Perimeter Beam Bracing 1

[KootK]

This should be an easy one. Here's the situation:

1) Perimeter wide flange beam running parallel to steel deck flutes.
2) Steel deck is topped with concrete (typical composite Canam / Vulcraft set-up).
3) Steel beam is not composite with concrete topping (no nelson studs, only puddle welds).

Is the top flange of this beam continually braced?

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[HuckleberryFinn]

I'll bite. Sure, unless the beam is massively big with a huge required bracing force - The cast in slab braces the top flange due to gravity load. 9/10 it should be fairly easy to develop the required bracing force for the flange (~.02Mn) - even when using unconfined compression elements in the slab (using less than 0.35F'c of the slab).

-Huck

 

[CELinOttawa]

Nope. You need a positive attachment with at least minor rotation-resistent capacity. It may behave as stable and braced, but is not reliably so at Ultimate. Even for a lean-on style bracing solution you still require positive attachment.

Not to say these don't work, they just aren't safe enough and don't meet minimum practice standards. Never seen one fail, but I'm not interested in stamping one or otherwise trying my luck.

 

[JAE]

It is braced in my view. The puddle welds or deck screws can be checked if you are unsure - use AISC's beam stability section to verify strength of the deck-to-beam connections as well as stiffness (which I'm sure the deck is stiff enough). The main question is whether the lateral bracing force required is more than the deck connectors.



Check out Eng-Tips Forum's Policies here:
faq731-376

 

[KootK]

This is usually a moot point, of course, as such beams are often sufficiently braced by perpendicular framing members.

@Huck: I suspect that gravity / friction would usually prevent LTB. I haven't relied on that in the past however.

@CEL: I agree, I usually do not consider this condition braced. I'm curious about your requirement for rotation resistance. Why do you say that? I actually agree but I know of no code specified torsional restraint requirement. I've never understood why the tension chord of a simple span truss can buckle but the tension flange of a wide flange beam cannot. Also, would you change your answer if the beam in question was an interior beam rather than a perimeter one?

@JAE: based on your response, I think that you might be envisioning the deck flutes running in the wrong direction.

I see two scenarios:

1) If the beam tries to buckle toward the building exterior, the deck will engage the topping and then it is a matter of how well that topping edge is tied back to the rest of the slab.

2) If the beam tries to buckle toward the building interior, the deck may pull away from the topping and crinkle up accordion style.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[mtu1972]

I typically took them as braced. It may well be a moot point in that something has to support the deck perpendicular to the beam in question and those beams/purlins/joists would provide the lateral bracing at their spacing(s). These are generally close enough to provide the bracing such that the full capacity of the edge beam can be achieved.

gjc

 

[hokie66]

I agree with JAE, if the deck is a composite type, and filled with concrete, even if the deck runs parallel with the beam. The only consideration is the connection of the deck to the beam.

There may be additional detailing requirements for seismic of which I am unaware.

 

[KootK]

@Hokie / JAE (maybe): The only thing keeping the deck from going "accordion" is the ability of the deck deformations on the sides of the flutes to engage the adjacent concrete in tension, right? If so, how on earth would one come to know that capacity? Let me know if my use of accordian doesn't make any sense. It may be sketch time.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[KootK]

Oh, application is a very small, 10' girder that only has a heavily loaded transverse framing member coming in at midspan. It would be an academic problem if it didn't actually exist.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[hokie66]

Then why are you concerned? Short member, braced at midspan. But I still think the deck will brace it, provided it is connected well enough. Sometimes, I think we get too picky about what constitutes bracing, when the true force required is very small.

 

[CELinOttawa]

Yup, now I'm with Hokie... You have a nodal brace at mid span on a 10' span. Not likely to fail, particularly with the minor attachment to a slab element along the length. At that point I'd not be worrying about the puddle welds because they are secondary to the nodal brace.

While I still tend to think about this in terms of NZS 3404, I like the simplicity of this AISC answer:

http://aisc.org/DynamicTaxonomyFAQs.aspx?id=1646

and the clarity of this bridge resource:

http://www.academia.edu/6206943/Ste..._Highway_Administration_Bracing_System_Design

 

[JAE]

With concrete on the deck there is no accordion action that occurs in my view. There is natural bond on a deck (even non-composite types).

And the perpendicular framing members is a good point, KootK.

Check out Eng-Tips Forum's Policies here:
faq731-376

 

[KootK]

I have a rather special situation that has resulted in LTB needing to be checked between the points of nodal bracing (5' 1/2 span). I've already counted on the supported framing as nodal bracing. I only shared the particulars of the situation to provide context. My question, which you've all kindly weighed in on already, is whether or not this deck assembly, as described, could be counted on as beam top flange bracing in general. I'm still not so sure. Check out the attached sketch for a more explicit depiction of the failure modes that nag at me.

That bridge bracing manual is great -- thanks. Yura's pretty much got a monopoly on all things bracing these days. If he ever gets hit by a bus we may just have to settle for what he's figured out for us so far.



The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[JAE]

I can't see Mode 1 ever happening. Mode 2 perhaps but I'd have to think about it - there is some connectivity between the beam and concrete directly above the beam via irregular welds and friction (I know - you can't count on friction). And the bending stiffness of the deck itself might be adequate - I'd have to put numbers to it to see.



Check out Eng-Tips Forum's Policies here:
faq731-376

 

[KootK]

Normally, having a hot rolled angle as your deck edge element would preclude mode number two.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[DaveAtkins]

I don't believe what you show in your sketch is a realistic failure mode. If there is a potential for that, there are a lot of untopped decks in this world that are in trouble. My opinion--you are overthinking this.

However, if you want to be rigorous, you could calculate the capacity of the deck in weak axis bending (the webs will fail in double curvature).

DaveAtkins

 

[BAretired]

Personally, I would consider it unbraced between the midspan joist and the end supports, recognizing that I am being somewhat conservative.

BA

 

[DaveAtkins]

There is a limit to how far a deck can span, based on gravity loads (bending and deflection). So there is a limit to how far apart perpendicular members can be spaced. I just think that for typical deck spans the deck is able to brace the perimeter beam.

DaveAtkins

 

[KootK]

Thanks for the input Dave. I deny the existence of over thinking. Either I've thought about something to my satisfaction and it has become rote, or I haven't and further thinking is required. Nothing in between.

However, if you want to be rigorous, you could calculate the capacity of the deck in weak axis bending (the webs will fail in double curvature).

You could make a nearly identical argument for an un-topped steel deck. Of course, you never would. Is there not an inconsistency there?

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[DaveAtkins]

Again, we don't check an untopped deck in weak axis bending because we assume it provides adequate bracing for a beam top flange.

DaveAtkins