CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Why in invicid model it has vortex (http://www.cfd-online.com/Forums/fluent/98679-why-invicid-model-has-vortex.html)

Yannian March 16, 2012 05:45

Why in invicid model it has vortex
 
hi,
when I do a invicid model in which there are pressure jump(fan model), it generate vortex on the wall. Can you tell me the reason?Thank you!

Amir March 16, 2012 06:18

Hi,

Your assumption is not right. Each inviscid flow is not necessarily irrotational!!! (e.g. rigid body rotation). Nevertheless, ensure that you've achieved converged solution.

Bests,

Yannian March 16, 2012 06:35

Quote:

Originally Posted by Amir (Post 349805)
Hi,

Your assumption is not right. Each inviscid flow is not necessarily irrotational!!! (e.g. rigid body rotation). Nevertheless, ensure that you've achieved converged solution.

Bests,

Hi Amir,
The defenition of vorticity in my thread is the curl of velocity. If it is invicid,du/dy,du/dz,dv/dx ... are all zero, so the curl of velocity is naught. Please comment.

Amir March 16, 2012 06:50

Quote:

Originally Posted by Yannian (Post 349813)
Hi Amir,
The defenition of vorticity in my thread is the curl of velocity. If it is invicid,du/dy,du/dz,dv/dx ... are all zero, so the curl of velocity is naught. Please comment.

No!
The definition of inviscid flow is that the diffusion term of momentum equation can be neglected. As I said, in some cases like rigid body rotation this condition is satisfied but obviously the vorticity is not zero! i.e. :
Any irrotational flow is inviscid but not all the inviscid flows are irrotational.
I just wanted to clear this physical point for you but as a numerical point of view, when you use the inviscid model, you've ignored the elliptic term of momentum equation and it may cause some instabilities; consequently, you need more iterations for convergence. But as a tip, you can solve your case with full navier stokes equation and then use the converged results as your initial guess for inviscid flow. (This procedure is not common but can be used just for damping instabilities)

Bests,


All times are GMT -4. The time now is 09:18.