BLF safety factors with real world problems.

[bill8123]

When calculating buckling load factors with FEA programs such as Pro/Mechanica what type of safety factors do you guys use for real world problems? e.g. Thin walled cantilevered beam under static bending loads.

I understand FEA does not consider real world manufacturing inaccuracies etc. Hence the question.

Are there any codes or standards that cover this?

Thanks

 

[EngAddict]

AS1210 pressure code uses 3 for theoretical elastic buckling and 2 against the lower bound of experimental buckling results. For structural work you could probably reduce this a little but if you know it is prone to buckling then keep it around 3 for those items. I have seen BLF as low as 1.2 to 1.5 used but I think this is best left for conservative hand calculations only.

 

[bill8123]

Thanks fellow Aussie.

 

[ESPcomposites]

It would depend on the industry. In the aircraft industry, where weight is critical, a factor of 1.0 to 1.20 might be used.

The lower factor is more relevant for the short/intermediate column range where plasticity (Johnson part of the Johnson-Euler curve) is often slightly conservative compared to empirical data.

A FEM could account for the real world scenarios (manufacturing tolerances, imperfections, eccentricities, etc.), but only if you tell it to. But because it is not usually practical (nor are the scenarios always easy to quantify), we often just solve the nominal solution and apply a margin of safety based on empirical data.

Brian

http://www.espcomposites.com

 

[crisb]

If you are talking about load factors obtained from an Eigenvalue buckling analysis (as against a non linear analysis), you need to remember that these are only the ratio of applied loads to elastic critical loads, and do not consider the yield stress if this gives a capacity less than the elastic critical load.
For structural work this is codified in Eurocode EN1993-1-1 and similar codes. In simplistic terms the buckling load factor is back calculated to give a slenderness ratio which is then applied to the strut curves that include allowances for initial imperfections and residual stresses. There is no one simple additional factor that can be applied and the final design load factor in structural work can often be many times less than that obtained from an Eigenvalue buckling analysis.

 

[rb1957]

it shouldn't be a matter of what load to apply to an FEM ... it should be a matter of what load are you designing to ? what code ? (a/c generally use 1.5)

 

[EngAddict]

Buckling analysis is performed on top of stress analysis, it has not been suggested it would be used instead.

As you say permissible stress design will give you a slenderness ratio that can be compared to allowables that incorporate allowances. Limit state will use the elastic buckling load factor to calculate a moment amplification factor. Both of which are not useful in numerical analysis and so a suitable BLF must be selected.

I don't know of many structural codes that give BLF for FEA. There aren't many that even cover FEA with allowables for local stresses, and classification of primary and secondary stresses. When modelling a structure in FEA generally the imperfections in material and manufacturing are not modelled or able to be modelled so this must be included in the BLF. Unfortunately a lot of this comes down to engineering judgement without specific design codes.