ansys1.pdfzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz

LOW VELOCITY IMPACT ANALYSIS OF
LAMINATED FRP COMPOSITES
B. KRANTHI KUMAR 1
M.Tech student, Department of Mechanical Engineering,
Godavari Institute of Engineering and Technology,
Rajahmundry, Andhrapradesh, India.
E-mail:Kranthikumar301@gmail.com
LAKSHMANA KISHORE.T 2
Associate professor, Department of Mechanical Engineering,
Godavari Institute of Engineering and Technology,
Rajahmundry, Andhrapradesh, India.
tlkishore.giet@gmail.com
Abstract
Fiber reinforced composites have become increasingly important over the past few years and are now the
first choice for fabricating structures where low weight in combination with high strength and stiffness are required.
Fiber Reinforced Plastics (FRP) composites are in greatest commercial use. They have been extensively used in
aerospace, automotive, marine and construction industries due to their inherent advantages over conventional
metals. Failure modes of such laminated structures are also different than those of conventional metallic materials.
Impact is one such great design limitation criteria involved in designing new composite products.
The present work is aimed at gaining an initial understanding of the impact behavior of fiberglass
reinforced laminates with vinylester and is polyester resins. The purpose of this research is to characterize the
damage done to fiberglass laminates subjected to low velocity, high mass impact. The effect of adding a protective
layer of rubber to the laminates is also investigated.
Finite element models are created with ANSYS/LS-DYNA onlinear finite element software. These models
are used to simulate the drop tower tests and extended to thicker laminates as well as different impact speeds and
impactor mass.
These models are able to predict approximate stresses and strains induced in the laminates during the
impact which are compared to the damage from the drop tower tests.
Keywords: FRP, composites, vinylester, polyester resins, stress and strain.
1. INTRODUCTION
A Composite material is a material brought about by combining materials differing in composition or form on a
macro scale for the purpose of obtaining specific characteristics and properties. The constituents retain their identity such
that they can be physically identified and they exhibit an interface between one another. In addition, three other criteria
are normally satisfied before we call a material as composite. Firstly, both the constituents have to be present in
reasonable proportions. Secondly, the constituent phases should have different properties, such that the composite
properties are noticeably different from the properties of the constituents. Lastly, a synthetic composite is usually
produced by deliberately mixing and combining the constituents by various means. The constituent that is continuous
and is often, but not always, present in the greater quantity in the composite is termed as the Matrix. The aim is to
improve the properties of the matrix by incorporating another constituent. Generally used Matrix materials in composites
are Polymers, Ceramics and Metallic Matrix. Polymers have low strength and Young’s moduli, ceramics are strong, stiff
and brittle, and metals have intermediate strength and Young’s moduli, together with good ductility. The second
constituent is known as the reinforcement, as it enhances the mechanical properties of the matrix. The diameter of the
fibers varies from 0.1- 100 µm. Strength and stiffness properties of reinforced elements are generally 10- 1000 times
B. Kranthi Kumar et al. / International Journal of Engineering Science and Technology (IJEST)
ISSN : 0975-5462 Vol. 4 No.01 January 2012 115

higher than those of the matrix. At least one of the reinforcement is small, say <500µm and mechanical properties of
composites are function of the shape, dimensions, orientation and quantity of the reinforcement.
2. Finite Element Analysis
Some of the first research on the modeling of impact damage resistance in Composite materials was on the
subject of bird strikes. In 1981, R.E McCarty [1] used the nonlinear finite element software MAGNA to model birds
striking the canopy of the F-16A and in 1982; M.S. Hirschbein [2] used NASTRAN to model the effects of bird
strikes on turbine blades. Both of these authors claimed reasonable results.
There have been many damage criterions created for composite damage analysis in Livermore’s LS-DYNA
transient non-linear FEA software [3]. One of the earliest was established by Chang and Chang [4] in 1985. They
created a damage model to predict damage in notched laminate composites under tensile loading. The model could
be used to assess the type and the extent of the damage, evaluate residual strengths, and predict ultimate strengths.
The damage was assessed with a proposed failure criteria and proposed property degradation model. They found
excellent agreement between their predicted results and experimental data. The current composite damage model
that is used by LS-DYNA is based on this model.
They identified the three main in plane failure modes as matrix cracking, fiber matrix shearing, and fiber
breakage. The following failure criterion (equation 1) was used to predict matrix cracking:
2
0
0
2
m
XY
XY
XY
XY
t
Y
e
d
d
Y
u
XY
XY


















 (1)
Where σy is the transverse tensile stress
σxy is the shear stress in each layer.
Yt is the transverse tensile strength,
γu
xy is the ultimate shear strain and
em is the failure strain of the matrix material.
By adding the shear stress-shear strain relationship:
3
1
XY
XY
XY
XY
G


 








 (2)
where Sc is the ply shear strength.
For laminates with linear elastic behavior (a = 0),
As before, for linear elastic materials the equation can be reduced to:
2
2
2
f
c
xy
t
x
e
S
X

















 
 (3)
This failure criterion states that when the stresses σx and σxy, in any one of the plies, satisfy one of the previous
equations (ef = 1), either fiber breakage or fiber-matrix shearing occurs in that layer.
Once damage has occurred the material looses strength in that area and the reduction is dependent on the
mode of failure. For matrix cracking in a layer, the transverse modulus Ey and Poisson’s ratio γy are reduced to
zero. The longitudinal modulus Ex and the shear-stress strain relation of that layer are unchanged
When fiber breakage and/or fiber-matrix shearing are predicted, the extent of the property degradation
depends on the size of the damage. For fiber failure, both Ey and Ex are reduced to zero. The longitudinal modulus
Ex and the shear modulus Gxy degenerate according to the Weibull distribution:



















0
1
exp
A
A
E
E
x
d
x (4)
2.1 Elements Used
Due to advantages and disadvantages of one element over the other, as well as the parameters to be
determined in view, two elements are used in the total analysis. They are Shell 163 element and Solid 164 element.

2.1.1 SHELL163 Element
SHELL163 is a 4-node element with both bending and membrane capabilities [5]. Both in-plane and
normal loads are permitted. The element has 12 degrees of freedom at each node: translations, accelerations, and
velocities in the nodal x, y, and z directions and rotations about the nodal x, y, and z-axes. This element is used in
explicit dynamic analyses only.
Fig. 1 Shell 163 Element.
2.1.2 SOLID164 Element
SOLID164 is used for the 3-D modeling of solid structures. The element is defined by eight nodes and nine
degrees of freedom at each node: translations, velocities, and accelerations in the nodal x, y, and z directions. This
element is used in explicit dynamic analyses only. This element cannot be used for models in which inter laminar
stresses are to be predicted. This solid element uses an effective modulus through the thickness of the layer by
averaging the elastic modulus of all the layers. This becomes a problem when the layers have drastically different
moduli from layer to layer as in fiberglass laminates. This element will not take into effect the position in the
laminate of the very high or very low modulus layers, which plays a key role in the bending of the laminate. It will
also give inaccurate inter laminar stresses since the adjoining moduli are the key parameter for that calculation.
Fig.2 Solid 164 Element.

3 .Methodology :Statement of Problem
The problem is to characterize the damage occurred to fiberglass laminates subjected to low velocity, high
mass impact. This subject is a crucial design question that appears frequently in the design of new composite
products. Impact loads affect almost all commercial composite products whether by design or by accident. This
investigation attempted to provide initial insight into impact damage done to the fiberglass laminates by performing
several drop weight impact analyses with finite element models and predicted the behavior of the laminates under
different impact situations. Further research is needed to evaluate the effects of impact damage on specific
applications.
3.1 Non- Linear Analysis
The non- linearity arising from the nature of material is called ‘Material Non- linearity’. All non- linearities
are solved by applying the loads slowly (dividing it into a number of small load increments). The model is assumed
to behave linearly for each load increment, and the change in model shape is calculated at each increment. Stresses
are updated from increment to increment, until the full applied load is reached.
4. Finite element analysis
Finite element analysis is used to gain information about the behavior of the composite laminates subjected
to impact loading. Simple models are analyzed and more complex models are used to simulate additional parameters
such as thick laminates and rubber protecting the composite. FEA is used to predict the stresses and strains induced
in the laminate by a low velocity projectile. These stresses and strains can later be used to predict the life span of the
composite under various loading conditions. ANSYS and ANSYS/LS-DYNA finite element codes are used for the
simulations.
4.1 Static Model
Initially a static model is created to get an idea of the loads caused by a 1.25 m. diameter granite rock.
Several simple shell models are made with pressures in various locations. The resulting stresses and strains are very
low and well within the capability of the composite even without a rubber liner. From this it is decided that the
impact is going to be the limiting factor in the design.
4.2. Results and Discussions
The block is placed at several distances from the shell and dropped onto it. The maximum deflection in the
center of the shell converged to –0.3048 μm as the initial distance is decreased, table This final deflection compared
to the static deflection represents an impact factor of 2.2, which is close to the 2.0 predicted by mechanics of
materials.
Table 4.1 Convergence of deflection and stress
Distance from
target (mm)
0.254 0.0254 0.00254 0.000254 0.0000254
Deflection (μm) -0.11684 -0.4826 -0.03175 -0.03048 -0.03048
Stress Intensity
(N/m2)
21.7 9.1 5.95 5.775 5.705
4.3 Drop Tower Model
The next step is to model the drop tower test in order to predict stress and strain information. Two separate
ANSYS/ LS-DYNA models are created in order to compare the results of the two and pick one to use to model the
tests. One target is created with shell 163 elements and one with solid 164 elements. The shell element has the
advantage of fewer elements and degrees of freedom, which lowers the required computer time. The solid elements
are advantageous due to their capability to capture the through thickness effects of the deflection and stress since
more than one element is placed in the thickness direction.

4.3.1 Shell 163 Model
4.3.1.1 Problem Description
A 12.5 cm x 12.5 cm shell target whose thickness is 0.6 cm and with the laminate properties is used for the
target and a sphere with 1.8 cm diameter is used to represent the round end of the impactor tup, figure.3. The shell is
constrained in the Z direction around the edges to simulate the clamp device and also constrained in all directions
along two sides to prevent rigid body motion.
Fig.3 Meshed model of Square shell target Fig.4 Cylindrical target with quadrilateral mesh with
boundary conditions specified.
4.3.1.2 Problems with Target Plate Shape
The above model used very similar boundary conditions as the test jig in the drop weight tower. However,
these boundary conditions gave some strange results in the clamped region due to the bending and membrane
actions of the shells. In order to remove the possibility of these effects and also to lower the number of degrees of
freedom, a new model is created. This model consisted of a 7.5 cm diameter flat cylindrical shell with the z
displacements and out-of-plane rotations fixed around the outer edge.
4.3.1.3 Problems with Target Plate Mesh
A uniform mesh is often difficult with a cylindrical plate. If a triangular mesh is used, singularities are
created at the center of the circle where the elements internal angles are very small. ANSYS also does not
recommend the triangular mesh option with shell 163 elements therefore the quadrilateral mesh is used for all
models. With the quadrilateral mesh the center elements have a slightly different shape than the elements on the
outer rings of the circle; however, no irregularities are noticed with this mesh, figure.4.shown above
4.3.1.4 Results and Discussions
In order to verify the accuracy of the model, the impact force and the total energy was compared with the
experimental results. Both were within 15% of the experimental results. Figure.5 shows the impact force predicted
along with the force history from the experimental results with an impact velocity of 2 m/s.

Fig.5 Force history and prdicted FEA Fig .6 Energy- History Graphs for Impactor
Velocity of 2 m/s shell 163 element.
In dynamic finite element analysis, the hourglass energy is a large concern. A representative energy plot
from ANSYS/ LS-DYNA is shown in Figure 6.
4.3.2 Solid 164 Model
4.3.2.1 Problem Description
A 7.5 cm diameter solid cylinder is used for the composite target and a 1.75 cm diameter sphere is used for
the steel impactor, figure.7. The solid elements are constrained with all translations fixed around the edges to
simulate the clamped boundary conditions in the drop tower tests. The spherical solid is given an initial velocity (2
m/sec) in the direction of the target as in the other models. The material properties used for this model are shown in
the tables listed below table 4.2
Table 4.2 Material Properties used for Composite Laminates.
Distance from target
(mm)
0.254 0.0254 0.00254 0.000254 0.0000254
Deflection (μm) -0.11684 -0.4826 -0.03175 -0.03048 -0.03048
Stress Intensity
(N/m2)
21.7 9.1 5.95 5.775 5.705
Fig.7 Finite Element Model of the Impactor Fig.8 Energy- History Graphs for Impactor
Velocity of 2 m/s using Solid 164 element

4.4.1 Drop Tower Tests Simulations :Test 1 – Vinylester Resin
4.4.1.1 Description
For this drop tower test, the impact mass is 5.5 kg and the impactor velocity just before impact is 3.81 m/s.
In the FEA model, the density of the sphere is adjusted to achieve the same mass as the impactor. The target is
modeled with solid 164 elements with a thickness of 6 mm. A linear orthotropic material model is used for the
composite laminate and linear isotropic properties for the steel impactor. These properties are listed in table 4.3
Table 4.3 Material properties used for composite laminate.
Distance from target
(mm)
0.254 0.0254 0.00254 0.000254 0.0000254
Deflection (μm) -0.11684 -0.4826 -0.03175 -0.03048 -0.03048
Stress Intensity (N/m2) 21.7 9.1 5.95 5.775 5.705
These material properties are determined with the aid of the shareware program The Laminator [6]. The
individual material properties for each layer are given in Table 4.4,4.5,4.6. These layer properties are inputted into
the program along with the layer thickness and stacking sequence to determine the apparent material properties.
Table 4.4 Material properties for each layer.
E1 E2 G12 V12
Roving (E-glass) 1.00 ×107
2.00 ×106
2.00 ×106
0.2
Mat material (isotropic) 1.00 ×106
1.00 ×106
4.00 ×105
0.28
Matrix (vinylester) 5.00 ×105
5.00 ×105
1.00 ×105
0.3
Table 4.5 Lay-up of Isopolyester laminate.
Layer Thickness (mm)
317.5 mm Nexus 0
311.15 mm – 2 oz. OC 0.419
Resin 0.482
90 – 113 yield roving 0.787
Resin 0.482
311.15 mm – 2 oz. OC 0.419
Resin 0.482
Resin 0.482
311.15 mm – 2 oz. OC 0.419
342.9 mm Nexus 0
Total 4.378
The impact force is first compared to the experimental data. From a curve fit of the experimental
acceleration data, the maximum impact force is found to be approximately 7116.8 N. Five samples are tested with
the same impact velocity and the maximum impact force is almost identical in all samples. Figure.9 shows a
comparison between the experimental data and the predicted impact force verses time. The higher predicted impact
force can be attributed to the model not including the energy release due to the damage (delamination and matrix
cracking).

Fig.9 Force history and FEA force for an impactor velocity of 3.81 m/s.
4.4.2 Test 2 – Vinylester Resin
4.4.2.1 Description
The second test consisted of the same material with an impactor velocity of 2 m/s. Since the impact
velocity is much lower, the impact force is also reduced. Figure.10 shows the predicted and actual force data for this
test.
Table 4.6 Lay-up of vinylester laminate.
Layer Thickness(mm)
342.9 mm Nexus 0
311.15 mm – 2 oz. OC 0.381
Resin 0.482
Resin 0.482
311.15 mm – 2 oz. OC 0.381
Resin 0.482
Resin 0.482
311.15 mm – 2 oz. OC 0.381
Resin 0.482
Resin 0.482
311.15 mm – 2 oz. OC 0.381
342.9 mm Nexus 0
Total 6.396

Fig.10 Force history and FEA force for impactor velocity of 2 m/s.
The predicted time span that the impactor is in contact with the target did not change with different
impactor speed in the ANSYS models. The contact times used in the tests are 5.3 ms. for the first test (3.81. m/s) and
4.4 ms. for the second test (2 m/s). The FEA model predicted a contact time of 3.7 ms. for all impactor speeds.
4.4.2.3 Comparisons
Three tests are performed with the vinylester resin and impact speeds of 2, 3, and 4 m/s respectively. These
three analyses are compared in order to examine the effect of velocity on the specimens; all other parameters are
kept the same. Table 4.7 is a summary of these three tests.
Table 4.7 Comparison between test with impact speeds of 2, 3, and 4 m/
Impact
Velocity
(m/s)
Predicted maximum in-
plane stress on
top (N/m2
)
Predicted maximum in-
plane stress on bottom
(N/m2
)
Backside crack
length (m)
1 300.3 x106
228.1 x106
0.0152
2 477.4 x106
292.6 x106
0.0533
3 651.7 x106
312.8 x106
0.0610
Based on comparisons of the predicted stress patterns on the bottom of the FEA models and the crack
length, the cracks extend to the (175 – 245) x106
N/m2
range on all three of these models. Further work could
determine a stress range that initiates and continue matrix cracking and de-bonding.
4.4.2.4 Stress and Strain
The maximum predicted stress in the z direction for these three impactor speeds is shown in Figure.11

Fig.11 Maximum in-plane stress for three different impactor speeds
(viz. 2, 3, and 4 m/s).
As before, the predicted stress jumped quickly from the 2 to the 3 m/s case then leveled off as the speed
went to 4 m/s. The stress contour patterns were almost identical in each case with only the values of the stress
regions changing. The strain in the X direction (maximum in-plane) also showed nearly identical contour patterns
for all three cases. These strains are shown in Figure.12.
Fig .12 Maximum in-plane strains for three different impactor speeds
(viz. 2, 3, and 4 m/s)
The “wavy” pattern in the high strain area is very interesting. This pattern is also seen in the external
backside crack on some of the test samples. This is most likely due to the anisotropic nature of the laminate.

5. Conculsion
The purpose of this research is to examine the impact response of fiberglass reinforced laminates. Several
drop weight tower impact analyses are performed on 12.5cm x 12.5 cm FRP plates with resins including vinylester
and iso polyester. These test simulated high mass (5.5 kg) and low velocity (2-4 m/s) impact. ANSYS/ LS-DYNA
non-linear finite element analysis program is used to predict the stresses in the laminates during the impact. These
stresses are then compared to the damage from the drop tests. The critical stress range for impact damage is in the
range of (175-245) x106
N/m2
, which is very close to the static ultimate strength of the Duraspan laminates.
References
[1] R.E. McCarthy, “Finite Element Analysis of a bird-Resistant Monolithic Stretched Acrylic Canopy Design
[2] for the F-16A Aircraft.” AIAA Aircraft Systems and technology Mtg. Dayton, Ohio, Aug. 11-13 1981.
[3] Hirschbein, M.S., “Bird Impact Analysis Package for Turbine Fan Blades.” 23rd
Structures,
[4] Structural Dynamics, and Materials Conference, New Orleans, LA, May 10-12 1992.
[5] LS-DYNA, Livermore Software Technology Corporation, 7374 Las Positas Rd., Livermore, CA 94550.
[6] Chang, F.K., and K. Y. Chang, “A Progressive Damage Model for Laminated Composites Containing
[7] Stress Concentrations.” J. Composite Mater. 21 (1997): pp. 834-855.
[8] Ansys Theory Manual, “Elements Reference”.
[9] The Laminator – Classical Analysis of Composite Laminates, http://www.thelaminator.net
[10] Ansys Theory Manual, “Elements Reference”.
[11] Cantwell, W.J., P. T. Curtis, and J. Morton, “Low Velocity Impact Damage Tolerance in CFRP
[12] Laminates Containing Woven and Non-Woven Layer.” Composites. 14 (1983): pp. 301-305.
[13] Lui, D. “Impact Induced Delamination – A View of Bending Stiffness Mismatching.”
[14] Journal of Composite Materials. . 22
[15] Hou, J. P., N. Petrinic, and C. Ruiz, “A Delamination Criterion for Laminated Composites Under
[16] Low- Velocity Impact.” Composite Science and Technology. 61 (2001): pp. 2069-2074.
[17] Cantwell, W.J., and J. Morton, “The Impact Resistance of Composite Materials – a Review”
[18] Composites. 22 (1991): pp. 347-362.

ansys1.pdfzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz

More Related Content

Similar to ansys1.pdfzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz (20)

More from aljabbery (18)

Recently uploaded (20)

ansys1.pdfzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz