Stability analysis - European codes vs US codes

[stillfan]

I am running through some calculations on an aluminum structure. In the longitudinal direction it is a braced frame and the lateral direction it is a moment frame. The frame is not the typical column/ beam connection. The column I believe would be what they refer to as a lean to. Where it is a beam/column and then the brace of course is still analyzed as a beam. In order to comply with stability requirements, I have chose to use 2nd order analysis in combination with the allowable stress design equations supplied in the Aluminum design guide. My counterpart in Europe is telling me a couple of things that just don't make sense to me.

First he is stating that their code states that if they perform 2nd order analysis in FEA (global analysis) then they do not have to perform individual component checks (local). Of course they are using von mises stress analysis, so his actual stress, what he refers to as the section stress. He then compared his section stress ratio to LTB stress ratio/ local buckling stress ratio/ torsional stress ratio. In a manner laid out below;

section stress = .97
local = .85
lateral torsion = 1.91
torsional = 1.75

Now the results I am getting for my ratios are similar to the 1.91 ratio (way overstressed). He is saying that because he performed the FEA analysis he only has to look at the section stress and not the buckling effects as this is taken care of in the global stability analysis. I honestly think that this is a misinterpretation of the European code procedure, how could the european code be so radically different than the american code. Has anyone else had this problem? How is it that the american code specifically states that you must also perform section checks based on formulas that account for eulers buckling, elastic vs inelastic deformation requirements, section properties and the european code says the exact opposite?

Another red flag is that for lateral buckling bracing, in a frame that see's double curvature, you must brace both the top and bottom flange. This is not in his current design, obviously the FEA analysis does not account for proper bracing. It has been brought to his attention but again since he performed 2nd order FEA analysis he is again under the assumption that he does not have to account for this.

Anyone encounter these items? Any feedback would be appreciated.

 

[stanmorek]

I don't have a lot of experience with aluminium as a structural material and have been getting to grips with Eurocode 3 Steel structures. But yes as I understand it what your European colleague says is correct but it is only one of a number of approaches allowed which also includes what you're doing. Given the stiffness of Al p-delta effects would be a concern to me. I would be asking questions about his QA procedures in modelling and interpreting the analysis and perhaps want to see some back up calculations.

Appendix A in the linked document has a flowchart on the analysis process. Its for steel but the suite of Eurocodes are written with a common approach to designing with the major materials.

Link

 

[stillfan]

Thank you, this will help alot. The first thing I see at a glance is that they have a diagram with both flanges braced of the I-beam. Now I know if it is a compact section it won't be a concern but the aluminuim I am working with is slender and even though it is a box beam/column it is aluminuim, which makes it a terrible material for buckling.

The aluminuim design manual says that you must have bracing on both flanges anytime a beam goes into double curvature. Do you know if the EN1999 says anything about this?

This is going to seem like an elementry question but I have never compared the Von mises stress analysis to the working stress analysis. Do you happen to know if there is a huge difference between the actual stresses results between the two? I would assume not, since they are both looking at the combination of local stresses.

Thank you,