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 The information provided may be out of date.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.

Modelling a 3D coil geometry for mechanical simulations

Please login with a confirmed email address before reporting spam

Hi all,

I am doing some research on coils' mechanical properties under Joule heating. I am having problems creating models for coils with a cross-section that is not circular. See the attached image for an example.

Does anyone have a good method of creating a helix with a non-circular cross section?

I tried a few 3D parametric curves, but was unable to come up with one that was meshable and computationally efficient. I tried modelling as a series of parallel layers, but this leaves me without any pitching on the layers nor connections between them.


6 Replies Last Post 2013年9月10日 GMT-4 13:09
Edgar J. Kaiser Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2013年9月10日 GMT-4 10:41

David,

there had recently been a couple of posts about helices with certain cross sections. I would recommend to search the forum. Your picture shows something that goes far beyond a helix.

Cheers
Edgar

--
Edgar J. Kaiser
emPhys Physical Technology
David, there had recently been a couple of posts about helices with certain cross sections. I would recommend to search the forum. Your picture shows something that goes far beyond a helix. Cheers Edgar -- Edgar J. Kaiser emPhys Physical Technology

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2013年9月10日 GMT-4 11:06
Hi Edgar,

I did see a few related posts before I wrote mine -- they seem to seek different things, however.

Perhaps I shouldn't have used the term cross-section. I'm happy with the wire being circular. I want the shape into which it's *wrapped* to be non-circular. So really, I don't want a helix so much as I want a spiralling shape -- the kind of shape you get when wrapping wire around an arbitrarily-shaped former.

Do you have any ideas on how to achieve this? I suspect it might be wise to use another package to make the model, then import this into COMSOL for the simulation phase.
Hi Edgar, I did see a few related posts before I wrote mine -- they seem to seek different things, however. Perhaps I shouldn't have used the term cross-section. I'm happy with the wire being circular. I want the shape into which it's *wrapped* to be non-circular. So really, I don't want a helix so much as I want a spiralling shape -- the kind of shape you get when wrapping wire around an arbitrarily-shaped former. Do you have any ideas on how to achieve this? I suspect it might be wise to use another package to make the model, then import this into COMSOL for the simulation phase.

Edgar J. Kaiser Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2013年9月10日 GMT-4 11:28
David,

you could try to find a suitable parameterization of the wire trajectory of your coil and then sweep a circular cross section (the wire) along this trajectory.
You would probably have to do something similar in any CAD software too, wouldn't you? It is in general better to generate the geometry in COMSOL rather than importing it. Importing may be an advantage in case the geoemtry already exists in a CAD format.
Explicitely modeling a coil as the one you showed in the picture will result in a huge mesh and respective memory requirements.

Can you approximate the coil by a multiturn coil domain?

Cheers
Edgar

--
Edgar J. Kaiser
emPhys Physical Technology
David, you could try to find a suitable parameterization of the wire trajectory of your coil and then sweep a circular cross section (the wire) along this trajectory. You would probably have to do something similar in any CAD software too, wouldn't you? It is in general better to generate the geometry in COMSOL rather than importing it. Importing may be an advantage in case the geoemtry already exists in a CAD format. Explicitely modeling a coil as the one you showed in the picture will result in a huge mesh and respective memory requirements. Can you approximate the coil by a multiturn coil domain? Cheers Edgar -- Edgar J. Kaiser emPhys Physical Technology

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2013年9月10日 GMT-4 11:40
Hi Edgar,

I did attempt to model the coil by sweeping through a parametric curve. Unfortunately, I was unable to find a curve that was easily meshable and computationally affordable. This could have been because I mostly limited myself to considering variants on the superellipse. Do you have any other ideas for parametrised shapes I might try?

I looked into the multi-turn coil domain when I began this project. As I understand it, the multi-turn coil domain is useful for electromagnetic properties, but does not give turn-by-turn simulation of mechanical properties. Please, though, correct me if I am mistaken --in this case, it would be wonderful to find that I am wrong!
Hi Edgar, I did attempt to model the coil by sweeping through a parametric curve. Unfortunately, I was unable to find a curve that was easily meshable and computationally affordable. This could have been because I mostly limited myself to considering variants on the superellipse. Do you have any other ideas for parametrised shapes I might try? I looked into the multi-turn coil domain when I began this project. As I understand it, the multi-turn coil domain is useful for electromagnetic properties, but does not give turn-by-turn simulation of mechanical properties. Please, though, correct me if I am mistaken --in this case, it would be wonderful to find that I am wrong!

Edgar J. Kaiser Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2013年9月10日 GMT-4 11:48
Ok, I realize, that it is all about mechanics. Well, regardless of the parameterization you will end up with cross sections of many circles touching each other. This is always difficult to mesh.
What mechanical properties exactly are you after? Is the coil a loose bunch of windings or is it embedded into a resin matrix? Do you need to resolve the wire structure or can you approximate by a solid body?

If you really need to explicitely resolve the wires, can you start with a 2D axially symmetric model and look how it behaves?

--
Edgar J. Kaiser
emPhys Physical Technology
Ok, I realize, that it is all about mechanics. Well, regardless of the parameterization you will end up with cross sections of many circles touching each other. This is always difficult to mesh. What mechanical properties exactly are you after? Is the coil a loose bunch of windings or is it embedded into a resin matrix? Do you need to resolve the wire structure or can you approximate by a solid body? If you really need to explicitely resolve the wires, can you start with a 2D axially symmetric model and look how it behaves? -- Edgar J. Kaiser emPhys Physical Technology

Please login with a confirmed email address before reporting spam

Posted: 1 decade ago 2013年9月10日 GMT-4 13:09
Good point RE meshing.

I am interested in tryino get thermally induced stress and strain on a wire by wire basis. I am interested in motors with and without resin.

I have managed something like a 2D simulation. This is my attempt to extend it.
Good point RE meshing. I am interested in tryino get thermally induced stress and strain on a wire by wire basis. I am interested in motors with and without resin. I have managed something like a 2D simulation. This is my attempt to extend it.

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.