Richard Mechanics: Initial condition, Units of parameter and source term

Hello everyone,
I am learning OGS6, my intention to use OGS is to simulate the rainfall-induced landslide. I have a few queries and difficulties while learning Richard Mechanics benchmark.

1. About modelling existing groundwater table (as an initial condition)
I want to define the groundwater table as given in the following figure. Is defining initial condition using ParaView filter is enough or there is any special procedure for that? (because I feel that my model is not giving right result)


2. About a unit of the parameter used in the Richard Mechanics benchmark file??
It looks like the Internation System (IS) is used in OpenGeoSys but in the Richards mechanics benchmark; it is weird because both liquid and solid phase has the same 1 unit density. Could anybody explain to me that? (I wanted to change the parameter as per my model requirement and when I have tried to change the parameters, the iteration does not converge).

3. About source term definition
I want to define the rainfall flux boundary as a source term on top of the model. Has anybody already done rainfall simulation please help me?

Thank you for your help in advance.
Here are the input files for my model.
model.zip (53.3 KB)

Hi Laxman,
you can prescribe a mass flux through the top by specifying a Neumann boundary condition for the liquid pressure. Some of the benchmarks have indeed fantasy units for testing reasons. So you should set physical values according to a self-consistent unit scheme.

I’m very interested in your benchmark once it runs. If you define an initial pore pressure field, you should try to prescribe an initial effective stress field as well. Or you run a calculation of the initial gravity step to until your water table is equilibrated. Then you can run a restart from there and adapt your boundary conditions.

You may also be interested in this example in Fig. 2 here:

It’s for a dry slope but it would be cool to try that with your model eventually (once you get the elastic model to run).

Best,
Thomas

I added {:meshname} to the <prefix>, s.t. the output files don’t overwrite each other. Now the line looks like so
<prefix>Project_2d_Richard_{:meshname}</prefix>

  1. The input looks good and is read correctly as you can verify by looking into the first output file Project_2d_Richard_project_ts_0_t_0.000000.vtu, pressure_interpolated field.

  2. You probably mean the SI system. Actually the units are free for the user to choose. Only in some constitutive relations there is expected input say in Kelvin.
    The solid phase density is set to 1.5 in the given project file, which is somewhat low.
    Changing it to 2500 works for me. (At least it does not crash.) See my version of the project file attached, for other changes.
    The result does not look nice, I think increasing mesh resolution would be good because the pressure gradients on the hilltop are too large for a single element.

project.zip (2.1 KB)

Thanks for your reply. But I don’t know how to set up an initial effective stress zone or run a calculation of the initial gravitational steps to get a stable water table. Can you please elaborate or give some examples?
Furthermore, thank you for providing me with your publication. I am planning to incorporate failure process once my model runs.
Regards,
Laxman Kafle

Thank you very much for clearing my doubt and enhancing my model. But still, I have some doubt especially on this part of the code. I understand their function on the code but I did not found an appropriate parameter that I could use. Could you suggest me literature or tentative values that I could use for my model (for clay and sand)? If you have any literature for understanding Richards Mechanics benchmark, I will be thankful to you.


I have changed the mesh as per your suggestion, now result looks a little nicer.
GWT_ConstOutlet.zip (149.6 KB)

Regards,
Laxman Kafle

Hi Laxman,
I also want to establish an initialization to get a stable state. Have you sloved this problem? Can you tell me more details? Thanks!

Hi Rui Feng,
Yes, I have solved this problem by running the initial gravity step.

To generate the in-situ pore water pressure and stress field you need to run an initial-steady calculation (initial gravity-step). There are different ways you can get an initial stable state. The first is that you can gradually apply gravity in your model to get a stable initial state. You can achieve this by gradually applying density (example).

In more complicated problems this method may not work because of large plastic deformation because of transient effects during groundwater settling. Hence, I divided this step into three more sub-steps. First, run the groundwater settling step to generate the equilibrium groundwater table. In this step, I have avoided plastic deformation by introducing large strength parameters for soil (Cohesion and FrictionalAngle). In the second step, I gradually reduce the strength parameters to the desired value and at last run the verification step to check whether the model has attended the equilibrium or not?

It depends on which process you want to model and the complications of your problem.

Best,
Laxman Kafle

1 Like

Thanks, Laxman!
As you said, I think I need to divide my calculation into more sub-steps. So, I want to know is there any guidance or manual to help complete these steps?

I am afraid there is no manual to help complete these steps. You have to connect each step with the restart option.

If you need my model, please contact me via email: lksm19@mails.tsinghua.edu.cn

Best,
Laxman Kafle

There is now a webpage on some basic conventions like units of BCs and STs: OGS Conventions