Intercomparison of Synthetic Inflow Turbulence Generation Methods for Large-Eddy Simulation Models in Thermally Driven Convective Boundary Layer Simulations

抄録

In this study, synthetic inflow turbulence generation methods developed in computational fluid dynamics (CFD) and meteorological fields were applied to thermally driven convective boundary layer (CBL) simulations. Methods developed in the CFD field include the Reynolds stress Cholesky decomposition and digital filter-based method (DF method), and the cell perturbation method (CPM) is a method developed in the meteorological field. Intercomparison results show that both methods can reproduce turbulence in thermally driven CBLs when a proper driver region is ensured. The turbulence reproduced using the DF method in a thermally driven CBL has a shorter driver region than that reproduced using CPM. However, CPM can be applied to a simulation without limiting the inflow boundary, although it requires a longer driver region than the DF method. Therefore, it was confirmed that both methods have unique merits that can be useful for downscaling from meteorological mesoscale models to microscale large-eddy simulation models.