Unsaturated soil formulation

[Francisliii]

Hi Dear all,

I am working on modeling unsaturated soil behavior using the built-in formulation "unsaturated material with 2 phase suction effects". My first model is based on the "Shallow foundation" and guided by the "Rapid drawdown" examples in the tutorial manual. The model went through eventually, but there are a few questions I would appreciate any help to clarify.

Regarding the calculation data, gravity load is added immediately in the first calculation step, and quasi-static convergences and k0 procedure are applied.

Questions:

1. Is hydraulic boundary condition necessary in this type of problem where transient flow is not presented? In the very first trial, I added the hydraulic boundary conditions as guided in the tutorial but the calculation returns "forrtl: severe (157): Program Exception - access violation". Then I deleted the hydraulic boundary condition and kept everything else the same and the model went through.

Quoted from tutorial manual (11.2.1.7): "Since the body is expected to experience large displacements, initial boundary nodes may differ from boundary nodes at a later stage of the simulation, thus it is necessary to define a potential area/volume of research of boundary nodes ...".

Since the pressing foundation caused large deformation to the unsaturated soil, deleting hydraulic boundary condition is not aligned with the quotation above. Could anyone help me understanding this part?

2. In the case of manually adding phreatic surface to the soil, the matric suction is defined by the SWRC model (say Van Genuchten). Is there a way to maintain the matric suction value in the soil body to a relatively constant value through out the simulation? I wondered if the "initial water pressure" option in calculation data or "2D-Liquid Pressure" option in loading conditions would do the job?

3. What's the difference between "prescribed velocity and initial velocity". My current understanding is that the initial velocity applies in the first calculation step and the rest will be taken care of by gravity. Prescribed velocity will eventually be reached and maintained thereafter during the simulation. Is this correct?

Thank you all for any helps!
Francis

[Francisliii]

Well I figured out the third questions.

[FrancescaCeccato]

Dear Francis,
I'm not sure to get your problem correctly.

1. What hydraulic boundary conditions are you using? Hydraulic head, seepage fare or infiltration? These are not always necessary, it depends on the problem you want to simulate. Remember that if you do not prescribe any condition at the boundary for the liquid pahse, zero pressure is automatically applied.
2. the formulation is fully coupled; this means that soil deformations will modify pressures (or suctions) and pressure changes will generate displacements. It is not possible to assign a constant suction in the model in a fully coupled formulation. You may try to assign suctions at the boundaries to try keeping the suction more stable, but when the soil deforms the suction will inevitably change.
3. prescribed velocity only applies to structures. it can be used, for example,for pile penetration, CPT, foundations ect.

[CPLA81192]

Thank you for your reply. I am trying to find a way to set it. Thank you very much.

[CPLA81192]

I'm very sorry to bother you. When simulating rainfall infiltration, when the initial liquid pressure is high, such as 2000kPa, the soil is prone to expansion and loosening, and it is easy to disconnect from the bottom. May I ask what may be the reason for this situation

[FrancescaCeccato]

Dear @CPLA81192,
Sorry but I do not clearly get the property of the model. Are you applying a suction at the top surface? What is the initial state?

[CPLA81192]

The soil I simulated had an initial matric suction of 2000kPa, so the INITIAL_WATER_PRESSURE of my facility was 2000kPa. However, the soil would swell during the infiltration simulation

[FrancescaCeccato]

Initial suction is 2000kPa, is it initially in equilibium with gravity?
Then water infiltrates from the top, suction decreases and effective stress decreases, thus soil swell. If swelling is much larger than expected it may be because of the constitutive model. Probably the elastic modulus is too small and deformation due to swelling is too large.