Development of a simulation tool for MHD flows under nuclear fusion conditions

Resumen

In Nuclear Fusion Technology, MHD flows can be encountered in liquid metal (LM) breeding blankets, the part of a fusion reactor where tritium, one of the fusion fuels, is to be produced. There are several types of LM breeding blankets, which can be classified according to the fraction of the thermal load extracted by the LM. Such classification provides valuable information on liquid metal flow properties. For instance, if no heat removal is carried out by the LM, its velocity can be quite low, what makes buoyancy the predominant force in front of inertia. The flow inside breeding blanket channels can be very complex, particularly in those blanket types where buoyancy plays a relevant role. The understanding of the flow nature, including the possible instabilities that might appear, the exact knowledge of flow profiles for tritium control purposes, and the prediction of thermal fluxes for thermal efficiency analysis are of great interest for blanket design optimization. In this direction, a thermal-MHD coupled simulation tool has been implemented in the OpenFOAM toolkit. The resultant code can be understood as a preliminary predictive tool for liquid metal breeding blanket channel design. The developed code is a transient 3D tool that accounts for thermal-MHD coupling and can deal with several layers of materials. Various MHD modeling strategies have been studied, starting with the implementation of an induced magnetic field formulation and continuing with an electric potential formulation based on the low magnetic Reynolds approximation, in this case using the conservative formula of the Lorentz force proposed by Ni et al. (2007). Two pressure-velocity couplings have been analyzed. The first one is based on a projection method whereas the second one, which has proved to be more robust, follows a PISO-like algorithm (Weller et al. 1998). The thermal coupling has been achieved by means of the Boussinesq hypothesis. The developed tool accounts for the linear wall function for Hartmann boundary layers from Leboucher (1999), which reduces substantially the CPU time of the simulations. The code also accounts for fluid-solid thermal and electrical coupling by means of implicit coupling of fluid and solid grids. Special attention has been placed in correctly coupling liquid-solid energy transport equations by means of the conservative form of the equations in both domains. All along the development process, validation steps have been carried out with successful results. An alternative thermal-MHD tool has also been implemented following the 2D approach from Sommeria and Moreau (1982). Such code accounts for the 0-equation Q2D turbulence RANS model from Smolentsev and Moreau (2006). Three application cases are considered. In the first case, the integrated effect of volumetric heating and magnetic field on tritium transport in a U-bend flow, as applied to the EU HCLL blanket concept, is studied. The second application case corresponds to the thermal analysis of the blanket design that is being developed in the framework of the Spanish National Project on Breeding Blanket Technologies TECNO_FUS (through CONSOLIDER-INGENIO 2010 Programme). The third and last case includes the instability analysis of a pressure-driven MHD flow in a horizontal channel with a constant thermal load. The application cases have not only shown the code capabilities to simulate liquid metal channels in breeding blankets but, also, have provided a useful know-how on flow properties inside those channels.