Planetary Dragging Cutting Edge Treatment and Edge Rounding Prediction on Carbide Tools

Abstract

This paper describes the use of the “Dragging” (DF) cutting edge preparation process with 2 grit sizes and three mixing ratios. Both the immersion depth of the tool in the abrasive medium and the dragging duration time were manipulated. A Repeatability and Reproducibility (R&R) analysis and edge radius (ER) prediction were carried out using Machine Learning by Artificial Neural Network (ANN). The results achieved were that the influencing factors on the ER in order of importance were drag depth, drag time, mixing percentage and grain size respectively. Furthermore, the reproduction accuracy of the ER is reliable in comparison with traditional processes such as brushing and blasting and the prediction accuracy of the ER of preparation with ANN was 94% showing the effectiveness of the algorithm. Finally, it is demonstrated that DF has reliable feasibility in the application of cutting-edge preparation on carbide tools.