Problem with tin melting...

Marek

Hi. I'm new in Fluent and like I've mentioned in subject I have a problem with melting. I'm trying to model a problem fully described at:

http://www.fast.u-psud.fr/~gobin/bench.html

I have set everything, like it is described there (2d, incompressible, unsteady, laminar, no slip, all thermophysical properities and Boussinesq approximation) but I get incorrect results. I've tried to play with mesh, time step and under relaxation coefficients, but still nothing. My results are similar in general, but the melting process takes place to quick.

There is only one thing which I'm not sure. Source terms added to momentum and energy equations. I've read about that in few studies about melting in rectangular cavity. They are added in momentum eq. to account for the velocity suppression. I assumed, that Fluent calculates them itself, but now I'm not sure. So my new idea is tu prepare an UDF and add them to equations. Am I correct? ;-)

If yes, there is a little problem With the exception of last few hours spent on reading manual, i have no idea about writing UDFs and programming in C. So I need some help.

The forms of source terms are:

- x momentum:

S= -A*u, where A=(C*(1-f)**2)/(f**3+q) ;f-liquid fraction

- y momentum:

S= -A*v - r*g*b*(h-h ref)/cp, where r-density, b - thermal exp. coef.

- energy

s= r*L*(df/dt)

Full description is available as pdf file at

http://tinyurl.com/ax7tq (e.g. Appendix II)

Do someone know which of them I need to specify in UDF, and which of them Fluent calculates itself?

As I've mentioned I have almost no idea about C, so every help will be appreciated. My conception is to change the example macro (DEFINE_SOURCE) from manual a bit, but I don't now how to deal with f - liquid fraction from my formula.

Thanks in advance.

Marek