Design and strength analysis of glass fiber-reinforced epoxy composite shelter.
Wang, Tie ; Xu, Jingyu
1. Introduction
In today's world, automobile lightweight has been the best
effective way to solve energy saving and emission reduction problems.
Recently, an extensive interest within vehicles industries to develop
and make use of lightweight composite materials and structures has been
generated by the need of reducing energy consumption. Clear benefits of
using fiber reinforced plastic (FRP) composites have been reported
within aeronautical, rail, naval and automotive industries [1]. Besides,
glass fiber-reinforced epoxy composite (GFRP) application on the shelter
can reduce the quality and enhance the strength. The lighter vehicle
body means less fuel consumption, more power and higher transport
efficiency. The higher shelter strength means more different battle
environments and stronger operational ability. Therefore, it is
especially important to study GFRP shelter currently.
The study object in this paper is the shelter of a special vehicle,
whose carriage panel skins were replaced by GFRP laminates incorporating
E-glass fibers within an epoxy matrix. But many of the available
publications are the study of containers and the application of glass
fiber-reinforced epoxy composites. Kevin Giriunas et al. [2]
investigated the ISO shipping container's structural strength using
finite element computer modeling and the computer simulations
demonstrated the effectiveness of the container walls and roof to resist
the loads, which were beneficial for the shelter modelling and analysis.
Genelin and Salim [3], Borvik et al. [4], and Borvik et al. [5]
performed blast load structural tests on actual ISO containers. The
available information is relevant and important to
[FIGURE 1 OMITTED]
2. The shelter computer model information
The material properties of the plates of the shelter were analyzed
with classical lamination theory [10] and structurally define and
evaluate performance for the shelter. M.A. Badie et al. [6] examined the
effect of fiber orientation angles and stacking sequence on the
torsional stiffness, natural frequency, buckling strength, fatigue life
and failure modes of composite tubes, which incorporates hybrid
carbon/glass fiber with an epoxy matrix, and the results obtained in
this article could provide the reference for GFRP shelter study. Craig
W. Hudson et al. [7] made use of carbon fiber-reinforced composite for
rail vehicle floor panels, it was concluded that the use of lightweight
material could indeed get a lot of weight loss.
Table 1 shows the overall dimension of the shelter studied in this
paper, which conforms to ISO [8] and national standards [9]. Structural
components and framework (details are shown in Table 2) of the shelter
are seen from Fig. 1. In the article, the structural strength analysis
of various aluminium shelters using FEM was performed firstly to select
the optimal structure model for GFRP shelter design. Then, the impacts
of different lay-up design and different fiber thickness of GFRP
laminate to the strength and stiffness of the shelter were carried out
to find the better lay-up design method based on the shelter
lightweight. The loading scenario was choose as the helicopter lifting
condition according to ISO [8] and national shelter standards [9], which
could determine the ability to withstand over loading ability of the
shelter.
designed by laminate lay-up design points [11]. Sandwich panels
utilize flexural strength of a system composed of outer stiff skins
spaced by a softer core of low density. Spacing between skins is
increased to improve flexural resistance, thermal insulation, and
minimize relative slip from shear transfer. Softer foams can be better
insulators and will generally result in better continuous strain
transfer, minimizing de-bonding failure [12].
The shelters were modelled and analyzed using the programs CATIA
[13] and ANSYS Workbench [14]. CATIA [13] is a three-dimensional (3D)
Computer Aided Design (CAD) program used to model 3D objects. ANSYS
Workbench [14] is the finite element analysis (FEA) program used to
apply a finite element meshing and analyze the meshed shelter models
imported from CATIA [13]. Because of the integral bearing characteristic
of the stiffener plate, the skeleton and the polyurethane foams were
divided by hexahedral elements, and the inner and outer skins of the
carriage panels were meshed with quadrilateral elements. Q235 is used
for the skeleton, corner fittings and skids, and its main properties are
as follow: density is 7850 kg/[m.sup.3], elastic modulus is 2100 MPa,
poisson's ratio is 0.3 and the yield stress is 235 MPa. However,
the skeleton is divided into three kinds of square tubes and the
concrete parameters are shown in Table 2. As to the rigid polyurethane
foams, its density is 60 kg/[m.sup.3], elastic modulus is 10 MPa, and
poisson's ratio is 0.3. 2A12 is selected for the aluminum plate,
and its properties are as follow: density is 2700 kg/[m.sup.3], elastic
modulus is 70000 MPa, Poisson's ratio is 0.3. The material
properties of selected E-glass/ epoxy composite used in GFRP shelter
carriage panel are displayed in Table 3, which are provided by Shenyang
Tongchuang FRP company.
3. The shelter carriage panel material simulation
In this part, four simplified shelter models were established,
which were named M1, M2, M3 and M4. Similar assumptions were made. The
rear and right side of the shelter containing the doors, locking
assembly, and hinges were replaced by an identical wall used for the
other side wall section with similar properties. It was assumed that the
rear door and the right door and window assembly could withstand the
same loads as the other walls. All of the connections were modelled to
represent fully welded connections which could not fail. Therefore, the
sandwich panels of M2 include the rigid polyurethane foams compared with
M1. The corner posts were added on M3 based on M2. And then the corner
fittings were set upon M4 based on M3. Simplified models of the shelter
were used to verify model assumptions, and show which components of the
shelter could be simplified without sacrificing accuracy. The optimal
structure model would be for GFRP shelter design. The GFRP shelter with
the initial design [45[degrees], 90[degrees], 0[degrees], 45[degrees],
-45[degrees]] was named as M5(5). Number 5 in the bracket presented that
the laminate of GFRP shelter was designed for five-layers.
Through the linear analysis in ANSYS [14], M4 was the most strength
and stiffness, which would be used for the composite simulations. Based
on M4, the material of the carriage panel skin was replaced by E-glass
fiber epoxy laminate. [45[degrees], 90[degrees], 0[degrees],
45[degrees], -45[degrees]] for the laminate was finished by the ACP
module in ANSYS. From Fig. 2 and Fig. 3, the maximum deformation of
M5(5) was smaller compared with M1, M2 and M3, 0.6 mm larger than M4.
Meanwhile, the maximum equivalent stress of M5(5) was the smallest and
reduced about 40% than the maximum value. Fig. 4 was the deformation
cloud pictures of four simplified shelter models. Therefore, GFRP can be
applied on the shelter carriage panel skin to get the light-weight
purpose, increase the shelter structural strength and have better
carrying capacity.
4. GFRP shelter lay-up design and simulation
The GFRP shelter analysis includes mainly two parts. One is the
effects of different lay-up schemes on GFRP shelter strength and
stiffness. Giving twelve kinds of lay-up schemes, the laminates were
designed with 5 layers in order to find five better lay-up schemes.
Another is the impacts of fiber thickness on GFRP shelter strength and
stiffness. It can be realized by controlling lay-up scheme the same and
changing fiber thickness. Based on five better lay-up schemes, the
laminate was designed with ten-layer and twenty-layer.
4.1. The effects of different lay-up schemes
Figs. 5 and 6 correspond to the maximum deformation and equivalent
stress of twelve shelter models with different lay-up schemes. The model
designed with the lay-up scheme of [-45[degrees], 45[degrees],
-45[degrees], 45[degrees], -45[degrees]] had the best deformation 1.9697
mm. [90[degrees], 45[degrees], 90[degrees], -45[degrees], 90[degrees]]
had the worst deformation 2.1791 mm. Maximum stress was between 148 to
173 MPa. According the comparison , we can conclude that [45[degrees],
90[degrees], 0[degrees], 45[degrees], -45[degrees]], [45[degrees],
90[degrees], 0[degrees], 90[degrees], 45[degrees]], [-45[degrees],
45[degrees], 0[degrees], 45[degrees], -45[degrees]], [-45[degrees],
45[degrees], -45[degrees], 45[degrees], -45[degrees]] and [0[degrees],
45[degrees], 90[degrees], -45[degrees], 0[degrees]] are five better
lay-up schemes.
[FIGURE 2 OMITTED]
[FIGURE 4 OMITTED]
[FIGURE 5 OMITTED]
4.2. The impacts of different fiber thickness
Based on [45[degrees], 90[degrees], 0[degrees], 45[degrees],
-45[degrees]], [45[degrees], 90[degrees], 0[degrees], 90[degrees],
-45[degrees]], [-45[degrees], 45[degrees], 0[degrees], 45[degrees],
-45[degrees]], [-45[degrees], 45[degrees], -45[degrees], 45[degrees],
45[degrees]] and [0[degrees], 45[degrees], 90[degrees], -45[degrees],
0[degrees]], M5(5), M6(5), M7(5), M8(5) and M9(5) were established
respectively. Then, ten-layer and twenty-layer design for the laminate
were completed and the established models were M5(10), M6(10), M7(10),
M8(10), M9(10) and M5(20), M6(20), M7(20), M8(20), M9(20). Just take
M5(10) and M5(20) as examples, M5 represents the model with the lay-up
scheme of [45[degrees], 90[degrees], 0[degrees], 45[degrees],
-45[degrees]]. Besides, number 10 in the bracket presents ten-layers
design, that is [45[degrees], 90[degrees], 0[degrees], 45[degrees],
-45[degrees]] 2. 20 indicates [45[degrees], 90[degrees], 0[degrees],
45[degrees], -45[degrees]] 4.
[FIGURE 3 OMITTED]
[FIGURE 6 OMITTED]
Through the simulation, ten-layer design can improve the GFRP
shelter strength and stiffness partly. The equivalent stress of M5(10),
M7(10) and M9(10) declined slightly compared with M5(5), M7(5) and
M9(5). M6(10)'s and M8(10)'s unchanged. Moreover, the
deformation of M5(10), M7(10) and M8(10) decreased a little than M5(5),
M7(5) and M8(5). But the deformation of M6(10) and M9(10) increased some
than M6(5) and M9(5).
In addition, twenty-layer design can preferably improve the
mechanical properties of GFRP shelter. As compared to five-layers design
models, the reduced maxi mum deformation of M8(20) was the least and
M9(20) decreased the most. The reduced deformation of M5(20) and M7(20)
were respectively 0.0027 and 0.0047 mm. But M6(20)'s deformation
increased 0.0012mm. And the reduced range of the maximum equivalent
stress is between 0.01 to 0.03 MPa. M6(20) had no changes, M7(20) and
M8(20) both fell 0.01 MPa, M5(20) fell 0.02 MPa and M9(20) decreased up
to 0.03 MPa. While compared with ten-layers design models, the maximum
equivalent stress of M8(20) decreased 0.01 MPa and the other
twenty-layer models had no changes. The deformation of M5(20) and M8(20)
reduced and the others had some increase instead. Figs. 7 and 8 were the
maximum deformation and equivalent stress of different fiber thickness
shelter models with five better lay-up schemes.
[FIGURE 7 OMITTED]
[FIGURE 8 OMITTED]
5. Conclusions
Following the trend of composite application in the automotive
industry, this paper has made use of E-glass fiber epoxy composite in
the shelter. Under the helicopter lifting condition, the simulation and
comparisons were performed according ISO and industry standards, and the
following conclusions have been drawn.
1. Through the analysis of lightweight materials, E-glass fiber
composite can be used as the shelter carriage panel skin instead of
aluminum in order to achieve the lightweight design. Comparing with the
aluminum shelter, GFRP shelter has higher strength and wider carrying
capacity.
2. By analyzing the effects of different lay-up schemes on the
shelter strength and stiffness, we see that [+ or -] 45[degrees] and
0[degrees] are more suitable for the laminate design, and [+ or -]
45[degrees] make the best mechanical performance. [+ or -] 45[degrees]
should be put outside of the laminate, and then 0[degrees] and
90[degrees] could be put inside. With the thickness of glass fiber
decrease, more [+ or -] 45[degrees] mean higher strength of the
laminate. [+ or -] 45[degrees] and 0[degrees] can effectively improve
the shelter structural strength.
3. According to the research of the impacts about different fiber
thickness on the shelter strength and stiffness,
[45[degrees],90[degrees], 0[degrees], 45[degrees], -45[degrees]] and
[-45[degrees], 45[degrees], -45[degrees], 45[degrees], 45[degrees]] are
more appropriate for twenty-layer design. [-45[degrees], 45[degrees],
0[degrees], 45[degrees], -45[degrees]] and [0[degrees], 45[degrees],
90[degrees], -45[degrees], 0[degrees]] are fit for ten-layer design.
[45[degrees], 90[degrees], 0[degrees], 90[degrees], -45[degrees]] should
be designed with five-layer.
4. After using lightweight materials, the quality of the shelter
has decreased from 1119.9 kg to 956.5 kg, which has reduced about
14.59%.
http://dx.doi.org/10.5755/j01.mech.2L2.8624
Acknowledgements
This work is supported by Natural science fund project of Liaoning
Province under Grant #20102189, fund project of Shenyang Tongchuang FRP
company under Grant #2012040503, Science and technology plan projects of
Liaoning Province under Grant #2011220055.
References
[1.] Velea, M.N.; Wennhage, P.; Zenkert, D. 2014. Multi-objective
optimization of vehicle bodies made of FRP sandwich structure, Composite
Structure 111: 7584. http://dx.doi.org/10.1016/j.compstruct.2013.12.030.
[2.] Giriunas, K.; Sezen, H.; Dupaix, R.B. 2012. Evaluation,
modeling, and analysis of shipping container building structures,
Engineering Structures 43: 48-57.
http://dx.doi.org/10.1016/j.engstruct.2012.05.001.
[3.] Genelin, C.L.; Salim, H.A. 2010. Evaluation of modular
re-locatable buildings subjected to blast. Tyndall (FL): Air Force
Research Laboratory.
[4.] Borvik, T.; Hanssen, A.G.; Dey, S.; Langberg, H.; Langseth, M.
2008. On the ballistic and blast load response of a 20 ft ISO container
protected with aluminum panels filled with a local mass-Phase I: Design
of protective system, Engineering Structure 30: 16051620.
http://dx.doi.org/10.1016/ j.engstruct.2007.10.010.
[5.] Borvik, T.; Hanssen, A.G.; Dey, S.; Langberg, H.; Langseth, M.
2008. On the ballistic and blast load response of a 20 ft ISO container
protected with aluminum panels filled with a local mass-Phase II:
Validation of protective system, Engineering Structure 30: 621-1631.
http://dx.doi.org/10.1016/j.engstruct.2007.10.011.
[6.] Badie, M.A.; Mahdi, E.; Hamouda, A.M.S. 2011. An investigation
into hybrid carbon/glass fiber reinforced epoxy composite automotive
drive shaft, Materials & Design 32:1485-1500.
http://dx.doi.org/10.1016/j.matdes.2010.08.042.
[7.] Hudson, C.W.; Carruthers, J.J.; Robinson, A.M. 2010. Multiple
objective optimisation of composite sandwich structures for rail vehicle
floor panels, Composite Structure 92: 2077-2082.
http://dx.doi.org/10.1016/j.compstruct.2009.10.018.
[8.] Souza, V.A.; Kirkayak, L.; Watanabe, I.; Suzuki, K.; Ando, H.;
Sueoka, H., et al. 2013. Experimental and numerical analysis of
container multiple stacks dynamics using a scaled model, Ocean
Engineering 74: 218-232.
http://dx.doi.org/10.1016Zj.oceaneng.2013.05.013.
[9.] GJB2093-94. 1995. General test methods for military shelter.
National Science and Technology Commission. 25p.
[10.] Talib, A.R.A.; Ali, A.; Badie, M.A.; Lah, N.A.C.; Golestaneh,
A.F. 2010. Developing a hybrid, carbon/glass fiber-reinforced, epoxy
composite automotive drive shaft, Materials & Design 31: 514-521.
http://dx.doi.org/10.1016/j.matdes.2009.06.015.
[11.] Omkar, S.N.; Khandelwal, R.; Yathindra, S.; Naik, G.N.;
Gopalakrishnan, S. 2008. Artificial immune system for multi-objective
design optimization of composite structures, Engineering Applications of
Artificial Intelligence 21: 1416-1429.
http://dx.doi.org/10.1016/j.engappai.2008 .01.002.
[12.] Dong, H.; Wang, J.; Karilhaloo, B.L. 2014. An improved
Puck's failure theory for fibre-reinforced composite laminates
including the in situ strength effect, Composites Science and Technology
98: 86-92. http://dx.doi.org/10.1016/j.compscitech.2014.04.009.
[13.] Zhang, L.L. 2008. CATIA V5 application course. Beijing:
Tsinghua University Press. 61-205p.
[14.] Li, B.; He, Z.J.; Chen, X.F. 2008. ANSYS Workbench design,
simulation and optimization. Beijing: Tsinghua university Press. 33-86p.
Tie Wang, Jingyu Xu
School of Automobile and Transportation, Shenyang Ligong
University, Shenyang110159, China,
[email protected]
Received November 07, 2014
Accepted March 12, 2015
Table 1
Typical specification for a standard CAF35 shelter
ID codes Length L, mm Width W, mm Height H, mm
CAF35 350 2100 1900
Table 2
The material properties of square tube
Material Specification, Elastic Density, Location
type mm modulus, kg/
H B t MPa [m.sup.3]
100 80 4 70000 1204 Longitudinal
beams of the
floor plate
Q235 80 40 2 65100 1020 Other beams of the
floor plate and
inner beams of
the top plate
60 40 2 67400 1120 Inner beams of the
side plates
Table 3
The material properties of E-glass/epoxy composite used in the
laminates of GFRP shelter carriage panels
Parameters [rho], kg/ [E.sub.x], [E.sub.Y]
Value [m.sup.3] MPa MPa
1800 34000 6530
Parameters [[gamma]. [G.sub.YZ], [G.sub.XY],
Value sub.YZ] MPa MPa
0.366 1698 2433
