Beam Bracing per AISC 360-16 Appendix 6 1

[StrEng007]

AISC's Appendix 6, Section 6.3, the commentary says this:

Beam bracing must control twist of the section, but need not prevent lateral displacement.
Both lateral bracing, such as a steel joists attached to the compression
flange of a simply supported beam, and torsional bracing, such as a cross-frame or
vertical diaphragm element between adjacent girders, can be used to control twist.

If we taken a W-beam, how does the beam experience lateral only buckling? I understand how the LTB occurs, but I'm not sure why this section of the appendix separates the two. I guess the question is, how does a flexural member experience lateral only buckling? Wouldn't any form of lateral buckling instigate LTB?

Another question, why is panel bracing defined by Vbr, while point bracing is Pbr? For any brace, I see the force that enters the brace as an axial force. So why does panel bracing use the term shear?

 

[StrEng007]

The following text is from "Buckling of Beams with Inflection Points" by Yura.

In the statement, " Lateral bracing prevents lateral movement at the point on the cross section where the brace is attached. When applied to only one flange in a beam with inflection points, lateral bracing will not prevent lateral-torsional buckling, but the buckling capacity will be improved."

Does this mean in beams WIHTOUT inflection points, the lateral brace WILL prevent LTB?
AISC Section 6.3 allows for the use of lateral bracing, but based on the appendix statement above, this lateral bracing does not actually prevent twisting of the cross section.

 

[JoshPlumSE]

I think StrEng007's post answers this much more completely than I could. But, I thought I would add the more simplistic way that I view that appendix commentary:
1) Beam bracing must control the twist of the section.
2) Torsional bracing will control the twist of the section.
3) Lateral bracing of the compression flange will often prevent twist as well. Especially if that compression flange is connected to a diaphragm.
4) If the top compression flange is ONLY braced against deflection (like a bottom flange), then the flange could still twist torsionally in LTB. That's why a diaphragm (which generally prevents the flange from twisting) improves the restraint.

 

[Smoulder]

"Does this mean in beams WIHTOUT inflection points, the lateral brace WILL prevent LTB?"

It better because that's how we design in Australia. And UK I guess because our rules probably came from them. The reason is because point of twisting is the bottom flange for s/s beams so top flange brace is enough to keep it all straight. On condition there's proper lateral and twist bracing each end. Which is why it doesn't work for cantilevers with tip just flapping around.

 

[JoshPlumSE]

"Does this mean in beams WIHTOUT inflection points, the lateral brace WILL prevent LTB?"

Not necessarily.... There are a lot of caveats. But, it usually does (IMO).

Bracing the top flange will prevent lateral deflection, but not twist. I believe is there is a full depth shear plate it will prevent twist. Not sure about partial depth. Especially if it does NOT connect to the top flange. Though I'd have to review references before I said this definitively. Also, if you have a diaphragm on the compression flange, then it will likely prevent twist as well.

 

[human909]

Not necessarily.... There are a lot of caveats.

For 'regular' hot rolled sections I can't think of any caveats for a beam restrained avast twist at both ends.

You don't need stiffeners to restrain the bottom tension flange.

As mention by smoulder and in the first discussion, AS4100 relies on this fact.

 

[TLHS]

The lateral torsional mechanism happens because of the combination of lateral displacement with torsion. The lateral movement increases the torsional moment arm and creates a feedback loop. The strength failure tends to come from the torsion, but without the lateral movement the torsion doesn't increase in an unstable way. The section needs to be free to twist *and* move for for the classical LTB mechanism to happen.

 

[dik]

It gets even more interesting when you include the location of the load relative to the shear centre. I had a long cantilevered monorail hoist a few years ago, to lower large equipment down a mine shaft, that I used the Eurocode approach for this loading.

-----*****-----
So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[human909]

Dik. Do you remember which code reference this was. When I get a chance I'd like to have a look.

 

[StrEng007]

All very interesting. This is the first time I actually noticed the text in the appendix (shown in my OP). I have always taken Lb as what is defined in the manual.

"Lb = length between points that are either braced against lateral displacement of the compression flange or braced against twist of the cross section".

So for simple spans, I always took this as any perpendicular beam that attached near the top flange, any diaphragm attachment along to top of beam, or any specific brace to prevent twisting. What really threw me off here was how the appendix implies that you don't need to brace for lateral displacement... but lateral bracing prevents twist of the overall cross section. So basically it's not required, but it's a direct solution to prevent twisting.

Maybe another way to ask my original question is, when would someone use lateral bracing of simple span flexural member to prevent something other than LTB. As in, can you have lateral displacement of a W-beam without LTB?

 

[human909]

"As in, can you have lateral displacement of a W-beam without LTB?"

Maybe I'm misinterpreting thing but it seems that the obvious answer is a simple YES.

Any lateral load on a beam (or any member) can and will produce lateral deflection without LTB.

A simple gantry without a rigid deck subject to cross wind load may be subject to excessive deflection if it isn't suitably restrained with in plan lateral bracing.

 

[StrEng007]

human,
I meant in terms of flexure where bending is about the strong axis of the beam. I was referring to that 'loop' where buckling of the compression flange in strong-axis flexure creates lateral displacement of that flange, which is then restrained by the tension flange, which then creates that lateral torsional buckling. I think the answer to that is NO, lateral ONLY displacement will not occur unless an independent lateral load acts on the beam as you mentioned. For any other scenario with a W-beam, most cases of the initial compression flange lateral displacement due to flexure will result in twisting of the cross section.

Thank you and everyone else for your feedback.

Are you familiar with the brace force requirements? Why is panel bracing defined by Vbr, while point bracing is Pbr? For any brace, I see the force that enters the brace as an axial force. So why does panel bracing use the term shear?

 

[Lomarandil]

Human909, there are a number of alternative formulations (mostly not codified, but accepted in the engineering literature) of Cb modifications for cases where the load is applied at a height above or below the shear center (e.g. Dik's monorail) or the support is applied at a height above the bottom flange (e.g. suspended beams).

SSRCs Guide for Stability Design Criteria of Metal Structures and the research published by Yura and Helwig are the sources that come to mind.

 

[human909]

I'm aware of the effects of loads above below or even to the side of the shear centre. But unless we start getting to extremes I'd argue my assertion stands.

This assertion based off and supported by;
-my own rational inquiry
-AS4100 treatment of LTB

I can't certainly conjure up odd member geometries or odd loading positions that would make LTB highly likely unless the member is extremely well restrained. But for 'normal' geometries and loadings restraining of the compression flange is sufficient for non cantilevers.

 

[Lomarandil]

Oh, I jumped into the discussion midway and misunderstood your query. Disregard.

 

[KootK]

I guess the question is, how does a flexural member experience lateral only buckling? Wouldn't any form of lateral buckling instigate LTB?

There's no such thing as lateral only buckling in this context. A purely lateral motion allows a vertical load to do no additional work. Ergo the load has not shed potential energy and buckling has not occurred.

Beam bracing must control twist of the section, but need not prevent lateral displacement.

What AISC is getting at, and doing a poor job of it, is referring to the bracing condition shown below where the beam can translate but not rotate. Because it can't rotate, the movement of the beam is constrained to pure lateral translation which, per the above, cannot constitute buckling. The most common case of this is where the LTB bracing is provided by an incoming roll beam that is, itself, not braced against lateral translation by a diaphragm etc.

 

[KootK]

The reason is because point of twisting is the bottom flange for s/s beams so top flange brace is enough to keep it all straight.

I believe that statement to be slightly in error. Unbraced, simple span beams will twist about a point in space that may be located anywhere between the beam shear center and an infinite distance below the shear center. And that may be well below the bottom flange. Beams that have a high Ix:Iy ratio will LTB buckle in a pattern that is very nearly pure lateral displacement. This can happen with large bridge girders. In such scenarios, the center of twist is way, way below the bottom flange.[/quote]

 

[StrEng007]

Thank you all.

Koot,
Your quick detail should be pasted into the text of Appendix 6. I hope to see that exact sketch in the next edition.

 

[KootK]

Your quick detail should be pasted into the text of Appendix 6. I hope to see that exact sketch in the next edition.

I'll send a strongly worded email... as one does.