Live Load + Snow Load at Assembly Roof

[JSMHA]

I was wondering how people handle the snow load at assembly roofs where the live load is 100 PSF per the code. Are people applying full snow load on the roof with the assembly live loading?
Based on ASD design 2015 IBC Equation 16-11 provides a load combination of D+.75L+.75S. ASCE 7-10 provides a similar load combination D+.75L+.75S but clarifies that the S is flat roof snow, not drifting and unbalanced cases. If we call the flat roof snow load 30 psf, the 100 psf controls over 97.5 PSF we get from the load combination. My approach has been to design for just the 100PSF unless there are unbalanced snow load cases that exceed 100 in which case I would design for full snow load separately.
But what about for concrete member design. Based on ACI 318-14, the load combination 5.3.1b (1.2D+1.6L+0.5S), which provides no clarity that the S does not need to include unbalanced snow load cases. Based on this we would need to design concrete members for 1.6*100+.5*30=175PSF (or more depending on unbalanced conditions).
Does anyone know of a reference that clarifies this? How are people handling these loading cases? It seems impractical to apply full snow drift and full assembly live load together.

 

[Guppyz]

Canadian codes specify that the live load and the design snow load do not occur concurrently and that you would take the greater of the two to design with. Not many people on patio/roofs when a design level snow storm is occurring.

 

[phamENG]

I would follow the load combinations as stated. I agree that you won't have full assembly live load during a design snow storm...but we're not really worried about during the storm. Those load combinations also don't consider full live and snow load at the same time. The ASD load combination assumes 3/4 of each, while the LRFD load combination assumes full live and half snow.

Live load isn't just people. It can be furniture, temporary equipment, and other things that don't form a part of the permanent structure but also stay in place when the snow comes. And are you really telling me that a bunch of people aren't going to walk onto a snow covered terrace in the morning at a ski resort? Assembly areas also tend to be large, and some places will use mechanized equipment to clear it.

 

[Guppyz]

I think there are exceptions to every rule, but to say you have a 1/50 year storm followed by a complete 100 psf occupant live load is a stretch.

 

[phamENG]

Who is saying that? The load combination reduce one or both of the transient loads to less than the full design magnitude....

 

[BAretired]

It's true that you don't get a full load of people when the deck is covered with snow, but snow removal can cause large piles of snow to accumulate in some areas while waiting for it to be trucked away. After a heavy snowfall, heavy equipment may be moving about if the area is large. It is not a bad idea to discuss with the owner his plans for removing snow.

BA

 

[Enable]

Just to clarify the Canadian approach for a second. The latest NBCC load combinations are based primarily on the work of Bartlett, Hong & Zhou (2003) in the Canadian Journal of Engineering Vol 30, No.2, pp 429-439, and 440-448.

You'll note that they do not say loads of certain types will not, for sure, occur together. Rather, they use a statistical approach, which shows they are unlikely at their full magnitude to do so. They go into detail how this effects the target reliability index.

Consequently, we do NOT ignore the combination of multiple types of loads but rather we break loads down into 'principal load' and 'companion loads' such that the loading factors are adjusted for each to meet a given reliability index.

For example, if I am looking at a snow loading case being my principal load I might have the following combination (note: [1.25,0.9] to be read as factor is either 1.25 or 0.9, whichever is worse for the loading case under review)

Principal Loads
[1.25,0.9] Dead Load + 1.5 * Snow Load

Companion Loads
0.5 * Live Load OR 0.5 * Wind Load

On the other hand, if Live Load is my principal load we get

Principal Loads
[1.25,0.9] Dead Load + 1.5 * Live Load

Companion Loads
0.5 * Snow Load OR 0.5 * Wind Load

 

[Guppyz]

The intended use is the largest factor for sure. Open air walk-way from one space to another will have a chance to occur both design level snow load and portion of design live load and vice versa, while I doubt the sunbathing roof terrace is getting large volume of visitors with a design level snow sitting on it.

 

[skeletron]

For the member design, I would run a baseline analysis with 100psf first to satisfy live load deflection. Modify the design to satisfy building geometry. Then re-run the analysis with a full factored combination (principal/companion). Step back and assess where your design is being pushed and where it is not an issue. Then consider the other factors with the area's geometry and obstructions.

If it is an assembly area, there is definitely some value in designing it for full code-load. Yes, there is always that nagging questions of "does the full LL also come with full SL, and what is the probability of that" and I think the above process somewhat helps calm that question with some minimal analysis. In the end, if you are adding a couple more bars, or a couple inches to satisfy the full factored LL and SL, then it really shouldn't be an issue. If the baseline design significantly has to change to accommodate the full code load then start refining your numbers.