Modeling of bubbly and slug flow behavior under microgravity conditions

Abstract

Two-phase flow systems for space applications have advantages over single-phase flow systems. Due to increased heat transfer coefficients, they can achieve the same energy management as single-phase systems with lower mass, size and pumping requirements. In order to correctly design a two-phase system for space, knowledge of the flow behavior in a zero-gravity (0-g) environment is necessary. Development of pressure drop correlations are, specifically, required. To meet this need, the Interphase Transport Phenomena Laboratory Group (ITP) of Texas A&M University conducted two-phase flow experiments aboard the National Aeronautics and Space Administration's (NASA) KC-135 aircraft. The present work is concerned with modeling of two-phase pressure drop under 0-g conditions, for bubbly and slug flow regimes in an adiabatic pipe. Three sets of data were used in this study. The first set, from the ITP group, included three bubbly points, nine bubbly/slug points, nine slug points, and three single-phase points. These single-and two-phase pressure drop data were collected in 1992 aboard NASA's KC-135 aircraft. They were measured in a 10.4 1-mm ID 1. 25 1-m long glass tube during periods with acceleration levels in the range 0︢.05 g. The working fluid used was freon R-12. A methodology to correct and validate the data was developed to achieve high levels of confidence. Two other sets of data, Dukler in 1988 and in 1993 from experiment in the NASA Lewis Research Center Learjet aircraft, are also analyzed. A homogeneous model was developed to predict the pressure drop for particular flow conditions. This model, which uses the Blasius Correlation, was found to be accurate for single-phase, bubbly and bubbly/slug flows, with average errors not larger than 28%. For slug flows, however, the errors were greater, with a value in average of 41 %. The 1-g data were found to be approximately equal to 0-g data for single-phase, bubbly and bubbly/slug flows. However, for slug flows, the 0-g data were approximately 100% greater. Other models based on digital imager observations were also investigated. These included a quasi-homogeneous model, a separated flow model, and the Suo & Griffith model, which studied the transition from bubbly to slug flow. The applicability of these models are shown. These last models were aimed at obtaining some initial scoping results in order to guide future experimental procedures and hardware design. The next flights are expected in spring 1995.