Use of an ANN–Levenberg–Marquardt optimization for radiative hybrid nanofluids and heat transfer efficiency over non-isothermal wedge and cone surfaces with chemical reaction effects

Abstract

The advancement of thermal engineering technologies is driven by the increasing requirement for efficient energy utilization and enhanced heat transfer performance in various industrial sectors. The present study emphasizes the systematic examination of hybrid nanofluid magnetohydrodynamic flow across the geometries of two different surfaces—a cone and a wedge—within a porous medium while considering thermal radiation and chemical reactions. This study is governed by PDEs, which are consequently transformed into ODEs through suitable similarity transformations. Mathematical solutions are derived utilizing the bvp5c method within a MATLAB environment. The upsurge in thermal radiation and Brownian motion parameter enhances the thermal profile of the hybrid nanofluid. The mass transfer rate of the hybrid nanofluid decreases with respect to the Schmidt number, and the chemical reaction enhances for thermophoresis. An artificial neural network based on the Levenberg–Marquardt backpropagation scheme (ANN-LMBPS) was employed, and its accuracy was validated against the bvp5c results for the effects of magnetism, porosity, thermophoresis, Brownian motion, and chemical reaction.