estimating subgrade modulus 2

[durtboy505]

I'm in geotech and I'm trying to give a useful number to the structural engineer for designing a mat slab on very soft saturated clays. If using conventional footings and isolated floor we are estimating primary settlements of six inches max and 3 inches differential, and deep foundations not an option with the client and site location. Using some models I'm getting K values between 1pci and 10pci. Yeah. I know. that sucks!

I found a closed thread from January '10 and Mr. OCI had the following to say: "Here is a graph that I have used occasional for a quick and dirty subgrade modulus. Obviously not a replacement for a more accurate value, but better than nothing.

http://files.engineering.com/getfile.aspx?folder=dd044c6a-fca4-46eb-bb16-bf"

Using OCI's graph I'm getting 25 pci. Thanks OCI! If you are out there, Can you tell me where that graph is from?

Anyone else want to chime in on estimating subgrade modulus?
Cheers
Durtboy505

 

[rowingengineer]

Talk about sticking your thumb up your proverbial and smelling it twice. I assume your claiming that the 25 pci/in can be extracted from the graph that stops at 95pci/in, there is a reason it stops at 95pci/in because after that dirty guesses/global rules of thumb are no longer acceptable.

Just because the computer programme is giving you answer you don't like doesn’t mean you can justify trying some rules of thumb. Start looking at local results, unless this is for some oil rig in the middle of the deep sea, I doubt you’re the first to do a soils test at in your area and probably not the first to run into this situation, find out what they did, look at how there building is performing. Maybe consider a plate load test or similar, but this again will only get you within the ball park. At the low bearing pressure you are talking about you coefficient of variation will be probably 200%. And if you are not the man whom signs off on your reports, go see the man that does.

Maybe Big H will stop by and give you some more testing methods/geotehcinal engineering advice rather than the some young structural engineer. And for what it is worth here is another useless graph for you.

http://www.slideruleera.net/CBR_Correlations.pdf

ANY FOOL CAN DESIGN A STRUCTURE. IT TAKES AN ENGINEER TO DESIGN A CONNECTION.”

 

[Ron]

durtboy505...25 pci is relatively low, but in the ballpark for the material you've described. Oddly, the subgrade modulus for structural considerations is considerably different than the subgrade modulus for pavement purposes. Given the same materials, the subgrade modulus for pavement purposes will be 2 to 2-1/2 times the structural value.

As RE noted...BigH...where are ya?

 

[BigH]

No correlation is useless; as with all correlations, one needs to understand the underlying assumptions and have a wealth of judgement to determine the usefulness and reasonableness of correlations.

In reading this, I wonder if the subgrade modulus is even applicable to very soft saturated compressible clays - that take the load initially by the incompressible water; then will undergo consolidation under the loads. While in materials that, under load, will behave in an "elastic" fashion, it makes sense in that the ks value is applicable to elastic materials, clays that will undergo consolidation in the virgin range - is this really a compatible design assumption?

How stiff is the slab? is there any flexibility (check out Polous and Davis's book on Elastic Solutions for Soil and Rock Mechanics. Is the kv analysis using springs that are all taking equal support or does the analysis allow for non-uniform spring values? (e.g., intermediate spring concept).

Indian codes take into account the "relative stiffness factor" and column spacing to determine if a mat is applicable or a beam and slab is a better choice of analysis. (IS:2950 Part 1-1981 - Design of Raft Foundations (C31)) There is a text by Naiman Kurian "Design of Foundation Systems" that might be of interest - he goes into many aspects of subgrade modulus - it is not a "single" system concept. In this book, they don't even attempt to give a value range of kv if the undrained shear strength is less than 50 kPa (1000 psf) (and 50 is the upper limit of firm (medium stiff) clay and lower limit of stiff clay).

If one truly has a soft (very soft) clay condition, it might be better to subexcavate for a depth, backfill with engineered fill and use spread foundations or grade beams as the load transfer support system letting the engineered fill to act as a flexible raft.

I suppose that I am judging that the subgrade modulus approach (Winkler) is really applicable to the conditions as stated in the OP - a Sunday night observation!

 

[dik]

I designed a 474 taxi and hanger about 20 years back with a subgrade reaction of 25 pci... pretty soft stuff... and it's still 'kickin'.

If I don't have real geotech info, for our firm clays here I typically use 50 pci and for real stiff ones use 75 pci. For good granular material, use 150 pci and for really good granular use 200. For compacted granular sub-base and well compacted good granular base with quality concrete and excellent quality control I've used as high as 300 pci... I've not had any problems with these values... might be that they are overly conservative, but it's a price the client pays for saving on a geotech report...

Dik

 

[rowingengineer]

Sory Big H, I did mean the graph was useless to his problem. I agree with Big H, the Kv no longer applies.

ANY FOOL CAN DESIGN A STRUCTURE. IT TAKES AN ENGINEER TO DESIGN A CONNECTION.”

 

[dik]

That should be 747... Dik

 

[rowingengineer]

Dik,
just a quick note, I think he is the geotech.

ANY FOOL CAN DESIGN A STRUCTURE. IT TAKES AN ENGINEER TO DESIGN A CONNECTION.”

 

[B16A2]

No kidding, I thought the geotech was supposed to be the expert in soils, why is this question being posed to structural engineers?

 

[dik]

We're good! That's why!

Dik

 

[Ron]

Yeah, dik! That's it.

I worked for many years (18) in a large, predominately geotechnical firm (at least that's what started it in the 1940's). Our geotechs did not understand subgrade modulus. Keep in mind that these guys were (and still are) considered some of the better geotechs in the world. They could not get the difference between static subgrade modulus (mat foundations) and dynamic subgrade modulus (pavement design). A question that was routinely posed to me (one of only one structurals in the branch office, and one of only 20 or so in the company)was "why do the structural engineers always say that our subgrade moduli are too high for mat design"? My response was that they were only considering the modulus as if it were applied to pavements.

While it goes against conventional logic, the pavement modulus is typically higher because the dynamic loads are so transient. For a mat foundation, the load are there, ad infinitum.

I suppose it helps for my understanding that I ran a geotechnical lab for a few years....was able to make the connection between structural and geotechnical applications.

 

[dik]

We just had the project manager from Toronto for a large project in Saskatchewan in our office, reviewing the progress of the project. During his review he mentioned that the client was especially pleased with the work coming out of the Winnipeg office.

I rose to the occasion and to the group clearly stated it was because we are that good! It was well received.

Dik

 

[durtboy505]

Thank you everyone for all the insights. I work in a land of dessicated hard dry clays, that are usually poised to swell. This site was not something we see even once in a blue moon. I posted this to the structural engineer forum because I was hoping to get some feedback from structural engineers as well as geotech gurus. I was having difficulty formulating the words to explain to my structural associate on this project what BigH so skillfully described for me in his response - that the Kv was not applicable, and giving him one to plug into his black box may not be appropriate. Thanks again everyone!