Menu

Discussion Forum

New Discussion
Filtern

Note: This discussion is about an older version of the COMSOL Multiphysics® software. The information provided may be out of date.

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.

laminar flow in k-epsilon model

Posted 04.10.2010, 03:06 GMT-4 Version 4.2 7 Replies

Xsu
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Hi,

I have tried incompressible flow for laminar flow condition.However, when I set the inlet velocity to 1 or more than that,
it occur the error: maximum of iterations reached. Return solution have not converged.Issit I set the inlet velocity too high?

Then it came to me the idea of using the k-epsilon model.However, this model is for turbulent. Just wonder can it be used for laminar flow as well?
7 Replies Last Post 15.08.2012, 12:06 GMT-4
Danial Tz
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 04.10.2010, 10:18 GMT-4
I am a confused about what you want to do. If you want laminar flow just use laminar flow model, and make sure you use step functions for your inlet to assure a convergence (look at flow around cylinder example in comsol). Using k-e for laminar is an overkill, and does not make sense.

good luck ;)
I am a confused about what you want to do. If you want laminar flow just use laminar flow model, and make sure you use step functions for your inlet to assure a convergence (look at flow around cylinder example in comsol). Using k-e for laminar is an overkill, and does not make sense. good luck ;)
Xsu
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 04.10.2010, 21:13 GMT-4
ok.Thanks.I will try it out
ok.Thanks.I will try it out
Xsu
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 04.10.2010, 22:35 GMT-4
Hi,

I manage to solve the convergence problem with incompressible flow.
I try and error the meshing and finally it works.

Also I include the step function which suggested by Danial,which I include umax*flc1hs(t[1/s]-0.1,0.1)-provide by cylinder flow tutorial, it give me the same result.(which I varies the meshing)

Thanks Danial.

However, when I implement the step function, it is in time dependent solver.
So, I just take the final time,t=1, in order to find the steady-state response.

Just for sharing.

Thanks.
Hi, I manage to solve the convergence problem with incompressible flow. I try and error the meshing and finally it works. Also I include the step function which suggested by Danial,which I include umax*flc1hs(t[1/s]-0.1,0.1)-provide by cylinder flow tutorial, it give me the same result.(which I varies the meshing) Thanks Danial. However, when I implement the step function, it is in time dependent solver. So, I just take the final time,t=1, in order to find the steady-state response. Just for sharing. Thanks.
Danial Tz
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 05.10.2010, 02:09 GMT-4
Hi,

If you are doing steady state, you won't need step function. You only need it in transient analysis.

Best,
Danial
Hi, If you are doing steady state, you won't need step function. You only need it in transient analysis. Best, Danial
Xsu
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 05.10.2010, 02:24 GMT-4
Hi,

Thanks for Danial advice.

Means I have to varies the mesh in order to convergence in high velocity input? (for example, Vin=1.5ms-1)
Hi, Thanks for Danial advice. Means I have to varies the mesh in order to convergence in high velocity input? (for example, Vin=1.5ms-1)
Danial Tz
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 05.10.2010, 07:51 GMT-4
There are many parameters that influence convergence, one being mesh size. That's why you have to do mesh studies to see if your results do not vary anymore with mesh size (so you have small enough mesh). Other parameters can be the tolerances, correctness of boundary conditions, and physical correctness of the model (say you are not solving a Re of 1e6 with laminar model).

You're on the right way ;)
There are many parameters that influence convergence, one being mesh size. That's why you have to do mesh studies to see if your results do not vary anymore with mesh size (so you have small enough mesh). Other parameters can be the tolerances, correctness of boundary conditions, and physical correctness of the model (say you are not solving a Re of 1e6 with laminar model). You're on the right way ;)
Peng Xie
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 15.08.2012, 12:06 GMT-4
Hi, I recently have encountered the similar problems like yours. Can you enunciate how you change the mesh that led to the solution to the problem?
Thank you very much
Hi, I recently have encountered the similar problems like yours. Can you enunciate how you change the mesh that led to the solution to the problem? Thank you very much

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