Simple single element wing convergence
 

Hello, I have been doing an angle sweep of a wing profile I drew.

I using a 2D steady analysis using Spalart-Allmaras model.

I then look at how the lift and drag forces change with AoA. The behaviour so far is very linear but the convergence values are not particularly tight so the actual values could be a little bit off.

When doing the 15deg AoA, again the forces are a good fit with a linear curve. I decided to run the model overnight without the convergence stopping the simulation. Now the residuals cycle up and down along with the surface pressure integral monitor.

I don't know if by now the case should be unsteady as the plots indicate it could possibly be related to a vortex shedding type of cyclic behaviour, but the flow behind the wing isn't very aggressive.

Does anyone have any ideas how I should proceed?
How do people normally monitor convergence of a wing? Ideally I guess we would monitor perhaps a Cl plot but I don't think this would ever converge to a satisfactory level.

What is weird is the fact that plotting the lift and drag force after 500 itterations of each case yields a predictable linear wing characteristic curve.

By the way, the chord of the wing is 0.18m, it's in a rectangular domain with 1.5m length and 1m height, pressure inlet 2000pa, pressure outlet 0pa, which causes an inlet velocity of around 57m/s. There are around 300,000 elements in the grid.

Do you think that k-epsilon model would be a better alternative?

I hope someone can help me so I can start having some more confidence in the results.

Thanks a lot.