Academic validation of multi-phase flow codes

Transport equations solved in the multiphase codes are more complex than those in single phase codes. The origin of this difference is mainly due to the complex interactions between the different length and time scales involved in the different phenomena. It is therefore difficult to estimate the potentialities of the numerics in multiphase codes without the help of computational results. For the validation, simple situations relevant for two phase phenomena are chosen. They must be as simple as possible, mainly with single effects and related to the scope of codes devoted to safety analysis. Cases in which one physical phenomenon is uppermost are preferred. Heuristic criteria which must be fulfilled by the computational results are then defined on the basis of our experience in multiphase flow computational analysis. They are illustrated with results from benchmarks. The way each criterion is fulfilled in a given calculation and its relative importance are underlined. This approach enables to separate the numerical and physical part of the results. The results allow to exhibit a correct behavior of the numerical method of the SIMMER-IV code, developed jointly by JAEA (Japan), FzK (Germany), and CEA (France). More generally, the approach is convenient to investigate the potentialities of new developed codes. It allows to avoid unwanted effects like “black box” ones, when unexpected results are obtained in non-academic situations.