Theory of ion temperature gradient (ITG) and trapped electron modes (TEMs) is applied to the study of particle transport in experimental conditions with central electron heating. It is shown that in the unstable domain of TEMs, the electron thermodiffusive flux is directed outwards. By means of such a flux, a mechanism is identified likely to account for density flattening with central electron heating. Theoretical predictions are compared with experimental observations in ASDEX Upgrade. A parameter domain (including L-and H-mode plasmas) is identified, in which flattening with central electron heating is observed in the experiments. In general, this domain turns out to be the same domain in which the dominant plasma instability is a TEM. On the contrary, the dominant instability is an ITG in plasmas whose density profile is not affected significantly by central electron heating. The flattening predicted by quasi-linear theory for low density L-mode plasmas is too small compared to the experimental observations. At very high density, even when the dominant instability is an ITG, electron heating can provide density flattening, via the coupling with the ion heat channel. In these conditions the anomalous diffusivity increases in response to the increased ion heat flux, while the large collisionality makes the anomalous pinch small and the Ware pinch important.

This paper analyses the properties of a critical gradient transport model based on a few assumptions: electrostatic gyroBohm scaling law, existence of an instability threshold and finite background transport below the threshold. A quantitative criterion of stiffness is proposed, which provides the means for a quantitative assessment and inter-machine comparison. It is also shown that this transport model is compatible with a two-term scaling law of global confinement, as proposed recently by the International Tokamak Physics Activity-Confinement Data Base and Modelling Topical Group. This model has also been applied to analyse a variety of experiments mostly using electron heat modulation on JET, ASDEX-Upgrade, TORE SUPRA and FTU. The thresholds are found to be in the expected domain for micro-instabilities in tokamaks. However, the stiffness factor is found to cover a broad range of variation.

This paper is an overview of recent results relating to turbulent particle and heat transport, and to the triggering of internal transport barriers (ITBs). The dependence of the turbulent particle pinch velocity on plasma parameters has been clarified and compared with experiment. Magnetic shear and collisionality are found to play a central role. Analysis of heat transport has made progress along two directions: dimensionless scaling laws, which are found to agree with the prediction for electrostatic turbulence, and analysis of modulation experiments, which provide a stringent test of transport models. Finally the formation of ITBs has been addressed by analysing electron transport barriers. It is confirmed that negative magnetic shear, combined with the Shafranov shift, is a robust stabilizing mechanism. However, some well established features of internal barriers are not explained by theory.

The efficiency of generating a helical current in magnetic islands for the purpose of suppression of neoclassical tearing modes (NTMs) by electron cyclotron current drive (ECCD) is studied experimentally in the ASDEX Upgrade tokamak. It is found that the efficiency of generating helical current by continuous current drive in a rotating island drops drastically as the width 2d of the co-ECCD driven current becomes larger than the island width W. However, by modulating the co-ECCD in phase with the rotating islands O point, the efficiency can be recovered. The results are in good agreement with theoretical calculations taking into account the equilibration of the externally driven current on the island flux surfaces. The result is especially important for large next-step fusion devices, such as ITER, where 2d>W is expected to be unavoidable during NTM suppression, suggesting that modulation capability should be foreseen.