Methodology Development of a Gas-Liquid Dynamic Flow Regime Transition Model

Current reactor safety analysis codes, such as RELAP5, TRACE, and CATHARE, use flow regime maps or flow regime transition criteria that were developed for static fully-developed two-phase flows to choose interfacial transfer models that are necessary to solve the two-fluid model. The flow regime is therefore difficult to identify near the flow regime transitions, in developing two-phase flows, and in transient two-phase flows. Interfacial area transport equations were developed to more accurately predict the dynamic nature of two-phase flows. However, other model coefficients are still flow regime dependent. Therefore, an accurate prediction of the flow regime is still important.;In the current work, the methodology for the development of a dynamic flow regime transition model that uses the void fraction and interfacial area concentration obtained by solving three-field the two-fluid model and two-group interfacial area transport equation is investigated. To develop this model, detailed local experimental data are obtained, the two-group interfacial area transport equations are revised, and a dynamic flow regime transition model is evaluated using a computational fluid dynamics model.;Local experimental data is acquired for 63 different flow conditions in bubbly, cap-bubbly, slug, and churn-turbulent flow regimes. The measured parameters are the group-1 and group-2 bubble number frequency, void fraction, interfacial area concentration, and interfacial bubble velocities. The measurements are benchmarked by comparing the prediction of the superficial gas velocities, determined using the local measurements with those determined from volumetric flow rate measurements and the agreement is generally within +/-20%. The repeatability four-sensor probe construction process is within +/-10%. The repeatability of the measurement process is within +/-7%. The symmetry of the test section is examined and the average agreement is within +/-5.3% at z/D = 10 and +/-3.4% at z/D = 32. Revised source/sink terms for the two-group interfacial area transport equations are derived and fit to area-averaged experimental data to determine new model coefficients. The average agreement between this model and the experiment data for the void fraction and interfacial area concentration is 10.6% and 15.7%, respectively. This revised two-group interfacial area transport equation and the three-field two-fluid model are used to solve for the group-1 and group-2 interfacial area concentration and void fraction. These values and a dynamic flow regime transition model are used to classify the flow regimes. The flow regimes determined using this model are compared with the flow regimes based on the experimental data and on a flow regime map using Mishima and Ishii's (1984) transition criteria. The dynamic flow regime transition model is shown to predict the flow regimes dynamically and has improved the prediction of the flow regime over that using a flow regime map. Safety codes often employ the one-dimensional two-fluid model to model two-phase flows. The area-averaged relative velocity correlation necessary to close this model is derived from the drift flux model. The effects of the necessary assumptions used to derive this correlation are investigated using local measurements and these effects are found to have a limited impact on the prediction of the area-averaged relative velocity.