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.

Magnetic flux density map in 2D - resolution

Please login with a confirmed email address before reporting spam

Hi, I'm a fairly new COMSOL user. I am modeling a simple solenoid with active shielding coils in Multiphysics AC/DC module and have validated my model (I know it is solving for the correct field profile). My problem is when performing the 2D surface mapping of the magnetic flux density (mf.normB): specifically I would like to visualize the 10 G (gauss) line. I attempted to manually alter my "range" in the surface plot settings, however when I set 10 G (0.001 T) as my minimum for the data range, my entire volume remains visible (I know this is not the case, based on my 1D analysis). When I play with the minimum, alternating between 0.001 and 0.02 in 0.001 increments (0.001, 0.002, 0.003, etc), I see a change at 0.004 and then another change at 0.015, but no change to the boundary in between. I know physically this is not accurate, so I am trying to figure out what sort of parameters I need to change in order to improve precision. I include some images showing the field map with the default data ranges, as well as when setting the limit to 0.001, 0.004, 0.015 T. I also include my mesh, which is set as 0.002 m in the "critical" volumes (coils, magnet aperture, and beamline to neighboring component). I tried one iteration with an extremely fine mesh over the full volume but this did not solve the problem and made solutions very time consuming. Thanks, Melanie!!



1 Reply Last Post 15.07.2021, 23:06 GMT-4
Robert Koslover Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 3 years ago 15.07.2021, 23:06 GMT-4
Updated: 3 years ago 15.07.2021, 23:07 GMT-4

First of all, I suggest you consider using a contour plot. You should be able to specify a 10 G contour (among any others you wish), and if all goes well, it will plot the contour(s) sensibly (ideally, not too ragged looking). Second, it seems that you have a very large range of field values here (going up to several Tesla) and values of 10 Gauss are very small in comparison. So, you may have a numerical noise issue. You didn't post your model (the .mph file) so it is difficult to guess your settings in detail, but... in addition to employing a finer mesh, you might want to consider using a higher discretization order to the elements? Also, there are some smoothing options for these plots that you might want to experiment with. Good luck. If you want better advice, I suggest you post your .mph file (you can clear the solution first) to the forum.

-------------------
Scientific Applications & Research Associates (SARA) Inc.
www.comsol.com/partners-consultants/certified-consultants/sara
First of all, I suggest you consider using a *contour* plot. You should be able to specify a 10 G contour (among any others you wish), and if all goes well, it will plot the contour(s) sensibly (ideally, not too ragged looking). Second, it seems that you have a very large range of field values here (going up to several Tesla) and values of 10 Gauss are very small in comparison. So, you may have a numerical noise issue. You didn't post your model (the .mph file) so it is difficult to guess your settings in detail, but... in addition to employing a finer mesh, you *might* want to consider using a higher discretization order to the elements? Also, there are some smoothing options for these plots that you might want to experiment with. Good luck. If you want better advice, I suggest you post your .mph file (you can clear the solution first) to the forum.

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.