Scholarlink

<p>This analysis investigates the magnetohydrodynamic (MHD) boundary layer flow and heat transfer characteristics of single-walled carbon nanotubes (SWCNTs) dispersed in Maxwellian and non-Maxwellian nanofluids with water between two disks. Numerical methods incorporated thermal radiation effects and convective boundary conditions, and similarity transformations were employed for analysis. Key findings reveal that the Nusselt number, skin friction coefficient, and Sherwood number decrease with increasing material fraction, while the Deborah number significantly impacts velocity and temperature profiles. The integration of the Finite Element Method (FEM) with neural networks, specifically the Levenberg–Marquardt supervised method, ensured precise and efficient computations. The absolute error between numerical and predicted results remained minimal, validating the accuracy and reliability of the approach. These results enhance the understanding of thermal radiation effects in nanofluids and demonstrate the potential of machine learning in modeling complex fluid systems. Applications include advanced cooling systems, heat exchangers, and MHD aerospace technologies, offering significant advancements in thermal management solutions.</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

Neural network-assisted analysis of MHD boundary layer flow and thermal radiation effects on SWCNT nanofluids with Maxwellian and non-Maxwellian models

  • 期刊论文

  • 发表日期：

    2025年1月31日

查看全文
  • K. Jyothi,
  • Annareddy Sailakumari,
  • Ramachandra Reddy Vaddemani,
  • Raghunath Kodi

摘要

This analysis investigates the magnetohydrodynamic (MHD) boundary layer flow and heat transfer characteristics of single-walled carbon nanotubes (SWCNTs) dispersed in Maxwellian and non-Maxwellian nanofluids with water between two disks. Numerical methods incorporated thermal radiation effects and convective boundary conditions, and similarity transformations were employed for analysis. Key findings reveal that the Nusselt number, skin friction coefficient, and Sherwood number decrease with increasing material fraction, while the Deborah number significantly impacts velocity and temperature profiles. The integration of the Finite Element Method (FEM) with neural networks, specifically the Levenberg–Marquardt supervised method, ensured precise and efficient computations. The absolute error between numerical and predicted results remained minimal, validating the accuracy and reliability of the approach. These results enhance the understanding of thermal radiation effects in nanofluids and demonstrate the potential of machine learning in modeling complex fluid systems. Applications include advanced cooling systems, heat exchangers, and MHD aerospace technologies, offering significant advancements in thermal management solutions.