[TowerTalk] Fwd: Fwd: The Need for Grounding

[jimlux]

On 1/14/16 11:13 AM, Hans Hammarquist via TowerTalk wrote:
>
>
> No, the current does not follow "the shortest path". Yes, my mother
> always said that but Ohms' law predict more accurately how the
> current divide. More current will go through a lower resistant/ high
> conductive path but there will be current going through any
> conductive path, high or low.
>

One needs to really consider the inductance and capacitance as well. 
It's sort of a transmission line problem.


>
> Keep in mind that due to ionization conductivity can change during a
> lightning event. The current will even follow paths generated
> directly through air.

Sure enough.  There's a fascinating book called "Spark Discharge" by 
Bazelyan and Raizer that talks all about this.  There's a nice photo of 
a >100 meter long spark.


I guess you might also experience where
> grounding rods get insulated from ground due to the glazing effect
> due to the very high current density and following high temperature.
> I don't know if this ever happens but like to hear if anybody knows.
>

I don't think the rod will become insulated during the stroke.. molten 
glass is a conductor (it's how they keep it molten in glass factories). 
  If you've got enough current density to get hot enough to fuse soil, 
you probably have a fully ionized arc.

Arcs have a well known negative resistance characteristic: the more
current, the lower the resistance.   (roughly proportional to sqrt(I)
for free burning arcs in a gas).

AS K1TTT pointed out, conductivity generally increases as the current
increases.  Lots of measurements, lots of theory behind this.

Modeling it is non trivial.  There are people who have used NEC to try 
and model the RF properties of the grounding system, but I think by now, 
everyone uses purpose designed FEM codes. That way you can incorporate 
all the nonlinear effects.

Things like flashover are still more probabilistic.


>
> All the above events will lead to the unpredictable path of lightning
> damages.


yes indeed.