Analytical Estimation of Elastic Properties of Polypropylene Fiber Matrix Composite by Finite Element Analysis
Analytical Estimation of Elastic Properties of Polypropylene Fiber Matrix Composite by Finite Element Analysis
Analytical Estimation of Elastic Properties of Polypropylene Fiber Matrix Composite by Finite Element Analysis
= +
( )
12 f f m f
v v V v 1 V
1
(2)
Major Poissons Ratio
= + (3)
In-Plane Shear Modulus
( )
f f
12 f m
1 V V 1
G G G
= +
1 f f m m
E E V E V
(4)
2.1.2. Semi-Empirical Model (Halphin-Tsai)
Halphin and Tsai developed their models as simple equa-
tions by curve fitting to results that are based on elasticity.
The equations are semi-empirical in nature since involved
parameters in the curve fitting carry physical meaning.
Longitudinal Youngs Modulus
(5)
= +
Transverse Youngs Modulus
f
m 2 f
m f f
m
E
1
E E 1 V
,
E 1 V E
E
q
q
q
( (
( (
( +
(
= =
(
(
(
+ (
(
(
12 f f m m
v v V v V
(6)
The term is called the reinforcing factor and de-
pends on the following;
Fiber geometry, Packing geometry, Loading condition
Major Poissons Ratio
(7) = +
In-Plane Shear Modulus
f
m 12 f
m f f
m
G
1
G G 1 V
,
G 1 V G
G
q
q
q
( (
( (
( +
(
= =
(
(
(
+ (
(
(
(8)
2.2. Compliance Matrix
In composite material fibers may be oriented in an arbi-
trary manner. Depending on the arrangements of the fi-
bers, the material may behave differently in different
directions. According to their behaviour, composites may
be characterized as generally anisotropic, monoclinic,
orthotropic, and transversely isotropic. In this paper,
transversely isotropic characteristics have been consid-
ered for the fiber reinforced composite and fiber ar-
rangement as shown in the Figure 1.
3
2
1
Figure 1. Arrangement of fiber direction for transversely
sotropic composite. i
Copyright 2012 SciRes. AMPC
B. PAL ET AL.
Co es. AMPC
25
( )
x 1 12 2 12 2
y 12 1 2 23 2
z 12 1 23 2 2
23 23 2
13 12
12
1 E E E 0 0
E 1 E E 0 0
E E 1 E 0 0
0 0 0 2 1 E 0
0 0 0 0 1 G
0 0 0 0 0
23
13
12 12
0
0
0
0
0
1 G
x
y
z
o
o
o
t
t
t
( (
( (
( (
( (
=
( (
( (
( (
( (
( (
e (
(
e
(
( e
(
+
(
(
(
(
Here 1-2-3 orthogonal coordinate system is used where
the directions are taken as follows:
v
1
dV
V
ij ij = e
}
(10)
e
The 1-axis is aligned with the ber direction.
3. Computational Details
The 2-axis is in the plane of the layer and perpen-
dicular to the bers.
In this present work, finite element method is used to
approximate the different elastic property of the fiber
reinforced composites by ANSYS 12.
The 3-axis is perpendicular to the plane of the layer
and thus also perpendicular to the bers.
Stress strain relationship for compliance matrix for
transversely isotropic matrix given below.
Assumptions made for the present analysis were
The composite is
Macroscopically homogeneous;
2.3. Representative Volume Element (RVE)
Linearly elastic;
Primary to use of numerical approximations of the effec-
tive properties of composite is the concept of representa-
tive volume element (RVE). Square or cubic RVEs are
used for most numerical approximations because of the
ease of numerically solving boundary values problems
with these geometries. The difficulties involved in gen-
erating statistical information about particle distributions
and concentrations leads to difficulties in the rigorous
determination of RVE sizes. Hence, for most applications,
RVE sizes have been rather arbitrary. In this paper a
RVE model of 420 m 420 m has considered which
consists of different volume fraction of fiber density. In
the model calculations, cuboid will be matrix and cylin-
der will be fiber. Figure 2 shows the typical RVE which
consists of 10%, 17%, 27%, 40% and 50% of fiber con-
tent.
Macroscopically transversely isotropic;
Initially stress free (no thermal stress).
The fibers are:
Homogeneous;
Linearly elastic;
Isotropic;
Regularly spaced;
Perfectly aligned.
The matrix is:
Homogeneous;
Linearly elastic;
Isotropic.
3.1. Modelling
A regular three-dimensional arrangement of short fibre in
a matrix was adequate to describe the overall behaviour
of the composite, was modelled as a regular uniform ar-
rangement, as shown in Figure 1. This model assumed
that the fibre was a perfect cylinder of length 250 m,
and diameter (d = 50 m) in a cube (420 420 250
m
3
) of matrix. It is assumed that the geometry, material
and loading of the unit cell are symmetrical with respect
to x-y-z coordinate system as shown in Figure 2. There-
fore, 10% volume fraction i.e. 9 fibers has been inserted
in a cubic matrix (420 420 250 m
3
) uniformly as
shown in Figure 2. 3D finite element meshing for the
10% volume fraction of the fiber has been done, simi-
larly the finite element modelling and meshing were
done by varying the fiber volume fraction from 10% to
54% and as shown in Figure 3.
2.4. Homoginization
In classical lamination theory the composite lamina is
modelled as a homogeneous orthotropic medium with
certain effective moduli that describe the average ma-
terial properties of the composite. To describe this mac-
roscopically homogeneous medium, macro-stress and
macro-strain are derived by averaging the stress and
strain tensor over the volume of the RVE. The average
stress and strain quantities defined in Equations (9) and
(10) thus ensure equivalence in strain energy between the
equivalent homogeneous material and the original het-
erogeneous material. These average quantities will be
used in the subsequent analysis to determine composite
moduli
3.2. Element Type
v
1
dV
V
ij o o =
}
ij (9)
The element SOLID 45 of ANSYS V12.0 used for the
pyright 2012 SciR
B. PAL ET AL. 26
10% 17% 27% 40% 54%
Figure 2. Unit cell of square array fibre packing geometry for RVE model.
10% 17% 27% 40% 54%
Figure 3. Finite element mesh for 10% volume fraction.
present analysis which is based on a general 3D state of
stress and is suited for modeling 3D solid structure under
3D loading. The element has 8 noded brick element with
three degrees of freedom per node (UX, UY and UZ).
3.3. Boundary Condition
In this work the boundary condition with normal strain
applied in x direction are as follows [11]:
( ) ( ) ( ) ( ) u LF 0, v BF 0, w BKF 0 and u RF o = = = =
( )
x
x x
y
z
u 0, y, z 0
5.0E 05
0
onstant
0
constant
where LF, BF, BKF and RF stand for left face, bottom
face, back face and right face of the RVE model. All
other faces are free of any displacement.
In this paper, axial loading is modelled by a displace-
ment acting on the plane yz at x. For such loading condi-
tions, the boundaries of the RVE also correspond to lines
of symmetry. Thus, normal displacements of the bounda-
ries of the quadrant are restricted to those that cause the
boundary to displace only parallel to the original bound-
ary.
The displacement constraints applied to the finite ele-
ment model to determine E
1
are as follows and as shown
in Figure 4.
( )
( )
( )
( )
( )
y
y
z
z
u x, y, z
u x, 0, z
u x, y, z c
u x, y, 0
u x, y, z
o
o
o
=
= =
= =
= =
=
=
3.4. Material Property
The different material properties of both matrix and fiber
has shown in the Table 1.
Material properties are used to determine the several
elastic properties of the composite that is Longitudinal
youngs modulus (E
1
) and Transverse youngs modulus
(E
2
) by varying the volume fraction of the fiber.
3.5. Sample Analysis
Figure 5 shows the analysis of 40% of volume fraction
of fiber reinforced composite in which the von misses
stresses for the matrix fiber and composite has been in-
dicated. The stresses are obtained by applying the dis-
placement loading in longitudinal direction.
4. Results and Discussion
4.1. Longitudinal Youngs Modulus (E
1
)
Figure 6 shows comparison of finite element data, rule
of mixtures and Halphin-Tasi results for composite mo-
dulli E
1
at different volume fractions. The finite-element
solution gives identical results to the rule of mixtures and
semi empirical analytical formulation. The linear de-
pendence of E1 on fibre volume fraction is demonstrated
and, as expected, the modulus increases while increasing
the fibre volume fraction.
4.2. Transverse Youngs Modulus (E
2
)
Figure 7 shows comparison of finite element data, rule
of mixtures and Halphin-Tasi results for composite mo-
dulli E
2
at different volume fractions. Finite element re-
sults are as expected; on the assumption that the compos-
ite is macroscopically transversely isotropic the values
obtained for the whole unit cells investigated are per-
fectly coincident. The Rule of mixtures model, the Hal-
pin-Tsai equation and data from the finite-element cal-
culations are compared. The term is called the rein-
forcing factor in Halpin Tsai equation and depends on the
fiber geometry, packing geometry, loading condition. As
he fibers are circular and packing density is increasing t
Copyright 2012 SciRes. AMPC
B. PAL ET AL. 27
Figure 4. Boundary condition for longitudinal modulus of composite in x direction.
Table 1. Shows the material properties of composites.
Material Properties Polypropylene co ethylene (PPE) Polypropylene (PP)
Longitudinal Modulus E
1
1.05 Gpa 4.5 Gpa
Transverse Modulus E
2
1.05 Gpa 4.5 Gpa
Poissons Ratio 0.33 0.2
ANSYS 12.0.1
ELEMENT SOLUTION
STEP = 1
SUB = 1
TIME = 1
SEQV (NOAVG)
DMX = 0.597E-04
SMN = 204.969
SMNB = 168.386
SMX = 909.209
SMXB = 1283
204.969
283.217
361.466
439.715
517.964
596.213
674.462
752.711
830.96
909.209
ANSYS 12.0.1
ELEMENT SOLUTION
STEP = 1
SUB = 1
TIME = 1
SEQV (NOAVG)
DMX = 0.587E-04
SMN = 904.134
SMNB = 898.861
SMX = 909.209
SMXB = 914.222
904.134
904.698
905.262
905.826
906.39
906.953
907.517
908.081
908.645
909.209
ANSYS 12.0.1
ELENENT SOLUTION
STEP = 1
SUB = 1
TINE = 1
SEQV (NOAVG)
DNX = 0.597E-04
SMN = 204.969
SMNB = 194.669
SMX = 212.096
SMXB = 221.558
204.969
205.76
206.552
207.344
208.136
208.928
209.72
210.512
211.304
212.096
Figure 5. Stress distribution of matrix, fiber and composite during computing of E
1
.
ROM
H-T Model
FEA
3.5
3
2.5
2
1.5
1
0.5
0
0 10 20 30 40 50 60
Volume Fraction in %
E
1
i
n
G
p
a
Figure 6. Comparison of finite element data, rule of mixtures and Halphin-Tasi results for composite modulli E
1
for different
volume fration of fiber.
Copyright 2012 SciRes. AMPC
B. PAL ET AL. 28
ROM
Halohin-Tsai
FEA
2.5
2
1.5
1
0.5
0
0 10 20 30 40 50 60
Volume Fraction in %
E
2
i
n
G
p
a
Figure 7. Comparison of finite element data, rule of mixtures and Halphin-Tasi results for composite modulli E
2
for different
volume fration of fiber.
accordingly the = 2 is adapted to consider those
changes in the composites. In this Halpin-Tsai equation
shows a good agreement between finite element values.
But in Rule of mixture there is no such parameter which
can be considered for those changes. Due to this reason
The Rule of mixture is not showing the good agreement
with finite element data.
4.3. Poissons Ratio
The major Poissons ratio
12
0 for the composite is de-
fined as minus the ratio of strain in the y-direction di-
vided by the strain in the x-direction when only the stress
x
o is applied. The Rule of Mixtures expression for the
composite Poissons ratio is similar in form to the ex-
pression for E
1
. It evident from the Figure 8 that ana-
lytical and numerical results are in perfect agreement
although data from the Rules of Mixtures are slightly
higher than the finite-element results.
The Poissons ratio
23
0 describes the contraction in
the z-direction on
23
0 applying loads in the y-direction.
No analytical formulas have been applied to compare
with numerical results of Poissons ratio. It is worthwhile
to show finite element results of both with different fibre
volume fractions. As expected from the Figure 9,
23
0
appears to be higher than
12
0 and the comparison be-
tween the two Poissons ratios shows that Poissons ra-
tios
12
0 are less sensitive to the fibre volume fraction.
In particular, the values of
12
0 are quite close to the
Poissons ratio of polypropylene (PP) especially at higher
fibre volume fraction. On the contrary,
23
0 is strongly
dependent from the fibre volume fraction and, its values
approaches the Poissons ratio of the Polypropylene co
ethylene at lower values of the fibre volume fraction.
5. Conclusions
1) The 3D unit cell of the unidirectional fiber rein-
forced composite has been identified (RVE) as a basic
building block for estimating overall mechanical com-
posite properties.
2) Various micromechanical methods to determine the
elastic behaviour of composite materials have been dis-
cussed.
3) Finite element analysis has provided an implicit
means of modelling polymer composites.
4) Numerical homogenization tools have been devel-
oped for the evaluation of the effective material proper-
ties of the short fiber composites.
5) The appropriate constraints on the RVE under vari-
ous loadings have been determined from symmetry
boundary conditions, obtaining a nearly complete set of
elastic constants for a three-dimensional unidirectional
composite.
6) The results of elastic moduli E
1
, E
2
, E
2
,
12
,
23
,
compared with the results of analytical solution and it is
found that the results from FE simulation are in good
agreement with the analytical results employed in this
exercise namely the Halpin-Tsai semi-empirical expres-
sion and to some extent the Rules of Mixtures.
7) As this mechanical property of fiber-filled compos-
ites are affected by a number of parameters such as fiber
type, matrix type, fiber orientation, fiber geometry, vol-
ume fraction of the fibers and the degree of interfacial
Copyright 2012 SciRes. AMPC
B. PAL ET AL. 29
Theoritical
FEA
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
0 10 20 30 40 50 60
Volume Fraction in %
V
2
3
Figure 8. Comparison of finite element data and rule of mixtures results for composite poissons ratio
12
0 .
FEA.V12
FEA.V23
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
0 10 20 30 40 50 60
Volume Fraction in %
V
1
2
a
n
d
V
2
3
Figure 9. Comparison of finite element results for composite poissons ratio
12
0 and
23
0 .
adhesion between the fiber and the polymer matrix, the
ROM has not been adopted by this property expect the
volume fraction. So the results are also not so accurate.
8) The Halpin-Tsai (HT) model is also a theoretical
model. This model, besides the modulus of elasticity of
the polymer, E
m
, and of the fiber, E
f
, includes a geomet-
rical parameter (aspect ratio) of the fiber as well. The
model has a complicated mathematical structure with
fiber geometry, packing geometry, loading condition. This
factors help to showing the good agreement with finite
element data.
9) It is observed that the change in volume fraction of
fiber has a significant effect on elastic properties.
REFERENCES
[1] D. F. Adams and D. R. Doner, Longitudinal Shear Load-
ing of a Unidirectional Composite, Journal Composite
Materials, Vol. 1, No. 1, 1967, pp. 4-17.
[2] D. F. Adams and D. R. Doner, Transverse Normal Load-
ing of a Unidirectional Composite, Journal Composite
Materials, Vol. 1, No. 2, 1967, pp. 152-164.
doi:10.1177/002199836700100205
[3] S. Houshyar, R. A. Shanks and A. Hodzic, Modelling of
Polypropylene Fibre-Matrix Composites Using Finite Ele-
ment Analysis, Express Polymer Letters, Vol. 13, No. 1,
2009, pp. 2-12.
[4] S. Houshyar, R. A. Shanks and A. Hodzic, The Effect of
Fibre Concentration on Mechanical and Thermal Proper-
ties of Fibre Reinforced Polypropylene Composites, Jour-
nal of Applied Polymer Science, Vol. 96, No. 6, 2005, pp.
2260-2272. doi:10.1002/app.20874
[5] G. Kalaprasad, K. Joseph and S. Thomas, Theoretical Mo-
delling of Tensile Properties of Short Sisal Fibre-Rein-
forced Low Density Polyethylene Composites, Journal
Copyright 2012 SciRes. AMPC
B. PAL ET AL. 30
of Materials Science, Vol. 32, No. 16, 1999, pp. 4261-
4267. doi:10.1023/A:1018651218515
[6] C. T. Sun and R. S. Vaidya, Prediction of Composite
Properties from a Representative Volume Element, Com-
posites Science and Technology, Vol. 56, No. 2, 1996, pp.
171-179. doi:10.1016/0266-3538(95)00141-7
[7] S. Li, General Unit Cell for Micromechanical Analyses
of Unidirectional Composites, Composites Part A, Vol.
32, No. 6, 2000, pp. 815-816.
doi:10.1016/S1359-835X(00)00182-2
[8] B. R. Kim and H. K. Lee, An RVE-Based Microme-
chanical Analysis of Fiber-Reinforced Composites Con-
sidering Fiber Size Dependency, Composite Structures,
Vol. 90, No. 4, 2009, pp. 418-427.
doi:10.1016/j.compstruct.2009.04.025
[9] P. De Buhan and A. Taliercio, A Homogenisation Ap-
proach to the Yield Strength of Compositematerials, Eu-
ropean Journal of Mechanics, Vol. 10, No. 2, 1991, pp.
129-154.
[10] Y. Pan, L. Iorga and A. A. Pelegri, Numerical Genera-
tion of a Random Chopped Fiber Composite RVE and Its
Elastic Pproperties, Composites Science and Technology,
Vol. 68, No. 13, 2008, pp. 2792-2798.
doi:10.1016/j.compscitech.2008.06.007
[11] M. Porfiri and N. Gupta, Effect of Volume Fraction and
Wall Thickness on the Elastic Properties of Hollow Parti-
cle Filled Composites, Composites Part B: Engineering,
Vol. 40, No. 2, 2009, pp. 166-173.
doi:10.1016/j.compositesb.2008.09.002
[12] K. Hbaieb, Q. Wang, Y. H. J. Chia and B. Cotterell, Mo-
delling Stiffness of Polymer/Clay Nanocomposites, Po-
lymer, Vol. 48, No. 3, 2007, pp. 901-909.
doi:10.1016/j.polymer.2006.11.062
[13] A. G. Facca, M. T. Kortschot and N. Yan, Predicting the
Elastic Modulus of Natural Fibre Reinforced Thermo-
plastics, Composites Part A: Applied Science and Manu-
facturing, Vol. 37, No. 10, 2006, pp. 1660-1671.
doi:10.1016/j.compositesa.2005.10.006
[14] L. Peponi, J. Biagiotti, J. M. Kenny and I. Mondragon,
Statistical Analysis of the Mechanical Properties of Na-
tural Fibers and Their Composite Materials. II. Composite
Materials, Polymer Composites, Vol. 29, No. 3, 2008, pp.
321-325. doi:10.1002/pc.20386
Copyright 2012 SciRes. AMPC