Racking System on Slab on Grade

[SteelPE]

I am trying to design a slab on grade for a project that I am working on (see other post about forklifts). Attached you will find a design chart and example from an old text I have called “designing Floor Slabs on Grade”. For reverence, this same chart is used in ACI 360R.

I am a bit confused as to what loads to apply to the posts/slab. The design example assumes a 5-level shelf at 100 psf with a base plate of 20 square inches (16,000 lbs). However, this appears to be based on an assumption that the column in the middle of two back to back racking stems will take all of the load onto one 20 square inch plate (to me is should be supported on two plates). I can’t say that I have seen a racking system built like this before as every racking system I have seen has two sets of columns in the middle separated by 6”-8”.

So what load would you used when designing a racking system with two back to back set of shelves (two separate columns on two separate base plates)? Would you use 8 kips on 20 square inches or would you use the full 16 kips on 20 square inches?

 

[SteelPE]

So no one has ever used to charts provided?

I have a similar problem where the my rack load is 14,000 lbs. The racking system has a base plate of 36 square inches however, there are two posts separated by a few inches that are supporting 14,000 pounds (7,000 lbs per post). Each one with a 36 square inch base plate. So I have a few options:

1) Design the slab for a 7,000 lb load on a 36 square inch base plate
2) Design the slab for a 14,000 lb load on a 72 square inch base plate
3) Design the slab for a 14,000 lb load on a 36 square inch base plate

I think option 1 and 2 will yield the same results due to the way the calculation is performed.

I have talked to two colleagues and they think option 1 or 2 is the way to go and this is the direction I am leaning.

 

[lexeng18]

I would sum the combined area of the two base plates and square root for an equivalent single square base plate, assuming your design reference doesn't cover rectangular bearing areas. So I'd look at this as a single 14k load on a 50 square inch base plate.

 

[jayrod12]

I'm on board with 1 or 2. I think 3 is overly conservative and not indicative of the way it has been constructed.

 

[Ron247]

I have used it years ago. Page 43 is the sketch they are using I think. It shows 2 racks back to back. In that scenario, 16 kips is correct. The center post would be 4'x8'x100 psfx5 shelves = 16k. The sketch on the nomograph is really just definining X and Y.

 

[SteelPE]

Rummaging through some old paperwork looking for a detail for another project I came across a slab design my old mentor did 20 years ago (he has since passed on). He basically uses method #1 to design for the slab design. Inside his paperwork he has on old PCA document from 1976 that again, reproduces the same chart outlined above. I still have the same confusion reading the 1976 article as I do with the example I posted above.

I also had a long conversation with my client about this issue. They have now stated that the 14k load is the load in each racking bay including the load placed uniformly on the slab.... so I can reduce the load on the racks by 25% (racks have 3 shelves). I still think I might look at it as a 7k load on a 36 square inch base plate.

FWIW, they are in the process of blasting to put the foundation in. They are reprocessing the blast material to use under the slab. So I think the sub base will be quite stiff.

The client is putting pressure on me to make a decision soon as they would like the issue put behind them.

 

[TLHS]

I quite like this set of charts. It's more aggressive, but the references are all there. Honestly, though, I never get to this sort of compression values on an unreinforced slab on grade. Around here there'd always be some sort of seismic situation that causes uplift or shear problems.

https://www.structuremag.org/wp-content/uploads/2014/08/C-StructuralDesign-Azzi-Apr081.pdf

 

[Once20036]

Ah, the old Structure Magazine article strikes again.
Please read all of the text. It is only applicable for a single point load in the middle of a slab.
Any time that you have two point loads within a distance "r" of one another, you are outside the scope of the article.
Refer to the center portion of the right column on page 2. This certainly isn't the case for any racking system that I`ve ever seen.

I`ve documented my concern and frustration with the references in several previous posts. I`m not sure if TLHS means it when he/she says the references are all there.

I think that your option 1 is not appropriate to use. Think about punching shear on a flat slab. Your shear failure planes will overlap and you don't get double your capacity just because you doubled your plate area.
I`m not sure which design procedure you're using, but if there's no way to account for adjacent loads, I like Lexeng18's suggestion of using 14k on 50 sqin.

 

[SteelPE]

I'm not quite sure how he arrives at 50 square inches in his example.

I am referencing a number of different sources that all go back to the PCA method for racking design. I have "Designing Floor Slabs on Grade" by Ringo and Anderson. I have an older copy of ACI 360R. I also have an old PCA article from 1976. All of these methods are the same (PCA) and are specifically related to racking systems.

I did see the Structure Magazine article but after a quick review I didn't bother with an in depth read because it appeared to only apply to point loads (which is also covered in my first reference).

FWIW, punching shear is is more than adequate in this example (assuming d = slab thickness - 1"), so that is not an issue at all.