Scholarlink

<p>The internal structure of a turn-mill machine is highly complex, comprising numerous components that generate considerable heat during operation. This heat causes localized temperature rises, leading to thermal deformation across the machine structure and significantly affecting the accuracy of the Tool Center Point (TCP). Accurately identifying thermally sensitive regions that are highly correlated with the TCP is therefore essential for developing targeted thermal compensation algorithms and sensor deployment strategies. Such efforts enhance machining precision and shorten the trial production cycle. To address this challenge, this study proposes a systematic method that integrates Silhouette score evaluation, hierarchical clustering, and Spearman correlation analysis to effectively segment temperature regions and assess their thermal impact. By analyzing correlations between regional temperatures and the spindle and sub-spindle TCP under various operating conditions, 17 thermally critical regions were successfully identified. Through this combined workflow, we deliver, in advance, a comprehensive evaluation of each region’s thermal impact on the TCP across diverse machining conditions, accurately pinpointing thermal critical areas. This method is particularly suited to the early development stage of a machine tool: by embedding sensors at optimal locations prior to factory shipment, it ensures monitoring coverage even in hard-to-reach or non-serviceable areas, thereby potentially improving further thermal error prediction accuracy.</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

A simulation-based clustering framework for identifying thermal critical areas in turn-mill machines

  • 期刊论文

  • 发表日期：

    2025年9月17日

查看全文
  • Woei-Tyng Chang,
  • Chien-Hsiang Hung,
  • Yung-Chuan Chang,
  • Jyun-Rong Zhuang

摘要

The internal structure of a turn-mill machine is highly complex, comprising numerous components that generate considerable heat during operation. This heat causes localized temperature rises, leading to thermal deformation across the machine structure and significantly affecting the accuracy of the Tool Center Point (TCP). Accurately identifying thermally sensitive regions that are highly correlated with the TCP is therefore essential for developing targeted thermal compensation algorithms and sensor deployment strategies. Such efforts enhance machining precision and shorten the trial production cycle. To address this challenge, this study proposes a systematic method that integrates Silhouette score evaluation, hierarchical clustering, and Spearman correlation analysis to effectively segment temperature regions and assess their thermal impact. By analyzing correlations between regional temperatures and the spindle and sub-spindle TCP under various operating conditions, 17 thermally critical regions were successfully identified. Through this combined workflow, we deliver, in advance, a comprehensive evaluation of each region’s thermal impact on the TCP across diverse machining conditions, accurately pinpointing thermal critical areas. This method is particularly suited to the early development stage of a machine tool: by embedding sensors at optimal locations prior to factory shipment, it ensures monitoring coverage even in hard-to-reach or non-serviceable areas, thereby potentially improving further thermal error prediction accuracy.