Evaluation of a full-scale helium-filled soap bubble generator

Various aspects of the design and operation of a full-scale helium-filled soap bubble generator are studied. Shadowgraphy, particle image/tracking velocimetry, hotwire anemometry, and Monte Carlo simulations are employed to investigate bubble production regimes, diameters, production rates, time responses, and the flow quality downstream from the full-scale system. Modifications to internal nozzle geometry are found to greatly impact the production regimes that the nozzles operate within. Specifically, improving the axisymmetry of the air flow within a nozzle leads to desirable bubble formation over a larger range of input combinations and the ability to operate at larger input rates in general. The input of bubble film solution (BFS) is also found to be important for ensuring proper operation, as both small and large inputs lead to undesirable production. A previously defined theoretical relationship (Faleiros et al., Exp Fluids 60:40, 2019) between input parameters and the mean bubble time response is confirmed but found to vary depending on nozzle operation, as spilled BFS and leaked helium during bubble formation cause deviation from theoretical operation. Monte Carlo simulations reveal that the spatial filtering of particle image velocimetry (PIV) reduces the standard deviation of the effective distribution of the bubble time responses by a factor of ${\text{PPIR}}^{-1/2}$, where PPIR is the number of particles per interrogation region. This power law is used to derive an equation for estimating the minimum time scale of the flow that can be resolved using the bubbles from a given generator during applications of PIV. Finally, the wind tunnel flow downstream from a full-scale generator is found to be affected by the blockage of the structure, with the freestream deficit increasing by at most 1.2% of the mean and the freestream turbulence intensity increasing by at most 0.3% for freestream velocities of 6 m/s or greater.