ORIGINAL PAPER

Testing and Modeling the Short-Term Behavior of Limeand Fly Ash Treated Sulfate Contaminated CL Soil

Ahmed Mohammed . Cumaraswamy Vipulanandan

Received: 15 December 2013 / Accepted: 22 April 2015

� Springer International Publishing Switzerland 2015

Abstract In this study, the effects of calcium sulfate

contaminated soil treatment methods on the index

properties, compacted soil properties, free swelling and

compressive stress strain relationship of a CL soil

obtained from the field was investigated. Calcium

sulfate concentration in the soil was varied up to 4 %

(40,000 ppm) and the soil samples were cured for seven

days at 25 �C and 100 % humidity before testing. With

4 % sulfate contamination the liquid limit and plasticity

index of the soil increased by 44 and 81 % respectively.

The free swelling increased from 7 to 18 % and the

compressive strength decreased by 65 % with 4 % of

calcium sulfate, Also the study investigated the effects

of fly ash (class C) and hydrated lime treatment on the

behavior of treated sulfate soils. X-ray diffraction

(XRD)was used to characterize the soil and the reaction

products of lime and fly ash treated soils. Based onXRD

analyses,major constituents of theCL soilwere calcium

silicate (CaSiO3), aluminum silicate (Al2SiO5), magne-

sium silicate (MgSiO3) and quartz (SiO2). Addition of

calcium sulfate resulted in the formation of calcium

silicate sulfate [Ternesite Ca5(SiO4)2SO4] and alu-

minum silicate sulfate [Al5(SiO4)2SO4]. Treatment

with lime resulted in the formation of ettringite

[Ca6Al2(SO4)3(OH)12�26H2O]. Treating with fly

ash resulted in the formation of calcium silicate

hydrate (CaSiO3H2O) and magnesium silicate hydrate

(Mg3SiO3H2O), cementing by products. Contaminated

soil treatment with lime and fly ash reduced the index

properties and free swelling and increased the short-

term compressive strength of the soil. Behavior of the

compacted sulfate soils, with and without treatment,

has been quantified using a unique model that was used

to represent both linear and nonlinear responses. Also

the model predications were compared with other

published data in the literature. Compressive stress–

strain relationships of the sulfate soil, with and without

lime and fly ash, have been quantified using a nonlinear

constitutive model. The constitutive model parameters

were sensitive to the calcium sulfate content and the

type of treatment. Compared to the fly ash treatment,

the lime treatment reduced the strain at peak stress

making the lime treated soil more brittle.

Keywords Calcium sulfate � Free swelling �Compaction � Compressive strength � Lime � Fly ash

1 Introduction

In the past few decades, it has been reported that

sulfate rich soils have become a challenge in

A. Mohammed � C. Vipulanandan (&)

Center for Innovative Grouting Material and Technology

(CIGMAT), Texas Hurricane Center for Innovative

Technology (THC-IT), Department of Civil and

Environmental Engineering, University of Houston,

Houston, TX 77204-4003, USA

e-mail: cvipulanandan@uh.edu

A. Mohammed

e-mail: Asmohammed2@uh.edu

123

Geotech Geol Eng

DOI 10.1007/s10706-015-9890-8

engineering projects because of swelling and also

chemically attacking the construction materials such

as steel and concrete (Mitchell and Dermatas 1992;

Petry and Little 1992; Kota et al. 1996; Rollings et al.

1999; Puppala et al. 2002). Sulfate heave distress

problems have been reported in Texas, Nevada,

Louisiana, Kansas, Oklahoma, and Colorado where

lime, fly ash and cement have been traditionally used

to treat the natural soils rich in sulfates (Kota et al.

1996; Rollings et al. 1999).

Hunter (1988) indicated that lime treated sulphate

bearing clay is risky even at relatively low sulphate

concentrations. Based on short-term (1–30 days)

studies it has been observed that treating the sulfate

soils with calcium-based chemicals such as lime and

ordinary Portland cement has resulted in increased

heave (Mitchell and Dermatas 1992; Petry and Little

1992).

Bell (1996) studied the linear shrinkage of mont-

morillonite clay with different percentages of lime

additive and showed that the shrinkage decreased with

the addition of lime, but the decrease was not a linear.

The study also showed that the unconfined compres-

sion strength did not increase linearly with the addition

of lime and excessive addition of lime actually

reduced the strength. During this treatment, miner-

alogical and microstructural changes occurred, which

lowered the plasticity and also enhanced the load-

bearing capacity and enable soil compaction to higher

density. Arabani and Veis (2007) observed that any

increase in lime content beyond 6 % had a negligible

effect on the compressive strength of a treated CL soil.

However, an increase in lime content up to 6 percent

resulted in a noticeable increase in compressive

strength and the results reported are for 1 day of

curing. Reaction of lime with alumino-silicate in

clays, produced hydrated cementitious products that

bonded to the soil particles and the results reported are

for 7 and 28 days of curing (McCarthy et al. 2009).

During this treatment, mineralogical and microstruc-

tural changes occurred, which lowered the plasticity

and also enhanced the load-bearing capacity and

enable soil compaction to higher density. Ramesh

(2009) concluded that adding barium chloride to lime

treated soil prevented the formation of ettringite and

thaumasite in lime treated soils containing sulphate.

Cokca (2001) used low calcium and high calcium

fly ash to stabilize expansive CH soils. The study also

included comparison with lime and cement treated

soils. The results showed that a the soil treated with

20 % fly ash had nearly the same stabilizing effect on

the swelling potential as a soil treated with 8 % lime.

Nalbantoglu (2004) investigated the effectiveness of

class C fly ash as a stabilizer to expansive CH soils and

the results reported are for 28 days. This study showed

that the plasticity index of high plasticity soils was

reduced but the effect was minimal with low plastic

soils. The study also found that fly ash was an effective

method to reduce the swelling potential for one of the

tested soil but had adverse effects with other soils.

Little et al. (2010) stated that civil engineers are at

times required to stabilize sulfate-bearing clay soils

with calcium-based stabilizers. Deleterious heaving in

these stabilized soils may result over time. Recently,

the potential of polymer treatment on a sulfate CL soil

was investigated and the polymer was effective in

modifying the soil properties in less than a day

(Mohammed and Vipulanandan 2014).

According to the studies summarized in Table 1

on clay soils stabilized with lime and fly ash, most of

the specimens were prepared and tested near opti-

mum moisture content (OMC %) from 1 to 30 days

of curing. Both 6 % lime and 10 % fly ash have been

used to treat the soil (Tables 1, 2). Calcium compo-

nents in the stabilizers such as lime, cement and fly

ash are known to react with free alumina and soluble

sulfates in the soils to form ettringite mineral a weak

sulfate mineral that will undergo significant heaving

when subjected to hydration. This heave, termed as

sulfate-induced heave in the literature, is known to

severely affect the performance of highways, run-

ways, parking lots, residential and industrial build-

ings, and other earth structures built on fly ash and

lime treated sulfate rich soils (Hunter 1988; Mitchell

and Dermatas 1992; Petry and Little 1992; Rajendran

and Lytton 1997; Rollings et al. 1999). The formation

of ettringite mineral (Ca6(Al(OH)6)2(SO4)326H2O) in

treated soils and its exposure to moisture variations

from seasonal changes result in differential heaving,

which in turn causes cracking of pavement struc-

tures built on the treated soils. If not addressed

immediately, this heave will further deteriorate the

structures to a condition where they need immediate

and extensive rehabilitation (Mitchell and Dermatas

1992; Petry and Little 1992). Hence there is a need

to develop relationships to quantify the effect of

calcium sulfate on the behavior of treated and

untreated soils.

Geotech Geol Eng

123

2 Objectives

The overall objective was to quantify the changes in

the properties of a field CL soil contaminated with

varying percentage of calcium sulfate up to 4 %

(40,000 ppm). Also of interest was to investigate the

treatment of sulfate contaminate soil with lime and fly

ash. The specific objectives are as follows:

1. Identify the changes in the soil clay mineralogy

due to calcium sulfate contamination and treat-

ment with lime and fly ash.

2. Quantify the changes in the index properties,

compaction properties, free swell and compres-

sive strength of a CL soil with vary amount of

calcium sulfate with and without treatment.

3. Quantify the compressive stress–strain relation-

ship of clay soil contaminated with calcium

sulfate up to 4 %, and treated with 6 % lime and

10 % fly ash.

3 Materials and Methods

3.1 Soil

Field clay soil sample was used in preparing the sulfate

soil. Atterberg limits, grain size distribution and

hydrometer, compaction, swelling and compressive

strength tests were performed according to ASTM

Standard. The results are summarized in Table 3.

3.2 Hydrated Lime

Lime for soil treatment applications is typically used

in the form of quicklime (CaO) or Hydrated lime

(Ca(OH)2). Quicklime (CaO) is manufactured by

chemical processes transforming calcium carbonate

(limestone—CaCO3) into calcium oxide (CaO). When

quicklime reacts with water it transforms into hydrated

lime. The hydrated lime (Ca (OH)2) reacts with the

clay particles and modifies the clay based on its

mineralogy.

3.3 Fly Ash Class C

ASTM C618-91 defines fly ash as the finely divided

residue that results from the combustion of ground or

powdered coal and is transported from the boiler byTable

1Summaryofclay

soilsstabilized

withlimeandflyash

Reference

Soiltype

Stabilizer

Application

%ofStabilizer

(bydry

weight)

Curingtime

(days)

Curing

temperature,

humidity

Water

content

forthestudy

Sivapullaiah

etal.(2000)

CL

Lim

eSulfatesoil

67and30

Notspecified

OMC

Edilet

al.(2002)

Clay

Fly

ash(class

C)

Roadway

over

softsubgrade

10

14and28

100%

humidity

Naturalwater

content

Acostaet

al.(2003)

Clay

Fly

ash(class

C)

SoilStabilization

0,10,18,30

14and28

100%

humidity

OMC

Harriset

al.(2004)

Clay

Lim

eSulfatesoil

61

25�C

OMC

PhaniandSharma(2004)

CH

Fly

ash(class

C)

Expansiveclay

0,5,10,15,20

––

OMC

Puppalaet

al.(2006)

Clay

Lim

eSulfatesoil

4Notspecified

(25–40)�C

OMC

Kumar

etal.(2007)

CH

Fly

ash

Expansiveclay

0,5,10,15,20

7,14and28

25�C

OMC

Aravindet

al.(2011)

CH,CL

Lim

eExpansiveclay

6and8

240�C

OMC

Rem

arks

Claysoils

Fly

ashand

limehavebeenused

Expansiveand

sulfatesoils

Mainly

6%

oflime

and10%

flyash

1–30days

Upto

40�C

and

100%

humidity

Mainly

OMC

was

used

OMCoptimum

moisture

content—

(standardcompaction)ASTM

D698

Geotech Geol Eng

123

flue gases. The physical and chemical characteristics

of fly ash vary greatly and mainly depend on the

combustion method, coal properties, and particle

shape of the fly ash. According to the literature, coal

from different sources will produce different combus-

tion products and these differences will influence the

effectiveness of fly ash as a soil stabilizer (Karim et al.

2011).

3.4 Test Methods

Soil was first dried in the oven at a temperature of

60 �C, crushed, pulverized and sieved to get sizes finerthan # 4 (0.075 mm) sieve. The pulverized soil was

then mixed with different percentage of calcium

sulfate and water. Soil samples were placed in

moisture tight bags and cured for 7 days at room

temperature before testing.

Four tests, Atterberg limits, compressive strength,

standard compaction and swell tests were conducted

on contaminated soil with different percentages (by

weight) of calcium sulfate up to 4 %. Sulfate soils

were treated separately with 6 % of lime and 10 % of

fly ash. Based on the compaction results, optimum

moisture content was used to prepare soil specimen for

the strength, swell and shrinkage tests. Cylindrical

specimens (1.400 dia. 9 2.800 length) were used for

UCS tests. After 7 days of curing, treated soil

specimens were subjected to unconfined compression

strength (UCS) and free swell tests.

3.4.1 XRD Analysis

AnX-ray diffraction (XRD) analyses was performed in

order to determine the effect of calcium sulfate

contaminated and treatment with lime and fly ash.

X-ray diffraction (XRD) was used to characterize the

field soil and reaction products of soilwith 4 %calcium

sulfate and treated contaminated soils with 6 % of lime

and 10 % of fly ash at 25 �C. The XRD pattern of the

particleswasobtainedbyusingSiemensD5000powder

X-ray diffraction (Jenkins and Snyder 1996) (Fig. 1).

Specimens for XRDwere prepared from air-dried soils

and ground with a mortar and pestle until the material

passed through sieve No. 200 (75 lm). The powder

(&2 g) was placed in an acrylic sample holder (3 mm)

Table 2 Literature review on sulfate soil stabilization using lime

Reference Soil type Lime (%) Sulfate content (ppm)

Mitchell (1986) Silty clay 4 Up to 15,000

Hunter (1988) Silty clay 4.5 43,500

Perrin (1992) Clays 5 2000–9000

Perrin (1992) Clays 6–9 % 14,000–25,000

Mitchell and Dermatas (1992) Sand mixed with 30 % clay content 3 3000–62,000

Rajasekaran et al. (1994) Marine clay with 52 % clay fraction 3 30,000

McCallister and Tidwell (1997) Expansive clays NA 2775

Sridhran et al. (1995) Black cotton soil 6 5000–30,000

Kota et al. (1996) Clayey subgrades 6–7 % [12,000

Burkart et al. (1999) Clays 6–9 % 233–18,000

Remarks Mainly clay soils Lime content varied

between 3 and 9 %

Sulfate content varied between

223 and 30,000 ppm (0.023–3 %)

Table 3 Test methods and physical properties of the CL soil

Property Test method Value

Passing sieve #200 (%) ASTM D 6913 64

Specific gravity ASTM D 854 2.66

LL (%) ASTM D 4318 40

PI (%) ASTM D 4318 19

OMC (%) (standard compaction) ASTM D 698 16.5

Max. dry density (g/cm3) ASTM D 698 1.52

Sand (%) ASTM D 6913 36

Silt (%) ASTM D 6913 45

Clay (%) ASTM D 6913 19

Free swelling (%) ASTM D 4546 7.0

Compressive strength (kPa) ASTM D 5102 573

Soil type ASTM D 2487 CL

Geotech Geol Eng

123

depth. All samples were analyzed by using parallel

beam optics with CuKa radiation at 40 kV and 30 mA.

Sampleswereground toafine textureusingamortarand

pestle andweremountedonplastic sample holders. The

average grain size of the soil crushed with an agate

mortar and pestle is expected to be in the 10–15 lmrange. All samples were scanned for reflections (2h)from 0� to 80� in steps of 0.02� and a 2 s count time per

step.

3.4.2 Free Swelling Test (FS)

The optimum water contents of clay soil con-

taminated with different percent of calcium sulfate

with and without 6 % lime and 10 % fly ash

treatment were determined by standard proctor tests

in accordance with ASTM D 698. For cyclic free

swelling tests, cylindrical samples were prepared at

optimum water content for the natural clay soil and

clay soil treated using 6 % lime and 10 % fly ash.

The swelling tests were carried out in the conven-

tional oedometer apparatus for natural soils and

treated soils using lime and fly ash. The consolida-

tion ring was pushed through the compacted soil

samples at optimum moisture content, using stan-

dard proctor and the extra material was carefully

trimmed in accordance with the ASTM Test

Method D 2435-96 to match the height of the

consolidation ring. The inside of the ring was

lubricated with silicone grease to minimize side

friction between the ring and the soil specimen.

After 7 days curing period at 100 % humidity and

25 �C temperature, the swelling tests were per-

formed based on ASTM D 4546. Filter papers were

placed on top and bottom of the soil specimen to

prevent finer particles from being forced into the

pores of the porous stones placed on both sides of

the specimen. Soils trimming were used to obtain

the moisture content data of the specimens before

start of the test. Each ring was weighed to cross

check the maximum dry unit weight. The specimens

were inundated with water and were allowed to

swell freely under 7 kPa (1psi) pressure. Free swell

measurements for first day were made at various

time intervals similar to that used for consolidation

test (ASTM D 4546-08). Each test was performed

for at least 7 days. Times versus swelling measure-

ments were then recorded. The percent free swell

may be expressed as:

FS% ¼ DHHo

� �� 100 ð1Þ

where: DH is height of swell due to saturation; Ho is

the original height of specimen.

3.4.3 Unconfined Compression Tests (UCS)

Unconfined compression tests were conducted ac-

cording to ASTM D-5102. The unconfined compres-

sive strengths were determined from the stress–strain

relationships. The natural soil contaminated with

calcium sulfate up to 4 % and treated sulfate soils

using 6 % of lime and 10 % of fly ash were compacted

at optimum moisture content and maximum dry

density in standard compaction molds. Thin wall

cylindrical steel samplers (1.400 dia.* 2.800 length) werecarefully pushed in- the standard compaction mold,

the soil samples were then extruded using a hydraulic

jack. Soils trimming were used to obtain the moisture

content data of the specimens before starting the test.

Soil samples were weighed to crosscheck the max-

imum dry unit weight. The specimens were cured for

7 days period at 100 % humidity and 25 �C tem-

perature. Loading was continued until the load

decreased with increasing strain, or until 20 % strain

was reached (Fig. 1).

4 Modeling

Hyperbolic relationship has been used to present the

behavior of cement and polymer modification soils

(Ata and Vipulanandan 1998; Usluogullari and

Vipulanandan 2011; Mohammed and Vipulanandan

2014). Relationship between index properties, com-

pacted soil properties, free swell strain and compres-

sive strength of the sulfate-contaminated soil with and

without treatments was investigated. Based on the

inspection of the test data following relationship is

proposed.

Y � Yo ¼X

Aþ B � X ð2Þ

where, Y is the soil property with varying sulfate

contaminating. Yo is the soil property without con-

tamination with calcium sulfate (natural CL soil). A

and B: are model parameters (Table 4). X is the

calcium sulfate concentration.

Geotech Geol Eng

123

Fig. 1 XRD Pattern of: a natural CL soil. b Soil contaminated with 4 % calcium sulfate. c Soil contaminated with 4 % calcium sulfate

treated with 6 % of lime. d Soil contaminated with 4 % calcium sulfate treated with 10 % of fly ash

Geotech Geol Eng

123

Fig. 1 continued

Geotech Geol Eng

123

Based on the experimental results the trends were

either linear or nonlinear with the calcium sulfate

content. As shown in Fig. 2 relationship proposed in

Eq. 2 can be used to represent various linear and

nonlinear trends based on the values of the parameters

A and B. When parameters A and B are positive the

relationship was hyperbolic. Linear relationship is

represented by Eq. 2 when B = 0 and A will take any

value. When parameters A and B are negative the

inverse hyperbolic relationship is obtained.

5 Results and Analyses

5.1 XRD

The soil had calcium silicate (CaSiO3) (2h peaks at

26.33�, 35.12�, 36.20� and 73.53�), aluminum silicate

(Al2SiO5) (2h peaks at 39.55�, 57.19� and 60.09�),magnesium silicate (MgSiO3) (2h peak at 43.17�) andquartz (SiO2) (2h peaks at 14.02�, 20.92�, 27.98�,45.87�, 64.21�, 68.36� and 75.68�). Addition of 4 %

calcium sulfate modified the clay composition with the

formation of ternesite (Ca5(SiO4)2SO4) (2h peaks at

20.85�, 28.43� and 29.88�) and aluminum silicate

sulfate (Al5(SiO4)2SO4) (2h peaks at 40.03� and

60.09�) and merwinite (Ca3Mg(SiO4)2) (2h peak at

46.0�). XRD analyses also showed that the fly ash

class C composition was quartz (SiO2), mullite

(Al2O3SiO2), aluminum oxide (Al2O3), iron oxide

(Fe2O3), calcium silicate (Ca2SiO4) and magnesium

silicate (Mg2SiO4). The major quartz peaks were

found at 2h of 19.75�, 20.92�, 23.23�, 27.59� and

50.21�. The main mullite peaks were at 2h of 12.77�,31.21�, 35.16�, 55.12�, 60.09� and 63.83�. Duchesneand Reardon (1999) also reported mullite (Al2O3SiO2)

peak for fly ash in their study. Addition of lime to

Table 4 Model parameters

for treated and untreated

soils contaminated with

calcium sulfate

Treatment Soil property (Y) Figure Model parameters

Yo A B R2

Untreated LL 3a 40 0.04 0.05 0.94

PI 3b 19 0.04 0.06 0.92

OMC (%) 5a 17 0.21 0.20 0.95

cdmax. (g/cm3) 5b 1.52 -6.45 -9.18 0.92

Free swell (%) 9 7.0 0.094 0.065 0.95

Compressive strength (kPa) 11 553.6 -0.0019 -0.002 0.88

6 % Lime LL 3a 37 0.285 0.0 0.97

PI 3b 15.7 0.085 0.069 0.95

OMC (%) 5a 18.4 0.84 0.0 0.98

cdmax. (g/cm3) 5b 1.64 -14.48 -7.26 0.82

Free swell (%) 9 5.54 0.223 0.240 0.89

Compressive strength (kPa) 11 716 -0.003 -0.002 0.91

10 % Fly ash (C) LL 3a 31.5 0.282 0.0 0.92

PI 3b 12 0.079 0.044 0.94

OMC (%) 5a 20 0.802 0.0 0.95

cdmax. (g/cm3) 5b 1.59 -10.64 -9.278 0.94

Free swell (%) 9 5.85 0.181 0.172 0.91

Compressive strength (kPa) 11 807.4 -0.002 -0.002 0.96

Fig. 2 Modeling the linear and non linear responses of treated

sulfate soils

Geotech Geol Eng

123

sulfate contaminated soil resulted in the formation of

ettringite which is hydrous calcium aluminum sulfate

mineral [Ca6Al2(SO4)3(OH)12�26H2O)] (2h peaks at

9.01�, 22.06�, 27.62�, 35.04�, 37.45� and 75.48�),calcium silicate hydrate (CaH2O4Si) (2h peaks at

39.7�, 42.7� and 46.0�) and aluminum silicate hydrate

(AlSiO3H2O) (2h peak at 60.0�).

5.1.1 Liquid Limit (LL)

Additional of calcium sulfate to the natural CL soil

increased the liquid limit and the change was nonlinear

as shown in Fig. 3a. When the calcium sulfate content

in the soil was 4 %, the liquid limit increased from 40

to 57 % and can be attributed to the changes in the clay

mineralogy. The change in the LLwith calcium sulfate

concentration was represented using hyperbolic rela-

tionship Eq. (2) and the parameters A, B and Yo are

summarized in Table 4, and the coefficient of deter-

mination (R2) for the relationship was 0.94. Total of 19

data were collected from the literature and the liquid

limit varied from 31 to 73 % with a mean and standard

deviation of 52.3 % and 13.2 respectively. The

collected data from the literature are compared to the

model prediction and 47 % of these data located above

the model prediction as shown in Fig. 3. Addition of

6 % of lime and 10 % of fly ash to the sulfate soil with

4 % of calcium sulfate decreased the liquid limit by 12

and 22 % respectively. Linear trends were observed

between the LL and calcium sulfate concentration of

sulfate soils modified using 6 % of lime and 10 % of

fly ash (by dry weight) as shown in Fig. 4 (a). Fly ash

treatment showed greater reduction in LL than lime

treatment. The model parameters for lime and fly ash

treatment are summarized in Table 4 with parameter B

being zero and the coefficient of determination (R2)

were 0.97and 0.92 respectively.

5.1.2 Plasticity Index (PI)

Plasticity index of natural CL soil increased from 19 to

34 % by increasing calcium sulfate content to 4 %

because of the formation of calcium silicate sulfate

and aluminum silicate sulfate in the soil. Total of 17

data were collected from various research studies in

the literature and the plasticity index varied from 14 to

48 % with a mean and standard deviation of 22.2 and

11.3 respectively. About 65 % of the research data

located below the model prediction as shown in

Fig. 3b. Plasticity index of the natural soil contaminat-

ed with 4 % of calcium sulfate decreased by 25 and

16 % when the sulfate soil modified using 6 % of lime

and 10 % of fly ash (by dry weight) respectively as

shown in Fig. 4b. In this study, total of 21 soil samples

were tested. Hyperbolic relationship was used to relate

plasticity index to calcium sulfate concentration for

treated and untreated sulfate soil as shown in Fig. 4b.

The parameters A, B and Yo for untreated sulfate soil

and treated using 6 % of lime and 10 % of fly ash are

summarized in Table 4. The coefficients of determi-

nation (R2) for the hyperbolic relationships for

untreated and lime and fly ash treated sulfate soils

were 0.92, 0.95 and 0.94 respectively.

5.1.3 Compacted Soil

Optimum moisture content for the field CL soil

increased from 17 to 21.8 % when the calcium sulfate

concentration increased from 0 to 4 % as shown in

Fig. 5a. Total of 17 data on optimum moisture content

25

35

45

55

65

75

0 1 2 3 4 5

LL (%

)

Calcium Sulfate Concentra�on (%)

No. of Data=19

Sivapullaiah (2002) Puppala (2004)Harris (2004) Chakkrit (2008)Cerato (2011) Current StudyModel

10

15

20

25

30

35

40

45

50

0 1 2 3 4 5

PI (%

)

Calcium Sulfate Concentra�on (%)

No.of Data=17

Puppala (2004) Arvind (2007)Harris (2008) Chakkrit (2008)Cerato (2011) Current Study

(a) (b)

Fig. 3 Variations of index properties with calcium sulfate content. a Liquid limit (LL). b Plasticity Index (PI)

Geotech Geol Eng

123

(OMC) for sulfate contaminated soils from various

research studies were obtained and the mean and

standard deviation were 20 and 4.5 %respectively.

About 33 % of the literature data were located below

the model predication as shown in Fig. 5a. Additional

of 6 % lime and 10 % of fly ash (by dry weight) to the

sulfate soil with 4 % of calcium sulfate in the current

study increased the (OMC) by 6 and 12 % respectively

as shown in Fig. 6a. Linear trends were observed

between the (OMC %) versus calcium sulfate con-

centration for soils treated with 6 % of lime and 10 %

of fly ash (by dry weight) as shown in Fig. 5a and the

parameter B was zero. The model parameters A, B, Yo

and coefficient of determination (R2[ 0.9) for un-

treated sulfate soil and treated with 6 % of lime and

10 % of fly ash are summarized in Table 4.

As shown in Fig. 5b, when the calcium sulfate

concentration increased from 0 to 4 % the dry density of

the CL soil decreased by 7 %. Total of 16 data on

maximum dry density of calcium sulfate contaminated

soils were collected from the literature and themean and

standard deviation were 1.66 g/cm3 and 0.11 respec-

tively.All the data from the literaturewere located above

the results from the current study as shown in Fig. 5b. In

the current study, the maximum dry density of sulfate

soil with 4 % calcium sulfate concentration increased 7

and 6 % by using 6 % lime and 10 % of fly ash

respectively as shown in Fig. 6b. The model parameters

A, B, Yo and coefficient of determination for untreated

sulfate soil and treated with 6 % of lime and 10 % of fly

ash are summarized in Table 4.

5.1.4 Free Swelling (FS)

Total of 15 soil samples were tested in this study.

Percent of free swelling with time for the CL soil with

10

15

20

25

30

35

40

0 1 2 3 4 5

PI (%

)

Calcium Sulfate Concentra�on (%)

Field Soil 6% lime 10% Fly ash Model

20

25

30

35

40

45

50

55

60

65

0 1 2 3 4 5

LL (%

)

Calcium Sulfate Concentra�on (%)

Field Soil6% lime 10% Fly ash Model

(a) (b)

Fig. 4 Variations of Index properties of untreated and treated calcium sulfate soil with 6 % of lime and 10 % of fly ash. a Liquid limit

(LL). b Plasticity Index (PI)

1.2

1.3

1.4

1.5

1.6

1.7

1.8

1.9

2

Max

imum

Dry

Dens

ity(g

m/c

m3 )

No.of Data=16

Gerald (2000) Harris (2008)Michael (2011) Cerato (2011)McCarthy (2012) Current Study

12

14

16

18

20

22

24

26

28

0 1 2 3 4 50 1 2 3 4 5

Op�

mum

Moi

stur

eCo

nten

t(%

)

Calcium Sulfate Concentra�on (%) Calcium Sulfate Concentra�on (%)

No.of Data=17

Harris (2008) Michael (2011)Cerato (2011) McCarthy (2012)McCarthy (2012) Current StudyModel

(a) (b)

Fig. 5 Variations of compacted soil properties with calcium sulfate content. a Optimum moisture content (OMC). b Maximum dry

density

Geotech Geol Eng

123

different percentage of calcium sulfate are shown in

Fig. 7. Free swelling percent of natural CL soil

increased by 160 % when the calcium sulfate changed

from 0 to 4 % as shown in Fig. 7 because of changes in

the clay mineralogy. Free swelling versus calcium

sulfate concentrations were compared with the several

data from literature as shown in Fig. 8. About 50 % of

the literature data were located above the model

predication as shown in Fig. 8. Additional of 6 % of

lime and 10 % of fly ash decreased the free swelling at

4 % of calcium sulfate concentration by 53 and 46 %

respectively as shown in Fig. 9. Hyperbolic relation-

ship between the percentage of free swelling and

calcium sulfate concentration for untreated and treated

sulfate soils were observed as shown in Fig. 9. The

model parameters A, B and Yo are summarized in

Table 4 and the coefficient of determination (R2)

varied from 0.89 to 0.95.

5.1.5 Stress–Strain Relationship

Stress–strain relationships for untreated and treated

sulfate soils are shown in Fig. 10. With the increase in

sulfate content the unconfined compressive strength

and the initial modules of the treated and untreated

compacted soils decreased as shown in Fig. 10.

Total of 15 unconfined compression tests were

performed in this study. Compressive strength of the

CL soil used in the current study decreased by 65 %

when the calcium sulfate content was increased from 0

to 4 % as shown in Fig. 11. The variation of strength

with calcium sulfate content was represented using the

proposed model and the parameters are summarized in

Table 4. Additional of 6 % of lime and 10 % of fly ash

(by dry weight) to the sulfate soil with 4 % calcium

sulfate increased the 7 days cured unconfined com-

pressive strength by 130 and 120 % respectively as

shown in Fig. 11.

5.1.6 Stress–Strain Behavior Modeling

Soils are generally modeled as linear elastic, linear

elastic—perfectly plastic or as strain hardening mate-

rials. In this study the soil, with and without treatment,

exhibitored strain softening behavior as shown in

Fig. 10.

5.2 p–q Model

Based on experimental results, model proposed by

Mebarkia and Vipulanandan (1992), was used to

predict the stress–strain behavior of treated sulfate

contaminated CL soil with different percentage of

polymer solution (Eq. 3). The model is defined as

follows:

r ¼eec

qþ ð1� p� qÞ eec

� �þ p � e

ec

� �pþqp

0BB@

1CCA� rc ð3Þ

where, r = compressive strength, rc; ec = compres-

sive strength and corresponding strain, p,q = material

parameters.

16

18

20

22

24

26

0 1 2 3 4 5

Op�

mum

Moi

stur

e Co

nten

t (%

)

Calcium Sulfate Concentra�on (%)

Field Soil6% Lime10% fly ash Model

1.35

1.4

1.45

1.5

1.55

1.6

1.65

0 1 2 3 4 5

Max

imum

Dry

Den

sity

(gm

/cm

3 )

Calcium Sulfate Concentra�on (%)

Untreated Soil6% Lime10% Fly ash Model

(a) (b)

Fig. 6 Variations of compaction characteristics of treated soils with different calcium sulfate content. a Optimum moisture content

(OMC). b Maximum dry density

Geotech Geol Eng

123

Parameter q was defined as the ratio of secant

modulus at peak stress to initial tangent modulus.

Parameter p was obtained by minimizing the error in

the predicated stress–strain relationship. Hence, pa-

rameters p and q in (Eq. 3) were determined based on

the stress–strain behavior of sulfate soil treated with

6 % of lime and 10 % of fly ash and the values and

coefficient of determination (R2) are summarized in

Table 5. In the Fig. 10, the predicted values of

compressive strength for sulfate contaminated CL soil

0

2

4

6

8

10

12

14

16

18

20

0 2 4 6 8 10 12

Free

Sw

ellin

g (%

)

Time (Days)

Untreated Soil

0% 1% 2% 4%Calcium Sulfate Concentra�on

0

1

2

3

4

5

6

7

8

9

10

0 2 4 6 8

Free

Sw

ellin

g ( %

)

Time (Days)

6% Lime

Calcium Sulfate Concentra�on

(a) (b)

(c)

0

2

4

6

8

10

12

42 30 1 5 6 7 8

Free

Sw

ellin

g (%

)

Time(Days)

10% Fly Ash

Calcium Sulfate Concentra�on

0% 1% 2% 4%

0% 1% 2% 4%

Fig. 7 Free swell–time relationships. aUntreated calcium sulfate soil. bModified soil using 6 % of lime. cModified soil using 10 % of

fly ash

0

5

10

15

20

25

30

35

40

45

0 1 2 3 4 5 6

Free

Sw

ellin

g (%

)

Calcium Sulfate Concentra�on (%)

Li�le (2010)Cerato et al. (2011)McCarthy (2012)Current StudyModel

Fig. 8 Variation of free swell with calcium sulfate content

4

6

8

10

12

14

16

18

20

0 1 2 3 4 5

Free

Sw

ellin

g ( %

)

Calcium Sulfate concentration (%)

Free Swelling % at 7days

Untreated Soil6% Lime10% fly ashModel

Fig. 9 Variation of free swell of sulfate soil after 7 days of

curing with lime and fly ash stabilizer

Geotech Geol Eng

123

treated with 6 % lime are compared to the 10 % fly ash

treated soil. The lime treated soils were stronger than

fly ash treated soils. As summarized in Table 5 the

parameters p and q influence by sulfate content and the

method of treatment.

5.2.1 Parameter q

Parameter q represents the nonlinear behavior of the

material up to peak stress. For calcium sulfate

contaminated CL soil the parameter q was in the

range of 0.56–1.0 (Table 5). Treating the CL soil with

6 % lime and 10 % fly ash decreased the q parameter

to be in the range of 0.53–0.6 and 0.53–0.71 respec-

tively (Table 5).

5.2.2 Parameter p

For untreated soils the parameter p varied from 0.02 to

0.51. Soil treated with 6 % lime and 10 % fly ash the

parameter p varied from 0.2 to 0.65 and 0.07 to 0.22

respectively.

5.3 Failure Strain

The compressive strain at peak stress was affected by

the level of sulfate contamination and also the type of

0

100

200

300

400

500

600

700

0 0.5 1 1.5 2 2.5 3 3.5 4

Stre

ss (k

Pa)

Axial Strain (%)

Calcium Sulfate=0%

Calcium Sulfate=1%

Calcium Sulfate=4%

Model

0

100

200

300

400

500

600

700

800

900

0 0.5 1 1.5 2 2.5 3

Stre

ss (k

Pa)

Axial Strain (%)

Calcium Sulfate=0%

Calcium Sulfate=1%

Calcium Sulfate=4%

Model

(a)

(b)

0

100

200

300

400

500

600

700

800

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5

Stre

ss (k

Pa)

Axial Strain (%)

Calcium Sulfate=0%

Calcium Sulfate=1%

Calcium Sulfate=4%

Model(c) 10% Fly Ash Treatment

Untreated Soil

6% Lime Treatment

Fig. 10 Compressive stress–strain relationship for a untreated sulfate soil, b treated sulfate soil with 6 % lime, c treated sulfate soil

with 10 % fly ash

0

100

200

300

400

500

600

700

800

900

0 1 2 3 4 5

Unc

onfin

ed C

ompr

essi

ve S

tren

gth

(kPa

)

Calcium Sulfate Concentra�on %

Untreated Soil6% Lime10% fly ash Model

Fig. 11 Variation of compressive strength with sulfate content

for untreated and treated sulfate contaminated CL soil

Geotech Geol Eng

123

treatment. Compared to the 10 % fly ash treatment

where failure strain was over 2.5 %, 6 % lime

treatment reduced the strain at failure stress to about

1.5 %.

5.4 Nonlinear Model (NLM) Parameters

The model parameters rc, ec, p and q were influenced

by the composition of the soil. It is being proposed to

relate the model parameters to the independent

variables [sulfate concentration (S%), lime (L%) and

fly ash (FA%)] using a nonlinear power relationship as

proposed by Demircan et al. (2011).

The effect of sulfate content, lime and fly ash were

separated as follows:

Model parameters ðrc; ec; p; qÞ¼ A0 þ a � ðSÞb þ c � ðLÞd þ e � ðFAÞf ð4Þ

where: Ao Initial model parameter for uncontaminated

and untreated soil (field soil). a, b, c, d, e, f, g and h are

the nonlinear model parameters.

The NLM parameters were obtained from multiple

regression analyses using the least square method. The

NLM model parameters are summarized in Table 6.

6 Conclusions

In this study the effect of sulfate content on a CL soil

was investigated. XRD analyses showed that the

composition of clay was calcium silicate (CaSiO3),

aluminum silicate (Al2SiO5), magnesium silicate

(MgSiO3) and quartz (SiO2). Addition of calcium

sulfate to the soil resulted in the formation of calcium

silicate sulfate (Ternesite) and aluminum silicate

sulfate. Over 140 tests were performed during this

study. Also data from the literature was compared to

the data from the current study. Based on the

experimental and analytical study on a sulfate con-

taminated CL soil with and without treatment follow-

ing conclusions are advanced.

1. Liquid limit of natural CL soil increased from 40

to 57 % with the addition of 4 % calcium sulfate

because of the modification of clay mineralogy.

Adding 6 % lime and 10 % fly ash to the 4 %

sulfate soil decreased the LL by 12 and 22 %

respectively.

2. Plasticity index of the CL soil increased by 79

with 4 % calcium sulfate content. The plasticity

index for 4 % sulfate contaminated soil was

Table 5 Stress–strain model parameters for sulfate soil treated using lime and fly ash

Calcium sulfate, S (%) Lime, L (%) Fly ash, FA (%) p–q Model parameter

rc (kPa) ec (%) p q R2

0 – – 573 2.77 0.22 0.69 0.99

1 – – 285 1.88 0.51 0.56 0.99

4 – – 196 3.72 0.02 1.0 0.99

0 6 – 810 1.33 0.22 0.53 0.97

1 6 – 533 1.49 0.47 0.60 0.98

4 6 – 450 1.94 0.65 0.58 0.98

0 – 10 715 2.69 0.22 0.53 0.99

1 – 10 494 3.10 0.41 0.64 0.99

4 – 10 433 3.81 0.07 0.71 0.97

Table 6 Nonlinear model parameters for sulfate contaminated CL soil treated with 6 % lime and 10 % fly ash

Model parameters Ao a b c d e f R2

rc (kPa) 573 -2.81 0.26 240 0.16 8.9 1.2 0.96

ec (%) 2.77 0.65 0.21 -0.37 0.01 7E-5 4.2 0.94

p 0.22 0.08 0.22 0.02 1.13 0.5 2.3 0.85

q 0.69 0.10 0.62 2E-4 1.13 3E-4 1.0 0.83

Geotech Geol Eng

123

reduced by 25 and 16 % when treated with 6 %

lime and 10 % fly ash respectively.

3. Addition of lime and fly ash increased the

maximum dry density and optimum moisture

content of the compacted sulfate soils.

4. Free swelling increased from 7 to 19 % with the

increase of calcium sulfate content from 0 to 4 %

because of the modification of the clay mineralogy.

Additional of 6 % lime and 10 % fly ash decreased

the free swelling percent to 9 and 10 % respectively

after 7 days of swelling measurement.

5. Compressive stress–strain relationship was af-

fected by sulfate content in the soil. Unconfined

compressive strength of the CL decreased with

increased sulfate content. Addition of 4 % calci-

um sulfate to the soil decreased the strength by

65 %. The unconfined compressive strength of

sulfate soils increased with 6 % lime and 10 % fly

ash.

6. The improvement in the soil strength due to lime

and fly ash treatments can be attributed to the

formation of calcium silicate hydrate as cement-

ing compound.

7. The hyperbolic model was effective in predicting

the changes in the sulfate contaminated CL soil

with and without treatment.

8. Lime treatment (6 %) showed much higher

enhancement in short-term strengths of sulfate

contaminated soils compared to 10 % fly ash

treatment. Also lime treatment reduced the com-

pressive strain at failure stress to 1.5 % compared

to fly ash treated soil of 2.5 %.

9. Nonlinear compressive stress–strain model was

used to predict the behavior of lime and fly ash

treated sulfate contaminated soils. The model

parameters were sensitive to the type of

treatment.

Acknowledgments This study was supported by the Center

for Innovative Grouting Materials and Technology (CIGMAT)

with funding from various industries and Underground

Construction Technology Association (UCTA). Sponsors are

not responsible for any of the findings.

References

Arabani M, Veis M (2007) Geomechanical properties of lime

stabilized clayey sands. Arab J Sci Eng 32(1B):11–25

Aravind P, Chittoori S, Puppala AJ (2011) Influence of miner-

alogy and plasticity index on the stabilization effectiveness

of expansive clays. Transp Res Rec 91–99

Ata A, Vipulanandan C (1998) Cohesive and adhesive proper-

ties of silicate grout on grouted—sand behavior. J Geotech

Geoenviron Eng 124(1):38–44

Acosta HA, Edil TB, Benson, CH (2003) Soil stabilization and

drying using fly ash. Geo Engineering Report No. 03-03,

Department of Civil and Environmental Engineering,

University of Wisconsin-Madison

Bell G (1996) Lime stabilization of clay minerals and soils. Eng

Geol 42:223–237

Burkart B, Goss CG, Kern PJ (1999) The role of gypsum in

production of sulfate-induced deformation of lime-stabi-

lized soils. Environ Eng Geosci 5(2):173–187

Cokca E (2001) Use of class C fly ashes for the stabilization of

an expansive soil. J Geotech Geoenviron Eng 127(7):568–

573

Demircan E, Harendra S, Vipulanandan C (2011) Artificial

neural network and nonlinear models for gelling time and

maximum curing temperature rise in polymer grouts.

J Mater Civil Eng 23(4):372–377

Duchesne J, Reardon EJ (1999) Lime treatment of fly ash:

characterization of leachate composition and solid/water

reactions. Waste Manag 19(3):221–231

Edil T, Bin-Shafique M, Tanyu B, Kim W-H, Senol A (2002)

Field evaluation of construction alternatives for roadway

over soft subgrade. Transportation Research Board,

Washington, DC, pp 36–48

Harris P, Tom S, Stephen S (2004) Hydrated lime stabilization

of sulfate-bearing soils in Texas. Texas Department of

Transportation, FHWA/TX-04/0-4240-2, pp 1–36

Hunter D (1988) Lime-induced heave in sulfate-bearing clay

soils. J Geotech Eng 114(2):150–167

Jenkins R, Snyder RL (1996) Introduction to X-ray powder

diffractometry. Wiley, New York

Karim R, Zain M, Jamil M, Lai C, Islam N (2011) Strength

development of mortar and concrete containing fly ash: a

review. Int J Phys Sci 6(17):4137–4153

Kota P, Hazlett D, Perri L (1996) Sulfate-bearing soils: prob-

lems with calcium based stabilizers. Transportation Re-

search Record 1546, Transportation Research Board,

Washington, DC, pp 62–69

Kumar A,Walia B, Bajaj A (2007) Influence of fly ash, lime, and

polyester fibers on compaction and strength properties of

expansive soil. J Mater Civil Eng ASCE 19(3):242–248

Little D, Syam N, Herbert B (2010) Addressing sulfate-induced

heave in lime treated soils. J Geotech Geoenviron Eng

ASCE 110–118

McCallister LD, Tidwell L (1997) Double lime treatment to

minimize sulfate-lime induced heave in expansive clays.

US Army Engineers, Waterways Experiment Station,

Vicksburg

McCarthy M, Csetenyi L, Sachdev A, Jones M (2009) Role of

fly ash in the mitigation of swelling in lime stabilized

sulfate—bearing soils. World of coal ash (WOCA) con-

ference, pp 1–18

Mebarkia S, Vipulanandan C (1992) Compressive behavior of

glass-fiber-reinforced polymer concrete. J Mater Civil Eng

4(1):91–105

Geotech Geol Eng

123

Mitchell JK (1986) Practical problems from surprising soil be-

havior. J Geotech Eng ASCE 112:259–289

Mitchell K, Dermatas D (1992) Clay soil heave caused by lime-

sulfate reactions. Innovations in uses for lime. ASTM STP

1135, American Society for Testing and Materials

(ASTM), Philadelphia, PA, pp 41–64

Mohammed A, Vipulanandan C (2014) Compressive and tensile

behavior of polymer treated sulfate contaminated CL soil.

J Geotech Geol Eng 32(1):71–83

Nalbantoglu Z (2004) Effectiveness of class C fly ash as an ex-

pansive soil stabilizer. Construct Build Mater 18:377–381

Perrin L (1992) Expansion of lime-treated clays containing

sulphates. In: Proceedings of the seventh international

conference on expansive soils, ASCE Expansive Soils

Research Council, New York, 1, pp 409–414

Petry M, Little D (1992) Update on sulfate-induced heaven

treated clays; problematic sulfate levels. Transportation

Research Record 1362, National Research Council,

Washington, DC, pp 51–55

Phani BR, Sharma RS (2004) Effect of fly ash on engineering

properties of expansive soils. J Geotech Geoenviron Eng

130(7):764–767

Puppala AJ, Kadam R, Madhyannapu R, Hoyos LR (2006)

Small-strain shear moduli of chemically stabilized sulfate-

bearing cohesive soils. J Geotech Geoenviron Eng 132(3):

322–336

Puppala AJ, Viyanant C, Kruzic. AP, Perrin L (2002) Evaluation

of a modified sulfate determination method for cohesive

soils. Geotech Test J 25(1):85–94

Rajasekaran G, Essaku S, Mathews PK (1994) Physico-chemi-

cal and mineralogical studies on Cochin marine clays.

Ocean Eng 21(8):771–780

Rajendran D, Lytton R (1997) Reduction of sulfate swell in

expansive clay subgrades in the Dallas district. Texas

Transportation Institute, Rep. No. TX-98/3929-1, Bryan,

TX, pp 1–159

Ramesh HN (2009) Effect of sulphate content in soils as a po-

tential geohazard. IGC 2009, Guntur, pp 976–979

Rollings R, Burkes J, Rollings M (1999) Sulfate attack on ce-

ment-stabilized sand. Geotech Geoenviron Eng 125(5):364–

372

Sivapullaiah P, Sridharan A, Ramesh HN (2000) Strength be-

havior of lime treated soils in the presence of sulphate. Can

Geotech J 37:1–10

Sridhran A, Sivapullaiah PV, Ramesh HN (1995) Consolidation

behavior of lime treated sulphatic soils. In: Proceedings

international symposium compression consolidation

clayey soils, Hiroshima, Japan, I, pp 183–188

Usluogullari O, Vipulanandan C (2011) Stress–strain behavior

and California bearing ratio of artificially cemented sand.

Test Eval ASTM 39(4):637–645

Geotech Geol Eng

123