[Joao Bomfim]

Hi friends,
I am trying to validate a simulation of a VAWT starting with a 2d case. I'm doing the grid independence test, but the grid doesn't converge satisfactorily over a four-mesh range of 40,000 elements to 350,000 elements. The default boot is the same for all meshes. The y+ values for the cells over blade surfaces are smaller than 1, and the turbulence model used is ke realizable with enhanced tratament in the walls. My finnest mesh is in the picture. When I was using CFX, I remember that for a case with two nodes on the z axis (pseudo 2d), the grid converged at about 150 000 to 200 000 elements. So what I would like to know is:
- Doesn't fluent use turbulence models or the laminar case to test the grid?
- Whatever use static or rotating mesh? (Sliding mesh)
- I use second order equations, but can I use first order?
If anyone can advise me more on this subject in fluent I appreciate it because I don't see anything similar to what I got in cfx. Thanks!

[MIDHUN@CFD]

try to get aspect ratio below 600 and dont use RKE with mrf or rotating domains, its not compatible. use sst k w instead .. you can use first order if convergence is slower and try using PRESTO for pressure

[Joao Bomfim]

Thanks for your reply!!! Could you tell me why rke is not compatible with rotating mesh? This is first time I see it. Best regards

[MIDHUN@CFD]

becoz the swirl component get cancelled