heat transfer
 

Hey guys,
Im learning how to use the standard solvers of OpenFoam and had some succes when running channel w/o heat transfer and turbulence.
Im trying to simulate a laminar flow in a pipe with heat transfer between the walls and the fluid. Im using buoyantPimpleFoam and I copied the thermophysical properties from tutorials as fvSchemes and Solution.
When I run the solver, the follow error appears:

--> FOAM FATAL ERROR:
Maximum number of iterations exceeded

From function specieThermo<thermo>::T(scalar f, scalar T0, scalar (specieThermo<thermo>::*F)(const scalar) const, scalar (specieThermo<thermo>::*dFdT)(const scalar) const) const
in file /home/opencfd/OpenFOAM/OpenFOAM-1.7.x/src/thermophysicalModels/specie/lnInclude/specieThermoI.H at line 67.


I dont know if Im defining wrong boundary conditions or if I missed something...

Thanks for the attention, hope someone can help...

I have the same error when using chtMultiRegionSimpleFoam to model 3D compressible flow in a duct. From what i read, it means that at some point you are dividing by zero. See if your iterations before the error have negative temperatures or if it is doing 1000 iterations when solving for pressure.

most likely, you are going to have to double-check your boundary conditions.

I suggest you post your boundary conditions and hopefully somebody can help you out.

Also, if you are modeling laminar flow... why not use a "simple" heat transfer solver?

Anyway... best of luck to you

Hi,

I'm facing the same issue with my cht setup.
The error appears without any prior sign - no negative temperature, no iteration barrier reached.
Running my case with different parallel setup (number of cores allocated) results in the exact same log file values, however the simulation blows up at different time steps, strange.

My case consists of a solid (two region connected) surrounded by air. The boundary setup can be found in the attachment - my goal is to simulate the airflow around a heated object, so I tried to mimic an open boundary around the solid which sits on a "table" and the bottom part is heated.

Any input appreciated!

Time steps?
 

Hi Toorop

have you tried reducing the time steps? You are using 1 sec. fixed time step. I suggest using variable time steps with Courant about 0.5. That might help but I only did a fast check of your data.

Regards

Fabian

Hi Fabian,
my solver is chtMultiRegionSimpleFoam, so its a steady state solver. Thus, the time step is just an iteration counter, correct me if I'm wrong.

Checked my case without parallelization and the simulation was OK. No clue what is going wrong in parallel execution.

Hi Toorop

right. I should have read your post more closely. Have you had a glimpse on the bug reports?

http://www.openfoam.com/mantisbt/my_view_page.php

As the solver works without parallelization this might be a bug.

Regards

Fabian

Thank you for the hint, I'm running a few more cases just to be sure it is indeed a bug or something else. If the problem can be reproduced I will make a report about it and post here the conclusions.

Hi Toorop,
did you try chtMultiRegionSimpleFoam with :

?
I tried it and it crashes when it solves hEqu because it calculates rho/psi.
I opened a new thred:

http://www.cfd-online.com/Forums/ope...implefoam.html

I'd like to use the buoyantSimpleFoam with the incompressible thermo type to simulate liquid water.
Thanks for any help
Andrea

Hi Andrea,
unfortunately I have no experience with the incompressible thermo type, but I will follow your thread on the issue.

I'm currently working on the strange blow up of my simulation, and investigating the reproducibility of the situation.

Best of luck to your project!