Hi Frank,
as @jairogut said in the other forum thread I would also consider this as “numeric noise”.
And also as long as you don’t see any clear oscillating frequency I would not consider doing a transient analysis, since those require a different approach and also can be a bit more tricky than the steady-state analysis. As long as you are not interested in transient effects and steady-state results work for you I would always first go with the steady-state simulation, which can also work well for wings depending on the wall turbulence model and the right mesh.
Best regards Sebastian