Construction and Building Materials 30 (2012) 814–825

Contents lists available at SciVerse ScienceDirect

Construction and Building Materials

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

Mechanical and dynamic properties of coconut fibre reinforced concrete

Majid Ali ⇑, Anthony Liu, Hou Sou, Nawawi ChouwDepartment of Civil and Environmental Engineering, The University of Auckland, Private Bag 92019, Auckland 1142, New Zealand

a r t i c l e i n f o a b s t r a c t

Article history:Received 21 July 2011Received in revised form 12 December 2011Accepted 23 December 2011Available online 21 January 2012

Keywords:Coconut fibresConcrete compositesMechanical and dynamic propertiesBeam

0950-0618/$ - see front matter � 2012 Elsevier Ltd. Adoi:10.1016/j.conbuildmat.2011.12.068

⇑ Corresponding author.E-mail address: mali078@aucklanduni.ac.nz (M. A

Coconut fibres have the highest toughness amongst natural fibres. They have potential to be used as rein-forcement in low-cost concrete structures, especially in tropical earthquake regions. For this purpose, themechanical and dynamic properties of coconut fibre reinforced concrete (CFRC) members need to be wellunderstood. In this work, in addition to mechanical properties, damping ratio and fundamental frequencyof simply supported CFRC beams are determined experimentally. A comparison between the static anddynamic moduli is conducted. The influence of 1%, 2%, 3% and 5% fibre contents by mass of cementand fibre lengths of 2.5, 5 and 7.5 cm is investigated. To evaluate the effect of coconut fibres in improvingthe properties of concrete, the properties of plain concrete are used as a reference. Damping of CFRCbeams increases while their fundamental frequency decreases with structural damage. CFRC with higherfibre content has a higher damping but lower dynamic and static modulus of elasticity. It is found thatCFRC with a fibre length of 5 cm and a fibre content of 5% has the best properties.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

Researchers have used plant fibres as an alternative of steel orsynthetic fibres in composites such as cement paste, mortar andconcrete [1–26]. These natural fibres include coconut, sisal, jute,hibiscus cannabinus, eucalyptus grandis pulp, malva, ramie bast,pineapple leaf, kenaf bast, sansevieria leaf, abaca leaf, vakka, date,bamboo, palm, banana, hemp, flax, cotton and sugarcane fibres. Nat-ural fibres are cheap and locally available in many countries. Theiruse, as a construction material, for improving the properties of thecomposites costs a very little when compared to the total cost ofthe composites. Compared to steel fibres, they are also easy to useor handle because of their flexibility, especially when high percent-age of fibres is involved. However, in such a case, a methodology forcasting needs to be developed. For expressing the quantities offibres, volume fraction and fibre content are often used [11–16].Volume fraction can either be part of total volume of compositeor part of volume of any ingredient to be replaced. Fibre contentcan be part of total weight/mass of composite or any ingredient tobe replaced. Researchers often investigated the optimum quantityand length of fibres [11–15] to achieve maximum strength of thecomposite; any further increase or decrease in volume fractionand/or fibre length may reduce the composite strength.

Coconut fibre is extracted from the outer shell of a coconut. Thecommon name, scientific name and plant family of coconut fibreare coir, cocos nucifera and arecaceae (Palm), respectively. Thereare two types of coconut fibres, brown fibre extracted from

ll rights reserved.

li).

matured coconuts and white fibres extracted from immature coco-nuts. Brown fibres are thick, strong and have high abrasion resis-tance, while white fibres are smoother and finer, but alsoweaker. Coconut fibres are commercially available in three forms,namely bristle (long fibres), mattress (relatively short) and decor-ticated (mixed fibres). These different types of fibres have differentuses depending upon the requirement. In engineering, brown fi-bres are mostly used. According to official website of InternationalYear for Natural Fibres 2009 [27], approximately, 500,000 tonnesof coconut fibres are produced annually worldwide, mainly in Indiaand Sri Lanka. Its total value is estimated at $100 million. India andSri Lanka are also the main exporters, followed by Thailand, Viet-nam, the Philippines and Indonesia. Around half of the coconut fi-bres produced is exported in the form of raw fibre. The generaladvantages of coconut fibres include moth-proof; resistant to fungiand rot, provide excellent insulation against temperature andsound, flame-retardant, unaffected by moisture and dampness,tough and durable, resilient, spring back to shape even after con-stant use. Coconut fibre is the toughest fibre (21.5 MPa) amongstnatural fibres (Munawar et al. [28]). They are also capable of takingstrain 4–6 times more than that of other fibres (Munawar et al. [28]and Satyanarayana et al. [29]).

Abiola [30] evaluated the mechanical properties (load-extensionand stress–strain curves, Young’s modulus, yield stress, stress andstrain at break) of inner and outer coconut fibres experimentally,and the results were verified by finite element method using a com-mercial software ABAQUS. The author found that the inner coconutfibre had a higher mechanical strength as compared to that of outerfibre, but the outer coconut fibre had a higher elongation propertywhich enables it to absorb or withstand higher stretching energy.

M. Ali et al. / Construction and Building Materials 30 (2012) 814–825 815

Ramakrishna and Sundararajan [31] investigated the variationin chemical composition and tensile strength of four natural fibres,i.e. coconut, sisal, jute and hibiscus cannabinus fibres, when sub-jected to alternate wetting and drying and continuous immersionfor 60 days in water, saturated lime and sodium hydroxide. Chem-ical composition of all fibres changed because of immersion in theconsidered solutions. Continuous immersion was found to be crit-ical due to the loss of their tensile strength. However, coconut fi-bres were reported best for retaining a good percentage of itsoriginal tensile strength in all tested conditions.

2. Previous works on coconut fibre reinforced composites

Slate [11] investigated compressive and flexural strength ofcoconut fibre reinforced mortar. Two cement-sand ratios byweight, 1:2.75 with water cement ratio of 0.54 and 1:4 with watercement ratio of 0.82 were considered. Fibre content was 0.08%,0.16% and 0.32% by total weight of cement, sand and water. Themortars for both design mixes without any fibres were also testedas reference. Cylinders of 50 mm diameter and 100 mm height andbeams of 50 mm width, 50 mm depth and 200 mm length weretested. The curing was done for 8 days only. It was found that, com-pared to that of plain mortar of both mix designs, all strengthswere increased in the case of fibre reinforced mortar with all con-sidered fibre contents. However, a decrease in strength of mortarwith an increase of fibre content was also observed.

Cook et al. [12] reported the use of coconut fibre reinforced ce-ment composites as low cost roofing materials. The parametersstudied were fibre lengths (2.5, 3.75 and 6.35 cm), fibre volumes(2.5%, 5%, 7.5%, 10% and 15%) and casting pressure (from 1 to2 MPa with an increment of 0.33 MPa). They concluded that theoptimum composite consisted of fibres with a length of 3.75 cm,a fibre volume fraction of 7.5% and is casted under the pressureof 1.67 MPa. A comparison revealed that this composite was muchcheaper than locally available roofing materials.

Aziz et al. [13] cited the work of Das Gupta et al. [14,15] whostudied the mechanical properties of cement paste compositesfor different lengths and volume fractions of coconut fibres. Azizet al. concluded that the tensile strength and modulus of ruptureof cement paste increased when fibres up to 38 mm fibre lengthand 4% volume fraction were used. A further increase in lengthor volume fraction could reduce the strength of composite. Thetensile strength of cement paste composite was 1.9, 2.5, 2.8, 2.2and 1.5 MPa when it was reinforced with 38 mm long coconut fibreand the volume fractions of 2%, 3%, 4%, 5% and 6%, respectively. Thecorresponding modulus of rupture was 3.6, 4.9, 5.45, 5.4 and4.6 MPa, respectively. 4% volume fraction of coconut fibres gavethe highest mechanical properties amongst all tested cases. With4% volume fraction, they also studied the tensile strength of ce-ment paste reinforced with different lengths of coconut fibres.With the fibre lengths of 25, 38 and 50 mm, the reported tensilestrength was 2.3, 2.8 and 2.7 MPa, respectively. The results indi-cated that coconut fibres with a length of 38 mm and a volumefraction of 4% gave the maximum strength.

Paramasivam et al. [16] conducted a feasibility study of coconutfibre reinforced corrugated slabs of 915 mm � 460 mm � 10 mmfor low-cost housing. A cement–sand ratio of 1:0.5 and water–ce-ment ratio of 0.35 were used. Test for flexural strength using thirdpoint loading was performed. For producing required slabs having aflexural strength of 22 MPa, a fibre length of 2.5 cm, a volume frac-tion of 3%, and a casting pressure of 0.15 MPa were recommended.The thermal conductivity and absorption coefficient for low fre-quency sound were comparable with those of asbestos boards.

Agopyan et al. [17] studied coir and sisal fibres as replacementof asbestos in roofing tiles. The dimensions of the tiles were487 mm � 263 mm � 6 mm. Three-point bend test specimen with

2% total fibre volume fraction, support span of 350 mm, deflectionrate of 5 mm/min was employed for determination of the maxi-mum load. After the ageing periods of 16 and 60 months, the cor-responding maximum load taken by coir tile were 235 and 248 N,respectively while that by sisal tiles were 237 and 159 N, respec-tively. The major benefit of reinforced tiles was their at least 22%higher energy absorption than that of the unreinforced tiles whichcould help to avoid fragile rupture of tiles during transportation orinstallation.

John et al. [18] studied the coir fibre reinforced low alkaline ce-ment mortar taken from the internal and external walls of a12 year old house. The panel of the house was produced using1:1.5:0.504 (cement:sand:water, by mass) mortar reinforced with2% of coconut fibres by volume. Fibres removed from the old sam-ples were reported to be undamaged. No significant difference wasfound in the lignin content of fibres removed from external andinternal walls, confirming the durability of coconut fibres in ce-ment composites.

Luisito et al. [19] of PCA-Zamboanga Research Center in Philip-pines invented coconut fibre boards (CFB) for applications such astiles, bricks, plywood and hollow blocks. It is used for internal andexterior walls, partitions and ceiling. CFB consisted of 70% cementand 30% fibre by weight. It has water absorption of 32%, waterswelling of 4.2% and bending strength of 0.81 MPa, respectively.

Mohammad [20] tested wall panels made of gypsum and ce-ment as binder and coconut fibre as reinforcement. Bending andcompressive strength, moisture content, density and water absorp-tion were investigated. As expected, coconut fibres did not contrib-ute to bending strength of the tested wall panels. Compressivestrength increased with the addition of coconut fibres. There wasno considerable change of moisture content with coconut fibres.However, moisture content increased with time. Water absorptionof panels was not significantly affected with an increase in fibrecontent.

Ramakrishna and Sundararajan [21] carried out the experimentson impact resistance of slabs using a falling weight of 0.475 kg froma height of 200 mm. The slabs consisted of 1:3 cement–sand mortarwith the dimension of 300 mm � 300 mm � 20 mm. They werereinforced with coconut, sisal, jute and hibiscus cannabinus fibreshaving four different fibre contents of 0.5%, 1.0%, 1.5% and 2.5%by weight of cement and three fibre lengths of 20, 30 and 40 mm.A fibre content of 2% and a fibre length of 40 mm of coconut fibresshowed the best performance by absorbing 253.5 J impact energy.At ultimate failure all fibres, except coconut fibres, showed fibrefracture while coconut fibre showed fibre pull-out. The ultimatefailure was determined based on the number of blows requiredto open a crack in the specimen sufficiently and for the propagationof the crack through the entire depth of the specimen.

Li et al. [22] studied untreated and alkalized coconut fibres withthe lengths of 20 mm and 40 mm as reinforcement in cementitiouscomposites. Mortar was mixed in a laboratory mixer at a constantspeed of 30 rpm, with cement: sand: water: super plasticizer ratioof 1:3:0.43:0.01 by weight, and fibres were slowly put into the run-ning mixer. The resulting mortar had a better flexural strength (in-creased up to 12%), higher energy absorption ability (up to 1680%)and a higher ductility (up to 1740%), and is lighter than the conven-tional mortar.

Reis [23] performed third-point loading tests to investigate theflexural strength, fracture toughness and fracture energy of epoxypolymer concrete reinforced with coconut, sugarcane bagasse andbanana fibres. The investigation revealed that fracture toughnessand energy of coconut fibre reinforced polymer concrete werethe highest, and an increase of flexural strength up to 25% was ob-served with coconut fibres.

Asasutjarit et al. [24] determined the physical (density, mois-ture content, water absorption and thickness swelling), mechanical

Table 1Properties of plain concrete (PC) and coconut fibre reinforced concrete (CFRC) according to Baruah and Talukdar [25].

Fibre volume fraction(%)

Compressive strength(MPa)

Split tensile strength(MPa)

Modulus of rupture(MPa)

Shear strength(MPa)

Toughness index(I5)

Toughness index(I10)

– 21.42 2.88 3.25 6.18 1.934 1.9340.5 21.70 3.02 3.38 6.47 2.165 2.2701.0 22.74 3.18 3.68 6.81 2.109 2.7731.5 25.10 3.37 4.07 8.18 2.706 4.2742.0 24.35 3.54 4.16 8.21 2.345 3.452

816 M. Ali et al. / Construction and Building Materials 30 (2012) 814–825

(modulus of elasticity, modulus of rupture and internal bond) andthermal properties of coir-based light weight cement board after28 days of hydration. The physical and mechanical properties weremeasured by Japanese Industrial Standard JIS A 5908-1994 and thethermal properties according to JIS R 2618. The parameters studiedwere fibre length, coir pre-treatment and mixture ratio. 6 cm longboiled and washed fibres with the optimum cement:fibre:waterweight ratio of 2:1:2 gave the highest modulus of rupture andinternal bond amongst the tested specimens. The board also hada thermal conductivity lower than other commercial flake boardcomposite.

Baruah and Talukdar [25] investigated the mechanical proper-ties of plain concrete (PC) and fibre reinforced concrete (FRC) withdifferent fibre volume fractions ranging from 0.5% to 2%. Steel, syn-thetic and jute and coconut fibres were used. Here, the discussionis limited to the coconut fibres reinforced concrete (CFRC) only. Thecement:sand:aggregate ratio for plain concrete was 1:1.67:3.64,and the water cement ratio was 0.535. Coconut fibres havinglength of 4 cm and an average diameter of 0.4 mm with volumefraction of 0.5%, 1%, 1.5% and 2% were added to prepare CFRC.The sizes of specimens were (1) 150 mm diameter and 300 mmheight for cylinders (2) 150 mm width, 150 mm depth and700 mm length for beams, and (3) 150 mm cubes having a cut of90 mm � 60 mm in cross-section and 150 mm high for L-shapedshear test specimens. All specimens were cured for 28 days. Thecompressive strength r, splitting tensile strength (STS), modulusof rupture (MOR) using four point load test and shear strength s,are shown in Table 1 for PC and CFRC. It can be seen that CFRC with2% fibres showed the best overall performance amongst all volumefractions. The compressive strength, splitting tensile strength,modulus of rupture and shear strength of coir fibre reinforced con-crete with 2% fibres by volume fraction were increased up to 13.7%,22.9%, 28.0% and 32.7%, respectively as compared to those of plainconcrete. Their research indicated that all these properties wereimproved as well for CFRC with other fibre volume fractions of0.5%, 1% and 1.5%. Even for CFRC with small fibre volume fractionof 0.5% the corresponding properties were increased up to 1.3%,4.9%, 4.0% and 4.7%, respectively.

Li et al. [26] studied fibre volume fraction and fibre surfacetreatment with a wetting agent for coir mesh reinforced mortarusing nonwoven coir mesh matting. They performed a four-pointbending test and concluded that cementitious composites, rein-forced by three layers of coir mesh with a low fibre content of1.8%, resulted in a 40% improvement in the maximum flexuralstrength. The composites were 25 times stronger in flexural tough-ness and about 20 times higher in flexural ductility.

To the best knowledge of the authors the only research [25] onthe static CFRC properties is done with only one coir fibre length of4 cm. With regard to dynamic properties of CFRC, no study hasbeen reported. Dynamic tests had been performed only for con-crete reinforced by other fibres, e.g. polyolefin fibres [32] or rubberscrap [33]. To reveal the consequence of fibre length for CFRC prop-erties, thorough investigations involving more fibre lengths andother parameters are required in order to have reliable insights.To be able to apply CFRC in cheap housing in tropical earthquake

regions, the knowledge of static and dynamic properties of CFRCis necessary. This study is the first step in filling this knowledgegap. CFRC can be used in blocks, parking pavements to avoidshrinkage cracks. Even it can also be used in normal reinforcedconcrete to improve its behaviour during earthquake. But it needsto be properly investigated before implementation.

3. Experimental work

In this study, fibre lengths of 2.5, 5 and 7.5 cm and fibre contents of 1%, 2%, 3%and 5% by cements mass are considered to investigate the mechanical and dynamicproperties of CFRC.

3.1. Preparation of coconut fibres and CFRC

Ordinary Portland cement, sand, aggregates, water and imported brown coco-nut fibres were used for preparation of CFRC. The maximum size of aggregateswas 12 mm (passing through 12 mm sieve and retained at 10 mm sieve). The meandiameter of coconut fibres is 0.25 mm.

Since fibres were in hydraulic compressed form, preparation of fibres into therequired length was a time consuming and laborious task. Different approacheswere tried to get fibres into the required length quickly without much success. Fi-nally, coconut fibres were loosed and soaked in tap water for 30 min to soften thefibres and to remove coir dust. Fibres were washed and soaked again for 30 min.Washing and soaking were repeated three times. Fibres were then straightenedmanually and combed with a steel comb. To accelerate drying process, wet long fi-bres were put in an oven at 30 �C for 10–12 h where for the most part moisture wasremoved. The fibres were then completely dried in the open air, combed again andfinally cut into the required length with a guillotine. It may be noted that the pre-cut fibres are also commercially available at relatively high cost, as these are pre-pared for special purposes like brushes, mats etc. This cost can be reduced if fibresare mechanically prepared at large scales.

3.2. Mix design

For plain concrete, the mix design ratio for cement, sand and aggregates was 1,2 and 2, respectively with water cement (W/C) ratio of 0.48. The mix design forCFRC was the same as that of plain concrete, except that (1) more water was added(stepwise to avoid bleeding) because of fibre addition to make CFRC workable and(2) different lengths and contents of fibres were added and the same amount ofaggregates was deducted from the total mass of aggregates. All materials weretaken by mass of cement. It is well known that the W/C ratio has an influence onproperties of concrete, but compaction is also an important factor. The increasedW/C ratios for CFRC were to ensure its proper compaction with workable mix sothat the good strength could be achieved. The obtained properties of CFRC withrespective W/C ratio can be taken as optimum one, because any addition of waterwill cause bleeding, ultimately reducing its strength in hardened state. Whereas,reduced W/C ratio can lead to improper compaction, again resulting in lessstrength.

3.3. Casting procedure

A pan type concrete mixer was used in preparing plain concrete. All materialswere put in the mixer pan along with the water, and the mixer was rotated for threeminutes. The slump test was 50 mm.

For preparing CFRC, a layer of coconut fibres was spread in the pan, followed byspreading of aggregates, sand and cement. The first layer of fibres was hidden underthe dry concrete materials with the help of a spade. Then, another layer of coconutfibres followed by layers of aggregates, sand and cement was spread. This process isrepeated until the rest materials were put into the mixer pan. Approximately, threequarters of the water (according to a water cement ratio of 0.48 which was thesame as that of plain concrete) was added, and the mixer was rotated for 2 min.Then the remaining water was added and the mixer was again rotated for 2 min.All CFRCs were not workable at this stage, so more water was added in small

0.45

0.5

0.55

0.6

0.65

0.7

W/C

rat

io

2.5 cm 5.0 cm7.5 cm PC

0.45

0.5

0.55

0.6

0.65

0.7

1% 2% 3% 5% 2.5 cm 5 cm 7.5 cm

W/C

rat

io

1% 2% 3%5% PC

(a) (b)

Fig. 1. Effect of fibre (a) content and (b) length on water cement ratio.

10

40

70

100

Slum

p (m

m)

2.5 cm 5.0 cm7.5 cm PC

10

40

70

100

1% 2% 3% 5% 2.5 cm 5 cm 7.5 cmSl

ump

(mm

)

1% 2% 3%5% PC

(a) (b)

Fig. 2. Influence of fibre (a) content and (b) length on slump.

Table 2Specimens labelling scheme.

Fibres (%)a Fibres length 0 cm 2.5 cm 5 cm 7.5 cmSymbol O 1 2 3

0 O OO – – –1 A – A1 A2 A32 B – B1 B2 B33 C – C1 C2 C35 D – D1 D2 –

a By mass of cement.

M. Ali et al. / Construction and Building Materials 30 (2012) 814–825 817

increments to make the CFRC workable. The mixer was rotated for one minute aftereach increment of water. The water cement (W/C) ratio of CFRC varied from 0.49 to0.62. It was to make sure that there should not be any bleeding. Finally, the mixerwas rotated for three minutes to get CFRC well mix.

The effect of fibre content and length on W/C ratio of CFRC is shown in Fig. 1. Itcan be observed that increasing fibre content or length results in an increased W/Cratio. The W/C ratio of all CFRCs was more than that of PC.

A slump test for PC and CFRC was always performed before pouring it intomoulds. The slumps for CFRCs were 10–40 mm, but CFRC was workable inspite ofthis low slump. The slump is usually decreased with increased fibre content. Theslump of CFRC with 5 cm long fibres increased as compared to that with 2.5 cm longfibres for all considered fibre contents. The slump decreased when the fibre lengthincreased from 5 to 7.5 cm. In Fig. 2, the effect of fibre content and length on slumpis shown. The slump of all CFRCs was less than that of PC.

CFRC was poured into the moulds and the moulds were lifted up to a height ofapproximately 200–300 mm and then dropped to the floor for self compaction ofthe fibre concrete and to remove air voids from CFRC. All specimens were curedfor 28 days before testing.

3.4. Specimens

Cylinders 100 mm in diameter and 200 mm in height and beamlets 100 mmwide, 100 mm deep and 500 mm long were prepared for PC and CFRC. Beams withthe dimensions of 100 mm width, 100 mm depth and 915 mm length, were pre-pared only for CFRC, because PC beams broke during demoulding. Cylinders andbeamlets were used for material properties while beams were used to determinethe fundamental frequency, damping ratio and dynamic modulus of elasticity. Aset of three samples for each particular test was produced. A total of 6 cylinders,3 beamlets and 3 beams were prepared for one combination of a particular lengthand fibre content. Table 2 shows the specimens labelling scheme. In total 72 cylin-ders, 36 beamlets and 11 beams were tested.

A, B and C (also D, E and F for cylinders splitting tensile strength testing) alongwith labels depicted the mark of each specimen. All specimens were white washedbefore testing to enable a clear identification of cracks.

3.5. Testing procedure

3.5.1. Cylinder and beamlet testsAll cylinders were tested in a compression testing machine to determine static

modulus of elasticity Estatic, compressive strength r, corresponding strain e, com-pressive toughness Tc and splitting tensile strength STS. Each cylinder was cappedwith plaster of Paris for uniform distribution of load before testing of the Estatic, rand Tc.

All beamlets were tested in a universal testing machine of capacity 100 kNusing 4-point loads to obtain modulus of rupture MOR, corresponding deflectionD, flexural toughness Tf and cracking load PCrack. Cracking load is the load takenby fibres and part of concrete after the first visible crack is produced.

3.5.2. Beam testsThe experimental set up for testing beams is shown in Fig. 3. A small impact load

PMax-impact was applied three times at mid-span of the beam with the help of a cal-ibrated hammer. The response was recorded by accelerometers, located near tomid-span (Fig. 3a). Then the same beam was put under a static load PMax-static of1 kN in the universal testing machine. Deflection was noted using a linear variabledifferential transformer (LVDT), and Fig. 4 shows a typical load–displacementcurves. Again, a small impact load was applied three times and the response was re-corded. The static load was increased by 1 kN, followed by the impact load forrecording response. This procedure was repeated until the first crack of beam. Thestatic load before producing the first crack was taken as the reference for the justbefore crack stage. Note that the impact load was applied three times to take theaverage of resulting three values of a particular dynamic property. The magnitudeof impact load was kept small so that no additional damage was produced sincethe goal was to identify the fundamental frequency and damping at different dam-age stages. Four stages were considered: (i) Uncracked beam [S1], (ii) just beforecracking [S2], (iii) cracked beam [S3] and (iv) after cracks occurred following 2–3 cy-cles of static load [S4a, S4b and S4c]. Crack development is shown in Fig. 5. Each cy-cle of loading consists of applying a static load on the cracked beam up to a certaindeflection and then releasing the load for measuring its dynamic properties.

4. Mechanical properties

4.1. Static modulus of elasticity Estatic

Estatic is calculated as the ratio of stress change to strain changein the elastic range. Stress–strain curves of PC and CFRC with 5%and 5 cm long fibres are shown in Fig. 6a. Stress–strain relationship

Fig. 3. Experimental set up for dynamic tests (a) applying hit with calibrated hammer at mid-span and (b) introducing damage by four-point static loading.

0

1

2

3

4

5

Loa

d (k

N)

Displacement (mm)

Just before crack [S2]

At cracking [S3]

0.0

0.5

1.0

0.0 0.2 0.4 0.6 0.8 1.0 0 2 4 6 8 10

Loa

d (k

N)

Displacement (mm)

Cycle 1 [S4a]

Cycle 2 [S4b]

Cycle 3 [S4c]

After cracks

Fig. 4. Typical load displacement curves at different damage stages of CFRC beams.

At cracking [S3] After crack - cycle 1 [S4a] After crack - cycle 2 [S4b]

Fig. 5. Crack development.

818 M. Ali et al. / Construction and Building Materials 30 (2012) 814–825

for each sample shows the average of strain readings taken by twoLVDTs attached to the specimens.

Crushed PC and CFRC cylinder specimens with fibre content 5%and 5 cm long fibres for determining Estatic and r are shown inFig. 6b. It can be noted that the spalling of concrete is observedin case of PC cylinders, whereas only crack are produced for CFRCcylinders.

Fig. 7 shows the influence of fibre content and length on Estatic.The solid straight line is Estatic of PC. Estatic of CFRCs decreased withincreasing fibre content and length. However, Estatic of CFRC having2.5 cm long fibres with increasing fibre contents showed a differ-ent trend, i.e. it first increased and then decreased (Fig. 7a), andthese values were higher than that of PC. Compared to PC value,an addition of fibres caused about 15% increase or decrease of Estatic

of CFRC.The following simple equation is developed for estimating static

modulus of elasticity in GPa:

Estatic ¼ Xs þ Ysc þ Zsc2 ð1Þwhere c is fibre content parameter of the values of 0, 1, 2, 3 or 5 andXs, Ys and Zs are constants corresponding to fibre length Lf. The

values can be taken from Table 3. The percentage error in Estatic withEq. (1) is less than one percent (Table 5).

4.2. Compressive strength r

The maximum stress value of stress–strain curves is taken ascompressive strength r. Fig. 8 shows the influence of fibre contentand length on r. The straight line is r of PC. r decreased with high-er fibre content, however it first increased and then decreased withlonger fibres. Because of fibres, r can increase up to 24%. Comparedto that of PC, in most cases, r increases. Only with 7.5 cm long fi-bres having 2% or 3% fibre content, r is less than that of PC. Thismay be caused by creation of air voids because of long fibres withrelatively high fibre content.

4.3. Compressive toughness Tc

Tc is calculated as the total area under the stress–strain curve.Other researchers have taken it as area after the maximum stressup to 1% strain [34]. Fig. 9 shows the effect of fibre content and

(a) Stress-strain curves for Estatic (b) Tested cylinders for Estatic and σ

0

10

20

30

40

0.0 0.5 1.0 1.5 2.0St

ress

(M

Pa)

Strain (%)

PC

CFRC

Fig. 6. Compressive test.

(a) (b)

25

30

35

40

45 2.5 cm 5.0 cm7.5 cm PC

25

30

35

40

45

1% 2% 3% 5% 2.5 cm 5 cm 7.5 cm

Est

atic

(GPa

)

Est

atic

(GPa

)

1% 2% 3%

5% PC

Fig. 7. Influence of fibre (a) content and (b) length on static modulus of elasticity Estatic.

Table 3Constants for static modulus of elasticity Estatic.

Fibre length Constants

Lf Xs Ys Zs

– 33 0 02.5 32.6 �2.2 0.375 34 �1 07.5 33 3.6 �0.66

Table 4Constants for dynamic modulus of elasticity Edynamic.

Fibre length Constants

Lf Xd Yd Zd

2.5 39.2 0 �0.285 37 �2.6 0.147.5 29.7 2.4 �0.86

M. Ali et al. / Construction and Building Materials 30 (2012) 814–825 819

length on Tc. The solid straight line indicates Tc of PC. Compressivetoughness increased with higher fibre content. CFRC with 5% fibrecontent and 5 cm long fibres has the highest toughness of 0.32MPa.

4.4. Splitting tensile strength STS

Load-time histories, obtained during STS tests of PC and CFRCwith 5% and 5 cm long fibres, are shown in Fig. 10a. The maximumload from these curves is taken for the calculation of STS. PC cylin-ders were broken into two halves at maximum load, while CFRCcylinders were held together after cracks and even when the testwas continued up to more than 800 s for all CFRC cylinders in orderto observe the post peak load behaviour. The two pieces were heldtogether up to the end of the test. One of the tested PC and CFRCcylinder specimens are shown in Fig. 10b.

Fig. 11 shows the influence of fibre content and length on STS ofCFRC. The solid straight line is STS of PC. STS decreases with higherfibre content, however it first increases and then slightly reduceswith increasing fibre length. In the case of 1% long fibres, STS iskept on increasing with the length of fibre (Fig. 11b).

Compared to the STS of PC, an addition of fibres can increase ordecrease the splitting tensile strength up to 11%.

4.5. Modulus of rupture MOR

Fig. 12a displays the load–displacement curves of PC and CFRCwith 5% and 5 cm long fibres. The maximum load from thesecurves is taken for the calculation of MOR.

PC beams were broken into two pieces at maximum load(Fig. 12b). However, CFRC beams were held together even aftermaximum load (Fig. 12c). Fig. 12d shows the cross-section of aCFRC beam. CFRC beams were intentionally broken into two halvesto observe fibre failure. Two types of fibre failure were observed:(i) fibre breaking and (ii) fibre pull-out. More pull-out fibres wereobserved with higher fibre content, whereas this pull-out was de-creased as expected with increasing fibre length.

Fig. 13 displays the influence of fibre content and length onMOR. The solid line is MOR of PC. With higher fibre content andlonger fibre MOR increases. However, MOR of PC beams is mostlyhigher than that of CFRC beams. Compared to PC value MOR ofCFRC with 5% fibre content and 5 cm long fibres increased slightlyup to 4%.

4.6. Flexural toughness

Flexural toughness is measured as the total toughness index(TTI). It is the ratio of area under load–displacement curve up tomaximum deflection to the area under curve up to first-crack load.In Fig. 12a it is the ratio of area under curve up to 18 mm deflectionto area under curve up to 1.5 mm deflection. Usually, toughness in-dex is taken as the area under the curve up to 3, 5.5 or 10.5 times

(a) (b)

30

35

40

45

50

σ(M

Pa)

2.5 cm 5.0 cm7.5 cm PC

30

35

40

45

50

1% 2% 3% 5% 2.5 cm 5 cm 7.5 cm

σ(M

Pa)

1% 2% 3%5% PC

Fig. 8. Influence of fibre (a) content and (b) length on compressive strength r.

(a) (b)

0.2

0.25

0.3

0.35

0.4

Tc

(MPa

)

2.5 cm 5.0 cm

7.5 cm PC

0.2

0.25

0.3

0.35

0.4

1% 2% 3% 5% 2.5 cm 5 cm 7.5 cm

Tc

(MPa

)

1% 2% 3%5% PC

Fig. 9. Influence of fibre (a) content and (b) length on compressive toughness Tc.

(a) (b)

3

3.5

4

4.5

5

STS

(MPa

)

2.5 cm 5.0 cm7.5 cm PC

3

3.5

4

4.5

5

1% 2% 3% 5% 2.5 cm 5 cm 7.5 cm

STS

(MPa

)

1% 2% 3%5% PC

Fig. 11. Influence of fibre (a) content and (b) length on splitting tensile strength STS.

Fig. 10. Splitting test.

820 M. Ali et al. / Construction and Building Materials 30 (2012) 814–825

the first-crack deflection to area under curve at first-crack deflec-tion, and they are donated as I5, I10 and I15, respectively [35].

Fig. 14 shows the effect of fibre content and length on TTI. Thesolid straight line is TTI of PC. TTI increases with higher fibre con-tent (Fig. 14a). As far as fibre length is concerned for fibre contentof 1%, 2% and 3%, TTI increases a little when fibre length changesfrom 2.5 to 5 cm and then it decreases slightly when fibre length

increases to 7.5 cm (Fig. 14b). The possible reason can be explainedas follows: (i) when fibre length is 2.5 cm, more fibres are availablefor bridging the crack, however a shorter fibre embedment resultsin pull-out of fibres; (ii) when fibre length is 5 cm, relatively lessnumber of fibres is available but sufficient embedment length isthere to hold the cracks. This results in a higher TTI as comparedto that with 2.5 cm long fibres; (iii) when fibre length is 7.5 cm,

(a) Load-displacement curves for MOR (b) Tested PC beam

(c) Tested CFRC beam (d) Cross-section of a tested CFRC beam

0

4

8

12

0 5 10 15 20

Loa

d (k

N)

Displacement (mm)

CFRCPC

Fig. 12. Beamlet test.

(a) (b)

3

3.5

4

4.5

5

MO

R (

MPa

)

2.5 cm 5.0 cm7.5 cm PC

3

3.5

4

4.5

5

1% 2% 3% 5% 2.5 cm 5 cm 7.5 cm

MO

R (

MPa

)

1% 2% 3%5% PC

Fig. 13. Influence of fibre (a) content and (b) length on modulus of rupture MOR.

Table 5Comparison of modulus of elasticity.

Concrete Fibre content (%) Modulus of elasticity (GPa)

Edynamic Estatic % Difference (A and D)

Experimental Eq. (1) % Difference (A and B) Experimental Eq. (2) % Difference (D and E)A B C D E F

PC – – – – 33.10 33.00 0.30 –CFRC (2.5 cm) 1 38.56 38.92 �0.94 35.85 35.94 �0.25 7.0

2 38.80 38.08 1.86 37.85 37.56 0.77 2.53 35.96 36.68 �1.99 37.55 37.86 �0.83 �4.45 32.12 32.2 �0.25 34.32 34.50 �0.52 �6.8

CFRC (5 cm) 1 34.46 34.54 �0.24 32.94 33.00 �0.18 4.42 32.38 32.36 0.05 32.15 32.00 0.47 0.73 30.31 30.46 �0.50 31.02 31.00 0.06 �2.35 27.37 27.5 �0.47 29.28 29.00 0.96 �7.0

CFRC (7.5 cm) 1 31.25 31.24 0.05 30.80 30.77 0.10 1.52 31.11 31.06 0.15 29.69 29.68 0.03 4.63 29.24 29.16 0.26 29.31 29.33 �0.07 �0.3

M. Ali et al. / Construction and Building Materials 30 (2012) 814–825 821

number of fibres are further reduced resulting in a lower TTI ascompared to that with 5 cm long fibres.

CFRC with 5% fibre content and 5 cm long fibres has the highesttoughness index of 10.1.

4.7. Density

Fig. 15 shows the effect of fibre content and length on thedensity of CFRC. The solid straight line indicates the PC density.

(a) (b)

0

3

6

9

12

TT

I (-

)

2.5 cm 5.0 cm7.5 cm PC

0

3

6

9

12

1% 2% 3% 5% 2.5 cm 5 cm 7.5 cm

TT

I (-

)

1% 2% 3%5% PC

Fig. 14. Influence of fibre (a) content and (b) length on total toughness index TTI.

(a) (b)

2200

2250

2300

2350

2400

Den

sity

(kg

/m3 )

2.5 cm 5.0 cm7.5 cm PC

2200

2250

2300

2350

2400

1% 2% 3% 5% 2.5 cm 5 cm 7.5 cmD

ensi

ty (

kg/m

3 )

1% 2% 3%5% PC

Fig. 15. Influence of fibre (a) content and (b) length on density.

822 M. Ali et al. / Construction and Building Materials 30 (2012) 814–825

As expected, the density of CFRC decreased with higher fibre con-tent and increased with shorter fibre length. In general, the densityof CFRC decreased up to 4% as compared to that of PC. A smallerdensity is significant since less inertia forces will be activated inearthquakes and thus smaller structural dimension is required towithstand the reduced earthquake impact.

5. Dynamic properties

5.1. Damping ratio n and fundamental frequency f

Four damage stages were considered: (i) uncracked beam [S1],(ii) just before cracking [S2], (iii) cracked beam [S3] and (iv) aftercracks occurred following some cycles of static load [S4a, S4band S4c]. Typical recorded acceleration time-histories at stagesS1, S3 and S4b are shown in Fig. 16.

A logarithmic decrement is used for calculating the damping ra-tio of simply supported CFRC beams. f is calculated from the periodof the recorded acceleration time histories.

n of CFRC beams increased and f decreased with the formationof cracks. The effect of fibre content on n and f is displayed inFig. 17 for CFRC beams having a fibre length of 7.5 cm at differentdamage stages. CFRC beam with 3% fibre content has the highestdamping and the lowest fundamental frequency in an uncracked

(a) Uncracked beam [S1] (b) Cracked

0 0.1 0.2-1

0

1

Time (sec)

Ver

tical

acce

lera

tion

(g)

0 0.-1

0

1

Time

Ver

tical

acce

lera

tion

(g)

Fig. 16. Typical recorded acceleratio

and cracked stage as compared to that of CFRC beams with fibrecontents of 1% and 2%.

The effect of fibre content and length on the dynamic propertiesof CFRC beams before and after cracks is shown in Fig. 18. As ex-pected, compared to the properties at a damaged stage, the fibreeffect on the dynamic properties before damage is not so pro-nounced. However, after cracking n increases and f decreasesconsiderably.

5.2. Dynamic modulus of elasticity Edynamic

Fundamental frequency is used to define the actual beam dy-namic Young’s modulus Edynamic (Zheng et al. [33]). Fig. 19 showsa comparison of Edynamic of CFRC for considered parameters. It de-creases with increasing fibre content and length.

The following empirical equation is developed to predict dy-namic modulus of elasticity in GPa:

Edynamic ¼ Xd þ Ydc þ Zdc2 ð2Þ

where c is fibre content (0, 1, 2, 3 or 5) and Xd, Yd and Zd are con-stants corresponding to fibre length Lf. The values are given in Ta-ble 4. The percentage error in Edynamic with Eq. (2) is less than 2%(Table 5).

beam [S3] (c) After crack - cycle 2 [S4b]

1 0.2(sec)

0 0.1 0.2-1

0

1

Time (sec)

Ver

tical

acce

lera

tion

(g)

n time histories of CFRC beam.

(a) Damping ratio (b) Fundamental frequency

0

5

10

15

20

Dam

ping

rat

io (

%)

1% 2% 3%

0

50

100

150

S1 S2 S3 S4a S4b S4c S4d S1 S2 S3 S4a S4b S4c S4d

Fund

amen

tal f

requ

ency

(H

z) 1% 2% 3%

Fig. 17. Effect of fibre content on dynamic properties of CFRC beam with 7.5 cm long fibres at different damage stages.

Dam

ping

rat

io

(unc

rack

ed b

eam

) D

ampi

ng r

atio

(c

rack

ed b

eam

) Fu

ndam

enta

l fre

quen

cy

(unc

rack

ed b

eam

) Fu

ndam

enta

l fre

quen

cy

(cra

cked

bea

m)

Fibre content Fibre length

3

6

9

12

15

18

ξ(%

)

2.5 cm 5.0 cm7.5 cm

3

6

9

12

15

18

ξ (%

)

1% 2%3% 5%

3

6

9

12

15

18

ξ(%

)

2.5 cm 5.0 cm7.5 cm

3

6

9

12

15

18

ξ(%

)

1% 2%3% 5%

75

100

125

150

175

f (H

z)

2.5 cm 5.0 cm7.5 cm

75

100

125

150

175

f (H

z)

1% 2%3% 5%

75

100

125

150

175

f (H

z)

2.5 cm 5.0 cm7.5 cm

75

100

125

150

175

1% 2% 3% 5% 2.5 cm 5 cm 7.5 cm

1% 2% 3% 5% 2.5 cm 5 cm 7.5 cm

1% 2% 3% 5% 2.5 cm 5 cm 7.5 cm

1% 2% 3% 5% 2.5 cm 5 cm 7.5 cm

f (H

z)

1% 2%3% 5%

(a) (b)

Fig. 18. Influence of fibre (a) content and (b) length on dynamic properties of CFRC beams.

M. Ali et al. / Construction and Building Materials 30 (2012) 814–825 823

(a) (b)

25

30

35

40

45

Edy

nam

ic(G

Pa)

2.5 cm 5.0 cm7.5 cm

25

30

35

40

45

1% 2% 3% 5% 2.5 cm 5 cm 7.5 cm

Edy

nam

ic(G

Pa)

1% 2%

3% 5%

Fig. 19. Influence of fibre (a) content and (b) length on dynamic modulus of elasticity Edynamic.

Table 6Consequence of CFRC for mechanical properties.

Concrete type Estatic (GPa) r (MPa) STS (MPa) MOR (MPa) Density (kg/m3) Tc (MPa) TTI (–)

PC 33.1 34.7 3.82 4.34 2338 0.265 1CFRCs with minimum values 29.38 (5%, 5 cm) 31.3 (3%, 7.5 cm) 3.42 (5%, 2.5 cm) 3.41 (1%, 2.5 cm) 2242 (5%, 5 cm) 0.23 (1%, 5 cm) 3.21 (1%, 2.5 cm)CFRCs With maximum values 37.8 (2%, 2.5 cm) 43.2 (1%, 5 cm) 4.27 (1%, 7.5 cm) 4.51 (3%, 7.5 cm) 2298 (1%, 2.5 cm) 0.32 (5%, 5 cm) 10.1 (5%, 5 cm)Recommended CFRC 29.38 36.1 3.74 4.43 2242 0.32 10.1

(5% fibre content, 5 cm long fibres)

Table 7Consequence of CFRC for dynamic properties.

Concrete type Uncracked CFRC beams Cracked CFRC beams Edynamic (GPa) Estatic (GPa)

n (%) f (Hz) n (%) f (Hz)

PC – – – – – 33.1CFRCs with minimum values 2.5 (1%, 2.5 cm) 109.3 (5%, 5 cm) 5.8 (1%, 2.5 cm) 84.3 (5%, 5 cm) 27.4 (5%, 5.0 cm) 29.3 (5%, 5 cm)CFRCs with maximum values 6.9 (5%, 5 cm) 145.2 (1%, 2.5 cm) 14.1 (5%, 5 cm) 120.4 (1%, 2.5 cm) 38.8 (2%, 2.5 cm) 37.9 (2%, 2.5 cm)Recommended CFRC 6.9 109.3 14.1 84 27.4 29.3

(5% fibre content and 5 cm long fibres)

824 M. Ali et al. / Construction and Building Materials 30 (2012) 814–825

6. CFRC with best overall properties

Tables 6 and 7 summarise all static and dynamic results, respec-tively. The maximum values of different properties are generatedwith different combinations of fibre lengths and contents as canbe observed from the tables. Since the CFRC is prepared with a par-ticular fibre length and content, therefore that combination forCFRC should be selected, for which most properties are better thanthe plain concrete and the CFRC with other fibre lengths and con-tents. From the obtained results, CFRC with fibres length of 5 cmand 5% fibre content has the best overall properties. By using fibresin concrete the compressive strength, compressive toughness,modulus of rupture and total toughness index can be increasedfrom 34.7 MPa, 0.265 MPa, 4.34 MPa and 1 to 36.1 MPa,0.32 MPa, 4.43 MPa and 10.1, respectively. It also has the highestdamping. At damaged stage the damping is approximately morethan twice than that in uncracked stage.

7. Conclusions

Experiments have been performed to investigate the mechani-cal and dynamic properties of coconut fibre reinforced concrete(CFRC). The mechanical properties are static modulus of elasticityEstatic, compressive strength r, compressive toughness Tc, splittingtensile strength STS, modulus of rupture MOR, total toughness in-dex TTI and density. These properties are also compared with thoseof plain concrete. The dynamic properties are damping ratio n, fun-damental frequency f and dynamic modulus of elasticity Edynamic ofCFRC beams. The considered fibre lengths are 2.5, 5 and 7.5 cm andthe fibre contents are 1%, 2% and 3% for all fibre lengths, and 5% for

2.5 and 5 cm long fibres. Three specimens of CFRC are tested foreach combination of fibres to get reliable average results.

The static investigation reveals:

� The properties can increase or decrease depending on fibrelength and content, and CFRC strengths can be greater or smal-ler than that of plain concrete.� The testing confirmed that coconut fibres in concrete can

improve its flexural toughness considerably for all consideredcases.� The CFRC with 5 cm long fibres having 5% fibre content has an

increased r, Tc, MOR and TTI up to 4%, 21%, 2% and 910%, respec-tively, and decreased Estatic, STS and density up to 6%, 2% and 3%,respectively, as compared to that of plain concrete.

The dynamic tests show that:

� Damping of CFRC beams has, as expected, growing and the fun-damental frequency has declining trend with increasing damage.� The increase of fibre content has resulted in a higher damping

ratio and a lower fundamental frequency. The change is morepronounced after cracking.� CFRC beams with 5 cm long fibres have higher damping as com-

pared to those with other fibre lengths. It can be clearlyobserved in cracked beams.� The static and dynamic modulus of elasticity decreases with an

increase in fibre content or fibre length.� From the considered cases, CFRC with 5 cm long fibres and 5%

fibre content has the best overall mechanical and dynamicproperties.

M. Ali et al. / Construction and Building Materials 30 (2012) 814–825 825

� Only 7% difference is observed in static and dynamic modulus ofelasticity. This small difference indicates that non-destructivemodal testing may be used to determine modulus of elasticityof built members.

It is important to note that the free vibration testing of the un-cracked CFRC beams gives some idea of changes in material damp-ing due to the introduction in fibres, but these measurements arefor low amplitude excitations and behaviours would likely be dif-ferent during a seismic event. Therefore, further research on con-figuration and behaviour of CFRC structures under earthquakeloading is necessary.

Acknowledgements

The authors would like to thank all persons who helped themthroughout the research, particularly Mark Byrami, Noel Perin-panayagam, Claudio Oyarzo-Vera and Golden Bay Cement andWinstone Aggregates for support of this research. The careful re-view and constructive suggestions by the anonymous reviewersare gratefully acknowledged. The first and last authors also wishto thank Pakistan Higher Education Commission for supportingthe PhD study of the first author at the University of Auckland.

References

[1] Aggarwal LK. Studies on cement-bonded coir fibre boards. Cem Concr Compos1992;14(1):63–9.

[2] Al-Oraimi SK, Seibi AC. Mechanical characterisation and impact behaviour ofconcrete reinforced with natural fibres. Compos Struct 1995;32(1–4):165–71.

[3] Aziz MA, Paramasivam P, Lee SL. Concrete reinforced with natural fibres. NewReinf Concr 1984;1:106–40.

[4] Corradini E, De Morais LC, De Rosa MF, Mazzetto SE, Mattoso LHC, Agnelli JAM.A preliminary study for the use of natural fibers as reinforcement in starch–gluten–glycerol matrix. Macromol Symp 2006;245–246:558–64.

[5] Fernandez JE. Flax fiber reinforced concrete – a natural fiber biocomposite forsustainable building materials. High Perform Struct Mater 2002;4:193–207.

[6] Flower PA, Hughes JM, Melias R. Review bio composites: technology,environmental credentials and market forces. J Sci Food Agric 2006;86:1781–9.

[7] Mwamila BLM. Natural twines as main reinforcement in concrete beams. Int JCem Compos Lightweight Concrte 1985;7(1):11–9.

[8] Ramaswamy HS, Ahuja BM, Krishnamoorthy S. Behaviour of concretereinforced with jute, coir and bamboo fibres. Int J Cem Compos LightweightConcr 1983;5(1):3–13.

[9] Rao KMM, Rao KM. Extraction and tensile properties of natural fibers: vakka,date and bamboo. Compos Struct 2007;77(3):288–95.

[10] Toledo FRD, Ghavami K, England GL. Free, restrained and drying shrinkage ofcement mortar composites reinforced with vegetable fibres. Cem ConcrCompos 2005;27(5):537–46.

[11] Slate FO. Coconut fibers in concrete. Eng J Singapore 1976;3(1):51–4.[12] Cook DJ, Pama RP, Weerasingle HLSD. Coir fibre reinforced cement as a low

cost roofing material. Build Environ 1978;13(3):193–8.

[13] Aziz MA, Paramasivam P, Lee SL. Prospects for natural fibre reinforcedconcretes in construction. Int J Cem Compos Lightweight Concr 1981;3(2):123–32.

[14] Das Gupta N C, Paramsivam P, Lee S L. Mechanical properties of coir reinforcedcement pastes composites. Housing Sci 1978;2(5):391–406.

[15] Das Gupta NC, Paramsivam P, Lee SL. Coir reinforced cement pastescomposites. In: Conference proceedings of our world in concrete andstructures; 1979. p. 111–6.

[16] Paramasivam P, Nathan GK, Das Gupta NC. Coconut fibre reinforced corrugatedslabs. Int J Cem Compos Lightweight Concr 1984;6(1):19–27.

[17] Agopyan V, Savastano Jr H, John VM, Cincotto MA. Developments on vegetablefibre-cement based materials in São Paulo, Brazil: an overview. Cem ConcrCompos 2005;27(5):527–36.

[18] John VM, Cincotto MA, Sjostrom C, Agopyan V, Oliveira CTA. Durability ofslag mortar reinforced with coconut fibre. Cem Concr Compos 2005;27(5):565–74.

[19] Luisito JP, Neil JM, Rolendio N. <http://www.ecocoboard.net/dwnloads/Presentations%20workshop/Penamora_Coconut%20fibre%20cement%20boards.pdf>; 2005 [accessed 02.12.10].

[20] Mohammad HBMH. Coconut fiber reinforced wall panelling system. Masterthesis, Faculty of Civil Engineering. Universiti Teknologi, Malaysia; 2005.

[21] Ramakrishna G, Sundararajan T. Studies on the durability of natural fibres andthe effect of corroded fibres on the strength of mortar. Cem Concr Compos2005;27(5):575–82.

[22] Li Z, Wang L, Wang X. Flexural characteristics of coir fiber reinforcedcementitious composites. Fiber Polym 2006;7(3):286–94.

[23] Reis JML. Fracture and flexural characterization of natural fiber-reinforcedpolymer concrete. Constr Build Mater 2006;20(9):673–8.

[24] Asasutjarit C, Hirunlabh J, Khedari J, Charoenvai S, Zeghmati B, Shin UC.Development of coconut coir-based lightweight cement board. Constr BuildMater 2007;21(2):277–88.

[25] Baruah P, Talukdar S. A comparative study of compressive, flexural, tensile andshear strength of concrete with fibres of different origins. Indian Concr J2007;81(7):17–24.

[26] Li Z, Wang L, Wang X. Cement composites reinforced with surface modifiedcoir fibers. J Compos Mater 2007;41(12):1445–57.

[27] <http://www.naturalfibres2009.org/en/fibres/coir.html> [accessed 02.11.10).[28] Munawar SS, Umemura K, Kawai S. Characterization of the morphological,

physical, and mechanical properties of seven non-wood plant fibre bundles. JWood Sci 2007;53(2):108–13.

[29] Satyanarayana KG, Sukumaran K, Mukherjee PS, Pavithran C, Pillai SGK.Natural fibre–polymer composites. Cem Concr Compos 1990;12(2):117–36.

[30] Abiola AWBOA. Mechanical Property Evaluation of Coconut Fibre. Masterthesis, ISRN: BTH-AMT-EX-2008/D05-SE. Department of MechanicalEngineering, Blekinge Institute of Technology, Kariskrone, Sweden; 2008.

[31] Ramakrishna G, Sundararajan T. Impact strength of a few natural fibrereinforced cement mortar slabs: a comparative study. Cem Concr Compos2005;27(5):547–53.

[32] Yan L, Jenkins CH, Pendleton RL. Polyolefin fiber-reinforced concretecomposites part I: damping and frequency characteristics. Cem Concr Res2000;30(3):391–401.

[33] Zheng L, Sharon H, Yuan Y. Experimental investigation on dynamic propertiesof rubberized concrete. Constr Build Mater 2008;22(5):939–47.

[34] Chan R, Bindigana vile V. Toughness of fibre reinforced hydraulic lime mortar.Part 1, quasi-static response. Mater Struct 2010;43:1435–44.

[35] Richardson AE, Conventry K, Landless L. Synthetic and steel fibres in concretewith regard to equal toughness. Struct Surv 2010;28(5):355–69.