Torsional Bracing Requirement in 13th edition 3

[3doorsdwn]

The 13th edition now addresses bracing strength (and stiffness) for points along columns and beams. For beam bracing, criteria for lateral bracing and torsional bracing are given (as opposed to using some of the old rules of thumb (i.e. 2% of the compressive force, etc.)). My question is: If my section (a doubly symmetric I-beam) cannot fail by flexural-torsional buckling (only by flexural buckling), does torsional bracing need to be considered? Or just lateral restraint for the compression flange?

Most I-beams (in the steel manual) are controlled by flexural buckling (at the typical lengths used in design). So I was uncertain if they meant this criteria primarily for singly symmetric shapes, or was it intended to take care of initial twist imperfections in the beam.

Any thoughts are appreciated.

 

[271828]

I don't know what you mean by "flexural buckling" and "flexural-torsional buckling."

Flexural-torsional buckling is something that happens to singly-symm columns.

If you're referring to lateral-torsional buckling, then Appendix 6 indicates that you can consider a location along your beam braced if the compression flange can't translate OR if the entire section at that location can't twist.

 

[271828]

Oh yeah, forgot to type this. Flexural buckling is a column limit state also.

Sorry to nit-pick your terminology. Considering that App 6 has column and beam stuff, it needs to be clear which we're talking about.

 

[3doorsdwn]

I am referring to Section 6.3.2 (in the appendix of the 13 edition).

I guess what is confusing me is: what torsion am I bracing it against? Initial imperfections? Flexural torsional buckling? (Not a mode of failure for doubly symetric I-beams in the manual, except at short lengths.) Torsion caused by external loading (not likely)?

 

[271828]

6.3.2 is for torsional bracing of a steel beam against lateral-torsional buckling.

Nothing special, just a regular unbraced steel beam subjected to flexure only. Doesn't matter if it's doubly-symm--the section still applies. Actually, the equations were derived for doubly-symm beams, so try applying them to mono-symm ones if you want to really have fun, LOL.

You can say your beam is braced at a point if that point is restrained from twisting or if the compression flange is restrained from translation. 6.3.2 is for the former--restraint against twist.

A beam doesn't have to have any initial imperfection to buckle. Neither does a column.

Real members do have initial imperfections, though, and those go into the derivations of the equations in Appendix 6.

If you want to say your beam is braced against twist at a section, then you have to have something bracing it there, whether it be another beam framing in, kickers, or whatever. This section just gives you a way to calculate how strong and stiff the bracing element(s) must be.

This is, by far, the most confusing set of provisions in the AISC Spec. The deeper you look into it, the less sense it makes also. If you want to research the background, look at the Fundamentals of Beam Bracing paper by Yura. There are various seminar notes floating around out there too.

Warning, though: Yura & Helwig have made this stuff so confusing that many people, steel gurus included, pretty much throw up their hands when trying to dig into the background.

 

[3doorsdwn]

Funny you mention that (about Yura), because I have attended his seminar (a few years back), and when I look at the notes from it, I was getting somewhat confused.

What I suspected from looking at this spec was that it applied mainly to flexural-torsional buckling, thanks for confirming that for me.

The thing about it is: typically, I am dealing with a floor system (consisting of nothing but girders and intermediate beams) that does not have much of a restraint for the top flange (i.e. grating). I’ve always made it a point to tie my beam lines back to a lateral force resisting system (some times using in-plane bracing or a concrete slab) to give it the necessary stiffness.

Anyway, thanks again.

 

[271828]

Cool. No problem.

Hate to nit-pick again, but flexural-torsional buckling is something that happens to columns. So is flexural buckling.

In this case, we have lateral-torsional buckling.

 

[Bagman2524]

Just from prelinary reading of AISC 360-05 it seems a lot more technical and complicated than the 89 ASD that I'm used to. I'm not sure if this is a good thing or not. Is trying to trying to get steel design as close to 100% efficient as possible worth making the design much more complex and thus increasing design time and possibly introducing error?

 

[3doorsdwn]

That’s an excellent point Bagman: we don’t have time to turn projects into research projects. However if you want to consider a point braced, it must meet this criteria 9for strength AND stiffness).

What prompted this question was the language of section 6.3 (p. 16.1-193):

“At points of support for beams, girders and trusses, restrain against rotation about their longitudinal axis shall be provided. Beam bracing shall prevent the relative displacement of the top and bottom flanges, in other words, the twist of the section. Lateral stability of beams shall be provided by lateral bracing, torsional bracing or a combination of the two.”

The way that read, I was not sure if the first sentence meant that torsional bracing was MANDATORY. (And bracing girders for torsion is difficult where you just have beams with simple shear connections going from girder to girder.) And the last sentence was somewhat ambiguous (at least to me) as well. I was not sure if they meant that lateral bracing could be provided by either one of those methods or not. The code is a legal document; so interpretation of it has to be taken with care.

 

[271828]

The 2005 AISC Spec. wasn't made more complicated in an attempt to increase efficiency. It reflects 16 years worth of research since 1989 and is directly applicable to many more situations than the 89 Spec.

For example, if you're designing a singly-symm I-shaped beam using the green book, then you use equations that were derived specifically for doubly-symm shapes. The 2005 Spec. has specific sections for singly-symm I-shapes. For those with the need and/or inclination to dig in and use the most modern info, the 2005 Spec. is greatly preferable.

Also, I really don't think it's as bad as it looks. Take Chapter F for example. Sure it looks horrible, but if one only deals with compact W-shapes, then F1 & F2 are the only applicable sections. Compare the simple format of F2 to the spaghetti-like format of the green book's Section F1!!

 

[271828]

Focus on your last quoted sentence and you know what to do.

"Lateral stability of beams shall be provided by lateral bracing, torsional bracing or a combination of the two.”

You can read the Fundamentals of Beam Bracing Paper or seminar notes to get further backup on this. There are lots of examples in teh seminar notes, some that show lateral bracing only and some that show torsional bracing only.

 

[3doorsdwn]

Right. The thing about that sentence (i.e. "Lateral stability of beams shall be provided by lateral bracing, torsional bracing or a combination of the two.”), is it didn’t say: ‘shall be provided by EITHER lateral bracing, torsional bracing, or a combination of the two’. So that was kind of throwing me off.

I attended Yura’s seminar on this (years ago) and looking back at the notes, it was clear HOW to provide each type of bracing, just not which one is supposed to be used.