Learning Forum

[monsboerns]

Hello Fluenters!

I really need someones help. I am currently simulating a multiphase flow in a rotating pipe. I am doing this simulation in 2d with Ansys Fluent.

Setup: There are 50 vol% droplets within a 50 vol% continuuous phase. Because of the rotation of the pipe around its axis, I have a backflow at the outlet. The outflow is against the atmosphere. I am using the mixture model, where i define my continuous phase with the dispersed phase as a second phase. In case of backflow, air should enter the domain (see also attached .png-file). The continuous and dispersed fluids are plastics with very high zero shear viscosities (continuous phase: 6500 Pas and droplets: 100 Pas)

Can somebody help me to set the backflowing air in a right way? How should this be properly done in Ansys Fluent, that it acts like a second continuous phase? I am currently not getting satisfying results as shown in the attached .png-file, where the air flows through my other two phases like they would not exist.

Thanks in advance for your help!

Best regards,

monsboerns

[Rob]

Why wouldn't it behave like you're seeing? The denser material moves towards the wall due to the swirl, and as the inflow area is smaller than the outlet you'll get a core forming (swirl will assist in this too).
What are the phases that you're using? The description isn't clear which is primary, and what are the secondary phases.

[monsboerns]

Hello Rob!
First of all, thank you for your comment! You are right, my description is not complete, sorry for that.
The red area in the picture above is 50 vol% LDPE (continuous) which contians 50 vol% PET (droplets). At the inlet this 50/50 vol% mixture is constantly entering the domain with u_m=0.1 m/s. I initialized the field with 0 of every parameter, except of the volume fraction, which i set to0.5, i.e. that the pipe is already filled with the LDPE/PET mixture.
Another thing is, that as soon as the backflow occurs, the residuals of continuity immediately start to grow to a high level (about 1e+00).
Before that i simulated just water in the rotating pipe. There everything worked fine if I turn on the CSF-model in the mixture model with the surface tension of water of 0.072 N/m. However, if I do the same simulation, now with 50 vol% water and another dispersed fluid phase of 50% with similar physical properties, i get the same weird shape of the core as shown in the second picture. In this case I set the surface tension of both phases to the value of water. So I assume that I am doing something wrong in the setup of the mixture model?
Best regards monsboerns

[Rob]

To clarify, you have a continuous phase LDPE, and two secondary phases, one is PET and the other is air?

[monsboerns]

Hello Rob!
Thats is right, yes.
My problem now is, how to model the backflowing air properly, that it acts like another continuous phase?
Best regards
monsboerns

[Rob]

If you want a continuous phase you'll need VOF or multifluid VOF. Does the PET separate in the LDPE or can it be modelled as a single phase?

[monsboerns]

Thank you for your answer!
Yes, the PET should separate from the LDPE due to the density difference and the rotation of the pipe. This is what I am actually interested in and it works pretty well so far. The air at the outlet should just interact with the mixture in a more or less physically correct way, so I am not interested in the interface between the mixture and the air.
Do you think such a behaviour of the air can be achieved in the mixture model?
Best regards
monsboerns

[Rob]

I think you'll be better off with the multifluid VOF.

[monsboerns]

Hello Rob!
Thank you for your help! I think this is truly a much better idea, so I will continue with the multifluid VOF. I am sure that this will give me better results regarding the backflowing air.
Have a nice day.
Best regards
monsboerns

[monsboerns]

Just one thing: Is it also possible to use the Schiller-Naumann- or other drag force models in the Euler Multifluid VOF Model to model the droplet movement of PET in the continuous LDPE phase?
Best regards
monsboerns

[Rob]

You'll find all of the phase interaction options in the panels, and guidance on their applicability in the documentation.