[Zhouzhou]

Hello everyone,

I am dealing with the identification of vortex in simple shape using Q-criterion, but it does not work. As shown in attachment 1, there are two eddy obviously (Circle out with a box). But according to the Q-criterion, the determined area of the eddy are not exactly, as shown in attachment 2. Moreover, there are two additional area on the two sides of this shape, which are totally wrong, because they are not eddy region. So I want to know is there anything wrong when I apply the Q-criterion?
eddy1.JPG
eddy2.JPG

[FMDenaro]

do you have more infos about the way the criterion is applied in COMSOL?
Have you tried to compare with existing test published in literature

[Zhouzhou]

Originally Posted by FMDenaro

do you have more infos about the way the criterion is applied in COMSOL?
Have you tried to compare with existing test published in literature

Dear FMDenaro,

Thanks for your reply. I calculate Q in the COMSOL using the equations as below, and determine the eddy region by setting Q>0:

s11=ux
s22=vy
s12=0.5*(uy+vx)
s21=0.5*(vx+uy)
ss=s11*s22-s12*s21
omga12=0.5*(uy-vx)
omga21=0.5*(vx-uy)
omga=-o12*o21
Q=0.5*(omga*omga-ss*ss)

Based on this method, the red region in attachment 2 can be separated. But it seems the determined region is not the correct eddy/vortex region which I want. I have not tried to compare with existing test published, but the eddy/vortex region what I understand should be the region circled in the figure below. So how can I identify to get this?

[FMDenaro]

The Q-criterion uses also the pressure constaint ....
have a look here
https://www.researchgate.net/publica..._JFM_285_69-94

[Zhouzhou]

Originally Posted by FMDenaro

The Q-criterion uses also the pressure constaint ....
have a look here
https://www.researchgate.net/publica..._JFM_285_69-94

I looked the pressure constraint before, but I really don't know how to use this additional constraint. Moreover, the identified region using the only constraint Q>0 has already mismatched the eddy/vortex region, so is it helpful to add anther constraint?

[alain.wokusch]

a pretty good solution is provided in the following post. Look for the answer which mentions the 1995 Jeong & Hussain paper. It exaclty shows what needs to be implemented as as a Component Variable for postprocessing of vortex cores in Comsol according the Q criterion:


https://www.researchgate.net/post/Ho...n_Ansys_Fluent


(question is about Ansys but this one answer is also valid for Comsol)

[Kell]

Originally Posted by Zhouzhou

Hello everyone,

I am dealing with the identification of vortex in simple shape using Q-criterion, but it does not work. As shown in attachment 1, there are two eddy obviously (Circle out with a box). But according to the Q-criterion, the determined area of the eddy are not exactly, as shown in attachment 2. Moreover, there are two additional area on the two sides of this shape, which are totally wrong, because they are not eddy region. So I want to know is there anything wrong when I apply the Q-criterion?
Attachment 51452
Attachment 51453

Hey, were you able to get the required result? If yes, what did you change? I am also working with COMSOL and need to apply this criterion but am not getting the required result.

[Moreza7]

Originally Posted by Kell

Hey, were you able to get the required result? If yes, what did you change? I am also working with COMSOL and need to apply this criterion but am not getting the required result.

You can export results to a VTK file and then plot QCriterion using Paraview, Tecplot, etc.