Designing a solid–fluid interface layer and artificial neural network in a nanofluid flow due to rotating rough and porous disk

The present research analyzes the impact of nanoparticle diameter and the interfacial layer on the nanofluid flow over a rough rotating disk with melting. Additionally, homogeneous and heterogeneous reactions play an essential part in comprehending the dynamics of mass transfer. Appropriate similarity variables are utilized to convert nonlinear governing equations into ordinary differential equations. The reduced equations are solved numerically by using a shooting approach and Runge–Kutta–Fehlberg fourth-fifth (RKF-45)-order method. In addition, an advanced intelligent numerical computing solver that interprets heat transfer and surface drag force is offered. This solution uses artificial neural networks with multilayer perceptron, feed-forward, back-propagation, and the Levenberg–Marquardt method. The plotted histograms display the error distribution for each of these predicted values from a zero-error point. More values that are close to the zero-error line will be present in a solution method that is more exact and precise. The results reveal that the radial velocity profile’s oscillatory behavior is shown to diminish close to the disk as the viscous force rises with higher slip parameter values. The axial component of velocity decreases as the slip parameter upsurges, which is to be expected as less fluid is radially released. The increase in melting parameter diminishes the temperature profile.