[Stuntmanbob]

Hi,
I want to simulate the flow around a turbine blade in a linear cascade. So I built a structured Q3D mesh with gmsh (hexaedra cells). The cascade only contains one blade, but has a periodic interface on the channel walls. I also built an O-Grid around the blade, with mesh refinement for the boundary layer.
The problem I have is that the pressure distribution shows around the blade shows an oscillating behaviour downstream of the stagnation point on the Trailing edge. See picture.

What I’ve done until now:
1) Increase convergence. I started the simulation with the SST turbulence model. The convergence behaviour was pretty poor. Using the k-omega turbulence model the residuals decreased below 1e-5, apart from U_z (I think this doesn’t matter in Q3D simulations).
2) Of course I checked my mesh. I’ve tried a finer mesh and made some adjustments on my O-Grid. I know have a maximum yPlus of around 0.77 on the suction side, the average is around 0.31. CheckMesh says “Failed 1 mesh checks”, but that’s because of the High aspect ratio of the cells in my boundary layer (Max aspect ratio: 3567, number of cells 2249). Before my adjustments the maximum aspect ratio was around 20k, but the pressure distribution looked the same.
The other mesh parameters are:
Mesh non-orthogonality Max: 80.63584, average 22.32671
Number of severely non-orthogonal (>70 degrees) faces: 58
Max skewness = 3.607935
3) I also tried the standard wall function on PS+SS instead of the Low Re Wall function, but it had also no effect.

Additional information:
1) I’m a beginner with openFoam, so I use the free Version of simFlow on Windows 10 for the setup. The Version is 3.1.22. It uses the OpenFoam Version 1612+. This is what it says when I click on advanced Settings.
2) The solver is rhoSimpleFoam
3) I’m using paraFoam for postprocessing with Ubuntu on Windows10, with the WSL (Windows subsystem for Linux).
4) For the pressure distribution I activated the pressure and suction side mesh parts in paraFoam and created a slice through the surfaces. So I could use the ExtractSurface Filter. Then I write the data with File -> Write Data and read and plot it with python. I also tried to plot the data directly in paraFoam, but this hasn’t worked yet.

Besides the picture I will also upload my system folder in a zip file. If anyone has any ideas how to fix the problem I would be very glad.


Thank you for reading my long post,
Stuntmanbob