In order to explain the results of the non-inductive current produced in the lower hybrid current drive (LHCD) experiments, a broadening of the radiofrequency (RF) power spectrum coupled to tokamak plasma needs to occur. The presented modelling, supported by diagnostic measurements, shows that the parametric instability (PI) driven by ion sound quasimodes, which occur in the scrape-off plasma layer located near the antenna mouth, produces a significant broadening of the launched LH spectrum. Considering the parameters of LHCD experiments of JET (Joint European Torus), and other machines as well, the PI growth rate is high enough for producing the compensation of the convective losses and, consequently, the broadening of a small fraction (of the order of 10%) of the launched power spectrum. Such a phenomenon is identified to be intrinsic to the RF power coupling in the LHCD experiments. As the principal implication of considering such spectral broadening in modelling the LH deposition profile, experiments of LHCD-sustained internal transport barriers in JET were successfully interpreted, which evidenced the effects of a well-defined LH deposition profile. The present work is important for addressing the long-lasting debate on the problem of the so-called spectral gap in LHCD. The design of LHCD scenarios relevant to the modern fusion research programme, an important requirement of which is the control of the plasma current profile in the outer half of plasma, can be properly achieved by considering PI-induced spectral broadening.

Abstract. The stationary ELM-free "Quiescent H-mode" (QH-mode) regime, obtained with counter neutral beam injection, is studied in ASDEX Upgrade (AUG) and JET. QH-mode plasmas have high pedestal and core ion temperatures together with good H-mode confinement. ELMs are replaced by continuous MHD oscillations, the "Edge Harmonic Oscillation" (EHO) and the "High Frequency Oscillation". Stationarity of particle and impurity densities is linked to the occurrence of these MHD modes. The EHO location in the steep-gradient region and its appearance with increasing edge pressure points at the edge pressure or pressure gradient as possible drivers for the EHO. Injection of small cryogenic pellets can raise the plasma density somewhat without triggering ELMs. Orbit-following calculations of the slowing down distribution show the presence of an enhanced fast particle density in the H-mode barrier region despite the large loss currents with counter-injection.

Results from the first measurements of a core plasma poloidal rotation velocity (upsilontheta) across internal transport barriers (ITB) on JET are presented. The spatial and temporal evolution of the ITB can be followed along with the upsilontheta radial profiles, providing a very clear link between the location of the steepest region of the ion temperature gradient and localized spin-up of upsilontheta. The upsilontheta measurements are an order of magnitude higher than the neoclassical predictions for thermal particles in the ITB region, contrary to the close agreement found between the determined and predicted particle and heat transport coefficients [K.-D. Zastrow, Plasma Phys. Controlled Fusion 46, B255 (2004)]. These results have significant implications for the understanding of transport barrier dynamics due to their large impact on the measured radial electric field profile.

We present an ultrafast neural network (NN) model, QLKNN, which predicts core tokamak transport heat and particle fluxes. QLKNN is a surrogate model based on a database of 300 million flux calculations of the quasilinear gyrokinetic transport model QuaLiKiz. The database covers a wide range of realistic tokamak core parameters. Physical features such as the existence of a critical gradient for the onset of turbulent transport were integrated into the neural network training methodology. We have coupled QLKNN to the tokamak modelling framework JINTRAC and rapid control-oriented tokamak transport solver RAPTOR. The coupled frameworks are demonstrated and validated through application to three JET shots covering a representative spread of H-mode operating space, predicting turbulent transport of energy and particles in the plasma core. JINTRAC-QLKNN and RAPTOR-QLKNN are able to accurately reproduce JINTRAC-QuaLiKiz T i,e and n e profiles, but 3 to 5 orders of magnitude faster. Simulations which take hours are reduced down to only a few tens of seconds. The discrepancy in the final source-driven predicted profiles between QLKNN and QuaLiKiz is on the order 1%-15%. Also the dynamic behaviour was well captured by QLKNN, with differences of only 4%-10% compared to JINTRAC-QuaLiKiz observed at mid-radius, for a study of density buildup following the L-H transition. Deployment of neural network surrogate models in multi-physics integrated tokamak modelling is a promising route towards enabling accurate and fast tokamak scenario optimization, Uncertainty Quantification, and control applications.