Table of contents

Initial hydrogen pellet injection experiments have been performed in plasmas characterized by a low particle re-cycle fraction resulting from the action of a poloidal magnetic divertor. As in past experiments, the interaction of the fuel with the plasma is observed to be adiabatic. Moreover, the pellet mass is accounted for in the core plasma density increase, indicating only a small loss of fuel while the pellet transits the divertor scrape-off plasma. The effect of edge re-cycling on the density was studied by comparing divertor (low-re-cycle) and limiter (high-re-cycle) plasmas; the distinction between the two cases is clearest in the edge plasma region where the density decay rates differ most. Particle transport subsequent to pellet injection is less ambiguous for divertor cases, and, by comparing the density profile relaxation and the electron temperature recovery with an empirical transport model that closely approximates the pre-injection plasma conditions, it is concluded that the plasma confinement properties do not deteriorate as a result of pellet injection. The principal difference between central and edge fuelling is demonstrated by a peaking of the density profile and an extended decay time for the density perturbation.

The paper considers the effect of anisotropie pressure on plasma stability in closed magnetic traps against internal kink mode (m = 1) perturbations. It is shown that for p_|| > p_⊥ anisotropie pressure can suppress this instability mode.

A model analyses the effect of secondary electron emission and gas recycling on open-ended fusion devices. Secondary emission is assumed to be limited by space-charge effects. It is also assumed that cold plasma is created by the ionization of gas recycled from the end wall. The axial power flow in the device was increased by these phenomena. This article shows that fairly simple techniques can control the enhancement of the power losses. The results of this model are in agreement with measurements made on the TMX experiment. Estimates made for the TMX Upgrade and MFTF-B indicate that these effects will not significantly enhance the axial power losses.

It has been shown that the minimum-energy state of a diffuse pinch discharge is that described by the so-called 'Bessel Function Model'. The structure of this state is related to a limiting case of the 'Tangled Discharge Model', and it is shown that random currents normal to the magnetic field will prevent the attainment of this state unless the conductivity is large (in the strict sense, infinite). Estimates of the minimum-conductivity required are given.

The dependence of the measured radiation losses in ETA-BETA II on plasma density and plasma current rise time is discussed. A high-density limit is found, corresponding to large radiation. Above this limit, burning through the light impurities (mainly oxygen) does not take place and a cold, radiation-dominated discharge is obtained. This high-density limit scales with the plasma current, indicating a lower limit on the I/N parameter similar to the tokamak high-density limit. With longer current rise times the radiation energy loss during the set-up phase increases about linearly with rise time, corresponding always to about 50% of the total energy losses during this phase.

A Poynting's Theorem method is used for evaluating the volt-second consumption in a tokamak discharge. The method accurately identifies the inductive and resistive components of the volt-second consumption, and allows both quantities to be determined from magnetic measurements made outside the plasma. Only simple computational techniques are required. Application of the method to typical Doublet III near-circular plasmas (R = 1.43 m, a = 0.44 m, b/a ≅ 1.2) indicates that the flux at the plasma surface required to establish the current flat-top is 2.0 ± 0.2 V·s/MA. Approximately 40% of this flux is consumed in resistive dissipation. This division between resistive and inductive flux differs significantly from that obtained using an alternative data analysis method in which the resistive loss is evaluated at the plasma axis. The reasons for the difference are discussed.

An array of emitting and collecting probes was used to measure the characteristics of the plasma at the end wall of TMX. The emitting probes indicated a plasma potential with a characteristic value of approximately 8 V, measured at 2 cm and 10 cm from the grounded wall. The potential in the plug varied from 300 to 1000 V. The collecting probes indicated a cold plasma present near the wall, with an electron temperature of about 5 eV and a cold-electron density of about 2 × 10⁹ cm⁻³, while the hot-ion density at the wall was about 6 × 10⁸ cm⁻³. The density of electrons in the plug was about 2 × 10¹³ cm⁻³. The electron temperature measured at the wall was decoupled from the temperature of the electrons in the plug and the central cell. The plasma wall sheath was established to balance the flux of hot ions and cold electrons. An additional species of electrons, i.e. secondary electrons produced at the end wall, was also identified. These secondary electrons had a directed energy comparable to the local plasma potential, an energy spread of less than 7 eV, and a density of about 4 × 10⁸ cm⁻³. The experimental results are shown to be consistent with a simple model involving cold-gas recycling from the end wall.

Efficient operation of a moving-ring reactor requires inductive energy extraction from the plasma after the thermonuclear burn is complete. The process of inductive energy extraction from a plasma is studied in this paper. An energy conservation theorem is proven, whose content is that the amount of energy which may be extracted inductively from a plasma is the sum of three terms: the self-inductance energy of the plasma, the energy associated with the mutual inductance between the plasma and the external field coils, and the plasma thermal energy. These terms are evaluated for a spherical Hill's vortex equilibrium. For this equilibrium, the sum of the first two terms is negative, i.e. for fixed external field the magnetic energy is lower when the plasma is present than when it is absent. This surprising fact is discussed and shown not to lead to any unphysical consequences. A careful analysis of a conceptual energy extraction device is given, with emphasis on the role of inductive interactions between the plasma and the external coils.

First measurements of the charge-exchange ion temperature in the Heliotron-E device are described. The limiting temperatures calculated from a simple energy balance equation are presented and compared with the measured temperatures. In low-current plasma, the measured temperatures agree well with the thermal conduction limit based on neoclassical theory, which includes an effect of the electron-to-ion temperature ratio T_e/T_i. In high-current plasma, however, the ion temperature is about one-third of the limiting temperature. This discrepancy may be due to anomalous heat and particle transport driven by the plasma current.

Detailed numerical simulations of the radial temperature and density profiles and energy confinement times measured in Pulsator (2 × 10¹³ < ⟨n_e⟩ < 7 × 10¹³ cm⁻³), and in the high-density regime of Alcator C (10₁₄ <⟨n_e⟩ < 4.5 × 10¹⁴ cm⁻³) and FT yield a uniform scaling of the transport in Ohmically heated tokamak discharges. The anomalous electron thermal conductivity can be described by a modified Coppi-Mazzucato law, χ_e [cm²·s⁻¹] = 4.5 × 10¹⁰ (B_t [G], n_e [cm⁻³], T_e [ke V], A_i ion atomic mass), where the l/(T_e q) dependence cannot be distinguished from a law. The particle diffusion is given by the neoclassical term and an anomalous contribution of (0.1–0.3) χ_e. The transport due to sawtooth oscillations is treated by an enhanced, Bohm-like diffusion within the q = 1 surface. In this model, the apparent saturation of the energy confinement time with density observed especially in Alcator C can be explained without anomalous enhancement of the neoclassical ion heat conductivity. An explanation of the saturation under the usual Alcator-scaling χ_e ∼ l/n_e would require a five-times enhancement above the neoclassical value. The increase in ion heat conductivity due to the toroidal-field ripple is found to be negligible in the regimes considered.

Experimental data for ion cyclotron resonance heating in the RFC-XX machine in IPP-Nagoya are presented. The achieved ion temperature is as high as 100 eV at n = 10¹³ cm⁻³ and 1 keV at n = 10¹² cm⁻³. The ion energy confinement becomes worse by the application of a longer pulse, which is found to be due to the enhanced charge-exchange loss and/or electron drag. Axially and azimuthally arrayed antennas are used in the heating, and the importance of the phasing is demonstrated. A simple model of the multiple-antenna problem is also given and used to interpret the experimental data.

The feasibility of pumped limiters has been demonstrated on the Macrotor tokamak using both small neutral collection probes (2.5 cm diameter) and a large-area limiter (20 cm × 60 cm). Neutral pressures due to particle accumulation have reached as high as 8 microns. The particle removal efficiency of the neutral collection probes was measured to be 17% and estimated to be greater than 50% in the large-area limiter case. A quadratic dependence of the pressure on plasma density is observed.

It is shown that two typical kinds of experimental evolution of a re versed-field-pinch plasma (RFP plasma) can be interpreted fairly well by using a partially-relaxed-state model. The RFP plasma tends to evolve along the curve of constant beta on axis in the F — θ diagram. A proposal of how to achieve a stable RFP plasma with a higher β-value by using an F — θ stability diagram is presented.

Simple and relatively accurate relations describing collisional slowing-down of ions by monoenergetic electrons are derived, and the dependence of the slowing-down of ions on the electron velocity distribution is investigated. It is shown that the commonly used stopping-power formula is inapplicable for the description of slowing-down of slow ions in non-equilibrium plasmas.

It is known that during the heating of a warm-background ion species by a more energetic ion species, the warm-ion distribution function departs from that of a Maxwellian. It is also known that the velocity dependence of the particle and energy loss mechanisms is a crucial factor in determining the manner in which the bulk ions deviate from Maxwellians. The question of the extent to which these distortions cause an enhanced fusion reactivity in a TFTR-like plasma is addressed. For the cases presented here, no enhancement is found.

The plasma scrape-off layer in a tokamak limiter-controlled discharge involves sonic ion flow down a long narrow flux tube. Small perturbations to such flows can cause large fluctuations which can in turn induce local charge separation, i.e. double sheaths, and further strong fluctuations. It is suggested that these various fluctuations may contribute to the anomalous transport generally observed in scrape-off plasmas. Recent Langmuir probe measurements of fluctuations in the DITE tokamak are consistent with this conjecture.

The velocity space loss regions of a heliotron system under the influence of an electric field are studied. The adiabatic invariants J and are used to study the drift surfaces of localized, blocked and transit particles. A radial electric field limits some of the drift surfaces inside the plasma column and thus reduces the area of the loss region.