논문 상세보기
Discretization Uncertainty Estimation for Thermal Analysis of the Spent Fuel Transport Cask
p.119
Thermal analysis and safety assessment of spent fuel transport cask are mainly conducted using commercial Computational Fluid Dynamics (CFD) codes based on Finite Volume Method (FVM). The reliability and predictability of CFD codes have greatly been improved by the development in the computer systems, and are widely used to calculate heat flow in complex structures that cannot be analyzed theoretically. In the field of thermal analysis using the CFD code, it is important to clearly reflect the physical model of the transport cask, and a grid configuration suitable for the physical model is essential for accurate analysis. However, since there are no clear standard and guidelines for grid configuration and size, it is highly dependent on the user’s insight. Spatial discretization errors result from the use of finite-width grids and the approximation of the differential terms in the model equations by difference operators. Since the user usually cannot change the truncation error order of a given discretization scheme, spatial discretization errors can only be influenced by the provision of optimal grids. Therefore, it is necessary to quantify the spatial discretization errors caused by the grid. In the case of Orano TN’s NUHOMS® MP197 transport cask, considering four grids for two sets, the temperature uncertainty of the neutron shield, which has the lowest margin at the limit temperature among transport cask components, was quantified by applying 5-step procedure of the Grid Convergence Index (GCI) method for the uncertainty estimation presented in ASME V&V 20-2009. In the case of domestic spent nuclear fuel transport cask (KORAD21), neutron shield among the transport cask components has the lowest margin at the limited temperature. Accordingly, in this study, the temperature uncertainty of the neutron shield was quantified by applying GCI to three sets considering seven grids. As a result of the calculation, the uncertainty was less than ± 1°C, and the temperature of the neutron shield including the uncertainty was evaluated to be maintained below the limit temperature of 148°C.
