Do we need expansion joint between main building and elevator tower

[jdengos]

Wondering whether we should place an expansion joint between main building and elevator tower due to seismic load. Please see attached plan, 4-story building. Even this building may not need since we have shear walls in the building, deflection is small. But for moment framed building, or for PT building, do we need expansion joint?
Thanks

 

[jdengos]

There is one little slab between them. I am afraid it will add torsion to the elevator tower, also damage the little slab.

 

[milkshakelake]

There is an argument to be made that the connecting slab has enough strength to transfer the lateral loads, and the elevator tower can be designed for the additional torsion.

 

[jdengos]

The tower is columns frame only with cmu between levels. it's said the main building stair walls each end limit any deflections at end, so no need for this case.
We have a similar PT slab building with concrete moment framing, it's said we need expansion since any force on PT slab would damage the PT building, not sure it's exactly true.

 

[JoshPlumSE]

I think it's good practice to have an ISOLATION joint between stairs and the main building. The fear is that the diagonal members in a stair tower create enough stiffness that the stairs will absorb a good amount of lateral load that was intended for the main lateral force resisting system. Especially when you have moment frames as your main lateral force resisting system.

Regardless, if the stair tower is designed as an outside vendor (as is common practice) then if it is going to receive load from the building then you need to let the vender know how much additional load to design for.

 

[jdengos]

Yes, the elevator column frame tower is outside of the main building. Main building has one concrete shear wall stair tower each end.
I feel we still need expansion, but boss said any tower outside of main building never need isolation. Only the PT slab building need since any force on PT slab would collapse the PT building, not sure exactly why.

 

[KootK]

it's said the main building stair walls each end limit any deflections at end, so no need for this case...Main building has one concrete shear wall stair tower each end.

I agree with that logic in support of not installing a movement joint.

I am afraid it will add torsion to the elevator tower, also damage the little slab.

Let the numbers be your guide on this. Design the the elevator frame for the anticipated movement of the stair shaft.

Only the PT slab building need since any force on PT slab would collapse the PT building, not sure exactly why.

The post tensioning will tend to shrink the slab between the stair shafts. That need not affect your decision making regarding the elevator shaft however. This issue is normally dealt with via intelligent building planning:

1) Delay strips between the shafts.

2) Slab to wall joints that temporarily allow differential, lateral slip.

3) Don't infill the elevator shaft with block until after most of the slab movement has taken place.

... probably a bunch of other stuff.

 

[Greenalleycat]

I'm a tad confused due to the Americanese on the plan, so I don't really understand whether this is a wall or frame building

The important thing is to be aware of the engineering fundamentals - a large eccentric stiffness can cause highly undesirable performance in an earthquake (see the CTV collapse in Chch)
If large stiffness is present then it is best usually to isolate the two structures to avoid inducing large torsional responses
It sounds like, in this case, the external stairwell is a pretty lightweight structure that won't significantly affect the primary structure?
In this case it may be fine to connect the two, but be mindful of the potential to induce large torsions onto the stairwelll - it is possible and maybe even likely that the end connected to the main building will go one direction whilst the unconnected side tries to go the other way

Also, I will second the comments about the stairs
This was a big lesson from building performance here - many stairs became stiff struts between floors and had highly undesirable performance, changing structural performance and often times collapsing the stairwells
Now, all stairs in seismic areas have sliding details at the base to prevent this

 

[KootK]

The important thing is to be aware of the engineering fundamentals - a large eccentric stiffness can cause highly undesirable performance in an earthquake (see the CTV collapse in Chch)

While I agree with that statement in general, I do not feel that it's relevant to this building. I would expect the awesome, torsional lever arm between the two internal stair shafts to iron out any stiffness eccentricity problems regardless of how stiff the elevator shaft is or isn't once the block is installed.

 

[Greenalleycat]

It was meant to be a general statement, trying to loosely respond to the general statement from his boss that I do not agree with

I feel we still need expansion, but boss said any tower outside of main building never need isolation.

As I said, I am a bit confused by the specifics of this building so I don't actually understand the lateral system well enough to be sure what approach I would take in analysing this building

 

[milkshakelake]

I don't think OP clearly explained if they're designing the elevator shaft or not.

If they are not designing it: The torsional forces and slab strength can be designed for.

If they are designing it: It's better to isolate the building because there's no way to account for the stiffness transfer and compatibility. I mean the interaction between your building and the exterior core designed by others. The exterior core could be overloaded, even if you give them the loading, because that interaction is affected by their design. I'd personally isolate it with a combination of pins and long slotted holes through detailing, using steel to carry that little slab.

Edit: Fixed a significant typo; reversed what I said

 

[BAretired]

I would not use a separation joint, and would design the Elevator tower to be flexible, at least in the north-south direction, so that it relies on the main building for lateral support in that direction.

 

[jdengos]

The main building has two shear wall towers.
The outside elevator tower is only column-beam frame with infill block walls. We didnt specially design the tower since we design the main building and the elevator separately.
I agree no expansion in this case since both are stiff with little drift.

But boss said even a flexible building can use the outside elevator tower without expansion. Only case we need expansion is when we have large drift PT building has a stiff tower. I dont quite understand how the rigid tower can collapse a big PT slab building.

 

[KootK]

If they are not designing it: It's better to isolate the building because there's no way to account for the stiffness transfer and compatibility.

I feel that this would be pretty easy to handle. "Hey shaft dude. Plan your shaft thing to be able to ride along with my building for the 8 mm drift that I expect at the top of this adjacent stair shaft."

Functional movement joints are expensive in enclosed buildings. You'd need a movement joint at the roof, all of the floors, and two, multistory wall joints. And the presence of the movement joints introduces durability and maintenance risks. Lastly, if the structures are to be isolated, then they will have to be separated enough to prevent seismic pounding.

I feel that the costs and risks associated with movement joints should only be incurred where one really expects differential movement. And this isn't that in my view.

 

[Greenalleycat]

Jdengos, the engineering principle is clear - eccentric stiffness influences the building response and can cause increased displacements, larger loads on lateral elements, and potentially high local forces such as an in diaphragms and wall-floor connections
Any one of those aspects, if not considered properly, could lead to poor structural performance, with the very worst case outcomes being the initiation of a collapse mechanism
I do agree that full collapse of a building is unlikely though - or at least, full collapse of this building because, as Kootk said, it has good structural systems elsewhere

The CTV building I referenced earlier is a very high profile failure in my country that suffered from all of these issues (and more)
It was a crappily designed and badly built multi-storey building that relied on an eccentric shear core to provide lateral stability
The high eccentricity of the shear core meant that the outside 'gravity frames' suffered very large displacements and ended up failing in shear
The diaphragm was also terribly connected to the walls and the diaphragm connections subsequently failed

Do be careful with infill block walls though, they will add massively to your stiffness in that external core unless the infills are isolated from the surrounding frames
Effectively that structure is a rigid block wall structure not a frame structure if all bays are masonry infilled
The issue that this building could have is that you have a rigid infill structure off to one side connected by a small section of diaphragm
To maintain displacement compatibility your building needs to push large amounts of load through that diaphragm and into the core - this will influence the building response to be more torsional than it otherwise would be
This could also lead to failure of that section of the floor due to the large loads going through it

 

[milkshakelake]

I dont quite understand how the rigid tower can collapse a big PT slab building.

A theoretical super stiff elevator tower wouldn't collapse a building. It would take unintentional lateral forces that it may not be designed for. If it survives, it could break the connecting slab. But even in this theoretical scenario, it could all be designed, though it might not be economical.

Edit: KootK agreed, that makes sense. Keep it flexible and add bracing or moment stiffness in the left-right direction.

 

[KootK]

Only case we need expansion is when we have large drift PT building has a stiff tower. I dont quite understand how the rigid tower can collapse a big PT slab building.

It wouldn't collapse the PT building. If proper planning measures are not taken this might happen:

1) The PT would cause the ends of the main building to draw towards one another.

2) The stiff elevator shaft -- if it were stiff -- would try to resist this movement and, possibly, be damaged in the process.

But boss said even a flexible building can use the outside elevator tower without expansion.

It sounds as though you need to have your boss explain / qualify that statement. If the elevator tower is made very flexible relative to the main building, then that statement is probably true-ish. However, unless your CMU infill is detailed to allow movement, the elevator shaft will probably be quite stiff.

 

[jdengos]

@KootK, I am so thankful and feel you ever replied every of my posts!
The main building has symmetric stair towers each end, but once connected to the outside elevator tower, will things change? We checked both individual buildings are very rigid with little drift, so I can bear with no expansion. But still worried it will change the lateral system once they are combined since outside elevator tower and the little connection slab between is not designed with any lateral force/torsion from main building.

 

[Greenalleycat]

Jdengos, you still seem to be missing the point. It is up to you to assess and analyse whether this will cause a change in the performance of the structure, not us
We have all repeatedly been in agreement that this -could- happen and may even be -likely- due to the block infill in the stairwell..but we are not the ones doing the design

You say you know the performance of the building neglecting that external core
Great, run a comparative model that incorporates the external core and see how the performance changes
Alternatively, build a separate model that accurately reflects the full structure (including masonry infill) of your infilled external core and apply the floor displacements to it at each level
Then see how much force materialises in that structure - how does that compare to the main structure?
That won't tell you the torsional influence but at least will give you a feel for the comparative stiffnesses of the two systems and what the magnitude of loads through the floor are likely to be

Torsion can be a funny beast. I once did a concept design for a 15 storey building that stepped in at the 8th and 11th floors
I had a very long 45 degree wall going from basement to the 8th storey - it was the biggest wall I had so I wanted it in there
However, my building was highly torsional and I had big displacements at floor edges, I couldn't make it work no matter what I did
So I deleted the 45 degree wall in despair to see if anything happened and suddenly my building had beautiful translational displacements, the torsion disappeared, and the displacements reduced to within acceptable limits
Completely counter intuitive but real nonetheless

So, sometimes you just have to do the numbers, there's only so much we can share on the theory

 

[KootK]

@KootK, I am so thankful and feel you ever replied every of my posts!

You're most welcome jdengos. It takes a village.

The main building has symmetric stair towers each end, but once connected to the outside elevator tower, will things change? We checked both individual buildings are very rigid with little drift.

Things will certainly change. The only question is whether or not they will change to an extent that is significant. Here, I doubt it. The reason that I feel this way is shown below. The building will posses an absolutely monstrous ability to resist torsion as a result of the large lever arm between the two stair cores. This torsional resistance should dwarf any undesirable eccentricity effects by including the elevator shaft in the structural system.

outside elevator tower and the little connection slab between is not designed with any lateral force/torsion from main building.

So design it for one of the following:

1) The forces that would be generated by the two structures remaining contiguous OR;

2) The displacement capacity that would be required of the elevator shaft for it to ride along with the main building.

Your design assumptions need to be consistent with your detailing assumptions. If you mean for your analysis of the elevator shaft remain independent of the building as a whole, then you probably do need to isolate the structures. If you're going to tie the structures together structurally, then I feel that you ought to do some proper accounting of the implications of that.