Search

Discharge scenarios and control schemes in ASDEX Upgrade (AUG) are evolving more and more complex. Especially in physics investigations for ITER and DEMO sophisticated scenarios exploit the operational space. This increases the probability of design flaws or human errors in the pulse configuration, but also aggravates the potential damage in the failure case. The ASDEX Upgrade Flight Simulator Fenix, which is currently under construction, will provide a fast and efficient simulation tool for testing and validating discharge scenarios, as well as control and monitoring functions, during their development and immediately prior to experimental pulse execution. This ensures, that the scenarios and settings are adequate to reach the experimental goals and that the margins to operational limits are sufficiently large also during the dynamic evolution of the discharge. Simplified physics and plant system “control” models combined with a representation of the ASDEX Upgrade Discharge Control System (DCS) allow for fast simulation runs with reasonable prediction quality. In the simulation an event generator can trigger plasma instabilities, technical failures and external events to test the resilience of the designed pulse against unplanned incidents. The granularity of modelling shall be customizable, such that the simulator can also be used for detail investigations with elaborate physics at the cost of longer simulation time. As a basis for implementation the ITER Plasma Control System Simulation Platform (PCSSP) has been chosen. The flight simulator, extends PCSSP with an ASTRA co-simulator for the ASDEX Upgrade tokamak model and with custom modules for its actuators, diagnostics and control system. Plugins will enable reading original AUG discharge programs and configuration files, as well as storing the results in the AUG shot file database.

Fenix, the ASDEX Upgrade (AUG) flight simulator under development, is based on the Plasma Control System Simulation Platform (PCSSP) developed for ITER, the ASTRA transport code and the SPIDER equilibrium code. Fenix will give a session leader the possibility to check whether the discharge will meet experimental goals prior to execution. It is also designed to facilitate the development of control system features and the validation of physical models. It reads the AUG discharge program and checks if all the parameters and reference waveforms are reasonable during the discharge simulation. ASTRA serves as a plant model (physical model of AUG tokamak) which outputs idealised diagnostic signals (temperature, density, etc.). ASTRA calculates these data from the particle and energy transport in the plasma core, plasma edge, Scrape of Layer (SOL) and divertor. It also includes particle balance, L-H transition and sawtooth models, and it is equipped with the 2D equilibrium reconstruction code SPIDER. The second component of Fenix is a model of the Discharge Control System (DCS) and AUG actuators. The DCS model processes diagnostic signals from ASTRA, computes commands for actuators and sends them back to ASTRA closing the feedback loop. Controllers for coil currents to control plasma current, position and shape are implemented in the model as well as gas puff valves (divertor and midplane) and the pellet injector to control electron density. The DCS model also simulates external heating actuators such as NBI, ECRH and ICRH. Fenix has methods for reading configuration files and archiving. This article presents modelling of the ASTRA and DCS components, and the first results from the simulations of the flattop phase.

openaire: EC/H2020/633053/EU//EUROfusion An Electron Cyclotron Wave (ECW) system will be used in the European DEMO for the stabilization of Neoclassical Tearing Modes (NTMs). In order to avoid movable mirrors in the harsh environment close to the plasma and to simplify the NTM launcher integration, the tuning of the ECW deposition location can be achieved by launching frequency-tunable ECWs from fixed mirrors while the frequency is tuned in discrete steps of 2–3GHz. An overview of the frequency step-tunable ECW system for NTM stabilization is presented. The design considerations are discussed based on the current DEMO baseline parameters and the status of technologies. A simulation of NTM stabilization with an idealized frequency tunable ECW system on an analytical NTM model is shown. The simulation takes into account a realistic tuning speed based on the present technology and considers the current NTM launcher configurations in DEMO. A simple sweeping strategy is adapted for the control of frequency. Various uncertainties, which will affect the feasibility, need to be further investigated.

Efficient fuelling will be an essential task in the EU-DEMO. The basic requirement here is to establish the target plasma core density with a minimum particle flux by injecting mm-size solid fuel pellets. Modelling showed this requires a pellet launch from the vessel inboard. Optimization can be achieved by the pellet parameters and the injection geometry; the latter however taking into account boundary conditions resulting from system integration needs. Design activities integrating the pellet transfer system into the vessel and the breeding blanket unveiled several possible variants requiring different levels of technical efforts. Basically, all extra efforts bear the benefit of a deeper and hence more favourable particle deposition. To quantify the potential gain, a full closed loop modelling was performed calculating the required pellet particle flux for any solution considered. Results allow now to balance potential advantages against related efforts required. Furthermore, the analysis tools developed can be employed for even more refined optimization of the pellet actuator tool by e.g. taking into account the interplay of pellet fuelling with burn control requirements.

DEMO is a proposed demonstration fusion power plant which is currently being designed. Fusion power, Pfus, has to be controlled to stay above a minimum level to produce sufficient net electricity. However, this has an impact on the power through the separatrix, Psep, and thus can produce excessive heat flux to the divertor which can lead to damage. This gives a limit for the maximum allowable Psep. The aim of this work is to develop a framework which will give the possibility to design and to test controllers by simulating DEMO plasma transport. Therefore, we coupled the ASTRA transport modelling code for fusion devices with the Simulink simulation framework. ASTRA is equipped with equilibrium, transport, fuelling, heating and current drive modules. Simulink is a powerful tool to model and to simulate different dynamic systems. It allows fast and simple development of controllers using its built-in blocks. Therefore, coupling of ASTRA with Simulink gives the advantage of fast development of controllers for the power plant modelled with sophisticated physics based on the transport codes. This coupling was used to run the first basic simulations, where we simulated feedback control of fusion power, power through separatrix and pedestal top Greenwald density fraction, n GW ped top . n GW ped top is controlled via the pellet frequency to keep it below one. Fusion power is controlled with external heating (NBI). To prevent increasing Psep and heat load on the divertor while increasing Pfus, xenon gas can be puffed into the vessel. However, Psep has to be kept above a threshold to stay in H-mode. Therefore, feedback control of the xenon gas puff is modelled. These controllers were used to control fusion power, power through separatrix and pedestal top Greenwald density fraction with two different external disturbances. In the first one, we mimicked changes of the H-factor and in the second one we mimicked the fall of a tungsten flake. Results show that we can control these parameters already with a combination of simple controllers.

Integrated edge-localized-mode (ELM)-free or small ELM scenarios for the demonstration fusion power plant (DEMO) are investigated in ASDEX Upgrade using argon seeding for radiative power removal mainly in the pedestal region. An important aspect is the modification of the electron pressure in the pedestal by the additional radiative power losses. Full ELM suppression could be achieved in a no-ELM H-mode scenario featuring an edge electromagnetic quasicoherent mode up to a heating power of 5 MW, where argon radiation allowed the extension of the heating power operational space. At higher powers up to 12 MW (reaching the beta limit), ELMs of reduced size prevail and detachment is obtained by the argon seeding. Control of the position of a radiating zone localized inside the X-point was found to be favorable compared to the control of the separatrix power for low P s e p / P h e a t . Integration of pedestal cooling with Ar and ELM suppression by resonant magnetic perturbations (RMP) allowed an increase of the core radiation and a partial recovery of normalized confinement to H98 = 1. This favorable behavior is finally limited by the loss of the RMP density pump-out effect, followed by an ELM-free H-mode phase and re-occurrence of ELMs. For an active tailoring of the pedestal pressure profile, precise knowledge of the radiation profile is required. Modeling of the argon radiation profile with the STRAHL code showed good agreement with bolometry only if charge exchange of neutral deuterium with argon ions was taken into account, highlighting the importance of the neutral density in the pedestal region. In view of best integration of a no-ELM scenario with divertor detachment and high heating power, the divertor compression and enrichment of argon and neon were compared using dynamic gas puff experiments. Argon shows more than a factor of 3 higher divertor enrichment compared to neon, but the absolute values decrease with higher neutral deuterium pressure.

The ITER first wall is designed for start-up and ramp-down in limiter configuration. The wall panels are toroidally shaped in order to spread the incident parallel power flux q|| uniformly, assuming a single decay length λq whose value is not known from first principles. In order to study the scaling of q|| with plasma parameters, infra-red viewing of specially-designed limiters has been used on the COMPASS tokamak in ∼100 discharges with scans in Ip, ne and for all combinations of magnetic field and Ip directions. The IR measurement clearly shows that in addition to the main SOL heat flux profile with λq > 40 mm, a steep gradient (λqnear = 4 ± 2 mm) dominates q|| near separatrix. This appears independently of limiter shaping, insertion with respect to neighbors and incident field-line angles. Good agreement is found between the measured λqnear and the prediction of a heuristic drift-based model.

The Thomson scattering diagnostics (TS) at the COMPASS tokamak (operated in the Institute of Plasma Physics ASCR, Prague, Czech Republic) is based on two high-power Nd:YAG lasers, which have a pulse energy of 1.5 J and repetition rate of 30 Hz each. A new timing and synchronization unit (TSU) based on a Field Programmable Gate Array (FPGA) was designed and constructed in order to assure reliable and synchronized control of flash lamps and Q-switches of both lasers. Correct delay between laser pulses is necessary to obtain the highest possible laser energy and to prevent the active medium, the Nd:YAG rod, from thermal damage. An FPGA development kit was chosen as core hardware for the TSU and dedicated input and output circuits were manufactured to adapt the kit interfaces for COMPASS TS needs. In addition, the FPGA firmware, the control software, and the graphical user interface were developed to control the TSU hardware. This paper describes the TS regimes that are addressed by the new TSU. Further, the TSU design, the operation flow, and the firmware and software development are presented. Finally, the results of commissioning are shown.

This paper reports on probe measurements of the electron energy distribution function and plasma potential in the divertor region of the COMPASS tokamak during D-shaped plasmas. The probe data have been processed using the novel first-derivative technique. A comparison with the results obtained by processing the same data with the classical probe technique, which assumes Maxwellian electron energy distribution functions is presented and discussed. In the vicinity of the inner and outer strike points of the divertor the electron energy distribution function can be approximated by a bi-Maxwellian, with a dominating low-energy electron population (4-7 eV) and a minority of higher energy electrons (12-25 eV). In the private flux region between the two strike points the electron energy distribution function is found to be Maxwellian with temperatures in the range of 7-10 eV. The comparative analysis using both techniques has allowed a better insight into the underlying physical processes at the divertor region of the COMPASS tokamak. (© 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim)

The COMPASS plasma control system is based on the MARTe real-time framework. Thanks to MARTe modularity and flexibility new algorithms have been developed for plasma diagnostic (plasma position calculation), control (shaping field control), and protection systems (central solenoid protection). Moreover, the MARTe framework itself was modified to broaden the communication capabilities via Aurora. This paper presents the recent upgrades and improvements made to the COMPASS real-time plasma control system, focusing on the issues related to precision of the real-time calculations, and discussing the improvements in terms of discharge parameters and stability. In particular, the new real-time system has given the possibility to analyze and to minimize the transport delays of each control loop.

This paper presents an overview of the Control, Data Acquisition, and Communication system (CODAC) at the COMPASS tokamak: the hardware set-up, software implementation, and communication tools are described. The diagnostics and the data acquisition are tailored for high spatial and temporal resolution required by the COMPASS physics programme, which aims namely at studies of the plasma edge, pedestal, and Scrape-off-Layer (SOL). Studies of instabilities and turbulence are also an integral part of the programme. Therefore, the data acquisition consists of more than 1000 channels, sampled at rates from 500 kS/s up to 2 GS/s. Presently, the feedback system controls the plasma position and shape, plasma current, and density and it includes 32 analogue input channels as well as 1 digital input/output channel and 8 analogue outputs. The feedback control runs within the Multi-threaded Application Real-Time executor (MARTe) framework with two threads, a 500 μs cycle to control slow systems and a 50 μs cycle to control the fast feedback power supplies for plasma position control. In this paper, special attention is paid to the links between the systems, to the hardware and software connections, and to the communication. The hardware part is described, the software framework is addressed, and the particular implementation – the dedicated software modules, communication protocols, and links to the database are described.

The main magnetic fields in COMPASS – i.e. The Toroidal, Magnetising, Equilibrium, and Shaping Fields – are created by a set of four corresponding thyristor power supplies controlled in a 0.5 ms loop. The plasma position has to be controlled both radially and vertically by two additional magnetic fields provided by two fast amplifiers (FAs) based on MOSFET technology, each supplying ±100 V and up to ±5 kA. Currently, an ongoing project aims at ELM triggering by fast changes of the vertical position of the plasma column, also referred to as vertical kicks. For this purpose, a new Vertical Kicks Power Supply (VKPS) capable of quick change of vertical plasma position is being constructed. This power supply should operate at up to 1.2 kV with switching frequency up to 5 kHz. It is designed as a H-bridge but based on IGBT transistors which can be operated at higher voltages than MOSFETs. In this paper, we focus on the FAs and VKPS engineering design and required output parameters. Both the power supplies are based on modern components with highest available ratings in their categories. Unique design of the power supplies takes advantage of the short duration of the COMPASS discharge by overloading the transistors above their maximal steady-state rating. The FA is regularly operating, so that in addition to describing its design, we also describe the achieved performance parameters. Finally, the common controller unit, communication, and error handling is described.

The COMPASS tokamak is expected to produce several gigabytes of data per shot in near future. A new storage system is needed to accommodate and access all the data. It should be scalable, fault-tolerant, and parallel. It should not be based on proprietary solutions to maintain independence from hardware and software manufacturers and preferably it should be built on inexpensive commodity hardware. One of the promising distributed parallel fault-tolerant file systems, GlusterFS, was selected for testing. The aim of the work was to make initial tests of a particular small GlusterFS setup to confirm its aptitude for the COMPASS storage system. Aggregated reading throughput from multiple NFS clients was one of the most important figures that were benchmarked, it scaled well with the number of clients, starting just above 60 MiB/s 1 for 1 client and going slightly over 300 MiB/s for 6 clients.

The COMPASS tokamak, a divertor device with clear H-mode and ITER-relevant geometry, has been re-installed in IPP Prague after its transport from CCFE in UK. The first plasma was achieved in December 2008. Many new diagnostic tools with both high temporal and spatial resolutions have been developed to address the scientific programme of COMPASS focused on H-mode physics and pedestal investigations. In the paper, an overview of existing and in near future planned diagnostics (magnetic, spectroscopy, microwave, probe and beam/particle) on COMPASS is given including their basic technical specifications and achievements.

An efficient horizontal and vertical stabilization of the plasma column position are essential for a reliable tokamak operation. Plasma position is generally determined by plasma current, plasma pressure and external vertical and horizontal magnetic fields. Such fields are generated by poloidal field coils and proper algorithm for the current control have to by applied, namely, in case of fast feedback loops. This paper presents a real-time plasma position reconstruction algorithms developed for the COMPASS tokamak. Further, its implementation in the MARTe (Multithreaded Application Real-Time executor) is described and the first results from test of the algorithm for real-time control of horizontal plasma positions are presented.

Modern fusion experiments require the presence of several subsystems, responsible for the different parameters involved in the operation of the machine. With the migration from the pre-programmed to the real-time control paradigm, their integration in Control, Data Acquisition, and Communication (CODAC) systems became an important issue, as this implies not only the connection to a main central coordination system, but also communications with related diagnostics and actuators. A subsystem for the control and operation of the vacuum, gas injection and baking was developed and installed in the COMPASS tokamak. These tasks are performed by dsPIC microcontrollers that receive commands from a hub computer and send information regarding the status of the operation. Communications are done in the serial protocol RS-232 through fibre optics. Java software, with an intuitive graphical user interface, for controlling and monitoring of the subsystem was developed and installed in a hub computer. In order to allow operators to perform these tasks remotely besides locally, this was integrated in the FireSignal system. Taking advantage of FireSignal features, it was possible to provide the users with, not only the same functionalities of the local application but also a similar user interface. An independent FireSignal Java Node bridges the central server and the control application. This design makes possible to easily reuse the Node for other subsystems or integrate the vacuum slow control in the other CODAC systems. The complete system, with local and remote control, has been installed successfully on COMPASS and has been in operation since April this year.

The COMPASS tokamak has started its operation recently in Prague and to meet the necessary operation parameters its real-time system, for data processing and control, must be designed for both flexibility and performance, allowing the easy integration of code from several developers and to guarantee the desired time cycle. For this purpose an Advanced Telecommunications Computing Architecture based real-time system has been deployed with a solution built on a multi-core x86 processor. It makes use of two software components: the BaseLib2 and the MARTe (Multithreaded Application Real-Time executor) real-time frameworks. The BaseLib2 framework is a generic real-time library with optimized objects for the implementation of real-time algorithms. This allowed to build a library of modules that process the acquired data and execute control algorithms. MARTe executes these modules in kernel space Real-Time Application Interface allowing to attain the required cycle time and a jitter of less than 1.5 μs. MARTe configuration and data storage are accomplished through a Java hardware client that connects to the FireSignal control and data acquisition software. This article details the implementation of the real-time system for the COMPASS tokamak, in particular the organization of the control code, the design and implementation of the communications with the actuators and how MARTe integrates with the FireSignal software.

The Compass tokamak is currently being installed in Prague, Czech Republic. Its control and data acquisition system is being redesigned and built from scratch and it will be based on novel digital technologies. The proposed hardware for the controller is based on the PICMG 3.0 Advanced Telecommunications Computing Architecture ™ (ATCA) standard. The platform contains one ATCA controller with a Gigabit Ethernet interface, up to 12 ATCA Digitizer-Generator-Processor (DGP) cards and trigger and clock inputs, all on a 12U shelf. The multi-core × 86-based General Purpose Processor (GPP) controller will be connected to the DGP cards by Peripheral Component Interconnect Express ™ (PCIe) point-to-point links through the ATCA backplane. Multiple-input multiple-output (MIMO) signal processing will be shared by the DGP cards using the built-in Field Programmable Gate Arrays (FPGAs) and the controller’s × 86 general processor. Eleven Aurora ™ 2.5 Gbit/s links allow to further parallelize the execution of code among several FPGAs. In order to guarantee real-time execution of the control codes a framework based on Linux and the Real-Time Application Interface (RTAI) will be used. This will explore the features provided by the new multi-core technologies. Synchronization between the subsystems will be guaranteed by a real-time event network. The interface to the system will be provided by the FireSignal control and data acquisition system. This will allow the operators and diagnostic coordinators to configure the hardware, prepare the discharges, pre-program events of interest and follow results from the discharge. FireSignal will also orchestrate the data flow coming from the different diagnostics into the database and to registered data clients.

The electron density on COMPASS is measured using 2 mm microwave interferometer. Interferometer signal is used as an input for the feedback control loop, running under the MARTe real-time framework. Two different threads are used to calculate (fast 50 μs thread) and to control (slow 500 μs thread) the electron density. The interferometer measures a line averaged density along a measurement chord. This paper describes an approach to control the line-averaged electron density in a real-time loop, using a correction to the real plasma shape, the plasma position, and non-linear effects of the electron density measurement at high densities. Newly developed real-time electron density control give COMPASS the chance to control the electron density more accurately which is essential for parametric scans for diagnosticians, for physics experiments and also for achieving plasma scenarios with H-mode.

The core Thomson scattering diagnostic (TS) on the COMPASS tokamak was put in operation and reported earlier. Implementation of edge TS, with spatial resolution along the laser beam up to ∼1/100 of the tokamak minor radius, is presented now. The procedure for spatial calibration and alignment of both core and edge systems is described. Several further upgrades of the TS system, like a triggering unit and piezo motor driven vacuum window shutter, are introduced as well. The edge TS system, together with the core TS, is now in routine operation and provides electron temperature and density profiles.

Recent results of measured fast neutrons created in the D-D reaction on the COMPASS tokamak during ohmic discharges are presented in this paper. Two different type detectors were used during experiment. He-3 detectors and bubble detectors as a support. The measurements are an introduction for neutron diagnostic on tokamak COMPASS and monitoring neutrons during discharges with Neutral Beam Injection (NBI). The He-3 counters and bubble detectors were located in two positions near tokamak vacuum chamber at a distance less than 40 cm to the centre of plasma. The neutrons flux was observed in ohmic discharges. However, analysis of our results does not indicate any clear source of neutrons production during ohmic discharges.

The scrape-off-layer (SOL) parameters in the COMPASS tokamak are studied by using a Langmuir probe mounted on a horizontal reciprocating manipulator. The radial profiles of the plasma potential, the electron energy distribution function and the electron densities are derived from the measured current-voltage probe characteristics by applying the firstderivative probe technique (FDPT). It is shown that close to the tokamak wall the electron energy distribution function is Maxwellian, while in the SOL, in the vicinity of the last closed flux surface and inside the confined plasma, the electron energy distribution function is bi-Maxwellian with a low-energy electron fraction dominating over a higher energy one. The radial profiles of the electron pressure and the parallel electron power flux density in COMPASS are also presented.

Real-time plasma electron density profile estimation and control are essential in the operation of future tokamaks. In particular, the robustness against diagnostics failure and disturbances is important for long pulse operation., A model-based approach to profile estimation is implemented on the control systems of ASDEX Upgrade and TCV, which is able to merge information from various diagnostics for both core and edge density, as well as systematically handling diagnostic failure. The model used for profile estimation is employed to tune a feedback controller before an experiment, thereby reducing the experimental time required for tuning., Subsequently, this observer and controller have been employed in scientific experiments on ASDEX Upgrade and TCV. On ASDEX Upgrade, the density profile estimator was used in high-density pellet-fuelled discharges, providing a more reliable real-time estimate of the core density for feedback control than previously achieved. On TCV, in experiments on integrated pressure and safety factor profile control, the density profile estimator and feedback controller provide a constant density despite disturbances from time-varying ECCD power. Additionally, the real-time density profiles provide an essential input for other real-time plasma state estimation codes including Electron Cyclotron ray tracing codes, contributing to a complete real-time estimation of the entire plasma state.