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 The information provided may be out of date.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.

Time convergence issue

Please login with a confirmed email address before reporting spam

Hi

Im working in the acoustic domain for quite some time.
Im using the time dependent model for simulation (generalized alpha).

The log file is attached below .
I cant understand what the out refers to in this case.
I occurs randomly at fixed intervals

The range(0,0.5e-7,24e-6) is used at the times field for time dependent study .
The time stepping is 5e-9.

How do i check for whether the data is converging or not?
How should the convergence plots look ?
Help in any kind will be really really useful.
Thanks a lot for your time and effort!!

Ive attached the snap shot of the time dependent study and the log file.

Regards
Glenston


5 Replies Last Post 29 mar 2012, 02:50 GMT-4
Nagi Elabbasi Facebook Reality Labs

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 27 mar 2012, 09:30 GMT-4
The analysis is converging according to your log file. “Out” means results are saved at this point. COMSOL saves results at the times that correspond to your specified range(0,0.5e-7,24e-6). To see convergence plots you should check the “Generate convergence plots” box in the settings window of the Study node.

Nagi Elabbasi
Veryst Engineering
The analysis is converging according to your log file. “Out” means results are saved at this point. COMSOL saves results at the times that correspond to your specified range(0,0.5e-7,24e-6). To see convergence plots you should check the “Generate convergence plots” box in the settings window of the Study node. Nagi Elabbasi Veryst Engineering

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 28 mar 2012, 02:35 GMT-4

The analysis is converging according to your log file. “Out” means results are saved at this point. COMSOL saves results at the times that correspond to your specified range(0,0.5e-7,24e-6). To see convergence plots you should check the “Generate convergence plots” box in the settings window of the Study node.

Nagi Elabbasi
Veryst Engineering


Dear Nagi

Thanks a lot for replying!

Firstly how do we know looking at the log file that the analysis is converging?
Secondly if you have noticed at some time instances like 1e-7,1.5e-7,2e-7,2.5e-7,3.0e-7,3.5e-7,4e-7,4.5e-7,5e-7 the line is a blank followed by -out. But instances such as 5e-8,5.5e-7,6e-7,6.5e-7,7e-7 has out different from that mentioned earlier.

Ive highlighted the points in the log file that i have attached in this post.
Is there something that im not reading properly.

Thanks a lot again!

Regards
Glenston
[QUOTE] The analysis is converging according to your log file. “Out” means results are saved at this point. COMSOL saves results at the times that correspond to your specified range(0,0.5e-7,24e-6). To see convergence plots you should check the “Generate convergence plots” box in the settings window of the Study node. Nagi Elabbasi Veryst Engineering [/QUOTE] Dear Nagi Thanks a lot for replying! Firstly how do we know looking at the log file that the analysis is converging? Secondly if you have noticed at some time instances like 1e-7,1.5e-7,2e-7,2.5e-7,3.0e-7,3.5e-7,4e-7,4.5e-7,5e-7 the line is a blank followed by -out. But instances such as 5e-8,5.5e-7,6e-7,6.5e-7,7e-7 has out different from that mentioned earlier. Ive highlighted the points in the log file that i have attached in this post. Is there something that im not reading properly. Thanks a lot again! Regards Glenston


Nagi Elabbasi Facebook Reality Labs

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 28 mar 2012, 08:06 GMT-4
You're welcome!

The analysis goes all the way so it must be converging based on the convergence tolerances selected for both spatial convergence and time integration convergence. If it did not converge at a certain time COMSOL would stop, and also give an error message. In your case, it also seems that COMSOL is not checking time integration errors because you specify manual time stepping.

I’m not sure why the –out lines look different sometimes. My understanding is that if –out is on the same line as other text it means that COMSOL outputted that solver step directly. If it is on a line of its own then COMSOL did some interpolation between solutions. That doesn’t fit your case since you manually specified a constant step size. Maybe some numerical round-off errors.

Nagi Elabbasi
Veryst Engineering
You're welcome! The analysis goes all the way so it must be converging based on the convergence tolerances selected for both spatial convergence and time integration convergence. If it did not converge at a certain time COMSOL would stop, and also give an error message. In your case, it also seems that COMSOL is not checking time integration errors because you specify manual time stepping. I’m not sure why the –out lines look different sometimes. My understanding is that if –out is on the same line as other text it means that COMSOL outputted that solver step directly. If it is on a line of its own then COMSOL did some interpolation between solutions. That doesn’t fit your case since you manually specified a constant step size. Maybe some numerical round-off errors. Nagi Elabbasi Veryst Engineering

Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 29 mar 2012, 01:55 GMT-4
Hi

I can confirm that my understanding is that several out grouped comes from COMSOL interpolating the results you demanded, while as in your log file COMSOL is running in "automatic mode and gets to a point just above/below your values and interpolates to get at precisely your value, so as Nagi says its an interpolation issue.

But on the other side, depending on what you are simulating, you might not need all these intermediate steps, try a "strict" stepping, you might gain time, but check carefully the results by comparison as if your steps are coarse w.r.t. the gradients involved, the results might be less precise

--
Good luck
Ivar
Hi I can confirm that my understanding is that several out grouped comes from COMSOL interpolating the results you demanded, while as in your log file COMSOL is running in "automatic mode and gets to a point just above/below your values and interpolates to get at precisely your value, so as Nagi says its an interpolation issue. But on the other side, depending on what you are simulating, you might not need all these intermediate steps, try a "strict" stepping, you might gain time, but check carefully the results by comparison as if your steps are coarse w.r.t. the gradients involved, the results might be less precise -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 29 mar 2012, 02:50 GMT-4
Hi Ivar,

Thanks a lot for replying.
Checking the case when i use strict and intermediate time stepping.
A comparison should give me a good understanding on how the time stepping works.

Regards
Glenston
Hi Ivar, Thanks a lot for replying. Checking the case when i use strict and intermediate time stepping. A comparison should give me a good understanding on how the time stepping works. Regards Glenston

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.