Thin porous membrane

General discussion about Elmer
Post Reply
anttik
Posts: 3
Joined: 05 Jul 2022, 13:53
Antispam: Yes

Thin porous membrane

Post by anttik »

Dear all,

I'm new to Elmer. I've tried to google this but haven't found a clear answer.

Is it possible to simulate gas flow through a very thin (thickness in the order of nanometers) and porous membrane?

Thanks,
Antti
raback
Site Admin
Posts: 4812
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: Thin porous membrane

Post by raback »

Hi Antti,

The only thing helping you in small dimensions is some slip boundary conditions (as opposed to no-slip). This helps you until Knudsen number ~0.1, see https://en.wikipedia.org/wiki/Knudsen_number. For nanometer scale you probably better use some Monte Carlo methods as continuity assumption breaks down.

-Peter
anttik
Posts: 3
Joined: 05 Jul 2022, 13:53
Antispam: Yes

Re: Thin porous membrane

Post by anttik »

Hi Peter,

The scenario I'm working can be described as a pipe with its cross-section fully covered by a membrane that lets through a part of the gas. Gas comes in on one side, encounters the membrane and only part of it goes through it and continues down the pipe. I'm interested in the properties of the gas that was able to get through to the other side. I think I could simplify the situation and consider the membrane as a simple 2D-wall. Is there a boundary condition I could use or could I for example define it as a porous material?

Sincerely,
Antti
raback
Site Admin
Posts: 4812
Joined: 22 Aug 2009, 11:57
Antispam: Yes
Location: Espoo, Finland
Contact:

Re: Thin porous membrane

Post by raback »

Hi Antti,

There is a condition that allows to add Darcy type of resistance for flow. However, I don't see how it could help you here.

The continuity assamption behind Navier-Stokes is pretty limiting here. You cannot separate the different constituents and even less treat their transport differently.

You should use some model that can deal with individual gas particles and their interaction with each other and the walls. Some kind of particle dynamics code with interaction potential maybe could do this. I know people use codes like Gromacs to model cell membranes, for example. Whether it is applicable to this, I don't know.

Another approach could be to find some experimental data of such membranes. If you have a new membrane maybe you can create some heuristical model when you know the thickness and porosity. I could imagine that there are such simplified models that work pretty ok for known membranes.

Some problems are just so difficult for ab initio type of computation that experiments may still prevail.

-Peter
anttik
Posts: 3
Joined: 05 Jul 2022, 13:53
Antispam: Yes

Re: Thin porous membrane

Post by anttik »

Good points, thanks for the advice!

Cheers,
Antti
Post Reply