Scholarlink

<p>Robotic milling of ultra-high-molecular-weight polyethylene (UHMWPE) offers flexibility for machining complex geometries, but reliable process parameter selection remains challenging because conventional graphical analysis may misidentify dominant factors. This study employs a face-centred central composite design with response surface methodology (FCCD-RSM) to evaluate the effects of spindle speed, feed rate, and depth of cut on surface roughness (Ra, Rz) and machining time in robotic milling of UHMWPE using a KUKA KR120 R2700 industrial robot. Parameter rankings derived from main effect plots and analysis of variance (ANOVA) are consistent for Ra but diverge for Rz and machining time, indicating that graphical interpretation may misrepresent parameter importance in compliance-sensitive robotic systems. This discrepancy arises because main effect plots reflect mean response magnitude, whereas ANOVA quantifies statistical contribution relative to residual variability. Multi-response optimisation using a desirability function, with Ra minimised and material removal rate (MRR) maximised, identified an optimal parameter combination of spindle speed 12500&#xa0;rpm, feed rate 1260&#xa0;mm/min, and depth of cut 0.34&#xa0;mm, achieving Ra = 0.940&#xa0;μm and MRR = 3519&#xa0;mm³/min with a composite desirability of 0.857. Experimental validation confirmed strong agreement between predicted and measured responses, with prediction errors below 5%. The results demonstrate that ANOVA-informed parameter prioritisation, integrated with RSM-based optimisation and experimental validation, provides a more reliable framework for process decision-making in robotic milling of UHMWPE than graphical analysis alone.</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

Improving parameter selection reliability in robotic milling of UHMWPE using multi-response RSM optimisation and statistical validation

  • 期刊论文

  • 发表日期：

    2026年6月30日

查看全文
  • Wan Nor Shela Ezwane Wan Jusoh,
  • Mohamad Irwan Yahaya,
  • Mahamad Hisyam Mahamad Basri,
  • Shukri Zakaria

摘要

Robotic milling of ultra-high-molecular-weight polyethylene (UHMWPE) offers flexibility for machining complex geometries, but reliable process parameter selection remains challenging because conventional graphical analysis may misidentify dominant factors. This study employs a face-centred central composite design with response surface methodology (FCCD-RSM) to evaluate the effects of spindle speed, feed rate, and depth of cut on surface roughness (Ra, Rz) and machining time in robotic milling of UHMWPE using a KUKA KR120 R2700 industrial robot. Parameter rankings derived from main effect plots and analysis of variance (ANOVA) are consistent for Ra but diverge for Rz and machining time, indicating that graphical interpretation may misrepresent parameter importance in compliance-sensitive robotic systems. This discrepancy arises because main effect plots reflect mean response magnitude, whereas ANOVA quantifies statistical contribution relative to residual variability. Multi-response optimisation using a desirability function, with Ra minimised and material removal rate (MRR) maximised, identified an optimal parameter combination of spindle speed 12500 rpm, feed rate 1260 mm/min, and depth of cut 0.34 mm, achieving Ra = 0.940 μm and MRR = 3519 mm³/min with a composite desirability of 0.857. Experimental validation confirmed strong agreement between predicted and measured responses, with prediction errors below 5%. The results demonstrate that ANOVA-informed parameter prioritisation, integrated with RSM-based optimisation and experimental validation, provides a more reliable framework for process decision-making in robotic milling of UHMWPE than graphical analysis alone.