I'm trying to solve a Thermal and Fluid Dynamic problem where there are the following issues. I have a rotating wall (which is a rotor). This steel rotor has its own thickness (so, I'm solving a conduction problem too). Outside this rotor, I have the cooling flow (air) which has to cool down the rotor wall (so, on this side, I have a convection problem). When I setup the boundary conditions and I consider this rotor rotating wall, I should choose (for the thermal side)one of the following options (but no one is correct for my problem): 1. Adiabatic (my problem is not adiabatic) 2. Heat flux (I don't know it, because for my problem the conduction problem is not mono-directional only; the heat transfer is also present through different stationary parts around the rotor) 3. Temperature (I don't know it, because it is the result I would like to find) 4. Heat transfer coefficient (I cannot setup it, because it is a function of the flow field from the convection point of view).

How might I setup a rotating and non-adiabatic wall (at the same time) as boundary condition? Does anybody can help me, please? I hope so!

Have a good day !!! Davide

Hi,

The rotor to fluid interface should be an interface, not a boundary condition.

Glenn Horrocks

Dear Glenn, thank you very much. Your suggestion is completely right but, in my case, a problem persists because I don't know if I can define an interface as following: "rotating surface" and "interface between fluid and solid" at the same time. If yes, I don't know how I might do it. Thanks again Have a good day Davide

You shouldn't have a rotating surface. Rotating walls are only valid if the wall velocity is tangent to the wall. In the case of a rotor you have a normal component to the wall velocity - in other words the wall "pushes" the fluid. Therefore you must use a rotating fluid domain around the rotor; the rotor wall will be stationary in the rotating frame. Your shroud can be a counter-rotating wall, since it will not have a normal component to the wall velocity.

I would like to thank you very much for all your suggestions.

I'm asking if I might consider another opportunity I saw...but I'm still investigating on it. I can consider every part of my domains on a stationary frame (this is the way I used for meshing my domains, specifying every single material). I can define a solid-fluid interface between fluid (air) and rotor solid. At this point, the following things happen: 1. Creation of the domains interface (we can see it on the bottom side of the pre-processing window...where also periodicities and GGIs are defined) 2. At the same time, I can see the interfaces effects on each domain. I mean: I see the interface as "boundary condition" on the corresponding domain (with the same name I gave to the interface but with the different side number specified). 3. If I open this "kind of boundary condition", I can specify a wall velocity value on the corresponding wall (my rotor wall but...on the fluid side).

But another problem might persist ! Is it numerically correct? How does the code solve this kind of setup?

I hope you can apologize me for my wild definitions.

Thanks Davide

Hi,

If the domains on both sides of an interface are rotating then doesn't the interface rotate as well?

Glenn Horrocks