[saisanthoshm88]

CFX has got two forms of convergence criteria:

1. The residuals of continuity, momentum and energy equations.

2. The Domain Imbalances.

It is possible for the residuals to have converged well but the imbalances to not converge; and it is also possible for the residuals to not have converged but the imbalances have converged.

So to consider a solution to have good convergence what should be the percentage of domain imbalance in general? (For a Steady state simulation and a Transient simulation)

I know that some quantities of interest should also be monitored to judge the convergence.

But I’m particularly interested in knowing how the domain imbalances are indeed monitored.
I mean reasonably of what order they should seem to be to say that the solution had converged well in terms of the domain imbalances.

Is it always required to have them converged by 1%.

[ghorrocks]

It is possible for the residuals to have converged well but the imbalances to not converge;

Yes

and it is also possible for the residuals to not have converged but the imbalances have converged.

Yes

Different simulations need different convergence. Read the CFX documentation for a guide, but ultimately you need to do a sensitivity analysis to check you have sufficiently converged for you specific case.