Explosion of a LH2 tank in space

The event terminated an unmanned flight. One of the goals of the mission was to gain experience from re-firing engines in space. It was important to understand better the behaviour of the liquid hydrogen in the fuel tank in space, under reduced gravity conditions. The engineers were intentionally venting the two fuel tanks (hydrogen and oxygen) to give momentum to the spaceship and were following the sloshing of the liquid hydrogen in the tank by means of a camera.
By the fourth orbit, experiments were performed to obtain data applicable to an alternate venting system and to the design of future cryogenic stages. The tests were to observe: (i) the behaviour of liquid hydrogen when vehicle longitudinal thrust is terminated, (ii) its behaviour under rapid de-pressurisation of the hydrogen tank in a low gravity environment, and (iii) the rate of pressure rise in a closed cryogenic container in space.
When the fuel tank vents were closed for this last test, approximately 16,000 Ib (7.3 metric tons) of LH2 remained in the tank and the tank pressure was 12.4 psia (90.9 bar). The pressure raised to 37.7 psia (2.6 bar) 1.5 hours later, which corresponds approximately to one orbit.
The pressure was to be maintained below the 42 psia (2.9 bar), the activation value of the fuel tank pressure relief system. By the start of the fifth orbit, just before the expected loss of ground signal, the last reading of the pressure was 39.9 psia (2.8 bar), and all data showed that the vehicle was intact and functioning normally. The loss of signal was a normal below the horizon signal loss. After recapturing the signal by the following radar, approximately two minutes later, the radar signal was showing several pieces and indicated that the vehicle had exploded during the two-minute in the loss of signal. The most probable explanation of this occurrence is that the common bulkhead ruptured, and a spark or impact detonated the propellants.

The pressure raise rate in the liquid hydrogen tank was almost 6 time higher than the calculated value based on a model assuming homogeneous distribution of the heat input within the liquid bulk. The significantly higher measured value was attributed primarily to heating of the ullage gases and to inhomogeneous heating of the liquid.
Ehen the pressure in the thank reached the value of the pressure relief system activation, a release was probably initiated, but it is plausible to assume that the pressure raise was too high for the system to be able to release flow to mitigate it.
Although the rupture of the tank was an unwanted event classified as incident, it was part of an experimental and pioneering programme aiming at testing components under extreme conditions. The only ROOT CAUSE possible is a shortcoming of the testing sequence an of the design calculation.

stage, rocket