[AlexRonto]

Hello all,

I do a CHT analysis at an electric generator. So I have 2 rotating disks (rotors) with a protrusion at their periphery and a stationary rim between them (stator). Inside the rim I have there are the coils of the machine. The rotating discs' diameter is 0.7m and the air gap between each disk and the stator is 4mm. The meshing is done at ICEM.

I firstly tried to put a sizing control with an automatic method so the elements are small enough near the walls (rotating and stationary) and the mesh I achieved is the one at the picture 5. (I suppose the triangles is not a good choice, as they create skewed elements and they do not capture the boundary layer region of the air gap properly.) However, the results are no good for the heat transfer. I have much higher temperatures than the ones I expect.

I then tried to put inflation layers near the wall boundaries, but ICEM could not create them for some reason.

So, my question is 1), if the the mesh that I have created could be the reason for the bad calculation of Heat transfer from the solver (I use Automatic wall function for the turbulence) .
And 2), supposing the problem of the bad results is the mesh, does anybody have any suggestions about the meshing options for the current geometry?

Thanks in advance!

[Gweher]

Hi Alexandros,

As you're using ICEM you should take advantage of the blocking capabilities of the software. Start with an initial 3D blocking, slice it and assign the edges to the curves and faces to the corresponding surfaces.
I would also recommend reading the cfd wiki for basic CFD analysis guidelines, especially for BL.