DEVELOPMENT OF NUMERICAL METHODS FOR
VULNERABILITY ASSESMENTS

The protection against the release of radioactive material through sabotage is one of the primary objectives of the state´s nuclear security regime. Among other things this also applies to transports of such radioactive materials, which is why the state should specify which transports require protection against sabotage because of the potentially caused unacceptable radiological consequences. The potential radiological consequences of a successful act of sabotage can be determined by a quantitative estimate as part of a vulnerability assessment for a transport. In Germany a nuclear transport is classified into a security category that combines threats against unauthorized removal and sabotage, so the assessment of the potential consequences is part of this classification. The potential radiological consequences of sabotage are calculated by using both analytical and numerical tools, and these calculations are based on established and verified methods of calculating release and dispersion processes. In order to develop such methods for potential sabotage scenarios and release processes, both an experimental basis and a development of numerical models are essential. Within the scope of two research projects, the German expert and research association for nuclear safety and security GRS investigates release processes caused by the impact of a shaped charge in the form of a two-phase (particle loaded) discontinuous free jet, which is emitted from a package. The goal of our research activities is to analyze whether numerical simulations using CFD (Computational Fluid Dynamics) software enables an improvement of the accuracy and flexibility of the calculation methods. The outflow behavior of a discontinuous free jet is considered in the simulations, systematically evaluated, and then compared to experimental data collected in the course of one of the research projects. It is shown that a good data basis could be created based on experiments and that the results of the simulations to date can reproduce them well. An application of these methods in future vulnerability assessments is discussed.