[Mengqi_LIU]

Hi,

I am working on a flat fan spray modeling task using the VOF-Spray One-Way Coupling method in CONVERGE v3.0.

I noticed that the drop size distribution of one-way coupling output concentrates on a narrow range, which is quite different from the drop size measurement results by laser diffraction.

I tried to turn on the user-defined drop distribution in injector setting. But it did not work. It seems that the drop size in one-way coupling method only depends on the nozzle diameter. I can modify the drop size by turn on “10-Nozzle diameter specified” in injector setting and change the nozzle diameter, but the drop size distribution is still quite far away from experiments.

I have several questions:
In CONVERGE manual, the parcel radius of one-way coupling method is calculated using below equation:

Equation 16.35: parcel radius = Ca^0.5*0.5*Diameter

Is diameter in equation 16.35 related to the cell size at region interface in VOF simulation? How can I define Ca in this equation?

Is there any other way to tune the drop size distribution in flat-fan spray modeling? Is it better to use ELSA model instead of one-way coupling method?

Thanks.

[Lu_CONVERGENT_SCIENCE]

Hi Mengqi

I am very glad that you are using the VOF-Spray One-Way coupling method in our CONVERGE solver.

You are right for VOF-Spray One-way coupling, the parcel radius at the coupling interface is uniform, and calculated by Equation 16.35.

Diameter in Equation 16.35 is calculated by the area of the nozzle from VOF simulation (assume a circular nozzle shape), so it will be slightly affected by the cell size in VOF simulation.

Ca here means the concentration coefficient, it is the ratio of discharge coefficient and nozzle velocity coefficient, which can be set in the spray model.

The drop size distribution at the interface may differ from the size distribution you got from laser diffraction where there is a distance from our interface. So to compare the CONVERGE simulation result with your laser diffraction result, you'd better check the size within the chamber, not just after the interface.

If you want to have a better control of the size distribution, you can try our normal spray modeling, where we have different size distribution models (Page 340, Eq 11.7).

ELSA is more computational expensive then the vof one-way coupling, but it has better control of size distribution.

Sincerely

Lu

[Mengqi_LIU]

Hi Lu,

Many thanks for your very clear answer!

I have tested the approach you mentioned, and I have three questions left:

First one, I changed the discharge coefficient (cd) in spray model after getting the vof_spray.dat file. But the parcel radius in simulation output did not change. I noticed that there is a Ca value in each time step of vof_spray.dat. Does that mean the Ca is coming from vof_spray.dat file instead of the spray.in file in One-Way coupling simulation? If so, how can I tune the Ca in vof_spray.dat file?

Second question, based on my test results, although User-defined button is an available option in CONVERGE studio. Turn on this option and then define injdist.in does not change the parcel radius in simulation output. It means the drop size distribution cannot be user defined in One-Way coupling simulation, is that right?

Third question, if I use normal spray modeling. The spray shape in CONVERGE is either full cone or hollow cone. Can I define the spray shape as flat fan spray?

Best Regards,

[Lu_CONVERGENT_SCIENCE]

Hi Mengqi

Those are all good questions, it took time for me to answer those question. I am sorry for the delay here.

For Ca value in vof one-way coupling simulation, Ca value is from vof simulation, seems calculated based on the void fraction, so if you want to change this value, you need to change your VOF simulation settings.

For VOF one-way coupling simulation, it seems user definded drop size distribution is not activated, so we cannot change drop size distribution by this feature. I will talk with development team or GUI team to grey out this option or add user defined feature in our CONVERGE solver.

For vof one-way spray simulation, if you want to have your own distribution (like your flat fan), we suggest that you define the VOF-Lagrangian simulation interface at downstream of the nozzle (currently we suggest place this interface within nozzle to generate uniform drop size), and use a box shape interface (a box with all surrounding surface set as interface). With this setting, the droplet will keep the original distribution and velocity, which is determined by the shape of your nozzle.

Please let me know if you have further questions.

Sincerely

Lu

[Mengqi_LIU]

Hi Lu,

Thank you for your kindly reply,

Using interfaces to set the VOF to one way coupling transition is a good suggestion. I actually tested similar way and it can make the results better.

I have no further questions for now. I am looking forward to the new CONVERGE version and I believe that this powerful software can become better with the excellent work from you guys. Thank you for your help!

Best regards,

Mengqi