Architecture comparison and scaling laws

A synthesis of the architecture debate: which shape wins, under which assumptions.

Comparison table

Architecture Minimum size Stability Mass efficiency (thick shielding) Assembly Docking Best fit
Tethered bola smallest mass, long cable poor without compression member good simple deploy, hard operations at hub only early demos
Rigid barbell / stick ~67 m³ Twark_Main, 2022-10-30 needs ≥3 arms or in-plane mass Twark_Main, 2024-09-06 best (spheres) modules at two ends, one connecting rod Twark_Main, 2024-07-29 axis or off-axis testbeds, variable-g, interplanetary
Wheel / torus ~3100 m³ at 1 g Twark_Main, 2022-10-30 inherently stable poor for thin tori, ~2.7× shielding Twark_Main, 2024-07-05 repeat identical modules mikelepage, 2024-07-23 hub, gantry work stations, hotels, LEO
Drum / cylinder tens of m radius only if L < 1.3 R lamontagne, 2026-09-19 good in-space fabrication mikelepage, 2023-03-09 axis settlements
Shell + internal centrifuges large L = 0, excellent best in-space fabrication anywhere on shell large stations
Starship-based 1–4 ships depends on configuration poor (dead mass) Coastal Ron, 2022-12-31 none to minimal axis / tail early tests

Scaling laws the thread relied on

  • Shielding, pressure hull and MMOD all scale with surface area. Volume ("performance") ÷ area ("cost") favours compact shapes Twark_Main, 2023-02-21; Paul451, 2023-06-05. Effective diameter D_eff = 6V/A Twark_Main, 2023-09-28.
  • Interior mass often dominates. At ISS-like interior density (100–150 kg/m³), outfitting outweighs the pressure structure. Shielding dominates only for small D_eff or very thick shields Twark_Main, 2023-09-28. Typical ISS density is ~100–150 kg/m³ Twark_Main, 2023-02-04; Twark_Main, 2023-10-02.
  • Rotation load from interior mass in a torus: P_t = ρ·a·R. It grows with major radius, so large tori carry more "weight" than air-pressure load Twark_Main, 2023-09-29; Twark_Main, 2023-09-29.
  • A thick shield on a torus of minor radius r scales as ((r+t)² − r²)/r², so a larger minor radius makes shielding cheaper Twark_Main, 2023-09-30.
  • Above ~150 m radius, mass efficiency stops being decisive and usability and manufacturability dominate. Early stations are limited by Earth payload volume Paul451, 2023-09-17.
  • Modularity beats geometry when the difference is under an order of magnitude mikelepage, 2023-03-09.
  • Figure of merit: usable volume at a given g per unit moment of inertia (kg·m²) mikelepage, 2022-09-28.
  • Many small stations or one big one? "Better 100 stations × 100 people than one 10,000-person station": mass production and redundancy mikelepage, 2022-10-12. Paul451 counters that maintenance labour may scale sub-linearly (ISS: 4 person-days of maintenance per 2 of research on a 6-person crew), favouring larger stations Paul451, 2024-01-30.

Positions at the end of the thread (2026)

  • Twark_Main: double barbell, ideally 100+ m radius, cellular internal pressure-isolation structure, 20–30 cm polyethylene + boron shielding. For very large volumes, a non-rotating sphere with internal centrifuges (the "Pokeball") Twark_Main, 2023-09-16; Twark_Main, 2026-09-20. "Tumblers for small stations, Death Stars for large ones" Twark_Main, 2025-04-13.
  • mikelepage: modular Mars-g torus assembled from Starship modules in LEO, with a zero-g cage core mikelepage, 2024-07-23. Bolas for deep space and asteroids mikelepage, 2023-09-15.
  • Coastal Ron: 100 m Mars-g wheel of 8 m tubes, paired with a zero-G depot station, for asteroid-belt refuge Coastal Ron, 2026-09-18.
  • lamontagne and BIS: Island Zero (variable-g research) → Kalpana II-type drums → Avalon lamontagne, 2026-09-18.
  • Twark_Main's meta-point: the goal should be "an algorithm for designing the best station from requirements, instead of the One True Station Design" Twark_Main, 2025-09-20. Before him, Paul451: "form follows function" Twark_Main, 2022-10-30.

Open questions

  • Resolved only conditionally: barbell versus torus depends on shielding thickness (deep space vs LEO), assembly cost per unit and circulation needs.
  • The minimum viable product for each use case (see use cases).

Sources

Forum posts

  • mikelepage, 2022-09-28 (reply #1797, 1 like): Figure of merit: volume at g per MoI. Post.
  • mikelepage, 2022-10-12 (reply #2023, 1 like): 100 × 100 better than 1 × 10,000. Post.
  • Twark_Main, 2022-10-30 (reply #2257): What size does the station want to be. Post. Attachments: solid_black_line_is_the_one_to_look_at.jpeg.
  • Twark_Main, 2022-10-30 (reply #2258): Min barbell vs torus volume. Post.
  • Coastal Ron, 2022-12-31 (reply #2745): Dead mass ratio. Post.
  • Twark_Main, 2023-02-04 (reply #3237): ISS density ~100 kg/m³; parallel axis theorem. Post.
  • Twark_Main, 2023-02-21 (reply #3381): Volume = performance, area = cost. Post.
  • mikelepage, 2023-03-09 (reply #3537): Spheres/cylinders need in-space fab. Post.
  • Paul451, 2023-06-05 (reply #3784, 3 likes): Shielding, pressure, MMOD scale with area. Post. Attachments: Modules as dumbbell.jpg.
  • mikelepage, 2023-09-15 (reply #3943): Asteroid-pit bola. Post.
  • Twark_Main, 2023-09-16 (reply #3949): Twark_Main's optimal design recap. Post.
  • Paul451, 2023-09-17 (reply #3956): Above ~150 m mass efficiency not decisive (Paul451). Post.
  • Twark_Main, 2023-09-28 (reply #4037): D_eff = 6V/A; interior mass dominates at ISS density. Post.
  • Twark_Main, 2023-09-29 (reply #4080): Weight exceeds pressure in modest tori. Post.
  • Twark_Main, 2023-09-29 (reply #4083): P_t = ρaR. Post.
  • Twark_Main, 2023-09-30 (reply #4085): Shield mass factor; integrated tension grid. Post.
  • Twark_Main, 2023-10-02 (reply #4098): ISS >150 kg/m³. Post.
  • Paul451, 2024-01-30 (reply #4437, 1 like): Maintenance scaling favours larger stations. Post.
  • Twark_Main, 2024-07-05 (reply #4646): Torus 2.7× shielding. Post.
  • mikelepage, 2024-07-23 (reply #4656): Torus assembly by berthing. Post.
  • Twark_Main, 2024-07-29 (reply #4666): Bola: one connecting rod vs dozens. Post.
  • Twark_Main, 2024-09-06 (reply #4783, 3 likes): 3-arm stability criterion. Post. Attachments: moment_of_inertia_ratio_by_arm_angle.png.
  • Twark_Main, 2025-04-13 (reply #4859, 1 like): Design principles (1 like). Post.
  • Twark_Main, 2025-09-20 (reply #4997): Design algorithm from requirements. Post.
  • Coastal Ron, 2026-09-18 (reply #5140): 8 m tubes built on Earth = truly near-term. Post.
  • lamontagne, 2026-09-18 (reply #5142, 6 likes): BIS OSP roadmap (6 likes). Post.
  • lamontagne, 2026-09-19 (reply #5159, 2 likes): Globus L < 1.3 R. Post. Attachments: 2007KalpanaOne (1).pdf.
  • Twark_Main, 2026-09-20 (reply #5160, 1 like): Pokeball. Post.