An aluminium heatsink is used to conduct heat away from electronic components. The heatsink fins are in a duct, through which air is blown.
Heat flux input to the solid domain is handled by SYRTHES. Fluid flow in the duct is handled by Code Saturne. The two solvers are coupled to handle heat transfer at the common boundaries.
Some transient variations are included, to represent intermittent heat source, and intermittent air-flow
Included here:
README.pdf - notes about the study cases
heatsink - the study directory (the meshes are missing from the MESH directory)
Mesh_fluid.med, Mesh_solid.med, Mesh_solid.syr.gz (the meshes are too big to upload in the study, so had ho be uploaded separately. Mesh_solid.syr.gz must be uncompressed. Move all meshes into the MESH directory in the heatsink study. )
Coupling with SYRTHES - forced cooling of a heat exchanger
Forum rules
As this forum is intended to post modelling examples, for better clarity, posts may be edited, and questions may be removed in some cases once the parent is updated.
As this forum is intended to post modelling examples, for better clarity, posts may be edited, and questions may be removed in some cases once the parent is updated.
-
Brennan Sharp
Coupling with SYRTHES - forced cooling of a heat exchanger
- Attachments
-
- heatsink-tar.gz
- (1.15 MiB) Downloaded 931 times
-
- README.pdf
- (81.75 KiB) Downloaded 1002 times
-
- Mesh_fluid.med.gz
- (3.94 MiB) Downloaded 942 times
-
Brennan Sharp
Re: Coupling with SYRTHES - forced cooling of a heat exchanger
And the remaining Mesh files ...
- Attachments
-
- Mesh_solid.med.gz
- (5.42 MiB) Downloaded 906 times
-
- Mesh_solid-syr.gz
- (4.66 MiB) Downloaded 896 times
-
Claus Andersen
Re: Coupling with SYRTHES - forced cooling of a heat exchanger
I haven't had a chance to run the test case yet, but thank you for sharing a working case :)
Regards and happy holidays
Claus
Regards and happy holidays
Claus
-
Charles Warner
Re: Coupling with SYRTHES - forced cooling of a heat exchanger
Brennan-
Thanks for the case study- most helpful, especially since there are some apparent discrepancies in the documentation...Specifically, the User's manual instructs one to use "link_syrthes" to build the executable, while one finds in the "Readme" file instructions to use "make". There is also mention of a GUI called "Edamox" for editing the parameters file which apparently no longer exists.
It would seem that, from your "Readme.pdf" you have not run Syrthes independently of Code_Saturne, and I have been trying to do this, running in to one problem after another- the latest about 2 pages of undefined references. You note that you only get coupled surface temperature data for the solid domain, and I would think that running Syrthes independently would give you results for the entire solid domain, based on the documentation.
Of course, one could always run Code_Aster to get the solid domain characteristics...
I am using Syrthes 3.4.2 with Code_Saturne 2.0 Beta 2 in the Salome-MECA 2010:x86_64 environment as built on CAELinux 2010. The fortran compiler I am using is gfortran, but it seems Syrthes wants g77. Is this a problem?
There does seem to be a dearth of information on Syrthes out there...
Thanks for the case study- most helpful, especially since there are some apparent discrepancies in the documentation...Specifically, the User's manual instructs one to use "link_syrthes" to build the executable, while one finds in the "Readme" file instructions to use "make". There is also mention of a GUI called "Edamox" for editing the parameters file which apparently no longer exists.
It would seem that, from your "Readme.pdf" you have not run Syrthes independently of Code_Saturne, and I have been trying to do this, running in to one problem after another- the latest about 2 pages of undefined references. You note that you only get coupled surface temperature data for the solid domain, and I would think that running Syrthes independently would give you results for the entire solid domain, based on the documentation.
Of course, one could always run Code_Aster to get the solid domain characteristics...
I am using Syrthes 3.4.2 with Code_Saturne 2.0 Beta 2 in the Salome-MECA 2010:x86_64 environment as built on CAELinux 2010. The fortran compiler I am using is gfortran, but it seems Syrthes wants g77. Is this a problem?
There does seem to be a dearth of information on Syrthes out there...
-
Brennan Sharp
Re: Coupling with SYRTHES - forced cooling of a heat exchanger
Hello Charles,
Glad it's useful.
Regarding SYRTHES alone, I have run it that way although not for a long while due to my application interest.
In a coupled simulation, each solver writes its results for its own domain, in its own output files. After a coupled run, you should (or at least potentially) have the solid domain results too, but they'll be in the SYRTHES output, which is there somewhere in the study/case/results. I allude to this, vaguely. in the last section of my readme, about post-processing. So, for the sake of clarifying this point here: from the results of the coupled simulations I've given, I can get full fluid domain and full solid domain results, within the scope of each solver, but they're not all in one results file.
It seems that nonetheless you've found a weakness in my notes, so I'll try to elaborate on getting the SYRTHES data, but please bear in mind that I've set out to demonstrate coupled simulation setup, with an expectation that readers are familiar with both solvers independently.
Regarding the build environment you have, I compiled SYRTHES manually, with gfortran as the fortran compiler, and I don't have g77 anywhere. I'm guessing that the FORTRAN compiler used to build SYRTHES is also required at run time. I can't comment on the CAELinux build of SYRTHES as I'm not using that distribution, but on my own installation the Make file exists in the installation directory e.g. /opt/syrthes3.4.2/bin/Makefile. With it is Makefile.in, which is where the FORTRAN compiler is specified, dependent on machine architecture. Mine is set to gfortran for the relevant architecture. As you need to run make when you set up a stand-alone SYRTHES case, I suspect the same Make files are used.
Regards,
Brennan
Glad it's useful.
Regarding SYRTHES alone, I have run it that way although not for a long while due to my application interest.
In a coupled simulation, each solver writes its results for its own domain, in its own output files. After a coupled run, you should (or at least potentially) have the solid domain results too, but they'll be in the SYRTHES output, which is there somewhere in the study/case/results. I allude to this, vaguely. in the last section of my readme, about post-processing. So, for the sake of clarifying this point here: from the results of the coupled simulations I've given, I can get full fluid domain and full solid domain results, within the scope of each solver, but they're not all in one results file.
It seems that nonetheless you've found a weakness in my notes, so I'll try to elaborate on getting the SYRTHES data, but please bear in mind that I've set out to demonstrate coupled simulation setup, with an expectation that readers are familiar with both solvers independently.
Regarding the build environment you have, I compiled SYRTHES manually, with gfortran as the fortran compiler, and I don't have g77 anywhere. I'm guessing that the FORTRAN compiler used to build SYRTHES is also required at run time. I can't comment on the CAELinux build of SYRTHES as I'm not using that distribution, but on my own installation the Make file exists in the installation directory e.g. /opt/syrthes3.4.2/bin/Makefile. With it is Makefile.in, which is where the FORTRAN compiler is specified, dependent on machine architecture. Mine is set to gfortran for the relevant architecture. As you need to run make when you set up a stand-alone SYRTHES case, I suspect the same Make files are used.
Regards,
Brennan