Learning Forum

[cbra_8]

Hello,

I'm wondering if someone would be able to explain a bit more about the theory behind the mixture model for multiphase flow in Fluent and how it works when the Shnerr-Sauer cavitation model is employed. After reading through the Fluent Theory Guide, I can understand that a transport equation for the vapor volume fraction is solved along with Navier-Stokes and continuity. I also understand that there is some consideration of bubble dynamics according to the Rayleigh-Plesset equation when the Schnerr-Sauer cavitation model is selected. However, I'm confused as to how the Rayleigh-Plesset considerations fit into the solution procedure and how it connects to the solution of the vapor volume fraction.

There is a setting for the "Diameter" in the definition of the secondary phase, which is treated as a constant value unless a user-defined function is specified. There is also a bubble number density input for the cavitation model. So, does the program initialize a certain number of bubbles within a local region according to this bubble number density and diameter? Then from this is the vapor volume fraction calculated?

I understand that the bubbles are not tracked and interfaces are not resolved. However, if the diameter value is constant over time, how can the local vapor volume fraction change?

Many thanks,

cbra_8

[Rob]

The bubble diameter is used to calculate drag on the bubbles, the amount of vapour is based on the phase change and isn't linked to a number of bubbles. 

[cbra_8]

 

 

Ok, so then if I understand correctly:

 

 

• At a given time step the program first solves the equations of motion for the mixture phase according to the multiphase model (mixture model in this case) along with the turbulence.
• Next, the cavitation model separately solves for the vapor volume fraction by accounting for evaporation and condensation mass transfer with the Rayleigh-Plesset equation for a specified nucleation site density. The pressure field from the multiphase model is an “input” to this calculation, and the radius of the bubbles is allowed to change according to the Rayleigh-Plesset equation.
• The vapor volue fraction data is then passed from the cavitation model back to the multi-phase model, which uses a fixed bubble diameter along with the vapor fraction from the cavitation model to determine a number of bubbles. The drag on these bubbles is then calculated and the motion of the secondary phase relative to the mixture phase is obtained.

 

 

Is my understanding correct? If so, then there are two separate considerations of bubbles, one within the cavitation model determined by the nucleation site density where the bubble size is allowed to change, and one within the multiphase model where the size of the bubble is not allowed to change and is only used for calculation of the drag between phases. Is this also correct?

Thank you,

cbra_8

 

 

[Rob]

Mostly. The individual bubble drag is calculated, but the solver doesn't calculate a number of (multiphase) bubbles, just a volume fraction. DPM uses parcels, the Mixture & Eulerian models use a volume fraction. 

https://ansyshelp.ansys.com/account/Secured?returnurl=/Views/Secured/corp/v231/en/flu_th/flu_th_sec_multiphase_cavitation.html  

[cbra_8]

Thanks, so how does the single bubble drag scale with the volume fraction then?

I can't seem to follow the link you provided. I work in a lab with an academic/research license so it seems that I don't have access to some customer support features when I try to follow the link. Is there any other way to access that information?

Many thanks.

[Rob]

Click on Help in Fluent and then copy the link into the browser. The documentation is not "public" in that it's for users only, so the solvers have a token/cookie/biscuit to bypass the log in page. 

The large bubble drag is corrected based on the phase interaction options from the phase bubble size - they'll be covered in the theory guide. Multiphase is deceptively easy to switch on, but then rather more difficult to understand.