[Flexi1095]

Hey,

after I searched quite a while through the forum, I am a bit confused. Maybe someone can help me out.
Following Issue: I want to simulate a pipe with two inlets and two gases (each gas entering at one inlet). I already run a simulation with interFoam, which worked fine, just deltaT is very small.
But I'm confused, if I really have to use a multiPhaseSolver (like interFoam) for a flow with 2 gases (a steady-state-solve would be also fine for me)?
Would it be also work with e.g. simpleFoam? Can I also use alpha.* there like in interFoam?

Edit: Just tried to set it up with simpleReactingParcelFoam, which worked also quite good. Is this the right dircetion? But there I can't see the gas percentage in each cell of each gas, is this normal?

Sorry for my quite basic questions.
Thank you in advance and best regards

(Similar Thread, but I couldn't find the answer there:
Tracing a gas in a gas)

[crubio.abujas]

Originally Posted by Flexi1095

Hey,

Following Issue: I want to simulate a pipe with two inlets and two gases (each gas entering at one inlet). I already run a simulation with interFoam, which worked fine, just deltaT is very small. But I'm confused, if I really have to use a multiPhaseSolver (like interFoam) for a flow with 2 gases (a steady-state-solve would be also fine for me)?

Would it be also work with e.g. simpleFoam? Can I also use alpha.* there like in interFoam?

Well, typically multiphase solvers are based on the assumption that the phases are not miscible, this is, they have a clear interface separating them. In the application you mention of two gases mixing it is quite obvious that they are going to interact and this is not the case. So I would say that multiphase solvers are not the best option for your problem.

I think what you're looking for is to model different concentration of species inside a single phase fluid. All the fluid is a gas, but in any point there are different concentrations of the gases conforming it. Each specie (02, H2, CH4...) is model as a scalarField, a single number which determine de concentration in that specific point. The thermo-physical properties (cp, density, viscosity) may vary according with the composition on each point and that can be taken into account.

In this kind of simulation is frequent that you want to model some kind of chemical reaction, so you may want to have a look onto reactingFoam (inside the combustion category), which allows to do all I've told you in the paragraph before. Of course the reactions can be disabled if you are just working with inert substances.

However, if your case can be simplified (there is one main gas a small traces of the other, for example) you can try to model the main current with simpleFoam, and up on these results run a scalarTransportFoam, to see how the second-gas is distributed along the flow. This approach considers that the secondary-gas does not affect the thermo-physical properties of the main-gas, which may be consistent if there are small traces or their properties are alike.

Forgot about the alpha.* that is for the multiphase solvers, simpleFoam would simply ignore these files while running.

Concerning the steady-state, it is perfectly fine to work with it in any of the approaches to the problem I've discussed, if you're just interested in the final state.


I hope that makes it a little bit more clear.
Good luck!

[Flexi1095]

Hey crubio.abujas,

thank you very much for your detailed explanation! I will have a look onto reactingFoam and simpleFoam+scalarTransportFoam. If more questions come up, I'll come back.

Best regards