NEC-LIST: Wire mesh vs. surface patches

A structure that I'm modeling in NEC-4 includes a (perfectly)
conducting closed surface enclosing a nonzero volume. Therefore
either wire mesh or surface patches can be used to represent the
surface. I've tried both and both seem to work well -- at least, they
did when I was conservative in modeling. My wire mesh and my
surface-patch tiling were both square, 0.027 wavelength on a side.

With about 2500 current unknowns (=NS+2*NP) in each case, the
wire-mesh job took about 2.5 times longer to run than the patch job.
Is this disparity expected?

For a given number of current unknowns (and less conservatism in
modeling, i.e., a coarser grid or larger patches) should I expect
better accuracy using wire mesh or using surface patches? The surface
being modeled is a stick-shaped rectangular solid, larger than a
wavelength in one dimension and smaller than a wavelength in the two
others.
Received on Fri Jul 10 1998 - 09:37:39 EDT