Table of contents

A fine scale drift wave mode possessing unusually high frequency 2–3 MHz and radial wave number is observed using the correlative upper hybrid resonance backscattering technique at the FT-2 tokamak under conditions when the ETG mode should be unstable. The radial wave number spectrum of turbulence is measured and shown to be maximal at values 120–170 cm⁻¹ corresponding to the largest ETG instability growth rate.

Two-dimensional edge plasma structures of the l = 2 helical system CHS have been investigated with the use of a neutral lithium beam probe. Two different types of edge magnetic configuration have been compared, namely, an inboard-wall limiter configuration and an ergodic magnetic divertor configuration which is intrinsic to non-axisymmetric helical devices. The scrape-off plasma in the limiter configuration is essentially one-dimensional and the radial scale length is well described by a simple diffusion model with finite connection length of the magnetic field line. While in the ergodic divertor configuration, edge plasma structure shows up–down asymmetry in the steady state near the outer separatrix region regardless of the symmetric magnetic structure. The asymmetry reverses when the magnetic field direction is reversed. Plasma shift is apparently in the direction of ion B × ∇B drift. The result suggests that plasma equilibrium in the ergodic layer is not determined simply by the magnetic field structure. The effect of plasma flow and drift motions would be important. Such asymmetry in plasma structure may induce non-uniform heat deposition on the divertor plates. Modelling of plasma equilibrium in the ergodic layer is necessary for practical helical divertor design.

Second harmonic 60 GHz electron cyclotron (EC) pre-ionization using low field side (LFS) X-mode launchers has been effective in producing target plasmas for startup of the DIII-D tokamak with electron densities comparable to fundamental LFS pre-ionization plasmas using the same EC system. A visible Bremsstrahlung array showed that breakdown occurred at the 2nd harmonic resonance location and after a few milliseconds the EC driven plasma filled the entire vessel, independent of the resonant location, which was varied from near the inner wall to the centre of the torus. The power threshold for ionization of ∼0.4 MW was observed in DIII-D for these 2nd harmonic pre-ionization experiments. An orbit following code calculated that cold (0.03 eV) electrons could be heated to energies above 20 eV where the ionization of the neutral deuterium gas can occur. Scaling from DIII-D to ITER indicates that electron cyclotron heating (ECH) 2nd harmonic pre-ionization and initial plasma formation is possible if ITER operation at reduced toroidal field is desired, but additional experiments are required to extrapolate the entire ECH start-up scenario from DIII-D to ITER.

IPP Garching has successfully developed a RF-driven negative ion source for the ITER neutral beam injection system. The RF source is now an interesting alternative to the reference design with filamented sources due to its in principle maintenance-free operation. Current densities of 330 A m⁻² and 230 A m⁻² have been achieved for hydrogen and deuterium, respectively, at a pressure of 0.3 Pa and an electron/ion ratio of 1 for a small extraction area (7.0 × 10⁻³ m²) and short pulses (<4 s). Reliable deuterium operation with more than 150 pulses in the required parameter range was obtained by an improved caesium operation utilizing a control of all source temperatures (grid as well as source body) and monitoring the amount of caesium in the source. The development now concentrates on extending the pulse length to up to 1 h and extending the size of the source. The long pulse test bed went into operation last year; pulses of up to some 100 s with more or less stable conditions in terms of extracted currents and Cs dynamics have been achieved with current densities in the range of 150–200 A m⁻² in hydrogen operation. The pulse length, however, is still limited by non-sufficient cooling of some parts of the source and the RF circuit. The commissioning of the so-called half-size source test facility started recently with the first plasma pulses; this large RF source with the width and half the height of the ITER NNBI source is dedicated to demonstrate the homogeneity of a large RF plasma—extraction is foreseen in a later phase—and the tests of an ITER-relevant RF circuit.

The excitation mechanism and stabilization of the observed m/n = 1/1 instability during the sawtooth ramp phase in the lower hybrid current driven plasmas on the HT-7 tokamak (Sun Y et al2005 Plasma Phys. Control. Fusion47 745–56) have been investigated in this paper. This mode is most likely destabilized by the current density gradient at the q = 1 surface. Based on the local transport model, it is shown that the saturation and stabilization of the m = 1 mode before the subsequent sawtooth crash may result from the flattening of the current profile because of the locally enhanced transport of the suprathermal electrons due to the magnetic island.

Internal transport barriers (ITBs) in the ion channel in the tokamak ASDEX Upgrade allow for high energy confinement but collapse after only several energy confinement times. In this paper we show that in most cases the ITB phase is terminated clearly before the first ELM burst, thereby ruling out the ELMs as the main trigger of the ITB collapse. For the first time, the ITB formation and sustainment are found to be associated with a mechanism of transport suppression based on thermal ions dilution by the injected fast ions. Interestingly, such ITBs do not require reversed magnetic shear. The linear growth rate of the ion temperature gradient driven mode is computed as a function of the fast ion fraction with gyrokinetic stability analysis. Monte Carlo simulations predict the fast ion population to be above the gyrokinetic critical fraction in a region consistent with the experimental ITB width. The density threshold documented for the onset of ASDEX Upgrade ion ITBs is explained. The role of T_i/T_e and of the plasma sheared rotation for ITB sustainment are analysed. The stabilization mechanism presented here is consistent with the observed ITB lifetime of the order of the beam slowing down time. A possible runaway mechanism leading to ITB collapse is described. Finally, the relevance of this particular ITB scheme for ITER is discussed.

Pellet injection is considered the most promising technique to fuel ITER plasmas. However, standard low field side (LFS) pellet injection seems inadequate for an efficient fuelling of the plasma core because it is too demanding in terms of injection velocity. Alternative configurations, as injection from the high field side (HFS) or vertically from the top of the machine (VHFS) are technologically less demanding due to a drift of the ablated material down the magnetic field gradient, which helps its penetration. In the past year LFS and VHFS pellet injection experiments have been performed on the Frascati Tokamak Upgrade (FTU: B_T = 8 T, Ip = 1.5 MA, R = 93.5 cm, a = 30 cm), a circular cross-section device with a molybdenum toroidal limiter located on the HFS. The rationale of these experiments was to study the advantages of VHFS injection in a high toroidal field machine, where the value of B_T can be considered as representative of that in ITER. The results demonstrate that the main advantage of VHFS injection (with respect to LFS) is that it allows one to obtain the same post-pellet density profile with an injection speed 2–3 times lower than that required in the case of LFS injection. However no significant displacement of the ablated material was observed, possibly due to the small value of the ∇ B-induced drift in these ohmic, low β plasmas. The experimental measurements were compared to the predictions of the drift model (Pégourié B. et al2007 Nucl. Fusion47 44) showing a reasonable agreement in both the magnitude of the drift and the main features of the line integrated density during pellet ablution.

The stability of peeling, ballooning, and peeling–ballooning modes, which relate to edge-localized modes, is investigated numerically with the linear ideal magnetohydrodynamic stability code MARG2D. The effects of 'sharpness' on the stability of the peeling–ballooning mode are examined, where the sharpness is defined in terms of the curvature at the top or bottom of the outermost flux surface. It is found that the stability limit of the pressure gradient significantly improves as the sharpness increases even when the ellipticity and the triangularity are unchanged. The sharpness is an important parameter for high performance H-mode operations with high pedestal pressure.

Transitions to improved core electron heat confinement are triggered by low order rational magnetic surfaces in TJ-II electron cyclotron heated (ECH) plasmas. Experiments are performed changing the magnetic shear around the rational surface n = 3/m = 2 to study its influence on the transition; ECH power modulation is used to look at transport properties. The improvement in the electron heat confinement shows no obvious dependence on the magnetic shear. Transitions triggered by the rational surface n = 4/m = 2 show, in addition, an increase in the ion temperature synchronized with the increase in the electron temperature. Ion temperature changes had not been previously observed either in TJ-II or in any other helical device. SXR measurements demonstrate that, under certain circumstances, the rational surface positioned inside the plasma core region precedes and provides a trigger for the transition.

The prediction of impurity peaking in future fusion devices such as ITER necessitates the study of the dependence on Z of the impurity transport in present devices. In this paper we describe a novel technique to determine the transport of impurities with different atomic numbers independently. A technique has been developed that allows simultaneously the measurement of the transport of Ne and Ar in the same discharge while minimizing the systematic errors in the spectroscopic measurements. The reproduction of the charge-exchange measured densities, absolute vaccum ultra-violet line intensities and absolute soft x-ray intensity is achieved in an impurity transport simulation. The method used to estimate the errors on the transport coefficients of neon (Ne) and argon (Ar) is presented. In the plasma region where the diffusion and convection coefficients are determined for hybrid discharges, the transport of Ne and Ar is observed to exceed neoclassical predictions. In the same regions, the diffusion coefficients of both impurities are similar. The convection coefficients are also comparable for Ne and Ar. The peaking of Ne and Ar density profiles are comparable during the period where the intermittent slow reconnecting n = 1 mode is stable in these hybrid discharges.

A better energy-confined state with modified current density profile is achieved by fast biasing (applied voltage rise time ∼ the ion–ion collision time) an electrode placed inside the limiter in very low q_a discharges of the Saha Institute of Nuclear Physics–tokamak (SINP–Tokamak) (Lahiri et al1996 Nucl. Fusion36 254). The plasma current increases and lasts for an extended period. Profile modification of the radial electric field (E_r) in the edge region is also observed in the better-confined state. However, j_Φ-profile modification as well as energy confinement time improvement occurs prior to E_r-profile modification.

The impacts of plasma disruptions on ITER have been investigated in detail to confirm the robustness of the design of the machine to the potential consequential loads. The loads include both electro-magnetic (EM) and heat loads on the in-vessel components and the vacuum vessel. Several representative disruption scenarios are specified based on newly derived physics guidelines for the shortest current quench time as well as the maximum product of halo current fraction and toroidal peaking factor arising from disruptions in ITER. Disruption simulations with the DINA code and EM load analyses with a 3D finite element method code are performed for these scenarios. Some margins are confirmed in the EM load on in-vessel components due to induced eddy and halo currents for these representative scenarios. However, the margins are not very large. The heat load on various parts of the first wall due to the vertical movement and the thermal quench (TQ) is calculated with a 2D heat conduction code based on the database of heat deposition during disruptions and simulation results with the DINA code. For vertical displacement event, it is found that the beryllium (Be) wall does not melt during the vertical movement, prior to the TQ. Significant melting is anticipated for the upper Be wall and the tungsten divertor baffle due to TQ after the vertical movement. However, its impact could be substantially mitigated by implementing a reliable detection system of the vertical movement and a mitigation system, e.g. massive noble gas injection. Some melting of the upper Be wall is anticipated at major disruptions. At least several tens of unmitigated disruptions must be considered even if an advanced prediction/mitigation system is implemented. With these unmitigated disruptions, the loss of the Be layer is expected to be within ≈30–100 µm/event out of a 10 mm thick Be first wall.

The Large Helical Device (LHD) is not only the largest stellarator for the research of fusion plasma near a reactor region but also the largest superconducting system. Availability higher than 98% has been achieved in the long-term continuous operation both in the cryogenic system and in the power supply system. It is due to the robustness of the systems and efforts of maintenance and operation. One big problem is the shortage of cryogenic stability of a pair of pool-cooled helical coils. Composite conductors had been developed to attain sufficient stability at high current density. However, it was revealed that a normal-zone could propagate below the cold-end recovery current by additional heat generation due to the slow current diffusion into a thick pure aluminium stabilizer. Besides, a novel detection system with pick-up coils along the helical coils revealed that normal-zones were initiated near the bottom of the coil where the field is not the highest. Therefore, the cooling condition around the innermost layers, the high field area, will be deteriorated at the bottom of the coil by bubbles gathered by buoyancy. In order to raise the operating currents, methods for improving the cryogenic stability have been examined and stability tests have been carried out with a model coil and small coil samples. We selected a method to lower temperatures of the coil and an additional cooler has been installed at the inlet of the coil. The outlet temperatures of the coil have been successfully lowered to 3.8 from 4.4 K of the saturated temperature, as planned.

This report summarizes the contributions presented at the 11th EU–US Transport Task Force workshop on transport in fusion plasmas, held in Marseilles, France, 4–7 September, 2006⁸The present workshop: http://www-fusion-magnetique.cea.fr/ttf2006.. There were sessions on momentum transport, multi-scale physics, electron transport, particle transport and transport in the scrape-off layer.