The Impact of Conjugate Heat Transfer in Flow Over a Vertical Plate and Application of Artificial Neural Network

Heat conduction properties are important in the flow area for the simulation of engineering applications where heat transfer is needed between both liquid and solid. Conjugate Heat Transfer (CHT) refers to thermal problems involving both conduction within the wall and convection within the fluid. CHT is crucial for heat exchangers, gas turbine blades, nuclear reactor cooling pipes, aircraft engines, and spacecraft. Additionally, the influence of the magnetic field is significant in fields such as electrostatic precipitation, MHD power generators and pumps, aeroheating, and polymer science. The motivation for this study is to understand the combined effects of CHT, mixed convection, magnetic fields, and viscous dissipation on velocity, temperature profiles, local skin friction, and heat transfer parameters over a vertical plate. The boundary layer equations have been derived from the Navier-Stokes and energy equations using similarity methods and solved numerically with the Keller Box technique. A new correlation for local skin friction and heat transfer parameters has been developed. Moreover, Artificial Neural Network (ANN) models have been applied to forecast desired numerical values. The optimal ANN model for local heat transfer has one hidden layer and nine neurons, achieving an R² value of 0.9077607 and an MSE of 0.0003101. For local skin friction, the best-performing model has one hidden layer and fifteen neurons, with an R² value of 0.9470261 and an MSE of 0.0250369.