Re: ScanTesla7.62 :-))

["Tesla list" <tesla@xxxxxxxxxx>]

Original poster: Vardan <vardan01@xxxxxxxxxxxxxxxxxxxxxxx>

Hi Antonio,

You are right about V3 needing to be changed with C3.  I noted a 
energy leak "somewhere", but did not know where ;-)) - An easily 
fixed bug.  I did do it for ground strikes though (probably after it 
obviously messed up ;-)).  Here is what I do for the streamer model:

=======================
        LeaderLength = 27.19734 * sqrt(0.5 * VCsec_max * VCsec_max * 
(C2 + C3));  //Secondary energy length.

        if (LeaderLength > StrikeDistance) LeaderLength = StrikeDistance;

        if (T1 != T1_start && C3_inc1 < 0.0)  //Do this if active 
streamer modeling is enabled and the model has run once.
                {

                C3 = (LeaderLength - TerminalDiameter) / 6 * 
1.0e-12;  //Streamer capacitive load.

                if (C3 < 0.0) C3 = 0.0;
=======================

The LeaderLength is calculated from the maximum recorded secondary 
voltage using a Freau style equation and a constant I just "made up" 
based on observed coils.  VCsec_max never goes down and hopefully 
it's maximum value does indeed represent the LeaderLength.  C3 is 
simply (LeaderLength - Terminal Diameter) / 6  picoFarads.  Sort of 
to match this:

http://drsstc.com/~sisg/files/scantesla/StreamerCapacitance.pdf

It is very simple really.  I am sure there are far better ones.  I 
needed to make "something" for the model at this point, and this is 
what I came up with.  It works OK for now and does take far more into 
account than a fixed 1pF/foot + 220k model.

I never really mentioned much about the above before.

Most posts these last few days have been about the ball lightning 
thing and fairly quiet otherwise.  I assume everyone was up last 
night drilling porous ceramics like I was :o))

http://drsstc.com/~sisg/files/plasma-ball/Electrode-Big-Hole.JPG

Cheers,

        Terry


At 08:26 AM 6/11/2006, you wrote:
> Tesla list wrote:
> .................
> Can you comment about how is the dynamic streamer model? I think that
> I didn't see it described here. By the way, it appears that I am not
> receiving some (maybe most of) the messages.
> Looking at the source code, it seems that C3 is being changed during
> the calculations. This can be done, but the state variable must be
> changed to q3 (charge) instead of v3, because if c3 changes, v3
> changes immediately.
>
> Antonio Carlos M. de Queiroz