calibration of mohr coulomb and/or hypoplasticity material model

[awa5114]

I would like to derive material parameters based for a simulation which will implement the mohr coulomb and/or hypoplastic material model based on the following information:

• Heavily overconsolidated
• Very dense, fine sand
• In situ preconsolidation pressure of 1 MPa
• Bulk unit weight of 20 kN/m3
• Relative density 100%
• Groundwater table 6 m below piling platform level
• In situ water content ranging between 10-12%
• Degree of saturation = 71%
• D50 (mean particle size) between 0.1 and 0.15 mm

I also have averaged CPT information with friction sleeve and end resistance as a function of depth. Unfortunately I don’t have information about the dimensions of the CPT apparatus used.

I am new to constitutive models and the calibration of their parameters. Therefore I would like to ask, where can I start in terms of calibrating the parameters?

Thanks

 

[LRJ]

You need soil laboratory test data, prepared at the conditions you state, to calibrate your model against. You have to run a single element driver for the appropriate test boundary conditions and match the response with the response from the lab test.

For the soil you describe, you are unlikely to get a good match using those constitutive models. It may therefore be a case of ensuring you match the critical features for your analysis.

 

[ATSE]

1. If you are new to numerical modeling, you have a long road ahead of you.
2. What software are you using? Most numerical questions need to be framed around what software platform is being used, and which program (2D or 3D)
If Plaxis, then use the "Soil Test" feature to check your theoretical constitutive model behavior against lab results.
3. A Mohr-Coulomb (primarily bi-linear) model will be much simpler, quicker, and much more forgiving than any non-linear (usually hyperbolic) model. If you're new to this, stick with MC for a while.
4. I've found that populating a parameters table before you start is a helpful reference (in a single view), and makes model input more systematic. Like this (taken from a Plaxis tutorial example):

Good luck

 

[fattdad]

what's the end game? Slope stability? Shoring design?

f-d

ípapß gordo ainÆt no madre flaca!