Some observations on the status of two-phase critical flow models

Abstract

INTRODUCTION Bour6 (1978), Weisman & Tentner (1978), Lahey & Moody (1977), and Brittain (1977) have presented recent reviews on critical two-phase flow. The paper by Henry (1979), while concentrating on critical two-phase flow in nozzles, presents an excellent organization of the literature distinguishing studies involving subcooled and saturated stagnation conditions and flow geometries. Wallis (1979) develops an overview of the model developments, seeks to assess whether, for given restraints, some models are performing well enough, and comments on the degree of sophistication being used in model developments. This paper offers a few brief observations on some selected topics dealing with the acceptability and the deficiencies in the modelling of critical two-phase flow. To prepare the reader for the particular objectives of this paper, the following points are made: (1) There is no single, best estimate model for predicting critical two-phase flow mass rates. Even after more than three decades of research, critical two-phase flow is still an active area for research; however, there are models, as Henry (1979) notes, which appear to be adequate for carefully specified conditions. (2) A primary need for calculating critical two-phase flow is linked with the safety and the performance of emergency core cooling systems for nuclear power reactors. As will be explained in the paper, this linkage has created some artificialities and thus some confusion in what should be the appropriate research. (3) The special attention being given to critical two-phase flow grows out of the need for realistic predictions of the transient mass discharge rates from postulated breaks in the primary system of nuclear power reactors. Safety margins now embedded in the licensing evaluation models are believed to be adequate, but quantification of the safety margins remains to be ascertained. Even under carefully controlled experimental conditions with simulated breaks and depressurization of systems scaled to nuclear power systems, the data reduction for interpreting transient critical two-phase flows and the comparisons among models have had some problems. (4) The paper by Wallace (1979) present judgments on which models should receive preference and which appear to be ill-advised. Though such heuristics may have merit, there is another way of addressing this problem and that is to be develop a more thorough process for model selection through comparisons with data. First, the data base for comparing model predictions needs to be reevaluated for determining the experimental errors. The selection process should present the inherent limitations used in the model, conditions for applicability, model errors, and uncertainties in the predictions. Conditions need to be agreed upon for specifying when a model is acceptable. In an example to be presented on code verification or code assurance, this process of model selection could not be completed. Other examples taken from the literature will illustrate that though comparisons of model predictions with experimental data provide an interim guide, much more work is needed to resolve acceptability of models. (5) This paper, together with papers by Henry (1979) and Wallace (1979), serves to