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.

No abstract

Data on erosion and melting of beryllium upper limiter tiles, so-called dump plates (DP), are presented for all three campaigns in the JET tokamak with the ITER-like wall. Highresolution images of the upper wall of JET, show clear signs of flash melting on the ridge of the roof-shaped tiles. The melt layers move in the poloidal direction from the inboard to the outboard tile ending on the last DP tile with an upward going waterfall-like melt structure. Melting was caused mainly by unmitigated plasma disruptions. During three ILW campaigns around 15% of all 12376 plasma pulses were catalogued as disruptions. Thermocouple data from the upper dump plates tiles showed a reduction in energy delivered by disruptions with fewer extreme events in the third campaign, ILW-3, in comparison to ILW-1 and ILW-2. The total Be erosion assessed via precision weighing of tiles retrieved from JET during shutdowns indicated the increasing mass loss across campaigns of up to 0.6 g from a single tile. The mass of splashed melted Be on the upper walls was also estimated using the high-resolution images of wall components taken after each campaign. The results agree with the total material loss estimated by tile weighing (~130 g). Morphological and structural analysis performed on Be melt layers revealed a multilayer structure of re-solidified material composed mainly of Be and BeO with some heavy metal impurities Ni, Fe, W. IBA analysis performed across the affected tile ridge in both poloidal and toroidal direction revealed a low D concentration, in the range 1-4 x 10 17 D atoms/cm 2 .