Menu

Discussion Forum

New Discussion
Filter

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.

Non-Linear Magnetic Material - Forced Mu_R

Posted Aug 23, 2024, 8:54 a.m. EDT Electromagnetics, Materials, Model Manager Version 6.0 5 Replies

JoP12321
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Hi,

Relatively new to COMSOL, did try and Google this and used GPT but still haven't found a work around.

I'm modeling a cylindrical solenoid and a cylindrical piece of iron (custom BH curve) and trying to do a force calculation. I used the default Soft Iron material and updated with custom BH curve.

When I set this up, it forces me to input a constant mu_r value (from Amepere's law), even though it has a specified non-linear magnetic property.

I set this to 100, hoping that it was just a dummy input, however whenever I play around with this value, it is very clear that COMSOL is just using this value, not the BH curve. Can say that with certainty, as if I set this mu_r to 1, B is approximately 0.1T and if I set mu_r to 100, B is approximately 3T, which is well above the custom BH curve (Saturation 2T).

If anyone has any advice on how I can bypass the forced mu_r input, or a way of properly overriding this with the given BH curve that would be great. Model attached.

If you have any questions please let me know.

Many Thanks in Advance

Attachments:

5 Replies Last Post Aug 30, 2024, 9:49 a.m. EDT
Edgar J. Kaiser Certified Consultant
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 6 months ago Aug 23, 2024, 10:20 a.m. EDT

You need to add an Ampere's law node for the nonlinear domain with the BH constitutive relation.

-------------------
Edgar J. Kaiser
emPhys Physical Technology
www.emphys.com
You need to add an Ampere's law node for the nonlinear domain with the BH constitutive relation.
JoP12321
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 6 months ago Aug 29, 2024, 6:36 a.m. EDT

Hi All,

Many thanks for you quick help.

Can confirm adding the Ampere's Law node with a BH relationship allowed the simulation to run.

Unfortunately, I cannot get the simulation to converge at all.

I have tried many different meshes (my default attempts are 'finer' mesh quality with adaptive mesh. I've tried using my custom BH curve, as well as the default 'soft-iron without losses' non-linear material. Can't get either to converge at all.

I have also tried iterative and direct solvers and increasing/decreasing the tolerance. I have tried everything I can think of, and have yet to get a study to converge towards a solution. Have tried running higher quality meshes overnight, but to no avail either.

Might be worth noting, that if I use a linear BH relationship with a constant mu_r, I can get a 'finer' mesh with adaptive mesh to converge to a solution in less than 5 minutes.

I've attached my updated model.

Does anyone have any advice on how to get my study to complete?

Thanks again.

Hi All, Many thanks for you quick help. Can confirm adding the Ampere's Law node with a BH relationship allowed the simulation to run. Unfortunately, I cannot get the simulation to converge at all. I have tried many different meshes (my default attempts are 'finer' mesh quality with adaptive mesh. I've tried using my custom BH curve, as well as the default 'soft-iron without losses' non-linear material. Can't get either to converge at all. I have also tried iterative and direct solvers and increasing/decreasing the tolerance. I have tried everything I can think of, and have yet to get a study to converge towards a solution. Have tried running higher quality meshes overnight, but to no avail either. Might be worth noting, that if I use a linear BH relationship with a constant mu_r, I can get a 'finer' mesh with adaptive mesh to converge to a solution in less than 5 minutes. I've attached my updated model. Does anyone have any advice on how to get my study to complete? Thanks again.

Attachments:

Edgar J. Kaiser Certified Consultant
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 6 months ago Aug 30, 2024, 1:55 a.m. EDT

As this is a stationary study you may need to add a 'Gauge Fixing for A-Field node'. Your BH-curve has quite a kink after the first linear piece. Try to apply piecewise cubic interpolation to achieve a smoother curve. With these changes the model converged here. I am running version 6.2. The maximum normB values are a little high. You may have some issues in the geometry, because I get warnings about inverted elements. Cheers Edgar

-------------------
Edgar J. Kaiser
emPhys Physical Technology
www.emphys.com
As this is a stationary study you may need to add a 'Gauge Fixing for A-Field node'. Your BH-curve has quite a kink after the first linear piece. Try to apply piecewise cubic interpolation to achieve a smoother curve. With these changes the model converged here. I am running version 6.2. The maximum normB values are a little high. You may have some issues in the geometry, because I get warnings about inverted elements. Cheers Edgar
JoP12321
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 6 months ago Aug 30, 2024, 6:46 a.m. EDT

Hi Edgar,

I cannot thank you enough!

I added the gauge fixing node. This is one of those solutions that I would never have been able to figure out by myself.

Simulation now completes in 2m:42s, leagues faster than my biggest hopes!

The default soft iron (without losses) BH curve is indeed quite a strange shape. I updated my simulations with my Vibrating Sample Magnetometer data, so fully represents my lab work.

Thanks once again, hopefully this thread will be useful for anyone experiencing similar issues in the future!

All the best.

Hi Edgar, I cannot thank you enough! I added the gauge fixing node. This is one of those solutions that I would never have been able to figure out by myself. Simulation now completes in 2m:42s, leagues faster than my biggest hopes! The default soft iron (without losses) BH curve is indeed quite a strange shape. I updated my simulations with my Vibrating Sample Magnetometer data, so fully represents my lab work. Thanks once again, hopefully this thread will be useful for anyone experiencing similar issues in the future! All the best.
Edgar J. Kaiser Certified Consultant
Send Private Message Flag post as spam

Please login with a confirmed email address before reporting spam

Posted: 6 months ago Aug 30, 2024, 9:49 a.m. EDT

I am happy I could help. The Gauge Fixing is a frequent pitfall. There are a few blog contributions about the background of the subject.

Cheers, Edgar

-------------------
Edgar J. Kaiser
emPhys Physical Technology
www.emphys.com
I am happy I could help. The Gauge Fixing is a frequent pitfall. There are a few blog contributions about the background of the subject. Cheers, Edgar

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