Little Joe 6

When the Space Task Group began designing the Mercury capsule it ran into a problem no ground rig could settle: it had to be shown that an astronaut could survive a launch-vehicle explosion at the worst possible moment, that of maximum aerodynamic load. Reproducing that condition meant flying for real, and flying often. Spending a Redstone or an Atlas on every test was slow and ruinously expensive, so Langley engineers proposed a deliberately crude vehicle: solid propulsion, proven hardware wherever possible, no electronic guidance and control system at all, and performance close to what a Redstone would give with the capsule on top. It came in at roughly a fifth of the Redstone's basic cost, was far cheaper to operate and, above all, could be fired from the facilities that already existed at Wallops Island, without competing for pads at Cape Canaveral.

The name came off the drawings. The first engineering sketches showed four holes for the main motors, and someone read in them the double deuce of the dice: "Little Joe", the craps throw. Four smaller circles were added later for the Recruit motors, along with the four large stabilising fins protruding from the airframe, but the nickname had stuck.

Little Joe was one of the pioneering operational launch vehicles built on the rocket-cluster principle. It carried four modified Sergeants — called Castor or Pollux depending on the modification — plus four supplemental Recruit motors, arranged to fire in different sequences according to what each mission demanded, so that take-off thrust varied a great deal from flight to flight. Maximum design thrust was almost 230 thousand pounds. That was enough, in theory, to loft a spacecraft of some two tons on a ballistic path over a hundred miles high and to reproduce the take-off environment a crewed Atlas would impose on the capsule.

With no guidance, any thrust asymmetry became a stability problem, and the design was amended early on: the nozzles of the forward-thrusting motors went from straight to canted, to minimise upset from unsymmetrical thrust. The second substantive change was a booster destruct system. Range safety demanded some way of terminating the thrust of the main motor units on command, and Charles H. McFall and Samuel Sokol of Langley devised a blowout system that North American and Thiokol Chemical Corporation, the makers of the motor components, fitted to the forward end of each combustion chamber. That system, added almost as a piece of paperwork, would end up as one of LJ-6's declared objectives.

On 21 August 1959, at Wallops, a Little Joe stood ready for the first attempt at a maximum-dynamic-pressure escape test. Thirty-five minutes before launch, with the area being evacuated and the batteries for the programmer and the destruct system on charge, a flash changed everything: when the smoke cleared, only the capsule-and-tower combination had been launched, on a trajectory resembling an off-the-pad abort. The booster and the adapter clamp ring were still intact on the launcher. That flight went into the books as Little Joe 1 and as a failure, leaving the programme with an unflown booster and the uncomfortable feeling of not knowing whether the launch vehicle worked at all.

The answer was to lower the ambition. On 4 October 1959 the very booster that had been left standing in August was finally fired, now carrying what the programme's official history calls a double dummy: an uninstrumented boilerplate capsule fitted with an inert escape rocket system. The purpose, far more modest than August's, was to prove the reliability of the whole booster propulsion cluster. All four Pollux motors and the four smaller Recruits were set to fire in sequence.

That is the trait separating LJ-6 from almost all its stablemates. The other Little Joe flights existed to test the escape tower and the capsule under extreme conditions; LJ-6 existed to test the vehicle that would carry out those tests. It was, in programme jargon, a qualification flight for the test rig rather than for the article under test. Hence a deliberately mute payload: no instrumentation, no live escape motor, no capsule separation and no recovery planned.

The project's official chronology lists precisely what was sought: to check the integrity of the launch vehicle airframe and motor system; to check the operation of the launcher; to check the validity of the calculated wind corrections; to obtain performance and drag data; and to check the operation of the destruct system. None of those five items has anything to do with astronaut survival, and all of them are preconditions for the flights that did test it to mean anything.

The wind-correction objective deserves a note. An unguided rocket fired from a steerable launcher can only compensate for winds aloft by pointing differently before ignition: computing that correction was itself a technique that had to be validated in flight. And the destruct objective meant, by definition, destroying the vehicle: LJ-6 was planned from the outset not to come back.

The stack stood about 55 feet tall and weighed twenty tons at lift-off. It climbed to a peak altitude barely short of forty miles according to the programme's official history, and of 37.12 statute miles according to the project chronology, with a range of 79.4 statute miles. Maximum speed was 3075 miles per hour and maximum dynamic pressure 3400 pounds per square foot; peak acceleration reached 5.9 g. Declared payload was 1134 kilograms.

About two and a half minutes into the flight, and in any case well after apogee, the destruct packages carried on board were successfully initiated and the vehicle broke up as intended. Debris fell more than seventy miles from Wallops Island; the project chronology gives a total range of 79.4 statute miles. Nothing was recovered, and nothing was meant to be. The whole flight lasted five minutes and ten seconds.

There was a malfunction, and it was recorded: the two Pollux motors due to light in the cluster's second stage did so before the exact planned instant. The effect on the trajectory was slight. The effect on the structure was greater than planned, because the vehicle had to withstand a harsher load regime than the designed test imposed, and since the test was precisely a structural one, the anomaly ended up working in the programme's favour: more was proved than had been set out to prove. That is why LJ-6 appears both as a successful launch in the chronology and as a partial success in NASA's tally of uncrewed missions. Both are true depending on what is measured: the vehicle delivered, the sequence did not.

With the booster judged sound, NASA's supervisory team — John C. Palmer from Wallops; Roland English, James Mayo, Clifford Nelson and Charles McFall from Langley; and William M. Bland and Robert O. Piland from the Space Task Group — set about preparing the serious attempt: checking the correct operation of the abort escape system at maximum loading conditions. The region called max q, the portion of the flight path where relative speed between vehicle and atmosphere produces the greatest air resistance, was the critical point Little Joe existed to reproduce, and both Little Joe and the Mercury-Atlas were expected to experience comparable dynamic pressures.

That attempt came a few weeks later with Little Joe 1A and, after it, Little Joe 2, Little Joe 1B, Little Joe 5, Little Joe 5A and Little Joe 5B, by then with McDonnell production capsules instead of boilerplates. In all, Little Joe flew eight times between 1959 and 1961 from Wallops Island, with six successes and two failures. LJ-6 was the flight that turned that cluster of motors into a tool anyone could rely on.

The number misleads. LJ-6 was not the sixth flight of the series but the first the Little Joe booster actually made, and the designation was applied afterwards: the programme's official history says in so many words that this test "later became known as Little Joe 6". Series numbering tracks vehicle and test-configuration assignments, not the chronological order of the flights, a pattern repeated across the family, with designations such as 1A, 1B, 5A and 5B interleaved in the calendar. When reading any table of the series it pays to sort by date before sorting by number.