Scholarlink

<p>This paper utilizes limited simulated data on the spatial distribution of thermal load to achieve rapid prediction of the transient thermal load distribution on the rake face of gear hob, thereby elucidating the time-varying characteristics of thermal load spatial distribution. By leveraging the analytical expression of the metal cutting heat distribution coefficient, the primary influencing factors of thermal load spatial distribution are explored, alongside an exposition of methods for numerically analyzing chip formation and the rake angle during hobbing. For addressing the spatial distribution of thermal load, a&#xa0;fitting model for thermal load spatial distribution is established based on the Hippo optimization algorithm. Building upon this, a&#xa0;prediction model is developed using convolutional neural networks, with basic parameters as inputs and hyperparameters of the thermal load fitting model as outputs. Finally, a&#xa0;comparative analysis is conducted between the prediction model and finite element simulation data, demonstrating that the average relative error does not exceed 3.16%. This model provides support for research on the optimization of hobbing process parameters and tool wear.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Predicting thermal load spatial distribution on high-speed dry gear hob

  • 期刊论文

  • 发表日期：

    2025年4月28日

查看全文
  • Lin Li,
  • Yongpeng Chen,
  • Jin Dai,
  • Wenqiang Qin

摘要

This paper utilizes limited simulated data on the spatial distribution of thermal load to achieve rapid prediction of the transient thermal load distribution on the rake face of gear hob, thereby elucidating the time-varying characteristics of thermal load spatial distribution. By leveraging the analytical expression of the metal cutting heat distribution coefficient, the primary influencing factors of thermal load spatial distribution are explored, alongside an exposition of methods for numerically analyzing chip formation and the rake angle during hobbing. For addressing the spatial distribution of thermal load, a fitting model for thermal load spatial distribution is established based on the Hippo optimization algorithm. Building upon this, a prediction model is developed using convolutional neural networks, with basic parameters as inputs and hyperparameters of the thermal load fitting model as outputs. Finally, a comparative analysis is conducted between the prediction model and finite element simulation data, demonstrating that the average relative error does not exceed 3.16%. This model provides support for research on the optimization of hobbing process parameters and tool wear.