[sasenbb]

Hi,

Recently I'm doing my project about Rayleigh-Benard convection in binary mixtures.
In my work, the Soret effect(thermal diffusion) is considered. In the equation of Soret coefficient ,St=Dt/D, where St is Soret coefficient (1/k), Dt is thermal diffusion coefficient(m*m/s*k) and D is mass diffusivity(m*m/s) . Here comes the problem: when I try to input the thermal diffusion coefficient, I found the unit of thermal diffusion coefficient in FLUENT is (kg/m*s), not the same as (m*m/s*k). For example, the thermal diffusion coefficient Dt I want to input is -1.143e-11(m*m/s*k), so how much should I input in the unit of (kg/m*s)?

Any help would be appreciated!


Thanks, Sasenbb.

[Amir]

Dear 李炳威,

Any diffusion coefficient dimension is (m2/s); like momentum (nu), thermal (alpha) and mass (D) coefficients. In FLUENT when the dimension required is (kg/m.s), you have to use (rho* diffusion coefficient) which rho is base fluid density in which the property would be diffused. But as I know, thermal diffusion coefficient is alpha and its dimension is (m2/s) not (m2/s.k)!!!

Bests,

[sasenbb]

Thanks Amir for the response.But the thermal (alpha) diffusion coefficient you mentioned which dimension is (m2/s) is not the same parameter as I mentioned. Usually we call the thermal (alpha) diffusion coefficient (m2/s) thermal diffusivity, and call the Dt in Soret effect thermal diffusion coefficient(m2/s.k). And this can be found in many paper about Soret effect(thermal diffusion). And according to the Flunet 6.3 user's guide, thermal diffusion coefficient should be (m2/s.k), but when comes to the input interface, it change into (kg/m.s).

[Amir]

Originally Posted by sasenbb

Thanks Amir for the response.But the thermal (alpha) diffusion coefficient you mentioned which dimension is (m2/s) is not the same parameter as I mentioned. Usually we call the thermal (alpha) diffusion coefficient (m2/s) thermal diffusivity, and call the Dt in Soret effect thermal diffusion coefficient(m2/s.k). And this can be found in many paper about Soret effect(thermal diffusion). And according to the Flunet 6.3 user's guide, thermal diffusion coefficient should be (m2/s.k), but when comes to the input interface, it change into (kg/m.s).

Are you sure?! I just find one definition in manual; Eq. 8.9-1, and as I said before, its dimension is: kg/m3 * m2/s = kg/m.s
According to this equation, maybe you're considering D/T as a new coefficient! is it the case?!

Bests,

[sasenbb]

Thank you again Amir, I think I get it. According to <The soret effect: A review of recent experimental results>Eq.(1) ,and Eq. 8.9-1 in the Fluent 6.3 user's guide, DT,i in the Eq. 8.9-1 is not the parameter in the St=DT/D. And DT,i = rho*DT*C0(1-C0)*T, C0 is the mass fraction , DT,i(kg/m.s) DT(m2/s.k)

[ngothientu]

Hello Sasenbb
I also face similar to your problem. I set the mixture includes 2 components, the problem is that only the first specie appears in Thermal diffusion coefficient. If I add more species in the mixture, then go back the thermal diffusion coefficient, the first and the second species are appeared.
Could you please help me when I keep 2 species in a mixture and can input both thermal diffusion coefficient.??? If only 1 species can input, should I input for the solute (particle) (4% vol. fraction) or the solvent (96% vol. fraction)?
Thank you in advance.
Regards.