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.

Segregated Solvers mechanism?

Please login with a confirmed email address before reporting spam

Hello,
Can anybody explain to us the mechanism of the segregated solvers and how they work?! The help documentation did not illustrate if you still have a "multiphysics" or "coupled" solution if you did not group components in the "Solver parameters" box.
The name (segregated) implies solving individual modules separately, not coupled, especially when the solver does not proceed with grouping of model components!!

This leads me to another question, if somebody can help, which is the validity of a solution and its dependence on the solver type... In other words, you may try a solver sometimes and it does not work while some other solver(s) might work, does this mean if a solver works properly you got a valid and a correct solution OR the solution you have may change depending on the solver type?

Thank you for you help!

4 Replies Last Post Jul 18, 2009, 4:28 p.m. EDT

Please login with a confirmed email address before reporting spam

Posted: 2 decades ago Jul 17, 2009, 12:17 a.m. EDT
This is my personal thinking. if you have 3 groups, COMSOL MP solves your problem like this ; group 1 --> group 2 --> group3 --> group 1, 2, 3 --> group 1 --> .....

This is my personal thinking. if you have 3 groups, COMSOL MP solves your problem like this ; group 1 --> group 2 --> group3 --> group 1, 2, 3 --> group 1 --> .....

Please login with a confirmed email address before reporting spam

Posted: 2 decades ago Jul 17, 2009, 2:36 a.m. EDT
Thank you Tae for your response.
I am not sure about your suggestion because we can manage the order of solving the problem by choosing the "manual specification of segregated steps" checkbox which allows us to decide the order of the solution.

Here is what I tried which is similar to what you just said... I included the functions U1, U2 and U3 separately in three groups then add U1+U2 in another group, U1+U2+U3 in fifth group and the solver did not proceed with the solution if I "coupled" different functions together!!!

Any thoughts about that??

Thanks
Thank you Tae for your response. I am not sure about your suggestion because we can manage the order of solving the problem by choosing the "manual specification of segregated steps" checkbox which allows us to decide the order of the solution. Here is what I tried which is similar to what you just said... I included the functions U1, U2 and U3 separately in three groups then add U1+U2 in another group, U1+U2+U3 in fifth group and the solver did not proceed with the solution if I "coupled" different functions together!!! Any thoughts about that?? Thanks

Please login with a confirmed email address before reporting spam

Posted: 2 decades ago Jul 18, 2009, 10:21 a.m. EDT
what do you mean functions U1, U2, U3? are they applications included dependent variables or functions you made?

If they are just dependent variables, you just need 3 groups not 4th, 5th.
what do you mean functions U1, U2, U3? are they applications included dependent variables or functions you made? If they are just dependent variables, you just need 3 groups not 4th, 5th.

Please login with a confirmed email address before reporting spam

Posted: 2 decades ago Jul 18, 2009, 4:28 p.m. EDT
Yes, U1, U2 and U3 are applications' dependent variables such as Temperature, Concentration....etc.

I am sorry for confusion!

Thanks
Yes, U1, U2 and U3 are applications' dependent variables such as Temperature, Concentration....etc. I am sorry for confusion! Thanks

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.