Generation Of Tokamaks Research Articles

Automated experimental design of safe rampdowns via probabilistic machine learning

Typically the rampdown phase of a shot consists of a decrease in current and injected power and optionally a change in shape, but there is considerable flexibility in the rate, sequencing, and duration of these changes. On the next generation of tokamaks it is essential that this is done safely as the device could be damaged by the stored thermal and electromagnetic energy present in the plasma. This works presents a procedure for automatically choosing experimental rampdown designs to rapidly converge to an effective rampdown trajectory. This procedure uses probabilistic machine learning methods paired with acquisition functions taken from Bayesian optimization. In a set of 2022 experiments at DIII-D, the rampdown designs produced by our method maintained plasma control down to substantially lower current and energy levels than are typically observed. The actions predicted by the model significantly improved as the model was able to explore over the course of the experimental campaign.

Control and Protection Strategy of Toroidal Field Magnet Power Supply System in HL-2M

The HL-2M device is a new generation tokamak at the Southwestern Institute of Physics that is used to develop nuclear fusion research. To confine the plasma, the toroidal field (TF) coil of HL-2M is needed to output 140-kA direct current to generate a 2.2-T magnetic density flux. Therefore, the TF power supply system was developed. It is powered by two flywheel pulsed motor-generator sets, and a cophase counter parallel connection bridge rectifier was used to provide more current. According to the TF power supply system topology, the amplitude of the magnetic flux density of the TF coil can be tuned by adjusting the excitation system of the flywheel pulse generator; the control strategy and mathematical formula are given. Then, the TF power supply protection system is established, and protection strategies under different faults are designed. The control and protection performance of the TF power supply system has been successfully applied and verified in the first plasma discharge of the HL-2M experiment.

Magnetic control of DTT alternative plasma configurations

One of the main challenges concerning next generation tokamaks (such as DEMO) will be the development of a heat and power exhaust system able to withstand the large loads expected in the divertor region. A dedicated Divertor Tokamak Test (DTT) facility has been proposed in the EUROfusion Roadmap, with the aim of testing unconventional solutions, such as advanced magnetic configurations and liquid metal divertors. Magnetic control of alternative plasma configurations, such as the X-Divertor, will play a key role in the solution of the heat exhaust and yet can be a challenging point, due to increased sensitivity introduced by secondary x-points. To overcome the complications introduced by secondary x-points in advanced plasma shapes, magnetic control in DTT is achieved by resolving to the eXtreme Shape Controller, in order to control both the plasma shape and the secondary x-point position.

A methodology for discriminating phase and amplitude effects on synchronization in tokamak pacing experiments

The control of macroscopic instabilities, such as Edge Localised Modes (ELMs) and sawteeth, is becoming an essential ingredient in the optimisation of scenarios in preparation for the next generation of tokamaks and the demonstrative reactor. Various pacing experiments have been indeed successfully carried out in many devices but various details of their interactions with the plasma remain poorly understood, in particular the assessment of the relative contribution of driver phase and amplitude to frequency synchronization. In this paper, a data analysis methodology for investigating the details of pacing experiments is proposed. The approach is based on the wavelet decomposition of the signals and the following implementation of information theoretic indicators, to determine the basic form of the interactions. The main analysis tool deployed is the conditional mutual information, equivalent to the informational transfer entropy, which can detect actual causal influences and not only statistical correlations between signals. The most flexible type of ELM pacing, the vertical kicks, permits to clearly confirm the potential of the analysis tools developed. In the dedicated experiments to optimise the ELMs pacing with pellets and sawteeth synchronization with ICRH modulation, the main effect is due to the influence of the amplitude of the external perturbations. Some evidence of phase synchronization has been found, in both JET and AUG, which indicates the direction of possible future optimization of the interventions. Indeed, adjusting the phase of the pacing schemes would free density and ICRH power to give more leeway to other feedback loops, an important aspect, particularly for the integrated control of the next generation of devices. The long-term objective of this line of research is therefore twofold. On the one hand, it is expected to contribute to better physical interpretation of the pacing experiments and possibly also of their target instabilities. From an operational perspective, hopefully these insights will help in the development of reactor relevant scenarios, as exemplified by examples of ELM a pacing with pellets carried out in preparation for the DT campaign on JET.

Scenario adaptive disruption prediction study for next generation burning-plasma tokamaks

Next generation High Performance (HP) tokamaks risk damage from unmitigated disruptions at high current and power. Achieving reliable disruption prediction for a device’s HP operation based on its Low Performance (LP) data is a key to its success. In this letter, through explorative data analysis and dedicated numerical experiments on multiple existing tokamaks, we demonstrate how the operational regimes of tokamaks can affect the power of a trained disruption predictor. First, our results suggest data-driven disruption predictors trained on abundant LP discharges work poorly on the HP regime of the same tokamak, which is a consequence of the distinct distributions of the tightly correlated signals related to disruptions in these two regimes. Second, we find that matching operational parameters among tokamaks strongly improves cross-machine accuracy which implies our model learns from the underlying scalings of dimensionless physics parameters like q 95, β p and confirms the importance of these parameters in disruption physics and cross machine domain matching from the data-driven perspective. Finally, our results show in the absence of HP data from the target devices, the best predictivity of the HP regime for the target machine can be achieved by combining LP data from the target with HP data from other machines. These results provide a possible disruption predictor development strategy for next generation tokamaks, such as ITER and SPARC, and highlight the importance of developing baseline scenario discharges of future tokamaks on existing machines to collect more relevant disruptive data.

ELM and inter-ELM heat and particle flux to a secondary divertor in the DIII-D tokamak

DIII-D edge-localized-modes (ELM) discharges with a secondary X-point inside the vessel were used to evaluate the power distribution between the primary and secondary divertors. The magnetic balance, defined by the distance between the primary and secondary separatrices at the outer midplane (dRsep), was varied from −5 mm (lower-single-null, LSN) to +16 mm (upper-single-null, USN). We find that the secondary divertor receives up to 1/3 of the total heat flux (primary + secondary divertors), and it is dominated by the ELM-driven flux. Both the radially-integrated and the peak heat flux to the secondary divertor decay below ∼50% of the maximum as dRsep is varied from −5 to +6 mm, but the integrated heat flux decay flattens out and it would take dRsep above ∼25 mm to reduce the ELM heat flux to the secondary divertor below 10% of the one deposited to a well-defined SN. Both the secondary inner strike point (ISP) and outer strike point receive heat flux during ELMs. The peak heat flux is comparable in both strike points for dRsep < 10 mm, but the ISP receives less than 15% of the total flux. This is experimental evidence that the secondary inner divertor region receives significant ELM flux which should be taken into account for the design of future generation tokamaks.

Observation of two threshold fields for runaway-electron generation in tokamaks

Two different threshold electric fields, characterizing a lower field required for significant seed RE generation and sustainment and a higher field required for the RE avalanche onset, have been experimentally observed in the flattop of EAST discharges. These results demonstrate that the threshold electric field for the RE avalanche onset is 1.2-fold higher than the RE detection onset field required for the primary generation. An updated 0-D model, including the delay in the RE avalanche onset, suggests that the RE population should be less than the previous prediction. The delay in the RE avalanche onset also opens a possibility for RE suppression at relatively low electron density.

The particle tracking code BBCNI for large negative ion beams and their diagnostics

Heating and current drive in the next generation tokamak ITER requires the use of large and powerful neutral beams, generated by a precursor ion beam from an ion source around 1 m × 2 m in cross-section. To avoid energy losses and component damage, strict requirements are placed on the divergence and uniformity of this ion beam, which is comprised of many individual beamlets. Understanding, controlling, and predicting the behaviour of these large ion beams requires knowledge of these individual beamlets and their interactions with one another. This is hindered by available experimental diagnostics on these large beams typically only having access to volume averaged information. A forward simulation of beam diagnostics would allow the connection of experimental results with otherwise unobtainable individual beamlet properties. The particle tracking and ray tracing code Bavarian Beam Code for Negative Ions was developed for this reason, and takes into account the interaction of individual component beamlets with whole-beam diagnostics to produce synthetic data that can be compared with experimental results. In this work a significantly reworked and upgraded version of the code is presented and example results are given and analysed for the ITER relevant test facility BATMAN Upgrade. It is shown how the simulation can recreate experimental results, and that one must consider the whole beam in order to do so. The impact of beamlet mixing on beam emission spectroscopy results is shown, as is the importance of long range magnetic fields on the beam transport. The capabilities and limitations of the code are discussed with a view toward application to ITER size ion sources.

Self-driven current generation in turbulent fusion plasmas

Global gyrokinetic simulations with self-consistent coupling of neoclassical and turbulent dynamics show that turbulence can significantly affect plasma self-driven mean current generation in tokamaks. The current amplitude, profile and associated phase space structures can all be modified. Turbulence can significantly reduce the current generation in the collisionless regime, generate current profile corrugation near the rational magnetic surface and nonlocally drive current in the linearly stable region—all these are expected to have a radical impact on broad tokamak physics. Both electron parallel acceleration and residual stress from turbulence play crucial roles in turbulence-induced current generation.

Tokamak evolution and view to future

The article begins with a brief review of the achievements of Russian tokamaks in the active period of their development from 1962 to 1973, under the leadership of academician L.A. Artsimovich. During these years, the following basic issues were solved: the equilibrium problem, the MHD stability of the plasma column, and creation of the hot plasma with intense DD neutron radiation. It was shown that the ion confinement in tokamaks is close to the neoclassical model, and the electron confinement is abnormal. It improves with the increasing frequency of collisions, the opposite of the case with ions, in what is known as the alternative model of confinement along the magnetic field. Finally, the first scaling law for the energy lifetime of plasma was obtained, which accurately predicted the plasma parameters of the next generation of tokamaks (the so-called T-4 scaling). The subsequent movement in this direction (the ‘Artsimovich vector’) led to the creation of DT reactors with a fusion power of up to 10 MW (TFTR, JET) and to the ITER project. The main objective of the further development of tokamaks is their transition to steady-state fusion operation, which is a prerequisite for their use in industrial power generation. This makes it necessary to re-evaluate the achievements and obstacles that have to be overcome. The first limitation which thus arises is the so-called PH/S limit, which limits the value of the plasma heating power in a tokamak, as well as the discharge duration ▵t (the ‘TRIAM vector’) in current tokamaks (▵t ~ 1/(PH/S)1.7). Analysis of the existing experimental data shows that the most probable reason for the limitation of PH in existing tokamaks is the breakdown of the plasma sheath in the places of direct contact of the plasma with the wall. The reason for limiting the discharge duration ▵t may be the gradual accumulation of the erosion products in the contact zones of the plasma with the tokamak first wall, which can facilitate such a breakdown. Creating a closed circulating lithium flow between the first wall and plasma is proposed as the solution to the problem of accumulation of the products of the first wall erosion. Preliminary studies (appendices and ) have shown that the undesirable accumulation of tritium in the protective lithium films can be avoided if the temperature of the wall of the tokamak discharge chamber does not exceed 400 °C.

Optimization of the PF coil system in axisymmetric fusion devices

The tokamak is nowadays the most promising fusion device for the plasma confinement and the production of fusion energy. It exploits magnetic fields to confine plasma inside a torus shaped chamber. The magnet system of a tokamak device is mainly composed by Toroidal Field (TF), Central Solenoid (CS) and Poloidal Field (PF) coils. In this paper, a new approach is described for the optimization of the PF coil system. The proposed procedure allows to optimize the number, position and dimension of the PF coils reducing, at the same time, currents and forces on the coils while fulfilling the machine technological constraints. The method exploits the linearized relation between the plasma-wall gaps and the PF coil currents. The procedure effectiveness has been tested on the PF coil system of the ITER tokamak and exploited for the design and optimization of the PF coil system for the next generation tokamak DEMO, as shown in the last section of the paper.

Dust generation in tokamaks: Overview of beryllium and tungsten dust characterisation in JET with the ITER-like wall

Operation of the JET tokamak with beryllium and tungsten ITER-like wall provides unique opportunity for detailed studies on dust generation: quantity, morphology, location, etc. The programme carried out in response to ITER needs for safety assessment comprises: (i) remotely controlled vacuum cleaning of the divertor; (ii) local sampling of loosely bound matter from plasma-facing components (PFC); (iii) collection of mobilized dust on various erosion-deposition probes located in the divertor and in the main chamber. Results of comprehensive analyses performed by a number of complementary techniques, e.g. a range of microscopy methods, electron and ion spectroscopy, liquid scintillography and thermal desorption, are summarized by following points:(a) Total amount of dust collected by vacuum cleaning after three campaigns is about 1–1.4 g per campaign (19.1–23.5 h plasma operation), i.e. over 100 times smaller than in JET operated with carbon walls (i.e. in JET-C).(b) Two major categories of Be dust are identified: flakes of co-deposits formed on PFC and droplets (2–10 μm in diameter). Small quantifies, below 1 g, of Be droplets and splashes are associated mainly with melting of beryllium limiters.(c) Tungsten dust occurs mainly as partly molten flakes originating from the W-coated tiles.

Density control in ITER: an iterative learning control and robust control approach

Plasma density control for next generation tokamaks, such as ITER, is challenging because of multiple reasons. The response of the usual gas valve actuators in future, larger fusion devices, might be too slow for feedback control. Both pellet fuelling and the use of feedforward-based control may help to solve this problem. Also, tight density limits arise during ramp-up, due to operational limits related to divertor detachment and radiative collapses. As the number of shots available for controller tuning will be limited in ITER, in this paper, iterative learning control (ILC) is proposed to determine optimal feedforward actuator inputs based on tracking errors, obtained in previous shots. This control method can take the actuator and density limits into account and can deal with large actuator delays. However, a purely feedforward-based density control may not be sufficient due to the presence of disturbances and shot-to-shot differences. Therefore, robust control synthesis is used to construct a robustly stabilizing feedback controller. In simulations, it is shown that this combined controller strategy is able to achieve good tracking performance in the presence of shot-to-shot differences, tight constraints, and model mismatches.

Stress Analysis of the K-DEMO Vacuum Vessel

Components that make up the central column of a tokamak have a strong impact on the overall sizing of the reactor. In most of the next generation tokamaks being considered at PPPL, the vessel is separate from the blanket support structures. A substantial structure is provided as nuclear and electromagnetic shielding to protect the inner legs of the TF and the vessel pressure boundary. The K-DEMO reactor uses a version of this concept.This technical note addresses the structural adequacy of the K-DEMO vacuum vessel design as of November 2015. The vessel surrounds the internal vacuum components of the reactor and its primary purpose is only to provide the vacuum boundary for the rest of the internals. Static vacuum pressure stresses, stresses due to static magnetic loads, and approximate disruption stresses have been evaluated.

Modelling of advanced three-ion ICRF heating and fast ion generation scheme for tokamaks and stellarators

Absorption of ion-cyclotron range of frequencies waves at the fundamental resonance is an efficient source of plasma heating and fast ion generation in tokamaks and stellarators. This heating method is planned to be exploited as a fast ion source in the Wendelstein 7-X stellarator. The work presented here assesses the possibility of using the newly developed three-ion species scheme (Kazakov et al (2015) Nucl. Fusion 55 032001) in tokamak and stellarator plasmas, which could offer the capability of generating more energetic ions than the traditional minority heating scheme with moderate input power. Using the SCENIC code, it is found that fast ions in the MeV range of energy can be produced in JET-like plasmas. The RF-induced particle pinch is seen to strongly impact the fast ion pressure profile in particular. Our results show that in typical high-density W7-X plasmas, the three-ion species scheme generates more energetic ions than the more traditional minority heating scheme, which makes three-ion scenario promising for fast-ion confinement studies in W7-X.

Preemptive data distribution infrastructure for data centric analysis and modelling

The next generation of tokamaks, e.g. ITER, will have extremely large data collection rates significantly larger than those experienced today in present tokamaks, with consequential new challenges in data management, data analysis and integrated modelling. One of these challenges is to ensure that appropriate data is efficiently made available when it is required and where it is consumed. Data volumes with limited network capabilities mean not all data can be distributed in time when a data-object is requested. One possible solution is to preemptively identify and distribute efficiently the data across the storage services before a user or an application requests it. Preemptive data distribution rely on analysis of historical access patterns to identify a set of rules whereby following a data-object request the most probable set of next requests can be inferred. Implementation of these rules requires the inferred sets of data to be moved close to data-object consumer. The work presented will describe the Apache Spark Machine Learning tools, the results of the analysis, and an implementation of the preemptive distribution experimental platform at CCFE, together with plans for its future integration and testing on the upcoming SAGE platform.

The behavior of runaway current in massive gas injection fast shutdown plasmas in J-TEXT

Runaway currents following disruptions have an important effect on the first wall in current tokamaks and will be more severe in next generation tokamaks. The behavior of runaway currents in massive gas injection (MGI) induced disruptions have been investigated in the J-TEXT tokamak. The cold front induced by the gas jet penetrates helically along field lines, preferentially toward the high field side and stops at a location near the q = 2 surface before the disruption. When the cold front reaches the q = 2 surface it initiates magnetohydrodynamic activities and results in disruption. It is found that the MGI of He or Ne results in runaway free shutdown in a large range of gas injections. Mixture injection of He and Ar (90% He and 10%Ar) consistently results in runaway free shutdown. A moderate amount of Ar injection could produce significant runaway current. The maximum runaway energy in the runaway plateau is estimated using a simplified model which neglects the drag forces and other energy loss mechanisms. The maximum runaway energy increases with decreasing runaway current. Imaging of the runaway beam using a soft x-ray array during the runaway current plateau indicates that the runaway beam is located in the center of the plasma. Resonant magnetic perturbation (RMP) is applied to reduce the runaway current successfully during the disruption phase in a small scale tokamak, J-TEXT. When the runaway current builds up, the application of RMP cannot decouple the runaway beam due to the lower sensitivity of the energetic runaway electrons to the magnetic perturbation.

Dynamics of intrinsic axial flows in unsheared, uniform magnetic fields

A simple model for the generation and amplification of intrinsic axial flow in a linear device, controlled shear decorrelation experiment, is proposed. This model proposes and builds upon a novel dynamical symmetry breaking mechanism, using a simple theory of drift wave turbulence in the presence of axial flow shear. This mechanism does not require complex magnetic field structure, such as shear, and thus is also applicable to intrinsic rotation generation in tokamaks at weak or zero magnetic shear, as well as to linear devices. This mechanism is essentially the self-amplification of the mean axial flow profile, i.e., a modulational instability. Hence, the flow development is a form of negative viscosity phenomenon. Unlike conventional mechanisms where the residual stress produces an intrinsic torque, in this dynamical symmetry breaking scheme, the residual stress induces a negative increment to the ambient turbulent viscosity. The axial flow shear is then amplified by this negative viscosity increment. The resulting mean axial flow profile is calculated and discussed by analogy with the problem of turbulent pipe flow. For tokamaks, the negative viscosity is not needed to generate intrinsic rotation. However, toroidal rotation profile gradient is enhanced by the negative increment in turbulent viscosity.

Prediction of density limit disruptions on the J-TEXT tokamak

Disruption mitigation is essential for the next generation of tokamaks. The prediction of plasma disruption is the key to disruption mitigation. A neural network combining eight input signals has been developed to predict the density limit disruptions on the J-TEXT tokamak. An optimized training method has been proposed which has improved the prediction performance. The network obtained has been tested on 64 disruption shots and 205 non-disruption shots. A successful alarm rate of 82.8% with a false alarm rate of 12.3% can be achieved at 4.8 ms prior to the current spike of the disruption. It indicates that more physical parameters than the current physical scaling should be considered for predicting the density limit. It was also found that the critical density for disruption can be predicted several tens of milliseconds in advance in most cases. Furthermore, if the network is used for real-time density feedback control, more than 95% of the density limit disruptions can be avoided by setting a proper threshold.

Impurity-induced divertor plasma oscillations

Two different oscillatory plasma regimes induced by seeding the plasma with high- and low-Z impurities are found for ITER-like divertor plasmas, using computer modeling with the DUSTT/UEDGE and SOLPS4.3 plasma-impurity transport codes. The oscillations are characterized by significant variations of the impurity-radiated power and of the peak heat load on the divertor targets. Qualitative analysis of the divertor plasma oscillations reveals different mechanisms driving the oscillations in the cases of high- and low-Z impurity seeding. The oscillations caused by the high-Z impurities are excited near the X-point by an impurity-related instability of the radiation-condensation type, accompanied by parallel impurity ion transport affected by the thermal and plasma friction forces. The driving mechanism of the oscillations induced by the low-Z impurities is related to the cross-field transport of the impurity atoms, causing alteration between the high and low plasma temperature regimes in the plasma recycling region near the divertor targets. The implications of the impurity-induced plasma oscillations for divertor operation in the next generation tokamaks are also discussed.