Tool condition monitoring - An intelligent integrated sensor approach

Abstract

The main objective of this project was to integrate two sensors, i.e., vision and force, to create an on-line TCM system in milling with better performance. The successive image analysis method was developed to measure flank wear and detect breakage in-cycle while the spindle rotates. Features extracted in time domain from cutting force were used to detect breakage and estimate flank wear in-cycle with a SOM network, which was trained with the flank wear measured by vision. On-line experimental results show that vision could provide accurate and robust wear measurement, independent of cutting conditions. Furthermore, the SOM gave good estimation result by tracking the wear increment that was used to train it. The integrated system also detected breakage successfully. The monitoring results suggest that this TCM system may be used to detect breakage and estimate flank wear on-line in milling in the industry.