Timescale convergence?!
 

Hi guys,

I'm running a NACA0015 profile at low reynolds numbers (10000-1000000) with steady sst including gamma-theta transition model. I varied my timescale to become a converged (steady state) solution. But I found out, that in some cases both decreasing and increasing the timescale from my converged timestep gives me an oscillating (transient) result.

My question is, does the "good" timescale maybe damp transient effects so the solution can convergate?

Cheers

at what AOA?

AoA=0, critical Re from 10000 til 100000. above 100000 it's ok.

If the flow is not in the gross separated flow regime (which I suppose it is not at AOA=0), then this is probably a laminar separation bubble. They tend to be unstable and jiggle about quite a bit. If the bubble is small it probably does not affect things much but at low-moderate Re (50k-200k for many airfoils) the separation bubble can be large. It might be important in this Re range.