General Discussion on Skyscraper Drift 1

[IlliniPE]

The conversation started in the office about skyscrapers being designed to sway in the wind, some argued to reduce wind pressure or the classic "if it doesn't bend, it will break" argument. That argument seems more of a ductility conversation rather than stating that buildings are designed to sway as in intended to sway. I argue that they are "allowed" to sway, but not necessarily encouraged to. I understand this is largely a semantics issue, but, in my mind, all buildings would be designed to be rigid so long as they resist the loads and cost not being a factor. I was curious of everyone else's thoughts on this. Are buildings "supposed" to sway, or are they just "allowed" to?

 

[3DDave]

They are designed to resist the loads applied to them and to limit deflection to acceptable amounts. Swaying does not relieve loads in buildings, but making them stiff enough not to sway can be extremely expensive. A popular control is the use of a tuned mass that counters swaying and converts the motion into heat.

This is an excellent video on the small scale swaying of traffic light poles:

https://www.youtube.com/watch?v=oYW_3R1yN8Y

If the swaying is in time with the natural frequency, it can be quite destructive.

 

[XR250]

ome argued to reduce wind pressure or the classic "if it doesn't bend, it will break" argument.

I think they are getting buildings confused with trees - where this actually may apply.

 

[IlliniPE]

If the swaying is in time with the natural frequency, it can be quite destructive.

I completely forgot about this factor, as well. They had me second guessing myself, but I agree that zero sway would be best, money not a factor of course.

I think they are getting buildings confused with trees

I think they are just speaking to ductility without really realizing what they're saying. IOW, brittle vs ductile failure.

 

[WesternJeb]

What is the end goal of the conversation / debate? Will this change design philosophy? IBC lays out deflection limits and there are some acceleration limits out there as well used as rules for general design.

 

[IlliniPE]

What is the end goal of the conversation / debate?

The goal is to make sure I am not intended to include sway/drift in design purposefully, but rather restrict it as much as possible within allowable limits. It's a "supposed" to do it vs "allowed" to do it discussion. It's similar to how the FDA allows bugs in food, but they aren't supposed to be there purposefully.

 

[KootK]

In certain seismic applications you are supposed to have ductility and therefore, usually, inelastic drift capacity. This is sensible for a dynamic load.

In wind design, you are allowed to have drift. And that's appropriate for a load that we treat psuedo statically in many cases.

My understanding is that the wind design of very tall buildings is mostly governed by occupancy comfort and, therefore, the frequency of the drift more so than the magnitude of it. In this regard, performance based design for wind seems to be gaining traction as it has for seismic design.

I've been to the top of the John Hancock building on a windy day and had a chance to survey the motions of the surrounding skyline. Those buildings are moving significantly, upwards of a meter I'm pretty sure. However, because the period of the vibration is kept within limits, they are still successful structures from the perspective of their occupants.

 

[IlliniPE]

In certain seismic applications you are supposed to have ductility and therefore, usually, inelastic drift capacity.

I can definitely agree with this, but I don't take that to mean we "want" it to drift. Just that we accommodate the inevitable.

Those buildings are moving significantly, upwards of a meter I'm pretty sure. However, because the period of the vibration is kept within limits, they are still successful structures from the perspective of their occupants.

Period is the big one to consider, as it can quickly turn a successful structure into a pile of rubble. It still surprises me that they are allowed to drift to any noticeable amount. Structurally, fine, but I have to imagine some motion sickness would set in for many.

 

[ggcdn]

In the 'medium-tall' design space (45-60 stories) wind design basically amounts to making the LFRS stiff enough to limit drifts (H/500 for instance), strong enough for the equivalent static loads, and then limiting accelerations to levels deemed acceptable for the occupancy. This is usually done in conjunction with wind tunnel testing. In many towers, the accelerations are governed by vortex shedding causing movement perpendicular to the main wind direction. The mitigation strategies include changing building shape & stiffness to break up the vortex shedding (rarely done), adding a sloshing tank (common), tuned mass damper (less common), or other source of supplemental damping (viscoelastic headers, etc).

Typically, equivalent wind demands increase with building period, and likewise it is harder to meet drift limits.

 

[driftLimiter]

The drift is supposed to happen (because any deformable material resisting loads deforms). If you are increasing the drift "on purpose" then you are probably looking at the dynamics of the structure and perceived accelerations, or perhaps the spectral acceleration (a longer period building get a smaller seismic load than a short period structure).

 

[bugbus]

I’ve noticed there is a misconception popular among the public that a tall building ‘needs to be’ flexible enough to survive an earthquake. You see this in a lot of YouTube comments on videos where tall buildings are noticeably swaying during earthquakes. “If it didn’t sway, it would collapse.” I don’t know where this idea came from, but it’s pretty pervasive what from I’ve seen, although probably not entirely without truth. A more flexible structure will normally see less base shear than a stiffer structure, but it is not necessarily a ‘requirement’.

 

[LittleInch]

A more flexible structure will normally see less base shear than a stiffer structure,

There's your answer. Stiffer building, both in normal wind and earthquake needs stronger foundations / stronger connection between foundations and structure. Also a stronger building is heavier and probably costs more.

This is therefore like a lot of engineering, getting an acceptable compromise between conflicting issues.

As flexible as a tree and people living at the top would get injured by hitting walls and flying internal objects. Super stiff and no one could afford the appartments.

So getting back to the OP I would say buildings, especially tall buildings, are "supposed to sway within allowable limits".

How's that for an engineering fudge??

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[cliff234]

IlliniPE: If an engineer makes the building too stiff, the structure may not be built (because it costs too much) and they may be fired. It costs money to increase stiffness. More stiffness costs more money. The structure must obviously be designed for strength to withstand the code-mandated wind speeds (so the building does not blow over in a storm). But lower wind speeds can be used to check building drift. Drift is a serviceability check, not a strength check. Most engineers strive for a drift limit of about H/500 under a mean recurrence interval wind speed lower than the code mandated wind speed for strength. In addition to designing for strength and drift, designers must strive to limit acceleration in taller buildings so occupants do not get seasick due to the building swaying. That is usually the biggest challenge. In high rise buildings, it is often cross-wind acceleration that occupants feel(rather than drift). Cross-wind acceleration is acceleration perpendicular to the direction of the wind. An example of cross-wind acceleration can be demonstrated if you stir a cup of coffee using a plastic straw. Stir the coffee. When the coffee is swirling around fast, hold the straw still near the perimeter of the cup. You should feel the straw vibrating perpendicular to the flow of the coffee. That's cross-wind acceleration. Dampers are used in some high rise buildings to limit acceleration (not to limit drift).

 

[bookowski]

A more flexible structure will normally see less base shear than a stiffer structure

For seismic yes, for wind a more flexible structure sees higher wind load, i.e. longer period = higher wind. To control occupant comfort you theoretically want more stiffness and more mass.

When I was building my house and looking at windows in a showroom I was pushing on some thinking that they seemed flimsy. The salesman explained to me that this was intentional, because "if they were stronger they'd break".

 

[IlliniPE]

The salesman explained to me that this was intentional, because "if they were stronger they'd break".

You were sold on them weren't you? I'm not sure where this misconception started and took off from.

In high rise buildings, it is often cross-wind acceleration that occupants feel(rather than drift). Cross-wind acceleration is acceleration perpendicular to the direction of the wind.

I learned a new term today. Thank you. I do agree that stiffness comes at a cost. It was more of a question as to if we WANT the building to sway. I will have to do some more research on limiting acceleration (not that I'll be designing a skyscraper anytime soon).

A more flexible structure will normally see less base shear than a stiffer structure, but it is not necessarily a ‘requirement’.

This would largely be due to the increased weight of the structure though correct? More weight = more base shear from seismic. Wind should, theoretically, decrease with a more rigid structure because you're limiting your dead load eccentricity.

(because any deformable material resisting loads deforms). If you are increasing the drift "on purpose" then you are probably looking at the dynamics of the structure and perceived accelerations, or perhaps the spectral acceleration

Of course it is inevitable that it will sway, but the question is, to limit the sway to a minimum or purposefully cause it for design purposes. I do like your point about long vs short period. If I'm remembering correctly, your base shear is inversely proportional to period correct? So longer period would in fact result in a lower base shear. But do people rely on this longer period for seismic design, or just use a dampening system?

Typically, equivalent wind demands increase with building period, and likewise it is harder to meet drift limits.

Could you expand on this? Why would wind demands increase with building period? Would it be realized when the building sways "back" against the wind?

I greatly appreciate all the input from everyone!

 

[kissymoose]

In static loading, there's no difference in base shear. If you've got two buildings, one more stiff than the other, and the lateral static load is the same, the base shear is the same for both.
If anything, the flexible building will have more second order effects to account for.
In dynamic loading, stiffness matters, and a more flexible building in general can distribute an impulse load to more members than a more stiff building can.

Typically, equivalent wind demands increase with building period, and likewise it is harder to meet drift limits.

I haven't done wind design for tall structures before so I don't know, but are there dynamic code provisions for wind loads?

 

[Trenno]

for wind a more flexible structure sees higher wind load, i.e. longer period = higher wind

Not always. If you start off with a relatively flexible tower, sometimes you end up on the uphill side of the Stiffness v Wind Response curve - thus making the structure stiffer you end up with higher wind loads.

 

[cliff234]

IllinPE: You asked, Do we want a building to sway? The answer is, we have no choice. All beams deflect. All buildings sway. The key is to control the sway (i.e, lateral deflection0 and to control the acceleration.

 

[IlliniPE]

thus making the structure stiffer you end up with higher wind loads.

I've always known it to be longer period=more wind base shear. Is that not the case?

 

[IlliniPE]

You asked, Do we want a building to sway?

I asked it as, do we purposefully want them to sway. Of course it's inevitable that they will.
Do we want concrete to crack? No, but it's going to. I agree that we just seek to minimize sway.