The onset of turbulence in a simulation of the oscillating flow over a burning propellant

In a rocket motor, acoustic oscillations can affect the burning rate of the propellant, leading to combustion instabilities. It is very difficult and expensive to study experimentally the behavior of oscillatory burning in solid propellant motors. For this reason, a simulation facility was constructed that uses solid carbon dioxide (dry ice) to simulate the internal flow field of a solid propellant rocket motor. The use of dry ice makes it possible to study the fluid mechanical aspect of rocket motors without combustion. The research includes modifying the existing simulation facility to improve certain experimental deficiencies and implementing new experimental techniques. The work involves quantifying the occurrence of acoustically introduced turbularization and investigating flow phenomena that may contribute to possible acoustic instability sources.; In these experiments the onset of turbulence in a rectangular geometry was investigated with and without side-wall injection. Three techniques were used to define the transition from laminar to turbulent regimes: statistical analysis, spectral analysis, and flow visualization. Calibrated hot film anemometry and a computer data acquisition system were used to record and analyze acoustic flow data. Four classifications of flow regimes were achieved: (a) laminar, (b) distorted laminar, (c) weakly turbulent, and (d) conditionally turbulent. A fully turbulent flow was not developed at any of the driving frequencies tested. A relaminarization always occurred within a periodic cycle. The transition between the flow regimes was defined by the standard deviation of velocity data as a function of acoustic Reynolds number.; The oscillatory velocity data was compared to the wave amplitude predicted by one-dimensional acoustic theory. For the case without side-wall injection, the experimentally measured velocity amplitude matched closely the theoretical wave amplitude since the hot film probe was located outside the Stokes boundary layer region. For the case with side-wall injection, the experimentally measured velocity above the transpiring surface exhibited an overshoot of approximately 1.5 times the theoretical one-dimensional wave amplitude. This overshoot was expected since for unsteady and conditionally turbulent flow, the hot film probe was shown to be located within the Stokes boundary layer region formed at the transpiring surface.