Learn from tests and failed flights

In July 2022, JAXA described LE-9 tests addressing turbopump blade vibration and collecting data about an additively manufactured injector. Those activities examined particular design questions. A test can produce useful data without qualifying every part of the complete launch vehicle.

Think of an original test plan with three questions: does a pump meet its vibration limit, does a stage start correctly, and does a payload separate? A pump test directly addresses the first question. The other two need relevant integrated-system or flight evidence. Evidence becomes more useful when its scope is stated clearly.

H3’s first test flight on March 7, 2023 failed after its second-stage engine did not ignite. JAXA’s initial report described the failure and the investigation it was starting. That report does not justify assigning the earlier first-stage LE-9 turbopump development issue as the cause of the second-stage event.

Test Flight 2 launched on February 17, 2024. JAXA’s March account reported completion of its demonstration objectives and secondary-payload separations. The photograph here shows liftoff; the later report supplies the outcome. Success after a failure matters, but it does not make every subsequent flight certain to succeed.

Flight 8 also failed, on December 22, 2025. Its initial report described abnormal behavior during the second stage’s second ignition and premature shutdown, leaving QZS-5 short of its planned orbit. An observed symptom identifies what happened at a particular point; it is not automatically the final root-cause explanation.

By May 2026, JAXA described formal review of the investigation and countermeasures, with further validation before Flight 6. Flight numbers need not follow launch order. The useful lesson is a cycle of observation, investigation, corrective work and relevant testing—not a claim that a public announcement or a single later success guarantees all future flights.