Hi, Can the experts here give me their view on what they consider good convergence for engineering flows. The CFX documentation mentions it to be residual RMS=1 E -4. Is this appropriate? Does this change convergence criterion change from flow to flow? Adrian

Hi,

you can check up the following in help:
:Solver Modelling
:>Advice on Flow Modelling
:>>Monitoring and Obtaining Convergence
:>>>MAX Residual Level

Basically, 1e-4 is said to be "sufficient for most engineering applications", but the convergence criteria depends on the problem geometry and boundary conditions. I would recommend you to make a test on a typical flow problem that you are interested in: just run a few simulations with different residual levels, and check out the dependency of the flow quantity that you are looking for.

Taner Baytekin