Coefficient of active earth pressure for clay vs sandy clay 3

[elperko]

Hi,

I'm trying to calculate active earth pressure on a retaining wall. One thing is bugging me; it makes sense to me that a more cohesive soil will 'hold itself together' better than a less cohesive soil (i.e. a clay vs a sandy clay) and will therefore apply a lower lateral load to a retaining structure. Using a table for approximate soil properties I am getting a higher angle of effective internal shear resistance for a sandy clay, which in turn leads to a lower coefficient of active earth pressure and lower lateral force. Could someone explain why this is the case?

Thanks,

Bradley

 

[GeoPaveTraffic]

Cohesive soils will have a higher earth pressure than non-cohesive soils because they have a lower effective (long term) friction angle.

You comment about "hold itself together" actually refers to undrained strength (short term).

Mike Lambert

 

[Ron]

It appears that you are using the simplified Rankine equation of Ka=((1-sin(phi))/(1+sin(phi)). This equation is valid for cohesionless soils, but not necessarily so for cohesive soils. Take a look at the Mohr-Coulomb approach for cohesion. You'll see the difference.

As an exercise, plot the Rankine values for Ka against tan(phi)....an approximation of cohesion. You'll see that the lower the phi angle, the higher the cohesion and thus the greater reduction in Ka under that premise. That premise is only valid for Ka-tan(phi)> 0, so you'll see a break point at phi=23.5 degrees......which implies the break point between cohesive and non-cohesive soils with respect to their effect on Ka.

 

[LRJ]

Ron, what would you recommend for Ka for cohesive soils then? What's the equation for that? Is it a case of substituting the function involving phi for a function involving undrained shear strength?

 

[oldestguy]

Before going too far, please describe the PI, LL and what climate? High shrink swell clay can be real difficult to depend on. Changes in moisture content can raise heck with things. You might want to use granular backfill.

 

[Doctormo]

Bradley - I look at "Ka" as a positive force resulting from the phi angle calculation and c being a negative force or tension holding back the soil.

From Bowles, the earth pressure at a depth:

σh = Ɣ•Z•Ka - 2•c*sqrt(Ka)

The first part is the typical triangular distribution from the phi angle and the second part is a fixed amount of negative pressure that the cohesion provides. The frictional part has to overcome the cohesive part at some depth to provide a net driving force thus there is no simple Ka relationship for cohesive soil earth pressure.

However, as Mike noted above, there is short term and long term strength properties to be concerned about. Cohesion decreases over time and may eventually go to zero at some effective frictional strength. Soil also cracks when in tension thus the cohesion benefit is relieved thus increasing pressure, The crack can also fill with water which increases lateral pressure.

Most people use low phi angles to approximate cohesive pressures so a simple triangular pressure diagram can be used at a high enough magnitude to be safe. A Ka = 0.50 (60 psf/ft) might be a reasonable value if one was to guess a number for a decent clay.

As oldestguy notes, clays are quite difficult with different issues depending on where one is at. Many use granular backfill material to avoid the difficulties of predicting the lateral earth pressure from clay.

 

[fattdad]

I'm not going to agree or disagree on Dr.Mo's assertions on what, "Most People" do.

For long-term stress conditions, we have long term strength conditions. Such long-term conditions relate to the effective stresses and the drained strength. We measure the actual drained strength via the triaxial or drained direct shear test. We also correlate it to soil classification and for granular soils relate the drained strength to SPT N-value.

In both direct shear and triaxial strength tests, we will detect a drained cohesion intercept. Some folks like to determine the drained earth pressures allowing for some measure of cohesion, where you will realize some benefit in fine-grained soils or course-grained soils with some activity in the fines.

I ignore drained cohesion, but use the actual friction angle after allowing some consideration for strain compatibility. I mean I would not assemble a failure envelope on two points where failure was at 4 percent strain and one point where failure was at 15 percent strain.

I do not rely on the laboratory certificate's interpretation of friction angle. It is professionally negligent to let a lab provide such professional service! (Bear in mind the ASTM Specification excludes data interpretation for phi and c.

f-d

ípapß gordo ainÆt no madre flaca!

 

[Okiryu]

I just recommended an equivalent fluid density for clays very similar to what Doctormo noted above for a permanent retaining wall. For temporary excavations, we use the undrained shear strength of clay (c different to zero and phi=0).

Also, do you think that UU triaxial tests for compacted clays can provide acceptable/reasonable values for phi'? A simple reasoning (but not sure if it is true) for this is because compacted clays are not saturated so if we do drained triaxial tests "there will be not too much to drain" so will be some similitude with UU tests. Because sometimes we use clays to backfill against retaining walls, I was trying to check the phi' of compacted clays...

 

[Doctormo]

fattdad - "I'm not going to agree or disagree on Dr.Mo's assertions on what, "Most People" do." - Should I have said "many people"?

I have looked at thousands of geotechnical reports from around the US and almost every one provides an equivalent fluid pressure table or text for retaining wall pressures that are free to rotate. Granular backfill is typically 35-40 pcf/ft and site soils are 50-70 pcf/ft and many recommendations just mimic the IBC values. Thus I concluded that "most people" do this.

That being said, I agree with your technical approach assuming one has enough information to make such an assessment of the retained or backfill soils. The same reports I refer to rarely have enough testing done to determine the parameters necessary and the contractors seems to find the worst fill on a site to backfill walls when granular is not required.

 

[LRJ]

Also, do you think that UU triaxial tests for compacted clays can provide acceptable/reasonable values for phi'? A simple reasoning (but not sure if it is true) for this is because compacted clays are not saturated so if we do drained triaxial tests "there will be not too much to drain" so will be some similitude with UU tests. Because sometimes we use clays to backfill against retaining walls, I was trying to check the phi' of compacted clays...

UU triaxials do not measure effective stress (they are a total stress test) so it is not possible to derive effective stress parameters such as phi.

 

[Okiryu]

LRJ, yes I understand your point, however for unsaturated and compacted clays I got phi values of around 15 degrees which made me think that since the samples were unsaturated, the UU Tests in this particular case may simulate a drained triaxial tests. But anyway, best to run drained triaxial tests to get the correct long term drained parameters for phi'. Thanks !!

 

[LRJ]

To get effective stress parameters you need to measure pore-water pressure so you can derive effective stresses. Unconsolidated Undrained (UU) triaxials do not measure pore-water pressure - at least no apparatus I've ever seen does. I'm therefore unsure how you would've derived the phi angle since you wouldn't have an effective stress path.

I'm not an expert in unsaturated soil testing but I would've thought if the soil was not fully saturated then the pore space would be occupied by (compressible) air, thereby undermining the requirement for constant volume in an undrained test.

 

[Okiryu]

LRJ, that's exactly my point. Because air is "filling" the voids changes in volume due to loading may simulate consolidation / drained conditions... I remember that one of my professors was explaining something like this also. Thanks again for your response.

 

[Okiryu]

LRJ, also just want to mention that I understand that my statement above is not theorically correct, but I was looking for answers about why I am getting phi values in an UU test. For my unsaturated clays the above was one of the thoughts I had...

 

[LRJ]

What I really don't understand is how you have derived a phi value from a UU at all. Do you have a stress path? If so, what does it look like? What values are presented on the axes?

 

[fattdad]

if the soil is unsaturated and you run three UU tests (unconsolidated, but at different cell pressures), you will get an increase in shear strength directly associated with the increase in sigma3 - the total stress of the cell pressure. You see if there is air in the soil, the undrained increase in cell pressure will result in changes to the soil stress as air is squeezed into solution.

The phi=0 condition is very typical for saturated clays in UU loading. Heck, we get one or two lectures on this topic in graduate school! Reality is though, many soils below the water table remain unsaturated. So, what's the soil strength for design?

In my practice, I require all UU tests be performed at 3 different cell pressures. That way, I know to what extent the phi=0 condition is realistic, and to what extent the soil is actually saturated.

f-d

ípapß gordo ainÆt no madre flaca!

 

[LRJ]

Thanks for the explanation, fattdad.

 

[Okiryu]

Also, Holtz and Kovacs book has a nice explanation about why you can get phi value in UU tests in unsaturated clays. Basically, as LRJ originally mentioned, air voids get compressed and strength is increased...BTW, the UU tests we run also had 3 different cell pressures as f-d mentioned...I am attaching the Mohr envelope figure from H & K book for your reference...