Page 8 of 8

Re: thermal expansion

Posted: 07 Oct 2022, 08:52
by Roland
Hello Rich,
Thank you I will try this version and we keep us informed.
Roland

Re: thermal expansion

Posted: 07 Oct 2022, 17:20
by Roland
Hi,
In order to enter the "impedance boundary condition" BC's for the inductor, I have to select one after the other about 200 boundaries corresponding to the whole surface of the inductor, which takes a lot of time!
So is there in Elmer any possibility of making a "multiple" selection of a group of boundaries, for example by selecting this group of boundaries with a kind of "select box" command with the mouse?
Thanks a lot if anybody has a trick for doing that!
Roland

Re: thermal expansion

Posted: 07 Oct 2022, 18:31
by kevinarden
If you make a physical group of them in gmsh which can be done with a script, then they will translate to Elmer as one boundary.

Physical groups of volumes become bodies in Elmer
Physical groups of surfaces become boundaries in Elemer

Re: thermal expansion

Posted: 07 Oct 2022, 22:33
by Roland
Thank you Kevin, I will try this and we keep us informed.
Roland

Re: thermal expansion

Posted: 09 Oct 2022, 16:36
by Roland
Hi,
Just to see if the Joule heating result is better, I finally meshed the plate with a very fine mesh thanks to the gmsh "transfinite" command (see the here attached file).
Unfortunately, after solving over 2 seconds in transient state, as seen in the attached Joule_heating result, this Joule heating in the plate is still very crazy, unrealistic and completely uneven(with still even negative values which I skipped in the plot!) (while it seems much better and more even on the inductor surface...).
I struggle to find what's the reason of this completely unrealistic Joule heating result and now I am out of ideas about how to fix that! And as long as this is not fixed, the heat_transfer and linear elastic physics cannot give correct results, as you can understand!
So if anybody has an idea of how to improves this (Kevin, Rich, and perhaps Peter...?) it would be very wellcome!
Thanks in advance for any help!
Roland