Large Eddy Simulation Of Turbulent Atmospheric Boundary Layer Based On Recycling-Rescaling-Revising Method

As lots of super-high rise and long-span structures shows, the wind-induced fluctuating response received more and more attention. In Computational Wind Engineering, the study of steady wind characteristics using Reynolds-Averaged Navier-Stokes equation(RANS) has been so developed. But another essential problem is still need to be studied, which is to high-fidelity simulate the characteristics of the Turbulent Atmospheric Boundary Layer(turbulent ABL).As one of the special turbulence boundary layer s, turbulent ABL has similar fundamental theories and also some same characteristics with spatially developing flat-plate turbulent boundary layers. In some extent, the atmospheric boundary layer of earth could be seen as a flat-plate boundary layer in planetary scale. Thus the successful simulation results of low-Reynolds boundary layer can be a good basement for high-Reynolds cases. In this paper, we analyzed the basic principle of the recycling-rescaling method proposed by Thomas Lund, which is one of the efficient simulation methods for spatially developing turbulent boundary layers, and use this methodology to implement the UDF parallel programs of ANSYS-FLUENT. Then we finished the large-eddy simulation of spatially developing flat-plate boundary layer, and focus on the effects of initial condition, the way of velocity re-introducing and the station of recycle plane.In practical application, the smooth wall condition can be difficult t o fit and thus we proposed an active control method by adjusting the fluctuating velocity within the boundary layer. By constructing a fluctuating coefficient, the flat-plate boundary layer at low Reynolds number is simulated.Based on the same control method proposed in low-Reynolds turbulent boundary layer simulation, we finished the large-eddy simulation of A-type and B-type turbulent atmospheric boundary layers by introducing the assumption that the boundary layer thickness will keep constant along with the streamwise direction. At last, we give a brief result comparison between active-control method and passive-control method for A-type ABL simulation.