Why does the entire fluid domain attain the pressure as vaporization pressure?

[abhu123patil]

Dear Team,

I am a research student and modelling a propellent flash-boiling flow inside a chamber of which only one side is open to atmosphere.

I want to simulate the propellent is initially stored with a liquid volume fraction of propellent (at 6 bar), filling only 1/3 rd part of the chamber. And rest 2/3 rd part of the chamber contains vapour volume fraction of propellent (at 1 bar). To achieve that, I want to patch the liquid volume fraction with the vapour pressure of 6 bar and rest of the part with vapour volume fraction at atmospheric pressure of 1 bar.

But after initialization, the entire chember attains the pressure equal to vapour pressure (i.e., 6 bar, an input value of caviation model for vapour pressure), and the Fluent does not allow to patch my desired 2/3 rd part of the chamber at atmospheric pressure of 1 bar. But, Fluent does allow me to patch the desired volume fraction.

Other Modelling Setup details as follows:

Fluent Version - 2024R2

State - Transient

Models - 3D, VOF, 2-Phase, Schnerr-Sauer caviation model.

Turbulence Model - k-omega SST

Mesh - Hexahedral

Boundary conditions - Only Pressure Outlet with gauge pressure of 1 bar

Material - Propellent liquid and Propellent vapour

Best Regards

[Rob]

Did you initialise and then patch? What zone/register did you patch? 

[abhu123patil]

Dear Rob,

Thank you for the response.

Yes, I initialized (standard and/or hybrid) and then patch.

Sorry, I could not understand your second question clearly.

I have entire fluid domian (no solid cell zone or no moving zone) and I used cell registers to patch.

 

Regards

[Rob]

The latter part was to check you were patching a region which you are. 

You should be able to patch pressure at the mixture level, other quantities may be at the phase level. 

[abhu123patil]

Dear Rob,

Thank you for the response.

Yes, I agree with your thoughts. But, somehow I am unable to patch pressure at the mixture level.

I am able to patch pressure which is above vapour pressure (greater than 6 bar) of the propellent but not below (lower than 6 bar) of it.

Regards

[Rob]

Please can you post screen shots? 

[abhu123patil]

 

Dear Rob,

Sure, attaching pictures for your reference.

Kindly let me know if you would like to see any specific picture apart from the attached one.

Regards

 

 

 

[Rob]

Initialise at 1bar and then patch liquid to 6bar. How can you patch the gas region to 1bar when there's not a register? Make sure you save before iterating as I suspect it's not going to be stable. 

[abhu123patil]

Dear Rob,

Thank you for sharing the suggestion.

I have already tried changing the initialization (gauge) pressure ranging from -10 bar to 5.99 bar, the setup always gets initialized at 6 bar.

The setup gets initalized to higher pressure (e.g., 6.1 bar and beyond) successfully. But, it fails to get initialized at any lower pressure.

Correct me if I am wrong, these trials are possible with standard initialization, which I have already conducted. The setup is not respecting my initial values for lesser pressure (< 6 bar), it is only respecting for higher values (> 6 bar).

In case of hybrid initialization, the results are the same. The setup always gets initialized at 6 bar (i.e., at vaporization pressure of 6 bar) automatically.

In addition to that, I am not able to conduct FMG initialization using TUI.

 

 

On the second point, I am not not at all patching gas region at 1 bar.

I want to initialize the flow field at 1 bar and patch the liquid region at 6 bar using the register. (Reverse is also not working for me just for a trial purpose).

 

I highly suspect that the initialization issue is somehow connected with the cavitaion modelling.

 

Regards

[Rob]

Possibly, switch it off and see. Cavitation assumes you have a liquid and then get gas as pressure drops, can't say I've tried a model that's already below the cavitation pressure when starting. That's not going to work well as everything will try and flash at once, will then try and condense (latent heat is checked) intantly and likely fall over. 

[abhu123patil]

Dear Rob,

Switching off the cavitation does resolve the issue of initailization and patching as per the requirement.

But as you mentioned, by switching off the caviation the flow field will get a sudden pressure drop from 6 bar to 1 bar without mass transfer from liquid volume fraction to vopour volume fraction, won't it?

 

Let me mention my desired goal, I want to analyze flash boiling by modelling mass-transfer. Where the propellent is at 6 bar in liquid state (at vapour pressure) and surrounding is at 1 bar.

Therefore, I want to initalize the setup at 1 bar and patch the liquid volume fraction at 6 bar. (Which I am unable to do).

 

Hence, I can not switch off the cavitation model.

 

Regards

[Rob]

I don't think cavitation is going to give you what you want. Flashing is a really nasty phase change problem, and you're likely going to need temperature and pressure effects, potentially on both phases. It's a topic we approach with extreme caution, and my starting point would probably be Eulerian multiphase and a very careful look at what was actually needed to hopefully avoid modelling it. 

[abhu123patil]

Dear Rob,

I agree with your thoughts, but that could be a different topic of discussion.

Currenlty, I am more intrested for solving a hurdle to simply patch the pressure. I guess that is the minimal initialization setup of a model.

My question would remian the same, why Fluent is not allowing me to patch the lower pressure to the liquid volume fraction when the caviation model is switched on? or Why does the entire fluid domain attain the pressure as vaporization pressure?

 

Regards

[Rob]

Initialise as liquid at 1 bar, then patch gas phase & then liquid pressure. I wonder if there's a check so you can't patch a pressure on the domain that's lower than cavitation value. Cavitation wasn't designed to mimic flashing and there are a few hard coded checks in Fluent to prevent impossible set ups: they're not always as sensible as you'd expect. 

[abhu123patil]

Dear Rob,

I guess there is a check for patching a pressure on the domain that's lower than cavitation value.

Thank you for your support.

 

Best Regards