FEA problem in Solidworks 2011-2012

[povelocj]

I am having problems doing FEA with the new Solidworks edition. I was able to FEA my chassis and analyze torsional rigidity in last years Solidworks, but I'm having problems with the new version.

The main problem is in the front rollhoop where I have a 3 inch radius in a few sections. The attached picture shows the errors I get.

I get an error message on the upper bends, something about neutral axis (see picture).

Also, the lower bend isn't being treated as a beam either.


Has anybody had similar issues and found workarounds? I had no problems with this in Solidworks 2010-2011, but I'm having issues now with the 2011-2012 software

http://i52.photobucket.com/albums/g14/povs9/problems.jpg

[povelocj]

I am having problems doing FEA with the new Solidworks edition. I was able to FEA my chassis and analyze torsional rigidity in last years Solidworks, but I'm having problems with the new version.

The main problem is in the front rollhoop where I have a 3 inch radius in a few sections. The attached picture shows the errors I get.

I get an error message on the upper bends, something about neutral axis (see picture).

Also, the lower bend isn't being treated as a beam either.


Has anybody had similar issues and found workarounds? I had no problems with this in Solidworks 2010-2011, but I'm having issues now with the 2011-2012 software

http://i52.photobucket.com/albums/g14/povs9/problems.jpg

[RobbyObby]

Sometimes Solidworks gets buggy when running beam analysis n a tube with a small bend radius. Somtimes all you have to do is go back to Model tree and delete the Structural member in question, then remake it and it fixes itself. Iv'e also found sometimes that changing the mesh size can often fix it as well.
However, what I usually do is model a very simplified frame initially, with very little bends, and mitered joints instead. (In your case you might only need to simplify the lower bend. Having bends with very shallow angles usually ends up giving you the meshing error you have run into.) Once the initial iterations are done, I then perform a final FEA of the correct frame using shell analysis.
Dont know if that was any help as it is a very compromised solution.

[Andrew07]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by RobbyObby:
However, what I usually do is model a very simplified frame initially, with very little bends, and mitered joints instead.
</div></BLOCKQUOTE>

SW simulation looks at that piece as a separate beam but it doesn't meet the required length/diameter ratio for the beam approximation. The result are the problems your having. Another possible solution in addition to the one offered above is to use mixed meshing. Beam elements for the beams and solid meshing (or shell) for the stubby beams/corners. I happen to hate mixed meshing in SW though, it is very frustrating and it seems to take away the ability to select joints for loads and fixtures.

[RollingCamel]

I remember something about reassigning all segments as beam parts and their material, along with creating a new simulation.