Spatial dispersion in finite element models for ion cyclotron resonance heating
Theory and applications for toroidal plasmas
Time: Wed 2025-01-29 09.00
Location: F3 (Flodis), Lindstedtsvägen 26 & 28, Stockholm
Video link: https://kth-se.zoom.us/j/67880732648
Language: English
Subject area: Electrical Engineering
Doctoral student: Björn Zaar , Elektromagnetism och fusionsfysik
Opponent: Senior researcher Remi Dumont, Commisariat à l'energie atomique et aux énergies alternatives (CEA), Paris, France
Supervisor: Associate professor Thomas Jonsson, Elektromagnetism och fusionsfysik; Professor emeritus Jan Scheffel, Elektromagnetism och fusionsfysik
QC 20241230
Abstract
Nuclear fusion can provide large amounts of energy from earth-abundant elements,with no carbon emissions and little radioactive waste. For the nuclei to fuse under earth-relevant conditions, temperatures in excess of 100 000 000 °C are needed. At these temperatures, the fuel is in a plasma state. A common method to heat the plasma is ion cyclotron resonance heating (ICRH), where radiofrequency waves are launched from an antenna on the vessel wall into the plasma to resonate with the gyrating ions. Wave propagation and dissipation in hot magnetized plasmas is a nonlocal process, where the plasma response at a given point depends on the particles' cumulative acceleration along their orbits. To quantify how the plasma is heated, numerical simulations are required. This thesis aims to provide a numerical framework that can simulate the coupling of the wave from the antenna to the plasma, the wave propagation and dissipation inside the plasma, as well as the acceleration of individual ions and how they deposit their energy in the plasma.
To this end, an iterative scheme that adds nonlocal effects to an otherwise local finite element (FE) model is developed. FE models are suitable for modeling irregular geometries and wave coupling through the cold scrape-off layer plasma, but not necessarily the hot core plasma. Examples of nonlocal effects that are added iteratively are mode conversion from the fast magnetosonic wave to the ion Bernstein wave (IBW) and up- and downshift of the parallel wavenumber. Further, the wave solver is coupled to a Fokker-Planck solver that evaluates the effect of ICRH on the ion distribution function. The models presented in this thesis are in 1D or 2D axisymmetry, but are not conceptually different from a generalization to 3D.