Observations of bulk plasma rotation in radio frequency (RF) heated JET discharges are reported. This study is concentrated on RF heated L-mode plasmas. In particular, the toroidal rotation profiles in plasmas heated by ion cyclotron resonance frequency (ICRF) waves and lower hybrid (LH) waves have been analysed. It is the first time that rotation profiles in JET plasmas with LH waves have been measured in dedicated discharges. It is found that the toroidal plasma rotation in the outer region of the plasmas is in the co-current direction irrespective of the heating scenario. An interesting feature is that the toroidal rotation profile appears to be hollow in many discharges at low plasma current, but a low current in itself does not seem to be a sufficient condition for finding such profiles. Fast ion transport and finite orbit width effects are mechanisms that could explain hollow rotation profiles. This possibility has been investigated by numerical simulations of the torque on the bulk plasma due to fast ICRF accelerated ions. The obtained torque is used in a transport equation for the toroidal momentum density to estimate the effect on the thermal bulk plasma rotation 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.

Plasmas heated by ICRF only in the JET tokamak show distinct structures in the toroidal rotation profile, with regions where dω/dr>0 when the minority cyclotron resonance layer is far off-axis. The rotation is dominantly co-current with a clear off-axis maximum. There is only a slight difference between a high-field side (HFS) or a low-field side position of this resonance layer: the off-axis maximum in the rotation profile is modestly higher for the HFS position. This is in contrast to the predictions of theories that rely mainly on the effects arising from ICRF-driven fast ions to account for ICRF-induced plasma rotation. The differences due to the direction of the antenna spectrum (co- or counter-) are small. A more central deposition of the ICRF power in L-mode and operation in H-mode both lead to more centrally peaked profiles, both in the co-direction. Strong MHD modes brake the rotation and lead to overall flat rotation profiles.