Little Joe 5

When NASA began planning Project Mercury's test flights it ran into an uncomfortable piece of arithmetic: an Atlas cost on the order of two and a half million dollars apiece and a Redstone about a million, figures that could not sustain the long series of qualification flights the project required. The answer was to build a booster of its own — simple, solid-propellant, and costing in the region of two hundred thousand dollars per unit. That booster was Little Joe.

The idea predated the agency itself. In January 1958 Max Faget and Paul Purser had worked out on paper how to cluster four solid-fuel Sergeant rockets, then in standard use at Wallops Island, to boost a crewed nose cone above the stratosphere. The proposal was shelved, but in August 1958 William Bland and Ronald Kolenkiewicz returned to those preliminary designs looking for a cheap cluster able to lift full-scale, full-weight capsules. Drop tests of boilerplate capsules were yielding aerodynamic data on the dynamic stability of the configuration in free fall, and the equivalent data for the powered phase was plainly missing. In October 1958 an agency team drew up new engineering layouts and estimates for the booster structure and a suitable launcher.

The name was born at that drawing board. The first cross-section drawings showed four holes, and the nickname "Little Joe" followed, taken from the craps throw of a double deuce; the four large stabilising fins protruding from the airframe helped it stick. Four smaller circles were added later for the Recruit motors, but by then the name was fixed. Little Joe was the first rocket designed purely to qualify crewed spacecraft and one of the pioneering operational vehicles built on the motor-cluster principle. Between 1959 and 1961 it flew eight times from Wallops Island, always for the same purpose: testing the escape system and heat shield of the Mercury capsule, and above all solving the problem of escaping an explosion on the pad or during ascent.

The early flights of the series — Little Joe 1, LJ-6, LJ-1A, LJ-2 and LJ-1B — had all used boilerplate articles: shape- and mass-representative structures, some carrying primates, good enough to verify the escape-system concept, the spacecraft's descent dynamics and the operation of the parachutes. What that hardware could not do was say anything about the behaviour of the real vehicle, with its production electronics, sequential system and separation mechanisms.

Little Joe 5 broke through that barrier. It was the first of the series to fly a production Mercury spacecraft manufactured by McDonnell Aircraft, capsule No. 3. That capsule already had a history inside the programme: it had been delivered to Langley on 29 July 1959 for a noise and vibration test, and more than a year later, on 27 September 1960, it was erected at the Wallops Island launch site for the LJ-5 flight. Flying production hardware changed the meaning of the test: this was no longer about whether the concept worked, but about certifying that the assembly which would carry an astronaut behaved as expected at the worst conceivable moment of ascent.

The stated purpose of the test was to check the spacecraft in an abort simulating the most severe launch conditions. In engineering terms that meant demonstrating the structural integrity of the spacecraft and escape system during an escape manoeuvre initiated at the highest dynamic pressure anticipated during an Atlas launch for orbital flight; demonstrating the performance of the escape system, the sequential system and the recovery system; determining the spacecraft's flight dynamic characteristics during the escape manoeuvre; and establishing the adequacy of recovery and prelaunch checkout procedures. That same block of objectives reappeared, almost word for word, in the plans for LJ-5A and LJ-5B — precisely because LJ-5 satisfied none of them.

The critical point deserves emphasis. An abort at maximum dynamic pressure is the limiting case for the escape system. If the tower pulls the capsule away at that instant, the structure simultaneously takes the escape motor's thrust, the deceleration of the stack and the aerodynamic loads of the airflow. It is the case that no drop test and no pad abort can reproduce.

The autumn of 1960 was a busy stretch for the programme. While capsule No. 3 was being prepared at Wallops, spacecraft No. 5 went to the Marshall Space Flight Center for booster compatibility checks and from there to Cape Canaveral for the Mercury-Redstone 2 mission; flight-type pressure suits arrived from B. F. Goodrich and went straight onto the human centrifuge; and in October the astronauts completed their third centrifuge training programme at the Aviation Medical Acceleration Laboratory, regarded as the last major centrifuge preparation before the first crewed Mercury-Redstone flight. LJ-5 was not an isolated experiment but one item on a calendar already pointing at crewed flight.

Little Joe 5 lifted off from pad LA-1 at Wallops Island on the morning of 8 November 1960. The ascent was normal for the first seconds. At 15.4 seconds after lift-off the escape rocket motor and the tower jettison motor ignited prematurely, entirely outside the planned sequence.

What followed was worse than a mistimed ignition. The capsule did not separate from the booster. Rocket, spacecraft and escape tower stayed mated through the whole ballistic trajectory and hit the water that way: the stack was destroyed on impact with the Atlantic. The flight reached an apogee of 10.1 miles (16.2 km) and a range of 13 miles (20.9 km) and lasted two minutes and twenty-two seconds from lift-off to impact. Capsule and booster debris was afterwards salvaged from the sea floor for post-flight analysis.

Not one test objective was achieved. Spacecraft No. 3, which had survived more than a year of ground testing, was lost in a little over two minutes.

The post-flight analysis attributed the failure to several possible causes without settling on one. The first pointed to failure of the limit switches on the clamp ring between spacecraft and adapter. The second, to failure of the limit switches on the escape-tower clamp ring. The third, to improper rigging of those switches in either assembly.

All three explanations share one denominator, and that is the lesson of the flight: a handful of microswitches whose only job was to report that a ring had released could, if badly rigged or badly adjusted, fire the escape system early and at the same time prevent the very separation that system was supposed to produce. This was not a conceptual failure of the escape system or of the capsule structure; it was a failure of the sensing chain and the wiring that governed the sequence. With the hardware destroyed on impact and only debris recovered from the sea floor, the investigation could not close the case on a single proven cause.

The immediate consequence was that the test had to be flown again. On 20 January 1961 spacecraft No. 14 was delivered to Wallops Island for the Little Joe 5A maximum dynamic pressure abort test, the sixth mission of the Little Joe series. It flew on 18 March 1961 with the explicit aim of satisfying the objectives LJ-5 had left unmet because the spacecraft failed to separate from the booster.

Lift-off was normal, but nineteen seconds later the escape tower fired prematurely, in a near-replica of the November event. The abort signal was given and the launch vehicle-to-adapter clamp ring released as intended, yet the spacecraft stayed on the booster because the escape motor was already expended. Separation was achieved with the retrorockets, but the command went up before the flight had reached its apex, where separation had been planned; the result was a rather violent separation. Even so, the parachutes deployed at about forty thousand feet and, once recovered, the capsule proved to have suffered only superficial structural damage — so little that the same spacecraft was reused for the Little Joe 5B flight.

The flight reached an altitude of 7.7 miles (12 km) and a range of 18 miles (29 km), lasted five minutes and twenty-five seconds and subjected the spacecraft to 8 G. Despite the recovery, the test objectives were again not considered met: the abort had not occurred under the conditions sought.

Spacecraft No. 14A — the same vehicle as before, refitted — arrived at Wallops Island on 4 April 1961 for the Little Joe 5B mission. It was launched on 28 April 1961 from pad LA-4, with the same aim as always: to test the Mercury escape system under maximum dynamic pressure conditions.

There was an anomaly from the first instant. One of the booster's rocket motors did not ignite until four seconds had elapsed, which pitched the vehicle onto a lower trajectory than planned. The side effect was that the abort manoeuvre took place at greater dynamic pressures than the flight test plan had specified. Apart from that, every other sequential system worked to plan and a normal helicopter recovery followed the landing. All test objectives were met and in fact exceeded, since the spacecraft had withstood higher dynamic pressures than required.

LJ-5B reached an apogee of 2.8 miles (4.5 km) and a range of 9 miles (14 km), lasting five minutes and twenty-five seconds with an acceleration of 10 g. The capsule flown on LJ-5A and LJ-5B is on display at the Virginia Air and Space Center in Hampton, Virginia.

Judged by its own objectives, Little Joe 5 was a complete failure: it lost the spacecraft, it did not separate the capsule, it verified none of what it set out to verify and it did not even yield a closed cause. Judged by what it gave the programme, the balance reads differently. The flight put production hardware through real abort conditions for the first time and found that the weak link was not where attention had been focused — the escape motor, the structure, the heat shield — but in the instrumentation of the clamp rings and how it was rigged. That LJ-5A repeated the failure four months later confirms how subtle the problem was and that the first correction did not cure it.

It also illustrates the logic of Mercury's entire uncrewed series. Little Joe existed precisely so that failures like this would happen over the Atlantic with an empty cabin, at a cost per flight far below that of an Atlas. A premature escape-motor ignition fifteen seconds into flight, with the capsule unable to separate, would have been a fatal accident on a crewed mission. When Alan Shepard flew the first crewed Mercury-Redstone on 5 May 1961, the escape system above him had been debugged in part by what happened to spacecraft No. 3 on 8 November 1960.