Systematic Investigation of Tool Wear Monitoring in Turning Operations

Abstract

This study investigates the feasibility of monitoring tool wear and surface finish based on force signal by conducting a set of experiments in turning operations. Our study shows that the force signal appears curvi-linear with both uniform and maximal tool wear and shifts systematically across different cutting conditions; the force frequency spectrum contains extra information yet proves difficult to interpret and utilize; the surface finish behaves rather erratic. Three modeling techniques, multi-layer neural networks, adaptive neural-fuzzy inference system and multiple linear regressions are all capable of capturing the underlying force-wear dependency. Inferences are made regarding some essential issues, e.g. wear classification versus continuous wear estimation, force, vibration, acoustic emission versus sensor fusion. The study also suggests that a monitoring scheme should ultimately link measured signals to workpiece geometry inaccuracy rather than just to tool wear parameters. Additionally greater efforts should be directed to the integration of tool wear monitoring and geometry tolerances control.