R. Drube

Review of the ASDEX Upgrade data acquisition environment - present operation and future requirements

The data acquisition environment of the ASDEX Upgrade fusion experiment was designed in the late 1980s to handle a predicted quantity of 8 Mbytes of data per discharge. After 7 years of operation a review of the whole data acquisition and analysis environment shows what remains of the original design ideas. Comparing the original 15 diagnostics with the present set of 250 diagnostic datasets generated per shot shows how the system has grown. Although now a vast accumulation of functional parts, the system still works in a stable manner and is maintainable. The underlying concepts affirming these qualities are modularity and compatibility. Modularity ensures that most parts of the system can be modified without affecting others. Standards for data structures and interfaces between components and methods are the prerequisites which make modularity work. The experience of the last few years shows that, besides the standards achieved, new, mainly real-time, features are needed: real-time event recognition allowing reaction to complex changing conditions; real-time wavelet analysis allowing adapted sampling rates; real-time data exchange between diagnostics and control; real-time networks allowing flexible computer coupling to permit interplay between different components; object-oriented programming concepts and databases are required for readily adaptable software modules. A final assessment of our present data processing situation and future requirements shows that modern information technology methods have to be applied more intensively to provide the most flexible means to improve the interaction of all components on a large fusion device.

Real-Time Diagnostics and their Applications at ASDEX Upgrade

Abstract For applications of advanced plasma control schemes, many computers that execute complex algorithms need to communicate with low latency so that result data are promptly available for operating adequate actuators that can directly influence the plasma behavior. ASDEX Upgrade has completed the commissioning phase of its real-time diagnostic framework serving that purpose. Several applications were successfully tested, and progress toward a full feedback neoclassical tearing mode stabilization loop is evident. The new real-time diagnostics comprise several new diagnostics capable of acquiring raw data (up to 1 MHz, up to 60 channels), processing the raw data (calibrate, transform, evaluate, etc.) and transmitting the results over suitable networks to other computers, all in real time. Projects for machine safety (divertor cooling and hot spot detection), physics studies [regulation of density peaking by application of electron cyclotron resonance heating (ECRH)], and real-time state monitors (ECRH deposition calculation) have demonstrated the capabilities of the new diagnostics and the control framework. The control system can now operate its actuators in line with decisions based on algorithms with rather high complexity. Adding new control algorithms has become a distributed effort with manageable overhead.

Impact of a pulsed supersonic deuterium gas jet on the ELM behaviour in ASDEX Upgrade

The possibility for pacing of type-I edge localized modes (ELMs) in H-mode plasmas by intermittent gas injection was investigated in ASDEX Upgrade as a possible alternative to, and in comparison with, ELM control by pellets. A Laval nozzle type molecular deuterium injector was used, delivering 1.7u2009ms long jets with up to about 1020D per pulse at a supersonic flow velocity of 2.2u2009kmu2009s−1. With a repetition rate of 2u2009Hz and a fast rise time of ≈25u2009µs, comparable to typical ELM rise times, the injector seemed to be well-suited for single ELM trigger tests. When applied to H-mode discharges with a moderate type-I ELM frequency of 40–60u2009Hz, no prompt (<0.5u2009ms) ELM triggering could be achieved, in contrast to the experience with pellets. There was, however, clear evidence for a delayed effect in the form of an inverse correlation of the gas pulse amplitude with the time interval between the gas pulse and the next ELM. The apparent lack of prompt ELM triggering seems to be due to a self-blocking of the gas jet by an extremely fast formation of a high density plasma layer in the separatrix vicinity, while the delayed effect may be simply caused by the jet-induced axisymmetric edge profile modification, similar to the delayed ELM cascade observed after a prompt ELM in case of large pellet injection. The delayed trigger effect observed might still be useful for ELM control in future machines, but the related high gas fuelling at elevated pulse frequency could make it unattractive in view of overall plasma performance.

Recent Developments in the ASDEX Upgrade Data Acquisition Environment

Abstract ASDEX Upgrade today delivers approximately 25 GBytes of data per week. To manage this demand, which in fact is a growth by a factor of two in the last 2 years, several improvements to the data acquisition (DAQ) system have been made to avoid bottlenecks and to enhance the usability. Modifications were done to the diagnostic clients to speed up the storage of big diagnostic files to the central analysis server. The diagnostic synchronization server has been modified to handle wait requests not only for raw but for any level of evaluated data files. The central analysis server has been upgraded to deliver the power to do synoptic data analysis on up to 500 MBytes/shot on a single multiprocessor machine in shared memory. Additionally a cluster of ten workstations for parallel applications has been built up for MHD equilibrium calculations and other CPU-intensive tasks. The Andrew File System (AFS) archive servers have been upgraded to more disk capacity, a redundant storage architecture and faster network connections. However, as a basis for these improvements the network backbone and the server connections have been moved from FDDI to Gigabit-Ethernet and single workstation connections from Ethernet to Fast-Ethernet. Performance analysis results give an impression of the achieved improvements. Other projects in conjunction with the DAQ system at ASDEX Upgrade are the ‘hotlink’ interface system development for the Soft-X-Ray and Mirnov-Probes diagnostics and the ‘S-link’ development for an enhanced electron cyclotron emission (ECE) diagnostic. Both will serve as prototypes for future real-time diagnostics, which shall be able to deliver processed data in real-time to other systems — especially experiment control — to achieve a possibly better experiment performance.

ASDEX UPGRADE: THE FIRST PERIOD OF OPERATION

The first plasma in ASDEX Upgrade (AUG) was produced on 21stMarch 91 after more than 8 years of design, manufacture, and assembly. During the first operational period until August 92, with 1949 shots, the control and safety systems were commissioned in parallel with plasma discharges. Most discharges were run with He, but also H and D discharges were executed up to a plasma current of 800 kA at Bo = 2 T. With additional heating by ICRH, of up to 2.2 MW power and up to 2 s duration, the H-mode was obtained at a threshold of 1.2 MW after boronization. Discharges with line-densities of up to ne = 4*1019m−3and central electron temperatures of up to Teo =1.5 keV could adequately be diagnosed (ne-, Te-profiles, radiated power, ELM and MHD-activities, impurity influx …). With a CCD-camera the plasma boundary and critical in vessel components were observed. The magnetic diagnostics provided input for elaborate feedback control and reconstruction of the flux surfaces. The main effort of the experimental program was put on establishing the SN-divertor configuration, investigating the vertical n = 0 instability (VDE) and ICR heating scenarios. This paper describes the control systems, discharge essentials, vessel conditioning and the specific VDE behaviour.

Real-time information exchange between data acquisition and control systems at ASDEX Upgrade

Todays data acquisition systems and control at the ASDEX Upgrade fusion experiment communicate during the setup of an experimental discharge only. While control executes the discharge based on recipes and a fixed number of input and output channels, diagnostic measurements are done under a predefined schema. Considerable benefits for scenario specific discharge execution are expected to be drawn from a real-time integration of data acquisition and discharge control systems: Information from the control system can be used for the real-time adaptation of the data acquisitions configuration, such as channel selection, time resolution, signal filtering, analysis process selection and protocolling functions. Information from data acquisition processed from the full set of input channels and diagnostic-specific knowledge can be used to select and fine-tune control configurations. Access to the data acquisition systems as external information sources potentially reduces the complexity of the discharge controls peripheral hardware and preprocessing software. Information exchange among these systems on a peer-to-peer basis allows freely to communicate the results. This is preferable to a master-slave concept where strict communication protocols have to be observed. In this paper we present the current work on the integration of discharge control and data acquisition at ASDEX Upgrade and give an outlook on future developments.