CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   surface tension model (http://www.cfd-online.com/Forums/cfx/106684-surface-tension-model.html)

sjtusyc September 5, 2012 09:05

surface tension model
 
These days I do a multi-phase simulation using free surface model and homogeneous model.

At first i do not turn on the surface tension model, it converges badly,and it converges good when i take the surface tension model into consideration.

In my simulation,the water and gas interface is well0defined.

I am confused why the surface tension force has such a significant role in this simulation.

ghorrocks September 5, 2012 18:39

Usually the surface tension makes convergence harder, so your application is unusual.

Can you describe what you are modelling, what fluids and what geometry?

sjtusyc September 6, 2012 02:36

Quote:

Originally Posted by ghorrocks (Post 380447)
Usually the surface tension makes convergence harder, so your application is unusual.

Can you describe what you are modelling, what fluids and what geometry?

I wanna to simulate the internal flow of the pressure-swirl injector with fluctuation pressure boundary.Near the exit of the injector,air flows in as the result of the low pressure caused by the tangential motion of water.So the air and water are stratified,the interface is well-defined.

I wonder why should we take the surface tension force into consideration.
What will it influences?

ghorrocks September 6, 2012 06:09

What type of injector? How big? Is it producing a spray of droplets or some other flow?

sjtusyc September 6, 2012 20:52

Quote:

Originally Posted by ghorrocks (Post 380520)
What type of injector? How big? Is it producing a spray of droplets or some other flow?


Pressure-swirl injector.Yes,it is intended to produce spray of droplets finally.

ghorrocks September 6, 2012 20:54

Do you want to simulate the spray break up process directly, or use empirical models (eg LISA)?

sjtusyc September 6, 2012 20:56

Quote:

Originally Posted by ghorrocks (Post 380520)
What type of injector? How big? Is it producing a spray of droplets or some other flow?

I am doing the internal simulation at present to get the transient massflow and spray angle. Then do the spray simulation,using the LISA model.

ghorrocks September 6, 2012 21:09

I do not know the details of your model so cannot answer your original question.

However surface tension is critical to spray formation. The droplet formation process is driven by surface tension. So your comment about why you need surface tension is puzzling.

sjtusyc September 6, 2012 21:22

1 Attachment(s)
Quote:

Originally Posted by ghorrocks (Post 380658)
I do not know the details of your model so cannot answer your original question.

However surface tension is critical to spray formation. The droplet formation process is driven by surface tension. So your comment about why you need surface tension is puzzling.

I am not clear enough.
Now ,i am doing the INTERNAL flow simulation,like a pipie, not the spray process.The process before spray in the injector.

ghorrocks September 6, 2012 21:30

If this is an internal flow then isn't it single phase? Where does the gas/liquid mix come from?

sjtusyc September 6, 2012 21:38

1 Attachment(s)
Quote:

Originally Posted by ghorrocks (Post 380660)
If this is an internal flow then isn't it single phase? Where does the gas/liquid mix come from?

Hi,thanks for you patience.
It is a pressure-swirl injector, SWIRL means the liquid phase has a tangential velocity after it flows through swirl part (as the picture attached)of the injector.
According to the fluid mechanics,the fluid will have low static pressure having tangential velocity. So, in my case, the pressure of the liquid at the exit is lower than the atmospheric air,causing the air flow in the injector,and that helps the atomization process

The the mechanism of the pressure-swirl injector.

sjtusyc September 6, 2012 21:39

I wish am clear enough.
Thank you again.

ghorrocks September 6, 2012 21:48

I see. It is hard to know whether surface tension will be important in this or not. If you are close to the injector and before the surface tension has started to form the wobbles which grow and break the sheet up then you might not need it.

So if unsure then I would include it.

Be aware that surface tension models require very good mesh quality and much smaller time steps.

sjtusyc September 6, 2012 21:57

Thanks sincerely.
CFD is not easy.

ghorrocks September 6, 2012 22:51

No, it is not easy. That is why it is interesting.

sjtusyc September 7, 2012 02:02

Yes.
But,sometimes it drive me crazy when i can not figure it out.
Sometimes a small change will lead to much difference and i can not understand it.


All times are GMT -4. The time now is 22:29.