Adaptive threshold discrimination and synchronous squeezing transform for high-speed milling chatter detection

The term "chatter" in the manufacturing process refers to the self-excited vibrations that arise during machining operations, which include drilling, turning, and milling. These vibrations can lead to several negative outcomes, such as low surface quality, shorter tool lifespans, and even possible machinery damage. In order to address these issues, this paper presents a novel technique called Adaptive Threshold Discrimination and Synchronous Squeezing Transform (ATD-SST). Adaptive thresholding and synchronous squeezing are innovatively combined in the ATD-SST technique to improve chatter detection and characterization, particularly in noisy and non-stationary conditions. The adaptive threshold discrimination components dynamically adjust threshold values to enhance signal discrimination, while the synchronous squeezing transform components synchronise the signal processing to maintain consistency throughout the dataset. It was discovered through a comparison between acceleration and sound signals that acceleration signals produced better chatter detection results. To efficiently identify chatter, eight different indicators were established. The proposed method yielded early detection durations of 0.46 s, 0.31 s, and 0.52 s for varying machining conditions (speeds, cutting widths, and feed rates per tooth), respectively. The experiment results validate the method's effectiveness and highlight its potential to identify chatter promptly, thereby lowering the risk of workpiece damage while improving product quality.