Floor Plate in Lateral Stability Model 2

[SamETABS]

Dear All,

I'm doing a lateral stability study on a 40 story building. Should I model RC floor plates in the lateral stability model? I know that there would be a beneficial effect by including the floor plate on lateral sway of the tower. But, is it a logical thing to do?

many thanks.

 

[Lion06]

Are you modelling the entire building or just a single frame?

If you're modelling the whole building then you need it to distribute lateral loads. If you're only modelling a single frame then you don't need it.

What beneficial effect would it have on drift?

 

[SamETABS]

StructuralEIT,

Thanks for your post. Since, the plates are modelled as shell elements, their presence in 3D model will enhance the lateral stiffness of the tower. I have set up two models:

i)frames and shearwalls (i.e. w/o any fllor plate), which are part of the lateral stability system

ii)the whole building

the lateral displacement that I'm getting in the second model is considerably less than the first model.

is this rational?
thanks.

 

[Lion06]

How are you distributing lateral loads in the first model?

 

[SamETABS]

StructuralEIT,

I apply the lateral loads on Diaphragm CM, either wind or seismic. I've also attached a section from "Reinforced Concrete Design of Tall Buildings by B. Taranath" that emphasizes on modelling the floor plate!!

 

[Lion06]

That's only talking about rigid or flexible diaphragms. If your first model has a diaphragm, I'm not sure what adding a plate element as a second diaphragm is doing or why it's reducing the drift. What is the drift for each model?

 

[Lion06]

One other thing to think about - maximum drift is usually located at a corner under a torsional lateral case. If you're looking at a straight up wind case then you're drift is probably showing much lower than it should be.

Are you doing a wind tunnel study?

What kind of natural frequency do you have? I'm working on a 30 story building right now and we are kicking butt on drift (on the order of h/700 for a 50 yer wind), but are just squeaking by for accelerations at the top occupied story for a 10 year wind. Part of that is because we are using steel so we have a light structure and a smaller damping ratio. I'm just saying that you can work for drift but still fail accelerations.

 

[WillisV]

SamETABS,

Inclusion of the out of plane (flexural stiffness) of the floor plate will definitely stiffen the building up due to the outrigger action that results from coupling the shearwalls with the perimeter columns. If the floor plate was designed for the bending moments induced from this coupling action, then sure you can use it to check drift. If the floor plate was designed independently of the lateral loads (i.e. a gravity only floor plate design), then you should not include its beneficial effects for drift as the slab would crack under lateral loads since it is not reinforced to resist them.

Which way to design (as you asked which one is logical) is a matter of engineering judgment. By ignoring the slab flexural stiffness in the lateral model the design of the floors is a bit easier since you can do them independently, and the rebar will be a bit less. But the slabs are there, so with a bit more engineering effort you can use them to your benefit.

 

[Lion06]

Willisv-

What do you mean by outrigger action and how is modelling a "plate element" different than designating it as a rigid diaphragm?

 

[hokie66]

SEIT,

Perhaps I can help before Willis comes back to you. "Outrigger" is typical tall building speak for using the out of plane bending capacity of the floor systems to distribute loading, which otherwise would be taken by the cores, out to the perimeter. It usually involves stiff beams or floor to floor walls or trusses to make the columns work in push-pull like truss chords, but in some cases, some benefit can be shown just by using the slab bending capacity. I am not much in favor of using slabs as outriggers, but certainly stiff elements do a good job of reducing core requirements.

A plate element and a diaphragm have the same meaning to me as horizontal distribution members.

 

[Lion06]

Does anyone have an article or can point me to something I can read on the subject? I'm not seeing the difference between a diaphragm for a 4 story building and a diaphragm for a 40 story building in terms of limiting drift, but it does sound like something I should read up on.

Presumably this would help to raise the natural frequency of the structure as well, since it is stiffening it, right?

 

[WillisV]

Thanks hokie66 for helping out while I was sleeping on the job! Yes by outrigger action I meant what hokie said, using the out of plane bending capacity of the floor system to stiffen up the building. For a standard flate plate design with internal shearwalls and perimeter columns you can probably reduct drifts by 10% or so (rough numbers) with just the floor slabs.

For taller buildings it is very common to have one or two levels of deep outrigger beams (generally full story-height beams at the mechanical levels) connecting the center core to the perimeter columns. This helps widen the structural base of the building (same as it is harder to push a person over if they spread their legs out to widen the base of resistance vs. putting their feet together) to reduce overturning moments at the foundation and reduce drifts. It also reduces the natural frequency (helps with accelerations).

The difference betweeen a 4 and a 40 story building in this regard is that for a 4 story building with the same type force resisting system (shearwalls) the predominant resistance is from shear. For 40-story+ buildings this mode moves to flexural resistance of the core, so by engaging the perimiter columns you are providing a greater depth of flexural resistance for your cantilevered system, thus reducing deflections etc.

Other methods of helping tall building drift include the use of a hat outrigger system (basically outriggers at the roof) or using a virtual outrigger system by providing a belt of trusses or shearwalls around the perimeter of the building, which then engages the stiffness of the floor diaphragms in shear, though this is not as efficient as standard core to perimeter outriggers.

You can read some more about outriggers and virtual (belt) outriggers in attached file.

 

[WillisV]

What do you mean by outrigger action and how is modelling a "plate element" different than designating it as a rigid diaphragm?

Hope Hokie's answer and my follow up above answered this - we are talking about the out of plane flexural resistance of the slab here - designating the slab as a rigid diaphragm for in-plane resistance has nothing to do with its out of plane flexural resistance.

Using a plate element in conjunction with a rigid diaphragm constraint will work for modeling out of plane stiffness and handling in plane load distribution, as will using a shell element and leaving the diaphragms as semi-rigid in-plane (you can use rigid constraint constraint with shells too - but then no reason to use shells in the first place as opposed to plates).

 

[hokie66]

This article has decent generic descriptions of the various systems used in high rise buildings of past and present. Outrigger buildings are included, but the reference Willis gave goes into more detail on those.

http://sydney.edu.au/architecture/d...Structural Developments in Tall Buildings.pdf

 

[SamETABS]

WillisV,
Thanks for your response. I was actually wanted to point out this phenomenon at the beginning of my thread, that when you define floor plates, it starts to engage peripheral columns as outrigger action and stiffening up the building.

1) I agree with your approach as to design the floor plate for the moments induced by the lateral loads

2) In reality there would be a moment connection between vertical elements and floor plates which transfers some moments; How do you justify the approach of exclusion of floor plates and designing them just under gravity loads.

 

[SamETABS]

StructuralEIT,

Just to add to valuable posts by WillisV and hokie66, if you build two models; one with floor plates included and the other one w/o floor plates; you will notice from reading the axial force in the peripheral columns that in the first model (i.e. floor plates included)you will get higher axial loads in the peripheral columns under lateral load cases, due to outrigger action.

 

[WillisV]

SamETABS - regarding your point 2) - you can justify it because the slab will crack under large lateral loads and revert back to the way it was designed (as if no moment connection). It is advisable to use many smaller rebars in the slab in this condition to help mitigate crack propagation.

 

[Lion06]

ok, that makes a lot more sense now. We are doing something similar - we have two braced frames separated by a 9' corridor that can't have a brace through it so we connected the frames with a stiff beam moment connected at each end. This helped these two frames act more like a single frame with a much larger "d". Do you gain that much help with just a floor plate? We needed pretty stiff beams to get that to work out.

Also, I don't know how it is for other software, but for RAM you would have to designate all fo the exterior columns as "lateral" columns.

 

[slickdeals]

I think RAM has the gravity/lateral option. But there is no such distinction in ETABS/SAP. In case you don't want a particular column to participate in the lateral system, then you will have to assign moment releases to make sure it does not participate.

 

[hokie66]

SEIT,
My opinion is that a flat plate does not provide enough assistance to make up for the additional reinforcement required, and if outriggers are used, they should be major elements, with floor height trusses or walls being the most beneficial.

Connecting the braced frames with a stiff beam (I assume you mean steel) is similar to using concrete header beams to make two 'C' shaped lift cores work together.