Acoustic Tool Condition Monitoring with Angular Resolution on the Cutting Edges

This paper presents a new computational approach to the implementation of acoustic monitoring of precision CNC milling processes. The audio data is processed in the phase domain, in contrast to past solutions focusing on the time and frequency domains. This yields a high angular spatial resolution in the monitoring of wear around the perimeter of the tool; with this, each cutting edge can be evaluated individually. The results obtained are similar to those of direct measurement methods, despite the acoustics monitoring being indirect. Metrics are defined to monitor both the tool as a whole and to characterize the individual cutting edges. The methods were verified by applying them to significant volumes of data collected during the milling of numerous parts with multiple tools. Here, the detailed results for one tool are presented.