[Drunken_SheeP]

Hello everyone - my first (but surely not last) post in this forum! I have been working a lot in my studies with CFD and have now started a PhD doing some CFD - unfortunately I am the only simulation guy in my department! So I have literally nobody that I can talk to about some issues - likes this one:

I have been running 2D airfoil simulations for a long time now. And I have been mostly been told which mesh type to use. However, now that I'm doing my own work, I'm wondering which type is the best (or what are the advantages?). Below are some example of what I've used. Mostly I've used a hybrid mesh with a structured resolution of the boundary layer and an extension of the wake region. I also have used an O-type grid which is based on a hyperbolic extrusion - it was for me one of the fastest meshes to create. The C-type seems to be very common but I've not used it too much and it takes some work to make it fit nicely. Last but not least I've worked with H-grids (no example on hand) which were extremely cumbersome to create and I only used them for a very specific topic (acoustic wave propagation).

I'm mainly looking for rather simple calculations of lift, drag and maximum angle of attack at low Reynolds number (Re= 0.2 .. 1E6). Any advice on what to use or what the advantages / disadvantages are?

Thanks!

[scipy]

I'd go with your last picture. I'm guessing one would call that a complete O-grid?

In any case, my logic with airfoils was always as follows:

1. For small values of AoA you can use either your first grid or second grid since flow will mostly be aligned with your structured elements.

2. If you are going up to higher values of AoA then both 1st and 2nd picture run into problems once the flow direction is at >20-30° inclined vs. your structured elements direction. Basicallly, you lose the advantage of structured elements. In this case, there'd be no difference to just using tria/tetra once you're outside the boundary layer.

3. I'd go one or two directions:

a) low AoA = C-grid is ok, fully structured.
b) higher AoA = full O-grid (fully structured) - since you have a circular domain basically any AoA you choose, you still have most of your elements aligned with the flow direction - OR - you go with a structured (quad) layer around the airfoil and unstructured (tria) everywhere else and just use refinement in the solver (based on pressure gradient, shock location etc).