Multi-modal Sensing for Machine Health Monitoring in High Speed Machining

Abstract

Optimum performance of machining process relies on the availability of the information about process conditions for process monitoring and feedback to the process controller. Tool condition is the most crucial and determining factor to machine tool automation, hence online tool condition monitoring is of great industrial interest. A research work of tool condition monitoring for high speed machining is introduced in this paper. It employs multi-modal sensing which includes accelerometer, acoustic emission (AE) sensor and dynamometer, and advanced signal processing to monitor a high speed milling process. The results show that the frequency bands of wavelet decomposition which cover the frequency of cutter revolution are the most important bands among the spectrum. The energy distribution of signal shifts from low frequency to high frequency while tool wear develops. Wavelet analysis has the advantages of going deeper to the nature of physical phenomenon. The results based on time-frequency domain analysis are not so easy to be influenced by the noise and the cutting parameters which has always been a big problem for time-domain analysis.