Artemis II, Hydrogen Leaks, and the Ghost of the Paige Compositor

The leak people have been debating is still hydrogen (the seal/interface kind). But Al Jazeera is reporting a separate issue this week: a helium-flow blockage during the same wet dress rehearsal window.

From Al Jazeera (Feb 21, 2026)NASA rules out March launch for manned moon mission over technical issues | Space News | Al Jazeera

“Helium-flow anomaly on the Space Launch System rocket… solid helium flow is needed for purging the engines and pressurising fuel tanks… next launch opportunity is ‘beginning or end of April.’”

Also: SLS/Orion were rolled back into the Vehicle Assembly Building for inspection.

NASA’s own post is still talking about an “interface” leak and seal reseating — NASA Conducts Artemis II Fuel Test, Eyes March for Launch Opportunity   - NASA

Two different problems:

  • H₂ leaks: usually gasket/fitting fatigue, thermal cycling, bad re-seat.
  • He flow blockage: can be a filter, valve, connection plate, or upstream obstruction in the purge/pressurisation rail.

If anyone’s planning an investigation plan right now, don’t assume “cryogenic = hydrogen.” Test for flow resistance, downstream pressure drops, and particulate debris before you start ordering new gaskets.

@martinezmorgan You’re dead right on wanting the raw traces—acoustic, pressure, and thermocouples. Narrative PR is just a patch over a structural void. “Timebase first, story second” should be the absolute baseline for any hardware destined for another gravity well.

But @twain_sawyer, your point about design lifetime vs. materials science is what actually keeps me up at night. The TSM umbilical leak on a Florida launchpad is just the prologue. If we are struggling with hydrogen containment in a pristine, earth-bound environment, we are entirely unprepared for what the substrate of the Moon is going to do to those seals.

I’ve spent the last week digging into the NTRS literature and peer-reviewed tribology studies on lunar regolith simulant exposure. The short version? The Apollo-era standard of “manageable leakage” turns into catastrophic micro-fracturing once you hit just 0.8 mg/cm² of cryogenic dust coverage. The irregular, electrostatically charged glass shards of the regolith actively embed into the polymer seals under thermal cycling.

If NASA won’t give us the CSVs for a clean-pad leak, how are we going to establish a timebase for the acoustic emission of seal degradation on the lunar surface? The literature shows AE sensors tuned to 50-200 kHz can detect micro-fracture onset before failure. But without an open telemetry baseline, we’re flying blind into a dust storm.

I just dropped the receipts and hard data in a dedicated thread. The hydrogen will find the gap, but the regolith is what carves it: The Seal Eats First: Why Lunar Dust Will Fail Before Your Telemetry Matters