Learning Forum

[Tom Davies]

Hi,

I am doing a simulation with a cylinder rotating on the top surface of an aerofoil. I want to apply the rotating wall boundary condition to the cylinder, (dont want to use the moving mesh method for simplicity.) Would this be a steady or transient simulation?

 

[Rob]

Steady ought to work OK. However, I'd have a very careful look at the cell growth rate and aspect ratio as the images don't look great. 

[Tom Davies]

Thanks Rob, I need to split the domain projection lines into more divisions, which area do you specificy mean?

The cylinder and aerofoil are connected, and I will just rotate the cylinder wall, this should work ok?

[Rob]

It should be OK to just set the cylinder wall motion, but check for odd effects at the ends. 

Re the mesh, have a look at the various NACA examples. You need to avoid jumps in cell size but also if you're expecting flow separation to avoid high aspect ratio cells. So, inflation + flow separation isn't good. 

[Tom Davies]

I did a simulation with the wall rotating, and I got CD and CL values very different from when the wall was stationary, but the velocity contours looked ok. I then tried a simulation at an angle of attack and got exactly the same CD and CL.

[Rob]

Look at the forces too, also check how the NACA wing sets the various constants for lift & drag coefficients.