Heat Transfer. Cooling instead of heating.

Please login with a confirmed email address before reporting spam

Dear community,

I am trying to model the self-heating of a PCB. My model has a CPU as heat source and I want to know the temperature at 3 different spots on the board.

Even though I have the initial temperature values and the ambient temperature at 37°C, and am adding energy to the system through the heat source, the temperature drops below 37°C at one of my points of interest. How is that possible? What am I doing wrong?

Best, Matthias



5 Replies Last Post 2023年12月18日 GMT-5 11:07
Robert Koslover Certified Consultant

Please login with a confirmed email address before reporting spam

Posted: 4 months ago 2023年12月15日 GMT-5 15:29
Updated: 4 months ago 2023年12月15日 GMT-5 15:28

I think you have a time-stepping problem. I re-ran your problem, manually forcing it to take 20s time steps and it gave far more believable results. Go to Study --> Solver Configurations --> Solution 1 --> Time-Dependent Solver 1. In the Settings panel, click on Time Settings and change the value of "Steps taken by solver:" from "free" to "Manual." Then change "Time step:" from its default (.01) to some number like 20 (in your case, this is 20 seconds). Then re-run your model and you'll notice a very different result! I encourage you to also experiment with other different time steps and different meshings. Also, your present mesh is too coarse (in my humble opinion), in some places.

-------------------
Scientific Applications & Research Associates (SARA) Inc.
www.comsol.com/partners-consultants/certified-consultants/sara
I think you have a *time-stepping* problem. I re-ran your problem, manually forcing it to take 20s time steps and it gave far more believable results. Go to Study --> Solver Configurations --> Solution 1 --> Time-Dependent Solver 1. In the Settings panel, click on Time Settings and change the value of "Steps taken by solver:" from "free" to "Manual." Then change "Time step:" from its default (.01) to some number like 20 (in your case, this is 20 seconds). Then re-run your model and you'll notice a very different result! I encourage you to also experiment with other different time steps and different meshings. Also, your present mesh is too coarse (in my humble opinion), in some places.

Jeff Hiller COMSOL Employee

Please login with a confirmed email address before reporting spam

Posted: 4 months ago 2023年12月15日 GMT-5 16:03

I second Robert's view on the coarseness of your mesh. You may want to try a swept mesh, your geometry lends itself to it.

Jeff

-------------------
Jeff Hiller
I second Robert's view on the coarseness of your mesh. You may want to try a swept mesh, your geometry lends itself to it. Jeff

Please login with a confirmed email address before reporting spam

Posted: 4 months ago 2023年12月16日 GMT-5 05:52

Thank you very much for both of your replies!

Unfortunately both suggestions don't work out for me. I tried really hard to create a swept mesh but am unable to. I always get errors. I read through the tutorial and tried to gain knowledge from the bipolar transistor example but I simply can't produce a swept mesh.

As for the time-stepping: Once I change it to 20 seconds (I also tried 10s), it produces very weird jumps in temperature, that are also not explainable. I attached an image of the table.

I would be very grateful for additional help.

Thank you very much for both of your replies! Unfortunately both suggestions don't work out for me. I tried really hard to create a swept mesh but am unable to. I always get errors. I read through the [tutorial](https://www.comsol.com/support/learning-center/article/Introduction-to-Swept-Meshing-51861/152) and tried to gain knowledge from the [bipolar transistor example ](https://www.comsol.com/model/3d-analysis-of-a-bipolar-transistor-19699) but I simply can't produce a swept mesh. As for the time-stepping: Once I change it to 20 seconds (I also tried 10s), it produces very weird jumps in temperature, that are also not explainable. I attached an image of the table. I would be very grateful for additional help.


Please login with a confirmed email address before reporting spam

Posted: 4 months ago 2023年12月17日 GMT-5 10:26

I rebuilt the whole model step by step and found out that the weird results came by adding the vias. Because when adding the vias, I could not create a swept mesh anymore as the vias throw a "Unsupported topology of linking face" error.

That's why I exchanged the vias with cylinders with the same volume as the ring of the via. I can now create a swept mesh and get results that seem plausible.

This is not a solution however, just a workaround.

I rebuilt the whole model step by step and found out that the weird results came by adding the vias. Because when adding the vias, I could not create a swept mesh anymore as the vias throw a "Unsupported topology of linking face" error. That's why I exchanged the vias with cylinders with the same volume as the ring of the via. I can now create a swept mesh and get results that seem plausible. This is not a solution however, just a workaround.

Jeff Hiller COMSOL Employee

Please login with a confirmed email address before reporting spam

Posted: 4 months ago 2023年12月18日 GMT-5 11:07

Hello Matthias,

Page 850 of the Reference Manual, version 6.1, explains the requirements that your geometry needs to meet for a swept mesh to be possible. This will guide you on how to create your vias correspondingly.

Best,

Jeff

-------------------
Jeff Hiller
Hello Matthias, Page 850 of the Reference Manual, version 6.1, explains the requirements that your geometry needs to meet for a swept mesh to be possible. This will guide you on how to create your vias correspondingly. Best, Jeff

Reply

Please read the discussion forum rules before posting.

Please log in to post a reply.

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.