I'm attempting to solve a 2D, axi-symmetric, unsteady problem. I wan't to use mass-velocity-inlet boundary condition at a curved external boundary and I wan't the direction of the mass flux to be normal to the boundary and inside the domian. Since by default FLUENT takes the positive normal direction to a face as pointing outside the domain, I should give a negative mass flux, but the FLUENT doesnot accept negative mass fluxes. One method I have used to overcome this problem was to write udfs giving the direction vector for the mass flux at each point of the curved boundary. The udf uses the F_AREA command to calculate the components of the normal vector at each point, multiplies them with a minus sign and then returns them to FLUENT. When I use this method in FLUENT everything goes OK for 1 or 2 timesteps but then the direction reverses. I don't know why this is happening? I'll be glad to accept any help in this regard.

Sincerely,

Saad

1)without using mass-velocity-inlet, u better use velocity with vector direction.

2)If you know the x,y velocity components for the x,y coordinates of your inlet, then

Define-> Pofiles -> Write ... 3)for longer inlet domain, you may want to solve for a developed inlet domain and use that as an input of the original domain.. that will save ur computational time..

MNHK