I am just wondering how to know that my transition simulation has converged. actually, I am running a steady state simulation on airfoil case, and I found local convergence problem which, I guess, is caused by seperation bubble moving, i.e. it is actually a transient problem. So I start running transient simulation. what 's your comments on my case, and how do I know it is already converged (I have no experience on transient simulation).

tnanks, anyway

littlelz

when your lift and drag curve reaches to some periodic fluctuations.

Hi,

Assuming lift and drag are your key outputs, do a sensitivity check of mesh size, convergence tolerance and time step size (if transient) to assess the level of accuracy of your answer.

Glenn Horrocks

Thanks for your reply.

That's what I am concerning; the timestep size in transient simulation. In my case, as I set the time step size very small, say 1e-8, it is converged quickly, 3-5 coeff loops. as I increase it, say 5e-7, it is very slow to reach convergence. What's that mean? how do I decide a proper tiemsetp size? Please advice

Hi,

You find the correct timestep size by a sensitivity study. If you are getting 3-5 coeff loops at a timestep of 1e-8 then that is a good place to start as that is about the point where the transient solver usually works best - but check with a sensitivity analysis.

Glenn Horrocks

oh, I see.

many thanks for your help. all the best