Interpretation of results (buckling analysis) 1

[jeram123]

Hi,
I have tried to perform a buckling analysis for a simple hollow box (1000*450*300mm)using PATRAN. One end of the box is fixed and other end is loaded giving a transverse load 0f 900 kg (450 each at the ends of top and bottom panel)from down side. I got a value nearer to .06 through eigen value analysis.The buckling mode runs upto a length of 600mm and stops.
How can i verify the credibility of my FEA results.Please help me to sort out this problem.

 

[ESPcomposites]

- The eigen solution algorithms can find both positive and negative eigenvalues if you do not specify that you only want positive values. For example, if you had a uniaxial tension load on a beam, the eigenvalues would be negative. In your case, the applied load creates both tension and compression stresses on the box, so you have positive and negative eigen solutions. As previously mentioned, they occur at about the same values if the box is symmetric.

- Yes, Nastran can do a post buckling solution, but you will need to use a nonlinear solution (as opposed to the SOL105 eigen solution). SOL106 is the basic one, but you can also use SOL600 (Nastran has another one as well...400 I believe, but have not used it)

Brian

http://www.espcomposites.com

 

[jeram123]

Thanks rb,esp comp & taz...

Did anyone tried some postbuckling analysis? Is that possible with a simple euler beam?

 

[rb1957]

post-buckling analysis is long but not stunningly Hard.

i don't think you'll get much mileage out of a post-buckled comlun ... the point of post-buckled analysis is that one element (typically a thin skin) can buckle (typically in shear) but the structure doesn't have to collapse ... it can redistribute the loads and carry (limp) on.

 

[jeram123]

Thank you rb....can you give some idea on how to carry out a postbuckling analysis in patran?

I have tried MSC's tutorial,the following link......

http://www.mscsoftware.com/.../Lec13_Lesson19_Post-Buckle_Plate.pdf

but I didn't find it useful.

 

[ESPcomposites]

From my previous post:

"- Yes, Nastran can do a post buckling solution, but you will need to use a nonlinear solution (as opposed to the SOL105 eigen solution). SOL106 is the basic one, but you can also use SOL600 (Nastran has another one as well...400 I believe, but have not used it)"

As far as "how" to do perform the analysis, you have not yet defined what type of post buckling you are interested in. A few forms are:

- Local instability (such as a cap or web of a section)
- Skin buckling via compression (and redistribution to stringers)
- Shear buckling that results in diagonal tension (the load redistributes as a bending).

But in general, you would apply your "load" and perform a geometric nonlinear analysis. I say "load" because you may be more interested in a fixed displacement. For example, take bullet point (2) from above. You may want to apply a fixed displacement and let the load redistribute (i.e. the skin load will not be linear). You must perturb all of the modes in general.

But first, familiarize yourself with the classical approaches for:

- local buckling
- crippling
- the concept of effective
- diagonal tension

See the books by Bruhn, Niu, or Flabel. You probably don't want to perform a nonlinear FEM until you understand the physical behavior.

Brian

http://www.espcomposites.com

 

[rb1957]

NACA 2661 and 2662 are the basic references, if you're interested in post shear buckled strength. Bruhn and Niu, of course, cover it too. i'd start there, as they explain briefly (readably) what's happening.

 

[jeram123]

gr8 work rb & ESP...

I will try to make my question clear..

When talking about a simple euler column, can we say that there will be entirely no post buckling capacity?

 

[rb1957]

as in everything ... that depends !

if this is truly single loadpath structure then i'd say yes there is no post-buckling strength.

but if there is an alternate loadpath that would pick up more load once the primary loadpath became soft (ie yielded/buckled) then there is post-buckling strength in the structure (though not in the column).

 

[jeram123]

well said rb1957!
But in my problem the buckled part have it's unloaded edges SS; one loaded edge Clamped and other SS.

Can i find solutions for this BC's from any books or through any other means?

 

[rb1957]

?, how's that relate to an euler column ? sounds like you're back talking about the original problem ?

my 2c, FWIW, ...
1) use 4 sides SS (as being slightly conservative), or
2) an FE model of the panel, or
3) test something.

personally, i'd expect 3) to be impractical, but either 1) or 2) should get you an answer.

 

[jeram123]

i did (1) & (2), But both are giving different answers...
i think the problem is with b.c's (4 side ss)

So i need to know if we can solve 1 side ss & other clamped loading case through any theory

 

[rb1957]

i'd expect 1) and 2) would give different answers.

if you have to solve this from 1st principles, i'd suggest timoshenko "plates and shells"; also a thick pad of paper, lots of pencils, and lots (and lots) of coffee.

 

[SWComposites]

agree - 1) and 2) should give different results

but start with Timoshenko and Gere, Theory of Elastic Stability ...