Bauschinger effect correspondence of experimental tests

X. Lemoine1,2, A. Aouafi3
1
ArcelorMittal Maizières, Research and Development Automotive Products – Voie Romaine, BP30320, 57283
Maizières-lès-Metz Cedex France
URL: www.arcelormittal.com e-mail: xavier.lemoine@arcelormittal.com
2
LPMM, UMR CNRS 7554, ENSAM 4 rue Augustin Fresnel, 57078 Metz Cedex 3, France
URL: www.metz.ensam.fr; www.lpmm.fr
3
LPMTM-CNRS, UPR9001, Université Paris 13, 99 av JB Clément, 93430 Villetaneuse, France
URL: www-lpmtm.univ-paris13.fr e-mail: aouafi@lpmtm.univ-paris13.fr

ABSTRACT: Many studies have shown that improvements in the numerical prediction about rolled sheet
forming is done through the laws of behaviour increasingly complex, particularly by the combination of
isotropic and kinematics hardening (mixed hardening) to take on the Bauschinger effect. But there are several
types of test experimental characterization Bauschinger effect: simple shear in one direction and then reverse,
tension and compression, bending unbending test… In the case of steel, this paper will answer these two
questions: Are these experimental tests are equivalent in terms of Bauschinger effect? What is the impact of
these different tests on the identification of parameters of the law of behaviour?

Key words: Bauschinger Effect, Steel, experimental tests, Isotropic and kinematics hardening

metal, compression test resulting after a first plastic

1 INTRODUCTION deformation of tensile test, a decrease in the yield
stress. Since that day, this is known as Bauschinger
Increasingly in the forming and in-use properties effect. The figure below shows what we get in terms
prediction, we use the complex behaviour laws by of stress-strain in Bauschinger shear test.
introducing a kinematic hardening combined with a
hardening isotropic. In rheology, there are several
experimental tests to show the Bauschinger effect
allowing to determinate the parameters of this type
of laws. But what do these various tests define the
same characteristic of the Bauschinger effect? In this
paper, we will try to answer this question. As a first
step, it will be determined what the Bauschinger
effect and different experimental tests on
characterizing. In a second step, it will be different
experimental results on a wide range of steel with a
comparison of characterization on different types of Fig. 1. Stress-strain curve for Bauschinger tension-compression
tests. Finally, we discuss the implications for the test [2].
identification.
The Bauschinger effect (see fig. 1) is characterized
by two stages. The first stage consists to the
2 BACKGROUND ON THE BAUSCHINGER transient Bauschinger deformation composed by
EFFECT early re-yielding and workhardening stagnation. The
workhardening stagnation appears at a certain range
2.1 Description of prestrain. The second stage is the permanent
softening defined by stress offset in a region after
In 1881, J.Bauschinger [1] discovers realizing on a transient period.
2.2 Quantifications where Ei are different area under the curve, defined
in the figure 3.
To quantify the Bauschinger effect, we have got
several possibilities. The first is used by A. Aouafi at σ+

Stress
the last conference ESAFORM [2] (see fig. 2). He E3
quantified the bounds of stagnation region ( γ 1 and
|σ-| E2
γ 2 are respectively the beginning and the end of σ0 E1
workhardening stagnation). We can add the Epredef
quantification of the permanent softening.
Cumulative
Offset
strain
Fig.3. Schematic definition of energy parameters for
Bauschinger test.

2.3 Experimental tests

Hardening rate ș [MPa]

Inflexion Points
Before applying these quantification parameters, we
Transitory regime will list the experimental tests to achieve the
θ2
Bauschinger effect. For a Bauschinger effect in an
θ1
Cumulative shear strain
experimental test, it is enough to solicit the material
in one direction and then in the reverse direction
with the possibility of registering a variable linked to
stress in function on other variable linked to the
Fig.2. Schematic definition of yl and y2 parameters for strain. The experimental tests the more used are
Bauschinger shear test [3]. compression-tension [6], tension-compression [5, 7,
In the literature, there are others parameters to 8], simple Bauschinger shear test [9, 10, 11, 12] or
quantify the Bauschinger effect [3, 4, 5]. Here are bending test (3-points [13] or 4-points [14] bending
some examples about Stress parameters: test, cyclic bending tests [15, 16]).
For compression-tension, tension-compression or
σ+ −σ− Bauschinger shear test, we get a direct stress-strain
X 1σ = (1) curve. In contrast to the bending test, we have got
2 either force-displacement (3-points or 4-points
σ− bending) or moment-angle (cyclic bending).
X 2σ = (2) Accordingly for the bending tests, it is not possible
σ+ define the parameters of Bauschinger effect
σ+ −σ− quantifying and it is necessary to make an
X 1σ 1
X 3σ = +
= +
= (1 − X 2σ ) (3) identification by inverse method with FEA Code.
2σ σ 2 The difficulty of tension-compression or
where σ + denotes the flow stress in forward compression-tension tests is the buckling; the
advantage is to have got directly stress-strain curve
straining and σ − the early re-yielding defined at on reference axis of hardening about FEA code
offset given (see fig.3). behaviour implementation.
Here are some examples about Energy parameters: About shear test the difficulty is the Yield loci
parameters influence the hardening parameters
E1 identification. The advantage we can attempt large
β E1σ = (4)
E predef strain.

E2
β E 2σ = (5) 3 BAUSCHINGER EFFECT
E predef
CORRESPONDENCE OF EXPERIMENTAL
E3 TESTS ABOUT STEELS
β E 3σ = (6)
E predef For different steel grades, Bauschinger shear tests
were carried out in various laboratories of university 450

as LPMTM [3, 9] or M&S [10]. In the framewoek of 400

ArcelorMittal project, two specimens of steel grade 350

were testing by the compression-tension tests of 300

X1σ (MPa)
T.Kuwabara [6]. 250

200
450
150
400
100
350
50
300
0
X1σ (MPa)

250 0 5 10 15 20 25 30
Prestrain Equivalent VM (%)
200
DP Steel BS DP Steel CT Mild Steel BS Mild Steel CT
150

100 Fig.6. For different prestain of 5% (equivalent of Von Mises),

50
comparison of Bauschinger shear test and Compression-
Tension on same sample of 2 steels grades about Bauschinger
0
0 200 400 600 800 1000 1200 1400 1600 stress parameter X 1σ .
UTS (MPa)
TC (S.Sadagopan & al.) BS ArcelorMittal CT ArcelorMittal

Fig.4. For prestain of 5% (equivalent of Von Mises), curve of 60%

Bauschinger stress parameter X 1σ versus UTS, about 50%

Bauschinger shear test, Compression-Tension and Tension-

40%
Compression test [5].
X3σσ (MPa)

30%
60%

20%
50%

10%
40%

0%
X3σσ

30% 0 5 10 15 20 25 30
Prestrain Equivalent VM (%)
20% DP Steel BS DP Steel CT Mild Steel BS Mild Steel CT

10%
Fig.7. For different prestain of 5% (equivalent of Von Mises),
comparison of Bauschinger shear test and Compression-
0% Tension on same sample of 2 steels grades about Bauschinger
0 200 400 600 800 1000 1200 1400 1600
UTS (MPa)
stress parameter X 3σ .
TC (S.Sadagopan & al.) BS ArcelorMittal CT ArcelorMittal

Fig.5. For prestain of 5% (equivalent of Von Mises), curve of

Bauschinger stress parameter X 3σ versus UTS, about 4 MODELLING
Bauschinger shear test, , Compression-Tension and Tension-
Compression test [5]. To illustrate the impact of identification, we realize
In the figures 4 to 7, we compared 2 parameters of 3 different identifications on the DP steel with non-
the Bauschinger effect quantifying identified in the linear kinematic hardening using Amstrong-
preceding paragraph applied to these Bauschinger Frederick law [17].
shear tests (BS) and compression-tension tests (CT) X = C X ( X sat n − X )λ (5)
with tension-compression tests (TC) directed by
S.Sadagopan & al [5]. The first parameter (see The first identification uses only the Bauschinger
figures 4 and 6) shows a good correlation between shear tests, the second only the compression-tension
the different types of tests, however the second tests and the last third on 2 types of tests. In figure 8,
parameter not (see figures 5 and 7). The Figures 6 we find classical result, ie that the first two cases of
and 7 show also the comparison on the same sample identification showing a good correlations on the
of 2 steels grades, 2 parameters for Bauschinger type of test that was used for identification and bad
shear tests and compression-tension tests. Only the for others type of test and correlation average for the
first parameter shows a concordance of 2 types of third case. This is due in part because of yield loci
tests on the Bauschinger effect. (here Von Mises) and partly due to dispersion
measurement experimental tests. Figure 9 confirms
that the Lemaitre & Chaboche hardening law not ESAFORM2007, Saragossa, (2008)
enough to properly describe the “elbow” and 4. J. Yan, ‘Study of Bauschinger Effect in Various Spring
Steels’ Degree of Master of Applied Science, 1999,
Bauschinger effect of the curves, and the need Department of Metallurgy and Materials Science,
models containing doubles kinematics variables. University of Toronto,
300 5. S. Sadagopan, D. Urban, ‘Formability Characterization
Of A New Generation Of High Strength Steels’ Final
250
report of AISI/DOE Technology Roadmap Program,
200 March 2003
6. T. Kuwabara, Y. Morita, Y. Miyashita, S. Takahashi,
X1σ (MPa)

150
“Elastic-Plastic Behavior of Sheet Metal Subjected to In-
100 Plane Reverse Loading”, in Tanimura, S. and Khan, A.S.
(eds), Dynamic Plasticity and Structural Behaviors,
50
(1995) 841-844
0
7. H.Gese, ‘Report on determination of yield locus, fracture
6 12 17 5 9 11 17 diagram and Bauschiner’effect parameters of steel
BS BS BS CT CT CT CT
DP780’ in the test report 148-1G for MATFEM
Prestrain Equivalent VM (%)
Partnerschaft Dr. Gese & Oberhofer on behalf of Arcelor
DP Steel identification 1 identification 2 identification 3
Research S.A., TsPO TEST, Voronezh (2007)
Fig.8. Comparison of 3 possibilities of identification 8. W. Schmitt, O. Benevolenski, T. Walde, A. Krasowsky,
1500
‚Material Characterization for simulation of sheet Metal
Forming’, in VIII International Conference on
Stress
(MPa)

B 1250

1000
A Computationa Plasticity COMPLAS VIII, E. Oñate and
D.R.J. Owen (Eds) © CIMNE, Barcelona, 2005
750
9. S. Bouvier, V. Tabacaru, M. Banu, C. Maier, C. Girjob,
500
B. Gardey, H.Haddadi, C. Teodosiu, ‘Selection and
Strain
250
Identification of Elastoplastic Models for the Materials
0 used in the Benchmarks’, Digital Die Design Systems
-0.2 -0.1 0 0.1 0.2 0.3 0.4

BS
-250
-500
-750
-1000
Exp. BS10 Exp. BS20 Exp. BS30
Solver BS10 Solver BS20 Solver BS30
Fig.9. For identification 1, comparison stress-strain curve
between experimental and model for Bauschinger shear test
and Compression-Tension test
5 CONCLUSIONS
This paper has listed some tests and parameters to
quantify the Bauschinger effect on steels. A
comparison of 2 tests (Bauschinger shear and
compression-tension) has been made in terms of
parameters to quantify the Bauschinger effect and
identification of the kinematic hardening model.
REFERENCES
1. J. Bauschinger, ‘Ueber die Veranderung der
Elasticitatsgrenze und elastcitatsmodul verschiedener’,
Metal Civiling N.F., 27, (1881) 289-348
2. F. Yoshida, T. Uemori, ‘A model of large-strain cyclic
plasticity and its application to springback simulation’,
Int. Journal of Material Science, 45 (2003) 1687–1702
3. A. Aoufi, S. Bouvier, M. Gaspérini, X. Lemoine and O.
Bouaziz ‘Phenomenological Analysis of Kinematic
Hardening of HSLA and IF Steels Using Reverse Simple
Shear Tests’, Proceedings of the Int. Conf
(3DS) Contract IMS 1999 000051, 18 Months Progress
Report, November 2001.
10. P. Flores, ‘Development of Experimental Equipment and
Identification Procedures for Sheet Metal Constitutive
Exp. CT05 Exp. CT10 Laws’, PhD Thesis of University of Liege, Novembre
Solver CT05 Solver CT10
2005
11. R. Greze, H. Laurent, P. Y. Manach ‘Springback study
in aluminum alloys based on the Demeri Benchmark
Test: influence of material model’, Proceedings of the
Int. Conf ESAFORM2007, Saragossa, (2008)
12. E.F. Rauch, J.J. Gracio, F. Barlat, A.B. Lopes, J.F.
Fereira Duarte, ‘Hardening Behavior and Structural
Evolutions upon Strain Reversal of Aluminum Alloys’,
Scripta Mater., 46, (2002) 881-886
13. F. Campana, L.Cortese, F. Placidi, ‘FEM Evaluation of
Springback after Sheet Metal Forming: Application to
High Strength Steels of a Combined Isotropic-Kinematic
Hardening Model.
14. C. Lange, ‘Etude Physique et Modélisation Numérique
du Procédé de Sertissage de Pièces de Carrosserie’ PhD
Thesis of Ecole Nationale Supérieure des Mines de Paris
(2006)
15. M. Brunet, F. Morestin, S. Godereaux, ‘Nonlinear
kinematic hardening identification for anisotropc sheet
metals with bending-unbending tests’, Journal of
Engineering Material and Technology, ASME, 2001,
Vol.123, pp.378-383
16. G. Arnold, ‘Identification de Lois de Comportement
pour la Simulation de Procédé Mise en Forme à Froid’,
PhD thesis of Ecole Nationale Supérieure de Cachan
(2002).
17. Armstrong, P.J., Frederick, C.O., 1966. A mathematical
representation of the multiaxial Bauschinger effect.
G.E.G.B. Report RD/B/N731.