Hi all,

I am trying to simulate the effect of crosswind on a open combustion flame (vertical flame). The wind entrances the domain by the left side of the domain, so a specifi velocity inlet is assumed as boundary condition in this left side.

However, I don't really know what boundary condition to use for the right side (or wind outlet): I think I have 3 options:

1) Pressure outlet (e.g., pressure known)

2) Outflow (e.g., pressure and velocity unknown?)

3) Negative velocity inlet?

Notice that my domain is only 0.3 m width, and the vertical flame is burning inside that domain.

Any advise would be helpful. Thanks!!

David,

It sounds like you have a canned code and are limited by the BC menu of the code. If the code is a compressible code, you're looking for a subsonic, nonreflecting outflow BC for gas mixtures. If you have any control, there is one characteristic coming into the domain at the outflow. That quantity must be imposed by you. The boundary permits N_dim + 1 + N_species BCs. For the other BCs,

n dot tau dot t1 = 0

n dot tau dot t2 = 0

n dot q = 0

n dot V_i = 0

or

n dot [n dot (Grad tau)] dot t1 = 0

n dot [n dot (Grad tau)] dot t2 = 0

n dot [n dot (Grad q)] = 0

n dot [n dot (Grad V_i)] = 0

where

q = heat flux vector

tau = viscous stress tensor

V_i = diffusion velocity

n = boundary normal

t1, t2 = boundary tangents

None of your 3 choices are remotely close to what I have suggested.

You could try a subsonic hard inflow. Specify u,v,w,T, and Y_i. Use the last BC to either impose a condition on

(n dot tau dot n = 0) OR (n dot [n dot (Grad tau)] dot n = 0)

or try one of various approaches to specify density (or pressure).

Thanks for the information!