Hello,

I have a question. My residual monitors for continuity are oscillating in a fixed pattern between 1e-4 and 1e-5. The x,y,z,energy,epsilon and k residuals all oscillate below 1e-5.

I am modeling an unsteady incompressible flow over an airfoil. Is this okay? The oscillations amplitude have decreased with the number of iterations.

IS THIS OKAY?

ANY HELP WOULD BE GREATLY APPRECIATED. THANKS YOU!

-Manish

Hi,

I think is ok, all depends what you want, oscillations in the residuals are very common, an oscillation of constant amplitude tell you that you reach some degree of convergence, a decreasing amplitude with more iterations tell you that with the initial chosen number of iteration is not enough, here you have two options 1) increase the number of iterations until a constant amplitude (or a x value of residual - some cases need values about 1-e6 other 1e-4 or more - all depends what you want, remember in theory a case is converged with the residuals are zero wich for most practical cases is never reached) 2) improve your mesh, a fine mesh gets better results, and remeber that the solution most be independent of the grid, all this things are in theory, in pratice you fix your values to what is acceptable in cpu time and precision.

I hope this helps, comments about the message are very welcome, I'm not a CFD guru, yet!

FC