PCB thermal analysis revisited

About the development of the FEM module/workbench.

Moderator: bernd

User avatar
easyw-fc
Veteran
Posts: 3248
Joined: Thu Jul 09, 2015 9:34 am

Re: PCB thermal analysis revisited

Post by easyw-fc »

easyw-fc wrote: Thu Jul 14, 2022 8:34 pm does mention code work? @easyw-fc

Code: Select all

[mention]user-name[/mention]
ok, I'm re-trying to mention people using the old way:
bernd wrote: ping
HarryvL wrote: ping
thschrader wrote: ping
thschrader
Veteran
Posts: 2685
Joined: Sat May 20, 2017 12:06 pm
Location: Germany

Re: PCB thermal analysis revisited

Post by thschrader »

NewJoker wrote: Fri Jul 15, 2022 7:40 pm - this geometry is quite complex and mesh should be created carefully and more refined, Netgen might be better in this case
...
Yes, meshing is the fun part.
Even when meshing the copper part alone, I get 470000 cells, and the mesh is still ill-shaped.
My idea was to use the copper alone, using 2D-elements (with arch-facebinder tool).
But that doesnt work, thermomech sim with 2D-elements wont run with FC, right?
And you need the board-solid to connect all copper parts. When using copper alone,
you get several not connected meshes. Or you must clean up the copper part, which means
redesign...Mmmh...
mesh_pcb.JPG
mesh_pcb.JPG (257.06 KiB) Viewed 687 times
User avatar
mfro
Posts: 359
Joined: Sat Sep 23, 2017 8:15 am

Re: PCB thermal analysis revisited

Post by mfro »

Kunda1 wrote: Thu Jul 14, 2022 12:51 pm Oooh, really cool topic being discussed here!
I feel tempted to disagree. Rather hot topic indeed ;)
Cheers,
Markus
User avatar
Kunda1
Veteran
Posts: 13022
Joined: Thu Jan 05, 2017 9:03 pm

Re: PCB thermal analysis revisited

Post by Kunda1 »

:D
Alone you go faster. Together we go farther
Please mark thread [Solved]
Want to contribute back to FC? Checkout:
#lowhangingfruit | Use the Source, Luke. | How to Help FreeCAD | How to report FC bugs and features
User avatar
easyw-fc
Veteran
Posts: 3248
Joined: Thu Jul 09, 2015 9:34 am

Re: PCB thermal analysis revisited

Post by easyw-fc »

thschrader wrote: Sun Jul 17, 2022 8:56 am Yes, meshing is the fun part.
a friend of mine, did something interesting some time ago...
fem-fc-tests.jpg
fem-fc-tests.jpg (416.85 KiB) Viewed 590 times
I have to find out his suggestions
david69
Veteran
Posts: 1125
Joined: Wed Jan 01, 2014 7:48 pm

Re: PCB thermal analysis revisited

Post by david69 »

nice. @easyw-fc with FreeCAD?
thschrader
Veteran
Posts: 2685
Joined: Sat May 20, 2017 12:06 pm
Location: Germany

Re: PCB thermal analysis revisited

Post by thschrader »

easyw-fc wrote: Mon Jul 18, 2022 9:03 am
thschrader wrote: Sun Jul 17, 2022 8:56 am Yes, meshing is the fun part.
a friend of mine, did something interesting some time ago...
fem-fc-tests.jpg
I have to find out his suggestions
Impressive!
Maybe he can share the file?
Background/colorbar looks like FC...
User avatar
easyw-fc
Veteran
Posts: 3248
Joined: Thu Jul 09, 2015 9:34 am

Re: PCB thermal analysis revisited

Post by easyw-fc »

david69 wrote: Mon Jul 18, 2022 11:46 am nice. @easyw-fc with FreeCAD?
definitively yes! :D
thschrader wrote: Mon Jul 18, 2022 12:52 pm Impressive!
Maybe he can share the file?
Background/colorbar looks like FC...
He sent me some details, not the file, 2 years ago... I pinged him to know if he improved his method to have something to be usable...
Anyway I can share some of his thought:
The steps are:
  • Make the layout
  • Import the copper into FreeCad using ‘fcad_pcb’ https://github.com/realthunder/fcad_pcb
    • It is smart to set the copper thickness to multiple of actual thickness as meshing thin layers is a PITA.
      For 35um board I used 200um. You can account for this later lowering the copper thermal resistance and specific heat by the same ratio.
  • Import only copper which will have some thermal effects (you can specify which nets you want to import)
  • Import parts into FreeCad. If needed arrange and align them as needed
    • For FEM analysis you have to prepare the model
    • Select all the parts then Part Workbench/Split/Boolean Fragments
    • In the data tab change Boolean fragments mode from Standard to CompSolid
    • Select all then Part Workbench/Compound/Compound Filter
  • Create FEM analysis
    • Set the analysis to thermomechanical
    • Add all the materials (copper, FR4). Add solids to specific materials.
    • Add boundary conditions (initial temperature, heat dissipation through all surfaces, heat generation, …)
    • Add mesh. This the weakest link. With Gmesh I could make a mesh for simple designs (without via’s) while with complex designs I had better luck with netgen. But Netgen generates big meshes even with Very Coarse settings.
  • Write .inp file (this will take some time)
  • Run the simulation (this will also take some time)
Unfortunately I'm not confident in FEM, so I cannot reproduce it on a test file (which I don't have)
User avatar
easyw-fc
Veteran
Posts: 3248
Joined: Thu Jul 09, 2015 9:34 am

Re: PCB thermal analysis revisited

Post by easyw-fc »

I'm attaching a similar 3d pcb model if someone would dare to test
fem-test-1.png
fem-test-1.png (364.28 KiB) Viewed 470 times
.
fem-test-fcad-pcb.FCStd
(155.91 KiB) Downloaded 9 times
kamocat
Posts: 17
Joined: Mon Jul 11, 2022 10:39 pm

Re: PCB thermal analysis revisited

Post by kamocat »

easyw-fc wrote: Mon Jul 18, 2022 5:47 pm I'm attaching a similar 3d pcb model if someone would dare to test
I had trouble with Calculix - it said some of the solids didn't have a material assigned.
However, I got some semblance of a result from Elmer. Sadly, I had to delete the results to get it small enough to upload. But here's a picture.
elmer_heat.png
elmer_heat.png (20.11 KiB) Viewed 399 times
There's still that weirdness with absolute zero on the FR4, but the copper does appear to be conducting and radiating very well.
Attachments
fem-test-fcad-pcb.FCStd
(694.21 KiB) Downloaded 6 times
Post Reply