Hi,

I need from someone that has applied the algorithm and understands it very well.

In this moment Im starting very small, at:

http://www.students.ncl.ac.uk/f.j.he...owexample.html

There is a small diagram of my actual process. It is the flow of a gas through a pipe. The black arrows are known velocities and the blue cells represent known pressures (in this case both are zero).

My solution procedure is this one:

1) First I assume a pressure distribution on the black cells as well as a velocity flow

2) Solve the momentum equations for the red arrows (horizontal and vertial) Six for the horizontal and 3 for the vertical velocities

3) With the new velocities calculate the pressure correction.

And here is where the problems start.

I don't know why, but when I solve the system of equations for the pressure corrections the results are unrealistic (negative pressures)

What would you do for the solution of this small system? What boundary conditions you would use? Is the system too small and therefore not well defined?

a system is not well defied for small sized just results are unrealistic. a negative pressure is not wrong, the equation give you the relative pressure field that you should correct it later. if you are using the staggered grid you should have 4 equation for the vertical solution. inflow is clear and just put a very low viscousity for the outflow cells as patankar mentiond it. it won't affect your solution. finally take a look ant 'patankar' or 'malalaskar' books they'll be good guid for you.