SELF

40

and S.B. Karavashkin

The second of possible processes of metal bridge formation is caused by possible crossing the plasma jets ejected from cathode during the pulsing discharge. The cathode jets were studied by Taev [35], Mick and Krags [10], Bron [36] and other researchers. In particular, Taev wrote: "Vapour streams consisting, as spectroscopically established, of mercury atoms can arise both on anode and cathode or on both electrodes… When colliding, streams disperse" [35, p. 333- 334]. In Fig. 22 we show the images of such crossing plasma streams.

Fig. 22. Typical shape of plasma flares crossing in the gap [35]

In Fig. 23 Mick and Krags [10] show a sequential series of images of pulse breakdown in hydrogen medium [10, Fig. 272, image 28].

a)                                                     b)                                                  c)

d)                                                   e)

Fig. 23. "The image of sparking discharges in hydrogen obtained at different instants after discharge beginning with the help of Kerr cell. Images have been obtained in the following time intervals in microseconds: a - 0; b - 4; c - 5,5; d - 8,5; e - 11" [10]

The authors classified these images as an evidence of streamer breakdown of gas medium in gap. None the less, some features tell that the plasma filament produced in the centre of images is not the streamer by several reasons. First, streamer precedes the breakdown, develops in non-ionised gap and provokes production of new avalanches [37, p. 29]. But in Fig. 23a we see a partially ionised gleamy gap where germinates a central column which can be and rather is a plasma jet. Second, when closing a central column between the electrodes in Fig. 23b, we see around this column a decaying plasma forming a dark space, which also is not typical for streamer breakdown after it strapped the electrode gap. And in Fig. 23d we see the fall of gleam, which is even more irrelevant to streamer and basically inconsistent with gas discharge, which has to develop in the channel of streamer consisting of plasma. And the final stage of discharge shown in Fig. 23d, within conventional understanding, cannot develop after deionisation of plasma gap, it undoubtedly develops after the burst of bridge formed by plasma jets. Anyway, the sequence of visual representations shown in Fig. 23 fully corresponds to formation and burst of plasma jet bridges. Such bridges usually can form at the final stage of showering arcs with enough large distance between electrodes, when the energy accumulated in the inductance is still sufficient and liquid holes on the surface are quite extensive to form plasma jets.

Thus, as we can see from this analysis, there exist two real possibilities to form metallic bridges strapping the electrode gap. The first of them is caused by electric pressure affection, it is mostly typical for practically whole time of showering arcs growth, so it determines the main mass transfer and contact erosion. The second process can arise in high-current showering arcs when the process in them is protracted to the region of long inter-electrode gaps. Its affection onto erosion will be prevalent in this region of high-current discharges.

From the considered mechanism of bridge formation, the main criterion for contact protection from showering arcs follows. It means smoothed overvoltage in the oscillation sections in order to prevent bridge formation, which as a consequence will lessen the erosion; below we will speak of it in more details.