TOOL WEAR MONITORING IN DIAMOND TURNING BY FUZZY PATTERN-RECOGNITION

This paper introduces a fuzzy pattern recognition technique for monito ring single crystal diamond tool wear in the ultraprecision machining process. Selected features by which to partition the cluster of patter ns were obtained by time series AR modeling of dynamic cutting force s ignals. The wear on a diamond tool edge appears to be classifiable int o two types, micro-chipping and gradual, both very small compared to c onventional tool wear. In this regard, we used a fuzzy technique in pa ttern recognition, which considers the ambiguity in classification as well as the weakness of the cutting force variation, to monitor the di amond tool wear status, with satisfactory results.