Investigation and Correction of Actual Microphone Response for Chatter Detection in Milling Operations

Integrating sensors in machine tools for monitoring purpose entails dealing with different issues, not only related to accessibility and safety but also to measureable bandwidth and linearity of the sensors. Those factors could be related to the sensor itself but also to sensor-machine interaction that could drastically affect sensor performances and reliability. This paper presents a dedicated experimental investigation of the actual response of microphone transducer inside the machine-tool chamber, highlighting the effects of the machine-tool chamber in altering response linearity. The identified response is then processed with specifically developed equalization filters to correct the measured response and rescale the amplitude of frequency contributions, as required by most chatter detection techniques. The main aspect of both the experimental identification procedure and the development of an effective correction approach are presented and discussed. Finally, the technique is tested in processing signals acquired in experimental chatter tests to estimate the achievable improvements.