Timely and efficient chatter onset detection for end-milling

Abstract

Machining stability and chatter are issues that significantly limit the efficiency of metal removal processes, and for this reason they have been intensively investigated in relation to end milling and turning operations. Regardless of the advancements in machining stability prediction and modeling, there is still great need for the development of an effective chatter detection and monitoring method. Timely chatter detection prevents the damaging of the cutting tool and workpiece, as it allows for immediate intervention in the cutting process. This paper focuses on right-on-time detection of chatter onset which can be achieved by utilizing computationally efficient methods and by minimizing signal processing time lag. The first step of the proposed method is to decompose the measured vibration signal by using Gabor filter bank. Then, for the detection of the transition in cutting dynamics at the onset of chatter, the Teager’s nonlinear energy operator is used which can track variations in instantaneous frequency and amplitude. The main advantage of the energy operator is its computation efficiency with minimal time delays. The proposed method is tested on data obtained from slot end-milling of an inclined surface (incremental depth of cut) for different level of spindle speed.