Machine Vision-based automated Tool Wear Inspection System
Author(s): Prof. Robert Schmitt, Philip Hafner, Bjoern Damm
Affiliation(s): Laboratory for Machine Tools and Production Engineering WZL
Topic: 4. Image Processing and Simulation
Disturbances occurring within mill-cutting processes are most likely to affect the produced parts. Especially deviations due to wear of the used cutting tools have significant influence on the workpiece features. Primary causes for tool wear are mechanical abrasion, diffusion processes, built-up edges and scaling. In order to take effective and robust actions for process optimization, the tool deviations must not only be known qualitatively, but also have to be quantified. This results in the necessity of flexible, direct measurements on the applied tools. For this purpose, a machine vision system has been developed at the Laboratory for Machine Tools and Production Engineering and is presented in this paper. This system enables quick inspections on the shop floor avoiding any subjectiveness of human operators. One vital component of this system is a newly developed illumination device, which can be flexibly adapted to reflection properties of different cutting materials and coatings. By taking two pictures using different lighting conditions, the illumination device allows the machine vision software to segment and measure the wear area of the cutter’s perimeter edges. Using the developed illumination strategy and customized algorithm chains it is possible to enhance existing commercialized pre-setting and measurement tools for the classification of tool wear, such as flank wear and broken dies.