Stochastic Modelling and Numerical Simulations of Radial Filament Motion and Average Profiles at the Boundary of Fusion Plasmas

Abstract

<p>The boundary region of magnetically confined plasmas is well known to have inherently fluctuating parameters. Advances in experimental measurements and numerical computations have revealed that the fluctuation-driven transport of particles and heat across the magnetic field lines is mainly dominated by the radial motion of blob-like filament structures. These structures, which are localized in the drift plane perpendicular to the magnetic field and elongated along the field lines, can deposit intermittent loads on the first wall of the fusion reactor leading to destructive plasma- wall interactions. As such, an accurate description of all the statistical properties of the intermittent fluctuations is necessary for both understanding and predicting its causes and effects on the reactor and walls. <p>Recently, a stochastic model for intermittent fluctuations has been constructed based on a super-position of uncorrelated pulses moving in the radial direction with a random distribution of amplitudes, sizes and velocities. Numerical simulations of this model are utilized to investigate the implication of distribution and correlation between pulse sizes, velocities and amplitudes. The statistical analysis presented in this thesis includes lower-order statistical moments, probability density functions, auto-correlation functions and power spectral densities. <p>The results explain the mechanisms behind broad and flat time-average radial profiles of particle density in the scrape-off layer, the increase of the relative fluctuation amplitude as well as skewness and flatness in the far scrape-off layer. This proves valuable for describing intermittent fluctuations in the boundary region of magnetized plasmas.