[Kalabagh]

Hey there,


I want to simulate a Venturi pipe with a geometry a little bit complex. Fluid is water and Reynolds is fully turbulent. For now I'm working on a simplified geometry that consists just in two pipes forming a T.


This is my workflow:


1. Solve for 1000 cPs (laminar) - linearUpwind, converges perfectly

2. Solve for 100 cPs (laminar) using previous data - linearUpwind, converges perfectly

3. Solve for 10 cPs (turbulent) using previous data - upwind, converges perfectly
4. Solve for 10 cPs (turbulent) using previous data - linearUpwind, does not converge
5. Solve for 1 cPs (turbulent) using previous data - upwind
6. Solve for 1 cPs (turbulent) using previous data - linearUpwind

As you can see, I cannot overpass point 4 because it does not converge...


Mesh is done by blockMesh and snappyHexMesh and checkMesh with all checks points just 1 failed check that is "concaveCells". Nonorthogonality is max 61, av 9.


I'm using simpleFoam with kOmega SST model and 2 nonorthogonal correctors.


I think it is a problem related with yPlus because I have been refining carefully the surface zones and this has allowed me to decrease the viscosity but now I have a max yPlus of 2,6 with wallFunctions and I can't improve the results for the point above said.


Of course I can send all sim files anyone need.


Thanks in advance.