Building expansion joint or not?

[ars001]

I'm designing a single story office/warehouse with a mezzanine in MN. Walls are precast panels. Roof is metal deck with beam and column lines. Building is approx. 165'x480'. Should I break up the 480' dimension with an expansion joint? Obviously for ease of design I'm trying to justify avoiding it. Thanks

 

[frv]

Any time you get above 200-300 ft, you really should provide an expansion joint.. I'd say definitely for 480 ft.

 

[msquared48]

I would be very curious how this would work for the roof diaphragm... Is there an interstitial, internal shearwall or moment/braced frame line that could be used at the juncture of the proposed joint?

Mike McCann
McCann Engineering

 

[hokie66]

I agree with frv, your building deserves a roof expansion joint. Two separate roof sections, each 165 x 240 is appropriate.

 

[strguy11]

MSQUARED -

You wouldnt need a internal shearwall. You can still provide an exp joint at the middle, and transfer the shear across. The SDI diaphragm manual has a great example on this. The attached page from the manual shows to not provide the EJ continuous across the entire width and leave the middle connected. Shear will be smaller at the middle, so your decking should still work.

Or you could still provide the joint all the way across the building, and provide tensions ties on double beams (on on each side of the joint). This should be slotted to allow for movement in one axis, but transfer the shear parallel to the joint.

 

[lkjh345]

ars001:

FWIW: We have done similiar warehouses ( Precast walls with columns and joist/joist girder roofs, some with full length mezzanine floors) with approximately 500 feet between expansion joints in the midwest (USA). Can't say it doesn't make me nervous every time, but we have had no reported problems with these buildings and several of them have been in place for many years.

If you can easily add an expansion joint, I would just to ease the mental stress, but it can be done otherwise.

If the internal layout allows, you can use a double row of columns with cable cross bracing for the lateral system. Very cheap and efficient lateral system inside a warehouse structure.

 

[ars001]

Thanks for all the responses. Its all good input. I'm in lkjh345's boat. I've designed warehouse/office buildings up to 500'x500' without the use of expansion joints. And no reports of any problems. I also too get nervous about it. I just thought since this building was more of a rectangular structure that it just "looks" like it shall have an expansion joint. According to the Tech Report No. 65 that the AISC references, I fall below the limits for needing an expansion joint. Obviously I'm on the upper limits. Plus, I think its going to be hard since I have an internal mezzanine that runs the full 480'. If I do decide to put a expansion joint in, I will probably treat my diamphragm as a "3-sided diamphragm" and do the analysis that way.

 

[structuralnerd]

Could you do the expansion joint with the double column line and provide a deck angle at each side and treat it like two separate buildings all together? So, no shear transfer through the two buildings? Would the precast panels need to provide an expansion joint as well? What are the issues that come up if you do it that way?

 

[civilperson]

Is your buiding heated and cooled? What temperature change do you envision to require an expansion joint? If not restrained by rigid connections to other structures, a stand alone building, (less than 500 feet dimension), can be built with no expansion joint.

 

[hokie66]

While I don't like the idea of a 480' long building without a roof expansion, I like the idea of two cee sections each 240' long even less. I had assumed the building was a series of portal frames, but if you are depending on the deck as a diaphragm, I wouldn't cut it in two.

 

[frv]

Maybe I'm missing something.

What's the problem with breaking the diaphragm in two. You can detail it so that the diaphragm still allows for shear transfer while allowing it to expand in the long direction.

Additionally, you are breaking the diaphragm in the middle, where the shear is negligible anyway.

 

[hokie66]

Sure, the shear would be small, but the chord forces would be large. You can't have both continuity of chords and a movement joint.

 

[frv]

hokie66-

I'm not sure I follow.

are you talking about the chord forces in the long direction? (i.e. perpendicular to the expansion joint at the edge of the deck, running the full length of the building?)

If so, I don't really think a building with a 3 to 1 aspect ratio would be that troublesome.

 

[hokie66]

Yes, that is what I was talking about. In the absence of a bracing element in the vicinity of the expansion joint, the diaphragm has to span between the end walls. Therefore, the chords can't be broken, and if they are not broken, you don't have an expansion joint.

 

[frv]

hokie66-

At the expansion joint/chord interface, you typically attach one side of the joint to the chord and allow the other one to move laterally; the chord itself is unbroken.

You are still able to transfer shear across the expansion joint (by, for example providing a 1" bolt across the joint, without allowing the bolt to move in the short direction.

I still fail to see where the problem lies.

 

[hokie66]

frv,

Perhaps the reason we don't understand each other lies in the nomenclature we are using or that I just can't visualise your detail.

Fundamentally, the OP is trying to distribute the entire lateral load at roof level to the end walls which are 480' apart. He wants to do this by some type of diaphragm. Now if this diaphragm is broken in the middle and the two halves are prevented from moving differentially in the transverse direction but allowed to move in the longitudinal direction, you have a longitudinal expansion joint. But I fail to see how the chords on the two long sides can be unbroken and maintain a movement joint. Crossing the joint, they lock up the joint.

If there is an expansion joint and there is no bracing in the vicinity of the joint, the two ends of the building will behave like C sections, and a 240' wide channel flange is too much.

Perhaps someone else reading this thread can mediate the misunderstanding frv and I seem to have.

 

[csd72]

A reference:

http://www.modernsteel.com/Uploads/Issues/April_2005/30740_fisher.pdf