[vaina74]

My aim is to study the performance of a 'hydro' foil with OpenFOAM (and later of a propeller) and I'm going to create the computational domain with Salome. The Reynolds number is about 1.87E6, so I'll use a turbulent model. I have some questions:

1. is simpleFoam suitable?
2. k-e, k-w or an other turbulent model?
3. if i'm right, k-w doesn't need a wall function. are there any mesh generation guidelines for this model?
4. if i'm right, k-e needs a wall function: how can i evaluate a rough y+ value (and the appropriate distance of the first nodes by the wall) by the Re number?

Thanks for your help.

[linnemann]

Hi

Yes simpleFoam works great.

k-wSST model works fine for this type of simulation. k-e realizable also works fine.

Wall functions depends on what type of mesh you are able to generate. If you can create a mesh with a y+ below 1 in the first cell you do not need wall functions and should use a lowRe turb model.

If y+ is above 15 in the first cell you should use wall functions and a high Re model.

I havent come across a method of estimating the first cell thickness based on the overall Re since y+ is very flow driven. I have tried using some estimations but they never end up fitting anyway. I suggest creating an initial coarse mesh and do some iterations, check the y+ value and adjust your mesh accordingly.

Just for inspiration here is a mesh created with Salome with a momentum source (actuator disk) in the middle of the foil.

[vaina74]

Thanks for your reply. I generated the domain around my foil, as in the pictures - I can further refine the mesh close to the wall. The blade section is about 300 mm. I think I will first apply the k-wSST model, then I will use different models, as k-e or Low Re.
If I'm right, I can't apply bc with Salome. How can specify patches and apply bc to an imported mesh? Are there any OpenFOAM utilities?

UPDATE