CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Dirichlet and Nuemann (http://www.cfd-online.com/Forums/fluent/101177-dirichlet-nuemann.html)

 Zacc April 28, 2012 14:26

Dirichlet and Nuemann

Hi everyone,

I want to simulate turbulence over a flat plate. Now I have met a problem about the boundary conditions. I would like to apply Dirichlet Boundary conditions, velocity in the direction y and Nuemann Boundary conditions, stress, for one boundary at the same time. The most important is " at the same time for one boundary"

Could you please tell me how can I do this in Fluent? Thank you so much!

 LuckyTran April 29, 2012 11:27

Quote:
 Originally Posted by Zacc (Post 358210) Hi everyone, I want to simulate turbulence over a flat plate. Now I have met a problem about the boundary conditions. I would like to apply Dirichlet Boundary conditions, velocity in the direction y and Nuemann Boundary conditions, stress, for one boundary at the same time. The most important is " at the same time for one boundary" Could you please tell me how can I do this in Fluent? Thank you so much!
Which direction is the y direction? Is that plate-normal or plate-parallel?

The plate-normal velocity is 0 by kinematic constraints, flow can't flow through the plate.
The plate-parallel velocity at the wall is also 0 by no-slip conditions, here is where wall functions jump in.

 Zacc April 29, 2012 13:28

Quote:
 Originally Posted by LuckyTran (Post 358329) Which direction is the y direction? Is that plate-normal or plate-parallel? The plate-normal velocity is 0 by kinematic constraints, flow can't flow through the plate. The plate-parallel velocity at the wall is also 0 by no-slip conditions, here is where wall functions jump in.
Thank you! Actually I didn't make it clear. The boundary conditions are applied for the farfield where there is a suction in the direction y (normal to the plate)

 LuckyTran April 29, 2012 16:27

Quote:
 Originally Posted by Zacc (Post 358348) Thank you! Actually I didn't make it clear. The boundary conditions are applied for the farfield where there is a suction in the direction y (normal to the plate)
I'm having a hard time imagining this boundary. Can you clarify?

What is the extent of this boundary? And you intend to specify the velocity across this boundary + shear? Since flow is in y, you want to specify velocity in (x,z) and shear in (x,z). Is the velocity uniform? What about shear?
I am also unclear on another point: Once the velocity distribution is specified, the shear stress is automatically known. How can you specify it further?

Providing a sketch would help.

 Zacc April 29, 2012 23:36

1 Attachment(s)
Quote:
 Originally Posted by LuckyTran (Post 358371) I'm having a hard time imagining this boundary. Can you clarify? What is the extent of this boundary? And you intend to specify the velocity across this boundary + shear? Since flow is in y, you want to specify velocity in (x,z) and shear in (x,z). Is the velocity uniform? What about shear? I am also unclear on another point: Once the velocity distribution is specified, the shear stress is automatically known. How can you specify it further? Providing a sketch would help.
Thank you! First, what I want to do is to simulate the turbulent separation bubble. So I have to do my suction-Velocity in the direction y which is normal to the flow as showed in the image. But we have met a problem of rotation. So the second thing is to simulate the situation irrotational by applying the shear in the farfield. Because with the condition of irrotation, we have du/dy=dv/dx, knowing that v is the velocity of suction.

 All times are GMT -4. The time now is 19:57.