Yet another LTB question 3

[bpstruct]

Don't want to get into theory...very specific question here. Have a metal building with purlins on top of main frames- presumably only minimally connected to the top flange of the main frame beams at about 4 feet oc. Thinking that the unbraced length for positive flexure (top flange in compression) would be 4 feet. If I were to weld plates to the webs at each purlin location (connecting bottom flange to top flange), would the unbraced length for negative flexure at the columns and at center for uplift match the unbraced length for positive flexure? I'm thinking no, since translation/rotation of that bottom flange would be dependent on the connection between the purlin and the top flange of the beam.

 

[BAretired]

I agree with you. The bottom flange would not be braced.

BA

 

[JAE]

I agree with BAretired.

 

[hokie66]

It depends. If the purlins are connected to cleats with 2 bolts and there are web stiffeners, then it has been shown for some sizes of rafters that this type connection provides adequate bracing. I will try to find the references and post later. If the purlins are "only minimally connected", then I would agree that there is no chance that the bottom flange can be considered braced.

 

[JAE]

...or if the purlins are small, 8" deep light gage tinfoil trying to brace a huge, long span steel girder then the connection would be meaningless.

 

[hokie66]

This has been discussed a few times previously, including in this thread:

thread507-321425

I will try to find the research report which I remember.

 

[dik]

...be appreciated Hokie...

Dik

 

[amec2004]

From this AISC Engineering FAQs

http://www.aisc.org/DynamicQuestion.aspx?Grpid=6&QueId=1646&ste=_Preview&id=2100&type=0

and Yura's AISC Engineering Journal Fundamentals of Beam Bracing

We can summarize that

1. Lateral brace can be provided through one or two of the options
1) Brace the flange subject to compression directly OR
2) Prevent twist of the cross section.

2. Secondary beam pin connecting to main beam locating within the one-third depth of the beam web that is closest to the compression flange can be considered to provide a direct brace.

[RED]3. Transverse stiffening can be provided to prevent twist and transfer the bracing effectiveness from a properly attached floor system to the compression flange[/RED]

Please refer to attached PDF file for details

anchor bolt design per ACI 318-11 crane beam design

http://www.civilbay.com

 

[hokie66]

Thanks for that, amec. I still haven't found the report which I was talking about, but will look some more.

It is stated or implied in the reference which amec gave that bracing needs to be on the loaded flange, e.g. top or tension flange in a cantilever with gravity loading. It would follow that for uplift of a roof beam which is loaded on the top flange, the purlins would provide bracing, as they are connected to the tension flange. The principle is that tension flange loading acts to self correct any tendency of the opposite flange to translate. Comments?

 

[BAretired]

I agree that the position of the load relative to the centroid of the beam is a major factor in LTB and is not usually taken into account by engineers. A continuous roof or floor beam is normally loaded on the top, so the compression flange becomes critical in the vicinity of the columns.

If gravity load were suspended from the beam, lateral torsional buckling would occur at a higher load, but it could still occur. "Theory of Elastic Stability" by Timoshenko and Gere takes this into account.

BA

 

[bpstruct]

Thanks to all for your comments.