Lateral Stability of Beam with Compreesion Load 1

[Milco]

Hi,
I am sure this question has been answered in the past but I cannot find a clear answer to it.
In my example, I am looking at the collector beam in a simple one-story braced bay.
The axial compression force in the beam from the lateral seismic forces is, say 40kips.
The joists run parallel to the beam.
The beam top flange is continuously supported by the 1 1/2" steel roof deck.
The beam bottom flange is unsupported.

My question is;
When checking this beam for axial compression, is this member consider continuously laterally supported by the deck even though the bottom flange isn't supported? Would adding a stiffener at the mid-span of the beam reduce Ky to 0.5?

In the past, I have added kicker braces from the bottom flange of the beam to the top flange of the adjacent OWSJ to cut down Ky but I was wondering if this is necessary.

Thanks!

 

[JAE]

For a collector with that much load, I'd tend to not depend only on the deck.

 

[RPMG]

It is braced in the Ky direction, but not against torsional buckling. However, the most engineers analyze it as minor axis buckling and brace the bottom flange because no one wants to analyze the beam flange for torsional buckling.

Some engineers assume the resultant force is higher than mid-depth, in such a manner that the tension flange remains in tension. The logic for this is similar to the uniform force method for connection design, in that the forces can be resolved anyway that the problem can be solved, because any solution will be more conservative than the true answer, which is unknown to us.

 

[JAE]

RPMG - that makes perfect sense.

 

[Milco]

Thanks for the replies!

So adding a stiffener still would not restrain the member against torsional buckling and a brace is needed to prevent rotation?

My concern about assuming the resultant force is higher than mid-depth is;
Since this beam is parallel to the OWSJ there is not much gravity load on it so I am concerned that the bending forces will not keep the bottom flange in tension.

 

[structSU10]

you could have a note on the drawings to have to joist supplier provide certain strength and stiffness of bridging to try and brace the bottom flange?

 

[phamENG]

structSU10 - that's asking a lot, especially if it's a simple building. Probably better off looking at the joist spec and, if needed, defining a unique bridging size yourself if one is needed. Your idea may work in a complex building with a lot of custom designed joists where the joist engineer is already heavily involved. If you're just picking standard designations out of a manual, a detailer will lay out the shops and it will only get a cursory look from the engineer before it makes its way to your desk.

Milco - Look at a section and imagine the beam section not as part of a beam running out of the page but as member in a FBD in the plane of the page. Is there anything in that "frame" to stop it from rotating? No, there isn't. Adding stiffeners doesn't prevent rotation. It will prevent warping of the cross section, but there is nothing stopping the section as a whole from rotating. So you either need to add a discreet brace to stabalize it, or you have to bring the length of the beam back into the picture and consider the stability of the entire beam, which without a brace will be a consideration of Torsional Buckling as RPMG stated.

 

[RPMG]

No, a stiffener is not adequate.

I think that concern is valid, so add the kicker.

 

[WillisV]

It’s called constrained axis torsional buckling.

This limit state is described fairly well in AISC’s Seismic Design Manual.

The following two papers address the calculations involved (with tables) as well as (in the second paper) the utilization of the deck to assist further with preventing this limit state.

https://www.aisc.org/Torsional-and-...ling-Tables-for-Steel-W-Shapes-in-Compression

https://www.aisc.org/Continuous-Bracing-Requirements-for-Constrained-Axis-Torsional-Buckling

 

[JLNJ]

If this happens just a few places, you are going to spend more money trying to find an exact solution than you would spend just making the beam more stout. A costly design effort might still result in a heavier beam.

Kickers to joists for stability makes me nervous as we were involved in a failure of a big PEMB girder which was “braced” to roof joists. Do this as a last resort.

I would consider making the beam heavier, even if you end up considering all the compression force in the top flange only. It probably gets dragged out of the roof diaphragm anyway.

 

[JoshPlumSE]

Constrained Axis Torsional buckling (CAT Buckling as I call it) is also well described in the AISC design guide on tapered members.

I think a kicker to braced the bottom flange can work fine. However, there are some caveats. If you're kicking back to the joist, does it have sufficient stiffness and strength in that direction to really brace the member?

I might prefer a full depth member (or nearly so) between this beam and the joist which also continuously connects to the deck. Doing the same thing as a kicker (bracing back to the joist), but also distributing the load into the deck along the way.