Hi there

i´m trying to simulate a impeller in a duct. I´ve defined two domains (the duct and the fan) and a domain interface between them... the fan domain is rotating. Everything is ok but the solver displays a error

ERROR #001100279 has occurred in subroutine ErrAction. |

| Message: |

| No control surfaces have been found at all for Domain Interface 1- |

| . This is not supported. ????? What´s wrong??

thanks Pedro Lima

hi, could you precise how u defined your interface in CFX pre. Also, are you making a steady-state study or transient? What about the meshes ...it is a one to one interface? by

its fluid-fluid, with automatic pitch change.. they are both modeled using cfx build as surfaces

i forgot, its a steady-state simulation

Did you define control surfaces for Domain Interface 1? If yes, maybe you can try "None" for the pitch change if you are modeling the whole machine.

that's the problem.. how can i define control surfaces??

I´m new in CFX

Now i´ve tried to do a more simple simulation.. two cubes, one bigger and the other a small one inside the big one, each one is a domain... defined the boundary interface, fluid-fluid and the error is the same (no control surfaces found...)

You can't rotate a cube in this manner. The interface must be a surface of revolution. Try running the rotor-stator tutorial, it should give you a reasonably good idea of how to proceed.

-CycLone

thanks, i´ve already tried with two cylinders.. i´ve completed the turbo-wizard tuorial but what i need is very different. i found the "control surface" error, i was selecting "none" in the frame change option. now with the "frozen rotor" or "stage" option there´s a error

ERROR #555000001 has occurred in subroutine SRF_CORRECT. |

| Message: | GGI intersection error: The theta extent is too small on |

| Domain Interface: Interface 1 |

| Please examine your domain interface and look for the following: |

| 1) surface with more than 360 degrees of revolution |

| 2) surface intersects R=0 |

| 3) surface has faces normal AND parallel to the rotation axis |

| Change Transformation Type to None instead of Automatic in the |

| solver command file if any of the 3 situations apply. |

+--------------------------------------------------------------

???????????????????????????????????????????????i´m lost, i don´t know what is theta extend..

sorry to ask this, but can someone send me a a working cfx-pre file with a thing rotating inside other thing, so i can see whats wrong with my model? i´ve lost two days already with this problem, and i just don´t know whay the problem is... i´ve tried to change the rotacional domain to a stacionary one, but it gives me the "no surface control found" and i can´t use a frozen rotor or stage option with a stacionary domain (i think..)

Thanks Pedro Lima

Hi Pedro,

As CycLone points out, you cannot rotate a cube within a cube!!! The steady state assumption is that one domain is sliding past the other, but with only the change in reference frame being active, not the actual motion. This requires that the interface is a surface of revolution (i.e. the surface must be the result of rotating a curve about an axis). Unless you have re-invented geometry, a cube does not have a surface of revolution.

If you have already done a cylinder within a cylinder and the rotor-stator tutorial, you should already have the example you need. Just stop and think about what you are trying to do.

Regards, Robin

ok thanks for the advice

but i can´t do a steady-state simulation with a rotating impeller?only transient?

Hello Pedro Lima,

Ofcourse you can do a steady state simulation for a rotating impeller. By the way one basic check... did you make sure that the mesh of both domains are exactly lying in the same plane at the interface. I mean there should not be any small intrusion of one mesh in to the other. This may happen if you do the mesh of two domains separately.

Regards Anantha

He's trying to put a cube within a cube. It doesn't matter if they match perfectly, they need to be a surface of revolution, which they are not.

Hello, yes I understood that. What I have referred is to his actual impeller problem. There I just wanted to check if the meshes of two domains (rotating and non rotating) are lying exactly on the same plane with out intruding in to each other. Sorry if I have confused in my previous posting.

Regards Anantha

[Jaiganesh S]

I have also got the same problem in cfx solver for diffuser analysis.Can any one tell me how to finding the control surface of interface or solving problem.I had used same procedure in impeller analysis but i didn't get the problem.