CFD Online Discussion Forums - [swak4Foam] dynamic no-flux BC for electrokinetics, can groovyBC do this?

CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)

- OpenFOAM Community Contributions (https://www.cfd-online.com/Forums/openfoam-community-contributions/)

- - [swak4Foam] dynamic no-flux BC for electrokinetics, can groovyBC do this? (https://www.cfd-online.com/Forums/openfoam-community-contributions/194909-dynamic-no-flux-bc-electrokinetics-can-groovybc-do.html)

dynamic no-flux BC for electrokinetics, can groovyBC do this?

Hi,
I have a solver for a Poisson-Nernst-Planck system, that is, the Poisson equation gives me an electric field $-\nabla \phi$ and that electric field pushes the ions according the the Nernst-Planck equation, together with diffusion . The ion flux would be something like this

$J=-D(\nabla c + z_i c \nabla \phi )$

now, as a boundary condition, I need this flux to be zero in one of my boundaries. Obviously, making the condition $\nabla c = 0$ isn't enough in this case, since $\nabla \phi$ has a constant value (that comes from the boundary conditions to the Poisson equation), and additionally, the simulation starts from a non-uniform

field, so making that zero isn't an option either. I would need to specify a condition for the gradient in the form

$\nabla c = -z c \nabla \phi$

where the

in the right hand side would be the previous concentration value at the boundary (from the previous timestep) and $\nabla \phi$ is known (since this happens to be a boundary condition for the Poisson equation).

I was looking at my options and apparently It looks like could use groovyBC, or otherwise I would need to write a library specifically for that boundary condition.
I have no experience whatsoever with swak4foam or writing custom boundary conditions so I would like to take the easiest path.
Could anyone enlighten me on whether this can be done with groovyBC? is there any other alternative?

Hi OmarEGB,

did you manage to add that type of boundary condition? I would like to add a similar BC.

I would appreciate any hint :)

Quote:

Originally Posted by Snapy (Post 680886)

Hi OmarEGB,

did you manage to add that type of boundary condition? I would like to add a similar BC.

I would appreciate any hint :)

Code:

snGrad(phi)

gives you the gradient in the normal direction of the patch

Thanks for the fast response and hint :)

So I could use snGrad() in refCast to set the BC for a specific patch:

Am I correct?

Quote:

fixedGradientFVPatchScalarField& c =

refCast<fixedGradientFvPatchScalarField>

(c.boundaryFieldRef()["patchID"]);

c.gradient() = phi.boundaryField()["patchID"].snGrad();