Flow over cube arrays of different packing densities

Measurements by 120° x-wire anemometry over uniform urban-type surfaces of two different area densities were performed in a wind tunnel, together with direct measurements of the surface drag. The aerodynamic characteristics of each surface were determined and compared, the influence of area density and array geometry on these parameters was examined. Various approaches were discussed for the determination of the roughness length (z0) for a given surface. The surface shear stress (determined from form drag measurements by pressure tapping a roughness element or from the total surface drag determined by a floating drag balance) and the shear stress (determined from spatially averaged vertical profiles of Reynolds shear stress) were compared. The surface shear stress was found to be about 25% greater than the measured Reynolds shear stress in the inertial sub-layer over the surfaces. There was, however, no constant stress region and extrapolation of the shear stress profiles in the inertial sub-layer to the zero-plane displacement provided a much better estimate of the surface shear stress. The results did not support the argument put forward in the literature that the zero-plane displacement could be reliably predicted from the height of the centre of drag force. Finally, the accuracy of existing geometrical methods of determining the aerodynamic properties of rough surfaces was shown to be limited by the use of inappropriate wind tunnel data in their formulation.