Study of Cs transport in SPIDER plasma using a test-particle Monte Carlo code

Large negative ion sources are used in heating neutral beam for additional heating in fusion devices, when a high particle energy is required. Surface processes for negative ion production are enhanced by caesium evaporation in the plasma chamber. The low work function of the alkaline metal increases the probability of H- (D-) ions production, which are mostly produced via the surface conversion of positive and neutral particles. The energy of these H- precursors influences beam properties, such as beam divergence and uniformity. Tracking the caesium particle trajectories and predicting their distribution on the source walls and grids, hence, is a key task to optimize H- production, extraction and acceleration; on the other hand, the amount of injected caesium shall be minimized, so that only a limited quantity leaks to the accelerator where it causes the production of secondary charges and an additional breakdown probability. In this thesis work, a Monte Carlo-based code written in a C++ framework is updated to trace caesium transport in SPIDER, which is the full size ion source prototype of the ITER injectors. Starting from a model developed for tracking positive ions, the whole experiment is simulated, considering electrical and magnetic fields, and treating elastic and inelastic collisions. The updated version of the code derives caesium particle trajectories in the SPIDER plasma as a function of experimental conditions. Combining caesium transport with the main sputtering processes leads to an estimation of the caesium distribution on the plasma chamber surfaces, studying its uniformity as a function of space and time. A model for caesium deposition and redistribution was developed in MATLAB, with the aim to determine the optimal caesium coverage in terms of uniformity and amount. The code uses the transport matrices output from the C++ code to calculate the array of arrival points for caesium particles in the SPIDER mesh. The code includes also a virtual measurement of the caesium density made by Laser Absorption Spectroscopy diagnostics. A visual representation of the caesium coverage time evolution is possible using the output of the code in Paraview software. Finally, experimental measurements for power calibration of the work function diagnostics were performed at CATS (Consorzio RFX) using a light power sensor. The aim of this diagnostics will be the measurement of the sample work function of SPIDER molybdenum surfaces coated with caesium by using the Fowler Method. In summary, the thesis work approached the study of caesium dynamics with computational as well as experimental methods, outlining the future steps for a predicting model for ceasium transport inside negative ion sources.