Method
How JaPax turns NASA experiments into an answer
Every number on this site is one of four kinds. This page says which, where each one comes from, and what the model does and does not know.
Four kinds of number
- Measured
- A value recorded in a PSI dataset, cited with its PSI ID, DOI and sample size.
- Reported
- A value stated in a peer-reviewed paper or NTRS report that we did not re-derive from raw data.
- Estimated
- An engineering assumption or hypothesis with no experimental source. It is always shown as such.
- Simulated
- The output of the JaPax component-level spread model (a graph, level A).
From the archive to the page
- NASA's Physical Sciences Informatics (PSI) repository is queried through its public API. Investigation metadata and file inventories come from there.
- Experiment tables (CSV) and facts we curated by hand, each with its source, are merged into one dataset with a build script.
- Every fact in the spacesuit, ISS and environment component catalogues carries a tag and a source, and automated tests (npm run test:lab) fail if one is missing.
- The totals on the home page (19 investigations, about 12.7 TB, 12,146 files) are summed from the sizes PSI declares (decimal terabytes, 1 TB = 10¹² bytes), queried on 15 Nov 2026; the per-investigation sizes are saved in data/psi-size-total.json.
What the simulator computes
- Spread speed = anchor × oxygen × airflow × gravity × material.
- The anchor, 2.0 mm/s, is a representative value inside the concurrent-flow spread rates NASA measured on thin fabric in microgravity (1.8-2.6 mm/s, SAFFIRE I-III, NTRS 20180005168 Table 1; PSI-98, PSI-99, PSI-100).
- The four factors that scale it to other atmospheres are modelling choices, not fitted to data, and each one shows its source in the simulator.
- The component layout and the rupture window of the oxygen tank are estimates. The simulator labels them as such and offers a view with only measured values.
- The result is a model, not a test. It ranks what to protect first; it does not certify a design.
What the data cannot say yet
- No dataset in PSI was measured at lunar gravity (0.166 g). LUCI reported the first extended-duration burns in simulated lunar gravity in 2025 (NTRS 20250010653), and FM2 is a planned lander mission. Everything lunar here is an extrapolation.
- FLEX burned fuel droplets, not fabric, plastic or cables. BASS-II reached only 22.2% oxygen, below the 26.5% of a lunar habitat.
- The BASS-II table has no pressure column; 1 atm is an assumption and is labelled so.
- 100% oxygen at 4.3 psi, the spacesuit, is outside every tested atmosphere.
- The ranking weights (35, 25, 20, 10, 10) are a prioritisation aid built by JaPax from cited facts, not an official NASA metric.
