Composites: Part A 42 (2011) 310319

Contents lists available at ScienceDirect

Composites: Part A
journal homepage: www.elsevier.com/locate/compositesa

Improvement of mechanical performance of industrial hemp bre reinforced polylactide biocomposites

Moyeenuddin A. Sawpan a,, Kim L. Pickering a, Alan Fernyhough b
a b

Department of Engineering, University of Waikato, Hamilton, New Zealand Biomaterials Engineering, Biopolymer Network/SCION, Rotorua, New Zealand

a r t i c l e

i n f o

a b s t r a c t
In this work, mechanical properties of chemically treated random short bre and aligned long hemp bre reinforced PLA composites were investigated over a range of bre content (040 wt.%). It was found that tensile strength, Youngs modulus and impact strength of short hemp bre reinforced PLA composites increased with increased bre content. Alkali and silane bre treatments were found to improve tensile and impact properties which appears to be due to good bre/matrix adhesion and increased matrix crystallinity. A 30 wt.% alkali treated bre reinforced PLA composite (PLA/ALK) with a tensile strength of 75.5 MPa, Youngs modulus of 8.18 GPa and impact strength of 2.64 kJ/m2 was found to be the best. However, plane-strain fracture toughness and strain energy release rate decreased with increased bre content. The mechanical properties of the PLA/ALK composites were increased further due to alignment of long bres. 2010 Elsevier Ltd. All rights reserved.

Article history: Received 14 August 2010 Received in revised form 30 October 2010 Accepted 5 December 2010 Available online 10 December 2010 Keywords: A. Polymermatrix composites B. Mechanical properties E. Injection moulding E. Compression moulding

1. Introduction In recent years, polymer matrices from renewable resources are gaining ground over conventional petroleum based matrices (e.g. polyethylene, polypropylene, etc.) because of environmental problems related to their disposal as well as concerns over petroleum availability [1]. Polylactide (PLA) is the rst commodity plastic produced from annually renewable resources. For a long time, PLA was mainly used in biomedical applications because of its high production costs. However, recent developments in the manufacture of its monomer (i.e. lactic acid) economically from agricultural products (e.g. corn, potato and cane sugar) have placed this material at the forefront of the emerging plastic industries [25]. Due to the commercial potential for natural bre reinforced polymer composites in automotive applications and building construction as well as demands for environmentally friendly materials, the development of PLA based composites for many applications is an interesting area of research. For instance, Oksman et al. [6] produced ax bre reinforced PLA composites by extrusion and compression moulding. They found that tensile strength of the composites increased slightly at 30 wt.% bre content (53 MPa) compared to PLA only samples (50 MPa). However, the tensile strength of 40 wt.% bre samples decreased to 44 MPa. This decrease in tensile strength at
Corresponding author. Present address: Composite Materials Research, Pultron Composites Ltd., 342 Lytton Road, Gisborne 4040, New Zealand. Tel.: +64 6 867 8582; fax: +64 6 867 8542. E-mail address: moyeen@pultron.com (M.A. Sawpan).
1359-835X/$ - see front matter 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.compositesa.2010.12.004

higher bre content may be due to: (i) inadequate amounts of matrix to wet the bres and (ii) reduction of bre length during processing. Youngs modulus at 30 and 40 wt.% bre content was found to be 8.3 and 7.3 GPa, respectively, which was signicantly higher than that of PLA only samples (3.4 GPa). In another study, Bax and Mussig [7] also used ax bres to reinforce PLA (PLA was in the form of bre) by hot-pressing followed by pelletising and injection moulding. It was found that tensile strength of the composites increased from 44.5 to 54.1 MPa and Youngs modulus increased from 3.1 to 6.31 GPa as the bre content increased from 0 to 30 wt.%. The above ndings were fairly consistent with other research work [6]. It was also observed that impact strength (unnotched samples) of the composites at all bre contents was lower than the PLA only samples. Garcia et al. [8] fabricated PLA composites reinforced with kenaf bre using extrusion and injection moulding. They also added maleated-PLA in the composites as compatibiliser. It was found that tensile strength of the composites reinforced with 30 wt.% bre decreased about 10% compared to unreinforced PLA samples, which was inconsistent with other works [6,7]. This could be due to the difference in reinforcement type and processing method. However, for the same composites, Youngs modulus increased about 71%, which was consistent with other reports [6,7]. They also found that impact strength of notched samples increased from 7.4 to 12.2 J/m with increased bre content (030 wt.%) but decreased in the case of unnotched samples (from 76 to 52 J/m). This trend in impact strength of the unnotched composite samples was in general agreement with other researchers [7]. Serizawa et al. [9] also used

M.A. Sawpan et al. / Composites: Part A 42 (2011) 310319

311

kenaf bres to fabricate PLA composites by extrusion and injection moulding. They found that impact strength of the notched samples decreased from 4.4 to 3.1 kJ/m2 as the bre content increased from 0 to 20 wt.%, which did not agree with Garcias ndings [8]. Suggestion was made that instead of using a twin screw extruder, composite impact strength can be improved by compounding the materials with a single-screw extruder which prevents the bre from being ground (crushed particles) during processing. Mathew et al. [10] used microcrystalline cellulose (MCC), wood our and wood pulp to reinforce PLA using similar processing methods to Oksman et al. [6]. They found that the tensile strength of MCC reinforced composites decreased (from 49.6 to 36.2 MPa) with increased MCC content (025 wt.%) whereas Youngs modulus increased signicantly (from 3.6 to 5 GPa) as for Garcia et al. [8] with kenaf bre. They also observed that the tensile strength, Youngs modulus and storage modulus of wood our and wood pulp reinforced composites were higher (for wood our composites tensile strength and Youngs modulus were 45.2 MPa and 6.3 GPa, respectively, and for wood pulp composites tensile strength and Youngs modulus were 45.2 MPa and 6 GPa, respectively) than those of MCC reinforced composites at similar level of reinforcement. In another study, Lee et al. [11] reinforced PLA with bamboo bres using batch mixing and compression moulding. They also found that Youngs modulus increased with increased bre content (1050 wt.%) but tensile strength decreased. In addition, they found that tensile strength and Youngs modulus improved at all bre contents when maleic anhydride treated bamboo bres (5 wt.%) were used as a compatibiliser and dicumyl peroxide as a free radical initiator. In a later report, Lee and Wang [12] applied a bio-coupling agent (lysine-based diisocyanate) as compatibiliser in the PLA/bamboo bre composites using similar processing method of [11]. As for Lee et al. [11], they observed that Youngs modulus of the composites increased with increased bre content (050 wt.%) but tensile strength decreased. In addition, tensile strength and Youngs modulus improved at all bre contents in the presence of coupling agent. Vila et al. [13] used eucalyptus wood bre and rice husks to reinforce PLA using extrusion and injection moulding. They did not observe any notable increase in tensile strength for the composites reinforced with 30 wt.% wood bre or rice husks compared to PLA only samples but Youngs modulus increased signicantly (57% increase for PLA/wood bre and 45% increase for PLA/rice husks composites). Pilla et al. [14] used silane treated pine wood our (PWF) to fabricate PLA composites by kinetic mixing and compression moulding. It was observed that tensile strength of the untreated bre composites was unchanged compared with that of PLA matrix (55.5 MPa) at 20 wt.% PWF content but decreased (51.7 MPa) at 40 wt.% PWF. Youngs modulus was found to increase signicantly with increased PWF content (0.63 GPa for PLA, 0.86 GPa for 20 wt.% PWF/PLA composites and 1.18 GPa for 40 wt.% PWF/PLA composites). Notable change in tensile strength and Youngs modulus was not found for the composites reinforced with 20 wt.% silane treated bre compared to the untreated bre composites. However, tensile strength of the silane treated bre reinforced composites (57.1 MPa) slightly increased at 40 wt.% bre content but with no signicant change in Youngs modulus compared to the untreated bre composites. Iwatake et al. [15] prepared micro-brillated cellulose (10 wt.%) reinforced PLA composites by kneading and compression moulding. They found that tensile strength and Youngs modulus of the composites increased by 25% and 40%, respectively, compared to PLA only samples. They also observed that further addition (15 wt.%) of micro-brillated cellulose caused a decrease in tensile strength. Plackett et al. [16] fabricated aligned long jute bre mat (40 wt.%) reinforced PLA composites, which were rst pressed and consolidated under vacuum at different temperatures (180 220 C) then compression moulded. They found that tensile

strength and Youngs modulus of the composites increased significantly compared to PLA only samples (tensile strength 55 MPa and Youngs modulus 3.5 GPa) at all processing temperatures. Composites processed at 210 C had the highest tensile strength (100.5 MPa) and Youngs modulus (9.4 GPa). This signicant increase in tensile strength and Youngs modulus compared to other ndings [6,7,17] appears to be due to the alignment of bres in loading direction. The above studies indicate that generally Youngs modulus of PLA composites can be improved by adding bres. This is because natural bres are very stiff compared to PLA matrix. However, tensile and impact strength of composite are greatly inuenced by bre type and processing method. Natural bres such as hemp, sisal, ax, kenaf and jute are covered with waxy materials, thus hindering the hydroxyl groups from reacting with polymer matrices. This can lead to the formation of ineffective interfaces between the bres and matrices, with consequent problems such as debonding and voids in resulting composites. Chemical treatments provide an important and effective means to remove non-cellulosic components in cellulose bres and add functional groups to enable better bonding in polymer composites. In this study, industrial hemp bres were subjected to different chemical treatments, namely alkali, silane and acetic anhydride, in an attempt to produce high strength and stiff PLA composites. 2. Materials and methods 2.1. Materials NatureWorks PLA (polylactide) polymer 4042D, from NatureWorks LLC, USA was used as a thermoplastic matrix. The industrial hemp bres were supplied by Hemcore Ltd., UK. [3-(2-Aminoethyl amino)propyl]trimethoxy silane was purchased from Aldrich and Sigma, respectively. All other chemicals used were of analytical grade obtained from local commercial sources. 2.2. Methods 2.2.1. Fibre treatment Prior to treatment, untreated bres (FB) were washed with hot water (50 C) to remove dirt. Afterwards, bres were dried in an oven at 80 C for 48 h. 2.2.1.1. Alkali treatment. Pre-dried bres were soaked in 5% sodium hydroxide solution at ambient temperature for 30 min. After treatment, bres were copiously washed with water to remove any traces of alkali on the bre surface and subsequently neutralised with 1% acetic acid solution. The treated bres (ALK) were then dried in an oven at 80 C for 48 h. 2.2.1.2. Silane treatment. A solution of 0.5 wt.% silane coupling agent [3-(2-aminoethyl amino)propyl trimethoxy silane] was prepared in acetone. The pH of the solution was adjusted to 3.5 with acetic acid and stirred continuously for 5 min. Fibres (67 wt.% moisture content) were then immersed in the solution for 45 min. After treatment, bres were removed from the solution and dried in oven at 65 C for 12 h. Finally, the bres (SIL) were thoroughly washed with water to remove chemical residues until a pH of 7 was obtained and then dried in an oven at 80 C for 48 h. Similar silane treatment procedures also employed for bres that were previously alkali treated (ALKSIL). 2.2.2. Fabrication of composites Chopped dried short bres (average length 4.9 mm) and PLA pellets were compounded (10, 20 and 30 wt.% bre) by using a

312

M.A. Sawpan et al. / Composites: Part A 42 (2011) 310319

ThermoPrism TSE-16-TC twin screw extruder for good mixing of bre and polymer. The extruded composite material was pelletised and dried at 80 C for 24 h and then injection moulded using a BOY15-S injection moulding machine. No processing aids or other additives were used. PLA and aligned long bre (average length 65 mm) composites were produced by compression moulding using lm-stacking at three different bre contents (30, 35 and 40 wt.%). Dried long bres were aligned using a hand carding machine from Ashford Handicrafts Limited, New Zealand. PLA lms (0.5 mm thick) were produced from dry pellets, using an extruder equipped with a coat hanger die. PLA sheets and bres were weighed prior to composite fabrication to determine the weight percentage of bres and matrix of the resulting composites. Stacks of PLA lms and bres were prepared by placing alternately PLA lms and aligned bre mats in a parallel array. Before pressing, these were placed between two Teon sheets in a stainless steel matched-die mould (220 150 3.5 mm3). The stacks of PLA and bres were pre-pressed at 185 C for 5 min keeping a constant pressure of 2 MPa using a hot press machine and afterwards compacted at elevated pressure of 5 MPa for 3 min. The assembly was consolidated under a pressure of 5 MPa until the mould was naturally cooled down to ambient temperature. Composite plaques were cut to desired shapes using a computer numerical controlled (CNC) mill. 2.2.3. Measurement of PLA crystallinity Crystallinity of PLA in the composites was measured using a DSC 2920-TA Instruments machine. All DSC scans were carried out at a scan rate of 10 C/min from room temperature to 200 C in the presence of air using samples of about 10 mg. The percent crystallinity (XDSC) of PLA was calculated by using the following equation [14]:

fore testing. Four samples were evaluated for each batch of samples. 2.2.5. Scanning electron microscope (SEM) Composite fracture surface morphology was studied using a Hitachi S-4000 and a S-4700 eld emission scanning electron microscopes. Hitachi S-4000 was operated at 5 kV and Hitachi S4700 was operated between 5 and 20 kV. All samples were ion sputter-coated with platinum and palladium to provide enhanced conductivity. Samples were mounted with carbon tape on aluminium stubs and then sputter-coated with platinum and palladium to make them conductive prior to SEM observation. 3. Results and discussion 3.1. PLA crystallinity in composites In Fig. 1, it is apparent that the crystallinity of PLA in composites increased with increased bre content which could be due to the increased availability of nucleation sites leading to the formation of increased transcrystallinity [21,22]. However, the increase of crystallinity was smaller as the bre content increased from 20 to 30 wt.% which could be due to the decrease of PLA content or space limitation. It is apparent that the crystallinity of PLA in the PLA/ALK, PLA/ALKSIL and PLA/SIL composites increased compared with that of the PLA/FB composites. This could be because impurities (i.e. wax and pectin) were removed from the bres after treatment [23], which in turn increased the number of nucleating sites (i.e. crystalline portion of cellulose [22]) of the bres. 3.2. Tensile properties Typical stress versus strain records of the PLA and composites are presented in Fig. 2a. It may be observed that the curves initially show linear deformation, followed by a small non-linear deformation prior to the attainment of the maximum strength. It can also be seen that curves became steeper with increased bre content. In Fig. 2b, it can be seen that the two sample halves were completely separated virtually perpendicular to the applied load, typical of unstable fracture. All the samples showed similar fracture behaviour.

X DSC % fDHf DHcc 100g=DHo f w

where DHo f = 93 J/g for 100% crystalline PLA, DHf is the enthalpy of melting, DHcc is the cold crystallisation enthalpy and w is the weight fraction of PLA in the composite. 2.2.4. Mechanical properties measurement 2.2.4.1. Tensile testing. Tensile testing was carried out according to the ASTM D 638-03 Standard Test Method for Tensile Properties of Plastics [18]. Five samples of each type were tested by an Instron 4042 tensile test machine. The cross-head speed was 5 mm/min. 2.2.4.2. Impact testing. The impact testing was carried out according to the EN ISO 179 Plastics Determination of Charpy impact strength [19] using a Ray-Ran Pendulum Charpy Impact System. The impact velocity was 2.9 m/s and the hammer weight of 0.475 kg. Dimensions of the samples were 80 8 3.5 mm3 with a single notch of 0.25 mm. Five replicates were evaluated for each batch of samples. 2.2.4.3. Fracture toughness testing. Mode I fracture toughness testing was carried out using single-edge-notched-bend (SENB) specimen according to the ASTM D 5045-99 Standard Test Methods for Plane-Strain Fracture Toughness and Strain Energy Release Rate of Plastic Materials [20]. LLOYD LR 100 K universal testing machine was used for this purpose. The Length (L), Span length (S), Width (W) and Thickness (B) of the specimens were 126, 56, 12.7 (0.03) and 3.3 (0.03) mm respectively, which satises the condition 2B < W < 4B standardised for SENB specimens used in planestrain fracture toughness of plastics. The initial crack length was 6.35 mm (0.005). The loading speed was 10 mm/min and the notch root of the specimens was sharpened using a razor blade be-

30

25

Crystallinity, XDSC (%)

20

15

10

PLA/FB PLA/ALK PLA/SIL PLA/ALKSIL

0 10 20 30

Fibre content (wt%)

Fig. 1. DSC crystallinity of PLA and composites. (For interpretation of the references to color in this gure legend, the reader is referred to the web version of this article.)

M.A. Sawpan et al. / Composites: Part A 42 (2011) 310319

313

70 60
D C B

50

Stress (MPa)

40 30 20 10 0 0.0
A

A: PLA B: PLA/10 wt% FB C: PLA/20 wt% FB D: PLA/30 wt% FB

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

Strain (%)

(a)

(b)

Fig. 2. (a) Typical stress versus strain curves for tensile testing of PLA and composites (PLA/FB), (b) photograph of the specimens after testing. (For interpretation of the references to color in this gure legend, the reader is referred to the web version of this article.)

80 75

Tensile strength (MPa)

70 65 60 55 50 45 0 10 20 30

Young's modulus (GPa)

PLA/FB PLA/ALK PLA/SIL PLA/ALKSIL

3 0 10 20

PLA/FB PLA/ALK PLA/SIL PLA/ALKSIL

30

Fibre content (wt%)

Fig. 3. Tensile strength of untreated and treated hemp bre reinforced PLA composites as a function of bre content. (For interpretation of the references to color in this gure legend, the reader is referred to the web version of this article.)

Fibre content (%)

Fig. 4. Youngs modulus of untreated and treated hemp bre reinforced PLA composites as a function of bre content. (For interpretation of the references to color in this gure legend, the reader is referred to the web version of this article.)

The average tensile strength and Youngs modulus of the untreated and treated hemp bre/PLA composites are depicted in Figs. 3 and 4, respectively. As can be seen, the tensile strength and Youngs modulus of the PLA/ALK, PLA/ALKSIL and PLA/SIL composites increased compared with those of the PLA/FB composites. This could be attributed to good adhesion in between treated bres and PLA as shown by decreased bre pull-out (see Fig. 5) and increased PLA crystallinity in the treated bre reinforced PLA composites. Among all the samples, tensile strength and Youngs modulus of the PLA/ALK composites were found to be the highest. The tensile strength and Youngs modulus of the PLA/ALK composites were 75.5 MPa and 8.2 GPa, respectively, about 10.5% and 8.2% higher than those of the PLA/FB composites. These results were better than any reported tensile properties for the short natural bre/PLA composites [8,14,24]. As can be seen in the fracture surface of PLA/ALK composite (see Fig. 6), ALK bres are tightly connected with the PLA matrix. It can also be seen that some bres were broken and/or torn. The brillation of bre was another feature, which was most probably caused by regions of the bre, well bonded to the matrix, being torn from

underlying layers of the bres. All these observations supported strong bonding in the PLA/ALK composites. As can also be noted in Fig. 3, the relationship between tensile strength and bre content was not linear, which indicated that at higher bre content, the benet to composite strength by adding bre was somewhat decreased. It is well known that bre shortening inevitably occurs during extrusion and injection moulding of the composites containing both natural [25,26] and synthetic bres [27], due to the strong shear stresses that act in the viscous molten polymer. So, in the present case, as the bre content increased, the probability of the bre/bre interaction and bre/ equipment wall also increased, resulting in an increase of short bre (i.e. bres below the critical length) populations in the composites. This is evident by the fact that bre pull-out increased with increased bre content as can be seen in Fig. 7. Thus, the non-linear relationship at higher bre content could be explained by the increase of population of the shorter bres. As for untreated bre/PLA composites, the failure strain of the treated bre/PLA composites also decreased with increased bre content as can be seen in Fig. 8. This behaviour could be due to

314

M.A. Sawpan et al. / Composites: Part A 42 (2011) 310319

Poor fibre/matrix adhesion

Good fibre/matrix adhesion

(a) PLA/FB

(b) PLA/ALK

Good fibre/matrix adhesion Good fibre/matrix adhesion

(c) PLA/SIL

(d) PLA/ALKSIL

Fig. 5. SEM micrographs of tensile fracture surface of untreated and treated hemp bre reinforced PLA composites.

20 wt% fibre

Fibrils

Fig. 6. SEM micrographs of tensile fracture surface of PLA/ALK composites.

the lower failure strain of the bres compared to that of PLA. Stress concentrations brought about at the broken bre ends could promote fracture of the matrix, leading to overall failure of the composites at strains below that of the unreinforced PLA itself. It is evident that the variation of failure strain for different composites did not follow a similar order as the bre content increased from 10 to 30 wt.%. It is also found that failure strain of all the samples varied insignicantly (less than 1%) for 1030 wt.% reinforcements. It must be accepted that experimental error as well as inconsistent bre dispersion will have inuenced variability. As the alkali treated short hemp bre/PLA composites had the best tensile strength and Youngs modulus, the investigation was expanded to produce high strength alkali treated hemp bre/PLA

composites using aligned long bres. The composites were fabricated according to the method described in Section 2.2.2. The average tensile strength, Youngs modulus and failure strain of the alkali treated aligned long bre/PLA composites as a function of bre content are shown in Fig. 9. Tensile properties of alkali treated short bre/PLA composites are presented for comparison. As can be seen, the average tensile strength and Youngs modulus of the long bre composites (30 wt.% bre) were higher than those of the short composites, which would be expected due to higher reinforcement efciency for the aligned long bres. In short bre composites, a high amount of composite fracture near the bre-end positions of the short bres might occur which could reduce the effective bre length and hence the tensile strength and Youngs modulus [28]. It is also apparent that the tensile strength and Youngs modulus of the long bre PLA composites increased with increased bre content up to 35 wt.% and further increment of bres (40 wt.%) caused a decrease in the tensile strength and Youngs modulus. At 35 wt.% bre content, the average tensile strength and Youngs modulus of the long bre/PLA composites was 85.4 MPa and 12.6 GPa, respectively. This tensile strength of the long hemp bre/PLA composites was found to be lower (about 14 MPa) than the reported tensile strength of aligned jute bre mat/PLA composites [16]. This could be due to some misalignment of hemp bres in the composites. The hemp bres used in this work were in bales of tangled and twisted strips. Despite the fact that the bres were carded, they appeared to be somewhat crimped due to spring-back during composite processing. However, Youngs modulus of the long hemp bre/PLA composites was found to be better than any reported long bre/PLA composites [16]. In the case of 40 wt.% bre content (see Fig. 10), bres were not thoroughly wetted due to the insufcient amounts of matrix being available to cover the bres. As a result, bres were not well connected with the matrix, and some gaps between the bres

M.A. Sawpan et al. / Composites: Part A 42 (2011) 310319

315

(a) 10 wt% fibre

(b) 20 wt% fibre

(c) 30 wt% fibre

Fig. 7. SEM micrograph of the tensile fracture surface of PLA/FB composites at different bre contents (scale bar = 100 lm).

3.2

Failure strain (%)

2.8

PLA/FB PLA/ALK PLA/SIL PLA/ALKSIL

2.4

2.0

1.6

1.2 0 10 20 30

Fibre content (%)

Fig. 8. Failure strain of untreated and treated hemp bre reinforced PLA composites as a function of bre content. (For interpretation of the references to color in this gure legend, the reader is referred to the web version of this article.) Fig. 10. SEM micrograph of the tensile fracture surface of long aligned bre ALK/ PLA composites (40 wt.% bre).

3.3. Impact strength and matrix were evident. In this situation, load on the composites was not distributed evenly from the bre to bre through the matrix, and catastrophic failure of the composites was observed because of poor wetting of the bres. The average impact strength of short hemp bre (untreated and treated) reinforced PLA composites as a function of bre content is depicted in Fig. 11. As can be seen, impact strength of all compos-

90

15 3.0

Young's modulus (GPa)

Tensile strength (MPa)

80

12

Failure strain (%)

0 10 20 30 40

2.5

70

2.0

1.5

60

6 1.0

50 0 10 20 30 40

3 0.5 0 10 20 30 40

Fibre loading (wt%) Long fibre Short fibre

Fibre loading (wt%) Long fibre Short fibre

Fibre loading (wt%) Long fibre Short fibre

Fig. 9. Tensile properties of alkali treated aligned long bre and random short bre reinforced PLA composites. (For interpretation of the references to color in this gure legend, the reader is referred to the web version of this article.)

316

M.A. Sawpan et al. / Composites: Part A 42 (2011) 310319

8 2.7 Long fibre Short fibre

Impact strength (kJ/m )

2.4

2.1

Impact strength (kJ/m )

1.8

PLA/FB PLA/ALK PLA/SIL PLA/ALKSIL

0 10 20 30

1.5 0 10 20 30 40

Fibre content (wt%)

Fig. 11. Impact strength of untreated and treated hemp bre reinforced PLA composites as a function of bre content. (For interpretation of the references to color in this gure legend, the reader is referred to the web version of this article.)

Fibre loading (wt%)

Fig. 13. Impact strength of long and short bre reinforced PLA composites (PLA/ ALK) as a function of bre content. (For interpretation of the references to color in this gure legend, the reader is referred to the web version of this article.)

ites increased with increased bre content. This was because as the bre content increased, more interfaces exist on the crack path, and more energy was consumed. In fact, the concentration of short bres would have increased with increased bre content, which could lead to increased pull-out and also increased impact strength. Similar to the tensile strength and Youngs modulus, it can be observed that the alkali and silane treatments enhanced the impact strength of composites. This nding is in agreement with other studies [26,29,30]. The PLA/ALK composites with 30 wt.% bres had the highest impact strength (2.64 kJ/m2), which was approximately 12% higher than that of the PLA/FB composites (2.34 kJ/m2). Fibrillation of bres, which is associated with high energy absorption [31], can be observed in the impact tested fracture surfaces of alkali and silane treated bre reinforced PLA composites (see Fig. 12). The increased PLA crystallinity of the alkali and silane treated bre composites compared to untreated bre composites could be another factor leading to increased impact strength. Indeed, impact strength of the composites increased in the following order: PLA/FB < PLA/SIL < PLA/ALKSIL < PLA/ALK which is consistent with the order of PLA crystallinity in composites. Perego et al. [32] and Todo et al. [33] also showed that the impact strength of PLA increased with increased crystallinity of PLA. Fig. 13 presents the average impact strength as a function of bre content, for alkali treated long bre/PLA composites. Impact strength of the alkali treated short bre/PLA composites are presented for comparison. Similar to tensile strength of the long bre composites, it may be observed that the impact strength of

Fig. 14. Photograph of the impact tested long hemp bre reinforced PLA composites (35 wt.% bre). (For interpretation of the references to color in this gure legend, the reader is referred to the web version of this article.)

long bre composites increased as the bre content increased and attained the maximum value (7.4 kJ/m2) at 35 wt.% bre content. Further increment of bres caused a decrease in impact strength of the composites. For an equivalent amount of bres (30 wt.%), the impact strength of the long bre composites was 101% higher than that of the short bre composites.

(a) PLA/ALK

(b) PLA/SIL

(c) PLA/ALKSIL

Fig. 12. SEM micrographs showing brils (indicated by arrow) in the impact fracture surface of PLA composite samples (30 wt.% bre).

M.A. Sawpan et al. / Composites: Part A 42 (2011) 310319

317

Fibrils Fibre pull-out Fibre fracture

(a)

(b)

Fig. 15. SEM micrographs of the impact fracture surface of long bre PLA/ALK composites (35 wt.% bre).

2.2 6 2.0

1.8

K Ic (MPa-m )

1/2

1.6

GIc (kJ/m )

1.4

3 1.2
PLA/FB PLA/ALK PLA/SIL PLA/ALKSIL PLA/FB PLA/ALK PLA/SIL PLA/ALKSIL

1.0 0

10

20

30

10

20

30

Fibre content (wt%)

Fibre content (wt%)

(a)

(b)

Fig. 16. KIc and GIc of untreated and treated hemp bre reinforced PLA composites as a function of bre content. (For interpretation of the references to color in this gure legend, the reader is referred to the web version of this article.)

A photograph of impact tested long bre PLA composites is presented in Fig. 14. As can be seen, samples were not completely separated into two pieces but bres bridged the gap to hold the sample together. This mode of failure was associated with high energy absorption [34]. In addition, examination of the impact fracture surfaces showed bre pull-out due to the fracture of long bre during impact loading (see Fig. 15a). Fibrillation was also evident in the impact fracture surface (see Fig. 15b), which was consistent with the short bre composites. 3.4. Fracture toughness Fig. 16 presents the average KIc and GIc of short hemp bre (untreated and treated) reinforced PLA composites as a function of bre content. As can be seen, KIc and GIc of all the composites decreased with increased bre content. This could again be due to increased stress concentrations (bre ends) and PLA crystallinity of the composites compared to PLA only samples. ALK, SIL and ALKSIL bres were found to decrease the KIc and GIc of the composites compared to the FB bre composites. This could be attributed to the greater PLA crystallinity and improved interfacial adhesion in the PLA/ALK, PLA/SIL and PLA/ALKSIL composites compared with that of the PLA/FB composites [35].

Fig. 17 shows the average KIc and GIc of the alkali treated long and short bre composites as a function of bre content. As can be observed, KIc and GIc of the long bre composites was higher than those of the short bre composites. This was because in the long bre composites the bres were oriented perpendicular to the loading direction thus had greater resistance to crack propagation. Fig. 18 indicated that crack propagation was suppressed somewhat due to the bre bridging. Like short bre composites, KIc and GIc of the long bre composites decreased with increased bre content. This could again be due to the crystalline interface of PLA/hemp bre composites through which cracks can propagate easily. At 30 wt.% reinforcement, KIc and GIc of the long bre composites were 36.4% and 25.1%, respectively, higher than those of the short bre composites. 4. Conclusions Tensile strength, Youngs modulus and impact strength of short hemp bre reinforced PLA composites were found to be increased with increased bre content (1030 wt.%). It was found that PLA could be reinforced with a maximum of 30 wt.% bres using conventional injection moulding, but could not be processed at higher bre contents due to poor melt ow of the compounded materials.

318

M.A. Sawpan et al. / Composites: Part A 42 (2011) 310319

7 2.0

Long fibre Short fibre

Long fibre Short fibre

1.6

K Ic (MPa-m )

1/2

1.2

G Ic (kJ/m )

0.8

1 0.4 0 10 20 30 40 0 10 20 30 40

Fibre loading (wt%)

Fibre loading (wt%)

Fig. 17. KIc and GIc of long and short hemp bre reinforced PLA composites (PLA/ALK) as a function of bre content. (For interpretation of the references to color in this gure legend, the reader is referred to the web version of this article.)

References
[1] Sawpan MA, Pickering KL, Fernyhough A. Hemp bre reinforced poly(lactic acid) composites. Adv Mater Res 2007;2930:33740. [2] Mohanty AK, Misra M, Drzal LT. Natural bers, biopolymers, and biocomposites. Taylor and Francis: CRC press; 2005. [3] Sinha Ray S, Okamoto M. Biodegradable polylactide and its nanocomposites: opening a new dimension for plastics and composites. Macromol Rapid Commun 2003;24:81540. [4] Solarski S, Ferreira M, Devaux E. Characterization of the thermal properties of PLA bers by modulated differential scanning calorimetry. Polymer 2005;46:1118792. [5] Sawpan MA, Pickering KL, Fernyhough A. Characterisation of hemp bre reinforced poly(lactic acid) composites. Int J Mater Product Technol 2009;36:22940. [6] Oksman K, Skrifvars M, Selin JF. Natural bres as reinforcement in polylactic acid (PLA) composites. Compos Sci Technol 2003;63:131724. [7] Bax B, Mussig J. Impact and tensile properties of PLA/Cordenka and PLA/ax composites. Compos Sci Technol 2008;68:16017. [8] Garcia M, Garmendia I, Garcia J. Inuence of natural ber type in ecocomposites. J Appl Polym Sci 2008;107:29943004. [9] Serizawa S, Inoue K, Iji M. Kenaf-ber-reinforced poly(lactic acid) used for electronic products. J Appl Polym Sci 2006;100:61824. [10] Mathew AP, Oksman K, Sain M. Mechanical properties of biodegradable composites from poly lactic acid (PLA) and microcrystalline cellulose (MCC). J Appl Polym Sci 2005;97:201425. [11] Lee S-H, Ohkita T, Kitagawa K. Eco-composite from poly(lactic acid) and bamboo ber. Holzforschung 2004;58:52936. [12] Lee S-H, Wang S. Biodegradable polymers/bamboo ber biocomposite with bio-based coupling agent. Compos Part A: Appl Sci Manuf 2006;37: 8091. [13] Vila C, Campos AR, Cristovao C, Cunha AM, Santos V, Parajo JC. Sustainable biocomposites based on autohydrolysis of lignocellulosic substrates. Compos Sci Technol 2008;68:94452. [14] Pilla S, Gong S, ONeill E, Rowell RM, Krzysik AM. Polylactide-pine wood our composites. Polym Eng Sci 2008;48:57887. [15] Iwatake A, Nogi M, Yano H. Cellulose nanober-reinforced polylactic acid. Compos Sci Technol 2008;68:21036. [16] Plackett D, Andersen TL, Pedersen WB, Nielsen L. Biodegradable composites based on L-polylactide and jute bres. Compos Sci Technol 2003;63: 128796. [17] Ganster J, Fink H-P. Novel cellulose bre reinforced thermoplastic materials. Cellulose 2006;13:27180. [18] Standard Test Method for Tensile Properties of Plastics. ASTM D 638; 2003. [19] Bledzki AK, Reihmane S, Gassan J. Properties and modication methods for vegetable bers for natural ber composites. J Appl Polym Sci 1996;59:132936. [20] Standard Test Methods for Plane-Strain Fracture Toughness and Strain Energy Release Rate of Plastic Materials. ASTM D 5045; 1999. [21] Mathew AP, Oksman K, Sain M. The effect of morphology and chemical characteristics of cellulose reinforcements on the crystallinity of polylactic acid. J Appl Polym Sci 2006;101:30010.

Fibre bridge

Fig. 18. SEM micrograph of the long bre PLA/ALK composites showing the bre bridging in SENB tested sample (30 wt.% bre).

KIc and GIc of the composites decreased with increased bre content which could be due to the increase of stress concentration (number of bre ends) and crystallinity of PLA in composites. Tensile properties and impact strength of the composites were increased further with bre treatments (e.g. alkali and silane) which could be due to improved bre/matrix adhesion and increased PLA crystallinity. Alignment of long bres was found to be an effective technique to improve the mechanical properties of PLA/hemp bre composites compared to those of short hemp bre/PLA composites. The highest mechanical properties were obtained with a 35 wt.% aligned long alkali treated bre composites with tensile strength of 85.4 MPa, Youngs modulus of 12.6 GPa and impact strength of 7.4 kJ/m2.

Acknowledgement The nancial support from Biopolymer Network Ltd., New Zealand for this work is greatly acknowledged.

M.A. Sawpan et al. / Composites: Part A 42 (2011) 310319 [22] Eichhorn SJ, Baillie CA, Zafeiropoulos N, Mwaikambo LY, Ansell MP, Dufresne A, et al. Current international research into cellulosic bres and composites. J Mater Sci 2001;36:210731. [23] Sawpan MA. Mechanical performance of industrial hemp bre reinforced polylactide and unsaturated polyester composites. PhD Thesis, The University of Waikato; 2009. [24] Pengju P, Bo Z, Weihua K, Serizawa S, Iji M, Inoue Y. Crystallization behavior and mechanical properties of bio-based green composites based on poly(Llactide) and kenaf ber. J Appl Polym Sci 2007;105:151120. [25] Bos HL, Mussig J, van den Oever MJA. Mechanical properties of short-ax-bre reinforced compounds. Compos Part A: Appl Sci Manuf 2006;37:1591604. [26] Tokoro R, Vu DM, Okubo K, Tanaka T, Fujii T, Fujiura T. How to improve mechanical properties of polylactic acid with bamboo bers. J Mater Sci 2008;43:77587. [27] Fu SY, Lauke B, Mader E, Yue CY, Hu X. Tensile properties of short-glassberand short-carbonber-reinforced polypropylene composites. Compos Part A: Appl Sci Manuf 2000;31A:111725. [28] Thygesen A, Thomsen AB, Daniel G, Lolholt H. Comparison of composites made from fungal debrated hemp with composites of traditional hemp yarn. Ind Crops Prod 2007;25:14759.

319

[29] Huda MS, Drzal LT, Mohanty AK, Misra M. Effect of chemical modications of the pineapple leaf ber surfaces on the interfacial and mechanical properties of laminated biocomposites. Compos Interfaces 2008;15:16991. [30] Huda MS, Drzal LT, Mohanty AK, Misra M. Effect of ber surface-treatments on the properties of laminated biocomposites from poly(lactic acid) (PLA) and kenaf bers. Compos Sci Technol 2008;68:42432. [31] Hill CAS, Abdul Khalil HPS. Effect of ber treatments on mechanical properties of coir or oil palm ber reinforced polyester composites. J Appl Polym Sci 2000;78:168597. [32] Perego G, Cella GD, Bastioli C. Effect of molecular weight and crystallinity on poly(lactic acid) mechanical properties. J Appl Polym Sci 1996;59:3743. [33] Todo M, Sang Dae P, Arakawa K, Koganemaru M. Effect of crystallinity and loading-rate on mode I fracture behavior of poly(lactic acid). Polymer 2006;47:135763. [34] Williams T, Allen G, Kaufman MS. The impact strength of bre composites. J Mater Sci 1973;8:176587. [35] Sawpan MA, Pickering KL, Fernyhough A. Inuence of loading rate and bre content on plane-strain fracture toughness of random short hemp bre reinforced polylactide (PLA) bio-composites. In: Proceedings of the 10th international conference on wood and biober plastic; 2009. p. 1328.