Scholarlink

<p>This study aims to revolutionize the prediction and optimization of rolling parameters in metal strip rolling processes by addressing the critical issue of edge cracks. The research focuses on enhancing product quality while promoting sustainable manufacturing through reduced material waste and improved resource efficiency. A hybrid framework integrating Finite Element Analysis (FEA), Artificial Neural Networks (ANN), and Genetic Algorithms (GA) was developed. Detailed simulations were performed using ANSYS and ABAQUS to analyze rolling dynamics and edge crack behavior. Critical parameters, including roller speed, rolling stages, and slip factor, were optimized using ANN–GA modeling based on FEA results. Validation was carried out through microstructural analysis, mechanical testing, and stored energy evaluation. The optimized parameters reduced the crack area by 11.3%, improved product quality with tensile strength increasing from 253&#xa0;MPa (as-received) to 325&#xa0;MPa (rolled), and enhanced resource efficiency with a 9.5% reduction in material waste. The ANN (LM-Levenberg–Marquardt &amp; BR-Bayesian Regularization)–GA hybrid model showed excellent agreement with FEA results, with an error margin of &lt; 2% in crack area and &lt; 5% in crack angle, confirming the reliability of the hybrid framework. Overall, the integration of FEA, ANN, and GA provided more accurate predictions than conventional trial-and-error methods. The model’s applicability is currently limited to specific materials and controlled industrial settings. Broader validation under varied materials and rolling conditions is required. Optimized parameters derived from the hybrid model can reduce production defects, minimize material loss, and improve sustainability in industrial rolling operations. This study introduces a novel hybrid FEA–ANN–GA framework for rolling process optimization. The approach offers a predictive and optimization tool that advances sustainable manufacturing and improves industrial productivity.</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

Enhanced Prediction and Optimization of Metal Strip Rolling Parameters Using Hybrid FEA-ANN-GA Models to Minimize Edge Cracks

  • 期刊论文

  • 发表日期：

    2025年10月25日

查看全文
  • S. P. Sundar Singh Sivam,
  • V. G. Umasekar,
  • Stalin Kesavan,
  • J. Rajprasad

摘要

This study aims to revolutionize the prediction and optimization of rolling parameters in metal strip rolling processes by addressing the critical issue of edge cracks. The research focuses on enhancing product quality while promoting sustainable manufacturing through reduced material waste and improved resource efficiency. A hybrid framework integrating Finite Element Analysis (FEA), Artificial Neural Networks (ANN), and Genetic Algorithms (GA) was developed. Detailed simulations were performed using ANSYS and ABAQUS to analyze rolling dynamics and edge crack behavior. Critical parameters, including roller speed, rolling stages, and slip factor, were optimized using ANN–GA modeling based on FEA results. Validation was carried out through microstructural analysis, mechanical testing, and stored energy evaluation. The optimized parameters reduced the crack area by 11.3%, improved product quality with tensile strength increasing from 253 MPa (as-received) to 325 MPa (rolled), and enhanced resource efficiency with a 9.5% reduction in material waste. The ANN (LM-Levenberg–Marquardt & BR-Bayesian Regularization)–GA hybrid model showed excellent agreement with FEA results, with an error margin of < 2% in crack area and < 5% in crack angle, confirming the reliability of the hybrid framework. Overall, the integration of FEA, ANN, and GA provided more accurate predictions than conventional trial-and-error methods. The model’s applicability is currently limited to specific materials and controlled industrial settings. Broader validation under varied materials and rolling conditions is required. Optimized parameters derived from the hybrid model can reduce production defects, minimize material loss, and improve sustainability in industrial rolling operations. This study introduces a novel hybrid FEA–ANN–GA framework for rolling process optimization. The approach offers a predictive and optimization tool that advances sustainable manufacturing and improves industrial productivity.