Hello,

I was wondering if anyone has ever had any experience with residuals that intially start to converge and then after several iterations (i.e. 25,000+), they start to diverege???

I am working with the Reynolds Stress Model and I have decent convergence (1e-05) on everything except continuity (1e-2). In trying to get continuity to converge I let the simulation run and now at 25,000+ iterations I'm noticing that my Reynolds stresses, y-velocity, k and epsilon are starting to diverge slowly...Is the simulation adjusting itself in order to get the continuity to converge???

If it helps at all I'm simulating a flat plate boundary layer with super super fine mesh and low freestream intensity (0.2%).

Any ideas???

Thanks so much, Erika

It is normal dont worry

So then should I allow the solution to progress??? Will it eventually converge???

yes. But not in absolute terms. It may start fluactuating between some value i.e diverging and converging. check the mass flow balance(it is ur gloval convergence)

Thanks for your help!

wats ur msn or yahoo id?

I don't have one.

which type of flow u r working on

I'm simulating a boundary layer experiment that was conducted in a wind tunnel with very low free stream turbulence intensity (0.02%). I'm actually trying to duplicate the turbulence intensity curves that Klebanoff published in his experiment from 1955 (Characteristics of Turbulence in a Boundary layer with Zero Pressure Gradient).