This document is intended for publication in the open literature. It is made available on the clear understanding that it may not be further circulated and extracts or references may not be published prior to publication of the original when applicable, or without the consent of the Publications Officer, EUROfusion Programme Management Unit,

Access conditions for full suppression of edge localised modes (ELMs) by magnetic perturbations (MP) in low density high confinement mode (H-mode) plasmas are studied in the ASDEX Upgrade tokamak. The main empirical requirements for full ELM suppression in our experiments are: 1. The poloidal spectrum of the MP must be aligned for best plasma response from weakly stable kinkmodes, which amplify the perturbation, 2. The plasma edge density must be below a critical value, 3.3 × 10 19 m −3 . The edge collisionality is in the range ν * i = 0.15−0.42 (ions) and ν * e = 0.15−0.25 (electrons). However, our data does not show that the edge collisionality is the critical parameter that governs access to ELM suppression. 3. The pedestal pressure must be kept sufficiently low to avoid destabilisation of small ELMs. This requirement implies a systematic reduction of pedestal pressure of typically 30% compared to unmitigated ELMy H-mode in otherwise similar plasmas. 4. The edge safety factor q 95 lies within a certain window. Within the range probed so far, q 95 = 3.5−4.2, one such window, q 95 = 3.57−3.95 has been identified. Within the range of plasma rotation encountered so far, no apparent threshold of plasma rotation for ELM suppression is found. This includes cases with large cross field electron flow in the entire pedestal region.

This document is intended for publication in the open literature. It is made available on the clear understanding that it may not be further circulated and extracts or references may not be published prior to publication of the original when applicable, or without the consent of the Publications Officer, EUROfusion Programme Management Unit,

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.