Table of contents

Two effects leading to significant modifications of the theory of colloidal plasmas are discussed. The first concerns the fact that depending on the sign of the electric potential on a droplet in a colloidal plasma, the electrons or the ions impact on the droplet with a much higher effective temperature than their background plasma temperature. The analytical expression for this temperature is derived. It is shown that the electron-droplet collision rate in a wet alkali-metal vapour plasma may be substantially modified by the effect of the attractive potential between a droplet and an electron because of the image of the electron in the droplet. This effect is found to be important when the droplet sizes are in the submicron range.

Jump conditions are derived separately for electrons and for ions. The resulting equations are generalizations of previously derived jump conditions and are shown to reduce to these in appropriate limits. However, even in these limits, these new results are more general, since the derivation eliminates a number of restrictive assumptions previously invoked.

It is shown that the generalized form of the sheath criterion corresponds to 'sonic flow' at the plasma boundary.

A single-fluid one-dimensional radial compression theory has been developed to predict the densities and temperatures in the continuous flow pinch. An interpretation of the discrepancies between the theoretical and experimental results has been given. The efficiency of the focused flow device has been discussed for both normal and large outer electrodes.

If the expansion velocity of a laser-produced plasma is around the ion sound speed, then in the laboratory frame, it is possible to have standing ion sound waves together with pump and backscattered waves of equal frequency. Partial reflection of the pump wave from the high-density regions of the plasma can then lead, via stimulated Brillouin scattering to enhanced reflection.

An exact numerical method has been used to study a probe stagnation point in a continuum flowing plasma. The solutions for the 'low-density continuum' plasma, i.e. with mean free path<<probe radius<<Debye length are calculated for Reynolds numbers Re from 0 to 400, charged diffusion Reynolds numbers Ra_c from 0 to 10, and scaled probe potentials phi _p/ epsilon from 0 to +or-32, where phi _p=eV_p/kT_e, epsilon =T/T_e. The numerical results show that (i) the usual semi-log method for temperature determination leads to large errors; (ii) the effect of flow leads to large errors in the space potential determination; (iii) for a sufficiently large accelerating potential, the currents always show a linear dependence; (iv) approximate current expressions are suggested at smaller Reynolds numbers (Re<or=0.1) and diffusion Reynolds numbers (Ra_c<or=1).

From measurements of the relative abundances of O^- and CO₃^- ions produced in carbon dioxide by electron attachment and subsequent ion-molecule reactions, and from current-growth data, values of the primary ionization coefficient, attachment coefficient and rate coefficient for the reaction O^-+2CO₂ to CO₃^-+CO₂ have been obtained for 60*10^-17<E/N<152*10^-17 V cm² at gas number densities N between 0.2 and 5.0*10¹⁶ cm^-3.

A new method of measuring discharges at the surface of a dielectric is presented. It is based on observations of the flash accompanying the discharges with a linear photomultiplier. The measurements were made for C₄H₄O₆KNa.4H₂O, KH₂PO₄, NH₄H₂PO₄, glass (G) and organic glass (PG). The measured flash-energy spectra reveal two parts which vary strongly with the dielectric used, the applied electric field and the time. The flash energies, geometrical picture and flash-energy spectra are discussed.

The composition of cathode mass loss was analysed for cadmium, copper and molybdenum vacuum arcs. It showed that two dominant flows are present, one consisting of ions, the other of molten droplets which have sizes in the order of microns to tens of microns. The droplet flow is mainly oriented along the cathode plane and is strongly dependent on the fusion temperature of the cathode metal and the charge transfer by the arc. The cathode mass loss in vapour form is considered to be small.

It was arranged that any particles would traverse a system of fringes formed by the interference of two laser beams and the scattered light would be detected by a photomultiplier. Single particles above 5 mu m could be detected. No particles were observed in a variety of gaps between 0.5 and 2 mm stressed at up to 200 MV m^-1 whether or not subsequent breakdown occurred. If particles in this range do exist the field required to detach them is apparently greater than that which the gap will withstand before breakdown occurs by other mechanisms.

Measurements are reported of the ion density and voltage drop of a short pulsed arc. Results obtained while the arc is running show broad agreement with calculations based upon a fully ionized plasma which loses heat to the electrodes by electron conduction and convection. Results obtained after the arc has ceased abruptly are explicable on the grounds of ion diffusion and cooling due to heat conduction by neutrals. Results obtained with a current decreasing steadily to zero (to simulate an AC 'current zero') indicate that non-equilibrium effects are significant.