Digital modeling-driven chatter suppression for thin-walled part manufacturing

Abstract

Thin-walled parts are widely used in various industries such as aerospace and automotive, but the manufacturing processes are often harmed by chatter which is a self-excited vibration because of the poor rigidity in the direction perpendicular to the wall surface. The traditional stability lobe diagram (SLD) method can predict chatter based on the manufacturing system and workpiece parameters. However, these parameters could vary along with the manufacturing execution, compromising SLD's accuracy and even feasibility. To enable effective chatter suppression in thin-walled part milling, this study proposes a digital twin model, where two sub-models including the cutting parameters optimization and chatter detection are established. In the sub-model of cutting parameters optimization, a real-time SLD considering the time-varying modal parameters at the cutting region of the workpiece is generated as the optimization criteria. The sub-model of chatter detection can recognize chatter by a fusional analysis of the multiple sensors' signals, including vibration, force, and sound. Considering the bias of real-time SLD, these two sub-models are combined to output optimized cutting parameters to avoid chatter. Besides, a monitoring window to visualize the milling scenario and a database to record the manufacturing data are implemented in the digital twin model. According to the milling experiments, the digital twin model is validated to perform more effectively in chatter suppression than the traditional stationary SLD method.