[highpressure]

Hello everyone,

I'm attempting to replicate the simulation case presented in this article as a validation for my model. However, I'm encountering difficulties with the meshing process.

Each blade has been split into three parts, from apex to central zone. This allowed to keep thinner cells on the edge, crucial zone for vortex generation, and reducing the number of cells in the central zone. In the rest of domain, a gradual transition from zone around the rotor to boundaries has been constructed.

According to the authors, they divided the geometry into different sections to achieve refinement around curvatures. Additionally, the attached figure indicates they used layers with "relativeSizes" turned off (see attached figure).

I am facing the following issues:

1. I am unsure how to split the geometry in SHM to refine specific regions, such as blade tips.
2. When I disable "relativeSizes," SHM crashes. My average y+ is around 20, and I am uncertain how to configure the layer parameters to achieve optimal results.

I have attached the case file for reference.

Thank you for your assistance!

[Yann]

Hello Ilyes,

1. You should have several regions (named surfaces) in your STL/OBJ file in order to specify different refinement levels for each surfaces in snappyHexMeshDict. You should find resources about this on the forum. For instance: Naming regions in stl file for snappyHexMesh
2. When relativeSizes is set to true, the layers parameters such as finalLayerThickness are expressed as a fraction of size of the first cell next to the layer cells. If relativeSizes is set to false, this parameter is expressed in meters. So if you switch from relative to absolute size, you will need to adjust your layers settings accordingly. (see documentation: https://doc.openfoam.com/2306/tools/...-relativeSizes)

Regards,
Yann

[highpressure]

Bonjour Yann,


Thank you for your reply !

You should have several regions (named surfaces) in your STL/OBJ file in order to specify different refinement levels for each surfaces in snappyHexMeshDict. You should find resources about this on the forum. For instance: Naming regions in stl file for snappyHexMesh

I think the provided link doesn't directly answer the question. Could you please show me a tutorial on how to do that ?


I still have two other questions regarding the 3D simulation :

1. In a 2D simulation, how do I choose the thickness (extrusion) in the 3rd direction (Z) ?
2. While running a 2D case, I'm still getting forces and coefficients according to a third direction. Why ?

Regards,

[Yann]

Bonjour Ilyes,

I don't have a specific tutorial for that because it depends on the software you use to prepare your geometries. Some software supports regions in STL/OBJ files, others do not. So you will have to dig a bit more about that on your own.

Regarding your questions:

1. the thickness value does not matter, you can pick whatever you want. However, you might have to use the mesh thickness when defining your lRef/Aref values for the forceCoefficient function object in order to get the proper coefficients.
2. OpenFOAM works in 3D (your mesh still has a thickness in 2D mode) so it will output 3 force components. But in 2D the 3rd component should be 0.

Regards,
Yann