Undrained analysis A, B (and C)

Question

Hi all, 
 I have some questions in these types. 
 
 01. Only consolidation analysis is possible after the undrained analysis in A and if it is B then manual said Su is needed to update. Please explain this statement.

02. Based on undrained A, it over estimates the Cu of the soil in MC but not in advanced model like HS method A (as it matches like real soil behaviour) , can you explain how ?

03. MC model is not correct model and lets say if I'm going to use MC model, which type is more suitable as every types has pros and cons for example A=> correct PWP calculation but over-estimates the Cu , B=> PWP calculation is not correct, C => total stress analysis only.

04. How PWP calculation in undrained B is not correct? and why C is not available in HS?

One literature paper indicated like this for MC, A and B

Thanks.

Dennis Waterman · Accepted Answer

Hi Nitha, 
 01) Undrained shear strength is specific to a certain effective stress state. So if the effective stresses change, the undrained shear strength changes. A consolidation analysis increases the effective stresses and therefore also increases the undrained shear strength. When using Undrained B with consolidation that means that the user must update the undrained shear strength to match the new stress state due to consolidation. 
 02) It has to do with the fact that Mohr-Coulomb considers the soil to be elastic and other models like HS to be elastoplastic. A full explanation would fall outside the scope of this forum... 
 03) Mohr-Coulomb doesn't give correct PWP in neither Undrained A nor Undrained B, so your initial assumption is not correct. So in principle M-C can be used for Undrained B and Undrained C for just undrained stability analysis - no consolidation, and not for deformations as deformations are unreliable and sometimes completely wrong. 04) Undrained B can only be used with the elastic Mohr-Coulomb model, and the assumption that soil is elastic is what gives wrong PWP. So it's not so much that Undrained B gives wrong pore pressures, but the Mohr-Coulomb model that gives wrong pore pressures. 
 The HS model is an elastoplastic model and the elastoplasticity as used in the HS model depends on effective stresses. Therefore the HS cannot be used with Undrained C as Undrained C only calculates total stresses. If you want to calculate with elastoplasticity and Undrained C we recommend using the NGI-ADP model. 
 With kind regards, 
 Dennis Waterman

Dennis Waterman · Answer

Dear Nitha, 
 With an Undrained B analysis the user specifies the strength of the soil by the undrained shear strength. So for the strength it doesn't matter if the effective stresses are correct or incorrect, because independent of the stresses the soil will always fail at the same undrained shear strength. So if soil failure is important, Undrained B with Mohr-Coulomb works. If you're interested in structural forces, deformation etc, then Undrained B is not good. So using Undrained B with Mohr-Coulomb for a dam or embankment or a slope stability without reinforcement should work fine, but for excavations it is not since the structural elements are important there. 
 We see a lot of people doing deep excavations using Mohr-Coulomb and Undrained B. However, when comparing results with for instance Hardening Soil and HSsmall we see that if the factor of safety on the soil is matched, the Mohr-Coulomb calculation typically gives lower structural forces and deflections - we have seen structural forces using Mohr-Coulomb that were 20-30% lower than using HSsmall. Considering that HSsmall has a far better prediction of the soil behaviour it's reasonable to say that using Mohr-Coulomb for excavations the user should take into account that the structural forces in the wall and anchors are underestimated. 
 By the way, I mention HSsmall here .... generally the HS model gives larger structural forces than the HSsmall model. Again, if we reason that HSsmall models the soil behaviour in a better way than HS does it seems logical to conclude that the Hardening Soil model often overestimates the structural forces in a deep excavation. How much exactly highly depends on the ratio between stiffness of the wall and stiffness of the soil. A stiff wall in soft soil seems to give more overestimation than a very flexible wall in stiff soil, but this is an observation based on limited data. 
 With kind regards, 
 Dennis Waterman.

Dennis Waterman · Answer

Dear Keviin, 
 No, that is not correct. M-C will for sure overestimate the undrained strength for a normally consolidated soil. The Hardening Soil model may be able to predict the correct undrained shear strength, but depending on the stiffness parameters may still overestimate it or even underestimate it. And that is not unlikely, it happens. 
 So if you chooise certain stiffness parameters you should check with for instance SoilTest what the resulting undrained shear strength is for the appropriate stress path (since undrained shear strength is not a single value, but different for different stress paths) and you may even want to double-check after your calculation if for instance the resulting (sig1 - sig3)/2 in your model doesn't exceed the undrained shear strength you got from a field or lab test. 
 With kind regards, 
 Dennis Waterman

Dennis Waterman · Answer

Hi Kelvin, 
 Mohr-Coulomb is an elastic-perfectly plastic model, which means it's am elastic model with a failure criterion. As long as the soil doesn't fail, the soil behaviour is elastic, whereas in the Hardening Soil model the soil already behaves elasto-plastic when the stress state is not at failure. 
 So Mohr-Coulomb is nothing like the Hardening Soil model and Mohr-Coulomb is just not a realistic representation of soil behaviour. According to Mohr-Coulomb if you compress soil and then release it again the soil bounces back into it's original shape as if it was rubber without any permanent deformation. That is just not how soil behaves... 
 With kind regards, 
 Dennis Waterman

Undrained analysis A, B (and C)

Top Replies