# LW33 — calm-water numerical check

11 September 2026. **Approximate browser motion model only; not CFD or validated boat performance.**

## Setup and result

4,100 kg assumed load; no waves; free heave, pitch and roll; speed prescribed. Each run begins from the mass-only zero-trim rest position and runs for 60 seconds. Primary time step 1/240 s; repeated at 1/480 s. Default dynamic lift and assumed 0.35 m thrust lever.

| Speed (kn) | Instantaneous trim at 60 s (° bow-up) | Rise from rest (mm) | Displaced volume (m³) | Weighted contact (m²) | Run |
|---|---|---|---|---|---|
| 0 | 0.000 | -0.0 | 4.000 | 14.84 | Numerically settled |
| 10 | 0.274 | 6.4 | 3.880 | 14.53 | Numerically settled |
| 15 | 0.876 | 32.9 | 3.445 | 13.41 | Numerically settled |
| 20 | 1.225 | 60.6 | 3.038 | 12.35 | Numerically settled |
| 25 | 1.557 | 109.2 | 2.392 | 10.61 | UNSETTLED — reject baseline |
| 30 | 0.680 | 309.7 | 0.528 | 3.82 | UNSETTLED — reject baseline |

Rise is the change in vertical translation at the assumed CG, not bow or stern rise. Weighted contact is the model's immersion-weighted projected column area, not true wetted hull surface area.

Final-ten-second trim ranges: 2.15e-17, 0.00e+0, 0.00e+0, 3.03e-10, 6.35e-1, 5.59e+0 degrees, in speed order. Maximum time-step refinement differences: 1.61e+0 degrees trim and 5.68e-2 m rise. All figures are numerical model outputs.

Hydrostatic-only control at prescribed 30 kn: 0.0000° trim, -0.000 mm rise. Speed has no effect in this control because the dynamic lift and assumed thrust moment are disabled.

## Interpretation and limits

**The 25 and 30 kn cases remain unsettled in calm water. Their final values are instantaneous samples, not running-attitude predictions. At 30 kn the time-step change materially alters the sampled response. Do not accept this speed range as a baseline.** This may reflect the approximate lift/contact/damping formulation; these runs do not establish real-boat porpoising. The internal settling screen requires final-10-second heave range below 1 mm, trim range below 0.05°, final force imbalance below 0.5% and pitch moment below 10 Nm, with no stop at either time step. These are diagnostic thresholds, not a validation standard.

A converged time step or settled motion does not establish physical accuracy. The model assumes CG at the reference buoyancy-centre x, 0.8 m vertical CG, simplified inertias, damping, constant beam/deadrise and a partial Savitsky-inspired lift extension. It does not solve engine power, forward acceleration, full drive forces or propeller thrust. Artificial closure patches reach the loaded waterline, and registration to the display CAD remains provisional. The 2,750 mm photo-based CG estimate has no confirmed datum and is not used as a measured CG.

Do not tune the model merely to force the desired 2–3° running trim. Compare these results with independently measured drafts and running attitude before calibration.

## Checks still open

- Desktop/mobile browser visuals: unavailable in the current supported preview environment; automated and minimal-DOM checks are not device testing. Check afloat geometry, free flight, re-entry, fullscreen, touch controls and save/reload on actual devices.
- Physical inputs: LW team to confirm sealed hull, mass inventory, CG datum/direction/height and drive installation geometry; provide bow/stern drafts at known load.
- CFD: no new flow solution was run. Historical R18 is only 0.025 s of startup, with unresolved mesh connectivity defects; a settled 4,100 kg CFD baseline remains outstanding.
- Sea trial: collect reciprocal GPS speed, RPM, fuel flow, drive trim, load and measured running trim; reserve independent runs for validation.

Machine-readable results: calm-water-check.json. Reproduce with tools/check_calm_water.mjs. Full motion assumptions: motion-model.md.