Life prediction system using a tool’s geometric shape for high-speed milling

Abstract

In recent years, machinery and tool technology has been developing rapidly. The accuracy of operations have also become more and more exact. Elsewhere, raw materials have also been honed, hoping to provide more useful properties than previously. Thus, how to find the best way to prolong the life of a tool subjected to hardened material cutting is the target of this research. Three kinds of tool angle of the endmill are used in this research; clearance angle, rake angle, and helical angle. The cutting conditions are the same; we only change the tool angle for all the cases studied. We attempt to discover better tool geometrical angles for the high-speed milling of NAK80 mold steel. The tool wear rate was measured through a toolmaker’s microscope and the roughness of the machined surface was measured by the roughness-measuring instruments after several complete surface layers were removed from the workpiece in the experiment. Also, a noise-mediator was used to detect the level of cutting noise during each surface layer workpiece removal of the high-speed milling process, and different noise levels were then compared with the tool wear rates for identifying noise characteristics in the case of an over-worn tool state. An abductive network was applied to synthesize the data sets measured from the experiments and the prediction models are established for tool-life estimation and over-worn situation alert under various combinations of different tool geometrical angles. Through the identification of tool wear and its related cutting noise, we hope to consequently construct an automatic tool wear monitoring system by noise detection during a high-speed cutting process to judge whether the tool is still good or not, and, so, the cost of milling can be reduced.