We present state-of-the-art computations of propagation and absorption of electron cyclotron waves, retaining the effects of scattering due to electron density fluctuations. In ITER, injected microwaves are foreseen to suppress neoclassical tearing modes (NTMs) by driving current at the q = 2 and q = 3/2 resonant surfaces. Scattering of the beam can spoil the good localization of the absorption and thus impair NTM control capabilities. A novel tool, the WKBeam code, has been employed here in order to investigate this issue. The code is a Monte Carlo solver for the wave kinetic equation and retains diffraction, full axisymmetric tokamak geometry, determination of the absorption profile and an integral form of the scattering operator which describes the effects of turbulent density fluctuations within the limits of the Born scattering approximation. The approach has been benchmarked against the paraxial WKB code TORBEAM and the full-wave code IPF-FDMC. In particular, the Born approximation is found to be valid for ITER parameters. In this paper, we show that the transport in ITER is diffusive unlike in present experiments, thus causing up to a factor of 2 to 4 broadening in the absorption profile. However, the broadening depends strongly on the turbulence model assumed for the density fluctuations, which still has large uncertainties.

A finite-difference time-domain code is used to obtain the full-wave solution of the O-X mode conversion process for typical parameters of the TJ-II stellarator in a cylindrical geometry. This reduction of the complicated stellarator geometry to a cylindrical geometry is chosen since the conversion process occurs only over a limited radial plasma volume. In the calculations, Gaussian antenna beams are studied with the option of different beam waists in the poloidal and toroidal direction. Optimum conversion efficiency is found if the wavefront of the incident antenna beam is matched to the local curvature of the O-X conversion layer. Finally, the code is used to calculate the complete O-X-B conversion process into a Bernstein wave.

Abstract. It is not fully understood how electromagnetic waves propagate through plasma density fluctuations when the size of the fluctuations is comparable with the wavelength of the incident radiation. In this paper, the perturbing effect of a turbulent plasma density layer on a traversing microwave beam is simulated with full-wave simulations. The deterioration of the microwave beam is calculated as a function of the characteristic turbulence structure size, the turbulence amplitude, the depth of the interaction zone and the size of the waist of the incident beam. The maximum scattering is observed for a structure size on the order of half the vacuum wavelength. The scattering and beam broadening was found to increase linearly with the depth of the turbulence layer and quadratically with the fluctuation strength. Consequences for experiments and 3D effects are considered.

In the stellarator TJ-K, overdense low-temperature plasmas are created by means of microwaves at 2.45 GHz. Extensive studies have been carried out to understand the heating process. The plasma breakdown at the cyclotron resonance layer has been directly observed with a multiple Langmuir probe array. Profile measurements indicate power deposition at the plasma boundary, where the upper hybrid resonance (UHR) is located. This result is confirmed by full-wave simulations which emphasize the importance of the vacuum vessel to increase the absorbed microwave power due to multiple reflections. Further indications for heating at the UHR layer are found by measurements of the wave electric field of the incident microwave and by power-modulation experiments. In contrast to similar experiments, no indication for heating by electron Bernstein waves was found.

Simulations using 3D and 2D full-wave codes have shown that edge filaments in tokamak plasmas can significantly affect the propagation of microwaves across a broad frequency spectrum, resulting in scattering angles of up to 46 • . Parameter scans were carried out for density perturbations comparable in width and amplitude to MAST filaments and the effect on the measured emission was calculated. 3D effects were discovered in the case of an obliquely incident beam.In general, the problem of EM propagation past wavelength-sized 3D inhomogeneities is not well understood, yet is of importance for both heating and diagnostic applications in the electron cyclotron frequency range for tokamaks, as well as atmospheric physics. To improve this understanding, a new cold-plasma code, EMIT-3D, was written to extend full-wave microwave simulations in magnetized plasmas to 3D, and make comparisons to the existing 2D code IPF-FDMC. This work supports MAST experiments using the SAMI diagnostic to image microwave emission from the plasma edge due to mode conversion from electron Bernstein waves. Significant fluctuations in the SAMI data mean that detailed modelling is required to improve its interpretation.