Composites Part B 89 (2016) 143e154

Contents lists available at ScienceDirect

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

Numerical modelling of ax short bre reinforced and ax bre fabric

reinforced polymer composites
Janis Sliseris a, Libo Yan b, c, *, Bohumil Kasal b, c
a
Riga Technical University (RTU), Kalku Street 1, LV-1658, Riga, Latvia
b
Department of Organic and Wood-Based Construction Materials, Technical University of Braunschweig, Hopfengarten 20, 38102, Braunschweig, Germany
c
Centre for Light and Environmentally-Friendly Structures, Fraunhofer Wilhelm-Klauditz-Institut WKI, Bienroder Weg 54E, 38108, Braunschweig, Germany

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

Article history: The ever-increasing demand of ax short bre-reinforced and ax bre fabric-reinforced polymer
Received 10 October 2015 composites in various engineering applications calls for accurate predictions of their mechanical be-
Received in revised form haviors. In this study, numerical methods to generate and simulate mechanical properties of ax short
17 November 2015
bre-reinforced and ax bre fabric-reinforced polymer composites are proposed. The microstructures
Accepted 28 November 2015
Available online 14 December 2015
of short ax bres with different bre length-to-diameter ratios are generated by algorithm taking ber
defects (e.g. kink band) and ber bundles into account. Bidirectional ax fabric is generated and dis-
cretized by tetrahedron 4-node nite elements. A brittle material law for bre defects and interfacial
Keywords:
A. Discontinuous reinforcement
zones of bre bundles is proposed. Flax short bre/polypropylene and ax fabric/epoxy composites are
A. Polymer-matrix composites (PMCs) modeled by a non-linear plasticity model considering an isotropic hardening law and non-local con-
C. Finite element analysis (FEA) tinuum damage mechanics. The numerical modelling results are compared with the experimental results
of these composites. This study shows that the simulation can capture the main damage mechanisms of
the composites such as bre breakage initiated at the ber defects, damage of polymer matrix and the
bre debonding at bre/matrix interface accurately. In addition, the simulation results exhibit good
agreements with the experimental results in the aspects of elastic properties and nonlinear tensile
stress-strain behavior of the short bre and bre fabric reinforced polymer composites.
2015 Elsevier Ltd. All rights reserved.

1. Introduction
Because of an increasing environmental concern, it is urgent for
the development of light-weight, cost-effective and sustainable
materials as alternatives to conventional materials. In recent years,
the use of bio-bres to replace glass bres as reinforcement ma-
terials in polymer composites has gained popularity for various
engineering applications [1e8]. Natural bres are widely available
in most countries and are cost-effective with low density. They are
biodegradable, renewable, non-hazardous and non-abrasive. In
addition, their specic mechanical properties are comparable to
those of synthetic glass bres used as reinforcement materials
[9e12]. Among all the natural bres, ax bre offers the best
* Corresponding author. Department of Organic and Wood-Based Construction
Materials, Technical University of Braunschweig, Hopfengarten 20, 38102,
Braunschweig, Germany. Tel.: 49 531 220 7725.
E-mail addresses: janis.sliseris@gmail.com (J. Sliseris), l.yan@tu-braunschweig.
de, libo.yan@wki.fraunhofer.de (L. Yan).
potential combination of low cost, light weight, and high strength
and stiffness for structural application [3,11,13,14], which is a
promising alternative to glass bre in engineering applications,
such as civil [8,15e21] and automotive [22e25]. Most recently, ax
bres are also used with other bres in hybrid structures for
different engineering applications. The studies on hybridization of
ax bres with basalt [46], carbon [47] and glass bres [48e51]
show promising structural performance of ax bre reinforced
polymer composites in the aspects of durability, mechanical
properties and re performance. To better utilize natural bre
reinforced polymer composites (i.e. ax) in engineering elds, a full
understanding the knowledge of the mechanical properties of their
composites will be necessity, which can be achieved with a help of
computer simulation [26]. The use of numerical simulation can
create different components using such materials in shorter time
and less expensive, if appropriate numerical models can be created.
Since the natural bres (i.e. ax) are embedded in the polymer
matrix in their composites, it is quite difcult to measure the crack
initiation and propagation, damage development and failure
mechanisms by experimental studies on a micro-scale level, thus,

http://dx.doi.org/10.1016/j.compositesb.2015.11.038
1359-8368/ 2015 Elsevier Ltd. All rights reserved.
144 J. Sliseris et al. / Composites Part B 89 (2016) 143e154
the use of simulation tool will be helpful to determine the pro-
gressive failure of the composites [27].
Flax short bre is one of the most commonly used congura-
tions in polymer composites. Flax short bre reinforced polymer
composites made by extrusion, resin transfer and injection molding
techniques have become popular in automotive industry for
different components mainly because of their light-weight, excel-
lent energy absorbing properties, sustainability and low costs
[13,14]. For example, ax bre reinforced polypropylene (PP)
composites were produced by injection molding using short ax
bers with length up to 25 mm and average diameter about 30 mm
[13]. Tensile tests show that the failure mechanisms of short ax/PP
composites on micro-scale are ber breakage, bre debonding,
matrix cracking and split of ber bundles [28e30]. Semi-analytical
approach with orientation averaging technique has been used to
numerically model the mechanical properties ax bers and their
short bre reinforced composites [31]. In this method, it was
necessary to create a unit-cell single ber nite element model.
Then, the result of single ber model was used to obtain the me-
chanical behavior of multi-ber model by orientation averaging
technique. This technique did not include the effects such as split of
ber bundles, bre interlocking, bre defects and multi-axial stress
state in the modeling program. To have a more accurate modelling
result, indeed, it is important to take the randomness in micro-
structure and the actual geometry with defects of the bres into
account. In the work done by Mattrand et al. [32], the randomness
Fig. 1. Flow chart of representative volume element (RVE) generation for short ax ber reinforced polymer composite, where N(x,y)e random number according to normal
distribution with mean value x and standard deviation y, U(x,y)e random number according to uniform distribution in interval [x,y]. Only ber bundle generation algorithm is
shown explicitly. However ber and ber defect generation algorithm is analogical.
of ax bre cross section was modeled by randomized version of
the Fourier expansion of its complex coordinate function. In addi-
tion, two bre interface models were developed by using cohesive
zone elements. However, the model [33] was very computationally
intensive and worked ne only on few bre models, which cannot
represent the actual stochastic microstructure of ax bre rein-
forced polymer composite. Moreover, for short ax bre reinforced
polymer composites produced injection molding, the bre orien-
tation is governed by ow direction, which does not form a periodic
microstructure. Thus, appropriate constitutive models for ax bre,
interface zones, bre bundles and matrix were not well docu-
mented in the literature.
Flax fabric is another conguration which is widely used in
polymer composites because the woven fabric allows the control of
bre orientation and quality control, good reproducibility and high
productivity [34]. For ax fabric reinforced polymer composites,
several numerical modeling studies can be found in the literature,
e.g. Refs. [35,36]. Liang et al. [35] modelled the tensile properties of
unidirectional ax fabric reinforced laminates. Poila ^ne et al. [36]
modelled the viscoelastic behavior of unidirectional ax fabric
polymer composite successfully. In addition, the simulation of vi-
bration damping of multi-layer ax-carbon composite in the frame
of plate theory was also documented by Assarar et al. [37]. Flax
bres and PP or epoxy matrix have highly nonlinear material
behavior. Even for small strains the experimental results show a
nonlinear stress-strain curve [30]. In despite of many attempts to
 J. Sliseris et al. / Composites Part B 89 (2016) 143e154 145

Fig. 2. Tetrahedron discretization of ax ber composite.

model ax fabric reinforced polymer composites, a consistent non-
linear numerical model on micro-scale level cannot be found in the
literature.
In this study, we develop micromechanical numerical model for
both ax short bre-reinforced PP composite and ax bre fabric-
reinforced epoxy composite, respectively. The numerical model
consists of stochastic bre generator, including bre defects and
bre bundles effects. Appropriate constitutive models for ax -
bres, interface zones and matrix are also proposed. To avoid non-
convergence and mesh dependency problems, the non-local the-
ory is applied.
2. Methods
2.1. Finite element model
The stochastic structures of ax bre are generated by algo-
rithm, which generate bres, bre bundles and bre defects
randomly. Fig. 1 gives the ow chart of representative volume
element (RVE) generation for short ax bre reinforced polymer
composite and Fig. 2 shows the tetrahedron discretization of ax
short bre reinforced polymer composite. Fibres are not allowed to
overlap. Fibre length and bre diameter are specied with normal
distribution functions as input parameters. A random defect on the
ax bre is generated, where the average distance between defects
are specied. The algorithm generates the necessary number of
bres and bre bundles by specifying the bre solid volume frac-
tion (SVF). The bre bundle always consists of seven bers with a
ber diameter of 23% of bundle diameter. This geometry of bre
bundle allows to avoid poor mesh problems, if the distance be-
tween adjacent bers is too small.
The geometry is discretized by 4-node tetrahedron nite ele-
ments (Fig. 2). The mesh quality is controlled by limiting the
maximal aspect ratio of tetrahedron to 5. The aspect ratio is pro-
portional to the ratio of maximal edge length and radius of tetra-
hedron [38]. The tensile load is applied by applying a constant
displacement at one side of the sample and the opposite side is
xed. The symmetry conditions are used for the representative
volume element of ax/epoxy composite.
Geometry of short ber reinforced composites and ax fabric
reinforced composites is generated by creating a python script
within Salome-Meca 2015 software [38]. Fig. 3 shows the geometry

Fig. 3. Geometry of ax ber and ber bundle.

146 J. Sliseris et al. / Composites Part B 89 (2016) 143e154

The resins are modeled by Von Mises plasticity model [42] with
isotropic hardening given by curve of tension. The deformation
consists of elastic, plastic and temperature caused parts:

e p th (1)
The rate of plastic deformation is calculated by solving the
problem:

3 e s 1
_ p p_ _  A_ s  _ th ; (2)
2 seq

seq  Rp  0; (3)

p_ 0; if seq  Rp < 0; (4)

Fig. 4. Material models for resin-, ber- and ber defects-damage in a ber cell wall.
p_  0; if seq  Rp 0; (5)

where p cumulated equivalent plastic strain, R(p) hardening func-

of ax ber and ber bundle. The geometry module in this software q
generates primitives-points, lines, surfaces and solids. Afterwards, tion, e
s s  13 skk dij deviatoric stress, seq 32 e
sij esij equivalent
the crop, cutting and gluing operation with primitives have been stress of Von Mises, A elasticity tensor.
done. To specify different material properties and boundary con- The hardening function is dened as below:
ditions, a specic group for each geometrical object is assigned. The
nonlinear, quasi-static mechanical problem is solved by Code_Aster si1  si
Rp si p  pi ; p2pi ; pi1 : (6)
solver STAT_NON_LINE [39]. This solver search for balance of in- pi1  pi
ternal and external forces. The nonlinear system of equation is
The tangential matrix is obtained by solving equation
solved by using NewtoneRaphson method [40].

2.2. Non-linear material models
The main non-linearity in mechanical behavior of ax bre
reinforced polymer composite comes from the material non-linear
behaviors of ax bre and the resins. According to previous study
[41], PP or epoxy resins can be modeled with a non-linear plasticity
model. However, the tensile test of single ax ber showed a brittle
behavior and the damage usually occurs in the ber defects [30].
The ax ber bundles tend to split in separate bers, when the
tensile transversal load is applied to ber bundle [30]. These in-
uences are taken into account in the development of material
constitutive models in this study (Fig. 4).
_
s_ eq  Rp 0: (7)
Defects of ax bers and interface between ber bundles are
modeled by using a brittle material law with the continuum
damage mechanics. The state of material can be characterized by
the strain and damage d [42]. In the case of totally damaged
material d 1, and in the case of non-damaged d 0.
The stressestrain relationship of ax ber and ax ber defects
is linear and the stiffness is reduced linearly by damage variable d
s 1  dE : : (8)
In addition, the evolution of the damage is always increasing
and controlled by the following function threshold:

Table 1
Mechanical properties of ax bers and matrix.

Material Parameter Value Material Parameter Value

Flax ber Constitutive law Linear elastic, isotropic Polypropylene Constitutive law Plasticity with isotropic hardening
Young's modulus 54 GPa [8] E 1.627 GPa [30]
Poisson's ratio 0.2 Poisson's ratio 0.35 [29]
Defects of ax bers Constitutive law Brittle material law sy 1.627 MPa [30]
E 54 GPa s1,1 5.227, 0.005a
Poisson's ratio 0.2 s2,2 7.927, 0.01a
ET 2000 MPab s3,3 9.927, 0.015a
sy 50 MPab s4,4 11.427, 0.02a
Lc 1.0 s5,5 12.427, 0.025a
Interface zones in ber bundles Constitutive law Brittle material law s6,6 12.927, 0.03a
E 1.627 GPa
Poisson's ratio 0.35
ET 300 MPab
sy 2 MPa
Lc 1.0
a
Obtained by tting experimental data shown in Ref. [30].
b
Values obtained by tting experimental data [44] of single ber tensile test and also taking into account numerical convergence issues that does not allow to use too big
negative Young's modulus.
 J. Sliseris et al. / Composites Part B 89 (2016) 143e154 147

Z (   2 )
 2 1 2 1 v v v
1 1g Q min  Lc dU; (13)
f ; d : E :  w ; (9) 2 2 vx vy vz
2 1gd U

where coefcients w and g are both material characterization
constants. They can be obtained directly from tensile stress-strain
curve of single ber.
The coherence conditions completely determines the rate of
damage:
where Lc the characteristic length, U is the volume which is po-
tential localization of cracks and the non-local theory is applied. In
this case, U is volume, where brittle material law is dened e de-
fects of ax bres and the interface between bre bundles. The
higher value of characteristic length the less local behavior is
obtained.

f ; d  0; (10) 3. Results and discussion

In Sections 3.1e3.3, the simulation results of ax short ber

d_  0;
_ ; d 0:
df
The progressive damage and plastic deformation growth in
matrix or bers can lead to serious numerical convergence prob-
lems. The progressive damage growth and negative Young's
modulus (negative slope in the stress-strain curve) can lead to loss
of ellipticity in inelastic problems of ax composite. The damage
variable is computed in discrete integration points that leads to
mesh dependency. To avoid of these problems, we use a nonlocal
formulation with regularized strains [43].
The regularized strains are computed by solving additional
minimization problem:
(11) reinforced PP composites are presented. The effects of ax ber
orientation, ber length and diameter ratios on the material
properties of the short ax/PP composites are analyzed. The
simulation results of ax fabric reinforced epoxy composites are
(12)
presented in Section 3.4.
3.1. Effect of ber orientation
The numerical parameters and mechanical properties of ax
ber and matrix used for ax short bre/PP composites are listed in
Table 1.
Fiber orientation plays a crucial role in the mechanical proper-
ties of ax short ber reinforced polymer composite. The random
microstructures with different ber orientations are generated as
shown in Fig. 5. The size of representative volume elements (RVEs)

Fig. 5. Microstructures of short ax bres in matrix: (a) orientation coefcient 0.32, (b) orientation coefcient 0.59, (c) orientation coefcient 0.72 and (d) orientation coefcient
0.84.
148 J. Sliseris et al. / Composites Part B 89 (2016) 143e154

Fig. 6. Young's modulus of short ber reinforced ax/PP depending of ber orientation.

is 2 mm, which is larger than the maximal length of the ax ber of
1.9 mm, by doing so, there are no bers that completely cross the
RVE. The average diameter of ax ber used is 73.4 mm [30] and the
ax ber solid volume fraction is 13.7%.
The ber orientation is measured as the ratio of ber orientation
vector projections on coordinate axis (Figs. 3 and 5). The isotropic
orientation corresponds to an orientation coefcient of 0.33. If the
bers are completely oriented in Z direction, then the orientation
coefcients in X,Y,Z directions are 0, 0, 1, respectively (we specify
only the main orientation, so the orientation coefcient is 1 for this
microstructure).
In short ax bre reinforced polymer composites, a precise
measurement of the ber orientation is impossible due to the
complicated shape of the bre in the polymer matrix. However, it
has shown that the short ax/polymer composites manufactured
by injection molding have more oriented bers at skin layers and
less at the core due to the directions of the main streamlines [30].
Fig. 6 shows the Young's modulus (in linear elastic case) as a
function of the ber orientation for short ax/PP composites by
injection molding. We also plot the experimentally measured
Young's modulus for the bre reinforced polymer composite at the
skin and core layers given in Ref. [30]. Fig. 6 indicates that the ber
orientation coefcient at the core layer is around 0.46 and that at
the sink layer is 0.63, respectively.
Then, the microstructure with ber orientation coefcient of
0.63 is generated, corresponding to the skin layer and after that the
nonlinear tensile stress-strain curves of the short ax bre/PP
composites is plotted and given in Fig. 7. As shown in the gure,
there is a good agreement between the simulated and the experi-
mental tensile stress-strain curve of the short ax/PP composites
presented in Ref. [30].
3.2. Effect of ber length/diameter ratio on stiffness
The ratio of ber length-and-diameter also has a great impact on
mechanical properties of short ax ber reinforced polymer

Fig. 7. Flax/PP composite with oriented bers in skin layer (experiment values obtained from Ref. [30]).
 J. Sliseris et al. / Composites Part B 89 (2016) 143e154 149

Table 2 3.3. Progressive failure and plastic strains

Properties of ax ber and microstructure representative element.

L/D D (mm) L (mm) RVE size (mm) Orientation SVF To design ax bre reinforced polymer composites with favor-
10 30 300 600  600  600 0.18:0.19:0.63 0.1379
able properties, it is important for researchers to have a good un-
25 30 750 600  600  1500 0.19:0.20:0.61 0.1364 derstand on its damage initiation, development and failure
50 30 1500 400  400  3000 0.21:0.18:0.61 0.1367 mechanisms. In reality, it is impossible to record the cracking
initiation, damage development, and plastic deformation of ax
bre reinforced polymer composites by experimental study on a
composites. As this ratio decreases, the mechanical properties of microscale level because the bres are embedded in the polymer
the composites also decrease. Up to a certain level, there is an in- matrix. Hence, the numerical results of damage and plastic de-
crease in mechanical properties with an increase of length-to- formations of short ax bre reinforced polymer composites are
diameter ratio [13]. Stochastic microstructures with ax bres presented. Figs. 9e11 show the different damage mechanisms of
with various ber length-to-diameter ratios are generated. The the short ax/PP composites using the numerical modelling.
geometry of bers, dimensions of RVEs, ber orientation and solid Generally, it is realized that the plastic deformation of the short
volume fraction of bers are shown in Table 2. The relationship ax/PP composites initiates at the ber endings and around the
between the Young's modulus of the composites and ber length- ber defects [30]. As shown in Figs. 9 and 10, these failure mech-
to-diameter ratios is illustrated in Fig. 8. The considered micro- anisms are successfully captured by the simulation.
structures of the ax bres have following length-to-diameter ra- Natural bres such as ax have many defects and imperfections
tios: 10, 25 and 50. As illustrated in Fig. 8, the comparison with the along the ber length [13,28,45]. Normally, the damage of bre
experimental results [30] show a good agreement in the modulus of initiates at the defect points of the bre [13,28]. To capture this
elasticity. effect, the randomly distributed ber defects are generated in the

Fig. 8. Modulus of elasticity depending on ratio of ber length and diameter.

Fig. 9. Failure mechanisms 1: plastic strains at ber endings: (a) at 0.6% strains, (b) at 1.8% strains, (c) at 3% strains.
150 J. Sliseris et al. / Composites Part B 89 (2016) 143e154

Fig. 10. Failure mechanism 2: damage in ber defects (SEM photo is reproduced from Ref. [30]).

numerical simulation. The brittle material constitutive law is used
in the simulation of ber defects. The average distance between
ber defects are studied by Andersons et al. [45] and this value is
used as input parameter to the random ber generator in this
study. The ber defects are shown in Fig. 10(b).
The conguration of ber bundles also has an inuence on the
deformation mechanisms of short ax/PP composites. Experi-
mental investigations shown that ax ber bundles can be easily
cracked [30], as shown in Fig. 11. In the numerical model, the ber
bundles consist of seven individual bers. The interface zone be-
tween the bers are modeled by a brittle material law. As shown in

Fig. 11. Failure mechanism 3: split of ber bundles.

 J. Sliseris et al. / Composites Part B 89 (2016) 143e154 151

Fig. 12. Microstructure failure mechanisms at different stress-strain states in the tensile response of short ax/PP composite. The red color is ber and ber bundle damage and the
blue color indicates plastic deformation growth. (For interpretation of the references to colour in this gure legend, the reader is referred to the web version of this article.)

the modeling result, the damage of interface zone in the ber

Table 3 bundles is clear, further damage and plastic deformation growth in
Mechanical properties of epoxy and ax fabric based on [4].
matrix near the ber bundles are observed.
Properties Epoxy Flax fabric Fig. 12 shows the damage development at different strain-stress
Tensile modulus, GPa 3.89 16.25 states of short ax/PP composites. The rst damage appears at a
Tensile strength, MPa 78 145.55 strain of 0.1% due to the ber and ber bundle defects. Afterwards,
Tensile strain at failure, % 3.5 3.15 with an increase in applied load, plastic deformation in the matrix
Poisson's ratio 0.35 0.2 near ber damage and ber endings is developed.

Fig. 13. Stressestrain plot of single ax yarn and epoxy.

152 J. Sliseris et al. / Composites Part B 89 (2016) 143e154

Fig. 14. Finite element model of ax-epoxy composite.

Table 4
Simulation and experimental results of ax fabric/epoxy composite.

Value Experimental results Simulation results Difference (%)

Tensile modulus, GPa 8.7 8.738 0.4

Tensile strength, MPa 108.0 118.7 9.9
Transversal Young's modulus, GPa e 5.7
In-plane shear modulus, GPa e 6.3
Out of plane shear modulus, GPa e 4.0
In-plane Poisson's ratio e 0.24
Out of plane Poisson's ratio e 0.37

3.4. Flax fabric/epoxy composite
The numerical simulations of ax fabric reinforced epoxy com-
posites are based on the experimental investigations done by Yan
et al. [4]. The ax fabric has a plain weave structure with a density
of 550 g/m2 and a yarn diameter of 1 mm. Flax yarns are virtually
generated and simulated. The ax yarn middle line is assumed to be
a sinusoidal function. Flax yarns have circular cross section. In the
contact zones of ax yarns, a little overlapping of yarns is allowed.
Other parameters of the ax fabrics and the epoxy resin are shown

Fig. 15. Experimental and simulated tensile stress-strain curves of ax fabric/epoxy composites.
 J. Sliseris et al. / Composites Part B 89 (2016) 143e154
Fig. 16. Failure zones of ax fabric/epoxy composite at 2.5% strain. Tensile load is
applied in x direction.
in Table 3. The tensile stress-strain curves of single-strand ax yarn
and the epoxy resin are illustrated in Fig. 13 [4], which are used as
input parameters for the numerical simulation. Flax yarns and
epoxy are modeled by using a non-linear plasticity model with an
isotropic hardening law, which is described in the tensile stress-
strain curve. Flax fabric/epoxy composite with a single layer of
fabric is analyzed. The thickness of the composites 1.625 mm and
the ber volume fraction is 60%. The nite element model of RVE is
discretized by tetrahedron 4-node nite elements (Fig. 14). Con-
stant static deformations are applied in direction of yarns. The
symmetry conditions are applied to the sides of RVE.
The rst simulation is performed to calculate the linear elastic
properties of the composite. We homogenized the material using a
volume averaging technique. As shown in Table 4, the difference of
tensile Young's modulus of the ax fabric/epoxy composite be-
tween the numerical modeling and the experimental result is less
than 1%. The homogenization also gives the transversal isotropic
properties of the composite. Other elastic constants, which are
difcult to measure experimentally, are shown in Table 4. Then,
non-linear simulation of the quasi-static tensile test of ax fabric/
epoxy composites up to 2.5% strain is performed. The obtained
tensile stress-strain curve by numerical modelling showed a nearly
perfect agreement with the experimental curve within the strain
up to 2%, as given in Fig. 15. The calculated tensile strength by the
numerical modelling is around 10% larger than that obtained from
the experiment (Table 4).
As shown in Fig. 15, the damage of plastic deformation initiation
zones is located on the surface of yarns and starts to grow at a strain
of 0.25%. The damage and possible crack growth is perpendicular to
the applied load (Fig. 16).
4. Conclusions
In this study, numerical methods to generate and simulate
mechanical properties of short ax ber reinforced polymer com-
posite and ax fabric reinforced polymer composite are proposed.
The microstructures of the short bres are generated including -
ber defects and ber bundles effect. A brittle material law for ber
defects and interface zones of ber bundles is proposed for the
simulation. The polypropylene and epoxy matrices used for short
153
ax bre and ax fabric reinforcements are modeled by a non-
linear plasticity model with an isotropic hardening law. The nu-
merical results can capture the main damage mechanisms such as
ber breakage at ber defects, matrix damage and damage of the
cohesive zone quite well. Good agreements between simulation
results and experimental results of elastic properties and non-
linear tensile stress-strain curves up to ultimate strains are
achieved.
References
[1] Ku H, Wang H, Pattarachaiyakoop N, Trada M. A review on the tensile prop-
erties of natural bre reinforced polymer composites. Compos Part B
2011;42(4):856e73.
[2] Yan L, Su S, Chouw N. Microstructure, exural properties and durability of coir
bre reinforced concrete beams externally strengthened with ax FRP com-
posites. Compos Part B 2015;80:343e54.
[3] Dittenber DB, GangaRao HVS. Critical review of recent publications on use of
natural composites in infrastructure. Compos Part A 2012;43:1419e29.
[4] Yan L, Chouw N, Yuan XW. Improving the mechanical properties of natural
bre fabric reinforced epoxy composites by alkali treatment. J Reinf Plast
Compos 2012;31(6):425e37.
[5] Koronis G, Silva A, Fontul M. Green composites: a review of adequate mate-
rials for automotive applications. Compos Part B 2013;44(1):120e7.
[6] Yan L, Chouw N, Jayaraman K. Effect of UV and water spraying on the me-
chanical properties of ax fabric reinforced polymer composites used for civil
engineering application. Mater Des 2015;71:17e25.
[7] Yan L. Plain concrete cylinders and beams externally strengthened with nat-
ural ax fabric reinforced epoxy composites. Mater Struct 2015. http://
dx.doi.org/10.1617/s11527-015-0635-1.
[8] Yan L, Chouw N, Jayaraman K. Lateral crushing of empty and polyurethane-
foam lled natural ax fabric reinforced epoxy composite tubes. Compos
Part B 2014;63:15e26.
[9] Kabir MM, Wang H, Lau KT, Cardona F. Chemical treatments on plant-based
natural bre reinforced polymer composites: an overview. Compos Part B
2012;43(7):2883e92.
[10] Yan L, Chouw N. Effect of water, seawater and alkaline solution ageing on
mechanical properties of ax fabric/epoxy composites used for civil engi-
neering application. Constr Build Mater 2015;99:118e27.
[11] Shah DU, Schubel PJ, Clifford MJ. Can ax replace E-glass in structural com-
posites? A small wind turbine blade case study. Compos Part B 2013;52:
172e81.
[12] Tan H, Yan L, Huang L, Wang Y, Li H. Behavior of sisal ber concrete cylinders
externally wrapped with jute FRP. Polym Compos 2015. http://dx.doi.org/
10.1002/pc.23761.
[13] Yan L, Chouw N, Jayaraman K. Flax bre and its composites e a review.
Compos Part B 2014;56:296e317.
[14] Pickering KL, Aruan Efendy MG, Le TM. A review of recent developments in
natural bre composites and their mechanical performance. Compos Part A
2015. http://dx.doi.org/10.1016/j.compositesa.2015.08.038.
[15] Assaedi H, Alomayri T, Shaikh FUA, Low IM. Characterization of mechanical
and thermal properties in ax fabric reinforced geopolymer composites. J Adv
Ceram 2015. http://dx.doi.org/10.1007/s40145-015-0161-1.
[16] Yan L, Chouw N, Jayaraman K. Effect of column parameters on ax FRP
conned coir bre reinforced concrete. Constr Build Mater 2014;55(31):
299e312.
[17] Yan L, Chouw N. Natural FRP tube conned bre reinforced concrete under
pure axial compression: a comparison with glass/carbon FRP. Thin Walled
Struct 2014;82:159e69.
[18] Xia YY, Xian GJ, Kafodya I, Li H. Compression behavior of concrete cylinders
externally conned by ax bre reinforced polymer sheets. Adv Struct Eng
2014;17(12):1825e33.
[19] Yan L, Chouw N. A comparative study of steel reinforced concrete and ax
bre reinforced polymer tube conned coconut bre reinforced concrete
beams. J Reinf Plast Compos 2013;32(16):1e10.
[20] Yan L, Chouw N, Jayaraman K. On energy absorption capacity, exural and
dynamic properties of ax/epoxy composite tubes. Fibers Polym 2014;15(6):
1270e7.
[21] Cevallos OA, Olivito RS, Codispoti R, Ombres L. Flax and polyparaphenylene
benzobisoxazole cementitious composites for the strengthening of masonry
elements subjected to eccentric loading. Compos Part B 2015;71:17e25.
[22] Meredith J, Ebsworth R, Coles SR, Wood BM, Kirwan K. Natural bre com-
posite energy absorption structures. Compos Sci Technol 2012;72:212e7.
[23] Meredith J, Coles SR, Poe R, Coing E, et al. On the static and dynamic properties
of ax and cordenka epoxy composites. Compos Sci Technol 2013;80:31e8.
[24] Yan LB, Chouw N, Wood BM, Karwar K. Crashworthiness characteristics of ax
bre reinforced epoxy tubes for energy absorption application. Mater Des
2013;51:629e40.
[25] Yan LB, Chouw N, Jayaraman K. Effect of triggering and polyurethane foam-
ller on axial crushing of natural ax/epoxy composite tubes. Mater Des
2014;56:528e41.
154 J. Sliseris et al. / Composites Part B 89 (2016) 143e154
[26] Sliseris J, Andra H, Kabel M, Plinke B, Wirjadi O, Frolovs G. Numerical pre-
diction of the stiffness and strength of medium density berboards. Mech
Mater 2014;79:73e84.
[27] Sliseris J, Rocens K. Optimal design of composite plates with discrete variable
stiffness. Compos Struct 2013;98:15e23.
[28] Bos HL, Van Den Oever MJA, Peters OCJJ. Tensile and compressive properties of
ax bres for natural bre reinforced composites. J Mater Sci 2002;37:
1683e92.
[29] Baley C. Analysis of the ax bres tensile behaviour and analysis of the tensile
stiffness increase. Compos Part A 2002;33:939e49.
[30] Bourmaud A, Ausias G, Lebrun G, Tachon M, Baley C. Observation of the
structure of a composite polypropylene/ax and damage mechanisms under
stress. Indus Crops Prod 2013;43:225e36.
[31] Modniks J, Andersons J. Modeling the non-linear deformation of a short-ax-
ber-reinforced polymer composite by orientation averaging. Compos Part B
2013;58:188e93.
[32] Mattrand C, Be akou A, Charlet K. Numerical modeling of the ax ber
morphology variability. Compos Part A 2014;63:10e20.
[33] Charlet K, Be akou A. Mechanical characterization and modeling of interfacial
lamella within a ax bundle. Proc Eng 2011;10:906e11.
[34] Yan L. Effect of alkali treatment on vibration characteristics and mechanical
properties of natural fabric reinforced composite. J Rein Plast Compos
2012;31:887e96.
[35] Liang SX, Gning PB, Guillaumat L. Quasi-static behaviour and damage
assessment of ax/epoxy composites. Mater Des 2015;67:344e53.
[36] Poil^ane C, Cherif ZE, Richard F, Vivet A, Ben Doudou B, Chen J. Polymer
reinforced by ax bres as a viscoelastoplastic material. Compos Struct
2014;112:100e12.
[37] Assarar M, Zouari W, Sabhi H, Ayad R, Berthelot JM. Evaluation of the damping
of hybrid carboneax reinforced composites. Compos Struct 2015;132:
148e54.
[38] Salome-Meca. Finite element simulation environment including preprocess-
ing, post processing and solver. 2015. http://www.code-aster.org/V2/spip.
php?article303 [accessed on 06.09.15].
[39] Code_Aster 12.4. Finite element solver for solid mechanics. <http://www.
code-aster.org/V2/spip.php?article272> [accessed on 10.09.15].
[40] Broumand P, Khoei AR. X-FEM modeling of dynamic ductile fracture problems
with a nonlocal damage-viscoplasticity model. Finit Elem Analy Des 2015;99:
49e77.
[41] Spahn J, Andra H, Kabel M, Mller R. A multiscale approach for modeling
progressive damage of composite materials using fast Fourier transforms.
Comput Methods Appl Mech Eng 2014;268:871e83.
[42] Kossa A, Szabo  L. Exact integration of the von Mises elastoplasticity model
with combined linear isotropic-kinematic hardening. Inter J Plast 2009;25(6):
1083e106.
[43] Jackiewicz J, Kuna M. Non-local regularization for FE simulation of damage in
ductile materials. Comput Mater Sci 2003;28:684e95.
[44] Spahn Johannes. An efcient multiscale method for modeling progressive
damage in composite materials [PhD thesis]. Kaiserslautern University of
Technology; 2015.
[45] Andersons J, Porike E, Spa rnins E. The effect of mechanical defects on the
strength distribution of elementary ax bres. Compos Sci Technol
2009;69(13):2152e7.
[46] Fiore V, Scalici T, Calabrese L, Valenza A, Proverbio E. Effect of external basalt
layers on durability behaviour of ax reinforced composites. Comp Part B
2016;84:258e65.
[47] Fiore V, Valenza A, Di Bella G. Mechanical behavior of carbon/ax hybrid
composites for structural applications. J Compos Mater 2012;46:2089e96.
[48] Zhang YL, Li Y, Ma H, Yu T. Tensile and interfacial properties of unidirectional
ax/glass ber reinforced hybrid composites. Compos Sci Technol 2013;88:
172e7.
[49] Rao S, Bhardwaj A, Beenhag A, Bhattacharyya D. Fire performance of ax
laminates and their hybrids. Adv Mater Res 2012;410:114e7.
[50] Arbelaiz A, Fernandez B, Cantero G, Llano-Ponte R, Valea A, Mondragon I.
Mechanical properties of ax bre/polypropylene composites. Inuence of
bre/matrix modication and glass bre hybridization. Compos Part A
2005;36(12):1637e44.
[51] Morye SS, Wool RP. Mechanical properties of glass/ax hybrid composites
based on a novel modied soybean oil matrix material. Polym Compos
2005;26:407e16.