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 The information provided may be out of date.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.

problem with solution convergence with a T junction

Please login with a confirmed email address before reporting spam

Hi!

I'm doing a CFD study of two phase flow split in a T junction at transient with initialization (2D geometry), but I have problem with solution convergence. My model is at attach. I'm a new user of comsol, I'm not familiarized with the software. Can anyone help me??

I tried to see the t junction model of comsol, but it's so complex to me.


Gustavo.


4 Replies Last Post 10 nov. 2011, 12:03 UTC−5

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 8 nov. 2011, 02:30 UTC−5
Hi,

You need to multiply your inlets by some time-varying function such as 'step' or 'ramp' (see other threads or manual). Also you may want to change your outlet condition to a pressure or, better, a normal stress condition, because having velocity at inelt and outlet does cause computational problems.

Cheers
Hi, You need to multiply your inlets by some time-varying function such as 'step' or 'ramp' (see other threads or manual). Also you may want to change your outlet condition to a pressure or, better, a normal stress condition, because having velocity at inelt and outlet does cause computational problems. Cheers

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 8 nov. 2011, 10:33 UTC−5
Hi Amir,

I added the step at the inlets and changed the outlet condition. If I set the range time at the time dependent study to until 1e-3 s the simulation dont have computational problems, but if I use the range until 1 s I have computational problems. Can you help me? the file of my simulations is at attach.

thanks
Hi Amir, I added the step at the inlets and changed the outlet condition. If I set the range time at the time dependent study to until 1e-3 s the simulation dont have computational problems, but if I use the range until 1 s I have computational problems. Can you help me? the file of my simulations is at attach. thanks


Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 8 nov. 2011, 11:34 UTC−5
Hi,

I changed the outlet boundary condition to a zero normal stress, slightly better, then I reformulated the step function which was not properly defined and I introduced a relative tolerance as well as slighlty modified the time step and the overall time domain.

It works, however you can see how the water is also flowing into the air conduit, therefore I don't think it will converge for any time longer than what I used, it seems there is something wrong with some of the conditions imposed on the two phases.

Cheers
Hi, I changed the outlet boundary condition to a zero normal stress, slightly better, then I reformulated the step function which was not properly defined and I introduced a relative tolerance as well as slighlty modified the time step and the overall time domain. It works, however you can see how the water is also flowing into the air conduit, therefore I don't think it will converge for any time longer than what I used, it seems there is something wrong with some of the conditions imposed on the two phases. Cheers


Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 10 nov. 2011, 12:03 UTC−5
thanks Amir,


thanks Amir,

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.