Deep Drawing 2

JOURNAL DE PHYSIQUE IV
Colloque C7, supplkment au Journal de Physique 111,Volume 3, novembre 1993
Strain hardening of aluminium alloy 3004 in the deep drawing and

ironing processes
J. COURBON and J.L. DUVAL*
Pechiney CRYBE1 27, 38340 Voreppe, France
* Laboratoire des Propriktks Mkcaniques et Thermodynamiques des Matkriaux (LPMTM-CNRS), Univ.
Paris-Nord, avenue J.B. Clkment, 93430 Villetaneuse,France
ABSTRACT
The evolution of material hardening resulting from the canmaking operations on aluminium beverage cans
has been investigated. Tensile tests in cup walls revealed that deep drawing induced softening in the
hoop direction and hardening in the meridian direction. This anisotropy is retained in the ironing operation.
Changes in strain path on a heavily cold-rolled material probably cause such a complex behaviour. To
determine hardening laws for deep drawing, simple shear tests were thus performed because of the strain
path similarity. They allowed to determine hardening laws over larger strains than tension could reach and
revealed a saturation of stress. Altogether they proved adapted to the understanding of deep drawing.
INTRODUCTION
Aluminium alloy 3004 is commonly used for the production of beverage can bodies. It consists of
commercial purity aluminium (0,4 wt% Fe and 0,2 wt % Si) alloyed with Mn : 1 wt%, Mg : 1 wt% and
Cu : 0,15 wt% to give strengthening by solid solution, by Mn-based dispersoids and mostly by dislocations. Strain-hardening is especially important in this alloy : the elastic limit is about 70 MPa in the
annealed state and 280 MPa in the cold-rolled sheets delivered to canmakers.
Cups are deep-drawn and redrawn from these sheets, then their wall is ironed through three dies,
increasing the height of the can body while reducing its thickness. The deep drawing and ironing
operations with their specific strain paths modify the hardening of the material. Besides the impact on the
can body final properties such as vertical crush resistance, such effects must be incorporated into
numerical simulations of the process to make them realistic. Thus a specific study of the evolution of
hardening at the various stages of the process has been developed.
CHANGES IN THE WALL MATERIAL

Deep drawing results in a strain path close to pure shear (fig.l), there is a hoop shrinking and a meridian
elongation. The shear directions vary with respect to the orthotropic rolling axes all around the can. The
texture anisotropy resulting from the rolling operations, though kept as low as possible, results in light
earing and thickness heterogeneities all around the cup wall.
Ironing strains are close to those of rolling with a through-thickness shear of the wall (fig.1).
Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jp4:1993739
punch
Fig.1 :sketch of the strains in the deep drawing and ironing operations.
Fig.2 :evolution of the 02 %proof stress and tensifestrength of cup walls,

showing the dzfference between meridian (left)and hoop (right)specimens.
0 :as-rolled sheet ;I :draw ;2 :redraw ;3 , 4 and 5 :ironing passes
The resulting changes in hardening have been evidenced by cutting hoop and meridian tensile specimens in
cup walls. Care has been taken to measure the thickness at many points of the specimens, to account for
heterogeneities. The averaged features are shown on Fig.2. The process parameters were :diameter at first
draw 60%, at second draw 48% of initial flange diameter ; wall thickness at first ironing pass 78%, at
second pass 58%, at third pass 36% of initial sheet thickness. The meridian specimens tend to workharden all along the stages of the process. On the other hand, hoop specimens soften during drawing and
redrawing.
Strain softening during deep drawing of aluminium alloys had previously been reported by Grimes et al.
[l] and Merchant et al. [2].Yet the situation is more complicated since only hoop specimens are actually
softened. The change in strain path between rolling and deep drawing with subsequent dislocation
annihilation probably accounts for the phenomenon. This is backed up by the recent findings of Shen [3] :
a 3004 partially recovered after rolling has a more stable microstructure and does not soften during
drawing. Afterwards, ironing work-hardens the material again. But the difference created between hoop
and meridian specimens does not disappear.
Such tensile tests on cup walls already bring useful information on the material hardening ;but they cannot
explain the material behaviour in the deep drawing process :the only use of an arithmetic equivalent strain
appears dubious since the hardening becomes negative. There is an effect of dislocation microstructure,
and such an effect is linked with the change in strain path from rolling to deep drawing. To understand the
softening, one has to develop a mechanical test close to this specific strain path, then develop macroscopic
models including kinematic hardening, for isotropic hardening would never give rise to strain softening.
Pure shear tests appear adapted to the situation :they are developped in the next section.
PURE SHEAR TESTS FOR DEEP DRAWING

As stated, defining stress-strain laws adapted to the deep drawing operation is not straightforward. Strains
of 50-60% are reached in cup walls. Tensile tests are not very useful since the strain path differs very
much and the homogeneous ductility is limited to 5% because of the necking instability. Simple shear tests
appear more adapted because of the strain similarity. Fig.3 shows the principle of the test, essentially
developed by Rauch et al. [4], that can be adapted to any tensile device. The instability to avoid is not
necking but buckling, since one diagonal of the thin specimen is undergoing a compression. A proper
choice of dimensions for the sheared zone is able to prevent it [5] : the specimens were 0,30mm thick,
38mrn long and the sheared zone was 2mm broad.
The specimens were taken at different angles to the rolling direction and a grid was printed parallel to the
shear direction, to compare the strain monitored from the move of the machine to the one measured from
grid deformation.
Fig.4 shows a typical stress-strain curve. Uniform strains are very much increased and the stress-strain
laws differ from the ones observed in tension. Most noticeable is the absence of hardening after a certain
amount of shear strain (-10%). This evolution could not be guessed from tensile testing, since the uniform
tensile ductility of 3004 is too low (4%) : on the contrary, a Swift-type hardening cr = K (E + E O ) ~was
observed up to tensile necking.
\ Specimen
Fig.3 :shear testing device. The sheet is in the sketch plane, fastened in thefmed and mobile grips.
Y
Fig.4 :shear stress z / shear strain ycurve (3004 sheared at 0,45,90 to the rolling direction).
Strain localization provokes an early rupture at 45 Oand 90 "
when compared with shear at O0 to the rolling direction.
250
In order to translate the T/Y (shear stress /plastic shear strain) curves into G/E tensile curves, or vice-versa,
a plasticity model is needed. Even though cold-rolled 3004 is not perfectly isotropic, we may use the von
Mises yield criterion, which brings the equivalences : o = GT
and E = y / f l . A closer analysis of the
curves in terms of the plastic power allowed to replace fl by 1,9.
Fig.5 shows the comparison of tensile and shear curves in terms of Vy. The Swift extrapolation of tensile
curves clearly results in an overevaluation of the flow stress in the material, since no saturation of
hardening was predictable in tension.
Fig5 :comparison of equivalent shear stresses and strains broughtfrom tensile tests and from pure
shear tests. Tensile hardening has been extrapolated in the large strain range from the Swij? law.
IRONING
A direct evaluation of the curve using force measurements in an ironing operation was not tried, since
friction on both punch and die sides play a major role in the resulting forces when compared to workhardening. Tensile tests on can walls with different ironing reductions allow to define a stress-strain curve.
After redrawing, there is a steep hardening but it then tends to saturation at ironing strains above 1.
CONCLUSION
Simple tensile tests performed on cup walls between stages of the canmaking operation allowed to detect
hardening or softening effects linked with the strain path used. The mere use of isotropic hardening is not
adapted to these forming operations, especially deep drawing. Ironing is more simple but comes on a
structure modified by the deep drawing stages.
Simple shear tests proved useful to give an experimental stress-strain law because of their similarity to
deep drawing and the large range of homogeneous deformation. A saturation of hardening was evidenced
and the danger of extrapolating tensile curves to large deformations shown. Altogether, the simple shear
test proved useful for sheet material characterization for deep drawing purposes.
REFERENCES
[I] Grimes R. and Wright J.C., J. Inst. Metals 9 5, (1968), 182.
[2] Merchant H.D., Hodgson D.S., O'Reilly I. and Embury J.D., Mats Character.25 (1990), 251.
[3] Shen T.H., TMS Fall Meeting, Symp. on Aluminum Alloys for Packaging, Chicago, (1992).
[4] Rauch E.F. and G'Sell C., Muter. Sci. Eng. A 1 11, (1989), 71.
[5] Genevois P., PhD Thesis, INP Grenoble, France (1992).

Deep Drawing 2

Uploaded by

Copyright:

Available Formats

Deep Drawing 2

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Deep Drawing 2

Uploaded by

Copyright:

Available Formats

JOURNAL DE PHYSIQUE IV

Colloque C7, supplkment au Journal de Physique 111,Volume 3, novembre 1993

Strain hardening of aluminium alloy 3004 in the deep drawing and

CHANGES IN THE WALL MATERIAL

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jp4:1993739

Fig.2 :evolution of the 02 %proof stress and tensifestrength of cup walls,

PURE SHEAR TESTS FOR DEEP DRAWING

You might also like