Materials and Design 56 (2014) 9–19

Contents lists available at ScienceDirect

Materials and Design

journal homepage: www.elsevier .com/locate /matdes

Technical Report

Mix design for self-compacting palm oil clinker concrete based onparticle packing

0261-3069/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.matdes.2013.10.086

⇑ Corresponding author. Tel./fax: +60 379675233.E-mail addresses: jegathish@siswa.um.edu.my (J. Kanadasan), hashim@um.

edu.my (H.A. Razak).

Jegathish Kanadasan, Hashim Abdul Razak ⇑StrucHMRS Group, Department of Civil Engineering, Faculty of Engineering, University of Malaya, 50603 Kuala Lumpur, Malaysia

a r t i c l e i n f o

Article history:Received 6 July 2013Accepted 30 October 2013Available online 7 November 2013

a b s t r a c t

The consumption of waste materials in self-compacting concrete (SCC) in the construction industry willnot only help to conserve the natural resources but also promote sustainability in preserving theenvironment. Palm oil clinker (POC) is a waste by-product from the incineration process of oil palm shellsand fibres. They are porous and lightweight in nature, which makes them suitable for use as a lightweightaggregate (LWA). In this study, a new procedure was employed to obtain the mix design based on theparticle packing (PP) concept to ensure the fresh and hardened properties of SCC are achieved. The actualpacking level of aggregate and paste volume is integrated into the proportioning method to obtain thefinal mix design. The proposed procedure was verified by evaluating the SCC formed for self-compacta-bility and mechanical properties. Based on the overall performance of fresh and hardened properties, itcan be deduced that the procedure satisfied the requirements for SCC. The satisfactory results indicatethat the mix design can be employed not only for POC but also for a variety of combinations of aggregate.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Self-compacting concrete (SCC) has progressively developed tobecome a concrete that is both efficient as well as providing a qual-ity product without comprising its durability. Reduced labourneeds, good surface finishing, extended workability and excellenthardened properties are some of the key elements of SCC. SCC isdesigned in a way that it will be able to flow under its own weightand produce good self compactability features without the need ofany external vibration. The study of the particle packing (PP) givesa good understanding concerning the consumption of aggregateand paste volume for a given unit volume of concrete. Hence, opti-misation of particle packing in SCC is vital to improve the flowbehaviour and also enhance the hardened properties. Over theyears, the application of lightweight self-compacting concrete(LSCC) has been progressing rapidly with proven versatility inthe reduction of the dead load of the overall structure. The mainconcern of LSCC is the consumption of lightweight aggregate,which is usually lower in density coupled with high porosity.

The agricultural industry has been a mainstay of the Malaysianeconomy for the past two decades with millions of hectares beingplanted with oil palm, rubber, paddy, sugar cane, coconut andcocoa. The extraction of useful material from these plants gener-ates various types and forms of waste material, which need to bedisposed of appropriately. Generally, they comprise ash, grains,

wastewater, shells and large combined chunks. Malaysia is one ofthe largest producers and exporters of palm oil products to variouscorners of the world. Plenty of waste by-products are collectedthroughout the palm oil processing phases. One of them is palmoil clinker (POC), which is obtained in large chunks during the oilpalm shell and fibre incineration process. Nowadays, POC isdisposed of by landfill. This not only causes soil pollution but alsoaffects the ground water supply source. Therefore, the consump-tion of these abundantly available waste products in the construc-tion industry will not only help to reduce the environmentalproblems but also provide an alternative to the diminishing naturalaggregates. The sustainability of these aggregates, which is one ofthe major issues in the construction industry, can be addressedaccordingly.

2. Background studies

Concrete generally comprises aggregate, cementitious materialsand some chemical admixtures. The bond or interaction betweenthem plays a vital part in determining most of the fresh and hard-ened properties of concrete. The proper arrangement of particu-lates in concrete not only enhances the interlocking effectsbetween them but also promotes some of the key hardening anddurability features. Hence, optimisation of PP in concrete wouldpromote enhanced performance due to the dense arrangement ofthe structure. Increasing the packing density of the aggregate canreduce the binder content as well as decrease the cost of the con-crete [1]. It has also been reported that improved aggregate pack-ing has the potential to enhance the properties of concrete in terms

10 J. Kanadasan, H.A. Razak / Materials and Design 56 (2014) 9–19

of strength, modulus of elasticity, creep and shrinkage [2]. Theinterlocking effect and interaction between the aggregate is alsoclosely related to the aggregate size whereby addition of aggre-gates greater than 2.36 mm induces good aggregate contact [3].Kwan and Fung [4] explained that increasing the packing densityhas a significant improvement on the rheological behaviour of ce-ment paste and mortar. While Jalal et al. [5] reported that theincrease in paste volume, due to the additional binder content,may enhance the rheological properties of SCC. Girish et al. [6]reported that an increase in paste volume elevates the slump flowvalues generally for varying water content. While on the otherhand, when utilising two types of lightweight aggregate with dif-ferent densities for SCC production, it was found that the lighterones gives larger slump flow [7]. Incorporation of lightweightaggregate also has the capability to reduce the viscosity of themix as the movement of aggregates gets easier through the mortarphase [8,9].

Research has diverged from the production of ordinary SCC tothe utilisation of abundantly available waste materials as aggre-gate replacement and binder substitution. As the world is movingtowards the concept of 3R (Recycle, Reuse, Reduce) and a sustain-able environment, the use of these waste materials not only helpsto save the environment but also preserves the depletion of naturalresources. Plenty of waste products from the agricultural industryare currently utilised for the production of normal concrete andSCC, such as rice husk ash (RHA), sawdust ash (SDA), palm oil fuelash (POFA), rubber aggregate and bagasse ash. Memon et al. [10],found that by incorporating 10% RHA, the strength of SCC improveswith the increase in superplastisizer (SP) content as an effect of theenhanced self-compacting behaviour. Furthermore, the addition ofbagasse ash up to 30% produces concrete with a good compressivestrength compared to control concrete [11]. The use of rubberaggregate produces SCC that is less prone to cracking due to lesserrestrained shrinkage [12]. Elinwa et al. [13] reported that SDAreacts during the cement hydration process to produce gel that en-hances the overall structure of the concrete. In addition, otherwaste by-products that are used as aggregate as well as binderreplacements in SCC production include crumb rubber, recycledglass (RG), marble dust (MD), rubble powder and waste liquid crys-tal glass (LCD). Topcu et al. [14] reported that utilising MD at200 kg/m3 can be considered as maximum and ideal to producesignificant positive effect on both the fresh and hardened proper-ties of SCC. SCC incorporating LCD glass sand at 30% substitutionlevel produced higher ultrasonic pulse velocity (UPV) and electricalresistance values compared to control mix [15]. With the additionof fine recycled concrete aggregate, the resistance of SCC againstchloride ion penetration increases [16].

Self-compacting rubberised concrete (SCRC) shows a lowerYoung’s modulus of elasticity (E) for all mixes but their flexuraltoughness improves [17]. Generally, the use of quarry waste lime-stone powder in SCC at constant cement content decreases the SPcontent and enhances the 28 days compressive strength [18]. Fur-thermore, the use of waste concrete powder (WCP) as a replace-ment for the cement generally reduces the viscosity of the pasteand compressive strength compared to the one with only ordinaryPortland cement (OPC) [19]. An investigation carried out for650 days to study the chloride induced corrosion resistance in nor-mal concrete with cement content substituted with volcanic ash(VA) and volcanic pumice powder (VPP) indicate that with 20%replacement levels, the concrete performs better in resisting corro-sion. [20].

This study is divided into two different phases. The first phasecomprised developing methods to study the PP for a combinationof POC aggregate and ordinary aggregate. The techniques weremodified to cater for the porous and lightweight physical charac-teristics of POC. The second phase of the experimental work

consisted of developing a mix proportion for SCC, incorporatingthe PP results. Selected mix proportions were verified by preparingthe actual POC SCC and their performance was evaluated for freshand hardened properties.

The mix design for SCC proposed earlier by various researchersdiffers considerably with regards to the ordinary concrete.Currently, there are no fixed methods or definite values or factorsavailable to supplement the SCC design process. Okamura and Ou-chi [21] proposed a mechanism for achieving self-compactabilityby limiting the aggregate content, lowering the water powder ratioand using SP. Generally, the method involved designing a mix pro-portion considering the coarse aggregate phase and mortar phase.Initially the coarse aggregate level was kept at 50% of the total con-crete volume and the fine aggregate ratio at 40% of the mortar vol-ume. Then, the water/powder and SP dosage was determined toachieve the required flow of SCC. Trials were carried out by testingthe mix design against the SCC properties compliance test.

Su et al. [22] proposed a new methodology to develop mixdesign for SCC through integration of the packing factor (PF) be-tween the aggregate in different states. They suggested the PFcould be defined as the ratio of aggregate in the packed state tothat of the loose state. Higher values of PF indicate a higher levelof aggregate content with a reduction in mortar volume. While,conversely, lower values of PF relate to a higher binder content.Hence, the selection of optimum PF is vital to create a balance be-tween cost and the concrete properties. In addition, in their studyChoi et al. [23] modified the Su et al. [22] mix design process byattempting a different view of the packing model. They performedthe determination of PF separately for fine and coarse aggregate.The PF obtained is related through the volume ratio of fine to totalaggregate (S/a). Furthermore, Dinakar et al. [24] designed SCC forground granulated blast furnace slag (GGBS) applying the effi-ciency concept methodology whereby in their study the efficiencyof slag activity was studied at varying replacement levels and theslag replacement was chosen according to the strength required.

The cement content, as proposed by the method was altered inthis mix design by the trial and mix process. A suitable cementcontent for a fixed aggregate ratio was studied by varying thePOC substitution ratio and fine/total aggregate (F/A) ratio. In addi-tion, the PP method, which focuses on aggregate packing and pastevolume, were integrated into the study to form the mix propor-tions. Throughout the process of mix design, POC powder wasutilised as additional binder content rather than as the percentagereplacing cement to enhance the SCC characteristics.

3. Materials used

3.1. Aggregate

Four different types of aggregate were used for this studynamely POC coarse, POC fine, gravel and sand. The characteristicsof POC itself, which is lightweight, promotes its utilisation as anaggregate in SCC. It is crushed in three different stages to obtainthe coarse, fine and powder to suit SCC requirements. Fig. 1 showsa large chunk of POC, which was obtained from the palm oil mill,while Fig. 2 shows POC coarse and POC fine after the crushing pro-cess of large chunks of POC. The physical characterisation test per-formed on each type of aggregate and the results are presented inTable 1 below.

3.2. Cementitious materials

OPC Type I was used throughout the experimental work. POCpowder was obtained by grinding POC into fine powder using a ballmill. In this study, POC powder was added into the mix to further

Fig. 1. A large POC chunk collected from the palm oil mill.

Fig. 2. POC fine and POC coarse after crushing.

Table 1Physical characteristics of the aggregate.

Properties Aggregate

Fine POC Coarse POC Sand Gravel

Aggregate size (mm) <5 5–14 <5 5–14Specific gravity 2.15 1.73 2.66 2.63Moisture content (%) 0.5 ± 0.25 1 ± 0.5 0.08 0.28Water absorption (%) 10 ± 5 3 ± 2 0.39 0.58Aggregate crushing value (%) – 56.44 – 17.93Aggregate crushing value – 16.99 – –(Ten per cent fines)

Table 2Chemical composition of cement and POC powder.

Oxides Cement POC powder

Chemical propertiesCaO 64.00 6.37SiO2 20.29 59.90SO3 2.61 0.39Fe2O3 2.94 6.93A12O3 5.37 3.89MgO 3.13 3.30P2O5 0.07 3.47K2O 0.17 15.10TiO2 0.12 0.29Mn2O3 0.12 –Na2O 0.24 –Others 0.94 0.36Loss on ignition 1.40 1.89Bogue compound % by mass –Composition of cementCompoundC3S 58.62C2S 13.95C3A 9.26C4AF 8.95

Physical propertiesSpecific gravity 3.15 2.59

J. Kanadasan, H.A. Razak / Materials and Design 56 (2014) 9–19 11

maximise the usage of palm oil waste materials. The good waterabsorption characteristic of POC powder plays a vital role in con-trolling the viscosity behaviour of fresh concrete. The chemicalcomposition of cement and POC powder are tabulated in Table 2.

3.3. Admixtures

A polycarboxylate based SP was used throughout the study. Thedensity of the SP provided by the manufacturer was 1.08 kg/L andcomprised 40% solid particles.

4. Experimental programme

4.1. Mixing ratios

Different proportions of POC aggregate and ordinary aggregatecombinations were prepared to test for PP. Fine aggregates ratios(F/A) of 0.5 and 0.6 were evaluated to ensure a wider range of ra-tios for SCC were taken into consideration. Different levels of POCsubstitution of 0%, 25%, 50%, 75% and 100% were selected for each

F/A. Fig. 3 shows the flow of the mix design process of SCC for thePOC aggregate. Since the use of POC as aggregates for SCC is onlyintroduced in this work, the variation of possible combinationswere kept wide enough to allow for the determination of the high-est and lowest PP range.

4.2. Preparation of aggregate

Four types of aggregate were used in this study i.e. POC coarse,POC fine, gravel and sand. POC coarse and gravel were grouped to-gether into the same classification called coarse aggregate. WhilePOC fine and sand were both considered as fine aggregate. Thecoarse aggregate was prepared in the range of 5–14 mm. Similarsizes of aggregate were selected to promote a good interlocking ef-fect between them to enhance the packing characteristics andflowability of SCC. For the fine aggregate, the maximum size wasfixed at 5 mm. POC fine and sand were prepared to contain aggre-gate below 5 mm sieve size. The fineness modulus (FM) of POC fineand sand were prepared to have almost similar values since theyplay a very important role in controlling fresh SCC behaviour.

The aggregate was soaked in water for 24 h at room tempera-ture to permit maximum water absorption. It was later allowedto dry until the saturated surface dry (SSD) condition was achieved.The saturated state of the aggregate prevents the ingress of waterinto the aggregate while testing for PP. Performing the test under

FAIL

PASS

• 100% POC / 0% ORD • 75% POC / 25% ORD • 50% POC / 50% ORD • 25% POC / 75% ORD • 0% POC / 100% ORD

Paste volume determination

Determination of cement content

SELECTION OF MATERIAL 1) POC coarse 2) POC fine 3) Gravel 4) Sand

Physical Characterization Tests

1) Specific gravity 2) Unit weight 3) Moisture content 4) Water absorption 5) Aggregate crushing value (ACV)

Particle Packing Measurement 1) Void volume

Particle packing

Different Combinations 1) POC coarse + POC fine (F/A – 0.5 - 0.6) 2) Gravel + sand (F/A – 0.5 - 0.6) 3) POC coarse + POC fine (F/A – 0.5 - 0.6) +

Gravel + sand (F/A – 0.5 - 0.6)

Water and additional powder content determination

Selection of correction lubrication factor (CLF)

Determination of aggregate content

Verification test Trial mix & fresh SCC

POC SCC Design Check EFNARC

requirements Excess paste effect

Determination of aggregate substitution ratio

Fig. 3. Flowchart of achieving and verifying the mix design for SCC using POC aggregate.

12 J. Kanadasan, H.A. Razak / Materials and Design 56 (2014) 9–19

SSD condition is important to avoid any loss of fluid throughabsorption during the PP test. The porous and angular nature ofPOC coarse aggregate has to be given due consideration as it sub-stantially affects the amount of voids within the concrete struc-ture. The pores on POC are expected to be filled with pasteduring the SCC production.

4.3. Determination of particle packing (PP)

Phase 1: All the aggregate particles are checked to ensure theyhave been soaked in water for 24 h. They are later brought into theSSD condition.

Phase 2: A clean and dry baseplate is prepared for mixing theaggregate. The combination of aggregate is prepared on the base-plate, as per the ratio in Table 3. They are thoroughly mixed usinga scoop and trowel for 5 min. Once a homogenous mixture ofaggregate is achieved, it is later placed into the container in a

loosely packed state. The container is checked to ensure that it isclean from any type of impurity.

Phase 3: Clean potable water is prepared of known volume. It ischecked to ensure that it is warm and under room temperature tomaintain the average density of water. Subsequently, the water ispoured slowly into each corner of the container filled with aggre-gate. Fig. 4 depicts the schematic diagram for determining the PPvalues.

Phase 4: Once the water level reaches the top surface of the con-tainer, the time on the stopwatch is started. The water level ischecked consecutively every 30 s for a period of 2 min. This is basi-cally to allow for saturation of water into the pores on the POCaggregate. Water is constantly added if there is a reduction inwater level. For complete ordinary natural aggregate substitution,this process is also carried out to ensure the same level of consis-tency. The amount of water utilised which represents total amountof voids recorded.

Table 3Ratio of aggregate combinations for PP determination.

Fine aggregate ratio Mix proportion Ratio (by volume)

Coarse aggregate Fine aggregate

POC coarse Gravel POC fine Sand

0.5 POC – 0/Ord 100 0.000 0.500 0.000 0.500POC – 25/Ord 75 0.125 0.375 0.125 0.375POC – 50/Ord 50 0.250 0.250 0.250 0.250POC – 75/Ord 25 0.375 0.125 0.375 0.125POC – 100/Ord 0 0.500 0.000 0.500 0.000

0.6 POC – 0/Ord 100 0.000 0.400 0.000 0.600POC – 25/Ord 75 0.100 0.300 0.150 0.450POC – 50/Ord 50 0.200 0.200 0.300 0.300POC – 75/Ord 25 0.300 0.100 0.450 0.150POC – 100/Ord 0 0.400 0.000 0.600 0.000

Table 4Particle packing and void ratio for different aggregate combinations.

Fine aggregateratio

Mix proportion Voidsratio

Particle packing(PP)

0.5 POC – 0/Ord – 100 0.2901 0.7099POC – 25/Ord – 75 0.3257 0.6743POC – 50/Ord – 50 0.3553 0.6447POC – 75/Ord – 25 0.3848 0.6152POC – 100/Ord – 0 0.4204 0.5796

0.6 POC – 0/Ord – 100 0.2872 0.7128POC – 25/Ord – 75 0.3425 0.6575POC – 50/Ord – 50 0.3774 0.6226POC – 75/Ord – 25 0.4028 0.5972POC – 100/Ord – 0 0.4145 0.5855

Table 5Final mix proportion analysis using PP for SCC (0.5 F/A).

Mix ID. PPCLF Paste volume (m3/m3)

TR 1 (POC 0/ORD 100) 0.5857 0.4143TR 2 (POC 25/ORD 75) 0.5563 0.4437TR 3 (POC 50/ORD 50) 0.5318 0.4682TR 4 (POC 75/ORD 25) 0.5076 0.4924TR 5 (POC 100/ORD 0) 0.4782 0.5218

Water

Mixture of aggregates

Volume known container

Fig. 4. Schematic diagram of PP test.

Fig. 5. Coating of binders onto the POC aggregate during the mixing stage.

J. Kanadasan, H.A. Razak / Materials and Design 56 (2014) 9–19 13

4.4. Proposed mix design

The PP results for various combinations of mixes are tabulatedin Table 4. The mix proportioning of SCC is carried out based onthe obtained PP results and corresponding void volume. The voidvolume which was calculated from the amount of water utilisedfor a fixed total volume of container governs the volume of pasterequired. The basic idea of this method is to determine the re-quired minimum volume of paste that would help to lubricatethe aggregate. Generally, it can be observed that the void ratio

increases with the increase in the POC aggregate. This confirmsthat a high amount of surface voids are available within andbetween the coarse and fine POC aggregate. The highest PPrequires less paste content as the arrangement of aggregate istightly packed thereby only allowing limited penetration of thepaste between the aggregate. Each different aggregate combinationmay also have a different paste level to achieve its maximum con-crete optimisation level.

Table 5 shows the final mix proportion analysis using PP. Thelowest PP represents the availability of an immense amount ofspace for the paste to provide good coating and lubrication. Theshape of the POC aggregate itself, which is flaky and porous createsa substantial number of voids to be filled with other particles. Fur-thermore, the pores on the POC aggregate could also be filled withbinder and form a paste within the aggregate skeleton. Fig. 5 showsthe coating of cement and POC powder on the coarse POC aggre-gate during the initial dry mixing process. As for SCC, the exactstandard methods or procedures for obtaining the mix design arenot available. Hence, the mix proportion developed has to be ver-ified by conducting laboratory tests to ensure they meet therequirements of EFNARC [25].

Trials were conducted to obtain the optimised cement contentlevel based on the overall range of aggregate combinations. Havinga fixed cement content is vital to ensure that good comparisonstudies can be conducted while varying other parameters in termsof combinations of aggregate. Extensive studies were performed todetermine the feasible cement content that can produce concretewith enhanced strength and SCC properties without compromisingthe effects on the environment. Three different cement contentswere examined with an alternating water binder ratio (W/B) to en-sure they meet the required SCC criteria. The amount of cementconsumed in the mix was kept low at 380 kg/m3, 400 kg/m3 and

Table 6Final mix proportion for SCC (0.5 F/A).

Mix ID. Cement (kg/m3) Water (kg/m3) POC powder (kg/m3) SP dosage (%) Aggregates (kg/m3)

POC coarse POC fine Gravel Sand

TR 1 (POC 0/ORD 100) 420 180 223 1.20 – – 770 779TR 2 (POC 25/ORD 75) 420 192 267 0.90 120 150 549 555TR 3 (POC 50/ORD 50) 420 203 304 0.75 230 286 350 354TR 4 (POC 75/ORD 25) 420 213 340 0.65 329 409 167 169TR 5 (POC 100/ORD 0) 420 225 384 0.40 414 514 – –

(a) PP at 0.6 F/A

(b) PP at 0.5 F/A

Fig. 6. Relationship between PP and POC aggregate at 0.6 F/A and 0.5 F/A.

14 J. Kanadasan, H.A. Razak / Materials and Design 56 (2014) 9–19

420 kg/m3 to maximise the use of waste materials as well as toenhance the sustainability in the mix proportion. From numeroustrials carried out, it was found that 420 kg/m3 cement content pro-vides the optimum performance of SCC. Table 6 depicts the finalmix proportion of the verification trials.

This study focuses mainly on obtaining a mix design for POC.Adjustments were made continuously to tailor the combinationsof aggregate with the highest and lowest PP values. Importancewas also given to ensure that the POC 0 mix with only naturalaggregate also satisfies the SCC range. This was carried out byapplying similar correction lubrication factor (CLF) values to allmix proportions. Since this new technique of mix proportion is car-ried out based on the aggregate characteristics, verification of themix is important to confirm the capability or effectiveness of thedesign. Hence, five trials were selected to have the control andPOC aggregates with the highest and lowest PP, respectively. Thechoice of trials is purely to evaluate the performance of SCC at peakPP regions. The cement content was kept constant throughoutwhile designing the mix. The increase in paste volume as thePOC replacement increases is balanced with POC powder. The pow-der provides additional paste volume when combined with waterand cement. Consequently, the need for additional cement can beavoided as well as limiting the effects on the environment. Fig. 6shows the three dimensional plot of PP at varying C/A ratios. Thisplot confirms the increase in the volume of the voids when the ra-tio of POC aggregate is increased. Hence, the introduction of POCaggregate at any amount needs excess paste to ensure the SCC cri-teria are achieved.

4.5. Procedure

Step 1: Determination of correction lubrication factor (CLF).A suitable CLF is chosen based on the PP values obtained and

applied to the PP values to obtain the highest paste volume avail-able for a given concrete volume. This fosters achievement of theSCC mix proportion.

PPCLF ¼ PP� CLF ð1Þ

where PPCLF = particle packing applied with correction lubricationfactor; PP = determined particle packing from experiment;CLF = correction lubrication factor.

Step 2: Calculation of coarse and fine aggregate content.In this mix design process, the aggregate content can be ob-

tained using Eq. (2). The subscript AGG refers to specific aggregatetype considered for SCC. For this study, it involves POC coarse, POCfine, gravel and sand. For each mix design, the volume ratio ofaggregate to total aggregate content is included. The adjustmentin mix design should be done according to the moisture contentand water absorption of aggregate during mixing on site.

WAGG ¼ PPCLF � A:RAGG � S:GAGG � 1000 ð2Þ

where PPCLF = particle packing applied with correction lubricationfactor; WAGG = aggregate content of in SCC (kg/m3); A.RAGG = ratioby volume of respective aggregate to total aggregate; S.GAGG = spe-cific gravity of aggregates.

Appropriate ratio of respective aggregate to total aggregate hasto be taken into consideration depending on the number of aggre-gate combination used for the study.

Step 3: Selection of cement content.The cement content has to be selected rationally to ensure that

the requirements of SCC are met. This includes the passing ability,flowability and segregation resistance. Trials have to be carried outthoroughly for a given range of cement content to satisfy the class

EFNARC Slump Flow Range (SF 2)Maximum Minimum

Fig. 7. Effect of paste volume on slump flow.

EFNARC Slump Flow Range (SF 2)Maximum Minimum

Fig. 8. Relationship between slump flow, POC substitution ratio and paste volume.

J. Kanadasan, H.A. Razak / Materials and Design 56 (2014) 9–19 15

of SCC properties required. In addition, the cement content alsodepends on the amount of powder materials added to the overallconcrete skeleton. Good adjustment between the cement contentand powder materials are vital to provide the most optimum SCCperformance.

VCEMENT ¼WCEMENT=S:GCEMENT ð3Þ

where VCEMENT = volume of cement (m3/m3); WCEMENT = cementcontent (kg/m3); S.GCEMENT = specific gravity of cement.

Step 4: Calculation of paste volume.

VPASTE ¼ 1� PPCLF ð4Þ

where VPASTE = volume of paste (m3/m3); PPCLF = particle packingapplied with correction lubrication factor.

Step 5: Calculation of water and powder content.

VPASTE � VCEMENT � VAIR ¼ VPOC Powder þ VWATER ð5Þ

where VPASTE = volume of paste (m3/m3); VCEMENT = volume ofcement (m3/m3); VAIR = air content in SCC (%); VPOC Powder = volumeof POC powder (m3/m3); VWATER = volume of water (m3/m3).

VWATER=ðVPOC Powder þ VCEMENTÞ ¼W=B ð6Þ

where VWATER = volume of water (m3/m3); VPOC Powder = volume ofPOC powder (m3/m3); VCEMENT = volume of cement (m3/m3);W/B = water/binder ratio (by volume).

Since the values of VCEMENT and W/B are known, the values ofVPOC Powder and VWATER can be obtained by solving Eqs. (7) and(8) simultaneously. The respective volume of paste, POC powder,cement and SP can be converted to weight by applying their ownspecific gravity values.

WPOC Powder ¼ VPOC Powder � S:GPOC Powder � 1000 ð7Þ

WWATER ¼ VWATER � S:GWATER � 1000 ð8Þ

Step 6: Calculation of SP dosage.Sufficient SP dosage can enhance the flowability and passing

ability of SCC. But caution also has to be given to ensure excessiveSP is not used as it would result in segregation. The right dosage ofSP has to be selected to ensure optimum performance. The watercontent of SP has to be calculated as part of the water content.

WSP ¼ SP%� ðWCEMENT þWPOC PowderÞ ð9Þ

where WSP = superplastisizer content (kg/m3); SP% = superplastisiz-er dosage (%); WCEMENT = cement content (kg/m3); WPOC Powder =POC powder content (kg/m3).

5. Results and discussion

5.1. Fresh properties

Verification of the mix design method is carried out by takinginto consideration the mix proportion for five mixes, as shown inTable 6, whereby full and partial replacements of the control andPOC aggregate are made. These mixes are selected as they repre-sent the highest and lowest PP for the given F/A ratio of 0.5. Thefresh SCC properties tests carried out include slump flow test,V-funnel, L-Box test and T500. The outcomes are later comparedwith the EFNARC [25] standard to ensure they meet the SCC crite-ria. Both compressive strength and UPV test results are presentedto establish the relationship between the PP and hardenedproperties.

5.1.1. Slump flowFig. 7 depicts the effect of the paste volume on the slump flow.

It is evident that the slump flow increases with the increase in

paste volume. As the paste volume increases, the PP valuedecreases. POC 0 recorded the lowest paste volume compared tothe other POC mixes. The bonding between the aggregate and mor-tar phase is vital for the performance of the slump flow. The highlypacked structure of POC 0 has a greater tendency to hold the par-ticles together resulting in a lower slump flow. Fig. 8 illustrates thevariation in slump flow with the increase in POC substitution ratio.As the POC substitution increases, the rise in paste volume resultedin higher flow diameter. Previous study also indicate that incre-ment in paste volume at different water content produces increas-ing slump flow which is similar to the current results [6]. Besidesthat, reduction in mass of concrete due to lightweight aggregateincorporation may also have contributed to higher slump flow.The observation made is almost similar to findings by other re-searcher for two lightweight aggregates with different densities[7]. The results above comply with the EFNARC [25] range for theslump flow class 2 (SF2) from 660 mm to 750 mm.

5.1.2. V-funnelFig. 9 shows the effect of the paste volume on the V-funnel flow

time. EFNARC [25] limited the V-funnel time to be less than 9 s and9–25 s for VF 1 and VF 2 respectively, as shown by the dashedregion. It can be seen that all the results are within the range of6–15 s and lay in both VF1 and VF2 region. Generally, a decreasingtrend of flow time can be observed as the POC substitution level in-creases. As the POC replacement level increases the paste volumeincreases which influences greatly the viscosity of the mix. While

EFNARC Viscosity (VS 2)

EFNARC Viscosity (VS 1)

Fig. 11. Effect of paste volume on T500 time.

EFNARC Viscosity (VF 1)

EFNARC Viscosity (VF 2)

Fig. 9. Relationship between V-funnel and paste volume.

16 J. Kanadasan, H.A. Razak / Materials and Design 56 (2014) 9–19

for POC 0 mix fully composed of ordinary aggregates, the highlypacked state with lesser paste volume increases the viscosity ofthe mix to produce longer flow time. Fig. 10 depicts relationshipbetween V-funnel flow time, POC substitution level and paste vol-ume. The existence of high volume of voids between POC aggre-gates due to shape and porous nature greatly contributes to thehigher paste volume. This eventually affects the viscosity of thepaste. Besides that, previous studies indicate that utilisation oflightweight aggregates also reduces the viscosity of the mix asthe transportation process gets improved [8,9].

5.1.3. L-BoxThe amount of paste available within the SCC is vital to coat and

lubricate the aggregate to enable it to flow across the obstacles. Inthis study, it is a deciding factor as it helps to provide the muchrequired coating around the porous POC aggregate. It acts as atransportation medium to migrate them according to the force offlow. From the experiment, it was found that all five mixes per-formed similarly by maintaining the height level ratio at 1.0. Thisagain proves that high F/A enhances the flowability criteria ofSCC. EFNARC [25] specifies that the passing ability range for L-Box test should be between 0.8 and 1.0.

5.1.4. T500

Fig. 11 presents the effect of the paste volume on the T500 time.The viscosity and density of the overall SCC structure also plays avital role in determining the flow time to reach the 500 mmboundary. POC 100 with maximum lightweight aggregate content

EFNARC (VF 1)

EFNARC (VF 2)

Fig. 10. Relationship between V-funnel, POC substitution ratio and paste volume.

produced a significantly shorter T500 time. The high amount ofpaste volume coupled with the significant reduction in the weightof SCC shortens the flow time, and, consequently, increases theslump flow diameter. While POC 0 with the highest PP valuerestricts the flow movement and produces a slower flow time.The slump flow time reported also shows a similar pattern. Allthe mixes produced a T500 time greater than 2 s. Fig. 12 showsthe relationship between T500 time, POC substitution ratio andpaste volume. It can be seen that the increasing paste volumedue to rise in POC aggregate replacement level greatly affects theviscosity of the mix to produce shorter flow time.

5.2. Hardened properties

Hardened properties tests were also carried out to ensure thatthe SCC satisfies the minimum requirements. This study providesan insight and correlations between some of the fresh SCC proper-ties and mechanical characteristics. The changes diagnosed withrespect to the substitution level have an input on the optimisedreplacement levels.

5.2.1. DensityFig. 13 shows the effect of POC substitution level on the density

of the SCC produced. POC aggregate which are lightweight and por-ous in nature contributes significantly to reduce the overall densityof the concrete structure. POC 100 SCC situates in the region below2000 kg/m3 which can be considered as lightweight concrete.Besides that, POC 75 situates very close to the lightweight region

EFNARC (VS 2)

EFNARC (VS 1)

Fig. 12. Relationship between T500 time and POC substitution ratio.

Fig. 15. Relationship between ultrasonic pulse velocity (UPV) test and POCsubstitution ratio.

LIGHTWEIGHT CONCRETE

Fig. 13. Effect of POC substitution ratio on density of concrete.

Fig. 16. Effect of compressive strength on paste volume.

J. Kanadasan, H.A. Razak / Materials and Design 56 (2014) 9–19 17

between 2000 and 2100 kg/m3. Approximately 18% of reduction indensity was observed for the range POC 0 and POC 100.

5.2.2. Ultrasonic pulse velocity (UPV) testFig. 14 represents the effect of the paste volume on the UPV test

results. POC 0 with higher PP attains higher UPV readings and pro-duces excellent UPV readings at 28 days compared to the othermixes. The packed structure of this mix combined with the strongbond between the aggregate and paste provides a dense mediumgiving higher velocity of the pulse through the samples. As thePOC substitution ratio increases, the void volume rises with pastevolume. Fig. 15 shows the plot of UPV readings and POC substitu-tion ratio at the age of 3 days and 28 days. A trend can be observedwith the increasing POC replacement ratio. POC 100 with the high-est paste volume significantly reduces the packing level of the SCC;hence, producing lower pulse travel velocity. The existence ofempty voids leads to a reduction in the packed matrix formationwithin the aggregate surface, which, eventually, reduces the veloc-ity of the ultrasonic pulse across the sample. Researchers agreethat a concrete can be classified as good quality if the UPV valueis between 3660 m/s and 4575 m/s [26].

5.2.3. Compressive strengthFig. 16 illustrates the relationship between the compressive

strength and paste volume. The test was performed according toBS EN 12390-3 [27]. Generally, the presence of a higher amountof paste affects the overall strength performance. POC 0 with

Fig. 14. Effect of paste volume on ultrasonic pulse velocity test (UPV) test.

Fig. 17. Relationship between compressive strength and ultrasonic pulse velocity(UPV) test.

higher PP and lower paste volume produces SCC with higherstrength compared to all other mixes. For POC SCC it can be ob-served that all of them attained early strength between 35 MPaand 40 MPa. At 28 days, all the samples obtained strength of al-most above 50 MPa. With the same W/B ratio, increasing the POCsubstitution ratio generally reduces the compressive strength of

18 J. Kanadasan, H.A. Razak / Materials and Design 56 (2014) 9–19

SCC. However, it can be observed that POC 100 performed satisfac-torily by attaining almost 68% of the strength compared to POC 0.Fig. 17 depicts the relationship between the UPV test reading andcompressive strength. The correlation carried out shows thatincreasing the POC substitution ratio reduces the UPV readings.The uniformity and packing level of the mix greatly influences boththe UPV and compressive strength. POC 100, with loosely packedstructure and high amount of paste, showed a lower compressivestrength and UPV readings.

6. Conclusions

From the investigation conducted, the following conclusionswere made based on the fresh and hardened properties of POC SCC:

� From the verification test results, it is observed that the mixdesign satisfied the required fresh SCC range requirementsaccording to EFNARC [25]. PP can be utilised as a mediumnot only to understand the interaction of aggregates withinthe SCC skeleton but also to predict the hardened propertiesof SCC.

� This new method of establishing the mix design for SCC canbe used for POC aggregate and is also applicable for othertypes of aggregate. By carrying out simple proportioningand the PP test, the mix design for SCC can be obtained.

� The early strength of POC SCC mixes were similar to that ofthe control mixes. The strength of the SCC mixes at variousreplacement rates reduced with increasing POC aggregatecontent. Almost 68% of the strength could be achieved usingfull replacement of POC compared to POC 0.

� The variation in paste volume for each mix according to PPhas a significant impact on the hardened properties of SCC.A trend can be observed between the compressive strengthand UPV values as the paste volume altered.

� Although POC as a waste material could not produce theappearance and performance on par with ordinary aggregate,generally they satisfied the fresh properties criteria accord-ing to the EFNARC [25] range and have adequate hardenedproperties.

� The POC aggregates, which are a waste material to be land-filled, perform satisfactorily as aggregate materials. Withthe escalating environmental pollution due to waste materialfrom the agricultural industry, utilising POC in concrete willbenefit the construction industry. The sustainability ofaggregate can be prolonged since the source of naturalaggregates are depleting.

Acknowledgments

The authors would like to sincerely thank University of Malayaand the Ministry of Higher Education, Malaysia for supporting thisstudy through research grants UM.C/625/1/HIR/MOHE/ENG/56 andPV079/2012A. Moreover, the assistance provided by the staff of theConcrete Laboratory are also appreciated.

Appendix A. Sample calculation of the proposed mix designmethod

The following example is a mix design for SCC mix incorporat-ing both POC and natural aggregates. Aggregate substitution ratiowith respect to POC aggregates kept at 50% substitution level.The F/A ratio fixed at 0.6. The physical characteristics of the mate-rials to be used are provided below;

PP ¼ 0:6226

VAIR ð%Þ ¼ 1:5

S:GPOC Powder ¼ 2:59; S:GPOC:C ¼ 1:73; S:GPOC:F ¼ 2:15;S:GGRAVEL ¼ 2:63; S:GSAND ¼ 2:66

A:RGRAVEL ¼ 0:2; A:RSAND ¼ 0:3; A:RPOC:C ¼ 0:2; A:RPOC:F ¼ 0:3

Step 1: Determination of correction lubrication factor (CLF).

PPCLF ¼ 0:6226� 0:825 ¼ 0:5136

Step 2: Calculation of coarse and fine aggregate contents.In this mix design process, the aggregates are proportioned by

volume and determined its particle packing.

WPOC:C ¼ 0:5136� 0:2� 1:73� 1000 ¼ 177:71 kg=m3

WPOC:F ¼ 0:5136� 0:3� 2:15� 1000 ¼ 331:27 kg=m3

WGRAVEL ¼ 0:5136� 0:2� 2:63� 1000 ¼ 270:15 kg=m3

WSAND ¼ 0:5136� 0:3� 2:66� 1000 ¼ 409:85 kg=m3

Step 3: Selection of cement content.

VCEMENT ¼ 420=3150 ¼ 0:1333 m3=m3

Step 4: Calculation of paste volume.

VPASTE ¼ 1� 0:5136 ¼ 0:4864 m3=m3

Step 5: Calculation of water and powder content.

VPASTE � VCEMENT � VAIR ¼ VPOC Powder þ VWATER

0:4864� 0:1333� 0:015 ¼ VPOC Powder þ VWATER

VWATER þ VPOC Powder ¼ 0:3381

A predetermined W/B ratio (by volume) selected initially to ob-tain a fixed W/C ratio (by weight). Adjustment could be made laterto obtain a gradually increasing or fixed ratios tailored according toresearchers preferences. In this example, W/B 0.8657 was obtainedto produce a W/C ratio (by weight) of 0.30.

VWATER=ðVPOC Powder þ 0:1333Þ ¼ 0:8657

VWATER � 0:8657VPOC Powder ¼ 0:1154

VPOC Powder ¼ 0:11937 m3=m3

VWATER ¼ 0:21873 m3=m3

WPOC Powder ¼ 0:11937� 2:59� 1000 ¼ 309:17 kg=m3

WWATER ¼ 0:21873� 1:00� 1000 ¼ 218:73 kg=m3

Step 6: Calculation of SP dosage.

WSP ¼ 0:75%� ð420þ 309:17Þ ¼ 5:4688 kg=m3

References

[1] He H, Guo Z, Stroeven P, Stroeven M, Sluys LJ. Characterization of the packingof aggregate in concrete by a discrete element approach. Mater Charact2009;60:1082–7.

[2] Sobolev K, Amirjanov A. Application of genetic algorithm for modeling of densepacking of concrete aggregates. Constr Build Mater 2010;24:1449–55.

J. Kanadasan, H.A. Razak / Materials and Design 56 (2014) 9–19 19

[3] Shen S, Yu H. Characterize packing of aggregate particles for paving materials:particle size impact. Constr Build Mater 2011;25:1362–8.

[4] Kwan AKH, Fung WWS. Packing density measurement and modelling of fineaggregate and mortar. Cem Concr Compos 2009;31:349–57.

[5] Jalal M, Mansouri E, Sharifipour M, Pouladkhan AR. Mechanical, rheological,durability and microstructural properties of high performance self-compactingconcrete containing SiO2 micro and nanoparticles. Mater Des2012;34:389–400.

[6] Girish S, Ranganath RV, Vengala J. Influence of powder and paste on flowproperties of SCC. Constr Build Mater 2010;24:2481–8.

[7] Kim YJ, Choi YW, Lachemi M. Characteristics of self-consolidating concreteusing two types of lightweight coarse aggregates. Constr Build Mater2010;24:11–6.

[8] Uygunoglu T, Topçu _IB. Thermal expansion of self-consolidating normal andlightweight aggregate concrete at elevated temperature. Constr Build Mater2009;23:3063–9.

[9] Wu Z, Zhang Y, Zheng J, Ding Y. An experimental study on the workability ofself-compacting lightweight concrete. Constr Build Mater 2009;23:2087–92.

[10] Memon SA, Shaikh MA, Akbar H. Utilization of rice husk ash as viscositymodifying agent in self compacting concrete. Constr Build Mater 2011;25:1044–8.

[11] Rukzon S, Chindaprasirt P. Utilization of bagasse ash in high-strength concrete.Mater Des 2012;34:45–50.

[12] Turatsinze A, Garros M. On the modulus of elasticity and strain capacity of self-compacting concrete incorporating rubber aggregates. Resour Conserv Recycl2008;52:1209–15.

[13] Elinwa AU, Ejeh SP, Mamuda AM. Assessing of the fresh concrete properties ofself-compacting concrete containing sawdust ash. Constr Build Mater2008;22:1178–82.

[14] Topçu _IB, Bilir T, Uygunoglu T. Effect of waste marble dust content as filler onproperties of self-compacting concrete. Constr Build Mater 2009;23:1947–53.

[15] Wang H-Y, Huang W-L. Durability of self-consolidating concrete using wasteLCD glass. Constr Build Mater 2010;24:1008–13.

[16] Kou SC, Poon CS. Properties of self-compacting concrete prepared withcoarse and fine recycled concrete aggregates. Cem Concr Compos 2009;31:622–7.

[17] Najim KB, Hall MR. Mechanical and dynamic properties of self-compactingcrumb rubber modified concrete. Constr Build Mater 2012;27:521–30.

[18] Felekoglu B. Utilisation of high volumes of limestone quarry wastes inconcrete industry (self-compacting concrete case). Resour Conserv Recycl2007;51:770–91.

[19] Kim YJ, Choi YW. Utilization of waste concrete powder as a substitutionmaterial for cement. Constr Build Mater 2012;30:500–4.

[20] Anwar Hossain KM. Chloride induced corrosion of reinforcement in volcanicash and pumice based blended concrete. Cem Concr Compos 2005;27:381–90.

[21] Okamura H, Ouchi M. Self-compacting concrete. J Adv Concr Technol2003;1:5–15.

[22] Su N, Hsu K-C, Chai H-W. A simple mix design method for self-compactingconcrete. Cem Concr Res 2001;31:1799–807.

[23] Choi YW, Kim YJ, Shin HC, Moon HY. An experimental research on the fluidityand mechanical properties of high-strength lightweight self-compactingconcrete. Cem Concr Res 2006;36:1595–602.

[24] Dinakar P, Sethy KP, Sahoo UC. Design of self-compacting concrete withground granulated blast furnace slag. Mater Des 2013;43:161–9.

[25] EFNARC. The European guidelines for self-compacting concrete: specification,production and use. 2005:p. 68.

[26] Hwang C-L, Bui LA-T, Lin K-L, Lo C-T. Manufacture and performance oflightweight aggregate from municipal solid waste incinerator fly ash andreservoir sediment for self-consolidating lightweight concrete. Cem ConcrCompos 2012;34:1159–66.

[27] BS EN 12390-3. Testing hardened concrete – compressive strength of testspecimens. 2009.