Specific requirements for bepu methods using system thermalhydraulic codes with 3d-pressure vessel modelling
Résumé
The 3D modules of system thermalhydraulic codes were first used for transient analysis with a very coarse nodalization including only a few hundreds of meshes in the whole pressure vessel. Today the increased computer power allows 3D simulations with a much finer nodalization for many new transients. A core modelling with one mesh per assembly may become a standard practice in near future. This allows to look at much finer multi-dimensional physical processes and may provide a better accuracy of predictions. Specific requirements are necessary for such finer simulations. First, a more detailed PIRT exercise should identify dominant phenomena at a smaller scale than for previous nodalizations. Based on porous body 3D equations, one can identify a list of phenomena to model and to validate on specific separate effect tests. This includes the turbulent diffusion of heat and momentum, the void dispersion, and the heat and momentum dispersion due to space averaging. In addition, 3D wall friction and interfacial friction tensors have to be developed and validated for non-isotropic media like core rod bundles. An order of magnitude analysis may allow to neglect some of these processes or simplify some models, depending on the sub-component (core, lower plenum, annular downcomer, upper plenum) and depending on the physical situation encountered in accidental transients. Attention will be drawn on LOCA situations in PWRs and on the core component of the Pressure Vessel. The degree of completeness of the physical modelling and of the related validation is evaluated. The impact of numerical errors and of the node size is discussed. Directions are given for future BEPU methods applied to a core modelling with one mesh per assembly by considering a multi-scale validation of the dominant 3D processes.
Origine : Fichiers produits par l'(les) auteur(s)
Loading...