Design of a facade system to accommodate building seismic movements

[jsharkbait]

Hi guys,

We're doing a 12 story commercial building with a Seismic Design Category of B, and have recently been questioned on the building movements under seismic loads so that they can test the window wall system they are using (glass and terracotta).

We gave them the anticipated building drifts from our etabs model for wind drift and ultimate seismic models, but they are baulking at the seisimic drift, which we're saying is 1.6" for the typical floors- well within the H/50 limit in ASCE7. We've also indicated that the code says "architectural" components do not need to be designed for seismic forces in SDC B- See 13.1.4 in ASCE7-10.

I don't have much experience in facades, so I don't know what they are normally designed for. I know the whole point of the drift limit is to make sure there they don't fail and there isn't facade raining all over people who are trying to exit a building under a seismic event- so I feel we've only caused more confusion by giving them a large drift, but also saying they can just ignore it (perhaps we as engineers are used to seeing such apparent contradictions but others aren't).

Just wanted to get a feel for how this is normally handed by others. Thank you.

 

[r13]

You should asking then to provide the specific parameters they are looking for, which can be beyond the scope of typical structural engineering service. Do you have architect on this project, as he might have clues.

 

[jsharkbait]

It was the architect who was looking for the drifts in the first place, along with the facade engineer and manufacturer. They have some criteria of their own that they apparently typically use, but aren't sure where those numbers have come from. Those numbers are more stricter and wouldn't correlate with our expected movements.

 

[r13]

Yes. I am not surprised that they are looking for specific parameters usually overlooked in the structural engineering, but have immense importance in their trade. Asking for the list, and reference materials, I have no doubt there are something you can provide, or come up a good estimate. Even have to say "sorry, nothing to offer" at the end, you would benefit from knowing what are the concerns. Don't be shy, nobody knows everything, especially things not directly tied to our training/practices.

 

[KootK]

1) Where is your project? NJ? I'll assume so for now.

Those numbers are more stricter and wouldn't correlate with our expected movements.

2) Can you share what those number are? I'm pretty curious myself.

3) In my experience, in western Canada, accommodating a story drift of 1.6" will be difficult for many conventional systems, especially in a market like NJ that probably isn't especially used to having to deal with this kind of thing. I find that story drifts in excess of 0.5" - 0.75" will get me into trouble in my area.

4) In low seismic environments, you often get let off of the hook somewhat because normal design practices for wind (~h/400) preclude getting anywhere close to (H/50) kinds of numbers. This must not be true for your project, though, as you're, what... H/75?

5) When seismic performance governs, it becomes important to make sure that everybody has a clear understanding of what constitutes "failure". Failure probably is not glass breakage and, as you've implied, is more about keeping thing from flying out on to the street below. I'm no expert on this part of the story, however, so hopefully others will chime in. The book below has been very useful to me in working through these kinds of issues. Here's related paper by the author: Link.

6) I suspect that your code "out" is the way forward here.

 

[KootK]

For what it's worth, a common specification that I see in seismic jurisdictions is:

a) No glass breakage at 1% drift.

b) No glass fallout at 2% drift.

 

[r13]

I think they want/need story drift rather than global.

 

[Agent666]

That's what the 1.6" is if I'm not mistaken, the interstorey drift.

I'd say its common practice to allow for separation to allow the drift to occur without imposing loads on the facade from the building movement. Obviously facade itself has to deal with its own inertia loads or wind loads under earthquake or wind respectively. Although it does depend on how rigid the facade elements are. Timber framing with linings maybe able to handle considerable drift without failure if fixed to each level above and below. On the other end of the spectrum, fixing a precast concrete cladding panel between levels with no means of accommodating the drift is asking for the panels to depart the building as they participate in the lateral force resisting systems response.

Or you design a facade to accommodate any building movement and the induced loads that come from this. This is inherently much harder to do as the drifts you work out are an estimate at best.

 

[Agent666]

Keep in mind at an ultimate limit state event, damage mitigation isn't usually a design objective. People/code expect there to be damage, what they don't expect is facades landing on the street and munching people. This might be violating the life safety objectives at the ultimate limit state.

Fixings and means of dealing with any deformation is key to making sure the facade can achieve the desired level of performance.

Determining the clients expectations or industry best practice around expected damage following a design level event or at least communicating possibilities to them is a good starting point. It's their risk if event occurs and building cannot be occupied for any reason (weather tightness for example).

 

[mronlinetutor]

I know the requirement for elevator lift is H/1000 for high rise projects. That would translate to 1000' to 1". You did not mention that what the material of the front facade would be. I am afraid of glass. I did a one-story commercial concrete building with parking below and above the roof. I use brick laid on a reinforced concrete panel. Do not use brick because brick and cement do not have any tension at all. My final year project is on the tension between the brick and cement and I find that brick and cement literally do not have any tension.

disclaimer: all calculations and comments must be checked by senior engineers before they are taken to be acceptable.

 

[Agent666]

1000" to 1" I think you mean for 1:1000. ... That would tend to be the serviceability requirement I would have thought,not for ultimate design.

 

[jsharkbait]

Thanks for your replies.

KootK:

1) The project is in NYC, and the (2) facade engineers have typically seen seismic movements at H/267- we asked for clarification on their number and it seems they dig some digging but weren't able to come up with too much- it's just something they've used for a long time and have't been questioned on it much.

3) This is where I'm confused. I would have thought that in a higher seismic area designing for such drifts would have been commonplace and the joint systems would be much better thought out, as I'm assuming drifts would be much higher on the regular.

4) Because we design for H/400 wind service those numbers are acceptable, but the higher seismic drift at ultimate levels is killing us. We're at about H/90. So I don't see that they could be designing for no glass fallout at 2% drift if they can't stomach the 1.6". I would have thought they'd need the whole joint to take up that movement and not load up the glass as Agent666 suggested.

 

[KootK]

There's no question that some markets are more highly evolved than others when it comes to accommodating drift in the building skin. Certain products are better suited to it than others too. Nothing that I'd think of as common "window wall" would get top hung and permit large, free, in plane translation at the bottom. In my area, a window wall is framed much like the platform framed stud wall of my house and the drift accommodation is via 1/2" movement plus the local crushing of something behind that.

 

[KootK]

1/2 x 267 = 11'... about what I'm used to for window wall in the upper height range for residential and hospitality.