Large iges files fea

[bstork]

I am trying to run an fea on an iges file exported from mimics. It is a 3-D model of a skull and a little above 150 mb. None of the current computers I have access to can import this file into ProE or Abaqus. I was wondering what the specs should be on a computer purchased and what would be the best FEA software to run it on.

 

[GBor]

IGES file size and FEA file size are based on two different aspects, so the FEA file may be much smaller depending on the level of detail to which you will mesh the skull. As for the "best FEA", what are you trying to do? Fracture mechanics has some specialty codes that may be useful. Abaqus is certainly a good product. Event simulation (impact analysis), Algor and NENastran have good, relatively inexpensive products...depends on your budget and what you're trying to do.

Garland E. Borowski, PE

 

[johnhors]

bstork

Assuming you are talking PC's for hardware, then a top spec machine from any manufacturer will suffice. That spec being a machine with at least a 3GHz processor, 2 Gigabytes of ram, 128 Mbyte graphics card from ATI or nVidia and 200+ Gigabyte hard disc space.

Garland made specific statements on various FEA softwares, I'll just add that any software is only restricted by the hardware that you run it on, the software simply grabs enough disc and memory resources (if available) to deal with the job that you throw at it. You'll find that ProE or Abaqus should be able to import your model on a machine with enough resources. If you have a PC with the above capabilities and it still fails to import then you'll have to look at unix machines.

I'm familiar with Abaqus/CAE and I know that it's IGES import translator is particularly weak, it's much happier with SAT, Parasolid or Elysium neutral file conversions from ProE etc.

 

[gwolf]

bstork,

If the file has come from a scan of a skull I'm guessing that the IGES file is comprised of tens/hundreds of thousands of triangular facets, the number relating to the scan resolution. The file size probably exceed the interface limits and/or the analysis package limits. Also, even if you do get this huge model in, remember that the resulting FE model will be 100-1000 times bigger (but not 100-1000x the IGES file, IGES is an inefficient file format).

I strongly suggest that you put a much smaller resolution model through, say 1000-3000 facets for a skull, do a simple analysis and then review your situation in terms of analysis time and result quality. You will then have a sensible benchmark against which to make decisions.

I strongly disagree with Johnhors that you can buy your way out of trouble in this particular case with a bigger, faster computer, although you may need one anyway to make your life easier.