[azoppi]

I'm new here. Hi to everybody.
Here's my question.

I try to simulate with starccm_double_precision the configuration in the attachments file. [7 clark-y airfoils]

As you can see from the images i have a background mesh, with 7 overset meshes. [All 3D] The grids are similar as the tutorial said.

Each airfoil is free to rotate only in z-axix under the action of the only Fluid Dynamics Forces.
Gravity option enabled.

My first try was to start simulate with stationary airfoils in the beginning and let them free to rotate after 0.1 seconds [release time] with TimeStep = 0.001 s
Everything went ok.

Now I have to link all the airfoils each other: I want same angle of rotation for every airfoil.
I can't find any option for "rigid link" so i try:
Body Coupling --> Linear Spring with Spring Constant = 1.0E20 N/m and Relaxation Lenght = 0.89 m [0.89 m is the distance between airfoils center of rotation]
So i link airfoil1 with airfoil2, airfoil2 with airfoil3, ... ... , airfoil6 with airfoil7.
I try to enable the ramp time [release time 0.1 s -- ramp time 5.0 s]
I try also with TimeStep = 0.0001 s

At the end of the Release Time, after 2 or 3 iterations, the simulation stops running because an overlapping grid problem.

Could anyone help me to costrain the airfoils to rotate with the same angle ?


Many thanks for considering my request. Andre.

[rwryne]

Are you using seperate "boxes" for each airfoil for your overset mesh? If so, I do not believe these are allowed to overlap at this point. I have been told by someone at cd-adapco that this is a feature they are working on (perhaps for v8.04?)

You might try making the overset boxes for each airfoil smaller so that they have less of a chance of overlapping.

[azoppi]

Thanx for answer.

I know that i can't overlap, but it isn't a matter of mesh size as you suggest.

The image below shows what happens at the first (or second) iteration immediatly after the release time (when the 7 airfoils are free to rotate).
Like a blow up, and only for some airfoils.

It depends (i think) from the setting of "Coupling Spring"

I thought that a spring with infinity stifness could be seen like a "rigid link".
I can't explain the final configuration of the airfoils.

I only want to know if i have to study the problem in another way or if it's possible to connect the airfoils each other (same rotation for all).

[ryancoe]

Your approach with the spring sounds smart. When I first read this I thought maybe you could set the airfoils as a single rigid-body, but I don't think this would allow them to rotate about different centers.

Have you tried different configurations of linking the airfoils? Also, I would try a relatively small spring stiffness and see what that might do.

Good Luck!

[azoppi]

They can't be a single body. Technically it's a louver, so every airfoil must have a different center of rotation.

Yes, I try also with low stiffness, relaxation length = 0, disabled repulsive forces, etc etc...even if it's thinking about nonsense, but only to try to know the behaviour of the "coupling".

Anyway: distance between centers of rotations is 0.89 so i put
- relaxation length = 0.89
- "infinity" rigidity
and this is a rigid link, really.

The overlapped configuration really surprised me. I can't explain it fisically....

[rwryne]

Originally Posted by anderzop

Thanx for answer.

I know that i can't overlap, but it isn't a matter of mesh size as you suggest.

The image below shows what happens at the first (or second) iteration immediatly after the release time (when the 7 airfoils are free to rotate).
Like a blow up, and only for some airfoils.

It depends (i think) from the setting of "Coupling Spring"

I thought that a spring with infinity stifness could be seen like a "rigid link".
I can't explain the final configuration of the airfoils.

I only want to know if i have to study the problem in another way or if it's possible to connect the airfoils each other (same rotation for all).

Ah, I misread your original post..... I thought you were wondering what "overlapped grid problem" meant. After seeing your picture, I think its obvious to everyone that its not a simple overlapped grid issue but rather the linking

I don't have much experience with body coupling (or DFBI in general) so I'll let others comment