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.

Equilibrium Discharge Model Not Converging

Please login with a confirmed email address before reporting spam

Hi, I am running a case for Equilibrium Discharge model with convergence criteria (1e-5). I am trying to simulate a thermal plasma torch. But model is not converging and has residual value is fluctuating around 0.1 for each iteration. Also, my temp/velocity profiles(results while solving) are also not changing and have attained a certain profile after 30-40 iterations. Could anyone please suggest a technique to get the model converged. I am attaching a picture of convergence graph for your reference. Thanks in advance.



1 Reply Last Post Apr 16, 2021, 5:19 p.m. EDT
Robert Koslover Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 3 years ago Apr 16, 2021, 5:19 p.m. EDT
Updated: 3 years ago Apr 16, 2021, 5:23 p.m. EDT

I can't speak to the particulars of your model. But if you haven't already, take a good look at your mesh quality statistics and see if you can remesh the problem in a manner that improves the mesh quality, especially if/where it is low. Also, consider using a finer mesh, especially where there may be strong gradients in computed quantities. Also, consider a different element discretization (e.g., if presently linear, consider quadratic). Also, consider trying a different solver or different pre-conditioner. Recheck all your boundary and initial conditions to be sure they make sense everywhere. If none of that works, consider posting your model to the forum. There's always a chance someone here will find a way to fix it. :) See also: https://www.comsol.com/support/knowledgebase/1260 and https://www.comsol.com/support/knowledgebase/103 or just search the Comsol Knowledge Base for "convergence" and follow the links.

-------------------
Scientific Applications & Research Associates (SARA) Inc.
www.comsol.com/partners-consultants/certified-consultants/sara
I can't speak to the particulars of your model. But if you haven't already, take a good look at your mesh quality statistics and see if you can remesh the problem in a manner that improves the mesh quality, especially if/where it is low. Also, consider using a finer mesh, especially where there may be strong gradients in computed quantities. Also, consider a different element discretization (e.g., if presently linear, consider quadratic). Also, consider trying a different solver or different pre-conditioner. Recheck all your boundary and initial conditions to be sure they make sense everywhere. If none of that works, consider posting your model to the forum. There's always a chance someone here will find a way to fix it. :) See also: https://www.comsol.com/support/knowledgebase/1260 and https://www.comsol.com/support/knowledgebase/103 or just search the Comsol Knowledge Base for "convergence" and follow the links.

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.