Search

Significant progress has been made on ASDEX Upgrade during the last two years in the basic understanding of transport, in the extension of the improved H-mode in parameter space and towards an integrated operating scenario and in the development of control methods for major performance limiting instabilities. The important features were the understanding of particle transport and the control of impurity accumulation based on it, the satisfactory operation with predominantly tungsten-clad walls, the improved H-mode operation over density ranges and for temperature ratios covering (non-simultaneously) the ITER requirements on ν*, n/nGW and Te/Ti, the ELM frequency control by pellet injection and the optimization of NTM suppression by DC-ECCD through variation of the launching angle. From these experiments an integrated scenario has emerged which extrapolates to a 50% improvement in n T τ or a 30% reduction of the required current when compared with the ITER base-line assumptions, with moderately peaked electron and controllable high-Z density profiles.

First results are reported on the prediction of disruptions in one tokamak, based on neural networks trained on another tokamak. The studies use data from the JET and ASDEX Upgrade devices, with a neural network trained on just seven normalized plasma parameters. In this way, a simple single layer perceptron network trained solely on JET correctly anticipated 67% of disruptions on ASDEX Upgrade in advance of 0.01 s before the disruption. The converse test led to a 69% success rate in advance of 0.04 s before the disruption in JET. Only one overall time scaling parameter is allowed between the devices, which can be introduced from theoretical arguments. Disruption prediction performance based on such networks trained and tested on the same device shows even higher success rates (JET, 86%; ASDEX Upgrade, 90%), despite the small number of inputs used and simplicity of the network. It is found that while performance for networks trained and tested on the same device can be improved with more complex networks and many adjustable weights, for cross-machine testing the best approach is a simple single layer perceptron. This offers the basis of a potentially useful technique for large future devices such as ITER, which with further development might help to reduce disruption frequency and minimize the need for a large disruption campaign to train disruption avoidance systems.

The plasma parameters required for the occurrence of (3, 2) FIR-NTMs and thus good confinement properties at high normalized plasma pressure (βN) are investigated. Both on ASDEX Upgrade and on JET it is found that, with remarkable consistency between the devices, for a variety of plasma parameters in conventional current profile scenarios, this regime establishes itself for discharges with sufficient high normalized plasma pressure at mode onset, βN,onset > 2.3. An active triggering of this regime has been achieved by lowering the magnetic shear at the q = 4/3 rational surface. That way an ideal (4, 3) MHD mode is being destabilized, forcing the required three-wave coupling between the (3, 2) NTM, (1, 1) and (4, 3) mode activity.

In this paper, we present numerical tools convenient for the investigation of tokamak discharges with current holes. These tools allow the computation of fixed- and free-boundary equilibria with current holes, their representation in Boozer co-ordinates, the determination of the corresponding magnetic fields, field line tracing and also tracing of guiding centre orbits inside the hole. The numerical methods are tested by using two example configurations with different hole sizes. Furthermore, an equilibrium reconstruction of an ASDEX Upgrade discharge is performed. The motional Stark effect polarization angles of this discharge, which indicate a small current hole in the plasma centre, are reproduced by a numerical fit of various equilibrium parameters including pressure profile, rotational transform profile and poloidal coil currents. The resulting plasma parameters lie well inside the experimental error margins.

The MFBE_2001 code computes the magnetic field in a form suitable for field line tracing inside and outside the plasma boundary of two- and three-dimensional finite-β equilibria. It is a modified version of the Magnetic Field Solver for Finite-β Equilibria (MFBE) code which was developed for the investigation of stellarator equilibria without net toroidal current. The numerical method used in MFBE_2001 is based on the `virtual casing' principle which also allows the computation of magnetic fields of equilibria with net toroidal current, such as quasi-axisymmetric and tokamak equilibria. The numerical accuracy of the code is tested for an axisymmetric tokamak equilibrium and the corresponding three-dimensional one including the toroidal field ripple. For this purpose, the results obtained with the code system VMEC/NEMEC+MFBE_2001+GOURDON are compared with those computed with the axisymmetric equilibrium code DIVA.

An on-line predictor of the time to disruption has been installed on the ASDEX Upgrade tokamak. It is suitable either for avoidance of disruptions or for mitigation of those that are unavoidable. The prediction uses a neural network trained on eight plasma parameters and their time derivatives extracted from 99 disruptive discharges. The network was tested off-line over 500 discharges and was found to work reliably and to be able to predict the majority of the disruptions. The trained network was installed on-line, tested over 128 discharges and used to inject killer pellets to mitigate the disruption loads.

Disruption is a sudden loss of magnetic confinement that can cause damage to the machine walls and support structures. For this reason, it is of practical interest to be able to detect the onset of such an event early. A novel technique is presented of early prediction of plasma disruption in tokamak reactors which uses neural networks and `fuzzy' inference. The studies carried out in the work make use of an experimental database of disruptive shots made available by the ASDEX Upgrade Team. The main result of the work is that, in the limit of the available database, it is possible to predict the onset of the disruptive event sufficiently in advance in order to put the control system into action. The proposed system is a modular scheme that exploits a decomposition of the original database carried out in a proper way.

Three-dimensional, ideal MHD tokamak equilibria are computed with the free-boundary NEMEC code, and their stability properties are investigated with the CAS3DN code, which is an ideal, linear MHD code. Numerical problems are discussed in detail by means of a test equilibrium with a large helical core. Further computations are performed for AUG-type equilibria perturbed by resonant magnetic perturbation (RMP) fields, and an ITER scenario taking into account the toroidal field ripple and perturbation fields caused by test blanket modules (TBMs). Moreover, a quantitative measure of the corrugation of the flux surfaces is introduced. It is found that the contour of the corrugation on equilibrium flux surfaces reflect the ideal kink structures of rational-q surfaces, and the periodicity of the external perturbation fields. Both stabilizing and destabilizing effects of the RMP fields, and the influence of the TBMs on the stability properties are observed.