Self Weight of Deep Foundations and Structural Slabs

[abusementpark]

The traditional method for designing any deep foundation system is to treat the entire system like an elevated platform and all the loading (including self-weight) is ultimately transferred down the load path to the piles or shafts.

However, I wonder how much of the self-weight truly gets transferred to the deep foundation elements (i.e. piles or shafts). When the concrete for the foundation superstructure (i.e. grade beams and slabs) is initially poured, nearly 100% of the dead load has to be soil supported since the foundation superstructure has no spanning capabilities until some curing starts to occur. So, if the soil is initially taking all of the dead load, what is going to cause a transfer in load once the foundation superstructure cures? I can understand how it could easily occur when you are dealing with highly expansive soils, but it soil movement enough of a concern for typical, non-expansive soils?

I'm curious to get your thoughts on this matter. I'm not saying our current practice isn't the right/safe way to do it. However, I wonder about reality.

 

[JAE]

I'm sure the load starts out on the soil, but if you are using a deep foundation/structural slab system in the first place, that tells me the underlying soil isn't adequate to support the loads.

The soil conditions that exist for structures like this are surely quite varied. But these parameters would be tough to know actually and even tougher to use to determine what the load sharing (between soil and structure) would be.

I think you are correct that we have to design the safe way and put all of the load in the structure.

Most times the "bad" soil won't have much stiffness or capacity to take load at all.

 

[Ron]

The initial load on the soil is actually quite low....somewhere on the order of 100 to 200 psf. Even poor soils can usually take that amount. As the structural slab/grade beams cure and achieve a load carrying capacity, then the additional load as added by subsequent construction will be picked up the the transfer mechanism. No additional load, except through deflection, will be tranferred to the soil, except at the deep foundation system.

 

[hokie66]

As structural engineers, we have to make sure there is a load path. But structures are often smarter than we are, and select alternative load paths. Isn't that a good thing?

 

[SteelPE]

I imagine the amount of load the soil can take would be site specific. If you have a layer of compressible organic material between the foundation system and the bearing strata, over time the weight of the foundation will be transferred to the piles themselves.

I have seen slabs that were initially cast against the soil but over time the soil compressed and there was a gap between the slab and the soil. In this instance, the foundation system performed properly.

 

[ishvaaag]

Obviously how the loads are shared will be dependent upon the case and advance in time of the settlement. One special ACI publication (SP-152) has an article referring to the thing for the piled mat foundation of the Messeturm, that must be one of the tallest buildings in Germany. The foundation was instrumented and I post how the load was measured to be shared between mat and piles, fig.17.

 

[JoshPlumSE]

While the slab may be in direct contact with the soil when it is initially poured, there is not guarantee that it will remain there. It may settle locally under the slab. That's the assumption we usually make. In general, I would say that is good practice.

That being said, I have seen a project where the geotech report allowed load sharing between the slab and the piles. I'm sure the geotech did an extensive investigation to justify this. But, it's not hard to understand conceptually. In this case, they could easily justify the Geotech's expense because it was an existing structure that could never have been modified without it.