Hi All,

I am a CFX 5 user and I am attempting to solve a flow where the "effective density" is a function of an additional variable. I am having all kind of troubles with it (I get by CFX-PRE but CFX-SOLVER crashes). Did anyone ever succesfully implemented such formulation in CFX 5?

Sebastien,

Doesn't the density have to be a function of the thermodynamic state of the fluid? Are you assuming ideal gas? What exactly do you mean by the "effective density"?

Perhaps I don't properly understand your problem, but maybe I can be of help with more information.

I am a relatively new CFX 5.6 user and am working on a somewhat similar problem (I want the additional variable to contribute to the overall pressure). I might be wrong, but I believe an additional variable is considered massless. Does anyone know if that is correct?

Jonathan

Jonathan, when the fluid is an ideal gaz, yes the density is thermodynamic variable and it is not a good idea to change it at will. But In my case, I am considering an incompressible fluid with a constant density. The "effective density" would me defined as rho*add_variable where the add_variable is a volume fraction. It is for a very special model.

I have tried several things with fortran subroutines, but it appears that CFX is in a way very well built: it may even crach when the density is a function of variables that are not built-in. (No there was not any bugs in my routines..).

I am about to consider another approach to implement my physical model.

Probably not the same, but we have used gases with densities that depended on P and T in a non-ideal way. Then we had to give in that density varied on 'Pressure' and 'Temperature'. I think CFX than makes a look up table for the different states. Just guessing, but could it be that your look up table is not generated automatically. Call your local representative...........

Astrid

Hi Astrid,

I have tried making the density of function of pressure and tempereture: it works well. The only problem comes when the density is a function of an additional variable.

Thanks anyways.