Project
ToolSpy – Automated tool wear measurement

 

Problem

Nowadays the reduction of tool costs offers an enormous potential to increase the competitiveness. The reasons are that many measuring methods cannot be applied or their usage is too costly. Therefore tool wear is measured rarely and only by manual subjective optical measuring systems (e.g. microscopes). Thus exchange and regrinding times are only based on statistical calculations and not on the current tool wear. Hence optimisation potentials in tool costs can be exploited.

The solution is a machine-vision-system. Because of the optical characteristics of the different cutting materials interferences can appear. Reflexions and partially shadings constrain a reliable image acquisition and thus a robust image processing.



Solution

Based on an illumination unit developed at WZL, which is adapted on the geometrical and optical conditions of milling tools, these perturbations are reduced.

Thus it is possible to realise optimised input data for the following image processing steps. Within the pre-processing the cutting edges and their positions are detected as well as the wear area is segmented by the use of adjusted reflexions.




Subsequent to these image processing steps the wear type can be classified and the flank wear (VB, VBmax) can be measured.