Table of contents

Large scale, DC type plasma convection is investigated in the JFT-2M tokamak. In the plasma edge region, E*B type convective flows are observed, which are caused by interactions with both a poloidal limiter and an RF wave injection. Observation of changes in the edge potential during lower hybrid (LH) wave injection indicated that the confinement improvement in the LH phase is due to the reduction of an intrinsic E*B type convection. These results are used in the construction of a model that explains both the L and H modes found in the tokamak NBI experiments. The model consists of both a deterioration mechanism based on the obstacle induced convection and a recovery due to the increase of the electric conductivity along the magnetic field. Agreement between the model and typical experimental results is shown

The open divertor in DIII-D has been modified by the installation of a baffled plenum with a narrow entrance aperture. Neutral particle pressures in the range of 2-20 mtorr have been measured in this new plenum during experiments in L and H mode plasmas in which the neutral beam power was varied and the divertor target location was scanned with respect to the entrance aperture. These pressures exceed the predicted minimum required to provide adequate particle exhaust for controlling the density in H mode discharges when pumping experiments are conducted in the future. The pressure measurements presented here, which were carried out in the absence of a pump, were modelled with the aid of two transport codes: an edge plasma code and a neutrals code. It was found in this modelling process that the divertor flux amplification factor required to explain the measured pressures is of the order of 30 to 40, much higher than the factors used in the modelling of data in the absence of the baffle. This higher flux amplification factor results from increased recycling at the divertor owing to the presence of the baffle

An account is given of experiments on the DITE tokamak, using a fast reciprocating set of combined Langmuir and magnetic probes in the same discharge to compare the propagation characteristics of the turbulent electrostatic and magnetic perturbations with those expected from the measured local electric field and pressure gradient, and to investigate the effect of shear on cross-field transport. In addition, the relation between the electrostatic and magnetic aspects of the turbulence is investigated using direct cross-spectral measurements. The results are consistent with a picture in which magnetic fluctuations in the edge are due to the currents driven by the basically electrostatic turbulence, whose own propagation behaviour is dominated by the effects of the sheared E*B/B² velocity

A light ion minority heating scheme based on long wavelength fast magnetosonic wave (FMSW) global resonance excitation appears to be promising for open traps. This scheme has been studied numerically using the 2D code PLFEM. This code uses both local and nonlocal approximations and can search for global resonances. The global FMSW resonance width (TE_1,0,2 mode), which is caused by hydrogen minority cyclotron energy absorption, is found to be a maximum for a minority concentration of C_H=1-2%. For these concentration values a substantial discrepancy between the local and nonlocal approximations occurs. This discrepancy is a result of local Alfven resonance excitation in the vicinity of the hydrogen cyclotron zone in a cold plasma. The Alfven resonance disappears when even a slight thermal motion of the ions (nonlocality) is taken into account. By proceeding with global resonance field patterns a suitable antenna system consisting of two loops fed with opposite phases has been chosen. Calculations have demonstrated that the antenna system has good electrotechnical qualities

The material transport by erosion and redeposition at the plasma facing wall areas in high temperature plasma experiments has been studied using limiter-like carbon probes with well defined surface deposits and depth markers. The probes have been exposed in the scrape-off layer (SOL) of the Joint European Torus (JET) during single discharges. For the evaluation of these experiments a computer program, ERO, has been developed. The calculated erosion-deposition rates for carbon as a function of the distance to the last closed flux surface (LCFS) agree well with the experimental results. For a single ⁴He JET discharge, erosion yields of 530 Å for the silicon deposit and 80 Å for vanadium have been measured near the LCFS. A large amount of redeposited silicon (about 17% of the sputtered atoms) has been found on the probe surface in co-deposition with carbon on an area not favoured by the proposed model. This observation can be explained by an additional force on the impurity ions (e.g. the existence of a local electric field), which may cause the deposition

The penetration of deuterium pellets into ohmically heated TORE SUPRA plasmas has been studied for various plasma conditions (n_e = 1 × 10¹⁹-6 × 10¹⁹ m³, T_e = 1-4 keV) and pellet characteristics (N_p = 2 × 10²⁰-1.3 × 10²¹ atoms/pellet, V_p = 0.6-2.4 km/s - i.e. an increase of about 1 km/s from the velocity range covered by the available data for pellet studies). The measured penetration depths compare well with the predictions of the NGS model. A refined NGPS model is presented, in which the plasma channel radius is computed self-consistently and the heating of the neutral cloud by the 'cold' plasma sheath is taken into account explicitly. When the value of n_e is sufficiently high, it fits the experimental penetration values well, and the computed matter deposition profile compares well with the measured H_α signal. The same model has been shown to fit the experimental scaling laws deduced from the JET and ASDEX data

A new version of indirect drive target is proposed for heavy ion inertial fusion (HIIF), in which the cavity for radiative symmetrization is filled with a low density low Z material and is called a 'hotraum'. When heated to a temperature T ⩾ 100 eV, the hotraum becomes transparent to thermal X-rays and ensures radiative smoothing of the energy flux which implodes the fuel capsule enclosed in the cavity. Ion beams are focused on the full target cross-section and deposit their energy in the high Z casing and in the outer region of the hotraum. The initial target configuration is spherical and requires no special orientation in the reactor chamber. Enhanced hydrodynamic efficiency of the capsule implosion, resulting from tamped ablation inside the hotraum, compensates for the high losses in the hotraum and leads to a beam-to-fuel energy coupling of about 5% and target gains in the range of 50-100. Hydrodynamic mapping of deposition nonuniformities onto the fusion capsule can be avoided by fast initial heating of the hotraum. A regime is found in which the ablating capsule insulates itself from hydrodynamic disturbances in the hotraum. The low density required for the hotraum fill allows stopping ranges of up to 100 mg/cm², corresponding to 6 GeV ²⁰⁹Bi ions

The symmetry of implosion of the hotraum target that has been proposed by J. Meyer-ter-Vehn and the author is analysed in the case when the target is driven by a nonspherical ensemble of heavy ion beams. The ion beams, each with a final focus radius r_bf ⩾ R (R is the target radius), are all assumed to irradiate the target at the same angle α₀ and symmetrically with respect to the equatorial plane. Separate one dimensional (1-D) simulations for the pole and equator target sectors indicate that the l = 2 mode in the asymmetry of fuel implosion vanishes for α₀ ≈ 36°. The l = 4 mode can also be eliminated by adjusting the ion current profile across each beam. It is conjectured that the higher asymmetry modes with l ⩾ 6 will be smoothed out by radiative symmetrization to ≲ 1%. A 1-D average (over the target latitude) energy gain of G = 57 is calculated

Antenna loading measurements carried out during high power ion Bernstein wave (IBW) heating experiments on the DIII-D tokamak indicate that efficient, direct coupling to the IBW at ω ≲ 2ω_ci as predicted by linear coupling theory did not occur. Whereas strong peaking in the loading resistance near cyclotron harmonics is predicted for high edge densities (ω < ω_LH|_edge), the observed loading resistance is nearly independent of the toroidal magnetic field. The loading anomaly can be explained by invoking the ponderomotive force which decreases the edge density immediately in front of the antenna, thus allowing coupling to the cold plasma lower hybrid wave (LHW). A linear LHW coupling code including the modified density profile due to the ponderomotive force reproduces the measured dependence of antenna loading on toroidal magnetic field, edge density, antenna frequency and antenna phasing. Further evidence for the ponderomotive force is obtained from reactive loading measurements which indicate that the plasma is pushed away from the antenna as the radiofrequency power level is increased. The results indicate that the lack of central ion heating observed during DIII-D IBW experiments may be due to a lack of efficient mode transformation from the coupled LHW to a centrally propagating IBW, possibly as a result of nonlinear mechanism(s)

Recent far infrared collective scattering measurements on the DIII-D tokamak during H mode operation have provided a more complete understanding of the L to H transition, and have confirmed the interpretation of previous reflectometry data. The frequency spectra of fluctuations measured during the neutral beam heated L and H mode discharges in DIII-D are dominated by E × B Doppler shifts. This fact is utilized to enhance the spatial localization of scattering data through knowledge of the measured variation in the radial electric field, E_r. Scattering data interpreted in this way indicate a rapid (≈100 μs) reduction in the edge density fluctuation level coincident with the L to H transition, thereby quantifying and confirming the interpretation of previously obtained reflectometer data. The reduction in the RMS electron density fluctuation level (n_e) is ⩾50%. Additionally, the presence of increased electric field shear is observed in the region where the fluctuations are strongly suppressed, as determined from the reduction and broadening of the scattered auto-power spectrum associated with the plasma edge inside the separatrix. Finally, the poloidal wavenumber spectrum has been characterized across the L to H mode transition, indicating that there are no significant modifications to the spectral width over the H mode parameter range investigated

The Fokker-Planck equation is solved for alpha particles in a tokamak, including both pitch angle scattering and slowing down. Losses of the toroidally trapped particles due to stochastic diffusion related to magnetic ripple are accounted for by assuming that all the stochastic orbits leave the device immediately. This assumption permits the problem to be solved by the introduction of a loss cone corresponding to the stochastic region of the velocity space. The eigenvalue problem is solved numerically, to show that the additional stochastic loss of energy due to scattering into the loss cone substitutes about 30% of the birth energy for typical fusion parameters

A database of disruptive discharges in ASDEX has been set up. It contains data for all 33509 ASDEX shots. The statistics of disruptions on ASDEX are evaluated for all these shots; in a more detailed analysis, different operation regimes are regarded separately. It is found that after eliminating the most dangerous operation regimes, 17% of all shots end with a disruption. There is an indication that a well established discharge may be repeated several times at a lower disruption rate (less than 5% for ELMy H modes). The statistics of the flux expulsion at a disruption indicate that a major disruption is characterized by a typical current profile change of Δl_i ≈ 0.37. It is found that about 50% of all disruptions have clear precursors lasting for more than 50 ms, namely locked modes or minor disruptions which in a future machine, might be used to trigger the application of stabilization or shutdown schemes. From the statistics of the radial and vertical displacements during a disruption the average forces on the vacuum vessel are calculated which, in the case of ASDEX, are not critical, but may be of concern for future large devices

Simultaneous measurements of the poloidal magnetic field distribution in JET by means of the Faraday and motional Stark effects are reported. The experiments were designed to minimize the errors resulting from a noncircular flux surface geometry and from a limited spatial resolution. Both techniques show that full reconnection of the flux inside the q=1 surface does not occur during a sawtooth collapse