MHD turbulence models for fusion reactor blankets

Turbulence models, considering the effect of magnetohydrodynamics (MHD), were derived for canonical flows and can't predict complex engineering flows. For this reason, most researchers either consider the quasi-two-dimensional approximation when appropriate or perform high-resolution DNS and LES of MHD turbulence. However, with continued interest in fusion reactors, timely predictions of MHD turbulent heat transfer in fusion blankets and vacuum vessel cooling channels are crucial for design. In fact, the superconducting magnets used to confine the plasma affect the turbulent fluid flow through Lorentz forces. In this work, we implement MHD turbulence in the open-source code OpenFOAM. First, the laminar MHD solver is improved and benchmarked with existing data. Next, existing MHD RANS turbulence models are incorporated and compared against commercial codes such as HIMAG. The fidelity of these models is tested in simplified geometries of conceptual fusion reactor designs like the ARC reactor from Commonwealth Fusion Systems. It is found that MHD turbulence significantly hinders heat transfer in fusion blankets, and robust 3D models are required for accurate predictions and design improvements.