A time-frequency acoustic emission-based monitoring technique to identify workpiece surface malfunctions in milling with multiple teeth cutting simultaneously

In the view of implementing automated solutions for monitoring complex machining processes such as milling, the usage of acoustic emission (AE) in machining is regarded as a promising way for assessing machine tool condition and for in-process detection of workpiece malfunctions. Correlating AE signal events with the occurrence of workpiece surface anomalies (e.g. laps, material drag) can be a powerful method for scrap reduction of expensive components such as those employed in aerospace industry. This paper proposes new methods for supervising cutting processes with multiple teeth cutting simultaneously, i.e. milling, by use of AE signals backed-up by force data. This is done by taking into account signals patterns when one, two or three teeth are cutting simultaneously, situation that often occurs in real milling applications. The research shows for the first time that identification of milling conditions (i.e. cutting with one/two/three teeth) is possible using only AE signal in time-frequency (T-F) domain. Moreover, detection of surface anomalies, such as folded laps that are generated by damaged cutting edges can be successfully identified in various milling conditions. The paper demonstrates that time-frequency analysis of AE signals empowered with advanced processing techniques has great potential to be used in flexible and easily to implement monitoring solutions to enable milling of anomaly-free workpiece surfaces in difficult-to-cut aerospace materials.