CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (https://www.cfd-online.com/Forums/openfoam-solving/)
-   -   ReactingFoam specie transport equation (https://www.cfd-online.com/Forums/openfoam-solving/59474-reactingfoam-specie-transport-equation.html)

 alberto August 28, 2007 19:15

During a discussion on IRC. a

During a discussion on IRC. a doubt about reactingFoam implementation came out.

In reactingFoam, the transport equation for the specie mass fration contains the diffusional term in the form:

fvm::laplacian(turbulence->muEff(), Yi)

which means that the effective specie diffusivity is assumed to be equal to the effective viscosity:

Deff = D_lam + D_turb = nu_lam/Sc_lam + nu_turb/Sc_turb = nu_lam + nu_turb

which means that

Sc_lam = Sc_turb = 1

Why this assumption is done?

For example, in a laminar calculation, if we consider water, the viscosity of the fluid and the diffusivity of a specie, for example the oxygen, can be very different. At 25°C we have:

D_02 ~ 2.0 * 10^-10 m^2/s
nu_H20 ~ 1.0 * 10^-4 m^2/s

In turbulent flows, the turbulent Schmidt number is set to 0.7 in other codes, but however it can change too. For example, in a pipe with passive transport it's between 0.6 and 0.9.

Thanks in advance for any clarification.

With kind regards,
Alberto

 ly.foam March 27, 2009 08:44

I guess the reason is that when dealing with turbulent combustion, the molecular transport are often small compared to their tubulent counterparts and so are neglected.
But I think it is very important of you to have pointed it out that when dealing with laminar combustion with OpenFoam, special attention may have to be paid to the issue. But I am not very sure about this.

 smehdi609 March 27, 2009 13:22

You are right alberto, I usually use fvm::laplacian(turbulence->muEff()/Sc, Yi) and set Sc=0.7.