Menu

Discussion Forum

New Discussion
Filtern

Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

Modelling a Flow across a boundary

Posted 17.10.2014, 00:41 GMT-4 5 Replies

Tan Chong
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Hi guys,

I'm dealing with a flow across a boundary problem, as visualized in the picture below. The geometry consists of 2 rectangles, 1 for each zone. Both use Transport of Diluted Species.

One of the species will diffuse from zone 2 to zone 1.

To model this, I added a Negative Flux term, at the boundary of zone 2, and added another flux term at the boundary of Zone 1, Positive Flux term. But I am unable to get convergence of solution.

Is there anything wrong with this implementation?

Attachments:

5 Replies Last Post 17.10.2014, 04:43 GMT-4
Lasse Murtomäki
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 10 years ago 17.10.2014, 02:10 GMT-4
You have inflow from below to the both domains. Is that convective flow? Then the situation would be unphysical. Perhaps I got it wrong, pleasse send the model file.

Lasse
You have inflow from below to the both domains. Is that convective flow? Then the situation would be unphysical. Perhaps I got it wrong, pleasse send the model file. Lasse
Tan Chong
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 10 years ago 17.10.2014, 03:35 GMT-4
Hi Lasse,

Thanks for offering your help. Unfortunately the file is not on me now (at college), so I can only provide a description.

Yes, both sides have convective flows (Transport of Dilute Species), and are from the bottom. The method of mass transfer from the outer (right) to the inner zone (left) is via simple diffusion. There's no convective effect.

I have created 2 adjacent rectangles, one for the Outer, and 1 for the inner zone. In addition, on each side, I have added a 'Flux' node. Both terms are equal, but not coupled. The outer side has a negative sign (to reflect outward flux), the inner side, vice versa.

The error for the plot remains more or less static around 10^0 to 10^-1.

Hi Lasse, Thanks for offering your help. Unfortunately the file is not on me now (at college), so I can only provide a description. Yes, both sides have convective flows (Transport of Dilute Species), and are from the bottom. The method of mass transfer from the outer (right) to the inner zone (left) is via simple diffusion. There's no convective effect. I have created 2 adjacent rectangles, one for the Outer, and 1 for the inner zone. In addition, on each side, I have added a 'Flux' node. Both terms are equal, but not coupled. The outer side has a negative sign (to reflect outward flux), the inner side, vice versa. The error for the plot remains more or less static around 10^0 to 10^-1.
Lasse Murtomäki
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 10 years ago 17.10.2014, 03:46 GMT-4
If you have convective inflow to both domains, the volumes of the domains are not constant. And if the flow rates are different, a pressure difference is created between the domains, creating convective flow. That is why I said that the model in unphysical.

I have replied in the morning to another question, where I explained in detail how to apply the flux boundary condition at the interface of two domains.
If you have convective inflow to both domains, the volumes of the domains are not constant. And if the flow rates are different, a pressure difference is created between the domains, creating convective flow. That is why I said that the model in unphysical. I have replied in the morning to another question, where I explained in detail how to apply the flux boundary condition at the interface of two domains.
Tan Chong
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 10 years ago 17.10.2014, 04:29 GMT-4
Hi Lasse,

Thanks for your comments. I'll go take a look at your comments.

The domains are actually bounded. This is a 2D axisymmetric reactor, where the outer side of the outer domain is bounded by a wall, and the region in-between is a membrane.

I understand where you are coming from, which is why the flux term is actually dependent on the pressure term at the outer end of the reactor. More explicitly-speaking, J = Permeability * (Pressure Difference between outside and inside) at the boundary. Simple diffusion still manifests along the radial direction, but at the boundary, the mode of transportation across to the other domain rests solely on the 'J' term.
Hi Lasse, Thanks for your comments. I'll go take a look at your comments. The domains are actually bounded. This is a 2D axisymmetric reactor, where the outer side of the outer domain is bounded by a wall, and the region in-between is a membrane. I understand where you are coming from, which is why the flux term is actually dependent on the pressure term at the outer end of the reactor. More explicitly-speaking, J = Permeability * (Pressure Difference between outside and inside) at the boundary. Simple diffusion still manifests along the radial direction, but at the boundary, the mode of transportation across to the other domain rests solely on the 'J' term.
Lasse Murtomäki
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 10 years ago 17.10.2014, 04:43 GMT-4
Proper meshing is very important with convective flows. In my experience, too a sparse mesh gives very inaccurate results with the Navier-Stokes equation. I would still like to look at your model :)

Lasse
Proper meshing is very important with convective flows. In my experience, too a sparse mesh gives very inaccurate results with the Navier-Stokes equation. I would still like to look at your model :) Lasse

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.

Suggested Content