Comparative analysis of MHD HFE-7100, HFE-7200 and HFE-7500 based nanofluids with Soret and Dufour effects using artificial neural network

Abstract

Nimonic 80A/HFE-7100, HFE-7300, and HFE-7500 nanofluids with Marangoni convection, Joule heating, and thermal-solutal gradients are compared using the Xue thermal conductivity model and the computational capability of supervised neural networks with Levenberg Marquardt algorithm (SNN-LMA). It is primarily used in high-temperature thermal management and advanced energy systems. These nanofluids are useful for cooling gas turbine blades, aerospace components, and nuclear reactor parts, with Nimonic 80A providing mechanical strength and HFE-based nanofluids improving heat transfer. The model is useful for improving surface-tension-driven flows in microchannels, electronic cooling systems, and electromagnetic processing units where Joule heating and double-diffusive convection are important. It is particularly beneficial in material processing, thin-film coating, and metallurgical procedures, allowing for more precise control of heat and mass transport under coupled thermal and solutal effects. The thermo-solutal Marangoni-driven Darcy-Forchheimer flow of MHD nanofluids across a sheet is investigated in this study. The model includes the properties of Joule heating, viscous dissipation, thermal radiation, and Soret-Dufour influence. By relating shear stress to temperature and concentration gradients, the Marangoni boundary condition is supposed at the surface. Using HAM (Homotopy Analysis Method), the model is analytically solved after the reduction of PDEs to a system of coupled nonlinear ODEs through suitable similarity transformation.