Hybrid numerical and ANN modeling of unsteady MHD HMT over a rotating porous plate

This study presents a numerical investigation of unsteady magnetohydrodynamic (MHD) flow and heat and mass transfer (HMT) in a rotating, electrically conducting fluid over a heated vertical porous plate. The effects of key physical parameters—including permeability (N), Soret number (S0), Schmidt number (Sc), Prandtl number (Pr), thermal radiation (R), and heat source (Q)—are analyzed through graphical representations of velocity, temperature, and concentration profiles. The results reveal that thermal diffusion (via higher Soret numbers) enhances velocity, while increased permeability and Schmidt number tend to suppress it. Temperature profiles are elevated by thermal radiation but decline with higher values of Pr and Q. Similarly, concentration profiles decrease with increasing Sc and R, although Soret effects improve mass transfer rates. The influence of magnetic fields and thermal gradients on HMT is quantified using Nusselt (Nu) and Sherwood (Sh) numbers, where Sh increases with the Soret number and Nu diminishes as the plate radius increases. In addition, an artificial neural network model was employed to validate the numerical predictions, achieving an outstanding prediction accuracy of 99.99%. The findings provide valuable insights into the coupled thermal and solutal transport mechanisms in MHD systems and propose a reliable computational framework for future research and engineering applications involving complex flow phenomena.