Chatter identification in flank milling of thin-walled blade based on fractal dimension

Chatter is a kind of self-vibration with unstable behavior, and chatter identification is crucial in the machining process to ensure a better surface quality. In this work, vibration signals are obtained in different chatter conditions through a milling experiment, and a dimensionless indicator named fractal dimension is employed to reflect the chatter severity level. The fractal dimension in time domain is calculated directly and the fractal dimension in frequency domain is calculated after wavelet de-noising. With the increase of chatter severity level, the fractal dimension in time domain decreases and the fractal dimension in frequency domain increases. The experiment results show that the two indicators can detect chatter timely and effectively in the flank milling of thin-walled blade.