Development of a high-fidelity multi-physics simulation tool for liquid-fuel fast nuclear reactors

The Molten Salt Reactor (MSR) is one of the six Generation-IV nuclear reactor designs. It presents very promising characteristics in terms of safety, sustainability, reliability, and proliferation resistance. Numerous research projects are currently carried out worldwide to bring this future reactor technology to a higher maturity, and in Europe efforts are focused on developing a fast-spectrum design: the Molten Salt Fast Reactor (MSFR).

Numerical simulations are essential to develop MSR designs, given the scarce operational experience gained with this technology and the current unavailability of experimental reactors. However, modeling an MSR is a challenging task, due to the unique physics phenomena induced by the adoption of a liquid fuel that is also the coolant: transport of delayed neutron precursors, strong negative temperature feedback coefficient, distributed generation of heat directly in the coolant. Moreover, the geometry of the core cavity of fast-spectrum designs often induces complex three-dimensional flow effects. For these reasons, legacy codes traditionally used in the nuclear community often prove unsuitable to accurately model MSRs, in particular fast-spectrum designs, and must be replaced by dedicated tools.

This thesis presents the development of one of these multi-physics codes, which aims at accurately modeling the three-dimensional neutron transport, fluid flow, and heat transfer physics phenomena characterizing a fast-spectrum liquid-fuel nuclear reactor. The coupling is realized between an incompressible Reynolds-Averaged Navier-Stokes model and a discrete ordinates neutron transport solver, both based on a discontinuous Galerkin Finite Element space discretization which guarantees high-quality of the solution.

As the research was carried out in the context of the Euratom SAMOFAR project, the MSFR is taken as reference case study. We extensively analyze its behaviour at steady-state and during several transient scenarios, assessing the safety of the current design and thus deriving useful information on its further development.