LH2 release and explosion from a failed weld
The event occurred during a pressure testing of a LH2 pump discharge system. A weld joint failed at an internally tied bellows. The weld failed at 1100 psi (ca. 76 bar) while the test pressure planned was 2000 psi (ca. 138 bar). The rupture of the line caused the release of large amounts of liquid hydrogen. The explosion was heard at 10 000 ft. (ca. 3 km) distance, but there was little evidence of scorching. The piping (part of a triple gimbal) had been hydrostatically tested to 2400 psia (ca. 165 bar) and with helium gas to 2000 psia several days before the failure. The explosion has been evaluated equivalent to 5 lb TNT (ca. 2. 3 kg), based on damage to equipment and buildings. [Ordin, NASA (1974)]

The INITATING CAUSE was the weld failure at cryogenic temperature. The most likely ROOT CAUSES were considered (i) the improper design of the gimbal and (ii) the material selection . The design shortcoming resulted in serious weaknesses. The material used (17-4PH) undergoes ductile to brittle transition near -100° F. The steel was also notch sensitive and could not withstand the non-axial loading at cryogenic temperatures. The ignition source was considered to be due to the energy from the tearing and moving of metal.