Skip to content

Simulator

Pick an environment and watch the fire behave

Step into each place in 3D, then watch the same burning droplet under its own atmosphere and gravity. The readouts below are computed from the data; the animation shows what that atmosphere does to the flame, molecule by molecule.

Tap any surface to start a fire there.

Touch to burn

Start a fire anywhere and watch it spread

Pick a place, then tap any surface in the 3D scene: that is where the fire starts. It grows and creeps across the surfaces around it while the numbers below the scene change in real time.Model: a flame front propagating over a surface. Its spread rate depends on the atmosphere (O₂ fraction, total pressure), the ventilation flow, gravity and the fuel.

Lunar gravity = extrapolation. No experimental combustion data in PSI; LUCI (2025) and FM2 are the first.

SIMULATION · flames and spread are illustrative

Real fires here creep at millimetres per second, so time is sped up. Both clocks are shown. 3D scene: a visualization, not a measurement. Models: NASA 3D Resources (ISS, EMU, Lunar Module), procedural terrain and Earth textures from three.js examples. References

Live data

all values simulated

Waiting for ignition

Time

—

simulated

Starts when you tap a surface.

Spread speed

—

simulated

Flame-front spread rate over the surface.

Flame front

—

simulated

Radial distance from the ignition point to the flame front.

Burned area

—

simulated

Surface already pyrolysed or incandescent. Scene units are approximately metres.

Smoke / soot

—

simulated particles

Soot and smoke particles rendered as a visual aid.

Burning spots

—

simulated

Flame sprites on the front and surface cells already involved.

Surfaces reached

—

simulated

Number of distinct model surfaces reached by the fire.

Intensity

—

simulated

Normalised flame intensity (dimensionless, not a temperature).

Detection
—No dataset gives detector response times. The delay below is yours, not a measured time.
Extinction
Ready: press Activate
How is the spread speed built?

Speed = anchor × oxygen × airflow × gravity × material. Only the anchor is measured; the rest are modelling choices, shown here so you can judge them.

  • Anchor2.0 mm/smeasured

    Representative value inside the measured concurrent spread on thin fabric (SAFFIRE I-III, 1.8-2.6 mm/s).

    Source: NTRS 20180005168 Table 1; PSI-98, PSI-99, PSI-100

  • Oxygen×1.50estimated

    Spread rises with O2 in microgravity (qualitative). The exponent 1.3 is a modelling choice, relative to 22 % O2.

    Source: NTRS 20160000593 (BASS-II, PMMA rods, O2 16-21 %)

  • Airflow×0.70estimated

    Flame quenches below about 1-5 cm/s at 16-21 % O2; the steps 15/5/1 cm/s are a discretisation. With gravity, buoyant convection supplies air, so low forced flow matters less.

    Source: NTRS 20150008962 (SIBAL, flow 1-40 cm/s); PSI-20 ACME BRE

  • Gravity×1.13estimated

    1 g spread was faster than microgravity in PMMA rods and silicone tests. The slope 0.8 per g is a modelling choice; partial gravity (Moon) is an extrapolation.

    Source: NTRS 20160000593; PSI-99 (silicone, 1 g vs 0 g)

  • Material×1.00estimated

    Thicker or non-fabric solids spread slower (PMMA rods: slower with diameter). The exact factor is a modelling choice; cable insulation has no flame-spread data at all.

    Source: NTRS 20160000593; no cable flame-spread dataset

Component engine

simulation

Tap a part of the model to identify it and see how a fire could move between components. The timeline, the colours on the model and the numbers come from the same engine.

No part selected yet.

Measures

Applied at the current simulated time. Every measure here is an engineering proposal (hypothesis); none has test data for this model.

Measures unlock once a fire is burning: tap a part of the model to start one.

  • Shut off the fans and ventilation so the air stops moving.

    Why and source

    estimated Engineering proposal (hypothesis for the engine): with less airflow the flame spreads slower or starves (ACME BRE, PSI-20: flames went out in still air at 21 % O2 in microgravity). In this model the effect depends on oxygen and gravity; it is not tested for this place. Source: Hypothesis for the engine; effect basis PSI-20 ACME BRE, DOI 10.60555/vb93-3y72 (flames in still air at 21 % O2 in microgravity). No lunar-gravity combustion data exists (LUCI and FM2 are the first)

  • Use a portable extinguisher on the burning part before the fire grows.

    Why and source

    estimated Engineering proposal (hypothesis): the engine puts out the fire where it started at that moment. No extinguisher of the lunar module is documented in the sources used here. Source: Hypothesis; no test in the sources used

  • Put a fire-blocking barrier between the ascent stage and the descent stage so a fire cannot cross.

    Why and source

    estimated Engineering proposal (hypothesis): the engine removes this connection. A barrier breaks the continuous fuel path that the model assumes. No test of such a barrier exists in the sources used. Source: Hypothesis; flame spread needs continuous fuel (stated, not measured, for this place)

No simulation recorded yet: tap a part of the model to start a run, then download its data.

Data gaps
  • The lunar module is labelled by mesh-name prefix (the 134 meshes have opaque names): 6 components were identified from highlight views of the model, with medium to low confidence; the rest are listed as unidentified.
  • Gravity at 0.166 g is an extrapolation: PSI holds no partial-gravity combustion data (LUCI 2025 and FM2 are the first).
  • No PSI/NTRS combustion test exists for the lunar module or its materials; spread speeds are the SAFFIRE anchor scaled by estimated factors.
  • Energetic components: none verified. The lunar module carried oxygen, propellant and helium tanks, but none was verified with a source here and none is a separately labelled part of the model, so there is no rupture.
  • The landing gear and the antennas are treated as non-combustible metal (assumption): a fire does not spread into them in the model.
  • The ascent and descent stage labels (polySurfa*, group13_p*) come from the model's shapes by eye; no drawing of the lunar module was used. Detection and crew actions are not modelled.

Lunar base: fixed conditions

  • Oxygen30 % or moreestimated

    Planned lunar habitats combine a reduced total pressure with O₂ enrichment to keep a breathable O₂ partial pressure. A higher O₂ mole fraction increases flammability.

    Source:

    Challenge statement; FM2 plans 21 % at 14.7 psia and 34 % at 8.2 psia (NTRS 20240015307)

  • Pressure10.2 psi or lessestimated

    Reduced total pressure lowers the structural mass of the pressure vessel.

    Source:

    Challenge statement; FM2 (NTRS 20240015307)

  • Gravity0.166 gestimated

    About 1/6 of terrestrial g. No flame-spread data exist at this level, so the model interpolates between the 0 g and 1 g regimes.

    Source:

    Lunar gravity is a constant, but there is no PSI combustion data at this g (LUCI 2025 and FM2 are the first)

  • Airflowabout 10 cm/s (assumed)estimated

    Assumes habitat ventilation comparable to the ISS cabin. This is an assumption, not a measurement.

    Source:

    Assumed equal to the ISS cabin; no lunar-habitat data

Want your own numbers? Press Custom.

Evidence basis: BASS / BASS-II, thin-fabric tests.

Fire suppression

Choose how to fight this fire, then activate it. The effect is applied to the fire you started.

Tested up to 70 % in FLEX.

Delay before it acts

Start a fire first: tap a surface in the 3D scene, then Activate unlocks.

Why is this happening?

Each statement carries its evidence tag and source. Counts for CO₂ and helium come from measured FLEX droplet tests (PSI-69, DOI 10.60555/mbq8-0451, 274 tests); FLEX burned droplets, not fabric, plastic or cables.

  • measuredThin fabric burning with the airflow spreads steadily; the flame goes out when the flow drops below about 1-5 cm/s (higher flow needed at lower O2). Wider samples, faster flow and more O2 spread faster.Physical basis: Concurrent-flow spread is oxidiser-limited: below the quench velocity the flame is starved.Source: NTRS 20150008962 (BASS/BASS-II on the ISS; SIBAL samples 1.2 and 2.2 cm wide, O2 16-21 %, flow 1-40 cm/s)
  • reportedWith a moderate flow, a thin fuel burning with the airflow spreads faster as the flow rises. Opposed thin fuel spreads fastest at an intermediate flow.Source: NTRS 20140011099 (BASS on the ISS; flow 0-40 cm/s)
  • reportedDanger: a suddenly stronger airflow, or touching another fuel, can make a small flame flare up.Source: NTRS 20140011099 (BASS on the ISS; flow 0-40 cm/s)
  • measuredFewer than 3 tests fall in this window, so a comparison would be noise.Physical basis: FLEX ignited fuel droplets with and without the agent and recorded the extinction outcomes.Source: PSI-69 FLEX, DOI 10.60555/mbq8-0451, 274 droplet tests
  • reportedIn microgravity helium can be the better agent: CO2 glows and re-radiates heat to the fuel, and helium works about 11 times better per kilogram. On the ISS the extinguishers are CO2.Source: NTRS 20040161188 (foam fuel; drop tower and ZGF, short tests)
  • reportedSpraying an agent only at the flame is not enough: the jet drags air in and keeps the flame alive even with the ambient airflow off. Dilute the whole space.Source: NASA challenge guide p.31 (BASS)
  • measuredFLEX burned fuel droplets, not cloth, plastic or cables, so this shows the mechanism, not a safe dose for this material.Source: PSI-69 FLEX, DOI 10.60555/mbq8-0451, 274 droplet tests
  • reportedIn a large spacecraft fire the rise in CO2 and CO was bigger than the rise in temperature and pressure, which favours detecting gases early.Source: NTRS 20240002981 (Saffire VI, 1.5-3.2 kW fires in 17-19 m3)

What is data and what is simulation: the spread you watch is a simulation anchored to measured SAFFIRE points (about 1 to 2.6 mm/s on thin fabric in microgravity). The factors that scale it to other atmospheres (oxygen, airflow, gravity, material) are estimated, not fitted. Particles, area and speed shown above are produced by this engine. The measured rates are listed below with their PSI ID and DOI. No experiment gives a spread curve, and FLEX has no suppressant test above 25 % O₂.

The rates NASA actually measured, with their PSI ID and DOI, are in the Evidence tab.

Live readouts

Oxygen30%
Earth 21%
Pressure10.2 psi (0.694 atm)
Earth 14.7
Gravity0.166 g
Earth 1 g

30% O2 or more at 10.2 psi or less (challenge statement). FM2 will test 21% O2 at 14.7 psia and 34% O2 at 8.2 psia, atmospheres of interest to the Human Landing System and Lunar Rover (NASA NTRS 20240015307). Lunar gravity: no data in PSI; LUCI reported first simulated-lunar-gravity burns in 2025.

Compared with Earth

+43%

oxygen concentration

O₂ partial pressure: 3.1 psi (Earth 3.1 psi). Almost the same partial pressure, yet 43% more oxygen by concentration; the challenge notes that flammability depends strongly on oxygen concentration.

Computed from the figures above.

Gravity evidence

0.166 g: no data in PSI. LUCI reported first lunar-gravity burns in 2025 (NASA NTRS) and FM2 will test on the Moon; microgravity and 1 g results bound the answer.

What measured FLEX tests say for this atmosphere

Of the 5 closest FLEX tests, 0 ended in extinction. The animation follows that majority. It describes nearby measurements; it is not a prediction.

Hi, I'm Pax, your guide!

I can help you: