Construction and Building Materials 50 (2014) 499–507

Contents lists available at ScienceDirect

Construction and Building Materials

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

Impact resistance of hybrid fibre-reinforced oil palm shell concrete

0950-0618/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.conbuildmat.2013.10.016

⇑ Corresponding author. Tel.: +60 379677632; fax: +60 379675318.E-mail address: johnson@um.edu.my (U.J. Alengaram).

Kim Hung Mo, Soon Poh Yap, U. Johnson Alengaram ⇑, Mohd Zamin Jumaat, Chun Hooi BuDepartment of Civil Engineering, University of Malaya, 50603 Kuala Lumpur, Malaysia

h i g h l i g h t s

� Hybrid fibre of 0.9% steel fibre + 0.1% PP fibre produced the highest impact energy.� Uncrushed OPSC had better impact resistance compared to crushed OPSC.� The lowest crack width was observed in the mix with 1.0% steel fibre.� The highest compressive strength and compressive energy were found for 1.0% steel fibre in uncrushed OPSC.

a r t i c l e i n f o

Article history:Received 20 May 2013Received in revised form 30 September 2013Accepted 4 October 2013

Keywords:Hybrid fibreFibre-reinforced lightweight concreteDrop hammer impactImpact energyCrack growth resistanceImpact ductility index

a b s t r a c t

This paper presents the results of an experimental impact test conducted using drop hammer on the plainand the fibre reinforced oil palm shell concrete (FROPSC) panels. The variables investigated are differentcontents of steel (0.75%, 0.9%, 1%) and polypropylene fibres (0.1%, 0.25%, 1%), with uncrushed and crushedOPS. The FROPSC with uncrushed OPS developed higher initial and final impact resistance compared tospecimens with crushed OPS. The specimen with 0.9% steel + 0.1% polypropylene (PP) hybrid-FROPSC,developed excellent impact energy of about 17 kJ that was 60 times higher than the plain OPSC. Itsimpact ductility index (li) of 42 is double the value compared to other specimens. It also showed excel-lent crack growth resistance due to secondary cracks formation. Final crack widths of FROPSC rangedbetween 0.079 and 0.507 mm. Further, the compressive energy of 785 J was found for specimen with1% steel fibres.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

The application of concrete is not only limited to structuralmembers, but it is also used in the construction of barriers and pro-tective structures in places that are subjected to high repeated im-pact loads. The examples of impact loadings are vehicular and shipcollisions with structure, rock falls, missile impacts, explosions,machine dynamics, wind gusts and earthquakes [1,2]. Concretesubjected to high impact loads will experience significant damagein the structural stability and integrity. The residual strength ofconcrete may be questionable once damage has occurred [1].Therefore, impact resistant concrete has to be designed to sustainrepeated impact loading effectively. Most impact resistant struc-tures require the use of mass concrete or high performance con-crete; which leads to higher construction costs of the structures.As such, lightweight concrete (LWC) could provide a more eco-nomic-friendly alternative for impact resistant structures.

LWC can be produced using oil palm shell (OPS) as light-weight coarse aggregates [3–6]. In Malaysia, the huge oil palmindustry generates 4 million tons of oil palm shell (OPS) as wastematerials annually [7]. The vast amount of shells is stockpiled inopen air which might contribute to air, water and land pollu-tions. Conversion of OPS into potential replacement for conven-tional crushed granite aggregate contributes to sustainability asit might reduce both the extraction of granite stones and theenvironmental impact. A number of investigations had beencarried out in the last decade to produce structural grade OPSconcrete (OPSC) [8–11]. Recent investigations were focused onthe production of high strength OPSC by using crushed OPS, withcompressive strength exceeding 40 MPa being successfullydeveloped [12–14].

The role of coarse aggregate in impact-resistant concrete is vitalas it acts as a barrier to the crack propagation [15]. Published re-searches showed that the lightweight OPS has lower aggregate im-pact value (AIV) than the crushed granite aggregate, an indicationof high impact resistance of OPS [4,9]. However, LWC is consideredas a brittle material [16]. The higher the compressive strength ofLWC, the higher is its brittleness. To compensate the brittlenessof LWC, addition of fibres into concrete is desirable to improve

500 K.H. Mo et al. / Construction and Building Materials 50 (2014) 499–507

both the ductility and impact strength of concrete. The inclusion ofsteel fibres in concrete improves ductility and energy absorptioncapacity under repeated impact loading as reported in publishedstudies [17–20]. Steel fibres volume of above 0.5% was mostly usedto enhance the impact resistance of concrete [21–24]. Theenhancement is attributed to the crack bridging mechanism wherethe propagation of crack is blocked by the presence of steel fibres.

Apart from the steel fibres, there are increasing applications ofsynthetic fibres such as polypropylene (PP) and nylon fibres in con-crete and these fibres were found to improve the impact resistanceof concrete [25–28]. Although synthetic fibre reinforced concreteprovides lower impact resistance than that of steel fibres, theadded advantages of synthetic fibres are light and non-corrosive.Hence the idea of fibre hybridization is introduced to improvethe impact resistance and crack growth resistance of concrete[20,24,29–31]. The hybrid fibres in impact resistance structuresserve to preserve the impact strength of such structures over alonger period of time. This is especially important for structuresthat are subjected to high possibility of steel fibre corrosion, suchas in coastal and marine structures and in humid condition whichis common in tropical countries like Malaysia. Further, for a givenvolume of fibres, increasing the amount of synthetic fibres with arelatively lower density than steel fibres in the hybrid fibre systemreduces the dead load of the structural members. Investigation onfibre reinforced OPSC (FROPSC) had been done by Shafigh et al. [32]and Yap et al. [33], using steel and synthetic fibres individually butwork has yet to been carried out on hybrid FROPSC (h-FROPSC).

The beneficial effects of hybrid fibres in concrete contributed tothe investigation of h-FROPSC in this study. The addition of hybridfibres in the OPSC might enhance both the impact and crackgrowth resistances of the OPSC. The combined effect of LWC andhybrid fibres could greatly reduce the construction costs. However,there is uncertainty to what extent the fibres combination can beused to obtain the optimum impact resistance. Therefore, the focusof the present study is to investigate the effects of steel–PP hybridfibres of different percentages on the impact behaviour of h-FROPSC. The other characteristics investigated in this study includeslump, oven-dry density (ODD), compressive strength and ultra-sonic pulse velocity (UPV). The effect of the crushed and uncrushedOPS in the h-FROPSC is also investigated and reported.

This study on h-FROPSC could enhance the understanding onthe effect of hybrid fibres in OPSC. The enhancement of the impactand crack growth resistance would enable the possibility ofFROPSC to be used as sacrificial protective barriers such as highwaycrush cushions, mountain rock fall barriers, heavy industry walls orfloors, and dykes.

2. Experimental program

The main objective of the experimental study was to evaluate the impact resis-tance of steel fibre reinforced OPSC and hybrid steel-polypropylene fibre reinforcedOPSC. The other tests include compressive strength and modulus of elasticity.

2.1. Materials

2.1.1. CementASTM Type I Ordinary Portland Cement with Blaine specific surface area and

specific gravity of 335 m2/kg and 3.10, respectively was used in this investigation.The cement content was kept constant at 550 kg/m3 for all the mixtures. Silica fume(SF) of 10% by weight of cement was used as a supplementary cementitious mate-rial with a specific gravity of 2.10.

2.1.2. Coarse and fine aggregateThe coarse aggregate used in this study was OPS collected from local palm oil

factory, in both uncrushed and crushed conditions, with maximum sizes of14 mm and 9 mm, respectively (Fig. 1). The uncrushed OPS has concave and convexshape with smooth surface on the outer convex side. The crushed OPS has more spi-ky edges than the uncrushed OPS (Fig. 1b). The physical properties of the uncrushed

and crushed OPS are given in Table 1. Both the crushed and the uncrushed OPS havelower aggregate impact value (AIV) and bulk density than the crushed graniteaggregate. The OPS content in all the mixes was kept constant at 360 kg/m3.

Mining sand with specific gravity and fineness modulus of 2.67 and 2.70 respec-tively was used as fine aggregate. The fine aggregate content was kept constant at780 kg/m3 for all mixes.

2.1.3. Water and superplasticizerPotable water with pH value of 6 was used in all the mixes. In order to improve

the workability of the concrete mixes, a polycarboxylate ether-based superplasti-cizer of 1.2% of cement weight was used. The water to binder ratio of 0.30 was usedfor all the mixes.

2.1.4. FibresThe hybrid fibres in this study include (i) hooked-end steel fibre (aspect ra-

tio = 65 and length = 35 mm) and (ii) fibrillated PP fibre (length = 12 mm). The spe-cific gravity of steel fibre and PP fibre are 7.9 and 0.9, respectively.

2.2. Mixing procedure

A total of 10 mixes were prepared. The mix proportions of all the concretemixes are shown in Table 2 with different steel and PP fibres combinations. Initiallythe coarse and fine aggregates were mixed in the rotary mixer followed by cementand silica fume for about 5 min. This was followed by the addition of water and SPand the mixing continued for another 6 min. Finally, the fibres were added andmixed for a further 2 min.

2.3. Specimen moulding and testing

The concrete was cast in 100 mm cubes, 150 / � 300 mm cylinders and600 � 600 � 50 mm panels for testing the compressive strength, modulus of elas-ticity/compressive energy and drop hammer impact test, respectively. The de-moulding was done after 24 h and the concrete specimens were cured in water tillthe age of testing. The compressive strength and modulus of elasticity tests weredone in accordance to British Standard (BS) 1881: Part 118. The cube compressivetest was tested at 1-, 3-, 7-, 28-, 56-, 90- and 180-day, while the modulus of elas-ticity was tested at the age of 28-day.

2.3.1. Compressive energyThe compressive energy of concrete was taken as the area under the load–

deflection curve obtained by subjecting the concrete cylinder under compressiveloading. The rate of loading was kept at 4.42 kN/s similar to that of the modulusof elasticity test.

2.3.2. Drop hammer impact testThe drop hammer impact test was done based on modification of the recom-

mendations by ACI Committee 544 in which an impact specimen is subjected to re-peated blows on the same spot. In this modified impact test, a 10 kg drop hammerwas released from a height of 600 mm on the panel specimen (Fig. 2). The numberof blows to cause the first visible crack and failure was observed and used to calcu-late the first crack and failure impact energy of the concrete, respectively. The im-pact energy is given in the following equation:

Eimpact ¼ mgh� N ð1Þ

where Eimpact = impact energy in Joule (J); m = mass of drop hammer = 10 kg;g = 9.81 m/s2; h = releasing height of drop hammer = 600 mm; N = number of blows.

The ratio of the number of blows to cause failure, Nf to the number of blows tocause the first crack, Nc is defined as impact ductile index, li = Nf/Nc [29]. The crackwidths were also measured using a high magnification crack microscope. Fig. 3shows one of the crack width measurements observed using the microscope. Fur-ther, the microscope could provide clear illustrations on the pull out of fibres withinthe crack as shown in the Section 3.6.4.

3. Results and discussion

3.1. Workability

3.1.1. Slump valuesThe slump values of the OPSC mixes are given in Table 3. In gen-

eral, the addition of fibres in concrete caused significant loss in theworkability of the concrete. The large surface area of fibres ad-sorbed more cement mortar around the fibres and hence the vis-cosity of the concrete increased, resulting in low slump values[16]. The slump values of FROPSC in the range of 20–50 mm wasfound 60–70% and 40–60% lower than control mixes for uncrushed

Fig. 1. (a) Uncrushed and (b) crushed OPS.

Table 1Comparison of the physical properties between OPS and crushed granite [4].

Physical property Uncrushed OPS Crushed OPS Crushed granite [4]

Maximum size (mm) 14 9 –Compacted bulk density (kg/m3) 635 658 1470Specific gravity 1.37 1.35 2.6124-h water absorption (%) 24.3 19.1 0.76Aggregate impact value (%) 2.11 2.63 17.29

Table 2Fibre combination of OPSC and FROPSC mixes.

No. Mix designation Steel fibre (%) PP fibre (%) Total volume (%)

Uncrushed OPS1 A1 0 0 02 A2 1.00 0 1.003 A3 0.90 0.10 1.004 A4 0.75 0.25 1.005 A5 0 1.00 1.00

Crushed OPS6 B1 0 0 07 B2 1.00 0 1.008 B3 0.90 0.10 1.009 B4 0.75 0.25 1.0010 B5 0 1.00 1.00

K.H. Mo et al. / Construction and Building Materials 50 (2014) 499–507 501

and crushed OPS mixes, respectively. However in the case of LWC,the slump values in the range of 0–25 mm produce satisfactorycompaction [33]. Further, the mixes A5 and B5 with 1% PP fibresproduced the lowest slump. The dispersion of the fibres in the con-crete was not uniform and clogging of fibres was encountered inthe FROPSC. The clogging of fibres is commonly associated withPP fibres [34,35].

Fig. 2. Impact test set up.

3.1.2. Effect of fibre hybridization on slumpResearches on hybrid fibre-reinforced concrete showed that the

effect of steel fibres on the loss of slump was more significant thanPP fibres [16,36]. Chen and Liu [16] reported that the ‘holding ef-fect’ of steel fibres reduces surface bleeding and sedimentation ofaggregates, resulting in higher viscosity of concrete. However, inthe present study, the mixes A5 and B5 with 1% PP fibres producedlower slump compared to mixes A2 and B2. This might be attrib-uted to the high dosage of PP fibres that caused poor fibre disper-sion during the mixing. Thus, the poor dispersion effect of PP fibrescontributed to higher loss of slump compared to the slump loss forsimilar dosage of steel fibres; however, the effect of PP fibres in theworkability was insignificant in hybrid fibre system due to lowdosage of fibres. The slump values of mixes with hybrid fibres(A3 and A4, B3 and B4) exhibited close slump values to those ofmixes with only steel fibres (A2 and B2) and this might be attrib-

uted to the high volume of steel fibres (0.75–0.9%) used in theconcrete.

3.1.3. Effect of uncrushed and crushed OPS on workabilityThe mix B1 using crushed OPS showed 25 mm reduction in the

slump value compared to the mix A1 with uncrushed OPS. Thecrushed OPS has larger surface area compared to the uncrushedOPS [12]. Hence, the crushed OPS requires more quantity of mortarand for a given mortar content, it led to low workability. However,the remaining mixes (A2–A5 and B2–B5) with fibres showed neg-ligible effect as the effect of fibres on workability was moreevident.

Fig. 3. Crack width measurement.

0

10

20

30

40

50

60

0% fibre 1.0% steel 0.9% steel + 0.1% PP

0.75% steel + 0.25% PP

1.0% PP

Com

pres

sive

str

engt

h (M

Pa)

Fibre content

Uncrushed OPS

Crushed OPS

Fig. 4. Comparison of 28-day compressive strength of OPSC and FROPSC.

502 K.H. Mo et al. / Construction and Building Materials 50 (2014) 499–507

3.2. Oven-dry density (ODD)

EN 206-1 [37] defines lightweight concrete (LWC) as concretehaving an oven-dry density (ODD) of not less than 800 kg/m3 andnot more than 2000 kg/m3 produced using lightweight aggregatefor all or part of the total aggregate. As seen from Table 3, all theFROPSC in this investigation produced ODD of less than 2000 kg/m3 and thus it could be categorized as LWC. The addition of steelfibre into OPSC significantly increased the density of the concrete(Table 3). The higher the steel fibre content, the higher the densityof the concrete. This is evident in the mixes with 0.75–1% steel fi-bre (A2–A4 and B2–B4) that produced a density increment as highas 240 kg/m3 (or 14% increment) compared to the control mixes.This was due to high specific gravity of steel fibre [32,38]. In con-trast, the addition of PP fibres had no influence on the densitydue to its low specific gravity of 0.90.

3.3. Compressive strength

Table 3 shows 28-day compressive strength of the controlmixes of OPSC with uncrushed (A1) and crushed (B1) OPS as37 MPa and 43 MPa, respectively. These values were found to bequite similar to the reported values for OPSC [12,32]. The highestand the lowest 28-day compressive strengths were obtained formixes with 1% steel and PP fibres, respectively. In general, theenhancement of compressive strength between 28 and 180 dayswas not significant.

3.3.1. Development of compressive strengthTable 3 shows that all the mixes achieved high early strength

at the age of 7-day. The mixes reached 55–70%, 70–86% and80–96% of 28-day compressive strength at the age of 1-, 3- and

Table 3Workability, oven-dry densities and compressive strengths of OPSC and FROPSC.

Mix Slump (mm) Oven-dry density, ODD (kg/m3) Compressive

1-day

A1 75 1770 24.4A2 30 1900 30.6A3 30 1870 27.5A4 30 1810 24.9A5 20 1770 21.9B1 50 1700 29.9B2 30 1940 30.2B3 30 1900 28.9B4 30 1820 21.0B5 20 1700 15.8

7-day, respectively. The high early strength might be attributedto the addition of silica fume. The highly reactive and micro-fillereffects of silica fume contributed to the early strength of OPSC[8].

Most of the studies on the OPSC show the compressive strengthup to 28-day. In this investigation, the long term compressivestrength of OPSC and FROPSC were found up to the age of 180-day. The results showed that there was no retrogression of strengthafter the age of 28-day and hence the discussion in this article isbased on the 28-day compressive strength.

3.3.2. Effect of fibre hybridization on compressive strengthThe comparison of 28-day compressive strength of all mixes is

shown in Fig. 4. The addition of the PP fibres beyond 0.25% reducedthe compressive strength of OPSC and similar findings have beenreported by researchers [16,28,39]. The low strength of mixes withPP fibres could be attributed to its low stiffness compared to thesteel fibres. Further, the dispersion problem caused poor compac-tion of the cement matrix and fibre and consequently reducedthe compressive strength of FROPSC with PP fibres more than0.25%.

The PP fibres had no effect in enhancing the compressivestrength of the OPSC, thus it can be inferred that the compressivestrength in h-FROPSC as reported in Table 3 could be the resultof steel fibres. Previously published results showed that the addi-tion of steel fibres increased the compressive strength of concrete[22,23,32,36,38,40,41]. The steel fibres are effective in arresting themicro-cracks and limit the crack propagation; and hence the steelfibres allow the concrete to sustain additional compressive loadbefore failing. This was evident in A2/B2 and A3/B3 mixes whichcontain 1% and 0.9% steel fibre, respectively; these mixes with 1%and 0.9% steel fibre respectively increased the compressivestrength of OPSC significantly. The highest enhancement of com-pressive strength of 34% was found for the mix A3, relative tothe control mix A1. Campione et al. [42] also reported compressive

strength (MPa)

3-day 7-day 28-day 56-day 90-day 180-day

31.8 35.4 37.0 37.1 37.3 37.336.2 39.4 49.2 49.9 49.8 49.936.1 40.8 49.4 49.5 49.5 49.930.6 33.1 37.8 38.3 38.4 38.424.3 28.5 34.9 35.7 35.7 35.736.9 41.0 42.9 43.9 44.5 44.537.5 43.4 46.6 47.2 47.1 47.336.1 39.5 45.7 45.9 46.5 46.426.7 32.3 37.3 37.9 38.1 38.319.6 22.7 26.8 26.9 26.9 27.0

Table 4Development of ultrasonic pulse velocity (UPV) of OPSC and FROPSC.

Mix Ultrasonic pulse velocity, UPV (km/s)

1-day 3-day 7-day 28-day 56-day 90-day 180-day

A1 3.22 3.58 3.91 3.91 3.92 3.95 3.94A2 3.68 3.80 3.86 4.07 4.11 4.11 4.11A3 3.64 3.89 3.97 4.05 4.06 4.07 4.08A4 3.61 3.88 3.91 4.02 4.03 4.05 4.05A5 3.57 3.83 3.92 3.98 4.02 4.03 4.03B1 3.29 3.56 3.82 3.92 4.03 4.07 4.07B2 3.66 3.80 3.85 3.98 3.95 4.03 4.05B3 3.53 3.79 3.89 3.98 3.98 3.98 4.00B4 3.17 3.57 3.80 3.80 3.80 3.82 3.84B5 3.51 3.61 3.67 3.68 3.68 3.68 3.68

K.H. Mo et al. / Construction and Building Materials 50 (2014) 499–507 503

strength increase of up to 30% when steel fibres were added intolightweight expanded clay aggregate concrete and they attributedthis to the effectiveness of fibres when there is less aggregate inter-ference [42]. In LWC, the aggregate to binder ratio is usually lowand therefore there is reduced interference from the aggregates.Further, Campione et al. [42] reported that fibre reinforced LWCperformed better than NWC with the same percentage of fibres.Duzgun et al. [38] reinforced that finding and found higher com-pressive strength improvement with the addition of steel fibreswhen an increased amount of lightweight aggregate was used toreplace the conventional aggregates. It was noteworthy that therewas marginal difference in the compressive strengths between themixes with 0.9% to 1% steel fibre. Increase of steel fibre contentabove 0.9% resulted in less than 2% difference in compressivestrength for both A2/A3 and B2/B3 mixes. A possible explanationcould be the non-uniform dispersion of the fibres [41]. The addi-tion of steel fibre of 0.75% with 0.25% PP fibres in the mixes A4and B4 produced slight difference in the compressive strength,compared to the control concrete. Therefore, in terms of the com-pressive strength, the optimum PP content in the h-FROPSC is<0.25%. Any increase in the PP fibre beyond this value would resultin the decrease of the strength. This finding is comparable to thereported results on the effect of steel–PP fibres in normal weightconcrete [34].

3.3.3. Comparison between uncrushed and crushed OPSThe non-fibrous OPSC with crushed aggregates produced high-

er compressive strength compared to OPSC with uncrushedaggregates. The crushing of OPS reduces the smooth concaveand convex surfaces and increases the rough and spiky brokenedges of OPS. This enhanced the bond between the OPS and thecement paste [12]. However, when fibres were added, the OPSCwith crushed OPS produced the compressive strength close tothat of uncrushed OPS. In general, the addition of fibres slightlyimproves the compressive strength of concrete [16,32]. All thefour cases of FROPSC with crushed aggregates produced lowerstrength than the corresponding FROPSC with uncrushed aggre-gates. It should be noted that in the FROPSC produced withcrushed OPS, the quantity of crushed OPS aggregates used is high-er than the corresponding FROPSC with the uncrushed OPS. Thiscould be due to the smaller size of crushed OPS, which leads tolarger number of crushed OPS particles compared to that ofuncrushed OPS for a given weight of OPS. The high quantity ofcrushed OPS hindered the effective dispersion of fibres betweenthe OPS aggregates. Thus, the effect of fibres on the enhancementof the compressive strength in the FROPSC with crushed aggre-gates was not evident.

Table 5Modulus of elasticity and compressive energy of OPSC and FROPSC (28-day).

Mix 28-day modulus ofelasticity (GPa)

28-day compressiveenergy (Nm)

A1 10.03 279A2 14.17 785A3 11.65 678A4 11.08 591A5 10.44 248B1 9.55 294B2 16.17 617B3 13.94 450B4 11.73 301B5 10.04 155

3.4. Ultrasonic pulse velocity (UPV)

Ultrasonic pulse velocity (UPV) is an indicator of compressivestrength of concrete and quality of aggregates used [43]. LowUPV value indicates the presence of the internal voids or porousaggregates in the concrete. The values of UPV for all the mixes werefound within the range of 3.17–4.11 km/s and these values in-creased with the age as shown in Table 4. The high early strengthdevelopment of the OPSC due to the addition of silica fume is re-flected in the UPV values. There was insignificant improvementin the UPV values between 28- and 180 days. In general, the addi-tion of fibres makes the compaction of concrete difficult and likelyto lead to pores within the concrete with reduced UPV values;however, all the mixes with fibres produced 7-day UPV values be-tween 3.67 and 3.97 showing the quality of concrete as ‘‘good’’[44].

3.5. Modulus of elasticity (MOE) and compressive energy

The modulus of elasticity (MOE) of the control OPSC wasfound to be about 10 GPa (Table 5). The values were withinthe range reported by previous researcher for OPSC [7]. Similarto the compressive strength, the addition of steel and hybrid fi-bres improved the MOE; however, the mix with mono PP fibresdid not have any influence on the MOE of FROPSC. In contrast,the MOE of all crushed OPS mixes showed 6–20% higher valuesthan their corresponding uncrushed companions. This could beattributed to the stronger aggregate-paste bond from theincreasing spiky edges in crushed OPS compared to the uncru-shed OPS. The effect of steel fibres on the MOE is evident asthe highest MOE of 16.2 GPa was obtained for the mix B2, 70%more than the control mix, B1. In hybrid mixes, the strongerand stiffer fibres (steel fibre) improve the first crack stress andultimate strength [34].

The observation on the compressive energy of FROPSC wasfound similar to that of MOE as shown in Table 5. The additionof steel fibres in OPSC considerably enhanced the compressiveenergy of the concrete. The steel fibres of 0.75%, 0.9% and 1%in the uncrushed and crushed OPS mixes enhanced thecompressive energy by 110–180% and 3–110% relative to thecontrol mixes, respectively. The ability of steel fibres in bridgingthe micro-cracks enabled the specimen to absorb more energybefore the development of macro-cracks. In contrast to theMOE, the compressive energy of the specimens with uncrushedOPS (with lower AIV than the crushed OPS (Table 2)) was foundmore than that of specimens with crushed OPS. The uncrushedOPS has convex and concave surfaces that gave rise to moreabsorption of compressive energy compared to crushed OPS.The specimens with only PP fibres produced the lowest compres-sion energy due to its low stiffness and non-uniformity in distri-bution of fibres.

Table 6Impact test results.

Mix Blow number to cause firstcrack

Impact energy (first crack),Eimpact,1st,cr (J)

Blow number to cause specimenfailure

Impact energy (specimen failure),Eimpact,fail (J)

Impact ductileindex, li

A1 1 59 5 294 5.0A2 7 412 181 10,654 22.6A3 7 412 293 17,246 41.9A4 6 353 120 7063 20.0A5 2 118 34 2001 17.0B1 1 59 4 235 4.0B2 6 353 124 7299 21.7B3 6 353 135 7652 22.5B4 5 294 94 5533 18.8B5 1 59 12 706 12.0

0% fibre 1.0% steel 0.9% steel + 0.1% PP

0.75% steel + 0.25% PP

1.0% PP

No

of b

low

s to

fir

st c

rack

Fibre content

Uncrushed OPS

Crushed OPS

Fig. 5. Relationship between fibre content and blow number to cause first crack.

R² = 0.81

0

100

200

300

400

500

0 200 400 600 800 1000Firs

t cra

ck im

pact

ene

rgy,

Eim

pact

,1st

cr (J

)

Compressive energy, E comp (Nm)

Fig. 6. Correlation between compressive toughness and the first crack impactenergy of FROPSC.

0

50

100

150

200

250

300

350

0% fibre 1.0% steel 0.9% steel + 0.1%

PP

0.75% steel +

0.25% PP

1.0% PP

No

of b

low

s to

fai

lure

Fibre content

Uncrushed OPS

Crushed OPS

Fig. 7. Relationship between fibre content and blow number to cause failure.

Fig. 8. Crack propagation process through OPS aggregates [45].

504 K.H. Mo et al. / Construction and Building Materials 50 (2014) 499–507

3.6. Impact energy

3.6.1. First crack impact energyThe first crack impact energy and the number of blows to cause

the first crack are shown in Table 6 and Fig. 5, respectively. Thedrop weight hammer caused visible cracks to occur upon the firstblow in the control OPS concrete (A1 and B1). Likewise, the speci-mens with only PP fibres developed first crack with just single im-pact. In contrast the effect of steel fibres in both the impact energyand the first crack is quite significant. Generally LWC is brittle [16],but OPSC has ductility characteristics [9]. The first crack impact en-ergy of OPSC was found as high as 50% compared to that of highstrength concrete reported by Yew et al. [30].

The observation on the first crack impact strength of FROPSC issimilar to that of the compressive energy, as both measures the en-ergy absorption capacity of FROPSC (Fig. 5). Steel fibres mainlycontributed to the first crack impact resistance of FROPSC as theaddition of steel fibres enhanced the first crack strength of the con-crete by at least 5–7 times, compared to the control OPSC. The steelfibres were found highly effective in inhibiting the growth of mi-cro-cracks and blunting the propagation of these cracks beforethe cracks joined up to form macro-cracks.

The first crack impact strengths of uncrushed OPS mixes werefound higher than that of the crushed OPS mixes. This could beattributed to the low AIV of uncrushed OPS as these aggregates re-sist impact due to their shape and orientation of the aggregate dur-ing the impact test.

Fig. 9. (a) Primary cracks in OPSC panel and (b) primary and secondary cracks in FROPSC panel.

Fig. 10. Microscopic view of pulling out of (a) steel fibre and (b) PP fibre.

Table 7Crack widths and number of secondary cracks prior to failure of all mixes.

Mix Crack width (mm) Number of secondary cracks

First crack Final crack

A1 0.166 0.983 0A2 0.060 0.079 9A3 0.090 0.124 11A4 0.092 0.219 6A5 0.157 0.507 3B1 0.280 1.085 0B2 0.045 0.094 8B3 0.080 0.163 8B4 0.082 0.171 4B5 0.263 0.461 0

K.H. Mo et al. / Construction and Building Materials 50 (2014) 499–507 505

3.6.2. Correlation between compressive energy and first crack impactstrength

The static compressive energy and the first crack impactstrength, which was due to the dynamic impact load showed sim-ilar behaviour and the addition of steel fibres enhanced both theseproperties. Fig. 6 shows the relationship between these two prop-erties. The Eq. (2) has been proposed to predict the first crack im-pact energy based on the linear correlation (R2 = 0.81) betweenthese two properties. The high compressive energy of the speci-mens led to high impact energy absorption of OPSC before thecommencement of initial cracks.

Eimpact;1st;cr ¼ 0:532ðEcompÞ þ 39:61 ð2Þ

where Eimpact,1st,cr = first crack impact energy in Joule (J); Ecomp =compressive toughness

3.6.3. Final impact strengthEffect of fibre bridging determines the post-crack impact energy

absorption and hence the impact ductility of concrete. The impactductile index (li), expressed as the ratio between the final and ini-tial impact energies and shown in Table 6 provides a good indica-tion to the ductility of the concrete subjected to impact load. It wasevident that the final impact energy of OPSC and FROPSC was sig-nificantly higher than the first crack impact energy. Even after theformation of the first cracks, the specimen was able to sustain largeamount of impact load before it failed. The final impact energy ofboth the OPSC and the FROPSC exceeded the published results onhigh strength concrete [30]. It implied that OPSC and FROPSC exhi-bit excellent impact resistance and they have high potential to beutilized as impact resistant members.

The effect of uncrushed OPS aggregates became more signifi-cant during the post-crack stage when subjected to impact load.Fig. 7 indicated that li of all crushed OPS mixes were significantlylower than the corresponding uncrushed OPS mixes. The finalimpact energy of OPSC and FROPSC with uncrushed OPS was25–183% higher than the corresponding mixes with the crushedOPS. The uncrushed OPS with low AIV had better shock absorbingcharacteristic compared to the crushed OPS. When the crackspropagated and encountered with the uncrushed OPS aggregates,more energy was required to force the cracks past the aggregates(Fig. 8) compared to the crushed aggregates. The highest value ofli of about 42 was recorded for the mix A3; however the valuewas almost halved for the corresponding mix, B3 with crushedOPS.

Table 6 shows that all the FROPSC produced the ductility indexwithin the range of 4.5–8 times compared to the control mixes.The highest final impact energy was observed for the hybrid fibrecombination of 0.9% steel fibre + 0.1% PP fibre, followed by 1%

506 K.H. Mo et al. / Construction and Building Materials 50 (2014) 499–507

steel fibre, 0.75% steel fibre + 0.25% PP fibre, 1% PP fibre and 0%fibre. In the hybrid system where 0.1% PP fibre was added with0.9% steel fibre, the shorter PP fibres were able to provide addi-tional crack bridging without causing much dispersion problem.This was more beneficial compared to addition of 1% steel fibrewhere difficulty in dispersion might lead to poor fibre bridging.Nevertheless, when steel fibre content was decreased to 0.75%and PP fibre increased to 0.25%, the impact energy absorbed uponfailure was reduced by 59% and 30% for the uncrushed andcrushed OPS, respectively. For the OPSC reinforced with 1% PP fi-bre, an obvious decrease in the final impact resistance was ob-served compared to the OPSC with the addition of steel fibresas shown in Fig. 7. This might be attributed to the short PP fibresas these fibres were not able to cross the crack region in the con-crete as effectively as the steel fibres; further, the PP fibres didnot have the stiffness or the anchorage to bridge cracks comparedto that of the hooked-end steel fibres. But both the PP-FROPSCwith uncrushed and crushed mixes produced higher final impactenergy compared to the control OPSC mixes. The final impactenergies absorbed by PP-FROPSC were five-times for the uncru-shed OPS and two-times for the crushed OPS, respectively higherthan the corresponding plain OPSC.

3.6.4. Failure patternThere were two distinct types of failure pattern observed in the

OPSC panel specimens. For the plain non-fibrous OPSC, the con-crete panel broke into four pieces upon failure (Fig. 9a). The OPSClost its structural geometry and integrity upon reaching the impactenergy capacity. However the failure of the FROPSC was due to per-foration of the panels by the drop weight hammer and the speci-men was not broken into pieces, unlike OPSC panels (Fig. 9b).This behaviour indicated that the FROPSC panels remained struc-turally integral, and also ductile.

Further, pulling out of fibres was observed for all the FROPSCpanels at failure as shown in Fig. 10. A number of secondary crackswere found to form around the vicinity of the point of impact inFROPSC panels.

3.6.5. Crack growth resistanceAll the crack widths were measured soon after the first crack

formation and during the propagation of cracks to the paneledge (Table 7). The initial crack width was used as a compara-tive study to determine the effectiveness of the fibres in bridgingmicro-cracks in the concrete. Initial crack width of the controlA1 and B1 OPSC was found to be around 0.160 mm. The increasein the steel fibre volume up to 1% led to reduced initial crackwidth whereas PP fibres were found ineffective in crack bridging.The reduction in the crack width in the FROPSC specimens wasfound up to 65% and 85% in the uncrushed and crushed OPSspecimens, respectively compared to the plain OPSC. The lowestcrack widths of 0.06 mm and 0.045 mm were found in the spec-imens with 1% of steel fibres of uncrushed (A2) and crushedspecimens (B2), respectively. On the other hand, the crackwidths of PP-FROPSC specimens were similar to that of the plainOPSC.

The higher stiffness and hooked end of steel fibres were effec-tive in providing resistance to cracks compared to PP fibres. Conse-quently, the increase in the steel fibre volume led to decrease inthe crack width. The comparison of the initial crack widths fromTable 7 shows that there was insignificant difference betweenthe control (A1 and B1) and the A5 and B5 specimens. However,the final crack widths of the A5 and B5 specimens are nearly halfof the respective control specimens (A1 and B1). Thus, it can beconcluded that though the PP fibres were less effective in arrestingthe crack width compared to steel fibres, these fibres were usefulin bridging macro-cracks compared to micro-cracks. Moreover,

secondary cracks began to appear just prior to the failure in theFROPSC specimens as shown in Fig. 9b; however, as seen fromFig. 9a the plain OPSC specimens had no secondary cracks uponfailure. The formation of the secondary cracks is an indication ofthe effect of fibres in arresting and prohibiting the crack growth.Thus, the initial cracks had not propagated further due to the for-mation of the secondary cracks. A large number of secondarycracks were formed in specimens with steel only and hybrid fibrescompared to control and PP only fibres. The specimen A3 had moresecondary cracks than the specimen A2 although the former hadlower steel fibre content. This might be attributed to the hybrid fi-bre used in the specimen A3 that had the optimum impact resis-tance as discussed earlier in Section 3.6.3.

4. Conclusion

The following conclusions can be drawn from the results of theexperimental study conducted:

i. The oven-dry densities of all the FROPSC specimens fellwithin the range of 1700–1940 kg/m3 and hence fulfilledthe requirement for LWC as stipulated in EN206-1.

ii. The OPSC specimens with crushed OPS aggregates producedlower slump values compared to those of uncrushed OPSaggregates due to the high mortar demand. The addition ofsteel fibre and high amount of PP fibre had negative effecton the slump values of FROPSC.

iii. The FROPSC mixes A2, A3, B2 and B3 with crushed anduncrushed OPS and 0.9% and 1% steel fibres produced highstrength OPSC of about 40 MPa. However, the usage of highdosage of PP fibres of 1.0% decreased the compressivestrength of FROPSC much lower than the control concrete.The long term compressive strength up to 180-day of allmixes showed no retrogression of strength.

iv. Though the OPSC specimens with crushed OPS aggregatesproduced higher compressive strength; its impact resistancewas lower compared to the corresponding mixes of theuncrushed OPS aggregates. However, once the fibres wereadded in the OPSC, the uncrushed OPS series producedhigher compressive strength due to the stronger bondbetween the fibres and the aggregates.

v. The 7-day UPV test results indicated that all the OPSC andFROPSC achieved the lower limit of 3.66 km/s stipulatedfor ‘good’ concrete.

vi. There was a good correlation between the compressivetoughness and the impact energy absorbed at the formationof the first impact crack. The higher the compressive tough-ness, the higher the first crack impact energy absorbed bythe concrete.

vii. All the FROPSC had higher first crack impact energy com-pared to the OPSC. The FROPSC with fibre content of 0.9%to 1% had the highest first crack impact energy due to themicro-crack bridging of the fibres.

viii. All the FROPSC specimens resisted high impact loads beforefailure and produced smaller crack widths, compared to theOPSC. The specimens with hybrid fibres (0.9% steel + 0.1%)produced the highest impact resistance.

ix. The uncrushed OPS aggregate produced higher final impactenergy due to its energy absorption capability compared tothe corresponding crushed OPS specimens. The crack widthsof the FROPSC specimens with steel fibres were found lowerthan the specimens with PP fibres. The impact energy distri-bution in the FROPSC specimens was clearly evident due tothe formation of more number of secondary cracks in thespecimens.

K.H. Mo et al. / Construction and Building Materials 50 (2014) 499–507 507

Acknowledgement

This research work was financially supported by University ofMalaya under High Impact Research Grant (HIRG) No. UM.C/HIR/MOHE/ENG/02/D000002-16001 (Synthesis of blast resistantstructures).

References

[1] Lok TS, Pei JS. Impact resistance and ductility of steel fibre reinforced concretepanels. HKIE Trans 1996;3(3):7–16.

[2] Nataraja MC, Dhang N, Gupta AP. Statistical variations in impact resistance ofsteel fiber-reinforced concrete subjected to drop weight test. Cem Concr Res1999;29:989–95.

[3] Basri HB, Mannan MA, Zain MFM. Concrete using waste oil palm shells asaggregate. Cem Concr Res 1999;29:619–22.

[4] Mannan MA, Ganapathy C. Engineering properties of concrete with oil palmshell as coarse aggregate. Constr Build Mater 2002;16:29–34.

[5] Mannan MA, Alexander J, Ganapathy C, Teo DCL. Quality improvement of oilpalm shell (OPS) as coarse aggregate in lightweight concrete. Build Environ2006;41:1239–42.

[6] Teo DCL, Mannan MA, Kurian VJ, Ganapathy C. Lightweight concrete madefrom oil palm shell (OPS): structural bond and durability properties. BuildEnviron 2007;42:2614–21.

[7] Alengaram UJ, Al-Muhit BA, Jumaat MZ. Utilization of oil palm kernel shell aslightweight aggregate in concrete – a review. Constr Build Mater 2013;38:161–72.

[8] Alengaram UJ, Jumaat MZ, Mahmud H, Shirazi SM. Effect of aggregate size andproportion on strength properties of palm kernel shell concrete. Int J Phys Sci2010;5(12):1848–56.

[9] Alengaram UJ, Jumaat MZ, Mahmud H. Ductility behavior of reinforced palmkernel shell concrete beams. Eur J Sci Res 2008;23(3):406–20.

[10] Alengaram UJ, Jumaat MZ, Mahmud H, Fayyadh MM. Shear behaviour ofreinforced palm kernel shell concrete beams. Constr Build Mater 2011;25:2918–27.

[11] Alengaram UJ, Mahmud H, Jumaat MZ. Enhancement and prediction ofmodulus of elasticity of palm kernel shell concrete. Mater Des 2011;32:2143–8.

[12] Shafigh P, Jumaat MZ, Mahmud H, Alengaram UJ. A new method of producinghigh strength oil palm shell lightweight concrete. Mater Des 2011;32:4839–43.

[13] Shafigh P, Jumaat MZ, Mahmud H. Oil palm shell as a lightweight aggregate forproduction high strength lightweight concrete. Constr Build Mater 2011;25:1848–53.

[14] Shafigh P, Jumaat MZ, Mahmud H, Hamid NBAA. Lightweight concrete madefrom crushed oil palm shell: tensile strength and effect of initial curing oncompressive strength. Constr Build Mater 2012;27:252–8.

[15] Zhang MH, Shim VPW, Lu G, Chew CW. Resistance of high-strength concrete toprojectile impact. Int J Impact Eng 2005;31:825–41.

[16] Chen B, Liu J. Contribution of hybrid fibers on the properties of the high-strength lightweight concrete having good workability. Cem Concr Res 2005;35:913–7.

[17] Banthia N, Yan C, Sakai K. Impact resistance of fiber reinforced concrete atsubnormal temperatures. Cem Concr Comp 1998;20:393–404.

[18] Ong KCG, Basheerkhan M, Paramasivam P. Resistance of fibre concrete slabs tolow velocity projectile impact. Cem Concr Comp 1999;21:391–401.

[19] Marar K, Eren O, Celik T. Relationship between impact energy and compressiontoughness energy of high-strength fiber-reinforced concrete. Mater Lett 2001;47:297–304.

[20] Erdem S, Dawson AR, Thom NH. Microstructure-linked strength properties andimpact response of conventional and recycled concrete reinforced with steeland synthetic macro fibres. Constr Build Mater 2011;25:4025–36.

[21] Nataraja MC, Nagaraj TS, Basavaraja SB. Reproportioning of steel fibrereinforced concrete mixes and their impact resistance. Cem Concr Res 2005;35:2350–9.

[22] Mohammadi Y, Carkon-Azad R, Singh S, Kaushik S. Impact resistance of steelfibrous concrete containing fibres of mixed aspect ratio. Constr Build Mater2008;23:183–9.

[23] Nili M, Afroughsabet V. Combined effect of silica fume and steel fibers on theimpact resistance and mechanical properties of concrete. Int J Impact Eng2010;37:879–86.

[24] Alavi Nia A, Hedayatian M, Nili M, Afroughsabet V. An experimental andnumerical study on how steel and polypropylene fibers affect the impactresistance in fiber-reinforced concrete. Int J Impact Eng 2012;46:62–73.

[25] Alhozaimy AM, Soroushian P, Mirza F. Mechanical properties of polypropylenefiber reinforced concrete and the effects of pozzolanic materials. Cem ConcrComp 1996;18:85–92.

[26] Wang N, Mindess S, Ko K. Fibre reinforced concrete beams under impactloading. Cem Concr Res 1996;26:363–76.

[27] Toutanji H, McNeil S, Bayasi Z. Chloride permeability and impact resistance ofpolypropylene-fiber-reinforced silica fume concrete. Cem Concr Res 1998;28:961–8.

[28] Song PS, Hwang S, Sheu BC. Strength properties of nylon- and polypropylene-fiber-reinforced concretes. Cem Concr Res 2005;35:1546–50.

[29] Deng Z, Li J. Mechanical behaviors of concrete combined with steel andsynthetic macro-fibers. Int J Phys Sci 2006;1(2):57–66.

[30] Yew MK, Othman I, Yew MC, Yeo SH, Mahmud H. Strength properties of hybridnylon-steel and polypropylene-steel fibre-reinforced high strength concrete atlow volume fraction. Int J Phys Sci 2011;6(33):7584–8.

[31] Banthia N, Nandakumar N. Crack growth resistance of hybrid fiber reinforcedcement composites. Cem Concr Comp 2003;25:3–9.

[32] Shafigh P, Mahmud H, Jumaat MZ. Effect of steel fiber on the mechanicalproperties of oil palm shell lightweight concrete. Mater Des 2011;32:3926–32.

[33] Yap SP, Alengaram UJ, Jumaat MZ. Enhancement of mechanical properties inpolypropylene- and nylon-fibre reinforced oil palm shell concrete. Mater Des2013;49:1034–41.

[34] Qian CX, Stroeven P. Development of hybrid polypropylene-steel fibre-reinforced concrete. Cem Concr Res 2000;30:63–9.

[35] Bentur A, Mindess S. Fibre reinforced cementitious composites. London,UK: Elsevier Applied Science; 1990.

[36] Libre NA, Shekarchi M, Mahoutian M, Soroushian P. Mechanical properties ofhybrid fiber reinforced lightweight aggregate made with natural pumice.Constr Build Mater 2011;25:2458–64.

[37] European Committee for Standardization. Concrete – Part 1: specification,performance, production and comfirmity. European Standard CEN, EN206-1;2000.

[38] Duzgun OA, Gul R, Aydin AC. Effect of steel fibers on the mechanicalproperties of natural lightweight aggregate concrete. Mater Lett 2005;59:3357–63.

[39] Yao W, Li J, Wu K. Mechanical properties of hybrid fiber-reinforced concrete atlow fiber volume fraction. Cem Concr Res 2003;33:27–30.

[40] Adeen MB, Al-Attar AA, Ra’ouf SM. Determination of mechanical properties ofhybrid steel-nylon fiber reinforced concrete. Mod Appl Sci 2010;4(12):97–108.

[41] Gao JM, Sun W, Morino K. Mechanical properties of steel fiber-reinforced,high-strength, lightweight concrete. Cem Concr Comp 1997;19:307–13.

[42] Campione G, Miraglia N, Papie M. Mechanical properties of steel fibrereinforced lightweight concrete with pumice stone or expanded clayaggregates. Mater Struct 2001;34(4):201–10.

[43] Solis-Carcano R, Moreno EI. Evaluation of concrete made with crushedlimestone aggregate based on ultrasonic pulse velocity. Constr Build Mater2008;22(6):1225–31.

[44] Leslie JR, Cheesman WJ. An ultrasonic method of studying deterioration andcracking in concrete structures. ACI J Proc 1949;46(1):17.

[45] Teo DCL, Mannan MA, Kurian VJ. Structural concrete using oil palm shell (OPS)as lightweight aggregate. Turk J Eng Environ Sci 2006;30:1–7.