Composites: Part A 42 (2011) 993–999

Contents lists available at ScienceDirect

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

Effect of manufacturing process on the flexural, fracture toughness,

and thermo-mechanical properties of bio-composites
Abdullah A. Kafi ⇑, Kevin Magniez, Bronwyn L. Fox
Centre for Material and Fibre Innovation, Deakin University, Geelong, VIC 3217, Australia

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

Article history: This study evaluated the performance of jute reinforced polyester bio-composites cured with an out-of-
Received 8 October 2010 autoclave curing process, Quickstep™. The mode I fracture toughness (GIc), flexural, and thermo-
Received in revised form 10 February 2011 mechanical properties of the composites were measured for cure cycle times of 5, 30, 60 and 90 min
Accepted 4 April 2011
at an optimum cure temperature of 95 °C, and were compared to a reference composite laminate cured
Available online 9 April 2011
at room temperature. Generally, the Quickstep process was found to improve the degree of cure by up to
55% and the resistance to crack propagation of the resulting composites was found to be more stable. A
Keywords:
balance in properties was obtained after 30 min of cure where an improvement in thermo-mechanical
A. Bio-composites
B. Fracture toughness
and fracture properties was observed, but on the other hand, was accompanied by a decrease in flexural
D. Thermo-mechanical properties strength and modulus compared to the reference sample. Shorter (i.e. 5 min) and longer cure cycle times
E. Quickstep (above 30 min) used by the Quickstep were found to be detrimental to the properties of the resulting
composites.
Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction and epoxy [8]. Unsaturated polyesters (UP) possess good

mechanical properties, flexible cure cycle and lower in cost com-
Composite materials have rapidly become a material of choice pared to other thermosetting resins. These resins are widely used
as an alternative to other traditional materials such as metals with glass fibres for automotive and building application [9] but
and have found applications in many sectors. Nonetheless, in light also more recently with natural fibres. Natural fibre based thermo-
of the environmental impact that the composite industry has, cou- set composites have been successfully manufactured by several
pled with the issues associated with the disposal and recycling of methods, including wet lay-up , resin transfer moulding (RTM)
waste composite, there has been a growing interest for bio-com- [10], and finally vacuum assisted resin infusion [11]. Nonetheless,
posites as an eco-friendly alternative [1]. As a result, a large the inherent incompatibility between natural fibres and synthetic
amount of research has been dedicated to the use of natural fibre matrix resulting in poor fibre–matrix adhesion is often an issue
as a substitute for glass, carbon and synthetic fibres; driven by po- which needs to be addressed in order to improve the mechanical
tential weight saving, lower raw material price, and ecological performance of the resulting bio-composites. The quality of fi-
advantages of using resources which are renewable [2,3]. Many re- bre–matrix adhesion can be enhanced either through surface treat-
search articles have reported the use of various natural fibres, ment or via improvement of manufacturing processes. For
mostly derived from the bast or stem of the plants, including jute, instance, curing of composites at higher temperatures will further
sisal, flax and hemp. Among these, jute, the lignocellulosic bast fi- lower the viscosity of the resin, which in turn will increase the
bre has been shown to be a promising candidate for composite wetting of the fibres and improve the fibre–matrix adhesion in
applications because of its low specific gravity (1–1.45), high spe- composites [12].
cific modulus (19 GPa), and low cost [4]. One of the major limita- It has also been reported that the fibre–matrix interfacial adhe-
tions which have slowed down the implementation of jute fibres sion has a significant influence on the interlaminar fracture tough-
into bio-composite materials are its poor interfacial adhesion with ness of composites [13]. The resistance to delamination growth has
resin matrices, resulting in poor mechanical properties. been characterized by stress intensity factors (K) or strain energy
Natural fibres have been previously used with various thermo- release rates (G) under different loading conditions (i.e. modes I–
setting resins including phenolic [5], vinylester [6], polyester [7], III). Increasing the level of adhesion at the fibre–matrix interphase
generally yields improvements in the fracture toughness of com-
posites [14]. The work described in this paper utilizes a rapid
⇑ Corresponding author. Tel.: +61 3 5227 038. out-of-autoclave curing process, namely the Quickstep process,
E-mail address: aakaf@deakin.edu.au (A.A. Kafi). which has been shown to improve the fibre–matrix adhesion as

1359-835X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.compositesa.2011.04.002
994 A.A. Kafi et al. / Composites: Part A 42 (2011) 993–999

well as the mode I interlaminar fracture toughness properties of

carbon fibre reinforced composites [15]. This study addresses this
gap by setting out to investigate the performance of natural fibre
based composite using Quickstep based on flexural, fracture tough-
ness and thermo-mechanical properties of the resulting jute/poly-
ester bio-composites. The effect of both cure temperature and time
on the degree of cure of composites is analyzed by differential
scanning calorimetry.

2. Materials and methods

Fig. 2. Schematic representation of Quickstep process (www.quickstep.com.au).
(For interpretation of the references to colour in this figure legend, the reader is
2.1. Materials referred to the web version of this article.)

The unsaturated polyester resin used was a ESCON 62-333 pro-

cured from fibreglass International, Australia, was premixed man- flexible silicon bladders containing a heat transfer fluid (HTF) as
ually with a cobalt napthenate promoter solution (1.5 wt.%) and shown in Fig. 2. During process, the heating and cooling rates are
36 wt.% styrene cross-linking agent. The curing agent was methyl controlled by quickly circulating the HTF contained in tanks at var-
ethyl ketone peroxide (MEKP). Unbleached woven jute fabrics ious temperatures. The Quickstep process has therefore the ability
were sourced from Bangladesh Jute Mills Corporation. to cure composite panels with a reduced risk of exothermic reac-
tions. The pressure created by the HTF within the bladder system
2.2. Experimental procedures during the curing process was 10 kPa at maximum. The standard
diagram for the QS tooling showed the vibration source in Fig. 2,
2.2.1. Composite fabrication however, the work did not utilise that option.
Square samples of woven jute fabrics (250 mm in length) were Two thermocouples were inserted one on each side of the lam-
used to produce the composites. Polyester resin premixed with inate before vacuum bagging to record resin temperature and to
1 wt.% MEKP was poured evenly by hand onto each layer of woven control the cure cycle. Four different cure cycles were used at con-
jute fabric with an approximate resin to jute weight ratio of 3:1. stant dwell temperature (95 °C) and heating rate (3 °C/min) where
For the overall stacking sequence, a total of 12 layers of fabric were dwell time was varied (e.g. 5 min, 30 min, 60 min and 90 min). The
placed with a [0/90] orientation (in both warp and weft directions) resultant composites were approximately 6 mm thick and were
on a stainless steel mould coated with a poly (vinyl alcohol) (PVA) stored in the fumehood overnight to remove unfixed styrene vapor.
release agent. The overall sequence of lay-up at room temperature For clarity of the discussion, the samples cured using the Quickstep
was as follows: layer of release film, sandwich structure made of process at 95 °C for 5, 30, 60 and 90 min will be referred to as 95/5,
resin and fabric, a layer of peel ply, and a breather on the top 95/30, 95/60, and 95/90 respectively.
(Fig. 1). The stack contains 12 layers of fabric impregnated with re-
sin was finally vacuum bagged at 85 kPa before curing.
2.2.2. Density and void content
A reference laminate was produced from an overnight cure with
The densities of the jute fibre and polyester resin given by the
no pressure applied at room temperature (as recommended by the
manufacturer were 1.48 g/cm3 and 1.10 g/cm3, respectively. The
manufacturer) and for clarity of discussion, the specimen tested
density of the composites were determined by the buoyancy meth-
from this laminate will be referred to as RT throughout the text.
od (Archimedes principle) using water as the displacement med-
A number of laminates were also produced using the Quickstep
ium according to the ASTM D 792-00. The void content of the
process at various cure cycle times (for details see next section be-
composites was determined according to the ASTM D 2734-94.
low) and their properties were compared to the ones of the refer-
ence laminate.
2.2.3. Fibre volume fraction
2.2.1.1. Quickstep process. Composite panels cured using Quickstep The fibre volume fraction was derived from the fibre/resin
QS5 Technology were positioned in a clamshell style mould with weight ratio and the densities of both fibre and the resin systems

Vacuum bag
Breather Peel ply
Release film

To Vacuum
Thermocouples
Thermocouples

Tickytap Release film Resin layer

Aluminium mould plate Fabric layer

Fig. 1. Schematic representation of the vacuum bagging process. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version
of this article.)
 A.A. Kafi et al. / Composites: Part A 42 (2011) 993–999 995

[16]. The fibre volume fraction was calculated using the following 5 Hz, with a strain of 0.05% and at a heating rate of 3 °C/min. The
equation: storage modulus (E0 ) and loss tangent (tan d) of the composite
specimen were measured as a function of temperature (30–
1
Fibre volume fraction ¼ 1     Vm 180 °C). The size of the specimen was approximately
V fibre
1þ V resin 55  12  3 mm. The glass transition temperatures were defined
as the maximum of the loss tangent (tan d).
where VV fibre is the fibre to resin volume ratio and Vv is the void
resin
content.
2.2.6. Flexural tests
Flexural strength (FS) and flexural modulus (FM) were deter-
2.2.4. Differential scanning calorimetry (DSC)
mined using the 3-point bending method as per ASTM D790-84a.
The degree of cure of the composites was measured from heat
The test was conducted at a crosshead speed of 2 mm/min1 with
flux DSC analyzes (Perkin–Elmer DSC) in a nitrogen atmosphere.
a Lloyd LR30 K universal tensile tester. Samples were conditioned
Approximately 6–7 mg samples of premixed polyester resin mixed
afterwards according to the ASTM D5229/D5229 M-92(2010) stan-
with 1 wt.% MEKP curing agent was subjected to a scan rate of
dard (65% relative humidity, 20 °C for at least 48 h). After condi-
10 °C/min from 25 °C to 180 °C. The objective was to calculate
tioning, at least six specimens were tested for per laminate.
the total heat of reaction of the material (Ht) which was found to
be approximately 145.8 J/g. A reference sample cured at room tem-
perature overnight was scanned at 10 °C/min from 25 °C to 180 °C 2.2.7. Interlaminar shear strength
to determine its residual heat flow. Approximately 6 to 7 mg sam- Short beam shear test was conducted in accordance with ASTM
ples of pre-impregnated jute materials (resin to jute weight ratio of D2344 at a crosshead speed of 1 mm/min with a Lloyd LR30 K uni-
3:1) were sealed in aluminum pans. In order to replicate the Quick- versal tensile tester. Interlaminar shear strength (ILSS) of the com-
step cure cycles in the DSC, the samples were heated up to 95 °C at posites was determined according to the span to thickness ratio of
3 °C/min and were left at that temperature for various duration 4 and length to thickness ratio of 6. The formula ILSS = 0.75Pb/bd
times of 5 min, 30 min, 60 min and 90 min). The samples were (MPa) was used, where Pb represents the breaking load, b the width
then subjected to a second scan rate of 10 °C/min from 25 °C to of specimen and d the thickness of specimen. All specimens were
180 °C to determine the residual heat flow (Hr). The residual heat simply supported in a fixture and loaded at mid span. A minimum
flow was used to calculate the degree of cure of the resin, a, accord- of 8 specimens were prepared and were conditioned according to
ing to the following equation: the ASTM D5229/D5229 M-92(2010) standards prior to testing.

Ht  Hr
a¼ 2.2.8. Mode I double cantilever beam (DCB) tests
Ht
DCB tests which characterizes the resistance to crack propaga-
tion in an opening mode were performed in accordance with the
2.2.5. Dynamic mechanical analysis (DMA) protocol of the European Structural Integrity Society using a Lloyd
Dynamic mechanical analysis (DMA Q800) was performed in a LR30 K universal testing machine fitted with a 1 kN load cell [17]. A
dual-cantilever bending flexural loading mode at a frequency of thick starter film of 65 mm in length was inserted during the wet
lay-up process to facilitate crack initiation during testing. DCB
specimens were cut to an approximate size of 20 mm in width
and 172 mm in length. The cut edges of each specimen were pol-
ished before one side of the specimen was coated with white paint
to aid observation of the propagating crack. Two rectangular
shaped aluminum tabs were glued to opposite sides of the speci-
men at the end containing the crack. All specimens were condi-
tioned to 20 ± 2 °C and 65 ± 2% relative humidity for 24 h prior to
the testing. The mode I critical energy release rate GIc was calcu-
lated by the corrected beam theory (CBT) [17]. GIc-initiation values
were calculated when the pre-crack was visually observed to start.
Values of GIc-propagation were calculated from the maximum or
plateau of the resistance curve, commonly known as the R-curve.
Five specimens were tested from each panel.

2 100
Fig. 3. Effect of curing temperature on the degree of cure of composites (5 min
Flexural Modulus (GPa)
curing time). 1.8 90
Flexural Strength (MPa)
Flexural Strength, MPa
Flexural Modulus, GPa

1.6 80
1.4 70
Table 1 1.2 60
Properties of the laminates after different cure cycles. 1 50
Curing Degree of cure Tgb Fibre volume Void content 0.8 40
process (%)a fraction (%) (%) 0.6 30
95/5 87 87 25 (±0.03) 6 (±0.003) 0.4 20
95/30 99 90 29 (±0.009) 5 (±0.003)
0.2 10
95/60 98 90 27 (±0.002) 5 (±0.001)
95/90 97 83 28 (±0.0013) 5 (±0.001) 0 0
RT 64 85 22 (±0.06) 5 (±0.004) 95/5 95/30 95/60 95/90 RT

a
Determined from the residual heat of reaction (DSC). Fig. 4. Effect of the curing process on the flexural strength (FS) and modulus (FM) of
b
Determined by the peak of tan d (DMTA). composites.
996 A.A. Kafi et al. / Composites: Part A 42 (2011) 993–999

2.2.9. Scanning electron microscope (SEM) analysis The RT sample, which showed both the lowest degree of cure
SEM observations were carried out on a SEM4 Jeol Neoscope at a and the lowest fibre volume fraction but the highest interfacial
magnification of up to 500 times at a working distance of 10 mm and strength, achieved the best results in terms of flexural strength
EHT of 10 kV on mode I fractured surface between 0 and 5 mm crack and modulus. The 95/5 sample which displayed comparable ILSS
length of the DCB specimen. All samples were mounted and gold and fibre volume fraction values than the other Quickstep samples,
sputtered under vacuum prior to the observations. under-performed most likely as a results of its lower degree of
cure. Both the 95/30 and 95/60 samples for which the interfacial
strength, degree of cure and fibre volume fraction (FVF) parameters
3. Results and discussion
were found to be reasonably similar performed comparably. On the
other hand, the flexural strength decreased significantly (but
3.1. Degree of cure
approximately 50%) in the particular case of the 95/90 sample,
while its modulus remained approximately the same (as a result
The polyester resin was cured at various temperatures and its
of uniform fibre volume fraction between the Quickstep samples).
degree of cure was monitored after 5 min using DSC. The results
Degradation of fibre–matrix adhesion (confirmed by ILSS results)
which are displayed in Fig. 3 indicated that the degree of cross-
most likely resulted in the observed loss of flexural strength.
linking increased rapidly with increasing temperature, and was
found to plateau above 95 °C to approximately reach 87%. By rais-
ing the curing temperature up to 130 °C, the degree of cure only in- 3.3. Dynamic mechanical properties
creased by 1%. In light of these results and in order to minimize the
curing temperature to be used with the Quickstep process, a curing The storage moduli (E0 ) as a function of the temperature for the
temperature of 95 °C was chosen as the optimum curing tempera- various composite samples are shown in Fig. 6. It has been ob-
ture. The effect of curing time on the extent of cross-linking at this served that storage modulus increased significantly after 30 min
temperature was investigated and the results are displayed in Ta- of cure, but decreased gradually thereafter. It is suggested here that
ble 1. It was found that the degree of cure further increased to the observed trends which are slightly different to the static flex-
reach a maximum value of 99% after 30 min of cure which indi- ural results are also strongly influenced by both the degree of cure
cated that the maximum degree of cure achievable at 95 °C should and interfacial strength of the composites. In light of the results,
in principle be reached after 30 min of cure. The glass transition the degree of cure however seemed to have had an overbearing
temperatures results (not shown here for brevity) on the samples
determined by DSC shortly after the manufacturing process sug-
gested that the Quickstep process accelerated the cross-linking 3500 95/5 95/30 95/60 95/90 RT
process. The glass transition of the RT sample shortly after manu-
facturing was found to the significantly lower (by approximately 3000
Storage Modulus, MPa

20 °C) than that of the other Quickstep samples.

2500

3.2. Flexural and inter laminar shear strength (ILSS) properties 2000

The flexural and ILSS results are displayed in Figs. 4 and 5, 1500
respectively. It is important to highlight that direct correlation be-
1000
tween the interfacial strength, degree of cure, and fibre volume
fraction (FVF) parameters with the flexural properties of a compos- 500
ite is rather complex.
It has been previously reported by Thomason et al. [18] that the 0
interfacial adhesion strength in a composite has a dominating ef- 30 50 70 90 110 130 150 170
fect on the flexural strength of composites but has minor effects Temperature,ºC
on the flexural modulus. On the other hand, the flexural modulus
Fig. 6. Effect of curing process on the storage modulus of composites.
is directly influenced by the fibre volume fraction of a composite
[19]. Our results correlate well to some extent with the above
literature.
35
95/5 95/30 95/60 95/90 RT

14 30
Interlaminar Shear Strength, GPa

12 25

10 20
Load, N

8 15

6
10
4
5
2
0
0 20 40 60 80
0
RT 95/5 95/30 95/60 95/90 Displacement, mm

Fig. 5. Effect of the curing process on the Interlaminar Shear Strength (ILSS) Fig. 7. Effect of curing processes on the load–displacement (L–D) curve load
properties of composites. capacity.
 A.A. Kafi et al. / Composites: Part A 42 (2011) 993–999 997

3000 95/5 95/30 95/60 95/90 RT had ample time to further crosslink until the testing was carried
Fracture Toughness,GIc (kJ/m2)

out. The Tg was found to increase by 2 °C after only 5 min of cure

2500 and increased further after 30 and 60 min of cure. The shift of Tg
to higher temperatures can be associated with the decrease in
2000 mobility of the matrix chains, better cross-linking in resin matrix
or improved fibre–matrix interfacial adhesion as mentioned before
1500 [20]. The exposure of resin to a temperature of 95 °C for extended
period of times possibly induced some thermal degradation notice-
1000 able from the significant shift to lower Tg’s after 90 min of cure.

500 3.4. Mode I fracture toughness

0 Both the fibre volume content and void content were measured
0 5 10 15 20 25 30 35
on the composites samples (Table 1). The fibre volume fraction was
Crack Length, mm
found to be similar for composites processed with Quickstep but
Fig. 8. Effect of curing processes on the crack-resistance curve of composites. slightly higher than the RT reference sample. The effect of void
content has been reported to have a significant effect on the mode
I fracture toughness of composites [21]. In this work, however, a
effect over the interfacial strength. Both 95/30 and 95/60 displayed similar void content was measured on all the samples, and there-
significantly better thermo-mechanical response (E0 ) than the RT fore this parameter could not be accounted for any variation in
sample but evidently the degradation of the fibre matrix interface mode I results. The mode I interlaminar fracture toughness (GIc)
negatively affected the storage modulus (E0 ) for curing times properties are shown in Figs. 7 and 8. Fig. 7 displays the load–dis-
exceeding 60 min. The corresponding glass transition tempera- placement (L–D) curve and Fig. 8 displays the crack-resistance
tures (Tg) taken from the peak value of loss tangent (tan d) curve curve (R-curve), respectively.
are shown in Table 1. It is important to highlight at this point of The sample cured at room temperature withstood load of up to
the text that the testing of the samples was carried out testing a 35 N and showed displacement of about 31 mm. The highest load
few weeks after manufacturing and therefore the Tg measured by capacity for this sample was attributed to excessive plastic defor-
DMTA does not correlate any longer to the degree of cure (mea- mation noticeable from high extension value. When comparing
sured shortly after manufacturing). It is believed that the samples the samples produced using Quickstep, it can be seen from Fig. 7

Fig. 9. Surface morphology of the RT composites after mode I testing. Poor interfacial adhesion is visible from the surface cracking as well as extensive numbers of fibre pulled
out.

Fig. 10. Surface morphology of the 95/5 composites.

998 A.A. Kafi et al. / Composites: Part A 42 (2011) 993–999

Fig. 11. Surface morphology of the 95/30 composites. The images show improvement in fibre alignment.

that the load capacity was found to be significantly higher for the present in incompletely cured resin will causing internal plastici-
95/30 and 95/60 samples (up to 25 N) compared to the 95/90 and zation, which could probably explain why the breaking load was
95/5 counterparts. It can nonetheless be highlighted that instabil- much higher for RT sample (Fig. 7). The samples cured using Quick-
ity in the load–displacement (L–D) curve for the 95/60 sample is step process displayed uniform crack propagation along the inter-
visible. The reason could be inconsistency within the crack initi- laminar interface indicating good distribution of resin onto the
ated surface which is normal for natural fibre composites. fibre surface. It was also found that the 95/30 sample had the larg-
When analyzing the R-curve (Fig. 8) it can be noted that com- est GIc-initiation and GIc-propagation values compared to other
posites cured at room temperature (RT) failed prematurely after samples and this improvement is in line with the previous results.
only 10 mm crack growth associated with an abnormally high The SEM fractography images of the mode I fractured surfaces
GIc-propagation and lowest GIc-initiation value. The high fracture are displayed in Figs. 9–13. Most samples displayed some level of
toughness value for this sample could be due to the higher plastic distortion, fibres pull out and local cracking as a result of the load-
deformation of this sample which was only partially cured (Ta- ing. It is interesting to point out in some of the Quickstep samples,
ble 1). It is well documented that low molecular weight species the wetting of the fibres seemed to have been improved as a result

Fig. 12. Surface morphology of the 95/60 composites.

Fig. 13. Surface morphology of the 95/90 composites.

 A.A. Kafi et al. / Composites: Part A 42 (2011) 993–999 999

of decreased viscosity [12]. Regardless of that visual observation, [6] Ray D, Sarkar BK, Bose NR. Impact fatigue behaviour of vinylester resin matrix
composites reinforced with alkali treated jute fibres. Compos Part A Appl S
however, it has been demonstrated that there was no improve-
2002;33(2):233–41.
ment of the ILSS in these samples. [7] Shibata M, Takachiyo KI, Ozawa K, Yosomiya R, Takeishi H. Biodegradable
polyester composites reinforced with short abaca fibre. J Appl Polym Sci
2002;85(1):129–38.
4. Conclusions [8] Gassan J, Gutowski VS. Effects of corona discharge and UV treatment on the
properties of jute-fibre epoxy composites. Compos Sci Technol 2000;60(15):
2857–63.
The work reports the use of Quickstep, an out-of-autoclave pro- [9] Martin JL. Kinetic analysis of an asymmetrical DSC peak in the curing of an
cess, which was used for the first time to effectively cure a jute/ unsaturated polyester resin catalysed with MEKP and cobalt octoate. Polymer
polyester bio-composite, in a timely manner. This process was 1999;40:3451–62.
[10] Sreekumar PA, Joseph K, Unnikrishnan G, Thomas S. A comparative study on
found to improve the degree of cure and the resistance to crack mechanical properties of sisal-leaf fibre-reinforced polyester composites
propagation of the resulting composites which was found to be prepared by resin transfer and compression moulding techniques. Compos
generally more stable. In addition, based on mode I interlaminar Sci Technol 2007;67:453–61.
[11] Williams C, Summerscales J, Grove S. Resin infusion under flexible tooling
fracture toughness, degree of cure, and storage modulus results, (RIFT): a review. Compos Part A Appl S 1996;27A:517–24.
the performance of a jute/polyester bio-composite could be im- [12] Davies LW, Day RJ, Bond D, Nesbitt A, Ellis J, Gardon E. Effect of cure cycle heat
proved through some basic optimization of the cure cycle temper- transfer rates on the physical and mechanical properties of an epoxy matrix
composite. Compos Sci Technol 2007;67:1892–9.
ature and time. The work also concluded that the degree of cure [13] Kim JK, Mai YW. High strength, high fracture toughness fibre composites with
did not always play a major role in improving the flexural perfor- interface control – a review. Compos Sci Technol 1991;41(4):333–78.
mance of composites but rather the fibre volume fraction and the [14] Madhukar MS, Drzal LT. Fiber–matrix adhesion and its effect on composite
mechanical properties: IV. mode I and mode II fracture toughness of graphite/
fibre/matrix interfacial strength had a greater influence.
epoxy composites. J Compos Mater 1992;26(7):936–68.
[15] Zhang J, Fox B. Manufacturing influence on the Delamination fracture behavior
of the T800H/3900-2 carbon fiber reinforced polymer composites. Manuf
Acknowledgement
Process 2007;22(6):768–72.
[16] Heslehurst RB. Composite structures engineering design vs. fabrication
The authors would like to acknowledge Dr. Mohammad Saiful requirements. In: Developments in composites: advanced, infrastructural.
natural, and nano-composites. UNSW, Sydney, Australia; 2006.
Islam from CSIRO Materials Science and Engineering, Geelong, Aus-
[17] Protocols for the interlaminar fracture testing of composites. European
tralia for his assistance during the progression of this project. Structural Integrity Society (ESIS) Delft; The Netherlands: 1993.
[18] Thomason JL, Schoolenberg GE. An investigation of glass fibre/polypropylene
interface strength and its effect on composite properties. Composites
References 1994;25(3):197–203.
[19] Bisanda ETN, Ansell MP. The effect of silane treatment on the mechanical and
[1] Netravali AN, Chabba S. Composites get greener. Mater Today 2003;6(4):22–9. physical properties of sisal–epoxy composites. Compos Sci Technol
[2] Bledzki AK, Faruk O, Sperber VE. Cars from bio-fibres. Macromol Mater Eng 1991;41(2):165–78.
2006;291:449–57. [20] Aziz SH, Ansell MP. The effect of alkalization and fibre alignment on the
[3] John MJ, Anandjiwala RD, Pothan LA, Thomas S. Cellulosic fibre-reinforced mechanical and thermal properties of kenaf and hemp bast fibre
green composites. Compos Inter 2007;14(7–9):733–51. composites: part 1-polyester resin matrix. Compos Sci Technol 2004;64(9):
[4] Mohanty AK, Misra M, Hinrichsen G. Biofibres, biodegeadable polymers and 1219–30.
biocomposites: an overview. Macromol Mater Eng 2000;276/277(1):1–24. [21] Ricotta M, Quaresimin M, Talreja R. Mode I strain energy release rate in
[5] Singh B, Gupta M, Verma A. The durability of jute fibre reinforced phenolic composite laminates in the presence of voids. Compos Sci Technol 2008;68:
composites. Compos Sci Technol 2000;74(6):526–9. 2616–23.