Table of contents

The laser-triggered switching (LTS) of high-voltage spark gaps is considered. The basic theory is presented which predicts dependencies of the delay to breakdown and switching jitter on such variables as fill gas mixture and pressure, gap spacing, polarity, and geometry. It is shown that electrical arcs of several metres length can be directed by laser action. A complete set of experiments is reported which adequately support the proposed theory. The performance of LTS is considered and results are reported on multiple gap triggering, multiple channel triggering, triggering of voltages in excess of 3 mV, repetitive switching at rates up to 50 pps with subnanosecond jitter, as well as various geometries, pulse forming demonstrations, and output voltage selection on a Marx generator.

The so-called plasma polarisation shift effect is reviewed both theoretically and experimentally. This effect involves the shift of a line under the influence of perturbers in the vicinity of a radiator. As this is a typical n-body problem, only approximate treatments are available. These are reviewed critically with particular emphasis on the case of strong interactions, typical of, e.g., resonance lines. Experiments devoted to the observation of this effect are described in detail and the results compared with the different theoretical estimates. It is established that plasma polarisation leads to blue shifts of resonance lines but the experimental results appear to be partly contradictory for the lower series members and increasingly more consistent when considering the higher members.

The energy transfer from positive ions and electrons around a negative probe immersed in a plasma is analysed semi-theoretically. Probe theory is used to establish the spatial profiles of electrical field and charged particle number densities. Energy transfers from ions and electrons accelerated to, and away from, the probe are analysed. By these means the gas heating of neutral molecules is computed, as well as the temperature profiles around the probe. At a sufficiently high potential the surrounding gas is heated appreciably, and the energy transfer to the probe surface is enhanced. The principal contribution to the latter is from the heated surrounding gas, rather than from the energetic ions striking the surface.

Measurements of transport parameters of pulsed electron swarms moving through a low-pressure gas by observation of the photon flux resulting from electron-molecule collisions have been recently reported (Blevin et al., 1976). One of the possible sources of error in this kind of experiment is the variation of mean electron energy through the swarm. This effect is considered here along with the resulting variation of ionisation and excitation frequency through the swarm. The validity of the experimental method is considered in the light of the above factors.

A non-stationary cathode spot model is proposed which is based on an interpretation of experimental data recently obtained. The significance of ion mass generation by Joule heating and the separation of energy flows associated with electron and ion currents in the cathode spot are the main differences with other known cathode spot models. The ion energy balance at the metal-vacuum transition is evaluated quantitatively and applied to a calculation of the ion fraction in the cathode current. For a broad range of metals and independent of the electron emission mechanism, minimum and maximum ion current fractions are found of around 10% and 20% respectively.

A consistent system of gas-dynamic and Maxwell equations which includes the ponderomotive force is formulated and discussed with respect to momentum and energy conservation in the interaction of radiation with a fully ionised plasma. The role of polarisation and compressibility of the plasma is pointed out.

A generalised validity condition for local thermodynamic equilibrium (LTE) relevant to a laser-produced plasma is proposed and this new validity criterion is used to calculate the LTE domain for an aluminium-plasma. The applicability of the various equilibrium models, namely LTE, collisional-radiative model (CR) and coronal equilibrium (CE) is then checked for such a plasma. It is found that for a laser pulse of 1 J, 300 ps pulse width, and 3*10¹³ W cm^-2 peak power, irradiating an aluminium target, more than 87% of the X-ray radiation is emitted for the CR domain of the plasma.

In dense arc plasmas (T approximately=12500K, N_e approximately=1-6*10¹² cm^-3) the high continuum emission, the strong broadening and self-absorption of the lines, as well as more fundamental problems of non-ideality (N_D=⁸/₃ pi N_e lambda _D³ not >>1) renders the usual diagnostics difficult. Infrared schlieren is shown to be a good alternative. Results are in good agreement for electron density with Stark broadening measurements using Griem's data. Absorption measurements also show that the plasma does not exhibit abnormal behaviour compared to the usual case (N_D>>1).

The relationship of the singular current which flows within one Larmor radius of the axis of a Z-pinch to the current due to guiding centre and diamagnetic motion in the magnetised region is found. In particular for zero local ion centre-of-mass velocity it is shown that ions flow on-axis from the anode to the cathode and there is a return flow by guiding centre drift in the off-axis magnetised region. The analogous electron flow indicates that in a steady pinch held under pressure balance the heat losses to the electrodes will be dominant and indeed must be balanced by Ohmic heating for an equilibrium to hold.

The electrical breakdown of a gas at radio frequencies in the HF band (f approximately 10 MHz) has been studied in some detail over a period of about ten cycles of the applied field immediately prior to and following a breakdown event. Discharge current and voltage are monitored as a function of time, together with the photon output from electronically excited species in the gas. The development of the discharge over the first few cycles of the applied field following the initiation of the discharge is determined by parameters such as overvoltage and the dynamic properties of the radio frequency generator. These two properties in particular control to a large extent the early breakdown phase.

Motion of the wavecrests is calculated for the case when a frequency modulation has been imposed locally on a coherent ionisation wave with saturated (large) amplitude. Different possibilities of continuation of the solution into the region of multivaluedness are discussed and compared with the experimentally observed pattern.

The processes and errors in determining the ion temperatures of an inhomogeneous plasma from Doppler profiles of radiation are discussed. When adequate procedure is taken with the Abel inversion, errors in the process are minimised.

Radiation transport within wall-stabilised nitrogen arcs has been investigated for pressures of 1 and 10 atm. The computations allow for continuum and N I and N II line transport, and some molecular band absorption. The spectral data used gives reasonable agreement with experiment for the radiative balance on the arc axis at 1 atm. The calculated radiative arc loss accounts for a significant fraction of the electric power input, and it is found that losses at wavelengths less than 1000 AA are most important over a wide range of conditions. The dominant contribution to the loss is made by line radiation transport. Experimental temperature profiles were used for the 1 atm calculations, and for 10 atm the profiles were predicted using an approximate value for the local radiative balance determined from isothermal-arc calculations.

The electron swarm development in nitrogen is studied for E/N values from 56.6 to 1131 Td by a Boltzmann equation method in which the effect of ionisation is considered properly. Alteration of the electron collision cross-sections from the literature is limited to a minimum, although minor amendments are found necessary for the ionisation coefficient to agree well with previous measurements. The cross-sections for vibrational excitation are considered separately for its quantum number v=1-10. The electron swarm parameters are calculated for the time-of-flight, pulsed Townsend and steady-state Townsend experiments. The results show that the deduced electron drift velocity, diffusion coefficient and mean energy are generally in good agreement with previous experimental and theoretical values.