Vibration Signal Acquisition and Chatter Feature Analysis in Thin-Walled Part Milling

Chatter often occurs during the machining process of thin-walled parts, leading to surface defects, reduced machining accuracy, and even equipment damage. This paper focuses on the detection of milling chatter in thin-walled parts. For data acquisition, an experimental platform is built for the milling process of a thin-walled workpiece, and a series of machining experiments are set up with different cutting speeds and depths of cut. Time-domain and frequency-domain features are extracted from the dataset obtained from the milling experiments to recognize the machining state through features. These features are then analyzed under both stable cutting and chatter conditions. Based on logistic regression classification, the recognition accuracy of various features for identifying milling states is analyzed, and the relationship between features and machining states is discussed.