CFD Online Discussion Forums

CFD Online Discussion Forums (
-   CFX (
-   -   Flow boundary conditions (

Turbulent April 19, 2010 08:22

Flow boundary conditions

I would like to know what are the preferred boundary condition to use in CFX in case of 2-D external flow past a body (e.g. cylinder, airfoil) in case I use a rectangular domain mesh.

The two possible options I know are:

1. Definition of the top and bottom of the domain as opening and relative pressure of 0. Here the default option for the flow direction is normal to boundary (which is not correct since the top and bottom boundaries are relatively far from the body so the flow suppose to be parallel to it).

2. Definition of the top and bottom as an inviscid wall - the problem here is that if the boundaries are not too far from the body the flow might be constrained

What do you recommended and why?


zandi April 19, 2010 16:03

could you specify the characteristics of your problem more clear?
as I know for a 2d problem you have to model it 3d with a thin thickness with a width of an element and side of the domain would be being symmetry.if i'm right
but for bottom and top i couldn't understand where is your inlet or direction of flow. Opening is used when the flow come in and go out and we donít know the direction.

ghorrocks April 19, 2010 16:25

I recommend option 3. Assuming the flow is from left to right, the LHS, top and bottom is a velocity inlet and the RHS is a pressure outlet. Note this suffers the blockage factor issue just like your option 2.

Your option 1 will have troubles converging and option 2 is workable and will give essentially the same result as my option 3.

If you must have the top and bottom close to the object so you can't use this approach as the blockage factor is too high the best option is to do an analytical inviscid solution of the flowfield and impose this flowfield as the inlet flow condition. Assuming the inviscid flow field is close to the viscous flow then this will allow you to move the top and bottom quite close with a much reduced blockage factor effect.

Turbulent April 20, 2010 04:15

Some clarification
Thanks for the help.

The problem I talk about is 2-D but the modeling in CFX is of course 3-D with one cell extrusion in the third dimension.

The flow is from left to right.

Thanks again!

All times are GMT -4. The time now is 13:53.