Simple case on numerical diffusion with CfdOF - "RuntimeError: Only isothermal simulation"

[carlocal]

Hi all,

I'm new to FreeCad/CfdOF forum, even if I'm kind of a "silent" user of it. First of all I'd like to sincerely kindly thank all the developers of the great CfdOF which has become for me a fundamental tool through which start using OpenFOAM (I've always been a Fluent user). I think the value of your work is inestimable, and - as little as I can - I would like to spread the use of this tool more and more - which in a way is revolutionary.

I open this discussion because I would like to replicate the classic example of false diffusion testing by Versteeg, H.K. and Malalasekera (to understand: something like this https://nimasamkhaniani.ir/openfoam/openfoam-schemes/) with different spatial discretization schemes.

This classic problem can be expressed by transporting a scalar, namely temperature. Therefore, I set up the simulation with two velocity inlets (activating the energy equation, and therefore setting two different temperatures) and with two pressure outlets. When I try to generate the file, on "Write case" it tells me that it is impossible to do so: "RuntimeError: Only isothermal simulation currently supported for incompressible flow."

Hence, I've tried different solutions: from incompressible to compressibile, from single phase to free surface, from steady to transient and so on. Nothing is actually working. How can a write a case under such conditions? I'm attaching here my simple simulation file.

A second question would then be: to change the discretization schemes, do I act directly on the files generated by the solver? Or is there a way to do this directly from the CfdOF workbench?

Until now I have always simulated cold-flow problems, and I have encountered difficulties that I have always solved through this forum and/or tutorials. It never occurred to me to work on temperature, that's why I try to ask here. Thank you very much in advance!

Carlo

[oliveroxtoby]

This classic problem can be expressed by transporting a scalar, namely temperature. Therefore, I set up the simulation with two velocity inlets (activating the energy equation, and therefore setting two different temperatures) and with two pressure outlets. When I try to generate the file, on "Write case" it tells me that it is impossible to do so: "RuntimeError: Only isothermal simulation currently supported for incompressible flow."

Hence, I've tried different solutions: from incompressible to compressibile, from single phase to free surface, from steady to transient and so on. Nothing is actually working. How can a write a case under such conditions? I'm attaching here my simple simulation file.

I would suggest you use the task panel to set up the PhysicsModel rather than changing the values in the Data panel. This will guide you to the allowed combinations. Setting Single Phase and Compressible is what you want.

Please make sure that your workbench is updated - some of the warnings I see suggest the model might have been created with quite an old version.

[oliveroxtoby]

A second question would then be: to change the discretization schemes, do I act directly on the files generated by the solver? Or is there a way to do this directly from the CfdOF workbench?

No, you will have to edit them directly.

[carlocal]

This classic problem can be expressed by transporting a scalar, namely temperature. Therefore, I set up the simulation with two velocity inlets (activating the energy equation, and therefore setting two different temperatures) and with two pressure outlets. When I try to generate the file, on "Write case" it tells me that it is impossible to do so: "RuntimeError: Only isothermal simulation currently supported for incompressible flow."

Hence, I've tried different solutions: from incompressible to compressibile, from single phase to free surface, from steady to transient and so on. Nothing is actually working. How can a write a case under such conditions? I'm attaching here my simple simulation file.

I would suggest you use the task panel to set up the PhysicsModel rather than changing the values in the Data panel. This will guide you to the allowed combinations. Setting Single Phase and Compressible is what you want.

Please make sure that your workbench is updated - some of the warnings I see suggest the model might have been created with quite an old version.

Thanks a lot! I updated the workbench and managed to write the case (compressible, single phase, RANS). Still I've got some issues: when running the simulation it stops because of radiative modeling not present. Should I adjust the files editing them by hand?

Thanks again.

[oliveroxtoby]

Thanks a lot! I updated the workbench and managed to write the case (compressible, single phase, RANS). Still I've got some issues: when running the simulation it stops because of radiative modeling not present. Should I adjust the files editing them by hand?

Thanks again.

I needed to correct a settings error, so please update your workbench. Also, you will need to set the pressure at the outlet boundaries to a positive value (e.g. 100 kPa) as the absolute pressure is used by the compressible solver.

[carlocal]

Thanks a lot! I updated the workbench and managed to write the case (compressible, single phase, RANS). Still I've got some issues: when running the simulation it stops because of radiative modeling not present. Should I adjust the files editing them by hand?

Thanks again.

I needed to correct a settings error, so please update your workbench. Also, you will need to set the pressure at the outlet boundaries to a positive value (e.g. 100 kPa) as the absolute pressure is used by the compressible solver.

Thanks a lot! Now it works. Maybe I will upload some results just for the sake of completeness.
Cheers