[mozkan26]

Dear all,

I`m trying to simulate the flow in an enclosed rotor-stator system. The attached figure represents the quarter of my geometry. The axis of rotation is y-axis. Faces are the rotor (red), the stator (blue) and the shroud (yellow) in the figure. The shroud is attached to stator and hasn`t got any motion/rotation.

My question is that I couldn`t find any info from Fluent User`s Manual about the discontinuity between the rotor and the shroud. Does anyone know how does Fluent resolve this discontinuity problem on the line between these stationary and rotating walls?

Regards,

Musa

[amin.z]

Originally Posted by mozkan26

Dear all,

I`m trying to simulate the flow in an enclosed rotor-stator system. The attached figure represents the quarter of my geometry. The axis of rotation is y-axis. Faces are the rotor (red), the stator (blue) and the shroud (yellow) in the figure. The shroud is attached to stator and hasn`t got any motion/rotation.

My question is that I couldn`t find any info from Fluent User`s Manual about the discontinuity between the rotor and the shroud. Does anyone know how does Fluent resolve this discontinuity problem on the line between these stationary and rotating walls?

Regards,

Musa

Hey
Is rotor rotary and shroud stationary?
Is them have a contact? Contact between rotary part and stationary part?

[mozkan26]

Hi Amin,
Thanks for your response. Yes the rotor is a rotating wall and the shroud is stationary wall. I have just modeled a 3D cylinder and defined its faces as my rotor, shroud etc. So yes there is a contact between rotating and stationary walls. And there is a discontinuity on that contact which is the line between these faces in my case.

[ghost82]

This can't be a real machine: why do you want to simulate a non real object?

[mozkan26]

Thanks for your reply Daniele,
Yes, you are right. In reality there should be a thin gap between rotor and shroud, say, the gap is 0.01*r (r is the radius of the rotor). I found in some papers, for example Severac & Serre (2007), that this kind of thin gap hasn`t got a significant effect on the flow and they used the geometry I attached in their own simulations. They said the discontinuity is regularized by employing a BL function into their code. So I just wonder if Fluent has any specific capability to resolve that kind of issues. Or should I consider to implement a UDF to that boundary somehow?

[amin.z]

Do you have any heat transfer in solid domain?
because for solid domains only energy equations is solving, if you don't want to calculate temperature distribution, I think you can assume the part of the shroud that have a contact with rotor, as rotary part!

And why would you want to model a contact between them? Why you don't model a gap between the parts?

[mozkan26]

Thank you Amin,

There is no heat transfer. I tried to run it and it seems working without any error notification. But I guess I need to carefully consider the results since I don`t know what is happening there.

Regarding your suggestion, if I model a thin gap between these parts, would it be just defined as an opening or what? There would be no inlets/outlets apart from that thin gap, so would it create a convergence problem in continuity equation?

Many tanks with regards,

Musa

[amin.z]

Originally Posted by mozkan26

Thank you Amin,

There is no heat transfer. I tried to run it and it seems working without any error notification. But I guess I need to carefully consider the results since I don`t know what is happening there.

Regarding your suggestion, if I model a thin gap between these parts, would it be just defined as an opening or what? There would be no inlets/outlets apart from that thin gap, so would it create a convergence problem in continuity equation?

Many tanks with regards,

Musa

I think you can define boundry conditon for fluid in the gap position as a wall! And I guess it's will working!

[mozkan26]

I do not understand your suggestion, I'm afraid. If I define the gap as a wall, why should I remodel my domain? Do you mean that I should separate my rotary part and shroud with another little stationary wall? If so there is still singularity between my rotary part and the thin gap (defined as a stationary wall), isn't it?

[amin.z]

Originally Posted by mozkan26

I do not understand your suggestion, I'm afraid. If I define the gap as a wall, why should I remodel my domain? Do you mean that I should separate my rotary part and shroud with another little stationary wall? If so there is still singularity between my rotary part and the thin gap (defined as a stationary wall), isn't it?

It's will be better If you have 2 seperate domains as rotary and stationary but becuase the gap size is too small, generating this 2nd part and meshing its need great time and memory! So you can have a fluid domain, rotor, shroud and a hollow gap! So walls of rotor and shroud side of gap will be adiabatic and the fluid BC side of gap will be a simple wall!

[mozkan26]

Thank you very much Amin. If there is any further approach to this topic, it will be appreciated. Thank you all for your kind responses.

Regards