Impact of Wavelength Scale Density Variation on Microwave Propagation in Tokamak Plasmas

The propagation of microwaves through magnetised plasmas in the presence of wavelength scale density variation poses an interesting physics problem. Microwaves have many uses in tokamaks, from diagnostics that help characterise the plasma to high-power beams used for heating and current drive. It is therefore important to be able to accurately predict the path that they will take in tokamak plasmas, even in the presence of fluctuations.

To this end, a full-wave cold-plasma code utilising the FDTD method (EMIT) has been developed both in 2D and 3D. EMIT-2D minimises computational cost, allowing full simulation of the beam from the antenna to the absorption region. A benchmark of the code was carried out before it was applied to the problem of OX-mode conversion. In plasmas with steep density gradients, mode conversion efficiency was found to decrease sharply due to the converted X-mode tunnelling back out of the plasma.

EMIT-2D was also used in a study of ECRH beam broadening by turbulence on DIII-D. Significant beam broadening was measured experimentally in three operating scenarios. Diagnostic data was used to generate synthetic turbulent density profiles for simulations. The simulations agreed with experiment, providing a direct comparison between simulation and experimental measurements of beam broadening for the first time, but diagnostic uncertainty led to significant uncertainty in the simulated results, motivating the need for future turbulence diagnostics of better spatial resolution.

To further characterise how beam broadening by electrostatic turbulence depends on plasma and beam parameters, a series of parameter scans were carried out covering tokamak relevant parameter ranges. The parameter scans were conducted in pairwise combinations of the parameters in order to determine the separability of the dependencies, and an empirical formula was found for fusion-relevant scenarios allowing the prediction of beam-broadening in microseconds instead of the hours required for full-wave simulations.