Wall deposition of scalar quantities - BC treatment and general modelling
 

Hello!

I'm working on transport and deposition of scalar quantities.
At the moment the approach is very simplified (incompressible, single-phase, single-region).

Before starting to tweak existing things my way, it would be great to hear some opinions.
Let me share some personal thoughts first, as a starting point:

1) What I need is first of all some physical models to implement boundary conditions, i.e. a way to express the flux of the transported quantities from the liquid domain to the walls.
I've found that in many works "perfectly absorbing walls" are assumed (i.e. 0 concentration at the walls), which can work as a starting point but if someone knows some CFD-oriented work which deals with this kind of problems in more detail it could be very useful to me.

2) Once I know how to compute gradients at the walls, the job is done from the liquid viewpoint. What is left is how to model the accumulation of deposited quantities on the walls. In my perspective, deposited quantities stick to the walls and cannot be released in any way, but can undergo some decay process. Therefore I need to keep track of deposited quantities, and have them evolve as though they are subject to some differential equation.

The wall is simply a boundary, so I'm not sure on how to deal with this problem using standard fvPatchFields and usual OF stuff. I've noticed that the ESI version implements some kind of Finite Area Method, that is something which may be related to my problem but I'm not familiar with (also, I'm working with the Foundation version).

Other approaches may include the definition of some auxiliary variable fields to store and evolve deposited quantities, but I didn't try this approach yet.

Does anyone have previous experience, or any kind of comment of suggestion?
Thank you in advance, any contribution would be greatly appreciated.

Andrea

Some brief update:

I've "found out" that indeed also the Foundation versions include functionality related to surface equations, i.e. the surfaceFilmModels library.
I've seen that it allows in principle to model the presence of a thin film, together with mass/heat transfer between the film and the bulk flow.

Unfortunately, I see that almost only solvers from the lagrangian family are using this functionality and little documentation can be found. Is it readily compatible with simpler solvers? I'm working on a custom solver which is derived from buoyantBoussinesqSimpleFoam (OF6).
If not, does anyone have some idea about the effort required to make it compatible?

Thanks,
Andrea