Table of contents

A simple model, based on the diffusion of impurity particles in the momentum space, is presented to simulate the distribution of laser blow-off injected aluminium species in the plasma of the MT-1M tokamak. The main emphasis is placed on the toroidal dispersion of the low-ionised particles just after the injection. The model is realized in a Monte Carlo computer code. Taking into account only two physical processes, namely the diffusion in the momentum space and the ionization of the injected atoms, good agreement with the experimental results can be demonstrated. A new method for the plasma edge electron density determination, using the widths of the toroidal distributions, is also proposed.

Expressions suitable for numerical evaluation, are given for the parameters characterizing the locations of the R-cutoff (upper X-mode cutoff), the L-cutoff (lower X-mode cutoff) and the O-mode cutoff for electromagnetic waves in relativistic plasmas. The ion motion is ignored, and it is assumed that the plasma is collisionless and in thermodynamic equilibrium. A new derivation of the relativistic cutoff expressions, starting from first principles, is given, together with a heuristic explanation for the relativistic shift of the O-mode cutoff.

The decay of an ion temperature discontinuity in a collisionless plasma is considered. Theoretical analysis and simulation are used. A shock is formed in the cold plasma region and a rarefaction wave is formed in the hot plasma region. Profiles of density and potential are presented. A comparison of the theoretical dependence on the initial hot ion temperature of the hot ion density, temperature, velocity, current and energy flux at the plateau region and the shock speed with the simulation data shows good agreement.

Inhomogeneities in Tokamak peripheral plasmas due to localized recycling of neutrals (plasma fuelling) are considered. A dense cloud of cold charged particles arises in the vicinity of such a localized particle source, for example a gas-puffing valve, as the plasma density exceeds a certain critical level. A dependence of the critical density on such parameters as dimensions of the device, heat flux from the discharge core and diffusivity of charged particles at the edge is found. The results of calculation are compared with experimental data.

The particle and energy reflection coefficients of fusion alpha -particles with energies from 100 eV to 100 keV from the surfaces of some first wall materials are calculated by the improved bipartition model for light ion transport. The influence of the surface barrier upon the reflection of alpha -particles is investigated. The calculated data also include the dependence of the reflection coefficients on the angle of incidence. The comparisons of present results with those of relevant experiments and Monte Carlo simulations show that the bipartition model may yield reasonably accurate reflection data for light ions.

A theoretical analysis of radio-frequency current drive in a tokamak at frequencies much less than the electron cyclotron frequency is presented. The global model includes kinetic hot plasma effects and collisions. All kinds of wave-plasma interactions (resonant wave-electron interaction, resonant wave-ion interaction and collisions) can contribute to the current drive. The analysis discusses the following current drive components: the momentum-transfer current and polarization current, the helicity current, the resonant and non-resonant currents, the currents caused by kinetic and collisional effects. The two most important new current drive components are introduced, the resonant helicity current and the collisionless part of the non-resonant current. Helicity current drive is proportional to the parallel helicity flux. The helicity conversion coefficient, expressing the conversion of the parallel helicity flux into the parallel induced current, is defined by the wave-plasma interaction. The collisionless part of the non-resonant current is proportional to wave damping on ions. However, this is an electron current, and wave-ion interaction just creates necessary polarization of the wave to drive a current. Circularly polarized waves drive a current differently from linearly polarized waves, and wave polarization is an important factor for current drive efficiency.

The radial electric field structure in the fusion reactor having the Torsatron/Heliotron configuration is discussed, taking into account the effects of the orbit loss of alpha -particles. The direct loss of alpha -particles has only a small influence on the profile of the radial electric field. The trapping efficiency of the alpha -particle is also evaluated by using the self-consistent radial electric field.

The total radiated power has been measured under different operating conditions in TdeV. Radiative losses from the bulk plasma in standard discharges, which amount to about 45% of the ohmic input power in the limiter mode of operation and about 25% in the divertor mode of operation, could be reduced by up to 70% by boronizing the chamber walls. In limiter mode, the boronization alters the radiation profile from a flat profile to a hollow one. This change results from elimination of the metallic impurities by masking the exposed metallic surfaces in the chamber with boron. In experiments conducted to control the plasma impurity content by plasma biasing, negative (or positive) biasing can decrease (or increase) the radiated power from the bulk plasma by up to 25% (40%) within the biasing range of about -200V to +300V. This effect is believed to have resulted from the influence of plasma biasing on the particle transport in the scrape-off layer, which in turn alters the impurity source. The observed trend in variations of the impurity content in the main plasma, which is the main source for the radiated power, is reproduced, using a simple 1D transport model.

A ray-tracing model for studying ion cyclotron resonance heating at higher cyclotron harmonics in very hot large plasmas is presented. It includes the full hot dielectric tensor for bi-Maxwellian distribution functions with parallel drift retaining temperature effects to all orders and allows for non-Maxwellian distribution functions. The model only retaining first-order temperature corrections underestimates the electron damping on the Bernstein wave. Highly energetic tails efficiently absorb fast wave power at high cyclotron harmonics. Significant alpha -particle absorption is to be expected in D-heating D-T scenarios.

The analysis of extended X-ray absorption fine structure (EXAFS) data by Djaoui et al. (1989) led to normalised estimates of the value of Lambda , the ion correlation parameter, partly due to the uncertainty in the ionization state of the Al plasma. The author adopts simplified version of the interatomic force model of Laughlin (1986) to include the effect of electron screening and shows how to estimate the temperature dependence of the Debye-Waller factor for a dense plasma system. This allows the experimental values of the Debye-Waller factor to be used to give limits on both the temperature and, via the Thomas-Fermi ionisation formula of More (1985) on Lambda . A subsequent assumption allows estimates of the actual correlation parameter, which are physically reasonable for the experiment in question. The author discusses how it will be necessary to develop a model using radial distribution functions to examine properly the question of temperature estimation from EXAFS experiments in the dense plasma regime.

The existence of poloidal flow transforms the elliptic Grad-Shafranov-Schluter (GSS) equation into an EGSS system (Extended GSS) of partial differential equation and an algebraic Bernoulli's equation. The EGSS system becomes alternatively elliptic ad hyperbolic as the Mach number of the poloidal flow increases with respect to the Alfven speed of the poloidal magnetic held. A computer program for solving EGSS equations in elliptic regions using the inverse method and Fourier decomposition has been prepared. The solutions in the first and second elliptic regions have been found for different plasma cross-sections, not necessarily up-down-symmetric. The solutions in different elliptic regions exhibit the significant differences in the poloidal magnetic field configuration and the shifts between magnetic axis and density axis differ in sign and magnitude.

A computer simulation is carried out to evaluate the accuracy of the ion temperature determined from the collective Thomson scattered spectrum, when a high-power, multimode D₂O laser beam is incident. The simulation results reveal the per cent bias error B<1% and per cent standard deviation D<6% when a single-mode incident beam of duration tau =1 mu s is scattered by a plasma with n_e=10¹³ cm^-3, T_e=T_i=300 eV, and the scattered spectrum is detected with a frequency resolution Delta f=10 MHz. The values of B and D increase considerably when the scattered spectrum for a multimode incident beam is fitted with theoretical spectrum for single-mode incidence. However, they reduce to almost the same values as those for the single-mode incident beam when the theoretical spectrum for the same multimode incidence is fitted in estimating T_i.

Considers the temporal behaviour of fast ions produced in a tokamak plasma by the injection of a periodically pulsed neutral beam. If the temporal distance between the pulses is properly chosen, then a dense bunch of fast ions is formed near a certain resonant magnetic surface. The temporal and spatial behaviour of these bunches is studied in some detail for the simple model of a tokamak with large aspect ratio. Possible implications of this phenomenon for diagnostics are discussed.

It is shown how one can measure at the same time both the electron line density and the poloidal magnetic field in a Tokamak, using only a polarimeter in order to determine the full state of polarization of radiation transmitted across the plasma. The proposed polarimeter is simpler and more flexible than the more conventional interferometer/polarimeter used on various Tokamaks.

A strong instability has been discovered caused by the interplay of two factors: (i) the gradient of the electron temperature across the magnetic field lines and (ii) the formation of Debye sheaths at the conducting end walls terminating the plasma along the magnetic field. The instability can be of some importance for the physics of tokamak scrape-off layers as it can determine the layer thickness. The authors generalize a previous analysis to the case of non-orthogonal intersections of the magnetic field with the wall surface, i.e. to the situation which is most typical for a tokamak with a limiter. The new effects which appear are caused by the change of the flux-tube length (between the walls) in the course of its displacement in the radial direction. Depending on the sign of the scalar product n.grad T_e, where n is a unit vector normal to the wall, there can occur either suppression of the initial instability or its considerable enhancement. The possible influence of these effects on the transport coefficients in the scrape-off layer is evaluated. Under realistic assumptions on plasma parameters, this influence can be quite significant.

Transport-relevant fluctuations of the electron temperature can in principle be measured by radiometry of the electron cyclotron emission (ECE), but they might be buried completely in natural fluctuations of the ECE due to the thermal nature of this radiation. The spatial coherence properties of thermal radiation can be utilized in correlation experiments to overcome the principal difficulties and to make temperature fluctuations accessible to measurement. On the basis of these considerations an intensity interferometer was built and installed on the Wendelstein 7-AS stellarator. The interferometer consists of two identical but independent multichannel heterodyne radiometers viewing the same emitting plasma volume along crossed lines of sight. The angle between the sightlines is chosen sufficiently large to decorrelate the natural fluctuations of the radiation field. A quasithermal microwave radiation source is used to test the intensity interferometer and to simulate the correlation experiment. Measurements of the spatial coherence and the statistical properties are in agreement with theoretical predictions.