Stabilization of Organic Soils with Fly Ash · Stabilization of Organic Soils with Fly Ash ... Soil organic content is a detrimental ... organic matter instead of reacting with pozzolanic

Stabilization of Organic Soils with Fly AshErdem O. Tastan1; Tuncer B. Edil, F.ASCE2; Craig H. Benson, F.ASCE3; and Ahmet H. Aydilek, M.ASCE4

Abstract: The effectiveness of fly ash use in the stabilization of organic soils and the factors that are likely to affect the degree of stabilizationwere studied. Unconfined compression and resilient modulus tests were conducted on organic soil–fly ash mixtures and untreated soil spec-imens. The unconfined compressive strength of organic soils can be increased using fly ash, but the amount of increase depends on the type ofsoil and characteristics of the fly ash. Resilient moduli of the slightly organic and organic soils can also be significantly improved. Theincreases in strength and stiffness are attributed primarily to cementing caused by pozzolanic reactions, although the reduction in watercontent resulting from the addition of dry fly ash solid also contributes to strength gain. The pozzolonic effect appears to diminish asthe water content decreases. The significant characteristics of fly ash that affect the increase in unconfined compressive strength and resilientmodulus include CaO content and CaO=SiO2 ratio [or CaO=ðSiO2 þ Al2O3Þ ratio]. Soil organic content is a detrimental characteristic forstabilization. Increase in organic content of soil indicates that strength of the soil–fly ash mixture decreases exponentially. For most of thesoil–fly ash mixtures tested, unconfined compressive strength and resilient modulus increased when fly ash percentage was increased. DOI:10.1061/(ASCE)GT.1943-5606.0000502. © 2011 American Society of Civil Engineers.

CE Database subject headings: Fly ash; Soil stabilization; Stiffness; Organic matter.

Author keywords: Organic soil; Fly ash; Stabilization; Strength; Stiffness; Stabilization.

Introduction

Construction of roadways on soft organic soils can be problematicbecause organic soils typically have low shear strength and highcompressibility (Edil 1997). Current practice for construction ofroadways over organic soil subgrades mostly involve the removalof the organic soil to a sufficient depth and replacement withcrushed rock (referred to as “cut and replace”) or preloading toimprove engineering properties. Chemical stabilization withbinders such as cement, lime, and fly ash can be undertaken rapidlyand often at low cost, and therefore chemical stabilization isbecoming an important alternative (Keshawarz and Dutta 1993;Sridharan et al. 1997; Kaniraj and Havanagi 1999; Parsons andKneebone 2005).

Chemical stabilization of soft soils involves blending a binderinto the soil to increase its strength and stiffness through chemicalreactions. The binder is intended to cement the soil solids, therebyincreasing strength and stiffness. The binders are generally addedas dry solids. In practice, reducing the water content of high-water-content soils to the optimum water content (OWC) is difficult andtime-consuming. Therefore, addition of dry solids and cementitious

materials is preferable. Thus, addition of a binder reduces both thewater content and binds the soil particles, which results in anincrease in strength and stiffness. Common binders include cement,lime, fly ash, or mixtures thereof. The use of fly ash as a binder isattractive because fly ash is an industrial by-product that isrelatively inexpensive, compared with cement and lime (FederalHighway Administration 2003). Additionally, using fly ash for soilstabilization, particularly fly ashes that otherwise would be land-filled, promotes sustainable construction through reduction ofenergy use and reduction of greenhouse gases.

Fly ash has been shown to effectively stabilize soft inorganicsoils (Ferguson 1993; Acosta et al. 2003; Prabakar et al. 2004;Bin-Shafique et al. 2004; Trzebiatowski et al. 2005), but little isknown regarding the effectiveness of stabilizing soft organic soilswith fly ash. Organic soils are known to be more difficult to sta-bilize chemically than inorganic soils (Hampton and Edil 1998;Janz and Johansson 2002). The objectives of this study were(1) to determine if fly ashes can stabilize organic soils, and, ifso, (2) to quantify the improvement in the unconfined compressivestrength (UCS, qu) and resilient modulus of the organic soil asadmixed with fly ash, and (3) to investigate potentially importantfactors affecting the stabilization process, such as fly ash andsoil characteristics, fly ash percentage in the mixture, and watercontent.

Background

Chemical Stabilization

When binders such as lime, cement, and fly ash are blended withsoil in the presence of water, a set of reactions occur that result indissociation of lime (CaO) in the binders and the formation of ce-mentitious and pozzolanic gels [calcium silicate hydrate gel (CSH)and calcium aluminate silicate hydrate gel (CASH)]:

CaOþ H2O ⇒ CaðOHÞ2 ð1Þ

1Assistant Project Engineer, Paul C. Rizzo Associates, Inc., Monroe-ville, PA 15146.

2Professor, Geological Engineering Program, Dept. of Civil and Envir-onmental Engineering, Univ. of Wisconsin, Madison, WI 53706. E-mail:[email protected]

3Wisconsin Distinguished Professor and Chairman, Dept. of Civil andEnvironmental Engineering, Univ. of Wisconsin, Madison, WI 53706.E-mail: [email protected]

4Associate Professor, Dept. of Civil and Environmental Engineering,Univ. of Maryland, 1163 Glenn Martin Hall, College Park, MD 20742(corresponding author). E-mail: [email protected]

Note. This manuscript was submitted on September 17, 2010; approvedon January 6, 2011; published online on January 8, 2011. Discussion periodopen until February 1, 2012; separate discussions must be submitted forindividual papers. This paper is part of the Journal of Geotechnicaland Geoenvironmental Engineering, Vol. 137, No. 9, September 1,2011. ©ASCE, ISSN 1090-0241/2011/9-819–833/$25.00.

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / SEPTEMBER 2011 / 819

Downloaded 15 Jan 2012 to 129.2.71.160. Redistribution subject to ASCE license or copyright. Visit http://www.ascelibrary.org

http://dx.doi.org/10.1061/(ASCE)GT.1943-5606.0000502




CaðOHÞ2 ⇒ Ca2þ þ 2½OH�� ð2Þ

Ca2þ þ 2½OH� þ SiO2 ⇒ CSH ð3Þ

Ca2þ þ 2½OH�� þ Al2O3 ⇒ CASH ð4ÞThese reactions are referred to as cementitious and/or pozzo-

lanic reactions that result in the formation of cementitious gels.The increase in strength was found to be roughly related to the typeand quantity of possible reaction products (i.e., cement reactionproduct, CSH for short-term strength and pozzolanic reaction prod-uct, CASH for long-term strength gain).

The source for the pozzolans (a siliceous or aluminous material)is either the soil or the binding agent. These reactions contribute tostabilization of soils in two ways. First, plasticity of the soil is re-duced by the exchange of calcium ions in the pore water withmonovalent cations on clay surfaces and by compression of the ad-sorbed layer because of the elevated ionic strength of the pore water(Rogers and Glendinning 2000). Second, the CSH or CASH gelsformed by cementitious and pozzolanic reactions bind the solid par-ticles together, and this binding produces a stronger soil matrix(Arman and Munfakh 1972). For organic soils, reactions areexpected to be inhibited or delayed by the existence of organiccompounds (Hampton and Edil 1998; Tremblay et al. 2002).Mechanisms of organic matter interference with strength gain inchemical stabilization are not fully understood, but the followingmechanisms are suggested (Hampton and Edil 1998; Axelsson et al.2002; Janz and Johansson 2002): (1) organic matter can alter thecomposition and structure of CSH gel, a cementing compound thatforms bonds between particles and also the type and amount ofother hydration products, e.g., ettringite; (2) organic materials oftencontain materials such as humus or humic acid, which retardstrengthening reactions; (3) organic matter holds 10 or more timesits dry weight in water and may limit water available for hydration;and (4) organic matter forms complexes with aluminosilicates andwith metal ions, and such complexes interfere with hydration.

Some fly ashes contain lime and pozzolans, such as Al2O3 andSiO2, and therefore are self-cementing. The effectiveness of a givenfly ash is expected to depend on the relative abundance of CaO andoxides providing pozzolans. For example, Class C fly ashes (i.e.,fly ashes meeting the requirements in ASTM C618 (ASTM 2008)for use in ready-mix concrete) have a CaO content > 20% (byweight) and a Al2O3 þ Fe2O3 þ SiO2 content of 50–70%. In con-trast, Class F fly ashes have < 10% CaO. Consequently, Class Cashes generally are more effective at forming CSH and CASH gelsthan Class F ashes (Sridharan et al. 1997).

Janz and Johansson (2002) indicate that the CaO=SiO2 ratio,which stands for relative abundance of CaO and SiO2, is an indi-cator of the potential for pozzolanic reactions and that binderswith larger CaO=SiO2 ratios are likely to be more effective stabi-lizers. For example, C3S clinker, which is a strong binder, hasa CaO=SiO2 ratio = 3. Similarly, the ratio of CaO=ðSiO2 þAl2O3Þ can also be used as an indicator of the potential to formCSH and CASH gels (Odadjima et al. 1995). However, binderswith a high CaO=SiO2 or CaO=ðSiO2 þ Al2O3Þ ratio can stillbe ineffective if pozzolanic reactions are limited by the availabilityof CaO pozzolans (e.g., too little CaO, SiO2, and/or Al2O3)

Inhibition of Cementing Reactions by Organic Matter

Fly ash specifications for concrete applications usually include anupper bound on the organic carbon content of the fly ash. Thisupper bound is normally characterized by the loss on ignition(LOI) measured with ASTM C311. Clare and Sherwood (1954)indicated that the organic matter in organic soils adsorbs Ca2þ ions.

When cement, lime, or fly ash (any source of Ca2þ ions) is added toorganic soils, following the hydration of lime [Eqs. (1) and (2)],released Ca2þ ions are likely to be exhausted by the organic matter,which limits the availability of Ca2þ ions for pozzolanic reactions.Thus, the amount of CaO in fly ash should be large enough to com-pensate for the consumption of Ca2þ ions by the organic matter inthe soil. The possible interactions of organic compounds with poz-zolanic minerals (Ca2þ or Alþ3) or CaðOHÞ2 are summarized asfollows (Young 1972): (1) calcium ions can be adsorbed by theorganic matter instead of reacting with pozzolanic minerals; (2) or-ganic compounds react with CaðOHÞ2 and precipitate, which formsinsoluble compounds and limits the availability of Ca2þ ions forpozzolanic reactions; (3) alumina can form stable complexes withorganic compounds, and calcium ions can also complex with or-ganic compounds, but Young (1972) stated that complexes formedby Ca2þ ions were not stable and would not affect the calcium ionequilibria; and (4) organic compounds can adsorb on CaðOHÞ2nuclei, which inhibit the growth of nuclei and formation ofCSH. Hampton and Edil (1998) indicated that the organic matterin soils can also retain large amounts of water, which can reduce theamount of available water for hydration reactions when a cementi-tious additive is blended with soil.

Similarly, organic matter in soil is known to affect stabilizationusing cements or fly ashes. For example, Tremblay et al. (2002)evaluated how cement stabilization of an inorganic soil [a clay withplasticity index (PI) = 26] was inhibited by organic content by add-ing organic compounds to the soil, such as acetic acid, humic acid,tannic acid, ethylenediaminetetraacetic acid (EDTA), and sucrose.Tremblay et al. (2002) also suggested that pozzolanic reactions arelikely to be inhibited if the pH of the soil-cement mixture is lessthan 9.

Materials and Methods

Soils

Three soft organic soils with different organic contents were used inthe study: Markey (silty, sandy peat), Lawson (low plasticity or-ganic sandy clay), and Theresa (moderately plastic organic clay).All soils were collected within 1 m of the ground surface and aretypical of organic soils encountered as a subgrade during roadwayconstruction in Wisconsin. Index properties of the soils (and com-paction parameters) are summarized in Table 1. All three soils hadbell-shaped compaction curves, but the maximum dry unit weightof these soils is less than the typical for soils from Wisconsinwith similar plasticity (Edil et al. 2006). An inorganic silt fromBoardman, Oregon (Boardman silt) was also used in the testingprogram. Index properties of the silt are summarized in Table 1.This silt, which has similar particle-size distribution as the fly ashesin the study, was used as a nonreactive binder in some of the mix-tures to separate the effects of cementing and reduction in watercontent by adding dry solid.

Fly Ashes

Six fly ashes and Type I portland cement were used as binders in thestudy. The fly ashes were obtained from electric power plants in theupperMidwesternUnited States andwere selected to provide a broadrange of carbon content (0.5–49%), CaO content (3.2–25.8%),and CaO=SiO2 ratio (0.09–1.15). General properties of the fly ashesare summarized in Table 2.

The Stanton and Columbia fly ashes classify as Class C ashand the Coal Creek fly ash classifies as Class F ash, accordingto ASTM C618 (ASTM 2008). The remainders are referred to

820 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / SEPTEMBER 2011


as off-specification fly ashes because they do not meet the require-ments for either Class C or Class F fly ashes in ASTM C618. Inaddition, the fineness of the Dewey and Columbia ashes exceedsthe maximum for Classes C and F, and the pozzolanic activity at7 days of the Presque Isle ash does not meet the minimum forClasses C and F. The Dewey, King, and Columbia fly ashes arederived from subbituminous coals, the Presque Isle fly ash isderived from bituminous coal, and the Coal Creek and Stantonfly ashes are derived from burning lignite. All of the fly ashes,except for the Presque Isle fly ash, which was collected by fabricfilters, were collected by electrostatic precipitators and stored dryin silos.

Among the six fly ashes, Dewey has the highest carbon content(LOI ¼ 49%) and Coal Creek has the lowest carbon content(LOI ¼ 0:5%). King has the highest CaO content (25.8%) and Pre-sque Isle has the lowest CaO content (3.2%). Dewey and King havethe highest CaO=SiO2 ratios (1.15 and 1.08), Stanton and Colum-bia have midrange CaO=SiO2 ratios (0.5 and 0.7), and Presque Isleand Coal Creek have the lowest CaO=SiO2 ratios (0.1 and 0.2). Allof the fly ashes have less CaO and a smaller CaO=SiO2 ratio thanthe Type 1 portland cement (CaO content = 62%, CaO=SiO2 ratio =2.9). The fly ashes generally are comprised of silt-size particles(< 75 μm and > 2 μm), with a coarse fraction between 5% and50% and a < 2� μm fraction between 10% and 67%. Deweyand Columbia fly ashes have similar grain-size distributions andare somewhat finer than King, Coal Creek, and Stanton, which havesimilar grain-size distributions. Presque Isle fly ash has mostlyuniform size particles (∼0:03 mm).

pH

The pH of each soil was measured using both ASTM D4972(ASTM 2007e, for inorganic soils) and ASTM D2976 (ASTM2004b, for peats). These methods differ in the ratio of dry solidto distilled water that is used (1∶1 for D4972, 1∶16 for D2976).All three soils had near-neutral pH, and both test methods yieldeda similar pH.

The pH of each fly ash was measured using ASTM D5239(ASTM 2004a) and the procedure described in Eades and Grim(1966). ASTM D5239 uses a solid to distilled water ratio of 1∶4and a 2-h lag between mixing and pH measurement. The Eadesand Grim method uses a solid to distilled water ratio of 1∶5, alag of 1 h, and requires the use of CO2-free water. The pH ofthe each fly ash was also measured at 1, 2, 6, 24, 48, and 96 h aftermixing to assess the pH change over time; however, the pH did notvary significantly with time. All pH results at 1 h after mixing aregiven in Table 2.

Unconfined Compression Testing

Unconfined compression tests were conducted on specimens pre-pared from the soils and soil–fly ash mixtures following ASTMD5102 (ASTM 2009b). The strain rate was 0:21%=min, whichis the same rate used by Edil et al. (2006) for evaluating soil–fly ash mixtures prepared with inorganic soils. Test specimens wereprepared by first mixing the dry soil and the dry fly ash at the speci-fied fly ash content on dry weight basis. Subsequently, the amountof water required was added, and after a wait of 2 h (to simulatefield conditions), the mixture was compacted in a steel mold with adiameter of 33 mm and height of 71 mm. The compactive effort forspecimen preparation was adjusted in such a way that the sameimpact energy per unit volume, as in the standard Proctor effort[ASTM D698 (ASTM 2007a)], was applied. After the compaction,the specimens were extruded with a hydraulic jack, sealed inplastic, and cured for 7 days in a room maintained at 100% rela-tive humidity and 25°C. Although the tests were performed onT

able

1.IndexPropertiesandClassifications

ofSo

ilsTested

Soilname

LL

PIFines

content(%

)Activeclay

content

(<2μm

)(%

)OC

(%)

Gravelcontent

(>4:75

mm)(%

)G

s

Classification

pH

wN

γ d(kN=m

3)

wopt

USC

SAASH

TO

AST

MD4972

AST

MD2976

Markeypeat

531

2515

278

2.23

PtA-8

(0)

5.9

6.3

5710.3

47

Theresa

soil

318

7536

6—

2.57

OL

A-4

(5)

7.6

7.1

2015.2

21

Law

sonsoil

5019

9755

5—

2.58

OL-O

HA-7-5

(23)

6.9

6.8

2813.3

28

Boardman

silt

221

7912

1—

2.67

ML

A-2-4

(0)

——

1117.3

17

Note:

LL=liq

uidlim

it;PI

=plasticity

index;

OC=organiccontent[A

STM

D2974,(AST

M2007b)];G

s=specific

gravity

;wN=naturalwater

content;γ d

=maxim

umdryunitweight(A

STM

D698);wopt=

optim

umwater

content[(A

STM

D6698,(AST

M2007c)];USC

S=unifiedsoilclassificatio

nsystem

;AASH

TO=AASH

TOclassificatio

nsystem

(num

bersin

parenthesesindicatethegroupindex).F

ines

content

andgrainsize

diam

eters(for

Cucalculations)arebasedon

AST

MD422(A

STM

2007d).A

ctiveclay

content,specificgravity,and

liquidandplasticity

indexweredeterm

ined

follo

wingtheprocedures

inAST

MC837,

D854,

andD43

18,respectiv

ely(A

STM

2009a,

2010b,

2010a).



Table 2. Properties and Classifications of Fly Ashes Tested

Parameter Dewey King Presque Isle Coal Creek Columbia Stanton Typical Class C Typical Class F

SiO2 (%) 8.0 24.0 35.6 50.4 31.1 40.2 40.0 55.0

Al2O3 (%) 7.0 15.0 18.0 16.4 18.3 14.7 17.0 26.0

Fe2O3 (%) 2.6 6.0 3.5 7.2 6.1 8.7 6.0 7.0

CaO (%) 9.2 25.8 3.2 13.3 23.3 21.3 24.0 9.0

MgO (%) 2.4 5.3 1.0 4.3 3.7 6.6 5.0 2.0

CaO=SiO2 1.15 1.08 0.09 0.26 0.75 0.53 0.60 0.16

pH 9.9 10.9 11.3 11.9 12.8 11.7 — —Specific gravity 2.00 2.66 2.11 2.59 2.63 2.63 — —Fineness, max (%) 57 18 26 28 58 23 34 34

Strength activity at 7 days, min (%) 83 78 49 83 96 111 75 75

Loss on ignition, max (%) 49.0 12.0 34.0 0.5 0.7 0.8 6 6

Classification Off-spec Off-spec Off-spec Class F Class C Class C Class C Class F

Note: Off-spec = off-specification. Loss on ignition was measured per ASTM C311 (ASTM 2011) at 550°C.

Fig. 1. Unconfined compressive strength (qu) of mixtures prepared with various fly ashes, Type I portland cement, and Boardman silt at very wetwater content: (a) Markey peat; (b) Lawson soil; (c) Theresa soil



specimens cured 7 days to simulate the early curing conditions dur-ing construction, both inorganic and organic soils are expected tohave significant strength gains with increasing curing time forcalcium-based additives (Edil et al. 2006; Sakr et al. 2009).

Resilient Modulus Test

The resilient modulus is a widely used property in flexible pave-ment design, as explained in the AASHTO Guide for Design ofPavement Structures (AASHTO 1993), and it indicates the stiffnessof a soil under a confining stress and a repeated axial load. Resilientmodulus, Mr , is calculated based on the ratio of deviator stress andthe recoverable strain. Different confining and deviator stresses areapplied on the test specimens to cover the range of expected in situstresses.

Specimens for the resilient modulus test were prepared in a pol-yvinyl chloride (PVC) mold with a diameter of 102 mm and aheight of 203 mm in the same manner as the unconfined compres-sion test specimens were prepared. Compactive effort was adjustedin such a way that the same compaction energy per unit volume asthe one specified in the standard Proctor compaction method(ASTM D698) was applied (i.e., 600 kN=m3). Required compac-tive effort was obtained when the number of blows with the stan-dard Proctor hammer was 22 and the number of compacted layerswas 6. After compaction, specimens were cured for 7 days in a wetroom, maintained at 25°C and 100% humidity. Specimens were ex-truded from the PVC molds after curing and tested according toAASHTO T292 (AASHTO 1991). Side friction during extrusionwas minimized by applying a very thin grease layer between the

Fig. 2.Resilient moduli (Mr) of soil–fly ash mixtures prepared with various fly ashes and Boardman silt: (a) Markey peat; (b) Lawson soil; (c) Theresasoil (FA = fly ash, wet = wet of optimum, opt = optimum water content)



PVC mold and the soil. The loading sequence for cohesive soilswas followed, and the conditioning stress was applied as 21 kPainstead of 41 kPa because some specimens were too soft to with-stand 41 kPa conditioning stress. Confining stress was 21 kPa forall loading sequences, and the deviator stress was increased in stepsof 21, 34, 48, 69, and 103 kPa and applied 50 repetitions at eachstep. The reported Mr is the modulus obtained at the initial state ofstress, i.e., at 21 kPa confining and deviator stress because thestresses are relatively small in the subgrade level, and the modulusof the stabilized material does not depend strongly on stress level.

Results and Analysis

Soil–fly ash mixtures were prepared with fly ash contents (based ondry weight) of 10, 20, and 30%. Most of the tests were conductedon specimens prepared at a very wet condition, corresponding to6–14% wet of the OWC for the Lawson soil, 5–22% wet of theOWC for the Theresa soil, and 5–18% wet of the OWC for theMarkey peat. This very wet condition is intended to simulatethe natural water contents of soft subgrades in the upper Midwest-ern United States (Edil et al. 2006). Additional tests were con-ducted with the soil fraction at OWC per standard Proctor.These tests were conducted as well-defined control conditionsand to assess the effect of water content. For the specimensprepared at OWC, fly ash contents were only 10% and 20%(the specimens were unrealistically dry for reactions with 30%

fly ash). Soil-cement mixtures were prepared at the very wetcondition with 10% cement, and only unconfined compression testswere conducted on these mixtures. The cement dosage chosen(10%) is greater than the typical dosage for inorganic soils becauseof the organic content and also to provide a direct comparison with10% fly ash content.

General Effectiveness of Fly Ash Stabilization

Unconfined compressive strengths (qu) of the soil–fly ash mixturesprepared at the very wet condition are shown as a function of fly ashtype in Fig. 1. The qu of mixtures prepared with organic soil andBoardman silt (nonreactive additive) or Type 1 portland cement(a highly reactive binder) are also included in Fig. 1 for comparison.Also shown in Fig. 1 are qu of each soil alone (without fly ash)when compacted at the very wet condition. Triplicate specimenswere tested for unconfined compressive strength as quality control,and the averages of these tests are reported as results. Addition offly ash to the organic soils resulted in significant increase in qurelative to that of the unstabilized soil in the very wet condition.Once stabilized with fly ash, both the Lawson and Theresasoils classify as at least stiff subgrade [qu between 100 and200 kPa (Bowles 1979)], instead of soft (25–50 kPa) or very soft(0–25 kPa) in their unstabilized very wet conditions. qu exceeding100 kPa was not always obtained for the Markey peat in the verywet conditions, but adding fly ash to the Markey peat did increasethe qu by a factor of up to 10. It is clear from Fig. 1 that the final qu

0

100

200

300

400

500

600

0 5 10 15 20 25 30 35

Lawson Soil, OWC+ (8-14)%

DeweyKingP. IsleCoal CreekColumbiaStanton

Unc

onfin

ed C

ompr

essi

ve S

tren

gth

(kP

a)

Fly Ash Percentage

(a)

0

20

40

60

80

100

120

140

0 5 10 15 20 25 30 35

Lawson Soil, OWC + (9-14)%

DeweyKingP.IsleCoal CreekColumbiaStanton

Res

ilien

t Mod

ulus

(M

Pa)

Fly Ash Percentage

(b)

0

100

200

300

400

500

600

0 5 10 15 20 25 30 35

Theresa Soil, OWC+ (8-22)%


Unc

onfin

ed C

ompr

essi

ve S

tren

gth

(kP

a)

Fly Ash Percentage

(c)

0

20

40

60

80

100

120

0 5 10 15 20 25 30 35

Theresa Soil, OWC+ (5-11)%


Res

ilien

t Mod

ulus

(M

Pa)

Fly Ash Percentage

(d)

Fig. 3. Engineering properties of organic soil–fly ash mixtures as a function of fly ash percentage in the mixture: (a) qu of stabilized Lawson soil;(b) Mr of stabilized Lawson soil; (c) qu of stabilized Theresa soil; (d) Mr of stabilized Theresa soil



achieved varies depending on the organic soil and the fly ash. Thisis in contrast to the findings reported for inorganic soils stabilizedwith different fly ashes by Edil et al. (2006), for which finalstrengths were comparable, although strength factors varied.

Fig. 2 shows the resilient moduli of soil–fly ash mixtures as afunction of binder type and content. The Mr for Markey peat andLawson and Theresa soils are reported at their OWCs because thesesoils were too soft to be tested at very wet conditions, i.e., 13% wetof OWC for Markey peat and 10% wet of OWC for Lawson andTheresa soils. The resilient modulus of Markey peat, even with30% fly ash, never reached 35MPa at very wet conditions, meaningthat Markey peat can be considered a very soft subgrade, i.e.,Mr < 35 MPa (Asphalt Institute 1999). Markey peat–Boardman

silt mixtures were too soft at very wet conditions to withstandthe conditioning stress, indicating that the addition of fly ash ismore effective than the addition of silt at very wet conditions.

Lawson soil admixed with 20% Dewey or Columbia fly ashes atvery wet conditions was medium-stiff, i.e., Mr ∼ 85 MPa (AsphaltInstitute 1999). When stabilized with 30% Dewey, King, Stanton,or Columbia fly ash, Lawson soil had a resilient modulus as highas 110 MPa at very wet conditions. At OWC, the resilient modulusof stabilized Lawson soil was always, even with 10% fly ash,higher than 50 MPa. Theresa soil admixed with 20% Dewey, King,Stanton, or Columbia fly ashes at very wet conditions had resilientmoduli of 50–70 MPa. When the percentage of these fly ashes wasincreased to 30% at very wet conditions, the resilient modulus

0

100

200

300

400

500

600

0 10 20 30 40 50

30% Fly ash

Markey Peat, OWC+ (10-14)%Lawson Soil, OWC+ (8-14)%Theresa Soil, OWC+ (8-11)%

Unc

onfin

ed C

ompr

essi

ve S

tren

gth

(kP

a)

LOI (%) of Fly Ash

(a)

r= -0.20t= -2.47

0

20

40

60

80

100

120

140

0 10 20 30 40 50

30% Fly ash


Res

ilien

t Mod

ulus

(M

Pa)

LOI of Fly Ash (%)

(b)

r= -0.03t= -0.20

0

100

200

300

400

500

600

9.5 10 10.5 11 11.5 12 12.5 13

Unc

onfin

ed C

ompr

essi

ve S

tren

gth

(kP

a)

pH of Fly Ash

(c)

r= 0.11t= 1.33

0

20

40

60

80

100

120

140

9.5 10 10.5 11 11.5 12 12.5 13

Res

ilien

t Mod

ulus

(M

Pa)

pH of Fly Ash

(d)

r = -0.08t = -0.64

0

100

200

300

400

500

600

10 20 30 40 50 60

Unc

onfin

ed C

ompr

essi

ve S

tren

gth

(kP

a)

Fineness of Fly Ash (%)

(e)

r= -0.12t= -1.47

0

20

40

60

80

100

120

140

10 20 30 40 50 60

Res

ilien

t Mod

ulus

(M

Pa)

Fineness of Fly Ash (%)

(f)

r = 0.1t = 0.76

Fig. 4. Engineering properties of soil–fly ash mixtures: (a) qu as a function of LOI of fly ash; (b)Mr as a function of LOI of fly ash; (c) qu as a functionof pH of fly ash; (d) Mr as a function of pH of fly ash; (e) qu as a function of fineness of fly ash; (f) Mr as a function of fineness of fly ash



varied between 65 and 105 MPa, indicating that the stabilizationprocess produced significant improvement in resilient modulus(i.e., medium-stiff subgrade consistency) considering that untreatedsoil (no fly ash) was too soft to be tested. At OWC, the resilientmodulus of stabilized Theresa soil varied between 50 and 130 MPa,depending upon the fly ash type and percentage used. Admixing10% fly ash with any of the three soils at very wet conditions failedto yield a resilient modulus greater than 50 MPa.

Comparison of the qu or resilient moduli obtained with differentfly ashes indicates that the criteria used to define fly ashes for con-crete applications (Class C) are not necessarily indicative of theeffectiveness for soil stabilization. For example, in some casesDewey and King fly ashes (both are off-specification fly ashes)resulted in comparable or greater strength and stiffness gain thanColumbia and Stanton fly ashes, which are Class C ashes andqualify for use as concrete additives.

0

100

200

300

400

500

600

0 0.2 0.4 0.6 0.8 1 1.2

30% Fly ash


Unc

onfin

ed C

ompr

essi

ve S

tren

gth

(kP

a)

CaO/SiO2 of Fly Ash

(a)

Coal Creek Fly Ash

DeweyFly Ash

r= 0.38t=4.84

0

20

40

60

80

100

120

140

0 0.2 0.4 0.6 0.8 1 1.2

30% Fly ash


Res

ilien

t Mod

ulus

(M

Pa)

CaO/SiO2

of Fly Ash

(b)

r= 0.30t=2.33

0

100

200

300

400

500

600

0 0.2 0.4 0.6 0.8 1

Unc

onfin

ed C

ompr

esiv

e S

tren

gth

(kP

a)

CaO/(SiO2+Al

2O

3)

(c)

r=0.43t=5.60 0

20

40

60

80

100

120

140

0 0.2 0.4 0.6 0.8 1

Res

ilien

t Mod

ulus

(M

Pa)

CaO/(SiO2+Al

2O

3)

(d)

r= 0.31t=2.45

0

100

200

300

400

500

600

0 5 10 15 20 25 30

Unc

onfin

ed C

ompr

essi

ve S

tren

gth

(kP

a)

CaO Content (%) of Fly Ash

(e)

DeweyFly Ash

Coal CreekFly Ash

r= 0.46t= 6.10

0

20

40

60

80

100

120

140

0 5 10 15 20 25 30

Res

ilien

t Mod

ulus

(M

Pa)

CaO Content of Fly Ash (%)

(f)

Coal Creek FA

Dewey FA

r= 0.22t=1.69

Fig. 5. Engineering properties of soil–fly ash mixtures, (a) qu as a function of CaO=SiO2 ratio of fly ash; (b)Mr as a function of CaO=SiO2 ratio of flyash; (c) qu as a function of CaO=ðSiO2 þ Al2O3Þ ratio of fly ash; (d)Mr as a function of CaO=ðSiO2 þ Al2O3Þ ratio of fly ash; (e) qu as a function ofCaO content of fly ash; (f) Mr as a function of CaO content of fly ash



The effect of reactivity of the binder can be evaluated by com-paring the qu of the soil–fly ash mixtures to the qu obtained usingcement or nonreactive Boardman silt as the additive in Fig. 1. quobtained with 10% cement at the very wet conditions was alwayshigher than those obtained with 10% fly ash at the same water con-tent, and in many cases 10% cement resulted in higher qu than ob-tained with 30% fly ash. In contrast, the mixtures prepared withBoardman silt had lower qu and resilient moduli than comparablesoil–fly ash mixtures. Thus, the increase in strength or resilientmodulus obtained by fly ash stabilization generally is attributableto chemical reactions and the reduction in water content obtainedby adding dry solids, but the significance of the reactions dependson the type of fly ash and the soil.

The importance of reactivity is also illustrated through the effectof fly ash content. For most of the mixtures, the qu and resilientmodulus increased as the fly ash content increased (Fig. 3). Theexceptions are the mixtures prepared with the less reactive fly ashes(Presque Isle and Coal Creek). Additionally, qu and resilient modu-lus do not increase linearly with fly ash content. In most cases, theincrease in qu and resilient modulus obtained as the fly ash contentincreased from 0–10% or 10–20% was larger than those obtainedwhen the fly ash content was increased from 20–30%. Thus, thebenefits accrued by adding more fly ash diminish as the fly ashcontent increases.

Effects of Fly Ash Characteristics

Graphs relating qu and resilient modulus to properties of the fly ash(LOI, pH, fineness, CaO=SiO2 ratio, CaO=ðSiO2 þ Al2O3) ratio,

and CaO content,) were prepared to identify characteristics ofthe fly ashes that have an important role in improving the strengthand stiffness of the organic soils (Figs. 4 and 5). qu and resilientmoduli of mixtures prepared at the very wet condition are shownbecause this condition is of practical interest for field situations(Edil et al. 2006). The resilient modulus data from both cells werecompared using a paired t-test at significance level of 0.05, corre-sponding to tcr ¼ 1:96 for unconfined compression test results andtcr ¼ 2:01 for resilient modulus test results.

Fig. 4 suggests that qu and resilient modulus are not affected byLOI, pH, or fineness (percentage retained on 45 μm sieve) of the flyash. This observation is consistent with the statistical analysis,which shows that qu and Mr are not correlated with LOI, pH, orfineness (t < 1:96 for qu and t < 2:01 for Mr). In contrast, quand resilient modulus suggest a correlation with CaO=SiO2 andCaO=ðSiO2 þ Al2O3), and the statistical analysis supports this ob-servation (Fig. 5). Relatively strong relationships exist between quor resilient modulus and these parameters for the Lawson andTheresa soil, whereas weaker relationships exist for the Markeypeat, thus the Markey peat data are excluded for calculation of cor-relation coefficient (r) and t. The relationships between qu andCaO=SiO2 and CaO=ðSiO2 þ Al2O3Þ for the Lawson and Theresasoils are illustrated with second-order nonlinear regressions, shownas solid lines in Figs. 5(a) and 5(c). Statistically, CaO content as-sociated with qu but not with resilient modulus.

Fig. 5 suggests that both CaO and CaO=SiO2 or CaO andCaO=ðSiO2 þ Al2O3Þ are important variables affecting the qu ofthe soil–fly ash mixtures prepared with the Lawson and Theresa

0

100

200

300

400

500

600

0 5 10 15 20 25 30

10% Dewey FA10% King FA10% Columbia FA20% Dewey FA20% King FA20% Columbia FA

Unc

onfin

ed C

ompr

essi

ve S

tren

gth

(kP

a)

OC of Soil (%)

(a)

r= 0.02t= 0.19

0

20

40

60

80

100

120

140

0 5 10 15 20 25 30

Res

ilien

t Mod

ulus

(M

Pa)

OC of Soil (%)

(b)

r= 0.17t=1.32

0

100

200

300

400

500

600

0 5 10 15 20 25 30 35 40

(c)

Unc

onfin

ed C

ompr

essi

ve S

tren

gth

(kP

a)

Pl

r= -0.02t= -0.19

0

20

40

60

80

100

120

140

0 5 10 15 20 25 30 35 40

Res

ilien

t Mod

ulus

(M

Pa)

PI

(d)

r= -0.17t=-1.32

Fig. 6. Engineering properties of soil–fly ash mixtures prepared at very wet conditions, (a) qu as a function of OC of soil; (b)Mr as a function of OCof soil; (c) qu as a function of PI soil; (d) Mr as a function of PI soil



soils. The highest qu and resilient moduli were obtained when theCaO content was at least 10, CaO=SiO2 ratio was between 0.5 and1.0, and CaO=ðSiO2 þ Al2O3Þ ratio was between 0.4 and 0.7.A similar conclusion can be drawn for the Markey peat, althoughthe trends in qu and resilient modulus for the Markey Peat aremodest. As illustrated in Figs. 5(e) and 5(f), CaO content aloneis not sufficient to evaluate whether fly ash will cause an increasein qu or resilient modulus. The circled data in Figs. 5(e) and 5(f)correspond to mixtures prepared with the Lawson and Theresa soilsand Coal Creek (CaO content = 13.3%) or Dewey fly ash (CaOcontent = 9.2%). Appreciably higher qu and resilient moduli areobtained with Dewey fly ash. This is attributed to its significantlyhigher CaO=SiO2 ratio (1.15 versus 0.26 of Coal Creek) eventhough Coal Creek fly ash has greater CaO than that of Deweyfly ash. The results indicate that CaO of 10% by weight is neededas a threshold value for strength gain, and CaO content andCaO=SiO2 ratio play a combined role on qu and resilient modulusof soil–fly ash mixtures.

Effects of Soil Type

The influence of organic soil type was evaluated by graphing qu andresilient modulus against organic content (OC) and PI (Fig. 6). SoilpH was not included in the analysis because the pH varied over anarrow range (6.1–7.3). As in the analysis of fly ash properties, the

qu and resilient moduli of mixtures prepared shown in Fig. 6correspond to the very wet condition.

Data for soil–fly ash mixtures from the study conducted by Edilet al. (2006) were also included in the analysis to increase the gen-erality of the findings. Edil et al. (2006) used Dewey, King, andColumbia fly ashes that were obtained from the same source asthe fly ashes used in this study. Edil et al. (2006) used a varietyof soils with OCs ranging from 1–10% and PIs ranging from15–38, and they mixed these soils with the three fly ashes. qu datawere adopted from their study for differentmixtures, each having oneof the following soils: inorganic clay (OC = 2%, PI = 38), slightlyorganic clay (OC = 4%, PI = 35) and organic clay (OC = 10%and PI = 19).

As shown in Fig. 6(a), qu decreased significantly as the OC in-creased to 10%, and then leveled off for higher OCs. A sharp de-crease in resilient modulus in response to an increase in OC of soilwas also observed in Fig. 6(b). This inverse relationship betweenqu or resilient modulus and OC may reflect the inhibition ofpozzolanic reactions by organic matter. Alternatively, the inverserelationship between qu or resilient modulus and OC mayreflect the weakness of organic solids relative to mineral solids.

0

100

200

300

400

500

0 100 200 300 400 500

Markey, OWC+ (10-18)%Markey, OWC± 5%Lawson, OWC+ (8-14)%

Lawson, OWC± 5%Theresa, OWC+ (7-22)%Theresa, OWC± 3

Unc

onfin

ed C

ompr

essi

ve S

tren

gth

of M

ixtu

res

with

Boa

rdm

an S

ilt (

kPa)

Unconfined Compressive Strength of Mixtures with Fly Ashes(kPa)

(a)

0

20

40

60

80

100

120

140

0 20 40 60 80 100 120 140

Markey Soil, OWC± 5%Lawson Soil, OWC+ (8-14)%Lawson Soil, OWC± 5 %Theresa Soil, OWC+ (8-22)%Theresa Soil, OWC± 2%

Res

ilien

t Mod

ulus

of S

ampl

es

with

Boa

rdm

an S

ilt (

MP

a)

Resilient Modulus of Samples with Fly Ashes (MPa)

(b)

Fig. 8. Comparison of engineering properties of mixtures preparedwith Boardman silt and fly ashes at the same binder content and similarwater content: (a) qu; (b) Mr

0

100

200

300

400

500

600

0 100 200 300 400 500 600

Markey & 10% FAMarkey & 20% FALawson & 10% FALawson & 20 % FATheresa & 10% FATheresa & 20% FA

Unc

onfin

ed C

ompr

essi

ve S

tren

gth

(kP

a)at

aro

und

OW

C

Unconfined Compressive Strength (kPa)at Wet of OWC

1:1 Line(a)

0

20

40

60

80

100

120

140

0 20 40 60 80 100 120 140

Markey PeatLawson SoilTheresa Soil

Res

ilien

t Mod

ulus

(M

Pa)

at a

roun

d O

WC

Resilient Modulus (MPa)at Wet of OWC

1:1 Line

(b)

Fig. 7. Engineering properties soil–fly ash mixtures prepared at opti-mum and wet of optimum water contents with the same binder type andpercentages: (a) qu; (b) Mr



Regardless, the trend in Fig. 6(a) suggests that the effectivenessof fly ash stabilization is significantly reduced when the OCexceeds 10%.

The effect of PI on qu and resilient modulus is shown inFigs. 6(c) and 6(d). Greater qu and resilient moduli are obtainedwhen the PI is eight or more. However, the apparent effect ofPI in Fig. 6(c) is probably spurious. The trend is more likely relatedto OC because the Markey peat had the highest OC and the lowestPI of the soils that were tested. A broader range of soils is needed toadequately assess the effect of PI.

Effects of Water Content

The effect of water content on the stabilization was investigated byplotting the qu and resilient modulus of the soil–fly ash mixtureprepared at very wet of OWC condition against the qu and resilientmodulus of the soil–fly ash mixture prepared at OWC (Fig. 7).The resilient moduli of soil–fly ash mixtures prepared at OWCwere almost always higher than those of prepared at very wet ofOWC. When the fly ash percentage was 10%, the soil–fly ash mix-tures prepared at OWC usually had higher qu, as opposed to thoseprepared at wet of OWC.On the other hand, as the fly ash percentage

0

100

200

300

400

500

7 8 9 10 11 12 13 14

Lawson Soil(a)

10% FA, OWC+ (8-13)%20% FA, OWC+ (8-12)%30% FA, OWC+ (8-14)%

Unc

onfin

ed C

ompr

essi

ve S

tren

gth

(kP

a)

pH of the mixture

0

20

40

60

80

100

120

140

7 8 9 10 11 12 13 14


Res

ilien

t Mod

ulus

(M

Pa)

pH of the mixture

Lawson Soil (b)

0

100

200

300

400

500

7 8 9 10 11 12 13 14

Theresa Soil(c)


Unc

onfin

ed C

ompr

essi

ve S

tren

gth

(kP

a)

pH of the mixture

0

20

40

60

80

100

120

140

7 8 9 10 11 12 13 14


Res

ilien

t Mod

ulus

(M

Pa)

pH of the mixture

Theresa Soil (d)

0

50

100

150

7 8 9 10 11 12 13 14

Markey Peat(e)

10% FA, OWC+ (13-18)%20% FA, OWC+ (11-17)%30% FA, OWC+ (10-14)%U

ncon

fined

Com

pres

sive

Str

engt

h (k

Pa)

pH of the mixture

5

10

15

20

25

30

7 8 9 10 11 12 13 14

10%FA20%FA, OWC+ (5-8)%30%FA, OWC+ (7-13)%

Res

ilien

t Mod

ulus

(M

Pa)

pH of the mixture

Markey Peat(f)

Fig. 9. Engineering properties of soil–fly ash mixtures prepared at very wet water content as a function of mixture pH after 1 h: (a) qu for stabilizedLawson soil; (b) Mr for stabilized Lawson soil; (c) qu of stabilized Theresa soil; (d) Mr of stabilized Theresa soil; (e) qu for stabilized Markey peat;(f) Mr for stabilized Markey peat



increased to 20%, soil–fly ash mixtures prepared at OWC usuallyhad lower qu than mixtures prepared at very wet of OWC, unlikethe 10% fly ash case. The shear strength of a cohesive soil generallyis inversely related to water content (Seed and Chan 1959; Khouryand Zaman 2004). On the other hand, observed increase in qu aswater content increases can be attributed to the use of more waterin hydration that can increase the amount of cementitious products.

The qu values of mixtures prepared with Boardman silt (non-reactive binder) are plotted against qu of mixtures prepared withthe fly ashes in Fig. 8. In nearly all cases, the soil–fly ash mixtureshad higher qu and resilient moduli than the mixtures prepared withBoardman silt for the very wet condition. However, at OWC, the qutended to be more similar for the soil–fly ash mixtures and themixtures prepared with Boardman silt. That is, the reactivity effectappears to diminish as the water content decreases. It appears that ifinitial water content is less than a critical amount needed forhydration reactions, strength gain may be limited.

Effects of pH of the Soil–Fly Ash Mixture

For soils stabilized with cement and blast furnace slag, if organiccontent is below 15%, in general there is significant strength gainonly if humic acid is less than 0.9% or pH higher than 5 (Kitazume2005). For organic soils, it is well known that humic acids consumethe calcium ions in the binder. When the acids are neutralized, theremaining binder quantity contributes to strength gain. Tremblayet al. (2002) mixed 14 different organic compounds with thesoil-cement mixture (the two soils were a clay and a silt, andthe two cements were ordinary portland Type 10 and sulfate-richgeolite 20) and investigated the effect of organic compound on thesoil stabilization. They reported that if an organic compoundcaused a pore solution pH of less than 9, no strength gain wasnoted. However, they also mentioned that a pore solution pH ofmore than 9 did not always indicate significantly high strengths.

pHs measurements conducted 1, 2, 24, 48, and 96 h after mixingwere not significantly different. When Lawson and Theresa soilswere mixed with a fly ash with CaO content higher than 10, thepH of the mixture reached above 9, which indicates that cementi-tious reactions are not likely to be inhibited (Tremblay et al. 2002).The pH of the mixtures involving Markey peat were also above 9 asthe percentage of fly was increased to 30%. The effect of pH on thequ and stiffness of the soil–fly ash mixtures is shown in Fig. 9.There is no apparent relationship between qu or resilient modulusand mixture pH. Fig. 9 seems to verify Tremblay et al.’s conclusionthat pH higher than 9 does not necessarily indicate higher qu.

Correlations between UCS and Resilient Modulus TestResults

The relationships between unconfined compressive strengths andresilient moduli at 21 kPa deviator stress for organic soil–flyash mixtures with the same fly ash type and percentage, preparedat the same water content, and cured for the same length of time aregiven in Fig. 10. Fig. 10 includes qu data from two different tests:(1) tests on small-size specimens (33 mm in diameter and 72 mm inheight) that were not subjected to resilient modulus testing, and(2) tests on large specimens (102 mm in diameter and 203 mmin height) that were previously tested in a resilient modulus test.However, only one set of resilient modulus test data was usedin correlation with both sets of unconfined compression test datafor a given soil–fly ash mixture. According to Fig. 10(a), whichincludes qu for small-size specimens, the conversion factor forqu (kPa) to obtain resilient modulus (kPa) varies from 70–570,and the best fit is 270. In Fig. 10(b), in which qu testing is per-formed on larger samples subjected to resilient modulus testingprior to testing, there is much less dispersion of the data, and

the conversion factor from qu (kPa) to resilient modulus (kPa) is213 and close to the best fit given in Fig. 10(a). The coefficientscorresponding to the slope of curve fit in Figs. 10(a) and 10(b)are close.

The secant modulus at 50% (E50) was obtained by dividing halfof the peak strength (qu=2) with the strain observed at that stresslevel in the unconfined compression test. Comparison of E50 withresilient modulus is given in Fig. 11. Fig. 11(a) shows the compari-son of E50 obtained from the unconfined compression tests per-formed on small specimens and resilient moduli obtained fromthe tests performed on large specimens. In Fig. 11(a), resilientmodulus varies between 1:6E50 and 20E50. Fig. 11(b) depictsthe comparison of E50 and resilient moduli results that were ob-tained by using the same specimens (larger specimens) in uncon-fined compression and resilient modulus tests. In this case, resilientmodulus varies between 1:8E50 and 12E50. In both cases, resilientmodulus is higher than E50.

Model for Stabilization of Organic Soils with Fly Ashes

The important factors in stabilization of organic soils with fly ashcan be summarized as follows: (1) fly ash properties: CaO contentand CaO=SiO2 ratio; (2) soil properties: OC; and (3) mixture char-acteristics: fly ash content and water content. Each of these vari-ables was included in a nonlinear regression analysis to find an

0

50000

100000

150000

200000

250000

300000

0 100 200 300 400 500

Markey SoilLawson SoilTheresa Soil

Res

ilien

t Mod

ulus

(kP

a)

Unconfined Compressive Strength (kPa)

(a)

Mr=570 qu

Mr=270 qu

R2=0.54

Mr=70 qu

0

20000

40000

60000

80000

100000

120000

140000

0 100 200 300 400 500 600 700

(b)


Res

ilien

t Mod

ulus

(kP

a)

Post-Mr Unconfined Compressive Strength (kPa)

Mr=213qu

R2=0.84

Fig. 10. Relations between Mr and qu of soil–fly ash mixtures: (a)unconfined compression tests performed on 133-mm-diameter and72-mm-high specimens and resilient modulus tests performed on102-mm-diameter and 203-mm-high specimens; (b) 102-mm-diameterand 203-mm-high specimens for both unconfined compression andresilient modulus tests



equation that can be used to predict the qu of organic soil–fly ashmixtures. The equation was derived statistically as follows: (1) trialsof linear and quadratic curvilinear regression models between sig-nificant characteristics and unconfined compressive strength werepreformed to see which model best described the relation based onF-test; and (2) a multiple regression model including all significantcharacteristics. Multiple regression models included second-orderor transformed functions of significant characteristics investigated.Possible correlations between independent variables were alsochecked, and highly correlated variables were dropped from themodel. Only data for the very wet condition were included becausethis condition is of practical importance. The qu of the soil alonewas also included in the analysis. The following regression modelwas developed:

qu-treated ¼ �320þ 795ðCaO=SiO2Þ � 573ðCaO=SiO2Þ2� 125;673ðe�OCÞ þ 6ðFApercÞ þ 25ðqu-untreatedÞ� 33ðpHmixtureÞ ð5Þ

where FAperc = fly ash percentage; pHmixture = pH of the soil–fly ashmixture after 1 h; qu-treated = stabilized unconfined compressive

strength of soil (kPa) after 7 days of curing; qu-untreated = unconfinedcompressive strength of untreated soil (kPa); and OC = organiccontent of soil (%). The developed model was intended to representthe data obtained from a wide range of organic soils and fly ashes,but it has not been validated on independently obtained data. Acomparison of the predicted versus measured unconfined compres-sive strength is shown in Fig. 12. According to Fig. 12, the regres-sion model represents the qu data reasonably well, with R2 = 0.71.

According to Eq. (5), the following inferences can be made:(1) there is an optimum ðCaO=SiOÞ2 ratio that maximizes the sta-bilized strength of the soil; (2) increase in the fly ash percentageincreases the qu of the soil–fly ash mixture; and (3) higher organiccontent of the soil indicates less qu of the soil–fly ash mixture. Themodel does not include CaO because it is highly correlated withother terms in the model. However, the physical effect of CaOcontent is still reflected in the model by mixture pH term, whichis controlled by the ½OH�� ions liberated after disassociation ofCaðOHÞ2 (formed by hydration of CaO). The unconfined compres-sive strength can be correlated to resilient modulus within a rangeof uncertainty. Alternatively, resilient modulus tests can be per-formed on promising mixtures.

Conclusions

The objective of this study was to determine if unconfined com-pressive strength and resilient modulus of soft organic soils canbe increased by blending fly ash into the soil. Tests were conductedwith three organic soils and six fly ashes. Portland cement and aninorganic silt were also used as a stabilizer for reference purposes.Fly ashes were mixed with soils at three different percentages andtwo different water contents (OWC and 9–15% wet of the OWC).The following conclusions are advanced:1. Unconfined compressive strength of organic soils can be in-

creased using fly ash, but the amount of increase dependson the type of soil and characteristics of the fly ash. Large in-creases in qu (from 30 kPa without fly ash to > 400 kPa withfly ash) were obtained for two clayey soils with an OC lessthan 10% when blended with some of the fly ashes. More mod-est increases in qu (from 15 kPa without fly ash to > 100 kPawith fly ash) were obtained for a highly organic sandy siltypeat with OC ¼ 27%. Resilient modulus tests could not be per-formed on organic soils without fly ash stabilization at wetconditions because the specimens were too soft. The addition

0

20

40

60

80

100

120

140

0 20 40 60 80 100 120 140


Res

ilien

t Mod

ulus

(M

Pa)

E50

(MPa)

Mr=1.6E

50

Mr=20E

50

(a)

0

20

40

60

80

100

120

140

0 20 40 60 80 100 120 140


Res

ilien

t Mod

ulus

(M

Pa)

E50

(MPa)

Mr=1.8E

50

Mr=12E

50

(b)

Fig. 11. Relations between secant modulus at (E50) andMr of soil–flyash mixtures: (a) unconfined compression tests performed on 133-mm-diameter and 72-mm-high specimens and resilient modulus testsperformed on 102-mm-diameter and 203-mm-high specimens;(b) 102-mm-diameter and 203-mm-high specimens for both uncon-fined compression and resilient modulus tests

0

100

200

300

400

500

600

0 100 200 300 400 500 600Pre

dict

ed U

ncon

fined

Com

pres

sive

Str

engt

h (k

Pa)

Measured Unconfined Compressive Strength (kPa)

1:1 Line

Fig. 12. Predicted versus measured unconfined compressive strengthof soil–fly ash mixtures



of fly ash, at wet conditions, to the slightly organic soils,Lawson and Theresa (OC = 5 and 6%, respectively) producedMr varying from 10–100 MPa, depending on the type and per-centage of the fly ash. At OWC, Mr for these soils could beimproved up to 120 MPa with the addition of fly ash. However,for Markey peat (OC ¼ 27%), stabilization with fly ash neverproducedMr > 30 MPa no matter which fly ash type and per-centage (up to 30%) was used.

2. The significant characteristics of fly ash affecting the increasein qu and Mr include CaO content and CaO=SiO2 ratio [orCaO=ðSiO2 þ Al2O3Þ ratio]. The highest qu and Mr wereobtained when the CaO content was greater than 10% andthe CaO=SiO2 ratio was 0.5–0.8. Comparable increases inqu andMr were obtained with the Class C ashes, normally usedin concrete applications, and the off-specification fly ashesmeeting the aforementioned criteria for CaO content andCaO=SiO2 ratio. However, much lower qu and Mr wereobtained with one off-specification fly ash primarily becauseof its low CaO content and CaO=SiO2 ratio. Carbon content ofthe fly ash (i.e., loss on ignition) seemed to have no bearing onthe qu and Mr of the soil–fly ash mixtures.

3. For most of the cases qu and Mr increased when fly ash per-centage was increased. Exceptions were mixtures with the lessreactive Presque Isle and Coal Creek fly ashes (CaO < 10%and CaO=SiO2 < 0:5)

4. The reactivity effect appears to diminish as the water contentdecreases, i.e, improvement in the qu of the soil due to the ad-dition of fly ash or inorganic silt to the soil was approximatelythe same for the mixtures prepared at OWC. When the fly ashpercentage in the mixture was 10%, the expected trend of high-er qu when water content decreased was observed. On the otherhand, as the fly ash percentage increased to 20% (more reduc-tion in water content compared to 10% fly ash case), soil–flyash mixtures prepared wet of OWC usually had greater qu thanthe ones prepared at OWC. The trend of stronger mixtures atwet conditions as opposed to the mixtures prepared at OWC isattributable to the requirement for more water for hydrationreactions of the higher amount of fly ash.

5. Soil organic content is a detrimental characteristic for stabili-zation. An increase in the organic content of soil indicates thatthe strength of the soil–fly ash mixture will decrease exponen-tially. No effect of soil pH and plasticity could be discerned onresilient modulus of the soil stabilized with fly ash. However,more research on the effect of these characteristics is requiredbecause the variation in pH and plasticity of the soils in thisstudy was not sufficient.

6. Fly ash stabilization of soils at OWC always resulted in greaterresilient moduli than at wetter conditions. Resilient moduluscan be estimated from unconfined compressive strength usinga multiplication factor between 70 and 570. Estimation of re-silient modulus based on static E50 obtained from the uncon-fined compression test can be made using a multiplicationfactor in the range of 1.6–20, which shows that the lower-strainresilient modulus is always higher than the high-strain E50.

Acknowledgments

Financial support for this study was provided by the NationalScience Foundation (NSF) and the U.S. Federal Highway Admin-istration Recycled Materials Resource Center (RMRC). Fly ashesused in the study were provided by Alliant Energy, Xcel Energy,We Energy, Great River Energy, and LaFarge North America. Thefindings and opinions in this report are solely those of the authors.

Endorsement by NSF, RMRC, or the fly ash suppliers is notimplied and should not be assumed.

References

AASHTO. (1993). Guide for design of pavement structures, Washington,DC.

AASHTO. (1991). “Standard method of test for resilient modulus of sub-grade soils and untreated base/subbase materials.” T292, Washington,DC.

Acosta, H. A., Edil, T. B., and Benson, C. H. (2003). “Soil stabilization anddrying using fly ash.” Geo Engineering Rep. No. 03-03, Dept. of Civiland Environmental Engineering, Univ. of Wisconsin, Madison, WI.

Arman, A., and Munfakh, G. A. (1972). “Lime stabilization of organicsoils.” Highway Research Record 381, Highway Research Board,Washington, DC, 37–45.

Asphalt Institute. (1999). “Thickness design—Asphalt pavements forhighways and streets.” Manual series No. 1, Lexington, KY.

ASTM. (2004a). “Standard practice for characterizing fly ash for use in soilstabilization.” D5239, West Conshohocken, PA.

ASTM. (2004b). “Standard test method for pH of peat materials.”D2976-71, West Conshohocken, PA.

ASTM. (2007a). “Standard test methods for laboratory compaction char-acteristics of soil using standard effort.” D698, West Conshohocken,PA.

ASTM. (2007b). “Standard test method for moisture, ash, and organicmatter of peat and other organic soils.” D2974, West Conshohocken,PA.

ASTM. (2007c). “Standard test method for on-line measurement of turbid-ity below 5 NTU in water.” D6698, West Conshohocken, PA.

ASTM. (2007d). “Standard test method for particle-size analysis of soils.”D422-63, West Conshohocken, PA.

ASTM. (2007e). “Standard test method for pH of soils.” D4972-01, WestConshohocken, PA.

ASTM. (2008). “Standard specification for coal fly ash and raw or calcinednatural pozzolan for use in concrete.” C618, West Conshohocken, PA.

ASTM. (2009a). “Standard test method for methylene blue index of clay.”C837, West Conshohocken, PA.

ASTM. (2009b). “Standard test method for unconfined compressivestrength of compacted soil-lime mixtures.” D5102, West Consho-hocken, PA.

ASTM. (2010a). “Standard test methods for liquid limit, plastic limit, andplasticity index of soils.” D4318, West Conshohocken, PA.

ASTM. (2010b). “Standard test methods for specific gravity of soil solidsby water pycnometer.” D854, West Conshohocken, PA.

ASTM. (2011). “Standard test methods for sampling and testing fly ash ornatural pozzolans for use in portland-cement concrete.” C311, WestConshohocken, PA.

Axelsson, K., Johansson, S. E., and Andersson, R. (2002). “Stabilization oforganic soils by cement and puzzolanic reactions—Feasibility study.”Rep. 3, Swedish Deep Stabilization Research Center, Linköping,Sweden.

Bin-Shafique, S., Edil, T., Benson, C., and Senol, A. (2004). “Incorporatinga fly ash stabilized layer into pavement design—Case study.” Proc. ICEGeotech. Eng., 157(4), 239–249.

Bowles, J. E. (1979). Physical and geotechnical properties of soils,McGraw-Hill, New Jersey.

Clare, K. E., and Sherwood, P. T. (1954). “Effect of organic matter on set-ting of soil-cement mixtures.” J. Appl. Chem., 4(11), 625–630.

Eades, J. L., and Grim, R. E. (1966). “A quick test to determine limerequirements for lime stabilization.” Highway Research Record 139,Highway Research Board, Washington, DC, 61–72.

Edil, T. B. (1997). “Construction over peats and organic soils.” Proc.,Conf on Recent Advances in Soft Soil Engineering, Vol. 1, Kuching,Malaysia, 85–108.

Edil, T. B., Acosta, A. A., and Benson, C. H. (2006). “Stabilizing soft fine-grained soils with fly ash.” J. Mater. Civ. Eng., 18(2), 283–294.

Federal Highway Administration (2003). “Fly ash facts for highway engi-neers.” Technical Rep. FHWA-IF-03019, 4th Ed., Washington, DC.



http://dx.doi.org/10.1680/geng.2004.157.4.239

http://dx.doi.org/10.1680/geng.2004.157.4.239

http://dx.doi.org/10.1002/jctb.5010041107

http://dx.doi.org/10.1061/(ASCE)0899-1561(2006)18:2(283)

Ferguson, G. (1993). “Use of self-cementing fly ashes as a soil stabilizationagent.” Fly ash for soil improvement (GSP 36), ASCE, New York.

Hampton, M. B., and Edil, T. B. (1998). “Strength gain of organic groundwith cement-type binders.” Soil improvement for big digs (GSP 81)ASCE, Reston, VA, 135–148.

Janz, M., and Johansson, S. E. (2002). “The function of different bindingagents in deep stabilization.” Rep. 9, Swedish Deep StabilizationResearch Centre, Linköping, Sweden.

Kaniraj, S. R., and Havanagi, V. G. (1999). “Compressive strength ofcement stabilized fly ash-soil mixtures.” Cem. Concr. Res., 29,673–677.

Keshawarz, M. S., and Dutta, U. (1993). “Stabilization of south Texas soilswith fly ash.” Fly ash for soil improvement (GSP 36), ASCE, New York,30–42.

Khoury, N. N., and Zaman, M. M. (2004). “Correlation between resilientmodulus, moisture variation, and soil suction for subgrade soils.” Trans-portation Research Record 1874, Transportation Research Board,Washington, DC, 99–107.

Kitazume, M. (2005). “State of practice report—Field and laboratory in-vestigations, properties of binders and stabilized soil.” Proc., Int. Conf.on Deep Mixing Best Practice and Recent Advances, Swedish DeepStabilization Research Centre, Linköping, Sweden

Odadjima, H., Noto, S., Nishikawa, J., and Yamazaki, T. (1995). “Cementstabilization of peaty ground with consideration of organic matters.”Proc., Int. Workshop on Engineering Characteristics and Behaviorof Peat, Sapporo, Japan.

Parsons, R. L., and Kneebone, E. (2005). “Field performance of fly ashstabilized subgrades.” Ground Improv., 9(1), 33–38.

Prabakar, J., Dendorkar, N., and Morchale, R. K. (2004). “Influence of flyash on strength behavior of typical soils.” Constr. Build. Mater., 18(4),263–267.

Rogers, C. D. F., and Glendinning, S. (2000). “Lime requirement forstabilization.” Transportation Research Record 1721, TransportationResearch Board, Washington, DC, 9–18.

Sakr, M. A., Shahin, M. A., and Metwally, Y. M. (2009). “Utilization oflime for stabilizing soft clay soil of high organic content.” Geotech.Geol. Eng., 27(1), 105–113.

Seed, H. B., and Chan, C. K. (1959). “Structure and strength characteristicsof compacted clays.” J. Soil Mech. and Found. Div., 85(5), 83–108.

Sridharan, A., Prashanth, J. P., and Sivapullaiah, P. V. (1997). “Effect of flyash on the unconfined compressive strength of black cotton soil.” Proc.ICE Ground Improve., 1(3), 169–175.

Tremblay, H., Duchesne, J., Locat, J., and Lerouil, S. (2002). “Influence ofthe nature of organic compounds on fine soil stabilization with cement.”Can. Geotech. J., 39(3), 535–546.

Trzebiatowski, B. D., Edil, T. B., and Benson, C. H. (2005). “Case studyof subgrade stabilization using fly ash: State Highway 32, PortWashington, Wisconsin.” Recycled materials in geotechnics (GSP127), ASCE, Reston, VA, 123–136.

Young, J. F. (1972). “A review of the mechanisms of set-retardation inportland cement pastes containing organic admixtures.” Cem. Concr.Res., 2, 415–433.



http://dx.doi.org/10.1016/S0008-8846(99)00018-6

http://dx.doi.org/10.1016/S0008-8846(99)00018-6

http://dx.doi.org/10.1680/grim.2005.9.1.33

http://dx.doi.org/10.1016/j.conbuildmat.2003.11.003

http://dx.doi.org/10.1016/j.conbuildmat.2003.11.003

http://dx.doi.org/10.1007/s10706-008-9215-2

http://dx.doi.org/10.1007/s10706-008-9215-2



http://dx.doi.org/10.1139/t02-002

http://dx.doi.org/10.1016/0008-8846(72)90057-9

http://dx.doi.org/10.1016/0008-8846(72)90057-9

Documents

Stabilization of Organic Soils with Fly Ash · Stabilization of Organic Soils with Fly Ash ... Soil organic content is a detrimental ... organic matter instead of reacting with pozzolanic