On all-FLR order numerical modelling of RF wave propagation and damping

Dirk Van Eester,Ernesto Lerche,J Hillairet

doi:10.1051/epjconf/201715703060

Abstract

Determining the fate of waves in hot magnetized plasmas in magnetic confinement fusion machines close to arbitrary cyclotron harmonics retaining finite Larmor radius effects requires solving the relevant integro-differential wave equation for waves of arbitrary wavelength. Anticipating exploitation in as realistic geometry as possible this requires massive computer resources. Work is ongoing to attempt reducing the required CPU memory and time. As Morlet wavelets are localised in x-space as well as k-space, they potentially offer a means to solve the relevant wave equation at reduced CPU requirement cost. Encouraging first results are obtained.

Highlights

Ion cyclotron resonance heating (ICRH) is a routinely used method to bring plasmas to fusion relevant temperatures in magnetic confinement fusion machines [1]
A new solver is under development to help understand the ICRH wave propagation and damping in multiple dimensions
It is inspired on the basic philosophy pioneered by Jaeger [2] for the AORSA code, and intends to provide a rigorous computation of the wave polarisation at arbitrary cyclotron harmonics

Summary

Introduction

Ion cyclotron resonance heating (ICRH) is a routinely used method to bring plasmas to fusion relevant temperatures in magnetic confinement fusion machines [1]. More often than not simplifications are made, first of all by truncating the dielectric tensor at the leading order terms in finite Larmor radius (FLR) corrections and secondly by simplifying the geometry. A new solver is under development to help understand the ICRH wave propagation and damping in multiple dimensions. It is inspired on the basic philosophy pioneered by Jaeger [2] for the AORSA code, and intends to provide a rigorous computation of the wave polarisation at arbitrary cyclotron harmonics. The key idea at this initial exploratory stage is to assess the importance of higher order finite Larmor radius corrections while keeping the zero order distribution functions simple (Maxwellian) and the geometry prescribed

Finite Larmor radius effects

Wavelet approach

Hot plasma wave equation

Numerical example

Final remarks