Dual-Mode Type Algorithm For Chatter Detection In Turning Considering Beat Vibration

The self-excited chatter is a complex dynamic behavior of the cutting process, which results in severe damage to the machining system. Therefore, an effective approach to chatter detection is critical for preventing adverse effects caused by chatter. Analysis of signals from turning processes shows that the beat vibration occasionally arises, which renders those chatter detection methods based on energy change invalid. In this paper, a dual-mode type algorithm for chatter detection in turning is proposed when considering the presence of beat vibration. The wavelet packet entropy is selected as the chatter detection indicator and its standard deviation in a beat period is utilized to determine whether beat vibration occurs. Accordingly, the effective chatter indicator under two signal modes is attained in different ways. As a result, chatter vibration can be detected even when the measured signal possesses the beat characteristics. Simulations and experimental study explain the reason for the emergence of the beating signal. The results of simulations and experiments demonstrate that the developed chatter detection approach can recognize chatter when the beat phenomenon appears.