Abstract
Immediate detection of abnormalities which occur in a production machine and taking a prompt and appropriate action are essential to maintain. Repair is required before the defective product is made in an actual production plants. This study aims to develop a condition monitoring methodology for a punching machine based on wavelet analysis of the impulsive acceleration response during the punching process. The acceleration response on the die in a turret punch press was measured during a repetitive punching test and analyzed with wavelets whose mother wavelet was the third derivative of the Gaussian function. The signal concentration measure evaluated from the extrema of the wavelet transform showed a considerable correlation with the wear transition, thus, it was suggested that it could be a signature of the tool wear progress and useful as a potential measure for the tool wear monitoring.
