Buckling Factor Results Differ Between ASim & Nastran

[bryan_mcnatt]

I have a beam model network that I'm analyzing for buckling (Linear Critical Buckling Load).  For grins, I decided to run the same model using Nastran.  I get wildly different results.

I went back to the Beam Tower tutorial and ran that buckling analysis under both ASim and Nastran solvers.  This time, both results were the same.

So, I suspect there's something about my model that kicks ASim off in one direction and Nastran in another.  Or, maybe, something in my model that ASim deals with better that Nastran stumbles over (or vice versa).

I can't yet attach the model to this post due to proprietary info in it.  However, I thought I'd ask for hints or advice about what to look for in my model or in the setup of the analysis that might lead to different results in ASim and Nastran.  Anyone have any experience or insight that might apply?

Thanks!

[AstroJohnPE]

Hi blaeiouk,

I have two ideas, but these are both just a guess.

1. The orientation of the beam axis would be the simplest way to get different results - if one processor interpreted the beam's strong axis to be pointed in one direction and the other processor interpreted the strong axis to be pointed in another direction. But I think that any problems or discrepancies with the beam orientation is resolved before the model gets to the processor. In other words, the orientation is created during the "Check Model" or "Verifying Geometry" phase, not during the actual analysis.
2. I do not know if the shear area has an effect on the buckling analysis or not. If one processor uses the shear area and the other does not use it, that could lead to a difference. You should be able to test this by entering zero for the shear areas (SA2 and SA3).