Scholarlink

<p>This study investigates the magnetohydrodynamic (MHD) flow of a viscous, electrically conducting fluid over a cylinder undergoing hyperbolic stretching. The effects of a magnetic field, porous medium, thermal radiation, and Joule heating are considered. The governing nonlinear coupled partial differential equations are transformed using similarity variables, simplified via the local non-similarity method, and numerically solved using the Runge–Kutta–Fehlberg 4–5th order technique. Additionally, an Artificial Neural Network (ANN) model is developed to predict key thermal and flow parameters, including skin friction and the Nusselt number. The ANN results closely align with the numerical solutions, demonstrating its reliability as an efficient alternative to traditional numerical methods. The concept of entropy generation is incorporated to analyze the effects of Joule heating, viscous dissipation, and heat transfer on system irreversibilities. The results indicate that entropy production is strongly influenced by thermophysical parameters, significantly affecting system efficiency. Furthermore, a comparative study with existing literature shows good agreement, validating the accuracy of the present approach. Quantitatively, as the non-dimensional axial coordinate (x) increases from 0.5 to 3 (i.e., by 500%), the magnitude of skin friction increases by approximately 82.6%, while the Nusselt number enhances by about 187.3%, indicating a substantial rise in both shear stress and heat transfer rate along the stretching cylinder. Overall, the findings highlight the efficiency of the ANN in optimizing computational performance with high accuracy. The present study provides valuable insights into entropy generation and heat transfer in MHD flows over stretching surfaces, with applications in industrial heat exchangers, energy systems, and electronic cooling. Future studies may extend this work by incorporating temperature-dependent viscosity, Hall and ion slip effects, and AI-driven predictive models for real-time thermal system optimization.</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

Nonlinear thermal analysis of magnetohydrodynamic boundary layer flow over a hyperbolic stretching cylinder with thermal radiation using Levenberg–Marquardt backpropagation neural networks

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年7月9日

查看全文
  • M. N. Pooja,
  • Gunisetty Ramasekhar,
  • V. Anitha,
  • S. K. Narasimhamurthy,
  • Abbaireddy Divya

摘要

This study investigates the magnetohydrodynamic (MHD) flow of a viscous, electrically conducting fluid over a cylinder undergoing hyperbolic stretching. The effects of a magnetic field, porous medium, thermal radiation, and Joule heating are considered. The governing nonlinear coupled partial differential equations are transformed using similarity variables, simplified via the local non-similarity method, and numerically solved using the Runge–Kutta–Fehlberg 4–5th order technique. Additionally, an Artificial Neural Network (ANN) model is developed to predict key thermal and flow parameters, including skin friction and the Nusselt number. The ANN results closely align with the numerical solutions, demonstrating its reliability as an efficient alternative to traditional numerical methods. The concept of entropy generation is incorporated to analyze the effects of Joule heating, viscous dissipation, and heat transfer on system irreversibilities. The results indicate that entropy production is strongly influenced by thermophysical parameters, significantly affecting system efficiency. Furthermore, a comparative study with existing literature shows good agreement, validating the accuracy of the present approach. Quantitatively, as the non-dimensional axial coordinate (x) increases from 0.5 to 3 (i.e., by 500%), the magnitude of skin friction increases by approximately 82.6%, while the Nusselt number enhances by about 187.3%, indicating a substantial rise in both shear stress and heat transfer rate along the stretching cylinder. Overall, the findings highlight the efficiency of the ANN in optimizing computational performance with high accuracy. The present study provides valuable insights into entropy generation and heat transfer in MHD flows over stretching surfaces, with applications in industrial heat exchangers, energy systems, and electronic cooling. Future studies may extend this work by incorporating temperature-dependent viscosity, Hall and ion slip effects, and AI-driven predictive models for real-time thermal system optimization.