[kpax]

Hi everybody,

i am currently trying to solve the flow of fluid in a simple geometry with one velocity inlet and two pressure outlets. I want to implement a modified icoFoam solver that adjusts the pressure values at the outlets based on certain criteria (e.g., to achieve a certain predescribed flow split).
My main problem is that PISO seems to converge terribly slow. E.g., if I use a constant inlet velocity and constant pressure values at the outlets (i.e., steady-state), a relatively steady state is reached after around ten time steps, but both inlet pressure and outlet fluxes do not reach a constant value even after 100 time steps, although the changes are small in the end. Unfortunately, it seems that that many time steps are necessary whenever I change the outlet pressures.
So here are my questions:
- why does it take so long to converge? is there a way to accelerate this? I used timestep = 0.001; piso corrector steps did not have a significant effect.
Is it allowed to use PISO in this way?

Anyone who has something to chip in, please feel free!

thx in advance

[kpax]

just a small update: i also tried timestep = 0.0001s and constant inlet velocity. the inlet pressure converged to a constant value after around 1000 time steps!! this cant be right.


edit: the result of checkMesh:
Checking geometry...
Overall domain bounding box (-0.00500013 0.0737801 0.0766678) (0.0261582 0.0874276 0.0917605)
Mesh (non-empty, non-wedge) directions (1 1 1)
Mesh (non-empty) directions (1 1 1)
Boundary openness (-9.2092e-17 -2.65308e-16 -5.09871e-17) OK.
Max cell openness = 2.76457e-16 OK.
Max aspect ratio = 6.53503 OK.
Minumum face area = 1.51201e-09. Maximum face area = 5.16808e-08. Face area magnitudes OK.
Min volume = 2.2457e-13. Max volume = 5.80311e-12. Total volume = 3.88672e-07. Cell volumes OK.
Mesh non-orthogonality Max: 59.2528 average: 17.1145
Non-orthogonality check OK.
Face pyramids OK.
Max skewness = 1.52552 OK.
Coupled point location match (average 0) OK.

seems alright doesnt it? maybe non-orthogonality could be a problem?