On Fri, 2 Jan 2009 16:31:08 -0800 (PST), wrote:

Does NEC-2 model "end caps" at free ends, which is equivalent to
assuming wires are solid, or does it just set the current equal to
zero at the free ends, which is equivalent to assuming wires are
hollow? Is it possible that it does both, but the specific model is
determined by the choice of computational kernel (extended vs.
standard)?

This presumes you are seeking a wire that has no current anywhere in
its interior. Clearly there is no such conductor as any conduction
via electrons (or the charge carrier of holes) must circulate about an
atom's nucleus which necessarily imposes an interior current for half
the orbit. The only exemption would be unless the conductor is one
atom thick where there would be no interior. This would then become a
quantum wire which would have problems of its own (called a coulomb
blockade). As such a topic is clearly beyond the scope of discussion
here, for all practical purposes wire is considered to be a bulk, even
for NEC-n and the notion of "end caps" is an artifact of other,
external considerations.

I've tried looking through some of the NEC-2 documentation, but I
can't find a definitive answer.

NEC-n designers are not interested in anticipating the questions to a
vast multitude of speculative scenarios exploring the edges of QED.
For instance, modern economics doesn't provide a definitive answer to
this question either (economists don't really find those same issues
germane either).

* * * * * * * * *

However, this question can be put to any NEC-n modeling package, and
an answer provided quickly with some effort - if you are a skilled
modeler. Further, the same question AND its NEC-n modeled answer can
be weighed at the bench for validity.

At the risk of introducing a practical example that can be tested at
the bench (knowing full well how that can tax the practical skills of
many arm-chair theorists), let's proceed with a simple experiment.

First, we approach the situation with a radiator that is both hollow
AND solid. Yes, a contradiction on the face of it, but explanation
will serve. A coax that is terminated at both ends with female BNC
connections has those ends capped with male shorting plugs. The net
effect is that the "conductor" has two paths, one that is the exterior
shield, and one that is the interior conductor. The shield and
conductor are shorted together at both ends. With such a connection,
we necessarily have a solid.

We drive the exterior shield with a direct connection. Let's simply
make it the vertical radiator against a field of radial ground wires.
To all intents and purposes, the coax is a slightly thickish radiator.
We can physically measure the current excitation along the length of
the exterior path quite simply. When we take the numbers and compare
it to the NEC-n model for a vertical radiator of equal thickness
against a field of radial wires, we find a very close agreement in
results.

NEC-n has been validated in the field.

If we break one of the male shorting caps open to insert a current
meter (replacing the shorting cap with the meter inside the coax);
then we discover there is no practical current inside the coax along
the inner conductor. This should be no mystery as it is a classic
expectation of Freshman physical science. {I will note here for the
purists that I have explicitly stated no "practical" current. There
is no one in this group who could possibly measure the impractical
current.}

To create the NEC-n model would require constructing a cage of wires
to simulate the shield, implanting an interior conductor to simulate
the inner conductor and shorting both ends. This is not a
intellectual leap, merely an hour's worth of careful design. I will
leave it to someone ELSE who cares about the issue to report the
current distribution of that interior wire. A very similar example is
already available at:
http://home.comcast.net/~kb7qhc/ante.../Cage/cage.htm
so complaints of the lack of resources, time, effort, understanding,
and the rest are hollow.

Barring reports to the contrary (and speculation counts for naught),
NEC-n remains validated in the field. I would like to see that
failure demonstrated, but untutored Arthru wholly lacks the skill in
the matter, so I won't hold my breath for his demonstration.

Then the only step that remains is to open the interior wire (in
either/both the physical real model and/or the model) and take new
current distribution numbers. This fully qualifies as an hollow
conductor (you can even remove the interior wire entirely to fully
qualify to the question).

The question is solved through the model without need for browsing
documentation.

If this doesn't serve, then the question wasn't all that important as
an issue in the first place. If this is a serious question, then it
is a necessary test. Barring reports of enumerated results, it then
the question becomes yet another troll. I would note this same test
fully quashes Arthru's speculations of interior currents of a
radiator.

No appeals to Gauss, Newton, Maxwell, or Einstein were necessary in
the production of this posting as results are self evident to the
skilled and naming those dead white scientists are employed only by
hucksters trying to validate patents (aka ego certificates).

73's
Richard Clark, KB7QHC