RE: NEC-LIST: Low angle radiation?

Hello Achim,

Since my name came up in Jack Belrose's message, and I have some
differences with your reply to him, I feel compelled to add my own
$0.02.

>
> I do not quite agree with your explanation of radiation. The
> acceleration of charges happens along the whole antenna. The drift
> velocity of charges in metals is very low (some mm/s or maybe cm/s).

I don't believe that the drift speed of charges (electrons) in a metal
has anything directly to do with radiation unless the ohmic resistance
becomes large enough that for purposes of a boundary-value solution
the metal can no longer be considering perfectly conducting. The
propagating fields in the region outside the conductor move at very
nearly light speed. Obviously, physical charges in a real metal can't
move that fast. Furthermore, a charge injected at the feedpoint of an
antenna does not accompany the propagating E-field normal to the
antenna surface down the wire. Instead, the normal field is "passed
off" from charge to charge in much the same fashion as a row of
dominoes "propagates" a wave when one falls into the next. The wave
speed in the domino model can propagate much faster than the actual
falling speed of the individual dominoes.

>
> So if the antenna is excited with a source of e.g. 1 MHz, each charge
> oscillates around its position of rest with 1 MHz and very low
> amplitude. It never sees the end of the dipole.

While I agree with your statement that each charge never sees the end
of the dipole (except) those right at the end, I don't see each charge
oscillating around some rest position. Consider the standing Q/I wave
on a dipole. The standing wave is comprised of two
counter-propagating waves moving at essentially contsant speed; there
is no oscillating charge associated with the standing wave as a
result. Thus, there is no significant amount of radiation from
oscillating charges per se. Whatever radiation comes from a dipole
between its feedpoint and ends appears instead to come from another
cause, which I believe to be a partial reflection of the propagating
waves due to the fact that the wire has a wave impedance that varies
spatially.

> Since the oscillation
> is harmonic, the same holds for the acceleration (second derivative
> w.r.t time). To inquire as to the causes of radiation from the
> viewpoint of charges is not easy, since the cause for the oscillation
> of the charges is the guided wave along the antenna and all the
> charges themselves cause radiation.

I do emphatically agree that inquiring as to the causes of radiation
from the viewpoint of charges is not easy; we solve our boundary-value
problems in terms of equivalent Q/I and never have to deal explicitly
with physical charge in a real material. I've recently been exploring
an approach I call FARS (Farfield Analysis of Radiation Sources) that
seems to offer a way to determine where and how much power is radiated
from an object like a wire antenna or scatterer on a per-unit-length
basis. Some results for FARS were published in the IEEE AP-S Magazine
in the June and October 1999 issues in "PCs for AP and Other EM
Reflections." FARS indicates that for a wire antenna several
wavelengths long the largest sources of radiation are the ends and the
feedpoint, the amount of the latter determined by the wire length.
But FARS also shows that a smaller amount of radiation occurs all
along the antenna. A time-domain version of FARS seems to be
consistent with this picture (thanks to Jerry Burke for implementing
FARS in NEC and TWTD). Discussion of various other aspects of
radiation were included in a number of previous issues of that column.

I must admit to still not understanding the phenomenon of radiation
very well. Further, various others in the EM community would strongly
disagree with the claim above that radiation even occurs along a wire.
Rather, their view is that the radiation is confined to the feedpoint
and ends. Considering that collectively we, as a community, still
don't seem to have a good quantitative handle on this problem, we
ought to be thankful that a thorough understanding of radiation seems
not to be necessary to analyze and design antennas and scatterers.

I'd be interested in any responses from NEC-LIST readers to any of the
above.

Best wishes,

Ed Miller

--
Dr. Edmund K. Miller
3225 Calle Celestial
Santa Fe, NM 87501-9613
505-820-7371 (Voice & FAX)
e.miller_at_ieee.org