CFD Online Discussion Forums (https://www.cfd-online.com/Forums/)
-   CFX (https://www.cfd-online.com/Forums/cfx/)

 cuteapathy March 25, 2012 06:15

about Taylor-Green Vortex problem in CFX

Hi everyone,

I am trying to compare cfx and openfoam by solving taylor-green vortex problem on a cube in both of them. I applied the same initial conditions in both of them. For boundary conditions, in openfoam, I applied both of pressure and velocity boundary conditions on my boundaries. In cfx, I applied pressure on three of boundaries and velocity on the other three ones. But, I got totally different results and figures from them. I am sure that I got correct results and figures in openfoam, but I do not know why it is wrong in cfx. I am wondering whether anyone tried this problem before and could kindly give me some advice about how to fix it. Thanks for any response.

 ghorrocks March 25, 2012 06:32

Why does the Taylor-Green vortex need flow boundary conditions at all? It is the decay of a vortex so why not put the outer boundary far away and simply use walls?

 cuteapathy March 25, 2012 09:07

Dear Glenn,

My advisor asks me to apply both of bondary conditions on six faces. I did it in openfoam and it worked. In cfx, do you mean I could use three walls instead of three outlets and remain to apply velocity on the other three inlets? Thank you for your response.

 ghorrocks March 25, 2012 18:05

Do you know how many times on the forum the reply to "why are you doing something bizarre?" is that my supervisor told me to do so......

I am no expert on the Taylor-Green vortex, but from what I have just read it is simply the decay of a vortex. There are no flows in or out of the domain. To accurately model this you will need your boundaries far enough away to not affect the dissipation. So the domain which makes sense to me is one with slip walls quite a way away from the central vortex. I cannot see how inlet or outlet boundaries are required here.

Also note there is an analytical solution for this flow so you have precise answers to compare against.

If you insist on using inlet/outlet/opening boundaries can you post an image of what you are modelling as in that case I am clearly missing something.

 cuteapathy March 26, 2012 06:44

Dear Glenn,

When I check some online information again, I think you are right. I am gonna try your suggestions. Thank you for your detailed advice. I appreciate your help.

 All times are GMT -4. The time now is 23:33.