Existing slab on grade, new loads

[penpe]

I'm seeking the best method to evaluate the point load capacity of an existing 6" thick slab on grade, reinforced with one layer WWF6x6-W2.9xW2.9. The client needs to double or triple the size (volume) of the water tank. Currently it's 10' diamter, with capacity of about 2500 gallons of "washwater" that weighs 10.3 lbs/gallon. It is supported on 4 legs. The load to each leg is about 6500 lbs. We're proposing a 13' diameter (taller) tank with 8 legs, but the point loads could be up to 9500 pounds.

I've tried a couple of methods: Shentu's "Simplified Analytical Method" was found in Structure magazine April 2008, and yields far greater capacity than ACI 360R-06 Design of Slabs on Ground reported by ACI committee 360 (copyright 2006). Neither method accounts for reinforcing steel. Shentu says the capacity is 44 kips with factor of safety of 2, while ACI method yields 5 kips with same factor of safety.

I'm using a 15" by 15" baseplate, 3000 psi concrete, modulus of subgrade reaction 50 pci, soil bearing capacity 1500 psf, tensile strength in flexure of concrete = 10% of f'c = 300 psi. (I've read it can be 8% to 15%). The ACI method relies heavily on concrete tension at the bottom surface of the slab.

The Shentu method seems geared toward rack loads in warehouses, and the ACI method is more aimed at wheel loads. Considering that the existing tank exceeds the ACI result (with F.S.=2), and has been there for 50+ years without failing (also much smaller baseplates), I assume the ACI method is overly conservative for this application. But the Shentu method yields such drastically different results I'm not believing I can trust it completely either.

Does anybody have suggestions for the best method to analyze existing slabs on grade for concentrated loads capacity? Thanks!

 

[OldDawgNewTricks]

A couple comments:
1. The traditional design methods that are derived from Westergaard are based on linear elastic theory. The Shentu method is based on elasto-plastic analysis and takes advantage of additional capacity after the onset of cracking. This should be taken into consideration if serviceability issues like cracking and deflection are important in addition to just the bearing capacity of the slab.
2. Stationary post loads vs. moving wheel loads. You usually need to be more conservative for wheel loads because of fatigue problems with cyclical loading.

 

[OldDawgNewTricks]

Another thought... Your design criteria (subgrade modulus, concrete flexural strength) seem quite conservative. I'm guessing that you don't have better information for the existing conditions and so you just made conservative assumptions? If that's the case, then it's probably the reason why the existing capacity doesn't pencil out for the current loads even though it has been in service for many years. If you are confident of the magnitude of the current loads, then you could calibrate your design criteria to match the "load-tested" capacity.

 

[dhengr]

Penpe:
Why not use a 13' dia. tank with a flat bottom? Cast a 2' wide by 12" thk. conc. ring beam, with rebar, and fill the center, level with compacted sand. Then the flat bottom distributes the load and eliminates the leg point loads, and the fab. details and problems of the legs. Probably saves some tank bottom fab. cost and detailing too. Saw cut the original slab around the ring beam to provide some crack control and to manage the potential for some differential settlement.

 

[Lomarandil]

I agree with OldDawg, the most likely explanation for the existing performance gap is in subgrade modulus.

----
just call me Lo.

 

[Settingsun]

Shentu gives 20 tonnes ultimate for 4" of the worst concrete on the worst ground. Seems it doesn't stand up to lab testing so doubtful out in the wild.

On the other hand, the old way is conservative for static loads away from edges/joints.

 

[Settingsun]

This is from an earlier paper by two of the same authors but different tests. Same basic conclusion that Shentu's method is unreliable and the elastic method doesn't need a safety factor applied.

 

[dik]

Thanks, Steve...

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Do you feel any better?

-Dik

 

[penpe]

Thanks, everybody, very much for the replies. dhengr, that would make the problem easier to solve, but the tank has to be elevated to allow for sloped bottom, a valve, and piping under the tank.
oldawg, yes to conservative assumptions. I did also consider that capacity would probably be higher for stationary loads than for wheel loads.
steveh, thanks for the article. The safety factor of 6 for Shentu makes sense, given the very high capacities found.
oldawg, I have found existing loading gives safety factor of 1.06, so I'll make the client aware of this result, for comparison. The larger tank options with more legs and bigger base plates are giving safety factor results ranging from 1.05 to 1.46. I'm thinking that any of those options are acceptable. It's also comforting that the reinforcing isn't being included in the analysis, and I think it must contribute something to the load capacity.