A Diagnostic Approach for Turning Tool Based on the Dynamic Force Signals

Abstract

In the current work it is proposed a simple, and fast softwired tool wear monitoring approach, based upon the features of the time series analysis and the Green’s Function (GF) features. The proposed technique involves the decomposition of the force signals into deterministic component and stochastic variation-carrying component. Then, only the stochastic component is processed to detect the adequate autoregressive moving average (ARMA) models representing the tool state at every wear condition. Models are further reduced to form a more representative parameter, the “Green’s Function (GF).” This reflects the dynamic behavior of the tool prior to failure and, may provide a comprehensive and accurate measure of the damping variation of the cutting process subsystem at different forms of tool’s edge wear. As wear enters the high rate region, the cutting process is forced toward the instability domain where it tends to have less damping resistance. It is also explained how a system response surface can be generated based on its Green’s function. It is proposed that this concept can be the basis for a diagnostic technique for use with many systems.