Table of contents

The Magnetically Insulated Inertial Confinement Fusion system combines many of the favourable aspects of both magnetic and inertial fusion in that physical containment of the hot plasma is provided by a metallic shell, while its energy is insulated from the solid wall by a strong self-generated magnetic field. The reactor potential of such a deuterium-tritium (D-T) burning system is examined by utilizing a quasi-one-dimensional, time dependent set of particle and energy balance equations for the thermal components plus an arbitrary number of fast alpha energy groups. Classical and anomalous diffusion is incorporated for particles and energy crossing the magnetic field that separates the core plasma from the 'halo' region, and the energy gain factor Q is calculated. It is shown that when proper choices are made for the size of the ablating fuel shell and the metallic outer shell, Q values in the hundreds and perhaps thousands are obtainable when reasonable values of initial plasma density, temperature, and radius are assumed. These encouraging results are contingent on certain assumptions concerning refuelling and metallic wall interactions that must await further experimentation and simulation studies for validation.

The problem of wave induced resonant ion transport in ICRH tokamak plasmas is considered. A diffusion coefficient which describes the transport in the radial dimension in configuration space is first derived and a bounce averaged Fokker-Planck equation formulated. Then, using the method of multiple time-scales, a diffusion equation for the resonant ion density is obtained. The diffusion equation is numerically solved for parameters appropriate to the JET tokamak. The results indicate that for high RF power input and long heating times (t > τ_s) there is a significant depletion of resonant ions from the centre of the discharge, a reduction in thermonuclear reactivity and a degradation in heating efficiency.

The beam target fusion reactivity for a monoenergetic beam in a Maxwellian target is approximately evaluated for non-resonant reactions. The approximation is accurate for the D-D and T-T fusion reactions to better than 4% for all beam energies up to 300 keV and all ion temperatures up to two-thirds of the beam energy.

The paper presents an extended theory of neoclassical impurity transport, starting from the parameters of bulk plasma toroidal and poloidal rotation. Analytic expressions demonstrating the influence of compressible flow on the perpendicular momentum balance and on the neoclassical Braginsky parallel viscosity are given. The predicted impurity transport is extensively compared with results published previously.

Particle and energy confinement of ions during ion cyclotron heating (ICH) in the centre cell of a tandem mirror is simulated by a Monte Carlo code. The interaction of the ion cyclotron wave with the ions is calculated using reduced equations of motion, and Coulomb collisions are modelled using a Monte Carlo algorithm. An axial electrostatic potential profile is generated by an iteration procedure, assuming Maxwellian electrons. The axial profile of neutral gas which fuels the ions from a localized gas box is determined from a neutral particle Monte Carlo code. Steady state calculations are reported for the Tandem Mirror Experiment Upgrade (TMX-U). The calculated energy confinement time τ_E is about twice that measured for off-midplane fuelling. The calculated τ_E can be decreased to the experimental value by either introducing extra gas at the ICH antenna to model the effect of ions striking the antenna or assuming anomalous radial loss. However, the enhanced radial loss also increases the end loss current well above that measured. Experimental and calculated τ_E agree more closely for midplane fuelling with heating from one antenna.

Numerical simulations of sawteeth in tokamaks have been carried out using reduced magnetohydrodynamics and a simple transport model. The electron temperature is evolved self-consistently, including Ohmic heating and a highly anisotropic thermal diffusivity. The sawtooth period and collapse time found in the simulations for Lundquist numbers S below 10⁷ compare favourably with experimental results from small and medium size ohmically heated tokamaks. The sawtoothing is found to be sensitive to the values of the transport coefficients. In particular, the perpendicular viscosity must be comparable to, or larger than, the perpendicular thermal diffusivity for distinct relaxation oscillations to occur. To study the scaling with S, the viscosity and perpendicular thermal conductivity have been scaled as 1/S, and S has been varied. Modifications of the equilibrium, caused by the sawteeth, play an important role in the scaling of the collapse time and period with the Lundquist numnber. The self-consistently computed q-profiles are very flat in the central region where q ≈ 1. Outside the low-shear region, the shear rises sharply. This q-profile allows a resistive mode to be turned on quickly with a high growth rate. The deviation of the central q from unity over the sawtooth cycle decreases with increasing S, and the collapse time shows a weaker dependence than the S^½ scaling of Sweet and Parker.

A linear resistive magnetohydrodynamic (MHD) stability analysis of a finite beta, cylindrical reversed field pinch (RFP) plasma is presented. The equilibrium distributions are specified by a model describing both the parallel and the perpendicular current density components. The average beta (⟨β⟩) of the configurations is varied by using the parameter χ (0 ≤ χ ≤ 1) which ensures that Suydam's necessary condition for stability is satisfied. It is found that there are configurations which are stable with respect to ideal pressure driven modes at rather high ⟨β⟩ values (up to 30%), although small changes in the equilibrium can produce a drastic reduction in the maximum achievable beta value. In this model, in which viscosity and Hall term effects are not taken into account, resistive g-modes are in general found to be always unstable, independent of the beta value. However, by considering only modes with 'sufficiently' high growth rates, it is possible to deduce some stability boundaries and 'beta limits' for relatively low values of the magnetic Reynolds number.

The stability of high (m, n) ideal ballooning modes in a toroidal plasma experiencing a rigid rotation about its axis of symmetry is investigated. Centrifugal forces induce poloidal asymmetry in the plasma pressure and affect the position and form of the magnetic surfaces. This leads to stabilization of the ballooning modes in all but the outer plasma regions (i.e for r < L_p, where L_p is the pressure scale length). Stabilization is possible for rather modest values of the parameter , where ∊ is the inverse aspect ratio, V₀ is the rotation velocity and C_s is the sound velocity. The analysis also indicates that plasma rotation can help gain access to the second stability regime in high beta experiments.

Relaxed (force-free) toroidal plasmas ('Taylor states') are of interest in fusion physics. The zero flux condition yields eigenvalues connected with the maximum toroidal current. Up to now, these eigenvalues have only been calculated for infinite aspect ratios and circular cross-sections or for rectangular cross-sections of the torus. Here, the zero flux and some zero field eigenvalues are calculated for toroids of arbitrary aspects ratios and arbitrary cross-sections such as circles, ellipses or Cassini curves.

The energy loss by radiation is estimated in REPUTE-1 by bolometric measurements. Burning out of the radiation barrier is observed in both ultra-low-q (ULQ) and reversed field pinch (RFP) discharges. The conditions for burning out of the radiation barrier are studied for various operation parameters. A reduction of the radiation loss is observed after wall conditioning by carbonization.

For current profiles with a slope discontinuity at ∊₂ close to ∊₁[q(∊₁) = 1] and a small shear ₁(∊₁), the jump of the derivative of the magnetohydrodynamic (MHD) solution across the singular q = 1 layer is of the form ∊ is the finite inverse aspect ratio, μ ≡ ₁∊₁/8(∊₂–∊₁) measures the ratio of the two small parameters. The growth rate of the sawtooth trigger can accordingly pass from resistive values either negative and finite or positive) to MHD values for slight variations of either ₁ or (∊₂–∊₁)/∊₁. The above profile properties have been observed in sawtoothing TEXTOR discharges.

Long wavelength resistive MHD modes are studied using a three-dimensional stability code and experimentally determined equilibrium profiles of DIII-D H-mode discharges. Calculated magnetic signals at more than two dozen poloidal locations show good agreement with the experimental measurements, both in relative amplitudes and phases, when the vacuum vessel is treated as conducting and its exact non-circular shape is taken into account. Simulations with a simple finite Larmor radius model suggest that the prevailing mode structure of the pressure driven resistive mode might change from short wavelength to long wavelength (from toroidal mode number n = 3 to n = 1) as the plasma pressure is increased, consistent with experimental observations.

A layer of TiD₂ (5 × 10¹⁹ TiD₂ cm⁻²) was prepared by deuterium implantation into a polycrystalline titanium foil at room temperature. The foil was subsequently installed in front of a large area surface barrier detector in a vacuum of 10⁻⁷ mbar to measure the reaction products of the two possible D-D fusion reactions. After a counting period of 30 hours, at most two fusion products were found, which limits the reaction probability to below 8 × 10⁻²⁴ fusion processes per second and per deuterium pair in the implanted layer.

Early in 1988, an agreement was reached by the European Community, Japan, the United States of America and the Soviet Union to jointly conduct conceptual design activities for the International Thermonuclear Experimental Reactor (ITER) until the end of 1990, under the auspices of the International Atomic Energy Agency. Since May 1988, participants from the four Parties meet regularly in Garching, Federal Republic of Germany, to carry out the design work. On the basis of the investigation results obtained so far, a concept for ITER has been defined which incorporates the maximum possible flexibility of the device and allows a variety of plasma configurations and operating scenarios to be adopted. For the technology experiments, with a full breeding blanket, the device can be operated typically with a plasma carrying a current of 18 MA at a major radius of 5.5 m. For the plasma physics experiments, the device can be configured with a thinner shield, if required, and it can produce a plasma of 22 MA with fully inductive operation and higher currents under limited technical conditions. A list of important specific physics and technology research and development tasks for ITER has been prepared and these tasks are being