Hello everybody, i´m working in thermaldiffusion and i want to use Fluent for this. After creating a liquid binary mixture (water-ethanol), i include the thermal diffusion coeficient of the ethanol and the solver is for species transport with the thermal diffusion. My question is, when must i include the mixture density? at the initialization or in the material propertys with the "incompressible-ideal-gas" option? And the mass-fraction of the ethanol?

Thank´s

"Incompressible-ideal-gas" is NOT the right density formulation for a mixture of two liquids. Use "volume weighted mixing law" and define the density of your two liqudids as a function of temperature (piecewise linear or polynomial). For a mixture of two species only one specie will appear in the boundary conditions as the mass fractions of both will sum up to unity.

RoM

Thanks RoM. I´d changed it. It looks better, but not yet. I´m working in 2D and I´ve 4 boundary walls which 2 of them are adiabatic and the other have 5ºC temperature difference. I just changed to temperature dependence of density and initialized to 0,329 mass fraction of one component. After solving, it says that more density component fall down by the hottest wall and raise by the coolest wall. I think that´s not ok!!. It must be the opposite.What´s happen??? And another question, can i see the velocity vector of one component?

Is your domain closed (no inlets and outlets)? Did you turn on gravity and is the gravity vector pointing in the "right" direction?

RoM

Hi RoM. My domain is closed, there is no inlet and outlet. Only four walls. The gravity vector is pointing down the y axis. What must i watch more? It seems to be everything all right, but we don´t know why the liquid densest loss by the wall, when it would have to be the opposite.

Once again, I am here.

I have two questions for you.

First: What is understood by "Thermal Diff coef"? It is that it is called, the thermal difusivity? We understand, it like the DT whose unit is (m2/sK). It is the correct thing? I suppose that no. The units are not the same. What can you say to me? Second: what would have to identify in Cp of the mixture? Mixing-law/Cte/ or Piecewise-linear?

For thermal diffusion please read chapter 8.8.2 of the users guide (multicomponent diffusion). Thermal diffusion is explained in equation 8.8-7, it basically means that molecules will diffuse toward heated surfaces. I dont have much experience in this detailed subject, but from a personal feeling i would say that thermal diffusion is at least one order of magnitude smaller than mass diffusion. If it would dominate we could just heat our alcohol, wait for the separation due to the thermal diffusion, decant the water ... and voila pure alcohol. Unfortunatly i have never seen this happen but i may be wrong about it .

RoM