Beam stability against LTB 2

[BMart006]

In AISC, the equations used for determining beam strength based on LTB are developed from the assumption that the beam is restrained adequately against torsional rotation at the supports (ends for a simply supported case). Suppose a beam is laid across a single span and supported by the ground on either end, but it is in no way mechanically fastened (bolted/welded) to the ground. Would you consider that to be restrained? How would you modify the allowable moment (or Cb factor) to accommodate such a case?

 

[WARose]

Without rotational restraint at the ends, to me, this violates one of the requirements of LTB as per the code.....so the code equations (as far as buckling goes) are no longer applicable. Therefore, in the situation you are describing, I would treat it like a suspended/lift beam. There have been a lot of discussion about this here over the years.......here is one good thread:

http://www.eng-tips.com/viewthread.cfm?qid=410584

 

[hokie66]

Unless the beam is restrained by other means, I would consider it unstable, so buckling is the least of your problems.

 

[WARose]

Good point hokie.....I wondered that myself....but kept the focus on buckling.

 

[handofthelion]

Unless the beam is restrained by other means, I would consider it unstable, so buckling is the least of your problems.

Does this depend on how the load is applied? If the load is applied below the rotational axis (e.g. neutral axis for an I-beam) is this still not stable for vertical loading?

 

[hokie66]

The main issue with stability is the way the end supports are configured. The OP's beam is just sitting on the ground, so nothing to prevent it rolling over.

It is worse than a suspended lift beam, as lift beams are supported at their ends from the top. They can swing, stably.

 

[BMart006]

I realize what I'm asking wouldn't meet building code. Actually what got me thinking about it was the situation where you raise a load, supported by two or more beams, with jacks. For example, lifting an office trailer or other load with two beams underneath the load for jacks to raise it. The jacks wouldn't restrain the ends and it seems unstable, yet I've seen it done countless times. So I'm looking for a good way to predict the behavior.

 

[dik]

Hokie... the location of the load application point relative to the shear centre may have an influence on the stability.

Dik

 

[Lomarandil]

For a beam resting on supports, but not torsionally restrained (e.g. your jacks, or cap beams that are just sitting on each other) -- there was a British standard that proposed using Lb' = Lb+2d. Has seemed to provide reasonable results for me in the past.

Can't find my reference for it, but the internet says BS5950 4.3.2 Table 13.

----
The name is a long story -- just call me Lo.

 

[bearjew]

For a beam resting on supports, but not torsionally restrained (e.g. your jacks, or cap beams that are just sitting on each other) -- there was a British standard that proposed using Lb' = Lb+2d. Has seemed to provide reasonable results for me in the past.

Can't find my reference for it, but the internet says BS5950 4.3.2 Table 13.

Would this approach also work for checking a lifting beam for LTB?

 

[Ingenuity]

Would this approach also work for checking a lifting beam for LTB?

For LTB of lifting beams here are two reference papers:

1] "Buckling of Suspended I-Beams" by Dux and Kitipornchai, ASCE, Journal of Structural Engineering, Vol 116, No. 7, July 1990.

2] "Stability of I-beams Under Self Weight Lifting" by Dux and Kitipornchai, Australian Steel Construction Journal, Volume 23, No. 2, 1989.​

 

[Lomarandil]

bearjew -- no, very different problem.

A beam resting on supports (but not connected) is rotationally restrained by the reaction extending across the beam width. This is then discounted as I mentioned above for web flexibility (versus a traditional bearing stiffener) between the bottom and top (compression) flange.

A lifting beam is rotationally restrained only by its own self weight relative to the point of rotation imposed by the rigging.

Ingenuity's two references are some of the best I've found to address the subject, although they're a little too academically focused to apply to some situations.
For other references, check out WARose's post above. I'm not sure I'm entirely sold on Helwig's method, but it's very practical to apply.
Regardless -- this is a problem without a rigorously defined solution (as far as I'm aware). And I've been looking for years.

----
The name is a long story -- just call me Lo.