Engineering properties and fracture behaviour of high ... · Engineering properties and fracture behaviour of high ... The concrete with high compressive strength ... geopolymer lightweight

Construction and Building Materials 111 (2016) 286–297

Contents lists available at ScienceDirect

Construction and Building Materials

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

Engineering properties and fracture behaviour of high volume palm oilfuel ash based fibre reinforced geopolymer concrete

http://dx.doi.org/10.1016/j.conbuildmat.2016.02.0220950-0618/� 2016 Elsevier Ltd. All rights reserved.

⇑ Corresponding authors at: Department of Civil Engineering, Faculty ofEngineering, University of Malaya, Kuala Lumpur 50603, Malaysia.

E-mail addresses: [email protected] (I.I. Bashar), [email protected](U.J. Alengaram).

Iftekhair Ibnul Bashar a,b,⇑, U. Johnson Alengaram a,b,⇑, Mohd Zamin Jumaat a,b, Azizul Islam a,Helen Santhi c, Afia Sharmin a

aDepartment of Civil Engineering, Faculty of Engineering, University of Malaya, Kuala Lumpur 50603, MalaysiabCentre for Innovative Construction Technology, Department of Civil Engineering, Faculty of Engineering, University of Malaya, Kuala Lumpur 50603, Malaysiac School of Mechanical and Building Sciences, VIT University, India

h i g h l i g h t s

� High volume POFA (90%) was used in development of geopolymer concrete.� Aspect ratio of steel fibre had significant effect on tensile properties.� Tensile properties due to steel fibres are more significant in OPSGC than NWGC.� Bonding between matrix and aggregate influenced tensile and fracture properties.

a r t i c l e i n f o

Article history:Received 19 May 2015Received in revised form 5 February 2016Accepted 6 February 2016Available online 23 February 2016

Keywords:Palm oil fuel ashMetakaolinGeopolymer concreteFibre reinforced concreteAspect ratioFracture energyFlexural toughness

a b s t r a c t

This article reports investigation carried out using high volume palm oil fuel ash (POFA-90%) basedgeopolymer concrete using oil palm shell (OPS) as coarse aggregate. The tensile and fracture propertiesof OPS based lightweight geopolymer concrete (OPSGC) with the addition of steel fibres (SF) of two aspectratios (AR80, AR65) and three volume fractions (0.25%, 0.50%, 0.75%) were investigated. The results werecompared with that of normal weight geopolymer concrete (NWGC) with and without steel fibre of AR80and AR65 (0.50% volume fraction). The higher values of flexural and splitting tensile strengths of OPSGCcould be attributed to stronger bond between the rougher surfaces of the crushed OPS and matrix. Theaddition of SF(AR80) produced higher splitting strength, flexural strength and total fracture energy of5%, 6% and 50–80%, respectively compared to the corresponding values of SF(AR65). The toughness andequivalent flexural strength ratio of OPSGC were found higher than the corresponding values of NWGCand this could be attributed to the ductility of OPS. The values of residual load and residual strengthin two-deflection limits of L/600 and L/150 indicated the progressive failure, which reflected the ductilityof the OPSGC with fibres.

� 2016 Elsevier Ltd. All rights reserved.

1. Introduction

Geopolymer concrete (GC) could be an alternative of ordinaryPortland cement concrete (OPC) [1]. The greenhouse effect by theemission of CO2 from the calcination during cement productioncan be relegated through the application of geopolymer technologyin the construction industry. The production of one tonne ofcement directly generates 0.55 tonnes of chemical CO2 andrequires the combustion of carbon fuel to yield an additional

0.4 tonnes of CO2. Conversely, there is no CO2 production in thegeopolymerisation process. The polymeric reaction between silicaand alumina exploits in geopolymer through the use of alkalineactivators. The alkalinity of the activator can be low to mild orhigh. The main contents in geopolymerisation process are the sili-con and the aluminium. The binder can be produced by a poly-meric synthesis of the alkali-activated material from eithergeological origin or any by-products consisting of silica and alu-mina as known as pozzolanic material. Hence, the application ofgeopolymer technology could also be advantageous in term ofmaking use of waste by-product materials from industry.

Palm oil fuel ash (POFA) is one of the largest industrialpozzolanic by-products in the South-East Asia. The abundance andavailability of POFA created an ideal platform for the researchers

I.I. Bashar et al. / Construction and Building Materials 111 (2016) 286–297 287

to work on this pozzolanic material as source material in the devel-opment of cleaner and sustainable material like GC. Researchworks on POFA-based geopolymer concrete are limited and theincorporation of metakaolin (MK) along with POFA is another ave-nue for research towards utilisation of local waste materials in thedevelopment of GC.

Besides the environmental pollution due to the production ofcement, the consumption of large quantity of natural sand andcoarse aggregates (47 billion tonnes/year [2]) has led to a decreasein ground water levels and natural disaster in many countries. Inorder to ensure sustainable development, researchers, from allaround the world have focused their research on applying wastematerials to replace the conventional materials [3–6]. Recently,one industrial by-product from the quarry industry as known asmanufactured sand (M-sand) and another agro-waste materialnamed oil palm shell (OPS) have drawn the attention of manyresearchers in the production of lightweight concrete and as fullreplacement of mining sand [7–9].

Balamurugan and Perumal [10], Ji et al. [11] and Raman et al.[12] reported the potentiality of using quarry fines as a replace-ment of river/mining sand. The quarry industries produce millionsof tons of wastes in the form of quarry dust (QD). These wastes aredumped in the factory yards and hence reuse of QD might help inreducing the overuse of mining and quarrying. The sophisticatedtechnology known as Vertical Shaft Impact Crusher System allowsQD to be centrifuged to remove flaky and sharp edges. The end pro-duct is commonly known as manufactured sand (M-Sand) and it ispopular in some of the developing countries [13].

Apart from the initiative to utilise M-sand to wholly replaceconventional sand, OPS has been used to replace conventionalcrushed granite aggregate. Researchers [14–16] explored the suit-ability of OPS as lightweight aggregate and found that structuralgrade lightweight concrete could be produced using OPS as coarseaggregate [17,18]. During the last three decades, many researchworks have been carried out using OPS in OPC concrete as light-weight aggregate to replace conventional crushed granite aggre-gate [8,15,19–23]; Yap [14] reported the possibility of significantcost saving due to density reduction as the OPS concrete has about17–25% lesser density compared to conventional normal weightconcrete.

In this research work, high volume of POFA combined withM-sand and OPS as fine and coarse aggregates, respectively wereused in the development of a cleaner GC. Like OPC, GC was foundgood in compressive resistance and weak in tensile properties. Itsweakness in tensile resistance could be overcome by using thesteel fibres [24,25]. The concrete with high compressive strengthshows brittleness characteristics due to the low tensile strengthwhich affects a weak bond in the transition zone of the cementmatrix [26,27]. The bond weakness due to low tensile strength ofconcrete could be mitigated by adding steel fibres in concreteand the use of single type of fibres could improve the tensile prop-erty of concrete [28–30]. Knight et al. [31] reported the contribu-tion of fibres in the enhancement of strength and toughnessretention. The usage of steel fibres in concrete also enhances thetoughness as well as impact resistance of concretes [32]. Regard-less the improvement of tensile properties of concrete using steelfibres, a weak transition zone between steel fibres and paste wasobserved by the researchers [33,34] and this weak bonding wascaused due to a lot of porosity in the transition zone. By using

Table 1Chemical composition of POFA and MK (%).

Chemical compounds CaO SiO2 Al2O3 MgO

POFA 5.57 67.72 3.71 4.04MK 0.04 52.68 42.42 0.12

appropriate material, the porosity could be reduced as well as consol-idate the transition zone [35–37]. Madhkhan et al. [38] suggested theusage of pozzolanic material to reduce the porosity in concrete.

The mechanical properties of GC using conventional crushedgranite aggregate were reported [39–41]. However, the use ofPOFA and MK as binders and OPS and M-sand as coarse and fineaggregates in GC is entirely a new area of research; The use of highvolume of POFA of 90% and MK as source materials in GC wouldenable researchers to entirely utilise POFA as sole binder in thedevelopment of GC in future; further, the utilisation of M-sandand OPS as fine and coarse aggregates would pave way for thedevelopment of lightweight concrete as OPS is considered as light-weight aggregate [3,7,20,42]. On the other hand, the higher aggre-gate impact resistance of OPS would enable to test the ability ofOPS to resist other engineering properties such as fracture. In addi-tion, to overcome the brittle behaviour of GC, the effects of steelfibres with two different aspect ratios have been investigatedand reported through this research work. Ten concrete mixes wereprepared and tested for mechanical properties and fracture beha-viour of the hardened OPS geopolymer concrete (OPSGC) and theexperimental results were compared with normal weight geopoly-mer concrete (NWGC). The fracture behaviours investigatedthrough four-point bending test include fracture strength, fracturetoughness and peak deflection.

The novelty of this research work lies in the development ofgeopolymer lightweight concrete using high volume POFA, withlow content of MK; the addition of two local waste materials-M-sand and OPS as fine and coarse aggregates is another aspect ofnovelty. This is the first time the effect of M-sand and OPS in GCwas investigated. The enhancement of ductile properties of GCby replacing crushed granite aggregates by OPS and by the additionof steel fibres with two aspect ratios and different volume fractionsis another innovative aspect of this research work.

2. Materials and methods

2.1. Materials

The test result based on X-ray fluorescence (XRF) analysis of both POFA and MKis shown in Table 1 and it shows that both contained more than 70% of SiO2, Al2O3 &Fe2O3 and hence both could be categorised as Class F in accordance with ASTMC618. Fig. 1 represents the particle size distribution of POFA and MK. The physicalproperties of POFA and MK are shown in Table 2 and it was found that MK was finerand more uniformly graded than POFA. The percent of POFA and MK passingthrough 45 lm sieve were recorded as 95% and 80%, respectively (Fig. 1) whichcomplies with the recommendation by ASTM C618-12a. M-sand of Grade C (particlesize grading) was used (Fig. 2, Table 3). Lightweight crushed OPS were used as thewhole replacement of conventional crushed granite aggregate. The fineness modu-lus, specific gravity, water absorption and aggregate impact value (AIV) of OPS andgranite were determined based on ASTM C127-12, ASTM C131-06 and ASTM C136-06. The 24 h water absorption of OPS was found 25% (Table 4); since the AIV of OPSwere 3 times lower than that of granite, it has been proved that OPS has goodimpact resistance due to natural fibres. OPS were soaked in water for 24 h in waterbefore being used for casting. Hooked-end type steel fibres of length 60 mm and35 mm with aspect ratio 80 and 65, respectively were used. The steel fibres had aminimum tensile strength of 1100 MPa as specified by the manufacturer. Potablewater was used for all concrete casting.

2.2. Preparation of Alkaline solution and mix design

The alkaline activator using 14 M sodium hydroxide (NaOH) and liquid sodiumsilicate (Na2SiO3, ratio of SiO2/Na2O = 2.5) was prepared at least 24 h before thecasting. The sodium silicate liquid was mixed with 14 M NaOH solution by weightproportion of 1:2.5.

Na2O SO3 K2O Fe2O3 LoI

0.16 1.07 7.67 4.71 6.200.07 0.05 0.34 2.02 1.40

0

10

20

30

40

50

0

20

40

60

80

100

0.01 0.1 1 10 100 1000

% b

y V

olum

e re

tain

ed

% b

y V

olum

e pa

ssin

g

Particle Size µm

% by Volume passing POFA% by Volume passing MK% by Volume retaining POFA% by Volume retaining MK

Fig. 1. Particle size distribution of POFA and MK.

Grade CUpper limit

0

20

40

60

80

100

0.1 1 10

% P

assi

ng

Sieve Size (mm)

Grade C based on BS EN 933-8-2012Fineness Modulus: 2.88, 2.66, 2.67 for N-sand, M-sand & QD respectively

Grade CLower limit

Fig. 2. Particle size distribution curve of M-sand.

288 I.I. Bashar et al. / Construction and Building Materials 111 (2016) 286–297

Steel fibres of aspect ratio (AR) 65 and 80 were added with OPSGC in a rate of0.25%, 0.50% and 0.75% and with NWGC in a rate of 0.50% as a control mix (Table 5).

2.3. Casting, specimen preparation, curing and testing of fresh and harden concrete

The binders were mixed with the aggregates and steel fibres in the drum mix-ture of capacity 0.2 m3 (Fig. 3). The alkaline activator was then added and mixedproperly and then the additional water was added to the mixture. The fresh densityof the concrete was then measured. The specimens were cured at 65 �C temperaturein curing chamber for 48 h after 3 h of casting and then kept in room temperature of28 �C and relative humidity of 79%. Three specimens were prepared for each testand the average values are reported. The dimensions of specimens and testing stan-dards are shown in Table 6. The fracture testing arrangement is shown in Fig. 4while Fig. 5 shows the test specimen with the groove.

3. Results and discussion

3.1. Workability and density of concrete

The slump value is one of the important characteristics of theworkability of fresh concrete. Though there is no code of practiceavailable for workability tests for geopolymer concrete [43], slumptest was carried out to check the workability of the mixes. Asexpected, zero slump was obtained for the mixes; one of the rea-sons attributed to zero slump is the use of large quantity of POFAas binder; since POFA has high loss on ignition value it will absorbmore water and hence could reduce the workability. Further highsilica content in POFA makes the mix more viscous. Generallygeopolymer concretes with high binder content such as POFA orGGBFS result in low slump. Further, OPS is lightweight and hencezero slump was expected, nevertheless using appropriate vibra-tion, the concrete was compacted and upon demoulding no honey-comb was found.

In normal Portland cement concrete the addition of water orsuperplasticiser enhances the workability of concrete. On the con-trary, the usage of excessive water decreases the rate of geopoly-merisation [44,45] as well as the strength of concrete. Therefore,

Table 2Physical properties of POFA and MK.

Physical properties Colour S.G. S.S.A. D [4

POFA Black 2.2 1.72 20.0MK Off-white 2.5 4.33 10.9S.G. = Specific gravity

S.S.A. = Specific surface area, m2/gD [4,3] = Volume moment mean, lmD [3,2] = Surface area moment mean, lm

D (vD (vD (v

minimum required water was used in consistent rate and thisattributed to zero slump.

The fresh densities of all mixes are shown in Table 7. The differ-ence among the fresh densities of OPSGC with and without steelfibres is solely attributed to the proportion of additional steel fibresas all other material proportions were kept constant. As shown inTable 5, three mixes of NWGC-one without steel fibres and twomixes with steel fibres of AR 65 and 80 of 0.5% (volume fraction)were used for comparison. As expected, higher specific gravity ofcrushed granite aggregate resulted in higher fresh density ofNWGC compared to OPSGC. The fresh densities of all OPSGCs withsteel fibres fulfilled the requirement for lightweight aggregate con-crete as defined in BS EN 1992-1-1:2004 and the variation of thedensities among the mixes was due to different proportions ofsteel fibres.

3.2. Compressive strength

The compressive strengths development at the ages of 3-, 7-,14- and 28-day are shown in Table 7. The early strength develop-ment was found 82–97% and 85–98% of the 28-day compressivestrength at the ages of 3- and 7-day, respectively for all geopoly-mer concrete with or without fibres and this could be attributedby geopolymerisation process during heat curing [46]. The28-day compressive strength results show that OPSGC achieved 28-day compressive strength of 30 MPa and according to ACI 213R-14,this strength is within the range for structural grade concrete.

Despite the addition of steel fibres in the mixes, the increase incompressive strength was found minimum. In addition, thevolume of steel fibres and the aspect ratio (AR) did not have anysignificant influence on the compressive strength as the 28-daycompressive strengths of all these OPSGC mixes were found within30–32 MPa (Table 7). Though research works [47,48] showed thathigher AR provided better performance, it was not evident in the

,3] D [3,2] D (v, 0.1) D (v, 0.5) D (v, 0.9)

6 3.49 3.23 18.46 38.275 1.39 0.48 4.26 32.67, 0.1) = 10% of particle by volume below this size, 0.5) = 50% of particle by volume below this size, 0.9) = 90% of particle by volume below this size

Table 3Physical properties of M-sand.

Specific gravity Absorption (%) Fineness modulus D10, mm D30, mm D60, mm Cu = D60/D10 Cc = D302 /(D10 � D60)

OD SSD

1.97 2.63 0.91 2.66 0.30 0.55 1.42 4.80 0.73

OD = Oven Dry; SSD = Saturated Surface Dry; Cu = Uniformity coefficient; Cc = Coefficient of curvature.

Table 4Physical properties of coarse aggregate.

Coarse aggregates Fineness modulus Specific gravity Water absorption, % Aggregate impact value, %

Conventional granite aggregate 7.43 2.67 <1 11.9Crushed OPS 6.25 1.32 24.74 3.93

Table 5Mix design with variables of steel fibres proportion and aspect ratio.

Concrete designation POFA MK M-sand OPS Gravel Steel fibres*-65/35 Steel fibres*-80/60 Alkaline/binder Water/binder

Proportion (kg/m3) by weight

OPSGC 508 56 636 212 – – – 0.5 0.11OPSGC65/0.25 508 56 636 212 – 19.75 (0.25) –OPSGC80/0.25 508 56 636 212 – – 19.75 (0.25)OPSGC65/0.5 508 56 636 212 – 39.50 (0.5) –OPSGC80/0.5 508 56 636 212 – – 39.50 (0.5)OPSGC65/0.75 508 56 636 212 – 59.25 (0.75) –OPSGC80/0.75 508 56 636 212 – – 59.25 (0.75)NWGC 510 57 380 – 753 – – 0.4 0.2NWGC65/0.5 510 57 380 – 753 39.50 (0.5) –NWGC80/0.5 510 57 380 – 753 – 39.50 (0.5)

* Proportions in percent in bracket ().

Fig. 3. Casting of geopolymer fibre reinforced concrete.

Table 6Specimens for testing of concrete.

Test (age) – standards Specimens and dimensions (mm)

Compressive strength test (3-, 7-, 14- & 28-day) - BS EN 12390-3:2009 Cubes: 100 mm (concrete); 50 mm (mortar)Indirect tensile strength test (28-day) – BS EN 12390-6:2009 Cylinders: /100 mm � 200 mm heightFlexural strength test (28-day) – BS EN 12390-5:2009 Prisms: 100 mm � 100 mm � 500 mmStatic modulus of elasticity (28-day) – ASTM C469 – 14 Cylinders: /150 mm � 300 mm heightFracture toughness test (28-day) – ASTM C1609/C1609 M-12 Prism of 100 mm � 100 mm � 500 mm with 30 mm groove

I.I. Bashar et al. / Construction and Building Materials 111 (2016) 286–297 289

mixes with AR 80 and AR 65. Mo et al. [49] related the occurrenceof micro-crack in the lightweight aggregate during loading as anexplanation of how the steel fibres helped improving compressivestrength. It is known that lightweight aggregate has less aggregatecrushing value than normal weight aggregate. Therefore, cracksoccurred in the coarse aggregate before extending into the mortar[49].

Gao [47] reported that a sufficient bonding between steel fibresand binding material increased the compressive strength of light-weight concrete. However, insignificant improvement of compres-sive strength was noticed for OPSGC and NWGC with steel fibres inthis research work. This could be due to the porosity in geopolymerconcrete which was induced by the viscous geopolymer bindingmatrix.

P/2P/2

Vertical LVDT

Fig. 4. Fracture test according to ASTM C1609/C1609M-12.

Fig. 5. Specimens for fracture test.

290 I.I. Bashar et al. / Construction and Building Materials 111 (2016) 286–297

OPSGC with steel fibres of AR 65 (OPSGC-65) achieved slightlyhigher compressive strength than OPSGC 80 for the same volumeof steel fibres in concrete. Steel fibres of AR 65 (SF(AR65)) had 3.33times more surface area than same volume of SF(AR80). The largesurface area of steel fibres might enhance the bond resistanceand crack propagation due to stiffness of fibres [48]. However, areverse result was noticed for NWGC; NWGC-80/0.50 achievedslightly higher compressive strength than NWGC-65/0.50. Themortar quantity to total surface area of steel fibres and aggregates

Table 7Development of compressive strength.

Mix designation Density (kg/m3) Cube compressive stre

Fresh 28-day 3-day 7-day

OPSGC 1830 1782 28.7 29.0OPSGC-65/0.25 1845 1816 28.3 30.6OPSGC-80/0.25 1845 1811 28.0 29.4OPSGC-65/0.5 1880 1819 26.8 29.0OPSGC-80/0.5 1860 1815 26.1 27.2OPSGC-65/0.75 1890 1814 30.5 30.9OPSGC-80/0.75 1875 1813 29.2 28.5NWGC 2045 2020 26.4 27.0NWGC-65/0.5 2084 2022 27.9 28.2NWGC-80/0.5 2084 2023 30.3 30.5

Table 8Flexural, indirect tensile strengths and modulus of elasticity.

Mix designation 28-day density(kg/m3)

28-day Compressivestrength (MPa)

Flexural strength(MPa) (% of 28-daycompressive streng

OPSGC 1782 30.00 4.33 (14.43)OPSGC-65/0.25 1816 31.35 4.69 (14.96)OPSGC-80/0.25 1811 29.96 4.65 (15.52)OPSGC-65/0.5 1819 30.90 4.72 (15.28)OPSGC-80/0.5 1815 31.86 4.80 (15.07)OPSGC-65/0.75 1814 31.30 4.86 (15.53)OPSGC-80/0.75 1813 30.54 5.14 (16.83)NWGC 2020 27.62 3.41 (12.35)NWGC-65/0.5 2022 29.17 3.61 (12.38)NWGC-80/0.5 2023 31.67 3.67 (11.59)

might have a relation that altered the result. As seen in Table 5, thetotal amount of mortar ingredients (quantity of binder and fineaggregate) was 1.27 times less in NWGC than OPSGC.

Compressive strength and density are related to each other.Higher value of the ratio of compressive strength to density(fcu/d) ratio was found for lightweight concrete compared to nor-mal concrete. As seen from Table 7, OPSGC, OPSGC-80/0.25 andNWGC-65/0.50 achieved compressive strength of about 30 MPa,but the strength to density ratios of OPSGC were found higher thanthe NWGC due to high density of the latter.

3.3. Flexural strength (ff)

The steel fibres play an important role in the development offlexural strength and its significance in the development of flexuralstrength is more than that in the development of compressivestrength. The experimental values of flexural strengths of OPSGCand NWGC with different volumes of steel fibres and AR 65 and80 are shown in Table 8. The increment of flexural strength dueto the addition of steel fibres was noticed in the range of 7–18%in OPSGC. The highest flexural strength of 5.14 MPa was achievedfor OPSGC-80/0.75. The improvement of flexural strength due tothe increase in steel fibres was noticed for both mixes with AR65 and 80; further, this improvement was found higher in the caseof AR 80 compared to that of AR 65.

OPSGC-65/0.25 achieved about 8% more flexural strength(4.69 MPa) compared to OPSGC (4.33 MPa); with further increasein the fibres volume (OPSGC-65/0.50 and OPSGC-65/0.75), theincrement was found higher (9% and 12%). Thus, the volume ofsteel fibres added in the mixes had direct influence on the flexuralstrength (Fig. 6). Higher volume of steel fibres achieved higher flex-ural strength. Gencel et al. [50] studied the mechanical perfor-mance of fibre-reinforced self-compacting concrete with fly ash

ngth, fcu (MPa) 28-day Compressive strength todensity ratio [(N/m2)/(kg/m3)]

14-day 28-day

30.0 30.0 16.8429.4 31.4 17.2628.5 30.0 16.5430.2 30.9 16.9931.9 31.9 17.5526.8 31.3 17.2530.5 30.5 16.8527.6 27.6 13.6729.1 29.2 14.4231.1 31.7 15.65

th)

Indirect tensile strength(MPa) (% of 28-daycompressive strength)

Modulus ofelasticity (GPa)

Flexural/indirecttensile strength

Experiment Eq. (1)

2.22 (7.40) 6.25 NA 1.952.81 (8.96) 6.56 NA 1.672.55 (8.51) 8.68 7.64 1.822.88 (9.32) 8.69 NA 1.642.89 (9.07) 9.10 8.72 1.662.93 (9.36) 8.72 NA 1.663.07 (10.05) 10.00 7.97 1.672.31 (8.36) 22.35 NA 1.482.62 (8.98) 22.70 NA 1.382.60 (8.21) 23.21 NA 1.41

ff,65 = 0.34(%S) + 4.5867R² = 0.8775

ff,80 = 0.98(%S) + 4.3733R² = 0.9523

4.4

4.6

4.8

5

5.2

1 1.4 1.8 2.2 2.6 3

Flex

ural

stre

ngth

, MPa

% volume steel fibre, %SAR65 AR80Linear (AR65) Linear (AR80)

Fig. 6. Development of flexural strength for different % volume of steel fibres.

I.I. Bashar et al. / Construction and Building Materials 111 (2016) 286–297 291

and reported the significance of steel fibres in the improvement offlexural strength of concrete.

The influence of steel fibres in concrete and the volume of steelfibres could be understood from the mechanism of flexuralstrength behaviour of concrete prism under loading. The concreteprism specimens were subjected to bending during flexuralstrength test; the tensile and compressive strength in prism sec-tion depend on the position of the neutral axis [51–53]. Gencel[50] reported that the compression phase of the prismwas stressedabout 11–17% of the compressive strength of the material at thefailure loading. Therefore, compression zone had an insignificantrole to failure. By contrast, tension zone played role in the develop-ment of flexural strength. Randomly distributed steel fibres con-trolled the tension zone by withstanding crack generation andthus enhanced the flexural strength [54]. High volume of steelfibres might withstand more crack generation in tension zoneand thus enhanced the flexural strength of concrete.

The aspect ratio (length/diameter) of steel fibres haddirect influence on the development of flexural strength. The flex-ural strength enhancement of 8–12% and 7–18% was found forAR65 and AR80, respectively with 0.25%, 0.5% and 0.75% of steelfibres volume. As seen from Fig. 6, the mixes with AR 80 showedhigher gradient than the corresponding mixes with AR 65. A com-parison between mixes with same volume fraction, but differentAR shows that mixes with higher AR produced higher flexuralstrength. Since AR is the ratio of length to diameter of the steelfibres, AR80 has more surface area than AR60 and this would influ-ence stronger bond between the steel fibres and binding matrix byimproving the stiffness to resist crack propagation. Other researchworks [47,55] also support this finding and Köksal [55] investi-gated the effectiveness of higher AR and concluded that longersteel fibres had higher capability of delaying crack propagation inconcrete.

Bond failure on OPS 1

2 3

4

Fig. 7. Bond failure between OPS

NWGC produced the lowest flexural strength among all the mixproportions. On contrary, OPSGC performed better flexuralstrength. The improvement of flexural strength due to the OPSwas reported by other research work [56]. As seen in Table 8,NWGC and OPSGC achieved flexural strength of 12.35% and14.43% of 28-day compressive strength. Alengaram [42] reportedthese values for cement based normal weight and OPS concreteas 14% and 10%, respectively. According to Alengaram [42] andShafigh et al. [57], percent of flexural strength of 28-day compres-sive strength was more in normal weight concrete than OPS con-crete. The slight reduction of flexural strength in OPS basedconcrete was due to lower strength and stiffness of OPS [42]. How-ever, as seen in current research work, the reverse result might beaffected by the volume of M-sand content in OPSGC and NWGC.OPSGC which contained more volume of M-sand than as mixedin NWGC, might have better compactability by the fine particlesof M-sand and thus enhanced the percent of flexural strength of28-day compressive strength. Alengaram [21] also found that theincrement of fine aggregate content enhanced the flexural perfor-mance of concrete.

Generally, the failure in tension occurs due to the breakdown ofbond between the matrix and the surface of the aggregate or byfracture of the matrix itself and not by fracture of the aggregate.Since there is less evidence on aggregate fracture compared tobond failure between binding matrix and aggregate (Fig. 7), thebond failure govern the flexural strength.

3.4. Indirect tensile strength (Splitting tensile strength, ft)

The splitting tensile test, which is known as the indirect tensilestrength test, provides the maximum tensile load which may beinduced on the concrete member before cracking. The results onthe effects of volume and aspect ratio of steel fibres in the develop-ment of indirect tensile strength (ITS) are shown in Table 8.

The addition of steel fibres increased ITS of both OPSGC andNWGC and this was supported by previous research works[58,59]. The increments of ITS of OPSGC due to the addition of0.25, 0.50 and 0.75% of fibres were found as 27–32% and 15–38%,for mixes with AR of 65 and 80, respectively. The mix OPSGC-80achieved higher ITS than OPSGC-65 for 0.5% and 0.75% steel fibresvolume. The load that induced overstress in non-fibre reinforcedconcrete and led to develop crack and substantial failure in con-crete, would withstand further crack propagation due to the addi-tion of steel fibres [50]. Wang and Wang [58] reported theimprovement of ITS up to 92.50% due to addition of fibres from0% to 2% of volume.

The mechanism of crack propagation in fibre-reinforced con-crete could be explained from Fig. 8 [60,61]. Once crack propagated

1 2

4

3

surface and binding matrix.

Fig. 8. Fibres avoidance mode of crack propagation; schematically reproduced from[60,61].

292 I.I. Bashar et al. / Construction and Building Materials 111 (2016) 286–297

in the matrix, the effective stiffness was reduced due to the frac-ture [61]. The tensile stress was then transferred through steelfibres. The steel fibres bridged micro cracks and prevented furtherpropagation to macro cracks as well as expansion and thus sub-stantially improved the ITS.

The ITS values of both mixes-NWGC-65/0.50 and NWGC-80/0.50 show an increase of about 13% compared to the corre-sponding control concrete-NWGC. The ratios between the 28-dayITS and compressive strength for OPSGC and NWGC were foundas 7% and 8%, respectively. The highest ITS in OPSGC-80/0.75 couldbe attributed to both steel fibres and increase in volume of steelfibres. The ratio of flexural to splitting tensile strength (ff/ft) wasfound about 1.67 for OPSGC-65 and OPSGC-80.

Fig. 9 represents a relationship between ITS and percentage ofsteel fibres volume with AR of 65 and 80. The intersection betweenthe slopes of AR65 and 80 as shown in Fig. 9 shows that the ITS ofabout 2.9 MPa could be achieved by the addition of about 0.55%steel fibres volume.

3.5. Static modulus of elasticity

The Young’s modulus of elasticity (MoE) is shown in Table 8 forthe different volume of steel fibres addition in OPSGC and NWGC.The MoE of OPSGC was 3.5 times less than NWGC. The highestvalue of MoE in OPSGC was found for OPSGC-80/0.75 as 10 GPa.It is noted that flexural and indirect tensile strength also showedthe highest result for OPSGC-80/0.75. The addition of steel fibres

ft,65 = 0.24(%S) + 2.7533R² = 0.9908

ft,80 = 1.04(%S) + 2.3167R² = 0.9694

2.5

2.6

2.7

2.8

2.9

3

3.1

3.2

0.25 0.35 0.45 0.55 0.65 0.75

Indi

rect

tens

ile st

reng

th, M

Pa

% volume steel fibre, %S

AR65 AR80

Fig. 9. Development of indirect tensile strength for different % volume of steelfibres.

improved the MoE 4–60% and 1–4% of OPSGC and NWGC, respec-tively; higher volume of steel fibres in concrete resulted higherMoE. Kurugöl et al. [62] reported similar outcome that additionof steel fibres increased the MoE.

A comparison of volume of steel fibres and MoE is shown inFig. 10. MoE increased with increment of % volume of steel fibresin both AR65 and 80. Mo [49] proposed following Eq. (1) to corre-late the MoE with the compressive strength of fibre reinforced OPSbased lightweight concrete containing steel fibres of AR80.

E ¼ 6:31ffiffiffiffiffiffif cu

q� 26:90 ð1Þ

A comparison between the experimental result and the aboveequation is shown in Table 8. The deviation of the estimated resultcould be due to the usage of different proportion of OPS and bind-ing material.

A typical stress–strain curve of concrete specimen is shown inFig. 11(a) and (b) shows the experimental stress–strain curves ofsix selected concrete specimens with and without fibre in light-weight and normal weight geopolymer concrete. Fig. 11(a) definesthe different stages of concrete before crushing under compressionload. The elastic behaviour of concrete lasts up to about 40% of theultimate load and then the plastic behaviour of concrete takesplace. The non-linear plastic state is a function of the micro-cracks at paste-aggregate interface. The concrete reaches the ulti-mate stress through non-linear strain-hardening followed by strainsoftening once the micro-cracks in concrete accumulate in largeamount.

In general, all specimens behaved elastically within the elasticlimits as observed from Fig. 11(b). Fibred and non-fibred OPSGChad higher strains compared to fibred and non-fibred NWGC atultimate stress. Hence OPSGC with and without fibres producedlower MoE than the corresponding NWGC. For Portland cementbased normal weight concrete, the strain of 0.002 mm/mm corre-sponding to the ultimate stress is well established and the strainduring crushing of concrete reaches between 0.0028 mm/mmand 0.0045 mm/mm. In this investigation, the strains at ultimatestress for both fibred and non-fibred OPSGC and NWGC were foundabout 0.006 mm/mm and 0.0035 mm/mm, respectively. The higherstrain in OPSGC is attributed to the low stiffness of OPS aggregatesand its ductile property. Islam [43] reported a strain of about0.003 mm/mm at ultimate stress for POFA (40%)-GGBS (60%) basedOPSGC. However, in this research, the POFA content is 90% with10% MK and hence the high content of POFA could have an effecton strain in the specimens

Similar to Portland cement concrete, steel fibres enhanced theductility of OPSGC and NWGC. The elastic limit of fibred andnon-fibred OPSGC and NWGC was obtained at about 33% of the

E65 = 4.32x + 5.83R² = 0.7604

E80 = 2.64(%S) + 7.94R² = 0.9578

6.5

7.5

8.5

9.5

10.5

1 1.4 1.8 2.2 2.6 3

Mod

ulus

of e

last

icity

, GPa

% volume steel fibre, %S

AR65 AR80Linear (AR65) Linear (AR80)

Fig. 10. Development of MoE for different % volume of steel fibres.

(a) (b)

0

5

10

15

20

25

30

0 0.002 0.004 0.006 0.008 0.01 0.012 0.014

Stre

ss (M

Pa)

Strain (mm/mm)

OPSGC OPSGC-65/0.50 OPSGC-80/0.50NWGC NWGC-65/0.50 NWGC-80/0.50

Fig. 11. (a) Typical Stress–Strain curve of concrete in compression; (b) stress–strain curves from this experiment.

I.I. Bashar et al. / Construction and Building Materials 111 (2016) 286–297 293

ultimate stress. After initial tangent modulus, the curves showyielding of concrete in non-linear manner and this could be definedas plastic state of concrete. During the plastic stage, the strainhardening and strain softening took place and this was followedby the crushing of concrete. The aggregate and binder matrix havesignificance in the development of ductility. The higher ductilitycharacteristic of OPSGC is governed by OPS and high volume ofPOFA in binding matrix.

The effect of AR of steel fibres on ductility in both the OPSGCand NWGC is noticeable as shown in Fig. 11(b). The mixes withfibres of AR65 showed higher ductility than the correspondingmixes with AR80. The low ductility of mixes with fibres of AR80could be attributed to the number of fibres as the fibres withAR65 has higher number compared to AR80 for a given volumeof fibres; thus, with lower number of fibres in the mixes withAR80, the distribution of fibres was not uniform throughout thecohesive geopolymer matrix.

3.6. Effect of steel fibres on fracture toughness

3.6.1. Characteristics of fracture toughnessThe steel fibres have significant effect on the fracture behaviour

of concrete. The fracture failure is shown in Fig. 12. Table 9 showsthe characteristic properties of fracture toughness. These proper-ties were derived from the load deflection curve and shown theperformance of fibre-reinforced concrete. According to ASTMC1609/C1609 M-12 [63], toughness properties are recommendedto determine for two deflection limits as L/600 and L/150, whereasL is defined as beam span. The end point deflection which was cal-culated at L/150, was a constant value of 2.0 for all types of con-crete material in this experiment (Table 9).

The first peak load (P1) is the load for the first point of zero slopein the load–deflection curve. The non-fibred OPSGC and NWGCachieved first peak load of 6.99 and 6.09 kN, respectively. The

Fig. 12. Fractu

higher value of P1 signified the ductile property of OPSGC as wellas hardened composite of OPS and geopolymer matrix. The valueof P1 varied from 6.39 to 11.16 kN and 6.13 to 7.05 kN with theaddition of 0.25, 0.50 and 0.75% volume of SF(AR65) and SF(AR80),respectively. The highest value of P1 (11.16 kN) was achieved forOPSGC-65/0.75 and an insignificant deviation of P1 was noticedfor SF(AR80). The first peak load is influenced by bond between steelfibres and mortar.

The effect of steel fibres in OPSGC was found more significantthan that in NWGC. NWGC-65/0.5 and NWGC-80/0.5 achieved P1as 5.85 and 3.22 kN, respectively. It should be noted that NWGCwas designed as a control mix to compare with fibred and non-fibred OPSGC. It was noted that both OPSGC and NWGC showedenhanced P1 due to the addition of SF. The first peak strength (f1)was related directly to P1 by arithmetical equation of P1/cross-sectional area and thus f1 gave an idea of ductility performanceof concrete as shown in Table 9. The first-peak deflection (d1) ofa beam under flexure could be related to P1 and ductile perfor-mance of concrete. The higher deflection to achieve the same valueof P1 connoted that concrete absorbed more energy to gain P1;however, this information is not sufficient to decide about the duc-tility of a concrete. The ductility could be understood from thetoughness of a concrete. ASTM C1609/C1609 M-12 calculates thetoughness at the deflection of L/150. Both the OPSGC-80/0.25 andOPSGC-80/0.50 needed similar P1 of value 6.13 kN to reach d1 to1.51 and 0.26 mm, respectively; however OPSGC-80/0.50 absorbedmore energy and gained 1.6 times higher toughness than OPSGC-80/0.25 due to the presence of higher volume of steel fibre.OPSGC-80/0.25 reached maximum d1 of 1.51 mm at P1 of 6.13 kN.

3.6.2. First peak deflection (FPD)The first peak deflection (FPD) is the measure of the deflection

during the first peak load. Fig. 13 shows a comparison of the firstpeak deflection of all the mix designs with and without steel fibres

re failure.

Table9

Fracture

characteristicsfordifferen

tmix

design

s.

Fracture

characteristics

OPS

GC

OPS

GC-65/0.25

OPS

GC-80/0.25

OPS

GC-65/0.50

OPS

GC-80/0.50

OPS

GC-65/0.75

OPS

GC-80/0.75

NW

GC

NW

GC-65/0.50

NW

GC-80/0.50

Endpo

intde

flection

–2.00

2.00

2.00

2.00

2.00

2.00

–2.00

2.00

First-pe

akload

,P1(kN)

6.99

6.10

6.13

6.39

6.25

7.12

7.05

6.05

6.01

6.00

First-pe

akde

flection

,d1(m

m)

1.14

1.94

1.51

0.86

0.26

1.06

1.03

0.39

1.62

1.35

First-pe

akstrength,f

1(M

Pa)

2.61

3.79

3.75

3.54

3.34

4.16

3.84

3.64

3.68

3.67

Peak

load

,Pp(kN)

6.99

6.10

6.13

6.39

6.25

7.12

7.05

6.05

6.01

3.86

Peak

-loa

dde

flection

,dp(m

m)

1.14

1.94

1.51

0.86

0.26

1.06

1.09

0.39

1.62

6.32

Peak

strength,f

p(M

pa)

2.61

3.79

3.75

3.54

3.34

4.16

3.84

3.64

3.68

2.37

Residua

lload

,PD 60

0(kN)

1.74

0.09

0.68

3.92

4.43

5.07

2.90

0.04

2.54

1.31

Residua

lload

,PD 15

0(kN)

6.99

3.83

2.92

0.36

3.15

4.63

5.42

0.18

2.35

3.24

Residua

lstrength,f

D 600(M

Pa)

0.65

0.06

0.41

2.17

2.37

2.97

1.58

0.02

1.56

0.80

Residua

lstrength,f

D 150(M

Pa)

2.61

2.38

1.79

0.20

1.68

2.71

2.95

0.11

1.44

1.99

Specim

entough

ness,TD 15

0(J)

3.03

3.41

4.86

4.22

6.82

6.37

8.19

0.94

3.01

4.40

Equivalen

tflex

uralstrengthratio,

RD T;150

0.02

0.03

0.04

0.03

0.06

0.04

0.06

0.01

0.03

0.04

294 I.I. Bashar et al. / Construction and Building Materials 111 (2016) 286–297

of AR65 and 80 in FOPSGC and NWGC. The FPD was measuredabout 34% higher for OPSGC than NWGC and this could due tohigher ductility of the former. It should be noted that OPSGC andNWGC have no steel fibre, therefore, no residual deflection couldbe found and an immediate failure of the specimens was observedafter the first peak load. An addition of 0.25% and 0.5% of steelfibres of AR 65 has insignificant effect and reduces the FPDwhereas, 0.75% of steel fibres of AR 65 has slight improvement ofFPD.

The highest FPD was observed for the addition of 0.25% of steelfibres with AR 80 in OPSGC. The longer fibre size and an appropri-ate mix proportion of 0.25% of AR 80 in OPSGC could be attributedto the enhanced performance of mix with AR 80. The addition ofsteel fibres with NWGC improves FPD and AR 80 has better rolethan AR 65 in the FPD.

3.6.3. Fracture strengthThe first-peak strength (FPS) has the significance in characteri-

sation of the fibre-reinforced concrete [63]. The behaviour of firstpeak load and the first peak strength of fibred and non-fibredOPSGC and NWGC are shown in Fig. 14.

The FPS of non-fibred NWGC was 1.4 times higher than that ofnon-fibred OPSGC. A slight different The FPS was improved on theaddition of steel fibres in both OPSGC and NWGC. The addition of0.75% steel fibres of AR65 shows the highest FPS. The OPSGC con-taining 0.75% volume of steel fibres with AR80 produced 41% lowerstrength than the corresponding volume of fibres with AR 65. Thehigher FPS of AR65 was found for 0.25% and 0.5% volume of steelfibres. This could be attributed by the total surface area of the steelfibres that has an important role in the development of bondstrength. For a given volume of steel fibres, the number of fibreswith AR65 is more than that of AR80 and hence in an identical %volume of fibres, AR65 consists of more surface area than AR80.

It is noticed that 0.75% of steel fibres gives the highest FPSamong 0.25%, 0.5% and 0.75% of volume in both case of AR65 andAR80. A slight reduction of FPS was found for 0.50% volume of steelfibres.

The residual load (RL) and residual strength (RS) are two impor-tant characteristics of fracture behaviour and are determined gen-erally for the limiting deflection of 1/600 and 1/150 of beam span.The RS at specified deflections characterises the residual capacityof the specimen after cracking. Fig. 15 shows the RL and RS of allmix proportions of OPSGC and NWGC with/without fibres.

OPSGC, which contains OPS as aggregate and no fibre, achievedmore RL and RS than the NWGC without fibre (NWGC). The RL(L/150) and RS (L/150) of OPSGC are relatively 62% and 75% higherthan the RL (L/600) and RS (L/600); however, this variation wasfound lower in NWGC.

The addition of steel fibres influences the RL as explainedbelow. The highest RL (L/150) and RS (L/150) were obtained for0.75% volume of steel fibre of AR65. The OPSGC with steel fibreAR65 of 0.75% volume achieved about 74% higher RL (L/150) andRS (L/150) than that of OPSGC-80/0.75. Similar reduction in RL(L/150) and RS (L/150) was also found in OPSGC-80/0.25 andOPSGC-80/0.50.

The values of RL and RS in two-deflection limit of L/600 andL/150 indicate the progressive failure, which reflects the ductilityof the concrete material.

3.6.4. Fracture toughnessThe toughness and equivalent flexural strength ratio are shown

in Fig. 16. The toughness determines the capacity of energy absorp-tion of the specimen and is expressed in terms of Joules (J). Thehigher toughness indicates the better post-cracking response [49]and the better ductility performance of the fibre reinforcedconcrete.

3

4

5

6

7

8

0

0.5

1

1.5

2

2.5

OPS

GC

NW

GC

OPS

GC

-65/

0.25

OPS

GC

-65/

0.50

OPS

GC

-65/

0.75

NW

GC

-65/

0.50

OPS

GC

-80/

0.25

OPS

GC

-80/

0.50

OPS

GC

-80/

0.75

NW

GC

-80/

0.50

Firs

t pea

k lo

ad (k

N)

Firs

t pea

k de

flect

ion

(mm

)

Mix designationFirst peak deflection First Peak Load

Fig. 13. First peak deflection for all mix designs.

1

2

3

4

5

6

3

4

5

6

7

8

OPS

GC

NW

GC

OPS

GC

-65/

0.25

OPS

GC

-65/

0.50

OPS

GC

-65/

0.75

NW

GC

-65/

0.50

OPS

GC

-80/

0.25

OPS

GC

-80/

0.50

OPS

GC

-80/

0.75

NW

GC

-80/

0.50

Firs

t pea

k st

reng

th (M

Pa)

Firs

t pea

k lo

ad (k

N)

Mix designationFirst peak load First peak strength

Fig. 14. First peak load and strength.

0

1

2

3

4

5

0

2

4

6

8

10

OPS

GC

NW

GC

OPS

GC

-65/

0.25

OPS

GC

-65/

0.50

OPS

GC

-65/

0.75

NW

GC

-65/

0.50

OPS

GC

-80/

0.25

OPS

GC

-80/

0.50

OPS

GC

-80/

0.75

NW

GC

-80/

0.50

Res

idua

l stre

ngth

(MPa

)

Res

idua

l loa

d (k

N)

Mix designation

Residual load (L/600) Residual load (L/150)Residual strength (L/600) Residual strength (L/150)

Fig. 15. Residual load and strength.

I.I. Bashar et al. / Construction and Building Materials 111 (2016) 286–297 295

The non-fibred OPSGC and NWGC achieved toughness as 3.03and 0.94 J, respectively and it indicated the ductility of OPSGCcompared to NWGC. Further, the toughness as well as the ductilityincreased by the addition of steel fibres. Similar findings werereported by Yap [14] and Mo [49]. The stronger ductile propertyof OPS compared to crushed granite aggregate enabled OPSGC toproduce up to 3.2 times higher toughness than NWGC. Mo [49]

reported an increment in toughness of 6 to 17 times compared tothe control concrete by increasing volume from 0.5% to 1%. S�ahinand Köksal [64] reported an increase in the fracture energy of2.2–3.6 by adding SF(AR 80) in grade 45 of normal weight concrete.

Results show that OPSGC-65/0.50 and �/0.75 achieved up to 2times higher toughness compared to OPSGC-65/0.25. With higheraspect ratio, similar trend in the enhancement in the toughness

0.005

0.02

0.035

0.05

0.065

0.08

0

2

4

6

8

10

OPS

GC

NW

GC

OPS

GC

-65/

0.25

OPS

GC

-65/

0.50

OPS

GC

-65/

0.75

NW

GC

-65/

0.50

OPS

GC

-80/

0.25

OPS

GC

-80/

0.50

OPS

GC

-80/

0.75

NW

GC

-80/

0.50

Equi

vale

nt fl

exur

al st

reng

th ra

tio

Toug

hnes

s (Jo

ule)

Mix designationToughness Equivalent flexural strength ratio

Fig. 16. Toughness and equivalent flexural strength ratio.

296 I.I. Bashar et al. / Construction and Building Materials 111 (2016) 286–297

was noticed in OPSGC-80/0.25. The increment of volume of steelfibre enhanced the stiffness of concrete matrix which led to pro-duce stronger bridge to withstand crack propagation in concrete.As seen from Fig.15, both the increase in the volume and aspectratio enhanced the toughness of concrete.

ASTM C 1609 relates the toughness and first peak load in thecalculation of equivalent flexural strength ratio and the equationto calculate this ratio is given below:

Equivalent flexural strength ratio; RDT;150 ¼ 150 � TD

150

f 1 � b � d2 � 100%

where, TD150 – specimen toughness in Joules, f1 – First-peak strength

in MPa, b and d are breadth and width of the specimens in mm.The equivalent flexural strength ratio OPSGC-80/0.75 was the

highest and as seen from Fig. 16, it changes relatively with changein toughness.

4. Conclusion

The use of high volume POFA in the development of OPSGC wasinvestigated; in addition, the effect of volume fraction and aspectratio (AR) of steel fibre on the mechanical properties of OPSGCand fracture behaviour of fibre reinforced OPSGC and NWGC wasalso investigated and reported in this research. Based on the exper-imental investigation, the following conclusions are drawn:

1. The volume fraction and aspect ratio (AR) of steel fibre hadlittle or no influence on the compressive strength of bothOPSGC and NWGC.

2. The use of steel fibres with higher aspect ratio enhanced theflexural and splitting tensile strengths and this could beattributed to due to the capability of fibres with longerlengths in delaying crack propagation in concrete.

3. The bonding between steel fibre and binder influencedthe development of flexural and splitting tensile strengthsand this is in turn related to the surface area of the steelfibre.

4. It has been observed from the failure mode of OPSGC, bondfailure of OPS along the convex surface governed the failureof flexural specimens; however, the flexural and splittingtensile strengths of OPSGC were found higher than that ofthe corresponding NWGC. This could be attributed to theeffect of stronger bond on the concave face and along therougher surfaces of the crushed OPS.

5. The mixes with fibres of AR65 showed higher ductility thanthe corresponding mixes with AR80, as the former hadhigher number of fibres compared to the latter for a givenvolume of fibres; the lower number of fibres in the AR80mixes reduced the ductility.

6. Though the addition of fibres in OPSGC slightly influencedthe modulus of elasticity of the concrete, it was significantlylower than the corresponding NWGC. On the other hand,lower modulus of elasticity had higher ductilitycharacteristics.

7. The addition of high volume of steel fibre with low aspectratio in OPSGC showed higher energy absorption.

8. The toughness and equivalent flexural strength ratio ofOPSGC were found higher than the corresponding NWGCand this could be attributed to ductility of OPS.

9. The values of residual load and residual strength in two-deflection limits of L/600 and L/150 indicated the progres-sive failure, which reflects the ductility of fibred OPSGC.

10. OPSGC produced higher residual load and strengths than thecorresponding NWGC which signified the better ductileproperty of OPSGC.

Acknowledgement

This research work was funded by the University of Malayaunder the High Impact Research Grant (HIRG) No. UM.C/625/1/HIR/MOHE/ENG/02/D000002-16001 ‘‘Synthesis of Blast ResistantStructures”.

References

[1] J. Davidovits, Geopolymer Chemistry & Application, third ed., 16 rule Galilee, F-02100, Institute Géopolymèr, Saint-Quentin, France, 2008.

[2] J.K. Steinberger, F. Krausmann, N. Eisenmenger, Global patterns of materialsuse: a socioeconomic and geophysical analysis, Ecol. Econ. 69 (5) (2010) 1148–1158.

[3] U.J. Alengaram, B.A.A. Muhit, M.Z. Jumaat, Utilization of oil palm kernel shell aslightweight aggregate in concrete – a review, Constr. Build. Mater. 38 (2013)161–172.

[4] S. Rukzon, P. Chindaprasirt, Use of disposed waste ash from landfills to replacePortland cement, Waste Manage. Res. 27 (6) (2009) 588–594.

[5] M. Safiuddin, U.J. Alengaram, M.A. Salam, M.Z. Jumaat, F.F. Jaafar, H. Saad,Properties of high-workability concrete with recycled concrete aggregate,Mater. Res. 14 (2) (2011) 248–255.

[6] M. Safiuddin, M. Abdus Salam, M.Z. Jumaat, Utilization of palm oil fuel ash inconcrete: a review, J. Civil Eng. Manage. 17 (2) (2011) 234–247.

I.I. Bashar et al. / Construction and Building Materials 111 (2016) 286–297 297

[7] I.I. Bashar, U.J. Alengaram, M.Z. Jumaat, A. Islam, The effect of variation ofmolarity of alkali activator and fine aggregate content on the compressivestrength of the fly ash: palm oil fuel ash based geopolymer mortar, Adv. Mater.Sci. Eng. 2014 (2014) 1–13, http://dx.doi.org/10.1155/2014/245473.

[8] K.H. Mo, S.P. Yap, U.J. Alengaram, M.Z. Jumaat, C.H. Bu, Impact resistance ofhybrid fibre-reinforced oil palm shell concrete, Constr. Build. Mater. 50 (2014)499–507.

[9] M.Y.J. Liu, C.P. Chua, U.J. Alengaram, M.Z. Jumaat, Utilization of palm oil fuelash as binder in lightweight oil palm shell geopolymer concrete, Adv. Mater.Sci. Eng. 2014 (2014) 1–6, http://dx.doi.org/10.1155/2014/610274.

[10] G. Balamurugan, P. Perumal, Use of quarry dust to replace sand in concrete-anexperimental study, Int. J. Sci. Res. Publ. 3 (12) (2013).

[11] T. Ji, C.-Y. Chen, Y.-Z. Zhuang, J.-F. Chen, A mix proportion design method ofmanufactured sand concrete based on minimum paste theory, Constr. Build.Mater. 44 (2013) 422–426.

[12] S.N. Raman, T. Ngo, P. Mendis, H.B. Mahmud, High-strength rice husk ashconcrete incorporating quarry dust as a partial substitute for sand, Constr.Build. Mater. 25 (7) (2011) 3123–3130.

[13] M. Westerholm, B. Lagerblad, J. Silfwerbrand, E. Forssberg, Influence of fineaggregate characteristics on the rheological properties of mortar, CementConcr. Compos. 30 (4) (2008) 274–282.

[14] S.P. Yap, C.H. Bu, U.J. Alengaram, K.H. Mo, M.Z. Jumaat, Flexural toughnesscharacteristics of steel–polypropylene hybrid fibre-reinforced oil palm shellconcrete, Mater. Des. 57 (2014) 652–659.

[15] M.Y.J. Liu, U.J. Alengaram, M.Z. Jumaat, K.H. Mo, Evaluation of thermalconductivity, mechanical and transport properties of lightweight aggregatefoamed geopolymer concrete, Energy Build. 72 (2014) 238–245.

[16] R.H. Kupaei, U.J. Alengaram, M.Z. Jumaat, H. Nikraz, Mix design for fly ashbased oil palm shell geopolymer lightweight concrete, Constr. Build. Mater. 43(2013) 490–496.

[17] P. Shafigh, H.B. Mahmud, M.Z.B. Jumaat, R. Ahmmad, S. Bahri, Structurallightweight aggregate concrete using two types of waste from the palm oilindustry as aggregate, J. Clean. Prod. 80 (2014) 187–196.

[18] K.H. Mo, U.J. Alengaram, M.Z. Jumaat, S.P. Yap, Feasibility study of high volumeslag as cement replacement for sustainable structural lightweight oil palmshell concrete, J. Clean. Prod. 91 (2015) 297–304.

[19] U.J. Alengaram, B.A.A. Muhit, M.Z. Jumaat, M.L.Y. Jing, A comparison of thethermal conductivity of oil palm shell foamed concrete with conventionalmaterials, Mater. Des. 51 (2013) 522–529.

[20] U.J. Alengaram, M.Z. Jumaat, H. Mahmud, M.M. Fayyadh, Shear behaviour ofreinforced palm kernel shell concrete beams, Constr. Build. Mater. 25 (6)(2011) 2918–2927.

[21] U.J. Alengaram, M.Z. Jumaat, H. Mahmud, Influence of sand content and silicafume on mechanical properties of palm kernel shell concrete, in: Internationalconference on construction and building technology (ICCBT 2008), 2008, pp.251–262.

[22] D.C. Okpala, Palm kernel shell as a lightweight aggregate in concrete, Build.Environ. 25 (4) (1990) 291–296.

[23] A.A.A. Abdullah, 10 – Palm oil shell aggregate for lightweight concrete, in: S.Chandra (Ed.), Waste Materials Used in Concrete Manufacturing, WilliamAndrew Publishing, Westwood, NJ, 1996, pp. 624–636.

[24] Guo. Y-c, Zhang. J-h, Chen. G-m, Xie. Z-h, Compressive behaviour of concretestructures incorporating recycled concrete aggregates, rubber crumb andreinforced with steel fibre, subjected to elevated temperatures, J. Clean. Prod.72 (2014) 193–203.

[25] L. Kidalova, N. Stevulova, E. Terpakova, A. Sicakova, Utilization of alternativematerials in lightweight composites, J. Clean. Prod. 34 (2012) 116–119.

[26] D.J. Oehlers, M. Mohamed Ali, M. Haskett, W. Lucas, R. Muhamad, P. Visintin,FRP-reinforced concrete beams: unified approach based on IC theory, J.Compos. Constr. 15 (3) (2011) 293–303.

[27] P. Visintin, D.J. Oehlers, C. Wu, M. Haskett, A mechanics solution for hinges inRC beams with multiple cracks, Eng. Struct. 36 (2012) 61–69.

[28] P. Visintin, D.J. Oehlers, M. Haskett, C. Wu, Mechanics-based hinge analysis forreinforced concrete columns, J. Struct. Eng. 139 (11) (2012) 1973–1980.

[29] P. Visintin, D.J. Oehlers, C. Wu, M.C. Griffith, The reinforcement contribution tothe cyclic behaviour of reinforced concrete beam hinges, Earthquake Eng.Struct. Dynam. 41 (12) (2012) 1591–1608.

[30] P. Visintin, D.J. Oehlers, M. Haskett, Partial-interaction time dependentbehaviour of reinforced concrete beams, Eng. Struct. 49 (2013) 408–420.

[31] D. Knight, P. Visintin, D.J. Oehlers, M. Jumaat, Incorporating residual strains inthe flexural rigidity of RC members with varying degrees of prestress andcracking, Adv. Struct. Eng. 16 (10) (2013) 1701–1718.

[32] P. Rashiddadash, A.A. Ramezanianpour, M. Mahdikhani, Experimentalinvestigation on flexural toughness of hybrid fiber reinforced concrete(HFRC) containing metakaolin and pumice, Constr. Build. Mater. 51 (2014)313–320.

[33] T. Zhang, P. Visintin, D.J. Oehlers, M.C. Griffith, Presliding shear failure inprestressed RC beams. II: behavior, J. Struct. Eng. 140 (10) (2014) 04014070.

[34] D.J. Oehlers, P. Visintin, M. Haskett, W.M. Sebastian, Flexural ductilityfundamental mechanisms governing all RC members in particular FRP RC,Constr. Build. Mater. 49 (2013) 985–997.

[35] T. Zhang, P. Visintin, D.J. Oehlers, M.C. Griffith, Presliding shear failure inprestressed RC beams. I: partial-Interaction mechanism, J. Struct. Eng. 140 (10)(2014) 04014069.

[36] D.J. Oehlers, P. Visintin, M.H.T. Haskett, The ideal bond characteristics forreinforced concrete members, in: International Symposium on Bond inConcrete, Brescia, Italy, 2012.

[37] D.J. Oehlers, P. Visintin, M.H.T. Haskett, J. Chen, Consequences and solutions toour abysmal neglect of the bond-slip behaviour in reinforced concrete, in:International Symposium on Bond in Concrete, Brescia, Italy, 2012.

[38] M. Madhkhan, R. Azizkhani, E. THM, Effects of pozzolans together with steeland polypropylene fibers on mechanical properties of RCC pavements, Constr.Build. Mater. 26 (1) (2012) 102–112.

[39] P. Sarker, Bond strengths of geopolymer and cement concretes, Adv. Sci.Technol. 69 (2011) 143–151.

[40] A. Rajerajeswari, G. Dhinakaran, M. Ershad, Compressive strength of silicafume based geopolymer concrete, Asian J. Appl. Sci. 2013 (2013) 1–8, http://dx.doi.org/10.3923/ajaps.2013.

[41] P. Posi, C. Teerachanwit, C. Tanutong, S. Limkamoltip, S. Lertnimoolchai, V.Sata, et al., Lightweight geopolymer concrete containing aggregate fromrecycle lightweight block, Mater. Des. 52 (2013) 580–586.

[42] U.J. Alengaram, H. Mahmud, M.Z. Jumaat, Development of lightweight concreteusing industrial waste material, palm kernel shell as lightweight aggregate andits properties, in: 2nd International Conference on Chemical, Biological andEnvironmental Engineering (ICBEE 2010), IEEE, 2010, pp. 277–281.

[43] A. Islam, U.J. Alengaram, M.Z. Jumaat, I.I. Bashar, S.M.A. Kabir, Engineeringproperties and carbon footprint of ground granulated blast-furnace slag-palmoil fuel ash-based structural geopolymer concrete, Constr. Build. Mater. 101(Part 1) (2015) 503–521.

[44] J. Davidovits, Geopolymers, J. Therm. Anal. Calorim. 37 (8) (1991) 1633–1656.[45] J. Davidovits, Properties of geopolymer cements, in: First International

Conference on Alkaline Cements and Concretes, 1994, pp. 131–149.[46] D. Hardjito, B.V. Rangan, Development and properties of low-calcium fly-ash

based geopolymer concrete, Research Report GC 1, Engineering Faculty, CurtinUniversity of Technology, Perth, Australia, 2005.

[47] J. Gao, W. Sun, K. Morino, Mechanical properties of steel fiber-reinforced, high-strength, lightweight concrete, Cement Concr. Compos. 19 (4) (1997) 307–313.

[48] R.D. Neves, J.Fd. Almeida, Compressive behaviour of steel fibre reinforcedconcrete, Struct. Concr. 6 (1) (2005) 1–8.

[49] K.H. Mo, K.K.Q. Yap, U.J. Alengaram,M.Z. Jumaat, The effect of steel fibres on theenhancement of flexural and compressive toughness and fracturecharacteristics of oil palm shell concrete, Constr. Build. Mater. 55 (2014) 20–28.

[50] O. Gencel, W. Brostow, T. Datashvili, M. Thedford, Workability and mechanicalperformance of steel fiber-reinforced self-compacting concrete with fly ash,Compos. Interfaces 18 (2) (2011) 169–184.

[51] R.J. Hoyle, Design of composite beams, J. Mater. Ed. 6 (1984) 927–938.[52] M. Criswell, Design of columns, J. Mater. Ed. 8 (1986) 305–315.[53] A.M. Neville, Properties of Concrete, fourth ed., Pearson Education Limited,

Essex, England, 1995.[54] C.D. Atis�, O. Karahan, Properties of steel fiber reinforced fly ash concrete,

Constr. Build. Mater. 23 (1) (2009) 392–399.[55] F. Köksal, F. Altun, _I. Yigit, Y. S�ahin, Combined effect of silica fume and steel

fiber on the mechanical properties of high strength concretes, Constr. Build.Mater. 22 (8) (2008) 1874–1880.

[56] S.P. Yap, U.J. Alengaram, M.Z. Jumaat, Enhancement of mechanical propertiesin polypropylene– and nylon–fibre reinforced oil palm shell concrete, Mater.Des. 49 (2013) 1034–1041.

[57] P. Shafigh, U.J. Alengaram, H. Mahmud, M.Z. Jumaat, Engineering properties ofoil palm shell lightweight concrete containing fly ash, Mater. Des. 49 (2013)613–621.

[58] H. Wang, L. Wang, Experimental study on static and dynamic mechanicalproperties of steel fiber reinforced lightweight aggregate concrete, Constr.Build. Mater. 38 (2013) 1146–1151.

[59] H.-T. Wang, L.-C. Wang, Experimental study on mechanical properties of steelfiber reinforced lightweight aggregate concrete, J. Build. Mater. 10 (2) (2007)188–194. Jianzhu Cailiao Xuebao.

[60] J.F. Mandell, F.J. McGarry, C.-G. Li, Fatigue crack growth and lifetime trends ininjection molded reinforced thermoplastic. High modulus fiber composites inground transportation and high volume applications, ASTM STP 873 (1985)36–50.

[61] H. Huang, R. Talreja, Numerical simulation of matrix micro-cracking in shortfiber reinforced polymer composites: initiation and propagation, Compos. Sci.Technol. 66 (15) (2006) 2743–2757.

[62] S. Kurugöl, L. Tanaçan, H.Y. Ersoy, Young’s modulus of fiber-reinforced andpolymer-modified lightweight concrete composites, Constr. Build. Mater. 22(6) (2008) 1019–1028.

[63] ASTM C1609/C1609M-12, Standard test method for flexural performance offiber-reinforced concrete (using beam with third-point loading), ASTMInternational, West Conshohocken, PA, 2012.

[64] Y. S�ahin, F. Köksal, The influences of matrix and steel fibre tensile strengths onthe fracture energy of high-strength concrete, Constr. Build. Mater. 25 (4)(2011) 1801–1806.