SWW 2005 - The coolest FEA things.

CosmosWorks added an impact analysis to the list of analysis types. I
played with it for quite some time to get a feel for what it could do.
It should come in handy for those who have to meet some kind of impact
or drop criteria. It is limited to impact with an infinite flat plane
and the area of contact is also limited to a small patch on the model
closest to the impact plane. It can't do evaluation of an item in a box
just yet. That would require moving over to GeoStar and a lot more
work.

NEiNastran is a real up and comer with the addition of assemblies and
the flexibility of using hex mesh where appropriate. What I really like
about NEiNastran is that the analysis part runs completely separate
from SW and there is a file called a bulk file that documents the
analysis. This means that a large analysis can be run without tying up
SW. It also means that a permanent document of what was analyzed
remains (the bulk file) regardless of any subsequent changes to the
geometry in SW. SRAC's DesignStar also has this standalone capability.

It makes a lot of sense to do the meshing within SW. After all, the
meshing is tied intimately with the geometry. But it really doesn't
make sense for the solver to run within SW because the solution phase
is just a big calculation. The mesh should be documented separate from
any changeable geometry so that what was calculated can always be
traced back to a particular geometry for which it is valid.

Also, because NEiNastran writes out the same data that its big brother
NE Nastran uses, a knowledgeable analyst can have the ease of use of
the SW interface with the power of a full blown analysis package for
advanced analysis types. They have combined the ease of use that
designers and beginners need with extendability to a fully featured and
fully supported high end FEA package.

Algor and ANSYS were also there. I didn't bother looking at either
booth, but for very different reasons. I have a long standing bias
against Algor based on the comments of other analysts. On the other
hand I have a lot of respect for ANSYS and know that it is a solid
product. I just didn't have the time.