Bayesian updating of modal parameters for modeling chatter in turning

Abstract

Variations in the mechanics and dynamics of the machining process under operational conditions cause inaccuracies in chatter model predictions. The parameters of chatter models therefore require re-calibration based on experimental observations during machining. This paper presents a new Bayesian method for the in-process calibration of chatter model parameters in turning. The new method comprises two stages. First, we identify the dominant closed-loop poles of the machining system from in-process vibration signals; then, we use those poles in a Bayesian model updating method to determine the probability distributions of the model parameters. Compared to existing methods, which require experimental observations under both stable and unstable conditions, the presented method requires a limited set of vibration measurements during stable conditions only. Moreover, the updated probability distributions are used to establish credibility bounds around the Stability Lobe Diagrams (SLD). An experimental example is presented to demonstrate the efficiency and effectiveness of the presented method in enhancing the accuracy of chatter predictions in turning.