A Modified Model For Thermal Mixing And Stratification Under Natural Circulation In Passive Containment

Thermal mixing and stratification often appears in a passive containment cooling system (PCCS). The stratification phenomena tend to block the establishment of natural circulation, which results in deleterious effects on heat removal function of the containment which only relies on natural circulation to cool down. Therefore it is important to accurately predict the temperature and density distributions both for design optimization and accident analysis.However, current major reactor system analysis codes such as RELAP5and TRAC only provide lumped parameter models which can only give very approximate results. The traditional2-D or3-D CFD methods require very fine grid resolution to resolve thin substructures. Previous scaling study has shown that stratified mixing processes in large stably stratified enclosures can be described using one-dimensional differential conservation equations, with the vertical transport by free and wall jets modeled using standard integral techniques. This allows very large reductions in computational effort compared to three-dimensional numerical modeling. The BMIX++code was developed at University of California at Berkeley, basing on the above idea.In this thesis, the BMIX++code is used to solve heat transfer and fluid mixing problems in large enclosures which can give the temperature and velocity profiles in a very short time without complicate meshing. However, the code had only been used to simulate rectangular or cylindrical enclosures with constant cross sectional area. In order to simulate enclosures as the containment with a dome, variable cross section area model is added and suitable physical models are selected for containment simulations.At the same time, a series of small scaling containment experiments with air injection are conducted to simulate the LOCA accidents. Simple variable Method is adopted in the experiments to study the effects of five different factors to the flow field. By studying the experiments results, the temperature fields tend to form two layers, one with an ascending temperature region in the lower space and the other with a stable temperature region in the upper space. A new boundary condition assumption is used in the new code to perform calculations, and the results of the experiments are compared with the numerical results from the new code. It is concluded that the new code gives good results when strong stratification phenomenon occurs, especially for the stable temperature region in the upper space; For the lower space, the calculated temperature profile is steeper than the experimental data. One possible explanation is that the simulation did not consider the downward flow along walls which impairs the increase of temperature, so that the temperature increases faster in the results of new code. In the end, two kinds of boundary condition assumptions were conducted in the new code and the results are compared with the CFD results and experimental results. Finally, the conclusions include that the new code runs much faster than CFD codes and the good agreement with experimental data shows the advance of the models used in the new code."...