Table of contents

A new method is described for the preparation of spherical high-pressure hydrogen targets to be used in laser-produced fusion experiments.

The relevance of Hilbert transforms to phase determination in bright-field electron optics using complementary half-plane apertures has recently been pointed out by Misell et al. An alternative approach based on symmetry arguments is presented which yields equivalent results but has conceptional and computational advantages.

The sharp, reversible transition of a superconductor from the normal to the superconducting state gives it obvious switching properties. One use for such a switch would be the protection of power-system components, especially superconducting cables and generators, which are now being developed. The bulk of the paper is concerned with a preliminary design study of such a device.

The required lengths of superconductor are such (similar 1 km) that burn-out at regions of low critical current (I_c) would be probable before the rising fault current in the whole superconductor reaches the average I_c and switching occurs. The proposed switch avoids this difficulty by backing the superconductor with a semiconductor which has a negative temperature coefficient of resistance in the relevant temperature range, thus preventing the catastrophic temperature rise that would occur in the normal superconductor alone because of its positive temperature coefficient. The crucial requirement of the semiconductor is that it should have, just below its melting point, the correct intrinsic resistivity, for which an expression is derived.

An illustrative design for a high-power generator-protection switch is worked out, and problems and incentives are outlined.

It is shown experimentally that a molecular beam maser may be operated with a state selector coated with an electret material, without recourse to an external EHT supply subsequent to electretization.

Neutron wave propagation in a polycrystalline prism is analysed via a velocity-dependent transport equation with a simple scattering kernel. Particular attention is devoted to examining the behaviour of sub-Bragg neutrons, and the effects of temperature and transverse buckling on the neutron waves. The contribution from the sub-Bragg continuum (pseudo-mode) is made possible to compute by the use of the analytical continuation of the dispersion function in the case of 1/v-form cross sections. Numerical results of the energy spectrum and detector response of neutron waves are presented for beryllium and graphite. The interference phenomenon between the discrete modes and pseudo-modes is also demonstrated for various temperatures and transverse bucklings.

This paper considers the physical processes which may occur when a liquid weld metal drop comes in contact with a liquid metal pool. Experiments with drops of aqueous KCl solutions and mercury, combined with a theoretical analysis of the effects of surface tension and Lorentz forces, lead to the following conclusions. Drops larger than a critical size, which is a function of the capillary constant and wire diameter, will transfer to the pool and smaller drops will lead to stable menisci. Surface tension alone will complete the interaction in approximately k(M/4πT)^1/2 s, where M and T are the drop mass and surface tension in CGS units and k similar 3·7. An electric current I can result in a repulsive force on drops of radius less than approximately μ₀I²/16π²T (MKS units). The magnetic pinch instability can dominate surface tension effects under extreme conditions, and magnetically induced streaming can make otherwise stable menisci unstable. Applied to short-circuiting MIG welding systems, these results show how proper attention to the current and weld metal surface tension can lead to reduced spatter and better process control.

In an earlier paper the effects of the thermal boundary layers on the electron temperature distribution near a flat-plate Langmuir probe in an incompressible flow were discussed. It is shown that by the use of the Illingworth-Stewartson transformation these results may be simply extended to deal with compressible flows. The conditions for the transformation are examined and found to be in reasonable agreement with the known behaviour of real gases. A modified electron thermal profile consistent with the transformation is calculated.

The plasma current and wall temperature distributions are evaluated in the precollapse phase using a one-dimensional representation of the current and thermal diffusion equations. It is assumed that the plasma electrical conductivity and the time derivative of the total current are both constant in this early stage of the discharge. The great advantage of a metallic wall construction over a nonmetallic one is immediately obvious from the dependence of wall heating on the physical properties of the wall material. Finally the calculations are applied to estimate the rise in wall surface temperature during the precollapse phase of documented linear, theta and toroidal pinches.

The floating potential (ie the potential assumed by a probe when the net current collected by it is zero) of a spherical electrostatic (Langmuir) probe in a collision-dominated plasma is investigated. The plasma is considered to be either static or flowing at subsonic velocities. The experimentally realistic situation is also considered where the probe is cold compared with the bulk plasma. The two extreme probe temperature conditions are discussed: that in which the electrons remain at their uncooled temperature, characteristic of the bulk plasma; and that in which they cool to the temperature of the neutrals near the probe. It is concluded that only in the case when the electrons cool and the sheath is thin does the floating potential vary appreciably from its value obtained in the absence of thermal gradients and thermoelectric effects. Experimental results are presented which support the concept that, for the opposite of the above condition (ie uncooled electrons, thick sheath), the floating potential is unaffected. In addition, the data have also been analysed in a manner suggested by Lam to yield an accurate value for the electron temperature.

Experimental methods are presented for the detection and elimination of instrumental effects arising from field penetration and space charge for an orifice probe in low-pressure (<2×10⁻² Torr) RF discharge plasmas in argon. A channel orifice, compared with a sharp-edged orifice, provides number density values which are too low. The creation of a dual-peak ion distribution by the insertion of a grid into the plasma is illustrated.

A study has been made of cathode spot structure and dynamics in copper vacuum arcs for currents between 25 and 100 A using a high-magnification optical technique. The cathode spot consists of one or more electron emission zones or fragments with areas fluctuating between 10⁻¹⁰ m² and 5×10⁻⁸ m². The spot generally exists in one of two types. In the second type the average current density is about 5×10¹⁰ A m⁻² and in the first type it is about (0·5-1)×10¹⁰ A m⁻². The cathode spot undergoes continuous rebuilding in the form of fluctuations in the number and size of the fragments such that transitions in spot type, from type 1 to type 2 and vice versa, occur on average every 100 μs.

A possible explanation of the observations in terms of a cathode spot model based on temperature-field electron emission is discussed.

A theoretical investigation of the heating and ionization of low-pressure air by microwave signals is presented in this paper, demonstrating how partially ionized air is first heated and then further ionized by a transient microwave field. Nonlinear models of collision, attachment and diffusion processes are used to calculate the transient change in the electron energy and number density when a microwave field is suddenly applied to low-pressure air. Characteristic times for the heating and ionization processes are presented as functions of field amplitude and air pressure, and are compared with existing theories and experimental data.

Estimates have been made of the absolute photoelectric yield from a pure mercury cathode in a low-pressure (0·9<P₀<5·6 Torr) Townsend discharge in mercury vapour over the range of 160<E_s/P₀<422 V cm⁻¹ Torr⁻¹ These estimates lie between 2·77×10⁻³ and 4·7×10⁻³ electrons/quantum, when non-resonance photons of energy 5·43 eV are incident on a pure mercury pool cathode of workfunction 4·52 eV.

In a number of devices, transverse magnetic fields are used to rotate arcs to minimize electrode erosion. Measurements are reported here for 45 A arcs rotating in air at atmospheric pressure which show that the cathode loss, which exceeds that of the anode, can be determined with good accuracy if allowance is made for oxidation. These results, which have been made over a much wider range of arc velocity than any work reported previously, show that the variation of cathode erosion rate with arc velocity is not a simple one, and that an abrupt fall or increase can occur for a very small change in arc velocity, whereas it has commonly been assumed that increasing the magnetic field will continuously reduce the loss of electrode material. It is suggested that these considerable variations in cathode erosion are associated with changes in electron emission caused by differing oxide film conditions on the cathode surface.

An experimental arrangement for the production of a quasi-stationary, high-current arc is described. It is stabilized by an axial gas flow in a high-pressure environment (current: 1900 A, pressure: 23 atm, gas: nitrogen). The conditions are described under which part of the arc assumes a cylindrical form. For such an arc the radial temperature distribution was measured. Because of the cylindrical shape a relatively simple evaluation yields quantitative data about the local radiative energy balance and the other energy transport mechanisms. This leads, for the first time, to a quantitative understanding of the different physical processes in this arc, and makes it possible to formulate a simplified arc model which is useful in practical applications.

A new mechanism for the development of the spark channel in gases is proposed, in which it is assumed that the shock wave, known to exist after the channel is complete (Freeman and Craggs 1969), actually develops while the channel is still in formation. The paraboloidal shock is thought to derive its power from the transfer of energy only from the electrons pre-heated in front of the developing channel by field concentration to the neutrals and ions behind it (detonation wave mechanism). Neutral gas in front of the channel is considered to be cold. Analytical expressions for the velocity and position of the detonation wave are derived, and experiments by the authors and by previous workers tend to confirm the mechanism for certain low-density gases. The `radiation overshoot' characteristic of shock transmission through some gases has been observed, suggesting that a shock indeed exists prior to completion of the spark channel.

The average velocity and the average degree of ionization of ions emitted from a copper-vapour arc in vacuum have been determined from ion current, energy, and impulse measurements. The average ion velocity found was 1·4×10⁴ m s⁻¹, and on the average these ions are ionized 1·7 times.

Assuming single ionization, about 50% of the atoms emitted are ionized.

It was found that the angular distributions of the emitted neutrals and ions were not equal.

An experimental investigation of the temporal growth, in cylindrical geometry, of ionization currents in crossed electric and magnetic fields has been made in molecular hydrogen at gas pressures where the electron mean free path is much greater than the dimensions of the discharge chamber. An extension of the spatial growth model advanced previously (Blevin and Stock 1973) is proposed which satisfactorily accounts for most of the phenomena observed.

The natural electrical conductivity of several organic liquids has been measured using transformer-coupled ratio arm bridges over a range of temperatures and pressures. Among these, n-hexane, diethyl hexyl sebacate and several alcohols showed a tendency for conductivity to fall with pressure. However, diethyl ether, carbon disulphide and, above 2500 bar, castor oil showed a tendency for conductivity to rise with pressure.