[Giunto94]

Hello everyone, I am working on a combustor for my PhD project and I want to use the partially premixed combustion model for my simulations.
For this combustion model, Fluent automatically creates a "pdf-mixture" in the Material section where the viscosity and thermal conductivity are set constant by default.
However, I would like to define a UDF to make these two properties temperature dependent. I wrote an UDF like this:

DEFINE_PROPERTY(thermal_conductivity,c,t){
real k;
real temp;
temp = C_T(c,t);
k=function of temp;
return k;
}

The UDF is successfully compiled, with no errors, but when I click on "Recalculate properties", in the Partially premixed model window, FLuent crashes. Instead, If I write k=constant in the UDF, it works well. I can't figure out why it doesn't work...
I need your help, please.

Lorenzo

[LoGaL]

In partially premixed model the temperature is not a primary variable (i.e. you calculate it as post processing step, you don’t solve for it)

[LuckyTran]

C_T is a cell thread, needs a mesh, which has nothing to do with the creation of a pdf table

[Giunto94]

I understand what you mean and It makes sense, but...if there is the possibility to use polynomial/piecewise polynomial/etc... correlations to define thermal conductivity as a function of temperature, why with the UDF doesn't work?

[LuckyTran]

It's really simple... except it contradicts the partially premixed combustion model. There's two things going on

First is that the PDF table is a lookup table generated once. You are defining the thermal conductivity using a UDF that cells for local cell temperatures which makes no sense. To appreciate what this means, open up microsoft excel and type in C_T into a cell. Nothing happens, it makes no sense to call for a cell thread in a PDF table. And even if it id, the local cell temperatures would change each iteration, whereas the PDF table should never change.

Second is the PDF is generated as a function of mixture fraction. Even for non-adiabatic cases what Fluent does is it generates a stack of PDF's each functions of mixture fraction (one table at each enthalpy) and then interpolates between different enthalpies.


You need to get more creative if you want temperature dependent thermal conductivities in your simulation. And just FYI there is a set of macro's for overwriting the PDF table ( of the form DEFINE_PDF_TABLE) that might be of some use to you.