Multi-feature extraction and analysis for boring chatter monitoring

Chatter will cause a series of changes in machining state during the boring process. Due to the difference in the sensitivity and ability of various sensors to resist external interference, the change of machining environment can lead to variation in response and make difference to the characteristics of signal. Therefore, it is essential to extract the features related to chatter from the machining signals. In this study, the dynamic model considering the regenerative effect is constructed, based on which the relationship between the limited cutting width and the spindle speed is determined. In addition, eddy current displacement sensor, vibration acceleration sensor, and acoustic pressure sensor are applied to propose a feasible boring chatter monitoring scheme and collect the machining signals. The time and frequency domain of the collected signals are analyzed to determine the pattern of time–frequency variation for three kinds of sensor signals generated with chatter.