The Fusion Nuclear Science Facility (FNSF) is an intermediate step in the path to commercial fusion energy that will accommodate the extreme fusion nuclear environment and the complex integration of components and their environment as well as the relevant nuclear science and plasma physics. The transient thermal and electromagnetic loads on plasma-facing components in FNSF have been shown to offer significant design challenges that are difficult to meet with solid walls. Hence, the project team is investigating the feasibility of using liquid walls to ameliorate some of the risk associated with solid wall designs.
In this paper, we examine the effects these transient loads will have on a liquid wall. Mass loss is considered using standard evaporation models accounting for transient surface temperatures. The heat transfer is modeled with a one-dimensional transient conduction model that accounts for evaporative losses. No liquid motion is considered. Loss rates of tens of microns per edge-localized mode (ELM) are predicted. Peak heat fluxes are treated parametrically to help address the substantial uncertainty inherent in models for the timing and spatial distribution of the heat deposited during the ELM. Boiling is considered but is found to not be of consequence, as the temperatures required for homogeneous nucleation of bubbles are substantially higher than a conventional boiling point. It should be noted that all evaporation calculations are for evaporation into a vacuum. In the future, we intend to incorporate these evaporation rates into an edge physics code to self-consistently model the net mass flows at the liquid surface in a tokamak.
Electromagnetic effects due to ELMs and disruptions are accounted for by assuming a stationary plasma quench. ELMs are addressed assuming a small fluctuation in the plasma current during an event, while disruptions are addressed assuming a full quench of the current. The variation in the plasma current induces currents in the conducting fluid, leading to forces on the liquid (and subsequent motion). A commercial finite element code is used to calculate the induced currents and forces associated with a static liquid divertor. Liquid motion is not considered in this calculation, so no magnetohydrodynamic (MHD) currents are addressed, but a simplified model is presented to estimate the impact of these currents on the liquid motion. Based on these calculations, the acceleration of the liquid is expected to be quite high, and containment of the liquid is likely not possible. The MHD effects appear to be relatively minor.