Compression Flange Buckling

[jdm86]

I'm investigating the integrity of a steel building that is already built. The ability for the building to resist the uplift forces from strong winds on the leeward side has been called into question. It is a large a-frame building that has glulams at 4' OC for framing members. The glulams are attached to the steel beam by means of a stiffener plate.

Q1: Can the stiffener plate acts as compression flange bracing in reverse loading? The beam is a W33 and the plate extends down 2'5 5/8" from the top flange. The plate is 3/8"x9 1/2"x2-5" and is welded to the web of the beam.

I'm really tight on time, we need to make a decision as to how we will move forward in the next few days and my boss wants my input on this by Monday. Thanks for your help.

 

[rowingengineer]

jdm,
A detail would be nice, and how the compression force is handled once in the glulam will also be of interest.


An expert is a man who has made all the mistakes which can be made in a very narrow field

 

[jdm86]

Here is a rough detail of what I'm looking at.

The glulams are supporting the plywood roof, they span 28'. They are spaced at 4' OC going up the steel, the steel is on a 12/12 pitch.

When the leeward side of the structure experiences uplift, it puts the bottom of the W33 in compression. I don't know how to figure out how much the 3/8" plate that stiffens the glulam connection acts as compression flange bracing to reduce my unbraced length.

 

[Lion06]

Why is the detail different for the two glulams framing into the steel?

My gut reaction is to say that it is not braced for uplift at the glulams. If that were a WF framing into a stiffener of shown height My gut would say yes. The problem with the detail as it stands is that the three bolt connection into the wood likely doesn't have the moment capacity or ratational stiffness to be considered a brace point per AISC -especially since they're so closely spaced. I know it seems counter-intuitive, but the closer the brace points are the stronger and stiffer they need to be to be considered brace points. The reason is as the get closer together it is trying to force the beam to buckle in a higher mode, which takes more force to brace it.

Can you add a bent plate from the bottom flange of the WF to the underside of the glulam at maybe a 2H to 1V? Without running any numbers that seems to feel a lot better. You'd still have to do the strengh and stifness checks, but at least it would be axial strength and stiffness and not rotational. It's tough to get moment connections in wood and the biggest reason is that the rotational stiffness is difficult to achieve so even is the moment capacity is adequate there is so much rotation that the connection never sees the moment.

 

[hokie66]

The connection does brace the bottom flange, but maybe not enough. It would be a lot better if you could get a lag screw up through the seat plate into the glulam.

 

[jdm86]

@ StructuralEIT - The connection is a-symmetric because it was assumed that the stiffener on the one side would adequately take the load. Each glulam has one side of the bent plate and one side with a stiffener connection. What you're getting at is exactly what I'm investigating: what bracing is provided and what is the failure mechanism for the compression on that flange. I understand the high frequency failure mode, that makes sense. Good point!

@Hokie66 - I like your idea about lagging through the seat of the glulam connection. I am leary of considering the bolted connection a moment connection because I know that the bolts can shift in the oversized holes slightly before taking any moment resistance. If I lag the underside, I will not have any shifting. The alternative perspective to that is that the plywood sheathing will act as the tensile reinforcing for the moment transfer across the top flange of the beam if the beam goes into uplift. You just have the glulam compressing against the web of the beam and the tension taken in the plywood.

It still comes down to how much the stiffeners can transfer to the glulam.

Thanks for all your input

 

[BAretired]

I would consider the bottom flange unbraced as it stands, particularly if the holes are oversized. I like the idea of adding a diagonal brace, but I would not do it at 4' centers...maybe 12' centers would be more appropriate.

Calculate the compression in the bottom flange due to wind uplift and ensure that the brace can provide 2% of that force laterally.

BA

 

[ToadJones]

In my opinion the continuous plywood sheathing, if running over top the W33 as shown, will add to the rotational stiffness of the glulam-to-girder connections.
You might be leary of counting on this, but it is there.

 

[BAretired]

I am leary.

BA

 

[Lion06]

I'm not sure I believe the diaphragm helps. I believe it increases the moment capacity (though we would never count on it), but it still isn't changing the rotational stiffness of the WF to glulam connection unless the diaphragm is attached to the glulam AND the WF, which I'd be very surprised if it is. It might help the stiffness of the actual glulam, but that's not the problem, it's the connection of the glulam to the WF.

 

[ToadJones]

Having the sheathing does nothing whatsoever.
The glulam connection is identical with or without the sheathing. The glulams are free to rotate with the sheathing there as the wood be without the sheathing there.
No force will be transferred through the sheathing across the beam.
If the glulam experiences uplift, the sheathing will not see any force and will not add to the global stiffness of all members involved.
Having built some similar structures with my hands and not just on paper, I'd be very surprised if there isn't some good ole' PL 400 subfloor adhesive between the steel beam and sheathing....just a hunch.

 

[ToadJones]

if we are counting on the rotational stiffness of the glulam-girder connection to provide bracing to the bottom flange via a stiffener, how can we ignore the sheathing/diaphram?

 

[rowingengineer]

Hokie is on the money, a lag screw in the bottom of the glulam and maybe one in the top.

I don’t agree with SEIT on the rational that if you have more braces these must be stronger. I would think you would be able to spread the load over a few connections.


An expert is a man who has made all the mistakes which can be made in a very narrow field

 

[ToadJones]

BA-
Often your replies go only towards furthering your engineering rectitude and do nothing for the furthering of the topic and their condescending tone is a bit off-putting.
....Just a peanut-gallery observation.

 

[ToadJones]

Throw some Tek screws though the sheathing into the top flange of the girder.

 

[BAretired]

Perhaps there is a gap between the deck and the top flange to allow for shrinkage of the wood beam (he said condescendingly).

BA

 

[BAretired]

There is another issue with respect to the existing beam hanger. It is a deep hanger with fasteners installed through the side plates near the top of the supported beam. This may promote splits in the beam should the member shrink and lift from the bottom of the beam hanger.

I do not believe it can provide reliable bending resistance required to stabilize the bottom flange of the steel beam.

BA

 

[tclat]

I agree with BARetired that the bottom flange is unbraced. Check the moment capacity of the beam at various unsupported lenghts and unstall a diagonal brace as required.

 

[Lion06]

Rowing-

I know it's counter-intuitive, but the closer spacing of braces forces the beam to buckle in a higher mode. This requires each brace to take more force. Check out the equations in App. 6. Part of the reason is that, typically speaking, more braces means greater beam capacity which means greater brace force.

 

[hokie66]

SEIT,

It may work that way in the US, but not in Australia. Sometimes I wonder about the way things are going in my home country...