Analysis of signals for monitoring of nonlinear and non-stationary machining processes

An investigation of cutting process monitoring based on dynamic force and acceleration signals in the frequency and time–frequency domains is presented in this paper. The performance of a new data analysis technique, the Hilbert–Huang Transform (HHT), is used to analyze this process in frequency and time–frequency domain. This technique is also compared with the traditional Fourier transform method power spectra in the frequency domain approach. A comparison is made of the analysis of two commonly used signals: acceleration and dynamic force. The shift of the main frequency peak into lower frequencies and higher frequency fluctuations is considered as a cutter tool wear indicator. The appearance of new frequency indicates a cutter tool fault. The performance of dynamic cutting force signal using both HHT and Fourier transform shows better results within this study. The HHT (which deals with nonlinear and non-stationary signals) has been shown to be a robust tool for estimating cutter tool wear/fault.