Modeling RF waves in hot plasmas using the finite element method and wavelet decomposition
Theory and applications for ion cyclotron resonance heating in toroidal plasmas
Time: Fri 2020-01-17 14.00
Subject area: Electrical Engineering
Doctoral student: Pablo Vallejos , Fusionsplasmafysik
Opponent: Philippe Lamalle, ITER Organization
Supervisor: Thomas Jonsson, Fusionsplasmafysik, Alfvénlaboratoriet; Torbjörn Hellsten, Fusionsplasmafysik
Fusion energy has the potential to provide a sustainable solution for generating large quantities of clean energy for human societies. The tokamak fusion reactor is a toroidal device where the hot ionized fuel (plasma) is confined by magnetic fields. Several heating systems are used in order to reach fusion relevant temperatures. Ion cyclotron resonance heating (ICRH) is one of these systems, where the plasma is heated by injecting radio frequency (RF) waves from an antenna located outside the plasma.
This thesis concerns modeling of RF wave propagation and damping in hot tokamak plasmas. However, solving the wave equation is complicated because of spatial dispersion. This effect makes the wave equation an integro-differential equation that is difficult to solve using common numerical tools. The objective of this thesis is to develop numerical methods that can handle spatial dispersion and account for the geometric complexity outside the core plasma, such as the antenna and low-density regions (or SOL). The main results of this work is the development of the FEMIC code and the so-called iterative wavelet finite element scheme.
FEMIC is a 2D axisymmetric code based on the finite element method. Its main feature is the integration of the core plasma with the SOL and antenna regions, where arbitrary geometric complexity is allowed. Moreover, FEMIC can apply a dielectric response in the SOL and in the region between the SOL and the core plasma (i.e. the pedestal). The code can account for perpendicular spatial dispersion (or FLR effects) for the fast wave only, which is sufficient for modeling harmonic cyclotron damping and transit time magnetic pumping. FEMIC was used for studying the effect of poloidal phasing on the ICRH power deposition on JET and ITER, and was benchmarked against other ICRH modeling codes in the fusion community successfully.
The iterative wavelet finite element scheme was developed in order to account for spatial dispersion in a rigorous way. The method adds spatial dispersion effects to the wave equation by using a fixed point iteration scheme. Spatial dispersion effects are evaluated using a novel method based on Morlet wavelet decomposition. The method has been tested successfully for parallel and perpendicular spatial dispersion in one-dimensional models. The FEMIC1D code was developed in order to model ICRH and to study the properties of the numerical scheme. FEMIC1D was used to study second harmonic heating and mode conversion to ion-Bernstein waves (IBW), including a model for the SOL and pedestal. By studying the propagation and damping of the IBW, we verified that the scheme can account for FLR effects.