periodic flows
 

can periodic flows be used to model angled vanes? I have a tube which contains 12 angled vanes which are used to create swirl. Can I model a 1/12 segment of the cylinder and set the segment planes as periodic boundaries. The fluent manual says that for periodic flows the geometry must be translationally periodic. I would describe the situation as rotationaly periodic, does this matter?

Re: periodic flows
 

Yes, you can model this easily in Fluent - Mesh your domain with two identical periodic faces and define them as walls initially. In Fluent first specify your rotational axis of the fluid domain (in the boundary condition panel) and then use the text interface to create the rotationally periodic interface with /grid/mofify-zone/make-periodic

Re: periodic flows
 

Thanks for that Jonas, it dosn't seem to be working but im not entirely sure that the face meshes are identical so I will try that first. Is this generally a better strategy than linking the face meshes and creating periodic boundaries in Gambit?

Re: periodic flows
 

The faces meshes must be identical and you should link them in Gambit. However, there used to be a bug in Gambit which broke things if you defined the faces as periodic already in Gambit - this might have been fixed now but defining them as walls has always worked for me so I'll stick with that.

If you have problems to get all nodes to match you can increase the matching tolerances slightly inside fluent with something like

/grid/modify-zone/set-tolerance

You should not need to increase it a lot - then something is wrong with your grid. Another reason for failing to make the faces periodic in fluent is if you have not set the rotational axis.

Re: periodic flows
 

I think that "Gambit-periodic bug" has been fixed a while ago as I used this feature successfully. At least since release 1.2 and later I hadn't any problems with it...

Re: periodic flows
 

thanks for the help so far, I am still having problems now when i check the grid i am getting messages like

WARNING: node on face thread has multiple shadows.

What does this mean?

Thanks in advance

Re: periodic flows
 

You get this warning if a thin wall (i.e. a two sided wall) hits on a periodic boundary. Then Fluent warns for multiple shadows.

Ignore the warning for this case.

Volker

Re: periodic flows
 

You should define the rotation axis of fluid in boundary condition if it not is Z.

Re: periodic flows
 

i have a similar geomtry set-up for a combustion chamber, What I would like to know and may also apply to phil, is that once i have a 60 degree segment of swirling flow with periodic boundaries can i use fluent to give me a particle track for the full 360 degrees to see the particle progress through the full chamber. Thankx

Re: periodic flows
 

In the new manual of fluent 6.0 chapter 5.6 p.5-41 I found s.th concerning duplicate shadows:

for coupled walls you can repair them by: grid/modify-zones/repair-duplicate-shadows

for periodic-bc: grid/modify-zones/repair-periodic

Re: periodic flows
 

Yeah it is possible

Display/views then set the number of periodic repeats you require 5 in your case.

Re: periodic flows
 

"NO" Displaying the 360 degree view will show the same particles track 6 times, not the actual particle track.

DC

periodic flows,repeating in time or geometry?
 

Hi all,

I am modeling a respiratory mask for adsorption of air pollutants,because the act of respiration is not continuous at the time I thaught that I should use periodic flow model in Fluent,But after studing the user guide I saw that apparently in this models the flows should be repeat with geometry of our system not with the time ,Is it right?

Shall i use periodic flows model or not?

Help me please.

Yes, the term "periodic" here refers to the geometrical periodicity, ie - if the actual geometry is such that the actual fluid domain in the simulation repeats itself.

For your case, you can of course model the flows that are periodic with time without using the periodic boundary conditions of course since your respiratory mask doesn't have repeating patterns of geometry. However such transient simulations are tricky to model you will need to do a lot of homework before venturing into it.

OJ

Hi,you mean that in spite of the expression "The geometry must be translationally periodic" in the 9.2 section of user giude about limitations of model,I should enable this model?

Isn't there another part in the fluent for periodic time flows?