M. Zilker

Max Planck Institute for Plasma Physics, Arching, Bavaria, Germany

Are you M. Zilker?

Claim your profile

Publications (52)32.05 Total impact

  • [Show abstract] [Hide abstract]
    ABSTRACT: The Control and Data Acquisition Station (short: CoDaStation) is a modular application for continuous data acquisition. It is based on the idea of abstract signal sources and signal sinks that are connected to a signal processing network. The structure of this network is defined by a configuration description and the behavior is controlled by a control system. Due to the well-defined interface definitions for signal processing units, configuration sources and control systems it is possible to use various implementations of these components and easily switch between them. This means less effort on environment changes which is especially important for experiment sites with operation times of several years and even decades. Furthermore, it provides room for tests using mock-up implementations and so an improved testability and stability. The CoDaStation will be an integral part of the control and data acquisition system of Wendelstein 7-X. However, the modular design allows the integration into very different scientific environments.
    Fusion Engineering and Design. 01/2014;
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The medium size divertor tokamak ASDEX Upgrade (major and minor radii 1.65 m and 0.5 m, respectively, magnetic-field strength 2.5 T) possesses flexible shaping and versatile heating and current drive systems. Recently the technical capabilities were extended by increasing the electron cyclotron resonance heating (ECRH) power, by installing 2 × 8 internal magnetic perturbation coils, and by improving the ion cyclotron range of frequency compatibility with the tungsten wall. With the perturbation coils, reliable suppression of large type-I edge localized modes (ELMs) could be demonstrated in a wide operational window, which opens up above a critical plasma pedestal density. The pellet fuelling efficiency was observed to increase which gives access to H-mode discharges with peaked density profiles at line densities clearly exceeding the empirical Greenwald limit. Owing to the increased ECRH power of 4 MW, H-mode discharges could be studied in regimes with dominant electron heating and low plasma rotation velocities, i.e. under conditions particularly relevant for ITER. The ion-pressure gradient and the neoclassical radial electric field emerge as key parameters for the transition. Using the total simultaneously available heating power of 23 MW, high performance discharges have been carried out where feed-back controlled radiative cooling in the core and the divertor allowed the divertor peak power loads to be maintained below 5 MW m−2. Under attached divertor conditions, a multi-device scaling expression for the power-decay length was obtained which is independent of major radius and decreases with magnetic field resulting in a decay length of 1 mm for ITER. At higher densities and under partially detached conditions, however, a broadening of the decay length is observed. In discharges with density ramps up to the density limit, the divertor plasma shows a complex behaviour with a localized high-density region in the inner divertor before the outer divertor detaches. Turbulent transport is studied in the core and the scrape-off layer (SOL). Discharges over a wide parameter range exhibit a close link between core momentum and density transport. Consistent with gyro-kinetic calculations, the density gradient at half plasma radius determines the momentum transport through residual stress and thus the central toroidal rotation. In the SOL a close comparison of probe data with a gyro-fluid code showed excellent agreement and points to the dominance of drift waves. Intermittent structures from ELMs and from turbulence are shown to have high ion temperatures even at large distances outside the separatrix.
    Nuclear Fusion 09/2013; 55:104003. · 2.73 Impact Factor
  • Source
  • [Show abstract] [Hide abstract]
    ABSTRACT: Wendelstein 7-X is intended to demonstrate steady state high performance plasma operation and to explore the physics basis of the Helias reactor concept. From there, the W7-X CoDaC concepts aim for steady state plasma stabilization at favorite operation points and largely or even fully documented plasma experiments for achieving efficiently scientific results. Several concepts are being employed that are not common to the fusion research so far. The plasma control system is based on segments consisting of control parameters and transition conditions to other segments. The data acquisition system aims at streaming all acquired data for archiving, control and monitoring purposes and tries to prevent early data reduction. This leads to data streaming rates of up to 30 GBytes/s during plasma operation of up to half an hour. The data analyses framework is intended to support largely automation with analyses chains based on service oriented architectures. A coupling to real time systems is envisaged later in the project for solving complex control problems. The envisaged model based data analyses rely on the existence of unanalyzed raw data in combination with a systematic documentation of the experimental setup, in particular for the diagnostics.
    Fusion Engineering (SOFE), 2013 IEEE 25th Symposium on; 01/2013
  • [Show abstract] [Hide abstract]
    ABSTRACT: For a complex experiment like W7X being subject to changes all along its projected lifetime the advantages of a formalized software development method have already been stated [1]. Quality standards like ISO/IEC-12207 provide a guideline for structuring of development work and improving process and product quality. A considerable number of tools has emerged supporting and automating parts of development work.On W7X progress has been made during the last years in exploiting the benefit of automation and management during software development:–Continuous build, integration and automated test of software artefacts.∘Syntax checks and code quality metrics.∘Documentation generation.∘Feedback for developers by temporal statistics.–Versioned repository for build products (libraries, executables).–Separate snapshot and release repositories and automatic deployment.–Semi-automatic provisioning of applications.–Feedback from testers and feature requests by ticket system.This toolset is working efficiently and allows the team to concentrate on development. The activity there is presently focused on increasing the quality of the existing software to become a dependable product. Testing of single functions and qualities must be simplified. So a restructuring is underway which relies more on small, individually testable components with standardized interfaces providing the capability to construct arbitrary function aggregates for dedicated tests of quality attributes as availability, reliability, performance.A further activity is on improving the development cycle. The use of release cycles has already provided favourable concentration of work and predictability of delivery times. However, the demand has risen, to react quickly on priority changes from W7X-project management. So a more agile development cycle is being prepared relying on smaller working packages, shorter release cycles and an associated release plan giving the software development responsible the possibility to react on a shorter time scale.
    Fusion Engineering and Design. 12/2012; 87(12):2232–2237.
  • [Show abstract] [Hide abstract]
    ABSTRACT: The superconducting stellarator Wendelstein 7-X (W7-X) is a fusion device, which is capable of steady state operation. Furthermore W7-X is a very complex technical system. To cope with these requirements a modular and strongly hierarchical component-based control and data acquisition system has been designed.The behavior of W7-X is characterized by thousands of technical parameters of the participating components. The intended sequential change of those parameters during an experiment is defined in an experiment program. Planning such an experiment program is a crucial and complex task. To reduce the complexity an abstract, more physics-oriented high-level layer has been introduced earlier. The so-called high-level (physics) parameters are used to encapsulate technical details.This contribution will focus on the extension of this layer to a high-level component model. It completely describes the behavior of a component for a certain period of time. It allows not only defining simple value ranges but also complex dependencies between physics parameters. This can be: dependencies within components, dependencies between components or temporal dependencies.Component models can now be analyzed to generate various views of an experiment. A first implementation of such an analyze process is already finished. A graphical preview of a planned discharge can be generated from a chronological sequence of component models. This allows physicists to survey complex planned experiment programs at a glance.
    Fusion Engineering and Design. 12/2012; 87(12):1949–1953.
  • [Show abstract] [Hide abstract]
    ABSTRACT: The set-up of experiment programs for the complex fusion device Wendelstein 7-X has to define a multitude of parameters which have to obey large number of rules arising from physics and technical constraints. Since this is hard to automate as long as the dependencies are not known sufficiently, the W7-X CoDaC team decided to implement an editor following a constructive approach: starting from an established experiment program the user is able to modify parts of it – thus complying the usual workflow of experimenters.Already the very first implementation has been deployed at the W7-X CoDaC prototype, the WEGA stellarator. Driven by agile programming principles the weighting of the requirements has been influenced by the editor usage in the daily experiment routine, thus ensuring client-oriented development steps and short release cycles. At present, a stable program editor implementation with graphical preview, immediate feedback on user actions and instantaneous warnings about incorrect settings is under continuous operation at the CoDaC prototype. It has potential to improve together with growing knowledge about the physical and technical constraints. The experiences gained give certainty that the editor is suitable for future use during the start-up phase and the first years of W7-X operation.
    Fusion Engineering and Design 12/2012; 87(12):1954–1957. · 0.84 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: ITER, Wendelstein 7-X, LHD, and TORE SUPRA are experimental facilities designed to lead the way to steady state fusion devices. These experiments require strategies to sustain a discharge in case of unforeseen events, e.g. heat overloads of plasma facing components or the failure of a plasma heating source. A recovery strategy is needed to get the discharge back for physics exploitation. For this purpose the W7-X segment control framework provides means for automated event detection along with options to formulate and initiate a recovery strategy. Besides handling of failures and degradation there are events that represent a desired plasma physical effect. An example for this kind of event is a transition to from Low to High-Confinement mode. These events indicate that a certain plasma state is reached and scientific examination can be altered thus enabling event-driven multiple experiments per discharge. Examples of both kinds of events will be presented and compared to other approaches in the community.
    Fusion Engineering and Design 12/2012; 87(12):1958–1960. · 0.84 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: With the present construction, in France, of the International Thermonuclear Experimental Reactor (ITER), nu-clear fusion, one of the more promising solutions for the sus-tainable large scale production of energy, is one step closer of becoming a player in the future energy mix. In ITER and in future commercial fusion reactors, highly energetic nuclear fusion reactions will be produced when a deuterium and tritium (D-T) plasma is heated up to temperatures ten times higher than the Sun's core temperature. The position control of this extremely hot column of plasma, inside the reactor's fusion chamber, is one of the most critic issues in the operation of these power generating devices. So far, real-time feedback systems used magnetic measurements to determine and control the plasma position. However, during the reactor's long pulse operation, the fast neutrons and radiation emitted by D-T plasmas may affect the magnetic sensors, likely leading to the accumulation of errors in the feedbacked measurements. Microwave reflectometry, a radar technique for determining the radial distribution of the plasma density, was proposed to provide an alternative to the magnetic measurements used for position control. However, producing reliable measurements from waves propagating in a turbulent fusion plasma has proved to be a difficult and complex task that needed to be successfully tamed before being validated as an alternative for measuring the plasma position. Herein are described the developments required to integrate this radar-like technique in the control systems of a fusion machine. These developments led to the first demonstration of plasma position control made in a fusion device using an alternative to the usual feedback of magnetic measurements. This ITER relevant demonstration was performed in the second largest European machine, the ASDEX Upgrade tokamak, located in Garching, Germany.
    6º Congresso do Comité Português da URSI Aplicações das ondas eletromagnéticas: da eficiência energética à bioengenharia; 11/2012
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: In fusion experiments, real-time feedback control of the plasma position plays a vital role for machine protection and disruption avoidance. This control task is presently performed using magnetic measurements that, in future long pulse tokamak devices of the ITER class, may be affected by drifting integrators or radiation induced voltages in the magnetic pickup coils. These effects could have an impact on the magnetic equilibrium reconstruction, causing potential losses of position control and, consequently, leading to premature discharge termination or plasma-facing component damage. Frequency modulated continuous wave O-mode reflectometry, a non-magnetic dependent technique used to measure the density profile, was proposed to backup or complement the standard magnetic-based control in such devices. This new control scheme has just been successfully demonstrated for the first time on the ASDEX Upgrade (AUG) tokamak. The location of the plasma boundary, used in the control of the plasma column position, was tracked in real-time (RT) using dedicated algorithms and a new approach that combines the reflectometry edge profile and a scaled line integrated density measurement from interferometry. Although feasibility studies on the viability of this method had been previously conducted at AUG, the capabilities required to produce this on-line demonstration were only incorporated in the diagnostic after a recent upgrade of its data acquisition and processing hardware. The results herein presented show the first successful demonstration of the reflectometry plasma position application as proposed for ITER.
    Nuclear Fusion 03/2012; 52(3):32003-6. · 2.73 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Achieving higher levels of plasma performance control in present fusion experiments requires that diagnostics be upgraded to deliver processed physical parameters in real time (RT). A key element in a diagnostic RT upgrade is the data-acquisition system (DAS) that should be capable of delivering the acquired data to the data-processing resources with very low latencies and in the shortest possible time. Adequate standard commercial solutions with these characteristics are not easily found in the market, which leads most of the time to the development of complex custom high-performance designs from ground-up. A mixed solution, partially based on commercial off-the-shelf (COTS) components has been developed to upgrade the existing ASDEX upgrade broadband reflectometry diagnostic so that a full demonstration of plasma position control using RT reflectometry density profile measurements can be performed. The designed 8-channel (12-bit/105 MSPS) DAS features a PCI Express 1.1 (PCIe) x8 interface to enable direct memory access (DMA) data transfers with an effective throughput in excess of 1 GB/s. The use of COTS components resulted in a faster hardware design cycle without compromising system performance and flexibility. The architecture of the system and its main design constraints are herein discussed. Benchmark results for data throughput and overall latency measurements are also presented.
    IEEE Transactions on Nuclear Science 09/2011; · 1.22 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Future fusion experiments, aiming to demonstrate steady state reactor operation, require physics driven plasma control based on increasingly complex plasma models. A precondition for establishing such control systems is widely automated data analysis, which can provide integration of multiple diagnostic on a large scale. Even high quality online data evaluation, which is essential for the scientific documentation of the experiment, has to be performed automatically due to the huge data sets being recorded in long discharge runs. An automated system that can handle these requirements will have to be built on reusable software components that can be maintained by the domain experts: diagnosticians, theorists, engineers and others. For Wendelstein 7-X a service oriented architecture seems to be appropriate, in which software components can be exposed as services with well defined interface contracts. Although grid computing has up to now been mainly used for remote job execution, a more promising service oriented middleware has emerged from the recent grid specification, the open grid service architecture (OGSA). It is based on stateful web services defined by the web service resource framework (WSRF) standard. In particular, the statefulness of services allows to setup complex models without unnecessary performance losses by frequent transmission of large and complex data sets. At present, the usability of this technology in the W7-X CoDaC context is under evaluation by first service implementations.
    Fusion Engineering and Design 07/2010; · 0.84 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Achieving higher levels of plasma performance control in present fusion experiments requires that diagnostics be upgraded to deliver processed physical parameters in real-time (RT). A key element in a diagnostic RT upgrade is the data acquisition system (DAS), that should be capable of delivering the acquired data to the data processing resources with very low latencies and in the shortest possible time. Adequate standard commercial solutions with these characteristics are not easily found in the market, what leads most of the times to the development of complex custom high-performance designs from ground-up. A mixed solution, partially based on commercial off-the-shelf (COTS) components, is under development to upgrade the existing ASDEX Upgrade (AUG) broadband reflectometry diagnostic so that a full demonstration of plasma position control using RT reflectometry density profile measurements can be performed. The 8-channel (12-bit/100 MSPS) DAS being designed features a PCI Express (PCIe) x8 interface to enable direct memory access (DMA) data transfers with throughputs in excess of 1 GB/s. The use of COTS components resulted in a faster hardware design cycle without compromising system performance and flexibility. The architecture of the system and its main design constraints as well as the system integration in the AUG RT diagnostic network are herein discussed. Preliminary benchmark results for data throughput and overall measurement latency are also presented.
    Real Time Conference (RT), 2010 17th IEEE-NPSS; 06/2010
  • [Show abstract] [Hide abstract]
    ABSTRACT: Development of a real-time (RT) data acquisition (DAQ) system for ASDEX Upgrade was started years ago. Based on a modular front-end configuration with serial input/output computer interface a small number of powerful new diagnostics were commissioned.In the course of these developments we concluded that “direct to memory” DAQ concepts hold promise for future DAQ tasks including those with a subsequent RT data processing step. A survey of the requirements of existing and future diagnostics at ASDEX Upgrade was conducted to assess the wider applicability of these DAQ techniques to the upcoming needs for new diagnostics or the refurbishment of old ones.The paper presents the results of the requirements survey and indicates in which cases the developed diagnostic concept can be applied. A series of new diagnostics implemented using these techniques is presented in detail. This covers developed front-end modules as well as whole diagnostic configurations in hardware and software.The achieved DAQ data rates, data transport latencies, and fine-tuning steps to maintain a reliable RT operation are described explicitly. An outlook is given upon the definition of a RT diagnostic standard for ASDEX Upgrade based on the developments described previously.
    Fusion Engineering and Design 04/2010; · 0.84 Impact Factor
  • The Review of scientific instruments 03/2010; 81(3):039903. · 1.52 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: We discuss the circuit design of a digital multiradian phase detector that measures the phase difference between two 10 kHz square wave TTL signals and provides the result as a binary number. The phase resolution of the circuit is 1/64 period and its dynamic range is 256 periods. This circuit has been developed for fusion plasma interferometry with submillimeter waves on the ASDEX Upgrade tokamak. The results from interferometric density measurement are discussed and compared to those obtained with the previously used phase detectors, especially with respect to the occurrence of phase jumps. It is illustrated that the new phase measurement provides a powerful tool for automatic real-time validation of the measured density, which is important for feedback algorithms that are sensitive to spurious density signals.
    The Review of scientific instruments 03/2010; 81(3):033507. · 1.52 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The new quality of the superconducting fusion device Wendelstein 7-X (W7-X) is its potential of steady state operation. W7-X is also a very complex technical system. In these two aspects W7-X compares to other super conducting fusion devices (ITER, KSTAR, EAST, SST-1, Tore Supra, JT-60SA, and LHD). The modular and strongly hierarchical control, data acquisition, and communication (CoDaC) system has been designed to cope with these two aspects, unprecedented for the control systems used by most operating fusion devices.The CoDaC system for W7-X will be thoroughly tested in a prototype installation at the WEGA stellarator in order to minimize the risks before commissioning. WEGA is a classical stellarator which allows steady state plasma pulses at a magnetic field of 0.06T. WEGA can run pulses up to 20s at a magnetic field up to 0.5T. Despite its lesser complexity WEGA has the same main components as W7-X, e.g. magnetic coil systems and ECRH. It is therefore considered to be a suitable test-bed for the control system.This paper focuses on discharges controlled by the new control system and the enhanced surveillance mechanisms introduced during the first year of operation. It shows characteristic curves for the key control parameters of the major components. Taking additionally dependencies and constraints between and within components into account a segmented discharge for each experiment is presented as result.
    Fusion Engineering and Design, v.85, 520-524 (2010). 01/2010;
  • [Show abstract] [Hide abstract]
    ABSTRACT: The long Wendelstein 7-X experiment programs will be segmented into arbitrary time slices. For each of these segments the planned behaviour is stored in a configuration database – and that for every component within the W7-X hierarchical layout. Up to now we have an editor to set up these segments for each single component. But generating the compound program description out of the components’ segments is time-consuming experts’ hand work. It is planned to implement a top-down program editor which is able to break down a high level physics program proposal to component segments while observing all the physical and technical constraints – a quite ambitious intention.This paper describes the first step to an experiment program editor: The “eXpress program editor” is able to modify parameters of a single, a subset, or all segments within a given program, thus adapting its behaviour in an easy way. A graphical program structure overview has been implemented giving fast access to program parameters for comparison and editing. Appropriate parameter editors regard type and constraints. In a single transaction all changes are saved, resulting in a new experiment program ready-to-use for the next experiment run.Even though changing the program's structure is not supported in this version, the eXpress program editor serves as a tool for quick program adaptations in the daily experiment routine. Furthermore this implementation will be used as a design and usage study for the later high-end W7-X program editor.
    Fusion Engineering and Design. 01/2010;
  • [Show abstract] [Hide abstract]
    ABSTRACT: Retrieval of scientific data at large fusion experiments is an important but surprisingly complicated task. There are many different data storage systems that have to be accessed using different software interfaces on varying platforms. With MDS+ a respectable approach towards a more standardised and simplified data access interface was started. However, it has its drawbacks regarding continuously acquired data and provides a whole data storage system rather than only an interface, which is not always demanded. Therefore, a data access interface based on service oriented technologies has been developed aiming for a more standardised and platform independent solution. The contribution will describe the current status of this development. It will explain the interface itself including the WSDL (Web Service Description Language) definition and show currently existing service implementations for the Wendelstein 7-X and the Wendelstein 7-AS data storage systems. Furthermore different possibilities to access the services on the client side will be discussed.
    Fusion Engineering and Design. 01/2010;
  • [Show abstract] [Hide abstract]
    ABSTRACT: One discharge at the steady-state experiment Wendelstein 7-X will last up to 30 min and will be divided into many elementary experimental program phases, so-called segments. Physics programs need to be planned and technical parameters need to be adjusted well before the discharge. The technical parameters provide a complete description for the behavior of all participating components.Physicists, however, aim at specifying the state and behavior of both plasma and machine in terms of physics. Therefore, an abstraction layer above the technical parameters has been developed, the so-called high level parameters. These specify a subset of the technical parameters on a higher, more physics-oriented level and are therefore used to encapsulate technical details. During ordinary program planning, the physicists and engineers only see and work with these high level (physics) parameters.In comparison to shot-based and other existing experiments, this is an entirely new scientifically oriented method to plan experiments in terms of physics parameters (supported by tools) which are automatically mapped onto technical parameters. Besides the introduction of the high level parameter concept, this paper shows the planning of experiment programs with high level parameters for the example of the WEGA (a small stellarator experiment) magnetic field. For this case, details about defining and editing these parameters as well as about the mapping of high onto low level (technical) parameters will be given. Furthermore, the support of these processes by specialized tools is described. In addition, other fields of high level parameter applications will be mentioned.
    Fusion Engineering and Design. 01/2010;