= we N N 
tec 
oD ~~ 
t 
bd 
Godding, Robert 
  
The size of the measuring cell and the number of cameras is flexible and depends on the size and the complexity of the 
stamping. The current cell consists of 4 Pulnix TM 1300 cameras with 1300 x 1030 sensor elements and different 
illumination components. The cameras are controlled by a Mikrotron framegrabber and allow a synchronous image 
acquisition. 32 cameras can be added to the system. For the measuring of surfaces a projection unit has been integrated. 
The whole system is controlled by a NT-Workstation for the image processing, data evaluation and visualization of 
results. 
The software includes modules for the calibration and orientation of the cameras. In a first step all necessary parameters 
for interior and exterior orientation are determined, using a calibration field with coded targets, which is integrated in 
the measuring cell /26/. For the calibration process a resection with additional determination of camera parameters is 
used. After calibration a fully automatic measuring of the printed or projected grids can be done. 
The similar algorithms for the strain and surface measurement are object based. Starting from one point the algorithm 
predicts new points at the surface and does a measurement in at least two images. With the results of this measurement 
object points are predicted, projected to the images, measured and directly re-projected to the object. This measuring 
strategy allow to consider the local shape of the object and to predict the local shape for the further measurements. One 
advantage of the system is the use of more than two cameras. Critical or disturbed object details can be measured from 
different positions. This possibility make the system more robust against reflections. 
The current measuring time of approx. 400points/second is sufficient for the foreseen application. The accuracy of the 
system depends on the object size and the number of images, which are used for the determination of the grid points and 
is between 1:20000 and 1:50000 of the object size. 
5 APPLICATION EXAMPLES OF THE DEVELOPED SYSTEM 
In the following the main features of the new photogrammetric measurement system are demonstrated with two 
examples representing typical applications in sheet metal forming. In the first example the acquisition of the surface 
geometry of an aluminum car seat component with grid projection is described. The second example exhibits the strain 
analysis of a rectangular model stamping part utilizing visioplasticity. The verification of the measured data is 
performed with results of a finite element (FEM) simulation. 
5.1 Surface geometry measurement of a seat component 
The measurement of the stamping part surface geometry is important to achieve information about shape deviations 
caused by dents, sink marks and especially springback of the sheet material. In this case the springback behavior of a 
car seat component made of the aluminum alloy AA5182 with a sheet thickness so = 1.0 mm and a surface area of 500 x 
500 mm was examined. Since analyzing springback only requires the surface geometry instead of a fixed grid projected 
light was applied. 
Therefore, a square grid was vertically projected onto the flat part's surface (Figure 9a). The applied customary slide 
projector was additionally equipped with a light source control to adjust contrast between the shaded grid lines and the 
illuminated part’s surface. A special sized slide (30 mm x 30 mm) was used to achieve the largest possible projection 
area. The distance of the vertically projected grids was approx. 2.5 mm. 
  
Figure 9: Square grid projection on a seat component (a) and measured 3D-point cloud (b) 
The images were taken from different directions using four CCD-cameras. The 3D-surface shape data (approx. 30.000 
measured points) were gathered by the photogrammetric evaluation of the images (Figure 9b). The comparison of the 
  
International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B5. Amsterdam 2000. 295