Fast fluid dynamics simulation of airflow around a single bluff body under different turbulence models and discretization schemes

Abstract

Fast and accurate simulation of outdoor airflow distribution is important for studying urban microclimate. Choosing a reasonable turbulence model and discretization scheme is not only related to the computational accuracy but also to efficiency. However, conventional CFD methods are computationally intensive and slow for unsteady problems, and thus cannot meet the demand for fast simulation of urban microclimate. In this paper, three pressure-correction schemes (i.e., NIPC, SIPC, and NSPF) for solving the N–S equation item by item are implemented in OpenFOAM, and then the differences are compared when applying different turbulence models and discretization schemes to quickly simulate the airflow distribution around a single 1:1:2 bluff body. All pressure-correction schemes can accurately predict the main airflow characteristics around the bluff body. The three schemes are about 2.5–3.5 times faster than the PISO algorithm, and they take the shortest computational time when applying RKE, followed by SQKE and RNG, while the longest computational time is required when applying SKE and LBKE. NIPC and SIPC have similar computational speeds, while NSPF is about 10–16% faster than them. The pressure-correction scheme with the first-order upwind scheme is about 6–10% faster than the second-order discretization scheme. Considering both computational accuracy and efficiency, the combination of NSPF with RNG or SQKE turbulence model and first-order upwind scheme may be a reasonable choice to quickly simulate the urban airflow distributions.