Scholarlink

<p>To precisely assess the dynamic milling force during the process of thin-wall complex-surface machining, a new cutting force model was established by taking into account both the effects of curved toolpath and tool wear. To acquire the experimental data for the modelling, an orthogonal experiment was conducted during which the cutting force signal was recorded, and the characteristic value of tool wear state was obtained by machine vision. Subsequently, the toolpath was partitioned into a series of subregions and original force signals were sub-processed to derive corresponding milling force values. Finally, a BP neural network was utilized to construct a dynamic milling force model that relates to positional region sequences, cutting depths, spindle speeds, cutting feed rates, and tool wear state. To validate the accuracy of the model, a comparative experiment was conducted. The results demonstrated that the absolute prediction error was less than 7 N, confirming the model’s reliability.</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

Cutting force modeling method for milling process of thin-wall complex-surface workpieces considering effects of tool wear and varying-curvature feed path

  • 期刊论文

  • 发表日期：

    2026年1月15日

查看全文
  • Zheng Zou,
  • Liang He,
  • Hao Zhang,
  • Benyuan Fu,
  • Ruizhi Shu,
  • Hui Jin,
  • Yong Yang

摘要

To precisely assess the dynamic milling force during the process of thin-wall complex-surface machining, a new cutting force model was established by taking into account both the effects of curved toolpath and tool wear. To acquire the experimental data for the modelling, an orthogonal experiment was conducted during which the cutting force signal was recorded, and the characteristic value of tool wear state was obtained by machine vision. Subsequently, the toolpath was partitioned into a series of subregions and original force signals were sub-processed to derive corresponding milling force values. Finally, a BP neural network was utilized to construct a dynamic milling force model that relates to positional region sequences, cutting depths, spindle speeds, cutting feed rates, and tool wear state. To validate the accuracy of the model, a comparative experiment was conducted. The results demonstrated that the absolute prediction error was less than 7 N, confirming the model’s reliability.