A Study on the Cutting Tool Wear by Time Series Approach

시계열분석 방법 에 의한 절삭공구 의 마멸 에 관한 연구

A new indirect tool wear sensing technique is proposed using time series analysis. The acceleration measured on the tool post in the vertical direction during turning is sampled with uniform interval, and fitted following the ARMA modeling procedures. Various signal characteristics are observed in respect with the progress of flank wear. From those observations it is believed that; *The variance of the signal is approximately proportional to the increase of flank wear. *The absolute power of a dynamic mode decreases in the beginning of cutting until the maximum flank wear increases up to 0.4-0.5mm, and then increases. *The other characteristics are not so much related to the tool wear as the signal variance and the absolute power of a dynamic mode. Hence, the absolute power of a dynamic mode seems to be a good factor for the indication of tool-change time regardless of tool material or cutting conditions.

본 연구에서는 그러한 간접적인 방법의 하나로서 공구대로부터 가속도신호를 절삭시간의 경과에 따라 측정한 후, 시계열자료를 모델화하는 전산프로그램을 사용하 여 적합한 모델을 구하고 그로부터 유도되는 여러 특성들을 관찰함으로써 보다 현실적 인 방법론을 제시하고자 하였다.