I'm doing some transient analysis of pipes. If the gas escape from the inlet boundary,what will happen? Can Star-CD deal with this kind problem?

If you have outflow you can use inlet or pressure boundary conditions.

I've used inlet boundary conditions.It seems that Star-cd hasn't processed the problem.

Hi Miles, since upwind scheme has been used, inlet boundary as i see could not handle reverse flow. why not extrude the inlet for a distance.

If you mean that you INTEND to use a velocity inlet as outlet boundary (i.e. apply a negative inlet flow - like a suction point) it should be alright. That is if you put a pressure boundary where the flow should enter, to be able to specify turbulent quantities.

It seems that Star-CD cannot deal with this problem.

It seems to me that you are not able to describe your problem in a way that we know what you are doing and what you intend to do.

I'm sorry. I'm trying to simulate the gas flow through pipes. And the boundary conditions is that the flow of the inlet boundary can be positive and negative.

You have about 3 possibilities to do this.

1.) use a user defined subroutine and describe the velocity as a function of time

2) use a table with the values of the velocity

3) use the transient calculation with different load steps. For each load step you can select you boundary values and ramp between them.

If you have an analytical function for the vel vs. time the subroutine should be the most straightforward way to go. Otherwise use the second or third way.

I'll use the first method. Thank you very much.