code_saturne User's Forum

Hello,

Does the CDO solidification model being compatible with the conjugate heat transfer within the V9.0 ?

Thank you in advance.

Kind regards.

Hi SS2X,

Normally it is. Is there a specific problem / difficulty ?

Hello Chai,

Thank your for your feedback. I think that I noticed in that forum that it was not possible, hence the question.

Kind regards.

Hello Chaï and SS2X,

Using CHT and CDO solidifcation together is possible. The way to setup this problem is possible with user-defined source files since it is a full CDO computation (meaning that there is no call to the legacy Finite Volume functions). I don't think there is a example of such a case settings.

The spirit is closer to an "internal coupling" than a code-code coupling. So, you a need a single mesh (if you have two meshes, you can perform a joining). A solid volume zone has to be defined on a part of the computational domain. This can be done using the GUI (page called "Volume conditions").

There are still some limitations (this functionality is recent), for instance, a constant mass density (a Boussinesq approximation is done in the fluid part) and no turbulence modelling.

Best regards,
Jérôme.

Hello Jérôme,

Thank you for your reply . Here below some complement from y side :

1) Regarding the mesh, I think that a bug or a functionality limitation occured, since the mesh joining (I have to get the operation's name afterwards and will give it here) was already performed on SALOMÉ and not double-face was detected at the fluid/solid boundary. and the same operation was performed along the boundary and despite this, the internal coupling was only activated on one of the border;

2) Using the soldification model in combination with the internal thermal coupling, an error arised by defining a value for the heat capacity in the fluid or solid zone. Moreover, would it be possible to have variable lambda and cp within the solidification model ? It seems that a case with CHT without CDO-solidification and fully-dedfined from the GUI writes in the ouput source order a file named "cs_meg_volume_function.cxx" (automatically written by CS) where the given law for thermal conductivity or heat capcity in each fluid or solid zone are implemented. I then tried to copy that file into the SRC order for the same case with solidification, the calculation ran, but the calculation was unstable. Would you then have a patch to combine the solidification model and varying lambda / cp (especially within the medium being simulated by the CDO-solidification model) or is it up-to-dae only compatible with CS properties defined as scalar ?

Thank you in advance for your intention .

Kind regards,

SS2X.