"Buckling" critical load case

[airmail]

Hello all!

I want to select the most critical RF for a composite panel submitted to axial, transversal and shear flow. It is modeled by several CQUADs.

I've run the FEM in Nastran and I have the flows in each CQUAD. But the question is: How to guess the most critical load case of buckling? To take lowest flows (mixing load cases) is too conservative, isn't it? Is there anyone who has faced on the same problem? Which was his criterion?

Thank you all in advance.

 

[GregLocock]

Usually the lowest buckling mode is the one that is practically important. There is nothing conservative about it.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[rb1957]

i think you are considering a diagonal tension problem, ie post-buckling strength of a skin/stringer panel.

you've meshed a panel in some detail, but i think you can really only apply one number, like an average, for each stress (bi-axial and shear). i'd pick either the numerical average or the geometrical average (closest to the center of the panel), or conservatively the largest stresses (maybe too conservative). you can do this on a case consistent basis (probably a lot of number crunching, but then that's why we have computers), or (as you note) a conservative envelop.

good luck

 

[airmail]

Greg, I'm not referring to buckling modes.

I'd like to know which is the proper criterion to read flows when a panel is modeled by several QUADs in order to select buckling critical loads.

I mean, which load case has the most desfavourable components (Nx, Ny, Nxy) that give me the lowest buckling failure?

Thank you very much for your interest.

 

[rb1957]

airmail,
can you calrify "composite" ... skin supported by stringers, non-metallic (plastic/carbon), or sandwich panel?

thx

 

[SWComposites]

airmail -

In most cases you can not look at individual element loads to determine the buckling load of a structure. You have a couple of options:
1) run an eigenvalue analysis in the FE code. This will give a moderately to severely unconservative result depending on the structure, loading, etc. A minimum reduction factor of 0.75 times the minimum eigenvalue is generally recommended.
2) run a nonlinear FE analysis with an initial imperfection (which can be calculated using the eigenvalue analysis and automatically applied by some FE codes) and determine the maximum load capability
3) use a hand analysis with classical buckling equations (which are based on eigenvalue analyses so you need to apply appropriate reduction factors). This sounds like what you are trying to do. The problem here is what loads to use. If for instance you have a flat plate modelled with lots of elements, then you either have to use the maximum loads from all of the elements or you have to determine an "average" loading across the panel (which requires some experience with the type of structure that you are analyzing). In either method, you cannot use the Nx, Ny and Nxy loads separately, rather you have to either combine them into the buckling equations or have to use an interaction formula.

You need to describe in much more detail what you are trying to analyze for us to be able to give you more specific advice. What is the structure, adn how was it modeled in the FE analysis? How specifically are you calculating buckling loads?

 

[airmail]

Hello again:

The panel is made up of carbon fiber (fabric). It is one of the panels that are between the leading edge and the wing torsion box. For the moment, I'm only considering in-plane loads.

What I was thinking to do is to take loads from the big FEM (it includes the entire wing) and then, insert the flows obtained (with a proper criterion) in a detailed panel. SOL105 should give me the buckling RF.

Thank you all for your interest.