Thermal Analysis of an Exponential Longitudinally Porous Fin with Hybrid Nanoliquid Using Physics-Informed Neural Networks

Fins are widely used in industry, especially in heat exchangers, because they are small, light, and inexpensive. Artificial neural networks (ANNs) have become an important tool for modelling complex, nonlinear heat and flow behaviour, especially where conventional numerical methods are too costly or impractical. This study uses an artificial intelligence (AI)-based numerical approach to analyse steady-state heat transfer in an exponentially profiled longitudinal porous fin wetted with a hybrid nanoliquid. The governing ordinary differential equation (ODE) is formulated using Darcy's law and Fourier's law of heat conduction, and it is reduced to dimensionless form using suitable dimensionless variables and boundary conditions. The physics-informed neural network (PINN) solution is validated against the bvp5c solver, and the two are in close agreement. The results show that the hybrid nanoliquid and the temperature-dependent properties of the exponential fin have a significant effect on the temperature distribution. Furthermore, the PINN and bvp5c results agree well with earlier published studies.