Fluids

Fluids

Topics related to Fluent, CFX, Turbogrid and more.

Inconsistent results in VOF simulation including species transport equations

TAGGED: ,

    • Sangwoo Kim
      Subscriber

      Hello!

      I am running a bubble column reactor.

      • The column is filled with liquid metal. CH4 gas is injected from the nozzle which is submerged in the liquid metal, and is going to be bubbled. So, probably a VOF model is a good choice in this case.
      • CH4 gas is decomposed by the reaction CH4(g) -> 2H2(g) + C(s) in the liquid metal, because the liquid metal acts as a catalyst. So, without considering C(s) for the time being, CH4 and H2 (and other inert gases) can be treated as a gas mixture, constituting one "phase."
      • For the first simulation, I ignored chemical reactions and reaction products (H2, C(s)). 

      So my setup is as follows:

      • ANSYS Fluent 2020R1
      • 2D VOF simulation using two phases
      • Primary phase: Liquid metal
      • Secondary phase: Gas mixture (CH4, N2)

      This figure shows a contour of the volume fraction of phase-2 (gas mixture). The reactor is filled with liquid metal, N2 occupies over the free surface, and CH4 is injected from the submerged nozzle. Bubbling phenomenon is well captured.

      This figure shows a contour of the mass fraction of CH4 gas, which does not match with the contour of the volume fraction of phase-2.

      It looks like all the paths swept by CH4 is remained as CH4 region, even though the region was already covered by phase-1.

      However, Fluent Theory Guide says that when the species transport equation is solved with the multiphase flow model, the species mass fraction in the species transport equation is multiplied by the volume fraction of the phase. So, the mass fraction of CH4, which is a component of phase-2, should be zero at the cell occupied by phase-1.

      Am I missing some step?

      Thank you for your consideration. 

    • SRP
      Ansys Employee

      Hi,

      Make sure that the boundary conditions for the species transport equation are correctly set. Check the injection boundary conditions, particularly at the submerged nozzle, and ensure that the mass fraction of CH4 is properly specified.

      I suggest to use monitor to track the mass fractions of CH4 and N2 during the simulation. This can help you identify where the discrepancies are occurring and how they evolve over time.

      • Sangwoo Kim
        Subscriber

        SRP,

        thank you for your reply.

        This is the boundary conditions at the inlet. There is only one inlet.

        The boundary conditions in this case are too simple to get some hint..

        And, I don't think the boundary conditions caused this problem, because if the boundary conditions for CH4 were the issue, the contour of volume fraction should also reflect those things. However, it seems that it does not.

        This is the contours of volume fraction of phase-2 (gas mixture) and mass fractions of CH4 and N2 after more calculations.

         

        In your second suggestion, what is going to be the meaning of the mass fractions of CH4 and N2 in this case?

        When VOF model is used, both phases share all the pressure, density, velocity, and so on.

        Then, when I want to calculate and monitor, for example, the total mass of CH4 introduced into the domain, what is going to be the right formula?

        In my problem, the mass fraction of CH4 at some cells occupied by liquid metal is given by one. Then, at those cells, the mass fraction of CH4 is going to be multiplied by the mass of liquid metal, which does not make sense.

        Anyway I am going to monitor some variables related with the mass fractions, and will post it later.

        Thank you for your suggestions, SRP. Any other replies will be highly helpful and appreciated.

Viewing 1 reply thread
  • The topic ‘Inconsistent results in VOF simulation including species transport equations’ is closed to new replies.