I have a mixture gas flow, and know the pressure of the flow outlet. However, the species mass fractions at the outlet are unknown. First, I define the BC of the outlet as pressure-outlet. but I find that I have to define the mass fractions. Then, I have to define the BC as outflow. But I think it is not so accurate.

Somebody help me. thanks.

sean

Check 6.8.1 "Inputs at Pressure Outlet Boundaries" in the Fluent manual. The species mass fraction in this panel is used for backflow only! good luck, Paul

Outflow BCs are supposed to be Neumman type BCs, that correspond to no normal gradients to the boundary surface (or fully-developed flow), but I am not quite sure this is really the case in Fluent

Outflow could sometimes give surprising solutions.

My opinion is to avoid as often as you can to use outflow BCs, you may definitively prefer pressure outlet, as it is pure and precise Dirichlet condition, and you really know what Fluent is doing!!

Good luck

Dear Paul, thanks for your help. I am still fonfused about the so called backflow. Need I input values for these fractions? Or just leave them like that? How can I determine their values?

sean

Dear Anthony Wachs,

I am so glad that you gave me so much advice. Would you please tell me how to define the mass fractions in the pressure-outlet panel. Mr. Paul said that they are for backflow, But I am still not clear how to make it done.

These values are only used if you get backflow, that is flow coning in to your domain from the outlet - set the values to something reasonable. Hopefully you will not have any backflow in your final solution and then these values have no effect.

I am very intrest in yout disscussion because I have a similar problem.I am modeling a smoke diffusion from the chimney to the surrounding.The space is very large in relate to the chimney.I think the outlet is outflow because I do not know the pressure of the boundary.But when I set the outflow boundary,it can not converge,beside,I get two different solutions with two boundary condition. How can I deal with it?thank you.

As it had been said, Outflow is homogeneous Neuman condition for pressure. That is to say that pressure is extrapolated from the domain to the BC.

This condition should only be used for developped flow.

For external flow you should use Pressure outlet condition with a pressure equal to atmospheric static pressure at stagnation point.

best regards

Thank you for your help.But I don't know where the stagnant point is,is it reasonable to set the boundary which is relatively far away from the chimney "atmospheric static pressure"?

Hi,

The pressure outlet b.c demands only the backflow values. In case your flow situation has backflow after the converged solution, then only the values matter.

Other way of doing this is to substitute the backflow mass fraction value to be equal to the previous iteration if backflow exist during iteration. This can be done using udf.

Dont go for OUTFLOW If your flow is not a fully developed one