Wind loads on rooftop equipment 2

[JAE]

A small debate in our office. We have two 12' x 24' long (in plan view) condensers sitting above a roof of a large building. The condensers are 16 feet tall. They will be set parallel to one another such that the 24 ft lengths will align.

The two 12 ft. widths will be separated by a 4 foot gap.

The units will sit on a steel framework of beams supported by columns extending through the roof of the building.

We are using ASCE 7-05 for the wind load on roof-top equipment. In ASCE 7 it has a section titled "shielding" that states that for buildings and other structures you cannot use one element to shield another.

The debate is whether, in this case with two large condensers sitting only 4 feet apart, does the "projected area of the units" mean we have to put the wind load on both of these or does the small width of the gap (4 feet) mean that in essence we have "one" unit and receive wind on one face only?

See the attached sketch.

 

[WillisV]

For the design of the building structure below, engineering judgment would seem to indicate you would treat it as one unit, ignoring the gap. In the past when I have discussed similar situations with "wind experts" the consensus seems to be that if the gap between elements is smaller than the 1 to 2 times the height of the elements the wind will not have a chance to re-attach to the surface and speed up to the degree necessary to provide significant wind pressure to the second obstruction.

 

[slickdeals]

JAE: I thought that ASCE did not allow for shielding effect because of the uncertainty of whether a building/structure would be permanently there to provide shielding during the lifetime of either structure.

In your case, based on the aspect ratio of the gap, it might act as a single unit.

I am curious to see the responses.

 

[JAE]

slickdeals - yes - for buildings that is a consideration - that the "front" building might be torn down at one point, leaving the rear building to take more wind than it was originally designed for.

However, in this case, if one of the units is taken down, the framing that supports that unit would not see any increase in wind area.

The ASCE 7 commentary states that no research exists that would tell us what actually happens between physical elements like this.

Per WillisV, there could be some factor of gap width to wind pressure - if you take it to ridiculous extremes:

1" gap - I'm certain that the rear unit sees no windward wind.

50 ft gap - I'm certain that the rear unit sees 100% wind force.

In between gap - ??

 

[msquared48]

If you want to treat them as one unit with no questions, just modify the design and enclose the 4 foot space with removeable light gage metal framing.

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask

 

[a2mfk]

I'd check the framing for both units in place, wind hitting one side or the other, assume that they act as one unit for wind purposes. Then do a quick check if either were removed. Depending on the weight of each unit, area, etc., it is not always super obvious which will control.

As to what gap would I be comfy with before I started worrying about it personally, in absence of anything from code or testing? About 10ft, gut feeling, FWIW

 

[Ron]

JAE...assuming that the framing is independent for each unit, though only 4 feet apart, if you design one for the exposure then the other meets the requirement, assuming you duplicate the framing.

If shared framing...who knows? Engineering judgment prevails so just document what you decide, why you decided it, and move on.

 

[JAE]

Ron - shared framing - same "platform" so to speak.

We have, for the time being, settled on 100% wind on the windward unit and 50% wind on the shielded, leeward unit.

Still open to comments - just wondering if anyone else deals with or has dealt with this.

(Mike - we can't close off the gap - condensers need minimum space between for free air flow and the client needs free access)

 

[BAretired]

JAE,
I would agree with your proposal, although I think it is likely a bit conservative.

BA

 

[dhengr]

JAE:
I’d use a projected area of 16'x24' and apply + & - pressures to a cube shape which was 28' long in the wind’s direction, with appropriate near edge pressure increases. Then for good measure I would take those code pressures and apply them to the faces perpendicular to the wind in the 4' gap, both + & -, but only on a strip around the edge of the gap, say (4' gap width)/2 = 2' wide at the top and the two sides, but maybe the bottom if wind can get under your support frame, thus you have a downward open “U” shaped loaded areas on the gap faces, or maybe a narrow four sided loading area.

If the condition of one of the units being removed is real, I agree that the wind loading won’t change by much, but you better check to see that you don’t literally have uplift on the unloaded legs under some wind load orientations. I think that is what a2mfk was suggesting.

I don’t know that my approach is absolutely correct or not, but these damn codes these days and our general changing approach to engineering inhibits the use of any engineering judgement or common sense in the engineering process. We have more questions than answers if the code doesn’t cover the condition explicitly. Does every condition need a wind tunnel test? Or are we allowed to use reasonable engineering judgement and experience at times, or will the codes ultimately cover every detail under every possible condition?

 

[southard2]

Perhaps I'm wrong, but if your are using ASCE 7-05 for example the section on roof top units would have you apply the full wind load to one side of the unit. Since wind can blow in two opposite directions each of those units would get the full load no matter what. There is no leeward and windward component of the total wind pressure. So both unit will have to designed for the full horizontal forces and each anchored adequately.



John Southard, M.S., P.E.

http://www.pdhlibrary.com

 

[SAIL3]

Since it is difficult to get a handle on how much effective shielding is going on, I would tend to go with what you suggested...100% on windward unit and 50% on leeward unit.
This partially addresses other questions ie.,
what happens with wind 10 to 20 deg off-perpendicular to the 24'side?.
Is there any wind tunnel effect with wind paralel to the 24' side?

 

[SAIL3]

There is an excellent and practical book published by ASCE that addresses various shielding effects of open frame structures, pipe racks, vessels, access, etc.

ISBN 0-7844-0262-0

"Wind Loads and Anchor Bolt Design for Petrochemical Facilities"
1997

 

[SteveGregory]

Southard is correct. The anchorage for each unit should be designed for the full wind load prescribed in Section 6.5.15.

And don't forget the 1.9X multiplier in 6.5.15.1.

 

[dik]

For the overall load on the structure, I would ignore the gap and treat it as a single larger unit with an equivalent to an internal pressure on the internal face of each unit (likely covered by regular wind pressure assuming a single smaller unit)... The units are too close together.

Attachment would be predicated by treating them as individual smaller units.

Dik

 

[JAE]

Southard, and SteveGregory - thanks for your help but I think I didn't communicate the question clear enough. I'm not concerned here with the units themselves or their connections to the supporting frame....I do not design the units nor do I specify the connection to my framework below.

The framework below is the issue and the other comments above are very helpful - thanks all.

Other comments are still welcome. SAIL3 - I'll check out that book.

We will check the case with one unit on the near or far location alone.

 

[southard2]

Jae,

One thing that concerns me abit about the ASCE code and roof top equipment is how we define roof top equipment. Some of these roof top units aproach the sizes of small buildings and in some cases are larger than stair well towers. It would be nice if at some point in the future ASCE addresses some kind of size limitation. At what point do you say OK this is part of the building and apply the traditional leeward and windward wind loads. That said with the 1.9 multiplier noted above the roof top unit loading in many cases might be more conservative than the MRWRS calcs.

And yeah I was totally thinking about the unit and its connections and not the framing below.

John Southard, M.S., P.E.

http://www.pdhlibrary.com

 

[hetgen]

Venturing in to other codes … 0.30 reduction factor should be appropriate.

When there are other walls or fences upwind that are equal in height or taller than the wall or fence height h an additional shelter factor can be used with the net pressure coefficients for boundary walls and lattice fences. The degree of the shelter depends on the spacing between the walls or fences and the solidity of the upwind (sheltering) wall or fence.. Shelter remains significant up to spacings of 20 wall heights. At very close spacings the net pressure coefficient on the downwind (sheltered) wall can be zero or can reverse in sign. A minimum limit to the shelter factor of 0.3 has been set to cover this case safely.

See attached Values of shelter factor to reduce the net pressure coefficient

 

[dik]

hetgen:

interesting... do you have the other 26 charts or can you quote a source?

Dik

 

[slickdeals]

I believe it is from the British Code BS 6399. I don't want to cite sources, but you can freely download them from one of those file sharing sites.