High-fidelity numerical investigation of liquid metal magnetoconvection

As global energy demands continue to rise, the search for sustainable, low carbon energy sources has become increasingly urgent. Nuclear fusion represents a promising solution, offering the potential for safe, virtually limitless energy with minimal environmental impact. Liquid Metals, particularly lithium (Li) and its alloys, have emerged as strong candidates for cooling in nuclear fusion reactors due to their high thermal conductivity, low vapour pressure, and ability to breed tritium (critical fuel for sustaining fusion reactions). However, the behaviour of liquid metals under the combined influence of heat loads and magnetic field, a phenomenon known as magnetoconvection, introduces complex fluid dynamics that differ significantly from those observed in conventional fluids.

Magnetoconvection arises from the interplay between thermal convection and electromagnetic forces in electrically conductive fluids. In fusion reactors, liquid metals experience significant Lorentz forces due to their motion within magnetic fields, leading to highly non-uniform flow distributions, altered turbulence structures, and even flow laminarisation under certain conditions.

These effects can drastically impact the efficiency of heat transfer and the overall performance of the cooling system. The interaction of buoyancy forces with magnetic fields further complicates the flow behaviour, creating regimes that range from laminar flows to unsteady mixed convection, depending on the specific operating conditions.

Despite extensive research on Magnetohydrodynamics (MHD) and mixed convection phenomena separately, the combined effects of these forces in liquid metals remain poorly understood. This gap in knowledge poses a significant challenge to the optimisation of all applications within the field of Liquid Metal (LM) magnetoconvection.

This dissertation aims to advance the understanding of magnetoconvection in liquid metals through the development and application of high-fidelity numerical methods. By solving the MHD equations, which combine the Navier-Stokes and Maxwell equations, using Large Eddy Simulation (LES) paradigm, this research provides detailed insights into the fluid dynamics of liquid metals.

The simulations are conducted using a customised module of the opensource software OpenFOAM, specifically developed for this project. The findings of this research aim to inform the design and optimisation of cooling systems in future thermonuclear reactors. By enhancing our understanding of magnetoconvection, this work may provide valuable insights that could assist in selecting materials and refining operating parameters to improve heat transfer efficiency under challenging conditions. While the results contribute to addressing some of the complexities involved in liquid metals magnetoconvection, further research and validation will be necessary to fully assess their implications for the practical development of fusion as a viable energy source.