According to HorneTD :

Several posters have said here that impedance or rather the reactive
component of the total impedance could not possibly make a difference at
60 hertz.

I was involved in that discussion, and as you may recall, I agreed
_completely_ with you over considering cable armor alone as
"adequate grounding".

The only point where I disagreed with you is the suggestion that
the "coil reactive impedance" is likely to play a significant factor
between tripping the OCPD and not tripping the OCPD. Corrosion is
going to be the main factor in virtually every real world case if
you're dealing with a mere 60Hz.

At 400Mhz, on the other hand, a couple turns of wire is essentially
an open circuit.

I don't know how to explain away the testing that was done by
UL that showed that a wireman's failure to cut a slot between two
knockouts of a metal box that each pass a conductor of an AC circuit
into that box will result in inductive heating and that this inductive

Remember that inductive heating isn't the same thing at all as
reactive impedance. So one doesn't imply the other and v-v.

heating can, over time, cause the pyrolysis of the supporting structural
element to which such a box is attached resulting in the eventual open
flaming ignition of the structure. That process takes months rather
than hours so it is not well understood in the industry because the
effect is not readily discernible over a short time.

I'd love to see this test report. What current level/voltage
was that at?

In order for this to occur at anything remotely resembling the
power/frequency levels that occur in people's homes, there'd have to be
some pretty subtle metalurgical changes going on in the metal box.

There ain't no way that 15A at 60hz going through separate holes in,
say, 16ga steel is going to induce enough current flow to generate
appreciable heat in said steel. No matter how long you waited.
Unless you had the whole assembly in a dewar flask and didn't mind
waiting the months for micro-degree daily increments to get to dangerous
levels.

Indeed, inductive heating _still_ occurs _even if_ the whole circuit
goes through one hole. Thus, the same thing would happen (albeit take
longer) even if the whole cable went through one hole.

Unbalanced circuits (eg: hot plus neutral) radiate, period. Thus
they induce current in surrounding plate penetration, one or two hole.
Thus they generate heat.

But, the frequency is VERY important.

The amount of inductive heating at 60hz on simple plate penetration
is small. Not non-existant, but _extremely_ small. So small to be
dwarfed by conduction/convection in any real-world situation we're
likely to see. Now, if we're talking million amp plus flows, that's
different.

I do not lay any claim to fully understanding the physics involved and
according to some posters that makes me a "codebot."

I'm not one of those posters. I consider you a highly skilled and
knowledgeable professional, with very valuable real experience. Just
don't know RF very well - s'alright, you know a _lot_ more about code
than I do.
--
Chris Lewis, Una confibula non set est
It's not just anyone who gets a Starship Cruiser class named after them.