Steel beam lateral torsional buckling 4

[Gus14]

In the pictrue attached.
1. Do the floor beams provide enough lateral torsional buckling resistance to the girders to be assumed continuously braced?

2. How to provide lateral buckling resistance at the red mark without assigning a column ? what type of a beam and connection is required ?

3. Any red flags to check if an all welded double angle connection is assaigned at the blue circle?

4. Any refrences to design continuous beam over column connection (cap plate) ?

 

[r13]

1. From the layout, it seems the girder can be considered adequately braced, thought the floor beams should be better arranged to center in columns.
2. Is there a concentrated load? Otherwise, why needs brace?
3. It should be fine, as long as the girder is aligned with the floor beam, and the hosting beam is adequate to take the loads.
4. I think AISC has many examples, unless there is particular concerns.

 

[Gus14]

Thanks for reply,
Thankfully I dont have to brace it. I'm just curious about the process of bracing a location generally and the reasons for it. And I can't find an example on Cap plate design in the AISC design examples (is the connection called something else?)

 

[r13]

1. The bracing is required for lateral torsional buckling concerns. As the unbraced length increases, the permissible beam strength decreases. Suggest to review the code provisions concerning "lateral torsional buckling", and be familiar with it.
2. Try AISC Manual Vol. II - Connection, Link

 

[Enhineyero]

Hi Gus, here are my thoughts.
1. the girders looks laterally restrained (un-restrained length = to joist spacing, unless you have a floor slab), I'm not sure if its torsionally restrained can't see the cross-section (you have to provide torsional restraint to the compression flange, i.e. fly brace, full depth cleat).
2. locate a joist on the red mark to restrain it laterally
3. Check welds against eccentric loads, check beam web failure including any eccentric load effects (they are usually thinner than the angle so they fail first).
4. Sorry I don't think there is one, you have to check this using first principles.

 

[human909]

I must be looking at a different picture to everybody else. I don't see any mid girder lateral restraints.

You can't provide LATERAL restraint to one girder simply by linking it laterally to another girder without lateral restraint. Otherwise you have a fairly recursive situation. Such member might be able to provide torsional restraint on both girders but it isn't as straight forward as saying that the beam is restrained at each joist.

 

[r13]

The adequacy of bracing is always subject to debate. In this case, I think the conditions indicated below are met, thus is considered fully braced.

The designer may use one or both of two general options to provide a beam brace: (1) brace the flange subject to compression directly or (2) prevent twist of the cross section. A direct brace may be provided for a primary member by a properly attached floor system itself or by a secondary framing member. Generally, a brace connection, such as the simple shear connection for an infill beam, that is located within the one-third depth of the beam web that is closest to the compression flange can be considered to provide a direct brace. If this is not the case, transverse stiffening can be provided to prevent twist and transfer the bracing effectiveness from a properly attached floor system to the compression flange.

A long-standing rule of thumb is to provide bracing for two percent of the compressive force in the flange or member being braced. Although it lacks an explicit consideration of the required bracing stiffness, this approximation is typically conservative. Note that the two percent rule applies only to compression members that are considered straight within ASTM tolerances. AISC Specification, Appendix 6, addresses requirements for stability bracing of beams and columns.

 

[phamENG]

Gus, I'll add my thoughts to the mix:

1) No. I would not consider them continuously braced by secondary framing/joists/infill beams/<insert your regional terminology here>. They may be adequately braced to preclude LTB, but they are not "continuous." In terms of whether or not they provide enough LTB restraint, you have to look at the strength and stiffness of the girder in compression/tension. Where are you practicing? You mentioned AISC - if you're using AISC, you can look in the appendices for stability bracing of columns and beams. It lays out a procedure for determining the minimum requirements for a brace point. I think you'd be in the relative bracing category with this one - not nodal.

2) Not sure why this is a concern? Doesn't look like you have a beam connecting there. Can you elaborate a little more on this location and why you're worried about it?

3) All welded double angle should fine, though shop welded/bolted or bolted/bolted is likely to be less expensive.

4) AISC has a connection example like this for a wide flange beam over HSS column to use as a moment connection (though lots of people don't like it). You can adapt it to your case - the checks for the beam will be the same. Checks on the column will be a little different, especially where you're crossing perpendicular to the weak axis. Those should not be considered moment connections unless your column aspect ratio is approaching 1 and you use a REALLY thick plate. Even then it may be questionable.

 

[human909]

The adequacy of bracing is always subject to debate. In this case, I think the conditions indicated below are met, thus is considered fully braced.

Subject to debate doesn't get us anywhere. Lets debate it if you disagree!

And onto that. Your quoted paragraph doesn't support your contention. To reiterate. In order for something to be laterally braced it must be stopped from moving laterally. The you an unrestrained flange cant brace another unrestrained flange!!

 

[Gus14]

Thank you everyone for feedback.
Apparantly the contractor wants the floor beams to rest on top of the girders (like the attached pictrue) this means that they won't provide any lateral support. I have to add lateral bracing but the resources mentioned are very helpful.

Also on site columns are aligned about the strong axis. I had to draw the sketch quickly.

For future projects what is the name of a connection other than "cap plate" that serves the same function ( continuous beam ) ?

 

[KootK]

Are your infill members here joists or beams? I'm having a tough time parsing that out based on the terminology and sketches so far.

Apparantly the contractor wants the floor beams to rest on top of the girders (like the attached pictrue) this means that they won't provide any lateral support.

They will provide competent lateral support to the top flange of the supporting beams. You'll likely want to run kickers down to brace the bottom flange in the zones of negative, hogging moment.

 

[KootK]

Ok so beams will provide lateral support with kickers. But joists won't right?

Not quite.

1) Beams will provide top flange lateral AND rotational LTB restraint WITHOUT kickers so long as the depth of the supported beam and it's connection are reasonably deep relative to the depth of the supporting member. AISC has a specific requirement for this. Something the tune of 50% of the girder "T" dimension I think. Do verify that.

2) Joists will provide top flange lateral LTB restraint WITHOUT kickers.

3) Joists will provide top flange lateral AND rotational rotational LTB restraint WITH kickers.

Note that these two conditions are equivalent for LTB:

a) Top flange lateral restraint + Bottom flange lateral restraint and;

b) Rotational restraint + Either flange lateral restraint.

Both [a] and will effectively brace your bottom flange in negative moment, hogging regions.

 

[Gus14]

By floor beams I mean floor joists. They are not beams and they are closely spaced. Sorry for any confusion. I will look into running kickers. Again thanks for feedback.

 

[r13]

Human909,

I remember you know how to, and can perform buckling analysis, similar to that one in the discussion on the "twin girder" thread a while ago. Why not do another one to see the buckling mode of this system, by assuming all floor beams are simply connected to the end girders without topping floor. Appreciate if you don't mind.

 

[r13]

human909,

Also you might want to review this paper, or maybe you have read it many times. Link

In the paper, Yura has identified lateral translation and twisting are the two modes of buckling concerns, but the twisting is the most significant of the two, and is used to develop lateral torsional buckling modification factor C[sub]b[/sub], which in turn is required to determine the elastic buckling moment with respect to the unbraced length. In this paper, he also briefly addressed restraint of decking, though I think it is not considered in the development of those formulations.

 

[human909]

2) Joists will provide top flange lateral LTB restraint WITHOUT kickers.

What provides the lateral resistance? What is the load path to a point of restraint?

I remember you know how to, and can perform buckling analysis, similar to that one in the discussion on the "twin girder" thread a while ago. Why not do another one to see the buckling mode of this system, by assuming all floor beams are simply connected to the end girders without topping floor. Appreciate if you don't mind.

This cannot be readily done without consideration of connection stiffness between the joists and the beams. Throw in connection stiffness and a you get torsional restraint and buckling is reduced. (though usually effectively length will still be greater than joist spacing).

My point has been NOT that joists don't reduce LTB (if there is some rotational stiffness at the connection). My point is that unrestrained joists don't provide LATERAL restraint.

 

[KootK]

What provides the lateral resistance? What is the load path to a point of restraint?

The low aspect ratio floor diaphragm.

Load path = flange --> joist top chord --> deck fasteners --> deck diaphragm (probably concrete) --> diaphragm collectors --> vertical load resisting system --> foundations --> mother earth.

If this setup can't be relied upon to laterally brace a flange, we might as well just forget about the concept altogether.

My point has been NOT that joists don't reduce LTB (if there is some rotational stiffness at the connection).

Joists will help to restrain LTB even with no rotational stiffness at the joint at all so long as the joist are attached to some manner of diaphragm that restrains axial joist translation. Luckily, the overwhelming majority of joists are so restrained laterally.

 

[human909]

Thanks Kootk. For stepping me through that and making the differences my assumptions clear. I agree with your statements above.

I was operating on the assumption there was no viable deck diaphragm. My reason was that none was mentioned and if there was a deck diaphragm then the solution is trivial. (I do a quite a few gantry structures with grating which I certainly don't rely upon for diagram action.)

 

[KootK]

You're most welcome human909. It is certainly true that, in the industrial arena, you see a lot less reliance on deck diaphragms and a lot more reliance on discrete, horizontal trussing. And the discrete trussing certainly has a visceral, mechanical appeal.

The products that would typically be called "open webbed steel joists" in North America almost have to have competent diaphragms. The chords are usually so slender that they'd be worthless without either a capable diaphragm or 1/8th point cross bridging.