Gaspar García, CFD portfolio
NACA 0012 at 10 degrees: lift extrapolated from three grids sits +2.14 % above a reference code and +3.4 to +3.8 % above experiment
CaveatedUpdated 2026-11-14
Objective. Predict the lift of a NACA 0012 at 10 degrees and Re 6 million with k-omega SST on a three-grid ladder, with a numerical uncertainty, against the NASA TMR reference computation and experiment.
| Reynolds number (chord) | 6 million |
|---|---|
| Angle of attack | 10° |
| Fluid | Incompressible, unit speed and chord |
| Solver | OpenFOAM simpleFoam, steady |
| Turbulence model | k-omega SST, fully turbulent |
| Wall y+ max | 1.65 coarse, 0.85 finest |
| Cells | 48,120 to 193,200 |
| Free-stream velocity | (0.984808 0.173648 0) |
Comparison with the references
| Quantity | Reference | Simulation | Signed difference |
|---|---|---|---|
| Lift, extrapolated, vs CFL3D (computation) | 1.07781 | 1.10083 | +2.14 % |
| Lift at 10 degrees, grid-corrected, vs Ladson (experiment) | 1.0628 | – | +3.59 % |
| Lift, grid-corrected, 8 to 14 degrees, vs Ladson | – | – | +3.41 to +3.77 % |
| Suction peak Cp vs CFL3D, coarse grid | −5.60 | −5.5924 | 0.14 % smaller in magnitude |
Lift coefficient against Cells
- Simulation, k-omega SST
- Extrapolated to zero cell size
- CFL3D, NASA TMR (computation)
- Ladson experiment, interpolated to 10 degrees
Verdict
Lift extrapolates to 1.10083, +2.14 % above CFL3D and +3.4 to +3.8 % above the Ladson experiment after grid correction, with a GCI of about 1.00 %. The finest grid fails one checkMesh test and drifts above the convergence gate, so the case is listed with caveats.
Limits
- Finest grid fails checkMesh (aspect ratio, 12 undocumented cells).
- Three grids only: the order cannot be checked; drag is not grid-converged.
- The NACA 4412 run is not blind and is not compared here.