The research presented in this thesis aims at understanding how gases influence the cavitating vortex generated by a wing tip in a stationary flow. Gases in the flow occur as non-condensible microbubbles as well as gases dissolved in the liquid. They are typically abundant in surface flows. For studying how dissolved gases influence a cavitating vortex, a set of diffusion models are developed that predict how gases dissolved in water are transported into a vortex cavity. The predictions of the growth rate of a cavitating vortex using these models is then compared with the growth and development of a cavitating wingtip vortex generated by a NACA 662 −(415) foil. The cavitating vortex is measured using single-view shadowgraph imaging in a cavitation tunnel. Some experimental practices on how the flow conditions are prepared prior to a measurement are also investigated, which is instructive in guiding experimental practices that promote repeatability of the results. A good agreement is achieved between the measured growth rate of the cavity, and that predicted by the diffusion models. The results also explain previously observed behavior of the development of a cavitating vortex from literature.

Following their inception, vortex cavities emanating from stationary wing tips in cavitation tunnels are often observed to grow. These effects are usually attributed to the free and dissolved non-condensable gases in the liquid. However, a detailed mechanism for the cavity's growth is not known. Consequently, the repeatability of vortex cavitation in different flow facilities is generally poor. The main aim of our work is to highlight the contribution of dissolved gases to the cavity's growth, hence addressing water-quality influence in nuclei-depleted conditions. A model is provided for a steady-state diffusion-driven mechanism that transports dissolved gases from the surrounding liquid into the vortex cavitation through a diffusion layer located outside its interface. The model results show that the cavity grows uncontrollably when the dissolved gas concentration in the liquid is saturated or oversaturated relative to its saturation level at ambient pressure conditions (c_∞/c_sat≥1). In addition, it is shown that stable cavity sizes can be achieved when the c_∞/c_sat<1. The predictions in the range 1.04≤c_∞/c_sat≤1.33 are compared with experimental data and infer either of the two geometries for the diffusion layer: (i) a 5μm thin film approximated by a hollow cylinder around the cavity, or (ii) one that evolves like a boundary layer along the axis of the cavity. For the latter modeling approach, the observed length of the cavity was much larger than that required to match with the experimental data, skewing a preference to the thin-film assumption. In the undersaturated regime (c_∞/c_sat=0.14 & 0.39), the proposed model has a qualitative agreement with the data of Briançon-Marjollet and Merle (1996).