NEC-LIST:NEC 4 Glitch - UPDATE

>Date: Wed, 12 Dec 2001 14:10:25
>To: nec-list_at_gweep.ca
>From: Robert Flanagan <rflanagan_at_anc.com>
>Subject: NEC 4 Glitch
>
>Code hangs on a model when NGF used.
>Model of a waveguide 24 x 48 x 164 inches @200MHz using multiple surface
patches pointed inward. Surfaces are divided into 6, 11, 41 multiple patches
on the respective dimensions, 1526 patches total. Using NGF file intending
to added feeds and other things inside later.
>The code seems to hang with "writing N.G.F. file" displayed. A (partial?)
144kb file is written but creates an end of file error if I attempt to use
it. An alternate version that does the entire model in one file (w/o NGF
file) works just fine. Also, the NGF version works fine when partitioning is
reduced to 2,3,11 patches per dimension ... but that is way coarser than I
think I need.
>Platform: P4-1GHz, 512Mb, W2000-pro (NT), parameter file set to allow 5000
segments in ram.
>Has anyone else seen this problem? Can anyone reproduce it? Am I exceeding
some undocumented limit on NGF file size, number of patches, ?
>
>Bob Flanagan
>rflanagan_at_anc.com
>

Update:
These symptoms were present with both Nittany GNEC and the LLNL NEC4D3600
when run on machines using Win2000-Pro. They appear not to be present when
run on a Win98 machine.
Many thanks to Jerry Burke for support through the ordeal of this problem.
Bob

-- 
The NEC-List mailing list <nec-list_at_gweep.ca>
http://www.gweep.ca/mailman/listinfo.cgi/nec-list