Shao, Siyao2020-10-262020-10-262020-08https://hdl.handle.net/11299/216831University of Minnesota Ph.D. dissertation.August 2020. Major: Mechanical Engineering. Advisor: Jiarong Hong. 1 computer file (PDF); xv, 121 pages.Ventilated supercavitation has gained considerable research interests over last half century owing to its potential applications in the high-speed underwater transportations. A critical problem to apply such technique is the controlling of the supercavity behaviors in unsteady flow conditions. Compared to the traditional fin-based control, using ventilation to control supercavitation eliminates the additional planing force caused by the controlling surface which limits the speed of the underwater vehicle. However, the lingering problem of employing such control method is the understanding of the gas leakage mechanisms of the ventilated supercavity especially its temporal characteristics. Despite vigorous investigations of the supercavity gas leakage, there is still a dearth of a systematic study of the temporal characteristics of gas leakage due to immense technical difficulties. In this dissertation, we propose to use low cost digital inline holography (DIH) to measure the 3D bubble distribution in the wake of the supercavity to infer its instantaneous gas leakage. This study is two-folded. First, major improvements are made in the conventional DIH processing method which allow us to precisely measure the clustered particles (e.g., bubbles) and estimate their 3D inclinations. With the help of emerging machine learning techniques, DIH processing is further accelerated to accommodate the needs of mass processing holograms captured from the supercavitation wake bubble measurements. This technique development has been applied to many different particle diagnostic tasks such as using DIH to measure oil droplets and pesticide sprays. Second, a systematic quantification of instantaneous gas leakage of ventilated supercavitation is carried out at the high-speed water tunnel at Saint Anthony Falls Laboratory. Both magnitude and temporal occurrence of the gas leakage fluctuation have been revealed and quantified from the experimental results. The experimental findings are further connected with the temporal variations associated with the supercavity gas leakage mechanisms including re-entrant jet gas leakage, vortex tube gas leakage, and bubble shed-off gas leakage. The temporal characteristics of supercavity gas leakage can be used to estimate the controllability of the ventilated supercavity under different flow and ventilation conditions.enDigital inline holographyMachine learningSupercavitationVentilation controlThesis or Dissertation