Machine learning thermal analysis of Boger C3H8O2 hybrid nanofluid for industrial heat transmission applications: Cattaneo-Christov flux model

Abstract

This work uses the Levenberg-Marquardt with neural networks (LMNN) approach to investigate the impact of the Cattaneo-Christov flux model on the surface tension gradient flow of Boger hybrid nanofluid over the Riga plate with heat production. This concept is extremely important for modern industrial heat transfer systems that require increased thermal efficiency and exact control over mass and heat transfer. The machine learning-based analysis of Boger propylene glycol-based hybrid nanofluid employing the surface tension gradient and Cattaneo-Christov flux model can be used in coating and thin-film processes, polymer and chemical manufacturing, electronic and microelectromechanical system cooling, and materials processing industries. By taking into consideration non-Fourier heat flux and surface-tension-driven fluxes, the model aids in the optimization of process parameters, reduction of thermal losses, and improvement of product quality in advanced thermal management and energy-efficient industrial applications. The suggested model has been assessed for excellence after the estimate solution of multiple scenarios has been verified utilizing the training, testing, and validation procedure of LMNN. The proposed LMNN is then validated using regression analysis, mean square error, and histogram explorations. The velocity profile increases with an increase in the solvent friction parameter, while the thermal profile decreases.