Experimental and analytical results of a liquid-gas separator in microgravity Available to Purchase

The microgravity phase separator designed and fabricated at Texas A&M University relies on centripetally driven buoyancy forces to form a gas-liquid vortex within a fixed, right-circular cylinder. Two phase flow is injected tangentially along the inner wall of this cylinder. Centripetal acceleration is produced from the intrinsic momentum of the resulting rotating flow and drives the buoyancy process. Gas travels under density gradients through the rotating liquid, eventually forming a gaseous core along the centerline of the cylinder. Gas core stability, the presence of liquid in the air line, and the presence of air in the liquid line determine whether a successful core results. To predict separation failure, these three factors were examined both analytically and empirically with the goal of determining what operating circumstances would generate them. The centripetal acceleration profile was determined from angular velocity measurements taken using a paddle wheel assembly. To aid in understanding the nature of the rotating flow, these results were compared to analytical results provided by solving simplified Navier-Stokes equations. The theoretical velocity profile indicated a linear dependence on radius, which with the experimental data agreed, although two distinctly different slopes were observed. As injection nozzle width increased, the difference between the slopes lessened. For all three nozzles tested, the discontinuity between the linear sections occurred at a radius of approximately 3.8 cm. The maximum centripetal acceleration generated by the flow was greatest for the 0.0635 cm wide, 0.516 cm tall injection nozzle and least for the 0.102 cm wide, 1.02 cm tall injection nozzle. The circumstances leading to carry-under are dictated by the relationship between axial and radial bubble transit times. To determine the radial and axial transit times, the radial velocity profile was solved analytically by relating the buoyancy and drag forces for a 0.0635 cm radius bubble. This velocity profile was then used to produce a numerical solution for the radial transit time. Volumetric flowrate analysis provided the axial velocity and bubble transit time. 33.4, 50.1, 66.8, and 83.5 cm³/s flowrates were tested and only the 33.4 cm³/s flowrate resulted in conditions which would lead to carry under.