NEC-LIST: wire meshing in NEC

I'm trying to use NEC-2 to model solid metal bow-tie monopole antennas
using a square wire mesh as part of my thesis and wanted to see if
anybody here could shed some light on some of the issues that I have
been running into.  I am interested in simulating accurately the
input impedance of bow-tie antennas of varying heights and flare
angles over a wide bandwidth, 0 - 6 GHz for instance.

I have seen in past postings that generally a wire density of
lambda/20 is necessary to accurately model a solid metal surface. Has
this been the case for others or is a higher density generally
required? I have also heard that NEC-2 might have problems with this
dense of a mesh and that NEC-4 might give better results- is this
true?

One last problem - for a given size of antenna, I have found a mesh
density that will give me the accuracy in input impedance that I need
over a given bandwidth of 0 to 4 GHz. This was for a short bow-tie
and a lambda / 20 density for the highest frequency was required. Now
I want to use the same mesh to simulate a bow-tie with the same flare
angle but twice as high. I would expect that with the same density,
the impedance would be accurate over the same bandwidth as before. In
running it, however, I find that the impedance results are only
accurate out to say 3 GHz. Has anybody else experienced this
phenomena? How do you explain it physically?

If anybody has any insight at all, it would be very much appreciated.
Thanks in advance.

Sincerely,
 

Aaron Kerkhoff
Applied Research Labs
University of Texas - Austin
Received on Thu Mar 15 2001 - 05:12:30 EST