buckling "safety" factor 1

[Ing.Vedder]

Hi all,
I've done a buckling analysis in order to investigate the lateral torsional buckling of a particular cantilevered steel beam on two support (here the schema:Link)
I've found a buckling factor of 8.
I would like to know which is the "minimum" buckling factor to take in order to be sure that the steel beam will not buckle during the functioning.
I would like to be totally sure (it is an heavy functioning).

I don't know if there is another method to investigate the stability due to the particular section of the beam and to its statical schema. I've tried with eurocode 3 but i failed, so buckling analysis is my last chance.

thank you

 

[Eagleee]

Any chance you can do a static non-linear analysis? With a linear buckling analysis, the effects of imperfections (geometrical and residual stresses) are not accounted for. Depending on what software you have access to and in how much detail you want to go, one option would be to do a FE analysis on your beam in which the initial geometry includes an imperfection with the shape extracted from your LBA (worst case scenario). In this case, you can choose a conservative value for the amplitude of this initial imperfection. In the EN1993-1-1 you can check 5.3.2 (3)b which is the (much simpler) case for flexural buckling.

 

[klaus.]

Well, there is no limit number for a buckling factor ...of course need to be more than 1
You need to do a nonlinear analysis .... including imperfection
and then 2 things need to be fulfilled
1 ) stable solution
2 ) max stress smaller than design limit








best regards
Klaus

 

[Ing.Vedder]

thank for your help.
the item is a steel part of a heavier machine. Which are the imperfection that i have to consider? I must imagine some imperfection?

Furthermore, i would like to use Eurocode3 for checking the lateral torsional stability of the beam, but i don't know in this case which is the Lenght between two lateral torsional supports.
The restrain on the left is a very rigid pinned restrain that block the beams to twist (so is a torsional restrain), while the restrain before the cantilevered part is a small wheel unde the beams that support the beams and guide it against lateral movement but don't restrain the beams to rotate on its longitudinal axis and so can twist.

 

[klaus.]

Are you doing 'nad ' calculation or FE Model ?
Using FE Model you do not need to worry about length....if the FE Model ist correct and considering the correct boundary conditions

Imperfection ..well yes...a buckling shape can help to find the relevant imperfection shape
if you are not sure...you need to consider different imperfections




best regards
Klaus

 

[Eagleee]

You do the buckling analysis to get the imperfection so you dont have to imagine it. The first eigenmode will give the lowest buckling factor and will be the worst. Of course, things can get much more complicated in reality, if loads do not act at the same time, etc...

The Eurocode is notoriously bad in prescribing anything that has to do with LTB. I would suggest you have a look at 'Rules for member stability in EN1993-1-1' which is ECCS publication No. 119. As I remember, they have towards the end of the book a section which touches upon beams with overhangs, as yours.

That being said, with LTB it is becomes very difficult very fast as you move away from the standard simply supported beam case. You have a varying section, beam with overhang, loads acting both in the strong and weak plane of the c.s, filled RHS...you are as far as it gets from the standard case. The only thing that would have missed was to have the beam made out of some exotic material.

 

[Ing.Vedder]

@Klaus:
what is "nad" calculation? I've done a FEM buckling analysis using Sap2000 and schematizing the beam using Shell (Thin). The lenght is necessary if i perform a calculation using Eurocode 3 standar method to check stability.

@Eaglee:
i totally agree with you, thank you for the tip abou ECCS publication, i'll try to find it even if seems to be "unavalable"..
Just a note: the unique loads acting on my beam are the self weight of the beam and the concentrated load of 30 kN at the end of the cantilevered part.I've a varying section, that's true but no loads are acting on the weak plane; the arrow you see on the side of the beam are only leader to indicate the section of the beam, not forces! Fortunately the steel is normal carbon steel for structural purposes!

thank again both

 

[Eagleee]

 

[klaus.]

Sorry...I meant 'hand' calculation


best regards
Klaus

 

[WARose]

I've worked backwards myself (from the [American] code to the LTB & Euler buckling equations) to see what kind of saftey factor is built in to the code for AISC steel shapes. If memory serves (and this is eliminating any possiblity of localized buckling from thin flanges or whatever).....the buckling saftey factor was on the order of about 2-3.

Obviously that changes when you get into thin walled structures (i.e. imperfection sensitive structures).