Milling chatter detection for a discrete-edge end mill based on optimized VMD

Milling chatter is a dynamic instability that deteriorates surface integrity, accelerates tool degradation, and limits productivity. For discrete-edge end milling, the intermittent edge geometry makes the measured vibration response strongly nonstationary, and chatter-related components are easily masked by spindle harmonics and broadband noise. This article develops an Improved Pelican Optimization Algorithm-Variational Mode Decomposition (IPOA-VMD) framework for chatter detection. IPOA searches for the VMD mode number and penalty factor directly from the measured signal, thereby avoiding empirical parameter assignment. After decomposition, the chatter-sensitive mode is selected according to frequency- and energy-based screening criteria, and its energy–entropy–center–frequency feature is then used to discriminate the machining state. Simulation and milling tests demonstrate that the proposed framework achieves accurate modal separation, enhances chatter-related feature representation, and provides reliable chatter identification under the investigated discrete-edge end-milling conditions.