Abstract
The CFD simulation accuracy mostly depends on the appropriate setting of boundary conditions and numerical simulation parameters. This study shows the influence of two types of boundary condition settings on the CFD simulation results of Double-Skin Facade (DSF) for a specific problem. These two boundary settings are the constant temperature on the DSF surfaces called Boundary A, and Boundary B is defined via solar radiation using the Discrete Ordinate radiation Model (DOM). The paper verified both the numerical simulations using the experimental data. Comparing the numerical results of two types of boundaries with experimental data shows that both cases underestimated the values lower than 5.2 K and 0.1 m/s for the temperature and velocity respectively at the regarded measured points. Boundary A gives more accurate temperature prediction results, while Boundary B shows velocity magnitude closer to the measurements in the middle height of the cavity; the average temperature and velocity differences between the two boundary types are 0.6 K and 0.003 m/s respectively which are negligible. Finally, the selection of boundary conditions depends on study purposes, however, when the DSF is equipped with blinds and if there is not enough data in hand but the exact value of solar irradiation, using the Boundary B approach is suggested; it can provide reasonable results associated with multi-type of thermal boundary conditions at the same time. Furthermore, if the goal is to investigate the flow pattern in the DSF, Boundary B is argued to perform better than the constant temperature boundary condition.