ISSN 0853-8697

BENCHMARK PROBLEM FOR CRUSH ANALYSIS OF

PLASTIC PARTS FOR AUTOMOTIVE

Faisal R. M.
Department of Chemical Engineering, Faculty of Industrial Technology
Islamic University of Indonesia,
Jl. Kaliurang Km. 14 Yogyakarta 55501
Phone. (0274) 895287, Facs. (0274) 895007 ext. 148
E-mail: [email protected]

ABSTRAK
Dengan mengurangi biaya dan waktu pengembangan desain proses, permintaan
atas CAE meningkat. Untuk mengembangkan teknik simulasi komputer pada analisis
pengrusakan, diadakan kelompok kerja untuk menyeleksi masalah-masalah benchmark dan
dilakukan percobaan-percobaan simulasi di antara anggota-anggota. Beberapa uji tegangan
dan kempaan yang dilakukan untuk menilai pengaruh dari kadar ketegangan dan suhu
pada sifat-sifat yang berhubungan dengan mekanik dari bahan plastik. Kemudian, uji-uji
pengrusakan dari bahan percobaan yang terbuat dari polipropilen yang dkuatkan oleh karet
dilakukan untuk menyediakan benchmark data. Melalui percobaan dari simulasi komputer
dan membandingkan dengan benchmark data., beberapa parameter kunci dari simulasi
pengrusakan di temukan.

Kata-kunci :Polipropilen; simulasi komputer; uji pengrusakan; benchmark data.

1. INTRODUCTION
Recently, to cut development time, reduce prototyping cost and improve
quality, CAE is being commonly used at the design stage. However, the
accumulation of know-how for analysis, e.g. procedure to measure material
properties of thermoplastic materials and treatment in numerical simulation, are
not sufficient to evaluate impact performance of plastic parts by CAE.
Mechanical properties of solid thermoplastic materials consisting of millions
of monomers differ from that of metals consisting of arranged atoms of metal.
Furthermore, since semi-crystalline polymers such as Polypropylene consist of a
crystalline phase and an amorphous phase, these mechanical properties have high
non- linearity.
Although, some papers explain their detailed mechanism [1,2], further
research on generating material data and numerical treatment in numerical
simulation [3, 5] is necessary to use CAE software properly.
This paper describes the detail of a benchmark test selected to develop these
analysis know-how and a brief overview of case studies by crush analysis code.

TEKNOIN, Vol. 10, No. 2, Juni 2005, 85-94 85

2. MECHANICAL PROPERTIES OF THERMOPLASTIC MATERIALS
Generally, mechanical properties of thermoplastic materials are strongly
dependent on strain rate, temperature and direction of load. Furthermore they are
dependent on molding or process conditions to make plastic parts and humidity
for some plastics.
Some factors affecting mechanical properties are shown in following
paragraph.

2.1 Effect of Molding Condition

Injection molding process is often used to make plastic parts of ABS or
Polypropylene. Since thermoplastic material is melted and injected in a mold
under high pressure in this injection molding process, it is difficult to maintain the
temperature or pressure history of material uniformly over the part. Mechanical
properties often vary even in the same part. Figure 1 shows the distribution of
Young's modulus measured from results of compression tests of a specimen, cut
from dumb bell test piece (JIS K6810) made by ABS resin (UBE cycon, VW10), by
universal tester.
This result shows that mechanical properties are distributed even in such a
simple shape specimen, and this implies that some consideration of this
distribution is necessary when generating mechanical properties for CAE analysis,
e.g. averaging by typical mesh size.

Figure 1. Result of compression test

Figure 2. Appearance of test piece after tensile test

86 Faisal – Benchmark Problem for Crush Analysis of Plastic Parts for Automotive
2.2 Effect of Strain Rate
Figure 2 shows a photograph of a test piece made of transparent ABS resin
(Ube cycon, CTB3505) after tensile test by universal tester. Whitening by void or
crazing, which is a peculiar phenomenon of thermoplastics, is observed in this
photograph. The outward appearance of whitening varies strongly with a slight
change of strain rate, i.e. stripe-like whitening under lower strain rate and uniform
or mist-like whitening under higher strain rate. This can be related to the sharing
mechanism of load in the specimen. So, it is easily understood that the effects of
strain rate on mechanical properties of thermoplastics are significant.

2.3 Effect of Ambient Temperature

Figure 3 shows the result of tensile test varying ambient temperature. This
plot shows that the yield stress is significantly dependent not only on strain rate
but also on ambient temperature within the range of experience in service
conditions and this dependency cannot be ignored.
Furthermore, temperature rises during test due to visco-elastic phenomenon
and low conductivity of thermoplastics added further complexity. Figure 4 shows
the temperature rise inside the specimen during compression test.

Figure 3. Yield strength-strain rate curve (ABS)

Figure 4. Temperature rise during test (ABS)

TEKNOIN, Vol. 10, No. 2, Juni 2005, 85-94 87

3. BENCHMARK PROBLEM FOR CRUSH ANALYSIS
3.1 Stress of Strain Curve of Material
Polypropylene (rubber toughened PP by grand polymer) which is widely
used for automotive parts is selected for benchmark problem. Stress}strain
relationships necessary for crush analysis are measured by injection molded JIS
K71113 type 2 dumb bell specimen and compression/tension tester specially
designed for thermoplastics and made by the author's laboratory.
Figure 5 shows the result of true stress}true strain relationship obtained from
extension from gauge length of 25mm and load from load-washer during tensile
test assuming volume conservation conditions. The lines of strain rate of 0 (static)
and 133 (1/s) are extrapolated from experimental data by Malbern type
constitutive function.

Figure 5. True stress-true strain curves

3.2 Test Piece for Benchmark Problem

Test piece for benchmark problem is cut from a plate with ribs injection
molded by polypropylene, and its width, height and depth are 30, 32.3, 30 mm,
respectively. This has an open box shape and its sidewalls are tapered; the
thickness at the top end is 1.7mm and that at bottom end is 1.2 mm. Figure 6
illustrates the geometry of the test piece.

Figure 6. Test piece for benchmark test

88 Faisal – Benchmark Problem for Crush Analysis of Plastic Parts for Automotive
3.3 Test Apparatus for Benchmark Test
The test drop weight type impact test apparatus designed to maintain
repeatability shown in Figure 7 is used for benchmark test.

Figure 7. Schematic illustration of crush test apparatus

4. RESULTS OF EXPERIMENT
4.1 Result of Crush Tests
As an example of a benchmark test, the load} displacement relationship
when a weight of 26.26 kg is dropped from 10 cm height is shown in Figure 8. A
photograph of the test piece after a crush test is shown in Figure 9.

Figure 8. Load-displacement curve

TEKNOIN, Vol. 10, No. 2, Juni 2005, 85-94 89

 Figure 9. Photograph of test piece after test

5. RESULT
Several case studies on this benchmark problem by using PAM-CRASH',
RADIOSS' and LS-DYNA' are carried out. Through these case studies, effect of
each numerical parameters on accuracy of analysis result are investigated and
methods to improve accuracy are developed by comparing these analysis results
and experiments. The following paragraph will describe some examples of case
studies by PAM-CRASH [6].

5.1 Case Ctudy 1: Effect of Mesh Size

First of all, the effect of mesh size which determines computing cost of
analysis is investigated. Figure 10 shows fine mesh used in these case studies.
Coarse mesh is generated by doubling the mesh size from this fine mesh. The
analysis result of this case study and effect of mesh size are shown in Figure 11.
Strain rate dependency was considered by specifying all curves shown in Figure 5.
Although both analysis results underestimate the first peak of load by about 20% at
displacement of 3mm, the displacement of the first peak and displacement and
load of the 2nd peak at displacement of 9mm show good agreement with
experimental results. Overall behavior is re-produced by analysis, except for the
displacement of over 12 mm.

90 Faisal – Benchmark Problem for Crush Analysis of Plastic Parts for Automotive
 Figure 10. Fine mesh for finite element analysis

.
Figure 11. Effect of finite element mesh size

5.2 Case Study 2: Effect of Strain Rate Dependency

Consideration of strain dependency of mechanical properties is also effective
in computing cost as mesh size. Comparison of analysis results with and without
considering strain rate dependency of stress}strain relationship is shown in Figure
12. Consideration of strain rate on stress}strain relationship is done by using all
curves shown in Figure 5. Consideration is also done by using single curve (select
strain rate of 0.002 or 3.42 as typical strain rate) from Figure 5. Coarse mesh size
shown in the previous paragraph is selected for this case study.

TEKNOIN, Vol. 10, No. 2, Juni 2005, 85-94 91

 Figure 12. Effect of strain rate dependency

Whereas the result of typical strain rate"0.002 (1/s) under-predicts the

experiment, the result of typical strain rate"3.42 (1/s) and result considering strain
rate dependency shows good agreement with the experiment. Although there is no
procedure to determine the typical strain rate, this case study shows that by
estimating appropriate strain rate, computing cost can be reduced.

5.3 Case Study 3: Effect of Imperfection

The specimen used in this benchmark problem has imperfections; within its
four open ends, two ends are 0.2mm higher than others. In previous case studies,
these imperfections are considered at finite element modeling. In this case study,
analysis results without considering this imperfection are investigated. This may
occur with use of CAE at design stage. Figure 13 shows the result of the case study.
There are significant over estimates of load over a wide range of displacements in
the analysis results without considering initial imperfections. This shows that it is
important to consider such a small initial imperfection, especially in the contact
region in crush analysis. Other case studies about mesh geometry, modeling of
contact or friction, etc. are carried out in the same manner as the above paragraph.

92 Faisal – Benchmark Problem for Crush Analysis of Plastic Parts for Automotive
 Figure 13. Effect of initial imperfection

6. CONCLUSION
Factors affecting mechanical properties for crush analysis of thermoplastics
are investigated. Through several case studies by ‘PAM-CRASH' upon benchmark
problem by a box-shape specimen cut from plate with ribs made from
Polypropylene, the following are confirmed: Molding conditions to make the test
specimen, strain rate and ambient temperature during test, significantly affect
mechanical properties within the range of service conditions. So, it is important to
consider these dependencies in crush analysis to design plastic parts. Effects of
mesh size are not significant within the range of size of mesh selected in this
report. Further case studies are necessary to investigate the effect of mesh size.
There are possibilities to reduce computational time by selecting appropriate
typical strain rates. Further case studies are necessary to establish the procedure to
estimate typical strain rate. Consideration of small initial imperfection at the
contact area significantly affects the analysis result. Detailed representation of this
geometry at finite element modeling is necessary in crush analysis. Further case
studies to investigate other numerical solution parameters, e.g. time step at explicit
analysis, shape of element, modeling of contact and friction are necessary to
establish understanding of crush analysis for thermoplastic parts. Further case
studies, investigation and enhancement of benchmark tests are being carried out
among members of the working group. Through these case studies, benchmark
data, which can contribute to develop techniques or know how for crush analysis
of thermoplastic part will be prepared. Techniques or know how developed
through these case studies will be distributed and shared.

REFERENCES
[1] Narisawa, I. (1992) Selection and Design Stiffness of Plastics, Kogyo-Chyosakai (in
Japanese).
[2] Narisawa, I. (1993) Fracture Toughness of Plastics, Sigma-Syuppan (in Japanese).

TEKNOIN, Vol. 10, No. 2, Juni 2005, 85-94 93

[3] Hirao, and Fujii et al. (1994) Effect of Strain-Rate and Temperature on The
mechanical Properties of ABS Resin, Journal of JSPP, Vol. 6, No. 4 (in Japanese).
[4] Hirao, and Fujii et al. (1995) Stress Analysis Considering the Strain-Rate
Dependency on Mechanical Properties of Thermoplastics, Journal of JSPP, Vol.
7, No. 6 (in Japanese).
[5] Fujino, Kawano, and Fujii. (1997) An Experimental Consideration on Fracture
Toughness of ABS Resin, Journal of JSME, No. 96-0688 (in Japanese).
[6] PAM-CRASH User's Manual Version 1998.

94 Faisal – Benchmark Problem for Crush Analysis of Plastic Parts for Automotive