Download as pdf or txt
Download as pdf or txt
You are on page 1of 6

Materials Science Forum Vol. 566 (2008) pp.

255-260
online at http://www.scientific.net
© (2008) Trans Tech Publications, Switzerland
Online available since 2007/11/20

New Impact Testing Methods for Sheet Metals Based on


SHPB Technique

Tadashi Hasebe1a and Yutaka Imaida2b


1
Faculty of Engineering, Kobe University
1-1 Rokkodai, Nada, Kobe 657-8501, Japan
2
Faculty of Engineering, Doshisha University
Kyotanabe, Kyoto 610-0321, Japan
a b
hasebe@mech.kobe-u.ac.jp, yimaida@mail.doshisha.ac.jp

Keywords: Split Hopkinson Pressure Bar Method, FEM, Impact Tension/Compression

Abstract. This paper proposes new impact testing methods applicable to sheet metals both under
tension and compression based on widely used split Hopkinson pressure bar (SHPB) technique.
Explicit dynamic finite element simulations by using LS-DYNA 3D are systematically conducted
for several specimen clamping conditions to seek the appropriate methodologies to realize the two
tests. For the tensile test, a method which can reduce stress oscillations that usually appear in the
measured stress-strain curves is proposed and is devised to be used with SHPB technique. For the
compression test, a candidate which can restrict buckling of the specimen is proposed. The method
uses a special die-set sandwiching the sheet metal specimen which is simultaneously compressed
without disturbing the planar stress wave to be propagated.

Introduction
Under the strong needs toward more accurate simulation on crashworthiness of motor vehicles and
estimation of sheet stamping processes in steel and automotive industries[1], tensile tests on sheet
metals in impact regimes have been extensively conducted[2], particularly on countermeasures to
reduce the stress overshoot phenomena, which tend to be pronounced in the one-bar method[3].
There should inevitably exists oscillations substantially caused by lateral vibration of the
input/output bars known as Poch-Hammer effect combined with dispersion of the elastic wave
during transmission. The authors[2] identified the major reason is the additional vibration caused at
the clamping parts, i.e., pin-and-hole assembly for the one-bar method and screw part for the split
Hopkinson pressure bar (SHPB) method[4]. For sheet metals, the pin-and-hole type would be most
suitable methodology for clamping the sheet specimen. Compression test is also important
especially in evaluating Bauschinger effect under load reversal, since sheet metals more or less
encounter such cyclic straining during stamping processes. No attempt, however, has been made on
this to date.
This paper aims at proposing SHPB-based new testing methods both for tension and
compression of sheet metals utilizing systematic FEM simulations. A minor modification for the
tensile test and a completely new idea to realize impact compression of sheet specimen are
systematically examined.

Analytical Model and Procedure


In the present study, LS-DYNA3D[5] is used for the dynamic explicit finite element simulations.
Many FEM-based studies on SHPB[6,7] have been reported so far, demonstrating the accuracy of
the FEM analysis. The accuracy of the present dynamic analysis in terms of elastic wave
propagation and its application to SHPB method has been carefully checked to be satisfactorily
good. For the contact analysis, e.g., for specimen-elastic bar and pin-hole interfaces, the penalty
method[5] is used with penalty coefficient f=0.1and contact spring coefficient K= 7 ×103 MPa.

All rights reserved. No part of contents of this paper may be reproduced or transmitted in any form or by any means without the
written permission of the publisher: Trans Tech Publications Ltd, Switzerland, www.ttp.net. (ID: 133.30.115.6-29/01/08,14:59:29)
256 Explosion, Shock Wave and Hypervelocity Phenomena

Shapes and dimensions of the specimen together with the incident/transmission bars (referred to
as “elastic bars” altogether, hereafter) for the SHPB method are depicted in Fig.1. For the pair of
elastic bars and pins for clamping the sheet specimen, isotropic elasticity is assumed as Young’s
modulus E=2×105MPa, Poisson’s ratio ν = 0.3 , and density ρ = 7.8×10-6kg/mm3, assuming High
Cr-Steel, whereas, for the material to be tested, commercially pure Al with isotropic hardening
elasto-plasticity is assumed, i.e., E=7 × 104MPa, ρ = 2.7 × 10-6kg/mm3 for elasticity and yield
stress σ Y = 80MP, hardening modulus EP=50MPa for plasticity. Regarding the constitutive equation
for the Al specimen, Cowper-Symonds model[6] is used to take into account the strain-dependent
stress response, i.e.,
 ε& 1 
σ y = 1 + ( ) P  (σ 0 + E pε effP ) , (1)
 C 
where the strain rate sensitivity parameters are C=6500 and P=4.
A trapezoidal impulse shown in Fig.2 is applied to the left edge of the incident bar as an initial
velocity profile to generate planar incident stress wave. Figure 3 schematizes two measurement
methods for stress-strain (S-S) response of the specimen in the present analysis, i.e., a direct method
detecting from the center of the specimen and that based on SHBP technique from incident,
reflected and transmitted stress waves detected at elements 500mm apart from the
specimen-clamping edges corresponding to the strain gauge positions. The former is regarded here
as the “true” response and is used as a reference.

Analytical Results and Discussions


Impact Tensile Test. Figure 4(a),(b) illustrates the “conventional” pin-type model of the clamping
part and the newly proposed “insert-type” model, respectively. For the clamp part of the elastic bar,
normally a slit is introduced as in Fig.4(a) into which a sheet specimen is inserted. Since this part
tends to oscillate just as a pair of flippers, a simple idea here is to connect the upper and lower parts
at the both edges as shown in Fig.4(b). Comparison of the vertical displacement of the elastic bar as
a function of time for both the models is shown in Fig.4(c), clearly showing a drastic reduction of
the oscillation amplitude in the new model. Also the effect of friction at the elastic bar-specimen
interface is investigated for the new model. Figure 5 compares the results with and without the
friction effect, demonstrating the discrepancy between the true S-S response and that obtained by
SHPB method is almost negligible when the friction effect is removed. This implies the importance
of lubrication in the pin-type clamping methods in practice.

Compressive Test. For impact compression tests, two conditions must be satisfied, i.e.,
(1)restriction of buckling and (b)preservation of planar stress wave propagating through the sheet
specimen. Static tests require the former condition alone, however, impact tests must additionally
satisfy the latter. A static compression test for sheet specimens has been successfully carried out
under static loading conditions [7] by introducing a pair of die-sets sandwiching the specimen
without causing additional stressing to the specimen during the tests. To satisfy the above two
conditions simultaneously, the die-sets sandwiching the sheet specimen should be compressed at
once. As a result of preliminary examinations to seek the appropriate conditions, the smaller cross
section of the die-sets than that of the elastic bars was found to be required for sufficient plastic
deformation of the sheet specimen to occur.
Materials Science Forum Vol. 566 257

Fig.1 Shape and dimensions of specimen and incident/transmission bars.

Fig.2 Trapezoidal velocity pulse applied to


the incident bar to generate planar stress
wave in it.

Fig.3 Schematics of stress-strain curve


measurements, i.e., direct method from
specimen center and that based on SHPB
method from stress waves.

Here we prepare roughly two types of the die-sets based on the pin type as depicted in Fig.6; i.e.,
types A and B. The type A has no screw part, while the type B has screw parts to fix the die-set and
specimen assembly with the elastic bars. For the type A, two cases are further considered, i.e., (1)no
bonding and (2)perfect bonding at the interface between the die-sets and specimen. In the former,
friction effect is ignored.
Figure 7 shows the results for the type A comparing the effect of the interfacial bonding, while
Fig.8 is the result for the type B. Figure 9 displays contour plots of stress distribution for the two
types , i.e., the type A without interfacial bonding and the type B. As can be seen in Fig.9(a), the
stress wave is evidently disturbed, i.e., planar wave is not maintained. This is simply because that
the pins are bent due to interfacial displacement between the assembly and the elastic bars.
Consequently, the evaluated S-S curve for this case is significantly deviated from the specimen
response as shown in Fig.7(a). The perfect interfacial bonding, on the other hand, can essentially
improve this situation as depicted in Fig.7(b). This is an ideal condition and technically difficult to
achieve. One simple as well as practically available alternative is the screw clamp, i.e., the type B.
As shown in Fig.8, the type B exhibits excellent evaluated S-S response agree well with the
real response, which is based on undisturbed planner stress wave propagating through the
die-set-specimen assembly as can be confirmed in Fig.9(b). The agreement between the two
responses is almost comparable with that in Fig.7(b), the ideal condition. Note the stress overshoot
and the following oscillation becomes slightly pronounced in this case due to the oscillation
inevitably caused at the screw part.
258 Explosion, Shock Wave and Hypervelocity Phenomena

Improved
Part

(a)Conventional Pin-type (b)New- type (c)


Fig.4 FEM models for (a)conventional type and (b)new ”insert” type, together with
(c)comparison of vertical oscillation of elastic bar between two types.

(a)With Friction (b)Without Friction

Fig.5 Comparisons of true stress-true strain curves for new type between specimen
response and that based on SHPB method: (a)with friction between
specimen-die slit interface and (b)without friction.

As is clarified above, the pin-type clamping is required to appropriately apply the compressive
load to the specimen. In addition to this, a fixture via screw of the die-set-specimen assembly to the
elastic bars is indispensable to keep the propagating stress wave planar, which can be a practical
alternative to the perfect interfacial bonding.
The next step is to separate the specimen response alone out of the overall S-S response for the
die-specimen assembly. Since the present case yields simple strain-constant model, the separation is
straightforward, meaning a simple law of mixture can be applied, i.e.,
σ = σ specimenV fspecimen + σ dieV fdie
 specimen (2)
V f + V fdie = 1

from which σ specimen can be obtained. It is found first of all that materials having similar mechanical
properties to the specimen are appropriate for the above yields sufficient accuracy, e.g., bulk mild
steels for IF steel and commercially pure aluminum for high purity aluminum.
Materials Science Forum Vol. 566 259

(a)Type A (b)Type B

Fig.6 FEM model for impact compression test using special die set;
(a)without screw (type A) and (b)with screw (type B) for clamping.

(a)Without Interfacial Bonding (b)With Interfacial Bonding

Fig.7 Comparison of true stress-true strain curves Fig8 True stress-true strain curve for
under impact compression for type A between type B with screw fixture for
two fixing conditions for die-set-specimen die-set-specimen assembly.
assembly.

(a)Type A

(b)Type B

Fig.9 Stress distributions for (a)type A with no interfacial bonding and (b)type B.

Fig.10 True stress-true strain curves evaluated based on proposed method, using
die-set material having (i)smaller and (ii)larger yield stress than that for
sheet metal tested, comparing with true response.
260 Explosion, Shock Wave and Hypervelocity Phenomena

Two cases are further examined in terms of the appropriate material properties for the die-sets,
i.e., those with yield and tensile stresses, σ Y , σ B , larger and smaller than the specimen material.
Figure 10 shows comparison of the evaluated S-S curves based on Eq.(2) with the true response
obtained separately using solid specimen for the material tested. The die-set material with smaller
σ Y , σ B exhibits better agreement, whereas the larger σ Y , σ B result in underestimation of the flow
stress level. The good agreement between the evaluated and true responses for the former case
demonstrates the effectiveness of the proposed method.
For a closed concave die, however, air-cushion effect often causes incomplete forming,
Conclusion
This paper made an attempt to propose new impact testing methods for sheet specimen under
tension and compression based on conventionally as well as widely used SHPB technique. For the
tensile test, a closed slit type clamping part was proposed to significantly reduce the stress
oscillation. For impact compression tests, an appropriate combination of pin and screw clamps for
specimen-die-set assembly was shown to be suitable to restrict the buckling of the specimen and
maintain the propagating planar stress wave undisturbed.

References
[1] A. Uenishi, H.Yoshida, Y. Kuriyama and M. Takahashi: Material characterization at high strain
rates for an organization of car body structure, Nippon Steel Technical Repert, Vol.378 (2003),
p.21.
[2]New materials Center affiliated with Osaka Science & Technology Center: Kinzoku zairyo no
kousoku henkei tokusei hyouka houhou no kenkyu kaihatsu jigyo gyomu seika houkokusho,
(2000) (in Japanese).
[3]K. Kawata, S. Hashimoto, K. Kurokawa and N. Kanayama: A new testing method for the
characterization of materials in high velocity tension, Mechanical Properties at High Rates of
Strain 1979, Inst. of Physics, Bristol and London, (1979), p.71.
[4]H. Kolsky: An Investigation of the Mechanical properties of materials at very high rates of
loading, Proc. Physical Society, Vol.B62, No.11, (1949), p.676.
[5]J. O. Hallquist: LS-DYNA Theoretical Manual”, Livermore Software Technology Corporation,
(2006).
[6]L. D. Berrhold, and C. H. Karnes: Two-dimensional analysis of the split Hopkinson pressure bar
system, J. Mech. Phys. Solids, Vol23, No.1-19 (1975).
[7]J. Rodriguez, C. Navarro and V. Sanchez-Galvez: Numerical Assessment of the Dynamic
Tension Test Using the Split Hopkinson Bar, J. Testing and Evaluation,Vol.22, No.4 (1994),
p.335.
[8]G. R. Cowper and P. S. Symonds: Strain hardening and strain rate effects in the impact loading of
cantilever beams, Brown Univ. App. Math. Report, (1958) p.28.
[9]T. Kuwabara,Y. Morita, Y. Miyashita and S. Takahashi: Elastic-plastic behavior of sheet metal
subject to in-plane reverse loading, J. Japan Soc. Technol. Plasticity, Vol.36, No.414 (1995),
p.768.

You might also like