Deep neural network-based prediction and computational fluid dynamics analysis of convective heat transfer in dusty fluid flow over heated surface

Abstract

Due to the necessity of creating customized nanomaterial designs, advanced coating processes have been developed on the surface of an isothermal sphere. To achieve accuracy in these processes, it is necessary to understand thermodynamic behavior, the rheology of materials, and the chemical reactions involved. The novelty of the present research is analyzing heat transfer in dusty fluid flow over a heated surface in the cross section of the cylinder. We develop a mathematical model including these factors for both phases and demonstrate how these factors affect the flow field. Thermal radiation propagation is simulated using Rosseland's flux diffusion model. Equations describing flow transport and magnetic body force are expressed using a Cartesian coordinate system. A set of boundary conditions has been established for momentum, thermal energy, and concentration particles to formulate conservation equations for the surface of an isothermal sphere. Using a nonsimilar transform, nonlinear partial differential equations (PDEs) are transformed into dimensionless PDEs. Using the PDSolve technique, dimensionless PDEs are solved. A significant factor in coating engineering is analyzing and calculating velocity distributions, mass, and heat transfers, which were the results of this research. The calculation is carried out using MAPLE 2024, a computational software tool. A deep neural network program was designed, which emphasizes machine learning for predicting the nature of physical phenomena for fluid applications. As part of the validation process of the proposed research, some statistical metrics were taken to assess the degree of error between true values and anticipated values. Based on the results presented, the presented approach is the most efficient method to predict physical quantities for the surface of an isothermal sphere. These results are therefore recommended for the development of industrial device setups.