ASTRID is a project for an industrial prototype of a 600 MWe sodium cooled Fast Reactor, led by CEA. An important program is in progress for the development and the validation of numerical tools for the simulation of the dynamic mechanical behavior of the Fast Reactor cores, with both experimental and numerical parts. The cores are constituted of Fuel Assemblies (of FA) and Neutronic Shields (or NS) immersed in the primary coolant (sodium), which circulates inside the Fluid Assemblies. The FA and the NS are slender structures, which may be considered as beams, form a mechanical point of view.

The dynamic behavior of this system has to be understood, for design and safety studies. Two main movements have to be considered: global horizontal movements under a seismic excitation, and opening of the core. The fluid leads to complex interactions between the structures in the whole core. The dynamic behavior of the core is also strongly influenced by contacts between the beams and by the sodium, which limit the relative displacements. Numerical methods and models are built to describe and simulate this dynamic behavior. The validation of the numerical tools is based on the results of different experimental programs, already performed or in progress.

The paper is mainly devoted to the modeling of the Fluid Structure Interaction phenomena in the Fast Reactor cores.

Tubes bundles immersed in a dense fluid are very common in the nuclear industry (reactor cores and steam generators). In the case of an external excitation (earthquake or shock) the presence of the fluid leads to “inertial effects” with lower natural frequencies, and “dissipative effects”, with higher damping. The geometry of a tubes bundle is complex, which may lead to very huge sizes for the numerical models. Many works have been made during the last decades to develop homogenization, in order to simplify the problem.

Theoretical analyses are presented on different simplifications and assumptions which can be made in the homogenization approach. The accuracy of the different assumptions depends of the conditions of the system: fluid flow or fluid at rest, small or large displacements of the structure.

In the general case, it is theoretically necessary to consider the Navier Stokes equations: the fluid flow is fully nonlinear. Models have been developed during the last years, based on the Euler linear equations, corresponding to a fluid at rest, with small displacements of the structure. Only the inertial effects are theoretically described but the dissipative effects may be taken into account by using a Rayleigh damping.

Different theoretical analyses show that, even in the case of a nonlinear fluid flow, the linear potential flow models may be used as linear equivalent models. In the cases with an important head loss in the fluid flow through the tubes, the fluid movement is mainly driven by the important forces exchanged with the structure and by the pressure gradient. The global equations of the system are close to the equations used for porous media, like the Darcy equations.

An important condition to get a relevant model is to describe globally the energy balance in the system. The energy given to the fluid by the solid correspond to a variation of kinetic energy in the fluid and to energy dissipation in the fluid. Attention will be paid to the cases where the tubes bundle is in interaction with free fluid, without tubes. The global equation of the system has to be accurate for the tubes bundle and for the free fluid also.