Edge Localized Modes Behaviour Research Articles

Magnetic X-points, stochasticity, and Edge Localized Modes

The confinement and stability properties of the edge of a magnetically confined fusion plasma have long resisted theoretical explanation. Numerical simulation with the M3D extended MHD code shows that large periodic Edge Localized Modes (ELMs), associated with otherwise good confinement, represent a new type of nonlinear MHD plasma instability, in which a coherent plasma instability couples to the favorable part of a chaotic magnetic field. The magnetic surface bounding the plasma typically contains at least one X-point. Under small perturbations, such a Hamiltonian surface splits into two different asymptotic limits. A ballooning-type plasma instability can couple to the "unstable" field manifold that forms bulges around the original magnetic surface and then grow to penetrate deep into the plasma. The magnetic field becomes chaotic over the affected region, allowing plasma loss from the core to the outside. Eventually the instability saturates and the plasma relaxes back towards its original axisymmetric shape. The interaction between the plasma and the homoclinic field perturbation drives a multi-stage process that reproduces many experimental observations. This type of instability may help to explain the wide range of ELM and ELM-free behavior observed in fusion plasmas with X-points, as well as properties of plasma confinement.

Simulation study of density dynamics effect on the ELM behavior with TOPICS-IB

Density dynamics effect on the edge-localized-mode (ELM) behavior has been studied by using an integrated simulation code TOPICS-IB based on a 1.5 dimensional transport code with a stability code of peeling-ballooning modes and a transport model of scrape-off-layer (SOL) and divertor plasmas. Neutral models of core and SOL-divertor regions have been integrated with the TOPICS-IB. The TOPICS-IB successfully simulates the behavior of the whole plasma with density dynamics. It is confirmed by varying the density and the temperature instead of artificially enhancing the collisionality in the previous study that the experimentally observed collisionality dependence of the ELM energy loss is caused by the bootstrap current and the SOL transport. Additionally, ion convective and charge-exchange losses are found to enhance the collisionality dependence by the equipartition effect. On the other hand, the ELM particle loss is found to be almost independent of the collisionality, as shown in experiments, due to the density profile before the ELM and the increase of the SOL density during the ELM in addition to the bootstrap current effect.

Chapter 3: ELMy H-Mode Operation in JET

A wide range of studies on JET have contributed greatly to the development of the ELMy H-mode as a high-performance scenario for fusion devices and to the understanding of the physical processes that underlie it. Development has focused on the production of a high-performance, high-density, stationary scenario suited to deuterium-tritium operation and with small edge energy loads. Physics studies have made strong progress in the understanding of the L-H threshold, energy confinement, pedestal physics, and edge-localized mode behavior. A strong focus of this work has been providing a basis for extrapolation to future machines, such as ITER, for which, as the largest existing tokamak, JET has been of particular importance.

Fast imaging of edge localized mode structure and dynamics in DIII-D

Fast-framing images of CIII and Dα emission in the low-field-side plasma boundary of the DIII-D tokamak [J. L. Luxon, Nucl. Fusion 42, 614 (2002)] show that edge localized modes (ELMs) rapidly eject multiple field-aligned filaments from the plasma edge. The toroidal and poloidal mode numbers of these filaments depend on normalized plasma density, with measured ELM toroidal mode numbers ranging from ⩽10 to 20 in low-density plasmas and 15 to 35 in high-density plasmas. In high-density plasmas with moderate collisionality νped*=0.50, ELMs originate at the low-field-side midplane region and the ion parallel velocity in the scrape-off layer is faster for ELMs with larger Dα divertor emission, suggesting that large ELMs eject higher-temperature ions from deeper within the plasma compared to small ELMs. In low-density plasmas with collisionality νped*=0.25, the midplane and divertor ELM signals appear simultaneously, indicating that ELM behavior depends on collisionality. At all νped*, ELMs drive parallel fluxes to the divertor; in addition, ELMs drive cross-field propagation of filaments, which results in plasma-wall interactions that are poloidally localized within 15cm of the midplane. Using the wall interactions as signatures of the filaments in the scrape-off layer, the measured poloidal width of the filament ranges from 1to5cm.

Effect of plasma shaping on performance in the National Spherical Torus Experiment

The National Spherical Torus Experiment (NSTX) has explored the effects of shaping on plasma performance as determined by many diverse topics including the stability of global magnetohydrodynamic (MHD) modes (e.g., ideal external kinks and resistive wall modes), edge localized modes (ELMs), bootstrap current drive, divertor flux expansion, and heat transport. Improved shaping capability has been crucial to achieving βt∼40%. Precise plasma shape control has been achieved on NSTX using real-time equilibrium reconstruction. NSTX has simultaneously achieved elongation κ∼2.8 and triangularity δ∼0.8. Ideal MHD theory predicts increased stability at high values of shaping factor S≡q95Ip∕(aBt), which has been observed at large values of the S∼37[MA∕(m∙T)] on NSTX. The behavior of ELMs is observed to depend on plasma shape. A description of the ELM regimes attained as shape is varied will be presented. Increased shaping is predicted to increase the bootstrap fraction at fixed Ip. The achievement of strong shaping has enabled operation with 1s pulses with Ip=1MA, and for 1.6s for Ip=700kA. Analysis of the noninductive current fraction as well as empirical analysis of the achievable plasma pulse length as elongation is varied will be presented. Data are presented showing a reduction in peak divertor heat load due to increasing in flux expansion.

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 1.7 ms long jets with up to about 1020D per pulse at a supersonic flow velocity of 2.2 km s−1. With a repetition rate of 2 Hz and a fast rise time of ≈25 µs, 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 40–60 Hz, no prompt (<0.5 ms) 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.

Edge localized mode control with an edge resonant magnetic perturbation

A low amplitude (δbr∕BT=1 part in 5000) edge resonant magnetic field perturbation with toroidal mode number n=3 and poloidal mode numbers between 8 and 15 has been used to suppress most large type I edge localized modes (ELMs) without degrading core plasma confinement. ELMs have been suppressed for periods of up to 8.6 energy confinement times when the edge safety factor q95 is between 3.5 and 4. The large ELMs are replaced by packets of events (possibly type II ELMs) with small amplitude, narrow radial extent, and a higher level of magnetic field and density fluctuations, creating a duty cycle with long “active” intervals of high transport and short “quiet” intervals of low transport. The increased transport associated with these events is less impulsive and slows the recovery of the pedestal profiles to the values reached just before the large ELMs without the n=3 perturbation. Changing the toroidal phase of the perturbation by 60° with respect to the best ELM suppression case reduces the ELM amplitude and frequency by factors of 2–3 in the divertor, produces a more stochastic response in the H-mode pedestal profiles, and displays similar increases in small scale events, although significant numbers of large ELMs survive. In contrast to the best ELM suppression case where the type I ELMs are also suppressed on the outboard midplane, the midplane recycling increases until individual ELMs are no longer discernable. The ELM response depends on the toroidal phase of the applied perturbation because intrinsic error fields make the target plasma nonaxisymmetric, and suggests that at least some of the variation in ELM behavior in a single device or among different devices is due to differences in the intrinsic error fields in these devices. These results indicate that ELMs can be suppressed by small edge resonant magnetic field perturbations. Extrapolation to next-step burning plasma devices will require extending the regime of operation to lower collisionality and understanding the physical mechanism responsible for the ELM suppression.

Effect of B-field dependent particle drifts on ELM behavior in the DIII-D boundary plasma

Edge localized mode (ELM) effects in the DIII-D pedestal and boundary plasmas were measured with multiple fast diagnostics in matched, lower single-null, ELMing H-mode discharges with the ion B×∇B drift either toward or away from the divertor. Data show a strong dependence of the delay in inner vs. outer divertor ELM Dα on drift direction, and a weaker drift dependence of the inner vs. outer delay of Prad, in addition to the strong density dependence seen in previous work. Time dependent boundary plasma modeling during an ELM was done with the UEDGE code including a six-species fluid carbon model and the effect of B-field induced particle drifts. The ELM was modeled as an instantaneous, outer midplane peaked, increase of diffusion coefficients from the top of the pedestal to the SOL. The simulations show delays in the inner vs. outer divertor ELM transients that are similar to the measured delays at moderate density.

Calculation of edge toroidal current density distributions from DIII-D lithium beam measurements using Ampère’s law

The local edge current density j(r) is a parameter of basic importance in understanding the stability of high performance tokamaks, as well as the dynamics of edge localized mode behavior. On DIII-D, the lithium beam polarimetry diagnostic provides precise measurements of the local magnetic field projection along the field of view at 32 radial locations in the plasma edge. Using these measurements, the known spatial calibration and a minimal amount of information about the magnetic field shape from equilibrium reconstructions, Ampères law may be used to provide a straightforward parameterization for the edge toroidal current density in terms of the measured magnetic field and its radial derivative. This approach is relatively insensitive to errors in the reconstruction and is simple to apply.

Impact of toroidal rotation on ELM behaviour in the H-mode on JT-60U

The mitigation of the large pulsed heat loads induced by edge-localized modes (ELMs) on the divertor plates is one of the most important issues for a tokamak fusion reactor. However, ELMs have been completely suppressed in the quiescent H-mode (QH-mode) plasmas produced in the DIII-D tokamak (see Burrell K H et al 2002 Plasma Phys. Control. Fusion 44 A253). One of the key conditions for producing QH-mode plasmas is that the direction of neutral beam injection (NBI) should be opposite to that of the plasma current (i.e. ctr-NBI), which then leads to the toroidal rotation velocity being in the counter direction to the plasma current. By using various combinations of NBI lines in JT-60U, it has been possible to investigate the impact of the toroidal rotation velocity on ELM behaviour by changing the toroidal momentum input in a detailed manner for similar absorbed NB heating power. It has been determined that the ELM frequency decreases with increased counter toroidal rotation velocity at the plasma edge even to the point of the ELMs disappearing. In addition, the magnitude of the pulsed Dα signal at the divertor decreases with decreasing ELM frequency. These results indicate that it is possible to control the ELM behaviour through the toroidal momentum input.

Characterization of peeling–ballooning stability limits on the pedestal

The peeling–ballooning model for edge localized modes (ELMs) and pedestal constraints, based on ideal MHD instabilities driven by pressure gradients and current in the edge barrier region, has been broadly applied toward understanding ELM and pedestal behaviour in a number of tokamak experiments. Due in part to multiple driving terms, multiple wavelengths and second stability access physics, the peeling–ballooning stability limits which are proposed to constrain the pedestal and trigger ELMs depend sensitively on many details of the tokamak equilibrium. Here we present a technique for characterizing these stability constraints as a function of important parameters, using carefully constructed model equilibria. We discuss comparisons of calculated stability constraints to observed pedestal behaviour, in which an encouraging level of agreement is found. We then present results of an extensive series of calculations which characterize the peeling–ballooning stability constraints as a function of pedestal width, magnetic field, plasma current, density, and triangularity.

H-mode threshold and dynamics in the National Spherical Torus Experiment

Edge parameters play a critical role in high confinement mode (H-mode) access, which is a key component of discharge optimization in present day toroidal confinement experiments and the design of next generation devices. Because the edge magnetic topology of a spherical torus (ST) differs from a conventional aspect ratio tokamak, H-modes in STs exhibit important differences compared with tokamaks. The dependence of the National Spherical Torus Experiment (NSTX) [C. Neumeyer et al., Fusion Eng. Des. 54, 275 (2001)] edge plasma on heating power, including the low confinement mode (L-mode) to H-mode (L-H) transition requirements and the occurrence of edge-localized modes (ELMs), and on divertor configuration is quantified. Comparisons between good L-modes and H-modes show greater differences in the ion channel than the electron channel. The threshold power for the H-mode transition in NSTX is generally above the predictions of a recent International Tokamak Experimental Reactor (ITER) [ITER Physics Basis Editors, Nucl. Fusion 39, 2175 (1999)] scaling. Correlations of transition and ELM phenomena with turbulent fluctuations revealed by gas puff imaging and reflectometry are observed. In both single-null and double-null divertor discharges, the density peaks off-axis, sometimes developing prominent “ears” which can be sustained for many energy confinement times, τE, in the absence of ELMs. A wide variety of ELM behavior is observed, and ELM characteristics depend on configuration and fueling.

Influence of plasma–wall interactions on the behaviour of ELMs in JT-60U

In JT-60U, extensive plasma–neutral interaction during type I edge localized modes (ELMs) leads to a transient increase in the edge plasma density, seen as spikes of a few ms duration on the inner vertical interferometer channel (FIR1). The spikes can reach up to ∼40% of the pre-ELM level of the FIR1 signal, in the case of giant ELMs. Fast edge measurements revealed that these density spikes are caused by ionisation of neutrals circulating between the plasma and the wall, as a result of the ELM particle and heat load onto the target. The increase in the edge density is more or less equally divided between the scrape-off layer (SOL) and outer core region, but is outside of the top of the H-mode pedestal. Prompt ionisation of neutrals and the increase in the plasma density around the separatrix position may affect the edge MHD stability, as often manifested by the occurrence of a second, satellite ELM triggered at lower pedestal temperature and pressure, clusters of type III ELMs or periods of L-mode. Both the magnitude of the density spikes and the observed changes in the ELM behaviour were found to depend on the wall conditions.

Active control of edge localized modes by radio frequency waves

The effect of radio frequency wave driven torque on edge localized mode (ELM) activity is studied. It is shown that the radio frequency driven torque causes transition of type I giant ELMs to the benign grassy ELM behavior. The efficiency of this process scales directly with minor radius and inversely with the plasma density. It is argued that the technique of active ELM control will be efficient in the present for moderate-sized machines and in the future for large machines.

A minimal dynamical model of edge localized mode phenomena

A simple, low-dimensional model of edge localized mode (ELM) phenomena is presented. ELM dynamics are determined by the interaction of few basic processes at the edge of tokamak plasma; these include the evolution of magnetohydrodynamic (MHD) pressure gradient driven instabilities, the low–high (L–H) transition (mediated by electric field shear induced suppression and generation), the fueling of the edge by neutral particles, and edge heating by thermal flux from the core plasma. In the parameter regime characteristic of an H-mode plasma, the model exhibits a transition to stationary relaxation oscillations (i.e., stable limit cycle behavior) corresponding to ELMs. The dependence of ELM frequency, amplitude, etc. on the heating power Pin and other control parameters is studied. The transition from giant to grassy ELM behavior appears as a consequence of increasingly marginal stability to ideal ballooning as Pin increases. Several details of the dynamics of the ELM onset and behavior close to the L–H transition threshold are revealed, as well.

JET relevance to ITER, new trends and initial results

Critical ITER physics requirements and R&D issues are considered and, the relevance and capability of JET in investigating these issues are discussed. In particular, JET gives its highest priority to a coherent programme in support of the design of an ITER divertor operating in a regime consistent with a small scrape-off length. JET intends to perform critical tests of such divertor regimes in D-T plasmas. Initial results from JET after its extensive modifications are presented. These show that the new divertor in JET has good power-handling capability and that carbon blooms, which often plagued the earlier campaign of high performance, have not yet been observed. Good progress in divertor characterization and, in achieving the H and VH modes and full lower hybrid non-inductive current drive has been made. Also, valuable insight into edge localized mode behaviour and their occurrence is being gained.

ELM studies on DIII-D and a comparison with ASDEX results

In previous analysis of edge localized modes (ELMs), three different types (types I, II and III) of ELM were observed in DIII-D, whereas only one type of ELM has been found in ASDEX. The type III ELMs observed on DIII-D are similar to the ELMs observed on ASDEX, whereas the type I ELMs did not occur on ASDEX. ELMs are found to be MHD events of short duration (1 to 2 ms), which, on DIII-D, may be followed by a transient L mode phase. This compound event was not observed on ASDEX. A stability analysis of DIII-D discharges indicates that in the standard DIII-D configuration, type III ELMs occur well below the ideal ballooning limit, whereas Type I ELMs generally occur close to this limit. In DIII-D discharges where the plasma parameters are close to the ASDEX values, type I ELMs are not observed, even at the beta limit. This suggests that the ELM behaviour in DIII-D is not uniquely governed by ballooning stability

H-modes under steady-state conditions in JET

Two H-mode regimes have been identified in JET in which edge localized modes (ELMs) maintain steady-state conditions. In the first regime, strong gas puffing was used in combined (ICRF plus NBI) heating experiments at powers of up to 20MW. Rapid ELM activity occurred and at moderate powers ( approximately 8MW) steady-state H-modes with durations of up to 18s and energy confinement times of up to 95% of the JET/D-IIID scaling were established. At high toroidal beta ( beta N>or=1.5) H-mode plasmas were also found to exhibit regular ELM behaviour which resulted in steady-state H-modes with confinement enhancement of approximately 90% of the JET/D-IIID scaling. This paper examines the plasma properties of these regimes and assesses their implications for steady-state H-mode operation in ignited plasmas.

Plasma shaping, edge ballooning stability and ELM behaviour in DIII-D

The occurrence of giant edge localized modes (ELMs) in DIII-D has previously been correlated with the violation of the ballooning stability criterion at the plasma edge. These results are extended in the paper. It is demonstrated theoretically that flux surfaces near the separatrix of properly elongated and triangulated plasmas may be moved into the connection region between the first and second stability regions for ideal MHD ballooning modes, when q is high and the shear is low near the plasma surface. The edge flux surfaces are then predicted to have no limit to the sustainable pressure gradient. Experimentally, giant ELMs disappear in these highly shaped plasmas. However, the edge pressure gradient does not increase and ‘grassy’ ELM behaviour appears instead. These results lend further support to the hypothesis that giant ELMs in DIII-D are triggered by ideal ballooning mode instabilities, but they indicate that giant ELM and grassy ELM behaviour may arise from somewhat different mechanisms.