Gaspar García, CFD portfolio
A small propeller: tracing a +27.8 % thrust error to the blade pitch, with power still unexplained
CaveatedUpdated 2026-11-14
Objective. Reproduce the measured thrust and power of a small retail propeller at one advance ratio with a steady rotating-frame RANS model, and find where the disagreement comes from before tuning anything.
| Advance ratio J = V / (n D) | 0.523 |
|---|---|
| Rotational speed | 4998 rpm (523.389336 rad/s) |
| Diameter | 0.2794 m |
| Free-stream speed | 12.172312 m/s |
| Chord Reynolds number | 6.3 × 10⁴ at three-quarter radius, 2.0 × 10⁴ at the tip |
| Fluid | Incompressible air, kinematic viscosity 1.5e-05 m²/s (test temperature not published), density 1.225 kg/m³ |
| Solver | OpenFOAM v2412, simpleFoam (SIMPLEC) with a multiple-reference-frame rotor zone |
| Turbulence model | k-omega SST, fully turbulent |
Comparison with experiment
| Quantity | Experiment | Simulation | Signed difference |
|---|---|---|---|
| CT, re-pitched to 5.5 in (blades and hub) | 0.0156 | 0.01558 | −0.11 % |
| CP, re-pitched to 5.5 in (blades and hub) | 0.0152 | 0.01696 | +11.6 % |
| Efficiency, re-pitched to 5.5 in | 0.535 | 0.481 | −10.2 % |
| CT, table loft 5.80 in | 0.0156 | 0.01994 | +27.8 % |
| CP, table loft 5.80 in | 0.0152 | 0.01932 | +27.1 % |
Power coefficient CP against Advance ratio J
- UIUC propDB (kt0470, 4998 rpm)
- Simulation, re-pitched to 5.5 in
- Simulation, table loft (5.80 in)
Verdict
The table loft over-predicted thrust and power by +27.8 % and +27.1 %, the signature of too much pitch; re-pitching to the nominal 5.5 in removed the thrust error. Power stays +11.6 % high, unexplained, on one grid with a waived mesh check.
Limits
- The thrust match is not a precision number: 1 % in J is 5.4 % in CT, the pitch is uncertain and the hub is estimated.
- One grid, no numerical uncertainty; checkMesh failed on 2 blade faces (waived, recorded).
- Near-wall cells aft on the chord are far coarser than designed, cause unknown; the case is parked.