[mfim88]

Hi,

1. Can anyone tell me when to use the option "from neighboring cell" and when not to? assuming we don't know initially whether backflow will occurs or not.

2. It is normal to have backflow when choosing "from neighboring cell" option and no backflow occurs when choosing "normal to boundary" option.

p/s: i am doing fluent simulation on a cyclone separator. (high swirling flow).

[LuckyTran]

If the backflow direction is known, then you should specify it explicitly. If not known then there are two approaches:

The default is normal to the boundary where any back-flow will have a direction that is perpendicular/normal to the outlet boundary.

I am not sure of the specific calculation procedure, but when From Neighboring Cell is selected, the direction of the back-flow will be calculated based on the nearby cells. This means that the back-flow can enter at an angle with respect to the outlet boundary (it is not fixed to normal angles) but flow will be reversed in the sense of the local flow-field. This is useful when flow exiting the boundary is not normal to the boundary because of swirl, or you are trying to simulate compressor/fan discharge, etc.

[mfim88]

ok thanks LuckyTran for the insight.

Yes, my case have high swirling at the outlet, thus, from what i understand, "from neighboring cell" will give more accurate results.

[nedanikou]

Hi
Would you please tell me how I can implement Neighboring Cell B.C. in code?
I wrote the outlet B.C. (pressure outlet B.C.- Neighboring Cell method) as follow:
Uxi(j,nz,i)=Uxi(j,nz-1,i)
Ueta(j,nz,i)=Ueta(j,nz-1,i)
Uzeta(j,nz,i)=Uzeta(j,nz-1,i)
i=1,nx
j=1,ny
k=1,nz