Table of contents

The injection of 10-100 keV Li⁰ diagnostic beams into magnetically confined fusion plasmas causes collisionally induced Li I emission at 670.8 nm, in close relation to the edge plasma electron density. A numerical method for quantitative reconstruction of the plasma density exclusively from relative Li I 670.8 nm emission profiles as measured along the diagnostic beam has been developed, involving all relevant collisional interactions of the Li atoms with plasma constituents. The applicability of the described algorithm is illustrated by experimental results obtained for the TEXTOR Tokamak edge plasma at KFA Julich.

Azimuthal spectra of unstable curvature-driven flute modes in the gas-dynamic trap (GDT) are found to be consistent with theory when including the finite larmor radius (FLR) terms. Growth rates of the modes have been measured for different values of pressure-weighted curvature of the field lines inside the trap and also for different ion pressures changed with auxiliary ion cyclotron heating. A transition from the spectra dominated by the mode of rigid displacement of the plasma column to that where a few modes have approximately the same growth rates was observed.

Electron density oscillations in a linear RF discharge operated with helium as the filling gas are measured by means of a 4 mm microwave interferometer, the microwave beam passing the discharge along a diameter. The azimuthal wave numbers and the direction of rotation are determined from ICCD camera images which are shot end-on. The wave number parallel to the magnetic field is obtained from the phase relation of modulated light emission received side-on at different axial positions. The oscillation frequency is typically lower than both the ion cyclotron frequency and the frequency of momentum transfer collisions between ions and neutrals. A sharp onset of oscillations accompanied by a weakening of plasma confinement is observable provided some critical magnetic field strength has been exceeded.

Several present day Tokamak experiments have reported large plasma rotations (of the order of sound speed) induced by neutral beam injection during auxiliary heating. The measured rotation profiles indicate that velocity shear can be significant. The authors examine the effect of such sheared flows on the equilibrium of Tokamak plasmas. Using the generalized MHD equilibrium equations for axisymmetric rotation and adopting a variational formulation, the authors carry out a detailed numerical investigation of the form of the magnetic surfaces and other equilibrium quantities as a function of the flow function profile. In certain limits they also obtain analytic expressions for the ellipticity and finite shift of the magnetic axis.

A general analysis is presented of the sensitivity of reflectometry to perturbations of the plasma profile, using a full-wave description in one-dimension. The square of the wavenumber is allowed to have an imaginary part, to account for absorption or divergence of the wave. Correlation reflectometry is investigated. It is found that the phase correlation is substantial, regardless of the correlation length of the fluctuations, unless either the wave attenuation is substantial or else the fluctuation correlation function is non-monotonic, corresponding to narrow-band turbulence. Curves are presented that provide fairly general information for the purpose of experimental interpretation.

An ion-temperature-gradient driven instability is investigated in the Columbia Linear Machine (CLM). A transit-time RF technique is used to heat the core of the plasma column. This produces the peaked ion temperature profile with the parameter eta _i= delta ln T_i/ delta ln n exceeding the critical value ( approximately 1.4) required for onset of the instability. Along with the peaked temperature profile, a new feature in the density fluctuation spectrum is observed. This mode has an azimuthal mode number m=2 and propagates in the ion diamagnetic drift direction. It is identified as an eta _i mode. The mode is similar to the one studied previously in the CLM using DC biased meshes for heating the ions.

It is shown that parallel ion viscosity and thermal conduction can drive instabilities in a plasma with a sheared parallel flow delta (U.B⁰)/ delta r; the thresholds are generally well below that of the ideal Kelvin-Helmholtz instability; the relationship between mode number and frequency is compatible with that of the Toi mode (Toi et al., 1989, Phys. Rev. Lett, 62, 430). A quasi-linear calculation shows that the variation of the total, i.e. laminar and turbulent, parallel momentum and energy inside a thin annular volume, localized in the vicinity of the plasma edge, are matched by the differences of anomalous momentum and energy fluxes carried by turbulence autocorrelations through the boundaries: this corresponds to total momentum and energy conservation. If the rotation velocity is close to threshold, two instabilities are found to be self-stabilizing when the theory is particularized to self-similar solutions; they indeed slow down the rotation. As for the other two, the turbulence at first destabilizes the plasma further while spinning up the rotation.

Empirical studies of the scaling of Tokamak energy confinement times with machine parameters constitute a useful point of contact with physics-based transport theories. They also form the basis for the design of next-step and reactor grade Tokamaks. In most cases a simple power law (or sometimes offset linear) functional form is fitted to the data. Such linear regression techniques have the advantage of computational simplicity, but otherwise have little a-priori justification. Neural networks provide a powerful general purpose technique for nonlinear regression which exhibits no essential limitations on the functional form which can be fitted. The authors apply neural networks to the problem of energy confinement scaling and they illustrate the technique using the data from the JET (Joint European Torus) Tokamak. The results show that the neural network approach leads to a substantial improvement in the ability to predict the energy confinement times as compared with linear regression. The significance of this result is discussed.

The electrostatic character of the self-excited drift wave in a magnetized plasma column is demonstrated. The level of the self-excited magnetic fluctuation is small. With the help of external coils fixed onto the glass tube a rotating field component perpendicular to the steady state axial magnetic field is produced and the influence on the self-excited electrostatic drift wave spectrum is studied. Positive frequency shifts, modulation of the drift-wave amplitude and phase, suppression of the drift wave and mode competition and locking effects are studied. Furthermore, enhanced plasma losses are observed which increase approximately with the square of the amplitude of the oscillating magnetic field.

The effect of nonaxisymmetry on Alfven modes in toroidal geometry is investigated numerically. A model equation is used which simplifies the analysis on the resonant surfaces. Alfven modes, characterized by their poloidal and toroidal node numbers (m,n), are found to exist in moderate nonaxisymmetry as well. For fixed (m,n), however, the mode collapses from a global feature on the resonant surface into an infinitely thin Alfven ballooning mode along a field line, if the nonaxisymmetry exceeds a critical threshold or, with given nonaxisymmetry, if the poloidal variation does so. Alfven balloonings are polarized within the magnetic surfaces and are stable.

Deals with the possibility of generation of longitudinal oscillations in the field of low frequency dipole pumping. Analysis is carried out using Kaw's dispersive equation (1976). An incremental and nonlinear frequency shift for oscillations in the electron-ion plasma, has been obtained. In the concluding section of the paper, the authors analyze the instability of certain concrete modes on the basis of the obtained results.