Dependence of transport in high-beta low collisional H-modes on ExB-shear and q-profile

Thermonuclear fusion power plants could at some point be an attractive way for humanity to harness energy in a sustainable, reliable and safe manner. One promising design for a fusion reactor is the tokamak, in which plasma is confined by helically twisted magnetic field lines in a toroidally shaped vessel. A limiting factor on the core temperature that can be achieved in such a reactor – and therefore also on the power that can be produced in it – is turbulence, since it can cause a significant transport of heat out of the plasma. In previous experiments, temperatures have been observed in the core of tokamaks, that exceed what state of the art numerical models of turbulent transport would predict. In this thesis, experimental and numerical studies are presented, that investigate what effects are responsible for this observed suppression of turbulence. One possible cause are sheared flows in poloidal direction that decorrelate turbulence structures. The cause for this so-called E×B-shear is a drift in the plasma created by radial electric fields Er. In this work, experiments have been conducted that vary the E×B-shear, by changing the heating sources of the plasma such that its rotation – the main cause for the radial electric field in the core – is reduced. Since this change in heating mix also affects other potentially relevant parameters, non-linear simulations with the gyrokinetic code GENE are performed, to disentangle between these effects. Both in the experiments and simulations, the E×B-shear was not found to play a role in observed reduction of turbulent transport in the plasma core. Instead, in further experiments a strong dependence on the so-called q-profile – a measure for the helicity of the magnetic field lines – became evident. By elevating q, and shaping it such that the magnetic shear s – a normalized gradient of q – reaches more strongly negative values in the core, the ion temperature profiles become more peaked. As with the investigations regarding the E×B-shear, nonlinear GENE simulations were performed to disentangle effects of the q-profile from other potentially competing effects. More detailed studies to determine which aspects of the q-profile are responsible for the reduction of transport did unfortunately not lead to a conclusive result, since it is challenging to measure the q-profile with high precision. Nevertheless, from the limited amount of experiments where a variety of q-profiles could be determined with great precision, some general trends could be observed: For one, experiments seem to indicate that transport is reduced near locations where s exhibits negative values; Theory explains this with turbulent structures being bent towards the high-field side, which is inherently stable towards turbulence. Furthermore, the experiments are in general consistent with numerical studies that find the reduction of transport to be dependent on the values of q and s. The mechanism here is that the instability causing turbulence interacts with other instabilities driven by suprathermal ions and transfers energy into them. Through these interactions, energy is siphoned away from the main instability, reducing the turbulence. These fast ion driven instabilities are sensitive to the exact values of q and s. Based on these findings, a heuristic model has been implemented in the transport code TGLF, that allows it to better model experiments for which it otherwise strongly overestimates turbulent transport. This not only paves a path to be able to better study and develop such scenarios with reduced transport, it also sheds some further light on the role q, s and fast ions have on the stabilization of turbulence.