Re: Convergence using Solidworks Express

This has been my contention all along with these so called "spell checkers". There is no real way to do a validity check on these models. If there is anyone out there using this junk as a real design tool they are in deep trouble.

FEA is neither easy nor cheap. The FEM vendors would like you to think that it's all handled, but it isn't. There are real issues to be dealt with when doing FEA analysis.

I've used NE/Nastran for years and stick by their tools. Yes, FEMAP is a royal pain, but I can do a thorough check of my boundary conditions and results when I'm done. Noran Engineering had embarked upon a SolidWorks add-in development before the inclusion of CosmosExpress, but decided to stop because of it's inclusion in 2003. I'm told that they realized shortly after introduction that CosmosExpress was nothing more than a loss leader to get people to buy real Cosmos, so they have restarted their development. I'm really looking forward to having this tool available. Having an easy to use front end for one of the most advance FEA tools on the market.

Chris

I didn't see your original, just Chris's response (which brings up very valid points).

What is the wall thickness on your channel? I seem to recall that a rough rule of thumb is that you need at least 3 solid elements across a wall to get reasonable results.

Do you have sharp corners or very small radii? Perhaps you are seeing "real" stress concentration effects showing up with the smaller elements.

How did you apply the force couple? If you apply forces over very small areas, then the stresses are going to be very high locally and perhaps the smaller elements are showing that effect.

I would guess from what you wrote that you've already thought about these types of issues, but it's all I can think of.

Jerry Steiger Tripod Data Systems