FE slab patterned loading 4

[dangardinerx]

Y'all
having previously done FE slab analyses on buildings with somewhat irregular grids (column spacings) and having found that patterned loading will sometimes give me my worst combinations of vertical shear and bending both for slab shear and forces for column design (where the bending is more critical), i find my company about to buy software (Orion) which not only does not pattern automatically but won't permit it manually. I am told anecdotally of supposed studies suggesting that patterning for flat slabs is not relevant.
Does anyone have an educated opinion on this, please?

 

[rapt]

Most Design Codes require patterning for FE analysis as well as equivalent frame analysis.

It does not matter what supposed studies suggest.

 

[DRW75]

An interesting topic, I've also been educated to the fact that patterned loads produce worst case conditions, and it doesn't take a genius to come up with conditions where this can occur.

The only difficulty that I have, and perhaps may be related to what you've heard anecdotally, is that since live loads are generally functions of probability & statistics, and that if a worst case scenario is having (take for example a triple span) max loads on either end, and no load in the middle, what are the statistical chances of this happening... in my opinion, highly unlikely if there is an equal chance of load being in any one place. But, if statistics are used, it may come out that say there is a reasonable chance of probability (say 5% in 20 years) that maybe 70% of the maximum loads on the ends can occur with 10% of the maximum load in the midspan... but then this loading condition may not govern the design compared to 100%/100%/100%. It becomes a statistical nightmare, and I think that is reason why people use the patterned loading as a conservative design philosophy. As far as it being required by design codes, you have to design for the worst case event, and if that happens to be a freakish oddly loaded structure, then so be it, but if you didn't design for this, and the building (or whatever) collapses, you better have very good reasons for NOT considering the now obvious worst case scenario.

Hence to design without considering this is professional suicide (in my opinion)...

DRW75

 

[jhardy1]

DRW75 is correct, in that the absolute worst pattern load case (e.g. alternate spans fully loaded and totally unloaded respectively) is statistically very unlikely. However, there is usually at least one "reasonably probable" load case, which often generates moments and shears etc quite close to the worst case pattern loading - and that is to apply the full load on one span only, with no load on any other span.

The "real world" scenario where this might arise is in an interior refurbishment for example. It is not uncommon for all furniture, partitions, etc to be stacked in one part of a floor area, while carpets are ripped up, new partitions re installed, etc, on the adjacent floor areas. Even if the loading on the loaded span does not exceed the normal design floor load, such load cases can often generate shears and moments that are 80% to 90% of the extreme pattern loading case results.

You should always design for pattern loading, if the relevant codes dictate so.

 

[UcfSE]

The ACI provides information on pattern loading and some simplifications in the body of the code such as limits on different patterns you should consider and even a reduction of live load in some cases, as well as requirements. Now I don't remember if those reductions applied to slabs in general or if it's only for code methods of analysis (direct design or equivalent frame). Limitations of course are optional in that if you choose to do some more conservative that isn't required by code you are permitted to do so. That (the code) should be your guide as to what patterns you must investigate and which are upto your judgment. I say design with patterns in mind.