[engineerm]

Hi friends,

I have a project about 2D translational rotor-stator periodic analysis on ANSYS Fluent. I have set the interfaces without problem. When I tried to set periodicity to stator bottom-top edges and rotor bottom-top edges, I have wrote the codes in the console, but the periodicity failed with the message "number of elements are not equal between the zones" as follows:

/mesh/modify-zones/make-periodic
Periodic zone [()] 9
Shadow zone [()] 10
Rotational periodic? (if no, translational) [yes] no
Create periodic zones? [yes] yes
Auto detect translation vector? [yes] yes
Error: Failed to make zones periodic.
Error Object: #f

Warning: Number of elements does not match between zone 9 and 10.

I have changed the default matching tolerance from 0.05 to 0.5 using the code " /mesh/modify-zones/mathing-tolerance ", but it hasn't worked unfortunately.

I will be very glad to have helps from you about my problem. You could see the 2D geometry and fluent fail messages in the attached files.

My Best Regards

[sufjanst]

Like I wrote in the other thread, check your flow axis in Cell Zone Conditions

[engineerm]

Originally Posted by sufjanst

Like I wrote in the other thread, check your flow axis in Cell Zone Conditions

hi sufjants, I don't think so about the problem occuring due to the axis. Because my former geometry was very like the current one and I hadn't faced any problem about the periodicity although I had used the same axis.

Thanks for your feedback. If you have further solutions, I will be very thankful for you.

[LuckyTran]

Don't use auto-detect but manually input the translation vector.

You need a conformal mesh on the periodic pair (means the mesh must be exactly the same).


It says you have wrong number of faces so leads me to think you don't have a conformal mesh. You can change tolerance all you want, but if there are different number of faces, then it's obvious....

[engineerm]

Originally Posted by LuckyTran

Don't use auto-detect but manually input the translation vector.

You need a conformal mesh on the periodic pair (means the mesh must be exactly the same).


It says you have wrong number of faces so leads me to think you don't have a conformal mesh. You can change tolerance all you want, but if there are different number of faces, then it's obvious....

thank you very much. I'll try it and hope it'll work!

[praveencyr]

Have you matched the mesh for those two edges?

Use match control option during meshing.

[engineerm]

Originally Posted by praveencyr

Have you matched the mesh for those two edges?

Use match control option during meshing.

No, I do not match the mesh for the edges. Are they going to have the same number of elements after matching mesh? If yes, sounds it could work. I will try it. thanks!

[praveencyr]

Originally Posted by engineerm

No, I do not match the mesh for the edges. Are they going to have the same number of elements after matching mesh? If yes, sounds it could work. I will try it. thanks!

Yes, the faces will be meshed with similar number of elements and with same pattern.
So, it will match without having higher tolerance value!

[engineerm]

Originally Posted by LuckyTran

Don't use auto-detect but manually input the translation vector.

You need a conformal mesh on the periodic pair (means the mesh must be exactly the same).


It says you have wrong number of faces so leads me to think you don't have a conformal mesh. You can change tolerance all you want, but if there are different number of faces, then it's obvious....

I did not enter "yes" for auto detect as you offered and entered 0,1513266 m which is the the distance between the edges on y direction, also tried other directions but still does not work. I couldn't overcome it.

[engineerm]

Originally Posted by praveencyr

Yes, the faces will be meshed with similar number of elements and with same pattern.
So, it will match without having higher tolerance value!

It sounds very good!! I hope it provides the periodicty!

[praveencyr]

Originally Posted by engineerm

It sounds very good!! I hope it provides the periodicty!

Yes, I'm doing similar kind of project but with rotational periodic BC. So, this will work for sure!

[engineerm]

Originally Posted by praveencyr

Yes, I'm doing similar kind of project but with rotational periodic BC. So, this will work for sure!

hi dear praveencyr, I have tried it but I have faced meshing problem as you can view from the picture as "2". I have searched the literature about match control, it says the geometry of the edges must be the same as you can see the requirements of match control in the other picture as "1" I attached or from the link below.

What's wrong with it?

https://www.sharcnet.ca/Software/Ans...Face_Mesh.html

[engineerm]

Originally Posted by engineerm

hi dear praveencyr, I have tried it but I have faced meshing problem as you can view from the picture as "2". I have searched the literature about match control, it says the geometry of the edges must be the same as you can see the requirements of match control in the other picture as "1" I attached or from the link below.

What's wrong with it?

https://www.sharcnet.ca/Software/Ans...Face_Mesh.html

hi dear friends,

I have solved my problem and overcame the periodicity problem. It was caused due to the creating names for the edges in meshing. I had created the periodic names for the all length of the edges which was incorrect as seen in the photo "incorrect name selection". I have changed the name selections making blade profiles as wall, you can see in the photo " correct name selection". Aften then, I was able to see the velocity vectors in the results.

Thanks your helps

I just demand some literature articles about 2D or 3D turbomachinary axial water or air flow analysis using Fluent. If you have some articles about it, it will be very helpful for me to decide blade chord length, inlet velocity/pressure, turbulent viscosity values selected on Fluent.

best regards