Entropy Analysis of Magnetohydrodynamic Laminar Boundary Layer Flow over a Flat Plate with Viscous Dissipation Medium Using Various Regression Analysis

The present study employs an entropy analysis to investigate the boundary layer flow and heat transfer characteristics of an electrically conducting fluid subjected to the influence of a constant transverse magnetic field over a flat plate with viscous dissipation. The Keller box method is utilized to solve non-similar boundary layer equations. The effects of magnetic interaction, viscous dissipation (Eckert number), and group parameters on heat transfer, entropy generation, Bejan number, velocity, and temperature profiles are calculated. The findings demonstrate that a reduction in the magnetic parameter and Eckert number is associated with an enhancement in the heat transfer parameter. Furthermore, the surface behaves as a significant source of irreversibility due to higher entropy generation parameters near the surface. A new correlation is also obtained for the local skin friction and heat transfer parameters. An Artificial Neural Network is constructed to forecast the desired output values. Finally, the problem is mathematically defined with symbolic regression trees using genetic programming and artificial bee colony programming. The performances of the algorithms are evaluated according to coefficient of determination and root mean squared error values. The best model is obtained by ABCP algorithm with three arithmetic operators (+, -, *). The model’s coefficient of determination is found to be 0.9762, while its root mean squared error value is calculated as 0.0129.