[hconel]

Hi,

I am running simpleFoam with wall-resolved kOmegaSSt on an axisymmetric geometry and Uz residual won't go below 1e-4 while other variables are way below 1e-5. To be exact:

Time = 2612

smoothSolver:  Solving for Ux, Initial residual = 1.47806e-05, Final residual = 4.27788e-06, No Iterations 1
smoothSolver:  Solving for Uy, Initial residual = 0.000255141, Final residual = 2.18455e-05, No Iterations 2
smoothSolver:  Solving for Uz, Initial residual = 0.00580407, Final residual = 0.000204739, No Iterations 2
GAMG:  Solving for p, Initial residual = 7.11716e-05, Final residual = 6.89229e-06, No Iterations 15
time step continuity errors : sum local = 8.9928e-07, global = 1.83346e-08, cumulative = -0.0018131
smoothSolver:  Solving for omega, Initial residual = 1.03051e-05, Final residual = 3.00373e-06, No Iterations 1
smoothSolver:  Solving for k, Initial residual = 5.23285e-05, Final residual = 4.50814e-06, No Iterations 2
ExecutionTime = 815.15 s  ClockTime = 817 s

I might be doing something wrong in this simulation but this situation has raised some questions for me.

I have noticed that in planar (2D) geometries (with empty BCs) the Uz component is not calculated at all, whereas in axisymmetric geometries (with wedge BCs) the Uz component is calculated. First, this did not make sense to me, but if you consider a swirling flow with axisymmetric azimuthal component everywhere, it would still be an axisymmetric flow in polar coordinates. Am I wrong? In that case, the Uz component (azimuthal component actually) should converge.
On the other hand, I do not think that OpenFOAM switches to polar coordinates in an axisymmetric simulation; it just carries out 3D computations with special BC treatments (right?).

So my question is, is setting the residual control for Uz to a large value completely safe in axisymmetric geometries in OpenFOAM?

Thanks in advance,

Huseyin