Snow drift: flat roof adjacent to gable

[walterbrennan]

Howdy...

Given: a gabled roof with an attached, flat lower roof immediately adjacent to the low end of the gable on one side.

Desired: is the drift height (hd) specifically limited by the vertical difference between the top of the low (flat) roof and the bottom of the gable above; or can (should) it "creep" up past the end of the gable, if the calculations indicate such a drift height? ASCE's diagram (fig. 7-8, chapter 7.0, ASCE 7-02) only demonstrates an aerodynamic drift situation between flat roofs of different heights, separated only by vertical distance. And the language in section 7.7 doesn’t seem particularly definitive on this matter, either.

A logical extension of that question would be whether or not sliding snow would be included on top of the drift; as was indicated by figures back in the good ol' UBC. Although ASCE 7 states, in section 7.9, “Sliding loads shall be superimposed on the balanced snow loads," it seems to me like if there were a substantial (vertical) difference between upper roof eave and lower roof surface, you could expect (aerodynamic) drifting AND sliding loads to occur at the lower roof. Would you consider such an opinion horrendously over-conservative?

Opinions…?

Thanks,

walterbrennan

 

[DaveAtkins]

I don't assume that the drift on the lower roof can be higher than the eave of the upper roof. But I do assume snow can slide onto the lower roof, and I use the worse case (drifted snow, or slided snow). ASCE 7 does not make you put slided snow onto drifted snow ("Sliding loads shall be superimposed on the balanced snow load.")

DaveAtkins

 

[geoffdale]

Hi all.
Up here in Canada, we have a clause that allows us to max out our snow load at 3x ground. There are still cases though, similar to what you described here, that we don't implement this clause.

 

[walterbrennan]

DaveAtkins,

What if there is NO eave height above the low (flat) roof (i.e. the gable "dies" into the lower flat roof... does that mean there's no drift...? And, according to the sliding provisions (7.9, I think), we seem to be able to determine for ouselves whether sliding might be reduced or eliminated, under just such circumstances... (i.e. "...unless a portion of the snow on the upper roof is blocked from sliding onto the lower roof or is expected to slide clear of the lower roof.")

Whadayathink?

walterbrennan

 

[walterbrennan]

I just got off the telephone with Prof. Michael O'Rourke (Rensselaer Polyetchnic Institute), the moderator for ASCE's most recent snow load web seminar ("Snow Provisions in ASCE 7-02"). He admitted that ASCE does not yet have load cases or provisions for dealing with a gabled roof dying into a flat low roof. He suggested that:

1. From a leeward direction (i.e. wind blowing away from the gable across the flat roof) the wind would likely be deposited on the gable just as it passed the ridge, resulting in an unbalanced load condition as in ASCE 7-02 Figure 7-5; with little or no snow actually drifting down onto the flat roof.

2. From a windward direction (i.e. wind blowing across the flat roof toward the gable), with a steep enough roof gable, the situation might be treated as in Figure 7-8, with the gable peak taken as the high roof elevation. On the other hand, he said, if it were a “flat” enough gable, there would probably be no windward drifting at all. When pressed on what he might consider a reasonable range of slopes to validate this approach, he responded that 75 or 80 degrees seemed reasonable for the steep condition. When asked what should be done when the situation fell between a “flat” enough gable and the “steep” enough gable, he demurred; once again flatly stating that ASCE simply did not address snow loading for a gable roof with no vertical eave distance above an attached flat roof.

We are left, I believe, to draw our own conclusions. What I’m wondering is, what are the rest of you doing for this (relatively) common geometric scenario…?

Thanks,

walterbrennan

 

[miecz]

My home in New England has a gable (4 on 12?) roof that dies onto a nearly flat roof. That area sometimes drifts in a snowstorm, with waist high snow drifts where the ground snow is no more than 12 inches. My neighbors must get a kick out of me shovelling my roof! The tendancy to drift depends on which way the wind blows, but I've never paid attention to wind direction vs. drift.

 

[DaveAtkins]

Maybe I am becoming more liberal as I get older, but I would not worry about drifting or sliding in the situation that you describe, walterbrennan.

DaveAtkins

 

[walterbrennan]

DaveAtkins...

Did you read the response posted just ahead of yours...?! This guy only has a 4/12 slope and can confirm waist-deep snow drifting where the gable dies into the flat roof...

I'm not sure where you live, but having spent more than 12 years in Fairbanks, I've also shoveled a few roofs. It’s all the more curious to me that this scenario hasn’t been covered by ASCE, yet; considering that they treat so many others.

Just thinking out loud...

walterbrennan

 

[miecz]

If I were to design the roof I mentioned in my above post, in cross section, I would draw a line from the tip of the gable to the end of the flat roof, raise that line a foot, and consider the area under the line filled with snow.

 

[vmirat]

I would determine what the drift would be at the ridge of the pitched roof to the flat roof, as if both were flat, then taper the drift to where the pitched roof meets the flat roof. The drift would wind up looking like a roof cricket (chirp, chirp).

 

[walterbrennan]

Thanks for the opinions. Good to know that there are others who have given this some thought...

Have a good 4th.

walterbrennan