An experimental and computational study of tokamak plasma turbulence

Turbulence driven by small-scale instabilities results in strong heat and particle transport, which significantly shortens the confinement time and prevents the formation of a self-sustained plasma reaction in magnetic confinement devices. Control and possible mitigation of this transport is critical to achieve controlled fusion, which requires a proper understanding of the underlying physics of turbulent processes.

This thesis explores both computational and experimental methods to study turbulent fluctuations, focusing on modelling and the use of the Tangential Phase Contrast Imaging (TPCI) diagnostic that measures electron density fluctuations. On the experimental side of this work, the TPCI diagnostic is upgraded with new equipment to perform ion scale measurements. The diagnostic is used in an initial study of turbulence when the shape of the plasma changes from positive to negative triangularity in a diverted configuration, and to study the effects of an electron cyclotron beam on turbulence.

However, the main part of the thesis is devoted to gyrokinetic GENE simulations of turbulence and modelling of TPCI signals. The simulated density fluctuations are post-processed with a synthetic diagnostic that has been developed to model measurements from TPCI, taking into account the effect of a spatial filter for selecting the measured wave-vector directions. Modelling, with GENE and the synthetic diagnostic, is performed for several TCV plasma scenarios with the aim of improving the interpretation of the TPCI signals. The synthetic diagnostic is used to establish a link between the TPCI signals and the underlying modes, and to find new ways to apply the TPCI diagnostic to measure Ion Temperature Gradient (ITG) and Trapped Electron Mode (TEM) turbulence. It is also demonstrated how computationally cheap linear simulations can be used to predict certain aspects of the TPCI signals.

In the final part of this work a computational study of turbulent fluctuations in the JT-60SA machine is also carried out, motivated in part by the concomitant design of a TPCI diagnostic for that device. Turbulence in a representative, planned, high-performance JT-60SA plasma discharge, scenario 1, is modelled with local gyrokinetic flux-tube simulations. The discharge features a double-null separatrix, 41 MW of combined Neutral Beam Heating (NBH) and Electron Cyclotron Heating (ECH), and a high predicted ratio β of the plasma kinetic to magnetic pressure. Electromagnetic effects, in particular compressional magnetic field fluctuations, are essential in this scenario due to the large β. A detailed linear and nonlinear study of the fluctuations is performed and conclusions are drawn on the capability of reduced transport models to predict profiles in reactor-relevant conditions and on the projected performance of the JT-60SA machine. Finally, predictions are provided of the characteristics of the TPCI measurement, using the results from the gyrokinetic study and the