[yongxiang]

Hi everyone,

I have a weired simulation, you can see the following log. The Uy and Uz takes 1000 iterations, and it start with a smaller residual but end with a higher residual. Do you have any hint for this problem? I am using icoFoam.

Courant Number mean: 0.00395302 max: 0.501596
smoothSolver: Solving for Ux, Initial residual = 2.0073e-07, Final residual = 2.0073e-07, No Iterations 0
smoothSolver: Solving for Uy, Initial residual = 4.25849e-05, Final residual = 0.000558133, No Iterations 1000
smoothSolver: Solving for Uz, Initial residual = 9.46643e-05, Final residual = 0.000694397, No Iterations 1000
DICPCG: Solving for p, Initial residual = 0.00259512, Final residual = 0.000119937, No Iterations 9
time step continuity errors : sum local = 6.14191e-12, global = 5.95492e-15, cumulative = 2.63548e-11
DICPCG: Solving for p, Initial residual = 0.00256588, Final residual = 0.00011219, No Iterations 9
time step continuity errors : sum local = 5.74491e-12, global = 8.04627e-15, cumulative = 2.63629e-11


the setup for U in fvSolutionDict is:

U
{
solver smoothSolver;
smoother symGaussSeidel;
tolerance 1e-05;
relTol 0;
}

Regards
Yong

[Tobi]

Yes, the problem here is (in the first guess), a bad set-up. In other words your BC lead to a set of equations that have infty solutions.