Consistency improved random flow generation method for large eddy simulation of atmospheric boundary layer

Abstract

Realistic turbulence is the key component to obtain accurate simulations for large eddy simulation of atmospheric boundary layer (ABL). At present, synthetic turbulence methods have been widely applied, but there are still some defects such that spatial correlations and cross-correlations between different velocity components cannot be strictly reproduced. Thus, we propose a new inflow turbulence generation (ITG) method, the consistency improved random flow generation (CIRFG) method. The proposed method is derived to model the spatial correlations and cross-correlations between different velocity components by explicitly embedding turbulence characteristics, thus it can be used widely in arbitrary situations without any empirical parameters. Meanwhile, the CIRFG method is maintained to satisfy arbitrary mean velocity, turbulence intensity, frequency spectra, temporal correlation, etc. Further, according to the simulation of ABL flows, it is verified that the new method has good self-sustainability of wind profiles. Finally, compared with the simulations of flow around high-rise building model with wind tunnel tests, it is demonstrated that the proposed method has high accuracy and feasibility.