A statistical analysis of acoustic emission signals for tool condition monitoring (TCM)

The statistical properties of acoustic emission signals for tool conditionmonitoring (TCM) applications in mechanical lathe machining are analyzed inthis paper. Time series data and root mean square (RMS) values at various toolwear levels are shown to exhibit features that can be put into relation withageing in both cases. In particular, the histograms of raw data show power-lawdistributions above a cross-over value, in which newer cutting tools exhibitmore numerous larger events compared with more worn-out ones. For practicalpurposes, statistics based on RMS values are more feasible, and the analysis ofthese also reveals discriminating age-related features. The assumption thatexperimental RMS histograms follow a Beta (b) distribution has also beentested. The residuals of the modeling b functions indicate that the search fora more appropriate fitting function for the experimental distribution isdesirable.