roof expansion joint and drag strut

[a2mfk]

I'm one of the first to say post a sketch, so if this is the case with this question let me have it...

New Building:
Wind but no seismic, in Florida
35ft tall tilt panels including parapet, roof height @ 30ft/
Steel framing interior, front side only will have offices with a second floor (steel frame and composite deck-slab) and the rest is open light industrial and warehouse.

I have selected two locations at plan geometry changes for expansion joints in the roof, and because the length of the roof requires them, dividing the roof into 3 sections. For the purposes of this discussion it can be one long rectangle with an approximate 4:1 ratio for the whole building.

1. Forgive me if this is a silly question, do I need to continue the expansion joint at the second floor also? I only have a second floor along the front side of the building that extends back about 1/4 of the building width.

My immediate response is no, the expansion joint is for the roof not the entire building. Worse case I have only 3/4" of total expansion at each joint, which includes the end and center sections.

The expansion joint(s) detail right now is along a single column line with one roof WF beam. This beam will also act as a drag strut and at one end of the building as part of a braced frame.

Both sides of the deck are mounted to two continuous plates that can slide on top of the WF with a gap in between obviously to accommodate the movement. I assume I need to let both sides of the roof float.

2. BUT- Can I rigidly connect the middle roof section along both edges and design for expansion while allowing the end sections to slide at the expansion joint? Or must both sides of an expansion joint be unrestrained.

3. At the columns along my drag strut/ expansion joint, I assume I also need to detail the roof girders to slide at the seated connection to the column? I would imagine with the dead load alone the friction would act as a restraint at the slotted seat connection, so if I want it to slide this would have to have some type of low-friction slide pad? Is this all even worth it for 3/4" expansion, or can I just take this into account and design for an additional P-delta of 3/4"?


Sorry if this got wordy, and I will post a couple of sketches if that helps. I have an excellent expansion joint article but it doesn't cover these details. Any other articles or links so I can do my own research are also welcome.

Thanks in advance.

 

[Lion06]

A sketch might help. The way I'm picturing this, I can't see why you wouldn't need an expansion joint at second floor. Otherwise you're asking that diaphragm to negotiate the differential movement of frames across an expansion joint which essentially is bridging two separate structures.

 

[Ron]

No need for the joint to be at the floor level...roof only.

Slip pads not that common in construction down here. I'd be shocked if they got the installation right. For only 3/4" theoretical, I would probably just consider the p-delta and move on.

Maybe I'm missing something here, but if you break the structure, there's no need for the deck to slide. You'll have an expansion joint in your roof membrane. Attach the deck either side, gap it about 1-1/2" and don't worry about it.

 

[Lion06]

Ron-
Do you have any concern about the second floor diaphragm negotiating the differential movement between the two structures?

 

[hokie66]

In my opinion, if expansion joints are required, the structures each side of the joint should be separate. Joints relying on slots and slide bearings are prone to locking up. Forces can be transmitted between the structures parallel to the joint, but not perpendicular.

 

[Ron]

PA...no. The floor is significantly below the roof and the movement occurs mostly at the roof.

hokie66 stated my point better than I.

 

[a2mfk]

Thanks guys, sorry I should have posted a sketch to begin with, I hate it when people don't do it and then I am that guy...

Ron- I agree with your joist girder slide connection comment, I will design for P-delta as its only 3/4" theoretical. As you can see in the sketch, its not a lot of expansion, but there are some odd angles to the roof plan..

I know ideally you always have two rows of columns at an expansion joint, but that is not really an option here as I need to keep it to one column line. It is going to act as a drag strut and its going to be part of a braced frame taking shear from both roof diaphragms.

Ron- you answered one of my most important questions, do I need to have both sides of an expansion joint be unrestrained. I did not think so for a steel frame and deck roof. I was thinking I would fix the center section of roof, and let the outside sections be the unrestrained sides. Its no different then three different buildings pushed together, the ends of buildings are never unrestrained...

 

[Ron]

a2...you might consider locating the EJ along the building corner angles. You will have a deck direction change there anyway, so it is convenient and easy to construct. There you could probably use a 2-column setup as well.

Also, I understand you probably don't have a lot to do with the roof membrane, but you might suggest to the architect that there be a roof membrane expansion joint to coincide with your structural joint, but also have a membrane expansion joint halfway between on either side as well. Those would only have to go through the roof system, not the decking.

 

[a2mfk]

Thanks Ron, I will have to take a look at the column grids, where there is second floor, etc. for my braced frames to align with this joint. They just did some major revisions and I haven't seen all of the revised drawings, just want to nail down this concept so I can move on with the frame design. I'll definitely have the architect detail an expansion joint at the same location, and coordinate that with them.

Why do you like the joints from corner to corner? Not that it is a bad idea, but you have an odd angle at some point where the deck will have to change orientation and be the ends will have to be cut at angles... Sorta seems six of one, half dozen of the other. But I will take a closer look at that possibility.

 

[Ron]

I like the corner-to-corner because it allows the deck direction change to coincide with the structural and roof joint locations. Otherwise, you have a roof membrane being installed over a deck direction change, which is hell on the membrane for the long term.