Active Stabilization Feasibility

01 · Verdict

Yes — this is the right tool for this hull

Your own spec says a 1-foot rise in water level changes buoyancy by 1/7 of displacement. That means it takes only 3,930 lb of force to hold the deck perfectly still against a 1-foot wave. Three reversible 7 kN mixers deliver ±4,720 lb — full cancellation up to ~2.4 ft waves, graceful partial cancellation beyond that, plus pitch and roll control from differential thrust at the three corners.

80–90%

heave & pitch reduction · typical protected anchorage (1–2 ft waves)

±5–15cm

residual heave left over, where uncontrolled motion would be ±0.5–0.6 m

<5%

motion-sickness incidence, vs 20–50% uncontrolled in trade-wind chop

~17%

of displacement as vertical thrust — 2–3× more than most active-heave systems

How much could it reduce motion?

80–90% of heave and pitch in the 1–2 ft seas where most Caribbean anchorages live; 60–75% in 3–4 ft seas. It also kills the resonant amplification near your 3.7 s natural period, which is where passive designs suffer most.

How much motion remains?

±5–15 cm of heave in a normal anchorage, with vertical accelerations of 0.05–0.12 m/s² RMS — a slow, small-amplitude breathe rather than a bob. In 1 m seas, roughly ±15–22 cm remains.

How comfortable is that?

Comparable to a large expedition yacht at anchor or a docked floating home: sleep uninterrupted, cooking and desk work normal, coffee stays in the cup. Sensitive guests may notice a gentle sway; nobody plans their walk.

Is the curved track practical?

Feasible — with seven specific design requirements (below). A rack-and-pinion electric tilt is the more robust moving option; a fixed 45° mount is the strongest value if propulsion stays on your rim drives.

02 · Interactive model

Watch it work

A single-degree-of-freedom heave model built from your numbers: waterplane stiffness 3,930 lb/ft, effective heave mass ≈ 20 t (with plates), plate damping ζ≈0.28, 0.35 s prop response lag, feed-forward + feedback control. Flip the system on and off; drag the wave height; try wave periods near and far from the 3.7 s natural period.

Live simulation Heading into waves · deep water · vertical scale ×2.5 MODE: FULL

Stabilization mode

Significant wave height H_s

0.60 m · 2.0 ft

Wave period T

Your heave natural period is ≈3.7 s. Try 4 s (resonance) vs 7 s — resonance is the whole story for an uncontrolled SWATH-style hull.

Residual heave

— m

± — in

Vert. accel RMS

— m/s²

— g

Motion reduction

— %

vs uncontrolled ghost

Thrust in use (peak)

— %

Elec. power draw

— kW

— kWh/day

Comfort

—

Heave displacement trace · last 25 s stabilized uncontrolled ghost

The orange ghost integrates the identical hull with the mixers permanently off, so the reduction figure is a like-for-like comparison in the exact same seaway. Model is concept-grade (±30%): quasi-static wave forcing, linearized plate damping. Directional control (pitch/roll) is handled separately in §04.

03 · Why it works

Two facts do all the heavy lifting

Fact 1 — your hull is resonant

Heave stiffness comes only from the waterplane: 1 ft of sink adds 1/7 of displacement, so k = 3,930 lb/ft (57.4 kN/m). With ~20 t of effective heave mass (structure + added mass of the foils and your bolt-on heave plates), the natural period lands at ≈ 3.6–4.0 s.

That sits directly inside the Caribbean trade-wind wave band. An uncontrolled hull doesn't just follow these waves — it amplifies them near resonance. The heave plates buy damping, but a resonant soft ride stays a resonant soft ride.

Fact 2 — holding still is cheap

To pin the deck at a fixed level while the sea surface moves ±Δ, the mixers must absorb exactly the buoyancy change: F = k × Δ. For a wave of height H that's a peak force of k·H/2 — which equals your full 21 kN authority at H ≈ 0.73 m (2.4 ft). Below that: total cancellation. Above it: proportional, graceful degradation. No cliff.

Force budget: what the sea demands vs what you have

Quasi-static heave force demand k·H/2 at each wave height, against available mixer authority.

H = 0.3 m (1 ft)

8.6 kN · covered 2.4×

H = 0.6 m (2 ft)

17.2 kN · covered 1.2×

H = 1.0 m (3.3 ft)

28.7 kN · 73% cancelled

H = 1.5 m (5 ft)

43.1 kN · 49% cancelled

H = 2.0 m (6.5 ft)

57.4 kN · 37% cancelled

Most of the thrust is reactiveBecause the platform barely moves, the mixers behave like an active spring: force oscillates at wave frequency but does almost no net work. You pay for propeller losses, not for lifting the seastead. That's why the power bill (§06) is far smaller than "hovering on thrusters" intuition suggests.

Is the response fast enough?

Waves run 3–7 s period; a VFD-driven 2.5 m prop develops thrust with a ~0.5–1 s time constant, and can reverse in roughly the same. Pure feedback would be marginal — but with a wave sensor on the forward leg giving 4–8 s of warning, feed-forward control makes the lag nearly irrelevant. This is exactly how active-heave-compensation cranes and dynamic positioning already operate.

The resonance overlap, drawn

Caribbean trade-wind waves · T ≈ 3–6 s Your Tn ≈ 3.7 s 2s345678910s

Where the cyan band and orange band overlap, passive motion is amplified ~2–3× by the Q of the system. The mixers' first job is deleting that peak; their second job is cancelling the remaining forced motion.

04 · Predicted motion & comfort

What you would actually feel

Heading into the waves, moored with your tension-leg helical screws, deep water. Values are ± single-amplitude heave; concept-grade ±30%. Uncontrolled figures assume your heave plates fitted (they cap the resonant response); the mixers remove what's left.

H_s · Setting	Uncontrolled heave	Feel, uncontrolled	Stabilized heave	Reduction	Feel, stabilized
0.3 m (1 ft)glassy morning, light chop	±0.25–0.30 m	Restless if the period is near 4 s — resonance makes even 1 ft chop annoying	±0.03–0.05 m	85–90%	Effectively still. Reading, fine work, standing showers all normal.
0.6 m (2 ft)typical trade-wind anchorage	±0.45–0.60 m	Classic anchored-catamaran night: clinking galley, interrupted sleep	±0.06–0.12 m	80–85%	A slow, faint breathe. Sleep unaffected; seasickness essentially absent.
1.0 m (3.3 ft)lively, first-reef conditions	±0.8–1.0 m	Miserable to live aboard; MSI 30–60% over hours	±0.15–0.22 m	70–75%	Noticeable gentle motion, zero restrictions on daily life. Force saturation begins here.
1.5 m (5 ft)exposed bay, brisk trades	±1.1–1.4 m	Not livable at anchor	±0.35–0.45 m	60–65%	Like a calm day on a regular boat — fine by day, still better than any passive option.
2.0 m (6.5 ft)wrong side of the island	±1.5–1.9 m	Deck wet, loads on mooring, everything stowed	±0.55–0.70 m	~55%	System saturated; the correct move is to relocate to the lee. Mixers buy you comfort, not weather immunity.

Comfort, on the ISO 2631 scale

Human comfort at sea is driven by vertical acceleration RMS, not amplitude. Uncontrolled resonance in 2 ft chop produces ~0.4–0.7 m/s² RMS — deep in the "uncomfortable, seasick within hours" bands. Stabilized, you sit at 0.05–0.15 m/s² RMS in the same water: the bottom two bands, where long-term habitation is comfortable.

< 0.10 comfortable · sleep OK
← stabilized, typical

0.10–0.20 slight motion
most activities fine

0.20–0.40 uncomfortable over hours
stabilized, 1 m+ seas

0.40–0.63 quite uncomfortable
uncontrolled lives here

> 0.63 severe · MSI climbs fast

Pitch & roll authority

The mixers sit at the triangle's corners, 25.4 ft from the centroid. Running one corner against the other two gives:

max pitch moment ≈ 7 kN × 7.74 m + 2 × 7 kN × 3.87 m ≈ 108 kN·m
max roll moment ≈ 2 × 7 kN × 6.7 m ≈ 94 kN·m

Wave pitch demand in 1 m / 5 s head seas is on the order of 30–60 kN·m — so pitch control carries ~1.5–2× margin up to about H_s = 1 m, after which the controller sensibly prioritizes heave. Expect residual pitch < 1° in typical conditions; heading into the waves, as you already plan, halves the excitation before the controller even starts. The same differential thrust can actively trim the platform during transit.

Mooring stays safeYour 3 ft tension-leg preload stores ≈ 3 ft × 3,930 lb/ft ≈ 11,800 lb of pretension — more than double the mixers' full upward pull of 4,720 lb. Lines never go slack; the screws just see a smaller, smoother cyclic load, which is good for fatigue life.

05 · Energy budget

Use it as the sun allows — your instinct is right

A 2.5 m prop at 7 kN is lightly loaded (≈1,425 N/m² disc loading, ~50–80 rpm): quiet, efficient, fish-friendly. Actuator-disc ideal power at full thrust is 5.8 kW; with propeller, motor and VFD losses, budget ≈11 kW electrical per mixer at full, ~33 kW all three — a peak you'll rarely hold for long.

Typical draws, because thrust oscillates around zero while the deck stands still:

Mode	Authority	Avg. draw	Energy / day	Result
Comfort	100%	6–12 kW	140–280 kWh	Full flat deck in 1–2 ft chop
Eco	60%	2.5–5 kW	60–120 kWh	~60–70% reduction; duty-cycles through wave groups
Night / cloudy	Eco	~3 kW	~75 kWh	≈ 20% of your battery bank — fine for one night

Your roof carries roughly 60–70 m² of solar (the 44 ft triangle interior) — about 13–15 kWp, worth 60–85 kWh/day in Caribbean sun. Read honestly:

Eco mode is roughly solar-neutral — run it 24/7 on good days with little bank draw.
Comfort mode is a battery conversation — your 25%-of-displacement LFP bank (≈3.1 t ≈ 350–400 kWh) covers 1.5–3 days of it, recharged over sunny days.
The controller should tier automatically: battery SoC > 80% and sun up → Comfort; SoC < 50% or after sundown → Eco; SoC < 30% → Off. Nothing about the hull needs the mixers to be safe — they are pure comfort, so shedding them costs nothing but motion.

Free redundancy storyEach mixer hangs off its own leg's battery/inverter bus — the same triple-redundant split you already planned for the rim drives. One dead leg and you keep two stabilizers (66% authority, full heave, degraded pitch/roll). In transit, the same hardware is your emergency propulsion: ~5–6 kn on mixers alone if all six rim drives ever quit.

06 · Mounting study

The curved track — verdict: feasible, with caveats

Putting the same mixer on a 90° arc — vertical for stabilization, horizontal for propulsion, any angle between for blended station-keeping — is mechanically sound. One geometry trick makes it clean: keep the propeller thrust line passing through the center of the arc's radius. Then the track carries the full 7 kN as a radial load at every angle, and the winch only ever fights friction and carriage weight. Seven requirements make it real:

React propeller torque. At ~8–9 kW delivered the prop reacts ≈1.8–2.5 kN·m on the carriage. A single rail would spin it — use twin rails or a keyed center guide.
End stops take full thrust. In both terminal positions the 7 kN pushes the carriage into a hard stop — design stops and their welds for ±7 kN at 10⁷ cycles fatigue, plus a storm-factor of 2.
Storm cradle & lashing. A 2.5 m propeller dangling under a leg in any seaway is fragile. Every mode change ends in a positive cradle; transit = cradled and lashed, or prop removed.
Assume Caribbean fouling. Rails grow algae and barnacles in weeks. Favor wiper-sealed wheel bogies, hard-anodized or 2205 duplex running surfaces with UHMW pads, and a brush the divers can pass over in 10 minutes.
Materials & CP. 316L minimum, 2205 preferred, isolate dissimilar metals, and bond the track into each leg's cathodic protection zone.
Cable without a through-hull. Festoon/energy-chain or spiral cable from the leg conduit to the carriage — consistent with your no-through-hulls rule; VFD stays dry, up top.
Position feedback + auto-hold. A simple limit/encoder pair and a self-braking winch; if power dies mid-arc the carriage must stay put, not slide to the bottom of the curve.

Two alternatives worth pricing against it

Option B · moving mount

Electric tilt — curved rack & pinion

Same 90° arc, but a gearmotor on a curved rack drives the carriage — positive drive both ways, holds any intermediate angle under thrust with no winch brake load.
No cable dragging through the water along the track.
One more submerged gearbox to maintain; costs more than winch + cable.

Option C · strongest value

Fixed 45° mounts — zero moving parts

Tilt each mixer 45°, azimuth them 120° apart pointing inward (or outward). Horizontal components cancel by symmetry; vertical authority is cos45° → 14.9 kN total (71%), still force-matched to ~1.7 ft waves.
Bonus: like a tilted-rotor drone, biasing the three thrusts unequally creates a net horizontal force anywhere in the ±45° cone — ~5–15 kN of slow surge/sway station-keeping. The wind on your 44′×7′ wall is only ~1.2 kN at 15 kn and ~3.5 kN at 25 kn: easily covered, so the boat stops snubbing on the mooring.
Can't serve transit propulsion — but your six rim drives already do that.

Option A · your proposal

Curved track + winch

Cheapest bill of materials; reuses proven winch gear; genuinely delivers propulsion and stabilization from one machine.
Any intermediate angle available for blended thrust.
Most fouling-sensitive of the three; winch cable lives in seawater; needs the full seven-item checklist above.

RecommendationIf you want the mixers to double as cruising propulsion, build the tilting mount — and consider the rack-and-pinion drive over cable-and-winch for holding torque at intermediate angles. If propulsion stays on the rim drives (it's a fine system), the fixed 45° mount is the best dollar: nothing underwater moves, nothing fouls into failure, and you keep essentially everything you asked the stabilizer to do. Either way, mount the mixer ~0.8–1 prop-diameter clear of the leg bottom: expect 10–20% thrust deduction from jet/hull interaction, and verify the final strut shape with CFD or a tow-tank run.

07 · Control system

The brain is the cheap part

Sense → predict → cancel

Sense. Marine IMU at the center of gravity (heave/pitch/roll rates), GNSS, wind vane — plus a wave-elevation sensor on the forward leg (pressure transducer or upward-looking ranging sensor).
Predict. Waves arrive at the boat with their shape visible 4–8 s ahead. A short prediction filter turns measured elevation into an anticipated force profile F(t) = k·ζ(t).
Cancel + damp. Feed-forward commands −F(t) to the mixer trio; feedback (LQR or well-tuned PID) mops up model error and adds active damping; a 3×3 allocation matrix splits heave/pitch/roll commands into per-mixer setpoints, saturating gracefully with heave priority.
Fail safe. Any sensor dropout, comms loss or watchdog trip ramps thrust to zero over ~5 s. The hull's passive behavior — with heave plates — is always safe; the mixers only ever add comfort, never safety-critical function.

Numbers the loop must hit

control rate ............ 20 Hz
sensor→command latency . < 200 ms
prop thrust time const . 0.5–1.0 s (handled by prediction)
thrust slew ............ ±14 kN/s achievable w/ VFD torque control
wave band covered ...... 0.14–0.35 Hz (T = 3–7 s)

Plays well with your other ideas

Tension-leg parking: the controller sees the mooring as extra stiffness and just works less. It can also gently trim mooring tension seasonally.

Two connected seasteads: this is where it gets genuinely lovely. Share wave sensor data between the two boats, phase-lock the controllers, and the walkway between them sits inside a doubly-calmed zone. The same coordinated thrust that flattens each deck flattens the relative motion the walkway cares about — your two-computer scheme and this system are the same software problem.

Propulsion mode: thrust reversal by motor reversal is fine at harbor speeds; if you want crisp wave-canceling, specify controllable-pitch or accept the ~1 s reversal lag and let prediction absorb it.

08 · Caveats & path forward

What to verify before cutting steel

Frequency-domain seakeeping model of the actual geometry (strip theory / boundary-element: Capytaine, Aqwa, or similar). Firm up excitation forces, plate damping, and RAOs; everything on this page is ±30% until then.
CFD on the jet–leg interaction (or a tow-tank afternoon): quantify thrust deduction with the mixer this close to the leg bottom and heave plates; shape the mount strut from the answer.
1:8–1:10 scale rig: three small reversible props under a ballasted triangle, in a wave tank or a calm-water pond with a wavemaker board. Validate the control law and measure residual motion directly. This is a weekend-scale experiment and it de-risks everything.
Mount fatigue review: ±7 kN, 10⁷ cycles at whichever mount wins (track, rack-pinion, or fixed 45°).
Sea trials: flat water tuning → trade-wind anchorage → 1 m seaway. Bring a phone on the deck and a phone on the dock; the accelerometer comparison will sell the system better than any report.

Keep these margins honest

Ventilation: keep the propeller ≥1.5–2 m below the waterline at the lowest heave excursion. At your draft that's comfortably satisfied; re-check if you ever lighten the boat.
Weight: mixers + mounts + tracks run ~1.5–2.5 t total. Audit it against the "humans and their stuff" margin inside 27,500 lb before committing — and put the hardware mass where the batteries live, low.
Thrust deduction: assume 15% lost to hull interaction until proven otherwise.
Saturation etiquette: above H_s ≈ 1.5 m the controller should announce it's power-limited and suggest relocating, rather than silently giving 55% and draining the bank.
Marine life: ducted props at ≤8–10 m/s tip speed are quiet and slow — good. Add a guard ring anyway; it doubles as a docking fender for the gear.