[poliwang]

Hi, friends
I am using non-reflecting boundary condition at the turbulent subsonic outflow. As you know, if all the normal component u is positive (flow out), we need only P_infinity. However, locally u can be negative. If so, what kind of additional information can I use at the boundary?

Many thanks for any help.

[FMDenaro]

Originally Posted by poliwang

Hi, friends
I am using non-reflecting boundary condition at the turbulent subsonic outflow. As you know, if all the normal component u is positive (flow out), we need only P_infinity. However, locally u can be negative. If so, what kind of additional information can I use at the boundary?

Many thanks for any help.

if the flow is subsonic (u+a) is always positive, (u-a) is always negative therefore you shoud consider only the convective wave. If you have reverse flow at the outflow you can set Neumann conditions

[poliwang]

Thanks for the useful suggestions. Could you please show more details? For example I need to update u,v,w, rho, T and p. Which variables (all of them) use the Neumann conditions?

Originally Posted by FMDenaro

if the flow is subsonic (u+a) is always positive, (u-a) is always negative therefore you shoud consider only the convective wave. If you have reverse flow at the outflow you can set Neumann conditions

[FMDenaro]

Originally Posted by poliwang

Thanks for the useful suggestions. Could you please show more details? For example I need to update u,v,w, rho, T and p. Which variables (all of them) use the Neumann conditions?

have you already read the paper of Lele & Poinsot on JCP?

[poliwang]

Yes, I have read the Poinsot&Lele 1992 paper, where there is no such discussion. Do you mean some different papers? If so, please let me know which they are. Many thanks.

Originally Posted by FMDenaro

have you already read the paper of Lele & Poinsot on JCP?

[FMDenaro]

Originally Posted by poliwang

Yes, I have read the Poinsot&Lele 1992 paper, where there is no such discussion. Do you mean some different papers? If so, please let me know which they are. Many thanks.

that paper is ok, more recent paper appeared, you can find those by means of google...

I would try first setting Neumann condition only for the velocity