Large-eddy simulation of turbulent flows over an urban building array with the ABLE-LBM and comparison with 3D MRI observed data sets

In this article we describe the details of an ABLE-LBM (Atmospheric Boundary Layer Environment-Lattice Boltzmann Model) validation study for urban building array turbulent flow simulations. The ABLE-LBM large-eddy simulation results were compared with a set of 3D magnetic resonance image (MRI) velocimetry data. The ABLE-LBM simulations used the same building layout and Reynolds numbers operated in the laboratory water channel. The building set-up was an evenly spaced orthogonal array of cubic buildings (height = H) with a central tall building (height = 3H) in the second row. Two building orientations, angled with 0°and 45° wind directions, were simulated with ABLE-LBM. The model produced horizontal and vertical fields of time-averaged velocity fields and compared well with the experimental results. The model also produced urban canyon flows and vortices at front and lee sides and over building tops that were similar in strength and location to the laboratory studies. The turbulent kinetic energy associated with these two wind directions were also presented in this simulation study. It is shown that the building array arrangement, especially the tall building, has a great effect on turbulent wind fields. There is a Karman vortex street on the lee side of the tall building. High turbulent intensity areas are associated with the vortex shedding motions at building edges. In addition, the wind direction is a very important factor for turbulent wind and kinetic energy distribution. This validation study indicated that ABLE-LBM is a viable simulation model for turbulent atmospheric boundary layer flows in the urban building array. The computational speed of ABLE-LBM using the GPU has shown that real-time LES simulation is realizable for a computational domain with several millions grid points.