In 1967 a fire inside a pure-oxygen cabin killed three astronauts; the hatch needed about ninety seconds to open even under ideal conditions. Today spacesuits and planned lunar habitats are still full of oxygen, and in space there is no one to call. JaPax turns NASA's fire experiments into one answer: where a fire starts, where it goes, how long it takes, and what must be protected so the crew comes home.
sum of the total size PSI declares for the 19 investigations (PSI API, 15 Nov 2026)
12,146
public files catalogued
tables, images and flame videos, summed across all investigations
0
datasets at lunar gravity
The first fire experiment on the Moon has not flown yet; FM2 is planned for late 2026
Why we exist
Fire in space is a question of human lives
1967
A hazard that was underestimated
During a ground test of Apollo 1 a fire started in a cabin full of pure oxygen. The hatch needed about ninety seconds to open even under ideal conditions and could not be opened against the rising pressure. Gus Grissom, Ed White and Roger Chaffee died inside. It was not bad luck: it was an underestimated risk.
Today
The danger moved, it did not leave
An astronaut who steps outside the station wears a suit filled with 100% oxygen. The habitats planned for the Moon would hold more oxygen than the air we breathe on Earth. Oxygen keeps the crew alive, and it turns any spark into a threat.
In space
No firefighters, no exit, no one to call
If a fire starts, the crew has only what someone designed before launch. Every choice, which material, where to put a barrier, which component to shield, becomes the difference between coming home or not.
The blind spot
Decisions made with very little light
Fire does not behave in space as it does on Earth. NASA has run combustion experiments in microgravity for years, but its archive holds zero datasets at lunar gravity. Whoever designs a suit, a ship or a base has no answer to the key question: if a fire starts, what do I protect first?
Our answer
Protect what keeps you alive. Not everything.
We lit the fire from each of the six parts of a spacesuit and let the engine run. A fire that starts in the upper garment reaches the oxygen tank in under a minute, and the tank bursts at 2 minutes 13 seconds.
0:00
Fire starts in the upper garment
0:58
The oxygen tank catches fire
2:13
The tank bursts
No protection
The tank bursts in 6 of 6 fires
Oxygen tank shielded
The tank bursts in 1 of 6 fires
In the model, shielding the tank turns six bursts out of six into one: the one that starts at the tank itself. You do not need to protect the whole suit. You need to protect what keeps the astronaut alive.
Simulated by the JaPax component engine, which starts from a flame spread rate measured by NASA on thin fabric (SAFFIRE I-III, 1.8-2.6 mm/s). The factors that scale it to the suit (oxygen, airflow, gravity, material), the component layout and the rupture window are estimates, labelled as such. Suit conditions: 100% O₂ at 0.29 atm, no airflow. This is a model, not a test.
Ventilation shutdown may not be enough to put out a fire in O₂-enriched atmospheres
In ACME BRE, flames in O₂-enriched atmospheres were sustained in microgravity under quiescent conditions, with no forced flow; at 21% O₂ they self-extinguished. Cutting ventilation is an established step for extinguishing spacecraft fires; the authors suggest it may not be enough in oxygen-rich atmospheres.
NASA-STD-6001 Test 1 is not conservative for microgravity
BASS indicates that Test 1, an upward-propagation test at 1 g, does not bound the worst case: the limiting oxygen concentration is lower in microgravity. In NASA's Researcher's Guide (2015) the minimum sits near an opposed-flow velocity of 10 cm/s (a qualitative schematic), typical of spacecraft ventilation.
Terrestrial inert-agent standards give a reduced margin in low gravity
FLEX found that the flammability limit is lower in reduced gravity, so terrestrial design standards for inert-gas suppressants (CO₂, He) do not provide the same safety margin.
Each region or marker is an atmosphere (O₂ mole fraction, total pressure) that was actually tested. The orange region is the expected lunar-habitat envelope: oxygen-enriched and sub-atmospheric. Coverage there is sparse, and no test was run at lunar gravity (0.166 g).
Hover (or focus with the keyboard) any element of the chart. Diamonds = real environments · areas and dots = atmospheres actually tested · margin marks = partial data (n/v = not verified).
View as table
Element
Detail
Lunar habitat region
O₂ ≥ 26.5% and pressure ≤ 0.69 atm (10.2 psi). This is where materials are more flammable than on Earth and where data matters most.
FLEX (PSI-69): 273 of 274 tests plotted (1 with an invalid pressure of 0 left out)
Each circle is a group of tests at the same O₂ and pressure (bigger = more tests). Measured O₂ 12–34% and pressures from 0.70 to 3.05 atm, with heptane and methanol droplets and CO₂/He suppressants. The 0.70 atm column sits right at the lunar habitat pressure.
BASS-II (PSI-25): initial O₂ 13.6–22.2%
Measured initial oxygen across its 129 tests. None reached the 26.5% habitat threshold, so solid materials are not yet tested at lunar-habitat oxygen. Pressure is not recorded in the table, so only the O₂ span is marked.
ACME BRE (PSI-20): habitat atmospheres tested
Three atmospheres NASA considers for missions: 14.7 psia/21% O₂, 10.2 psia/26.5% O₂ and 8.2 psia/34% O₂. With elevated O₂ the flame kept burning with no air flow; at 21% it self-extinguished.
CFI (PSI-39): pressure 0.25–5 atm
Cool flames in n-dodecane droplets. The O₂ concentration is not verified in the sources, so it is drawn as a margin bar rather than an area.
ACME CFI-G (PSI-159): up to 85% O₂
The microgravity tests with the most oxygen (challenge statement). Pressure is not verified, so only the O₂ maximum is marked.
Earth / ISS
21% O₂ at 14.7 psi. The ISS has the same atmosphere but in microgravity.
Planned lunar base
30% O₂ or more at 10.2 psi or less (challenge statement). +43% oxygen versus Earth: more flammable.
Spacesuit
100% O₂ at only 4.3 psi. Outside the range of every verified test.
Why truth is part of saving lives
When lives are at stake, an invented number is a danger
In JaPax every number says where it comes from and whether it was measured, reported, estimated or simulated. Where there are no data we say it out loud, because a gap nobody sees can also cost a life. Our honesty is not a limitation: it is the reason you can rely on us when you have to decide.
Measured
Value recorded in a PSI dataset, cited with PSI ID, DOI and sample size.
Reported
Value stated in a peer-reviewed paper or NTRS report and not re-derived from raw data.
Estimated
Engineering assumption or hypothesis with no experimental source; shown explicitly as such.
Simulated
Output of the JaPax component-level spread model (graph, level A).
Open gaps
Planned experiments and missing data
FM2
1/6 g (Moon)
Flammability of Materials on the Moon
A robotic combustion chamber for the lunar surface on CLPS mission CP-21, planned to launch in late 2026. It will test 21% O2 at 14.7 psia and 34% O2 at 8.2 psia on SIBAL fabric and acrylic rods burned upward and downward. SIBAL burns only downward in lunar gravity in air but extinguishes at once in 1 g (NASA NTRS 20240015307; launch date, NTRS 20260000646).
Flew on a spinning Blue Origin New Shepard sounding rocket (11 RPM, 1.22 m effective radius). First-ever extended (>25 s) burns in simulated lunar gravity: a cotton-fiberglass fabric and a 4 mm PMMA rod burned together in air at normal pressure for over 2.5 minutes, with oxygen falling until extinction. Results reported in 2025 (NASA NTRS 20250010653); not in PSI.
An insert for the ISS Combustion Integrated Rack that burns solid spacecraft materials in practical geometries at oxygen and pressure levels of current and planned NASA Space Exploration Atmospheres. It will test whether the 1-g flame test (NASA-STD-6001) is conservative in low gravity, improve understanding of early fire growth, and compare suppression by diluents, flow reduction and venting. Built to serve five investigations and open to future researchers (NASA NTRS 20200000361).
Combustion science JaPax is built to ingest these datasets as soon as they are published in PSI.
Lessons paid for with lives
Mishaps that shaped the requirements
1967
Apollo 1
A flash fire swept the Command Module during a pre-launch test in a 100% oxygen cabin. Grissom, White and Chaffee died. The three-piece hatch needed about 90 seconds to open even under ideal conditions and could not be opened against the rising pressure.
Lesson learned: A unified hatch that can be opened from inside or outside in 3 seconds, tightly restricted combustible materials and fire-resistant spacesuits.
Oxygen tank no. 2 exploded en route to the Moon after its heaters overheated and degraded the Teflon insulation; heater voltage had been raised from 28 to 65 V DC without upgrading the thermostats.
Lesson learned: Oxygen system redesign: a third tank, upgraded thermostats, no stirring fans and stainless-steel-sheathed wiring.
On 24 February a solid-fuel oxygen generator canister caught fire in the Kvant-1 module. The flame burned for several minutes and the crew fought it with foam extinguishers.
Lesson learned: Changes to canister preparation and emergency training; the ISS uses dissimilar, redundant oxygen generation.