Table of contents

Observations were made of the temporal variations of conduction currents through point-to-sphere and sphere-to-sphere gaps of insulating oil and of n-hexane under high direct voltages. Current bursts were distinguished from the normal fluctuations of the quasi-steady current component. Long-duration bursts were similar to some previous observations, but many short bursts had the polarity opposite to that of the quasi-steady component.

A low-velocity microparticle gun has been developed for a controlled study of the momentum and charge reversal associated with particle bouncing in a high-voltage vacuum gap for impact velocities of 5-200 m s^-1. It has been demonstrated that there are a range of experimental conditions for which this mechanism can lead to an enhancement in the kinetic energy of a microparticle. This observation is used in support of a multi-transit model for the initiation of electrical breakdown by the impact of microparticles having velocities in excess of 1500 m s^-1.

A shock wave is found to occur behind a driven planar detonation in a shock tube where the piston motion is provided by the contact surface between the detonated and expanding driver gas. Pressure and luminosity records taken in an oxyacetylene mixture, at an initial pressure of 30-50 Torr, show that the secondary shock has similar properties to those predicted by Lee's theory (1965); it is concluded, therefore, that the shock arises because of flow divergence behind the detonation front. Divergent flow in detonations confined in tubes is known to originate through the influence of the wall boundary layer, which effectively causes a negative or outward displacement of the flow boundary. By using this nozzle flow model, which is the basis of velocity deficit theory, measured secondary shock strengths and positions are found to be in agreement with the theoretical predictions over the range of values which are experimentally accessible.

A study is made of the conditions under which multigroup diffusion theory reasonably approximates multigroup transport theory for the fast neutron penetration problem. Defining the probability that a neutron born in a given energy group should suffer its next collision in that group, it is demonstrated that decreasing the size of the energy group decreases the relative accuracy of a diffusion theory calculation. Examples are given for light water and natural iron.

Experimental measurements of sheath thickness and sheath growth in an unseeded atmospheric-pressure flame plasma of ionization density approximately=10¹⁵ m^-3 and electron temperature approximately=1800K have been compared with analytical solutions for these parameters. The analysis is an extension of that of Clements and Smy (1969) which allows for finite probe size and plasma flow velocity. The results, which are applicable for large sheath formation on the upstream side of a cylindrical probe, are in good agreement with theoretical predictions.

Describes measurements of the temperature distribution and the composition of the plasma-MIG arc system using optical spectroscopy. It is found that the central part of the arc contains the elements Fe, Mn, Cu, Ca and Ar at a temperature of about 7000K. The outer regions of the arc show only argon spectral lines and have much higher temperature value ( approximately=13000K). From these measurements a number of conclusions can be drawn about the physical properties of the plasma-MIG welding arc. The most important conclusion is that only a very small part of the wire current flows through the bright inner arc.

An experimental and theoretical study of the behaviour of a vacuum arc on Cu and Mo electrodes in an axial magnetic field at currents between 150 A and 16 kA has been carried out. The maximum magnetic field strength was 0.25 T. It was found that the influence of the magnetic field on both the high- and low-current arc was determined by the ratio between the cyclotron frequency and collision frequency of the electrons. The relatively small difference in behaviour of the Mo and Cu arcs was explained by the presence of neutrals in the Cu arc, while in the Mo arc these particles were practically absent. Some conclusions have been made about the electron temperature and the energy losses of the individual arc columns observed at high magnetic field strengths.

Formative time lags have been measured in uniform-field gaps of up to 3 cm in air and nitrogen at pressures of 300 and 500 Torr. Recorded time lags were in the range 20 ns<or=t_f<or=1.7 mu s for overvoltages Delta V of up to 50%. For each pressure and gap spacing, a discontinuity was found in the log-log plot of t_f against Delta V for air but not for nitrogen. Consideration has been given to the possibility of a change in the breakdown mechanism as the overvoltage is increased. Observations of the form of voltage collapse at breakdown and comparison of measured and calculated time lag data suggest that the same basic breakdown mechanism, involving the production of photosecondary electrons at the cathode, exists over the entire overvoltage range of the present study. At high overvoltages, however, this mechanism is probably strongly influenced by distortion of the applied electric field due to the space charge produced in the early stages of ionization growth.

Experiments have been made on the electron emission properties of flat oxide cathodes in a 3 Torr argon discharge, using a calorimetric method. The emission properties of thermionic cathodes depend on the gas discharge parameters, particularly on the ion bombardment. Even so, reproducible measurements have been made in a restricted current range of the order of twice the zero-field emission current. The measuring method is described, and experimental results are given. From these the work-function e phi ₀, the zero-field current density j₀, the field dependence of the emitted electron current and the temperature dependence of phi ₀ were calculated. These values are compared with results of measurements on oxide cathodes operated in vacuum. Low-intensity ion bombardment appeared to activate the cathodes.

It is possible to have a mode of electron collection with a somewhat reduced fall voltage at the anode, by producing a cathode jet with sufficient velocity at the anode surface. This has recently been shown to give a transition from a spotted to a continuous track on a rotating cylindrical anode (Reeves-Saunders, 1971). In the paper, which deals with this second continuous mode, the copper anode has been divided into two cylindrical parts insulated from one another by a small butt gap and rotating together in air at atmospheric pressure. The current distribution between the two halves has been determined for various positions of the stationary tungsten rod cathode. From these data the current density distribution at the anode surface has been determined.

The limit of absolute thermodynamic stability and tensile strength of liquids (n-hexane, n-pentane and iso-pentane) have been determined using the phenomenon of electric-field-induced nucleation. An extrapolation of the curves of critical voltage applied to the liquid against temperature to the points of zero critical voltage gives the limiting value of the superheat temperature at a given pressure. The locus of these limiting values in the p-t plane, terminating at the critical point, gives the boundary of absolute thermodynamic stability (spinodal) of the liquid. The electrical forces at the interface corresponding to various critical voltages have been proposed to give the limiting negative pressures (tensile strengths). The energy contributed by the electric field over the volume of a vapour nucleus of critical size is in agreement with the value of the thermal energy required for the formation of the nucleus of the same size de novo in a liquid.

The reading of an electromagnetic flowmeter usually depends on the velocity profile as well as on the total flow rate. A numerical method is given for designing the magnetic field so as to reduce this effect. Fluid velocity is assumed to be everywhere parallel to the axis of the flow tube but not necessarily axisymmetric. The results show that velocity profile effects can be greatly reduced. Details are given of an electromagnet designed to produce an improved magnetic field. In situations where flow is non-axisymmetric (e.g. near junctions) the maximum error expected in the reading of a flowmeter with the improved field is about 10% compared with about 50% for a conventional point electrode flowmeter. However, to achieve the same sensitivity about 25 times the power is required. The method of analysis is likely to lead to improvements in the design of both industrial and medical flowmeters.