ANALYSES TO DEMONSTRATE THE STRUCTURAL PERFORMANCE OF THE KN18 IN HYPOTHETICAL DROP ACCIDENT SCENARIOS

The KN18 is a new cask design by KONES for KHNP for the dry or wet transportation of up to 18 PWR spent nuclear fuel assemblies in South Korea. The containment vessel consists of a cylindrical thick-walled forged steel body, closed by a stainless steel lid with bolts. Spent fuel assemblies are located in a basket which consists of a tube disc system. Two pairs of trunnions are attached for lifting, manoeuvring and tie-down. A pair of impact limiters manufactured from wood and encased in steel cladding provide impact energy absorption during the hypothetical accident conditions. The package complies with the requirements of 10 CFR Part 71 for Type B(U)F packages. It received its transport license from the Korean Competent Authority KINS in early 2010 and is expect to enter service in 2011. Structural performance of the package in the normal and accident conditions were demonstrated against the requirements of 10 CFR Part 71 by analysis including extensive calculations by state-ofthe-art finite element methods, and confirmed by tests carried out on a 1/3 scale test model which were also used to verify the numerical tool and methods used in the analyses. For the analyses of the hypothetical accident drop conditions, the models consisted of the complete package - including the impact limiters, the containment structure and the basket – which was modelled explicitly in detail and in three dimensions, to take into account the complex interaction between the components and the non-linearities in the geometry, the material behaviour and overall behaviour. The analyses were carried out using the explicit transient finite element method so that the transient behaviour could be robustly simulated. This paper presents two of the analyses from the suite of analyses for demonstrating the performance of the package in the hypothetical accident drop scenarios, discussing the analyses methodology, modelling technique and evaluation methodology, as well as analyses results and package response. 1/3 scale model drop testing and benchmarking of the model to the scale model tests are the subject of a separate paper.