No abstract

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

FAST is a new machine proposed to support ITER experimental exploitation as well as to anticipate DEMO relevant physics and technology. FAST is aimed at studying, under burning plasma relevant conditions, fast particle (FP) physics, plasma operations and plasma wall interaction in an integrated way. FAST has the capability to approach all the ITER scenarios significantly closer than the present day experiments using deuterium plasmas. The necessity of achieving ITER relevant performance with a moderate cost has led to conceiving a compact tokamak (R = 1.82 m, a = 0.64 m) with high toroidal field (B T up to 8.5 T) and plasma current (I p up to 8 MA). In order to study FP behaviours under conditions similar to those of ITER, the project has been provided with a dominant ion cyclotron resonance heating system (ICRH; 30 MW on the plasma). Moreover, the experiment foresees the use of 6 MW of lower hybrid (LHCD), essentially for plasma control and for non-inductive current drive, and of electron cyclotron resonance heating (ECRH, 4 MW) for localized electron heating and plasma control. The ports have been designed to accommodate up to 10 MW of negative neutral beams (NNBI) in the energy range 0.5-1 MeV. The total power input will be in the 30-40 MW range under different plasma scenarios with a wall power load comparable to that of ITER (P /R ∼ 22 MW m −1). All the ITER scenarios will be studied: from the reference H mode, with plasma edge and ELMs characteristics similar to the ITER ones (Q up to ≈1.5), to a full current drive scenario, lasting around 170 s. The first wall (FW) as well as the divertor plates will be of tungsten in order to ensure reactor relevant

The ITER Scenario Modelling Working Group (ISM WG) is organized within the European Task Force on Integrated Tokamak Modelling (ITM-TF). The main responsibility of the WG is to advance a pan-European approach to integrated predictive modelling of ITER plasmas with the emphasis on urgent issues, identified during the ITER Design Review. Three major topics are discussed, which are considered as urgent and where the WG has the best possible expertize. These are modelling of current profile control, modelling of density control and impurity control in ITER (the last two topics involve modelling of both core and SOL plasma). Different methods of heating and current drive are tested as controllers for the current profile tailoring during the current ramp-up in ITER. These include Ohmic, NBI, ECRH and LHCD methods. Simulation results elucidate the available operational margins and rank different methods according to their ability to meet different requirements. A range of ‘ITER-relevant’ plasmas from existing tokamaks were modelled. Simulations confirmed that the theory-based transport model, GLF23, reproduces the density profile reasonably well and can be used to assess ITER profiles with both pellet injection and gas puffing. In addition, simulations of the SOL plasma were launched using both H-mode and L-mode models for perpendicular transport within the edge barrier and in the SOL. Finally, an integrated approach was also used for the predictive modelling of impurity accumulation in ITER. This includes helium ash, extrinsic impurities (such as argon) and impurities coming from the wall (including tungsten). The relative importance of anomalous and neo-classical pinch contributions towards impurity penetration through the edge transport barrier and further accumulation in the core was assessed.