curved columns

[bolt04]

I need to design a curved column to be supporting a eccentric load.
any help in the calculations and best formulas to use in working the max force and stress it can take
cheers

 

[rb1957]

look up "columns with eccentric loads". look at how they solve the general case (a straight column with an off-set load), and you'll have to modify the calc for your curved column. maybe you're a curved "beam column" ?

i don't think this is like the standard "euler" column analysis, where a straight column snaps through; i think this is more like critical stress = yield stress. ie, the beam is initially curved and the loading only increases the curvature.

 

[structSU10]

If you have a structural program that can do P-delta analysis, if you input your column geometry, segmenting it properly, and apply your eccentric loads and such, you should get a its response. If you want a hand calculation, i know i have seen buckling formulations for initially crooked columns and eccentrically loaded columns, and you may be able to derive the solution for a combination of the two.

 

[JoshPlumSE]

I'd agree with StructSU10. But, would add the following:
1) This is an example where AISC's Direct Analysis method will really help. The stiffness reductions included in the DA Method are intended to capture the INelastic buckling whereas a pure P-Delta analysis would only capture the elastic buckling.

2) If the column is curved only in one direction, you will probably want to apply a notional load in the other direction. Because, a small eccentricity could cause some de-stabilizing torsion in the cross section that you would want to be taken into account for the P-Delta analysis.

 

[bolt04]

actually its going to look more like a curve beam column. Iv done some FEA on the part just need to do some manual calcs and work out where the least stressed areas are as it needs to be as light as possible while still holding the load.

 

[IFRs]

Doesn't the FEA tell you where the least stress is?