Probabilistic failure assessment of an irradiated DEMO breeding blanket component under ferromagnetic loads: Impacts of material data scattering and uncertainty in local stress intensities

Abstract

The most significant technical challenge in developing structural materials for in-vessel components in the Demonstration Power Plant (DEMO) fusion reactor is to qualify the lifetime of material to guarantee the lifetime of components. The qualification needs to be conducted theoretically rather than empirically because it is conducted based on the knowledge and data acquired in fission neutron irradiation experiments and various simulation irradiation experiments rather than in experiments performed in the actual fusion environment. Such experiments will not be achievable until the DEMO fusion reactor becomes operational. Consequently, various uncertainties are expected in irradiation data and loading conditions.The probabilistic approach, where the probability of failure is calculated based on the probability density function of postulated load distribution and material property distribution, is one path to mitigate such uncertainties. In this approach, those probabilities constitute the subjective probability based on the theoretical understanding of fusion neutron irradiation effects on materials to maximize the confidence in the estimated value.A case study of the Japanese DEMO breeding blanket is the basis for the discussion of the technical issues involved in the probabilistic approach. The breeding blanket design was simplified, the failure probability was estimated, and the impact of electromagnetic force on ferromagnetic material caused by a strong magnetic field induced by toroidal field coils was considered. The discussion of the impact of the irradiation damage is based on the irradiation data for F82H reduced-activation ferritic/martensitic steel. A direction for the development of fusion structural materials is suggested in terms of design technology and the statistical quality of the irradiation database.