Compromise between the orbital station's radius, crew comfort, and structure mass
Artificial gravity on a rotating station is not a choice of 'turn on 1g', but an engineering compromise with physiology. A smaller radius means less structure, less mass to launch and fewer welds – but to achieve the same acceleration requires the station to spin much faster, which the human vestibular apparatus perceives very poorly.
The centripetal acceleration on the ring of a rotating station is defined by a simple formula: a = ω²·r, where ω is the angular velocity (rad/s), and r is the radius of rotation. Rewriting through rotations per minute (RPM): RPM = 60/(2π) · √(a/r).
This formula dictates a hard compromise: to achieve a specified acceleration (e.g., 9.8 m/s² for 1g), at small r, a large ω is needed—meaning the station must rotate quickly. A radius of 25 m for full 1g requires about 6 rpm; a radius of 250 m for the same 1g only needs about 1.9 rpm.
This means that a cheaper, more compact station (less structure, less mass in orbit) necessarily spins faster — and spin speed is precisely the parameter that most agitates the human vestibular apparatus.
Doubling the radius reduces the required number of rotations per minute by only √2 ≈ 1.41 times for the same g-level — so the gain in comfort from increasing the radius is slowed down, while the loss in construction mass accelerates.
Classical studies of rotating rooms (Graybiel et al., 1960–70s, NASA/Naval Aerospace Medical Institute) showed that unadapted individuals start to experience boredom and illusions of head tilt (cross-capling effects) at 3 RPM during rapid head movements. At speeds over 6 RPM, discomfort is felt by practically all test subjects, even with careful movements.
The main reason is the Coriolis force, which arises when a person moves their hand, foot, or head inside a rotating system. This force 'pushes' motion off its expected trajectory, and the vestibular apparatus (semicircular canals) registers an opposing signal that conflicts with visual perception — the classic cause of space motion sickness.
The second small radius effect is the gravitational gradient: the difference in 'weight' between an astronaut's head and feet. At a radius of 25 meters, the gradient can exceed 10-15% of the orbital g-level, while at 250 meters it falls below 1%.
At the other end of the spectrum are giant stations like the O'Neill cylinder or Stanford Torus proposed by O'Neill. Rotating slowly, they practically cannot be felt as 'rotation' — the vestibular apparatus receives a clean, stable signal of weight. The cost is a structure that grows much faster than proportionally to radius.
At 1 RPM (typical for large concepts like the Stanford Torus or O'Neill Cylinder), angular velocity is so low that Coriolis forces from typical human movements (walking, raising an arm) become much smaller than the gravity felt by the body. Studies suggest that at ≤2 RPM most people do not experience significant discomfort even without prior adaptation.
The head-to-feet gravity gradient also becomes negligible: with a radius in the hundreds of meters, the difference in 'weight' over a person's height (≈1.7 m) is only a fraction of a percent. In fact, at large radii, artificial gravity becomes almost indistinguishable from natural.
But physics of pressure and structural strength work against large radii. A sealed torus or cylinder must withstand internal atmospheric pressure (like a boiler), and according to Barlow's law, wall thickness for a given material stress grows proportionally to the radius of the cross-section. The perimeter length increases linearly with radius, surface area quadratically, and combined with radiation shielding requirements (a meter-thick layer of regolith or water around living quarters), mass grows approximately as r^2.2-2.5 based on empirical NASA-conceived concepts from the 1970s.
This means that doubling the radius—along with only about 1.4 times improvement in comfort criterion by RPM—can increase the required construction mass by 4–6 times. Such nonlinear cost growth is the main restraining factor for large rotating stations.
The O'Neill cylinder (Project 'Island Three', 1976) at a scale of 3200 m radius would have required millions of tons of structural material according to estimates — realistic only if raw materials are mined on the Moon or asteroids, not brought from Earth.
Regardless of the physiologically optimal radius, practical design is limited by what can be delivered into orbit and assembled there within a reasonable number of launches. The volume of the main rotating cylinder, payload capacity, and complexity of in-space assembly form a strict 'ceiling' on the practical radius.
No current launch vehicle can deploy a complete structure of hundreds of meters in radius as a single block — large-scale rotating stations must be assembled in orbit from separate modules, trusses, and cable elements delivered by dozens of individual launches. Even the cheapest reusable launchers currently cost thousands of dollars per kilogram to low Earth orbit.
This transforms the 'mass ~ r^2.2' curve from an abstract engineering dependency into a linear mission cost dependency: doubling the radius means not just more metal, but much more launches, docking operations, and extravehicular activity (EVA) hours to construct the structure.
NASA programs for in-space assembly (constructing structures directly in orbit using robotic manipulators and autonomous welding/3D printing of trusses) consider a realistic range of orbital demo radii within the approximate 100–200 meters — much smaller than O'Neill or Stanford torus concepts, but already sufficient to significantly reduce RPM compared to compact stations (e.g., ~3 rev/min at 1g for a radius of 100 m).
Such projects rely on tethered configurations – two masses connected by a tether and rotating around a common center of mass – allowing for a large effective radius of rotation without requiring a rigid and massive structure across the entire diameter, significantly reducing mass requirements compared to a solid ring or torus.
Tether-based artificial gravity systems (tether-based artificial gravity) are a popular near-term solution: two capsules at the ends of a 100–500 m tether, rotating around their common center of mass, provide the required radius without the mass of a solid ring.
If you overlay a comfort curve (which increases with radius but saturates) on a cost/mass curve (which increases with radius far more), a characteristic 'sweet spot' emerges – a range of radii where additional comfort from increasing size still justifies the extra mass, but beyond which the benefit becomes negligible compared to the sharp rise in costs.
Comfort as a function of radius (with fixed g) has the shape of a saturation curve: it quickly increases at small radii (where each additional meter significantly reduces RPM), then levels off — beyond approximately 200-250 meters, further increase in radius yields progressively smaller increases in comfort, since RPM is already close to an imperceptible threshold ≈2 rev/min.
Cost/mass as a function of radius has an opposite form: it slowly increases for small radii and then sharply accelerates—doubling the radius from 200 to 400 m can increase the structure's mass by 4–5 times due to nonlinear scaling of pressure, radiation shielding, and structural strength.
The intersection of these two curves is the point where the marginal comfort benefit equals the marginal cost of mass—defining the theoretically 'optimal' radius for a given mission budget.
For most station design analyses, the 'sweet spot' for the next generation lies roughly in the range of 100–250 meters radius: here RPM drops below ~3, which is considered tolerable even without long-term adaptation, while the structure mass remains at an order of magnitude achievable with 10–30 heavy launches, not hundreds.
Beyond ~300-400 m, further RPM reduction (down to the level of the O'Neill cylinder, ~0.5–1 rpm) provides a psychologically pleasant, almost 'terrestrial' experience but requires space-scale extraction and manufacturing (lunar regolith, asteroid raw materials), currently out of reach for purely terrestrial supply in the nearest decades.
Practical design rule: aim for RPM ≤ 3 at the smallest radius that allows — this is the essence of the 'sweet spot' between physiology and engineering.
The final step is transforming the abstract compromise curve into a concrete engineering solution: chosen radius, RPM, and g-levels aligned with mission duration, launch budget, and crew physiological requirements. Historical and contemporary concepts illustrate the full spectrum of possible solutions.
The history of artificial gravity is one of gradual shift from 'ideal comfort' to 'realistic engineering'. The Stanford Torus (1975) and O'Neill Cylinder (1976) are academic concepts from an era when space industrialization with mining on the Moon was considered; they chose a large radius for comfort, ignoring the Earthly cost of launch.
Braun's wheel (1952) took a different path: a compact, realistic 1950s-era station with relatively fast rotation and noticeable physiological compromises.
Current research (NASA, commercial stations like Vast concepts, Gateway spin-modules) leans towards a 'hybrid' approach: moderate radius (100-500 m), often achieved through cable systems or rotating modules, with target RPM in the 2-4 range and partial gravity (0.3-0.6g) instead of full 1g—reducing mass requirements while maintaining sufficient physiological protection against microgravity effects (bone and muscle loss, cardiovascular deconditioning).
No rotating station with humans on board has yet flown. Every number in this simulation is an engineering estimate based on academic research and conceptual NASA/ESA projects, not a flight-proven fact.
Choosing the radius of the rotating station allows the designer to balance three interrelated variables: radius (mass, cost), RPM (physiological comfort) and g-level (protection duration from muscle atrophy). None of them can be optimized separately without compromising others.
A practical conclusion for missions over the next decade: partial gravity (0.3–0.6g instead of full 1g) on a moderate radius (100–300 m) with RPM around 2–4 is the most realistic compromise – sufficient to significantly reduce medical risks of long-duration flight, and sufficiently easy to be realistically built with current or near-term launchers.
| Product | Indication | Trial Design | Key Result |
|---|---|---|---|
| Von Braun Wheel | ≈38 m, ~4–5 rev/min, 1g | Compact wheel, inflatable envelope, 1950s | Technologically simplest; strong Coriolis discomfort |
| Stanford Torus | ≈900 m, ~1 rev/min, 1g | Large torus, mirror ring for solar light | Practically imperceptible rotation; mass — millions of tons |
| O'Neill Cylinder | ≈3200–6400 m, ~0.5–0.8 rev/min | Paired cylinders rotating in opposite directions for stabilization | Maximum comfort and space; requires off-Earth resource extraction |
| Tethered / NASA in-space assembly | 100–500 m, ~2–4 rev/min | Cable capsules or assembled trusses in orbit | Most realistic for the next ten years; moderate mass |