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

net/publication/332072324

Numerical study of mechanical behavior of Agave fibers reinforced composites

Conference Paper · March 2019

CITATIONS READS
0 70

4 authors, including:

Imen Mâatouk Slah Msahli

15 PUBLICATIONS   5 CITATIONS   
Institut Supérieur des Etudes Technologiques de Ksar Hellal
167 PUBLICATIONS   717 CITATIONS   
SEE PROFILE
SEE PROFILE

Zouari Rym
National Engineering School of Monastir
18 PUBLICATIONS   30 CITATIONS   

SEE PROFILE

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

Textile and para-textile Valorization of natural fiber (kenaf, alfa, waste of cotton, posidonie...) View project

Three Dimensional Numerical Simulation of Continuous Filaments Migration in Polypropylene Twisted Yarn View project

All content following this page was uploaded by Zouari Rym on 29 March 2019.

The user has requested enhancement of the downloaded file.

Numerical study of mechanical behavior of
Agave fibers reinforced composites

Yosra Glouia1, Asma El Oudiani1, Imen Maatoug 1, Rim Zouari2 and Slah
Msahli 1

1 Laboratory of Textile Engineering, University of Monastir, Ksar Hellal 5070, Tunisia

gyosra1@yahoo.
2 Mechanical Engineering Laboratory, University of Monastir, Tunisia

Abstract. An investigation was carried out on local natural cellulosic fibers which
have gained interest in the composite field due to their superior specific properties.
A multi scale finite element (FE) model of unidirectional natural fiber composite
materials with reasonable dimensionality was developed. The mechanical behav-
ior of the composite at macro scale as well as meso-scale was simulated. In partic-
ular the response to tensile and three-points bending test was studied. Linear mate-
rial properties are obtained by using pure strains assumptions in the implicit
analysis of the composite, while the non-linear behavior and viscoelastic parame-
ters require the explicit dynamic analysis. Simulation is performed thanks to
Abaqus finite element software. Comparison of Experimental and FEM tensile
and three-point bending Strength shows very good agreement.

Keywords: Finite Element Analysis, natural fibers, biocomposites, mechanical

behavior, viscoelasticity, plasticity, Abaqus

1 Introduction

Composites are the multi-phase material and wonder materials becoming an es-
sential part of today’s materials due to their advantages such as low weight, corro-
sion resistance and high fatigue strength (Wilson 2003). They are extensively used
as materials in making aircraft structures, automotive industry, electronic, packag-
ing, medical equipment, space vehicle and homebuilding (Berreur et al. 2002).
Despite their interesting performance, these materials are polluting and harmful
to nature. And considering the increase in environmental constraints such as Kyo-
to protocol, it was necessary to think of ecofriendly materials as substituents, for
instance natural fibers (Wambua et al. 2003). Natural fibers made polymer com-
2

posites are gaining more interest in engineering applications due to appreciable

structural properties at a relatively low cost which finds its attraction towards the
field of research. Composite microstructures are determined by the physical and
mechanical properties of the individual materials. Analytical methods provide rea-
sonable prediction for relatively simple configurations of the phases. Complicated
geometries, loading conditions and material properties often do not yield analyti-
cal solutions, due to complexity and the number of equations.
In this case, numerical methods are used for approximate solutions, but they
still make some simplifying assumptions about the inherent microstructures of
heterogeneous multiphase materials, one such method is finite element analysis.
Literature reveals that researchers have not been interested till now to viscoelasto-
plastic behavior law of composite materials. In this context, the present work aims
to simulate the mechanical behavior of natural cellulosic fibers reinforced compo-
sites using finite element modeling method.

2 Materials and methods

2.1 Published Data

To be able to model and simulate the mechanical behavior of any material

with Abaqus finite element package, several essential parameters must be availa-
ble especially material geometry and dimensions as well as behavior law. In litera-
ture, several researchers (Tunisia, France, Germany…) have worked on natural fi-
bers reinforced composites. Chaabouni studied composites made from
thermosetting matrix and Agave fiber (Chaabouni 2005), Ghali worked on polyes-
ter luffa composites (Ghali 2011), Ben Brahim studied tensile properties of unidi-
rectional Alfa-polyester composite (Ben Brahim and Ben Cheikh 2007), Allègue
worked on mechanical properties of Posidonia oceanica fibers reinforced cement
5Allègue 2014).
Following this bibliographic synthesis and published data, we notice that we
have the maximum of information and data (Curves, the max of tests ... etc.) for
the Agave fibers, the polyester and epoxy matrices and the composite resulting
from these latter components (Chaabouni 2005). To this is added the researches of
(Msahli 2007) and (El Oudiani 2003) in which they conducted respectively a
physical characterization (diameter, length, morphology…) and a rheological
study on Agave fibers. Thanks to these three researches, it is possible to perform a
finite element analysis of unidirectional aligned Agave-polyester and Agave-
epoxide composite.
Based on (Chaabouni 2005) and (Msahli 2007) researches, Agave Americana
L. fibers density is equal to 1.36 which is considered as a low weight compared to
 3

other hard vegetable fibers. For these reasons, Agave Americana L. fibers are very
interesting. We resume in Table 1 elastic parameters (Young’s modulus and Pois-
son coefficient) as well as tensile mechanical characteristics (resistance, elonga-
tion) of the performed composites with diverse ratios of fibers.

Table 1. Elastic parameters of Agave based Composites (Chaabouni 2005)

Matrix Agave E (Gpa) ν Stress Strain

fraction (Mpa) (%)
(%)
PES 15 1.48 0.32 28 4.8
PES 24 1.07 0.32 31 9
PES 31 1.164 0.32 33 8.5
EPX 10 2.34 0.333 48 3.2
EPX 20 2.317 0.333 24 1.1
EPX 30 2.307 0.333 22 1.1

2.2 Finite Element Modeling

The choice of the length scale, together with the constitutive material models
and the related parameters is a key point especially in the fields of composites,
which are made of different materials. Several approaches were used and devel-
oped during the last years and are reported in the literature. In particular, two dif-
ferent techniques of modelling have been employed in the literature, one technique
does not reproduce all the fibers and the matrix but considers the layer as a unique
homogenous mean. This approach most likely does not allow the detailed investi-
gation of complex phenomena. The Meso heterogeneous model (Da Selva et
al.,2015; Bavan and Kumar, 2013; Gehring, 2013; Sun and Vaidya, 1995; Kumar
et al.,2016; Prasad et al.,2014; Rajesh et al., 2016; Houshyar et al., 2009) instead
reproduces every single fiber and the matrix, as well as the interaction among
these. Modeling of the mechanical behavior of fiber-reinforced polymer matrix
composites is presented by the example of Tunisian Americana. L Agave fibers in
a polyester and epoxide based matrix respectively. Our study contains two differ-
ent scales macro and meso scale models because of the heterogeneity in composite
materials when analyzing them at numerous scales.
4

3 Results and discussion

3.1 Macro scale modeling, tensile test simulation

Unidirectional Agave reinforced composite is considered as a three-

dimensional plate (rectangular shape) when analyzing it at macro scale. Dimen-
sions of the plate are: 100*20*2 mm3.
Concerning plastic behavior, we used experimental data and we calculate true
stress and true strain thanks to the following formulas:
(1)
(2)
(3)

(4)
The following figures represent the results of the simulation of the tensile test
of Agave polyester composites as well as Agave epoxide composites using various
weight ratios. We draw for each composite the numerical and the experimental
curve. Then, we conclude from the two curves the error generated by the model.
 5

Fig. 1 Comparison between experimental and numerical stress-strain response under ten-
sile test of respectively 15% Agave polyester composite (a) and 24% Agave polyester com-
posite (b)

For these composites, experimental and numerical curves present the same
shape with an error in the range of 5-15%. This error may be due to the assump-
tion of homogeneity of composite material which does not exist concretely. We
summarize the error value for each composite in table 2.

Table 2. Error (%) between experimental and numerical results

Composite Error
(%)
PES 15% Agave 7.62
PES 24% Agave 7.39
EPX 10% Agave 16.25
EPX 30% Agave 5.52

Error value is acceptable since it doesn’t exceed 16%. We conclude thus that
model converge and may be validated.

3.2 Macro scale modeling, 3 points bending test simulation

We consider the same constitutive law for the model used in the simulation of
the tensile test since we are working with the same materials already described.

Numerical flexural modulus was calculated applying the following formula (4).
6

(4)

Where P1 is the maximum load (N), L the range (mm), e the specimen thick-
ness (mm), b the specimen width (mm) and y1indicates the defection (mm).
Results are summarized in table 3.

Table 3. Experimental and numerical values of flexural modulus (MOE) of Agave based
composites under 3 points bending test

Composite NumericalExperimentalError
MOE MOE (GPa) (%)
(GPa)
PES 15% 1.4 1.6 12
Agave
PES 24% 2.12 2.5 15
Agave
PES 31% 2.4 2.6 7
Agave
EPX 10% 2.34 2.4 6
Agave
EPX 20% 2.4 2.5 4
Agave
EPX 30% 2.7 2.8 3
Agave

Comparison of Experimental and FEM tensile and three-point bending Strength

shows very good agreement since error doesn’t exceed 15%.

3.3 Meso scale modeling, tensile test simulation

The material id modelled and analyzed for mechanical properties with finite el-
ement method software (ABAQUS 6.14). Our model is a three-dimensional de-
formable body. We created a rectangle that is then extruded to form a parallelepi-
ped. The modeling of the fibers is created by their partition in the matrix in order
to ensure a perfect adhesion. Thus, we can model the composite on the mesoscop-
ic scale, without having to create different pieces. The number of circles corre-
sponds to the mass fraction considered in the study. Dimensions of sample are
100*20*2 mm3. Fiber’s diameter is about 0.263 mm. Elastic parameters of Agave
fibers and matrices are summarized in table 4.
 7

Table 4. Fibers and matrix properties (Chaabouni,2005) (Carma, 2006).

E (MPa) ν
Agave fibers 4000 0.33
Polyester matrix 870 0.37
Epoxide matrix 1924 0.386

In order to apply the boundary conditions, we have coupled the two faces z = 0
and Z = 100 mm to a rigid body with well-defined reference points (RP1 and
RP2). First, we perform a static analysis and we develop a model in which the
composite material is assigned as unidirectional composite by assuming the fiber
and matrix materials are linearly elastic. The numerical simulation results appear
in figures 2 and 3.

Fig. 2 Comparison between experimental and numerical response under tensile test consid-
ering elastic behavior law for epoxide 10% Agave composite

Fig. 3 Comparison between experimental and numerical response under tensile test consid-
ering elastic behavior law for polyester 15% Agave composite

The curves in figures 2 and 3 present the same shape with the presence of slight
difference. The difference between experimental and numerical simulated curves
is equal to 14% for the case of polyester/Agave composite. While, for epoxide
Agave composite, the difference between numerical and experimental curves is
8

about 2%. Hence the model converges. We calculate also numerical Young’s
modulus and we compare it with the experimental value as mentioned in the next
table.

Table 5. Numerical Young’s modulus vs Experimental Young’s modulus.

Composite E simula- E experi-

tion (GPa) mental
(GPa)
PES 15 1.2 0.96 [4]
Agave
EPX 10 2.25 2.2 [4]
Agave

Numerical and experimental results have very good agreement.

Although the previous model shows good results, it doesn’t reflect the real be-
havior law of Agave fibers as well as matrices employed in this study. That is
why, we propose a second model in which we consider a viscoelastic behavior law
for Agave fibers and an elastoplastic behaviour for polyester and epoxide matrix
as it is mentioned in literature (Perrot et al., 2006).
The viscoelastic law of fibers is identified by the load–extension test added
with the relaxation test. These tests have been then implemented in the ABAQUS
finite element package via Prony series method. They define shear moduli and
volumetric compression moduli over time as illustrated in formula 5:

(5)
Where GR(t) is the shear modulus over time, G0 the shear modulus at the very
beginning of the test, g R(t) is the ratio between the shear modulus at the beginning
of the relaxation test, τi is a relaxation time corresponding to the increment i, t is
the time during the relaxation test and gi is the shear modulus corresponding to the
increment i and corresponds to the relaxation time τi.

Abaqus allows to calculate Prony series from the results of the dimensioned re-
laxation tests.
In this method, we conduct an explicit analysis. The mechanical properties
values are the same mentioned in table 4. Results of simulation under tensile test
for Agave polyester and Agave epoxide composites are illustrated in figure 4.
 9

Fig. 4 Comparison between experimental and numerical response under

tensile test using explicit analysis and viscoelastoplastic behavior law of re-
spectively polyester 15% Agave composite (a) and epoxide 10 % Agave (b)

The curves of figure 4 are very similar and present the same shape with a slight
difference that may be due to many factors (presence of porosity, discontinuity,
presence of voids in fibers and matrix, anisotropic behaviour of Agave fibers….).
Difference between experimental and numerical simulated curves is equal to
5.88% for the case of Agave PES composite and 9% for the case of Agave EPX
composite. Hence the model converges. Numerical and experimental results have
very good agreement. The error value is smaller for the case of the second model,
considering a viscoelastoplastic behavior law.

5 Conclusion

Finite element analysis is performed in two different scales: macro and meso
(scale of fibers). For macro scale modeling, we have predicted the behavior and
the response of polyester Agave composites as well as epoxide Agave composites.
Numerical and experimental results were globally in agreement even so the pres-
10

ence of a slight difference which varied depending on the material studied. This
error was in the interval of 5% and 22% concerning tensile test simulation results.
Moreover, we have simulated the behavior of composites under three points
bending test. In the one hand, we have calculated flexural modules. And in the
other hand, we showed the influence of fibers addition to polyester and epoxide
matrices on flexural properties of composite materials. Simulation results were in
very good agreement with literature.
Concerning the meso scale modeling, we have predicted the behavior and the
response of polyester Agave composites as well as epoxide Agave composites tak-
ing into account the properties of matrix and fibers. We started by a simple elastic
behavior law for Agave fibers, polyester and epoxide. Even though, this model
was in concordance with experimental results, it didn’t reflect accurately the be-
havior law of the materials.
A second model in which we consider a viscoelastic behaviour for Agave fibers
and an elastoplastic behaviour for polyester and epoxide matrices was thus per-
formed to remedy to deficiencies of the first simulation model.
The error percentage for the second model does not exceed 9%, which is evalu-
ated as a very satisfying result. The difference in the results, obtained from Finite
element analysis and experimental analysis, could be explained by the presence of
porosity, discontinuity and presence of voids in fiber and matrix and also due to
the anisotropic behavior of Agave fiber (Prasad et al., 2014).
The error between experimental and numerical results of Agave fiber compo-
sites could be also attributed to the non-uniformity and heterogeneity of the fibers.

References

Wambua P, Ivens J, Verpoes, I (2003) Natural fibres: Can they replace glass in fibre rein-
forced plastics? Composites science and technology, 63(9), 1259-1264
Ghali L, Msahli S, Zidi M., Sakli F (2011) Effects of fibre weight ratio, structure and fibre
modification onto flexural properties of Luffa –polyester composites. Advances in Ma-
terials Physics and Chemistry 1(3), 78 – 85
Ben Brahim S and Ben Cheikh R (2007) Influence of fibre orientation and volume fraction
on the tensile properties of unidirectional Alfa-polyester composite. Composites science
and technology 67(1), 140-147
Msahli S, ChaabouniY, Sakli F ,Drean J.Y (2007) Mechanical behavior of Agave America-
na L. Fibres: Correlation between fine structure and mechanical properties. Journal of
Applied Sciences 7(24), 3951-3957
Da Silva L.J, Panzera T.H, Christoforo A.L, Durão L.M.P, Rocco Lahr F.A (2012) Numer-
ical and Experimental Analyses of Biocomposites Reinforced with Natural Fibres. In-
ternational Journal of Materials Engineering 2(4): 43-49
Sarvana Bavan D, Mohan Kumar G.C (2012) Finite element analysis of a natural fiber
(Maize) composite beam. Journal of Engineering, 2013, 1-7
Sun C.T, Vaidya R.S (1996) Prediction of composite properties from a representative vol-
ume element. Composites Science and Technology 56(2), 171-179
View publication stats

11

Kumar J, Praveen D.N, Thara R, Irfan G (2016) Experimental & finite element analysis of
Sisal fibre reinforced composites. International Journal of Recent Trends in Engineering
& Research 2(7), 155-160
Prasad G, Venkatachalam, Akshat, Rathi, Rajakuma S (2014) Finite element analysis of
jute fibre made hybrid polymer matrix composite. Applied Mechanics and Materials
592(594), 363-367
Rajesh M., Srinag T, Phani Prasanthi, P, Venkataraovenkatrao, K (2016) Finite Element
Analysis of Coir/Banana Fiber Reinforced Composite material. International Journal of
Advanced Research in Mechanical Engineering & Technology 2(4), 29-33
Houshyar S, Shanks R.A, Hodzic A (2009) Modelling of polypropylene fibre-matrix com-
posites using finite element analysis. EXPRESS Polymer Letters 3(1), 2–12
Perrot Y, Baley C, Davies P (2006) Influence of low styrene emission polyester resins on
the aging behavior of composites in a marine environment. Applied Composite Mate-
rials 13(1), 1-22
Berreur L., Maillard B, Nösperger S (2002). L’industrie française des matériaux compo-
sites.
Wilson M.J (2003) Finite Element Analysis of Glass Fibre Reinforced Thermoplastic Com-
posites for Structural Automotive Components. Dissertation, University of Nottingham
Chaabouni Y (2005) Microstructure Characterisation of Agave Americanan L. fibre ; study
of Agave fibre reinforced composites. Dissertation, University of Mulhouse
El Oudiani A (2003) Study of the mechanical behavior of Agave fibers Americana L. Dis-
sertation, University of Monastir
Ghering F (2013). Mechanical behaviour and damage study of short hemp fibre reinforced
thermoplastics: experimental approach and modelling. Dissertation. University of Lor-
raine

http://stibeziers.fr/tsipm/spip_tsipm/IMG/pdf/composite3.pdf