[zachnap]

Hello:

I made a fully-structured mesh of a generic projectile with a wall-spacing calculated from this site = 1e-6 m. I ran the mesh in Fluent with a kw-SST turbulence model with 1% turbulence intensity and 1 viscosity ratio at 100 m/s... my results were acceptable, qualitatively speaking; quantitatively off but... within reason. I ran the same mesh at 30 m/s, same settings, etc. and I get strange results - completely off in every respect... well, for half the length of the projectile. the first half of the projectile looks reasonable (and physical), yet the second half is completely separated flow which I know should not be there since we have PIV to use as a validation tool and is just nonsense anyway.

Is there anything that I should be accounting for when using a turbulence model and transition to siginificantly different velocities that could be causing this result? The mesh works well for Laminar runs at both 100 and 30 m/s.

I cannot include pictures...