Levenberg-Marquardt算法
灵敏度(控制系统)
算法
过程(计算)
基质(化学分析)
面子(社会学概念)
计算机科学
控制工程
工程类
人工智能
材料科学
人工神经网络
电子工程
社会科学
操作系统
社会学
复合材料
作者
Khoe-Qui Le,Yu-Ren Wu,Trong-Thuan Luu
出处
期刊:Journal of Manufacturing Science and Engineering-transactions of The Asme
[ASME International]
日期:2024-06-12
卷期号:146 (9)
被引量:1
摘要
Abstract Currently, numerous studies have applied gear skiving processes to produce face gear. However, there remains a significant challenge in achieving a flexible computing model for manufacturing a precise tooth surface for face gear. This study proposes a novel mathematical model that combines the cutter modification method and computer numerical control (CNC)-axis motion modification methods within a unified “closed-loop optimization.” This approach aims to enhance the tooth surface accuracy of skived helical face gears by determining optimal coefficients. Applying the Levenberg–Marquardt algorithm and sensitivity matrix enables the calculation of new polynomial coefficients, ensuring the attainment of gear surfaces with an accuracy grade of B6 (according to the ANSI/AGMA 2009-B01 standard) for each target surface. The proposed methodology involves the generation of a helical skiving cutter using a corrected rack. Subsequently, the cutting path on the CNC machine is optimized by incorporating additional motions expressed in polynomials. A comprehensive skiving simulation is conducted to achieve the desired face-gear surface, which is corrected by specified polynomial coefficients. The proposed model is validated through numerical and machining simulations using vericut software. The results affirm the practicality and efficacy of our approach in achieving the desired accuracy in producing helical face gears through power skiving processes.
科研通智能强力驱动
Strongly Powered by AbleSci AI