How to set domain translation and rotation in CFX

ricardo.halfeld · September 2, 2009, 23:25

Hi everybody,

I'm currently trying to make a 2D simulation of a wing profile rotating about an axis somewhat far from it AND around it's own axis. My current approach is to have a static square mesh with a hole in it, so in this hole goes another circular rotating mesh which also has an excentric hole where the wing profile mesh goes. All meshes are connected with GGI Interfaces.

The rotation of the second mesh can be easily defined by clicking domain containing it's parts and setting the domain motion to rotation. But what about the smaller domain containing the wing profile? Using the same setting I'd get it to move around it's primary axis of rotation (the axis of rotation of the second mesh), but not around it's own axis. Any ideas?

I have an Expression that defines the X and Y coordinates of the center of the smaller domain and another one with it's angle of rotation, both funtions of time, but I can't find out how to link these expressions to the domain motions.

ghorrocks · September 2, 2009, 23:42

Some options:
1) Model the airfoil mesh using a normal deforming grid (not a rotating domain) and use a CEL expression to move the mesh how you wish. You then connect everything up with GGIs as normal.
2) Just have a static background mesh and use the new V12 immersed solid functionality. This would be quite easy to do, but it does mean that you will not be able to have a nice boundary layer grid on the airfoil. If a good boundary layer grid is important then this approach is not recommended.

ricardo.halfeld · September 4, 2009, 10:06

Glenn, thanks for your input. The boundary layer is indeed important and making a refined mesh throughout the blade excursion path would be "too expensive". And the guy orienting my project has had bad experiences with deforming mesh in forced oscilations.

I can't believe there's no way to define the mesh position as a function of time as well as rotation with CEL. I believe this is the simplest way to run this simulation. Doesn't anyone have a clue on how to do it?

ghorrocks · September 5, 2009, 03:24

OK, if you need a good boundary layer mesh then you only have one choice - being a deforming mesh. Properly implemented it should work fine.

Quote:

I can't believe there's no way to define the mesh position as a function of time as well as rotation with CEL.

Yes, you can do it - using deforming mesh.

If you look in the tt.txt file and/or the RULES file in <CFXROOT>/etc you will see that translating meshes is in there. However as it is not documented (not even as a beta feature) so it probably either is only in the CEL and not the solver or doesn't work properly. I don't know, I have never used it. You will need to talk to a support person to get a definite answer on this.

Even if you can get a combined rotating and translating mesh thing working then I guarantee it does not have the full equations of motion in there, meaning some accelerations of the frame of reference are not correctly modelled. So to be sure I recommend you go to a deforming mesh as that has all this stuff covered.

ricardo.halfeld · September 6, 2009, 09:40

Quote:

Originally Posted by ghorrocks

You will need to talk to a support person to get a definite answer on this.

OK. I will.

I never used deforming mesh, so I will try some tutorials. Thanks for all the support!

ghorrocks · September 7, 2009, 08:27

It would be great if you could post the response you get from CFX support, then we can all learn a bit.

ricardo.halfeld · September 8, 2009, 12:40

Quote:

Originally Posted by ghorrocks

It would be great if you could post the response you get from CFX support, then we can all learn a bit.

I will. I tried getting in contact through Ansys' website. I had to fill some forms and they should contact me soon... that was 3 days ago...

Is there a support email address?

ckleanth · September 8, 2009, 19:22

Quote:

Originally Posted by ricardo.halfeld

I will. I tried getting in contact through Ansys' website. I had to fill some forms and they should contact me soon... that was 3 days ago...

Is there a support email address?

well there is but don't you have an ansys sales contact through whoever sold the software to you/company/university?

bear in mind if your oranisation does not have a support contract.. well you will wait for quite a long time...

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Translation and rotation for a moving wall	Stéphane Champmartin	FLUENT	1	November 28, 2006 11:57
Relief Plots in CFX	Michael	CFX	15	June 30, 2006 09:01
what the result is negatif pressure at inlet	chong chee nan	FLUENT	0	December 29, 2001 06:13

September 2, 2009, 23:42		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,871 Rep Power: 144	Some options: 1) Model the airfoil mesh using a normal deforming grid (not a rotating domain) and use a CEL expression to move the mesh how you wish. You then connect everything up with GGIs as normal. 2) Just have a static background mesh and use the new V12 immersed solid functionality. This would be quite easy to do, but it does mean that you will not be able to have a nice boundary layer grid on the airfoil. If a good boundary layer grid is important then this approach is not recommended.

September 4, 2009, 10:06		#3
ricardo.halfeld New Member Ricardo Halfeld Join Date: Sep 2009 Posts: 4 Rep Power: 17	Glenn, thanks for your input. The boundary layer is indeed important and making a refined mesh throughout the blade excursion path would be "too expensive". And the guy orienting my project has had bad experiences with deforming mesh in forced oscilations. I can't believe there's no way to define the mesh position as a function of time as well as rotation with CEL. I believe this is the simplest way to run this simulation. Doesn't anyone have a clue on how to do it?

September 7, 2009, 08:27		#6
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,871 Rep Power: 144	It would be great if you could post the response you get from CFX support, then we can all learn a bit.