International Journal of Heat and Mass Transfer | 2021
Toward a local drift flux model for high-pressure, subcooled, convective boiling flows
Abstract
Abstract Forced convective boiling is of great interest for several applications in the power and process industry, particularly in nuclear plants. Under certain nominal, incidental or accidental conditions, a boiling crisis may occur resulting in the meltdown of the heating surface. It is then essential to predict as accurately as possible the thermal-hydraulic conditions leading to the occurrence of this boiling crisis. Such an objective cannot reasonably be achieved without a good description of the associated two-phase flow. The objective of the present study is twofold: (1) to identify the necessary key parameters for correctly describing boiling flows, and (2) to present in a didactic way an original stationary and local model involving these parameters. This new model is primarily based on four mixture balance equations, a submodel for the local vapor generation rate, and a turbulence submodel inspired by the pioneering work of [25]. The results obtained with this original boiling flow model are then compared to an extensive experimental data set obtained on a R12/R134a experimental facility. The comparison clearly demonstrates that this new model contains the fewer necessary submodels to describe the structure of a boiling two-phase flow under pressurized water reactor conditions. Subcooled boiling is acceptably described by the model. However, for higher values of void fraction, the model always predicts a nonexistent void fraction peak near the heating wall and overpredicts the wall and liquid temperatures. This behavior may be explained by: (i) the inadequacy of the radial turbulence modeling, (ii) the use of Prandtl s analogy whose validity under boiling conditions is questionable, and (iii) too simplistic a model for the vapor generation rate.