Greetings,

I am a computer scientist (not experient on CFD field) and I'm facing a problem that I believe to be related to velocity-pressure coupling.

I am writing a simple code and I'm trying to validate it against a simple laminar and incompressible air flow in a pipe (to begin with). The problem is I'm facing mass conservation problems.

I am using SIMPLE and was previously facing the odd-even decoupling problem. After I solved it (by changing from colocated to staggered grid) I started having this severe mass conservation problem and I don't know what to look for or where to begin with.

The equations in the discretized form as well as the steps I take in the algorithm are presented in this file: http://www.panic-lab.rutgers.edu/~diego/cfd/question.ps

Could you please take a quick look and see if I'm missing something?

Since I have limited knoledge in the area, the more I check these equations and their discretization, the more I think it's correct, though the problem is still there!

ANY comment would be helpfull,

Thanks in advance!!

I did not understand, why do you use the pressure in the (1) step? In that way the (2) is not correct. The outlet BC looks doubtful as well. It will be good to look in a CFD hand-book and find out the appropriate method.

Regards

I though I was supposed to use pressure values from the previous iteration for the momentum equation (1) and then correct the velocity (4) using the new calculated pressure.

Is your suggestion something like this? http://www.panic-lab.rutgers.edu/~di...d/question2.ps

I made the modifications and it didn't make much of a change ... So I'm guessing it might be the BC's then.

I also tried, for the BC's in the outlet, to simply copy the value from the free cell neighboring the outlet cell (for both pressure and velocity) but it didn't work as well.

If I watch the velocity profile as the simulation runs, what I see is that the (expected) parabolic profile advances through the pipe but when it gets close to the outlet, the pressure profile shifts all of a sudden and pushes the velocity parabolic profile backwards again ... It keeps on oscilating like this until it "converges" to some no sense velocity profile.

Any suggestion for the BC's?

Thanks for the help!

This is a snapshot of the velocity profile close to convergence ... The representation is a little different from the usual. The velocity values follow the color scale shown underneath the view of the pipe.

http://www.panic-lab.rutgers.edu/~di...el-profile.jpg

There was supposed to be a "red stripe" going all the way to the outlet area (right). I hope it helps to explain my problem.

Thanks.