Scholarlink

<p>The Silicon carbide (SiC) particles with different sizes (30, 60, and 90&#xa0;μm) reinforced Al 1200 alloy composites offer enhanced properties (hardness, density, and strength) suitable for engineering applications. The finer (30&#xa0;μm) SiC particles exhibit better physical and mechanical properties, as validated by microstructural analysis. SiC particles pose machining difficulties, including cutting resistance, increased surface roughness (SR), and tool wear. Optimizing the machining parameters (reinforcement size: RS, cutting speed: CS, feed rate: FR, and depth of cut: DOC) to improve machining quality characteristics (material removal rate: MRR, SR, and chip thickness ratio: CTR) is of industrial relevance. Box-Behnken Design (BBD) analysis confirms that all individual factors are statistically significant (FR &gt; CS &gt; DOC &gt; RS) and exhibit a nonlinear effect on MRR, SR, and CTR. The significance of cutting variables (CS, DOC, and FR) interactions confirms their primary importance, rather than microstructure variation with RS. The highest coefficient of determination, 0.9896, 0.9603, and 0.9817, are established for CTR-MRR, SR-CTR, and SR-MRR, confirming their strongly dependent nonlinear relationships. Artificial neural networks (ANNs) reduce prediction error to 4.01%, compared with 4.7% for the BBD model, across three responses across ten random experimental cases, resulting in a 14.7% improvement in prediction accuracy. Grey Wolf Optimizer (GWO) integrated desirability function approach (GWO-DFA) recommends optimized conditions, confirming higher MRR (401.4 mm<sup>3</sup>/s), lower SR of 3.32&#xa0;μm (validated by 3D and 2D surface profiles), and a nominal CTR of 0.678, with a reduced prediction error of less than 5%. The proposed systematic framework ensures industry-ready model deployment for widespread engineering applications.</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

Multiobjective optimization of milling performance characteristics of Al 1200 SiC metal matrix composites using statistical and artificial intelligence tools

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年3月13日

查看全文
  • Pavitra Ajagol,
  • Rajaneesh N. Marigoudar,
  • Anjan B. N,
  • Manjunath Patel G. C.,
  • Manjunath Shettar,
  • Avinash Lakshmikanthan

摘要

The Silicon carbide (SiC) particles with different sizes (30, 60, and 90 μm) reinforced Al 1200 alloy composites offer enhanced properties (hardness, density, and strength) suitable for engineering applications. The finer (30 μm) SiC particles exhibit better physical and mechanical properties, as validated by microstructural analysis. SiC particles pose machining difficulties, including cutting resistance, increased surface roughness (SR), and tool wear. Optimizing the machining parameters (reinforcement size: RS, cutting speed: CS, feed rate: FR, and depth of cut: DOC) to improve machining quality characteristics (material removal rate: MRR, SR, and chip thickness ratio: CTR) is of industrial relevance. Box-Behnken Design (BBD) analysis confirms that all individual factors are statistically significant (FR > CS > DOC > RS) and exhibit a nonlinear effect on MRR, SR, and CTR. The significance of cutting variables (CS, DOC, and FR) interactions confirms their primary importance, rather than microstructure variation with RS. The highest coefficient of determination, 0.9896, 0.9603, and 0.9817, are established for CTR-MRR, SR-CTR, and SR-MRR, confirming their strongly dependent nonlinear relationships. Artificial neural networks (ANNs) reduce prediction error to 4.01%, compared with 4.7% for the BBD model, across three responses across ten random experimental cases, resulting in a 14.7% improvement in prediction accuracy. Grey Wolf Optimizer (GWO) integrated desirability function approach (GWO-DFA) recommends optimized conditions, confirming higher MRR (401.4 mm3/s), lower SR of 3.32 μm (validated by 3D and 2D surface profiles), and a nominal CTR of 0.678, with a reduced prediction error of less than 5%. The proposed systematic framework ensures industry-ready model deployment for widespread engineering applications.