ASCE 7-10 Drift on Adjacent Structures 1

[RWW0002]

I am looking through the ASCE 7-10 requirements for leeward snow drift on adjacent (lower) structures (horizontal separation s less than 20'), and it appears there is a significant difference in the approach from ASCE7-05 to 7-10. To summarize, per 7-05 drifts were applied to all adjacent structures (within 20') with lower roofs, but this drift could be reduced by the factor (20-s)/20 where s is the separation distance. Per ASCE 7-10, buildings within the aerodynamic shade region receive drift with some limitations on drift height for buildings relatively close in height.

This all appears to make sense, but can have a very large impact for buildings nearing the 20' limit. Basically, based on 7-10, for a building with 19.5' separation from a taller building (h>s/6) the roof must be designed for a full leeward drift from the fetch of the taller structure. This is not reduced based on the separation distance. The same building with a 20.5' separation does not have to designed for any drifting snow from the taller building.

Per ASCE7-05 the drift would have been reduced based on the size of the separation to a very small percentage of the full leeward drift.

Making things even more confusing is that if the low roof continues over to the the high roof (no separation) the same portion of roof we were referring to does not have to be designed for any drift.

See attached sketches for clarity.

Am I missing something here? I get that drift is complicated and hard to predict, but I am having a hard time telling a client that we must design a roof for 100 psf drift loads unless we adjust the separation by 1 additional foot (in which case there are no significant drift loads required).

 

[JAE]

The drift is still truncated, not 100% at the edge of the lower roof - see commentary Figures C7-2 and C7-3

Check out Eng-Tips Forum's Policies here:
faq731-376

 

[structSU10]

Have you looked into the commentary? The figures C7-2 and C7-3 seem to help clarify the issue - it seems the drift is still reduced based on a 1 on 6 projection of the wind shadow from the taller building.

Edit: JAE beat me by seconds!

 

[RWW0002]

Only the windward drift appears truncated to me. The figure for Leeward drift notes that the drift height at the edge of the lower roof is hd or (6h-s)/6.

 

[RWW0002]

@ structSU - Yes the leeward drift is reduced based on the 1:6, but this may not help much for large roof steps.

 

[Once20036]

RWW, I ran into this recently as well, and it was definitely significant when we were required to change from ASCE 7-05 to ASCE 7-10 midway into a small building project.

To illustrate what RWW is describing, look at the leeward drift on a building.
I used Pg = 30, Ce, Ct, I=1. Lu=500' and the step height is 10'
Pf = 21 psf for ASCE 7-05 and ASCE 7-10

Take an adjacent building that's 10' lower with no horizontal distance between the buildings.
Both ASCE 7-05 and ASCE 7-10 indicate a drift height of 7.08' and drift width of 28.3'
Looking at a point 19' away, the drift height should be 2.32'.

Now let's add a 19' alley. We're talking about the same point in space.

Per ASCE 7-05, hd = 7.08 *(20-19)/20 = 0.354'
note that the height of separation doesn't matter.

Per ASCE 7-10 figure C7-2, hd is the smaller of hd (7.08') and (6*h-s)/6, or 6.83'

This is a huge difference!
ASCE 7-10 is saying that the drift 19' away from a step is 96% of what the max drift would be if there were no separation!
ASCE 7-10 is saying that the drift is 20x more than ASCE 7-05 said.
ASCE 7-10 is saying that the drift is 3x what the drift would be if there was no alley between the buildings.
Further more, if we went one foot more, there would be absolutely no drift.
Like RWW implied, this is tough to believe.

 

[JAE]

OK - I think you are correct that in 7-05 the drift was simply truncated and in 7-10 it changed.

We had a guy come by our office for a lunch and learn a few months ago and he did mention that there'd been a lot of new research into snow drifting and things were being adjusted.

I wrote a spreadsheet for 7-10 that confirms all your numbers Once20036. What is really weird is that as we diminish the 19 ft. gap, the drift grows to the 7.08 ft. height but the
width of drift stays at the larger width (41 ft. = smaller of (6h-S) and 6hd). So when the gap goes to 0 the drift width doesn't match the "typical" drift width.



Check out Eng-Tips Forum's Policies here:
faq731-376

 

[RWW0002]

Thanks for the response Once. The example you point illustrates my point perfectly. The only thing I might point out is that, per ASCE 7-10 the gap actually makes the drift MORE than the drift with no separation (same max drift but spread over 6hd = 41' instead of the typical drift length).

If it were not for the figure in the commentary, I might be tempted to think that the intent of the section is to truncate the leeward drift similar to the windward (i.e. hd calculated per 7.7.2 at the face of the high roof step and extended over the increased drift distance of 6hd. Whatever portion fell past the gap and over the low roof would be applied to that roof). This would lead to about half the calculated drift height at the low building (3.5')in your example This would still potentially be an increase over ASCE 7-05, but not near the extent pointed out above.

 

[RWW0002]

Thanks JAE, I did not see your post before replying.

If there is additional research I can buy the new drift criteria, but such a drastic change between codes is always a tough sell. It is even worse in this case due to the 20' limit. If it is an issue with 19.5' gaps, I would think it would be an issue with 20.5' gaps also..

Here is a sketch of my alternate (and potentially justified if it were not for the commentary figure) interpretation of the code section with the truncated drift.

 

[RWW0002]

Lets try that again. See attached.

 

[JAE]

Check out this pdf online slide presentation - specifically slide 31 (page 16) where it mentions an inconsistency.

http://oksea.org/wp-content/uploads/2018/03/Handout..3-29-18.ORourke.TwoPerPageGrayscale2.pdf

I think the sense here is that 7-10 tried to be simple and created a weird problem in their development of the drifting.

Check out Eng-Tips Forum's Policies here:
faq731-376

 

[RWW0002]

Thanks all. Sounds like, inconsistent or not, we are stuck with the new drift criteria for the near future. Hopefully this will be modified in future code cycles.