Palf PDF

Mechanical Properties of Pineapple Leaf Fiber-Reinforced
Polyester Composites
L. UMA DEVI, 1, * S. S. BHAGAWAN, 2 SABU THOMAS 1
1
School of Chemical Sciences, Mahatma Gandhi University, Priyadarshini Hills P.O., Kottayam-686 560, Kerala, India
Propellant Engineering Division, Vikram Sarabhai Space Centre, Thiruvananthapuram-695 002, Kerala, India
Received 13 March 1996; accepted 3 November 1996
ABSTRACT: Pineapple leaf fiber (PALF) which is rich in cellulose, relatively inexpen-
sive, and abundantly available has the potential for polymer reinforcement. The present
study investigated the tensile, flexural, and impact behavior of PALF-reinforced polyester composites as a function of fiber loading, fiber length, and fiber surface modification.
The tensile strength and Youngs modulus of the composites were found to increase
with fiber content in accordance with the rule of mixtures. The elongation at break of
the composites exhibits an increase by the introduction of fiber. The mechanical properties are optimum at a fiber length of 30 mm. The flexural stiffness and flexural strength
of the composites with a 30% fiber weight fraction are 2.76 GPa and 80.2 MPa, respectively. The specific flexural stiffness of the composite is about 2.3 times greater than
that of neat polyester resin. The work of fracture (impact strength) of the composite
with 30% fiber content was found to be 24 kJ m02 . Significant improvement in the
tensile strength was observed for composites with silane A172-treated fibers. Scanning
electron microscopic studies were carried out to understand the fibermatrix adhesion,
fiber breakage, and failure topography. The PALF polyester composites possess superior
mechanical properties compared to other cellulose-based natural fiber composites. q
1997 John Wiley & Sons, Inc. J Appl Polym Sci 64: 17391748, 1997
INTRODUCTION
As fiber-reinforced structural plastics are taking
the central stage in almost every sphere of material science, lignocellulosic natural fibers like the
pineapple leaf fiber come as a viable and abundant substitute for the expensive and nonrenewable synthetic fiber. These fibers with high specific strength improve the mechanical properties
of the polymer matrix.
In tropical countries, fibrous plants are available in abundance and at least some of them are
agricultural crops. Pineapple is among them.
Pineapple leaf fibers at present are a waste prodCorrespondence to: S. Thomas.
* Present address: Lecturer, St. Thomas College, Kozhencherry-689 641, Kerala, India.
q 1997 John Wiley & Sons, Inc. CCC 0021-8995/97 / 091739-10
uct of pineapple cultivation. Hence, without any

additional cost input, pineapple fiber can be obtained for industrial purposes.
Among various natural fibers, pineapple leaf
fibers (PALFs) exhibit excellent mechanical properties. These fibers are multicellular and lignocellulosic. They are extracted from the leaves of the
plant Ananus cosomus belonging to the Bromeliaceae family by retting. The main chemical constituents of pineapple fiber are cellulose (7082%),
lignin (512%), and ash (1.1%).1 The superior
mechanical properties of pineapple fiber are associated with its high cellulose content and comparatively low microfibrillar angle. Among matrix
resins, unsaturated polyesters have been commonly used for making thermoset composites, especially with glass fibers.2
Natural fibers like jute, coir, sisal, sunhemp,
straw, etc., 3 7 are incorporated into the polyester
1739
/ 8e65$$4072
03-28-97 17:35:12
polaa
W: Poly Applied
4072
1740
DEVI, BHAGAWAN, AND THOMAS
matrix to make composites. Thomas and co-workers 8 10 reported on the use of sisal fiber as a potential reinforcing agent in polyethylene, polystyrene, natural rubber, and various thermosets
such as epoxy, phenolformaldehyde, and polyester. Although composites based on polyester matrix resin are well known, there have been only
limited studies on the properties of PALF-reinforced polyester composite. Mukherjee and Satyanarayana 11 reported on the mechanical properties
of PALF. Pavithran et al.12 studied unidirectionally aligned PALFpolyester composites and the
toughness of these composites were found to increase with the microfibrillar angle of the fibers.
Recently, George et al.13,14 reported on the processing characteristics, viscoelastic properties,
and mechanical behavior of PALFLDPE composites. The present investigation was aimed at
analyzing the mechanical properties of a PALFreinforced polyester composite. The influence of
fiber length, fiber loading, and coupling agents on
the tensile, flexural, and impact properties of the
composites was analyzed. The properties of the
composites were compared with other natural fiber-reinforced polyester composites.
EXPERIMENTAL
Pineapple leaf fiber (PALF) was obtained from
South India Textile Research Association, Coimbatore. The fiber was chopped to the desired
length, washed with water, dried in air, and used
for composite fabrication. Room-temperature cure
general-purpose polyester resin (HSR 8131, supplied by Bakelite Hylam Ltd., Hyderabad, India)
was used for the study. Vinyltri(2-ethoxy methoxy)silane (Silane A172) and g-aminopropyltrimethoxysilane (Silane A1100) were supplied
by Union Carbide Co., Montreal, Canada. The
physical and mechanical properties of the pineapple fiber and polyester are given in Table I.
Table I Physical and Mechanical Properties of
Pineapple Fiber and Polyester Resin
Property
PALF
Density (g/cm3)
Tensile strength (MPa)
Elongation at break (%)
Youngs modulus (MPa)
Specific strength (MPa)
Specific modulus (MPa)
1.526
170
3
6210
110
4070
/ 8e65$$4072
Polyester
1.159
22.9
1.6
580
19.7
502
03-28-97 17:35:12
For NaOH treatment, the chopped fiber was

dipped in 2% NaOH solution for 1 h, washed with
water, and dried in air. For acetylation, 10 g of
pineapple fiber was soaked in glacial acetic acid
for 1 h at room temperature, decanted, and then
soaked in acetic anhydride (50 mL) containing 2
drops of concentrated sulfuric acid for 5 min. The
fiber was filtered and washed with water until it
was free from acid and then dried in air for 24 h.
For silane treatment, the fibers were treated
with 0.3% solution of silane in a waterethanol
mixture (40 : 60) for 1.5 h, dried in air for 30 min,
followed by drying at 507C for 1 h.
Randomly oriented pineapple fiber polyester
composites with varying fiber length and fiber volume fraction were prepared by a hand layup
method. Composite sheets of size 150 1 150 1 3
mm3 were prepared using a closed mold. First,
the mold was polished and then a mold-releasing
agent was applied on the surface. General-purpose polyester resin was mixed with 1 wt % cobalt
naphthenate (accelerator) and 1 wt % methyl
ethyl ketone peroxide (catalyst). The resin mixture was degassed in a vacuum desiccator and
then poured onto the mat placed in the mold.
When the mat was completely wet by the resin,
the mold was closed, pressed, and cured at room
temperature for 24 h. Sheets were prepared with
fibers of varying length (5, 10, 20, 30, and 40 mm)
and of varying fiber loading. The specimens of required dimensions were cut from the sheets,
smoothed by sanding, and used for testing.
Tensile testing of pineapple fibers was carried
out in an Instron universal testing machine Model
1121 at a crosshead speed of 1 mm per min. Specimens were prepared by mounting single fibers on
a stiff cardboard piece with a 50 mm window. The
ends of the fibers were fixed on the cardboard. The
diameters of the fiber specimens were measured
using an optical stereomicroscope. Since the diameter of the fiber is found to be different in different places, an average of about six readings
were taken for diameter determination.
Tensile testing of the composite specimens was
carried out using an Instron universal testing machine Model 1121 at a crosshead speed of 5 mm
min 01 and a gauge length of 60 mm. Rectangular
specimens of size 150 1 10 1 3 mm3 were used
for testing. Flexural tests were performed according to ASTM D 790. Rectangular strips of size
75 mm long, 15 mm wide, and 3 mm thick were
used for these tests. A crosshead speed of 5 mm
min 01 was used. The flexural strength and flex-
polaa
W: Poly Applied
4072
PINEAPPLE LEAF FIBER/POLYESTER COMPOSITES
1741
ural modulus were calculated using the following

relationships:
Flexural strength
3PL
2bd 2
(1)
Flexural modulus
L 3m
4bd 3
(2)
where L is the support span; b, the width of the

specimen; d, the thickness; P, the maximum load;
and m, the slope of the initial straight line portion
of the load-deflection curve.
Charpy-type impact tests on unnotched specimens were performed using a pendulum impact
testing machine PSW 0, 4. The test specimen was
50 mm long, and the cross-sectional area, 24 mm2 .
For evaluation of tensile, flexural, and impact
properties, five specimens were tested and average values are reported.
The surfaces of the fractured specimens under
tensile, flexural, and impact tests were examined
using a scanning electron microscope (SEM)
JEOL Model JSM-35C and a Cambridge 250 MK3
stereoscan.
RESULTS AND DISCUSSION

Mechanical Properties in Tension
The uniaxial stressstrain diagrams of PALF
polyester composites with varying fiber length
Figure 2 Stressstrain curve of PALFpolyester

composites at various fiber loadings (fiber length 30
mm).
and fiber loading are given in Figures 1 and 2.

The deformation behavior of the composites under
an applied load can be understood from the
stressstrain curve (Fig. 1). The stress increases
linearly with strain at low elongation. However,
at higher elongation, nonlinear behavior is observed. This is true for all lengths of fibers used.
It is evident from the figure that for a given strain
level stress increases with fiber length up to 30
mm and then decreases, indicating an optimum
fiber length.
From the stressstrain curve of the composites
with varying fiber loading (Fig. 2), it is found
that the stressstrain curve of pure polyester is
similar to that of brittle materials. The behavior is
perfectly elastic, i.e., the stress increases linearly
with strain. However, addition of fibers makes the
matrix more ductile. This is evident from the high
elongation-at-break values of the composites.
Effect of Fiber Length
Figure 1 Dependence of stressstrain behavior of

PALFpolyester composites on fiber length (fiber loading 30 wt %).
/ 8e65$$4072
03-28-97 17:35:12
The effect of fiber length on the tensile properties

of PALFpolyester composites (containing 30%
fiber by weight) can be readily assessed from the
data shown in Table II. In general, the polyester
composites showed an increasing trend in the mechanical properties up to a fiber length of 30 mm.
At a fiber length of 30 mm, the Youngs modulus
polaa
W: Poly Applied
4072
1742
Table II Effect of Fiber Length on Tensile

Properties of PALF-reinforced Polyester
Composites (Fiber Content 30 Wt %)
Fiber
Length
(mm)
Youngs
Modulus
(MPa)
Tensile
Strength
(MPa)
Elongation
at Break
(%)
5
10
20
30
40
815
1870
1990
2290
1970
15.6
35.0
39.2
52.9
38.4
3.0
3.0
3.0
3.6
3.0
of the composites is 180% higher than that of a 5

mm length due to effective stress transfer between the fiber and the matrix. When the fiber
length is increased to 40 mm, a decrease in moduli
is observed.
The change in tensile strength of the composites with fiber length followed the same pattern
as that of the modulus. The tensile strength value
of the composites with 30 mm-long fibers is about
240% higher than that with a fiber length of 5
mm. When the length of the fiber is increased
from 30 to 40 mm, the tensile strength is decreased by 38%. The decrease in strength above
a fiber length of 30 mm is due to fiber entanglements that occur above an optimum size of the
fibers. However, no significant changes in the
elongation at break of the composites are noticed
with respect to variation in fiber length.
Observation of the composite specimens have
shown that long fibers tend to bend or curl during
molding and cause a reduction in the effective
length of the fiber below the optimum length (30
mm), which results in a decrease of properties.
Thus, there exists a critical fiber length that is
required for the fiber to develop its fully stressed
condition in the matrix in these composites. From
Table III Variation of Tensile Properties of
PALFPolyester Composites as a Function of
Fiber Loading (Fiber Length 30 mm)
Fiber
Content
(Wt %)
Youngs
Modulus
(MPa)
Tensile
Strength
(MPa)
Elongation
at Break
(%)
0
10
20
30
40
580
1770
1830
2290
2520
20.6
17.1
40.0
52.9
63.3
1.6
1.3
3.0
3.6
5.0
/ 8e65$$4072
03-28-97 17:35:12
the overall mechanical properties, a fiber length

of 30 mm was found to be the optimum length for
effective reinforcement in a polyester. In the case
of fibers shorter than the critical length (30
mm), the stressed fiber will debond from the matrix and the composite will fail at low strength.
Effect of Fiber Loading
The effect of fiber loading on the randomly oriented composites is shown in Table III. It is observed that the Youngs modulus of the composites showed an increase with increase in fiber content. Addition of 40% fiber increased the modulus
by 340%. The high stiffness of these materials
makes them suitable for structural applications.
The tensile strength of the composite is slightly
decreased with the addition of 10% fiber. This is
because at low fiber loadings the fibers act as
flaws. However, further increase in fiber content
up to 40% increases the tensile strength. Addition
of 40% fiber increases the tensile strength of the
composite by about 210% than the neat polyester.
The percentage elongation at break is quite low
in the case of pure polyester resin. The elongation
at break decreases for 10% fiber composites and
then increases with increasing concentration of
fiber loading as observed from Table III. The brittle nature of the polyester resin decreases with
the addition of pineapple fiber and, therefore, the
elongation at break increases. It is interesting to
note that the failure elongation of the composite
is much higher than that of the individual components (polyester and PALF) at higher fiber loadings, indicating a synergistic effect.
The tensile fractographs of PALFpolyester
composites containing 10 and 40 wt % of PALF
examined with SEM are presented in Figures 3
and 4. Figure 3 shows fiber pullout, debonding,
and fibrillation. At high fiber loading, fiber-to-fiber contact is greater, which is evident from the
SEM photograph (Fig. 4). The presence of polyester particles on the fiber surface indicates better
interaction between the fiber and the matrix.
Mechanical Properties in Flexure
The stressstrain behavior of PALFpolyester
composites under flexure is shown in Figure 5.
In the case of pure polyester, the flexural stress
increases linearly with strain. But the flexural
behavior of the composites is nonlinear. These
curves show an initial linear portion followed by
a nonlinear portion from the middle of the curve
polaa
W: Poly Applied
4072
1743
Figure 3 SEM of tensile failure surface of PALF

polyester composites (fiber content 10 wt %).
up to the fracture point. The addition of fibers

makes the composite more ductile.
The dependence of flexural properties of the
composites on fiber length is shown in Figure 6.
The flexural strength of the composites containing
30 mm-long fibers was 23% higher than that of
composites containing fibers of 5 mm length. The
flexural stiffness of the composites containing 30
mm-long fibers was also found to be higher than
that of 5 mm-long fibers by 57%. However, the
flexural properties showed a decrease for composites having a fiber length of 40 mm. The best flexural properties of the composite are obtained at a
fiber length of 30 mm.
Figure 7 shows the variation of flexural
strength and flexural modulus with fiber loading.
From the figure, it is seen that the flexural
Figure 4 SEM of tensile fracture surface of PALF

polyester composites (fiber content 40 wt %).
/ 8e65$$4072
03-28-97 17:35:12
Figure 5 Variation of flexural stress with deflection

of PALFpolyester composites.
strength values of the PALFpolyester composites are found to be less than the neat resin at
low weight fractions of the fiber. But increase in
fiber content from 10 to 30% increases the flexural
strength by 120% for composites containing 30
Figure 6 Variation of flexural strength and flexural

modulus with fiber length (fiber loading 30 wt %).
polaa
W: Poly Applied
4072
1744
Figure 7 Variation of flexural strength and flexural

modulus with fiber loading of PALFpolyester composites.
mm-long fibers. Further increase to 40 wt % results in a slight lowering in flexural strength by

about 12%. The flexural strength of the composites is significantly higher than the corresponding
tensile strength. The flexural modulus increases
with fiber content up to 30 wt % and the value for
30 wt % is 105% greater than that of the neat
polyester resin. However, further increase of fiber
loading to 40% decreased the value by 13%. The
flexural properties of the 10 mm fiber composite
vary in the same manner. The decrease of the
flexural modulus at 40 wt % fiber loading is
clearer in the system with short fiber lengths (i.e.,
10 mm). The best flexural properties are observed
in composites with 30% fiber content. The decrease in the flexural properties at higher wt %
(40%) of fiber loading is due to the increased fiberto-fiber interactions and dispersion problems.
The specific flexural stiffness is plotted against
fiber wt % in Figure 8. The specific flexural stiffness increases with the addition of fiber and is
maximum for 30 wt % composites. Its value decreases for 40 wt % composites. The 30 wt % composite has a specific flexural stiffness of 0.25 (m
1 10 6 ), which is 2.3 times greater than that of
polyester.
Scanning electron micrographs of the fractured
surfaces of the composites after the flexural test
are given in Figures 9 and 10. Large fiber break-
/ 8e65$$4072
03-28-97 17:35:12
Figure 8 Variation of specific flexural stiffness of

PALFpolyester composites with fiber content (fiber
length 30 mm).
age and delamination are seen in Figure 9. Figure

10(a) and (b) shows a high extent of fiber pullout
and fibrillation.
Impact Behavior of PALFPolyester Composites
The work of fracture (impact strength) of PALF
polyester composites at 30% loading as a function
of fiber length is shown in Table IV. It is seen
that a comparatively higher impact strength is
Figure 9 SEM of fracture surface of PALFpolyester

composite under flexural failure (fiber content 30 wt %).
polaa
W: Poly Applied
4072
1745
(2) are at odds with those required for good stiffness and strength of the composite.
In most fiber-filled composites, a significant
part of the energy absorption during impact takes
place through the fiber pullout process. The energy involved and, hence, toughness are greatest
when the length of the fibers is equal to the critical
length (lc ) in accordance with (1). Again, it is
apparent that maximum strength and maximum
toughness cannot be achieved simultaneously and
that composites must be designed for an optimum
combination of the desired mechanical properties.
The variation of the work of fracture with fiber
length was calculated by Copper 16 using a model
based on (1) given above. Fibers shorter than lc
(30 mm) will be pulled out from the matrix rather
than broken when a crack passes through the
composite. The fracture energy will then largely
be a combination of the work needed to debond
the fibers out of the matrix and the work done
against friction in pulling the fibers out of the
matrix. The fracture energy (U) arising from fiber
pullout is given by the following expressions due
to Cottrell 17 :
For l lc ,
U1
Figure 10 (a) SEM of fracture surface of PALFpolyester composites under flexural failure (fiber content
10 wt %). (b) SEM of magnified view of a pulled-out
fiber during flexural failure (fiber content 10 wt %).
observed for composites of fiber length 10 and 30

mm (critical fiber length). In fact, there is a decrease in impact strength for composites of higher
fiber length (40 mm).
The energy-absorbing mechanisms built in the
composite include
1. Utilization of the energy required to debond the fibers and pull them completely
out of the matrix.
2. Use of a weak interface between the fiber
and the matrix.15
03-28-97 17:35:12
(3)
where d the fiber diameter, t the interfacial

frictional stress, and v the volume fraction of
the fiber. Hence, U1 } l 2 for l lc .
The energy reaches a maximum when l lc
and will have a value
Umax
vtl 2c
12d
(4)
Strictly speaking, the work of fracture arising

from the fibers and matrix and the energy to debond the fibers should be added to this energy.
Table IV Variation of Work of Fracture

(Impact Strength) with Fiber Length for PALF
Polyester Composites (Fiber Content 30 Wt %)
Fiber Length
(mm)
Work of Fracture
(kJ m02)
10
20
30
40
24.4
19.9
24.2
22.0
It is clear that the conditions leading to (1) and
/ 8e65$$4072
v tl 2
12d
polaa
W: Poly Applied
4072
1746
Figure 12 SEM of fractograph of the composite during impact test (fiber content 30 wt %).
Figure 11 Variation of work of fracture of PALF

polyester composites with fiber content (fiber length 30
mm).
However, in many cases of practical interest, the

energy of pullout considerably exceeds the other
contributions. Thus, for maximum toughness, it
is desirable to use fibers having a length l lc
based on eqs. (3) and (4) since Umax } l(lc ) when
l lc . The observed high impact strength of the
composites with fiber length 10 and 30 mm (critical fiber length) (Table IV) can thus be explained.
Also, progressive delamination occurs in these
cases, resulting in high energy absorption. However, for fibers having length greater than the critical length, only a proportion of the fibers will
pullout and then the energy is given by
U2
vtl 3c
12dl
and the work of fracture for a 30 wt % composite

is found to be 24 kJ m02 . It is about 1200% greater
than that of pure polyester resin.
It is generally accepted that the toughness of
a fiber composite is dependent mainly on the fiber
stressstrain behavior.18 Pineapple fibers are
comparatively strong fibers with high failure
strain and therefore impart a high work of fracture to the composites. The fiber pullout, interface
fracture, and delamination are the major contributions toward the high toughness of these composites.
SEM of the impact fracture surfaces of PALF
polyester composites containing 30 and 10 wt %
fibers are shown in Figures 12 and 13. Figure 12
shows mainly the corrugated surface of the fiber
and Figure 13 shows debonding between PALF
and polyester.
(5)
The impact energy or strength decreases as

U2 }
1
l
The slight decrease in impact strength for composites with fibers of length 40 mm can thus be explained.
The work of fracture (impact strength) of the
composites is found to increase linearly with the
weight fraction of the fiber as shown in Figure 11
/ 8e65$$4072
03-28-97 17:35:12
Figure 13 SEM of fracture surface (impact) of

PALFpolyester composites (fiber content 10 wt %).
polaa
W: Poly Applied
4072
1747
Table V Effect of Chemical Treatment on Mechanical Properties (Fiber Length 30 mm, Fiber
Content 30 Wt %)
Type/Nature of Treatment
Untreated
2% NaOH (treated for 1 h)
Silane A172
Silane A1100
Glacial acetic acid and
acetic anhydride
Tensile
Strength
(MPa)
Elongation at
Break (%)
Youngs Modulus
(MPa)
Flexural
Strength
(MPa)
Flexural Modulus
(GPa)
52.9
55.4
73.5
52.7
3.6
4.7
4.3
3.5
2290
1460
2690
80.2
77.4
85.6
79.6
2.76
2.93
2.84
42.7
3.6
2260
71.5
2.55
Effect of Chemical Treatment on Mechanical

Properties
The performance of composites is to a greater extent influenced by the interfacial interaction between the fiber and the matrix. The interfacial
bond may be improved by surface treatment.
Table V shows the mechanical properties of
composites obtained with chemically treated
PALF. Reagents such as NaOH, silane coupling
agents (silane A172 and A1100), and glacial acetic acid were used for treatment.
NaOH treatment showed marginal improvement in tensile strength. However, flexural
strength and the Youngs modulus decreased. The
slight enhancement in tensile strength in these
composites is attributed to the improved wetting
of alkali-treated PALF with polyester. NaOH
treatment is mainly a process of surface activation. It leads to the formation of a rough surface
which would increase the mechanical interlocking
between the fiber and the polyester.
A comparison of the effect of two silane coupling agents on the mechanical properties of the
composites was carried out. A 40% increase in
the tensile strength was observed when the fibers
were treated with silane A172 [vinyltri(2-ethoxy
methoxy)silane]. The flexural strength of these
composites was also increased by about 7%. In
silane A172-treated composites, the alkoxy
groups of silane hydrolyze to form silanols
({OH). This {OH group interacts with the
{OH groups of lignocellulosic PALF, forming hydrogen bonds, and the vinyl group reacts with the
polyester. This would cause the resin to be less
interconnected, resulting in a higher elongation
of the silane-treated composites. Addition of the
coupling agent silane A1100 has improved the
Youngs modulus of the composites marginally.
However, other properties were unaffected. Thus,
/ 8e65$$4072
03-28-97 17:35:12
the treatment of the fibers with silane A1100 has

a negligible effect on the properties of the composite as it is a nonreactive type on the basis of its
reactivity with the polyester matrix. The mechanical properties of the composites with acetylated
fibers were found to be less than those of the untreated fiber composites.
Physical Properties
The density and hardness of the composites with
varying fiber content is given in Table VI. The
density of the composites slightly decreased with
fiber content for a low weight fraction (10 wt %),
but remained steady for a high weight fraction of
the fiber. Incorporation of the fiber reduces hardness at a low weight fraction of the fiber. However,
at high fiber concentration, the hardness values
of the composites are found to increase.
Comparison with Other Short Natural Fiber
A comparison of the mechanical properties of the
PALFpolyester composite with other natural fiberpolyester composites at 30 wt % fiber content
Table VI Variation of Density and Hardness of

PALFPolyester Composites with Fiber Content
(Fiber Length 30 mm)
Fiber Content
(Wt %)
Density
(g/cm3)
Hardness
(Shore D)
0
10
20
30
40
1.15
1.03
1.10
1.09
1.09
82
73
75
83
84
polaa
W: Poly Applied
4072
1748
Table VII Properties of Natural Fiber Polyester Composites Prepared by Hand Layup Method (Fiber
Content 30 Wt %)
Property
Strawreinforced
Polyester
Sisal Polyester
PALFPolyester

Flexural strength (MPa)
Impact strength (unnotched) (kJ/m2)
47
2.6
28
53
11
52.7
80.2
24.2
Source from Refs. 5 and 7.
is given in Table VII. The mechanical properties

(tensile, flexural, and impact) of short PALF
polyester composites are significantly higher than
those of straw and sisal fiber polyester composites. This is due mainly to the high cellulose content of PALFs.
CONCLUSION
The results of the present study showed that a
useful composite with good strength could be successfully developed using pineapple fibers as a reinforcing agent for the polyester matrix. The optimum length of the fiber required to obtain PALF
polyester composites of maximum properties was
found to be 30 mm. The stressstrain behavior in
tension reveals that neat polyester is brittle and
the addition of fibers makes the matrix more ductile. The tensile strength and Youngs modulus of
these PALF polyester composites increased linearly with the fiber weight fraction. But in the
case of flexural strength, there is a leveling off
beyond 30%. The impact strength also increased
linearly with the weight fraction of the fiber. The
composite with 30 wt % fiber content exhibits an
impact strength of 24 kJ/m 2 . The high toughness
of this natural fiber polymer composite places it
in the category of tough engineering materials. A
significant increase in the strength of the composites was observed after treatment of the fibers.
The best improvement was observed in the case of
silane A-172-treated fiber composites. The PALF
polyester composites exhibit superior mechanical
properties when compared to other natural-fiber
polyester composites and can be used as structural composites.
/ 8e65$$4072
03-28-97 17:35:12
REFERENCES
1. P. S. Mukherjee and K. G. Satyanarayana, J.

Mater. Sci., 21, 51 (1986).
2. A. N. Shah and S. C. Lakkad, Fiber Sci. Technol.,
15, 41 (1981).
3. H. Wells, D. H. Bowen, I. Machphail, and P. K. Pal,
in Proceedings of the 35th Annual Technical Conference on Reinforced Plastic Composites, Society of
the Plastics Industries, New York, 1980.
4. D. S. Varma, M. Varma, and I. K. Varma, Text. Res.
J., 54, 827 (1984).
5. K. Joseph, PhD. Thesis, Mahatma Gandhi University, Kottayam, 1992.
6. A. R. Sanadi, S. V. Prasad, and P. K. Rohatgi, J.
Mater. Sci., 21, 4299 (1986).
7. N. M. White and M. P. Ansell, J. Mater. Sci., 18,
1549 (1983).
8. K. Joseph, S. Thomas, C. Pavithran, and M. Brahmakumar, J. Appl. Polym. Sci., 47, 1731 (1993).
9. K. Joseph, C. Pavithran, and S. Thomas, to appear.
10. S. Varghese, B. Kuriakose, S. Thomas, and A. T.
Koshy, Ind. J. Nat. Rubb. Res., 4, 55 (1991).
11. P. S. Mukherjee and K. G. Satyanarayana, J.
Mater. Sci., 21, 51 (1986).
12. C. Pavithran, P. S. Mukherjee, M. Brahmakumar,
and A. D. Damodaran, J. Mater. Sci. Lett., 6, 882
(1987).
13. J. George, K. Joseph, S. S. Bhagawan, and S.
Thomas, Mater. Lett., 18, 163 (1993).
14. J. George, S. S. Bhagawan, and S. Thomas, J. Appl.
Polym. Sci., 57, 843 (1995).
15. M. J. Folkes, Short Fiber Reinforced Thermoplastics, Wiley, New York, 1982.
16. G. A. Copper, J. Mater. Sci., 5, 645 (1970).
17. A. H. Cottrell, Proc. R. Soc. A, 282, 2 (1964).
18. T. Fujii, J. Jpn. Soc. Comp. Mater., 1, 35 (1975).
polaa
W: Poly Applied
4072
Ageing Studies of Pineapple Leaf FiberReinforced

L. Uma Devi,1 Kuruvila Joseph,2 K. C. Manikandan Nair,3 Sabu Thomas3
1
St. Thomas College, Kozhencherry, Kerala, India

St. Berchmans College, Changanacherry, Kerala, India
3
School of Chemical Sciences, Mahatma Gandhi University, Priyadarshni Hills P.O., Kottayam, Kerala, India
2
Received 7 November 2003; accepted 29 April 2004

DOI 10.1002/app.20924
Published online in Wiley InterScience (www.interscience.wiley.com).
ABSTRACT: This experimental study evaluated the water
absorption characteristics of pineapple leaf ber (PALF)
polyester composites of different ber content. The degree of
water absorption was found to increase with ber loading.
The mechanism of diffusion was analyzed and the effect of
ber loading on the sorption kinetics was studied. The diffusion coefcient was calculated and found to increase with
ber content. Studies were also made to correlate water
absorption with the cross-sectional areas of the specimens.
INTRODUCTION
Composite materials involving ber reinforcement are
replacing traditional materials at a fast rate. Because
natural bers such as sisal, jute, banana, and pineapple are renewable and relatively less expensive than
their synthetic counterparts, these promising bers are
used as a substitute for synthetic bers in polymer
composites. Natural bers possess moderately high
specic strength and stiffness and can be used to make
useful structural composites.1 8
Synthetic berreinforced composites impart good
long-term behavior to various aggressive environments and an enhancement in strength and stiffness.
However, it is found that natural berreinforced
polyester materials are more or less sensitive to humidity through absorption of water, leading to physical degradation such as plasticization of the matrix
with water and the differential swelling between the
bers and the resin. The uptake of water can also lead
to chemical degradation such as hydrolysis of the
matrix. Consequently, the matrix structure and ber
matrix interface are affected, resulting in changes of
bulk properties such as dimensional stability and both
mechanical and electrical properties.
Pineapple leaf bers (PALF) are easily available and
they possess excellent mechanical properties.9 These
Correspondence to: S. Thomas (sabut@sancharnet.in).
Journal of Applied Polymer Science, Vol. 94, 503510 (2004)
2004 Wiley Periodicals, Inc.
The effects of ageing on the tensile properties and dimensional stability of PALF polyester composites were studied
under two different ageing conditions. Ageing studies
showed a decrease in tensile strength of the composites. The
composite specimens subjected to thermal ageing showed
only a slight deterioration in strength. 2004 Wiley Periodicals, Inc. J Appl Polym Sci 94: 503510, 2004
Key words: diffusion; ageing; polyesters; bers; composites
bers show high ultimate tensile strength and initial

modulus because they have high cellulose content and
comparatively low microbrillar angle. Also its composites possess superior mechanical properties.10 Because they are lignocellulosic, PALF bers are hydrophilic in nature. It is well established that natural
bers that are hydrophilic in nature are susceptible to
moisture absorption.11 Water transport is strongly affected by the nature of bermatrix interface. If the
interface is strong, it is difcult for the water molecules to diffuse into the composite system. Upon various treatments of the ber and matrix resin, one can
increase the interfacial interaction, thereby making the
composite more resistant to water transport. To
achieve the full potential of polymer composites, they
must have good environmental conditions. Environmental conditions such as exposure to humid environment or change in temperature and exposure time
cause dimensional instability and loss of mechanical
properties. Absorbed moisture has many detrimental
effects on material performance.12,13 Sapieha et al.14
studied the kinetics and equilibrium of water sorption
in polyethylene/cellulose composites. Several other
studies have been reported concerning the effect of
different ageing conditions on the physical and mechanical properties of natural berreinforced polymer composites.1517
PALF is obtained from the leaf of the plant Ananas
cosomos belonging to Bromeliaceae family. PALF is used
as potential llers for thermosets, thermoplastics, and
elastomers. George et al.18 20reported on the mechanical properties, rheological behavior, and viscoelastic
504
UMA DEVI ET AL.
TABLE I
Physical and Mechanical Properties of Pineapple Leaf
Fiber and Polyester Resin
Ageing studies
Property
PALF
Polyester
Density, g cm3
Tensile strength, MPa
Elongation at break, %
Youngs modulus, MPa
1.526
170
3
6210
1.159
22.9
1.6
582
properties of PALF-reinforced polyethylene composites. Uma Devi et al.21 reported the mechanical properties of PALFpolyester composites. The present
study was aimed at evaluating the water absorption,
ageing characteristics, and the effect of different ageing conditions on the mechanical properties of PALF
polyester composites.
Water absorption studies of the composites at room

temperature were carried out according to ASTM D
570-81. The test specimens were 75 mm long, 25 mm
wide, and 4 and 3 mm thick. The specimens were
dried in an air oven for 24 h at 80C, cooled in a
desiccator, and immediately weighed. The conditioned specimens were placed in a container of distilled water maintained at room temperature and the
samples were completely immersed. At the end of 2,
24, and 96 h, the specimens were removed from water
one at a time, the surfaces were wiped off with a dry
cloth, and samples were weighed to the nearest 0.0001
g. Weight measurements were repeated at the end of
the rst week and every 2 weeks thereafter until the
increase in weight per 2-week period, by three consecutive measurements, remained more or less constant.
The water absorbed by the sample was then determined.
EXPERIMENTAL
Immersion in boiling water
Materials
Pineapple leaf ber (PALF) was obtained from South
India Textile Research Association (Coimbatore).
Washed and dried bers were used for the study.
General-purpose polyester resin (HSR, 8131, supplied
by Bakelite Hylam Ltd., Hyderabad, India) was used
for the study.The properties of PALF and polyester
resin are given in Table I. Composites made from
PALF and polyester resin of 30 wt % were used for
this study.21 The properties are detailed in Table II.
Preparation of PALFpolyester composites
Randomly oriented pineapple berpolyester composites of 30 mm ber length, with varying ber loadings, were prepared by hand layup method. Composite sheets (size 150 150 3 mm3) were prepared
using a closed mold. Polyester resin was mixed with 1
wt % cobalt naphthanate (accelerator) and 1 wt %
methyl ethyl ketone peroxide (catalyst), degassed, and
then poured onto the mat placed in the mold. When
the mat was completely wet by the resin, the mold was
closed and kept under pressure for 24 h. Samples were
postcured and test specimens of the required size were
cut from the sheets and used for testing.
Test samples were immersed in boiling water for 7 h at

atmospheric pressure. These samples were then dried
between two sheets of lter paper and the weight and
dimension of each specimen were measured. The samples were then subjected to tensile measurements.
Thermal ageing
Test samples were heated at 100C in an air oven for
48 h. After being conditioned at room temperature, the
tensile properties of these samples were determined
by an Instron (Canton, OH) UTM model 1121.
Mechanical testing
Tensile testings were carried out using an Instron
Universal testing machine model 1121, according to
ASTM D 638, at a crosshead speed of 5 mm min1 and
a gauge length of 60 mm. The specimens were
clamped with screws.
THEORY
The percentage increase in weight of the specimens
after immersion in water for various time intervals
was calculated to the nearest 0.01% as
TABLE II
Properties of PALFPolyester Composites of Different Fiber Loading
Fiber content
(wt %)
Density
(g cm3)
Tensile
strength (MPa)
Elongation
at break (%)
Youngs
modulus (MPa)
10
30
40
1.03
1.09
1.09
17.1
52.9
63.3
1.3
3.6
5.0
1767
2290
2519
PINEAPPLE LEAF FIBERREINFORCED COMPOSITES
Increase in wt %
Net wt conditioned wt 100
Conditioned wt
(1)
(2)
To study the mechanism of water sorption, the kinetic parameters n and k were analyzed from the
following relationship:
(3)
where Q is the mol % uptake at equilibrium, t is the

time, and k is a constant characteristic of the polymer,
which indicates interaction between polymer and water.
The values of diffusion coefcients were calculated
from the relationship
D h /4Q 2
(4)
where is the slope of the initial linear portion of the

sorption curves, h is the initial thickness of the sample,
and Q is the mol % uptake at equilibrium.
The sorption coefcient (S) or solubility can be calculated from the equilibrium swelling using the relation
S
M
Mp
(5)
where M is the mass of the solvent taken up at

equilibrium and Mp is the initial mass of the polymer
sample. The permeability coefcient, which could be
considered as the total effect of sorption and diffusion,
can be estimated using the following equation:
P DS
Ddt
b2
0.75
1 Vd 1 exp 7.3
Q t (mol %)
log Qt/Q log k n log t
Qt V d 1 exp 7.3
The mol % uptake Qt is given by
Weight of water taken up at time t

100
Molecular wt of the solvent
initial wt of the sample
505
(6)
where D is the diffusion coefcient and S is the solubility.

The JacobsJones model is used to describe the diffusion processes in two-phase materials consisting one
dense and one less-dense phase. According to the
model the diffusion proceeds by parallel Fickian processes in both phases.22,23 The variation of normalized
water content Q(t) with time can be described by the
morphology-dependent equation:
Dlt
b2
0.75
(7)
where Q(t) Qt/Q; Dd and Dl are the diffusion

coefcients of the dense and the less-dense phases,
respectively; Vd is the volume fraction of the dense
phase; and b is the thickness of the sample. In the case
of PALFpolyester composites, the material can be
considered as a two-phase system24 with the nonpolar
polyester matrix as one phase (phase 1) and the ber
as second phase (phase 2). In both phases, the diffusion takes place through the formation of hydrogen
bonds between the hydroxyl groups of ber and water
molecules. In this case the JacobsJones model can be
modied to eq. (8), as follows:
Qt Q 1 1 exp 7.3
D1t
b2
0.75
Q2 1 exp 7.3
D2t
b2
0.75
(8)
where D1 and D2 and Q(1) and Q(2) are the diffusion

coefcients and the normalized water content of absorbed water at saturation in the two phases, respectively.
Effect of water absorption
Water absorption in a brous composite depends on
such factors as ber loading, area of exposed surfaces,
and diffusivity. The results of diffusion experiments
were expressed as mol % uptake of water by 100 g of
the polymer. The mol % uptake (Qt %) was plotted
against the square root of immersion time (in minutes)
to obtain the sorption curves in Figure 1. At equilibrium, Qt was taken as Q.
Water absorbed by the polyester sample was found
to be very small in the initial stage. However, if
crosslinked unsaturated polyester specimens are subjected to a lengthy immersion, the water absorption
was found to be about 2% at equilibrium.
For PALFpolyester composites, the degree of water
absorption increased with ber loading. From the
sorption curves it is seen that the uptake increased
linearly at rst, after which it leveled off, thus indicating the attainment of equilibrium. The percentage of
water absorbed by 40% composites was 123% higher
than that of composites containing 30% ber at saturation point. For hydrophilic bers such as PALF, an
increase in water-sorption properties of the composite
may be expected with the increasing concentration of
506
Figure 1
UMA DEVI ET AL.
Sorption curves showing the mol % uptake of PALF/polyester composites having different ber loadings.
ber. These bers swell with absorption of water,

which results in the development of shear stress along
the matrixber interface, leading to delamination and
debonding. Also the absorption of water is attributed
to the capillary phenomenon, although the presence of
hydroxyl groups enhances the absorption of water by
forming hydrogen bonding.
Figure 2 shows the variation of Q with ber loading. From the gure, the rate of absorption is found to
depend on cellulose content. This is because, with an
increase in ber loading, the rate of water absorption
increases as a result of the increase in total cellulose

content. The sample containing lower ber concentration (i.e., with 10 and 20% weight fraction of ber)
reached equilibrium more rapidly (i.e., after 1 week),
whereas the samples with 30 and 40% ber content
reached the saturation point after 3 weeks. Furthermore, a very rapid initial sorption of water was observed in the case of composites with high ber content (40%), which could be attributable to the uptake
of water by the exposed cellulosic bers located at the
surface of the composite. The transport of moisture
Figure 2 Variation of Q with ber loading.
507
TABLE III
Values of n, k, Q, Diffusion Coefcient (D), Sorption Coefcient (S), and Permeability Coefcient (P), for
PALFPolyester Composites with Different Fiber Loading
Fiber
loading
(wt %)
D (cm2 s1)
S (g/g)
P 1010
(cm2 s1)
0
10
20
30
40
0.320
0.256
0.243
0.274
0.268
1.570
1.180
1.094
1.264
1.193
0.1156
0.1996
0.2386
0.4503
0.9909
1.0 109 1.3 1010

1.3 109 2.7 1010
1.6 109 4.4 1010
2.3 109 0.424 1010
2.0 109 1.3 1010
0.020
0.03
0.05
0.08
0.18
0.248
0.42
0.84
1.79
3.64
below the glass-transition temperature (Tg) is a threestage process in which the absorbed moisture rst
occupies the free volume present. In the second stage,
water becomes bound to network sites causing swelling. Finally, water enters the densely crosslinked regions. According to the equilibrium theory, a polymer
will absorb solvent until the solvent chemical potential
in the polymer and that in the free solution are equal.25
Kinetics of water sorption
The equilibrium sorption values Q of the composites
and the values of n and k, determined by linear regression analysis, are given in Table III. The value of n
provides information about the mechanism of solvent
transport under all experimental conditions. If n 0.5,
the mechanism is said to be Fickian.26 Here at all ber
loadings, n is less than 0.5. Thus a deviation from
Fickian behavior is observed in all cases.
Diffusion, sorption, and permeation
percentage increase in the cross-sectional area of the

specimens after immersion in distilled water for 7
weeks is shown as a function of ber loading in
Figure 3.
The polyester samples showed no measurable area
change, even after immersion in water for 7 weeks.
The cross-sectional area of PALFpolyester composites increased after immersion in water for 7 weeks,
reaching a maximum for 30 wt %, after which it leveled off.
Modeling of water sorption: JacobsJones model
The experimental results of water diffusion in PALF
polyester composites are tted by eq. (8) by a nonlinear least-square t method and Figure 4 shows a typical experimental and tted diffusion curve for 30%
composite at 28C. The tted values of Q(1), Q(2), D1,
and D2 are listed in Table IV. The theoretical model
shows good agreement with the experimental values.
This conrms the existence of two different phases
and mode of diffusion in PALFpolyester composites.
The values of diffusion coefcients obtained are also

given in Table III. The D values increased with increase in ber loading up to 30 wt % and thereafter
decreased. The increase is attributed to the inherent
hydrophilic nature of bers. Thus the ability of water
molecules to diffuse through the polymer increases.
Sorption describes the initial penetration and dispersal of penetrant molecules into the polymer matrix.
Solubility is a thermodynamic parameter that depends
on the strength of the interaction in the polymer/
penetrant mixture. The values of S and P are given in
Table III. The sorption coefcient values increased
with increasing ber loading and is maximum for 40
wt % composites. The permeation coefcient also increased for composites with 10 to 40 wt % loading.
Compared to D and P, the value of S is higher, which
implies an increased tendency of the penetrant to dissolve into the polymer.
Effect of water sorption on dimensional stability
Studies have been made to correlate water absorption
with the cross-sectional area of the specimens. The
Figure 3 Effect of ber content on the increase in area of

PALF/polyester composites after immersion in water for 7
days.
508
UMA DEVI ET AL.
TABLE V
Tensile Strength of PALFPolyester Composites After
Immersion in Boiling Water for 7 ha
Sample
Polyester
Unaged
Aged
Composite
Unaged
Aged
a
Figure 4 A typical experimental and tted diffusion curve

for 30% composites at 28C.
The slight deviation from the experimental curve results from the deviation of diffusion processes from
Fickian behavior. This was very clear from the n values.
Immersion in boiling water
Table V shows that the tensile strength of pineapple
berpolyester composites decreases on immersion in
boiling water. The samples subjected to water ageing
showed a decrease of 34%. The natural bers absorb
water and swelling of the ber increases with exposure time in boiling water. As a result of the swelling
of bers, which are surrounded by the polymer matrix, cracks may be formed at the matrix. This may also
Tensile
strength (MPa)
Elongation
at break
(%)
Modulus
(MPa)
20.6
13.7
2
1
580
1600
52.9
34.5
4
3
2290
1900
Fiber length: 30 mm; ber loading: 30 wt %.
contribute to the penetration of more water into the

composite during prolonged exposure in boiling water. The modulus of the composite was found to decrease with exposure in boiling water. The natural
bers absorb water because of its hydrophilic nature.
Therefore as a consequence of the high water content
after exposure in boiling water, the stiffness of the
cellulose ber decreased substantially.27 At 30 wt %
loading, because of the excessive absorption of water,
the intramolecular hydrogen bonding in the cellulose
ber might be reduced by the formation of intermolecular hydrogen bonding between cellulose molecules and water molecules, thus leading to deterioration of the modulus.28 The elongation at break values
of neat polyester samples and PALFpolyester composites decreased slightly after exposure in boiling
water, although the modulus of the neat polyester
increases. The increased temperature leads to an increase in crosslinking, which thereby increases the
modulus.
Dimensional stability of the composites was studied
by measuring the changes in weight and cross-sectional area of the composites after exposure in boiling
water for 7 h.
From Table VI it is clear that the absorption of water
in the polyester samples was quite low (0.4%), although in the case of PALFpolyester composites, the
change in weight was found to be 6%. An increase in
cross-sectional area was also found as a result of exposure in boiling water. The polyester specimens
showed no measurable area change after exposure;
TABLE IV
Fitting Values as per Eq. (8) for Composites at 28C
Fiber content
(wt %)
log D1
Q(1)
log D2
Q(2)
0
10
20
30
40
12.1344
11.8464
11.2448
11.4730
11.4152
0.10467
0.01304
0.89905
0.54737
0.44619
11.8742
11.3806
11.6818
11.3798
11.4168
0.89533
0.98696
0.10095
0.45265
0.5536
0.11559
0.19485
0.23599
0.4503
0.9906
however, in PALFpolyester composites, an increase

in cross-sectional area was observed after exposure in
boiling water. The increases in weight and cross-sectional area are attributed to absorption of water by the
hydrophilic bers in the composites. Because raw bers were used in the fabrication of these composites,
the water absorption was high, thus leading to deterioration in mechanical properties as well as dimensional stability.
Thermal ageing
The mechanical properties of the composites after
thermal ageing at 100C are shown in Table VII. In the
case of tensile strength values a decrease of 6% was
observed upon thermal ageing; however, the modulus
is decreased by 51%. Therefore the decrease of tensile
strength is lower, whereas in the boiling water ageing,
a decrease of 34% in tensile strength is observed. This
may be a result of the loss of strength of ber during
prolonged exposure at elevated temperature (100C)
caused by the decomposition of volatile extractables
present on the ber surface.
Although the components of cellulose are supposed
to remain stable up to 160C the degradation reaction
may be initiated slowly on prolonged heating at
100C.16 The degradation of ber will develop voids at
the interface and this leads to poor bermatrix adhesion. The modulus of the composite was also found to
be less than that of the unaged specimens, although
the elongation at break was not affected by ageing at
elevated temperature. The loss in mechanical properties of PALFpolyester composites after water ageing
and thermal ageing are comparable with the results of
banana ber polyester composites.29
CONCLUSIONS
Results of this study indicate the environmental effects
on the tensile properties of PALF-reinforced polyester
composites. Ageing studies showed a decrease in tensile strength of about 6% for samples subjected to
thermal ageing, whereas the composite specimens
subjected to water ageing showed considerable loss in
tensile strength, about 34%. The water uptake of the
composites was found to increase with ber loading
TABLE VI
Variation in Weight and Cross-Sectional Area of the
Polyester and PALFPolyester Composite Specimensa
Sample
Increase in wt %
Increase in area %
Polyester
PALFpolyester
composites
0.4
Nil
1.3
Fiber length: 30 mm; ber loading: 30 wt %.
509
TABLE VII
Tensile Properties of PALFPolyester Composites After
Thermal Ageing 100C, 48 h; Loading 30 wt %
Sample
Tensile
strength
(MPa)
Elongation
at break
(%)
Modulus
(MPa)
Unaged
Aged
52.9
49.8
4
4
2290
1110
and leveled off at longer times. The kinetic parameters

of water sorption of composites of different ber loadings were calculated. The diffusion coefcient and
sorption coefcient were found to increase with increasing ber content of the composites. Water absorption leads to a slight deterioration in mechanical
properties and dimensional stability. Models were
used to t experimental data, which showed good
agreement with theoretical values initially but slight
variation at equilibrium. The slight deviation from
experimental values may be attributed to the deviation of diffusion from Fickian behavior.
References
1. Bisanda, E. T. N.; Ansell, M. P. Compos Sci Technol 1991, 41,
165.
2. Pavithran, C.; Mukherjee, P. S.; Brahmakumar, M.; Damodaran,
A. D. J Mater Sci Lett 1987, 7, 882.
3. Varma, I. K.; Ananthakrishnan, S. R.; Krishnamoorthy, S. Composites 1989, 20, 383.
4. Bismark, A.; Aranberri Askargorta, I.; Springer, J.; Lampke, T.;
Wielage, B.; Shendervich, A.; Limbach, H. H. Polym Compos
2002, 23, 872.
5. Rong, M. Z.; Zhang, M. Q.; Liu, Y.; Zhang, Z. W.; Yang, G. C.;
Zeng, H. M. Polym Polym Compos 2002, 10, 407.
6. Rozman, H. D.; Saad, M. J.; Ishak, Z. A. M. J Appl Polym Sci
2003, 87, 827.
7. Joseph, P. V.; Rabello, M. S.; Mattoso, L. H. C.; Joseph, K.;
Thomas, S. Compos Sci Technol 2002, 62, 1357.
8. Sreekala, M. S.; Kumaran, M. G.; Thomas, S. Composites A 2002,
33.
9. Mukherjee, P. S.; Satyanarayana, K. G.; J Mater Sci 1986, 21, 51.
10. Pavithran, C.; Mukherjiee, P. S.; Brahmakumar, M.; Damodaran,
A. D. J Mater Sci Lett 1987, 6, 882.
11. Mckenzie, A. W.; Yuritha, J. P. Appita 1979, 32, 460.
12. Komorowskj, J. P. National Aeronautical Establishment, National Research Council of Canada, Aeronautical Note NAEAN-10, NRC No. 22700, 1983.
13. Weitsman, Y. In: Comprehensive Composite Materials; Talreja,
R.; Manson, J. A. E., Eds.; Polymer Matrix Composites, Vol. 2;
Elsevier: New York, 2000.
14. Sapieha, S.; Caron, M.; Schreiber, H. P. J Appl Polym Sci 1986,
32, 5661.
15. Maldas, D.; Kokta, B. V.; Polym Plast Technol Eng 1990, 29, 419.
16. Maldas, D.; Kokta, B. V.; Daneault, C.; Koran, Z. Drevarsky
Vyskum 1990, 125, 1.
17. Joseph, K.; Thomas, S.; Pavithran, C. Compos Sci Technol 1995,
53, 99.
18. George, J.; Bhagwan, S. S.; Prabhakaran, N.; Thomas, S. J Appl
Polym Sci 1995, 57, 843.
510
19. George, J.; Janardhanan, R.; Anand, J. S.; Bhagwan, S. S.

Thomas, S. Polymer 1996, 37, 5421.
20. George, J.; Joseph, K.; Bhagawan, S. S.; Thomas, S. Mater Lett
1993, 18, 163.
21. Uma Devi, L.; Bhagawan, S. S.; Thomas, S. J Appl Polym Sci
1997, 64, 1739.
22. Jacobs, P. M.; Jones, F. R. J Mater Sci 1989, 24, 2331.
23. Jacobs, P. M.; Jones, F. R. J Mater Sci 1989, 24, 2343.
24. Maggna, C.; Pissis, P. J Polym Sci Part B: Polym Phys 1999, 37, 1165.
UMA DEVI ET AL.
25. Omidian, H.; Hashemi Fahimeh Askari, S. A.; Nasi, S. V. Ind J

Polym Sci Technol 1994, 32, 115.
26. Lucht, L. M.; Peppas, N. A. J Appl Polym Sci 1987, 33, 1557.
27. Ray, R. G.; Kokta, B. V.; Danaeault, C. J Appl Polym Sci 1990, 40,
645.
28. Sapreha, S.; Pupo, J. F.; Scheriber, H. P. J Appl Polym Sci 1989,
37, 233.
29. Pothen, L. A.; Thomas, S.; Neelakanthan, N. R. J Reinforced
Plast Compos 1997, 16, 744.
Dynamic Mechanical Analysis of Pineapple Leaf/Glass

Hybrid Fiber Reinforced Polyester Composites
L. Uma Devi,1 S.S. Bhagawan,2 S. Thomas3

1
Department of Chemistry, St. Thomas College, Kozhencherry, Kerala 689641, India
2
Department of Polymer Engineering, Amrita Vishwa Vidyapeetham, Ettimadai, Coimbatore,

Tamil Nadu 641105, India
School of Chemical Sciences, Mahatma Gandhi University, Priyadarshini Hills, Kottayam, Kerala 686560, India
The dynamic mechanical properties of randomly oriented intimately mixed hybrid composites based on
pineapple leaf bers (PALF) and glass bers (GF) in unsaturated polyester (PER) matrix were investigated.
The PALFs have high-specic strength and improve
the mechanical properties of the PER matrix. In this
study, the volume ratio of the two bers was varied by
incorporating small amounts of GF such as PALF/GF,
90/10, 80/20, 70/30, and 50/50, keeping the total ber
loading constant at 40 wt%. The dynamic modulus of
the composites was found to increase on GF addition.
The intimately mixed (IM) hybrid composites with
PALF/GF, 80/20 (0.2 Vf GF) showed highest E0 values
and least damping. Interestingly, the impact strength
of the composites was minimum at this volume ratio.
The composites with 0.46 Vf GF or PALF/GF (50/50)
showed maximum damping behavior and highest
impact strength. The results were compared with
hybrid composites of different layering patterns such
as GPG (GF skin and PALF core) and PGP (PALF skin
and GF core). IM and GPG hybrid composites are
found more effective than PGP. The activation energy
values for the relaxation processes in different composites were calculated. The overall results showed that
hybridization with GF enhanced the performance properties. POLYM. COMPOS., 31:956965, 2010. 2009 Society
of Plastics Engineers
INTRODUCTION
In recent years, cellulose bers have become an important class of reinforcing materials [13]. Natural bers are
considered to be a potential alternative to glass bers
(GFs) for use in composites applications. Uses of natural
bers as reinforcement in polymers have gained imporCorrespondence to: L. Uma Devi; e-mail: umaharitham@yahoo.com
DOI 10.1002/pc.20880
Published online in Wiley InterScience (www.interscience.wiley.com).
C 2009 Society of Plastics Engineers
V
POLYMER COMPOSITES-2010
tance because of their eco-friendly nature. Their advantages also include low density, biodegradability, high
stiffness, and relatively low price. The properties of the
natural ber composites can be improved by hybridizing
these bers with GFs. The hybridization enhances the
strength, stiffness, and also improves the moisture-resistant behavior of the resulting composite. Several studies
are being undertaken in these days to replace GF partially
or fully by natural ber. GF is reported to have good reinforcing effect along with natural bers like sisal, jute,
hemp, etc [49].
Many studies have been reported on the dynamic mechanical behavior of hybrid composites of natural bers
and GF by Thomas and coworkers [10, 11]. Dynamic mechanical properties of randomly oriented intimately mixed
short banana/sisal hybrid ber reinforced polyester composites were reported by Idicula et al. [12]; maximum
stress transfer between the ber and the matrix was
obtained for composites having banana and sisal bers in
the ratio 3:1, which showed lowest impact strength and
minimum tan d value. Thermoplastic natural rubber was
recently reinforced by the short carbon and kenaf ber
hybrid systems [13]. The storage modulus, loss modulus,
and tan d of the untreated hybrid composite were found
better than those of the treated hybrid composites.
This work aims at developing high-performance, costeffective, and lightweight hybrid composites. It describes
studies on hybrid composites containing pineapple leaf
ber (PALF) and GF as the reinforcement. The PALFs
exhibit superior mechanical properties because of its
high-cellulose content (7082%) and comparatively lowmicrobrilar angle (148).These are extracted from the
leaves of the plant Ananus cosomos belonging to bromeliacease family by retting. Our earlier studies proved
PALF to be an effective reinforcement in polyester (PER)
and LDPE systems [14, 15].
TABLE 1. Physical and mechanical properties of PALF, GF, and

polyester resin.
Properties
PALF
GF
Polyester
Density (g/cm3)
Diameter (lm)
Youngs modulus (GPa)
1.526
3060
413
6.5
1.6
2.54
525
2500
70
2.5
1.16
22.9
0.97
1.6
The PALF is hybridized with small quantities of stronger GF and randomly oriented intimately mixed hybrid
composites and other layered hybrid composites GPG (GF
skin and PALF core) and PGP (PALF skin and GF core)
were prepared. The purpose of the study is to develop
new cost-effective structural composite material having
better performance properties using PALF and small
quantities of GF. Our previous study indicated that the
tensile properties of these composites are excellent [16].
In this article, their dynamic mechanical properties were
evaluated with respect to hybrid-ber ratio by varying the
volume fractions of GF and PALF.
EXPERIMENTAL
Materials
PALF was obtained from South India Textile Research
Association, Coimbatore. Isophthalic polyester HSR 8131
obtained from m/s Bakelite Hylam, Hyderabad, India is
used (viscosity at 258C-650 cps, Sp. gravity at 258C-1.11,
styrene content 3035%). Multidirectional E-glass strand
mat is supplied by Ceat Ltd. (Hyderabad, India). The
physical and mechanical properties of PALF, GF, and
polyester resin are given in Table 1.
posites were prepared by uniformly mixing the two bers

PALF and GF of required weights corresponding to the
different volume fractions of each. GFs of length 50 mm
(from the woven mat) were used for the preparation of
IM composites. The polyester resin was mixed with
1 wt% cobalt naphthenate (accelerator) and 1 wt% methyl
ethyl ketone peroxide (catalyst). The resin mixture was
degassed and poured on to the ber mat of uniform thickness prepared and placed in the mold. The closed mold
was pressed, allowed to cure at room temperature for
24 h under constant pressure (10 kg/cm2) and postcured
at 508C for 24 h.
Hybrid composites of different layering patterns such
as trilayer GPG (GF/PALF/GF), and PGP (PALF/GF/
PALF) were also prepared by hand lay-up technique mentioned earlier. GF mats were sized; both PALF and GF
mats of required weights were used in the preparation of
the layered composites. Trilayer PALF/GF hybrid ber
(GPG) composites were prepared by placing GF mats on
both sides and PALF in between and PGP by placing
PALF mats on both sides of the GF mat. The mats were
arranged in such a manner so as to get the two types of
bers maximally intermingled. The relative volume
fraction of the two bers is varied in all these patterns
keeping the total ber content constant at 40 wt%. A
schematic representation of different hybrid conguration
used in the study is given in Fig. 1.
Characterization of Composites. The dynamic mechanical analysis was carried out using TA Instruments
Thermal Analyzer. Dual cantilever method was used.
Measurements were made at three frequencies (0.1, 1, and
10 Hz).
RESULT AND DISCUSSION
Method
Preparation of PALF/GF Hybrid Polyester Composites. The PALF was chopped to their optimum length
30 mm [14], washed with water, dried in air, and used for
composite preparation. Hand lay-up method followed by
compression molding was adopted for composite fabrication. Composite sheets of size 150 3 150 3 3 mm3 were
prepared. Randomly oriented intimately mixed (IM) com-
Effect of Hybrid-Fiber Ratio

From a study of static mechanical properties [16], the
total ber loading of PALF/GF hybrid composite is optimized as 40 wt%. The relative volume fraction of GF is
varied from 0.1 to 0.46 Vf. In other words, the PALF/
GF ratio is varied as 90/10, 80/20, 70/30, and 50/50.
The description of PALF/GF ber ratio in the different
FIG. 1. Schematic representation of different hybrid congurations.
DOI 10.1002/pc
POLYMER COMPOSITES-2010 957
TABLE 2. Description of composite samples with different hybrid

congurations.
TABLE 3. Values of constant C, activation energies, tan d peak height

and impact strength of IM hybrid composites.
PALF/GF ratio
Volume
fraction of GF
Volume fraction of GF (Vf)
Intimately mixed (IM) and PGP PALF/GF hybrid

90/10
80/20
70/30
50/50
GPG hybrid
90/10
80/20
70/30
0.1
0.20
0.26
0.46
0.1
0.20
0.26
hybrid composites is given in Table 2. As this study

concentrates on preparing cost-effective composite, small
volume fractions of GF are incorporated along with
PALF. Hybridization with small amounts of GFs also
makes these composites more suitable for technical
applications such as automotive interior parts.
The improvement of the dynamic mechanical properties observed for the hybrid composites depends on the
hybrid-ber ratio and properties of the individual components of the composite. It is also affected by the strength
and efciency of the interfacial adhesion and interaction
balance between the PALF, GF, and the bulk PER matrix.
Behavior of Intimately Mixed Composites. The variation of E0 with temperature of intimately mixed (IM)
PALF/GF hybrid composites at increasing volume fraction of GF at a frequency of 10 Hz is shown in Fig. 2.
Comparing the different hybrid laminates, E0 increases
with the increase in the volume fraction of GFs and a
maximum value is obtained for 0.46 Vf (GF/PALF ratio
0
0.1
0.20
0.26
0.46
Constant
(C)
Activation energy
(kJ/mol)
Peak
height
Impact
strength (J/m)
0.0230
0.0206
0.0234
0.0276
106.1
210.9
314.9
207.7
104.4
0.129
0.126
0.128
0.135
0.150
236
899
862
1203
1128
50:50) in the glassy region. This is due to the fact that

the modulus of elasticity of GF is greater than that of
PALF. However, E0 values of the hybrid composite 0.2 Vf,
i.e., GF/PALF (20/80) is also very close to 0.46 Vf. This
has maximum value of storage modulus in the rubbery
region also. The storage modulus of 0.46 Vf GF is
decreased with increase in temperature and becomes close
to that of the unhybridized composite at high temperature.
However, comparatively low-volume fraction of GF 0.2 Vf
showed higher stiffness at all temperatures. The storage
modulus values of the composites are signicantly higher
than the resin. The extent of reinforcing action was evaluated by measuring the coefcient of reinforcement (C)
based on Eq. 1. The C values are given in Table 3. The
lower the value of C, the higher is the reinforcement.
C
E0G =E0R composite

E0G =E0R resin
where E0G and E0R are the storage modulus values in the
glassy and rubbery region respectively.
The C value decreases as GF loading is increased till
0.2 Vf and an increase is noticed there after. This implies
that the effectiveness of ber reinforcement is maximum
at this volume fraction (0.2 Vf) of the composite. On further increasing the volume fraction of GF, agglomeration
of GF takes place which decreases the effective stress
transfer between the ber and the matrix. The storage
modulus of the intimately mixed hybrid composites are
higher than the neat PALF/PER composite (0 Vf) at all
temperatures. The activation energy E for glass transition
of the hybrid composites can be calculated from the
Arrhenius equation
f f0 expE=RT
FIG. 2. Variation of storage modulus with temperature of PALF/GF

hybrid polyester (IM) composites of varying volume fractions of GF.
958 POLYMER COMPOSITES-2010
where f is the measuring frequency, fo is the frequency

when T approaches innity. The value of fo is a constant
characteristic of the substance, T is the absolute temperature corresponding to the maximum of tan d curve and R
is the gas constant. According to Arrhenius equation, a
plot of log f against 1/T should give a straight line with
slope proportional to the activation energy for the relaxation process. The higher value of activation energy
observed for 0.2 Vf (Table 3) showed effective stress
transfer between the ber and the matrix which decreases
the polymer chain mobility.
DOI 10.1002/pc
FIG. 3. (a) Variation of tan d with temperature of intimately mixed

composites at 10 Hz. (b) Variation of tan d with temperature of intimately mixed composites at 10 Hz (excluding neat resin).
The variation of tan d with temperature (10 Hz) for the

intimately mixed hybrid composites of different volume
fractions of PALF and GF is shown in Fig. 3a. The neat
polyester shows the highest damping. The tan d for the
0.46 Vf GF hybrid composites is comparatively high. The
mobility of the polymer molecules is decreased for composites having lesser GF content. The tan d values corresponding to the Tg is decreased when GF concentration is
lowered and the composite of 0.1 and 0.20 Vf GF showed
DOI 10.1002/pc
least damping. Strong interaction of ber and matrix takes

place at small volume fractions of GF which reduced the
mobility of molecular chains at the interface and as a
result damping is reduced.
Table 4 shows the tan d maximum and the Tg values
from tan d maximum of intimately mixed hybrid composites at different frequencies. The tan d maximum showed
a gradual increase with increasing volume fraction of GF
at all frequencies. A maximum value of tan d is observed
for GF/PALF ratio 50:50 at all frequencies. At higher
concentrations of GF, agglomeration takes place resulting
in incompatibility between the ber and the matrix. The
Tg from tan d shows a positive shift with increase of frequency and the mechanical damping of the composites is
decreased.
The PER resin showed the tan d peak at 1168C at
higher frequency. All the composites showed two relaxation peaks. The primary transition peak is shifted to
higher temperature 161 and 1638C for 0.1 and 0.2 Vf of
GF at 10 Hz. This is due to immobilization of the polymer molecules which increased the Tg of the composites.
The higher value of Tg in the case of 0.1 and 0.2 Vf
hybrid at 10 Hz shows better ber-matrix adhesion and
lower damping compared with composites of higher volume fraction of GF. These results are consistent with the
storage modulus measurements. The tan d peak height
given in Table 3 is also found to be lower for 0.1 and 0.2
Vf GF composite. As the Tg of the hybrid composite 0.2
Vf GF showed more positive shift than 0.1 Vf, it has least
damping and better interfacial bonding. Elevation of Tg is
taken as a measure of the interfacial interaction. The
broadening of the transition region is also more for 0.2 Vf
GF composite. Thus, at lower volume fractions of GF,
better interfacial interaction is observed in the composites.
The low-impact strength of 0.1 and 0.2 Vf GF composite
given in Table 3 is consistent with this result.
A small hump because of additional relaxation is
observed at a temperature lower than Tg in all the hybrid
composites. The additional relaxation is clearly observed
for lower volume fractions 0.1 and 0.2 Vf and more for
0.2 Vf composites (Fig. 3b). We believe that the additional relaxation could be due to the effect of the presence
of strong interphase in the composite. This is schematically shown in Fig. 4. The effect of the interlayer
TABLE 4. Tg and tan dmax values of IM hybrid composites.
Tg (8C) from tan
dmax frequency
(Hz)
tan d maximum
frequency (Hz)
Volume fraction of GF
Neat resin
0
0.1
0.20
0.26
0.46
0.1
10
0.1
10
0.350
0.166
0.170
0.166
0.170
0.180
0.557
0.133
0.133
0.131
0.135
0.150
0.798
0.129
0.126
0.129
0.135
0.150
114
113
116
106
107
110
116
143
130
115
121
139
116
160
161
163
156
154
TABLE 6. Values of E00 max and Tg from E00 max of IM hybrid polyester
composites.
E00 max (MPa),
frequency (Hz)
FIG. 4. Schematic representation showing the ber, matrix and the

interlayer.
becomes more prominent for 0.2 Vf GF composite. Eklind

and Maurer [17, 18] have reported on an interlayer model
to simulate the dynamic mechanical properties of lled
blends. The ller particles have been reported to be surrounded by an interlayer attached to the ller surface.
Similar results have been shown by Pothan et al. [19]. It
is noteworthy that the appearance of a relaxation process
below the glass transition of the polymer matrix improves
the toughness and impact strength of the material at low
temperatures [20]. The additional relaxation temperatures
of the hybrid composites are given in Table 5.
The Table 6 shows the E00 max values and Tg from
00
E max at various frequencies. Figure 5 shows the variation
of loss modulus E00 of the composites with the temperature at 10 Hz. Loss modulus also showed a similar trend
as in the case of storage modulus with variation of relative volume fraction of the bers. Maximum value of E00
is obtained for composite of higher GF content (0.46 Vf).
However 0.2 Vf GF composite showed higher loss modulus values at higher temperature. The energy loss per
cycle is directly related to loss modulus. A gradual
decrease in loss modulus is observed with increase of frequency. The Tg obtained from loss modulus is found to
be much lower than that from tan d curves. An additional
hump was observed in the E00 verses T curves of all composites even though the main transition is at a lower temperature. This hump became more prominent and the
curve is broader for 0.2 Vf composite.
The PALF and GF were found to be more effective independently in PER matrix [14, 21]. The distinct viscoelastic changes observed with the incorporation of hybrid
bers in the matrix were attributed to the formation of an
interphase because of strong interactions between the
reinforcing bers and the matrix. Several authors have
indicated the idea of an interphase immediately surround-
Tg from E00 max (8C),

frequency (Hz)
0.1
10
0.1
10
Neat resin
0
0.10
0.20
0.26
0.46
249
272
309
243
307
342
260
229
257
293
263
301
235
222
243
283
257
294
55
87
94
90
87
95
62
86
92
86
89
96
62
92
96
92
90
97
ing the reinforcing phase with properties intermediate of

the ber and the matrix [22]. This argument is also
applied to explain shifts in the Tg of the polymer matrix
with the addition of reinforcing phase. For composites
investigated in this work, the occurrence of chemical
bonding during curing is also assumed. A schematic diagram showing the probable bonding between PALF, GF,
and resin is shown in Fig. 6. In addition to physical interactions, the resulting covalent and hydrogen bonds
enhance the interfacial adhesion and improve the dispersion of bers in the matrix.
Effect of Layering Patterns. The volume fraction of
GF in the trilayer GPG (GF as the skin and PALF as the
core) is also varied up to 0.26 Vf as in the intimately
mixed hybrid composites. Figure 7 shows the variation
of storage modulus with temperature of GPG composite
TABLE 5. Tg from the additional relaxation in the IM hybrid

composites.
0
0.1
0.20
0.26
0.46
Tg (8C)
120
128
116
116
113
FIG. 5. Variation of loss modulus with temperature of PALF/GF hybrid

composites at 10 Hz.
DOI 10.1002/pc
FIG. 8. Variation of storage modulus with temperature of PGP hybrid

composites (10 Hz).
FIG. 6. Schematic diagram showing the probable bonding between the

PER resin, PALF, and GF.
of different volume fractions of GF at a frequency of

10 Hz.
On comparing the initial storage modulus values which
is the rst storage modulus recorded in the DMA modulustemperature plot of the GPG composite of different
volume fractions of GF, it is observed that the hybrid
FIG. 7. Variation of storage modulus with temperature of GPG hybrid

composites (10 Hz).
DOI 10.1002/pc
composites of comparatively lower volume fractions of

GF, i.e., 0.1 and 0.2 Vf have the highest values compared
with others. The GPG composite of 0.2 Vf exhibits comparatively higher storage modulus values all through the
temperature range. The values are found to be much
higher than the neat resin. The GPG composite of higher
volume fractions of GF, i.e., 0.26 Vf (PALF/GF 70/30)
had lower initial modulus values. However its E0 value is
observed to be higher above Tg.
The value of C given in the table decreases as GF
loading in the composite increases till 0.26 Vf reached.
The activation energy is maximum for 0.26 Vf composite
which is also higher than the intimately mixed hybrid
composite. Hence, the effectiveness of reinforcement of
GF is highest for 0.26 Vf hybrid composite. The increase
in values of activation energy with ber loading indicates
that mobility of the polymer chains is decreased in the
presence of bers [23] and thereby Tg is shifted to high
temperature. It can be concluded that viscoelastic properties are considerably affected by adding GF to PALF.
Maximum stress transfer between the ber and the matrix
takes place at this volume fraction.
On observing the variation of storage modulus with
temperature of trilayer PGP (PALF as the skin and GF as
the core) in Fig. 8, the storage modulus decreases with
increase in temperature for all volume fractions of GF.
The storage modulus is maximum for higher volume fractions of GF, i.e., 0.46 Vf (PALF/GF, 50:50) at all temperatures. This is consistent with the results observed in the
static mechanical property measurement of these composites [16]. Because PALF has lower tensile properties
compared with GF, the stiffness of the composite is
decreased when PALF is used as the skin material at
comparatively lower concentrations of GF. However, 0.1
sion between the ber and the matrix is evident in Fig. 9a

and b compared with 9(c).
The variation of tan d values of the layered hybrid
composites (GPG and PGP) with temperature at a frequency of 10 Hz is shown in Figs. 10 and 11. For GPG
composites, addition of 0.2 Vf GF slightly reduces the tan
d value of the hybrid composite when compared with
other volume fractions of GF which show that these
composites possesses low damping and interfacial adhesion. A wider and lower tan d peak is obtained for these
composite of GF/PALF ratio 20/80. Thus, the hybrid
FIG. 9. Scanning electron micrographs of tensile fracture surface of

(a) IM Vf (0.2Vf), (b) GPG (0.2 Vf), and (c) PGP (0.46 Vf) composites.
Vf GF composite exhibits comparatively higher storage

modulus values at higher temperatures. The storage modulus values of the PGP composites are found to be much
less than the GPG composites. The SEM of the tensile
fracture surface of the IM and GPG composites of 0.2 Vf
and PGP of 0.46 Vf are shown in Fig. 9ac. Good adhe962 POLYMER COMPOSITES-2010
FIG. 10. (a) Variation of tan d with temperature of GPG hybrid composites at 10 Hz. (b)Variation of tan d with temperature of GPG hybrid
composites at 10 Hz (excluding resin).
DOI 10.1002/pc
TABLE 7. The coefcient (C), tan dmax (Tg), tan d peak height and
activation energies of GPG hybrid composites.
Volume
fraction
of GF
Coefcient of
reinforcement
(C)
tan dmax
(Tg) at 10 Hz
Peak height of
tan d curve
at 10 Hz
Activation
energy
(kJ/mol)
0.022
0.021
0.010
157.5
157.9
163.6
0.129
0.128
0.146
212.8
368.3
503.3
0.10
0.20
0.26
The Tg from tan dmax is increased for 0.1 and 0.46 Vf

PGP composites compared with other volume fractions.
The peak height is also less for these composites (Table
8), which is clearly observed from Fig. 11b. The storage
modulus values of these compositions are also found to
be high compared with others. The shifting of Tg of neat
resin to higher temperature is associated with the
decreased mobility of the polymer chains due to the ber
matrix interaction.
Figures 12 and 13 show the temperature dependence of
loss modulus for the layered composites, GPG and PGP
of different volume fractions of GF. Loss modulus
showed similar trend as in the case of storage modulus
with variation of relative volume fraction of GF as in the
case of intimately mixed hybrid composites.
ColeCole Analysis
FIG. 11. (a) Variation of tan d with temperature of PGP hybrid composites at 10 Hz. (b) Variation of tan d with temperature of PGP hybrid
composites at10 Hz (excluding resin).
composite of GF content 0.2 Vf reduces the damping values over a wide temperature range.
The peak height of the GPG composites at 10 Hz is
given in Table 7. The peak height is less for 0.2 Vf showing better interfacial bonding. However for higher volume
fractions of GF (0.26 Vf), the peak height is more, which
is clear from Fig. 10b. It is already mentioned that 0.2 Vf
composite has higher initial storage modulus and comparatively higher storage modulus at all temperatures.
DOI 10.1002/pc
The changes in viscoelastic properties have been demonstrated by means of colecole method in the literature,
which had been developed for the study of dielectric dispersions.
It is used to examine the structural changes occurring
in cross linked polymers after ber addition to polymeric
matrices [24]. Figure 14 shows colecole plot where the
loss modulus E00 data are plotted as a function of storage
modulus E0 for PALF/GF/hybrid polyester composite(40%), but of different hybrid congurations showing
good dynamic mechanical behavior.
The nature of the colecole plots is reported to be indicative of the homogeneity of the system. Homogeneous
polymeric systems are reported to show semicircle diagrams [25]. In this study, we obtained imperfect circles
indicative of the heterogeneity of the system. However,
TABLE 8. The tan dmax (Tg) and tan d peak height of PGP hybrid
composites.
Volume
fraction of GF
0.10
0.20
0.26
0.46
tan
dmax (Tg) at
10 Hz (8C)
Peak height
of tan d curve
(10 Hz)
Activation
energy (kJ/mol)
162.6
157.7
158.8
160.6
0.130
0.131
0.136
0.131
115.8
242.2
203.1
223.6
FIG. 14. Colecole plot of PALF/GF hybrid composites.

FIG. 12. Variation of loss modulus with temperature of GPG hybrid
composites at 10 Hz.
Using Einstein equation, Ec Em (1 Vf), the storage

modulus of the composite is determined theoretically,
where Ec and Em are the storage modulus of the composite
and the matrix, respectively. The experimental and theoretical storage modulus of the hybrid composites intimately
mixed (IM), GPG and PGP of different volume fractions of
GF at 508C are given in Fig. 15. The experimental values
are higher than the theoretical values showing good rein-
forcing effect of bers in the glassy state of the matrix for

intimately mixed hybrid composites and at lower volume
fraction of GF for GPG and PGP hybrid composites. The
PGP hybrid composites showed good agreement with theoretical predictions compared with other hybrid composites.
The damping factor tan d values of the composite is
determined theoretically using Nielson equation tan dc
tan dm (1 2 Vf), where c and m refer to the composite
and the matrix, respectively. The experimental and the
theoretically calculated tan d values of the hybrid composites at 508C are shown in Fig. 16. The experimental
damping values are lower than the theoretically predicted
values at lower volume fraction for GF for intimately
mixed (IM) and GPG hybrid composites. This indicates
high-interfacial interaction. It is to be noted that IM and
GPG hybrid composites of lower volume fraction of GF
showed good mechanical as well as dynamic mechanical
FIG. 13. Variation of loss modulus with temperature of PGP composites at 10 Hz.
FIG. 15. Experimental and theoretical storage modulus of hybrid composites having different volume fractions of GF.
the shape of the curve shows relatively good ber/matrix

adhesion.
Theoretical Modeling
DOI 10.1002/pc
REFERENCES
FIG. 16. Theoretical modeling of tan d of hybrid composites having

different volume fractions of GF.
behavior. However, the PGP hybrid composites showed

higher damping values than the theoretically predicted
values at all volume fractions of GF.
CONCLUSION
Dynamic mechanical properties of PALF/GF hybrid
ber polyester composites have been studied with special
reference to hybrid-ber ratio and hybrid conguration. It
can be concluded that dynamic mechanical properties of
polyester are affected by adding PALF and GF. The
increase of GF content to 0.2 Vf enhances the material
stiffness and reduces the damping values for intimately
mixed and GPG hybrid composites. However, 0.26 Vf
GPG composite showed higher stiffness at higher temperatures. The stiffness of PGP composite is increased with
increase in GF loading. Dynamic mechanical analysis
showed that intimately mixed and GPG PALF/GF hybrid
polyester composites gave the highest storage modulus
values compared with PGP composites. The high-damping
parameter (tan d) of the resin was drastically reduced on
incorporation of PALF and GF. It is to be noted that considerable variation is not noticed among the different
types of hybrid composites with respect to tan d values.
Colecole plots show an imperfect semicircle showing the
heterogenity of the system and good interfacial adhesion
for the composites of different hybrid congurations
showing good dynamic mechanical behavior. Activation
energy calculated for intimately mixed and GPG showed
the highest values for composites of high-interfacial interaction. The higher effectiveness of reinforcement of IM
and GPG composites compared with PGP is also observed
from the activation energy values. Thus by hybridizing
PALF with small amounts of GF in polyester resin a
unique cost-effective composite possessing appropriate
stiffness and damping behavior could be prepared.
DOI 10.1002/pc
1. D. Ray, S. Sengupta, S.P. Sengupta, A.K. Mohanty, and M.

Misra, Macromol. Mat. Eng., 292, 1075 (2007).
2. A. Pietak, S. Korte, E. Tan, A. Downard, and M.P. Staiger,
Appl. Surf. Sci., 253, 3627 (2007).
3. S. Leduc, J.R.G. Urena, R. Gonzalez-Nunez, J.R. Quirarte,
B. Riedl, and D. Rodrigue, Polym. Polym. Comp., 16, 115
(2008).
4. S. Panthapulakkal and M. Sain, J. Comp. Mater., 41, 1871
(2007).
5. K. John and S.V. Naidu, J. Reinf. Plast. Comp., 26, 373
(2007).
6. E.M.F. Aquino, L.P.S. Sarmento, W. Oliveira, and R.V.
Silva, J. Reinf. Plast. Comp., 26, 219 (2007).
7. K.S. Ahmed and S. Vijayarangan, Ind. J. Eng. Mater. Sci.,
13, 435 (2006).
8. Abdullah-AL-Ka, M.Z. Abedin, M.D.H. Beg, K.L. Pickering, and M.A. Khan, J. Reinf. Plast. Comp., 25, 575 (2006).
9. G.C. Yang, H.M. Zeng, N.B. Jian, and J.J. Li, Plast. Indus.,
1, 79 (1996).
10. M.S. Sreekala, S. Thomas, and G. Groeninckx, Polym. Compos., 26, 388 (2005).
11. L.A. Pothan, P. Potschke, R. Habler, and S. Thomas,
J. Comp. Mater., 39, 1007 (2005).
12. M. Idicula, S.K. Malhotra, K. Joseph, and S. Thomas, Comput. Sci. Tech., 65, 1077 (2005).
13. H. Anuar, S.H. Ahmad, R. Rasid, A. Ahmad, and W.N.W.
Busu, J. Appl. Polym. Sci., 107, 4043 (2008).
14. L.U. Devi, S.S. Bhagawan, and S. Thomas, J. Appl. Polym.
Sci., 64, 1739 (1997).
15. J. George, S.S. Bhagawan, and S. Thomas, J. Therm. Anal.,
47, 1121 (1996).
16. L.U. Devi, S.S. Bhagawan, and S. Thomas, Mechanical
Characteristics of Pineapple Leaf/Glass Fibre Reinforced
Unsaturated Polyester Hybrid Composites, in Proceedings
of the ISAMPE National Conference on Composites
INCCOM-1V, December 2005, 329.
17. H. Eklind and F.H.J. Maurer, Polymer, 38, 1047 (1997).
18. H. Eklind, and F.H.J. Maurer, J. Polym. Sci. Part B: Phys.
Ed., 34, 1569 (1996).
19. L.A. Pothan, Z. Oommen, and S. Thomas, Comput. Sci.
Technol., 63, 283 (2003).
20. T. Murayamma, Dynamic Mechanical Analysis of Polymeric
Material, Amsterdam, Elsevier (1978).
21. G. Camino, M.P. Luda, A. Ya. Polishchuk, M. Revellino, R.
Blancon, G. Merle, and J.J. Martinez-Vega, Comput. Sci.
Technol., 57, 1469 (1997).
22. J.J. Arnold, M.P. Zamora, and A.B. Brennan, Polym. Compos., 17, 332 (1996).
23. M. Ashida, T. Noguchi, and S. Mashimo, J. Appl. Polym.
Sci., 30, 1011 (1985).
24. B. Harris, O.G. Bradella, D.P. Hamed, C. Lefevre, and
J. Verbert, J. Mater. Sci., 28, 3353 (1993).
25. J.D. Ferry, Viscoelastic Properties of Polymers, New York,
Wiley, 264 (1980).
Water Absorption Behavior of PALF/GF Hybrid

L. Uma Devi,1 S.S. Bhagawan,2 K.C. Manikandan Nair,3 S. Thomas3

1
St.Thomas College, Kozhencherry, Kerala 689641, India
2
Department of Polymer Engineering, Amrita Vishwa Vidyapeetham, Ettimadai, Coimbatore 641105, India
School of Chemical Sciences, Mahatma Gandhi University, Kottayam, Kerala 686560, India
The water absorption characteristics of pineapple leaf

ber (PALF)/glass ber (GF) hybrid polyester(PER)
composites, and chemically modied PALF/polyester
composites were evaluated by immersion in distilled
water at 28, 60, and 908C. The diffusion properties of
the intimately mixed (IM) and the layered hybrid composite GPG (Glass skin and PALF core) of different
PALF/GF ratio at the three temperatures were compared in order to identify the environmental ageing
mechanism at different temperatures. The effect
of temperature on the kinetics and thermodynamics of
diffusion were also examined. The water uptake of
both IM and GPG hybrid composites was decreased
with increase in glass ber content; the lowest water
uptake was observed for 0.46 Vf GF hybrid composite.
Among the chemically modied composites, vinyl tri
2-methoxy ethoxy silane treated composites showed
the lowest water uptake. Finally, parameters like diffusion, sorption, and permeability coefcients were
determined. It was observed that equilibrium water
uptake is dependent on the nature of the composite
and temperature. Experimental results were also compared with theoretical predictions. POLYM. COMPOS.,
32:335346, 2011. 2011 Society of Plastics Engineers
INTRODUCTION
Understanding water absorption in polymer composites
is important for many applications such as waste water
treatment, packaging and building industry. The way in
which ber reinforced composite materials absorb water
depends upon many factors, such as temperature, ber volume fraction, orientation of reinforcement, ber nature,
area of exposed surfaces, diffusivity, and surface protection.
Water absorption into composites occurs mainly by one
Correspondence to: L. Uma Devi; e-mail: umaharitham@yahoo.com
DOI 10.1002/pc.21034
Published online in Wiley Online Library (wileyonlinelibrary.com).
V
major mechanism of direct diffusion of water molecules

into matrix and to much lesser extent into the bers. The
other common mechanisms are capillarity and transport by
microcracks. The capillarity mechanism involves ow of
water molecules along the ber-matrix interface, followed
by diffusion from the interface into the bulk resin.
The important factors that affect water sorption of
composites are the hydrophilicity of the individual components and the structural arrangement of the bers within
the matrix [1]. It is well established that lignocellulosic
natural bers that are hydrophilic in nature are susceptible
to moisture absorption [2]. Water transport is strongly
affected by the nature of bermatrix interface. If the
interface is strong, it is difcult for the water molecules
to diffuse into the composite system. Upon various treatments of the ber and matrix resin, one can increase the
interfacial interaction, there by making the composite
more resistant to water transport. To achieve the full
potential of polymer composites, they must have good
environmental conditions. Environmental conditions such
as exposure to humid environment or change in temperature and exposure time cause dimensional instability and
loss of mechanical properties. Temperature inuences
moisture absorption in polymers and composites in a
complex manner. The temperature may change the fundamental diffusion mechanism in a material. The moisture
diffusion is coupled with relaxation process. A change in
temp will vary the relative rates of diffusion and relaxation, thus changing the overall absorption behavior.
Absorbed moisture has many detrimental effects on
material performance [3, 4]. Two possible mechanisms
have been suggested for the explanation of this effect
occurring in most composite systems; matrix plasticization or degradation of the ber-matrix interface [5, 6].
Water absorbed by epoxy matrix composites plays the
role of plasticizer, as evidenced by the reduction in the
matrix glass transition temperature, Tg [7, 8].
The moisture content of a plastic is very intimately

related to properties such as mechanical strength, appearance, dimensions, electrical insulation resistance, and
dielectric loss. The effect upon these properties of change
in moisture content due to water absorption depends
largely on the type of exposure and inherent properties of
the plastic. Synthetic ber reinforced composites impart
good long term behavior to various aggressive environments and an enhancement in strength and stiffness. Castaing and Lemoine [9] studied the effect of water sorption
and osmotic degradation on long-term behavior of glass
ber reinforced polyester. They found that plasticization
of the matrix with water remains the main consequence
of absorption and osmotic delamination of the plies may
happen only on long-term immersion. One of the major
disadvantages of thermoset resins is their tendency to
absorb water when exposed to humid environment.
Camino et al. [10] studied the kinetic aspects of water
sorption in polyester resin/glass ber composites and they
have developed a method that is based on the use of thermodynamic and kinetic parameters of water absorption/
desorption for the physio-chemical characterization of
sheet-moulding compounds.
Natural ber reinforced polyester materials are more or
less sensitive to humidity through absorption of water and
the differential swelling between the bers and the resin.
The uptake of water can also lead to chemical degradation
such as hydrolysis of the matrix. Consequently, the bermatrix interface is affected, resulting in changes of bulk
properties such as dimensional stability and both mechanical and electrical properties. The water absorption and
ageing studies of PALF/polyester(PER) composites carried out by us have been reported earlier [11].
According to Li and Luo [12] the physical mechanism
of moisture diffusion into highly hygroscopic bers such
as wool and cotton can be described by a two stage moisture diffusion process; a fast Fickian diffusion with a concentration diffusion coefcient and a slow diffusion with
a time dependent diffusion coefcient. Interfacial adhesion and resistance to moisture absorption of natural-ber
composites can be improved by treating these bers with
suitable coupling agents. Hybridization of natural ber
with stronger and more corrosion-resistant synthetic ber
like glass can also improve the stiffness, strength, and
moisture resistance of the composite. Recently there have
been quite a number of studies on hybridization of synthetic ber and natural lignocellulosic ber. Hybridization
also offset the disadvantage of one component by the
addition of another ber and a balance of properties could
be achieved [13]. Several studies have been reported concerning the water absorption in chemically treated natural
ber composites and natural ber/glass ber hybrid composites [1417]. Joseph et al. [18] have studied the water
sorption behavior of phenol-formaldehyde hybrid composites reinforced with banana ber and glass ber. The
water uptake of these composites was found to decrease,
and the composites with glass ber at the periphery and
banana ber at the core have the maximum resistance to

water absorption. Effect of hybridization and chemical
modication on the water-absorption behavior of banana
ber-reinforced polyester composites was studied by
Pothan and Thomas [19]. Resistance to water absorption
properties of the hemp ber composites was found
improved by hybridization with glass bers [20].
In the present study, the water absorption characteristics of high performance, cost-effective hybrid composites
prepared from high modulus glass ber (GF) and pineapple leaf ber (PALF) in polyester resin were studied. The
dependence of water sorption on GF content, layering pattern, chemical treatment, and temperature have been analyzed. The diffusion coefcient, sorption coefcient, permeability coefcient, and the kinetic parameters of water
diffusion of the composites were calculated. It is important to add that to the least of our knowledge, no detailed
studies have been reported on the water absorption of
hybrid composites of PALF and GF.
EXPERIMENTAL
Materials
PALF was provided by South India Textile Research
Association, Coimbatore, India. The ber is obtained
from the leaves of the plant Ananus cosomus belonging
to Bromeliaceae family. Neatly separated PALF, cut manually to their optimum length 30 mm [21] was used for
the study. Unsaturated isophthalic polyester HSR 8131
was obtained from M/S Bakelite Hylam, Hyderabad (Viscosity at 258C-650 cps, Sp. gravity at 258C-1.11, Tensile
strength 23 MPa, Impact strength 11 J m21). The multidirectional glass strand mat was supplied by Ceat (Hyderabad, India). Methyl ethyl ketone peroxide (MEKP) and
cobalt naphthanate were of commercial grade.Vinyl tri (2methoxy ethoxy silane) and c-methacryloxy propyl trimethoxy silane were obtained from Sigma-Aldrich, India.
Sodium hydroxide, glacial acetic acid, and polystyrene
maleic anhydride (PSMA) were of commercial grade.
Table 1 gives the characteristics of pineapple leaf ber
and glass ber used.
Chemical Modication of the Fiber

Silane Treatment. About 0.6% of the respective silane
was mixed with an ethanol/water mixture in the ratio 6:4,
mixed well and allowed to stand for 1 hr. The pH of the
solution was carefully controlled to bring about the complete hydrolysis of the silane by the addition of acetic
acid or NaOH. The chopped PALFs were dipped in the
above solution for one and half hour. The ethanol/water
mixture was drained out and the ber was dried in air followed by drying at 608c for 1 hr.
DOI 10.1002/pc
TABLE 1. Physical and mechanical properties of PALF and GF [22].

Properties
Density g cm23)
Diameter (lm)
Youngs modulus (GPa)
PALF
1.526
60 6 6
413 6 8
6.2
1.6
GF
2.54
15
2,500 6 8
5672
2.5
Polystyrene Maleic Anhydride (PSMA) Treatment. A

xed amount of cut ber was kept in 5% solution of
PSMA in toluene and reuxed for half an hour The ber
was then dried at 60708C in the oven.
Sodium Hydroxide(NaOH) Treatment. Chopped PALF
was treated with 2% solution of NaOH for 2 hr. The bers
were then washed with very dilute acid to remove
alkali. Washing was continued till the bers were free of
alkali. The washed bers were then dried in the oven at
60708C.
Composite Processing
Composite sheets were prepared using hand lay up
method. The chopped PALFs were uniformly spread in a
mold measuring 150 3 150 3 3 mm3 and made into a
mat. The mat was impregnated by pouring the polyester
resin to which 1 vol% cobalt naphthanate and 1 vol%
MEKP were added. The air bubbles were removed carefully with a roller. The closed mold was cured at room
temperature for 24 hr under constant pressure (1 MPa)
and post cured at 608C for further 24 hr. Composite
sheets were also prepared using chemically modied
bers in a similar manner.
Hybrid composites of two different congurations, intimately mixed (IM) and Glass/PALF/Glass (GPG) hybrid
composites (40 wt%) containing small volume fractions
of GF 0.1 and 0.46, PALF/GF(90/10 and 50/50), were
selected for the study. IM composites were prepared by
uniformly mixing the respective quantities of PALF (30
mm) and glass bers (50 mm). Randomly oriented GF
mats and PALF mat were arranged in the mold to prepare
trilayer GPG hybrid composites; the GF mats placed on
both sides and PALF mat in between.
specimens of the same material were tested. The mole

percent uptake, Qt of water by 100 g of polymer was calculated using the equation,
Qt Mw =Mrw =Mis 3100
(1)
Where M(w) is the mass of water at time t, Mr (w) the relative molecular mass of water, 18, and Mi(s) is the initial
mass of the sample. When equilibrium was reached, Qt
was taken as the mole percent uptake at innite time Q!.

Water Uptake of PALF/GF Hybrid Polyester Composites
Changes in the water sorption characteristics of PALF/
PER composites upon hybridizing with GF at different
temperatures were studied. Sorption curves of IM and layered GPG hybrid composites having different volume fractions of GF at 28, 60, and 908C are shown in the Figs. 1
3. It is observed that the neat polyester resin shows very
low water absorption at all temperatures. The observed
absorption of water in this case could have occurred
through the micro cracks present inside the material.
All the composite samples absorbed water rapidly at
the initial stage and later a saturation level was attained
without any further increase in water sorption. The water
absorption curves show a multistage mechanism. The initial portion of the moisture absorption curve is linear,
after which the mechanism changes. Water diffusion in
polymers was found to lead to typical phenomena of
Sorption Experiments
Samples of approximate dimensions 10 3 10 3 3
mm3 were used for the measurement of water absorption.
The corners of the samples were curved to avoid nonuniform water diffusion. The thickness and weight of the
sample were measured. Samples were immersed in water
at different temperatures 28, 60, and 908C in a thermostatically controlled air oven. The specimens were periodically taken out of water, surface dried with absorbent
paper and reweighed. This process was continued till
equilibrium was reached. At each temperature, three
DOI 10.1002/pc
FIG. 1. Molar uptake of PALF/GF hybrid polyester composites having

different GF content as a function of immersion time at 288C.

composite swelling and physical relaxation. According to

Florys two-stage theory the swelled polymer chains
induce increased elasticity of chain structure and thus
increased chemical potential. The increased chemical
potential inhibits further absorption of water, which may
be observed as the rst equilibrium of water uptake. However, the swelled polymer chains start relaxing with time
and subsequently reduce chemical potential. Consequently,
the second equilibrium is attained by the decreased elasticity of the polymer chains. In the hybrid composites, in
addition to the above phenomena, the change in mechanism can also be attributed to the delamination occurring
in the composites. Penetration of water into the matrix
causes absorption of water by the bers as well.
The rate of absorption of water in the hybrid is different for the two bers. Glass bers principally consist of
silicates of various metals. When immersed in water, the
nucleophilic attack of the OH2 at silicon takes place, with
the formation of a transition metal complex followed by
the breakage of the bonds, and formation of new bonds. In
the pineapple bers, hydrogen bonding through the OH
of the glucose is the principal water-absorption method.
The absorption of water causes delamination of the two
layers of the ber attributed to the differential swelling
especially in GPG hybrid. This causes further absorption
of water into the free voids created by delamination. The
matrix resin also absorbs water. The terminal
OH of the
polyester, oxygen of the ester linkage, or residual cobalt
ions are all sites for formation of the hydrogen bond.
From the sorption curves, it is evident that a rapid initial uptake occurs in the unhybridised composite and both
IM and layered hybrid composites of lower GF content.

This is more predominant at higher temperature of 908C.
However, all the hybrid composites showed lower water
uptake than the neat PALF/PER composite except IM
composite of lower (0.1) GF content. The moisture uptake
for both hybrid systems IM and layered GPG decreases to
a greater extent with increase of glass ber content. It is
because the moisture uptake of glass ber is very less.
The IM and GPG hybrid composite of 0.46 Vf of GF
showed minimum water uptake. As a result, moisture
absorption during ageing can be reduced by replacing
PALF with GF in the composite. Thus the durability of
natural ber composite can be improved by hybridizing
with GF.
The saturation level for PALF/PER composite is much
higher than for the hybrid composites at higher temperatures, especially at 908C. The water absorption of the
PALF/GF hybrid composites of higher GF loading is
much lower than that of neat PALF/PER composite. The
IM hybrid composite of 0.46 Vf exhibits lower water
uptake at high temperatures 60 and 908C compared to the
layered GPG hybrid composites. As the glass ber loading increases, the composite surface becomes more and
more hydrophobic and water can penetrate only through
the interlayers in the composite leading to reduced water
absorption. Similar decrease of water absorption is
observed in the banana/glass hybrid polyester composites
[19] and bamboo glass ber reinforced polypropylene
hybrid composites [23].The decrease in moisture content
for samples with high glass ber content can be attributed
to the role of the glass bers which serve as a barrier for
preventing water diffusion.

DOI 10.1002/pc
increases, the rate of diffusion of the composites also

increases. Diffusion (D) is related to the velocity of the
diffusing molecules by the equation given below.
D
FIG. 4. Variation of molar uptake with root time of PALF/PER composite (40 wt%) at different temperatures.
Effect of temperature on sorption can be understood

from the respective sorption curves. The effect of temperature on the diffusion process of PALF/PER, IM, and
GPG layered hybrid composites is shown in Figs. 46.
It is observed that the initial rate of water sorption and
the equilibrium rate of water uptake of PALF/PER
increases with increasing temperature. As temperature
FIG. 5. Variation of molar uptake with root time of IM hybrid composite (0.46 Vf) at different temperatures.
DOI 10.1002/pc
1
kc
3
(2)
where c mean velocity of molecules, k mean free

path (distance traveled by molecules between two consecutive collisions). Since the mean velocity increases with
temperature, the diffusion also increases with temperature.
Maximum absorption is observed at 908C for the neat
PALF/PER. However, a decrease in absorption is
observed for the hybrid composites with higher GF content at 908C. This is observed from the Figs. 5 and 6 and
the Q! values of the composites given in Table 2. The
effect of GF loading on the Q! values for the composites
at different temperatures is evident from Table 2. The
neat polyester shows low water uptake. For the neat
PALF/PER composite, the rise of temperature leads to
enhanced water uptake at equilibrium. The Q! values of
both IM and GPG hybrid composites are found to be
highly decreased at all temperatures with increase in relative volume fraction of glass ber. IM and GPG hybrid
composites of 0.46 Vf show very low water uptake compared to 0.1 Vf hybrid composites. A decrease in Q! is
observed at 908C for all the hybrid composites except
for IM with 0.46 Vf. The decrease in Q! at higher
temperature of 908C shows that the temperature has a
strong effect on the ber/matrix interaction in these com-
FIG. 6. Variation of molar uptake with root time of GPG hybrid composite (0.46 Vf) at different temperatures.
TABLE 2. Effect of hybrid conguration and GF loading on the Q!,

n, and k values of PALF/GF hybrid composites at different temperatures.
Sample
Neat polyester
PALF/PER
IM (0.1 Vf)
IM (0.46 Vf)
GPG (0.1 Vf)
GPG (0.46 Vf)
Temperature (8C) Q! (mole g21)%

28
60
90
28
60
90
28
60
90
28
60
90
28
60
90
28
60
90
0.11
0.14
0.15
0.75
1.04
1.03
0.70
0.79
0.63
0.33
0.28
0.31
0.85
0.86
0.82
0.32
0.53
0.37
k (g/g/min2)
0.32
0.28
0.25
0.28
0.28
0.24
0.26
0.28
0.23
0.24
0.30
0.25
0.26
0.39
0.28
0.26
0.25
0.36
21.57
21.43
21.1
21.10
21.0
20.90
21.20
21.10
20.70
20.79
20.79
20.65
21.10
21.10
20.86
21.10
20.79
20.86
posites. This is associated with the changes that occur in

interface at high temperatures in composites. At high temperatures, IM composites show comparatively lower sorption. This is also lower than the GPG hybrid. Similar
decrease in Q! values is observed at 908C for the banana
glass hybrid ber phenol-formaldehyde composites [18].
The water uptake decreases with increasing temperature
from 50 to 708C for the banana glass hybrid ber polyes-
FIG. 8. Effect of PALF treatment on mole% uptake of water for

PALF/PER composites as a function of immersion time at a temperature
of 608C.
ter composites and a reverse trend is noticed at 908C

[19]. The temperature increase to 708C promotes
improved molecular motion. The increased water uptake
at higher temperature can be attributed to the relaxation
mechanism in the polymer whereby the polymer network
is distorted. This inevitably leads to large-scale segmental
motion, allowing further penetration of water. The reason
for the slight higher water uptake at 708C for the PALF/
GF hybrid composites can also be attributed to the combined effect of the increased water uptake in the voids
created by the degradation of the resin, as well as to resin
cracking and ber debonding. At 908C, leaching of the
material leads to chemical degradation, which results in a
slight lowering of the composite weight.
Effect of Chemical Treatment
FIG. 7. Effect of treatment on mole% uptake of water for PALF/PER

composites as a function of immersion time at a temperature of 288C.
The water absorption behavior of chemically modied

PALF/PER composites at 28, 60, and 908C are given in
Figs. 79. It is observed from the gures that the chemically treated PALF/PER composites show a decreased
rate of water absorption compared to the unmodied
PALF/PER composites at all temperatures. The initial
uptake of water due to capillary action is also reduced.
The increased water uptake for the untreated composite
can be attributed to the increased capillary action. The
treated composites showed a decrease in water absorption
which is due to the improved ber/matrix adhesion. In
chemically-treated bers, rearrangement of cellulose
brils leads to free spaces for the matrix resin to squeeze
in and lesser space for the water molecules. As the free
hydroxyl groups of cellulose and lignin become chemiDOI 10.1002/pc
FIG. 11. FTIR spectrum of NaOH treated PALF.
FIG. 9. Effect of PALF treatment on mole% uptake of water for

PALF/PER composites as a function of immersion time at a temperature
of 908C.
cally bonded in treated bers, the hydrophilic character of

the ber is reduced which favors a strong interfacial adhesion between PALF and the matrix. The coupling agents
form chemical bonds and hydrogen bonds and this
reduces the ber/matrix debonding due to moisture. In
addition chemical modication covers some of the surface
pores as well in the ber. Shrinkage of the micropores,
collapse of capillaries upon chemical treatment may also
block the capillary absorption.
The water absorption is highly decreased for the silane
composites and NaOH-treated composites. The slope
change after the initial fast absorption can be attributed to
the over all change in the absorption mechanism. Mercer-
FIG. 10. FTIR spectrum of untreated PALF.
DOI 10.1002/pc
ization results in the leaching out of the amorphous waxy

cuticle layer which reduces the water sorption capacity of
the ber. The lower water uptake of the alkali treated ber
composite is due to the improved ber/matrix adhesion.
The composites containing the NaOH treated bers exhibit
the lowest water sorption. The water sorption is found to
follow in the order Untreated [ PSMA treated [ Vinyl silane treated [ methacrylate silane treated [ NaOH treated
composites. However vinyl silane treated composite shows
decreased water sorption than methacrylate silane treated
composite at 908C. In the vinyl silane treated composite,
the adhesive bonding is also not deteriorated upon heating.
The decreasing absorption of the methacrylate silane
treated composite is due to the better adhesion between the
methacrylate silane treated bers and the resin. The water
uptake of the PSMA treated ber composite is comparatively higher than the other treated composites. This may
be attributed to the interactions between water molecules
and maleic anhydride groups present in PSMA.
The FTIR spectrum of untreated and NaOH-treated
PALF are given in Figs. 10 and 11. In the FTIR spectrum
of NaOH treated ber, the relative intensity of OH peak
decreases indicating the participation of some free
hydroxyl groups. The structure of cellulose is very complex undergoing lattice transformation into various polymorphic forms under NaOH treatment and this becomes
more complex due to the presence of lignin and hemicellulose. The IR band around 2,923 cm21 is assigned to
CH2 anti-symmetric stretching [24]. In the 2% NaOH
treated ber, the band shifts to lower frequency side at
2,914 cm21. The total intensity of the band which can be
measured by the absorption band area is slightly decreased
with mercerization. The decrease of intensity of the band
with NaOH treatment concludes that carbon atoms
attached to carbon or hydrogen (CC or CH)
decrease. The decrease of lignin content on the surface is
clear from this. The IR band at 1,431 cm21 in the
untreated PALF which is assigned to the CH2 symmetrical
logQt =Q1 log k n log t
FIG. 12. TG curves of untreated and NaOH-treated ber.
deformation [25, 26] shifts to low frequency 1,429 cm21

with out any signicant change in the intensity of band.
The effect of change of this band is related to the change
in the environment of the C6 carbon atom due to the formation of breaking of hydrogen bond involving the atom
O6. This indicates the change in lattice from cellulose 1
into cellulose 11. The lattice transformation from cellulose
1 to cellulose 11 is restricted due to the presence of lignin
[27]. In the case of lattice transformation, the intensity of
the band should change sharply, but that is not taking
place here. Thus the crystal conversion of cellulose 1 in to
cellulose 11 is only partial.
The slight improvement in thermal stability of NaOHtreated ber compared to untreated is observed from the
TG curves (see Fig. 12). Therefore it can be noticed that
the modied pineapple leaf bers displayed a higher thermal stability as evidenced by the increase in the temperature of degradation. The percentage weight loss at 1008C
is lower for the NaOH-treated ber compared to raw ber
which indicates lower moisture content (weight loss at
1008C for raw PALF, 9.8 and NaOH-treated PALF, 8.3).
This is due to the presence of hydrophobic groups present
on the ber surface after treatment. These results showed
that alkalization modies plant bers promoting the development of ber-resin adhesion which then will result
in improved water resistance of the NaOH treated PALF/
PER composite.
where Qt is the mole percent increase in water uptake at

time t, Q!, the mole percent water uptake at equilibrium,
t the time, and k and n are constants. The constant k indicates the interaction between polymer and water and the
value of n provides information about the mechanism of
solvent transport under all experimental conditions. When
n 1/2, diffusion obeys Ficks law and the mechanism is
said to be Fickian [28]. This occurs when the segmental
mobility of the polymer chains is faster than the rate of
diffusion of permeant molecules. Ficks law is normally
applied for the diffusion of water in a homogeneous material and where there is no chemical interaction between
the absorbed water and the material.
Effect of hybridization of pineapple leaf ber with glass
ber on the kinetic parameters was analyzed. The values
of n and k of the hybrid composites and chemically modied composites determined by linear regression analysis
are given in Tables 2 and 3. For all the hybrid composites
and chemically-treated PALF/PER composites, n is less
than 0.5. Thus, a deviation from Fickian behavior is
observed in all cases. k values are found to increase with
increase of temperature for all the composites.
Diffusion, Sorption, and Permeation

The values of diffusion coefcient were calculated
from the relationship
D phh=4Q1 2
TABLE 3. Effect of surface modication of n and k values of the

treated PALF/PER composites at different temperatures.
Fiber treatment
Untreated
Vinyl silane
Methacrylate silane
Kinetics of Water Absorption
(4)
where h is the slope of initial linear portion of the sorption curves and h is the thickness of the sample. Q! is
the mole % at equilibrium. The values of diffusion coefcient D of the hybrid composites and chemically modied PALF/PER composites are given in Tables 4 and 5.
The diffusion coefcient of the hybrid composites
NaOH
To follow the mechanism of sorption, the kinetic parameters n and k were calculated from the following
equation
(3)
PSMA
Temperature (8C)
k (g/g/min2)
28
60
90
28
60
90
28
60
90
28
60
90
28
60
90
0.28
0.28
0.24
0.30
0.40
0.30
0.22
0.39
0.24
0.19
0.35
0.32
0.28
0.38
0.31
21.10
21.0
20.90
21.10
21.30
21.03
20.79
20.99
20.63
20.77
20.76
20.55
20.84
20.87
20.98
DOI 10.1002/pc
TABLE 4. Transport properties of PALF/PER and hybrid composites

(D, diffusion coefcient, S and P, sorption and permeation coefcient).
Samples
Temperature
(8C)
Neat resin
PALF/PER
IM (0.1 Vf)
IM (0.46 Vf)
GPG (0.1 Vf)
GPG (0.46 Vf)
28
60
90
28
60
90
28
60
90
28
60
90
28
60
90
28
60
90
D (31029) 6
S.D (31029)
(cm2 s21)
0.47
0.54
0.40
0.74
2.17
0.21
0.53
0.88
0.33
0.74
3.60
0.21
0.34
1.53
0.21
0.13
0.67
0.14
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
P (310210)
(cm2 s21)
S (g g )
0.01
0.02
0.05
0.01
0.11
0.004
0.07
0.04
0.09
Nil
1.9
0.06
0.04
0.09
0.04
0.03
0.05
0.03
21
0.11
0.10
0.08
0.13
0.14
0.06
0.15
0.15
0.03
0.06
0.05
0.04
0.15
0.15
0.04
0.05
0.08
0.04
S M1 =Mp
0.51
0.54
0.32
0.99
3.09
0.13
0.79
1.32
0.09
0.45
1.80
0.08
0.51
2.29
0.08
0.06
0.53
0.05
P D:S
(5)
in which Ea is the activation energy of diffusion, D0 the

preexponential factor and R is the gas constant. The exponential correlation of diffusivity with 1/T results in a very
strong temperature dependence.
DOI 10.1002/pc
(6)
where M! is the mass of the solvent taken up at equilibrium and Mp is the initial mass of polymer sample. Solubility is a thermodynamic parameter which depends on
the strength of interaction in the polymer-penetrant mixture. From the S values of the composites for IM and
GPG hybrid composites (Table 6), the sorption coefcient
decreases with increase of GF loading. For a particular
sample the sorption coefcient decreases with increase of
temperature.
Permeability coefcient, which could be considered as
the total effect of sorption and diffusion, can be estimated
using the equation [23].
changes with increase in volume fraction of GF. This is

evident in both the patterns IM and GPG hybrid composites. This is due to the lower hydrophilicity of GF than
PALF. The D value increases up to 608C and further
increase of temperature decreases the diffusion coefcient. A greater decrease is observed at 908C. This is
true for the composites of varying GF concentration.
The diffusion coefcient values are lowest for GPG
hybrid composites of 0.46 GF content at all temperatures. This is due to the lower hydrophilicity of GF than
PALF which forms the surface layer. Diffusion, which is
a thermally activated process, is very sensitive to temperature. It is a general phenomenon that low diffusion
coefcient of materials in polymers shows considerable
dependence on any change of condition of concentration,
temperature etc.
The diffusion coefcient D which characterizes the
ability of water molecules to diffuse into the ber is
decreased at higher temperature. At higher temperatures,
the hydrogen bond formation tends to disappear and the
molecular packing controls the diffusion in bers. The
absorptiondiffusion mechanism in this case will be slow
as the entrance of water molecules into the crystalline
region of the ber is difcult. This means that the diffusion coefcient of the composites is lower at higher temperature (908C). Diffusion coefcient obeys the activated
transition theory and its temperature dependence can be
expressed by the Arrhenius equation [4, 29].
D D0 exp Ea =RT
The permeability of small molecule into a polymer is

dependent on diffusivity as well as the sorption or solubility of a liquid in the polymer. The sorption coefcient (S)
or solubility can be calculated from equilibrium swelling
using the relation [30].
(7)
where D is the diffusion coefcient and S, the solubility.

On increasing the temperature to 608C, P increases and a
decrease is observed thereafter.
The treated composites show lower diffusion coefcient values at 28 and 908C; however an increase is
noticed at 608C. The sorption and permeability coefcient
of the treated composites behave in a similar manner as
the hybrid composites with temperature. The sorption
coefcient decreases with respect to temperature whereas
the permeability coefcient varies in a similar manner as
the diffusion coefcient. Among the treated composites
vinyl silane-treated composites showed minimum sorption
which is in agreement with the good mechanical properties and better interfacial adhesion.
TABLE 5. Effect of chemical modication of bers on D, S, and P

values of PALF/PER composites at different temperatures.
Samples
Temperature
(8C)
PALF/PER
Vinyl silane
Methacrylatesilane
NaOH
PSMA
28
60
90
28
60
90
28
60
90
28
60
90
28
60
90
D (3109) 6
S.D (31029)
(cm2 s21)
S (g g )
P (310210)
(cm2 s21)
6
6
6
6
6
6
6
6
6
6
6
6
6
6
6
0.13
0.14
0.06
0.16
0.15
0.06
0.15
0.13
0.05
0.13
0.13
0.05
0.15
0.15
0.21
0.99
3.09
0.13
0.74
4.95
0.06
0.74
4.81
0.05
0.40
2.86
0.05
1.02
5.10
0.44
0.74
2.14
0.21
0.46
3.30
0.10
0.49
3.70
0.10
0.29
2.20
0.10
0.68
3.40
0.21
0.01
0.11
0.004
0.02
0.28
0.07
0.04
0.14
0.03
0.02
0.40
0.01
0.22
Nil
0.007
21
TABLE 6. Activation energy and thermodynamic parameters of PALF/

GF hybrid ber polyester composites (40 wt%).
Samples
PALF/PER
IM hybrid
GF (0.1 Vf)
GF (0.46 Vf)
GPG hybrid
GF (0.1 Vf)
GF (0.46 Vf)
Activation energy
(kJ mole21)
DH
(kJ mole21)
2DS
(J mole21)
42.5
26.0
61.3
20.0
62.5
28.8
5.4
67.3
10.7
59.4
64.8
27.2
216.0
14.4
101.1
Thermodynamic Parameters
The thermodynamic parameter for diffusion, viz the
enthalpy and entropy of sorption DH and DS can be calculated using vant Hoffs relation
ln Ks DS=R DHs =RT
(8)
where Ks number of moles of solvent sorbed at equilibrium/mass of the polymer. The values of DS and DHs of
the chemically treated and PALF/GF hybrid composites
obtained by regression analysis of the plots of log Ks vs.
1/T are given in Tables 6 and 7.
Activation energy of diffusion is high for all the hybrid
composites except IM composites of low GF content (0.1 Vf)
and it lies in the range 2065 kJ mol21 (Table 6). As glass
ber loading increases, activation energy increases. This
shows that diffusion of water is hindered by the presence of
glass bers. GPG hybrid composites where glass bers are
on both surface show comparatively higher activation energies. DS is negative in all cases and DH is also negative for
all the composites except IM hybrid of higher GF content.
The activation energy values of the chemically treated
composites are given in Table 7. It is found that the chemically treated composites have higher activation energies
than the untreated composites. This is due to the presence
of a strong interface which hinders the entry of water. The
water molecule can easily form bonds with maleic anhydride groups present in PSMA and so the energy needed for
this process is low compared to other treated composites.
FIG. 13. Experimental and tted diffusion curve for PALF/PER composite at 908C.
and one less-dense phase. According to them the diffusion

proceeds by parallel Fickian processes in both phases [31,
32]. The variation of normalized water content Q(t) with
time can be described by the morphology-depended Eq. 9
(
"
0:75 #)
Dd t
Qt Vd 1 exp 7:3 2
b
(

0:75 )
Dl t
1 Vd 1 exp 7:3 2
(9)
b
Theoretical Modeling
Jacob-Jones Model. This model describes the diffusion
processes in two-phase materials consisting of one dense
TABLE 7. Activation energy and thermodynamic parameters of treated
PALF/PER composites.
Fiber
treatment
Neat PALF/PER
Vinyl silane
Methacrylate silane
NaOH
PSMA
Activation energy
(kJ mole21)
DH
(kJ mole21)
2DS
(J mole21)
42.5
77.9
79.9
80.1
63.6
26.0
24.6
29.9
5.2
8.8
61.3
54.5
71.8
23.7
10.9
FIG. 14. Experimental and tted diffusion curve for IM (0.46 Vf) composite at 908C.
DOI 10.1002/pc
where Q(t) Qt/Qa, Dd and Dl are the diffusion coefcients of less dense and the dense phase respectively, and
Vd is the volume fraction of the dense phase. In the case
of PALF/PER and PALF/GF hybrid ber composites, the
material can be considered as two phase system [33] with
polyester matrix as one phase (Phase 1) and the ber as
second phase (Phase 2). The diffusion of water in the matrix phase is mainly attributed to thevoids present in the
composite material. In the second phase (ber) the diffusion takes place through the formation hydrogen bonds
between the hydroxyl groups of ber and water molecules. In this case the Jacobs-Jones model can be modied to Eq. 10, as follows:
(
"
0:75 #)
D1 t
Qt Q1 1 exp 7:3 2
b
(

0:75 )
D2 t
Q2 1 exp 7:3 2
10
b
where D1 and D2 and Q(1) and Q(2) are the diffusion coefcient and the normalized water content of absorbed
water at saturation in the two phases respectively. Since
Q(2) 1 2 Q(1) Eq. 10 can be modied as
(
"
0:75 #)
D1 t
Qt Q1 1 exp 7:3 2
b
(

0:75 )
D2 t
1 Q1 1 exp 7:3 2
10
b
The experimental results of water diffusion in PALF/PER
and PALF/GF composites are tted by this equation by
nonlinear least square t method and Figs. 13 and 14
shows typical experimental and tted diffusion curves of
40% PALF/PER composite and PALF/GF hybrid composite at 908C. The tted values of Q1, Q2, D1, and D2 are
listed in Tables 8 and 9. Theoretical prediction shows
good agreement with the experimental values. This conrms the existence of two different phases and mode of
diffusion in PALF composites. The slight deviation from
the experimental curve results from the deviation of diffusion processes from Fickian behavior. The deviation from
the Fickian behavior is clear from the values of n given
in Table 2.
TABLE 8. Fitting values as per Eq. 10 for PALF/PER composite.

PALF/PER
composite
temperature (8C)
28
60
90
DOI 10.1002/pc
Q1
log D1
logD2
Q2
Qa
0.167828
0.23193
0.123337
212.6696
212.2426
212.0054
212.3846
212.2523
212.0126
0.5791
0.8113
0.9159
0.7469
1.0433
1.0393
TABLE 9. Fitting values as per Eq. 10 for PALF/GF hybrid PER

composite.
Sample
temperature (8C)
Q1
log D1
logD2
Q2
Qa
GPG (0.1 Vf) 28

GPG (0.1 Vf) 60
GPG (0.1 Vf) 90
GPG (0.46 Vf) 28
GPG (0.46 Vf) 60
GPG (0.46 Vf) 90
IM (0.1 Vf) 28
IM (0.1 Vf) 60
IM (0.1 Vf) 90
IM (0.46 Vf) 28
IM (0.46 Vf) 60
IM (0.46 Vf) 90
0.0442
0.5649
0.1160
0.1070
0.4204
0.0355
0.1876
0.1404
0.3339
0.0160
0.0002
0.1687
212.5421
212.3633
212.0702
212.7414
212.3896
213.0316
212.701
212.5531
211.9487
212.3141
212.3118
211.7535
212.541
212.3629
212.1161
212.7219
212.3864
211.9952
212.5542
0.4344
211.9806
212.3123
212.3118
212.0135
0.8096
0.2981
0.7049
0.2156
0.1841
0.3362
0.5212
0.6569
0.2970
0.3179
0.2792
0.1469
0.8538
0.8631
0.8210
0.3227
0.6045
0.3718
0.7089
0.7974
0.6310
0.3339
0.2794
0.3157
CONCLUSION
The water absorption kinetics of PALF/GF hybrid ber
reinforced polyester composites were analyzed at different
temperatures (28, 60, and 908C). The effect of chemical
modication of PALF, hybrid pattern and GF loading on
the water absorption of the composites were evaluated.
The changes in temperature were found to alter the diffusion mechanism during water absorption. IM and GPG
hybrid composites of higher GF content showed minimum
water uptake at all temperatures. The NaOH and vinyl silane treated composites showed lower water uptake among
the chemically modied composites. The diffusion coefcient and permeation coefcient of all the composites
were low at 28 and 908C and the solubility coefcient
decreased with increase of temperature. Modied JacobsJones Model was used to study the diffusion process in
PALF/PER and PALF/GF hybrid polyester composites
and showed good agreement between experimental values
and theoretical predictions. The model prediction also
conrms the existence of two different phases and modes
of diffusion in the composites.
REFERENCES
1. M.S. Sreekala, M.G. Kumaran, and S. Thomas, Comp. Part
A, 33, 763 (2002).
2. A.W. Mckenzie and I.P. Yuritha, Appita, 32, 460 (1979).
3. J.P. Komorowskj, National Aeronautical Establishment,
National Research Council of Canada, Aeronautical Note
NAEAN-10,NRC No22700 (1983).
4. Y. Weitsman, Comprehensive Composite Materials, in:
Polymer Matrix Composites, Vol. 2, R. Talreja and L.A.E.
Manson, Eds., Elsevier: New York (2000).
5. C.W. Browning, Polym. Eng. Sci., 18, 16 (1978).
6. C. Shen and G.S. Springer, J. Comp. Mater., 11, 2 (1977).
7. L. Banks and B. Hills, Polym. Bull., 1, 337 (1979).
8. W.W. Wright, Composites, 12, 201 (1981).
9. P.H. Castaing and L. Lemoine, Polym. Comp., 16, 349
(1995).
10. G. Camino, M.P. Luda, A.Y Polishchuk, M. Revellino, R.

Blancon, G. Merle, and J.J. Martinez-Vega, Comp. Sci.
Technol., 57, 1469 (1997).
11. L.U. Devi, K. Joseph, K.C.M. Nair, and S. Thomas, J. Appl.
Polym. Sci., 94, 503 (2004).
12. Y. Li and Z.X. Luo, J. Text. Inst., 91, 1 (2000).
13. B. Harris, Engineering Composite Materials, Institute of
Metals, London (1986).
14. S. Alix, E. Philippe, A. Bessadok, L. Lebrun, C. Morvan,
and S. Marais, Bioresource. Technol., 100, 4742 (2009).
15. C. Baley, F. Busnel, Y. Grohens, and O. Sire, Comp. Part
A, 37, 1626 (2006).
16. G. Cicala, G. Cristaldi, G. Recca, G. Ziegmann, A. El-Sabbagh,
and M. Dickert, Mater. Des., 30, 2538 (2009).
17. K. Jarukumjorn and N. Suppakarn, Comp. Part B. Eng., 40,
623 (2009).
18. S. Joseph, M.S. Sreekala, P. Koshy, and S. Thomas, J. Appl.
Polym. Sci., 109, 1439 (2008).
19. L.A. Pothan and S. Thomas, J. Appl. Polym. Sci., 91, 3856 (2004).
20. S. Panthapulakkal and M. Sain, J. Appl. Polym. Sci., 103,
2432 (2006).
Sci. 64, 1739 (1997).
22. M. Idicula, A. Boudene, L.U. Devi, L. Ibos, Y. Candau, and

S. Thomas, Comp. Sci. Technol., 66, 2719 (2006).
23. M.M. Thwe and K. Liao, Compo. Part A., 33 43 (2002).
24. J. Blackwell, P.D. Vasko, and J.L. Koening, J. Appl. Phys.,
41, 4375 (1970).
25. C.Y. Liang and R.H. Marshessault, J. Polym. Sci., 43, 71
(1960).
26. C.Y. Liang and R.H. Marshessault, J. Polym. Sci., 39, 269
(1959).
27. J.F. Revol and D.A. Goring, J. Appl. Polym. Sci., 26, 1275
(1981).
28. L.M. Luchit and N.A. Peppas, J. Appl. Polym. Sci., 33, 1557
(1987).
29. J.P. Komorowskj, National Aeronautical Establishment.
National Research Council of Canada, Aeronautical Note
NAE-AN-4, NRC No. 20974 (1983).
30. S.B. Harogoppad, T.M. Aminabhavi, J. Appl. Polym. Sci.,
42, 2329 (1991).
31. P.M. Jacobs and F.R. Jones, J. Mater. Sci., 24, 2331 (1989).
32. P.M. Jacobs and F.R. Jones, J. Mater. Sci., 24, 2343 (1989).
33. C. Maggna and P. Pissis, J. Polym. Sci. Part B Polym.
Phys., 37, 1182 (1999).
DOI 10.1002/pc
Dynamic Mechanical Properties of Pineapple Leaf Fiber

L. Uma Devi,1 S.S. Bhagawan,2 S. Thomas1

1
School of Chemical Sciences, Mahatma Gandhi University, Priyadarshini Hills, Kottayam 686560, Kerala, India
2
Department of Polymer Engineering, Amrita Vishwa Vidyapeetham, Ettimadai, Coimbatore 641105,

Tamil Nadu, India
In the recent years, lignocellulosic bers that originate

from a renewable source have been found to provide
good reinforcement in polymer matrices. Among the
natural bers, pineapple leaf ber (PALF) exhibits
excellent mechanical properties, besides possessing
low density, high stiffness, and low cost. The dynamic
mechanical properties, storage modulus (E0 ), and loss
tangent of PALF-reinforced polyester (PER) composites
were evaluated at three frequencies 0.1, 1, and 10 Hz
and temperatures ranging from 30 to 2008C. Addition
of PALF of 30 mm length (aspect ratio 600) was found
to increase the storage modulus leading to a maximum
value at 40 wt%. The glass transition temperature (Tg)
of the composite of 40 wt% showed a positive shift
indicating high polymer/ber interaction. A new relaxation is observed at 40 wt% showing the presence of a
strong interphase at all aspect ratios. SEM photographs of fracture surfaces of composites conrm the
results obtained from static and dynamic mechanical
analysis. POLYM. COMPOS., 32:17411750, 2011. 2011
Society of Plastics Engineers
INTRODUCTION
Incorporation of renewable natural bers like sisal, banana, ax, and pineapple in polymeric materials helps to
obtain new materials of high performance. Their advantages include low density, exibility, low cost, and highspecic mechanical properties, besides their environment
friendly character and are replacing traditional materials at
a fast pace. They are found to provide effective reinforcement in various thermosets and thermoplastic matrices [1
6]. Among these, pineapple leaf ber (PALF) exhibits
excellent mechanical properties. These bers have high cellulose content (7082%) and comparatively low microbrilar angle (148). Our earlier studies have proved PALF to
Correspondence to: L. Uma Devi; e-mail: umaharitham@yahoo.com
DOI 10.1002/pc.21197
Published online in Wiley Online Library (wileyonlinelibrary.com).
V
be an effective reinforcement in polyester (PER) matrix

[7]. The water absorption and aging studies of PALF/PER
composites have also been reported by us [8]. Thomas
et al. [9] also studied the thermogravimetric and dynamic
mechanical thermal analysis of PALF-low density polythene composites. The natural ber composites are cost
effective and can be used in different applications where
static and dynamic loading place a key role.
Dynamic mechanical analysis (DMA) over a wide
range of temperatures is quite useful in studying the polymer composite structure [1013]. Polymers are usually
subjected to cyclic deformation, and therefore DMA is
very important as these tests measure the response of a
material to sinusoidal or other periodic stress. The DMA
is thus useful in studying composite structure and performance. Mohanty and Nayak [14] studied the mechanical and dynamic mechanical behavior of sisal ber-reinforced high-density polyethylene composites. Bledzki and
Zhang [15] investigated the dynamic mechanical behavior
of jute ber-reinforced epoxy foams. The dynamic
mechanical properties of jute/polyester composites were
evaluated by Saha et al. [16]. Gouanve et al. [17] have
carried out DMA to understand the efciency of the
helium plasma treatment on ax ber-reinforced unsaturated polyester resin composite. The DMA of various natural ber-reinforced composites was carried out by several
workers [1820]. Mohanty et al. [21] manufactured green
composites from soy-based plastic and pineapple leaf ber
and evaluated their thermal and mechanical properties. Joseph et al. [22] found that the viscoelastic properties of sisal ber-lled low-density polyethylene composites were
inuenced by ber length and ber orientation. Liu et al.
[23] studied the impact of ber length and the processing
method on the DMA and mechanical properties of the
kenaf ber-reinforced soy-based biocomposites.
Natural ber-lled polymer composites have attracted
great interest due to increasing environmental concerns,
low costs, and their potential to compete with glass-ber
composites. The purpose of this study is to create a natu-
TABLE 1. Physical and mechanical properties of pineapple ber and

polyester resin.
Properties
PALF
Polyester
Diameter (lm)
Density (g/cm3)
Youngs modulus (MPa)
50
1.526
413
1.6
6210
1.159
22.9
1.6
970
Composite sheets of varying ber length and ber loading

were prepared. Unlled polyester resin samples were also
prepared by curing the resin with methyl ethyl ketone peroxide and cobalt naphthenate. GF/PER composites of different ber loading were also prepared using glass ber
mat.
Characterization
The DMA was carried out using TA Instruments Thermal Analyser in Dual cantilever mode. Measurements
were made at three frequencies (0.1, 1, and 10 Hz). Rectangular specimens of size 35 mm 3 10 mm 3 3 mm
were used for the study. The mechanical measurements
were carried out using an FIE universal testing machine
at a cross head speed of 50 mm/min and a gauge length
of 100 mm. Mechanical measurements of the composites
of different ber loading at optimum ber length 30 mm
[7] were also carried out. Tensile tests were done according to ASTM D 638-76 and exural properties according
to ASTM D 790. Izod impact tests on notched specimens
were carried out using International Equipments, Mumbai
as per ASTM-D256. The results reported are the average
of ve parallel tests.
ral ber-reinforced composite with excellent dynamic mechanical properties using PALF, which is abundantly
available in India. This work details a systematic study of
dynamic mechanical behavior of neat polyester resin and
short PALF-reinforced polyester composites with special
reference to ber length, ber loading, and frequency.
Fiber/matrix interactions were analyzed from the dynamic
mechanical data. The results were also compared to composites using standard glass bers as reinforcement.
EXPERIMENTAL
PALF was obtained from South India Textile Research
Association, Coimbatore. Woven E-Glass bers were supplied by Ceat, Hyderabad, India. [Tensile strength 2500
(MPa), tensile modulus 70 (GPa), elongation at break 2.5
(%).] The polyester resin (HSR, 8131 supplied by Bakelite Hylam, Hyderabad, India) was used for the study (at
258C viscosity is 650 cps and specic gravity is 1.11).
Methyl ethyl ketone peroxide (MEKP) and cobalt naphthanate were of commercial grade supplied by Sharon
Enterprise, Cochin. The physical and mechanical properties of the materials are given in Table 1.

Static Mechanical Properties
The mechanical properties of the composites at different ber loading are given in Table 2. The tensile strength
is linearly increased with ber loading, and a maximum is
observed for 40 wt% composites. Addition of 40% ber
increased the tensile modulus by 180%. The high stiffness
of these materials makes them suitable for structural
applications. The 40% ber composite exhibits higher
elongation due to increased ber matrix adhesion and better ber dispersion. All the tensile properties showed a
decrease for 50 wt% showing poor dispersion of the bers
at higher ber loading due to high ber to ber contact.
The increase in mechanical strength is primarily attributed
to reinforcing effect imparted by the bers. The best exural properties are also observed for composites with 40%
ber content. The exural modulus for 40 wt % composite is 150% greater than the neat polyester resin. However, further increase in ber loading to 50% decreased
Composite Processing
The bers were chopped to desired length, washed
with water, dried at 60708C for 2 h. Hand lay up method
was used for composite preparation. Composite sheets of
size 150 3 150 3 3 mm3 were prepared. The polyester
resin was mixed with 1 vol% cobalt naphthenate and 1
vol% methyl ethyl ketone peroxide. The resin mixture
was poured on to the ber mat prepared of denite
weight, which is placed in the mold, and the air bubbles
were removed carefully with a roller. The closed mold
was cured at room temperature for 24 h under constant
pressure (1 MPa) and postcured at 508C for further 24 h.
TABLE 2. Static mechanical properties of PALF/PER composites (ber aspect ratio 600).
Fiber loading
(wt%)
0
15
30
40
50
Tensile
strength (MPa)
23
36
49
65
59
6
6
6
6
6
3
3.6
2.8
1.2
3.7
Tensile
modulus (MPa)
970
2030
2510
2730
1900
6
6
6
6
6
95
115
160
47
100
Elongation
at break (%)
1.6
2.8
3.6
9.2
6
6
6
6
6
6
0.25
0.43
0.56
0.3
0.77
Flexural
strength (MPa)
50
46
55
70
62
6
6
6
6
6
1.5
1.2
0.9
1.8
1.5
Flexural
modulus (MPa)
1500
2100
3400
3790
3360
6
6
6
6
6
16
27
50
57
57
Impact
strength (J/m)
11
106
202
236
365
6
6
6
6
6
1
9
16
13
17
DOI 10.1002/pc
FIG. 1. Variation of storage modulus with temperature of PALF/PER

composites of different ber aspect ratio at a frequency of 10 Hz (ber
loading 40 wt%).
the value by 11%. The decrease in exural properties at

higher loading (50%) is due to the increased ber-to-ber
interactions and dispersion problems. The impact strength
of PALF/PER composite is found to increase linearly
with weight fraction of ber. As cellulose bers are porous, impact strength increases with loading. It has been
proved by Gordon and Jeronimidis [18] that the toughness
of composites increases with the microbrillar angle of
the bers and reaches a maximum at 158208. It will then
decrease with increasing angle. The orientation of the cellulose macromolecule is higher in the central layer of the
secondary wall of the cells of natural bers, and, here,
they form long lamentary brils inclined at a small angle
to the ber axis, which is the microbrillar angle. The
lower the microbrillar angle, the better will be the tensile properties (tensile modulus and strength) in the ber
axis. However, the toughness and exibility are improved
with the increase of microbrillar angle showing a maximum followed by a decrease. Pineapple bers have
microbrillar angle of 148, and so its composites exhibit
good impact properties.
Storage Modulus. Variation in dynamic modulus of
PALF-polyester composites of varying ber length (aspect
ratio) were considered as a function of temperature and
frequency. To analyze the effect of ber aspect ratio, the
ber content was set constant at 40 wt%. The variation of
storage modulus (E0 ) of composites as a function of temperature of different ber aspect ratio (l/d:length divided
by diameter of ber) at a frequency of 10 Hz is shown in
Fig. 1. The E0 values are decreased as temperature
increased for the neat resin and the composites. The composites of different ber aspect ratio exhibited similar
DOI 10.1002/pc
trend. The difference in the storage modulus of neat resin

and the composites is small at low temperature showing
that ber incorporation does not contribute to stiffness of
the material at low temperatures. The drop in modulus is
reduced on increasing temperature for reinforced composites. The composites exhibit higher storage modulus
when compared with the neat resin, and this clearly shows
the reinforcement effect of PALF in spite of the difference in the ber aspect ratio. The storage modulus, which
is a measure of the maximum energy stored in the material during one cycle of oscillation, gives an indication of
the stiffness behavior of the sample. Fiber incorporation
increases the storage modulus at high temperatures as the
bers retain their stiffness at elevated temperatures. As l/d
ratio increases, the properties increase, reach a maximum
at critical l/d ratio, and then a decrease. The composites
of 30-mm ber length (l/d 600) showed higher modulus
values at all temperatures. The stiffness, which is strongly
dependant of ber dispersion, is maximum in the composites of aspect ratio 600. The probable explanation for the
decrease in modulus of 40 mm composite (aspect ratio
800) is due to the entanglement of the longer bers. This
in fact changes the effective aspect ratio also. Viscoelastic
properties of low-density polyethylene reinforced with
short pineapple bers have been studied by George et al.
[24] and found that the storage modulus of composites
containing pineapple bers of length 2 mm showed highest value than containing 5 mm long bers.
Fiber length, which is one of the main ber-related
variables, has got a determining role at the interface. In
view of the high l/d ratio, the pineapple leaf bers are
potential reinforcements for use in polymer composites.
In this context, it is important to mention that our earlier
studies on the static mechanical properties of short PALF/
polyester composites indicated that 30-mm long bers (aspect ratio 600) were optimum for the best balance of the
mechanical properties [7].
The temperature dependence of storage modulus of
neat resin and PALF-polyester composites (ber aspect
ratio 600) of different ber loading at a frequency 10 Hz
is shown in Fig. 2. There is a gradual decrease in the storage modulus values for the resin as well as for the composites as temperature is increased. This corresponds to
the transition from the glassy to the rubbery state. Similar
decrease is observed in the storage modulus of several
natural ber reinforced composites [2527]. The E0 values
are increased with the incorporation of bers in the polyester matrix and is maximum for 40 wt% composites at
all temperatures. The above behavior is clearly observed
in Fig. 3 in which the storage modulus values are plotted
as a function of ber loading at different temperatures.
The 40 wt% composite showed highest storage modulus
values above the Tg region in the rubbery plateau, and the
ber matrix interface is not much deteriorated at higher
temperatures. Banana polyester composites depicted a
similar increase in storage modulus at 40% loading [28].
Notable decrease in modulus is observed for 50 wt %
1743
FIG. 2. Variation of storage modulus with temperature of PALF/PER

composites of different ber loading at a frequency of 10 Hz (ber aspect ratio 600).
composites at all temperatures. At higher ber loading,

dispersion becomes poor leading to a decrease in ber
matrix interaction [24, 27]. Thomas and coworkers [27]
observed similar trend up to 40 vol% of banana/sisal 1:1
in short banana/sisal hybrid ber-reinforced polyester
composites while a decrease in modulus was observed
there after for 0.48 vol % ber.
SEM studies also conrm the results obtained from
DMA. SEM of the tensile fracture surface of PALF/PER
composites of different ber weight fraction of aspect ratio 600 are shown in Fig. 4. The degree of reinforcement
and ber/matrix interfacial characteristics are clear from
FIG. 4. Scanning electron micrograph of fracture surface of PALF-PER

composites (ber aspect ratio 600; (a) 40 wt %; (b) magnied view of
40 wt %; (c) 50 wt%.
FIG. 3. Variation of storage modulus with ber loading of PALF/PER

composites at different temperatures (frequency 10 Hz, ber aspect ratio
600).
the micrographs. Better ber/matrix adhesion is observed

for 40 wt % composite in Fig. 4a and its magnied view
4b. Matrix particles are seen sticking to the ber surface.
Poor ber/matrix adhesion is observed in Fig. 4c of 50%.
Pulled out bers are also seen in 4(c). The effectiveness
of llers on the moduli of composites can be represented
by a coefcient C [28] such as
DOI 10.1002/pc
TABLE 3. The ller effectiveness coefcient C, peak height and peak

width at half height of PALF/polyester composite at 10 Hz (ber aspect
ratio 600).
Fiber loading
(wt%)
Peak height (cm)
Peak width at
half maximum (cm)
8.1
3.9
3.5
1.2
1.3
2.1
2.6
2.6
6.1
6.1
0.095
0.057
0.022
0.025
Gum
15
30
40
50
E G =E R composite
E0 G =E0 R resin
where E0 G and E0 R are the storage modulus values in the

glassy and rubbery regions, respectively. The higher the
value of the constant C, the lower the effectiveness of the
ller on the moduli. The measured E0 values at 45 and
1508C for the composite and the resin were used as E0 G
and E0 R, respectively. The values obtained for the different systems at a frequency of 10 Hz are given in Table 3.
The C values are decreased as ber loading is increased
up to 40 wt%, and an increase is observed for 50 wt%.
The effectiveness of the ller is found to be maximum for
40 wt% composite. Maximum stress transfer takes place at
this optimum ber weight fraction. At high ber loading,
there might not be enough matrix to properly cover the
bers, which would also decrease the load transfer.
The viscoelastic properties of a material are dependent
on temperature and time (frequency). Hence it is interest-
FIG. 5. Effect of frequency on the variation of storage modulus with

temperature of PALF/PER composites (ber loading 40 wt%, ber aspect ratio 600).
DOI 10.1002/pc
TABLE 4. Experimental and theoretical values of storage modulus and

tan d of PALF/PER composite of different ber loading at 508C (ber
aspect ratio 600).
Fiber loading
(wt%)
15
30
40
50
Tan d
Storage modulus (MPa)

Experimental
Theoretical
Experimental
Theoretical
3,370
3,850
3,670
1,275
3,270
3,650
3,880
4,114
0.087
0.074
0.067
0.059
0.075
0.064
0.060
0.051
ing to observe the inuence of frequency on the dynamic

properties of the composites. Figure 5 shows the inuence
of frequency on the storage modulus of PALF/PER composites of 40% loading. The storage modulus of 40 wt%
composite is increased with the increase in frequency.
This is attributed to the lesser mobility of polymeric
chains at higher frequency.
Using Einstein equation, Ec Em (1 Vf) [29], the
storage modulus of the composite is determined theoretically, where Ec and Em are the storage modulus of the
composite and matrix, respectively. The experimental values given in Table 4 are in agreement with the theoretical
values except for 50 wt%.
Damping Behavior. The variation of mechanical damping parameter tan delta (tan d) with temperature for
PALF polyester composites of different ber aspect ratio
is shown in Fig. 6. The neat resin showed higher damping
peak. The composites of 10-mm ber length (aspect ratio
200) showed maximum damping among the composites,
followed by the 40 mm (aspect ratio 800) composites
whereas 30 mm composites (aspect ratio 600) showed
least damping. Composites of 40% loading (all ber
FIG. 6. Effect of ber aspect ratio on tan delta with temperature of

PALF/PER composites at a frequency of 10 Hz (ber loading 40 wt%).
1745
TABLE 5. Values of tan dmax and Tg values of neat resin and PALF/
PER composites of different aspect ratio having 40 wt% ber loading.
Fiber length
(mm)
Gum
10
20
30
40
Tan d maximum
Frequency (Hz)
Tg from tan d
maximum (8C)
Frequency (Hz)
Fiber
aspect ratio
0.1
10
0.1
10
200
400
600
800
0.35
0.22
0.17
0.17
0.19
0.55
0.22
0.14
0.13
0.17
0.79
0.22
0.14
0.13
0.17
114
106
102
115
110
116
125
112
145
130
116
130
157
160
132
aspect ratio) showed two relaxation peaks, the higher temperature relaxation corresponds to the glass transition. We
could see the new relaxation at all aspect ratios; however,
this is more predominant at aspect ratios 200 and 800.
The additional relaxation indicates a strong interphase in
the composites of all aspect ratios. A more broadening of
the curve is observed for composites of aspect ratio 600
indicating improved interaction between the ber and the
matrix. The peak height also gets reduced, and the glass
transition temperature (Tg) showed a positive shift for
composites of l/d ratio 600. Thus, the optimum stress
transfer between the ber and the matrix takes place in
the composites of aspect ratio 600 (30-mm ber length).
At highest aspect ratio, ber entanglements occur which
reduces the ber dispersion [24].
Table 5 gives the values of tan d maximum, and Tg
values of neat polyester resin and PALF/PER composites
of different ber aspect ratio at different frequencies. The
damping shows its highest value for composites at 0.1 Hz
and decreases as frequency is increased. The tan d is
found to decrease as ber aspect ratio is increased to 600
(ber length 30 mm). This can be attributed to better
ber/matrix adhesion. However, an increase is noticed for
composites of 40 mm (aspect ratio 800). Sreekala et al.
[30] have observed similar trend in the study of the
dynamic mechanical properties of oil-palm ber/phenol
formaldehyde composites. The tan d maximum is found
to be less for 40-mm oil-palm ber composite and further
increase in ber length to 50 mm increases the tan d
value. As expected, the transition temperatures are shifted
to higher temperatures as the frequency is increased for
PALF/PER composites of different ber length. The Tg of
the neat polyester is also found to be increased with
increase of frequency. A broadening effect of glass transition region and an increase in peak temperature to the
high temperature side for composites of 30 mm (aspect
ratio 600) are indications of the high degree of reinforcement. The composites of aspect ratio 800 (40 mm long
bers) showed lower value of Tg at all frequencies.
A mechanical damping term tan d is the ratio of loss
modulus to the storage modulus E00 /E0 . It is related to the
degree of molecular mobility in the material, and damping
is affected through the introduction of bers. This is due
to shear stress concentration at ber ends in association

with viscoelastic energy dissipation in the matrix material.
Another reason could be the elastic nature of the ber. In
a composite, the molecular motions at the interface contribute to the damping of the material. The determination
of the magnitude of damping makes it possible to quantify the interface bonding. Figure 7 shows the variation of
tan d with temperature for composites of different ber
loading at a frequency of 10 Hz.
The incorporation of PALF to the polyester matrix
decreased the mechanical damping as the tan d value gets
lowered. The reinforcing bers restricted the mobility of
polymer molecules, raised storage modulus values, and
reduced the viscoelastic lag between the stress and strain,
and hence the tan d values are decreased in the composites
[15, 31, 32]. This is also due to the fact that there is less
matrix by volume to dissipate the vibrational energy [33].
This may be taken as an effect of interaction between ber
and polymer matrix. Nielson [34] reported that incorporation of reinforcing llers usually decreases damping. Also,
elasticity is increased upon addition of bers.
A small hump due to secondary/additional relaxation is
observed at a temperature lower than Tg in the composites
having higher amount of PALF. The additional relaxation
is clearly observed for 40 wt% composites (Fig. 7). The
additional relaxation could be due to the effect of the
presence of strong interphase in these composites. The
effect of the interlayer becomes more prominent for 40
wt% composite. Several authors have indicated the idea
of an interphase immediately surrounding the reinforcing
phase with properties intermediate of the ber and matrix
[26, 35]. Eklind and Maurer [36, 37] have reported on an
interlayer model to simulate the dynamic mechanical
FIG. 7. Effect of ber loading on tan delta with temperature of PALF/

PER composites at a frequency of 10 Hz (ber aspect ratio 600).
DOI 10.1002/pc
TABLE 6. Values of tan dmax and Tg values of neat resin and PALF
polyester composites of different ber loading (ber aspect ratio 600).
Tan dmax
Frequency (Hz)
Tg from tan dmax (8C)

Frequency (Hz)
Fiber loading
(wt%)
0.1
10
0.1
10
Gum
15
30
40
50
0.35
0.16
0.18
0.17
0.13
0.55
0.21
0.22
0.13
0.12
0.79
0.30
0.27
0.13
0.13
114
118
115
115
165
116
122
115
145
168
116
122
117
160
168
properties of lled blends. The ller particles have been

reported to be surrounded by an interlayer attached to the
ller surface.
The transition peak is shifted to highest temperature
upon the introduction of higher amount of ber (40%)
indicating decreased mobility of chains by the addition of
bers. This shows better polymer/ber interaction. When
the ber concentration is lower, the packing of the bers
is irregular in the composite, leading to matrix rich region
and consequently easier failure of the composite takes
place. When there is optimum concentration of the bers,
the neighboring bers will prevent crack propagation.
The values of tan d maximum and Tg of neat resin and
composites of different ber loading are given in Table 6.
The tan d value is found to be low for 40 wt% composite.
This is due to the increased interactions between the ber
and the matrix. The Tg is shifted from 115 to 1608C for
the 40% composite as frequency is increased to 10 Hz
(Table 6). The positive shift in Tg value shows the effectiveness of the ber as a reinforcement.
The Tg of the composites increases with the increase in
the frequency. The molecular motions at the interfacial
region generally contribute to damping of the material
apart from those of the constituent [38]. From the tan d
temperature curve (Fig. 7), it is evident that the addition
of ber decreases the area under the curve. The greater
area under the curve gives an indication of the total
amount of energy that can be absorbed by the material
during an experiment.
The peak height and peak width at half maximum measured from Fig. 7 are given in Table 3. The width of the tan
d curve is increased for 40 wt% composite, and this indicates that the volume of the interface is increased. The
height of the tan d peak of the composites is smaller than
that of the neat resin. The amount of polymeric resin
decreases when PALF concentration increases in the composite. Thus, a reduction in height of tan d peak is expected.
The lowering of the peak height and the increase in the
peak width at half maximum of 40 wt% composite can be
attributed to the decrease of the mobility of the polymeric
chains. This indicates better ber/matrix interaction.
It is observed that with the increase in frequency, the
value of tan d or damping is decreased for 40 wt% comDOI 10.1002/pc
FIG. 8. Effect of frequency on the variation of tan d with temperature

of PALF/PER composites (ber loading 40 wt%, ber aspect ratio 600).
posite (Fig. 8). At higher frequencies, the material exhibits elastic behavior and behaves more like a solid. At
lower frequencies increased damping results due to the
liquid like behavior as the molecule get time to move
under cyclic loading. Thus, an increase in tan d values is
observed at low frequencies. It is interesting to note that
at higher temperatures, the tan d values are not affected
by the change of frequency, and the curves overlap one
another for 40% composite (Fig. 8). The observations during mechanical measurements given in Table 1 are in
good agreement with dynamic mechanical test results.
The composite of 40 wt% loading has maximum properties both mechanical and dynamic mechanical that
showed a uniform stress distribution from the matrix to
the ber and better ber/matrix interaction.
The damping factor tan d values of the composite is
determined theoretically using Nielson equation tan dc
tan dm (1 2 Vf) [39] where c and m refer to the composite
and matrix, respectively. The experimental damping values
as a function of ber loading and the theoretically calculated values are given in Table 4, and the experimental values are in good agreement with the theoretical values.
Comparision With GF/PER Composites

Mechanical Properties. An examination of the tensile
properties of GF/PER composites reveals that addition
of 16% ber causes 96% increase in tensile strength
(Table 7). The composite with 40% GF shows maximum
tensile strength as in PALF/PER composites; however, it
is much higher, 313% than the neat resin. The tensile
modulus also increased with the increase in ber loading
up to 40 wt%. The stiffness of the material decreases for
higher ber loading 50 wt%. The percentage elongation at
break of GF/PER composite increases linearly up to 40
1747
TABLE 7. Mechanical properties of GF/PER composites.

Fiber loading
(wt%)
0
16
24
35
40
50
Tensile
strength (MPa)
23
45
73
94
95
89
6
6
6
6
6
6
3.0
3.1
2.9
4.9
3.2
1.9
Tensile
modulus (MPa)
970
2,951
2,056
4,348
4,401
2,834
6
6
6
6
6
6
95
125
88
145
95
52
Elongation
at break (%)
1.6
4.3
4.9
7.1
7.1
5.4
wt% and declines after that. At very high ber loading,

ber agglomeration results lead to a decrease in adhesion
between the ber and the matrix. Easy debonding and
ber failure occur at lower strain. At high ber loading,
there might not be enough matrix to properly cover the
bers, which would also decrease the load transfer. The
exural strength values of these composites were also
found to be less than the neat resin at low weight fraction
of the ber similar to PALF/PER. At higher concentrations of GF, the exural strength is considerably
increased. About 204% increase in exural strength is
observed on incorporation of 40% GF in polyester matrix
after which a decrease is noticed. The increase in exural
strength due to increase in ber loading is mainly due to
increased resistance to shearing. The variation in exural
properties of GF/PER composites is similar to PALF/PER
composites. The impact strength of GF/PER composite
increases largely with incorporation of ber. The composite of 16 wt% shows maximum value, which is 58 times
greater than the neat resin. At this lower loading, bers
are found embedded in the matrix. Hence, ber breakage
and ber pull-out occur on application of a sudden force.
FIG. 9. Variation of storage modulus with temperature of GF/PER

composites of different ber loading at 10 Hz.
6
6
6
6
6
6
0.25
0.15
0.30
0.35
0.25
0.20
Flexural
strength (MPa)
50
48
116
144
152
119
6
6
6
6
6
6
1.5
1.9
2.5
1.5
1.2
2.7
Flexural
modulus (MPa)
2,753
3,484
7,365
9,862
9,221
8,390
6
6
6
6
6
6
32
58
85
98
81
78
Impact
strength (J/m)
11
638
592
538
558
502
6
6
6
6
6
6
1
21
22
20
13
21
But at higher ber loadings, ber-to-ber contact

increases and ber breakage will be the predominant failure mechanism. The interber interaction decreases the
effective stress transfer between the ber and the matrix.
This decreases the impact strength.
Dynamic Mechanical Properties. Examining the variation of storage modulus with temperature of GF/PER
composites of different ber loading at 10 Hz (Fig. 9), it
is found that the storage modulus increases with the
increase in volume fraction of glass bers in the polyester
matrix even though a slight decrease is noticed for 35%.
The modulus of elasticity of GF is much higher, and
therefore 40% composite shows higher storage modulus
values. As the temperature is increased, the storage modulus exhibited a sharp drop in the temperature region from
90 to 1208C. E0 values are low for GF/PER composites at
high temperature as the GF becomes rubbery. The curves
of E0 in the rubbery state became atter, and the rubbery
range extends more to the higher temperature side when
ber loading increases. This suggests that the thermal stability of the composites increases with ber loading.
FIG. 10. Variation of tan d with temperature of GF/PER composites at

10 Hz.
DOI 10.1002/pc
Figure 10 shows the variation of tan d with temperature

for GF/PER composites of different ber loading. From
the loss tangent curves, it is seen that the neat polymer
exhibits highest tan d maximum. It is also observed that
introduction of bers reduced the peak height of the matrix. GF, a well-known reinforcing ber can be considered
as a pure elastic material, so that there is not much difference in damping with increase in GF content.
It is interesting to note that GF/PER composite (40 %)
shows higher damping behavior than PALF/PER composites.
CONCLUSION
The mechanical and dynamic mechanical properties
of short randomly oriented PALF/PER composites have
been investigated. The dynamic mechanical behavior of
PALF/ PER composites is greatly dependent on ber aspect ratio and volume fraction of the ber. Incorporation
of PALF increased the modulus and decreased the damping characteristics of the neat sample. The composite of
ber aspect ratio 600 showed higher modulus values at
temperatures ranging from 30 to 2008C. The storage
modulus is increased with increase in the ber loading,
and a maximum value is observed for composite of 40%
loading. The incorporation of 40 wt% PALF in polyester
resin resulted in a positive shift of the Tg, indicating restricted mobility of polymer molecules at the interface.
Additional relaxation is also observed for 40% composite showing the presence of a strong interphase. As the
frequency is increased, the tan d is decreased, whereas
E0 increased for the composite. The peak height is
decreased, and peak width is maximum in this case. The
ndings were correlated with morphological evidence.
Hence it can be concluded that PALF could effectively
reinforce polyester matrix when used in an optimum
concentration of bers. The tensile and exural properties of GF/PER composites vary in a similar manner as
PALF/PER composites though the properties are higher
for GF/PER composites. The impact strength increased
linearly with ber loading for PALF/PER composites
while 16 wt% showed maximum value for GF/PER composite. Incorporation of GF in to polyester matrix has
also resulted in an increase of storage modulus for 40
wt% composites. However, PALF/PER composite (40
wt%) showed higher storage modulus values above the
Tg region in the rubbery plateau. Thus, we can understand that the ber/matrix interface in PALF/PER composite is not much deteriorated even at higher temperatures.
REFERENCES
1. F.G. Torres, O.H. Arroyo, and C.J. Gomez, Therm. Plast
.Comp. Mater., 20, 207 (2007).
2. E. Bodros, I. Pillin, N. Montrelay, and C. Baley, Compos.
Sci. Technol., 67, 462 (2007).
DOI 10.1002/pc
3. B.A. Acha, M.M. Reboredo, and N.E. Marcovich, Polym.

Int., 55, 1104 (2006).
4. C.N. Zarate, M.I. Aranguren, and M.M. Reboredo, J. Appl.
Polym. Sci., 107, 2977 (2008).
5. S. Leduc, J.R.G. Urena, R. Gonzalez-Nunez, J.R. Quirarte,
B. Riedl, and D. Rodrigue, Polym. Polym. Comp., 16, 115
(2008).
6. N.E. Zafeiropoulos, G.G. Dijon, and C.A. Baillie, Compos.
A., 38, 621 (2007).
Sci., 64, 1739 (1997).
8. L.U. Devi, K. Joseph, K. C. M. Nair, and S. Thomas, J.
Appl. Polym. Sci., 94, 503 (2004).
9. J. George, S.S. Bhagawan, and S. Thomas, J. Therm. Anal.,
47, 1121 (1996).
10. J.D. Ferry, Viscoelastic Properties of Polymers, Wiley, New
York (1961).
11. T. Murayamma, Dynamic Mechanical Analysis of polymeric
Material, Amsterdam, Elsevier (1978).
12. B.F. Read and G.D. Dean, The Determination of Dynamic
Properties of Polymers and Composites, Wiley, New York
(1978).
13. L.E. Nielsen, Mechanical Properties of Polymers and Composites, Marcel Dekker, New York (1975).
14. S. Mohanty and S.K. Nayak, J. Appl. Polym. Sci., 102, 3306
(2006).
15. A.K. Bledzki and W.Y. Zhang, J. Reinf. Plast.Comp., 20,
1263 (2001).
16. A.K. Saha, S. Das, D. Bhatta, and B.C. Mitra, J. Appl.
Polym. Sci., 71, 1505 (1999).
17. F. Gouanve, M. Meyer, J. Grenet, S. Marais, F. PoncinEpaillard, and J.M. Saiter, Comp. Interf., 13, 355 (2006).
18. N. Boquillon, J Appl. Polym. Sci., 101, 4037 (2006).
19. M. Taividi, R.H. Falk, and J.C. Hermanson, J. Appl. Polym.
Sci., 101, 4341 (2006).
20. S. Mohanty, S.K. Verma, and S.K. Nayak, Compos. Sci.
Technol., 6, 538 (2006).
21. W. Liu, M. Misra, P. Askeland, L.T. Drzal, and A.K.
Mohanty, Polymer, 46, 2710 (2005).
22. K. Joseph, S. Thomas, and C. Pavithran, Mater. Lett., 15,
224 (1992).
23. W. Liu, L.T. Drzal, A.K. Mohanty, and M. Misra, Compos.
B, 38, 352 (2007).
24. J. George, K. Joseph, S.S. Bhagawan, and S. Thomas,
Mater. Lett., 18, 163 (1993).
25. B. Wielage, Th. Lampke, H. Utschick, and F. Soergel, J.
Mater. Process. Technol., 139, 140 (2003).
26. L.A. Pothan, Z. Oommen, and S. Thomas, Compos. Sci.
Technol., 63, 283 (2003).
27. M. Idicula, S.K. Malhotra, K. Joseph, and S. Thomas, Compos Sci. Technol., 65, 1077 (2005).
28. L.E. Nielson, J. Polym. Sci., Part B: Polym. Phys., 17, 1897
(1979).
29. A. Einstein, Investigation on Theory of Brownian Motion,
Dover, New York (1956).
30. M.S. Sreekala. Ph.D. Thesis, Mahatma Gandhi University,
Kottayam,1999.
1749
31. A.K. Rana, B.C. Mitra, and A.N. Banerjee, J. Appl. Polym.
Sci., 71, 531 (1999).
32. P. Ghosh, N.R. Bose, B.C. Mitra, and S. Das, J. Appl.
Polym. Sci., 64, 2467 (1997).
33. S. Das and A.K. Rana, Compos. Sci. Technol., 62, 911
(2002).
34. L.E. Nielson, Mechanical Properties of Polymers and Composites, Vol. 2, Marcel Dekker, New York (1974).
35. J.J. Arnold, M.P. Zamora, and A.B. Brennan, Polym. Compos., 17, 332 (1996).
36. H. Eklind and F.H.J. Maurer, Polymer, 38, 1047 (1997).
37. H. Eklind and F.H.J. Maurer, J. Polym. Sci., Part B: Phys
Ed., 34, 1569 (1996).
38. S. Dong and R. Gauvin, Polym. Compos., 14, 414 (1993).
39. L.T. Drazal, M.J. Rich, M.F. Koining, and P.F. Llyod, J.
Adhes., 16, 133 (1983).
DOI 10.1002/pc

Palf PDF

Încărcat de

Informații document

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Palf PDF

Încărcat de

Drepturi de autor:

Formate disponibile

Mechanical Properties of Pineapple Leaf Fiber-Reinforced

Received 13 March 1996; accepted 3 November 1996

uct of pineapple cultivation. Hence, without any

DEVI, BHAGAWAN, AND THOMAS

For NaOH treatment, the chopped fiber was

PINEAPPLE LEAF FIBER/POLYESTER COMPOSITES

ural modulus were calculated using the following

where L is the support span; b, the width of the

RESULTS AND DISCUSSION

Figure 2 Stressstrain curve of PALFpolyester

and fiber loading are given in Figures 1 and 2.

Figure 1 Dependence of stressstrain behavior of

The effect of fiber length on the tensile properties

DEVI, BHAGAWAN, AND THOMAS

Table II Effect of Fiber Length on Tensile

of the composites is 180% higher than that of a 5

the overall mechanical properties, a fiber length

PINEAPPLE LEAF FIBER/POLYESTER COMPOSITES

Figure 3 SEM of tensile failure surface of PALF

up to the fracture point. The addition of fibers

Figure 4 SEM of tensile fracture surface of PALF

Figure 5 Variation of flexural stress with deflection

Figure 6 Variation of flexural strength and flexural

DEVI, BHAGAWAN, AND THOMAS

Figure 7 Variation of flexural strength and flexural

mm-long fibers. Further increase to 40 wt % results in a slight lowering in flexural strength by

Figure 8 Variation of specific flexural stiffness of

age and delamination are seen in Figure 9. Figure

Figure 9 SEM of fracture surface of PALFpolyester

PINEAPPLE LEAF FIBER/POLYESTER COMPOSITES

observed for composites of fiber length 10 and 30

where d the fiber diameter, t the interfacial

Strictly speaking, the work of fracture arising

Table IV Variation of Work of Fracture

It is clear that the conditions leading to (1) and

DEVI, BHAGAWAN, AND THOMAS

Figure 11 Variation of work of fracture of PALF

However, in many cases of practical interest, the

and the work of fracture for a 30 wt % composite

The impact energy or strength decreases as

Figure 13 SEM of fracture surface (impact) of

PINEAPPLE LEAF FIBER/POLYESTER COMPOSITES

Effect of Chemical Treatment on Mechanical

the treatment of the fibers with silane A1100 has

Table VI Variation of Density and Hardness of

DEVI, BHAGAWAN, AND THOMAS

Tensile strength (MPa)

Source from Refs. 5 and 7.

is given in Table VII. The mechanical properties

1. P. S. Mukherjee and K. G. Satyanarayana, J.

Ageing Studies of Pineapple Leaf FiberReinforced

St. Thomas College, Kozhencherry, Kerala, India

Received 7 November 2003; accepted 29 April 2004

Key words: diffusion; ageing; polyesters; bers; composites

bers show high ultimate tensile strength and initial

UMA DEVI ET AL.

Water absorption studies of the composites at room

Test samples were immersed in boiling water for 7 h at

PINEAPPLE LEAF FIBERREINFORCED COMPOSITES

where Q is the mol % uptake at equilibrium, t is the

where is the slope of the initial linear portion of the

where M is the mass of the solvent taken up at

log Qt/Q log k n log t