Application of Discrete Wavelet Decomposition in Monitoring of Hole-making Inconel 718

The paper describes an application of discrete wavelet decomposition of monitored signals from Profibus data originating from spindle and from the feed drive of machine tool. Discrete wavelet decomposition is a most efficient method to separate uniform signal pattern from disordered signal. It was demonstrated that - compared to the evolution of average spindle torque over the tool path - extracted signal variance from the spindle torque through application of wavelet transformation has a significantly higher correlation with maximum flank wear on the tool. Within the framework of tools and cutting conditions chosen in this work, this is true both for drilling with spiral drills and for finishing with face-cutting finishers.