Redirection of radio-frequency power flow by filaments

Abstract

When radio-frequency waves propagate through magnetized plasmas, they scatter at density filaments, which are usually aligned with the magnetic field lines. We show that this scattering can redirect part of the power from the perpendicular direction to the parallel direction (along the field lines, along the filament). In magnetic confinement fusion devices, such as tokamaks, plasma is often heated by Ion Cyclotron Range of Frequency (ICRF) antennas. For such antennas, our simulations show that a parallel Poynting flux in the level of 1 MW/m 2 can be generated by this power redirection mechanism. The increase of the magnitude of electric and magnetic fields of the wave in the filament is responsible for the parallel redirection of power flow. Simulations with constant density in the filament and background plasma show that the dominant parallel wavenumbers of the power spectrum in the filament are increased and additional wavenumbers can be generated. In the cases with an experimental density profile, only the power spectrum inside the filament is significantly affected while the rest remains almost the same.

The importance of realistic plasma filament waveforms for the study of resonant wave-filament interactions in tokamak edge plasmas

“…Ram et al 1 pointed out circumstances in which the filament "behaves like a transmission line with power flowing in its interior" (i.e. in the parallel direction), a phenomenon that was also observed numerically in 3D Finite Element simulations 7 . Extreme versions of this phenomenon are sometimes described as "mode conversion resonances" 6 because the mode conversion can be resonant i.e.…”

“…In this paper we considered predictions regarding resonant ICRF wave-filament interactions in tokamak edge plasmas, based on analytic lines of reasoning such as the Mie series 1,2,6,7,9 , and compared them with results of Finite Element calculations with a more realistic plasma filament waveform. Qualitatively, many of the analytic predictions remain valid in the more realistic cases considered here, especially those concerning the conditions under which mode conversion resonances are expected to be common (both the prediction that they occur under NSTX-like high harmonic fast wave conditions, and the prediction that they do not occur under AUG-like conditions, are confirmed numerically).…”

“…( 22) predicts resonances close to n F n B = 1 when ξ is small (small k ) and ω Ω i is large (high harmonic heating), that is, operating parameters common in NSTX but uncommon in most other tokamaks. Indeed, 3D numerical studies on AUG 7 we believe resonant wave-filament interactions to be among the factors responsible for the experimentally observed parallel RF power losses on NSTX 8 . This is the reason why we are especially interested in NSTX-like parameters.…”

“…One possible adverse effect of this mode conversion, even under generic conditions, is that increased |E | may increase sheath rectification, and thereby impurity production, where a filament intersects a plasma-facing component surface 7,28 .…”

Recent progress in modeling ICRF-edge plasma interactions with application to ASDEX Upgrade

“…A full tokamak geometry, where the filament can be traced all the way to the divertor, is rarely feasible to simulate, and even then, perfect electric conducting BCs at the divertor would not be appropriate [19]. The use of perfectly matched layers [97] to toroidally terminate the filaments, as in [98], requires the surface waves to be toroidally forward [94,99].…”