lable at ScienceDirect

Journal of Cleaner Production xxx (2015) 1e9

Contents lists avai

Journal of Cleaner Production

journal homepage: www.elsevier .com/locate/ jc lepro

Performance evaluation of palm oil clinker as coarse aggregate in highstrength lightweight concrete

Rasel Ahmmad a, Mohd Zamin Jumaat a, *, U. Johnson Alengaram a, **, Syamsul Bahri a,Muhammad Abdur Rehman b, Huzaifa bin Hashim a

a Department of Civil Engineering, Faculty of Engineering, University of Malaya, Kuala Lumpur, 50603, Malaysiab Department of Geology, Faculty of Science, University of Malaya, Kuala Lumpur, 50603, Malaysia

a r t i c l e i n f o

Article history:Received 9 January 2015Received in revised form24 July 2015Accepted 12 August 2015Available online xxx

Keywords:Oil palm shellPalm oil clinkerLightweight aggregateHigh strength lightweight concreteModulus of elasticity

* Corresponding author. Tel.: þ60 379675203.** Corresponding author. Tel.: þ60 379677632.

E-mail addresses: zamin@um.edu.my (M.Z. Ju(U.J. Alengaram).

http://dx.doi.org/10.1016/j.jclepro.2015.08.0430959-6526/© 2015 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Ahmmad, Rconcrete, Journal of Cleaner Production (201

a b s t r a c t

The enhancement in the mechanical properties of high strength lightweight concrete (HSLWC) utilizingpalm oil clinker (POC) as a replacement for oil palm shell (OPS) as lightweight coarse aggregate has beeninvestigated and reported. A series of concrete mixes was prepared with 25%, 50%, 75% and 100%replacement of coarse aggregate by POC in HSLWC, while setting other parameters as constant. Theparameters investigated include slump value, compressive strength, stress-strain behaviour, modulus ofelasticity and its normalization, ultrasonic pulse velocity (UPV) and failure modes. The results showedthat the replacement of OPS by POC as coarse aggregate has significant positive impact on compressivestrength, modulus of elasticity and UPV. The highest compressive strength of about 63 MPa obtained forthe mix with POC was about 43% higher than the control mix. Moreover, the enhancement in modulus ofelasticity up to 2.5 times could significantly control the deflection.

© 2015 Elsevier Ltd. All rights reserved.

1. Introduction

High strength and lightweight concretes are commonly desiredfor different structural applications including slabs and joists inhigh rise buildings, offshore and marine structures, and bridgedecks in highway bridge structures (Malier, 1992). In the case ofconstruction using conventional concrete, self-weight represents asignificant proportion of the total load of the structure; however, areduction in the density of the concrete would reduce the overalldead load to the foundation that could result in subsequentreduction in construction cost. Al-Khaiat and Haque (1998) re-ported that structural lightweight concrete has obvious advantagese higher strength/weight ratio, better tensile strain capacity, lowercoefficient of thermal expansion, and superior heat and soundinsulation e due to air voids in the lightweight aggregate. Inaddition, Topçu (1997) reported that the reduction in the dead loadof a construction by the use of a lightweight aggregate in concretecould result in a decrease in the cross section of columns, beams,

maat), johnson@um.edu.my

., et al., Performance evaluati5), http://dx.doi.org/10.1016/

slabs and foundations. Since the load due to earthquake is pro-portional to the mass of civil engineering structures and buildings,reducing the mass of the structure or building is of utmostimportance in the design of structural elements (Kılıç et al., 2003).In line with the design of structures using lightweight aggregate,high strength lightweight concrete (HSLWC) is one of the mostimportant research areas in concrete materials and structural en-gineering. A decrease in the density of the concrete for the samestrength category permits a saving in the dead load in the structuraldesign and foundation. For producing HSLWC, different types ofnatural and artificial materials can be used as coarse aggregateincluding expanded clay, Leca, Lytag, basaltic-pumice, etc.Furthermore, in recent times, agricultural waste, such as oil palmshell (OPS) (Mo et al., 2015b), palm oil clinker (POC) (Kanadasanand Razak, 2015) and coconut shell (Gunasekaran et al., 2013)have been used as the replacement aggregate instead of conven-tional crushed granite aggregate in the development of lightweightconcrete.

Malaysia is the second largest palm oil producing countries inthe world and hence the production of palm oil results in manywaste products. Thewastes from palm oil industries are OPS, emptyfruit bunches, palm oil fibres and palm oil fuel ash. Research workson the use of OPS as lightweight coarse aggregate in the develop-ment of lightweight aggregate concrete has been well established

on of palm oil clinker as coarse aggregate in high strength lightweightj.jclepro.2015.08.043

Fig. 1. Smooth (a) concave and (b) convex surface of waste OPS aggregate.

R. Ahmmad et al. / Journal of Cleaner Production xxx (2015) 1e92

since 1984 (Abdullah, 1984). However, POC is relatively a newsustainable material in the development of lightweight concrete. Itis an industrial waste material available produced mainly in palmoil industries in South East Asia. In order to run the generator toproduce electricity in palm oil industries, palm oil fibre and OPS areburnt at 850 �C. Generally, the proportion of palm oil fibre and oilpalm shell is 70: 30. POC (Shafigh et al., 2014) is available inboulders of sizes ranging from 100 to 300 mm. These boulders arethen crushed into suitable sizes as required in the development oflightweight concrete.

Shafigh et al. (2012) obtained a maximum compressive strengthof 53 MPa using OPS as the aggregate in lightweight concrete;however, the modulus of elasticity (MOE) obtained was only18.46 GPa, and the MOE/fcu ratio was only 348. The structuralbehaviour of OPS concrete (OPSC) has been analysed with respectto flexure by researchers (Alengaram et al., 2008); although OPSC isstructural lightweight concrete, in the case of doubly reinforcedbeams, it exceeds the code limit for deflection, as shown by Teoet al. (2006). The deflection of reinforced concrete flexural mem-bers depends on numerous factors, including the strength of theconcrete, the tensile reinforcement ratio balanced reinforcementratio (r/rb), the amount of longitudinal compressive reinforcementand the amount of lateral tie steel (Ahmad and Shah, 1985; Shinet al., 1989). Therefore, it is necessary to enhance the mechanicalproperties of OPS lightweight concrete as well as the MOE. Incontrast, using POC as the lightweight aggregate in concrete,Mohammed et al. (2011) showed that the ratio ofMOE/fcuwas in therange of 480e670 and with the increase in the concrete strength,the ratio improves. Also, the deflection of a POC concrete beam isless than the desired code limit, as shown by Mohammed et al.(2014). Consequently, POC aggregate improves the properties oflightweight concrete compared to the OPS lightweight concrete.

The aim of this study is to enhance the mechanical properties ofHSLWC incorporating waste POC for producing novel lightweightconcrete. The variable studied includes different percentages ofPOC as a replacement for OPS coarse aggregate. The mechanicalproperties investigated include the compressive strength, stress-strain behaviour, modulus of elasticity of HSLWC. Moreover, theinternal bonding property of concrete has been studied by the ul-trasonic pulse velocity test.

2. Materials used

As the study aims to enhance the mechanical properties ofhigh strength lightweight concrete, the effect of POC replace-ment for OPS as coarse aggregate in different proportions asindicated in the mixture proportion was investigated. Details ofthe materials used in the study and experimental works per-formed to achieve the objectives are given in the subsequentsections.

2.1. Cement

Type I 42.5 grade ordinary Portland cement (OPC), with a spe-cific gravity of 3.14 g/cm3, was used in the preparation of the testspecimens. The Blaine's specific surface area of the cement was3510 cm2/g. In lightweight concrete, this OPC acts as the bindermaterial and fills the pores of the coarse and fine aggregates.

2.2. Superplasticiser (SP)

A high range water-reducing admixture, Sika ViscoCrete®-2199,had been found to be suitable for the concrete mixes in the presentinvestigation. In accordance with BS 5075, the admixture is chlo-ride free and is compatible with all types of Portland cement,

Please cite this article in press as: Ahmmad, R., et al., Performance evaluaticoncrete, Journal of Cleaner Production (2015), http://dx.doi.org/10.1016/

including sulphate resistant cement. The use of SP content was keptconstant at 1.74% of the cement weight.

2.3. Aggregate

The fine aggregate comprises local mining sand with a specificgravity, fineness modulus, water absorption and maximum grainsize of 2.66, 2.89, 1.17% and 4.75 mm, respectively. Moreover, OPSand POC coarse aggregates are considered as a renewable source ofaggregate from waste materials. OPS, which, according to theexisting literature (Alengaram et al., 2011a,b), comes in differentshapes with a smooth convex and concave surfaces, was used as thecoarse aggregate, as shown in Fig. 1.

The OPS collected from a local crude palm oil producingmill waswashed to remove oil coating and then dried. The dry OPS wascrushed using stone crusher and sieved using 4.75 mm and12.5 mm sieves. The particles between 4.75 mm and 12.5 mmwereconsidered to be suitable as coarse aggregate in the preparation ofconcrete. In contrast to the preparation of OPS, POC needs onlycrushing and similar size as that of the OPS. It should be noted thatthe OPS had low specific gravity of 1.22, but had high 24 h waterabsorption of 23.52%. Its fineness modulus and compacted bulkdensity were found as 6.23 and 683 kg/m3. The measured range ofshell thickness was between 0.62 and 6.08 mm, which was withinthe range as stated by Alengaram et al. (2013). The POC is the resultof burning OPS and fibre during incineration process and obtainedin boulders of sizes between 100 and 300 mm (Kanadasan andRazak, 2014). The crushed POC aggregate contains numerouspores and has a rough and undulated surface, as shown in Fig. 2.

The OPS was replaced with POC as coarse aggregate at 0%, 25%,50%, 75% and 100% by volume. Furthermore, POC had a specificgravity of 2.08 which is higher than that of OPS; similarly its fine-nessmodulus and compacted bulk density of 6.29 and 782.10 kg/m3

were also found higher compared to OPS. . In contrast, the lower24 hwater absorption of 3.56% obtained for POC aggregates is muchlower compared to very high water absorption of about 24% forOPS. Table 1 illustrates the chemical composition of the POC andOPS waste materials.

2.4. Mix proportions

Five concrete mixes were prepared to investigate the enhance-ment of the mechanical properties of HSLWC containing OPS andPOC as coarse aggregates. The cement content and water to cementratio for thosemixtures were kept constant at 466 kg/m3 and 0.353,correspondingly. A control mix with only OPS as coarse aggregatewas cast for comparison purpose. Since the control concrete con-tained no POC, this mix was designated as P0 (no partial replace-ment POC as coarse aggregate). The other four mixes contained POCcoarse aggregate by volume replacement of OPS coarse aggregate at25%, 50%, 75% and 100% is named as POC based HSLWC, which were

on of palm oil clinker as coarse aggregate in high strength lightweightj.jclepro.2015.08.043

Fig. 2. Irregular and rough surface of POC aggregate.

R. Ahmmad et al. / Journal of Cleaner Production xxx (2015) 1e9 3

designed based on the particle packing concept of concrete mixdesign (Kanadasan and Razak, 2014). The mixes are referred to asP25, P50, P75 and P100. The details of the concrete mixes withslump are presented in Table 2.

2.5. Specimen preparation

Concrete cube specimens of 100 mm in size were used for thedetermination of the ultrasonic pulse velocity (UPV), oven drydensity and compressive strength. A cylinder of 100 mm diameterand 200 mm length was used to determine the modulus of elas-ticity. After casting, these specimens were covered with plasticsheathing and kept in the moulds for 24 h at a room temperature of28 ± 2 �C. These specimenswere then cured inwater at 24± 2 �C for28-d. Three concrete specimens were prepared and tested for eachtest condition to obtain average values.

3. Results and discussion

In this work, for the improvement of the mechanical propertiesof lightweight concrete, the partial replacement of the OPS light-weight coarse aggregate with the POC as lightweight coarseaggregate was at 0%, 25%, 50%, 75% and 100% by volume. Table 2shows the mix proportions and slump values. Though the surfaceof POC is rougher compared to OPS, the slump values of all themixes show consistent values. It can be seen from the slump valuesthe required workability is achieved in the range of site appliedlightweight concrete (Mehta and Monteiro, 2006) for all the mixeswith the use of water to cement ratio of 0.353 and SP content of1.7% of cement weight. The results on improved hardened concreteproperties of HSLWC are discussed in the following sub-sections.

3.1. Compressive strength of hardened OPS and POC based HSLWC

The compressive strength of the concrete mixes in water curingup to 56-d is provided in Table 3. As shown in Table 3, the increasein the POC content enhanced the compressive strength of theconcrete. The gradual replacement of OPS with POC by 25%, 50%,75% and 100% is reflected in the 28-d compressive strength as theseincrements enhanced in the sequence of 14.7%, 19.8%, 28.5% and43.1%, compared to the strength of the control mix P0. Moreover,the highest 28-d cube compressive strength of 61.67 MPa for themix P100 with a 100% replacement of OPS by POC in contrast tomixP0, which shows the effect of POC as coarse aggregate.

Table 1Chemical composition of POC and OPS waste materials (Ahmmad et al., 2014).

Oxides SiO2 K2O CaO P2O5 MgO Fe2O

POC 59.63 11.66 8.16 5.37 5.01 4.62OPS 46.61 9.88 14.76 1.95 2.91 10.19

Please cite this article in press as: Ahmmad, R., et al., Performance evaluaticoncrete, Journal of Cleaner Production (2015), http://dx.doi.org/10.1016/

The strength capacity of lightweight aggregate concrete (LWAC)is based on strength of the lightweight aggregate (LWA) used andthe hardened mortar, in addition to the mortar bonding in theinterfacial zone of aggregate (Lo et al., 2007). According to Okafor(1988), the OPSC failure at 28-d was dependent upon the failurebetween the aggregate and mortar bonding. In OPS, the weak bondbetween the smooth concave and convex surfaces is one of themain drawbacks. However, in the case of POC, the effect of the bondbetween the matrix and the POC plays a significant role in theenhancement of the strength.

The compressive strength of 55.37MPa obtained for themix P75in this investigation is higher than the previously reported for OPSbased lightweight concrete (Alengaram et al., 2013; Shafigh et al.,2011). Further, the cement content of 466 kg/m3 used in thisinvestigation is lower than some of the previous study (Shafighet al., 2012), and the 28-d compressive strength was 62 MPa,which was higher than all the previous studies for OPS based highstrength lightweight concrete. Shafigh et al. (2012), reported forOPSC the highest compressive strength of 53 MPa using maximumcement content in the range of 540e550 kg/m3. However, it couldbe seen from this investigation, the cement content of 466 kg/m3

for the mix P100 was about 14e16% lower compared to previousstudies, but it produced 19% higher compressive strength thanOPSC. This implies that compared to previous studies, the use ofPOC as the aggregate replacement in concrete could significantlylead to reduction in the cost of the high strength lightweight con-crete due to lower cement content.

Generally, the coarse aggregate is the weakest constituentmaterial in HSLWC due to the weaker bond between the LWA andthe mortar. The LWA controls the ultimate load carrying capacityof the concrete. However, researchers (Basri et al., 1999; Mannanet al., 2006) opined that in the OPS based concrete, the weakerinterfacial bond is due to the smooth concave and convex sur-faces of the OPS aggregate. In the case of OPS concrete, the failureoccurs due to the bond between the aggregate surface and mortarand hence the strength depends on the bonding strength of theaggregate and mortar. Fig. 3 shows the failure of OPSC containingOPS as the aggregate at the age of 28-d. The failure occurs due tothe crushing of the aggregate and the simultaneous debonding atthe interface between the mortar and aggregate surface. Thoughbreaking of OPS is evident in the failure mode, most of thefailure is seen due to the debonding of the mortar and aggregatesurface that causes the lower strength in concrete containingOPS.

In contrast to failure of OPS in OPSC, the porous surface textureof POC aggregates improves the bonding of OPSC as the mortar fillsinto the pores and strengthens the bond with the aggregate issimilar to the recycled lightweight aggregate concrete (Bogas et al.,2015). Therefore, the failure of the OPSC containing the POCaggregate only occurs due to the crushing of the aggregate in theconcrete, as shown in Fig. 4. However, in the case of POC basedconcrete, the load carrying mechanism is enhanced by the sus-tained action of POC due to adhering mortar that transfers the loadto POC. Instead, in the case of the OPS aggregate, the load carryingmechanism depends on the bond between the surface of the OPSand the mortar. Therefore, it causes an improvement in thecompressive strength of POC based HSLWCwith the increase in POCaggregate.

3 Al2O3 SO3 Na2O TiO2 Cr2O3 Others

3.7 0.73 0.32 0.22 e 0.583.33 7.84 1.15 e 1.38 e

on of palm oil clinker as coarse aggregate in high strength lightweightj.jclepro.2015.08.043

Table 2Mix proportion (kg/m3) and basic property of concrete.

Mix Cement Water W/C SP Fine aggregate Coarse aggregate Slump (mm)

Normal sand OPS POC

P0 (OPS) 466 166 0.353 8.1 847 373 e 55P25 (POC) 466 166 0.353 8.1 847 280.5 143.1 50P50 (POC) 466 166 0.353 8.1 847 187 286.2 55P75 (POC) 466 166 0.353 8.1 847 93.5 429.3 51P100 (POC) 466 166 0.353 8.1 847 e 565 53

Table 3Compressive strength of OPS and POC based HSLWC at different ages.

Mix Cube compressive strength fcu at different ages

1-d 3-d 7-d 28-d 56-d

P0 28.37 37.44 42.11 43.09 43.52P25 29.35 39.35 45.37 49.45 49.73P50 31.60 45.11 48.07 51.62 51.47P75 32.10 47.63 51.76 55.37 56.82P100 41.85 51.64 53.86 61.67 63.20

R. Ahmmad et al. / Journal of Cleaner Production xxx (2015) 1e94

In this research, OPSC was prepared using a combination ofcrushed OPS and POC as the aggregate in different mix proportions;hence, the correlation between the initial age (1-, 3-, 7-d) strengthand 28-d strength, as shown in Fig. 5 is important. Generally, thelinear correlation gives high values of coefficient between theinitial and 28-d compressive strength. All the mixes produced87e97% of their 28-d strength at the age of 7-d and similar corre-lation of 82e90% was reported (Lo et al., 2004) for LWAC usingexpanded clay as the coarse aggregate. On the other hand, Wilsonand Malhotra (1988) reported lower correlation of 76e87% forhigh strength lightweight concrete using expanded shale aggre-gate. The ratio observed for 1- and 3-d compressive strength to 28-d compressive strength, was 57e67% and 79e87%, respectively. Theequations showing the relationship between 1-, 3-, and 7-, and 28-d compressive strengths are given below:

�f 0cu

�28 ¼ 1:16

�f 0cu

�1 þ 14:2 (1)

�f 0cu

�28 ¼ 1:13

�f 0cu

�3 þ 2:06 (2)

�f 0cu

�28 ¼ 1:43

�f 0cu

�7 � 16:62 (3)

where ðf 0cuÞ1, ðf 0cuÞ3, ðf 0cuÞ7 and ðf 0cuÞ28 are the 1-, 3-, 7- and 28-dcompressive strength for OPS concrete, correspondingly.

Fig. 3. Failure pattern of co

Please cite this article in press as: Ahmmad, R., et al., Performance evaluaticoncrete, Journal of Cleaner Production (2015), http://dx.doi.org/10.1016/

3.2. Strength to density ratio

The additional benefit of using crushed POC aggregate asreplacement to conventional crushed granite aggregates wouldenable lower dead load in structural elements. Another aspect ofthe oven dry density (ODD) is the strength to density ratio (Table 4)of the POC based concrete; this ratio for POC based concrete wasfound in the range of 27e31 � 10�3 MPa/kg/m�3 and this is highercompared to previously published values of 17e22 � 10�3 MPa/kg/m�3 for OPS based concrete (Mo et al., 2015a). In addition, thecomparison between POC based concrete and conventionalcrushed granite concrete shows that the former has superior per-formance. Conventional concrete has the lowest value in the rangeof 14e19� 10�3 MPa/kg/m�3 (Mo et al., 2015a) due to high density.However, the reduction in the density of POC based concrete showsabout 63% higher strength to density ratio compared to conven-tional concrete. Though both the mixes P0 and P100 fall underlightweight category, the later mix shows superior ratio comparedto the former and this could be attributed to the stronger bondbetween the aggregate and matrix that enabled the concrete toproduce higher strength. Also, the POC based mixes (P25eP100)show higher ratio of about 40% compared to the published strengthto density ratio of 17e22 � 10�3 MPa/kg/m�3 for OPSC.

3.3. Strength increasing factor at 28-d

One of the strength characteristics of concrete is the aggregateproportion in concrete. This characteristic was analytically calcu-lated based on the proportion of POC in the mixes and the strengthenhancement is reported as factor. The strength increasing factor(SIF) is the ratio of the compressive strength of the mixes P25, P50,P75 and P100 containing different proportions of OPS and POCwiththe compressive strength of the control concrete mix P0 as given inEqn. (4).

SIF ¼ Strength of concreteStrength of control concrete

(4)

ncrete containing OPS.

on of palm oil clinker as coarse aggregate in high strength lightweightj.jclepro.2015.08.043

Fig. 4. Failure pattern of concrete containing POC as coarse aggregate.

Fig. 5. Relationship of short-term (1-, 3- and 7-d) compressive strength with long-term (28-d) compressive strength.

R. Ahmmad et al. / Journal of Cleaner Production xxx (2015) 1e9 5

Using Eqn. (4), the SIF for themixes obtainedwere 1.15,1.20,1.28and 1.43 for the POC replacement as the coarse aggregate by 25%,50%, 75% and 100%, respectively. SIF is used here to quantify theincrease in the compressive strength as higher the SIF greater thestrength of the concrete. The SIF increases gradually with the in-crease in the POC content in the concrete.

By regression analysis, experimental data were used to establisha valid relationship between the SIF and the consumption of POC, p,as shown in Eqn. (5) established from Fig. 6. Eqn. (5) shows therelationship between the SIF and replacement of OPS by POC in OPSand POC based HSLWC.

SIF ¼ 0:004pþ 1:0124 (5)

where, p is the percentage of OPS replacement by POC.

3.4. Stressestrain relationship of OPS and POC based HSLWC

Fig. 7 shows the stressestrain relationships for all the speci-mens of OPS and POC based HSLWC containing different

Table 4Strength to density ratio and modulus of elasticity (MOE) of OPS and POC basedHSLWC.

Mix no. fcu at 28-d Oven drydensity at28-d, r (kg/m3)

Strength todensity ratio ( f cu

r) at

28-d (10�3 MPa/kg/m�3)

MOE(GPa)

MOE/fcu

P0 43.09 1787 24.11 14.1 327.2P25 49.45 1833 26.98 16.4 331.6P50 51.62 1879 27.47 20.9 404.9P75 55.37 1925 28.76 27.6 498.5P100 61.67 1971 31.29 34.8 564.3

Please cite this article in press as: Ahmmad, R., et al., Performance evaluaticoncrete, Journal of Cleaner Production (2015), http://dx.doi.org/10.1016/

proportions of OPS and POC aggregates. The value of the strain ispresented with the increase in the compressive load. Domagała(2011) reported that for normal weight concrete (NWC) thelinear stress-strain curves stretches up to 30e45% of its ultimatestrength. Akbar (2008) showed that for NWC the strain at the peakstress was in the range of 0.0015e0.002. From the experimentalvalues, the strain at maximum stress in the range of0.00173e0.00401 was found higher than the NWC. It was foundthat the specimen of the mix P0 showed more cracks compared tothe other specimens. On contrary, the addition of more amount ofPOC as coarse aggregate reduced the number of cracks and thiscould be attributed to the brittle nature of the POC aggregate.Usually, LWAC tends to be more inelastic than the NWC, whichcauses sudden failure after the ultimate load (Balendran et al.,2002). The brittleness of the POC aggregate is one of the limita-tions of the lightweight concrete. However, the OPSC mixes con-taining 50% of OPS or more as the coarse aggregate showed ahigher strain rate than the NWC; and this could be advantageousto overcome the temperature and shrinkage cracks on the struc-tural element.

Fig. 7 also shows that the effectiveness of the ultimate loadcarrying capacity increased with the increase in the POC asreplacement for the OPS aggregate. It can be seen that an increasein the POC coarse aggregate content increased the ultimate stress aswell as the slope of the stress vs. strain curve. However, the ε valuedecreased due to the increase in the POC aggregate content. In thestress-strain diagram, it is seen that the strains at the failure pointof the P100 and P75 specimens were found very close to maximumstress. However, in the HSLWC with up to 50% of replacement ofOPS by POC aggregate, the failure occurred steadily as seen from theFig. 7.

Fig. 6. SIF with increase of POC content in concrete.

on of palm oil clinker as coarse aggregate in high strength lightweightj.jclepro.2015.08.043

Fig. 7. Stressestrain curves for OPS lightweight concrete.

R. Ahmmad et al. / Journal of Cleaner Production xxx (2015) 1e96

3.5. Modulus of elasticity (MOE) of OPS and POC based HSLWC

The modulus of elasticity and compressive strength of theHSLWC increased with the increase of the POC content as seen fromFig. 8. The highest MOE of about 34.8 GPa was obtained for the mixP100 and it is higher than previous studies (Gao et al., 1997; Kayaliet al., 2003) in HSLWC. Though MOE of structural LWC is in therange of 10e24 GPa, (FIP, 1983) the mix P100 with whole replace-ment of OPS with POC shows significant improvement and this isvital in the design of structural elements in serviceability limitstate. Furthermore, the values of MOE of P50, P75 and P100 werehigher compared to the previous study (Shafigh et al., 2012) of OPSbased HSLWC. Generally, stiffness of lightweight concrete is lesscompared to normal weight concrete (Zhang and Poon, 2015). Theuse of POC increased the stiffness of OPS based lightweight con-crete and resulted in lower deflection as observed by Alengaramet al. (2008).

The modulus of elasticity of the concrete could also be corre-lated with the compressive strength of the concrete (Alengaramet al., 2011a,b). The relationship between the MOE and the cubecompressive strength of HSLWC using POC aggregate is shown inFig. 9 and compared with the equation previously used to predictthe modulus of elasticity of lightweight concrete in CEP/FIP manual(Short and Kinniburgh, 1978) (Eqn. 6) and ACI code (ACI 318, 2008)(Eqn. 7).

Fig. 8. Relationship of MOE and compressive streng

Please cite this article in press as: Ahmmad, R., et al., Performance evaluaticoncrete, Journal of Cleaner Production (2015), http://dx.doi.org/10.1016/

ESðpreÞ ¼ ½r=2400�2xðfcuÞ13x9:1 (6)

where, Es(pre) (kN/mm2), fcu (N/mm2) and r (kg/m3) are the pre-dicted elastic modulus, cube compressive strength and air-drydensity.

ESðpreÞ ¼ 0:043W1:5c

ffiffiffiffiffif 0c

q(7)

where, Es(pre) (kN/mm2), f 0c (N/mm2) and r (kg/m3) are the predictedelastic modulus, cylinder compressive strength and air-dry density,respectively.

From Fig. 9, it is observed that CEB/FIP manual and ACI codeunderestimated the MOE of the POC based HSLWC in case of mixesP75 and P100 containing higher amount of POC. Also, these equa-tions overestimate MOE in the case of mixes P0 and P25 containinglower amount of POC. An equation is proposed to predict the MOEbased on the compressive strength from the experimental data andit is shown below:

ESðpreÞ ¼ 0:0005ðfcuÞ2:69 (8)

where, Es(pre) (kN/mm2), is the predicted elastic modulus, and fcu isthe cube compressive strength.

As known, OPS has smooth surface and low specific gravity (Teoet al., 2007) and hence it produces low MOE. However, theincreased stiffness (Guo et al., 2014) of POC along with more POCcontent, the stiffness of the concrete increases. This is evident in theMOE values as it increases from 16 to 34 GPa for whole replacementof OPS. The comparison of MOE between P0 and P100 shows themix with whole replacement of OPS has 140% increase compared tothe mix with OPS. Similar trend could be seen between the mixesP25 and P100, as the later shows 100% increase in theMOE. Anotherimportant aspect of material characteristics is the normalization ofthe material properties; thus, the ratio between the MOE and thecorresponding compressive strength shows (Table 4) the increasingtrend as the POC content increases. Based on the previously pub-lished data, the MOE/fcu ratio for high strength OPSC was found inthe range 329e360 (Yew et al., 2014). Due to superior MOE values,the conventional concrete shows higher MOE/fcu ratio of 390e450(Sarıdemir, 2013) compared to high strength OPSC. However, thePOC based mixes (P50, P75 and P100) with MOE/fcu ratio in the

th in HSLWC with the increase in POC content.

on of palm oil clinker as coarse aggregate in high strength lightweightj.jclepro.2015.08.043

Fig. 9. Relationship with MOE with cube compressive strength of OPS concrete.

R. Ahmmad et al. / Journal of Cleaner Production xxx (2015) 1e9 7

range of 400e560 is higher than the OPS based concrete. The MOE/fcu ratios of P75 and P100 show these mixes could structurallyperform similar to that of conventional concrete with lowerdeflection.

4. Non-destructive test

4.1. Ultrasonic pulse velocity (UPV) of OPS and POC based HSLWC

UPV is a non-destructive test that is conducted to evaluate theconsistency and quality of concrete to specify the existence ofpours, cracks and to determine the crack depth in the concrete.The UPV values of control mix along with POC based HSLWC isshown in Table 5. The concrete with UPV values between 3.66and 4.58 km/s is termed as ‘good’ (Leslie and Cheesman, 1949).The concrete within the above said range is considered as havingno large voids or cracks that reduce the structural integrity.Shafigh et al. (2011) showed the failure in OPSC arising at theinterfacial zone between the aggregate and mortar because of theleaky mortar at the interface. Therefore, UPV is the essentialmethod for detecting large pores in the aggregate-mortar inter-face. The UPV values for all the mixes investigated in thisresearch were found higher than 3.66 km/s from the age of 1-dand above.

The trend of increase in the UPV with the increasing age ofconcrete could be attributed to the bonding between the mortarand aggregate; this enables the concrete to enhance its strengthwith reduced pores in the solid skeleton of concrete. It is alsoimportant to note that the increase in the POC content resulted inenhanced UPV values. The rate of increase in the UPV values showsslower trend due to slow rate of hydration after 3-d.

In addition the above mentioned points, it is important todiscuss the correlation between short and long term UPV valuesand the relationship between the compressive strength and UPV.These are explained in the following sub-sections.

Table 5Ultrasonic pulse velocity (UPV) of OPS and POC based HSLWC.

Mix UPV at different ages in km/s

1-d 3-d 7-d 28-d 56-d

P0 3.68 3.95 4.10 4.13 4.21P25 3.78 4.02 4.22 4.25 4.34P50 3.81 4.13 4.27 4.32 4.41P75 3.87 4.20 4.32 4.44 4.48P100 4.02 4.31 4.36 4.51 4.58

Please cite this article in press as: Ahmmad, R., et al., Performance evaluaticoncrete, Journal of Cleaner Production (2015), http://dx.doi.org/10.1016/

4.1.1. Correlation between short term (1-, 3-, 7- d) and long term(28-d) UPV

The comparison between 1-, 3-d and 28-d UPV values show87e95% of UPV values was achieved at early ages. Further, all themixtures produced 96e99% of their 28-d UPV values at the age of7-d.

The correlation between the UPV values at the ages of 1-, 3-, and7-d and 28-d of the concrete mixes is shown in Fig. 10. As seen, allthe mixes show linear correlation between early and later age UPVvalues.

The following equations indicate linear relationship between 1-,3-, 7-d and 28-d UPV values and the 28-d UPV values can be pre-dicted using this equation:

ðUPVÞ28 ¼ 1:0746ðUPVÞ1:03721 (9)

ðUPVÞ28 ¼ 1:0478ðUPVÞ1:00173 (10)

ðUPVÞ28 ¼ 0:5384ðUPVÞ1:43937 (11)

where (UPV)1,(UPV)3,(UPV)7 and (UPV)28 are the 1-, 3-, 7- and 28-d UPV of OPS concrete, respectively.

4.1.2. Relationship between compressive strength and UPVAlthough there is no unique correlation with UPV and the

compressive strength of the concrete (Neville, 2005), the hard-ened mortar depends on the water/cement ratio and subsequentlythe UPV values is affected. When the voids are filled with water,the UPV travels faster than when the voids are filled with air(Neville, 2005), which indicates that the moisture also affects theUPV of the concrete. However, the strength of the concrete can beassessed using the UPV test. Moreover, with the change in thecoarse aggregate in the concrete, the strength property of theconcrete is changes. The relationship between the compressivestrength and UPV of the concrete containing various proportionsof OPS and POC coarse aggregate has been developed. Fig. 11shows the correlation between the cube compressive strengthand UPV of the OPSC mixes. From the figure, it is seen that therelationship between the compressive strength and UPV of OPSC isexponential. From the relationship, the following equation isdeveloped:

f 0c ¼ 0:1982ðUPVÞ3:79 (12)

Fig. 10. Correlation between initial age (1-, 3- and 7-d) and 28-d UPV.

on of palm oil clinker as coarse aggregate in high strength lightweightj.jclepro.2015.08.043

Fig. 11. Compressive Strength vs. UPV of OPS and POC base HSLWC.

R. Ahmmad et al. / Journal of Cleaner Production xxx (2015) 1e98

5. Conclusion

The effect of replacement of lightweight OPS aggregate withanother industrial waste material, POC at various percentages todevelop HSLWC was investigated. The properties investigatedinclude slump, ODD, compressive strength, stress-strain behaviour,MOE and UPV. Based on the investigation, the following conclu-sions can be drawn:

1. Generally the replacement of OPS with POC shows positive in-fluence on both the workability and the compressive strength.The enhancement in the compressive strength with thereplacement of 25%, 50%, 75% and 100% aggregate volume inOPSC by POC coarse aggregate shows 14.7%, 19.8%, 28.5% and43.1%, respectively higher than the control mix.

2. The strength to density ratio of POC based HSLWC is in the rangeof 27e31� 10�3 MPa/kg/m�3 and it is about 40% and 63% highercompared to OPS based lightweight concrete and normal weightconcrete, correspondingly. This increasing strength to densityratio indicates, POC based HSLWC reduce theweight of structuresignificantly compared to previous OPS based lightweight con-crete and normal weight concrete.

3. The failure in HSLWC containing OPS as the aggregate showstwo types of failure mode e debonding between the aggregate-mortar interface and the breaking of the aggregate. On thecontrary, the concrete containing POC fails due to the crushingof the aggregate.

4. The strain of OPS and POC based HSLWC at maximum stress inthe range of 0.00173e0.00401 was found higher than the NWC,which indicate it may have more shrinkage crack resisting ca-pacity compared to normal weight concrete.

5. The increase in the MOE of POC based HSLWC by about 2.5times compared to the control mix (P0) could be attributed tostronger bond between the mortar and aggregate. Thenormalization of MOE shows that an increase in the POC con-tent enhances the MOE/fcu values. The highest MOE of 34.8 GPaachieved for the mix with whole replacement of OPS with POC(P100) is about 66% higher than the mix with 50% partialreplacement of OPS with POC. Though MOE of structural LWC isin the range of 10e24 GPa, the mix P100 with whole replace-ment of OPS with POC shows significant improvement and thisis vital in the design of structural elements in serviceabilitylimit state.

6. The increment of POC in the mixes shows improvement in UPVvalues with the mix with whole replacement of OPS with POCshowing about 9% higher value than the control mix. Althoughthe increment in UPV shows only about 9%, the compressive

Please cite this article in press as: Ahmmad, R., et al., Performance evaluaticoncrete, Journal of Cleaner Production (2015), http://dx.doi.org/10.1016/

strength improvement 43% could be attributed to good inter-locking between POC and mortar.

7. The regression analysis between UPV and the compressivestrength shows good correlation between these two with the R2

value of 0.97.

Acknowledgement

The authors would like to thank the University of Malaya (UM)for funding through the High Impact Research Grant H-16001-00-D000036 to conduct the present research.

References

Abdullah, A., 1984. Basic strength properties of lightweight concrete using agri-cultural wastes as aggregates. In: Paper Presented at the Proceedings of Inter-national Conference on Low-cost Housing for Developing Countries, Roorkee,India.

ACI 318, 2008. Building Code Requirements for Structural Concrete (ACI 318-08) andCommentary. American Concrete InstituteInternational Organization forStandardization.

Ahmad, S.H., Shah, S.P., 1985. Structural properties of high strength concrete and itsimplications for precast prestressed concrete. PCI J. 30 (6), 92e119.

Ahmmad, R., Jumaat, M.Z., Bahri, S., Islam, A.B.M.S., 2014. Ductility performance oflightweight concrete element containing massive palm shell clinker. Constr.Build. Mater. 63, 234e241. http://dx.doi.org/10.1016/j.conbuildmat.2014.04.022.

Akbar, H., 2008. Design of Reinforced Concrete Structures. Basic Topics, vol. 1. SimayDanesh Publ, Iran.

Al-Khaiat, H., Haque, M., 1998. Effect of initial curing on early strength and physicalproperties of a lightweight concrete. Cem. Concr. Res. 28 (6), 859e866.

Alengaram, U.J., Jumaat, M.Z., Mahmud, H., Fayyadh, M.M., 2011a. Shear behaviourof reinforced palm kernel shell concrete beams. Constr. Build. Mater. 25 (6),2918e2927.

Alengaram, U.J., Jumaat, M.Z., Mahmud, H., 2008. Ductility behaviour of reinforcedpalm kernel shell concrete beams. Eur. J. Sci. Res. 23 (3), 406e420.

Alengaram, U.J., Mahmud, H., Jumaat, M.Z., 2011b. Enhancement and prediction ofmodulus of elasticity of palm kernel shell concrete. Mater. Des. 32 (4),2143e2148.

Alengaram, U.J., Muhit, B.A.A., Jumaat, M. Z. b, 2013. Utilization of oil palm kernelshell as lightweight aggregate in concrete e a review. Constr. Build. Mater. 38,161e172. http://dx.doi.org/10.1016/j.conbuildmat.2012.08.026.

Balendran, R., Zhou, F., Nadeem, A., Leung, A., 2002. Influence of steel fibres onstrength and ductility of normal and lightweight high strength concrete. Build.Environ. 37 (12), 1361e1367.

Basri, H., Mannan, M., Zain, M., 1999. Concrete using waste oil palm shells asaggregate. Cem. Concr. Res. 29 (4), 619e622.

Bogas, J.A., de Brito, J., Figueiredo, J.M., 2015. Mechanical characterization of con-crete produced with recycled lightweight expanded clay aggregate concrete.J. Clean. Prod. 89, 187e195. http://dx.doi.org/10.1016/j.jclepro.2014.11.015.

Domagała, L., 2011. Modification of properties of structural lightweight concretewith steel fibres. J. Civ. Eng. Manag. 17 (1), 36e44.

FIP, 1983. FIP Manual of Lightweight Aggregate Concrete. Surry University Press.Gao, J., Sun, W., Morino, K., 1997. Mechanical properties of steel fiber-reinforced,

high-strength, lightweight concrete. Cem. Concr. Compos. 19 (4), 307e313.Gunasekaran, K., Annadurai, R., Kumar, P.S., 2013. Plastic shrinkage and deflection

characteristics of coconut shell concrete slab. Constr. Build. Mater. 43, 203e207.http://dx.doi.org/10.1016/j.conbuildmat.2013.02.019.

Guo, Y.-c., Zhang, J.-h., Chen, G.-m., Xie, Z.-h., 2014. Compressive behaviour ofconcrete structures incorporating recycled concrete aggregates, rubber crumband reinforced with steel fibre, subjected to elevated temperatures. J. Clean.Prod. 72, 193e203. http://dx.doi.org/10.1016/j.jclepro.2014.02.036.

Kanadasan, J., Razak, H.A., 2015. Engineering and sustainability performance of self-compacting palm oil mill incinerated waste concrete. J. Clean. Prod. 89, 78e86.http://dx.doi.org/10.1016/j.jclepro.2014.11.002.

Kanadasan, J., Razak, H.A., 2014. Mix design for self-compacting palm oil clinkerconcrete based on particle packing. Mater. Des. 56, 9e19.

Kayali, O., Haque, M., Zhu, B., 2003. Some characteristics of high strength fiberreinforced lightweight aggregate concrete. Cem. Concr. Compos. 25 (2),207e213.

Kılıç, A., Atis, C.D., Yasar, E., €Ozcan, F., 2003. High-strength lightweight concretemade with scoria aggregate containing mineral admixtures. Cem. Concr. Res. 33(10), 1595e1599.

Leslie, J., Cheesman, W., 1949. An ultrasonic method of studying deterioration andcracking in concrete structures. J. Am. Concr. Inst. 21 (1), 17e36.

Lo, T., Cui, H., Li, Z., 2004. Influence of aggregate pre-wetting and fly ash on me-chanical properties of lightweight concrete. Waste Manag. 24 (4), 333e338.

Lo, T.Y., Tang, W., Cui, H., 2007. The effects of aggregate properties on lightweightconcrete. Build. Environ. 42 (8), 3025e3029.

Malier, Y., 1992. High Performance Concrete: from Material to Structure. CRC Press.

on of palm oil clinker as coarse aggregate in high strength lightweightj.jclepro.2015.08.043

R. Ahmmad et al. / Journal of Cleaner Production xxx (2015) 1e9 9

Mannan, M., Alexander, J., Ganapathy, C., Teo, D., 2006. Quality improvement of oilpalm shell (OPS) as coarse aggregate in lightweight concrete. Build. Environ. 41(9), 1239e1242.

Mehta, P.K., Monteiro, P.J., 2006. Concrete: Microstructure, Properties, and Materials.Mo, K.H., Alengaram, U.J., Jumaat, M.Z., 2015a. Utilization of ground granulated blast

furnace slag as partial cement replacement in lightweight oil palm shell con-crete. Mater. Struct. 1e12.

Mo, K.H., Alengaram, U.J., Jumaat, M.Z., Yap, S.P., 2015b. Feasibility study of highvolume slag as cement replacement for sustainable structural lightweight oilpalm shell concrete. J. Clean. Prod. 91, 297e304. http://dx.doi.org/10.1016/j.jclepro.2014.12.021.

Mohammed, B.S., Al-Ganad, M.A., Abdullahi, M., 2011. Analytical and experimentalstudies on composite slabs utilising palm oil clinker concrete. Constr. Build.Mater. 25 (8), 3550e3560. http://dx.doi.org/10.1016/j.conbuildmat.2011.03.048.

Mohammed, B.S., Foo, W., Abdullahi, M., 2014. Flexural strength of palm oil clinkerconcrete beams. Mater. Des. 53, 325e331.

Neville, A.M., 2005. Properties of Concrete, fourth ed. Pearson Education Limited,Harlow, UK.

Okafor, F.O., 1988. Palm kernel shell as a lightweight aggregate for concrete. Cem.Concr. Res. 18 (6), 901e910.

Sarıdemir, M., 2013. Effect of silica fume and ground pumice on compressivestrength and modulus of elasticity of high strength concrete. Constr. Build.Mater. 49, 484e489.

Shafigh, P., Jumaat, M.Z., Mahmud, H.B., Alengaram, U.J., 2011. A new method ofproducing high strength oil palm shell lightweight concrete. Mater. Des. 32 (10),4839e4843.

Please cite this article in press as: Ahmmad, R., et al., Performance evaluaticoncrete, Journal of Cleaner Production (2015), http://dx.doi.org/10.1016/

Shafigh, P., Mahmud, H.B., Jumaat, M.Z., 2012. Oil palm shell lightweight concrete asa ductile material. Mater. Des. 36, 650e654.

Shafigh, P., Mahmud, H.B., Jumaat, M.Z.B., Ahmmad, R., Bahri, S., 2014. Structurallightweight aggregate concrete using two types of waste from the palm oilindustry as aggregate. J. Clean. Prod. 80, 187e196. http://dx.doi.org/10.1016/j.jclepro.2014.05.051.

Shin, S.-W., Ghosh, S.K., Moreno, J., 1989. Flexural ductility of ultra-high-strengthconcrete members. ACI Struct. J. 86 (4).

Short, A., Kinniburgh, W., 1978. Lightweight Concrete. Applied Science Publishers.Teo, D., Mannan, M.A., Kurian, V., Ganapathy, C., 2007. Lightweight concrete made

from oil palm shell (OPS): structural bond and durability properties. Build.Environ. 42 (7), 2614e2621.

Teo, D.C., Mannan, M.A., Kurian, J.V., 2006. Flexural behaviour of reinforced light-weight concrete beams made with oil palm shell (OPS). J. Adv. Concr. Technol. 4(3), 459e468.

Topçu, I. l. B., 1997. Semi lightweight concretes produced by volcanic slags. Cem.Concr. Res. 27 (1), 15e21.

Wilson, H., Malhotra, V., 1988. Development of high strength lightweight concretefor structural applications. Int. J. Cem. Compos. Lightweight Concr. 10 (2),79e90.

Yew, M.K., Mahmud, H.B., Ang, B.C., Yew, M.C., 2014. Effects of heat treatment on oilpalm shell coarse aggregates for high strength lightweight concrete. Mater. Des.54, 702e707.

Zhang, B., Poon, C.S., 2015. Use of furnace bottom ash for producing lightweightaggregate concrete with thermal insulation properties. J. Clean. Prod. 99,94e100. http://dx.doi.org/10.1016/j.jclepro.2015.03.007.

on of palm oil clinker as coarse aggregate in high strength lightweightj.jclepro.2015.08.043