CMU Shaft (Non-lateral force resisting) connection to Composite Steel Podium Slab Debate

[MidwestSE]

Hey Everybody!

Here is a question we discuss internally frequently and I'm curious what everybody else does.

Situation: We have a composite steel podium with several stories of wood on top (+/- 50' tall building). We have a CMU elevator shaft that we're not using as a lateral system. On the majority of our jobs, we do not have any steel beams supported from the CMU (for construction scheduling purposes) and do not connect the wood to the CMU. Long story short, the CMU shaft is along for the ride.

Question 1:
Does anybody else routinely do it this way where the CMU shaft is along for the ride and not participating as a lateral element? Any comments on why/why not are welcome.

Question 2:
Would you tie the concrete slab into the CMU shaft?

Question 3:
If you wouldn't tie the slab into the shaft, would you detail a movement gap around the shaft at the slab or pour the slab up to the CMU and force the cmu to go along with any movements?

All thoughts/perspectives welcome!

 

[jayrod12]

If I'm understanding correctly, the CMU shaft isn't even along for the ride, it's free standing in the centre of a separate building? It's only connection is at the base of the shaft?

 

[MidwestSE]

Correct - if you leave a gap at the podium level - for H/400 on a 12' story - say 1/2" compressible fill all the way around, then the CMU won't move except for what the wood stories push on it up above (unless you detail a movement gap around the wood up top...) so it is basically never moving. Is that how you've seen it done, or do you normally see the shaft tied into the slab?

 

[phamENG]

I thought about this a little bit a couple years ago designing a similar building (though the podium was all concrete and connected to the shaft).

Start at the top of the building:
Wind pushes on the wood superstructure. It's going to deflect. If you have a movement joint large enough, then sure. Wind isn't going to push the CMU tower (this assumes there is no over-run space poking up and exposed directly to wind above the roof) and the wood will move while it stays still. But, if there's no gap, you will get movement. The CMU is probably going to be the stiffest thing around, so it will attract the load - even if you don't fasten to it. (The diaphragm will push on it one way or the other - without fastening you could end up with some unpleasant stress concentrations).

Now move down to the podium:
You have a gap here. The podium is going to move with some period in the wind, and the CMU will either stand still or move with some other period from the wind loading on the upper structure being transferred to it. The former (CMU standing still) would be easy. The other sounds like it would be a nightmare to predict with any accuracy, though you could probably get away with the gap designed for the CMU standing still.

But wait - this is just wind. Bring seismic into the mix and it all goes crazy. If you're going to give them a gap, you need to give them a gap such that no load is transferred, they need to be seismically separated. This is really punishing as you have to use your amplified deflections at the level in question (ASCE 7-10, 12.12.3). I wouldn't be surprised if your gap at the upper floors is 6 inches wide or more. What architect is going to accept that around the elevator/stair tower? So then you tie them together. Now you're CMU is participating at the upper floors but not at the podium. That means the base of your CMU will now have to span vertically for at least 2 floors before it reaches a diaphragm. That makes the design of the lower portion of the shaft that much harder.

Long story short - I considered a similar arrangement and quickly dismissed it. The diaphragm detailing was a bit more complicated, but it was worth it in the long run.

 

[MidwestSE]

Thanks, phanENG. That's essentially the exact debate we've had before and are having now.

So you tied the CMU into the concrete podium, correct?. What about at the wood upper stories? Did you explicitly tie CMU to wood or not? Our typical practice is to build a bearing wall around the shaft and "isolate" the wood from the CMU (with a small gap..."small" is decided by the GC) so we don't have any issues with differential movement.

 

[phamENG]

Tied it all together and detailed the diaphragm accordingly.

If you just do a "gap" with no dimension, you have to way to predict the behavior. If small is a quarter inch, what happens if you if you get 5/16" movement at the top of the building? In a strong storm I'd say that's reasonable with 3-5 stories of wood over a level or two of podium (probably underestimating it). That means that the floor diaphragm is pressing into the shaft on one side. You may have ignored it in the analysis, but now you have a shaft with a very high relative stiffness compared to your wood shear walls preventing the diaphragm from moving. That throws a bunch of load into the diaphragm in an area you didn't expect it, and there's no mechanism to disperse that load to a manageable level. If the shaft is near the edge of the diaphragm, you could end up with a real shear failure in your diaphragm there. Since you have enough shear walls elsewhere, you're unlikely to have a collapse - but you're more prone to extensive damage that could be easily prevented. Not only that, you're technically in violation of the structural separation clause I quoted earlier. You're intentionally creating a situation where "portions of the structure are not designed and constructed to act as an integral unit in resisting seismic forces." So you can't just leave that gap up to the contractor - it has to be defined based on predicted elastic drift amplified for approximated inelastic behavior.

The other issue is differential vertical movement. Your wood structure is free to move 100% independently from the CMU - this is good in that it prevents stresses from building up, but it's bad because your sills between the two are going to come out of alignment. You have to work closely with the architect to make sure it's detailed to account for that.

 

[phamENG]

Oh, and another thing to think about - you have to look at the compressibility of your compressible material. If it is 50% compressible, your gap needs to twice as large as you calculated. Once the gap has been closed by 50%, the filler can no longer compress freely and will begin to directly transfer the load to the adjacent floor.

 

[jayrod12]

My experience is the same as Pham's except that we don't deal with seismic, so I haven't seen that level of total deflections. II've seen people structurally separate it, however this was for lower rise, i.e. 30ft. At 50ft tall can the cmu wall be standalone? I guess if it sees no lateral loads it's probably fine. But then you would need to isolate it everywhere on every floor. 1.5" gap for wind only is tolerable with standard detailing. For seismic, if the drifts expected are in the range phamENG is indicating, then that's unreasonable. Once you start connecting it, you may as well do it for the whole way down.

What's the lateral system for the podium? Braced frames?

 

[phamENG]

One thing to remember - just because the wind load is higher than the seismic base shear, you can't just ignore seismic. There are still detailing and other prescriptive requirements that must be met. Structural separation is one of them. Most of us in low seismic zones can get away with ignoring it because low rise buildings won't quite reach a point where it matters. And if they do, the likelihood of ever seeing the event is low enough that we get away with it. But once you start getting into midrise - it's time to pay attention.

Don't believe me? Think about the quake near Richmond, VA a few years ago that broke the Washington Monument or the Caribbean quake a few weeks ago that set downtown Miami swaying. Both of these occurred in places where most structural engineers with respond to the word earthquake with a "huh?" and a cocked head. (I live in the Mid-Atlantic area, so I know - but I've had an opportunity to design a few buildings in Southern California and had a professor who, oddly enough, is studying base isolation systems.)

 

[MidwestSE]

hey Jayrod,

Thanks for the responses! Our lateral system varies (conc shear walls/brace frames/moment frames). Depending what system we use, the lateral frame is either as stiff as the CMU shaft or significantly less (i.e. moment frames)

Thank you both for the responses!

 

[Jonathan Monge Cubillo]

Hi all!

PhamENG I´ve a question. Whan you say detailed the diaphragm accordingly, do you think its possible to detail the collectors in a metaldeck diaphragm?
I´m currently reviewing a mid rise on cold formed steel (5 stories) over a concrete podium.

 

[jayrod12]

Jonathon,

You're best off starting your own thread outlining your question in detail. Include as much detail and information as possible to receive the most appropriate answers.

 

[phamENG]

As jayrod said - start your own thread. And include a lot more detail in your question. For instance, what aspect of collector detailing are you trying to address. Lots of people will be willing to help you once you start your own thread.