RE: THOR Bang energy vs. streamer length measured

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "Denicolai, Marco" <Marco.Denicolai-at-tellabs-dot-com> 

Hi Jim,

What make problematic treat this subject is that the so-called
"streamer" formation is such a multi-timescale phenomena. At every scale
level the mechanisms are possibly different. I can see at least:

1. The single semiwave level. That is, for my TC resonant freq. (about
62 kHz) the 8 us time the voltage rises (positive or negative). This
happens many times within a single bang. We have corona, then the
streamer-leader effect. Event at this level we know we have several
filaments generating from a single semiwave (!).

2. The single bang level. Includes the effect of a bunch and
positive/negative alternation of level 1. How they do affect each other?
I don't know. The number of oscillations within a single bang depends on
the TC tuning and on the RSG quench time, of course.

3. The bang sequence level. This is the level I'm working at. A series
of successive bangs elongates the leader. They are spaced according to
the BPS rate. Note that we suppose this is an isolated sequence. The TC
was just turned on.

4. The discharge sequence level. Now the TC is permanently on. Bangs
come continuously at BPS rate. Streamers swing all directions. Now we
can speculate how well a streamer will stay "on the same track" of a
previous streamer. We can now have more or less "fat" streamers. I would
actually call them breakdowns or "final jumps".

Now, certainly at some of the above levels channel warmup will play its
role. But there are many alternatives for already taken "ways" to become
a leader. In my understanding the static charge left will direct the
final choice. In any case the elongating leader end, where the filaments
are swinging, is in virgin territory and will be directed by the gap
electric field and its deformation due, again, to static charge
deposited in it.

Do you agree?

 > I was thinking that as a new spark starts to develop, it can
 > more rapidly move through the previous channel, depositing
 > the new charge as it goes.  Clearly, in a TC, with HV AC on
 > th
ck and forth through
 > the spark channel during the "bang", but that's a fairly fast
 > (i.e. speed of light) propagation, as opposed to leader
 > growth, which is a 0.01c kind of thing.
 >
 > I'm thinking of a faster time scale version of the dielectric
 > treeing phenomenon, for instance.

-