Mat foundation settlement

[Okiryu]

Hi,

We are planning a mat foundation (48m x 75m) on a site with heavily overconsolidated fat clay (CH) soils. These clays are located starting from the existing ground surface. Layers thicknesses vary from 2.5 to 5 meters. Underneath these clays are dense to very dense gravels and occasionally outcrops of limestone rock.
Fat clay properties are: SPT N-values=6~9. G=0%, S=6~13%, Fines=87~94%. W%=30~40%, LL=56~94%, PL=27~34%. Shear strengths=65~77 kPa. Cc=0.39~0.42, Cr=0.05~0.08. Po=1000~1300 kPa

The mat will be placed on top of 1~2m of fill. The preliminary thickness of the mat is 750mm. Maximum floor load is approx. 50 kPa.

Bearing capacities are okay but when checking consolidation settlements, they are between 50mm~100mm (allowable settlements are 25mm). So, I am planning to remove between 1~2m of clay and replace it with granular soils in order to reduce these settlements. Good thing is that after place the 1~2m of fill and install the mat, we have 4~5 months to start with the superstructure. At the end, all these conditions (fat clay removal, the 4~5 months prior the superstructure construction) will help to limit the settlements to 25mm.

My questions are:

1. Is any relation between shear strength and consolidation indexes (Cc and Cr)? I was expecting based on SPT N-values and shear strength values, these clays are could have less Cc and Cr values. Perhaps it was the influence of sampling disturbance?
2. In your experience, other than surcharge, do you have other options in order to reduce consolidation settlements?
3. Do I need to consider elastic settlement of the fat clay layers? Based on SPT N-values and shear strength values, I think that elastic modulus is high so I am not considering elastic settlements.

Thanks for your help !!

 

[oldestguy]

You might also comment on climate and potential for moisture change, such as increase (via that granular layer) or nearby trees reducing it, with resulting differential shrinkage or heave. Are there area problems with this clay damaging buildings with seasonal changes?

 

[Okiryu]

We are located in a tropical area. The clays are not expansive. Trees will not be placed around the area. But the site used to be a farm area. Nearby structures looks okay without signs of structural damages.[tt][/tt]

 

[BigH]

Why is the allowable settlement set at 25 mm? Is this because that is what is typically used for spread footings? If you pick up most recognized geo texts, they will indicate a prescriptive allowable settlement of 50 mm. Also, it seems that for a mat of that size, the thickness of the mat is on the thin side. You should check out the mats rigidity. You might wish to thicken the slab on column and wall lines.

 

[Okiryu]

Thank you for your replies. I agree that for mat foundations an allowable settlement of 50mm is shown in more of thextbooks (I have checked Bowles), however my project requirements are for an allowable settlement of 25mm. The bldg is one-story office building with very few heavily loaded mechanical rooms. The structural is checking the thickness of the mat.

 

[Okiryu]

I am talking with the Project Manager and Structural and they are considering to allow 50mm max settlements for the mat. Thanks.

 

[fattdad]

Structural loading for slab-on-grade design typically accounts for some "per square foot" load that influences the structural design of the slab. If you take this, "per square foot" load and assume that load is sustained dead load, you'll caclulate MUCH more settlement that will actually develop.

Often the slab thickness is determined by the modulus of subgrade reaction. What value are you using for that modulus?

I'm not sure that subgrade improvement for 2 ft will have much effect on the settlement of a large areal load.

f-d

ípapß gordo ainÆt no madre flaca!

 

[Okiryu]

The structural loading are:
1. Total Dead load (mat plus superstructure)
2. Floor live loads
3. Roof live loads
I am using the sum of all the above loads (and also the weight of the fill) as the additional stress (delta sigma) when calculating consolidation settlements. I am assuming that the floor and roof live loads are sustained loads. However, these live loads are much lower than the dead and fill loads, so they actually do not affect the consolidation settlement calculations. The parameters which affect more the final settlement are the thickness of the fat clay layer and Cr.
The modulus is 100 pci (although we are specifying granular soils for the fill, so "k" could be higher)

Thanks !

 

[aeoliantexan]

If I understand, you have the weight of the mat, about 375 psf, 6ft of fill, about 750 psf, and a floor load of 1000 psf, so the increase in soil pressure is about 2125 psf. At the middle of the clay layer, the effective stress is about 3375 psf. The preconsolidation pressure is about 4 to 5 times the undrained strength, say 5600 psf, so the clay is well overconsolidated, unless the superstructure loads are substantial. The compressibility of the gravel should be low.

My experience would suggest the total settlement would not exceed an inch by very much. Why not support the structure on spread footings and use a slab-on-grade floor? I have never seen a one-story office building built on a thick mat like this.

For the cost of the mat, you can probably use piers or piles bearing in the gravel, a structural floor, and a crawl space, and movements should be very small. You should still have money left over.

What am I missing?

 

[Okiryu]

Thanks for your reply, aeoliantexan.

As you mentioned, for the mat option, the increase in soil pressure is about 2125 psf and the clay is well overconsolidated. However, if we assume a 5m thick layer of clay from the surface, is not the initial effective stress about 20 kPa (420 psf) (σo’=2.5x(18-10)=20 kPa)?

Also, since the clay is overconsolidated, I am using Cr obtained from the recompression portion of the consolidation test curve (Cr=0.08) and not distributing pressures since the loaded area is large. Plugging this numbers into the 1-D Terzaghi's consolidation equation I got settlements about 150mm (6 inches). My supervisor also thinks that these settlements are too high based on the SPT values we got for this clay (6~9). I do not rely too much in the SPT for clays, so I decided to do some consolidation and triaxial tests. Shear strengths came back as expected, so no issues regarding bearing capacities, however settlements are high and controlling the foundation design, as I explained above.

As you suggested, we will be exploring also the spread footings option and see how it goes.

Thanks!

 

[aeoliantexan]

Yes, using the classical equation for recompression settlement, and Cr=.065, I get about 5 inches. I inferred from your 420 psf that the water table is close to the ground surface.

It is pretty widely accepted that such calculated recompression settlements are substantially overestimated, and that is my experience. My experience is for lower plasticity clays than yours, however. If you haven't applied Schmertmann's procedure to the lab consolidation curves, it would be worthwhile. 0.08 seems like a high value for Cr.

Perhaps you could get three inches or so of settlement. Because it is recompression, I expect that it will occur rather quickly, and most of it will be finished before you build the structure on the mat. This is consistent with your comment about the timing.

50 kPa is a large floor load for an office building. Is this correct?

One thing that bothered me was your comment about the limestone outcrops. If your mat really settles 3 inches, these hard points could create quite a problem.

I still like piles and a supported floor if you want to be conservative.

 

[Okiryu]

aeoliantexan, thanks for taking the time with this.

I am located in a tropical area and although the water table was not high, the clays presented more than 90% of saturation and I expect that they will become saturated at some point of the life of the structure, specially during heavy storms/rainy seasons. I will also take a look at the Schmertmann's procedure. When I calculate Cr, I just assume that it is the slope of the recompression portion of the consolidation curve.

The large floor load only occurs in some rooms, where we have heavy equipment storage rooms.

I did not think about the limestone outcrops. How they can affect the structure (if I get high settlements)? The outcrops were encountered occasionally where the gravel soils are. No outcrops within the clay soils layer. This is very typical geology in my area.

I was checking the spread footing option, but I still getting high settlements (more than 25mm). I like the idea of soil-treatment (since the gravel layer is at shallow depths) or piles as you suggested. For floor slab, structurally supported slab might be the solution.

Thanks again !!

 

[aeoliantexan]

I was thinking that you could have essentially zero settlement over the limestone outcrops, which would make your differential settlement equal to your maximum settlement, and possibly occurring over short horizontal distances.

If you are sure the clay is not expansive (the Atterberg limits make it suspect), you could use a reinforced concrete structural floor cast on grade and supported on piles. If you still think there will be settlement of the ground due to the fill load, you should design the plumbing, etc to accommodate the ground settling away from the floor.

Slab-on-grade floors and footings or a beam-stiffened thin slab supporting the structure are usually the lowest-cost solution for a one-story building and should be considered first.
If you will really have a 1000 psf floor load over room-size areas, you could have trouble with a slab-on-grade floor. Local surcharges could be used to simulate the heavy floor loads and reduce the settlement.

 

[Okiryu]

Aeoliantexan, thanks again...fortunately, in my area, clays are non-expansive. Also, we are planning to make recommendations for flexible connections for utilities. We are still working with the structural for this foundation. I think that will be some settlement after place the fill. I am assuming that the fill won't settle since we are specifying granular engineered fill. Will post more info here after talk with the structural.

In the meantime, do you have any good reference for slab on grade design? Specially I would like to know how the structural uses the geotechnical parameters to design the slab.

Thanks!!

 

[aeoliantexan]

I don't usually design them structurally, but a helpful reference is, "Designing Floor Slabs on Grade" by Boyd C. Ringo and Robert B. Anderson, Copyright 1992 The Aberdeen Group, ISBN 0-924659-34-3.

For lightly loaded floor slabs my advice is usually:
1. If the foundation is separate, such as foundation walls and footings or grade beams and deep foundations, provide isolation joints to allow some foundation settlement without cracking the slab.
2. If non-load bearing partion walls are supported on the slab, isolate them from the structural walls and second floors.
3. Provide a vapor barrier between the subgrade and the slab.
4. Provide nominal reinforcing such as #4 bars spaced 16" each way and provide chairs to keep them in the middle of the slab. This helps forgive minor lapses in construction quality as well as shrinkage, temperature changes, etc.

 

[Okiryu]

Aeoliantexan,Thanks for the advice!! Will look for that book..

 

[BigH]

In the meantime, you might find the following useful:

http://www.concreteconstruction.net/images/Soil Information Needed for Slab Design_tcm45-342340.pdf

http://www.buildingsguide.com/calculators/structural/FCSGPCAPL/

http://www.excelcalcs.com/repository/strength/civil-engineering/grdslab.xls/

http://www.ce.memphis.edu/6136/PDF_notes/h_slabs.pdf

 

[Okiryu]

Thanks for the help!! I purchased already the Ringo and Anderson book...

Also, I was reading again this post and it was a question there that I would like to follow up:
1. Is any relation between shear strength and consolidation indexes (Cc and Cr)?
my question is based on the fact that sometimes one does shear strength tests (unconfined compression, UU) and then consolidation tests for a similar sample a the clays result [pre][/pre]more compressible than you can expect.

Thanks again!!

 

[aeoliantexan]

Not that I know of. In my understanding, Cc is a fundamental property of the soil that does not change with water content or stress history. Cr also, although it is harder to pin down. There are empirical relationships between indices such as liquid limit and Cc.

Undrained shear strength varies with water content and stress history. There are relationships between undrained strength and preconsolidation pressure. A crude one that is easy to remember is Pc = (4 to 5) times Su.

The drained strength of clay is, in simplest terms, Tan Phi times the effective normal stress. There are relationships between the drained Phi angle and plasticity index. So perhaps one could derive a relationship between Cc and Phi.