Menu

Discussion Forum

New Discussion
Filter

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.

Convergence problem with Open Boundary condition - No viscous stress

Posted May 6, 2013, 3:41 p.m. EDT 5 Replies

Toyin Aseeperi
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Hi,

I'm modeling fluid flow through a 2D porous media using Darcy's Law. I intend to estimate fluid flux values across the top boundary of model, however,my desired outputs will be one in which the results obtained at this boundary are dependent on the domain solution.

For my simulations so far, I have observed that the domain solution depends on the prescribed values of the boundary condition, i.e I have to specify Dirichlet boundary conditions of stress, pressure or velocity along the top and bottom boundaries. The Side boundaries are specified as No- flow.

The Comsol user guide suggested that the use of Open Boundaries ( with No viscous stress) might be useful in obtaining solutions in cases where pressure cannot be predicated ahead of time, as such I am trying to run the model using the Brinkman's Interface in order to take advantage of the open boundary physics for the top boundary, along with a specified condition along the bottom of the model. However, I have been unable to get my model to converge with this specification. (The mph file and model diagram file is attached)

I do not understand why this is happening and would appreciate insights/tips into this problem, either under the original Darcy's Law formulation or using the Brinkman's equations.

Thanks.



5 Replies Last Post Oct 7, 2014, 2:01 p.m. EDT
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago May 7, 2013, 2:12 a.m. EDT
Hi

I would say it's mainly that it converges slowly, you do not have any "boundary layers (the boundaries are not defined) and when you look at the results when it stops after 25 iterations you see very steep velocity gradients/singularities close to the extreme boundaries, these are probably driving the convergence and slowing down the process.

Try to add correctly the boundary layers and or just use the default physics induced mesh, and you could increase the "maximum number of iterations" entry under the "Solver - ... - Fully coupled" node, Methods and Termination tab

--
Good luck
Ivar
Hi I would say it's mainly that it converges slowly, you do not have any "boundary layers (the boundaries are not defined) and when you look at the results when it stops after 25 iterations you see very steep velocity gradients/singularities close to the extreme boundaries, these are probably driving the convergence and slowing down the process. Try to add correctly the boundary layers and or just use the default physics induced mesh, and you could increase the "maximum number of iterations" entry under the "Solver - ... - Fully coupled" node, Methods and Termination tab -- Good luck Ivar
Toyin Aseeperi
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago May 16, 2013, 1:47 p.m. EDT
Thanks Ivar.

I tried both of the suggestions described, i.e, increasing the maximum number of iterations and making use of the default physics induced mesh but still couldn't get the simulation to converge using the open boundary with no viscous stress boundary condition.
Thanks Ivar. I tried both of the suggestions described, i.e, increasing the maximum number of iterations and making use of the default physics induced mesh but still couldn't get the simulation to converge using the open boundary with no viscous stress boundary condition.
Amber Whelsky
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Jul 30, 2013, 11:19 a.m. EDT
Were you able to solve this problem?
Were you able to solve this problem?
Toyin Aseeperi
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Sep 24, 2013, 3:51 p.m. EDT
No Amber, I am yet to find a solution to this problem
No Amber, I am yet to find a solution to this problem
An Le
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago Oct 7, 2014, 2:01 p.m. EDT
Hi,

I have the same problem. Have you ever solved this?

Thanks.

An
Hi, I have the same problem. Have you ever solved this? Thanks. An

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