Owen Duffy wrote:
"John Gilmer" wrote in
net:

Do you expect it to vaporize and set your roof on fire?

Well, engineering of lightning protection is about design of a protection
system that will, amongst other things, survive most events so as to
continue to provide protection, and to minimise incidental damage.

So, yes, down conductors adequately sized to manage the risk of the
conductor "vapourising" is part of the scope, and physical design to
minimise the risk of side flash causing damage is also part of the scope.

It is interesting, no confusing, that you have two guides that give such
different guidance. In Australia, we too have a standard for house wiring,
and another standard for lightning protection, but they are not in conflict
and our standard for lightning protection is well aligned with NFPA 780 on
the downconductor size issue.

Owen




The thing is, AWG 6 wire won't vaporize or even melt or even get warm to
the touch. There's not enough "action" (I^2 T) in a lightning stroke to
do it. Remember that the current is high, but only lasts a matter of a
50-100 microseconds.

Say you are using AWG 10 wire which has a resistance of 1 milliohm per
foot. a 50 kA strike will dissipate 50E3^2*1E-3 = 2.5 MegaWatts.. which
is big.. but for 50 microseconds, that's only 150 joules. That same
foot of wire weighs about 1/2 an ounce (I'm sorry for the customary
units, but they are what I remember off the top of my head AWG 10 is
1/10th inch in diameter, 1 ohm/kft, and 32 ft/lb).. or about 14 grams.

Specific heat of copper is 0.38, so we have deltaT = 150/14 * 0.38
let's call it about 4 degrees C.

I should note that this is a bit optimistic.. the AC resistance for a 50
microsecond pulse will be higher than for DC because of skin effect
(skin depth at 1 MHz is 65 microns, 2.5E-3 inches, and it goes as the
square root, so even at 100kHz, it's still not much more).. so the
dissipation will be higher.

But, you've got a long ways from 30C to 1000C (melting point of copper)
and even farther to "vaporization"...

(as a practical matter, you need kiloJoules to explode a 1 meter AWG 30
copper wire.. hundreds of joules just "melts" it. )

(note also that while the peak current might be 50kA or 100kA, the
average current is substantially less..)

Mechanical stresses from magnetic fields are a bigger concern, as well
as "sideflash".