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 The information provided may be out of date.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.

Insulation Layer

Please login with a confirmed email address before reporting spam

Dear Ivar,

I want to split a solid conductor (as a winding) into parts in axisymmetric environment and study that in frequency domain. In this way a thin gap (or electrical insulation) should be represented between each neighboring sections. The model is in Magnetic Fields (mf) physics. I tried “thin low permeability gap” boundary condition but unfortunately it doesn’t work. In my simulations, it seems that the boundaries don’t have any effect on calculations. Please find attached my mph file.

FYI, i am using the new version of COMSOL "4.2.1.110"

Thank you in advance,
Warms Regards,
Saeed


5 Replies Last Post 30 mar 2012, 07:26 GMT-4
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 21 nov 2011, 09:40 GMT-5
Hi

there is one feature one must/should not forget: material are (currently) applied per element type: i.e. you have domain material properties and boundary material properties etc. So if you add a "thin feature on a boundary you must also add (or de-double) a material for this boundary, and specify in the material properties: "Geometrical Entity level: Boundary"

But as you defined the "user property, I believe its not 100% correct that COMSOL should flag a "material omission error"

Then if you want to resolve details (i.e. skin effects) in the coil, you need a mesh fine enough in the material (Cu part) to resolve the gradients, and I would propose that you use boundary mesh elements there too (see the doc)

Then I would start with a 4x or 5x smaller air radius to have a smaller model, you could later add a layer and a spherical infinite layer, to improve the results

i.e. plot a contour line and see if you have at least 1-2 elements per line

--
Good luck
Ivar
Hi there is one feature one must/should not forget: material are (currently) applied per element type: i.e. you have domain material properties and boundary material properties etc. So if you add a "thin feature on a boundary you must also add (or de-double) a material for this boundary, and specify in the material properties: "Geometrical Entity level: Boundary" But as you defined the "user property, I believe its not 100% correct that COMSOL should flag a "material omission error" Then if you want to resolve details (i.e. skin effects) in the coil, you need a mesh fine enough in the material (Cu part) to resolve the gradients, and I would propose that you use boundary mesh elements there too (see the doc) Then I would start with a 4x or 5x smaller air radius to have a smaller model, you could later add a layer and a spherical infinite layer, to improve the results i.e. plot a contour line and see if you have at least 1-2 elements per line -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 21 nov 2011, 16:10 GMT-5
Dear Ivar,

Thank you for your precious information. As you can see in the new *.mph attached file, i implemented all your aforementioned points, but the "thin low permeability" boundary condition still doesn't have any effect on calculations.
I expect that by defining this boundary the first winding divides into two separate parts so i can see induced currents separately in both parts. But when you look at the current density (phi component) it seems that the software still considers the first winding as a continuous solid winding (please correct me if i am wrong).
Maybe i am using an inappropriate feature to have an "electrical insulation layer" between two solid windings. Let me ask my question in another way and i would really appreciate it if you can help me with my problem:
what do you suggest me to do to model an "electrical insulation layer" which electrically separate a solid winding into two parts?

Thank you in advance,
Warm Regards
Saeed
Dear Ivar, Thank you for your precious information. As you can see in the new *.mph attached file, i implemented all your aforementioned points, but the "thin low permeability" boundary condition still doesn't have any effect on calculations. I expect that by defining this boundary the first winding divides into two separate parts so i can see induced currents separately in both parts. But when you look at the current density (phi component) it seems that the software still considers the first winding as a continuous solid winding (please correct me if i am wrong). Maybe i am using an inappropriate feature to have an "electrical insulation layer" between two solid windings. Let me ask my question in another way and i would really appreciate it if you can help me with my problem: what do you suggest me to do to model an "electrical insulation layer" which electrically separate a solid winding into two parts? Thank you in advance, Warm Regards Saeed


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 22 nov 2011, 02:15 GMT-5
Hi

indeed not sure why I see no difference with or without your boundary layer (and the suppression of the air boundary definition). One other way to study this behaviour is to add two "layers of i.e.18 mm on your first rectangle (select left and right)

then to model the air as a "true" thin air domain and then to compare the results (mesh the Cu + thin layer as tri before the boundary layer)

But I do not see any major difference there, so I wonder if it is not correct

--
Good luck
Ivar
Hi indeed not sure why I see no difference with or without your boundary layer (and the suppression of the air boundary definition). One other way to study this behaviour is to add two "layers of i.e.18 mm on your first rectangle (select left and right) then to model the air as a "true" thin air domain and then to compare the results (mesh the Cu + thin layer as tri before the boundary layer) But I do not see any major difference there, so I wonder if it is not correct -- Good luck Ivar

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 30 mar 2012, 03:05 GMT-4
Hi,
I'm facing the same problem as Saeed. I need to model an electical separation layer between conductors, but the "thin low permeability " boundary condution has no effect on results.
I solved the problem by creating a physical thin air layer between the conductors, that of course requires a very fine mesh. But I still don't understand why the original b.c. doesn't wotk.

Regards
Massimo
Hi, I'm facing the same problem as Saeed. I need to model an electical separation layer between conductors, but the "thin low permeability " boundary condution has no effect on results. I solved the problem by creating a physical thin air layer between the conductors, that of course requires a very fine mesh. But I still don't understand why the original b.c. doesn't wotk. Regards Massimo

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 30 mar 2012, 07:26 GMT-4
Hi,

in MEF you can check, if Transition Boundary Condition is working for you.
For frequency domain I have modified your example (MEF is not for transient).

Ralf
Hi, in MEF you can check, if Transition Boundary Condition is working for you. For frequency domain I have modified your example (MEF is not for transient). Ralf

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.