Composites Part B 106 (2016) 88e98

Contents lists available at ScienceDirect

Composites Part B
journal homepage: www.elsevier.com/locate/compositesb

The effect of surface treatment on the performance of flax/

biodegradable composites
Panayiotis Georgiopoulos a, Aggelos Christopoulos a, Stefanos Koutsoumpis b,
Evagelia Kontou a, *
a
Mechanics Department, National Technical University of Athens, Iroon Polytechniou 9, Zografou, 15780, Athens, Greece
b
Physics Department, National Technical University of Athens, Iroon Polytechniou 9, Zografou, 15780, Athens, Greece

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

Article history: In the present study, three series of natural fiber-polymeric composite materials have been prepared and
Received 1 July 2016 experimentally studied, in terms of Scanning Electron Microscopy, Differential Scanning Calorimetry,
Received in revised form Tensile-Flexural testing, Dynamic mechanical analysis and Creep. The materials are unidirectional flax
31 August 2016
fiber composites, prepared with the “film stacking method”, based on three types of biodegradable
Accepted 8 September 2016
Available online 10 September 2016
polymeric matrices. The addition of flax fibers with this method leads to a significant toughening of the
polymers under investigation. In order to further improve the composites thermomechanical perfor-
mance, three different surface treatments of flax fibers, namely silanization, plasticization and treatment
Keywords:
Polymer-matrix composites (PMCs)
with maleic anhydride, were employed. The effect of the fibers treatment on the mechanical properties
Fibres of the composites has been comparatively studied and discussed. The surface treatment with maleic
Mechanical properties anhydride resulted to the highest Young's modulus increment, whereas silanization and plasticization
Rheological properties improved both tensile and flexural properties. Similar effect was obtained for storage modulus incre-
ment, as well creep resistance of the composites with modified flax fibers.
© 2016 Elsevier Ltd. All rights reserved.

1. Introduction
In recent years, due to the growing environmental awareness
and the necessity for developing sustainable materials [1], there is a
trend of replacing conventional fibers (glass, carbon), which are
predominantly used as reinforcements in polymer matrices, with
bio-fibers, i.e. fibers derived from natural sources such as plants,
animals and minerals [2]. Low cost, low specific weight, flexibility
in processing, competitive mechanical properties, recyclability,
safer handling and working conditions but principally their
contribution to the environmental issue, are the advantages of bio-
fibers, in comparison with glass fibers, and the reason for the
growing trend of using them [2e5]. However, there are com-
pounding difficulties due to the inherent polar and hydrophilic
nature of natural fibers, as a result of the strongly polarized hy-
droxyl groups of lignocellulose [6], and the non-polar features of
most thermoplastics. This results to non-uniform dispersion of the
fibers, while their high moisture absorption leads to bad adhesion
* Corresponding author.
E-mail address: ekontou@central.ntua.gr (E. Kontou).
with the matrix with the creation of voids at the interface [2]. Plant-
based natural fibers consist of cellulose, hemicellulose, lignin,
pectins and waxes. Their mechanical properties, thus, their rein-
forcing efficiency, depend on the content, nature and crystallinity of
cellulose, and the angle of microfibrils [2e7]. Cellulose is a natural
polymer, consisting of D-anhydroglucose (C6H11O5) repeating units,
joined by 1,4-b-D-glycosidic linkages at C1 and C4 position [8].
Among the various natural fibers, flax, kenaf, jute, hemp, coir,
bamboo, ramie and sisal [4,9], are of particular interest.
In order to preserve the environmental advantages and the eco-
efficiency of the natural fibers, biodegradable or recyclable poly-
mers are considered as an ideal matrix for this kind of composites
[10]. Biodegradability is an alternative eco-efficient solution for the
end-of-life disposal of polymer products, if recyclability is not an
option [11]. The so-called “green composites”, by reinforcing bio-
polymers with natural fibers [12], due to their significant envi-
ronmentally beneficial properties, create a new perspective for the
composites technology. Construction, decoration, decks, roofs,
auto-motive and soundproofing are some of the continuously
increasing applications, or fields of application of biocomposites,
exploiting their high stiffness and strength [2,3]. By using a
biodegradable material as a matrix, the derived bio-composite is

http://dx.doi.org/10.1016/j.compositesb.2016.09.027
1359-8368/© 2016 Elsevier Ltd. All rights reserved.
 P. Georgiopoulos et al. / Composites Part B 106 (2016) 88e98
totally biodegradable, thus, there is a new perspective by producing
limited life-time polymer composites, designed to biodegrade after
use. This kind of materials may find interesting applications as
temporary structural material, mainly in agriculture and con-
struction field [13]. Hydrophilic nature of most of the biodegradable
polymers leads to the assumption of better interaction with natural
fibers. Specifically for Poly-Lactic Acid (PLA), its polar structure is
expected to provide improved adhesion of the fibers, thus,
improved properties for the bio-composite [14]. This is not
confirmed by Oksman et al. [15], as pull-outs and clean fiber sur-
faces were noticed in the fracture surface.
In recent years, material systems, based on a biodegradable
matrix reinforced with flax fibers have been studied [11,13,15e23].
Polylactic acid (PLA) in particular, due to its competitive properties,
has been examined as a matrix reinforced with kenaf fibers
[24e26], bamboo fibers [27], hemp [28], jute [29,30] etc. The study
of biodegradation procedure of a variety of composites, is focused
on the special behavior exhibited by each specific matrix [11,25,27].
In a PLA/bamboo fiber composite, it was found that its degradability
could be controlled by the amount of the compatibilizing agent,
namely LDI (Lysine based diisocyanate). The degree of fibers
adhesion on the matrix was effected by the LDI content that has
been employed [27]. In a composite material, based on PLA and
kenaf fibers, it was found that the biodegradation rate is increasing
due to the presence of the fibers [25]. Another interesting topic is
the role of plasticizers, employed for the compatibilization of PLA,
and their absorption by the natural fibers [15].
Limited studies are reported, to our knowledge, for unidirec-
tional natural fiber composites, regarding mostly polypropylene
matrix [31e33] and epoxy resin [34e36]. Regarding biodegradable
unidirectional natural fiber composites, studies have been made for
PLA composites [24,33] and starch composites [22,37]. A major
problem of the effectiveness of natural fiber/polymer composites is
the incompatibility between hydrophilic fibers and hydrophobic
matrix. This can lead to a poor fiber/matrix adhesion due to the
presence of pendant hydroxyl and polar groups in the components.
As a consequence, high moisture uptake takes place leading to a
deterioration of the mechanical properties [3] Therefore, an
important issue in studying polymer/natural fiber composites is the
improvement of the fiber-matrix interaction, either by modifying
the fiber's surface or by the matrix modification employing addi-
tives called coupling agents. Combination of the two approaches
has been also reported [38]. In Ref. [39] five different chemical
treatments were applied to examine their effect on the composites
mechanical performance. Referring to chemical modification, it is
meant as a chemical reaction between some constituents of the
natural fiber and a chemical reagent, forming a covalent bond be-
tween them. This results in both, a better fiber dispersion, which is
usually restricted due to the hydrogen bonding between fibers, and
bond formation between fiber and polymer matrix [9].
The scope of this study is the evaluation of the effectiveness of
three different biodegradable polymers as matrix materials for
unidirectional flax fiber composites, prepared with the “film
stacking method”, a procedure which has not been studied suffi-
ciently so far. Previous work [40] has been focused on
manufacturing unidirectional plant fiber reinforced composites, as
well as on the use of a novel plant fiber yarn surface treatment
method to enhance mechanical properties and fiber-matrix adhe-
sion. Therefore, orientation of the fibers and less thermo-
mechanical processing, as mixing is avoided, are considered to be
the advantages of this method employed for the preparation of
composites. In the majority of previous studies on these compos-
ites, the reinforced matrix is polypropylene, due to its low cost and
thermal stability. The composites prepared in the present study, are
essentially novel materials due to both, the matrix material and the
89
unidirectional character of the fibers incorporated. In addition,
three different methods improving the compatibility between
matrix and fiber were applied, namely treatment of flax fibers with
silane, plasticizer and maleic anhydride. Only few works have been
previously done, to comparatively analyze the effect of similar fiber
treatments on the thermomechanical performance on the mate-
rials. The thermal and mechanical properties of the prepared
composites were comparatively investigated, providing informa-
tion about the fibers/polymers efficiency to the mechanical
enhancement, for each fiber's modification procedure employed.
2. Experimental part
2.1. Materials
Natural fiber polymer composites were prepared, based on
three different types of polymer biodegradable matrices as follows.
Poly-lactic Acid (PLA), which was supplied by NatureWorks LLC.
The selected grade 2002D has a D content of 4.25%, a residual
monomer content of 0.3% and a density of 1.24 g/cm3. The material
in pellets form was dried at 45
C for a minimum of 8 h prior to use
in a desiccating dryer. The second matrix under the commercial
name Ecovio® was supplied by BASF SE (Ludwigshafen, Germany).
The selected grade Ecovio® L BX 8145 (EC), is basically a blend of
poly(butylene adipate-terephthalate) copolyester (Ecoflex® F BX
7011), which is based on non-renewable resources, and PLA
(NatureWorks). Because of the PLA content, Ecovio® L BX 8145
consists of 45% of renewable resources. The material in pellets form
was dried at 75
C for a minimum of 4 h prior to use in a desiccating
dryer.
The third matrix employed has the commercial name Bionolle®
1001, and is based on Poly-butylene succinate, of an average density
of 1.26 g/cm3, kindly provided by Flexopack S.A. (Athens, Greece).
Very common technologies for natural fiber composite mate-
rials are resin transfer molding, vacuum injection molding, struc-
tural reacting injection molding, injection molding, and
compression molding. Regarding PLA-natural fiber composites,
their preparation usually involves a mixing of PLA matrix and fi-
bers, after specific conditions of drying, and a hot pressing pro-
cedure [21]. Another type of manufacturing is the melt mixing
procedure using a twin-screw extruder, or a solution mixing pro-
cedure, followed by a compression molding process hereafter, with
the processing conditions being a subject of research [19,25].
In our work, unlike the most published works so far, the pro-
duction procedure for natural fiber composites adopted, is quite
similar to those for the production of conventional unidirectional
laminate fiber composites. The three types of polymeric matrices
examined, were prepared into thin films of an average thickness of
0.2 mm, by a hot press treatment. Hereafter, following the well-
known film stacking method [16], alternating layers of polymeric
film and flax fibers-tissue were placed and hot pressed for 6 h, at a
pressure of 5 MPa, and at a temperature equal to 100
C for Bionolle
and 124
C for Ecovio® and PLA. According to the followed pro-
cedure, the fiber breakage due to shear stresses imposed during the
melt-mixing procedure, could be avoided, while an optimization of
the fiber orientation could be achieved.
By this method, plates of unidirectional fiber polymer compos-
ites were prepared, at the same average weight fraction 22% and an
average thickness of 1 mm. The corresponding volume fraction was
calculated to be equal to 19.5%. The prepared composites are
designated as BIONflax, ECflax and PLAflax, based on the three
matrices, Bionolle®, Ecovio® and PLA, correspondingly. In addition,
three different methods improving the compatibility of flax fibers
with the matrix were employed. These methods are: A) silaniza-
tion, where the flax fibers are immersed in a 2 wt% Vinyl-Triethoxy-
90 P. Georgiopoulos et al. / Composites Part B 106 (2016) 88e98

Silane (VTES) solution in distilled water for 24 h. Hereafter, fibers
were washed with distilled water and dried at 60
C overnight [41].
It is assumed that by this procedure, the provided hydrocarbon
chains influence the wettability of the fibers and improve their
chemical affinity to the polymer matrix. The effect of moisture on
the materials properties is thus minimized [9]. B) Plasticization of
the fibers. After drying at 100
C for 2 h, a weighed amount of fibers
added to a neat plasticizer, namely tributyl citrate (TBC) and the
mixture heated at 120
C for 8 h. Hereafter, the fibers were cooling
at room temperature, washed with acetone to remove the plasti-
cizer, and dried in a vacuum at room temperature. The amount of
plasticizer present was determined by the weight difference [21].
The plasticizer is expected to have two functions, i.e. to overcome
weakening of fibers as a result of the moisture lost and increase the
ductility of the matrix. C) Treatment with maleic anhydride (MA).
The maleic anhydride (MA) was applied in a 10 wt% with respect to
the fiber weight. The flax fibers were esterified for 25 h with MA,
dissolved in acetone at 50
C, in a solution fiber/solvent ratio equal
to 1:25 wt/volume. Hereafter the flax fibers were washed in cold
acetone and distilled water [42].
The nomenclature of the flax composites, produced as above, is
as follows: MATRIXflax (no treatment), MATRIXflax-Silane,
MATRIXflax-TBC, MATRIXflax-MA for each treatment
correspondingly.
2.2. SEM analysis
2.4. Tensile testing
Tensile measurements were performed with an Instron 1121
type tester, at room temperature. The pure matrices were tested in
the form of dumbbell type specimens, of a gauge length of 20 mm,
and 3 mm wide and the applied crosshead speed was 0.5 mm/min.
This value corresponds to an effective strain rate of 4.16  104 s1.
Regarding the composites, square tabs made from fiberglass were
bonded to the each end of specimens in order to perform tensile
stress experiments. For this type of loading, the primary purpose of
the tabs is to protect the composite material from damage by the
grips that clamp the specimen ends to apply the axial tensile load.
This axial load is introduced into the composite via shear, due to the
frictional forces developed between the grip faces and the surfaces
of the tabs. The tensile specimens were of 50 mm in gage length
and 6 mm in width.
The deformation could be measured very accurately with an
experimental procedure, which is based on a noncontact method
with a laserextensometer, described in detail in a previous work
[43]. Five specimens of each material type were tested, and the
average scatter is presented in Table 2. In addition, tensile testing
of flax fibers has been performed, and the Young's modulus,
tensile strength and elongation at break were found equal to
85 GPa, 700 MPa and 1.6% correspondingly. These elastic prop-
erties are within the values reported in the literature for flax fi-
bers [9].

Scanning Electron Microscopy (SEM) was performed on a 2.5. Flexural testing

NanoSEM 230 (FEI) using an EverharteThornley Detector (ETD). In
order to secure conductivity of the surface for clear imaging, gold
sputtering was applied with a nominal thickness of 7 nm using an
EMS 550X Sputter Coater.
2.3. DSC experiments
Calorimetric measurements (DSC) were carried out using a
Setaram DSC 141 instrument, calibrated with an Indium standard.
Each sample was heated at a constant heating rate of 5
C/min from
20 up to 180
C, and the thermogram was recorded. DSC samples
were taken from the centre and the edge of a 2 mm thick material
sheet. Three specimens of each sample were measured and the
results were averaged and summarized in Table 1.
All composites and the polymeric matrices were subjected to
the same heating rate.
This mode of deformation is very close to the end-applications
of the fiber composites and therefore very important for their
design and optimization. A three-point bending procedure was
applied on the materials studied, and the load-displacement
curve was obtained. The specimens' dimensions were 1 mm in
thickness and 12.7 in width, while the distance between the
supports was 30 mm. The cross-head speed applied was 0.5 mm/
min.
2.6. Dynamic mechanical analysis (DMA)
Dynamic mechanical analysis experiments were performed
using the TA Instruments DMA Q800 instrument. The mode of
deformation applied was the single cantilever beam, and the mean

Table 1
DSC results. Table 2
Tensile and flexural results.
Material Tg (
C) Tm (
C) Tcc (
C) Heat of Crystallinity
fusion DHa (J/g) change (%) Material Tensile Strain Young's Flexural
strength (MPa) at break modulus (GPa) strength (MPa)
BION 114 e 64.10 e
BIONflax 114 e 67.20 4.70 BION 35.0 ± 1.8 0.3 0.5 ± 0.055 47 ± 1.6
BIONflax/Silane 116.4 e 61.96 3.30 BIONflax 119.0 ± 6.6 0.018 12.0 ± 0.72 118 ± 6.9
BIONflax/TBC 114 e 54.86 14.4 BIONflax/silane 114.0 ± 6.7 0.015 9.70 ± 0.5 116 ± 5.4
BIONflax/MA 115 e 64.30 e BIONflax/TBC 113.5 ± 5.6 0.015 13.22 ± 0.83 91 ± 4.3
BIONflax/MA 55.0 ± 2.7 0.005 14.4 ± 0.82 64 ± 3.8
EC 60 146/151 110 4.7 e
ECflax 60 146/151 107 5.8 23.4 EC 18.0 ± 1.0 0.2 0.4 ± 0.06 25 ± 1.2
ECflax/Silane 58 146/153 106 1.51 67 ECflax 95.0 ± 5.1 0.015 8.0 ± 0.6 90 ± 4.8
ECflax/TBC 60 139/147 e 13.0 176 ECflax/silane 95.5 ± 5.2 0.013 13.5 ± 0.69 82 ± 4.9
ECflax/MA 57 146/153 106 8.27 76.0 ECflax/TBC 68.5 ± 4.1 0.010 10.0 ± 0.56 74 ± 4.5
ECflax/MA 50.0 ± 1.8 0.005 25.0 ± 1.5 67 ± 4.0
PLA 58.5 149/155 e 19.5 e
PLAflax 62.0 149/155 e 31.5 61.5 PLA 55.0 ± 2.9 0.07 3.5 ± 0.40 110 ± 6.0
PLAflax/Silane 62.0 147/157 e 25.23 29.3 PLAflax 99.0 ± 4.7 0.016 16.0 ± 0.80 140 ± 6.9
PLAflax/TBC 58.0 141/151 e 30.86 58.2 PLAflax/silane 102.5 ± 5.2 0.015 12.5 ± 0.90 117 ± 5.7
PLAflax/MA 62.0 149/156 e 36.0 84.6 PLAflax/TBC 63.0 ± 4.1 0.010 12.7 ± 0.78 83 ± 4.1
a PLAflax/MA 35.0 ± 1.9 0.003 25.0 ± 1.4 63 ± 3.1
Heat of fusion normalized to pure polymer mass.
 P. Georgiopoulos et al. / Composites Part B 106 (2016) 88e98
dimensions of sample plaques were 12.6 mm in width and 17.5 mm
in length. The temperature range varied from 20 up to 180
C. The
temperature dependent behavior was studied by monitoring
changes in force and phase angle, keeping the amplitude of oscil-
lation constant, and the strain value was 0.1%. The experiments
were performed at a frequency value of 1 Hz, and the heating rate
was 3
C/min. The storage and loss modulus curves versus tem-
perature were then evaluated.
2.7. Creep experiments
Creep experiments have also been performed with TA Q800, at a
single cantilever beam mode of deformation, for a specific time
period equal to 30 min, at various temperatures varying from 37 up
to 180
C, with an interval of 5
C. A constant stress level of 20 MPa
has been applied for all material types, and applying the time-
temperature superposition (TTS) principle, the creep strain curves
were evaluated, for all materials examined.
2.8. Water uptake experiments
Water absorption was determined according to ASTM D570
method. After drying for 24 h, the weight and the thickness of
the samples were measured. Then, they were immersed in
distilled water at room temperature. Each sample was removed
from the water, dried by wiping with blotting paper and
weighed immediately in order to calculate the water uptake.
After each measurement, the samples were immersed in the
water again.
Fig. 1. SEM pictures of fractured surfaces of: (a) BIONflax, (b) BIONflax/MA.
Fig. 2. SEM pictures of fractured surfaces of: (a) ECflax/Silane, (b) ECflax/MA.
91
3. Results and discussion
3.1. SEM fractography analysis
In Figs. 1e3 representative SEM images of fracture surfaces, as
obtained after tensile testing, are depicted for the flax composites
examined. In all pictures the unidirectional orientation of flax fibers
is confirmed. In Fig. 1a, it can be seen that most of the fibers are well
bonded to the BION matrix and those which are broken seem to
quite short, having matrix material on their surfaces. In Fig. 1b,
BIONflax/MA fracture surface is shown, with a number of fibers
covered with substantial amount of polymeric matrix. Further-
more, a number of fibers appear to have a good adhesion with the
matrix, and a lower number of pulled out fibers is detected.
ECflax/Silane image is presented in Fig. 2a. Most of the fibers are
well adhered on the EC matrix, whereas no holes revealing poor
adhesion are shown. In Fig. 2b, ECflax/MA composite's image is
presented. The majority of the fibers are well connected with the
matrix, and are surrounded by polymeric material. The quality of
fiber-PLA matrix connection is shown in Fig. 3a, whereas flax fibers
are covered with PLA matrix and no pull out effect is observed.
Compared with the PLAflax/TBC image in Fig. 3b, a non-
homogeneous fiber distribution is observed, but the fiber-matrix
adhesion seems to be satisfactory.
3.2. DSC results
The DSC results are summarized in Table 1. For the composites
with untreated flax fibers, it is obtained that the glass transition
92 P. Georgiopoulos et al. / Composites Part B 106 (2016) 88e98

Fig. 3. SEM pictures of fractured surfaces of: (a) PLAflax, (b) PLAflax/TBC.

temperature Tg and the melting point Tm of the materials is not

affected by the flax fibers presence, with the exception of the Tg of
Ecflax. The Tg of BION and BION/composites is lower than ambient
temperature and not presented in this table. On the other hand, the
heat of fusion DHm is increased due to fibers, namely 5, 23 and 61%
for BION, EC and PLA correspondingly. This result is an evidence
that the flax fibers act as nucleation regions and enhance crystal-
linity formation into the bulk polymer. During heating, EC and
ECflax exhibit a cold-crystallization temperature (Tcc), not shown
by the other material types. When cold-crystallization occurs, less
perfect crystals are formed, which are driven to melting at higher
temperatures. When flax fibers are subjected to surface treatment,
as described in the previous paragraphs, the following results are
obtained: Glass transition temperature generally is not affected,
with the exception of PLAflax/TBC, where Tg is lower than that of
pure matrix. The same effect is observed for the melting temper-
ature Tm, which exhibits about 8 to 4
lowering for PLAflax/TBC,
due to the plasticization occurred by the presence of TBC. Regarding
melting region, the composites based on PLA and EC exhibit two
melting points, (Tm1 and Tm2) between 141 and 157
C, for each
type of flax fibers treatment. This effect was previously obtained for Fig. 4. Tensile Young's modulus of all flax fiber composites studied.
composite materials based on EC [44].
The first melting peak Tm1 is constant (146
C) for all ECflax
composites, with the exception of ECflax/TBC which is 139
C, due
to the presence of plasticizer. Tm1 is related to the fusion of
imperfect crystallites, formed during cold crystallization upon
heating, and Tm2 related to the fusion of new crystallites formed
through the melt-recrystallization process [44].
Following Table 1, EC and ECflax composites exhibit a cold
crystallization region Tcc around 106
C, characterized by a cold
crystallization enthalpy DHcc. The lowering of Tcc for all ECflax
composites with respect to pure EC, denotes that flax fibers pro-
mote kinetics and extend crystallization on heating [44]. The
overall heat of fusion DH, which is the difference (DHm  DHcc),
normalized to the weight of pure polymeric matrix, is shown in
Table 1. From these values, the percentage crystallinity change for
all composites studied, was evaluated and presented in this table.
The highest crystallinity increment (about 176%) is obtained for
ECflax/TBC, followed by an 84% increment for PLAflax/MA, and a
61% increment for PLAflax composites. This trend is not observed in
the composites based on BION matrix.

3.3. Tensile and flexural results

The tensile and flexural properties of the matrices and the

composites, are summarized in Table 2, and Figs. 4e7. From these Fig. 5. Tensile strength of all flax fiber composites studied.
 P. Georgiopoulos et al. / Composites Part B 106 (2016) 88e98
Fig. 6. Flexural modulus of all flax fiber composites studied.
Fig. 7. Flexural strength of all flax fiber composites studied.
results it is revealed that regarding Young's modulus, and flax fiber
composites with no fiber treatment, the highest increment is ob-
tained for BIONflax, while substantial increment is also obtained for
ECflax and PLAflax. This degree of enhancement is comparable to
that of conventional polymer-glass/carbon reinforced composites,
where the fiber volume fraction is usually three times higher than
that in the present work. Similar behavior has been reported in
Ref. [22], where Mater-Bi® as a matrix was reinforced by unidi-
rectional flax fibers. Among the three types of composites with no
fiber treatment examined in the present study, PLAflax attains the
highest stiffness (16 GPa), but not the highest percentage incre-
ment, while its tensile strength is lower than that of BIONflax.
Given that PLA matrix has better elastic properties than Ecovio®
and Bionolle®, it is extracted that the PLA matrix-flax fiber adhesion
is not adequate, since the tensile strength is strongly affected by the
quality of interphase between fiber and matrix [45]. However,
compared to previous studies [15] where the natural fibers are
randomly dispersed in the PLA matrix, a higher mechanical
enhancement of PLA with unidirectional flax fibers, is obtained in
93
the present work. After the fibers treatment, the highest percentage
increment of Young's modulus is exhibited by ECflax composites,
with the ECflax/MA materials showing a modulus over 60 times
higher than that of the matrix. In a similar way, for BIONflax and
PLAflax composites the highest modulus is observed after the
treatment with maleic anhydride, revealing this way that MA im-
proves the stiffness of the composites, whereas this effect is more
intense for PLA and EC composites. On the other hand, strain at
break is reduced for all flax-fiber composites, due to the presence of
fibers, which make the materials more elastic, whereas the most
dramatic lowering is obtained for the MA-treated materials. This is
in accordance with the lowest tensile strength exhibited by these
composites. Young's modulus is measured at low strains, when
stress transfer from the matrix to the fibers through the interface is
less critical than in the subsequent stages in the nonlinear regime of
the stress-strain curve, where final breakage takes place [38].
Therefore, the improved stiffness is not always an indication of a
better fiber-matrix adhesion, if the material is driven to failure at a
low stress. Combining the tensile results, it can be extracted that
the highest enhancement is obtained for BIONflax, followed by
BIONflax/TBC, ECflax/Silane, PLAflax and PLAflax/Silane. The
various polymeric matrices employed, seem to collaborate in a
different way with the flax fibers (treated or no-treated). These
remarks appear to be in accordance with the above discussed SEM
images.
The flexural strength of the fiber composites in comparison to
the one of the pure matrix is shown in Table 2. For untreated fibers,
a high improvement is obtained for the bending strength of flax
composites, compared to the matrice's one. Again, PLAflax attains
the highest (nominal) bending strength, but the percentage incre-
ment is of the order of 27%, while ECflax and BIONflax appear to
have an improvement of the order of 260 and 150% correspond-
ingly. BIONflax, BIONflax/Silane appear to have almost the same
flexural strength, and the same is observed for ECflax and ECflax/
silane, whereas PLAflax exhibits the highest flexural strength. In
conclusion, with the exception of silane, the fiber-treatment was
found to have a detrimental effect on the flexural strength of the
composites.
As it is mentioned in Refs. [13e15], the differences in chemistry
between fibers and matrix play an important role to the mechanical
performance of the composites. The use of a compatibilizer should
improve the adhesion quality, but it is not known what kind of
damage should cause to the matrix material. It has been reported
that for BION matrix [13,46] and PLA matrix [15] the tensile
strength could not be improved. This effect may be attributed to the
mixing procedure followed, due to which the fibers are short and
not oriented as in the present study. Regarding PLA, its inherent
brittleness prevents the adhesion of the natural fiber. This fact is
inversed when a plasticizer is employed in the mixing procedure
[15], while the interpretation of silane effectiveness on stiffness
improvement is not so clear [38]. Regarding unidirectional natural
fiber composites, there is no adequate information so far about the
fiber modification on the composite's mechanical performance.
In order to evaluate the different reinforcing mechanisms in the
composites studied, a set of well known semi-empirical equations
developed by Halpin and Tsai [47] for the Young's modulus were
employed, such as the rule of mixture for a stress imposed parallel
to the fibers direction (parallel spring model):
 
Ec ¼ Vf Ef þ 1  Vf Em (1)
Where Ec, Ef and Em are the Young's moduli of the composite, the
fiber and the matrix correspondingly and Vf is the fiber volume
fraction. Eq. (1) is the upper bound of a composite's modulus.
94 P. Georgiopoulos et al. / Composites Part B 106 (2016) 88e98

An improved version of Eq. (1) is given by the modified
expression [48,49]:
 
Ec ¼ kVf Ef þ 1  Vf Em
Where k is a parameter related to the quality of fiber-matrix
adhesion.
The general form of the longitudinal modulus, developed by the
Halpin eTsai equation is given by:
 
1 þ zhVf
Ec ¼ Em
1  hVf
Ef
1
With h ¼
Em
Ef
Em þz
Where z is a measure of reinforcement geometry, which depends
upon loading conditions and geometries of the inclusion, that is,
the aspect ratio in the case of fiber [47] and the limiting values of z
are zero and Infinity. When parameter z takes the value of zero as a
lower bound, Eq. (3) gives:
1 V Vm
¼ fþ
Ec Ef Em
Where Vm is the matrix volume fraction and Eq. (5) represents the
series connected model, which is the lower bound of a composite
modulus. When z tends to infinity, the Halpin eTsai Eq. (3) reduces
to Eq. (1).
In Table 3 the experimental data of Young's modulus are pre-
sented in comparison with the rule of mixtures (Eq. (1)), and the
Halpin-Tsai (Eq. (3)). Eq. (3) does not work for the composites
which exhibit higher elastic stiffness enhancement than that of the
upper bound (Eq. (1)), namely ECflax/MA and PLAflax/MA com-
posites. From Table 3, as it was expected, the experimental data
have a deviation from the rule of mixtures, given that this equation
assumes perfect adhesion between fibers and matrix and neglects
the interactions due to the difference in their Poisson ratios. The
lowest deviation between experiment and rule of mixtures is
exhibited by the BIONflax/MA, ECflaxSilane and PLAflax compos-
ites. The values for PLAflax/MA and ECflax/MA are not included,
given that the Young's modulus of these materials is higher than
that by the rule of mixtures. Regarding Eq. (3), it is observed that
the highest parameter z and the lowest parameter h values are
referring to the BIONflax/MA, ECflax/Silane and PLAflax compos-
ites, which denotes that the fiber effectiveness is improved for
these material types, and the physical meaning of parameter z is
(2) that it is regarded as a reinforcement measure. This result is in
accordance with the tensile and flexural experimental data dis-
cussed in the above paragraph.
3.4. DMA results
From the storage modulus variation versus temperature
(Fig. 8), the mechanical reinforcement, observed in tensile testing,
(3) is obvious for all fiber composites studied. For the PLAflax, the
transition region at about the same temperature domain 60e75
C
with PLA matrix, appears to be smoother, and the material flows
above 125
C. The storage of the PLA matrix at high temperatures
(4) exhibits some increment at the region of 100
C. This effect is
attributed to the cold crystallization temperature, that PLA ap-
pears to have at the same temperature and the consequent crys-
tallinity increment of PLA matrix [50]. In an earlier work [15], an
analogous step at the storage modulus curve of PLAflax at 80
C
was found, which was attributed to the cold crystallization tem-
perature of PLAflax. In the present work such an effect has not
been obtained, however the PLAflax remains mechanically active
up to 130
C. This is due to the crystallinity increment of PLAflax of
the order of 62%, as extracted by the DSC results. The transition
(5)
region of the composite is slightly shifted to lower temperature
range, and hereafter, a “leathery” plateau is observed between 80
and 100
C. Similar behavior is shown for ECflax ECflax/Silane and
ECflax/MA (Fig. 8). This behavior can be attributed to both, the
increased crystallinity of ECflax composites (with the exception of
ECflax/Silane), and the very good adhesion established between
matrix and fibers and the consequent ability of load transfer from
matrix to fibers, as it is also found from Young's modulus incre-
ment for ECflax. These two reasons improve the materials stiffness
and stability. In addition, this strengthening effect could be con-
nected with the cold-crystallization of ECflax at about 106
C
obtained in the DSC experiments. For both composites, PLAflax
and ECflax it can be summarized that they are mechanically active,
due to the flax fibers, almost above 100
C, while both unrein-
forced matrices are getting weak above 80
C. The storage
modulus of ECflax/TBC composite exhibits a gradually decreasing
trend, with no transition apparent, probably to the plasticization
effect of TBC.
As far as BION matrix is concerned, with its Tg below ambient
temperature, the BIONflax materials are more stable at the tem-
perature region examined, and the flow region of the composite

Table 3
Model parameters for the Young's modulus of the composites.

Material Young's modulus (exp) (GPa) Young's modulus by Eq. (1) (GPa) Halpin-Tsai Parameters

z h
BIONflax 12 17.4 286 0.37
BIONflax/silane 9.70 17.4 160 0.51
BIONflax/TBC 13.2 17.4 410 0.29
BIONflax/MA 14.4 17.4 625 0.21

ECflax 8 17.3 137 0.60

ECflax/silane 13.5 17.3 579 0.27
ECflax/TBC 10.0 17.3 221 0.48
ECflax/MA 25.0 17.3 e e

PLAflax 16.0 19.3 60 0.27

PLAflax/silane 12.5 19.3 22 0.50
PLAflax/TBC 12.7 19.3 23 0.49
PLAflax/MA 25.0 19.3 e e
 P. Georgiopoulos et al. / Composites Part B 106 (2016) 88e98
Fig. 8. Storage modulus of all flax fiber composites studied: (a) PLAflax composites, (b) ECflax composites, (c) BIONflax composites.
occurs at about 105
C. This temperature range is quite lower than
that for the PLAflax and ECflax, and this is due to the low glass
transition temperature of the Bionolle matrix. By summarizing the
storage modulus/temperature variation of the three composites, it
is revealed that best improvement was achieved for PLAflax
composite.
Regarding the effect of fiber treatment on the storage modulus
variation, an improved enhancement of the maleated and silane
treated flax fibers is obvious, for all material types. EC and PLA fiber
composites appear to have two transition regions, one in the range
of 70e80
C and the other at 130
C. In general, the flax fibers
pretreatment, causes a surface modification increasing the surface
energy of the fibers, provides better wettability, and thus enhances
the polymer-fiber adhesion, due to chemical bonding formation
with the matrix [41].
The storage modulus increment with fiber treatment is a
stronger evidence of mechanical enhancement, given that in tensile
experiments, stiffness can be improved, even without good inter-
facial adhesion. This is because modulus is measured at low strains,
where stress transfer is less crucial.
3.5. Creep results
Creep strain of all materials studied has been evaluated for a
time and temperature range, not presented here. Hereafter, the
master curves of the creep strain, obtained by applying the time-
temperature superposition (TTS) principle in the creep-
95
temperature experiments at a reference temperature of 37
C,
were constructed and depicted in Figs. 9e11, exhibiting a good
overlap of the isothermal curves. For reasons of clarity, the creep
strain experimental data of the pure polymers are not presented
here, due to the high strain difference between matrices and
composites. A substantial decrease of the creep strain with fibers is
obtained, meaning that creep resistance was enhanced by the
addition of fillers, and revealing the solid-like behavior of the
composites. By comparing the master curves of the three types of
untreated fiber composites, the best creep resistance (lower creep
strain) is exhibited by BIONflax, followed by PLAflax, while ECflax
appears an incremental trend of creep strain at longer times. This
may be related to an inadequate stress transfer from the matrix to
the fibers in the case of ECflax. In Fig. 9 the creep strain data of
BION/flax composites are illustrated, where the BIONflax/MA
composite exhibiting the higher creep resistance, reveals a more
elastic material, in accordance with the mechanical improvement,
shown in the previous paragraphs. The BIONflax, BIONflax/Silane
and BIONflax/TBC composites show a similar creep response, with
no essential differences. In Figs. 10 and 11, for the ECflax and PLAflax
composites a quite similar trend is obtained, with the ECflax/MA
being the more solid-like material having the lowest creep strain
value.
3.6. Water uptake results
The water absorption WA was calculated by Eq. (6):
96 P. Georgiopoulos et al. / Composites Part B 106 (2016) 88e98

MðtÞ  M0
WAð%Þ ¼  100% (6)
M0

where M(t) is the sample weight at time t, and M0 the initial weight.
Figs. 12 and 13 show the variation of the water absorbance as a
function of square root of immersion time for Bionflax and PLAflax
composites representatively. Water content increased with time
due to the high hygroscopicity of flax fibers. A rapid water absor-
bance during the first hours of immersion is observed, until a stage
where a plateau (or very lower increment rate) is reached,
revealing a Fickian behavior [44]. However, not all samples have
reached a saturated moisture level. The percentage water uptake is
higher for PLAflax composites (Fig. 13), compared to that of Bionflax
(Fig. 12) composites. Regarding the effect of flax fibers treatment on
the water uptake, the trend is that the fiber-treated composites

Fig. 9. Creep master curve at a reference temperature of 37
C for BIONflax

composites.

Fig. 12. Water uptake as a function of square root of immersion time for BIONflax
composites.

Fig. 10. Creep master curve at a reference temperature of 37
C for ECflax composites.

Fig. 13. Water uptake as a function of square root of immersion time for PLAflax
Fig. 11. Creep master curve at a reference temperature of 37
C for PLAflax composites. composites.
 P. Georgiopoulos et al. / Composites Part B 106 (2016) 88e98
exhibit a lower water content. In general, strong intermolecular
fiber-matrix bonding decreases the rate of water absorbance of the
material [41]. For improved adhesion between matrix and fibers,
the rate of water uptake is expected to decrease since there are
fewer gaps in the interfacial region, while more hydrophilic groups
may be blocked by the coupling effect [51,52]. For PLA/flax com-
posites, after the initial rapid water absorbance, the slope of water
uptake is lower for PLAflax/TBC and PLAflax/MA, whereas it is
almost the same for PLAflax and PLAflax/Silane. Also, in absolute
values, the water uptake is lower for the PLAflax/TBC and PLAflax/
MA treated samples, compared to PLAflax and PLAflax/Silane. This
effect reveals a better fiber-matrix adhesion for maleated and
plasticized flax fibers.
4. Conclusions
The evaluation of the effectiveness of three different biode-
gradable polymers as matrix materials for unidirectional flax fiber
composites, prepared with the “film stacking method” has been
extensively studied. Due to the hydrophilic nature of flax fibers and
the hydrophobic one of the polymeric structure, poor fiberematrix
adhesion may occur, reducing the mechanical enhancement of the
composites. Therefore, three different chemical treatments of flax
fibers, namely silanization, plasticization and treatment with ma-
leic anhydride were applied, and the produced flax fiber compos-
ites were investigated. A significant mechanical enhancement was
obtained with the presence of flax fibers, while hereafter,
depending on the surface modification type, Young's modulus,
flexural modulus, as well as storage modulus was found to be
further increased. These results were supported by the SEM photos,
where good adhesion between fibers and matrix has been detected,
related with a lower number of pulled out fibers. Another evidence
for the mechanical enhancement is the crystallinity increment
observed in some material types. Regarding Young's modulus, its
values varied between 8 GPa for ECflax up to 25 GPa for ECflax/MA
and PLAflax/MA, whereas the highest tensile strength was
observed in the untreated flax fiber composites, followed by the
Silane and TBC treated ones. A similar response was obtained by the
flexural experiments.
In addition, following the creep strain data, it can be concluded
that all the flax composites, especially these with surface treated
flax fibers, appear to have a substantial resistance to creep, and
therefore reveal as promising materials for load-bearing applica-
tions at long time periods.
Regarding the water uptake, the highest moisture levels were
exhibited by the PLAflax composites, whereas the maleated and
plasticized flax fiber/composites had the lowest percentage water
content, revealing this way an improved fiber-matrix adhesion.
References
[1] Yan L, Chouw N, Yuan X. Improving the mechanical properties of natural fibre
fabric reinforced epoxy composites by alkali treatment. J Reinf Plast Compos
2012;31(6):425e37.
[2] John MJ, Thomas S. Biofibres and biocomposites. Carbohydr Polym
2008;71(3):343e64.
[3] Yan L, Chouw N, Jayaraman K. Flax fibre and its composites - a review. Compos
B 2014;56:296e317.
[4] Faruk O, Bledzki AK, Fink HP, Sain M. Biocomposites reinforced with natural
fibers: 2000e2010. Prog Polym Sci 2012;37(11):1552e96.
[5] Joshi SV, Drzal LT, Mohanty AK, Arora S. Are natural fiber composites envi-
ronmentally superior to glass fiber reinforced composites? Compos A
2004;35(3):371e6.
[6] Kalia S, Kaith BS, Kaur I. Pretreatments of natural fibers and their application
as reinforcing material in polymer composites e a review. Polym Eng Sci
2009;49(7):1253e72.
[7] Sanadi AR, Prasad SV, Rohatgi PK. Sunhemp fibre-reinforced polyester. J Mater
Sci 1986;21(12):4299e304.
[8] Nevell TP, Zeronian SH. Cellulose chemistry and its applications. New York:
97
John Wiley and Sons; 1985.
[9] Bogoeva-Gaceva G, Avella M, Malinconico M, Buzarovska A, Grozdanov A,
Gentile G, et al. Natural fiber eco-composites. Polym Compos 2007;28(1):
98e107.
[10] Li XH, Meng YZ, Wang SJ, Varada Rajulu A, Tjong SC. Completely biodegrad-
able composites of poly(propylene carbonate) and short, lignocellulose fiber
Hidegardia populifolia. J Polym Sci B 2004;42(4):666e75.
[11] Barkoula NM, Garkhail SK, Peijs T. Biodegradable composites based on flax/
polyhydroxybutyrate and its copolymer with hydroxyvalerate. Ind Crops Prod
2010;31(1):34e42.
[12] Mohanty AK, Misra M, Drzal LT, Selke SE, Harte BR, Hinrichsen G. Natural fi-
bers, biopolymers, and biocomposites: an introduction. In: Mohanty AK,
Misra M, Drzal LT, editors. Natural fibers, biopolymers, and biocomposites.
Boca Raton, FL, USA: CRC Press; 2005.
[13] Baiardo M, Zini E, Scandola M. Flax fibre-polyester composites. Compos A
2004;35(6):703e10.
[14] Shanks RA, Hodzic A, Wong S. Thermoplastic biopolyester natural fiber
composites. J Appl Polym Sci 2004;91(4):2114e21.
[15] Oksman K, Skrifvars M, Selin JF. Natural fibres as reinforcement in polylactic
acid (PLA) composites. Compos Sci Technol 2003;63(9):1317e24.
[16] Van de Weyenberg I, Chi Truong T, Vangrimde B, Verpoest I. Improving the
properties of UD flax fibre reinforced composites by applying an alkaline fibre
treatment. Compos A 2006;37(9):1368e76.
[17] Wong S, Shanks RA, Hodzic A. Effect of additives on the interfacial strength of
poly(l-lactic acid) and poly(3-hydroxy butyric acid)-flax fibre composites.
Compos Sci Technol 2007;67(11e12):2478e84.
[18] Lee SG, Choi SS, Park WH, Cho D. Characterization of surface modified flax
fibers and their biocomposites with PHB. Macromol Symp 2003;197(1):
89e100.
[19] Bax B, Müssig J. Impact and tensile properties of PLA/Cordenka and PLA/flax
composites. Compos Sci Technol 2008;68(7e8):1601e7.
[20] Shanks RA, Hodzic A, Ridderhof D. Composites of poly(lactic acid) with flax
fibers modified by interstitial polymerization. J Appl Polym Sci 2006;99(5):
2305e13.
[21] Wong S, Shanks RA, Hodzic A. Poly(L-lactic acid) composites with flax fibers
modified by plasticizer absorption. Polym Eng Sci 2003;43(9):1566e75.
[22] Romha
ny G, Kocsis JK, Cziga
ny T. Tensile fracture and failure behavior of
thermoplastic starch with unidirectional and cross-ply flax fiber re-
inforcements. Macromol Mater Eng 2003;288(9):699e707.
[23] Ma X, Yu J, Kennedy JF. Studies on the properties of natural fibers-reinforced
thermoplastic starch composites. Carbohydr Polym 2005;62(1):19e24.
[24] Ochi S. Mechanical properties of kenaf fibers and kenaf/PLA composites. Mech
Mater 2008;40(4e5):446e52.
[25] Yussuf AA, Massoumi I, Hassan A. Comparison of polylactic acid/kenaf and
polylactic acid/rise husk composites: the influence of the natural fibers on the
mechanical, thermal and biodegradability properties. J Polym Environ
2010;18(3):422e9.
[26] Nishino T, Hirao K, Kotera M, Nakamae K, Inagaki H. Kenaf reinforced
biodegradable composite. Compos Sci Technol 2003;63(9):1281e6.
[27] Lee SH, Wang S. Biodegradable polymers bamboo fiber biocomposite with
bio-based coupling agent. Compos A 2006;37(1):80e91.
[28] Sawpan MA, Pickering KL, Fernyhough A. Hemp fibre reinforced poly(lactic
acid) composites. Adv Mater Res 2007;29e30:337e40.
[29] Cho D, Seo JM, Lee HS, Cho CW, Han SO, Park WH. Property improvement of
natural-fiber reinforced green composites by water treatment. Adv Compos
Mater 2007;16(4):299e314.
[30] Plackett D. Maleated polylactide as an interfacial compatibilizier in bio-
composites. J Polym Environ 2004;12(3):131e8.
[31] Bledzki AK, Fink HP, Specht K. Unidirectional hemp and flax EP- and PP-
composites: influence of defined fiber treatments. J Appl Polym Sci
2004;93(5):2150e7. http://dx.doi.org/10.1002/app.21131.
[32] Madsen B, Lilholt H. Physical and mechanical properties of unidirectional
plant fibre composites-an evaluation of the influence of porosity. Compos Sci
Technol 2003;63(9):1265e72.
[33] Khondker OA, Ishiaku US, Nakai A, Hamada H. A novel processing technique
for thermoplastic manufacturing of unidirectional composites reinforced with
jute yarns. Compos A 2006;37(12):2274e84.
[34] Rong MZ, Zhang MQ, Liu Y, Yang GC, Zeng HM. The effect of fiber treatment on
the mechanical properties of unidirectional sisal-reinforced epoxy compos-
ites. Compos Sci Technol 2001;61(10):1437e47.
[35] Oksman K, Wallstro € m L, Berglund LA, Filho RDT. Morphology and mechanical
properties of unidirectional sisal-epoxy composites. J Appl Polym Sci
2002;84(13):2358e65. http://dx.doi.org/10.1002/app.10475.
[36] Ratna Prasad AV, Rao KM. Mechanical properties of natural fibre reinforced
polyester composites: jowar, sisal and bamboo. Mater Des 2011;32(8e9):
4658e63.
[37] Ochi S. Development of high strength biodegradable composites using Manila
hemp fiber and starch-based biodegradable resin. Compos A 2006;37(11):
1879e83.
[38] Sobczak L, Brüggemann O, Putz RF. Polyolefin composites with natural fibers
and wood-modification of the fiber/fillerematrix interaction. J Appl Polym Sci
2013. http://dx.doi.org/10.1002/APP.36935.
[39] Alix S, Philippe E, Bessadok A, Lebrun L, Morvan C, Marais S. Effect of chemical
treatment on water sorption and mechanical properties of flax fibres. Bio-
resour Technol 2009;100(20):4742e9.
98 P. Georgiopoulos et al. / Composites Part B 106 (2016) 88e98
[40] Shah DU, Schubel PJ, Licence P, Clifford MJ. Hydroxyethylcellulose surface
treatment of natural fibres: the new ‘twist’ in yarn preparation and optimi-
zation for composites applicability. J Mater Sci 2012;47(6):2700e11.
[41] Wang B, Panigrahi S, Crevar W, Tabil L. Application of pretreated flax fibers in
composites. In: CSAE/SCGR 2003 Meeting, Montre
al Que
bec; July 6e9, 2003.
[42] Cantero G, Arbelaiz A, Liano-Ponte R, Mondragon I. Effects of fibre treatment
on wettability and mechanical behaviour of flax/polypropylene composites.
Compos Sci Technol 2003;63:1247e54.
[43] Georgiopoulos P, Kontou E, Niaounakis M. Thermomechanical properties and
rheological behavior of biodegradable composites. Polym Compos 2014;35(6):
1140e9.
[44] Georgiopoulos P, Kontou E. The effect of wood-fiber type on the thermo-
mechanical performance of a biodegradable polymer matrix. J Appl Polym Sci
2015;132(27):42185. http://dx.doi.org/10.1002/app.42185.
[45] Avella M, Buzarovska A, Errico ME, Gentile G, Grozdanov A. Eco-challenges of
bio-based polymer composites. Materials 2009;2(3):911e25.
[46] Wollerdorfer M, Bader H. Influence of natural fibres on the mechanical
properties of biodegradable polymers. Ind Crops Prod 1998;8(2):105e12.
[47] Halpin JC, Kardos JL. The Halpin-Tsai equations. A review. Polym Eng Sci
1976;16(5):344e52.
[48] Cox LH. The elasticity and strength of paper and other fibrous materials. Br J
Appl Phys 1952;3:72.
[49] Krenchel H. Fibre reinforcement. Copenhagen: Akademisk Forlag; 1964.
[50] Georgiopoulos P, Kontou E, Meristoudi A, Pispas S, Chatzinikolaidou M. The
effect of silica nanoparticles on the thermomechanical properties and
degradation behavior of polylactic acid. J Biomater Appl 2014;29(5):662e74.
[51] Arbelaiz A, Fernandez B, Ramos JA, Retegi A, Llano-Ponte R. Mechanical
properties of short flax fibre bundle/polypropylene composites: influence of
matrix/fibre modification, fibre content, water uptake and recycling. Compos
Sci Technol 2005;65:1582e92.
[52] Espert A, Vilaplana F, Karlsson S. Comparison of water absorption in natural
cellulosic fibres from wood and one-year crops in polypropylene composites
and its influence on their mechanical properties. Compos Part A 2004;35:
1267e76.