Validation of a novel downdraff outflow simulator: A slot jet wind tunnel.

摘要

Downdraft outflows are known to inflict catastrophic damage to built structures. An experimental test facility is needed, in which models of built structures are subjected to this type of wind load to ensure adequate structural design. This thesis investigates the physical simulation of a downdraft outflow using a novel approach. It is hypothesized that a downdraft outflow could be adequately represented, for wind engineering purposes, by a plane wall jet arising from a rectangular slot nozzle. A physical test facility was developed and validated against available downdraft outflow observations, and results from numerical and experimental simulation of wall jets.;The wall jet findings aided the design of a downdraft outflow simulator based on generating a plane wall jet with a rectangular slot nozzle. The outflow gust was modelled with an actuated gate that generated a dominant roll vortex. Wind speed measurements were compared to predictions by a sub-cloud numerical model. Simulator performance was compared with field observations of a 2002 Texas event. Wind speed records were decomposed according to classical time series analysis. Length scales, characterising the coarse and fine flow structure, were determined from the time-varying mean and residual components, respectively. The simulated downdraft outflow was approximately 1200 times smaller in spatial extent than the 2002 Texas event. Keeping in mind the variability of downdraft size, microburst outflows would be modelled at 1:100 to 1:500 length scaling in the Wind Tunnell Downdraft Outflow Simulator at The University of Western Ontario, i.e. a range comparable to that of conventional boundary layer wind tunnels.;Keywords: downdraft outflow, downburst, microburst, thunderstorm gust, turbulent wall jet, slot jet, plane wall jet, hot-wire anemometry, wind tunnel, time series analysis.;Wall jets are typically generated in a plane or radial configuration. The feasibility of approximating one configuration with the other was investigated. The discrepancy between the normalised mean velocity profiles for the developed plane wall jet and the developing radial wall jet was generally less than the range attributable to total measurement uncertainty. Where maximum turbulence intensities and turbulent kinetic energy production occurred, turbulence quantity profiles for the developing radial wall jet provided the best match with those of the developed plane wall jet.