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.

Mesh resolution for RF problem

Please login with a confirmed email address before reporting spam

Hello,

I am trying to solve an electromagnetic wave propagation problem with a frequency of 140 GHz (which corresponds to a wavelength of 2.14 mm).

My issue is the results seem to have quite bad resolution (attached screenshot of the wave power). I am using an adaptive mesh, which scales from wavelength / 2 to wavelength / 4 (attached screenshot with the mesh of the model). Since my problem is axis-symmetric, I tried to create a symmetric mesh, by copying the domain of the one half to other half (the orange curve in the graph). Even with this trick, however I was not able to increase the mesh resolution, as the model was running out of memory while solving the problem (when I tried to scale the mesh from wavelength / 4 to wavelength / 6 for example).

Is there any idea how can I increase the resolution of the results without necessarily pushing the mesh resolution way too high?

Best wishes, Christos Vagkidis



2 Replies Last Post 22.01.2024, 11:13 GMT-5
Edgar J. Kaiser Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 1 year ago 21.09.2023, 03:52 GMT-4

Christos,

your model is mirror symmetric, so you can cut away half of the geometry and set a symmetry boundary at the mirror plane. Check the documentation or examples to learn how to use symmetries. Typically you need six mesh elements per wavelength in RF models.

Cheers Edgar

-------------------
Edgar J. Kaiser
emPhys Physical Technology
www.emphys.com
Christos, your model is mirror symmetric, so you can cut away half of the geometry and set a symmetry boundary at the mirror plane. Check the documentation or examples to learn how to use symmetries. Typically you need six mesh elements per wavelength in RF models. Cheers Edgar

Please login with a confirmed email address before reporting spam

Posted: 10 months ago 22.01.2024, 11:13 GMT-5

Christos,

your model is mirror symmetric, so you can cut away half of the geometry and set a symmetry boundary at the mirror plane. Check the documentation or examples to learn how to use symmetries. Typically you need six mesh elements per wavelength in RF models.

Cheers Edgar

Hello Edgar,

Thank you very much for your help. This approach indeed helped a lot.

Cheers, Christos

>Christos, > >your model is mirror symmetric, so you can cut away half of the geometry and set a symmetry boundary at the mirror plane. >Check the documentation or examples to learn how to use symmetries. Typically you need six mesh elements per wavelength in RF models. > >Cheers >Edgar Hello Edgar, Thank you very much for your help. This approach indeed helped a lot. Cheers, Christos

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.