Figure I'll chime in since I am the aforementioned instructor, and am curious about this topic as well as I've yet to find an official resolution to it.
JAE said:
I don't believe that is what the code says. I'll have to look but I've never done that ever.
Now not reducing storage live loads makes sense of course. But I've never heard of using different tributary areas for different types of live loads.
I based this off of my interpretation of ASCE 7-10 4.7.2 (and fully suggest that I could be incorrect, this is just how I read it).
The definition of L
o = unreduced design live load per ft
2 of area supported by the member. (Essentially L
o = design pounds per square foot)
The definition of A
t = tributary area in ft
2
Now I take the ft
2 from L
o's definition to be equivalent to the ft
2 in A
t's definition. This is because the definition of L
o states that it refers design load per square foot of area supported by the member but if we take this "area supported by the member" as the entire tributary area for a multi-story column with differing loading at each story then it's not correct. Thus, I interpret this as we have to calculate L (reduced) for each loading type. I further rationalize this as what if K
LL changes types between stories? For example a column is located centrally in a lower story but the building tapers as you go up in height so at some point this column becomes an edge column. In that case it seems clear to me that you must calculate K
LL*A
t for each K
LL region and it would not make sense to use the total A
t for these two calculations.
My one final rational is that each loaded area will have different statistical probabilities for maximum load. For example, let us say we have a building with mixed offices and residential floors. The chances of all residential floors being loaded up fully in a building is low; so we reduce the live load for the residential floors as we would expect. Now, the offices below will likely be loaded at different times than residential (but will have some long-term live load in each). Thus, the chances that all the residential floors are fully loaded AND the office floors are fully load is low, but I'd say it to be no lower than each area analyzed separately.
Do the statistics of typical office loading change based on how many apartments are occupied above it? In my opinion I don't believe so. Is there a larger tributary area that the column supports when we include both the office stories and residential stories? Yes, which ignoring type of live load, statistically should have be lower uniform load. So, this is why I could see this working both ways.
Again though, I've yet to find any official clarification and I do agree that I've seen a few examples do it my way (though I question these examples quality as we found other errors in it) and a few examples do it the other way.
Professional and Structural Engineer (ME, NH)
American Concrete Industries