Abstract:
This study conducted a validation for the lattice Boltzmann method-based large-eddy simulation (LBM-LES) applied to flow around a single 1:1:2 building. The flow structures and distribution of time-averaged velocity and turbulent kinetic energy (TKE), particularly at pedestrian height, were examined against data obtained by an experiment and finite volume method-based LES (FVM-LES), in addition to checking the grid independence and solution accuracy of LBM-LES. Results of validation metrics showed that the grid resolution of b/16 and b/32 (b: building's width) achieved similar simulation accuracy and satisfied the acceptance range. However, b/32 gave better reattachment lengths. Furthermore, low accuracies of the velocity and TKE was observed in several areas while using b/16, such as above the roof, around the strong wind area near the sidewall, and in the wake flow area at the pedestrian height. LBM-LES generally showed similar fluctuating velocity characteristics as FVM-LES, including power spectra, and probability densities. However, numerical oscillations probably occurred above the building on the windward side. This study showed that LBM-LES could obtain accurate results for both time-averaged and fluctuating velocities. Moreover, at least b/32 is recommended while simulating pedestrian wind or the reverse flow structures, even though b/16 can obtain satisfactory validation metrics.
