Foundations on Rock and Soil

[MTNClimber]

I'd like to take a survey of how you folks typically to deal with strip footings that are supported on rock and soil. I've seen reports require:

-Over excavation of the rock by 12-inches and replaced with granular fill.
-Transition zones of about 20' long that include over excavating about 18" of bedrock and tapering the rock excavation to 6".
-Vertical joints in the foundation and wall at the location of change in bearing material.
-Additional reinforcement in the footing at the change in bearing material.

Just interested if anyone had any other options or even experience (good or bad) with the options above.

Also, do you have any issues with part of the interior isolated footings bearing on rock while other are bearing on soil?

 

[EireChch]

to be honest, I dont think I have ever encountered such an instance.

But to me it would come down to the structure, the loading and differential settlement between the two.

If differential settlement between the two was significant then I would consider widening the portion on granular to result in a differential settlement that was acceptable. Maybe 10mm??

 

[TigerGuy]

The only time I am comfortable with bearing directly on rock, is when drilled shafts are socketed into said rock. For shallow foundations, even if completely on rock, I usually like at least 6 inches of cushion between rock and concrete.

If the soil is properly prepared, and the rock is excavated for a minimum 6-inch cushion, I've not experienced problems in any of my past projects.

 

[EireChch]

Tiger Guy - Why are concerned with bearing directly on rock?

More from a construction point of view where you can create a level foundation base as excavated rock may be uneven?

 

[dik]

I did a bank project in northern Manitoba where they had to remove some rock for the SOG floor. I use the cut at the wall and filled the space with formed concrete and placed the masonry bearing wall directly on it (no foundation). The SOG had a min 6" buffer of compacted gran fill.

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

 

[TigerGuy]

Buildings move, ever so slightly, no matter how we design them. If the concrete is in hard contact with the rock, that's a rigid point of contact causing potential stress to the structure. If the rock surface is overly jagged, there is a point of stress concentration. Things like that

 

[geotechguy1]

I've seen recommendations as you described but also having the part on rock on shallow foundations and then the part where soil is present require piles founded in the rock.

I think it's really a soil-structure interaction problem.

 

[le99]

We usually specify "Excavate to the competent rock. Fill voids with lean concrete."

 

[MTNClimber]

I dont think I have ever encountered such an instance.

I'm really surprised to hear that. Top of rock elevation can be highly variable where I practice.

The only time I am comfortable with bearing directly on rock, is when drilled shafts are socketed into said rock. For shallow foundations, even if completely on rock, I usually like at least 6 inches of cushion between rock and concrete.

If the soil is properly prepared, and the rock is excavated for a minimum 6-inch cushion, I've not experienced problems in any of my past projects.

Interesting. Seems conservative but as you said zero problems is a great track record.

I did a bank project in northern Manitoba where they had to remove some rock for the SOG floor. I use the cut at the wall and filled the space with formed concrete and placed the masonry bearing wall directly on it (no foundation). The SOG had a min 6" buffer of compacted gran fill.

Thanks for the story dik. I agree with the slab being on compacted fill or crushed stone. Maybe I'm wrong but I feel like a slab on rock could be asking for cracks to develop.

I've seen recommendations as you described but also having the part on rock on shallow foundations and then the part where soil is present require piles founded in the rock.

I've recently heard of this being done on a large warehouse facility with a very large building footprint. While I feel like this is conservative and can be expensive, I also can see how this can approach can be justified with certain scenarios and clients.

We usually specify "Excavate to the competent rock. Fill voids with lean concrete."

That's definitely a good way to get a relatively flat surface for the foundation construction too.

Thanks for the responses everyone. What brought this up was a project we have with a building footprint that has sloping bedrock underneath. Most of the building was going to be on either bedrock or weathered rock and small portion was going on glacial till. All pretty stiff material but we had concerns at the transition points along the continuous footings where rock would transition to glacial till. Unfortunately it would of been overkill to have them excavate down to bedrock everywhere so we had to figure out what everyone was comfortable with.

 

[geotechguy1]

Suppose as a thought experiment you put a book next to a sponge of the same height on your counter, and then put an object like a brick on top, say half on the rock and half on the sponge.

There's a good chance the brick will stand up, and it should be intuitive that the book is carrying all of the load and the sponge none or very close to none. Now instead suppose you have an apartment building sitting half on rock and half on soil: The same thing should apply, BUT (and I'm not a structural engineer so don't claim to know the answer to this question) are current methods of analysis adequate to design the structure to handle the business of redistributing the loads towards the stiffer area (rock) - and are current methods of site investigation adequate to characterize the site to the point that such modeling could credibly be done during the design phase?

I think the answer is, yes, the techniques and software exist to do both of those things but the people likely to be investigating and designing these things in practice are no where close to commonly understanding and using those techniques and software. Therefore, footings on rock and piles through soil to rock in other areas so that overall behaviour is the same.

 

[EireChch]

we to have variable rock head where I’m working now but no one has asked me that question yet!

I wouldnt be concerned with bearing directly on rock at as long as it was flat and smoothed out with site concrete.

If you have a big site and are forced to excavate 300mm and bring in gravel to refill, seems like extra work for not much good reasoning.

Re the issue of half on rock half on soil, i think the only thing that you need to worry about is differential settlement. This can be quantified and dealt with easily.

 

[GeoEnvGuy]

I live in hard rock country. We don't excavate rock we have to blast it. Foundation recommendations indicate to make sure the angle of the rock is flatter than 45 degrees or fill it with concrete or backfill with aggregate.

 

[Mccoy]

I've come across this issue quite a few times, there has even been a discussion on it about 20 years ago. There may be a couple of significant cases.

1= building on rock AND on soft, compressible soil. In this instance a proper solution is not easy, what has been experienced here is a slab partly supported by rock, partly by drilled piles or micropiles to underlying rock basement or more competent soil, with the expectation of differential settlements, hence a very rigid foundation. There would also probably be issues related to different seismic behaviour on rock and on the soil. A more thorough but expensive solution would be to excavate and replace the soft soil, but this would depend on depth.

2= building on rock AND on stiff, not-so-compressible soil. The solution adopted has been to excavate the soil to some thickness (about 2 feet), replace it with well compacted granular fill, extend the fill thickness above the rock surface by a suitable measure (at least one foot), and lay a rigid slab or a rigid grillage foundation. This also addresses in part the seismic issue, the layer immediately below the foundations being homogeneous.

Solution #2 has been found to be valid during a Mw 6.3 earthquake in Central Italy.

 

[EireChch]

Mccoy,

Point 1 - I would be very comfortable with, everything is taken to rock. I imagine that the pile length is tapering so piles eventually reduce to nothing. This is a good way of minimizing the potential differential.

Point 2 - if the levels do not permit the 1ft raise with fill over rock, will you force the client to excavate rock and replace. If this is hard rock that needs blasting....personally I would not. (regardless of hard rock/blasting or not).

 

[Mccoy]

Point 2 - if the levels do not permit the 1ft raise with fill over rock, will you force the client to excavate rock and replace. If this is hard rock that needs blasting....personally I would not. (regardless of hard rock/blasting or not).

Right, if there is a level issue, I too would insist on engineering a super-stiff fill on the soil side and on designing an adequately stiff foundation. And maybe I would recommend a daily prayer on top of it...