Table of contents

It is pointed out that the accidentally resonant charge-exchange reaction, O⁺ + H⁰ ⇄ O⁰ + H⁺, is an important mechanism for causing the loss of singly charged oxygen ions from oxygen-contaminated hydrogen plasmas. Results of a Monte Carlo simulation are presented which show that the fraction of oxygen lost because of charge exchange exceeds 1/3 when the parameters n_e ∼ 10¹³ cm⁻³, n_H⁰ ∼ 10¹¹ cm⁻³ and T_e ∼ 3 eV are attained. This process may be important in determining impurity losses from the edge plasmas of tokamaks.

For a toroidal sheet pinch, analytic relations are derived for β_p, the position of the magnetic axis, the poloidal magnetic field and the critical β_p values. Comparisons with numerical solutions show that the Garching Belt Pinch equilibria can very well be approximated by an infinite cylindrical slab model. The β_p limits for the occurrence of current reversal and a biaxial configuration may provide a better understanding of the whole class of strongly elongated equilibria with "flat current distributions".

Experimental results from a large number of tokamaks have been collected. It is shown that the measured energy confinement times are consistent in the sense that they may be represented (accurately) by simple empirical formulas. These formulas are used in predicting the Ohmically heated temperatures in future tokamaks. The predicted temperatures are similar to but slightly lower than those in present machines. T_e0 is about 0.5 to 1 keV. This result emphasizes the point that substantial additional heating will be required for the attainment of ignition temperature.

Measurements of temperature rise of the wall and of a movable limiter have been made in the Oak Ridge Tokamak (ORMAK). For a normal discharge without injection, approximately 70% of the energy in the discharge is deposited on the wall. Very little energy is deposited on a point limiter until it is moved several centimetres inside the circular limiter, indicating that the body of the plasma is not in contact with the limiter and that diffusion and heat conduction do not transfer a large fraction of the plasma energy to the limiter.

The dispersion relation for the dissipative trapped-electron mode including the effects of magnetic shear, finite ion Larmor radius, banana-plateau velocity space separation, magnetic curvature drift resonance, ion-ion collisions, etc. is used to examine the plasma stability of recent tokamak experiments. Regimes of importance of various effects in these experiments are also discussed. Instabilities mostly occur at k_θ ρ_i ∼ 1, where k_θ is the azimuthal wave number and ρ_i is the ion Larmor radius. The often used anomalous heat conduction coefficient, , predicts losses comparable to those observed in TFR, ST, and ATC.

Re-cycling is studied in the DITE tokamak by changing the working gas from hydrogen to deuterium and back to hydrogen. The change in the composition of the plasma is followed both during the discharge and from shot to shot by using a scanning Fabry-Perot interferometer, a mass spectrometer and a carbon thermal desorption probe. The results of the three techniques are consistent and show that the gas trapped and desorbed from the wall plays a major part in the re-cycling process. A detailed analytical modelof the principal re-cycling mechanisms is proposed.

The loss rate of singly charged ions from an end-stoppered mirror trap is discussed by using a linearized Fokker-Planck equation and the associated variational principle. A comparison is made with the earlier work of Pastukhov and with numerical results.

Multigroup discrete ordinates methods avoid many of the approximations that have been used in previous neutral transport analyses. Of particular interest are the neutral profiles generatedas an integral part of larger plasma system simulation codes. To determine the appropriateness of utilizing a particular multigroup code, ANISN, for this purpose, results are compared with the neutral transport module of the Düchs code. For a typical TFTR plasma, predicted neutral densities differ by a maximum factor of three on axis and outfluxes at the plasma boundary by ∼ 40%. This is found to be significant for a neutral transport module. Possible sources of the observed discrepancies are indicated from an analysis of the approximations used in the Düchs model. Recommendations are made concerning the future application of the multigroup method.

A summary is given of the linear theory of the ideal and resistive hydromagnetic stability of tokamaks. The first section provides an introductory account of the various aspects of the stability problem, and the subsequent sections provide a survey of the subject and a review of the literature. For aperfectly conducting plasma the modes of instability are of three types: kink, internal, and axisymmetric. When resistivity is introduced the kink and internal modes have significantly modified forms. The analysis of the standard tokamak, having a large aspect ratio, circular cross-section and low β, is almost complete but the study of small aspect ratio, high-β configurations and the optimization of such configurations are at an early stage.

The prospects for relatively "clean" D-³He reactors (lean D, rich ³He) are evaluated with significant gains over a comparable power DT reactor (14-MeV neutrons reduced by 135, all neutrons by 26, plasma tritium by ∼ 10³, tritium handling by ∼ 10³.