Wind tunnel modelling of buoyant plumes
The short range dispersion in the atmosphere of buoyant gases, such as hot air or natural gas, may be hazardous and dangerous. The available methods for studying this problem were reviewed. Wind tunnel studies were considered to be the most suitable method for studying near-field dispersion, and methods for accurately modelling the nearfield behaviour of a buoyant plume of gas were examined. The experiments were performed in the Oxford University 4m x 2m low speed wind tunnel at a model scale of 1:200. The mean trajectory and rate of spread of a buoyant plume from a 60 m high (full-scale) stack were measured in the presence of a simulated natural wind. The exact similarity requirements were derived from dimensional analysis and from the equations of motion. In practice, it is not possible to match all the necessary dimensionless groups and exact scaling of the exit gas density ratio and the exit Reynolds number is often relaxed. A series of experiments was performed to examine the effect of these two groups on mean plume behaviour, with the intention of providing guidance for correct simulation of plume dispersion at reduced-scale. The exit density ratio was found to have little effect on the near-field plume behaviour, provided all the other dimensionless groups were matched. Plumes with low Reynolds number were found to rise significantly higher than plumes with higher 'turbulent' Reynolds numbers. This difference in trajectory could not be correlated with the plume exit momentum flux. The effect of the cross-flow on near-field dispersion was examined by performing experiments in four different simulations of the earth's atmospheric boundary-layer. The behaviour of the plume was found to be sensitive to both the velocity profile and the turbulence intensity of the cross-flow. To study dispersion in the wind tunnel, the cross-flow should be an accurate simulation of the velocity profile and turbulence intensity components of the natural wind.