Tool condition monitoring: unscented Kalman filter for tool flank wear estimation in turning of Inconel 718

Abstract

Abstract This article presents flank wear estimation during the turning process of Inconel 718 in dry cutting condition using unscented Kalman filter (UKF). A discrete flank wear model is developed where two components of flank wear due to abrasion and diffusion are considered as state variables and the cutting force is taken as the output variable. The proposed model can be implemented for online tool wear monitoring and the UKF predicts the actual states of tool flank wear in real-time by measuring the cutting force variation. The simulation result of flank wear estimation using UKF is compared with the extended Kalman filter (EKF) under a similar cutting environment. The error of estimation for UKF is obtained less than that of EKF indicating better accuracy of UKF than EKF for tool condition monitoring of the proposed model. Hardware experiments are performed to validate simulated results of both UKF and EKF on the proposed model correlating the cutting force and flank wear components.