Scholarlink

In the present study, natural convection heat transfer is investigated numerically in a pseudo-plastic fluid filling a square cavity’s four walls are cooled isothermally, and it contains a circular isothermal heat source. The Finite Difference-Lattice Boltzmann Method (FD-LBM) is employed as the computational technique to simulate the effects of controlling physical parameters, specifically the Hartmann number ( \(0\le Ha\le 50\) ) and the CMC concentration ( \(0\le \xi \le 0.4\%\) ), on the thermal and dynamic characteristics of the working fluid. This analysis is conducted for a fixed Rayleigh number of \(5\times {10}^{5}\) . . The numerical results obtained are utilized to train, test, and validate an artificial neural network (ANN) model. The findings indicate that the dynamic and thermal behaviors of the pseudo-plastic fluid are significantly influenced by variations in the physical parameters within their specified ranges. The increase in \(Ha\) /( \(\xi \) ) negatively/(positively) impacts the fluid flow intensity and heat transfer within the cavity. Furthermore, the effect of the magnetic field on the pseudo-plastic fluid is more pronounced than that on the Newtonian fluid. A correlation for the mean Nusselt number has been proposed, and the heat transfer rates predicted by the ANN are found to be in very good agreement with those simulated by the FD-LBM.

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

Optimizing Natural Convection Heat Transfer of Pseudo-Plastic Fluids in Magnetic Fields Using Artificial Neural Networks

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Khalid Chtaibi,
  • Mohammed Hasnaoui,
  • Abdelkhalek Amahmid,
  • Haïkel Ben Hamed,
  • Youssef Dahani,
  • Abdelghani Raji,
  • Safae Hasnaoui,
  • Abdelfattah El Mansouri

摘要

In the present study, natural convection heat transfer is investigated numerically in a pseudo-plastic fluid filling a square cavity’s four walls are cooled isothermally, and it contains a circular isothermal heat source. The Finite Difference-Lattice Boltzmann Method (FD-LBM) is employed as the computational technique to simulate the effects of controlling physical parameters, specifically the Hartmann number ( \(0\le Ha\le 50\) ) and the CMC concentration ( \(0\le \xi \le 0.4\%\) ), on the thermal and dynamic characteristics of the working fluid. This analysis is conducted for a fixed Rayleigh number of \(5\times {10}^{5}\) . . The numerical results obtained are utilized to train, test, and validate an artificial neural network (ANN) model. The findings indicate that the dynamic and thermal behaviors of the pseudo-plastic fluid are significantly influenced by variations in the physical parameters within their specified ranges. The increase in \(Ha\) /( \(\xi \) ) negatively/(positively) impacts the fluid flow intensity and heat transfer within the cavity. Furthermore, the effect of the magnetic field on the pseudo-plastic fluid is more pronounced than that on the Newtonian fluid. A correlation for the mean Nusselt number has been proposed, and the heat transfer rates predicted by the ANN are found to be in very good agreement with those simulated by the FD-LBM.