Published maths. Every input printed.
No black box. Every result lists its inputs, its assumptions and the valid range of the maths, and warns when you leave it. Here is exactly what runs underneath.
What runs underneath.
Standard, documented methods only. Replaced by better models as calibration data allows; the version is printed on every report.
| Model | Method | Source | Valid range |
|---|---|---|---|
| Propeller | Wageningen B-series open-water polynomials | Oosterveld & van Oossanen, 1975 | Z 2–7 · EAR 0.30–1.05 · P/D 0.5–1.4 |
| Hull, displacement | Gerr speed–length relation, steep wall past max S/L from D/L ratio, ITTC-57 friction | Gerr · ITTC-57 | Displacement and semi-displacement hulls |
| Hull, planing | Savitsky planing surface: trim, wetted length, ITTC-57 friction, appendage drag | Savitsky, 1964 | Planing hulls · blended with the displacement model over the hump |
| Hull–prop interaction | Wake fraction w and thrust deduction t | Empirical, per drive | Per drive and speed–length ratio, single or twin screw |
| Engine | Rated power and rpm, torque curve per engine type, governor to high idle | Maker data | High idle 4–7 % above rated · gear ratio and efficiency |
| Seaway | Added resistance in waves, wind drag on windage, fouling factor | Kreitner-type | Current changes speed over ground only |
| Cavitation | Keller minimum EAR, Burrill-type index at 0.7R, blade loading, tip speed | Keller · Burrill | Immersion estimated from drive and weight · ventilation not modelled yet |
What is checked, and what isn't yet.
Accuracy beats features. Where something is still an estimate, the result says so.
| Area | What we check | Status |
|---|---|---|
| B-series implementation | Matches published curves: B4-70, P/D 1.0, J 0.8 → η₀ ≈ 0.69 | Checked |
| Hull model | Top speed against 127 measured boats (calibration v0). On 18 locked test boats the tuning never saw: mean error 11.7 %, bias −1 %. Cross-validation on the rest: 6.9 %. Pro calibrates each boat from its own sea trial. | Measured · v0 |
| Speed ranges | The printed range (±12.2 %) is meant to hold 8 in 10 boats. On the 18 test boats it held 61 %. | Below target |
| Stock props | Many stock props are segmental, not B-series; B-series tends to over-pitch them | Not corrected yet |
| New tools | Each tool checked against known cases before it goes live | Rule |
How a verdict is found.
Torque balance
Find the rpm where engine torque equals propeller torque at full throttle.
Thrust balance
At that rpm, propeller thrust equals hull resistance, including waves, wind and fouling.
Compare to rated
Where that rpm lands against the engine's rated range gives the verdict.
Size the prop
For each diameter, solve the pitch that lands full throttle on rated rpm, 0–5% reserve.
Rank
Best speed with a cavitation penalty. Payment never enters this step.
Inputs, assumptions, valid range and disclaimer go with every result.
A decision aid, not a guarantee.
Propbench gives engineering estimates. For shops, the shop or engineer makes the recommendation to their customer and stands behind it. For owners, confirm with a qualified prop shop before buying a propeller. Results lean to slightly underpropped rather than overloaded, and show ranges, not false precision, in the free version.