[bephi]

Hi at all,
i am doing some simulation of a flow through a pistol valve using the simpleFoam-solver containing k-w-SST turbulence model. unfortunately my calculations often crash or when they run from start- to endtime, solution are diverging.

maybe you have an idea, what can be improved to avoid such errors?! timeStep is calculated with CFL-No. <1.

examples:
1) here the velocity component U_z increases.

Time = 0.74498

DILUPBiCG:  Solving for Ux, Initial residual = 2.50978e-05, Final residual = 8.49602e-07, No Iterations 2
DILUPBiCG:  Solving for Uy, Initial residual = 2.54388e-05, Final residual = 2.2754e-06, No Iterations 2
DILUPBiCG:  Solving for Uz, Initial residual = 8.85837e-06, Final residual = 1.02878e+13, No Iterations 1001
DICPCG:  Solving for p, Initial residual = 1, Final residual = 0.00995054, No Iterations 108
DICPCG:  Solving for p, Initial residual = 0.113132, Final residual = 0.00109658, No Iterations 32
DICPCG:  Solving for p, Initial residual = 0.018198, Final residual = 0.00017058, No Iterations 98
time step continuity errors : sum local = 3.2839e+07, global = -124559, cumulative = -124377
63 additional processes aborted (not shown)

2) here just the cumulative time step error is big.

Time = 0.36633

DILUPBiCG:  Solving for Ux, Initial residual = 0.0126915, Final residual = 0.000657026, No Iterations 2
DILUPBiCG:  Solving for Uy, Initial residual = 0.00954027, Final residual = 0.000545504, No Iterations 2
DILUPBiCG:  Solving for Uz, Initial residual = 0.00808909, Final residual = 0.000426527, No Iterations 2
DICPCG:  Solving for p, Initial residual = 7.40292e-07, Final residual = 9.93734e-08, No Iterations 45
DICPCG:  Solving for p, Initial residual = 5.64208e-07, Final residual = 9.98093e-08, No Iterations 4
DICPCG:  Solving for p, Initial residual = 1.38451e-07, Final residual = 9.08471e-08, No Iterations 1
time step continuity errors : sum local = 11.0104, global = -0.956176, cumulative = 62676.9
63 additional processes aborted (not shown)

And what can in general be said about the duration the user should define (startTime /endTime)? Is there any relation between this defined time and the physical time which the fluid needs to flow trough the geometry? or is most priority that the variable we are looking for, is converged?

[herbert]

Hi bephi,

simpleFoam is a solver for steady state problems. Therefore the time step has no effect on solution, convergence, etc. The first step to improve stability is to reduce relaxation factors.

or is most priority that the variable we are looking for, is converged

Yes of course. You should only trust your solution if its converged.

Regards,
Stefan

[bephi]

Hi thanks for your answer.

Which URF should be decreased in your opinion?

Regards!
Philipp