LTNE effect on non-linear radiative MHD mixed convective flow in an annular porous medium: Intelligent Computing Paradigm

Abstract

In this study, the back-propagated Levenberg-Marquardt algorithm-based neural network intelligent computing paradigm is used to examine the influence of non-linear thermal radiation, induced magnetic field and local thermal non-equilibrium effect on the magnetohydrodynamic mixed convective flow in the vertical annular porous region. In the annular porous region the inner cylinder is considered as electrically conducting and the outer cylinder as electrically non-conducting. The heat transfer is expected due to mixed convection and non-linear thermal radiation. The Galerkin finite element method is adopted to solve the governing equations. In the vertical annulus an increase in the magnetic Prandtl and Hartmann number reduces the velocity and induced magnetic field due to a strong external magnetic field and the Lorentz force. The higher Brinkmann values decelerate the conduction of heat generated by viscous dissipation hence, the temperature increases. The fluid and solid temperature increase with the buoyancy parameter. The non-linear radiation term enhances the fluid temperature, whereas it decelerates the solid temperature. The ANN model utilises finite element method output for training, validating and testing to predict the heat transfer characteristics. The performance of the BLM-NN algorithm is assessed using mean square error, error histogram and regression.