Defence Research Establishment Suffield Medicine Hat, Alberta Canada
Two canopy windflow models basic model and alternative model are described and assessed against measurements of mean wind and turbulence in a wide range of canopies both vegetative and urban. The windflow models are based on a first-order turbulence closure for the description of both the mean wind U and kinetic energy of turbulence kappa, viz. Reynolds equations are closed using eddy-viscosity K lambda kappa 12, where lambda is an algebraically specified turbulence length scale. The windflow models are compared with eleven different canopy flows aeroelastic plant canopy corn canopy square and staggered arrays of cubical obstacles at three different frontal area densities 0.0625, 0.16, AND 0.44 square and staggered arrays of billboard-shaped obstacles at a frontal area density of 0.16 and, the Tombstone canopy. Comparisons of model predictions with these different canopies show that the alternative windflow model gives the best overall conformance with respect to both mean flow and turbulence profiles. The basic windflow model was found to underpredict the turbulence kinetic energy within the canopy, although this underprediction does not appear to affect its reasonably good predictions of the mean wind speed and the shear stress. However, both the basic and alternative windflow models underpredict severely the magnitude of the mean wind shear at the canopy top for urban canopies at the highest frontal area density 0.44 investigated. It is suggested that the alternative windflow model can be used probably unmodified viz., without resorting to tuning of any model constants for the simulation of a wide range of canopy flows provided the frontal area densities of the urban canopies are not so large as to provoke skimming flow.