Lie symmetry based neural networking analysis for Powell–Eyring fluid flow with heat and mass transfer effects

Abstract

Heat exchangers, drying, dehydration, pollutant dispersion, separation processes, and biological individualities to name a few are significant phenomena in numerous fields and industries subject to both mass transfer and heat transfer aspects. Owing to such importance the aim of the present study is to offer a machine learning remedy for mass transfer rate at the surface subject to the Powell–Eyring (PE) fluid model. The flow field is rooted with suction and injection effects. The heat transfer aspects are considered by using the energy equation. The developed differential system is reduced by using Lie symmetry and the shooting scheme is used to get numerical data corresponding to four inputs namely suction parameter, PE fluid parameter, Schmidt number, and power law index. Levenberg–Marquardt backpropagation-based neural networking model is developed along with MSE and regression analysis to approximate the ShD number. It is observed that the mass transfer rate shows inciting values towards the suction parameter, Schmidt number, and concentration power law index. Further, the magnitude of Powell–Eyring fluid concentration is higher in the case of injection.