Table of contents

The full text of this introduction is available in the PDF.

An arbitrary beam shaping-zooming concept is proposed for direct drive fusion on the Laser Mégajoule and the National Ignition Facility. It is based on the intrinsic conception of these facilities which use four beamlets per beam. A substantial improvement of the laser–target coupling efficiency is achieved with a lowering of ignition threshold. A design providing a thermonuclear energy of about 30 MJ with less than 1 MJ energy-laser has been investigated. The control of low-modes irradiation uniformity is achieved dynamically with an enhancement of the hydrodynamic stability of the acceleration and deceleration phases.

Eventual elimination of the in-board Ohmic heating solenoid is required for the spherical torus (ST) to function as a compact component test facility and as an attractive fusion power plant. An in-board Ohmic solenoid, along with the shielding needed for its insulation, can dramatically increase the size and, hence, the cost of the plant. Advanced tokamak reactor designs also assume no or a small in-board solenoid to reduce the size and cost of the plant. In addition, elimination of the in-board solenoid greatly reduces the coil stresses and simplifies the coil design. Here, we investigate using static as well as dynamic codes in ST geometries with two complementary solenoid-free plasma start-up approaches: one utilizes only the outer poloidal field coils to create a relatively high quality field null region while retaining significant poloidal flux, and the other takes advantage of the poloidal flux stored in the conducting centre-post to create a start-up condition similar to that of the conventional Ohmic solenoid method. We find that it is therefore possible to come up with a promising configuration, which produces a quality multi-pole field-null and sufficient loop-voltage needed for plasma initiation and significant poloidal flux for subsequent current ramp-up. The present solenoid-free start-up concept, if proved feasible, can be readily extended to higher field devices due to relatively simple physics principles and favourable scaling with the device size and toroidal field.

To investigate energetic ion transport induced by bursting modes in the frequency range of Alfvén eigenmodes, which is called abrupt large-amplitude events (ALEs) driven by negative-ion-based neutral beam (N-NB) injection, neutron emission profile measurement and charge exchange (CX) neutral particle (flux) measurement using a natural diamond detector have been performed simultaneously in JT-60U. It is found from the CX neutral particle (flux) measurement that energetic neutral particles in a limited energy range (100–370 keV) are enhanced due to ALEs, and the neutron radial profile is flattened. The change in the energetic ion density profile inferred from these measurements indicates that ALEs expel energetic ions from the core region of the plasma and induce both redistribution and loss of energetic ions. It has been shown that the energy range of transported energetic ions is consistent with a resonance condition between energetic ions and ALEs, and the energetic ion transport results from the resonant interaction between energetic ions and ALEs. Further, a fraction of the energetic ion loss has been quantitatively estimated to be ∼4%.

The development of scenarios characterized by H-mode confinement and internal transport barriers (ITBs) in high triangularity, δ ∼ 0.4–0.5, discharges is of particular interest for ITER advanced tokamak operation. Previous JET experiments have shown that high triangularity favours H-modes which are ELM-free or develop type I edge localized mode (ELM) activity, which inhibits long lasting ITBs. The recent experiments reported here concentrate on integrated optimization of edge and core conditions. The stability of the edge pedestal was controlled using gas injection, deuterium or light impurities, and plasma current ramps. Both methods yield more ITB-friendly edge pedestal conditions, varying from small type I to type III ELMs and, in extreme cases, resulting in L-mode. In parallel, the conditions for triggering and sustaining ITBs encompassing a large proportion of the plasma volume (outer ITBs) were optimized, as opposed to less performing ITBs located closer to the plasma centre (inner ITB). These plasmas have deeply reversed target current profiles with q_min ∼ 3 and a narrow inner ITB, located typically at a small normalized radius ρ < 0.5 and close to the reversed shear region, is routinely observed. Large radius outer ITBs are only triggered at an input power in excess of 20 MW, but they do not usually survive the transition into H-mode. The best results, in terms of sustained high confinement, have been obtained with neon injection; an outer ITB is triggered during the phase with L-mode edge and survives into H-mode for about 2 s, corresponding to ∼10 times the global energy confinement time τ_E, at q₉₅ = 7.5, H₈₉β_N ∼ 3.5–4 and ∼60% of the Greenwald density limit. In summary, a high triangularity scenario has been developed, which combines the desirable characteristics of controlled edge, long lasting wide ITBs and high performance at density higher than the low triangularity JET scenarios.

Experiments are described that have increased our understanding of the transport and stability physics that set the H-mode edge pedestal width and height, determine the onset of Type-I edge localized modes (ELMs) and produce the nonlinear dynamics of the ELM perturbation in the pedestal and scrape-off layer (SOL). Models now exist for the n_e pedestal profile and the p_e height at the onset of Type-I ELMs, and progress has been made towards models of the T_e pedestal width and nonlinear ELM evolution. Similarity experiments between DIII-D and JET suggested that neutral penetration physics plays an important role in the relationship between the width and height of the n_e pedestal. Plasma physics appears to dominate in setting the T_e pedestal width. Measured pedestal conditions including edge current at ELM onset agree with intermediate-n peeling–ballooning (P–B) stability predictions. Midplane ELM dynamics data show the predicted (P–B) structure at the ELM onset, large rapid variations of the SOL parameters and fast radial propagation in later phases, similar to features in nonlinear ELM simulations.

Due to the high neutron yield and the large plasma size many ITER plasma parameters such as fusion power, power density, ion temperature, fast ion energy and their spatial distributions in the plasma core can be measured well by various neutron diagnostics. Neutron diagnostic systems under consideration and development for ITER include radial and vertical neutron cameras (RNC and VNC), internal and external neutron flux monitors (NFMs), neutron activation systems and neutron spectrometers. The two-dimensional neutron source strength and spectral measurements can be provided by the combined RNC and VNC. The NFMs need to meet the ITER requirement of time-resolved measurements of the neutron source strength and can provide the signals necessary for real-time control of the ITER fusion power. Compact and high throughput neutron spectrometers are under development. A concept for the absolute calibration of neutron diagnostic systems is proposed. The development, testing in existing experiments and the engineering integration of all neutron diagnostic systems into ITER are in progress and the main results are presented.

Neoclassical tearing modes, triggered by the long-period sawteeth expected in tokamaks with large non-thermal α-particle populations, may impose a severe β limit on experiments with large fusion yields and on reactors. Sawtooth destabilization by localized current drive could relax the β limit and improve plasma performance. ³He minority ion cyclotron current drive around the sawtooth inversion radius has been planned for ITER. Several ion species, including beam injected D ions and fusion born α particles, are however also resonant in the plasma and may represent a parasitic absorption of RF power.

Modelling of minority ion cyclotron current drive in an ITER-FEAT-like plasma is presented, including the effects of ion trapping, finite ion drift orbit widths, wave-induced radial transport and the coupled evolution of wave fields and resonant ion distributions. The parasitic absorption of RF power by the other resonant species is concluded to be relatively small, but the ³He minority current drive is nevertheless negligible due to the strong collisionality of the ³He ions and the drag current by toroidally counter-rotating background ions and co-rotating electrons. H minority current drive is found to be a significantly more effective alternative.

The behaviour of high energy runaway electrons during the flat-top phase of lower hybrid (LH) heated discharges has been investigated in the Frascati Tokamak Upgrade. It is found that, as in electron cyclotron (EC) heated discharges, the fall of the electric field during the wave injection phase yields a reduction of the runaway energy (frequently leading to runaway quenching). However, in contrast to EC heated discharges, after the heating phase, in LH discharges the acceleration of the wave-resonant suprathermal tail allows the runaway population to be restored.

Monte Carlo simulation based on test particle representation is carried out in order to investigate the neoclassical effect on a strike point pattern of ions on the local island divertor (LID) head in the Large Helical Device configuration. According to collisionality, the particle orbits mainly contributing the particle flux from the core region to the LID head vary; in the present paper we call it the neoclassical effect on the edge transport phenomena. The pattern on the LID head is numerically observed by tracing the guiding centre orbits of the test particles under the effects of Coulomb collision and anomalous diffusion. We find that the strike point patterns on the LID head are varied according to collisionality. For the case of λ_mfp/L_c > 1, the pattern is characterized by the passing particle orbits along the field lines of the island separatrix, where λ_mfp is the mean free path and L_c ≈ 250 m is the connection length given as a length along a field line connecting the core region to the LID head. On the other hand, for the case of λ_mfp/L_c ≪ 1, the particles intensively strike the edge of the LID head, and the pattern is given by the trapped particle orbits. The trapped particle orbits mainly contribute the particle flux from the core region through the inside of the island, and finally the particles arrive at the edge of the LID head.

Glow discharge cleaning (GDC) is a widely used technique for wall conditioning in fusion experimental devices. Though the cleaning effects of GDC are essentially related to the microscopic modification of the wall surface, there are few reports about it. In the present study, specimens of wall materials were exposed to GDC plasma of helium, hydrogen and neon in the Large Helical Device (LHD) by using the retractable material probe transfer system and irradiation damage was examined by transmission electron microscopy to understand the underlying microscopic mechanism of GDC. In the case of Ne-GDC, the specimen surface was covered with a thick deposited layer of Fe and Cr but no radiation induced defects were observed. Due to the high sputtering efficiency and very shallow penetration, it is likely that neon atoms effectively sputter the surface contamination without leaving serious damage or remaining in the sub-surface region. After the Ne-GDC phase, retained Ne can be successfully removed by the following short hydrogen GDC. It was shown that a two-step GDC with Ne and H is very effective to clean the metallic surface of the LHD.

This paper reports results on measurements of the density profiles. Here we analyse the behaviour of the electron density for a set of experiments in type I ELMy H-mode discharges in ASDEX Upgrade where the plasma current, plasma density, triangularity and input power were varied. Detailed measurements of the radial extent of the perturbation on the density profiles caused by the edge localized mode (ELM) crash (ELM affected depth), the velocity of the radial propagation of the perturbation as well as the width and gradient of the density pedestal are determined. The effect of a type I ELM event on the density profiles affects the outermost 20–40% of the plasma minor radius. At the scrape-off layer (SOL) the density profile broadens while in the pedestal region the density decreases resulting in a smaller density gradient. This change in the density profile defines a pivot point around which the density profile changes. The average radial velocity at the SOL is in the range 125–150 ms⁻¹ and approximately constant for all the density layers far from the pivot point. The width of the density pedestal is approximately constant for all the ELMy H-mode discharges analysed, with values between 2 and 3.5 cm. These results are then compared with an analytical model where the width of the density is predominantly set by ionization (neutral penetration model). The width of the density profiles for L-mode discharges is included, since L- and H-mode have different particle transport. No agreement between the experimental results and the model is found.

The L–H transition in a helical-axis heliotron, Heliotron J, is investigated. For electron cyclotron heating (ECH), neutral beam injection (NBI) heating and ECH + NBI combination heating plasmas, the confinement quality of the H-mode is examined with an emphasis on its magnetic configuration dependence. The vacuum edge rotational transform, ι(a)/2π, is chosen as a label for the magnetic configuration where ι/2π is the rotational transform and a is the average plasma minor radius in metres. The experimental ι(a)/2π dependence of the enhancement factor over the L-mode confinement reveals that specific configurations exist where high-quality H-modes (1.3 < H_ISS95 < 1.8) are attained. is the experimental global energy confinement time and is the confinement time scaling from the international stellarator database given as . R is the plasma major radius in metres, is the line-averaged plasma density in 10¹⁹ m⁻³, P_L is the power loss in megawatts that accounts for the time derivative of the total plasma energy content and B_t is the toroidal magnetic field strength in tesla (Stroth U. et al1996 Nucl. Fusion36 1063). The ι (a)/2π ranges for these configurations are near values that are slightly less than those of the major natural resonances of Heliotron J, i.e. n/m = 4/8, 4/7 and 12/22. To better understand this configuration dependence, the geometrical poloidal viscous damping rate coefficient, C_p, is calculated for different values of ι(a)/2π and compared with the experimental results. The threshold line-averaged density of the H-mode, which depends on the configuration, is in the region of 0.7–2.0 × 10¹⁹ m⁻³ in ECH (0.29 MW) + NBI (0.57 MW) operation. As for the edge plasma characteristics, Langmuir probe measurements have shown a reduced fluctuation-induced transport in the region that begins inside the last closed flux surface (LCFS) and extends into the scrape-off layer. In addition, a negative radial electric field E_r (or E_r-shear) is simultaneously formed near the LCFS at the transition.

Using a gyrokinetic toroidal particle code with global profile effects, the toroidal electron temperature gradient driven (ETG) turbulence in positive and reversed shear tokamaks is studied. In the simulation, initial saturation levels of the ETG mode are consistent with the mixing length theory, which shows a Bohm (gyro-Bohm) like ρ^*-scaling for a ballooning type (slab like) ETG mode in a positive (reversed) shear configuration, where ρ^* is the electron Larmor radius ρ_te divided by the minor radius a. In a realistic small ρ^* positive shear configuration, the ETG mode has a higher saturation level than the large ρ^* positive shear configuration and the reversed shear configuration. In the nonlinear turbulent state, the ETG turbulence in the positive and reversed shear configurations shows quite different structure formations. In the positive shear configuration, the ETG turbulence is dominated by streamers which have a ballooning type structure, and the electron temperature T_e profile is quickly relaxed by enhanced heat transport in a turbulent time scale. In the reversed shear configuration, quasi-steady zonal flows are produced in the negative shear region, while the positive shear region is characterized by streamers. Accordingly, the electron thermal diffusivity χ_e has a gap structure across the q_min surface, and the T_e gradient is sustained above the critical value for a long time. The results suggest a stiffness of the T_e profile in positive shear tokamaks and a possibility of the T_e transport barrier in reversed shear tokamaks.

We present significant advances in the performance and understanding of the sustained spheromak physics experiment (SSPX). High performance discharges are obtained by adjusting the injected current and bias magnetic flux to produce a flat to slightly peaked radial current profile, consistent with reducing the drive for tearing and other MHD modes. Wall conditioning and density control are also essential for obtaining high performance discharges. With these optimizations, plasmas with peak electron temperatures of ∼350 eV, edge magnetic fluctuation levels less than 1% and core thermal diffusivities (χ_e) <10 m² s⁻¹ are produced. Most SSPX temperature data are bounded by β_e ∼ 5% but differentiating between a stability limit and simple energy transport, until now, has not been possible. Achieving these high performance discharges enables us to consider the possibly of auxiliary heating of a spheromak for the first time.

Far Scrape-Off Layer (SOL) and near-wall plasma parameters in DIII-D depend strongly on the discharge parameters and confinement regime. In L-mode discharges cross-field transport increases with the average discharge density and flattens far SOL profiles, thus increasing plasma contact with the low field side (LFS) main chamber wall. In H-mode between edge localized modes (ELMs) the plasma–wall contact is weaker than in L-mode. During ELM fluxes of particles and heat to the LFS wall increase transiently above the L-mode values. Depending on the discharge conditions, ELMs are responsible for 30–90% of the net ion flux to the outboard chamber wall. ELMs in high density discharges feature intermittent transport events similar to those observed in L-mode and attributed to blobs of dense hot plasma formed inside the separatrix and propagating radially outwards. Though the blobs decay with radius, some of them survive long enough to reach the outer wall and possibly cause sputtering. In lower density H-modes, ELMs can feature blobs of pedestal density propagating all the way to the outer wall.

Extensive study of ripple diffusion in strongly inward shifted LHD configurations has revealed the crucial role of the bumpy component. The bumpy component plays a role in increasing surface element around half of the toroidal period through a decrease of the magnetic field strength, which increases [decreases] local radial drift per surface element around half [the beginning and the end] of the toroidal period in such strongly inward shifted configurations. This characteristic causes the lowest ripple diffusion level around a case of R_ax = 3.53 m at vacuum among a wide range of R_ax, which provides the physics basis for the important finding of 'neoclassical transport optimized configuration' in Murakami et al (2002 Nucl. Fusion42 L19). Here R_ax denotes the magnetic axis position.

Stabilization of an m = 3/n = 2 neoclassical tearing mode (NTM) by a local electron cyclotron current drive (ECCD) has been studied in high β_p ELMy H-mode plasmas in the JT-60U tokamak. In the stabilization experiment, the electron cyclotron (EC) injection timing is scanned from the 'before mode onset' phase (preemptive ECCD) to the mode saturation phase (late ECCD). It has been demonstrated that the ECCD is more effective when it is applied before the mode onset and the critical timing for effective ECCD is related to the mode excitation phase. The EC power to prevent the 3/2 NTM excitation by preemptive ECCD is lower by about 20% compared with complete stabilization by late ECCD. Even when the 3/2 NTM is excited, the stabilization effect is stronger than for late ECCD. For a high performance plasma of the normalized beta, β_N = 3.0, the H-factor H_89PL improves from 1.9 to 2.1 when the 3/2 NTM is suppressed. A numerical study has been made based on the modified Rutherford equation coupled with a 1.5D transport code and an EC code. The simulation reproduces well the time evolution of the magnetic island both at the growing and stabilizing phases. The simulation results show that it is important to adjust not only the EC current position but also its width to stabilize the NTM effectively. The reduction of EC power for NTM prevention in experiments for preemptive ECCD can be explained by the increase of the minimum seed island size for NTM excitation.

Compatibility of advanced tokamak plasmas with high density and high radiation loss has been investigated in both reversed shear (RS) plasmas and high β_p H-mode plasmas with a weak positive shear on JT-60U. In the RS plasmas, the operating regime is extended to high density above the Greenwald density (n_GW) with high confinement (HH_y2 > 1) and high radiation loss fraction (f_rad > 0.9) by tailoring the internal transport barriers (ITBs). With a small plasma-wall gap, the radiation loss in the main plasma (inside the magnetic separatrix) reaches 80% of the heating power due to metal impurity accumulation. However, high confinement of HH_y2 = 1.2 is sustained even with such a large radiation loss in the main plasma. By neon seeding, the divertor radiation loss is enhanced from 20% to 40% of the total radiation loss. In the high β_p H-mode plasmas, high confinement (HH_y2 = 0.96) is maintained at high density ( ) with high radiation loss fraction (f_rad ∼ 1) by utilizing high-field-side pellets and argon (Ar) injection. The high is attributed to the formation of strong density ITB. Strong core-edge parameter linkage for confinement improvement is observed, where the pedestal pressure and the core plasma confinement increase together. The measured radiation profile including contributions from all impurities in the main plasma is peaked, and the central radiation is ascribed to the contribution from Ar accumulated inside the ITB. Impurity transport analyses indicate that the Ar density profile, twice as peaked as the electron density profile, which is the same level as that observed in the high β_p H-mode plasma, can yield an acceptable radiation profile even with a peaked density profile in a fusion reactor.

Analysis based on a gyro-kinetic ballooning stability code predicts that both the ion and electron thermal diffusivities, due to the ion temperature gradient (ITG) and electron temperature gradient (ETG) modes, respectively, increase with the safety factor q almost linearly. In the case of ITG driven ion thermal diffusivity, the q dependence originates from the coupling to the ion acoustic mode, and in the case of the electron thermal diffusivity due to the ETG mode, it emerges from the coupling to the skin size drift mode. In the ETG mode, charge neutrality does not hold for typical tokamak discharges, and mixing length estimates yield a thermal diffusivity large enough to be relevant to experiments.

A dominant particle pinch is observed in Ohmic plasmas of the electric tokamak (ET) associated with enhanced poloidal rotation. The density increases dramatically, with strong profile peaking. In pinch dominated particle transport, the pinch velocity profile is determined from the Thomson density profile data. The pinching rate is controlled with soft gas puffing. Hard puffing produces inverted density profiles that do not pinch due to the MHD instabilities. The build-up time of the density is typically 1 s. Due to density accumulation in the absence of significant core fuelling, the characteristic Troyon limit (β_N = βaB/I ∼ 3,%, m, T, MA) is reached even in Ohmic plasmas. Density ramps are terminated by internal disruptions due to beta collapse without any significant radiative energy loss. The loop voltage remains low (0.4 V) during the ramp. Prior to, and during the ramp, we observe no reduction in the electrostatic fluctuations in the present experiments. Electrode biasing, using the J × B force, shows that the density accumulation can be reduced and even stopped through slowing the poloidal rotation (reducing the magnitude of the background negative E_r). This observation is consistent with the presence of a 'viscous' pinch driven by the dominance of the radial electric field through ion mobility. Other neoclassical pinch mechanisms (i.e. Ware and thermoelectric) contribute to the density accumulation and are shown to be secondary effects as revealed by the radial current modulation effects. The easily achieved thermal Mach numbers are ±0.15 for poloidal and ±0.2 for toroidal rotation using the present biaser. There is no significant spontaneous toroidal rotation. The spontaneous poloidal rotation seen in Ohmic plasmas has the thermal Mach number M_p ∼ 0.15 across the measured profile where r/a > 0.5. This rotation is sufficient to account for the observed radial pinch velocity.

The rapid formation of an electron internal transport barrier (eITB) is observed during a slow evolution (≃200 ms) from a centrally peaked to a hollow current density profile, while all external actuators remain constant. The time constant for the barrier formation appears to be shorter than the electron energy confinement time. The improved confinement associated with the barrier formation occurs first in a localized region off-axis. Then the effects propagate to inner and outer flux surfaces on a confinement time scale. The temporal and spatial localizations of the barrier formation suggest a threshold in the magnetic shear profile that triggers the onset of the eITB. The threshold-like behaviour can be attributed to the sudden appearance of a minimum q (or zero-shear) flux surface in the safety factor profile. Based on modelling of the current profile evolution using the ASTRA transport code, the appearance of the zero-shear flux surface is correlated with the temporal and spatial formation of the barrier.

Following the 19th IAEA Fusion Energy Conference (Lyon, 2002), (1) three-time progress in the formation of ion-confining potential heights ϕ_c including a record of 2.1 kV in comparison to those attained 1992–2002 is achieved for tandem-mirror plasmas in the hot-ion mode with ion temperatures of several kiloelectronvolts. (2) The advance in the potential formation gives the bases for finding the remarkable effects of radially produced shear of electric fields E_r, or non-uniform sheared plasma rotation on the suppression of intermittent vortex-like turbulent fluctuations. (i) Such a shear effect is visually highlighted by x-ray tomography diagnostics; that is, spatially and temporally intermittent vortex-like fluctuated structures are clearly observed as two-dimensionally reconstructed visual structures for the first time in kiloelectronvolt order ion-cyclotron heated plasmas having a weak shear in GAMMA 10. (ii) However, during the application of plug electron-cyclotron heatings (ECH), the associated potential rise produces a stronger shear (dE_r/dr = several 10 kV m⁻²) resulting in the disappearance of such intermittent turbulent vortices with plasma confinement improvement. X-ray observations also show elongation of a vortex structure from a circular into an ellipsoidal shape, as depicted in H-mode theories, with an outward shift. (3) For the physics interpretations and control of such potential and the associated shear formation, the validity of our proposed theory of the potential formation is extensionally tested under the conditions with auxiliary heating. The data described above fit well to the extended surfaces calculated from our proposed consolidated theory of the strong ECH theory (plateau formation) with Pastukhov's theory on energy confinement.

Plasma profiles across the separatrix and scrape-off layer (SOL) in Alcator C-Mod are examined for a range of plasma densities, currents and magnetic fields in Ohmic L-mode discharges and for a subset of conditions in Ohmic H-mode discharges. In all plasmas, electron pressure gradient scale lengths ( ) exhibit a minimum value just outside the separatrix (i.e. in the near SOL), forming the base of a weak (strong) pedestal in L-mode (H-mode) plasmas. Over a wide range of conditions in Ohmic L-mode discharges, at this location are found to track with a monotonic function of electron collision frequency, when this quantity is normalized according to the framework of electromagnetic fluid drift turbulence (EMFDT) theory. Moreover, at fixed values of normalized collisionality parameter (characterized as the 'diamagnetic parameter', α_d), electron pressure gradients in the near SOL increase with plasma current squared, holding the MHD ballooning parameter, α_MHD, unchanged. Thus, the state of the near SOL is restricted to a narrow region within this two-parameter phase-space. An implication is that cross-field heat and particle transport are strong functions of these parameters. Indeed, as α_d is decreased below ∼0.3, cross-field heat convection increases sharply and competes with parallel heat conduction along open field lines, making high plasma density regions of α_MHD–α_d space energetically inaccessible. These observations are consistent with the idea that the operational space of the edge plasma, including boundaries associated with the tokamak density limit, is controlled by EMFDT.

Design studies are shown on the National Centralized Tokamak facility, formerly called JT-60SC. The machine design is carried out to investigate the capability for flexibility in aspect ratio and shape controllability for the demonstration of the high-β steady state operation with nation-wide collaboration, in parallel with ITER towards DEMO. Two designs are proposed and assessed with respect to the physics requirements such as confinement, stability, current drive, divertor and energetic particle confinement. The operation range in the aspect ratio and the plasma shape is widely enhanced consistent with sufficient divertor pumping. Evaluations of the plasma performance towards the determination of the machine design are presented.

International collaboration on development of a stellarator confinement database has progressed. More than 3000 data points from nine major stellarator experiments have been compiled. Robust dependences of the energy confinement time on the density and the heating power have been confirmed. Dependences on other operational parameters, i.e. the major and minor radii, magnetic field and the rotational transform , have been evaluated using inter-machine analyses. In order to express the energy confinement in a unified scaling law, systematic differences in each subgroup are quantified. An a posteriori approach using a confinement enhancement factor on ISS95 as a renormalizing configuration-dependent parameter yields a new scaling expression ISS04; . Gyro–Bohm characteristic similar to ISS95 has been confirmed for the extended database with a wider range of plasma parameters and magnetic configurations than in the study of ISS95. It has also been discovered that there is a systematic offset of energy confinement between magnetic configurations, and its measure correlates with the effective helical ripple of the external stellarator field. Full documentation of the International Stellarator Confinement Database is available at http://iscdb.nifs.ac.jp/ and http://www.ipp.mpg.de/ISS.

High performance reversed shear (RS) discharges with strong internal transport barrier (ITB) and flat pressure profile in the plasma core region disrupt frequently even with low beta. We analysed MHD instabilities leading to low beta disruption with measuring fluctuations and current profile with MSE measurement. We mainly observed two types of disruptions. One is the disruption without precursor at surface q ∼ integer. The other is the disruption with n = 1 precursor of γ > 100 ms. The poloidal mode number of the n = 1 mode is equal to the outermost integer of q. The n = 1 mode exists continuously from the peripheral region to the ITB layer or the n = 1 modes exist separately at the peripheral region and at the ITB layer. The phase has a difference of 180° between at the peripheral region and at the ITB layer. To explain these characteristics of disruption, we introduce a simple model where disruption occurs when both MHD instabilities at the plasma surface and at the q = m/n surface in the RS region, where m and n are the poloidal and the toroidal mode numbers of the surface MHD instability, are simultaneously unstable. This simple model can explain almost all observed disruption by two processes. One is the surface mode triggered disruption, which occurs when surface q changes; the corresponding q = m/n surface at ITB layer changes discretely. The other is the internal mode triggered disruption, which occurs when the corresponding q = m/n surface becomes unstable gradually.

With the dynamic ergodic divertor in TEXTOR fundamental effects of the coupling of external magnetic field perturbations to the confined plasma have been studied. The non-linear coupling between external (m/n = 12/4) and internal modes (m/n = 3/1) has been investigated. The critical perturbation field (m/n = 3/1) for the excitation of an m/n = 2/1 tearing mode depends not only on the magnitude but also on the direction of the toroidal angular momentum input by neutral beam injection (NBI). Below the excitation threshold of this mode a toroidal spin-up of the plasma has been observed, which only depends on the strength of the perturbation field. It is independent of both the rotation direction of the external perturbation field and the toroidal angular momentum supplied by the NBI.

The quiescent H (QH) mode, an edge localized mode (ELM)-free and high-confinement mode, combines well with an internal transport barrier to form quiescent double barrier stationary state, high performance plasmas. The QH-mode edge pedestal pressure is similar to that seen in ELMing phases of the same discharge, with similar global energy confinement. The pedestal density in early ELMing phases of strongly pumped counter injection discharges drops, and a transition to QH-mode occurs, leading to lower calculated edge bootstrap current. The plasmas current ramp experiment and ELITE code modelling of edge stability suggest that QH-modes lie near an edge current stability boundary. At high triangularity, QH-mode discharges operate at higher pedestal density and pressure and have achieved ITER level values of β_PED and ν^*.

Internal coils, 'I-Coils', were installed inside the vacuum vessel of the DIII–D device to generate non-axisymmetric magnetic fields to act directly on the plasma. These fields are predicted to stabilize the resistive wall mode (RWM) branch of the long-wavelength external kink mode with plasma beta close to the ideal wall limit. Feedback using these I-Coils was found to be more effective as compared to using external coils located outside the vacuum vessel. Locating the coils inside the vessel allows for a faster response and the coil geometry also allows for better coupling to the helical mode structure. Initial results were reported previously (Strait E.J. et al 2004 Phys. Plasmas11 2505). This paper reports on results from extended feedback stabilization operations, achieving plasma parameters up to the regime of C_β ≈ 1.0 and open loop growth rates of γ_openτ_w ≳ 25 where the RWM was predicted to be unstable with only the 'rotational viscous stabilization mechanism'. Here C_β ≈ (β − β_{no−wall.limit})/(β_{ideal.wall.limit} − β_{no−wall.limit}) is a measure of the beta relative to the stability limits without a wall and with a perfectly conducting wall, and τ_w is the resistive flux penetration time of the wall. These feedback experimental results clarified the processes of dynamic error field correction and direct RWM stabilization, both of which took place simultaneously during RWM feedback stabilization operation. MARS-F modelling provides a critical rotation velocity in reasonable agreement with the experiment and predicts that the growth rate increases rapidly as rotation decreases below the critical. The MARS-F code also predicted that for successful RWM magnetic feedback, the characteristic time of the power supply should be limited to a fraction of the growth time of the targeted RWM. The possibility of further improvements in the presently achievable range of operation of feedback gain values is also discussed.

The PDF file provided contains web links to all articles in this volume.