Proposal of Robust/Precise Tool-Condition Monitoring by “Multi-Color/Orientation Lighting (M-COL)” Especially for Tool Flank Wear: Experimental Verification in Two Colors/Orientations

This paper proposes a novel tool-condition monitoring method by “Multi-Color/Orientation Lighting (M-COL)” to identify conditions such as the flank wear width and adhesion/recession. Conventionally, tool wear is identified by either direct observation using a white light source and a camera or indirect observation utilizing signals such as cutting force. However, the former methods are often affected by environmental conditions, and the latter methods are impractical, for example, due to high-cost dynamometers; therefore, further improvement in the precision, robustness, and practicality of tool-condition monitoring is desirable. To address these issues, the proposed method utilizes multiple-color light sources from different orientations. In this paper, two are used, focusing on the point that the wear direction and the clearance angle are generally known, where one lightens the worn area and the other lightens the non-worn area. Due to this lighting setup, the wear boundary can be detected even under disrupted environments, e.g., additional ambient illumination. In addition, even the adhesion/recession can be detected since the position of the initial cutting edge can be measured by the proposed method. Cutting experiments of steel were conducted to verify the accuracy and robustness of the proposed method. Specifically, the flank wear widths identified by the proposed method ranged from 84.0% to 93.4% of those measured by the conventional microscope and profilometer, the minimum peak-to-sidelobe ratio for the proposed method was 2.84 times that for the conventional method, and the standard deviations of the identified wear width due to image-position shifts and angular misalignment were 4.13 μm and 4.51 μm, respectively; the proposed method is accurate and robust. The overall results show that the proposed method is effective for practical tool-condition monitoring.