Status and challenges of CFD-modelling for poly-disperse bubbly flows

A clear progress was achieved during the last 20 years in the qualification of CFD-codes for problems of the nuclear safety research. Especially two-phase flows are important, e.g. for LOCA scenarios, but up to now the predictive capabilities of CFD-methods for such flows are limited. Two-phase flows are determined by complex interactions between the phases. Some of them are not yet well understood at the local scale and, therefore, CFD models are limited. This paper discusses such local phenomena and their reflection in presently available CFD-models. It turns out that most of the assumptions in the formulation of closure models for the multi-fluid approach reflect the real phenomena only in a coarse way. Possible uncertainties are listed. Nevertheless, the simulation results obtained by the HZDR baseline model for poly-disperse flows in which all models including model parameter are fixed show in general a rather good agreement with experimental data. One sensitive issue seems to be how to handle the bubble size. In case of poly-disperse flows the sub-division of the gas phase with respect to the bubble size is important and the exact choice of the limits for this division sensitively influences the simulation results.