Soil shear strength in general 2

[Kereo]

Hey all:

Probably didn't phrase it properly

For a typical Triaxial test, is it possible to determine the shear strength of the entire soil layer lets say 5 meters homogeneous, just by sampling at different depth and interpret a "critical" value from the stress path?

Secondly, I'm trying to relate Mohr-Coulomb theory to the results obtained from Triaxial test; If I recall correctly, for a typical CU test on NC Clay, this should be the stress path:

But I'm quite confused when the MC model is mentioned, I mean shouldn't the result from CU test resemble MC theory already? How come a straight vertical stress path is deduced (which consequently overestimates shear strength):

Cheers

 

[GeoEnvGuy]

For NC clay see attached post to look at the stress path in 3d q, log p, and eo space.

https://www.eng-tips.com/viewthread.cfm?qid=447196

The MC theory draws a straight line to show that as shear strength increases, the effective stress path doesn't generate any positive or negative pore pressures to increase or decrease the effective stress.

The first stage of site investigation is desktop and it informs the engineer of the anticipated subsurface conditions. By precluding the site investigation the design engineer cannot accept any responsibility for providing a safe and economical design.

 

[Kereo]

Hi GeoEnvGuy, thank-you for referring to the previous post.

As for the straight line in MC theory with the constant effective stress, is that the numerical model derived from using software such as Plaxis? If that's the case, why do we have to use software to derive the shear strength when we have fairly accurate readings from Triaxial test?

It seems to me that the Triaxial test is only for deriving strength parameters such as cohesion and friction angle. However, in most online resources I've seen, there's hardly any that mentions the use of software to derive shear strength after the test is done, they seem to have just taken the result from the test as "true" shear strength of the soil.

 

[GeoEnvGuy]

Triaxial testing does provide an estimate on shear strength but you need to take it with a grain of salt. See effects of sample disturbance figure 3 and 5 which expand on Charles ladd and DeGroots work. Also the differences between triaxial compression, extension and direct simple shear need to be considered as they pertain to a failure geometry.

https://www.google.com/url?sa=t&sou...FjALegQIAxAB&usg=AOvVaw0JGE_br11mQdI_PEZy9dSS

I am not following your comment on numerical models deriving shear strength, from my experience you tell the software how to determine the strength of a particular stratum or specify a strength in cohesion in thin layers. If you tell it MC and give it a phi it determines the specified effective stress or total stress and multiplies by tan phi then adds cohesion. With all software models garbage in and garbage out and a skilled tech savy person can get the software to show you whatever result you want. The hard part in numerical models is knowing what pore pressure to use on which stratum and how to get the software to implement it, example trying to use a b-bar for clays following construction.

Determining what shear strength to use is a combination of field in situ testing and lab testing results. How you apply the findings depends on how the material was deposited, placed fill, alluvial. Glacial, etc.

The first stage of site investigation is desktop and it informs the engineer of the anticipated subsurface conditions. By precluding the site investigation the design engineer cannot accept any responsibility for providing a safe and economical design.

 

[LRJ]

You're right that constitutive models should replicate the observed behavior from laboratory element tests. Before running a full numerical model a single element driver should be run to confirm that this is the case. All too often this is not done (e.g. particularly when structural engineers think they can analyse soil...), which is a recipe for all sorts of problems.

The Mohr-Coulomb model is a very basic model and the predicted stress path is very unlikely to be the real stress path experienced by the soil. To better replicate the stress path you'd need to consider better constitutive models. Ultimately you could get a reasonable prediction with any constitutive model if you alter the input parameters accordingly to replicate the behavior which is most critical for your design.

Also, it is worth noting that while the Mohr-Coulomb model indicated above would overpredict the strength of normally consolidated (i.e. contractant; moving to the left on the stress path) soil, it would underpredict the strength of overconsolidated (i.e. dilatant; moving to the right on the stress path) soils.