[hmdl]

Hello everyone,

has anyone some experience in modelling melting and solidification in conjugate heat transfer in CFX? I am trying to model the phase change as a single step reaction. I defined the solid and liquid phases as two distinct materials, the liquid material having the reference enthalpy equal to the latent heat of the material. I created single step reaction, the reactant the solid, the product the liquid. For the reaction rate, I defined a CEL with an if condition, 1000 over the melting temperature, 0 otherwise. I created a reacting mixture and defined it for the default domain. The thing is, when I run it, the reaction doesn't take place . What am I doing wrong here?

Regards

[ghorrocks]

CFX does not have a solidification/melting model, so if you are going to model this you are going to have to set it up yourself. And I would not expect it will be easy..... You might have an easier time using a CFD software which has a solidification/melting model.

I have no experience in it myself, but many people over the years have raised questions on solidification/melting models. Do a search of the forum and see if they help.

Also: There are a few possible approaches.
* You can use a single phase approach but a momentum source term to "freeze" material which you deem to have solidified. This is simple to do but modelling subcooling and latent heat effects is a bit tricky.
* You can have a liquid domain with single phase liquid and a solid domain, and use moving mesh to move the interface. Again getting the heat right at the phase change is challenging.
* Your idea of using a reaction to model it is interesting. I can't see how you model subcooling/superheat with this method but if that is not significant then good. But I would expect some hoops to jump through and convergence difficulties as you try to get this to work.

So I have not answered your question but hopefully this is some useful background.

[maher9149]

Hmdl. Did you solve this problem?