# LW33 sensitivity checks

2026-09-11. Unvalidated reduced-order model; no new CFD solution.

## Findings

- Settled: 33 cases
- Within numerical screen: 18 cases

51 paired cases: 15 load/speed cases, 18 assumed-CG cases and 18 wave cases. The calm-water screen checks settling; the wave screen checks time-step sensitivity, not settling. Passing either does not establish physical accuracy or safety.

## Protocol

CG shifts are assumed offsets from the reference LCB, positive forward; VCG is assumed height above historical keel. Mass changes hold CG fixed; no payload distribution is inferred. Waves ramp in over 3 s. Numerical screens are diagnostics, not safety or physical validation. Wave RMS differences cover matching 20 Hz samples over the whole run, including startup; displayed traces are downsampled to 2 Hz. Crossing waves use a 90 degree chop offset. Dry means weighted contact below 0.1%; re-entry counts require 0.05 s dry.

Each case starts with a 30-second calm, zero-speed relaxation for the specified mass and assumed CG. Then 60 seconds at prescribed speed. All cases are repeated at 240 and 480 Hz. Inertias and the 0.35 m drive lever remain fixed assumptions.

Calm settling requires final 10-second trim range <0.05°, heave range <1 mm, final vertical force residual <0.5% and pitch moment <10 Nm at both steps. Final mean differences must be <0.05° trim and <1 mm heave. Wave RMS differences must be <0.5° pitch/roll and <50 mm heave over matching 20 Hz samples, and both runs must complete. These are internal diagnostic screens, not validation criteria from a standard.

## Cases

| Case | Group | Speed kn | Mass kg | CG shift mm / VCG m | Sea | Finding |
|---|---|---|---|---|---|---|
| 1 | load | 20 | 3600 | 0 / 0.8 | Calm | Settled |
| 2 | load | 25 | 3600 | 0 / 0.8 | Calm | Settled |
| 3 | load | 30 | 3600 | 0 / 0.8 | Calm | Settled |
| 4 | load | 35 | 3600 | 0 / 0.8 | Calm | Settled |
| 5 | load | 40 | 3600 | 0 / 0.8 | Calm | Settled |
| 6 | load | 20 | 4100 | 0 / 0.8 | Calm | Settled |
| 7 | load | 25 | 4100 | 0 / 0.8 | Calm | Settled |
| 8 | load | 30 | 4100 | 0 / 0.8 | Calm | Settled |
| 9 | load | 35 | 4100 | 0 / 0.8 | Calm | Settled |
| 10 | load | 40 | 4100 | 0 / 0.8 | Calm | Settled |
| 11 | load | 20 | 4500 | 0 / 0.8 | Calm | Settled |
| 12 | load | 25 | 4500 | 0 / 0.8 | Calm | Settled |
| 13 | load | 30 | 4500 | 0 / 0.8 | Calm | Settled |
| 14 | load | 35 | 4500 | 0 / 0.8 | Calm | Settled |
| 15 | load | 40 | 4500 | 0 / 0.8 | Calm | Settled |
| 16 | cg | 25 | 4100 | -300 / 0.6 | Calm | Settled |
| 17 | cg | 35 | 4100 | -300 / 0.6 | Calm | Settled |
| 18 | cg | 25 | 4100 | -300 / 0.8 | Calm | Settled |
| 19 | cg | 35 | 4100 | -300 / 0.8 | Calm | Settled |
| 20 | cg | 25 | 4100 | -300 / 1 | Calm | Settled |
| 21 | cg | 35 | 4100 | -300 / 1 | Calm | Settled |
| 22 | cg | 25 | 4100 | 0 / 0.6 | Calm | Settled |
| 23 | cg | 35 | 4100 | 0 / 0.6 | Calm | Settled |
| 24 | cg | 25 | 4100 | 0 / 0.8 | Calm | Settled |
| 25 | cg | 35 | 4100 | 0 / 0.8 | Calm | Settled |
| 26 | cg | 25 | 4100 | 0 / 1 | Calm | Settled |
| 27 | cg | 35 | 4100 | 0 / 1 | Calm | Settled |
| 28 | cg | 25 | 4100 | 300 / 0.6 | Calm | Settled |
| 29 | cg | 35 | 4100 | 300 / 0.6 | Calm | Settled |
| 30 | cg | 25 | 4100 | 300 / 0.8 | Calm | Settled |
| 31 | cg | 35 | 4100 | 300 / 0.8 | Calm | Settled |
| 32 | cg | 25 | 4100 | 300 / 1 | Calm | Settled |
| 33 | cg | 35 | 4100 | 300 / 1 | Calm | Settled |
| 34 | waves | 30 | 4100 | 0 / 0.8 | regular, 0.6 m, 0° | Within numerical screen |
| 35 | waves | 30 | 4100 | 0 / 0.8 | regular, 0.6 m, 90° | Within numerical screen |
| 36 | waves | 30 | 4100 | 0 / 0.8 | regular, 0.6 m, 180° | Within numerical screen |
| 37 | waves | 30 | 4100 | 0 / 0.8 | regular, 2 m, 0° | Within numerical screen |
| 38 | waves | 30 | 4100 | 0 / 0.8 | regular, 2 m, 90° | Within numerical screen |
| 39 | waves | 30 | 4100 | 0 / 0.8 | regular, 2 m, 180° | Within numerical screen |
| 40 | waves | 30 | 4100 | 0 / 0.8 | irregular, 0.6 m, 0° | Within numerical screen |
| 41 | waves | 30 | 4100 | 0 / 0.8 | irregular, 0.6 m, 90° | Within numerical screen |
| 42 | waves | 30 | 4100 | 0 / 0.8 | irregular, 0.6 m, 180° | Within numerical screen |
| 43 | waves | 30 | 4100 | 0 / 0.8 | irregular, 2 m, 0° | Within numerical screen |
| 44 | waves | 30 | 4100 | 0 / 0.8 | irregular, 2 m, 90° | Within numerical screen |
| 45 | waves | 30 | 4100 | 0 / 0.8 | irregular, 2 m, 180° | Within numerical screen |
| 46 | waves | 30 | 4100 | 0 / 0.8 | cross, 0.6 m, 0° | Within numerical screen |
| 47 | waves | 30 | 4100 | 0 / 0.8 | cross, 0.6 m, 90° | Within numerical screen |
| 48 | waves | 30 | 4100 | 0 / 0.8 | cross, 0.6 m, 180° | Within numerical screen |
| 49 | waves | 30 | 4100 | 0 / 0.8 | cross, 2 m, 0° | Within numerical screen |
| 50 | waves | 30 | 4100 | 0 / 0.8 | cross, 2 m, 90° | Within numerical screen |
| 51 | waves | 30 | 4100 | 0 / 0.8 | cross, 2 m, 180° | Within numerical screen |

## Still needed

Measured CG datum, direction and height; drive geometry and inertia; confirmation of closed hull surfaces; measured drafts and weight inventory; sea-trial calibration and independent validation. The model does not solve engine power, full drive forces, added mass, horizontal wave orbital velocity or breaking waves. Real browser/device visual QA is not completed.

Model: cad-relative-incidence-v2. Geometry SHA-256: 96a1430c84b328aef57315beb8d6fda5cd279252ce4e9546babc16881a172708. Motion source SHA-256: 58d816f793f9c348e190f275a6deb0fbc810ab5f42df6b7cfde8531279e162d8. Reproduce with tools/run_sensitivity.mjs, then tools/build_sensitivity_review.mjs.