WF Beam - HSS Column Connection for Flexible Moment Connection (FMC) Moment Frame

[KootK]

We'll get the preliminaries out of the way here:

1) I probably hate FMC more than you do. Save your breath.

2) Yes, I very much wish that I were using wide flange columns.

I need to come up with an FMC moment connection between an HSS column and a wide flange beam. The detail shown below is the best I've come up with so far. Think Empire State but with HSS. Because it's FMC, my goals here are:

A) kinda stiff so the building doesn't flop over.

B) Not so stiff that I'm back to having to consider gravity moments in my columns.

C) Reasonable to erect. Not so sure I've achieved this. Might need to field weld the upper WT.

My questions:

D) What do you hate about what I've done?

E) Got any better ideas?

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[jayrod12]

Care to explain to a stupid layman like me why that's any more flexible than a standard moment connection?

 

[WARose]

Given any thought to a end plate connection? They can be just as flexible as something like this and (I would think) be faster to build.

With either one you are going to pick up some gravity load.

 

[KootK]

Care to explain to a stupid layman like me why that's any more flexible than a standard moment connection?

WT's might flex a little? Kinda looks like what they show for WF columns in the literature? This is a bit part of what I dislike about this. Weak qualification for the degree of flexibility needed/provided.

Given any thought to a end plate connection? They can be just as flexible as something like this and (I would think) be faster to build.

I hadn't. I've only got about 1" of clearance on the side though so bolts outboard would be a problem.

With either one you are going to pick up some gravity load.

Preaching to the choir. Tell it to the FMC proponents.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[structSU10]

might be funky, but could you slot the top WT at the web and put it under the beam top flange? leave "tabs" wide enough to accept the bolts. Not sure if that is still a pain to erect. could also try to hollo bolt on the WT or something.

 

[jayrod12]

To me, a flexible moment connection could be as simple as a good shear connection. There's no way a simple shear connection couldn't transfer some load from racking. But as you indicated, we aren't here to debate the use/effectiveness of FMCs. Is this a delegated design item? (And I mean you are doing the delegated design and the EOR has indicated the column isn't capable of any gravity moment)

 

[Klitor]

Can you make it into (probably cant but ill still ask) some sort of cap-plate connection?

 

[SAIL3]

FMC may only be flexible for a certain range of loading and it is difficult to accurately establish that range.....I would make it a moment conn and design the col accordingly....that way, I take all the uncertainty
out of the problem.....

 

[canwesteng]

What's the lateral load look like? If you just need to resist notional loads I like shear tabs, following the extended shear tab procedure per AISC/Muir, just designed for a little more moment. Knife plate through the HSS or reinforce the face of the column if required.

Otherwise, I don't like the weld from your T's to the HSS. They don't look strong enough to fully develop the bolts or the stem of the T. I think the key is that you have a ductile failure mode if you're counting on some partial fixity, so welds should fully develop the material they connect to. Likewise, bolt failure should govern your anchor bolts.

 

[JoshPlumSE]

Why not run the WF over the top of the HSS and then cap plate the HSS and bolt it to the beam? Kind of like a rotated end plate moment connection.

I have to think that's an easier and more robust connection. Though you'll get some gravity moments in the column.

 

[dik]

as Josh noted or a wide end plate on the beam and column face and bolts would likely be the least expensive.

Dik

 

[KootK]

might be funky, but could you slot the top WT at the web and put it under the beam top flange? leave "tabs" wide enough to accept the bolts.

That's an interesting concept if I could make the geometry flush out. I had considered it previously but rejected it because I was concerned that severing the tee web like though would compromise stiffness too much. However, the FMC method has you essentially ignore closing moment connections and rely predominantly on opening moment connections. And this version of the connection might be uniquely suited to that. Flexible for opening and stiff for closing...

could also try to hollo bolt on the WT or something.

This may be the way to go really. Perhaps this is what WArose had in mind too. With a 10x10 HSS left open, do you think it might be possible to use conventional bolts rather than Hollow? To a degree, I'd think that one could reach there arm in there to get the job done. Maybe provide an erection seat. Not sure if that would violate some OSHA rule or something though.

To me, a flexible moment connection could be as simple as a good shear connection. There's no way a simple shear connection couldn't transfer some load from racking. But as you indicated, we aren't here to debate the use/effectiveness of FMCs.

I'm EOR here but I'm dealing with an unusually sophisticated client which limits my options. I agree that a good shear connection can transfer some moment, particularly joint opening moment, but I'm reluctant to rely on such a connection as my entire lateral system if it doesn't match the handful of connection styles that I've seen presented and tested in the literature. And what I've seen in the literature is overwhelmingly bolted on clip angles and tees at the flanges. I may already be off reservation in that I've not seen any HSS examples. The HSS situation tends to favor welding rather than bolting and, in my mind, that makes it tougher to provide the nebulous, ill-defined degree of connection ductility required for FMC.

FMC may only be flexible for a certain range of loading and it is difficult to accurately establish that range.....I would make it a moment conn and design the col accordingly....that way, I take all the uncertaintyout of the problem.....

You take out the uncertainty but, additionally, also the primary benefits of using the system in the first place (beams, columns, and connections not designed for gravity moments). I'm with you in spirit as I'd very much prefer to do conventional moment frames. That said, with the FMC path decision already in the rear view mirror, I'll not be neutering any of the beneficial aspects going forward.

Can you make it into (probably cant but ill still ask) some sort of cap-plate connection?

Why not run the WF over the top of the HSS and then cap plate the HSS and bolt it to the beam? Kind of like a rotated end plate moment connection.

The whole idea here is to be able to disregard gravity moments with some degree of credibility. I don't see how that would be possible with this connection style at interior column joints. And, again, it's not a scheme that's gotten any air time in the literature.

What's the lateral load look like?

It's not notional as it's the primary lateral system for the building. That said, nothing's working to hard and it's mostly drift control governing things.

Otherwise, I don't like the weld from your T's to the HSS. They don't look strong enough to fully develop the bolts or the stem of the T.

Thanks, I'll check it out once I move past the concept stage. I'm not trying to fully develop anything really. That said, it would be nice if weld failure could not be the governing mode.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[KootK]

Care to explain to a stupid layman like me why that's any more flexible than a standard moment connection?

I'm afraid that I can't. It's actually a very stiff flexural connection that would likely fail in a non-ductile fashion. So, in retrospect, my detail sucks and accomplishes little of what I need it to. I submit the detail below as an improvement. I'd be intending for the vertical angle leg to yield flexurally prior to tearing everything else apart. Goofy FMC...

I feel that this gives me a measure of predictability at the top of the connection. At the bottom, I guess my options are:

1) Design the angle leg to yield prior prior to HSS wall buckling.

2) Design HSS wall buckling for lateral induced moments but ignore gravity effects (*barf*).

3) Design the HSS wall for gravity moments as well as lateral moments, perhaps adding a vertical plate to the seat and extending it above the bottom flange as required to distribute the load.

I'm leaning towards #3.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[jayrod12]

For what it's worth, I like that detail much better.

I feel that regardless of direction of moment, i.e. opening or closing, since your flange force would be equal the limiting factor (and easiest to design) would be the upper angle leg yielding and bending laterally.

To me, the bottom seat can be infinitely stiff both vertically and laterally and the weak point of the connection would always be the top angle.

 

[EngineeringEric]

If trying to eliminate the gravity loading.... Can you just place a horizontal steel plate wider than the your beam flange first, then weld that (overhead) to the outer edges of the top flanges and similar to bottom flanges after the dead load is placed? Thus all future loads will be live and lateral and not mess with FMC... cause nobody wants that.

Basically shear tab only for DL, OMRF for Live and Lateral.

Your second detail is only flexible for one direction, the first detail is flexible for both positive/negative moments. Does that matter?

I think using angles for both flanges is pretty easy to construct, easier than welding up your own WT. L4x8 x bf wide and LLH

 

[KootK]

I feel that regardless of direction of moment, i.e. opening or closing, since your flange force would be equal the limiting factor (and easiest to design) would be the upper angle leg yielding and bending laterally.

I agree, thanks for the bounce. I was hoping to design the vertical angle leg to yield in bending before the bolts etc gave way. Poor man's capacity design of sorts.

Can you just place a horizontal steel plate wider than the your beam flange first, then weld that (overhead) to the outer edges of the top flanges and similar to bottom flanges after the dead load is placed?

I could, and it would be awesomely stiff. Unfortunately, even the live load moments would pretty much kill off the spirit of the thing.

and not mess with FMC... cause nobody wants that.

It's not you or me but I'm afraid that somebody does want that. Rest assured that if it comes up again, I'll be steering things in another direction.

This has unfolded in a rather interesting manner.

1) I went with the welded tee thingys shown at the top.

2) Steel fabber pointed out that, with the tees present, he would have to cope the top and bottom flanges to maintain a normal column to beam gap.

3) Before I could respond, fabber said "never mind, I see that note blah blah says to use WT shear connnection for this situation instead of knife plate and now, because of the tee flange thickness the top and bottom, there is no interference." I didn't even realize that I had such a note.

4) I know feel that a simple, WT shear connection is the optimal version of this, sans a direct flange connection. It's flexible, it's got plenty of precedent not causing problems as a simple shear connection, and a direct flange connection isn't necessary to mobilize the small amount of moment needing to be transferred. And it's cheap and easy to erect. I'll just make the tee to column welds nice and stout.


I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[Prestressed Guy]

I would be concerned about the lack of web stiffening in the HSS. Many years ago a coworker tried to design a lifting bracket out of thick walled square HSS sections. The bracket was shaped like a squared off up-side-down question mark. the joints were mitered with full-pen welds and the engineer assumed that he would get something close to the capacity of the section.
When tested, it folded up at about 10% of the section capacity. The webs bowed out and the flanges collapsed. as I see it, there is no way for the compression load in the beam flange to be transferred to the HSS section.

 

[dik]

mitigated a bit by the use of Ts... they can span from one side of the tube to the other. Could consider using steel shims to eliminate the snugness and avoid field welding.

Dik

 

[KootK]

Fun, wasn't expecting this to come back to life. For what it's worth, the sketch shown below is what I went with in the end, mostly out of deference to jayrod's point that my original flexible moment connections weren't all that flexible. Topside, I basically tried to replicate the source of flexibility typically found in the more ubiquitous versions of the detail involving WF columns.

When tested, it folded up at about 10% of the section capacity. The webs bowed out and the flanges collapsed. as I see it, there is no way for the compression load in the beam flange to be transferred to the HSS section.

Thanks for sharing your experience Haydenwise, I'm grateful to have access to it. I've long been interested, philosophically, in whether or not engineers should really be allowed to do anything for which relevant testing is not available. The Northridge issues, everything associated with ACI appendix D, and the approach now being taken in high seismic areas would seem to suggest otherwise. I'm in no hurry to have my flexibility limited as a designer but historical data seems to suggest that we really can't be trusted to make up our own stuff based on first principles. There, I said it.

mitigated a bit by the use of Ts... they can span from one side of the tube to the other.

My thoughts as well and I've retained that feature for the predominantly compression part of the connection. As Haydenwise suggested, I doubt that one could count on developing the full flexural capacity of the tube column. However, with judicious Packer-esq checking of HSS wall failure modes, I believe that one could rely on the flexural strength associated with the outcome of those checks.

I like to debate structural engineering theory -- a lot. If I challenge you on something, know that I'm doing so because I respect your opinion enough to either change it or adopt it.

 

[BadgerPE]

I just Googled "Tube Column WF Beam Moment Connection" and an image popped up from this thread. Apparently I spend too much time here, because without seeing the website it was at, I knew it was KootK's handwriting.