Menu

Discussion Forum

New Discussion
Filtern

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 The information provided may be out of date.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.

Meshing problem with very thin Layer, sourrounded by air.

Posted 06.03.2013, 05:41 GMT-5 Low-Frequency Electromagnetics, Mesh Version 4.3 3 Replies

Marius Rohde
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam


I'm modelling a spiral inductor on a PCB. The height of the copper layer is only 70 µm. Between the turns is an air gap of 200 µm. The inductor is sourrunded by an air domain and an infinite element domain.

The copper can meshed fast with a free Triangular mesh and a swept mesh. But I'm having problems with meshing the sourrunding air. I tried various mesh sizes but none of them worked.

I read that i might be possible to reduce the dimensions of the copper and build a layer with a zero thickness. But I need also the ohmic resistance of the inductor. So i think this not a solution.

Please help me with this problem.

Attachments:

3 Replies Last Post 20.03.2013, 16:24 GMT-4
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 06.03.2013, 10:51 GMT-5
Hi

you should slice your "air" at the thickness of the coil (in Z direction) so you have a thin slice in the middle of your model then mesh that one fully wit a boundary triangle and a sweep, finalise with thet for the upper and lower part. Normally the PML / INF mesh section should be structured mesh elements

Its often better to cut up your model (you could work with a symmetry plane two, in the middle of the coil height, it makes the model smaller, solve quicker and debug quicker

--
Good luck
Ivar
Hi you should slice your "air" at the thickness of the coil (in Z direction) so you have a thin slice in the middle of your model then mesh that one fully wit a boundary triangle and a sweep, finalise with thet for the upper and lower part. Normally the PML / INF mesh section should be structured mesh elements Its often better to cut up your model (you could work with a symmetry plane two, in the middle of the coil height, it makes the model smaller, solve quicker and debug quicker -- Good luck Ivar
Marius Rohde
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 07.03.2013, 05:16 GMT-5
Hi Ivar,

thank you very much for your advise. It worked! Finally I was able to mesh the model with the additional air slice.

But now im fighting with the convergence of my solution. I'm trying to calculate the inductivity with the mef module.
Unfortunately the solution does not converge. I tried various refinements, but this didnt work.

Dou you know other possibilties to improve the model to reach convergence ?

kind regards

Marius
Hi Ivar, thank you very much for your advise. It worked! Finally I was able to mesh the model with the additional air slice. But now im fighting with the convergence of my solution. I'm trying to calculate the inductivity with the mef module. Unfortunately the solution does not converge. I tried various refinements, but this didnt work. Dou you know other possibilties to improve the model to reach convergence ? kind regards Marius
Ivar KJELBERG COMSOL Multiphysics(r) fan, retired, former "Senior Expert" at CSEM SA (CH)
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 20.03.2013, 16:24 GMT-4
Hi

as often a fine enough mesh (= reasonable discretization) is a "must" else, often its easier to start with 1-2 turns, then extrapolate by "addition" to N turns

--
Good luck
Ivar
Hi as often a fine enough mesh (= reasonable discretization) is a "must" else, often its easier to start with 1-2 turns, then extrapolate by "addition" to N turns -- 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.

Suggested Content