Scholarlink

<p>The study focuses on modeling and optimization of tungsten inert gas (TIG) cladding parameters to develop Al-10TiC-10SiC composite coatings on Al 6061 alloy. A full factorial design was employed for evaluating the influence of current and scanning speed on microhardness and wear characteristics. Regression analysis and artificial neural network (ANN) were utilized for predictive modeling, with ANN exhibiting greater predictive accuracy compared to regression, as evidenced by error metrics and correlation coefficients. A multi-objective genetic algorithm (MOGA) was then employed to derive the Pareto front comprising 105 non-dominated solutions representing trade-offs between maximizing microhardness and minimizing wear. Multi-criteria decision-making (MCDM) methods such as MOORA, TOPSIS, and COPRAS have been implemented on the Pareto solutions to find the most suitable compromise solution. The predicted optimal parameter range (current: 85.13-85.44 A; scanning speed: 3.33-3.34&#xa0;mm/s) was adjusted to practical machine settings (85 A and 3.50&#xa0;mm/s), and subsequently verified experimentally. The experimental results at optimum condition exhibited minimum deviations from predictions, with 2.79% for microhardness and 2.59% for wear. Surface characterization confirmed that optimum parameters produced improved microstructural bonding, reduced wear grooves, and enhanced resistance to abrasion. The outcomes reveal that the hybrid method of ANN-MOGA-MCDM provides a robust and dependable framework for optimizing TIG cladding process parameter to achieve superior mechanical and tribological performance.</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

Hybrid Modeling and Optimization of TIG Cladding Parameters for Al-10TiC-10SiC Coatings on Al 6061 Alloy

  • 期刊论文

  • 发表日期：

    2026年6月2日

查看全文
  • Pritam Paul,
  • Subrata Kumar Ghosh,
  • Siddhartha Kar,
  • Pabitra Maji,
  • R. K. Bhogendro Meitei,
  • Sujoy Chakraborty

摘要

The study focuses on modeling and optimization of tungsten inert gas (TIG) cladding parameters to develop Al-10TiC-10SiC composite coatings on Al 6061 alloy. A full factorial design was employed for evaluating the influence of current and scanning speed on microhardness and wear characteristics. Regression analysis and artificial neural network (ANN) were utilized for predictive modeling, with ANN exhibiting greater predictive accuracy compared to regression, as evidenced by error metrics and correlation coefficients. A multi-objective genetic algorithm (MOGA) was then employed to derive the Pareto front comprising 105 non-dominated solutions representing trade-offs between maximizing microhardness and minimizing wear. Multi-criteria decision-making (MCDM) methods such as MOORA, TOPSIS, and COPRAS have been implemented on the Pareto solutions to find the most suitable compromise solution. The predicted optimal parameter range (current: 85.13-85.44 A; scanning speed: 3.33-3.34 mm/s) was adjusted to practical machine settings (85 A and 3.50 mm/s), and subsequently verified experimentally. The experimental results at optimum condition exhibited minimum deviations from predictions, with 2.79% for microhardness and 2.59% for wear. Surface characterization confirmed that optimum parameters produced improved microstructural bonding, reduced wear grooves, and enhanced resistance to abrasion. The outcomes reveal that the hybrid method of ANN-MOGA-MCDM provides a robust and dependable framework for optimizing TIG cladding process parameter to achieve superior mechanical and tribological performance.