Prediction of MHD flow and entropy generation by Artificial Neural Network in square cavity with heater-sink for nanomaterial
2019 Elsevier B.V. Numerical analysis of magneto-hydrodynamic flow has been a matter of concern for research engineers and scientists. In this paper, magneto-hydrodynamic convection in square tank occupied with Cu-H2O nanomaterial is investigated for different configurations of heater-sink, in which Artificial Neural Network (ANN) model was used as an advanced predictive tool. The active semi-circular thermal location (heater and sink) at the left- right vertical sides are kept constantly at high and low temperatures respectively, whereas other walls are kept adiabatic. To reach the solution, Galerkin residual finite element analysis has been implemented. The investigation has been done for Hartmann number (Ha = 0 - 100), Rayleigh number (Ra = 103-107) and nanomaterial concentration (φ=0 - 0.05) and finally, streamlines, isotherm contours and entropy generation contours are discussed thoroughly. The overall heat transfer and generation entropy are quantitatively investigated by overall Nusselt number (Nu) and Bejan number (Be), respectively. Existence of external Lorentz forces affects on both non-dimensional performance parameters, Nu and Be. Finally, the higher heat transfer is found for middle-middle configuration of heater-sink walls. The impact of Ha and φ on Nu and Be found from the numerical heat transfer analysis has been predicted and compared with ANN prediction model. To be noted, ANN is widely used technique to compare and predict different experimental and numerical data accurately in many engineering applications.