Composites: Part A 37 (2006) 1473–1479

www.elsevier.com/locate/compositesa

Tensile mechanical properties, morphological aspects and

chemical characterization of piassava (Attalea funifera) fibers
a,*
J.R.M. dÕAlmeida , R.C.M.P. Aquino b, S.N. Monteiro b

a
Materials Science and Metallurgy Department, Pontifı́cia Universidade Católica do Rio de Janeiro, Rua Marquês de São Vicente,
225, Gávea – 22453-900, Rio de Janeiro, RJ, Brazil
b
Science and Technology Center, Universidade Estadual do Norte Fluminense, Av. Alberto Lamego, 2000, Horto – 28015-820,
Campos, RJ, Brazil

Received 5 August 2004; received in revised form 15 March 2005; accepted 16 March 2005

Abstract

The tensile mechanical properties of piassava fibers, as well as their chemical composition and morphological aspects, are reported.
The values obtained showed that piassava has a mechanical behavior and chemical composition comparable to that of coir fibers. The
difficulties related with a reliable way of measuring the true elastic modulus of these slender fibers are discussed and a simple correction of
the experimental data is presented. As main characteristic surface features piassava fibers present a well arranged pattern of silicon rich
star-like protrusions. Its chemical composition reveals that piassava are lignin rich fibers, 48.4 wt%. X-ray diffraction showed that cel-
lulose I is their main crystalline constituent. Their thermal degradation begins at 225 °C, and the whole thermal degradation behavior of
piassava fibers has many aspects, like the initial water loss and the content of residues, close to that shown by pure lignin.
Ó 2005 Elsevier Ltd. All rights reserved.

Keywords: A. Piassava fibers; B. Mechanical properties; C. Elastic properties; D. Scanning electron microscopy

1. Introduction composites has the benefit of less wear and deterioration

Natural fiber reinforced composites have yet many chal-
lenges to overcome in order to become largely used as reli-
able engineering materials for structural elements.
However, their use is steadily increasing and many large
industrial corporations are planning to use, or have yet
commencing to use, these materials in their products [1].
One of the great advantages of these composites is the large
quantity of natural fibers available worldwide. Further-
more, as highlighted on many texts, e.g. [2], natural fibers
are a biodegradable and renewable resource. Their softness
is also an advantage in relationship to the common syn-
thetic fibers, like glass fiber. Natural fibers are much less
abrasive and, therefore, the manufacture of natural fiber
*
Corresponding author. Tel.: +55 021 31141234; fax: +55 021 3114
1236.
E-mail address: dalmeida@dcmm.puc-rio.br (J.R.M. dÕAlmeida).
of machine member parts.
Flax and hemp are the natural cellulosic fibers that at-
tract the most attention in Europe [3]. On the other hand,
jute, sisal, and to a lesser extent coir, are more extensively
reported in the literature; see for example [4–9]. There are,
however, many other less studied cellulosic fibers that are
attractive, whether by their morphology, by their intrinsic
properties or by their cost. Luffa cylindrica is one fine
example of such fibers. Its fibersÕ natural mat like spatial
arrangement provides a detouring path for cracks, and pro-
duces tougher composites [10]. Another of such less studied
fibers is piassava (Attalea funifera).
An earlier study from our group has shown that it is a
promising candidate as fiber reinforcement for resin matrix
composites [11]. Therefore, in this work, the chemical com-
position, structural characteristics and tensile mechanical
properties of piassava fibers are presented and its main fea-
tures are highlighted and discussed. One has to point here,
that although piassava has some characteristics common to

1359-835X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.compositesa.2005.03.035
1474 J.R.M. d’Almeida et al. / Composites: Part A 37 (2006) 1473–1479
other lignocellulosic fibers, their properties have not yet
been fully described. And, in order to enable their reliable
use in composites, the basic aspects of their structure and
morphology, such as the lignin content and surface topo-
graphic features, and aspects related to their thermal
behavior, such as the onset of thermal degradation, must
be analyzed. This work aims to fulfill this gap of informa-
tion about these traditional fibers.
2. Piassava (A. funifera)
Piassava is a stiff fiber extracted from the leaves of a
palm tree of natural occurrence at the Brazilian Atlantic
rain forest. The natural incidence range of the piassava
palm tree is close to the coastal zone (up to 60 km inland)
and it is mainly concentrated between 13°S and 17°S lati-
tudes. Its grow is favored at regions with hot average tem-
perature (24 °C) and high relative humidity (>80%). The
fibers, up to 4 m long with an average width of 1.1 mm,
are harvested once by year. The palm leaf is cut, and the
fibers are mechanically separated from the petiole. Each
palm tree can yield 8–10 kg of fibers. Fig. 1 shows typical
cross-sections of piassava fibers, where one can see the
usual variability of form and size common to all naturally
occurring cellulosic fibers.
The production of piassava fibers is steadily increasing
over the past years and, nowadays, it amounts to about
100,000 metric tons per year, as shown in Fig. 2 [12]. The
main use of these fibers is for industrial and domestic
brooms, industrial brushes, ropes and baskets, carpets
and roofs. It is estimated that around 20% of the fiber pro-
duction is disregarded by the transformation industries.
These leftover fibers, that do not meet the requisites for
the uses cited above, are, however, undamaged and long
Fig. 1. Transversal cross-section of piassava fibers. Observe the variability
of shape and size.
100000
80000
production, ton.
60000
40000
20000
0
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999
year
Fig. 2. Annual production of piassava fibers.
enough to be used as reinforcement in fiber composites
[11].
3. Experimental methods and materials
Piassava fibers disregarded by a broom industry were
used in this work. Their tensile properties were evaluated
following the procedures recommended by the ASTM stan-
dard D-2101. Since it is not possible to attach mechanically
gripped extensometers to the single fibers under test, sev-
eral gage lengths were used in order to measure the elastic
compliance of the test apparatus and to correct for the true
modulus of elasticity. The correction procedure used is out-
lined in Appendix A. The gage lengths used were: 150, 120,
80, 50, 25 and 15 mm. The tests were performed on a
mechanically driven machine with 10 kN of capacity and
equipped with pneumatic action grips. An optimum hold
pressure of 0.3 MPa was experimentally found to give reli-
able test data. With this clamping pressure neither fiber
slippage during tests nor fiber crushing at the grips were
found to occur. The test speed used was 1 mm/min and
15 specimens were tested per gage length.
The morphological characterization of the fibers was
done by scanning electron microscopy (SEM). The analysis
was performed on gold sputtered samples using secondary
electrons, and with a beam voltage of 15–20 kV. The com-
position of topographic features identified at the surface of
the fibers was determined by energy dispersive X-rays spec-
trometry (EDS).
The thermal stability of the fibers was evaluated by
thermogravimetric analysis (TGA). This experiment was
conducted under nitrogen atmosphere, from ambient
temperature (25 °C) to 400 °C at a heating rate of 10 °C/
min.
The crystalline character and the chemical composition
of the fibers were determined by X-rays diffraction and
by the van Soest method, respectively. To perform the X-
rays analysis a sample was calcinated at 400 °C by 2 h. This
sample was then scanned from 2h = 5° to 2h = 125°, with
 J.R.M. d’Almeida et al. / Composites: Part A 37 (2006) 1473–1479
increments of 0.05°, using the Cuja radiation. The equip-
ment was set at 35 kV and 40 mA. The van Soest method
is one of the several acid-detergent fiber methods used to
evaluate lignin and cellulose content of plant materials. It
involves the hydrolysis of the cellulose component with
72% w/v sulphuric acid solution with deionised water. A
complete description of this method, as well as a descrip-
tion of its precision in relationship to other acid-detergent
fiber methods, can be found elsewhere [13,14].
4. Experimental results and discussion
The surface morphology of the piassava fibers is shown
in Fig. 3(a). As observed at higher magnification, Fig. 3(b),
areas of the surface of the fibers are covered by randomly
distributed organic material. This organic material is prob-
ably a residue left by the extraction of the fibers from the
Fig. 3. (a) General aspect of the piassava fibers; (b) Organic residue
attached to the surface of the fibers.
1475
palm tree, since many areas appear clean, and there one
can observe the common parenchyma cells seen at the out-
er surface of lignocellulosic fibers [15,16], Fig. 4. Fig. 4 also
shows, besides parenchyma cells (!), an array of pro-
truded features. These protrusions are shown in detail in
Fig. 5 and were identified by EDS as Si rich particles
(Fig. 6). Similar features were also reported for coir fibers
[8,17].
A typical stress–strain curve of piassava fibers is shown
in Fig. 7. One can see a non-linear region following the ini-
tial linear portion of the graph. Similar behavior was also
reported for other lignocellulosic fibers like sabai [16], coir
Fig. 4. The cleaner parts of the fibersÕ surface show the common
topographic aspects characteristic of natural fibers (!) and an array of
protrusions.
Fig. 5. Magnifying view of the protrusions.
1476 J.R.M. d’Almeida et al. / Composites: Part A 37 (2006) 1473–1479

Table 1
Mechanical properties of the piassava fibers
Gage length (mm) ry (MPa) rr (MPa) er (%)a E (GPa)a
15 75.8 ± 18.0 134.8 ± 22.5 21.9 ± 4.3 1.07 ± 0.27
25 81.7 ± 12.4 147.3 ± 25.6 16.3 ± 2.4 2.27 ± 0.52
50 78.4 ± 14.3 131.1 ± 27.1 11.9 ± 4.3 2.58 ± 0.39
80 75.6 ± 8.5 132.8 ± 12.5 8.4 ± 1.9 3.57 ± 0.52
120 69.8 ± 12.6 108.5 ± 16.0 6.4 ± 2.3 3.55 ± 0.86
150 80.2 ± 25.0 142.9 ± 46.8 7.8 ± 3.9 4.59 ± 0.55
Tensile test.
a
Uncorrected values due to test apparatus compliance. See the text for
explanation.

Young´s modulus, GPa 4

Fig. 6. EDS spectrum of the protrusions, showing that they are Si rich
compounds.
3

160 2

140
1
120
0 20 40 60 80 100 120 140 160
100 gage length, mm
srtessM,Pa

80 Fig. 8. Values of the corrected YoungÕs modulus as a function of the gage

length (j). Uncorrected values (H). The value predicted by the procedure
60 described in Appendix A is shown as a dashed line.

40

20 more prominent. Therefore, the elastic modulus decreases

0 contribution of the test machine, the values of the YoungÕs
0 1 2 3
strain, %
Fig. 7. Typical stress–strain curve of the piassava fibers.
[17,18] and cellulose [19], and to a lesser extend banana fi-
bers [20]. The experimental results obtained are shown in
Table 1, where ry is the 0.2% offset yield stress, rr is the
stress at rupture, er is the strain at rupture and E is the
YoungÕs modulus. Despite a fairly large variation, that in-
deed are common to lignocellulosic fibers, some clear
trends could be established. At first, one can see that both
the YoungÕs modulus and the deformation at rupture are
varying with the gage length. Since the YoungÕs modulus
is actually not dependent from gage length, these results
must be due to the contribution of the compliance of the
test machine. In fact, from the data shown in Table 1,
one can see that as the gage length is reduced the contribu-
tion of the elastic elongation of the test apparatus becomes
while the deformation increases. To eliminate the elastic
4 5
modulus were corrected using the procedure summarized in
Appendix A. Fig. 8 shows the effect of the correction on the
fibersÕ modulus as a function of the gage length. As could
be expected, an average constant value is obtained, inde-
pendent of the gage length. Moreover, the corrected values
approach the one obtained using the linear fit described in
Appendix A.
The corrected value obtained is, however, low when
compared with other common natural fibers, as for exam-
ple jute (E = 13–26.5 GPa [2,21]) or flax (E = 50 GPa [3]),
although it is on the range of cotton (E = 1.1–12.6 GPa
[2,22]) and coir (E = 3–6 GPa [22]) and near the lower
bound for sisal (E = 9.4–22 GPa [2]).
From Table 1, it is also apparent that both the yield
stress and the stress at rupture were not dependent on
the test apparatus compliance, as one should expect. There-
fore, average values could be calculated from the results of
all tested specimens. The computed values for these proper-
ties are ry = 76.9 ± 4.2 MPa and rr = 133.0 ± 13.5. To cal-
 J.R.M. d’Almeida et al. / Composites: Part A 37 (2006) 1473–1479
culate these values, the actual area of each fiber was mea-
sured at five different points along the gage length. The
average area value of each fiber was then used on the
calculations.
As for the YoungÕs modulus, the tensile strength of the
more common natural fibers are higher than that found
for piassava fibers. For instance, jute (rr = 550–900 MPa
[22]), sisal (rr = 507–640 MPa [21,22]) or banana (rr =
529–914 MPa [23]). However, the well reported coir fiber
has tensile strength level comparable to the one found here
for piassava, viz. rr = 106–270 MPa [22,23].
The results of the thermogravimetric analysis are shown
in Fig. 9. One may note that following a small low temper-
ature loss of weight between 75 and 85 °C, the onset of the
fiber thermal degradation does not occur until 225 °C. The
low temperature weight loss of 5.18% can be attributed to
water in the form of absorbed moisture or combined water.
Similar results for this low temperature loss of weight were
obtained for other lignocellulosic fibers like luffa, 5% [10],
jute, with a fairly higher value of 10.2% [24], flax (6.3%)
and wheat straw (7.3%), as well as for a-cellulose (7.6%)
and lignin, 5.8% [25]. The temperature determined for the
onset of thermal degradation of piassava fibers also agrees
with the values reported on the literature for other lignocel-
lulosic fibers and materials. Table 2 compares these data.
Fig. 9. TGA curves: (a) weight loss vs. temperature; (b) derivative of the
weight loss vs. temperature.
Table 2
Temperature of the onset of lignocellulosic fibers degradation TGA
Fiber Temperature (°C)
Piassaba 225
Sisal [9] 302
Luffa [10] 250
Sabai [16] 244
Jute [24] 282
Wheat straw [25] 283
Flax [25] 330
a-cellulose [25] 322
Lignin [25] 213
1477
One can see that the temperature at which piassava fibers
begin to degrade is close to that of lignin, being situated
at the lower bound of the temperatures of the other com-
mon lignocellulosic fibers. The TGA also shows that a res-
idue of about 40% remains at 400 °C. This value is similar
to the value of 32.8% reported for lignin [25].
In Fig. 9, it is also shown the curve of the first derivative
of the weight loss versus temperature. One can clearly no-
tice the presence of two peaks at high temperatures, occur-
ring at 276 and 347 °C, respectively. The peak at 347 °C
can be associated with the decomposition of a-cellulose.
A value of 345 °C was, indeed, found for the decomposi-
tion of a-cellulose in jute fibers [24]. Both values are, how-
ever, slightly higher than the one listed in Table 2 for a pure
a-cellulose sample [25]. The peak at 276 °C could be asso-
ciated to the decomposition of hemicellulose. Values rang-
ing from 260 [26] to 281 °C [27] are found in the literature.
A clear peak for lignin was not observed, but this can be
attributed to the upper limit temperature of the test appa-
ratus used, that reaches only 400 °C. Values of 375 °C or
higher are reported for degradation temperature of lignin
[26,28], and acid treated lignin, as investigated in this work,
are expected to have higher degradation temperatures due
to condensation [26].
The X-ray diffraction spectrum for piassava is shown in
Fig. 10. The result obtained is very similar to those of other
lignocellulosic fibers, like sisal [29], and reflects the crystal-
line lattice of cellulose I [29].
The chemical composition of piassava fibers is shown in
Table 3, where the composition of the most commonly
used lignocellulosic fibers are also listed for comparison.
The results obtained revealed that piassava has a higher lig-
nin content than any of the other common lignocellulosic
fibers. This could be responsible for its inherent flexural
rigidity and water proof resistance. The figures here re-
ported, however, agree very well with previously reported
data for piassava fibers [30,31]. One can note from Table
3, and from the previously listed data of the mechanical
properties, that piassava has many similarities with coir
fiber.
Fig. 10. X-ray spectrum. The peak at 24.7° corresponds signs the presence
of cellulose I.
1478 J.R.M. d’Almeida et al. / Composites: Part A 37 (2006) 1473–1479
Table 3
Chemical composition, in wt%, of piassava and other lignocellulosic fibers
Fiber Lignin (%) Cellulose (%) Hemicellulose (%)
Piassava 48.4 31.6 – The chemical analysis performed showed that piassava
Jute [21] 13.0 61.0 20.4
Sisal [21] 8.0 78.0 10.0
Coir [21] 45.0 43.0 0.3
Flax [21] 2.2 71.0 18.6
Hemp [21] 3.7 74.4 17.9
5. A note on piassava–polyester composites
Since the mechanical properties obtained place piassava
on the group of low strength–low modulus lignocellulosic
fibers, one has to ask if they are worth for being used as
reinforcement. Preliminary results on unidirectional pias-
sava–polyester composites show that, in fact, piassava pro-
duces an outstanding reinforcement effect. For composites
with mass fraction of fibers ranging from 0.30 to 0.50,
tested in three-point bending at a crosshead speed of
1 mm/min, values of up to 42 MPa were obtained for the
stress at rupture [11]. This is a very promising result, be-
cause it is comparable to the flexural data from other ligno-
cellulosic reinforced composites, such as jute–polystyrene
(r = 45–55 MPa) [32], coir–polyester (r = 29–41 MPa)
[17], straw–polyester (r = 34–56 MPa) [33] or banana–
polyester (r = 28–97 MPa) [34].
It is worth mentioning that the value obtained, i.e.,
42 MPa, for the stress at rupture of the piassava-polyester
composites is also comparable to the values shown by com-
mercial wood based materials and bagasse boards that are
largely used as furniture. For example, the stress at rupture
of plywood typically ranges from 15 to 21 MPa [35], and
that of low density particle board from 5 to 10 MPa [36].
Moreover, it can also be emphasized that the low cost of
piassaba fibers associated with their large diameter, which
can facilitate handling, are valuable characteristics of these
fibers.
One can also expect that chemical treatments could im-
prove the stress transfer at the fiber/matrix interface.
Therefore, composites are now being prepared with surface
treated fibers and their tensile and flexural behavior will be
presented in a future work.
6. Conclusions
An investigation on the mechanical properties, surface
morphology, chemical composition and crystalline charac-
ter of piassava (A. funifera) fibers was presented. The re-
sults of the tensile mechanical tests showed that piassava
fibers have lower stress at rupture and lower YoungÕs mod-
ulus than the more common jute or sisal fibers, but their
mechanical properties are comparable to that of coir fibers.
The fibers present a regular array of silicon rich star-like
protrusions, which could help their mechanical interlock
with resins when used as reinforcements in resin matrix
composite materials. With this regard, it is interesting to
note that coir fibers have also a regular array of structural
features at their surface.
is a lignin rich fiber, showing again similarities with lignin
rich coir fibers. Their thermal degradation behavior dis-
played many similarities with the thermal behavior of pure
lignin. X-ray analysis revealed that cellulose I is the main
crystalline constituent of piassava fibers.
The strong similarities between piassava and coir are a
promising sign that todays relatively large amount of pias-
sava leftover can be used as an effective reinforcement for
resin matrix composite materials.
Acknowledgments
The authors thank the Brazilian Agencies CNPq and
FENORTE for their support.
Appendix A. Correction of the elastic contribution of the test
machine
The elastic deformation of the grips and frame from the
test apparatus can result on misleading data when fibers are
tested. Mechanical or electrical extensometers cannot be
used for these slender specimens. Therefore, a correction
has to be made, otherwise one can find a dependence of
the YoungÕs modulus, E, with the gage length of the tested
specimen. The correction here used was earlier described by
Guimarães and Chawla [37].
The stiffness, ks, of the system formed by the test speci-
men and the test machine is given by:
1=k s ¼ 1=k m þ 1=E  L0 =A0 ; ðA:1Þ
where km is the stiffness of the test machine and E, L0 and
A0 are, respectively, the YoungÕs modulus, the gage length
and the cross-section of the test specimen. The value of ks is
obtained from the linear portion of the load times elonga-
tion curve and, thus, is directly related to the uncorrected
YoungÕs modulus. Namely, ks = P/D‘ where P and D‘
are elements of a coordinate pair on the linear part of the
load–deflection curve. The plot of this value against L0/
A0, when varying the gage length of the test specimens
while maintaining a constant cross-section, will permit
one to obtain E and km. For cellulosic fibers, as shown in
Fig. 1, the cross-section also varies. Therefore, each fiber
had its mean cross-section measured by a micrometer along
its length, in order to minimize errors. At least five mea-
sures were made per fiber.
The correctness of this approach can be established by
the adherence of the experimental points to a straight line.
Fig. A1 shows the plot obtained using the above method.
The coefficient of correlation obtained was 0.973. The val-
ues obtained for km and E were:
k m ¼ 82 kN=m;
E ¼ 6:2 GPa.
 J.R.M. d’Almeida et al. / Composites: Part A 37 (2006) 1473–1479
0.045
0.040
0.035
1 / ks ,mm/N
0.030
0.025
0.020
0.015
0 20 40 60 80 100 120 140 160 180 200
-1
L0 / A 0, mm
Fig. A1. Fit of the experimental data points to Eq. (A.1).
References
[1] Schuh T, Gayer U. Automotive applications of natural fiber
composites. Benefits for the environment and competitiveness with
man-made materials. In: Leão AL, Carvalho FX, Frollini E, editors.
Lignocellulosic–plastics composites. Botucatu, Brazil: Unesp Pub-
lishers; 1997. p. 181–95.
[2] Bledzki AK, Gassan J. Composites reinforced with cellulose based
fibers. Prog Polym Sci 1999;24:221–74.
[3] Peijs T. Natural fiber based composites. Mater Technol
2000;15:281–5.
[4] Chand N, Tiwary RK, Rohatgi PK. Bibliography: resource structure
properties of natural cellulosic fibres – an annotated bibliography. J
Mater Sci 1998;23:381–7.
[5] Chand N, Hashmi SAR. Mechanical properties of sisal fibre at
elevated temperatures. J Mater Sci 1993;28:6724–8.
[6] Mohanty AK, Misra M. Studies on jute composites – a literature
review. Polym-Plast Technol Eng 1995;34:729–92.
[7] Costa FHMM, dÕAlmeida JRM. Effect of water absorption on the
mechanical properties of sisal and jute fiber composites. Polym-Plast
Technol Eng 1999;38:1081–94.
[8] Calado V, Barreto DW, dÕAlmeida JRM. The effect of a chemical
treatment on the structure and morphology of coir fibers. J Mater Sci
Lett 2000;19:2151–3.
[9] Li Y, Mai YW, Ye L. Sisal fibre and its composites: a review of recent
developments. Comp Sci Technol 2000;60:2037–55.
[10] Boynard CA, dÕAlmeida JRM. Morphological characterization
and mechanical behavior of sponge gourd (Luffa cylindrica) –
polyester composite materials. Polym-Plast Technol Eng 2000;39:
489–99.
[11] Aquino RCMP, dÕAlmeida JRM, Monteiro SN. Flexural mechanical
properties of piassava fibers (Attalea funifera) – resin matrix
composites. J Mater Sci Lett 2001;20:1017–9.
[12] http:www1.ibge.gov.br/estatistica/economia/agropecuaria/censo-
agro/defaut.shtm.
[13] van Soest PJ. Use of detergents in the analysis of fibrous feeds II. A
rapid method for the determination of fibre and lignin. J Assoc Off
Agric Chem 1963;46:829–35.
[14] Rowland AP, Roberts JD. Lignin and cellulose fractionation in
decomposition studies using acid-detergent fibre methods. Common
Soil Sci Plant Anal 1994;25:269–77.
1479
[15] Mattoso LHC, Ferreira FC, Curvelo AAS. Sisal fiber: morphology
and applications in polymer composites. In: Leão AL, Carvalho FX,
Frollini E, editors. Lignocellulosic–plastics composites. Botucatu,
Brazil: Unesp Publishers; 1997. p. 241.
[16] Chand N, Rohatgi PK. Potential use, mechanical properties and
thermal studies of sabai grass fibre. J Mater Sci Lett 1992;11:578–80.
[17] Prasad SV, Pavithram C, Rohatgi PK. Alkali treatment of coir fibres
for coir–polyester composites. J Mater Sci 1983;18:1443–54.
[18] Venkataswamy MA, Pillai CSK, Prasad VS, Satyanarayana KG.
Effect of weathering on the mechanical properties of midribs of
coconut leaves. J Mater Sci 1987;22:3167–72.
[19] Kompella MK, Lambros J. Micromechanical characterization of
cellulose fibers. Polym Test 2002;21:523–30.
[20] Kulkarni AG, Satyanarayana KG, Rohatgi PK, Vijayan K. Mechan-
ical properties of banana fibres (Musa sepientum). J Mater Sci
1983;18:2290–6.
[21] Bledzki AK, Reihmane S, Gassan J. Properties and modification
methods for vegetable fibers for natural fiber composites. J Appl
Polym Sci 1996;59:1329–36.
[22] Baley C, dÕAnselme T, Guyader J. Matériaux composites renforcés
par des fibres naturelles. Composites 1997;37(22):28–34 [in French].
[23] Satyanarayana KG, Sukumaran K, Kulkarni AG, Pillai SGK,
Rohatgi PK. Fabrication and properties of natural fibre-reinforced
polyester composites. Composites 1986;17:329–33.
[24] Das S, Saha AK, Choudhury PH, Basak RK, Mitra BC, Todd T,
et al.. Effect of steam pretreatment of jute fiber on dimensional
stability of jute composite. J Appl Polym Sci 2000;76:1652–61.
[25] Hornsby PR, Hinrichsen E, Tarverdi K. Preparation and properties
of polypropylene composites reinforced with wheat and flax straw
fibres. J Mater Sci 1997;32:443–9.
[26] Ramiah MV. Thermogravimetric and differential thermal analysis of
cellulose, hemicellulose, and lignin. J Appl Polym Sci 1970;14:
1323–37.
[27] Órfão JJM, Antunes FJA, Figueiredo JL. Pyrolysis kinetics of
lignocellulosic materials – three independent reactions model. Fuel
1999;78:349–58.
[28] Raveendran K, Ganesh A, Khilar KC. Pyrolysis characteristics of
biomass and biomass components. Fuel 1996;75:987–98.
[29] Rong MZ, Zhang MQ, Liu Y, Yang GC, Zeng HM. The effect of
fiber treatment on the mechanical properties of unidirectional sisal-
reinforced epoxy composites. Comp Sci Technol 2001;61:1437–47.
[30] Schuchardt U, Bianchi ML, Gonçalves AR, Curvelo AAS, Biscolla
FC, Peres LO. Piassava fibers (Attalea funifera). I. Chemical analysis,
extraction and reactivity of its lignin. Cellulose Chem Technol
1995;29:705–12.
[31] Gonçalves AR, Schuchardt U, Bianchi ML, Curvelo AAS. Piassava
fibers (Attalea funifera): NMR spectroscopy of their lignin. J Braz
Chem Soc 2000;11:491–4.
[32] Snijder MHB, Reinerink EJM, Bos HL. Polypropylene and polysty-
rene reinforced with annual plant fibers: screening of coupling agents.
In: Proceedings of the 2th international symposium on natural
polymers and composites. Atibaia - Brazil, September, 1998. p. 135–8.
[33] White NM, Ansell MP. Straw-reinforced polyester composites. J
Mater Sci 1983;18:1549–56.
[34] Zhu WH, Tobias BC, Coutts RSP. Banana fibre strands reinforced
polyester composites. J Mater Sci Lett. 1995;14:508–10.
[35] Carter GF, Paul DE. Materials science and engineering. Metals
Park: ASM International; 1991, p. 281.
[36] Monteiro SN, Rodrigues RJS, de Souza MV, dÕAlmeida JRM. Sugar
cane bagasse waste as reinforcement in low cost composites. Adv
Perform Mater 1998;5:1–9.
[37] Guimarães JR, Chawla KK. Characteristics of the stiffness of
a universal tensile testing apparatus. Metalurgia 1978;34:549 [in
Portuguese].