industrial slab on grade (comparaison between ACI360R-92 and plain concrete)

[structural87]

Hello,
I am designing a slab on grade of an industrial warehouse where some high uniform loading will be located and forklift circulation is expected daily.
The slab on grade is 100x100 meters and will be separated from the concrete pedestal by isolation joints. The concrete pedestals are spaced at a regular interval of 20x20 meters.

I followed the ACI 360R-92 "design of slabs on grade" and the method of PCA and found that 20cm slab thickness with 5 x 5 meters contraction joints is appropriate in condition to provide some minimal top reinforcement steel (As = 1cm2/meter).

What i found strange is that the above document treats contraction and expansion joints the same.
In other words, if i provide 5 x 5 meters contraction joints at a regular interval, can i ommit expansion joints in the 100x100 meters warehouse?

In a separate exercise, I would like to verify using the recommendations of ACI318-11 chapter 22 plain concrete the above findings.

Which panel should I consider ? shall I consider a panel of 5x5 and model the uniform load spread over 2x2m or should I increase the panel? Results are quite different between a small and a big panel which is logical.

thank you for your opinions.

 

[phamENG]

In most building applications, your slab isn't going to expand more than its initial contraction from curing. If you have very long and relatively narrow stretches of concrete (such as a retaining wall or pavement section) or concrete with especially high temperature swings (such as pavement) then an expansion joint is prudent. But we rarely put expansion joints in a slab on grade. For instance, if you use 10x10^-6 (m/m)/C * 5m * 32C = 0.0016m. This would the the worst case expansion from 0C to 32C (32F to 90F). The only way you'd see that much expansion into the joints is if the concrete is poured at 0C - which is a terrible idea. So your worst case (and largely unrealistic) case puts about 1.6mm of thermal movement at the joint. Shrinkage strains are a little harder to determine, but in most cases its a wash.

As for plain concrete modeling of a 5x5 section, that's not the approach I would take. You should have dowels crossing your joints to minimize differential vertical movement, and this will "spread" your load out past the bounds of the 5x5. You'll also need to check for middle, edge, and corner wheel loads from your forklifts.

I've found the attached spreadsheet helpful. It's in Imperial units, though - sorry.

 

[r13]

Just a note, a contraction/control joint is meant to draw shrinkage crick to the weakened section, but does not prevent the concrete to expand.

You can explore the use of plain concrete slab, but I personally don't think it's a wise choice for warehouse with traffic loading.

 

[structural87]

hi retired13. why isnt it a wise choice ?
the slab thickness is appropriate to handle the corresponding load caused by the forklift and the top reinforcement is placed in the upper half to reduce the shrinkage / thermal stresses along with contraction joints.

 

[r13]

You can explore the use of plain concrete slab, but ....

Repeated dynamic/traffic load is not very kind to plain concrete slab.

 

[structural87]

I agree but there is a recommended safety factor of 2 which u divide by your modulus of rupture.
2 permits unlimited traffic load on your slab.

 

[r13]

Yes, at a cost if you can justify with certainty.

 

[SlideRuleEra]

I am designing a slab on grade of an industrial warehouse where... forklift circulation is expected daily.

If high-lift forklifts will be used, the floor's finish and construction tolerance should be considered. The results will influence joint design and spacing.

http://www.SlideRuleEra.net

 

[Settingsun]

I wouldn't bother with expansion joints. As Phameng said, the expansion has to exceed the shrinkage before there's any stress, and then the stress is a small compression which is good for the concrete anyway. Expansion joints are wide and subject to damage from wheels.

1980s guidance from Britain was that approx 15% of load can be transferred across a joint by dowels in plain concrete floor slabs. Check a 5x5m panel with 85% of wheel load near edge.

For comparison with your calcs, the old British guidance was 200mm slab is good for 10 tonne forklift axle (4 to 5 tonne lift capacity) and minimum 60kPa distributed load.