AI-driven modeling and simulation for entropy optimized radiative Darcy-Forchheimer flow of micropolar tri-hybrid nanofluid at a stagnation point

The Levenberg-Marquardt (LM) backpropagated artificial neural network (ANN) approach can increase simulation speed and precision, enabling better heat transfer in the design and improvement of solar collectors. The aim of this study is to provide adequate mathematical model for optimized thermal transport and to identify entropy degradation mechanism in a boundary layer flow (BLF) at a stagnation point flow utilizing micropolar tri-hybrid nanofluid (tri-HNF) over a embedded in a porous medium. The contribution of dissipative heat, radiative heat permeable media and microrotation are assumed in the model. The tri-HNF composed of A l 2 O 3 , Cu and Ti O 2 nanoparticles dispersed in water. A shooting methodology in conjunction with the Runge-Kutta Fehlberg (RKF-45) method is used to numerically solve the governing equations. The governing equations include thermal radiation, Darcy-Forchheimer model, viscous dissipation, and micropolar effects. The ANN model built on the LM algorithm is used to accurately estimate entropy optimized flow, temperature, and angular momentum. The findings show that the thermal radiation substantially accelerates the heat transmission in tri-HNF. Raising the micropolar parameter increases the fluid temperature and entropy. The Brinkman number increases entropy formation. The results suggest that tri-HNF performs better thermally than NF and HNF. The ANN model achieves remarkable prediction accuracy, with mean squared error (MSE) values ranging from 10 − 8 to 10 − 10 .