Material Requirements for O'Neill Cylinder Habitats

Date: 2026-03-20
Scope: Quantitative mass estimates for two cylinder scenarios
Inputs from: Phase 2 constraint analysis (002), NASA SP-413, O'Neill (1977)

---

Design Scenarios

Parameter	Scenario 1: Minimum Viable	Scenario 2: O'Neill-Class
Radius	982 m	3,200 m
Length	2,000 m (see §A)	32,000 m
Target gravity	1.0 g	1.0 g
RPM	0.95	0.53
Rim speed	98.1 m/s	177.1 m/s
Population (est.)	~8,000	~1,000,000+
Surface area (\(2\pi r L\))	12.34 km²	643.4 km²
End cap area (\(2 \times \pi r^2\))	6.06 km²	64.34 km²
Total shell area	18.4 km²	707.7 km²
Interior volume (\(\pi r^2 L\))	\(6.06 \times 10^9\) m³	\(1.03 \times 10^{12}\) m³
Usable floor area (~50% of barrel)	6.17 km²	321.7 km²

§A. Minimum Viable Length

The NASA SP-413 cylinder design used a 10:1 length-to-radius ratio. For a minimum viable habitat, a 2:1 ratio (L = 2,000 m) is defensible: it provides enough length for a meaningful community while keeping end-cap structural penalties manageable. Below ~1 km length, the end-cap mass fraction becomes prohibitive and the interior feels claustrophobic. A 2 km cylinder at 982 m radius provides ~6 km² of barrel surface, comparable to the Stanford torus's 6.8 × 10⁵ m² projected area but with far more volume.

---

1. Structural Shell Mass

Hoop Stress Analysis

A rotating cylinder under internal atmospheric pressure and centripetal loading experiences hoop stress. For a thin-walled cylinder:

Atmospheric hoop stress:

\[ \sigma_{\text{atm}} = \frac{P \cdot r}{t} \]

where \(P\) = internal pressure, \(r\) = radius, \(t\) = wall thickness.

Centripetal hoop stress (from the shell's own mass and interior loading):

\[ \sigma_{\text{cent}} = \rho_{\text{shell}} \cdot \omega^2 \cdot r^2 \]

where \(\rho_{\text{shell}}\) is the areal mass density of everything at the rim.

For the structural shell alone under atmospheric pressure, solving for minimum wall thickness:

\[ t_{\min} = \frac{P \cdot r}{\sigma_y} \]

Material Properties

Material	Density (kg/m³)	Yield Strength (MPa)	Specific Strength (kN·m/kg)
Mild steel (A36)	7,850	250	32
High-strength steel (maraging)	8,000	1,400	175
Aluminum 6061-T6	2,700	276	102
Aluminum 7075-T6	2,810	503	179
Titanium Ti-6Al-4V	4,430	880	199
Carbon fiber composite (CFRP)	1,600	1,500	938

Wall Thickness Calculations

Scenario 1 (r = 982 m, P = 51 kPa at half-atmosphere per SP-413):

Material	t_min (mm)	t with SF=2 (mm)	t with SF=4 (mm)
Mild steel	200	400	800
High-strength steel	36	71	143
Aluminum 6061-T6	181	361	722
Aluminum 7075-T6	99	199	397
Titanium Ti-6Al-4V	57	114	228
CFRP	33	67	133

Scenario 2 (r = 3,200 m, P = 101.3 kPa at full atmosphere):

Material	t_min (mm)	t with SF=2 (mm)	t with SF=4 (mm)
Mild steel	1,297	2,594	5,187
High-strength steel	231	463	926
Aluminum 6061-T6	1,174	2,348	4,696
Aluminum 7075-T6	644	1,289	2,578
Titanium Ti-6Al-4V	368	737	1,473
CFRP	216	432	864

Note: Full atmosphere (101.3 kPa) doubles the pressure load compared to the SP-413 half-atmosphere design (51 kPa). The "comfortable" scenario uses full atmosphere for Earth-normal conditions.

Structural Shell Mass Estimates

\[ M_{\text{shell}} = \rho_{\text{material}} \times t_{\text{design}} \times A_{\text{surface}} \]

Scenario 1 (982 m radius, 2 km length, half-atmosphere): Using high-strength steel at SF=3 (\(t = 107\) mm):

\[ M_{\text{shell}} = 8{,}000 \times 0.107 \times 18.4 \times 10^6 = 15.7 \; \text{Mt} \]

This is excessive. The NASA SP-413 found that cylinders require ~4× more structural mass per unit area than tori, which is why they rejected pure cylinders.

Using the SP-413's preferred aluminum at SF=2, with half-atmosphere: For the barrel only (12.34 km²), aluminum 7075-T6, \(t = 199\) mm:

\[ M_{\text{barrel}} = 2{,}810 \times 0.199 \times 12.34 \times 10^6 = 6.9 \; \text{Mt} \]

However, the more relevant calculation follows the SP-413 approach. The NASA study found structural masses empirically:

Configuration	Radius (m)	Structural Mass (kt)	Source
Stanford torus (r=830m)	830	150	SP-413 Table 4-1
Cylinder (r=895m, L=8950m)	895	42,300	SP-413 Table 4-1
Sphere (r=895m)	895	3,545	SP-413 Table 4-1
SP-413 torus (built)	895	156	SP-413 Table 5-2

The SP-413 cylinder at r=895m, L=8950m had structural mass 42,300 kt = 42.3 Mt. Scaling:

Scenario 1 — scaling from SP-413 cylinder by surface area ratio:

\[ A_{\text{SP-413}} = 2\pi \times 895 \times 8{,}950 = 50.3 \; \text{km}^2 \]

\[ A_{\text{S1}} = 2\pi \times 982 \times 2{,}000 = 12.3 \; \text{km}^2 \]

\[ M_{\text{struct,1}} \approx 42{,}300 \times \frac{12.3}{50.3} \approx 10{,}350 \; \text{kt} \approx 10.4 \; \text{Mt} \]

But this overestimates because the SP-413 cylinder used full atmosphere. At half-atmosphere:

\[ M_{\text{struct,1}} \approx 10.4 \times 0.5 \approx 5.2 \; \text{Mt} \]

More realistic estimate using the torus shell density as a baseline: The Stanford torus shell was 156,000 t for a surface area of \(2.1 \times 10^6\) m² = 74 kg/m². A cylinder at the same radius faces higher hoop stress (larger radius of curvature in the longitudinal cross-section is infinite vs. 65m for the torus). The SP-413 found cylinders need ~4× more mass per unit projected area. Applying a 2× structural penalty over the torus:

\[ M_{\text{struct,1}} \approx 150 \; \text{kg/m}^2 \times 12.34 \times 10^6 \; \text{m}^2 = 1.85 \; \text{Mt} \]

Best estimate for Scenario 1 structural shell: 2 - 5 Mt depending on material and safety factor choices.

Scenario 2 — scaling from SP-413:

\[ A_{\text{S2}} = 2\pi \times 3{,}200 \times 32{,}000 = 643 \; \text{km}^2 \]

Direct calculation with high-strength steel, SF=3, full atmosphere, barrel only:

\[ t = \frac{3 \times 101{,}300 \times 3{,}200}{1{,}400 \times 10^6} = 0.695 \; \text{m} \]

\[ M_{\text{barrel}} = 8{,}000 \times 0.695 \times 643 \times 10^6 = 3{,}575 \; \text{Mt} \]

With CFRP, SF=3:

\[ t = \frac{3 \times 101{,}300 \times 3{,}200}{1{,}500 \times 10^6} = 0.649 \; \text{m} \]

\[ M_{\text{barrel}} = 1{,}600 \times 0.649 \times 643 \times 10^6 = 668 \; \text{Mt} \]

With aluminum 7075-T6, SF=3:

\[ t = \frac{3 \times 101{,}300 \times 3{,}200}{503 \times 10^6} = 1.934 \; \text{m} \]

\[ M_{\text{barrel}} = 2{,}810 \times 1.934 \times 643 \times 10^6 = 3{,}493 \; \text{Mt} \]

Best estimate for Scenario 2 structural shell: 700 Mt (CFRP) to 3,500 Mt (steel/aluminum)

Key insight: structural shell mass scales as \(r^2 \times L \times P / \sigma_y\). The O'Neill-class cylinder is enormously more massive due to the 3.26× radius increase and 16× length increase combined with doubled pressure.

---

2. Radiation Shielding Mass

Shielding Requirements

Protection Level	Areal Density	Equivalent	Source
Earth atmosphere	1,033 g/cm² = 10,330 kg/m²	10.3 t/m²	Standard physics
SP-413 design	4,500 kg/m²	4.5 t/m²	NASA SP-413 Ch. 4
Minimum for 0.5 rem/yr	~4,500 kg/m²	4.5 t/m²	SP-413 (accounts for oblique incidence)
Minimum for 5 rem/yr (worker limit)	~2,000 kg/m²	2.0 t/m²	Estimated from SP-413 scaling
"Barely survivable"	~1,500 kg/m²	1.5 t/m²	See note below

Critical physics note from SP-413 Chapter 2: At intermediate shielding depths (a few t/m²), cosmic ray dose increases to ~20 rem/yr due to secondary particle production (spallation). You must push through this "dose bump" to thicker shielding to reach the protective regime. The SP-413 found 4.5 t/m² was the minimum to get below 0.5 rem/yr. Going thinner than ~3 t/m² is actually worse than no shielding at all for cosmic rays.

"Barely survivable" minimum: Accepting 5 rem/yr (the radiation worker limit), and using water or hydrogen-rich materials (which produce fewer secondary neutrons than regolith), ~2 t/m² may be acceptable. But this is the occupational limit, not safe for families/children. For a true minimum with children, 4.5 t/m² is the floor.

Shielding Geometry

The shielding must cover the projected cross-sectional area against isotropic cosmic rays. For a cylinder, the total surface requiring shielding is the full outer surface (barrel + end caps) since cosmic rays come from all directions.

Scenario 1 (\(r = 982\) m, \(L = 2{,}000\) m):

\[ A_{\text{shield}} = 18.4 \times 10^6 \; \text{m}^2 \]

\[ \text{At } 4.5 \; \text{t/m}^2: \quad M_{\text{shield}} = 4{,}500 \times 18.4 \times 10^6 = 82.8 \; \text{Mt} \]

\[ \text{At } 2.0 \; \text{t/m}^2: \quad M_{\text{shield}} = 2{,}000 \times 18.4 \times 10^6 = 36.8 \; \text{Mt} \]

Scenario 2 (\(r = 3{,}200\) m, \(L = 32{,}000\) m):

\[ A_{\text{shield}} = 707.7 \times 10^6 \; \text{m}^2 \]

\[ \text{At } 4.5 \; \text{t/m}^2: \quad M_{\text{shield}} = 4{,}500 \times 707.7 \times 10^6 = 3{,}185 \; \text{Mt} \]

\[ \text{At } 10.3 \; \text{t/m}^2: \quad M_{\text{shield}} = 10{,}330 \times 707.7 \times 10^6 = 7{,}313 \; \text{Mt} \]

Shielding Material Options

Material	Density (kg/m³)	Thickness for 4.5 t/m²	Notes
Lunar regolith	1,500	3.0 m	Primary SP-413 choice
Water	1,000	4.5 m	Better H content, fewer secondaries
Lunar slag (processed)	2,500	1.8 m	Denser, thinner layer
Polyethylene	950	4.7 m	Best H density, expensive to produce

The SP-413 Stanford torus used 9.9 Mt of lunar regolith shielding — this was by far the dominant mass component (~95% of total). The shielding was 1.7 m thick at the torus surface.

Shielding Mass Summary

Scenario	Minimum (2 t/m²)	SP-413 Standard (4.5 t/m²)	Earth-Equivalent (10.3 t/m²)
1: Minimum	36.8 Mt	82.8 Mt	190 Mt
2: O'Neill	1,415 Mt	3,185 Mt	7,313 Mt

---

3. Atmospheric Mass

Atmospheric Pressure Options

Atmosphere	Total Pressure	O₂	N₂	Notes
SP-413 design	51 kPa (0.5 atm)	22.7 kPa	26.6 kPa	Half-pressure, safe with enriched O₂
Earth standard	101.3 kPa (1 atm)	21.2 kPa	79.0 kPa	Full atmosphere

Atmospheric Mass Calculation

For a cylinder, the atmosphere fills the entire volume. Using the ideal gas law:

\[ M_{\text{atm}} = \frac{P \cdot V \cdot \bar{M}}{R \cdot T} \]

where \(\bar{M}\) is the mean molar mass (28.97 g/mol for Earth mix, ~26 g/mol for SP-413 mix), \(R = 8.314\) J/(mol·K), and \(T \approx 293\) K.

At sea level density \(\rho_{\text{air}} = 1.225\) kg/m³ (Earth) or \(\approx 0.7\) kg/m³ (SP-413 mix at 51 kPa):

Scenario 1 (\(V = 6.06 \times 10^9\) m³):

\[ \text{At 51 kPa:} \quad M_{\text{atm}} \approx 0.7 \times 6.06 \times 10^9 = 4.2 \; \text{Mt} \]

\[ \text{At 101.3 kPa:} \quad M_{\text{atm}} \approx 1.225 \times 6.06 \times 10^9 = 7.4 \; \text{Mt} \]

Scenario 2 (\(V = 1.03 \times 10^{12}\) m³):

\[ \text{At 101.3 kPa:} \quad M_{\text{atm}} \approx 1.225 \times 1.03 \times 10^{12} = 1{,}260 \; \text{Mt} \]

\[ \text{At 51 kPa:} \quad M_{\text{atm}} \approx 0.7 \times 1.03 \times 10^{12} = 721 \; \text{Mt} \]

Cross-check with SP-413: The SP-413 cylinder (\(r = 895\) m, \(L = 8{,}950\) m, \(V = 2.25 \times 10^{10}\) m³) had atmospheric mass 14,612 kt = 14.6 Mt at 51 kPa. This gives \(\rho \approx 0.65\) kg/m³, consistent with the lower molecular weight SP-413 mix.

Atmospheric Mass Summary

Scenario	SP-413 mix (51 kPa)	Earth standard (101.3 kPa)
1: Minimum (982m × 2km)	4.2 Mt	7.4 Mt
2: O'Neill (3200m × 32km)	721 Mt	1,260 Mt

Note: atmospheric mass is large but not dominant compared to shielding. For Scenario 2, it becomes a very significant fraction.

---

4. Soil and Water Mass

Soil Requirements

Parameter	Minimum	Comfortable	Source
Soil depth for agriculture	0.5 m	1.0 - 2.0 m	SP-413, agricultural science
Dry soil density	1,300 kg/m³	1,500 kg/m³	Typical topsoil
Water content in soil	10% by mass	20% by mass	Agricultural standard
Agricultural area per person	20 m²	50 m²	SP-413 Table 5-4
Residential area per person	43 m²	67 m²	SP-413 Ch. 3 & 4

Soil Mass Estimates

Scenario 1 (usable floor area \(\approx 6.17\) km², ~50% agricultural):

\[ A_{\text{ag}} = 3.1 \; \text{km}^2 = 3.1 \times 10^6 \; \text{m}^2 \]

\[ M_{\text{soil}} = 0.5 \times 1{,}300 \times 3.1 \times 10^6 = 2.0 \; \text{Mt} \]

\[ M_{\text{water in soil}} = 0.1 \times M_{\text{soil}} = 0.2 \; \text{Mt} \]

Scenario 2 (usable floor area \(\approx 321.7\) km², generous allocation):

\[ M_{\text{ag soil}} = 1.5 \times 1{,}500 \times 10^8 = 225 \; \text{Mt} \]

\[ M_{\text{parks}} = 0.5 \times 1{,}500 \times 10^8 = 75 \; \text{Mt} \]

\[ M_{\text{water in soil}} = 0.2 \times 300 = 60 \; \text{Mt} \]

\[ M_{\text{soil, total}} = 225 + 75 + 60 = 360 \; \text{Mt} \]

Water Requirements

SP-413 reference: 42,000 t water (20,000 t free water + 22,000 t in soil) for 10,000 people = 4.2 t/person.

Water Category	Scenario 1	Scenario 2
Soil moisture	0.2 Mt	60 Mt
Rivers/lakes/reservoirs	0.5 Mt	200 Mt
Industrial/recycling	0.1 Mt	50 Mt
Humidity (in atmosphere)	~0.1 Mt	~10 Mt
Total water	~1 Mt	~320 Mt

Soil + Water Summary

Component	Scenario 1	Scenario 2
Dry soil	2.0 Mt	300 Mt
Water (all forms)	1.0 Mt	320 Mt
Total	3.0 Mt	620 Mt

---

5. Interior Mass (Buildings, Infrastructure, People)

SP-413 Reference Data (10,000 population)

Component	Mass (kt)	Per Capita (t/person)
Structures	77	7.7
Machinery	40	4.0
Biomass (plants)	5	0.5
Furnishings	25	2.5
People (avg 70 kg)	0.7	0.07
Total interior	148	14.8

Scaled Estimates

Scenario 1 (population ~8,000):

\[ M_{\text{interior}} \approx 8{,}000 \times 14.8 = 118 \; \text{kt} \approx 0.12 \; \text{Mt} \]

Rounding up for safety: ~0.2 Mt

Scenario 2 (population ~1,000,000):

\[ M_{\text{interior}} \approx 1{,}000{,}000 \times 14.8 = 14{,}800 \; \text{kt} \approx 14.8 \; \text{Mt} \]

With more generous infrastructure (parks, transport, industry): ~30 Mt

---

6. Total Mass Estimates

Scenario 1: "Barely Survivable" Minimum Cylinder (982m × 2km)

Component	Minimum Mass	Notes
Structural shell	2 Mt	CFRP, SF=2, half-atmosphere
Radiation shielding	37 Mt	2.0 t/m², worker-limit dose
Atmosphere	4.2 Mt	SP-413 mix at 51 kPa
Soil + water	3.0 Mt	Minimal agriculture
Interior	0.2 Mt	Spartan infrastructure
TOTAL	~46 Mt

With SP-413 standard shielding (4.5 t/m²): ~92 Mt

Scenario 2: "Most Comfortable" O'Neill-Class Cylinder (3200m × 32km)

Component	Mass	Notes
Structural shell	700 Mt (CFRP) to 3,500 Mt (metal)	SF=3, full atmosphere
Radiation shielding	3,185 Mt (SP-413) to 7,313 Mt (Earth-equiv.)	4.5 - 10.3 t/m²
Atmosphere	1,260 Mt	Full Earth atmosphere
Soil + water	620 Mt	Deep soil, lakes, rivers
Interior	30 Mt	Full urban infrastructure
TOTAL (CFRP + SP-413 shielding)	~5,800 Mt
TOTAL (metal + Earth-equiv. shielding)	~12,700 Mt

Comparison with Literature

Reference	Design	Total Mass	Notes
SP-413 (1975)	Stanford torus (r=830m)	~10.5 Mt	Dominated by 9.9 Mt shielding
SP-413 (1975)	Cylinder (r=895m, L=8950m)	~80+ Mt	42.3 Mt structure + 23.3 Mt shield + 14.6 Mt atmo
O'Neill (1977)	Island Three (r=3200m, L=32km)	~several thousand Mt	Order-of-magnitude consistent
ISS	LEO station	0.00042 Mt (420 t)	For scale reference

The ISS comparison: The "barely survivable" minimum cylinder is approximately 100,000× the mass of the ISS. The O'Neill-class cylinder is roughly 10 million × the ISS mass. This illustrates the extraordinary scale difference between current space construction and permanent habitats.

---

7. Material Sourcing

Lunar Regolith Composition (by mass)

Oxide	Fraction	Useful Elements
SiO₂	45%	Silicon, oxygen
Al₂O₃	15%	Aluminum
FeO	15%	Iron
CaO	10%	Calcium
MgO	10%	Magnesium
TiO₂	5%	Titanium

Lunar regolith provides: structural metals (Al, Fe, Ti), radiation shielding (bulk regolith/slag), oxygen (for atmosphere), silicon (for glass/solar cells). What it lacks: hydrogen (for water), carbon, nitrogen — these must come from asteroids, comets, or Earth.

Source Requirements

Scenario 1 (46 Mt minimum): - Shielding (37 Mt): Lunar regolith, minimally processed - Structure (2 Mt): Lunar aluminum or asteroid iron/nickel - Atmosphere N₂ (2.6 Mt): Asteroid volatiles or Earth import - Atmosphere O₂ (1.6 Mt): Lunar regolith extraction - Water (1 Mt): Asteroid ice or lunar polar ice - Soil (2 Mt): Lunar regolith with organic amendments from asteroidal carbon

Scenario 2 (5,800 Mt minimum): - Shielding (3,185 Mt): Lunar regolith — at SP-413's 1.2 Mt/yr extraction rate, this alone would take 2,654 years. Multiple orders of magnitude increase in mining infrastructure required, or asteroid capture. - Atmosphere N₂ (~1,000 Mt): Cannot come from the Moon (nitrogen-poor). Requires massive asteroidal/cometary sources. - Water (320 Mt): Asteroid ice capture on industrial scale.

The Nitrogen Problem

Earth's atmosphere is 78% N₂ by volume. The SP-413 half-atmosphere design reduces N₂ requirements, but even so, the Moon contains negligible nitrogen. For Scenario 2, ~1,000 Mt of nitrogen is needed. Carbonaceous chondrite asteroids contain 1-3% nitrogen by mass, so capturing and processing 30,000 - 100,000 Mt of asteroid material would be required for nitrogen alone.

---

8. Key Formulas Summary

Hoop stress (thin-walled cylinder):

\[ \sigma = \frac{P \cdot r}{t}, \qquad t_{\min} = \frac{P \cdot r}{\sigma_y}, \qquad t_{\text{design}} = \frac{\text{SF} \cdot P \cdot r}{\sigma_y} \]

Shell mass:

\[ M_{\text{shell}} = \rho_{\text{material}} \cdot t_{\text{design}} \cdot A_{\text{surface}} \]

\[ A_{\text{barrel}} = 2\pi r L, \qquad A_{\text{endcaps}} = 2\pi r^2 \]

Atmospheric mass:

\[ M_{\text{atm}} = \rho_{\text{air}} \cdot V = \rho_{\text{air}} \cdot \pi r^2 L, \qquad \rho_{\text{air}} = \frac{P \cdot \bar{M}}{R \cdot T} \]

At 101.3 kPa, 293 K: \(\rho_{\text{air}} = 1.225\) kg/m³. At 51 kPa, 293 K: \(\rho_{\text{air}} \approx 0.65\) kg/m³.

Shielding mass:

\[ M_{\text{shield}} = \sigma_{\text{areal}} \cdot A_{\text{total}} \]

where \(\sigma_{\text{areal}}\) is the areal density requirement (kg/m²).

Centripetal acceleration (artificial gravity):

\[ a = \omega^2 r = g_{\text{target}}, \qquad \omega = \sqrt{\frac{g}{r}} \]

---

References

• Johnson, Richard D., and Charles Holbrow, editors. Space Settlements: A Design Study. NASA SP-413, National Aeronautics and Space Administration, 1977. Chapter 4 (habitat selection), Chapter 5 (colony design), Chapter 6 (construction).

• O'Neill, Gerard K. The High Frontier: Human Colonies in Space. William Morrow and Company, 1977. Island Three specifications: paired cylinders, 3.2 km radius, 32 km length, ~1 million population.

• O'Neill, Gerard K. "The Colonization of Space." Physics Today, vol. 27, no. 9, 1974, pp. 32-40. Original proposal for L5 colonies using lunar materials.

• NASA SP-413, Chapter 2: "Physical Properties of Space." Cosmic ray flux of ~10 rem/yr unshielded; secondary particle production at intermediate shielding depths; 4.5 t/m² minimum for 0.5 rem/yr.

• NASA SP-413, Chapter 3: "Human Needs in Space." Atmospheric requirements (51 kPa total, 22.7 kPa O₂, 26.6 kPa N₂), gravity (0.95 ± 0.05 g), radiation (<0.5 rem/yr), area per person (67 m²).

• NASA SP-413, Chapter 4, Table 4-1: Comparison of habitat configurations — structural mass, shielding mass, and atmospheric mass for torus, cylinder, sphere, and dumbbell geometries at 1 rpm, 0.95g.

• NASA SP-413, Chapter 5, Tables 5-2 and 5-3: Stanford torus mass breakdown — 9.9 Mt shielding, 156 kt shell, 220 kt soil, 42 kt water, 530 kt interior total.

• Globus, Al, and Tom Marotta. "The High Frontier: An Easier Way." NSS Space Settlement Journal, 2018. Updated analysis arguing for smaller habitats at lower Earth orbits within the Van Allen belts (reducing shielding requirements).

---

Conclusions

1. Shielding dominates everything. In both scenarios, radiation shielding is the largest or second-largest mass component. The SP-413 found this in 1975, and it remains true. Any design optimization must start with shielding.

2. The structural shell is the second crisis. At O'Neill-class scales with full atmosphere, even CFRP requires ~700 Mt of structural shell. The hoop stress scales linearly with both radius and pressure. The SP-413's decision to use half-atmosphere (51 kPa) and a torus (smaller radius of curvature) were driven by this structural reality.

3. Atmospheric mass is non-trivial. The O'Neill cylinder contains 1,260 Mt of air at full atmosphere — more than the structural shell in some material scenarios. The nitrogen sourcing problem is severe.

4. The minimum viable cylinder (~46 Mt) is achievable with lunar resources in principle, though it requires industrial-scale lunar mining far beyond current capabilities.

5. The O'Neill-class cylinder (~5,800 Mt minimum) requires asteroid mining in addition to lunar resources, particularly for nitrogen and water. At current launch costs (~$2,700/kg to LEO with Starship), even moving 1 Mt from Earth would cost $2.7 trillion. All mass must come from space resources.

6. The SP-413 torus at 10.5 Mt was the most mass-efficient design among the options studied in 1975. It achieved comparable living area to much larger cylinders at a fraction of the mass. The cylinder's structural penalty is severe.