Convergence not reached/ k epsilon

Hey together,

so iam confused, why the Simulation didn't stop, after I reached the convergence criterion. So iam using the cpu booster, 3 multigrid levels, with CGI, 50 Iterations per Grid Level or convergence criterion -6 ( in Simulation/numerics. For the final mesh in Simulation/control 600 iterations or the convergence criterion of -8 ( in Simulation/control). As you can see in the picture for the fluid global residual, iam reaching the convergence criterion of -8 but the simulation didn't stop. Do every residuals needs to reach the criterion. Because as you can see in the second picture just k max have a value of -6.9.

And where can I see the simulation time after it's finished. During the Simulation I can see it in the command manager, but if the simulation is finished I can't see the final time anymore. I use Omnis 5.2, do you recommend me to set the convergence criterion to -6.8 to  safe computing time, or  at least to reduce the total iterations to 578 because here the residual for k max starts to become constant.  Iam using -8 because I read its a recommendation for the cpu booster.

Otherwise iam using a CFL Number of 100 and CFL number ( in CGI with CPU Booster ) also 100 ( in domain numerics)

Besides, I'm still a little stuck on calculating k and epsilon. I use the open: density-based solver and k-omega SST. I expect high velocities, above Mach 0.6 or an axial velocity above 500 m/s.  Is it sufficient to use for my case the shown formulas in the appendix for estimation ? Are Lref the height of the channel or domain ? Since I have a turbo machinery flow  I use the ratio    .  In the formula for epsilon, where the circle is, the value for this is 1 ? Otherwise, so far I have used an online tool to estimate these two values. They turn out very high, for k 3750 m2/s2 and epsilon 17.000.000 m2/s3. I calculate with a turbulence intensity of 5%.

thx and greetings Hussein