How to simulate the rainfall Infiltration

[shiyezia]

I now use the groundwaterFoam to simulate the groundwater flow，but i do not know how to simulate the rainfall infiltration.
Rainfall infiltration is a neumann boundary condition.
I'm trying to modify the U values in the Utheta file and to use the U to simulate rainfall conditions，but it seems to be ineffective and wrong.
How can i simulate the rainfall Infiltration ?
thanks.

[gerlero]

Are you setting velocity values on the boundaries/patches? If so, you should set darcyGradPressure as the boundary condition on the same patches in the h field (otherwise the groundwaterFoam solver will simply ignore the prescribed velocity).

[shiyezia]

Is setting velocity values equivalent to setting flow boundary conditions（like rainfalling infiltration）?

I set darcyGradPressure as the boundary, but I don't know the physical meaning of this.In the tutorials 1Dinfiltration_Ufixed/0/h

,Why does it need to be set -0.75 gradient for darcyGradPressure.
Does it give a fixed gradient for h？

[gerlero]

I don't think value and gradient are used (or required) by groundwaterFoam. I believe value existes so that ParaView doesn't complain when you open the case; gradient could probably be omitted. Regardless of what's set for those, the darcyGradPressure boundary condition will update the gradient of h on each timestep so that the boundary flux matches the prescribed value for Utheta on that boundary.

[phorgue]

@gerlero is right, value in the darcyGradPressure is necessary for paraview visualization only.

Couple U/h boundary conditions available for groundwaterFoam solver are :

• fixedValue / darcyGradPressure for imposing flux
• zeroGradient / fixedValue for fixed potential.

[shiyezia]

Thank you for your answers!
I found that when I use a cube, simulating infiltration works fine, but when I use a slope model, I cannot transmit the head pressure downwards,the results seem to be wrong.

The internalField of h is -5m and the fixedValue of Utop is -1e-5 m/s. Calculate non convergence in the slope model with the same boundary conditions and time steps.