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.

Ill-conditioned preconditioner

Please login with a confirmed email address before reporting spam

Hello,

I am getting this error. Can someone please explain what is the reason of this error and how can I fix it?

Thank you


1 Reply Last Post 27 giu 2013, 17:14 GMT-4
Josh Thomas Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 27 giu 2013, 17:14 GMT-4
Onder-

1) To be precise, the "error" you are asking about is actually just a warning. You still have a converged solution within the tolerances you have specified.

2) An ill-conditioned system is sensitive to small changes in the inputs and susceptible to roundoff errors. For this reason, COMSOL returns a warning to you and with it recommends that you increase the "factor in error estimate" which is another way of decreasing the tolerance on your convergence criteria. This will ensure you have an accurate result. Aside: In your case, apparently it is the preconditioner matrix associated with your iterative linear system solver that is ill-conditioned (not the main linear system you are trying to solve).

3) All of this might be a moot point, however, because for your problem I'd recommend using a Direct solver rather than an iterative solver with a multigrid pre-conditioner. You won't need any preconditioner for the Direct solver. Is there any particular reason you decided upon an iterative linear system solver? You have a relatively small RAM problem that solves with a Direct solver with less than 1GB of RAM in about 2 seconds on my machine which is an average Windows 7 workstation.

Hope these are helpful points to consider.

Best regards,
Josh Thomas
AltaSim Technologies
Onder- 1) To be precise, the "error" you are asking about is actually just a warning. You still have a converged solution within the tolerances you have specified. 2) An ill-conditioned system is sensitive to small changes in the inputs and susceptible to roundoff errors. For this reason, COMSOL returns a warning to you and with it recommends that you increase the "factor in error estimate" which is another way of decreasing the tolerance on your convergence criteria. This will ensure you have an accurate result. Aside: In your case, apparently it is the preconditioner matrix associated with your iterative linear system solver that is ill-conditioned (not the main linear system you are trying to solve). 3) All of this might be a moot point, however, because for your problem I'd recommend using a Direct solver rather than an iterative solver with a multigrid pre-conditioner. You won't need any preconditioner for the Direct solver. Is there any particular reason you decided upon an iterative linear system solver? You have a relatively small RAM problem that solves with a Direct solver with less than 1GB of RAM in about 2 seconds on my machine which is an average Windows 7 workstation. Hope these are helpful points to consider. Best regards, Josh Thomas AltaSim Technologies

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.