Mercury-Scout 1

Before geostationary communications satellites existed, following a spacecraft in low Earth orbit meant ringing the planet with stations beneath its ground track. The Mercury Tracking Network was exactly that: a series of ground stations and tracking ships, owned and operated by the United States, positioned under the flightpath of the Mercury spacecraft. When the spacecraft came within several hundred miles of a station, it could hold line-of-sight voice and telemetry communications over HF (shortwave), VHF or UHF, and be tracked by C-band and S-band radar. Each pass lasted only a few minutes, until the station dropped below the horizon; between stations the spacecraft was out of contact except for the occasional unreliable HF message. The stations were tied back to Mercury Control in Florida by land lines, undersea cables and, in some cases, HF radio.

The network was commissioned in 1961 with eighteen ground tracking stations and two ships, one in the Atlantic and one in the Indian Ocean, to close the gaps between ground sites. The hub was the Mercury Control Center at Cape Canaveral, Florida, with downrange antennas at Grand Bahama and Grand Turk. Eastward along the track came Coopers Island in Bermuda, the Atlantic ship, Maspalomas Station on Grand Canary Island, Kano in Nigeria, Zanzibar, the Indian Ocean ship, Muchea and the Woomera Test Range in Australia, Canton Island, Kauai in Hawaii, Point Arguello in California, Guaymas in Mexico, White Sands in New Mexico, Corpus Christi in Texas and Eglin in Florida. The line-up varied between flights: between MA-6 and MA-7, for instance, the Mid-Atlantic Ship was removed and the Indian Ocean Ship was repositioned to the Mozambique Channel.

Running the network was not the job of the organisation running Mercury. On 1 July 1961 responsibility for operating the Mercury global network was assigned to the Goddard Space Flight Center, with the proviso that during active mission periods control reverted to Space Task Group personnel. That split ownership — Goddard as standing operator, the Mercury team as wartime tenant — goes a long way towards explaining why the programme insisted on rehearsing the whole network before trusting an astronaut's life to it.

A tracking network cannot be verified on paper. Radars have to acquire a real target and hand it off to one another; telemetry receivers have to lock onto the right signal at the right instant; operators have to run acquisition, tracking and handover procedures inside the few minutes a pass allows; and the links back to Florida have to deliver the data in time for flight controllers to act on it. None of that can be exercised properly without something genuinely passing overhead.

In 1961 the programme was short of both time and opportunities. Mercury orbital flights were few and expensive, and each was devoted first of all to proving the Atlas launch vehicle and the spacecraft itself. Launching a small, cheap satellite whose only purpose was to be tracked promised many more repetitions without consuming a capsule. Hence the proposal of 5 May 1961 to use Scout rockets to launch small satellites for evaluating the worldwide Mercury Tracking Network in preparation for crewed orbital missions. The concept was approved on 24 May, and on 13 June the Space Task Group forwarded to NASA Headquarters the details of the Mercury-Scout instrumentation system, describing the mission as a check on the operational effectiveness of the Mercury global tracking network.

The effort was not confined to the flight itself. Network exercises were run in July 1961 with the MNTV-1 test vehicle, and NASA recorded what was observed at the ground stations during the exercises of 25 to 31 July 1961. The orbital vehicle would have been the natural continuation of that work: once the procedure had been rehearsed in simulation, what remained was a real target.

The rest of Mercury flew on Little Joe, Redstone and Atlas. For MS-1 NASA took a different route and used what already existed: the United States Air Force was already launching Blue Scout rockets from Cape Canaveral, so a Blue Scout II, vehicle D-8, was modified for the mission and the Air Force flew it.

The Scout was a four-stage solid-propellant launcher, with an Aerojet General Algol first stage, a Thiokol Castor second stage, an Allegany Ballistics Laboratory Antares third stage and an Altair fourth stage from the same laboratory. By the standards of the programme it was a small vehicle, about twenty-one and a half metres long against the bulk of an Atlas. That modesty was the point: a compact, solid-fuelled, inexpensive rocket able to place a light payload into low orbit was the logical carrier for a satellite whose entire job was to transmit and be tracked.

The MS-1 satellite was a small rectangular box massing 67.5 kg. Inside were two command receivers, two minitrack beacons, two telemetry transmitters and an S- and C-band beacon with their antennas, all powered by a 1500 watt-hour battery. The fourth-stage instrument package was attached as well.

The duplication of almost everything — two receivers, two beacons, two transmitters — was deliberate: the network had to practise against the same bands and the same kinds of signal a crewed Mercury spacecraft would radiate, and a single failure had to not cost the whole mission. The battery would run the electronics for about eighteen and a half hours before going flat, which would have exhausted the satellite in under a day. The operating plan stretched that endurance: after the first three orbits, roughly five hours, the equipment would be commanded off and the results analysed on the ground; then it would be powered back up for another three orbits and another five hours; and the cycle would be repeated a third time. Mercury planners reckoned that by shutting the satellite down and bringing it back, the network would gain data and experience equivalent to three Mercury orbital missions.

Mercury-Scout 1 was readied for launch on 31 October 1961 at Launch Complex 18B, Cape Canaveral. The prelaunch countdown proceeded normally, but the engine failed to ignite. Pad crews checked out and repaired the ignition circuits, and the flight was reattempted the following day.

On 1 November 1961 D-8 lifted off. Control began to fail only seconds after liftoff: the vehicle developed erratic motions that imposed high aerodynamic loads, and at T+28 seconds the first stage started to disintegrate. The Range Safety Officer issued the destruct command at 43 seconds into the flight. There was no orbit, no satellite and no tracking data.

The same day brought a change of a different order. The Space Task Group, the organisation charged with directing Project Mercury and the other crewed spaceflight programmes, was redesignated the Manned Spacecraft Center, with Robert R. Gilruth as its Director. The day the programme lost its network-calibration vehicle was also the day its organisation became a NASA field centre in its own right.

The failure lay neither in the rocket's design nor in the payload. The investigation traced it to an assembly error: a technician had accidentally transposed two wiring connectors in the guidance system. As a result, pitch errors were interpreted as yaw errors and yaw errors as pitch. Every correction the guidance commanded to straighten the vehicle pushed it off in the wrong axis, so the control loop degraded the trajectory with each iteration instead of stabilising it. That is why the motions were erratic from the first seconds, and why the airframe could not survive the loads that followed.

It is a failure mode characteristic of its era and its class: not a technological limit but a human integration error capable of destroying an entire vehicle in under a minute. For the programme it carried an uncomfortable irony, since the flight whose purpose was to verify that ground procedures worked was lost to a ground procedure carried out wrongly.

NASA cancelled the remaining Mercury-Scout missions. The decision cost less than it might appear, because the schedule had already done much of the work. By the time MS-1 flew, MA-4 had orbited: on 13 September 1961 the first uncrewed orbital Mercury-Atlas flew with only minor anomalies — a small leak in the oxygen system, loss of voice contact over Australia, and the failure of an inverter in the environmental control system — but with a highly successful outcome overall, in which the Mercury global tracking network and its telemetry performed excellently and were declared ready to support crewed orbital flight.

Confirmation followed shortly after MS-1 was lost. On 29 November 1961, twenty-eight days later, MA-5 carried the chimpanzee Enos into orbit and put the network through its paces again under real conditions; for that mission the Mercury astronauts were assigned as spacecraft communicators at several of the global network stations, rehearsing the role they would fill during crewed flights. Between MA-4 and MA-5 the network was checked out and found to meet all requirements, which made further Mercury-Scout missions unnecessary.

The calibration MS-1 was meant to supply was therefore obtained the expensive way: by using the programme's own orbital missions as tracking targets rather than a dedicated, inexpensive satellite. The price was that every network verification came bolted to a complete Mercury flight, with its Atlas and its spacecraft, and that there were no cheap repetitions for training operators. In exchange, the data came from the real object rather than a stand-in.

MS-1 appears in NASA's own census of the uncrewed developmental launches of Project Mercury, slotted between MA-4 and MA-5 and listed as a failure, with zero orbits and forty-three seconds of flight. It is the least-told mission of the programme, partly because it lasted under a minute and partly because it carried nothing recognisable: no capsule, no window, no occupant.

Its interest lies precisely in what it reveals about the programme. Mercury was never only a capsule and a rocket. It was also a planet-spanning infrastructure of stations, ships, radars, telephone lines and undersea cables that had to be built, operated and rehearsed, and whose reliability bore on the astronaut's safety as directly as any onboard system. MS-1 is the flight the programme devoted to that infrastructure, and the way it was lost — two connectors put in the wrong sockets — sums up the constraints of an age in which the big failures rarely came from physics and almost always from integration.