Improved modelling of downburst outflows for wind engineering applications using a cooling source approach

Large eddy simulations (LES), with a range of different practical ground roughness lengths ( z 0=0.001–0.1 m), are used to compare near surface outflow features of a physically realistic cooling source downburst model, previously validated by meteorological observations, with those of the more commonly used transient impinging impulsive jet. A scaling procedure is proposed, based on length, velocity, and vorticity scales from within the outflow, allowing for direct comparison between outflows from the two models. Five scaling parameters are presented, capturing the horizontal and vertical position of maximum velocity, the ring vortex aspect ratio, the height of the ring vortex above the surface, and a non-dimensional vorticity term representative of the relative contribution of the ring vortex to the near surface wind field. It is shown that the impinging jet model is not capable of capturing the outflow features predicted by the cooling source model, due to its unrealistic forcing parameters, and is, therefore, unable to capture the physics of an actual downburst event. This difference dominates the non-dimensional vorticity term, showing that impinging jet results deviate by at least 56% from the cooling source results, at times when all other scaling parameter differences are minimized.