I'm finding it difficult to get converged results when using LES and DES, my lift and drag are very unstable over the whole simulation. How important is it to obtain convergence with these methods compared to RANS? I simulate flow around a vertical square cylinder at moderate to high Re with Fluent.

Any suggestions and thoughts are welcome.

Jim

This flow would be unsteady. So the forces oscillate in time.

Do the statistcs. The means must converge.