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 The information provided may be out of date.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.

Model of 3D electromagnet

Please login with a confirmed email address before reporting spam

Hi all,

Thank you all for reading this in advance! Let me get started by saying I'm completely new to Comsol, but have been working with some CAD (SolidWorks) and simulation/physics (Matlab) software before.

My project is about finding flaws in metal structures through various techniques. Right now I am considering to purchase a license of Comsol to use for our simulation purposes. I have just received the trial package (version 4.2a) and did some tutorials (permanent magnet and 3D inductor) but to get a better feeling about Comsol would like to start with some basic simulations of my own.

Therefore I have successfully build a very basic 3d model of a steel cylinder with copper coils (Helix) wrapped around it, all encapsulated by a sphere of air. I would like to see what the magnetic field generated by the coils around the steel would like. I am using the Magnetic and Electric Fields (mef) package and from the 3d inductor example (in the library) I think I have to specify the terminal and ground of the coil in order to send a current through it. Though I am not sure why I can not seem select these, when I select the terminal and ground point (two ends of the copper coil, right?) then it says not applicable in the boundary selection box.

Also I think I might be doing something wrong with the mesh as it takes up all of my RAM to solve the study (I can build the mesh using less then 1GB of both RAM and virtual memory) and my computer basically hangs. Just as a note, the computer I'm using has 8 GB of RAM and a high-end graphics card. It might be to fine for what i'm trying to simulate, but If I choose it bigger then I end up getting errors?

Hopefully someone here is able to let me know if I'm going in the right direction and maybe let me know what I have to look at to get this simulation working. The model that I've build is also attached.

Thank you for your time!


1 Reply Last Post 2011年11月30日 GMT-5 02:37
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 2011年11月30日 GMT-5 02:37
Hi

you should do some estimation on the model size before starting. Estimate the total volume, the mes density, the number of DoF (degrees of freedom) ... In 3D a reasonnable ACDC model can easily use a few tens of GB of RAM and solve for a few hours on a large WS. It all depends on what you are looking for.

The easiest is to simplify your model, i.e. not resolve the individulas turns of the coil, but replace thm by a bulk conductor (except if you are studying what is going on in detail, inside the coil itself (skin effects, etc...)

And then use symmetry, i.e. 2D-axi is far lighter (RAM and time wise) than full 3D, BUT your model and the BC (boundary conditions must respect the symmetry

When you excite a coil in MEF (the heaviest of all ACDC models) you need to ensure you are respecting the current and filed loops, you cannot ask the physics to have current appearing in the middle of your model, and dissapearing somewhere else, otherwise nor you model, nor your results will respect the laws of physics

It is worth to do several, if not almost all models per physics, as there are, at the beginning when starting with COMSOL, a lot to learn and to get used too. COMSOL covers all physics, and all physics require more than a week or two to rehearsal, even after a few years at an university, we have all passed through these steps, you will see once mastered, COMSOL is real fun ;)

--
Good luck
Ivar
Hi you should do some estimation on the model size before starting. Estimate the total volume, the mes density, the number of DoF (degrees of freedom) ... In 3D a reasonnable ACDC model can easily use a few tens of GB of RAM and solve for a few hours on a large WS. It all depends on what you are looking for. The easiest is to simplify your model, i.e. not resolve the individulas turns of the coil, but replace thm by a bulk conductor (except if you are studying what is going on in detail, inside the coil itself (skin effects, etc...) And then use symmetry, i.e. 2D-axi is far lighter (RAM and time wise) than full 3D, BUT your model and the BC (boundary conditions must respect the symmetry When you excite a coil in MEF (the heaviest of all ACDC models) you need to ensure you are respecting the current and filed loops, you cannot ask the physics to have current appearing in the middle of your model, and dissapearing somewhere else, otherwise nor you model, nor your results will respect the laws of physics It is worth to do several, if not almost all models per physics, as there are, at the beginning when starting with COMSOL, a lot to learn and to get used too. COMSOL covers all physics, and all physics require more than a week or two to rehearsal, even after a few years at an university, we have all passed through these steps, you will see once mastered, COMSOL is real fun ;) -- 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.