A comparison study on the capabilities of turbulence models and density models of CFD on realistic terrain

Abstract

Two groups of numerical tests were carried out in the mountainous area adjacent to the Hong Kong International Airport (HKIA) in order to test the wind field simulation capability of computational fluid dynamics (CFD) over realistic complex terrain. The first group of tests was carried out for a typhoon case and the simulation results were used to compare the performance of different turbulence models. The comparison between the simulation results and the observational data from a Terminal Doppler Weather Radar (TDWR) shows that the unsteady Reynolds-averaged Navier–Stokes (URANS) method cannot describe the fluctuating airflow triggered by mountains, while the Large Eddy Simulation (LES) method can effectively reproduce the cyclical fluctuations triggered by the mountains, showing a better simulation capability on realistic terrain than URANS. The second group of tests was conducted for a vortex/wave shedding process under stable stratification to test different density models. The comparison between the simulation results and the observed data from the Doppler Light Detection And Ranging (LIDAR) system in HKIA shows that the simulation results using the incompressible ideal gas model were closer to the observational ones than those obtained by using the Boussinesq model. In particular, the incompressible ideal gas model has certain advantages because it was able to give a more detailed description of several weak perturbations than the Boussinesq model in the current study.

Keywords

• CFD
• wind current simulation
• complex terrain
• turbulence model
• density model