Monitoring brittle/ductile material removal mechanisms in micro-machining using wavelet packet analysis and green's function solutions of AE signals

摘要

Micro-machining of brittle materials such as Silicon, glass and ceramics involves both plastic deformation and brittle fracture [1]. Monitoring and identifying the dominant material removal mechanisms in such processes is important for achievinghigh productivity and surface finish with controlled tool wear. As the transient elastic energy that is spontaneously released when materials undergo deformation or fracture, acoustic emission carries rich information on the type, location, severity, andother characteristics of the deformation mechanisms at a microscopic scale. Therefore, acoustic emission signals can be regarded as signatures of the source events occurring at the location of deformation. This study examines the different elastodynamicrepresentations and characteristic waveforms for brittle and ductile material removal mechanisms based on Green's function solutions. Earlier work on this was done by Daniel [2]. A feature extraction scheme is then developed using wavelet packet analysis with best basis selected with respect to the characteristic waveforms to achieve high representation efficiency.