Tool condition monitoring based on vibration signal from an on-rotor sensor in CNC turning process

Aimed at improving manufacturing accuracy and efficiency, online condition monitoring of manufacturing turning processes have received wide research interests. This paper investigates dynamics of a CNC system, which paves fundamentals for applying a novel On-Rotor Sensing (ORS) approach as the online tool condition monitor in the turning process. The CNC lathe system is modeled as a multi degree of freedom (MDOF) system combining workpiece, chuck and spindle system. Both steady and stochastic excitations are modeled to understand the main responses behaviors of the systems. It has shown that the first several modes in the frequency range are lower than 2kHz and are the main responses and can be effectively captured by a low-cost ORS system is fabricated for monitoring the turning process. Especially, the modal responses are examined as masses of workpieces decrease during the turning process. Subsequently, the root mean squared (RMS) and spectral centroid of the resonant frequency band of the spindle system are found to be good indicators for different tool wear conditions. These results show that the ORS technique is efficient, and the proposed method does not require any protracted computing time which show a promising prospect for online automatic manufacturing monitoring.