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.
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
Answer Verified By: Nitharshan
Dear Mr. Dennis,
Thanks for your valuable information and I have small doubt in your Reply 03 => If the PWP calculation is not correct which means the effective pressure calculation is not correct, then how B is used for the stability analysis?
Normally people using MC model for its simplicity (input parameters) so please tell based on your experience if anyone does the deep excavation using MC Undrained B does it falls in the safer side (over-estimate the forces and deflection) or dangerous (underestimate the forces and deflection)?
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.
Dennis Waterman.
Really thanks for your valuable information.
Sorry only 1 last question, if use undrained A, let's say we are using laboratory results
For undrained A => effective parameters from CU test effective parameters
For drained => effective parameters from CD test
am I correct?
Drained and Undrained A both need effective strength and effective stiffness, which can only be obtained directly from a CD test.
A CU test gives effective strength, but undrained stiffness. It is possible to indirectly (through simulation in SoilTest) determine the effective stiffness from the undrained stiffness.
With kind regards,Dennis Waterman
Thanks for the reply, so it is better to use CD test eventhough we have some correlations from undrained stiffness to effective stiffness because effective strength values vary for CD and CU tests.
It is better to do a CD test, but for clay that is almost impossible ....
Why would effective strength values vary for CD and CU tests? They should give the same strength values, unless the CU test was not properly consolidated. (hence the test started when there were still excess pore pressures in the sample).
By referring to this paper.
PDF
Thanks,
Nitha.
Ok, so you found 1 paper for 1 specific type of clay somewhere in the world ....
And there are many reasons why not to trust this paper:
- The use the assumption in advance that the CU results are not good and the CD results are while that should have been investigated. So they're not investigating, they're looking to prove an assumption without explaining where that assumption comes from.
- There is no information on whether the CU test was properly consolidated and the CD test excess pore pressures.
- When doing a 15-21 days CD test on soft clay no doubt there are effects from for instance creep (and creep induces excess pore pressures, for instance)
- There may have been effects of suction around the failure area for the CD test, and suction may lead to higher strength for soft clays. But such effects cannot just be considered for general cohesion and friction angle values.
The biggest danger of using data from publications is that they're often for 1 specific soil under very specific conditions, and any conclusions cannot be used for other soils in other parts of the world under different conditions.
Noted, thanks.