Re: NEC-LIST: Memory constraints in NEC2

> From root_at_FC20.flashcomm.com Tue Feb 11 08:34:58 1997
> Date: Tue, 11 Feb 1997 08:27:08 -0500
> From: Phat Man <root_at_FC20.flashcomm.com>
> Subject: Re: NEC-LIST: Memory constraints in NEC2
> To: jkeller_at_gauss.engga.uwo.ca
> Mime-Version: 1.0
>
>
> I do not know much about the 6000. When I got my copy of NEC4 from LLNL
> it had the source code with it ... FORTRAN. I started using it on a SUN
> station but I found I could get better results and cheaper results from a
> standard PC. I have more expirance working with NEC4 than compiling it.
> If you have a fortran compiler I bet you can make it work.
>
> Or if you have a c compiler and you can use the f2c program that comes
> with linux I bet you could get it to work. If you can get a copy of NEC4.
>
> NEC4 input and output is almost identical to NEC2 and NEC3. There are a
> lot of other improvements in the code.
>
> What are you trying to model?
>
> Adam
>
>

Adam,

Actually, I am trying to model an RF birdcage resonator (used for Magnetic Resonance Imaging). The model must be as accurate as possible in order to determine the optimal lumped capacitive loads that the resonator requires for efficient operation. Unfortunately, the model requires a great number of segments (for a cylindrical shield), and not only am I experiencing significant time delays between executions (2.5 minutes, and that is the fastest computer we have - rs6000), but now I am not able to run NEC at all due to limitations on the number of allowable segments (not to mention lumped loads). Furthermore, NEC2 seems to be swapping memory on the hard-drive.

However, most of these problems are eliminated when NEC2 is used on my Pentium computer. It is just as fast and can handle the excess segments, however, there is nothing like the power of 50 RS6000 PowerPc computers at your fingertips! You know?

Thanks,

Jeff
Received on Wed Feb 12 1997 - 09:08:24 EST