Can you reinforce slab on grades to engage more soil? 1

[EngDM]

Going to cross-post this thread over here as it's not necessarily just a SAFE issue but more of a theory issue.

thread803-523278 - Please lock the other one if similar threads are not allowed.

When designing a raft slab, is it possible to stiffen the slab with additional horizontal reinforcing to have the slab engage a wider strip of soil below my line load and thus lower the bearing pressure? I have an issue where my load exterior bearing walls are exceeding my factored bearing capacity due to how the load propagates thru the concrete down to the soil, but I have existing drawings of an identical building on the same site with the same bearing allowances, that was stamped and sealed. My counter argument is that for the slab to start engaging the soil further away from the load point, it would first have to deflect a bit which, soil deflection under load is a failure (settlement is a different animal). But it was suggested to me to stiffen up the slab to go from, for instance, 3' of engaged width to 4'.

I'm just curious as to what approach would be considered reasonable? I've tried playing around in SAFE by increasing reinforcing and it doesn't appear that it will help in any way; only changing slab thickness or subgrade modulus has any affect.

 

[KootK]

I think that you are on the right track here:

1) Adding rebar alone doesn't do much to alter stiffness and, thus, does little to "redirect" load.

2) Adding rebar can allow you to lay claim to mobilizing more slab but the cost of that mobilization will potentially be:

a) More settlement.

b) Higher local bearing pressures that would need to be redistributed.

I feel your pain. Engineers in central Canada have traditionally be very aggressive with this kind of thing. I've seen strap beam foundations supporting property line columns where the strap beam spanned 25' and was only 16" thick.

 

[Greenalleycat]

In short: yes

When you say raft slab I'm assuming you mean a constant thickness slab?
This is important as you'd want all your load to be on the same bearing layer

I wouldn't describe soil deflection under load as a 'failure' - more a fundamental principle...
Excessive settlement (particularly differential settlement) is the failure - there's nothing that stops you allowing minor soil deflection to engage a wider area
Typically I'd expect the foundation to be much stiffer than the soil so engaging a wider band shouldn't be too hard

You do want to be careful when considering how your foundation is reinforced though - is there a punching shear or vertical shear issue if you have a thin raft?
How are you going to estimate the loads in the foundation with this model?

 

[TLHS]

Reinforcing isn't going to add any reasonable stiffness for this kind of situation. Thickness will. If you just need a foot more bearing width, have you considered trying to account for a layer of nice granular base? A well compacted granular material over a softer foundation soil will spread bearing across the subgrade somewhat.

Soil deflection under load isn't a failure, it's just reality. Also, depending on the soil and the slab geometries, the subgrade could be such that it's fairly flexible compared to the slab itself. It all depends on the system.

What does this slab look like? How are your propagating your load and how are you modelling your foundation soils?

And yeah, you likely aren't in these situations, but bearing stress failure may not really be a hugely clear failure mechanism, especially in mat style foundations. It depending on what the geotechnical engineer was trying to communicate. If it's settlement based, then it becomes a question of allowable system deflections. If the foundation holds the movement in a way that doesn't damage the superstructure then you're likely okay. This is somewhat typical in really bad soils. You're more likely to be using a pretty qualitatively thick slab in those conditions though. In theory with the Canadian system they're supposed to give you a serviceability pressure and a ULS factored pressure, which would help with this, but I haven't seen that done consistently in practice in a way that makes sense (and I don't blame the geotechs)

Side note, since it may affect how you're thinking about this. If you're using factored loads in Canada and are comparing against existing foundations, be aware that there are some conditions that are going to be more conservative. Anything that's resisting an overturning moment in the individual foundation with bearing is going to be magnified if you're assuming everything responds elastically. You'll also generally find that a lot of resistances from geotechs didn't scale the way you'd expect them to, when looking back at old reports. There are some ways to deal with some of this, but there's not great literature to point at. So the other foundation working may be an instance of the other guy just having different numbers to work with.

 

[EngDM]

You do want to be careful when considering how your foundation is reinforced though - is there a punching shear or vertical shear issue if you have a thin raft?

I've got a 16" thick slab to work with, punching and one way have already been analyzed and are nowhere near close to failure.

Side note, since it may affect how you're thinking about this. If you're using factored loads in Canada and are comparing against existing foundations, be aware that there are some conditions that are going to be more conservative. Anything that's resisting an overturning moment in the individual foundation with bearing is going to be magnified if you're assuming everything responds elastically. You'll also generally find that a lot of resistances from geotechs didn't scale the way you'd expect them to, when looking back at old reports. There are some ways to deal with some of this, but there's not great literature to point at. So the other foundation working may be an instance of the other guy just having different numbers to work with.

Existing building is less than 10 years old, and the new geotech report provided gives the same bearing pressure allowances. Old building was also ULS/SLS, and not the cursed allowable capacity.

Foundation soils are modeled in SAFE as a defined spring support with given subgrade modulus, and then that spring is assigned to the slab.

 

[JAE]

I get the idea that adding rebar isn't "affecting the stiffness that much."

But I think the idea to keep in mind is that adding rebar does indeed increase the stiffness of the slab (i.e. transformed area of steel adds to Ix)...but granted - very inefficient.
I agree that increasing slab thickness is the way to go - especially under concentrated or line loads.

The amount of spread, between a line load on a flat slab-on-grade depends a lot on the RELATIVE stiffness between the soil and the slab.

If you have total mush for soil, then the spread will be very wide (with high overall deflection).
If you have rock, and a less stiff slab - the spread will be almost nothing (i.e. the load goes directly into the soil below the load.

I'm not sure how SAFE handles this - sounds like it doesn't.

I might be inclined to model the slab as a finite element mesh with soil "springs" and see how it behaves with a parametric range of soil/slab values.
The slab stiffness is hard to pin down with FEA due to variable cracking but with a line load the behavior is fairly uniform and you might be able to do a step-wise procedure - altering the stiffness in a manual 2nd order procedure if your software can't do it for you. Analize - find which elements have stress beyond the cracking stress, alter the element's stiffness, re-analyze - rinse repeat. Or add the rebar - assume all is cracked - and go from there.

The soil stiffness is also hard to pin down but we typically use a range between 75 pci and 150 pci to get a feel for how much spread is really possible.

 

[EngDM]

I'm not sure how SAFE handles this - sounds like it doesn't.

If I change the subgrade reaction modulus it does indeed spread the load out further than just transfering the load down at a 45° angle and taking that bearing area.

AFAIK SAFE does it as a finite element mesh, that is how I set up he slab.