Dynamics of magnetized hybrid solid particles in a rotating cone with gradient descent optimization

Abstract

Background Recent developments in nanoscience and nanotechnology have led to innovative hybrid nanofluids with superior thermal attributes to traditional nanofluids. These developments act as inspiration for the current framework, which focuses on an unsteady Casson hybrid nanofluid flow over a rotating cone in the presence of nonuniform heat source/sink factor. Methodology Two distinct shaped hybrid nanoparticles-hosted water is emphasized in the present work, namely spherical shaped CoF e 2 O 4 and cylindrical shaped Cu . The contribution of the heat source/sink variable is incorporated into the regulating equations utilizing boundary layer approximations. The resulting expressions of dimensionless forms of transmuted ODEs are computed through an NDSolve scheme. The performance of the pertinent parameters is accentuated to exhibit the relevance of the considered flow model. Furthermore, Multiple Linear Regression is introduced to generate appropriate formulations for the prediction of physical quantities on nanofluid and hybrid nanofluid. Core findings The volume fraction particles have a propensity to boost the temperature distribution, while the azimuthal and tangential factor follows the opposite trend in both the nanofluid and hybrid nanofluid. The skin friction coefficients rise for unsteady parameters, but the Nusselt number gets diminished due to the unsteadiness. The heat transfer rate of CoF e 2 O 4 nanoparticles hosted H 2 O is increased in Rudyak model than the Brinkman viscosity model by the involvement of an effective viscosity coefficient in terms of volume concentration of dispersed particles in nanofluid. Validation The outcomes are validated by comparing the present exploration with previously published investigations in certain instances. The multiple linear regression accurately anticipated the physical quantities with minimal error 10 − 2 . Applications An implementation of machine learning as an approach to do thermal evaluation with reliability and precision has emerged as a growing trend. The results of this investigation will benefit plenty of heat transfer applications, such as heat exchangers, heat pipes, and micro-channels.