Godding, Robert
measured 3D-point cloud with the CAD model of the stamping part supplies a statement about geometrical deviations
of which relevant springback parameters (e.g. the bottom curvature) can be derived.
The vertical projection of square grids using one projector is sufficient for the surface acquisition of flat-shaped parts.
For more complex stamping part geometries additional projectors and cameras are necessary. With respect of error-free
image processing overlapping of the projections have to be prevented. Therefore, further investigations have to be made
to synchronize the CCD-cameras and projector units. An alternative method may be the contemporaneous overlapping
of colored grids, which presupposes the use of different optical filters and modifications of the image acquisition
software algorithms.
5.2 Strain analysis of a rectangular model part
This second example demonstrates the strain analysis of a rectangular model stamping part utilizing the method of
visioplasticity.
For the calculations of the strain-state it is necessary to process the 3D-coordinates of points, which are fixed on the
parts surface. The software algorithms for the surface measurement (either projected or fixed grid) are similar. An
additional software module for the visioplastic analysis using algorithms similar to those published in /21,22/ was
developed at IFUM and implemented in the proposed system.
To apply fixed grids on the sheets different marking techniques (e.g. electrochemical etching, laser marking and
printing) have been investigated at IFUM. As a result of the tests silkscreen printing yields the best results concerning
applicability for different sheet materials. Surface reflection behavior due to different sheet coatings and illumination,
wear resistance, which are of particular importance to achieve accurate grids and a sufficient contrast for the image
processing, were taken into consideration as well.
For the investigations experimental deep drawing tests as well as FEM simulations were performed. Following the
procedure of the investigations is explained with one example:
A zinc coated blank sheet out of the soft steel grade DCO5 (sheet thickness so = 1.0 mm) was marked with a 2.5 mm
square grid using silk-screen printing technique. For the deep drawing a rectangular model tool (220 mm x 110 mm)
was used. For the strain analysis especially the transition area between the punch radius and the side walls in the corner
areas are of special interest, since the highest strain rates and - in case of failure — the occurrence of optimum cracks
/27/ can be observed here.
A comparison between the major strains ¢, calculated with measured data and by FEM-simulation is shown in
Figure 10. For the simulation of the deep drawing process the FEM software "PAM STAMP" was used.
punch geometry 220 x 110 mm sheet material DCO5
drawing depth: — h *45mm sheet thickness: 8, 1.0 mm
punch radius: x= 12mm blankholder force: — £F, 100 kN
die radius: n,7 12mm drawing ratio: BL, = 18
Corner radius: ft.cd2mm lubrication: Raziol CLF 240
0.323
dôtE
major strains calculated with
optical measured data EM-simulation © FÜM
Figure 10: Verification of a FEM-simulation by automated strain analysis
As shown in Figure 11 and 12 the principal strains Q9; and q» were plotted in forming limit diagrams (FLD). As
expected all strains are located below the forming limit curve (FLC), which was determined in former investigations.
Therefore, in the investigated corner area no critical strain state is determined. A comparison of both forming limit
diagrams shows a comparatively high coherence (Aq « 4 9). Major strain @,, minor strain q», principal strain direction,
effective strain @y (von Mises), sheet thickness s and thinning As can either be visualized with a 3D-viewer or exported
into ASCII files.
296 International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B5. Amsterdam 2000.
