Re: Relativistic Runaway Breakdown Model for Lightning

["Tesla list" <tesla@xxxxxxxxxx>]

Original poster: boris petkovic <petkovic7@xxxxxxxxx>

Firstly,my apology for the late response to this
interesting post (was on vacation).

--- Tesla list <tesla@xxxxxxxxxx> wrote:

> Original poster: Greg Leyh <lod@xxxxxxxxxxx>
>
> Hi All,
>
> An interesting paper by A.V. Gurevich and K.P. Zybin
> appears in the May
> issue of Phys Today, titled "Runaway Breakdown and
> the Mysteries of
> Lightning."  If you have access to a copy of Physics
> Today I strongly
> recommend reading this article.
>
> One of the mysteries they address is how natural
> lightning can originate
> and propagate with an average electric field of only
> [200kV/m] -- an order
> of magnitude below what is needed to breakdown air
> at STP [2MV/m].
-----
More accurately stated the problem is a lightning
origin in clouds ,not a propagation.It is known from
lab experience that 1MV+ leaders can propagate in
quite a low external fields once the leader is formed.

As concerns problem of a lightning generation in
clouds ,it should be born in mind that cloud volume is
filled by hydrometeorites (common name for water
droplets,ice crystals etc).
For start,consider a metalic ball when placed in
external uniform field E.
There is a closed form solution (it can be obtained
using various techniques of solving symmetrical
electrostatic task) in such case.Maximum field at the
ball surface rises to Em=3E.
Almost to the same factor (Ec~3E) enhanced enhanced
will be the field at spherical water droplet
dielectric surface owing to a very high relative
dielectric constant of water (~80).
This gives the reason to believe that corona on
hydrometeorites keeps the field well bellow ionisation
threshold Ei in the cloud.At ,say, 3km altitude where
Ei~2MV/m the max field allowed would  be Ec=600...700
kV/m.
Indeed,max cloud fields ever measured by direct rocket
 probing of a charged cloud regions are close to that
range {{1}}.
Other field in situ msms give generally lower values
:100...300 kV/m  {{2}},{{3}}.
-----
  They
> provide a compelling argument that a relativistic
> electron chain reaction,
> initially triggered by a cosmic ray-induced shower,
> is the principal
> mechanism for generating the ionization necessary to
> trigger the main
> lightning strike.
-----
Than they must have explained phenomena of K processes
in the cloud?
It is hard to judge about mechanisms triggering
preliminary breakdown in clouds.Especially,if one
considers some observations that the location of the
charge change due to lightning does not necessarily
coincide with the location of the charge region
responsible for the lightning {{4}}.If nothing
else,such observations may go in favour of runaway
electron mechanism theory.
But,one thing is certain:Following the ionisation
onset,positive streamers in standard atmosphere can
propagate in average field 450-500 kV/m,giving rise to
possible leader.For *negative* streamers the average
field is twice as large -900-1000 kV/m.{{5}}
At 3 km altitude and Ei~2MV/m,the average field may be
as low as Ec=600 kV/m.
If this is so,a field in cloud can be enhanced above
Ec only for a very short time,say,as a result of eddy
concentration of charged hydrometeorites.The
enhancement will be reduced to zero by corona and
discharge may stop or under favorable condition
successive corona bursts may form a lightning leader.
Needless to say,the researcher has no practical chance
to guess where field may be locally enhanced to
introduce his probes there.
The favorable conditions are unknown or still poorly
understood.
Even some suggestions dealing with piezoelectric
effects orignating from collisions of ice particles in
thunderstorms can't be completely excluded.Surely air
streams can be very intensive near thunderstorm cells
and "mechanical" effects are good candidates for
lightning "lighters" as well.

One way or another,investigation of what has caused an
explosion or fire in industrial premises containing
charged particles / droplets provide evidence for a
spark discharge arising in volume with dispersed
charge.IIRC,there have been reports of studies with
gas jet generators ejecting small charged clouds.
Sometimes,extended bright structures up to 5" in  size
were observed along a charged spray boundary.
-----

> Assuming classical electron behavior, there is a
> critical maximum size for
> a disruptive coil, beyond which it's ability to
> produce longer sparks falls
> to zero.  This is due to the Fo larger coils
> becoming so low that the
> minimum possible time between firings starts to
> exceed the ion
> recombination times in air, allowing the arc channel
> to fully extinguish
> between firings.
-----
Sounds OK.
-----


>
> However if one allows for relativistic runaway
> breakdown effects, there is
> a critical *minimum* coil size that reaches the
> state where only 200kV/m is
> needed to generate the arc channel.
-----
I don't understand this statement.What do you mean by
term " generate" the arc?
-----


> Milikh suggests that the electric field needs to be
> supported over a length
> several times this, about 150m.
-----
This is nothing but a making pure conjectures.Nobody
knows prior to experimental checkup what can happen.
-----
  To achieve this
> scenario, a twin coil
> system is required with a physical scale of about
> 240ft high, 50ft diameter
> each tower, and a tower to tower spacing of  about
> 450ft.  Each tower would
> need to generate a peak voltage of about 15MV.
-----
Hmm..to extrapolate a giant TC operation parameters
just on the basis of that scenario is a gamble.
TC pulses are neiher monopolar nor the arc grow in a
single shot fashion.

15 MV??
Gaps  150 m long can be closed by 5MV monopolar pulse
even if the waveform ain't completely the optimum one.
Links to photographs of big russian VPG doing the
bussiness has been posted to TCML  several times.
According to {{5}} the depedence is roghly
Vmin::L^0.4.

150 m -> 5 MV
15 MV -> ~3000 m

3 km is already a lightning discarge gap.
Speaking of the lightning,some TCML listees ,every
once in awhile ,mention lightning parameters in
comparation with TC  parameters.
These discharges look and sound similar but here every
comparation stops.
Not just in waveforms,but there is a difference in
physical principles involved.
Natural lightning leader is an Electrodeless discharge
in a true meaning of the word.Cloud has almost no
conductivity,and its charge is attached to
hydrometeoric particles   sprayed in space.
The mechanism of leader propagation is principally the
same as that of bipolar leader excited in  lab E-field
from ends of wire floating in the external field.
Analoguosly,storm cloud charge serves just as source
of external field for binary (bipolar) creation and
energy supply for covering its propagation
expenditures.Corresponding considerations where leader
channel charge is treated as polarization charge along
conductor immersed in a cloud dipole field can be
found in {{6}}.
During the lightning development,binary leader system
undergoes significant time evolution.
Therefore,natural lightning by related physical
principle differs from marx-bank spark too.
The missunderstanding of the principal difference
between lab spark being galvanically connected to HV
electrode and natural lightning can be source of
various errors:From gross overestimation of axial
E-field in the lightning channel,to attributing too
low potentials to charged thunderstorm regions.
The latter are ,judging by atmospheric field
msms,huge.
Estimation are in 100..500 MV.Not just uncertainity up
to an order of magnitude in estimating that
parameter,but a mere problem of definition how it
should be related to the lightning spark is
troublesome.
Fortunatelly,in light of everything said,this
potential is of little relevance.
It is my understanding that the potential of fully
developed leader should be termed *lightning
potential*.
More reasons to justify such standpoint:
-lightning protection specialists are primarily
interested in return stroke stage of a lightning flash
and its hazardous effects to protected objects
-magnitude of peak current and considerations
concerning so called stricking radius and electrode
geometry of the objects at ground are all standing in
relationship with leader potential.

Registrations and msms of dynamic E-field variation at
ground surface as leader approaches gave roughly
typical negative leader potentials:
30-50 MV for stepped leader of first lightning
component (representative is ~30 kA)
10-15 MV for dart leader of subsequent component
(representative ~ 12 kA).

I think  the interval <10...50 MV> refered for
lightning has been already mentioned on this list,but
it should be clarified what is meant by that.
At least ,it doesn't harm to repeat it here once more
since the season of summer thunderstorms on North
hemisphere is about to begin ,and some people on the
list  may start wondering.

Regards,
Boris

References:
{{1}}
Winn et al;1974/J.Geophysics Res;79
{{2}}
Gunn;1965/J.Atmos.Sci;22
{{3}}
Evans;1969/J.Geophysics Res,88
{{4}}
Vonnegut;1983/J.Geophys.Res,88
{{5}}
Bazelyan,Raizer;1997/"Spark Discharge",Boca Raton
Press
{{6}}
Kasemir;1960/J.Geophys.Res,65