Windbreak aerodynamics: Is computational fluid dynamics reliable?

To investigate the suitability of computational fluid dynamics (CFD) with regard to windbreak aerodynamics, simulations are performed with a state-of-the-art numerical scheme (Fluent) and compared against experimental data for two- and three-dimensional disturbances, namely the case of a long straight porous shelter fence and the case of a shelter fence erected in a square about an enclosed plot. A thorough sensitivity study quantifies the impact of numerical choices on the simulation (e.g. grid-point density, domain size, turbulence closure), and leads to guidelines that should ensure objective simulation of windbreak flows. On a fine grid Fluent's "realizable k-epsilon closure" gives results that are in qualitative accord with the observed mean winds.