Mr.Navid,

As you mentioned that liquefaction is not considered in the analysis so i feel Mohr - Coulomb model is simple and easy to work..now, even coupled water - soil analysis can be performed using this model .. many literature are available related to your question..may you also refer to the paper 'Goertz-Allmann, B. and Wiemer, S. (2013). Geomechanical modeling of induced seismicity source parameters and implications for seismic hazard assessment. Geophysics, 78(1), KS25- KS39' . ok all the best in your research.

The MC is a failure model. I think you refer to "elastic-perfectly plastic" constitutive model with MC failure law. The answer is that is possible.

Please refer for example to the input tab of plaxis: you can use the undrained models (then allows the calculation of overpressure) with several models, or in flac fixing the flow=off.

In any case, I suggest you make a comparison between an ideally plastic MC model and an Hardening Soil Model: the use (almost) the same input data, because the results could be very different, and connected to the deformation results (then to the constitutive model)

Dear Navid,

I would advise against using the MC constitutive model to simulate undrained behaviour, especially if you will shear the soil domain beyond the "elastic" limit of the soil. The MC-model is a good first approximation for drained monotonic loading of granular soils but tends to overestimate effective stress values when going beyond the small-strain limit. Additionally, the original model does not include any stress- or strain-dependency of stiffness (Brinkgreve, 1997) and does not include a strength reduction component (Brinkgreve, 1997) which is needed to simulate cyclic behaviour.

Therefore, I would advise against using this model for complex simulations such as the one you are proposing. However, the MC-model will provide reasonable results within the small-strain range with monotonic loading. Furthermore, and as said before, if you are not going to include liquefaction as part of the anlaysis, then the HS/HSs model would be more suitable for this purpose, since it incorporates stress- and stress-path-dependency of soil stiffness, it includes an expression for pre-consolidation stress and can model hysteretic behaviour(Brinkgreve, 1997).

I would also recommend you to refer to the state-of-the-art publications regarding constitutive modelling to find out precisely which type of calibrations you would need to do before embarking in modelling your project conditions in a geotechnical engineering software (such as PLAXIS).

Hope this helps,

Regards,

Hugo Portugal

The drained strength limits are captured by MC rather well. And the drained limits can be used to make rough estimates.

In soils with relatively good drainage, loading cycles within drained limits will produce gradual increase in soil stiffness together with some excess Pp. This can be especially notable in sand which was recently liquefied or freshly deposited, where loading cycles within drained limits can cause rapid recovery of both stiffness and dilativity. Thus, potentially increasing resistance to liquefaction, as well as increasing soil stiffness.

When drained strength limits are crossed, soil will start losing stiffness, but generate extra strength. The water has to "seep into" the voids, which takes time. Thus, loads exceeding the drained limits can produce less deformation in undrained than drained, as negative (vacuum) pore pressures locks the soil grains together. However, great caution must be taken to check for one way and two way loading. One way loading cycles will produce slow increments of deformation accumulation, while two way loads, exceeding drained limits, can cause extremely fast liquefaction.

So, MC allows to approximate some general "trends" in soil behavior during earth quakes or other cyclic loads. Obviously, it has to be done with great caution. No model existing today can predict pore pressure accumulation fully correctly. Nevertheless, MC can be used to make quick approximations in early design stages:

1. Within the yield limits - frictional soils tend to shrink (excess Pp), which causes reduced undrained stiffness (lower damping in undrained mode), but potentially recover drained stiffness.

2. When yielding, frictional soils tend to dilate (negative Pp, the "boot effect"), dampen the oscillation and lose stiffness in the process. And the soil is more stable during "one way" loading, and extremely unstable during "two way" loading.

P.S. the mechanisms governing changes in drained soil stiffness is a poorly understood. It is a phenomenon which is not accounted for by contemporary soil models. Nevertheless, sand stiffness can change not only due to Pp buildup. Drained soil stiffness does change, and it is plausible to generate lower drained stiffness at higher density within controlled laboratory environment (a paradoxical disturbed soil state, currently not captured by numerical models).

P.P.S. A potential option is to run simulations with half the friction angle. This is often described as position of "Phase transformation line" (PTL). PTL is a stress angle used in more advanced models, and it can be isolated using MC, to check if it's triggered. However, bare in mind PTL assumption is mostly theoretical. Test results show PTL stress angle changes depending on loading history. Sand starts to dilate at a different stress angle depending on loading history. Thus, using MC to check for PTL is a way for quick estimating of a modelling concept, but the theoretical concept itself can be questioned through testing (file 3 and 4 show measurements where PTL position is observed to move in triaxial testing).

Hi,

I would suggest that you do not use the MC model for assessing pore pressure buildup simply, because the model fails to simulate the site response under cyclic loading. In case you would evaluate the response acceleration you can use MC in conjunction with backbone functions available in FLAC for reproducing hysteretic damping and modulus reduction within the elastic part.