To provide a tool abnormality detection device and tool abnormality detection method whereby tool abnormality can be detected with a high accuracy.
A CPU measures oscillation given an X-axis ball screw or an Y-axis ball screw which is stopped during processing of a work-piece W with use of a disturbance observer (S1). Then, a component same as a frequency of number of rotations of a tool T is detected and analyzed by wavelet transformation of the disturbance force estimated by the disturbance observer (S2). Further, on the basis of the analysis result, the intensity of a spectral component same as the frequency of number of rations of the tool T is calculated (S3). Then, the calculated spectral intensity is compared with a first threshold and a second threshold (S4, S6), whereby presence or absence of a failure can be detected with a high accuracy and the level of such a failure can be determined.
KOJIMA TERUHISA
NOMURA HIROAKI
AOYAMA TOSHIRO
KAKINUMA YASUHIRO
KOIKE AYA
ONISHI KOHEI
Masatoshi Yasui