Validation of CONTAIN-LMR code for accident analysis of sodium-cooled fast reactor containments
CONTAIN-LMR 1 is an analytical tool for the containment performance of sodium cooled fast reactors. In this code, the modelling for the sodium fire is included: the oxygen diffusion model for the sodium pool fire, and the liquid droplet model for the sodium spray fire. CONTAIN-LMR is also able to model the interaction of liquid sodium with concrete structure. It may be applicable to different concrete compositions. Testing and validation of these models will help to qualify the simulation results. Three experiments with sodium performed in the FAUNA facility at FZK have been used for the validation of CONTAIN-LMR. For pool fire tests, calculations have been performed with two models. The first model consists of one gas cell representing the volume of the burn compartment. The volume of the second model is subdivided into 32 coupled gas cells. The agreement between calculations and experimental data is acceptable. The detailed pool fire model shows less deviation from experiments. In the spray fire, the direct heating from the sodium burning in the media is dominant. Therefore, single cell modeling is enough to describe the phenomena. Calculation results have reasonable agreement with experimental data. Limitations of the implemented spray model can cause the overestimation of predicted pressure and temperature in the cell atmosphere. The ability of the CONTAIN-LMR to simulate the sodium pool fire accompanied by sodium-concrete reactions was tested using the experimental study of sodium-concrete interactions for construction concrete as well as for shielding concrete. The model provides a reasonably good representation of chemical processes during sodium-concrete interaction. The comparison of time-temperature profiles of sodium and concrete shows, that the model requires modifications for predictions of the test results. (authors)