Wind-tunnel experiments on atmospheric heavy gas dispersion: Metrological aspects

We propose a new experimental protocol to investigate the atmospheric dispersion of a dense gas in wind-tunnel experiments. Simultaneous measurements of concentration and velocity are performed by coupling a flame ionisation detector (FID) and a hot-wire anemometer (HWA), whose probes are located sufficiently close to each other so that the respective measurements correspond to a same sampling volume. The heavy gas emission consists in a mixture of air, carbon dioxide, and ethane. Carbon dioxide is used as a buoyant agent, and ethane is used as a tracer, its concentration being detected by the FID. In this context, the main metrological issue of the setup concerns the response of the FID and the HWA to different proportions of the two gases in the mixture. Notably, the presence of carbon dioxide affects the linear response of the FID, which can attain a saturation level, as well as the response of the HWA, due to the varying density of the gas. This can be compensated by determining a non-linear calibration curve of the FID, which is used to determine the instantaneous fluid density, and to apply a time-dependent correction to the HWA response. To analyse the performances of the coupled HWA-FID system we realised wind-tunnel experiments of tracer dispersion in a turbulent boundary layer. The estimate of mass fluxes through different sections downwind the source enlightens the reliability of the experimental results, and show that the measurements within a negatively buoyant gas plume are as accurate as those within a passive scalar plume.