Jeff wrote:

Surely it is your model that is an aberration. Did it include such things as
the loss tangent of the dielectric and the effects of the resistivity of the
liquid ?

Di-electric loading is a valid concept, but it has to be done sensibly, and
modelled even more carefully, taking ALL the effects of the di-electric into
account. GIGO.

The flaw is that the "wire insulation" feature of EZNEC is valid only
for thin coatings, such as those typically found on insulated wire. I
suspect it may also be invalid also for extreme values of dielectric
constant. I don't at present know exactly at what thickness or
dielectric constant the calculation becomes invalid, so results from
models with exceptionally thick and/or high dielectric insulation should
be viewed with some skepticism.

These limitations aren't spelled out in the EZNEC manual, an oversight
on my part for which I apologize. They will be included in the next
update to the manual, which will probably be included with the next
program update (v. 4.0.27).

Roy Lewallen, W7EL