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 The information provided may be out of date.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.

Solution does not converge when changing from spherical to coaxial port

Please login with a confirmed email address before reporting spam

Can anyone help me out? I have a very simple model and when I change the entrance port for the power input, I don't get a converged solution. If I choose circular port, it works.. Any ideas?


4 Replies Last Post 2 ott 2012, 14:27 GMT-4

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 1 ott 2012, 22:00 GMT-4
bump
bump

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 2 ott 2012, 01:01 GMT-4
Hi

I'm not sure if it's related to 4.2, but when I open it in 4.3 your PEC BC is not applied, I need to add a new one as LAST node.

Then I assume that it applies to surface 6, just as the copper should apply to the BOUNDARY 6 (if it was called)
Finally it looks like you port on a boundary, is not injecting energy into the PEC, as they are NOT connected via the edge vertex.

This is what I see, now as I haven't used EMW in 4.3 yet, and things seem to have changed (in the way we need to respect the order of the nodes, check the override matrix in the doc for 4.3) I cannot come with the "evident little hint that we are missing booth of us here ;)

It's worth to revisit a few model library examples and the doc, and if possible updated to 4.3 latest patch

What certainly works is to replace your PEC surface by a thin domain
Last thing, have you checked the ACDC or RF hypothesis, w.r.t. model size, compared to the wavelength 100 kHz is rather low

--
Good luck
Ivar
Hi I'm not sure if it's related to 4.2, but when I open it in 4.3 your PEC BC is not applied, I need to add a new one as LAST node. Then I assume that it applies to surface 6, just as the copper should apply to the BOUNDARY 6 (if it was called) Finally it looks like you port on a boundary, is not injecting energy into the PEC, as they are NOT connected via the edge vertex. This is what I see, now as I haven't used EMW in 4.3 yet, and things seem to have changed (in the way we need to respect the order of the nodes, check the override matrix in the doc for 4.3) I cannot come with the "evident little hint that we are missing booth of us here ;) It's worth to revisit a few model library examples and the doc, and if possible updated to 4.3 latest patch What certainly works is to replace your PEC surface by a thin domain Last thing, have you checked the ACDC or RF hypothesis, w.r.t. model size, compared to the wavelength 100 kHz is rather low -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2 ott 2012, 01:53 GMT-4
Thanks for the advice. For some reason under "PEC 1" all of the boundaries are overridden, even though they shouldn't be. I added a second "PEC 2" to make sure that the input port was connected to a PEC, but I still get the same error.

Interestingly, if you look at the example model "microwave_cancer_therapy", they have empty space surrounding the PEC, and I'm not sure why. I'm attaching this example from the model library for reference.

Sorry if this is basic or I'm just overlooking something. I'm new to COMSOL and trying to get a basic model to work so I know I can use it for a school project.

Interestingly, If I increase the input frequency the relative error gets smaller and eventually converges on a result, but not one that I would expect.

Thanks for any guidance you can give.
Thanks for the advice. For some reason under "PEC 1" all of the boundaries are overridden, even though they shouldn't be. I added a second "PEC 2" to make sure that the input port was connected to a PEC, but I still get the same error. Interestingly, if you look at the example model "microwave_cancer_therapy", they have empty space surrounding the PEC, and I'm not sure why. I'm attaching this example from the model library for reference. Sorry if this is basic or I'm just overlooking something. I'm new to COMSOL and trying to get a basic model to work so I know I can use it for a school project. Interestingly, If I increase the input frequency the relative error gets smaller and eventually converges on a result, but not one that I would expect. Thanks for any guidance you can give.


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 2 ott 2012, 14:27 GMT-4
Hi

the order of the nodes has also it's importance, therefore depending on what you have added, you might need to add another "later" node i.e. PEC in this case.

Think of the node tree as a top down application of the BC, any thing later might override the previous one BUT not the opposite

--
Good luck
Ivar
Hi the order of the nodes has also it's importance, therefore depending on what you have added, you might need to add another "later" node i.e. PEC in this case. Think of the node tree as a top down application of the BC, any thing later might override the previous one BUT not the opposite -- Good luck Ivar

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.