Void fraction measurement and analysis at normal gravity and microgravity conditions

Abstract

As the frequency and length of space flights increase, a better understanding of the physical phenomena associated with reduced gravity is needed. One of these processes is two-phase flow. Two-phase gas-liquid flows are encountered in many space applications such as; boiling and condensation, thermal-hydraulic power cycles for space stations and satellites, and in the transfer and storage of cryogenics. An experimental approach is needed to provide the background for accurate modeling and designing of equipment. Measurements of pressure, temperature, flow and void fraction at μ-g were made during several microgravity flights onboard NASA's KC-135 aircraft. High-speed video images were also recorded. The results were later compared to those obtained at 1-g.

This study is focused on the measurement and analysis of the volumetric void fraction in water-air, two-phase flow at μ-g and 1-g. The water used for the tests was either de-ionized and distilled, or filtered through activated carbon and distilled. The volumetric void fraction can be found from the ratio of the volume occupied by the gas to the total volume of the gas and liquid. Two capacitance void fraction sensors were used. In the early stages of void fraction measurement, a helical-wound-electrode void fraction sensor was designed and tested in February 1994. Over the course of this research, some of the problems associated with this sensor were identified and a new concave-plate-electrode capacitance sensor was developed having a linear response over the flow settings and 10 times the sensitivity of the helical wound sensor. Data was collected covering a wide range of void fraction, from approximately 0.1 to 0.9 at both 1-g and μ-g conditions. The flow regimes encountered included bubble, slug, transitional flow, and annular flow.

Void fraction values for slug flow appear to be slightly higher at μ-g. The average void fraction values for the remaining flow regimes do not appear to show any discernible difference. The development of the void fraction and flow profiles conducted by Zuber and Findlay (1965), was used to compare the profiles found at 1-g and μ-g. These results indicate that the void fraction profile is slightly flatter at 1-g for slug flow. Using this model, the results for bubble flow at 1-g agree with the results reported by other researchers where the well known "saddle" shape profile was found. A statistical approach was used by plotting probability density functions for the 1-g and μ-g void fraction data. A wider fluctuation in void fraction was found for bubble and slug flows at 1-g compared to μ-g. The probability density functions for the highly inertia dominant transition flow and annular flow regimes at 1-g and μ-g were comparable.