Buckling of a thin-walled beam

[Zaliel]

Hi to all, this is my first post, I hope I am not breaking any rules.

I'm tring to better understand how to model correctly the buckling of a beam with patran/nastran so i created a simple model for wich i already have the correct result of primary and secondary buckling.

I have made a 1D model and a 2D model, and I have a couple of questions:

1) Is there a way to filter the results of the 2D model to recognize which factor is relative to the primary and which is relative to the secondary buckling load?
(or do I have to plot the result and check the results from the shape of the deformation?)

2) In the results of the 1d model, the first 20 results I obtain are buckling factors with a value that is way lower than the one I have calculated through the euler equation (also they are all the same), but the 21st is the right value (the shape of the deformation of this result is correct the others are not). Is this normal or is there something wrong with my model?

For reference:
I'm working on a C section beam.
The web of the beam is 0.08 meter long
the 2 flanges are 0.04 meter long
the thickness is 0.0002 meter for the flanges and for the web
the lenght of the beam is 1 meter
the material is alluminium (E of 71*10^9 Pa)
one ended of the beam is fixed the other is free
the beam is loaded with a axial load on the free edge

Thanks

 

[crostolo]

1) You have to plot the result and check the results from the shape of the deformation

2) I suppose your beam, in the model, is made by 20 elements. Right?

The reason is due to the shape of beam, and the model used by the solutor (Tymoshenko) that is different from what the theory use (Euler). Try to modify the value of shear along X and Y on the beam section in order to use the Euler model and the first twenty "false" values disappears.

I had a very very similar example when I use student at university a couple of year ago.

 

[Zaliel]

thanks crostolo, I will try to do as you said.
It didn't cross my mind that the cause of that problem was that the model used by the solutor is different (neither i thought about the correlation between the number of elements and the number of "false" values).