[vdeschuy]

Hi all,

It might be useful to start this thread with a warning that I am quite new to OpenFOAM. I am running OpenFOAM version 5.0 on Windows 10 with Bash on Ubuntu on Windows.

I am doing a 2D case of incompressible flow around a simple cylinder with pimpleFoam, currently at Reynolds 1000. The goal is to test different RANS models and see what gives the best results for CD and St (Strouhal number). From literature, I found that the average CD should be around 1.0 and St should be around 0.21 at this Reynolds number. I was working with kOmegaSST and I was running on a fine mesh, without wall functions as yPlus seemed very low, but the obtained CD was 1.3.

I now started over with a very very coarse mesh and wall functions. The idea here is to gradually increase the density of the mesh and see what changes. On the very coarse mesh, to my surprise, the calculated average CD was around 1.1 and St was 0.213. I then ran yPlus to check whether using wall functions was a good option and here's the problem:

The maximum yPlus value at the cylinder was calculated to be < 1 for every timestep?

I'll give some basic information about the case and then I'll include the files:

• flow velocity at inlet: 0.2 m/s
• diameter of the cylinder: 0.2 m
• kinematic viscosity: 4x10^(-5) m^2/s
• The meshing was done with snappyHexMesh

I added case files including the postProcessing folder which includes the yPlus values as an attachment. (I sadly had to leave out the .stl file as it was too large)

Steps ran in bash to reproduce my results:
1. blockMesh
2. surfaceFeatureExtract
3. snappyHexMesh -overwrite
4. extrudeMesh
5. pimpleFoam | tee log
6. pimpleFoam -postProcess -func yPlus

Thanks in advance for the help!