Composites: Part B 58 (2014) 259–266

Contents lists available at ScienceDirect

Composites: Part B

journal homepage: www.elsevier .com/locate /composi tesb

Effects of extrusion temperature on the rheological, dynamic mechanicaland tensile properties of kenaf fiber/HDPE composites

1359-8368/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.compositesb.2013.10.068

⇑ Corresponding author. Tel.: +60 379674249.E-mail address: ahassan@um.edu.my (A. Hassan).

Fauzani Md. Salleh, Aziz Hassan ⇑, Rosiyah Yahya, Ahmad Danial AzzahariPolymer and Composite Materials Research Laboratory, Department of Chemistry, University of Malaya, 50603 Kuala Lumpur, Malaysia

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

Article history:Received 28 March 2013Received in revised form 14 September 2013Accepted 25 October 2013Available online 6 November 2013

Keywords:B. Rheological propertiesD. Mechanical testingE. ExtrusionKenaf fiber composites

The effects of extrusion processing temperature on the rheological, dynamic mechanical analysis and ten-sile properties of kenaf fiber/high-density polyethylene (HDPE) composites were investigated for low andhigh processing temperatures. The rheological data showed that the complex viscosity, storage and lossmodulus were higher with high processing temperature. Complex viscosities of pure HDPE and 3.4 wt%composite with zero shear viscosity of 62340 Pa s were shown to exhibit Newtonian behavior whilecomposites of 8.5 and 17.5 wt% with zero shear viscosity P30,970 Pa s displayed non-Newtonian behav-ior. The Han plots revealed the sensitivity of rheological properties with changes in processing temper-ature. An increase in storage and loss modulus and a decrease in mechanical loss factor were observed for17.5 wt% composites at high processing temperature and not observed at low processing temperature.Processing at high temperature was found to improve the tensile modulus of composites but displayeddiminished properties when processed at low processing temperature especially at high fiber content.At both low and high processing temperatures, the tensile strength and strain of the composite decreasedwith increased content of the fiber.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

The interest in using natural fibers in polymer matrix composite(PMC) has increased in recent years. This is due to the lightweight,non-toxic, low cost and biodegradable properties of natural fibers.The use of natural fibers, derived from renewable resources, asreinforcing agent in both thermoplastic matrix composites pro-vides positive environmental benefits with respect to disposabilityand raw material utilization [1].

Kenaf fiber has potential as reinforcement filler in PMC. Thepurpose of producing PMC is to create a new material that has bet-ter properties compared to their individual material. Kenaf fibercan generally be classified into two types. The first type is the out-ermost layer known as bast while the second type is the inner part,known as core. The core is very soft, hollow and suitable for appli-cation as organic filler in plastic, while bast fiber has hard proper-ties and is suitable for blending with plastic, textile industry andalso fiberglass technology applications. Kenaf fiber is also used asreinforcement for plastic and synthetic product, cosmetic product,organic filler and medicine. Besides that kenaf fiber is also environ-mental friendly.

The performance of a composite material varies with the fiber–matrix bond strength and to some extent, depends on the choice of

suitable processing techniques. There are various methods of pro-cessing natural fiber–polymer matrix composites. Methods such asextrusion, compression and injection molding are used to intro-duce fibers into the thermoplastic matrix. Twin-screw extrudersystem, with a feeding and a mixing zone, consists of screws witha multiplex shape. In this system, the fillers are very well blendedwith the matrix polymer because the mixtures pass through anumber of mixing blocks [2]. Potente et al. [3] studied the impactof speed, melt throughput, continuous-phase viscosity, screw con-figuration, and disperse-phase content on the melting behaviorand morphology development in the melting zone of a twin-screwextruder of polypropylene/polyamide-6 composite. Their resultshowed a finely dispersed morphology at the start of the meltingsection and the screw can feed and mix the melted mixture simul-taneously. The use of unmodified and modified sugarcane bagassecellulose/HDPE composite with zirconium oxychloride formed bycompounding 10% by weight of fiber using extrusion and compres-sion molding showed that the modified composites present bettertensile strength compared to unmodified composites. The celluloseagglomerations were responsible for poor adhesion between the fi-ber and matrix in unmodified composites [4].

Rheological measurement conducted in various steady stateand dynamic environments is a widely used technique for deter-mining the sensitivity of a material during processing [5]. Fromthe findings conducted on multi-wall carbon nanotube(MWCNT)/poly(ether ether ketone) (PEEK) composites containing

260 F.Md. Salleh et al. / Composites: Part B 58 (2014) 259–266

up to 17 wt% filler compounded by using a twin screw extruder,the complex viscosity and moduli as determined by the linear vis-coelastic measurements increase with increasing MWCNT concen-tration and the storage modulus, G0 exhibits a dramatic seven orderincrease in magnitude around 1 wt%, leading to a solid-like behav-ior at low frequency especially at higher filler loadings [6]. Ranaet al. [7] have reported that the storage and loss moduli of shortjute fiber reinforced PP composites increased with fiber content.However injection molding, extrusion or mixing in an internalmixer and then molding involves high shear and therefore it mightdamage the natural fiber [8].

In this research, the processing parameter of kenaf fiber/HDPEcomposites using co-rotating twin screw extruder was investi-gated at two different temperatures defined as low processingtemperature (LPT) and high processing temperature (HPT). Themain aim of this work is to establish the effect of this extrusionparameter on the rheological, dynamic mechanical and tensileproperties.

2. Experimental

2.1. Materials

Kenaf bast fiber of 3 mm length with the average density of134.3 kg/m3 was obtained from the National Kenaf & TobaccoBoard, Malaysia. It was sieved and fibers with diameter of less than0.5 mm were collected. A semi-crystalline high density polyethyl-ene (HDPE), Titanzex HI 1100 with a density of 961 kg/m3 and meltflow index of 7 g/10 min, supplied by Titan Petchem (M) Sdn. Bhd.,Malaysia was used as the matrix.

2.2. Extrusion and compression molding

Kenaf fiber/HDPE composites were prepared by melt-com-pounding using a co-rotating twin screw extruder with gravimetricmetering device feeder (Brabender KETSE 20/40 Lab Compounder,Germany). The screw has a diameter of 20 mm and length to diam-eter ratio of 40. Extrusion was carried out at screw speed of 80 rpmat 2 kg/h feeding rate with two different temperature settings, i.e.160, 165, 170, 175 and 180 �C from the hopper to the die and 165,170, 175, 180 and 185 �C, designated as LPT and HPT respectively.These processing temperatures produced the actual melt tempera-tures of between 190–195 �C and 194–198 �C for LPT and HPTrespectively. Pure HDPE was loaded automatically into the feedhopper by using gravimetric metering device while kenaf fiberwas introduced to the barrel at the side feeder between zones 3and 4 (decompression zone). The strands leaving the extruder diewere pelletized and molded using the compression molding ma-chine at temperature of 155 �C and pressure of ca 98 MPa(100 kgf/cm2).

2.3. Rheological study

For the rheological study, materials were subjected under dy-namic frequency sweep (DFS) mode using a Physica MCR 301 rhe-ometer (Anton Paar, Germany). A cone-plate measuring systemwas used with geometry of 25 mm diameter, 0.051 mm gap andangle 2h. Measurements were performed over frequency rangefrom 0.05 to 500 rad/s with 5% strain amplitude at temperatureof 190 �C.

2.4. Dynamic mechanical properties

The dynamic mechanical properties of composites were ana-lyzed using a dynamic mechanical analyzer model Q800 (TA

Instruments, USA) in thin film mode. DMA specimens were cutfrom the compression molded sheet to a dimension of 30 mmlength, 6.3 mm width and 0.13 mm thick. Measurements wereconducted over a temperature range from �135 �C to 100 �C witha heating rate of 3 �C/min and constant frequency of 1.0 Hz.

2.5. Tensile properties

Tensile tests were conducted according to ASTM D-638 stan-dard by using a universal testing machine (Instron 5567, USA)equipped with a load cell of 10 kN at a constant cross-head speedof 5 mm/min, and a gauge length of 50 mm. Dumb-bell shaped ten-sile test specimens were cut to the dimension of 75 mm length by6 mm width by 1 mm thick. For each test, a minimum of sevensamples were tested.

3. Results and discussion

Since there is no application of fiber surface treatment or cou-pling agent materials during compounding to improve fiber–ma-trix interfacial adhesion, it is assumed that any changes inmaterials properties are due to the quality of fiber dispersion with-in the matrix, compounded at LPT and HPT.

3.1. Rheological properties (DFS)

Rheology is a study of flow behavior of liquids or deformationbehavior of solid material when it is being processed. Many factorsincluding temperature, pressure, screw speed, type of material,amount of filler loading and diameter of die influence the qualityof the extrusion product. Extrusion process requires that the poly-mer composite be subjected to temperatures above its meltingpoint to enable deformation and flow while the screw speed isresponsible for feeding and mixing the melted material simulta-neously so as to be forced into the extruder barrel and die. Rheo-logical properties can reflect the internal structure and processability of materials [9]. DFS, an oscillatory test, has been performedat variable angular frequencies, keeping the amplitude at a con-stant value to give the complex viscosity, storage and loss modulibehavior of the kenaf fiber/HDPE composites.

3.1.1. Complex viscosityThe complex viscosity, g� is defined by,

jg�j ¼ jG�j

xð1Þ

where G� and x are the complex shear modulus and angular fre-quency respectively. Fig. 1 shows the complex viscosity curves asa function of angular frequency for the kenaf fiber/HDPE compositesobtained at LPT and HPT with different fiber loadings. g� of pureHDPE and 3.4 wt% fiber composite shows a Newtonian plateau untilangular frequency of 101 rad/s, beyond which non-Newtonianbehavior prevails. On the contrary, composites with 8.5 and17.5 wt% fiber display non-Newtonian behavior over the entirerange of frequencies for both at LPT and HPT.

The zero-shear viscosity, go, is a very important value since it isproportional to the average molar mass, Mw. The Carreau-Yasuda 1regression from the complex viscosity curve was used to determinethe go of the composites. The Carreau-Yasuda 1 regression is de-fined by,

g ¼ go � g11þ ðkxÞa� �ð1�nÞ=a þ g1; go � g1 > 0 ð2Þ

where g, go, g1, k, a and n are the viscosity, zero shear viscosity,infinite-shear viscosity, relaxation time, power law exponent and

Fig. 1. The complex viscosity-angular frequency behavior of composites extrudedat different processing temperature and fiber loading.

F.Md. Salleh et al. / Composites: Part B 58 (2014) 259–266 261

width of the transition range respectively. Data of go from Carreau-Yasuda 1 regression is tabulated in Table 1. In relation with Table 1and Fig. 1, it shows that pure HDPE and 3.4 wt% fiber compositeswith go 62340 Pa s shows Newtonian behavior, while compositeswith 8.5 and 17.5 wt% fibers display non-Newtonian behavior withgo P 30,970 Pa s. From the go data, average of molar mass value, Mw

was calculated using equation,

go ¼ keMaw ð3Þ

where the parameters ke and Mw are the material constant andaverage of molar mass respectively. The relationship betweenmolecular features and rheological properties is a topic of long-standing discussion in polymer science, especially the go–Mw

dependence, where the power law exponent, a, takes values varyingbetween 3.36 and 3.64 for polyethylene, and the material constant,ke depends on temperature [10,11]. Recent work has shown that therelationship between go and Mw for the two species is indistin-guishable at 175 �C, and those results with a very minor tempera-ture adjustment to 190 �C using Ea as 6.4 kcal/mol for linearpolyethylene [12–14] lead to,

go ¼ 3:4� 10�14 M3:6w ð4Þ

for Mw P 3200 [15]. The ke and a value from the above literaturewas used to calculate the average molar mass and tabulated inTable 1. From the table, it can be seen that the average molar massincreases with increasing fiber loading.

Table 1Zero shear viscosity and average of molar mass data of extruded kenaf fiber/HDPEcomposites at two different setting temperature.

Fiber fraction(wt%)

Barrelsetting

Zero shear viscosity,go (Pa s)

Average molar mass,Mw (g/mol)

0 LPT 710 34,152HPT 960 37,125

3.4 LPT 2339 47,545HPT 2020 45,648

8.5 LPT 30,971 97,441HPT 36,750 102,184

17.5 LPT 123,980 143,244HPT 159,620 153,659

Extrusion at HPT also shows a higher go and higher Mw for allsamples compared to the one processed at LPT except for the com-posites at 3.4 wt% fiber loading (Table 1). Moreover, composites ex-truded at HPT show higher g� compared to those at LPT as observedin 0.0, 8.5 and 17.5 wt% fiber composites (Fig. 1). The compositeswith 3.4 wt% fiber did not show significant difference when theprocessing temperature was changed. Extruding the compositesat HPT shows higher g� compared to the ones at LPT measured atboth low and high frequencies. At HPT, there is enough heat andenergy to melt the matrix and at the same time have the sufficientmelt strength to produce a more homogeneous dispersion of fiberwithin the matrix. It is also believed that at this processing temper-ature, the optimum stress relaxation occur and therefore shows ahigher g� compared to that at LPT. However, LPT does not seemto have sufficient heat to fully melt the matrix. The matrix becomesmore viscous and the fiber will be difficult to disperse in the ma-trix. This consequently creates inhomogeneous composites, result-ing in reduction of melt strength and complex viscosity. At lowtemperature, the viscosity as well as the shear stress generatedin the mixture is very high, causing the breakdown of the fibersduring mixing [16].

Composites at 3.4 wt% fiber loading exhibit an almost similarpattern as pure HDPE, i.e. Newtonian plateau at the low frequencyregion. This is believed to be due to the low loading level of kenaffiber in HDPE matrix. In the higher filled composite systems, a sig-nificant increase in g� can be observed with increasing fiber load-ing at all frequency range, i.e. non-Newtonian behavior.Appearance of shear-thinning behavior of g� for 8.5 and 17.5 wt%fiber composites at low frequencies indicates the transformationfrom liquid-like to pseudo-solid like response. The presence of fi-ber with higher fiber loading perturbs the normal flow of polymerand therefore hinders the mobility of polymer chain segments. Thisphenomenon is attributed to shear thinning effect, which is pre-dominant in the composite materials compared to a pure matrix.This behavior is due to the fact that at low oscillatory frequencies,disoriented fiber collisions dominate the flow compared to the sit-uation at high frequencies, where matrix behavior is the dominantfactor [5].

3.1.2. Storage and loss moduliThe storage and loss moduli as a function of frequency of kenaf

fiber/HDPE composites are presented in Figs. 2 and 3 respectively.Both storage, G0 and loss, G00 moduli curves of pure HDPE and3.4 wt% fiber composites increase progressively with increasingfrequency, showing a terminal behavior. Meanwhile at 8.5 and17.5 wt% fiber loadings, the slope of G0 and G00 level off especiallyat the low frequency range indicating the nonlinear pattern.Fig. 2 shows the slope values of G0 for pure HDPE at LPT and HPTare almost the same (G0 �x1.1). However, pure HDPE extruded atLPT shows the minimum magnitude of G0 value all the way fromlow frequency to high frequency regions. In addition, the slope va-lue of G00 for pure HDPE at LPT gives the lowest value of G00 �x0.87

(Fig. 3). From the above results, the relaxation exponents of storageand loss moduli for pure HDPE are less than what is expected fromthe linear viscoelasticity theory, where G0 �x2 and G00 �x1

respectively. According to the linear viscoelasticity theory, thestorage and loss moduli of homogeneous polymer systems obeyscaling law behavior in the low frequency region, with slopes equalto 2 and 1 respectively. The linear viscoelasticity at low frequencies(terminal zone) reflects a fully relaxed polymer chain [6]. Similar tog� finding, composites with 3.4 wt% fiber did not show a significantdifference on G0 and G00 with changing of processing temperature.The development of plateau at low frequencies can be seen forthe 8.5 and 17.5 wt% fiber composites, where G0 and G00 values devi-ate from that of pure HDPE due to the effect of fiber loading.

Fig. 2. The storage modulus – angular frequency behavior of composites extrudedat different processing temperature and fiber loading.

262 F.Md. Salleh et al. / Composites: Part B 58 (2014) 259–266

At higher fiber loading, the magnitudes of G0 and G00 of compos-ites compounded at HPT are greater than those compounded atLPT, suggesting a synergistic effect on the elastic behavior of thecomposites. Sample extruded at HPT with 17.5 wt% fiber loadingshows the highest magnitude of G0 and G00, thus revealing the rein-forcing effect imparted by the fiber. Bangarusampath et al. [6] re-ported that for MWCNT/PEEK composites, G0 and G00 often deviatefrom the linear viscoelasticity behavior. The viscosity of the matrixpolymer increases with the incorporation of fiber, which can beseen with the increase of the storage modulus. At loading above1 wt%, the terminal behavior disappears and instead tends to a pla-teau-like region that is indicative of a transition from liquid-like tosolid-like viscoelastic behavior. This critical composition (around1 wt%), correlates with the transitions observed in the other rheo-logical data and can be identified as a rheological percolationthreshold concentration. An investigation on fiber-filled polymericmelt claimed that formation of a plateau at a very low applied fre-quency indicates that relaxation of the polymer composites is con-centrated in the low frequency region. The authors also observedthat an apparent yield stress is exhibited due to the formation ofa network structure within the matrix [6]. However, in the present

Fig. 3. The loss modulus – angular frequency behavior of composites extruded atdifferent processing temperature and fiber loading.

study, the storage modulus increases considerably in the compos-ites at low frequency and no yield stress is observed. Thisphenomenon is attributed to an increase in the rigidity of the ma-trix with the incorporation of the fibers [5].

The change in rheological properties with fiber content was alsostudied using a log–log plot of G0 versus G00 (Han plot). This plotindicates the sensitivity of rheological properties to compositionand temperature [17,18]. Fig. 4 shows the Han plot of the kenaf fi-ber/HDPE composites, plotted with the data extracted from Figs. 2and 3. As can be seen, the Han plot shows that at these two differ-ent processing temperatures, the curves are not fitted into a singlecurve. HPT shows an upward shift compared to the LPT for pureHDPE, 8.5 and 17.5 wt% fiber loadings composites. This indicatesthat the rheological behavior of these samples is affected by theextrusion temperature significantly. The strong temperaturedependence of logG0 versus logG00 plot over a range of tempera-tures is attributable to the occurrence of a thermally induced tran-sition from an ordered microdomain structure to a disorderedhomogeneous phase [19]. From this, it is also clear that rheologicalproperties change with increase in fiber loading, given the Han plotshows an upward shift from pure HDPE to the highest filler contentof 17.5 wt% fiber. Unfortunately all the Han plots have a slope ofmuch less than 2 compared to the linear viscoelasticity theory. Itwas reported [20] that the logG0 versus logG00 for homopolypropyl-ene/maleated composite (PPVC-PP-g-MA) and copolymer ethylenepropylene/maleated composite (PPSC-PP-g-MA), the curves shiftupwards as the PP-g-MA content increases. In the oscillatory test,the deformation of the dispersed phase with very low viscosityshould induce and improve the deformation of the continuous PPphase. The deformation will become more prominent with higherPP-g-MA content. Consequently, the ratio of the amount of energystored (G0) to the amount of energy dissipated (G00) increases withan increase of the PP-g-MA content in the blend, that is the ratiovaries with the blend composition.

3.2. Dynamic mechanical analysis

3.2.1. Storage modulus, E0

Storage modulus is the elasticity properties that describes theenergy stored in the system. Change in E0 indicates the changesin rigidity and hence strength of the materials. Fig. 5 shows thestorage modulus (E0) curves of composites compounded at LPTand HPT. The storage modulus decreases with increase in temper-

Fig. 4. The Han plots of composites extruded at different processing temperatureand fiber loading.

Fig. 5. Storage modulus curves of composites compounded at (a) LPT and (b) HPT.

F.Md. Salleh et al. / Composites: Part B 58 (2014) 259–266 263

ature, and this is associated with softening of the matrix [21,22].Measurements of E0 at �130, �100, �75 and 25 �C indicating thestorage moduli within the glassy, leathery, rubbery plateau androom temperature regions respectively are summarized in Table 2.At �130 �C, LPT composites do not show any positive reinforce-ment imparted by the fibers that can withstand stress transfer atthe interface even at high fiber loading as shown by the negativevalues. However, the composites with 8.5 wt% fiber loading stillshows an improvement in reinforcement compared to the3.4 wt% fiber composites. Meanwhile, the 17.5 wt% fiber loadinghas the lowest percentage of reinforcement imparted by the fiber(�54%). Fig. 5(a) shows that 3.4 and 8.5 wt% fiber composites com-pounded at LPT have a higher storage modulus compared to thepure HDPE. While, 17.5 wt% fiber composites shows a lower stor-

Table 2DMA storage modulus of extruded kenaf fiber/HDPE composites.

Fiber fraction (wt%) Barrel setting Storage modulus, E0 (GPa)

E0 at �130 �C E0 at �100 �C E0 at

0 LPT 4.81 3.51 3.0HPT 8.77 6.27 5.4

3.4 LPT 6.06 4.03 3.5HPT 8.76 6.31 5.5

8.5 LPT 7.98 5.87 5.0HPT 10.65 7.81 6.8

17.5 LPT 4.02 2.95 2.5HPT 16.15 11.86 10.4

age modulus compared to pure HDPE at room temperature(Table 2). Processing at LPT with highest fiber loading of17.5 wt% fiber increases the shear between fiber–matrix–barrelto give an insufficient melt strength, resulting in poor fiber–matrixadhesions, thus limiting stress transfer between the matrix and thefiber and therefore causes a reduction in the storage modulus[7,21,22].

On the contrary, the storage modulus increases with increasingfiber loading when processed at HPT, as shown in Fig. 5(b). Atglassy region of �130 �C (Table 2), composites with 17.5 wt% fibershows the highest storage modulus value of 16.15 GPa with 84%reinforcement, followed by the composites of 8.5 wt% fiber with21% reinforcement compared to the pure HDPE (8.77 GPa). This isdue to the reinforcement imparted by the fibers, which allowsstress transfer from the matrix to the fiber [22–24]. As for the com-posites with 3.4 wt% fiber no significant difference is observedcompared to the pure sample at this processing temperature be-cause of low fiber loading. Low fiber loading may have acted asnotches, thus reducing reinforcing effectiveness [25]. HPT im-proves the dispersion of fiber within the matrix and gives betterreinforcement effect to the system, leading to increasing stiffnessof the matrix and shows a high storage modulus value. The highE0 values with higher filler loading over the range of temperaturesare associated with better fiber dispersion within the matrix[26,27].

3.2.2. Loss modulus, E00

Loss modulus is the viscous component that describes the en-ergy dissipated during a process and is susceptible to molecularmotions [7,22]. Variations of loss modulus as a function of temper-ature for composite specimens at LPT and HPT are graphically plot-ted in Fig. 6. From Fig. 6(a), pure HDPE, 8.5 and 17.5 wt% fibercomposites exhibit three transition peaks (a, b and c) while3.4 wt% fiber composite exhibits only two transition peaks (a andc). The a-transition temperature within the temperature range of38.5–66.5 �C (Table 3) is associated to the motion of long-chain(–CH2–)n segments in the crystalline region of the HDPE. The b-transition temperature, which is around -32.0 to -26.0 �C, is relatedto the segmental motion of (-CH3) relaxation in the amorphous re-gion of the HDPE and the c-transition between �120.0 and�117.0 �C, is attributed to long chain (–CH2–)n crankshaft relaxa-tion in the amorphous polyethylene segments of the HDPE chain.Khanna et al. [28] reported that the b and c-transitions for HDPEwere at �45 �C and �107 �C (±1 �C) respectively. Transitions at b-relaxation are associated with the branch point. This relaxation istherefore generally attributed to segmental motions of methylgroups in the non-crystalline phase. At LPT, the b-transition peakof the composites was shifted from about 5.5 and 6.0 �C to a lowertemperature compared to the pure HDPE as observed in the case of8.5 and 17.5 wt% fiber composites. A lower b-transition tempera-ture implies the presence of some processes, which have possibly

Reinforcement imparted by the fiber at �130 �C (%)

�75 �C E0 at 25 �C

3 1.80 –2 3.20

4 2.25 �312 3.17 0

9 3.04 �90 4.24 21

9 1.47 �542 5.72 84

Fig. 6. Loss modulus curves of composites compounded at (a) LPT and (b) HPT.Fig. 7. Mechanical loss factor curves of composites compounded at (a) LPT and (b)HPT.

264 F.Md. Salleh et al. / Composites: Part B 58 (2014) 259–266

led to the softening of the matrix and increasing the ability of thematrix chains to move more freely [29].

Fig. 6(b) shows a variation of the loss modulus of compositescompounded at HPT. At HPT, only the 17.5 wt% fiber compositeexhibits three transition peaks (a, b and c) while compositeswith 3.4 and 8.5 wt% fiber exhibit only two transition peaks(a and c). The a, b and c-transition temperatures are foundto be within 43.5–62.0 �C, �22.5 to �2.5 �C and between�119.0 and �116.0 �C respectively (Table 3). Processing at high-er temperature with 17.5 wt% fiber composite shows betterproperties of loss modulus compared to the pure HDPE. At17.5 wt% fiber composites, processing at this condition givesthe highest magnitude of E00a (625 MPa), E00b (727 MPa) and E00c(665 MPa).

Table 3DMA thermomechanical data of extruded kenaf fiber/HDPE composites.

Fiber fraction (wt%) Barrel setting Loss modulus, E00 (MPa)

a-Transition b-Transitio

E00 max (MPa) Ta (�C) E00 max (M

0 LPT 173 55.0 107HPT 297 57.0 183

3.4 LPT 217 66.5 –HPT 289 58.0 –

8.5 LPT 288 58.0 158HPT 359 62.0 –

17.5 LPT 152 38.5 140HPT 625 43.5 727

3.2.3. Tan DeltaMechanical loss factor curves of composite specimens com-

pounded at LPT and HPT are shown in Fig. 7. The tandmax peakcan also provide information on the glass transition temperature,Tg, and energy dissipation of composite materials [30]. FromFig. 7(a), pure HDPE presents the highest magnitude of tandmax

(6.23 � 10�2) at LPT. Composites with 3.4 wt% fiber loading showthe highest magnitude of tandmax (5.82 � 10�2) compared to otherfiber loadings. The damping or tand in the transition regionmeasures the imperfections in the elasticity of a polymer [31].The higher damping factor relates to a poor fiber dispersion withinthe matrix. At 17.5 wt% fiber loading, the tandmax value is found tobe higher compared to the composite at 8.5 wt% fiber loading. Its Tg

value also shifted slightly to the lower temperature (�115.5 �C)

Tand

n c-Transition Tandmax (�10�2) Ta (�C)

Pa) Tb (�C) E00 max (MPa) Tc (�C)

�26.0 263 �119.0 6.23 �113.5�22.5 371 �116.0 4.97 �113.5

– 299 �120.0 5.82 �113.0– 385 �119.0 5.07 �113.5

�31.5 342 �117.0 4.94 �112.0– 429 �116.0 4.66 �112.0

�32.0 191 �120.0 5.38 �115.5�2.5 665 �117.5 4.33 �113.5

Fig. 8. The tensile modulus and tensile strength of pure HDPE and kenaf fiber/HDPEcomposites.

Fig. 9. The tensile strain of pure HDPE and kenaf fiber/HDPE composites.

F.Md. Salleh et al. / Composites: Part B 58 (2014) 259–266 265

compared to the others (Table 3), which indicates an increase inmolecular mobility of composite with LPT.

From Fig. 7(b) composites compounded at HPT show that themagnitude of tandmax values reduce with increasing fiber loadingexcept for the 3.4 wt% fiber composite. The higher reduction oftandmax values for composites with higher fiber loading is believedto be due to the strengthening effect imparted by the fibers. This iscaused by the limited mobility of polymer matrix which was af-fected by the improvement in fiber dispersion in the composites[24]. In composites, as the amount of incorporated fiber increases,the contact area between the fiber and the polymer matrix also in-creases, leading to a stronger interaction between the fiber andmatrix. Hence, the chain mobility of HDPE around the fiber isrestrained, resulting in the improvement in the hysteresis of thesystem and a reduction in the internal friction [21]. Therefore, withthe presence of kenaf fiber, the molecular mobility of the polymericmaterials decreases and the mechanical loss to overcome intermo-lecular chain friction is reduced [32]. The increasing magnitude oftandmax at 3.4 wt% fiber composites is expected because the highvolume of matrix is able to dissipate the vibrational energy andgives higher damping at the interface.

3.3. Tensile properties

3.3.1. Tensile modulusThe tensile modulus values and tensile strengths of pure HDPE

and kenaf fiber/HDPE composites are presented in Fig. 8. At HPT,pure HDPE and kenaf fiber/HDPE composites show a higher tensilemodulus compared to those processed at LPT. The tensile modulusvalues increase with increasing fiber loading when processed atHPT. The increment of modulus values with increasing amount offiber loading is in agreement with what was reported by Girunet al. [33], where a higher fiber loading composites are able to with-stand more load and can fabricate the composites to become stiffer[34]. In comparison to the pure HDPE at HPT, the tensile modulus ofcomposites containing 3.4, 8.5 and 17.5 wt% kenaf fiber increase by8.7%, 10.9% and 12.2% respectively. This shows that the incorpora-tion of kenaf fiber at HPT improves the tensile modulus of HDPE,indicating that transferring of stress from the polymer matrix tothe stiffer fiber has occurred [35]. On the contrary, LPT compositesshow a decrease in tensile modulus compared to the HPT compos-ites for all samples, especially for the highest fiber loading of17.5 wt%. The tensile modulus of composite at 17.5 wt% kenaffiber decreases by 29% when processed at LPT compared to the HPT.

3.3.2. Tensile strengthThe tensile strength of the composites is shown to decrease

with increasing content of kenaf fiber at both LPT and HPT(Fig. 8). This could be attributed to the poor adhesion between fiberand matrix, resulting in the poor interfacial interaction and deb-onding of the matrix from the fiber during the tensile deformation.The debonding results in void formation, which lowers the tensilestrength because cracks can easily propagate through regions con-taining the voids [21,31]. Nevertheless, the HPT composites showsbetter tensile strengh compared to the LPT composites. LPT resultsin inconsistent melt of the resin that can lead to non-uniform dis-persion of the fibers in the composites and eventually lowers thetensile strength [36].

3.3.3. Tensile strainFig. 9 shows tensile strain of pure HDPE and kenaf fiber/HDPE

composites. Generally, the addition of fiber reduces the tensilestrain of composites. The tensile strains of the LPT compositesare higher compared to the HPT composites and greatly decreasewith fiber loading. The elongation probably arises from the poly-mer matrix because the kenaf fiber is rigid relative to HDPE.

Increasing the amount of filler decreases the amount of polymeravailable for the elongation [31]. Moreover, adding higher amountof fiber increases the possibility of fiber agglomeration. Such anagglomeration can lead to the formation of stress concentratedregion where less energy is required for elongating the crack prop-agation [37].

4. Conclusion

Rheological data on DFS measurements of HPT specimensshowed higher complex viscosity, zero shear viscosity and averagemolar mass for all samples compared to the LPT composites, exceptfor the low fiber loading of 3.4 wt%. Complex viscosity of pureHDPE and 3.4 wt% fiber composite with go 62340 Pa s showedNewtonian behavior. However, composites with higher fiberloadings of 8.5 and 17.5 wt%, the complex viscosity curves withgo P 30,970 Pa s displayed non-Newtonian behavior. In addition,storage (G0) and loss (G00) moduli curves of pure HDPE and3.4 wt% fiber composites increased progressively with increasingangular frequency. Meanwhile at 8.5 and 17.5 wt% fiber loadings,the slope of G0 and G00 were found to level off especially at thelow frequency range, indicating the nonlinear pattern due to shearthinning. The Han plots showed that at different processing tem-peratures the curve were not fitted into a single curve but an

266 F.Md. Salleh et al. / Composites: Part B 58 (2014) 259–266

upward shift with HPT for pure HDPE, 8.5 and 17.5 wt% fiber load-ings compared to the LPT. Moreover, increasing the fiber loadingshowed an upward shift of Han plot from pure HDPE to the highestfiber content of 17.5 wt%.

DMA showed that the storage and loss moduli of composites in-creased with increasing fiber loading and 17.5 wt% fiber compositeshowed the highest value of storage and loss moduli at HPT. How-ever, composite with the highest fiber loading (17.5 wt%) com-pounded at LPT showed the lowest storage and loss modulicurves. The incorporation of the kenaf fiber at HPT reduced themagnitude of tandmax values with the increasing fiber loading ex-cept for the low loading fiber of 3.4 wt%. Storage and loss modulias a function of temperature appeared to be more sensitive tothe fiber loading at LPT while this was not observed in the storageand loss moduli as a function of angular frequency.

By increasing the compounding temperature, the tensile modu-lus presented an increasing pattern compared to the ones pro-cessed at LPT. The tensile modulus also increased with increasingfiber loading when processed at HPT. However, the tensile strengthof the composites decreased with increasing content of the fiber atboth LPT and HPT. Nevertheless, composites processed at HPTshowed better tensile strength compared to the ones at LPT. Thetensile strains of the composites decreased with the increasingamount of fiber loading while composites processed at LPT pre-sented a higher tensile strain compared to those at HPT. Fromthe above, it can be concluded that the composites compoundedat HPT provides best performance of rheological, thermo-mechan-ical and tensile properties in comparison with composites com-pounded at LPT.

Acknowledgements

This research was supported by the University of Malayathrough Grant Nos. IPPP (PS347-2010A and PV003-2011B) andRG150-11AFR.

References

[1] Narayan R. Biomass (renewable) resources for production of materials,chemicals, and fuels. In: Emerging technologies for materials and chemicalsfrom biomass. ACS symposium series, (American Chemical Society), vol. 476;1992. p. 1–10.

[2] Carneiro OS, Covas JA, Ferreira JA, Cerqueira MF. On-line monitoring of theresidence time distribution along a kneading block of a twin-screw extruder.Polym Test 2004;23(8):925–37.

[3] Potente H, Krawinkel S, Bastian M, Stephan M, Pötschke P. Investigation of themelting behavior and morphology development of polymer blends in themelting zone of twin-screw extruders. J Appl Polym Sci2001;82(8):1986–2002.

[4] Mulinari DR, Voorwald HJC, Cioffi MOH, da Silva MLCP, da Cruz TG, Saron C.Sugarcane bagasse cellulose/HDPE composites obtained by extrusion. ComposSci Technol 2009;69(2):214–9.

[5] Mohanty S, Nayak SK. Rheological characterization of HDPE/sisal fibercomposites. Polym Eng Sci 2007;47(10):1634–42.

[6] Bangarusampath DS, Ruckdäschel H, Altstädt V, Sandler JKW, Garray D, ShafferMSP. Rheology and properties of melt-processed poly(ether ether ketone)/multi-wall carbon nanotube composites. Polymer 2009;50(24):5803–11.

[7] Rana AK, Mitra BC, Banerjee AN. Short jute fiber-reinforced polypropylenecomposites: dynamic mechanical study. J Appl Polym Sci 1999;71(4):531–9.

[8] George G, Tomlal Jose E, Akesson D, Skrifvars M, Nagarajan E, Joseph K.Viscoelastic behaviour of novel commingled biocomposites based onpolypropylene/jute yarns. Compos Part A, Appl Sci Manuf2012;43(6):893–902.

[9] Lee SH, Cho E, Jeon SH, Youn JR. Rheological and electrical properties ofpolypropylene composites containing functionalized multi-walled carbonnanotubes and compatibilizers. Carbon 2007;45(14):2810–22.

[10] Aguilar M, Vega JF, Sanz E, Martinez-Salazar J. New aspects on the rheologicalbehavior of metallocene catalyzed polyethylenes. Polymer2001;42(24):9713–21.

[11] Vega J, Otegui J, Ramos J, Martínez-Salazar J. Effect of molecular weightdistribution on Newtonian viscosity of linear polyethylene. Rheol Acta2012;51(1):81–7.

[12] Raju VR, Smith GG, Marin G, Knox JR, Graessley WW. Properties of amorphousand crystallizable hydrocarbon polymers. I. Melt rheology of fractions of linearpolyethylene. J Polym Sci: Polym Phys Ed 1979;17(7):1183–95.

[13] Mendelson RA, Bowles WA, Finger FL. Effect of molecular structure onpolyethylene melt rheology. I. Low-shear behavior. J Polym Sci Part A-2:Polym Phys 1970;8(1):105–26.

[14] Saeda S, Yotsuyanagi J, Yamaguchi K. The relation between melt flowproperties and molecular weight of polyethylene. J Appl Polym Sci1971;15(2):277–92.

[15] Wasserman SH, Graessley WW. Prediction of linear viscoelastic response forentangled polyolefin melts from molecular weight distribution. Polym Eng Sci1996;36(6):852–61.

[16] Joseph PV, Joseph K, Thomas S. Effect of processing variables on themechanical properties of sisal-fiber-reinforced polypropylene composites.Compos Sci Technol 1999;59(11):1625–40.

[17] Azizi H, Ghasemi I, Karrabi M. Controlled-peroxide degradation ofpolypropylene: rheological properties and prediction of molecular weightdistribution (MWD) from rheological data. Polym Test 2008;27(5):548–54.

[18] Azizi H, Ghasemi I. Investigation on the dynamic melt rheological properties ofpolypropylene/wood flour composites. Polym Compos 2009;30(4):429–35.

[19] Han CD, Kim J, Kim JK. Determination of the order-disorder transitiontemperature of block copolymers. Macromolecules 1989;22(1):383–94.

[20] Li S, Järvelä PK, Järvelä PA. Melt rheological properties of polypropylene-maleated polypropylene blends. II. Dynamic viscoelastic properties. J ApplPolym Sci 1999;71(10):1649–56.

[21] Liang Z, Pan P, Zhu B, Dong T, Inoue Y. Mechanical and thermal properties ofpoly(butylene succinate)/plant fiber biodegradable composite. J Appl PolymSci 2010;115(6):3559–67.

[22] Huda MS, Drzal LT, Mohanty AK, Misra M. Chopped glass and recyclednewspaper as reinforcement fibers in injection molded poly(lactic acid) (PLA)composites: a comparative study. Compos Sci Technol 2006;66(11–12):1813–24.

[23] Huda MS, Drzal LT, Mohanty AK, Misra M. Effect of fiber surface-treatments onthe properties of laminated biocomposites from poly(lactic acid) (PLA) andkenaf fibers. Compos Sci Technol 2008;68(2):424–32.

[24] Hassan A, Rahman NA, Yahya R. Extrusion and injection-molding of glass fiber/MAPP/polypropylene: effect of coupling agent on DSC, DMA, and mechanicalproperties. J Reinforced Plast Compos 2011;30(14):1223–32.

[25] Hassan A, Yahya R, Rafiq MIM, Hussin A, Sheikh MRK, Hornsby PR. Interfacialshear strength and tensile properties of injection-molded, short- and long-glass fiber-reinforced polyamide 6,6 composites. J Reinforced Plast Compos2011;30(14):1233–42.

[26] Aziz SH, Ansell MP. The effect of alkalization and fibre alignment on themechanical and thermal properties of kenaf and hemp bast fibre composites:Part 1 – polyester resin matrix. Compos Sci Technol 2004;64(9):1219–30.

[27] Sanadi AR, Caulfield DF, Jacobson RE, Rowell RM. Renewable agricultural fibersas reinforcing fillers in plastics: mechanical properties of kenaf fiber-polypropylene composites. Ind Eng Chem Res 1995;34(5):1889–96.

[28] Khanna YP, Turi EA, Taylor TJ, Vickroy VV, Abbott RF. Dynamic mechanicalrelaxations in polyethylene. Macromolecules 1985;18(6):1302–9.

[29] Tajvidi M, Falk RH, Hermanson JC. Effect of natural fibers on thermal andmechanical properties of natural fiber polypropylene composites studied bydynamic mechanical analysis. J Appl Polym Sci 2006;101(6):4341–9.

[30] Kim H-S, Kim S, Kim H-J, Yang H-S. Thermal properties of bio-flour-filledpolyolefin composites with different compatibilizing agent type and content.Thermochimica Acta 2006;451(1–2):181–8.

[31] Bleach NC, Nazhat SN, Tanner KE, Kellomäki M, Törmälä P. Effect of fillercontent on mechanical and dynamic mechanical properties of particulatebiphasic calcium phosphate–polylactide composites. Biomaterials2002;23(7):1579–85.

[32] Fay JJ, Murphy CJ, Thomas DA, Sperling LH. Effect of morphology, crosslinkdensity, and miscibility on interpenetrating polymer network dampingeffectiveness. Polym Eng Sci 1991;31(24):1731–41.

[33] Girun N, Ahmadun FR, Rashid SA, Ali Atieh M. Multi-wall carbon nanotubes/styrene butadiene rubber (SBR) nanocomposite. Fullerenes, Nanotubes CarbonNanostruct 2007;15(3):207–14.

[34] Salleh MAM, Abd J, Razak NAI, Razi AFL, Suraya A. The influences of melt-compounding parameters on the tensile properties of low filler loading ofuntreated-MWCNTs-polypropylene (PP) nanocomposites. J Eng Sci Technol2008;3(1):97–108.

[35] Pan P, Zhu B, Kai W, Serizawa S, Iji M, Inoue Y. Crystallization behavior andmechanical properties of bio-based green composites based on poly(L-lactide)and kenaf fiber. J Appl Polym Sci 2007;105(3):1511–20.

[36] Ku H, Wang H, Pattarachaiyakoop N, Trada M. A review on the tensileproperties of natural fiber reinforced polymer composites. Compos Part B: Eng2011;42(4):856–73.

[37] Huda MS, Mohanty AK, Drzal LT, Schut E, Misra M. ‘‘Green’’ composites fromrecycled cellulose and poly(lactic acid): physico-mechanical andmorphological properties evaluation. J Mater Sci 2005;40(16):4221–9.