FLAMEFALLAnalytics

Interactive · estimates, never verdicts

Scenario Lab

Set an atmosphere, a material, a flow and a gravity level. The answer is the nearest real tests, an estimate where a model passed validation, and “no data” where nothing was measured.

Scenario

Every result below updates as you move a control.

21%
101.3 kPa

101.3 kPa = 14.7 psi (sea level). 70.3 kPa = 10.2 psi. 56.5 kPa = 8.2 psi.

10 cm/s

Oxygen partial pressure

21.3kPa

of oxygen for PMMA (acrylic) in this atmosphere.

Source: computed from the inputs: 21% × 101.3 kPa. Sea-level air is 21.2 kPa. Dial scale 0 to 65 kPa, the highest in the droplet tests.

14.7psi

total pressure

101.3 kPa = 1.00 atm

10cm/s

ventilation flow

forced flow past the fuel

0 g

gravity

0 g (orbit)

Nearest real tests

The eight solid-fuel test points closest to this oxygen level, pressure and flow, same material first.

Where the scenario sits among the real tests

Every solid-fuel test point with a stated outcome. The crosshair is your scenario.

  • sustained
  • quenched
  • blown off
  • no ignition
  • your scenario (not a test)

Source: NASA/TM-20210011385 (BASS-II); PSI-99; Olson et al. 2023; ICES-2024-365; ICES-2025-156; NTRS 20240007628. Solid fuels only.

Nearest solid-fuel test points
InvestigationTestMaterial (as printed)O2, %Pressure, kPaFlow, cm/sFlow directionOutcomeDistanceSame materialWhat the source saysSource
BASS-IIB132-cm, 0.4-mm thick film21.0101.3assumed10.0opposedsustained0.00yesnotes: 'Multiple velocities for steady spread'NASA/TM-20210011385, Table A.1 (PDF p. 111)
BASS-IIB152-cm, 0.1-mm PMMA film21.0101.3assumed10.0opposedsustained0.00yesnotes: 'Spread at 10 cm/s, did not blowoff'NASA/TM-20210011385, Table A.1 (PDF p. 111)
BASS-IIB82-cm, 0.2-mm PMMA film21.0101.3assumed10.0opposedsustained0.00yesnotes: 'Spread at 10 cm/s, did not blowoff'NASA/TM-20210011385, Table A.1 (PDF p. 112)
BASS-IIM32-cm-wide, 5-mm-thick, one-sided PMMA20.8101.3assumed10.0opposedsustained0.10yesSpread rate, mm/s: '0.029 and 0.024' at Velocity, cm/s: '10 and 20'NASA/TM-20210011385, Table 5.1 (PDF p. 65)
BASS-IIM142-mm-thick PMMA, one-sided, 2 cm wide21.6101.3assumed9.0opposedquenched0.31yesnotes: 'flame self-extinguished at an air reading of 9 cm/s'NASA/TM-20210011385, Table A.4 (PDF p. 116)
BASS-IIM23-mm-thick PMMA one-sided, 2 cm wide21.2101.3assumed14.0opposedquenched0.33yesnotes: 'flame got very dim, so increased flow to 14 cm/s, but it self-extinguished'NASA/TM-20210011385, Table A.4 (PDF p. 117)
BASS-IIM104-mm-thick, 1-cm-wide PMMA slab20.1101.3assumed10.0opposedsustained0.45yesSpread rate, mm/s: '0.0373 and 0.03' at Velocity, cm/s: '15 and 10'NASA/TM-20210011385, Table 5.1 (PDF p. 65)
BASS-IIM134-mm-thick PMMA, 2 cm wide20.8101.3assumed6.0opposedsustained0.46yesSpread rate, mm/s: '0.047, 0.035, and 0.033' at Velocity, cm/s: '6, 5, and 4'NASA/TM-20210011385, Table 5.1 (PDF p. 65)

243 of the 313 solid-fuel points have an assumed pressure. A pressure marked as assumed is not printed for that test. BASS-II tests outside the fabric table are taken as 1 atm (the memo states 1 atm for the fabric tests; the others ran in the same glovebox), and the Saffire VI Nomex sample is given its campaign's initial pressure.

Source: BASS-II summary memo NASA/TM-20210011385 (Tables 5.1, 7.1, A.1 to A.6); PSI Saffire tables; ICES-2024-365; NTRS 20240007628. Each row names its own source. · open source

Model estimate

A probability from a small logistic model, shown only where that model passed validation.

NO DATANo model estimate for solid materials

A logistic model was trained on the 179 verified solid-fuel test points exactly as specified, and it failed validation: on test series it had not seen, its AUC is 0.48 (95% interval 0.38 to 0.65), no better than a coin. Showing its probabilities would be inventing confidence, so none is shown.

Why it fails, and every metric, are in the Model Lab. The nearest real tests above are the answer.

What the nearest tests did: 6 sustained · 2 quenched. Their distance from your scenario ranges from 0.00 to 0.46. The larger the distance, the less comparable the test: read each row, not the tally.

Triage score

A stated formula for ordering scenarios by how much attention they deserve. It is not a risk rating.

Your scenario

triage score, from 0 to 1

Hazard = Σ weight × term. A design choice, for ordering only.

Term by term

Each term is between 0 and 1. Its contribution is its weight times its value.

  • Oxygen margin0.30 × 1.00 = 0.300

    O2 partial pressure 21.3 kPa against a minimum of 13.7 kPa (BASS-II concurrent PMMA rod, 13.6% O2 at 101 kPa (NASA/TM-20210011385 §6.1; ASGSR 2014 poster))

  • Flow margin0.20 × 1.00 = 0.200

    10 cm/s, inside the 1-30 cm/s ventilation band

  • Heat-release scale0.15 × 0.28 = 0.042

    about 50 W, the figure given for BASS-II-scale flames in ICES-2019-101 (not re-verified; the summary memo says most tests were well below 100 W)

  • Detection difficulty0.15 × 0.40 = 0.060

    no detector-miss evidence for this material; +0.1 for a free volume above 15 m3 (far-field alarm took 981 s in Saffire VI)

  • Gravity0.10 × 0.50 = 0.050

    0 g (microgravity).

  • Invisible burning0.10 × 0.00 = 0.000

    solid fuel, or a droplet outside the cool-flame regime

Adjust the weights

They are a design choice. Whatever you type is renormalised to sum to 1.

Defaults: Oxygen margin 0.30, Flow margin 0.20, Heat-release scale 0.15, Detection difficulty 0.15, Gravity 0.10, Invisible burning 0.10.

Order of scenarios under these weights

  1. 1Overheated wire insulation, station cabin0.742
  2. 2Pressurised rover, acrylic panel, cabin fan on0.682
  3. 3Station cabin baseline, acrylic rod0.652
  4. 4Your scenario0.652
  5. 5Lunar-base cabin, thin fabric, low ventilation0.556
  6. 6Large fabric fire in a cargo-vehicle volume0.460

Ranking is weight-sensitive

Moving a single weight by 0.10 changes the top three: Detection difficulty weight -0.10 → Pressurised rover, acrylic panel, cabin fan on, Overheated wire insulation, station cabin, Station cabin baseline, acrylic rod; Gravity weight -0.10 → Overheated wire insulation, station cabin, Station cabin baseline, acrylic rod, Your scenario.

Do not read the order as settled. It depends on a choice of weights that no experiment fixes.

Source: Formula, lookup values and their sources: flamefall/ranking.py, exported to the app unchanged. Weights are a design choice, not a measurement.

What the data says, and what is unknown

The two lists a reader should leave with.

What the data says

  • 6 of the 8 nearest test points: sustained; 2 of the 8 nearest test points: quenched.
  • Oxygen partial pressure here is 21.3 kPa. Saffire VI started at 54.1 kPa and 31.0% O2, an oxygen partial pressure of about 16.8 kPa. That is lower than sea-level air (21.2 kPa) even though the oxygen fraction is higher. Source: Niehaus et al., ICES-2025-156, p. 3 (NTRS 20250002675); partial pressure computed here.
  • Smoke from overheated wire insulation can be too fine for a light-scattering detector: see What We Don't Know.

What is unknown

  • Nothing flagged for these inputs: the scenario sits inside the tested ranges.
  • How any of this changes in partial gravity: no repository dataset exists yet.