Steel beam and metal deck unbraced length 4

[Elindio1]

I have a concrete slab supported on a secondary steel beam. The structural drawings that I have are not the final construction drawings. They do not show any studs or how the metal deck is attached to the steel beam. Now, as per " Fundamentals of Beam Bracing" by Joseph Yura "a cross member merely resting (not positevely attached) on the top flange can significantly increase the lateral buckling capacity. The restoring solution is sensitive to the initial shape of the cross section and location of the load point on the flange. Because of these difficulties, it is recommended that the restoring effect not be considered in design". My question is the following, Can deck puddle welds be assumed as the possitively attachment points and by doing so, can I develop the continuos lateral support of the beam to develope the full capacity of the section?

 

[271828]

Yes, in typical situations, puddle welds will be plenty to laterally brace the top flange. Because you typed "secondary beam," I assume that the deck spans perpenticular to the beam.

Without going back and finding that paragraph by Yura, I think he's talking about a situation like a conc slab just simply sitting on the steel beam top flange. In that case, if the beam starts to twist, the slab will be pushing down on the flange tip that's going up, providing a torsional restraint of sorts. Note that you only have to restrain twist OR lateral movement of the compression flange. I don't think that's what you have.

 

[Elindio1]

Exactly that is my point, We cannot assumne that the slab will be providing a torsional restraint of some sort. This effect it is recommended by the Yura paper not be considered in the design.
I think you are correct. I'll have to consider the puddle weld as providing lateral restraint to the top flange. Thanks!!

 

[271828]

No problem. I think your puddle welds and concrete slab will provide a lot more than what AISC 13th Ed. Appendix 6 requires for a beam continuous lateral brace force.

The only time this makes me nervous is if I have a roof deck only and am trying to brace a girder. With the deck flutes running parallel to the girder, it's not as obviously providing a lateral brace.

 

[Lion06]

271828-
Would you ever consider deck with the flutes parallel to the beam to brace the beam? I have only used the beams that the deck is spanning to as bracing the girder. Assuming it is deck only as you say above.

 

[271828]

StrlEIT, I almost never say never, LOL, but I can't think of a time right off the top of my head. Beams usually frame into the girder close enough to make it irrelevant anyway--as you alluded to.

OTOH, it might be completely irrational to not use the deck. The brace force and stiffness is very tiny for a continuous top flange brace. It would take some effort to quantify the resistance, so I'd probably never try to prove that it's adequate.

 

[haynewp]

I was told of a collapse that happened because an engineer assumed that the deck running parallel braced a composite floor beam. It was left as a parameter set in a popular software package (here to be left unnamed) so it buckled like an accordian prior to the slab hardening; it actually happened during slab placement while the concrete was wet. But I agree that I don't think it would typically be a problem with beams framing into a girder, only for a less than usual case.

The gist is that I would not assume that it braces the beam.

 

[WillisV]

The only failures I am aware of occured when the contractor did not actually connect the deck to the girder prior to placing the concrete (no puddle welds or studs at all...just deck bearing on the girder). If the deck is properly attached with puddle welds, studs, etc. then even deck running parallel to the beams has been shown to provide adequate stiffness for bracing in typical conditions.

 

[haynewp]

WillisV, do you have a reference I can look at for that?

 

[WillisV]

haynewp,

Nethercot did a lot of work on this in the 70's, however the most up to date paper I am aware of is by Helwig and Frank entitled "Stiffness Requirements for Diaphragm Bracing of Beams" which was published in the Journal of Structural Engineering, November 1999. It is a well written and informative piece.

What you have to keep in mind is that it is the DIAPHRAGM action of the deck that is bracing the top flange, not the actual bending of the deck (though this does contribute some it is actually negligable compared to the diaphragm stiffness component). For the beam to fail in a lateral torsional buckling mode, the top flange has to translate laterally. The diaphragm stiffness resists this lateral translation. Diaphragm stiffness (the G' term in deck catalogs) is determined irrespective of the direction of the deck span. It is based on fastening schedules and length of deck span.

The basic finding of the paper is that the shear modulus (G') required for a deck to brace a beam for an applied moment Mu is equal to:

G'required = (Mu-Mg)/[(tributary width of deck)(0.375)(d)]

Where Mu is the applied load, Mg is the unbraced moment strength of the beam, tributary width of deck is the beam spacing, 0.375 accounts for top flange loading, and d is the depth of the beam. Make sure all your units work out.

What I have found is that the G' required to develop the full plastic moment strength of beams in most general cases is easily provided by typical 2 and 3" composite deck profiles - again irrespective of flute direction.

 

[271828]

Excellent Willis. Thank you very much--I'll have to download that paper.

 

[haynewp]

Willis,

I just read through some of that paper and I am just wondering if it is assuming that locally there is not going to be a problem:

"The metal forms can provide lateral bracing to the top flange of the beams or girders on which they are fastened due to the forms large in-plane shear stiffness."

Another example is how many engineers provide cross bracing adjacent to cmu walls when the deck is running parallel to the wall, instead of just relying on the deck as bracing. Althought there is a tremendous diaphragm shear capacity in that direction, I am wondering if locally it has to be able to get into the diaphragm first.

I emailed the engineer that told me of the collapse I mentioned for more details. He is also a professor that owned his own company for 30 years. Perhaps I misunderstood something when he told me about that collapse.

 

[haynewp]

Also, is the FE modeling in that paper based on the flutes perpendicular to the beams?

"Beam elements with only out-of-plane stiffness
were used along the edges of each panel to prevent local diaphragm distortions. The out-of-plane moment of inertia of the stiffening beams was of the same order of magnitude as the
corrugations in the actual deck panels."

 

[WillisV]

haynewp,

1. I'm not sure of your thought here. The sentence you reference summarizes the whole paper.

2. Engineers do a lot of things that might not technically be necessary but are considered good practice. I would probably put in bracing in that situation too. I will be ineterested to here more about the failure you are referring to.

3. I do not believe the FE modeling was based on flutes perpendicular to the beams - it was meant to be general and applicable to different deck profiles irrespective of direction. They used essentially just a fake flat surface and varied values of "E" to provide G' values similiar to those from actual deck profiles (read the paragraph after the one you referenced regarding the beam elements). Those little beam elements were just put in to prevent this flat surface from erroniously experiencing plate buckling. They chose to use a moment of inertia equal to "the same order of magnitude" as the corrugations just to have something within the ballpark to prevent this plate buckling. It is my understanding that these beams were placed around all four sides of each FE panel.

 

[haynewp]

My thought is that it may be addressing the entire diaphragm resistance and not focusing very much on local effects. I am not trying to prove you wrong here, I only know of the single collapse I referenced earlier that I am trying to get more info on. It may be that there is plenty of immediate stiffness in that direction of the deck to brace the beam, but looking at it in a common sense way, locally it would seem that it cannot be as stiff as as if the flutes were perpendicular to beam flange.

 

[haynewp]

nevermind, I think the paper has to be considering local effects as well if it is relying on the entire diaphragm as bracing.

 

[haynewp]

Alright I am back. The reply from my friend (this is a highly respected guy who has done a lot of research with AISC and is listed at the front of the LRFD manual) who told me of the collapse is below. He is completely aware of the diaphragm action and the research that has been done by Nethercot.

"You are correct. Deck parallel to a girder flange does not brace the girder flange. Take the brace force required of the girder flange, 4% of flange force, and apply it as a load to the edge of the deck 20 - 22 ga deck and see that the deck cannot possibly take it in compression"

 

[271828]

Sounds like you posed the question as if it's a spandrel beam. That would be a lot weaker than an interior case because the deck could just buckle if the beam flange tries to translate toward the interior. If the beam was interior, then one side will be going into tension.

The required brace force should be spread over a fair portion of the beam, so might not be all that much.

It would be an interesting 3-4 hour task to create a beam and portion of deck using shell elements, give the beam an initial out-of-plumbness so it will try to roll, and then apply load.

 

[haynewp]

To me the point is whether or not you can rely on parallel deck bracing the beam.

 

[271828]

Right, but if you pose the question like:

I have a spandrel beam parallel to the deck flutes, can the deck brace the top flange of the beam?

In that case, if the top flange tries to buckle toward the inside of the bldg, then the deck goes into compression. You or I could probably put on a pair of gloves and manually bend/buckle the deck in this condition. It would be a lot stiffer and stronger at an interior beam.

BTW, I'm inclined to not use it also, but I doubt I can prove that it's always wrong to use it.