undrained shear strength

[AK92]

Has anyone tried to correlate clay undrained shear strength with basic index properties such as water content, LL, PL, overburden stress, etc...? How successful would that be?

 

[rilxouss]

You need to have a good amount of soil data to establish a good correlation. Less hassle if you are only looking at one soil type.

Depending on the project scale, sometime it would be easier to directly measure the actual properties that you are interested with rather than to develop a correlation?

 

[AK92]

@rilxouss.

Yes, I agree that it would be easier to directly measure the shear strength either with a vane or a cone, but for some reason, some jobs have SPT sampling with only extensive index testing and limited triaxial testing.

In many textbooks they mention that the undrained shear strength is not a fundamental soil property, but a function of the effective stress.

I guess that for saturated clays, it should be some function of the effective stress, water content/void ratio, PI/LL/liquidity index. I wonder if we can use the consolidation test data for some help as well.

I have not tried this before, but is it possible to plot the void ratios and estimated confining stresses onto the e-log p' curve, and extend those points to the virgin consolidation line using the recompression slope to obtain a guess at the OCR?

This would then give us the OCR at many different points and allow us to generate an OCR profile of the whole site.
Then, we can back-calculate the S factor in Su/p' = S(OCR)^m using whatever undrained shear strength points that we have.
After that we can estimate all the undrained shear strengths using Su = S*(OCR)^m * p' where OCR can be determined using the void ratios.

Would an approach like that work?

 

[fattdad]

I'd rather trust correlation to pocket penetrometer on SPT samples than N-value or correlation to index properties.

f-d

¡papá gordo ain’t no madre flaca!

 

[tsek]

If you want to use SPT's, you may divide the SPT N-value by 10 and that gives you the undrained shear strength in tsf. Some like to multiply this number by 2/3. Obviously doesn't work on WOR and WOH materials.

 

[BigH]

Check out Leonards Foundation Engineering Handbook - Chapter 2 I believe and he discusses the Liquidity Index . . . there has been some correlations to this in the past . . . As for the N value, any standard geotechnical text gives approximate Su vs N value (use with caution - best to have developed your own site specific correlation).

 

[MelCao]

Better not to use this correlations,It could give you a range and often big range. You can't use this formulated correlation to calculate and use the result.

I beleive the water content, LL, PL tells you how much pores in the soil, and the natural of the soil particles in terms of how cohesive they are. usually you can tell that a soil is pretty soft and weak if the water content is high which could imply that the soil is quite loose. and the PL, LL high could mean the plasiticity of the soil and consistance of the soil.

In our practice as a designer in a contracting company, we don't really have too much information from the client due to cost issues. we seriously just estimate the Cu based on the description in the geotechnical report, like firm, stiff, very stiff, hard, dense, very dense, highly weathered etc. of course also have a look at the SPT if they did that, which just double confirm your assumption of the soil.

 

[richardx1]

Use civillabs.com to check your calculation