Vane Shear Test Results 2

[rgspringer]

As part of a slope stability analysis, we performed field vane shear tests on lean clay soils (PI between 10 and 22) using a Geonor H-70. The vane shear testing was performed as a supplement to previous testing performed on Shelby tube samples which yielded lower than expected results, possibly due to disturbance of the more sensitive, saturated materials.

The vane shear device did not have a "torque" ouput, but instead, the device output was already calibrated to provide the undrained shear strength, in kPa. The test results were reduced using the correction factor suggested by Azzouz in the Journal of Geotechncial Engineering, Vol. 109, No. 5, May 1983.

My first question is how do you break down the undrained shear strength of the unsaturated materials into a cohesion and phi value? I estimated a phi based on the PI and then roughly tried to back calculate the cohesion using the following equation: c = su -(po * 0.618 *tan (phi))

where: c = cohesion
su = field vane undrained shear strength
po = at rest lateral earth pressure
phi = drained angle of internal friction
0.618 = circumference of the field vane

Is this approach reasonable?

Secondly, for the saturated clay materials, typically, phi = 0; therefore, su = c. So can I assume that the field vane shear strengths are equal to the cohesion? The problem I am having with this is that the soils in this area have been shown to be normally consolidated to slightly overconsolidated (OCR = 1 to 1.1) based on consolidation testing. For normally consolidated soils, typically, su = 0.2 * pv', where pv' is the effective overburden pressure. Based on this correlation, my undrained shear strength for the saturated clays should be on the magnitude of about 600 psf; however, I'm getting vane shear values as much as 500 psf higher than this.

Any help would be greatly appreciated. Thanks.

 

[dgillette]

No, in my opinion the approach you describe for estimating c' and phi' is not reasonable. In a sense, you have more unknowns than knowns - you have undrained strength (averaged over the sheared surface), but you don't have any measurement of the pore-pressure response at any particular location; hence, no knowledge of sigma' at failure.

Not sure having a measured strength that much larger than the predicted value should bother you. su/sigma'vc is more commonly greater than 0.2, like 0.23 - 0.26 in simple shear, and more than that in triaxial compression (up to 0.36 in CIUC, 0.32 in CKoUC).

I don't have Azzouz's vane adjustment handy, but by either Bjerrum's or Aas, LaCasse et al, the adjusted VST strength generally comes out at ~0.22 (as stated by Mesri back in the '70s). There can also be effects of aging in old soils, which may or may not show up in consolidation tests.

Finally, you say "soils in the area have been shown to be normally consolidated to slightly consolidated." What does area mean? 20 feet away? 100 feet away? in the same county? Since the apparent strength ratio is so sensitive to precon stress, you would need to know that with some precision.

Bon chance!
DRG

 

[ItsOnlyMud]

Hi, you need to be looking at effective stress parameters for slope stability, shear vanes give total stress parameters! Use effective cohesion (c') of 0kPa, and base your friction angle (Phi') on PI correlation (I'll try and find one for NC clays) or do expensive consolidated undrained triaxial tests on undisturbed samples, and treat the results with caution!

 

[KVgeo]

@rgspringer

1st question)

If you are asking this question you are not qualified to do a slope stability analysis.

This shows a fundamental lack of understanding in basic geotechnical engineering.

Read a text book or ask a senior engineer to explain to you the difference between undrained and drained shear.

I don't think this forum exists to answer basic questions that you do not have the time to research in any undergraduate geotechnical text book.

2nd question)

undrained strengths from a vane are snapshots of the material at the time of the test. Undrained strengths are influenced by:

- Effective stress conditions
- Degree of saturation
- Overconsolidation ratio

the strength you get from a vane in a material at a depth of 5m may not be the mobilised peak strength of that material when you strip of 3m of overburden in the process of constructing your slope.

Recommendations

Invest in a text book. Get your work senior reviewed. Look for a mentor.

 

[rgspringer]

@KVgeo

I appreciate your input. Perhaps a little more clarification will help.

In regards to my 1st question, I don't feel that I lack a general understanding of drained and undrained shear, but I more lack experience with the interpretation of the field vane shear test results, more specifically, the output of the vane shear device since I was not on hand to witness the vane shear tests.

After my review of publications regarding field vane shear testing, my general understanding is that field vane shear testing performed in saturated clays theoretically engages only the cohesion portion of the shear strength as the frictional component is theoretically offset by the pore water pressure.

In my review of publications (textbooks, journals, etc.), I couldn't find an equation for determining the cohesion component for field vane shear tests performed in relatively unsaturated clayey soils where the pore pressures are relatively small and the soil grains can press together, thus engaging the friction component. As previously mentioned, I can estimate a internal friction angle value based on the PI of the clayey soils, and simply assume that the cohesion is the difference between the undrained shear strength (as determined by the field vane shear test) and the frictional component (based on the lateral earth pressure at the depth of the test) since the lateral pressure acts normal to the vane shearing surface. Again, this is for a test performed in a relatively unsaturated soil; therefore, total stress conditions will apply. The question is, for vane shear test results, is this a reasonable way for back-calculating the total cohesion for unsaturated clay soils or am I oversimplifying?

In regards to the effective stress condtions, my reason for needing both drained & undrained shear strength parameters is because not all of my soils will be saturated and because sudden drawdown analyses require that I analyze failure for both drained and undrained conditions.

As stated in my original post, OCR of the slope materials, based on limited consolidation testing, is about 1.0 to 1.1, or normally consolidated to slightly overconsolidated.

I didn't think to consider the degree of saturation, which is an excellent point that I will take into consideration.

Additionally, the slope being analyzed is an existing slope, so no overburden removal is expected.

I hope this additional information provides some clarification, and any additional input would be appreciated. Thanks.

 

[dgillette]

"After my review of publications regarding field vane shear testing, my general understanding is that field vane shear testing performed in saturated clays theoretically engages only the cohesion portion of the shear strength as the frictional component is theoretically offset by the pore water pressure."

This, I believe, is a misunderstanding, and it is leading you to try to find a connection where none exists. The undrained shear strength is often referred to as "cohesion," because it is convenient to put it into stability and bearing calculations as c=X and phi'=0, and this sometimes causes confusion. The VST engages both phi' and c'. It has to; strength is always controlled by effective stress, and effective stress does not generally go to zero in the VST (even though there is usually high excess PWP). If you do undrained triaxial tests on NC clay with PWP measurements, you will typically find that the effective stress strength envelope so obtained has a cohesion pretty close to zero. (On OC clays, there may be somewhat of a cohesion intercept.) In theory, if you could somehow know the effective stress state surrounding the vane, you could figure out the friction angle. But you don't know that, so you are better off just trying to estimate phi' from published correlations.

The difference between the effective stress and undrained envelopes is the excess PWP generated in undrained shear. If the material is not saturated, you would expect little or no excess PWP, because the pore fluid, part of which is air and water vapor, is orders of magnitude more compressible than liquid water. In partially saturated material, you can also get a significant apparent cohesion caused by capillarity, so yes, the degree if saturation can matter quite a bit.

 

[rgspringer]

@dgillette

Thank you for your comments. I now have a much better understanding of the field vane shear test. I was basing most of my assumptions on the vane shear text from Spangler and Handy's "Soil Engineering". I now understand that even though the vane shear test results are "sometimes taken as a measure of cohesion", that doesn't necessarily mean the tests aren't engaging the friction component of the shear strength. Thank you again.

 

[moe333]

One additional point is that Duncan and Wright (2005) suggest that there no excess PWP for saturation levels below 70%.

 

[KVgeo]

@ all above...

While I can understand your desire to help educate young engineers, it has become clear to me that replying to questions such as the one above may not be the prudent course of action.

It worries me that somebody performing slope stability analyses does not understand the fundamentals behind undrained shear. Concepts which are explained at length in any undergraduate soil mechanics text book.

The best advice rgspringer needs is to be told to invest some time in understanding basic geotechnical principals or asking his senior supervising engineer.

If you disagree with me, then please reply with your reasons why.

@rgspringer,

I am not trying to pick on you. Unfortunately, this post has broken the proverbial camel's back for me. Geotechnical engineering is not easy and is based on experience and empirical relationships. As such, practising geotechnical engineers need to be aware of the fundamentals behind the basic concepts.

Your post would worry me less if I knew senior engineers were reviewing your work. If so, you would be better off asking them questions about undrained shear.

In future, if you are asking questions for clarification on fundamental geotechnical concepts they should be phrased as hypotheticals as opposed to specific projects. Otherwise, I will continue to get very worried.