Professional Documents
Culture Documents
Composites: Part A
journal homepage: www.elsevier.com/locate/compositesa
a r t i c l e
i n f o
Article history:
Received 29 October 2007
Received in revised form 24 March 2008
Accepted 26 May 2008
Keywords:
A. Long ber thermoplastics
E. Extrusion compression molding
C. Process modeling
B. Thermoplastic composites
a b s t r a c t
Long ber thermoplastics (LFTs) have witnessed rapid growth in thermoplastics matrix composites,
mainly due to developments in the automotive and transportation sector. In LFTs, pelletized thermoplastic polymer matrix is reinforced with long glass or carbon bers (325 mm) are processed by extrusioncompression molding. The current work focuses on the applied science and manufacturing of E-glass/
polypropylene (E-glass/PP) LFT composite material. Process simulation was conducted to evaluate the
ow of ber lled viscous charge during the compression molding of the LFT composite. Studies on optimum charge size and placement in the tool, press force, temperature of mold, shrinkage and warpage
were also conducted. The ow pattern of the molten charge in the mold and the resulting ber orientation predicted by process simulation are veried experimentally. The studies have been applied for a
mass transit/transportation component namely, a LFT battery box access door for form-t-function to
replace a heavy metal door. Weight reduction of 60% was achieved using 40% weight percent E-glass/
PP LFT over the metal design.
2008 Elsevier Ltd. All rights reserved.
1. Introduction
Long ber thermoplastics (LFTs) are being used extensively in
automotive and transportation industry due to their superior specic strength and modulus resulting in substantial weight savings,
combined with relative ease of fabrication and handling [1].
Weight reduction in a vehicle increases overall fuel efciency,
thereby reducing the operating costs and signicantly contributing
to environmental and economic benets [2]. Global use of LFTs is
expected to grow from around 40 million lbs in 2001 to 75 million lbs in 2007 [3]. In general, some of the advantages of using
LFT over metals include high impact resistance, superior toughness, improved damping and corrosion resistance in conjunction
with ease of shaping and recyclability [4,5]. The use of a thermoplastic matrix provides the molder the ability to modify and enhance the properties of the resin by blending additives, llers
and re retardants depending on the nature of the application
[6]. Various components have been designed and manufactured
using LFTs for the transportation industry including, dashboard
carriers, front ends, seat shells, battery trays, spare wheel dwells,
etc. [2,7,8]. The typical applications of LFT components in an automobile are shown in Fig. 1 [8].
The mechanical properties of a part made of reinforced thermoplastics are dened by the matrix system, type of bers, ber con* Corresponding author. Tel.: +1 205 934 9199; fax: +1 205 934 8485.
E-mail address: uvaidya@uab.edu (U.K. Vaidya).
1359-835X/$ - see front matter 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.compositesa.2008.05.017
1513
P1 aT
1 aT P2 jc_ jP3
where
1514
oT
o2 T
k 2
ot
oz
oT
oT
Convection term: qcp vx
vy
ox
oy
ovx
ovy
Diffusion term: sxy
syz
oz
oz
w f /; x; y; t
2
3
4
where
q is the density;
cp is the specic heat capacity;
k is the thermal conduction coefcient;
sxz; syz, shear stress is xz and yz plane, respectively;
vx; vy; vz are the velocity component in x, y, and z, respectively.
The simulation of ber orientation during the compression
molding is essential to accurately predict the thermo-mechanical
behavior and the nal mechanical properties of the molded part.
In general, the orientation of a particle, such as ber is described
by two angles namely; in plane orientation and out of plane orientation angle. These angles change in time as the melt ows thorough a die. In general the angular orientations of the bers are
represented only by the in plane orientation angle. FolgerTucker
model [24] is used to capture the ow induced ber orientation.
The model adopts a statistical approximation that is applied to
the entire domain to predict the ber orientation. The state of particle orientation at a point is described by an orientation distribution function, and is dened such that probability of a particle
located at x, y at time t, being oriented between two angles is given
by Eqs. (5) and (6). Assuming the ber density is homogenous
throughout, the continuity equation can be written as shown by
Eqs. (7)(9). The ber distribution model accounts for the ber vol-
5
/2
w/; x; y; t d/
/1
ow
o
_
w/
ot
o/
C 1 c_ ow
ovx
ovx
2
/_
cos / sin /
sin /
w o/
ox
oy
ov
ov
y
y
cos2 /
sin / cos /
ox
oy
ow
o2 w ow
ovx
ovx
2
C 1 c_ 2
cos / sin /
sin /
ot
ox
oy
o / o/
ov
ov
y
y
cos2 /
sin / cos /
ox
oy
o
ovx
ovx
2
cos / sin /
sin /
w
o/
ox
oy
ovy
ovy
2
cos /
sin / cos /
ox
oy
where
1515
Fig. 3. Flow patterns for different charge placement and congurations: (a)(c) shows ow patterns of two small charges placed in different orientations, and (d) shows the
ow patterns of one long charge placed horizontally.
1516
Fig. 4. Flow front comparison of molten charge under compression molding: (a) short shot of charge compressed partially and (b) predicted ow front of charge.
shows a representative deformation after the part has been demolded and cooled to ambient temperature.
4. Design, and analysis of the LFT battery box door
As explained earlier, a mass transit part was designed and manufactured using 40 wt.% LFT E-glass/PP. A battery box access door
(referred to as battery door) is an external part of the 20 m
(60 ft) articulated mass transit bus (Fig. 9) which functions to protect and house the several batteries needed for the regular operation of the electrical systems of the bus. It is currently comprised of
an all steel sheet metal fascia which is bent to shape and then
welded to a tubular steel frame which provides additional stiffness
to the part. The metallic battery door is approximately
1 0.6 0.003 m and currently weights about 12 kg.
1517
Fig. 7. Fiber orientation for a representative sample obtained from Micro CT: (a) top surface showing a random orientation, (b) mid-plane through half the thickness showing
a preferential orientation, (c) ber distribution graph obtained from modeling showing random orientation on top surface, with a preferential orientation through the
thickness, and (d) representative ber frequency plot obtained from Micro CT images using FiberScan.
1518
Value
Units
7900
100
2
8000
175
20
1210
MPa
MPa
%
MPa
MPa
kJ/m2
kg/m3
selected based upon the assembly of the part to the exterior shell
of a bus. Static FEA simulations were conducted for three different
thicknesses (3, 4, and 5 mm) of the door shell using the properties
of the LFT (PP-GF40-03) material. Fig. 11 shows a typical von Mises
stress plot and the maximum deection of the door for a shell
thickness of 5 mm.
The von Mises stress plot illustrates the stress concentration in
the region of loading and the stress prole in the region of the stiffening ribs. The mid-span deection for the 3 mm shell thickness
LFT battery box door was excessive, and hence failed based on failure criteria (FC = 2.4, where a value of FC exceeding 1.0 denotes
failure). A similar scenario was observed with the 4 mm shell
thickness case (FC = 1.7). With the 5 mm thick version, the safety
factor is close to the strength of the steel counterpart, and the
deection (19.2 mm) is not signicant. A summary of the results
from the FEA and the solid model is provided in Table 2. The weight
between the steel frame and the LFT battery door design for 5 mm
thickness shell is compared in Table 3. The percentage weight savings on the nal LFT molded design was calculated to be approximately 60% compared to the steel frame design.
Fig. 10. (a) Single component design of the battery box access door with rib stiffened structure and (b) view of the door from the cosmetic side.
1519
Fig. 11. (a) Shows the Von Mises stresses (MPa) and (b) shows the displacement (mm) plots of 5 mm thick Battery Box Access Door subjected to 2223 N in the center.
Table 4
Processing parameters for extrusion compression molding of E-Glass/PP
Table 2
FEA for a load of 2223 N (500 lbf) for various shell thickness
Shell thickness
(mm)
Displacement
(mm)
Von Mises
stress (MPa)
Maximum stress
criterion (MPa)
Calculated
mass (kg)
3
4
5
44.70
28.19
19.20
224
156
99
2.40
1.70
1.00
2.10
2.80
3.50
Process parameter
Predicted value
Actual value
Units
25
80
90
3500
650
170
120
25
80
130
3900
650
170
180
mm/s
C
C
kN
mm
mm
S
Table 3
Comparison of weights between steel frame design and LFT design (5 mm thick shell)
Physical property
Steel door
LFT door
Face sheet
Tubular frame
5 mm shell design
Volume (m3)
Density (kg/m3)
Mass (kg)
Total mass (kg)
9.40E04
7.86E+03
7.39
11.96
5.80E04
7.86E+03
4.57
3.14E03
1.21E+03
3.51
3.51
and the experimental set-up used to compare the predicted displacement. Load vs. displacement data obtained from both the
FEA model and experimentally are compared in Fig. 14.
The stiffness of the door as predicted by the FEA is consistent
with the experimental stiffness until a deection of 15 mm (78%
of failure deection). At this point the stiffness response transitions
from a linear to a nonlinear state. The onset of nonlinearity can be
attributed to the local plasticity effects and/or damage initiation.
Fig. 12. Top and Bottom tool clamped on the press with an extruded charge.
1520
Fig. 13. (a) FEA model with applied boundary conditions and (b) experimental set-up to measure load vs. displacement on the molded LFT battery door.
Table 5
Results of ber weight fraction at various regions of molded battery door
Fig. 14. Load vs. displacement obtained from FEA and experiment.
Sample ID
Length
(mm)
Width
(mm)
Height
(mm)
Fiber weight
fraction
Rib 1
Rib 2
Rib 3
Rib 4
Rib 5
Skin center 1
Skin center 2
Skin corner 1
Skin corner 2
25.75
26.25
26.30
29.25
27.50
22.50
22.50
22.50
22.50
4.50
4.50
4.50
9.00
9.00
22.50
22.50
22.50
22.50
14.50
14.50
14.50
19.88
19.88
5.00
5.00
5.00
5.00
40.15
39.80
41.04
38.28
38.42
39.25
39.93
40.28
39.23
Fig. 15. (a) Failure initiation on the ribs during mechanical testing and (b) FEA showing stress concentration (von Mises stress) on the ribs at the same location.
Fig. 16. Fibers separated from resin and dispersed for ber length analysis.
Acknowledgements
The authors gratefully acknowledge the support provided by
the Federal Transit Administration (FTA), Department of Transportation, Project No. AL-26-7002 and Program Manager Terrell Williams of FTA. Technical help received from Juan Serrano and
George Husman is also gratefully acknowledged.
References
[1] Huptli A, Winski J. Direct processing of long bre reinforced thermoplastics:
selecting a feeding system. Plast Additives Compound 2003;5(5):369.
[2] Steffens M, Himmel N, Maier M. Design and analysis of discontinuous long
ber reinforced thermoplastic structures for car seat applications. In:
Proceedings of the International Conference on Computer Methods in
Composite Materials, CADCOMP; 1998. p. 3544.
1521
[3] Schut JH. Long-ber thermoplastics extend their reach. Plast Technol
2003;49(4):5661.
[4] Bartus SD, Vaidya UK. Performance of long ber reinforced thermoplastics
subjected to transverse intermediate velocity blunt object impact. Compos
Struct 2005;67(3):26377.
[5] Vaidya UK, Sriram R. Vibration damping of glass/polypropylene thermoplastic
composites. In: International SAMPE Symposium and Exhibition (Proceedings),
SAMPE 2004 Conference Proceedings, vol. 49; 2004. p. 75762.
[6] Schut J. Why long-glass molders are compounding in-line. Plast Technol
2002;48(4):529.
[7] Bartus SD, Vaidya UK, Ulven CA. Design and development of a long ber
thermoplastic bus seat. J Thermoplast Compos Mater 2006;19(2):13154.
[8] Krause W, Henning F, Troster S, Geiger O, Eyerer P. LFT-D A process
technology for large scale production of ber reinforced thermoplastic
components. J Thermoplast Compos Mater 2003;16(4):289302.
[9] Fu S-Y, Lauke B. Effects of ber length and ber orientation distributions on the
tensile strength of short-ber-reinforced polymers. Compos Sci Technol
1996;56(10):117990.
[10] Chawla KK. Interfaces. New York, NY: Springer-Verlag, New York, Inc.; 1998. p.
126.
[11] Thomason JL, Vlug MA. Inuence of bre length and concentration on the
properties of glass bre-reinforced polypropylene: 1. Tensile and exural
modulus. Composites Part A: Appl Sci Manufact 1996;27(6):47784.
[12] Thomason JL, Vlug MA, Schipper G, Krikor HGLT. Inuence of bre length and
concentration on the properties of glass bre-reinforced polypropylene: Part 3.
Strength and strain at failure. Composites Part A: Appl Sci Manufact
1996;27(11):107584.
[13] Bockstedt RJ, Skarlupka RJ. Overview of long ber reinforced thermoplastics.
In: Technical Papers, Regional Technical Conference Society of Plastics
Engineers, Design for the 21st Century; 1996. p. 6975.
[14] Hartness T, Husman G, Koenig J, Dyksterhouse J. The characterization of low
cost ber reinforced thermoplastic composites produced by the DRIFT
[trademark]
process.
Composites
Part
A:
Appl
Sci
Manufact
2001;32(8):115560.
[15] Kang S, Hieber CA, Wang KK. Optimum design of process conditions to
minimize residual stresses in injection-molded parts. J Thermal Stresses
1998;21(2):14156.
[16] Wang VW, Hieber CA, Wang KK. Dynamic simulation and graphics for the
injection molding of three-dimensional thin parts. J Polym Eng
1986;7(1):2145.
[17] Foss PH. Coupling of ow simulation and structural analysis for glass-lled
thermoplastics. Polym Compos 2004;25(4):34354.
[18] Foss PH, Chiang HH, Inzinna LP, Tucker III CL, Heitzmann KF. Experimental
verication of C-mold ber orientation and modulus predictions. In: Annual
Technical Conference ANTEC, Conference Proceedings, vol. 1, 1995. p. 6748.
[19] Foss PH, Harris JP, OGara JF, Inzinna LP, Liang EW, Dunbar CM, et al. Prediction
of ber orientation and mechanical properties using C-MOLD and ABAQUS, In:
Annual Technical Conference ANTEC, Conference Proceedings, vol. 1, 1996. p.
5015.
[20] Shay Jr RM, Foss PH, Mentzer CC. Comparison of C-MOLD predictions and
experimental shrinkages: ber-lled materials. In: Annual Technical
Conference ANTEC, Conference Proceedings, vol. 1, 1996. p. 496500.
[21] VerWeyst BE, Tucker III CL, Foss PH. Performing an accurate ber orientation
analysis. In: Annual Technical Conference ANTEC, Conference Proceedings,
vol. 1, 1996. p. 72630.
[22] N.N., EXPRESS 6.1, User and Theroy Manual, Aachen, Germany, 2000.
[23] Heber
M.
Modell
zur
rheologischen
auslegung
faserverstarkter
thermoplastischer prebauteile. Aachen: RWTH; 1995.
[24] Folgar F, Tucker III CL. Orientation behavior of bers in concentrated
suspensions. J Reinforced Plast Compos 1984;3(2):98119.
[25] Halpin JC, Kardos JL. HalpinTsai equations: a review. Polym Eng Sci
1976;16(5):34452.
[26] Shen H, Nutt S, Hull D. Direct observation and measurement of ber
architecture in short berpolymer composite foam through micro-CT
imaging. Compos Sci Technol 2004;64(1314):211320.
[27] Christensen SK, Hutchinson B, Sun EM, Osswald TA, Davis BA. Fibermatrix
separation in ribbed SMC and BMC parts. In: Annual Technical Conference
ANTEC, Conference Proceedings, vol. 1, 1997. p. 7827.
[28] [cited 2007 August 28]; Available from: http://tools.ticona.com/tools/mcbasei/
product-tools.php?sPolymer=&sProduct=.