Artificial intelligence and numerical simulations for Cattaneo–Christov heat and mass fluxes of nano-encapsulated phase change materials in a zigzag porous cavity

An incompressible smoothed particle hydrodynamics (ISPH) method based on fractional-time derivative carried out numerical simulations of heat and mass transfer of nano-encapsulated phase change materials (NEPCMs) inside a zigzag cavity. The impacts of Cattaneo–Christov heat and mass fluxes and exothermic chemical reactions are conducted in this study. The closed domain is chosen as a zigzag cavity to represent wavy surfaces that can be implemented in enhancement heat and mass transfer in industrial fields. An artificial neural network (ANN) is developed to estimate the values of Nu¯ and Sh¯. The annulus is formed from the inner diamond with Th&Ch and outer zigzag domain containing triangle lines on horizontal boundaries at ∂T∂n=∂C∂n=0 and spline curves on vertical boundaries at Tc&Cc. The performed results are obtained at buoyancy ratio, Cattaneo-Christov heat/mass fluxes, Darcy, Dufour, Soret, Frank-Kamenetskii numbers, fusion temperature, and fractional-time derivative. The results achieved an enhancement in heat and mass transfer by employing buoyancy ratio, and Frank-Kamenetskii numbers. Darcy number supports the distributions of temperature and concentration, and it represents a porous struggle of a nanofluid flow. Fusion temperature controls the position and intensity of a heat capacity ratio between hot and cold sources. The velocity field augments by 21.74% as the Soret number increases from 0 to 1.2. With a strong correlation between the goal values and the prediction values derived by the ANN model, the built ANN model can forecast the Nu¯ and Sh¯ values.