Menu

Discussion Forum

New Discussion
查找

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

Posted 2022年3月2日 GMT-5 03:21 Electromagnetics, Charged Particle Tracing, Plasma Physics 1 Reply

Chiu Wing Chan
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Hello! Just a little survey, what will you do firstly if you find the solution does not converge of your model? What are the parameters that you will consider first to achieve the convergence of the solution in COMSOL?

1 Reply Last Post 2022年3月2日 GMT-5 08:58
Jeff Hiller COMSOL Employee

Please login with a confirmed email address before reporting spam

Posted: 3 years ago 2022年3月2日 GMT-5 08:58
Updated: 3 years ago 2022年3月2日 GMT-5 09:08

The first thing I'd do is ask myself whether the mathematical problem I am asking the software to solve actually has a unique solution. Many PDEs don't have solutions or have multiple solutions.

The above may seem like an obvious thing to say, but you'll find in this Discussion Forum instances of users being baffled by lack of convergence when they have either insufficiently constrained their problems (e.g. no reference pressure in a flow problem, forgetting to remove all rigid body motions in structural mechanics, no imposed potential anywhere in electric currents or electrostatics,...) or over-constrained them (e.g. conductive heat transfer with a heat source and no flux conditions on all boundaries), or are in some other way trying to solve a mathematical problem that simply doesn't have a unique solution.

So that's where I look first.

Next, I look at all my physics settings as well as all definitions that they depend on, to make sure that I really did set up the model to match exactly the mathematical problem that I intended to set up.

Only after checking those things do I look into the more technical aspects and dig into solver settings and mesh settings.

Best,

Jeff

-------------------
Jeff Hiller
The first thing I'd do is ask myself whether the mathematical problem I am asking the software to solve actually has a unique solution. Many PDEs don't have solutions or have multiple solutions. The above may seem like an obvious thing to say, but you'll find in this Discussion Forum instances of users being baffled by lack of convergence when they have either insufficiently constrained their problems (e.g. no reference pressure in a flow problem, forgetting to remove all rigid body motions in structural mechanics, no imposed potential anywhere in electric currents or electrostatics,...) or over-constrained them (e.g. conductive heat transfer with a heat source and no flux conditions on all boundaries), or are in some other way trying to solve a mathematical problem that simply doesn't have a unique solution. So that's where I look first. Next, I look at all my physics settings as well as all definitions that they depend on, to make sure that I really did set up the model to match exactly the mathematical problem that I intended to set up. Only after checking those things do I look into the more technical aspects and dig into solver settings and mesh settings. Best, Jeff

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