Heat and Mass Transfer Analysis for Unsteady Three-Dimensional Flow of Hybrid Nanofluid Over a Stretching Surface Using Supervised Neural Networks

Abstract

The application of hybrid nanomaterials for the improvement of thermal efficiency of base fluid has increasingly gained attention during the past few decades. The basic purpose of this study is to investigate the flow characteristics along with heat transfer in an unsteady three-dimensional flow of hybrid nanofluid over a stretchable and rotatory sheet (3D-UHSRS). The flow model in the form of PDEs was reduced to the set of ordinary differential equations utilizing the appropriate transformations of similarity. The influence of the rotation parameter, unsteadiness parameter, stretching parameter, radiation parameter, and Prandtl number on velocities and thermal profile was graphically examined. A reference solution in the form of dataset points for the 3D-UHSRS model are computed with the help of renowned Lobatto IIIA solver, and this solution is exported to MATLAB for the proper implementation of proposed solution methodology based on the Levenberg–Marquardt supervised neural networks. Graphical and numerical results based on the mean square error (MSEs), time series response, error distribution plots, and regression plots endorses the precision, validity, and consistency of the proposed solution methodology. The MSE up to the level of 10–12 confirms the accuracy of the achieved results.