Best tool for modeling heat distribution in electronic enclosures
- Marcel
- Topic Author
- Offline
- New Member
-
Less
More
- Posts: 3
- Thank you received: 0
14 years 10 months ago #4328
by Marcel
Best tool for modeling heat distribution in electronic enclosures was created by Marcel
Hi all,
I am new to CAELinux (CAELinux 2009 based on Ubuntu 8.04 LTS on 64 bit machine) and would like to model the heat distribution in a closed electronic enclosure with several time-dependent heat sources.
Does anyone of you have already some experience in modeling such a problem and give me some advice which tool in CAELinux would suit my problem best? Is there a dedicated solver for such problems?
Up to now I am using CF Design for my calculations which mainly employs the finite element method as discretisation method. Is there maybe a tool that uses the same discretisation method so that I would be able to compare the results (just to see how big the difference is..)?
Any help would be greatly appreciated!
I am new to CAELinux (CAELinux 2009 based on Ubuntu 8.04 LTS on 64 bit machine) and would like to model the heat distribution in a closed electronic enclosure with several time-dependent heat sources.
Does anyone of you have already some experience in modeling such a problem and give me some advice which tool in CAELinux would suit my problem best? Is there a dedicated solver for such problems?
Up to now I am using CF Design for my calculations which mainly employs the finite element method as discretisation method. Is there maybe a tool that uses the same discretisation method so that I would be able to compare the results (just to see how big the difference is..)?
Any help would be greatly appreciated!
- Matthew Bondy
- Offline
- Premium Member
-
Less
More
- Posts: 98
- Thank you received: 3
14 years 10 months ago #4331
by Matthew Bondy
Replied by Matthew Bondy on topic Re:Best tool for modeling heat distribution in electronic enclosures
You could probably run this with OpenFOAM. It would very suitable since you have time-dependent heat sources. I think OpenFOAM is designed for custom boundary conditions. Someone might have their input files posted somewhere on the net for a similar problem.
Code Saturne has a nice GUI but I have never run a problem with heat transfer. I do not recall what options you have but I suspect the GUI would be insufficient for this problem. If you have any familiarity with french this would be worth exploring.
Best regards,
- Matt Bondy
Code Saturne has a nice GUI but I have never run a problem with heat transfer. I do not recall what options you have but I suspect the GUI would be insufficient for this problem. If you have any familiarity with french this would be worth exploring.
Best regards,
- Matt Bondy
- Marcel
- Topic Author
- Offline
- New Member
-
Less
More
- Posts: 3
- Thank you received: 0
14 years 10 months ago #4332
by Marcel
Replied by Marcel on topic Re:Best tool for modeling heat distribution in electronic enclosures
Hi Matt,
thanks for your reply. I already took into consideration using OpenFOAM but I thought this program was more suitable for solving problems in fluid mechanics. I don't know if this would fit for my problem since I only have free convection in my enclosure and no fast air or other fluid flows where a CFD analysis would be necessary.
What would be about Elmer? As I mentioned before I would like to employ a program that uses the finite element method as discretisation method since I would like to compare the results with commercially available software (and I only have CF Desing available that employs the FEM, too..). Maybe anyone has already modeled a similar problem with Elmer?!
Best,
Marcel<br /><br />Post edited by: Marcel, at: 2010/06/10 18:35
thanks for your reply. I already took into consideration using OpenFOAM but I thought this program was more suitable for solving problems in fluid mechanics. I don't know if this would fit for my problem since I only have free convection in my enclosure and no fast air or other fluid flows where a CFD analysis would be necessary.
What would be about Elmer? As I mentioned before I would like to employ a program that uses the finite element method as discretisation method since I would like to compare the results with commercially available software (and I only have CF Desing available that employs the FEM, too..). Maybe anyone has already modeled a similar problem with Elmer?!
Best,
Marcel<br /><br />Post edited by: Marcel, at: 2010/06/10 18:35
- Matthew Bondy
- Offline
- Premium Member
-
Less
More
- Posts: 98
- Thank you received: 3
14 years 10 months ago #4334
by Matthew Bondy
Replied by Matthew Bondy on topic Re:Best tool for modeling heat distribution in electronic enclosures
I assume the enclosure is filled with air? I believe Salome Meca has basic abilities for determining temperature distribution in solids. If you assume that convection is negligible it might be suitable. However, I think there are only a few options for boundary conditions. You could just use Code Aster directly or possible even modify the files created by Salome Meca somewhat manually.
You mentioned CFDesign, I have no experience with that software but it certainly has the ability to consider convection as fluid flow governed by the Navier Stokes equations. Have you disabled that capability in any analysis with CFDesign? Does it have simpler model equations for convection?
I once attempted a similar comparison (Fluent, Code Saturne, OpenFOAM) and like you suggest I wanted to compare software using the same mathematical technique (finite volume). In retrospect I should not have concerned myself with that issue. In your case this might be somewhat difficult, I have seen very few finite element CFD packages. (I am currently reading a book on finite element in CFD that suggests that the finite volume method can be a very specific implementation of the finite element method.)
I would not worry which method is used (finite element or finite volume), they should be equally valid. My biggest concerns would be using the same mesh, ensuring the same equations are solved (ie Navier Stokes), identical boundary conditions, and use of the same post processor.
I dont think there are any Elmer examples in the wiki (ie it would pretty cool if you could make one). Sometimes the biggest difficulty is creating and meshing geometry. If your enclosure is simple enough this might not be a problem.
Good luck,
- Matt Bondy<br /><br />Post edited by: Matthew Bondy, at: 2010/06/13 03:58
You mentioned CFDesign, I have no experience with that software but it certainly has the ability to consider convection as fluid flow governed by the Navier Stokes equations. Have you disabled that capability in any analysis with CFDesign? Does it have simpler model equations for convection?
I once attempted a similar comparison (Fluent, Code Saturne, OpenFOAM) and like you suggest I wanted to compare software using the same mathematical technique (finite volume). In retrospect I should not have concerned myself with that issue. In your case this might be somewhat difficult, I have seen very few finite element CFD packages. (I am currently reading a book on finite element in CFD that suggests that the finite volume method can be a very specific implementation of the finite element method.)
I would not worry which method is used (finite element or finite volume), they should be equally valid. My biggest concerns would be using the same mesh, ensuring the same equations are solved (ie Navier Stokes), identical boundary conditions, and use of the same post processor.
I dont think there are any Elmer examples in the wiki (ie it would pretty cool if you could make one). Sometimes the biggest difficulty is creating and meshing geometry. If your enclosure is simple enough this might not be a problem.
Good luck,
- Matt Bondy<br /><br />Post edited by: Matthew Bondy, at: 2010/06/13 03:58
- Marcel
- Topic Author
- Offline
- New Member
-
Less
More
- Posts: 3
- Thank you received: 0
14 years 10 months ago #4377
by Marcel
Replied by Marcel on topic Re:Best tool for modeling heat distribution in electronic enclosures
I now tried Salome Meca in combination with Code Aster. The results of my calculations are quite the same as the ones I obtained using CFDesign.
I think about trying Elmer, too, but up to now I had no time unfortunately.. But sometime I may have time to write a tutorial in the Elmer wiki
Best,
Marcel
I think about trying Elmer, too, but up to now I had no time unfortunately.. But sometime I may have time to write a tutorial in the Elmer wiki

Best,
Marcel
Moderators: catux
Time to create page: 0.128 seconds