Wind Tunnel Analysis

[SKJ25POL]

May I ask in what cases we use wind tunnel analysis? For what structures?
What is the benifit of wind tunnel analysis?

Thank you

 

[glass99]

Code analysis of wind pressures is usually wrong either on the high side or low side by quite a bit. Its necessarily based on gross simplifications. Wind tunnel studies are most often done on tall buildings where the difference matters enough to pay for the study. Anything taller than probably 30 stories is a big enough project with high enough wind pressures at the top to make it count. I have also done a wind tunnel study on a flexible canopy structure. Anything with low frequencies and large torsion deflections should be analyzed.

 

[JStephen]

A couple of years ago, for continuing education, I took a 2-day seminar on ASCE 7. I found it to be very interesting and informative and recommend it, looks like it's coming up again this fall:

http://mylearning.asce.org/diweb/catalog/item/eid/200400623

Anyway, if I remember right, it was mentioned in the seminar that the cost of a wind-tunnel analysis was something on the order of $50,000 and up, if that helps put it into perspective.

 

[KootK]

Depending on what was included, I've had reports cost in the range of 20-80k.

Wind studies are sometimes needed even on shorter buildings for the design of cladding attachments for two reasons:

1) At street level, municipalities sometimes worry about localized high winds due to the aerodynamic interplay of the new building with the adjacent ones.

2) For buildings characterized by wacky geometry, there is a perceived need for better localized wind estimates.

When testing is done for these reasons, you can often get primary structure numbers added on for 5-10k which changes the cost benefit calculation some.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[glass99]

As a "political" point: we should all be a bit less timid about asking for wind tunnel studies. Code analysis drives me crazy because it is such a black box method which disconnects us from physical reality over the long term. We get bogged down in idiotic discussions about which GCp factor we should be referencing, but no one really understands the aerodynamics of their building.

 

[KootK]

@glass99: I understand your perspective but I quite disagree. I think we'd be better off with conservative, easy to calculate code wind loads. I'd hate to have to bother with wind load studies for routine projects.

When I first started out in WI, it was 20 psf to 25 ft, 25 psf to 40 ft etc. I've probably botched the numbers but the point is that it was awesome to work with. And, while wind studies certainly improve our ability to "know" our loads, we still don't really know.

I've worked on a few projects where the wind studies produced extremely low net wind pressures for MWFRS design (7 psf). At first blush, that sounds like a wonderful thing. However, the economy was offset by two factors:

1) Seismic loads.
2) I refused to design the building for a MWFRS wind pressure less than 15 psf. Most structural engineers will have a "floor" value no matter what the wind consultant comes up with.

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[glass99]

KootK: I agree that simplicity is nice. The old NYC Building code had a blanket allowance for all components and cladding of 30psf. It allows you to put your energy into something more productive like an elegant structural design rather than faffing about with stupid formulas. At the other end of the scale, I can't do all kinds of interesting things because of the supposed uniform suction that develops on the back side of tall buildings. This is a multiplier affect with the excessively conservative insulated glass deflection requirements, so my $300/sft storefront becomes $500/sft and gets value engineered out. Meanwhile, the Empire State Building was designed to 20psf even at the top of the building, and has survived a couple of hurricanes in the last few years.

 

[KootK]

Faffing about with stupid formulas. I'm filing that for future reference. I certainly take your point with regard to the economics aggressively designed cladding elements.

I have a 1920-ish book called "Wind Bracing" written by a NY structural engineer. It's glorious. The subject matter paraphrased: maybe we should have wind bracing on tall buildings?

The greatest trick that bond stress ever pulled was convincing the world it didn't exist.

 

[glass99]

Back in the day they didn't worry about lateral loads so much. Earthquakes had an allowance of zero. Kind of whacky of them.

My larger point though, is that there is no substitute for first principles engineering in terms of driving innovative design forward, and you can't do that without checking yourself through testing. Wind tunnel testing in particular is under specified for lightweight structures like canopies and long span roofs. Structural testing of glass to metal connections is something you learn a lot from, as it is with unorthodox materials like metal mesh and funky geometries. We should not be so afraid to tell owners to front for testing.

 

[SKJ25POL]

KootK (Structural), glass99 (Structural) please do not use this forum for your conversation ONE TO ONE conversation.
I apreciate practical information on Wind Tunnel Analysis for a new guy in this subject.

Thank you everybody.

 

[glass99]

SKJ: um, we were publicly discussing the question you asked. The tl;dr version: folks do wind tunnel studies for idiosyncratic reasons which include politics and building size, and the code is ridiculous sometimes. Maybe you were expecting some more straightforward answer, but that's how it goes in the real world.