Real-Time Control of Sawteeth and NTMs in TCV and ITER

The main goal of this thesis is to demonstrate the capability of magneto-hydrodynamic (MHD) instability control, particularly sawteeth and neoclassical tearing modes (NTMs), in order to achieve high performance operation. Experiments and simulations have been carried out to pursue this purpose on different tokamaks: TCV (in Switzerland), KSTAR (in Korea), AUG (in Germany) and ITER (in construction in France). Each tokamak has different features such as machine size, heating systems, operation scenarios and energy confinement time scale, all the tokamaks are equipped or will be equipped with an electron cyclotron heating/current drive (ECH/ECCD) system for plasma heating/current drive and control of MHD instabilities. Therefore, this work focusses on the feasibility of using the localised ECH/ECCD beams to control the instabilities; sawteeth and NTMs. For the experimental part, sawtooth and NTM control experiments have been carried out. In TCV, novel ways of sawtooth period control - sawtooth pacing and locking - have experimentally been demonstrated using the TCV real-time control system. Based on the successful application of these methods to sawtooth control in TCV, we have next focussed on the extension of these new sawtooth period control methods to other tokamaks: KSTAR and AUG. In the 2013 KSTAR experimental campaign, the applicability of sawtooth locking using EC power modulation has been tested for sawtooth period control in the presence of fast particles generated by neutral beam injection (NBI). The KSTAR real-time control system was not ready for sawtooth pacing thus only locking has been examined. These preliminary KSTAR experimental results have shown the possibility of sawtooth period control using sawtooth locking, although proper locking was not obtained yet. The locking parameters would still need to be adjusted for single period locking to occur. In order for the investigation of the capability of sawtooth locking on KSTAR tokamak to be complete, more experiments with different locking parameters should be carried out. The sawtooth locking technique has also been applied to AUG plasmas. As in the KSTAR tokamak, the real-time control for sawtooth pacing was not available, thus sawtooth locking has been tested. The AUG plasmas were more complicated compared to TCV and KSTAR cases due to the fast particles effect on the evolution of sawtooth from both NBI and ion cyclotron heating (ICH). Sawteeth did not lock to the EC modulation in AUG experiments, though in some discharges they became somewhat more regular. However, the application of sawtooth locking to the AUG tokamak has been well initiated and more experiments will follow to understand better the behaviour of sawteeth and to determine the sawtooth locking range. In addition, sawtooth control was demonstrated and used in other experiments studying the role of sawteeth on impurity transport. Concerning the NTM control experiments, we have focussed on the enhancement of the NTM