[Baum]

Hi, I am currently modeling new droplets appearing in a flow field based on condensation. My question is the following:

One of my species (KCl) is condensing from the gas phase mixture. I calculated the mass flow of KCl that condenses with one UDF, and subsequently added inert KCl-particles of the same mass flow as an injection in each cell. Now, I want to hook a source term UDF to the mixture to represent the amount of KCl that leaves the gas phase in each cell. Of course, I also need a source term for the energy of condensation, which is clear. But when it comes to the mass, do I need to add the source terms to kcl<g> only, or also add a source term for mass overall? Would doing both remove the mass twice? Would only doing it for kcl<g> remove the species, but not reduce the overall mass (however that would work?).

Thanks!

[vinerm]

You need to apply sources in both but there are situations when mass source could be 0. E.g., if there are three species, A, B, and C, with C being the last one or the most abundant one that does not require mass fraction equation to be solved, and A is converting into B, then total mass would remain same. There will be negative source for A and positive for B.

However, your case is different. You have a positive mass source for a particular specie. If you apply mass source for that specie, you will also have to add that to total mass, otherwise, due to the mass conservation, the most abundant specie will reduce. Therefore, if 0.1 is mass source for a specie, it should also be mass source for total mass.

[Baum]

Thanks Vinerm!

Two follow-up question for that:

1) If I set different under-relaxation factors for the species and density (which I guess represents the mass?), would problems occur because the species is getting removed more quickly than the overall mass?
2) would there be a momentum exchange as well? Or is that automatically introduced via the new particles (specifically their mass and velocity)?

[vinerm]

URF values certainly have effects, however, only if the solution is not converged. Once it has reached convergence, value of URF is irrelevant.

As far as sources are concerned, mass source should never be alone. It should always be accompanied with momentum, turbulence, and energy sources, depending upon whichever models are enabled. Otherwise, rest of the field gets redistributed and you will lose or gain temperature or momentum or turbulence, which could be incorrect.

[Baum]

But isn't it true that I am adding momentum by adding the particles (with a certain mass and velocity --> momentum) to the mixture? Thus, removing mass via sources (which reduces momentum) and adding particles with a certain mass mass (adding back the momentum) should cancel out, right?

[vinerm]

Mass is being removed from the continuous phase. If you do not remove the corresponding energy, momentum, and turbulence, then that field gets redistributed over the remaining mass. Particles do not belong to the Eulerian phase. Due to their interaction with the Eulerian phase, there will be momentum transfer but when you are injecting particles, you might be injecting those at a certain temperature, with a certain velocity, implying they have their energy and momentum. This has to have come from the Eulerian phase, otherwise, this is additional and will slowly be shared with the Eulerian phase, ultimately leading to increase in the momentum, energy, and turbulence.

When the mass is removed, only mass is removed. It causes redistribution of the momentum but its integral remains the same.