Learning Forum

[prebenjs]

I'm using VOF and want to model a outlet where only one of the phases is allowed to exit the domain. The phase exiting is a gas while the phase being held in the domain is a liquid. How do i model this in Fluent ? Is it possible?

[aitor.amatriain]

What are you modeling?

[prebenjs]

I'm doing some simulations where I want to remove air, while allowing for the liquid to stay inside the domain. This means that if water is on the outlet, it should not be "drawn" out of the domain. This might also be able to implement via a reaction ? Whats your though on this?

[DrAmine]

Best to achieve that through the deployment of source terms. You define your outlet as wall and you just remove the phase of interest (including the secondary fluxes) out of the domain via an UDF or Expressions ( better to use UDF to be able to provide source term linearization). Another alternative is to create surface reactions but this is bit complex.

If you guarantee that the outlet is only covered by the phase you want to remove then a velocity outlet can be used.

Cheers!

[Murari Iyengar]

Hi prebenjs To do this, you can edit the phase option under the outlet boundary condition. For more information on VOF, please refer the link below.
14.3. Volume of Fluid (VOF) Model Theory (ansys.com)

[prebenjs]

Interesting options. The outlets will be at some point be covering the outlets so velocity is not an option. How do you implement this through expressions ? I've actually thought about implementing as an expression but I couldn't figure out how to remove the one phase by doing this ?
Regarding the UDF option. How would this best be implemented ? Through the mass transfer UDF ? Would this only be hooked through the "source term" in the cell zone conditions ?
Surface reactions is another option I've thought about but i couldn't really figure out how to do this. My thoughts was that this needed to be defined as a heterogeneous reaction (since gas to liquid), and then i need to assign the specific wall in the UDF as well, but it stopped there.

[prebenjs]

What outlets are you specifically referring to here ? I have not seen anything like what you're talking about ?

[DrAmine]

UDF: is not a phase change so you require DEFINE_SOURCE to extract the phasic mass and bulk mass. Same can be done with expressions if you can separate some cell layers towards the boundary of interest or by just using cell register which will make the cells where you want to add a sink.
I prefer UDF as it allows for linearization (if you can linearize your sink)

[prebenjs]

Thank you for quick response. I don't know how your knowledge is within UDF and expressions in Fluent, but would it be possible to get a short example ? (of a UDF and an expression)
I thought that the easiest way would be to define it via expression and cell registers, but I'm struggling with how this should be implemented and how to write this in the expressions language.
How would such a DEFINE_SOURCE to extract these masses be set up ? I've seen an example of a degassing source (in the manual) which is almost the same principle. But since i'm using VOF i won't need all of the secondary _recoil terms ? (opposed to if using eulerian). Would be really nice with a little example here if you have the knowledge.

[DrAmine]

Examples are available in Fluent User's guide for Expressions and Customization manual for UDF. That example for degassing is good one. But you just need to provide a sink term when cell register is True (cell register are boolean).

[prebenjs]

Yes i've seen what is in the User's guide and customization manual but i feel like they lack quite some bit on sink terms and for my application.
That example for degassing must be for eulerian since there are functions for both phases, while in VOF this should be coupled ? or should it be the same ? Could you also explain what is the recoil functions that's in the example ? are these momentum sinks ?
My second question would be how to formulate the right syntax for the expressions since if implemented by expressions would be something similar to:
IF (outlet_cells, IF(volume_fraction="gas-phase", sink term, nothing happens), nothing happens)
But the IF condition should only take a vector so I'm uncertain if this actually would work ? How would a sink-term be formulated in expressions (not any examples of this) ? How do i formulate that nothing should happen? Any other comments on this expression is really appreciated.

[DrAmine]

You require a sink term for secondary phase and bulk. So that should be similar. For momentum you require just one sink term as you just solve for bulk momentum.
With the expressions: is not so correct. You need to tell that if you are in the outlet cells and that volume fraction of gas phase is larger than 0.95 (or something) then you apply a sink term otherwise nothing. Why vector if you can define the losses component wise?

[prebenjs]

Hmm okay. And to implement this at a surface I need to create a user defined memory of cells that is set before the sink (when looking at the degassing example) ? Or is it possible to assign this to a surface in the GUI after interpretation or compilation ? I've just seen that source terms needs to be hooked in the cell boundaries but is it possible to connect to surface ?
I'm just struggling with how to define at location in the expression. So take an example:
IF(Outletcells AND ((Volumefraction = gas-phase)@wall > 0.95, sink term, nothing happens)
How do i define that the volumefraction should be at this location in Expression syntax ? And how do i define that nothign happens ? There isn't really any good examples of this in the user guide.