[marsmoonlight]

Hello! I'm new into SU2 and CFD in general and I'm trying to simulate flow over airfoils using SU2, I tried to use the incompressible inviscid hydrofoil .cfg file in tutorials to simulate air over a shape I modeled in GMSH (It's a first approach so it's not a great model), I only changed in the .cfg the density of the fluid, the speed, and the .su2 mesh name so it could use the mesh file I generated, I named my physical curves as they were in the .cgf file, inlet, outlet, airfoil, etc. But when I run the simulation the coefficientes, CD and CL are zero, when I reach the 200 iterations and I see the result in paraview I notice that the simulation is simply not doing what I want, I don't know what I'm doing wrong. The same happened when I tried to use the turbulent example of the naca0012 with spallart almaraz.

I'll attach my .cfg and .geo, as well as some screenshots of the geometry and the results in paraview. Thanks in advance for any reply! I'm sure that maybe I'm doing a lot of mistakes (And sorry for that) but I'll use them to learn.

[bigfootedrockmidget]

Your setup is not converging on this mesh, you can see this by looking at the pressure and velocity residuals. Try for instance a much smaller inlet velocity (U=0.1 m/s) and a smaller initial velocity, using only a velocity in the x-direction. Also try a much smaller value for your CFL number (CFL_NUMBER=1).

You will need much more iterations for such a small CFL number, but at least it will converge (a bit). In this case, the lift will be 0 of course.

Since your airfoil is in a very small domain, the inlet and outlet, and walls will severely influence your results. The boundaries are actually so close to the airfoil that it will probably be difficult to get a fully converged solution unless the velocity is very low.

If you want to investigate lift and drag for different angles of attack, have a look at the mesh size of the naca0012 - the distance from the airfoil to the edge of the domain is quite large. Also notice the use of a circular domain using free-stream conditions on the edge.

[marsmoonlight]

Thank you for your reply! I did what you recommended, I created a way bigger mesh, decreased courant number and tried to run the simulation, but anyway the coefficients are zero. I'm not reaching convergence yet, so maybe working with the mesh could fix the problem.