Stability of SS Beam w/o bracing

[NS4U]

I am evaluating a structure that has an existing simply supported steel wide flange girder. There are timber joists that sit on the girder, but are not rigidly connected to it. The girder sits on a bearing plate, and has no stiffeners, nor any lateral support at the ends.

Essentially the beam has no lateral bracing, but is OK- no signs of instability, it could be getting stability from the friction with the timber joists, or the section could be stiff enough all together where lateral stability does not need to be provided.

AISC pp. 2-13 says something to that effect: “beams with thick webs and relatively shallow depths could be properly designing without providing lateral stability”

How do I check this though? Appendix 6 doesn’t seem to contain any thing that address if stability is required, it just seems to assume that bracing is required. I’m not an expert on appendix 6 though, so maybe I am missing something.

Thanks

 

[Lion06]

AISC does require all beams to be torsionally restrained at points of support. That's the fundamental assumption for all of the strength equations.

That being said, if you have a shallow beam with a relatively wide flange, it could be stable without the aforementioned torsional restraint. Additionally,even though you can't quantify the bracing provided by the timber joists, it is providing some.

 

[NS4U]

Thanks. Where does AISC say that?

 

[Lion06]

App. 6.3 (pg 16.1-193) "At points of support for beams, girders, and trusses, restraint against rotation about their longitudinal axis shall be provided."

 

[Lion06]

Also, page 2-13, under the STABILITY BRACING section says, "Beams, girders, and trusses must be restrained against rotation about their longitudinal axes at points of support (a basic assumption stated in the preamble of Specification Chapter F)."

The preamble to chapter F states, "This chapter applies to member subject to simple bending about one principal axis. For simple bending, the member is loaded in a plane parallel to a principal axis that passes through the shear center or is restrained against twisting at load points and supports."

 

[NS4U]

Ok thanks... now my problem is that I am not modifying or adding load to the existing beam, I'm just switching out it's supporting column.

It my review of the framing, I noticed the beam doesn't have any positive bracing. So I was wondering if I should add some bracing, but if I can show the beam is OK as is (by the code) then I won't bother.

What do you think? How would/could I approach this?

 

[Lion06]

I think it wants to have bracing. Will the column have any lateral brace other than the beam? Does the column just come up the underside of the beam and stoop? Does it connect back to anything? If it does, then provide stifferners on the beam and a bolts from the beam flange to the cap plate. If it doesn't, come up with something to brace it, because that, in my opinion, is a stability problem.

 

[trainguy]

You could make a good connection to the new supporting column - that would help. You may end up with more of the required torsional restraint at supports.

tg

 

[paddingtongreen]

This is an interesting version of a question I've discussed on a couple of occasions, off the record, with engineers connected with AISC. Landing on bearing plates, the bottom flange is restrained against rotation, but not necessarily the top flange. The joists sitting on the top flange prevent rotation of that top flange. Whether the joist/beam interface has enough friction to hold the beam in place is irrelevant as long as rotation is prevented, as it surely is in the OP's case. The guys I spoke to didn't argue for or against my analysis.

Michael.
Timing has a lot to do with the outcome of a rain dance.

 

[Lion06]

paddington-

I don't believe that bottom flange bracing is adequate to consider the section braced. It wants to be braced at or near the compression flange. The only way that bracing the bottom flange works in a bearing connection is to provide stiffeners and a positive connection to the support.

 

[NS4U]

The beam will sit on bearing plate, and there will be 4 bolts through the beam flange into the bearing plate.

I don't want to add stiffners to the beam for two reasons:

1) I don't like the idea of welding to beam that supports timber framing (fire hazard).

2) I'd have to shore the beam to weld it, which would be expensive- since I can't weld a simply supported beam while it's under load.

 

[NS4U]

Correction- Meant to say I can't weld a simply supported beam while it's under load *cheaply*

this is really outside of my scope and budget as well as the owner's so I'm trying to get something simple done to convince myself it is ok or add a cheap (cost-wise) fix.

if I absolutely have to add bracing, I'm thinking of some sort kicker from the top flange (or possibly centroid?) back to an existing joist near the support.

 

[Lion06]

Why can't you add stiffeners while it's under dead load? We;ve done that before on retrofits where the original engineer left stiffeners off.

My biggest concern is that a column coming to the underside of a beam with no positive connection to anything is inherently unstable. How do you even just the column being braced at it's top at that point?

 

[NS4U]

The column is anchored to a footing (cantilever).

 

[cntw1953]

From old ASD point of view (a remote memory):

If there is no problem with web cripping (thick web), and the flange is capable of handling local bending due to rotation induced by incidental eccentricity, plus all atresses along the beam are within allowable that was derived without coniidering intermediate braces, you are good to go without end brace, nor bear stiffeners.

The requirement from the newer code could be different. Also, I could be wrong. Check it out.

 

[ishvaaag]

Also, bracing can also be obtained from wood elements. For the structure you seem to have, not an entirely disdainable option.

 

[BAretired]

Wood joists usually bear on a wood nailer bolted to the top flange of the beam. Joists are typically toe-nailed to the nailer. When built this way, the top flange of the beam is considered laterally braced.

If your joists bear directly on the steel flange with no physical connection, you could use a pair of steel straps spaced at about four feet on center nailed to the joists and bent around the beam flange to prevent lateral movement.

BA

 

[Stillerz]

This situation, as described, sounds a hell of a lot like 95% of homes in the U.S. that have basements.
95% of which have "unstable" lally columns supporting a steel beam which could not really be considered cantilivered even though the column is buried in a few inches of concrete.

Another requirement of AISC, if I remember correctly, is that all members with "unframed ends" must have a pair of stiffners. In a home, with joists 16" o.c., I would consider the ends framed.