Simulation and validation of MHD benchmark problems using ALMA

High resolution simulations are well known for their predictive capability in liquid metal cooling applications such as liquid metal blankets. ALMA (Anti- symmetric, Large-moment, Accelerated) is a high-performance computational engine designed to solve plasma and neutral gas models on hybrid supercomputers (Halpern et al, 2021). Since the anti-symmetric formulation conserves quadratic invariants of the system (such as energy and enstrophy), the solver is robust, numerically stable and easily scalable while tackling high resolution, multi-variate and unsteady neutral MHD flows. To demonstrate the solver capabilities in liquid metal blanket applications, benchmark problems proposed by Smolentsev et al [2015] are simulated. Firstly, the fully developed flow of a conducting liquid in a rectangular duct subject to uniform transverse magnetic field is considered. Insulating walls and walls of finite conductance are chosen and the flow profile is validated for a range of Hartmann numbers (varying from 500 to 15000). Secondly, the liquid metal flow in square duct subject to a spatially varying magnetic field is chosen. The velocity profile and pressure drop across the duct were validated with the experimental data and analytical expressions.