V.5 No 1

27

On physical processes in showering arcs

A special issue for practice, and first of all for the purpose of spark-extinguishing, is the issue of volt-ampere characteristic of short arcs. After Omelchenko [8], short arc distinction from the plasma long arc is that the regularity UI = const corresponds to the plasma arc, and U = const   - to the short arc which is independent of current. This statement is supported by the fact that the short arc mainly is not fed by switched current and therefore does not depend on it.

The fall of voltage at the short arc is thought to be some exceeding the sum of exit work and ionisation potential; it makes (10 - 20)  V [20, 21]. Though we can think it to be only a premise, as we have not found in literature the short arc oscillograms with the oscillograph resolution enough for measurement. In all published oscillograms, the sections at which the short arcs are supposed are too compressed, it is practically impossible to determine the fall of voltage enough accurately.

As to the value of voltage switching on the short arcs, data also is contradictive. The researchers [8, 9, 11, 21], having been referring to Germer [3, 14], state this voltage to be in limits (50 - 400) V. So Omelchenko [8] writes: "Cathode arcs arise at (100 - 400) V, and those anode - at lower voltage (50 V)". But as it follows from Germer's work [14], these voltages were determined after maximal voltages at the contacts shown in the oscillogram of unstable discharges, to which it also is worthy to pay a special attention from the view of following comparative analysis with experimental results presented in this work. Similar studies of other scientists [21] showed that the voltage bumps in unstable discharges can achieve (1,5 - 2) kV, and in Mills' oscillograms [2] (see Fig. 1) - even more than (800 - 1000) V. Such spread of values of breakdown voltage, even noting all amount of factors deforming the results in experimental studies of breakdown phenomena, and in so small gaps, must raise the question: what was the cause?

Finally, touching the point of short arc duration, we should mark, the main opinion of scientists, in that number [8, 9, 11], has been formed after the work by Germer and Boyle [14], according to which, the short arc lives 1 microsecond. But in another work by Germer [3] we can see the oscillograms with short arcs, both of 1 mcsec duration with the length of connected cable 60 m (capacitance 1200 pF) and of 0,2 mcsec duration with the length of connected cable 10 m (capacitance 200 pF). Hence, the short arc duration can be different, and it is dependent on switched current, parasitic capacitance (cable length), rate of contacts motion, material of contacts, nothing to say that the very short arc, as we saw it, is discontinuous and alternating in sign; with a poor resolution of oscillograph it was impossible to account it properly.

We carried out this brief analysis of today situation in the studied subject in order to collect a more or less integer outlook in the physics of discharge phenomena from disordered and sometimes, as we saw, even contradictive facts. The lack of integrity was caused, in particular, by lack of integer surveys. In some places of our survey we intentionally anticipated some things and emphasised some points, to make clearer the trend and causation of our investigation; its results are stated below.