Experimental Search for Signatures of Zonal Flow Physics in a Large Spherical Tokamak using Beam Emission Spectroscopy

Cross-field turbulent transport in magnetically confined plasmas is a highly effective loss mechanism of the heat and mass of fusion fuels and principally responsible for degrading the quality of confinement. Recently, an inter-dependence of turbulence and flows has been demonstrated in conventional tokamaks, including zonal flows that take energy directly out of turbulence. Data, in particular an experiment in a spherical tokamak, is scarce. This project uses a Beam Emission Spectroscopy (BES) diagnostic on the MAST spherical tokamak to measure local fluctuation of plasma density. We develop correlation and spectral analysis techniques to facilitate investigation of flow structure across the plasma profile. I first developed a velocimetry routine using cross-correlation time delay estimation (CCTDE) optimising the length of correlation functions of fluctuating MAST BES data to achieve high time resolution for zonal flow detection. Tests using surrogate data were included to improve the accuracy and reliability of the code, achieving up to a 98% successful measurement rate in broadband data. The new code’s performance was benchmarked using three realistic examples of spherical tokamak physics that previously had made velocimetry unusable. Each had an imposed mode mimicking zonal flow shown to be observable and measurable in my velocity spectra. Results from this spectroscopy identify four classes of spectra observed in real BES data. Spectra with the clearest coherent peaks are shown to exist concurrently with a long-lasting magnetic mode. A systematic test of zonal flow physics at shot times that produce velocity spectra with coherent peaks was developed. Automating and weighting scores for those tests created a framework for GAM detection; adaptable as expectations of GAM physics change. In this project it enabled the first search of a substantial MAST data set of 63 shots in which BES observed the plasma edge. No examples matched all expected physics. The observed spectrum patterns match GAM theory that accounts for safety factor and plasma rotation, but not elongation; the modal average extent of matches covers four columns (about 8 cm plasma radius) of the detector. A scan of 209 shots at L-H transition times found 5 cases with good matches to expected zonal flow physics but relatively weak velocity modes.