Weight of seawater displaced:
Seawater density ≈ 64.0 lb/ft³
Displacement = 716.7 × 64.0 = 45,871 lbs ≈ 45,900 lbs
Displacement = 20.8 metric tonnes ≈ 22.9 short tons
Total Displacement: approximately 45,900 lbs (20.8 tonnes)
This is the maximum load the platform can support before the legs submerge beyond the 2/3 waterline. Actual structural weight plus payload must stay safely below this figure, with margin for waves and dynamic loading. A typical design reserve would target using no more than 60–70% of maximum displacement for structure, leaving 30–40% for payload and safety margin.
Displacement Budget (Rough Estimate)
Item
Estimated Weight
3 Legs (aluminum version)
~5,400 lbs
3 Legs (duplex SS version)
~11,000 lbs
Triangle frame (60 ft side, aluminum)
~4,000–6,000 lbs
Pyramid superstructure + cladding
~3,000–5,000 lbs
Solar panels (~800 ft²)
~1,600 lbs
Batteries, systems, propulsion
~2,000–4,000 lbs
Cables (Dyneema + hardware)
~500–1,000 lbs
Furnishings, water, supplies
~2,000–4,000 lbs
Total (aluminum build)
~19,000–27,000 lbs
Remaining payload margin (aluminum)
~19,000–27,000 lbs
The aluminum build leaves a healthy margin. The duplex stainless build adds roughly 5,500 lbs to leg weight, tightening the budget but still workable. Either way, you have enough displacement for a small crew and supplies.
3. Material Comparison: Duplex Stainless Steel vs Marine Aluminum
We analyze two options for the 3 cylindrical legs. Each leg: 3.9 ft (46.8 in) diameter, 30 ft long, with dished (torispherical/ellipsoidal) end caps top and bottom.
Material Properties
Property
Duplex SS 2205
Marine Aluminum (5083-H321)
Density
0.283 lb/in³ (7,820 kg/m³)
0.096 lb/in³ (2,660 kg/m³)
Yield Strength
65,000 psi (450 MPa)
33,000 psi (228 MPa)
Tensile Strength
90,000 psi (620 MPa)
44,000 psi (305 MPa)
Modulus of Elasticity
29 × 10&sup6; psi
10.3 × 10&sup6; psi
Corrosion Resistance (seawater)
Excellent (PREN ≥ 35)
Very good (with proper alloy)
Fatigue (seawater)
Excellent
Good (lower fatigue limit)
Weldability
Requires skilled TIG; back-purge needed
Standard TIG/MIG; well understood
Raw plate cost (approx.)
$4–$7/lb
$3–$5/lb
Weight Calculations per Leg
Option 1: Duplex SS 2205
Cylinder wall: 0.25 in (1/4") thick
Circumference = π × 46.8 = 147.0 in
Wall area = 147.0 × 360 in = 52,920 in²
Volume of steel = 52,920 × 0.25 = 13,230 in³
Weight (wall) = 13,230 × 0.283 = 3,744 lbs
Two dished ends: 0.50 in (1/2") thick
Approx. surface area per cap ≈ 1.15 × π × (23.4)² ≈ 1,978 in²
Volume per cap = 1,978 × 0.50 = 989 in³
Weight (2 caps) = 2 × 989 × 0.283 = 560 lbs
3,744 + 560 = 4,304 lbs per leg
Total 3 legs: ~12,910 lbs (5.86 tonnes)
Option 2: Marine Aluminum 5083
Cylinder wall: 0.50 in (1/2") thick
Circumference = π × 46.8 = 147.0 in
Wall area = 147.0 × 360 in = 52,920 in²
Volume of aluminum = 52,920 × 0.50 = 26,460 in³
Weight (wall) = 26,460 × 0.096 = 2,540 lbs
Anodic protection recommended; inspect for pitting, especially at welds
Biofouling resistance
Slightly better
Needs antifouling coating
Impact resistance
Higher (stronger, tougher)
Good but dents more easily
Repairability at sea
Very difficult (specialized welding)
Easier (standard TIG welding)
Container shipping
Heavier but same dimensions
Lighter — easier logistics
Recommendation: Marine aluminum (5083-H321) is the more practical choice for this project. It saves ~4,150 lbs across all three legs, costs roughly half as much, is easier to fabricate and repair, and still offers 30–50 year life with reasonable maintenance. The duplex stainless steel is the "build it once for a century" option but at a significant cost and weight penalty. Given the tensegrity design philosophy and desire for container-shippable components, aluminum is strongly favored.
Note on thickness choices: The specified 1/4" duplex SS walls and 1/2" aluminum walls provide similar structural capacity at these dimensions because duplex SS has roughly double the yield strength of 5083 aluminum. Both specifications appear adequate for a vessel operating in open ocean at moderate depths and pressures, though a formal FEA analysis accounting for wave-induced bending at the joint and external hydrostatic pressure at full submergence depth should be performed before final construction.
4. Living Space Estimate
The superstructure is a triangular pyramid (tetrahedron) sitting on the 60 ft equilateral triangle frame, with the apex 25 ft above the base. We need to estimate usable floor area where ceiling height ≥ 7 ft.
Pyramid Geometry
Base: equilateral triangle, side = 60 ft
Base area = (√3/4) × 60² = 1,559 ft²
Apex height: 25 ft above base center
The pyramid tapers linearly. At height h above the base, the cross-sectional triangle has:
Side length = 60 × (1 − h/25) ft
Area at height h = 1,559 × (1 − h/25)² ft²
Floor Layout
Floor
Floor Height
Ceiling Height
Min Headroom
Side Length at Floor
Side Length at Ceiling
1st Floor
0 ft (base)
8 ft
8 ft (full)
60.0 ft
40.8 ft
2nd Floor
8 ft
16 ft
8 ft (full)
40.8 ft
21.6 ft
3rd Floor
16 ft
25 ft (apex)
9 ft (center max)
21.6 ft
0 ft (point)
Usable Area with ≥ 7 ft Headroom
For each floor, the usable area is the region where the sloping walls are at least 7 ft above the floor level. Because the pyramid walls slope inward, the perimeter of each floor loses some area near the edges where the ceiling is too low.
1st Floor (0–8 ft):
The ceiling at any point is defined by the pyramid surface. At the floor perimeter (side 60 ft), ceiling = 0+ ft. At the 7 ft headroom contour, we need the pyramid surface to be at height 7 ft, which means side = 60 × (1 − 7/25) = 43.2 ft.
Usable area = (√3/4) × 43.2² = 808 ft²
However, the floor itself extends to the full 60 ft base, so the area between the 43.2 ft contour and the 60 ft perimeter has less than 7 ft headroom. This outer zone (1,559 − 808 = 751 ft²) is usable for storage, built-in seating/beds, mechanical systems, etc. with reduced headroom.
2nd Floor (8–16 ft):
Floor side = 60 × (1 − 8/25) = 40.8 ft → floor area = (√3/4) × 40.8² = 722 ft²
At 7 ft above 2nd floor = height 15 ft: side = 60 × (1 − 15/25) = 24.0 ft
Usable area (≥7 ft headroom) = (√3/4) × 24.0² = 250 ft²
3rd Floor (16–25 ft):
Floor side = 60 × (1 − 16/25) = 21.6 ft → floor area = (√3/4) × 21.6² = 202 ft²
At 7 ft above 3rd floor = height 23 ft: side = 60 × (1 − 23/25) = 4.8 ft
Usable area (≥7 ft headroom) = (√3/4) × 4.8² = 10 ft²
The 3rd floor is essentially an attic/observatory with only a tiny patch of full standing room. Most of it (192 ft²) has 0–7 ft headroom, useful as a loft/sleeping area/storage.
Floor
Total Floor Area
Area with ≥7 ft Headroom
Reduced Headroom Zone
1st Floor
1,559 ft²
808 ft²
751 ft²
2nd Floor
722 ft²
250 ft²
472 ft²
3rd Floor
202 ft²
10 ft²
192 ft²
Totals
2,483 ft²
~1,068 ft²
~1,415 ft²
Usable living space with ≥ 7 ft headroom: approximately 1,070 ft²
The first floor provides the bulk of the full-height living space (~808 ft²). The second floor adds a good-sized room (~250 ft²). The reduced headroom areas add an additional ~1,415 ft² of useful space for sleeping lofts, storage, mechanical systems, workbenches, and built-in furniture. Total enclosed area across all floors is approximately 2,480 ft².
Solar panel area: The three sloping faces of the pyramid have a total surface area of approximately:
Each face ≈ slant height × base/2. Slant height from base edge midpoint to apex ≈ √(17.32² + 25²) ≈ 30.4 ft.
Area per face = (60 × 30.4)/2 = 912 ft². Total = 3 × 912 ≈ 2,736 ft².
At 80% coverage: ~2,189 ft² of solar panels.
At ~20 W/ft² (modern panels): ~43.8 kW peak capacity — a very generous power supply.
5. Ball Diameter Calculation (Option B)
In the revised design, each 30 ft cylinder is replaced by a 20 ft cylinder plus a sphere at the bottom. The sphere must have the same volume as the 10 ft of cylinder it replaces.
Volume of 10 ft of cylinder (diameter 3.9 ft):
Vcyl = π × (1.95)² × 10 = π × 3.8025 × 10 = 119.46 ft³
r³ = 119.46 / (4π/3) = 119.46 / 4.1888 = 28.52
r = 28.521/3 = 3.057 ft
Diameter of ball = 2 × 3.057 = 6.11 ft ≈ 6 ft 1 in
Required ball diameter: approximately 6.1 feet (1.86 m)
This is 1.57 times the cylinder diameter of 3.9 ft. The ball extends about 1.1 ft beyond the cylinder on each side.
Draft Comparison
Option A (straight 30 ft cylinders):
Legs at 45°, 20 ft submerged along the leg axis.
Vertical depth (draft) = 20 × sin(45°) = 20 × 0.707 = 14.1 ft
Option B (20 ft cylinder + 6.1 ft ball):
Total leg length along axis = 20 + 6.1 = 26.1 ft (shorter than 30 ft).
Submerged portion: We still need 2/3 of the volume submerged. The total volume is the same (238.9 ft³ per leg). At 2/3 submersion, we need 159.3 ft³ submerged per leg.
The ball volume = 119.5 ft³. So we need only 159.3 − 119.5 = 39.8 ft³ of cylinder submerged.
Cylinder submerged length = 39.8 / 11.95 = 3.33 ft of cylinder below waterline.
But the ball (diameter 6.1 ft) is below the cylinder.
Total submerged length along axis = 3.33 + 6.1 = 9.43 ft.
Vertical draft = (3.33 + 6.1) × sin(45°) = 9.43 × 0.707 = 6.67 ft
Wait — this doesn't seem right. Let me reconsider.
The 2/3 submersion refers to 2/3 of the leg length, not volume. Re-reading the design: "30 feet long and 2/3rds of the way in the water" — so 20 ft of the 30 ft length is submerged.
For Option B, the total leg length is 20 ft cylinder + 6.1 ft ball diameter = 26.1 ft total. If we keep the same waterline position (same depth of submersion), the cylinder has about 13.3 ft submerged (total axis length below water = 13.3 + 6.1 = 19.4 ft). Actually the submersion is set by weight balance, not a fixed ratio. Let's compute it properly:
The structure weighs the same and needs the same buoyancy (same total displacement). So the same 716.7 ft³ must be submerged. With Option B the ball is always fully submerged (it's at the bottom), and we solve for how much cylinder is submerged.
This makes sense — the ball replaces exactly 10 ft of cylinder volume, so 10 ft of cylinder is still submerged, plus the ball below it.
Total submerged length along axis = 10.0 + 6.1 = 16.1 ft
Vertical draft = 16.1 × sin(45°) = 16.1 × 0.707 = 11.4 ft
Draft reduction: 14.1 ft → 11.4 ft (saving 2.7 ft / 19%)
Parameter
Option A (Straight Cylinder)
Option B (Cylinder + Ball)
Total leg length (along axis)
30.0 ft
26.1 ft
Submerged length (along axis)
20.0 ft
16.1 ft
Draft (vertical)
14.1 ft
11.4 ft
Above-water leg length
10.0 ft
10.0 ft
Volume per leg
238.9 ft³
238.9 ft³ (same)
6. Drag & Speed Analysis
We estimate the drag of both configurations and compute cruising speed for 3,000 W and 4,000 W of shaft power delivered to the water.
Methodology
At the very low speeds involved (0.5–1.5 knots), wave-making resistance is negligible. Drag is dominated by viscous/form drag on the submerged leg structures. We use the standard drag equation:
Fdrag = ½ × ρ × CD × A × v²
Where:
ρ = seawater density = 1,025 kg/m³
CD = drag coefficient
A = projected frontal area of submerged components (perpendicular to flow)
v = velocity through water (m/s)
At equilibrium: Power = Fdrag × v = ½ × ρ × CD × A × v³
Solving: v = (2P / (ρ × CD × A))1/3
Frontal Area & Drag Coefficients
Important note on leg orientation: The legs extend at 45° outward from the triangle corners. When the seastead moves forward (toward the front corner), one leg is directly ahead (the front leg) and two are behind (port and starboard rear). The front leg's projected area and the rear legs' projected areas must all be considered. For simplicity, we assume the seastead moves forward with one corner leading.
Option A: Straight Cylinders
Each leg: 3.9 ft (1.19 m) diameter cylinder
Front leg (pointing forward into the flow at 45° down):
The flow hits the cylinder at an angle. Projected frontal area ≈ diameter × submerged length × sin(angle to flow).
The front leg extends forward and down at 45°. If moving straight ahead, flow hits it somewhat axially. Effective projected area ≈ 1.19 m × 6.10 m × sin(~45°) ≈ 5.13 m²
Actually, let's simplify: the cylinder is at 45° to horizontal. Water flows horizontally. The projected area perpendicular to flow is:
A = diameter × submerged_length × cos(45°) = 1.19 × 6.10 × 0.707 = 5.13 m² per leg
CD for a long cylinder in cross-flow (Re ~105 at these speeds) ≈ 1.0
3 legs total projected area:
Since the two rear legs are angled backward, they also present projected area to the flow, though partially shielded. For conservative estimate:
Total effective A ≈ 3 × 5.13 × 0.85 (slight shielding factor) = 13.1 m²
Effective CD × A = 1.0 × 13.1 = 13.1 m²
Option B: 20 ft Cylinder + 6.1 ft Ball
Cylinder portion submerged: 10.0 ft (3.05 m) at 3.9 ft (1.19 m) diameter
Projected area per leg (cylinder) = 1.19 × 3.05 × cos(45°) = 1.19 × 3.05 × 0.707 = 2.57 m²
Ball portion: 6.1 ft (1.86 m) diameter sphere
Projected area = π/4 × 1.86² = 2.72 m²
CD for a sphere (Re ~105) ≈ 0.47
Per leg effective CD×A:
Cylinder: 1.0 × 2.57 = 2.57
Sphere: 0.47 × 2.72 = 1.28
Total per leg: 2.57 + 1.28 = 3.85 m²
3 legs total (with shielding factor):
3 × 3.85 × 0.85 = 9.82 m²
Additional Drag: Frame & Cables
The triangle frame is above water — minimal hydrodynamic drag (some wind drag, neglected here).
Cables are submerged thin lines — small contribution.
Estimate cable drag: ~6 cables, ~40 ft submerged length, ~1 in diameter each.
Cable CD×A ≈ 6 × 1.2 × (0.025 m × 12.2 m × 0.707) ≈ 6 × 1.2 × 0.216 ≈ 1.55 m²
Total effective CD×A:
Option A: 13.1 + 1.55 = 14.65 m²
Option B: 9.82 + 1.55 = 11.37 m²
Speed Calculations
v = (2P / (ρ × CDA))1/3
Propeller efficiency assumed: ~50% for these submersible mixers used as thrusters (they are not optimized for thrust, but the banana blade design is reasonable at low speeds).
Effective shaft power to water: Peff = Pelectrical × 0.50
Option A: v = (12000 / 15,016)1/3 = (0.7991)1/3 = 0.928 m/s = 1.80 knots = 2.07 mph Option B: v = (12000 / 11,654)1/3 = (1.0297)1/3 = 1.010 m/s = 1.96 knots = 2.26 mph
Speed Summary Table
Power Input
Option A (Cylinders)
Option B (Cyl + Ball)
Speed Gain
3,000 W electrical
1.14 kts (1.31 mph)
1.24 kts (1.42 mph)
+8.6%
4,000 W electrical
1.25 kts (1.44 mph)
1.36 kts (1.57 mph)
+8.9%
12,000 W (all 4 full)
1.80 kts (2.07 mph)
1.96 kts (2.26 mph)
+9.0%
Key finding: At 3,000–4,000 watts, both designs achieve the target speed range of roughly 1.1–1.6 mph. Option B (cylinder + ball) provides approximately a 9% speed improvement at any given power level, or equivalently, achieves the same speed with about 22% less power. Both designs easily achieve the 0.5–1 mph design target with just 1,500–2,000 W.
Current & wind assist: Ocean gyres typically flow at 0.2–1.0 knots. With favorable currents and trade winds, effective ground speed could be 2–3 knots without using any propulsion power at all. The propellers are really for maneuvering and course corrections, which is a sound strategy.
Drag Force at Target Speeds
Speed
Option A Drag Force
Option B Drag Force
0.5 mph (0.22 m/s)
~370 N (83 lbf)
~287 N (65 lbf)
1.0 mph (0.45 m/s)
~1,502 N (338 lbf)
~1,165 N (262 lbf)
1.5 mph (0.67 m/s)
~3,355 N (754 lbf)
~2,603 N (585 lbf)
Thruster capacity check: Each 3,000 W mixer produces ~2,090 N thrust. Four units = 8,360 N total. At 50% efficiency = 4,180 N effective. This is well above the drag forces at target speeds, confirming the propulsion system is adequately sized. Even with only 2 operational thrusters (one per side for redundancy), you have ~4,180 N gross or ~2,090 N effective — still sufficient for 1+ mph in calm conditions.
7. Cost Estimates
Leg Cost Comparison: Option A vs Option B
The ball adds fabrication complexity. Spheres are more expensive to form than cylinders. For a 6.1 ft sphere in 1/2" aluminum or 1/4" duplex SS, the ball halves would likely be press-formed or spun, then welded at the equator and to the cylinder.
Weight per Leg
Component
Option A (30 ft Cylinder)
Option B (20 ft Cyl + Ball)
Duplex SS
Aluminum
Duplex SS
Aluminum
Cylinder shell
3,744 lbs
2,540 lbs
2,496 lbs (20 ft)
1,693 lbs (20 ft)
End caps (2)
560 lbs
380 lbs
280 lbs (1 top cap)
190 lbs (1 top cap)
Sphere shell
—
—
~930 lbs*
~635 lbs*
Transition ring/collar
—
—
~50 lbs
~35 lbs
Total per leg
4,304 lbs
2,920 lbs
~3,756 lbs
~2,553 lbs
Total 3 legs
12,912 lbs
8,760 lbs
~11,268 lbs
~7,659 lbs
* Sphere shell: surface area = π × 6.1² = 116.9 ft² = 16,833 in². Duplex SS at 1/4": 16,833 × 0.25 × 0.283 = 1,191 lbs, but reduced to ~930 lbs accounting for the cylinder junction opening. Aluminum at 1/2": 16,833 × 0.50 × 0.096 = 808 lbs, reduced to ~635 lbs.
Cost Estimates (Materials + Fabrication)
Configuration
Duplex SS 2205
Marine Aluminum 5083
Option A (30 ft straight cylinders × 3)
$105,000–$160,000
$51,000–$89,000
Option B (20 ft cyl + ball × 3)
$120,000–$185,000
$62,000–$108,000
Cost premium for Option B
+$15,000–$25,000
+$11,000–$19,000
The sphere fabrication adds roughly 15–20% to leg costs despite slightly less material weight, because spheres require forming on a press or spin-forming equipment, more complex welding, and tighter quality control. However, the balls could potentially be sourced as pre-made pressure vessel heads or tank components, which could reduce the premium.
Full System Cost Estimate
System
Estimated Cost Range
Legs (aluminum, Option A)
$51,000–$89,000
Legs (aluminum, Option B)
$62,000–$108,000
Triangle frame (aluminum, 60 ft sides)
$30,000–$55,000
Pyramid superstructure + cladding
$25,000–$50,000
Solar panels (~43 kW, with mounting)
$35,000–$55,000
Battery bank (100–200 kWh LiFePO4)
$25,000–$50,000
Propulsion (4 × mixers + spare + mounts)
$30,000–$45,000
Dyneema cables + hardware
$8,000–$15,000
Electrical systems, inverters, controls
$10,000–$20,000
Interior fit-out (basic)
$15,000–$35,000
Water maker, plumbing, marine toilet
$5,000–$12,000
Shipping & logistics
$10,000–$25,000
Assembly labor
$15,000–$30,000
TOTAL (aluminum, Option A)
$260,000–$480,000
TOTAL (aluminum, Option B)
$270,000–$500,000
8. Design Tradeoffs & Additional Analysis
Option A vs Option B: Comprehensive Comparison
Factor
Option A (Straight Cylinders)
Option B (Cylinder + Ball)
Winner
Speed at 3 kW
1.31 mph
1.42 mph
B
Speed at 4 kW
1.44 mph
1.57 mph
B
Draft
14.1 ft
11.4 ft
B
Fabrication simplicity
Simple cylinders
Sphere adds complexity
A
Cost (aluminum)
$51k–$89k legs
$62k–$108k legs
A
Weight (aluminum)
8,760 lbs
7,659 lbs
B
Container shipping
30 ft cylinders may need splitting
26 ft cylinders + balls — balls are 6 ft wide
Tie
Heave damping
Good (long cylinder)
Better (ball adds mass moment)
B
Roll/pitch stability
Good
Better (lower CG, more spread mass)
B
Seakeeping
Good
Better (smaller waterplane area ratio)
B
Biofouling surface area
~467 ft² submerged
~468 ft² submerged
Tie
Inspection/maintenance
Simpler geometry
Ball harder to inspect underneath
A
Repairability
Easy — flat cylinder patches
Sphere harder to patch
A
Seakeeping Analysis
Why Option B has better seakeeping:
Reduced waterplane area: Both options have the same cylinder diameter at the waterline (3.9 ft), but Option B has more of its buoyancy volume deep below the surface in the ball. This means less of the buoyancy is near the waterline, which reduces the vessel's response to passing waves (the "SWATH effect").
Heave natural period: A smaller waterplane area relative to displacement increases the natural heave period, pushing it further from typical ocean wave periods (6–12 seconds). This reduces resonant heave response.
Added mass: The sphere has a larger added mass coefficient than a cylinder of equivalent volume, providing additional inertial resistance to wave-induced motion.
Lower center of buoyancy: The ball concentrates buoyancy deeper, providing a stronger restoring moment in roll and pitch.
Waterplane Area Comparison
Waterplane area per leg = π/4 × 3.9² = 11.95 ft²
Total waterplane area = 3 × 11.95 = 35.8 ft²
This is the same for both options (same cylinder diameter at waterline).
For comparison, a typical 45 ft monohull sailboat has a waterplane area of ~200–300 ft².
Your design has only 12–18% of that, which is why the ride will be dramatically smoother.
Heave Natural Period Estimate
Theave = 2π × √(m / (ρg × Awp))
m = displacement mass = 20,800 kg
ρg = 1,025 × 9.81 = 10,055 N/m³
Awp = 35.8 ft² = 3.33 m²
Caution: A heave natural period of ~5 seconds falls within the range of common ocean wave periods (4–12 seconds). This means there could be resonant heave response in certain sea states. The small waterplane area limits the forcing, but adding heave plates (horizontal discs at the bottom of the legs) would add damping and increase the effective mass, pushing the natural period higher. This is highly recommended.
Heave Plate Recommendation
Adding a horizontal circular plate (heave plate) at the bottom of each leg — or around the equator of each ball in Option B — dramatically increases heave damping and added mass. A plate of diameter ~8–10 ft at each leg bottom would:
Increase the effective mass by 2–3× for vertical motion
Push the heave natural period to 8–12 seconds
Add significant viscous damping
Cost very little (simple flat plates)
In Option B, the ball itself partially serves this function
Tensegrity Cable Loads
Buoyancy force per leg: (at full 2/3 submersion)
Fbuoy = 238.9 ft³ × 64 lb/ft³ = 15,290 lbs per leg
Weight of leg (aluminum): ~2,920 lbs (Option A) Net upward force per leg: ~12,370 lbs
The leg is at 45°. Buoyancy acts vertically upward. The leg is held by two cables running from the leg bottom to the two opposite corners of the triangle.
Each cable carries a significant load. With two cables per leg at various angles, cable tensions will be in the range of 8,000–15,000 lbs depending on exact geometry and dynamic loading.
Cable recommendation: 16mm (5/8") jacketed Dyneema for the primary tensegrity cables provides an excellent safety factor. The backup loop cable could be 12mm. All Dyneema connections should use proper thimbles and splices (not knots), and cables should be inspected regularly for chafe and UV degradation (the jacket helps greatly with UV).
Propulsion Notes
The 2,500 mm banana blade submersible mixers are a creative and cost-effective choice. At $5,000–$8,000 each (salt water rated), four units plus a spare totals $25,000–$40,000.
Mount them in protective cages to prevent entanglement with lines, kelp, or marine life.
Consider adding kort nozzles (ducts) around the propellers to increase thrust efficiency by 20–30% at low speed — this could bring the effective efficiency from ~50% to ~65%.
Differential thrust for steering is excellent with the wide spacing (~60 ft between port and starboard legs). Even a 10% thrust differential will produce strong turning moments.
The front leg has no propulsion — this is fine. The vessel will turn around its center of resistance, which is near the centroid of the submerged legs.
Final Recommendation
Overall recommendation: Option B (cylinder + ball) in marine aluminum 5083.
Material: Marine aluminum saves ~4,150 lbs vs duplex SS (Option A) and ~$50,000–$70,000. It's easier to fabricate, repair, and ship. Life expectancy of 30–50 years is well suited to a vessel that will evolve over time.
Configuration: Option B costs ~$11,000–$19,000 more than Option A in aluminum, but provides:
9% better speed efficiency (or 22% power savings)
2.7 ft less draft (important for coastal approach and anchoring)
Better seakeeping (more SWATH-like behavior)
1,100 lbs weight savings
Improved heave resistance
The modest cost premium pays for itself in operational benefits. Add heave plates to the balls for best motion performance.
Risk Items to Address
Heave resonance: Add heave plates or damping fins. Natural period of ~5 seconds without them is a concern.
Cable chafe: Dyneema is strong but vulnerable to chafe. All contact points need proper fairleads, thimbles, and chafe protection.
Tensegrity joint design: The "flexible" joints need to accommodate small angular movements without fatigue failure. Universal joints or elastomeric bearings are typical solutions.
Container shipping: The 6.1 ft ball diameter exceeds standard container width (7.7 ft interior). It will fit inside a standard container but with minimal clearance. Alternatively, ship the ball in two hemispheres and weld on site.
Lightning protection: The metal pyramid is an excellent lightning rod. Ensure proper bonding and grounding to the sea.
Galvanic corrosion: If mixing metals (aluminum structure, stainless fasteners), use proper isolation and zinc anodes.
All estimates are preliminary engineering approximations. Final design should include professional naval architecture review, FEA structural analysis, and classification society consultation.