28.0

Fig. 11 Effects of additive

content and curing time on
cohesion

Fig. 12 Effects of additive

content and curing time on
internal friction angle
735 Page 12 of 14 Arab J Geosci (2018) 11: 735

condition is almost invisible and a large quantity of cementi-

tious matter appeared on the L + SF treatment (Fig. 13c), as
stated by Goodarzi et al. (2016). Such changes can be referred
to the beneficial effects of L + SF on the additional formation
of cementing phase. The superior influences of L + SF are
ascribed to the immediate growth of silicate gel and to extend
synthesis of the new cementitious compound (i.e., CSH)
which is mainly formed by Ca+2 of lime and SiO2 of silica
fume. This can rapidly block off the soil voids and greatly
interlock the clay particles together, resulting in the continu-
ous development of soil geotechnical properties as the addi-
tive content increases (Goodarzi and Salimi 2015). The grain
sizes have increased due to enlargement of soil structure and
agglomeration of the material (Fig. 13c). On treatment with
lime, clays aggregate to form a more coherent mass due to
pozzolanic reaction and cation exchange reactions. In further
treatment with silica fume along with lime, the cementitious
compounds are formed and the silica fume imparts a mechan-
ical bonding which forms well-developed floccules and shows
a more porous nature. Thus, this type of aggregation and im-
provement in porosity brings the desired improvement in the
engineering properties of the soil (Amina and Rani 2017).
Thus, depending on increasing grain size and dimensions
of aggregation, higher values were obtained with series having
3% lime + 10% silica fume additives when compared with
those having 3% lime additives in strength experiments.

Conclusion

It is determined that lime and silica fume additives have a

significant impact on dispersibility, swell potential, and
strength of the soil sample revealing swell and dispersive fea-
tures. On the other hand, it was seen that using silica fume,
which is a waste material, together with lime rather than being
used alone, has been much more effective and it has success-
fully improved the engineering features of this soil having
high plasticity. In this study, the following conclusions were
reached:

1. In the increased L + SF additive ratios, the plasticity index

decreased as a result of the decrease in the liquid limit
value and the increase in the plastic limit value. As a
Fig. 13 SEM image of the samples: a the natural clay sample, b clay + result, the workability of the soil has improved and the
3% L, and c clay + 3% L + 10% SF soil class has changed.
2. The optimum water content value of the soil increased and
addition of silica fume and the subsequent increase in particle the maximum dry density value (MDD) decreased with
size due to the rapid flocculation of particles. With the addition increasing the amount of silica fume added to the mixture
of silica fume, the arrangement of the clay particles transforms while keeping the 3% lime ratio constant. Significant re-
from a dispersed structure to a flocculated structure and some duction in MDD value was observed at 3% L + 10% SF
new cementing compounds are formed. The surfaces of parti- additive level.
cles have been generally coated by the hydration gels, and the 3. At 3% L + 10% SF additive level, a reduction of 95% was
soil pores have been mainly filled. Besides, the flocculated observed in swell percentage and a reduction of 94% was
Arab J Geosci (2018) 11: 735 Page 13 of 14 735

observed in swell pressure. No significant changes have Basma AA, Al-Hamoud AS, Husein A (1995) Laboratory assessment of
swelling pressure of expansive soils. Appl Clay Sci 9:355–365.
occurred in silica fume additive percentages that were
https://doi.org/10.1016/0169-1317(94)00032-L
more than 10%. Bharadwaj S, Trivedi MK (2016) Impact of micro silica fume on engi-
4. When the test results were compared in terms of neering properties of expansive soil. International Journal of Science
dispersibility, an improvement was seen in the dispersive Technology & Engineering 2(12):435–440
properties of the soil sample at silica fume additive level Buhler R, and Cerato AB (2007) Stabilization of Oklahoma expansive
soils using lime and class C fly ash. GeoDenver: New Peaks in
of 10%, with 3% lime content. The expected effect was Geotechnics. GSP 162: Problematic Soils and Rocks and In Situ
not observed when the additive percentage was increased Characterization. Denver, CO, Feb. 18–21, 2007. CD Proceedings
due to the presence of free silica fume, which is unable to Chen FH (1988) Foundation on expansive soils. Elsevier, Amsterdam, the
react with adequate lime. Netherlands
Djellali A, Houam A, Saghafi B (2017) Indirect estimation of swelling
5. Depending on improved structural integrity and dimen-
pressure of clayey subgrade under pavement structures. Arab J Sci
sion being observed as a result of SEM analysis, higher Eng 42(9):3991–3999. https://doi.org/10.1007/s13369-017-2546-7
values revealed with series having 3% lime + 10% silica Dua Y, Lib S, Hayashic S (1999) Swell–shrinkage properties and soil
fume additives when compared with those having 3% improvement of compacted expansive soil, Ning-Liang highway,
China. Eng Geol 53:351–358. https://doi.org/10.1016/S0013-
lime additives in strength experiments.
7952(98)00086-6
Eades JL, Grim RE (1966) A quick test to determine lime requirements
As the population increases, and with it the number of built for lime stabilization. Highw Res Rec 139:61–72
structures, the dimensions of the swelling and dispersive soil El-Aziz A, Abo-Hashema M, and El-Shourbagy M (2004) The effect of
problems will also grow. The combination of silica fume and lime silica fume stabilizer on engineering properties of clayey sub-
grade. Paper presented at the Engineering Conference of Mansoura
lime can be successfully utilized as an additive to increase the University Faculty, Sharm: 96–104
efficiency of soil stabilization from the economic, technical, Fattah MY, Al-Saidi AA, Jaber MM (2015a) Improvement of bearing
and environmental point of views. capacity of footing on soft clay grouted with lime–silica fume mix.
Geomech Eng 8(1):113–132. https://doi.org/10.12989/gae.2015.8.
1.000
Fattah MY, Al-Saidi AA, Jaber MM (2015b) Characteristics of clays
stabilized with lime–silica fume mix. Ital J Geosci 134(1):104–
113. https://doi.org/10.3301/IJG.2014.36
References Goodarzi AR, Goodarzi SH, Akbari HR (2015) Assessing geo-
mechanical and micro-structural performance of modified expansive
Abbasi N, Nazifi MH (2013) Assessment and modification of Sherard clayey soil by silica fume as industrial waste. Transactions of Civil
chemical method for evaluation of dispersion potential of soils. Engineering 39(C2):333–350
Geotech Geol Eng 31:337–346. https://doi.org/10.1007/s10706- Goodarzi A, Salimi M (2015) Stabilization treatment of a dispersive clay-
012-9573-7 ey soil using granulated blast furnace slag and basic oxygen furnace
Abd El-Aziz MA, Abo-Hashema MA, and El-Shourbagy M (2004) The slag. Appl Clay Sci 108:61–69. https://doi.org/10.1016/j.clay.2015.
effect of lime–silica fume stabilizer on engineering properties of 02.024
clayey subgrade. Fourth Mansoura International Engineering Goodarzi R, Akbari HR, Salimi M (2016) Enhanced stabilization of high-
Conference (4th IEC), Faculty of Engineering, Mansoura ly expansive clays by mixing cement and silica fume. Appl Clay Sci
University, Egypt, Paper No. 96 132-133:675–684. https://doi.org/10.1016/j.clay.2016.08.023
Abdullah WS, Alshibli KA, Al-Zou’bi MS (1999) Influence of pore Harichane K, Ghrici M, Missoum H (2010) Effect of the combination of
water chemistry on the swelling behavior of compacted clays. lime and natural pozzolana on the durability of clayey soils.
Appl Clay Sci 15:447–462. https://doi.org/10.1016/S0169- European J. Geotech. Eng. 15:1194–1210
1317(99)00034-4 Harichane K, Ghrici M, Kenai S (2018) Stabilization of Algerian clayey
Al-Rawas AA, Taha R, Nelson J, Al-Shap B, Al-Siyabi H (2002) A soils with natural pozzolana and lime. Periodica Polytechnica Civil
comparative evaluation of various additives used in the stabilization Engineering 62(1):1–10. https://doi.org/10.3311/PPci.9229
of expansive soils. Geotech Test J 25(2):199–209. https://doi.org/10. Hassanlourad M, Rokni MN, Hassanlo M, Badrlou A (2017) Dispersive
1520/GTJ11363J clay stabilised by alum and lime. International Journal of
Alrubaye AJ, Hasan M, Fattah MY (2017) Stabilization of soft kaolin GEOMATE 12(29):156–162. https://doi.org/10.21660/2017.29.
clay with silica fume and lime. Int J Geotech Eng 11(1):90–96. 93287
https://doi.org/10.1080/19386362.2016.1187884 Huang R, Wu L (2007) Stability analysis of unsaturated expansive soil
Al-Zairjawi MK (2009) Effect of adding cement and silica fume with slope. Earth Science Frontiers 14(6):129–133. https://doi.org/10.
cement on compaction properties and shear strength of clayey soil. 1016/S1872-5791(08)60007-X
Al-Qadisiyah Journal for Engineering Sciences 2(2):253–261 Indraratna B, Nutalaya P, Kuganenthira N (1991) Stabilization of a dis-
Amina SM, Rani V (2017) Control of dispersivity of soil using lime and persive soil by blending with fly ash. Q J Eng Geol Hydrogeol
silica fume. International Journal for Research in Applied Science & 24(3):275–290. https://doi.org/10.1144/GSL.QJEG.1991.024.03.03
Engineering Technology 5(5):386–391 Kalkan E (2011) Impact of wetting–drying cycles on swelling behavior of
ASTM (1994) Annual book of ASTM standards. Soil and rock, vol. 0 clayey soils modified by silica fume. Appl Clay Sci 52(4):345–352.
4.08. American Society for Testing and Materials, Philadelphia https://doi.org/10.1016/j.clay.2011.03.014
Attoh-Okine NO (1995) Lime treatment of laterite soils and gravels- Knodel PC (1991) Characteristics and problems of dispersive clay soils.
revisited. Constr Build Mater 9(5):283–287. https://doi.org/10. United States Department of the Interior Bureau of Reclamation,
1016/0950-0618(95)00030-J Materials Engineering Branch, Denver, USA, R91-09, p.17
735 Page 14 of 14 Arab J Geosci (2018) 11: 735

Lambe TW (1960) The character and identification of expansive soils. Sherard JL, Dunnigan LP, Decker RS (1976) Identification and nature of
Soil PVC Meter, Federal Housing Administration, Technical Studies dispersive soils. J Geotech Eng Div 102(4):287–301
Program, FHA 701 Shi B, Jiang H, Liu Z, Fang Y (2002) Engineering geological character-
Latifi N, Meehan C, Majid MZA, Horpibulsuk S (2016) Strengthening istics of expansive soils in China. Eng Geol 67:63–71. https://doi.
montmorillonitic and kaolinitic clays using a calcium-based non- org/10.1016/S0013-7952(02)00145-X
traditional additive: a micro-level study. Appl Clay Sci 132-133: Turkoz M, Tosun H (2011) A GIS model for preliminary hazard assess-
182–193. https://doi.org/10.1016/j.clay.2016.06.004 ment of swelling clays, a case study in Harran plain (SE Turkey). J
Moravej S, Habibagahi G, Nikooee E, Niazi A (2018) Stabilization of Environ Earth Sciences 63:1343–1353. https://doi.org/10.1007/
dispersive soils by means of biological calcite precipitation. s12665-010-0805-y
Geoderma 315:130–137. https://doi.org/10.1016/j.geoderma.2017. Turkoz M, Vural P (2013) The effects of cement and natural zeolite
11.037 additives on problematic clay soils. Sci Eng Compos Mater 20(4):
Murty VR, Praveen GV (2008) Use of chemically stabilized soil as cush- 395–405. https://doi.org/10.1515/secm-2012-0104
ion material below light weight structures founded on expansive Turkoz M, Savas H, Acaz A, Tosun H (2014) The effect of magnesium
soils. J Mater Civ Eng 20(5):392–400. https://doi.org/10.1061/ chloride solution on the engineering properties of clay soil with
(ASCE)0899-1561(2008)20:5(392) expansive and dispersive characteristics. Appl Clay Sci 101:1–9.
NRC (1983) Safety of existing dams, evaluation and improvement. https://doi.org/10.1016/j.clay.2014.08.007
National Research Council, National Academy Press Umesha TS, Dinesh SV, Sivapullaiah PV (2009) Control of dispersivity
Nelson JD, and Miller JD (1992) Expansive soils—problems and practice of soil using lime and cement. International Journal of Geology 1(3):
in foundation and pavement engineering. J. Wiley and Sons, New 8–16
York USBR 5400 (1989) Determining Dispersibility of Clayey Soils by the
Ouhadi VR, Goodarzi AR (2006) Assessment of the stability of a disper- Crumb Test Method, Earth Manual II, United States Department of
sive soil treated by alum. Eng Geol 85:91–101. https://doi.org/10. the Interior Bureau of Reclamation: 414–418
1016/j.enggeo.2005.09.042 USBR 5410 (1989) Determining Dispersibility of Clayey Soils by the
Papworth HF (1997) Production of silica fume. M.I.E. (Australia) Civil Pinhole Test Method. Earth Manual II, United States Department of
Engineering the Interior Bureau of Reclamation: 425–437
Penner D, Lagaly G (2001) Influence of anions on the rheological prop- Vakili AH, Selamat MR, Moayedi H (2013) Effects of using pozzolan
erties of clay mineral dispersions. Appl Clay Sci 19:131–142. and Portland cement in the treatment of dispersive clay. Hindawi
https://doi.org/10.1016/S0169-1317(01)00052-7 Publishing Corporation The Scientific World Journal, 10 pages,
Phanikumar BR (2009) Effect of lime and fly ash on swell, consolidation https://doi.org/10.1155/2013/547615
and shear strength characteristics of expansive clays: a comparative Yilmaz I, Civelekoglu B (2009) Gypsum: an additive for stabilization of
study. Geomech Geoeng 4(2):175–181. https://doi.org/10.1080/ swelling clay soils. Appl Clay Sci 44:166–172. https://doi.org/10.
17486020902856983 1016/j.clay.2009.01.020
Qi S, Vanapalli SK (2015) Hydro-mechanical coupling effect on surficial Yong and Warkentin BP (1996) Introduction to soil behavior. In: Chapter,
layer stability of unsaturated expansive soil slopes. Comput Geotech vol 7. Macmillan, New York, pp 151–175
70:68–82. https://doi.org/10.1016/j.compgeo.2015.07.006 Zhang Z, Zhang B, Yan P (2016) Comparative study of effect of raw and
Savas H (2016) Consolidation and swell characteristics of dispersive soils densified silica fume in the paste, mortar and concrete. Constr Build
stabilized with lime and natural zeolite. Sci Eng Compos Mater Mater 105:82–93. https://doi.org/10.1016/j.conbuildmat.2015.12.
23(6):589–598. https://doi.org/10.1515/secm-2014-0202 045