CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   OpenFOAM Running, Solving & CFD (http://www.cfd-online.com/Forums/openfoam-solving/)
-   -   Actuator disk (http://www.cfd-online.com/Forums/openfoam-solving/57841-actuator-disk.html)

saasebo February 26, 2009 10:20

Dear Forum I am trying to s
 
Dear Forum

I am trying to simulate flow around a ship propeller nozzle profile, and nozzle drag and lift are my objectives. A small axisymmetric wedge is used. My problem is assign correct boundary conditions at the actuator disk (AD) replacing the propeller.

The a priori information is the pressure jump (profile) over the AD. I have seen that it exists a fan bc in OF-1.5, but i can not find examples of how to use it.

Another option is to use two patches for the AD, and assign a velocity at the pressure side. But how should I then make the turbulent variables periodic over the disk?

Picture: http://steinar.aasebo.googlepages.com/

Thanks,
Steinar

antonio_ing March 2, 2009 16:56

why don't you try the porous b
 
why don't you try the porous bc condition? it should just gives you the pressure jump. There is a tutorial about porous media in the tutorial folder.

hope this helps

saasebo March 3, 2009 09:51

Well, is it then possible to g
 
Well, is it then possible to get a pressure rise over the porous wall?

Has nobody used the fan boundary condition with success?

Best Regards,
Steinar

arash March 5, 2009 11:29

have you tried applying veloci
 
have you tried applying velocity outlet on the inlet of the AD and velocity inlet on the outlet of the AD. You may use these conditions in case you have enough data for the velocity distribution on the inlet/oulet of the propeller. I'm not sure it will work for you or not.

Keep in touch and let us know your strategy for solving your problem.

Best regards,
Arash

saasebo March 6, 2009 05:29

It should be possible to find
 
It should be possible to find a velocity distribution such that the pressure jump distribution becomes correct.

But how could I then make the turbulence variables k and epsilon cyclic over the disk? If this is possible without to much programming, it may be a good approach.

Thanks a lot for the interest, I'll keep you informed.

Best Regards,
Steinar


All times are GMT -4. The time now is 11:23.