Construction and Building Materials 29 (2012) 201–205

Contents lists available at SciVerse ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier .com/locate /conbui ldmat

Use of steel and polypropylene fibers to improve flexural performance of deepsoil–cement column

Piti Sukontasukkul ⇑, Pitthaya JamsawangDepartment of Civil Engineering, King Mongkut’s University of Technology-North Bangkok, 1518 Pibulsongkarm Road, Bangsue, Bangkok 10800, Thailand

a r t i c l e i n f o

Article history:Received 29 April 2011Received in revised form 13 September2011Accepted 2 October 2011Available online 24 November 2011

Keywords:Flexural performanceFiber reinforced soil cementDeep cement mixingPolypropyleneSteel fibers

0950-0618/$ - see front matter � 2011 Elsevier Ltd. Adoi:10.1016/j.conbuildmat.2011.10.040

⇑ Corresponding author. Tel.: +66 2 913 2500x8621E-mail addresses: piti@kmutnb.ac.th, piti_sk@hotm

a b s t r a c t

In this study, steel and polypropylene fibers are used to improve flexural performance of soil–cementpile. Deep cement mixing (DCM) technique has been used for decades in Thailand to improve the strengthof soft clay in Bangkok and vicinity. However, since it is made of about 10–20% of cement, it also sharessimilar properties as the harden cement such as good compressive strength, low permeability, and poortensile and flexural strengths (brittle). In several occasions when subjected to bending moment caused byhorizontal loadings (induced by large embankment), the poor flexural strength can lead to a failure of thesoil–cement column. In order to improve its flexural strength and brittleness, a technique of mixing shortfibers (similar to that used in conventional concrete) is introduced to the soil cement mixture. Two typesof fiber are used to produce the fiber reinforced soil cement (FRSC): polypropylene and steel fibers atthree different volume fractions of 0.5%, 0.75% and 1.0%. Flexural performance of the FRSC is carriedout according to ASTM C1609. Results show improvements in the flexural performance as seen by theincrease in the equivalent flexural strength ratio and the residual strength when the fibers are incorpo-rated into the mix. The polypropylene fiber is found to perform better than the steel fibers. With theincreasing volume fraction, the toughness is also increase.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction

The properties of Bangkok soft clay are low strength, high com-pressibility and low permeability. Several techniques have beenadopted to improve the properties of soft clay for example, in situearth reinforcement and piles of various materials, partial or com-plete ground replacement, geotextiles, preloading, grouting, deepmixing pile or sand pile; timber or concrete piles: micro piles,cast-in situ piles. These techniques have their own advantagesand disadvantages [1].

Deep mixing (DM) techniques, developed during 1960s, werefirst reported in the literature in the early 1970s. The deep mixingstabilizing process typically takes place by mechanical dry mixing,wet mixing or, grouting [2,3].

The deep cement mixing (DCM) was introduced in 1999 [2]. Inthis technique, the cement powder is added into the soil duringthe mixing process. The properties of DCM are somewhat superiorto the DM in both of strength and permeability. The DCM pile be-comes widely used to improve the engineering properties of thickdeposits of soft ground and effectively reduce settlements of full-scale embankments [4,5]. However, similar to the conventional con-crete, the properties of the DCM is also affected by the properties of

ll rights reserved.

25; fax: +66 2 587 4337.ail.com (P. Sukontasukkul).

cement in terms of brittleness and poor flexural (or tensile)strengths [6]. When subjected to vertical and horizontal loads in-duced by embankment, large settlements may occur due to horizon-tal movements [7]. In addition, horizontal loads can also cause thebending moment in DCM pile and cause the pile to fail in tension(due to its brittleness).

Although the DCM technique is effective in improving the com-pressive strength of the DM column, the improvement on the brit-tleness is still needed. This study is aimed primarily to improvethat property and increase its toughness by mean of adding shortfibers into the DCM mixture.

Short fibers have been used for centuries to reinforce brittlematerials like cement or masonry bricks. There are many fibretypes available nowadays for commercial use such as steel, glass,synthetic materials (polypropylene, carbon, nylon, etc.) and somenatural fibers. For convention fiber reinforced concrete (FRC), thetypical fibers volume fraction is in the range of 0.5–1.0%. At thepractical volume fraction used in SFRC (<1%), the increase in com-pressive, tensile, or flexural strength is small because the matrixcracks essentially at the same stress and strain as in plain concrete.The real advantage of adding fibers is that, after matrix cracking,fibers bridge these cracks and restrain them. In order to further de-flect the beam, additional forces and energies are required to pullout or fracture the fibers. This process, apart from preserving theintegrity of concrete, improves the load-carrying capacity beyondcracking [8–17].

Fig. 1. Specimens wrapped in plastic sheet.

Table 1Properties of Bangkok soft-clay.

Property Value

Specific gravity 2.68Liquid limit 95%Plastic limit 42%Saturated unit weight 15.6 kN/m3

Natural water content 74%Undrained shear strength 15.6 kPa

P

L/600 L/150

P600

P150

δ

Deflection

Loa

d

Fig. 2. Schematic illustration of load–deflection curve for calculating first-peakstrength ( f ) and residual strengths (ASTM C1609).

202 P. Sukontasukkul, P. Jamsawang / Construction and Building Materials 29 (2012) 201–205

In this study, the fiber reinforced soil cement (FRSC) is preparedusing 2 different types of fibers (steel and polypropylene) at threedifferent volume fractions: 0.5%, 0.75% and 1.0%. The experimentbegins with determining the optimum moisture and cement con-tents to meet the unconfined compressive strength requirement.After that the specimens are cast in form of beams to test underthird point loading condition. Results in terms of peak strength,residual strength and area under load–deflection curve are evalu-ated and discussed.

2. Experimental procedure

2.1. Materials

Soil: Bangkok soft clay extracted from the depth between 5 and 8 m (Fig. 1,Table 1).

Table 2Properties of Fiber.

Materials Specific gravity Shape Length (mm)

PP 0.91 Fully crimped 58Steel-s 7.8 Hooked end 35Steel-l 7.8 Hooked end 60

Table 3Specimen designation and detail.

Designation Moisture content (%) Cement content (%)

P-SC 100 20

SFSC-L0.5 100 20SFSC-L0.75 100 20SFSC-L1.0 100 20

SFSC-S0.5 100 20SFSC-S0.75 100 20SFSC-S1.0 100 20

PFSC-L0.5 100 20PFSC-L0.75 100 20PFSC-L1.0 100 20

Total number of specimens

Cement: Portland type IFibers: Steel and polypropylene fibers with properties as shown in Table 2

2.2. Determining optimum mix proportion

According the requirement of the Highway Department of Thailand (HDT), theminimum unconfined compressive strength of soil–cement should be at least 6 ksc(�0.6 MPa). The process of determining optimum mix proportion consists of (1)determining the moisture content of the soft clay, (2) remixing the soft clay withwater to obtain the total moisture contents of 70%, 100%, 130% and 160%, (3) mixingcement into the remixed-clay at 10%, 15% and 20% by weight of dried clay, (4) cast-ing the soil cement into cylindrical shaped specimens, and (4) testing the specimensat the age of 28 days under unconfined compression condition (Fig. 1). The mix withlowest cement and water content that pass minimum requirement of 0.6 MPa is se-lected as the optimum mix proportion (according to the test results, the optimummoisture and cement content for the controlled mix that pass the requirement isfound at 100% and 20%, respectively, see Section 3.1).

2.3. Specimen preparation for flexural performance test

After the optimum mix proportion is determined, the specimens are then pre-pared in form of cylinder with dia.-50 � 100 mm and beam with 100 � 100 �350 mm. For plain soil–cement, all ingredients (cement, clay and water) are mixedin a mixer for 5 min and cast. For fiber reinforced soil cement, fibers are added intothe fresh soil–cement and mixing is continued for another 5 min until the fibers aredistributed thoroughly. The specimens are demolded after 24 h and wrapped withplastic sheet for 28 days prior the test date (Fig. 1). Details on mix proportion andnumber of specimen are given in Table 3.

Section (mm) Aspect ratio (l/d) Tensile strength (N/mm2)

Rect. 1.0 � 0.5 52 450Circle dia. �0.55 64 1100Circle dia. 0.90 67 1100

Fiber No. of specimen

Mat. Length (mm) Vol. frac. (%)

5

Steel 60 0.5 5Steel 60 0.75 5Steel 60 1 5

Steel 35 0.5 5Steel 35 0.75 5Steel 35 1 5

PP 58 0.5 5PP 58 0.75 5PP 58 1 5

50

(a) (b)

(c) (d)

Fig. 3. Unconfined compression response of soil–cement with moisture content of(a) 70%, (b) 100%, (c) 130% and (d) 160%.

Fig. 4. Unconfined compression strength of soil–cement with 70–160% moisturecontent and 10–20% cement content.

(a)

(b)

(c)

Fig. 5. Flexural response of (a) short steel FRSC (b) long steel FRSC and (c)polypropylene FRSC.

P. Sukontasukkul, P. Jamsawang / Construction and Building Materials 29 (2012) 201–205 203

2.4. Experiments

After curing for 28 days, the specimens are subjected to flexural performancetest (ASTM C1609). Results from the load–deflection curves are used for calculatingvalues such as the first-peak strength ( f ) (Eq. (1)), residual strengths at deflection ofL/600 ( f D

600) and L/150 ( f D150) (Fig. 2), area under the load–deflection curve from

deflection of 0 to L/150 (TD150), and equivalent flexural ratio (Eq. (2)).

First Peak Strength; f ¼ PL

bd2 ð1Þ

Equivalent Flexural Strength ratio; RDT;150 ¼

150 � TD150

f1bd2 100%: ð2Þ

3. Results and discussion

3.1. Optimum mix proportion

The optimum mix proportion is determined by testing soft claymixed with different proportions of cement and water contentsunder unconfined compression. The selected mix proportion musthave a minimum compressive strength of 0.6 MPa (according therequirement of Thailand Highway Department (THD)). Results as

(a)

204 P. Sukontasukkul, P. Jamsawang / Construction and Building Materials 29 (2012) 201–205

shown in Figs. 3 and 4 indicate that there are three mixes that passthe compressive strength requirement which are 70–15, 70–20and 100–20. Even though those with 70% moisture content seemedto exhibit higher strength than the one with 100% moisture con-tent, they are not chosen as a controlled mix proportion becauseof the dryness and the difficulty to mix uniformly with fibers. Thus,the mixture with 100% moisture content and 20% cement contentis selected as a controlled mix proportion.

3.2. Flexural performance

3.2.1. Load–deflection responseThe flexural responses of both PSC and FRSC are shown in Fig. 5.

For PSC, the load is found to increase linearly with the increasingdeflection from the beginning up to the peak load, after that asharp drop load is observed. This sharp drop indicates the brittle-ness of the PSC. In the case FRSC, the pre-peak response is similarto that of PSC. Unlike conventional concrete, for the PSC and FRSC,the non-linearity portion due to the toughening mechanism is notfound because of the lack of aggregate. Prior to the peak, the cracksbegin to initiated and quickly become localized at the peak. Thiscauses the matrix to fail. As the fibers begin to take over, the loadstarts to increase again.

It is the post-peak response that really differentiated the PSCfrom the FRSC. For PSC, once the strain energy is high enough to

Table 4Flexural performance.

Name Peakload(kN)

Deflection

L/600 L/150

PD600

(kN)f D600

(MPa)PD

150

(kN)f D150

(MPa)TD

150

(Nm)RD

T;150

PSC 0.53 0.53a 0.18a – – 0.07 6.8

PFSC-L0.5 0.58 0.41 0.13 0.39 0.13 0.81 67.5PFSC-L0.75 0.60 0.59 0.18 0.49 0.15 1.00 73.4PFSC-L1.0 0.62 0.59 0.18 0.66 0.20 1.24 93.2

SFSC-S0.5 0.45 0.34 0.13 0.22 0.08 0.54 46.4SFSC-S0.75 0.47 0.42 0.15 0.22 0.09 0.57 47.3SFSC-S1.0 0.53 0.53 0.19 0.39 0.14 0.84 67.3

SFSC-L0.5 0.60 0.18 0.06 0.12 0.04 0.37 39.5SFSC-L0.75 0.68 0.22 0.08 0.16 0.06 0.41 43.4SFSC-L1.0 0.66 0.38 0.14 0.32 0.11 0.68 64.3

a For plain concrete, value of first peak load is used for representing the load andthe strength at L/600.

Fig. 6. Relationship between fiber volume fraction and equivalent flexural strengthratio.

(b)

(c)

Fig. 7. Residual strengths at different deflections of (a) 0.5% FRC, (b) 0.75% FRC and(c) 1.0% FRC.

cause the crack to self-propagate, fracture occurs almost instanta-neously. For FRSC, the fiber bridging effect helps to control the rateof energy release. Thus, FRSC remains its ability to carry load afterthe peak (residual load). However, this also depends on the typeand content of fibers. In this case, the polypropylene FRSC performsbetter than the steel FRSC.

Comparing between short and long steel fiber, though both fi-bers have similar aspect ratios, they do not show the same perfor-mance. In general, long fiber would tend to perform slightly lowerthan short fiber. This is due to the fact that in the mixing and cast-ing process, long fibers are more difficult to achieve the uniformity.This causes the samples with long fibers to have slightly higherporosity and lower strength.

P. Sukontasukkul, P. Jamsawang / Construction and Building Materials 29 (2012) 201–205 205

3.2.2. Flexural performanceFlexural performance in terms of first-peak strength ( f ), residual

strengths at deflection of L/600 ( f D600) and L/150 ( f D

150), area underthe load–deflection curve from deflection of 0 to L/150 (TD

150), andequivalent flexural ratio are given in Table 4.

First, let consider the equivalent flexural ratio (RDT;150) which

represents the area under the load–deflection curve (or energyabsorption) per strength and volume, all FSC appear to have higherequivalent flexural ratios than plain soil cement. This shows theeffectiveness of the fibers to bridge across cracks and to enhancethe energy absorption ability of concrete under loading. Whencomparing between all types of fiber, the PFSC is found to performbetter than both SFSCs (Fig. 6). This can be explained by looking atthe bond between the fiber and the matrix. When a fiber is beingpulled out of the matrix, the bond strength comes from two parts,an interfacial bond and a frictional (or anchorage) bond. The inter-facial bonding depends strongly on the material properties whichin this case is the strength of the matrix (assuming that there isno chemical bond between fiber and matrix). The frictional bonddepends mainly on the shape of the fiber. Since the matrix strengthof soil cement is very small, thus the interfacial bond is quite neg-ligible. Therefore the main contributor is the fiber shape. And, sincethe shape of the polypropylene fiber is fully crimped, it is able toprovide better frictional bond than the steel fibers [18,19].

In the case of residual strengths at L/600 and L/150, these valuesrepresent the ability of fiber reinforced concrete to sustain loadafter the first crack at different specific deflection (Fig. 7). By look-ing at the results, all FRSC are able to maintain their load carryingcapacity after the first crack (peak load). In most cases, the residualstrengths decrease after the first crack. The highest decreases arefound mostly in the long steel fiber soil cement. This may be dueto the length of the fiber which causes the difficulty in mixingand in obtaining uniformity.

4. Conclusions

1. By mixing fibers into the soil–cement, the flexural performanceis enhanced as seen by the the long descending load–deflectionresponses after the peak, the increase in the equivalent strengthratios and the residual strengths.

2. The values of the equivalent strength ratio indicate the energyabsorption ability of FRSC. In this study, these values are foundto depend on the type and the fiber content. With the increasingvolume fraction, the equivalent strength ratios are alsoincreased. The polypropylene fiber performs better than thesteel fiber due to its highly deformed shape.

3. The fibers increase the load carrying capacity of the soil–cementafter the first crack as seen by the value of the residual strengthat both L/600 and L/150. In most cases, residual strengths are

found to decrease with the increasing deflection, though at dif-ferent degree depending the type and the fiber content.

Acknowledgements

The authors would like to thank S.R. Fibre Co., Ltd. for providingstudent scholarship for a graduate student. Thanks are also tosenior students for their assistances during the experiments andMr. David Simpson for proofreading this manuscript.

References

[1] Lin KQ, Wong IH. Use of deep mixing to reduce settlement at bridgeapproaches. J Geotech Geoenviron Eng, ASCE 1999;125(4):309–20.

[2] Porbaha A. State of the art in deep mixing technology, part I: basic conceptsand overview. Ground Improv 1998;2:81–92.

[3] Kamon M. Effect of grouting and DMM on big construction projects in Japanand the 1995 Hyogoken-Nambu Earthquake, grouting and deep mixing. In:Proceedings of IS-Tokyo’96, 2nd international conference on groundimprovement geosystems, Tokyo, 14–17 May 1996, p. 807–23.

[4] Bergado DT, Ruenkrairergsa T, Taesiri Y, Balasubramaniam AS. Deep soilmixing to reduce embankment settlement. Ground Improve J 1999;3(3):1–18.

[5] Lai YP, Bergado DT, Lorenzo GA, Duangchan T. Full-scale reinforcedembankment on deep jet mixing improved ground. Ground Improve J2006;10(4):153–64.

[6] Dong P, Qin R, Chen Z. Bearing capacity and settlement of concrete-cord DCMpile in soft ground. Geotech Geolog Eng 2004;22(1):105–19.

[7] Wu M, Zhao X, Dou YM. Application of stiffened deep cement mixed column inground improvement. In: Proceedings international conference on deep mixingbest practices and recent advances, Stockholm, Sweden, 2005. p. 463–8.

[8] Romualdi JP, Batson GB. Mechanics of crack arrest in concrete. In: Proceedings,ASCE, vol. 89, EM 3, June 1963. p. 147–68.

[9] Romualdi JP, Mandel JA. Tensile strength of concrete affected by uniformlydistributed closely spaced short lengths of wire reinforcement. ACI J Proc1964;61(6):657–71.

[10] Shah SP, Rangan BV. Fiber reinforced concrete properties. ACI J Proc1971;68(2):126–35.

[11] Johnston CD. Steel fibre reinforced mortar and concrete – a review ofmechanical properties. Fiber reinforced concrete. Detroit: American ConcreteInstitute; 1974. p. 124–42.

[12] Johnston CD, Gray RJ. Uniaxial tension testing of steel fibre reinforcedcementitious composites. In: Proceedings, international symposium ontesting and test methods of fibre-cement composites, RILEM, Sheffield, April1978. p. 451–61.

[13] Bentur A, Mindess S. Concrete beams reinforced with conventional steel barsand steel fibres: properties in static loading. Int J Cem Compos LightweightConcr 1983;5(3):199–202.

[14] Shah S. Concrete and fibers reinforced concrete subjected to impact loading,cement-based composites: strain rate effects on fracture. Mater Res Soc SympProc 1986;64.

[15] Rossi P. Mechanical behaviour of metal–fibre reinforced concretes. Cem ConcrCompos 1992;14(1):3–16.

[16] Mindess S. Fibre reinforced concrete: challenge and prospects. Fiber ReinfConcr: Mod Dev 1995:1–12.

[17] Sukontasukkul P, Pomchiengpin W. Post-crack (or post-peak) flexuralresponse and toughness of fiber reinforced concrete after exposure to hightemperature. Constr Build Mater (JCBM) 2010;24:1967–74.

[18] Bentur A, Mindess S. Fibre reinforced cementitious composites. Elsevier; 1990.[19] Burakiweicz A. In: Swamy RN, editor. Testing of fibre bond strength in cement

matrix, in testing and test methods of fiber cement composites. Lancaster: TheConstruction Press, Ltd.; 1978.