Thermal analysis of wet and dry porous fins using hybrid nanofluids under convective-radiative conditions with an artificial neural network and self-organizing map approach

초록

This study presents a comprehensive thermal analysis of wet and dry porous triangular fins containing aluminum oxide–copper/water hybrid nanofluids under combined convection and radiation effects. A one-dimensional transient mathematical model is developed and solved numerically using an implicit finite difference scheme to investigate the influence of thermal and geometric parameters on temperature distribution and fin efficiency under different thermal boundary conditions. Particular attention is given to the comparative performance of wet and dry fin surfaces subjected to sinusoidal and cosine thermal forcing. The numerical results reveal that wet fin configurations exhibit superior heat dissipation compared with dry fins, resulting in lower temperature levels and enhanced thermal performance. The effects of convection, radiation, internal heat generation, fin geometry, phase variation, and boundary temperature amplitude are systematically examined. To accelerate thermal prediction and facilitate design-space exploration, a machine-learning framework combining an Artificial Neural Network and a Self-Organizing Map is developed. The predictive model demonstrates excellent agreement with numerical results, while the clustering analysis identifies distinct thermal operating regimes and parameter interactions that are difficult to recognize through conventional parametric studies. The findings indicate that wet porous fins subjected to cosine boundary conditions provide the most effective thermal management performance among the configurations considered. The proposed numerical and machine-learning framework offers an efficient tool for thermal design and optimization of porous fin systems used in electronics cooling, energy systems, and thermal management applications.

키워드