On-Line Flank Wear Estimation Using an Adaptive Observer and Computer Vision, Part 2: Experiment

Abstract

An on-line flank wear estimation system, using the integrated method presented in Part 1 of the paper, is implemented in a laboratory environment, and its performance is evaluated through turning experiments. A computer vision system is developed using an image processing algorithm, a commercially available computer vision system, and a microscopic lens. The developed algorithm is based on the difference between the intensity of the reflected light from a flank wear surface and that from the background. The difference is very significant and an appropriate selection of the intensity threshold level yields an acceptable binary image of the flank wear. This image is used by the vision computer for the calculation of the flank wear. The flank wear model parameters that need to be known a priori are determined through several preliminary experiments, or from data available in the literature. Cutting conditions are selected to satisfy the assumptions made on the design of the adaptive observer presented in Part 1. The resulting cutting conditions are typical of those used in finishing cutting operations. The integrated method is tested in turning experiments under both constant and time varying cutting conditions, and yields very accurate on-line estimation of the flank wear development.