[RobertH]

Hello everyone,


I’m new to CFD and to ANSYS's software and I want to do a master thesis on bicycle wheel aerodynamics. In a first step, I want to investigate the drag force of the leading edge of a bicycle wheel in 2D using different tire shapes. In the future I want to switch to 3D, simulate a small section of the leading edge and maybe a full wheel.

At this moment I experience problems with the meshing. Up to now I have used DesignModeler for the geometries and Ansys Meshing for creating the meshes. Some colleagues already helped me on some aspects of my meshes and told me to get y+ = 1. Simulations were carried out using Ansys Fluent with k-w-SST-model and a lot of default settings. The flow in x-direction is 10 m/s.

1. With mesh (1), I calculated a “lift force” (y direction) of 0.2 N in ANSYS Fluent (compared to 0.6 N drag force). I think anything different from 0 N is nonsense. With meshes (2), (3), (4) the lift forces were 0.006 N, 0.035 N and -0.019 N respectively. Can this difference be explained only by the poor boundary layer of mesh (1)?
2. In (detail boundary layer) the boundary layers are parallel to the geometry except of the cell inside the corner (meshes 2, 3, 4). Is this an issue and if yes, how can the boundary layers be arranged so that they lay perfectly parallel to the geometry?
3. I tried to make a mesh like (5) and visualized it in MS Paint. However, by splitting up my body I was only able to generate meshes (3) and (4), I couldn’t generate a mesh like (5). My idea was to have a block structured mesh with a finer mesh inside the inner rectangle. Is the mesh (5) better than (3), (4) and if yes, how could I generate such a mesh? If required I can explain how I generated (3) and (4).
4. The geometry is symmetric in y-direction. Is it possible to specify a symmetry axis/plane and mesh symmetrically around that? Using just half of the geometry with a symmetry BC in the simulation is not an option because I want to change the angle of attack.
5. Could I use either of the meshes as a base for a 3D mesh with prisms of this shape in z direction? If yes, how does it work?

I’d be grateful for help on any of these questions! Anything that I’ve done so far is debatable since I have little to no experience with CFD.


Best regards
Robert