concrete tilt wall chord angle

[broncosfan]

We are designing a large 500000SF+ warehouse. The solid concrete tilt walls are probably going to be 24' wide, 9 1/4" thick, and 40' high. It's in a low seismic area - design category B. The joists (supporting 22GA metal deck)are at approximately 6' o.c. and sit on 8"x6"x12"long ledger angles welded to steel embeds. My question(s) have to do with the chord force. First, in calculating the chord force, it can be shown that each individual 24' wide concrete panel has enough overturning resistance at 0.6D to resist the whole chord force. Does this mean that a connection between the panels at the roof can be eliminated. From talking with a couple other engineers, this actually seems to be fairly common. Second, we plan to run a continuous angle along the wall, welded to the top of the joists with say a 4" fillet weld. Does anybody see a reason that this angle should be attached to the concrete wall as well - say with a fillet weld to another embed on the top leg? Could this extra connection be eliminated if loads are given on the drawings for the joist seats to carry from the bottom of the continuous angle down to the top of the 8"x6" joist seat? Finally, if splices are needed for the continuous angle on the top of the joists (see question 1), should the angle be broken say at the middle of each concrete panel (12' from the joints) so there is a little room to stretch as the concrete shrinks to avoid diagonal shrinkage cracks. WHEW!!! Sorry about the long post. That took longer to type than I thought it would. I'd appreciate any comments.

 

[UcfSE]

Did you mean to ask if the connection between the panel and the roof angle can be eliminated? I don't think you would eliminate the roof connection. Maybe footing?

At the connection of the roof diaphragm to the panel, you are supporting more than just the roof diaphragm shear. You are also supporting the out-of-plane wall reaction, assuming the roof deck is a point of support for the wall panel for out-of-plane loading. You will need some sort of connection to support that reaction. I think the easiest would be a deck angle unless you joist connection can handle that also.

On the lateral shear side, the shear needs to be able to reach the panel and eventually get into the foundation. If you can transfer the roof shear every 6 feet through the roll over capacity of the joist seat that's up to you. That could turn into a lot of shear every 6 feet to combine with other loads, something to look at.

ACI 318 chapter 14, 14.2.6 states that wall panels must be connected to intersecting elements, and footings. I've seen details where panels are connected with dowels to the lowest floor slab, with nothing to the footing itself, but we always provide a connection to the footing anyway. Maybe it depends on how you interpret the code and your connection detail to the slab.

 

[broncosfan]

The panels will be connected to the footings with clip angles and doweled into the slab. I'm more concerned about what happens at the roof. Let's say that we have a metal roof deck welded to joists at 6' spacing. The joist seats are welded to a steel ledger angle which is welded to steel embedments. Ok, the ledger angle has to transfer wind and seismic loads perpendicular to the face of the wall into the metal deck between embedments. It also has to transfer lateral shear from the metal deck into the embedments parallel to the length of the wall. It also has to carry a chord force from the metal deck acting as a deep beam, right? Would my ledger angle need to be continuous at the 3/4" joints between the 24' wide tilt wall panels because of this chord force? Could I eliminate a splice because the average tilt wall panel has a large overturning resistance due to .6 times it's selfweight? Does splicing the ledger angle between adjacent tilt panels increase the risk of diagonal shrinkage cracks at the upper corners of the panels? Thanks.

 

[UcfSE]

Panels will generally not experience overturning and chord forces at the same time. The walls parallel to the wind feel the lateral shear and overturning and the walls perpendicular to the wind force feel the chord forces.

Try checking out some of the information by the Tilt-Up Concrete Association.

http://www.tilt-up.org/default.htm

 

[broncosfan]

That makes sense to me. But the panels will have to support either the chord force or the lateral shear, whichever controls. If the ledger angle is not spliced at the 3/4" joints between tilt wall panels, I guess I have a long series of individual 24' wide concrete shear walls. Whereas if the ledger angle is spliced at the joints, I might have one single 400' long shear wall. Thanks for the link, I'll check that out.

 

[astructurale]

I'm in precast, and there are minimum requirements for diaphragm to wall panel and wall panel to footing connections (minimum loading requirements for structural integrity). You can't count on gravity and shear friction alone for connections. I would think that tilt-up has similar criteria, and ACI, as UcfSE has mentioned, has this criteria also.

The connection of the diaphragm to the walls is key to get the load into the walls, and even if the tilt-up walls can resist the overturning alone, they still should at least be tied back to the slab on grade (which is also a common detail in tilt-up), but I think they also would anchor the slab on grade to the footing.

Yes, the connection to the panels must resist the greater of the two loads. You do not need to splice the angle at the joints between the panels. The angle does not prevent the walls from acting indivudually. Each panel acts individually and will likely be enough to accomodate the overturning. If not, then you can group your panels together using panel to panel connections along the vertical joints. However, you will have shear build up in the connections along the vertical joints in those panels if you do have to design them as grouped. Those forces can really accumulate.

The PCI handbook has some good examples also with sketches that go into this concept. They also include good information on structural integrity connections.

 

[STR04]

please explain how "each individual 24' wide concrete panel has enough overturning resistance at 0.6D to resist the whole chord force". i keep thinking of a concrete beam with out tension reinforcing...and then snap! the wall falls over because of the diaphragm drifts without a chord. i'm a little confused with this word picture.

 

[whyun]

Typical chord continuity detail I am familiar with at panel joint consist of a blockout with the chord bars exposed. Then the two exposed chord bars from each panels are connected with a piece of angle welded to each bars.

In addition, the wall itself has to be "stitched-up" to the diaphragm to fully transfer the out-of-plane reaction from the wall panel. 0.6D and overturning do not come into play for this load direction.

 

[broncosfan]

If the overall building dimensions of the building between expansion joints is 400' x 400' and the shear force is 400lb/ft, the maximum chord force is 20kip. A 40' high x 24' wide x 9 1/4" tilt wall panel weighs 111kip. 111kip times 12' times 0.6 equals 800kip-ft overturning resistance. The overturning force due to the maximum chord force (at the center of the building) equals 20kip times 40 feet which equals 800kip-ft. This doesn't include the dead weight of the footings, soil on footings, attachment to slab, and the roof bearing on wall.

 

[UcfSE]

What we are trying to say is that overturning and chord force are not related. Chord force does not cause overturning. Chord force is an internal force caused by the external wind force. The tension or compression chord force is a result of the bending behavior of the diaphragm. It does not cause overturning on walls. If you draw a FBD of the chord then the force on one end of the chord will be balanced by the force on the opposite side, with nothing transferred to other members. It is simply a tension or compression felt within the member acting as a flange for the diaphragm. Overturning is caused by the external wind load coming from the diaphragm.

 

[broncosfan]

So right at the 3/4" joint between the individual 24' wide panels, the deck will start to "rip" apart due to the chord force in tension (or buckle due to compression). However, the deck really can't "rip" because it is attached to the wall and each individual wall panel has alot of overturning resistance and is not going anywhere according to the calculations above. I'm not trying to justify anything here. I'm just trying to get a better picture in my mind of what is actually happening. The numbers I ran above would not work out so good in a hurricane region or a moderate to high seismic area. This forum is great because I can ask these stupid questions without looking like a fool in front of my boss.

 

[broncosfan]

When you think about a warehouse roof, you have a metal deck, but you also have rows and rows of joists in one direction and rows of joist girders in the other direction. The metal deck is welded to the joists, the joists are welded to joist girders, and joist girders are welded to the columns. Isn't it conservative to say that the whole chord force is concentrated right at the very edge of the deck? Wouldn't each row of joists and joist girders will behave like a "chord?" I did some reading and in the 3rd edition of The Tilt-Up Design and Construction Manual there is a paragraph under chord forces that says "Some engineers in low seismic areas omit any connection between panels. Even for low to moderate chord forces an argument can be made that the in-plane overturning resistance of individual panels can resist chord forces without an actual connection between panels."

 

[astructurale]

The chord force is not concentrated at the corners of the deck. I don't believe that would be a good design practice, or a conservative one, and probably not the best way to handle the forces in your metal deck diaphragm. The most conservative approach is to attach every joist and girder to the tilt-up panels. Doing so also will keep the connection size minimal. Honestly, I have not seen even one project's structural drawings where only a certain number of joists/girders were connected to the panels, and I have done numerous warehouse structures of this type (just out of precast walls).

You need connections from your girders and joists to the tilt-up panels to transfer the load to the tilt-up panels. This more evenly distributes the load into the walls.

You may not need actual connections between each of the tilt-up panels though. If your panels can resist the overturning without uplift then then only reason to provide panel to panel connections would be for alignment purposes. For precast panels we provide panel to panel connections such that you have at least two or three per panel joint with one of the connections being closer to the top of the panel (1st 1-2ft). I don't know if it is common practice to provide panel to panel connections for alignment of tilt-up panels though. The 3rd edition of The Tilt-Up Design and Construction Manual should address the case for panel alignment I would think. If not if you do not expect to have differential bowing of your panels then you may not need any. In precast panel to panel connections are provided for several reasons: to account for un-even heating of one face of the panel due to sun (air conditioning on other face that would result in bow, bowing due to eccentric loading of joist/girders or eccentric prestressing, and bowing accidentally created by shipping or improper dunnage points for storage in the yard.

I hope that this helps to clear things up for you, and that I have not been too repetitive in my replies.

 

[prblmfxr]

Is the force caused by wind or earthquake? If wind, you will never, ever, ever, ever see these design forces in your building unless you're located within hurricane zone or by tornado. If by tornado, just smile and say building code does not require design for tornado (except for tornado shelter), and you will never have to worry about your building regardless of connection details used. If by hurricane, different story, study is worthwhile.