In the present study large-eddy simulations (LES) are employed to study turbulent boundary layers with two- and three-dimensional obstacles on the wall. The presence of the obstacle(s), which is a small portion of the boundary layer thickness, increases substantially the required number of grid nodes in all coordinate directions and therefore the cost. To overcome this problem a parallel multi-domain LES methodology on Cartesian grids has been developed. The equations governing the evolution of the large-scales are solved using a second-order projection method. Spatial derivatives are approximated with second-order central differences on a staggered grid. The large-band matrix associated to the Poisson equation is solved using Krylov subspace iterative solvers. The ratio between the obstacle height and the boundary layer thickness is approximately 7, and the Reynolds number based on the obstacle hight and freestream velocity is of the order 1500. This configuration resembles the conditions in the experiments reported by Vin¸ont et al. (J. Fluid Mech., 424:127-167). Detailed comparisons between the LES and the above experiments for the velocity statistics at various downstream locations will be reported.
Large-eddy simulations of the flow around surface mounted obstacles
Broglia;
2002
Abstract
In the present study large-eddy simulations (LES) are employed to study turbulent boundary layers with two- and three-dimensional obstacles on the wall. The presence of the obstacle(s), which is a small portion of the boundary layer thickness, increases substantially the required number of grid nodes in all coordinate directions and therefore the cost. To overcome this problem a parallel multi-domain LES methodology on Cartesian grids has been developed. The equations governing the evolution of the large-scales are solved using a second-order projection method. Spatial derivatives are approximated with second-order central differences on a staggered grid. The large-band matrix associated to the Poisson equation is solved using Krylov subspace iterative solvers. The ratio between the obstacle height and the boundary layer thickness is approximately 7, and the Reynolds number based on the obstacle hight and freestream velocity is of the order 1500. This configuration resembles the conditions in the experiments reported by Vin¸ont et al. (J. Fluid Mech., 424:127-167). Detailed comparisons between the LES and the above experiments for the velocity statistics at various downstream locations will be reported.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
