Table of contents

Experimental results are presented which show that the MHD surface waves or fast waves of m=+or-1 have been excited in the range of frequency less than ion cyclotron frequency using a helical antenna in a finite beta and cylindrical plasma. Their dispersion relation and spatial structure were successfully confirmed after the wave signals were separated from those of the local shear Alfven waves that propagated together with the surface waves. In order to obtain the dispersion relations from the signals, the authors adopted three kinds of methods: (1) conventional cross-power spectrum density using fast Fourier transform, (2) maximum entropy method, and (3) direct estimation of the phase velocity of wave packets. The observed properties of both the surface waves and the SAWs are in general agreement with theory.

The edge plasma density versus plasma radius function of the MT-1 Tokamak plasma is measured by a new laser blow-off method. A thin film of sodium evaporated on a glass substrate is brown off by a Q-switched ruby laser pulse. The enhanced shortening of the pulse of neutrals was observed along the beam propagation toward the plasma center by measuring the resonance light intensity of atoms excited by the plasma electrons. The density of the plasma is calculated from the measured exponential time decay of the blow-off-light pulse without any further calibrating measurements.

The author obtains a nonlinear amplitude equation in slab geometry which describes the growth and saturation of a magnetic island. This is done by extending the earlier work of P. Rutherford (1974) to include nonlinear terms. The width of the saturated island is shown to be determined by the current profile, for a nonsymmetric profile. The symmetric case is also considered.

The constitutive relation of a non-homogeneous plasma for h.f. waves in the ion cyclotron frequency domain is derived by integrating the Vlasov equation along the unperturbed particle orbits in the drift approximation. The integro-differential wave equations and the corresponding power balance equation in Tokamak geometry are obtained by expanding the h.f. current to second order in the ratio of the ion Larmor radius to the typical perpendicular wavelength; terms proportional to the much smaller ratio of Larmor radius to the length characterizing equilibrium gradients are neglected. This allows the author to cast the FLR wave equations into an explicitly vector form, and to identify clearly the physical meaning of each term. By specializing to a plane stratified model of the Tokamak, the author compared the complete set of FLR wave equations with a considerably simpler set previously derived by Swanson (1981) and by Colestock and Kashuba (1983), which has been extensively used for analytic and numerical studies of ICR heating. The author concludes that at least in the important scenario of a hydrogen minority in a deuterium plasma (and of pure first harmonic heating as a particular case), the use of the reduced set of equations is well justified, provided that finite Larmor radius contributions to electron absorption are taken into account.

There is a critical current I_PB of about 1 MA (the Pease-Braginskii current) at which Ohmic heating and Bremsstrahlung losses balance in a Z-pinch under pressure equilibrium. An analytic zero dimensional model shows the process of radiative collapse when the prescribed current exceeds the critical current. In particular for a linearly rising current radiative collapse is complete when the current is square root 3 I_PB. However in practice the voltage limitation imposed by an external circuit prevents such a total collapse, and by including this in the model a maximum density ( approximately 10³⁰-10³² m^-3) can occur followed by an expansion and damped oscillation about an equilibrium at which the current equals the Pease-Braginskii current. In the absence of alpha-particle pressure the maximum density is limited by the resistance of the narrow column, the large voltage across which ( approximately 10⁸ V) is balanced essentially by a large negative LI; it occurs when the current is I_PB (( delta -1)/( delta -2))^1/2 where delta =7/3+4/3 ln (R_w/a), where a is the pinch radius and R_w is the radius of the current return. The minimum current following maximum density is shown to be greater than I_PB/ square root 2. Degeneracy effects can be included in the model.

A charge exchange recombination spectroscopy (CXRS) diagnostic has been established on JET to study fully stripped low-Z species. Ion temperature in the plasma centre is measured from visible lines of helium, carbon and oxygen excited by charge exchange with heating neutral beam particles. Coincident cold components produced at the plasma edge are apparent on helium and carbon spectra and most spectra are subject to accidental blending from other species' edge plasma emission. The charge exchange feature can be isolated from the various composite lines and all three impurities agree on the same temperature within experimental error. Observed column emissivities are converted into absolute impurity densities using a neutral beam attenuation code and charge exchange effective rate coefficients. The models take detailed account of cascading and the influence of the plasma environment in causing l-mixing, and allow the n-dependence of the rate coefficients to be addressed experimentally. Some illustrative results are shown for helium (helium discharge or minority r.f. heating), carbon and oxygen concentrations monitored during characteristic operating regimes.

Radial electron thermal and particle diffusivities have been measured using transient methods yielding chi _e and D_e simultaneously in the same spatial region in the plasma interior. Three methods have been applied, analysis of inward propagation of temperature and density perturbations caused by injection of a small pellet into the plasma, the outward propagation of electron temperature and density pulses caused by sawteeth, and thirdly time-dependent transport analysis applied to non-stationary plasmas. In Ohmically heated deuterium plasmas, limited by the outer carbon belt-limiters, the authors deduce chi _e=2.9+or-0.4 m² s^-1 and D_e=0.4+or-0.2 m² s^-1, giving chi _e/D_e=7.2+or-3. The large value of chi _e/D_e would seem to exclude E*B convection as a dominant mechanism in the observed radial thermal transport in JET.

Experimental measurements using a long pulse relativistic electron beam of about 800 ns duration and peak energy of about 200 keV were used. Numerical calculations, for the determination of the trajectories of the beam electrons, were also carried out using a model for the self-fields based on measurements. The beam diamagnetic field, radial loss at the cusp etc. were calculated and compared with the experimental results. Axial beam diamagnetism was observed in the precusp region. The external cusp field was seen to be drastically modified by the self fields. The peak radial loss of beam electrons in the cusp region was seen to occur along the shifted cusp plane. Total radial loss in the cusp region was seen to depart from the single particle value for self fields comparable to the external field. The postcusp beam diamagnetism was also seen to be enhanced over the value estimated from the single particle description in the range of beam self fields comparable to and larger than the external field.

Emission and absorption of high harmonics radiation in a Tokamak plasma is investigated. The interplay of spatial inhomogeneities of the plasma parameters and toroidal variation of the magnetic field is discussed for horizontal and vertical propagation.

Very limited diagnostic techniques are available for the study of flux and charge state distribution of impurities and plasma ions in the Tokamak scrape-off layer. The author describes an in situ plasma ion mass spectrometer (PIMS) which provides real time information on the charge-to-mass ratio of the ions in the scrape-off layer. The instrument, which is being developed on the DITE Tokamak, is based on the cycloidal focusing system and achieves a mass resolution of approximately=20. Results are presented which demonstrate the performance of the device. The observed charge state distributions are compared with theoretical predictions and an explanation for the near equality of the oxygen and carbon fluxes is presented.

Flute-type fluctuations excited spontaneously at the edge of a strongly magnetized plasma column couple nonlinearly with an externally injected convective cell. During the interaction the azimuthal mode-number spectrum of the fluctuations is strongly deformed, and a broader spectrum appears. The phase and amplitude of the fluctuations vary after the interaction and eventually an ordered structure is formed with a phase locked to the injected cell. The macroscopic behaviour is analyzed numerically, using a simple model equation of the Van der Pol type. The observed ordering of the fluctuations is well explained by this type of equation, both for the case when the fluctuations consist of a single mode and when they consist of two modes of different frequencies.

Helical structure and hot spots were observed in a Mather-type plasma focus operated at 3 KJ of bank energy. The experiments were carried out with the help of a fast optical framing camera and two X-ray pin-hole cameras with different filters. It was observed that initially a conical pinch (with base diameter of 6 mm and length of 14 mm) with temperature of approximately=10² eV was formed. This pinch disintegrated after approximately=50 ns by a single lobe sausage instability into a central high temperature ( approximately=10³ eV) filament of 1 mm diameter and 8 mm length containing a high emissivity helical structure. This helix is probably responsible for generation of axial magnetic field and relaxation of the focus pinch. Hot spots of high X-ray intensity and temperature ( approximately=10³ eV) were also observed much beyond the filament region.