Tool Wear Monitoring in Diamond Turning by Fuzzy Pattern Recognition

Abstract

This paper introduces a fuzzy pattern recognition technique for monitoring single crystal diamond tool wear in the ultraprecision machining process. Selected features by which to partition the cluster of patterns were obtained by time series AR modeling of dynamic cutting force signals. The wear on a diamond tool edge appears to be classifiable into two types, micro-chipping and gradual, both very small compared to conventional tool wear. In this regard, we used a fuzzy technique in pattern recognition, which considers the ambiguity in classification as well as the weakness of the cutting force variation, to monitor the diamond tool wear status, with satisfactory results.