Scholar - SciOpen

The artificial neural network (ANN) in conjunction with the incompressible smoothed particle hydrodynamics (ISPH) approach, deals with exothermic reaction effects on Cattaneo-Christov (Ca-Ch) heat and mass transport of nano-enhanced phase change material (NEPCM) in a curvilinear cavity. The ANN model, trained on data obtained from ISPH simulations, accurately predicted the mean $\bar{N u}$ and $\bar{S h}$ values. Two cases of boundary conditions included $(T_{h} & C_{h})$ on top/bottom walls and $(T_{c} & C_{c})$ on vertical walls and inner ellipse for C1. The boundary walls of a curvilinear cavity were kept at $(T_{h} & C_{h})$ and the inner ellipse was maintained at $(T_{c} & C_{c})$ for C2. The pertinent parameters were scaled as Frank-Kamenetskii number $F k (0 - 1,)$ Ca–Ch heat, mass transfer parameters $(δ_{θ} & δ_{Φ}) (0 - 0.2),$ Hartmann number $H a (0 - 60),$ buoyancy ratio parameter $N (- 2 - 4)$ , power law index parameter $n (1.1 - 1.4)$ , Rayleigh number $R a (10^{3} - 10^{5})$ , Soret/Dufour numbers $(S r & D u) (0 - 0.5)$ , and fusion temperature $θ_{f} (0.1 - 0.9)$ . The simulation results demonstrated the effectiveness of Ca-Ch heat and mass transport parameters in lowering temperature and concentration within a curvilinear cavity at C1 and C2. Increasing $δ_{θ} & δ_{Φ}$ from 0 to 0.2 resulted in a $44.1 %$ and $48.9 %$ drop in velocity field at C1 and C2, respectively. Boundary conditions (C1 and C2) significantly affected mass, heat transfer, heat capacity ratio, and velocity field within a curvilinear cavity. An increase in Power law index $n$ from 1.1 to 1.4, reduced a velocity field by $64.68 %$ and $64.66 %$ at C1 and C2, respectively. Increasing $S r$ and $D u$ helped distribute concentration. When $S r$ and $D u$ were raised from 0 to 0.5, the velocity field increased by $34.17 %$ and $29.73 %$ , respectively, at C1 and C2.