See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/267630751

Acoustic emission characterization of damage in short hemp-fiber reinforced

polypropylene composites

Article in Polymer Composites · October 2014

DOI: 10.1002/pc.23272

CITATIONS READS

43 289

5 authors, including:

Mustapha Assarar Daniel Scida

Université de Reims Champagne-Ardenne Université de Reims Champagne-Ardenne
54 PUBLICATIONS 2,270 CITATIONS 40 PUBLICATIONS 1,611 CITATIONS

SEE PROFILE SEE PROFILE

Wajdi Zouari El Hadi Saidane

Université de Reims Champagne-Ardenne Université de Haute-Alsace
36 PUBLICATIONS 660 CITATIONS 15 PUBLICATIONS 337 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Ferroelectric shell finite element View project

Finite elements with rotational DOFs View project

All content following this page was uploaded by Wajdi Zouari on 12 January 2018.

The user has requested enhancement of the downloaded file.

Acoustic Emission Characterization of Damage in Short
Hemp-Fiber-Reinforced Polypropylene Composites

Mustapha Assarar, Daniel Scida, Wajdi Zouari, El Hadi Saidane, Rezak Ayad
University of Reims Champagne-Ardenne, Laboratory of Materials Science and Engineering (LISM, EA4596),
IUT de Troyes, 9 rue de Quebec, 10026 Troyes Cedex, France

This article aims at investigating the effects of hygro- advantage of design flexibility and exhibit interesting
thermal aging on the damage mechanisms of short recycling properties [4–6]. Unfortunately, natural fibers
white Hemp Fiber Reinforced Polypropylene (HFRP)
composites with various fiber contents (10, 20, 30, and
still have several problems that restrain their use in sev-
40 wt%). Injected molded specimens were subjected eral domains. In fact, they have hydrophilic character and
to hygrothermal aging with a relative humidity of 80% present broad dispersion in physical and mechanical prop-
and two temperatures, 25 and 50 C. The water absorp- erties [7]. Furthermore, the use of vegetal fiber reinforced
tion and its effect on tensile properties of HFRP com- composites for structural applications requires additional
posites were investigated. The Acoustic Emission (AE)
information about the evolution of damage mechanisms
technique combined with scanning electron micros-
copy observations was used to identify microstructural within these materials, in particular during their aging.
damage events leading to overall failure of the HFRP Consequently, many research works have been carried out
composites. This identification according to hemp- to investigate the performance of vegetal fiber composites
fiber content and hygrothermal aging was made with [3], in particular in a wet environment, because the prob-
an unsupervised method based on a statistical multi-
lems of environmental aging are essential for the develop-
variable analysis (k-means algorithm). The AE results
indicate that the quality of fiber-matrix interface plays ment of bio-composites. These research investigations
a major role in the damage process of HFRP compo- have shown that the exposure of vegetal fiber composites
sites, shown by the number of AE signals induced by to a hygrothermal environment leads to a decrease of
the interface failure and their amplitude ranges. their mechanical properties when water spreads in the
POLYM. COMPOS., 37:1101–1112, 2016. V C 2014 Society of
material [8–12]. For example, in the work of Dhakal
Plastics Engineers
et al. [13], it is shown that in the case of hemp fiber rein-
forced polyester composites, the moisture uptake
INTRODUCTION increased with the fiber volume fraction. They also
showed that moisture induced a significant degradation of
Nowadays, environmental problems are much more their composites mechanical properties at elevated tem-
visible and clearly threatening our world equilibrium. perature. Moreover, water absorption, mechanical, and
Therefore, new regulations have been implemented to thermal properties of short hemp/glass fiber reinforced
reduce the environmental footprint of the polluting indus- polypropylene hybrid composites were investigated in
tries [1]. For example, in manufacturing industries, such Ref. [14]. The authors have observed that thermal proper-
as automotive, construction and packaging, the use of nat- ties and resistance to water absorption properties of the
ural fibers instead of synthetic ones, presents an interest- hemp fiber composites were improved by hybridization
ing solution to respect these regulations. In fact, vegetal with glass fibers.
fibers are biodegradable; their resources are renewable With the aim of characterizing failure in vegetal fiber
and are considered neutral toward CO2 emissions [2]. reinforced composites, several studies have correlated the
Moreover, vegetal fibers are one of the cheapest and eas- Acoustic Emission (AE) parameters to the various dam-
ily available fibers and their use as reinforcement within age modes and their chronology during the failure
thermoplastic composites has gained a considerable [15–19]. De Rosa et al. [20] used AE to evaluate the
attraction, in particular, for automotive products [3]. Even effect of damage dissipation in jute/glass hybrid lami-
if they are low in strength performance, they have the nates. The level of impact loading and the residual
strength of the composites were evaluated by correlating
the AE activity to the applied cyclic stress on impacted
Correspondence to: M. Assarar; e-mail: [email protected]
DOI 10.1002/pc.23272
specimens at various energies. Dogossy and Czigany [21]
Published online in Wiley Online Library (wileyonlinelibrary.com). used the amplitude of AE signals as parameter, to corre-
C 2014 Society of Plastics Engineers
V late the acoustic patterns with the fracture modes in maize

POLYMER COMPOSITES—2016
 FIG. 1. Water uptake of HFRP composites according to aging time including standard deviations: (a) Aging
1: 80% RH and 25 C, (b) Aging 2: 80% RH and 50 C. [Color figure can be viewed in the online issue,
which is available at wileyonlinelibrary.com.]

hull filled polyethylene composites. These authors have fracture and failure behavior of jute cloth reinforced ther-
distinguished three failure modes within bio-composites moplastic copolyester composites shown in Ref. [24].
they studied: matrix deformation (below 25 dB), maize These research works indicate that characterization of
pull-out (26–40 dB) and maize breakage (over 41 dB). In damage in vegetal fiber composites is still difficult,
a similar work, Czigany [22] correlated the AE parame- because, in addition to their anisotropic behavior, vegetal
ters, such as amplitude and energy to characterize the fibers present a large variability in geometry, morphol-
evolution of damage over different composites. This cor- ogy, and properties. Consequently, damage classifications
relation has allowed characterizing the following damage based only on one parameter of AE signals such as ampli-
mechanisms at different range of loading: matrix defor- tude, energy and duration, could present overlaps between
mation at the beginning (21–35 dB), delamination (35–45 the detected AE signals as shown in Ref. 25.
dB), fiber pull-out (45–60 dB), and ultimately fiber break- To better discriminate the damage modes, several
age (over 60 dB). Anuar et al. [23] emphasized that sig- supervised and unsupervised analysis methods have been
nal propagation as well as its shape depend on several already applied to traditional composites [26–29], in order
parameters such as material properties, dimensions of to avoid the inaccuracy of mono parametric classifica-
specimens and distance between source and sensor. Con- tions. For example, Sause et al. [29] have considered an
sequently, they have proven for their studied specimens unsupervised technique to identify damage within nickel-
that AE amplitude is an essential parameter to identify copper-coated and uncoated carbon fiber reinforced plas-
the different failure modes as well as their chronology in tic (CFRP) specimens, used in coated CFRP tank systems.
the case of untreated and treated kenaf-fiber reinforced In particular, the most important AE signals originating
composites. The AE amplitude was also used to study the from coating failure (crack formation and crack growth

FIG. 2. Water uptake of HFRP composites according to the hemp fiber content (wt%): (a) Aging 1: 80%
RH and 25 C, (b) Aging 2: 80% RH and 50 C.

1102 POLYMER COMPOSITES—2016 DOI 10.1002/pc

 nals, obtained from the multivariable method, is based on
some literature results and SEM observations of fracture
surfaces.

MATERIALS AND EXPERIMENTAL PROCEDURE

Materials
The tested materials are constituted of short hemp-
fibers as reinforcements of a Polypropylene (PP) matrix.
They are distinguished by the weight ratio of hemp fiber
(10, 20, 30, and 40 wt%). All specimens were provided
by the AFT Plasturgie Company (Fontaine-Les-Dijon–
France). They were individually elaborated using the
injection molding process with a two-cavity mold. Tem-
FIG. 3. Stress–strain response of unaged HFRP composites under mon-
otonic tensile tests. [Color figure can be viewed in the online issue,
peratures of 185 and 25 C were used for melt and mold,
which is available at wileyonlinelibrary.com.] respectively. The Hemp fibers have undergone the retting
process and a mechanical fiber separation. After that, the
within the liner and liner delamination) have been deter- fibers were treated by a confidential chemical treatment,
mined by comparison between coated and uncoated speci- and no coupling agent was added during the injection
mens. In another work, Sause and Horn have proposed in molding of the samples.
Ref. [30] finite element simulations of AE signals origi- The observations by image analysis on about 40 fibers
nating from failure of CFRP structures. The signals of the inside several HFRP samples, showed that hemp fibers
most important damage mechanisms of fiber reinforced have a length of 1 6 0.2 mm, external diameter of 80 6
composites (matrix cracking, fiber breakage, and fiber- 30 lm and inner diameter of 32 6 24 lm. On the other
matrix interface failure) have been numerically simulated hand, from microscopic observations and a study by
and have shown a good agreement with experimental image analysis on similar composite, in this case a short
waveforms. glass FRP composite obtained using injection molding
Although they are useful, these methods have not been process, Bonnet [32] showed that the final product
extensively considered to discriminate the damage events presents a Heart-Skin structure. On skins, due to Fontaine
in vegetal fiber reinforced composites. effect, fibers which are less numerous are oriented into
In this article we will identify the damage mechanisms flow direction. On the heart, the injection rate results in a
in short Hemp-Fiber Reinforced Polypropylene (HFRP) transverse orientation of the fibers.
composites with different fiber contents (10, 20, 30, and
40 wt%) and subjected to two hygrothermal aging. The Experimental Testing
first aging, called Aging 1, is defined by a Relative
Humidity (RH) of 80% and a temperature of 25 C, while Each batch of tensile samples was subjected to hygro-
the second one, called Aging 2, is characterized by a RH thermal aging, at 80% RH/25 C and 80% RH/50 C. The
of 80% and a temperature of 50 C. This identification is specimens of each HFRP composite were weighed regu-
made by means of a multivariable analysis based on k- larly during 50 days. The choice of this duration aims
means algorithm. This method was successfully applied principally at clarifying the impact of the water absorp-
to discriminate damage events in the case of several com- tion on the mechanical properties and the evolution of
posites with synthetic fibers [31]. The main objective of damage mechanisms. Temperature of 50 C is used to
this work is to apply this method to vegetal fiber rein- accelerate the aging mechanisms. To determine the
forced polymer matrices. The correlation between damage weight change, the samples were, in a few seconds, wiped
mechanisms in HFRP composites and clusters of AE sig- dry with a tissue paper and weighed outside the environ-
mental chamber at room temperature. Unaged and aged
samples were subjected to tensile tests following the
TABLE 1. Mechanical properties of HFRP composites derived from standard ASTM D638-10 with a cross-head speed of 2
tensile tests. mm/min. This norm suggests dumbbell-shaped samples
with central length, width, and thickness of 105, 10, and
Hemp fiber Young’s modulus Maximal stress Failure strain
(wt%) (GPa) (MPa) (%) 4 mm, respectively. A clip-on extensometer with 50 mm
gauge length was used to measure strain. For each set of
10% 2.23 6 0.09 20.70 6 0.61 4.07 6 0.05 samples, all the results were taken as the average value
20% 3.15 6 0.10 21.35 6 0.51 3.69 6 0.21 and the standard deviation of five tested samples. To
30% 3.99 6 0.12 22.22 6 0.61 3.37 6 0.15
obtain information about damage evolution in the tested
40% 4.66 6 0.13 22.62 6 0.70 2.90 6 0.12
specimens, tensile tests were coupled with AE

DOI 10.1002/pc POLYMER COMPOSITES—2016 1103

 quency range of 100 kHz–1 MHz and presenting a reso-
nance peak of around 300 kHz. These piezoelectric
sensors are coupled on the specimen faces with a silicon
grease. The adjustment of the AE system (before perform-
ing tensile tests on the studied specimens) was made in
two steps. The first one aims at eliminating the AE sig-
nals caused by noise and was performed by increasing

FIG. 4. Evolution of mechanical properties as function of hemp-fiber

content (wt%): (a) Young’s modulus, (b) tensile strength, and (c) strain
at failure. [Color figure can be viewed in the online issue, which is
available at wileyonlinelibrary.com.]

measurements, achieved by two channel data acquisition

system provided by Mistras Group Corporation. A sam-
FIG. 5. Variation of mechanical properties according to hygrothermal
pling rate of 5 MHz and a 40 dB preamplification are aging: (a) Young’s modulus, (b) tensile strength, and (c) strain at failure.
used to record AE data. AE measurements were achieved [Color figure can be viewed in the online issue, which is available at
by using two resonant piezoelectric sensors with a fre- wileyonlinelibrary.com.]

1104 POLYMER COMPOSITES—2016 DOI 10.1002/pc

 progressively the threshold until 32 dB (32 dB was found
to be the optimal threshold to avoid noise signals). The
second step, and always before the tensile tests, is
intended to eliminate the AE signals induced by action
grips and to optimize the distance between the two sen-
sors. This was performed by considering a pencil lead
break procedure with a minimum amplitude of 32 dB.
These two steps enable to achieve the correct setting of
the AE system, i.e., the optimal position of the sensors
and the record of AE signals originating principally from
damage. Once the AE system is correctly adjusted, the
threshold was fixed at 32 dB during the tensile tests for
all tested HFRP composites. More details about this
experimental protocol (hygrothermal aging, tensile test)
can be found in previous publications [16, 17].

RESULTS AND DISCUSSION

Water Absorption
In composite materials, it is known that water uptake
depends on many factors, such as temperature, fiber vol-
ume fraction, reinforcement orientation, fiber nature (per-
meable or impermeable), area of exposed surfaces,
diffusivity, reaction between water and matrix, and surface
protection [33, 34]. In the case of vegetal fibers reinforced
polymer composites, the moisture absorption is strongly
related to fiber content because of their hydrophilic nature
[35]. Accordingly, we present in Fig. 1 the water uptake
for all HFRP composite samples according to the aging
time. At the end of the 50 days of exposure, the water
uptake of each composite does not reach an equilibrium
plateau. Figure 1 shows also that water uptake increases in
a quasi-linear way when increasing the hemp content as
illustrated in Fig. 2. The increase in water absorption is
caused, among other factors, by the highly hydrophilic
nature of hemp fiber compared to the PP matrix [36].
This water uptake is found to be lower than 1% for
the aging 1 and 2% for the aging 2, which shows that
temperature activates the diffusion process in HFRP com-
posites. In fact, the diffusion processes in aqueous envi-
ronment of polymer composites are activated by an
increase in temperature, because the Fickian diffusion
coefficient is related to temperature exponential by the
Arrhenius-type relationship. Similar observations were
suggested for other natural fiber-reinforced composites in
Refs. [37, 38].

Mechanical Properties
To study the influence of hemp fiber content on the
mechanical behavior of HFRP composites, Fig. 3 shows
FIG. 6. PCA visualization of the clustering during static tensile tests of typical stress–strain curves for the unaged samples under
HFRP composites: (a) two classes, (b) three classes, and (c) four classes. monotonic tensile test. After a linear behavior, the curves
[Color figure can be viewed in the online issue, which is available at reveal a substantial nonlinear response, which could be
wileyonlinelibrary.com.]
attributed to the matrix plastic deformation. This nonlin-
ear portion decreases proportionally when increasing the

DOI 10.1002/pc POLYMER COMPOSITES—2016 1105

 FIG. 7. Unaged samples. Variation of each cluster according to: (a) amplitude histograms, (b) duration his-
tograms, and (c) counts number histograms. [Color figure can be viewed in the online issue, which is avail-
able at wileyonlinelibrary.com.]

hemp fiber content (the increase of the fiber content up by 3%, 7%, and 9%. This low increase in the tensile
reduces the ductility of the HFRP composites). For each strength is likely to be attributed to the inefficiency of the
composite, we remark that the mechanical stress remains hemp fiber treatment, used to improve the quality of the
approximately constant after having reached the maxi- fiber-matrix interface. In addition to that, and as reported
mum stress while the strain increases until failure. The in Ref. [39], we notice that 30 wt% is presumably the
static strength properties of the different composites are optimum fiber content ratio for the reinforcement in
given in Table 1. HFRP composites, because the tensile strength increases
Compared to HFRP composite with 10 wt% of fiber by less than 2% when compared to the result of the 40
content, the Young’s modulus increases by 41%, 79%, wt% fiber content composite.
and 109% for composites with 20%, 30%, and 40 wt% To assess the influence of humidity and temperature
fiber content, respectively, and the tensile strength goes on the mechanical properties of HFRP composites, Figs.

1106 POLYMER COMPOSITES—2016 DOI 10.1002/pc

 FIG. 8. SEM pictures taken from the fracture surfaces of unaged HFRP composites with: (a) 10 wt%, (b)
20 wt%, (c) 30 wt%, and (d) 40 wt%. [Color figure can be viewed in the online issue, which is available at
wileyonlinelibrary.com.]

4 and 5 illustrate the evolution of the elastic and ultimate the strain at failure by 11% and 38%, respectively, for the
properties of unaged and aged materials in term of the 40 wt% HFRP composite. The plasticization of hemp
fiber content. These results show that the hygrothermal specimens caused by moisture absorption explain the
aging leads to significant variations in the Young’s modu- increase in failure strain. In fact, water molecules act as
lus, tensile strength, and strain at failure. Indeed, humidity plasticizer agent in the HFRP composites and temperature
induces a decrease in the Young’s modulus and in the accelerates water diffusion, which normally leads to a
tensile strength (Fig. 4a and b), while the strain at failure higher increase in the strain at failure.
is found to increase (Fig. 4c). These variations are also
amplified by the temperature and the hemp fiber content.
Damage Mechanisms Analysis
For composites with 40 wt% hemp fiber content aged at
80% RH/25 C, the Young’s modulus decreases by about To identify the different failure modes of bio-
10% after 50 days of aging and by 26% for the second composites, several methods have used the amplitude to
aging during the same period (Fig. 5a). On the other classify the AE signals [19]. Unfortunately, in the case of
hand, Young’s modulus increases linearly with the materials with very high damping characteristics, the clas-
increase in hemp fiber content (Fig. 4a), whatever the sification based on amplitude only presents several over-
aging. Besides, the decrease of the tensile strength in the laps between failure mechanisms. Consequently, it is
case of the aging 2 could be essentially related to thermal important to consider simultaneous characteristics of AE
dilatations that have the effect of accelerating the diffu- signals to evaluate different damage mechanisms and to
sion of the humidity in the PP, by affecting afterwards follow their evolution in time until the final global
the rigidity of the fiber and consequently the degradation failure.
of HFRP composites. The evolution is different for strain It is within this context that the AE results obtained
at failure because the hygrothermal aging 1 and 2 raise during mechanical tensile tests on our studied HFRP

DOI 10.1002/pc POLYMER COMPOSITES—2016 1107

 FIG. 9. Amplitudes of AE signals versus tensile tests time for unaged HFRP composites: (a) 10 wt%, (b)
20 wt%, (c) 30 wt%, and (d) 40 wt%. [Color figure can be viewed in the online issue, which is available at
wileyonlinelibrary.com.]

composites are analyzed, and by considering a multivari- modes present in HFRP composite materials. To be sure
able analysis by using Noesis Software. of the number of classes identified by this classification,
The objective of this part is to use the main character- it is necessary to combine it to other information from
istics of AE signals such as amplitude, duration, rise the literature results and microscopic observations.
time, and number of counts, to apply this multivariable The signals were separated in groups according to the
method in order to study failure mechanisms in the HFRP chosen parameters (amplitude, duration, rise time, and
composites. This section is divided into two parts. First, counts). Next, the classification was applied by varying
only the failure mechanisms related to the hemp fiber the number of classes to optimize the overlap rate
reinforcement will be discussed for the unaged compo- between the various groups. In addition, the obtained
sites. Second, the damage evolution according to the groups were analyzed with a principal component analysis
hygrothermal aging will be studied. (PCA) to examine the overlap rate as shown in Fig. 6.
The classification used in this work, considered also The obtained results show that the minimum of overlap
by Marec et al. [31] in the case of traditional composites, (12%) is obtained with three classes (Fig. 6b). The pro-
aims at having an additional information in real time on cedure of optimization of the number of classes as well
the initiation and propagation of the various damage as the choice of the AE signals parameters were detailed

1108 POLYMER COMPOSITES—2016 DOI 10.1002/pc

 FIG. 10. Number of hits versus tensile tests time for unaged and aged HFRP composites. [Color figure can
be viewed in the online issue, which is available at wileyonlinelibrary.com.]

in the work of Marec et al. [31]. After the optimization of attributing the relevant damage mechanisms to each group
the number of classes, the waveforms of each signal were of signals [19, 31]. Finally, some microscopic observa-
analyzed on the basis of literature results with the aim of tions on fracture surfaces of the studied specimens were

DOI 10.1002/pc POLYMER COMPOSITES—2016 1109

 necessary to note that AE signals induced by the matrix
cracking (class 1) are characterized by low amplitudes
(centered around 40 dB) with duration and counts number
lower than 30 ls and 10, respectively. Furthermore, the
characteristics of these signals are similar to other results
from Refs. [39–41]. The signals of fiber-matrix damage
mechanism (class 2) are identified by amplitudes, dura-
tion, and counts number centered at 50 dB, 50 ls, and
12, respectively. The fiber failure mechanism (class 3) is
mainly identified by signals of amplitudes between 38
and 77 dB with duration higher than 100 ls. The signals
characteristics of these last two classes are different from
those identified by several studies in the case of tradi-
tional composites [23, 41]. Indeed, the second class is
induced by different damage in the interface between
fiber and matrix such as pull-out mechanism (Fig. 8a)
and matrix debonding (Fig. 8c), resulting in emission of
signals with various values of amplitude [35–62 dB]. In
the same way, the third class is related to several failure
processes in hemp-fibers: micro-fibrils explosion (Fig.
8b), breakage of micro-fibrils (Fig. 8b) and hemp-fiber
breakage (Fig. 8c), that lead to AE signals with a
wide range of amplitude [38–77 dB] and count number
[10–50].
Figure 9 presents the amplitude scatter graph of the
three classes versus tensile tests time for 10, 20, 30,
and 40 wt% HFRP composites. The classification shows
that the damage within the studied materials is mainly
governed by the matrix cracking (Class 1), which
presents more than 73% of AE events. Besides, note
that AE activity increases with hemp fiber content. In
spite of this increase, the rates of AE events induced by
matrix-fiber interface failure (lower than 21%) and fiber
breakage (lower than 5%) remain low, which could
explain the moderate increase in the tensile strength
observed in the mechanical behavior (Fig. 3 and Table
1). Finally, we remark that the fibers failure (third class)
is detected earlier when the hemp fiber content
increases.
Figure 10 presents the evolution of the cumulative
number of hits of unaged and aged HFRP composites.
These graphs show that the damage mechanisms evolu-
tion is different for unaged and aged composites. In
fact, for unaged composites (Fig. 10a, d, g, and j), the
matrix cracking and fiber debonding are detected in the
beginning of the test while the fiber failure is indicated
after few seconds (13 s for 10 wt% vs. 4 s for 40
wt%). Concerning the aged composites, we observe an
FIG. 11. Number of hits versus hemp content: (a) Matrix cracking, (b) offset of the acoustic activity which is slightly amplified
Fiber matrix debonding, and (c) Fiber failure. with the aging at 50 C (Fig. 10c, f, i, and l). We also
remark that the number of acoustic events decreases
made to be sure of the presence of these damage mecha- with aging in particular when the temperature and the
nisms in HFRP composites. hemp fiber content increase, which confirms that tem-
To better illustrate the evolution of damage mecha- perature amplifies the degradation of the studied
nisms identified by the classification in the HFRP compo- composites.
sites, Fig. 7 presents the distributions of amplitude (Fig. To better illustrate the effect of hygrothermal aging on
7a), duration (Fig. 7b), and counts number (Fig. 7c). It is the AE activity, Fig. 11 presents the variation of the

1110 POLYMER COMPOSITES—2016 DOI 10.1002/pc

cumulative hits as a function of the hemp fiber content ACKNOWLEDGMENTS
for each class (matrix cracking, fiber matrix debonding,
The authors would like to thank the society AFT Plas-
and fiber failure). For the matrix cracking (Fig. 11a), we
turgie Company for provided compounds. They would
notice a relatively low decrease of acoustic activity for
also like to thank the University of Technology of Troyes
the 10 and 20 wt% aged composites with respect to
for the use of their SEM facility.
unaged ones. For the 30 and 40 wt% composites aged at
50 C, the reduction is found to be equal to 78 and 83%, REFERENCES
respectively. On the other hand, the acoustic activity
induced by matrix-fiber interface failure and fiber break- 1. Directive 2000/53/EC of the European Parliament and of
age decreases more significantly than the matrix cracking the Council of 18 September 2000 on End-of-Life Vehicles,
damage for all HFRP composites. This result shows that Off. J. Eur. Commun., L 269/34 (2000).
the studied aging further degrades the fiber-matrix inter- 2. M. Pervaiz and M.M. Sain, Resour., Conserv. Recycl., 39,
face, which explains the important reduction of the acous- 325 (2003).
tic activity (Fig. 11b). Indeed, hemp fiber contributes to 3. J. Holbery and D. Houston, JOM, 58, 80 (2006).
more water penetrating into the fiber/matrix interface
4. C. Baley, Compos. A, 33, 939 (2002).
through the micro-cracks generated by swelling of fibers
5. R. Heijenrath and T. Peijs, Adv. Compos. Lett., 5(3), 81
[13]. As capillarity and water transport along fiber–matrix
(1996).
interfaces become more favored, the water molecules
cause damage on the interface and inside the elementary 6. A. Bourmaud and C. Baley, Polym. Degrad. Stabil., 94, 297
(2009).
hemp fibers. It results in fibers pull-outs and fiber matrix
debonding, as shown in the Fig. 8a and c. Finally, the 7. H.P. Fink, J. Ganster, J. Fraatz, and M. Nywlt, Relations
Between Structure and Mechanical Properties of Cellulosic
low number of fibers breakage (Fig. 11c) could be
Man-Made Fibres, in the Proceedings of Akzo-Nobel Vis-
explained by degradation in the fiber-matrix interface cose Chemistry Seminar Challenge in Cellulosic Man-Made
induced by water absorption (for aged HFRP composites) Fibres, Stockholm (1994).
or interface problems caused by the inefficiency of fiber
8. C.P.L. Chow, X.S. Xing, and R.K.Y. Li, Compos. Sci. Tech-
treatment (for unaged HFRP composites). nol., 67, 306 (2007).
9. H. Chen, M. Miao, and X. Ding, Compos. A, 40, 2013
CONCLUSIONS (2009).
10. A. Le Duigou, P. Davies, and C. Baley C, Polym. Degrad.
This article presents the results of experimental investi- Stabil., 94, 1151 (2009).
gations of hygrothermal aging effects on HFRP compo-
11. A. Arbelaiz, B. Fernandez, J.A. Ramos, A. Retegi, R.
sites. Four hemp fiber contents were considered (10, 20,
Llano-Ponte, and I. Mondragon, Compos. Sci. Technol., 65,
30, and 40 wt%) and the hygrothermal aging was defined 1582 (2005).
by a RH of 80% at different temperatures (25 and 50 C).
12. M.S. Islam, K.L. Pickering, and N.J. Foreman, J. Polym.
The static strength and residual stiffness of aged speci- Environ., 18, 696 (2010).
mens were evaluated with tensile tests after an exposure
13. H.N. Dhakal, Z.Y. Zhang, and M.O.W. Richardson, Com-
of 50 days to hygrothermal aging.
pos. Sci. Technol., 67, 1674 (2007).
The obtained results show that the mechanical prop-
14. S. Panthapulakkal and M. Sain, J. Appl. Polym. Sci., 103,
erties (Young’s modulus, tensile strength, and strain at
2432 (2007).
failure) of the studied HFRP composites were affected
15. G. Romhany, J. Karger-Kocsis, and T. Czigany, J. Appl.
by the chosen hygrothermal aging, in particular at
Polym. Sci., 90(13), 3638 (2003).
50 C. The multi-variable analysis of the AE signals
16. M. Assarar, D. Scida, A. El Mahi, C. Poil^ane, and R. Ayad,
and Electron Microscopy observations enabled, on the
Mater. Des., 32, 788 (2011).
one hand, to identify the chronology of the various
17. D. Scida, M. Assarar, C. Poil^ane, and R. Ayad, Compos. B,
damage mechanisms in HFRP composites and, on the
48, 51 (2013).
other hand, to characterize their chronology according
to the fiber content and the hygrothermal aging. The 18. C. Santulli, NDT&E Int., 34(8), 531 (2001).
obtained results revealed that the number of AE signals 19. I.M. De Rosa, C. Santulli, and F. Sarasini, Compos. A, 40,
induced by matrix-fiber interface failure remains low, 1456 (2009).
for both aged and unaged composites, which could be 20. I.M. De Rosa, C. Santulli, F. Sarasini, and M. Valente,
principally related to the inefficiency of hemp fibers Polym. Compos., 30(12), 1879 (2009).
treatment. In addition to that, it was shown that the 21. G. Dogossy and T. Czigany, Polym. Test., 25(3), 353 (2006).
result obtained from the classification can be used as 22. T. Czigany, Compos. Sci. Technol., 66, 3210 (2006).
an indicator to evaluate the quality of the fiber-matrix 23. H. Anuar, S.H. Ahmad, R. Rasid, S.N. Surip, T. Czigany,
interface, by focusing on the number of AE signals and G. Romhany, J. Compos. Mater., 41(25), 3035 (2007).
induced by the interface failure and on their amplitude 24. B.A. Acha, N.E. Marcovich, and J. Karger-Kocsis, Plast.
ranges. Rubber Compos. 35(2), 73 (2006).

DOI 10.1002/pc POLYMER COMPOSITES—2016 1111

 25. G. Romhany, J. Karger-Kocsis, and T. Czigany, Macromol. 34. G.S. Springer, Environmental Effects on Composite Mate-
Mater. Eng., 288(9), 699 (2003). rial, Vols. 1–3; Technomic, Westport (1981/1984/1988).
26. N. Godin, S. Huguet, R. Gaertner, and L. Salmon, NDT&E 35. P.V. Joseph, M.S. Rabello, L.H.C Mattosoc, J. Kuruvilla, and
Int., 37(4), 253 (2004). T. Sabu, Compos. Sci. Technol., 62(10–11), 1357 (2002).
27. M. Moevus, N. Godin, D. Rouby, M. R’Mili, P. Reynaud, 36. M. Assarar, S. Buet, R. Ayad, A. Cherouat, and X.L. Gong,
and G. Fantozzi, Compos. Sci. Technol., 68(6), 1258 (2008). JNC 18, Nantes 12–14 Juin 2013.
28. S. Momon, N. Godin, P. Reynaud, M. R’Mili, and G. 37. A. Retegi, A. Arbelaiz, P. Alvarez, R. Llano-Ponte, J. Labidi,
Fantozzi, Compos. Part A, 43, 254 (2012). and I. Mondragon, J. Appl. Polym. Sci., 102, 3438 (2006).
29. M.G.R. Sause, D. Schultheiß, and S. Horn, J. Acoust. Emis- 38. A. Espert, F. Vilaplana, and S. Karlsson, Compos. A, 35,
sion, 26, 1 (2008). 1267 (2004).
30. M.G.R. Sause and S. Horn, J. Nondestruct. Eval., 29, 123 39. A. Bravo, L. Toubal, D. Koffi, and F. Erchiqui, Int. J.
(2010). Mater. Sci., 3(3), 1 (2013).
31. A. Marec, J.-H. Thomas, and R. El Guerjouma, Mech. Syst. 40. A. Laksimi, S. Benmedakhene, and L. Bounouas, Monitor-
Signal Process., 22, 1441 (2008). ing Acoustic Emission During Tensile Loading of Thermo-
32. B. Bonnet, Impact Behavior of Composites for Automotive plastic Composites Materials, Paper presented at the
Applications (in French), Phd Thesis, Ecole Nationale Proceeding of ICCM (1999).
Superieure des Mines de Paris, 22–23 (2005). 41. G. Kotsikos, J.T. Evans, J.M. Hale, and A.G. Gibson,
33. C.H. Shen and G. Springer, J. Compos. Mater. 10, 2 (1999). Mater. Eval., 58, 1320 (2000).

1112 POLYMER COMPOSITES—2016 DOI 10.1002/pc

View publication stats