Host: The Japan Society of Mechanical Engineers
Name : [in Japanese]
Date : October 04, 2024 - October 05, 2024
In recent years, gear skiving has gained attention as a highly efficient machining method for internal gears used in reducers. This process involves the synchronized rotation of the workpiece and tool, enabling the continuous machining of multiple teeth and thus achieving high machining efficiency. However, a significant challenge with this method is the reduction of tool life. To address this issue, it is necessary to calculate tool wear conditions and cutting resistance. Previous studies have estimated the cutting thickness of various parts of the cutting edge and predicted cutting resistance during gear skiving through numerical analysis. In these analyses, the intersection of the region where the tool's cutting edge passes once for each groove of the workpiece and the region where the cutting edge passed previously is determined by sequential processing. However, this method imposes a high computational load due to the numerous intersections that need to be derived, resulting in calculation times several times longer than the actual machining time. To overcome this challenge, I developed a method to reduce calculation time by replacing sequential calculations with parallel processing. The developed method utilized a GPU, an ultra-parallel computing device, to achieve high-speed computation. As a result, the execution time of the program was reduced to approximately 50% compared to previous studies, and similar cutting force data was obtained.
