Revisiting machinability assessment: Towards total machining performance

Abstract

The term “machinability”, introduced over hundred years ago, is vague and cannot fully describe the performance of machining systems. Machinability databases established over many decades are outdated: missing recent advances, e.g., cutting tool grades, geometry, coatings, and cutting fluids effects. This keynote paper summarizes findings of a CIRP-sponsored three-year collaborative study in five interrelated topics. The paper presents a critical review of the state-of-the-art on these topics, the results of two major round robin tests, three industry-based case studies, and a novel predictive system of machining performance, utilizing advanced deep learning methods. Outlook and future directions are also presented.