Snow Drift Calcualtions 3

[SteelPE]

I am in the process of designing a small office addition to an existing warehouse.

Code = IBC 2015 /ASCE 7-10

The office addition is at a much lower elevation to the warehouse, this is going to cause snow drift from the warehouse to the lower office. The total length of the building above the office is 840 feet which is producing a considerable amount of drift on the new office addition. However I am wondering if I am being too harsh with my calculation.

The higher roof (at 840') is broken up to two different elevations..... The first 520' is at elevation 28' the last 320' is at elevation 38', so we already have a snow drift accumulating at this change in elevation that is accumulating drift on the warehouse roof

What distance is correct in ASCE-7?

 

[Celt83]

For only a 10ft elevation change I'd probably look at it the way you did using the total 840' as the upper roof length. If it was a more substantial step I'd probably look at the two upper roofs as buildings with a S=0 separation for the first tier and then S=first tier up fetch length for the top tier roof.

this book may be worth a purchases I've seen it mentioned on here a few times: Link

My Personal Open Source Structural Applications:

https://github.com/buddyd16/Structural-Engineering

Open Source Structural GitHub Group:

https://github.com/open-struct-engineer

 

[bones206]

Check out this thread: thread507-379232. Especially if you are in MA.

 

[straub46]

The snow guide book Celt83 mentioned definitely has an example of the multi-step roof. My previous employer had a copy of it, so I don't have access to it anymore. I think I remember that the length of both roof steps need to be used as the total fetch on the lower roof, but I could be mistaken.

Go Bucks!

 

[Finn]

Never dealt with this, but I would be inclined to use the combined length. Snow drifts often get quite large - I am wrapping up a design with a peak snow load of 140 PSF (roof+drift).

I believe this stepped condition may be covered in the commentary of ASCE 7-16. Take a look..

 

[Once20036]

Last time I looked into this, I think that it was poorly defined in ASCE, but better defined in some of the FM Global snow design documentation.
They indicated that the drift adjacent to a stepped up roof area was based on the length of the high bay, not the total length of the building.

I believe their documents are available for free online.

 

[phamENG]

I'd be inclined to use the entire length. The only thing that I can see reducing it would be the fact that a wind eddy at the first step would drop much of the snow right there and it would be trapped in the wind shadow of the step. BUT...what about after that first drift forms? The wind shadow will be gone, and you'll have a relatively smooth fetch the full 840'. Now, if the first drift doesn't come all the way to the elevation of the high roof, I might reduce it since the math is procedure is suggesting all of the snow that would drift from the upper roof already has and landed in the wind shadow.

 

[bones206]

Just to expand on my previous post in case you didn't;t click the link, it is a code requirement in MA to consider the entire stepped roof length as upwind fetch.

 

[JLNJ]

I would probably go with 320 + half of the 520, for a total of 580 feet of fetch.

For something which has such a large effect on designs and causes so many failures, the code is quite vague when it comes to snow drift scenarios. Code writers would rather spend their time wringing their hands about P little-delta theory or squeezing the parking structure live load down to a preposterously low 40 instead of an already uncomfortably low 50 psf.

 

[phamENG]

bones - your file doesn't seem to want to open. Good find, though. Sounds like it confirms what most of us are thinking.

 

[bones206]

Here’s a better link:

https://www.mass.gov/doc/780-cmr-ninth-edition-chapter-16-structural-design-amendments/download

 

[phamENG]

Nice. That's exactly what SteelPE is asking about. (Page 10/15 for those not interested in reading the whole thing.)

 

[SteelPE]

So, bones was right..... I was using the right lu with my initial calculations ...... Being from MA, I should have known about this section of the code, but I didn't remember it at the time of posting this question... which is quite embarrassing.

 

[phamENG]

SteelPE - that's the trouble with those "addition and revision" documents. It can be tough to keep everything straight and know when your state has changed some section of the model code. ICC and UpCodes have started posting coherent state-by-state codes to their websites, but it looks like MA is still missing from that.