Richards flow vs Richards mechanics

Hello all,
Today I was trying to model a rainfall infiltration using the Richards flow process. I have applied a rainfall intensity of 45.4mm/day for a 5-day period and then an extra 10 days to monitor post rainfall behaviour. I am not much satisfied with the result obtained from the Richards flow process. So, I prepared another model with the Richards mechanics process but with zero displacement boundary throughout the domain (which basically turns the Richards mechanics process into the Richards flow process).

I did a comparative study and observe, water seepage after the rainfall is unnoticeable in Richards flow but seen in the Richards Mechanics process.

Input files: RichardsFlow.zip (20.5 KB) RichardsMechanics.zip (29.0 KB)

Could anybody check my test model and suggest to me why this is happening?

Thanks

Hi Laxman, I did not check your model entirely, because I don’t have a real feeling for your specific problem (I have no idea what to expect exactly and how which quantities should look like).
However, I’m not sure about your assumption that zero-displacement BC turns your RM int RF. This depends on the model size, discretization and stiffness. The better way to turn RM into RF would be to use near-rigid elastic constants.

2 Likes

Hi Joerg,
That was my assumption about turning RM into RF may be wrong and I thought implementing zero displacement boundary condition is equivalent to infinite elastic modules. Obviously turning RM into RF is not a numerically good option (takes lots of time for simulation). I am working with the RM process for a long time, so I just prepared that for testing purpose.
Right now I am simulating rainfall infiltration using the Richards flow process, and I have a feeling that seepage water movement after the rainfall event is a little slower than expected. If you have time, could you check my Richards flow file and point out any mistake if I have made?

Hi Laxman,
I took a look at your files and I think the physics is pretty much the same in both files.
(Keep in mind that some properties moved now from phase to medium in RF process in the current OGS version)
However, you are using a different integration order for both cases. As your mesh is very coarse, this really matters.
I just checked the pressure and saturation for one point, but using an integration order of 3 also for RF gives more or less the same results in both cases.