Atlas Agena

The decommissioned Atlas-D missiles were fitted with Agena upper stages for use as space launchers. As the Agena was stretched to carry more propellant, and its engine was made both more powerful and capable of being restarted in space, it became available in A, B and D models (the planned C model was cancelled). The Atlas-Agena-A employed the MA-2 power plant.14 Two of four launches failed. On the first launch, on 26 February 1960, the Agena with the first satellite for the Missile Defense Alarm System (MIDAS) failed to separate from its booster, but the second was inserted into orbit on 24 May. On 11 October the Agena with the first of the Satellite and Missile Observation System (SAMOS) satellites failed, but the second was launched on 31 January 1961. The Atlas-Agena-B had the MA-3 power plant.15 Of 28 launches between 12 July 1961 and 21 March 1965, eight suffered problems. On 23 August 1961, on the second mission, the Agena attained low 'parking orbit' and then failed to restart to make the burn that was to insert the Ranger 1 spacecraft into a highly elliptical orbit. The next launch on 9 September with SAMOS 3 was a total loss when the vehicle exploded on the pad. MIDAS 4 was successfully dispatched on 21 October, but on 18 November Ranger 2 was stranded in parking orbit when an inoperative roll gyroscope prevented the Agena restarting. After SAMOS 4 was lost on 22 November, SAMOS 5 was successfully placed into orbit on 22 December but failed to return its film capsule. On 26 January 1962 the Agena

with Ranger 3 was able to restart to leave parking orbit, but a guidance fault made the probe miss the Moon by some by 37,000 kilometres. After a string of successes, Mariner 1 was lost on 22 July when the Atlas flew off course and at T + 293 seconds was destroyed by the range safety officer. Although this is often cited as the exemplar of the loss of a mission to a software error, in fact, as is often the case, two independent faults had interacted fatally.16 The guidance antenna on the Atlas performed below specifications. When the received signal became weak and noisy, the vehicle lost its lock on the ground reference that supplied steering commands. In the event that radio guidance was lost, the autopilot was supposed to reject the spurious signals from the antenna and proceed on its stored program. However, at this point a second fault took effect. Due to an oversight, a hyphen had been omitted from the program, and this had the effect of allowing the flawed signals to command the vehicle to veer left and drop its nose. It was found that the hyphen had always been missing, but had been benign since there had been no radio guidance failure on the previous flights. Mariner 2 was successfully dispatched on 27 August and became the first spacecraft to return data about another planet when it flew past Venus on 14 December 1962 at a range of 35,000 kilometres. There were another two successes, then MIDAS 6 was lost on 17 December 1962 and MIDAS 8 on 12 June 1963. The Atlas-Agena-D had the MA-5 power plant.17*18 Of 27 launches between 12 July 1963 and 7 April 1978 there were only two failures. On 5 November 1964 Mariner 3 was trapped when the shroud failed to separate. An investigation found that the new lightweight shroud had bonded to the probe. However, it was able to be revised in time to dispatch Mariner 4 on 28 November, a few days before the 'window' for a mission to Mars closed, and the pictures that it snapped during a 10,000-kilometre flyby on 15 July 1965 revealed the surface of the planet to be scarred by large impact craters. The other incident was on 4 December 1971, when an Atlas was lost with a Canyon communications intelligence-gathering payload.

In addition, in June 1962 NASA ordered a variant of the Agena on which the

Atlas models (left to right): Atlas-D, 20 February 1962 with John Glenn in a Mercury capsule; Atlas-Able, 26 November 1959 with Pioneer P3 (lost when the shroud collapsed after 45 seconds); Atlas-Agena-A, 26 February 1960 with MIDAS 1; Atlas-Agena-B, 9 September 1961 with SAMOS 3; Atlas-Agena-D, 30 September 1965 with the 22nd KH-7 satellite.

Atlas models (left to right): Atlas-D, 20 February 1962 with John Glenn in a Mercury capsule; Atlas-Able, 26 November 1959 with Pioneer P3 (lost when the shroud collapsed after 45 seconds); Atlas-Agena-A, 26 February 1960 with MIDAS 1; Atlas-Agena-B, 9 September 1961 with SAMOS 3; Atlas-Agena-D, 30 September 1965 with the 22nd KH-7 satellite.

engine could be fired as many as six times to enable it to undertake orbital manoeuvres while docked to a manned Gemini spacecraft. This Gemini-Agena Target Vehicle also had a pair of thrusters mounted in 'saddle bag' fashion on each side of the main engine. The first of these vehicles lifted off from Pad 14 at Canaveral on 25 October 1965. After the Agena was released, it coasted for 50 seconds and then fired its thrusters for ullage to settle the propellants in their tanks prior to firing the main engine for orbital insertion. The telemetry data ceased a fraction of a second after the engine ignited.19 Discovering what had gone wrong from the momentary telemetry was a daunting task. The multiple restart capability had been provided by replacing the solid-charge starter that generated the gas to drive the turbine with one that created the gas by vaporising liquid drawn from a tank. This was not a factor. However, in the Agena-D, which could be restarted only once, the flow of oxidiser was initiated prior to feeding fuel since trials had shown that this improved the engine's performance. But because NASA intended to use the vehicle for extensive orbital manoeuvring, it had rejected this 'waste' of oxidiser, and demanded that the fuel and oxidiser be injected simultaneously. Despite successful ground tests, the fuel in this case had been injected into the chamber ahead of the oxidiser, creating a combustion instability in which the fuel-rich mix had burned explosively, and this 'hard start' had created a pressure transient that blew the engine apart.20 This therefore serves as an excellent example of how a seemingly straightforward modification can give rise to complications. The next flight on 16 March 1966 was successful, and when Gemini 8 rendezvoused with the vehicle a few hours later Neil Armstrong and Dave Scott became the first people to see another spacecraft in orbit.21

The target vehicle for the next mission, Gemini 9, was lost on 17 May 1966 when, 10 seconds before the Atlas was to shed its side-mounted auxiliaries, one of the engines gimballed hard over and forced it to nose-dive into the Atlantic. As a contingency, it was decided to launch an ad hoc target comprising the docking system stabilised by a bolted-on attitude control system of the type used by the Gemini spacecraft. This Augmented Target Docking Adapter was placed directly into the rendezvous orbit on 1 June, but the telemetry during the ascent was ambiguous in that although the command to jettison the aerodynamic shroud had been issued, the confirmation that this had occurred was not received. When Tom Stafford and Gene Cernan drew along-

A GATV as seen from Gemini 8.
The ATDA with its fouled shroud.

side the vehicle, they found that the shroud had partially deployed - the pyrotechnics had separated its base from the vehicle and split it lengthwise, but a thin metal strap was preventing it from slipping off the aperture of the conical docking collar. Stafford likened the fouled shroud to the jaws of ''an angry alligator''. When the pictures taken by the astronauts were examined by the engineers, the fault was identified. The Agena shrouds were made by Douglas Aircraft, and the Gemini-Agena Target Vehicles were built by Lockheed, whose pad technicians were trained to install the shrouds. However, the improvised Augmented Target Docking Adapter had been made by McDonnell (which had built the Gemini spacecraft and the docking system) and in this case the shroud had been fitted by McDonnell's technicians, the Lockheed supervisor had been called away, and his verbal instructions for finishing the job had been misunderstood.

0 0

Post a comment