Estimation of vibration stability in milling of thin-walled parts using operational modal analysis

The system dynamic model is essential for chatter prediction in milling of thin-walled structures, which is largely affected by the estimation precision of dynamic modal parameters. This paper presents a method of dynamic modal parameter identification from response signals through operational modal analysis (OMA). To avoid omission of the modes and improve recognition accuracy, multichannel least square complex exponential (LSCE) method and multichannel autoregressive moving average (ARMA) method are applied to identify the modal parameters by processing multiple sets of response signals simultaneously. The convergence characteristics of the damping stability diagram at different natural frequencies are employed to eliminate the false modes caused by harmonics and model order. After that, the predicted stable boundaries of the milling system are estimated by an extended semi-discretization method, which incorporates the effects of multi-modes, multi-point contact, and regeneration chatter caused by the interaction between tool and thin-walled part. Through the milling experiment validation, it is shown that the dynamic modal parameters can be identified accurately, and chatter can be well predicted.